Fission Reactor Analysis

This program is concerned primarily with the behavior of neutrons in thermal and fast fission reactors and includes such topics as neutron diffusion and slowing down, criticality, numerical methods, and transport theory. A wide range of research activity is carried out in this area. It includes conception, design and analysis of advanced reactors such as novel concepts of power reactors (see paper), reactors for the transmutation of nuclear waste (see paper) and reactors for space exploration (see paper); conception and analysis of advanced nuclear fuel cycles (see paper), including proliferation resistant multi-recycling of the nuclear fuel; Investigation of possibilities for improving the design and performance of Light-Water Reactors (see paper); Development of improved computational methods for core design and analysis (see paper); development of intelligent methods for the optimization of the design of nuclear systems (see paper); criticality safety analysis (see paper); as well as radiation shielding design optimization (see paper) and design of facilities for medical applications of nuclear radiation (see paper). We are actively involved in the Generation-IV, Nuclear Energy Research Initiative, Nuclear Engineering Education Research and Advanced Fuel Cycle Initiative programs of DOE as well as in the NASA space nuclear power program. We have research collaboration with National Laboratories including Lawrence Berkeley National Laboratory, Lawrence Livermore National Laboratory, Los Alamos National Laboratory, Argonne National Laboratory, Idaho National Laboratory and Oak-Ridge National Laboratory. Professor Vujic is also involved in this project.

The solution of particle transport problems in many areas (atmospheric studies, cancer therapy, reactor core analysis) require computational tools that are more accurate, and that can efficiently handle and routinely analyze a variety of complex systems without any changes to the solution algorithm. A particularly difficult challenge is to be able to keep all geometrical details in large domains while preserving accurate mathematical models of the physical phenomena. In order to remove some of the most serious disadvantages of both stochastic Monte Carlo (MC) and non-stochastic computational methods in particle transport, the development of a unified computational method which will bring together the best features from both groups is under the way. Two deterministic methods are being used: the collision probability method (CP) and the method of characteristics (CM). The geometry-dependent ray-tracing part in both CP and CM methods is decoupled from the physics part, and replaced by a geometry-independent Monte Carlo type ray-tracing. We are also working on extension to general 3D geometries. Of particular interest are irregularly shaped cell or domain surfaces (for example, human brain). A new set of integral transport equations needs to be derived for this case.

There are several important topics that will be addressed in the future: higher orders of scattering, domain decompositions for large domains (full 3D reactor core, for example); improved interface-current coupling of newly formed subdomains; more general boundary conditions on the external domain surfaces (vacuum, reflective, periodic, albedo);
coarse mesh rebalancing over large domains; efficient integration over 3D irregular nodes, and efficient use of parallel computers or networks of workstations to speedup the calculation. This project is partialy support by the Committee on Research, University of
California, Berkeley.

The Encapsulated Nuclear Heat Source (ENHS) is an innovative reactor concept that features a number of unique features, including (a) >20 effective full power years (EFPY) of operation without refueling and fuel shuffling; (b) a nearly zero change of reactivity with operation; (c) 100% natural circulation; (d) autonomous operation and (e) superb safety.

The ENHS is based on the technology of lead-bismuth cooled small reactors the Russian developed for their most advanced nuclear submarines. The ENHS is to be factory manufactured and delivered already fueled and weld-sealed to a power plant, installed in the reactor pool and provide energy for ~ 22 EFPY without refueling or fuel shuffling. At the end of the core life this “nuclear battery” would be changed out for a replacement “battery" and transported to a regional center for backend fuel cycle services.

The ENHS would, in fact, constitute a totally new fuel management concept. It may be suitable for developing countries as well as for industrial countries. Following the postulation, in 1998, of this reactor concept by Ehud Greenspan and David Wade, its feasibility has been studied during 1999 through 2002 with the support of the DOE NERI program. This feasibility study was done in collaboration with Argonne National
Laboratory, Lawrence Livermore National Laboratory, Westinghouse and several Korean and Japanese organizations. Following this feasibility study the ENHS was selected as one of the reactor concepts that are candidates for Generation IV reactors. Its study continues with the support of the Generation-IV program as well as with the support of the Lawrence Livermore National Laboratory. 

This three-year study is sponsored by the DOE NERI program. It was initiated in September 2002. Collaborating with UC Berkeley are Prof. Neil Todreas of MIT, Westinghouse and Professor Yamawaki of the University of Tokyo. The objective of this study is to assess the feasibility of improving the design and performance of pressurizes (PWR) and boiling (BWR) water reactors by fueling them with hydride instead of oxide fuel. Several hydride fuel compositions are considered, including uranium-zirconium hydride, uranium-thorium hydride and plutonium-thorium hydride. The performance improvement expected from this novel design approach include a higher energy generation per core loading; longer core life and, hence, higher  capacity factor; higher core power level or, alternatively, smaller core and reactor vessel volume. The net outcome is expected to be improved economics, improved resource utilization, reduced waste, improved proliferation resistance and improved safety. Prof. Donald Olander of UCB NE is participating in this project.

We are participating in a team led by Lockheed-Martin that is performing trade studies for the development of a nuclear electric propulsion space system concept capable of supporting outer planet science missions. For the purpose of this study, a Jupiter Icy Moons Orbiter (JIMO) to explore the three Icy Moons of Jupiter has been chosen. Later use of the space system could include missions to Saturn, Neptune, and Pluto. The space system reactor power should have the modularity and/or flexibility to provide up to 300 kWe.  NASA and DOE are supporting this project. UCB involvement includes conception of inherently safe core design, assessment of the safety of core designs and optimizing the radiation shield design so as to minimize its weight. Our unique nuclear design optimization code SWAN is used for the latter task.

Many dozens of man-years have been invested over the last 40 years or so in the development of methods and computer codes for the solution of the neutron and photon transport equation. Presently, there are very efficient codes for solving the transport equation using either a deterministic or a stochastic approach; for many types of problems they can provide answers at desirable accuracy in acceptable computer running time. Be them as efficient solvers of the transport equation as they are today, these codes are not very efficient design tools. The designer of a nuclear system needs to solve the transport equation many times before he can identify the optimal (or, actually, a supposedly optimal) design. The search for the optimal design is being done, so far, using a "trial-and-error" approach. At no phase during this laborious process the designer knows (1) in what direction to change the system design variables so as to approach the optimal design, and (2) how close is the design to the optimal. Using perturbation theory formulation, we developed a nuclear design optimization method that is capable of an intelligent, efficient and reliable search for optimal nuclear designs. The SWAN code has been successfully applied to the optimization of radiation shields, fusion reactor blankets and, more recently, medical facilities for treating brain tumors based on boron neutron capture therapy (BNCT). Recently we developed a new version of SWAN that can search for minimum critical mass or for maximum keff; it can be very helpful for criticality safety analysis. Presently we have two projects involving the upgrading of SWAN: One is the incorporation of SWAN within the SCALE code package for criticality safety analysis (sponsored by Oak Ridge National Laboratory). SCALE-5, the new version of the SCALE code package soon to be released by ORNL will incorporate our SWAN code as one of its new calculation sequences. This will make our SWAN code available to the large international SCALE users community. The other is the development of a two-dimensional version of SWAN (was sponsored by the DOE NEER program).