Understanding the potential risk of stress corrosion cracking of spent nuclear fuel dry storage canisters has been identified as a knowledge gap for determining the safety of long-term interim storage of spent nuclear fuel. To address this, the DOE is funding a multi-lab DOE effort to understand the timing, occurrence, and consequences of potential canister SCC. Sandia National Laboratories has developed a probabilistic model for canister penetration by SCC. This model has been continuously updated at SNL since 2014. Model uncertainties are treated using a nested loop structure, where the outer loop accounts for uncertainties due to lack of data and the inner aleatoric loop accounts for uncertainties due to variation in nature. By separating uncertainties into these categories, it is possible to focus future work on reducing the most influential epistemic uncertainties. Several experimental studies have already been performed to improve the modeling approach through expanded process understanding and improved model parameterization. The resulting code is physics-based and intended to inform future work by identifying (1) important modeling assumptions, (2) experimental data needs, and (3) necessary model developments. In this document, several of the sub-models in the probabilistic SCC model have been exercised, and the intermediate results, as the model progresses from one sub-model to the next, are presented. Evaluating the sub-models in this manner provides a better understanding of sub-model outputs and has identified several unintended consequences of model assumptions or parameterizations, requiring updates to the modeling approach. The following updates have been made, and future updates have been identified.
Thermodynamic modeling has been used to predict chemical compositions of brines formed by the deliquescence of sea salt aerosols. Representative brines have been mixed, and physical and chemical properties have been measured over a range of temperatures. Brine properties are discussed in terms of atmospheric corrosion of austenitic stainless steel, using spent nuclear fuel dry storage canisters as an example. After initial loading with spent fuel, during dry storage, the canisters cool over time, leading to increased surface relative humidities and evolving brine chemistries and properties. These parameters affect corrosion kinetics and damage distributions, and may offer important constraints on the expected timing, rate, and long-term impacts of canister corrosion.
Stainless steels are susceptible to localized forms of corrosion attack, such as pitting. The size and lifetime of a nucleated pit can vary, depending on a critical potential or current density criterion, which determines if the pit repassivates or continues growing. This work uses finite element method (FEM) modeling to compare the critical pit radii predicted by thermodynamic and kinetic repassivation criteria. Experimental electrochemical boundary conditions are used to capture the active pit kinetics. Geometric and environmental parameters, such as the pit shape and size (analogous to additively manufactured lack-of-fusion pores), solution concentration, and water layer thickness were considered to assess their impact on the pit repassivation criterion. The critical pit radius (the transition point from stable growth to repassivation) predicted for a hemispherical pit was larger when using the repassivation potential (Erp) criteria, as opposed to the current density criteria (pit stability product). Including both the pit stability product and Erp into its calculations, the analytical maximum pit model predicted a critical radius two times more conservative than the FEA approach, under the conditions studied herein. The complex pits representing lack-of-fusion pores were shown to have minimal impact on the critical radius in atmospheric conditions.
This progress report describes work performed during FY21 at Sandia National Laboratories (SNL) to assess the localized corrosion performance of canister materials used in the interim storage of spent nuclear fuel (SNF). Of particular concern is stress corrosion cracking (SCC), by which a through-wall crack could potentially form in a canister outer wall over time intervals that are shorter than possible dry storage times. In FY21, modeling and experimental work was performed that further defined our understanding of the potential chemical and physical environment present on canister surfaces at both marine and inland sites. Research also evaluated the relationship between the environment and the rate, extent, and morphology of corrosion, as well as the corrosion processes that occur. Finally, crack growth rate testing under relevant environmental conditions was initiated.
Stress corrosion cracking (SCC) is an important failure degradation mechanism for storage of spent nuclear fuel. Since 2014, Sandia National Laboratories has been developing a probabilistic methodology for predicting SCC. The model is intended to provide qualitative assessment of data needs, model sensitivities, and future model development. In fiscal year 2021, improvement of the SCC model focused on the salt deposition, maximum pit size, and crack growth rate models.
During typical atmospheric conditions, cathodic reduction reactions produce hydroxyl ions increasing the pH in the cathodic region. Therefore, cathodic reduction reactions are investigated on platinum and stainless steel 304 L (SS304L) in NaOH solutions ranging in pH from 13.6 to 16.5. It was found that in solution pHs less than 16.5 the cathodic reduction reaction on Pt and SS304L was ORR with an electron transfer number less than two due to superoxide formation as an intermediate. Increasing pH decreased the number of electrons transferred. At a pH of 16.5, the cathodic reduction reaction on SS304L is no longer ORR and the cathodic current on the surface of the alloy is due to oxide reduction occurring on the surface as indicated by the creation of multi-component Pourbaix diagrams. The results of this study have important implications for predicting corrosion in atmospheric environments.
Pit growth and repassivation are complex, with many interconnecting geometric and environmental parameters to consider. Experimentally, it is difficult to isolate these individual parameters to study their effect on the stability of pits. To enable these studies, a finite element modeling approach has been developed to allow systematic testing of parameters that impact a pit’s stability. The specific parameters studied were the cathode diameter, the pit diameter and shape, and the water layer thickness. Hemispherical and rectangular-based pits were studied to determine the impact of the overall pit shape. Pit stability results were compared with mathematical calculations based on the Maximum Pit Model, for both 50% saturation and 100% saturated salt film coverage. Further studies expanded the range of pit geometry to those relevant to additively manufactured surfaces.
The natural convection boundary layer (δnc) and its influence on cathodic current in a galvanic couple under varying electrolytes as a function of concentration (1 - 5.3 M NaCl) and temperature (25 °C-45 °C) were understood. Polarization scans were obtained under quiescent conditions and at defined boundary layer thicknesses using a rotating disk electrode on platinum and stainless steel 304L (SS304L); these were combined to determine δ nc With increasing chloride concentration and temperature, δnc decreased. Increased mass transport (Sherwood number) results in a decrease in δnc providing a means to predict this important boundary. Using Finite Element Modeling, the cathodic current was calculated for an aluminum alloy/SS304L galvanic couple as a function of water layer (WL) thickness and cathode length. Electrolyte domains were delineated, describing (i) dominance of ohmic resistance over mass transport under thin WL, (ii) the transition from thin film to bulk conditions at δncnc and (iii) dominance of mass transport under thick WL. With increasing chloride concentration, cathodic current decreased due to decreases in mass transport. With increasing temperature, increased cathodic current was related to increases in mass transport and solution conductivity. This study has implications for sample sizing and corrosion prediction under changing environments.
Maximum pit sizes were predicted for dilute and concentrated NaCl and MgCl2 solutions as well as sea-salt brine solutions corresponding to 40% relative humidity (RH) (MgCl2-rich) and 76% RH (NaCl-rich) at 25 °C. A quantitative method was developed to capture the effects of various cathode evolution phenomena including precipitation and dehydration reactions. Additionally, the sensitivity of the model to input parameters was explored. Despite one's intuition, the highest chloride concentration (roughly 10.3 M Cl−) did not produce the largest predicted pit size as the ohmic drop was more severe in concentrated MgCl2 solutions. Therefore, the largest predicted pits were calculated for saturated NaCl (roughly 5 M Cl−). Next, it was determined that pit size predictions are most sensitive to model input parameters for concentrated brines. However, when the effects of cathodic reactions on brine chemistry are considered, the sensitivity to input parameters is decreased. Although there was not one main input parameter that influenced pit size predictions, two main categories were identified. Under similar chloride concentrations (similar RH), the water layer thickness (WL), and pit stability product, (i·x)sf, are the most influential factors. When varying chloride concentrations (RH), changes in WL, the brine specific cathodic kinetics on the external surface (captured in the equivalent current density (ieq)), and conductivity (κo) are the most influential parameters. Finally, it was noted that dehydration reactions coupled with precipitation in the cathode will have the largest effect on predicted pit size, and cause the most significant inhibition of corrosion damage.
Cathodic kinetics in magnesium chloride (MgCl2) solutions were investigated on platinum (Pt) and stainless steel 304 L (SS304 L). Density, viscosity, and dissolved oxygen concentration for MgCl2 solutions were also measured. A 2-electron transfer for oxygen reduction reaction (ORR) on Pt was determined using a rotating disk electrode. SS304 L displayed non-Levich behavior for ORR and, due to ORR suppression and buffering of near surface pH by Mg-species precipitation, the primary cathodic reaction was the hydrogen evolution reaction (HER) in saturated MgCl2. Furthermore, non-carbonate precipitates were found to be kinetically favored. Implications of HER are discussed through atmospheric corrosion and stress corrosion cracking.
This progress report describes work performed during FY20 at Sandia National Laboratories (SNL) to assess the localized corrosion performance of container/cask materials used in the interim storage of spent nuclear fuel (SNF). Of particular concern is stress corrosion cracking (SCC), by which a through-wall crack could potentially form in a canister outer wall over time intervals that are shorter than possible dry storage times. Work in FY20 further defined our understanding of the potential chemical and physical environment present on canister surfaces, evaluated the relationship between the environment and the resultant corrosion that occurs, and initiated crack growth rate testing under relevant environmental conditions. In FY20, work to define dry storage canister surface environments included several tasks. First, collection of dust deposition specimens from independent spent fuel storage installation (ISFSI) site locations helped to establish a more complete understanding of the potential chemical environment formed on the canister. Second, the predicted evolution of canister surface relative humidity RH) values was estimated using ISFSI site weather data and the horizontal canister thermal model used by the SNL probabilistic SCC model. These calculations determined that for typical ISFSI weather conditions, seasalt deliquescence to produce MgCl2-rich brines could occur in less than 20 years at the coolest locations on the canister surface, and, even after nearly 300 years, conditions for NaCl deliquescence (75% RH) are not reached. This work illustrates the importance of understanding the stability of MgCl2-rich brines on the heated canister surface, and the potential impact of brine composition on corrosion processes, including pitting and stress corrosion cracking. In an additional study, the description of the canister surface environment was refined in order to define more realistic corrosion testing environments including diurnal cycles, soluble salt chemistries, and inert mineral particles. The potential impacts of these phenomena on canister corrosion are being evaluated experimentally. Finally, work over the past few years to evaluate the stability of magnesium chloride brines continued in FY20. MgCl2 degassing experiments were carried out, confirming that MgCl2 brines slowly degas HCl on heated surfaces, converting to less deliquescent magnesium hydroxychloride phases and potentially leading to brine dryout.
A sensor to determine water layer (WL) thickness, ranging from 0−5 mm, in salt-spray testing is presented. WL thickness is based on electrical resistivity and sensor design was guided by Finite Element Modeling with validation under known WL thicknesses. WLs were measured in continuous salt spray testing and angle of exposure played the largest role in thicknesses. At angles greater than 20˚ from vertical, semi-periodic run-off decreased WLs up to 80 %. Finally, exposure angle determines if thin-film conditions are achieved, likely influencing corrosion rate and morphology. Allowances for sample angle in testing standards pose a potentially large source of variability.
Here, for the first time, we demonstrate the use of an in situ spectroelectrochemical Raman technique to explore simulated atmospheric corrosion scenarios with a variable boundary layer thickness (δ). The effects of solution flow rate on oxygen concentration and δ were explored. It was found solution regeneration is necessary to prevent oxygen depletion in the Raman cell. It was further shown that by increasing the solution flow rate, the effective δ decreases and allows for the investigation of atmospheric corrosion scenarios. Finally, the technique developed was utilized to explore the effect of precipitation on the cathodic behavior of SS304L in dilute MgCl2. During cathodic polarization, evidence supports previous observations that magnesium hydroxide species are kinetically favored over the thermodynamically predicted magnesium carbonate.