Proceedings of the

COMMITTEES

The International Conference Committee

Rodney Allam, UK

Alexey M. Arkharov, Russia

Stan Augustynowicz, USA

John G. Baust, Pres. Comm. C1, IIR, USA

John Campbell, USA

Walter F. Castle, UK

Didier Coulomb, Dir. IIR, France

Vaclav Chrz, Pres. Com. A2, IIR, Czech Republic

Ralf Herzog, Pres. Comm. A1, IIR, Germany

Boris A. Ivanov, Russia

Milos Kadlec, Czech Republic

Zdenek Kaiser, Czech Republic

Mohammad Kalbassi, UK

Philippe Lebrun, Head Section A, IIR, France

Omar Maiwand, UK

Locksley Mc Gann, Canada

Hans Quack, Germany

Stanislav Šafrata, Czech Republic

Ralph Scurlock (chairman), UK

Gabriele Spoerl, Germany

The Local Organizing Committee

Vaclav Chrz, (chairman)
Zdenek Kaiser
Radoslav Kundera
Martin Lánský (Org. Com. secretary)
Zdenek Machala
Pavel Měřička
Vera Musilová
Pavel Urban
Josef Ota
Jiri Pařízek

Stanislav Šafrata (vice-chairman)
Pavel Schustr (vice-chairman)
Martin Vinš

The Editorial Committee

Vaclav Chrz

Petr Duda

Romana Kočová

Milan Kouba
Tomas Králík

Pavel Měřička
Věra Musilová
Pavel Urban
Josef Ota
Stanislav Šafrata
Pavel Schustr

Logistics

ICARIS Ltd., Conference Management

Ladislav Červinka & Dalibor Červinka, Directors

Romana Kočová, Project manager

LIST OF EXHIBITORS

Registered till April 5, 2008

ACD CRYO AG

Gutenbergstrasse 1, CH-4142 Muenchenstein
Switzerland

Phone: +41 61 413 0230, Fax: +41 61 413 0233

E-mail: patrick.ravinel@acdcryo.com

www.cryoind.com

AIR LIQUIDE S.A.

Division Matériel Cryogénique

8 avenue Gutenberg, Parc Gustave-Eiffel,
F-77607 Bussy Saint-Georges, Marne la Vallée

France

Phone: +33 1 6476 1537, Fax: +33 1 6476 1699

E-mail: bjoern.sindermann@airliquide.com

www.dmc.airliquide.com

AUSTRIAN AEROSPACE GmbH.

8 Stachengasse 16, A-1120 Vienna,

Austria

Phone: +43 1 80199 3070, Fax: +43 1 80199 3060

E-mail: johannes.stipsitz@space.at

www.space.at

CHART FEROX, a.s.

Ústecká 30, 405 30 Děčín 5

Czech Republic

Phone: +420 412 507 343, Fax: +420 412 510 200

E-mail: sales@chart-ind.com

www.chart-ferox.com

CRYOSTAR SAS

2 rue de l´Industrie – ZI – BP 48, 68220 Hésingue

France

Phone: +33 389 70 27 27, Fax: +33 389 70 27 77

E-mail: info@cryostar.com

www.cryostar.com

HEROSE GmbH., Armaturen und Metalle

Elly-Heuss-Knapp-Str. 12, D-23843 Bad Oldesloe

Germany

Phone: +49 4531 5090, Fax: +49 4531 509 120

E-mail: info@herose.de

www.herose.de

LINDE AG, Linde Enginnering Division, Schalchen Plant

Carl-von-Linde-Strasse 15, 83342 Tacherting

Germany

Phone: +49 89 7445 6291 Fax: +49 89 7445 6291

E-mail: Gerald.Hecht@Linde-LE.com

www.linde-engineering.com

NEXANS Deutschland Industries GmbH. and Co.KG

Kabelkamp 20, 30179 Hannover

Germany

Phone: +49 (0) 511 676 3250, Fax: +49 (0) 511 676 2134

E-mail: Klaus.Schippl@nexans.com

www.nexans.de

VRV Group

Via Burago 24, 20060 Ornago (MI)

Italy

Phone: +39 039 6025 1 Fax: +39 039 6025 499

E-mail: cryo@vrv.it

www.vrv-group.com

CR08-63

The IIR, Yesterday, Today and Tomorrow

Coulomb D.

International Institute of Refrigeration (IIR),
177 boulevard Malesherbes 75017 Paris, France

Abstract

The second half of 19th century showed both a sharp increase in the demand for cold storage, refrigerated transport, needs of various factories and in the development of successful refrigerating machines. 1908 was the year of the creation of the International Institute of Refrigeration. The IIR had then to adapt to new challenges such as the protection of the environment, to new uses of refrigeration and scientific progress.

Introduction

The IIR is celebrating its hundredth anniversary. The first part of the text describes the creation the IIR, with the history of artificial cold, the 1908 event and the first years of this new international body. The second part explains what the IIR is today, the changes that have appeared regarding the challenges, the organization and the actions.

I – The Creation of the IIR

a) Introduction

At 3 p.m. on October 5, 1908, 3000 specialists in the field of artificial cold gathered in the Grand Amphitheatre of the Sorbonne in Paris. They had come to attend the formal opening of the First International Congress of Refrigeration. The Congress, which lasted a week and attracted a total of 5000 participants from 40 countries, was a resounding success. Representatives from the worlds of science, commerce, industry and government exchanged views on low temperatures, refrigeration technology, food, applications of artificial cold in trade and industry, and legislative issues. The Congress culminated in the founding of the International Association of Refrigeration in January 1909. It became an international organization, the International Institute of Refrigeration, in 1920.

While it was electrical engineering that had taken the world by storm in the last quarter of the 19th century, the baton passed to the cold industry in the early years of the new century. Buyers included breweries and ice-cream factories, cold storage and refrigerated transport companies, hospitals (for the conservation of dead bodies), dairy, chocolate, rubber and perfume factories, dyeworks and factories producing liquid carbonic acid, ammonia or air. Those involved in building mineshafts and subway tunnels soon saw the potential of artificial cold in their line of business: refrigerant pipes could be used to create a wall of frozen ground, after which it became far easier to dig out the space inside. In short, artificial cold was a growing market.

b) Artificial cold

The mid-19th century witnessed a sharp increase in the demand for natural ice during the summer months in Europe, North America (where ice was soon being used by the middle classes) and the colonies. The demand was perhaps greatest among breweries producing lager, which ferments at 5-8°C, unlike the 20-30°C of many English beers. The advent of railways and steamships boosted trade in natural ice from Scandinavia and Canada, but suppliers could not keep pace with the growing demand. Furthermore, rising concern about the sawing of blocks of ice from polluted rivers and lakes gave extra impetus to the development of machines that could manufacture clean artificial ice. The producers of natural ice lowered their prices in a fruitless effort to reverse the tide.

The first machine to produce a continuous output of ice was invented by the French businessman Ferdinand Carré. His idea was to release ammonia from a water solution by heating it, to condense the vapour under pressure until it was liquefied, and then to allow this liquid to evaporate and expand in a sealed space. This would extract heat from an adjoining space with water, which would immediately freeze. The vapour would be absorbed by the “aqua ammonia”, after which the cycle would be repeated. A prototype was placed in a brewery in Marseille in 1859. Carré’s ice machine became rather famous when it was displayed at the Paris World Exposition of 1867. He was already doing a brisk trade before then: the Confederates had bought several machines from him during the American Civil War (1861-1865). After some adjustments made by Mignon and Rouart in Paris, the vapour absorption device was one of the best-selling refrigerators in the years 1870-1885, especially in France. After that it was superseded by the vapour-compression refrigerator, which is based on a far simpler construction.

This system, which is still applied in household refrigerators, artificial ice rinks and industrial plants today, was invented by the French engineer Charles Tellier, earning him the title “le père du froid”. It uses a closed cycle. A compressor is used to compress methyl ether (which was later replaced by methyl chloride, sulphur dioxide, carbonic acid gas, and above all ammonia); a water-cooled condenser turns this into liquid, which evaporates in the space to be refrigerated (in a system of pipes – the main difference with regard to Carré’s system) and thus extracts heat from it. Tellier built his first refrigerator in Paris in 1863. Four years later he installed an improved version, using methyl chloride as the coolant, in an ice factory in Marseille, France.

Commercially speaking, the most successful machines were compression refrigerators using ammonia, launched in 1875 after theoretical studies carried out by the scientifically trained Carl von Linde. The Gesellschaft für Linde’s Eismachinen A.G., in Wiesbaden, supplied its first machine to a brewery in Munich and was soon the market leader. By 1890 the German company had sold about a thousand machines, and around the turn of the century the Wiesbaden factory was sending off one or two of its refrigerators every day[1].

A major innovation made possible by the new refrigerators was the export of frozen meat from Australia, New Zealand and South America to Europe. Cooling the meat with ice proved not to be an option; steamships were still slow in the 1870s, and clippers also took over 100 days to cross the ocean. The problem had to be solved with machines. In 1876, Tellier built a compression refrigerator on board the French ship Le Frigorifique. This steam-powered three-master sailed from Marseille to Buenos Aires with a cargo of frozen meat, to return to Le Havre a year later. Though not a commercial success, the voyage had demonstrated that shipping frozen meat across the oceans was technically feasible.

Bulk transportation imposed more stringent demands, and the problem with Tellier’s machine was that if built on a larger scale, it sometimes broke down. Besides this, the toxicity of the coolants and the risk of explosion deterred ship owners from taking the plunge. It was another type of refrigerator that made them change their minds: the air expansion machine patented by the Scottish butchers Bell and Coleman in 1877. This cooled the produce by the rapid expansion of compressed air, and in spite of poor efficiency – large steam engines were needed to compress the necessary quantities of air – and problems with frozen water vapour, the sailing vessel Strathleven transported 34 tons of frozen meat from Australia to England safely in 1879 using one of these machines. Things moved very fast after this. In 1907, Argentina exported 425 000 tons of frozen meat to England alone.

Low-temperature science, too, progressed in leaps and bounds. The last quarter of the 19th century witnessed the liquefaction of each of the “permanent gases” in turn. In 1877, the Frenchman Louis Cailletet and his Swiss colleague Raoul Pictet liquefied air. In 1883, the polish team Zymunt von Wroblewski and Karol Olszewski went a step further, by inducing the blue liquid of oxygen to boil gently. James Dewar, working in the Royal Institution, London, became the first to produce liquid hydrogen, in 1898, after which Heike Kamerlingh Onnes won the race for liquid helium in Leiden on 10 July, 1908[2].

c) The Paris congress

In this atmosphere of up-and-coming artificial cold, of new, hitherto unsuspected applications, of changing economies in countries such as Argentina, of a proliferation of technical problems crying out for solutions, of a scientific quest for absolute zero, the idea of holding a major international Congress of refrigeration, in Paris, emerged. A Congress of this kind had been held in Vienna in 1873, to coincide with World Exposition there, but it had focused on brewers and their need for natural ice. At the beginning of the 20th century, artificial cold produced by refrigerators revolutionized agriculture all over the world and offered an immense economic potential concerning national and international food trade by using large-scale refrigerators.

The engineer J. de Loverdo was the prime mover of the Paris congress, and in May 1907 a circular was distributed calling for participants. The initiative soon attracted a wide-ranging and distinguished band of supporters including l’Institut de France, the French Parliament, the Collège de France, l’Académie de Médecine, major transport companies. No-one interested in cold could afford to miss the Premier Congrès International des Industries Frigorifiques, which was finally held at the Sorbonne University on October 5-10, 1908 under the more appealing and inclusive name of Premier Congrès International du Froid, or First International Congress of Refrigeration[3]. The broad aim was to exchange ideas and discoveries in the field of cold technology.

To keep the Congress manageable, it was divided into six sections: low temperatures, refrigeration installations, applications of cold to foodstuffs, applications in other industries, applications in trade and transport, and a final section that would examine the relevant legislation. The name of the Congress made it clear that it was not to be a one-off initiative. Ideas for an international institute for cold and science, or for training courses in refrigeration technology, to be founded in Paris, soon proved overambitious. Instead, the preparatory committee offered to set up an International Association of Refrigeration. Its remit would be to perform research on scientific, technological and industrial applications, to set up a library covering all aspects of the field, to publish articles and inform its members, to provide courses, set up excursions, and organize a biennial Congress on the subject of cold, to be held in a different country each time. National committees were formed to ensure that all went smoothly. The latter coordinated the submission of reports for Paris, and once the Association of Refrigeration actually got off the ground, they were to have seats on its Executive Committee.

During the opening session on Monday October 5, 1908, the French minister of Agriculture, Joseph Ruau, emphasized that agriculture, being the dominant factor in the economic growth in the second half of the 19th century, profited a lot from the science of cold and its technical applications. After Ruau’s speech, the national committee chairmen were all invited to say a few words. Kamerlingh Onnes, who represented the Dutch government, took the opportunity to define the mission of the International Association of Refrigeration: “to bring together all knowledge bearing on low temperature”[4]. He also emphasized that research on artificial cold and its applications was of importance to all countries and all social classes. The congress on refrigeration, said Kamerlingh Onnes, could help to expand “international solidarity… that precious treasure of humanity”. In conclusion, he emphasized the importance of studying the physical properties of matter at extremely low temperatures. This would further clarify the relationship between matter and electricity, thus preserving the dream of “energy reservoirs of a size that passes imagination”. The French physicist Jacques-Arsène d’Arsonval, who spoke on behalf of the scientific community during the opening ceremony, also emphasized the importance of pure research. “All your machines’, he said, addressing the technicians around the hall, ‘rely on thermodynamic principles”. The scientific community in turn derived great benefit from experience gained in industry: a science-and-technology spiral avant la lettre.

During the closing session, he placed Kamerlingh Onnes in the limelight: his liquid helium made him the star of the Congress.

In the avalanche of recommendations that the Congress adopted on its final day, applied cold technology predominated, but there were also follow-up proposals to the goal that Kamerlingh Onnes had formulated at the opening session. The most striking was: “Given the crucial interest attached to pursuing and coordinating scientific and practical work in the field of low temperatures, the Congress emits the wish of the foundation of an International Association for the promotion of scientific and other studies, with its head office in Paris, which would pursue its study of the whole field of refrigeration and at the same time continue to strengthen the already specialized work centres”.

d) The International Association of Refrigeration

The International Association of Refrigeration duly materialized. It was founded on January 25, 1909 in the presence of delegates from 35 countries. Lebon was appointed President of the new association, and De Loverdo became its director. The Paris Congress led to the establishment of six international committees. Vice-president Kamerlingh Onnes was chosen to chair the “first committee”, which was to focus on scientific matters, and which also included Louis Cailleted, Charles-Edouard Guillaume (of the Bureau des Poids et Mesures) and James Dewar.

While the Association started life with a few dozen members, by the time of the 2nd Congress of Refrigeration, held on October 6-11 in Vienna, Austria, it had 1700. Argentina contributed most, with 1000 members – all because of its frozen meat – and the United States at 370, also had a strong contingent. For all America’s numerical preponderance, however, 92% of the financial contributions came from Europe. The second congress attracted over 3000 participants.

A proposal was adopted to set up a grants system enabling young physicists to perform research “relevant to cold technology” in Leiden’s cryogenic laboratory. The 3rd International Congress of Refrigeration was held in September 1913 in the dual venues of Washington and Chicago[5].

e) Restructuring Association

After the Great War, the Association was restructured into the International Institute of Refrigeration. This was triggered by the resignation of the president, André Lebon, on December 12, 1918. Following this, the director of the Association convened a meeting of the Executive Committee on February 6, 1919. The meeting in the “Crédit Foncier [mortage bank] d’Algérie et de Tunisie” in Paris was attended by only six of the 28 members of the executive committee.

Discussions on restructuring designed to place the Association on a solid financial basis were postponed until the end of the peace talks in Versailles. On June 21, 1920, the Association was replaced by the International Institute of Refrigeration. This had a far more tightly-knit organizational structure, based on that of the International Institute of Agriculture in Rome: instead of individual members it had participating countries in six categories, paying fixed contributions. And these rules are still valid in 2008.

The formal International Congress of Refrigeration, held in the premises of the Ministry of Trade in Paris, was preceded by a meeting of the provisional Executive Committee, to which Kamerlingh Onnes belonged. President Lebon, now back at his post, reviewed the organisational structure: the Bulletin, the grants, and the international committees, including the physics, chemistry and thermometry committee chaired by Kamerlingh Onnes. Lebon opened the meeting by congratulating the Dutchman on his recently acquired status of membre correspondant of the Académie.

In the Great Hall at Rue de Varenne, it fell to Kamerlingh Onnes, “un grand savant à l’avant-garde de la science”, to respond – on behalf of the 42 countries attending – to the welcome speech given by Ricard, the French Minister of Agriculture. Ricard emphasized that the war had truly brought home the benefits of refrigerating food. The next step was to make a concerted effort to improve the accessibility of the cold industry. A more far-reaching and useful goal than improving the refrigeration could scarcely be imagined, the minister concluded. In his response, Kamerlingh Onnes remarked, not without self-interest, that as long as the Institute followed in the Association’s footsteps, success was assured. The science of refrigeration had a golden future and developments had time and again exceeded their wildest expectations. What had begun with a little cloud of liquid air in a Cailletet test tube had grown into a cryogenic industry producing billions of tonnes of oxygen, nitrogen and argon annually both in liquid and gaseous state. This is a real fulfilling of the prophetic words of Jacques-Arsène d’Arsonval, delivered at the occasion of first liquefaction of air in 1877: “Industrial liquefaction of air is not only a scientific upheaval; it is also an economical and social upheaval. Preparation of oxygen and nitrogen by liquefaction of air brings forth an upheaval in illumination, metallurgy, chemical industry, health care and agriculture.” There is not much to be added after more than 100 years.

II – The IIR today

a) Fundamentals

The IIR is 100 years old. It has changed, but some principles are still alive:

The basic reasons for the creation of the IIR are still important: the role of refrigeration in agriculture and food, the importance of science and technology in refrigeration and especially cryogenics, and the need for scientists to share their research.

The IIR is still an intergovernmental organization with six member-country categories. However, it also has, as in the beginning, private and corporate members who receive its services.

The structure of the IIR, with congresses and with commissions or committees, is partially the same. The Bulletin, which was the first IIR publication, has been maintained.

The roots are still presents, but new branches have appeared and the world has changed. The IIR had to adapt to its environment and continues to adapt to new challenges.

b) The challenges

The most important challenges for humanity in the 21st century are health and the environment.

Diseases and mortality are still too widespread in developing countries. The aim to live as old as possible in good health is the goal of most people in developing countries. Refrigeration is one of the answers: it is necessary to guarantee sufficient quantities of food, available to everybody, and to preserve its quality, particularly in order to avoid contamination.

Refrigeration is also necessary to enable the storage and transport of health products (vaccines, certain drugs, diagnostic products…), and to preserve organ and tissue in hospitals, for cryobiology, surgery and medicine.

Refrigeration is at the core of two major threats to the environment: ozone depletion and climate change, because of the use of certain refrigerants and because of the energy it needs. We thus have to implement new refrigerants, to reduce electrical consumption and to develop new environmentally friendly technologies. Refrigeration is also an answer to global warming: air conditioning will be increasingly necessary in many cases and refrigeration is needed in several leading-edge energy sources: liquefied natural gas, liquefied hydrogen, thermonuclear fusion. Moreover, refrigeration will be needed for the capture of CO2 in energy plants, the steel industry…

The IIR is not only still necessary: it is more and more necessary.

c) Members

Our members have changed but not totally. The First and Second World Wars, and decolonization had an important impact. However, the main countries present when the IIR was set up are still there.

May I mention that Czechoslovakia, was one of the founding member countries (it joined in 1921) and both the successor countries, Czech Republic and Slovakia are still active in the IIR.

Originally, there were about 40 member countries; they are now 61. We should have more than 150 member countries according to our mission and global challenges.

The number of private and corporate members is also too low: there were about 500 such members 50 years ago and there are currently almost 600. We certainly need to attract more private and corporate members, and welcome suggestions.

The challenges we have are challenges for governments, but also for all public and private sectors.

d) The committees, commissions and working parties

The IIR is more sophisticated than at the beginning, perhaps too sophisticated. However, it reflects the various fields of actions and the broad mission we have. We have a General Conference, an Executive Committee, a Management Committee, a Science and Technology Council. The latter comprises ten commissions: each one has about 50 members from all paying IIR member countries. Three of them are involved in this conference: Commissions A1 (Cryophysics, Cryoengineering), A2 (Liquefaction and separation of gases) and C1 (Cryobiology, Cryomedicine).

We still have commissions dedicated to the cold chain, but also to issues that did not exist at the outset, such as Air Conditioning. We also regularly create and sometimes disband working parties, which do not have the same official status. However, they are very necessary to actively work on precise subjects. There are several projects in the field of the commissions involved in this conference. I hope we will be able to find here persons to handle them. Your field is important for the future and we need, as for existing working parties, conferences, workshops, publications and statements in your fields.

e) Publications

The first IIR publication was the Bulletin. It was first published in 1910; it still exists, but it has of course changed: it is now an electronic Bulletin (e-Bulletin). We have an electronic database, Fridoc, which now comprises more than 81 000 entries. It is the most important database in refrigeration technologies. The Bulletin now essentially comprises the new entries of abstracts of articles and documents published all over the world and is now merging with Fridoc.

We launched two other publications:

The International Journal of Refrigeration was created 30 years ago. We needed a scientific journal, with the same kind of selection and strict peer-review approach as the best ones. We have succeeded and its impact factor is the best in the refrigeration sector: it is the 26th out of 106 journals in mechanical engineering and 14th out of 42 journals in thermodynamics.

The Newsletter was created 8 years ago. It comprises selected news from all over the world in addition to IIR news and it is sent to our entire network (more than 3000 people).

The IIR progressively began to publish books and guides, thanks to its network of experts, which is our main wealth. We publish technical books, brochures, diagrams, training courses and recommendations. We publish reference documents that are used and recognized all over the world. For example, we publish the International Dictionary of Refrigeration in 11 languages.

Thanks to our intergovernmental status, we are invited to international events and are able to deliver statements, to participate in meetings and side events, to prepare international standards and recommendations for governments, the United Nations and its various bodies, for decision-makers. For instance, we regularly deliver statements during the United Nations Conferences on the ozone layer and on climate change.

f) Conferences and congresses

One of our main activities is now the holding of conferences. We organize about 3-4 IIR conferences per year and we co-sponsor 8 conferences per year on the average. Most of them are series of conferences, like this one, and it is important to have regular events on key subjects for scientists and engineers. I am sure that this one will be successful as usual. Abstracts of the papers presented will be inserted in our Fridoc database and some of them can be presented to the International Journal of Refrigeration. The proceedings will be inserted in our Catalogue of Publications and sold. We will do our best to raise the visibility of your work.

We also organize The International Congress of Refrigeration every 4 years, which covers all fields of refrigeration technologies, as was the case 100 years ago. It is a great pleasure for me that the next one will take place in Prague in 2011. I hope you will all attend it.

Conclusion

In conclusion, the IIR is 100 years old. However, it is still young: refrigeration preserves health and quality of life! We still have a lot of work to do throughout this century, together, for a better health, for a better environment, in a way of sustainable development.

I hope the coming generation will continue to celebrate the IIR in 2108, with many more people and many more countries and companies in healthier and safer world for everyone.

Nota

The first part of this paper partially comes from a text of Dirk van Delft, which will be the introduction of a brochure published for the IIR Centenary in 2008.

References

1 Dienel, H. L., Linde: history of a technology corporation 1879-2004 (2004).

2 Van Delft, D., Freezing Physics. Heike Kamerlingh Onnes and the Quest for Cold ((2007).

3 Museum Boerhaave, archives of Heike Kamerlingh Onnes, inv. No. 190.

4 Bulletin Officiel du Premier Congrès International du Froid, Nos. 1 & 2 (1909)

5 Bulletin Mensuel de l’Association Internationale du Froid (1910-1913)

CR08-62

THE CENTENARY OF THE FIRST LIQUEFACTION OF HELIUM

Scurlock R.G.

Kryos Technology, 22 Brookvale Road, Southampton, SO17 1QP, UK.

ABSTRACT

This paper marks 2008 as the Centenary Year since the first liquefaction of helium. On 10^th July 1908, Professor Heike Kamerlingh Onnes and his team first liquefied helium at his Cryogenic Laboratory, University of Leiden, The Netherlands.

Before describing this liquefaction event, the paper discusses the considerable difficulties Kamerlingh Onnes had to overcome in achieving his success. In contrast to his rivals, Dewar and Olszewski, he adopted the first ever “big science” approach to build up a large laboratory at Leiden, with the extensive infra-structure and expertise needed for his attempt. He also had a strong working relationship, through his experimental measurements on the low temperature properties of gases, with theoretical physicist van der Waals at the University of Amsterdam.

A brief chronology outlines how Kamerlingh Onnes’s success in liquefying helium helped to open the “door” from the classical physics of the 19^th century into the new scientific world of macroscopic quantum physics of the 20^th century.

1. INTRODUCTION

On 10^th July 1908, Professor Heike Kamerlingh Onnes and his team first liquefied helium in the Cryogenic Laboratory, University of Leiden, The Netherlands.

Figure 1. Building of the Cryogenic Laboratory, University of Leiden

Today, at Cryogenics 2008, we celebrate the centenary of this landmark achievement, which opened the door from the 19^th century classical world of the physical sciences, into the strange, new, quantum mechanical world of the 20^th century, with its many macroscopic manifestations at low temperatures.

However, before we look at the consequences of his achievement, let us examine how Onnes overcame many problems to achieve his success, when his main competitors, Sir James Dewar, at The Royal Institution, London, and Dr Karol Olszewski, at The Jagiellonian Unversity, Krakow, failed.

2. ONNES AND HIS COMPETITORS, DEWAR AND OLSZEWSKI

Dewar, having been the first to liquefy hydrogen in 1898 while working with a small team at the Royal Institution, set out to liquefy helium in the same way with the same 2 technical assistants and the same low level of funding.

Olszewski, working again with a small team, set out to liquefy helium in the same manner as his earlier successful first liquefactions, but again with a low level of funding.

Onnes realised shortly after his 1882 appointment as Professor of Experimental Physics at Leiden, at the age of 29 years, that the success of his new Cryogenic Laboratory required a new approach ( probably the first big science approach ). Over a continuous period of 26 years, he set about building up the extensive infra-structure to support a large laboratory and the cryogenic experience of his staff through many research projects on the properties of matter at low temperatures, before turning to

the liquefaction of helium [1]. He instituted a training school for instrument makers and glass blowers, an open house for visiting scientists, a new journal for the publication of all research results at Leiden, and obtained adequate government funding plus royal patronage. In addition, he built up a strong collaborative relationship with van der Waals, who was 16 years older than Onnes and had been appointed in 1877 as Professor of Physics at Amsterdam University.

Figure 2. Professor Heike Kamerlingh Onnes

3. THE RELATIONSHIP BETWEEN ONNES AND VAN DER WAALS

The influence of the friendship between Onnes and van der Waals cannot be underestimated in the successful contributions they both made to physics. Van der Waals was a powerful theoretician, making 2 major contributions: (i) pioneering the concept of a general equation of state to relate the PVT behaviour of a real gas from high temperature down to and below it’s critical temperature, by considering gas molecules to have (a) a finite size, and (b) a finite sphere of influence or interaction between each molecule and its near neighbours in the fluid state, (ii) hypothesising and testing a Law of Corresponding States linking the PVT behaviours of different gases in terms of their critical P, V and T parameters.

Over a period of years, Onnes carried out measurements to test and confirm van der Waals’s theories, including the helium gas isotherms, so as to predict the critical PVT parameters needed to attain liquefaction with his available liquefier equipment.

In fact, van der Waals’s contributions were recognised world-wide by his being awarded the 1910 Nobel Prize in Physics--- 3 years before Kamerlingh Onnes’s award.

Figure 3. Onnes with van der Waals at the helium liquefaction stand

4. RESEARCH FUNDING AT THE END OF THE 19^TH CENTURY

At the end of the 19^th century, in contrast with the Arts, funding was sparse for supporting growth in the Sciences, both as an academic activity and to meet the demands of the new industries of the Industrial Revolution. Only in Germany was the development ( with the royal sponsorship of the Kaiser ) of applied sciences being supported by government funding, through the introduction of the polytechnics. This policy had led to the domination in Europe by the German chemical industry, in the production of organic compounds and aniline dyes, before the end of the century.

Figure 4. Onnes with Flim and students

The Netherlands government adopted the same policy in order to compete (with the royal sponsorship of Queen Wilhemina) and Onnes was one of the fortunate beneficiaries when he was awarded continuous funding for his Cryogenic Laboratory from the beginning of his appointment in 1882.

In comparison in the UK, Royal Commissions had been set up and, for example, had advocated that “education of science in universities should not be specialised” and recommended government funding of £4000 for the endowment of research throughout the UK to be administered by the Royal Society. Dewar could expect little help from this source, having made so many enemies.

In Krakow, there was little or no government funding, and Olszewski and Wroblewski had to use their small resources to carry out their research, which had included the

first liquefaction of oxygen, carbon monoxide and nitrogen. They also achieved the production of liquid hydrogen in a transitory jet in 1884, some 14 years before Dewar’s achievement. However, Wroblewski was seriously burned, and died soon afterwards, from an experiment on the physical properties of hydrogen in 1888. Olszewski continued on his own, to condense and solidify argon. Onnes recognised him as the “precursor of cryogenics”, but that with limited funding he could not be a serious rival in the competition to liquefy helium.

Figure 5. Stand of the first liquefaction of helium

5. GASES AND MATERIAL RESOURCES IN 1908

It is impossible today to imagine how difficult was every aspect of low temperature science in 1908.

Gases

The gases, required in a pure state, for the study of their properties and for use as refrigerants, needed to be produced and purified on a DIY basis as required in the laboratory. They were not available commercially.

Hydrogen was made by the electrolysis of water, purified over heated catalysts and dried before storing at low pressure in gas holders, or at higher pressures in cylinders. Helium was a very rare gas in 1908, and the major recognised source was a rare mineral, monazite sand, which Onnes obtained from North Carolina, USA. When heated, the sand could produce about 1 litre of impure helium gas at 1 bar from each kg of sand. Elaborate purification steps, taking many months, were then needed to remove every contaminant including hydrogen, before the helium product was acceptable for a liquefaction attempt. Onnes used about 400 litres of gas in continuous

circulation for his first successful liquefaction, producing a volume of liquid of a few tens of millilitres.

The occurrence of helium in natural gases ( at 0.3 to 2.0 % ) was not publicised until 1907 by Cady and McFarland [2], while the first production did not start until late 1917 in Hamilton, Ontario, Canada, and early 1918 in Fort Worth, Texas, USA.

Instrumentation

Instrumentation was crude, to say the least, in comparison with today..

A constant volume gas thermometer was the most reliable, employing a small copper bulb filled with helium under pressure, with the absolute temperature varying linearly with pressure down to 10K, about twice the critical temperature; but pressure readings at lower temperatures were not reliable indicators, eg. for the NBP at 4.2K.

Platinum resistance thermometry was the alternative means of measuring temperature, but again below 10K, was not reliable.

Constructional materials for cryogenic use

Metallic materials of construction in the form of thin-walled tubing and cylinders were limited to low thermal conductivity nickel alloys like German silver, medium conductivity copper alloys like brass, and high conductivity copper. No steels or aluminium alloys existed. Joints were made with soft or hard solder--- oxy-acetylene welding had not yet been invented.

The fall back material was glass tubing in the form of high borosilicate Pyrex glass, which turned out to be permeable to helium at ambient temperature; it was later

replaced by low-borosilicate Monax glass, which is non-permeable. As a result, the major parts of laboratory cryostats, dewars, liquefiers and associated pipework were made of glass, with the aid of highly trained glass blowers.

This lack of materials and equipment resources for low temperature research did not change until the mid 1950’s. Until then, the idea of cryogenic applications using liquid helium appeared to be impossible.

6. THE FIRST HELIUM LIQUEFACTION, ON 10^TH JULY 1908

After 26 years of preparation, construction and trials, Kamerlingh Onnes and his team of technicians, led by Gerrit Flim, were ready for the first attempt to liquefy helium.

On 9^th July 1908, preparations began with the production of enough liquid air for the next day, via the cascade liquefier of methyl chloride, ethylene and oxygen. Some of the liquid air was then used as a precoolant for making sufficient liquid hydrogen for the liquefaction attempt.

On 10^th July 1908, work started at 0545, the first elaborate and tedious job being to remove the last of the impurities from the helium gas down to the lowest possible level. The gas was passed over copper oxide, after which oxygen and gases of similar volatility were removed by freezing them out in liquid hydrogen. The helium was then compressed and passed over charcoal, at successively liquid air and liquid hydrogen temperatures, for several times until all impurities had ben removed as far as practicable. The purpose of these purification stages was to prevent blockages caused by freezing out of residual impurities in the fine passages of the high pressure pipework of the liquefier.

At 1620, the purified helium was admitted into the liquefier cycle, which involved precooling with liquid air at 80K, liquid hydrogen at 20K, and pumped hydrogen at

14K, close to it’s triple point. At 14K, the compressed gas was well below it’s estimated JT inversion temperature for helium, and was led into a Hampson recuperative cooling spiral ending in a JT expansion valve.

By 1900, the gas thermometer had reached an apparent temperature of 5K and stopped moving down, but no liquid could be seen. It was then realised that the thermometer was behaving as though immersed in liquid.

Going beneath the cryostat, with an electric light, Onnes looked upwards into the glass cryostat and clearly saw the liquid meniscus “standing out sharply ( once seen ) like the edge of a knife against the glass wall”.

He had indeed liquefied helium [3].

His immediate regret was that his friend Professor van der Waals was in Amsterdam, and would have to wait several days to see the liquid for himself at the second liquefaction run.

On the same day, with his first liquefaction run, Onnes was able to measure the normal boiling point at 4.3K, and estimate the critical temperature at 5.0K . He also noted the unexpectedly low density of the liquid at 125 g/litre (125kg/m3), and the extremely low surface tension.

He also tried to reach the triple point, by reducing the vapour pressure. He reached 7mm Hg (corresponding to 1.7K ) but the helium remained liquid.

For the next 13 years until 1921, Onnes would use larger and larger vapour pumps, probably reaching a minimum temperature of 0.83K, but still the helium remained liquid. It was beginning to appear that helium remained liquid to absolute zero.

The same day, Onnes sent a telegram to Dewar announcing his success. Dewar’s reply shows his complicated feelings thus:

CONGRATULATIONS. GLAD MY ANTICIPATION OF THE POSSIBILITY OF THE ACHIEVEMENT BY KNOWN METHODS CONFIRMED. MY HELIUM WORK ARRESTED BY ILL HEALTH BUT HOPE TO CONTINUE LATER ON.

7. THROUGH THE “DOOR” TO MACROSCOPIC QUANTUM PHYSICS

Five years later in 1913, Kamerlingh Onnes was awarded the Nobel Prize in Physics for his successful liquefaction of helium. His discovery of superconductivity in mercury in 1911 was not even part of the reasons for his Nobel award. However, he made certain in his acceptance speech to mention superconductivity, expressing his wonder about “the abrupt loss of electrical resistance”. In addition, he mentioned the extremely low density of liquid helium, and suggested to the Nobel audience that explanations for these strange phenomena “could possibly be connected with the new quantum theory”. How right he was!

8. SUBSEQUENT DEVELOPMENTS: A BRIEF CHRONOLOGY [4]

Following on the work of Onnes, in 1926 Keesom solidified helium at Leiden by the application of pressures above 26 bar, and established the melting curve using an all-glass cryostat.

In the 1930’s, the phenomenon of superfluidity in liquid helium was identified and the associated effects of λ-specific heat, fountain effect, wall-film flow etc. were measured and formulated into various theories.

Before the mid 1950’s, liquid helium temperature research involved cryostats which included their own miniature helium liquefiers and glass dewars. Experimental work was slow, time consuming and sometimes dangerous with high pressures inside glass systems. Then the availability of 4-8 L/h Collins helium liquefiers, and the like, enabled the original few, and many new, low temperature physics laboratories to have a centralised liquid facility, or to have access to a commercial liquid helium facility and distribution system via 25, 50, 100 litre, or larger, portable metal dewars.

Superconductivity remained very much in the background (although it had been found in many non-magnetic solids), until 1957 when the BCS theory of Type I superconductivity via pairing of electron states was proposed and tested [5].

But it was not until 1961, when high current, high field, Type II superconductors were discovered, that any application of superconductivity could begin to be considered seriously [6].

A further 10 years of development led to the large scale manufacture of NbTi and Nb3Sn conductors suitable for power engineering applications in motors, generators, cables, fault current limiters and transformers. However, the power engineers have

hesitated since then, about their use of Type II superconductors, because of the large scale liquid helium requirements.

The 1980’s saw the new development of 5 T, liquid helium cooled, superconducting magnets with a 1m diameter bore, leading directly to the invention of Magnetic

Resonance Imaging and it’s use as a medical diagnostic tool. Since then, most hospitals around the world have invested in MRI, all using liquid helium, while higher fields, lower stray fields, and improved computer techniques are leading to functional MRI (eg. for studying brain functions ) when combined with liquid helium-cooled, Josephson-device, encephalography.

A spin off during the 1990’s has been the development of miniature closed cycle cryocoolers to replace the use of liquid helium altogether. Thus, laboratory cryostats for all kinds of application at liquid helium temperature can now operate in a liquid free state at 4K, just like a domestic refrigerator at 263K.

On the other hand, the large scale use of liquid helium has expanded rapidly since 1980 into space science and particle physics, employing ever larger and ever more sophisticated superconducting solenoid, dipole and quadrupole magnets.

The confidence gained during this period has led directly to the design, building and commissioning in 2008 of the Large Hadron Collider LHC particle accelerator at CERN.

The LHC is in a tunnel 26 km in circumference, with liquid helium cooled 3 Tesla superconducting magnets at 1.8K, around the whole length of the tunnel. The weight of the superconducting magnets is a total of 36,000 tonnes. 36 MW of power is needed for the refrigeration plant, in which the total helium inventory is 138 tonnes of liquid helium
(equivalent to 1.1 million litres of liquid; more than the total inventory for the rest of Europe ).

The future of liquid helium is bright as more and more natural gas streams are being stripped of their helium, before being exported as fuel gas or chemical feedstock. Even larger applications than the LHC can be envisaged, particularly in the development of fusion reactors with their requirement for very large volumes of high magnet fields, which can only be created by liquid helium cooled superconductors at present.

Following their discovery in 1986 [7], the new era of ceramic superconductors, operated at liquid nitrogen temperatures rather than liquid helium temperatures, now appear to be particularly suitable for power engineering applications, with high current densities at relatively low magnetic fields of 1 Tesla. Prototype cables, motors, generators, fault current limiters and transformers are presently being built and tested.

However, where high magnetic fields of 5 – 10 Teslas, or higher, are required as in MRI, liquid helium temperatures will continue to prevail for the present.

9. REFERENCES

1. Scurlock R.G, History and Origins of Cryogenics, (1992), Oxford University Press, Oxford.

2. Cady H.P. and McFarland D.F., J.Am.Chem.Soc.,(1907) 29 1523.

3. Kamerlingh Onnes H., Commun.Phys.Lab., Leiden (1908) 108 ; Proc.R.Acad., Amsterdam (1908) 11, 168.

4. Timmerhaus K.D. and Reed R.P.(Editors), Cryogenic Engineering:Fifty Years of Progress, (2007), Springer, New York.

5. Bardeen J., Cooper L.N. and Schrieffer J.R., Phys. Rev .,(1957) 108 1175.

6. Kunzler J.E., Buehler E., Hsu F.S.U. and Wernick J.H., Phys. Rev. Letters, (1961) 6 89.

7. Bednorz J.G. and Muller K.A., Z. Phys., (1986) B64, 189.

CR08-56

FIFTY YEARS FROM HELIUM LIQUEFACTION
IN CZECHOSLOVAKIA
AND A NEW TURBINE TECHNOLOGY

Kaiser Z.¹, Kouba M.², Kundera R.³, Prušák J.⁴, Šafrata S.⁴, Schustr P.⁵, Chrz V.²,

¹ Ingersol Rand, Prague, Czech Republic (Ferox Děčín formerly)
² Chart Ferox, Děčín, Czech Republic
³ PBS, Velká Bíteš, Czech Republic
⁴ Institute of Physics, Academy of Sciences of Czech Rep., Prague, Czech Rep.
⁵ ATEKO, Hradec Králové, Czech Republic

ABSTRACT

When celebrating 100 years of helium liquefaction at the conference Cryogenics 2008 in Prague, it is worth to look back onto the part of helium history in this part of world and in the Czech Republic specifically. It is just fifty years from first steps for helium liquefaction in the Institute of Nuclear Physics of the Czech Academy of Science. Technology of Helium liquefiers has been developing in the Ferox company for thirty years, then. This research required development of helium turbines, which was realized by ATEKO and PBS companies. Successful development of highly efficient turbines with magnetic brakes using dynamic gas bearings was a base of new production branch, which has been continuing up today with series supplies to the most recognized European manufacturers of helium liquefiers.

PIONEERING AFTER THE WORLD WAR II

Helium liquefaction was a long- time privilege of several laboratories in the world. Still in the year 1946, there were fifteen labs only in the world, who operated a helium liquefier with a performance of several liters per hour. These were results of several-year work of teams of physics and technologists of particular institutions.

Also in Czechoslovakia, research in the branch of low temperature nitrogen and helium physics and technology was developing rapidly after the World War II. After founding the Department of Low Temperatures of the Institute of Nuclear Physics of the Czechoslovakian Academy of Sciences in Prague in the year 1956 the need of access to liquid helium became critical. It was decided on building own liquefier using experience and documentation of the Moscow Institute of Problems of Physics with direct support of the Academy member P. L. Kapitsa (Fig. 1) and his team.

Under supervision of the Department of LT, the liquefier was designed and manufactured in the KSB works in Děčín (Chart Ferox, a.s., now) , in the years 1957 to1958. It was a cascade type using subcooling with liquid nitrogen and liquid hydrogen. The performance was 8 liters of LHe or 15 liters of LH2 respectively. It was built in the new premises of the LT Department in Řež near Prague, where in the night from 12 to 13 April 1960 helium was liquefied first. As the second installation, Linde liquefier with performance of 3 liters/hour in the same year.

CUSTOMIZING OF KAPICA’S LIQUEFIER

Increasing needs of other laboratories in Czechoslovakia and some other countries of the Central and Eastern Europe faced difficulties with import from the USA or Germany, where the helium liquefiers were already commercialized. Based on successful manufacturing of the first liquefier, it was decided that commercial production would be located at KSB, Děčín. This was again supported by P. L. Kapitsa. His Institute provided documentation as a base for the new design, already working as prototype, with a piston expander instead of hydrogen subcooling.

This type of liquefier was suggested and first time built by Kapitsa in 1934. During the World War II, liquefiers on this basis were built by Collins in USA and by Meissner in Germany.

Textové pole: Figure 3. Helium liquefier ZH9 with the oscilloscope and the records of the indicator diagrams from top: a) normal, b) with premature... c) ...and delayed closure of the exhaust valve. The period of development and manufacturing of these liquefiers, namely ZH4 and ZH9 (Fig. 3) with production 4,5 and 9 liters LHe/hour respectively started in Děčín in 1964. Both the types were equipped with expanders. Although manufacturing of apparatuses of copper and stainless steel was on high level in the works that time, manufacturing of expander was a new challenge for design and manufacturing technology.

The principal part of the liquefier, decisive for the correct function and especially for the performance of this unit, is the piston expander. The expander must operate with a high adiabatic efficiency and reliably. This is rather an exacting demand in view of the low temperatures (28 to 12 K) of the moving parts such as the piston with cylinder and valves. Since all known lubricants are unusable at the service temperature of the expander, gas lubrication is effected here directly by the working medium, i.e. helium. There is a small gap between the piston and cylinder, through which some of the expanded helium escapes and forms a gas layer preventing contact of the surfaces of the piston and the cylinder. The optimum width of this gap is given, on the one hand, by the admissible amount of the escaping gas, which should be as small as possible, and, on the other hand, by the necessity of preventing possible contact between piston and cylinder. The piston diameter amounts to 35 mm, clearance is 0,008 mm at the diameter at the working temperatures. Since the material of the cylinder is stainless steel, while the piston is made of carbon steel with a surface of plastic, both parts have different coefficients of thermal expansion and must consequently be made with a wider clearance in such a way that only after cooling to the working temperature is the mentioned value of 0,008 mm attained.

A part of the measuring system of the liquefier was an electric measuring system recording a continuous indicator diagram of the piston expander during operation. This diagram represents the dependence of the pressure in the working space of the expander on the position of the piston in the cylinder. It provides the best information on the function of the expander and it makes possible control of its operation on an oscilloscope (Fig. 3 a to c).

The first liquefier was finished in the Děčín works (part of the trust Chepos that time) in 1965. Two years later, the liquefiers ZH4 and ZH9 were installed in several laboratories in the Czechoslovakia, Poland and DDR (East Germany). Successively, other were built not only in these countries but in Poland, Hungary and Bulgaria. Availability of liquefiers enabled progress in basic research in physics, superconductivity and their applications in countries of the eastern block of that time divided Europe. Encompassing of that time difficult problems of manufacturing and operation of these liquefiers was only the first step, which opened way to further development.

However unique were the first liquefiers with their concept, the operation and maintenance required continuous attention. Similarly it was with the accessories. Oil lubricated piston compressors with insufficient oil separation did not allow long time operation. Helium, liquefied in a little separator inside the liquefier, was periodically and with thermal losses transferred to larger outside storage Dewars, those of relatively small size with nitrogen shielding. The expander worked on constant speed, the only way of control of performance was the filling of the cylinder by the timing of valves.

Modernization was the next step starting with 70’s. A new oil free piston compressor was developed by CKD Prague (Fig. 4). The new liquefier-refrigerator enabled liquefaction directly into a larger Dewar vessel with higher level of automatic operation and lower maintenance.

Progress in applications of superconductive magnets including first industrial applications required large quantities of liquid helium. Two types new liquefiers-refrigerators with oil free compressors and effective heat exchangers, as well as continuous modules of helium purification were the answer to the challenge. They were still equipped with the Kapitsa type expander, but of larger size and with electronic control of speed. Development of control computers allowed continuous automatic operation in adjustable regimes.

The largest liquefier ZHR50 was designed in Děčín, that time already the state enterprise Ferox, in the half of 70’s. It had two piston expanders, built-in helium purification and fully automated operation. The performance was 30 to 100 liters of helium/hour or 150W at 4,5K in refrigeration regime respectively. This liquefier is still in operation at the Slovak Academy of Sciences in Bratislava. (Fig. 5)

Another new type was a compact liquefier ZHR20 (Fig. 6), with a single expander. Its performance was 25 liters of liquid helium per hour. It was delivered for several important projects on applications of superconductivity. It was able to work in conditions of industrial surrounding in month’s long operation.

HELIUM DEWARS, CRYOSTATS AND CRYOGENIC SYSTEMS

Textové pole: Figure 6. ZRH 20 liquefier
with a helium Dewar
500 liter. A He³- He⁴ diluents refrigerator was designed also in 70’s for continuous cooling of solid samples to temperatures in the range of tens of mK ( Fig. 7 - with Stan Smrž).
A range of superinsulated vessels for storage and transport of helium, as well as purposely designed cryostats for the needs of research in low temperature physics were developed and delivered that time.

Top products of that kind were

- the cryostat for a superconductive quadrupole installed at the Institute of Nuclear Research in Dubna near Moscow (Fig. 8)

- cryostats for superconductive magnets of gyratrones for the Kurtchatov Institute of High Energies in Moscow

- a super large helium cryostat 3,2 m in diameter and 10 m in length for tests of large superconductive magnets for the IVTAN Moscow

- a cryostat for a NMR spectrometer developed by the Institute of Scientific Instruments of ASCR was distinguished by a very low evaporation of liquid helium. It was also manufactured at Ferox, then.

A very unique program was a rotating cryostat for a superconductive generator Škoda with a power of 5 MVA with a superconductive rotor. The rotating cryostat allowed cooling of winding at a speed of 3000 rpm with liquid helium delivered from the storage vessel over a rotating joint.

For an experimental motor Skoda 55 kW with a superconductive winding, a cryostat with minimum gap between the superconductive stator and the rotor and with ability to resist a high torque between the superconductive winding and the outer vessel of the cryostat was developed.

Textové pole: Figure 9. SC magnet of the separator of kaoline.
In front: Demagnetization blanket of the separation matrix. The most complex cryogenic system was developed, built and operated for a project of a superconductive magnetic separator of kaolin. By separation of oxides of iron and manganese, high whiteness of kaolin was achieved for production of top quality china.

An industrial-scale magnetic separator for cleaning of kaolin clay was designed, installed and tested during years 1983 – 1988 in a kaolin clay cleaning plant. The superconducting system of the separator worked with the magnetic field up to 5 T in the working space of the separator, whose active cavity dimensions are 560 mm diameter and 1320 mm lenght (Fig. 9). The cryogenic system comprises two helium liquefiers of different concept, one of them operating with a piston type expander, the other one with an expansion turbine. This project was one of the first real industrial applications of superconductive magnets in the world.

FROM THE PISTON EXPANDERS TO TURBINES

Further demand on industrial applications of superconductivity required highly reliable liquefiers. Rotating machines were the answer of designers. Screw compressors instead of the piston ones and radial turbines instead of piston expanders.

Development of expanders was started at the Research Institute of Food and Refrigeration technology (ATEKO, a.s., in Hradec Kralove, today) in co-operation with PBS, state enterprize, in the middle of 80’s. The result was implemen-tation of manufacturing systems of two types of turbines HEXT 0.5 with the speed 216 000 and 237 000 rpm at the inlet temperature 9K and 15K respectively, with cooling power 91 W and 220 W (Fig. 10). This design concept of ATEKO is original and unique in the world by now. Dynamic gas bearings lubricated by the working helium and eddy current magnetic brake were the distinguishing characteristics of the reliable and compact design. The other were small size, low weight, high thermodynamic efficiency and a very simple speed control, which allows to maintain optimum speed for achieving maximum thermodynamic and refrigeration efficiency according to actual process parameters. The first Ferox liquefier with two turboexpanders ATEKO (Fig. 10 a) was the ZRH30T was supporting the above described superconductive magnetic system of kaolin purification.

The last developed liquefier-refrigerator was again the two-turbine type ZRH3T, intended for continuous refrigeration of helium cryostats, including those for magnetic tomography. Oil lubricated screw compressor with highly efficient oil separator was used for helium circulation. The development of this first Czechoslovak system of helium liquefaction without any piston machines was finished by beginning of 90’.

After political and economical changes by beginning of 90’s the state supports of research in superconductivity were drastically reduced and the existing liquefier markets collapsed. This is why the design and manufacturing of helium liquefiers and cryostats was abandoned in Ferox and the company, in frames of privatization by Air Products and Chart later, concentrated fully onto the branch of storage and distribution equipment of liquefied air gases and natural gas (LNG), which resulted in new progress and successful entering the world market as one of the largest suppliers.

In total, Ferox delivered 45 helium liquefiers in the period 1964 to 1992.

Nevertheless, development and manufacturing of helium expansion turbines continued successfully in PBS with deliveries to the main helium liquefier manufacturers in the world.

HELIUM TURBINES AS AN INDEPENDENT PRODUCTION BRANCH

First of all PBS deliveries of turbines started for Linde A.G., Munich. 258 turbines were delivered for the liquefier LKKA for refrigeration of superconductive magnets Siemens during the period 1988 to 1992. Turbines were delivered in two pieces for each liquefier. The same it was with all the liquefiers described below except the liquefier Criotte Impianti with four turbines PBS.

ATEKO developed new larger turbines HEXT 1 and HEXT 1.8 with cooling power 1000W and 1800 W respectively during 1988 to 1991. 14 of them were delivered to liquefiers of Linde and Ferox. PBS, a.s. developed new type of 3D radial-axial wheels. Measurements, done by ATEKO at the customer, indicated high turbine efficiency 78 to 80%.

Since 1991 PBS continued their development independently. Step by step, new types were put over to other liquefiers.

Type HEXT 0.5 equipped with 3D wheels was delivered in 10 pieces for 5 liquefiers L5 of Linde.

New type HEXT 1,5, cooling power 1500 kW has been delivered in 56 pieces since 1995 for the Linde liquefier, marked TCF 10.

Another achievement as acceptance of the next type HEXT2 with maximum power 2000 W for the L’Air Liquide liquefier HELIAL1000 since the year 2002 in the amount of 30 pieces.

Another type HEXT1.1 was delivered in four modifications for a university liquefier of the company CRIOTEC Impianti in the year 2007.

Textové pole: Figure 11. The PBS HEXT turbine
a) (left): The turbine body.
b) The turbine 3D wheels 10 to 18,5 mm diam.

With systematic development and innovations, PBS achieved considerable results in increasing of the thermodynamic efficiency of turbines (Fig. 11) and reliability of their operation:

- The speed was increased up to 360 000 rpm, which enabled to design turbines with maximum efficiency for all the input parameters.

- Circumferential velocity on the outer diameter of the wheel achieved 370 m/s.

- Implementation of 3D design and manufacturing of axial wheels on CNC machine tools. (Fig. 11.b)

- Considerable increase of bearing capacity of the axial bearing. Circumferential velocity on the outer diameter achieved 520 m/s.

- Radial bearings were improved

- Method of dynamic balancing of turbine rotors at operation speed was implemented. Measurement of the amplitude of rotor vibrations proved considerable reduction at commercial series production.

- CFD simulations of helium flow and consequently new guiding baffle profiles and profiles of the channels of the low temperature part resulted in higher turbine efficiency.

- For reduction of parasitic heat flow into cold helium the low temperature part of the turbines and the cold end of the shaft were redesigned.

- The control unit delivered with the turbines makes possible to operate turbines with increased speed during the warm start and cooling down. This results in higher efficiency and shorter startup time.

Total number of helium turbines manufactured and delivered up today is 376.

Development and manufacturing of turbines in PBS continues.

After customizing all this improvements and achievements, PBS took active part at development of baffle type cryogenic compressors for helium refrigerating systems for accelerators of elementary particles for CERN, accelerator Elbe at Rosendorf and others.

Flow systems were delivered for radial compressors of Linde and Air Liquide (Fig. 12).

Portfolio of cryogenic products of PBS was enlarged by design and delivery of flow parts and the low temperature part of three circulators of supercritical helium for completion of the Linde delivery for the accelerator Wendelstein 7 in Greifswald .

CONCLUSIONS

Development of helium liquefiers in Czechoslovakia was stimulating research of physics of solid matters and superconductivity in the Central and Eastern Europe in the period 1964 to 1991. On this base, new branch of helium expansion turbines brought to a level of series production and top efficiencies has been successfully progressed up today.

REFERENCES

1. Kaiser Z.: Equipment for Helium Liquefaction, Czechoslovak Heavy Industry (Journal), RAPID, Prague, July 1968

2. Kaiser Z., Fojtek J., Kouba M., Kotva J., Šuma J.: Magnetic Separator with a superconducting magnet and a reciprocating matrix. Proceedings of the conference CryoPrague 86, published by IIF-IIR at Ceuterix s.a., Leuwen 1986

3. Z.Kaiser, P.Vykydal, J.Fojtek,, S.Smrž, M.Kouba, J.Šuma: Cryogenic Magnetic Separator, Proceedings of the 9th Int. Conf. on Magnet Technology, Zurych, 1985

4. Tuček L., Kundera R.: Design, Manufacture and Operational Characteristics of helium turbines used for helium liquefiers, Proceedings of the conference Cryogenics 94, House of Technology, Ústí nad Labem, Czech Rep. 1994

5. Kundera R., Červenka B., CFD application in cryogenic turbines design and testing, Proceedings of the conference CryoPrague 2006, ICARIS, Prague, 2006

CR08-17

Improvements of helium liquefaction / refrigeration plants and applications

Muehlegger M., Berdais K.-H., Wilhelm H., Ungricht Th.

Linde Kryotechnik AG, Daettlikonerstrasse 5, CH-8422 Pfungen, Switzerland

ABSTRaCT

Design features for a new range of helium liquefiers and refrigerators with capacities ranging from 30 to 280 l/h of liquid helium (LHe) and 100 to 900 Watt, respectively. The latest He cold box development shows an increased efficiency due to improved turbine and heat exchanger design. Other benefits of the new design include short cool-down times and a very compact design, which offers better flexibility and process control. The modularity of the system was designed in order to cover a wide range of applications like sophisticated shield cooling at different temperature levels or simultaneous operation modes for He liquefaction and refrigeration purposes. The presentation will highlight the individual improvements in the design.

During the presentation the influence of certain parameters like power requirement and cold box inlet pressure in relation to the liquefaction and refrigeration capacity shall be shown and discussed for the range of newly developed Helium liquefiers. In addition, the presentation will cover the latest results of recently installed liquefiers in comparison with the previous model.

Keywords: Liquefier, Refrigerator, L-Series, Turbine

Introduction

High reliability, availability, low operational costs and short delivery times have become key requirements for small scale helium liquefiers. In addition, the demands concerning product design and user friendliness have risen. By using standardised components, like heat exchangers, expansion turbines and control software of highest quality, the L-Series can be designed and manufactured according to customer’s requirements and specifications maintaining short delivery times. This paper shows the impact of these optimised plant components on liquefaction rate and specific power input.

Liquefaction process

L-series liquefiers are designed for liquefaction rates at 4.4 K from 20 l/h up to 290 l/h. The range is covered by three sizes of liquefiers - L70, L140 and L280, which are all based upon a Claude cycle. FIGURE 1 shows the process flow diagram of an L-Series plant.

High pressure (HP) helium gas supplied by the compressor system enters the cold box. It is cooled down in heat exchanger E3110 and E3120 by counter-flow low-pressure (LP) helium gas. At the cold end of E3110 a liquid nitrogen (LN2)-evaporator is integrated so that pre-cooling of the HP stream with LN2 becomes possible and the refrigeration or liquefaction capacity of the plant is enlarged. The heat exchanger E3120 has two sections. Between these two sections the high-pressure stream is split in two parts. The larger fraction expands in turbine X3130. After a further cool-down in heat exchanger E3140 it enters turbine X3150. It is expanded to low pressure and finally joins the returning JT-stream. The smaller fraction, called Joule-Thomson stream, continues to be cooled down in heat exchangers E3120 - E3160. After that it is throttled by the JT-control valve to dewar pressure and gets partially liquefied in the dewar. The gaseous fraction is returned as a LP stream to the cold box. It is warmed in the heat exchangers before returning to the intake of the compressor.

Impure helium is fed to the integrated purifier. By cooling down the impure gas in counter current with cold helium HP gas, impurities like nitrogen and traces of other gases condensate and/or freeze out. The purified gas is fed into the cold-box HP-inlet side. By warming up the purifier it will be regenerated and the impurities will be discharged.

Figure 1: Process Flow Diagram of an L-Series helium liquefier by Linde Kryotechnik AG

Design features and improvements

Based on the well proven TGL turbine design and technology, the new TED (Turbo Expander Dynamic) turbines have been developed. A TED 16 turbine is shown in FIGURE 2. Schoenfeld H., et al. [1] showed that efficiencies could be increased between 13 to 20% per turbine. Improved bearing design allows higher thrust capacities and wider operation range, leading to dramatically reduced cool-down times. The robustness of the TED turbine is based on special bearing materials. The TED turbines are absolutely maintenance free. Thus highest availability and reliability are ensured, providing an expected MTBF (Mean Time Between Failure) of around 250,000 operating hours.

The L-series is equipped with a heat exchanger block, consisting of five counter-flow brazed-aluminium plate-fin heat exchangers. All heat exchangers have the same block length and identical plate-fin design concept. The heat-exchanger surface has been increased compared to the TCF plants and pressure drops within the heat exchangers could be reduced. The L-Series heat exchanger can be used for both liquefaction and refrigeration plants. This concept ensures high operation flexibility in case of a combined liquefier and refrigeration (mixed-mode) plant.

The outer piping has been minimised. The liquefaction and purification process has been designed in such a way that no icing on outer pipe-work shall occur. The entire L-Series (L70, L140 and L280) have the same design layout. Their external appearance only differs by the cold box diameter. FIGURE 3 shows a L140 liquefier.

All L-Series plants use the same Siemens S7-300 control strategy and the same software (PLC). In addition, a state-of-the-art visualisation is available for the L-Series. The control panel is separated from the cold box, which provides additional flexibility for the installation of the plant. Plant components are operated with a decentred control system and are connected by PROFIBUS DA. Field wiring is minimised and maximum flexibility concerning placement is ensured.

Figure 2: TED 16 turbine with cooler

Figure 3: State of the art helium liquefier L140

Table 1 Performance data of L-Series plants with and without LIN precooling compared to TCF plants

Comparison of performance data: L-Series vs. TCF

TABLE 1 shows performance data of the L-Series liquefier in comparison to the TCF-Series. Liquefaction rates of TED equipped L-Series plants are based on constant isentropic efficiencies of 75% for turbine 1 and 80% for turbine 2, according to the measurements of Schoenfeld et al. [1]. The cold box inlet state is 13.0 bar and 313K for L-Series and TCF plants. Dewar pressure is 1.20 bar (4.4 K) for L-Series plants and 1.25 bar (4.45 K) for TCF plants. The assumed heat in-leak into the dewar and transfer line are 10 Watts for both types. Using the same mass flow and thus compression power, liquefaction capacities have been increased between 47 and 97% for operation without LN2 pre-cooling. For operation with LN2 pre-cooling, the increase of liquefaction capacity is between 51 and 117%. In [1], a TCF20 was equipped with TED turbines. The increase of the liquefaction rate was 52% compared to a TGL-equipped TCF20. The combination of TED turbines and improved heat exchangers in L140 lead to an increase of up to 102% compared to a TCF with the same compressor mass flow. Consequently, the specific compression power input per liquefied helium could be significantly reduced. For liquefaction, specific compressor power is defined as

(1)

where P_comp is the shaft power of the helium recycle compressor. In TABLE 2 data for the L-Series and TCF plants are presented. The required power for producing the LN2 is not considered for the specific compression power.

For liquefaction of one litre liquid helium at 4.4 K without LN2, between 1.54 and 2.22 kW of compression power is required for an L-Series plant. For TCF plants, between 2.62 and 4.06 kW are necessary (TABLE 2). This means that an L280 requires a third less compression input power per produced litre LHe than a TCF50. TABLE 2 shows p for liquefaction with LN2 pre-cooling as well. L-series plants require between 33% and 50% less specific compression power for operation with LN2 pre-cooling.

Results from the First Operating Plants

To date more than twenty L-series plants have been ordered and several have already been commissioned. The first four plants are already in operation. The L280 at Ibaraki, Japan is operating with 9.6 bar and 313K high-pressure inlet state.

Table 2. Specific compression power for L-Series plants compared to TCF plants

Liquefaction capacities of 220l/h without purifier operation have been achieved. Cool down time from ambient to operation temperature is less than 2.5 hours. To cool down a similar TCF50, approximately 4 hours were necessary.

Conclusion

By using high-efficient TED turbines and optimised heat exchangers, liquefaction capacities have been increased by 50% to 100% in comparison to the old liquefier generation. The power input per litre LHe produced has been decreased by 33% to 50%. The L-Series from Linde Kryotechnik AG established a new benchmark for small scale helium liquefiers and refrigerators.

References

1. Schoenfeld H., Cretegny D. and Loehlein K., "Standard liquefier-test results with improved turbines," Proceedings of ICEC – 20, Beijing, China, 2004, pp. 119-122

CR08-47

EVOLUTION OF THE STANDARD HELIUM LIQUEFIER RANGE

Caillaud A., Aigouy G., Crispel S., Delcayre F., Grabié V., Dauguet P.

AIR LIQUIDE, Advanced Technologies Division, Rue de Clémencière,
B. P. 15, 38360 Sassenage, France

ABSTRACT

The standard helium liquefier and refrigerator range, called Helial and designed by Air Liquide DTA, has recently been upgraded in order to improve the efficiency of these machines. Indeed, over the multi-range markets requiring these cryogenic systems, (international laboratories, aerospace applications, synchrotrons, HTS applications...), the technological solution has to provide increasingly high performances. The new range, equipped with very reliable DTA turbo-expanders, constitutes a highly efficient product for this wide application field. The optimizations, adaptations and results of the Helial Evolution series, doubling the performance for the same power consumption, will be presented.

INTRODUCTION

The Helial was born in the 1980s. These machines were revolutionary in that they constituted the first helium liquefiers to be fully automatic and therefore easily operable. Nearly 30 years later, their refrigeration and liquefaction capacities have grown enormously, but helium liquefiers-refrigerators still operate on the same principle. With a 30-year wealth of experience, Air Liquide decided to assess the situation, looking through all the projects. This study led in 2007 to the launching of a new range named Helial Evolution. Indeed, in the demanding high-tech markets, cryogenic systems such as standard liquefiers and refrigerators must provide increasingly high performances with strong reliability. This paper describes the capitalization of the last six years regarding small and medium liquefaction. The results of this study are presented as well as the evolutions they generated on the standard machines. Process and design improvements are detailed, and finally benefits and adaptations of these new standard machines are described.

1. CAPITALIZATION OF THE PAST SIX YEARS

In 2001, Air Liquide upgraded its standard helium liquefiers range constituted by the Helial 7, Helial 20 and Helial 50. Three new liquefiers/refrigerators were created and called Helial 1000, Helial 2000 and Helial 3000. Their performances, which are presented in TABLE 1, resulted from an evolution of the liquefaction market characterised by the arrival of third generation synchrotrons requiring dedicated cryogenic systems with refrigeration and mixed-mode operations. At the same time, small liquefiers were still required all over the world, particularly in Asia with the premises of helium liquefaction in new developed countries.

	Helial 1000	Helial 2000	Helial 3000
Liquefaction capacity without LN2	40 L/h	85 L/h	180 L/h
Liquefaction capacity with LN2	80 L/h	175 L/h	300 L/h
Refrigeration capacity without LN2	130 W	415 W	750 W
Refrigeration capacity with LN2	160 W	500 W	900 W

Table 1. Helial 1000/2000/3000 performances

Figure 1. Characterisation of past Helial projects.

The design of the Helial 1000/2000/3000 range was therefore adapted to these specific requirements, trying to find a good compromise between refrigeration and liquefaction, in order to fulfil specifications requiring various modes of operation.

After six years of experience with the Helial 1000/2000/3000 range, past projects were analysed. First, Air Liquide examined the field of applications for which these systems were installed. FIGURE 1 – upper left corner – shows the variety of customer applications for which the Helial systems provided a solution. Two main fields appear: synchrotron centres and liquefaction centres. They constitute about 75% of the needs in terms of helium liquefaction and/or refrigeration. Nevertheless, the remaining 25% reveal new markets like cold and ultra-cold neutron sources, HTS applications or neutral beam injectors for fusion applications, which should develop in the near future for helium refrigeration. Therefore, Helial machines should constitute a solution for cryogenic needs within these new fields. On the upper right corner of FIGURE 1, the operation modes for past projects have been distinguished highlighting a quasi-perfect repartition between liquefaction, refrigeration and mixed modes. These two repartitions, i.e. final applications and operation mode required, convey the difficulty to design a standard machine able to cover the whole range of applications, satisfying the diversity of users.

Regarding the operating temperatures, FIGURE 1 – bottom left corner – shows that most of the projects work at liquid helium temperature. Nevertheless, in recent years, more and more projects at temperatures above 10 K were born in different fields such as HTS applications, cold neutron sources and space chambers. The standard machines must remain flexible to allow, with minor modifications, operation at this temperature range.

Finally, the last part of the analysis of past projects concerns the on site performance. The graph on FIGURE 1 – bottom right corner – shows that more than 40% of the installations in operation exceed the guaranteed performance by more than 30%. These conservative results led to a better knowledge of the machines as well as of their components. This better control of the process and component limits encouraged Air Liquide to propose a solution closer to the customer needs.

Moreover, power consumption becoming a major topical concern, Air Liquide decided at the same time to analyse the efficiencies of the Helial range. Each machine can be characterised by a specific consumption, which is defined as the ratio between the power dissipated to ambient at warm end and the power absorbed at the cold temperature. TABLE 2 shows that the specific consumption of each Helial corresponds, for a liquefaction mode without liquid nitrogen (LN) pre-cooling, to about 8% of the percentage of the Carnot efficiency. For machines characterised by an equivalent cold power between 150 W and 900 W, the percentage of the Carnot efficiency should be between 10 and 15% (FIGURE 2).

	Helial 1000	Helial 2000	Helial 3000
Equivalent power for liquefaction w/o LN2	137 W	290 W	614 W
Specific consumption for liquefaction w/o LN2	967 W/W	862 W/W	732 W/W
% Carnot	7%	8%	9%
Specific consumption for refrigeration w/o LN2	1015 W/W	602 W/W	600 W/W
% Carnot	6%	11%	11%

Table 2. Helial 1000/2000/3000 efficiencies

This differences in performance can be explained by the multi-mode optimised design of these machines, which was determined by the strategy decided upon by Air Liquide seven years ago. This choice permitted to propose a single multi-purpose product in order to cover a large market of applications, but had the drawback of not being optimised for pure modes. The results of this capitalisation led Air Liquide to decide a new strategy characterised by dedicated machines, i.e. refrigerators and liquefiers, in order to improve the specific consumptions in a context of energy consumption reduction.

The process was then reviewed in order to dedicate and optimise a standard machine to one pure operation mode, taking into account the acquired experience on components limits and machines behaviour. This led to the upgrade of the Helial range.

2. HELIAL EVOLUTION RANGE

2.1 Process Optimisation

Helial machines can operate in a liquefaction mode or in a refrigeration mode. For both modes, the basic cycle is the same, remaining a Claude cycle with two turbines installed in series. Nevertheless, the components of the cold box do not have the same contribution depending on the operating modes. The process studies performed were based on the differences between refrigeration and liquefaction modes.

Figure 2. Efficiency for refrigerators and liquefiers compared to Carnot efficiency (Strobridge [1])

In refrigeration mode (Figure 3 – left side), the customer’s application is maintained at a constant temperature with a liquid helium bath, where evaporation absorbs the power injected into helium. Some gas is generated and recovered in the cold box heat exchangers. The boil-off participates to the cool-down of the high-pressure (HP) gas to be liquefied by counter-flow exchange. In this mode, the power extracted by the turbines compensates only for the heat losses of the system, (i.e. non-reversibility of heat exchangers), heat-in leaks to internal components of the cold box and the heat load transferred by the customer’s application to helium. As shown by the left picture, the refrigerator is aimed at removing 18 J/g from helium gas (which corresponds to the latent heat) to condensate again liquid helium in order to maintain the level.

Figure 3 : Differences between refrigeration mode (left side) and liquefaction mode (right side).

In liquefaction mode (Figure 3 – right side), no cold gas is returned through the heat exchangers except for the flash generated by the Joule-Thomson expansion. Therefore the turbines must balance not only the heat losses of the system but also must extract the power to cool down the helium to be liquefied. The liquefier must remove about 1540 J/g in order to cool the gas down to the liquefaction point, and then again 18 J/g in order to liquefy helium.

In that way, one can easily understand that in a liquefaction mode, turbines have to extract more power than in a refrigeration mode, whereas in a refrigeration mode, the heat exchangers must provide the largest surface, particularly for the last heat exchanger, so as to recover the maximum cold enthalpy of the cold gas generated by the application. When the Helial was designed for a mixed mode, the efficiency in case of a pure liquefaction mode or a pure refrigeration mode would not have been the maximum since turbines and heat exchangers are operated in off-design modes.

Within the diversity of applications, the requirements for liquefaction and refrigeration modes are very different in terms of process specificities, scope of supply, control system, and system operating approach. The recent evolution of the Air Liquide Helial range relies on the choice of the design mode in order to propose a standard system optimised for the desired operation mode.

The range is therefore split into two series: liquefiers and refrigerators. This separation permits one to obtain the maximum specific consumption with a given compressor for the operation mode chosen by the customer. The new Helial Evolution range is composed of three sizes of cold boxes : small, medium and large. For each size, the customer’s specification leads to the design mode for which the performance will be optimised: liquefaction or refrigeration. The new range now consists of three liquefiers: Helial SL, Helial ML and Helial LL; and three refrigerators: Helial SF, Helial MF and Helial LF. Even if the external parts of the cold boxes of Helial refrigerators and liquefiers of the same size are identical, the heat exchangers and the turbines will be different being optimised for the chosen range.

2.2 Turbines Improvements

Some studies have also been performed regarding the influence of turbines efficiencies. We estimated the impact of increasing efficiency on the liquefaction rate. The results of the calculation are given in Table 3. This information pushed Air Liquide to work on the efficiency of the turbines, trying to improve these figures on their smallest machines that only extract several hundreds of watts.

This has been managed thanks to the use of 3D-open wheels as it was already done on larger turbines. About ten efficiency points have been gained and demonstrated on our specific test bench in DTA, which permits testing the turbines under real cold conditions. These good results, in addition to the optimisation resulting from the choice of operating mode, also contributed to the improvement of the performances (as shown in Table 3). Additional studies are still in progress with the goal of gaining further increases in efficiency.

	Turbine 1	Turbine 2
Effect of one additional point upon the liquefaction rate (%)	+ 0.9	+ 1.25

Table 3. Influence of turbines efficiencies on liquefaction rate

2.3 New Range of Performances

The liquefier optimisation performance results are summarised in Table 4.

As pointed out in the table, the specific consumptions of the new liquefiers when compared to the former range are much lower. The specific consumption for liquefaction without LN pre-cooling as a percentage of Carnot is now around 12%, which constitutes a good position on the curve of Figure 2 and particularly considering the small size of these machines. In terms of performances, the Helial ML anticipates a doubled capacity with the same cycle compressor and size than the Helial 1000. The Helial LL offers 78% more capacity than the Helial 2000 when both use the same compressor, and a small adapted machine for low capacity requirements, the Helial SL, is born.

This improvement means lower operation costs thanks to a better adaptation to the customer’s needs, process optimisation and better use of the main components.

	HELIAL SL	HELIAL ML	HELIAL LL
Max. Liquefaction capacity without LN2	25 L/h	70 L/h	145 L/h
Max. Liquefaction capacity with LN2	50 L/h	150 L/h	330 L/h
Compressor electrical motor	55 kW	132 kW	250 kW
Specific consumption for liquefaction w/o LN2	645 W/W	552 W/W	505 W/W
% Carnot	10%	12%	13%

Table 4. HELIAL Evolution liquefiers performances

3. BENEFITS AND ADAPTATIONS OF THE STANDARD PRODUCT

The standard Helial units are dedicated to be used for a maximum range of applications, from liquefaction centers to refrigeration applications. Hence, this product must fulfil a lot of different requirements and should remain flexible.

3.1 Benefits of Standard Helial Range

First, this product must remain easy to operate, since the number of operators dedicated to cryogenics has been decreasing in most centers. The Helial is then totally controlled by a dedicated performing control system, in order to adapt automatically operations to the load variations of the system. It is equipped with suitable sensors and automatic valves which makes it easy to operate. The control system can also be exported through a supervisor with friendly interfaces. This supervisor can offer different levels of control from simple monitoring of parameters up to a total remote control with orders given through the supervising system. Diagnosis and permanent check with recording is then made easier. Air Liquide DTA also proposes a remote access to the PLC through a modem so that an office expert can give direct assistance to each customer.

Helial product also constitutes a high reliability solution with reduced maintenance, resting on the use of very robust components such as oil lubricated screw compressors, aluminium plate fin heat exchangers and Air Liquide own static gas bearings turbines, characterised by a calculated MTBF higher than 150,000 hours. These turbines are designed, manufactured and tested in real conditions in the Air Liquide DTA workshops, which allows control of the whole process, and qualification of the turbines before delivery. This secures the delivery schedule and the performances of the machines. The reduced maintenance enables one to propose a product adapted for a long and continuous operation time without shut-down, with a control system providing permanent monitoring of all the parameters, permitting auto-diagnostic and safe-guarding the system from unanticipated stops. Thanks to its worldwide experience, Air Liquide developed some partnerships with subcontractors and suppliers, and can propose a product which complies with all the regulations and norms.

This product is also designed in order to limit the operation costs. The new optimised Helial Evolution range led to lower power and utility consumptions due to improved system efficiencies. With the automated control system, a permanent operations teams is not necessary, which also leads to the operation cost reduction.

Finally, Air Liquide can propose a turn-key system in order to take responsibility for the entire cryogenic project. This includes the proposal of complete solutions including on-site services such as the installation work, and maintenance and operation contracts which have been developed. The delivery time has also been reduced by modifying the manufacturing process so as to comply with customer requests to shorten the project construction phase.

CONCLUSION

The Helial Evolution constitutes a standard product line providing high performances with high reliability and efficiency. Performances have been considerably increased for the new Helial Evolution range. This product is the answer to a multi-range markets and remains adaptable to specific requirements.

REFERENCES

Boissin, J.C., Gistau, G., Hébral, B., Pelloux-Gervais, P., Ravex, A. and Seyfert, P., “Cryogénie : Mise en oeuvre des basses températures,” in Techniques de l’Ingénieur, B2 382, pp. 2-4.

CR08-40

Liquid helium in laboratory use – practical remarks

Haberstroh Ch.

Lehrstuhl fuer Kaelte- und Kryotechnik, TU Dresden, Germany

Abstract

Within the last decades an uninterrupted increase in the use of liquid helium for laboratory applications was recorded. On the other hand increasing problems and excessive costs caused by incorrect use or lack of knowledge about helium specifics are reported.

In this contribution a number of typical aspects are addressed, like commercial supply of liquid helium vs. operation of an own liquefier, the use of helium dewar vessels and balancing of helium amounts, usual causes for leakage and for contamination as well as special features of helium recovery systems.

Introduction

Liquid Helium (LHe) is usually used for cooling tasks below liquid nitrogen temperature level, mostly in the range between 2 and 5 K. The LHe demand especially in Germany can be characterized as follows:

- There is an increasing number of universities as well as national and semi-national institutes with current utilization of LHe. Several of them had been founded within the last decades.

- The specific LHe consumption at the single points of use and thus the overall consumption of the whole institutes has raised appreciably, compared e.g. with the situation in the 70th or 80th.

Both aspects can be attributed – at least partially – to changes in the cryo technology within the last decades: in the 70th or 80th typically small table-top flow cryostats or bath cryostats with a capacity of a few liters only had been in use. Often these had been self-developed and self-fabricated devices. For experiments these were cooled-down and used for some hours at dedicated days only.

Meanwhile a wide range of bath, flow and magnet cryostats are commercially well available and are often delivered as a turn-key system within a complex experimental set-up. This does first broaden the subgroup of potential users: whereas in former times experiments at LHe temperatures had been subject for dedicated cryo specialists only, today as well chemists, material scientists or physicists find themselves among the LHe users, mostly not involved in cryogenics else wise. Often the attention paid to the LHe cooling is similar low as to cooling water of similar supplies, and the LHe handling is done according to the manual of the system supplier, without further knowledge of any LHe specifics.

Secondly, the dimensions of those commercial cryostats are appreciably bigger and more complex compared with earlier self-made devices. Improvements in thermal insulation are outbalanced by that by far. Moreover most of the commercial superconducting magnet, NMR and ESR systems are customarily or necessarily kept at LHe temperatures continuously. Thus a permanent LHe demand is generated all over the year, independent of the periods of use.

1. commercial LHe supply

As for Germany, meanwhile an excellent infrastructure can be stated concerning the supply of LHe. LHe can be ordered from several gas companies as a standard product and is delivered at short notice to almost all places. Transport dewar vessels with a capacity of 50 l – mostly 100 l – and up to 500 l are used. Prices are found between 6 and 12 € per liter LHe, including rental dewar vessel and transport fees. The helium gas released at the user is either blown to the atmosphere or collected and compressed to high pressure cylinders, in case a helium recovery system is installed in the respective institute.

For the institute administration such a commercial supply can be very attractive:

1. onset of LHe utilization almost instantaneously,

2. no or only minor invest cost in advance,

3. no additional space or staff requirements.

Moreover a LHe provision from an external supplier fits well to the common trend of outsourcing, what stands for reduction of all activities other than the institute’s own dedication.

Thus a LHe supply on commercial terms is practiced by many institutes, sometimes with remarkable increase of volume upon successful implementation. Sometimes up to 100 000 l LHe per year are obtained, i.e. a turn-over of 20 transport dewars per week and costs beyond ½ M€ per year.

Two aspects have led to a more critical estimation of a solely commercially-bases supply: on the one hand delivery bottlenecks had to be registered in the past, concerning all gas companies and the entire European market for periods of several weeks each time in the last years - obvious with increasing frequency. E.g. in the summer 2000 as well as in the autumn 2001, 2005 and 2006 or at the end of 2007 only parts of the necessary LHe quantities could be delivered, partly had test series to be abandoned and MRI devices allowed to warm-up.

Secondly: in institutes with large and regular LHe consumption there is generally a helium recovery plant installed at least. Re-compressed helium gas at typically 200 bar is returned to the gas company against an appropriate credit note. Thus the effective purchase price is reduced significantly. The gas companies show up, however, increasingly less interest in returned helium. The acceptance is rejected generally or only a poor refund of 1 ... 2 €/m³ is granted. Apart from high logistic expenditure the principal reason lies in the fact that there is nearly no need for further gaseous helium (besides easily contaminated).

Both has its cause in the same point: Helium is won exclusively as by-product from the natural gas extraction. Appropriate plants for this are installed at He-rich deposits, which are predominantly situated overseas in the USA, in North Africa and in the Middle East. For transport the helium is locally liquefied and transported in locked 11,000 gal overseas containers to Western Europe. Due to the heat input during the transport the helium is found in supercritical state at approx. 4 ... 5 bar_abs. finally. At the destination therefore only about half of the contents can be filled up in liquid form and delivered to the LHe customers. The remaining helium is merchandised as high-pressure helium gas, usually just sufficient to meet the complete gaseous helium demand. Therefore there is normally no reasonable use for further helium gas coming from the customer, moreover for a more or less impure one.

2. Helium Balancing

With the balance of helium conversion values apply in accordance with table 1. Special attention should be paid to the high density of cold helium gas at 4.2 K or at slightly elevated temperatures: in good approximation 10 % of the liquid density can be assumed. Therefore still a tenth part of the LHe capacity is contained in an “empty", but cold helium dewar.

Table 1: Conversion numbers for helium [1]

liquid @ 4.2 K / 1013 mbar	1 l
gaseous @ 4.2 K / 1013 mbar	7.2 l
gaseous @ 0°C / 1013 mbar	702 l
gaseous @ 15°C / 1 bar	751 l

The gaseous state at 0°C /1013.25 mbar is often denoted as standard temperature and pressure (STP) or T_np_n, in physics or science one mostly refers to that. At the gas industry against it the so-called reference state (“Bezugszustand”) at 15°C and 1 bar is used exclusively [1]. This leads in practice frequently to wrong conversions and misunderstandings.

For the use of commercially supplied LHe the scenario in an unfavorable case can be found as follows:

Table 2: Balance example of poor utilization of a LHe dewar vessel

	dewar filling level
order	100 l
filling ex factory and account	96 l
after transport	94 l
after 14 days stand-by	80 l
after cryostat refill	20 l
liquid return	20 l

In view of organizational and process uncertainties usually the LHe is ordered with some buffer time. With a typical boil-off rate of approximately 1 l LHe/day the level already sank first around 14 l. Secondly, against the wide-spread judgment, in this example only about 54 l instead of 60 l were transferred to the LHe cryostat. Approximately 10 % of the quantity must be vaporized and remain in the transport dewar to replenish the void volume. Remaining liquid is generally not credited with commercial supply. Actually in the example shown above thus the real price per l LHe has nearly doubled. Against it, in an optimized utilization down to the necessary 2 ... 3 l remaining liquid content and with minimized evaporation losses, a substantially better balance can be effected. It should not be forgotten however that a discrepancy of at least 20 % compared to the account balance still remains.

3. LHe supply by self-operated liquefier

In Germany actually nearly 50 helium liquefiers are in use, several of them taken into operation within the last years only. Liquefiers are operated mostly by universities or e.g. by institutes of the Max Planck or Leibniz society. Those plants generally comprise the following components:

§ cold box (liquefier) and cycle compressor

§ oil removal; line drier

§ gas management and gas analysis

§ stationary LHe vessel and dewar filling station.

For the recovery system in addition is needed:

§ feed-in ports and piping

§ a helium balloon (10 … 40 m³)

§ a high pressure compressor for 200 … 300 bar

§ a high pressure storage volume at 200 … 300 bar.

Approximately 200 m² of floor space are required, partly outside of the building. Additional supply of cooling water and of LN₂ if used for pre-cooling must be taken into account. The specific electric power consumption has dropped remarkable due to efficiency improvements in the liquefier technology. 2 kWh per l LHe now can be calculated contrary to 4.5 kWh per l LHe in the past. Besides that the additional power consumption of the high pressure compressors must not be neglected: for an at least one-time re-compression of each helium batch to e.g. 200 bar approx. 0.5 kWh/l LHe must be added. The total investment sum for the components specified above is at present around 2 M€. Two operators must be engaged for operation and for the decant service. The installation of an own liquefier plant therefore is a quite fundamental decision for the respective institute.

With an own plant the costs for mere liquefaction can be beat down to 1 € / l LHe or even appreciably lower (plant depreciation, building and personnel costs usually not directly allocated to the single institutes within national institutions). The re-liquefied helium is distributed next the places of use, the helium gas released from the experiment has to be recovered as far as possible (a recovery rate of 90 … 95 % stands for good practice). Losses must be compensated by purchased helium, actually at a price level equivalent to 5 … 10 € / l LHe (regardless whether in liquid state of as high pressure gas). Beyond that, for spring 2008 a further price increase of more than 40% was announced by the gas companies. Thus the operation of an own liquefaction plant makes only sense if a high recovery rate is sustainable. In case of losses up to 1/3 of the helium amount per cycle, as observed particularly at new installations sometimes, the price advantage in relation to a commercial supply shrinks dramatically or turns even into the opposite.

4. Causes for helium losses

Often the causes for helium losses in spite of an existing recovery system are unclear to the responsible persons. In most cases quite trivial causes can be identified by on closer inspection:

- transport dewars not connected to the recovery system

- bath cryostats cut off from the recovery system during refill (for insufficient dimensioned lines or instruments)

- LHe dewars are opened despite elevated pressure inside (due to its high density the released gas volume equals to an appreciable helium amount)

- excessive cool-down of the transfer lines before insertion into the cryostat

- forgotten ball valves or helium paths at the cryostat warmed-up and opened after completed series of measurement

- loose line fittings (e.g. widened ends of PVC-hoses)

The helium recovery systems are usually held on some mbar positive pressure. By latter point often unnoticed, slight leakages are caused, which can accumulate however to substantial losses.

5. Helium contamination

In certain way a reverse problem is additional gas – i.e. impurities – in the recovered helium. Mostly this concerns ambient air (perhaps in somewhat changed composition), in some cases also pure nitrogen (from pre-cooling). Practically all liquefiers therefore are equipped with freeze-out purifiers, which can remove such impurities within certain limits. Tolerable are generally concentrations up to some tenths percent by volume. Impurity levels exceeding 2 ... 3 vol-% become problematic, since from this level the purifiers normally are overtaxed.

As causes for infiltration of impurities are found mostly:

- Faulty operation of the cryogenic equipment (often from unawareness, often in case of acquired turn-key equipment or insufficiently understood cryostat technology, often with insufficiently instructed experimenters with different specialization)

- Leakages at pumped flow cryostats (despite the quiescent evaporation develops in the LHe supply dewar with longer cryostat runs a negative pressure; smallest leakages than result in sniffing in ambient air. For security the LHe dewars in such cases are to be kept pressure-free over a compensating line.

- Diffusion through hoses and balloon material (cf. 6.1)

- Local negative pressure within the recuperation system (cf. 6.2).

5.1 Helium balloon diffusion problems

Usually in today’s recovery plants a balloon is installed as temporary buffer storage before re-compression. Mostly these are ball- or zeppelin-shaped with a capacity of 10 m³ to 30 m³. As jacket material chloroprene rubber (CR, Neoprene) coated nettings are used (thickness d ≈ 0.9 mm, 1.1 kg/m²), newly as well PU-coated nettings (d ≈ 0.15 … 0.35 mm). All these materials have a more or less pronounced permeation for the different kinds of gas. Generally each partial pressure for its own tries to equalize. So helium would diffuse to the outside, the different atmospheric gases into the balloon. For a Neoprene coat this way helium losses of about 1.1 l gaseous (at 15°C/1 bar) per m² and per 24 hours are found. After all this sums up to about 16 m³ per year for a typical 20 m³ balloon with a total surface of ~ 39 m².

With a PU-foil material these losses are reduced to about 0.4 l He gaseous / m² × 24 hours. Meanwhile as well PU based material with additional diffusion barriers is available (nano platelets, 100 x 100 x 5 nm in size). Helium losses are quoted here with 0.03 l He gaseous / m² × 24 hours.

More problematic usually is the incorporation of atmospheric molecules into the helium balloon. Unfortunately only vague diffusion parameters are known here. Generally these should be lower by a factor 40 approx. for N₂, O₂ or CO₂. Against it for the polar H₂O-molecule comparably high diffusion numbers are found.

5.2 Sub-atmospheric pressure in the recovery system

With a certain configuration of a helium recovery system it can come to a negative pressure at certain feed points. This illustrates Fig. 1. As shown here, the recovery system extends over a laboratory building with several floors. The helium balloon is installed here – according to general practice – in the attic. The helium inside is held at ambient pressure (1000 mbar supposed). The pressure level of the atmosphere for different height computes itself from the barometric scale factor. Close to ground level in good approximation a linear change of 0.12 mbar/m can be set for air (15°C/1 bar; ρ ≈ 1.21 kg/m³). For helium at 15°C/1 bar (ρ ≈ 0,167 kg/m³) in the same way a pressure change of only 0.017 mbar/m can be calculated. This results, as indicated in Fig. 1, in an air pressure change of approximately 0.5 mbar from floor to floor. Within the helium system, however, a pressure change of only 0.1 mbar per floor is effected. In result according to the example given in Fig. 1 in the floors below the attic (without the check valve already drawn in) a negative pressure in the helium lines between 0.4 mbar and 1.6 mbar is found.

This represents a perfectly unacceptable condition:

- Small leakages or forgotten valves at the inlet ports lead immediately to immense impurities (with positive pressure naturally on the other hand to appropriate helium losses)

- There is a severe risk to suck in ambient air into LHe dewars or cryostats connected to the recovery system. Frozen air can plug neck tubes and blow-off lines, the latter being an extremely dangerous situation.

Figure 1: Scheme of a helium recovery system in a multi-storage laboratory building with sub atmospheric pressure at the inlet ports.

The self-weight of the balloon jacket in this respect does not play any relevant role. In the numerical example for a 20 m³ balloon from heavy neoprene material a jacket mass of 22.5 kg is distributed over a cross section of ~ 11 m². The resulting positive pressure p is

p = 22.5 kg × g / 11 m² = 0.2 mbar

For a reasonable overpressure of 4 mbar an additional weight of some 450 kg would be necessary. The right solution is of course the introduction of an adequate check valve with an opening pressure of some mbar, as already indicated in Fig. 1. Nevertheless a negative pressure still remains at the suction side of the recovery compressor.

References

1. 1 x 1 der Gase - Physikalische Daten für Wissenschaft und Praxis (fluid handling and data collection for science and practice), Air Liquide GmbH, 2005 (www.airliquide.de)

CR08-31

Effect of Alternate tube Characteristics on High Capacity Pulse Tube Cryocoolers Performance

Saidi M.H., Sarikhani N., Jafarian A., Hannani S.K.

Center of Excellence in Energy Conversion, School of Mechanical Engineering,
Sharif University of Technology, Tehran, Iran

Abstract

High capacity pulse tube cryocoolers offer the promise of the cooling capacity required for operation of superconducting devices. The purpose of this paper is to investigate the influence of the alternate tube characteristics on a high capacity pulse tube cryocooler performance, intended to achieve 250 W at 80 K. In this respect the hydrodynamic and thermal behavior of the cooler is explained by applying the mass and energy balance equations to different parts of the cryocooler. Nodal analysis technique is employed to simulate the tube section behavior numerically. Employing the proposed model the effect of essential characteristics such as tube aspect ratio, frequency of oscillations, reservoir volume and double inlet is considered. To determine the optimum operation conditions that maximize the enthalpy flow and thus the coefficient of performance, net enthalpy flow analysis is conducted. The results of this analysis are tested and validated by comparing them with the experimental data.

Introduction

The use of very low temperatures for application in superconducting devices, space, military and medical equipments requires the achievement of high performance cryogenic systems. Pulse tube refrigerator as a reliable cryogenic cooler, contains no moving parts at its cold head, and thus having considerable system advantages over the most other types of cryogenic refrigerators in terms of reliability, life time and low vibration and cost. However, the older types of cryocoolers like GM and Stirling coolers are going to be replaced by pulse tube refrigerators in a wide variety of applications.

Gifford and Longsworth [1] were the first who introduced pulse tube refrigerator in 1966. The modern pulse tube cryocooler, which was equipped with an orifice and a surge volume on the warm end, was developed by Mikulin [2] in 1984. Due to this modification the performance of the pulse tube increased and became comparable to the performance of practical coolers such as Stirling cycle, Gifford-McMahon and Joule-Thomson cryocoolers. In 1990 Zhu et al. [3] added a bypass to the device and introduced the double inlet pulse tube refrigerator. The improvement of the double inlet pulse tube refrigerator was achieved by implementing a bypass between the central zones of the Pulse tube and that of the thermal regenerator. This improvement led to a new configuration named Multi Bypass Pulse Tube Refrigerator [4]. By the end of the 1990s, temperatures below 2 K were achieved with a three stage pulse tube [5]. In recent years two major models of pulse tube refrigerators are currently under development. The first style, known as G-M type, is a variant of the Gifford-McMahon cryocooler. The second type of pulse-tube refrigerators is known as Stirling type.

In this paper we present the design and optimization of a high capacity Stirling type double inlet pulse tube refrigerator (DIPTR), which is intended to produce 250 watts of refrigeration at 80 K. Special precautions have to be considered in the design and functioning optimization of the high capacity DIPTR systems. In fact, the energy consumption of the high capacity cryocoolers is significantly greater that of the medium capacity ones. Thus, the coefficient of performance that considers both the cooling power and the rate of work transfer to the gas should be optimized to ensure the highest cooling capacity and minimum work. Here, we study the influences of the cooler geometry and operating parameters of the tube section on the cooling capacity and coefficient of performance by using the developed model. Furthermore, in order to optimize the performance of the pulse tube cryocooler, the effect of orifice valve and double inlet are explored as well.

Physical model and governing equations

A double inlet Stirling type pulse tube cooler contains several components such as, a piston which generates pressure oscillations, the stainless steel regenerator which acts as a porous medium, the tube through which the working fluid flows freely, heat exchangers which make desirable thermodynamic contact with the gas, an orifice, one bypass line and a reservoir. Figure 1 shows a schematic view of a single-stage double-inlet Stirling-type pulse tube refrigerator.

Figure 1: Shematic view of double inlet pulse tube refrigerator

The operating process of the pulse tube refrigerator is complicated due to the nature of unsteady, oscillating and compressible gas flow. To trace the process, we treat it as a one dimensional, periodic, unsteady compressible flow. Following assumptions are introduced in our model as well:

1. The working fluid is an ideal gas

2. Physical properties are functions of temperature

3. The outer surface of the regenerator and tube wall are adiabatic

4. Thermal behavior of the regenerator and heat exchangers are ideal

The hydrodynamic and thermal behavior of the cryocooler is predicted by classical thermodynamic model. In this respect the mass and energy balance equations are applied to six control volumes. Temperature and pressure are assumed to be spatially averaged in the control volumes except in the tube section and regenerator. Nodal analysis technique is employed to simulate the tube section behavior numerically. For the regenerator a linear trend is acquired for the pressure and temperature distribution along the regenerator. The implicit control volume method is used to perform the spatial discretization of conservation of mass and energy in tube section. The nodal analysis is used to divide the whole tube section into a series of control volumes. The tube section is divided into (n) control volumes in longitudinal direction.

The temperature, pressure, density, mass, and gas properties of each control volume are accounted in the center of that control volume. The node is supposed to be in the center of each control volume. The second order up-wind is used to perform the spatial discretization of conservation of mass and energy in the tube section.

Complete system of differential equations for DIPTR is as follows:


	(1)


	(2)


	(3)







	(4)

	(5)


	(6)


	(7)

In the above system of equations:

The subscript (i) designates the number of control volumes in the tube section which is varying from 1 to n.

The temperature at the boundaries is shown by superscript * and is defined as follows:

	(8)
	(9)
	(10)

The hydraulic conductance of the regenerator is obtained from the Kozeny law established for porous media [6].

(11)

where, permeability K, is expressed in terms of the pressure drop coefficient and the Reynolds number [6, 7].

(12)

The value of the coefficients a, b and c for different situations can be found in [8]. We have chosen the empirical correlation for the oscillating flow which was proposed by Tanaka [9].

Heat transfer between the gas and tube wall in tube section is obtained from the Nusselt number equation under oscillating flow, in complex form as described by Kornhouser [10].

Results and discussion

The geometry and operating parameters of the cryocooler, which has been considered in the present paper, are presented in Table 1.

compressor swept volume(m³)	1.75E-4	tube wall thickness (m)	1.00 E-3
compressor dead volume(m³)	1.5E-5	hot heat exchanger volume(m³)	1.25E-5
Frequency (Hz)	50	cold heat exchanger volume(m³)	2.5E-5
average pressure (bar)	20	reservoir volume (m³)	2.25E-3
after cooler volume(m³)	2.00E-5	cold end temperature (K)	80
regenerator diameter (m)	0.1	hot end temperature (K)	300
regenerator length (m)	0.08	compressor conductance	10
tube diameter (m)	0.04	orifice conductance	1.5
tube length (m)	0.2	double inlet conductance	2

Table 1: Geometery and operating parameters

Figure 2 shows the comparison of pressures along the pulse tube refrigerator. The amplitude of pressure decreases as the flow passes through the cryocooler toward the reservoir. The pressures are comparable in amplitude with an obvious phase shift, as the above figure displays.

MATLAB Handle Graphics

Figure 2: Comparison of pressures at different locations of the cryocooler

In figure 3 the instantaneous temperature along the alternate tube section has been represented. In this figure the temperature has been plotted at different cycle times in comparison with the cycle averaged temperature along the tube section.

MATLAB Handle Graphics

Figure 3: Instantaneous temperature along the tube

Figure 4 shows the influence of the cryocooler orifice conductance on cooling capacity. The effect of reservoir volume has been reported in this figure as well.

As expected optimize values are observed on the cooling capacity of the cryocooler, which depend on the reservoir volume. Cooling capacity of the cryocooler increases almost 10% by increasing the relative reservoir volume from 9 to 18. No significant increase is observed in the cooling capacity when the relative reservoir increases more. The significant effect of orifice conductance is noticeable in this figure. The trade off between the phase shift angle, between pressure and velocity at the cold end and pressure amplitude in the tube section results is as an optimum as figure 4 displays.

MATLAB Handle Graphics

Figure 4: Cooling capacity of the cryocooler vs. orifice conductance

Figure 5 represents the variation of cooling capacity vs. frequency of oscillations at different average pressures of the cryocooler. A significant increase is observed in cooling capacity as the average pressure of the cryocooler increases from 10 to 25 bar. This figure shows the influence of the frequency of oscillations on the cryocooler capacity as well.

MATLAB Handle Graphics

Figure 5: Effect of average pressure and frequency on cyocooler capacity

Figure 6 shows the effect of double inlet valve conductance on cryocooler coefficient of performance for different magnitudes of alternate tube aspect ratios. By increasing the double inlet valve conductance cooling capacity of the cryocooler decreases due to the decrease of the mass flow rate at the cryocooler cold head. However, by increasing the double inlet conductance, the rate of work transfer to the working fluid decreases as well. Thus, an optimum is observed on cryocooler coefficient of performance due to the trade off between the cooling capacity and transferred work to the gas.

To verify the results of the present model, the cooling capacity and the transferred power to the working fluid have been compared with experiment at the same operating parameters and geometry [11]. The computational model predicts 126 W cooling capacity at 80 K cold

MATLAB Handle Graphics

Figure 6: Effect of double inlet valve and alternate tube section aspect ratio on

cryocooler coefficient of performance

end temperature with 1.8 kW net power transfer to the gas. Results show that 2.4% deviation is found between the cooling capacity, predicted by the model and that of the experiment. The experimental model reports the cooling capacity of 123 W at 80 K with compressor power of 3.1 kW. The reasons of the difference between the transferred power to the gas predicted by our model and the experiment include the compressor efficiency and lost work due to the entropy generation, which have not been considered in the present study.

Conclusion

The hydrodynamic and thermal behavior of a high capacity Stirling PTR was predicted by control volume analysis model. Nodal analysis technique was employed to simulate the tube section behavior numerically. The influences of the tube section geometry and operating parameters on the cooling capacity and coefficient of performance have been studied. In order to optimize the performance, the effect of orifice valve and double inlet has been explored as well. The cooling capacity of 250 W at 80 K with 3.9 kW net power transfer to the gas has been predicted by the computational model. Results have been compared with experiments and a good agreement was observed.

Nomenclature

area	Subscripts &Superscripts
specific heat capacity		after cooler
inertia coefficient		cold end
compressor conductance		cold heat exchanger
orifice conductance		Compressor
double inlet conductance		Gas
diameter		hot end
heat transfer coefficient		Hydraulic
Heaviside function		hot heat exchanger
thermal conductivity		node number
permeability		double inlet junction
length		Tube
mass flow rate		constant pressure
number of nodes		Wall
Nusselt number		Regenerator
pressure		Passage
Pecklet number		boundary value
radius	Greeks
gas constant		Porosity
Reynolds number		dynamic velocity
time		angular velocity
temperature		rate of work transfer
volume		element length

References

[1] Popescu G., Radcenco V., Gragalian E. and Ramany Bala P., A Critical Review of Pulse Tube Cryogenerator Research, Int. J. Refrig. Vol. 24, 2001, 230-237.

[2] Mikulin EI., Tarasov AA. and Shkrebyonok MP., Low Temperature Expansion Tubes, Adv. Cryog. Eng., 1984, 629-637

[3] Zhu SW., Wu PY. and Chen ZQ., A Single Stage Double Inlet Refrigerator Capable of Reaching 42 K, ICEC 13 Proc. Cryogenics, 1990, 257-261.

[4] Cai JH., Wang JJ., Zhu WX. and Zhou Y., Experimental Analysis of Double Inlet Principle in Pulse Tube Refrigerator, Cryogenics, 1993 522-525.

[5] Xu M. Y., A. de Waele T. A. M., and Ju Y. L., A Pulse Tube Refrigerator Below 2 K, Cryogenics 39(6), 1999, 865-869.

[6] Nika Ph., and Bailly Y., Comparison of Two Models of a Double inlet Miniature Pulse Tube Refrigerator: Part B, Electrical Analogy, Cryogenics (42), 2002, 593-603.

[7] Nika Ph., and Bailly Y., Comparison of Two Models of a Double inlet Miniature Pulse Tube Refrigerator: Part A, Thermodynamics, Cryogenics (42), 2002, 605-615.

[8] Nika Ph., Bailly Y., Jeanot J.C. and Labachelerie M. D., An Integrated Pulse Tube Refrigeration with Micro Exchangers: Design and experiment, International Journal of Thermal Science, Vol. 42, 2003, 1029-1045.

[9] Tanaka M., Yamashita I., Chisaka F., Flow and Heat Transfer Characteristics of the Stirling Engine Regenerator in an Oscillating Flow, JSME Internat. J. Ser. (2) 33 (2), 1990, 283–289.

[10] Kornhauser A.A., and Smith J.L.Jr., Application of a Complex Nusselt Number to Heat Transfer during Compression and Expansion., Journal of Heat Transfer 116, 1994, 536-542.

[11] Imura J., Shinoki S., Sato T., Iwata N., Yamamoto H., Yasohama K., Ohashi Y., Nomachi H., Okumura N., Nagaya S., Tamada T., Hirano N., Development of High Capacity Stirling Type Pulse Tube Cryocooler, Physica C 463–465, 2007, 1369–1371.

CR08-58

GREEN CRYOGENICS: THE USE OF NATURAL CONVECTION TO IMPROVE THE EFFICIENCY OF CRYOGENS AND CRYOCOOLERS

Scurlock R.G.¹, Wang C.²

¹Kryos Technology, 22 Brookvale Road, Southampton, SO17 1QP, UK.
²Cryomech. Inc., 113 Falso Drive, Syracuse, NY 13211, USA.

ABSTRACT

Cryogenics and Cryogenic Engineering are major users of energy, by their nature in providing or using refrigeration or “cold ”at low temperatures.

The magnitude of this energy is inversely proportional to the lowest temperature and is far larger than the enthalpy absorbed by the cryogen produced, or by a cryocooler/condenser.

Economy and “green “ practice is becoming increasingly important in Cryogenics to:

1. reduce energy related emissions,

2. reduce cost of making cryogens, and the cost of losses in storage and use of them,

3. reduce losses in handling and transfer.

Natural convection can be used as a major factor for reducing radiative, conductive, and convective heat in-leaks, in the design of low loss dewars, and in minimising transfer losses.

Cryocoolers are becoming increasingly important as cryogen-free replacements for cooling large industrial, laboratory and medical systems. In many cases the cryocooler is used as a condenser of boil-off vapour in, for example, LHe cooled superconductivity magnet systems.

Recent work by Chao Wang, Cryomech Inc. Syracuse, NY , USA, has revealed how the helium liquefaction performance of cryocooler/ condensers can be significantly improved.

Simple modifications to the cold head of a Cryomech PT410 pulse tube cryocooler/condenser has led to a 67% increase in helium liquefaction rate, at the same compressor power. These modifications are the first experimental results from using natural convection to enhance the precooling of the helium boil-off vapour prior to condensation. Further improvements may be possible.

Additional attention to matching cryocooler head to the dewar neck geometry may yield even further beneficial results.

INTRODUCTION

In the current thinking on climate change, it is important to note that Cryogenics is a major user of energy, in providing and using refrigeration, or “cold”, at low temperatures. Even with a Carnot efficiency of 100%, the magnitude of this energy consumption is considerable, being inversely proportional to the lowest temperature; and is far larger than the enthalpy absorbed by the cryogen produced, or by a cryocooler/condenser. We therefore need to develop a “green” attitude, and examine ways of reducing energy consumption and losses to reduce “carbon footprints”.

This paper is, perhaps, a beginning, but it does illustrate how the use of natural convection can offer some ”green” credentials leading to improved efficiency.

The starting point is the fact that, with decreasing temperature, natural convection heat transfer coefficients increase by 10 – 100 fold in the vapour near the liquid boiling point. In addition, the local temperature gradient, or stratification, in the vapour can give rise to a further enhancement of the heat transfer by a factor of 10 or more, in, for example, helium vapour at 10K. Together, these heat transfer enhancements approach 1000 times the ambient convection heat transfer coefficients, so that, for example, helium vapour at 10K can transfer heat to, or from, solid surfaces as effectively as the subcooled liquid at 4K.

These very large enhancements can be used advantageously in several ways.

This paper reports on two sequential developments, namely

(a) low loss dewars and

(b) cryocooler/condensers.

1. NATURAL CONVECTION IN THE DESIGN OF LOW LOSS DEWARS

The boil-off vapour from a liquid bath does not flow uniformly up the neck. Some 20 years work at Southampton [1] and elsewhere, has demonstrated that

(a) the vapour flow is confined to a thin boundary layer flow at the neck wall,

(b) heat flux from the wall drives the convection flow up the wall,

(c) the magnitude of this upward flow is many times larger than the boil-off mass flow and forms part of a thermosyphon recirculation,

(d) the downward flow part of the thermosyphon occurs at the centre of the neck.

Studies have shown that some, or all, heat fluxes down the neck, whether radiative, convective or conductive, can be absorbed by the cold vapour rather than by the liquid. This simple concept can be applied directly to the design of low loss dewars, containers and tanks.

2. CRYOCOOLERS AS CONDENSERS

Placing a cryocooler cold head in the neck of a dewar can achieve two bonuses at the same time.

(a) The cooling produced by the cold head, of the central vapour core of the thermosyphon flow, will increase the mass flow of the recirculation and hence the upward mass flow at the neck wall. The heat influxes down the neck will be further absorbed, thereby yielding a significant reduction in the boil-off rate.

(b) The distributed cooling of the downward flow over the cold head can be used as a simple precooling function to increase the rate of condensation. This bonus in condensation rate cannot be achieved if the cryocooler head is mounted in a vacuum with spot cooling heat exchangers attached to the dewar vapour vent line.

3. ENHANCEMENT OF CONDENSATION RATE

A number of convection features need to be considered to maximise condensation.

3.1 Central position in cryostat neck

Since the thermo-syphon flow is cylindrically symmetrical, the cryocooler head should be placed on the axis in the centre of the down flow. No part of the cold head should project into the upward boundary layer flow at the neck wall.

3.2 Use of regenerator/recuperator for distributed cooling

While the regenerator is primarily for exchanging heat cyclically between warm and cold flows, the mean temperature of the outer wall has a non-linear temperature gradient between upper and lower ends. Thus the regenerator (or recuperator) wall can be used, accidentally or deliberately, as a continuous heat exchange surface for cooling the downward boundary layer flow of condensable gas.

Theoretical analysis and experimental tests by Wang in 2001, [2] and 2005, [3] and those by Zhu et al [4] and Ravex et al [5], have indicated that significant cooling effects are available.

One limit on the cooling effect may be the plain surface area of the regenerator. It is therefore postulated that improved heat transfer may be obtained by increasing the heat transfer area with a set of low thermal conductivity vertical stainless steel fins, or high thermal conductivity horizontal copper pins or rings, perhaps 1-2 mm high, along the vertical length of the regenerator (or recuperator).

3.3 Use of pulse/expander tube for distributed cooling

The primary source of cooling is provided by expanded gas at the lower end of the pulse tube/expander tube. After each expansion, the cold gas passes back via the cold end heat exchanger into the regenerator. It is interesting to consider whether some of the “cold” in the expanded gas can also be taken from the surface of the pulse tube or expander tube by natural convection; this would further increase the usable cold available for precooling. Since the pulse tube/expander tube also has a non-linear temperature gradient between top and bottom, the cooling effect would be distributed, and again could be used to provide significant precooling prior to condensation.

Intermediate precooling from the pulse tube wall is suggested by Wang [6].

3.4 Integration with spot cooling condensing performance

Clearly, a balance must be made between the distributed cold taken from the regenerator/recuperator and pulse tube/expander tubes, and the consequent reduction in spot cooling power of the cold head. Caution is therefore needed in not modifying the tubes with too much additional heat transfer surface by adding too many vertical fins, or horizontal pins or rings.

Reducing convective heat transfer can be achieved by applying layers of MLI superinsulation to the tube surfaces. This reduction is demonstrated by the difference in liquefaction rates between Tests 2 and 1 described below.

3.5 Radiation cooling

As mentioned earlier, the cryocooler/condenser increases the mass flow in the vapour thermosyphon loop. As a result, thermally floating horizontal radiation shields, (no colder than 70K) and not thermally attached to the cryocooler head, will be adequately cooled by the vapour flow so as to absorb all 300K radiation down the cryostat neck. Alternatively, if the radiation shields are thermally attached to the bottom of the first stage, then they may also assist the precooling of the downward boundary layer flow of condensable gas at the higher temperatures, as well as absorbing 300K radiation down the neck. These alternatives will be tested later on.

3.6 Matching cryocooler and cryostat neck geometries

For maximum precooling, the overall temperature gradients in the cryocooler cold head and cryostat neck should be closely similar. It follows that the length of the cold head should match the length of the cryostat neck, or be slightly shorter.

4. PRELIMINARY EXPERIMENTAL RESULTS

So far, 4 preliminary test results have been obtained [7], using modifications to the Cryomech PT410 helium liquefier/cryocooler described in Reference [3], as follows:

Test 1. Result achieved with modifications to cold head described in Reference [3], in 60L dewar with 180mm diameter neck.

Test 2. As Test 1 but with removal of G10 sleeve and superinsulation around cold head, also condensing coil, as shown schematically in Figure 1.

Figure 1. Helium liquefaction with pulse tube cold head inside dewar neck

1. dewar neck; 2. portion of down stream flow of helium gas to be liquefied;
3. a thermosyphon loop; 4. pulse head cold head; 5. 1^st stage cooling station; 6. 2^nd stage cooling station/condenser.

Test 3. Reduction of diameter of dewar neck from 180 mm to 150 mm in new 60L dewar, and addition of horizontal copper fins to 2^nd stage pulse tube and regenerator.

Test 4. As Test 3 but with addition of horizontal radiation shields/fins at the lower end of the 1^st stage regenerator, as shown schematically in Figure 2.

Figure 2. Latest helium liquefaction system with pulse tube cryocooler.

1. helium dewar; 2. dewar neck; 3. fins on the 2^nd stage regenerator; 4. 1^st stage cooling station; 5. radiation shields/fins on the 1^st stage regenerator; 6. pulse tube cold head;
7. flexible lines; 8. fins on the 2^nd stage pulse tube; 9. 2^nd stage cooling/condenser;
10. compressor.

The performance improvements, in terms of liquefaction rates per day, are listed in Table 1.

	Test 1	Test 2	Test 3	Test 4
Liquefaction rate	12.8 L/day	15.0 L/day	19.5 L/day	21.4 L/day

Table 1: Test results

It can be seen how the modifications led to significant improvements in the liquefaction rate, from 12.8 L/day to 21.4 L/day, an increase of 67 %.

It should be added that a newly made cold head with copper fins on the 2^nd stage regenerator and pulse tube was used for Tests 3 and 4.

Further tests should enable more improvements in condensing performance to be achieved using natural convective heat transfer, towards the theoretical limit, for example, of
34 L/day for a 1W/4.2K cryocooler.

CONCLUSIONS

Simple modifications are possible to improve the low loss boil-off performance and the condensing performance of cryocoolers, using the natural convection in the neck of a cryostat.

Natural convective heat transfer coefficients are very high in dewar necks and can be used advantageously both for cooling the dewar neck and for extracting precooling from both the pulse tube and regenerator, without major modification to the cryocooler head.

Clearly, there remains a great deal of optimisation to attain the maximum condensation rate with a given cryocooler.

REFERENCES

1. Scurlock R.G., Low Loss Storage and Handling of Cryogenic Liquids: The Application of Cryogenic Fluid Dynamics. Kryos Publications, 2006.

2. Wang C., Helium liquefaction with a 4K pulse tube cryocooler. Cryogenics 2001, 41, 491-496.

3. Wang C., Efficient helium recondensing using a 4K pulse tube cryocooler. Cryogenics 2005, 45, 719-724.

4. Zhu S.W., et al., 4K pulse tube refrigerator and excess cooling power. Adv. Cryo. Eng. 2002, 47A, 633-640.

5. Ravex A., Trollier T., Tanchon J., and Prouvé T., Increase in Performance of 4K pulse tube cryocooler. Proc.CryoPrague 2006/ICEC 21, 2006, 515-524.

6. Wang C., Intermediate cooling from regenerator and pulse tube in a 4K pulse tube cryocooler. Cryogenics, 2008, 48, to be published.

7. Wang C., and Scurlock R.G., Improvement in performance of cryocoolers as condensers. Cryogenics, 2008, 48, accepted for publication.

CR08-12

The Development of a Vuilleumier Cryocooler for New Zealand’s High Temperature Superconductor Industry

Gschwendtner M.A., Tucker A.S.

University of Canterbury, New Zealand

Abstract

A Vuilleumier cryocooler with a projected cooling capacity of 100 W at 77 K has been developed at the University of Canterbury in New Zealand. While most cryocoolers use mechanical compressors with associated vibration, leakage and lubrication problems, the Vuilleumier concept employs a thermal compressor with only a marginal mechanical work input requirement. This paper highlights its significant design advantages and presents results from a computer model that justifies this approach, despite the apparent disadvantage of the Carnot limitation arising twice.

Introduction

New Zealand has invested in high temperature superconductor technology (HTS) and is now in a significant position globally. Various local companies have already spun off this technology such as HTS-110 with laboratory magnets and Canterbury TX with transformers. It is anticipated that more companies will follow. The sheer number of possible applications of HTS technology necessitates the development of affordable, reliable and mobile cooling systems. This project originated from the need to provide an already existing and fast growing industry with a locally developed cooling system that can be specifically adapted and produced at lower cost.

1. background

1.1 Theoretical background

Any cold-producing devices such as Gifford-McMahon, Stirling or pulse-tube cryocoolers require a variation in the pressure of the working gas in order to achieve expansion and compression. This is usually done by a mechanical compressor in the form of a piston that reciprocates in a cylinder. However, a reliable seal design is required in order to avoid gas leakage past the piston and, unless precise and expensive clearance seals are employed, lubrication is involved. Lubrication, on the other hand, does not go well with machines that incorporate a regenerator such as Stirling and pulse-tube refrigerators as there is the risk of contaminating the fine pores of the regenerator matrix with the lubricant and thus significantly reduce the performance of those machines. Vibration and noise are further potential problems of mechanical compressors [1, 2, 3].

An even greater issue arises from the current demand of higher cooling capacities. The scaling-up of existing equipment to meet the required higher cooling capacities by simply increasing the physical size of these machines is often very difficult and typically disproportionately more expensive. One of the limiting factors in this exercise is the mechanical compressor that is usually driven by linear motors. It is common knowledge that the cost of linear motors increase disproportionately with their size and they become exorbitantly expensive. In addition to that, large mechanical compressors tend to be very noisy and heavy – quite often a decisive criterion in some cryocooler applications. A higher moving mass of a larger mechanical compressor also leads to more severe balancing issues.

A promising alternative is to use a thermal compressor that creates a pressure variation by a change in the temperature of the working gas instead of a change in volume as in a reciprocating piston machine. This principle is based on Rudolph Vuilleumier’s invention from 1917 [4] and basically consists of a Stirling heat engine that is coupled to a Stirling refrigerator while sharing the same working gas. While it can be argued that a considerable disadvantage of using a heat engine as a compressor lies in the fact that the Carnot penalty has to be paid twice, significant design advantages may still justify the implementation of the Vuilleumier concept. This is the case even more so if a design is chosen that promises to be as nearly as efficient as a mechanically driven as will be shown below.

1.2 Working principle

A Vuilleumier refrigerator operates between three temperature levels – hot, ambient and cold. Its working principle can be best understood by first looking at how a mechanical compressor-type Stirling refrigerator works. Figure 1 shows a gamma-type Stirling configuration.

Figure 1: Gamma-type Stirling refrigerator.

A power piston reciprocates in a cylinder and creates a variation in pressure by changing the volume of the working gas. Separate, but connected to the compression and expansion cylinder by a port, is the refrigerator part. Here a displacer shifts the gas back and forth between two heat exchangers – a heat rejector at ambient temperature and the absorber at the cold end temperature. The motion of the displacer is out of phase by 90° with the power piston, such that the gas in the refrigerator is exposed to the cold end during the expansion phase and is exposed to the heat rejector during the compression phase. In between the two heat exchangers the regenerator is located, serving two main purposes: Firstly, it acts as a thermal barrier between the two temperature levels in order to avoid short-circuiting of thermal energy. Secondly, due to the porous structure of its matrix through which the working gas oscillates, the regenerator acts as a temporary heat store for the gas which picks up heat when moving away from the cold end and rejecting heat when moving away from the heat rejector heat exchanger. A Vuilleumier refrigerator has a similar working principle; however, the mechanical compressor on the right hand side is replaced by a heat engine while the remaining components in the refrigerator are identical (Figure 2).

Figure 2: Vuilleumier refrigerator.

Here, the pressure variations are created by a change in temperature of the working gas. An additional displacer moves the gas between two heat exchangers and alternately exposes most of the gas to a hot temperature and to a cold temperature. Since the overall volume of the system remains constant at all times the gas pressure varies with the temperature according to the Equation of State. Again, compressor and refrigerator parts are connected via a port. Figure 3 shows a comparison between the Carnot Coefficients of Performance (COP) for a Vuilleumier machine and a Stirling refrigerator at given cold end temperatures. Ambient temperature is assumed to be 300 K whereas the hot temperature of the heat engine in the Vuilleumier configuration is 1000 K.

Figure 3: Comparison of the Carnot-Coefficient of Performance (COP) for a Vuilleumier and a Stirling refrigerator with three different mechanical efficiencies.

The diagram illustrates two things: the Carnot-COP of a Stirling is superior to a Vuilleumier where the Carnot efficiency of the thermal compressor has to be included as well, but the situation looks different if the efficiency of the Stirling’s mechanical compressor is also taken into account. The diagram shows two efficiencies (75% and 65%) that straddle the COP-curve of the Vuilleumier and the assumed temperatures. The point to note is that mechanical compressors also have a limited efficiency and, as long as reasonably high temperatures are realised in the thermal compressor of a Vuilleumier machine, the COPs of both configurations are in the same range. However, there are significant design advantages for the Vuilleumier configuration as will be shown below.

2. Design considerations

The fact that the overall volume of the working gas remains constant in a Vuilleumier machine means that sliding seals which have to maintain a large pressure differential between the inside and the outside of the system can be abandoned. Instead, the gas envelope can be hermetically sealed by either static seals or, if practical, even be a fully welded enclosure.

A further advantage is the fact that no pistons are required to operate between a large pressure difference (and small ΔT). Instead it is possible to use displacers which do not have to compress the gas, rather, move it from one side to the other, only having to overcome gas friction (small Δp, large ΔT). This comparatively small work requirement means that smaller and cheaper linear motors are sufficient to drive the displacers. Since displacers have only a small pressure difference across them, seal problems are much less severe and the manufacturing tolerances are less tight which reduces the cost to a great extent.

The absence of problematic sliding seals and the ease with which the whole system can be sealed results in another advantage for thermal compressors: the average pressure of the working gas can be increased without the requirement for stronger mechanical compressors. Since the cooling performance is an almost linear function of the average system pressure, the machines can either be built smaller or, alternatively, the cycle frequency can be reduced. This last possibility is beneficial for the efficiency of the cycle as more time is available for heat transfer processes and fluid friction losses can be reduced.

Lubrication with all its associated issues can be avoided by the use of flexure bearings which are common practice in today’s cryocooler technology. Flexure bearings are circular discs made of steel with slots in the form of a spiral. This allows the discs to flex normal to their faces while, at the same time, providing a high stiffness radially. A shaft that is supported by stacks of these discs is more or less centrally located and constrained, but is able to reciprocate along its axis. If the flexure bearings are designed such that occurring stresses remain well below the fatigue limit these components are not only maintenance-free but also exhibit a long lifetime.

Furthermore, the choice of flexure bearings and the need for displacers only instead of pistons facilitates the use of clearance seals. These are basically tight-fit displacers in cylinder liners that provide a gas seal through a very small clearance. Since displacers have to seal against a relatively small pressure difference across them, the machining tolerances are not as tight as in the case of pistons operating between a large pressure difference. This way of sealing makes redundant the use of sliding seals that are prone to wear and also eliminates the risk of seal material debris floating around in the gas cycle. Again, this contact-free gas seal in conjunction with lubricant-free flexure bearings are a very good match that promises reliability and longevity.

Using heat as the main energy input allows a high energy transfer rate per unit volume and per unit mass which is of particular importance for the demand of higher cooling capacities. Heat can be applied to the system without any noise or vibrations and the fact that different heat sources can be used depending on the requirements or the environment of the application may be an additional significant benefit. While, at first sight, it may seem irresponsible from a thermodynamic viewpoint to use electric heating, the situation looks different if the efficiency of mechanical compressors is taken into account (see diagram in Figure 3). Electric heating is very convenient and readily available and, after all, mechanical compressors also use electricity.

Thanks to recent design developments such as flexure bearings and linear motors, the Vuilleumier concept is now more attractive than it may have been several years ago. Also, advances in materials have pushed the boundaries in the use of light-weight components and high temperature insulation materials. Finally, and not least, a new territory has been opened up by digital electronic controls of linear motors. Instead of being constrained by a kinematic drive system, the position-versus-time relationships for the displacers can now be accommodated such that their motion follows more naturally the gas cycle in Stirling machines.

3. Modelling

3.1 Background

For modelling of the gas cycle in a Vuilleumier cryocooler it is not necessary to represent the above mentioned design aspects in great detail. For the analysis and optimisation David Gedeon’s Stirling simulation software package SAGE was used [5]. This allows the user to plug various modules together and connect them through relevant variables. For instance, the face of a piston is connected to the adjacent gas space through a volume variable. Or, a heat exchanger is linked to a connecting tube with a mass flow variable, whereas the heat exchanger wall is in thermal contact with a heat sink or source through a heat flow variable. All fundamental equations are solved for each module separately until the program converges to a final solution. SAGE has a mapping function which allows the specification of various parameters that are to be varied in given increments, as well as an optimisation function. All modules/components in SAGE may have two or more sub-levels where further input variables can be specified. For example, one sub-level of a heat exchanger could be the geometry of internal fins. Depending on how detailed the user wishes to model a system, SAGE takes heat conduction paths along solid and gaseous components into account, calculates fluid friction and pressure drops, and even shuttle and seal losses in piston/cylinder arrangements can be modelled.

Figure 4 shows how above mentioned Vuilleumier cryocooler configuration was modelled in SAGE. Both the refrigerator part in the lower half and the thermal compressor in the top half comprise the same components. Both double-acting displacers are neighboured by their adjacent gas spaces to the left and the right in the diagram, followed by connecting ports to the respective heat exchangers (upwards in the diagram). Both the refrigerator regenerator and the heat engine regenerator are sandwiched by the adjacent heat exchangers. The ‘connecting duct’ module at the far left at the bottom of the diagram connects the gas spaces of the refrigerator and the thermal compressor. Finally, the three temperature symbols at the bottom of the diagram represent the three temperature levels to which a number of heat flow variables are connected to.

Figure 4: SAGE-model of a Vuilleumier cryocooler configuration.

3.2 Procedure of analysis

From the approximately 150 input variables 25 could be indentified as most critical. In order to find an optimum configuration, however, if one assumed a variation of each variable in three increments and a computing time of 10 seconds for each solution, it would take some 250,000 years to test all combinations. The dreaded search for the elusive needle in the haystack turned out to be relatively simple, though. It was found that if each component was optimised separately , one at a time, then repeated after all components had undergone this process, the optimum performance of the system had almost converged after one loop only.

The optimisation of each component was achieved through mapping by which typically three, four or five critical parameters were varied in up to ten increments each. SAGE records the user-specified output values for each combination which can then be displayed graphically in a spreadsheet. The optimum combination is easily found by sight, although a compromise between efficiency (COP) and cooling capacity had to be sought almost all the time.

Figure 5 shows a typical result of a mapping process for the regenerator. The achieved cooling capacity of the Vuilleumier cryocooler and its percentage of Carnot efficiency are plotted over the length of the regenerator for various outer housing diameters. For design constraints the inner diameter of the concentric tubular shape of the regenerator housing was held constant. Also the wire diameter and the porosity of the regenerator matrix were fixed after a material had been chosen. The arrows indicate the increase of the regenerator’s outer diameter for both the cooling capacity plots and the Carnot-efficiency plots. Here it was relatively straight-forward to pick the smallest possible outside diameter of the regenerator since an increase in diameter had a detrimental effect on both the cooling capacity and the Carnot efficiency. The choice of the regenerator’s length, however, was less clear as an increased length resulted in an ambiguous performance. A compromise had to be found where the Carnot efficiency was reasonably high at a still-acceptable cooling performance. Quite often the selection process was influenced by design considerations and/or constraints.

Figure 5: Mapping result of regenerator analysis.

4. Analytical results

Once all components were optimised and the performance of the chosen configuration could not be increased further, a performance chart could be plotted (Figure 6). This was obtained in varying the cold space temperature in the SAGE model and calculating the respective cooling capacity and Carnot efficiency. It can be seen that the cooling capacity is an almost linear function of the cold space temperature, with a value of somewhat more than 100 W at the design point of 77 K, reaching down to around 30 K with almost zero cooling capacity.

Figure 6: Predicted performance chart of a Vuilleumier cryocooler configuration by SAGE.

The predicted Carnot efficiency also drops with a decreasing cold space temperature but in a non-linear fashion. At the design point the Carnot efficiency of 25% is almost twice as high as commercially available pulse-tube or Gifford-McMahon cryocoolers with similar cooling capacities. It should be noted, however, that this is a computer model only that is partially based on ideal assumptions despite the fact that it takes a number of ‘reality-effects’ into account such as heat conduction paths, shuttle losses and irreversible heat transfer processes.

Conclusion

In view of an increasing demand of higher cooling capacities in today’s cryocooler applications an alternative to mechanical compressor-type coolers is proposed. The use of thermal compressors as in Vuilleumier refrigerators offers a variety of design advantages which may result in quieter and more reliable operation, lower manufacturing cost and, as a computer model predicts, at least comparable efficiencies. The model predicts a cooling capacity of somewhat more than 100 W at a cold space temperature of 77 K and at a Carnot efficiency of 25%. The simulation results of the proposed Vuilleumier cryocooler configuration not only compare well with existing Stirling, Gifford-McMahon and pulse-tube cryocoolers, but also seem to justify the use of a thermal compressor as opposed to a mechanical one. While it should be noted that above presented results are only simulated and reality effects will take their toll, the outlook is very promising. The Stirling Group at Canterbury University has gained some confidence in SAGE’s predictions from earlier modelling and subsequent experimental performance testing of non-cryogenic Stirling refrigerators [6, 7]. A prototype of the described Vuilleumier cryocooler is being manufactured at the University of Canterbury according to the results of the optimisation procedure with SAGE. Extensive testing is planned for the first half of 2008 with the implementation of smaller modifications in the second half of this year. Results will be published elsewhere.

references

1. Walker, G., Cryocoolers, Part 1: Fundamentals, Plenum Press, New York and London (1983)

2. Walker, G., Cryocoolers, Part 2: Applications, Plenum Press, New York and London (1983)

3. Walker, G. and Bingham, E. R., Low-Capacity Cryogenic Refrigeration, Clarendon Press, Oxford (1994)

4. Vuilleumier, R., Method and Apparatus for Inducing Heat Changes, U.S. patent 1,275,507 (1918)

5. Gedeon Associates, 16922 South Canaan Road, Athens, OH 45701, U.S.A.

6. Haywood, D., Investigation of Stirling-type Heat-pump and Refrigerator Systems Using Air as the Refrigerant, Doctoral Thesis, University of Canterbury, Christchurch, New Zealand (2004)

7. Haywood, D., Raine, J. K., Gschwendtner M. A., Investigation of seal performance in a 4-a double-acting Stirling cycle heat-pump/refrigerator, Proceedings of the 10th International Stirling Engine Conference, Osnabrück, Germany (2001)

acknowledgment

The New Zealand Foundation for Research, Science and Technology funded this project and contributed to its success in a manner which was very supportive but, at the same time, quite unbureaucratic.

CR08-20

THE EFFICIENT MANAGEMENT OF LIQUID HELIUM AT SOUTH POLE STATION DURING THE AUSTRAL WINTER

Baker R.A., Sullivan P.

Raytheon Polar Services Company, 7400 South Tucson Way,
Centennial, CO 80112, USA

ABSTRACT

Liquid helium (LHe) is critical for the operation of various astrophysical projects at the South Pole. Getting large quantities of liquid helium to the South Pole is logistically very difficult and costly. In order to supply these experiments with a combined 40 liters per day of liquid helium South Pole Station has historically had to begin the austral winter viewing season with 34,000 liters. Of this amount 34% is supplied to the projects while 66% is boiled away. By capturing the boiled off gas and using pulse tube technology 65% of the liquid helium can now be delivered to the projects and only 35% is unrecoverable. Now, 34,000 liters can supply projects with more than 80 liters per day combined total. More projects that require liquid helium can be funded, or the logistical load on the South Pole can be reduced. South Pole Station has recently procured three Cryomech PT410 pulse tube cryo-refrigeration systems custom fitted to three 4,000 liter Wessington liquid helium storage dewars. These units are tuned to re-liquefy the boil-off inherent in the Wessington dewars. This technology does not come without cost. It takes power to operate the pulse tubes and at the South Pole that translates into fuel supplied by flights. The net trade-off was beneficial to the NSF, RPSC, and science projects.

KEYWORDS: South Pole, liquid helium, pulse tube, cryorefrigerator

INTRODUCTION AND HISTORY

For logistical purposes there are two seasons at Amundsen/Scott South Pole station in Antarctica, the austral summer and austral winter. The summer season begins around 1 November of a given year when the ambient temperatures warm up enough for aircraft to safely fly in and out of the station, and ends on 15 February each year when the ambient temperatures fall below the safe operating temperatures of the aircraft. The station closes to all incoming and outgoing traffic on this date. A small crew of station operations personnel and scientists remain behind and work throughout the austral winter. All logistics take place within the 3 ½ months while the station is open during the austral summer. All material, scientific equipment, food, fuel, personnel, etc. must be in place when the station closes on 15 February.

There has been a continuous scientific presence at Amundsen/Scott South Pole Station since 1957. It was originally a meteorological station but over the years has evolved into a station that supports cutting edge science projects, some of which can be performed at no other location on the planet. In 1988 liquid helium was used for the first time to cool astrophysical receivers that are looking at extremely faint energy signatures from the Cosmic Microwave Background (CMB) radiation and the origins of the universe. The low relative humidity (RH) of the atmosphere, no diurnal cycle, and the fact that a telescope can be pointed at a specific location in the celestial dome for long integration periods makes the South Pole an ideal place for astrophysical work. Since then liquid helium has been used increasingly for a variety of projects over the years, peaking in 2006 and 2007. See Figure 1.

Figure 1 LHe requirements at South Pole have increased yearly to the maximum capacity of the station. New technology is being incorporated into the design of new telescopes and cryorefrigeration is replacing cryostats.

There are logistical challenges involved with getting liquid helium to the South Pole and maintaining a supply there. Large volumes, usually in increments of 13,140 liter Gardner LHe dewars, must travel by ship or be flown approximately 10,000 miles via commercial and military aircraft from the continental USA (CONUS), through New Zealand, to the coast of Antarctica, and on down to the South Pole. This size transport dewar just fits in the cargo hold of the LC-130 Hercules military cargo airplane which flies from McMurdo Station on the coast of Antarctica to the South Pole. This logistics chain can only be accomplished during the supply window of the austral summer of a given year, because temperatures outside this time window prevent travel of any kind to and from the South Pole by these aircraft. See Figure 2

When the South Pole closes for the winter the quantity of liquid helium on station must be sufficient to provide a regular, uninterrupted supply to the astrophysical projects for 270 days. The harsh conditions of the South Pole winter present challenges that make this difficult. The air is dry (<1% RH), the elevation is high (~2835 meters), the atmospheric pressure is low (~68.3 kilopascals) and the median ambient temperature in the winter is <-65 C. A poor understanding of how these conditions affect instruments and equipment resulted in some early failures of the South Pole Cryogenics Program. There were seasons when the winter supply of liquid helium was exhausted or lost before the viewing season was over. See Figure 3.

Figure 2 The logistics chain from CONUS to South Pole.

Figure 3 Early in the cryogenics history of South pole the station would run out of LHe before the austral winter observing season was over. 3K refers to the 3,000 gallon transport dewar and W1, W2, W3 refer to the Wessington dewars.

The South Pole Astrophysical Program grew faster than the South Pole Cryogenics Program. Telescopes operated outside and liquid helium was stored in large transport dewars in the same inhospitable environment. Seals froze and failed, vacuum insulations were lost, and large quantities of LHe were rapidly evaporated. The United States Antarctic Program (USAP) and the NSF quickly learned that heated shelters for key pieces of equipment had to be built, but the planning and execution of such endeavors is a multi-year process. Ad hoc and temporary structures were built and used

until a comprehensive cryogenics facility could be designed, engineered, and constructed at the South Pole.

Resources at the South Pole are limited and finite, particularly in terms of power and fuel. Large helium liquefiers required more power than was available at the South Pole. As a result the station employed a passive approach to storing and supplying liquid helium to the astrophysical projects. It stored large volumes, enough to supply the needs of the experiments plus the natural boil-off inherent in the storage dewars, transfer losses, and some contingency for equipment failure. This is inefficient and wasteful but historically less costly for the station than employing an active method of supply, i.e., re-liquefying the boiled off helium gas.

THE ISSUE

The LHe requirements of the astrophysical projects at the South Pole grew in successive years. The South Pole had to find a way to supply as much as 60 liters per day total to the various projects over the 270 days of the austral winter. The supply plan also had to take into account the amount of helium boiled off from each of the storage dewars, transfer losses, cryostats, and a contingency amount to mitigate unforeseen problems or catastrophic failures, usually on the order of 20%. This translated to beginning the austral winter viewing season with 34,000 liters of liquid helium, the maximum capacity of South Pole Station. This left no room for further growth.

Passive storage of liquid helium outside is not a viable method for supplying cutting edge astrophysical projects with cryogens. Supplying liquid helium to the astrophysical projects at the South Pole is similar in scope and complexity to supplying liquid helium to experiments in spacecraft, but with a far smaller budget.

THE SOLUTION

A comprehensive South Pole Cryogenics Program was established and a three-pronged approach was taken. First, a working group, the Liquid Helium Working Group (LHeWG) consisting of grantees, the National Science Foundation (NSF), and Raytheon Polar Services Company (RPSC), the support contractor, was established. Considerations of budget, existing South Pole infrastructure, resource limitations, logistics, and emerging technology were made.

Already underway were plans to modernize the station in terms of size, infrastructure, logistics, and resources. The second prong was to include the construction of a facility specific for housing, tracking, and working with cryogens and equipment in a warm environment. Third, a search was initiated for technology that would reduce liquid helium waste from dewar boil-off but not bankrupt the station fuel and power supply.

The South Pole Station Modernization Project

The South Pole Station Modernization (SPSM) project was a multi-year project managed by the NSF that increased the size of the station, fuel capacity to 450,000 gallons, power output to 800 kilowatts, population, logistics, and generally upgraded the infrastructure. With these improvements it was possible to look at modern solutions to the South Pole Cryogenics Program.

The New Cryogenics Facility

SPSM made possible the construction of a new 2900 square foot, three-module cryogenics facility. All cryogenic equipment, instruments, and storage were moved out of the harsh South Pole environment into a spacious and heated structure. The station could have some assurance of the integrity of storing dewars and other cryogenic equipment during the austral winter. Seals would no longer fail due to the cold and vacuum insulations would not be compromised.

There is now room for two 13,140 liter Gardner liquid helium transport dewars and three 4,000 liter Wessington liquid helium storage dewars. There is also room for smaller dewars (250 liter and 100 liter) used to transport liquid helium from the storage facility to the telescopes. Permanent hard-pipe transfer lines were installed in the facility to make liquid helium transfers simpler and more efficient with less transfer loss. Space was allocated for maintenance, monitoring, and tracking. The new cryogenics facility also has the capacity for housing refrigeration units to reliquefy helium gas. For the first time the South Pole could explore the possibility of taking an active approach to liquid helium storage.

The first of three modules of the new South Pole cryogenics facility was brought online in February 2006. This module housed two Gardner 13,140 liter liquid helium transport dewars for the austral winter. Prior to this construction the large transport dewars wintered outside. The Wessington liquid helium storage dewars were housed in a temporary heated structure for the austral winter. In February, 2007 all three modules of the new cryogenics facility were brought online and all equipment related to cryogenics was for the first time brought inside under one roof.

Pulse Tube Cryorefrigeration

With the construction of the new cryogenics facility and the increase in overall station power it was now possible to investigate helium re-liquefaction at the South Pole. Several approaches were studied.

The facility was designed to house two 13,140 liter Gardner liquid helium transport dewars and three 4,000 liter Wessington liquid helium storage dewars. The measured static boil-off rate of the large transport dewars was 35 liters per day and the measured static boil-off rate of the Wessingtons was 15 liters per day for a combined loss of liquid helium of 115 liters per day, the biggest single user on station.

One option was to capture the boiled off helium gas into a bag and run it through a large liquefaction unit. A unit of sufficient size to liquefy this much helium gas would require in excess of 90 kilowatts of power to operate. This would be a very large burden to the overall station power budget.

Another option was to mate pulse tube cryorefrigeration units to the Wessington storage dewars. In order to match the evaporation rate of 15 liters per day (0.02 grams per second) for a Wessington dewar the cryorefrigerator would be required to provide an enthalpy change of 1563 Joules per gram from 300K to 4K and 21 Joules per gram for the latent heat of condensation.

The cooling power required to match this boil-off rate would be:

dm/dt * H_fg + dm/dt * C_p * DT (1)

0.02 * 20 + 0.02 * 5 * 300-4 = 30 Watts (2)

where:

dm/dt = 0.02 grams per second

H_fg = 20 Joules per gram

C_p = 5 Joules per gram Kelvin

This takes the temperature down to 4K and then the re-condensing would take place. Approximately 0.4 Joules per second or 30 Watts of heat must be removed from the helium gas to match the boil off rate of the Wessingtons and it must be done within the power budget of South Pole Station.

Cryomech, Inc manufactures a 4K pulse tube cryorefrigerator, the PT410 that could be custom fitted to the neck tube of a Wessington dewar, reliquefy the boiled off gas and drain the condensate back into the dewar, all at a cost of 8 kilowatts of input power. Another 4 kilowatts of power per unit would be required to cool the associated compressor for a total of 12 kilowatts per unit or 36 kilowatts for systems fitted to all three Wessingtons. This is a substantial power savings over larger all-encompassing units and is a reasonable compromise.

In February, 2005 a Cryomech PT410 cryorefrigeration unit was installed on one of the Wessington liquid helium storage dewars in the temporary cryogenics facility built in 2001, and operated throughout the austral winter as a proof of concept. The liquefaction rate of this unit successfully matched the evaporation rate of the Wessington it was fitted to.

In fact, the cooling power was slightly greater than the Wessington boil-off and gas from a second storage dewar was plumbed into the cryorefrigeration system and the total liquefaction rate was measured to be 17.5 liters per day, 2.5 liters per day more than the boil off rate of the Wessington dewar. These results were not totally unexpected since the inlet temperature of the gas into the system was less than 300K resulting in less heat removed from the vapor to be liquefied.

For the austral Winter of 2006 two more Cryomech PT 410 cryorefrigeration units were procured and mounted on the remaining two Wessington storage dewars. As indicated by Figure 4 all three units performed better than expected and some gas from the two 13,140 liter Gardner transport dewars was plumbed into the cryorefrigeration systems and that gas was re-liquefied along with the Wessington boil-off.

Figure 4 In 2007 3 cryorefrigeration units were fully operational. Helium gas from the transport dewars was plumbed into the system resulting in all 3 Wessington dewars increasing in volume over the course of the season.

The cumulative total of gas re-liquefied amounted to 52.5 liters of liquid helium per day, a total savings of product that, up until this year, had been lost to the atmosphere.

A manifold was designed, constructed, and installed that combined the boil-off of all five dewars and distributed it to the three cryorefrigeration systems so they could all liquefy helium gas at maximum capacity. Gas in excess of what the cryorefrigeration systems could liquefy was then recovered and compressed into gas cylinders for use by the station Meteorologists and other science projects. The difference for the 2006 austral winter was, in previous years a cumulative total of 115 liters of liquid helium per day was being lost as boil-off while in 2006 only 62.5 liters of LHe per day was being not being recondensed, a product savings of 54%. These savings were repeated in 2007.

Additional fuel and power savings were made in 2007 when the new cryogenics facility was modified to use the ambient temperature of the South Pole to cool the associated compressors and heat the building. A glycol cooling loop and heat exchangers were installed in the facility that removed waste heat from the compressors and distributed it through the three modules of the cryogenics facility. The mechanical chillers were taken offline and a savings of 4 kilowatts per system was realized for a total savings of 12 kilowatts. Less fuel is also used to heat the building for even more power and fuel savings.

2008 and the Future

The South Pole cryogenics infrastructure is now established. The capability to provide as much as 80 liters per day of liquid helium to astrophysical projects now exists. This is a 100% improvement over seasons without the new innovations. The program also has the capability to be flexible and to grow with developing technology. The South Pole cryogenics program as well as the astrophysical projects it supports has grown from infancy to maturity. The telescopes are now starting to incorporate pulse tube technology into their design and using cryorefrigerators instead of cryostats to keep their detectors cold. By keeping abreast of the latest technology, upgrading equipment and instruments as needed, careful monitoring, and being innovative it is likely that increased improvements will be realized among all stakeholders without a substantial increase in cost to the station.

CONCLUSION

Several factors were involved with the successful turn around of the South Pole Cryogenics program. First, all member of the Liquid Helium Working Group had a stake in the success of the program. All had expertise in areas that were relevant to the program. Conferences were attended and manufacturers were consulted to explore options and new technology.

Second, budgets were established for the construction of a comprehensive cryogenics facility. It was determined that the cost of lost scientific data far outweighed the cost of constructing the new facility and the purchase of helium liquefiers. By bringing the cryogenic equipment and instruments indoors and providing a warm space for the cryogenics technician to work the dewars, transfer lines, vacuum pumps and other equipment could be protected from the harsh South Pole environment and properly maintained.

Cryorefrigerators had been looked at in the past but their cost, complexity, and lack of reliability made them a questionable choice for use at the South Pole. Pulse tube technology is more reliable because of the lack of moving parts in the cold head and associated with that reliability is less downtime and less maintenance. They are less expensive to purchase, less expensive to operate, produce less vibration, and generally have longer lifetimes than traditional cryocoolers such as GM or Stirling cycles. They are more suited to the challenges presented to the Cryogenics Program at the South Pole.

ACKNOWLEDGMENTS

The authors would like to thank Jesse J. Alcorta at Raytheon Polar Services for his invaluable input to the South Pole Cryogenics program for over 15 years. Thanks also to Chao Wang, Peter Gifford, and Brent Zerkle at Cryomech, Inc. for the design, installation, and testing of the cryorefrigeration systems at the South Pole, Eddie Rowe at Wessington, and to the NSF for funding this venture.

REFERENCES

1. Wang, C., Efficient Helium Recondensing Using a 4K Pulse Tube Cryocooler, Cryogenics (2005) 45 719-724
2. Handbook of Chemistry and Physics, 69^th Edition, CRC Press, Inc, Boca Raton, FL (1989)

3. Van Sciver, S. W., Helium Cryogenics, Plenum Press, New York, New York (1986)

CR08-04

Cryogenic System
of the Swiss Ultra-cold neutron source

Anghel A.¹, Blau B.¹, Daum M.¹, Kirch K.¹, Grigoriev S.²

¹Paul Scherrer Institute, CH-5232 Villigen-PSI, Switzerland,
² Efremov Institute, St. Petersburg, Russia

Abstract

The ultra cold neutron source under construction at the Paul Scherrer Institute is a new facility dedicated to the production of UCN. An essential element of this source is a 5 K solid ortho-deuterium moderator. The cryogenic system of the UCN source is presented emphasizing on the thermal process, design criteria, and evaluation of heat loads in the D₂ condenser, para-ortho D₂ converter and moderator.

Introduction

Ultra cold neutrons (UCN) are neutrons with very low kinetic energies (less than ~300neV) that can be trapped in evacuated storage volumes. They are mostly used for studying fundamental properties of the neutron and its interactions. A new facility is under construction at the Paul Scherrer Institut (PSI), Switzerland, dedicated to the production of UCN and designed to produce almost two orders of magnitude larger intensities and densities of UCN than presently available [1]. The production of UCN takes place in steps starting with a several seconds long pulse of the PSI proton beam (590 MeV, 2 mA) hitting a Pb target [2] and generating about 10 spallation neutrons per proton. The neutrons are thermalized in a 3 m³ D₂O moderator tank surrounding the proton target. A cold moderator of 30 dm³ solid D₂ (Fig.1, left panel) further cools the neutron spectrum and generates UCN. Sufficiently cold neutrons can excite a phonon in the solid D₂ and, by that, loose almost all kinetic energy to become ultra cold. The reverse process of gaining energy due to collisions of UCN with phonons is suppressed by cooling the moderator at a temperature of about 5K. In order for UCN to leave the solid D₂ into a vacuum storage system and to experiments, the D₂ quality is crucial: besides the low temperature also a low hydrogen contamination (<0.1%) and a low (<1-2%) para-D₂ concentration is required [3]. The cryogenic system is not only used to cool the D₂ moderator to 5 K, but also to prepare the appropriate quality of D₂. Starting from 30m³ normal D₂ gas at 1bar in (warm) storage tanks, the D₂ is liquefied and solidified in a condenser vessel and transferred as liquid to a para-to-ortho converter (Fig.1, right panel). At a liquid temperature close to the triple point temperature of deuterium (18.7K), ortho D₂ is produced in a catalytic conversion process using OXISORB^® [4]. The ortho D₂ is then transferred to the moderator, either as liquid (~20 K) or vapor. The solid D₂ is formed in the moderator by slowly cooling from the liquid through the triple point to 5K or by directly solidifying from the gas phase, respectively. The cryogenic system is also used to operate a cryopump (5K) and the thermal radiation shield (60K) of the large UCN storage tank of the source. Therefore, a dedicated helium cryogenic system, built around an existing refrigerator, has been designed to deliver cooling at the three required temperature levels, i.e. at 5K, 20K, and 60K.

Converter

Phase-separator

Condenser

sD2 Moderator

Figure 1: The layout of the ultra-cold neutron source. The solid D₂ moderator (left panel) and the combined helium-deuterium cold-box (right panel) are shown.

1. cooling concept

1.1 System components and heat loads

The main part of the cooling system is installed in a cold box (“Cryobox”) which includes the cold helium system and the preparatory part of the D₂ system. The Cryobox, apart from valves, piping and sensors, consists mainly of three vessels provided with external (condenser, converter) or internal (phase separator) heat exchangers as shown in the right panel of Fig.1. The vessel at the top is the deuterium condenser. It is cooled down to about 5K by a 4.5K supercritical helium loop in order to condense and freeze the deuterium from the storage tanks. The reason we chose the freezing process instead of condensation is the relatively high vapor pressure of the liquid deuterium down to the triple point (194.6mbar at 18.71K). At this pressure a considerable amount of deuterium gas will remain in the 30m³ storage tanks (almost 1kg) and therefore lost for the process. At the end of the condensation process, the condenser is separated from the D₂ storage tanks by closing a valve. The solid deuterium is then melted by increasing the temperature of the condenser to about 20K. Liquid deuterium is transferred by gravity to a second vessel, the converter, already at ~20K. Here, with the help of a special catalyst material (OXISORB^®), a chromic oxide (Cr₂O₃) impregnated silica gel, the para-to-ortho conversion of deuterium takes place. Once the desired ortho-D₂ concentration ~98-99% has been reached, the user has two options: 1) ortho-D₂ crystal growth in the moderator vessel directly from the liquid phase or 2) ortho-D₂ crystal growth in the moderator vessel by resublimation.

In the first option a valve placed at the bottom of the converter is opened and liquid deuterium is transferred by gravity to the D₂ moderator vessel. The moderator vessel has already been cooled down before (i.e. during the para-ortho conversion phase) to about 20K and prepared to receive the liquid ortho-D₂. After the transfer by gravity, it is further cooled down slowly to 5K and solid deuterium is produced from the liquid phase. This cooling process is continued until all deuterium in the moderator vessel is frozen. The moderator vessel temperature is then maintained at the lowest possible temperature (~5K).

In the second option the ortho-D₂ crystal is grown by resublimation i.e. solidification from the gas phase. For this purpose, the moderator vessel is cooled down to 5K and then a valve is opened such that the saturated ortho-D₂ vapor from the converter can condense on the cold walls of the moderator vessel. The converter is kept at about 20K during this process.

A third vessel, the helium phase separator, will be placed in the Cryobox as well. It serves as an interface to the helium refrigerator. This vessel contains 5 helium heat-exchangers

Figure 2. PID of the UCN cooling system. The heat exchangers HX2-HX5 are actually placed in the phase separator in contact with the liquid helium bath.

(HX1-HX5 in Fig.2) in contact with the liquid helium bath where the warm helium coming from the process is re-cooled before being throttled in the Joule-Thomson valves and returned as cold gas to the refrigerator. The special heater EH1 in the phase separator is installed in order to simulate the full heat load and will consume the unused cooling power.

Other important components of the cryogenic system are the solid D₂ moderator vessel, the cryopump and the thermal shield of the UCN storage tank. The D₂ moderator vessel and the cryopump are located inside the UCN tank several meters below the Cryobox.

The complete Process and Instruments Diagram (PID) scheme of the cryogenic system for UCN is shown in Fig.2. The temperature, pressure, liquid level and mass-flow sensors used for the process control are indicated. A 10m long transfer line with a loss of about 0.5W/m connects the 4.5K and the 60K refrigerator ports with the Cryobox.

1.2 Solid D₂ moderator vessel

The main heat load on the solid deuterium moderator vessel occurs during the ultra cold neutron production at initially 5K when the proton beam hits the spallation target. In order to estimate the thermal load a thermal transient FEM model was built based on the moderator vessel geometry and the data for power deposition. The moderator consists of a vertical cylinder of solid-D₂ at 5K of diameter 474mm and height 157mm (V=27.7dm³, M=5.51kg). The bottom surface of the D₂ is 356.5mm above the target axis. The D₂ is surrounded by a wall of AlMg4.5 alloy with a thickness of 1.5mm on the outer and bottom side and 1mm on inner side. On the top surface a radiation boundary condition is applied and on the inner, outer and the bottom surface convective boundary conditions are

Figure 3: The time dependence of the moderator heat load and maximum (a) and minimum (b) temperature in the solid deuterium during the neutron pulse for h=100W/m²K .

applied corresponding to the cooling by 20g/s supercritical helium with an inlet temperature of 4.5K. Two scenarios were investigated: 1) a good, but conservatively low, heat transfer coefficient (h=100W/m²K) on the convective boundaries and 2) an adiabatically insulated vessel, corresponding to a poorly-cooled moderator. The power deposition in solid-D₂ was taken from [5]. Interpolated data tables for the temperature dependence of the material properties like specific heat, density, thermal conductivity of solid deuterium, AlMg4.5 alloy and supercritical helium at 3bar have been used in the calculation. The applied pulse is rectangular with a flat top of 8s and the temporal development is followed for 120s.

The time dependence of the temperature in the solid deuterium is shown in Fig.3. As can be seen the maximum temperature is below 13K i.e. the solid deuterium does not melt during the pulse. The total heat load on the refrigerator has a peak of 250W at the end of the pulse. The average load is 80W, well below the available cooling power of the refrigerator, 300W.

1.3 The condenser

The purpose of the condenser is to pump/condense as much gaseous D₂ from the storage tanks as possible. Due to the rather large volume of the storage tanks (30m³, ~5kg nD₂) and the relatively high vapor pressure of liquid deuterium (~194.6mbar at triple point) the condensation of deuterium even at the triple point temperature (18.71K) is not enough to extract the whole D₂ mass from the storage tanks. Deuterium solidification is therefore required. A heat exchanger helix around the condenser is used to cool it down to 4.5K during the condensation process. In this phase the condenser works practically as a cryogenic pump. In order to improve the condensation efficiency, the condenser is equipped with 6 internal fins to extend the condensing surface and improve the heat transfer. The condenser volume is around 40dm³ and is made of copper.

The heat load on the condenser during the 5K operation was estimated using a simple film condensation model. The equilibrium wall temperature was calculated by equating the heat deposited by the condensation process to the heat transferred to helium in the cooling coil wound around the condenser vessel

where is the condenser area including the fins, the heat transfer coefficient from wall to helium and the coil area. The heat-transfer coefficient from D₂ vapor to wall was calculated using the Nusselt correlation for the film condensation [6]

which depends on the liquid and vapor density , the liquid viscosity and the latent heat of vaporization of D₂. is the saturation temperature, the wall temperature and the height of the fins. is the helium logarithmic mean temperature difference. If on the helium side we assume a heat transfer coefficient of 500W/m²K and a temperature difference at the outlet of 1K we obtain a wall temperature of 23K and a heat load of 767W. Under these conditions the deuterium condensation speed will be 2.4g/s and the ~5kg deuterium will condense in about 0.6 hours. The calculated helium mass flow necessary to sustain this process is around 7g/s. Unfortunately, under this condition the heat load is higher than 300W, the maximum available cooling power of the refrigerator. The solution is to throttle the deuterium flow using a valve placed before the condenser such that during the condensation process the load on the refrigerator does not exceed the available cooling power. The drawback is a longer, but not prohibitive, condensation time.

1.4 The converter

The purpose of the converter is the catalytic para-ortho conversion of deuterium. The catalyst (OXISORB^®) is placed inside a cylinder with a volume of about 50dm³. About one third of the converter is filled with catalyst. It is not possible to fill the converter completely with catalyst because it will adsorb a big volume of D₂ and this volume will be lost. The converter vessel is provided with a demountable flange which allows the change of the catalyst for regeneration purposes. A cooling pipe, controlling the process temperature, is soldered on the outer wall of the converter vessel.

Because the generation of UCN is proportional to the amount of ortho deuterium and in order to estimate the heat load generated by the catalytic conversion a mathematical model has been developed. The underlying physics is an exoenergetic catalytic conversion reaction coupled to a non-homogeneous diffusion-convection process and a thermal conduction process with a volume heat generation problem. Initially, at 300K, deuterium is in the so called normal deuterium phase (nD₂) with the equilibrium composition: 66.67% ortho-D₂ and 33.33% para-D₂. After cooling to 5K and reheating to 20K the composition does not change appreciably with time because the natural conversion rate of para-D₂ to ortho-D₂ is very small (~0.06%/hr, time constant ~70days). The initial condition is therefore. The catalyst and the forced convection in the liquid accelerate dramatically the conversion. Measurements performed at PSI [4] show that the time constant of the conversion is in the range. The initial temperature is assumed to be 20K, the temperature of the heat sink at the converter wall. The conversion equation is a first order reaction for the ortho-D₂ concentration

where is the equilibrium concentration of ortho-D₂ at tempe-rature (K). At 20K the ortho-D₂ equilibrium concentration is 98.6%. Newly created ortho-D₂ diffuses out from the catalyst, a process described by a standard diffusion equation

where is the D₂ bulk self-diffusion constant [7]. Inside the catalyst the effective diffusion constant is probably smaller. We assume it to be three times smaller. The para-to-ortho conversion generates heat at a rate given by

where

is the latent heat of conversion and is the density of liquid D₂.

Finally, the thermal conduction process is described by the equation of heat conduction in liquid D₂ with a volumetric heat source Q, the heat generated by the para-ortho conversion

where and are the thermal conductivity and the specific heat of liquid deuterium.

The simulation result presented in Fig. 4 indicates that in the absence of forced convection the converter performance is poor even after 24 hours. Two thirds of the D₂ volume not in contact with the catalyst will practically not convert while the other one third will convert almost completely resulting in an average ortho-D₂ concentration of ~77.4% much lower than the required 98%. Therefore a gas-lift pump driven by a small heater (EH4 in Fig.2) to circulate the liquid D₂ trough the converter is used. The D₂ vapor generated by the gas-lift pump is recondensed in the condenser at 19K. With 50W of heating power, 0.15g/s D₂ can be circulated. The recycling time for the 5kg of D₂ contained in the converter is ~9hr.

1.5 The thermal shield

Due to the high level of neutron radiation we can not use super-insulation for the thermal insulation of the UCN storage tank and moderator vessel. We resort therefore to the classical solution of actively cooled thermal shield i.e. an aluminum screen cooled by a pipe wound around it. The optimum distance between the turns is calculated using the equation

where is the thermal conductivity of the shield material, the thickness, the maximum allowed temperature difference and the heat load from thermal radiation in W/m². A typical calculation for an Al shield of 3mm thickness shows that for a helium inlet temperature of 65K and a maximum temperature difference of 1K, the optimal distance between the turns is 0.26m which results in having ~10 turns with an overall length of 42m.

Figure 4: The elevation of the equilibrium (a) and non-equilibrium (b) ortho-D₂ concentration (right panel) after 24 hours along the vertical diameter of the converter. The time dependence of the maximum D₂ temperature (a) and the temperature in the center of the converter (b) (left panel).

The calculated radiation heat load is 550W and the total helium pressure drop is 120mbar. If the helium inlet is at the bottom of the shield then the bottom shield temperature is 65K and the top shield temperature will be 70.5K. The helium outlet temperature will be 70.3K. The Cryobox, which is placed 6m above the moderator vessel, has a low level of neutron radiation and super-insulation can be used without any problem of degradation.

2. Refrigerator

2.1 Specifications

The UCN source needs cooling at three temperature levels: 1) ~5K used to condense and keep the deuterium in the solid state in the condenser and moderator, 2) ~20K to liquefy the solid-D₂ in the condenser and to keep the liquid D₂ temperature in the converter during the para-to-ortho conversion process and 3) ~60-80K for cooling the thermal shields of the UCN storage volume. The cold source available is an existing helium refrigerator, which can deliver cooling power at two temperature levels, 4.5K and ~60-80K. For the 5K operation, the plant will be operated according to the regular specifications i.e. 300W at 4.5K with 32g/s. Typical for this operation is that the helium pressure before the last heat exchanger stage will be reduced by a special valve from 10-12bar to any other pressure down to 3bar. With this method we avoid an additional pressure reducing valve in the Cryobox and assure a moderate operating pressure in the moderator vessel. Operation in the 20K temperature range is performed using heaters (EH2, EH3, EH5 in Fig.2) placed before the cooling coils of the condenser, converter and moderator vessels. The warm gas coming out of these coils is cooled back to 4.5K in the phase separator’s heat-exchangers HX3, HX4 and HX5 and then throttled to low-pressure by valves as shown in the PID scheme of Fig.2. The calculation for operation at 23K is illustrated in the T-S diagram of Fig.5. We see that this operation can be performed with at most 5.5g/s while still ensuring that the refrigerator can operate in steady-state conditions. Q_heater is the power necessary to increase the helium temperature to 23K. Q_last and Q_HX1 are the heat transferred to the low pressure side of the last refrigerator heat exchanger and to the liquid helium bath of the phase separator respectively. Finally, Q_res is the rest cooling power which covers other thermal losses like those in the transfer line, heat conduction and thermal radiation.

Figure 5: The T-S diagram of the proposed 20K operation. Maximum 5.5g/s can be used to cool the condenser/moderator at maximum 23K. This mode requests 42g/s and operation at 1.35bar in the phase separator.

conclusionS

A helium cooling system for a new ultra cold neutron source readapting an existing refrigerator has been designed and optimized. The main processes responsible for the heat load: the neutron and gamma irradiation during a proton pulse, the D₂ condensation and the para-ortho conversion have been modeled and corresponding heat loads and time scales estimated. Special attention was given to the unconventional operation in the 20 K mode. Numerical simulations were essential to understand the different processes.

REFERENCES

1. Atchison, F. et al., The UCN Source at PSI, Proc. of the Int. Collab. On Advanced Neutron Sources, ICANS-XVIII, Guandong 2007

2. Wohlmuther, M. and Heidenreich, G., The spallation target of the ultra-cold neutron source at PSI, Nucl. Instr. Meth. A564, 51 (2006)

3. Morris, C.L. et al., Measurement of Ultracold-Neutron Lifetimes in Solid Deuterium, Phys. Rev. Lett. 89, 272501 (2002).

4. Bodek, K. van den Brandt, B. et al., An apparatus for the investigation of solid D₂ with respect to ultra-cold neutron sources, Nucl. Instr. Methods, A 533, 491 (2004)

5. Atchison, F., Calculated values for heating, particle fluxes and activation in components of the UCN source, Table IV, pag. 9, PSI Internal report TM-14-02-02 (2002)

6. Incropera, F. P. and DeWitt, D. P., Fundamentals of Heat and Mass Transfer, 4^th edition, John Wiley & Sons, Inc.

7. Souers, P.C., Hydrogen properties for Fusion Energy, University of California, Berkeley 1986

CR08-54

EXPERIMENTAL SET-UP OF HEAT TRANSFER MEASUREMENTS IN HE II

Chorowski M., Fydrych J., Strychalski M.

Wrocław University of Technology, Faculty of Mechanical and Power Engineering, Wybrzeże Wyspiańskiego 27, 50-370 Wrocław, Poland

Abstract

Superconducting magnets and cavities can be cryostated with superfluid helium under saturation or elevated pressure. To optimize heat transfer processes through electrical Rutherford superconducting cable insulation, a dedicated helium II cryostat has been designed, manufactured and commissioned. The cryostat enables reproduction of vast range of the helium thermodynamic states, including pressurization up to 6 bar of the superfluid helium. The cryostat construction, instrumentation and external equipment including vacuum pumping system are described. The paper also presents experimental results obtained during the cool-down of the cryostat.

Introduction

Currently built superconducting magnets, with the coils wound of low temperature superconductors, in most cases use NbTi alloy with its critical temperature equal to 9.6 K. There are attempts to replace the NbTi alloy with inter-metallic compound Nb₃Sn characterized by the critical temperature of 18.1 K, however the technological difficulties encountered in case of Nb₃Sn mechanical and thermal processing are still in favour
of NbTi [1]. Hence in case of accelerator magnets, that usually have to provide stable magnetic field at the level of 10 T, maximum acceptable temperatures can be as low as
1.8 K. The temperature can be practically achieved only with the use of superfluid helium.

In high-energy particle accelerators the superconducting magnet coils are exposed to the particles resulting from the beam losses and interacting with the magnets what leads to the energy dissipation in the coils and the helium. Additional amount of heat can be dissipated in the magnet structure due to small inter-movements of the cable wires caused by thermal stresses; moreover there are always residual heat fluxes from the environment through unavoidable thermal bridges. All these heat fluxes must be uninterruptedly transferred to the cooling helium within the allowable temperature margin which in case of superfluid helium can be of the order of 0.1 K.

In practice the accelerator magnets are usually immersed in liquid helium which penetrates and fills all voids in the magnet structure like for example: spaces in-between magnet yoke layers, longitudinal channels designed for magnetic field quality improvement, a distance between beam tube and magnet coil. As the NbTi superconducting coils do not need any resin impregnation, the helium can penetrate also through superconducting cables due to their inter-wire porosity, as well as through intentionally porous electrical insulation.

As mentioned above, all the integrated heat fluxes must be finally transferred to the helium. The efficiency of this process is determined by the heat transfer intensity between the magnet elements and the helium that can be in different thermodynamic states like two-phase saturated helium at normal pressure (He I), supercritical helium, superfluid helium at saturated pressure (saturated He IIs) or superfluid helium at about atmospheric pressure (pressurized He IIp).

The high energy physics accelerators with the superconducting magnets (Hera at DESY, Tevatron at FERMILAB, RHIC at Brookhaven National Laboratory and finally LHC at CERN) have relatively small heat influxes to the helium (not exceeding few Watts per meter). Sufficient heat transfer efficiency was guaranteed by magnet construction resulting from the field requirements. Therefore during the magnet design phase the heat transfer capability has not been the critical issue. However, because of planned LHC upgrade which will be focused on the improvement of the beam luminosity, it is predicted that the longitude density of the heat dissipated in the coils of some magnets can be as high as about 100 W per meter. To transfer this amount of heat to the helium requires reconsideration of the construction of superconducting coils with respect to the heat transfer intensity. It is very probable that the heat transfer can become a limiting factor in the increase of the energy and luminosity in upgraded and future accelerators.

In spite of the widespread use of liquid helium for cooling of superconducting magnets, RF cavities, cryogenic vacuum pumps and other devices and instruments, there is a lack
of sufficient knowledge that can be used for accurate calculation of heat transfer between magnet structure and helium remaining in different thermodynamic states. There is also
a lack of experimental data which allow reliable estimation of the heat fluxes from the Rutherford-type cables to supercritical or superfluid helium.

To perform heat transfer measurements a dedicated helium II cryostat was designed, manufactured and commissioned at the Faculty of Mechanical and Power Engineering of Wrocław University of Technology. The paper presents the cryostat construction and chosen reception test results.

1. EXPERIMENTAL SET-UP description

The experimental set-up for performing heat transfer measurements in superfluid helium is composed of three major parts (the scheme of the set-up is presented in Figure 1). The main part is a Claudet-type cryostat which enables to pressurise superfluid helium inside the measurement chamber [2]. During the cool-down and the measurements, the cryostat needs to be regularly supplied with liquid helium. For this purpose the experimental set-up
is equipped with a liquid helium supply system.

Figure 1: Scheme of the experimental set-up

The supply system is composed of a 100-litre liquid helium dewar and vacuum insulated helium transfer line. To force the liquid helium to flow into the cryostat, the dewar is pressurized with the gaseous helium supplied from the helium cylinder. The third part of the set-up is vacuum system which is composed of a high capacity vacuum pump, vacuum line and vacuum control valve. The vacuum system allows lowering temperature by decreasing the pressure of saturated helium in a dedicated volume in the cryostat.

1.1 Cryostat construction

The cryostat construction is based on the Claudet principle and its scheme is shown in Figure 2. Inside the cryostat there are three separated helium vessels that allow keeping helium simultaneously in three thermodynamic states: at normal boiling point temperature (He I), at superfluid helium saturated (He IIs) and superfluid pressurised (He IIp) states. Helium from the external dewar is regularly supplied to the He I vessel. The lambda plate separates thermally the He IIp volume from the He I vessel, while the lambda valve allows filling the He IIp vessel with helium and its replenishment during the subcooling phase.

Figure 2: The cryostat scheme and the helium thermodynamic state in the vessels

Liquid helium from the He I vessel can flow into the He IIs volume through a recuperative heat-exchanger followed by Joule-Thomson valve. During the subcooling phase the helium vapour from the He IIs volume is pumped away by a vacuum system and the helium vapour pressure is decreased to the level of a few millibars. The evacuated subcooled helium vapour preliminarily cools down the liquid helium streaming in the heat exchanger coil. Due to low pressure and isenthalpic expansion in the Joule-Thomson valve, the temperature of the helium is decreased below the lambda point temperature (2.17 K) and in the vessel He IIs a certain amount of helium is constantly kept in a superfluid saturated state.

The wall of the He IIp vessel, which is made of oxygen-free copper, does not cause
a significant heat resistance and the helium that is closed in the He IIp volume is smoothly cooled-down by the helium in the external He IIs vessel. As a result, the helium temperature in the He IIp volume also decreases below 2.17 K, but the pressure in this vessel can be kept at the stable level that is much higher than the equivalent saturated helium pressure.

1.2 Lambda plate and high pressure He IIp vessel

A lambda plate is a partition that separates the He IIp volume from the He I vessel. To decrease the heat transfer from liquid helium He I to superfluid helium a lambda plate is usually made of high performance insulator material, as for example composite G10. Then the pressure in the He IIp volume cannot differ significantly from the pressure in the He I vessel, which is usually a little bit higher than atmospheric pressure.

The lambda plate in the described cryostat, as well as the He IIp vessel have been designed to enable the measurements in superfluid helium which can be pressurised significantly above normal pressure level. The wall of the vessel is thick enough to withstand pressure difference up to 6 bars, whereas the lambda plate is composed of two rigid parallel metal sheets. Figure 3 shows a cross-section view of the high pressure He IIp vessel together with the lambda plate. The space between two metal sheets of the lambda plate is open to the He IIs volume where temperature is a little bit lower than in the He IIp vessel and pressure is of about few millibars. This solution helps with reducing heat flux from the He I vessel to
the He IIp volume. Heat is transfer only by thermal conduction through the lambda valve seat pipe and through a thin metal collar that connects two vessels in their circumferences.

Figure 3: Photo (a) and cross-section view (b) of the lambda plate and high pressure He IIp vessel

1.3 Vacuum pumping system

Vacuum pumping system is composed of a 4-meter vacuum pipe DN50, a vacuum control valve, a helium vapour heater and two vacuum pumps SOGEVAC SV 100 and SV 300 with the nominal pumping speeds equal to 97.5 m³/h and 280 m³/h, respectively. The system is presented schematically in Figure 1.

The vacuum control valve allows adjustment of the pressure in the He IIs volume and in this way the change and automatic control of the superfluid helium temperature. The helium vapour heater is made of a coil-shaped copper pipe with an electrical heater wound around it. It is installed in order to protect the vacuum pumps from damages that could be caused by cold helium vapour inflow.

To design correctly the vacuum system and especially to select a vacuum pump with
a sufficient pumping speed a dedicated mathematical model of the cryostat subcooling process has been applied.

1.4 Indispensable pumping capacity calculation

In the cryostat the subcooling process is based on the helium temperature lowering below 4.2 K in both He II vessels by decreasing the pressure in the He IIs volume. In the presented calculations we assumed a constant subcooling rate at the level of 1 mK/s. This rate should lead to reaching 1.8 K at about 2500 s. Depending on the presented in Figure 4 relationship between the saturated helium pressure and its temperature, and taking into account the assumed subcooling rate, we have obtained the evolution of the pressure in He IIs volume. Figure 5 shows the dynamic characteristics of helium temperature and pressure during the analysed subcooling process.


Figure 4: Helium saturated pressure versus temperature	Figure 5: Evolution of temperature and pressure in the He IIs vessel

We have also assumed that the content of helium in both He II volumes is V = 8 dm³ and total heat leaks Q_st to the helium through the vacuum insulation and thermal conduction bridges is equal to 3 W. Making use of the relationship between helium density and its temperature and pressure r = r(T, p), as well as between helium specific heat and its temperature and pressure c_p = c_p (T, p), we derived a formula for an instant heat flux that must be given up by the helium to guarantee assumed temperature drop rate:

. (1)

Figure 6 shows the evolution of the heat flux Q during a whole subcooling period. At the beginning of the process the heat flux is equal to 7.5 W only, and after decreasing gradually to 5.5 W during the first 1800 s, it rises rapidly to above 14 W. This rapid change is caused by an intense growth of the helium specific heat near the lambda line.

During the cryostat subcooling an instant heat flux Q is taken over by a certain amount
of evaporating helium Dm in a time Dt. This process can be described by the following equation:

, (2)

where r_He stands for heat of vaporization.

Due to the fact that the heat of vaporization for helium does not change significantly in analysed temperature range (1.8 K – 4.2 K), we have treated it as a constant and equal to 23 kJ/kg.

Combining the equations (1) and (2) yields a formula for the helium mass flow rate q_m:

. (3)

The variations of the calculated helium mass flow rate q_mduring the cryostat subcooling process are also presented in Figure 6. At the beginning of the process the mass flow rate
is lower than 0.4 g/s and decreases to 0.27 g/s within the first 1800 s. Then, because
of an intense growth of the specific heat near lambda line, to keep the assumed rate
of temperature drop, it has to rise rapidly to the value above 0.7 g/s. As soon as the final temperature (1.8 K) is reached, the mass flow rate decreases to 0.16 g/s, which is necessary to keep a stable temperature in both superfluid helium vessels.


Figure 6: Evolution of heat flux and helium mass flow rate	Figure7: Evolution of the volume flow rate of helium vapour warmed up to 310 K.

Estimation of both the helium mass flow rate (Fig. 6) and helium pressure evolution
(Fig. 5) was very useful for determining the volume flow rate of the helium vapour through the vacuum pumping system. Because the applied vacuum pumps can work only with warm gases it is necessary to warm helium vapour up to 310 K before the pump inlets. For this purpose we installed a coil with an electrical heater (see Fig. 1). This solution helps
to protect the pumps but it leads to the increase in an indispensable pumping speed.
The calculated evolution of volume flow rate of the helium vapour in the cross section just before the vacuum pumps is presented in Figure 7. In the analysed case the volumetric flow is affected not only by helium specific heat growth close to lambda line, but also by continuous pressure decrease. Therefore the volume flow rate, which is equal to a few liters per second at the beginning, rises to above 130 dm³/s at the end of the subcooling process. When the final temperature (1.8 K) is obtained the volume flow rate drops to 75 dm³/s.

The presented analysis proves that the indispensable pumping speed of the vacuum pumps applied for subcooling of the cryostat should be higher then 75 dm³/s. And of course the higher pumping speed the faster the final temperature is reached. Therefore we have decided to apply two pumps with the total nominal pumping speed of about 380 m³/h, which is equal to 105 dm³/s.

1.5 Instrumentation and data acquisition system

The cryostat described in the paper is equipped with two pressure sensors connected to the He IIp vessel (shown in Figure 3) and with six cryogenic temperature sensors - CernoxÔ. All the temperature sensors for the cryostat reception test were installed as shown schematically in Figure 8. Before the test the Cernoxes were calibrated in the temperature range 1.64 – 290 K and during the test they were connected to a specially designed 1mA DC source.

Figure 8: Schematical location of temperature sensors in He IIp and IIs vessels

To acquire measurement results we applied a dedicated LabVIEW™ code. The code controlled KEITHLEY 2000 multimeter and read 6 resistances every second and converted them on-line into corresponding temperature values.

2. THE CRYOSTAT RECEPTION TEST RESULTS

The analysed cryostat is dedicated for performing heat transfer measurement in superfluid helium environment. After construction and checking its overall behaviour first in nitrogen and then in helium temperature levels the cryostat was put to the final reception test.
The test was focused on subcooling helium and obtaining superfluid condition as well
as on decreasing temperature below 1.7 K. Measured temperature developments are shown
in Figure 9. The subcooling phase from the temperature 4.2 K to 2.17 K lasted 30 minutes. During this period there were noticeable differences among measured temperatures, even up to 0.4 K. Furthermore the helium inside the He IIp vessel was strongly stratified until the superfluid helium temperature was reached. The difference between temperatures measured in the top and bottom of the vessel exceeded 0.2 K. It was caused by a continuous small leak of liquid helium through the lambda valve that was driven by thermal compressibility of helium. This temperature difference disappeared when the helium in the He IIs vessel crossed lambda point – compare Figure 9.

Figure 9: Measured temperature courses during subcooling of the cryostat

Figure 10 presents temperature evolutions during the superfluid condition period with a few remarks concerning some important actions. The duration of this period was of about
60 minutes while the lowest temperature was equal to 1.6 K. The spread of measured values was rather stable along whole period. The difference between temperatures measured by coupled temperature sensors (T1& T2, T3& T4 and T5& T6) was lower than 25 mK.

Figure 10: Temperature changes during the superfluid helium state period
of the cryostat subcooling

The cool-down rate, expressed by the slope of the temperature evolution curve, indicates that it is feasible for the cryostat to reach a lower temperature level than measured during the commissioning. The subcooling phase was stopped due to the full use of the helium in the supply dewar.

conclusionS

The heat transfer phenomena may be the critical issue in future high-field accelerator magnet design and operation. There is a need of new experimental data enabling verification of new ideas of enhancing the heat transfer between the magnet elements, including superconducting cable and the cooling helium. A Claudet type cryostat oriented at the heat transfer measurements in superfluid helium at elevated pressure has been designed, manufactured and commissioned. A special attention has been paid to a high capacity helium pumping system and a simple analytical model of the required pumping process has been proposed.

REFERENCES

[1] Devred A. et al., High filed accelerator magnet R&D in Europe, IEEE Trans. Appl. Supercond (2004) 14 2 339-344

[2] Claudet G., Aymar R., Tore Supra and He II cooling of large high field magnets, Adv. in Cryogenic Eng. (1990) 35 55-67

ACKNOWLEDGEMENTS

The work has been partly supported within the cooperation agreement K944 signed between CERN, Geneva, Switzerland and Wrocław University of Technology, Poland.

CR08-39

S-N-S phase transitions of geometrically-metastable superconducting thin films

Ribeiro Gomes M.

Centro de Física Nuclear da Universidade de Lisboa, Av. Prof. Gama Pinto, 2,
1649-003 Lisboa, Portugal

Abstract

Phase transitions of superconducting films continue to be a frontier research area, not only as fundamental solid state issue but also due to the current potential applications of these devices as radiation detectors. A description of the S-N transition is provided in terms of flux penetration regimes and experimental characteristics fields; the N-S transition, being a continuous process, is associated with the different regimes of superconductivity nucleation.

Introduction

A complete theoretical description of Superconducting-Normal-Superconducting (S-N-S) phase transitions in laminar pure type-I/II superconductors is still. A key contribution to the better understanding of the S-N-S phase transition is achieved, using an unconventional experimental technique of real time flux motion detection, complementary to the conventional SQUID measurements which provide static magnetization results integrated over time scales ~3 orders of magnitude larger; or to static images, and more recently real-time high-resolution movies, with time scales ~0.1 s and spatial resolution of a few mm. The combination of these three techniques provides a promising laboratory for further investigation of the S-N-S phase transitions driven by an applied magnetic field, transport current, or radiation induced. Highly-sensitive energy detectors are continuously a challenging R&D topic, and cryogenic microcalorimeters using superconducting films are at the present time the best detector candidates for X-rays, electrons, nuclear recoils, etc, in many different fields of physics [1], such as mass spectroscopy of biomolecules, X-ray chemical analysis for industrial applications, cosmological dark matter searches, solar neutrino studies, X-rays astrophysics, double beta-decay and direct neutrino mass determination experiments. Superconducting microcalorimeters provide a large number of advantages over the other type of detectors. They can be relatively large and still be sensitive to small amounts of deposited energy, since the energy is sensed after it has been converted to heat (even interactions that produce little or no ionization can be detected). Being a thermal detectors, they do not depend on the charge transport properties of the absorber and a variety of superconducting materials can be used, in contrast to more conventional ionization detectors. While the microcalorimeter operating principle is simple, the full detector performance and optimization is not trivial. A full understanding behind the detection application rests on the solid state behavior of the superconducting films.

1. experimental technique

The measurements were performed in a single-shot He-3 refrigerator at temperatures 300 mK<T<4.2 K, with an overall experimental uncertainty of better than 0.5%. The magnetic field was applied perpendicularly to the sample by a coil external to the refrigerator, with an homogeneity of 1% over the sample area and relative precision of better than 2x10^-4. The applied magnetic field (H_a) sweep rate was varied from 0.5 G/s to 250 G/s, and it is PC-remote controlled allowing a continuous increase or decrease of H_a and its reversal to zero. The experimental technique is a real-time fast-pulse detection, described in detail elsewhere [2, 3]. The detection principle is based on Faraday’s law: the motion of a flux bundle within the sample, surrounded by a sensor loop, induces an electromotive force, which is proportional to the amount of flux quanta associated with the applied magnetic field strength and contained in a moving bundle of a given surface area. The flux motion induces a voltage-signal in the Cu pickup loop that is transformed-coupled to a fast amplifier. A 10 kHz frequency cutoff on the detection bandwidth allows a discrimination between applied magnetic field increments and flux bundle motions within the sample bordered by the sensitive loop. The acquisition system permits the synchronization of the magnetic field step-rise with the recording gate during which pulses are detected. The samples were mounted in a multi-layered design mechanically fastened to the cold plate of the He-3 refrigerator. The samples were cut from pure, annealed, pinhole-free 12.5-125 micron thick (L_y) polycrystalline metal foils provided by different commercially available suppliers. The purity was, in general, better than 99.99%. All samples were observed under different optical microscopes, with a < 1 µm spatial resolution, for surface dimension measurements and geometry imperfections. Any sample showing significant geometrical imperfections and non-parallel edges (maximum tolerance of 10% on the sample width) was not used in experiments. Prior to sample mounting, all samples were cleaned to eliminate any surface contamination by hand handling. The sample width (L_x) and length (L_z) were chosen in the majority of the cases to fit within the outer width of the Cu-loop (width of each Cu layer 150 mm and length of 3.2 cm; inner width of the loop 550 mm and outer 850 mm). Different type-I/II materials were studied: Re, Sn, Al, In, Pb, Ta, Zn,V and Nb, even if the majority of the studies were performed on Re and Sn.

2. Superconducting-to-Normal phase transition

As is commonly known, the S-N transition is characterised by three distinct regimes of flux penetration demarcated by two fields, denoted H_fdp and H_ldp for first and last detected penetration event, respectively. Typical integral and differential flux penetration curves are presented in Fig. 1(a) for a Sn and 1(b) for a Re film, respectively, where the number of events is plotted as a function of the applied magnetic field, H_a, which was varied from 0 G up to H_a>H_c(T).

Figure 1: (a) Integral S-N curves for 50-mm thick Sn film; (b) Differential S-N penetration curves for a 25-mm thick Re strip.

In the first regime, delimited by 0<H_a<H_fdp, no flux is admitted in the sample volume and a diamagnetic band is created along the sample volume. At a magnetic field given by, where is the demagnetization factor for a rectangular cross section, the superconducting state becomes unstable at the sample edges and flux is allowed to penetrate in the sample edges, up to an extent of the order of a few coherence lengths, ξ(T), creating therefore an edge intermediate state, well described by Landau [4]. This process is reversible and undetectable with the fast-pulse readout technique. The second regime, defined by H_fdp≤H_a≤H_ldp, is characterized by the migration of flux bundles across the diamagnetic band, containing a finite number of flux quanta, which diameter is determined by the curtain-like structure generated at the sample edges. Each flux bundle, circled by a superconducting current, moves to the center of the sample. This process is irreversible and detectable with the fast-pulse technique. As the magnetic field is further increased, the central region occupied by flux bundles expands and the diamagnetic band shrinks. The onset of this regime is geometry dependent and the end corresponds to the complete destruction of the diamagnetic band. The third and last regime is again reversible and delimited by H_ldp<H_a<H_c(T). In this regime, the normal central region has reached the curtain-structure at the edges, creating normal corrugated channels that expand reversibly with the magnetic field increase until the sample is in a complete normal state. No events are detectable.

In contrast with earlier wisdom [5, 6] this first flux penetration field severely depends on the trapped flux inside the sample volume after a first magnetic field scan. The so-called H_fdp in “virgin” (trapped flux free) penetration curves shows a good agreement with the predictions from Benkraouda [7] proposed for type-II or type-I superconducting flat strips. In the case of “non-virgin” curves the first entry field is defined by:

where is the aspect ratio and and depends on the amount of trapped flux; for the “virgin” case , and within the same measurement but different magnetic field cycle z is constant. The end of the irreversible regime of penetration, initiated by H_fdp, is defined by the magnetic field H_ldp where the last event is detected; no more events are detected for a further increase of the magnetic field. This field is generally smaller than the critical field H_c(T). However, when extrapolated to zero-threshold, and consequently eliminating the noise cut-off, H_ldp equals the critical field, in contradiction with the common knowledge but in agreement with published work [5]. Consequently, the H_ldp studies indicate that this observable field is coupled with the annihilation of the diamagnetic band surrounding the strip edges, and therefore has the same geometry dependence as H_fdp. The various experimental results yielded:

where depends on the superconducting material and threshold setting: for Sn independent of the temperature and for Re . The interpretation underlying these two constants is under investigation for a systematic study conducted in other materials undergoing a complete phase transition. For H_a>H_ldp, a continuous flux penetration driven by H_a occurs until H_a=H_c(T) and the strip is fully normal. In this regime no events are detected since they are outside the bandwidth of the pulse amplifier.

3. Normal-to-Superconducting phase transition

The N-S phase transition, on which fewer theoretical predictions exist and the disagreement among researches as to the existence of a barrier to flux expulsion is still open [8]. The measurements indicate that, although the N-S transition is a continuous process [9] characteristic fields defining the different regimes of flux expulsion are observable and depend on intrinsic superconducting properties of the material. In contrast with other experimental techniques and previous published work [3], it was possible to identify for the first time the two major characteristic expulsion fields: first expulsion field, H_fe, associated with the spontaneous nucleation field of superconductivity on a perimeter surface sheath, H_c3, suggested by Saint-James et al.[10] but never observed previously with this technique; and second expulsion field, H_se, the onset field of the spinodal regime, H_c2. Both fields were measured in different materials with different geometries.

In our experiments it was observed that the N-S phase transition is typically characterized by four regimes, demarcated by three characteristic fields: H_fde, and H_fe and H_se, as shown in Fig. 2. The integral curve corresponds to the expulsion part of the hysteresis cycle, following a complete field ramping to well above H_c(T).

Figure 2.:Typical expulsion curve with regimes and characteristic fields identification.

In the first regime for H_fde<H_a no events are detected, this regime may be extended to H_fewith the disappearance of H_fde; this field may be lower or higher than H_c(T) depending on the superconducting material and sample properties. The regime 2 of expulsion is defined by H_fe<H_a<H_fde may begin at fields ~H_c(T) depending on the superconducting properties of the material. The third and more perceptible then regime 2 occurs for H_se<H_a<H_fe, where a variable number of events may be detected. The majority of events are detected in regime 4; frequently this is the only detectable regime, and occurs for 0<H_a<H_se. The identification of the characteristic expulsion fields is also shown in Fig. 3, which represents typical a differential expulsion curves with four visible regimes of expulsion present. This measurement was conducted in a 50-mm Sn sample at 350 mK.

Figure 3: Typical differential expulsion curve with characteristic fields identification.

The flux expulsion onset occurs at H_fde and is characterized by a narrow signal. A further H_a decrease induced no additional expulsion events until H_a=H_fe where a small event rate is observed. The field H_se defines the beginning of a rapid event rate which persists throughout the remainder of the field decrease. H_c3 has a weak signature corresponding to ~0.05% of the total detected signal, consistent with a surface sheath of width 2x at the strip perimeter. This field is often not observed due to the noise level associated with that measurement. H_c2, in contrast, sets the beginning of the last regime of flux expulsion where ~99% of the signal is detected. While H_se does not depend on the sample geometry nor surface quality, providing a convincing agreement with H_c2(T), as can be seen in Fig. 4, and the extracted Landau-Ginzburg k parameter, as indicated in Fig. 5 for the case of Re; H_fe deviates from the theoretical predicted H_c3(T) or H_fe(T)/H_se(T) depending on the sample surface roughness.

Figure 4: Variation of h_se with H_c/ξ² for various materials.

Figure 5: Experimental k values as a function of sample thickness for various Re samples.

A third expulsion field, H_fde, is frequently observed at higher magnetic fields and in the vicinity of H_c(T) which is associated with heterogeneous nucleation resulting from metallurgy and/or impurities that stimulate nucleation in the vicinity of the defect, also observed with other experimental techniques [11].

conclusionS

In this work, a complementary approach of the S-N-S thin film phase transition is provided in terms of the magnetic flux motion within an intermediate state evolution process, achieved with an unconcentional real-time fast-pulse readout technique. Different superconducting materials were studied under various experimental conditions, and for different geometrical aspect ratios. The magnetic fields characterizing the energy barrier to flux penetration, of geometrical original, were clearly identified and compared with the existing theoretical models for the first flux penetration field. The N-S phase transition, being a continuous process, was shown to be geometry independent and directly related with the superconducting properties of the material. The two characteristic expulsion fields, were observed, for the first time with this technique, corresponding to the spontaneous nucleation field of superconductivity on a perimeter surface sheath and the onset of the spinodal regime field. A third expulsion field, associated with heterogeneous nucleation resulting from metallurgy and/or impurities stimulating nucleation near the defect, together the other two fields were proven to be an excellent tool to extract fundamental superconducting parameters, such as the Landau-Ginzburg k of a material. These are key features when superconducting systems are designed for radiation detectors [12, 13], in general. The achievements in the b-decay ¹⁸⁷Re spectrum for H_p<H_fdp for direct electron antineutrino mass determination yielded promising expectations [14]. On the other hand, superconducting cryogenic detectors are currently the best performing detectors, in particular for neutrino mass measurements [15], where thin superconducting films are used as thermal sensors. Specifically, transition edge sensors (TES) have demonstrated a sensitivity of a few eV. The theoretical prediction for its optimum performance has not yet been experimentally reached and the modeling of such a behavior is necessary. The intrinsic vortex dynamics and intermediate state evolution, within the superconducting volume, are fundamental features to clarify for pushing an experimental detector sensitivity to the sub-eV threshold.

Acknowledgements

This work was supported by the grants PRAXIS/10033/1998 and SFRH/BPD/36293/2007 of the Foundation for Science & Technology of Portugal. Part of this work was accomplished during the author PhD programme at the Center of Nuclear Physics of the University of Lisbon (CFNUL), Portugal and the experimental work supported by the scientific projects funds of the Advanced Detectors Group of the CFNUL.

REFERENCES

1. Proceedings of the 11^th International Workshop on Low Temperature Detectors, Nucl. Instr. And Meth. A (2006) 559

2. Jeudy, V., Limagne, D. and Waysand, G., Irreversible flux entry in tin superconducting strips: Geometrical Metastability, Europhys. Lett. (1991) 16 491-496

3. Jung, G., Girard, T.A.,Valko, P., Gomes, M.R., Jeudy, V., Limagne, D. and Waysand, G., Expulsion of magnetic flux in a type-I superconducting strip, Physica C (2002) 377 121-129

4. Landau, L.D., Zh Eskp. Teor. Fiz (1937) 7 371 and Jour. Phys. (1943) VII(3) 99

5. Jeudy, V., Jung, G., Limagne, D. and Waysand, G., Irreversible flux penetration regimes in type-I superconducting strips, Physica. C (1994) 225 331-336

6. Gomes, M.R., “Destruction of the geometrical energy barrier in type-I superconducting rhenium”, Master Thesis, Faculty of Science/University of Lisbon, Portugal (1998)

7. Benkraouda, M. and Clem, J. R., Magnetic hysteresis from the geometrical barrier in type-II superconducting strips, Phys. Rev. B (1996) 53 5716-5726

8. Valko, P. Gomes, M.R:, and Girad, T.A., Nucleation of superconductivity in thin type-I foils, Phys. Rev. B (2007) 75 140504-1-4

9. Huebener, R. P., “Magnetic flux structures in superconductors”, (1979), Springer-Verlag, Berlin Heidelberg, 2^nd ed., (2001)

10. Saint-James, D., and Gennes, P.G., Onset of superconductivity in decreasing fields, Phys. Lett. (1963) 7 306-308

11. Shoenberg, D., Superconducting cylinders, Proc. Camb. Phil. Soc. (1937) 33 260-276

12. Gomes, M.R., “Study of phase transitions in geometrically-metastable superconducting m-laminar systems”, PhD Thesis, Faculty of Science/University of Lisbon, Portugal (2005)

13. Jeudy, V., Collar, J.I., Girard, T.A. Limagne, D. and Waysand, G., S-35 beta irradiation of a tin strip in a state of superconducting geometrical metastability, Nucl. Instr. And Meth. A (1996) 373 65-67

14. Gomes, M.R., Girard, T.A., Oliveira, C., Jeudy, V. and Limagne, D., A superconducting measurement of the 187-Re beta decay spectrum, Nucl. Instr. And Meth. A (2000) 444 84-87

15. Nucciotti, A. for the MARE collaboration, The MARE Project, J. Low Temp. Phys. DOI 10.1007/s10909-008-9718-5

CR08-07

Black surfaces for cryogenic applications

Králík T., Hanzelka P., Musilová V., Srnka A.

Institute of Scientific Instruments of the ASCR, v.v.i
Academy of Sciences of the Czech Republic, Královopolská 147, 612 64 Brno

Abstract

Surfaces with high absorption or emission of thermal radiation are often needed in cryogenic and space applications. But it can be more difficult to realize a sufficiently black surface applicable in cryogenic systems than to make highly reflective surfaces. Experimental results on black surfaces concerning the thermal radiative properties in the temperature range from 20 to 300 K of the source of the thermal radiation are presented.

Introduction

Black surfaces which absorb or emit thermal radiation at low temperatures have found their typical application in space-borne devices [1] and cryopumps [2, 3]. They are realized as coatings on surfaces of metals that ensure the heat transfer towards or from the black surface. Especially in this region of applications it is not a trivial task to prepare sufficiently black surfaces. In addition to the “blackness” in infrared and far infrared, the properties of the surface like outgassing rates (vacuum compatibility), mechanical properties or possibly directional “residual” reflectivity (diffuse or specular) are of great importance. Primary motivation of this work was to find a suitable coating for baffles of a cryopump [4] and to develop a reference black surface in apparatus for emissivity and absorptivity measurements [5, 6]. For these reasons we have measured and compared low-temperature emissivity and absorptivity of various coatings on copper and aluminium.

1. Method and Apparatus

We have performed measurements of radiative heat flow between the opposite surfaces of two interchangeable concentric parallel discs (samples) of the same diameter, the radiator and the absorber [5, 6].

From the measured radiative heat flow Q_R, the mutual emissivity (emissivity factor) e_RA of radiator and absorber surfaces was evaluated:

(1)

In this relation, T_R and T_A (T_R>T_A) are the temperatures of the radiator and absorber, whereas A and s represent the area of the measured sample surface and the Stefan-Boltzmann constant, respectively.

Let e_R represent the emissivity of the radiator and a_A the absorptivity of the absorber, respectively. Then an approximate relation can be written:

(2)

This relation is exactly valid for spectral values of e_RA, e_R and a_A and when parallel surfaces are infinite. Thus for total hemispherical values e_R(T_R) and a_A(T_A, T_R), the formula (2) represents a good approximation when properties of the surfaces depend weakly on wavelength of the radiation (grey surfaces), i.e. when a_A(T_A, T_R) = a_A(T_A) = e_A(T_A). Here a_A(T_A, T_R) is the total hemispherical absorptivity of the absorber surface at the temperature T_A for incident radiation of blackbody at the temperature T_R. For finite area we can use the formula (2) when the diameter of radiator and absorber is much larger than the distance between them (view factor approaches unity).

The approximate values of absorptivity a_A or emissivity e_R of a sample can be calculated from (2) for the known values e_R or a_A of the surface opposite the sample (reference surface). When the emissivity or absorptivity of the reference surface is much higher than the absorptivity or emissivity of the sample, the measured e_RA may be considered, with a small error, to be the absorptivity or emissivity of the sample, respectively.

From the measurement of heat transfer between two identical samples, the emissivity of one surface can be evaluated if we assume that absorptivity does not depend on the temperature of the material or this dependence is weak, i.e. when a_A(T_A, T_R) » a_A(T_R). Then the emissivity of the radiator equals the absorptivity of the absorber and thus the emissivity e_R(T_R) = a_A(T_R) can be evaluated from the measured value e_RA using the approximate relation (2).

In our apparatus, the heat exchanged between radiator and absorber Q_R is conducted from the absorber through the thermal resistor into the LHe bath and its value is derived from the temperature drop on the thermal resistor in steady state. The range of measurable heat Q_R can be changed by adding of a thermal shunt parallel to the thermal resistor. A bronze tape or copper wire was soldered in parallel to the thermal resistor, which is made from a thin stainless steel tube. Consequently, for the same Q_R the temperature T_A of the absorber is different with different shunt, which we also used for testing of the influence of the absorber temperature on the result of measurement.

The accuracy of Q_R measurement is within 3 % of the measured values. In addition, a systematic error, reducing the measured values, corresponds to the value 0.96-0.97 of view factor for radiator and absorber configuration. This error reduces by about 3-4 % the measured value of e_RA for nearly black surfaces.

2. MATERIALS

Sample substrates are made of a Cu sheet (99.5%) and they have the form of discs with the diameter of 40 mm and thickness of approx. 1 mm. The side of the substrate on which coating is applied was usually treated by grinding with wet corundum sand or with steel wool.

Three types of the epoxies were used. The first of our samples was covered with epoxy Aralditeä LY 5210 filled with soot. Later we applied epoxy resin Spolimer with hardener HT1 and Epoxy Chs520 with hardener Telalit 600. Both epoxies are produced by the Czech firm Spolchemie, a.s [7]. In some coatings polyester screen printing fabric Sefar PET 1000 77-48Y PW (abbrev. SPF) was used as filling and reinforcement. This net-like fabric is 80 mm thick and its projected area is about 65%.

2.1 Samples

Epoxy Araldite LY 5210 + 1% soot - Samples 1, 3

The roughened substrate was brush-painted with the epoxy Aralditeä LY 5210 with hardener Aradurä HY 2954 (prod.: Ciba A.R.L. Ltd) filled with the 1 wt% soot Chezacarb (prod.: Chemopetrol a.s, Czech Rep.). The thickness of the coating is on average of 70 mm and it has glossy black colour.

Epoxy Spolimer with one layer of the SPF - Samples 32, 33

One layer of the SPF impregnated with the Epoxy Spolimer was placed on the substrate. The excess of the epoxy was wiped away. The epoxy was hardened 1 h at 100°C, 1 h at 140°C and 2 h at 180°C. The total thickness of the coating is 90 mm. The coating has red brown appearance with visible texture of the SPF and it is visually semitransparent. This coating was also used as an absorptive coating on chevrons in a small helium bath cryopump [4].

Epoxy Chs520 with one layer SPF covered with the Mylar^â foil - Samples 49, 50

The SPF impregnated with the Epoxy Chs520 was applied on the substrate. The SPF was covered with the Kapton^â foil coated with Teflon^â and a force 10 N was applied. After the successive epoxy curing for 1 h at each of the temperatures 60°C, 80°C, 100°C and 120°C, the Kapton^â foil was peeled off and the Mylar^â foil without aluminization (th. 12 mm) was glued on the epoxy surface with the adhesive Loctite^â480ä. The total thickness of the layer is 120-130 mm (Nr. 49) and 140-160 mm (Nr. 50). The coating has spotted black appearance due to black colour of the adhesive.

Epoxy Chs520 – 380 mm (four layers of the SPF) - Samples 57, 58

Four layers of the SPF impregnated with the epoxy Chs520 were placed on the Kapton^â foil coated with Teflon^â and than the substrate was pressed on this layer. The curing proceeded under a force of 10 N and at the same temperatures as by the samples 49 and 50. After curing the epoxy, the Kapton^â foil was peeled off. The samples have yellowish appearance. The total thickness of the coating on both samples is on average 380 mm.

Chemglaze Z306ä on aluminized Kapton^â - Samples 39, 40

Double side aluminized Kapton^â coated with the paint Chemglaze Z306ä on one side was applied on these samples. The Kapton^â foil has thickness 25 mm and the total thickness of the coated foil is 80 mm. The paint is based on a polyurethane binder filled with fumed silica and carbon [1]. This foil was glued with the Loctite^â480ä to the substrate covered with the similar layer as that on the sample Nr. 32 and 33. The foil was provided by the firm Austrian Aerospace GmbH, Austria [8]. The paint Chemglaze Z306ä (Aeroglaze) is used for thermal control in space-borne instruments and it is also often used as a well-tried black surface for cryogenic applications.

Black paints – Samples L1, L2 and L3, L4

Commercially available paints produced by the German firm Warnecke&Böhm GmbH (WB-Lacke) were applied on substrates. The samples L1 and L2 were coated with visually matt-black epoxy baking paint. The binder of the paint is acrylic epoxy resin. The samples L3 and L4 have the same appearance but the type of the paint binder is amino-alkyd. The thickness of the coating on samples L1 and L2 is approx. 20 mm and 25 mm for the samples L3 and L4. Both paints are one-component, they were applied by spraying and hardened by baking at 150°C.

Diamond Like Carbon (DLC) on copper substrate - Sample 65

The DLC coating was deposited by combined PVD/PACVD (Physical Vapour Deposition/Plasma Assisted Chemical Vapour Deposition) method in industrial coating equipment of the firm HVM Plasma s.r.o [9]. The base layer Cr/WC:H was deposited by magnetron sputtering to maintain good adhesion of the DLC, whereas the top DLC layer was deposited by pulsed plasma from acetylene. The total coating thickness is 3.3 μm. The thickness of the top DLC layer is 2.2 μm. This coating is commercially used in mechanical engineering for reduction of friction and wear. Maximum working temperature of this coating is 300°C.

Fractal Blackä on Al foil - Samples 60, 61

Fractal Blackä is a coating very black for visible and infra-red radiation. It was provided by the firm Acktar Advanced Coatings Ltd. (Israel) [10]. The coating was created by magnetron sputtering on a thick Al foil. The foil was glued with Loctite^â480ä to the substrate. Fractal Blackä is completely inorganic layer and its thickness is 10-12 mm.

Nanocomposite coatings on copper substrate - Samples 67, 68

MARWIN^â SI (sample 68) is a nanocomposite AlTiSiN system which is commercially used as a super-hard coating of the machinery cutting tools. The coating is created by using PVD in special coating facility in the firm SHM s.r.o. (Czech Rep., Šumperk) [11]. The micro hardness of the coating is 43 GPa, the thickness is 2-3 mm and the thermal stability is better than 1000°C. The colour of the coating is black.

LUBRIK^â SI (sample 67) consists of the gradient film TiAlN as the base layer and of the end lubricating layer TiAlCO. The total thickness of all layers is 3.8 mm. The coating has shiny black colour. This coating is used for working of nonferrous metals and for pressing and forming tools and it was also created by the firm SHM s.r.o.

Meas.	Radiator	Absorber	Coating and Configuration Radiator × Absorber	Thermal shunt
ME_P5	1	3	Epoxy Aralditeä HY 5210 + 1% soot	Shunt 0
ME03	32	33	Epoxy Spolimer + 1 SPF	Shunt 0
ME06	49	50	Epoxy Chs520 + 1 SPF + Mylar foil	Shunt 0
ME09	57	58	Epoxy Chs520-380 mm	Shunt 1
ME11	58	57	Epoxy Chs520-380 mm	Shunt 1
ME05	40	39	Chemglaze Z306ä on Kaptonä	Shunt 0
E28	40	58	Chemglaze Z306ä × Epoxy Chs520-380 mm	Shunt 2
MEL1	L1	L2	Acrylic epoxy paint, WB-Lacke	Shunt 2
MEL2	L3	L4	Amino-alkyd paint, WB-Lacke	Shunt 2
A62	58	65	Epoxy Chs520-380 mm × DLC	Shunt 2
A63	58	65	Epoxy Chs520-380 mm × DLC	Shunt 1
E25	65	58	DLC × Epoxy Chs520-380 mm	Shunt 2
E29	61	58	Fractal Black^â × Epoxy Chs520-380 mm	Shunt 2
E30	68	58	MARWIN^â SI × Epoxy Chs520-380 mm	Shunt 2
E32	67	58	LUBRIK^â SI × Epoxy Chs520-380 mm	Shunt 2

Table 1 : Description of individual measurements and their configuration.

3. Measurement RESULTS and discussion

Table 1 lists the measurements. The letters ME refer to the measurements in which the same surfaces are both on the radiator and absorber. Letter E refers to the emissivity measurement, i.e. when the studied sample is placed on the radiator and for the absorber is used a reference surface as black as possible (samples 57 and 58). Letter A refers to the absorptivity measurement with the sample mounted on the absorber opposite to “black” radiator. Measurement without shunt (Shunt 0) means that the basic thermal resistor made of a stainless steel tube is used. In most measurements, a bronze (Shunt 1) or copper shunt (Shunt 2) was soldered in parallel to the basic thermal resistor.

In following figures, the open symbols and dashed lines represent the emissivity e_R or the absorptivity a_A of the samples, evaluated from the measured values e_RA (full symbols) by using the relation (2) for known a_A or e_R of the opposite surface or using relation a_A = e_R.

3.1 Epoxies

No cracking or peeling of coatings after repetitive cool down from room temperature to 5 K was observed. Outgassing, very probably of water, was observed during several measurements of very low values of absorptivity of metals when the epoxy on the radiator achieved temperatures over 250 K. This effect was registered as a very slow increase in measured metal absorptivity, probably due to formation of a condensate on the metal.

Identical samples 57 and 58 were used as the radiator and absorber in the measurements ME09 and ME11. In each of these measurements, we used different thermal shunts and thus the influence of the absorber temperature T_A (Fig.1 – right hand side axis) on the measured mutual emissivity could be checked. Although the temperatures of material on the absorber differ at the same T_R in the experiments ME09 and ME11, the measured values of e_RA are identical within the accuracy of measurement. It is in agreement with expectation of weak temperature dependence of optical properties of dielectrics in infrared and far-infrared. From the measurement we can deduce that at least within the interval of material temperatures T_A from 5 to 60 K and radiation temperatures T_R from 40 to 160 K, the absorptivity of the samples 57 and 58 depends very weakly on the temperature of the coating.

Figure 1 : Epoxy coatings. Measured e_RA and emissivity e_R evaluated using relation (2) under condition e_R=a_A. Gray curves T_A(T_R) are temperatures of the absorber in measurements
ME09 and ME11.

3.2 Paints

The emissivity of the Chemglaze Z306ä at T_R=300 K, evaluated from our measurement, is 85 % (Fig. 2). This is a very good result in comparison with epoxies if we take into account that the thickness 55 μm of the Chemglaze Z306ä coating is nearly one order smaller than the thickness of the epoxy layer with emissivity 89 %. The data found in literature show higher values of the Chemglaze Z306ä emissivity [12, 13]. This may be caused by the higher thickness of the brush-painted coating (100 mm) in the case of the Jamotton’s data [12]. Nevertheless the values in [12] are unusually high in comparison with our results for the temperatures T_R<100 K. The data published by Fabron [13] are more consistent with our results.

For both paints from WB-Lacke and Chemglaze Z306ä, we obtained nearly the same value of emissivity at 300 K. For the paint with amino-alkyd binder, also the low temperature emissivities approach the values obtained for Chemglaze Z306ä. The thickness of this paint is half of the Chemglaze Z306ä paint thickness.

Figure 2 : Paint coatings. Measured e_RA, emissivity e_R evaluated from (2) and published data on emissivity of Chemglaze Z306ä.

3.3 Thin film coatings

The highest values of emissivity were achieved with the coating Fractal Black^â (Fig. 3). The total hemispherical emissivity e_R estimated from the relation (2) for the temperature region 230–300 K is approximately 90 %. For comparison, the spectral hemispherical absorptivity of the Fractal Black^â, according to the producer’s data [10], decreases from 99.5 % at wavelength 10 mm (300 K black body temperature) to 96 % at 13.5 mm (230 K).

The measured absorptivity of the Diamond Like Carbon is of about 60 % in the range T_R = 150-300 K (Fig.3). Similarly like for thick epoxy layers (see sec. 3.1), we checked the influence of the temperature of DLC layer on its absorptivity. In two measurements with different shunts, the different temperatures T_A of DLC layer at the same temperature T_R (Fig. 3) were achieved. The measured absorptivity did not change with temperature of the layer. We can conclude that in the region of temperatures T_A = 5–75 K and T_R = 60–190 K the absorptivity of the DLC layer does not depend on the material temperature. The independence on material temperature was confirmed by emissivity measurement (not presented in Fig. 3). The values of mutual emissivity e_RA obtained in “absorptivity” and “emissivity” measurements at T_R = 30–300 K are equal within a relative accuracy of 2 %.

The films MARWIN^â SI and LUBRIK^â SI have lower emissivity above 40 K than the DLC (Fig. 3).

Figure 3 : Thin films. Measured e_RA and from (2) evaluated emissivity e_R and absorptivity a_A. T_A(T_R) are temperatures of absorber in measurements A62 and A63.

Conclusions

We have measured total hemispherical emissivity and absorptivity of various types of coatings below room temperature up to 15 K. Coatings based on epoxy resins (thickness 70-380 mm), polyurethane and alkyd paint (thickness 20-55 mm) and also inorganic thin films (2-11 mm) were applied on Cu or Al substrate. All coatings withstood cryogenic and vacuum conditions. Also the films that are commercially used as super-hard coatings of machinery cutting tools lasted out the cool down to 5 K in spite the fact that the coatings were applied on Cu whereas the technology is originally developed for cutting tools coatings. The nanocomposite coatings and DLC film may be applicable in cryogenics if also high temperatures are needed.

For two coatings, thick epoxy layer and DLC layer, we have tested the dependence of radiative properties on the temperature of the coating in the region of material temperatures 5-80 K (temperatures of the source of radiation were 20-160 K). We did not observe any dependence of measured radiative properties on the temperature of material. The independence on material temperature was confirmed for DLC layer by measurement of both absorptivity and emissivity at temperatures from 20 to 300 K.

The increase of transparency for far infra-red electromagnetic radiation with increasing wavelength is the property expected for dielectric materials. This behaviour results in the decrease of emissivity of dielectric layers on metals with decreasing temperature. For all coatings we observed the decrease in measured emissivities at low temperatures, i.e. the surfaces are not grey at low temperatures. At higher temperatures, weak dependence on the temperature was obtained. The non-greyness at low temperatures could influence the values of low temperature emissivities and absorptivities that were evaluated under the assumption of grey surfaces.

The highest values of emissivity were achieved for thick epoxy composite layer (380 mm) and Fractal Black (11 mm). The epoxy coating is filled with polyester net (filling more than 50 %). At temperatures over 60 K its emissivity approaches 90 % and about 80 % at 30 K. For Fractal Black coating, we have measured values 85-88 % above 120 K and about 60 % at 30 K. In spite of very low coating thickness (approximately 3 mm), relatively high emissivity values were observed for DLC layer.

Acknowledgement

This work was supported by the Academy of Sciences of the Czech Republic (Project Nr. AV0 Z20650511). The authors thank the firms HVM Plasma s.r.o. (Czech Rep.), SHM s.r.o. (Czech Rep.), Spolchemie a.s. (Czech Rep.), Acktar Advanced Coatings Ltd. (Israel), and Austrian Aerospace GmbH (Austria) for providing the samples.

References

1. Persky, M. J., Review of black surfaces for space-borne infrared systems, Rev. Sci. Instrum. (1999) 70 2193-2217

2. Benvenuti, C.J., Characteristics, advantages, and possible applications of condensation cryopumping, J. of Vacuum Science and Technology (1974) 11 591-599

3. Haefer, R.A., Kryo-Vakuumtechnik, Springer-Verlag, Heidelberg (1981)

4. Musilova, V., Dupak, J., Hanzelka, P., Kralik, T. and Urban, P., Economical helium bath cryopump: design and testing, Vacuum (2004) 74 77-83

5. Kralik, T., Hanzelka, P., Musilova, V., Srnka, A., Device for measurement of thermal emissivity at cryogenics temperatures, 8th Cryogenics 2004 IIR international conference, Icaris, Praha (2004) 23-29

6. Musilova, V., Hanzelka, P. Kralik, T., Srnka, A., Low temperature radiative properties of materials used in cryogenics, Cryogenics (2005) 45 529–536

7. Spolchemie a.s. Czech Rep. [online]. [2008] <http://www.spolchemie.cz>

8. Austrian Aerospace GmbH. [online]. [2008] <http://www.space.at/htmldocs/2.html>

9. HVM Plasma s.r.o. Czech Rep. [online]. [2008] <http://www.hvm.cz>

10. Acktar Advanced Coatings Ltd. [online]. [2008] <http://www.acktar.com/category/FractalBlack>

11. SHM s.r.o. Czech Rep. [online]. [2008] <http://www.shm-cz.cz/en/products/pvd-coatings>

12. Jamotton, P. et al., Measurement of the total hemispherical emittance of different surfaces at temperature from 4.4 to 200 K, Sixth European Symposium on Space Environmental Control Systems, Noordwijk, The Netherlands, ESA SP-400 (1997) 543-547

13. Fabron, Ch., Meurat, A., Measurement of total hemispheric emissivity at low temperatures / designing a cryogenic test bench, Fourth International Symposium Environmental Testing for Space Programmes, Liège, Belgium, ESA SP-467 (2001) 51-59

CR08-11

25 TESLA HTS MAGNET INSERT COIL IN ZERO BOIL OFF CRYOSTAT

Good J., Bracanovic D.

Cryogenic Ltd, 30 Acton Park Industrial Estate, London W3 7QE, UK

ABSTRACT

There is increasing interest in magnetic fields for NMR at above 1 GHz (23.48 Tesla) but these fields are not available with commercial Low Temperature Superconductors (LTS) at either 4.2 or 2.2 K or reduced temperature.

Cryogenic Ltd has manufactured a coil which is designed to demonstrate the feasibility of a magnetic field of 25 Tesla in a working bore of 50mm. The magnet uses HTS conductors combined with LTS and is suited to solid state research for NMR and ESR. It is cooled by closed cycle cryostat and runs at 4.2K.

The application of High Temperature Superconductor (HTS) is additionally attractive because the magnet can run at 4.2K rather than being pumped to 2.2K as the critical field of HTS is much higher than 25 Tesla.

The magnet consists of 5 coils; the outer coil is of NbTi section, and the next of Nb3Sn section. Inside are 3 coils of a high-temperature superconductor HTS BiSrCaCuo-2223 tape.

The outer two coils have a 140mm bore and provide 15 Tesla at 4.2K. The target for the HTS is to provide up to 9 Tesla at 4K.

The performance of the magnet both LTS and HTS section is discussed together with the operating characteristics of the closed cycle cryostat.

1. introduction

Cryogenic Ltd has manufactured a coil which is designed to demonstrate the feasibility of a magnetic field of 25 Tesla in a working bore of 50mm at 4.2K.

At present the only way fields of 25 Tesla can be generated continuously is by a bybrid or resistive magnet. However, hybrid coils require 10 megawatts of DC power to operate and the field generated is not sufficiently stable [1].

The applications for extra high field magnet systems are many and varied. Some examples are:-

o Characterization of new material and semiconductors including nitrides and zinc oxides, by Hall effect resistivity, specific heat, magnetic moment, De Haas-van Alphen measuremnts.

o High field spectrometry using NMR at 1 GHz for research into structures of bio macro-molecules which is important for the development of new drugs. Both sensitivity and resolution can be improved by moving to higher frequencies of 1 GHz and beyond.

o Measurements of other Physical properties such as quantum fluids and plasma phyiscs.

2. the project

In this project the ultimate goal is to provide a superconducting magnet capable of generating a continuous magnetic field of 25 Tesla. The magnet was built under the “HIGINS” project and funded by the EU under the 6th framework for Research.

The magnet with a first generatioin HTS set of insert coils has been delivered to the Engineering Department of Cambridge University. Since the BISCO [2] tapes used in the HTS insert coils were relatively low current density after insulation, the field provided is modest so the magnet only provides 19 Tesla at 4.2K. New coils are now being build using second generation HTS tapes that have much higher currents and with these the magnet should achieve the expected 25 Tesla field performance.

The magnet had to meet strict requirements for stray field with a 5 Gauss line not more than 4 metres from the field centre and this motivates a compact design since it is the diameter of the outer windings which give use to a large diopole moment and stray field.

It was also required that the magnet should have low or zero helium consumption. Accordingly, the magnet was designed to run in liquid helium with a zero loss cryostat. The cryostat has its own 4K cryocooler [3] built in to recycle and recondence helium.

2.1.The Magnet

The outer magnet is designed to be as compact as possible but to provide 15 Tesla at 4.2K in a 140mm bore. Inside this diameter are three coils each made of HTS conductor, two coils have single lengths and one required a single joint within the winding. These coils ideally would have provided 9 Tesla at 4.2K giving a total field of 24 Tesla. The outer magnet is made with two winding, one of NbTi and one of NbSn.

Filamentary NbTi conductor is used for the outer winding and provides a field of 8.5 T at 4.2K in a bore of 220mm.

The inner coil is wound using bronze route Nb₃Sn. This conductor is widely used and very reliable. It can withstand high tensile stress and is well suited for fields up to the 17 Tesla. Above that the critical current decreases making alternative conductors more useful. To obtain good high field performance, the Nb₃Sn is alloyed with small quantities of tantalum and titanium, which both increases Jc and Bc.

The three inner sections are made of a high performance, high strength BiSrCaCuO tape in a silver alloy matrix. The project generates up to 9 Tesla in a background field of 15 Tesla. Research on silver alloys to provie a high strength matrix proceeded in Poland [4] with additional tests in Dresden [5] under the Higins Project. In the event, however, the best conductor available to wind the coils was purchased from American Superconductor. It was insulated by winding a thin Kapton tape as a lap around the BISCO tape at Trithor [6]. The specification of the conductor is shown in Table 1 below. The dimensions of the conductor winding are given below in Table 2.

Table 1: Specification of HTS tape
Average thickness:	0.4 mm
Minimum width:	3.9 mm
Maximum width:	4.3 mm
Minimum bend diameter:	50 mm
Maximum tensile stress:	65 Mpa
Critical current at 77K	100 A
Critical current at 4K	300 A

Table 2: Dimensions of the Conductor Winding
	Inner Radius (mm) (cm)	Outer radius (mm)	Magnet length (mm)	No. of turns	Wire length (km)	Inductance (H)
Nb₃SnN	72	108.9	320	12232	6.93	29.4
NbTi	114	168.7	360	23699	21.052	91.3
HTS 1	25	43	100	590	0.112	8^-3
HTS 2	43	49	120	532	0.125	10^-3
HTS 3	575	65	140	576	0.221	10^-2

Each of the three windings is made as conventional layer winding and not as a series of pancake windings as this reduces the number of joints and power dissipation. Connections are made to each end of the HTS tape by NbTi conductors using soft solder. These joints are at a low field point at one end of the magnet. Figure 1 shows one of the coils being wound. After winding each coil is impregnated with resin.

Textové pole:

Figure 1: HTS tape winding

2.2 The cryostat

The important requirement for the cryostat was to have the helium boil-off from the magnet system kept to a minimum and preferably zero. The cryostat is designed to be recondensing and uses a Gifford-McMahon refrigerator with a base tempertaure of below 4 K. The first stage of the cryocooler is used to cool a single radiation shield surrounding the helium bath at about 50K. The 2^nd stage re-condenses gas evolving from the helium reservoir using a special design of heat exchange.

The cryocooler and compressor require 6.5kW for operation. When operating normally the cryostat has no boil-off and will in fact condense gas from room temperature so that the helium reservoir slowly fills over time. The magnet is supported inside the cryostat with a standard support structure from the top plate. The whole assembly can be lifted out of the cryostat so that the magnet can be inspected or modified. Four sets of current leads are provided so that the inner HTS and outer LTS coils can be energised seperately. A lambda plate has been built into the cryostat to allow the magnet to be operated at 2.2K but it has not been used. A drawing is shown in Fig 2.

When running the magnet power supplies an additional 3kW is drawn, giving a total power consumption of less than 10 kW. This compares extremely favourably with the 10 mW power requirements of conventional resistive high field magnets.

Figure 2: Drawing of zero-boil off cryostat housing the magnet.

3. Test results

The NbSn / NbTi magnet was built and tested first together with the zero boil off cryostat. The cryostat, which has been installed at Cambridge [7], has proved very successful and without current in the magnet leads, it slowly condensed helium from room temperature at a rate of nearly 100cc per hour.

The outer LTS magnet has been tested at 4.2K to 14 Tesla at which point stress induced training quenches were observed.

The inner HTS coils were first seperately tested in liquid helium at 4K. The coils show good performance with low resistive losses up to about 300 Amps. All three coils have some losses which are believed to be due to the resistance of the joints between sections or to the NbTi connecting cables. The effective resistance appears to be about 1.5 µOhm on each coil as can be seen from the test results show of Figure 3.

Figure 3: Inner HTS coil test.

The rather low engineering current density achieved in these HTS windings limits the field available from the combined magnet to 18 to 19 Tesla.

4. conclusions

While the field achieved today by this magnet is not higher than can be achieved with LTS conductors at 4.2K it does demonstrate the practicality of the approach. Furthermore new 2nd generation YBCO conductors with up to four times the current density are now becoming available and with these conductors a field of 25 Tesla appears practical for this magnet provided the windings can be designed in such a way as to support the forces involved.

A magnet of 25 Tesla which is all superconducting will have unique advantages compared to the hybrid alternative. Firstly, both capital and running costs are very much lower. Secondly, the field is more stable and quieter. It will also be possible to make the magnet of high homogeneity more easily. Ideally, a true persistant coil would be made however this may prove difficult due to flux creep in the HTS material as well as the difficulty of making superconducting joints to HTS materials.

5. references

1] Zhehong Gan, Hyung-Tae Kwak, Mark Bird, Timoth Cross, Peter Gor’kov., wiiliam Brey, Kiran Shetty. High Field NMR using resistive and hybrid magnets, High Field NMR resistive and hybrid magnets

2]. American Superconductors Corportation, 64 Jackson Road Devens MA 01434

3] Sumitomo Heavy Industries Ltd, 2-1-1, Yato-cho,Nichitokyo-city, Tokyo, 188-8585

4] Maciej Chorowski, TTA Techtra Sp. z o.o.ul. Muchoborska 18, PL 54-424 Wroclaw, Poland

5] Wolfgang Hassler, IFW, Institut für Festkörper- und Werkstofforschung Dresden e. V., Helmholtzstr. 20, 01069 Dresden, Germany

6] Jan Wiezoreck, Trithor GmbH Heisenbergstr. 16, D - 53359 Rheinbach,

7] Archie Campbell, Tim Coombs, Univerysity of Cambridge From, Dept. of Engineering, Trumpington St., Cambridge CB2 1PZ, UK

CR08-59

LIQUID DISTRIBUTION FROM STRUCTURED PACKINGS AND DISTRIBUTORS UNDER TILT AND MOTION RELEVANT TO FLOATING CRYOGENIC AIR SEPARATION PLANTS

Kalbassi M.A.¹, Waldie B.², White V.¹, Bell C.²

¹Air Products PLC, Surrey, UK
²Offshore Processing Research Group, Herriot-Watt University, Edinburgh, UK

ABSTRACT

Adaptation of land-based gas-to-liquid processes to floating production plants is being proposed as a means of recovering large offshore reserves of “stranded gas”. The cryogenic air separation plants which supply oxygen to the process need to cope with the tilt and motion conditions experienced on large production ships or barges. Results for liquid distribution under tilt conditions are presented and this experimental data is used to estimate separation efficiency and product oxygen composition under static, tilt and motion conditions.

Introduction

There is considerable activity worldwide on the development of offshore floating production systems for conversion of natural gas into liquid hydrocarbons. Some large reserves of gas, so called “stranded gas”, are too far from land for a pipeline to be economic. In that situation chemical conversion to liquids would reduce drastically the volume of hydrocarbon to be moved and allow use of shuttle tankers. Large-scale plants for such conversion are being used on land to produce high value liquid products including clean diesel fuels for which there is increasing demand. Fischer Tropsch processes for conversion of the intermediate syngas to liquids are a key part of these plants. The cryogenic air separation plants which supply oxygen to the gasification step of the Fischer Tropsch reactors need to be adapted to cope with the tilt and motion conditions experienced on ships and barges proposed for offshore plants. The studies reported here are part of an ongoing investigation into how tilt and motion affect liquid distribution and consequently mass transfer performance in the packed columns used in cryogenic oxygen production and other separation processes.

Previous reports on the effects of tilt and motion have been mostly on columns of up to 0.5 m diameter using water or aqueous solutions of similar surface tension. For example, columns of 0.22 and 0.40 m diameter were used by Tanner et al [1] in a comparison of the mass transfer performance of different packings subjected to tilt and motion. Results from detailed studies of liquid distribution in the 0.40 metre column were applied [2] in a parallel column technique to model the observed mass transfer performance. Columns for oxygen production need to be larger in diameter, around 4 metres, and provide many more theoretical stages than those used offshore for water de-aeration or regeneration of glycol. Maldistribution has an increasingly severe effect on separation performance as the number of theoretical stages increases [3]. In addition the extent of deflection or motion increases with column height. Another factor in modelling cryogenic separation columns is the low surface tension of the liquids, typically only 20% of that for water. Present studies are therefore based on a larger, 1 m diameter, column and use of a lower surface tension liquid as well as water. The results from these tests provide input for parallel column studies to predict the effect of the measured distribution on distillation performance.

Packing Studies: Experimental Techniques

Experiments were done on the 1 metre diameter by 5 metre high transparent column shown in Figure 1. Metal structured packings were studied, these being used in cryogenic air separation to minimise pressure drop. This paper reports experiments on 4 m (20 layers) of Sulzer 500Y. Consecutive layers were rotated by 90°. Metal foil wall wipers were fitted at each layer. The column is mounted on pivots to allow tilting or near sinusoidal motion by a mechanical drive with variable speed and amplitude. In the usual vertical orientation the column was within 0.2° of true vertical. A pressurized distributor was used to avoid tilt or motion affecting the initial liquid distribution. This was a ladder type giving 150 holes/m².

Detailed data on the distribution of liquid from the column were obtained with a multi cell collection and online cell flow measuring system. Outgoing liquid passed first through 580 cells, mainly 35mm by 35mm in size, arranged in a 27 by 27 row array located immediately under the packing. Liquid streams were then led through flexible PVC tubes (visible in Figure 1) to a set of flow measurement cells. These contained wire electrodes for measurement of rate of fill by conductance. Groups of sixteen collection cells are connected in sequence to the sixteen measuring cells fitted with rapid acting fill and drain valves. Rates of fill are measured via a fast response conductivity meter, multiplexer and PC. Further details of a smaller version are available [2]. Mains water and a surfactant solution with defoamer were used to study the effect of surface tension. Surface tension of the solution, 34 mN/m, was significantly lower than that of water though still higher than that of liquid oxygen and nitrogen. Safety, materials incompatibility and cost precluded other lower surface tension liquids. Foam suppressant was essential to avoid foam effects in the column and measuring system. Mean liquid fluxes were 2 and 4 l/m².s, typical of the low liquid fluxes in some parts of an oxygen column.

Results: Vertical and Tilt

The influence of tilt on liquid distribution can be shown most clearly by grouping the cells into 27 rows running at 90^o to the plane of tilt (Figure 2). This is also useful for subsequent parallel column modelling. In Figures 3 and 4 mean flux per row is plotted against row position across the column.

For the vertical column with 4 metres of 500Y, flux distribution across the column is more even with the lower surface tension solution than with water Figure 3. With tilt of 4° the surface tension effect is reversed, the flux distribution being worse for the low surface tension liquid than for water (Figure 4). Both liquids suffer considerable maldistribution with the uppermost wall region devoid of liquid. Near the lower wall, fluxes over twice the column mean occur with water and surfactant.


Figure 1: 1 metre Diameter Column in Motion	Figure 2: Chordal Subdivision of Plan Area for Parallel Column Modelling


Figure 3: *Flux distributions for water and surfactant from 4m of 500Y in vertical orientation*	Figure 4: *Flux distributions for water and surfactant from 4m of 500Y at 4° tilt*

The low surface tension of cryogenic liquids and some hydrocarbons is well recognised as an important factor in packed column performance but there are few published reports on the effect of surface tension on distributions from actual columns and none apparently at the present scale. More data is now available on fundamental aspects of liquid flow on single sheets of packing materials including the influence of surface tension and contact angle but there remains the substantial task of applying that to predicting the distribution from an actual column.

Another, more concise, way of comparing distributions is the fractional standard deviation parameter, FSD. This though it is not of use in subsequent modelling. FSD was applied by Reiss [4] to co-current flow distributions in packed columns and by Waldie [5] to compare packings under tilt.

where,

v = flux in given chordal slab [kmol/m2] a = area of slab [m2],

V = mean flux over whole column [kmol/m2] A = total column area [m2]

n = number of slabs

FSD = 0 for a perfectly even distribution

The influences of surface tension and column orientation on distribution are summarised in terms of FSD in Table 1. The trends shown by the graphs are confirmed quantitatively. Table 1 also shows how FSD is affected by the degree of subdivision of the distribution data. The finer the degree of subdivision the greater is the FSD.

Column conditions*

500Y

3x9 rows

500Y

9x3 rows

500Y

27x1 row

4m/2/W/V

0.0164

0.0239

0.0314

4m/2/SA/V

0.00519

0.0103

0.0148

4m/2/W/4º

0.1269

0.1450

0.1583

4m/2/SA/4º

0.3164

0.3441

0.3555

Table 1: Liquid distributions in terms of flow distribution parameter FSD

*Code example: 4m/2/W/V means 4m packed height/ 2 l/m²s / Water/ Vertical

Results: Motion

Fluxes to selected cells were measured continuously either in 16 cells connected to the measuring cells through flexible tubes or in a single cell from which liquid fell freely into a measuring cell which moved with the column. The latter gave the best resolution but could only be applied to one collection cell at a time.

Volume/time plots for 16 cells spaced along a line near the middle of the base parallel to the motion plane are shown in Figure 5. There is some evidence of cyclical variations in slope, hence flowrate, in cells nearer the wall. The possibility that flexing of the tubes contributed to these variations cannot be ruled out. Definite confirmation of visual evidence of variations in flowrate is given in Figures 6 and 7 from the single cell measurements. Fluctuations are more pronounced at the higher feed rate (4 l/s m² mean flux), probably due to a lower proportion of the liquid remaining attached to the surface of the moving packing.

Figure 5: Volume/Time Plots for 16 Cells over Single Cycle of ±3°at 35 sec motion period for 500Y with Surfactant


Figure 6: Time Dependant Flowrate in Cell B over Single Cycle of ±3°/35 secs , 2 l/s m²	Figure 7: Time Dependant Flowrate in Cell B over Single Cycle of ±3°/35 secs , 4 l/s m²

Distributor Studies

In the above packing study a pressurised distributor was used to ensure a constant initial distribution pattern independent of tilt or motion. On an actual plant a gravity distributor would avoid the need for a pump and thus be preferable if it was not affected too much by tilt and motion. Conventional and proprietary designs have been studied on both the 1 m column and on a larger computer controlled motion simulator in the Heriot Watt Offshore Processing pilot plant. A thin slice model technique has been developed which allows elements of gravity distributors up to 4.5 m diameter to be studied at accelerations similar to those at the top of a 40-50 m column subjected to motions expected on a large production ship. This has provided data for example on the relative significance of hydrostatic and dynamic heads during motion and the development of proprietary designs of gravity distributors [6].

Interpretation of Packing Distribution on Column Performance

The results of the tilt and motion studies can be used to determine the performance of a cryogenic air separation distillation column system under these conditions. Here we studied specifically the bottom section of packing of the low pressure (LP) column and the top section of packing of the high pressure (HP) column. The results show the deterioration in performance of the packing section due to the liquid maldistribution in the packing caused by the motion. This was carried out using Aspen Plus to simulate the distillation.

First, the nominal cases were simulated. Two sections were to be studied. The first, the bottom of the LP column, is shown in Figure 8. The L/V in the packing is 1.4 and the feed composition is 0.95 O₂ and 0.05 Ar, product is pure Oxygen. The second section to be studied, the top of the HP column, is shown in Figure 9. Here the L/V is 0.6 and the feed, stream 3, is air and the product is pure Nitrogen. In both cases 20 layers of Sulzer 500Y were used.

To study the effects of motion, this flowsheet was adapted by creating 8 parallel columns. The feeds to the column, liquid and vapour, were split between these 8 columns, the split fractions depending upon the results of the experimental motion studies. In the case of the LP column, the liquid from the bottom of the packing was then mixed into the reboiler. For the HP column, the vapour is mixed into the condenser. The 8 parallel column arrangement is shown for the LP column in the Aspen flowsheet in Figure 10.


Figure 8: Aspen Flowsheet of LP Packing Simulation	Figure 9: Aspen Flowsheet of HP Packing Simulation

Figure 10: Parallel Columns used for motion simulation

Only the results from experiments in which surfactant was used are analysed using parallel column analysis since these experiments more closely match the expected behaviour of a cryogenic liquid.

For the stationary column cases, where data is available for all 27 rows, shown in Figures 3 and 4, neighbouring cells were grouped and averaged to give 8 average cells across the diameter that were used to determine the liquid flows into the eight parallel columns. The vapour splits were then determined by the areas of the chords over which the averaging had taken place.

For the results with the column in motion 16 cells were used to collect the data in the experiment and these cells were almost across a diameter, Figure 5. Therefore, the eight flows were taken by averaging Cell No. 1 & 16, Cell No. 2 & 15, etc. This gives 8 flows and dividing each by the total gives the liquid split fractions required.

Using the above calculated split fractions, the simulation was performed to determine the performance. In order to determine the effective efficiency of the packing under these conditions, we return to the Aspen flowsheet with the single column and performed a simulation to determine the section efficiency required to obtain the separation achieved with the motion results applied to the parallel columns. These results can be found in Table 2.

The section efficiency of the stationary packing is also surprisingly low. Note that the two column sections investigated are not pinched. The efficiency of the LP column simulations works out greater than that of the HP columns. Note that stationary data shows packing inherent tendency to maldistribute and thus reduced packing efficiencies shown in Table 2.

Packing	Column	Ideal	Stationary	Motion	4 degree tilt
500 Y	LP	100%	71.6 %	22.1 %	12.4 %
500 Y	HP	100%	45.4 %	19.1 %	11.1 %

Table 2: Separation Efficiency

Conclusions

Reducing liquid surface tension to about half that of water reduces maldistribution from 500Y structured packings in a vertical column but increases it when the column is tilted by 4º.

Tilt of 4º causes significant maldistribution from 500Y structured packings at a mean column flux of 2 l/sm². Other studies though on the same column have shown that these packings suffer less maldistribution than random packings under tilt.

On the very large ships now proposed for gas to liquids schemes tilt is not expected to exceed 1° to perhaps 2°. Present data therefore indicates that it will still have to be taken into account in column design especially as there is a further decrease in surface tension with cryogenic liquid.

With the column moving ±3° at 35sec period, cyclical variations in local flux occur, at least near the wall. These are more pronounced at a higher mean column flux.

Even though 20 layers of Sulzer 500Y has been shown to give low performance in a shipboard scenario, the techniques reported in this paper have been demonstrated to provide a method to compare the performance of different proprietary types of packing [7] or distributors in a shipboard scenario and relate this to the expected performance of the distillation column in which the packing is employed.

References

1. Tanner, R.K., Baker, S.A. and Waldie, B., 1992, Proc.Distillation & Absorption 92. I.Chem.E. Symp. Series No 128, B.111- B.118.

2. Tanner, R.K., Baker, S.A., Millar, M.K. and Waldie, B., 1996, Trans.I.Chem.E.74A.177-182

3. Billingham, J.F. and Lockett, M.J., Trans I ChemE 2002, 80A , 373-382

4. Reiss, L.P., 1967, Ind.Eng.Chem.Proc.Des.Dev. 6, 486

5. Waldie, B., 2002, Proc.Gas Proc.Assⁿ Europe Annual Conference, Rome

6. Kalbassi M.A and Zone,I.R. U S Patent 6,907,751 B2 ,June 21 2005

7. Armstrong P.A, Kalbassi, M.A, Miller D, US Patent 5, 984,282, Nov 16 1999

CR08-43

COMPLEX SEPARATION OF MULTICOMPONENT FLOWS TO EXTRACT INDUSTRIAL AND INERT GASES

Bondarenko V. L.¹, Losyakov N. P.², Simonenko O. Yu.²

¹Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,
107005, Moscow, Russia
²Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine

ABSTRACT

The paper analyzes the composition of the waste flows, which appear in ammonia production. The ways of the inflow of the inert gases into ammonia synthesis circuit have been shown. The potential volumes of helium, neon, argon, krypton and xenon, which can be extracted in chemical industry, have been calculated. The preferable sequence of the multicomponent mixtures processing has been reasoned. The conditions resulting in acceptable degrees of extraction and the given quality or rare gases and accompaniments have been detected.

INTRODUCTION

The rare gases volumes and applications are continually increasing. A significant part of them is extracted from the waste products of the oxygen production at the steel mills. In spite of the developed metallurgical industry in Ukraine and Russia, the raw sources of rare gases have been practically exhausted. Therefore, there is an acute task of finding alternative sources of raw mixtures. The chemical industry plants can be considered as such. In the technological cycle of ammonia production there are considerable quantities of waste gas containing all the range of inert gases. Separating these flows into individual components is possible by means of sequential separation at cryogenic temperatures.

COMPLEX TECHNOLOGY OF WASTE FLOW SEPARATION

To get hydrogen, needed in ammonia production the process of natural gas conversion is used. This reaction requires huge volumes of oxygen, which is the basis of atmospheric air. Together with the air flow considerable amounts of inert admixtures get into the synthesis device. Because they do not take part in the reaction, they are accumulated in the NH₃synthesis circuit and they are discharged into the atmosphere as a waste flow. This mixture, besides the rare gases, contains valuable products: methane and components of singas – nitrogen and hydrogen, as well as the product of synthesis – ammonia, which all, as by products of the separation of rare gases can be returned back to the synthesis process. Typical composition of by-product gas mixtures is shown in Table 1 on the example of two biggest enterprises of Ukraine. The presence of several substances in the waste flow allows, prima facie, many variants of separation. However, considering the physical properties of separate components and a number of technological limitations, a certain sequence of separation processes can be observed in the industry. Preferable variants of waste flow separation are shown in Figure 1. The width of lines, characterizing separate products on the diagram, corresponds to their content in the mixture. All the complex of technologies can conventionally be split into several processes: I – purification from ammonia; II – preliminary cryogenic separation of the mixture; III – recovery of “high-boiling” products; IV – hydrogen-helium mixture processing.

ENTERPRISE

NH₃

Xe+Kr

CH₄

N₂

H₂

«Azot» enterprise, Severodonetsk, Ukraine

1,7

<0,001

13,0

5,3

20,7

0,01

0,3

Priportovy Zavod,

Odessa, Ukraine

8,6

5,6

0,01

64,4

0,4

Table 1: Volume content of the components in the waste flows, %

Figure 1: The preferable sequence of complex waste flow processing

The necessity of the preliminary mixture separation on stage II into “light” and “heavy” fractions allows relieving circuit III from high hydrogen consumption, making 2/3 from the end product volume. Besides helium, hydrogen also contains traces of neon. The need of recovery of the trace content of Ne and He in the source flow on stage II makes it necessary to use the rectification column. Using simpler means as preliminary separators (phase or reflux condensers) is unacceptable. This leads to the light inert gases dissolving in the liquid nitrogen-methane fraction, which is equivalent to their loss. The other, not less important function of column II, is the partial decrease of the concentration of high-boiling admixtures (N₂) in its gas fraction, consisting mainly of nitrogen. As follows from Figure 2-a, the effective means of hydrogen-helium flow enrichment is lowering the temperature in the column II condenser (Figure 3). It is rational to additionally cool the mixture (Н₂-N₂-He) and lower the nitrogen concentration in it in a separate reflux condenser RC1. The autonomy of this device allows applying alternative variants of cooling. As such, nitrogen boiling at a reduced pressure (Р = 0,02 MPa, Т = 66 К) or the upstream of gaseous hydrogen (Т = 24 К) after separation in the circuit (VI-b, Figure 1) can be used.

As it is shown in the diagram (Figure 2-a), due to reducing the phase equilibrium temperature in RC1 from 83 to 64 K it is possible to condense and return to the column up to 70% of high-boiling components (mainly nitrogen). This leads to considerable reduction of loading on the final purification stage IV-a. Further cooling and enrichment of the flow during the continuous operation is difficult because of the danger of freezing of N₂, contained by the hydrogen fraction at Т £ 63,15 К. Total purification of hydrogen-helium mixture is achieved in the devices with periodically working process. This can be realized by means of adsorption at Т = 64 К or freezing at Т = 40 К. The research showed that the second variant is preferable because it allows achieving the same result at 45% less energy consumption. Reduction of operating costs of the cryogenic purification of the H₂-He mixture from nitrogen is conditioned by the narrow “corridor” of working temperatures in the freezer. Unlike the adsorbers, in this device, it is enough to raise the temperature from 40 to » 65 К for the admixtures disposal.

Figure 2: Phase equilibrium isotherms of the systems Н₂-N₂ (а) and Не-Н₂ (b) in the vapor phase.

1^V-2^V-3 - enrichment and purification of Н₂-fraction

in the additional reflux condenser RC1 and the freezer Fr (Figure 3);

6^V-12¢-12² - helium concentration process under different phase equilibrium conditions

Creating the circuit IV-b for H₂-He separation was partially based on the same physical principles and technological methods that are inherent to the process of hydrogen fraction purification from N₂ considered above (circuit IV-а, Figure 1). In spite of different temperature levels (64 К for the mixture Н₂-N₂ and 15…16 К for Не-Н₂), these problems have a common solution – creating favorable phase equilibrium conditions outside the respective rectification columns. This fact is illustrated by the phase equilibrium isotherms on Figure 2-b and it is proved by the closeness of schematic solutions of the circuits IV-a and IV-b (Figure 3).

Raw helium flow in point “12” is two order smaller than at the entry into HC (point “3”). This allows processing of efficient enrichment and final helium purification by use of helium refrigerators.

The specifics of the circuit III, in which pure argon and Kr-Xe concentrate from the methane basis are obtained, is introducing the additional column KC into the typical scheme (Figure 4). As the preliminary research showed, krypton enrichment in the cube of column MC cannot achieved within this step even in the case of liquid phase extraction on the “L” line. It is explained by the fact that the СН₄-Kr system has a rather low relative volatility a<1,6. Such coefficient is about 20 times smaller than the one of the extensivelly studied Kr-О₂ mixture, from which krypton-xenon concentrate is extracted in the air-separation plants.

The preliminary enrichment of krypton concentrate can also be achieved by the sorption method. This method, besides the concentration, also allows substituting methane for nitrogen. The reduction of СН₄ content from 95…99% to several percent simplifies the further nitrogen-krypton flow processing and obtaining Kr and Xe in pure form.

Figure 3: The scheme of H₂-He concentrate extraction and the flows parameters in characteristic points: SPS - system of preliminary separation; RC1 - N₂ reflux condenser; F - N₂ freezer (one of the two sections is shown); HC - hydrogen column; RC2 - H₂ reflux condenser; HC - helium column; HR - helium refrigerator; NA - neon adsorber (the circuits markings correspond to Figure )

№	Р, MPa	Т, K	Volume content, %
№	Р, MPa	Т, K	He	H₂	N₂
1^V	3,5	83	0,4	88,6	11,0
1^L	3,5	83	-	10	90
2^V	3,5	64	0,44	97,7	1,9
2^L	3,5	64	-	9	91
3	3,5	40	0,45	99,55	<0,0001
4	3,5	33	0,45	99,55	<0,0001
5	1,0	31	0,45	99,55	<0,0001
6^V	1,0	26	45	55	<0,0001
6^L	1,0	26	1,7	98,3	<0,0001
7	1,0	31	0,001	99,999	<0,0001
8	0,25	24	0,001	99,999	<0,0001
9	0,25	24	0,001	99,999	<0,0001
10	0,25	60	0,001	99,999	<0,0001
11	0,25	78	0,001	99,999	<0,0001
12	0,25	23	70	30	<0,0001

Figure 4: The preferable sequence of waste flow complex processing.

SPS - system of preliminary separation; MC - methane column; AC - argon column; KC - krypton column

CONCLUSION

1. The potential of chemical industry producing helium and argon are commensurable with the productivity of metallurgic industry oxygen plants.

2. The waste flows complex processing is possible only on basis of cryogenic separation methods.

3. There is a rather definite technological sequence of standard multi-component mixture processing. It includes ammonia redemption, gaseous fraction Н₂-Не separation in a column; purification of this fraction from nitrogen and further separation into hydrogen and helium.

4. Liquid fraction processing on the preliminary separation stage (II) takes place at least in two subsequent columns and is accompanied by the output of argon, N₂ and СН₄ product flows.

5. The methane fraction is the raw product for krypton and xenon extraction.

6. It is rational to extract rare gases in pure form in separate devices, not connected with the waste flow separation complex.

REFERENCES

1. Bondarenko V. L., Simonenko Yu. M., The method of xenon separation, (Variants) and installation for its realization. Patent of Russia №2134387. B.I.-22 (1999).

2. Arkharov A. M., Bondarenko V. L., Losyakov N. P. and all, A unit for the extraction of the krypton-xenon mixture from off-gases at the ammonia production. Proc. 6 Int. Conf. Cryogenics’2000, Praha (2000) 122-125.

CR08-41

Solubility of PROPANE AND ETHANE in liquid oxygen

Houssin-Agbomson D.¹, Arpentinier P.¹, Delcorso F.¹, Coquelet C.², Richon D.²

¹ Centre de Recherche Claude-Delorme Air Liquide, Jouy-en-Josas, France
² Laboratoire CEP/TEP Mines Paris, Fontainebleau, France

ABSTRACT

Industry is large consumer of air gases for many and varied applications. Among the various processes of separation of air components, the most employed remains fractional distillation at low temperatures. The presence of pollutants – like hydrocarbons – in the feed atmospheric air of air distillation units (ASU) can be at the origin of drastic dysfunctions. That is the reason why a more accurate knowledge of the solubility of hydrocarbons in liquid oxygen and of the thermodynamic behaviour of these flammable systems under process operating conditions (from 93 to 153 K) would improve both evaluation and control of the risks specific to ASU and their performances.

Introduction

Air gases, mainly oxygen, nitrogen, argon are essential for metallurgy, chemistry, petrochemistry, refining, energy, electronics, health… The fractional distillation which operates at cryogenic conditions is the most employed process. The feed atmospheric air must be cleaned, before liquefaction, by removing all components being potentially obstructive at low temperatures. In particular it is necessary to take special care of carbon dioxide and water, and of secondary pollutants that are either natural or produced by the various anthropic activities (industry, heating, road traffic…), like hydrocarbons (ethane, propane or ethylene), or ozone and nitrogen protoxide. The presence of pollutants in the feed of air distillation units can be at the origin of their drastic dysfunctions. Specifically, hydrocarbons can form highly flammable mixtures with oxygen [1]. The risk is controlled today through several means, which allow operating air distillation units in effective and safe way. However these aspects can be improved by a better knowledge of physical properties of air pollutants. Concerning “hydrocarbon-oxygen” binary systems scientific literature presents only few data, probably because of the danger with handling of such mixtures in laboratories. In order to be able to build the required database, Air Liquide and CEP/TEP Laboratory have designed, built and set up new experimental equipment allowing to work under safe conditions. The study of the solubility of propane in liquid oxygen was the really interesting first subject of investigation using new installation. Nevertheless, to validate equipment and procedure before working on propane-oxygen system, investigations began with the study of the non hazardous mixture: propane-nitrogen. And to improve our knowledge on “hydrocarbon-oxygen” systems behaviour, the experiments have been extended with the study of ethane-oxygen binary system.

1. Industrial context

Main steps of air distillation process are presented in the Figure 1 (A): compression of air, elimination of pollutants, generation and transfer of cold, compression of the products [2]. The main element of the process is the double fractionating column (see Figure 1 (B)), composed by the medium pressure column (MP-column: 0.5 to 0.6 MPa) and the low pressure column (LP-column: about 0.15 MPa). Compressed air is fed into the medium pressure column.

The two columns are connected by a reboiler-condenser which ensures at the same time the heating of the LP-column, the vaporization of oxygen and the backward flow of each column by condensation of nitrogen. In this reboiler-condenser, heated and vaporized liquid oxygen (by nitrogen, which condenses) contains dissolved impurities: nitrogen protoxide (N₂O), carbon dioxide (CO₂), hydrocarbons (C₂, C₃…).

Figure 1: (A) Unit operations involved in air distillation process.
(B) Conventional double column apparatus for air distillation [1].

However, taking into account the extent of the air flows, impurities, even if they are present only at very low amount in the air feed, can under particular conditions, in spite of the careful air cleaning steps, accumulate over long period of time in liquid oxygen to reach considerable contents. According to the operating conditions and technologies of vaporization used, they will settle then in a solid state or will form a second liquid phase (in addition to the oxygen rich phase). For industry, it is of primary importance to control the formation of these phases, solid or liquid, rich in impurities in order to maintain, on one hand, transfer efficiency by avoiding the clogging of the reboiler-condenser and, on the other hand a satisfactory safety level by minimizing ignition risk of hydrocarbons with oxygen. That is why, air processors have developed equipment (liquid oxygen filters), materials (new adsorbents for the air cleaning step) and strategies and procedures for check, analysis and follow-up of the impurities behaviour from feed air to the reboiler-condenser. The principal quantities whose knowledge is fundamental to control the formation of undesirable phases are: the solubility of the impurity in liquid oxygen, the partial pressure of the impurity in gaseous oxygen and the molar fraction of vaporized liquid oxygen.
Thus the solubility of hydrocarbons in liquid oxygen appears as a key variable. In fact its knowledge as a function of pressure and temperature will allow to predict the formation (or not) of a hydrocarbon rich second liquid or solid phase. Among hydrocarbons present in atmospheric air, propane is potentially one of the most critical to control taken into account the efficiency of the front-hand purification step. Unfortunately, up to now, its solubility was badly-known under industrial operating conditions.

2. State of the art

The study of the literature concerning the solubility of hydrocarbons in liquid oxygen shows that only few references are available on the subject: Karwat (1958) [3] and McKinley and Wang (1960) [4] for propane; Tsin (1940) [5], Cox and de Vries (1950) [6], McKinley and Wang [4], Amamchian et al. (1973) [7] and Bulanin (1973) [8] for ethylene; Cox and de Vries [6], Karwat [3] and McKinley and Wang [4] for ethane. Furthermore, experimental procedures are not sufficiently detailed in these references to make it possible to evaluate uncertainties on the measured values. In particular, few values of solubility in liquid oxygen are published at 90 K: 10 000 ppm [3] and 50 000 ppm [4] for propane, and 78 000 ppm [6], 128 000 ppm [3] and 215 000 ppm [4] for ethane. Moreover, temperature ranges are relatively limited (from 77 to 90 K) and do not cover the exploitation field of air distillation units. Experimental techniques used for the determination of the thermodynamic equilibria properties, are generally classified according to the method of equilibrium creation (static methods, dynamic methods) and according to the technique of determination of the compositions (synthetic methods, analytic methods) [9]. Measurements of solubility are based on “static-synthetic” or “static-analytic” methods.

3. Experimental aspects

3.1 Experimental method

The technique retained for this study is a “static-analytic” method with sampling of phases by ROLSI™ samplers (Rapid On-Line Sampler-Injector) [10], followed by gas chromatography analyses. This technique is based on a method described by Laugier and Richon [11]. The solute is introduced into the equilibrium cell and then it is diluted with the solvent. Once the compounds loaded, the cell is comparable to a batch reactor. Equilibrium is reached in a static way.

3.2 Description of the equipment

A 12-cm³ Hastelloy C276 cell is fixed inside a cryostat partially filled with liquid nitrogen. The cryostat used on this apparatus is a 55-dm³ double envelope vessel (L’Air Liquide GT55 model). The stability of the temperature, ± 0.05 K, is achieved using a heating resistance cable rolled up around a brass housing containing the equilibrium cell and connected to a proportional-integral-differential (PID) thermal regulator (West Mini 6100 model). The rounded heating resistance acts as the hot source and the vapour of liquid nitrogen as the cold one. Two ROLSI™ pneumatic samplers fitted on the top of the equilibrium cell (one for the liquid phase, the other for the vapour phase) and connected to a gas chromatograph (Varian 3800 model), allow direct injection of liquid and vapour phase samples into the carrier gas circuit of the gas chromatograph. Two pressure transducers measure the total pressure inside the equilibrium cell: a 0-1 MPa transducer (Drück PTX610) for low pressures and a 0-10 MPa transducer (Drück PTX611) for higher ones. After calibration, accuracies of pressure measurements are estimated to be better than ± 0.07 kPa for the low pressure transducer and ± 0.18 kPa for the other. Temperatures are measured by two four-wire 100-Ω platinum probes introduced inside wells managed in the walls of the equilibrium cell body (one at the top, another one at the bottom). Uncertainties on temperatures are lower than 0.02 K on the (100-140) K range. Once the system has reached equilibrium (i.e. P and T are constant), the analysis of the samples taken by the ROLSI™ samplers is carried out thanks to a Varian gas chromatograph (model 3800) equipped with two types of detectors fitted in series: a thermal conductivity detector (TCD) and a flame ionisation detector (FID). The TCD is used for the detection of oxygen, nitrogen and high quantities of propane, while the FID, more sensitive, allows detecting the very small quantities of propane. TCD and FID were repeatedly calibrated by introducing known amounts of each pure compound through a syringe into the injector of the gas chromatograph. Chromatograph calibrations realized for this type of mixtures lead to relative uncertainties around 1 % for each component.

3.3 Equipment and procedure validation

In order to validate our experimental equipment, nitrogen-oxygen system was studied at 110 K and the experimental results obtained have been compared to those of Baba-Ahmed et al. [12]. Baba-Ahmed used a Φ-Φ thermodynamic approach and adjusted the binary interaction parameter k_ij of the Soave-Redlich-Kwong equation of state [13, 14] on its experimental values (temperature range: 100 to 123 K) measured with similar equipment and procedure, but in an equilibrium cell of a volume 3.5 times larger than ours. The standard mixing rules [14], and the Mathias-Copeman alpha function [15] were used for this calculation. This adjustment (k_ij = -1.58.10^-2), using an “objective” function based on the total pressure and the nitrogen vapour phase composition, led to relative average deviations on these variables of respectively ± 0.6 % and ± 1.5 %. Average deviations obtained between our measured values and Baba-Ahmed computed values [12] concerning the vapour phase composition and the total pressure are about ± 0.5 %. These deviations, lower than the uncertainty of Baba-Ahmed adjustment, validate the experimental protocol and show that our equilibrium cell, in spite of its low volume, is still well adapted to the study of thermodynamic equilibria under cryogenic conditions.

4. Results and discussion

In this section, validation of the procedure on a similar but not flammable system is demonstrated by measurements performed on liquid compositions of the propane-nitrogen system under various experimental conditions (propane low and high composition areas). Then the study of propane solubility in liquid oxygen is presented with that of the ethane-oxygen binary system.

4.1 Propane-nitrogen binary system

Experimental values obtained between 109.98 and 125.63 K for the propane-nitrogen system at thermodynamic equilibrium allowed, according to a Φ-Φ approach, to adjust parameters of a thermodynamic model based on the Peng-Robinson equation of state (PR-EoS) [16] using the Mathias-Copeman alpha function and the Huron-Vidal mixing rules [17] coupled with NRTL activity coefficient model [18].

Figure 2: (A) Pressure-composition diagram for propane-nitrogen system at 109.98 K. (B) Zoom on diagram close to nitrogen vapour pressure. (¡): experimental data from this work; solid lines: calculated VLE on experimental data from this work and LLE prediction with PR-EoS.

Thanks to this work, four (P, x)-isothermal curves have been studied and modelling results have been compared to scientific literature (Cheung and Wang (1964) [19], Poon and Lu (1974) [20], Kremer and Knapp (1983) [21], Llave et al. (1985) [22]): good agreement is observed with Kremer and Knapp data while sets of data present systematic deviations with Poon and Lu data and Llave et al. ones that are consequently judged as suspicious [23]. Calculated complete phase diagrams (see Figure 2) show that they are of type III according to the van Konynenburg and Scott classification [24]: this system presents “vapour-liquid-liquid” phase separation.

4.2 Propane-oxygen binary system

Before determining solubility values with the apparatus previously described, another experimental device (fairly simple) has been set up in order to visualize propane-oxygen system at thermodynamic equilibrium (see Figure 3) and have more information on its behaviour in cryogenic conditions.

Čárový popisek 3 (se zvýrazněním): Vapour phase

Figure 3: Photography of the “vapour-liquid-liquid” equilibrium
for propane-oxygen system in cryogenic conditions at atmospheric pressure.

Thanks to this first experiment, some assumptions have been made and finally experimental protocol was modified in several times in order to better control the propane loading and succeed to determine propane solubility. These modifications, so simple in fact, have significantly contributed to improve the operability of the apparatus, and both the reliability and the repeatability of measurements.

Figure 4: Propane solubility in liquid oxygen as a function of temperature.
(¿) this work, (£) Karwat [3], (r) McKinley and Wang [4].

Data measurements have been performed in propane high and low composition conditions and in the “vapour-liquid-liquid” area allowing the determination of the propane solubility in the oxygen rich phase. Consequently, solubilities of propane in liquid oxygen have been determined at 110.22 and 120.13 K. These experimental values are presented in Figure 4, for comparison to literature data. At 110.22 K, the value of propane solubility in liquid oxygen is x_C3 = 0.0224 ± 0.009 for a total pressure of 0.536 ± 0.001 MPa. At 120.13 K, the solubility is higher and reaches a x_C3 value of 0.0570 ± 0.009 for a total pressure of 0.979 ± 0.001 MPa. Literature data presented in Figure 4 (Karwat [3], McKinley and Wang [4]) are the only available ones for the propane-oxygen binary system up to now. Figure 4 shows that the solubility value measured by McKinley at 90 K is not consistent with values of this work obtained at 110.22 and 120.13 K and not with that of Karwat at 90 K. Data from scientific literature were obtained with experimental techniques different from that developed in CEP/TEP Laboratory. McKinley and Wang [4], in particular, used a synthetic method coupled with a visual criterion of solubility detection. Unfortunately experimental technique and detection criterion of Karwat [3] are not clearly described. A detailed reading of these publications does not make it possible to conclude as for the repeatability and the accuracy of their corresponding measurements. It seems that the precautions taken in this work allow approaching solubility values with a more acceptable level of accuracy.

Figure 5: Pressure-composition diagram for propane-oxygen system at 110.22 and 120.13 K.
Solid lines: results of the modelling with adjustment on our experimental data;
L^I, L^II: respectively propane rich and lean liquid phases; V: vapour phase.

Based on our experimental data a modelling work was performed and complete phase diagrams have been calculated at two temperatures, as presented in Figure 5, with a particular adjusted thermodynamic model. These diagrams, as for propane-nitrogen system, are of type III according to the van Konynenburg and Scott classification [24].

4.3 Ethane-oxygen binary system

Data have been carried out at several temperatures and phase diagrams of ethane-oxygen system were calculated thanks to modelling after adjustment on these isothermal data. Only results obtained with ethane at 112.1 K are presented herein. At 112.1 K, ethane is completely soluble in liquid oxygen as shown on the pressure-composition diagram calculated displayed on Figure 6: no “vapour-liquid-liquid” equilibrium is observed, whatever liquid phase composition. Thus the phase diagram of ethane-oxygen system seems of type I according to the van Konynenburg and Scott classification [24].

Figure 6: Pressure-composition diagram for ethane-oxygen system at 112.1 K. Experimental data (¡) and result of the modelling with adjustment on our experimental data (solid lines);
L, V: respectively liquid and vapour phases.

Experiments and modelling are found in good agreement concerning the thermodynamic behaviour of ethane-oxygen system. Now it is important to emphasize that solubility values presented in scientific literature by authors at 90 K and at lower temperature, certainly are solubilities of solid ethane in liquid oxygen (note that melting temperature of pure ethane is about 90.4 K).

Conclusions

A new equipment was developed to study “hydrocarbon-oxygen” flammable systems and in particular propane-oxygen system. Preliminary studies, which were carried out on non dangerous systems for which solubility data are known or partially known (nitrogen-oxygen and propane-nitrogen systems), allowed to validate equipment and procedure.

Initially, composition measurements of the liquid phase, realized in propane low composition area, were dispersed according to propane loading rate. This was probably due to the properties of the two coexisting liquid phases (density, viscosity, interfacial tension, wettability) which are more influent with low quantities loaded in the equilibrium cell. Thus a great part of the study consisted in developing a loading protocol making possible the control of the quantity of propane injected into the cell in order to measure the propane solubility in the liquid phase with an acceptable level of accuracy. Propane solubilities in oxygen (mole fractions) determined during this study are: 0.0224 at 110.22 K and 0.0570 at 120.13 K. In comparison, the solubility value measured by McKinley and Wang at 90 K [4] is not consistent with values obtained in this work; this could be explained by the difference of experimental techniques or by the state of propane either a liquid or a solid close to pure propane melting point. However concerning Karwat value at 90 K [3], the agreement with our values can be considered as acceptable. Concerning ethane-oxygen system, experiments and modelling showed that ethane is completely soluble in liquid oxygen at 112.1 K.

Other data have been performed on few “hydrocarbon-oxygen” binary systems and will be presented in further communications (to be published). This work brings new information on thermodynamic behaviour of such systems (properties and phase envelopes), which will be used for accumulation phenomena interpretation highlighted in high capacity air distillation units.

References

1. Arpentinier, P., Cavani, F. and Trifirò, F., The Technology of Catalytic Oxidations, Tome 2: Safety Aspects, Editions Technip, Paris, France (2001) 471-670

2. Foerg, W., Refrigeration Science and Technology Proceedings, Munich, Germany (1996) 1-12

3. Karwat, E., Some Aspects of Hydrocarbons in Air Separation Plants, Chem. Eng. Prog. (1958) 54(10) 96-101

4. McKinley, C. and Wang, E. S. J., Hydrocarbon-Oxygen Systems Solubility, Adv. Cryog. Progress. (1960) 53 11-25

5. Tsin, N. M., Solubility of Ethylene and Propylene in Liquid Nitrogen and Liquid Oxygen, Zh. Fiz. Khim. (1940) 14(3) 418-421

6. Cox, A. L. and de Vries, T., The Solubility of Solid Ethane, Ethylene, and Propylene in Liquid Nitrogen and Oxygen, J. Phys. & Colloid. Chem. (1950) 54 665-670

7. Amamchian, R. G., Bertsev, V. V. and Bulanin, M. O., Infrared Spectral Analysis of Cryogenic Solutions, Zavodskaya Lab 4 (1973) 39(4) 432-434

8. Bulanin, M. O., Infrared Spectroscopy in Liquified Gases, J. Mol. Struct. (1973) 19 59-79

9. Coquelet, C., Etude des Fluides Frigorigènes. Mesures et Modélisations, PhD Thesis ENSMP, France (2003)

10. Guilbot, P., Valtz, A., Legendre, H. and Richon, D., Rapid On-Line Sampler-Injector. A Reliable Tool for HT-HP Sampling and On-Line Analysis, Analusis (2000) 28 426-431

11. Laugier, S. and Richon, D., New Apparatus to Perform Fast Determinations of Mixture Vapor-Liquid Equilibria up to 10 MPa and 423 K, Rev. Sci. Instrum. (1986) 57 469-472

12. Baba-Ahmed, A., Guilbot, P. and Richon, D., New Equipment Using a Static Analytic Method for the Study of Vapour-Liquid Equilibria at Temperatures down to 77 K, Fluid Phase Equilib. (1999) 166(2) 225-236

13. Redlich, O. and Kwong, J. N. S., On the Thermodynamics of Solutions. V. An Equation of State. Fugacities of Gaseous Solutions, Chem. Rev. (1949) 44 233-244

14. Soave, G., Equilibrium Constants for Modified Redlich-Kwong Equation of State, Chem. Eng. Sci. (1972) 4 1197-1203

15. Mathias, P. M. and Copeman, T. W., Extension of the Peng-Robinson Equation of State to Complex Mixtures: Evaluation of the Various Forms of the Local Composition Concept, Fluid Phase Equilib. (1983) 13 91-108

16. Peng, D. Y. and Robinson, D. B., A New Two-Constant Equation of State, Ind. Eng. Chem. Sci. (1976) 15 59-64

17. Huron, M. J. and Vidal, J., New Mixing Rules in Simple Equations of State for Representing Vapour-Liquid Equilibria of Strongly Non Ideal Mixtures, Fluid Phase Equilib. (1979) 3 255-271

18. Renon, H. and Prausnitz, J. M., Local Composition in Thermodynamic Excess Function for Liquid Mixtures, AIChE J. (1968) 14 135-144

19. Cheung, H. and Wang, D. I. J., Solubility of Volatile Gases in Hydrocarbon Solvents at Cryogenic Temperatures, Ind. Eng. Chem. Fundam. (1964) 3 355-361

20. Poon, D. P. L. and Lu, B. C. Y., Phase Equilibria for Systems Containing Nitrogen, Methane, and Propane, Adv. Cryog. Eng. (1974) 19 292-299

21. Kremer, H. and Knapp, H., Three-Phase Conditions during Gas Processing are Predictable, Hydrocarb. Process. (1983) 62 79-83

22. Llave, F. M., Luks, K. D. and Kohn, J. P., Three-Phase Liquid-Liquid-Vapor Equilibria in the Binary Systems Nitrogen + Ethane and Nitrogen + Propane, J. Chem. Eng. Data (1985) 30(4) 435-438

23. Houssin-Agbomson, D., Coquelet, C., Richon, D., Delcorso, F. and Arpentinier, P., Solubility of Hydrocarbons in Liquid Oxygen, 2007 AIChE annual meeting, Salt-Lake-City, Utah, USA (2007) 384c

24. Van Konynenburg, P. H. and Scott, R. L., Critical Lines and Phase Equilibria in Binary van der Waals Mixtures, Philos. Trans. R. Soc. (1980) 298 495-594

CR08-64

modeling heat-mass transfer Processes on regular PACKINGS of distilation plants

Arkharov I., Navasardyan E.

Bauman Moscow State Technical University, Moscow, Russian Federation

ABSTRACT

The paper presents the results of mathematical modeling heat-mass transfer processes on regular packing of two profiles – rectangular and parabolic types. The paper gives graphic illustrations of distribution of pressure, temperature and concentration in one cell for every type of packing. Modeling geometrical structure of the packing and heat-mass transfer processes is plotted in the program complex STAR-CD.

INTRODUCTION

Heat-mass transfer processes realized in column apparatuses are widely used in chemical, food, cryogenic, oil-processing and other industries. They are the base of many technological processes of separation, distillation, rectification of multicomponent mixtures, solutions and emulsions. So, their effectiveness is directly linked with technical and economic indexes of plants. It is evident that the intensification of heat-mass transfer processes in separation plants is one of the principle conditions of increasing efficiency and profitability of air separation plants (ASPs) as a whole. Creation of efficient heat-mass transfer apparatuses makes possible to increase thermodynamic efficiency of the ASPs, to decrease capital costs for their production as well as power inputs at exploitation, as a result to decrease cost price of oxygen, nitrogen, argon.

1. INCREASE OF AIR SEPARATION PLANT COLUMNS EFFICIENCY

Now the columns of 0.2 … 12,0 m diameter and 3,0 …70,0 m high at mass flowrate of raw material of 0.05 … 6,500 kg/s are widely applied in industrial plants. Before the trays of different type and model were used as a packing, many of which are used up to the present time. The rectification process in the tray column is stepped, it is characterized by high hydroresistance of a step, and the improvement of the process is limited. At present the performance of some types of trays is up to 80%, but power inputs indices of the ASPs are not decreased. New developments in this way are considerably curtailed; development of packing of other types is actively stimulated. Industrial packed columns for deuterium and tritium at 20K appeared in 70s of the last century. A continuous or film regime of rectification on packing is used in them. The packing is a contact surface formed by irregular fill of contact elements of different form or by a specially ordered structure of elements layers.

In the middle of 80s of the last century the Swiss company Sulzer Chemtech Ltd. has generalized successfully its own experience and know-how in the field of the packing in the chemical production, it elaborated and put on sale the packing Mellapack for air separation that is being used till now. Today the volume of world production of ASPs with packed columns reaches 45% of the total volume of ASPs production, and the tendency to increasing has been noticed.

The packed columns work at rather large range of pressure of 0.1 … 1.0 MPa and liquid loads of 5 … 20 m³/m²per hour, they have low hydro resistance (approximately 5-7 times less than a tray packing).

The upper columns (nitrogenous, low pressure), columns of raw and pure argon, columns of extraction and concentration of krypton–xenon mixture are packing apparatuses of modern ASP. Little by little some ASPs are equipped with lower (oxygen) packing columns. Today world producers of cryogenic heat-mass transfer equipment for apparatuses of average and high efficiency use a structure packet gofer packing. The creation of new type packing, perfection and renovation of production technology as well as determination of heat and mass transfer and exploitation characteristics inevitably is accompanied by a big volume of construction and technological developments. The experiment is still a decisive factor to solve this optimization task and select a type of packing. The interest to mathematical modeling heat-mass exchange processes in the layer of complex geometrical configuration of the packing is constantly growing.

Modeling and investigation of heat and mass exchange processes is done in several stages. At the first stage the problem of distribution of fluid and gas flows along the packing is resolved, dynamic retention is determined as well as hydrodynamic resistance of the packing. At the second stage on the base of the obtained result the process of mass transfer in each element of the packing taking into account inequalities of phases flow in it is modeling. Then the fields of a components concentration and their transfer in the volume of the packing are defined.

All the volume of the structure packing may be presented as some ordered in fixed manner combination of elementary cells or calculated elements. Such an approach allows applying a classic mathematical model of heat and mass transfer to the cells determined geometrically, and finding the solution of the system of hydrodynamic equations concerning the concrete cell. After calculation of the processes in a separate cell a layer of these cells is formed according to real dimensions of the separation section of the apparatus. Specifying initial values of concentration and desirable low of their distribution along the height we can define the necessary quantity of layers or the height of the packing.

2. MODELING GEOMETRY OF THE PACKING
AND HEAT-MASS TRANSFER IN IT

Let us examine a plate and parallel packing Mellpack 250 type without perforation. Its working surface is formed by gofer sheets at an angle of 45⁰ that lay in vertical position, for all this, even layers are turned at 180⁰ relative to odd layers on a vertical axis. Crossed sheets form a combination of cavities that are calculated elements or cells (Figure 1). A three-dimensional cell formed by adjacent plates of the packing is shown in Figure 2 (a), (b).

Coordinate axes coincide with the sides of the cell “y” and “z” in horizontal plane, “x” – in vertical plane; the origin of coordinates is at the top of the cell angle. A gravitational force is acting in the direction of the axis “x”.

When plotting a mathematical model let us assume the following assumptions:

-regime of flow of two phases is laminar and stationary;

-process of impulse, heat and component transfer is considered as steady;

-interaction of phases is one-sided;

-coefficients are homogenous and constant in the field of transfer;

-physical properties of two phases, mass flowrate and value of disturbing force along the axis “x” of the channel are adopted as constant;

-component in energetic balance is neglect. At the boundary of the section temperatures T_wall; T_i, T_vapor._¥; mass concentrations - W_1vapor.i, W_1vapor._¥ ; velocities of vapor and liquid U_vapor.i, U_liquid.i,U_vapor._¥ don’t depend on the coordinate “x”; here “i” is an index designating the parameters on the boundary of the film at y =.

The obtained systems of equations are solved by finite-difference methods. For example, by the method of finite elements. A finite-difference network consists of cubic or other type calculated cells (Figure 2, b). The type of calculated cells is chosen based on geometry of the cell. A recommended relation of network dimensions of the cell is 16x9x9, though decrease of the relation doesn’t influence much the precision of calculation.

Initial and boundary conditions:

Initial conditions

- for vapour

t = 0; 0£x£ ; 0£y£b; 0£z£d; (along “y”)

0£y£d; 0£z£a; (along “z”)

T = T₀, U_x = U_y = U_z = 0.

- for liquid

t = 0; 0£x£ ; d£y£b - d; d£z£а - d;

T = T₀, U_x = U_y = U_z = 0.

2.Boundary conditions (conditions of adhesion, solid impermeable walls):

t ≥ 0; x = 0, ; y = 0, b; z =0, a;

U_x = U_y = U_z = 0 for liquid and vapor

3.Heat boundary conditions: (adiabatic walls)

A convective movement is initiated by a constant gradient of temperature across a liquid layer:

T=T₁, at x=0 where T₁ – is warm side temperature, K;

T=T₂, at x=, at tá 0 where T₂ – is cold side temperature,K

y = 0		y ® ¥		y = d
	U_liquid_,_х = 0		U_vapor_{, y} = U_vapor._¥		a) U_{vapor. i}= U_{liquid. i} = U_i
	U_{liquid, y} = 0		T_vapor = T_vapor._¥		b) T_liquid._i = T_{vapor. i} = T_i
	T_liquid= T_wall		W_{1 vapor}= W_1vapor._¥;		c) W_{1 vapor}= W_{1 vapor. i}

Table 1: Boundary Coditions for vapour and liquid

Components and are neglect in energetic balance. The components of velocity are defined on the edges of calculated cells; the temperature and the pressure are determined in the center of the cells. To keep corresponding “transferable characteristic” it is recommended to use the form of finite-difference presentation of adequate members on “the cell-donor”. To apply necessary boundary conditions we assume that the liquid is surrounded by one layer of dummy cells. The difference scheme is explicit. The velocity distribution is calculated using the equations of momentum applying the values of the antecedent time pitch. Then a new distribution of velocity changes by circuit to perform the equation of mass conservation by means of changing pressures in calculated cells. After calculation of a new distribution of velocity it is used for calculation of a new distribution of the temperature applying the equation of energy. Then these new distributions of velocity, pressure, and temperature are used to define concentration, and as initial values they are used for calculation cycle at the next time pitch.

To illustrate this method the calculations of non-stationary heat-mass exchange processes in cells of several types using software “STAR-CD” and subprogram “Star-design” (to create a three-dimensional model of an elementary cell and a subprogram “Pro-amm” for generation of a 3-D finite-difference network from a modeled cell) were carried out.

The initial conditions are as follows: vapor – air (y_N2=0.79; y_O2=0.21); liquid – air (X_N2=0.79, X_O2=0.21), counterflow P_init=0.13 MPa, v_n=0.5 m/sec, T_init=83K. The stationary condition is modification of parameters at inlet no more than 0.1%. Figures 4 -7 show the distribution of velocity, pressure and concentration of O₂ in cells in steady process.

b)

а)

Figure 5. Field of velocities and pressure of vapor in the cell of rounds:

а) general view; b) middle section of the cell

Figure 6. Distribution of oxygen concentration in the cell of rounds

Figure 7. Distribution in the cell of triangular shape:

а) pressure change (minus corresponds to a zone of exhaustion);

b) vapor velocities relative to the liquid film

NOMENCLATURE

T- temperature [K]; d - thickness of a liquid film [mm]; D - thickness of a boundary vapor layer [mm]; n - kinematic viscosity [m²/s]; r - density [kg/m³]; K- heat exchange coefficient [W/m.K]; D – diffusion coefficient [m²/s]; ¥ - vapor region remote of the zone of the contact with a liquid film; W - mass concentration of the component; 1,2 – component.

CONCLUSIONS

It is clear from the given illustration that the field of velocities is more irregular. So, in the central region of the cells (Figure 4 b) a zone of slowdown of vapor is well seen, and a zone of acceleration is seen in nodal points. The zones of low and high pressure are distributed correspondently. But in laminar regime the field of concentration of one of the components (oxygen, Figure 6) practically has no features and its configuration is near to the field corresponding to the field in lengthy channel.

It is to note that the cited results of calculation according to the mathematical model obtained with software “STAR-CD” are our fist steps to apply modern software for modeling heat-mass transfer processes inside space structures at complex geometry.

REFERENCES

1.Mallinson G.D., De Vahl Davis G., Three-Dimensional Natural Convection in a Box: a Numerical Study, J.Fluid Mech. (1977) 83 1-31

2.Del Giudice S., Strada M., Comini G., Three-Dimensional Laminar Flow in Ducts, Numerical Heat Transfer (1981) 4 215-228

3.Chjan A.M.S., Benergi S., Numerical Simulation of Three-Dimensional cell Vortex in closed Cavities with Solid Impermeable Walls, Teploperedacha (1979) 101 2 52-57

CR08-42

OPERATION OF SMALL AND HIGH PRESSURE TANKS FOR LIQUEFIED AIR GASES

Hnízdil T., Suma J., Kouba M., Chrz V.

Chart Ferox, a.s., Ústecká 30, Děčín 5, 405 30, Czech Republic

ABSTRACT

Small vacuum storage tanks (up to 11 m³ capacity) are generally used for applications, where, for logistics optimization, the re-filling periods are longer, typically two-three weeks. With long retention time the bulk of liquid becomes warmer and doesn’t accept so much heat leak, which is transformed more into vaporization, then. In high pressure tanks with relief pressure set to 37 bar and operating pressure close to the critical pressure of nitrogen at 33 bar, as encountered at laser applications, e.g., the heat of vaporization is very low. Any heat leak causes large vaporization and holding times are very short. Minimization of heat leak is very important, then. Holding time and nominal evaporation rate was improved by using of multi-layer insulation at the new multi-layer insulated EVT tanks in sizes 3, 6 and 11 m3,. Measurement of temperature distribution in a cryogenic pallet tank 600 liters at high operating pressure proves warming up of the bulk of liquid. Recommendations for prolongation of the holding time of high pressure tanks are presented.

INTRODUCTION

Storage tanks used for storing of cryogenic liquefied gases are double-wall vacuum insulated vessels. Annular space has to be filled by special insulation material for to reduce rest gas convection and thermal radiation. This way, heat leak is minimized and optimum holding time achieved.

The outer jacket is only a container for the inner vessel and its thermal insulation. Its main purpose is to ensure a high vacuum in the insulation space between both the inner vessel and the outer jacket to provide good thermal insulation of the inner vessel.

Despite the good quality of insulation of cryogenic tanks, there is still a continuous heat flow (heat leak) from the environment to the fluid in the inner vessel, which increases the temperature of the fluid, causes partial evaporation and increases the pressure in the inner vessel. The withdrawal of fluid reduces the pressure in the inner vessel as a result of expansion of the vapor space. At best, thermal energy intake and fluid withdrawal rates are in balance. If the liquid withdrawal is too low so that it cannot compensate for the increasing quantity of evaporated gas, the pressure is growing until the pressure relief system set pressure is achieved; the relief system opens and releases gas from the tank from time to time to avoid over-pressurizing. In the majority of tanks the liquid withdrawal is larger and heat leak doesn’t compensate for the expansion of the vapor space. To prevent the pressure from falling below the required operation value, a pressure build up vaporizer (PBU) is installed at each tank. The pressure build-up regulator allows liquid from the bottom of the tank to flow into the PBU if the operating pressure falls below a set point. Evaporated gas is returned into the upper part of the tank, which results in a pressure increase, until the upper pressure set point is achieved. The lower and upper set points of the regulator are close to each other with the pressure difference 0,5 bar and far below the set relief pressure. This makes possible accumulation of vapor during interruption of liquid withdrawal.

Figure 1. Flow diagram of cryogenic storage tanks for air gases

The number of ordered and delivered HP tanks has greatly increased in the last few years. It is closely related to increased number of laser application requiring nitrogen at higher operation pressures around 30 bar. A few years ago the laser cutting equipment was very expensive, which was driving continuous operation of these tanks with no time for increase of pressure. Today, more and more lasers work at small companies under part-time operation. For laser cutting operation the tank operating pressure should be between 30 and 33 barg. In such cases, minimization of heat leak is very important. Regular shut-down – Friday evening to Monday morning (app. 60 hrs) can cause pressurization to the relief valve set pressure 37 bar. In case of shut-down throughout the weekend, these tanks use to pressurize up to either set pressure of the tank relief valves or set pressure of the boil-of regulator conected to the consumption net . When the quick pressure increase occurs, the typical first reaction of the end-user is that he is losing gas and the first impression is that the tank is faulty, because other tanks of the same type are working well.

1. SPECIFICS OF SMALL HIGH-PRESSURE TANKS

Holding time of a cryogenic tank is a time from a defined initial state of liquid up to achieving the relief valve set pressure due to heat leak and opening the relief valve of the tank. This situation is not desirable because of losses of the gas. At high pressure (HP) tanks with relief pressure set to 37 bar (3.7 MPa) and under operating pressure close to critical pressure of nitrogen at 33.97 bar (abs) = 33 bar (g=gauge) approx., the heat of vaporization is very low (Table 1), any heat leak causes large vaporization and holding times are very short. In the range of super-critical pressures 33 to 37 barg, the vapor and liquid in the upper layers become above-critical fluid, which expands according to its rising temperature. It should be accounted, that the heat leak of small tanks is relatively high because of large surface to volume ratio.

Forecast of pressure increase is especially difficult in case of small high-pressure tanks, because of following aspects:

a) Unknown average thermal state of liquid, depending on different temperature in various layers of liquid, with the lower layers being colder than the upper ones. The distribution and the average saturation of the liquid depends on the thermal history of the liquid, as it was transported, mixed with the rest of liquid in the tank, stored, the liquid withdrawn, etc. In consecutive steps of filling-emptying of the tank there is stratification of the bulk. Always the coldest liquid is being withdrawn and warmer liquid remains in the tank. When filling the tank with fresh cold liquid, the vapors are condensed and possible excessive pressure reduced. The mixed liquid gets colder. As the period between individual fillings is longer, the average liquid gets warmer.

b) Low heat of vaporization at high pressures, rapidly reducing with growing pressure

c) Uncertainty of the real physical level of liquid, which depends on the liquid density, which further depends on average saturation of the liquid. Consequently, there is large uncertainty on the real volume of the vapor space above the liquid. Smaller vapor space results in quicker pressure rise. Real level of liquid in the tank is higher than shown on the scale of the DP gauge. The level increase is rapid as the pressure is getting close to the critical point.

Common causes of rapid pressure increase:

1. When there is no withdrawal from full (95%) tank, liquid expands and fills in the ullage space (the vapor space above the liquid level). As liquid is almost incompressible this results in really fast pressurization (tens of bar per day)

2. Too high temperature of the liquid.

If the volume of the tank is oversized in relation to the average liquid consumption and if there is a long period between each two fillings. the liquid stored for a long time is gradually heated up close to its boiling point. Such a heated liquid has no capacity to absorb the heat leak as it had during warming up. Majority of the heat transferring through the insulation causes vaporization of liquid and consequently the tank is pressurized more quickly.

We can reach significantly different pressure increase in case of a single monitored high pressure tank by measurement, caused by minor changes of operating parameters.

(Long-term monitoring of VT6/37 in laser application, operated at 32 barg has shown increase by 1 to 3 bar/day, e.g.).

When the tank is equipped with the economizer line (Fig. 1, line J), minimum required product withdrawal (Fig. 1, line D) for compensation of the increase of pressure in the vapor space is equal to the vaporization which makes possible venting of the vapor space to the production line (Table 2.). Without any economizer line, the product withdrawal would have to be larger in the ratio of liquid to vapor density, which at saturated liquid at 30 barg is 2.7, but it can be much larger when some colder liquid is still at the bottom of the tank.

Table 2. Minimum withdrawal rates of the product for compensation of evaporation at constant pressure 30 bar in small high-pressure tanks, equipped with an economizer line.
(Compare to Table 1.)

When liquid nitrogen tanks are operated close to the critical point of nitrogen (33 barg) and the liquid is warmed up close to saturation pressure because of longer storage time, the evaporation rate is high due to low heat of vaporization. It is typically observed as quick pressurization when the tank is on stand by at a high pressure.

The situation with palletized tank in the range of sizes from 600 liter to 2000 liters and the relief pressure 37 is very similar to that, stated above on standard bulk storage tanks with the difference, that multi-layer insulation has been used for these tanks already for thirty years. The pressure and temperature changes are much quicker at these tanks because of their smaller volume and larger NER.

2. EXPERIMENTAL PART

We modeled the situation on a small 600 liter tank (Euro-Cyl 600/37). For to follow up temperatures in various layers of liquid and vapor space, two thermometers were at the bottom head, one was in the middle of the liquid bulk, the third was at the top head. The temperature on the surface of the liquid was calculated from the pressure in the vessel as vapor-liquid equilibrium. The tank was filled with liquid nitrogen and saturated to 1 bar abs. boiling point by venting the excessive gas to the atmosphere. (Measurement of temperature changes at constant surface temperature of the liquid was the aim of this experiment. As it can be seen from the course of the temperature curves, both liquid and vapor stratification was observed at beginning.

When the liquid saturated at a low pressure and corresponding boiling temperature (1 bar and 77 K e.g.) is exposed to higher pressure (28 bar, e.g.), condensation of vapors takes place at the liquid surface, the surface becomes “warm” and the heat penetrates down by thermal conductivity of liquid.

Figure 2. Temperature changes in a tank, exposed to periodical pressurization to 28 bar. Liquid level 84%.

Figure 2. shows how the pressure in the tank collapses as result of condensation of vapor on the surface of the cold liquid even, when impulse pressurization is repeated.

Collapsing of pressure is seen by decreasing of the equilibrium temperature in the top of the tank (T_p). The heat transfer into the bulk of the liquid is seen by increasing the liquid temperature (T1 – center of the bulk 50% height, T2 – 25%, T3, T4 – 0% -bottom of the tank). After 250 hours the entire volume of the bulk and the tank walls are on a temperature around 105 K corresponding to saturation at 10 barg, while the pressure above the liquid level was still 20 barg. This experiment proves warming up of the liquid due to the pressure in the tank and simultaneous collapsing of the pressure due to condensation on the cold liquid. .

Another experiment was done with the functioning pressure build-up (PBU) regulator function (Fig. 3). The regulator set pressure 28 bar is stabilized at the liquid surface.

Figure 3. Temperature changes in a tank, exposed to continuous pressure control at 28 bar. Liquid level 40%. The 50% thermometer measures the temperature of the vapor.

Than the tank was sealed and PBU, set to 31,7 barg (equilibrium temperature 125 K). The experiment was started by full opening of the pressure build-up regulator and increasing the pressure in the tank to the regulator set pressure. After achieving the set pressure, further inlet of vapor from evaporated liquid from PBU compensated for condensation of vapor on the surface of the liquid. Maintaining constant high temperature on the surface results in continuous warming of the bulk of the liquid. At warmer liquid the rate of condensation was reduced and tendency of pressure growth even at closed regulator was compensated by manual venting to the atmosphere. This way, constant pressure and, consequently constant temperature on the liquid level was maintained during the entire experiment time. All the liquid temperatures are equalizing to the temperature of equilibrium. The same happens in the vapor space. The original stratification of vapor is equalized. We presume that more important, than the conductivity of vapor and its possible circulation, thermal conductivity of the thick walls and head of the high pressure tank play an important role, when all the metal equalizes at the temperature of equilibrium.

When the bulk of liquid is colder than equilibrium, large part of heat leak can be absorbed by liquid on account of increasing its temperature and only a small part is converted into vaporization of part of the liquid, which causes increase of the pressure. But after the temperature of the liquid is close to equilibrium, all the heat leak is converted into the vaporization and the pressure is growing quickly. This cannot be seen in this diagram, because the pressure was maintained strictly constant by opening the PBU in the first period of the experiment during 95 hours, but by manual opening of the vent valve after this time. The purpose was to measure heating of the bulk of liquid at the constant surface temperature.

3. OPTIMUM OPERATION FOR REDUCTION OF PRESSURE INCREASE

3.1. Two important aspects to be considered, when operating high-pressure tanks:

a) The pressure increase is significantly influenced by the level of the liquid in the tank.
Differential pressure gauge does not show this real level of the liquid, because the temperature and density of the liquid should be close to equilibrium. As standard, the scale of the level gauge is calculated for the density of liquid, saturated at the atmospheric pressure, as the only state, which can be definitively reproduced in a real tank. But a typical operation state is a non-defined and continuously changing saturation to a temperature, which corresponds to medium thermal saturation of the liquid between the original cold state and the saturation at the tank pressure. This means that the liquid is always warmer, its density always lower and the real liquid level proportionally higher that what can be seen from the level gauge. The vapor space can be much lower than indicated from the nominal level, which results in quick vaporization.

b) The pressure increase is also directly influenced by the temperature and its stratification.
If the temperature of the liquid is low then the major part of heat input is consumed for the heating-up of the bulk.

3.2. Following recommendations can be applied to operation of high-pressure tanks

A) Maintaining the liquid temperature as low as possible.

Lower average temperature of the liquid results in larger absorption of heat leak into the liquid and reduction of vaporization. Quantity of vapor, which has to be accommodated in the vapor space, is lower, then.

- First rule comes out directly from the above described experiments. One part of liquid heating is caused by condensation of vapor, which is higher, the higher the pressure in the tank. The liquid level surface always has a temperature, which is the equilibrium one to the pressure. Heat is transferred downwards into the colder layers of the bulk of liquid. If the temperature gradient is large shortly after the tank filling, the tank pressure would drop as result of vapor condensation on the level. PBU comes into function then and maintains the pressure by vaporization of liquid. This should be prevented during stand-by periods like nights or weekends wherever are conditions for such a sophisticated control. When a tank is filled with fresh liquid and left unused over weekend, e.g., the PBU circuit should be shut and the pressure allowed to drop for the period not in use. This would reduce the warming up of the liquid and increase the non-vent holding time during this and possible next stand-by periods.

- Another of sources of heat is mixing of the fresh cold liquid filled in the tank with the rest of the previous liquid, which is warm after long time storage. The resulting temperature is lower, when the quantity of the rest of liquid is lower. The average temperature is lower, when the volume of the rest of liquid is small. This results in another useful rule: The rest of liquid before the filling of the tank should be minimized, with respect to specific conditions, by optimized logistics of liquid re-filling.

B) Maintaining the vapor space (ullage) sufficiently large.

Two aspects have to be considered:

- real vapor space is always smaller than what could be read from the scale of the level gauge, because the density of liquid is lower than the maximum one, considered for the scale. The real liquid height is inversely proportional to the liquid density, approximately. Average density of liquid is always somewhere between that one of mixed liquid after the tank filling and the equilibrium one corresponding to the tank pressure. More detail analysis find in [1].

- larger vapor space (in the range of liquid filling 80 to 95%) reduces the pressure raise rate. (At smaller liquid filling the pressurization rate is larger because of lower heat absorption in the liquid.)

C) Maintaining the optimum filling regime

Conclusion of the items A and B: It is better to fill the tank from 15 % to 50% rather than from 30 % to 65 %, e.g. We evaluated similar cases a the tan VT6/37:

- filling from 49,3% to 63 % - average pressure increase was 2,55 bar /day

(average from measured 2,80 and 2,39 and 2,47 bar increase per 24 hours).

- filling from 18,5% to 49 % - average pressure increase was 1,66 bar /day

(average from measured 0,99; 1,81; 1,61; 2,0 and 1,82 bar increase per 24 hours).

4. LOW HEAT LEAK TANKS

As it was reasoned and documented above, the general problem of rapid pressurization at HP tanks, especially the small ones, is caused by thermodynamics of the storage processes, For to reduce this problem, Chart Ferox, a.s. developed new range of multi-layer insulated small tanks of the range 3 to 11 m3 (marked EVT tanks), based on 20 years experience of Chart Distribution & Storage Division.

The heat leak of these tanks is lower than of those, insulated by perlite, by 20% approximately.

With respect to importance of the minimization of the heat leak at the high-pressure tanks, the heat leak is further reduced by increasing the thickness of the insulation by higher number of layers. The tank EVT6/37 (multi-layer insulated tank) with 50 % layers addition represents NER decrease by 38 % compared to the perlite insulated tank VT6/37.

Measurement of holding time of liquid, saturated initially at atmospheric pressure doesn’t represent the typical real operation conditions, but it is a precise high-reproducibility measurement, which, besides of NER measurement, proves the quality of insulation. The pressure increase in 400 hrs in case of EVT6/37 at 42 % of filling from 0 bar was 2,5 bar, while it was 6,5 bar in case of the VT6/37.

Substitution of perlite insulated tanks by multi-layer insulated EVT tanks with the addition of layers for high pressure tanks would eliminate majority of users’ problems with pressure increase at high pressure tanks.

Figure 4. Comparison of pressure rise in VT and EVT tanks

CONCLUSIONS

Specifics of small high pressure cryogenic tanks results in relatively short holding times between operation pressure and relief system set pressure, which causes loss of media during several-day stand by. Following measures can be recommended for prolongation of holding time.

Refilling of the tanks at minimum possible level of the rest liquid.

Not filling the tanks up to the trycock, but to lower level, when stand-by period is shortly after the refilling.

Shutting the PBU circuit during the stand-by period.

Range of special EVT tanks with multi-layer insulation was developed for maintaining low heat leak, which results in longer holding times.

REFERENCES

1. Chrz V., Suma J. : Dynamics of tank pressure during storage of cryogenic liquids. Proceedings of the 22 Congress of Refrigeration, Beijing (CD), International Institute of Refrigeration, Paris, 2007

CR08-52

40 FOOT cryogenic INTERMODAL ISO containers

Mátl P., Lánský M., Chrz V.

Chart-Ferox, a.s., Ústecká 30, Děčín 5, 405 30, Czech Republic

Abstract

As LNG has become more and more important energy source, demand for cost effective transport has arisen. The word effective relates to several features. First of all, it is a high payload, further capability to store and deliver liquefied natural gas without any losses by venting by providing long holding time and finally, very important, compatibility with requirements of intermodal transport. ISO containers make possible continuous rail, road, river, sea shore, over sea and ocean transport. Transport flexibility creates new 40 feet container opportunities. For instance, they can be used as a base of virtual pipelines that represent a viable alternative where capital and operational costs of fixed natural gas pipelines make the project too expensive.

Transport of LNG is not the only use of ISO containers. They can also be used as temporary source of LNG during pipeline maintenance or instead of stationary storage tanks at satellite LNG stations. Advantage of 40 foot containers is that they have large volume which is also attractive for transport of other low density cryogenic liquids, ethylene or ethane.

In order to meet all customers’ requirements Chart Ferox has successfully developed and built a new 10 bar 40 foot LNG ISO container with currently the largest volume in the world, 43,5 m3.

Chart-Ferox made the best of its experience of proven 20 foot cryogenic container design. It is essential that final element method of strength calculation is used to develop light container and retain its reliability and durability. Container approval meets ADR, IMDG, RID, EN 13530 and CSC requirements, for which, among other, the container was subjected complex static and dynamic load tests.

Currently, first series of the product was delivered to a customer in Pakistan, where it will be used in frames of a virtual pipeline project.

40 FOOT CONTAINER DEVELOPMENT

Firstly, usage of the container has to be defined. If the container is intended for intermodal transport, it must comply with specific codes and norms for road transport (ADR, DOT or equivalent national codes), for railway transport (RID) and for maritime transport (IMDG). Not all countries accept all design codes. For instance, the combination of ADR and EN 13 530 could not be used for ISO containers that are to be freely used in the USA because local regulations require that such equipment must have DOT approval when transported over roads. JB 150 coded containers are required in China. Each code has significant influence on container design and prototype testing.

Major weight difference is obvious on inner vessel. In order to keep same MAWP of the container ASME/DOT inner vessel is approximately 1.3 times thicker than ADR/EN coded containers. Other option is to keep same thicknesses of inner vessel but, due to the difference between the two codes, the maximum allowable working pressure of an ISO container manufactured according to ASME is lower compared to the same ISO container manufactured according to EN13530.

The container meets TPED and can be registered and operated in each country of the European Union, or countries, who accept this code. Other codes are applicable case by case.

The name “ISO container” comes from compliance with ISO 668 and ISO 1496-3, which clearly describes ISO containers dimensions, gross weights and testing. Meeting IMDG code, the container can be used for overseas or ocean transport. It is important to know that for cryogenics liquids like LNG, ethylene or liquefied gases container has to be designed according to the tank instruction T75 of IMDG and, consequently, the chapter 6.7.4 for the design of the container. Such containers have a specific name “portable tank”. Besides portable tanks we can see cryogenics containers called “tank container” on the market, which is according to ADR but not IMDG. Difference between portable tank and tank container is that the tank container can be designed without meeting requirements of the dynamic test.

Chart Ferox has developed and built a new 40 foot LNG 10 bar ISO container, marked TVS-43-PB-10, which meets requirements of the category of “portable tank”.

Chart Ferox Container features

ISO container is characterized by following parameters:

Parameter	Data
Product name	TVS-43-PB-10
Type of the container according to IMDG/ADR	UN Portable tank
Tank instruction for portable tank	T 75
Design codes	EN 13530, ISO 1496, ADR, RID, IMDG,
Approvals	CSC, TPED, Gost-R, Rozreshenie Rostekhnadzora
Gross capacity	43 500 Liters
Tare weight	12,000 kg
Maximum gross weight	30,480 kg
Max. payload	18,480 kg
MAWP	10 bar
NER	0.2% LNG per 24 hours
Max. Holding time	60 days at 81% LNG filling

Table 1 – Chat-Ferox container features

Because of low LNG density it makes possible using the entire available size of the tank, still in compliance with the maximum allowable gross weight of 30 480 kg. Possibly, this type is the most effective with the ratio of the payload 18480 kg to tare weight 12000 kg with 43.5 m3 volume .

The cryogenic ISO container is a vacuum insulated double walled vessel fitted into a forty-foot container frame. Highly efficient insulation and low heat leak of supports are critical for all cryogenics ISO containers. Container supports hold the inner vessel in its position inside the outer vessel. Non-metallic materials with low thermal conductivity are suitable for this type of supports. The most effective thermal insulation is a multilayer super insulation combined with high-level vacuum in the interspace. It consists of reflective aluminium foil and low conductivity glass-fibre-paper. These two components are laid in multiple layers providing thermal protection against heat leak.

Thermal performance is characterized by NER index, which shows how much liquid evaporates during 24 hours. The lower the NER index the better thermal performance the container has.

Other important indicator is the Holding Time. It specifies how long it takes the container to pressurize from atmospheric pressure up to the safety relief valve set pressure. As the relief valves should not vent during the whole transport time, the holding time determines the maximum transport time of a container without any product loss. Calculation method and measurement procedure is described in EN 12213 – Methods for performance evaluation of thermal insulation.

Design pressure of the ISO container is another technical aspect to be taken into account when selecting the right product for certain applications. The higher the operating pressure of the container, the longer the holding time. Therefore, it may seem to be better to select a container with the highest possible operating pressure. Higher operating pressure of the vessel results in higher tare weight, which may negatively affect the payload. Therefore design pressure should be set in order to ensure holding time and certain tare weight limit. The table below shows holding times for 10 bar and 7 bar container with different filling conditions but both with NER rate 0.2% LNG/day. 0 bar filling saturation pressure can be expected from marine terminals, e.g.. Then holding time is sufficient for both the design pressures. Most common case is that the inner vessel of the container arrives to filling station not with LNG temperature which increases saturation pressure of filled liquid. Moreover LNG from liquefiers or from intermediate pressure storage is often saturated to higher boiling pressure, 3 bar e.g.. We can consider from practical experience 5 bar saturation pressure. Holding time for 7 bar container would be probably too short, then.

		Holding time of 7 bar MAWP container	Holding time of 10 bar MAWP container
0 bar saturated LNG after filling	days	50	60
5 bar saturated LNG after filling	days	10	22

Table 2 – Holding time comparison

PRODUCT DESIGN AND TESTING

As mentioned above, this ISO container is of the category of “portable tank” according to IMDG. Quality and reliability of such a product is a fundamental requirement of every operator of cryogenic containers. This is why we paid maximum attention to the design concept. Its design arose from proven 20 foot version, but increased length brought new challenges. Finite element analysis (FEM), which was previously used during improvements of 20 foot container, made possible to compare stress in all the critical points and compare with those, which occur on the many years operated designs. Gained experience established qualification for design and optimization of new-longer container using FEM.

During design, it was necessary to optimize between contradictory requirements:

- Minimum weight

- Low heat losses, which mean suitable support system

- Stiff and strong construction able to withstand all operational loads

- Fatigue resistant construction.

All the design and testing states were calculated. It included static load of 1G in vertical upward direction and 2G in all other directions. Static tests included stacking, lifting from the four top corner fittings, lifting from the four bottom corner fittings, external restraint (longitudinal), internal restraint (longitudinal), internal restraint (lateral), rigidity (transverse), and rigidity (longitudinal). The most difficult state was the dynamic test;from the view of the assembly strength and from the view of evaluation as well.

According to ADR 2007 and IMDG each new portable tank has to meet requirements on dynamic test according to Manual of Test and Criteria, which is based on code „Transport Canada“. The container is fixed by all 4 corner fittings to a staying wagon. Another wagon strikes it several times until the impact acceleration meets the required criteria. Measured acceleration is analysed and Shock Response Spectrum (SRS) is calculated. The required impact is characterized by a defined SRS curve. The older method required maximum measured acceleration of 4g. The peak value of the acceleration record according to the new method during the test of the container prototype was 18g. Nevertheless the container was able to withstand that impact and dynamic test was recognized as successful.

One of the test cases simulated by FEA is shown in Figure 1. Linear elastic analysis was used which means that no deformation affecting model were considered. As deformations are negligible in comparison with structure dimensions they were multiplied by 150 in order to make them visible.

Picture 2 shows container during physical “Bottom lift static test”. Container is held by 4 arms whereas the angle between vertical line and arm axis is 45° which simulate lifting by ropes.

Figure 1 – FEM stress view

Deformed shape of ISO Container; Transverse stiffness test, ISO 1496-3. displacement enlarged 150 times, max displacement of 4.3 mm.

Figure 2 – Bottom lift static test according to EN 1496-3

40 FOOT Container usage

An advantage of ISO containers is the possibility of continuous transport over road, rail, river, sea or ocean directly to the end user without any liquid transfer. Another advantage is an easy implementation. The container of course undergoes all related regulations and therefore it doesn’t face possible problems with crossing frontiers. Road regulations, which relate just to the carrying lorries of general use for the container transport, are usually available at logistics companies. The third important advantage is the possibility of using the container not only as a means of transport, but also as a temporary satellite station or back up system in case of pipelines and storage tank maintenance.

If transport of LNG is regular and it’s defined by an exact amount (tons/day) we call it virtual pipeline. It has the same purpose as gas pipelines but with extra benefits. Gas pipelines have to be designed for expected flow rate but number of ISO containers operated to certain destination can be easily adjusted to the gas demand. At the region of destination, the containers can be directed to particular customers and unloaded to satellite stations or used as temporary installation by exchange full for empty. Owners of large logistics or gas companies can flexibly decide where containers are needed worldwide.

There are many typical examples, where it is effective to use LNG transport instead of gas pipeline.

If there are difficult geographic conditions for laying down pipelines (sea, mountains)

The existing pipeline is not sufficiently sized for growing demand of gas at the end site.

The gas consumption is irregular with high peaks caused by a batch process technology, weather, contract campaigns etc. The pipeline size may not be sufficient for covering the peaks, or the user is penalized for high or low consumption.

Isolated users, such as hotels in the country, factories, small towns, villages, where the energy consumption is of interesting size, but not sufficient for justifying a longer distance pipeline

Each new gas delivery project should be evaluated by feasibility study where there is comparison of all related costs.

CONCLUSION

40 foot ISO containers are specific products for LNG, liquid ethane and liquid ethylene, where the large volume can be used for the light liquid within the maximum container gross weight. It is an effective means of transport for complex logistics conditions when combining various paths like road, railway, river, sea and ocean without any transfer of product. This is based on meeting requirements of ADR, RID and of portable tank according to IMDG. It is also suitable for pilot projects, because the container may serve as temporary storage tank at satellite stations.

Figure 3 – Routine liquid nitrogen testing

Figure 4 – 40 FT ISO Containers

REFERENCES

1. Chrz V., Bures I.: ISO containers – modern means of cryogenic liquid transport. Proceedings of the 4th International conference on cryogenic technology and equipment, MirExpo, Moscow, 2007

2. Miroslav Cerny, Ivan Bures, Karel Mudroch: Design of cryogenics ISO containers – Cryogenics 2004

CR08-53

Cryogenic liquid transfer possibilities – focus on static vacuum insulated pipes

Takács D., Chrz V.

Chart-Ferox, a.s., Ústecká 30, Děčín 5, 405 30, Czech Republic

ABSTRACT

Liquefied gases can be transported in several piping systems. The most common systems are: naked pipe, mechanically insulated pipe, dynamic vacuum insulated and static vacuum insulated pipe. Each system is justifiable under certain conditions. Naked or mechanically insulated pipes can be used for periodical service with large flowrates, as filling pipelines of storage installations. Maximum insulation performance is required for pipelines under continuous operation with low flow rates, just to mention two extreme examples.

Chart has developed efficient static vacuum insulated pipe with patented features as bayonet-type connections for tighter seals using Invar alloy in the design. Several pre-designed solutions of joints of the pre-fabricated sections and several systems of thermal expansion compensation are presented with their particular advantages and suitability for different projects as well as impact on the project economy.

Several special accessories were developed as vapor separators. Experience from installation, testing of the final systems and follow-up service will be also presented.

INTRODUCTION - CRYOGENIC LIQUID TRANSFER POSSIBILLITIES

There are several possibilities to transfer liquefied gases at cryogenic temperatures such as:

- dewars

- piping systems

a) bare pipe

b) mechanically insulated pipe

c) dynamic vacuum insulated pipe

d) static vacuum insulated pipe

Dewars are suitable for small gas quantities. The filling losses are up to 40% (flash, cool down). There are handling labour costs and risk of injuries during handling

Bare pipes can be used for a larger bore, short length piping with high flowrates (tank filling, space craft LOX filling). They have high liquid losses. Frost and condensation can occur on the surface of the piping system. Condensation of air can take place on the pipe outer surface when transporting low pressure nitrogen. Enrichment of the condensate with oxygen is high, which could lead to an explosion.

Mechanical insulation is an economical option for short runs (< 5 m). A foam could degrade over the time and require periodical replacement. This cause operational cost. The insulation is bulky and it has 15x higher heat leak than static VI line.

The dynamically pumped system enables fast delivery and easy installation but the vacuum pump electricity consumption and maintenance cost can be critical. There is no radiation shield installed and in case of any failure the insulation of the system is lost. The dynamically pumped piping system has 3x higher heat leak than a static vacuum insulated line.

1. WHY MULTI LAYER VACUUM INSULATION?

Heat transfer is a transfer of mechanical energy between colliding molecules expressed as apparent mean thermal conductivity in [W/m_K] between boundary temperatures 300 K (ambient temperature) and 80 K (liquid nitrogen @ 0 barg)

air - 0.015 W/m_K (conductivity only)

solid insulations (cork, balsa wood) - 0.04 W/m_K

polyurethane foam - 0.02 W/m_K

unevacuated powder (perlite) - 0.02 W/m_K

evacuated perlite - 0.001 W/m_K

straight vacuum - 0.005 W/m_K

superinsulation - 0.0002 W/m_K

Case Study for liquid nitrogen transfer system (DN15):

Nitrogen for freezer filling Tank pressure: 2 bar

Freezer pressure: 0 bar (flash loss 11.1%) Distance: 15 m

End point quantity: 1000 kg Filling: 1x week
Project parameters see in the table 1.

Operation Loss in rigid pipe systems		Static VIP	Mechanical Ins.	"naked" pipe
Flow rate	[kg/h]	973	956	809
Heat leak	[W]	15	225	2250
Loss due to Heat Leak	[kg/h]	0,3	4,6	45
Time	[hr]	1,2	1,2	1,5
Operation Loss	[kg]	125	131	199
Operation Loss	[%]	11,1	11,6	16,6

Table 1: Operation Loss in rigid pipe systems

2. CHARACTERISTICS AND DESCRIPTION OF VIP

The vacuum insulated pipeline is composed of vacuum insulated sections. The sections are fully manufactured, evacuated and sealed in the factory. The individual sections are connected preferably my means of bayonet coupling or, alternatively, by evacuated field joints.

The vacuum insulated piping comprises a tube-in-tube conduit for transfer of mainly cryogenic fluids. The inner tube holds liquid, the outer tube keeps the vacuum insulation and bears the external loads. The annular space between the tubes is provided with insulation. The thermal contraction of the inner tube is compensated by inner line bellows.

Vacuum insulated pipes have very low heat in-leak because of the static vacuum insulation and the radiation shielding. Under normal circumstances (based on relative humidity and outside temperature) the VIP will be free of condensation of atmospheric water.

Standard piping sizes are listed in the Table 2 according to nominal DN sizes.

	Unit	Line nominal dimension
		DN15	DN25	DN40	DN50	DN80
Inner tube dimensions	mm	D21.3x1.65	D33.4x1.65	D48.3x1.65	D60.3x1.65	D88.9x2.11
Inner tube ID (approx.)	mm	18	30	45	57	85
Outer tube dimensions	mm	D60.3x1.65	D88.9x2.11	D101.6x2.11	D101.6x2.11	D141.3x2.77

Table 2. VIP line nominal dimensions

Design parameters of the vacuum insulated transfer systems are designed in accordance with the data in the table 3.

Parameter	Units
Design pressure	Bar (psi)	10.3 (150) *
Design temperature	K	73 to 338
Design code		ANSI Section B31.3, PED
Inspection authority		By code

* Other pressure ratings available upon request

Table 3. Design data of vacuum jacketed pipelines.

The inner and the outer pipe material is stainless steel AISI 304L or equivalent. The tube is of a longitudinally welded type. The compensating bellows are incorporated within the inner tube. Therefore the thermal dilatations of the process tube are not transferred to the outer tube.

The concentricity of the inner tube within the outer is ensured by spacers. The spacers also guide the compensators.

The line insulation comprises a combination of the laminar radiation shield with vacuum. Also a system of rest gas capturing by getters and the evacuation port is considered to pertain to the insulation.

The laminar radiation shield consists of 24-26 layers of spirally wound aluminum foil interleaved with glass paper.

The sections are fully manufactured, evacuated and sealed in the factory. There is no evacuation on site expected. The shipping vacuum level is < 1 Pa, the operating vacuum is expected to be of 100-1000x lower magnitude (0.01 - 0.001 Pa).

The rest gas capturing system comprises molecular sieve for absorption of most of the vacuum rest gas, which is mostly water and air, as well as the gases released to vacuum by construction materials or penetrating there from the ambient. Further, it comprises palladium oxide as a hydrogen converter.

Doses or these getters are installed in each section.

Each section can also be optionally provided with a Hastings DV-6 vacuum sensor (range 0.133 to 133 Pa) for quick vacuum checks.

3. PIPE JOINTS

The individual sections are connected preferably by means of bayonet couplings (Fig.1).

For DN15 and DN25 lines the bimetallic bayonets are proposed. The tip, or nose, of the male bayonet half is made from Invar 36, the female counterpart is from stainless steel. When cooled down to the cryogenic temperature, the stainless steel tip shrinks against Invar material, which stays nearly unchanged. The bayonet seal is thus achieved at its nose, the O-ring at the flange only prevents moisture from penetrating into the inner bayonet space. Both bayonet halves are fixed with a clamp.

a) b)

Figure 1., Bayonet joints female (a) and male (b)

The larger bore piping will use close tolerance bayonets (Johnston style). The fluid pressure is kept by the O-ring that is protected from the cold by the gas cushion. The gas layer in between both bayonet parts shall be very thin to prevent thermal oscillation (pulsation of the liquid/gas in the gap).

Both the above described types of joints minimize the installation work on site and the project lead time. On-site joints (Fig. 6. and 7) with locally perlite-insulated chambers offer the lowest cost solution, but the construction work lasts longer and the work exposed to weather may result in lower quality. The chambers are made from pipes of larger diameter than the outer pipe of the pipeline. After the overhanging inner pipes are welded together, the chamber pipe is set on and welded to vacuum tight chamber, filled by perlite and evacuated. Some on-site joints are mostly required even with bayonet joint system for compensation of tolerances in distances and for easier assembly of the system.

4. MODULAR VACUUM INSULATED PIPE

Modular Vacuum Insulated Pipe (MVIP) is a prefabricated interlocking system for cryogenic liquid service. Piping modules are connected with vacuum insulated bayonets for a simple and flexible system installation.

The following modules are available:

a) Straight lengths

b) Flexible module lengths

c) Vacuum Insulated Valve module

d) Cryovent module

e) Elbow module

f) Tee modules (male branch or female branch)

g) Drop modules with internal liquid trap

h) Bayonet adapters to adapt to Chart Vacuum Insulated Pipe design

Figure 2. Modular pipeline

5. PLANNING FOR VACUUM INSULATED PIPELINE SYSTEMS

The vacuum insulated pipeline system is assembled from sections that are fully factory manufactured. It means there is very little flexibility in this type of assembled lines. The piping routing and joint positions have to be carefully planned. Some degree of freedom can be obtained by using flexible hose sections, or field welded and evacuated joints.

From the process point of view, the VIP transports saturated liquid or even two-phase flow. The piping routing and dimensioning shall be planned for with this in mind (e.g. avoiding gas traps, vents use and distribution, etc.)

6. INSTALLATION

The VIP installation is very straightforward. Normally available supports and hangers can be used. Bayonet couplings require no welding on site and normally are preferred over the field joints. These are more laborious to make, however they could be of lower cost (say at diameters of DN50 and more), and provide additional flexibility for mismeasurements in planning and ease of installation.

7. APPLICATIONS OF SYSTEMS

Turn-Key packages of vacuum insulated pipelines are delivered for typical process installations:

Storage of cells and tissues

Long term storage of cells and tissues for artificial insemination, for transplants and other genetic and biological applications require liquid nitrogen temperature. Liquid nitrogen is distributed from the liquid nitrogen storage tanks to individual cryo-bio storage tanks.

Electronics industry

Inert atmospheres are required at some manufacturing procedures in microelectronic production. Continuous purging of nitrogen in small quantity is needed at individual manufacturing stands. For example, Chart Ferox delivered a turn key installation of 80 meters with phase separators and cryogenic vacuum insulated valves to a microchip factory at Roznov, CR.

Nitrogen Injection applications Nitrogen Injection:

Liquid Nitrogen Injectors deliver competitive advantages to food and beverages manufacturers.

Textové pole: Figure 3. Typical lay-out of a vacuum insultated pipeline application in an industrial installation Nitrogen gas in and around the product in a container displaces the oxygen normally in the atmosphere. Products packaged without hot process sanitation will degrade in the presence of oxygen. Nitrogen inert atmosphere prevents it. By dropping liquid nitrogen into the drink, oxygen is displaced and a protective atmosphere of nitrogen prevents oxidation and consequential degradation of the drink.

Oxygen Exclusion Applications in solid products:

Granular, round and irregular shape products such as: candy, nuts, potato chips, pills, and coffee, are saturated by nitrogen during processing for displacement of oxygen.

Pressurization Applications:

Beverages including water, juices, sweet drinks, low carbonation drinks. Usually bottle or can rigidity is necessary for bottle handling, distribution, and vending. A drop of liquid nitrogen is injected into the drink shortly before sealing, which makes overpressure by its rapid evaporation.

Typical accessories, used in the complete installation systems are:

Vacuum jacketed valves. Original Chart product is currently manufactured in co-operation by a subvendor.
Phase separators for removal of vapor from the two phase stream after reduction of liquefied gas pressure, when the liquid is transferred at its boiling point and throttled by the pipeline pressure drop. The APPS 160 (Adjustable Pressure Phase Separator) is used to lower the saturation point of liquid nitrogen. The APPS 160 will take liquid from a bulk tank, where the liquid is stored at a higher pressure (8,3 bar (120 psig) for example) and recondition it to the lower pressure, required for example, for test chambers (3,4 bar (50 psig) for example). This is achieved by separation of vapor from liquid and venting the vapor to the low pressure space.

· Liquid nitrogen injectors, which are dosing small quantities of liquid or gaseous phase for local cooling or inerting

· Flexible hoses are used for compensation of distances or connection of components, which can move during the use of the system.

Figure 4. Example of an industrial installation

8. LNG APPLICATIONS

Larger DN pipelines are often used for unloading or transfer of LNG at LNG terminals and satellite stations. 1 km long VIP was built by Chart at Trinidad for transfer of LNG from liquefier storage tanks into ocean-going LNG carriers.

Chart Ferox delivered an LNG re-fueling station for LNG fuelled ferries in Norway. A 100 m long DN 150 pipeline had to be built under a road to a sea shore. The transfer capacity is 1700 liters per minute. With respect to corrosion and impossibility of maintenance in an underground channel, the Vacuum insulated option was chosen for the project. The pipes are connected with on site joints with respect to the method of assembly inside the channel (Fig. 5 to 9)

Fig. 5. Twelve 10 m long sections were delivered to the site.

Fig. 6.
VIP DN 150 in the channel

Fig. 7 On-site joint with vacuum port

Figure 8. From LNG storage tanks to the undergroung channel entry. Transition form bare and polyurethane insulated pipes to vacuum insulated pipe.

Figure 9. During transfer of LNG:
Black part: The polyurethane foam insultated pipe in waterproof jacket winding.

White part: Bare pipe, frosted from atmospheric humidity

Grey part: Outer jacket of the vacuum insulated pipeline, No frost nor condensation.

CONCLUSIONS

Liquefied gases can be transported in several piping systems. Each system is justifiable under certain conditions.

Delivery of liquefied gases with minimum losses can be most efficiently done by using static vacuum insulated pipelines. This justifies VIP pipeline suitability for continuous or periodical transfer of smaller quantities, as it is typical at cryogenic storage, for refrigeration of continuous processes, etc. Another important advantage is resistance to weather and other ambient corrosion conditions. Bayonet joints make possible quick on site assemblies of the complete systems.

CR08-57

THERMODYNAMIC STUDY OF THE SIMULTANEOUS PRODUCTION OF ELECTRICAL AND COOLING POWER FROM LNG

Parise J.A.R.¹ , Esteves A.D.S.¹

¹Pontifícia Universidade Católica do Rio de Janeiro, 22453-900, Rio de Janeiro, Brazil

Abstract

The paper studies the use of LNG for electrical power production. Typical regasification systems, ORV (open rack vaporizers) and SCV (submerged combustion vaporizers), do not envisage the possibility of recovering the cryogenic energy of the Liquefied Natural Gas (LNG) for producing refrigeration power, which may be needed down the electrical power utilization chain. A simple thermodynamic analysis compares the energy utilization efficiency of a LNG electrical and cooling power cogeneration arrangement with conventional thermo-electrical power systems making use of LNG as fuel. Energy conservation fundamental principles are applied in the analysis to provide a description of the thermal behavior of the system, in terms of the electrical and cooling power demands. The energy efficiency of the cogeneration scheme is compared to those of traditional regasification systems, in terms of the electrical power to cooling load ratio. Typical system configurations, capable of matching cooling and electricity demands in a realistic situation, were devised for this preliminary analysis.

1. INTRODUCTION

Usage of natural gas as a primary energy source is growing in importance, due to large proven reserves and to its relatively lower air pollution and emission of greenhouse gases, if compared to other fossil fuels. It is expected that natural gas will account, by 2020, for 30% of the total electricity production. Liquefaction of natural gas (by condensation to a cryogenic temperature of about -161^oC) makes it possible to convey this energy source from remotely located reserves to the consuming markets. The maritime transportation of LNG, in double-hulled carriers, competes with long distance pipelines, now contributing to nearly a quarter of worldwide gas exports [1]. At the import terminal LNG is regasified, where several types of LNG vaporizers are commonly used. The following five types have either been used or demonstrated in LNG receiving terminals [2]: (i) Open Rack Vaporizers (ORV), (ii) Submerged Combustion Vaporizers (SCV), (iii) Shell and Tube type Vaporizers (STV) including modified designs such as the Reli-Vap type vaporizer, (iv) Combined Heat and Power unit with Submerged Combustion Vaporizer (CHP-SCV), and (v) other types, such as Ambient Air-Heated Vaporizers [3, 4]. LNG receiving terminals commonly use one of two types of LNG vaporizers: the ORV and the SCV. In general, the ORV system uses seawater as the heating medium. It has a lower operating cost than the SCV, but normally a higher capital cost due to the vaporizer equipment, the added seawater intake/outfall system, the large diameter seawater pipes, and the seawater pumping and treating systems The SCV requires fuel for the LNG vaporization, and the fuel consumption is about 1.5% of the send-out [2]. A common characteristic of these methods is that they have room for significant improvement on energy utilization, since no recovery is made of the “cryogenic energy” of LNG. Moreover, environmental impacts are expected from the combustion of the fuel gas for vaporization (SCV) and from the reduction of sea water temperature (ORV).

Two papers, among others, can be found in the literature reporting on studies of arrangements that recover part of the energy consumed in the liquefaction process of natural gas. Kaneko et al [5] make use of a combined cycle, comprised by a conventional gas turbine, working as the topping cycle, and an inverted Brayton cycle, for bottoming purposes. The authors claim a superiority of this system, if compared to traditional ORV systems, in terms of thermal efficiency and specific output. Deng et al [6] propose a new cogeneration power system, with two energy outputs rates (electricity and refrigeration), making use of both chemical and cryogenic energies of LNG.

This paper presents a basic study on the production of refrigeration power from the regasification process of LNG-fuelled thermal power plants for electricity production. The analysis is based on energy conservation fundamental principles and on the fact that, within the range of consumption of the produced electrical power, there is potential for concentrated refrigeration power demand, like from large-capacity refrigerated storage spaces.

2. system description

Three systems are considered for the production, from LNG, of electrical and refrigeration power. The first two systems, figures 1 and 2, are based on traditional thermal power plants, meeting electrical power demand as well as the electrical power consumption of a vapour compression chiller. Natural gas, produced from traditional regasification methods, ORV (figure 1) or SCV (figure 2), fuels the thermal power plants. The third system studied, figure 3, is based on cogeneration, where the refrigeration power is produced directly from the regasification process, thus diminishing the power required from the chiller. Neither heat recovery nor heat demand are considered in this study. Main components, or plants, of the system are: the thermal power plant (TP), the vapour compression chiller (VC) and the regasification plant (RG). The thermal power is characterized by an overall thermal efficiency, , and the vapour compression chiller, by the refrigerating coefficient of performance, . Two demands are to be met: the electrical and the refrigeration power loads, and , respectively. The LNG stream is represented by an energy rate equivalent of LNG consumption,, which, after regasification, converts to the energy rate equivalent of natural gas consumption, . Processes at the thermal power plant, chiller and regasification, involve unrecoverable rates of heat gain or loss, denoted by.

Figure 1: Simplified energy flow diagram of an electricity and refrigeration power production plant from LNG, using an ORV regasification scheme

Figure 2: Simplified energy flow diagram of an electricity and refrigeration power production plant from LNG, using an SCV regasification scheme

Figure 3: Simplified energy flow diagram of an electricity and refrigeration power production plant from LNG, using a cogeneration regasification scheme

The compressor power consumption of the chiller is . Specifically, the ORV regasification scheme, figure 1, has an overall seawater-to-LNG heat transfer effectiveness,. By its turn, the SCV regasification scheme, figure 2, is characterized by the ratio of imported gas used as fuel gas for LNG vaporization, . Finally, the cogeneration scheme of figure 3 allows for two possibilities: (a) the heat transfer rate for the vaporization of LNG, , is greater then refrigeration power demand, , which means that refrigeration power demand is met but heat transfer rate from an additional heat source, , is required to fully vaporize the LNG stream; otherwise, (b) a vapour compression chiller, of smaller capacity, if compared to non-cogeneration schemes, complements the LNG vaporization load, , with the evaporator power capacity, , in order to meet the refrigeration load,.

3. thermodynamic model

A set of parameters is taken from previous studies on heat recovery [7] and cogeneration systems [8]. The energy conversion ratio,, is here defined as the total energy delivery rate (electrical and refrigeration power) divided by energy input rate from LNG. In other words, is the ratio between the sum of all energy products and the total energy consumption.

Two other non-dimensional ratios, and , compare the magnitudes of the cooling to electricity loads and of cooling load to LNG vaporization, respectively.

Dividing equation by and introducing equation , one obtains:

The ratio between LNG energy equivalent consumption rate and electrical load will depend on the characteristics of the regasification system employed. Each regasification scheme, figures 1 to 3, will be dealt with separately. First, the energy balances applied for the thermal power plant and the vapor compression chiller, common to all three schemes, are, respectively:

3.1 ORV Scheme

Applying an energy balance applied to the ORV regasification plant control volume, depicted in figure 1, provides:

Substituting equations and into and dividing it by , the energy conversion ratio is obtained in terms of and system characteristics, and , respectively:

3.2 SCV Scheme

The use of imported gas as fuel for the vaporization of LNG reduces the flow of energy to the thermal power station by:

With the substitutions indicated in section 3.1, the energy conversion ratio becomes:

3.3 Cogeneration Scheme

Two situations are to be considered, regarding how the required vaporization heat rate, , compares with the refrigeration power demand, , equation . If , i.e., , it is assumed that the additional heat supply rate for LNG vaporization, , is provided by an ORV regasification system and, of course, a supplemental chiller is not required, i.e., . On the other hand, if , i.e., , the supplemental chiller will come into effect. For both cases, equation applies. Development of the energy conversion ratio equation, for both cases, follows:

a) Refrigeration load greater than LNG vaporization load, :

Equation applies. An energy balance applied to the vapour compression chiller control volumes provides:

Taking equation into and dividing it by , yields:

In the cogeneration scheme, the chiller only accounts for the refrigeration power that is not met by LNG vaporization:

Taking equations and into equation and

Writing in terms of the load ratios, equations and , and taking equation into , the energy conversion ratio equation becomes:

b) LNG vaporization load greater than refrigeration load, :

The chiller does not operate,

so that all the electrical power produced by the thermal power plant goes to the electrical power load:

Taking equation into :

4. results

Equations , , and were applied to provide an overview of the first-law performance in terms of the system characteristics and of the energy demands (refrigeration and electricity). Typical values for the system characteristics were chosen as follows: , [2],(Rankine cycle) and

Figure 4 shows how the energy conversion ratio varies with the two load ratios. One observes the superiority of the cogeneration scheme. The least energy efficient regasification arrangement is, of course, the SCV, because of the consumption of a fraction of LNG stream to convey vaporization, thus reducing , in relation to ORV, by a factor of , equation . The greater the cooling to electrical load ratio, , the greater the energy conversion ratio. Oppositely, greater energy conversion ratios are obtained when the LNG vaporization load surpasses the refrigeration load, , as no need of is made the electricity consuming chiller. Figure 5 shows the same results plotted in a different manner, emphasizing the reduction of the energy conversion ratio of the cogeneration scheme when the chiller comes into operation, in order to meet the refrigeration load. The limit of the curves, for very large values of , is the energy conversion ratios of the traditional regasification processes, in the present simulation, the ORV scheme. Such limit case would be the situation when the refrigeration load would be far greater than the vaporization load.

Figure 4: Variation of the energy conversion ratio for different regasification schemes, with varying cooling to electricity load and cooling load to LNG vaporization ratios, and , respectively.

Figure 5: Variation of the energy conversion ratio for different regasification schemes, with varying cooling to electricity load and cooling load to LNG vaporization ratios, and , respectively.

5. concluding remarks

The present analysis has proven, as expected, that the recovery of the cryogenic energy of LNG, in order to match a certain refrigeration load, can provide reasonable improvement on the energy utilization of LNG, here measured by a proposed non-dimensional parameter, the energy conversion ratio. Moreover, environmental impacts of the traditional schemes, like seawater temperature reduction (ORV) or additional CO₂ (SCV) emission, could be lessened with the application of a cogeneration scheme. However, attention should be brought to the fact that the present analysis did not take into consideration certain thermodynamic aspects, such as temperature requirements in the recovery arrangements, as well as technical, economical, cost-effectiveness and eventual operational limitations that may arise from the implementation of such scheme. For instance, a LNG receiving terminal with a C₂+ or C₃+ separation facility can import LNG feeds with varying compositions, in order to meet stringent calorific value export gas specifications and to decrease capital and operating costs. Studies found in the literature shows that process schemes for extracting C₂ or C₃ from rich imported LNG are feasible, effective, and economical [9]. The presence of such facilities may, then, impact the overall balances as proposed above. Furthermore, environmental changes due to global warming have caused various kinds of serious problems on a global scale. From this standpoint, the LNG industry must reduce harmful effects on the environment with the LNG related working chain, which includes exploitation, liquefaction, transportation of a natural gas and re-gasification issues [10]. Further study on the subject is recommended.

acknowledgements

The authors are indebted to CNPq (from the Brazilian Ministry of Science and Technology) and to FAPERJ (State of Rio de Janeiro Research Funding Agency) for the financial contribution to this study.

REFERENCES

1. Chrz, V., Liquefied Natural Gas: Current Expansion and Perspectives, 19^th Informatory Note on Refrigerating Technologies, International Institute of Refrigeration, Paris, France, November (2006)

2. Yang, C.C., Huang, Z., Lower Emission LNG Vaporization, LNG Journal (2004) November /December 24-26

3. Hubbard, B.S., Floating Storage and Regasification Concepts Using the LNG Smart Air Vaporization Technology, AIChem 2007 Spring National Meeting, Houston, USA (2007)

4. Davis, J.F., Understanding Ambient LNG Vaporizers, AIChem 2007 Spring National Meeting, Houston, USA (2007)

5. Kaneko, K., Ohtani, K., Tsujikawa, Y., Fujii, S., Utlization of the Cryogenic Exergy of LNG by a Mirror Gas-Turbine, Applied Energy (2004) 79 355-369

6. Deng, S., Jin, H., Cai, R., Lin, R., Novel Cogeneration Power System with Liquefied natural Gas (LNG) Cryogenic Exergy Utilization, Energy (2004) 29 497-512

7. Parise, J.A.R. and Cartwright, W.G., Experimental Analysis of a Diesel Engine Driven Water-to-Water Heat Pump, Journal of Heat Recovery Systems and CHP, (1988)8 75-85

8. Herbas, T.B., Dalvi, E.A., Parise, J.A.R., Heat Recovery From Refrigeration Plants Meeting Load and Temperature Requirements, International Journal of Refrigeration, (1990) 13 264-269

9. Yang, C.C., Kaplan, A., Huang, Z., Cost-effective design reduces C2 and C3 at LNG receiving terminals, Oil & Gas Journal, (2003) 101 issue 21

10. Kajitani, M., Efforts to minimize the environmental load at LNG receiving terminals, 23rd World Gas Conference (2006) Amsterdam

CR08-51

Gas impurities freezing out technologies.

Klepal J., Stoček P.

ATEKO Hradec Kralove Czech Republic

ABSTRACT

An experimental testing study verified the influence of small amount of epichlorhydrine vapour on character of the ice accretion formation during toluene freezing out from the nitrogen / toluene / water vapour mixture.

A toluene freezing out spiral wound heat exchanger with glass shell was installed into the testing line. Forms of the ice accretions were observed optically.

Epichlorhydrine impurities changed character and density of the ice accretion.

INTRODUCTION

One of methods used for separating impurities from off gases from chemical production plants consists in freezing out the impurities. However, the effectiveness of this process is dependent on the one hand on the concentration of an undesirable substance in a gas and on the other hand on the fact whether or not are present at the same time other impurities or other components (water vapour). In particular water vapour presence influences markedly the form of obtained substance frozen out from a off gas. Mostly they are so called hydrates formed by some hydrocarbons in the presence of water at a low temperature and at a high pressure eventually.

Particularly the form similar to hydrates formed at freezing out toluene vapours from off gases from the production line was the subject of the experimental measuring as described hereinafter.

1. EXPERIMENTAL APPARATUS AND PROCEDURE

A schematic diagram of the apparatus used in this work is given in Figure 1. The main component of the apparatus consists of a spiral wound vertical heat exchanger. The exchange is made by upward spiral pipe with a diameter of 6 x 1 mm wound with a diameter of 41 mm with a pitch of 20 mm.

The air with impurities was fed to the bottom end of the heat exchanger through a stainless steel tube with diameter of 17 mm fitted in the centre of the winding. For visual observation of the degree of formation of ice accretion on the wound heat transfer surface the heat exchanger was built in a glass tube with a diameter of 54 mm with the bottom end closed.

The glass tube with the wound tube exchanger was placed into a visual glass Dewar vessel for eliminating heat losses.

In the upper part the Dewar vessel was closed as a protection against penetration of the ambient atmosphere. The inter space of the heater was cooled with freezed ethanol. The circulation of ethanol is ensured with the help of a circulation pump and by LIN freezed bath filled by ethanol. The temperature of the circulating coolant was maintained at a value of –25 °C with the help of liquid N2. The off gas for freezing out was fed into the bottom

Textové pole: Legend: 1. Gas INPUT 2. Pressure reducing valve, 3 Gas flow meter, 4. Water saturation, 5. Toluene saturation, 6. Demister, 7. Glass Dewar vessel, 8. Experimental heat-exchanger, 9. Freezed ethanol bath part of the shell side of the heat exchanger through the tube fitted in the axis of the heat exchanger winding. As a carrying gaseous fluid it was used compressed air treated in a reduction station to a lower pressure. The gas was led through a gas meter and two vessels designed to serve for saturating the air with water and toluene. In this circuit it was installed also a separator with drop and overrun catcher.

In the course of the measuring they were read temperatures of the freezing out fluid and temperatures of the leaving air together with air rate of flow.

Textové pole: Fig. 1 The proper experimental verification of the freezing out process was divided in two experiments. In the course of the first experimental verification (Fig. 2) it was verified the course of freezing out of the pure toluene from the air and in the other experimental verification (Fig. 3) the course of freezing out of toluene from the air in the presence of impurities. In this case they were trace amounts of epichlorhydrine (1-chloro-2,3-epoxypropane) and 2,3-dichloro-1-propen.

2. RESULTS AND DISCUSSION

Already from the beginning of the measuring in both the cases a growth of ice accretion occurred in the form of needles on the proper winding of the heat exchanger tubes and then also on distance of wind.

A part of toluene presented in the air condensed in the feeding central tube and in the form of a condensate flown down to the bottom of the glass cylinder (Fig. 1).

The other part was then absorbed by the arisen ice accretion formed by hydrates on the heat exchanger winding (Fig. 2, 3).

Both the experiments ware carried out up to a complete “freezing and clogging” of the heat exchanger winding. Then the temperature of the circulating cooling liquid was increased and in the course of about 20 minutes all ice accretion was removed from the heat exchanger winding.

Textové pole:
Fig. 2 In the course of the experiments they were caught gradually liquid portions from single phases in the course of the freezing out and also in the course of thawing of the ice accretion resulted. After a stabilization and separation of single liquid components of the two-phase system (water + toluene) the measuring of volumes caught was carried out.

Textové pole:
Fig. 3 By carrying out a balance of inlets to the experimental equipment and outlets from the experimental equipment it has been found the effectiveness of the process of separating toluene from the air, in the first case (pure toluene) about 24 % and in the other case (contaminated toluene) about 43,5 %. A value of the theoretical efficiency of 48% has been calculated on the basis of a balance computation made with the help computing programme CHEMCAD with the use of a model for ideal gases.

3. CONCLUSIONS

Experimental results have shown that the formation of freezed out impurities (hydrates) in the course of freezing out influences mainly the character of ice accretion arising on the heat exchanger winding. The structure of the ice accretion arisen has not been characterized by a compact ice layer but rather as a layer of a wet snow. The influence of impurities has manifested itself as a “richer“ form of ice accretion with the ability to catch more the liquid toluene in its structure in the same time period when compared to the freezing out of “pure” toluene (compare Fig. 2 and 3). The resulting needle-like ice accretion (consisting mainly of water) is able to catch the condensing toluene and in this way it is increased its specific mass but also thermal conductivity.

In the course of the final comparison of measured values with values calculated on the basis of simulation programme CHEMCAD it has been attained a very good conformity of single parameters observed.

The results have been then used for optimising industrial plant in the line for liquidating off gases from resin production plant.

CR08-49

MULTISTAGE CRYOGENIC TREATMENT OF MATERIALS: PROCESS FUNDAMENTALS AND EXAMPLES OF APPLICATION

Alava L.A.

Cryobest International, S.L. Vitoria, Spain

ABSTRACT

The multistage cryogenic process is an evolution from the conventional cryogenic treatments of materials. It needs shorter process time achieving the same or even better results.

This paper introduces some basic fundamentals of these treatments, their effects and applications, the equipment, etc. Some examples with different materials and from different industrial sectors are also presented as well as some brief comments about R&D and future trends of the technology.

INTRODUCTION

People usually relates heat treatments with high temperatures, but thermal treatments can also involve cooling. Although it has been traditionally considered that deep cold temperatures have no permanent effect on the materials, it is not true at all.

Heat treatments were already known and used centuries ago but the access to really low temperatures was only possible in relatively recent days. Although the first experiences took place at the beginning of last century, it is not possible to properly speak about industrial cryogenic treatments of materials until the 70’s when the liquefied gases became more affordable and the treating equipment had more accurate process control systems.

During the 80’s and 90’s the use of this technology increased and some treating facilities were opened, mainly in the US. Nowadays it is possible to find cryogenic processing companies in many countries all over the world.

Although still hardly known and used in Europe, the cryogenic treatment of materials is a technology that is slowly getting acceptance in industry. It basically consists in submitting the materials to deep low temperatures for increasing some of their performance characteristics like wear resistance or fatigue life.

1. THE PROCESS

Cryogenic treatments basically consist in submitting the materials to deep cryogenic temperatures (below 120K) following predetermined time-temperature curves in order to enhance some of their physical or structural properties.

According to the previous definition, it must be noticed that the subzero processes (at about -80 ºC) that are used in many traditional heat treating facilities to reduce the austenite content in some tool steels cannot be considered cryogenic treatments.

1.1 Conventional cryogenic treatments

There is no a standard process for cryogenic treatments but most of them are quite similar. In conventional cryogenic treatments the materials are slowly cooled down to a temperature around -180 ºC and maintained for a period of time that lasts from eight hours to two or even more days. After the soak, the materials are slowly heated up to ambient temperature. Sometimes the treatment is completed with a soft tempering. The entire process typically needs two to three days to be completed.

There is a conventional process sub-category called “wet process“ where the soak is made by submersion in liquid nitrogen. Anyway, the previously described “dry process“ (no liquid nitrogen in the chamber) is more widely used.

The cryogenic treatments are performed in chambers designed for this purpose. The material is usually cooled using liquid nitrogen that is introduced in the processor through solenoid valves controlled by computer. Most of the modern chambers have heaters that also allow to control the temperature during the heating phases of the process.

1.2 Multistage cryogenic process

The multistage cryogenic treatment is a more advanced process that has been developed as an evolution from the conventional ones. In this treatment the isothermal soak at cryogenic temperature is substituted by several cryogenic cooling/heating phases. This process is more effective but its main advantage is that it is much faster (an average of fifteen hours for the whole process) than the conventional ones.

The cryogenic chambers that are used to apply a multistage cryogenic treatment are specially designed to perform this type of process.

Figure 1: Multistage cryogenic processor designed and manufactured by Cryobest International, S.L.

2. THEORIES AND FUNDAMENTALS

The cryogenic treatments and their applications have been developed mainly in an empiric way. This technology is becoming widely accepted and used in industry but there are still certain controversy and some mystery concerning the effects of deep cryogenic temperatures on the material microstructure. Some studies have been performed in recent years in this field and more are currently in progress all over the world.

Metallurgists know that, when a steel is quenched, usually the higher the carbon content the lower the temperature (M_f) at which the transformation of austenite into martensite finishes. In high carbon steels, cold temperatures lead to higher contents of martensite and therefore to harder and more stable structures (more desirable in most of the applications). More recently it has also been confirmed that cryogenic temperatures promote the precipitation of fine η-carbide particles in the steel matrix contributing to improve the material characteristics.

These facts could explain some of the improvements in material performance due to its submission to a cryogenic treatment but, basically, they are only valid for steel. Although many grades of steel (alloyed, cold working, hot working, HSS, stainless…) can be cryogenically treated, the cryogenic processes have also evident effects in other materials like casting, cemented carbide, cooper alloys, aluminium alloys, titanium, some ceramics and even certain polymers. That means that there must be something else, more general, that explains the changes in the properties of the materials.

Recent theories point to stress relieving in the microstructural level and more stable, continuous and homogeneous lattices as a key factor to achieve improved behaviours in the service performance of cryogenically processed materials. This point of view is becoming quite accepted nowadays but it must be more deeply investigated.

3. EFFECTS AND APPLICATIONS

3.1 Cryogenic treatment effects

We have seen that there is a wide range of materials that can be cryogenically treated with good results. The effects of these treatments depend on the material but, in general, it is possible to obtain several of the following improvements:

- better wear resistance

- improved fatigue life

- stress relieving and dimensional stability

- increased conductivity

- improved machinability

- slight increase of hardness

- better corrosion resistance

Apart from the material, the results depend on the considered application. Of course, the cryogenic treatments are not a cure-all but there are innumerable situations where the improvements are significant or even impressive. Some examples will be commented later.

3.2 Industrial applications

It is possible to find applications in practically every industrial sector: machining, casting, injection moulding, forging, welding, automotive, aerospace, electronics, steel, timber industries, mining, agriculture, motorsports, etc. Some examples of parts that can improve their performance and increase their lives are: knives, cutting tools (drill bits, carbide inserts, mills, hobs, broaches…), saws, punches, dies, rolls, moulds, electrodes, gears, shafts, bearings, springs, cables… When a wear or fatigue problem occurs or more life is needed there is usually a good chance for using cryogenic treatments.

Cryogenic processes do not substitute conventional heat treatments although sometimes could be considered as an extension of them. One important characteristic is that they permanently affect the whole mass of the components, not just the surface. If, for example, a cutting blade is cryogenically treated, it can be sharpen as many times as desired without loosing its improved performance. Another point to be taken into account is that cryogenic treatments are fully compatible with most of the surface treatments and coatings (nitriding, PVD, CVD, etc.) that are commonly used in industry to enhance the working performance of tools and components.

It is important to remark that cryogenic treatments are environmentally friendly. Absolutely no waste or residues are produced during the process. And even more, the use of this technology allows for significant reductions in energy and materials consumption.

4. EXAMPLES OF APPLICATION

The number of applications of this technology is practically unlimited. Many of the most typical applications of cryogenic treatments are in the field of perishable tools. Machining tools are usually made of different grades of HSS or cemented carbide (and very often are PVD or CVD coated). These materials usually react very well to the application of a multistage cryogenic treatment. Furthermore, machining tools are often sharpened (and sometimes also coated) several times during their life. As this processes only need to be performed once this is a key advantage of this technology in many applications.

Cryogenic treatment is a good way to achieve important cost reductions and improved productivities in gear making processes. A well-known automotive supplier company can certify it. In one of the plants they manufacture steering systems for cars and trucks. The truck steering racks manufactured in this facility are machined with specially designed Maag type cutters (Fig. 3). These cutters are made of ASP2030 (powder-metallurgical HSS) and are coated with TiN every time they are sharpened.

Figure 2: Maag type cutters (TRW, Spain)

In this application the use of the multistage cryogenic treatment has made a clear difference. Without the treatment, an average of 60 parts were machined between sharpening. Nowadays all the new cutters are treated and the average production is 160 parts between sharpening. Not only this; as the wear is more homogeneous, it is only necessary to remove 0.2 – 0.3 mm. of material to sharpen the cutters (0.3 – 0.4 mm. without the treatment) what means about 30% extra uses of the tool. In this case the savings due to the use of the cryogenic processes are really important.

Another typical application of this technology is the treatment of gear cutting tools like hobs. A company specialized in the manufacturing of flywheel starter ring gears for the automotive industry uses inserted blade hobs. These hobs are made of HSS (M35) and coated with TiN. This kind of tools cannot be coated every time they are sharpened because the coating temperatures are too high for the resins that accurately fix the blades in its position. Nevertheless, the multistage cryogenic treatment avoids this problem and, in this case, the cryogenically treated hobs cut between 50 % and 100 % more ring gears than the untreated ones.

A well known aircraft manufacturer uses thousands of drill bits to make holes in difficult to machine materials like stainless steel, titanium or nickel alloys. After some months doing tests with HSS and solid carbide tools they could check that the life of the cryogenically treated drill bits was, as an average, three times longer than the untreated ones (in some cases even five times longer). Obviously, nowadays this company is achieving important tool cost savings thanks to the use of the multistage cryogenic treatment.

Figure 3: Insert blade hobs after a multistage cryogenic treatment.

Timber and wood industries as well as pulp and paper industry can also benefit from the use of multistage cryogenic treatments. Just as an example, the big knives that are used to convert the wood logs in small chips suffer severe wear. In most pulp mills these knives have to be grinded (after being previously removed) once and even twice a day. These knives are usually made of HSS and increase their life between two and three times when cryogenically treated. Taking into account that the treatment is applied only once, it is easy to understand that a growing number of companies is using this process to reduce downtime and tooling costs in their chipping facilities.

Figure 4: 760 mm. long wood chipping knife.

A forging plant is using the multistage cryogenic treatment for the hot forging dies. These tools are made of hot working steel X40CrMov51 (H13). The dies are nitrided to achieve a higher hardness value in the surface. As nitriding is a surface treatment it has to be repeated after any die regrinding or restoration.

This company made tests using the multistage cryogenic treatment and the results were clear: the untreated dies made an average of 1430 parts while the cryogenically treated ones were able to make an average of 3550 (the nitrided dies have a similar performance). The next step would be to test a combination of nitriding + cryogenic but, meanwhile, the forging dies are cryogenically treated instead of nitrided.

A well known manufacturer of bearings uses steel rolls to laminate and calibrate the outer rings of the bearings. These rolls can be used to calibrate an average of 3000 units before needing grinding. The cryogenically treated rolls can make as much as 17000! units before grinding. Of course, all the rolls are now submitted to the multistage cryogenic treatment.

But not only steel or carbide tools can be bettered through the use of cryogenic treatments. Copper alloys also respond very well to the process. The resistance welding electrodes are usually made of alloys like CuCoBe or CuCrZr that have a good balance between strength and conductivity.

The influence of the multistage cryogenic treatment in the performance of resistance welding electrodes has been tested in several applications finding performance increases up to 500% in some cases. That is why nowadays there are some companies that are already using this technology to reduce their welding electrodes consumption and to increase their welding quality and productivity.

Other copper alloys like brass are also suitable for the multistage cryogenic treatment. A manufacturer of rolled profiles uses forming punches and wear parts made of brass. After two years of tests, now all the brass parts that are used in the factory are submitted to the cryogenic process. The life of those components has increased from two to three times compared with the untreated ones.

5. TECHNOLOGY USE AND EVOLUTION

After some decades seeking legitimacy, it seems that the cryogenic treatments will probably become accepted and used in industry. Nowadays it is possible to find cryogenic treatment facilities in many countries all over the world, but the growth in the use of these technologies is slower than expected (the present situation of these technologies in Europe is clear example).

5.1 Research needs

Cryogenic treatment fundamentals are not totally understood and this fact can explain the delay in the complete acceptance of these processes. Fortunately this situation is already changing due to the evident results of the treatments. Anyway, there is a clear need of improving the research activity in this field in order to achieve a more complete understanding of the mechanisms that are involved in the microstructural enhancement of the deep cold treated materials and also in a search of new applications.

There has been few research activity in this field of and, traditionally, most of it has been made in North America apart from some studies made in Europe. But during the last few years this situation has changed and, nowadays, there are more R&D projects concerning cryogenic treatments of materials. And, nowadays, most of the research activity takes place in Asia.

The European commission has approved the first big cooperative project in the field of cryogenic treatments. Its title is “Improvement of automotive tools and components trough the application of deep cryogenic treatments”. This three year project was launched in October, 2007 and the participants are research centers, universities and industrial companies from Austria, Italy, Germany and Spain. Hopefully, more new similar project proposals will also be presented in the near future in Europe.

Anyway, the empirical development of the technology will probably prevail in the development of new applications of the cryogenic treatments during the coming days.

5.2 Cryogenic treatments application forecast

Until now the use of this technology in industry has been mainly focused on the cryogenic treatment of all kind of tools and consumables: knives, saws, mills, inserts, dies, punches, welding tips, moulds... This is probably the most evident use for the technology because the increase of tooling life has a direct positive consequence in productivity and cost, something that is always interesting for the users.

The increase in wear resistance is the prevailing treatment effect when treating tools. But there are other treatment properties that will probably get more relevance in the coming years and one of them has special interest: the increase of fatigue life.

Probably, during the coming years there will be more and more applications where the cryogenic treatments will be used to increase the fatigue life of all kind of components. Apart from the tools the process will be applied to the manufactured components in order to increase their service life and their reliability. If this forecast is right, there will be necessary to treat much more quantities of materials. This would be a situation where the multistage cryogenic treatment has a clear advantage compared with the conventional ones: the process time is much shorter and, consequently, the treating facilities are much more productive.

CONCLUSIONS

The treatment of materials at cryogenic temperatures is a promising and cost effective technology that, although is not new, is still hardly known and used. Only a small part of its potential has been developed.

The metallurgical fundamentals are not fully understood and the cryogenic treatments have been, and still are, developed in an empirical way. During the last years the research activity in this field has significantly increased and nowadays it is also more global than in the past.

The multistage cryogenic treatment is an evolution from the conventional ones. It is more efficient and its process time is much shorter.

The wear resistance improvement and the increase in fatigue life are just two of the effects of the cryogenic treatments. A wide range of materials can be cryogenically treated and the number of applications is unlimited. They can be found in every industrial sector.

There are two main fields of application of cryogenic treatments:

- the treatment of tools is nowadays the more common application. It is a cost effective way to increase their productivity allowing less downtime and reduced costs.

- cryogenically treating components (bearings, gears, shafts, springs…) it is possible to greatly improve their performance and reliability and also to reduce their weight and size. This field of application will probably have a big increase in the future.

The multistage cryogenic treatment is an environmentally friendly technology that helps the materials to perform better. No doubt cryogenic treatments are an exciting technology that holds much future promise.

REFERENCES

1. Meng, F., Tagashira, K., Azuma, R. and Sohma, H., Role of Eta-carbide Precipitations in the Wear Resistance Improvements of Fe 12Cr-Mo-V-1.4C Tool Steel by Cryogenic Treatment, ISIJ International (1994) vol. 34, No 2 205-210.

2. Diekman, R., Cryogenics in the Thermoset Molding Industry, Thermosettings (2001) vol. 29.

3. Schiradelly, R. and Diekman, F.J., The Racer’s Edge, Heat Treating Progress (2001) nov. 43-50.

4. Molinari, A., Pellizzari, M., Gialanella, S. Straffelini, G. and Stiasny, K.H., Effect of deep cryogenic treatment on the mechanical properties of tool steels, Journal of Materials Processing Technology (2001) 118 350-355.

5. Huang, J.Y, Zhu, Y.T., Liao, I.J., Bourke, M.A. and Mitchel, T.E., Microstructure of cryogenic treated M2 tool steel, Material Science Engineering (2003) A339 241-244.

6. Zhiseng, W., Ping, S., Jinrui, L. and Shengsun, H., Effect of deep cryogenic treatment on electrode life and microstructure for spot welding hot dip galvanized steel, Materials and Design (2003) 24 687-692.

7. Huang, M.C., Gao, C.H., and Huang, L.G. Study on cryogenic phase change & wear characteristic of high speed steel, Acta Metallurgica Sinica, (2003) vol. 16, No. 6 524-530.

8. Manoj, V., Gopinath, K. and Muthuveerappan, G., Rolling contact fatigue studies on case carburized and cryogenic treated En 353 gear material, International Sympossium on Material Science and Engineering, Chennai (2004).

9. Singh, P.J., Mannan, S.L., Jayakumar, T. and Achar, D.R.G., Fatigue life extension of notches in AISI 304L weldments using deep cryogenic treatment, Engineering Failure Analisys (2005) 12 263-271.

10. Yong, A.Y.L., Seah, K.H.W., Rahman, M. Performance evaluation of cryogenically treated tungsten carbide cutting tool in turning, International Journal of Machine Tools& Manufacture (2006) 46 2051-2056.

11. Latas, Z., Ciski, A. and Suchmann, P., Cryogenic Treatment and Combination of Nitriding and Cryogenic Treatment of Hot Forging Tools, Proceedings of the 4^th WSEAS International Conference (2006) 133-139.

12. Zhirafar, S., Rezaeian, A., Pugh, M., Effect of cryogenic treatment on the mechanical properties of 4340 steel, Journal of material processing Technology (2007) 186 298-3

CR08-06

From the tissue bank to The tissue establisment

Měřička P., Straková H., Horynová A.

Tissue Bank, University Hospital, Hradec Králové, Czech Republic

Abstract

A review of changes in a role of a tissue bank in assuring clinical cell and tissue transplantation is presented. At the beginning the tissue bank was regarded a place to which the tissue collected by a surgeon was put until its use, mostly by the same physician. The role of a tissue bank was only to extend as much as possible the shelf life of the preserved tissue. Later the tissue banks started to introduce methods modifying the properties of the original tissue with the aim to lower its immunogenicity as well as to enhance or to prevent tissue rebuilding after transplantation. The chance to overcome the tissue rejection was enlarged after introducing methods combining the autologous cultured cells, e.g. epithelial keratinocytes with allogeneic or biosynthetic matrices. For a long period of time the activities of cell and tissue banks were not regulated by law, only the standards of voluntary organizations of tissue bankers have been available. The term tissue establishment was introduced by the Directive of the European Parliament and Council issued in 2004, that put high requirements on the safety and quality of the cell and tissue processing and banking procedures. The authors demonstrate the results of their effort to meeting these requirements at their workplace, Tissue Bank of the University Hospital, Hradec Králové.

1. Introduction

At the beginning of cell and tissue banking there was no regulation of these activities neither by voluntary standards nor by state health care authorities. The voluntary standards were first formed by the American Association of Tissue Banks in the 70´s, regulation by state authorities by the issue of the FDA interim regulations in 1993. In Europe the process was started later by standards of the EATB in 1993, issue of the Convention of Biomedicine in 1997, by recommendation of the European Council in 2002 and last but not least by the issue of the Directive of the European Parliament and Council in 2004 and of the European Commission Directives in 2006. Simultaneously the scope of tissue banking activities has been changing. Originally the tissue bank was regarded a place, where the collected tissue was stored for some time until its clinical use. The role of the tissue bank was to extend as long as possible the shelf life of the stored tissue. In living cells and tissues extending their shelf life was not possible without cryopreservation that has been available since the discovery of the cryoprotective action of glycerol in 1949. Later the tissue banks started to introduce methods modifying the properties of the original tissue with the aim to lower its immunogenicity and to enhance or prevent tissue rebuilding in the organism of the host. Different kinds of tissue grafts started to be prepared by tissue banks in large quantities, covering the needs of a particular region, or of the whole state. Because of these new approaches as well as of increased requirements on safety of the product the tissue banks needed to cover more activities than storage of tissues itself and developed into establishments with precisely defined requirements on premises, equipment, personell, control laboratories and their cell and tissue collection centres.

2. ORIGINAL project of the tissue bank, its definition and early history

The Tissue Bank of the University Hospital Hradec Králové was established in 1952 by prof. R. Klen on basis of his own concept presented at the occasion of the 14th Congress of the Czechoslovak Society for Orthopaedic Surgery and Traumatology in 1951 and published one year later (Klen 1952). Prof. Klen, who was in the position of the head of the department till 1984, defined the tissue bank as a facility specialized to collection, processing, preservation, storage and distribution of different kinds of tissues collected both in living and deceased donors and used for clinical or experimental purposes (Klen 1952). This concept was different from that of existing monobanks oriented to preservation of one tissue and being regularly parts of clinical departments. At the time of establishment the bank covered the needs of the departments of ophthalmology and orthopaedic surgery that performed cornea (Klen 1954) and bone transplantations (Lány, 1954). Cornea and/or the eye bulb was preserved in a moist chamber at +4°C. In bone tissue the low temperature preservation at –20°C was preferred to chemical preservation in merthiolate used predominantly in that time. Later the use of solid tissue grafts was started in neurosurgery ( Króo and Klen 1960, burn medicine (Klen et al. 1967) , stomatosurgery and face surgery. After moving of the bank to new premises at the building of Institute of Pathology in 1967 the technological background was enlarged, and freezing of tissues to –50°C and freeze- drying followed later by radiation sterilization were introduced(Klen et al 1977). Freeze-dried grafts were distributed to many clinical departments on the whole territory of Czechoslovakia and sent also abroad . Since the beginning of the 70´s cryopreservation methods bave been elaborated for storage of cells and tissues in a viable state. In the 80´s and 90´s regular programme of autologous and allogeneic transplantation of haematopoietic progenitor cells was introduced (Měřička et al. 1991, Bláha et al.1990). Simultaneously the methods of pretreatment of solid tissue grafts prior to freeze-drying leading to modifying of the process of their rebuilding in the organism of the host were introduced, such as glutaraldehyde stabilization in xenogeneic freeze-dried, radiation sterilized pericardium used in reconstruction of dura mater in neurosurgery (Měřička et al., 1986, Pařízek et al.1989) and demineralisation of allogeneic freeze-dried cancellous bone for use in stomatosurgery and dental implantology (Šimůnek et al. 1997). The burn surgeons used besides allogeneic skin grafts a big quantity of pigskin grafts as a temporary skin cover (Klein et al.,1995 Měřička et al.1995). In the beginning of the 90´s the culture of autologous epidermal keratinocytes as a permanent skin cover for use in burns or chronic skin defects was introduced under the financial support of the grants of Ministry of Health and of the Ministry of Defence (Klein et al.1997, Měřička et al. 1996-1997). In the second half of the 90´s this programme was replaced by culture of autologous chondrocytes for reconstruction of articular surface (Folvarský et al. 2002, Pavlata et al. 2003).

3. The legislative requirements on Tissue establishments set by the European Union And results of their implementation into the national legislation of the Czech republic

In the 90´s of the last century there was a rapid development of tissue banking and engineering technologies, there was also a rising criticism, however, focused on the following issues:

1. Respect of the cell and tissue donor rights.

2. Safety of the cell and tissue transplantation.

The solution for the first problem was included in the Convention of Medicine and its Additional Protocol.

The new high requirements on quality and safety of cell and tissue transplants were set recently by the European Parliament and Council Directive 2004/23/EC (European Union, 2004) as well as by the European Commission Directives 2006/17/EC and 2006/86/EC . All directives are to be implemented into the national legislation within 2 years after their issue. The National Transplantation Act approved in 2002 (Ministry of Health of the Czech Republic, 2002) included already the principles of voluntary and unpaid donation settled by the Directive 2004/23/EC. This act introduced a strong protection of the autonomy of the living or deceased donor as required by a Convention on Biomedicine (European Treaty Series, 1997). In living donors informed consent was introduced in all cases including harvest of surgical residues, e.g. femoral heads removed during hip-joint plasties. In deceased donors the presumed consent was preserved, the national register of persons rejecting post-mortem donation was established, however, and information of the next-of-kin became obligatory. This Act also defined the duties of tissue banks. The safety aspects were oriented to defining the contraindications of the cell, tissue and organ donation and traceability assurance. The new Act on Cells and Tissues of the Human Origin that sets the quality and safety norms for use of banked human cells and tissues in the clinical practice is expected to come into force in 2008. The main tool for achieving rapid implementation of the new quality and safety requirements into daily practice of tissue banks using the existing national legislation was starting the licensing process in 2004. The requirements set by the Ministry of Health in 2004 included compliance of the tissue banks system of work with the Directive 2004/23/EC and with the National Transplantation Act. In the years 2006 and 2007 when the licences were renewed, the compliance with the Directive 2006/17/EC and 2006/86/EC was required.

4. METHODS OF MEETING THE REQUIREMENTS IN THE TISSUE BANK

The similarities between tissue banking processing distribution and quality control practice and practice of manufacturing and control of sterile drugs were recognised by authors already in the 80´s ( Měřička 1983, Měřička et al. 1990). In that time it was obvious that meeting of these criteria was not possible within the tissue bank premises that were at the disposal of the authors. As the result of analysis of trends of world and European legislation the author proposed a radical rebuilding of the Tissue Bank of the University Hospital Hradec Králové in accordance with standards of the International Society for Pharmaceutical Engineering (The Society for Pharmaceutical and Medical Devices Professionals, 1999). A new concept of the cell and tissue bank as combination of cryogenic and clean-room technology was proposed and the bank was designed and built with the financial support of the Ministry of Health of the Czech Republic and of the University Hospital Hradec Králové in the years 1998-2002 (Měřička 2006,2007). The cryopreservation and storage facility with the aseptic processing rooms started operation in 2003, the new freeze-drying facility in 2007. Simultaneously the internal and external cell and tissue collection centres were established and their duties were defined. The collaboration with control laboratories and companies authorized to perform validation of the equipment and of clean rooms both at rest and operation conditions was settled as well. The bank was granted the multifunctional licence by the Ministry of Health in 2004. In the years 2006 and 2007 this broadest type of licence was regranted. In the year 2007 the model inspection of the State Institute for Drug Control, Prague was performed in the bank within the European Union twining project of this institute with AFSSAPS, France and MHRA, Great Britain.

5. Current organisation of the Tissue Establishment and results of its activity

The Tissue Establishment of the University Hospital Hradec Králové consists of internal and external cell and tissue collection centres, clean rooms serving for aseptic processing before cryopreservation or freeze-drying of tissues, cryostorage and freeze-drying facility and control laboratories. The number of living cell and tissue donors is presented in the Tables I and II. The number of deceased solid tissue donors is presented in the Table III. The number of solid tissue grafts delivered for clinical application is presented in the Table IV.

Table I: Number of living cell and solid tissue donors

Year	Living cell donors		Living solid tissue donors
	Autologous use	Allogeneic use	Autologous use	Allogeneic use
2004	95	14	15	84
2005	117	18	6	110
2006	80	10	10	140
2007	66	11	13	174

Table II: Number of living cell donors

Year	Autologous haematopoietic progenitor cells	Allogeneic haematopoietic progenitor cells	Cultured autologous chondrocytes	Sperm
2004	57	15	17	21
2005	66	14	26	25
2006	34	8	21	25
2007	22	6	22	22

Table III: Number of deceased solid tissue donors

Year	2004	2005	2006	2007
Number	16	19	17	20

Table IV: Number of preserved solid tissue grafts delivered for clinical transplantation

Year	Musculoskeletal tissue	Fascia lata
2004	231	165
2005	249	126
2006	288	138
2007	245	156

6. Discussion

Despite of the fact that the definition of the tissue establishment included in the Directive 23/2004EC does not differ substantially from the original definition of the tissue bank (Klen 1952), the legislative changes that were initiated by the issue of the Directives represented the most radical change in the position of tissue banks in the Czech Republic in their more than 50 years old history. The Transplantation Act made it possible to start and to run the tissue bank licensing process immediately after the issue of the Directives. Implementation of the principles of voluntary and unpaid cell and tissue donation led to considerable changes in the number of tissue donation (Tables I, II, III). From the tables it is clear that there was a considerable increase of the living solid tissue donations while the number of post-mortem donations remained relatively low. To secure satisfactory sources of the bone tissue collected in living donors it was necessary to establish external collection centres on a contract basis in four county hospitals. In post-mortem tissue donations a basis of the hospital donor management system was introduced in cooperation with the Regional kidney transplantation centre. On this basis all necessary measures are performed including contacting the donor´s family.

As the Transplantation Act was accompanied by a decree defining the conditions for health suitability of the donor, donor screening and serological testing (Ministry of Health, 2004), the implementation of the Directive 2006/17/EC did not represent a major obstacle for the majority of banks including our establishment.

Similarly the strict requirement of keeping traceability of the way from the donor to the host of the tissue graft did not represent any change of existing practice as this approach has been applied since the very beginning of the activity of the Tissue Bank of the University Hospital Hradec Králové (Klen 1954). The same case is a continuous monitoring of clinical results (Klen 1957, Měřička 1983) including registration of adverse reactions or graft failures.

On the contrary meeting of the requirements of the Directive 86/2006/EC setting high demands on the technical background of the tissue bank was not possible without radical rebuilding of premises of existing cell and tissue banks. It was proved on the example of our establishment that a concept of a tissue establishment based on combination of cryogenic and clean-room technology introduced as a priority in the Czech Republic represents an efficient tool how to meet the requirements of this directive. By strict keeping aseptic conditions during cell and tissue collection and processing of cell and tissue grafts a real alternative was established to the use of terminally sterilised grafts, which is important in the instances where terminal sterilisation impairs their mechanical and/or biological properties. Establishing of the system of internal and external quality control is an additional new feature of the practice of tissue banks as in the past the quality control was made predominantly by the tissue bank itself. The system is based on separating of personal responsibility for production and for the control, that is performed in majority of cases by laboratories not being a part of the tissue bank. The results of all control tests made by these laboratories are evaluated by a person responsible for release of the grafts for clinical application. This person, who must be an employee of the tissue bank, is nominated by the Ministry of Health and his/her name is included in the text of the licence. To assure higher involvement of the state in the graft safety and quality regular inspections of tissue banks are planned to be performed in the near future by the State Institute of Drug Control.

7. Conclusion

1. Despite of the fact that the definition of the tissue establishment included in the Directive 23/2004/EC does not differ substantially from the original definition of the tissue bank published in the 50´s, the legislative changes that were initiated by the issue of the European Union Directives represented the most radical change in the position of tissue banks in the Czech Republic in their more than 50 years old history.

2. Implementation of the principles of voluntary and unpaid donation into the practice of the tissue bank led to increased use of solid tissues collected in living donors.

3. The concept of a tissue bank based on combination of cryogenic and clean-room technology represents the useful tool to meeting the requirements of the Directive 2006/86/EC. This concept introduced by the authors as a priority in the Czech Republic is now being followed by other tissue establishments in the country.

8.References

Bláha M, Maisnar V, Jebavý L, Měřička P, Rondiak J, Šulc K, Vaňásek J. jr. 1990, Transplantation of blood progenitor cells collected from peripheral blood - first experience, Voj. Zdrav. Listy, LIX:218-223.
Bláha M, Měřička P, Žák P, Štěpánová V, Vávra L, Malý J, Toušovská K. 2003, The Risk of Infection Transmission from Blood Progenitor Cell Concentrates, J. Hemotherapy and Stem Cell Res., 12:161-164.
Convention for Protection of Human Rights and Dignity of the Human Being with Regard to the Application of Biology and Medicine: Convention on Human Rights and Biomedicine – Oviedo, 4.IV.1997 – European Treaty Series/164..
European Union, 2004, The European Parliament and the Council: Directive 2004/23/EC of the European Parliament and of the Council on Setting Standards of the Quality and Safety for the Donation, Procurement, Testing, Processing, Preservation, Storage and Distribution of Human Tissues and Cells, Strasbourg, 2004, 41 p.
Folvarský, J., Dědek,T., Dobeš, D., Frank, M., Adler, J.:Treatment of deep cartilage defects in the knee by mosaicplasty combined with autologous cultivated chondrocytes in Tissucol, 2002. European Journal of Trauma, 28, Suppl. P. 43
Klen, R. 1952, Tissue Centre (TC), Čs. nemocnice, 20:120-121 – in Czech.
Klen,R.: Collection and storage of grafts for transplantation,1954. In.Group of authors: Tissue Transplantation. SZdN, Prague, 13-21 (In Czech)
Klen, R. et al.: Experimental study of factors influencing take and survival of human skin homografts, 1968. Plast. Reconstr. Surg. 41,p.471-476
Klen, R., Metelka,M., Pařízek, J., 1977 : Freeze-dried homogeneous grafts of fascia lata in neurosurgery. J. Neurosurg. Sci 21/4, p.247-250
Klein, L., Měřička, P., Preis, J.,1995:Clinical experience with skin xenografts in burned patients. In:Maselis,M.,Gunn,S.W.A. (eds.) Proceedings of the Second International Conference on Burns and Fire Disasters:Perspectives 2000.-I.st ed., Palermo, Kluwer Academic Publishers, p.343-345
Klein, L., Měřička, P., Straková, H., Jebavý, L., Nožičková, M., Bláha, M., Talábová, Z., Hošek, F.:Biological skin covers in treatment of two cases of the Lyellś syndrome 1997. Annals of Transplantation, 2, p.45-48
Króo, M. and Klen,R.:Our experience with the dura mater replacement with frozen fascia and dura grafts, 1960. Lék.Zpr. LF UK v Hradci Králové, p. 80-85. In Czech
Lány, J.: Use of bone grafts in orthopaedic surgery,1954. In. Group of Authors: Tissue Transplantation. SZdN, Prague, p. 32-37 (In Czech).
Měřička, P.: Quality Control of Freeze-dried Tissue Grafts.In.: Proc. 16^th Int. Congr. Refrig., Paris 1983, Fr.3,1984, p.161-165
Měřička P, Vávra L, Hušek Z, Straková H, Špaček J. Pařízek J, Klein L, Levínská M.1990, Low Temperature Preservation of Tissues for Clinical Use. In: Bose A, Sengupta P (eds.) Advances in Cryogenics, Proceedings of the International Conference on Cryogenics, Calcutta, India, December 6-10,1988 (INCONCRYO), VII.Cryogenics in Life Sciences and Medicine, Surgery, Agriculture, McMillan, India Ltd., New Delhi, p.625-644.
Měřička P, Schustr P, Vinš M, Dudek A, Vávra L, Červinka M, Rondiak J. 1991, Containers for Freezing and Storage of Bone-Marrow Stem-Cells, Sb. Věd. Prací Lék. Fak. Karlovy Univ. Hradec Králové, 34:367-387.
Měřička, P., Klein, L., Preis, J., Ettlerová, E. 1995:The role of the tissue bank in disaster planning. In:Maselis, M., Gunn, S.W.A. (eds.) Proceedings of the Second International Conference on Burns and Fire Disasters:Perspectives 2000.-I.st ed., Palermo, Kluwer Academic Publishers, p.75-83
Měřička, P., Straková, H., Klein, L., Šubrtová, D: Practical aspects of establishing an allogeneic human keratinocyte bank 1996-1997.Roczniki oparzen, Annals of burns 7/8, p.105-109
Měřička P. 2000, Brief History of the Tissue Bank, Charles University Hospital, Hradec Králové, Czech Republic, Cell and Tissue Banking, 1,p.17-25.
Měřička P, Bláha M, Vávra L, Štěpánová V. 2003, Our System of Cross-contamination Prevention During Storage of Haematopoietic Progenitor-Cells - paper ICR 0302, In: Grof, G., Menzer, M. (eds): 21^st IIR International Congress of Refrigeration, Washington 2003.
Měřička P, Straková H, Čermák P, Štěpánová V, Hradecký Z, Drahošová M. 2002, New Safety-Assurance for Biological Skin Covers, Acta Chir. Plast., 44:23-29.
Měřička P, Vávra L, Vinš M, Schustr P. 2004, The Importance of Oxygen-Level Monitoring in the Cryostorage-Facilities In: Chrz V (ed.), The Eighth Cryogenics 2004, IIR International Conference: Refrigeration Science and Technology. International Institute of Refrigeration, Paris, Paper C 04-08.
Měřička P, Straková H, Vávra L, Schustr P, Vinš M, Popíšil J, Postupa J, Plasová B. 2004, The Cell- and Tissue- bank as a Combination of Cryogenic and Clean- room Technology. In: Programme and Abstracts International Congress of the European Association of Tissue Banks, October 13-16, 2004, p. 69.
Měřička P. 2006, Contribution to Safety Assurance in the Cryopreservation of Cells and Tissues used for Clinical Transplantation. Doctoral (Ph.D.) Thesis Summary, Nucleus HK, 34 p.
Měřička P., Bláha, M., Straková, H., Navrátil, P., Grófová, M., Počepcov, I., Folvarský, J., Trlica, J., Žvák, I., Šmejkal, K., Dědek,T. 2007: New Safety Assurance for Processing and Cryopreservation of Cell and Tissue Grafts used for Clinical Transplantation. Transplant International 20 Suppl.2, p.325.
Ministry of Public Health of Czech Republic, Decree number 437/2002 Code of law, from October 3^rd, 2002, which establishes in detail conditions for health suitability and extent of screening and testing of living or deceased tissues or organ donor for the use of transplantations. (Decree about health suitability of tissue and organ donors for the use of transplantations), Code of law number 437/2002 Part 153, pp. 8221–8223 (in Czech).
Ministry of Public Health of Czech Republic, Act number 285/2002 Code of law from May 30^th, 2002 about donation, recovery and tissue and organ transplants and about change of other Acts (Transplantation Act). Code of law of the Czech Republic No. 285/2002 Part 103, p. 6050 (in Czech).
Pařízek, J. Měřička, P., Špaček, J.,Němeček,S., Eliáš,P.,Šercl,M.,1989: Xenogeneic pericardium as a dural substitute in reconstruction of suboccipital dura mater in children. J. Neurosurg. 70,p. 905-909
Pavlata, J., Urban, K., Karpaš, K.,Měřička, P., Straková, H., Brtková, J 2003:. Reconstruction of the joint surface. Acta Medica (Hradec Králové), 46, p.57
The Society for Pharmaceutical and Medical Device Professionals - ISPE: Baseline. Pharmaceutical Engineering Guide. Pharmaceutical Engineering Guides for New and Renovated Facilities. Volume 3. Sterile Manufacturing Facilities, First Edition/January 1999. ISPE Headquaters, ISPE European Branch Office, 1999, 162 p.
Šimůnek, A.: Authorś own experience with using of freeze-dried demineralised bone in dental implantology 1997. In: Měřička, P., Straková, H., Stacey, G (eds.): Cryoprotectants in Medical Practice. Long Abstracts. International Institute of Refrigeration, Paris p.47

CR08-32

Ventilation of cryostorage facilities of tissue establishments

Lain M.¹, Měřička P.², Dvořák J.²

¹Technical University in Prague, Mech. Eng. Faculty, Prague, Czech Republic
² Tissue Bank, University Hospital, Hradec Králové, Czech Republic

Abstract

The liquid nitrogen is commonly used for long-term storage of viable cells and tissues in the cryostorage facilities of tissue establishments. As frequently large biological containers are used the continuous evaporation of nitrogen during storage as well as evaporation during refilling containers can lead to decrease of the air oxygen level below acceptable limits (20% for women, 18% for men - according to the Czech law and recommendation of producers). For this reason sufficient ventilation of such facilities is essential for safety of the staff. Ventilation should also eliminate the possible leak of liquid nitrogen in extraordinary situations. The character of nitrogen production is non-stationary due to above mentioned situations. The authors present the basic model based on non-stationary pollutant production. The nitrogen production in the model is calibrated according to on site measurements, and can be used for prediction of specific operation and emergency scenarios. The principles and recommendations for cryogenic storage room’s ventilation systems, its control and environment monitoring equipment is also presented.

Introduction

One of the main reasons for ventilation of spaces is maintaining the pollutants in the room within permissible concentration. In the cryostorage facilities of tissue establishments large quantities of liquid nitrogen are frequently used for long-term storage of viable cells and tissues. The continuous evaporation of nitrogen during storage as well as evaporation during refilling containers is the main source of the contaminant. The increasing concentration of nitrogen is not dangerous due to nitrogen itself, but due to decrease of the air oxygen level below acceptable limits.

The influence of oxygen level on man’s health can be classified in 4 stages:
1^st stage - oxygen level 20-14 % - changes in concentration, accelerated pulse.
2^nd stage - oxygen level 14-10 % - people stay conscious, loss of discernment.
3^rdstage - oxygen level 10-6 % - belly-ache, loss of movement control, swoon.
4^th stage - oxygen level below 6 %, spontaneous breathing stops, death in few minutes .

Up to 3^rd stage, it is difficult for unexperienced people to recognize any problem, there is not any smell or other warning. It can be very dangerous in nitrogen polluted places. In the hygienic standards usually 18 % is the bottom limit for oxygen level in working places, accordinbg to the Czech standard 288/2003. The bottom limit for oxygen level 20 % is set for women and teenagers.

1. The model

1.1. Contaminat concentration model

Textové pole:
Figure 1: Contaminant concentration balance The contaminant concentration model is based on the contaminant balance of the ventilated space under presumption of complete mixing of the air in the room (Fig.1).

(1)

M_Š          contaminant production [g/s]
V_p            volume flow rate [m³/s]
V_R           space volume [m³]
C             contaminant concentration [g/ m³]
t              time

Textové pole: (2) There are two possible solutions for non steady contaminant production or ventilation flow rates. The operation time can be divided into periods with constant productions and ventilation and commonly known analytical solution of equation (1) can be applied. For finite time intervals equation (1) can be directly applied. The presented model is based on equation (2) for nitrogen concentration. The oxygen level is calculated from nitrogen concentration.

1.2. Nitrogen production

Textové pole:

Figure 2: Model calibration and verification results There are two sources of contaminant in the storage room: The continuous evaporation of nitrogen from large biological containers and evaporation of nitrogen during refilling. In the Tissue Bank in University Hospital Hradec Králové, there were following containers at the time of our measurements:
2 x MVE XLC 1200,
1 x Cryocyl 230LP,
2 x MVE XLC 230,
1 x MVE Cryosystem 2000,
2 x Cryometal KL32,
1 x Eurocyl 230LP,
.
According the data of the producers of containers the total nitrogen evaporation is 22,5 m³/day.

1.3 Ventilation system

The existing ventilation system supplied constant amount of 1720 m³/h of conditioned fresh air into the storage room which was equal to the air exchange rate 8 /h (8 ACH). The supply of the fresh air should be close to the breathing zone of people, and the exhaust should be close to floor and at the largest nitrogen source (refilling valve).

1.4 Model calibration

The model vas calibrated (Fig. 2a) according to measured oxygen level in the room. Results of calibration are the nitrogen productions during the refilling and steady state. Finally according to calibration the 49,4 l/day of liquid nitrogen is evaporated and 1,54 l/min is evaporated during refilling. The calibrated model vas verified (Fig. 2b).

2. Tested cases

2.1 Controlled ventilation

Textové pole:
Figure 3: Oxygen level for 2 stage flow rate:
just storage - 1 ACH (left),
during refilling – 8 ACH (right). The various scenarios were tested on the model. To save energy the two stage control of the fresh air flow rate was recommended. During refilling and after it full flow rate (8 ACH) and during common operation reduced flow rate (1 ACH). The calibrated model was applied to check the oxygen level for such an operation (Fig.3).

2.2 Emergency scenarios

The model was applied to test various emergency scenarios as well. The oxygen level was modelled for a long-time maximum nitrogen production, that usually happens just during refilling period. The time of the modelled emergency situation was much longer than the normal duration of the re-filling period, which typically is 15 min. (Fig.4).

Textové pole:
Figure 4: Oxygen level for emergency situation 8ACH - left, 1 ACH - right.

conclusionS

The proper ventilation flow rate is very important for rooms with the cryostorage facilities of tissue establishments. The detailed model of the oxygen level can be very useful for system design and operation. In generally, two steps control of fresh air flow rate can be recommended. The low flow rate for just the storage period and the high flow rate for the refilling period. In the emergency situation the high ventilation is important as well.

REFERENCES

1. Dvořák, J.: Větrání prostor se zdroji dusíku, diploma work, CTU in Prague, Mech. Eng. Faculty, Department of Environmental Engineering, 2006.

2. Měřička, P., Vávra, L., Vinš, M., Schustr,P..: The importance of oxygen level monitoring in the cryostorage facilities C04-08). In: The Eight Cryogenics 2004 IIR International Conference .1^st edition. Refrigeration Science and Technology Proceedings, Paris, International Institute of Refrigeration, 2004, p.242-247

This research was supported by research plan MSM6840770011.

CR08-25

Special equipment for cryopreservation of tissue in a standard freezing unit

Spörl G.¹, Klingner E.², Quinger J.³

¹Institute for Air Conditioning and Refrigeration, Dept. of Applied New Technologies, Bertholt-Brecht-Allee 20, D-01309 Dresden, Germany
²edecto GmbH, Erlweinstraße 9, 01069 Dresden, Germany
³ Ingenieurbüro und Plastverarbeitung Quinger GmbH, Schwarzer Weg 7, 09557 Flöha

ABSTRACT

With forthcoming efforts in tissue engineering tissue banks will be established. For freezing tissues with different sizes standard freezing units will be taken up. Hence the effectiveness of such a unit was investigated. Results: The temperature homogenity was not sufficient for developing cryoprotocols for tissues. Commonly used vials and Petri dishes are not suited for storage of tissues during the cryopreservation procedure. Special hardware was developed to overcome this deficit.

introduction

Cryopreservation is a well established technique for long time storage of living biological materials, especially of cell suspensions. According to the-state-of-the-art of science and technology these suspensions are transferred to cryovials or bags (e. g. blood). For these cryopreservation can be done in two major ways. First, for small amounts of vials “Qualifreeze” vessels from VWR or “Mr. Frosty” vessels from Nalgene are often used. They guarantee a cooling rate of about 1°C per minute when filled with isopropanol and placed in a freezer (-80°C). [1] Second, controlled rate freezing can be carried out in the gaseous phase of an automatic Standard freezing unit. In this case the vials or bags are placed in special racks which allow controlled freezing and additional seeding if necessary. [2] For long time storage the vials and bags are transferred into a liquid nitrogen storage vessel.

Cryopreservation of living tissues is under way not only because of the forthcoming efforts in the field of tissue engineering. All in all regenerative medicine is in progress. In contrast to cell suspensions living tissues, especially tissue engineered constructs consist of a scaffold, seeded with different cell types, and are well defined in terms of the upper and lower surface. Therefore turning around of the tissue must be prevented. Furthermore the tissue area can be larger than 1 cm². [3] All these aspects refer to lacking applicability of vials for tissue cryopreservation. For that reason special equipment was developed for usage in a standard freezing unit.

1. Basic investigations on a freezing stage

1.1 Investigations of thin carrier for the supervised development of a cryoprotocol

At the beginning of our research, we have observed the behaviour of cells within a scaffold during cooling down to 120 K at a freezing stage (LINKAM). Therefore the constructs are placed on a glass cover slip which was mounted on the surface of the silver block of the freezing stage. In order to qualify the visual results thermocouples were placed at several points on the sample and on the stage (see figure 1, right). In respect to further application of the cryoprotocol glass cannot be used in practice. So other materials for cover slips were tested. The thickness of the glass cover slips ranges from 0,1 to 1 mm. Plastic cover slips were 1,1 mm thick. All measurements were done with the same cryoprotocol: 1. cooling with 5 K/min down to 273 K; 2. equilibration for 20 minutes;3. further cooling with 1 K/min down to 268 K; 3. equilibration again for 5 minutes; 4. continue cooling with 1 K/min down to 233 K; 5. final equilibrate for 5 minutes and 6. rewarming with different rates to 300 K. The appearance of spontaneous ice nucleation was used as a measure for the heat transfer from the silver block to the upper side of the tissue. Further cooling down to 120 K didn’t gave much more information concerning the first ice formation.

Tissue on cover slip

Silver block Thermal couples

Figure 1: Freezing stage on the microscopic stage (left)
freezing stage interior, sample preparation (right)

1.2 Results

At first very thin cover slips were used. The results are shown in figure 2. The temperature gradient between the silver block and the upper side of the tissue ranges from 5 to 7 K.

Figure 2: Temperature profiles of thermal couples, placed at different points (see fig. 1), using 0,1 mm thick glass cover slips

The spontaneous nucleation causes a very small anomaly because the latent heat can be dissipated quite well through the very thin cover slip. Thicker sample holders caused worse results. The worst values we got when using the plastic cover slips (see figure 3).

Figure 3: Temperature profiles of thermal couples, placed at different points (see fig. 1) using 1,1 mm thick plastic cover slips

Here, the temperature gradient between the silver block and the upper side of the tissue varies between 8 and 15 K. The spontaneous nucleation causes a very broad anomaly because the latent heat cannot be dissipated well through the thick plastic cover slip.

In order to manage the spontaneous ice formation, additional cold (about 80 K) was applied by cold forceps. The latent heat could be reduced or diminished at a large extent. But in case of tissue larger than 1 cm² the centre of it was still destroyed. The way out was found by a device, which brings cold inside the tissue “hedgehog”). A satisfying cryoprotocol could be developed as the proof of principle.

Cryopreservation of biological materials requires high cell respectively product viability and integrity at point of use. For the integration of this freezing process in a tissue bank, first the cryoprotocol has to be transferred to an automatic standard freezing unit. These units were developed for a controlled rate freezing with high accuracy and repeatability.

As mentioned before the temperature distribution within the tissue must be extensively homogeneous for a successful cryopreservation. For that reason the temperature distribution in the cooling chamber of the freezing unit should be well known.

2. Basic investigation on a standard freezing unit (Freezer)

2.1 Investigation of the temperature distribution in the cooling chamber

Cell type dependent cryoprotocols take care for an optimal cooling rate to avoid cell damaging during freezing. In case of controlled rate freezing the vials are placed in a rack inside the cooling chamber. If controlled seeding is necessary a special rack for autoseeding can be used. For both applications gaseous nitrogen cools the cryovials slowly. In case of seeding additional liquid nitrogen flows through the autoseeding rack at the distinct temperature. Figure 4 shows a standard freezing unit and an autoseeding rack for cryovials inside the cooling chamber.

Thermometers (Pt 100)

Gas inlet

Figure 4: ICE CUBE 15M from Sylab – overview – (left)
and autoseeding rack inside the cooling chamber (right)

The freezing unit is connected to a storage vessel for liquid nitrogen (LIN). The required size of that vessel depends on the flow rates of the freezing processes.

These freezers are often used for cryopreserving vials and bags, containing cell suspensions or blood. For that reason an arrangement of vials in an autoseeding rack was chosen in order to obtain the temperature distribution inside the cooling chamber under real conditions. Thirty cryovials of 2 ml volume have been placed in the rack, each filled with 1,5 ml of a cryoprotectant. Five characteristic positions were evaluated (see ¥ in figure 4) with thermocouples. One precision thermocouple was centred in each vial. An additional vial contained the sample thermometer (Pt 100: X) of the freezing unit. All temperatures were monitored during running a well established cryoprotocol without autoseeding. But this protocol initiated a small spontaneous ice formation. Here, the appearance of spontaneous ice nucleation was also used as a measure for the heat transfer from the chamber via the rack into the vials.

In order to eliminate errors of measurements the thermocouples were cyclic permuted. Furthermore all sensors were placed and fixed in one vial in all five positions for two runs.

2.2 Results

First, comparing the results from more than 20 measurements significant deviations could be found between the five positions. At the point of spontaneous ice formation temperature differences up to 10 K were measured (see insert in figure 5). Second, the sample thermometer of the freezer didn’t represent the real freezing process within the cell suspension (see figure 5).

Looking at cell suspensions, the cells are round and in an equilibrium state. Suspended cells freeze well and without visible deviation between different locations in the rack using cell specific cryoprotocols. In conclusion, the cell suspensions seem to tolerate the observed temperature differences.

Looking at tissues, the cells are adherent and spread and got in contact. When cryopreserving several tissue pieces larger than 1 cm² at the same time, the inhomogenous temperature distribution inside the cooling chamber can course undefined freezing and disturbtion of the tissue. This we have measured in the temperature supervision at the different positions as well as in an unfavourable effect on the tissue and cell integrity. This situation is not acceptable for cryopreservation of tissues and prohibit the transfer into tissue banks.

Left bottom

Right bottom

Middle

Left up

Right up

Pt 100 sample thermometer and

Pt 100 chamber thermometer of the freezer, equal to the programme

Figure 5: Representative temperature run at the place of the sample Pt 100 of the freezer
Insert: Temperatures of the thermocouples at the five measuring positions

3. Consequences for special equipment

Petri dishes and cryovials are not suited to store tissue during the cryopreservation process. The tissue shouldn’t be wrapped and turned. That’s why special tissue cryogenic vials with screw closure (can) similar to a small Petri dish were developed. The material for the tissue vials must tolerate low temperatures, be thin enough for a good heat transfer and be tight during the whole preservation process. In order to ensure a sufficient heat exchange between tissue and the surrounding two additional hardware components were designed. First, the device called “Hedgehog” takes care for an optimal and reproducible heat transfer through the tissue. Inside the tissue cryogenic vials some pins are placed as means for fixing tissue and for conducting cold into that tissue. This gives the opportunity to cryopreserve large-area tissue of more than 1 cm². Furthermore two kinds of cans were developed. The metallic one is comparable to a Petri dish with Æ = 35 mm and ensures a good thermal contact to the surrounding area inside the freezer but it’s only for laboratory use. The plastic one also fulfils the thermal requirements for cryopreservation and will be qualified for registration as a medical device in next future.

Second, the device called “Rack” ensures the heat exchange between cooling chamber or thawing unit, can and tissue. It allows efficient, defined and reproducible heat transfer to the special vials during cooling and rewarming. It guarantees a rapid thawing. Beside this it organises that some tissue pieces can be handled and stored in more than one can under equal conditions.

CONCLUSIONS

All developed parts for the cryopreservation procedure were tested also with tissue engineered constructs. Though tissue engineered mucosa of the mouth was cryopreserved and after thawing transplanted in an animal model [4] with good success.

The new equipment overcomes the disadvantages if inhomogeneous temperature distribution in freezer and allows in the same way an more precise rewarming of tissue with high temperature rates. In summary cryopreservation of the engineered mucosa becomes more successful with better results in biochemical tests and higher viability of the cells. Both special equipment - can with “hedgehog” and rack - are prepared for patent registration.

REFERENCES

1 Lehle, K., Hoenicka, M. et al., Cryopreservation of human endothelial cells for vascular tissue engineering, Cryobiology (2005) 50 154-161

2 Sputtek, A., Mingers B., Langzeitkonservierung von menschlichen roten Blutkörperchen, KI (1994) 30/9 441-443

3 Applegate, D.R., Liu, K., Mansbridge J., Practical considerations for large-scale cryopreservationof a tissue engineered human dermal replacement, Adv. In Heat and Mass Transfer in Biology (1999) HTD-Vol. 363/ BED-Vol. 44 77-91

4 Spörl, G., Eckelt, U., Lauer, G. and Klingner, E., Cryopreservation of Tissue Engineered Mucosa, Proceedings of ICMC´06 + 9th Cryogenics (2006) 359-363

CR08-55

THE LIQUID AIR CRYOCHAMBERS FOR WHOLE-BODY CRYOTHERAPY

Strnad P.¹ , Forýtková L.², Brojek W.³,

¹DN FORMED Brno s.r.o., Czech Republic
²Masaryk University - Faculty of Medicine, Department of Biophysics,Brno, Czech Republic
³METRUM CryoFlex , Sp. z o.o., Blizne Laszczyňskiego, Poland

ABSTRACT

The paper describes the liquid air cryochambers used for whole-body cryotherapy, namely the cryochambers exploiting the principle of the retention of cold. The cryochambers have already been installed in the Czech Republic and further studies leading to the fructification of the efficiency of this method with regard to diagnosing the patients have been started.

INTRODUCTION

The whole-body cryotherapy is one of the physical therapy methods of the application which is based on body reaction to the stimulation of the temperature lower than – 100 °C.

At present, the whole-body cryotherapy ranks the modern methods used in the rehabilitation and in physical medicine.

The introduction in 1978 of the whole-body cryotherapy to medical prophylaxis is attributed to Professor Toshiro Yamauchi and his team (liquid nitrogen cryochamber). [1],[3].

Having been inspired by the positive treatment results of Japanese scientists, professor Reinhardt Fricke and his team from Germany transferred the method in all its applied forms to Europe [2],[3].

Significant role in this field is played by the Polish specialists. The Polish cryotherapy originated in 1983 (Professor Zdzisław Zagrobelny). In the year 1989, the first cryochamber (liquid nitrogen cryochamber) was constructed in Poland (the second in Europe, the third in the world) [3].

In 2002 Wieslaw Brojek and Wlodzimierz Szmurlo first on the world used the liquid air in whole- body cryotherapy and built the first cryochamber with retention of cold on the world [4].

In 2006, the first cryochamber with retention of cold using the liquid air in Czech Republic was built and this cryochamber was the first cryochamber of this model which has been built outside of Poland. Evidently, the development of the cryochambers was influenced by the development of scientific knowledge and technology.

CRYOCHAMBER WITH RETENTION OF THE COLD

The cryochamber can be built in numerous designs. The selection of the appropriate design depends on local conditions and room available. Typical design of the cryochamber is presented in the Figure 1.

The operation principle consists in the application of the cold laying low in a trough and in using the cold to keep low temperature inside the treatment chamber. Liquid air is used in order not to apply heat exchangers as applied in traditional chambers with liquid nitrogen. The elimination of losses on heat exchangers allowed to increase the cryogenic liquid capacity by 50 %. The low temperature results from spraying the cryogenic liquid from
a system of jets inside of the chamber.

Figure 1. The cryochamber with retention of cold – cut-away view

Textové pole: Figure 2. Stairs into the cryochamber with retention of cold The operation principle consists in the application of the cold laying low in a trough and in using the cold to keep low temperature inside the treatment chamber. Liquid air is used in order not to apply heat exchangers as applied in traditional chambers with liquid nitrogen. The elimination of losses on heat exchangers allowed to increase the cryogenic liquid capacity by 50 %. The low temperature results from spraying the cryogenic liquid from
a system of jets inside of the chamber. Actuated valves are being closed one by one in
sequence as the temperature approaches the temperature pre-set at the controller, and the liquid air is being decreased till the main valve is closed. The main valve is the last one to cut off the liquid air flow into the cryochamber. Thermocouples inside of the chamber measure temperatures on different heights. The controller calculates the average thermocouple temperature indication, compares it against the temperature pre-set for the treatment carried out at the moment, and controls the operation of actuated valves. Via the controller, the software provides the correct temperature inside the chamber. Besides, the software monitors the oxygen concentration inside the chamber.

An acoustic alarm is generated as soon as the oxygen concentration reaches the hazard level. For patients’ safety, the chamber is equipped with two independent oxygen concentration meters.

The chamber for cryogenic treatment can be built in versions for two, three or four persons. The chamber is thermally insulated with glass foam panels of approx. 15 cm thickness fastened to a load bearing structure of wood. The inner surfaces of the chamber are finished with soft wood. Inside the chamber, between wooden louvers and chamber walls is a nozzle system for spraying the cryogenic liquid. The chamber is equipped with double leaf transparent door of organic glass. The chamber is also equipped with
a transparent roof (Fig. 2) to enable the operator (physiotherapeutist) watch the patients.

The electrical installation inside of the cryochamber is supplied from a safe voltage source (12 V AC, 24 V AC).

Figure 4. The cryochamber with retention of cold – rear view

The cryochamber is connected via a cryogenic pipeline with a tank which is the cryogenic liquid source. As for dimensions of the tank, the tanks for 6 or 11 tons of liquid air have been used usually. The cryogenic liquid is synthetic liquid air (mixed of pure nitrogen and oxygen). Liquid air contains 21% ± 2% oxygen. Each delivery is provided with a manufacturer certificate of the liquid air composition. The liquid air manufacturer is responsible for the composition, purity and for the procedure of the tank filling with the cryogenic liquid.

At present, liquid air is manufactured in Poland and it is delivered by two companies to the cryochambers in Czech Republic.

The cryogenic pipeline is either vacuum-insulated, or it is equipped with a traditional thermal insulation. The consumption of the liquid air is from 90kg/per hour to 120 kg per hour in dependence on the size of the cryochamber. A microprocessor control computer with appropriate software is part of the cryochamber system (Fig. 4). It provides the cryochamber control during treatments. Besides, the cryochamber is outfitted with an ozone generator and a blower; and it is also connected to the controller. The cryochamber control functions, drying procedures and sterilisation procedures have been divided into two independent circuits within the controller. An UPS is a component of the equipment, to enable the operation (treatment) to be safely finished in case of voltage failure.

DISCUSSION

An original solution, using a phenomenon of cold retention and direct injection of liquid air into the cryochamber, notified in the Europe Patent Office, provides an opportunity for widening the range of interventions offered so far in well-known cryo-chambers solutions.

A novelty in the construction of cryotherapeutic cabin is the location in the hollow below the level of the operative floor, which allows the use of the cold retention effect and advantage of direct liquid air injection into the cryochamber. The slope of the stairs is mild and the steps are wide which enables the patients who are less fit to enter the cabin without problems and can be used as an adaptation area (Fig. 2).

The staff is able to observe the patient permanently thanks to the fact that the cabin is well lit and that the ceiling is transparent. This makes the cryochamber very comfortable especially for patients who tend to claustrophobia.

APPLICATIONS OF THE WHOLE-BODY CRYOTHERAPY

Generally – in all models or types of cryochambers:

The cryochamber is applied for therapy of the following diseases:

· Acute and chronic arthropathy and diseases of articular cartilage, e.g. rheumatoid arthritis, Bekhterov’s deformans-ankylosing spondylitis, Reiter’s syndrome, lupus erythematosus, gout, collagen disease, soft tissue rheumatism, infectional arthritis, chronic arthropathy as e.g. degenerative arthropathy, secondary inflammation in degenerative arthropathy, periarticular inflammation of shoulder,

· Back pain, eg. pains in the lumbar region (after eliminating of kidney disorders),

· sciatic neuralgia, condition after the operation of nucleus pulposus,

· Soft tissues rheumatism as e.g. tendinitis, or peritendinous inflammation;

· Consequences of accidents and other injuries, f.e. sprain, dislocation, muscle strain, burn, oedema after injuries, Sudeck’s atrophy disease- post-traumatic osteoporosis;

· Surgical diseases, e.g. contracture in joints, oedema after the surgery of bresast, hand, jaw, pain in the scars, bone abscesses, fistula, local infections, haemorrhoids, pruritus of the anus, acute superficial inflammation of the veins, epistaxis, acute inflammatory conditions within abdominal cavity,

· Neurological diseases like spastic hemiparesis and spastic paraparesis, disseminated inflammation of brain and cord, myasthenia, Parkinson’s disease, different pains, inflammation of anterior horns of the spinal cord, acute neuritis;

It should be emphasised that cryotherapy can be used in case of paroxysmal tachycardia, diabetes varices of lower limbs, systemic lupus erythematosus, systemic scleroderma.

Because of strong stimulation of immunological and endocrine systems, cryotherapy of the whole body is more and more often applied in:

· Biological renewal, and in building of immunological immunity in patients, who often suffer from infectious diseases (angina, flu, catarrh, inflammation of respiratory tract) and it may be a perfect noninvasive alternative to different kinds of vaccines;

· In case of intensive exercise, e.g. sportsmen both before and after the competition;

· In case of reducing obesity and cellulite;

· Treatment of mood depression of patients with depression;

CONCLUSIONS

At present, four cryochambers with retention of cold, using liquid air, were installed in the Czech Republic. The interest in building the new ones is even increasing for to be used not only for rehabilitation centers but for medical applications and research too.

REFERENCES

1. Yamauchi, T.,Nogami, S.,Miura, K.: Various aplication of extreme cryotherapy and strenous excercice program – focusing on chronic reumatoid arthritis. Physiotherapy and rehabilitation 5. 1981.

2. Fricke R.: Ganzkörperkältherap in einer Kälter-Kammer mit Temperaturen – 110° Celsia. Z., Phys.Med.Bahns.Klin., 18 (1989) 43.

3. Zagrobelny Z. a kol.: Krioterapia miejscowa i ogólnoustrojowa, Wydawnictwo Medyczne Urban & Partner, Wroclaw 2003.

4. CryoFlex –Poland Sp. Z o.o.: The Usage Description of a Cryogenic Cabin ( Chamber for Cryotherapy -110 °C to -150 °C).

5. Brojek, W: Kryoterapia – uvagi ogólne. Balneologia Polska, tom XLVIII, nr.1, 2006

6. Strnad,P.,Forýtková,L.: Terapeutické aplikace nízkých teplot.In:Sborník konference

XXVIII. Dny lékařské biofyziky,Valtice,květen 2005.

7. Strnad,P.,Forýtková,L,: The Whole-Body Cryotherapy. In: Bratislava Medical

Journal,2006, 107(4), XXIXth Days of Medical Biophysics,Bratislava, May 2006.

8. Steinerová,A.,Korotvička,M.,Racek,J.,Zeman,V.,Strnad,P.,Bajgar,M.: Posouzení vlivu celotělové kryoterapie na lidský organismus.In:Sborník XIV.sjezdu Společnosti rehabilitační a fyzikální medicíny, Luhačovice,duben 2007

CR08-18

Exergetic analyses usable for control the operation parameters of the helium processing plant for different working conditions.

Gherghinescu S.

National Institute of R&D for Cryogenics and Isotopes Technologies (ICIT), Rm. Valcea, code 240050 Uzinei 4, CP10, Valcea, Romania

Abstract

The present work proposes several analyses and control models of a helium liquefaction plant.

The monitored functional parameters of a cryogenic plant are temperature and pressure; based on these two, enthalpy and entropy can be then easily derived. These will provide the feedback for the control loop which will control entire plant.

The LABWIEW software is used for data acquisition and has great stability and safety data acquisition.

Based on the experimental data, we can calculate the enthalpy and entropy for obtaining the variation of exergy in different working conditions.

The main objective of this paper is to develop a method to optimize the cryogenic cycle. The optimization would determine the working regimes (temperature, pressure, flow, power consumption, etc.).

The contribution of the present work consists in the fact that the soft is easy to use, provides real time response and convenient accessibility. This can be done by calculating the enthalpy and entropy by a simplified method based on two-variable function.

Key-Words: exergetic analyze, monitored functional parameters.

Theoretical model for the calculation of the non-reversible losses of expansion and compression

Fig. 1 Non-reversible process of expansion and compression in T-S coordinates

Model of calculation the non-reversible losses trough heat transfer

Fig.2 Variation of exergy in heat transfer process

, - exergy of the heat exchanged between the two fluids

- flow rate of cool and heat fluid

Model of calculation the non-reversible losses trough throttling (J-T effect)

The J-T effect is the main source of irreversible losses. In our installation we hawe only one J-T type proces (the transformation 10-11).

Fig. 3 Thermodynamic transformations of cryogenic cycle in T-S coordinates.

C

D

E

F

A

Fig. 4 The cryogenic cycle of helium.

Considering that the throttle process takes place between certain pressure limits we may write a polynomial equation which will help us to determine the temperature at the end of the throttle process.

Fig. 5. Fases separation in dewar of helium.

Knowing the input temperature in the Dewar vase with liquid helium we can determine the liquid percentage (y):

y (%) at P₁₁=1.3bara

LABWIEW PANEL FOR DATA ACQUISITION

Control panel for data acquisition and system management

The above presented control panel was realised in LABWIEW software and has two main functions:

- data acquisition from the system

- work process simulation on respect of the above equations

Fig. 7 Simulation of working process in Labwiew software based on experimental data.

CONCLUSIONS

The exergetic analyses allow a global evaluation of cold losses with impact on cryogenic installation performances. It also ensures a stationary regime by a on-line control of the working parameters using the LABWIEW software.

The mathematical model used to compute different parameters gives us results in good agreement with the experimental data. The exergetics analyses may be extended over complex cryogenic systems, where a precise parameters control is required to reaches the desired working parameters. This is necessary for all systems where very short setting-up time is required.

Bibliography

Adrian Bejan, George Tsatsaronis, Michael Moran „Thermal Design & Optimization” ISBN 0471584673 Wiley 1996.
Adrian Bejan „Entropy Generation Minimization” Wiley 1996.
Philip Thomas “Simulation of Industrial Processes for Control Engineers” Butterworth-Heinemann 1999.

CR08-09

Split Pulse Tube Cryocooler with Innovative Double-Piston Linear Compressor

Kaiser G., Albert S., Schmidt J., Heidrich R., Binneberg A., Klier J.

Institut für Luft- und Kältetechnik gemeinnützige GmbH, Dresden, Germany

Abstract

A split pulse tube cryocooler with double-piston linear compressor is developed. The compressor is driven via an innovative electro-dynamic linear motor operating by use of a moving coil and core principle. This new development offers several advantages compared to moving coil or moving magnet systems. It is possible to generate higher forces, power and efficiency like in the case of moving coil systems. The reluctance force of the magnetic system can be used as a substitute for mechanical springs. In this way the average position of the mover is assured as for moving magnet systems. After an introduction into the new driving principle, we present the theoretical and experimental results.

Introduction

There exist a market request for highly reliable medium power range cryocoolers for the application in the field of cryo-conservation (cryogenic storage and carrier vessels), cooling of high-T_c superconductor magnetic bearings for centrifugal machines and flywheel energy storages and also cryogenic test chambers. To comply with this request a project for the development of a 10 – 20 W @ 80 K split pulse tube cryocooler was elaborated.

During the project, the development of an innovative electrodynamic linear motor drive was emphasized. The idea behind was to combine the advantages of the moving-coil and the moving-magnet system and to eliminate their disadvantages. By use of moving-coil motors it is possible to generate large electrodynamic Lorentz-forces. The disadvantage is that there are additional means like mechanical springs, required in order to set the average position of the linear motor mover. Using a moving-magnet system this problem is possible to overcome. The reluctance force can be used to set the average mover position. The forces of a moving-magnet system are limited by the mass of the moving magnet.

An innovative moving-coil and -core system was therefore developed and tested in a double-piston linear motor driven inertance pulse tube cryocooler. The details of the drive and a simulation model of the cryocooler are described. The experimental results are presented and discussed.

1. The Linear Motor

1.1 Principle

The principle design of the moving-coil and -core (MC&C) motor is shown in Figure 1. The stator consists of a axially magnetized permanent magnet ring with two pole rings at both sides. The moving part consists of the central magnetic core and the coil for the drive.

Figure 1: Principle design of the MC&C motor

Without driving current, the mover is in its average position. Each shifting in axial direction leads to a force generated by the magnetic field, which tends to minimize the field energy of the magnetic system (reluctance force).

If a current flows through the magnetic coil, the windings are directed in such a way that a Lorentz-force is generated against the reluctance force of the magnetic system leading to axial movement. The combination of the mover mass and the reluctance spring lead to a resonant system. Inside the assembled cryocooler an additional pneumatic spring occurs leading to an increase of the resonant frequency.

1.2 Motor model and calculation

In order to calculate the motor performance a simulation model (EXCEL sheet) was developed. Two different versions of the motor were calculated in order to operate with an AC power supply of 12 V or 24 V respectively. The input data are presented in the following table:

Parameter	Value
Mover stroke	9 mm
Electric load	250 W

Figure 2: Motor input data

A NdFeB permanent magnet was given with the following parameters:

Parameter	Value
Remanence	1.20 T
Coercitive field strength	905 kA m^-1

Figure 3: Permanent magnet data

The calculation gives the following parameters for the motor:

Parameter	Value
Magnetic induction	438 mT
Number of windings, wire diameter	136, 1.4 mm (12 V) / 275, 1.0 mm (24 V)
DC resistance	0.206 Ohm (12 V) / 0.818 Ohm (24 V)
Force coefficient	8.23 N A^-1 (12 V) / 16.7 N A^-1 (24 V)
Mechanical power	166 W
Coefficient of performance	66.4 % (12 V) / 67.1 % (24 V)

Figure 4: Motor data

1.3 Motor measurements

Figure 5 shows the assembled linear drive and the motor mover. In order to get information about the actual performance of the manufactured motors the magnetic induction was measured locally inside the gap by use of a Model 8202 Honeywell Hall magnetometer. The results for the two motors are given in Figure 6.

Figure 5: Linear motor and MC&C mover

The magnetic induction inside the main part of the motor gap is just 10 % less than calculated. Considering the scattering field, which was not included in the calculation, this deviation represents a good result. Load operation at 500 W of both motors inside the cryocooler have shown a voltage range between 26.4 V and 28.0 V and currents between 22.8 A and 23.1 A in the frequency range between 45 Hz and 55 Hz. These current and voltage observations correspond with the observation of a 10 % less force coefficient.

Figure 6: Magnetic induction inside the motor gap for two different motors

2. The Cryocooler

2.1 Cryocooler model and calculations

The cryocooler was modelled by use of SAGE (Gedeon Assoc.). The calculation was performed under the following input conditions:

Parameter	Value
Swept volume	12.5 cm³ per cylinder / 2 cylinders
Average pressure	25 bar
Motor parameters	12 V (24 V) / 50 Hz, 500 W el. load

Figure 7: SAGE model input parameters

Under these given conditions one should obtain the following experimental results:

Parameter	Value
Cooling power @ 80 K	21 W
Minimum temperature	60 K

Figure 8: SAGE model results of calculation

2.2 Cryocooler testing and results

Figure 9 shows the MC&C pulse tube cryocooler integration in the test rig. The test setup consists of a frequency- and voltage-variable power supply for the linear motor operation, a vacuum system for the thermal insulation, a measuring system for temperature and pressure wave recordings during operation. A cooling water supply was used to remove heat from the motor housings and from the hot end heat exchanger and precooler.

Figure 9: MC&C pulse tube cryocooler test setup

Figure 10 shows the pressure variation, ΔP, inside the compressor at an electric load of 500W. The peek-to-peek value of the pressure variation inside the compressor is 4.1 bar. The simulation shows a pressure variation of 4.5 bar. This is in good agreement with the experimental results, taking into account additional void volume which is not considered in the simulation model.

Figure 10: Pressure variation at 500 W electric load for different average pressures and frequencies

First cool down measurements, at ΔP = 2.2 bar, have shown a minimum temperature of 138 K (see Figure 11). Next cool down runs with the higher pressure variation of 4.1 bar should lead to much lower temperatures. Further cold head optimization like the integration of flow-straightening means and the optimization of the inertance tube are currently under progress.

Figure 11: Average pressure dependence of the minimum temperature for different inertance tubes

conclusionS

A new type of moving coil and -core electrodynamic linear motor was developed. This motor is used as a drive of a double piston linear motor compressor for an intertance pulse tube cryocooler, designed and tested for the medium cooling power range. The experimental results for the magnetic and electric motor measurements show good agreement with the calculations. A deviation of 10 % less force coefficient was observed. The calculation for the inertance pulse tube cryocooler shows the possibility to generate 21 W cooling power at 80 K for an electric load of 500 W at the linear motors. The pressure variation inside the compressor cylinder is expected to be 4.5 bar.

Experimental observations during the test of the cryocooler show a maximum pressure variation of 4.1 bar which is in good agreement with the simulation results. During cool down experiments, at lower pressure variations, a minimum temperature of 138 K was reached. For higher pressure variations a lower temperature is expected. Improvement of the cooling performance data by use of flow-straighteners and the optimization of the phase shifter is currently under progress.

Acknowledgement

This work was financially supported by the German BMWI under contract no. IW041392.

CR08-01

Very-low temperature thermal conductivity of structural materials for large cryogenic experiments

Ventura G., Barucci M., Martelli V., Risegari L.

Department of Physics, University of Florence, Italy

Abstract

In large cryogenic experiments like CUORE (Cryogenic Underground Observatory for Rare Events) a mass of several tons is sustained by rods which must possess outstanding mechanical parameters together with a very low thermal conductivity. We have measured the thermal conductivity of some structural materials below 1K. A comparison is done among data for metallic alloys, polymers and reinforced materials.

Introduction

CUORE (Cryogenic Underground Observatory for Rare Events) [1] experiment consists of a large array of detectors for the search of ββ − (0ν) decay, to be installed in 2010 at the underground National Laboratory of Gran Sasso (LNGS). A packed array of 988 TeO₂ detectors will be cooled down to ~10 mK. The experiment is housed in a large cryogen-free cryostat cooled by five pulse tubes and one high-power specially designed dilution refrigerator [2].

The cryostat is ~3 m high and has a diameter of ~1.6 m. About 5·10³ kg of lead shielding are to be cooled to below 1 K and a mass of 1.5·10³ kg must be cooled to 10 mK. Several tie-rods sustain the different parts of the experiment. One end of each rod is at low temperature (10 mK for the detector frame, 50 mK for the coldest radiation shield and lead shield, 700 mK for the shield linked to the still), the other end being, in some cases, at room temperature. A thermalisation of the rods at the temperature of the first and the second stage of the pulse tubes will be realized. Hence also the value of the thermal conductivity of the material up to room temperature is important.

At the lowest temperatures, the thermal conductivity has great influence in establishing the thermal load on the dilution refrigerator. The thermal conductivity of the structural materials candidates for such tie-rods is usually known down to 4 K.

Here we present data of thermal conductivity below 1 K of PERMAGLAS ME771 [3], a new oriented glass epoxy laminate material. A comparison is also done with other materials, such as Torlon, Kevlar and metallic alloys, candidates for the realization of the CUORE tie-rods.

1. Experimental Set Up and Measurements

The thermal conductivity of PERMAGLAS ME771 was measured along the direction of the reinforcing fibres by the longitudinal steady heat flow method.

The experimental set up is shown in Fig. 1. The cylindrical sample has a length L=(89.6±0.2)mm and a sectional area A=(50±1)mm². At room temperature, the form factor is g=A/L=(0.561±0.016)mm.

Figure 1: Set up of the experiment (shorter sample)

The thermal contacts at the ends of the sample have been realized by means of two copper blocks and two copper screws 4 mm in diameter. Since the thermal contraction of PERMAGLAS ME771 is lower than that of copper [4,5], the thermal contact between the blocks and the two ends of the sample becomes better on cooling. A SMD (Surface Mount Device) NiCr heater and a RuO₂ thermometer were glued onto the two copper blocks at the ends of the sample.

The electrical connections to the heater and to the thermometer were made with NbTi wires (Ø= 25 μm). The NbTi wires were electrically connected by tiny crimped CuNi tubes. At the ends of the NbTi wires a four lead connection was adopted. The bottom copper block was screwed onto a copper support in thermal contact with the mixing chamber of a dilution refrigerator. Another RuO₂ calibrated thermometer was used for the measurement of the mixing chamber temperature. Both thermometers were calibrated by means of a SRD 1000 (Superconductive Reference Device) [6,7] and a NBS-SRM 767a fixed point device [8]. A copper shield, in thermal contact with the mixing chamber of the dilution refrigerator, surrounded the experiment.

A known power P was supplied to the upper end of the sample to establish a difference of temperature T₁ − T₀ between the ends of the sample. A LR-700 a.c. resistance bridge for the upper thermometer (measuring T₁) and a Picowatt for the lower thermometer (measuring T₀) were used. A four wire I−V source delivered the power P to the heater.

By derivation of the integrated power P(T₁) at constant T₀:

(1)

the thermal conductivity k(T) can be obtained.

A second set of measurements was carried out on a sample of g=(1.10±0.02)mm to cover the lowest temperature range. The second run on the shorter sample gave the same value of k in overlapping temperature range. This result guarantees that the effect of the contact thermal resistances do not influence the value of thermal conductivity.

2. Results and conclusions

The measured thermal conductivity of PERMAGLAS ME771 in the 100mK-1K temperature range is shown in Fig. 2. Data of k(T) can be fitted by the formula:

k(T) = a·T ⁿ= (9,00±0.01)·10^-5T ^{(1.880±0.002)} (2)

Figure 2: Thermal conductivity of PERMAGLAS ME771 in the 100 mK- 1 K temperature range

The maximum estimated error on k(T) is 4%.

Figure 3: Thermal conductivity of PERMAGLAS ME771 compared with other candidate materials for the CUORE supports

In Fig. 3 the thermal conductivity of PERMAGLAS ME771 is compared with that of other candidate materials for the CUORE supports [9-12]. From a purely thermal point of view, the PERMAGLAS ME771 is a good candidate for the realization of the supports. Of course mechanical data will play a decisive role in the choice of the material.

REFERENCES

1. Ardito, R., et al., CUORE: a cryogenic underground observatory for rare events, http://arxiv.org/abs/hep-ex/0501010 (2005)

2. Nucciotti, A., et al., Design of the cryogen-free cryogenic system for the CUORE experiment, J. Low Temp. Phys. (2008) 151

3. http://www.permali.com/english/pdf_files/Permaglas.pdf

4. Roechling, private communication

5. Pobell, F., Matter and Methods at Low Temperatures, Springer-Verlag, Berlin (1995)

6. Bosch, W.A., et al., Status Report on the Development of a Superconductive Reference Device for Precision Thermometry below 1 K, in Proceedings of the 8th International Symposium on Temperature and Thermal Measurements in Industry and Science TEMPMEKO 2001, VDE Verlag, Berlin (2001) 397-401

7. Schottl, S., et al., Evaluation of SRD1000 Superconductive Reference Devices, J. Low Temp. Phys. (2005)138 941-946

8. Schooley, J. F., Soulen, R. J., Jr., Evans, G. A., Jr., Preparation and Use of Superconductive Fixed Point Devices, SRM 767, NBS Special Publication 260-44, Washington, DC, National Bureau of Standards (1972) 1-35

9. Risegari, L., Barucci, M., Lolli, L., Ventura, G., Low temperature thermal conductivity of Ti6Al4V alloy, J. Low Temp. Phys. (2008) 151

10. Barucci, M., Lolli, L., Risegari, L., Ventura, G., Measurement of thermal conductivity of the supports of CUORE cryostat, submitted to Cryogenics (2007)

11. Ventura, G., Bianchini, G., Gottardi, E., Peroni, I., Peruzzi, A., Thermal expansion and thermal conductivity of Torlon at low temperatures, Cryogenics (1999) 39 481-484

12. Ventura, G., Barucci, M., Gottardi, E., Peroni, I., Low temperature thermal conductivity of Kevlar, Cryogenics (2000) 40 489-491

CR08-14

design, Fabrication and test results
on a conduction cooled HTS magnet

Joonhan B., Seokho K., Kideok S., Myunghwan. S.

Korea Electrotechnology Research Institute, Changwon, Korea

Abstract

This paper describes design, fabrication, and testing of the conduction cooled HTS magnet. The magnet is composed of 22 double pancake coils. The magnet was conductively cooled down to 5.6K with two stage GM cryocoolers. The temperatures of the HTS magnet were measured in the charging and discharging process. The successful operation of the magnet illustrates that the technology of cooling HTS magnet with GM cryocoolers is fully established.

Introduction

The superconductor can carry larger current without electricity loss because of having no resistance. The superconducting magnets using theses merits are utilized in superconducting equipments such as MRI, NMR, SMES, magnetic separator, superconducting generator and motor. Because the conventional superconducting magnets are cooled with coolants such as liquid helium or nitrogen to keep their superconducting properties, there are many disadvantages in this magnet type. The cryostats used to store the liquid cryogens and keep the magnet cold are rather complex. When a quench occurs in the magnet, sudden evaporation of a large amount of liquid coolants might be dangerous. Also, the liquid cryogen to run the magnet is very costly and there are annoying to refill coolants periodically. Recently, rapid progress of refrigerators and superconducting wires allows the superconducting magnets to be operated without cryogen use. Since Hoenig demonstrated a thermal design of the conduction cooled superconducting magnet combined with Gifford McMahon (GM) cryocooler in 1983, many researches on the refrigerator-cooled magnets have actively been performed world widely [1]. The most outstanding value of cryogen free magnet lies in easy and safe handling, low running costs, and compact system. Also, the cost for the electricity and the maintenance of cryocoolers would be less than 1/10 of that for the cryogens [2].

This paper describes the design, fabrication and test results on the conduction cooled high temperature superconducting (HTS) magnet for superconducting applications. The optimal design of the magnet was conducted using the objective function of minimizing the total required amount of the HTS conductor. The electromagnetic and mechanical behaviors of the magnet are analyzed. The prototype magnet composed of 22 double pancake coils was fabricated on the basis of design and analysis results. Finally, the performance of the magnet was evaluated during charging and discharging operation.

1. Optimal design of the HTS magnet

Generally, there are several types of HTS magnets such as solenoid, multiple solenoid, toroid, and so on. Solenoid type coils are simple to design and easy to fabricate, but they can not prohibit or confine stray field. Multiple solenoid coils show very good characteristics on stray field, but it has very poor energy density. Toroid coil can be a compromise proposal. A perfect toroid coil makes no stray field. The field is confined inside coil, but it is very hard to realize such kind of coil. Instead, coils wound in pancake or stacked pancake coils can be configured to simulate similar effect. Meanwhile, toroid coils require more conductors than solenoid coils but fewer wires than multiple solenoid coils [3]. In this study, the modular single pole double pancake coil (DPC) was selected in consideration of its installation environment.

The 4-ply HTS conductor is used to design the HTS magnet. The conductor is composed of two AMSC Bi-2223 tapes and two brass tapes which are soldered at each side of Bi-2223 tapes for the mechanical reinforcement. To minimize the minimum bending radius, the thickness of the brass tape and solder were controlled precisely. Finally, it was wrapped with kapton tape for electrical insulation. Table 1 shows the specifications of the 4-ply HTS conductor. It is well known that the critical current of the HTS tape depends on the direction and amplitude of the external magnetic field, so we confined the maximum operating current of the magnet within 70% of the critical current in the 4-ply HTS conductor. The main objective of the design is to find optimal dimensions of the magnet with minimum conductor length. There are several constraints to be considered for the design process, such as critical magnetic field, total HTS conductor length, and geometrical constraints for supporting and cooling equipments. Among them, the total HTS conductor length was selected as the main objective function.

Table 1: Specifications of 4-ply HTS conductor

Composition	2 BSCCO-2223 tapes and 2 brass tapes wrapped with 2 kapton tapes
Average width of the conductor	4.5 mm
Average thickness of the conductor	0.77 mm
Critical tensile stress	150 MPa at room temperature
Critical bend diameter	150 mm at > 95% Ic retention
Critical current	480 A at 20 K, 3 T

Figure 1: The across section of the HTS magnet

Table 2: Design Results of the HTS magnet

Operating temperature [K]	20
Inner diameter [mm]	500
Outer diameter [mm]	691
Number of turns per DPC	262
Number of DPCs	22
Gap between DPCs [mm]	4
Operating current [A]	275
Height [mm]	330
Maximum parallel field [T]	3.92
Maximum perpendicular field [T]	2.49
Central field [T]	2.98
Stored energy [kJ]	605
Inductance [H]	16
Length of HTS conductor [km]	10.8

The across section of the HTS magnet is shown in Figure 1. The aluminum bobbin to support DPC is coated with ceramic powder for insulation and plays a role of the cooling plate to cool down the HTS magnet. FRP (fiberglass reinforced plastics) spacer of 2 mm thickness is located in the middle of DPC for the winding arrangement. The design operating temperature of the HTS magnet is 20K, which is attained by conduction cooling with two GM cryocoolers. Table 2 shows the design results of the HTS magnet achieved by Auto-Tuning Niching Genetic Algorithm [4].

2. Electromagnetic and mechAnical analysis

HTS magnets experience the electromagnetic force, which cause the instability of the superconducting magnet and deformation of the conductors. The problem is more serious in large scale HTS magnet. Therefore, it is important to consider the mechanical forces on conductors of superconducting magnets caused by this electromagnetic force.

When electromagnetic forces and stress are balanced in a nonmagnetic material, the following equations are valid.

(1)

(2)

(3)

Where, J is current density, B is magnetic flux density and S is a stress tensor. When the operating current circulates in the solenoid HTS magnet, the magnetic flux density distributions in the magnet are obtained from expression (1) and (2). Also, Lorentz force interaction between the operating current and the magnetic field results in stress within the magnet, which tend to burst the windings radially outward and crush it axially and it could be computed by using equation (3).

Figure 2: The magnetic field distribution in the HTS magnet

Figure 3: The radial and hoop stress distributions in the DPC 11 of the magnet at current of 275 A

Figure 2 shows the magnetic field distribution in the HTS magnet. The strongest axial field in the middle of the magnet was produced and it caused the maximum hoop stress on the HTS conductor in outmost layer of the magnet. The evaluation of the maximum stress within the windings during charging the magnet is important in order to keep the magnet from the quench due to the movement of the HTS conductor. Assuming that the microscopic stress distribution due to the shielding current in the each HTS conductor can be neglected, the radial and hoop stress distributions in the solenoid HTS magnet are achieved through the axis-symmetric two dimensional numerical analysis using the finite element method. Poisson ratio of 0.35 and equivalent Young’s modulus were adopted to calculate the stress within the magnet because 4-ply HTS conductors are composite material [5]

The radial and hoop stress distributions in the DPC 11 of the magnet at the current of 275 A are shown in Figure 3. The maximum radial and hoop stress are produced in the DPC 11 because it is placed in the middle of the HTS magnet. The 3.61 MPa of the maximum radial stress was calculated at the center of the DPC 11. Also, 35.6 MPa of the maximum hoop stress was computed at the innermost radius of DPC 11.

3. fabrication and Assembly of the HTS magnet

3.1 Cooling system

The cryostat for the conduction cooled HTS magnet has the outer diameter of 1330 mm, the height of 924 mm, and the weight of 1100 kg. The two stage GM cryocoolers (Sumitomo, model 415D) were adjacently mounted on one side of the cryostat. The current leads and radiation shield are cooled by the first stage of the cryocooler and the HTS magnet is cooled down by the second stage of the cryocooler.

For improving cooling capacity, the HTS magnet was cooled down by using high purity cooper braid for a flexible thermal link and the thermal stress relaxation at the low temperature, which connected the magnet with the second stage of the refrigerator. 15 layers of MLI (multi-layer insulator) were wound on the HTS magnet to minimize radiation heat from the room temperature. A pair of brass current leads was anchored between the room temperature and the first stage of the cryocooler. Also, pairs of the HTS current leads composed of AMSC Cryoblock wires were inserted between brass current leads and the HTS magnet terminals for blocking conduction heat invasion through the brass current leads. In order to reduce the thermal contact resistance at the many mechanical boundary of the thermal path, each boundary surface was precisely controlled by mechanical machining and indium foil or thermal grease (Lakeshore, Apiezon N) was used at the boundary for providing good thermal contact.

3.2 HTS magnet

The 4-ply conductor was arranged on the aluminum alloy bobbin, which extract the heat generation from the magnet and was wound from the midpoint of the whole length of the conductor. The 4-ply conductor with length of 490 m was wound into a DPC of 262 turns. The HTS magnet is composed of 22 DPCs and they are cooled by two stage GM cryocoolers through conduction. The aluminum bobbins were anodized for the additional electrical insulation and the dielectric strength was measure as 2 kV at the flat surface and 500 V at the edge. To avoid the conductor displacement under electromagnetic force or detachment between the coil and the bobbin, a cryogenic resin was impregnated in vacuum after winding and they are hardened at room temperature for 24 hours. The configuration of the assembled HTS magnet with the top flange of the conduction cooled cryostat is shown in Figure4.

3.3 Monitoring the temperature and quench of the magnet

Monitoring the temperature and voltage of the conduction cooled HTS magnet is important issue to provide safe operation during the charge and discharge of the magnet. The total 24 Cernox temperature sensors and 21 E-type thermo-couples were installed along the thermal path to find out the poor thermal contact point. 22 DPC voltages, 22 joint voltages and current lead voltage were also measured through isolation amplifiers to detect the quench signal in the magnet in the energizing and deenergizing process.

Figure 4: The assembled HTS magnet with the top flange of the cryostat

4. test and discussion

4.1 Cool down

To reduce the initial cool down time, we circulated liquid nitrogen through the heat exchanger, which acted as thermal sink connected to the second stage of the cryocoolers. After the temperature of the heat exchanger reached around 170 K, supplying the liquid nitrogen was stopped and the HTS magnet was cooled down only using the two cryocoolers.

Figure 5 shows the cool down history for the major measuring points. It took 90 hours to cool down the HTS magnet to the saturation temperature. By minimizing the heat penetration through the support, the current leads and radiation, the final temperature was 5.6 K lower than the operation temperature in design. Before installing the cryocoolers, the cooling capacity of the cryocoolers was measured and the heat penetration could be estimated. According to the measured cryocooler temperature, the heat penetration by conduction and radiation to the HTS magnet is estimated to be about 6 W considering the two GM cryocoolers.

4.2 The temperature variations in the HTS magnet at different currnet charging rates

Figure 6 shows the temperature variations during charging the current to 275 A with ramp rate of 2 A/s and discharging with a 1 Ω dump resistor. The hot spot temperature at the HTS magnet was 12.9 K during discharging operation. Fig 7 shows the hot spot temperature in the HTS magnet when the current charging rate was varied from 0.2 to 2.5 A/s. In Figure 7, it was observed that the maximum temperature of the magnet rose as the current charging rate increased. This can be thought that the removed heat loss by cryocooler decreases as the ramp rate increases because the cooling time of cryocooler is longer than charging time of the magnet

Figure 5: The cool down history for the HTS magnet

Figure 6: The temperature variations during charging and discharging the HTS magnet

Figure 7: Hot spot temperatures in the HTS magnet at different current charging rates

conclusionS

The optimal design of the HTS magnet for superconducting applications was carried out with several constrain conditions and the prototype HTS magnet was fabricated on the basis of design results. The thermal and electromagnetic behaviors of the magnet were performed during charging and discharging operation. The results are as follows.

a) Maximum perpendicular and parallel field density in the designed HTS magnet are 2.49 T and 3.92 T, respectively.

b) On the basis of the field analysis results, The 3.61 MPa of the maximum radial stress was calculated at the center of the magnet and the maximum hoop stress was 35.6 MPa at the innermost radius of the magnet.

c) The temperature of the HTS magnet was cooled down to 5.6 K after 90 hours using two GM cryocoolers and the total heat penetration by conduction and radiation to the HTS magnet is calculated to be about 6 W using the cooling load map and the temperature of the cryocoolers

d) The hot spot temperature at the HTS magnet was 12.9 K in discharging with a 1 Ω dump resistor after charging current to 275A with ramp rate of 2A/s.

e) The maximum temperature of the magnet rise as the current charging rate increase because the cooling time of cryocooler is longer than charging time of the magnet.

The results through this research will be utilized in the optimal design and stability evaluation of the conduction cooled HTS magnet for superconducting applications.

Acknowledgement

This work was supported by Electric Power Industry Technology Evaluation & Planning, Republic of Korea.

REFERENCES

1. Heonig, M. O., Design concepts for a mechanically refrigerated 13K superconducting magnet system, IEEE trans. on Magn. (1983) MAG-19 3 880-883

2. Katano, S., Minakawwa, N., Hasebe, T. and Sakuraba, J., New cryocooler-cooled superconducting magnet: A 13.5T high-field split-pair coil magnet for neutron scattering, Physica B (2006) 385-386 1300-1302

3. Schonwetter, G., SMES solenoids with reduced stray field, IEEE trans. on Magn (1994) 30 2636-2639

4. Wooseok, K., Design of HTS magnets for a 600 kJ SMES, IEEE trans. Appl. Supercond. (1994) 16 2 620-623

5. Crandall, Dahl and Lardner, An introduction to the mechanics of solid-2^nd ed., McGraw-Hill Inc., New York, USA (1983)

CR08-23

Analysis on the quench at the conduction-cooled joint between HTS wire and normal conductor

Bae D.K.¹, Bae J.H.², Lee D.-Y.³, Lee S.-J.³, Park J.-S.³,

¹Department of Safety Engineering, Chungju National University, Chungju, Korea
²Korea Electrotechnology Research Institute, ChangWon, Korea
³Division of Energy & Electrical Engineering, Uiduk University, Kyongju, Korea

Abstract

The heat generated in the high-Tc superconducting (HTS) devices is related with the cost efficiency and safe factor of HTS devices. This paper deals with the quench at the conduction-cooled joint between the HTS wire and normal conductor. The 3-D numerical simulation of this phenomenon was implemented and compared with the experimental results. The experiment was implemented with the HTS wire mounted on the copper blocks cooled with a Gifford McMahon (GM) cryocooler.

Introduction

With the successful commercialization of the Bi-2223 powder-in-tube type HTS wire (1 G wire) and the YBCO coated conductor type HTS wire (2 G wire), the HTS magnet for high magnetic field and electric power applications has been researched & developed actively [1], [2]. To name a few, there are superconducting magnetic energy storage (SMES), magnetic resonance imaging (MRI), HTS fault current limiter, HTS motor, HTS cable and HTS transformer. Especially, the conduction-cooled HTS magnets are widely developed actively [3], [4]. The conduction-cooling system has many advantages such as easy operation, compact size, wide range of operating temperature and so on. The operating temperature range of recently developed conduction-cooled HTS magnets was from 4.2 K to 50 K [1]-[4]. But the thermal runaway occurred easily in many conduction-cooled HTS magnet systems. The thermal runaway could degrade or destroy the HTS coil. Although the possibility of the thermal runaway of the HTS magnet is small, because the most of the conduction-cooled HTS magnets are DC magnet and the operating current is usually much less than the critical current of the magnet, it should be noticed that AC loss, pulse current, larger operating current to reduce the cost, and electric power interruption may cause the thermal runway. The possibility of the thermal runaway should be considered when designing the conduction-cooled HTS magnets.

This paper deals with the quench at the conduction-cooled joint between the HTS wire and normal conductors, as an initial step in the analysis on the cause of the thermal runaway phenomenon in the conduction-cooled HTS magnet. The normal conducting copper block is usually used as the current lead in the HTS magnet system. The copper current lead may be heated by the AC loss of the HTS wire, pulse current and electric power interruption, at which point the thermal runaway phenomenon initiates. The minimum quench energy (MQE) of HTS wire is several orders higher than that of the LTS wire. However, because the normal zone propagation (NZP) velocity of the HTS wire is several orders slower than that of LTS wire and even very slow, the initiated thermal runaway causes damage to the HTS wire and finally could cause the burnout of the HTS wire. So it is important to evaluate the thermal runaway condition of the conduction-cooled HTS magnet.

In this study an YBCO coated conductor (CC) was mounted on the copper current leads. The mounted CC and copper leads were cooled down with a Gifford McMahon (GM) cryocooler. The temperature difference between the CC and the leads was made deliberately in the system so that the experiment of thermal runaway could be implemented.

1. Simulation of jointing part

1.1 Connecting methods between HTS wire and normal conducting lead

Figure 1 shows two connecting methods between HTS wire and normal conducting copper current lead. In method (a), both sides of HTS wire are in contact with the copper blocks. In method (b) one side of HTS wire is in contact with the copper block. The bottom side of the copper block is connected to the cold head of cryocooler. The distance between HTS wire and the bottom of the copper block of method (a) is closer than that of (b). So, method (a) is more efficient in cooling connecting parts than method (b) as a conduction-cooled connecting method.

1.2 Numerical simulation of jointing part

Because the numerical analysis of method (b) in Figure 1 can also simulate method (a) in Figure 1, method (b) was considered as the 3-D analysis model as shown in Figure 2. 3-D

(a)

(b)

Figure 1: Connecting methods between HTS wire and normal conducting lead, (a) HTS wire mounted between copper blocks, (b) HTS wire mounted on copper block

Finite element method (FEM) was used in the analysis. The boundary conditions of this analysis can be summarized as follows:

1) Perfect insulator boundary condition of whole outside surfaces except the bottom surface of the copper block.

2) Specified temperature boundary condition (connected to cryocooler) of the bottom surface of the copper block.

3) No thermal resistance between the HTS wire and the copper block.

The thermal conductivity of used copper was like Figure 3 [5]. The temperature of the cryocooler was set to 75 K. In this calculation, AC loss of HTS wire was not included.

Figure 2: 3-D analytic model with mesh

Figure 3: Thermal conductivity of copper [5]

Figure 4: Distribution of temperature in connecting part (75 K, 300 A-200 Hz)

Figure 5: Distribution of temperature in connecting part (75 K, 300 A-600 Hz)

Figure 4 shows the distribution of the temperature in the connecting part with the transport current of 300 A-200 Hz. The temperature of the bottom side was specified to 75 K. The temperature of the HTS wire was about 76.5 K, and then the difference between HTS wire and the bottom of copper block was about 1.5 K. Figure 5 shows the distribution of the temperature in the connecting part with 300 A-600 Hz. As the frequency of the current increases, the difference of the temperature became slightly bigger than that of 200 Hz. The temperature difference between the HTS wire and the bottom of the copper block was about 1.9 K. The results of this simulation mean that AC current and/or pulse current may cause the difference of the temperature between main HTS coil and normal jointing part, which will be the source of thermal runaway.

2. Experiment

2.1 Experimental setup

Figure 6 shows the experimental setup for the deliberate thermal runaway test. Three thermal sensors were mounted on the central part of HTS wire and two connecting parts, respectively. Three voltage taps were also mounted on the HTS wire. One was on the center of the wire (central voltage tap) and another was on the connecting part (Cu lead voltage tap) and the third was on the whole wire (whole wire voltage tap) Left and right copper blocks were connected to the cryocooler via thin glass fiber reinforced plastic (GFRP) plate to make deliberate temperature difference. The thickness of the GFRP was 1 mm. The copper blocks at the center of the system were connected directly to the cryocooler. The heater was installed on the cold head of the cryocooler to control the temperature.

The HTS wire used for this study was CC from AMSC Inc. The specifications of the used CC are shown in Table 1.

To make deliberate temperature difference between center point and connecting part, the whole system was cooled down to 10 K and warmed up to 80 K, and then re-cooled down the testing system to 75 K. Due to the GFRP plate between copper block and cold head, there was some delay in cooling the connecting parts. When the temperature of the center part reached 75 K, transport current began ramping-up.

Figure 6: Experimental setup for deliberate thermal runaway test

Average thickness	0.18-0.22 mm
Minimum width	4.27 mm
Maximum width	4.55 mm
Minimum double bend diameter (Room Temperature)	25 mm
Maximum rated tensile stress (RT)	250 MPa
Maximum rated wire tension (RT)	19.3 kg
Maximum rated tensile strain (77 K)	0.3 %
Minimum Ic (77 K, self-field 1 mV/cm)	80 A

Table 1: Specifications of used CC

3. Experimental results and discussions

3.1 Measurement of DC critical current

Figure 7 and 8 show the voltage and current characteristics of used CC at 75 K and 80 K respectively. The measured Ic at 75 K was 90 A and that at 80 K was 47 A. Four-probe method and 1 µV/cm criterion were used in the measurement. The whole voltage tap was used in these measurements. The Ic at 75 K was about two times larger than that at 80 K.

3.2 Experiment of deliberate thermal runaway

Figure 9 shows the temperature and transport current profile during the experiment. The delay of the temperature falling at the copper blocks was shown in Figure 9. When the transport current began ramping-up, the temperature at the connecting part was about 78.5 K and that at center was about 75 K. Block temperature was cooled down to 76.4 K at the ending of the experiment. The effective temperature difference between center and block was about 1.4 K.

Figure 7: Voltage and current characteristics of used CC at 75 K

Figure 8: Voltage and current characteristics of used CC at 80 K

Figure 9: Temperature and current profile

Figure 10 shows the experimental results of the deliberate thermal runaway. The length of center and Cu lead voltage tap was 4 cm and that of whole wire was 100 cm. Because the temperature of the Cu lead was of 1.4 K higher than that of center, the voltage on the Cu lead tap rose firstly. When the ramping-up transport current reached to 95 A, CC was burned out.

The purpose of the experiment was to verify the possibility of the burning out of the HTS conductor at the joint between HTS wire and normal conducting lead in the conduction-cooled HTS magnet system. High energy is concentrated into a small zone when quench occurs in the HTS wire cooled by cryocooler. The high energy concentrated into a small zone gives severe damage to HTS magnet. The temperature difference between HTS wire part and Cu lead part due to AC loss of HTS material, pulse current, power interruption and so on should be considered when design the conduction-cooled HTS magnets.

Figure 10: Deliberate thermal runaway

conclusionS

In this paper, temperature of connecting part between HTS wire and normal conducting copper lead was numerically calculated by using 3-D finite element method and the deliberate thermal runaway experiment using conduction-cooled CC was implemented. The temperature of normal conducting lead may be higher than that of HTS winding part so the critical current of the HTS conductor on the normal conductor decrease. The difference of temperature between HTS winding part and normal conducting part causes the degradation of HTS wire. The possible temperature difference between HTS wire part and Cu lead part should be considered in the design of conduction-cooled HTS magnets.

ACKNOLEDGEMENT

This work has been supported by KESRI (R-2005-7-068), which is funded by MOCIE (Ministry of commerce, industry and energy).

REFERENCES

1. Waynert, Joseph A., Boenig, Heinrich J., Mielke, Charles H., Willis, Jeffrey O., and Burley, Burt L. Restoration and testing of an HTS fault current controller, IEEE Trans. on Applied Superconductivity (2003) 13, 1984-1987

2. Kalsi, S. S., Aized, D., Connor, B., Snitchler, G., Campbell, J., Schwall, R. E., Stephanblome, Th., Tromm, A., and Kellers, J., HTS SMES Magnet Design and Test Results, IEEE Trans. on Applied Superconductivity (1997) 7, 971-976

3. Seong, K.C., Kim, H.J., Kim, S.H., Park, S.J., Woo, M.H., Hahn, S.H., Research of a 600 kJ HTS-SMES system, Physica C (2007), 463-465, 1240-1246

4. Obana, T., Tasaki, K., Kuriyama, T., Okamura, T., Thermal stability analysis of conduction-cooled HTS coil, Cryogenics (2003) 43, 603-606

5. Iwasa, Y. Case Studies in Superconducting Magnets, Plenum Press, New York, USA (1994) 385

CR08-37

ANALYSIS OF THE MAGNETIC PROPERTIES OF HTc SUPERCONDUCTORS AND APPLICATION THEM AS PERMANENT MAGNETS

Sosnowski J.

Electrotechnical Institute, 04-703 Warsaw, Pożaryskiego 28, Poland

Abstract

Among various present or near future applications of HTc superconductors one of more promising is utilizing these materials as permanent magnets. It is a very actual topic, especially because quite recently it has been revealed the giant magnetic trapped flux in these materials of the range of 17 T. This value is much higher than received for conventional permanent magnets, even based on rare earths, for which magnetic remanence does not exceed a few Teslas. In this paper the subject of the trapped flux will be considered taking into account the granular structure of the ceramic materials. The model describing the potential barrier height capturing the vortices, which determines the critical current density, magnetic induction distribution and trapped flux will be presented for HTc ceramic superconductors. The influence of ceramic material parameters on the trapped flux magnitude will be considered at an aim of indicating meaning these parameters, which should be useful for optimizing trapped flux value. Trapped flux has important meaning from the point of view of application of HTc superconductors in the magnetic levitation process, as superconducting bearings, which will be also briefly discussed in the paper.

Keywords: HTc superconductors, magnetic levitation, permanent magnets

1. Introduction

High temperature oxide superconductors have been discovered more than twenty years ago. Now it is time therefore for applications of these materials in electric devices. Exceptional electromagnetic phenomena appearing in these materials should be taken into account to this aim. In the paper mechanism of the trapped flux in HTc superconducting materials is considered, which giant value allows to treat these superconductors as promising permanent magnets, very useful in magnetic levitation process, for construction of magnetic bearings, future motors of new generation etc. The present work is partly stimulated by the recent experimental data, showing that in single macro-grain of YBaCuO trapped magnetic flux reaches giant value of 17 T at low temperature of 29 K (according to the results of Dowa Mining, Japan). It is just actual world record of the remanent magnetic induction among all materials presently known, including permanent magnets based on the rare earths. In the paper is performed theoretical analysis of these unique, from technical point of view magnetic properties of HTc ceramic superconductors, basing on modeling of the pinning interaction. The critical current and influence of it on trapped flux magnitude is considered.

2. outline OF THE PINNING INTERACTION IN HTc SUPERCONDUCTORS

Unique magnetic properties of HTc superconductors are essentially related to the pinning effects arising in these materials. Pinning phenomena describe the capturing of the magnetic vortices in the superconducting materials and determine therefore critical current density and trapped in vortices magnetic flux. Layered structure of the high temperature superconducting materials influences specific shape of the pancake type vortex in HTc superconductors, which is presented in Fig. 1. The core of that vortex retains the normal state, which is characterized by higher energy than superconducting one. As follows from theoretical analysis, especially based on the Ginzburg-Landau theory, the superconducting state is characterized by the order parameter, which denotes that this state is energetically more favourable than the normal one. It means further that the increase in the volume of the normal phase in system enhances its energy and therefore the amount of normal phase should be minimized [1]. This effect is considered just in the proposed pinning interaction model. The shift of the captured on the nano-sized pinning centre vortex possessing the normal core of the radius x, equal to the coherence length, enhances normal state energy of the superconductor. On the other hand the Lorentz forces, acting on the pinned vortices during current flow, as well as elasticity forces tear off the vortices from the pinning centres, which leads to their movement and dissipation effects. Potential barrier appears, to be a function of the material parameters, such as pinning centres kind and dimensions, elasticity shear modulus of vortex lattice, transport current density and as usually magnetic field and temperature. Critical current density for the flux creep process is reached, if the potential barrier height DU against tearing the vortices off vanishes, while the current satisfying the voltage criteria is achieved.

Figure 1: Scheme of the pancake type vortex in layered HTc superconductor

The model presented briefly above leads in its final form to the relation describing the potential barrier for flux creep process DU as the function of reduced current density i = j/j_C, where j_C is defined as critical current density for magnetic vortices creep process, when potential barrier disappears [2]:

(1)

Parameter a describing the elasticity forces of vortex lattice is defined in Eq. 3, while z appearing in Eq. 1 is determined according to the relation:

(2)

H_c is the critical thermodynamic magnetic field, d pinning centre width, l pinning centre thickness. In the paper the flat shape of the pinning centre has been considered.

Figure 2: The influence of the dimensions of the nano-sized pinning centres of width d, normalized to the coherence length x , on the critical current density versus applied magnetic field

As it follows from relation 1 the potential barrier height and therefore the critical current density are influenced significantly by the elasticity energy of the vortex lattice. Capturing of the vortices by the nano-sized pinning centres causes the deflection of the vortex from it’s equilibrium position in the regular vortex array, thus leading to an enhancement in the elasticity energy of the magnetic structure of the vortex lattice. This effect is the function of the shift of the single vortex from its equilibrium position in the lattice, described by parameter x, denoting the distance from the origin of the vortex core to the pinning centre edge. The elasticity energy increase is mathematically given by Eq. 3, assuming that the increase of the vortex elasticity energy is proportional to the square of the length of the vortex deflection from equilibrium position in lattice, with coefficient of proportionality expressed by the value of the parameter a.

(3)

Parameter c_s in equation 3 is the corresponding elasticity shear modulus, while l_a» l denotes the length on which the magnetic flux of vortices is distorted. We insert then expressions 1-3 into the constitutive relation describing generated electric field E in the flux creep process [2], just in the function of the potential barrier height. It allows us already to predict the form of the current - voltage characteristics and to determine then critical current density applying the appropriate electric field criterion. In Fig. 2 an example is shown of the computer calculations according to the elaborated model of the influence of the flat pinning centres width on the critical current density. This result indicates the meaning of the pinning centres parameters for analysis of the critical current density and then magnetic properties of the superconducting materials.

3. THEORETICAL ANALYSIS OF POSSIBILITY APPLYING HTc SUPERCONDUCTORS AS PERMANENT MAGNETS

The model presented in the previous section describes dependence of the critical current density of HTc superconductors on pinning centres parameters. The received results determine also the magnetic induction distribution in superconductors and describe therefore magnetic properties of the superconductors, which have very significant technical meaning, since ideal diamagnetism and trapped induction in superconductors is utilized already in magnetic levitation process, in superconducting bearings, trains and magnetic shields. This effect is utilized just in construction of the magnetic bearing built usually from a superconducting cylinder levitating around permanent magnets, as presents schematically Fig. 3, which shows six oppositely oriented permanent magnets in superconducting tube.

Figure 3: Schematic view of the cross-section of the magnetic bearing composed from 6 permanent magnet disks and levitating above them superconducting cylinder, with marked calculated by finite element method (FEM) magnetic field lines

The ideal diamagnetism of superconducting materials leads then to the expulsion of the magnetic field lines, as indicate the graphical results of calculations performed here using FEM numerical code. Fig. 4 presents results of the calculations using FEM method of the levitation force acting on the superconducting cylinder from the Fig. 3, as the function of the distance between the edge of the permanent normal magnets and the superconducting element. Very strong dependence of the levitating force on the distance is observed here.

Figure 4: Calculated levitation force versus distance b between superconducting cylinder and permanent magnets for magnetic bearing of the geometry shown in Fig. 3.

Figure 5. Calculated using FEM method of the magnetic induction distribution for the model of levitating magnetically train, built from the HTc superconductor placed above the permanent magnets arranged in the n-s-n magnetic configuration.

Another example of utilizing ideal diamagnetism of HTc superconductors appearing in weak magnetic field is shown in Fig. 5, which presents magnetic model of the track of levitating train, built from three oppositely arranged permanent magnets and the superconductor levitating above them. The expulsion of the calculated magnetic field lines for the model of this device seen in this figure results from an ideal diamagnetism of the superconducting element. It should be noticed in that point that although in present calculations ideal diamagnetism of superconducting materials has been utilized, in levitation process also remanent moment of superconductor can be applied in that aim what

Figure 6. Magnetic irreversible curve of sintered YBa₂Cu₃O_7-x sample. The arrows indicate the direction of magnetic field variation and finally magnitude of the trapped flux

brings even higher repulsion force. It follows directly from the giant remanent moment of HTc superconductor observed experimentally, as it was mentioned previously. Essential property of superconductor, which is the persistent current flow causes then that frozen magnetic flux theoretically should not change, while neglect creep process and leads further to the return of superconductor to its initial position in magnetic field, if shifted from it.

Figure 7: Magnetic induction profile in the HTc superconducting slab of the thickness 2x_m for magnetic induction cycle 0 → B_m → 0. The influence on induction distribution of the granular structure of the ceramic superconductors is presented in this picture.

The trapped magnetic induction magnitude discussed here is marked in Fig. 6 with arrow, as the value of remanent magnetization of measured according to presented in [3] method, loop of the magnetic hysteresis curve of sintered us YBa₂Cu₃O_7-x sample [3], in external magnetic induction cycle 0-B_m-0. Considered in theoretical model magnetic induction profile in this cycle of external magnetic induction, taking into account the granular structure of HTc ceramics is shown in Fig. 7, for the slab geometry of the superconducting sample exposed to the parallel, varying magnetic field. Basing on this figure we determine now mathematically flux trapped value normalized to the unit cross-section of the superconductor, it is average remanent magnetic moment, in function of amplitude of an external induction applied to the surface of superconducting slab B_m= B_e -B_c1 - D.

Figure 8: Dependence of the square root of the average trapped magnetic induction in HTc superconducting ceramic on the maximal magnetic induction in the magnetic induction cycle 0 → B_m → 0, versus grain’s radius.

Figure 9: Dependence of the square root of the average trapped magnetic induction in HTc superconducting ceramic on the maximal magnetic induction in the magnetic induction cycle 0 → B_m → 0, versus first critical magnetic field of superconducting matrix.

B_m is defined here as the difference between external magnetic induction B_e, first critical field B_c1/m₀, (m₀–vacuum permeability) and generally magnetic surface barrier D, appearing frequently on the surface of homogeneous superconductors. In present calculations however the surface barrier effects have been neglected. For B_m<0 trapped flux disappears:

(4)

For the next range of the magnetic induction increase, when B_p > B_m > 0 trapped flux normalized to the sample cross-section is given as:

(5)

where B_p = m₀j_cx_m - B_c1 is value of the first magnetic induction penetrating totally inside the sample. For next range of the magnetic field amplitudes described by the condition: 2B_p> B_m> B_p trapped flux is described by the following relation:

(6)

For higher values of the magnetic induction, it is satisfying the condition B_m> 2B_p the trapped flux saturates and is described then by the following formula:

(7)

Parameter n in the above equations describes the fulfillment of the ceramic material with superconducting grains, while B_g is the mean frozen magnet induction in individual grain:

(8)

Bc_1g denotes the first critical field in the grain of ceramic superconductor, while x_g is the product of the superconducting grain radius R_g and j_cg critical current density inside the grains, defined according to relation:

(9)

Results of computer calculations of the influence of the ceramic material parameters on the trapped flux magnitude are shown in Figs. 8 - 10 and indicate the influence of these parameters on the trapped flux, what has meaning from the point of view of the optimization trapped flux value. As it is seen trapped flux is the function of applied magnetic induction and for higher magnitudes finally saturates. Figure 8 presents the influence of the grain’s radius on the trapped flux value, while Figure 9 the dependence of the first penetration magnetic induction B_c1 of the superconducting matrix on the trapped flux. In Figure 10 is given the dependence of the trapped flux on the grain’s concentration. Performed analysis allows therefore to predict relevance of material parameters for receiving optimal flux trapping, essential parameter from the point of view of the application HTc superconductors as permanent magnets.

Figure 10: Dependence of the square root of the average trapped magnetic induction in HTc superconducting ceramic on the maximal magnetic induction in the magnetic induction cycle 0 → B_m → 0, versus grain’s concentration in cross-section of superconductor: (1) – 3*10 ⁵ cm^-2 , (2) – 10⁵ cm^-2 , (3) - 10 ³ cm^-2.

CONCLUSISONS

Technically important magnetic properties of the HTc superconductors have been described using the proposed new model based on pinning interaction, taking into account the granular structure of these unique ceramic materials. According to this has been considered the influence of the parameters of superconducting grains and surrounding them matrix on trapped magnetic flux. The importance of the interaction pancake type vortex - pinning centre for determining critical current and then trapped magnetic flux in HTc superconductors has been considered and relevance of this interaction emphasized.

REFERENCES

1. Sosnowski, J., Superconductivity and applications, Book Publisher of Electrotechnical Institute, Warsaw, Poland (2003) 1-200 – in Polish.

2. Sosnowski J., Vortex pinning in HTc superconductors, Studies of High Temperature Superconductors, Ed. A.Narlikar, Nova Science Publishers, Inc. New York (2006) 389-122.

3. Sosnowski J., Gajda D., Analysis of the flux trapped in HTc superconductors, Electrotechnical Institute Works, PL (2007) 231 126-134 - in Polish.

CR08-02

Gas flow through narrow gaps at low pressure in Super-insulation packages

Stipsitz J.¹, Dobrozemsky R.², Hirschl C.¹, Laa C.¹

¹Austrian Aerospace GmbH, Stachegasse 16, A-1120 Vienna, Austria
² Vienna University of Technology, Wiedner Hauptstrasse 8-10, A-1040 Vienna, Austria

Abstract

Super-insulation is the most effective thermal insulation for cryogenic applications and is employed in the vacuum space between the cold surfaces and the outer vacuum vessel. Satisfactory insulation performance can only be maintained if gas conduction is suppressed.

Conductance measurements have been performed by means of an ultra-high vacuum system. Based on these conductances and outgassing rate data from own measurements and literature, the pressure drop between the center of the Super-insulation and the chamber wall was calculated.

Introduction

Super-insulation (SI) is composed of alternate layers of reflector foils and spacer material. The reflector (aluminum or aluminized polyester foil) minimizes radiative heat transfer. The spacer prevents direct contact of, and minimizes solid conduction between adjacent layers. Satisfactory insulation performance can only be maintained if gas conduction is suppressed. Gas conduction becomes significant only for pressures higher than 1E-4 mbar (1E-2 Pa) and therefore the interstitial volume must be pumped and kept below this level, see Figure 1 and [1].

Figure 1: Typical SI performance with gas conduction [2]

The pumping of SI interstitial volume has been analyzed and measured by several authors [3-7]. Due to the low conductance of the narrow gaps in molecular flow, the gas pressure within the SI may be higher than the pressure measured at the chamber wall. Pumping of the SI internal volume depends on the outgassing of the used SI materials and the conductance of the narrow gaps between the foils for molecular gas flow. Knowing these two parameters, the pressure inside the SI can be estimated.

1. Measurement

1.1 Test Setup

The conductance measurements have been performed by means of an ultra-high vacuum (UHV) system equipped with Bayard-Alpert gauge (BAG) and quadrupole mass spectrometer (QMS). Flow rates Q of pure gases were admitted to the upstream side by a variable leak valve (VLV). Upstream pressures P_up were adjusted by means of a capacity diaphragm gauge (CDG) and the downstream pressures P_dwn were measured by the BAG with the gas composition monitored by the QMS. The experimental setup is illustrated in Figure 2.

Figure 2: Setup for conductance measurement

A cylindrical test specimen was prepared by winding a SI strip of length B_m = 2400 mm and width L_m = 50 mm around a 10 mm diameter stainless steel rod. This cylindrical specimen fits into the 32 mm inside diameter stainless steel test chamber in a way that the distance between layers is comparable to the actual situation in an extended SI blanket with no gap between the outermost layer and the wall of the test chamber (layer density: 3.3 layers/mm, distance between adjacent layers H_m = 0.3 mm), see Figure 3.

Figure 3: Test chamber with specimen

1.2 Measurements

The pressure reading instruments as well as the effective pumping speed S_eff on the downstream side were calibrated with respect to a secondary standard.

On pages 67-68 of [8] the throughput Q is defined as the product of the pumping speed S_p and the inlet pressure P, i.e.

, (1)

and the effective pumping speed S_eff obtained in a chamber connected by a conductance C to a pump having a pumping speed S_p is given by

. (2)

For equilibrium conditions the mass flow must be same for both the upstream and the downstream sides of the specimen. The conductance C of the SI test specimen can now be calculated from

. (3)

For all gases employed, equilibrium conditions were achieved within rather short periods of time. This means that gas flow rates Q as well as contributions due to outgassing and adsorption are constant during measurement periods. The upstream pressures were adjusted in the 1E-3 and 1E-2 mbar (1E-1 and 1 Pa) range to maintain molecular flow conditions with respect to the narrow spacing of the SI-layers, and the BAG-readings were evaluated. With the test specimen at 21°C (294 K) the following conductances of the SI package were derived (Table 1).

	C [liter/s]
H₂O	0.11
Ar	0.19
N₂	0.23
H₂	0.90

Table 1: Measured conductances for B_m=2400 mm and L_m=50 mm

For ideal gases in molecular flow, the conductance of an aperture is proportional to (1/M)^1/2, with M the molecular weight of the gas – see pages 80-81 of [8]. There is a good fit of the measurement results for Ar, N_2, and H₂ with this theoretical relation (±3%). Water vapor shows a lower value, because it is no ideal gas. It can be adsorbed to the MLI surfaces and diffuses more slowly through the specimen [9].

2. Analysis

Based on these conductances and outgassing rate data from own measurements and literature (see Table 2), the pressure drop between the center of the SI (P_up) and the chamber wall (P_dwn) was calculated.

	Outgassing rate [mbar.liter/(m².s)]	Main outgassing products
Glass fiber spacer	1E-9	H₂
Polyester spacer	2E-9	H₂
Aluminium [10]	1E-10	H₂
Aluminized polyester [11]	3E-7	H₂O

Table 2: Measured outgassing rates for spacer materials and literature values for foils

The unit area for the outgassing rates of the spacer materials is defined as the one-sided geometrical area. The unit area for the outgassing rates of the foil materials is the exposed surface area. These considerations depend essentially on pumping cycles, including baking and purging.

For the calculation it was assumed that the total outgassing of one foil plus one spacer layer occurs in the center plane of a SI blanket of 1 m² (see Figure 4), resulting into flow rates Q₀ from the center to either side.

Figure 4: Model for calculation of pressure drop in SI blanket of 1 m², having length 2L and width B

The pressure differences listed in Table 2 were calculated by means of

. (4)

For a slot with H<<B, the conductance is proportional to B·H²/L – see page 84 of [8]. Therefore the SI conductance C_SI for a SI blanket of 1 m² can be calculated from the measured conductance C, which relates to the specimen geometry (B_m, L_m), using the formula

. (5)

The calculated pressure drop caused by the outgassing of different SI materials is given in Table 3.

	Main outgassing products	C_SI [liter/s]	P_up – P_dwn[mbar]	P_upat P_dwn=1E-6 mbar
Glass fiber spacer	H₂	0.0375	1.7E-8	1.0E-6
Polyester spacer	H₂	0.0375	2.3E-8	1.0E-6
Aluminium [10]	H₂	0.0375	5.3E-9	1.0E-6
Aluminized polyester [11]	H₂O	0.0046	5.9E-5	6.0E-5

Table 3: Calculated typical pressure differences for SI blanket of 1 m²

conclusionS

The suggested method for the measurement of conductances of super-insulation in molecular flow yields consistent results.

Assuming an outside pressure P_dwn of 1E-6 mbar (1E-4 Pa) and the use of polyester foils without perforation, the pressure within the SI will come close to the level of 1E-4 mbar (1E-2 Pa), where gas conduction becomes important. In this case the influence of the spacer outgassing can be neglected.

For aluminum foils the spacer will contribute most to the outgassing and there is no significant pressure drop between the center of the SI and the chamber wall.

Acknowledgements

The work was carried out in a cooperation of the Vienna University of Technology and the Thermal Systems department of Austrian Aerospace (AAE), the largest supplier of space products and related ground support equipment in Austria, focusing on electronics, mechanisms and thermal insulation. The company of 130 employees is owned by the Saab Space Group. The business mission of AAE is to market, develop and manufacture satellite equipment for the global space industry and cryogenic insulation for terrestrial applications that strengthen and support the competitiveness of its customers. AAE produces high quality Multi Layer Insulation (MLI) since 1991, and is the leading supplier of MLI for spacecraft of the European Space Agency and a number of other European programs. AAE designs, manufactures and integrates thermal hardware products in-house or at the customer's site. As a product diversification, AAE has adopted space MLI for cryogenic applications, covered by the ‘Coolcat’ line of insulation products. ‘Coolcat’ is used for the insulation of superconducting magnets (MRI, NMR, and others); cryostats; transfer lines; and automotive liquid hydrogen tanks. AAE offers engineering consultancy and thermal analysis for the customer systems. In its state-of-the-art manufacturing facility for insulation, AAE is equipped with automated cutting machines and all processes for production on an industrial scale.

Prof. Dr. Rudolf Dobrozemsky is guest scientist at Institut fuer Allgemeine Physik (IAP) of Vienna University of Technology. Besides its program of introductory physics courses, IAP is well established in the fields of plasma, surface, nanotechnology, and ultrasonics research with a broad experience in the respective experimental and analytical techniques.

The project was funded by the Austrian Federal Ministry of Transport, Innovation and Technology (Austrian Research Promotion Agency), and Austrian Aerospace.

REFERENCES

1. Augustynowicz, S.D., and Fesmire, J.E., Cryogenic Insulation System for Soft Vacuum, Advances in Cryogenic Engineering, (2000) 45 1691-1698

2. Laa, C., Hirschl, C., and Stipsitz, J., Heat Flow Measurement and Analysis of Thermal Vacuum Insulation, presented at CEC/ICMC 2007

3. Keller, C.W., Cunnington, G.R., and Glassford, A.P., Thermal Performance of Multilayer Insulation, Final Report of NASA Contract NAS 3-14377 (1974) Section 5

4. Kaganer, M.G., Thermal Insulation in Cryogenic Engineering, Israel Program for Scientific Translations, Jerusalem (1969) 170-174

5. Bapat, S.L., Narayankhedkar, K.G., and Lukose, T.P., Performance prediction of multilayer insulation, Cryogenics (1990), 30 700-710

6. Bapat, S.L., Narayankhedkar, K.G., and Lukose, T.P., Experimental investigation of multilayer insulation, Cryogenics (1990), 30 711-719

7. Reiss, H., A coupled numerical analysis of shield temperatures, heat losses and residual gas pressures in an evacuated super- insulation using thermal and fluid networks. Part 1: Stationary conditions, Cryogenics (2004) 44 259-271

8. Roth, A., Vacuum Technology, Elsevier Science B.V., Amsterdam, The Netherlands, 3^rd Ed. (1990)

9. Dobrozemsky, R., Menhart, S., and Buchtela, K., Residence times of water molecules on stainless steel and aluminum surfaces in vacuum and atmosphere, J. Vac. Sci. Technol. A 25(3) (2007), 551-556

10. Ishimaru, H., Ultimate pressure of the order of 10^-13 Torr in an Al alloy chamber, J. Vac. Sci. Technol. A7 (1989) 2439

11. Glassford, A.P.M. et al, Effect of temperature and preconditioning on the outgassing rate of double aluminized mylar and Dacron net, J. Vac. Sci. Technol. A2 (3), (1984)

CR08-46

New developments of non-metallic cryostats for high sensitive electronic devices and other applications

Klier J., Spörl G., Schumann B., Binneberg A., Herzog R.

Institut für Luft- und Kältetechnik gemeinnützige GmbH,
Bertolt-Brecht-Allee 20,01309 Dresden, Germany

Abstract

For a number of high-precision devices working at low temperatures magnetic stray fields can drastically reduce their sensitivity. This problem can be avoided by the use of non-metallic cryostats. For most applications of high-T_c-SQUID devices in material investigations, geological exploration and medicine low noise cryogenic cooling systems are required. Under certain conditions, however, there is the demand of special geometries between sensor and testing area, and furthermore position independence and mobility. Examples are measurements of heart and brain biomagnetic signals or non-destructive evaluation on parts of air planes and reinforced concrete constructions. A series of special cryostats has been developed to meet these demands. The construction of the inner vessel allows reliable cooling of a sensor in vacuum referring to liquid level and stability of temperature and when the cryostat is turned around its transverse axis. For very small cryostats (< 150 mm) the sensor can be placed in vacuum on a sapphire rod mounted on a heat transfer area cooled by cryogenic liquids.

Introduction

Extensive work in the field of superconducting electronics has lead to the development of superconducting interference devices, known as SQUID magnetometers. These highly sensitive devices are able to measure very small magnetic fields down to 10^–14 Tesla. Their sensitivity, however, depends on the level of thermal noise, which can be minimized through the use of non-metallic materials for the cryostats providing the necessary low temperatures. SQUIDs made from conventional superconductor usually operate at liquid helium temperatures. Due to the need of liquid helium as cooling medium and the high costs of low-T_c SQUIDs there are some limits in their application. This situation changed after the discovery of the high-T_c superconductors in 1986 and the subsequent development of high-T_c-SQUIDs. Especially high-T_c superconductors like yttrium barium copper oxide (YBCO) with a transition temperature above 90 K allow the use of liquid nitrogen as cooling agent for high-T_c-SQUIDs.

For various applications of high-T_c-SQUIDs we have developed a variety of non-metallic cryostats. In this paper we present some of our most exceptional designs and latest developments.

General Properties of non-metallic cryostats

Usually standard laboratory cryostats are double-walled vessels, made from glass or metal, which are evacuated. In the case of glass the walls of the vessels are silver-plated in order to reduce thermal radiation into the cryostat. In the case of metal the spacing between these walls usually contain multilayer insulation. Most common the insulation material consists of thin polyester foils (superinsulation) coated with aluminium on one side. Such cryostats can be used when their magnetic influences do not affect the experimental results.

Changes of magnetic fields, however, induce eddy currents in metals. There is further thermo-magnetic noise in metals which can compromise the sensitivity, interaction with the measuring signal or reduction in the signal to noise ratio. Therefore cryostats free of any metal near the sensor are essential for applications with high magnetic field sensitivity. This requirement is fulfilled when glass fibre reinforced epoxy materials are used for the cryostat fabrication.

Foils of superinsulation can act as shields for higher frequency signals or induce additional noise. To avoid these problems we use polyester foils with partially demetallized aluminium face. The use of these modified foils does not affect the vacuum condition and evaporation rate of the cryostats. For some methods of geological exploration, however, the insulation must be replaced by completely non-metallic polyester foils.

Designs and Applications of non-metallic cryostats

For cooling highly sensitive measuring devices (e.g. to detect magnetic fields down to 10^–14 T) we have developed and manufactured special liquid nitrogen cryostats. The high sensitivity of the devices require effective means against outer electromagnetic disturbances. To meet this requirement the cryostat must be free of any metallic parts in the vicinity of the sensor. Composite materials such as glass fibre reinforced epoxy resin are used instead of metals. The design of thermal insulation and integrated shielding were adopted to the special requirements, such as transparency for high-frequency radiation, very low intrinsic noise, unusual geometrical conditions, long operating time, suitable for field tests, and low fabrication and maintenance costs.

Typical fields of applications are:

• medicine à e.g. measurements of currents of heart and brain (i.e., biomagnetic measurements with SQUIDs)

• material tests à e.g. detection of micro defects in materials by non-destructive evaluation

• geology à e.g. surveying of the earth in the vicinity of the surface, specification of building ground and ground water

• science à e.g. biophysical and standard investigation at and with SQUIDs, scanning tunnelling microscopy, etc.

1. Variable distance and position independent cryostats

The demand of using SQUID systems for the non-destructive evaluation on parts of air planes and reinforced concrete constructions initiated our reserach activities to develop variable distance and/or position independent cryostats. Having a cryostat in which the gap between SQUID and bottom of the tail of the vacuum vessel can be adjusted, opens the way for high resolution magnetometry of biological specimens. So it is aimed to optimize the local resolution of the measuring method by changing the distance between device and detector element outside the cryostat.

The SQUIDs are located in vacuum on a sapphire rod (outside of the liquid nitrogen vessel) mounted on a copper face cooled by the liquid nitrogen, see figure 1 and 3 (left). The smallest possible distance between sensor and object of measurement can be only 4 mm. For universal use of the cryostats special inserts are developed for filling and refilling, made of different non-metallic materials. The SQUIDs can be changed without warming-up the cryostat. Continuous measurements are possible by use of automatic filling stations.

Figure 1: Variable distance cryostat in z-direction.

The distance between the device and the outer wall of the vessel can be changed by spring bellows mounted on top of the cryostat. In this way the whole liquid nitrogen vessel is moved relative to the outer vacuum vessel, see figure 1. The bellows are pressed or streched by several timing bolts or precision bolts. The distance between SQUID and bottom of the outer vessel can be varied between 1 mm and several 10 mm. The isolation vacuum usually holds more than 6 months and the evaporation rate is less than 7% per day of LN₂.

For some applications it is necessary to vary the distance between sensor and outer wall of the cryostat not only in the z-direction but also in the horizontal plane, i.e., xy-direction. The realisation of such a demand is shown in figure 2.

Figure 2: (left) Variable distance cryostat: The position between sensor and cryostat itself can be varied both in horizontal and vertical direction (xyz-plane). (right) Sketch of the adjustment facility in horizontal direction (xy- direction).

The position independent cryostats are designed in such way, that they can be turned around their axis by 360°, see figure 3 (right). Hereby the construction of the inner vessel allows reliable cooling of the sensor with respect to liquid level and stability of temperature. Changes of temperature are less than 0.2 K. By turning the cryostat there are no significant changes of the evaporation rate and maximum working time.

Figure 3: (left) Drawing of a position independent cryostat, i.e., the cryostat can be turned by 360°. (right) Picture of a position independent cryostat which can be turned by 360°. In addition the distance and plane between SQUID and vacuum vessel is variable and can be inclined.

2. Very small mobile cryostats

SQUIDs can be used for magnetic non-destructive material tests in order to find very small failures (less than 1 mm) in thick material layers or hidden metals (e.g. steel in concrete). For this application the cryostat should be as small as possible and rotatable. A series of cryostat has been developed to meet these demands. The smallest of this type is shown in figure 4. The diameter is about 60 mm and the height about 140 mm. The weight of the cryostat is less than 500 g. The operational time is three hours in thermal equilibrium, when filled with about 36 ml of liquid nitrogen.

The construction of the inner vessel allows a reliable cooling of a sensor in vacuum according level and stability of temperature, when the cryostat is turned around its axis. The working temperature depends on the position and varies from 77.5 K to 78 K. The stability of the temperature in one position is about 0.1 K.

Figure 4: Drawing (left) and picture (right) of a very small mobile cryostat.

3. Lift cryostat – for superconductor-based magnetic levitation lift

The use of magnetic bearings has become of increasing importance. Although electromagnets permit the contactless transmission of power and, thus, rotation or motion, the generation of the required guiding forces in the magnetic bearings require costly measuring and control techniques. This problem could be overcome by the use of superconducting magnetic bearings.

For the realisation of a superconductor-based magnetic levitation lift the core of the superconducting magnetic bearings comprises a magnet rail made of conventional permanent magnets and superconducting blocks cooled to liquid nitrogen temperature. Arranged at a defined distance from the magnet rail, the superconductors ‘freeze’ the magnetic field of the permanent magnets. This puts the superconducting magnets in a position to maintain by themselves a certain distance from a magnet rail. Linear drives ensure synchronous, contactless lifting and lowering of the superconductor-based magnetic levitation lift.

The lift cryostat, especially developed for this project, is made from glass fibre reinforced epoxy material and cooled by liquid nitrogen. A predetermined space is kept between the superconductors and the surface of the magnet rails while the superconductors are cooled to below the transition temperature of about 92 K, see figure 5 (left). As soon as this temperature is undershot, the cryostat is non-positively and contactless anchored in the magnetic field of the magnet rails (at a distance of 1 mm). Since the configuration of the magnetic field does not change along the rail, the cryostat can move only along the rail at a given distance.

The cryostat holds 12 litre of liquid nitrogen. This quantity is sufficient for operating the magnetic levitation lift, see figure 5 (right), for approximately 10 hours without replenishing the nitrogen. Thereafter, some 5 litres of evaporated nitrogen must be replenished. Possible applications of such a system, superconductor-based magnetic levitation lift, are in clean rooms and pharmaceutics.

Figure 5: (left) Picture of lift cryostat and superconducting bearings. (right) First working model of a superconductor-based magnetic levitation lift.

4. Liquid Neon cryostat – cryo-collector

For climate and environmental research the analyses of trace gas within the atmosphere, up to heights of 40 km, is essential. The focus of such measurements are the halogenated trace gases, like Chlorofluorocarbon, which are responsible for the hole in the ozone layer. In these heights of the atmosphere the air pressure is only a few hPa. However, for gas chromatography certain amounts of gas samples are necessary. To overcome this problem one uses so-called cryo-collectors, which are transported to the appropriate height via large stratospheric balloons. The working principle of such cryo-collectors is to freeze the air in special sample containers. Due to the experimental conditions of such balloon flights, the time for cryogenic gas collection is quite short, sometimes only a few minutes. Therefore one needs a cryogenic fluid with low boiling point and high heat capacity. This makes liquid neon quite suitable as refrigerant. Figure 6 shows such a LNe cryo-colector.

Figure 6: Liquid Neon cryostat: A cryo-collector for air samples in the atmosphere.

conclusionS

We designed and fabricated a number of non-metallic cryostats, mainly using liquid nitrogen as cooling agent. The field of applications were high-precision devices working at low temperatures. Usually magnetic stray fields can drastically reduce the sensitivity of such devices. This can be avoided by the use of non-metallic materials, in our case glass fibre reinforced epoxy materials, for the fabrication of the cryostats. For some special applications of high-T_c-SQUIDs there is the demand of special geometries between sensor and testing area, and furthermore position independence and mobility. Therefore a series of special cryostats were developed.

CR08-03

Cryogenic Distillation Column Behavior
at the Variation of an External Factor

Pearsica C.,Stefan L.,Preda A.,Vasut F.

Institute of Isotopic and Cryogenic Technologies - ICIT Rm.Valcea,Romania

Abstract

Behavior of a cryogenic distillation column for a case of tritium-deuterium separation under periodic instability of the distillation process was studied by mathematical modelling. Influence of an external factor, analyzing a non steady state with the fluctuation of the hydrogen level in the column condenser was subject of investigation. The mathematical model is based on balance equations, column operated at total reflux and sinusoidal variation of the hydrogen level. There is studied the non steady state evolution in the distillation equipment. For several situations, results were presented in specific diagrams and plots.

Introduction

In order to realize correct operation of the distillation column from a liquid hydrogen isotopic plant it is necessary to maintain all the parameters constant, which corresponds to steady state. Practically, this cannot be realized and the operating column would suffer some perturbations, the regime would become non-steady. The effect of the non steady state is the decrease of the separation performance of the column [1].

We followed the way the variation of the liquid level in the condenser influences the operation of the cryogenic distillation column. Such a perturbation determines an immediate modification of the liquid flow in time and in length, but also modification in time of the condenser holdup and on column packing.

The system (the distillation column) was defined for the mathematical analysis of the distillation process of a tritium deuterium mixture.

The modell is analyzing the following situation: after attained normal operation steady state, there is applied to the system a sinusoidal perturbation by the variation of the liquid level from the column condenser. This moment represents the beginning of the study. The concentrations existing in the column at this moment of time there are be called in the paper as initial concentration.

A sinusoidal function that describe the liquid variation from the column condenser is defined by two variable parameters: the size (amplitude) and the frequency (period). This perturbation has a finite action, the intention being to analysis the influence of this kind of operation on the distillation process.

There are calculated the gas concentrations for every time step, on each plate. The results can give information about the process quality and the time for the entrance in the steady state (defined like the time between the perturbation initiations, till attaining the constant values corresponding to steady state).

The graphics for the entrance in steady state represent the moment from when the concentration has constant values.

1. Transfer Unit Height

A previous study analyzed the non-steady state for a case of a cryogenic distillation column condenser cooling circuit. The non-steady state determines the variation of the liquid level in the column condenser. This one induces the variation of the holdup at the top of the column, which determines the variation of the liquid down-flow.

The quality of the transfer element can be expressed using the Transfer Unit Height, IUT, where H_col is the column height and N is the number of plates corresponding to given separation [2].

The non-steady state involves the time variation of the Transfer Unit Height, so this parameter is calculated at every time interval and represents an indicator of the distillation process quality.

In order to show the non stationary character of this parameter, it was named Apparent Transfer Unit Height, . This factor is related to the Reference Transfer Unit Height: where N_T is the number of theoretical plates calculated for the steady state column operation. Input data for the hydrogen isotopic separation plant are: N_T = 30, H_col = 2m which means IUT_ref = 0.067m, for a tritium-deuterium mixture.

2. mathematic MODELL for studying the non-steady state of a separation column working at total reflux with perturbation in condenser

Due to the specific characteristic of the isotopic distillation column, big number of theoretical plates, so a big amount of package, a hypothesis was done, that the gas flow along the column is constant and the flow variation is all taken by the liquid flow variation.

We make the simplifying theory that the pressure variation from the column does not influence the separation factor α [3].

In a situation like this the distillation column is the one represented in Figure 1.

The holdup from the distillation column filling is uniformly distributed all the way. The equilibrium equation for the column:

(1)

where Hc and Hn are the quantities of liquid in the condenser and on the respective plate. In fact, the “plate” means a part of the column, corresponding to a single IUT_ref. Total vaporization is considered in the boiling vessel, while the column is operated at total reflux and total boil-off, x1=y0.

Balance of a stage n: (2)

(3)

For condenser: (4)

(5)

Sinusoidal variation of the liquid hold-up is considered in the condenser, according to the equation, with amplitude A and the period w :

(6)

After the calculation it is obtained: (7)

For the last plate it can be written:

(8)

For the stage n (9)

Making the next calculation the concentrations throughout the column can be obtained. Applying the method of finite time-differences Dt, the gas phase concentration throughout the column, on each respective stage:

(10)

The liquid phase concentration results from the relation of separation factor:

(11)

In a similar way, for the condenser:

(12)

(13)

The equation system that describes the process, which takes place in the column at the liquid level variation in the condenser, is formed of the equations (9)-(13), which together with the functions that describe the variation of the holdup in the condenser and on each plate will be solved throughout the column. A program, solving the equations system above was developed according to a logical diagram [4].

Results and discussions

For the analysis of the perturbation influence owed to the variation of the liquid level in distillation column condenser operated at total reflux, we considered the following cases [5]:

- initial concentrations of 25% and 50% tritium in the tritium-deuterium mixture

- variable sinusoidal amplitude function between 0 and 40% change of H_co

- sinusoidal function period between 0 and 30 minutes.

Some results are presented in the diagrams. Such as, in Figure 2 there are the concentration profiles along the column at the entrance in steady state for the case, when the initial concentration is 50% T/T+D.

There was dealt the situation when amplitude was kept constant at a value of 10% and takes place a variation of the perturbation frequency: 10, 15 and 30 minutes. The variation in time of the gas concentration in the top of the column, in the case of 50% initial concentration and constant period is shown in Figure 3.

A similar situation is dealt in the case when initial concentration is 25%. There are represented the gas concentrations at the entrance in steady state in the two cases: when is kept the amplitude constant (Figure 4) or the period constant (Figure 5).

For the latter case (Figure 5), there is observed that together with the growth of the amplitude value, the profile changes in the way of decreasing the separation performance, as a consequence, as the amplitude value grows, the more detrimental the situation is.

Noticing with attention the profiles drawn in figure 2, inversion of the concentration curve profile was observed. Considering that this result gets out of normality it is recommended to avoid to achieve or to overcome the value of 30% amplitude function and also it is recommended a period more than 15 minutes.

For the case when the amplitude is maintained constant but perturbation frequency varies (fig. 2 and 4 respectively), from the calculations we get to the conclusion that for acceptable values of the perturbation frequency, I mean a frequency of 2-4 times an hour, the concentration profile at the entrance in steady state is almost the same, while once overcome the frequency of 4 times an hour takes to a deep lack of balance, case when the distillation column manipulation becomes difficult and the time to achieve the steady state is longer.

The liquid level variation in the distillation column condenser causes decrease of the separation performance of the column, this one being lower as the perturbation is stronger.

The influence of the perturbation on the operation column can be followed through the IUT_calc parameter. There were considered for plotting the IUT values at the entrance in steady state. This parameter is represented as a function of amplitude at different values of frequency in the analyzed cases. In Figure 6 (the case of initial concentration 25%T/T+D) and 7 (the case of initial concentration 50%T/T+D), IUTcalc increases with the amplitude value and perturbation frequency. From calculations results that IUTcalc can increase 3.5 times towards IUTref value which means an increase of the column height with the same value if it is wanted the same separation.

The model allows the representation of this profile at any time. The file in format *.dat containing the necessary data for this kind of representations can be used if we want to analyze the evolution of the concentrations along the column in time. The profiles presented in the previous figures represent the moment of entrance in steady state.

CONCLUSIONS

Graphic representations can give information about the operation of a distillation column under conditions of non-steady state, resulted as a consequence of perturbation in sinusoidal shape of the liquid level in condenser. This way, concentrations profiles along the column at its evolution in time can be traced, but most of all, at the entrance in steady state, as is defined in the introduction, different values as S separation, Fenske number N_F or the height of the transfer unit IUT can be determined. Every case, by itself, may offer information upon the behavior of the distillation column operated at total reflux when the perturbation from condenser is described by sinusoidal function.

The models of calculation realized can be used to study more aspects of the evolution of the non-steady state in the distillation plant.

References

1. Peculea M., Instalatii criogenice, Ed. Conphys (1997)

2. Constantinescu D.M., Dimulescu A., Peculea M., Ursu I., Influenta instabilittaii vaporizatorului asupra functionarii coloanelor de distilare izotopica, St. Cerc.Fiz., Tom 31, nr.2, p 119-127, Bucuresti (1979)

3. Stratula C., Marinoiu V., Sorescu Gh., Metode si programe de calcul al proceselor de distilare, fractionare si absorbtie, Ed. Tehnica, Bucuresti (1976)

4. Toma M., Odagescu I., Metode numerice si subrutine, Ed. Tehnica, Bucuresti (1980)

5. CLAUDIA PEARSICA, Contributions on the Non-steady in a Liquid Hydrogen Isotopic Distillation Plant, Ph.D. Thesis (2007)

CR08-44

ANALYSIS of PERIODIC ADSORPTION PROCESSES,
USED In NEON And HELIUM PRODUCTION

Bondarenko V. L.¹, Simonenko Yu. M.²

¹Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,
107005, Moscow, Russia
²Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine

ABSTRACT

The analysis of inert gases purification in a single adsorber has been made. The criteria for the comparison of adsorption devices with different canal geometry have been offered. The dependence of the sorbent dynamic capacity on the operational and constructive factors has been studied. The complex of processes characteristic of the working phase and the regeneration period has been considered. The obtained information permitted to reduce the time of unproductive fragments of the cycle and increase the efficiency of the periodical purification in adsorbers.

INTRODUCTION

The sorption technologies are widely used in the technologies of rare gases extraction. There is a number of special requirements to the adsorbers used in this branch. The desire to keep the low content of residuals often increases the length of the working cycle and, in some cases, is accompanied with the increasing of the number of adsorbers necessary for ensuring the continuous purification.

The said specific phenomena are different in the adsorbers of different shape. It can be assumed that for each set of operational parameters there is a definite correlation of the adsorber canal length and its diameter, at which each kilogram of the sorbent will perform its functions most effectively. To find this optimum a sequence of process characteristics of a single adsorber work should be considered.

THE ANALYSIS OF MIXTURE SEPARATION PROCESSES IN THE ADSORBER

Figure 1 illustrates the temperature change in the adsorber during one cycle. The sorbent

Figure 1: The temperature change during one cycle of the adsorber work

regeneration on the stage t_H is achieved by the heat feeding through the walls of the device or by means of supplying the heated circulate flow through the sorbent layer. The shown operations sequence (or most of its stages) is characteristic for the neon and helium separation devices, the neon-helium mixture purification and a number of other technologies of production of rare gases and their isotopes.

An important utilitarian function of the adsorber is the admixture holding (1) and receiving the pure product on the outlet (2) during the time t_W. First of all, this property is characterized by the value of the sorbent adsorbing capacity а_adm [norm.m³/kg] with respect to the admixture. The value а_adm is equal to the volume of admixture-substance, captured by the sorbent mass unit. During the gas mixtures separation, their components mutually influence the individual values of adsorbing capacities.

For example, according to the Langmuir theory, the following correlation is acceptable for the estimation of the adsorption capacity in the binary mixture

, (1)

where: is the sorbent adsorbing capacity against the pure component of admixture under the conditions of total saturation;

y_adm and y_cl are the inclusion volume fractions of the components of the mixture;

Р is the working pressure;

b_adm and b_cl are constants, depending on the properties of the sorbent and the mixture components.

Reference sources give the sorbents characteristics for the cases of the sorbent “static” saturation. These absorption levels are usually rather higher than the characteristics obtained under real conditions. The measure of discrepancy between the adsorbing capacities of the dynamic А_adm and static а₁ depends of the canal shape, flow rate and a number of other operating and design factors. For the interpretation of the sorbent saturation we introduce the value b=A_adm/a_adm called the “sorbent usage degree” and the degree of the saturation of the working layer. Notwithstanding the indisputable reliability of А_adm, it characterizes only one (even though very important) cycle fragment. The dynamic adsorption capacity unambiguously determines the duration of the working phase (t_W), but cannot totally reflect the whole complex of the adsorption separation processes. Indeed, the task of getting the maximum of the saturation degree а_adm®А_adm (or b®1) is easy to realize. It is achieved, for example, by the multiple increase of the time of the working phase t_W. Though this regime is not attractive from the customer’s point of view as it is realized at negligibly small product recoveries, we believe that the inconsistency of b as the optimization object is compensated by the introduction of the time factor.

(2)

where: (t_S) is the duration of the whole period which includes the working time (t_W), the duration of heating (t_R) and pumping (t_V) during the regeneration and also the cooling time before putting into operation (t_C).

The value F is equal to the volume of the admixture adsorbed from the flow (y_adm), falling on the sorbent mass unit in a time unit. In terms of physics, in a freer interpretation, this coefficient can be interpreted as the “mean performance of the layer work”. The reduction of the F criterion to the mass unit (via specific adsorbing capacity а_adm, m³/kg) is not accidental. It is the quantity of the sorbent under the conditions of the periodical operation that in many respects determines the separation operating expenses. But one and the same mass of sorbent can be placed in the canals of different shape. The influence of the way of “packing” the sorbent in the device is evident if we compare the adsorbers with the unequal correlation of the diameter and length of the canal. In long devices (with small radial layer extent) the ideal conditions for the heat transmission through the wall will be created. This will result in the reduction of the time of the heating t_R and cooling t_C. At the same time, in such adsorbers, the processes, determined by the speed parameters will last longer (t_W, и t_V). In short canals the reverse situation will be observed: it is the processes of heat transfer through the wall of the adsorber that will become overextended (t_R и t_C). Besides, for the second case, a small degree of the sorbent usage will be typical (b ®0).

To calculate the cycle fragments duration a block of calculation models has been designed, which reflect the sequence of the single adsorber work. A complex of experimental stages has been created, which allowed obtaining the data, characterizing individual adsorption processes, in particular, the values of the static sorption capacity of silica gels and activated carbon during the pure nitrogen, neon and helium adsorption. The sorption heat of the mentioned gases has been estimated. The influence of the operating and design parameters on the sorbent dynamic capacity has been studied (these data are interpreted by the b factor). The heat phenomena in the sorbent layer during the regeneration and cooling have been investigated. The data, characterizing the pumping process of the adsorber, having additional admixtures, have been accumulated.

The obtained information allowed making resulting estimations of the characteristics of the devices with different size correlations. An adsorber for purifying the neon-helium mixture from nitrogen has been taken as an example object of optimization. Operation parameters, taken as the source data are characteristic of the typical technology of light inert gases (Ne and He) extraction. Two ways of heat feeding during the regeneration have been studied: through the wall of the device and by means of the heating gas supply (Figure 1). A series of devices of the same volume (v = 0,1 m³) have been analyzed. The adsorbers had different diameters and, naturally, different canal lengths L. For the indicated diameters the length ranged from L = 51 m tо L = 1,4 m correspondingly. The influence of the adsorber geometry on the duration of certain cycle stages is shown in Figure 2.

The dependence of the optimization criterion for different device geometry is shown on Figure 3. The value F allows forecasting the utilitarian factors of the adsorbers with the given capacity. The generalization of the obtained information showed that in the studied range and at the stipulated device volume there exists a definite correlation of L/D dimensions at which the F coefficient possesses the maximum value. The value of the said optimum is influenced, besides the operational and design factors, by the way of heat transfer in the process of regeneration. For the considered example, in the case of heat feeding through the wall of the device, the L/D optimal values are in the range from 50 to 100. When the sorbent is heated by means of the flow circuit, these values decrease to 20…50.

The information, shown on the graphs 2 and 3 is true for the following conditions: sorbent mass m = 45 kg; imaginary flow speed w = 0,04 m/s; static adsorption capacity of the admixture (N₂) а_adm= 0,377 norm.m³/kg; sorbent packed density (coal SKT-4) r = 430 kg/m³; admixture content (nitrogen – in neon-helium mixture) y_adm= 0,1; working pressure Р = 1,0 MPa; working temperature Т = 84 К.


a	b

Figure 2: The duration of certain phases of the adsorber working period

a - during the regeneration by means of heat transfer through the wall of the device;

b - during the regeneration by the heat flow

Figure 3: The influence of the canal diameter and the regeneration method

on the optimization factor F (upper curve – not taking into consideration the adsorber heating time)

CONCLUSION

The realized research proves the correctness and consistency of the optimization criterion F. It allows revealing the most successful “packing shape” of the given sorbent mass which will result in each kilogram of the sorbent performing its functions most efficiently.

REFERENCES

1. Arkharov A. M., Bondarenko V. L., Savinov M. Yu. et al. System of neon concentrate fine purification. Vestnik MGTU. Special issue «Refrigeration, cryogenic technology, systems of air-conditioning and life provision» (2005) 24-32.

CR08-45

NEON LIQUEFIERS AND THEIR USAGE IN THE INSTALLATIONS FOR RARE GASES EXTRACRION

Bondarenko V.L.¹, Diachenko Т.V.², Diachenko O.V.²

¹Moscow Bauman State Technical University, 5, 2-nd Baumanskaya Str.,
107005, Moscow, Russia
²Iceblick, Ltd., 29, Pastera Str., 65026, Odessa, Ukraine

ABSTRACT

We have shown throttle cycles for neon liquefaction with preliminary cooling at the temperature level T = 66…78 K in comparison. The classic Linde’s cycle and the variants of installations with the intermediate pressure working agent have been considered. The influence of the pressures on liquefaction coefficient, specific nitrogen flow and power consumption have been researched. The usage of diaphragm compressors used in rare gases extraction technologies has been justified.

INTRODUCTION

During the last stages of neon-helium mixture separation and extraction of Ne isotopes by rectification the temperatures of about T = 30K are used. Under these conditions it is preferable to use neon as a working body as an effective and safe refrigerant. Besides the task of cryogenic support of separation, the liquefaction of neon reduces the freight and warehouse costs. Liquid neon can be extracted by the direct recovery of the product from the cube of the neon-helium rectification column [1] or with the help of a separate installation [2]. The second alternative is more preferable since it allows efficiently producing a given volume of liquid Ne without hindering the operation of the installation for the neon-helium mixture separation.

THROTTLE LIQUEFIERS ON THE BASIS OF TWO PRESSURE CYCLE

In comparison with the classic Linde’s cycle, the throttle cycle with the working agent intermediate flow is less power-consuming. Consider two types of such installations (Figure 1, 2), different only in the way of the compressors connection. In scheme (I) the compressor of the type 1,6DC-10/12,5 is used as the machine of medium pressure (Table 1). For providing high pressure of the working medium one-stage boosters 1,6DC-16/12,5-200 or 4,0DC-60/12,5-200 with increased suction pressure (Р1 = 1,2 MPa) are suitable. The last unit has higher capacity and along with the standard high-pressure compressor 4,0DC-20/200 is used as C2 in the installation (II).

Compressor type	Capacity, nm³/h (g/s)	Overpressure, MPa		Engine power, kW
Compressor type	Capacity, nm³/h (g/s)	Initial	Final	Engine power, kW
1,6DC-8/200	9,8 (2,3)	0,02	20	5,4
1,6DC-10/12,5	11 (2,5)	0,02	1,25	2,0
1,6DC-16/12,5-200	21 (4,8)	1,25	20	6,7
4,0DC-20/200	20 (4,6)	0,02	22	11,4
4,0DC-60/12,5-200	70 (16,1)	1,25	20	15,0

Table 1: Diaphragm compressors produced by «Ural Compressor Plant» OJSC

I

II

Figure 1: Two pressure cycles with series (I) and parallel (II) connection of the compressors

Figure 2: T-s diagram of two-pressure cycle

With respect to the G consumption on the output of the second compressor the liquefaction coefficient is . In this correlation m is the content of liquid in the condensate collector СC₁, and z is the liquefaction coefficient with respect to the last throttle stage HE₃-Th₂

(1)

where: q₄ is the specific heat leakage to the throttle stage, kJ/kg, i₀, i₈, i₁₂ – enthalpy in the corresponding cycle points, kJ/kg.

The intermediate flow consumption is found from the circuit balance HE₂-Th₁-СC₁

. (2)

The specific nitrogen consumption, fed into the bath NB, is equal to

, (3)

where: G and G_N are the working medium expense on the outcome of the second compressor and the outer refrigerating medium (nitrogen), correspondingly, kg/s; q₁, q₂, q₃ – heat leakages to the corresponding cycle stages.

Specific power expenses of the installation with the intermediate pressure Р_IN according to the scheme (Figure 1, I) is

, (4)

where: r_L= 1206 kg/m³ – liquid neon density; T_A – temperature of ambient air; R = 411 J/(kg×К) – gas constant of neon; P₂, P₁ – pressures of the direct and reverse flow, MPa, correspondingly; h_C– isothermal coefficient of efficiency of the compressor; l_N = 4,3 MJ/kg – specific energy expenses for obtaining liquid nitrogen. The value l_VAC includes energy consumption for the vacuum pump driving gear during nitrogen boiling in the bath NB at reduced pressure (Т_N_В < 77,4 K). Depending on the type of the evacuation system and the temperature level l_VAC can reach 25% of the compressor capacity.

In case of parallel connection (Figure 1, II) specific energy expenses are

. (5)

The calculations have been made for the mode: P₁ = 0,15 MPa; P₂ = 20 MPa; T_N2 = 66 К. Analysis shows, that other things being equal, the intermediate flow pressure increase Р_IN leads to the increase of the content of the liquid m in the collector СC₁. At m®1 the intermediate flow expense is (1 - m)®0 and the cycle degenerates into Linde’s throttle cycle.

The information shown in the graphs (Figure 3) proves the obvious energy advantages of the two pressure cycle. At lower expenses for compression, the liquefaction coefficient of the cycle under study is in the same range as the Linde’s cycle. The extent of energy consumption depends on the compressors connection scheme. At the series connection specific energy consumption is on average 10…13% lower than for the Linde’s cycle. Parallel connection of the compressors reduces energy consumption 15…20% more.

Diaphragm compressors of «Ural Compressor Plant» OJSC (Yekaterinburg, Russia) are produced on two bases: 1,6DC и 4,0DC (Table 1). The outward appearance of several units is shown on Figure 4. Table 2 contains design parameters of real liquefiers with heat leakages and under recuperation taken into consideration. The basic criterion of the choice of the design is specific expenses per 1 dm³of liquid neon. This coefficient included the given equipment and operational costs.

Figure 3: a, b: two pressures cycle (for the temperature of neon

after the nitrogen bath Т₆ = 66К)

1,6DC-8/200

4,0DC-20/200

4,0DC-60/12,5-200

Figure 4: Diaphragm compressors of low (a), medium (b) and high (c) productivity

Table 2: The description of two pressures throttle liquefiers on basis of diaphragm compressors (Р = 20 MPa; P_IN = 1,25 MPa)

Compressor type		Productivity, l/h	Power inputs, kW×h/l l.Ne	Liquid N₂consumption, kg/h
At the series compressors connection (Figure 1, I)
1,6DC-10/12,5	1,6DC-16/12,5-200	3,8	3,8	5,0
4,0DC-20/200 (2)	4,0DC-60/12,5-200 + + 1,6DC-16/12,5-200	16,4	3,3	22,0
At the parallel compressors connection (Figure 1, II)
1,6DCК-8/200	1,6DC-16/12,5-200	3,2	3,9	4,3
4,0DC-20/200 (2)	4,0DC-60/12,5-200	12,6	3,8	16,8

High pressure throttle cycles with intermediate cooling on liquid nitrogen level are used for light inert gases separation [1] and getting neon isotopes [2]. The first installation (Figure 5, a) needs an additional neon refrigeration cycle with nitrogen vapour pumping (Т » 66К), as the throttle effect of the neon-helium mixture under separation is not enough to compensate the heat leakages to the cryogenic unit. In a relatively compact device for neon separation into isotopes ²⁰Ne and ²²Ne (Figure 5, b) a simpler alternative of the cycle is used (without the nitrogen vapour pumping). High pressures Р₂ = 16…18 MPa are practiced into start-up period lasting 16…20 h. In the steady-state regime lasting from 5 to 20 days, the compression pressure decreases to 10…12 МPа, and the required refrigerating capacity of the cycle does not exceed 30 W.

The considered cycles do not include all the alternatives of the installations we have researched and introduced. Particularly promising are the medium pressure designs using outer low-temperature refrigeration systems. As such it is possible to use two-stage gas cryogenic machines, CGM-100/20, helium refrigerators of the type KGS-600 and also the stages using the cold flow exergy in liquid helium evaporation devices.

Figure 5: Throttle neon cycles: a – in the unit of neon-helium mixture rectification separation; б – in the installation for getting neon isotopes (refrigerated cube compressor is used for pumping productional isotope ²⁰Ne). DC - diaphragm compressor; VP - vacuum pump; NВ - nitrogen bath; Т₁-Т₃ - heat exchanger; PS - helium phase separation ; RC - rectification column; Cn and E - condenser and neon cycle evaporator; Rs - working agent receiver (²⁰Ne)

CONCLUSION

For cryostatting the objects within the temperature range of Т = 28…40К it is preferable to use neon as the working agent as an effective and safe refrigerant. It is rational to create low-expense liquefiers on the basis of traditional high pressure throttle cycle with intermediate cooling. For the units with the productivity more that 15 l/h the transition to more economical schemes with using two pressures cycles is more preferable.

Using diaphragm compressors in neon cycles simplifies the system of working agent preparation. Such a solution ensures high quality of the products in saleable neon liquefaction units and isotopes separation systems ²⁰Ne и ²²Ne.

REFERENCES

1. Bondarenko V. L., Arkharov A. M., Golubev A. A. et al. Pilot-Commercial Plant for High Purity Neon Production. Preprints of the XX International Congress of Refrigeration, Sydney, Australia (1999) 1-4.

2. Arkharov A. M., Arkharov I. A. Bondarenko V. L. et al. Production of neon isotopes by rectification method at 28K Proc. 9 Int. Conf. «Cryogenics 2006», Praha (2006) 247-250.

CR08-35

GAS-CHROMATOGRAPHIC ANALYSIS OF MIXTURES OF HYDROGEN ISOTOPES USING DIFFERENT PARAMETERS

Preda A., Bornea A., Pearsica C., Vasut F.

Institute of Isotopic and Cryogenic Technologies – ICIT Rm.Valcea,Romania

ABSTRACT

Gas-chromatography is considered to be the most appropriate of many analytical techniques used to determine the composition or purity of gases.

The method for separating the isotopic species of hydrogen on a moderately large scale using low temperature is gas chromatography.

In this paper, for the analysis of gas mixtures containing hydrogen isotopes, we presented the analysis of mixtures of hydrogen isotopes using different parameters as: different temperature of the column oven and different sample loops. It is realized a comparative study to develop or improve existing methods for the qualitative and quantitative determination of the composition of gas mixtures of hydrogen isotopes.

As results, there are presented chromatograms for different H2, HD, D2 mixtures and different operated parameters.

INTRODUCTION

Chromatography is now an extremely versatile technique; it can be separate gases, and volatile substances by Gas-chromatography.

By classical definition, chromatography is a separation process that is achieved by distributing the substances to be separated between a moving phase and a stationary phase. Those substances distributed preferentially in the moving phase pass through the chromatographic system faster than those that are distributed preferentially in the stationary phase. As a consequence, the substances are eluted from the column in inverse order of their distribution coefficients with respect to the stationary phase [1].

Gas chromatography separation of hydrogen isotopes have been reported in the literature dating from the late 1950’s. Basically, three approaches have been employed to effect separations on an analytical scale, and these approaches may be distinguished on the basis of the column packing material used.

Three different column packing materials have been reported for the gas chromatographic separation of hydrogen isotopes. These are molecular sieves used for size exclusion chromatography, alumina used for gas/solid adsorption chromatography, and palladium such as palladium dispersed on alumina used for catalytic adsorption chromatography.

The species isotopic of hydrogen are: H₂, D₂, T₂, HD, HT, DT, ortho-H2, para-H2, ortho-D2, or para-D2, where D stands for ²H and T for ³H [2].

One of the objective of our laboratory is the enhancement and/or development the gas-chromatographic method for the separation of hydrogen isotope mixtures.

This paper discusses the specific design of the gas chromatograph using at our analysis. Details of flow inside the column, about the column, the utilized detector and chromatograms measured for a specific gas mixture at different parameters are presented.

EXPERIMENTAL

The gas chromatograph employed in this work was a type 3800 from Varian Analytical Instrument.

The Varian 3800 gas-chromatograph is equipped with a capillary molecular sieve 5A column with following characteristics:

- The length of the GC column is: 50 m;

- The inside diameter of the GC column is: 0,32 mm;

- The film thickness of the GC column is: 30 mm.

As detector, we used a Pulsed Discharge Helium Ionization Detector (PDHID).

The temperature of the filament of detector was fixed at 200⁰C.

The carrier gas used was: Helium (99,999% purity). It is recommended that a quality grade of helium 5.0 (99,999 % pure or better) be used at all times.

The sample loops used for these analysis was: 5 mL and 10 mL.

The operating temperature of the GC column for these experiments was -99 ^oC and -75^oC.

The temperature of the oven of the GC column was maintained in the range of 0° C to -99° C, by spraying liquid nitrogen into the oven. A temperature controller to control the liquid nitrogen flow and the heater was used [3].

RESULTS AND DISCUSSIONS

In this paper, for the analysis of gas mixtures containing hydrogen isotopes, we present the analysis of two sample of mixtures of hydrogen isotopes with different concentration, using different parameters as: different temperature of the oven column at -75⁰C and -99⁰C and different sample loops : 5mL and 10 mL.

The first sample has 10% concentration of deuterium in the hydrogen isotopes mixture and the second sample has 50% concentration of deuterium in the hydrogen isotopes mixture.

We present a comparative study between gas chromatograms obtained at these parameters.

The gas chromatograph column was conditioned before to use this for the separate of the hydrogen isotopes mixtures.

The method used was calibrated with standard gas of protium and deuterium by external standard calibration type.

In the Figure 1 and Figure 2 we analyzed the sample number one.

The method described in this was based on using a capillary molecular sieve 5A column which has been operated at -75⁰C.

Figure 1: Chromatogram of sample 1 at -75⁰C temperature of GC column and 5 mL sample loop

In Figure 1, the sample loop used for this analysis was 5mL and in the Figure 2, the sample loop used was 10 mL.

Figure 2: Chromatogram of sample 1 at -75⁰C temperature of GC column and 10 mL sample loop

In the both cases, the carrier flow rate was 3,0 mL/minute, the linear velocity was 30,2 cm/second and the pressure was 10 psi.

The retention times were relatively short, about 8-9 minutes, and the result is a bad separation of H₂ and D₂. In the Figure 1 when we used 5mL sample loop can observe a separation of mixture of hydrogen and deuterium but it is bad and in the Figure 2 we used 10mL sample loop the separation between hydrogen isotopes don’t exist.

Figure 3: Chromatogram of sample 1 at -99⁰C temperature of GC column and 5 mL sample loop

In the Figure 3 and Figure 4, we analyzed the same sample as in first two chromatograms, sample number one.

The method described in this was based on using a capillary molecular sieve 5A column which has been operated at -99⁰C.

In Figure 3, the sample loop used for this analysis was 5mL sample loop and in the Figure 4, the sample loop used was 10 mL sample loop.

Figure 4: Chromatogram of sample 1 at -99⁰C temperature of GC column and 10 mL sample loop

In the both cases, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8 cm/second and the pressure was 10 psi.

The retention times were about 8-9 minutes. In the both chromatograms we can observe a good separation of hydrogen isotopes mixture but better in Figure 3 when we used 5mL sample loop than in the Figure 4 when we used 10 mL sample loop.

For the next four chromatograms we used for analyzing the sample number two with a 50% deuterium concentration in the hydrogen isotopes mixture.

Figure 5: Chromatogram of sample 2 at -75⁰C temperature of GC column and 5 mL sample loop

In Figure 5, the sample loop used for this analysis was 5mL sample loop and in the Figure 6, the sample loop used was 10 mL sample loop. The temperature of column oven was -75⁰C in the both cases. Also, other parameters of method had the next values: the carrier flow rate was 3,0 mL/minute, the linear velocity was 30,2 cm/second and the pressure was 10 psi.

Figure 6: Chromatogram of sample 2 at -75⁰C temperature of GC column and 10 mL sample loop

In these chromatograms, we can observe a bad separation of hydrogen isotopes at this operating of column temperature.

In the next two figures we analyzed the sample number two and the method described in this was based on using a capillary molecular sieve 5A column which has been operated at -99⁰C.

Figure 7: Chromatogram of sample 2 at -99⁰C temperature of GC column and 5 mL sample loop

In Figure 7, the sample loop used for this analysis was 5mL sample loop and in the Figure 8, the sample loop used was 10 mL sample loop.

In the both cases, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8 cm/second and the pressure was 10 psi and the retention times were about 8-9 minutes.

In these chromatograms, we can observe a very good separation of the isotopic species of the hydrogen: para-hydrogen, ortho-hydrogen, HD and deuterium.

Molecular hydrogen occurs in two isomeric forms, namely with its two proton spins aligned either parallel (orthohydrogen) or antiparallel (parahydrogen). In the state of thermal equilibrium at room temperature dihydrogen contains 25 % of parahydrogen (nuclear singlet state) and 75 % of orthohydrogen (nuclear triplet state).

Orth- and para-hydrogen may easily be separated on column of molecular sieve. In all cases, the para-form appears before the ortho-form. Ortho- and para-deuterium are not so easily separated. Normal deuterium contains 66% ortho-deuterium and 33% para-deuterium.

Figure 8: Chromatogram of sample 2 at -99⁰C temperature of GC column and 10 mL sample loop

Chromatograms from Figure 7 and 8 are presented in Figure 9 to can see an evident difference between two sample loops: 5 mL sample loop is for the small chromatogram and 10 mL sample loop is for the high chromatogram.

Figure 9: Chromatogram of sample 2 at -99⁰C temperature of GC column and 5 and 10 mL sample loops

CONCLUSIONS

In this paper we could observe an unsatisfying separation of isotopic species of hydrogen at -75⁰C temperature of column and a good separation of isotopic species of hydrogen at -99⁰C temperature of column.

Also, we could observe a good separation of the two isomeric forms of hydrogen, orthohydrogen and parahydrogen.

At – 75⁰C temperature of column, the carrier flow rate was 3,0 mL/minute, the linear velocity was 30,2 cm/second and the pressure was 10 psi, and at -99⁰C temperature of column, the carrier flow rate was 3,7 mL/minute, the linear velocity was 32,8 cm/second and the pressure was 10 psi.

Between two sample loops used, a good separation was obtained using 5 mL sample loop.

Conclusively, the good method is the method based on using a capillary molecular sieve 5A column which has operated at -99⁰C and using 5 mL sample loop.

REFERENCES

SCOTT, R.P.W., Techniques and Practice of Chromatography, Marcel Dekker, INC, New York, USA (1995)10-15
LITTLEWOOD, A.B., Gas Chromatography, Second ed., Academic Press: New York and London (1970) 427-430
LASSE, R., GLUGLA, M., GUNTHER, K., PENZHOM, R.D., WENDEL, J., Fusion Science and Technology (2002), 41-43.

CR08-19

mathematical models concerning the design of column for isotopic exchange process in the Pilot Plant for Tritium and Deuterium Separation

Gherghinescu S.¹, Popescu G.¹

National Institute of R&D for Cryogenics and Isotopes Technologies (ICIT), ROMANIA,

Abstract

The present work has the purpose to determine the flow behavior of both phases, gaseous and liquid, of the hydrogen isotopes in order to obtain a better separation factor between hydrogen and water, a^D/T,in the D₂-DTO large scale isotopic exchange column.

Seeing that direct determination of the fractions of the hydrogen isotopic species is difficult and rather imprecise, the numerical estimation method with high precision is strongly required.

Therefore we shall use series of isotope exchange process characteristic equations to obtain the precise results needed in the process design for the achievement of the operating data target.

A special attention will be paid to the two first weight reactions in the isotopic exchange process:

Key-Words: separation factor, hydrogen isotopic species, isotopic exchange column.

Introduction

In the D₂-DTO isotopic exchange column, the tritium transfer occurs through three phases of hydrogen gas, water vapor and liquid water. Along the column the concentration of tritium will decrease from the column top to the bottom. Therefore, for the optimum plant design, it is necessary to determinate the D/T-isotopic separation factor and the dependence with the tritium concentration.

The object of this study is to establish a simple calculation method, in which the changes of overall D/T -isotopic separation factor on the operation temperature and the concentration of tritium in liquid water could be evaluated accurately.

A special attention will have the two first weight reactions in the isotopic exchange process between different phases (gas/vapor and vapor/liquid):

DT(g)+D₂O(v)ÛD₂(g)+DTO(v)

DTO(v)+D₂O(l)Û D₂O(v) +DTO(l)

Fig. 1

For these exchange reactions the separation factor a_g and a_l are defined as follows:

for the tritium transfer between hydrogen gas and water vapor

and for the tritium transfer between water vapor and water liquid, where x, y and z shows the atom fractions of tritium in liquid water, hydrogen gas and water vapor. In the end, the totally separation factor a^D/Tis determined as following:

a^D/T= a_ga_l

In the D₂-DTO isotopic exchange process, the transfer of tritium among hydrogen gas, water vapor and liquid water is described by following exchange reactions:

D₂(g)+T₂(g) Û2DT(g)

K1 (R1)

DT(g)+D₂O(v)ÛD₂(g)+DTO(v)

` K2 (R2)

D₂O(v)+T₂O(v)Û2DTO(v)

K3 (R3)

DTO(v)+D₂O(l)ÛDTO(l)+D₂O(v)

K4 (R4)

D₂O(l)+ T₂O(l)Û2DTO(l)

K5 (R5)

where K₁, K₂, K₃, K₄ and K₅ are the correspondent equilibrium constants, temperature-responsive, for the isotope exchange reactions, (R1) to (R5), respectively.

x) The expression of the atom fraction of tritium in liquid water,x is

where l_D2O, l_DTO and l_T2O represent the mole fractions of D₂O, DTO and T₂O in liquid water. Hence, l_D2O+l_DTO+l_T2O=1 and

where R_l is the molar fraction ratio of T₂O to DTO

therefore

And K₅ became

By solving the above second order equation, R_l can be expressed as function of x and K₅

and the atom fraction of tritium in liquid water x is given as

Then, the abundance ratio of tritium in liquid water is derived as

y) The expression of the atom fraction of tritium in hydrogen gas y is

Where h_T2, h_DT and h_D2represent the molar fractions of T₂, DT and D₂ in hydrogen gas, and R_g is the molar fraction ratio of T₂ to DT. Using the same calculation pattern we obtain the equilibrium constant for the gaseous reaction K₁ as

and the atomic fraction of tritium in hydrogen gas y as

Then, the abundance ratio of tritium in hydrogen gas is derived as

z) The expression of the atom fraction of tritium in water vapor z is

Where v_DTO represent the molar fraction of DTO in water vapor and R_v is the molar fraction of T₂O to DTO. The molar fractions of DTO(v_DTO), T₂O(v_T2O) and D₂O(v_D2O) will be calculated as functions of z and R_v and the equilibrium constant of the vapor phase reaction K₃ became

and from here z is expressed by

where the molar ratio R_v is represented as

By substituting the equilibrium constants K₄ and then K₅ we get the relation for z

Then, the abundance ratio of tritium in water vapor is derived as

Conclusions

Now that we have the description of x, y and z, which shows the atom fractions of tritium in liquid water, hydrogen gas and water vapor, the equilibrium constants K₁ to K₅ and the molar ratios R_g, R_v and R_l we can evaluate the separation factors a_g, a_l and a^D/T

Finally, the overall separation factor will be

where

With the temperature-responsive equilibrium constants K1, K2, K3, K4 and K5 and knowing the atom fraction of tritium in liquid water, x, it will be possible to determinate the theoretical shape of the overall separation factor according with the operation temperature and the concentration of tritium in liquid water. In the real operating conditions it will be necessary to determinate the efficiency of the catalyst which will determinate the real shape of the overall separation factor . By direct comparation with the theoretical shape of the overall separation factor between different operation temperatures and concentrations of tritium in liquid water it will be possible to determinate the optimal operating conditions for the isotopic exchange column.

Bibliography

(1) J.H.Rolston, J.dem Hartog and J.P.Butler: “The Deuterium Isotope Separation Factor between Hydrogen and Liquid Water”, J.Phys. Chem., 80(10), 1064-1067 (1976)

(2) Masami Shimizu, Kenji Takeshita "Simplified calculation method of deuterium separation factor between hydrogen and water (a^H/D) depending on D-atom fraction of liquid water" FIFTH CONFERENCE ON ISOTOPIC AND MOLECULAR PROCESSES PIM – 2007

(3) J.H.Rolston and K.L.Gale: “Deuterium-protium Isotopic Fractionation between Liquid Water and Gaseous Hydrogen”, J.Phys.Chem., 86(13), 2494-2498 (1982)

CR08-15

THE CREATION OF VEHICLES FOR MULTIMODAL TRANSPORTATION OF LIQUEFIED GASES

Zashlyapin R.A., Cheremnych O.Ya.

JSC “UralCryoMash”, Russian Federation, Sverdlovsk region, Nizhny Tagil

ABSTRACT

In this report there given information about the enterprise activities concerning the creation of effective means of transportation, tank-containers for multimodal transportation of LPG, (LH₂). Application analysis of different types of TC heat isolation is also given. For safety transportation of LNG and LH2 special drainage and locking-safety devices are used. Delivery experience of the first LNG consignment (using tank-container) from Russian Federation to European country is analyzed.

INTRODUCTION

JSC “UralCryoMash” is established as a development contractor and producer of vehicles for transportation and storage of liquefied low-temperature gases (nitrogen, oxygen, argon, ethylene, propane-butane mixtures, carbon dioxide, hydrogen, LNG) for different industrial branches [1, 2]. Enterprise has a great experience (gained in 1967-1990) in creating and safety operation of railway and motor vehicles for large-scale transportations of liquid nitrogen, oxygen, argon (8Г513, 15-558 С), liquid hydrogen (LHT-100, LHT-100 M), carbon dioxide (15-559), ethylene (15-147). This experience is used today at creation of new generation vehicles - tank-containers for multimodal liquefied gases transportation for the sake of space, energetic, oil-gas industrial branches.

Nowadays multimodal transportation (mixed) of liquefied gases (methane, hydrogen, oxygen, nitrogen, argon, ethylene, etc.) becomes more and more effective. Such transportation has got the following advantages:

- the ability of tank-container transportation by railway, motorway and seaway

- direct delivery from manufacture to consumer

- no liquid loss, arising from its being pouring from transportation vessels to stationary ones

- no necessity to use expensive terminals for filling/discharge

- high reliability and safety

- comparatively cheap transportation

When creating tank-containers the producer has the following objects:

- It’s a constructive solution that allows tank-container operating on present “filling” trestle bridges at liquefied gases filling works and operation on “discharge” trestle bridges at consumers’ site, particularly at European’s;

- Nomenclature widening of transported liquefied gas in the given type of tank-container or its modifications;

- Increasing of the transported mass of liquefied gas due to creation of 40 or 45 foot containers in accordance to international standards, which determine basic design and operating regulations at different ways of transportation [3,4,5];

- Generality and individuality of constructive solution from the point of view of similar to thermo physical and chemical properties of transported liquefied gases (“propane-butane-propylene”, “nitrogen-oxygen-argon”, “LNG-ethylene”, hydrogen-helium”);

- To ensure the maximum period of nondrainage transportation or control retainer time (time between the first filling condition at the pressure 0,13÷0,15 MPa and pressure raising as a result of heat leakage, i.e. pressure opening of safety valves);

- Observance of requirements for ecological and fire safety regulations in case of emergency (heat isolating cavity decapsulation or vessel MAWP exceeding)

1. TANK-CONTAINERS FOR NONCOOLED LIQUEFIED GASES TRANSPORTATION (LPG)

Liquefied noncooled gases (propane, butane, propylene and others) are widely used in industry and in private life. The enterprise has developed and mastered the production of different types of tank-containers for these gases: type size 1CC-TC-25/2,0, TC-25/2,2, TC-25/1,8; they have both combined discharge (“upper-bottom”) and “upper” or “bottom”; type size 1 AA TC- 52/1,8 with combined joint “filling-discharge” and joint of “gas drainage” (picture 1).

Technical characteristics of tank-containers are given in table 1.

Tank-containers are gathered with own safety locking valves or with valves of “Fort Valve” firm.

Tank-containers filling control is done with the help of level control block (in contrast to railway tanks) which is connected by the cable to transmitter set on a vessel; this system ensures the control of percentage filling of tank-container. When filling tank-containers with noncooled gases IMDG code is followed; By IMDG the maximum degree of TC filling with propane-0,42kg/l, butane – 0,51 kg/l, propylene – 0,43 kg/l.

According to table 1 at empty weight of TC-25/1,8 – 6,4 t and TC-52/1,8 – 11t, maximum allowable mass of transported liquefied gas is 17,6 t and 23 t. At normal filling with propane – 0, 42 kg/l its mass in TC-25/1,8 comes to 10,8 and 21,53. Thus, from the point of view of “steel intensity” 40-foot container is 16÷20% more effective than two 20-foot containers model TC-25/1, 8.

Tank-container model	Designation ISO	Empty weight	Maximum gross weight, t	Working pressure in vessel, MPa	Filling/ discharge method		Drainage method		Total capacity, m³
Tank-container model	Designation ISO	Empty weight	Maximum gross weight, t	Working pressure in vessel, MPa	Upper	Bottom	Upper	Bottom	Total capacity, m³
TC-25/2,0	1CC	9,6	24	2,0	B	-	B	-	25
TC-25/2,8 HC	1CC	96,	24	2,0	B	H	B	-	25
TC-25/2,0 HC-01	1CC	9,6	24	2,0	-	H	B	-	25
TC -25/2,2	1CC	9,6	24	2,2	B	-	B	-	25
TC -25/2,2 HC	1CC	9,6	24	2,2	B	H	B	-	25
TC-25/1,8	1CC	6,4	24	1,8	B	H	B	H	25
TC-52/1,8	1AA	11	34	1,8	B	H	B	H	52

Table 1. Technical characteristics of tank-containers for noncooled liquefied gases.

Tank-containers are certified by Russian Maritime Register and correspond to international standards IMDG;ADR;RID;CSC;CCC;ISO1496; ISO 668:1995; ISO 6346:1995; Russian national standards: GOST 14249-89; GOST25290-82; GOST 26291-94.

2. TANK-CONTAINER FOR LNG TRANSPORTATION

Creating tank-container TCM -35/0,6 constructors took into account requirements to its overall mass characteristics, determined by ISO 6346;1995. Regulations and instructions for big-volume heavy loads transportation by motor way in Russian federation, CIS, Baltic and European countries were also considered.

The analysis of allowable overall dimensions, full mass and axial loading of carriers in RF, CIS, Baltic and European countries at 40 foot TC transportation showed that allowable carrier mass including hauler with semi-trailer and 40 foot TC ISO can variate from 38 t (RF, Belorussia, Kazakhstan) to 42:44t (Ukraine, Baltic, Finland, Sweden, Spain and others).

If proceeding from normative requirements and taking into account that the mass of a five tractive unit is 11-12t and allowable gross mass up to ISO standarts -30,48 t, so TC mass for LNG for its exporting to European countries from RF mustn’t exceed 30,48 t.

These directions became the basis for creation of TCM 30/0,6 with capacity 35 m3, tare mass 14,95 and maximum mass of transported product 15,53 t (picture 2).

Technical characteristics of TC are given in table 2.

Parameters	Values
Type size ISO 1496÷1995	1AA
Maximum gross weight, t	30,48
Tare mass, t	14,95
Total capacity, m³	35,36
Maximum allowable working pressure in vessel, MPa	0,6
Loss of evaporation per day, l/day	0,39
Filling degree, kg/l not more	0,355
Period of nondrainage storage at pressure, in days, not less	35
Type of heat isolation	Fiber-vacuum
Reservoir material: - - vessel - - casing	12x18H10T 09Г2С-14
Discharge method	Reservoir blowing from the side source

Table 2. Technical characteristics of TC for liquefied natural gas (LNG) model TCM – 35/0,6

The enterprise has developed and uses different types of isolation for vehicles transported liquefied cryogenic gases: powder-vacuum (using of basalt fiber as heatisolating mats); multilayer screen-vacuum (using metalized film and glasscloth). For LNG tank-container both types of isolation can be used.

Powder-vacuum isolation on basis of pearlite has got some disadvantages such as: unavoidable pearlite shrinkage in cavity at transportation loadings and considerable vacuuming time of cavity filled with pearlite.

Screen-vacuum heatisolation is more effective than powder-vacuum one, but its cost is considerably higher than the latter.

Time of nondrainage transportation or the control time of product retaining is an important characteristic of cryogenic tank-container and comes out from the heatisolation effectiveness and maximum allowable value of working pressure in the vessel of TC.

An important moment must be taken into consideration when transporting LNG – requirements keeping for ecological and fire safety regulations.

Tank-container model TCM-35/0,6 is equipped with safety drainage device, that ensures safe LNG vapor dumping from the vessel due to gasdynamic destabilization of drainaged to the atmosphere gas burning at reaching the maximum allowable pressure in the process of transportation [6].

In case of an emergency at LNG transportation and vacuum loss in heatisolating cavity, safety devices protect the vessel from destruction and provide safety vapor dumping of LNG to the atmosphere.

For safety LNG discharge to the reservoir tank-container model TCM – 35/0,6 is equipped with the following safety devices:

- Safety-locking device intended for discharge pipeline by means of its automatic cutoff at a casual movement of tank-container during discharge.

- Safety-locking device together with fireproof device ensuring the discharge pipeline cutoff in case of fire

- High-speed valve, used for discharge pipeline cutoff in case of decapsulation of discharge system at consumer’s site.

For some technological questions working out, from the liquefied gas manufacturer plant (AGNKS, St. Petersburg, RF) to European consumer (Lingcheping, Sweden) LMG transportation in TC model TCM-35/0,6 on automobile container carrier (hauler with semi-trailer) was done: St.Petersburg – Helsinki (Finland) – sea ship – Stockholm – Lingcheping (Sweden); total way of transportation – 1100 km, way time – 48 hours, average speed 60-70 km/per hour.

At arriving LNG discharge from TC to stationary vessel-reservoir was done (capacity 53m, working pressure 1,5 MPa, picture 4).

For LNG exporting tank-containers model – 35/0,6 (developed by “UralCryoMash”) can be used.

3. TANK-CONTAINERS FOR LIQUED HYDROGEN TRANSPORTATION.

At present time liquid hydrogen is widely used in rocket- space techniques, aviation, energetic [2]. For LH transportation to consumer, railway transportation tank-containers model RHT – 100M, road tank model 17Г228 are created [8].

Parameters	Values
Type size ISO 1496-3÷1995	1BB
Gross weight, t	12,47
Tare mass, t	11,2
Mass of transported hydrogen, t	1,27
Total capacity, m³	20,0
Maximum allowable working pressure in vessel, MPa	1,2
Loss of evaporation per day, l/day	56
Period of non drainage storage at pressure increasing from 0,11MPa to 1,2 MPa, in days	53,6
Type of heat isolation	Vacuum-screen
Reservoir material: -Vessel - - Casing	12x18H10T 09Г2С-14
Discharge method	Reservoir blowing from the side source
Frame contacting with LH₂ and GH₂

Table 3. Technical characteristics of liquid hydrogen transportation tank

General view of LHT can be seen in picture 5, technical characteristic is given in table 3.

REFERENCES

1. Zashlyapin R.A., Cheremnych O.Ya., The creation of vehicles and stationeries for liquefied gases storage and transportation, Technical gases - 1, Odessa, Ukraine (2006) 20-126

2. Zashlyapin R.A., Cheremnych O.Ya.,Pavlenko S.T., The creation of vehicles and stationeries for lunar orbital complex filling with liquid hydrogen, Technical gases - 4, Odessa, Ukraine (2007) 15-19

3. IMDG CODE (International Maritime Dangerous Goods Code), St.Petersburg,ZNIIMF (2007) 512

4. IMDG CODE (International Maritime Dangerous Goods Code) (2004)

5. RID (International regulations for dangerous goods transportation by railway) (2005)

6. ADR (European agreement on international road transportation of dangerous goods) (2006)

7. Zashlyapin R.A., Cheremnych O.Ya., Organization and development of effective production means for LNG multimodal and railway transportation, Technical gases - 3, Odessa, Ukraine (2006), 32-36

8. Cheremnych O.Ya., Analysis of transportation peculiarities for LNG export in tank-containers and technologies of its discharge to reservoir, Technical gases – 6 Odessa, Ukraine (2007) 65-68

Picture 1. Tank-container model TC-25/2,0

Picture 2. Tank-container model TC-52/1,8

Picture 3. Road tank carrier with TCM – 35/0,6, at cryogenic complex LNG, Lingcheping, Sweden

Picture 4. LNG discharge from TCM -35/0,6 vessel to stationary reservoir.

Picture 5. General view of tank-container for liquid hydrogen.

CR08-28

THE INCREASE OF EFFICIENCY AND SAFETY OF LIQUID HYDROGEN TRANSPORTATION.

Cheremnych O.Ya.

JSC “UralCryoMash”, Russian Federation, Sverdlovsk region, Nizhny Tagil

ABSTRACT

Efficiency and safety criteria of vehicles for liquid hydrogen transportation are analyzed: rate of evaporation of cryogenic product, emergency situation in the process of transportation – decapsulation of heatisolating cavity of a tankage with liquid hydrogen.

INTRODUCTION

Liquid hydrogen has been widely used in different brunches of industry: aircraft-space techniques, aviation, power engineering [1, 2]. Liquid hydrogen delivery from cryogenic manufacture to consumer is carried by tank wagons, truck tanks and tank-containers for multimodal transportations [3].

More demands for liquid hydrogen transportation are made by “ Regulations for dangerous goods transportation” . As there exist a constant risk of explosion at hydrogen vapor dumping from the vessel, and risk of emergency situation at tanker breakdown at humps of a consist.

That’s why efficiency and safety criteria of liquid hydrogen transportation are considered by the example of a tank wagon model HTC-100M.

1. TANK WAGON FOR LIQUID HYDROGEN TRANSPORTATION.

Overview and schematic circuit of a tank wagon for LH transportation are given in picture 1; technical characteristic is given in table 1.

Parameters	Values
Geometrical volume, m	119
Hydrogen mass in the vessel, kg	7350
Working pressure in the vessel at transportation, MPa	0,25
Empty weight, t	77
Overall dimensions GOST 9238-83	1T
Length at coupler pulling faces, mm	25730
Axle loading, t	21,2
Heatisolation	Screen-powder-vacuum
Loss of liquid at evaporation, % per day	Not more than 0,8
Liquid hydrogen discharge	Side boosting
Time of nondrainage transportation, days	12

Table 1. Technical characteristics of tank wagon model HTC-100M

The peculiarity of tank container construction is in its fixing devices which must bear striking dynamic loads up to 12 g; this is achieved by fixing the vessel on sectors supports.

The main criteria of effectiveness of LHT is the rate of hydrogen (loss) evaporation in the process of transportation. It is this criteria that determines: transportation time without hydrogen gas release from the vessel and cryogenic product quality assurance.

In order to decrease the rate of evaporation of liquid hydrogen and also the components contamination by admixtures (nitrogen, oxygen) characteristics of different types of isolation were analyzed (layer-vacuum, powder-vacuum, layer-powder-vacuum); experiments were carried out on samples of the tank model LHT-100M.

In layer-vacuum isolation metallized film and glass fiber were used as gasket material. The quantity of screens in heatisolating place was 30, 20, 10; the number of layers of layer-vacuum isolation – 60, 90, 120.

Aerogel was used as heatisolating powder ( in comparison with perlite it doesn’t shrink). The quantity of aerogel was changing from 5,5 t.- 100% filling, 2,75 t – 50 % filling and 1,5 t – aerogel filling only strength elements parts (bearings, shafts, chains)

The results of analysis are given in table 2.

Type of heatisolation	Vacuum value in the vessel (mm merc. column)	Day loss, % per day
Powder-vacuum	5·10^-3	1,5
Layer-vacuum (90 screens)	5·10^-4	1,4
Layer-powder-vacuum, at aerogel mass: - 5,5 t (100%) - 2,75 t (50%) - 1,5 t (shaft and strength elements level)	5·10^-3 1·10^-3 8·10^-5	1,2 1,0÷1,1

Table 2. Rate of LH evaporation in LHT-100M depending on type of isolation.

Tank container testing in powder-vacuum isolation determined the rate of hydrogen evaporation at level 1,5% per day . It is explained by insufficient width of walls-between clearance , filled with heatisolating powder, increasing of which can be done only by decreasing of vessel diameter and this in its turn leads to transported hydrogen mass decreasing.

The results of testing with layer-vacuum isolation showed the rate of hydrogen evaporation at level 1,4 % per day. That is explained by low effectiveness of layer-vacuum isolation in supporting area and chains in lower part of the vessel. Analysis allowed to determine the optimal number of layers of layer-vacuum isolation (equals to 60), level of aerogel filling by mass 1,5t, vacuum value in cavity 8·10^{-5 mm merc column, that ensured hydrogen rate of
evaporation at the level 0,8% per day.}

RAILWAY TANK TESTING IN EMERGENCY CONDITIONS (LHT-100M)

LHT overview and its airhydraulics scheme are given in photo 1. In tank scheme safety devices are shown.

In operating conditions of HTC-100M real danger of casing crippling could arise (for example, because of its breakdown at railway consist reforming). Liquid hydrogen transportation was carried out by five cistern cars with attendant personnel. Total amount of transported liquid hydrogen – 36,7t. At such mass consequences of vapor loss in heat isolating space of tankage could be unpredictable.

That’s why it was decided to carry out an experiment in the testing area (photo 2). The experiment was carried out on a real tank LHT-100M with screen-vacuum heatisolation - vacuum loss imitation through a vacuum valve set on tankage casing (photo 3).

The main testing object was to determine operational reliability of safety devices in case of vacuum loss in heatisolation cavity because of casing leakage. While testing it was necessary to determine:

- Risk value at emergency situation as a result of decapsulation

- Rate of pressure increasing in the vessel with liquid hydrogen

- Propriety of safety valve throat calculation

For imitation of emergency situation caused by mechanical failure of casing, pneumatic valve 8 was open (photo 1, a). After the valve opening, air from the atmosphere started intensively coming to the heatisolation cavity. The noise from absorbing air could be heard during three hours at a distance of 300m. Moreover some cracks, strong and weak strokes could be heard. This noise was because of separate tank parts destruction. The inferior limit of operating temperature of casing (steel 09Г2С) was – 70C, and the lowest casing temperature during the experiment was 74 K.

The first acute sound was heard after 22 minutes, and the strongest noise could be heard in the period 2h – 2h 34 m after the beginning of testing. Pressure increasing in the vessel caused the disk bursting and safety valve response (photo 4).

Full evaporation of liquid hydrogen evoked by decapsulation of heatisolation cavity occurred during 6h 10 m. In an hour after the experiment beginning the casing surface (in bearing area) started frosting, then moisture from the atmosphere also started frosting on it (picture 5). In 2 hours 30 minutes moisture frosting could be seen along the full length of the casing.

At tank control examination after the experiment completion the following was revealed:

- along the full length of the casing cracks of different patterns and length were formed (0,2 -1,5 m);

- on a joint weld (head sticking to sidewall at the bottom part) there appeared a crack on a length 0,025 m;

- in the middle part of a center sill at the upper part of the H platform there also appeared a crack.

All that arouse because of liquid air accumulation in the bottom part of heatisolating cavity. It was here that cracks appeared. Some condensate flew through the cracks to the center sill, hence its destruction occurred.

During the experiment safety valve ensured the full release of increasing pressure in the vessel. At safety valve opening there was an acute clap with hydrogen vapor jet appearance (length about 8 meters). During the experiment fire risk situation did not occur. After the experiment some recovery works were done, so the tank could go to “UralCryoMash” at its own pace.

CONCLUSION

The results of experiments showed that even at emergency situation safety conditions of tank container LHT-100M can be ensured as well as to prevent negative influence on environment. That allowed to use the results, obtained during the unique experiment, in the process of new generation transport units creating.

REFERENCES.

1. JSC “UralCryoMash” Little land of Vagonka, SV – 96, Ekaterinburg, Russia (2004) 208

2. Shuttle space system “Energy-Buran”, SMF “OVM-LUCH”, Moscow, Russia (2004) 356

3. Zashlyapin R.A., Cheremnych O.Ya., Pavlenko S.T., The increase of efficiency and safety of liquid hydrogen transportation at railway and multimodal transportation, Technical gases - 6, Odessa, Ukraine (2007) 57-60

APPENDIX

Picture 1. Railway container for liquid hydrogen transportation, model LHT-100M: a - general view; b – schematic circuit: 1 – vessel; 2 – casing; 3 – vacuum valve; 4 – membranous safety device; 5 – filling-discharge line; 6 – side boost line; 7 – gas release; 8 – safety valve; 9 – gage board; 10 – control pane; 11- preheater; 12 – vapor drainage line in transit; 13 – end elements blowing by nitrogen; 14 – fire fighting system; 15 – balloons for nitrogen.

Picture 2. Railway container LHT – 100M on experimental stand of testing area.

Picture 3.Valve block for sudden vacuum loss imitation.

Picture 4. Hydrogen vapor dumping to the atmosphere through the safety devices at vacuum loss in heat isolating space of the tankage.

Picture 5. Casing frosting in the process of vacuum loss in heat isolating space of the tankage

CR08-29

THE CREATION OF VAPOR COOLING DEVICES FOR LIQUID OXYGEN IN STATIONARY RESERVOIRS USING LIQUID NITROGEN AS A COOLING REAGENT.

Cheremnych O.Ya., Korneva I.I.

Uralcryomash, Sverdlovsk region, Nizhny Tagil, Russia

ABSTRACT

Cooled liquid oxygen (non boiling) is used as a cryogenic component of fuel in space apparatus. It’s got lower temperature than its boiling temperature at atmospheric pressure, hence its density is high. Thus it allows to increase liquid oxygen reserves in a space apparatus tank. That’s to say that the equal capacity tank is filled with more product mass and product losses at storage and fillings are substantially decreased.

Liquid oxygen cooling in stationary conditions is carried out in both ways: stationary reservoirs and special devices (heatexchanging apparatus, cooling reservoirs).

The results of analysis of liquid oxygen cooling conditions and storage are given. In this case liquid nitrogen is used as a cooling reagent. Nitrogen is cooled by means of vapor space vacuuming on liquid surface with the help of ejectors.

The constructive solution of tank-reservoirs for cooled liquid oxygen is given.

INTRODUCTION

Low density is considered to be a grave disadvantage of some fuel cryogenic components (such as liquid hydrogen, LNG) as it leads to tank capacity increasing. Another disadvantage is their low heat per unit of volume. Moreover there is a necessity in compensation of losses as a result of component evaporation, maintenance of the given product quantity in rocket tanks, exclusion of nonequilibrium processes in surface filling stations.

The characteristics of rocket-carriers, acceleration blocks, and space ships can be improved by tanks filling with cooled (nonboiling) cryogenic products, which have lower temperature than their boiling temperature at atmospheric pressure and hence higher density. Thus it allows to increase fuel components recourses on board the rocket, i.e. the tank with equal capacity is filled with more product mass and its losses at storage and filling are considerably decreased. Moreover the use of cooled product ensures a single-phase liquid flow, some size decreasing, so the mass of pipelines and fittings that is particularly vital for aircrafts. Cooled liquid oxygen is more applied in space techniques.

Cryogenic liquids cooling in a start complex can be carried out both in filling systems reservoirs and in the process of rocket filling, acceleration block and a space ship in special facilities (heat exchanging apparatus, reservoirs-coolers), installed in a filling system [1].

For products cooling in tank-reservoirs cooling devices with relatively little productivity can be created as cooling is carried out in nontechnological time; it can be extended in time and not connected with technological preparation of rocket for start.

When cryogenic liquid is being cooled in the process of filling the productivity of cooling devices is determined by the rate of filling; in this case the equipment is some more complicated.

1.1 Cooling methods

Basic cooling methods.

With the help of cooling machine devices. The process is effective from the point of view of thermodynamics. Any temperature level can be reached. But the use of complicated machinery is its disadvantage.

With lower temperature products assisted. In this case refrigerating fluids are:

1.Cryogenic components, having boiling temperature at atmospheric pressure lower than that of necessary for fuel components cooling.

2.Cryogenic components, whose temperature in the reservoir with heatexchanger is carried to the required one by vacuuming of vapor space with the help of different devices.

The distinctive feature of these cooling methods is the loss of refrigerating fluid, evaporating in the process of fuel cooling. For collecting of refrigerating fluid vapors and their back condensation complicated additional equipment is required; in conditions of a start position it is considered to be inexpediently.

Due to liquid evaporation - by vacuuming of vessels vapor space or by barbotage through the liquid of low soluble noncondensable gas (usually helium). These methods allow reaching the products temperature up to the triple point. Moreover with their help an ice-like condition of some products can be got.

When products are cooled by barbotage practically helium is only used because of its safety, little solubility and low condensation temperature. This method is comparatively expensive. Under its realization at an open-ended scheme large quantity of expensive and deficit helium is required; when realizing at a closed scheme complicated system of helium refinement from cooled liquid vapors is required [2].

Cooling method by vapor space vacuuming above liquid surface is more applied. Cooling by vacuuming is a universal method of cryogenic products cooling; it doesn’t require complicated equipment, big capital and operating expenses.

1.2 The use of nitrogen as refrigerating fluid in the process of oxygen cooling

Having analysed all advantages and disadvantages of different cooling methods in conditions of a start position such method as vapor space vacuuming above liquid surface of a side refrigerating fluid (nitrogen) was chosen.

Figure 1: Scheme of oxygen cooling by vapor space vacuuming above liquid surface of a side refrigerating fluid (nitrogen).

1 –liquid oxygen, 2- liquid nitrogen, 3 – ejectors block

1.3 Cooling operating factors determination

1.Pressure determination at a pumping block absorbing (by nonlinear equation)

Where

Pn1 (T) – pressure at a pumping block absorbing, MPa;

Gn(Pn) - pumping consumption mass depending on pressure, kg/s;

P_N2(T) – nitrogen saturated vapors pressure in the vessel depending on temperature, MPa;

Pn – pressure at a pumping block, MPa;

– density of gaseous nitrogen depending on temperature, kg/m³;

A1 – local resistance equal to pumping line;

A2 - local resistance equal to pumping block;

Figure 2: Diagram of pumping consumption mass depending on temperature

2. Determination of pumping block heat rate

Q(T) = Gn(Pn) .r_N2(T)

Where

Q(T) - pumping block heat rate depending on temperature, J/s;

Gn(Pn1) – pumping consumption mass depending on pressure at pumping block absorbing, kg/s;

r_N2(T) - latent heat of nitrogen vaporization depending on temperature kJ/kg;

Figure 3: Diagram of pumping block heat rate depending on temperature

3. Heat leakage determination from environment

Where

Qinp(T) – heat leakage to a product depending on temperature, kJ;

Qinp - rated heat leakage to a product, K;

4. Product cooling time determination

Solution of differential equations

Where

T- current product temperature, K;

t - current time of product storage, sec;

Q(T) - pumping block heat rate depending on temperature, J/s;

Qinp(T) – heat leakage to a product depending on product temperature, kJ;

Mw – current nitrogen mass, kg;

Gn(T) – mass consumption of nitrogen depending on product temperature, kg/s;

Cp1_N2(T) – liquid nitrogen heat depending on temperature, J/kg*K;

Cp_O2(T) - liquid oxygen heat depending on temperature, J/kg*K;

M_O – oxygen mass;

M_m – metal mass (vessel mass in casing);

Figure 4: Diagram of product temperature depending on cooling time.

5. Plotting of product T temperature dependence on cooling timet.

Figure 5: Diagram of nitrogen mass depending on cooling time.

Conclusion

This type of cooling allows getting oxygen temperature (at a given rating of electors block) equal to 70K. The main advantages of such cooling method are: the ability of preparation of oxygen beforehand at nontechnological time, no filling fuel components losses, simplicity of equipment, efficiency.

REFERENCES

1. Filin N.V., Liquid cryogenic systems, Mechanical engineering, Leningrad, Russia (1985) 247-249

2. Arkharov A.M., Kunis I.D., Filling cryogenic systems of rocket-space complexes, MSTU by Bauman, Moscow, Russia (2006) 252-260

3. Sychyov V.V., Vasserman A.A., Thermodynamic properties of nitrogen, Standards publishing house, Moscow, Russia (1977)

4. Sychyov V.V., Vasserman A.A., Thermodynamic properties of oxygen, Standards publishing house, Moscow, Russia (1977)

Aigouy G.................................... 57

Alava L.A................................. 223

Albert S.................................... 267

Anghel A.................................. 107

Arkharov I................................ 173

Arpentinier P............................ 165

Bae D.K................................... 289

Bae J.H.................................... 289

Baker R.A.................................. 97

Barucci M................................ 273

Bell C....................................... 151

Berdais K.-H.............................. 51

Binneberg A..................... 267, 313

Blau B...................................... 107

Bondarenko V. L.............. 159, 325

Bornea A.................................. 337

Bracanovic D............................ 141

Brojek W................................. 253

Caillaud A.................................. 57

Cheremnych O.Ya.... 351, 359, 367

Chorowski M........................... 115

Chrz V............... 43, 183, 191, 199

Coquelet C............................... 165

Coulomb D................................. 25

Crispel S.................................... 57

Dauguet P................................... 57

Daum M................................... 107

Delcayre F.................................. 57

Delcorso F............................... 165

Diachenko O.V........................ 331

Diachenko Т.V......................... 331

Dobrozemsky R........................ 307

Dvořák J.................................. 241

Esteves A.D.S.......................... 209

Forýtková L............................. 253

Fydrych J................................. 115

Gherghinescu S................. 261, 345

Good J..................................... 141

Grabié V.................................... 57

Grigoriev S............................... 107

Gschwendtner M.A.................... 89

Haberstroh Ch............................ 65

Hannani S.K............................... 75

Hanzelka P............................... 131

Heidrich R................................ 267

Herzog R.................................. 313

Hirschl C.................................. 307

Hnízdil T................................... 183

Horynová A.............................. 233

Houssin-Agbomson D............... 165

Jafarian A................................... 75

Joonhan B................................ 281

Kaiser G................................... 267

Kaiser Z..................................... 43

Kalbassi M.A........................... 151

Kideok S.................................. 281

Kirch K.................................... 107

Klepal J.................................... 219

Klier J.............................. 267, 313

Klingner E................................ 245

Korneva I.I............................... 367

Kouba M........................... 43, 183

Králík T.................................... 131

Kundera R.................................. 43

Laa C....................................... 307

Lain M..................................... 241

Lánský M................................. 191

Lee D.-Y.................................. 289

Lee S.-J................................... 289

Losyakov N. P......................... 159

Martelli V................................. 273

Mátl P...................................... 191

Měřička P........................ 233, 241

Muehlegger M............................ 51

Musilová V............................... 131

Myunghwan. S.......................... 281

Navasardyan E......................... 173

Parise J.A.R............................. 209

Park J.-S.................................. 289

Pearsica C........................ 319, 337

Popescu G................................ 345

Preda A............................ 319, 337

Prušák J..................................... 43

Ribeiro Gomes M..................... 123

Richon D.................................. 165

Risegari L................................. 273

Šafrata S.................................... 43

Saidi M.H................................... 75

Sarikhani N................................ 75

Schmidt J.................................. 267

Schumann B............................. 313

Schustr P.................................... 43

Scurlock R.G........................ 35, 83

Seokho K................................. 281

Simonenko O. Yu..................... 159

Simonenko Yu. M.................... 325

Sosnowski J............................. 297

Spörl G............................ 245, 313

Srnka A.................................... 131

Stefan L.................................... 319

Stipsitz J................................... 307

Stoček P.................................. 219

Straková H............................... 233

Strnad P................................... 253

Strychalski M........................... 115

Sullivan P.................................... 97

Suma J..................................... 183

Takács D.................................. 199

Tucker A.S................................. 89

Ungricht Th................................. 51

Vasut F............................ 319, 337

Ventura G................................. 273

Waldie B.................................. 151

Wang C...................................... 83

White V.................................... 151

Wilhelm H.................................. 51

Zashlyapin R.A......................... 351

Equivalent power for liquefaction w/o LN2

Table 2. VIP line nominal dimensions

Design parameters of the vacuum insulated transfer systems are designed in accordance with the data in the table 3.

Autologous use

I

II