RRFM 2008 Transactions - European Nuclear Society


RRFM 2008 Transactions - European Nuclear Society

© 2008

European Nuclear Society

Rue de la Loi 57

1040 Brussels, Belgium

Phone + 32 2 505 30 54

Fax +32 2 502 39 02

E-mail info@euronuclear.org

Internet www.euronuclear.org

ISBN 978-92-95064-04-1

These transactions contain all contributions submitted by 29 February 2008.

The content of contributions published in this book reflects solely the opinions

of the authors concerned. The European Nuclear Society is not responsible

for details published and the accuracy of data presented.

2 of 435

Session I

International topics and overview on new

projects and fuel developments

3 of 435

Making the Nuclear Renaissance a Certainty


Director Business Development


Paul Gossen Str. 100

91058 Erlangen


In the following AREVA’s views and experience with the intensively discussed

renaissance on the market for nuclear power for electricity generation will be briefly

presented. For AREVA this renaissance started earlier than for anyone else by being

awarded with the first contract for a nuclear power plant project of the new generation.

Today we have two plants under construction, two more have been ordered and many

more are under discussion.

But what are the drivers for this renaissance of nuclear power and what are the real

challenges linked to it? It is expected that the world energy demand will grow for

15 000 TWh to 30 000 TWh by 2030. This is caused:

1. by an increasing demand of energy, due to several factors :

o Demography : there will be 2 billion more people on earth by 2030.

o The legitimate economic growth in fast developing countries, such as China,

South Africa, Brazil, India etc.

o Growth in developed countries: in spite of improvements in energy efficiency,

our modern way of life, with computers, air-conditioning and the like, is pulling


2. by security of supply, which comes in two components: reliable supply and at

affordable cost

There is a consensus that prices of oil and gas will remain high. And reliable supply

shall not be at the expense of affordability. Otherwise economical activities and jobs

are threatened, development of poor countries may never happen!


3. last but not least, the environmental concern: climate change is a new and

daunting global challenge.

In mitigating this increased demand caused by the three major drivers there is not a

single solution for the world. All available means need to be developed and must play

their role in a well balanced mix of energy sources for future electric power generation

and nuclear has to be part of it.

For AREVA meeting the challenge means being an integrated supplier with a global

infrastructure that is locally accessible with production and manufacturing in 41

countries and sales and marketing in over 100 countries. Our nuclear operations are

supported by 38 000 nuclear experts.

4 of 435

The worldwide nuclear capacity will grow in the coming decades. It should be noted

that in all scenarios, the nuclear stake remains constant in the mix at almost 15% of

the electricity generation. But even in the minimum scenario more than 100 new

reactors need to be built by 2030. Such a demand for new projects can only be

managed by the industrial standardization of reactor models.

Already in the beginning of the 1990ties AREVA with its industrial partners from the

electricity generating industry in Germany and France started the design of the EPR

from the well proven basis of their existing reactor fleet. The project was closely

monitored and supported by licensing authorities and independent inspection

agencies in both countries to ensure the EPR's licensability in France and Germany.

For the Finnish Olkiluoto 3 project, the EPR then underwent a complete design review

for the first time. Following a positive overall assessment by the Finnish authorities the

Government granted the construction license in February 2005. Before the customer

takes over the power plant, he must first apply for an operating license as part of the

second stage in licensing.

The EPR builds on proven technologies deployed in the two countries' most recently

built nuclear power plants – the French N4-series units and the German KONVOIseries

plants – and constitutes an evolutionary concept based on these designs. An

evolutionary design was chosen in order to be able to make full use of all of the

reactor construction and operating experience that has been gained not only in

France and Germany – with their total of more than 2100 reactor operating years –

but also worldwide. Guiding principles in the design process included the

requirements elaborated by European and US electric utilities for future nuclear power

plants, as well as joint recommendations of the French and German licensing


The EPR design as it is build now in Finland and France comprises and enhanced

safety level as compared to the former reactor generations and assures competitive

power generation cost with any kind of alternative power generation means, whether

fossil or renewable. It is the basis of a standard design that can be realized on almost

all available nuclear power plant sites around the world with only minor site specific


Safety levels at nuclear power plants have been constantly enhanced in the past. The

EPR, a nuclear reactor of the third generation, represents yet another step forward in

terms of safety technology, offering in particular the following features:

o Improved accident prevention, to reduce the probability of core damage even

further: This is provided by a larger water inventory in the reactor coolant

system, a lower core power density, high safety-system reliability thanks to

quadruple redundancy and strict physical separation of all four safety system

trains, as well as digital instrumentation & control systems and an optimized

man-machine interface.

o Improved accident control, to ensure that – in the extremely unlikely event of a

core melt accident – the consequences of such an accident remain restricted

to the plant itself: this is done by confining the radioactivity inside a robust

double-walled containment, by allowing the postulated molten core material

(corium) to stabilize and spread out underneath the reactor pressure vessel

and by protecting the concrete against meltthrough.

o Improved protection against external hazards (such as aircraft crash, including

large commercial jetliners) and internal risks (such as fire and flooding).

5 of 435

The EPR has a slightly higher reactor thermal output than other pressurized water

reactors currently in operation. The deployment of steam generators with economizer

sections along with an advanced steam turbine design lead to a higher efficiency.

Safety systems directly connected to the reactor coolant system serve to inject

coolant into the system and to remove residual heat in the event of a loss-of-coolant

accident (LOCA) are designed with a four fold redundancy. The in-containment

refueling water storage tank serves to store water for emergency core cooling and

accommodates any leakage water discharged via a pipe break in the reactor coolant


In addition to the systems for residual heat removal that are connected directly to the

reactor coolant system, a further system designed to assure heat removal in the event

of loss of normal feedwater supply is connected to the secondary system. This

consists of a four-train emergency feedwater system that supplies water to each

steam generator. In the steam generators, the heat generated in the reactor is used to

produce steam for driving the turbine. This steam is then condensed in the turbine

condenser. If the condenser should be unavailable due to loss of the main heat sink,

the excess steam can be directly discharged to the atmosphere from the steam

generators. The emergency feedwater system on the secondary side is equipped with

electric-motor-driven pumps that can be powered, if necessary, by the unit's four large

emergency diesel generators.

Full four-fold redundancy is provided for all safety systems and all of their auxiliary

systems. The risks associated with common mode failures – which can also affect

redundant systems of technically identical design – have been reduced by

systematically applying the principle of functional diversity. If one redundant system

should completely fail, there is always another system of diverse design that can take

over its tasks, thus enabling the EPR to be safely shut down and cooled. The

redundant trains of the safety-related systems are installed with strict physical

separation in four different buildings so that any interference between the redundant

systems is ruled out.

Not only the probability of occurrence of core damage states has been drastically

reduced, but the radiological consequences of severe accidents have additionally

been limited by means of a new containment design. This new design ensures that

the containment will retain its structural integrity under accident conditions. Any

radioactive leakages from the primary containment are collected in the space between

the two containment shells and can be directed through a filter system before being

discharged to the outside atmosphere. This means that even in the hypothetical event

of an accident causing melting of the core its consequences would be limited to the

plant itself so that no emergency actions in the vicinity of the plant would become


Besides the mitigation of hypothetical severe accidents EPR features in addition a

protection against the crash a commercial airliners. This protection is realized by thick

reinforced concrete walls covering the reactor, the fuel and two of the four redundant

safeguard buildings. In addition to the load effects, also induced vibrations need to be

considered. This is realized by the double wall of the reactor building, so that the

internal structures supporting safety related equipment are completely decoupled from

the outer concrete structure. Due to this design induced vibrations cannot directly

affect the component supports, but have to be routed via the basemat and being

damped on that way. Another consideration to be made when addressing the

protection against airplane crash is the effect of fuel fires caused by kerosene.

Consequently all building openings and ventilation ducts need to be protected in order

to avoid ingress of burning fuel into the building.

6 of 435

The high degree of redundancy does not only provide the required enhanced safety

level, but opens as well the chance to maintain redundant systems even during power

operation. This leads to shifting maintenance work from the shutdown period of the

plant for routine refueling operations to the operation period. As a consequence the

required annual shutdown time is reduced and the plant availability is increased,

which contributes to lower operation cost and improves the economic advantage of

the plant.

AREVA can claim today to be the first plant supplier with experience in constructing

Generation III nuclear power plants with these design features. This experience is

being gained through our projects in Finland and France. EPR is furthermore in

advanced licensing processes in the US and the UK by applying for a design

certification and by being subject to a generic design assessment, respectively.

In addition to these activities we are preparing for the projects in China for which the

contracts were recently signed and for the Constellation Energy project at Calverts

Cliff in the United States.

EPR is also under consideration for a number of emerging projects that are in an

earlier status of preparation. For ESKOM in South Africa we have just submitted bids

for two EPRs to be constructed as start of a fleet in this country. In the US EPR has

been selected by a number of utilities other than Constellation for their nuclear

programs to come in the short-term future. The GDA process of EPR in the UK is

supported by more than ten utilities that plan to invest into projects, once a prelicensing

statement of the British authorities has been granted. Also for the project of

the Baltic countries in Lithuania at the site of Ignalina, a plant with EPR technology is

under consideration.

The above gives just a list of projects that are in an advanced planning state. There

are many more countries and investors that started to reconsider nuclear power after

the Finnish and French projects had been launched. Should all these projects that are

under discussion to come on line by 2030 be realized, the nuclear industry will face a

big challenge. Not only the recruitment and training of young engineers will be

demanding, also the whole supply chain with its hundreds of subcontractors requires

a reassessment. Thanks to the early start with EPR, AREVA can benefit from the

advantages of an existing supply chain that had been established some years ago for

Olkiluoto 3. AREVA has invested into its own manufacturing workshops in particular

for upgrading its manufacturing capabilities for primary circuit equipment. In addition a

number of strategic partnerships with experienced subsuppliers were concluded to

ensure a reliable and timely delivery of components needed for all these projects.

The nuclear market is booming with a big number of new projects to be realized in the

short-term future. AREVA has made a lot of valuable experiences in the early

construction projects of EPR. We are well prepared and we continuing to adapt to the

needs of the market in the years to come.

7 of 435

The Karlsruhe Institute of Technology (KIT):

Research, Teaching and Innovation

Joachim U. Knebel

Forschungszentrum Karlsruhe GmbH

Programme Nuclear Safety Research (NUKLEAR)

Hermann-von-Helmholtz Platz 1

D-76344 Eggenstein-Leopoldshafen

Tel +49 (7247) 82 5510 • joachim.knebel@kit.edu

In the future, the Universität Karlsruhe (TH) and the Forschungszentrum Karlsruhe –

an excellence university and a national Helmholtz center – will pursue their missions

together at the Karlsruhe Institute of Technology (KIT). By consolidating their

capacities in research, teaching, and innovation, the two partners are laying the

foundations to become one of the internationally leading institutions for science and

technology. Their integrated executive, management, and codetermination bodies

will realize joint planning of strategy, structure, and development, following the

principle that research, teaching, and innovation constitute a unified entity and

introducing comprehensive lasting changes at both institutions. In Germany, the KIT

will serve as a model and meet the recommendation repeatedly expressed by the

Wissenschaftsrat “to intensify networking between universities and extra-university

research institutions " 1 .

Profile building and integration of the partners in the area of research will take place

on two levels: on the one hand through the competencies 2 , staff members of both

partners will bring to KIT, and on the other hand through concrete research work

conducted in projects of rather different scope and structure.



Wissenschaftsrat (Council for Science): Empfehlungen zur künftigen Rolle der Universitäten im

Wissenschaftssystem vom Januar 2006; Wissenschaftsrat, Drucksache 7067-06. S. 31 [Recommendations on

the future role of universities in the sciences, of January 2006; Council for Science, Print no. 7067-06, p. 31]

Competence means individual topic-related skills and the expertise of the staff members, including

methodological knowledge, to work on scientific and technological questions along generally valid quality


8 of 435

The expertise, skills, and research profiles of all KIT staff members will be organized

into joint areas and fields of competence. The resulting competence portfolio will

provide easy internal and external access to the scientific and technological

competencies of KIT and make them transparent. The generation of new projects will

be supported by seed money which is awarded to the best ideas emerging from

internal competition. The joint competence portfolio will be the basis for all ongoing

research at KIT and the breeding ground for new scientific ideas, projects, and

networks either formed among staff members themselves (“bottom-up”) or initiated

strategically (“top-down”).

Profiling of KIT research topics will take place at the institutional level through KIT

Centers and KIT Focuses which will combine and provide strategic support to

thematically related projects of different scope. KIT Centers stand out through their

unique characteristics in terms of scientific approach, strategic objectives, and tasks.

At the centers, national research objectives can be pursued in a better and more

comprehensive way as the program-related research of the Helmholtz Association

and the independent research of university groups will complement and strengthen

each other. KIT Focuses differ from KIT Centers with respect to the nature of their

socio-political mission, their size, and their duration. By consolidating research

capacities at KIT Centers and KIT Focuses critical mass is being achieved, enabling

KIT research to gain international competitiveness and visibility.

At KIT, excellent research is conducted outside of KIT Centers and KIT Focuses as

well and plays an important role in developing new research topics. This is why KIT

supports this research with measures laid down in the competence portfolio and

described in the Concept of the Future 3 .

Teaching and study at KIT are characterized by comprehensive supervision and care

of students, promotion of their early independence, and the extensive inclusion of

research. The integration of staff members from the Forschungszentrum into

teaching will drastically improve the student/instructor ratio, which will help reach

similar standards of international top-level universities in this respect as well. 4

Early independence and inclusion into research activities will be supported by

stronger integration of seminar, bachelor, diploma, and master’s theses into research

projects of different scopes throughout KIT, comprising even research projects of

major social relevance. Feasibility studies carried out by students and supported in

the context of the KIT Concept of the Future also serve this purpose. Establishing

KIT Schools will considerably extend interdisciplinarity in teaching. Being closely

related to and maintaining intense exchange with KIT Centers and KIT Focuses, they



Universität Karlsruhe (TH) (2006). A Concept for the Future of the Universität Karlsruhe (TH) – The

Foundation of KIT (Karlsruhe Institute of Technology).

Wissenschaftsrat (Council for Science): Empfehlungen zur künftigen Rolle der Universitäten im

Wissenschaftssystem vom Januar 2006; Wissenschaftsrat, Drucksache 7067-06, S. 87. [Recommendations

on the future role of universities in the sciences, of January 2006; Council for Science, Print no. 7067-06, p.


9 of 435

incorporate research, its methods, and its findings into teaching.

At KIT, the promotion of young scientific talent is based on excellent scientific

working conditions and aims at an adequate balance between early independence,

individual supervision and care, and training during the doctoral phase. This support

is provided by institutes and departments and is complemented by new

interdisciplinary elements of the KIT Schools and by promotion measures in the

context of the Concept of the Future 3 . The House of Competence (HoC) and the

Karlsruhe House of Young Scientists (KHYS) - overall structures at KIT - provide

support to young scientists in acquiring key qualifications and establishing

international networks.

The Forschungszentrum Karlsruhe and the Universität Karlsruhe (TH) already rank

among the leading innovative partners for business and industry in certain fields.

With KIT, this position will be expanded strategically. For this purpose, KIT will

introduce new instruments such as Shared Professorships and Shared Research

Groups as well as the KIT BusinessClub and the Karlsruhe Foundation for


KIT’s central idea is the integration of university and non-university research 5 ,

something that has been repeatedly demanded in the past. In implementing this idea,

KIT will consistently surpass every other model, thus setting new standards for

research, teaching, and innovation. In order for KIT to exploit its full potential, the

internal and external conditions for all those participating in the research, teaching,

and innovation process will need sustainable improvement.

Further specific information on KIT can be taken from the document ‘Concept for the

Karlsruhe Institute of Technology (KIT)’ and from http://www.kit.edu.

On February 22 2008 the Founding Ceremony of KIT took place in Karlsruhe, with

Federal Minister Dr. A. Schavan and Minister Prof. P. Frankenberg being present.

„Now, an important step towards the real merger is done: KIT will be set up as a

public body according to the Baden-Württemberg state law,“ announced BM A.

Schavan. Thus, KIT will be one legal entity with two missions: the mission of a state

research university and the mission of a national programmatic research centre

within the Helmholtz association.


Wissenschaftsrat: Empfehlungen zur künftigen Rolle der Universitäten im Wissenschaftssystem vom Januar

2006; Wissenschaftsrat, Drucksache 7067-06, S. 31. [Recommendations on the future role of universities in

the sciences, of January 2006; Council for Science, Print no. 7067-06, p. 31]

10 of 435

Research Reactor Coalitions

- First Year Progress Report

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe a , Kevin Alldred and

Nigel Mote b†


International Atomic Energy Agency (IAEA), Vienna, Austria


International Nuclear Enterprise Group, LLC, (INEG), USA

Abstract. The IAEA has initiated new activities with the objective of promoting formation of

coalitions of research reactor operators and stakeholders. The aim of this effort is to promote concrete

examples of enhanced regional cooperation, to form networks of research reactors conducting joint

research or other shared activities, and to form a voluntary, subscription-based, self-financed coalition.

The objective is to increase research reactor utilization and thus to improve sustainability at the same

time enhancing nuclear material security and non-proliferation objectives. This effort builds upon

existing IAEA efforts to enhance research reactor strategic planning, to encourage formation of

research reactor networks, and to promote regional and international cooperation.

This paper will describe progress in the first year of IAEA activities to assist the formation of research

reactor coalitions. This includes IAEA efforts to serve a catalytic and “match-making” role for the

formation of new commercial and other relationships to increase research reactor utilization, including

organizing various missions and meetings for exploratory and initial organizational discussions on

possible coalitions and networks .This also includes activities to assist research reactors in carrying out

strategic planning with a view to forming research reactor coalitions, training activities to assist in the

development of nascent coalitions, and development of arrangements to facilitate access to

stakeholders requiring irradiation services and for countries that are not operating a research reactor.

1. Background

Research reactors play a key role in developing the peaceful uses of nuclear energy. In order to

continue in this role, they need to be financially sound, with adequate income for safe and secure

facility operations and maintenance, including planning for eventual fuel removal and

decommissioning. However, in a context of declining governmental financial support, many research

reactors are increasingly challenged to generate additional income to offset their operational costs,

without making any provision for the liabilities that will be incured when their facilities reach the end

of their operating lives Reactors operating at low utilization levels have difficulty providing products

and services with the reliability demanded by potential users and customers, and this creates a

significant obstacle to increasing utilization.

These challenges are also occurring in the context of increased concerns about nuclear material safety

and security and the threat of nuclear proliferation, due to which research reactor operators are

compelled to substantially improve physical security and convert reactors from highly enriched

uranium (HEU) to low enriched uranium (LEU) fuel. Thus, there is today a complex environment for

† New Milford, Connecticut and Alpharetta, Georgia.


11 of 435

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe and Kevin Alldred and Nigel Mote

research reactors, and one in which underutilized, and therefore likely poorly-funded, facilities invoke

particular concern.

Many research reactors have limited access to potential customers for their products and services and

are not familiar with the business planning concepts needed to secure additional commercial revenues

or international program funding. This not only results in reduced income for the facilities involved,

but sometimes also in research reactors contracting for services at prices below those required to cover

their full costs, preventing recovery of back-end costs and creating unsustainable market conditions.

The research reactor community possesses the expertise to address these concerns. However, this

knowledge is not uniformly available as parochial attitudes and competitive behaviour restrict

information sharing, dissemination of best practices, and mutual support that could otherwise result in

a coordinated approach to market development, building upon strengths of facilities.

These attitudes are based, in part, on the belief that the markets for research reactor products and

services are “zero-sum,” with market gains by one research reactor resulting in losses by another

“competing” reactor. However, the formation of coalitions will likely stimulate new demand for

products and services, without reducing the demand from existing users.The success of user groups

and organizations such as WANO in the nuclear power generation sector show that the benefits of

cooperation can be obtained without sacrificing commercial interests.

Renewed interest in nuclear power and the worldwide expansion of diagnostic and therapeutic nuclear

medicine presents new opportunities to expand the use of research reactors – including by countries

without such a facility. However, a reactor constructed to meet a specific need might not have

sufficient identified utilization to fully occupy the facility, or to be adequately available for its

intended purpose. A potential solution to this dilemma would be the creation of one new multinational

facility rather than a number of national facilit ies, but this requires an increased level of

coordination between current and prospective operators.

To address the complex of issues related to sustainability, security, and non-proliferation aspects of

research reactors, and to promote international and regional cooperation, the Agency has undertaken

new activities to promote Research Reactor Coalitions and Centres of Excellence. This integrates

Agency regular and extra-budgetary funded program activities related to research reactors, national

and regional IAEA Technical Cooperation projects, especially “Enhancement of the Sustainability of

Research Reactors and their Safe Operation Through Regional Cooperation, Networking, and

Coalitions” (RER/4/029) and “Nutritional and Health-Related Studies Using Research Reactors”

(RAF/4/020; AFRA IV-12), and is also supported by a grant from the Nuclear Threat Initiative (NTI).

2. Concept outline

From the operational perspective, coalitions will facilitate peer group sharing of best practices,

improve information availability to members, and both reinforce and develop the operating disciplines

of safety, security and quality control. From the business perspective, coalitions will provide improved

market analysis and support for strategic and business planning. Where appropriate, coalitions may

jointly market services and increase contacts between research reactor operators and prospective

customers. By so doing, they will help increase reactor utilization, improve the services provided to

the communities they serve, generate additional revenues and thus justify additional investment in

operational improvements.

From the public perspective, coalitions will have the opportunity to enhance the information available

to help retain and build confidence in reactor operation.

There is not a “one size fits all” solution and coalitions can take several different forms according to

the needs and capabilities of their members. Possible coalition variants include: bilateral subcontracting,

joint venture and other supply arrangements between pairs of, or larger groups of,

research reactors; informal peer group networks that can share best practice information; and broader


12 of 435

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe and Kevin Alldred and Nigel Mote

coalitions that are capable of effectively marketing their members’ services and representing their

interests in common, as well as setting standards for all members. It is expected that some coalitions

will also offer access to members from non-reactor owning countries, with financial subscriptions paid

in return for access to reactor services. This will result in increased utilization of existing, or purposebuilt

facilities, thus avoiding construction of new reactors that will not be fully utilized or continued

operation of marginally supported reactors.

In most cases, it is envisaged that coalitions will not start with full scope implementation, but rather

will develop from relatively modest starting points (e.g. involving two or three members coordinating

a single activity), and will expand their scope of implementation as the confidence of the members,

and their governments, increases. For example, a simple, bilateral backup supply arrangement may

grow into an informal network, and eventually become a subscription-based coalition.

3. Concept benefits

A coalition is expected to have both general and specific benefits to participating research reactors.

The general benefits include such items as standardization of operating practices and security

procedures. The specific benefits of a coalition will derive from improved strategic and business

planning (using IAEA-TECDOC-1212 “Strategic Planning for Research Reactors” as a guide) and

joint marketing of the services of its participant reactors (commercial products and scientific/research

activities), with the coalition thus able to:

• Optimize the services offered (possibly including education and training, production of isotopes,

industrial irradiation services such as transmutation doping, neutron activation analysis and

other analytical services for industry and government) on a geographical basis, and reduce

operational costs.

• Maximize the use of specialized expertise or equipment at a particular facilities, and enable

facilities to specialize in services in which they could have a “comparative advantage.”

• Use the combined expertise of the participant facilities to best advise and serve their customers.

This would help increase customer knowledge of, and access to, the services and products the

coalition can provide, and support the customer with a more reliable and comprehensive

customer service.

• Improve the utilization and sustainability of individual research reactors, and increase overall

levels of demand to the mutual benefit of all market participants (suppliers and customers).

Increasing reactor utilization would generate additional revenues, or help make the necessary

justifications for additional local governmental support, thus improving sustainability. The

additional funding could assist individual reactors to pay for operational, safety and security


• Develop a common methodology for calculating costs of reactor services to include spent fuel

management and eventual decommissioning liabilities.

• Act as a coordinated entity in procuring new fuel and contracting for spent fuel management

services, thus reducing the costs of these activities incurred by each reactor operator and

benefiting from the economy of scale

• Provide assistance to reactors planning or undergoing conversion from HEU to LEU including

sharing of experience and planning expertise.

• Address needs of user groups without access to a research reactor in their Member State(s).

4. IAEA Activities and Progress

The Agency’s role is to serve as a catalyst and a facilitator of ideas and proposals. Meetings held by

the IAEA in August and September 2006 resulted in preparation of a grant request on research reactor

coalitions which was submitted to the Nuclear Threat Initiative (NTI) and approved in October 2006.

From October 2006 to January 2007, the IAEA conducted informal consultations with a wide number

of research reactor operators, commercial entities, users of research reactor irradiation services, and


13 of 435

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe and Kevin Alldred and Nigel Mote

other stakeholders. Approximately fifteen “notional proposals” for coalitions covering a range of

subjects and virtually all geographic areas were initiated, which became the basis of the Agency’s

initial activities in 2007. Following initial discussions with potential participants, several of the

notional proposals were further elaborated and then became the basis for exploratory meetings in fall


A. IAEA as “Matchmaker”

The IAEA identified several “matchmaker” opportunities. Two are described here as examples of how

coalitions can benefit both reactor operators and their customers. In both cases, the Agency’s initial

contacts led to direct meetings and negotiations between the various partners without the Agency’s


The first was between a well-utilized research reactor and another, less-well utilized but state-of-the–

art, research reactor in the same geographic region. In this case, the well-utilized reactor was seeking

additional irradiation capacity for its commercial business. In this coalition, the well-utilized reactor

will serve as the “lead reactor,” sub-contracting work to the second reactor based on the first reactor’s

orderbook. It will ensure that quality control and quality assurance procedures and standards are

adhered to by the sub-contracting reactor so that the products delivered to the lead reactor’s customers

meet the same standards as products irradiated in its own facility.

In the second example, the Agency brought together an existing research reactor supplier of industrial

isotopes , which is planning for cessation of operations, a commercial user of industrial isotopes/tracers

and an underutilized research reactor in a region where the commercial user had a growing demand for

industrial isotopes. In this case, the reactor is projected to be a direct contractor/supplier to the

commercial user, based on a non-exclusive contractual arrangement. The IAEA conducted a training

workshop at Imperial College U.K. from May 14-16, 2007 to assist staff of the underutilized research

reactor in understanding the management issues associated with supply of isotopes to a commercial


Following consolidation of these contractual arrangements, the IAEA will encourage the respective

partners to add additional members to the contractual arrangements, at a minimum to ensure back-up

production arrangements and to expand the “menu” of technical capabilities offered by the coalition.

B. Strategic planning for coalitions

Strategic planning assists research reactors to better understand their strengths and weaknesses, and

their stakeholders and stakeholder needs, and to adjust their activities to address national development

priorities as well as the commercial marketplace. Strategic planning can also assist research reactors in

developing ideas for alliances or coalitions based upon complementary strengths and weaknesses.

The IAEA organized an expert mission to Kazakhstan and Uzbekistan from 8-12 October 2007 to

assist the staff at the respective Institutes of Nuclear Physics to further develop strategic plans and to

consider formation of cooperative ties between the research reactors in the region. At an IAEA

Workshop on Advanced Strategic Planning for Research Reactor Coalitions (Europe region), Vienna,

17-19 December 2007, representatives of the two countries proposed formation of a Central Asia

Research Reactor Coalition, and a number of actions are contained in the meeting report with a view

toward concluding such an arrangement.

The workshop cited above was also attended by representatives of user organizations and research

reactor operators from Armenia, Austria, Azerbaijan, Czech Republic, Italy, Kazakhstan, Norway,

Romania, and Russia. The research reactor operators made presentations relating to their utilization

patterns and the development of strategic plans, based on a SWOT analysis (strengths, weaknesses,

opportunities, and threats), including the example of a research reactor which made a successful

transition from a state-supported institution to a fully commercial operation. Participants without

research reactors made presentations regarding their nuclear science, irradiation, nuclear power plant


14 of 435

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe and Kevin Alldred and Nigel Mote

support and training, and radiation protection needs for which access to, or services from, a research

reactor are necessary. The participants also visited the TRIGA reactor at the Atominstitut (ATI) of the

Vienna University of Technology for briefings on strategies and activities for the successful utilization

of a low-power research reactor, particularly for education and training purposes.

The final report of the workshop contains suggestions from each of the participants regarding ideas for

cooperation and collaboration with other research reactors and concrete proposals for research reactor

coalitions, with specific action items. In addition to the Central Asia Research Reactor Coalition noted

above, these include:

-Nuclear Education and Training Coalition (potentially involving Armenia, Azerbaijan, Austria/ATI,

Czech Republic/CTU, and Italy)

-Innovative Reactor Systems and Fuel Cycles (potentially involving Czech Republic/Rez,

Norway/Halden, Romania/INR, Russia/RIAR, and Ukraine.

-Central/Eastern Europe (via an external proposal from Hungary, and also involving Czech Republic,

Romania, and Poland)

The IAEA is currently pursuing a number of activities relevant to the first two proposals through both

regular budget and Technical Cooperation program mechanisms.

On the final proposal, the IAEA participated as an observer in an exploratory meeting organized by

KFKI in Budapest, Hungary on 28-29 January 2008 concerning the formation of an Eastern Europe

Research Reactor Coalition. The participants reached preliminary agreement to hold further

discussions with the objective of initiating enhanced cooperation in the field of neutron beam

experiments. .

C. Exploratory missions on forming research reactor coalitions

Missions and meetings were organized in fall 2007 to discuss forming specific coalitions:

1. Russian Federation experts and institutions, Dmitrovgrad, Russian Federation, 5-6 September

2007, and Vienna, Austria, December 13-14, 2007;

2. Instituto Peruano de Energia Nuclear (IPEN), Peru and Comision Chilena de Energia Nuclear

(CCHEN), Chile, with Missouri University Research Reactor (MURR) and McMaster Nuclear

Reactor (MNR), Lima, Peru and Santiago, Chile, 15-19 October, 2007;

3. CNEA (Argentina) and ATI, Buenos Aires, Argentina, 22-23 October, 2007;

4. ININ (Mexico) – Laguna Verde Nuclear Power Plant – ATI, Centro Nuclear ININ, 29 October


5. Caribbean region research reactor coalition (Jamaica-Mexico-Colombia), Centro Nuclear

ININ, 30-31 October 2007.

The meetings with Russian experts in September and December resulted in conclusion of meeting

protocols that cited a number of possible areas for coalitions among Russian research reactors and/or

with research reactors outside Russia. These include Russian coalitions for i) education in nuclear

science and engineering, and ii) industrial and medical radioisotopes; and international coalitions for

a) nuclear science and materials testing and b) LEU fuel conversion. Follow-up meetings and facility

visits to plan implementation steps are scheduled for March 12-14, 2008 in Russia.

The missions and facility visits that took place in October 2007 to Chile and Peru were led by the

IAEA. The team included representatives from MURR and MNR for discussions on possible

coalitions involving medical and industrial radioisotope research, development, and production.

Protocols with action items were agreed for both missions, which included a number of concrete ideas

for supply of radioisotopes between institutions and for transfer of production technology There has

been an extensive exchange of information in the following months, as well as arrangements


15 of 435

Ira N. Goldman, Pablo Adelfang and Shriniwas K. Paranjpe and Kevin Alldred and Nigel Mote

concluded for radioisotope supply. It envisaged that further meetings will be held in mid-2008 to

further formalize the coalition arrangements and to plan next steps.

IAEA-led missions to Argentina and Mexico in October 2007 included a representative from the

TRIGA reactor at ATI. These meetings focused on establishment of coalitions involving nuclear

education and training activities, including with the Insituto Dan Beninson (CNEA/Argentina), ININ

and the Laguna Verde Nuclear Power Plant (Mexico). Preliminary coalitions agreements were signed,

with specific follow-up steps defined. As a result of the meeting in Mexico, ININ is developing a

practical reactor operations training course for personnel from the Laguna Verde Nuclear Power Plant

to be held at its TRIGA reactor in 2008.

Preliminary agreement was reached at a meeting at ININ on 31 October 2007 to form a Caribbean

research reactor coalition between the three reactors in Colombia, Jamaica, and Mexico. It is

envisaged that this coalition will serve as a regional resource for users of nuclear science and

irradiation services in other countries in the Caribbean region that do not have research reactors. The

focus of its activities will initially be on neutron activation analysis, especially for environmental

applications, as well as training services . A draft Memorandum of Understanding for the coalition is

under review by the parties, a reactor operator certification course is being formulated by ININ (for

Colombia), and Jamaica is developing a course on neutron activation analysis.

Other proposals related to potential coalitions, including in Africa and East Asia and the Pacific are

still in the formulation stage, with exploratory meetings to be held in 2008. Of particular note, the

IAEA held a meeting in Vienna from 11 to 13 February, 2008, to explore the formation of a neutron

sciences/neutron scattering coalition with representatives primarily from the Europe region but also

from Australia and the U.S.

5. Conclusion

The Research Reactor Coalitions initiative has made considerable progress during its first year of full

activity. The IAEA has successfully played the role of “catalyst” and facilitiator of ideas. As a result –

and perhaps most importantly – the coalitions concept seems to be gaining international acceptance,

with the term frequently used in international research reactor meetings and discussions.

As further evidence of this, a number of countries and institutions have formulated, and more are

developing, their own proposals for coalitions.

The IAEA has also successfully identified a number of opportunities to act as “matchmaker” in

introducing and facilitating discussions between partners that led to new commercial arrangements for

increased utilization of specific research reactors. These arrangements are expected to form the basis

for broader research reactor coalitions in the future.

In addition, a significant number of exploratory missions and discussions were held, resulting in initial

or preliminary agreements for several coalitions. While these are still being developed, it is expected

that one or more formal research reactor coalitions will come to fruition in 2008 as a result of these


The IAEA invites suggestions and proposals for additional coalitions from other Member States and



16 of 435




CEA-Cadarache, DEN/DEC, 13108 St Paul Lez Durance Cedex - France


CEA-Saclay, DEN/DSOE, 91191 Gif sur Yvette Cedex - France


SCK•CEN, Nuclear Materials Science Institute, Boeretang 200, B-2400 Mol - Belgium.


ZWE FRM-II, Technische Universität München, D-85747 Garching bei München - Germany


AREVA-CERCA * , les Bérauds, BP 1114, 26104 Romans Cedex – France


This paper is an up date of the French IRIS program on high density UMo/Al

dispersion fuel. Some PIEs performed on the recent IRIS-3 and IRIS-TUM

experiments are presented and discussed. They confirm the good in-pile

behaviour of full size ground powder based plates up to high power and burn-up.

The positive effect of the Si addition to the Al matrix on the irradiation behaviour of

full size plates is also evidenced, in particular for atomised powder based

plates. Despite these good results and considering manufacturing and

reprocessing aspects, an oxide coated atomised UMo fuel is consequently

proposed as a promising solution.

1. Introduction

As alternatives to the very first fuel concept (dispersed atomised UMo in pure Al), the French

IRIS program has tested two improvements: modification of the matrix composition and a

change in the UMo powder characteristics [1]. Up to now, this program involves 4 full size fuel

plate experiments performed in the OSIRIS reactor on high density UMo dispersion fuel, IRIS1

[2], IRIS2 [3], IRIS3 [4] and IRIS-TUM [5]. The FUTURE plates [6] irradiated in the BR2 reactor

completed this program. The ground particle based fuels can show good in-pile behaviour, as

the IRIS1 experiment demonstrated. This was now confirmed by IRIS-TUM plate tests,

irradiated to higher equivalent burn-up at much higher load. The influence of the Si addition to

the Al matrix has been studied on both atomised (IRIS3) and ground (IRIS-TUM) UMo powder.

The Si benefit is obvious, especially for the plates made of atomised UMo powder. Postirradiation

examinations are in progress on the more recent irradiation tests. This paper gives

a preliminary comprehensive overview on the in-pile behaviour of these different fuels. The

predominant factors and their roles are discussed. In order to discriminate the different

parameters influencing the conservative in-pile behaviour of ground powder, a new experiment,

IRIS-4, with a fuel made of oxidised particles, is underway.

* AREVA-CERCA, a subsidiary of AREVA-NP, an AREVA and Siemens company

17 of 435

2. Main features of the IRIS experiments

The IRIS 1 to 3 experiments have been performed by CEA, within a close collaboration with

AREVA-CERCA for the manufacturing aspects. The IRIS-TUM experiment has been launched

in the framework of a collaboration between TUM, CEA and AREVA-CERCA.

All irradiations have been performed in the OSIRIS MTR reactor with the IRIS irradiation and

measuring device, originally developed to qualify the silicide fuel for the OSIRIS conversion

and FRM II [7].

All plates are full size and manufactured by AREVA-CERCA through classical rolling process.

The main manufacturing and irradiation features of the IRIS experiments are collected in Tab.


Manufacturing data

Irradiation data


UMo powder type ground atomised atomised ground atomised

Mo in UMo (wt%) 7.6 or 8.7 7.6 7.2 8.1 7

Enrichment ( 5 U wt%) 19.8 19.8 19.8 49.5 19.8

Si in Al matrix (wt%) 0 0 0.3 2.1 0 2.1 0 2.1

Matrix type A5 A5 AlSi0.3 AlSi2.1 A5 A5 AlSi2.1


Fuel loading (gU/cc) 7.9-8.3 8.2-8.3 7.8-8.0 7.3-8.4 7.9

As fab meat porosity (%) 11-13 1-2 0.8-2.4 8-9 1-2

Cladding material AG3NE AG3NE AG3NE AlFeNi AlFeNi

Year 2000-2001 2003 2005-2006 2005-2007 2008-2009

Number of plates 3 4 4 4 4

Status of experiment completed stopped stopped completed completed foreseen

OSIRIS core position 17 52 14 11 and 17 52

Max heat flux at BOL (W/cm 2 ) 123-145 238 201 250-258 290

Max clad surface temp. (°C) 68-73 93 83 97 100

Number of cycles 10 4 7 8 5-6

Duration (EFPD) 241 58 131 147 -

Plate average BU ( 5 U %) 46.9 32.5 48.8 35.3-59.3 LEU eq > 50

Average BU at MFP ( 5 U %) 54.0 39 56.5 43.4-69.8 LEU eq -

Max BU at MFP ( 5 U %) 67.5 39.7 58.8 56.3-88.3 LEU eq -

Average FD at MFP (f/cm 3 UMo) 3.2 10 21 2.2 10 21 3.4 10 21 4.2 10 21 -

Max FD at MFP (f/cm 3 UMo) 4.6 10 21 2.7 10 21 4.1 10 21 5.6 10 21 -

Tab. 1: Main features of the IRIS experiments

The main differences are related to :

• the type of UMo powder, atomised or ground,

• the type of matrix, either pure Al or added with silicon up to 2.1 wt %,

• the 49.5% enrichment of the IRIS-TUM plates to reach higher irradiation conditions,

• the maximum heat flux of about 120 W/cm 2 for IRIS1 to 258 W/cm 2 for IRIS-TUM (cf.

Fig. 2),

• the maximum clad surface temperature of 68°C for IRIS1 to 97°C for IRIS-TUM,

• the AG3NE or AlFeNi cladding.

3. Non destructive testing

The plate thicknesses have been measured before and after each cycle for all the IRIS plates.

The results are plotted as a function of fission density in Fig. 1. They demonstrate:

• the better in-pile behaviour of the plates made of ground particles up to high burn up

and heat flux, in comparison with the atomised UMo based fuel,

• the positive effect of Si addition to the Al matrix. This improvement is particularly

visible in the case of atomised UMo based fuel plates (IRIS-3). For the plates made of

18 of 435

ground UMo (IRIS-TUM), the effect of Si is covered by the features of the ground UMo

particles themselves (shape microstructure, defects, oxidised surface).



371 1237

local 1056 µm

IRIS1 ground 0%Si (U7MQ2003)

IRIS1 ground 0%Si (U9MQ2051)

IRIS-TUM ground 0%Si (U8MV8002)


Plate thickness increase (µm) ,











IRIS-TUM ground 0%Si (U8MV7003)

IRIS2 atom. 0%Si (U7MT2002)

IRIS2 atom. 0%Si (U7MT2003)

IRIS2 atom. 0%Si (U7MT2007)

IRIS3 atom. 0.3%Si (U7MV8011)

IRIS3 atom. 2.1%Si (U7MV8021)

IRIS-TUM ground 2.1%Si (U8MV8501)

IRIS-TUM ground 2.1%Si (U8MV8503)

Peak heat flux at BOL (W.cm -2 )





50 70 90 110

0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00

Max cladding temperature at BOL (°C)

Fission density (10 21 f/cm 3 UMo)

Fig. 1: Plate thickness increase with fission

density in UMo particles

Fig. 2: Irradiation conditions of the IRIS


4. Post Irradiation Examination

The IRIS-TUM plates U8MV8503 & U8MV8002 and the IRIS3 plate U7MV8021 (see Fig. 1 and

Tab. 2) have been recently examined by optical and scanning electron microscopy at the hot

laboratory (LHMA) of SCK•CEN in Mol, Belgium [8, 9].

IRIS-2 (0%Si) IRIS-3 (2.1%Si) IRIS-1(0%Si) IRIS-TUM (0%Si) IRIS-TUM (2.1%Si)

Fig. 3: Optical micrographs at MFP

Fig. 4: SEM images at MFP

Fig. 5: Detailed SEM images


Plate number (Si content) 2002 (0%) 8021 (2.1%) 2003 (0%) 8002 (0%) 8503 (2.1%)

Powder type atomised ground

As-fab porosity (%) 1.5 2.2 12.4 7.9 8.9

Max heat flux (W/cm 2 ) 238 201 124 254 258

19 of 435

T max clad surface (°C) 93 83 68 96 97

Max FD at MFP (10 21 f/cm 3 )


2.7 4.1 4.4 3.8 3.8

Max swelling (µm) 1237 90 77 104 93

Tab. 2: Main characteristics of the IRIS samples examined at the LECA and LHMA.

20 of 435

Some of the images collected are compared with those obtained at the LECA hot laboratory

of Cadarache, France. Their characteristics are gathered in Tab. 2. The different phases

existing in all the samples, determined by analysis of the SEM images, are plotted in Fig. 6.







as 1fab

IRIS-2 2 IRIS-3 3

atomized 0%Si 2%Si





Fig. 6: Surface fractions of the different phases for atomised (left) and ground (right) UMo.

The main observations derived from those images and plots can be formulated as follows:

• In all the samples, an interaction layer (IL) is formed at the UMo/Al interface at the

expense of the Al matrix and the UMo particles.

• The apparent volume of UMo particles is quite uniform. The UMo consumption is

compensated by its swelling due to fission products (FP) and fission gas (FG) bubble


• In the plates with few (0.3%) or no Si addition to the Al matrix, the IL is homogeneous

around all the fuel particles, while in the plates containing 2.1%Si, the IL is thinner,

irregular and jagged. In this latter case, the inter-diffusion Al/UMo seems to be partly


• In the plates with Si addition, Si particles are seen dispersed in the Al matrix except

close to the fuel particles probably because of fission track enhanced dissolution.

• An oxide layer (dark in the OM images) is clearly observed around ground UMo


• Fission gas bubbles, quite homogenous in size, are distributed in the fuel particles. In

atomised samples, these bubbles seem to reveal the cell boundaries (Mo depleted


• Some larger bubbles appear at UMo/IL interfaces and UMo/UMo inter-particle


• No or only very few crescent moon shape pores due to FG are detected at the Al/IL

interface. As these bubbles are the very start of the phenomenon leading to the large

pillowing observed in the IRIS-2 and FUTURE plates, their absence is a hint for a

more conservative behaviour of the IRIS-3 (2.1%Si) and IRIS-TUM plates.

5. Discussion











Recently, PIEs were performed on samples of IRIS-TUM plates U8MV8002 & U8MV8503.

These irradiations at high heat flux and BU confirm the observations already made on the

ground UMo based plates in the IRIS-1 PIEs. As discussed in our previous paper [1], several

characteristics of the ground fuel play a key role and are certainly at the origin of its

conservative in-pile behaviour. In random order, they can be listed as follows:

• Morphology/granularity:

o The irregular shape and size of ground particles could strengthen the cohesion

between the UMo particles and the Al matrix and increase plate mechanical




1 2 3 4

ground 0%Si 0%Si 2%Si





21 of 435

o The initial residual porosity in Al matrix, of about 10 vol. %, (against 1-2 vol. %

for spherical atomised powder), could act as a buffer for fission gases and

compensate part of the swelling.

o Another consequence lies in the amount of Al matrix available to react with

UMo particles. In ground fuel, the Al surface fraction is about 35%, much lower

than the 48% measured in atomised fuel plates.

• Microstructure:

o The high concentration of “defects” introduced by the mechanical grinding

process could also trap gas atoms.

o The UMo raw material, prior to powder production, is heat treated at high

temperature in order to avoid any Mo micro-segregation.

• Composition:

o Influence of Mo, O, Si on the IL composition, properties and stability at severe

irradiation conditions. Recent out-of-pile studies clearly showed the influence of

Si on the IL nature [10].

o The oxygen, introduced during grinding process as an irregular oxide layer

(UO 2 ) around UMo particles, and the Si particles, added to the Al matrix, seem

to act as a barrier to the inter-diffusion of Al/UMo, hindering the interaction

between UMo particles and Al matrix.

This positive effect of Si is particularly visible in the atomised UMo based IRIS-3 plates. For

the 0.3% Si containing plates, a pillowing occurred (cf. Fig. 1), as in the IRIS-2 and FUTURE

experiments, while in the case of the 2.1% Si plate, no abnormal swelling is observed [4]. The

PIEs performed on this 2.1% Si IRIS-3 plate U7MV8021 showed that 23% of the Al remains.

The IL represents only 22% of the volume (cf. Fig. 6), which is not enough for pillowing to

start. The Si particles close to the UMo fuel kernels act as obstacles to the inter-diffusion

Al/UMo [11] and IL growth. Various out of-pile heavy ion irradiation [12, 13, 14, 15] and

diffusion studies [10, 16, 17, 18] already showed this positive effect of Si in decreasing the

interaction rate between UMo and Al.

For a better quantification of the fission products (mainly gases) and IL/Al volume fraction

amounts and properties for breakaway swelling to occur, new measurements and

examinations (SEM, EPMA, XRD) of the IRIS-3 and high burned IRIS-TUM plates U8MV8501

& U7MV7003 are planned in 2008-2009.

6. Conclusion - Perspectives

The recent PIEs performed on the IRIS-3 (2.1%Si) and IRIS-TUM samples confirmed the

benefit of Si addition to the Al matrix. This effect is particularly visible in the case of plates

made with atomised UMo particles. For the ground UMo based fuel plates, this positive effect

is more difficult to evidence, because of the already good in-pile behaviour of ground UMo fuel

even without Si, which is related to its composition, microstructure and morphology.

To better discriminate the role of those different parameters, a new experiment, IRIS-4, with a

promising fuel made of oxidised particles has been launched. The objective is to test the

influence of an oxide layer coating on the UMo particles on the in-pile plate behaviour [19, 20].

The main specifications of this experiment are given in Tab. 1. Considering manufacturing

aspects and the difficulties to industrialise a grinding process, atomised particles have been

selected. The thermochemically controlled oxidation of the atomised UMo powder has been

done last autumn. The mean UO 2 thickness layer around UMo particles is 1.5±0.5 µm (cf. Fig.

7). The 4 full size plates, with or without Si addition to the Al matrix, have been already

produced at AREVA-CERCA [21] and will be irradiated in OSIRIS reactor from the middle of

2008. The fabrication and irradiation of test samples similar to IRIS-4 are planned by TUM.

22 of 435

Here the objectives are atomised powder of an enrichment of 49,8%, oxidized, with and

without Si addition and a heat load towards 400 W/cm 2 .

Fig. 7: Micrographs of the CEA/CERCA oxidised atomised UMo particles to be irradiated in

IRIS-4 experiment

7. References

[1] S. Dubois, J. Noirot, J. M. Gatt, M. Ripert, P. Lemoine, P. Boulcourt, RRFM, Lyon, France, 2007.

[2] F. Huet, V. Marelle, J. Noirot, P. Sacristan, P. Lemoine, RERTR, Chicago, Illinois, USA, 2003.

[3] F. Huet, J. Noirot, V. Marelle, S. Dubois, P. Boulcourt, P. Sacristan, S. Naury, P. Lemoine, RRFM,

Budapest, Hungary, 2005.

[4] M. Ripert, S. Dubois, P. Boulcourt, S. Naury, P. Lemoine, RRFM, Sofia, Bulgaria, 2006.

[5] A. Röhrmoser, W. Petry, C. Jarousse, J. L. Falgoux, P. Boulcourt, A. Chabre, P. Lemoine, RRFM,

Lyon, France, 2007.

[6] A. Leenaers, S. Van den Berghe, E. Koonen, C. Jarousse, F. Huet, M. Trotabas, M. Boyard, S.

Guillot, L. Sannen and M. Verwerft, J. Nucl. Mat. 335 (2004) 39-47.

[7] K. Böning, W. Petry, submitted to NIM A.

[8] A. Leenaers, S. Van den Berghe, S. Dubois, J. Noirot, M. Ripert, P. Lemoine, this meeting.

[9] A. Röhrmoser et al., this meeting.

[10] M . Cornen, M. Rodier, X. Iltis, S. Dubois, P. Lemoine, this meeting.

[11] A. Leenaers, S. Van den Berghe, E. Koonen, S. Dubois, M. Ripert, P. Lemoine, RERTR, Prague,

Czech Republic, 2007.

[12] H. Palancher, P. Martin, M. Ripert, S. Dubois, C. Valot, C. Proye, F. Mazaudier, RERTR, Boston,

USA, 2005.

[13] N. Wieschalla, K. Böning, W. Petry, A. Röhrmoser P. Böni, A. Bergmaier, G. Dollinger, R.

Großmann, J. Schneider, RERTR, Boston, USA, 2005.

[14] N. Wieschalla, A. Bergmaier, P. Böni, K. Böning, G. Dollinger, R. Großmann, W. Petry, A.

Röhrmoser and J. Schneider. J. Nucl. Mat. 357 (2006) 191-197.

[15] H. Palancher, P. Martin, V. Nassif, R. Tucoulou, O. Proux, J. L. Hazemann, O. Tougait, E. Lahéra,

F. Mazaudier, C. Valot and S. Dubois, J. Appl. Cryst. 40 (2007) 1064-1075.

[16] M. Mirandou, S. Balart, M. Ortiz and M. Granovsky, J. Nucl. Mat. 323 (2003) 29-35.

[17] C. Komar Varela, M. Mirandou, S. Arico, S. Balart, L. Gribaudo, RERTR, Prague, Czech

Republic, 2007.

[18] J.M. Park, H. J. Ryu, S. J. Oh, D. B. Lee, C. K. Kim, Y. S. Kim and G.L. Hofman, J. Nucl. Mat. In


[19] S. Dubois, F. Mazaudier, H. Palancher, P. Martin, C. Sabathier, M. Ripert, P. Lemoine, C.

Jarousse, M. Grasse, N. Wieschalla, W. Petry, RERTR, Cape Town, Republic of South Africa, 2006.

[20] F. Mazaudier, C. Proye, J. Miragaya, S. Dubois, P. Lemoine, C. Jarousse, M. Grasse, RERTR,

Cape Town, Republic of South Africa, 2006.

[21] C. Jarousse, G. Bourdat, S. Dubois, M. Ripert, P. Boulcourt, P. Lemoine, this meeting.

23 of 435







Idaho National Laboratory

P. O. Box 1625, Idaho Falls 83415 – U. S. A.


Argonne National Laboratory

9700 S. Cass Avenue, Argonne, IL 60439 – U. S. A.


Y-12 National Security Complex

Oak Ridge, TN 37830 – U. S. A.


Very high uranium density nuclear fuels are currently under development in the

U.S. to enable the conversion of many research reactors worldwide to LEU

based fuels. Significant progress has been made in both the uraniummolybdenum

based dispersion and monolithic fuel forms. The efficacy of silicon

additions to the matrix of dispersion fuel meats has been demonstrated. Full

size dispersion plates with loadings greater than 8.0 g-U/cc have been

fabricated with silicon additions to the matrix and are ready for irradiation testing.

Monolithic mini-plates with modified fuel/cladding interfaces (both silicon

enhanced and zirconium diffusion barriers) have been fabricated by both friction

bonding and hot isostatic pressing and have nearly completed irradiation to

demonstrate their impact on fuel/clad interface chemistry. Full size monolithic

plates have been fabricated with both types of interlayer by friction bonding and

are currently under irradiation to evaluate mechanical response at prototypic

scale. The plans for future development and qualification are discussed.

1. Introduction

The overall goal of the U.S. National Nuclear Security Administration’s (NNSA) Global Threat

Reduction Initiative is to minimize the use of highly enriched uranium worldwide. As part of

this initiative, the Reduced Enrichment for Research and Test Reactors (RERTR) program has

been charged with developing the nuclear fuels necessary to enable the conversion of civilian

research and test reactors. The program began development of dispersion type uraniummolybdenum

(U-Mo) based fuels in the early 1990’s. Although early testing demonstrated

very promising results, high power and burnup testing on U-Mo dispersed in aluminium

revealed that the fuel/matrix interaction product was prone to the formation of large fission gas

bubbles. Formation of these bubbles eventually lead to the onset of breakaway swelling.

Modifications to the fuel design were then sought to improve performance [1]. Adding silicon

to the matrix material was proposed as a way to form interaction products more similar to the

24 of 435

stable materials observed in U 3 Si 2 based dispersion fuel. A second U-Mo based fuel type was

also proposed at this time. The fuel meat was replaced by a solid (or ‘monolithic’) fuel foil that

eliminated the matrix material altogether. This fuel design would substantially increase the

net uranium density of the fuel and would consequently enable conversion of a new group of

reactors. However, implementation of the monolithic fuel form required significant fabrication

development before testing would be possible.

High density U-Mo based dispersion mini-plates (25 mm wide, 100 mm long, and 1.40 mm

thick) were fabricated for testing using standard roll bonding techniques. Mini-plates with the

silicon modified matrix material have been tested extensively at this scale in the RERTR-6

and RERTR-7 experiments. Several matrix materials were tested including Al-0.2% Si alloy,

Al-2.0% Si alloy, Al-6061 (~0.9% Si), and Al-4043 (~4.8% Si). These tests showed that for

silicon compositions greater than 2% a substantial reduction in interaction product thickness

was achieved and that the interaction product was stable under irradiation to very high fuel

phase burnups (>20% total uranium).

Fabrication techniques for very high density U-Mo based monolithic mini-plates were

developed to enable performance testing on the mini-plate scale. Mini-plates were fabricated

by friction bonding and were tested in the RERTR-6 and RERTR-7 experiments. These

experiments showed that the fuel phase remained stable and that the overall fuel performance

was good. However, behaviour similar to that observed in early dispersion tests was identified

at the fuel/clad interface. Although the interaction layer was very thin, void formation was

noted in regions of very high burnup. It was believed that formation of these structures might

weaken the bond strength between the fuel and cladding. Two approaches to improving the

bond behaviour were proposed including the application of a high silicon layer to the fuel/clad

interface (to hopefully yield the same response as in dispersion fuels) and the insertion of a

zirconium diffusion barrier between the fuel and cladding.

2. Recent Advances in Fuel Development

2.1 Fuel Fabrication

The implementation of monolithic fuel designs requires the development and demonstration of

three key fabrication aspects, foil fabrication, interlayer application, and fuel/clad bonding.

Significant advancements in all three areas were achieved in the last year.

In order to further strengthen the fuel/clad bond strength at the end of irradiation, the

incorporation of an interlayer material was proposed to either alter the chemistry of the

interaction product or minimize the amount of interaction. Adding silicon to the U-Mo/Al

interface has been shown to improve the irradiation stability of the interaction product in both

dispersion fuels and in monolithic fuel plates irradiated in the RERTR-7 experiment. A plasma

spray technique was used to apply a thin uniform layer of Al-Si or Si to the cladding pocket

prior to plate assembly thereby making it available in the fuel/clad interface region. The

formation of a U-Mo/Al interaction product could also be prevented by the insertion of a

diffusion barrier material between the fuel and cladding. A thin layer of zirconium has been

applied to the fuel foil during coincident hot rolling of the fuel coupon with a top and bottom

layer of zirconium [2].

Full size U-10Mo foils were successfully fabricated using two different processes [3]. Plate

shaped U-Mo ingots were cast at the Y-12 National Security Complex to simultaneously

dilute, alloy, and homogenize the fuel material. This plate was then hot rolled or machined to

an intermediate thickness (5.08 mm down to 2.29 mm) that was suitable for final reduction.

The plate was then sectioned into smaller coupons to simplify cold rolling into individual thin

25 of 435

foils (nominally 0.25 mm to 0.38 mm thick). The second process demonstrated at INL started

with the same Y-12 coupons and used a canned hot roll to enable interim annealing steps.

The resulting product from each process showed distinct differences that impacted

downstream processing. The grain structure of the cold rolled foils was equiaxed in nature

and the foils behaved in a very ‘soft’ manner. Alternatively, the grain structure in the hot rolled

foils was elongated in the rolling direction and the foils were stiffer and more brittle. These

properties proved to be important during subsequent friction bonding [4].

Several full size fuel plates (roughly 600 mm x 50 mm x 1.27 mm) were fabricated for

irradiation testing using the friction bonding process. Meaningful advances were made in the

design of the friction bonding tool piece and in the definition of critical process parameters.

These advances played a significant role in enabling the fabrication of two plates (without

interlayers) for ATR-Critical facility tests and two plates for the AFIP-2 irradiation experiment

in ATR. The AFIP-2 experiment consists of one fuel plate with a silicon enhanced fuel/clad

interface and one plate with a zirconium diffusion barrier between the fuel and clad. It was

observed during this fabrication campaign that hot rolled foils, which were more brittle, were

more likely to fracture and flake during friction bonding while the cold rolled foils, which were

softer, were more likely to deform and move in the cladding pocket during friction bonding. It

is believed that an optimum condition may lie somewhere in between these extremes.

Although the program is currently focusing most of its resources on development of the

monolithic fuel form, progress is still being made in the development of U-Mo based dispersion

fuels. Mini-plates (25 mm x 100 mm x 1.4 mm) were fabricated at 8.5 g U/cc loadings with

various high silicon matrix materials including Al-4043 (~4.8% Si), Al-2 Si alloy, and Al + 2 Si

mixture for testing in the RERTR-9A/B irradiation experiment in the ATR. Several full size

plates were also fabricated at BWXT following process development at ANL at >8.0 g U/cc

with Al-4043 and Al-2 Si alloy matrix materials.

2.2. Fuel Performance

The second key area of the fuel development program is fuel performance testing and

characterization. Three irradiation campaigns were completed in the last year, the RERTR-

7A, RERTR-7B, and RERTR-8. These experiments have provided the opportunity to further

assess the behaviour of U-Mo fuels under irradiation and to demonstrate the performance of

other key aspects of fuel design and fabrication.

The first mini-plates fabricated by hot isostatic pressing were irradiated in the RERTR-8

experiment [5]. The irradiation behaviour of the mini-plates was generally good and was

consistent with that of friction bonded fuel plates. The bond between the fuel and cladding

appeared robust and remained intact throughout irradiation. Fuel/clad interface behaviour

similar to that of the friction bonded fuel plates was observed (where small voids were seen in

the interaction product that formed between the fuel and cladding). Surface corrosion on the

cladding was comparable to that observed in both roll bonded dispersion fuels and friction

bonded monolithic fuels.

Additional understanding of the fission product retention and swelling characteristics of U-Mo

fuel was gathered through additional testing and modelling. Fuel plates were irradiated to

peak burnups in excess of 22% total uranium (fission density of approximately 8x10 21 f/cm 3 ) in

the RERTR-8 experiment. These tests showed that the fuel swelling rates remain consistent

with that of the recrystalization phase and that the threshold for the onset of breakaway

swelling has still not been reached. A breakthrough was also achieved in the ability to model

fission product swelling. Fracture surface specimens were examined by scanning electron

microscopy and the intergranular fission gas bubble size distribution for U-Mo fuels was

26 of 435

established. When coupled with recent transmission electron microscopy work [6] that

established the size of intragranular fission gas bubbles, a model to predict fission product

swelling [7] was developed and validated through the first stage of fission product swelling (up

to roughly 3x10 21 f/cm 3 ).

Additional analysis was also performed in order to evaluate the impact of silicon on the

interaction products that form at U-Mo and aluminium interfaces. Small punchings (~1 mm in

diameter) were removed from U-Mo dispersion fuel plates irradiated in the RERTR-6 campaign

and examined using scanning electron microscopy [8]. The fuel plates sampled contained Al-

0.2% Si and Al-4043 (4.8% Si) matrix materials. The examinations showed that the very thin

interaction layers associated with the higher silicon matrix materials was comparable in

thickness to the as-fabricated interaction layer thickness. It was also shown that the

interaction layer observed through x-ray mapping contained an appreciable amount of silicon.

It is believed that the presence of this silicon simultaneously limited the interaction product

growth and increased its irradiation stability. These observations are expected to translate

readily to the fuel/clad interface behaviour in monolithic fuels.

3. Results and Discussion

A significant amount of testing is necessary to achieve the goal of delivering a qualified fuel by

the end of 2011. The results from three key irradiation tests in 2008 will be used to evaluate

the readiness of U-Mo monolithic fuels for qualification testing. The RERTR-9A/B mini-plate

experiment will be used to determine the efficacy of fuel/clad interlayers (both silicon

enhanced and zirconium diffusion barriers) to control the formation of detrimental interaction

products. The AFIP-2 and AFIP-3 experiments will be used to evaluate the dimensional

stability of large plates under irradiation. At the conclusion of these tests, the performance of

the fuel will be evaluated and a decision to proceed with element testing will be made. The

first set of elements tested will consist of standard fuel designs (i.e. simple aluminium clad U-

Mo foils with the selected interlayer) and will be the basis of the report submitted to the NRC

for qualification. Additional development will continue in parallel to develop U-Mo based

monolithic fuels with burnable poisons and graded fuel zones (complex fuels). This

development will be reported in an addendum to the original qualification report to expand the

utilization envelope of the U-Mo monolithic fuel.

4. References

1. Lemoine, P. and Wachs. D. M., “High Density Fuel Development for Research Reactors,”

International Conference on Research Reactors: Safe Management and Effective Utilization,

November 5-9, 2007, Sydney, Australia.

2. Moore, G., et al., “Foil Fabrication and Barrier Layer Application for Monolithic Fuels,” 29 th

International Meeting on Reduced Enrichment for Research and Test Reactors (RERTR),

September 23-37, 2007, Prague, Czech Republic.

3. Dunavant, R., et al., “Update on Uranium-Molybdenum Fuel Foil Fabrication Development at

the Y-12 National Security Complex in 2007,” 29 th International Meeting on Reduced

Enrichment for Research and Test Reactors (RERTR), September 23-37, 2007, Prague,

Czech Republic.

4. Burkes, D.E., Rice, F.J, Jue, J.F., and Hallinan, N.P., “Update on Mechanical Analysis of

Monolithic Fuel Plates,” 12 th Annual Topical Meeting on Research Reactor Fuel Management

(RRFM), March 2-5, 2008, Hamburg, Germany.

5. Hofman, G.L., Kim, Y.K., Rest, J., and Robinson, A.B., “Postirradiation Analysis of the

Last High Uranium Density Miniplate Test: RERTR-8,” 12 th Annual Topical Meeting on

Research Reactor Fuel Management (RRFM), March 2-5, 2008, Hamburg, Germany.

27 of 435

6. Van den Berghe, S., Van Renterghem, W., and Leenaers, A., “Transmission Electron

Microscopy Investigation of Irradiated U-7 wt% Mo Dispersion Fuel,” 29 th International Meeting

on Reduced Enrichment for Research and Test Reactors (RERTR), September 23-37, 2007,

Prague, Czech Republic.

7. Rest, J., Hofman, G.L., Kim, Y.S., Shevlyakov, G., “Characterization of U-Mo Fission Gas

Bubbles on Grain Boundaries,” 12 th Annual Topical Meeting on Research Reactor Fuel

Management (RRFM), March 2-5, 2008, Hamburg, Germany.

8. Keiser Jr., D.D., Robinson, A.B., Janney, D.E., and Jue, J.F., “Results of Recent

Microstructural Characterization of Irradiated U-Mo Dispersion Fuels with Al Alloy Matrices

that Contain Si,” 12 th Annual Topical Meeting on Research Reactor Fuel Management

(RRFM), March 2-5, 2008, Hamburg, Germany.

28 of 435





Division of Nuclear Fuel Cycle and Waste Technology, International Atomic Energy Agency

Wagramer Strasse 5, A-1400 Vienna – Austria


Division of Physical and Chemical Sciences, International Atomic Energy Agency

Wagramer Strasse 5, A-1400 Vienna – Austria


In June 2007, the IAEA convened an international meeting of technical experts

from organisations with experience in the design and operation of aqueous

homogeneous reactors (AHRs), solution based fuel handling, radioisotope

production management as well as the recovery of 99 Mo from 235 U fission.

Participants discussed the current technology of AHRs and associated

radiochemical processes for radioisotopes separation; the technical and

economic feasibility of design, construction and operation of an AHR and

radioisotope processing facilities; and identified and defined future lines of

activity where the Agency’s effort will most effectively support related activities in

different member states.

This paper discusses the outcomes from the meeting. Specific detail is provided

on the principal advantages of the technology, as well as the challenges

associated with further development and deployment. The status of solution

reactors for fission-based medical isotope production is presented. A summary

of other areas of potential utilization is also included. Finally, future IAEA plans in

support of further development are presented.

1. Introduction

The use of aqueous homogeneous reactors (AHRs), also called solution reactors, for the

production of fission-based medical isotopes is potentially advantageous because of their

relatively lower cost; small critical mass; inherent passive safety; and simplified fuel handling,

processing and purification characteristics. These advantages stem partly from the fluid nature

of the fuel and partly from the homogeneous mixture of the fuel and moderator in that an AHR

combines the attributes of liquid-fuel heterogeneous reactors with those of water-moderated

heterogeneous reactors. If practical methods for handling a radioactive aqueous fuel system

are implemented, the inherent simplicity of this type of reactor should result in considerable

economic gains in the production of fission-based medical isotopes. In June 2007, the IAEA

convened a meeting of 10 technical experts from 7 institutions in 5 countries to review all the

relevant issues and make recommendations for future work and this paper presents the output

of this meeting.

2. Advantages of homogeneous aqueous reactors for the production of

fission-based medical isotopes

2.1 Reactor design flexibility and inherent nuclear safety characteristics

The flexibility of solution reactor design parameters is an important feature of the AHR

concept that allows customized design configurations to satisfy safety requirements and meet

29 of 435

or exceed isotope-production targets. The greater flexibility afforded by solution reactors with

respect to core operating power range is an important advantage with respect to 99 Mo

production demand. Solution reactors for isotope production could range from 50 to 300 kW.

The choice of fuel base and solution composition is contingent on core design, operating and

product isotope processing strategy. Traditionally, uranyl-sulfate fuel was preferred over

uranyl-nitrate because of its greater radiation stability. However, the distribution coefficient for

99 Mo extraction is higher from irradiated uranyl-nitrate solutions than from irradiated uranylsulfate

solutions; consequently a nitrate fuel base is clearly more advantageous from a

processing yield point of view. The fuel concentration is selected to minimize core

volume/fissile mass, optimize processing efficiency, or both. Solution reactors are typically

operated at 80°C and slightly below atmospheric pressure. The low operating fuel-solution

temperature, power density, and pressure provides thermodynamic stability, minimizes

potential safety risks and yet allow for sufficient flexibility to optimize 99 Mo production


The inherent nuclear-safety characteristics of solution reactors are associated with the large

negative density coefficient of reactivity in such systems. The reactivity effect resulting from

the operation of solution reactors at power may be thought of as the superposition of two

effects, namely: (1) an overall uniform volumetric expansion of the fuel solution due to the

increase in fuel temperature and the formation of gas bubbles due to radiolysis; and (2) a

corresponding density redistribution within the expanding volume in which the introduction of

an equivalent void volume displaces fuel from regions of higher reactivity worth to regions of

lower reactivity worth. The resulting density reduction is manifested in a large negative

coefficient of reactivity which provides a self-limiting mechanism to terminate a reactivity

excursion and provides inherent nuclear safety. Relevant experiments in the French CRAC

and SILENE facilities have demonstrated these phenomena.

2.2 Efficient neutron utilization, elimination of targets, less post-processing

uranium generated per curie of 99 Mo produced, and overall simpler

waste management

A unique feature of using the solution reactor for fission-based medical-isotope production

compared to conventional production is that the reactor fuel and target are one, consequently

a solution reactor can produce the same amount of 99 Mo at 1/100th the power consumption

and waste generation. Thus the potential advantages of utilizing solution reactor technology

are lower reactor power, less waste heat, and a reduction by a factor of about 100 in the

generation of spent fuel when compared with 99 Mo production by target irradiation in

heterogeneous reactors.

When one considers waste management in terms of both spent-reactor-fuel and spent-target

disposition, waste management for the solution reactor is far simpler. A solution reactor has

no need for targets and, therefore all processes related to the fabrication, irradiation,

disassembly and dissolution of targets are eliminated. Because these target-related

processes result in the generation of both chemical and radioactive wastes, 99 Mo production in

solution reactors can significantly reduce waste generation. Since the recovery and

purification of 99 Mo from conventional targets after dissolution will be quite similar (if not

identical) to that of a solution reactor, the solid and liquid wastes produced will be similar,

except for uranium disposition. Uranium from the solution reactor is recycled and only

disposed at the end of the fuel solution’s viability (up to twenty years).

2.3 Efficient processing of other isotopes using off-gas extraction

In addition to 99 Mo, other radioisotopes used by the medical community can be processed

more efficiently from a solution reactor. Radiolytic boiling enhances the off-gassing of volatile

fission products from the fuel solution into the upper gas plenum of the reactor. A number of

valuable radioisotopes such as, 133 Xe and 131 I, can be recovered from the off-gas. There is a

30 of 435

large demand for 131 I, as it continues to be widely used for therapy of thyroid disorders.

Further, higher specific activity achievable in the off-gas recovery makes it much more

effective for radiolabelling, compared to traditional uranium target irradiation technology. 89 Sr

and 90 Y are two more products of interest for similar recovery due to their proven therapeutic

utility and increasing demands, in particular for 90 Y. While the conventional source of 90 Y is

from a radioisotope generator housing the long-lived 90 Sr separated from the waste stream of

reprocessing plants, the AHR approach could be a potential new source for direct recovery

from irradiated uranium salt solution..

2.4 Less capital cost and potential lower operating costs

The core cooling, gas management, and control systems and auxiliary equipment will be

relatively small and simple compared to current research reactor target systems due to the

lower power of solution reactors. Isotope separation, purification and packaging systems

should be very similar to current target system facilities. The relatively smaller, less complex

solution reactor will be less costly to design and construct than traditional research type


Operating costs may be reduced through many of the improvement mechanisms mentioned

above. Specifically a target-free process eliminates all related costs, including the costs of

target waste handling and disposition. Any resources involved in the transport of the irradiated

target to a processing facility will be saved as will product losses due to any intermediate

cooling periods. Reactor control and operation is expected to be simpler potentially resulting

in reduced staffing requirements.

3. Design Challenges

Although AHR technology is well characterized in the research environment, the capability of

a solution reactor to perform a medical-isotope production mission in a long-term continuous

steady-state mode of operation in the 100 to 300 kW range is not guaranteed. Specifically,

many technical challenges must be addressed in transitioning the technology to a

commercial industrial environment.

3.1 Isotope separation technology

Solution reactor operation for medical isotope production could be challenged by the chemical

stability of the fuel solution induced by a high radiation environment without introducing new

undesired complex chemical structures in the product isotope and/or chemical reactions with

the solution being processed. Furthermore, the potential problems caused by the build-up of

adsorption and fission products, their effect on reactor operation, and the subsequent recovery

system is another challenge which must be addressed. In addition, the effects of build-up of

corrosion products, materials leached from the recovery system, and chemical additions must

also be analysed and optimized. If the fission product build-up and/or corrosion product effects

are important, a means to clean up the fuel solution in concert with waste-management and

economic considerations must be devised.

Another important effect that has not been fully characterized is the effect of molybdenum

redox chemistry of high radiation fields that will accompany fuel cooled for less time than

current practices. Because recovery is based on maintaining Mo in the (VI) oxidation state, its

reduction to lower oxidation states would diminish both its sorption in the loading phase and

it’s stripping from the column in alkaline solution, where the lower oxidation-state Mo species

precipitate in the column as hydrous metal oxides. Limited studies have shown that four hours

after irradiation, effects are seen by lowering of 99 Mo distribution ratios, especially in sulfate

media. Much more experimental work is required to understand and design for this effect.

31 of 435

3.2 Design optimisation

Several design parameters must be optimised during any specific design process. Two fuel

solutions are currently being considered for solution reactors dedicated to radioisotope

production, namely, uranyl-sulfate and uranyl-nitrate. As described above, sulfates facilate

easier reactor operation while nitrates tend to optimise 99 Mo recovery. Also the selected

uranium concentration in the fuel solution is a compromise between reactor optimization and

99 Mo separation efficiency. A lower uranium salt concentration in the fuel solution results in a

larger Kd for Mo(VI) and therefore a more effective and efficient recovery of 99 Mo. As a result,

the size of the recovery column can be smaller making washing of impurities more effective

and obtaining a more concentrated product solution of the raw molybdenum from the column.

However, a higher concentration of uranium in the solution will minimize the reactor fuel

solution volume leading to a more compact reactor.

3.3 Increasing power beyond current operating experience

Historically, solution reactors have been used either in a research capacity to: (1) study

nuclear kinetics phenomena associated with nuclear excursions; (2) as a neutron generator to

study the effects of irradiation on materials; or (3) to generate radioisotopes. As a result, most

reactor operations were transient in nature, or limited with respect to steady-state operation.

Physically, the radiolysis gas and vapour that form at high power densities create bubbles

that migrate to the surface of the solution. The resulting perturbations at the liquid surface

may cause reactivity variations, as well as waves and sloshing effects making it difficult for the

automatic rod control system to maintain steady state power conditions. These phenomena

are closely related to power density and need to be examined carefully to avoid potential

power instabilities or uncontrolled power transients. The design of the core tank may also

need to be reconsidered. These instabilities, while detrimental to predictable production

operations, pose a relatively small potential hazard provided the reactor vessel design can

accommodate pressure transients due to liquid perturbations. The use of Low Enriched

Uranium (LEU) fuel requires a greater volume of fuel and thus results in an increase in core

solution height which potentially diminishes the reactivity variations induced by perturbation of

the solution surface. Furthermore, a non-cylindrical core tank design would probably attenuate

the instability phenomena, thus further strengthening safety.

3.4 Licensing solution reactors

Since no operating license applications involving solution reactor facilities for isotope

production have been submitted, world-wide nuclear regulatory bodies have not developed

specific, relevant regulations. Hazard analyses for solution reactors have indicated

significantly lower hazard to workers, surrounding populations and the environment than those

reactors currently addressed by regulatory bodies. New regulations appropriately addressing

specific hazards associated with solution reactors for commercial isotope production will be

necessary. Until these regulations are formulated and issued, it may be feasible to address

these facilities in a manner similar to current research reactor standards with appropriate

modifications as needed.

4. Status of solution reactors for fission-based medical isotope production

Medical Isotope Production Reactors are under development in China, Russia and the United

States. Two fundamental technologies have been patented in the US, Europe and Russia.

These are solution reactors using LEU solutions of a) uranyl-nitrate salt and b) uranyl-sulfate

salt as the fuel. The ARGUS reactor, a 20 kW(th), High Enriched Uranium (HEU) solution

reactor has been operated as an experimental development activity by Kurchatov Institute in

Russia. Irradiated solution from this unit was processed to separate and purify 99 Mo to

European and US pharmacopoeia standards. It should be noted that meeting minimum

pharmacopoeia purity requirements alone may not be sufficient for specific formulations used

in the eventual medical imaging procedure.

32 of 435

Fundamental research on hydrated metal oxide sorbents continues both in the U.S. at

Argonne National Laboratory, and at the Kurchatov Institute and Ural Technological University

in the Russian Federation. Three sorbents have been considered for molybdenum recovery:

alumina (the classical inorganic sorbent for Mo recovery from acidic solutions), and two

sorbents specifically designed by Thermoxid (Thermoxid Scientific and Production Company,

Zarechnyi, Russia) for recovering 99 Mo from homogeneous reactor fuel solutions. There could

be scope for also exploring the use of a product called polyzirconium compound (PZC of

Kaken Co., Ltd., Hori, Mito-shi 310-0903 Japan) developed for replacing alumina in 99m Tcgenerators

for low-specific activity 99 Mo,

5. Conclusion

The current technology level is well established within the performed research tests. The next

step is to confirm that this new technology can be used in a day-to-day reliable production

environment. Active participation by both pharmaceutical and commercial nuclear reactor

industries will be necessary in order to successfully develop viable commercial applications of

this technology. While the advantages are numerous, commercial markets must be involved in

the establishment of an evolving technology in place of an existing well developed alternative.

5.1 Principal recommendations

• Formulate a scheme to address R&D needs and launch an IAEA Coordinated Research

Project (CRP) to share information on solution reactors and medical isotope processing


• Complete identified research activities based on documented technical challenges

associated with solution reactor technology, isotope separation technology, commercial

utilisation, economic/market analyses,

• LEU should be considered for all solution reactors for fission-based medical isotope


• Consider a bilateral or multilateral project to develop a prototype solution reactor for the

production of fission-based medical isotopes,

• Involve radioisotope technologists and regulatory and pharmaceutical agencies early in

any design process,

• Consider an IAEA Safety Guide on solution reactors for medical isotope production.

6. References and acknowledgments

As mentioned above, this paper represents the output of an IAEA meeting. Each of the below

participants presented papers during the meeting which will be included in an IAEA TECDOC

report being developed on this topic. The authors wish to acknowledge the participants’ input

and express our appreciation for their support.

Mr. W. Nui

Mr. X. Song

Mr. F. Barbry

Mr. M. M. L. A. Barbosa

Mr. V. A. Pavshuk

Mr. Y. D. Baranaev

Mr. E. Y. Smetanin

Mr. W. E. Reynolds

Mr. G. W. Neely

Mr. G. F. Vandergrift




Netherlands/Tyco Int.

Russia/RRC Kurchatov Inst.






33 of 435




Centre National de l’Energie des Sciences et des Techniques Nucléaires

CNESTEN, Rabat – Morocco

Phone : 00 212 37 81 97 50 - email : dg@cnesten.org.ma


The research reactor facility is located at the Nuclear Research Centre of Maamora

(CENM), located approximately 25 kilometres north of the city of Rabat. This facility

will enable CNESTEN, as the operating organisation, to fulfil its missions for the

promotion of nuclear Science and technology applications in various social and

economic sectors in Morocco, to contribute to the implementation of a national

nuclear power program, and to assist the National Nuclear Authorities in monitoring

nuclear activities for the protection of the public and the environment.

The reactor building includes a TRIGA Mark II research reactor with a nominal

power level of 2000 kW (t), and equipped for a planned future upgrade to 3,000-

kilowatts. This facility is the keystone structure of the Research Centre, which

contains, in addition to the TRIGA reactor, extensively equipped laboratories and

all associated support systems, structures, and supply facilities. The construction

of the Nuclear Centre was carried out in collaboration with AREVA-

TECHNICATOME of France and US GENERAL ATOMICS, and with the support of

the International Atomic Energy Agency.

The CENM with its TRIGA reactor and fully equipped laboratories will give the

Kingdom of Morocco its first nuclear installation with extensive capabilities. These

will include the production of radioisotopes for medical, industrial and

environmental uses, implementation of nuclear analytical techniques such as

neutron activation analysis and non-destructive examination techniques, as well as

carrying out basic research programs in solid state and reactor physics.

The TRIGA Mark II research reactor at CENM achieved initial criticality on May

2nd, 2007 at 13:30 with 71 fuel elements and culminated with the successful

completion of full power endurance testing on September 6th, 2007.

1. Introduction

A 2 MW type TRIGA Mark-II research reactor has been installed at Nuclear Research Centre

of Maamora (CENM), located at approximately 25 kilometres north of the city of Rabat. This

is the first nuclear reactor in the kingdom of Morocco. The reactor will be utilised for

research, manpower training and production of radioisotopes for their uses in medicine,

agriculture and industry. The fuel loading of the reactor started in May 1st, 2007 and the

reactor went critical in the May 02, 2007 at 1330 hours with 71 fuel elements. The reactor

achieved full power (2 MW) level and all the required reactor testing were completed in

September 2007. A key feature of the reactor is that the design has been developed with the

capability of being easily upgraded to a steady state power level of 3 MW.

2. Description of the reactor and design parameters

2.1 Reactor shield

The reactor shield is a reinforced concrete structure standing approximately 9.0 m above the

reactor hall floor. The beamports are installed in the shield structure with tubular penetrations

through the concrete shield and the reactor tank water and they terminate either at the


34 of 435

eflector assembly or at the edge of the reactor core. The reactor core and the reflector

assembly are located at the bottom of a 2.5 m diameter aluminium tank, 8.84 m deep.

Approximately 7.2 m of demineralised water above the core provides the vertical shield. The

radial shielding of the core is provided by 2.5 m of concrete having a minimum density of

2.88 g/cm3, water, ˜21 cm of graphite and 6.3 cm of lead.

The reactor is equipped with a thermal column. The outer face of this thermal column is

shielded by a track-mounted door approximately 1257 mm thick. The door is recessed into

the reactor shield structure, and is flush with the shield structure outer surface when closed.

2.2 Reactor Core

The reactor core is at the bottom of the reactor tank, which has an inside diameter of 2.5 m

and a depth of 8.84 m. The reactor core and reflector assembly is a cylinder approximately

1.092 m in diameter and 0.53 m high. The reactor core consists of a lattice of fuel- Moderator

elements, graphite dummy elements and control rods. The core is surrounded by a graphite

reflector and a 6.3 cm thick lead gamma shield. This entire assembly is bolted to a support

stand that rests on the bottom of the reactor tank. The outer wall of the reflector housing

extends 0.81 m above the top of the core to ensure retention of sufficient water for after-heat

removal in the event of a tank drain accident. Cooling of the core is provided by natural

circulation up to full power level. In case of loss of cooling water in the reactor tank there is a

provision of emergency core cooling system.

The top grid plate is aluminium plate of 3.17 cm thick. There are 121 holes of 3.82 cm

diameter in six hexagonal bands around a central hole for locating the fuel- moderator and

graphite dummy elements, the control rods and the pneumatic transfer tube. There are 6

holes of 1.58 cm near the G-ring of the grid plate for locating and providing support for the

neutron source holder at alternate positions.

A hexagonal section can be removed from the centre of the upper grid plate for inserting

specimens up to 11.2 cm in diameter. Two other sections are cut out of the upper grid plate,

for inserting specimens up to 6.1 cm in diameter.

The bottom grid plate is an aluminium plate 3.17 cm thick which supports the entire weight of

the core and provides accurate spacing between the fuel-moderator elements. The safety

plate of 2.5 cm thick aluminium is provided to preclude the possibility of control rods falling

out of the core.

The active part of each fuel-moderator element is approximately 3.63 cm in diameter and

38.1 cm long. The fuel is solid, homogeneous mixture of U-ZrH alloy containing 8.5% by

weight Uranium enriched to about 19.7% U-235. The H/Zr ratio is approximately 1.65. Each

element is clad with 0.051 cm thick stainless steel can. Two sections of graphite are inserted

in the can, one above and one below the fuel, to serve as top and bottom reflectors for the


2.3 Experimental and Irradiation Facilities

The reactor has extensive experimental facilities. It can be used to provide intense fluxes of

neutron and gamma for research, training and radioisotope production. The experimental

and isotope production facility of the reactor consists of the following:

(a) The rotary specimen rack assembly (Lazy Susan) located in the circular well in the

reflector assembly.

(b) Production of very short-lived radioisotopes is accomplished by a pneumatic transfer

system located in the G-ring of the core.

(c) One central experimental tube (Central Thimble) in the middle of the core (A-ring) for incore

irradiation at the region of maximum neutron flux.

(e) Three radial beamports, one of which pierces the graphite reflector and terminates

adjacent to the fuel.

(f) One tangential beamport.


35 of 435

(g) Other in-core irradiation facilities, such as hexagonal and triangular cut-outs etc.

3. Commissioning of the Reactor

The commissioning Program (CP) was prepared using applicable guidance provided in IAEA

Safety Series No. 35-S2 (Ref. 1) and the USNRC document NUREG 1537 (Ref. 2). The

tests are organized in the following stages:

3.1 Preoperational and pre-fuel loading tests;

Facility systems, auxiliary systems, reactor systems, and physical parameters were tested for

the appropriate operating conditions prior to fuel transfer into the reactor core.

Systems were tested according to designated specifications, when applicable, and

acceptable operation was established before core loading. Facility systems tested include

security, fire, communication, and ventilation systems. Auxiliary systems tested include

radiation monitoring, pool coolant, alarm, and interlock systems. Reactor systems tested are

the control system, and operation of reactor components.

The final preparation prior to loading fuel into the reactor for initial criticality was to inspect

and make dimensional measurements on each UZrH fuel element. The dimensional data for

each fuel element was recorded and will be retained for the life of the facility.

3.2 Fuel loading and low power tests

Certain verifications of instrumentation and control system functions were completed before

initialization of an approach to critical experiment by standard reciprocal source multiplication

factor measurements. The reactor achieved initial criticality on May 2nd, 2007 at 1330 hours

with 71 fuel elements with a reactor just supercritical by an excessive reactivity margin of

$0.042. Reactor configuration at criticality was as follows:

• Sixty four (64) standard fuel elements containing 8.5 wt% U,

• two (2) instrumented fuel elements containing 8.5 wt% U,

• five (5) fuel followed control rods elements containing 8.5 wt% U,

• eighteen (18) graphite reflector elements,

• fissile core mass of 2,653 kg U-235.

After criticality, fuel was safely added to the reactor core to achieve:

• An intermediate core loading of 86 fuel elements,

• calibration of control rods,

• verification of the required shutdown reactivity margin and other tests,

• final operational core loading of 101 fuel elements in preparation for conducting tests

and calibrations at intermediate thermal reactor power levels during the next phase of

the commissioning program,

• the reactivity control system was completely re-calibrated with the final, operational

core loading and the availability of an adequate shutdown safety margin was verified.


36 of 435

3.3 Power ascension and tests at low and intermediate power (

5. Conclusion

In meeting all of the objectives of the commissioning of the reactor, it has been demonstrated

that the CNESTEN Triga Mark II 2.0 MW reactor, with natural convection flow, is safe to

operate at all licensed powers.

Furthermore, this first research reactor will enable CNESTEN to fulfill its missions for

promotion of nuclear technology in the Kingdom of Morocco, contribute to the implementation

of a national nuclear power program, and assist the state in monitoring nuclear activities for

protection of the public and environment.

6. References

[1] Code on the Safety of Nuclear research reactors: Operations, Safety Series 35-S2,

International Atomic Energy Agency (June 1992)

[2] Guidelines for Reviewing and Preparing Applications for the Licensing of Non-power

Reactors, NUREG 1537, United States Nuclear Regulatory Commission (February 1996)


38 of 435



Research Reactor Design & Projects Division,

Bhabha Atomic Research Centre,

Trombay Mumbai 400085, India


Reactor Operations Division,

Bhabha Atomic Research Centre,

Trombay Mumbai 400085, India


India has formulated a country specific three stage nuclear power programme

which is essentially based on the availability of uranium and thorium deposits in the

country. At present three research reactors are in operation at BARC, India. These

reactors provide a wide platform to the scientists for conducting research in basic

and applied sciences & engineering. These reactors also meet the requirements of

radioisotopes. A Fast Breeder Test Reactor in operation at Kalpakkam, India has

provided overall insight into various aspects related to development of the first 500

MW Fast Breeder reactor. Kamini a 30 kW research reactor, at Kalpakkam, uses

U-233 fuel. A thorium fuel cycle-based Advanced Heavy Water Reactor (AHWR) is

being developed at BARC. Construction of a critical facility for experimental study

of core physics parameters of the AHWR has been completed and will be

operational soon. A programme for up-gradation and refurbishment of the 50 year

old Apsara reactor is being undertaken. To meet the increasing needs of research

& radioisotopes, construction of a 30 MW high flux research reactor is planned.

1 Introduction

The first Indian Nuclear Research Reactor went critical on August 4, 1956 and the event

marked the beginning of the success story of Indian Nuclear Programme. India has

formulated a country specific three stage nuclear power programme which consists of design,

construction and operation of Pressurised Heavy Water Reactors (PHWR) in the first stage.

During the second stage Fast breeder reactors would be developed and operated which will

also produce U233 from thorium. In the third stage of the programme, reactors using U233

based fuel would be developed. At present, Apsara, Cirus, and Dhruva research reactors are

in operation at Bhabha Atomic Research Centre (BARC). All three research reactors are

utilised for basic & applied research in science and engineering. These reactors also meet

the requirements of radioisotopes for applications in the fields of medicine, agriculture and

industry. Extensive refurbishment and safety up-gradation of the Cirus reactor was carried

out after more than 35 years of operation. For the ageing Apsara, detailed core design

changes and system modifications have been worked out to convert it into a LEU fuelled core

and to upgrade the reactor. Over 150 research reactor-years of operation has provided

valuable experience in the areas related to design, operation, maintenance of nuclear

reactors. These research reactors have provided valuable experience and inputs for

technology developments for the PHWRs of the first stage of our power programme.

Construction of a critical facility for experimental study of core physics parameters of the

Advanced Heavy Water Reactor (AHWR) has been completed and preparations for its first

criticality are underway.

2 Apsara

Apsara is a swimming pool type, light water cooled and light water moderated research

reactor with a maximum thermal neutron flux of 1.0 x 10 13 n/cm 2 /s at the rated power of 1.0

MW. The fuel used is High Enriched Uranium (HEU). The core is suspended from a movable

39 of 435

trolley in a pool 8.4 M long, 2.9 M wide and 8 M deep filled with de-mineralised light water.

The reactor core is supported by an aluminium grid plate having 49 positions on a 7 x 7

lattice for fuel elements, control elements, reflectors, irradiation holes, neutron source and

fission counter. Four cadmium rods function as control rods. Three of these rods serve as

coarse control rods and are also used to shut down the reactor. The fourth one is used to

regulate the reactor power.

3 Cirus

Cirus was the second research reactor built in India. Cirus uses natural metallic uranium as

fuel, heavy water as moderator and light water as a coolant. Cirus has a maximum thermal

neutron flux of 6.7 x 10 13 n/cm 2 /s. The fuel assemblies are placed in a vertical aluminium

reactor vessel having 199 lattice positions. Demineralised light water circulated in a closed

loop is used as the primary coolant. In case normal cooling circuit is not available shutdown

core cooling is ensured by gravity flow of water from a water storage tank, called “ball tank”.

This ensures shutdown core cooling for about 72 hours. Sea water is used as the secondary

coolant. The reactor is housed in a steel containment building.

4 Dhruva

Dhruva is a 100 MW (th) tank type high flux research reactor with natural metallic uranium as

fuel and heavy water as coolant, moderator and reflector. The maximum thermal neutron flux

is 1.8 x 10 14 n/cm 2 /sec. The reactor core is contained in a cylindrical stainless steel vessel

which is placed vertically in a light water filled stainless steel lined vault. There are a total of

146 lattice positions in the reactor vessel, out of which normally 127 positions are used for

loading the fuel assemblies and 9 positions contain the cadmium shut-off rod. The remaining

positions are used for isotope production and experimental facilities. Heat generated in the

fuel assemblies in the core during operation is removed by the heavy water coolant

circulated by main coolant pumps. For shutdown core cooling three auxiliary coolant pumps

are provided and each auxiliary pump has two prime movers (one an electric motor with

uninterrupted power supply and other a turbine driven by gravity flow of water). Also an

Emergency Core Cooling System is provided to take care of Loss of Coolant Accident. The

reactor and associated systems are housed in a rectangular concrete containment building.

5 Utilisation of Apsara, Cirus and Dhruva

All the three research reactors have been well utilised for basic and applied research,

neutron radiography, nuclear detectors testing, radioisotope production, material testing,

shielding experiments and human resource training and development.

The National Facility for Neutron Beam Research (NFNBR) has been created at BARC to

cater to the needs of the Indian scientific community. Scientists from universities and national

laboratories also use these facilities in research reactors through collaborative research

projects. Many of these collaborations are being supported by University Grant Commission-

DAE Consortium for Scientific Research (UGC-DAE CSR), Board of Research in Nuclear

Sciences (BRNS), and other agencies. These research reactors are also utilised for

conducting various engineering experiments. Some of the important experiments performed

in these research reactors are listed here:

• Experiments with different combinations of shield models were carried out at Apsara for

optimising the in-core shielding of the intermediate sodium heat exchanger of Prototype

Fast Breeder Reactor. Results obtained from these experiments have also been utilised

for validation of various computer codes used for shielding calculations. A large number

of shielding experiments were also carried out for radiation streaming studies through

penetrations and ducts of various shapes and sizes for the proposed AHWR.

• Flow pattern transition instability studies were carried out in Apsara by constructing a

loop similar to the geometry of AHWR coolant circuit. The neutron radiography facility

40 of 435

was utilised to visualise flow pattern and also to measure void fraction which is an

important parameter causing the flow pattern transition.

• Irradiation of various biological samples like plants, seeds, etc. was carried out in Apsara.

The experiments carried out at Apsara in the field of biosciences relate to studies on

different biological crop plants and ornamentals. These experiments have helped in the

development of high yielding varieties both in food crops and in ornamentals.

• Towards development of Mixed Oxide (MOX) fuel, UO 2 -PuO 2 fuel pins were test

irradiated for stipulated burn up in Pressurised Water Loop (PWL) of Cirus reactor.

Various design and manufacturing parameters were assessed through these tests.

Towards utilisation of Thoria based fuel in PHWRs, an experimental assembly containing

ThO 2 -PuO 2 fuel pellets was successfully irradiated to a burn up of more than 15000

MWD/Te in PWL. Irradiation of intentionally defected fuel pin was carried out for activity

transport studies. These studies have contributed significantly to the development of Nat

U oxide and Nat U-Pu MOX fuels for power reactors.

• Zircaloy calandria tubes manufactured by different routes were test irradiated in Dhruva

reactor to study their comparative in-pile growth behaviour. Assessment of radiation

induced creep of Zirconium materials has been carried out along with radiation

embrittlement studies of various structural materials used in Indian PHWRs. These

studies resulted in finalisation of manufacturing route for the PHWR pressure tubes and

calandria tubes

• Towards assessing the adaptability of the neutron noise measurement technique for

monitoring healthiness of the in-core components in PHWRs, an experimental assembly

with number of self-powered detectors was irradiated in Dhruva.

6 Safety Management of Research Reactors in BARC

Principal aim of research reactor safety is to keep radiation exposure of plant personnel and

members of public as low as reasonably achievable under all operational states and accident

conditions. To achieve this, the design process incorporates defence in depth philosophy

through multiple levels of protection. To ensure that the research reactors are operated

within the design limits and provisions for safe operation made in the design, do not degrade

during the life of the research reactor, a safety management system is established.

A well structured organisational set-up with clearly defined roles and responsibilities of its

constituents is an important ingredient of safety management system. There exists a well

defined hierarchical structure and line of communication, authority and regulation among

operating organisation, regulatory agency, health and safety organisation, maintenance and

services organisation and quality groups, and experimenters, to facilitate smooth and safe

functioning of the research reactors at BARC.

Documentation forms a vital part of the operational safety management of our research

reactors. The documents such as Design Basis reports, Safety Analysis Report, Technical

Specification, Quality Assurance manual, In-Service Inspection Programme, Emergency

Operating Procedures, Radiation Emergency Procedures, Plan for the regular emergency

exercises and tests, Operation & maintenance procedures for normal operation, etc. form

part of regular operating documents. Strict adherence to the technical specification for

operation ensures operational safety of the research reactor.

Approved Emergency Operating Procedures for postulated off-normal conditions are kept

available in the respective control rooms. The number of such procedures is kept to a bare

minimum to avoid dilution of their significance.

A detailed radiation emergency preparedness plan is prepared bringing out the

responsibilities of various agencies and the follow-up actions required are unambiguously

laid down.

41 of 435

Occupational health and safety is given prime importance to ensure that the radiation

exposure of plant personnel, members of public and environment is kept well within

prescribed limits and as low as achievable.

Area radiation monitors are provided in each research reactor in various areas of the plant

and the status is displayed in the control room. The areas around the plant are monitored by

periodically collecting samples of air, soil and plantation. All radiation workers are monitored

to keep their radiation exposure well within stipulated limits.

A comprehensive quality assurance programme covering all the operational and

maintenance activities is implemented to strengthen the safety culture and for enhancement

of safety by assessment of operational performance. Periodic Internal Regulatory Inspections

are carried out by services agency which is not reporting to O&M.

7 Ageing Management and Safety Upgrades

Ageing management aims at identifying refurbishment requirements and retrofit upgrades

that need to be implemented to qualify systems, structures and components to current safety

standards. After over 30 years of service, signs of ageing started showing in Cirus resulting

in its reduced availability and excessive efforts for maintenance. Detailed ageing studies

were conducted in a systematic manner for all systems, structures and components. Based

on these studies, refurbishment requirements were identified and refurbishing outage of the

reactor was taken from end 1997. After unloading of fuel from the core, further inspections

were carried out. Extensive refurbishing was then carried out and the reactor was made

operational again in October 2002.

Cirus Refurbishment

Ageing assessment mainly consisted of Samples/Coupons testing for material degradation,

Non Destructive Examination using various techniques such as Visual and Remote visual

inspection, Ultrasonic Tests, Eddy current tests, Radiography, etc. and Integrity tests such as

pressure testing, leak checking etc. This was carried out in two phases, in the first phase the

assessment that could be done with reactor in normal operating condition was completed. In

the second phase, the assessment that needed defuelling of core and/or draining of process

fluids was taken up. Few of the important works carried out are described briefly in the

following paragraphs.

Reactor Vessel: Visual inspection of the Reactor Vessel (RV) tubes and their eddy current

testing for wall thickness monitoring and volumetric examination for flaw detection was done.

Condition of the tubes had been found to be good and no unacceptable flaws were detected.

The expansion bellows of RV joins the vessel shell to top tube sheet by helium tight lap weld.

The fluctuating stresses in the bellows were assessed to be well below the endurance limit of

the material. From these studies it was concluded that there was no necessity to replace the

Reactor Vessel.

Graphite reflector: The air cooled graphite reflector around the RV is in two annular

segments and undergoes concurrent annealing with reactor in operation dissipating heat to

coolant air flowing between the inner and outer segments. A thermal safety analysis was

carried out by developing a computational model for predicting steady state and transient

temperatures of the graphite reflectors. Experiments were then carried out at different power

levels and the predictions were found to be in excellent agreement with the experimental

observations. Sample blocks were removed from the reflector and estimations of stored

energy using Differential Scanning Calorimetery was carried out. These studies showed that

the stored energy levels were within acceptable limits and there was no requirement of

carrying out annealing operation.

42 of 435

Flanged joints between aluminium extension pipes of RV and system SS piping: There

are nine flange joints with elastomer gaskets between the aluminium tubes extending from

the top of the reactor vessel and SS helium system pipelines, located above the top of upper

steel thermal shield. Leakages observed during reactor operation were arrested by installing

sealing clamps using remote handling gadgets.

Primary Coolant System Piping: The condition of inside surface of the coolant system

piping was assessed by metallurgical examination of a sample piece and was found

satisfactory. A new protective coating was applied in two layers on external surface of the

underground portion of piping. Couplings were separately coated with mastic compounds.

Civil Structures: All major civil structures were inspected. The tests comprised of visual

examination, Core sample analysis, Ultrasonic pulse velocity tests, Corrosion potential tests

and Rebound hammer tests. General condition of the Reactor building, Annulus building,

Reactor structure block, wet storage block, Reactor ventilation exhaust stack and dump

tanks of main coolant system was found to be satisfactory.

Water seepage noticed from ball tank was repaired by lining the inside of the tank with glass

fibre and epoxy. The ball tank was also qualified to meet the present seismic requirements

by incorporating additional reinforcements in the central shaft region.

Safety upgrades: A detailed seismic analysis of Reactor Containment Building, Ball tank,

Dump tanks, etc was carried out and was found that these structures meet the current

seismic standards. Physical separation of some of the safety related components was

carried out to guard against common cause failures due to fire, flooding etc.

Emergency ventilation exhaust system of the reactor building was earlier provided with an

alkali scrubber and silver coated copper mesh filters for trapping radioiodine under accident

conditions. These were replaced by the more efficient and easy to maintain combination

filters made of activated charcoal and high efficiency particulate air filters.

Desalination unit: A low temperature vacuum evaporation process based desalination unit

of 30 Te/day capacity, developed by the Desalination Division of BARC, has been coupled to

the reactor. This is done to serve as a demonstration of using low grade heat from a

research reactor for the purpose of desalination. The product water is being used to meet the

water requirements of the reactor.

8 Apsara Upgradation

Apsara has been in operation for the last five decades and an extensive refurbishment of the

reactor is planned to extend its useful life and also upgrade the reactor systems in line with

the current safety standards wherever possible.

As a part of up-gradation, it is planned to replace the existing HEU fuelled core by a LEU

fuelled core designed to operate at a higher power of 2.0 MW (th). This will enhance the

maximum available thermal neutron flux to 6 x 10 13 n/cm 2 /s. The refurbished reactor will thus

provide enhanced facilities for studies related to material irradiation, shielding studies,

isotope production, neutron detector testing etc. The core of the upgraded reactor will be

mounted on a grid plate having 64 positions arranged in 8 x 8 square array. The reactor core

consists of 11 standard fuel assemblies, 4 control fuel assemblies and one water hole for

irradiation/experiments. The core is surrounded by BeO assemblies which act as the

reflector. 8 irradiation positions are located in the reflector region. The upgraded reactor will

use U 3 Si 2 -Al dispersion fuel with enrichment limited to 19.75 % w/w U 235 .

9 Critical facility for AHWR

A low power Critical Facility is under construction as a part of the over-all technology

43 of 435

development program to support the design effort essential for evolution of new nuclear

reactor systems utilising the abundant reserves of thorium available in our country. As a step

in this direction, conceptual design and technical feasibility of the thorium fuel cycle-based

AHWR has been established and its detailed design is in progress.

The Critical Facility has been designed to facilitate study of different core lattices based on

various fuel types, moderator materials and reactivity control devices. The reactor is

designed for a nominal fission power of 100 W with an average flux of 10 8 n/cm 2 /sec. The

reactor can be operated at higher power levels of upto 400 W to obtain a neutron flux of 10 9

n/cm 2 /sec for short durations. One of the main features of the reactor is variable lattice pitch

which provides flexibility to arrange fuel inside the core in a precise geometry at the desired

pitch. For the initial set of experiments heavy water is used as the moderator and the

reflector. Reactor criticality is achieved by the manual control of moderator level in the core.

10 Multi Purpose Research Reactor

In order to meet the large requirement of high specific activity radioisotopes and to augment

the research and irradiation facilities available in the country a new Multi Purpose Research

Reactor (MPRR) with enhanced neutron flux is planned to be built.

MPRR is a 30 MW (th) research reactor with a maximum thermal neutron flux of 6.7 x 10 14

n/cm 2 /sec and fast neutron flux of 1.7 x 10 14 n/cm 2 /sec. The reactor is fuelled with Low

Enriched Uranium dispersion type fuel and uses light water as coolant and moderator. An

annular heavy water reflector tank surrounding the core provides highly thermalised neutron

flux region over a large radial distance from the core. The maximum thermal neutron flux

available in the reflector region is 3.5 x 10 14 n/cm 2 /sec. Most of the irradiation positions are

accommodated in the heavy water reflector tank. The core is cooled by light water flowing

from bottom to top across the core, which in turn is cooled by cooling tower water

recirculated in a closed loop. Reactor heat will be ultimately rejected to the atmosphere

through a cooling tower. A set of auxiliary pumps will be provided to remove the core decay

heat in case of failure of the main power supply or unavailability of the main coolant pumps.

Provision has been made for natural convection cooling of the core in the event of nonavailability

of both main and auxiliary pumps or during prolonged outage of the reactor.

There are five in-core irradiation positions and fifteen positions in the reflector region for

radioisotope production and material irradiation studies. Two fuel test loops, one cold

Neutron Source, five tangential beam tubes, two Pneumatic carrier facilities and two

positions for Neutron Transmutation Doping are located in the reflector region.

11 Fast Breeder Test Reactor

The Fast Breeder Test Reactor (FBTR) is a 40 MW (th), loop type, sodium cooled fast

reactor. FBTR uses a mixture of plutonium carbide and natural uranium carbide as fuel.

Heat generated in the reactor is removed by two primary sodium loops, and transferred to

the secondary sodium loops. Each secondary sodium loop is provided with two once-through

steam generator modules. The principal material of construction used for the reactor and

coolant circuits is Stainless steel (SS 316).

The reactor has been utilised for studying the irradiation creep behaviour of Zr-Nb. being

used in the Indian Pressurised Heavy Water Reactors. The present mission of FBTR is to

irradiate the MOX fuel (29 % Pu0 2 ) chosen for PFBR to the target burn-up of 100 GWd/t. In

the coming years, FBTR will be deployed for irradiation of advanced structural materials

contemplated for future fast reactors. The experience in construction, commissioning and

operation of FBTR for 20 years has provided sufficient feedback to enable the launch of the

Prototype Fast Breeder Reactor Project.

44 of 435

12 Kamini

Kamini reactor built in 1996 is probably the only reactor operating with Uranium-233 fuel. It is

tank type reactor with a power of 30 kW and maximum thermal neutron flux of ~ 10 13 n/cm 2 /s.

The reactor fuel is an alloy of Uranium-233 and aluminium in the form of flat plates. The

plates are assembled in an aluminium casing to form the fuel subassemblies. Demineralised

light water is used as moderator, coolant as well as shield. Cooling of the reactor core is by

natural convection. Provision has been made for cooling this water to maintain the water

temperature at a steady value when the reactor is operated for long durations at higher

powers. Start up and regulation of the reactor is done by adjusting the positions of two safety

control plates made of cadmium sandwiched in aluminium. The reactor is mainly used for

neutron radiography for fast reactor fuel development.

13 References

1. India’s First Research Reactor APSARA, Fifty Years of Operation and Utilisation,

Bhabha Atomic Research Centre, 2006

2. Dhruva Safety report volume I, II 1984.

3. Twenty years of FBTR, B Rajendran, National conference on Operating Experience

of Nuclear Reactors and Power Plants, Nov 2006

45 of 435




NRG, PO Box 25, 1755ZG Petten, The Netherlands


At present the High Flux Reactor, HFR, Petten, is involved in fission and fusion

power plant research and development, and carries out isotope production for

medical and technical applications. The power plant research and development

will continue to be focussed on new materials and components for higher

reliability and efficiency of Generation-4 fission power plants, and fusion power

plants. The HFR will be more than half a century old in 2015. To address these

future research requirements a replacement is inevitable.

The test reactor building industry is presently producing conceptual designs for

PALLAS for a research reactor in combination with isotope production. The

business plan has been drafted, supporting a design with the main parameters:

a 30-80 MWth flexible, reliable reactor core with neutron fluxes up to 5*10 18 n.m -2

at all power levels.

1. Introduction

The world population increases, but the growth might reach zero in this century. Even if

prediction materializes, the need for energy will multiply. The sources will have to change:

fossil fuels will decline sharply before the mid of the century. Depending on the world

development scenarios, [1], the contribution of fission and fusion energy must rise

considerable to satisfy the demand for electricity in the first place. Also in the EU the

predictions for the fission and fusion energy contribution are highly significant [2,3].

The HFR in Petten was designed and built in the fifties of the 20th century for the development

of fission energy. In 1984 the reactor vessel replacement prepared the HFR for the next 30

years of operation. In the period between 2015 and 2020 the second HFR reactor vessel will

near its end of its design life together with other major components that will have to be

refurbished in that period. Therefore, HFR replacement will be more economical.

The preparations for the replacement of the HFR with a new research reactor, PALLAS,

capable of isotope production in parallel, have started. This paper presents the major PALLAS

project steps. First it presents a projection of the demands from the fission and fusion

research areas, and the customers for isotopes. These culminate in the major requirements

for the research reactor. The reactor building industry has started producing a conceptual

design in 2008. The licensing situation for research reactors in The Netherlands will be shortly

addressed, as it sets the boundaries for the design.

2. Prospects for research reactors in the EU

Presently the HFR contributes greatly to the development of fission reactors and since the

latter decades of the previous century it delivers experimental results for the design and

construction of ITER and following fusion power plants. The production of isotopes was always

part of the production palette, but the production followed the increasing demand in recent


46 of 435

The research and development of power plants in the first half of this century will consist of the

development of generation-4 reactors in particular the high temperature reactor. The work will

encompass both structural and functional materials testing en demonstration of endurance

operation of (sub) components such as fuel elements of graphite and erosion and corrosion of

lead based coolants. Table 1 sums the main items for investigation in the next decades for

fission alongside the expected major items for fusion. Recently EFDA defined their missions

to build DEMO a first fusion power plant. These missions include testing of materials and

components in high flux fission research reactors. The 14 MeV neutron testing environment,

provided by IFMIF in the early twenties of this century, will not produce sufficient volume for

component testing. Therefore, component tests can be performed in Pallas to complement

IFMIF irradiations.

Material Fission: Gen-4, HTR Fusion: ITER, DEMO




Pressure vessel steel

Canning steel nano


Graphite, composite

Fuel particle element

Inert Matrix Actinides

Fuel element test in

all conditions

Pb-Bi compatibility

Graphite creep

Low activation steels

ODS steels


SiC composites

Lithium ceramics

Beryllium pebbles

Tritium release

Pb-Li behaviour

Bolt relaxation

First wall simulation

Tab 1: Fission and fusion power reactor research & development

Several studies, FEUNMARR and ESFRI [4,5] were recently completed to provide a roadmap

for test reactor capacity in the EU. In those studies the major test and isotope production

devices the EU needs are RJH, PALLAS, IFMIF and MYRRHA roadmap mission EFDA

science/technology. The demand for isotopes now centers around many, but in medical

isotopes the Mo production forms the majority. This need not be the case for the rest of the

century, but new medical applications of isotopes, lutetium for example, show there is a future

in medical isotope production. Tens of isotopes hold promise for new applications in treatment

and diagnosis. For the technical isotopes the market shows similar movements.

The business case for PALLAS, stretching far into this century, of course holds uncertainties.

At present the combination of research and development for power reactors and isotope

production is the best basis for the sound operation of PALLAS, if the requirements set for the

investment and the operation can be met.

3. Requirements for PALLAS

The main test reactor issues addressing the demands from the major R&D customers are

high neutron fluxes to simulate life times double or three times more quickly than the HFR. In

particular the end of life conditions in Generation-4 permanent structures and components

near the plasma of a fusion power plant require radiation damage in the order of 70 to 150 dpa,

accumulated in three to maximally five years. The production of isotopes has its own drive for

swiftly delivery of the necessary isotopes benefiting from high fluxes. The role of automated

production will have to be increased for higher productivity without compromising the reliability

of the production streams.

47 of 435

A core that can be rapidly re-configured to adapt the irradiation volume needed for each new

cycle satisfies the need for flexible operation. The core will have a nominal power of 40 MWth

with a reconfigurable minimum and maximum geometry allowing reduction and increase in

irradiation volume. These volumes should be made available through a 30 MWth minimum and

80 MWth power core. This flexible core should be well predictable with an appreciable

irradiation volume providing neutron fluxes up to 5*10 18 n.m -2 at all power levels. The resulting

flexibility of the core will allow a more economic use of the reactor fuel and limit the production

of waste to a minimum. This core property is an important issue for the broad support of the

utilization of PALLAS in the public domain.

Reactor coolant

make-up system

Reactor coolant

treatment system

Primary cooling system

Decay heat removal system







Heat sink










Heat exchanger


Emergency core cooling



Fig 1. Lay-out of the cooling systems showing the dedicated secondary system

The secondary cooling system, Fig. 1, could address the requirement not to spoil low

temperature heat generated by PALLAS. Different customers have an interest such as

glass house farmers, fish and shrimp growers, and the mining industry to heat expanding gas

from sources under the North Sea. In public information exchanges and hearings using the

low temperature 30 to 80 MW heat can be an important asset.




Exploitation Investment Rating

Double, threefold flux 2 to 3 times faster Neutral Neutral +++++


Flexible reflector Neutral Lower fuel cost Design scope +++++

Less waste complex

Team workload Neutral Attractive schedules Automated

less night shifts handling

Double hotcell

Automated isotope

production channels



Shorter term


More efficient

More efficient


1 more hotcell ++++


Channel +++


48 of 435

Tank in Plus-pool Smoother logistics Smoother handling Neutral +++

Tab 2: PALLAS advantages

Tab. 2 summarizes the advantages of the PALLAS requirements over the present HFR. The

high fluxes and flexible core arrangements will satisfy both the customers and the operator of

PALLAS. Extra high fluxes might be generated using boosters, but this is only feasible for a

limited number of experiments. Additional advantages, though of a lower overall impact, are

the effect of isotope production automation improving the schedules for the staff. A four lobe

pool concept (with the map of a plus sign) will improve the transfer and storage operations in

the pool. The operation of two separate poolside hot-cells will strengthen the reliability of postirradiation

services. The cells are situated to the left and the right of the lobe for the reactor

core. Each cell has its own lobe, with its own access.

Hot cell



Fuel storage


Transit pool





Hot cell

4. The design and build project

Fig 2. Schematic pool lay-out

JRC-IE, Petten, Mallinckrodt Medical, Technical University Delft, and NRG took the initiative

to replace the HFR with a new research reactor: PALLAS. NRG has the lead in drawing up

the requirements and initiate and manage the project leading to the conceptual design of

PALLAS. The business plan has been drafted to support the design requirements. Special

effort has been devoted to draw up the reactor requirements with optimal operation conditions,

both from the technical, safety and the budget points of view.

The technical specification, including the general layout of the core, main buildings, and

auxiliary equipment forms the basis for the conceptual designs to be provided by the research

reactor builders. The conceptual design phase will prove whether the existing set of the

requirements can be met, and will result in the optimal operation of PALLAS.

The tender approach is completely in line with the EU regulations for a restricted procedure

(procedure with pre-selection) with 3 phases under fair competition. Selected test reactor

builders are presently producing conceptual designs for PALLAS as a research reactor in

combination with isotope production according to the requirements, revised during the

consultation and dialogue phase of the tender procedure. After the delivery of three conceptual

designs a selection will be made for one offer. The phase of the pre-design will reach from

2009 to early 2010, followed (after formal approval) by the detailed design phase lasting to

49 of 435

2011. The building of the reactor would lead to first criticality in 2015, starting full power

regular operation, after a full year of test runs, in 2016.

It is expected that the financing of the detailed design and building will be firmly established in

2009. Prior to that the precise arrangements for reactor ownership, operator, and technical

scientific acquisition will be established.

The licensing will follow the IAEA guidelines [6,7] interpreted for The Netherlands situation in

the Nederlandse Veiligheids Regels, NVR. The present set of NVRs will be updated in the

2009, thus several issues will be discussed. Table 3 gives the five document clusters that

must be treated together with the comments made by NRG. An example of an issue to be

dealt with is the detailing of the requirements for the secondary shutdown system for


Document Cluster

Draft amendments to NS-R-4 by

the Netherlands regulatory body

NRG comments

NS-R-4 is the principal guide for PALLAS. Special

Netherlands requirements have been added.

Tentative list of standards Tentative selection by NRG of applicable (IAEA)

applicable to PALLAS standards.

Safety Codes and Guides in Review of the current NVRs in Netherlands will yield a

current regulation (2007) new set by the end of 2009.

Safety objectives in Netherlands

legislation; dose and risk criteria

Legislative and regulatory


5. Conclusion

Criteria for doses to population, mortality risks, core

meltdown frequency, and special safety requirements

new reactors.

Description of framework including the licensing

procedure and the governmental departments involved.

Tab 3: Research Reactor licensing issues in The Netherlands

The prospects for PALLAS, the contemporary replacement for the HFR, look bright. The

development of the Generation-4 fission reactors and fusion power plants will provide many

opportunities for PALLAS within the roadmap for research infrastructures in the EU.

The main requirements are neutron fluxes double to triple those of the HFR and a flexible core

reducing fuel cost and waste production. Other improvements over the HFR are more

automation in operation, double hot-cells alongside a plus shaped pool.

The tender process has reached the dialogue phase for fine tuning of the specifications

followed by the end of the year the selection of the best conceptual design that must lead to

first criticality of PALLAS in 2015.

The licensing procedure in The Netherlands has to be updated in 2009, that is a concern for

the conceptual design but during final design and building the licence requirements will be


6. References

[1] Energie en Samenleving in 2050, Nederland in wereldbeelden, Ministerie Van

Economische Zaken, Den Haag, 6 december 2002.

[2] F. De Esteban, The future of nuclear energy in the European Union, European Strategic

Exchange, Brussel, 23 May 2002.

[3] A. Bradshaw, Fusion and Energy supply, Contribution to the debate on the Green

Paper, EC, 2002, Brussel.

50 of 435

[4] D.P. Parrat, FEUNMARR Final Synthesis Report, contract nr. FIR1-CT-2001-20122,

October 2001.

[5] European Roadmap for research infrastructures 2006, European Communities,

Luxemburg, 2006, ISBN-92-79-2694-1, page 68.

[6] “Safety of Research Reactors”, International Atomic Energy Agency, Safety Standards,

Safety Requirements No. NS-R-4, Vienna, June 2005.

[7] “The Physical Protection of Nuclear Material and Nuclear Facilities”, International

Atomic Energy Agency, INFCIRC/225/Rev.4, Vienna, 1999.

51 of 435



C. Gonnier, D. Parrat * ; S. Gaillot, J.P. Chauvin, F. Serre,

G. Laffont, A. Guigon, P. Roux

Nuclear Energy Division

CEA Cadarache, France

* DEC/SA3C – Building 315 - CEA Cadarache

F - 13108 Saint Paul Lez Durance Cedex daniel.parrat@cea.fr


After a brief description of the Jules Horowitz Reactor (JHR) facility building status,

this paper will present in a first part the design work carried out on the irradiation


For materials, pre-design studies concern mainly capsules containing NaK, with or

without circulation. This type of device, inserted in the central hole of a JHR fuel

element, ensures very good temperature homogeneity on a batch of samples and a

high dpa rate. For LWR fuels, a set of loops and capsules adapted to PWR and

BWR conditions will fulfil expected needs. As examples, the Adeline loop will be

able to test a single experimental rod up to its operating limits. The Madison loop

will be devoted to long-term testing of up to 4 instrumented fuel rods under normal

conditions. The LORELEI-type capsule will implement LOCA tests.

In a second part, the paper describes the support facilities (laboratories,

examination benches) also present in the JHR and enhancing the quality of the

experiment. The conclusion underlines the international collaboration developed

around the JHR project.


The current development of nuclear energy will face in the first half of this century to a

specific situation characterized by:

• Operation of the standard water reactors up to their end of life, facing to ageing

process on irradiated materials and to maintenance of an expertise capability,

• Progressive commercial operation of new concepts of water reactors, using optimized

fuels and plant cycle management,

• Development of innovative concepts, mainly based on fast neutron systems, either for

energy production (electricity or heat) or for waste management (transmutation),

• Qualification of totally new components or materials for extreme conditions of use,

such as under high neutronic flux of for fusion systems.

To fulfil the experimental knowledge needs coming from the large variety of materials and

irradiation conditions to master, multipurpose research facilities are now key infrastructures

(see ref.[1]), in complement of prediction capabilities gained thanks to progresses in the

modelling. Within this frame, the JHR is designed to offer modern irradiation experimental

capabilities for studying material & fuel behaviour under irradiation, mainly due to:

• High values of fast and thermal neutron fluxes in the core and high thermal neutron

flux in the reflector+ (producing typically twice more material damages per year than

available today in European MTRs),

• A large variety of experimental devices capable to reproduce environment conditions

(pressure, temperature, flux, coolant chemistry…) of light water reactors (LWRs), of

52 of 435

gas cooled thermal or fast reactors, of sodium fast reactors, etc, including the

development of new types of embarked components and instrumentation,

• The possibility i) to test highly instrumented samples under normal conditions and up

to limits, in order to support advanced modelling for giving prediction on a broader

range, ii) to manage degraded fuel samples after soliciting tests (e.g. safety tests),

and iii) to perform a large variety of non destructive examinations on samples quickly

after their irradiation and with a minimum of handling.

The reference power of the JHR is 100 MW. Presentations of this project and of its main

features have already been done in several conferences [2], [3], [4].

The project is now at the development phase since beginning of 2006. After the construction

permit delivery gained in the first half of 2007, excavation works started mid-2007 on the

CEA Cadarache site in the southeast of France. Building construction is planned to start at

the beginning of 2009. The first criticality is expected during the year 2014. The lifetime of the

JHR will be at least of 50 years.

The safety assessment process, which is leading to the licensing of a reactor such as the

JHR, is mainly composed by two phases. The first phase involving the issuing of the

Preliminary Safety Report of the reactor and its analysis by the Safety Authority allows the

beginning of the construction of the Reactor. The second phase, which is constituted by the

issuing of the Provisory Safety Report of the reactor and its analysis by the Safety Authority,

allows the start up of the reactor to reach the first criticality.

At this time the project is ending the first phase of the safety assessment process allowing

the first concrete flooring at the beginning of 2009.


3.1 From the on-going device development to the JHR experimental capability

The design work of the JHR irradiation device park is driven by identified and expected future

experimental needs. The starting of the feasibility and/or the development phases is related

to the maturity of the demand and depends on the complexity of the device to set up.

Consequently the device studies presented in this paper correspond to the current view of

the long-term needs, which will be likely expressed during the coming decades. This

development is a first initiative towards the set-up of the whole JHR experimental device

park. It will also depend on the future irradiation market, and on the strategy applied by the

JHR Consortium members or by the International Joint Program Committee.

3.2 Devices for material studies in Gen II – Gen III conditions

3.2.1 The CALIPSO integrated loop

Experimental needs in the nuclear material irradiation science concern mainly the

characterization and the qualification of new cladding materials. They are characterized by i)

the minimization of the temperature gradients between samples constituting the experimental

batch, ii) the sample temperature stability versus time and iii) the possibility to apply a

controlled stress to the specimen with in-situ measurement of the resulting strain. This last

feature is a challenging technological issue, and is driven by both scientific knowledge (creep

kinetics quantification) and operational stakes (minimization of budget and time to results by

avoiding similar tests in hot cell).

To cope with this trend, the current design work is currently focused on the in-core CALIPSO

NaK integrated loop. Placed in the central hole of the fuel element, this device shall be

53 of 435

autonomous for long-term irradiations and embarks in a small volume all the components

needed to ensure a forced convection in the test section:

• The technological feasibility of the electromagnetic pump for the nominal operating

conditions (flow rate 2 m 3 /h, ∆P = 1,25 bar up to 600°C, with an outer diameter of 80

mm maximum located out of the core region) is now confirmed. Pump characteristics

include margins for sample-holders with a high pressure drop, and will ensure a

maximum axial temperature difference between samples of 7,5°C.

• The heat exchanger, placed under the sample, is designed to remove the gamma

heating deposited in the sample and the device structures. Different lengths will cover

the standard LWR operation temperature range (250°C to 450°C), and the 600°C

point will need a specific design.

• The head of the device holds the equipment box. It shall embark i) a connecting plate

for instrumentation (about 50 signals), ii) the electric power supply of the pump and of

the heater (to compensate if necessary an over-performance of the exchanger), iii)

the control of the gas gap pressures (NaK blanket, external thermal barrier) and iv)

the handling means. The objective is to design an upper head compatible with the

maximum of in-core devices.

The sample–holder designed so far for the CALIPSO loop is based on the “ZO” concept used

in the OSIRIS MTR. Designed firstly for LWR needs, it embarks 3 experimentation bases

holding 3 pre-pressurized tubular samples placed at 120° on each. The device allows gaining

quickly high fluencies thanks to a displacement per atom (dpa) rate up to 15 dpa/year. Strain

measurements are performed after specimen unloading at the intercycle.

The design phase of the CALIPSO loop is now finished and some critical components (such

as the electromagnetic pump or the embarked heat exchanger) have been studied more in

details with the aim to launch very soon the manufacturing of prototypes. A contract for the

detailed design and manufacturing of a CALIPSO prototype is under preparation and will be

normally signed during the second quarter of 2008.

3.2.2 Other types of material experimental devices

A simpler design of the CALIPSO loop is also under study, operated with natural convection

(no pump, but with an electrical heater). Called MICA, it will be able to hold a more

sophisticated and instrumented sample-holder containing about 10 tubular samples placed

vertically in the centre of the device. This sample-holder will allow applying a controlled

biaxial stress on tubular samples (axial and circumferential) and to measure on-line the

resulting strain. A first step towards this design is represented by the CEDRIC sampleholder,

designed to apply a controlled uniaxial stress on specimens made of SiC fibres. Its

operation in a CHOUCA device in Osiris is expected at mid-2008.

Stress corrosion cracking under irradiation in water coolant is taken into account through the

conceptual design of a specific sample-holder allowing in-pile irradiation assisted cracking

growth rate monitoring, thanks to the local electric potential drop measurement.

Other types of devices for material irradiation are planned, and mainly an in-reflector device

capable to irradiate large specimens representative of power reactor pressure vessels.

One can also mention the on-going design of an out-of-pile NaK technological loop, which

will be installed on the Cadarache site, as a common platform to test components belonging

to future irradiation devices.

3.3 Devices for LWR fuel studies

Different types of devices are currently designed, driven by the type of experimental

programme. As a first approach, one can classify the device design according to the

solicitation applied to the fuel sample.

54 of 435

3.3.1 Water reactor fuel studies under nominal conditions

When the fuel rod failure is not an experimental objective or a risk, and when LWR conditions

at the rod level are requested (temperature, pressure and coolant flow rate), the experiment

will be set up preferably in the MADISON water loop. This loop will be put on a moving box in

the JHR reflector, and will be capable to apply PWR or BWR conditions on the experimental

load. This load will be constituted by a sample holder embarking up to 4 instrumented PWR

or BWR-type geometry pre-irradiated fuel rods, with a fissile stack up to 600 mm and

irradiated in an very homogeneous way. The target is to have less than 3% heterogeneity on

the linear heat generation rate (LHGR) between any 2 rods. Of course 2 half-rods can

replace each rod, if comparative or statistical results are a stake.

The standard instrumentation of each rod will be a thermocouple (e.g. for fuel central

temperature measurement) and a Linear Variable Differential Transformer type (LVDT-type)

sensor connected to one end of the rod and measuring on-line a given parameter (e.g. clad

diameter, fission gas release…).

The feasibility study of this loop will be launched in Spring 2008 in collaboration with the

Institute for Energy Technology (IFE), operator of the Halden research reactor (HBWR,

Norway). Programs concerning fuel properties measurement versus burn-up or versus

LHGR, fission gas release, or corrosion studies will be performed in the MADISON device.

Some of these programmes will be long (several years), and the irradiation can be

accelerated compare to power reactors conditions, however with respect of scientific

constraints, in order to gain quickly high burn-ups and to gain knowledge on fuel end-of-life


The MADISON-type concept will likely represent the standard performing and commercially

attractive fuel irradiation service in JHR.

3.3.2 Water reactor fuel studies up to limits and under off-normal situations

Research of fuel product limits (e.g. class 2 ramps, internal over-pressurization, melting

approach…), and post-failure behaviour studies under normal conditions (failed rod

behaviour and fission product release studies), will be carried out in the ADELINE loop. The

pre-design study of this PWR loop, also placed on a moving box in the JHR reflector, will be

completed in Spring 2008. The in-pile part is based on the “jet-pump” flow-rate amplification

system, to minimize the contaminated coolant quantity and flow-rate going to the loop

components located in the experimental cubicle. The device head is designed for

management of a degraded fuel rod, by tight connexion to the JHR alpha hot cell.

The out-of-pile part comprises in particular the fission product and fissile material purification

system (resins, filters and degasser). Fission product concentration in the coolant can be

measured either on-line (by gamma spectrometry or delayed neutron detection) or by

sampling in the fission product laboratory, thanks to a specific line working at low flow-rate (a

few l/h).

The device neutronic and thermal-hydraulical design will offer high performances and a large


• The LHGR value of 500 W/cm with a 1% 235 U fresh UO 2 PWR fuel rod has been

confirmed after pre-design studies.

• The standard power ramp rate will be up to 660 W/cm.min, with accuracy on the

LHGR during the upper plateau, coming from the displacement system position, less

than 5 %

• The inlet coolant temperature will be precisely controlled and will range from 280°C

up to 320°C (from 150 W/cm). The outlet-inlet temperature difference will be of +5°C

maximum, thanks to a high water speed (about 5 m/s).

55 of 435

The sample is one PWR rod (but including a possible diameter evolution up to 12,5 mm).

The reference instrumentation is two sensors as for the MADISON sample. It can also be

gas minitubes for internal free volumes sweeping and routing to the fission product


3.3.3 Water reactor fuel studies under accidental situations

The safety experiments will constitute a key service offer by the JHR. For LOCA-type

experiments, the feasibility study of the dedicated capsule LORELEI has been started from

the end of 2007. The target is to be able to reproduce the typical temperature time history

and the quenching phase of a LOCA sequence on a single instrumented fuel rod, based on a

single-effect approach. The device itself will be heavily instrumented and capable to manage

the post clad burst and the post quenching phases.

As the experimental needs are closely linked to the model prediction capabilities, and as

these experiments are probably the most difficult ones to integrate in the JHR environment,

there is a strategic way for defining, as soon as the device pre-design phase, the future

experimental programmes. For this aim discussions or collaborations are being launched

with utilities or institutes (e.g. EDF, OECD and IRSN).

The current LORELEI design will be able to fulfil a part of the LOCA demand. Other designs

could be set up in a near future, depending on the physical mechanisms to explore. In

particular, it is expected to adapt it for tests on a small bundle, in order to point out some fuel

bundle effects. For these tests, the non-destructive examination benches (see § 4) will be a

crucial support to gain quickly a first detailed status of the tested sample.

The design of a capsule for fast transient implementation is also planned. Based on a singleeffect

strategy, the target is to gain basis data on the activated phenomenology (e.g. fission

gas release).

3.3.4 Other water reactor fuel device studies in progress or planned

The JHR reflector will also welcome simple boiling capsules for one instrumented LWR

experimental fuel rod. Placed either in a fixed location or on a displacement system, it will be

adapted to experiments, which don’t necessitate representative LWR conditions outside the

rod. The natural boiling conditions allow a large place around the rod, and this situation is

favourable to fragile or cumbersome instrumentation (e.g. on-line axial and circumferential

rod diameter measurement by metrology, as carried out by the DECOR sample-holder).

Small irradiation capsules with static gas gap around the sample are also foreseen. This type

of device will be suitable for irradiations on small fuel samples with adapted geometry, for

microstructure selection or material basis data obtaining.

Finally, it is worthwhile to point out that the MADISON-type concept (see § 3.3.1), could

evolved and be adapted to other environment conditions, such for example a unit placed in a

peripheral in-core position.

3.4 Fuel and material device studies for Gen IV power system conditions

Innovative development of a new generation of materials and fuels, which resist to high

temperatures and fast neutron flux in different environments, is necessary for the

development of these future reactors. There is a need to assess the behaviour under

irradiation of a wide range of structural materials such as graphite (VHTR and MSR),

austenitic and ferritic steels (VHTR, SFR, GFR, LFR), Ni based alloys (SCWR), ceramics

(GFR)… These innovative structural materials are often common to fission and fusion

applications. Experimental irradiations have to be carried out in order to study microstructural

and dimensional evolution, but also the behaviour under stress. New fuels for the different

Gen IV systems need also to be characterized or qualified in research reactors.

As the demand is less mature than for LWRs, the on-going studies address three topics:

56 of 435

• Materials behaviour under high temperature conditions: the conceptual design of an

helium gas loop in the JHR core, at high temperature (700-1200°C) and high fast

neutron flux (from 1 to 5 10 14 n/cm²/), has started. This loop will be dedicated to

separate effects experiments on selected materials, such as SiC/SiC or Oxide


• Gas thermal system fuels: This topic addresses high pressure and high temperature

gas rig designed for the irradiation of compact stacks in the JHR reflector. The stack

is swept by an inert gas at low flow rate to route the released fission gases to the

fission product laboratory for quantitative measurements. A feasibility study has been

performed in a European collaboration frame.

• Gas fast reactor fuels: The conceptual design of a gas rig or a gas loop in the JHR

core has started. The chosen design has to cope with JHR constraints and will

depend on the evolution of the demand. For this aim, the experimental feedback

gained from the IRRDEMO experiment planned in BR2 will constitute a great added



The JHR experimental process includes also non-destructive examination (NDE) stands

which aim is to increase the experiment quality through NDE on full devices or sample

holders by:

• Initial check of the experimental load state just before the beginning of irradiation

(after transportation or insertion in the device),

• Adjustment of the experimental protocol after a first irradiation run (sample evolution,

power tuning…),

• On the spot monitoring of the sample state after a test on the close-by stand located

in the reactor pool and with limited handlings (e.g. geometrical changes after an offnormal

transient, quantification of short half-life fission product distribution…).

The design phase of two underwater photonic imaging systems has just started end of 2007

in collaboration with VTT (FI). These systems will be respectively implanted in the reactor

pool (for experiments with short decay or for quick measurements) and in the storage pool of

nuclear auxiliary building (for longer examinations such as tomography). They should adapt

for all sorts of experimental devices, even still lead-connected to ground-based experimental

cubicles for some. Each system will accommodate on the same bench both quantitative

gamma-emission and X-ray transmission scans which will allow performing detailed 3D

images. Full device NDE can also be performed on an underwater neutron imaging bench

installed on the reactor pool flooring.

After extraction from their carrier, samples will be also scrutinized in fuel and materials NDE

hot cells, where one can find multipurpose test benches dealing with examination such as

visual checks, sizing, corrosion thickness measurement, crack inspection, gamma and X-ray

scans etc…



Besides the bilateral collaborations set-up for the development of some equipment already

mentioned in this paper, the JHR facility design and operation is largely open to international

collaboration. A first step was the signature in March 2007 of a Consortium Agreement for

the reactor construction and operation. This consortium associates the European Atomic

57 of 435

Energy Community (Euratom/JRC), European fundamental and applied research institutes

(CEA, CIEMAT, NRI, SCK­CEN, VTT), and two major companies: a utility (EDF) and a fuel

vendor (Areva). India also joined the Consortium in January 2008.

As an important subsequent step, a new FP6 project (MTR+I3, “MTR plus” integrated

infrastructure initiative) has been launched for 3 years from October 2006. This programme

reinforces a major evolution toward the following key objectives:

• Building up the European MTR Community, including new facilities as well as existing

ones (high performance MTRs as well as flexible small power facilities). Special

attention is paid on complementarities between MTRs: operators training with staff

exchanges, manufacturing practices, measurement best practices, opening accesses

for testing experimental devices innovations.

• Establishment of the JHR as a new European MTR, because cross fertilization with

existing European MTRs is important to take advantage of the available experience

and of the impetus provided by the JHR project.

• Support state of the art design, fabrication and test of innovative irradiation devices or

components with associated instrumentation. This addresses a comprehensive set of

topics strategic for both present and future power reactors.

From 2008 is launched an International Joint Program in collaboration with OECD/NEA, for

addressing issues of broad interest among the nuclear community, and gathering industry,

academic institutions, safety bodies and research centres.

Discussions are also in progress with other countries, either through public research

institutes or with industry, for joining the JHR Consortium or for defining a bilateral

collaboration (Sweden, Germany…).


[1] Future needs for material test reactors in Europe (Feunmarr findings)

C. Vitanza, D. Iracane and D. Parrat

Proc. of RRFM 2003, 9-12 March 2003, Aix-en-Provence, France

[2] The Jules Horowitz Reactor, a new Material Testing Reactor in Europe

D. Iracane

Proc. of the TRTR-2005 / IGORR-10 Joint Meeting, Sept. 2005, Gaithersburg, (MA), USA

[3] The Jules Horowitz Reactor: General layout, main design options resulting from safety

options, technical performances and operating constraints

JP. Dupuy et al.

Proc. of the TRTR-2005 / IGORR-10 Joint Meeting, Sept. 2005, Gaithersburg (MA), USA

[4] Generation IV systems R&D needs and research reactors policy

D. Iracane et al.

IAEA TCM on Research reactor support needed for Innovative power reactors and fuel

cycles, 20-22 November 2006, Vienna, Austria


58 of 435

Session II

Fuel development & fabrication

59 of 435

60 of 435

61 of 435

62 of 435

63 of 435

64 of 435

65 of 435

66 of 435

67 of 435

68 of 435

69 of 435




Argonne National Laboratory

9700 S. Cass Ave, Argonne, IL 60439, USA


Idaho National Laboratory,

Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415-6188, USA

1. Introduction


Results of destructive examination of fuel miniplates irradiated in the RERTR-8 test are

discussed . Metallographic features of dispersion fuel containing fuel particles of U-

7wt%Mo with 1wt% Ti or 2wt% Zr are analyzed. It is hypothesized that Zr, either as

alloy addition or fission product, may have a destabilizing effect on fission gas


The purpose of miniplate test RERTR-8 was to obtain irradiation performance data on

monolithic fuel plates fabricated by friction bonding (FB) and isostatic hot pressing (HIP), as well

as dispersion fuel plates that contain U-7Mo fuel particles alloyed with small amounts of Zr or Ti

(see Fig. 1). The results of the monolithic plates destructively examined to date were presented

at the 2007 RERTR meeting in Prague.

This paper presents the first results on the dispersion plates with Ti and Zr additions to U-7Mo.

2. RERTR-8 irradiation experiments

A total of 14 mini plates, seven monolithic plates and seven dispersion fuel plates, were

included in the RERTR-8 test irradiated in the ATR B-11 test position during cycle 138A, and B-

12 position during cycle 138B. Figure 1 shows the fuel plate positions in the test capsules as

well as a compositional description of the fuel plates. The total test duration was 104.7 EFPD to

achieve plate average fuel burnups in the range of 60 - 90 at% U-235 (LEU equiv). However,

the sides of the miniplates nearest to the ATR core achieved burnups in excess of 100% LEU

equivalent because of the flux gradient across the plates. The fuel enrichment was ~58.0% U-

235 in order to obtain the desired power densities and heat fluxes. RERTR-8 was designed to

have a peak heat flux ~355 W/cm 2 . The heat flux histories of the plates presented in this paper

are shown in Fig. 2.

* Work supported by US Department of Energy, Office of Global Threat Reduction, National

Nuclear Security Administration (NNSA), under Contract DE-AC-02-06CH11357. The submitted

manuscript has been authored by a contractor of the U. S. Government under contract NO.DE-AC-

02-06CH11357. Accordingly, the U. S. government retains a nonexclusive royalty-free license to

publish or reproduce the published form of this contribution, or allow others to do so, for U.S.

Government purposes.

70 of 435

3. Results

The transverse power shape is reflected in the gamma scan taken across the midsection of the

plates as shown in Fig. 3. Micrographs taken at the locations indicated in Fig. 3 are shown in

Fig. 4. For comparison, micrographs of a U-7Mo plate from the RERTR-7 test, which was

irradiated at similar condition, is also shown in Fig. 4, as well as a monolithic fuel plate in order

to facilitate the discussion.

Capsule Column 1 Column 2 Column 3 Column 4

A1 A2 A3 A4


U3Si2 U-10Mo FB U-7Mo-1Ti U-7Mo-2Zr

Al 0.01” foil Al-4043 Al-4043

U0R060 L1F200 D3R040 F3R030

A5 A6 A7 A8





C- Bottom


D- Bottom

B1 B2 B3 B4

B5 (112) B6 B7 B8

U-12Mo HIP

0.01” foil


C1 C2 C3 C4 (90)

U-10Mo HIP U-8Mo FB U-10Mo FB U-12Mo HIP

0.01” foil 0.01” foil 0.01” foil 0.01” foil

L1P020 J1F020 L1F190 H1P010

C5 (88) C6 (95) C7 C8 (87)

U-7Mo-1Ti U3Si2 U-7Mo U-7Mo-2Zr

Al-4043 Al Al-4043 Al-4043

D3R030 U0R040 R3R060 F3R040

D1 D2 D3 D4


Mg Matrix


D5 D6 D7 D8

Metallography performed

( )

Plate average burnup

Fig. 1 Test matrix of RERTR-8.

Heat flux (W/cm 2 )














0 20 40 60 80 100 120

Time (d)

Fig. 2 Plate-average heat fluxes of

RERTR-8 plates discussed in this


Cold side Midplate Hot side




Gamma scan count






Bu (%LEU)


0.0 0.2 0.4 0.6 0.8 1.0

Relative location (plate width)

Fig. 3 Gamma-scan result for D3R030 and F3R040.

The additions of Ti and Zr in combination with Si added to the matrix Al had been shown, in exreactor

diffusion couple tests [1], to reduce the extent of fuel-matrix interdiffusion more than with

Si alone. As indicated in Fig. 4, there appears to be a beneficial effect of adding Ti and Zr, albeit

71 of 435

small in absolute terms, because of the overwhelming reduction in interaction due to Si alone. It

remains to be seen if the effect of Ti or Zr is more pronounced at lower Si additions.

The fission gas bubble morphology is very similar for fuel with or without Ti. The bulk of the fuel

particles contain small, evenly distributed gas bubbles signifying stable swelling behavior

characteristic of metastable γ U-Mo. However, both fuels have developed large gas bubbles at

the periphery of several particles indicating unstable fission gas bubble behavior at high burnup.

An indication of perhaps similar unstable behavior appears to be occurring in the bulk of the fuel

containing Zr.

Fig.4 Post irradiation results of U-7Mo, U-7Mo-1Ti and U-7Mo-2Zr dispersion fuels in Al-4043

matrices and U-12Mo monolithic fuel.

4. Discussion

With the problem of excessive U-Mo/Al interaction and its commensurate break-away fission

gas bubble formation now well behind us, it appears that a new unstable fission gas behavior

has surfaced. As shown in figure 4, large bubbles form in the U-Mo fuel itself at the periphery of

the fuel particles, and not in the interaction product. Accelerated growth and interlinking appear

to take place at very high fission densities in these 58% U-235 enriched miniplates, at burnup

values of approximately 100% LEU equivalent. It appears, therefore, not to present an obvious

performance issue as LEU fuel is not driven to such high fission densities. However, this

phenomenon needs to be understood in order to establish appropriate margins for acceptable

fuel behavior, especially for monolithic fuel designs where such bubble formation at the fuel

72 of 435

periphery would weaken the fuel-cladding bond. The fission gas bubbles at the fuel-cladding

interface in monolithic miniplate H1P010, shown in Fig. 4, appear to be associated with the

interaction layer and represent the “old” porosity problem – not enough Si being available in the

Al 6061 cladding to stabilize the interaction layer. The morphology of the large gas bubbles at

the fuel particle periphery resembles that of U-Mo that has transformed from the metastable γ

phase to the equilibrium α+γ two-phase structure. As shown in Fig 5, this occurred in U-4Mo

irradiated in the RERTR-2 test. Apparently, at such a low Mo content the known γ stabilizing

effect of fission spikes is no longer sufficient to overcome the thermodynamic energy that drives

the alloy to its two phase equilibrium. However, the unstable behavior observed in Fig. 4 cannot

be due to a decrease in Mo content that causes this instability, as this clearly must increase as

U is burned, particularly with a Mo fission yield of 0.2 per U-235 atom fission.

20 µm

Fig.5 U-4Mo fuel specimen D005 after irradiation to 5.6x10 21 f/cm 3 .






Temperature ( o C)




10 100

Time (min)

Fig. 6 Effect of Zr additions to U-7Mo alloy on isothermal transformation [2].

There are, however, other high yield fission products that could change the chemistry of the fuel

alloy at high burnup. The most likely one in terms of yield (0.3 per U-235 fission), and its known

destabilizing effect on the U-Mo γ phase stability is Zr. This effect is shown in the TTT diagrams

in Fig. 6 and can be expressed in terms of the Zr/Mo ratio in the alloy. Because of the higher

fission yield of Zr compared to Mo, this ratio changes with burnup. This is shown in Fig. 7 for U-

Mo and U-Mo-2Zr together with the γ à α+γ’ transformation time derived from the TTT diagram

[2]. The TTT curves do not extend to the lower temperatures prevailing in the irradiation, and do

not include the stabilizing effects of the high fission rate in the fuel. However, they illustrate the

dramatic effect of the Zr/Mo ratio on the metastability of the U-Mo alloy. As far as the

73 of 435

appearance of gas bubbles at the fuel particle periphery, it has been shown by electron

microprobe examination that the fission product Zr migrates to the fuel particle surface during

irradiation [3,4]. Therefore, a high Zr/Mo ratio may occur initially at the fuel particle periphery. Of

the three miniplates shown in Fig. 4, F3R040 containing U-7Mo+2Zr is the only plate showing

large fission gas bubbles towards the center of the fuel particles, presumably because of its

initial high Zr/Mo ratio. Plate D3R030 containing Ti has a similar peripheral porosity as the pure

U-7Mo plate. This appears consistent with the negligible fission yield of Ti and its lack of effect

on the γ stability of U-Mo [5]. As the above discussion is merely a hypothesis, only detailed

micro-chemical analysis can improve our understanding of the metallurgy underlying this fission

gas bubble phenomenon. In addition, future tests with monolithic miniplates containing various

barrier layers between U-Mo foils and Al cladding, viz., Zr, Nb and Mo, should provide important

clues. In the interim, based on the above discussion, starting with as high a Mo content as

practicable, i.e., 10 or 12 wt% as a means of maintaining a low Zr/Mo ratio is advisable.

Time for γ-->α+γ' transformation (min)








0 20 40 60 80 100

Burnup (%LEU)

Fig.7 Change in Zr/Mo ratio as a function of burnup and time of γàα+γ’ transformation for U-

7Mo-2Zr and U-7Mo alloys at 450 o C. (No irradiation effects considered.)

5. Conclusions



The effect of Ti and Zr additions to U-7wt%Mo on the extent of fuel-aluminum interdiffusion,

although measureable, is small in absolute terms because of the overwhelming effect of the 5%

Si addition to the Al matrix. Ti additions to the U-7wt%Mo have no discernable effect on swelling

behavior of the fuel. However, there are indications that the addition of Zr may have a

destabilizing effect on fission gas behavior at high burnup.


[1] J.M. Park et al., J. Nucl. Mater., in print, 2008.

[2] C.A.W. Peterson, and W.J. Steele, UCRL-7824, 1964.

[3] F Huet et al., RERTR Conf., Chicago, 2003.

[4] F. Huet, RRFM Conf., Budapest, 2005.

[5] P.E. Repas et al., Trans. ASM, 57 (1964) 150.








74 of 435

Latest dispersed UMo fuel plate manufacturing results at AREVA-CERCA



Les Berauds, B.P. 1114, 26104 Romans Cedex – France


Commissariat à l’Energie Atomique / CEA-Cadarache

F-13115 – St Paul lez Durance Cedex – France


Commissariat à l’Energie Atomique / CEA-Saclay

F-91191 – Gif sur Yvette – France


Involved in the international UMo development program since 1999 this

paper aims at giving the recent manufacturing development results achieved

by AREVA-CERCA in the frame of collaborative efforts to overcome the

interaction layer formation during irradiation. Specifically a set of full size

UMo plates made of oxidized UMo powder was produced in order to

perform an irradiation test. This irradiation which is led by CEA and named

IRIS IV is scheduled in the CEA-OSIRIS reactor in 2008. This paper

presents, from a manufacturing point of view, the main information

gathered during the production.

1. Introduction

With an intrinsically good behaviour observed under irradiation, UMo alloy fuel is still

considered as a promising candidate in the frame of the worldwide reactor conversion


However, uncontrolled (UMo,Al) x interaction product formation occurring during irradiation,

which is defined as the initial cause of a detrimental process, has to be challenged [1].

Changing the aluminium matrix or either using coated UMo particles appear as some

potential remedies which are under evaluation [2].

Specifically, plates with a density up to 8 gU/cc manufactured through the CEA/CERCA

collaboration agreement, were irradiated in 2005 by CEA in the French OSIRIS reactor for

testing the benefits of Si addition in the aluminium matrix -0,3% and 2%- [3].

The irradiation conditions and the associated preliminary PIEs results are presented by CEA

and SCK-CEN in this conference [4, 5].

Moreover, four dispersed UMo plates using ground UMo particles were also successfully

irradiated in the frame of FRM II international working group program. The PIEs of these

plates are also presented during this conference by FRM2 [6].

As already observed through IRIS I irradiation [7] it seems that a modified UMo particles act

positively to the fuel behaviour during irradiation. These observations are consistent with

CEA out of pile results and FRM2 heavy ions investigations.

AREVA-CERCA, a subsidiary of AREVA-NP, an AREVA and Siemens company

75 of 435

In comparison with atomized UMo particles the properties of the fuel produced with ground

powder are the meat porosity ratio (~10%) and the UMo particles characteristics themselves

(uncontrolled oxidation layer and no Mo micro-segregation).

In order to evaluate independently the oxide effect over the particles, a set of full size plates

made of oxidized atomized UMo particles was produced at AREVA-CERCA. These plates

will be irradiated in the French OSIRIS reactor by mid-2008.

Alternative Al2%Si matrix will be also a part of this experimentation. The irradiation

conditions of IRIS IV will be similar to those used for the successful TUM plate irradiation.

IRIS IV plates were manufactured using oxidized U7Mo atomized powder in order to obtain

similar fuel properties to IRIS II and III.

This paper focuses on the description of IRIS IV UMo plates manufacturing and the main

plate’s characteristics reached within the scope of the study.

2. UMo particles preparation

Atomized low enriched U7%Mo particles were selected as initial raw material for the fuel


A specific heat treatment was defined in order to obtain the desired characteristics of the

oxide layer which has to be formed over the UMo particles surfaces.

As a compromise between the oxide layer uniformity around the fuel as well as to keep the

integrity of the barrier during the rolling steps, an oxide layer thickness of 1,5 µm was

selected for IRIS IV test.

In order to define the heat treatment, bilateral investigations were launched between CEA and


According to the CEAs investigations the Time-Temperature-Thickness diagram was

determined at a lab scale where a scaling up treatment feasibility was studied at AREVA-


Among the various Time and Temperature conditions, a lower temperature (220°C) was

chosen in order to finally get a homogeneous oxide layer around the particles as well as a

better adhesion to the substrate –Figure 2-.

Prior to the treatment, we calibrated and performed a temperature mapping of the furnace

used. Fine and large UMo particles were heat treated separately during the same time under

air atmosphere in Al 2 O 3 crucibles. According to the low annealing temperature of the particles

the γ phase of the UMo was not altered.

In order to have the same total mass per heat treatment batch, the UMo mass filled up in the

crucible was similar whatever the UMo particles size. In average, an oxide layer of 1,3 µm is

observed around UMo particles.

Size particles class

Oxidation layer thickness –Average-

Fine 1,4 µm

Large 1,3 µm

Figure 1: Oxide layer thickness according to the particle sizes

76 of 435

Further oxidized UMo powder characterizations are being carried-out by CEA.




Figure 2: Metallographic views of oxidized UMo particles

A) Main oxide layer aspect B) Altered oxide layer C) Homogenous oxide layer

3. IRIS IV fuel plates characteristics

The dimensions of the UMo full size plates for IRIS IV are identical to the ones already

produced in the frame of previous IRIS irradiations –IRIS I, II and III-.

The main IRIS IV plate characteristics are described in the figure 3 hereafter

Type A x 2 plates

Type B x 2 plates

Matrix Pure aluminium Al2%Si alloy

Ut and U 235 content 131,6 gUt & 26 gU 235


Cladding material

Nominal Fuel meat dimensions

Nominal Fuel Plate dimensions

~ 7,9 gU/cc


596,5 Length x 55 width x 0,51 thickness (mm)

641,9 Length x 73,3 width x 1,27 thickness (mm)

Figure 3: Characteristics of IRIS IV plates

4. Manufacturing and fuel plates characterizations

The manufacturing options of UMo dispersed fuel plates, with a density up to 8 gU/cc, is now

fully completed by AREVA-CERCA. Involved since 1999 in the UMo plate development, all

the fuel plate processing steps are suitably defined, and well mastered, so as to manufacture

UMo fuel plate with a high quality level. The main inspection criteria and results are

summarized figure 4 here below.

77 of 435


IRIS IV plates inspection results

Blister test

UT inspection

(For information)

X-Ray inspection: Stray particles &

white spot

Homogeneity inspection through

digital X ray processing (Figure 5)

No blister detected

One minor defect detected according to an inspection

carried out with the same criteria used for the routine

inspection of U 3 Si 2 OSIRIS fuel plates. UMo fuel plates


Few stray particles observed and accepted due to small

sizes which were less than specified. No white spot


U distribution inside the fuel meat area less than ± 16 %

Cladding thickness by metallographic

inspection (Figure 6)

Mid plan cross section : 0,38 mm

Dog-bone area : > 0,25 mm

Porosity (average / for information) Al matrix: 2 % and Al2Si matrix: 3,5 %

Figure 4: Main IRIS IV fuel plates inspection results

U7Mo within pure Al matrix –one longitudinal trace along the fuel meat / Density ~ 8 gU/cc-

Fuel meat area

U7Mo within pure Al2%Si matrix –one longitudinal trace along the fuel meat / Density ~ 8 gU/cc-

Dog bone area

Dog bone area

Figure 5: Homogeneity inspection results

The porosity variation in between the Al2%Si and pure Al matrix is coherent and explained

by the difference of the mechanical properties of the Al powders batches.

The homogeneity recorded on the IRIS IV UMo fuel plates, which is the U repartition over

the fuel (gU/cm 2 ), is very tight. In between plate to plate or over the same plate, the recorded

variations are less than ± 16 %. This result is found whatsoever the Al matrix.

Figure 6 exhibits a perfect fuel meat shape.

78 of 435

Figure 6: Metallographic inspection results

Due to the fuel meat deformation when the plate is rolled, and accentuated by the high

density, some UMo particles interpenetrate each others so as to reach, time to time, a local

deformation of the fuel particles. As shown on figure 7, even when it happens, locally, an

oxide layer is clearly visible. But sometime, as reported figure 2 and confirmed figure 7,

locally, the oxide layer is broken. Such effect was not yet quantified. This local and potential

detrimental effect will be a part of the IRIS IV experimentation and will be further analyzed

during the PIE investigations.

Local alteration of the oxide layer

Oxide layer integrity still remains even

under UMo particles penetration

Figure 7: Oxide layer behaviour after plate production by rolling

5. Conclusion

As part of our agreement with CEA, a new set of UMo full size plate with a specific oxide

barrier was produced successfully at AREVA-CERCA.

79 of 435

From a manufacturing point of view, and by using our highest state of art manufacturing

knowledge, dispersed UMo plate processing is considered as fully mastered. The quality of

IRIS IV UMo plate reached the standard of the high density silicide fuel plates. Without

major change, the switch to an industrial scale up plate production awaits the UMo fuel


The IRIS IV plates will be irradiated in OSIRIS reactor from mid 2008.


[1] F. Huet "Post irradiation examinations on UMo full sized plates –IRIS II experiment-"

9 th International Conference on Research Reactor Fuel Management (RRFM 2005),

Budapest, Hungary, April 10-13, 2005.

[2] JL.Snelgrove & All "High-density UMo fuels -latest results and reoriented qualification

programs-", 9 th International Conference on Research Reactor Fuel Management

(RRFM 2005), Budapest, Hungary, April 10-13, 2005.

[3] M. Ripert et al. “IRIS-3 experiment – status and first results of thickness increase”, 10 th

International Conference on Research Reactor Fuel Management (RRFM 2006), Sofia,

Bulgaria, 30 April –3 May, 2006.

[4] M. Ripert & All " Overview on high density UMo fuel in pile experiments in OSIRIS”,

this conference

[5] A. Leenaers et al. “Microstructural analysis of irradiated atomized U(Mo) dispersion

fuel in a Al matrix with Si addition”, this conference

[6] A.Röhrmoser & All "Reduced enrichment program for the FRM II, Status 2006/2007",

this conference.

[7] F. Huet & All “Full-sized plates irradiation with high UMo fuel loading –Final results of

IRIS I experiment- RERTR’03, Chicago

80 of 435





Nuclear Fuels and Materials Division, Idaho National Laboratory

P. O. Box 1625, Idaho Falls, Idaho 83403 USA


RERTR U-Mo dispersion fuel plates are being developed for application in research reactors

throughout the world. Of particular interest is the irradiation performance of U-Mo dispersion fuels

with Si added to the Al matrix. Si is added to improve the performance of U-Mo dispersion fuels.

Microstructural examinations have been performed on fuel plates with either Al-0.2Si or 4043 Al

(~4.8% Si) alloy matrix in the as-fabricated and/or as-irradiated condition using optical metallography

and/or scanning electron microscopy. Fuel plates with either matrix can have Si-rich layers around

the U-7Mo particles after fabrication, and during irradiation these layers were observed to grow in

thickness and to become Si-deficient in some areas of the fuel plates. For the fuel plates with 4043

Al, this was observed in fuel plate areas that were exposed to very aggressive irradiation conditions.

1. Introduction

The United States Reduced Enrichment for Research and Test Reactors (RERTR) Fuel

development program is actively developing low enriched uranium (LEU) fuels for the

world’s research reactors that are currently fueled by uranium enriched to more than 20%

235 U.

To assess the performance of U-Mo dispersion fuels with Si-doped matrices, different

reactor experiments have been conducted using the Advanced Test Reactor. Experiments

have been run with dispersion fuels that have Al-0.2Si, Al-2Si, Al-5Si, 6061 Al and 4043 Al

alloy matrices. This paper will discuss results of recent microstructural characterization that

was performed on fuel plates that were irradiated as part of the RERTR-6 and RERTR-7

experiments that have either Al-0.2Si or 4043 Al alloy as the matrix.

2. Experimental

2.1 Irradiation Testing

The RERTR-6 experiment was the first experiment to test “second generation” U-Mo fuels

designed to overcome the fuel performance problems encountered in U-Mo/Al dispersions

[1]. In this experiment, the fuel materials were tested to high burn-up under moderate flux

and moderate temperature conditions. The RERTR-7 experiment was a more aggressive

test and employed fuel enriched to 58% 235 U. RERTR-7 was divided into two parts: RERTR-

7A and RERTR-7B [2].

In Table 1, the irradiation conditions for some specific plates from the RERTR-6 and

RERTR-7 experiments are presented. These plates had either Al-0.2Si or 4043 Al as the

matrix and are the plates focused on in this paper. Chemical analysis of the 4043 Al

revealed a composition of 4.81Si-0.20Fe-0.14Ti-0.16Cu-0.01Cr-0.01Mn-bal Al. Less than

0.01 wt% of Zn and Mg was measured.

81 of 435

Table 1. Irradiation conditions for fuel plates R5R020, R3R030, R3R040, and R3R050.

Fuel Plate


Exper. Matrix Peak






(10 21 f/cm 3 )



Rate (10 14

f/cm 3 s)




(W/cm 2 )

R5R020 RERTR-6 Al-0.2Si 117.1 3.30 2.83 130.52

R3R030 RERTR-6 4043 Al 97.5 3.26 2.80 101.5

R3R040 RERTR-7 4043 Al N/A 5.03 6.46 N/A

R3R050 RERTR-7 4043 Al 139.9 4.90 6.30 299.3

2.2 Microstructural Characterization

For the as-fabricated fuel, microstructural characterization was performed on transverse

cross-sections using scanning electron microscopy with energy dispersive spectroscopy and

wavelength dispersive spectroscopy (SEM/EDS/WDS).

For the as-irradiated fuel plates, optical metallography (OM) was performed on transverse

cross section taken from the mid-plane of the fuel plate. For the SEM/EDS/WDS analysis of

the as-irradiated plates, a punching process was first used in the Hot Fuel Examination

Facility to generate one-mm-diameter cylinders that contained a sampling of the fuel meat,

and then in the Electron Microscopy Laboratory these cylinders were mounted, polished,

and coated with a thin layer of Pd [3]. SEM/EDS/WDS analysis was performed to

characterize the microstructure and determine how different fuel and matrix components

partitioned between the different phases during irradiation.

3. Results

3.1 As-Fabricated Plates

During the fuel fabrication campaign for the RERTR-6 experiment, archive fuel plates were

produced that were later characterized to determine the starting microstructure of the fuel

before irradiation. R3R020 was the fuel plate that was characterized to determine the

starting microstructure of a fuel plate with U-7Mo fuel particles and 4043 Al matrix. For the

Al-0.2Si matrix fuel, no as-fabricated fuel plate was produced to serve as an archive due to

the aggressive fabrication schedule being followed to get all the plates that comprised the

RERTR-6 experiment into reactor. Results from diffusion experiments using U-7Mo and

low-Si Al-Si alloys at temperatures representative of fuel fabrication temperatures can be

looked at to get an idea of how these plates would look after fabrication [4].

An SEM image of the microstructure of the R3R020 fuel plate is presented in Figure 1. Thin

fuel/matrix interaction layers are present around all the fuel particles. These interaction

layers were a result of the exposure of the fuel plates to relatively high temperatures during

the rolling and blister annealing steps that was a part of the fuel fabrication process. During

rolling, the plates were exposed to around 500°C for up to one hour. During blister

annealing the plates were exposed to 485°C ±20°C for 30 minutes [5]. Also, during fuel

fabrication, the original γ−phase U-7Mo alloy apparently decomposed to α-U and γ’. This

resulted in some localized fuel/cladding interaction, as shown in Fig. 1b. X-ray maps that

were produced (see Fig. 2) show that these interaction layers were enriched in Si. The

maximum Si content of the interaction layer was measured by SEM/EDS to be 45 at% with a

maximum (Al+Si) concentration of 69 at%, and the (Al+Si)/(U+Mo) ratio varied between 1.7

and 2.2. The layers were on the order of 1-2 µm thick. U, Mo, and Al were also mapped

and showed U and Mo in the fuel; U, Mo, and Al in the interaction layer; and, Al in the

matrix. No oxygen enrichment was observed in the interaction layers.

82 of 435



Fig. 1. SEM images of the microstructure for the as-fabricated fuel plate R3R020.



Fig. 2. (a) An SEM image of fuel particles in as-fabricated plate R3R020, and (b) a Si X-

ray map showing the enrichment of Si in the interaction layer.

For the RERTR-7 fabrication campaign, archive plates with U-7Mo-2Zr and U-7Mo-1Ti fuel

particles in 4043 Al matrices were examined with SEM/EDS/WDS. The observed

microstructures were very similar to those shown in Figs. 1 and 2, and the U, Mo, Al, and Si

partitioning behavior was very similar.

3.2 As-Irradiated Plates

3.2.1 Optical Metallography

Fuel plate R5R020 was a fuel plate with an Al-0.2Si matrix that was irradiated as part of the

RERTR-6 experiment. OM images that were taken in different areas of a full transverse

cross section taken at the mid-plane of the as-irradiated microstructure are presented in Fig.

3. Due to the fission density gradient that was present across the width of the fuel plates for

the RERTR-6 experiment, there was a variation in the interaction layer thickness that was

observed around the fuel particles. The thickest layers (∼10 µm) were observed at the

highest-burnup edge of the plate.

The fuel plates with 4043 Al alloy matrix that were irradiated in RERTR-6 or RERTR-7

experiments included R3R030, R3R040, and R3R050. Representative OM images of the

microstructures observed along full transverse cross sections taken at the mid-plane of

R3R030 and R3R050 fuel plates are presented in Fig. 4. Figs 4a and 4b show the relatively

narrow interaction layers observed across the mid-plane of R3R030 (around 1 to 2 µm).

Figs 4c and 4d show the thicker layers (up to 10 µm) that were observed across the midplane

of R3R050. R3R040 exhibited interaction layer thicknesses that were similar to those

observed for R3R050. The thickest layers were observed at the edge of the plates that

were exposed to the highest burnup.

83 of 435



Fig. 3. OM images of the R5R020 fuel plate microstructure observed at the edges of the

fuel plate with the (a) lowest and (b) highest burnups.





Fig. 4. OM images of the fuel microstructures observed for fuel plate R3R030 towards the

(a) highest and (b) lowest burnup edges of the plate and for plate R3R050 towards the

edges with the (c) lowest and (d) highest burnups.

3.2.2 Scanning Electron Microscopy

SEM images of the microstructure observed for the Al-0.2Si matrix fuel plate R5R020 are

presented in Fig. 5. Like was the case for the OM images (see Fig. 3), the thickness of the

interaction layer was observed to be around 10 µm. X-ray mapping was employed to

determine the partitioning behavior of fuel and cladding components (see Fig. 6 for Si). No

concentration gradients for U, Mo, Al, or Si were observed in the interaction layer, and pointto-point

composition analysis showed that the average composition (determined from

fourteen points), in at%, of the interaction layer was around 83.1Al-2.7Mo-14.3U (± ~2 at%).

As expected, based on the lack of Si in the generated Si X-ray maps, negligible Si was

measured in the interaction layer. The Si was observed in precipitates that were present in

the fuel meat matrix.

84 of 435

Fig. 5. SEM images of the microstructure observed for fuel plate R5R020. The arrows

indicate pores observed in the fuel.



Fig. 6 Secondary electron image (a) and Si X-ray map (b) for fuel plate R5R020.

For fuel plate R3R030, two different types of microstructure were observed: one had around

1 to 2 µm-thick interaction layers and the other had layers that were around 10 µm thick,

based on looking at the interaction layer thickness around the largest-diameter particles.

The thickest layers coincided with regions of the fuel plate that had achieved around 100%

LEU burnup. Fig. 7 shows the fuel plate microstructure, and Fig. 8 shows a Si X-ray map

where the thinner layers were observed. Figs. 9 and 10 show the same where the

microstructure displayed thicker layers. In the microstructure where the thinner layers were

observed, the layers were enriched in Si, and there were precipitate free zones (PFZ)

around many of the particles. These PFZs have been interpreted as the result of the recoil

damage zones that extend around each of the U-Mo particles to a distance of around 10

µm, and it has been suggested that the Si-containing precipitates in these regions dissolve

and the Si from the precipitates diffuses towards the fuel/matrix interface [6]. For the areas

of the microstructure with the thicker interaction layers, negligible Si was observed in the

layers. The original Si in the interaction layers appeared to have come out as precipitates in

the matrix. Point-to-point composition analysis at fifteen different locations within the ∼10

µm-thick-layer indicated an approximate composition, in at%, of 82.4Al-2.5Mo-15.1U (± ~2

at%). Because the Si-rich layer was smaller than the spatial resolution of the individual

composition measurements, the composition of this layer could not be measured.

85 of 435

(a) (b)

Fig. 7. SEM images of the microstructure for fuel plate R3R030 where relatively thin

interaction layers were observed.

(a) (b)

Fig. 8. SEM image (a) and Si X-ray map (b) for R3R030 where interaction layers were

relatively thin.



Fig. 9. SEM images of the microstructure where relatively thick interaction layers were

observed in fuel plate R3R030.



Fig. 10. SEM image (a) and Si X-ray map (b) in an area of the R3R030 fuel plate where

relatively thick interaction layers were observed.

86 of 435

4. Discussion

Based on the characterization that was performed on as-fabricated plates that went into the

RERTR-6 and RERTR-7 experiments, it is clear that Si-rich interaction layers were already

present around the U-7Mo fuel particles before any of the fuel plates with 4043 Al alloy

matrix were inserted into the Advanced Test Reactor. These interaction layers were a result

of the exposure of the fuel plates to relatively high temperatures during the rolling and blister

annealing steps that were a part of the fuel fabrication process. Based on interdiffusion

studies that have been performed using U-7Mo and low-Si Al alloys [4], there is a good

chance that the fuel plate with Al-0.2Si also had pre-existing Si-rich interaction layers.

Looking at the OM images for the irradiated fuel plates, it is clear that in some cases the

relatively thin interaction layers that were present in the fuel plates before irradiation have

grown in reactor, and in some cases have reached an approximate thickness of 10 µm. For

the fuel plates with 4043 Al matrices, the interaction layer thickness can approach 10 µm in

the areas of the fuel plates that achieved around 100% LEU burnup. Based on SEM

analysis, the 10-µm-thick interaction layers contain negligible Si. Conversely, when fuel

particles have retained the relatively thin interaction layers during irradiation and are

characterized using the SEM, appreciable Si is observed. This suggests that during

irradiation enough Si must diffuse to the interaction layers in order to keep the fabricationgenerated

layers stable (i.e., large pores do not form like for the U-Mo/Al matrix fuels). If this

does not transpire, then the Si-rich layers become unstable, and the U-Mo-Al interdiffusion

behavior that is typical during the irradiation of U-Mo dispersion fuels with Al as the matrix

takes over. Other researchers have also concluded that it is important to have sufficient Si in

the matrix of a U-Mo dispersion fuel in order to get good irradiation performance [7].

The Al-0.2Si matrix dispersion fuels do not appear to contain enough Si to keep the thin, Sirich

interaction layer stable. Only thick interaction layers were observed that contained

negligible Si. This in combination with the information from the R3R030, R3R040, and

R3R050 fuel plates would suggest that there is some Si concentration level between 0.2

wt% and 4.81 wt% where there would be enough Si in the matrix to keep the Si-rich

interaction layer stable, resulting in good fuel plate irradiation behavior to high burnups. It

has been shown that fuel plates with 2.0 wt% Si added to the matrix also exhibit good

irradiation performance [8]. Even with 4.81 wt% Si in the matrix of a fuel element, porosity

and 10-µm-thick interaction layers can be observed in some local areas of a fuel plate, but

this is only observed where the fuel plates had been exposed to extremely high burnup

levels (i.e., ∼100% LEU burnup). These high burnup levels are beyond what a typical

research reactor fuel would see, and even with these features present, the fuel plates

displayed overall good irradiation behavior.

5. Conclusions

Based on the characterization of as-fabricated and irradiated U-7Mo dispersion fuel plates

with either Al-0.2Si or 4043 Al alloy as the matrix, the following conclusions can be drawn:

1. Fuel plates that were inserted into the Advanced Test Reactor as part of the RERTR-6

and RERTR-7 experiments already had Si-rich interaction layers present around the fuel

particles, due to the fuel plate fabrication process.

2. After irradiation, the RERTR-6 fuel plate with Al-0.2Si alloy matrix appeared to have

developed only relatively thick fuel/matrix interaction layers that contained negligible Si.

3. The fuel plate with 4043 Al (4.81 wt% Si) matrix, irradiated as part of the RERTR-6

experiment, contained Si-rich interaction layers that were about the same thickness as those

that were produced during fabrication, along with relatively thick layers that contained

87 of 435

negligible Si. The thick layers seemed to form in areas of the fuel plate that were exposed

to the highest burnup. Thick interaction layers could also be found in fuel plates that had

4043 Al matrices that were irradiated in the aggressive RERTR-7 experiment.

4. In order for Si-rich fuel/matrix interaction layers to remain stable in U-Mo dispersion fuels

during irradiation, it appears there needs to be a sufficient supply of Si in the matrix, and the

optimal Si content is somewhere between 0.2 and the 4.81 wt%.


This work was supported by the U.S. Department of Energy, Office of Nuclear Materials

Threat Reduction (NA-212), National Nuclear Security Administration, under DOE-NE Idaho

Operations Office Contract DE-AC07-05ID14517. Personnel in the Hot Fuel Examination

Facility are recognized for their contributions in destructively examining fuel plates.


[1]. C. R. Clark et al., RRFM 2004, Munich, Germany, March, 2004.

[2]. D. M. Wachs et al., RERTR 2006, Capetown, South Africa, Oct. 29-Nov. 2, 2007.

[3] D. E. Janney et al., Hot Laboratories and Remote Handling Conference (HOTLAB

2007), Bucharest, Romania, Sep. 20-21, 2007.

[4] D. D. Keiser, Jr., Defect and Diffusion, Vol. 266 (2007) pp. 131-148.

[5] T. C. Weincek, Argonne National Laboratory Report, ANL/RERTR/TM-15, (1995).

[6] G. L. Hofman et al., RERTR 2006, Capetown, South African, Oct. 29-Nov. 2, 2007.

[7] G. L. Hofman et al., RRFM 2007, Lyon, France, March 11-15, 2007.

[8] G. L. Hofman et al., RRFM 2006, Sofia, Bulgaria, April 30-May 3, 2006.

88 of 435

UMo full size plate irradiation experiment IRIS-TUM – a progress report

W. Petry, A. Röhrmoser

Forschungsneutronenquelle Heinz Maier-Leibnitz (FRM II),

Technische Universität München, D-85747 Garching , Germany

P. Boulcourt, A.Chabre, S. Dubois, P. Lemoine

CEA Saclay - 91191 Gif-sur-Yvette Cedex – France

Ch. Jarousse, JL. Falgoux

CERCA Romans – France

S. van den Berghe, A. Leenaers

SCK•CEN, Institute for Nuclear Materials Science, Boeretang 200, B-2400 Mol - Belgium


Irradiation and swelling measurements for IRIS-TUM, an experiment to test large scale

UMo dispers fuel plates under elevated heat load, have been finished. UMo fuel made of

ground powder with an Al matrix with and without Si additives have been irradiated up to

a LEU equivalent burn-up of 88.3 %. In none of the fuel plates a failure of the first barrier

– the cladding – has been observed, even not at a thickness increase of 323 µm, which

corresponds to 66% of “swelling”. At low irradiation dose large in-build porosity delays

the onset of linear swelling. During the continuation of the irradiation, a period of almost

linear increase of thickness is then followed by one with more fast increasing. In the most

favourable case this nonlinear increase begins at about 2.0 10 21 f/cm 3 , in the case of no

additional Si at lower fission dose. Fuel with Si added to the Al in the dispersion swells

less than that without additives. First microscope images from samples cut out of plates

with medium irradiation level do not yet give a clear answer why this is the case. Growth

of the interdiffusion layer is – if at all - only slightly hindered by the addition of Si. The

progress achieved in this irradiation campaign is dominantly ascribed to the usage of

ground powder.

1 The collaboration CEA-CERCA-TUM

In 2003 the Technische Universität München (TUM) launched a program for the development of high density

fuel for research reactors with highest neutron flux. Principally this gain in density can then be used to

reduce the enrichment of the fuel. Still a single compact core like that of FRM II can for physical reasons not

be replaced by a compact core of low enriched Uranium (LEU). However, reduction to medium enrichment

(MEU) is conceivable [1]. In a collaboration with the French Commissariat à l’Energie Atomique (CEA) and

the company AREVA with its divisions NP and CERCA different metallurgical and methodological options

are persecuted: a) irradiation of full size fuel plates made of UMo alloy particles dissolved in an Al matrix

with an AlFeNi cladding, b) tests of modified UMo alloys in various dispersions by heavy ion irradiation, c)

development of manufacturing processes for full size UMo monolithic foils including cladding, d) calculation

of the neutronics and thermohydraulics of possible high density fuel elements for the high flux reactor

FRM II. Progress reports concerning b) & c) can be found in other contributions to this conference [2,3,4],

whereas the collaboration reported recently on the neutronics of an advanced fuel element design for FRM II


In this paper we present the status of a) test irradiations on full size plates on the basis of UMo fuel dissolved

in an Al matrix, the so-called irradiation experiment IRIS-TUM

2 TUM strategy on UMo dispersive fuel

To overcome the observed malfunction of the pure UMo dissolved in an Al matrix - firstly reported during

RRFM 2004 [6,7] - improvements came into focus that time as there are modifications in the UMo fuel respectively

the fuel matrix composition or coating of the fuel powder before mixing with the matrix. Both

89 of 435

aspects were the basis for several irradiation programs started 2004/2005 for this fuel. First results could be

shown in 2006 and were the ground for new optimism on UMo dispersive fuel. Main directions for future

fabrication pathways are modest additives to the fuel or matrix or an oxidization of the powder of this fuel,

both aiming on the suppression of the formation of the Al-rich interdiffusion layer around the UMo grains


The irradiation experiment IRIS-TUM incorporated this international experience in the definition of its

goals. I) Full size test plates manufactured under industrial conditions give much more reliable information

upon the irradiation behavior, when compared to mini-plates typically produced under idealized laboratory

conditions. II) Irradiation should happen at fission rates and heat load at least approaching those of high performance

research reactors. III) Continuous registration of the swelling in the course of the irradiation in

order to visualize swelling as function of build-up of the fission density. IV) Redundancy in the irradiation

program, i.e. two test plates of the same kind increase statistical evidence of the results and minimize the risk

of technical failures during the long lasting irradiation campaign. V) Only the French test program IRIS-1

did not fail at that time. Different to almost all other irradiation programs the full size plates for IRIS-1 were

manufactured on the basis of ground instead of atomized UMo powder. VI) Slight increase of the Mo content

to 8 wt % in order to be further away from the α- to γ-phase boundary. The γ-phase is supposed to show better

accommodation of the fission products. VII) Aim to a maximum Uranium density, but try also lower densities

in case the maximum density fails. VIII) Addition of Si in the Al matrix as diffusion blocker to suppress

the formation of the interdiffusion layer.

All these consideration resulted in the production of six full size fuel plates by CERCA, of which the essential

parameters are summarized in Table 1

Plate number 8001 8002 8501 8503 7002 7003

Uranium density gU/cm 3 8.5 8.4 8.3 8.3 7.3 7.3

Porosity vol. % 8.1 7.9 9.0 8.9 6.5 6.4

Si in Al content wt % 0.07 0.07 2.1 2.1 0.07 0.07

Vol. % of Al in the meat 38.2 38.0 38.7 38.6 45.0 45.2

Meat thickness mm 0.49 0.49 0.49 0.49 0.54 0.54

Mo in UMo wt % 8.1 8.1 8.1 8.1 8.2 8.2

Table 1: Parameters for the 6 test plates for IRIS-TUM. Common to all plates is the enrichment of

49.3(2)% 235 U, dimension of the fissile zone (meat) 558.5(1.5) × 55,5 × 0.49 mm 3 for the 8 gU/cm 3 density

and 558.5(1.5) × 55,5 × 0.54 mm 3 for the 7 gU/cm 3 density, and the dimension of the full plate including

cladding 641.45(5) × 73.3 × 1.3 mm 3 .

Great attention was given to the requirement to have conditions during the irradiations as close as possible to

those in a potential future FRM II fuel element. Particularly the maximum temperatures in the meat should

be comparable. To do so OSIRIS needed an extension of its irradiation license for heat flux values in the

order of 300 W/cm 2 . The authorization was granted mid of 2005, so that the irradiations could start September

2005. The outer cladding temperature was targeted to be above 100°C. For the FRM II with the actual

U 3 Si 2 fuel the nominal values of maximum temperatures at the cladding surface are given with 98°C for

BOL (‘begin of life’). Including safety margins a maximum temperature of 119°C at the cladding surface is

mentioned in the safety assessment of the FRM II core. With a MEU core the maximum temperatures are

expected to be slightly higher. And quite similarly it has to be added a margin to the maximum expected

fission density (FD) in a core, so that finally a FD = 2.3 . 10 21. cm -3 in the meat shall be reached by the test


3 Irradiation IRIS-TUM 2005-2007

Irradiation at the MTR reactor OSIRIS at CEA -Saclay started in Sept. 2005, the last irradiation cycle ended

March 2007 – see Fig. 1. The four plates with nominal 8 gU/cm 3 density were distributed into two irradiation

90 of 435

devices (core position 11 & 17) to respond to the effect of too severe self-shielding in the case of four adjacent

plates in one single irradiation device. The plates with 7 gU/cm 3 were not inserted in the core and served

as a reserve. The neutron spectra were rather identical for both positions, since they are at two similar edges

of the OSIRIS core. Measurements of the swelling were done in situ mechanically after each cycle # .

Fig. 1:

Irradiation schedule at OSIRIS

for five IRIS-TUM plates with

totally 8 reactor cycles at a

thermal power between 61 – 69









0 50 100 150

start: sept.05 FPI [days] end: march.07

3.1.1 Position 11

In June 2006 after five reactor cycles the flux calculations indicated a maximum fission density superior to

the target of FD max = 2.3 . 10 21. cm -3 . With that date the two plates at the position 11 (8503, 8002) have been

taken out of the irradiation programme in order to perform destructive Post Irradiation Examinations (PIE)

after a one year cooling down period. Only after detailed γ-spectroscopy several month after the suspension

of the irradiation in position 11 it turned out, that the calculated FD max overestimated the measured FD by

about 15%. As a consequence the achieved FD max in plates 8503 & 8002 slightly misses the target value –

see also chapter 4.1.1.

3.1.2 Position 17

Due to mechanical deformation plate 8001 could not be reinserted in position 17 after the 2 nd irradiation cycle.

Therefore irradiation of this plate had to be stopped and instead plate 7003 has been inserted in position

17. Because after the 5 th irradiation cycle non of the four plates showed break away swelling, it was decided

to irradiate the two plates in position 17 further. Even after the 8 th irradiation cycle no leakage of fission

products has been observed, however the swelling was such, that further irradiations have been abandoned.

After 1 year cooling time, i.e. earliest April 2008, also these plates will be examined by PIEs.

4 Results

4.1 Fission density (FD) distribution

The irradiation positions 11 & 17 are at the outer corners of the fuel array of OSIRIS. As a consequence

strong anisotropy in neutron flux in vertical as well horizontal direction of the test plates is expected. The

distribution of swelling over the hole surface of the test plates as shown in the subsequent figures gives an

idea of this anisotropy. The anisotropy itself is of no concern for the swelling tests. On the contrary, fuel

plates in real fuel elements experience similar anisotropic neutron fluxes. Ho wever, of great concern is the

exact knowledge of the anisotropic fission density and its absolute value. Therefore big efforts were undertaken

to reach most resilient data with respect to FD values and temperatures.

4.1.1 Expected and measured FD in the maximum flux plane (mfp)

The power density distribution for each irradiation cycle and for each plate have been calculated by 2dneutronic

flux calculations prior to the irradiation. Fig. 2 (left) shows the calculated fission density (FD)

# These mechanical measurements are done in the reactor pool with a device, originally developed for qualifying the

actually used U 3Si 2 fuel in FRM II [10].

91 of 435

distribution for the plate 8503 along the maximum flux plane (mfp), i.e. the horizontal line in the mid of the

plate length. When compared to the measured swelling in that same plane, see Fig 2 (right) a qualitative

agreement can be stated. Quite general the power in the mfp was calculated to be 20% higher than the total

plate average.







plate 8503, fission densities at 'mfp'


[cm -3 ]






-30 -20 -10 0 10 20 30

fuel zone (55,5 mm width)

5.cycle F210

4.cycle F208

3.cycle F207

2.cycle F205

1.cycle F204












swelling [µm]

plate 8503

-30 -20 -10 0 10 20 30

plate width [mm]

Fig. 2: Left: Calculated fission density distribution in the maximum flux plane for plate 8503 after up to 5

cycles of irradiation and compared to the measured fission density by γ-spectrometry after the last cycle

F210. Right: Measured swelling in the mfp after up to 5 cycles of irradiation.

To verify these beforehand calculated fission densities γ-spectrometry was performed with one plate of each

of the two core positions 11&17 some month after the respective last irradiation cycle. The FD as determined

by γ-spectrometry is regarded to have an absolute precision of about 3-5 %. A FD curve derived from the

measured activity of the isotope 137 Cs over the width of plate 8503 at position 11 is also shown in Fig 2 (left).

In fact comparing calculated and measured FD the latter reproduces much better the enhanced swelling at the

left as well as right corner. The concentration profile of the measured long living isotope 137 Cs over the plate

width is obviously a blueprint for the swelling profile.

Therefore the calculated power densities and the subsequently calculated FD have been calibrated by the

result of the γ-spectrometry. These calibration factors have been derived from the mean calculated and measured

fission densities for the positions 11/17 and are cf MC = 0.839/0.858. The so reached averaged and

maximal FD at the corners of the

mfp and after the respective last

irradiation cycle are noticed in

Table 2.












Table 2: Burn-up, FD and max.

swelling in the different irradiated

plates. FD values refer to the

meat as well to the UMo grains.

position 11 position 17

cf MC 0.839 0.858

burn-up (%)


(%) FD in meat (cm -3 )

average 14.1 14.6 5.5 16.7 23.7

max. 23.1 22.5 9.3 26.6 35.3

max. 57.8 56.3 23.3 66.5 88.3

average 1.3 1.3 0.5 1.3 2.1

max. 2.3 2.2 0.9 2.2 3.4

f meat/grain 0.533 0.539 0.54 0.466 0.532

FD in grain (cm -

33 )

max. 3.9 3.8 1.7 4.4 5.9

Swelling (µm) max. 93 104 31 178 323

Last irradiation cycle F210 F210 F205 F217 F217

92 of 435

As can be seen in Table 2 three plates miss the target of the maximal fission density FD max = 2.3 . 10 21. cm -3 in

the meat. However the irradiation of plate 8501 largely - and purposely - exceeds the target. By means of the

volume fraction of the UMo grains in the meat the fission density in the grain can be calculated. The factors

f meat/grain together with FD in the grain are noted in Table 2, too.

4.1.2 Heat load and temperatures at the hot spot

In a comparable manner the heat load and accordingly the temperatures at the outer side of the cladding had

to be corrected. The corrected maximal heat load and temperature at the hot spot in the mfp have been calculated

to 260 W/cm 2 resp. 98°C. Due to burn-up, but also due to variable core loading, heat load and temperature

varied with irradiation time. Time averaged heat load and temperatures at the hot spot are 230 W/cm2

and 90°C.

4.2 Thickness measurements

After each cycle the test plates have been extracted from their respective irradiation position, inserted in a

measuring device within the reactor pool and absolute thickness has been measured with an accuracy of ±2

µm along several traces in vertical and horizontal direction. Swelling data have been obtained by subtracting

the thickness of the respective plate before irradiation. Fig. 3 – 6 display the so measured swelling for four

test plates in the vertical direction – and Fig 2 in horizontal direction for plate 8503. Positions are given from

the lower end to the top of the fission zone, whereas in horizontal direction the distance is given from the

center and extends about 30 mm in each direction. Obviously the swelling is very anisotropic according to

the anisotropic flux distribution in the respective irradiation positions. In all plates maximum swelling happens

at a vertical position of about 300 mm and a horizontal position of about -25 mm or +25 mm. The

measured swelling includes also the thickness increase due to the build up of an oxidation layer during the

course of the irradiation.

4.2.1 Plates without matrix-additive

Plate 8002 has been irradiated during 5 cycles up to a maximum fission density of 2.2 10 21 f/cm 3 . Fig 3 (left)

shows the smooth build up of the swelling during the first two cycles. Essentially the growth of the UMo

particles consumes the build-in porosity. This is followed by a sine like build-up of the thickness increase

during the next two irradiation cycle. For comparison Fig. 5 (left) includes the measured fission product density

on a relative scale. Whereas swelling at low doses follows very well the shape of the fission product rate,

the increase in swelling is enhanced in the centre of the plate for the last cycle(s). This nonlinear increase of

swelling at higher total doses becomes more evident in Fig. 3 (right), where the increase in swelling from the

4 th to the 5 th cycle is shown alone, now along different traces on the plate. Traces far away from the









plate 8002, no Si

longitudinal trace - 26 mm




plate 8002 - only F210

trace + 26 mm

trace 0 mm

trace -13 mm

trace - 21 mm


trace - 26 mm



swelling (µm)



swelling [µm]









0 100 200 300 400 500 600

height (mm)


0 100 200 300 400 500 600

plate height [mm]

Fig.3: Left: Swelling for plate 8002 after up to 5 cycles of irradiation measured at the horizontal position

-26 mm along vertical direction. Right: Differential swelling, i.e. increase in swelling during the last irradiation

cycle at different horizontal positions.

93 of 435

hot spot indicate a swelling according the local fission rate, traces near the hot spot indicate an increase in

thickness beyond linearity. At the hot spot a total swelling of 104 µm or 21% with respect to the original

thickness of the meat has been observed.

Plate 7003 with a density of 7.3 gU/cm 3 has been irradiated to a similar maximum fission density of 2.2 10 21

cm -3 . Now a considerably larger swelling of maximal 178 µm or 33% has been observed. Fig. 4 clearly depicts

the nonlinear and even completely irregular swelling at traces of high dose. Particular Fig. 4 (right)

indicates pillowing, but it has to be mentioned that no breaking of the cladding has been observed, i.e. no

fission products have been released.













plate 7003, no Si

longitudinal trace - 26 mm





? swelling, only cycle F217

plate 7003

trace + 26 mm

trace 0 mm

trace -13 mm

trace - 21 mm

trace - 26 mm

swelling (µm)













0 100 200 300 400 500 600

height (mm)

swelling [µm]







0 100 200 300 400 500 600

plate height [mm]

Fig.4: Left: Swelling for plate 7003 after up to 5 cycles of irradiation measured at the horizontal position

-26 mm along vertical direction. Right: Differential swelling, i.e. increase in swelling during the last irradiation

cycle at different horizontal positions.

4.2.2 Plates with matrix-additive Si (2 wt%) Plate 8503









plate 8503, 2% Si

longitudinal trace - 26 mm








? swelling plate 8503,

only cycle F210

trace + 26 mm

trace 0 mm

trace -13 mm

trace - 21 mm

trace - 26 mm

swelling (µm)




FD fit



relative FD [any units]

swelling [µm]













0 100 200 300 400 500 600

height (mm)


0 100 200 300 400 500 600

plate height [mm]

Fig. 5: Left: Swelling for plate 8503 after up to 5 cycles of irradiation measured at the horizontal position

-26 mm along vertical direction. Also shown is the measured final fission product density on a relative

scale. Right: Differential swelling, i.e. increase in swelling during the last irradiation cycle at different

horizontal positions.

Plate 8503 has been irradiated during 5 cycles up to a maximum fission density of 2.3 10 21 cm -3 . The swelling

depicted in Fig. 5 resembles very much that shown for plate 8002. At low doses, i.e. the first 3 cycles the

94 of 435

vertical swelling follows reasonably well the shape of the fission product rate, whereas the increase in swelling

is enhanced in the centre of the plate for the subsequent cycles. At the hot spot a total swelling of 93 µm

or 19% with respect to the original thickness of the meat has been observed. This is a slightly less thickness

increase than observed for similar total doses in plate 8002.














plate 8501, 2% Si

longitudinal trace - 26 mm



? swelling, plate 8501

only cycle F217

trace + 26 mm

trace 0 mm

trace -13 mm

trace - 21 mm

trace - 26 mm

swelling (µm)








swelling [µm]









0 100 200 300 400 500 600

height (mm)


0 100 200 300 400 500 600

plate height [mm]

Fig. 6: Left: Swelling for plate 8501 after up to 8 cycles of irradiation measured at the horizontal position

-26 mm along vertical direction. Right: Differential swelling, i.e. increase in swelling during the last irradiation

cycle at different horizontal positions.

The twin plate 8501 – see Fig. 6 - was irradiated for a total of 8 cycles or a maximum fission density of 3.3

10 21 cm -3 . It shows a maximum swelling at the hot spot of 323 µm or 66% of the meat thickness. This is more

than three times the maximum swelling of plate 8503. Also for this plate the vertical swelling follows for the

first 3-4 cycles very well the shape of the fission product rate, whereas the increase in sweling enhances

more and more with the cycle number. A clear pillowing is observable, but also for this extremely high irradiation

dose no breaking of the cladding has been observed, i.e. no fission products have been released.

4.3 Discussion of the swelling

Fig. 7 summarizes the swelling at the hot spot of all irradiated IRIS-TUM plates. The following is easily


o All plates retain the fission products even at highest burn-up.

o Swelling is minimal during the first 2 irradiation cycles, most probably due to the consumption of the

build-in porosity of about 8 vol.%.

o A more or less linear increase up to a fission density of about 2.0⋅10 21 cm -3 is followed by a steeper

and steeper increase in the course of adding up fission densities.

o Plates with Si addition show a reduced swelling when compared to those without Si addition.

And in comparison to other full size tests with UMo dispersive fuel:

o The swelling is higher than in IRIS-1 (also ground powder) or IRSI-3 (atomized powder), presumably

because of the higher heat load and subsequent higher temperatures during the IRIS-TUM irradiation.

o The “best” UMo plate with Si addition swells at a the target FD of 2.3⋅10 21 cm -3 by 22 %, which is

21 % more than the silicide fuel with a density ρ = 3 gU/cm 3 .

95 of 435

max. swelling fuel layer
















0 1 2 3

fission density fuel layer [10 21 cm 3 ]

Fig. 7: Comparison of the swelling at the hot spot of all IRIS-TUM plates. For comparison the maximum

swelling observed for the irradiation programs IRIS-1 (ground powder) [6], IRIS-3 (atomized powder) [9]

and IRIS-U 2 Si 2 (ρ = 3 gU/cm 3 ) [10] are also shown.

5 Post Irradiation Examination (PIE)

After about 1 year of cooling time the plates 8002 and 8503, both irradiated during the first 5 cycles, could

be transported to CEA-Cadarache, where small samples have been cut out from the top corner and along the

mfp of the meat zone. These have been transported to SCK-CEN, Mol, Belgium, where the samples have

been prepared metallographically, and optical and scanning electron microscope examinations have been

performed in hot cells .

Fig. 8 (top) shows optical microscopy images of samples taken from the top end of the meat zone, i.e. a region

of lower fission density. The shredded shape of the ground powder particles is clearly discernable. Dark

lines within the UMo particles are presumably oxidized zones. In the top-right image the Si precipitates in

the Al meat are visible. In both samples an interdiffusion layer, kown to be rich in Al, has been formed

around the UMo particles. Scanning electron microscopy pictures with larger magnification - not shown

here – show the distribution of the fission gas bubbles within the UMo particles mainly along grain boundaries.

No fission gas bubbles are observed in the interdiffusion layer. The bottom part of Fig 8 displays the

average thickness of the interdiffusion layer measured along the mfp. Data have been grouped into 3 zones:

thickness of the interdiffusion layer at the interface between cladding and meat, separately for the top and

bottom interface (top and bottom with respect to the sample orientation) and in the centre of the meat. This

interdiffusion layer forms during irradiation and is suspected to be related to the break-away swelling observed

in previous irradiation tests of UMo fuel plates like IRIS-2 and FUTURE [6].

The PIEs of plates 8002 and 8503 will be continued, in particular electron probe micro-analysis is planned.

Further, plates 8501 and 7003 with higher fission densities are awaiting their transport to hot cells, once their

radiation level has lowered to tolerable values. A few preliminary conclusions can already be drawn at the

actual state:

o In the mfp the matrix material is consumed to a very high extent.

o From the metallurgical preparation of samples along the mfp it can be derived that the irradiated

meat becomes extremely brittle, that means has a high tendency for developing cracks.

96 of 435

No Si 8002L pos 1 2.1 wt% Si 8503L pos 1














interface meat-cladding top

meat mid

interface meat-cladding bottom


0 10 20 30 40 50 60















interface meat-cladding top

meat mid

interface meat-cladding bottom


0 10 20 30 40 50 60


Figure 8: Top: Optical microscopy images of samples taken from the top end of plate 8002 (left) and

8503 (right). Bottom: Measured mean thickness of the Al rich interdiffuion layer along the mfp plane for

three different positions: at the top interface between meat and cladding, in the middle of the meat layer

and at the bottom interface between meat and cladding.

o The interdiffusion layer is – if at all - only slightly reduced in the samples containing additional Si.

o For the irradiation doses achieved in plate 8503 and 8002 the fission bubbles are accommodated in

the UMo particles mainly along grain boundaries.


For the first time large UMo dispersion fuel plates have been irradiated to very high burn-up – up to 88.3

LEU equivalent – and at high heat load of 260 W/cm 2 . No failure of the first barrier – the cladding – has

been observed, even at a thickness increase of 323 µm which corresponds to 66% of “swelling”. Large buildin

porosity delays the onset of linear swelling. During the irradiation, a period of almost linear increase of

thickness is followed by a steeper, non linear increase of thickness. In the most favourable case this nonlinear

increase begins at about 2.0 10 21 cm -3 , in the case of no additional Si at lower fission densities. The beginning

of this nonlinear increase can be seen most clearly in the time and spatial dependence of the sweling.

Fuel with Si added to the Al matrix swells a little less than that without Si additive. The microscope images

from samples of plate 8503 and 8002 yet do not give a clear indication why this is the case. Growth of the

interdiffusion layer is – if at all - only slightly hindered by addition of Si.

97 of 435

The progress achieved in this irradiation campaign is dominantly ascribed to the usage of ground powder.

Why ground powder shows a more controlled swelling than atomized powder? A final answer has to wait for

more detailed PIEs, as they are in progress. Certainly the ground particles have a defect density orders of

magnitude higher than that of atomized particles. This higher defect density – and we explicitly include oxidation

and additional impurities - form seeds for the nucleation of medium large fission bubbles, which again

prevents diffusion of fission gases into the interdiffusion layer.

In spite of the progress reported here, we are still far away from high density fuel (ρ ≥, 8 gU/cm3) which

withstands the high irradiation doses and rates as they occur in research reactors with highest neutron fluxes

like FRM II. Also the best behaving fuel plate 8501 is far away from satisfying safety criteria as they are

achieved in the present U 3 Si 2 fuel. For instant it has to be examined, how UMo fuel behaves under higher

heat load because it is to suspect, that irradiation at higher temperature in the UMo grains will enhance diffusivity

of the fission products. Fig. 7 gives a first hint on that. Both, IRIS-1 and IRIS-3 show lesser swelling

than IRIS-TUM, and in both cases the temperature in the UMo grain has been much lower.

Therefore, TUM and its partners aim at future irradiation of large scale UMo dispers test plates at heat loads

in the order of 400 W/cm 2 . Further it seems to be unrealistic to produce ground powder with 50% enrichment

on an industrial scale as necessary to produce the annual needs of FRM II fuel element production [11].

Therefore we have to come back to atomized powder, but now with different metallurgical treatment like

oxidization, addition of diffusion blockers like Si in Al and/or modified defect structure.


[1] A. Röhrmoser, W. Petry, N. Wieschalla, Reduced Enrichment Program for the FRM-II, Status 2004/05, RRFM

2005, Budapest, Hungary

[2] R. Jungwirth, W. Petry, W. Schmid, L. Beck, A. Bergmaier, Progress in Heav-Ion Bombardment of UMo/Al

Dispersion Fuel, RRFM 2008, Hamburg, Germany

[3] W. Schmid, R. Jungwirth, W. Petry, P. Böni, L. Beck, Manufacturing of Thick Monolithic Layers in Cathode

Erosion Process, RRFM 2008, Hamburg, Germany

[4] C. Jarousse, P. Lemoine, P. Boulcourt, kw. Petry, A. Röhrmoser, Monolithic UMo Full Size Prototype Plates

for IRIS 5 Irradiation, RRFM 2007, Lyon, France

[5] A. Röhrmoser, W. Petry , Reduced Enrichment Program for FRM II, Actual Status & a Principal Study of

Monolithic Fuel for FRM II , RRFM 2006, Sofia, Bulgaria.

[6] P. Lemoine, J.L. Snelgrove, N. Arkhangelsky, L. Alvarez, UMo Dispersion Fuel Results and Status of Qualification

Programs, RRFM 2004, Munich, Germany

[7] G.L. Hofman, Y.S. Kim, M.R. Finlay, J.L. Snelgrove, S.l. Hayes, M.K. Meyer, C.R. Clark, F. Huet, Recent

Observations at the Post Irradiation Examination of low Enriched UMo Miniplates Irradiated to High Burnup,

RRFM 2004, Munich, Germany

[8] S. Dubois, F. Mazaudier, H. Palancher, P. Martin, C. Sabathier, M. Ripert, P. Lemoine, C. Jarousse, M. Grasse,

N. Wieschalla, W.Petry, A. Röhrmoser, Development of UMo/Al Dispersion Fuel: an Oxide Layer as a Protective

Barrier around UMo Particles, RERTR 2006, Cape Town, South Africa

[9] P. Lemoine, M.C. Anselmet, S. Dubois, French CEA Programs for the Development and the Qualification of

High Density Fuel for the JHR Project, RRFM 2008, Hamburg, Germany

[10] K. Böning, W. Petry FRM II Test Irradiations of Full Sized U 3 Si 2 -Al Fuel Plates up to Very High Fission Densities,

submitted to NIM A

[11] Communication by AREVA-CERCA, 2007

98 of 435




Department of Nuclear Materials

Chilean Commission for Nuclear Energy - CCHEN

Amunategui 95, Santiago 6500687, Chile

Author contact: jmarin@cchen.cl


CCHEN has developed, fabricated, and qualified MTR type fuel since 20 years,

all of them have been loaded in both Chilean research reactors. Recently, more

than 48 LEU uranium silicide fuel assemblies have been delivered to the

Chilean research reactor La Reina- RECH-1. New local development deals with

U-Mo fuel where, several activities has been completed such as casting of U-Mo

alloys, phase stabilization studies, techniques for powder production, interaction,

interdiffusion and out of pile swelling studies of standard and modified UMo/Al

system. In parallel, for fission Mo, UMo foil targets are under development in the

framework of an IAEA’s Coordinated Research Project, and some of the

achievements are included in this paper.

1. Introduction

CCHEN has been involved in development of fuel for research reactors since 1980’s. Actually

48 LEU high density dispersion fuel assemblies have been fabricated of U 3 Si 2 LEU with a

uranium density of 3.4 g/cm 3 for La Reina research reactor - RECH-1 (over 800 LEU fuel

plates). The work was launched in 1987 when was necessary to disassemble and reassemble

31 fuel elements for the other Chilean research reactor, RECH-2 at Lo Aguirre.

These task included inspection, X-ray examination of meat distribution, plates cold

examination, redesign of some fuel parts, and re-assemble of fuel elements.

In 1998 new LEU fuel was designed for conversion of RECH-1, starting with loading in the

reactor core four test fuel elements for irradiation behaviour surveillance. No fuel defects were

observed and no performance problems were observed. Complementary, a Chilean test fuel

element was fabricated for and irradiated in HFR, Petten, The Netherlands [1], achieving high

burn up performance and an excellent PIE results.

CCHEN continues on the development of new fuel designs and new fuel technologies. In 2003

has started a programme for developing U-Mo compound. As a result of it, several activities

have been carried out, [3], [4] such as casting of U-Mo alloys with Mo contents from 7 to

10wt%, phase transformations, gamma phase stabilization studies and several techniques for

powder production, including cryogenic milling, high energy milling and grinding milling of

machined chips. Particularly, interesting results from efficiency point of view, were obtained

through hydration – milling – dehydration or HMD process applied to an UMo with special

condition, deformed by cold rolling and crushed by impact. Also, they were carried out

interaction, interdiffusion and out-of-pile swelling studies. Last year, UMo foil manufacturing,

by means rolling, is under development. The final stage on this programme considers under

irradiation evaluation of dispersed and monolithic miniplates.

Based on the irradiation results, is necessary to evaluate the different solutions aimed to

stabilize an interaction layer zone produced by reactions of UMo fuel with standard Al matrix

[5]. as it is generally accepted. Among the different options studied, in this paper is included

an experimental evaluation of the effects of Si addition to Al matrix and/or addition of a third

element (Si, Zr, and Ti) to the UMo fuel.

99 of 435

In the other hand, suppressing the Al matrix employing just monolithic UMo as fuel meat

seems to be an promising alternative solution based on the hypothesis that this fuel type can

solve the U-Al interaction problem. In monolithic fuel the entirety of the fuel meat is comprised

of a single foil of the fuel alloy. This fuel configuration represents the optimum in fuel meat

density. The highly reduced fuel surface/volume relation and the fact that fuel-aluminium

interfaces are in the cooler region of the plates should minimize the fuel-aluminium reaction

[6]. Following this trend, CCHEN has started the development of technologies to obtain UMo

foil as the first stage of the final aim; to have their own methodology for UMo monolithic fuel


2. Experimental activities

UMo ingots were produced using an induction furnace placed inside a multipurpose chamber

with controlled atmosphere. Ingots were obtained by melting natural uranium and Mo metal

inside a high density alumina crucible and poured into a graphite mould. After casting, the

ingots were annealed at 950°C by 24 hours in vacuum atmosphere (10–5 Torr) and cooled in

argon in order to induce micro structural homogenisation and residual alpha phase

transformation for gamma phase stabilization.

For dispersion fuel and interaction studies, fuel grade fine powder was necessary. To produce

these powders, four techniques have been evaluated: Hydration - Milling – Dehydration (HMD),

cryogenic grinding and mechanical grinding using high speed rotating blades made of several


Fuel/Matrix interaction tests and out of pile swelling studies required more than twenty test

miniplates. These dispersion miniplates, of pure UMo or modified by third element addition

dispersed in Al matrixes, pure or alloyed with silicon, were manufactured employing the

powder metallurgy conventional method.

Interaction tests, which results have been reported previously [2]-[4], involved metallographic

preparation and inspection of samples extracted through punching of miniplates and annealed

in quartz capsules vacuum sealed. After annealing for diffusion tests, the samples were

analysed with SEM and EDS micro analyses of interaction layer (IL) regions formed by UMo

particles surrounded by aluminium matrixes. Following the kinetics considerations given by

the TTT curves of the U7Mo alloy, thermal annealing were performed to 550 °C for times up to

48 hours.

Based on interaction tests results and according to our experimental UMo program [7] the

following step was to develop the swelling tests were performed. Taken into account that the

swelling phenomenon produces thickness increasing, volume changes in miniplates can be

assumed as thickness changes. Then, for these studies, out of pile tests were applied to

dispersion miniplates. Air annealing carried out at 500°C followed by immersion density

measurements were applied after each annealing treatment to all miniplates. This

methodology permits leads to obtain global increasing trends for each fuel/matrix combination

and it’s comparisons with reference U 3 Si 2 miniplates.

In relation to monolithic UMo fuel, foil manufacture starts with casting of U-7Mo ingot. This

casting was performed employing an induction melting furnace using a gravity pour into a

graphite closed mould. Prior to hot rolling, the ingot is removed from the mould and annealed

in a vacuum atmosphere to homogenize the microstructure. Because the ingot has enough

thickness, the surfaces were machined under water to remove casting defects and to improve

the surface ruggedness. Finally, the ingot was divided in four sections to produce equal a

certain number of coupons. UMo foils for the monolithic test plates were produced by hot

100 of 435

olling of these coupons, which were sealed in a plain low carbon steel can (A37-24ES) to

isolate the UMo alloy from the atmosphere during processing. The coupon/steel assemblies

were repeatedly heated to 680°C and rolled at this temperature to reduce the thickness of the

fuel meat from 5,7 mm until an average value of 0,49 mm. In a previous rolling test the

thickness of a UMo fuel alloy was reduced from 2,5 mm to 0,32 mm also using only hot


3. Results and Discussion

3.1 Casting and microstructure homogenisation of U-7% wt Mo alloy

(a) (b) (c) (d)

Figure 1. Optical microscopy and SEM fracture surfaces (at room temperature) of U - 7% wt

Mo alloy (a), (b) As cast, (c), (d) homogenised by vacuum annealing.

In a cast alloy, Fig. 1(a), the presence of two phases, a light matrix of gamma phase and a

second phase, darker, precipitated in the gamma grain boundaries is observed and

accordingly to X-RD analyses, it corresponds to alpha phase. In 1(c) image, the presence of

the second phase is very few, product of its dissolution an homogenisation during the thermal

treatment. Related to fractography analyses of images (b) and (d) of as cast and annealed

samples respectively, the predominant fracture mechanisms corresponding to transgranular

ductile fracture via micro void coalescence combined with minor evidences of cleavage along

crystallographic planes (brittle fracture). According to Charpy tests carried out from –120 and

+20°C, U-7% Mo alloy shows a brittle-ductile transition temperature in the range of 10 to 15

°C. Any mechanic method for powder production could overcome easily this temperature,

promoting ductile fracture conditions.

3.2. Powder production

(a) (b) (c)

101 of 435



Figure 3. SEM images of powder produced by several methodologies from U7Mo previously

homogenised by annealing, cold rolled and crushed (R&C): (a) U7Mo-CTT-R&C cryogenic

grinding, (b) U7Mo-CTT-R&C grinded by Ti blades, (c) U7Mo-CTT-R&C grinded by WC blades,

(d) U7Mo-CTT-R&C Hydrated and (e) U7Mo-CTT-R&C Hydrated and Dehydrated.

In general terms, all grinding methods for powder production results with very low efficiency

and in grinding with WC blades, small amounts of Co contamination was detected in powders.

In the other hand, HMD process shown be efficient, specialty applied to cold rolled and

crushed UMo alloy. Anyway, in order to produce UMo powder for subsequent dispersion test

miniplates, enough amounts of UMo alloy were produced by means mechanical grinding using

Ti blades. The next stage will be powder production and characterization of UMo-Ti and UMo-

Zr alloys in R&C condition using HMD method.

3.3. Interaction tests in dispersion fuel miniplates

Figure 4. Morphology of Interaction

Layers after 48 hours/550°C annealing.

Comparison between UMo/Al (a)

and modified UMo+Si/Al (b).



(a) (b) (c) (d)

Figure 5. SEM images of (a) UMo, (b) UMo+Si, (c) UMo+Ti y (d) UMo+Zr particles dispersed

in Al matrixes after 48 hours/500°C (vacuum) annealing.

SEM combined with EDS concentration profiles analyses applied to UMo samples shown in

figure 5 reveal the occurrence of mechanisms of interdiffusion of U and Mo atoms from the fuel

particles toward UMo/Al interlayer zone. Evidences of Al atoms migration from the matrix

toward the outlying areas of UMo particles, where combines with U to form binary aluminides

(UAlx) or ternary compound U-Mo-Alx were detected. Towards the centre of the fuel particles,

also the presence of Al was detected in UMo+Si sample (b), which confirms the occurrence

of the interdiffusion phenomenon in the interlayer zone. The addition of a third element allows

to delay the interdiffusion phenomenon or at least to have some influence on the kinetics of

growth of the interface region. These effects are evident when observing the thickness and

morphology of the interface regions. (Figure 4). These results confirm the hypothesis outlined

in previous works [2] in the sense that the second phase formed by the addition of the third

element, and it’s preferable location in grain boundaries of UMo, it could constitute barriers to

diffusion or atomic migration of the UMo/Al system. Compositions analyses verify the

spontaneous migration of atoms of Si present in the Al-6061 cladding (0,6% wt%) toward the

particles of UMo where, probably it form compounds with U and/or Mo. For the UMo with Si

102 of 435

addition, this diffusive phenomenon is also verified but the presence of Si in the fuel particles

makes less strong the concentration gradient and the interdiffusion of this element, appears in

some grade, controlled. In consequence, if silicon atoms are presents in the UMo particles

and Al matrix, it’s mobility appears slowed, thus they can constitute diffusion barriers by

themselves. Other authors suggest that the addition of Si just has effect in the Al matrix [8]-

[9], while the outlined hypothesis is coherent with other authors [10] in the sense that the

addition of Si to the fuel phase (UMo) can help to avoid or limit the interdiffusion due to the

action of precipitates (second phase) and also as effect of decreasing of the silicon

concentration gradients. On the other hand, the Zr addition produces a very fine and

homogeneous dispersion of this element in the entire UMo microstructure, without preferential

location or segregations. By means of this mechanism the Zr could be causing restrictions to

movement of dislocations and vacancies and/or formation of precipitated in the grain

boundaries, all mechanisms that constitute barriers to the diffusion. Titanium act in very

similar form inside the UMo particles, with the difference that Ti experienced preferential

location in the interface, probably, for their affinity with the Al. The mechanisms for which the

third element is capable to control the thickness and the composition of the reaction layer are

relatively clear and keep certain relationship with disincentive, for some mechanism, the

atomic mobility.

3.4. Out-of-Pile swelling tests applied to dispersion fuel miniplates

Figure 6. Volume increase v/s annealing time

for UMo-Me alloy dispersed in Al matrixes

Summarized result for 500°C

According to the swelling test results, the volume

changes are directly related with the uranium

density, and in general, third element additions

result in improvements in swelling behaviours.

Comparatively, the best result was obtained for

dispersed miniplates made of UMo/AlSi-Mix

followed by UMoSi/AlSi alloy, both slightly better

than U 3 Si 2 for similar range of uranium density.

The volume increase for unmodified UMo/Al

system achieves levels almost three fold higher than those achieved with U 3 Si 2 ; however for

system UMo/AlSi mix, these undesired behaviour was reduced to values equivalent or slightly

lower than for U 3 Si 2 .

3.5. Development of monolithic U-Mo fuel

(a) (b) (c) (d)

103 of 435

(e) (f) (g)

Figure 7. Sequence of UMo foil manufacture in CCHEN. (a) Ingot casting, (b) underwater

machining of ingot surfaces, (c) UMo ingot, (d) ingot divided in four coupons, (e) UMo coupon

sealed in steel assembly for hot rolling, (f) foils manufactured by hot rolling (0,49 mm

thickness) and (g) exploratory test of cold rolling (0,32 mm thickness).

In figure 7, the sequence to obtain UMo foils includes ingot casting, machined and cutting in

four sections or coupons, which were encapsulated in steel assemblies. The coupons are hot

rolled and, after several rolling steps, the thickness was reduced from 5,7 to 0,49 mm

(91,4%). In a previous test, an UMo foil with thickness of 320 µm was achieved by hot rolling,

with total reduction of 86,9%. For the next step, new steel cans will be required to continues

hot rolling until reduce foil thickness to about 180-220 µm. Finally, limited cold rolling (5% or

less) would be applied to UMo foils just to improve the surface finish and stiffness increasing.

The following step would be UMo-10wt% alloy foil manufacture and, finally with the U7Mo and

U10Mo foils, to select a suitable UMo/Al6061 bonding methods in order to begin the

manufacture of monolithic fuel plate.

4. Conclusions

Based on results of characterization and testing described above for dispersion fuel miniplates

and monolithic fuel, the following conclusions can be drawn:

The volume changes are directly related with the uranium density and for similar annealing

condition, the unmodified UMo/Al system exhibited swelling levels almost three times higher

than those achieved with U 3 Si 2 . However for the system UMo/AlSi mix, this undesired

behaviour was reduced to values equivalent or slightly lower than for U 3 Si 2

Out of pile swelling results indicates that the modification by silicon addition is more effective

in the matrix than in the fuel alloy.

Manufacture of UMo foil for monolithic fuel has been achieved successfully.

5. Acknowledgements

The authors are grateful for the support received from CCHEN through it’s Nuclear Materials

Department and specially from technical staff members of Fuel Element Plant – PEC.

6. References

[1] P. M. Thijssen, J. Marin, J. Lisboa, L. Olivares, F. J. Wijtsma, R. H. J. Schuring and K. Bakker

“Irradiation Qualification of a Chilean Test Fuel Element.” Proceedings of the 10th International

Meeting on Research Reactor Fuel Management, RRFM, Sofia, Bulgaria, April 2006

[2] Luis Olivares, Mario Barrera, Jaime Lisboa, Jorge Marin, Klass Bakker, Fred Wijtsma,

“Results for the recent activities of reduced enrichment program for research reactors in Chile”

International Meeting on Reduced Enrichment for Research and Test Reactors, RERTR, Cape

Town, South Africa, 30 Oct-2 Nov, 2006.

[3] D. Fernández, L. Olivares, J. Lisboa, J. Marin, “Fragilización y Obtención de Polvos de

Aleación U-7Mo Mediante Hidruración-Molienda-Deshidruración” Jornadas SAM/CONAMET 2005,

MEMAT 2005, Mar del Plata, Argentina, October 2005

[4] C. Pozo, J. Lisboa, L. Olivares, and J. Marin, “Molienda Mecánica de Aleación UMo.

Interacción del Sistema UMo/Al” 4º Congreso Binacional de Metalurgia y Materiales, Santiago, Chile

28 Nov – 1 December 2006

104 of 435

[5] D. M. Wachs, R. G. Ambrosek, G. S. Chang, M. K. Meyer “Design and Status of RERTR

Irradiation tests in the Advanced Test Reactor”. International Meeting on Reduced Enrichment for

Research and Test Reactors, RERTR, Cape Town, South Africa, 30 Oct-2 Nov, 2006.

[6] C. R. Clark, G. C. Knighton, M. K. Meyer, G. L. Hofman. “Monolithic Fuel Plate Development

at Argonne National Laboratory” International Meeting on Reduced Enrichment for Research and

Test Reactors, RERTR, Chicago, Illinois, USA, October 5-10, 2003

[7] J. Marin, J. Lisboa, L. Olivares, M.A.C. van Kranenburg and F.J. Wijtsma, “Under Irradiation

Qualification of a Chilean Test Fuel” Proceedings of the XXVII International Meeting on Reduced

Enrichment for Research and Test Reactors, Boston, Massachusetts, USA, 6-11 November 2005.

[8]. G. L. Hofman, Yeon Soo Kim, Ho Jin Ryu, M. R. Finlay, D. M. Wachs, “Improved Irradiation

behaviour of uranium/molybdenum dispersion fuel”. Proceedings del 11th International Topical

Meeting of Research Reactor Fuel Management, RRFM, Lyon, France, 11-15 March 2007.

[9]. C. Komar Varela, M. Mirandou, S. Aricó, S. Balart, L. Gribaudo “The reaction zone in the

system U-Mo/Al6061 related with the decomposition of γ U-Mo”. Proceedings del 11 th International

Topical Meeting of Research Reactor Fuel Management, RRFM, Lyon, France, 11-15 March 2007.

[10]. D. M. Wachs, R. G. Ambrosek, G. S. Chang, M. K. Meyer “Design and Status of RERTR

Irradiation tests in the Advanced Test Reactor”. Proceedings del International Meeting on Reduced

Enrichment for Research and Test Reactors, RERTR, Cape Town, South Africa, 30 Oct-2 Nov, 2006.

105 of 435




SCK•CEN, Nuclear Materials Science Institute, Boeretang 200, B-2400 Mol, Belgium.


CEA-Cadarache, DEN/DEC, 13108 St Paul Lez Durance Cedex, France


CEA-Saclay, DEN/DSOE – 91191 Gif sur Yvette – Cedex – France

In the framework of the IRIS-3 irradiation, a full size, flat plate containing atomised

U(Mo) dispersion fuel in an aluminum matrix with addition of silicon, has been

irradiated in the OSIRIS reactor. The microstructural analyses of the irradiated fuel

from this project was performed at the hot laboratory (LHMA) of SCK•CEN in Mol,

Belgium. The obtained optical microscopy, scanning electron microscopy and electron

probe microanalysis results provide further insight in the effect of adding silicon to the

aluminum matrix.

1 Introduction

Fuel plate U7MV8021 was one of the 4 plates of the IRIS-3 experiment, irradiated in the OSIRIS

reactor [1]. The cladding of this plate is made of AG3NE Al-Mg alloy (2.81 wt% Mg) and the

meat consists of U7.3wt%Mo particles dispersed in an aluminum matrix to which 2.1 wt% Si has

been added. The fissile material density is 7.8-8.0 g U tot /cm 3 and the uranium enrichment is

19.8% 235 U. The fuel plate was kept in the reactor during 7 irradiation cycles (130.6 full power

days) and submitted to a heat flux of maximum 200 W/cm 2 , while the surface cladding

temperature is kept below 85 °C. At its EOL, the plate had an average burnup of 48.8 % 235 U

(3.4×10 21 fissions/cm 3 U(Mo)) with a peak burnup at the maximum flux plane of 59.3 % 235 U

(4.1×10 21 fissions/cm 3 U(Mo)).

After unloading and non destructive characterization at CEA, a slice of the fuel plate was cut at

the maximum measured burnup plane and transferred to the Laboratory for High and Medium

Activity (LHMA) at the SCK•CEN site for microstructural examination. The results of the PIE can

be compared to the microstructure results obtained on the FUTURE [2] and IRIS-2 plates [3].

2 PIE of IRIS-3

The slice cut from fuel plate U7MV8021 was subdivided in three samples. All samples were

embedded in the same mount in such a way that the complete section of the fuel (meat and

cladding) could be observed.

Fig. 1 Composite of micrographs showing a transverse cross section of the plate

at maximum flux plane. On the image the 8 analysis positions are indicated.

The composite image of optical macrographs gives an overview over almost the complete plate

width (fig. 1). It shows a homogeneously thick (~600 µm) meat layer in-between the cladding.

From the more detailed optical images, it is generally observed that plenty of the aluminum

matrix is left and the silicon particles dispersed in the matrix can be readily observed (fig.2). At

the interface between the meat and the cladding, a string of particles can be seen. The other

106 of 435


particles seen inside the cladding are Mg 2 Si precipitates, known to exist in AG3NE. It should be

noted that the matrix bordering the interface appears to contain less Si precipitates.

Interaction between the matrix and the fuel

particles has resulted in a layer formed around

each of the particles. The fuel agglomerates

contain numerous fission gas related bubbles,

all having roughly the same size. It is also seen

that near to some fuel kernels, the silicon

particles in the matrix have disappeared.

The obtained SEM images are used to quantify,

by image analysis, the surface fraction occupied

by the different phases, i.e. the U(Mo) fuel

particles, the interaction layer (IL) and the

matrix. The results are graphically represented

fig.3a. It is observed that the surface fraction

Fig. 2 Optical microscopy image (at pos 4)

showing a string of particles at the

interface with the cladding. The Si particles

added to the matrix are readily seen.

occupied by the different phases is nearly

constant over the complete width of the plate,

apart from the first (pos1) and last measuring

Fig. 3 Measured surface fraction of the different phases (a) and the thickness of the IL (b).

position (pos8) which can be related to the lower temperatures at the sides of the plate. It is

found that the surface fraction of U(Mo) fuel particles is ~55%, for the interaction layer ~ 22%

and for the matrix ~ 23%. It should be noted that the

values obtained for the surface fraction occupied by the

U(Mo) fuel also include the fission gas related bubbles.

Image analysis on the backscattered electron image

show that the bubbles occupy approximately 1% of the

surface of the fuel particle.

The measurements of the interaction layer thicknesses

are based on at least ten, randomly chosen points at

each position and location (interface outer cladding/meat,

middle of the meat, interface inner cladding/meat). It is

seen from fig.3b that also for the thickness of the

interaction layer a rather constant value, on average ~5

µm, is obtained over nearly the complete plate width

(pos 2 to 8). The lower values obtained at pos 1 and 8

are in agreement with the lower surface fractions found.

Fig. 4 SE image of some fuel kernels

revealing and the asymmetric

thickness and jagged edges of the IL.

107 of 435


The detailed secondary electron (SE) images of

the fuel kernels show that the layer thickness

around the particles is sometimes very

asymmetric and that the surface of the IL at the

interface with the matrix, has a jagged

appearance (fig.4). Furthermore, the typical

cellular structure of the U(Mo) fuel is reflected in

the distribution of the fission gas (FG) bubbles.

The secondary and backscattered electron

images show that the fuel particles contain

Fig. 5 SE image and x-ray maps covering an area

in the meat (pos 5). The result of the quantitative

linescan defined in the SE image is given in the

bottom graphs.

Fig. 6 SE image and x-ray maps covering a part

of a fuel particle (at pos 2). The results of the

quantitative linescans defined in the SE image

are given in the bottom graphs.

numerous fission gas related bubbles, all having

approximately the same size (100-300 nm). In

some fuel particles these (visible) bubbles are

located on the U(Mo) cell boundary, while in

others they can also be observed in the cell.

The Al, U and Mo X-ray maps obtained by

EPMA show the uneven growth of the IL. It is

seen that at some positions on the fuel kernel

periphery (point D in fig. 5) almost no interaction

between the fuel and the Al of the matrix has

taken place and a higher Si concentration is

measured. It should be noted that at this position

the typical fission product (Xe) related halo is

108 of 435


more blurred, as no sweeping of the fission products by the growing IL has occurred. It appears

that even some Si particles are present in the IL.

Quantification over the line as defined in the SE image of fig. 5 shows prior to point A a nearly

pure Al matrix with some Si particles dispersed in it (e.g. the peak at ~7 µm). Between point A

and C an IL is present. No large quantities of Si are measured in the IL except at point B, but it is

believed that this is a Si particle wedged in-between two IL’s as can be observed in the Xe map.

The typical halo of fission products around the kernel, usually seen as steep increase in the FP

concentration, is not reflected in this line scan but could be seen in others. The Nd signal

gradually reduces outside the fuel kernel boundaries into the Al matrix, while the drop in the Xe

signal inside the fuel particle indicates the loss of this fission gas by opening of the bubbles

during sample preparation. At point D no U-Al IL is measured but a small increase in the silicon

concentration (from 0 to 1 wt%) around point D is seen, as also visible in the Si mapping.

This is also observed in the linescan L2 over a fuel particle at position 2 (fig.6). At point e, very

limited U-Al IL can be measured but an increase in the Si concentration is observed. The

linescan L1 over the IL shows again that it does not contain large quantities of Si. A gradual

decrease of Si content from approx 1.5 wt% to nearly 0 over the IL (between point a and b) is

seen. Also here the typical sharp halo around the fuel particles is not reflected in the linescans.

Furthermore, it is interesting to see that, inside the fuel particles, patches of Xe are observed

and measured (between point c and d). These patches are matched by those U(Mo) cells that

are optically free from bubbles. It is almost certain that these Xe patches reflect the nanosized

bubbles ordered on a superlattice as observed by TEM [4].


3 Discussion

The microstructural PIE results show that the fuel plate has undergone the irradiation without

Vol% U(Mo)

Vol% IL










any important detrimental effects. The

typical features expected in irradiated

atomised U(Mo) fuel can be observed.

During the irradiation, an interaction layer

Vol% Al



22.5 has grown around each of the fuel particles

and in the fuel kernel, equisized fission gas

Thickness IL




related bubbles can be seen. No large

Max FD f/cm 3 meat

1.41×10 21

2.7×10 21

4.1×10 21

quantities of crescent moon shaped

Max BU % 235 U



porosities at the IL/matrix interface,

Max heat flux W/cm 2 340


201 indicating accumulation of fission gas and

Table 1 Surface fractions of the different phases considered responsible for fuel plate

found in the PIE of IRIS-2, IRIS-3 and FUTURE and

some irradiation characteristics .

pillowing, have been observed.

The measured surface fractions (table 1)

occupied by the different phases show again

that the IL has grown at the expense of the

Al matrix, while the reduction of the fuel kernel volume


is compensated by their swelling. Compared to the



IRIS-2 and FUTURE experiment (table 1), plenty of Al


matrix is left. Consistently, also in the thickness of the 9

IL a decrease can be observed, compared to the



obtained results in the other irradiations. If one looks at 7

the relationship between the maximum heat flux the 6 IRIS3

fuel has seen during the irradiation and the thickness of 5

the IL (Fig.7) a clear correlation can be seen. This

150 200 250 300 350

supports the notion that the thickness of the IL is at

least mainly temperature driven.

The 2.3 wt% Si added to the Al matrix is observed as

IL thikness (µm)

Max heat flux W/cm 2

Fig. 7 IL thickness as function

of the maximum heat flux.

109 of 435


particles randomly dispersed in the matrix, which is expected since Si is insoluble in Al. At the

interface between the cladding and the meat, a string of particles is seen. It is also viewed that

the matrix at the interface between cladding and meat contains less Si particles. One could

therefore assume that the string of particles might consist of silicon particles originating from the

matrix, possibly forming a secondary precipitate with the dissolved Mg from the AG3NE (2.81

wt% Mg) cladding. Also close to some of the fuel particles, the matrix appear to contains less

silicon particles. It is believed that this is the result of the destruction of the silicon particles by

fission fragment tracks (irradiation assisted dissolution).

The effect resulting from the addition of Si to the Al matrix is best seen in the EPMA

measurements. The X-ray maps show that the asymmetry in the IL thickness is related to the

presence of Si at the interface with the U(Mo) fuel kernel. It appears that only at those positions

where a Si particle was near to the fuel at the start of the irradiation, little IL has grown,

supporting the fact that the affinity of U for Si is larger than for Al [5, 6]. This results in the

formation of a U(Mo)-Al(Si) interaction layer away from Si particles or the formation of a Si-rich U

phase close to those particles. In case a Si-rich U phase has formed, no growth of an U-Al IL is

observed, supporting the notion of a U-Si layer as an anti-diffusion barrier [5, 6]. At positions at

the fuel kernel periphery that are not close to an Si particle, it appears that a near to "normal"

(U,Mo)(Al,Si) 4 (based on several quantifications) IL has grown. No large concentrations of Si in

the layer are measured which is contradictory to the result found in the out of pile experiments

[7]. A possible explanation for this difference could be the temperature at which both processes

have taken place. In case a Si particle is reached by the IL, at first the IL will incorporate the

particle (grow around it). This causes part of the jagged appearance of the outer periphery of the


The measured patches of Xe inside the fuel, show the stability of the nanosized bubbles even at

higher burnup. But the fact that only a few of such intact U(Mo) cells are seen could point out

that in some cases a critical concentration is reached after which the nanobubbles could

agglomerate to larger (i.e. 100-300 nm) stable bubbles, which are no longer on an ordered


4 Conclusion

The irradiation of AG3NE cladded fuel plates containing atomized U(Mo) powder dispersed in

an Al-Si matrix up to an average burn-up of 48.8 % 235 U has been successful.

The addition of 2.1 wt% Si to the Al matrix seems to have a positive result on the thickness of

the interaction layer, but only if there was close contact between the silicon particle and the UMo

fuel at the beginning of the irradiation.

5 References

[1] S. Dubois, J. Noirot, J. M. Gatt, M. Ripert, P. Lemoine and P. Boulcourt in: The proceedings of the 11th

International Topical Meeting on Research Reactor Fuel Management (RRFM), Lyon, France (2007).

[2] A. Leenaers, S. Van den Berghe, E. Koonen, C. Jarousse, F. Huet, M. Trotabas, M. Boyard, S. Guillot, L. Sannen

and M. Verwerft, J. Nucl. Mater. 335 (2004) 39-47.

[3] F. Huet, J. Noirot, V. Marelle, S. Dubois, P. Boulcourt, P. Sacristan, S. Naury and P. Lemoine in: The

proceedings of the 9th International Topical Meeting on Research Reactor Fuel Management (RRFM), Budapest,

Hungary (2005).

[4] S. Van den Berghe, W. Van Renterghem and A. Leenaers, accepted for publication in J. Nucl. Mater. (2008).

[5] A. Leenaers and S. Van den Berghe in: The proceedings of the 29th International Meeting On Reduced

Enrichment For Research And Test Reactors, Prague, Czech Republic (2007).

[6] A. Leenaers and S. van den Berghe, submitted for publication in J. Nucl. Mater. (2007).

[7] M. I. Mirandou, S. Arico, L. Gribaudo and S. Balart in: The proceedings of the 27th International Meeting on

Reduced Enrichment for Research and Test Reactors (RERTR), Boston, USA (2005).

110 of 435






13108 Saint Paul Lez Durance - France



91191 Gif sur Yvette - France


According to the latest international studies on UMo/Al dispersed fuel, Si and Ti

seem to be good candidates to reduce the interaction zone that appears between

the fuel particles and their surrounding matrix. This paper gives a better

understanding of the influence of Si and Ti on the U-Mo-Al system. The UMo

based raw materials have been arc melted and then widely used in diffusion

couples with Al based matrix. Si and Ti are respectively added in the range of [0.3-

12 wt%], in Al, and [1-2 wt%], in UMo. The interdiffusion experiments were

performed between 400°C and 550°C. Results of these experiments are mainly

based on the microstructural and physico-chemical characteristics of the

interaction products. Techniques used in this study are : arc melting, optical

microscopy, SEM, EDS, XRD and micro-hardness tests (Vickers).

1. Introduction

UMo/Al dispersed fuel is developed as high-uranium-density fuel in order to convert

Materials Testing Reactors (MTR) cores, currently working with U 3 Si 2 or UAl x fuel. This

conversion is foreseen to fulfil requirements of nuclear treaty of non-proliferation limiting the

use of 235 U in fuel to 20% in weight. In operating conditions, the reaction between UMo

particles and the Al matrix results in a large interaction zone [1-2] that surrounds the particles

and that sometimes leads to the failure of the fuel element because of its poor irradiation

behaviour (large porosities development, leading to pillowing and sometimes failure). That is

the reason why studies, aimed to stabilize and minimize (or avoid) the interaction zone

between fuel particles and the matrix, are performed by several teams. Remedies consist in

modifying the interaction layer (IL) composition [3-4] and thickness by adding a new element

either in the matrix or in the fuel. Based on thermodynamic calculations [5], on previous outof-pile

diffusion studies [6-7-8-9] and on latest irradiation tests [10-11-12], additions of Si into

the Al matrix and/or Ti in the fuel seem to be promising solutions [13].

2. Experimental details

2.1 U-Mo-Al-Si system

UMo alloys

Arc melted ingots of UMo, containing 7 or 10 wt.% Mo, were supplied from AREVA-CERCA ♦

fuel manufacturer. Thermal annealing (900°C, 72h, secondary vacuum) followed by an

helium quenching (2000°C/h) have been performed in order to homogenize the Mo content

and to retain the metastable γ phase of uranium.

♦ AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and SIEMENS company

111 of 435

Al alloys

Aluminium alloys have been chosen with a Si content ranging from 0.11 to 12 wt.%, known to

be the eutectic composition. These alloys compositions are detailed in Tab 1.

Al alloy 1050

Al98-Si2 4043 4343 4045 4047

(AlSi2) (AlSi5) (AlSi7) (AlSi10) (AlSi12)

Si (wt.%) 0.11 2 5 7.4 10 12

Other (wt.%) 0.21 Fe 0.29 Fe



Diffusion couples

Tab 1 : Al alloys compositions.

Diffusion couples are prepared with samples of approximately 2 x 5 x 5 mm 3 , cut out from

UMo ingots or Al alloys foils. Both parts are mechanically polished (grinding paper) and

chemically etched in diluted nitric acid before the annealing, in order to eliminate surface

contamination and oxide layer. Then the two parts of the couples are placed in intimate

contact and maintained under compressive stresses during the thermal treatment, thanks to

a clamping device. Following kinetics data given by the TTT curves [14] of UMo alloys,

thermal annealings were performed between 450 and 550°C for 0,5 to 3 hours, in order to

avoid or limit the influence of the eutectoid transformation of UMo. Thermal treatments were

performed under Ar + 5 % H 2 atmosphere.

2.2 U-Mo-Ti-Al system

The whole work performed on the U-Mo-Ti-Al system will not be describe here. We have

chosen to focus on two points :

- the U-Mo-Ti alloys elaboration,

- the UMo/Al-Ti interaction experiments, by thermal annealing.

U-Mo-Ti alloys elaboration

Two types of elaboration methods were used :

- arc melting (under an argon partial pressure),

- induction melting (under secondary vacuum).

In both cases, samples of about 1 g were obtained from an UMo8 (8 wt.% Mo) ingot supplied

by CEA and pure titanium and molybdenum wires (supplier : Goodfellow).

In the case of arc melting, different metals or alloys samples were melted before the U-Mo-Ti

alloy in order to trap the residual air (nitrogen and oxygen) in the furnace.

The U-Mo-Ti alloys studied compositions are : U-Mo8-Ti1 and U-Mo9-Ti2 (numbers : wt.%).

UMo/AlTi interaction experiments

UMo/Al-Ti interactions experiments were performed on two types of materials :

- UMo7/AlTi5 (7 and 5 : wt.% of Mo and Ti, respectively) miniplates, with atomized

UMo particles (supplier : AREVA-CERCA),

- UMo8/AlTi5 diffusion couples, prepared from an homogenized UMo8 ingot (supplier :

CEA) and from an AlTi5 mini-compact obtained by powder metallurgy (supplier :

AREVA-CERCA). These couples were prepared in the same way as the UMo/AlSi

ones (see previous section).

These different types of samples were annealed at 400 or 450°C for 2 hours, under Ar + 5 %

H 2 atmosphere.

112 of 435

3. Results and discussion

3.1 U-Mo-Al-Si system

After their diffusion annealing, samples are fully observed by means of optical microscopy

(OM) and scanning electron microscopy (SEM), in BSE mode (FEG-SEM-Philips XL 30

equipped with EDAX EDS detector). Main observations are :

- Morphology : a double layered interaction zone can be observed in each case. In most

cases the sub-layer located on UMo side is cracked (as it can be seen in figure 1).







Figure 1. UMo10/AlSi7 annealed at 550°C-0.5h (left image)and 450°C/3h (right image) [15].

- Thicknesses : vary from 5 to more than 700 µm. As expected on the basis of literature

data, the largest IL have been obtained in couples using Al 1050 (without Si). Si addition

tends to reduce the IL thickness. However, this trend encounters a limitation : indeed,

under 2 wt.%, Si has a negligible influence on interaction rate decrease. Above 5 wt.%,

additional Si doesn’t improve the IL reduction anymore as it can be seen on curves drawn

in figure 2.


IL thickness versus Si content






E [µm]










UMo7 AlSi 7




0 2 4 6 8 10 12

Si content [wt.%]

Figure 2. Influence of Si addition on IL thickness

Right images : UMo7/AlSi 2 (348µm) and UMo7/AlSi7 (159µm), annealed 1h at 550°C

(corresponding points can be seen on the blue curve)

- Composition : the IL sub-layer close to the UMo side of the couple is harder (1066 HV)

and richer in Si than the sub-layer located close to the Al side (936 HV). These observations

and the crack noted in each sample allow us to assume that a silicide phase could be

present on this side of the IL. EDS analyses show that the first sub-layer (on UMo side)

contains around 50 at.% of Si and 30 at.% U. These measurements slightly vary from one

sample to another, but the atomic ratio U/(Si+Mo+Al) remains between 0.43 and 0.53, which

could correspond to a USi 2 type phase, with Si accepting a few substitutions with Al and Mo

atoms. Close to the Al side, the ratio U/(Si+Mo+Al) indicates that this second sub-layer could

correspond to an UAl 3 type phase, with Al accepting a few substitutions with Si and Mo

113 of 435

atoms. XRD analyses on samples polished in edgewise direction (to obtain a larger area for

diffraction) are under progress. Preliminary results confirm the crystallographic nature of both

sub-layers. Further work is needed in order to determine atomic substitutions in each phase.

Close to the IL, the Al alloy is Si depleted. All these results are globally consistent with those

presented by Mirandou et al. [7-8].

3.2 U-Mo-Ti-Al system

U-Mo-Ti alloys elaboration

A more or less significant precipitation of titanium in the UMo matrix was evidenced in the U-

Mo-Ti alloys elaborated by arc furnace. When the getters used before the alloy melting (Ti

and/or Zr) were not efficient enough for trapping oxygen and nitrogen in the melting chamber,

a part of the titanium added in the alloy precipitated as titanium nitride (oxygen being

combined with uranium). When using a more efficient getter (such as an U-Zr alloy), very few

nitrides were found but a ternary Mo-Ti-U reach phase appeared (see figure 3). Due to the

presence of this ternary phase, which is characterized by a composition of the order of 40

at.%Mo, 40 at.%Ti and 20 at.%U, titanium concentration in solid solution does not exceed a

few wt.% in the U-Mo9-Ti1 alloy and about 1 wt.% in the U-Mo8-Ti2 alloy.

The induction melting, under secondary vacuum, leads to a more homogeneous "as cast"

state of the alloy, with titanium in solid solution in γ-UMo. This state seems to be more

favourable, assuming that titanium could play a beneficial role in U-Mo-Ti/Al interactions

when it is not precipitated [9, 16].

Further work is planned for optimizing the elaboration conditions of the alloys and studying

their thermal stability.

Mo-Ti-U ternary


TiN precipitate

Figure 3. U-Mo8-Ti2 alloy elaborated by arc melting, with an U-Zr getter.

UMo/AlTi interaction experiments

Thermal treatments were performed on UMo7/AlTi5 miniplates and on UMo8/AlTi5 diffusion

couples, in order to promote interactions.

In the case of the miniplates, the UMo/Al interaction layer was not affected, in terms of

thickness and morphology, by the direct vicinity of titanium-rich precipitates and a significant

porosity developed at the Ti/Al interface, due to the formation of the Al 3 Ti intermetallic

compound : see figure 4a. This porosity is due to a Kirkendall effect, Al diffusing faster than

Ti [17].

In the case of the diffusion couples, which were heat treated at an higher temperature, an

irregular interaction layer developed at UMo8/AlTi5 interface (figure 4b). Its thickness was

significantly lower (by a factor of about 10) than that of an UMo8/Al reference couple. A

careful examination of the AlTi5/interaction layer interface evidences an about 10 to 20 µm

thick continuous void along this interface (figure 4b). Voids are also present in the AlTi5 alloy,

as a consequence of a massive Ti/Al interaction with Al 3 Ti formation. Even if we cannot

114 of 435

exclude that these voids were enlarged when cooling the samples and preparing polished

sections, their presence tends to indicate that Al is massively consumed by both Ti and UMo

interactions mechanisms, which imply a double Kirkendall effect, leading to a physical

discontinuity at the fuel/Al interface which is probably at the origin of a decrease of the

interaction rate.


Interaction layer


20 µm

Ti rich







4. Conclusion

Interaction layer

Figure 4 : (a) UMo7/AlTi5 miniplate heat treated at 400° C for 2 hours,

(b) UMo8/AlTi5 diffusion couple, heat treated at 450°C for 2 hours.

In this study, we have shown that IL formed in UMo/AlSi diffusion couples are two-layered,

thinner and with elementary compositions different from those obtained in UMo/Al cases

(without Si). These results allow to conclude that, for at least out-of-pile experiments, the

addition of Si into the Al matrix is beneficial for the interaction rate decrease and that this

effect is linked to a modification of the interaction products nature.

U-Mo-Ti alloys were elaborated by arc melting and by induction melting. In the first case,

titanium tends to precipitate either as titanium nitride (when residual nitrogen is present in the

furnace chamber) or as a ternary Mo-Ti-U phase. In the second case, it seems to be nearly

homogeneously distributed in solid solution. Further work is needed for optimizing

elaboration conditions of such alloys and studying their thermal stability.

The study of UMo/AlTi interactions, on miniplates and on diffusion couples, shows that Al

massively interacts both with titanium-rich precipitates, in the Al-Ti alloy, and with UMo. This

interaction can lead to a lack of aluminium which physically slows downs the UMo/Al reaction

by creating voids. The behaviour of such a system, under irradiation, is to be checked.

5. Acknowledgments

We are pleased to acknowledge the AREVA-CERCA company, and especially Messrs.

Jarousse and Grasse for supplying some of the materials used in this study. We also want to

thank Messrs. Tougait, Pasturel and Noël, from the University of Rennes (France), for

induction melting experiments, for their help in determining the phases encountered in the

UMo/AlSi interaction layers and for many fruitful discussions. Finally, Mr. Miragaya and Mrs.

Silvestre and Rouquette are warmly acknowledged for their help in performing samples

preparations, heat treatments and XRD characterizations.

115 of 435

6. References

1 F. Huet et al., RRFM 2005, 10-13 April 2005, Budapest, Hungary.

2 Y.S. Kim et coll., JNM 245 (1997) 179-184

3 H. Palancher et al., J. Appl. Cryst. (2007), 40.

4 F. Mazaudier et al., RRFM 2006, Sofia, Bulgaria, April 30-May 3, 2006.

5 Y.S. Kim et al., RERTR 2005, Boston, USA, Nov. 6-10, 2005.

6 L.S. DeLuca, H.T. Sumsion, KAPL 1747, May 1957.

7 M. Mirandou et al., JNM 323 (2003) 29-35.

8 M. Mirandou et al., RERTR 2007, Prague, Czech Republic, Sept. 23-27, 2007.

9 J.M. Park et al, JNM, article in press.

10 G.L. Hofman et al., RERTR 2006, Oct. 29 – Nov. 2, 2006, Cape Town, Republic of South


11 Y.S. Kim et al., RERTR 2006, Oct. 29 – Nov. 2, 2006, Cape Town, Republic of South


12 M. Ripert et al., RRFM 2006, Sofia, Bulgaria, April 30-May 3, 2006.

13 Cornen et al. and Rodier et al, RRFM-2007, Prague, Czech Republic, Sept. 23-27, 2007.

14 P.E. Repas et al., Transactions of the ASM, Volume 57, 1964.

15 Cornen et al., Proceedings Symposium T, MRS Fall Meeting 2007, Boston, USA, 26-29


16 J.M. Park et al., RERTR 2007, Prague, Czech Republic, Sept. 23-27, 2007.

17 K. Nonaka et al., Materials Transactions 42 (2001) 1731-1740.

116 of 435




Nuclear Fuels and Materials Division, Idaho National Laboratory

P. O. Box 1625, Idaho Falls 83415 – U. S. A.


Results on the relative bond strength of the fuel-clad interface in monolithic fuel

plates have been presented at previous RRFM conferences. An understanding

of mechanical properties of the fuel, cladding, and fuel / cladding interface has

been identified as an important area of investigation and quantification for

qualification of monolithic fuel forms. Significant progress has been made in the

area of mechanical analysis of the monolithic fuel plates, including mechanical

property determination of fuel foils, cladding processed by both hot isostatic

pressing and friction bonding, and the fuel-clad composite. In addition,

mechanical analysis of fabrication induced residual stress has been initiated,

along with a study to address how such stress can be relieved prior to

irradiation. Results of destructive examinations and mechanical tests are

presented along with analysis and supporting conclusions. A brief discussion of

alternative non-destructive evaluation techniques to quantify not only bond

quality, but also bond integrity and strength, will also be provided. These are all

necessary steps to link out-of-pile observations as a function of fabrication with

in-pile behaviours.

1. Introduction

The overall goal of the Reduced Enrichment for Research and Test Reactors (RERTR)

program has been to develop fuels for nuclear research and test reactors that allow effective

conversion from highly enriched uranium (HEU) to low enriched uranium thereby reducing the

threat of nuclear proliferation worldwide [1]. Mechanical properties of the fuel have a

secondary impact on fuel behavior in terms of irradiation behavior. However, mechanical

properties of the fuel are extremely important for overall plate properties. Limited data exists

on the property-processing-structure relationship of metallic uranium monolithic fuel foils.

Most of the available literature involving properties, specifically for U-Mo alloys, were produced

in the 1950s and 60s, although processing methods and microstructural characteristics of

alloys in these investigations were significantly different than those of interest for the RERTR

program [2-4].

Characteristics of the monolithic fuel, both in terms of microstructure and properties, are

extremely important to a successful fuel plate irradiation. Two methods are currently being

aggressively investigated to encapsulate the monolithic fuel foils in 6061-T6 aluminum alloy

cladding: hot isostatic pressing (HIP) and friction bonding (FB) [5]. Both of these methods

can impose a significant amount of stress on the fuel foil, HIP thermally and FB mechanically,

in addition to creating residual stress in the fabricated plate leading to delamination before

irradiation, and significantly altering the mechanical properties of the precipitate hardened

aluminium alloy used as cladding. Therefore, the monolithic fuel must have optimum

characteristics to handle the thermally and mechanically induced stresses during plate

fabrication and a sufficient understanding of stress behaviour on the plate composite must be

gained, so that detrimental defects are not introduced prior to irradiation.

117 of 435

An example of the impact processing has on the monolithic foils is provided in Fig. 1. The

ultrasonic photographs in the figure show a foil that has clearly been affected by the process

(left) and one that has not been affected (right). Both foils were fabricated employing the

friction bonding process, using the same process parameters and fabricated in the same

assembly, i.e. one assembly contained two mini-foils. Clearly, there are differences in the

material properties. There appears to be a clean fracture surface at the bottom right corner of

the photograph on the left, suggesting that a concentration of impurities, most likely carbides,

are present in this area. These “stringers” are unable to accommodate the large processing

loads of friction bonding, and fracture occurs. In addition, along the upper edge of the foil on

the left small, high aspect ratio pieces of fuel have been removed and re-distributed away from

the fuel zone. It is believed that casting and quenching small lots of material results in a finer

grains and less homogeneous microstructure than that obtained from casting, and ultimately

slower cooling, of larger lots of material, i.e. that more characteristic of a large scale

fabrication campaign. Furthermore, warm rolling the finer grained, less homogeneous

microstructure will result in high aspect ratio grains, i.e. increased length to reduced width,

which results in exceptional mechanical properties in the longitudinal direction and reduced

mechanical properties in the transverse direction. Once again, the fuel foil in the photograph

on the left was unable to accommodate the lateral loads associated with the friction bonding

process, while such defects are rarely ever observed in the longitudinal direction.

Thus, the current update will provide results of studies that are underway and future plans to

investigate the mechanical properties of the fuel alloys and cladding material, processingparameter

relationships, composite behaviour and residual stresses induced by friction


Fig 1: Ultrasonic scans of fuel plates fabricated by FSW with a flawed HEU-10Mo foil (left)

and uniform HEU-10Mo foil (right).

2. Experimental Methods and Materials

1.1 Foil Preparation

Monolithic foil alloys of depleted uranium and ten weight percent (nominal) molybdenum were

investigated. A small scale arc melting and casting method was employed to homogenize

and fabricate the DU-10Mo coupons. Background on this method along with details relating

to the preparation of monolithic foils from the coupons, can be found in Ref. 6. Annealing

treatments were performed after rolling with varying temperature and time. Once foils were

prepared, dog-bone tensile specimens were prepared employing a hardened carbon steel

punch and die set. Scanning electron microscopy (SEM) was used to evaluate the fracture

surface of failed specimens.

1.2 Cladding Preparation

Effects of friction bonding applied load on the mechanical properties of aluminium alloy 6061

cladding were investigated. Two pieces of commercial 6061-T6 aluminium alloy, each 0.914

mm thick, were used for each experiment. The alloy had a typical elongated grain structure in

118 of 435

the rolling direction with an approximate area per grain of approximately 614 µm 2 . Each

aluminium alloy workpiece, both for the top and bottom sheets, measured 77.2 cm long by

7.94 cm wide. A single pass was made across the two sheets of aluminium to bond them

together, on one side only. Dog-bone tensile test specimens were prepared similarly to the

method discussed in Section 2.1. Thickness of each specimen varied along the length of the

test piece, but was nominally 1.56 ± 0.01 mm. Specimens were produced along the length of

the bond, parallel with the bond direction (stir-zone), so that a total of 6-8 tensile specimens

were obtained. Note that specimens represent properties under the tool pin in the current

experimental configuration.

1.3 Tensile Tests

Specimens were subjected to tensile loading employing an Instron 3366 universal testing

machine. All tensile tests were conducted at room temperature with a strain rate of 0.5

mm•min -1 . Engineering stress (σ) – engineering strain (ε) diagrams were employed to obtain

mechanical property information.

3. Results and Discussion

Results for the tensile tests performed on the DU-10Mo monolithic fuel foils are provided in

Table 1. Foils were subjected to two different annealing temperatures and three different

annealing times. Results in Table 1show that the annealing time has significant effect on

yield strength, elastic modulus and ultimate tensile strength. There is only a minor

dependence upon annealing temperature. Foils were found to fail in three different modes, a

ductile mode, a transgranular mode, and a mixed mode, examples of which are shown in

Figure 2. The failure mode is not dependent upon the annealing condition employed, but is

rather more dependent on impurity concentration, i.e. carbon, nitrogen and oxygen. Samples

that failed in an intergranular mode had relatively low concentrations of impurities (50 µg•g -1 C,

250 µg•g -1 C, >9 µg•g -1 N and >100 µg•g -1 O). Samples that

failed in a mixed mode manner had impurity concentrations bracketed by the previously listed

numbers, with the mostly ductile mixed mode concentrations being closer to that observed for

the purely ductile failure mode. The dependence upon impurity concentration rather than

annealing parameters is surprising and somewhat unexpected, especially based on the trends

observed. It is important to point out that these observations are based on single foils, and

reproducibility along with supporting experiments, have yet to be performed.


Temperature ( o C)

/ Time (min)


strength, sy



Modulus, E


Ultimate Tensile

Strength, UTS


Failure mode

650 / 30 741 ± 21 60 ± 3 745 ± 19 Mixed mode

650 / 60 783 ± 23 65 ± 2 783 ± 21 Ductile dimple

650 / 120 814 ± 27 70 ± 3 828 ± 21 Intergranular

675 / 60 810 ± 77 69 ± 6 815 ± 76 Ductile dimple

675 / 120 829 ± 47 71 ± 6 831 ± 47

Mixed mode;

mostly ductile

Tab 1: Mechanical properties of DU-10Mo foils as a function of annealing temperature and


119 of 435

Fig 2: Fracture surfaces of tensile test specimens showing a ductile dimple failure mode (top

left), an intergranular failure mode (top right) and a mixed mode (bottom)

Results of the tensile tests are summarized in Table 2 for 0.2% offset yield strength (σy),

modulus of elasticity (E), ultimate tensile strength (UTS) and percent of elongation (e f ).

Observation of the 0.2% offset yield strength shows that yield strength slightly increased as a

function of applied load for single bond passes made on one side of two aluminium alloy

sheets. However, yield strengths obtained for all four loads investigated are well below the

base material value (271 MPa). The decrease in the 0.2% offset yield strength compared to

the base material is attributed to both the loss of the strengthening precipitates that are

dissolved into the aluminum matrix during the temperature increase caused by the process,

and to the reduction of pre-existing dislocations in the parent material [7].

Observation of the modulus of elasticity results reveals that all values obtained are lower than

those obtained for the base material (81 GPa). This observation is attributed to the relative

thinness of the base material compared to the thickness of the samples tested, i.e. ~two

times thicker than the base material.

Ultimate tensile strength results show similar trends to those observed for the 0.2% offset

yield strength. Mainly, the UTS increased with increased applied load, but the experimental

values are significantly lower than the theoretical values or those obtained for the base

material (327 MPa). The UTS is observed to decrease 35% for an applied load of 62.3 kN and

38% for an applied load of 35.6 kN. This loss in tensile strength would be expected to

increase for multiple bond passes made over the assembly and bond passes made on both

sides of the assembly, as is the case for fabrication of the fuel plates.

One of the largest effects of the friction bonding application is on the percent of elongation of

the test specimens. The percent of elongation is significantly higher than the theoretical value

(~114%), while the increase in percent of elongation is moderately higher than that obtained

for the base material (~40%). The percent of elongation appears to be independent of the

applied load of the bond pass. Many FSW tensile test specimens reported in literature

contain microstructures from the different processing zones, i.e. nugget, HAZ and TMAZ. In

the current investigations, the specimens were taken from the processed area under the tool

pin, so that the microstructure is relatively homogeneous. Therefore, the tensile test

specimens contained only fully recrystallized grains, resulting in the significant increase in

material ductility. Minimal differences were observed between the samples in the average

area per grain under the pin, suggesting that there should be minimal differences in the

120 of 435

percent of elongation, as is the trend observed. Similar observations in the stir zone have

been made in other studies with mini tensile specimens [8,9].


Load (kN)

Yield strength,

sy (MPa)

Elastic Modulus,

E (GPa)

Ultimate Tensile

Strength, UTS (MPa)


e f (%)

35.6 167 ± 4 66 ± 6 255 ± 4 25 ± 2

44.5 170 ± 4 66 ± 5 264 ± 5 26 ± 1

53.4 171 ± 4 72 ± 9 273 ± 4 24 ± 3

62.3 177 ± 5 65 ± 4 275 ± 4 24 ± 4

Tab 2: Mechanical properties of friction bonded AA6061 cladding as a function of process


4. Future Plans for Mechanical Analysis

Future plans for mechanical analysis include residual stress analysis of both friction bonded

and hot-isostatic pressed fuel plates. This will be accomplished by using a combination of a

modified Sachs boring-out method, a deflection method and a Treuting-Read method. In

addition, composite tensile test specimens will be tested to evaluate overal structural

properties of the fuel plates. Combination of these tests, along with results presented, will

offer an acceptable baseline for beginning of life properties to be evaluated against irradiated


5. Conclusions

Mechanical properties of monolithic fuel and aluminium cladding processed by friction bonding

have been presented. Properties of the fuel appear to be more sensitive to impurity

concentration rather than annealing conditions. Properties of the aluminium cladding are

sensitive to the applied load used during the friction bonding process. Future plans for

mechanical analysis were discussed.

6. References

1 J. L. Snelgrove et al., “Devolpment of very-high-density low-enriched-uranium fuels,”

Nuc. Eng. Des. 178 (1997) pp. 119-126.

2 A. M. Nomine et al., "Grandeur, mecaniques associées à la corrosion sous contrainte

de I'alliage U-10Mo, " paper presented at the Coloque sur la rupture des materiaux, Grenoble,

9-21 January 1972.

3 M. B. Waldron et al., "Mechanical Properties of Uranium-Molybdenum Alloys," Atomic

Energy Research Establishment, Harwell, England, Report No. AERE-M/R-2554, 1958.

4 B. R. Butcher et al., "The Mechanical Properties of Quenched Uranium-Molybdenum

Alloys. Part I: Tensile Tests on Polycrystalline Specimens," J. Nucl. Mater., 11(1964), 149-


5 C. R. Clark et al., “Update on Monolithic Fuel Fabrication Methods,” Proceedings of

the RERTR Conference, Cape Town, South Africa (2006).

6 C. R. Clark et al., “Update on Monolithic Fuel Fabrication Development,” Proceedings

of the RERTR Conference, Boston, U. S. A. (2005).

7 M. W. Mahoney et al., “Properties of friction-stir-welded 7075 T651 aluminum,” Metall.

Mater. Trans. A 29 (1998) pp. 1955-1964.

8 A. von Strombeck et al. “Fracture toughness behaviour of FSW joints in aluminium

alloys,” in: Proceedings of the First International Symposium on FSW, Thousand Oaks, CA


121 of 435

9 R. S. Mishra et al., in: Proceedings of the International Conference on Joining of

Advanced and Specialty Materials III, ASM International (2000) pp. 157.

122 of 435




Comisión Nacional de Energía Atómica

Centro Atómico Constituyentes

Av. Gral. Paz 1499 (B1650KNA)

Buenos Aires. Argentina.


Ductile gamma uranium molybdenum alloys –?UMo– have excellent behaviour

under irradiation and are being qualified for their use as dispersed and

monolithic low enriched uranium –LEU– nuclear fuels. Nevertheless, excess

porosity growth has been detected in the interface between the aluminium and

the interaction zone when, at high neutron fluxes, amorphous phases are

present in the latter that cannot retain fission gas products. Hot colamination of

monolitihic plates is not possible because of the very different strenght of

aluminium and UMo.

Particularly, in monolithic fuels, this swelling issue and mismatch in

termomechanical properties can be simply avoided by using Zircaloy or

stainless steel alloys instead of the usual aluminium cladding. The growth

kinetics of the interaction zone with these materials is much slower. Additional

advantages are achieved in design capabilities by the possibility of reducing the

cladding thickness and simplicity is maintained in the fabrication process by hot

colamination above the decomposition temperature of the metaestable ?UMo.

The development and post-irradiation results of monolithic LEU plates of ?U-7Mo

(7% w/w Mo) with Zircaloy-4 cladding are described in this work performed in

collaboration and in the frame of international qualification efforts. New

alternatives of monolithic meat fabrication by powder metallurgy and stainless

steel cladding are presented. Plates with asymmetric meat thicknesses can be

easily obtained.

1. Introduction

Uranium alloys with a molybdenum content between 6 and 10 weight percent have excellent

performance under irradiation [1]. The alloy is used in the gamma body centered cubic phase

–?UMo– and can reach uranium densities as high as 16.5 gU/cm 3 for a nominal U-7Mo


It is desirable to use this ?UMo alloy as a monolithic kernel to convert high enriched uranium

–HEU– nuclear fuels to LEU without loosing reactor performance. It has been shown that this

task, at least, needs the development of new technologies since the big mismatch between

the thermomechanical properties of ?UMo and the aluminium alloys used as cladding

materials does not allow the usual picture and frame technique followed by a hot colamination


123 of 435

Different technologies have been tested or proposed to obtain meat-cladding welded plates

with aluminium coverage [2, 3, 4]. All of them, friction stir welding (FSW), transition liquid

phase bonding (TLPB), high isostatic pressure (HIP), swaging or hot colamination, need to

start with a near final size ?UMo meat.

Other complications can also appear since it is not discarded, that in the interface between

meat and cladding in monolithic plates, porosity can grow ought to the non retention of fission

gas products in amorphous phases as shown it happens in UMo dispersed fuels in contact

with aluminium.

A series of alternatives have been analyzed in which the aluminium cladding was changed by

zirconium and stainless steel alloys. Plates using Zircaloy-4 cladding have been developed,

fabricated and irradiated. Also monolithic ?UMo plates are being developed with AISI 304L

cladding material. In both cases the picture and frame method was used followed by hot


2. Zircaloy-4 cladding

An alternative method to avoid, or by-pass problems appearing in the fabrication and

performance of UMo fuels that are in contact with aluminium, begun its development in 2003

at CNEA studding the possibility of using Zr-4 as cladding material [5,6]. For miniplates

fabrication [7], the alloy was melted in an induction furnace and casted in a graphite mold to

obtain plates of 2 mm thickness. If initial smaller thickness was needed, the UMo plate was

hot rolled in air to the adequate thickness with intermediate etching passes. Coupons of 18 x

12 mm were machined with the required thickness. Lids and frames were machined with high

pressure water jet and, after assembling the sandwiches, they were TIG welded. Hot rolling

was performed in eight steps with heating temperatures in the stable gamma phase in an air

atmosphere furnace. The whole colamination process for each plate was optimized for

minimal time residence in the furnace and minimal possible temperature to reduce Zr-4

oxidation and UMo decomposition. Special precautions were developed for cleaning the UMo

and Zr-4 bonding surfaces; mechanical match between coupon and frame was optimized and

precautions had to be taken so as to take care of the big difference in thermal expansion

coefficient of both materials. Miniplates can be deformed without problem so as to obtain

curved plates. The monolithic miniplates were surface finished to a final thickness of 1 mm by

wet abrasion with SiC paper in a semiautomatic machine. Plates were quality checked by X-

ray radiography, ultrasonic scanning and other conventional methods. In figure 1 it is shown a

U-Mo monolithic miniplate with Zircaloy-4 cladding after cutting to final dimensions of 100 x 25

x 1 mm 3 in a guillotine machine. In the thickest meats, cladding thicknesses were as small

as 0.25 mm.

Two of these monolithic miniplates of ?UMo with Zircaloy-4 cladding [8] were irradiated by the

end of 2005 and beginning of 2006 [9] and performed post irradiation experiments in the frame

of international efforts of ?UMo fuels qualification. Irradiation was done in the Advanced Testing

Reactor (INL, USA) during the two cycle RERTR 7A experiment. The PIE took place at the

Hot Fuel Examination Facility (HFEF) (figure 2) of the Material and Fuel Complex (MFC) of

Idaho National Laboratory (INL), Idaho, USA on October 2006 (figures 3 and 4). The total burn

up reached 38 and 33 % respectively. The dielectric layer thickness of MZ25 was of 2.6

microns with a mean swelling of 3.6 %. No problems were detected in metallographic analysis

(figure 5). Other data of these miniplates can be seen in table 1 [10].

3. Stainless steel cladding

124 of 435

The possibility of reducing the cladding thickness allows the introduction of a new variable in a

conversion redesign strategy that can manage in a bigger extent the neutron moderation ratio

[11]. Also this variable can be managed by introducing a moderator, such as zirconium

hydride (ZrH 2 ) in the core of the plate. In this case it is necessary to use a stainless steel

cladding to avoid the possibility of hydriding a zirconium cladding.

Figure 1. Monolithic miniplate of ?U7Mo with

Zircaloy-4 cladding. (100 x 25 x 1 mm 3 ).

Figure 2. From left to right: Ross Finlay

(ANSTO), Enrique Pasqualini (CNEA), Julie

Jacobs (INL) and Curtis Clark (INL).

Figure 3. Miniplates inside the hot cell. MZ25

is the one at the right side of the photo.

Figure 4. Transverse cutting of irradiated

MZ25 miniplate in the hot cell.

125 of 435

Figure 5. Metallography of monolithic MZ25 after irradiation: ?UMo meat and Zr-4 cladding.

Table 1. Characteristics of the two monolithic miniplates of ?UMo/Zr-4 irradiated in RERTR

7A experiment. The weight composition of the core compound is 92,9 % U and 7,0 % Mo,

with a calculated density of 17,53 g/cm 3 . Uranium enrichment is 19,86 % 235 U.

Miniplate MZ25 MZ50

Thickness [mm] 0.99 1.01

Cladding thickness [mm] 0.36 0.25

Meat thickness [mm] 0.26 0.51

Meat width [mm] 18.8 18.6

Meat longitude [mm] 73.0 71.0

Total uranium [g] 5.9 10.9

Meat uranium density [gU/cm 3 ] 16.5 16.2

Surface uranium density [gU/cm 2 ] 0.41 0.21

Burn-up 235 U [%] 38 33

Fission density [f/cm 3 ] 2.7 x 10 21 2.3 x 10 21

Heat flux [W/cm 2 ] 135 217

Swelling [%] 3.6 -

Dielectric layer thickness [µ] 2.6 -

Prototypes of monolithic UMo plates with AISI 304L as cladding material were elaborated. The

UMo coupon was obtained by powder metallurgy methods by cold pressing HMD 12 powder;

standard picture and frame process was used, followed by hot colamination in a nitrogen


The use of a UMo monolithic meat elaborated by powder metallurgy allows the incorporation

of powdered moderators, such as high temperature stable hydrides, and also nanosized

porous powders to adsorb fission gases at grain boundaries so as to reduce overall swelling of

fuel plates.

126 of 435

4. Conclusions

The fabrication of monolithic ?UMo plates with non aluminium cladding (i.e.: Zircaloy-4, AISI

304L) and powder metallurgy methods using the traditional picture and frame technique is a

flexible and practical production scale technology for fuels with densities greater than 7

gU/cm 3 . Special geometrical shapes of meat and plates can be elaborated with the possibility

of incorporating fission gas adsorption materials and moderators in the nucleus of the plates.

This technology is the most promising one for the conversion of high flux reactors from HEU to

LEU. Minor modifications of usual equipment of plate production plants are needed for their

elaboration at industrial scale. Probable economic benefits can come out from thorough

evaluation of the whole fuel cycle, including storage and back end options. Higher surface

heat flows can be used accounting for higher flexibility in the materials used and the

possibility of higher reductions in cladding and plate thicknesses.

Several fabrication procedures are being thoroughly tested such as to improve control

on oxidation during heating, surface finishing of plates, elaboration of full size plates, fuel

assembly designs, cold pressing of UMo powders, etc.

5. Acknowledgements

This work is the result of the effort and commitment of technicians and professionals of

CNEA, INL and the RERTR program. I greatly appreciated the collaboration of my colleagues

at the International Fuel Development Working Group with a very special emphasis in Ross

Finlay, Silvia Balart, Jim Snelgrove, Curtis Clark, Mitch Meyer and Gerard Hofman.

6. References

[1] S. Van den Bergue, W. Van Renterghem and A. Leenaers. Transmission Electron

Microscopy investigation of irradiated U-7 wt.% Mo Dispersion Fuel. The RERTR-2007

International Meeting on Reduced Ewnrichment for Research and Test Reactors. September

23-27, 2007. Prague, Czech Republic.

[2] C.R. Clark, G.C. Knighton, M.K. Meyer, G.L. Hofman. Monolithic Fuel Plate Development

at Argonne National Laboratory. 25th International Meeting on Reduced Enrichment for

Research and Test Reactors (RERTR). Chicago, IL, USA. 5-10 Oct. 2003.

[3] B. W. Pace and G. R. Gale. LEU Fuel Development Progress and Programs, BWXT

Technologies, Inc. 25th International Meeting on Reduced Enrichment for Research and Test

Reactors (RERTR). Chicago, IL, USA. 5-10 Oct. 2003.

[4] C. R. Clark, J. M. Wight, G. C. Knighton, G. A. Moore and J. E. Jue. Update on Monolithic

Fuel Fabrication Development. 27th International Meeting on Reduced Enrichment for

Research and Test Reactors (RERTR). Boston, MA, USA. 6-10 de Nov., 2005.

[5] E. E. Pasqualini and M. López. Increasing the Performance of U-Mo Fuels. International

Meeting on Reduced Enrichment for Research and Test Reactors (RERTR-2004). Vienna,

Austria. 7-12 Nov. 2004.

[6] E. E. Pasqualini. Desarrollo de combustibles de U-Mo. XXXI Reunión Anual, AATN. Bs.

As. 23 al 25 de noviembre, 2004. AATN, Bs. As.

[7] E. Pasqualini. Dispersed (Coated Particles) and Monolithic (Zircalloy-4 Cladding) UMo

Miniplates. The 27th International Meeting on Reduced Enrichment for Research and Test

Reactors (RERTR). Boston, USA. Nov. 6-10, 2005.

[8] E. E. Pasqualini. Elaboración de miniplacas con U-Mo para irradiar en un reactor

de alto flujo. Núcleo disperso (partículas recubiertas) y monolítico (plaqueado con zircaloy-4).

XXXII Reunión Anual, AATN. Bs. As. 21 al 25 de noviembre, 2005. AATN, Bs. As.

127 of 435

[9] E. E. Pasqualini, J. Fabro and N. Boero. Dispersed and Monolythic Plate Type U-Mo

Nuclear Fuels. 10th International Topical Meeting on Research Reactor Fuel Management

(RRFM) . Sofia, Bulgaria. 30 April-3 May, 2006.

[10] E. E. Pasqualini. Irradiacion de miniplacas en el reactor ATR. (Advanced Testing

Reactor, Idaho, EEUU). XXXIV Reunión Anual. Asociación Argentina de Tecnología Nuclear.

19 al 23 de noviembre de 2007. Buenos Aires, Argentina.

[11] E. E. Pasqualini. Advanced Development in U-Mo Dispersed and Monolithic Fuels. The

28th International Meeting on Reduced Enrichment for Research and Test Reactors (RERTR).

Cape Town, Republic of South Africa. Oct. 29- Nov. 2, 2006.

[12] E. E. Pasqualini, J. Helzel Garcia, M. López, E. Cabanillas And P. Adelfang. Powder

Production of U-Mo Alloy, HMD (Hydriding-Milling-Dehydriding) Process. Proceedings RRFM,

March 17-20, 2002. Ghent, Belgium.

128 of 435




Argonne National Laboratory

9700 S. Cass Avenue

Argonne, IL 60439


SSCR RIAR, 433510 Dimitrovgrad, Ulyanovsk Region, Russia


PIE analyses were performed to characterize fission gas bubble development in

LEU U-Mo alloy fuel irradiated in the ATR using an analytical model for the

nucleation and growth of intra and intergranular fission-gas bubbles. Burnup was

limited to less than ~40 at%U-235 in order to capture the fuel swelling stage prior

to recrystallization. The model couples the calculation of the time evolution of the

average intergranular bubble radius and number density to the calculation of the

intergranular bubble-size distribution based on differential growth rate and

sputtering coalescence processes. Recent results on TEM analysis of

intragranular bubble distribution in U-Mo were used to set the irradiation induced

diffusivity and re-solution rate in the bubble swelling model. Using these values,

good agreement was obtained for intergranular bubble distribution compared

against measured data using a grain-boundary enhancement factor of 10 4 . This

value of enhancement factor is consistent with values obtained in the literature.

1. Introduction

Characteristic post irradiation morphology of LEU U-Mo fuel cross sections are

shown in Fig. 1 at several burnup levels [1]. Fission gas bubbles first appear on linear

features, decorated heterogeneously over the fuel cross section (shown in (a)). The

linear features are likely grain boundaries. There are virtually no visible bubbles in the

interior of the grains. As burnup increases (~40-50 %U-235), the bubble population

increases on the grain boundaries and additional bubbles progressively spread to the

interior regions (shown in (b)). At this stage, the fuel swelling rate increases. The

phenomenon underlying this increase in bubble nucleation and growth is grain

refinement or ‘recrystallization’ of the γ U-Mo. Eventually at higher burnup the entire

fuel cross section is uniformly decorated with bubbles (shown in (c)).

(a) 35 %U-235 BU (b) 65 %U-235 BU (c) 80 %U-235 BU

V6018G from RERTR-5 V6001M from RERTR-4 V6022M from RERTR-4

Fig. 1 SEM photos of irradiated U-Mo fuels from RERTR-4 and 5. The samples

shown in this figure were fabricated with the same batch of atomized fuel particles

and irradiated at similar temperatures [1].

*Work supported by US Department of Energy, Office of Global Threat Reduction, National Nuclear Security Administration (NNSA),

under Contract DE-AC-02-06CH11357. The submitted manuscript has been authored by a contractor of the U. S. Governmentunder contract

NO.DE-AC-02-06CH11357. Accordingly, the U. S. government retains a nonexclusive royalty-free license to publish or reproduce the

published form of this contribution, or allow others to do so, for U.S. Government purposes.


129 of 435

2. Characterization of grain and grain boundaries

The fuel particles used in the mini-plate tests were fabricated with the atomization

process. A “cellular” solidification structure is often found in rapidly cooled alloys that

have a pronounced solidus-liquidus gap. An additional feature of the rapid

solidification is a pronounced “coring” within the grains. As a result, the center of the

grains has a higher Mo content than the region surrounding the boundary. As shown

in Fig. 2, the size and shape of the grains vary in the particle; frequently columnar in

shape in the periphery whereas equiaxed and smaller in the interior.

Virtually all the grains at the periphery of C and D particles are columnar grains and

A also has a few, as shown in Fig. 2 (a). The columnar grains seem to have the

same size regardless of the particle size. The particle A is larger than B, but B has

larger grains in the interior part than A. This may be due to solidification and

interdiffusion. The grain size measurement from the SEM picture in Fig. 2 (b) is

consistent with the measurement for grains from the as-fabricated plate. Comparison

between the OM photo and SEM photo shows that the lines in the OM photo are

grain boundaries in the SEM photo. The grain size distribution measured from the

OM photo of Fig. 2 (a) for V03 shows that, although there are some large grains

observed, the predominant size is about 4 µm for this as-atomized fuel.





50 µm

(a) OM of V03. (b) SEM of V03.

Fig. 2 OM and SEM of Mini-Plate V03.

In order to obtain information on homogenous γ U-Mo, some powder was annealed in

the γ phase prior to fuel plate fabrication. As a result of γ-annealing, there are only

large grains in Z03 and the cellular or subgrain structure has been eliminated (Fig. 3).

50 µm

(a) Optical microscopy of Z03. (b) SEM of Z03.The scale bar is 10µm

Fig. 3 OM and SEM micrographs of mini-plate Z03: fuel powder was γ-phase

annealed for 100 hours at 800 o C before plate-fabrication.


130 of 435

3. The Model

The model presented here considers analytical solutions to coupled rate equations that

describe the nucleation and growth of inter- and intragranular bubbles under the

simultaneous effect of irradiation-induced gas-atom re-solution. The goal of the

formulation is to avoid a coupled set of nonlinear equations that can only be solved

numerically, using instead a simplified, physically reasonable hypothesis that makes the

analytical solutions viable. The gas-induced swelling rate is then assessed by

calculating the evolution of the bubble population with burn-up and subsequently the

amounts of gas in bubbles and lattice sites. Uncertain physical parameters of the model

are adjusted by fitting the calculated bubble populations at given burn-ups with

measured bubble size and density data.

Within the context of mean field theory, the rate equation describing the time evolution of

the mean density of gas in intragranular bubbles is given by



( t)


( t)]

= 16π fn




( t) cg

( t) + 4πr


( t) Dgc


( t) cb

( t) − bmb

( t)


( t)



The three terms on the right hand side of Eq. (1) represent, respectively, the change in

the density of gas in intragranular bubbles due to bubble nucleation, the gas-atom

diffusion to bubbles of radius r b

and the loss of gas atoms from bubbles due to

irradiation induced re-solution. Due to the strong effect of irradiation-induced gas-atom

re-solution, in the absence of geometric contact, the bubbles stay in the nanometer size

range. The density of bubbles increases rapidly early in the irradiation. Subsequently, at

longer times, the increase in bubble concentration occurs at a much-reduced rate.

Based on the above considerations, the left-hand side of Eq. (1) is set equal to zero.

This approximation will be more reasonable for larger values of t. A solution for cb


terms of m


and c g is then given by



n rg




= . (2)


( t)

For bubbles in the nanometer size range an approximate solution to the Van der Waals

(VdW) gas law is

1/ 3


( t)


( t)

= ⎜ ⎟

⎝ 4π

⎠ (3)

Using Eq. (1) and an argument similar to that used to derive Eq. (2), the steady-state

solution for m is given by


( )

3/ 2

1/ 2


b ⎞ ⎛ 4π

s v



t ⎞

m ( t ) = ⎜ ⎟




⎝ π ⎠ ⎝ b ⎠

According to Speight [2], the fraction of gas f s

that diffuses to the grain boundary of

grains of diameter d can be approximated by








( D t) −


D t

s g


g d g




131 of 435

Imposing gas-atom conservation, i.e., requiring that the sum of the gas in solution, in

intragranular bubbles, and on the grain boundary is equal to the amount of gas

generated, the term c g (t)

is determined as

c ( t)



( 1 + f ) + ( 1 + f )




32 π

+ 64 π

f r D

n g


f r D

n g

/ b


f β t / b

where β is the number of gas atoms produced per fission event.

Following the work of Wood and Kear [3], grain boundary bubble nuclei of radius R



produced until such time that a gas atom is more likely to be captured by an existing

nucleus than to meet another gas atom and form a new nucleus. An approximate result

for the grain-boundary bubble concentration is given by

1/ 2

⎛ 8 ⎞





1/3 2


π ξDgδ


where a is the lattice constant, z is the number of sites explored per gas-atom jump, δ

is the width of the boundary, ξ is a grain-boundary diffusion enhancement factor, and

K is the flux of gas-atoms per unit area of grain boundary. Under the above

considerations, the flux K of atoms at the grain boundary is given by

( f t)





K =

3 dt dt


Bubble coalescence without bubble motion can be understood on the basis of a

difference in the probability for an atom to be knocked out of the volume between a pair

of bubbles and the probability of an atom to be injected into this inter-bubble volume [4].

If the bubbles contained the same atoms as that comprising the inter-bubble volume, the

net flux of atoms out of the inter-bubble volume would be zero. However, since the gas

bubbles contain fission gas and not matrix atoms, the flux of atoms into the inter-bubble

volume is reduced by the bubble volume fraction, i.e., the net flux out of volume is equal

to λ V − λ( V −VB

), where λ is the atom knock-on distance, and V B

is the intergranular

bubble volume fraction. It is assumed that most of the impacted atoms receive enough

energy to travel distances λ on the order of the inter-bubble spacing. Thus, assuming

that the atom displacement rate is proportional to the fission rate, the overall net rate of

change of the concentration of bubbles in a given size range due to the balance

between growth due differential growth rate between bubbles of different size and

shrinkage due to bubble coarsening without bubble motion is given by

( )

dn r d ⎡ dr 6


dr = − ( )

n r

dr − λδ


f πr

n( r) dr = 0 . (9)

dt dr ⎣ dt ⎦ dg

The last equation is the condition for an equilibrium population of bubbles, where the

effective grain-face-bubble volume is assumed to be disk-shaped (lenticular) with


volume = δ πr , and where s



is the thickness of the material undergoing sputtering.

Equation (9) must be solved subject to the relevant boundary condition. In general, this

boundary condition concerns the rate at which bubbles are formed at their nucleation

size r 0

. The rate of bubble nucleation is provided by the Wood-Kear nucleation

1/ 2



132 of 435

mechanism [3] discussed above in the context of mean value calculations where on the

grain boundary the average time τ


for a gas atom to diffuse to an existing bubble is





πξ DgCb

. (10)

From a consideration of the growth rate of freshly nucleated bubbles it follows that

⎛ Cb

n ( r 0 ) dr = dr /( dr / dt) r = r 0


g ⎝ τ

b ⎠ . (11)

In general, the solubility of gas on the grain boundary is substantially higher than in the

bulk material. The gas concentration on the boundary will increase until the solubility

limit is reached (approximately given byτ b

), whereupon the gas will precipitate into

bubbles. Thus, the rate at which a grain boundary bubble adsorbs gas is approximately


( dm / dt) r= r

= bvC

g /( 4τ


/ 3)


, (12)

where C is the intergranular gas-atom concentration. Using the Van der Waals gas law









( kTr + 3γb

r )



( rkT + 2γb

) dt

Combining Eqs. (12) and (13)


( dr / dt)


r =






. (13)

( rkT + 2γb





( b 0 / 3)( kTr + 3γb






Subsequent to intergranular bubble nucleation, gas solubility on the boundary will drop

to a relatively low value and gas arriving at the boundary will be adsorbed by the existing

bubble population. The rate at which a grain boundary bubble adsorbs gas is

approximately given by

dm / dt = 12πrξD

g Cg

/ d g

. (15)

Combining Eq. (13) and (15)





( rkT + 2γbv

) 3b



dr / dt =





( kTr + 3γb


) d g r

. (16)

Using the approximation on the right-hand side of Eq. (16), Eq. (9) becomes



( )

( ) 6



bvξD gCg

dn r


n r −

− λδ f r n( r ) = 0

2 s




dr dg

, (17)

The solution of Eq. (17) subject to the boundary condition expressed by Eq. (11) and

(14) is

2 2 3 3


4 4


π r ( kTr + 3γbv

r ) exp[ −κ

( r − r0



r ) =


( 2 ) 2 vCg



rkT + γbv

, (18)


π f λδs

κ =

2bvξD gCg

. (19)


133 of 435

Comparison to the measured bubble-size distributions are made by integrating Eq. (18)

over the bubble size range, i.e.


+ i∆


N ( ∆i

) = ∫ n( r)dr

. (20)


∆ 0 + ( i−1)

4. Model Validation

Table 1 shows a description of fuel used in the analysis [1]. This data base consists of

both as-atomized and γ-annealed specimens. From table 1, the range of burn up is from

30 – 49 at% U-235, fission rate from 2.3 – 7 x 10 14 f/cm 3 -s, temperature from 66 – 185

º C, and Mo content from 6 – 10 wt.%. Table 2 shows the value of the key physical

parameters used in the model. As shown in Table 3, these values for D g and b were

estimated by comparing the calculated intragranular average bubble size and density to

measured results [5]. The remaining critical parameter ξ was determined by best

overall interpretation of the measured intergranular bubble-size distributions for the γ-

annealed and for the as-atomized specimens, respectively. The calculated results

shown in Table 3 can be brought more in line with the data by decreasing D ,

increasing b, or both. This then would require a commensurate decrease inξ . For this

exercise to be meaningful measured intragranular bubble-size distributions are required.


Table 1 Description of fuel used in the analysis [1]



Fuel Burn up, rate Total

Plate ID



(10 14 ) duration

at% U-235


f/cm 3 -s




( o C)

RERTR-3 580H Z03 U-10Mo(a,γ) 32 5.3 48 121

RERTR-3 580C Y01 U-10Mo(m,γ) 30 4.8 48 109

RERTR-1 - V002 U-10Mo(a) 39 3.8 94 66

RERTR-3 580G V07 U-10Mo(a) 30 5.1 48 122

RERTR-3 580W V03 U-10Mo(a) 38 6.3 48 149

RERTR-3 580Z S03 U-6Mo(a) 39 7.0 48 158

RERTR-5 600AG R6007F U-7Mo(a) 37 2.4 116 185

RERTR-5 600M V6019G U-10Mo(a) 49 2.9 116 142

RERTR-5 600AH V8005B U-10Mo(a) 37 2.4 116 170

a: atomized, γ: annealed at 800 o C for 70-100 hours, m: machined

Table 2 Values of key physical parameters used in the model

D g = 2.5 x 10 -31 cm 2 /s

b = 1 x 10 -18 • f s -1

? = 7 x 10 3 (γ-annealed powder) = 4 x 10 4 (as-atomized powder)

Table 3 Intragranular results

Calculated Data [4]

Bubble diameter (nm). 2.1 ≈ 2

Bubble density (cm -3 ) 1.5 x 10 18 ≈ 3 x 10 18


134 of 435

Grain-boundary bubble size distribution for Z03

Grain-boundary bubble size distribution for Y01









Bubble Density (cm -2 )






Bubble Density (cm -2 )








0.04 0.06 0.08 0.10 0.12 0.14 0.16

Bubble Diameter (µm)

0.04 0.06 0.08 0.10 0.12 0.14 0.16

Bubble Diameter (µm)



Fig. 4 Calculated and measured intergranular bubble-size distribution for γ-annealed plates

Grain-boundary bubble size distribution for V03

Grain-boundary bubble size distribution for V07










Bubble Density (cm -2 )




Bubble Density (cm -2 )









0.05 0.10 0.15 0.20 0.25 0.30 0.35

0.05 0.10 0.15 0.20 0.25 0.30

Bubble Diameter (µm)

Bubble Diameter (µm)



Fig. 5 Calculated and measured intergranular bubble-size distribution for as-atomized plates


135 of 435

Grain-boundary bubble size distribution for V8005B

Grain-boundary bubble size distribution for V6019G









Bubble Density (cm -2 )




Bubble Density (cm -2 )






0.05 0.10 0.15 0.20 0.25 0.30

0.05 0.10 0.15 0.20 0.25 0.30

Bubble Diameter (µm)

Bubble Diameter (µm)



Grain-boundary bubble size distribution for V002

Grain-boundary bubble size distribution for S03











Bubble Density (cm -2 )






Bubble Density (cm -2 )










0.05 0.10 0.15 0.20 0.25 0.30 0.35

0.05 0.10 0.15 0.20 0.25 0.30 0.35

Bubble Diameter (µm)

Bubble Diameter (µm)



Fig 5, continued


136 of 435

Bubble Density (cm -2 )







Grain-boundary bubble size distribution for R6007F

0.05 0.10 0.15 0.20 0.25 0.30

Fig 5, continued

Bubble Diameter (µm)




Figure 4 shows the calculated and

measured intergranular bubble-size

distribution for γ-annealed plates. Figure

4a is atomized whereas Fig. 4b is

machined. Figure 5 shows calculated

and measured intergranular bubble-size

distribution for as-atomized plates.

Figures 5a-5e are for 10 wt% Mo

whereas Figs. 5f and 5g are for 6 and 7

wt% Mo, respectively. As is evident

from Figs. 4 and 5, in general, the model

calculations are in remarkable

agreement with the data. The error bars

are shown on the measured data for V03

(see Fig. 5(a)). Similar uncertainties

can be considered for other plates. The deviation between calculated and measured

results shown in Figs 5f and 5g is most likely due to the lower Mo content and, thus,

requires different values for D and ξ .


5. Conclusions

Calculations of intergranular bubble size distribution made with a new mechanistic

model of grain boundary bubble formation kinetics is consistent with the measured

distributions. The gas-atom diffusion enhancement factor for grain boundaries was

determined to be 7 x 10 3 in order to obtain agreement with the measured distributions.

This value of enhancement factor is consistent with values obtained in the literature [6].

The enhancement factor is about six times higher for as-fabricated powder plates than

for the annealed plates, due to the lower Mo content on the boundaries. Model

predictions are sensitive to various model parameters such gas-atom diffusivity and resolution

rate. Improved prediction capability requires an accurate quantification of these

critical materials properties and measurement data.


1. S. L. Hayes, C.R. Clark, J.R. Stuart, M.K. Meyer, T. C. Wiencek, J. L. Snelgrove

and G. L. Hofman, Proceedings of the 2000 International Meeting on Reduced

Enrichment for Research and Test Reactors, Las Vegas, NV, 1-6 October 2000.)

2. M.V.Speight, Nucl. Sci. Eng. 37 (1969) 180.

3. M.H. Wood, K.L. Kear, J. Nucl. Mater. 118 (1983) 320.

4. R.C. Birtcher, S.E. Donnelly, C. Templier, Phys. Rev. B50 (1994) 764.

5. S. Van den Berghe, W. Van Renterghem, A. Leenaers, Proceedings of the 29 th

International Meeting on RERTR, Prague, Czech Republic (2007).

6. J.C. Fisher, J. Appl. Phys. 22 (1951) 74.


137 of 435




BP 1114, 26104 Romans-sur-Isère – France


This paper documents the developments undertaken at AREVA-CERCA to

manufacture silicide fuel plates of new designs, intended to answer the

needs of new tubular fuel elements. It emphasizes how we have managed

different development programs in order to improve our processing

parameters from a R&D scale until an industrial scale. Three examples are

more precisely developed: boron sheet insertion in a high density silicide fuel

plate, manufacturing of high density and high fuel meat thickness U 3 Si 2

bended fuel plates and manufacturing of high density U 3 Si 2 fuel plates with

over sizes.

1. Introduction

AREVA-CERCA has been involved in producing silicide fuel plates since 1982. Annually more

than 350 fuel assemblies are being delivered worldwide. Along the 50 years of existence of

CERCA, its manufacturing experience has increased significantly in mastering the production

of more complex fuel assembly designs requested by its costumers. The Development of

manufacturing and inspection processes as well as quality improvement were always a part of

our history and vision.

Specific requests of high density silicide fuel plates intended for tubular fuel elements have

dawned recently. The purpose was to develop solutions to answer the particular needs

stemmed from new reactor designs or reactor conversion. This means that we would have to

improve our processing parameters and that specific studies should be performed accordingly.

The recent developments carried out at AREVA-CERCA in the RR field are presented


- Boron sheet insertion in a high density U 3 Si 2 fuel plate,

- High density & thick fuel meat U 3 Si 2 fuel plates bending,

- Manufacturing of high density U 3 Si 2 fuel plates with over sizes.

The step to enriched uranium is developed as well.

2. Boron sheet insertion

2.1 Objectives

Till now, the AREVA-CERCA’s experience was limited to introducing a boron sheet next to an

aluminide fuel core. This technology was put in place at an industrial scale with success. The

transition to the same development with high density silicide fuel plates would induce to

master the difference of behaviour of a fuel core with a low density in uranium to a high density

in uranium (density increased by a factor 3) with regards to the proximity of the boron sheet.


AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

138 of 435

This kind of change put down an interesting challenge in mastering the complex interface

between the fuel and the boron cores taking into account the different mechanical behaviours

which lead to blister formation as well as possible low cladding thickness.

The purpose of the development program undertaken was to demonstrate our ability to shift

the manufacture of low density to high density fuel plates while maintaining the highest level of


2.2 Developments & results

This study was conducted as a project through a methodical approach which was declined in

several phases.

The aim of the first phase was to validate hypothesis made at R&D level and based on our

experiences gained with other fuels (exploratory phase).

The second phase consisted to a down selection of the most promising parameters obtained

from the previous stage (confirmation phase). So as to benefit from better representatives of

the tests performed, the number of samples used was increased early in stage 2.

The latest phase (validation phase) was performed through industrial batch in order to have a

full representation of the manufacturing reality.

More details about each phase are given below:

Exploratory phase:

Its objective was to test different selected combinations of parameters on full scale plates,

using depleted uranium.

Two kinds of parameters were retained:

- The constitutive materials of the boron sheet,

- Identified processing parameters.

All other plate’s characteristics remained identical.

As commonly used for development studies, a statistic Tagushi plan was designed. This first

phase lasted 6 months, at the end of which, a down selection took place.

Confirmation phase:

The solutions retained at the end of the first phase were tested with more consequent

quantities. This enabled us to compare in a more precise way the performances of each

solution retained and to benefit from more results to choose a unique solution: the best

combination for further manufacturing of the plates at a pre-industrial scale.

Validation phase:

Being aware that results on small quantities are no fully representative of a whole industrial

production, we have launched a manufacture of the retained solution at a large scale.

Depleted uranium was used for this purpose. The objective was to demonstrate that on a large

number of plates, a high quality level would be ensured.


AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

139 of 435

Several inspections were performed on the plates with successful results:

- Absence of blister confirmed also by no US indication,

- Minimum cladding thickness similar to other standard fabrications even at the boron / fuel

core interface – see figure 1.

- Excellent uranium surface distribution, even in the dog-bone area – see figure 2.

The other inspection results obtained were similar to those of other standard fabrications.

Following figures illustrate these results:

Fig 1: Metallographic inspection:

dog bone area

Fig 2: Surfacic uranium distribution


To ensure that all developments undertaken will be formalized and enhanced, detailed

synthesis reports are being established. Thanks to these documents, we can attest to the

qualification of the new processing developed. Further step will be the transfer to workshop, by

editing workshop level procedures and by training operators on these specific procedures.

3. High density and thick fuel meat U3Si2 fuel plates bending

3.1 Objectives

All along the 50 years, AREVA-CERCA has developed adapted tools to bend plates made of

silicide or aluminide alloys.

A large experience has been gained in this field and industrial process has been consolidated

through the significant number of fuel assemblies already produced routinely.

The purpose of the development program was to demonstrate that plate made of a thick fuel

meat in HD fuel could be bent while maintaining a perfect integrity of the core.

Technical challenges could be sum up as follows:

- Higher mechanical constraints while bending due to harder mechanical characteristics of

the fuel core,

- Distortions all along the plate.

3.2 Developments & results

The bending process was already mastered for plates with a standard thickness around

1,3 mm with a fuel core thickness around 0,5 mm. The necessary improvements to settle

consisted in developing new bending processing parameters in order to challenge the

production of plates made of an higher fuel meat thickness up to 0,8 mm.


AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

140 of 435

The study conducted has consisted in testing comparatively different combinations of

parameters on plates (first, on surrogate materials then on depleted uranium) and checking

their incidence on the final plate’s state quality.

For each step, tests were performed on a set of radius of curvatures, from the less demanding

to the more drastic. Each time, the verification of the correct shape of the plates, as well as

the integrity of the fuel core were analysed.

Examples of developments and inspections performed are presented on figures 3, 4 and 5.

Fig 3: Picture of a bended plate before

development of new bending processing

parameters – plate thickness = 2 mm

Fig 4: Picture of a bended plate after

development of new bending processing

parameters – plate thickness = 2 mm

Fig 5: Picture of a metallographic inspection – central zone of the plate

These results show that the bending of high density and thick fuel meat U 3 Si 2 fuel plates is

completely operational in AREVA-CERCA.

4. Over sized U3Si2 fuel plates

4.1 Objectives

AREVA-CERCA has manufactured several thousands of plates of so called “standard

dimensions” with an active length around 600 mm and an active width around 60 mm. So as

to increase our know-how, we have undertaken the following developments:

- Increase of the length and of the thickness of fuel plates: standard dimensions plus 60%,

- Diminishing of their width: standard dimensions minus 60%.

This means mastering a longer active length on a smaller width and keeping same regularity

in the cladding thickness while facing more difficult rolling conditions.


AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

141 of 435

4.2 Developments & results

To ensure that quality would be on the same level than our standards, we have tested

consequent quantities of depleted uranium plates with characteristics detailed previously.

Testing plates with non conventional dimensions led us to adapt our producing tools. Thus, we

have been facing the challenge to extend their capacity to extreme dimensions.

The main inspection results obtained are detailed in the table 1.

Blister test


X-Ray inspection: stray particles

& white spot

Uranium distribution inspection

Cladding thickness

See figures 6 and 7


Same level than for standard U 3 Si 2 plate fabrications

No stray particles observed

No white spot detected

Homogeneity less than ± 16 % in the standard area

Ratio mean cladding thickness / minimum cladding

thickness: equivalent to other standard fabrications.

Good regularity all along the plate.

Tab. 1: Main inspection results obtained on over size plates

Fig 6: Metallographic inspection:

cross section of the plate centre area.

Fig 7: Metallographic inspection:

cross section of the dog bone area.

These results show that the manufacturing of high density U 3 Si 2 fuel plates with overclassical

dimensions is also well mastered in AREVA-CERCA.

5. Step to enriched uranium

Another crucial aspect of these kinds of developments is the step to enriched uranium.

Indeed, as nuclear facility, we are anticipating the rules defined by the regulator. Modifying the

dimensions of the plates, their density and also the quantity of uranium 235, changes the

characteristics of the products and has a direct impact on the safety matter.

Such changes require specific studies, which have to be conducted by experts. Moreover, the

procedure may be subject to getting an authorization from the French safety authorities. Thus,

this is another parameter not to be sneezed at in this kind of study since its instruction can

take times and as a result extend the foreseen time schedule, and can lead to significant

adaptation of the working conditions. This is another aspect on which we pay a particular



AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

142 of 435

6. Conclusion

Improvements presented in this paper conducted AREVA-CERCA to enlarge its experience in

high density fuel plates and to master successfully complex fuel plates manufacturing


The result is that we are able to undertake consequent development programs and to deploy

all needed competencies so as to find adequate solutions to a given customer need.

Such capacity is an asset at a period where new research reactors are emerging with specific

technical demands and where others reactors are converting with necessity to adapt new

fuels to existing design. An example of our adaptability is the appliance of these

developments to tubular fuel elements for either HJR or MARIA fuel assemblies.


AREVA-CERCA, a subsidiary of AREVA NP, an AREVA and Siemens company

143 of 435






Laboratory for Mechanical Behaviour of Irradiated Materials


91191 Gif-sur-Yvette Cedex, France


ICSM – Paniscoule, Centre de Marcoule, University of Montpellier (UM2)

BP 17171, 30207 Bagnols-sur-Cèze, France


Laboratory of Aqueous Corrosion Studies


91191 Gif-sur-Yvette Cedex, France


For the Jules Horowitz new material-testing reactor (JHR), an aluminium base

alloy, called AlFeNi, will be used for the cladding of the fuel plates. This alloy was

developed for its good corrosion resistance in water at high temperatures.

However, few studies have been performed concerning its degradation process

and the relationships with irradiation effects. The conception of the JHR fuel

requires a better knowledge of the corrosion mechanisms.

Corrosion tests performed in autoclaves on AlFeNi plates and different techniques

show a duplex structure for the corrosion scale: a dense amorphous layer close to

the metal and a porous crystalline layer in contact with the water. The corrosion

process involves three mechanisms: inner growth of the amorphous scale, its

dissolution and the precipitation of the dissolved aluminium as hydroxide crystals.

The observation of corrosion scales formed under neutron flux shows that

irradiation increases the corrosion kinetics but also modifies the corrosion

morphology and probably the mechanism.

1. Introduction

Within the Jules Horowitz Reactor project, high performances for neutron fluences, for

experimental facilities and for its versatility are forecasted. To improve the reactor capabilities

with a low enriched fuel, as requested by IAEA, the fuel element conception has been

strongly optimized and the temperature of the reactor core will be higher than in older

experimental reactors.

The development of a thin oxide film on the fuel-plate clad can induce significant effects on

the cladding integrity due to the modification of the solid-liquid interface. Due to poor thermal

conductivity of such film and to the reduction of the water gap between fuel plates, fuel

cooling is reduced increasing the risk of fuel overheating (fuel expansion, increase of the

corrosion phenomena). Moreover, the thickness of the cladding decreases due to

consumption of metal associated with the oxidation reaction. Thus, safe use fuel requires a

good understanding of the aging phenomena under irradiation, in particular the corrosion


The first part of the work was performed on unirradiated plates. Static corrosion experiments

have been carried out in autoclaves to characterize the corrosion products and to identify the

associated corrosion mechanisms. The second part took into account the post-irradiation

examination of irradiated fuel plates to integrate the role of irradiation in the corrosion



144 of 435

2. Characterization of the AlFeNi alloy

The specifications for the AlFeNi alloy are summarized in Table 1. Our samples (20x20 mm)

were cut from rolled 1.4 mm thick sheets provided by CERCA (Romans, France) in an

annealed temper representative of the fuel cladding.

The samples were embedded in a conductive,

Bakelite resin with carbon filler and then polished

with SiC paper.The microstructure of our samples

was revealed by optical and scanning electron

microscopy. The chemical analysis profiles were

obtained by ElectronProbe MicroAnalysis (EPMA).

The alloy consists of micrometric isotropic

precipitates dispersed in an Al-Mg matrix. The

composition of these intermetallic precipitates is

very close to Al 9 FeNi as expected by published

references [1] [2] .

Addition element Specification

Fe 0.80 to 1.20

Ni 0.80 to 1.20


1.80 mini

Mg 0.80 to 1.20

Mn 0.20 to 0.60

Cr 0.20 to 0.50

Zr 0.06 to 0.14


0.30 maxi

Table 1: Chemical composition of the

AlFeNi alloy.

X-Ray diffraction confirmed that the matrix lattice parameter is in accordance with the Mg

content, when compared with other aluminium alloys containing magnesium in solid

solution [3] .

3. Characterization and description of the corrosion product

Static corrosion experiments were performed in autoclaves on fresh AlFeNi alloy plates

(20 mm x 20 mm). Two kinds of autoclaves have been used: V=0.5L – stainless steel and

V=5L – titanium. Experiments were done at 70, 165 and 250°C for different leaching times

(6–45 days). Deionized water was used for the experiments. The water pH, measured at

room temperature before and after each test, was in the range 5 to 8. The exposed samples

were examined through SEM, FEG-SEM, EPMA, XRD and µRaman spectroscopy.

To elucidate the sequential growth mechanism of corrosion products, a vapour gold coating

was deposited on the polished metal surface before the corrosion test. After corrosion, the

gold film was located between two different corrosion scales (Figure 1), thus revealing a

double inner and outer growth mechanism.

First examinations on the SEM pictures show a duplex structure (Figure 1):

• Close to the metal, a first amorphous scale contains all the alloying elements. The inner

growth mechanism of this layer does not seem to have any effect on the cathodic Al 9 FeNi

precipitates. Because of its low potential, magnesium is oxidized before aluminium.

According to XRD and Raman spectroscopy, the amorphous phase could result from a

disordered mixture of gibbsite Al(OH) 3 and brucite Mg(OH) 2 .

• The external layer, in contact with

the water, is constituted with pure

aluminium hydroxide crystals. At

165°C and 250°C, boehmite (AlOOH)

crystals were identified, as confirmed

from X-Ray and Raman analyses. No

additional element is detected in the

outer layer. The morphology of

aluminium hydroxide grains is strongly

dependent on the leaching conditions:

temperature, chemical environment,

water flow…

Figure 1: SEM micrograph of the leach sample

(BSE mode).

EPMA profiles (Figure 2) clearly show the differences in composition between both layers.

Moreover, magnesium diffuses from the metal to the amorphous layer. Iron and nickel are

clearly associated inside the intermetallic precipitates.


145 of 435

Figure 2: Quantitative elementary profiles through oxide scale determined from EPMA experiments.

4. Proposition for the mechanisms of AlFeNi corrosion

According to these observations, a description of the corrosion mechanism in three steps is

proposed, corresponding to three interfaces and associated kinetics of reactions:

• The redox reaction between magnesium and aluminium, on the one hand, and oxidative

species on the other hand, takes place at the interface between the metal and the

amorphous layer. This reaction follows its own kinetics and leads to the formation of the

amorphous layer.

• Near the interface amorphous layer-crystalline layer, the amorphous oxide dissolves in

the water, which penetrates through the porous crystalline scale.

• At the inter-layers interface, aluminium released in the solution precipitates to form

aluminium trihydroxide crystals with a third kinetics. From ICP-AES experiments, magnesium

remains in solution, as expected from thermodynamics. The fate of the other additive

elements is more difficult to underline.



Figure 3: FEG-SEM micrograph of the

corrosion product.

There are many indications of the dissolution-precipitation process:

• Firstly, the presence of two coordination kinds of Al atoms as detected by NMR analysis:

the octahedral coordination corresponding to already known aluminium oxide and hydroxide

and the tetrahedral coordination. Unknown in solids, this coordination signs the presence of a

polycation [4] during the condensation of aqueous aluminium species into aluminium hydroxide


• Secondly, as shown on Figure 3, the interface between both oxide layers goes

continuously from a compact material to a degradation area and another one full of very

small crystals to become a well crystallized scale. The largest crystals were thus the result of

many successive dissolution-precipitation processes.


146 of 435

• Thirdly, in some precise conditions at 70°C, isolated crystals on the amorphous scale

have been observed.

The degradation of AlFeNi alloy is hence the consequence of the competition between the

redox and dissolution-precipitation processes. Since the inner layer is always visible, the

formation rate of amorphous oxide is certainly more significant than that of the oxide

dissolution. New experiments will be developed to examine in more details each kinetics

depending on the leaching conditions.

5. Irradiation effects

An irradiation of U 3 Si 2 fuel plates in the Belgian BR2 reactor (SCK?CEN) [5] at Mol was

ordered by the CEA-Saclay to qualify the JHR fuel plates : average heat flux 256 W.cm -2 ;

average burnup 1.3 x 10 21 fissions.cm -3 meat. After three irradiation cycles of 20.5, 22.2 and

26.1 days, no change of microstructure was reported on the AlFeNi cladding.

But the overall oxide thicknesses measured at the hottest points (120-140°C) of the plates,

reached around 50 µm in 69 days; whereas less than 5 µm (averaged thickness) were

obtained after 34 days in autoclave at 165°C. Consequently, irradiation increases

significantly the corrosion rate.

Optical micrograph

SE scanning electron micrograph

Figure 4: Pictures of the outer cladding surface on AlFeNi cladded U 3 Si 2 fuel plate 6 .

The transverse micrographs of irradiated samples (Figure 4) exhibit morphology of the oxide

layers strongly different from that prepared in autoclaves: no crystal grains are visible in the

outer scale. Due to irradiation and/or water leaching flow in the reactor, the duplex structure

is not clearly observable. According to the optical micrograph, Al 9 FeNi precipitates can be

revealed in the layer adjacent to the cladding. However, no interface is visible with a second

layer. Moreover, on the SEM images, some precipitates can be observed in the outer part of

the layer. At this time, we cannot conclude about the nature of the oxide scale.

Besides, differences of oxide thickness have been observed between outer and inner

cladding surface. That could be relied to differences of temperatures or of water flow


6. Discussion from these observations

Most part of this work on the degradation process of AlFeNi alloy has been performed in

static autoclaves. This corrosion procedure is not representative of what happens in the core

reactor on the fuel cladding. Nevertheless, different points must be underlined.

Since the corrosion scale on AlFeNi alloy exhibits a duplex structure with two layers of

different densities, the weight measurement cannot be simply correlated to the oxide layer

thickness. Moreover, if the degradation scale results from a competition between the

formation of aluminium oxide and its dissolution, the concentrations of elements released in

the leachate do not traduce the real amount of corroded material: the weight gain is thus not

relevant to evaluate the corrosion rate and its kinetics cannot be directly obtained.


147 of 435

Secondly, the kinetics of corrosion is very dependent on the exposure conditions:

temperature, water uptake and leaching flow, water composition, pH... All these parameters

can modify the equilibrium of the kinetics competition. In order to illustrate this point, a

corrosion experiment was managed for 34 days at the same temperature (T=250°C) in the

same titanium autoclave with and without water renewal every 7 days.

Corrosion procedure Inner layer thickness outer layer thickness Weight gain

Without water renewal 24.0 ± 3.5 µm 13.1 ± 2.2 µm 329.5 mg.dm -2

With water renewal 11.6 ± 1.0 µm 6.1 ± 1.2 µm 260.5 mg.dm -2

Table 2: Comparison of results of the corrosion procedures.

Figure 5: SEM micrograph of the oxide layer

obtained for a sample leached for 34 days with

leachate renewal (BSE mode).

Figure 6: SEM micrograph of the oxide layer

obtained for a sample leached for 34 days without

any leachate renewal (BSE mode).

Figure 5 and Figure 6 show that inner and outer layers

are two times thicker without water renewing than with

it. Moreover, the weight gain is only 21% higher and

not 100% (Table 2). Even the oxide quality (density) or

the dissolution rates are different.

Another illustration of the aqueous media influence is

given by the presence of silicon in the water. During

the corrosion process in the autoclave, this silicon is

incorporated only in the amorphous layer, not in the

crystalline one (Figure 7).The role of silicon have to be

carefully examined since under irradiation, a

transmutation of aluminium into silicon occurs.

BSE-SE micrograph Si Kα X-Ray map

Figure 7: EPMA pictures of oxide layer

obtained in water contaminated with


To sum up, even if the conclusions about the effects of the irradiation on the corrosion

kinetics are not clear at this time, these effects can not be neglected to be as close as

possible of the reality in reactor.

7. References


H. Coriou, R. Fournier, L. Grall, J. Herenguel, J. Hure and P. Lelong, Al-Fe-Ni Alloys Corrosion

Resistant in Hot Water and Steam ; Proceedings of the second UN international conference on the

peaceful uses of atomic energy, Geneva 1958, P/1271, vol.5, pp. 128-152.

[2] .V. Raynor, V.G. Rivlin, Phase Equilibria in Iron Ternary Alloys, The Institute of Metals, 1988

[3] C. Vargel, Propriétés générales de l’aluminium et de ses alliages, Techniques de l'Ingénieur, M4661

[4] Jean-Pierre Jolivet, De la Solution à l’oxyde, Condensation des cations en solution aqueuse, Chimie

de surface des oxydes, Savoirs actuels, InterEditions / CNRS Editions, 1994

[5] A. Leenaers, S. Van den Berghe, E. Koonen, S. Dubois, M. Ripert, P. Lemoine, Post-irrradiation

examination of AlFeNi cladded U 3 Si 2 fuel plate irradiated under severe conditions, Transactions of 11 th

International Topical Meeting Research Reactor Fuel Management (RRFM) and Meeting of the

International Group on Reactor Research (IGORR), Lyon, France, 11–15 March 2007

[6] S. Van den Berghe, SCK?CEN, personal communication


148 of 435

AREVA-CERCA 10 years licence for fuel fabrication



Les Berauds, B.P. 1114, 26104 Romans Cedex – France


Every ten years, each French Nuclear Installation (referred here after as INB for

“Installation Nucléaire de Base”) shall be subject to a safety evaluation review in order to

obtain the operating licence for the next ten years period. The licence is delivered during a

so called “Factory Permanent Group” review whose participants are a group of experts

from the French Safety Authority (ASN), the French Institute for Radiation protection and

Nuclear Safety (IRSN) and the User of the plant. The safety evaluation is conducted by

both the User and the IRSN during at least a one year period before the Permanent Group

review. During this period, the User shall demonstrate the conformity with regards to

applicable standards of all the safety issues related to the factory operation such as

criticality, radioprotection, seism, fire, external risks, etc…

After more than one year of study, CERCA factory in Romans (France) referred as INB #

63 has succeeded its safety evaluation review in late 2006 and is now licensed to operate

safely till end of 2016.

The aim of this talk is to present the content of this project that has been conducted since

end of 2005 and whose purpose is to ensure the sustainability of CERCA fuel fabrication

factory in Romans (France), at least for the next ten years period.


1. Purpose

Every ten years, each French Nuclear Installation shall be subject to a safety evaluation review in order to obtain

the operating licence for the next ten years period.

As known, AREVA / CERCA is yearly manufacturing many types of Fuel Elements for Research Test Reactors

& Material Test Reactors as well as thousands of molybdenum targets for the nuclear medical market. The

factory is located in Romans (France) and is referred as INB 63 (Installation Nucléaire de Base # 63). The site is

shared with FBFC as LWR plants type fuel factory through INB 98.

To operate, the INB 63 is subject to the authorization of the French Nuclear Safety Authority (ASN).

Picture and map of the CERCA / FBFC site

149 of 435

“ASN is tasked, on behalf of the State, with regulating nuclear safety and radiation protection in order to protect

workers, patients, the public and the environment from the risks involved in nuclear activities. It also contributes

to informing the citizens.”

By end of 2006 and after a long preparatory period, CERCA was licensed by the ASN for ten years.

The purpose of this paper is to present the stakes of such an authorization and to highlight the main issues to

address during the project.

Be authorized

The authorization to run is subject to the prescriptions of the “Arrêté du 10 août 1984” (August 10 th 1984 decree)

related to the quality for the design, the construction and the operation of nuclear installations.

It is the responsibility of the operator to conform to the regulations. In front of the population, the ASN must

guarantee the conformance of the Nuclear Installation (INB) operation to the decree.

CERCA no more authorized to run would deprive many research reactors of fuel and would significantly disrupt

the production of molybdenum for medical exams. Therefore, be authorized is the challenge.

Show the ability to operate safely

So, it is CERCAs everyday responsibility to maintain a high level of safety and security in its facilities. For this,

a complete Safety, Security & Environment (SSE) system is deployed in order to ensure that all the practices

conform to the safety regulations requirements.

Be safe

The Nuclear Safety covers all the actions taken to prevent a nuclear accident or to limit its consequences.

Establishing and developing a strong safety organization is our priority for whole of our activities such as design,

fabrication, storage & shipment of nuclear products.

Particularly, this organization must take into account all the equipment changes.

2. The main steps of the authorization process

General project organization and planning

French State side

The Nuclear Safety Authority is in charge of validating the authorization to run. This authorization may be

delivered on the basis of a technical analysis which is conducted by the Institute for Radiation protection and

Nuclear Safety.

“The IRSN is the expert in research and specialised assessments into nuclear and radiological risk serving public

authorities”. The IRSN is appointed by the Safety Authority.

During the safety evaluation period, the IRSN has constituted a project organization with a project manager and

a team of experts on each discipline.


CERCA has also constituted a project type organization in order to prepare whole of the documentation and to

answer to the questions of the IRSN experts.

The team is leaded by the Safety, Security and Environment Management department, and is also composed of

personals from the operation department of CERCA and from several personals from different engineering

departments of AREVA.

Both teams always wanted to work closely in order to avoid any kind of misunderstanding. This spirit was a key

factor of success.

150 of 435

The overall schedule of the project was as follow:

N° 2003 2004 2005 2006 2007


1 INB 63 Safety evaluation notification by ASN



INB 63 FSAR revision

Safety analysis









FSAR Analysis by the IRSN

Safety evaluation project start by ASN

Technical exchange between CERCA & IRSN experts

Safety document delivery to IRSN experts

Safety evaluation by IRSN experts

Preparation of the ASN Experts Permanent Group

Experts Permanent Group meeting

Permanent Group pursue

Overall schedule of the project

Internal preparation period (Internal studies - FSAR revision)

The first step is to conduct internally a global safety analysis of the current situation in order to update the Final

Safety Analysis Report (FSAR) and the Operating Guidelines. These documents must be an accurate picture of

the factory at the beginning of the project in order to allow both parties to make their own diagnostic.

Doing the studies and updating the FSAR took about 1 ½ year. Obviously, the ideal would be to demonstrate

safe people with safe processes on safe machines in a safe building. But the regulation always changes in a safer

way and is more and more demanding. So, even if our level of safety is continuously upgraded, it remains still a

little gap between what is required and what is in place.

The CERCA FSAR is divided in 3 volumes

• 1 st volume : General description of the site and associated facilities

• 2 nd volume : Detailed description and safety analysis of each workshop and facility

• 3 rd volume : Global safety analysis

This structure allows anyone to easily access to the safety issues, either on the factory or at any work post.

The detailed evaluation review of each workshop and each process has permitted to show the strong points and

the weak points of our way to operate. So it was easy to draw up an improvement program that could be

submitted to the IRSN and implemented gradually.

Previously to the formal project start meeting, the revised FSAR as well as an improvement program proposal

was transmitted to the IRSN.

Project Start

The Safety evaluation review of the CERCA Nuclear installation is driven by the IRSN which scheduled a

formal “project start meeting” that took place on Wednesday December 5 th 2005 in Fontenay-aux-Roses (IRSN

head office).

During this meeting, it was reminded the duties of each party, the way to work together and the main milestones:

• Project organization on both sides (IRSN & CERCA)

• IRSN experts assignments in CERCA

• Discussions

• Safety files delivery by CERCA to IRSN

151 of 435

• Safety evaluation by the IRSN experts

• Factory Permanent Group meeting preparation

Evaluation by IRSN

This period took place between the project start and the safety files delivery to the IRSN by CERCA. It was a

favourable period for technical exchanges between IRSN and AREVA/CERCA.

In ten months we had about 30 technical joint meetings.

As decided before, the relationship between the people was maintained very open in order to avoid any


The following subjects were addressed:

• Criticality

Product sub-criticality follow-up during fabrication:

It is to demonstrate that, in any normal situation, the fabrication conditions allow to maintain

Keff < 0,950 and in any accident situation, Keff < 0,975.

No accident occurrence in case of single failure:

Specific sketches have been elaborated in order to ensure that a double check is systematically

done in case of a single criticality control mode.


Cellule SE3

Boîte Ø70

Valise ronde

de la matière


N° d’étape 3 4 5 6


Modes de

Masse + géométrie

Masse + géométrie Masse + modération


(H/U = 0)

Masse + géométrie Masse + géométrie

Moyens de contrôle

Masse totale P2’ obtenue par addition





« masse »

des masses mesurées P2

Moyens de contrôle

(Boîte Ø 70 : G1 )

Nacelle 30x30x4 : G2

Valise ronde : G3 (Valise ronde : G3)

« géométrie »





Valise ronde

6 charges de fusion

13 000g 235 U 350 g U

UAl : 630 g d’U

x 3500 g 235 U

U 3 Si x : 1230 g d’ U

Pesée e de la matière

Transfert unitaire des bocaux dans une

1 750 g 235 UAl


x : 630 g d’U

U 3 Si x : 1230 g d’U


valise ronde.

Limites procédé :

Limite : 6 bocaux / valise


Volume fixé (4,4 l) limitant la modération

UAl x : 105g d’ U / bocal

13 000g 235 U

5 000g U

U 3 Si x : 205 g d’U / bocal



Boîtes Ø 70

Déversement de la

Calcul de la masse totale

Umétal concass é-

13 000 g 235 U 5 000 g U

de masse

Étiquette d'identification du lot



Enregistrement de la masse

d’Al ré ellement ajoutée

Contrôle du

n°lingot renseigné sur la

respect des limites

Boîtes Ø 70

boîte + masse totale du

de masse

Bilan des charges préparées :

Umétal concassé-


Enregistrement de

•n° article de fusion

5 000 g U –

la composition des

•n° lot de fusion

13 000 g 235 U (masse d’ U du

charges (n° article,

lot de fusion)

n° lot de fusion,

•quantit é d’ 235 U par charge

composition r éelle des


•quantit é totale d’ 235 U du lot de fusion


(valise ronde)

n° lingot renseigné sur la boîte

Calcul de la masse d’Al

à ajouter


Système de suivi de masse

Enregistre dans le système de suivi de

•Quantité d’ 235 U dans les charges

masse :

• Les quantit és d ’ 235U dans les charges

•Quantité totale d’ 235 U dans le lot de fusion

• La quantité totale d’ 235 U du lot de fusion

• La masse totale d’U

Double chargement de la

Déversement de trop de Chute d ’un bocal

P1 Surchargement d’une valise : bocaux surchargés




matière : corrigé




Surchargement d’une valise : 8 bocaux au lieu de 6 :


Rapprochement de 2

immédiatement G2 Double chargement d’ un bocal : rendu impossible par l’exploitant, seules 6 alvéoles

P1 + G1


P1 + G1

détecté immédiatement


disponibles dans la valise



Erreur sur diamètre de la bo îte en


Matière non nivelée P1

MA2: détectée à l ’arrivée en





Contrôle de la


Défaillance balance en

de la

balance en

Rapprochement de 2 valises

P2’ + G3

Renversement d’une boîte

pesée ouverture de

P2’ + G3






P1 Chute de la nacelle P1


Chute de valise, bris de bocaux : corrigé immédiatement




matiè re dans une

nacelle sûre

par la géométrie

de matière fissile dans la

valise ronde (P2’)

Fiche suiveuse "fusion" :

Poids total de la valise ronde chargée


Contrôle du respect des limites

Example of specific sketch established to verify the presence of double check in case of single criticality control

mode – case of a part of the uranium alloy elaboration process

152 of 435

• Human factor (Tokaï-Mura accident experience feedback)

Consequences of high constraints on the safety during fabrication:

The purpose of this study is to identify the risk of overstepping the red line by the operator in case of

constraints in his work.

An investigation program has been launched in order, first, to determine the sensitivity of CERCA to

the human factor, second, to evaluate whether or not, specific measures should be taken. The

methodology is based on an interview of the operators.

Work post experience feedback evaluation

Establishing the safety / criticality basic requirements & rules applicable to the work post

Operator interviews



Action plan (if any)

Current conclusions are that CERCA is quite sensitive to the human factor (indeed, there is one operator

on each machine) but that the safety instructions are well understood and well observed.

• Radioprotection (internal exposure)

In CERCA, the internal exposure of the operators is very low. Every handling of material is done under

glove boxes or with the protection of a mask. Nevertheless, a few improvements are on going on some

work posts organisation.

• Radiological cleanness / Material dissemination

An evaluation was made on the safety of containments breaks during normal operation such as opening

of a glove box airlock. A few minor improvements may be implemented.

• Seism

The main seism issues were addressed during the previous evaluation review of the installation. A few

equipments like storage compartments, tables, etc. remain to be fixed in order to fit with the current


• Fire

A complete fire risks evaluation has been conducted and ends up in a calorific load clearance which is

on-going. Finally, the purpose of this study is to demonstrate that the local occurrence of a fire could

not spread everywhere so as to set fire to a large part of the workshop.

• Equipment ageing

Each automated machine was analysed in order to identify if a loss or a defect of the control system

could have consequences on the safety of the installation. The conclusions were that the safety is not

sensitive to our automatisms.

• External risks (rain, snow, wind, storm, …)

Series of risk evaluation have been requested by the IRSN to be conducted in the next 2 years. Those

evaluations are on-going now.

• Aggression risks (gas explosion, truck explosion, plane crash, …)

Same as above.

A gas delivery cabinet will be moved away from the CERCA building in order to remove any accident

due to a gas pipe breakdown.

153 of 435

• Hydrogeology

A survey plan has been initiated in order to improve our capability to detect a potential contamination of

the ground.

• Wastes management

This issue is managed at the site level. A global project is in charge of evacuating the wastes to the

specialized sites of the ANDRA in conformance with the applicable rules.

ANDRA is the National Radioactive Waste Management Agency. “ANDRA operates independently

from the waste producers. …. It is responsible for the long term management of the waste produced in


A selective sorting leads to direct the wastes, either directly to the storage sites, or to the compacting

facility of AREVA.

All those subjects were discussed with, and evaluated by the IRSN. Some of them where addressed during the

preparation period of the Factory Permanent Group of Experts meeting. Some others require more time and so, a

commitment from the INB 63.

The IRSN requested CERCA to produce nearly 20 safety analysis technical documents that were transmitted in

due time. The IRSN was satisfied with the quality of those documents.

Preparation of the Factory Permanent Group

It is the custom to organize a joint meeting between the IRSN and the operator in order to find acceptable

solutions for the items that have not been agreed during the safety evaluation period.

This meeting is very important as it states on most of the issues.

The meeting took place on October 17 th 2006. Its base of work was the IRSN report of INB 63 safety evaluation.

During the meeting, we confirmed the commitment of AREVA/CERCA to precise and improve the safety

system of reference of the installation where necessary. Also, we agreed together on several pending issues.

Factory Permanent Group meeting

The Factory Permanent Group of Experts meeting took place on November 29 th 2006 and was preceded one

week earlier by a visit of the installation by all the members (40 persons).

The purpose of this meeting is clearly to state on the “authorization to operate” renewal.

The expert members must be convinced by both the IRSN and CERCA that the installation and its organization

are in condition to allow a safe operation. Also, it is to ensure that the tool will be improved and maintained

during the next ten years.

During this meeting, the IRSN presented the conclusions of the INB 63 safety evaluation as well as the

commitment of the operator as discussed during the preparatory meeting. There were some discussions between

the members of the Permanent Group, the IRSN and AREVA/CERCA about pending issues. CERCA proposed

an improvement plan with regard to the recommendations of the Permanent Group. This improvement plan is in

progress now ad is very carefully followed by the ASN.


« A l’issue de l’examen des documents que vous avez transmis à l’ASN et ses appuis techniques, …, je n’émets

aucune objection à la poursuite de l’exploitation mentionnée en objet. »

The authorization to operate is delivered to CERCA.

Factory Permanent Group pursue

The project does not end. It is continuing!

Our authorization to proceed is bound with our wish to make progress.

154 of 435

For this, the CERCA project team has been maintained in order to perform all the improvements required by the

conclusions of the FPG. Whole of the actions, recommendations and commitments have been assessed and

scheduled with milestones to return to the ASN.

The top management of AREVA / CERCA is very committed.

Studies and works are on-going on line with the schedule. The ASN is in charge of checking the progress of the

project through regular inspections on the basis of the IRSN ratification of the CERCA files and works.

3. Conclusion

Getting the ASN authorization to proceed was a major issue for CERCA.

CERCA is authorized to operate till end of 2016. We were able to fit with the very high requirements level of the

ASN, provided some improvements and investments.

The key factors of success of this project were mutual comprehension, confidence, full transparency and

commitment between both parties.

The continuity of CERCA production is a reality in France but, why not anywhere else?

155 of 435

Session III

Reactor operation, fuel safety and core


156 of 435


Authorities, Activities and Plans for the Minimization of

High Enriched Uranium Through the Global Threat Reduction Initiative

Parrish Staples, John Creasy

Office of Global Threat Reduction,

National Nuclear Security Administration; Washington, DC 20585


The Office of Global Threat Reduction’s (GTRI) Conversion Program develops and

implements the technology necessary to enable the conversion of civilian facilities using

high enriched uranium (HEU) to low enriched uranium (LEU) fuels and targets. The

Conversion program mission supports the minimization and, to the extent possible,

elimination of the use of HEU in civil nuclear applications by working to convert

research reactors and radioisotope production processes to the use of LEU fuel and

targets throughout the world. During the Program’s 30 years of existence, 55 research

reactors have been converted from HEU to LEU fuels, and processes have been

developed for producing the medical isotope Molybdenum-99 with LEU targets. Under

GTRI, the Conversion Program has accelerated the schedules and plans for the

conversion of additional research reactors operating with HEU. This paper summarizes

the current status and plans for conversion of research reactors, in the U.S. and abroad,

the supporting fuel development activities, and the development of processes for medical

isotope production with LEU targets.


Nuclear research and test reactors have been in operation for over 60 years and have

served a variety of uses from pure nuclear science, to nuclear technology development, to

roles as research tools in non-nuclear scientific fields including medicine, agriculture, and

industry. To date, there are over 270 research reactors currently operating in more than

50 countries worldwide. The expanded use of research reactors began in 1954 under The

Atoms for Peace initiative. Initially, the majority of these research reactors were fueled

with low-enriched uranium (LEU), however as technology developed reactors began

requiring higher specific power and neutron flux, and to avoid costs associated with the

development of higher density LEU fuels, these reactors began using high-enriched

uranium (HEU) material. This change allowed existing fuel designs to be used.

As worries increased over the potential use of HEU in the manufacture of nuclear

weapons, concern grew about the potential of HEU-fueled research reactors becoming a

source of the material. In response, the U.S Department of Energy (DOE) initiated a

conversion program in 1978 to develop the technology necessary to reduce the use of

HEU fuel in research reactors by converting them to LEU fuel. Argonne National


157 of 435

Laboratory (ANL) and Idaho National Laboratory (INL) are the technical lead

laboratories for the program.

Beyond the research activities for research reactors described above, a significant purpose

of research reactors is the production of medical isotopes, Molybdenum-99 ( 99 Mo) in

particular. Although 99 Mo can be produced by neutron activation, it is more widely

produced by fission of 235 U, through the irradiation of HEU targets. In fact, a significant

fraction of the HEU that the U.S. exports every year is for the fabrication of targets for

the production of 99 Mo. In the mid-1980s the Conversion Program was expanded to

include, in addition to the conversion of research and test reactors, the development of

technology for the production of 99 Mo with LEU material.

Another expansion of the Conversion Program occurred in the early 1990s, when the

Program, which initially focused on reactors supplied with U.S.-origin HEU, began to

collaborate with Russian institutes with the objective of converting reactors supplied with

Soviet- or Russian-origin HEU to the use of LEU fuel. Since 1995, a fuel development

program specifically intended to support the conversion of Russian-supplied reactors,

including irradiation and qualification of fuels in Russian test reactors, has been


The ultimate objective of the Office of Global Threat Reduction (GTRI) is not only the

conversion of HEU-based reactors and 99 Mo production processes to use LEU, but to

remove the HEU material from the facilities and provide for its secure disposition. The

Conversion Program therefore coordinates its activities with programs which focus on the

secure disposition of HEU material, programs like GTRI’s Removal program, which

coordinates the repatriation of U.S.-origin and Russian-origin fresh and spent research

reactor fuel.


The Conversion Program has identified 207 research and test reactors worldwide that are

or were fueled with HEU fuel. The program has compiled a list of 129 of these research

reactors with the objective of converting them to LEU fuel. The current list contains

U.S.-supplied, Russian-supplied, and Chinese-supplied facilities. The selection of

facilities for inclusion in the list is based on the potential for converting the reactor to

LEU fuel (availability of LEU fuel, either already qualified or under development) and

the existence of a secure disposition path for the removed HEU fuel. The remaining 78

HEU-fueled reactors have been excluded from the Conversion Program scope for a

variety of reasons, including (1) classification as defense related facilities, (2) location in

countries that currently do not fully collaborate with the United States on reactor

conversion programs, or (3) requirements for very specialized LEU fuel which would be

too costly and time consuming to develop.

Since the inception of the Conversion Program, 55 of the 129 reactors have been

converted to LEU fuel or have shutdown prior to conversion. Under GTRI, DOE has


158 of 435

established targets for the conversion of 129 HEU-fueled research reactors. The current

goal is to convert the remaining 74

reactors in the list of candidates by the

year 2018. Of the 74 remaining research

reactors within the scope of the

Conversion Program, 46 can be

converted with existing LEU fuels, while

the remaining 28 require the

development of advanced high density

fuels to allow their conversion. A new

high-density UMo fuel is under

development that will allow the





Converted or verified as shutdown

Planned for conversion with existing fuels

Planned for conversion with new fuels

Beyond GTRI scope

conversion of 19 reactors, the remaining 9 reactors may be able to use the UMo fuel as

well, but further analysis is needed. The program is focusing much effort on the

development of these advanced high-density fuels, particularly UMo fuels, with the goal

of qualifying these advanced fuels by 2010.

The Conversion Program also coordinates with other agencies, including the State

Department, the Nuclear Regulatory Commission (NRC), and the International Atomic

Energy Agency (IAEA). The IAEA has supported the objectives of the Conversion

Program through departments concerned with nuclear security and technical cooperation.

The role of the NRC is important, as regulator for U.S. university reactors and as the

agency that approves the export of HEU material.

Current U.S. law authorizes HEU exports for reactors that have agreed to convert to LEU

fuel once a suitable fuel is qualified for their facility. This policy has been instrumental

in encouraging the conversion of research reactors with high utilization that require

significant annual amounts of fresh HEU fuel. Many reactors, however, have a very slow

rate of burn-up and require no new fuel in the immediate future. To encourage the

conversion of these reactors, the Conversion program has developed an incentive

program that allows the procurement of LEU fuel that would provide a service life

equivalent to that of the HEU fuel in the reactor. The number of conversions per year has

accelerated significantly since GTRI took over management of the Conversion program.

Since the announcement of GTRI the Program accelerated the conversion rate, with a

total of sixteen in the last three years.


From its beginning in 1978, the Reduced Enrichment for Research and Test Reactors

program, now the GTRI Conversion Program, has expanded its scope and strengthened

its mandate. Today the Program enjoys various levels of support from within the

Department of Energy up to the President, including several international agreements. In

1986, the Nuclear Regulatory Commission (NRC) issued a rule on “Limiting the Use of

Highly Enriched Uranium in Domestically Licensed Research and Test Reactors. This set

the mandate that research reactors must convert to use LEU if it is available and qualified


159 of 435

for use in the reactor. It also states that U.S. Government funds would be used to

implement the conversion. In 2004, Secretary Abraham committed the U.S. to converting

its domestic research reactors to use LEU in a speech to the IAEA, and created the Office

of Global Threat Reduction within the NNSA. RERTR became the Reactor Conversion

program and a pillar of this office. In 2007, in the third meeting of the Global Initiative to

Combat Nuclear Terrorism, the U.S. issues a joint statement with Russia. The Statement

calls for, among other things, “minimizing the use of highly enriched uranium…in

civilian facilities and activities”. Along with these political authorizations, the United

States Congress continually authorizes the expansion and increased funding of the

Reactor Conversions Program, which now includes 129 domestic and international



In the next few years the Conversion Program is expected to accelerate further, as many

reactor conversions will continue to occur. The technical efforts to establish agreements

with the reactor operators, and the development and procurement of LEU fuel will

increase rapidly to meet the challenges. Meeting this goal will also require increased

policy efforts to engage the governments and facilities that have not yet joined the

conversion effort as well as technical efforts to develop a conversion approach for

reactors that are technically more challenging.


160 of 435




Reactor Português de Investigação, Instituto Tecnológico e Nuclear

Estrada Nacional 10, 2686-953 Sacavém – Portugal


RERTR Program*, Argonne National Laboratory

9700 South Cass Avenue, Argonne, IL 60439 – USA


The 1 MW Portuguese Research Reactor (RPI) switched from high-enriched

uranium (HEU) to low-enriched uranium (LEU) in September 2007. The core

conversion was done under IAEA’s Technical Cooperation project POR4016,

with financial support from the US and Portugal. The safety analyses for the core

conversion were made with the assistance of the RERTR program. This paper

presents the measurements done during the start-up program and compares

them with an as-built MCNP model. The performance of the new LEU core is

compared to that of previous HEU cores.

1. Introduction

The Portuguese Research Reactor (RPI) is a 1 MW, pool-type reactor, built by AMF Atomics

and commissioned in 1961. The activities currently underway in the RPI cover a broad range

from irradiation of electronic circuits to calibration of detectors for dark matter search, as well

as by more classical subjects such as neutron activation analysis. Most of these activities

use in-pool irradiations.

The RPI was commissioned in 1961 with LEU fuel. However, it was later converted to HEU

fuel for economic reasons. In 1999 Portugal declared its interest to participate in the Foreign

Research Reactor Spent Nuclear Fuel Acceptance Program (FRRSNF). A commitment was

made to stop using HEU after May 12, 2006 and return all HEU fuel until May 12, 2009. The

core conversion to LEU was done within IAEA’s Technical Cooperation project POR4016 with

financial support of the US and Portuguese governments. An extension on the use of HEU

until May 31, 2007 was granted by the Department of Energy, in order to minimize the

downtime of the reactor. The actual conversion was done in September 2007. Table 1

summarizes the main milestones of the project.

A feasibility study was performed during 2005 with the assistance of the RERTR program at

Argonne National Laboratory. Uranium silicide (U 3 Si 2 -Al) dispersion fuel with a density of 4.8

g/cm 3 was selected because of its widespread use in research reactors and for the relatively

large number of manufacturers. The feasibility study also had the goal of minimizing the

number of assemblies required for operation during the current FRRSNF acceptance window.

The new LEU standard assembly has 235 U loading of 376 g vs. 265 g for an HEU standard

* Work supported by the U.S. Department of Energy, National Nuclear Security Administration, under

Contract No. DE-AC02-06CH11357.

161 of 435

assembly. With this design the core size remained unchanged, at 12 assemblies, and only

14 assemblies are required for operation until May 2016 [1]. The number of plates (18 for

standard and 10 for control assemblies) was kept the same as for the HEU fuel.

Milestone Planned Effective

Commitments for funding Mid 2005 As planned

Feasibility study End of 2005 As planned

Safety studies Mid 2006 End of 2006

Project and Supply Agreement Mid 2006 Early 2007

Fuel manufactured End of 2006 As planned

Regulatory Approval End of 2006 August 2007

Conversion Early 2007 September 2007

Tab. 1: Milestones for the conversion project

The results of neutronic studies, steady-state thermal-hydraulic analyses and accident

analyses demonstrated that the RPI could be operated safely with the new LEU fuel [2]. The

submission of the safety documentation for approval suffered a 6 month delay from planned.

The IAEA initiated the review of the documents shortly after their reception. Revised

documents were submitted in June 2007 addressing the issues raised during review. The

IAEA provided a letter of support for the conversion in late June and the licensing body of the

RPI approved the conversion in August 2007.

The most challenging aspect of this project was the conclusion of the required tripartite

agreement between the IAEA and the US and Portuguese Governments, which involved

several interactions with the two governments, the IAEA and the European Commission.

2. Conversion

Fig. 1 shows the initial LEU core configuration. LS1 through LS7 are standard assemblies and

LC1 through LC5 are control assemblies, NS is a Sb-Be neutron source, FC a fission

chamber and the DA are hollow dummy assemblies. The hollow dummy assemblies were

introduced in the LEU core in order to improve the thermal hydraulic safety margins [2].

Fig. 1. Initial LEU core configuration, adapted from MCNP model of core.

162 of 435

The shim-safety rods B1 to B4 are mounted in assemblies LC1 to LC4; the regulating rod,

BR, in LC5. The regulating rod was calibrated using the positive period method. The shimsafety

rods were calibrated in pairs B1/B2 and B3/B4 by comparison with a known

displacement of the regulating rod. At the end of these calibrations, the safety parameters of

Table 2 were determined, where B1 through B4 represent the shim-safety rod worth. The

quoted uncertainties of 3% derive directly from the uncertainty in the calibration of the

regulating rod and its propagation to the other parameters through the calibration process.





in OLC


1 Core Excess Reactivity E < 4.80 4.11 ± 0.12

2 Total Shutdown Subcriticality E – (B1+B2+B3+B4+BR) < -3.00 -9.09 ± 0.27

3 Min. Shutdown Subcriticality E – (B1+B2+B3) < -1.00 -4.73 ± 0.14

4 Regulating Rod Worth BR < 0.60 0.33 ± 0.01

Tab. 2: Compliance with Safety Parameters

All safety parameters obtained from the rod calibrations satisfy the requirements of the OLC.

3. Neutron fluxes

Thermal, epithermal and fast neutron fluxes were measured in 13 grid positions, including the

4 hollow dummy assemblies in positions 62, 63, 13 and 54, as shown in Fig. 2.

Fig. 2. Plot of core grid showing highlighted in bold and italic the

positions where neutron fluxes were measured.

The RPI does not have a regular fuel cycle, with a standard core configuration. Configurations

with up to 15 HEU assemblies were previously used; configurations up to 13 LEU assemblies

are now foreseen. For the purposes of flux comparisons, the best match with the current LEU

core is the first HEU core [3], implemented in February 1990; it is not a perfect match, since

the HEU core had one Be reflector in position 13 and the fission chamber in position 54.

Table 3 compares the measured thermal fluxes at core mid-height. Measurements were done

at 1 MW and 100 kW. The average ratio between the thermal fluxes measured in the HEU and

LEU cores is 0.9 ± 0.3, covering two orders of magnitude of the values. We are conservatively

163 of 435

assuming an uncertainty of 10% and 20% for the measured LEU and HEU flux values,

respectively. From the available data there is no clear loss or gain of thermal neutron flux with

the conversion to LEU. Furthermore, the LEU core has 2 additional irradiation positions, inside

the hollow dummy assemblies in positions 13 and 54, which have thermal neutron fluxes of

1.9x10 13 and 1.8x10 13 n/cm 2 /s, respectively.



LEU thermal

flux (n/cm 2 /s)

– 10%

HEU thermal

flux (n/cm 2 /s)

– 20%



(– 22%)

55 7.7E12 5.4E12 0.7

56 1.7E12 1.2E12 0.7

46 2.8E12 2.6E12 0.9

36 3.9E12 3.2E12 0.8

26 2.8E12 3.0E12 1.1

57 2.8E11 2.4E11 0.9

37 5.0E11 4.5E11 0.9

38 5.0E10 5.6E10 1.1

Tab. 3: Comparison between thermal neutron fluxes for HEU and LEU comparable cores.

Gamma dose rates were also measured in all free grid positions, at mid-height of the core,

using a Radiotechnique Compelec CRGA11 ionization chamber. The measurements were

done at a power of 100 kW and extrapolated to 1 MW using the 16 N linear channel. The ratio

of HEU to LEU values is 1.1 ± 0.2 covering one order of magnitude of the values.

4. Updated MCNP model

The MCNP core model used in the feasibility and safety studies [1,2] was updated using the

extensive data provided by the fuel manufacturer CERCA. Measured values for the uranium

isotopes, impurities in fuel meat and cladding were introduced, as well as measured values for

the plate and clad thickness.






Reactivity (%∆k/k)






0 20 40 60 80 100

Rod position (%)

Fig. 3. Integral rod worth curve of shim-safety rod 1: measured vs.

MCNP calculated values. The lines were drawn to guide the eye.

164 of 435

Since there is considerable shadowing between the shim-safety rods in this compact core,

the integral worth of the rods was calculated by simulating the actual rod positions that were

used in the measurement. The same procedure was applied before for the HEU cores with

excellent results [1]. Only preliminary results are shown here. A comparison of calculated and

measured values in determining the worth of shim-safety rod B1 is plotted in Fig. 3. The

integral worth was measured to be 2.6 ± 0.1 %?k/k and calculated to be 3.0% ?k/k.


MCNP thermal flux (n/cm 2 /s)





1E10 1E11 1E12 1E13 1E14

Measured thermal flux (n/cm 2 /s)

Fig. 4. Thermal neutron fluxes: measured vs. MCNP values. The top

line is a least-squares linear fit; the bottom line shows a 1:1 ratio.

Figure 4 shows preliminary results of the calculated thermal neutron fluxes vs. measured

values. Calculated values are along a straight line with a small offset to the 1:1 relationship

over nearly 3 orders of magnitude.


The RPI switched from HEU to LEU in September 2007 within IAEA project POR4016, with

financial support from the US and Portugal. For in-pool irradiations, the new LEU core has the

same performance as a comparable HEU core. The core change also allowed the introduction

of two high-flux positions which did not exist before, increasing the pool irradiation

capabilities. Work in progress includes the measurement of neutron fluxes and gamma dose

rates in the beam tubes and improvements in the as-built MCNP model of the core.


[1] J.G. Marques, N.P. Barradas, A.R. Ramos, J.G. Stevens, E.E. Feldman, J.A. Stillman,

J.E. Matos, “Core Conversion of the Portuguese Research Reactor: First Results”, Proc.

2005 International Meeting on Reduced Enrichment for Research and Test Reactors,

Boston, Massachusetts, November 6-10, 2005.

[2] J.E. Matos, J.G. Stevens, E.E. Feldman, J.A. Stillman, F.E. Dunn, K. Kalimullah, J.G.

Marques, N.P. Barradas, A.R. Ramos and A. Kling, “Core Conversion Analyses for the

Portuguese Research Reactor”, Proc. 2006 International Meeting on Reduced Enrichment

for Research and Test Reactors, Cape Town, South Africa, October 29-November 2.

[3] E. Martinho, I.C. Gonçalves, A.S. Oliveira, M.C. Lopes, C.R. Carlos, H. Silva, “Campo de

Radiações do Novo Núcleo do Reactor Português de Investigação”, Report LNETI/DEEN-

R-91/21 (1991) in Portuguese.

165 of 435


University Reactor

Conversion Lessons

Learned Workshop for

Texas A&M University

Nuclear Science Center

Eric C. Woolstenhulme

Dana M. Meyer

April 2007

The INL is a U.S. Department of Energy National Laboratory

operated by Battelle Energy Alliance

166 of 435


University Reactor Conversion Lessons Learned

Workshop for Texas A&M University Nuclear Science


Eric C. Woolstenhulme

Dana M. Meyer

April 2007

Idaho National Laboratory

Idaho Falls, Idaho 83415

Prepared for the

U.S. Department of Energy

Office of Nuclear Nonproliferation and Security Affairs

Under DOE Idaho Operations Office

Contract DE-AC07-05ID14517

167 of 435


The Department of Energy’s Idaho National Laboratory, under its

programmatic responsibility for managing the University Research Reactor

Conversions, has completed the conversion of the reactor at the Texas A&M

University Nuclear Science Center Reactor. With this work completed and in

anticipation of other impending conversion projects, INL convened and engaged

the project participants in a structured discussion to capture the lessons learned.

This lessons learned process has allowed us to capture gaps, opportunities, and

good practices, drawing from the project team’s experiences. These lessons will

be used to raise the standard of excellence, effectiveness, and efficiency in all

future conversion projects.


168 of 435


169 of 435




1. INTRODUCTION.............................................................................................................................. 1

2. BACKGROUND................................................................................................................................ 1

3. LESSONS LEARNED PROCESS..................................................................................................... 1

4. LESSONS LEARNED ....................................................................................................................... 2

4.1 General Conclusions.............................................................................................................. 2

4.2 Lessons Learned Meeting Summary ..................................................................................... 3

5. PRESENTATIONS ............................................................................................................................ 4

5.1 Texas A&M University Nuclear Science Center TRIGA Reactor Performance

Analysis................................................................................................................................. 4

5.2 TRIGA Fabrication Process .................................................................................................. 5

6. LESSONS LEARNED ....................................................................................................................... 5

6.1 Initiating Conversion Project................................................................................................. 5

6.1.1 Initiation .............................................................................................................. 5

6.2 Conversion Proposal Process ................................................................................................ 6

6.2.1 Proposal Preparation ........................................................................................... 6

6.2.2 Contract Negotiation ........................................................................................... 6

6.3 Fuel and Hardware Development and Procurement.............................................................. 7

6.3.1 Fuel Specifications and Drawings....................................................................... 7

6.3.2 Fuel Inspection.................................................................................................... 8

6.3.3 Preparation of Facility for Fuel Receipt.............................................................. 8

6.3.4 Reassembly ......................................................................................................... 9

6.4 Core Conversion.................................................................................................................... 9

6.4.1 Fuel Removal ...................................................................................................... 9

6.4.2 Refueling............................................................................................................. 9


170 of 435

6.5 Spent Nuclear Fuel Shipment.............................................................................................. 10

6.5.1 Cask Determination........................................................................................... 10

6.5.2 Transportation Plan/Security Plan..................................................................... 11

6.5.3 Route Assessment ............................................................................................. 11

6.5.4 Certification of University Quality Assurance Programs.................................. 12

6.5.5 Facility Preparations for Spent Nuclear Fuel Activities.................................... 12

6.5.6 Required Shipping Data Preparation................................................................. 12

6.5.7 Shipping Documentation................................................................................... 13

6.5.8 Cask Loading .................................................................................................... 13

6.5.9 Receipt Facility Preparation.............................................................................. 14

6.6 Other issues ......................................................................................................................... 15

6.6.1 Safeguards Information..................................................................................... 15

7. ROUND ROBIN .............................................................................................................................. 15


171 of 435













Argonne National Laboratory

U.S. Department of Energy

General Atomics

highly enriched uranium

Idaho National Laboratory

low-enriched uranium

National Nuclear Security Administration

Nuclear Regulatory Commission

Nuclear Science Center

spent nuclear fuel

Texas A&M University


172 of 435


173 of 435

University Reactor Conversion

Lessons Learned Workshop for

Texas A&M Nuclear Science Center


The Department of Energy’s (DOE) Idaho National Laboratory (INL), under its programmatic

responsibility for managing the University Research Reactor Conversions, has completed the conversion

of the reactor at the Texas A&M University Nuclear Science Center (TAMU NSC). This project was

successfully completed through an integrated and collaborative effort involving INL, Argonne National

Laboratory (ANL), DOE (headquarters and the field office), the Nuclear Regulatory Commission (NRC),

the universities, and the contractors involved in analyses, fuel design and fabrication, and spent nuclear

fuel (SNF) shipping and disposition. With this work completed and in anticipation of other impending

conversion projects, INL convened and engaged the project participants in a structured discussion to

capture the lessons learned. The objectives of this meeting were to capture the observations, insights,

issues, concerns, and ideas of those involved in the reactor conversions so that future efforts can be

conducted with greater effectiveness, efficiency, and with fewer challenges.


As part of the Bush administration’s effort to reduce the amount of weapons-grade nuclear material

worldwide, the National Nuclear Security Administration (NNSA) has established a program to convert

research reactors from using highly enriched uranium (HEU) to low-enriched uranium (LEU) fuel.

The research reactor conversion effort is a critical step under the Global Threat Reduction

Initiative’s Reduced Enrichment for Research and Test Reactors program. As part of this program, NNSA

is minimizing the use of HEU in civilian nuclear programs by converting research reactors and

radioisotope production processes to the use of LEU fuel and targets. The HEU is weapons-grade nuclear

material that can be used to make a nuclear weapon or dirty bomb. The research reactors are secure and

are used for peaceful purposes; however, by converting these reactors to use LEU, a significant step is

made toward ensuring that weapons-usable nuclear material is secure and safeguarded.

Among the list of research reactors targeted for conversion in 2006 were the University of Florida

and Texas A&M University.

Reactor conversions include analyses, LEU fuel fabrication, reactor defuel and refuel activities,

HEU packaging and transportation, and reactor startup.


The process for capturing the lessons learned from this project involved taking the schedule of the

project activities and focusing feedback and discussion on each respective activity. The feedback and

lessons learned discussions were held in an open discussion workshop, including all participating team

members and their representatives. To promote a more expedient discussion at the workshops and to help

the project team focus on the higher priority areas, a survey was developed and sent to project participants

before the workshops. The survey invited those involved in the project to score and offer comments with

regard to the projects activities in which they were involved. The survey was formatted with a 5-point

Likert scale, where 1 was low or “extremely challenging,” and 5 was high or “exceptional.” The surveys


174 of 435

were collected and scores were entered and averaged for each activity. The average score for each activity

is identified in Section 6 of this document.

Based on survey scores and comments, the workshop agenda was established and timeframes were

estimated. Consistent with expectations based on the survey results, the workshop discussions were brief

for the unremarkable areas and more extended and detailed in those areas of greatest significance. The

detailed lessons learned were captured and the themes and general conclusions were then drawn. The

general conclusions and themes tend to apply to all activities (almost as operating principles) and will

benefit future project teams and project managers. The more detailed lessons learned align to given

activities and apply to the project manager and those involved in the given activity, as that activity is



4.1 General Conclusions

This project was clearly a success. Nonetheless, there were many detailed lessons learned regarding

both technical and project management aspects. The specifics are provided in the following sections;

however, some general elements are key to the success of future conversion and spent fuel shipping

projects. Future projects will be conducted most effectively, efficiently, and with a minimum of risks,

interference, and interruptions if the following are an integral part of the project:

Project team composition, which includes a project team composed of individuals who are critical

thinkers, flexible, and committed to the project results (the following was extracted from the

comments submitted: “Having the right people who were willing to buy into the common vision

and mission was critical. Everyone had a great personal work ethic. Having a single person who is

solely dedicated to the project [allowing that person to stay in contact with all parties involved and

to identify and track issues] was instrumental in the success of the project.”).

Communication, including inclusive communications and exchange that provides for effective

sharing of needs, expectations, roles, responsibilities, data, assumptions, schedules, and facility and

equipment constraints.

Use of expertise, including confidence in and effective utilization of the varied expertise and

experience of the team members.

Proactivity and individual levels of initiative.

Early initiation includes the earliest possible initiation of planning and activities at every step in

the project process, thereby minimizing the likelihood of time-critical situations.

Verification and re-verification of data, analyses, specs, assumptions, performance expectations,

and equipment fit and function throughout the project.

Clear and common understanding, including clear expectations of roles, responsibilities,

technical variables, and technical results.

Knowledgeable and informed stakeholders who can advocate for the project, remove barriers,

and support decisions and adjustments needed to ensure project success (e.g., public, political, and



175 of 435

Compile reactor data includes assembly or compilation of the historical documents that reveal

what is known and unknown about the reactor.

Value-added government oversight, in which the public interests are served, objectivity is

retained, but NRC’s experience and expertise is available to the project.

The above list comprised the general themes of the lessons learned meeting. The detailed lessons

learned were discussed in the order of project activities, from initiation to closeout, and are provided in

the following sections.

4.2 Lessons Learned Meeting Summary

The Lessons Learned Workshop for the Texas A&M University Nuclear Science Center convened

on February 21, 2007, at the General Atomics (GA) facilities in San Diego, California. The following

were attendees at the workshop:

Dana Meyer, INL

Eric Woolstenhulme, INL

Doug Morrell, INL

Dale Luke, INL

Jim Wade, DOE-ID

Parrish Staples, DOE-NNSA

Scott Declue, DOE-SRS

Alexander Adams, NRC

Bill Schuser, NRC

John Bolin, GA

Jason Yi, GA

Ken Mushinski, GA

Pierre Colomb, CERCA

Helios Nadal, CERCA

Jim Matos, ANL

Jim Remlinger, TAMU

W Dan Reece, TAMU

Jamie Adam, NAC

Anthony Veca, GA

The following was the agenda for the workshop:

8:00 Welcome and introductory remarks, establish ground rules, and review agenda

8:30 Presentations

TAMU NSC TRIGA Reactor Performance Analysis—TAMU NSC

TRIGA Fabrication Process—TRIGA International


176 of 435

9:00 Discuss and collect lessons learned by each major activity area

Initiating Conversion Project

Conversion Proposal Process

10:15 Break

10:30 Discuss and collect lessons learned by each major activity area (continued)

Fuel and Hardware Development and Procurement

12:00 Lunch

1:00 Discuss and collect lessons learned by each major activity area (continued)

Core Conversion

SNF Shipment

2:20 Break

2:35 Discuss and collect lessons learned by each major activity area (continued)

Other areas needing to be addressed

3:35 Next steps and assignments

4:10 Closing remarks

4:30 Adjourn


5.1 Texas A&M University Nuclear Science Center

TRIGA Reactor Performance Analysis

Dr. Dan Reece summarized the TAMU NSC reactor conversion in his presentation. Dr. Reece

concluded that many things went very well, but there were a few problems. Dr. Reece also gave his

perspective on the lessons to be learned from the conversion work. Highlights from Dr. Reece’s

presentation include the following:

The difference between calculated values for fuel element temperatures and the actual measured

values of the new core

The apparent conflict between calculated values for neutron fluxes and the fluxes derived from foil

experiments in the new core

The importance of interactions and relationships with the various regulators and conversion team



177 of 435

The importance of planning and coordination for the project

The difficulty of locating specific details about the old core.

5.2 TRIGA Fabrication Process

This joint presentation covered the ongoing research concerning the difference between the

calculated values for fuel element temperatures and the actual measured values of the new NSC core.

Additionally, it was shown that the NSC fuel elements fabricated by CERCA were produced in

compliance with GA technical specifications and CERCA’s quality assurance requirements. The fuel

elements were delivered on time and in accordance with the initial manufacturing schedule.

The process for assembling TRIGA elements was discussed. The point was made that inserting the

meats into the cladding is a difficult process because of tight cladding tolerances. About 60% of the fuel

elements must have the fuel meats pressed into the cladding. Only meats and cladding with a large gap

actually just slide in.

For the instrumented fuel elements, the meat diameters were within tolerance, but at the small end

of the ID tolerance. The cladding ID was larger than is allowed per the drawings, but it was determined

that it was within the safety analysis report specifications and was cleared for use. This configuration

translated to a larger than nominal gap between the meat and the cladding. This gap reduces heat transfer

from the meat to the cladding and causes the fuel temperature to be higher than optimal. As the meat

swells from operating the reactor, the gap will decrease and the temperature will be lower.

The ostensible decrease in neutron flux was also discussed. The matter needs further investigation

and foil testing and the results will be documented in a report by GA.


The detailed lessons learned were discussed in order of project activities, from initiation to

closeout, and are provided in the following sections.

6.1.1 Initiation

The average survey score was 3.88.

6.1 Initiating Conversion Project


Some reactor specifications were difficult to

ascertain and came late in the project. Some of

this was because the contract with GA was

finalized later than optimum.


Early involvement of GA is imperative to better

understand the core and project implications

(e.g., fuel and hardware). Also, GA should be

invited to the reactor early in the process, with

procurement and analysis aspects being a key



178 of 435


The initial license amendment followed an old

example rather than following the NRC guidance

document, NUREG-1537. This resulted in some

unnecessary rewriting.


Follow NUREG-1537 rather than relying on

previous amendments. Reviewing past requests

for additional information from NRC may also be

of benefit.

6.2.1 Proposal Preparation

The average survey score was 2.83.

6.2 Conversion Proposal Process


An interactive request for additional information

resolution meeting with all parties involved was a

key activity. This was much more effective than

trading phone calls and emails. The face-to-face

and open, direct communication was key. This

reduced the required time to complete the process

by a factor of 10.


Teamwork is critical to success and efficiency of

the proposal process.

6.2.2 Contract Negotiation

The average survey score was 3.0.


The procurement process on both sides

(i.e., government and university) is problematic.

Lack of a mutual understanding in the

procurement process lends to bogging down the



Promote communications and negotiations

between the principle project parties before going

to the procurement agents. Once the terms are

understood, then the procurement people can be

brought in to complete the process.

Involve both procurement agents early on to

ensure that time is not lost negotiating differences

between processes and waiting for additional

information later.

Early initiation involvement and coordination of

contracts/procurement staff are crucial.


179 of 435

6.3 Fuel and Hardware Development and Procurement

6.3.1 Fuel Specifications and Drawings

The average survey score was 2.20.


Specifics about the fuel and hardware

procurement were confusing because of the varied

opinions and individual spreadsheets.

Specifics about the fuel and hardware

procurement were confusing because no cluster

assembly information was provided to the


The gram loading for the fuel elements was on the

low end of the required range.

Having the fabrication data for the new fuel

earlier in the process would be helpful.


It would be helpful to get everyone together at the

onset and create a format for presenting the fuel

and hardware information that everyone agrees to

and understands. Drawings and other historical

documents could be presented at the initial

meeting. The various parties could discuss the

data to ensure mutual agreement on what needs to

be ordered. One person could be charged with

keeping the fuel and hardware spreadsheet

updated and issued to the interested parties.

See above recommendation. Also, GA could

provide information about which upper and lower

adapters (and other hardware) are required for the

various cluster types.

The project should advise TRIGA International to

load the elements on the heavy side to maximize

the amount of fuel in the core. This maximizes the

per element value when considering the dollars

spent on fabrication, shipping, usage, and disposal

of a fuel rod.

This effort must be worked with the university to

ensure that all needed information is provided in

the data packages.

As a minimum, the data packages should be

included with the fuel shipment.

Caution must be taken to properly handle

proprietary information.


180 of 435

6.3.2 Fuel Inspection

The average survey score was 4.00.


The fuel receipt inspection worked well at the

reactor and at CERCA.

After inspection, it was unclear who took

ownership of the fuel.


The right people were involved in the inspection

(i.e., vendor, quality assurance personnel, and

receivers). A coordination meeting was held

before the inspection so that everyone involved

was well advised and clearly understood their

rolls. A source inspection was conducted at the

manufacturer site in France before shipment so

that the receipt inspection at the university was

less complex and time intensive.

There needs to be a clear transfer of responsibility

so that it is understood who owns the fuel at any

given time. A signature process could be devised

that formally documents and completes the

ownership transfer.

6.3.3 Preparation of Facility for Fuel Receipt

The average survey score was 3.60.


The truck/trailers arrived at NSC with the

containers positioned toward the front of the

trailers and with some of the containers turned

sideways; this precluded access with a pallet jack

or forklift.


Information about the shipping trucks and loading

configuration is important to expedite the receipt

of the fuel at the reactor. Ii would be best if the

trailers had a side-loading capability to make it

easier to unload the shipments with a forklift. The

INL should facilitate communications between the

shipper and reactor. The INL should consider

writing truck specifications into the contract with

the shipping company.


181 of 435

6.3.4 Reassembly

The average survey score was 3.33.


It may take specific training to open and

reassemble the shipping containers for return



Dave Capp at the INL was this person for the

TAMU NSC project. He did a great job. The INL

needs to secure a similar individual on all future


6.4 Core Conversion

6.4.1 Fuel Removal

The average survey score was 3.33.


Fuel removal went well at NSC.


Video taping of the processes will serve as a great

resource for those who must perform the tasks


It may be beneficial to have the core parameters

measured and documented before the reactor is

shutdown for refueling (i.e., fuel temperatures,

neutron flux, and control rod positions). The

measurements may be useful in analysis following


6.4.2 Refueling

The average survey score was 3.50.


Personnel turnover at the universities can

sometimes cause a loss of drawings,

specifications, and other documents. This can

make converting the reactor and SNF shipments a

significant challenge.


Early notification of the documentation needs by

the INL should be made to the university. This

will allow more time for locating the information.


182 of 435


Hardware for NSC had to be re-machined because

of lack of information. GA was quick to respond

to all issues identified; therefore, the issues were

resolved quickly.

The instrumented fuel elements read higher than

expected from the earlier analysis.

Thermocouple leads on the instrumented fuel

elements were too long for the NSC configuration.

The NSC cut the leads, but then required a half

day to re-work the lead wires.


An early start can also allow time for reactor

personnel to physically verify reactor components

before procurement of the parts.

Because of this issue, we must pay greater

attention to the details of the reactors.

Instrumented fuel elements cladding and fuel meat

gaps must be tighter to ensure that the actual

readings are more representative of the core


The correct length should be identified before

fabrication at CERCA. Cutting the thermocouple

leads is standard practice, but had it been

considered ahead of time, the materials and

capabilities could have been in place onsite to

significantly reduce the time and effort required.

6.5.1 Cask Determination

The average survey score was 3.67.

6.5 Spent Nuclear Fuel Shipment


The SNF shipment activities are very difficult for

universities that do not normally ship SNF.


Updated guidance from NRC regarding SNF

shipping would be helpful.

The INL should consider contracting with other

companies or experienced shippers to help the


The DOE could consider taking ownership of the

shipping rather than NRC.

It is important to field-verify all procedures, plans,

and such before shipping.

Not everyone with a need to know had copies of

the SNF shipping orders, specifically, some

information needed to be included in shipping

documents prepared by others. This was caused,

in part, by a Safeguards Information “blackout”

for information from NRC.

Safeguarded Information issues have been

resolved at NRC. This situation should not occur

in the future.


183 of 435


The cask was identified much later than

appropriate by INL. The tardiness of the contract

with the cask vendor caused delays in the facility

preparations. This caused unnecessary stress and

work for NSC.


The INL needs to make cask arrangements as

soon as possible.

The cask vendors need to make detailed site

assessments early in the project.

Drawings and procedures need to be supplied to

the reactor as soon as possible.

The project should make early visits to the

university and discuss the tasks associated with

SNF shipping.

6.5.2 Transportation Plan/Security Plan

The average survey score was 3.0.


Transport and security plans can be

time-consuming and labor intensive.

Guidance form NRC regarding HEU shipments

was not as clear or up-to-date as it could have



The project should get the most effective and

reliable sources to carry out the functions of

developing the plans.

The current guidance should be updated. The

NRC suggests we work with one of the current

licensees to get better understanding of the current


6.5.3 Route Assessment

The average survey score was 3.2.


Communication about the route assessment

documents was sometimes inefficient.


It was suggested to involve other subject matter

experts during the route assessment.

Communication lines between all parties

(i.e., shipper, INL, cask vendor, and other

facilitating companies) need to be open.


184 of 435

6.5.4 Certification of University Quality Assurance Programs

The average survey score was 3.0.


Certifying as an SNF shipper can be extensive.


Begin activities early and the program should

provide assistance to the facility, as needed.

6.5.5 Facility Preparations for Spent Nuclear Fuel Activities

The average survey score was 3.60.


The SNF shipping preparations are wide-ranging

and often difficult.


Need to ensure early, comprehensive planning

with attention to detail.

Start the process to procure support equipment

(e.g., cranes) early. This worked well for us.

6.5.6 Required Shipping Data Preparation

The average survey score was 2.5.


Required shipping data preparations can be

laborious and resource intensive.


Use of the parametric study on TRIGA fuel

burnups for completing the required shipping data

radioisotope and decay heat tables would be very


The university may need to check and validate the

applicability of the standard decay heat data.


185 of 435

6.5.7 Shipping Documentation

The average survey score was 3.0.


Shipping documentation, such as SNF

Transportation Plans and the Bill of Lading, were

very involved for an unfamiliar shipper.


The INL’s help was invaluable. The university

always felt that they had an ally and

knowledgeable resource to facilitate the process.

The project university also had confidence in the

experts and could trust their advice and

experience during document development.

6.5.8 Cask Loading

The average survey score was 3.67.


The SNF roles and responsibilities were well

defined going into the SNF shipping activities.


The NSC had been informed early in the project

that they were in charge and responsible for the

activities. All other entities also understood this at

the outset of the project. This hierarchy resulted in

effective working relationships between the

project entities.

We need to maintain this level of rigor and

discipline for future conversion projects.


186 of 435


The cask sat loaded at NSC over the weekend.

This was an unfavorable situation for the shipper.


Many notifications and logistics have to be

worked out for the moment the shipment leaves

the facility. Changes to planned shipping dates are

difficult if not impossible to effect. The SNF

loading was to begin on Monday. It was estimated

that loading would take about 5 days to complete,

thereby finishing on Friday. Weekends are not the

preferred times to start shipments; therefore, the

INL shipping coordinator felt that it was best to

leave the weekend for schedule contingency in the

case loading took longer than expected.

The project needs to fully communicate this

thinking and the firm shipping dates for the


In future shipments, the project needs to consider

the trade-off between shipping on a weekend or

leaving the loaded cask at the facility for the


6.5.9 Receipt Facility Preparation

The average survey score was 3.33.


There was some confusion on who was making

arrangements for the return shipments of the

Nuclear Assurance Corporation equipment. Just

days before the shipment, it was found that the

arrangement for a truck had not been made.


It needs to be clearly established, well in advance

of the cask loading dates, who is responsible for

planning and executing the tasks for all legs of the

shipments. This includes equipment shipment to

and from the various facilities.


187 of 435

6.6 Other issues

6.6.1 Safeguards Information

The average survey score was 3.0.


There was a bit of confusion regarding what

constitutes safeguards information and who can

have access to it.


The various entities involved with the project

need to clearly understand their responsibilities

and limitation under this order. The project should

consider holding an onsite meeting to clarify the

policies with the project team.


In concluding the discussion of the lessons learned, all participants were invited to reiterate,

summarize, or offer any other lessons learned. The following list provides their final thoughts:

Well defined goals and responsibilities are essential to success. All team members must understand

their responsibilities. Because of division of responsibilities at INL, it was confusing to NSC who

at INL was in charge of some tasks.

It is important for the project team to understand that if a task can be done early then it should be.

Performing tasks just-in-time would have caused the NSC conversion to fail because of

unexpected, last-minute tasks and issues. In other words, completing tasks early will allow the

project to be flexible enough to address the last minute challenges.

The NSC project went well in spite of the minor setbacks and challenges. The project will be held

to a higher standard of performance next time.

There will be some weeks/months after the project where parties will need to work together to get

some things accomplished and review present issues of conversion.

The next lessons learned analysis needs to include a specific “what went well” column so that we

can capture the things that worked.


This lessons learned process has allowed us to capture gaps, opportunities, and good practices,

drawing from the project team’s experiences. The process is inclusive and offers an opportunity for every

entity that “touched” the project to share from its experience. These lessons will be used to raise the

standard of excellence, effectiveness, and efficiency in all future conversion projects. Despite making

improvements to successive projects by addressing the lessons we have learned on this project,

conducting a lessons learned activity will be vital to each conversion project as technologies, regulations,

and other aspects of the environment change and influence success. It is recognized we cannot become

complacent, nor adopt a mindset that the process has been “perfected.”


188 of 435


University Reactor

Conversion Lessons

Learned Workshop for

the University of Florida

Eric C. Woolstenhulme

Dana M. Meyer

April 2007

The INL is a U.S. Department of Energy National Laboratory

operated by Battelle Energy Alliance

189 of 435


University Reactor Conversion Lessons Learned

Workshop for the University of Florida

Eric C. Woolstenhulme

Dana M. Meyer

April 2007

Idaho National Laboratory

Idaho Falls, Idaho 83415

Prepared for the

U.S. Department of Energy

Office of Nuclear Nonproliferation and Security Affairs

Under DOE Idaho Operations Office

Contract DE-AC07-05ID14517

190 of 435


The Department of Energy’s Idaho National Laboratory, under its

programmatic responsibility for managing the University Research Reactor

Conversions, has completed the conversion of the reactor at the University of

Florida. With this work completed and in anticipation of other impending

conversion projects, INL convened and engaged the project participants in a

structured discussion to capture the lessons learned. This lessons learned process

has allowed us to capture gaps, opportunities, and good practices, drawing from

the project team’s experiences. These lessons will be used to raise the standard of

excellence, effectiveness, and efficiency in all future conversion projects.


191 of 435


192 of 435




1. INTRODUCTION.............................................................................................................................. 1

2. BACKGROUND................................................................................................................................ 1

3. LESSONS LEARNED PROCESS..................................................................................................... 1

4. LESSONS LEARNED ....................................................................................................................... 2

4.1 General Conclusions.............................................................................................................. 2

4.2 Lessons Learned Meeting Summary ..................................................................................... 3

5. LESSONS LEARNED BY PROJECT ACTIVITY........................................................................... 4

5.1 Initiating Conversion Project................................................................................................. 4

5.1.1 Initiation .............................................................................................................. 4

5.2 Conversion Proposal Process ................................................................................................ 5

5.2.1 Contract Negotiation ........................................................................................... 5

5.2.2 Proposal Preparation ........................................................................................... 6

5.2.3 Submittal of Proposal.......................................................................................... 7

5.2.4 Requests for Additional Information................................................................... 8

5.2.5 Final Review and Comment on Proposal............................................................ 8

5.2.6 Conversion Order ................................................................................................ 9

5.3 Fuel and Hardware Development and Procurement.............................................................. 9

5.3.1 Fuel Specifications and Drawings....................................................................... 9

5.3.2 Fuel Fabrication Statement of Work and Procurement Documents.................. 10

5.3.3 Fuel Inspection.................................................................................................. 11

5.3.4 Preparation of Facility for Fuel Receipt............................................................ 11

5.3.5 Reassembly ....................................................................................................... 12

5.4 Core Conversion.................................................................................................................. 12

5.4.1 Fuel Removal .................................................................................................... 12

5.4.2 Refueling........................................................................................................... 13


193 of 435

5.5 Spent Nuclear Fuel Shipment.............................................................................................. 14

5.5.1 Cask Determination........................................................................................... 14

5.5.2 Transportation Plan/Security Plan..................................................................... 14

5.5.3 Route Assessment ............................................................................................. 15

5.5.4 Certification of University Quality Assurance Programs.................................. 15

5.5.5 Facility Preparations for Spent Nuclear Fuel Activities.................................... 15

5.5.6 Support Equipment/Tools for Spent Nuclear Fuel Activities ........................... 16

5.5.7 Appendix A Preparation.................................................................................... 16

5.5.8 Shipping Documentation................................................................................... 17

5.5.9 Cask Loading .................................................................................................... 17

5.5.10 Receipt Facility Preparation.............................................................................. 17

5.6 Other Issues ......................................................................................................................... 18

5.6.1 Safeguarded Information................................................................................... 18

6. ROUND ROBIN .............................................................................................................................. 18

7. ACTIONS......................................................................................................................................... 19

8. CONCLUSION ................................................................................................................................ 19


194 of 435











Argonne National Laboratory

U.S. Department of Energy

General Atomics

highly enriched uranium

Idaho National Laboratory

low-enriched uranium

National Nuclear Security Administration

Nuclear Regulatory Commission

spent nuclear fuel


195 of 435


196 of 435

University Reactor Conversion Lessons Learned

Workshop for the University of Florida


The Department of Energy’s (DOE) Idaho National Laboratory (INL), under its programmatic

responsibility for managing the University Research Reactor Conversions, has completed the conversion

of the reactor at the University of Florida. This project was successfully completed through an integrated

and collaborative effort involving INL, Argonne National Laboratory (ANL), DOE (headquarters and the

field office), the Nuclear Regulatory Commission (NRC), the universities, and the contractors involved in

analyses, fuel design and fabrication, and spent nuclear fuel (SNF) shipping and disposition. With this

work completed and in anticipation of other impending conversion projects, INL convened and engaged

the project participants in a structured discussion to capture the lessons learned. The objectives of this

meeting were to capture the observations, insights, issues, concerns, and ideas of those involved in the

reactor conversions so that future efforts can be conducted with greater effectiveness, efficiency, and with

fewer challenges.


As part of the Bush administration’s effort to reduce the amount of weapons-grade nuclear material

worldwide, the National Nuclear Security Administration (NNSA) has established a program to convert

research reactors from using highly enriched uranium (HEU) to low-enriched uranium (LEU) fuel.

The research reactor conversion effort is a critical step under the Global Threat Reduction

Initiative’s Reduced Enrichment for Research and Test Reactors program. As part of this program, NNSA

is minimizing the use of HEU in civilian nuclear programs by converting research reactors and

radioisotope production processes to the use of LEU fuel and targets. The HEU is weapons-grade nuclear

material that can be used to make a nuclear weapon or dirty bomb. The research reactors are secure and

are used for peaceful purposes; however, by converting these reactors to use LEU, a significant step is

made toward ensuring that weapons-usable nuclear material is secure and safeguarded.

Among the list of research reactors targeted for conversion in 2006 were the University of Florida

and Texas A&M University.

Reactor conversions include analyses, LEU fuel fabrication, reactor defuel and refuel activities,

HEU packaging and transportation, and reactor startup.


The process for capturing the lessons learned from this project involved taking the schedule of the

project activities and focusing feedback and discussion on each respective activity. The feedback and

lessons learned discussions were held in an open discussion workshop, including all participating team

members and their representatives. To promote a more expedient discussion at the workshops and to help

the project team focus on the higher priority areas, a survey was developed and sent to project participants

before the workshops. The survey invited those involved in the project to score and offer comments with

regard to the projects activities in which they were involved. The survey was formatted with a 5-point

Likert scale, where 1 was low or “extremely challenging,” and 5 was high or “exceptional.” The surveys

were collected and scores were entered and averaged for each activity. The average score for each activity

is identified in Section 5 of this document.


197 of 435

Based on survey scores and comments, the workshop agenda was established and timeframes were

estimated. Consistent with expectations based on the survey results, the workshop discussions were brief

for the unremarkable areas and more extended and detailed in those areas of greatest significance. The

detailed lessons learned were captured and the themes and general conclusions were then drawn. The

general conclusions and themes tend to apply to all activities (almost as operating principles) and will

benefit future project teams and project managers. The more detailed lessons learned align to given

activities and apply to the project manager and those involved in the given activity, as that activity is



4.1 General Conclusions

This project was clearly a success. Nonetheless, there were many detailed lessons learned regarding

both technical and project management aspects. The specifics are provided in the following sections;

however, some general elements are key to the success of future conversion and spent fuel shipping

projects. Future projects will be conducted most effectively, efficiently, and with a minimum of risks,

interference, and interruptions if the following are an integral part of the project:

Project team composition, which includes a project team composed of individuals who are critical

thinkers, flexible, and committed to the project results (the following was extracted from the

comments submitted: “Having the right people who were willing to buy into the common vision

and mission was critical. Everyone had a great personal work ethic. Having a single person who is

solely dedicated to the project [allowing that person to stay in contact with all parties involved and

to identify and track issues] was instrumental in the success of the project.”).

Communication, including inclusive communications and exchange that provides for effective

sharing of needs, expectations, roles, responsibilities, data, assumptions, schedules, and facility and

equipment constraints.

Use of expertise, including confidence in and effective utilization of the varied expertise and

experience of the team members.

Proactivity and individual levels of initiative.

Early initiation includes the earliest possible initiation of planning and activities at every step in

the project process, thereby minimizing the likelihood of time-critical situations.

Verification and re-verification of data, analyses, specs, assumptions, performance expectations,

and equipment fit and function throughout the project.

Clear and common understanding, including clear expectations of roles, responsibilities,

technical variables, and technical results.

Knowledgeable and informed stakeholders who can advocate for the project, remove barriers,

and support decisions and adjustments needed to ensure project success (e.g., public, political, and


Compile reactor data includes assembly or compilation of the historical documents that reveal

what is known and unknown about the reactor.


198 of 435

Value-added government oversight, in which the public interests are served, objectivity is

retained, but NRC’s experience and expertise is available to the project.

The above list comprised the general themes of the lessons learned meeting. The detailed lessons

learned were discussed in the order of project activities, from initiation to closeout, and are provided in

the following sections.

4.2 Lessons Learned Meeting Summary

The Lessons Learned Workshop for the University of Florida convened on February 22, 2007, at

the General Atomics (GA) facilities in San Diego, California. The following were attendees at the


Dana Meyer, INL

Eric Woolstenhulme, INL

Doug Morrell, INL

Dale Luke, INL

Jim Wade, DOE-ID

Parrish Staples, DOE-NNSA

Scott Declue, DOE-SRS

Alexander Adams, NRC

Anthony Veca, GA

Jason Yi, GA

Ken Mushinski, GA

Jim Matos, ANL

Ali Haghighat, UF

Benoit Dionne, UF

Roy Boyd, STS

Chip Shaffer, BWXT

Bill Schuser, NRC

The following was the agenda for the workshop:

8:00 Welcome and introductory remarks

Establish ground rules and review agenda

8:30 Discuss and collect lessons learned by each major activity area

Initiating Conversion Project

Conversion Proposal Process

10:15 Break

10:30 Discuss and collect lessons learned by each major activity area (continued)


199 of 435

Fuel and Hardware Development and Procurement

12:00 Lunch

1:00 Discuss and collect lessons learned by each major activity area (continued)

Core Conversion

SNF Shipment

2:20 Break

2:35 Discuss and collect lessons learned by each major activity area (continued)

Other areas needing to be addressed

3:35 Next steps and assignments

4:10 Closing remarks

4:30 Adjourn


The detailed lessons learned were discussed in order of project activities, from initiation to

closeout, and are provided in the following sections.

5.1.1 Initiation

The average survey score was 3.88.

5.1 Initiating Conversion Project


Open communication between the university and

the program went a long way in resolving a

question of roles and responsibilities. In this case,

the program analysts wanted to conduct the

analyses, while the university believed they

should perform them. The university saw it as an

opportunity to thoroughly understand their

reactor. A meeting was held to discuss the

university’s desires, rationale, and subsequently

their capabilities and scope of analyses, and it was

agreed to allow the university to do the analyses,

with the program analysts providing guidance and

expertise, as needed.


A valuable lesson learned in this regard was for

the program to understand and respect the

university’s objectives, and the related

programmatic benefits, and assist them as needed

to accomplish their goals.

With regard to the question of who would do the

analyses, we needed confidence in each others’

respective capabilities, clarity, and agreement of

roles based on those capabilities, and subsequent

demonstration of those capabilities in the

undertaking of the project.


200 of 435


The university team was segregated a bit and it

was not clear if all the necessary information was

being shared appropriately.

Insufficient coordination of reviews caused delays

and confusion.


A kick-off meeting with the university, designer,

fabricator, analyst, shipping support, and shipper

should take place as soon as possible to facilitate

formal and systematic documentation of ALL

technical and functional requirements for the

entire project in a technical and functional

requirements document. This would clarify roles,

expectations, and requirements, and especially

ensure that each piece of the design/specification

could be verified against those requirements.

Technical and functional requirements documents

would be signed and become the “binding”

document that everyone must abide by. Doing this

will help eliminate many of the design problems

that were experienced on this project. It would be

a living document that gets revisited at each


Explicitly discuss “who else” needs to be “on

board” to determine the support needed and

establish essential contacts for review and


Direct the university to provide, at the preliminary

meetings, a list of those individuals that they want

to review drawings, specs, and such.

5.2.1 Contract Negotiation

The average survey score was 3.0.

5.2 Conversion Proposal Process


Delays were experienced in the contracting

process due, in large part, to lack of understanding

of the work and time constraints by the contracts


Procurement and contracts personnel play a

pivotal role in managing risks and clarifying

obligations through the contracting process.

However, their effectiveness can be suboptimized

if they are ill-informed and are not involved early.


Involve contracts/procurement people early in the

process to promote an understanding of the work

that mitigates nonessential delays.

Start negotiations early to ensure the procurement

process is less troublesome. Involve procurement

personnel from both parties early, so that all

parties are informed and working together.


201 of 435

5.2.2 Proposal Preparation

The average survey score was 2.83.


The age and history of any given reactor

potentially allows for the likelihood that changes

have occurred in designs, equipment,

functionality, and such. These changes impact the

design, analysis, and any number of activities on

these projects.

Lots of time was spent up front trying to

determine format, content, and such. A clearer

guideline of what the format (and some

boilerplate) would be extremely helpful in

preparing the proposal.

Although proposals are not due until a specific

date, involvement of NRC to conduct upfront

negotiations and clarify expectations and

contractual obligations DURING proposal

development would greatly improve the process.

Proposal preparation went well. Lots of

interaction back and forth with a clear,

comprehensive plan and identification of who was

responsible for what.

The NRC oversight was value-added yet remained

objective. Several aspects of the proposal can only

be decided by NRC; therefore, early, open

involvement is crucial. Use NRC as a technical

resource/sanity check, and not just for answering

administrative-type questions (e.g., changes to

technical specifications), puts NRC in a position

to “advocate” the conversion proposal on behalf

of the university. Anytime the proposal preparer

questions how NRC might react to a point, he/she

needs to call and ask.


Advise university early (at the start of the process

or at the initial phase of the analysis) to recover

and provide any historical documents, geometries,

specifications, and such that are available. They

also need to identify what information is missing

so they can conduct whatever activities are

necessary to fill those data gaps.

Now that it has been published, we need to use the

NRC guide/template when preparing the proposal.

Involve NRC in the proposal process as soon as

reasonable regarding those areas where NRC

involvement is stipulated (i.e., before the postal

worker drops it off).

Embrace a collaborative and interactive operating

philosophy, yielding constructive and clear

communication and exchange.

Use NRC as a technical resource/sanity check and

not just for answering administrative-type

questions. Anytime the proposal preparer

questions how NRC might react to a point, he/she

needs to call NRC and ask.


202 of 435


There is a risk in preparing the conversion

proposal while developing the fuel, because gaps,

tolerances, and such must be known, documented,

and understood.


Complete the design before preparing the

conversion proposal. This will ensure the correct

design specs are included. The proposal can then

move forward with significantly minimized risk.

Transmit final drawings for fuel design to NRC to

support their review of the analyses.

Picking overly restrictive tolerances causes safety

limits to come down. Any future changes in

design means analyses have to be revisited and

sometimes revised. Over conservatism in

tolerances may make fabrication nearly

impossible. For example, the University of Florida

proposal asked for a ±1 mil tolerance across a

26-in. element. This was rigorously discussed

internally at the University of Florida and ANL

(who conducted the analysis), but was not

discussed with the designers at INL who would

have resisted such a limited tolerance.

Be less restrictive during the analysis so that we

are not so limited/restricted in the design.

The fabricator and the designer MUST collaborate

very closely at every phase of the process, almost

as if they were the same entity, so that nothing is