contents

Benchmark experiments in the VENUS critical facility

Advances in code validation for mixed-oxide fuel use in light-water reactors

Pierre D'hondt

A typical commercial thermal reactor produces around 250 kilograms of plutonium per gigawatt-year electric output. The majority of the spent fuel discharged from the world's 430 or so commercial nuclear reactors has been designated for interim storage and eventual direct disposal. Those utilities that have opted for the so-called "once-through"-or open-fuel cycle have chosen this option after considering the particular circumstances (political, economic, strategic, logistic, historic, etc.) that apply locally. With different local circumstances, other utilities have chosen to reprocess their spent fuel.

Current commercial reprocessing plants are all designed to separate the plutonium remaining at discharge for reuse. Historically, the rationale was to be able to recover sufficient plutonium to enable a buildup of fast reactors, which were expected to be deployed as uranium reserves became scarce and prices rose. For a variety of reasons, but principally the low price of uranium ore, fast reactors have not yet been deployed commercially and projected timescales for doing so have been put back everywhere.

There are several technical options for plutonium management involving reuse in existing light-water reactors (LWRs). Partial core loading of mixed-oxide (MOX) fuel assemblies in pressurized water reactors (PWRs) is already well established on a commercial basis, with thirty-seven reactors in Europe (two in Belgium, 22 in France, 10 in Germany, and three in Switzerland) currently operating with part-MOX loading. There is less experience in MOX usage in boiling-water reactors (BWRs), with just two BWRs in Germany currently using MOX fuel.

Partial MOX loading in a PWR involves the substitution of a fraction of the uranium dioxide (UO2) assemblies with MOX assemblies of the same mechanical design. This avoids issues of thermal-hydraulic and mechanical-handling incompatibility. The plutonium concentration in the MOX assemblies is usually adjusted so that the reactivity lifetime of the MOX fuel coincides with that of the UO2 fuel, in which case the average discharge burnups of the two fuel types will be the same.

The presence of MOX fuel batches in a LWR affects the nuclear design characteristics of the core in a complicated fashion. Consequently, a need was seen in the late 1980s to develop or to improve the in-core management codes and to benchmark and validate them against experimental data. An insufficient validation might induce a dramatic increase of the uncertainty factor with a possible reduction of the reactor power required to keep reactor operations within very stringent safety margins.

Based on experience accumulated during 25 years of collaboration, the Belgian Nuclear Research Center (SCK·CEN), together with Belgonucléaire, implemented a series of benchmark experiments in the VENUS critical facility at SCK·CEN in Mol, Belgium. The experiments were designed to provide organizations concerned with MOX fuel the ability to calibrate and improve their neutronic calculation tools.

A view of the VENUS critical facility showing the top of the nonpressurized vessel and the grid that holds the fuel rods in a vertical position.

VENUE capabilities

The VENUS critical facility is characterized by a high experimental flexibility:

The following parameters can be measured with high accuracy in VENUS:

The VENUS critical facility

The VENUS critical facility is a water-moderated zero-power reactor. (VENUS is short for Vulcain Experimental NUclear Study; Vulcain was a projected marine reactor.)

VENUS consists of an open (nonpressur-ized) stainless-steel cylindrical vessel and a set of grids that maintain fuel rods in a vertical position. After a fuel configuration has been loaded, criticality is reached by raising the water level within the vessel. Because the neutron flux is very low, no water circulation is needed to keep the fuel rods at low (room) temperature. The reactor shutdown is induced by emptying the vessel.

Experimental programs in VENUS

Because of its experimental flexibility, a series of experiments related to plutonium use in LWRs has been run in VENUS since 1990.

An early experiment investigated plutonium recycling in LWRs. The program, called VIP (VENUS International Programme), used fuel with high plutonium and gadolinium content. The aim of VIP was to validate reactor codes with respect to MOX fuel for both PWRs and BWRs. It focused on the criticality and fission-rate distribution calculation.

VIP was divided into two stages: VIP-BWR and VIP-PWR. VIP-BWR considered three mockups: an all-UO2 8x8 subassembly, an all-MOX 8x8 subassembly, and an island MOX 8x8 subassembly. VIP-PWR considered two mockups: an all-MOX 17x17 subassembly and a similar assembly containing burnable gadolinium absorber rods.

This program showed the validity of reactor codes such as DORT, TORT, GOG, TWOTRAN, and LWRWIMS for calculating criticality and fission-rate distributions in present-day fuel assemblies.

This experimental device, called a void box, simulates a void n a reactor core. This void box is transversed by highly enriched plutonium fuel.

To cope with future developments in the nuclear fuel cycle and the tendency of going to higher burnups, there was a need to investigate a possible positive void coefficient at high plutonium enrichments.

Therefore, an international program called VIPO (Void coefficient measurement In Plutonium mixed Oxide lattice) was established to determine the influence of a void bubble in a LWR reactor using high plutonium enrichment (i.e. from 10 percent to 15 percent) and the validation of related computer codes. A special experimental device, the so-called void box, was developed and constructed to simulate a void in the reactor core.

The VIPO program was successful in providing accurate benchmarks for validating reactor codes. It was shown that nowadays reactor codes can calculate fission-rate distributions with and without voids with an accuracy better than 2.5 percent.

After VIPO came VIPEX-PWR. This program, which was an extension of VIP-PWR, was aimed at determining core physics parameters of MOX assemblies that are of interest for reactor operations, such as the delayed neutron fraction (eff), the americium-241 effect, the moderator density effect, the control rod worth, and flux tilt within MOX rods.

The fraction of delayed neutrons is very important in reactor control; for plutonium-239 it is about three times as small as for uranium-235. So in principle, plutonium becomes prompt critical much faster than uranium. Although in a mixed plutonium-uranium configuration this effect is less pronounced, it is still important to have good quantification of the eff.

The aging of plutonium by the decay of plutonium-241 to americium-241 has a significant influence on the reactivity of a MOX assembly. By reloading pin by pin the same configuration used in VIP, the americium-241 effect was measurable after four years.

The present-day pitchÑthe distance between the fuel rodsÑof LWR fuel assemblies (12.6 millimeters) has been optimized for the use of uranium. This pitch results in a very undermoderated configuration for MOX fuel.

Investigation of moderator density effects on the reactivity can validate reactor codes for searches for the optimized pitch for MOX and for differences in hot and cold reactor conditions.

In VENUS, water-density reduction is simulated by inserting aluminum rods between fuel pins. To measure the effect of moderator density, the rods are pulled out from the configuration in several steps.

Because the neutron spectrum in a MOX assembly is harder-has higher energy-than in a UO2 assembly and the neutron-absorbing capacities of control rods are mainly based on the absorption of lower-energy thermal neutrons, the effectiveness of control rods-the control-rod worth-is less in a MOX fuel assembly. The control-rod worth can easily be measured in VENUS by inserting stepwise control rods at several positions and by determining the resulting difference in critical water level.

The reason for measuring the flux tilt is that the peripheral rods of the MOX assembly are subject to a large shift of the neutron spectrum because the neutrons from the UO2 assembly have a lower average energy than those from the MOX assembly.

The fission cross section for low energy is higher, resulting in a higher reaction rate for these neutrons. This is partially compensated by the higher neutron flux in the MOX assembly, but large power differences inside the fuel rod are still expected. Such flux tilt in VENUS is measured by inserting activation foils in between fuel pellets in a demountable rod or by direct measurements on the fuel pellet.

After VIPEX-PWR, we investigated recent BWR configurations in the BWR-NBN program, which was very similar to VIP-BWR, but for 9x9 BWR configuration. We also investigated the use of weapons-grade plutonium in LWRs. This was possible through the availability of plutonium fuel with plutonium vectors close to weapons-grade plutonium. This program, called IMP (Investigation of Military Plutonium), was an internal SCK·CEN program.

Recently, a new international program-REBUS-was implemented in VENUS. The program aims to establish an experimental benchmark for validation of reactor physics codes for the calculation of the loss of reactivity due to burnup for PWR fuel, both for UO2 and MOX fuel bundles.

The rationale for REBUS lies in the fact that present criticality safety calculations of irradiated fuel often have to model the fuel as fresh fuel because no precise experimental confirmation exists of the decrease of reactivity due to accumulated fission products.

Some useful definitions

Delayed neutron fraction (Betaeff): delayed neutrons appear in a fission event after an amount of time, as opposed to prompt neutrons, which appear almost immediately. While the fraction of delayed neutrons is small (less than 1 percent), accurate prediction of the effective delayed neutron fraction (the beta effect-or Betaeff) is vital for controlling the fission chain reaction in a nuclear reactor.

Americium-241 effect: The decay of plutonium-241 into americium-241 over time (plutonium-241 has a half-life of approximately 14.1 years).

Moderator density effect: The change in reactivity of the system when the density decreases from an increase in temperature or when the density increases from a decrease in temperature.

Control-rod worth: The effect on reactivity of inserting the control rods into the system. Flux tilt: A comparison of the flux in a MOX-fueled assembly (in particular, rods on the periphery vs. rods in the center) to that in surrounding uranium dioxide-fueled assemblies.

View of the REBUS container that will contain the spent fuel bundle

In other situations only actinide depletion is allowed to be taken into account and the influence of fission products has to be disregarded. The fact that this so-called "burnup credit" cannot (completely) be taken into account has serious economic implications for the transport, storage, and reprocessing of irradiated fuel. For long-term geological storage it is almost imperative to apply burnup credit.

In VENUS, the measurement of this burnup credit is based on comparative reactivity measurements between a fresh fuel bundle and an identical (initial composition) irradiated fuel bundle. Loading irradiated fuel bundles in a critical facility is rather new and for VENUS the necessary infrastructure for loading spent fuel in the REBUS container had to be installed.

Currently, an experimental program for three fresh fuel configurations has been executed and a dummy test for fine-tuning the loading procedure of irradiated fuel bundles is under way. The experimental program for the spent MOX fuel and for the spent UO2 bundles will be executed this spring.

Conclusions

The experiments carried out in the VENUS critical facility have demonstrated that VENUS is a very flexible and easy-to-use tool for the investigation of neutronic data as well as for the study of licensing, safety, and operation aspects.

Such data allow validation of the reactor physics codes for MOX use in LWRs. This validation has made it possible to safely operate thirty-seven PWRs and two BWRs with partial MOX-core loadings.

While MOX currently is not being used in a VVER-the Russian version of a PWR-a program to use weapons-grade plutonium in VVERs is under way. Based on the knowledge and experience resulting from VENUS' years of successful experimentation, researchers believe that the future use of MOX in a VVER is a viable option posing low technical risk.

This article was contributed by Pierre D'hondt, program manager for reactor safety at the Belgian Nuclear Research Center (SCK·CEN). D'hondt will be one of the plenary speakers at next month's Plutonium Futures-The Science Conference, set for July 6-10 in Albuquerque, N.M. (See article on Page 32). Here, he gives ARQ readers a preview of his talk.


NMT | LANL | DOE
Phone Book | Search | Help/Info

L O S  A L A M O S  N A T I O N A L   L A B O R A T O R Y
Operated by the University of California for the US Department of Energy

Questions? - Copyright © UC 1998-2000
- For conditions of use, see Disclaimer