Two presentations on fuel design for conversion were given by Panel 2.1 speakers: Daniel Wachs (Idaho National Laboratory) reported on efforts to develop LEU fuels for converting U. S.-origin reactors (Wachs, 2011),

and Yu. S. Cherepnin (Dollezhal Scientific Research and Design Institute of Energy Technologies [NIKIET]) described progress and prospects for reduction of fuel enrichment in Russian-origin reactors (Cherepnin, 2011).

Fuel Design for U. S.-Origin Reactors

Daniel Wachs

Highly enriched uranium (HEU) fuel elements in U. S.-origin research and test reactors consist of aluminum-clad plates (see Chapter 1) that contain a UAlx or U3O8-aluminum dispersion fuel meat clad in aluminum or a uranium-zirconium hydride (UZrHx) fuel meat clad in stainless steel (TRIGA fuel). Work carried out by Argonne National Laboratory and the Idaho National Laboratory, in cooperation with other American, European, and Korean organizations, has resulted in the development of three LEU dispersion fuel systems1 for conversion of plate-type reactors:

• UAlx (density = 2.3 grams of uranium per cubic centimeter [gU/cm[26]])

• U3O8 (3.2 gU/cm3)

• U3Si2 (4.8 gU/cm3)

These fuel systems are adequate for converting all but “high per­formance” research and test reactors.[27] [28] There are six HEU-fueled high — performance research reactors in the United States3 as well as four HEU-fueled high-performance research reactors in Europe that cannot be converted with these existing LEU fuel systems. The U. S. reactors are shown in Table 1-1 in Chapter 1; the European reactors are the following:

• Belgian Reactor 2 (BR2) at the Belgian Nuclear Research Centre in Mol, Belgium

• Forschungsreaktor Munchen II (FRM-II) at the Technical Univer­sity of Munich, Germany

• Jules Horowitz Reactor (JHR), under construction at the CEA Cadarache Research Centre in Cadarache, France (discussed in Chapter 4)

• Reacteur a Haut Flux (RHF) at the Institut Max von Laue-Paul Langevin (ILL) in Grenoble, France

Higher-density LEU fuel systems based on uranium-molybdenum (UMo) alloys are now under development for use in converting these U. S. and European reactors. Test irradiations have been carried out on several UMo alloys to assess their suitability for use as fuel for these reactors. Testing revealed that alloy phases with high U/Mo ratios (e. g., U-10Mo[29]) were most stable under irradiation because they suppressed the formation of fission gas bubbles.[30]

Two LEU fuel systems based on this alloy are now under development by Idaho National Laboratory and partners:

• UMo dispersion fuel: A UMo alloy dispersed in an aluminum ma­trix with uranium densities up to 8.5 gU/cm3. An LEU fuel system based on this material is being developed for conversion of BR2, RHF, and JHR.[31]

• Monolithic UMo fuel: Metallic UMo foils with a uranium density of 15.5 gU/cm3. An LEU fuel system based on this material is being devel­oped for conversion of ATR, HFIR, NBSR, MITR, and MURR (Figure 2-1).

Test irradiations of fuel elements containing both of these materials are now being carried out to develop and qualify these fuel systems.

A major challenge for MITR is to convert to LEU while still meeting the performance requirements for experiments in the reactor. Meeting these re­quirements will entail optimizing the fuel design to maximize heat transfer

and neutron flux. In particular, the neutron flux optimization is focused on maintaining HEU-equivalent fast neutron fluxes in in-core materials experi­ments and thermal neutron fluxes in out-of-core experiments. To prepare for conversion, the existing neutronics and thermal/hydraulics models for the MITR core were improved in several ways.

Several major improvements were made to the neutronics codes. The primary change enabled more accurate burnup modeling and benchmark­ing. The improvements entailed an extensive review of the model’s core structure and dimensions as well as an update of the cross-section li­braries, homogenized volume fractions, and discrete structures. Two ini­tial HEU cores were modeled, and the results compared favorably with measurements.

The neutronics codes were improved in other ways as well. A graphi­cal user interface was added, as was the capability to model HEU, LEU, and mixed core geometries. In addition, it is now possible to model all fuel movements, including flipping, rotating, and fuel storage. The models are also now able to track and plot the mass of isotopes as well as the power distribution in the core.

The improved burnup modeling has shown good results. Twelve recent cores have been modeled; the results were in good agreement with measured beginning — and end-of-cycle control blade positions. There was also good agreement among different models.

In addition to the neutronics modeling, thermal/hydraulics models were also updated and modified. Specifically, the models were modified to include the fuel’s longitudinal fins for the steady-state and loss-of-flow analyses. Initial results have shown that the LEU design core has a higher margin to onset of nucleate boiling and a lower peak cladding temperature with loss of flow.

The results of the neutronic and thermal-hydraulic analyses have been used to design an LEU fuel for this reactor. The LEU fuel assembly will contain more plates and use a thinner fuel meat (0.51 mm for UMo LEU fuel versus 0.76 mm for HEU fuel) and cladding (0.28 mm for UMo LEU fuel versus 0.35 mm for HEU fuel). Fuel developers have informed MITR staff that fuel and cladding of this thickness is feasible to manufacture.

As was noted in Chapter 2, the reactor’s operational power will need to be increased from 6 MW to 7 MW to counter the expected loss of per­formance after conversion. This will result in an increase in the cycle length from 40-50 days for the HEU fuel to 60-70 days for the LEU fuel.

The IVV-2M is a high-flux, 15 MW pool-type reactor with a hexagonal core containing 42 hexagonal fuel elements enriched in uranium-235 to 90 percent. Initial studies have been carried out to examine the feasibility of converting this reactor to 19.75 percent enriched fuel containing a UO2- aluminum dispersion fuel meat.

This reactor is being very effectively operated at present and has a high level of utilization, so any significant loss of consumer characteristics fol­lowing conversion would be problematic. Initial conversion studies have focused on identifying a fuel type that would meet consumer needs. Ana­lytical studies have examined the reactor characteristics that would result from conversion to dispersion fuels having uranium densities of 3.5 and 6.5 gU/cm3 as well as a UMo-aluminum dispersion fuel.

Conversion to a 3.5 gU/cm3 fuel that was manufactured using existing (extrusion-based) fuel fabrication technologies would result in insufficient reactivity reserve and the deterioration of other consumer characteristics such as burnup. Conversion using 6.0-6.5 gU/cm3 fuel would improve the feasibility of conversion if fuel elements with such material were able to be manufactured economically.

A. Zrodnikov

Research reactors in the United States and Russia serve a variety of in­dustrial and biomedical missions and enable research in fields such as phys­ics and nuclear engineering. Missions mentioned during the course of the symposium that seem likely to continue include silicon doping, radioisotope production, notably including molybdenum-99, and neutron therapy. It is essential to maintain the capability to meet these research and industrial needs. Other means (e. g., particle accelerators) may be developed in the future for generating some radioisotopes and producing neutron beams, but research reactors will be far more difficult to replace for some other applications. In particular, future research related to nuclear energy and the nuclear fuel cycle will necessitate maintaining and improving current research reactor capabilities in the United States and Russia as well as in other countries. Research reactors are especially needed to conduct basic research for nuclear power development.

Nuclear power generation faces major challenges in the coming de­cades. Increasing quantities of commercial spent nuclear fuel are being ac­cumulated around the world, and in the long-term, supplies of uranium-235 will begin to decrease. Fast neutron reactors (“fast reactors”) are being studied in the United States and in Russia for their potential to help meet these challenges. Such reactors have the potential to “burn” long-lived actinides in spent fuel and also to produce and operate using plutonium, thereby extending current fuel supplies. However, more research remains to be done on these topics to effectively design the needed facilities and processes.

Beyond the design and testing of future fast reactors, further research could also help to extend the capability of nuclear power plants to meet new tasks. For example, research on heat — and radiation-resistant materials could lead to the deployment of high-temperature nuclear plants to meet the needs of heat-intensive industrial processes, including water desalination, production of synthetic fuels, and hydrogen production. If fossil resources that currently fuel these processes are exhausted, nuclear power will be needed to fill the gap.

Several research problems related to these topics will need to be investi­gated in the coming decades, including improving the scientific understand­ing of: [77] [78]

3. Changes in macroscopic material properties caused by neutron and charged particle irradiation.

Research reactors will also be used in theoretical, computational, and experimental studies on thermo-physical, physical-chemical, corrosion, and physical-mechanical properties of advanced high-temperature coolants, fuel materials, and core structural materials. Moreover, data generated from such studies will help researchers to develop complete nuclear data libraries. This knowledge can be used to develop new nuclear technologies.

Much of the research work involving fast reactors may require capabili­ties that only a few current research reactors possess. A research reactor with a stationary steady-state fast neutron flux of about 1016 neutrons/ cm2-s will be required to support this research.

In the subsequent discussion, Thomas Newton (Massachusetts Institute of Technology [MIT]) agreed that this need for fast neutrons was also true at MIT, and observed that, after conversion, MIT plans to take advantage of the harder neutron spectrum that can be acquired with low enriched uranium (LEU) for fast neutron experiments.

G. Pshakin

The BFS-1 and BFS-2 critical assemblies[50] at the Institute for Physics and Power Engineering in Obninsk (Figure 2-11) provide a good example of reactors that cannot be converted to LEU fuel. These reactors, which are fueled with HEU and plutonium, were constructed in the late 1950s and early 1960s as part of the Soviet Union’s fast breeder program for nuclear energy development. Although these assemblies cannot be used for design­ing commercial-scale fast breeder reactors, they are useful for simulating fast breeder reactor cores, for fuel cycle research, and for studying the transmutation of minor actinides. This fuel used in these assemblies is not self-protecting[51] and therefore poses special security concerns.

Converting these facilities to LEU fuel cannot be accomplished without sacrificing the current mission. Moreover, even if the uranium enrichment of the fuel could be reduced, plutonium would still be required to simulate the cores of fast breeder reactors.

There are two options for addressing the security concerns associated with these facilities: (1) shut down the facility and remove all nuclear ma­terials; or (2) organize a state-of-the-art materials protection, control and accounting (MPC&A) system and enhance the culture of personnel through proper training, motivation, and support. The second option is obviously preferable.

The facility has cooperated with the United States to develop an MPC&A system. It includes a non-destructive analytical system based on high-resolution germanium detectors for isotopic measurement of ac­counted items; neutron coincident counters for nuclear material mass mea­surements; and specially designed access and monitoring systems. This program has to protect more than 100,000 HEU and plutonium discs that are used to model the cores of fast breeder reactors.

Initial irradiations of fuel elements containing UMo dispersions re­sulted in the formation of interaction layers between the UMo and Al particles and the development of porosity and distortion (pillowing). The addition of small amounts (~2 percent) silicon to the aluminum phase was

found to suppress the development of this interaction layer at burnups of up to 70 percent. However, test irradiations of this fuel material at high power (~ 500 watts per square centimeter [W/cm2]), high uranium loadings (> 8 gU/cm3), and high burnup (> 70 percent) resulted in the formation of small blisters on the fuel plates. Follow-up experiments are planned for the fall of 2011 to determine why such blistering occurs and how the fuel ele­ment can be modified to eliminate it. A bounding-case irradiation of this fuel material in BR2 is planned for 2011-2012.

Prior to beginning the conversion analysis, some safety analysis param­eters at MITR were not well known, particularly for the finned cladding.

To complete the safety analysis, further information is being gathered on the following three topics:

• Finned channel hydraulic pressure drop. A flow experiment has been built, measurements have been made, and a finned channel correla­tion[65] describing the relationship between the pressure drop and the flow rate has been developed.

• Adequacy of the onset of nucleate boiling (ONB) correlation for finned channels.[66] For this purpose, the MITR boiling flow loop is being constructed to measure ONB for the LEU channel geometry and validate the Bergles-Rohsenow ONB correlation[67] [68] [69] for finned channels. This facility will be operational later this year.

• Oxide distribution in the finned cladding. The current burnup limit of 1.7 x 1021 fissions per cubic centimeter is based on an even 50 microm­eter-thick aluminum oxide distribution on the cladding. However, particu­larly within the finned region, the actual oxide distribution is unknown. MITR is currently using an eddy current probe for fin-tip measurements of oxide thickness. This thickness will then be compared with the operational history of the fuel element. Finally, a selected fuel element will be shipped to Idaho National Laboratory for evaluation of the oxide distribution in the areas between the fins.

The WWR-M is an 18 MW pool-type reactor with a hexagonal core containing 145 fuel elements of WWR-5M design that are enriched in uranium-235 to 90 percent (Figure 2-6). Although this reactor entered ser­vice in 1959, it is still operating very effectively and has achieved several increases in power and flux densities since it was commissioned. It is now the highest power reactor of its type in existence.

Initial studies have been carried out to assess the feasibility of convert­ing this reactor to 36 percent enriched and 19.75 percent enriched fuel. It was observed that as enrichments decrease, burnups, thermal neutron flux densities, and fast neutron flux densities also decrease. These studies indi­cate that fuel having a uranium density of 8.25 gU/cm3 would be required to convert this reactor without sacrificing needed consumer characteristics. However, fuels with this density are not available at present.

R. P. Kuatbekov

Current research reactors are unlikely to meet all needed missions over the next few decades. Many of the currently operating research reac­tors are ageing, and many missions are projected to grow in importance. Consequently, there is a need to design and build new research reactors. In many cases, particularly for industrial applications, new reactors can be de­signed from the beginning to use LEU rather than HEU fuel. In other cases, particularly if HEU or even plutonium fuel is required to retain essential performance characteristics, alternative solutions may need to be found to meet nonproliferation goals.

The customers of research reactors do not care whether the reactor is fueled by HEU or LEU—they simply need the results within a reasonable period of time and at reasonable cost. This is true whether the results are completed research, produced materials, or medical isotopes. Consequently, two key qualifications for any new research reactor will be: (1) its ability to meet customer needs; and (2) economic and technical feasibility. With respect to economics, both initial costs and refueling costs of the reactor should be considered to be reasonable by the operator.

In addition, not all countries can afford to perform experiments to op­timize fuel for their research reactors, as the United States and Russia have done, and these countries are likely to be a major market for certain types of research reactors in the coming decades. For these reasons, NIKIET is using reliable and tested fuel types and design solutions in its new research reactor designs. At the same time, proliferation concerns will need to be accounted for.

NIKIET is in the process of designing several new types of LEU-fueled research reactors for industrial, biomedical, training, and research applica­tions. The focus is on the development of pool-type reactors with integrated passive safety systems. Pool-type reactors are convenient for the end-user because they allow for flexibility in the core configuration and easy access to experimental positions. NIKIET uses standardized components in its reactor designs, which reduces costs and simplifies future repairs.

NIKIET is focusing on narrow-purpose reactor designs that optimize each reactor for the customer’s primary end use. There are two end uses that are in highest demand, both mentioned elsewhere in this report: medical isotope production and silicon doping. NIKIET is focusing on optimizing designs of two reactor types for these applications: (1) a low-power (500 kW or less) reactor with natural circulation for silicon doping and (2) a 15 MW reactor for isotope production. Some preliminary computations have been carried out on these reactor designs, and NIKIET plans to improve these designs in the future with additional computations and design work.

It is feasible to meet most customer needs using LEU-based research reactors. Designing reactors to use LEU from the start will not be as much of a challenge as retrofitting some current HEU-fueled reactors. In fact, modifying research reactor cores that were originally designed to use HEU can be very expensive and technically challenging, as illustrated by the case studies in Chapter 3. If the core is optimized during the design stage, then it can simultaneously be optimized for its missions. For example, if the core is initially designed to use LEU fuel, then differences in neutron fluxes and spectra can be accounted for from the initial design stages.

In fact, NIKIET has found that several of its designs for new LEU reac­tors maintain high flux levels to meet customer requirements and achieve reliable operation with high fuel burnups. NIKIET has now proven com­putationally that these LEU reactors should operate as well as similar HEU reactors.

On the other hand, some cutting-edge research requires reactors with unique designs or higher fast or thermal neutron fluxes. This research may not be able to be carried out using the types of standardized designs de­scribed here. In many cases, unique LEU-fueled reactors can be designed from the start to meet these needs; however, it was suggested by some at the symposium that maintaining a small number of special-purpose reac­tors fueled by HEU or plutonium could have value, particularly for fast reactor research.

Time was set aside during this session for free discussion among sym­posium participants. Some of the key comments from that discussion are presented in this section. [52]

• Current work under way in Russia on monolithic fuel development could pave the way for conversion of many Russian research reactors. Jim

Matos commented that the densities of the LEU dispersion fuels described in the Russian presentations are too low to be used in converting many Russian reactors. Jim Snelgrove noted that monolithic pin-type LEU fuel is also being tested in Russia. This fuel is a potential replacement for the tube — type fuel that is now being used in Russian research reactors. The recent agreement between DOE and Rosatom to assess the feasibility of converting six Russian research reactors could play an important role in assessing the potential utility of this LEU fuel.

• There may be some research reactors that cannot be converted.

V. Ivanov noted that there may be some reactors with unique purposes that cannot be converted. For example, the multipurpose fast breeder reactor to be built in Dimitrovgrad will be fueled with HEU and plutonium. The concept of reducing risk by eliminating HEU does not make sense for this reactor because the HEU is used alongside plutonium. This is also true for critical assemblies. He also noted that the concept of “unique mission” has not yet been defined in Russia, and he suggested that there should be a limited list of parameters that could be applied to determine uniqueness. N. V. Arkhangelsky reminded symposium participants that it was recognized from the very beginning of the RERTR program that there are a number of research reactors that would not lend themselves to conversion, including fast breeders.

• Reactor ageing is a potential complication for conversion, but it can be managed. V. Ivanov noted that unless national regulatory require­ments dictate conversion, the decision to convert, upgrade, or shut down a reactor will be made by the operator/owner. The owner/operator must determine whether it makes sense to convert the reactor if the remaining lifetime is negligible. H.-J. Roegler commented that, in his experience, re­search reactor ageing problems can be cured, although in some cases it can take time. A. N. Chebeskov commented that different reactor facilities may have access to different resources to manage ageing. Having a set of best practices to manage ageing could be a topic for international cooperation.

• Reactor customers (users) are an important part of the conversion process. V. Ivanov commented that conversion work needs to be trans­parent to customers, not just designers and research reactor specialists. He suggested that it would make sense for the international community, including the customers of research reactors, to cooperate more closely on conversion.

• There may be economic advantages to conversion. Richard Meserve noted that conversion may have economic advantages that were not dis­cussed by any of the symposium presenters. In particular, LEU costs could be lower, depending on how that material is priced, and costs for securing

LEU fuel should be much lower than for HEU fuel. Jordi Roglans com­mented that transportation costs, especially international transportation costs, will be lower for LEU fuel because HEU is often transported by the military.

• A worldwide ethic on conversion should be developed. Yu. S.

Cherepnin suggested that the world community should develop a new ethic against operating reactors with HEU. Strong signals should be sent to operators of HEU reactors that they need to convert, and funding should be demanded from governments to support conversion.

• Working together, the Russian Federation and the United States have played and will continue to play important global roles in research reactor conversion. N. Laverov noted that the Russian Federation has decommissioned 200 nuclear submarines and, working with the United States, has returned 100,000 tonnes of natural uranium and 500 tonnes of HEU from foreign countries. The recent agreement between DOE and Rosatom to assess the feasibility of converting six Russian research reac­tors is an important step for eliminating HEU use in Russian research reactors. It is important that the Russian Federation and the United States serve as an example to countries by reducing the enrichments of their research reactors to lower levels.