Working Party on Pu-MOX fuel physics and innovative fuel cycles (WPPR)

Transcription

1 R&D Needs in Nuclear Science 6-8th November, 2002 OECD/NEA, Paris Working Party on Pu-MOX fuel physics and innovative fuel cycles (WPPR) Hideki Takano Japan Atomic Energy Research Institute, Japan

2 Introduction(1) NSC has been addressing issues related to plutonium utilization in LWRs and FBRs. Former days than 1992 For FBRs, HCLWR, FRs, much works of benchmark calculations has been conducted to validate the data and methods. As a result, nuclear codes and data is comparable to standard LWR benchmark calculations. Especially, for HCLWR application, computational codes based on ultra-fine group method which calculates accurately resonance shielding including intermediate neutron spectrum region were developed, and furthermore continuous energy Monte Carlo calculations were proposed as a reference solution. After this, the methodology is used as standard calculation for various benchmark calculations.

3 Introduction(2) In June 1992, the NSC set up a Working Party on Physics of Plutonium Recycling and in 1997 has started Working Party on Physics of Plutonium Recycling and Innovative Fuel Cycles (WPPR). In the Working Party, benchmark studies cover issues and perspectives, Pu-recycling in PWR, void reactivity effect in PWR, fast Puburner reactors, Pu-recycling in fast reactor, multiple Purecycling in standard and advanced PWR, MOX in BWR, and benchmark calculations of MOX fuel lattice experiments in LWR.

4 Plutonium-recycling in PWR(1) Three benchmarks were devised for MOX in PWRs. Two simple infinite pin cell problems investigated issues related to MOX utilisation in PWRs with plutonium, both of typical and poor isotopic quality. The typical type is a plutonium isotopic vector with higher fissile fraction using commercial PWR MOX. The latter type is a plutonium of low isotopic vector expected for a self-generation recycling mode in PWR MOX. The benchmarks were to investigate whether nuclear data and lattice codes require further development and validation to calculate the plutonium fuel core physics performance.

5 Plutonium-recycling in PWR(2) WPPR expected that good agreement between the various solutions would be obtained. In both benchmarks, the spread in k infinity was 3-5 % including the solutions of commercial established codes verified for uranium fuels, and if the solutions are not included, the spread decrease, but not less than 1%. Furthermore, this spread in k infinity increase with the plutonium contents. This means a need for improvement in both methods and basic data for higher plutonium isotopes and minor actinides. The calculational methods have to take into account resonance selfand mutual shielding treatment, especially including self-shielding treatment of Pu-242.

6 Plutonium-recycling in PWR(3) The third benchmark was a void reactivity effect in PWR. A supercell configuration of a 30x30 array of PWR UO2 and MOX fuel cells was carried out in the VENUS experimental reactor at MOL. The calculated results shows that the infinite lattice calculations gives a non-negligible spread of the results, but the void effect is positive for high content of MOX in the fuel. A similar tendency is derived from the results of the twodimensional calculations for an uranium assembly with a central moderated or voided region fuelled with MOX pins of different plutonium content.

7 Plutonium-recycling in PWR(4) As a result, the design of MOX fuel assemblies needs the use of the newest nuclear data bases and very detailed and sophisticated spectra and assembly codes. The use of codes only verified for UO2 fuel should be avoided. In the void reactivity effect benchmark, experimental validation would also be needed, in particular for reactivity coefficients, in the case of degraded plutonium isotopic composition. In recycling of plutonium in PWRs the important related physics issues would be the plutonium content limitation to avoid positive void effects and the minimisation of minor actinide production.

8 Plutonium-recycling in PWR(5) The summary conclusions from the PWR benchmark studies are as follows: The nuclear calculational methods have to take into account rigorous resonance self shielding, mutual shielding over the whole energy region for the fuel and cladding nuclides, and appropriate and well tested calculation methods should be used. Sufficient quality of basic nuclear data is needed for U-238 and the Pu isotopes, but also for higher acitinides and the major fission products, and modern nuclear data libraries such as ENDF/B-VI, JEF-2 and JENDL-3 are essential. Experimental verification related to maximum plutonium content in the case of degraded plutonium isotopic composition is needed in clean lattice configurations with different moderator-to-fuel ratios.

9 Pu-recycling in Fast Reactor (1) Two topics of benchmarks examined for fast reactor systems are the physics of plutonium recycle in both oxide and metal fuelled fast reactors to determine the potential of the fast reactors for consuming plutonium and fissioning minor actinides. In this benchmarks, the physics related to in particular consuming plutonium rather than establishing a breeding cycle and reducing the source of potential radio-toxicity has been widely investigated. Such fast reactor systems may have an important role in managing plutonium stocks until fast breeder reactors are needed.

10 Pu-recycling in Fast Reactor (2) The benchmark excises were considered to cases with non-breeding fast system with conversion ratios in the range 0.5 to 1.0. The fast burner reactors / LWR symbiosis was investigated. Results indicated a potential for significant nuclear waste toxicity reduction. In the metal fuelled FR benchmarks, high leakage cores, higher content of minor plutonium isotopes and higher actinide isotopes all need further validation work, including critical experiment performance.

11 Recycling of Plutonium in Advanced Converter Reactor As advanced converter reactors, the Japanese ATR, variant of the CANDU, the High conversion LWRs (HCLWR) were reviewed. The advanced converter reactors have a role as an intermediate step between today s thermal reactors and future fast breeder systems. For HCLWRs, in some lattice experiments such as LWHCR of PSI, ERASME/S (moderator ratio 0.5), ERASME/R (moderator ratio 0.9), the comparison of measured and calculated lattice parameters appear to be well knowledge of the neutronics and physics of such reactors.

12 Multiple Plutonium-Recycling in PWRs (1) A shortcoming of the first Pu-recycling benchmarks was that only one of the cases considered corresponded to plutonium of good isotopic quality, and to the scenario which might arise after many generations of MOX recycling (with extremely poor isotopic quality). Analysis of the intermediate steps was missed. CEA suggested a benchmark in which five consecutive generations of multiple recycling in a PWR would be followed. In the specification of the benchmark, attempts were made to make it as realistic scenario as possible, taking into account such details as the time delays in pond storage, MOX fabrication etc. and the dilution effect when MOX and UO 2 assemblies are co-reprocessed.

13 Multiple Plutonium-Recycling in PWRs (2) As in the previous benchmark exercise, the benchmark was restricted to the level of the lattice codes. This is the logical first step until the underlying nuclear data and lattice code calculations show adequate agreement. Two cases were considered; one for a standard 17x17 PWR lattice such as used in many of today's PWRs (designated the STD-PWR) and another lattice with an increased moderator to fuel ratio (3.5:1 compared with 2:1 for the STD-PWR). The latter case, designated the HM-PWR (for highly moderated), was intended to cover a proposed PWR design for use with MOX only.

14 Multiple Plutonium-Recycling in PWRs(3) The burnup reactivity, reactivity coefficients, microscopic crosssections, isotopic evolution and isotopic toxicity evolution with time were compared. The broad conclusions and observations are as follows: Since the earlier benchmarks, considerable progress has been made in nuclear data libraries and methods. The discrepancies between the different data libraries and lattice methods are now generally within reasonable bounds. The observed spread of results is now consistent with the uncertainties in the underlying nuclear data. Multiple recycle scenarios therefore appear to be practicable and feasible in conventional PWRs, at least in the near term. Questions are a possible positive void coefficient would almost certainly preclude recycle beyond the second generation.

15 Multiple Plutonium-Recycling in PWRs(4) The HM-PWR degrades the plutonium isotopic quality more rapidly than the STD-PWR, and this negates the benefit of the softer spectrum. The HM-PWR also seems to pose more difficulties for present nuclear data libraries and codes, as evidenced by the larger number of discrepant results seen in the HM-PWR benchmark. Therefore, even the HM-PWR is of questionable practicability with respect to later recycle generations. In view of these considerations, the WPPR agreed that there is no compelling reason to continue further benchmark studies at the level of the lattice codes.

16 Power Distribution Within MOX Fueled Assemblies (1) The main objective of the benchmark was to compare different techniques of fine flux assessment derived from coarse mesh diffusion calculations or transport calculations. Ten institutions contributed and more than 15 calculation schemes were examined including the majority of the methods used for reactor design: collision probability, Sn transport (finite difference and nodal), diffusion(finite difference and nodal), Monte Carlo, power reconstruction methods, etc.

17 Power Distribution Within MOX Fueled Assemblies(2) Larger discrepancies appear at U-MOX interfaces and it was pointed out that discrepancies may also appear in non-symmetric configurations. In conclusion, fine flux reconstruction can be achieved to a satisfactory precision except for local singularities. The second phase benchmark exercise has been undertaken for power distribution within MOX fueled assemblies. Concerning core calculations, a big spread of 860 pcm from the average value of keff was observed. This is due to cross section libraries and method used. As for the pin-by-pin power distributions, the average of participants overestimates the power in the central MOX assembly with a spread of about 3 % for each zone.

18 BWR MOX Benchmark The previous work of the WPPR addressed fast reactors and PWRs and did not consider BWRs. As BWRs were intended for MOX burning by several member countries, the WPPR felt that it was important to address the imbalance in its activities. The specified benchmark compares calculations for a modern 10X10 BWR design with assemblies containing MOX rods and U/Gd rods with a large internal water structure. The fissile content allows for an average discharge burn-up of > 50 GWd/t and the isotopic composition of the plutonium corresponds to Pu from burnt UO 2 fuel with discharge burn-ups of > 50 Gwd/t.

19 BWR MOX benchmark The results from participants showed that a similar accuracy is achieved for BWR as for the PWR benchmarks. For BOL (beginning of life), an agreement within 1% for k-infinity is observed, and a difference slightly more than 1 % is found when comparing ENDF/B-VI against JEF-2.2. For burn up cases, the agreement is better than that for BOL cases.

20 Innovative Plutonium fuel without uranium(1) For ex-weapons and civilian plutonium, the plutonium burning in reactors has been emphasized as a means of minimising the risks of proliferation. From this perspective, the most effective way is adopting plutonium without uranium. By using inert carriers for plutonium it is possible to avoid the production of fresh Pu-239 from U-238 captures. A specific task group established by PSI, JAERI, CEA and Politecnico Milano summarised the present status and major issues associated with R&D of plutonium fuels with non- uranic carrier, concentrating in particular on the physics.

21 Innovative Plutonium fuel without uranium (2) Some benchmarks have been carried out, in case of partially loaded IMF core in a PWR, the agreement among the solutions provided are well for UO2, but discrepancies are larger for MOX fuel. Further problem was encountered with the kinetic parameters in which the discrepancies are observed. Inert Matrix Fuel Work Shop is held every year since 1995, and research and development is under way to find potential fuel candidates.

22 On-Going Works In the 5-th WPPR meeting on November 1996, future works were discussed and the need for experimental analysis for MOX fuel critical experiments in LWRs was proposed in order to validate the calculations against the experiments as essential progression. Participants agreed and several experiments were candidated such as TCA (JAERI), VENUS-2(SCK/CEN), EOLE (Cadarache), and so on. These data sets are not available immediately, and would require the agreement of the original sponsors to release part or all of the data.

23 VENUS-2 The VENUS-2 programme carried out by SCK/CEN. This core consists of 3.3% and 4.0 % enriched UO2 rods and 2.7% Pu content MOX rods. The comparison of calculations and experiments are conducted for keff and pin-power distributions. The first benchmark is 2-D configuration model, and the second phase is 3-D model. Participants were adopted the blind benchmark procedure which the experimental values are not given to the participants. Thirteen solutions were submitted and the agreement is generally satisfactory(<500ppm reactivity, 5 percent on power in the UO2 regions and 10 percent on power in the MOX region). The higher spread of pin powers in the MOX region is partly a result of the MOX rods being positioned at the core periphery.

24 Benchmarks proposed presently(1) HTR Benchmark The WPPR address the issue of pebble bed reactors to avoid overlap for the IAEA prism type benchmark. The PBMR concept is more interesting for physics studies than the prismatic type. Proposals for HTR plutonium physics benchmark are presented by M.Delpech (CEA) and H.I, Rutten (Framatom). Furthermore, U. Kasemeyer presents the analysis of experiment carried out on HTR at the PROTEUS facility. The experimental data are of interest, but the PROTEUS geometry is complicated and difficult to model. So it is a candidate for work in a second phase study.

25 Benchmarks proposed presently(2) MOX core transient benchmark A benchmark exercise on Kinetic Parameters via Calculation/Experiment Comparison of Reactivity Effect in CROCUS is proposed. It is a 3-D problem and involve 4 water heights and two control rods, six cases in all. The first phase is the UO2 fuel benchmark, and a similar exercise on MOX will follow as a further step. In fact, large discrepancies due to data libraries have been found at PSI(20% in some cases). This benchmark was endorsed by the WPPR.

26 Benchmarks proposed presently(3) Benchmarks co-operated with TFRPD (The Expert Group on reactor-based Pu Disposition) Three benchmarks are co-operated with TFRPD, that is, 3-D VENUS-2 MOX Benchmark, which specifications have been prepared as Blind Benchmark on The 3-D VENUS-2 MOX Core Measurements at May 2001, KRITZ-2 Benchmarks consist of three low enriched uranium fuel cores and one MOX fuel core, and VVER-1000 MOX Benchmark proposed by KI. Furthermore, a study related to plutonium management in the medium term (i.e. beyond LWR MOX, but prior to fast reactors) had been addressed.