Atmospheric stress corrosion crack growth rates of 316 L stainless steel for nuclear waste containment

Highlights

• In situ high resolution (10⁻¹³ m/s) crack growth measurements of atmospheric SCC established.

• Effect of humidity under MgCl₂ deposits at 40 °C on crack growth rates evaluated.

• Drying and rewetting represents a severe exposure condition for SCC.

• Crack growth rate decayed with increasing humidity and crack depth.

Abstract

Atmospheric stress corrosion cracking (SCC) tests were conducted for a 316 L stainless steel (SS) at an applied stress of 1.1 σ_0.2 with a chloride deposition density of approximately 100 μg/cm² deposited as MgCl₂ from solution. The test specimens were exposed in an atmospheric chamber at a constant temperature of 40 °C with initial relative humidity (RH) of 40 %, followed by an increase to 60 % RH, then 80 % RH, and finally a decrease to 60 % RH. The total exposure period was 6500 h (about 9 months). The stress corrosion crack growth rate decreased with increasing RH and with increasing crack depth.

Introduction

Prior to the operation of a geological disposal facility (GDF), the UK’s intermediate level radioactive waste (ILW) is currently housed in interim surface storage facilities [1]. Generally ILW is encapsulated in a cement grout inside thin-walled (2.3 mm–6 mm thickness) containers which are commonly fabricated from 316 L SS [2,3]. It is important to ensure the integrity of the containers during this interim storage phase, which is envisaged to be at least for several decades [2], as well as during the subsequent operational phase of the GDF, which may also last several decades. The interim storage facilities in the UK, including those possible locations close to a marine environment, are generally fitted with high efficiency particulate inlet filters to reduce the ingress of possible external contaminants [4]. In addition, surfaces within the stores are monitored by an established swab methodology with a commercially available salt contamination meter to detect the ingress of contaminants [5]. While many stores may have excellent control of airborne species, information on every store is not currently available. Further, fault scenarios (where filters or facility barriers fail) cannot be ruled out over the storage lifetime. It is prudent to proactively establish bounding conditions within which integrity problems may arise.

Environmental data are collected and monitored from some larger, shielded, ILW stores and inland warehouses [[4], [5], [6]]. For naturally ventilated inland warehouse locations, temperature fluctuations are often found to vary between 5 °C–20 °C with an average temperature often found around 15 °C, generally following that of the external environment [4]. For larger well-shielded stores, temperatures are expected to be slightly lower, with less variation due to the greater thermal inertia of the building. The waste containers are non-heat generating; the activity of material within them is too low. Therefore, the temperature of the metal canister is reasonably close to the warehouse temperature. Relative humidity (RH) variations for the inland warehouses are generally found between 40 %–90 % with a yearly average close to 60 % [4,5]. Deposited aerosol species measured from surfaces representative of ILW container surfaces in inland warehouses suggest over the monitor period of about a decade, the presence of prevalent species including chloride, sulphate, nitrate and carbonate with deposit densities generally between 0.1 μg/cm² to 10 μg/cm² on horizonal surfaces [4]. Similar deposit levels were reported for cation species, with sodium and calcium being most prevalent from 1 to 10 μg/cm², and potassium and magnesium generally less than 1 μg/cm² during the same monitoring period [4]. The deposit level of species in ILW stores with ventilation inlet filters is typically expected to be lower than those reported for inland warehouse data [6].

However, there is a concern that the accumulative deposition of chloride-containing aerosols over the decades of storage would lead to localised corrosion developing on the SS containers, which in the presence of residual stress, e.g. at welds, could potentially lead to stress corrosion cracking. The goal of this present work is to refine understanding of the durability of ILW containers in chloride-containing atmospheres by obtaining quantitative data for stress corrosion crack growth rates as a function of RH under well-defined temperature and salt deposit conditions. The testing conditions adopted in this programme are designed to be highly aggressive relative to the real storage environments, making it more likely that cracking will occur and the effectiveness of the test method for crack growth rate measurement demonstrated. Future tests under more realistic, but feasible worst case deposition chemistry would then be used to inform inputs into the Atmospheric Corrosion of Stainless Steel in Stores (ACSIS) model under progressive development, which is designed to help evaluate the possibility of localised corrosion and stress corrosion cracking over long periods of exposure [7].

Specifically, the crack growth rate in 316 L SS specimens with a chloride deposition density of 100 μg/cm², deposited as MgCl₂ from solution, was measured in situ by the pulsed Direct Current Potential Drop (DCPD) method. This technique involves passing a constant current through the specimen and measuring the increase in potential drop generated by the increased resistance to current flow as the crack grows. The exposure condition is maintained at a constant temperature of 40 °C and varying RH conditions for a total exposure period of 9 months. An initial 40 % of RH is adopted to provide an aggressive condition of high [Cl⁻] to increase the possibility of cracking. Once the cracking is initiated and developed to a ‘quasi-steady’ growth state, the RH is subsequently increased to 60 % and then to 80 %, these RH values being closer to realistic conditions. A final decrease of RH from 80 % to 60 % is undertaken to assess the extent to which the crack growth rate recovers to the previous value at 60 % or whether exposure to 80 % RH causes the crack to arrest. MgCl₂ deposits with relatively high CDDs were chosen for these initial tests to enhance the likelihood of developing measurable SCC growth and to evaluate the efficacy of the testing methodology.

In most cases, crack precursors, i.e. pre-pits and pre-cracks, at varying depths were produced on specimens prior to the SCC testing following a methodology successfully deployed previously [8] and consistent with ISO 21153 [9]. The use of a crack precursor has the benefit of providing a known initial crack location, size and geometry (assuming the stress corrosion cracks would propagate from the crack precursors), optimal for DCPD probe location and data analysis. In practice, this method proved to be relatively ineffective in generating stress corrosion cracks for this application, with the cracks initiating preferentially from new pits formed during the test exposure, rather than from the crack precursors. Challenges in testing and data analysis are described and the calculated crack growth data from successful tests are discussed in context with published data compiled at Sandia National Laboratory [10].