C.L. Bentley

A deterministic method for transient, three-dimensional neutron transport

A deterministic method for solving the time-dependent, three-dimensional Boltzmam transport equation with explicit representation of delayed neutrons has been developed and evaluated. The methodology used in this study for the time variable of the neutron flux is known as the improved quasi-static (IQS) method. The position, energy, and angle-dependent neutron flux is computed deterministically by using the three-dimensional discrete ordinates code TORT. This paper briefly describes the methodology and selected results. The code developed at the University of Tennessee based on this methodology is called TDTORT. TDTORT can be used to model transients involving voided and/or strongly absorbing regions that require transport theory for accuracy. This code can also be used to model either small high-leakage systems, such as space reactors, or asymmetric control rod movements. TDTORT can model step, ramp, step followed by another step, and step followed by ramp type perturbations. It can also model columnwise rod movement can also be modeled. A special case of columnwise rod movement in a three-dimensional model of a boiling water reactor (BWR) with simple adiabatic feedback is also included. TDTORT is verified through several transient one-dimensional, two-dimensional, and three-dimensional benchmark problems. The results show that the transport methodology and corresponding code developed in this work have sufficient accuracy and speed for computing the dynamic behavior of complex multidimensional neutronic systems.

Analysis of a hypothetical criticality accident in a waste supercompactor

A hypothetical nuclear criticality accident in a waste supercompactor is examined. The material being compressed in the compactor is a homogeneous mixture of beryllium and {sup 239}Pu. The point-kinetics equations with simple thermal-hydraulic feedback are used to model the transient behavior of the system. A computer code has been developed to solve the model equations. The computer code calculates the fission power history, fission yield, bulk temperature of the system, and several other thermal-hydraulic parameters of interest. Calculations have been performed for the waste supercompactor for various material misloading configurations. The peak power for the various accident scenarios varies from 1.04 {times} 10{sup 17} to 4.85 {times} 10{sup 20} fissions per second (fps). The total yield varies from 8.21 {times} 10{sup 17} to 7.73 {times} 10{sup 18} fissions, and the bulk temperature of the system varies from 412 to >912 K.

Criticality Safety Evaluation of Shutdown Diffusion Cascade Coolers

A criticality safety study of diffusion cascade coolers in a shutdown state is presented. The coolers represent six typical cascade coolers at a gaseous diffusion plant with accumulated deposits of UO 2 F 2 . The study involves k eff calculations for the coolers with various distributions of UO 2 F 2 , which are assumed as part of several hypothetical accident scenarios. The results show that at least two independent failures must occur in order to have a criticality. Additionally, the distributions chosen represent the upper bounds for k eff . Individual results show that the k eff values for the cascade coolers designed for 80 and 97% enriched UF 6 with deposit amounts <2.409 and 2.185 kg, respectively, will not exceed 0.9 for the accident scenarios modeled. All other coolers require shell-side flooding with H 2 O in order to cause a criticality, which is possible only if two or more independent failures occur.

Validation of KENO V.a with ENDF/B-V cross sections for 233U systems

The multigroup Monte Carlo code KENO V.a and the 238- and 44-energy-group ENDF/B-V cross-section libraries were validated for {sup 233}U systems. Fifty-one critical experiments involving {sup 233}UO{sub 2}(NO{sub 3}){sub 2}, {sup 233}UO{sub 2}F{sub 2}, or {sup 233}U metal were selected for the validation. The H/{sup 233}U ratios for the experiments range from 0 to 1986. Each experiment was modeled with KENO V.a, and the effective multiplication factor k{sub eff} was calculated for each system using the 44- and 238-group ENDF/B-V, the 27- and 218-group ENDF/B-IV, and the 16-group Hansen-Roach cross-section libraries. The mean calculated k{sub eff} for all experiments using the 44- and 238-group libraries is 1.0090 {+-} 0.0021 and 1.0064 {+-} 0.0020, respectively. For comparison, the mean calculated k{sub eff} using the 27-, 218-, and 16-group libraries is 1.0142 {+-} 0.0038, 1.0125 {+-} 0.0038, and 0.9991 {+-} 0.0019, respectively. In general, an improvement exists in the agreement between the calculated k{sub eff}`s and the experimental results (i.e., k{sub eff} = 1.0) obtained with the newer ENDF/B-V libraries relative to ENDF/B-IV. This study is pertinent to {sup 233}U storage outside of the reactor.

Improved dose estimates for nuclear criticality accidents

Slide rules are improved for estimating doses and dose rates resulting from nuclear criticality accidents. The original slide rules were created for highly enriched uranium solutions and metals using hand calculations along with the decades old Way-Wigner radioactive decay relationship and the inverse square law. This work uses state-of-the-art methods and better data to improve the original slide rules and also to extend the slide rule concept to three additional systems; i.e., highly enriched (93.2 wt%) uranium damp (H/{sup 235}U = 10) powder (U{sub 3}O{sub 8}) and low-enriched (5 wt%) uranium mixtures (UO{sub 2}F{sub 2}) with a H/{sup 235}U ratio of 200 and 500. Although the improved slide rules differ only slightly from the original slide rules, the improved slide rules and also the new slide rules can be used with greater confidence since they are based on more rigorous methods and better nuclear data.