[en] This paper describes a model for determining the stability and associated radionuclide concentrations of colloids that might be present in the nuclear waste package environment from degradation of the nuclear waste forms. The model simplifies radionuclide-colloid behavior by assuming that all colloids can be defined as either smectite clay, a mixed actinide-bearing rare earth-zirconium oxide, iron oxyhydroxide (ferrihydrite (FeOOH), or uranophane (Ca(UO2)2(SiO3OH)2(H2O)5)). However, for the purposes of predictive stability modeling, the colloids are conceptually represented as montmorillonite, ZrO2, hematite, and meta autunite, respectively. The model uses theoretical calculations and laboratory data to determine the stability of modeled colloids with ionic strength and pH. The true nature of colloid composition and heterogeneity, generation, and flocculation will be extremely complex, involving the formation of numerous types of phases, often depending on the composition of the various waste forms and waste package materials. This model strives to capture the uncertainty of the real system using theoretical models. In this paper, one of the four representative colloids designed to capture the behavior of the spent fuel derived colloids is described in detail.

[en] This paper examines the problems associated with the analysis of low levels of neptunium (Np) in a uranium (U) matrix with electron energy-loss spectroscopy (EELS) on the transmission electron microscope (TEM). The detection of Np in a matrix of uranium (U) can be impeded by the occurrence of a plural scattering event from U (U-M5 + U-O4,5) that results in severe overlap on the Np-M5 edge at 3665 eV. Low levels (1600 - 6300 ppm) of Np can be detected in U solids by confirming the energy gap between the Np-M5 and Np-M4 edges is at 184 eV and showing that the M4/M5 ratio for the Np is smaller than that for U. The Richardson-Lucy deconvolution method was applied to energy-loss spectral images and was shown to increase the signal to noise. This method also improves the limits of detection for Np in a U matrix.

[en] Radiation portal monitors are being deployed, both domestically and internationally, to detect illicit movement of radiological materials concealed in cargo. Evaluation of the current and next generations of these radiation portal monitor (RPM) technologies is an ongoing process. 'Injection studies' that superimpose, computationally, the signature from threat materials onto empirical vehicle profiles collected at ports of entry, are often a component of the RPM evaluation process. However, measurement of realistic threat devices can be both expensive and time-consuming. Radiation transport methods that can predict the response of radiation detection sensors with high fidelity, and do so rapidly enough to allow the modeling of many different threat-source configurations, are a cornerstone of reliable evaluation results. Monte Carlo methods have been the primary tool of the detection community for these kinds of calculations, in no small part because they are particularly effective for calculating pulse-height spectra in gamma-ray spectrometers. However, computational times for problems with a high degree of scattering and absorption can be extremely long. Deterministic codes that discretize the transport in space, angle, and energy offer potential advantages in computational efficiency for these same kinds of problems, but the pulse-height calculations needed to predict gamma-ray spectrometer response are not readily accessible. These complementary strengths for radiation detection scenarios suggest that coupling Monte Carlo and deterministic methods could be beneficial in terms of computational efficiency. Pacific Northwest National Laboratory and its collaborators are developing a RAdiation Detection Scenario Analysis Toolbox (RADSAT) founded on this coupling approach. The deterministic core of RADSAT is Attila, a three-dimensional, tetrahedral-mesh code originally developed by Los Alamos National Laboratory, and since expanded and refined by Transpire, Inc. [1]. MCNP5 is used to calculate sensor pulse-height tallies. RADSAT methods, including adaptive, problem-specific energy-group creation, ray-effect mitigation strategies and the porting of deterministic angular flux to MCNP for individual particle creation are described in [2][3][4]. This paper discusses the application of RADSAT to the modeling of gamma-ray spectrometers in RPMs.

[en] Monte Carlo methods are typically used for simulating radiation fields around gamma-ray spectrometers and pulse-height tallies within those spectrometers. Deterministic codes that discretize the linear Boltzmann transport equation can offer significant advantages in computational efficiency for calculating radiation fields, but stochastic codes remain the most dependable tools for calculating the response within spectrometers. For a deterministic field solution to become useful to radiation detection analysts, it must be coupled to a method for calculating spectrometer response functions. This coupling is done in the RADSAT toolbox. Previous work has been successful using a Monte Carlo boundary sphere around a handheld detector. It is desirable to extend this coupling to larger detector systems such as the portal monitors now being used to screen vehicles crossing borders. Challenges to providing an accurate Monte Carlo boundary condition from the deterministic field solution include the greater possibility of large radiation gradients along the detector and the detector itself perturbing the field solution, unlike smaller detector systems. The method of coupling the deterministic results to a stochastic code for large detector systems can be described as spatially defined rectangular patches that minimize gradients. The coupled method was compared to purely stochastic simulation data of identical problems, showing the methods produce consistent detector responses while the purely stochastic run times are substantially longer in some cases, such as highly shielded geometries. For certain cases, this method has the ability to faithfully emulate large sensors in a more reasonable amount of time than other methods.

[en] Results for a radiolysis model sensitivity study of radiolytically produced H₂O₂ are presented as they relate to Spent (or Used) Light Water Reactor uranium oxide (UO₂) nuclear fuel (UNF) oxidation in a low oxygen environment. The model builds on previous reaction kinetic studies to represent the radiolytic processes occurring at the nuclear fuel surface. Hydrogen peroxide (H₂O₂) is the dominant oxidant for spent nuclear fuel in an O₂-depleted water environment. The most sensitive parameters have been identified with The most sensitive parameters have been identified with respect to predictions under typical conditions. As compared with the full model with about 100 reactions, it was found that only 30 to 40 of the reactions are required to determine [H₂O₂] to one part in 10^–5 and to preserve most of the predictions for major species. This allows a systematic approach for model simplification and offers guidance in designing experiments for validation. (author)

[en] The purpose of this study was to calculate a more accurate dose rate constant for the ¹³¹Cs (model CS-1, IsoRay Medical, Inc., Richland, WA) interstitial brachytherapy seed. Previous measurements of the dose rate constant for this seed have been reported by others with incongruity. Recent direct measurements by thermoluminescence dosimetry and by gamma-ray spectroscopy were about 15% greater than earlier thermoluminescence dosimetry measurements. Therefore, we set about to calculate independent values by a Monte Carlo approach that combined three estimates as a consistency check, and to quantify the computational uncertainty. The calculated dose rate constant for the ¹³¹Cs seed was 1.040 cGy h^-1 U^-1 for an ionization chamber model and 1.032 cGy h^-1 U^-1 for a circular ring model. A formal value of 2.2% uncertainty was calculated for both values. The range of our multiestimate values were from 1.032 to 1.061 cGy h^-1 U^-1. We also modeled three ¹²⁵I seeds with known dose rate constants to test the accuracy of this study's approach

[en] This report gives the details on the effect of iron chemistry on H₂O₂ decomposition under radiolytic condition at the surface of used nuclear fuel under repository conditions. Additionally, suggestions are offered on what further data or measurements would be required for model verification and applicability.

[en] A separation procedure was developed and implemented for the isolation of gold (Au) and platinum (Pt) radionuclides from deuteron irradiated Pt using tri-butyl phosphate (TBP) resin. Computational modeling and experimental results of the Au and Pt radionuclide production were reported and compared to literature values. Results of the separation procedure demonstrated a Pt recovery of 98 ± 2% and a Au recovery of 92 ± 7%, in their respective fractions (n = 4 ± 1σ) The decontamination factors were 2900 ± 500 for Au removal from the Pt fraction and 830 ± 30 for Pt removal from the Au fraction (n = 4 ± 1σ).

[en] A boron carbide capsule has been designed and used for spectral-tailoring experiments at the TRIGA reactor at Washington State University. Irradiations were conducted in pulsed mode and in continuous operation for up to 4 hours. A cadmium cover was used to reduce thermal heating. The neutron spectrum calculated with MCNP was found to be in good agreement with reactor dosimetry measurements using the STAY'SL computer code. The neutron spectrum resembles that of a fast reactor. Design of a capsule using boron carbide enriched in ¹⁰B shows that it is possible to produce a neutron spectrum similar to ²³⁵U fission.

[en] This report fulfills the milestone (M4SF-17PN010501041 Report on Radiolysis Modeling for the Defense Repository) to discuss continued integration of the PNNL Radiolysis Model and the ANL Mixed Potential Model. This work concerns the development of an integrated Radiolysis Model (RM) for evaluating defense waste materials (oxide and metal) degradation and radionuclide mobilization. Within an anoxic repository environment, primary oxidizing species (e.g., hydrogen peroxide (H₂O₂), OH• radicals, as well as chlorate and other oxidizing species depending on the disposal environment) will be generated at the surface of the nuclear waste forms as a function of their specific activity. RM development has included expansion of chemical environments considered to encompass species for various disposal environments. PNNL has been coordinating this effort with ANL on the integration of the radiolysis work with the fuel degradation model. In this study, we demonstrate and approximate possible effects of iodide on H₂O₂ generation. As has been shown these are conditions for which H₂O₂ generation is reduced. We find that the presence of the iodide ion reduces the steady-state H₂O₂ concentration, but not to the same degree as bromide at micro-molar concentrations.