Two presentations on modifications of research reactor cores to address the technical challenges of conversion were given by Panel 2.1 speakers: John Stevens (Argonne National Laboratory) provided a U. S. viewpoint on core modifications (Stevens, 2011), and I. T. Tetiyakov (NIKIET) provided a Russian viewpoint (Tetiyakov, 2011).

U. S. Viewpoint on Core Modifications

John Stevens

The conversion of a research reactor from HEU to LEU fuel can result in performance penalties in the reactor, primarily arising from the reduced density of uranium-235 and absorption of neutrons by uranium-238. Modi­fications to a reactor core may be required to overcome these penalties. Several core modification strategies have been used to overcome the penal­ties associated with the conversion of U. S.-origin research reactors; these include modifications to the following:

• fuel plate thickness and reflector locations;

• fuel meat thickness;

• uranium and burnable absorber loading; and

• fueled height of the core.

When making modifications to a reactor core one should strive to change as little as possible. Two particularly successful strategies for over­coming performance penalties that entail minimal changes are (1) tuning the burnable absorber to match the fuel composition; and (2) if cost is ac­ceptable, modifying reflector materials and/or geometries.

Of course, the fuel will, by definition, change from HEU to LEU during the conversion process, and the LEU fuel must be “acceptable” for conver­sion. An LEU fuel is considered to be acceptable for conversion when it meets the following criteria:

• Qualified: the fuel assembly has been successfully irradiation tested and is licensable.

• Commercially available: The fuel assembly is available from a com­mercial manufacturer.

• Suitable: The fuel assembly satisfies the criteria for LEU conver­sion of a specific reactor; safety criteria are satisfied; fuel service lifetime is comparable to current HEU fuel; and the performance of experiments is not significantly lower than for HEU fuel.

• The reactor operator and regulator agree to accept fuel assembly for conversion.

Successful conversion requires the involvement of reactor operators to un­derstand their needs and constraints.

The following examples were presented to illustrate some of the core modification options that are available to overcome conversion penalties. Some of the reactors described in these examples have already been con­verted, whereas others have not yet been converted.

David Cook

HFIR currently operates at 85 MW—following a derating from 100 MW in the early 1990s because of embrittlement of the reactor pres­sure vessel—using a U3Og-Al dispersion fuel that is 93 percent enriched in uranium-235. HFIR’s original primary mission was the production of transuranic isotopes. With the addition of the cold neutron source in 2007, the facility began hosting world-class cold and thermal neutron scattering research. The facility also meets critical needs for materials irradiation and the performance of neutron activation studies.

The reactor core is cooled and moderated by pressurized light water and is very small (50.8 cm active fuel height and 43.5 cm diameter). The HFIR core contains one inner and one outer cylindrical fuel element (see Figure 3-5). At the center of the inner fuel element is a 13 cm-diameter

hole (the “flux trap”) that contains vertical experimental targets for isotope production (californium). Outside the fuel elements is a beryllium reflector that contains additional experimental positions.

The inner fuel element contains 171 fuel plates, and the outer fuel ele­ment contains 369 fuel plates. The fuel plates, are involute-shaped, and the fuel meat is radially contoured along the involute—the fuel distribution is peaked in the center and thinner on the edges to suppress power peaking (see Figure 3-6). The inner element plates also contain a burnable poison (boron-10). These fuel plates are complex to manufacture because of the plate form and the welds at the sideplates.

HFIR has a number of unique design features that complicate conver­sion; consequently, it will be the most complex—and the last—of the U. S. research reactors to convert from HEU to LEU. Conversion will not occur until it is clear that the reactor’s primary operating missions and safety will not be significantly impacted. This includes maintaining the very high fluxes that HFIR is capable of generating, particularly in the flux trap region.

HFIR cannot be converted until an appropriate LEU fuel has completed development and qualification. Like MITR, HFIR’s unique fuel assemblies and highly compact core complicate conversion. Also like MITR, the use of high-density UMo monolithic LEU fuel along with additional changes in the reactor design is likely to allow for conversion. The new fuel will be 19.75 percent enriched in uranium-235 and have a density of 15.5 gU/cm3.

Radial

Contouring

Profiles

FIGURE 3-6 Radial contouring of fuel plates for reference LEU fuel design. The reference LEU fuel elements are shown in cross — section. The inner fuel element, shown on the left, will be more dramatically asymmetric than the outer fuel element, shown on the right, although both fuel elements are noticeably asymmetric. SOURCE: Cook (2011). Image courtesy of Oak Ridge National Laboratory.

The RBT-6 and RBT-10/2 reactors are pool-type reactors of similar design. The RBT-6 operates with 56 fuel elements at a power of 6 MW, whereas RBT-10/2 operates with 78 fuel elements at a power of 10 MW. Both reactors have neutron flux densities of about 1 x 1014 n/cm2-s. The fuel for both reactors is UO2 dispersed in a copper-beryllium matrix en­riched to 90 percent.

Although these reactors operate at full power most of the time, their experimental channels (up to 8 for RBT-6, slightly more channels for RBT- 10/2) only have about 50 percent utilization. There is interest in increas­ing the usage of these reactors. Possible additional experimental activities include silicon doping, isotope production (including molybdenum-99 pro­duction), testing of industrial materials, and neutron capture therapy. Some of these activities would require redesign of the experimental channels.

BOR-60

BOR-60 is a 60 MW sodium-cooled fast reactor that can produce up to 12 MW of electricity. It is fueled with UO2 or UO2-PuO2 fuel with uranium-235 enrichments of 45-90 percent and plutonium content of 70 percent. It has a maximum neutron flux density of 3.7 x 1015 n/cm2-s.

This reactor is currently used for test irradiations of reactor fuels and materials, including new fuels, cladding, and structural materials for fast reactors, water cooled reactors, and fusion reactors. It is also being used for transmutation research, other fuel cycle research, and isotope produc­tion. The experimental applications could be expanded to include advanced reactor and fuel cycle research.

A. N. Chebeskov

As noted in Chapter 1, the lack of availability of special nuclear mate­rial (SNM) that can be used to build a nuclear weapon is widely agreed to be a major barrier to nuclear proliferation (see Chapter 1). Thus, an essential part of understanding the proliferation risk associated with a research reactor involves understanding how straightforward it would be for a host state or terrorist organization to successfully misuse the reactor’s fuel material.

The attractiveness of a nuclear material from a proliferator’s point of view is determined in large part by a material’s ability to sustain a nuclear chain reaction. Material attractiveness is also influenced by whether it is necessary to process the material to make it usable in a nuclear weapon. To categorize fissile materials qualitatively, four categories (classes) might be used: very attractive, attractive, low attractive, and unattractive.

Several variables are relevant to the attractiveness of SNM. For ex­ample, for a given quantity of uranium, its attractiveness is proportional to both its enrichment and its mass. Higher-enriched materials are more attractive than lower-enriched materials; for example, HEU enriched to 90 percent uranium-235 is far more attractive than LEU, which is re­garded to be unattractive. Similarly, higher masses are more attractive than lower masses for a given level of enrichment. In general, the higher the enrichment, the less mass is required to obtain an equivalent amount of uranium-235.

Of course, nuclear weapons can be constructed using plutonium as well, but it is difficult to compare the attractiveness of different materials. Different grades of plutonium can be rated relative to one another as reac­tor grade (less attractive) and weapons grade (more attractive). However, very highly enriched uranium is the most desirable material for a potential proliferator, because of the relative simplicity of constructing a nuclear explosive device using HEU as opposed to plutonium.

As an example, at the MEPhI reactor, the small size and mass of very highly enriched fuel assemblies represent a higher theft risk than heavier power reactor fuel assemblies, especially for fresh fuel assemblies. The uranium contained in the MEPhI fuel assemblies would not need further enrichment to be usable in a nuclear explosive device. For irradiated fuel assemblies this risk is smaller because of the presence of strong radiation.

The availability of HEU—particularly in the civilian sector—is a sig­nificant proliferation and security concern. In 2001, the U. S. National Research Council stated in its report, Making the Nation Safer, that “(t)he primary impediment that prevents countries or technically competent ter­rorist groups from developing nuclear weapons is the [lack of] availabil­ity of special nuclear material (SNM),[8] especially HEU” (NRC, 2001). The availability of HEU in the civilian sector—as opposed to the military sector—is of particular concern, because resources may not be available or used to protect the material adequately during storage or transport.

If HEU is available, either stolen or purchased, it is plausible that a nuclear weapon could be built by either a state or a non-state actor.[9] The technical barriers to constructing such a weapon are not impassably high. As Pablo Adelfang of the International Atomic Energy Agency (IAEA) noted during the symposium (Adelfang, 2011), individuals with a basic knowledge of physics and machining could build a functioning bomb from stolen HEU. This is largely because HEU is only weakly radioactive—mak­ing it relatively easy to handle—and because such a device would not re­quire explosive testing to be assured of some yield.

In the civilian sector, HEU is primarily used to fuel research reactors and produce radioisotopes for use in medical procedures. The stockpiles of HEU held for these purposes and others are significant. At the end of 2003, the estimated global stockpile of HEU (both civilian and military) was around 1,900 metric tons. Although the vast majority of this HEU is under military control, about 175 metric tons is civilian HEU (ISIS, 2005). This quantity of HEU is sufficient to fabricate about 3,500 nuclear weapons.[10] The vast majority of this civilian HEU is located in the United States (124 metric tons) and in Russia (15-30 metric tons) (ISIS, 2005).

The potential proliferation risk associated with the use of HEU-fueled research reactors—the focus of the symposium and this summary report— arises from the need to transport and store both unirradiated and irradi­ated[11] HEU fuel. This fuel must be protected at all times and is potentially vulnerable to theft while in transit, including across national borders. Proliferation risk exists even in nuclear weapons states.

It is possible to replace HEU in many civilian applications with LEU, which is considered to have a lower proliferation risk because it is not suit­able for use in a nuclear device. Such replacements are possible using cur­rent technologies or technologies that are under development. For example, in 2009, the NRC found that the HEU targets used for the large-scale pro­duction of the medical isotope molybdenum-99 could be replaced by LEU targets (NRC, 2009). Similarly, many existing research reactors can operate using LEU fuel rather than HEU fuel (see Chapters 2 and 3 of this report). In fact, as discussed elsewhere in this report, many reactors have been suc­cessfully converted from HEU to LEU fuel, and many other conversions are under way. The continuation of this trend could significantly reduce the proliferation risk associated with the civilian trade in HEU.

As will be discussed in the next section, 40 percent of the world’s op­erating research reactors are located in the United States and Russia, and nearly all of the world’s research reactors are fueled with either U. S.- or R. F.-origin fuel. For these reasons among others, the United States and Russia combined have significant influence on the nature and extent of the worldwide trade in civilian HEU.

Paul Wilson

UWNR is a 1 megawatt (MW) TRIGA pool reactor (see Chapter 1) housed on the University of Wisconsin campus in Madison, Wisconsin. Its primary mission is the training of undergraduate and graduate nuclear engineering students; however, it is also used to perform research, including irradiation for neutron activation analysis.

The reactor first went critical as a 10 kilowatt (kW) LEU-fueled reac­tor in 1961 and, following several power upgrades, was converted to HEU fuel in 1979. It was converted back to LEU fuel 30 years later, successfully achieving criticality in 2009. At that time, UWNR was converted from us­ing 70 percent enriched TRIGA-FLIP (Fuel Life Improvement Program) fuel to TRIGA LEU 30/20 (30 percent uranium by weight, 20 percent enriched) fuel. The new LEU fuel is, like the previous FLIP fuel, a standard TRIGA — type fuel element containing erbium-doped uranium-zirconium hydride (UZrHx-Er) fuel (see Chapter 2).

The Oregon State TRIGA reactor is licensed to operate at a steady state power of 1.1 megawatts (MW) and can pulse to 2,500 MW with a peak steady-state thermal flux of about 1013 neutrons per square centimeter per second (n/cm2-s) in the B1 position. The reactor was originally fueled with a 70 percent enriched UZrHx fuel with a 1.6 weight percent erbium burnable absorber. The reactor was converted to a 19.75 percent enriched UZrHx fuel with a 1.1 weight percent erbium burnable absorber.

This reactor has a lifetime core, and it was important to the reactor operator to maintain a full grid of fuel assemblies in the converted core to maintain flexibility for conducting irradiation experiments. However, maintaining a full core reduced the shutdown margin (i. e., raised the excess reactivity) at the beginning of life of the new reactor core. Adjusting the erbium burnable poison to 1.1 percent in the converted core restored the shutdown margin and maintained the longevity of the core (Figure 2-4).

HFIR staff has developed a reference fuel design, but there is signifi­cant work remaining to evaluate its safety. HFIR’s conversion plan requires maintaining the fuel plates’ involute shapes, the overall core geometry, thermal/hydraulic system parameters, and key neutron fluxes, particularly in the flux trap region. To create a “proof of concept” reference LEU fuel design, state-of-the-art HEU-validated neutronics analyses were coupled to the original HFIR thermal design analysis. The current analysis uses the Nuclear Energy Agency’s Monte Carlo depletion interface code VESTA (MCNP/ORIGEN) that accounts for the zirconium interlayer between the UMo fuel and cladding (see Figure 2-1 in Chapter 2).

Current calculations indicate that essential neutron fluxes as well as fuel-cycle length can be preserved using UMo monolithic LEU fuel if the reactor power is increased from 85 to 100 MW. The fuel plates will need to be radially contoured (see Figure 3-6) and axially contoured on the lower edge to avoid flux peaking at the edges of the fuel.

Increasing the reactor power to 100 MW will require changes in the thermal/hydraulics safety basis, and new safety limits and protective sys­tem setpoints[70] must be derived from a revised thermal analysis. Transient analyses must also be reevaluated as well as fission product release, trans­port, and consequence analyses.

The return to 100 MW operation will also increase the heat flux from the fuel plates. However, ORNL plans to maintain the current primary coolant inlet temperature, flow rate, and pressure (pressure is constrained to 475 pounds per square inch atmospheric [psia] because of the embrittled vessel). Consequently, the safety limits and associated protection system setpoints will need to be changed, which will require the identification of additional safety margin, either through analysis or changes in fuel design. There are several resources that could be used to find the addi­tional required safety margin: (1) use a modern multidimensional physics analysis to evaluate the safety margin and demonstrate its adequacy; (2) revise the manufacturing uncertainties included in the safety analysis; and (3) revise the approach to the consideration of uncertainties (statistical versus multiplicative).

ORNL plans to begin the revised thermal/hydraulic analysis by per­forming a modern multidimensional analysis to analyze the safety margin at higher operating power. This analysis will use a COMSOL[71]-based, three-dimensional, detailed multiphysics LEU model to replace the existing HFIR steady-state heat transfer code. At present, ORNL staff is working to develop the integrated multiphysics modeling tools and place them into production. The new COMSOL model will require validation against the old (HEU) data, new (LEU) data, and separate effects testing, plus accep­tance by the regulator (the U. S. Department of Energy).

Two presentations on understanding and addressing the ageing and obsolescence of research reactors were given by Panel 2.2 speakers: H.-J. Roegler (an independent consultant from Germany, formerly with Sie­mens[40]) described an International Atomic Energy Agency (IAEA) initiative on research reactor ageing and ageing management (Roegler, 2011). E. P. Ryazantsev (Kurchatov Institute) provided a historical description of the research and test reactors at the Kurchatov Institute (Ryazantsev, 2011).

IAEA Initiative on Research Reactor Ageing and Ageing Management

H. — J. Roegler

The IAEA’s activities in ageing and ageing management for research re­actors began in the mid 1990s. In March 1995, the IAEA issued a TECDOC report (IAEA, 1995) on how to manage ageing in research reactors. Two months following the release of this report, the IAEA sponsored a confer­ence on research reactor ageing; the conference was held in Germany and involved more than 100 participants. In December 2008—more than a decade after publication of the TECDOC and sponsorship of the follow-up conference—the IAEA hosted an expert meeting at its Vienna headquarters to review the history of the agency’s efforts on ageing, including the ade­quacy of existing documentation, and to consider whether an initiative to collect additional information was warranted.

As the result of this expert meeting, the IAEA initiated the development of a database on research reactor ageing. This database is intended to ad­dress ageing as a technical and safety issue and explicitly excludes reactor conversion to LEU fuel. Information for the database was collected from research reactor operators using a standard template that was developed by the IAEA. The template permitted the reporting of a maximum of 3 ageing problems, classified by 13 possible ageing mechanisms in 76 reactor systems arranged in 9 groups. The template provided space for descriptions of ageing problems and actions taken to mitigate or cure them. A contact address for the reporting reactor was also required.

The templates were distributed in February 2009 to 133 research re­actor operators plus 28 other manufacturers and authorities. Responses from these organizations were incorporated into the database in October that same year. A total of 188 templates were initially submitted from 77 reactor facilities plus 6 other institutions (contributors were permitted to submit more than one template per facility). After review and revisions of the initial submissions, a total of 155 templates reporting on 367 ageing problems were included in the database.

There was a rather high-level of non-participation (43 percent) in this survey, which could have been caused by several factors, including language barriers, inexperience with completing these types of templates, or concern that the ageing problems might be publicly disclosed. One non-respondent justified the lack of participation as follows:

We do not have an ageing management program, because we do not have the funding for such a thing. We fix things when they break. That is un­fortunately the nature of our business here due to monetary constraints.

For me to fill out your template with something that is irrelevant is not worth your time, or ours. …We also do not necessarily wish to have this information be publicly available.

However, a convincing number of useful observations emerged from the template data that were submitted to the IAEA: [41]

Other frequently reported ageing problems included mechanical fatigue and wear and radiation-induced ageing (Figure 2-8).

• There were more ageing problems reported for younger reactors than for older reactors. This suggests that ageing problems begin with the initiation of operation of a research reactor.

Taken together, these data demonstrate the need for the future management of ageing in research reactors.

Although the database intentionally excluded information related to conversion, as noted previously, the information in the database is still potentially useful for conversion planning, because conversion needs to consider past as well as future ageing. The information in the database could be used, for example, to identify:

• Ageing systems and mechanisms to investigate

• Issues to discuss with the authorities

• Contacts for advice on addressing every type of ageing problems

The IAEA is planning to undertake a first update of this ageing da­tabase in August 2011. This will involve the reconfirmation of research reactor operator contacts, updates to the content of templates, and fresh approaches to the research reactor operators who did not provide informa­tion in 2009.

Robert Bari

Quantitative risk assessment has been used successfully to estimate safety risks, for example, at nuclear power plants. However, more research is needed before proliferation and terrorism risks can be effectively esti­mated using such a methodology. Such risk assessment methods are easier to apply to safety, for several reasons:

• The likelihood of an accident is more easily estimated than the like­lihood of a deliberate attack. A deliberate attack depends on the choices of an intelligent adversary, making likelihoods and methods of failure difficult to estimate.

• Inherent features and engineered systems with known character­istics provide safety, whereas both intrinsic (i. e., barriers intrinsic to the technologies themselves) and extrinsic (e. g., guns, guards, gates, safeguards) systems provide security. The effectiveness of some extrinsic measures, par­ticularly those that involve human action, can be difficult to estimate.

• For safety, defense in depth and safety margins are universally embraced.

Workable proliferation risk models still need significant development.

The methodology summarized here is one of several possible ap­proaches and is analogous to the approach developed for the Generation IV Forum: Proliferation Risk and Physical Protection (PR&PP).

To perform an effective risk assessment, it is important to gather a great deal of information about the research reactor facility as well as the country in which it is located. There are many countries with research reactors, each having its own national and geopolitical interests that could impact the potential for proliferation. In addition, a number of key assumptions need to be considered in the analysis. These include assumptions about potential threats, such as diversion, misuse, breakout, theft, and sabotage; extrinsic factors such as sources of fresh fuel supply, spent fuel disposition, and fuel transportation; and facility design and operational information that impact proliferation risk.[79]

The assessment itself involves building a range of scenarios by which proliferation could occur; analyzing specific scenarios to determine whether an attempted proliferation was successful and the barriers that were en­countered along the way; then using the responses to construct a risk estimate.

The key elements of an effective proliferation risk assessment include:

• Gather information on facility design.

• Define country (or countries) context.

• Establish/define international safeguards design.

• Establish/define physical protection design.

• Define adversary mission success.

• Identify facility targets (for adversary).

• Perform pathway analysis to define potential scenarios for proliferation.

• Evaluate pathways for each threat and measure.

• Assess and interpret results.

Further research will be needed before this type of analysis can be car­ried out in a dependable way for research reactors. The range of possible scenarios has not been explored in much detail. In addition, combining the information produced by each stage of the analysis described above to pro­duce an overall understanding of risk remains challenging. However, such a risk assessment process can still be worthwhile to perform. In particular, the process itself can provide useful insights, not just the final result.

Discussion

Many measures that can be taken to reduce the risk of proliferation from research reactors are already well known. Some measures mentioned by symposium attendees included avoiding the use of HEU fuel where possible in favor of LEU fuel; maintaining adequate nuclear materials protection control and accountability (MPC&A) and physical protection measures; and using appropriate insider prevention methods, as practical.

Many of the participants at the symposium observed that the principal means of reducing proliferation risk is conversion of research reactors to LEU; however, as noted previously, this may not be possible in all cases. Other participants noted that some risk also accompanies the use of LEU. Consequently, appropriate MPC&A and physical protection measures will continue to be needed, although to a lesser extent than with HEU fuel.

A symposium participant posed a question about the relative priorities between conversion to LEU and better physical protection. In particular, is it possible to compensate for HEU use through improved security? Robert Bari’s reply was that one cannot separate conversion from physical protec­tion. Clearly, maintaining HEU fuel poses a greater risk, but LEU use does not mean zero risk. Richard Meserve clarified that at a gross overview level, conversion lowers risks as well as the costs for physical protection.