Krešimir Trontl

Modeling of the ORNL PCA Benchmark Using SCALE6.0 Hybrid Deterministic-Stochastic Methodology

Revised guidelines with the support of computational benchmarks are needed for the regulation of the allowed neutron irradiation to reactor structures during power plant lifetime. Currently, US NRC Regulatory Guide 1.190 is the effective guideline for reactor dosimetry calculations. A well known international shielding database SINBAD contains large selection of models for benchmarking neutron transport methods. In this paper a PCA benchmark has been chosen from SINBAD for qualification of our methodology for pressure vessel neutron fluence calculations, as required by the Regulatory Guide 1.190. The SCALE6.0 code package, developed at Oak Ridge National Laboratory, was used for modeling of the PCA benchmark. The CSAS6 criticality sequence of the SCALE6.0 code package, which includes KENO-VI Monte Carlo code, as well as MAVRIC/Monaco hybrid shielding sequence, was utilized for calculation of equivalent fission fluxes. The shielding analysis was performed using multigroup shielding library v7_200n47g derived from general purpose ENDF/B-VII.0 library. As a source of response functions for reaction rate calculations with MAVRIC we used international reactor dosimetry libraries (IRDF-2002 and IRDF-90.v2) and appropriate cross-sections from transport library v7_200n47g. The comparison of calculational results and benchmark data showed a good agreement of the calculated and measured equivalent fission fluxes.

Long Term Sustainability of Nuclear Fuel Resources

The basic issue in considering the contribution of nuclear power to solving the world’s energy problem in the future is the availability of uranium resources and its adequacy in meeting the future needs of nuclear capacity. Increased interest in nuclear energy is evident, and a new look into nuclear fuel resources is relevant. In this chapter we address the issue of nuclear fuel resources long term sustainability in relation to the expected growth of the world nuclear power. Three main aspects have to be analyzed in order to estimate how long the world’s nuclear fuel supplies will last: nuclear fuel resources (uranium and thorium), technologies for nuclear fuel utilization, and energy requirements growth scenarios including different scenarios for nuclear share growth. Uranium nuclear fuel resources are analyzed based on joint OECD Nuclear Energy Agency (NEA) and the International Atomic Energy Agency categorization which classifies resources into conventional resources and unconventional resources. Conventional resources are further divided into identified resources (reasonably assured and inferred) with four cost ranges and undiscovered resources (prognosticated and speculative) with three cost ranges. Analyzed unconventional resources include phosphate deposits and seawater. Total resources are estimated to about 6.3 million tons of uranium in identified resources and 10.4 million tons of uranium in undiscovered resources. The amount of uranium contained in phosphate deposits and seawater is estimated to 22 million tons and 4 billion tons, respectively. Thorium nuclear fuel resources are analyzed based on slightly different categorization than uranium one. Four categories are used: reasonably assured resources, estimated additional resources of type I and II, and prognosticated resources. Total thorium resources according to the latest OECD-NEA report are estimated to about 6 million tons. Detailed analysis of potential technologies for improved nuclear fuel utilization is required in order to assess long term sustainability of nuclear fuel resources. Nowadays, thermal converter reactor technology with once-through nuclear fuel cycle is used. The effectiveness of the technology can be improved in the area of enrichment process as well as by introducing reprocessing of the spent fuel on larger scale. Other technologies are also on the development stage that allows their implementation in short or medium period of time. These include: thermal and fast breeder reactors of different kind, thorium based fuel cycle, and conversion of uranium or thorium by particle accelerators or fusion devices. Very important aspect of long term sustainability of nuclear fuel resources are scenarios for energy requirements growth, and scenarios for growth of nuclear share in electricity production resulting in overall nuclear capacity growth. On one hand, low growth scenario has a moderate continuous growth strategy of 1.3% per year. On the other hand, high growth scenarios are required if nuclear energy is to give an essential contribution to carbon emission control. Finally, there is a scenario based on a compromise between low and high growth assumptions. These scenarios have been used to estimate exhaustion time periods of nuclear fuel resources for different fuel utilization technologies. The obtained exhaustion time periods range from 50 years for the once-through thermal converter fuel cycle with high increase in world nuclear capacity to many thousands of years for breeder reactor technology or for other methods of releasing the energy of U238. A development of these methods would require several decades, in return for practically unlimited amount of nuclear energy. Our analysis shows that nuclear fuel resources are not a limiting factor for a long term large scale nuclear power development. Our study is based on the most recent data on nuclear fuel resources, energy growth predictions, and technologies for the nuclear fuel utilization improvement, as well as our own developed code for calculation of nuclear fuel demand for different nuclear energy growth scenarios.

PWR Containment Shielding Calculations with SCALE6.1 Using Hybrid Deterministic-Stochastic Methodology

The capabilities of the SCALE6.1/MAVRIC hybrid shielding methodology (CADIS and FW-CADIS) were demonstrated when applied to a realistic deep penetration Monte Carlo (MC) shielding problem of a full-scale PWR containment model. Automatic preparation of variance reduction (VR) parameters is based on deterministic transport theory ( method) providing the space-energy importance function. The aim of this paper was to determine the neutron-gamma dose rate distributions over large portions of PWR containment with uniformly small MC uncertainties. The sources of ionizing radiation included fission neutrons and photons from the reactor and photons from the activated primary coolant. We investigated benefits and differences of FW-CADIS over CADIS methodology for the objective of the uniform MC particle density in the desired tally regions. Memory intense deterministic module was used with broad group library “v7_27n19g” opposed to the fine group library “v7_200n47g” used for final MC simulation. Compared with CADIS and with the analog MC, FW-CADIS drastically improved MC dose rate distributions. Modern shielding problems with large spatial domains require not only extensive computational resources but also understanding of the underlying physics and numerical interdependence between -MC modules. The results of the dose rates throughout the containment are presented and discussed for different volumetric adjoint sources.

Active learning for support vector regression in radiation shielding design

Recently a novel approach based on support vector regression technique has been proposed and tested for the estimation of multi layer buildup factors for gamma ray shielding calculations, while for neutron shielding calculations some initial analyses have been conducted. During the development of the model a number of questions regarding possible application of active learning measures have been raised. In this paper general applicability of the active learning measures on the problem, in particular data transfer method used in the investigation, and testing of the active procedure are discussed.

Machine Learning of the Reactor Core Loading Pattern Critical Parameters

Correspondence should be addressed to Kreˇsimir Trontl, kresimir.trontl@fer.hrReceived 11 March 2008; Accepted 23 June 2008Recommended by Igor JencicThe usual approach to loading pattern optimization involves high degree of engineering judgment, a set of heuristic rules, anoptimization algorithm, and a computer code used for evaluating proposed loading patterns. The speed of the optimizationprocess is highly dependent on the computer code used for the evaluation. In this paper, we investigate the applicability of amachine learning model which could be used for fast loading pattern evaluation. We employ a recently introduced machinelearning technique, support vector regression (SVR), which is a data driven, kernel based, nonlinear modeling paradigm, in whichmodel parameters are automatically determined by solving a quadratic optimization problem. The main objective of the workreported in this paper was to evaluate the possibility of applying SVR method for reactor core loading pattern modeling. Weillustrate the performance of the solution and discuss its applicability, that is, complexity, speed, and accuracy.Copyright