The High Temperature Corrosivity of Radiolysed Nitric Acid Solution

Currently in the UK, spent nuclear fuel is reprocessed using the PUREX (Plutonium Uranium Reduction Extraction) process. This process generates large amounts of aqueous nitric acid based waste which is reduced in volume by evaporation before being stored in stainless steel tanks pending eventual disposal to a repository after conversion into a solid wasteform. The corrosivity of nitric acid solutions towards these stainless steel storage tanks is strongly affected by the presence of oxidants that can form in situ if certain dissolved metals such as cerium, chromium, ruthenium and neptunium are present, which is invariably the case in nuclear reprocessing plant liquors. Such liquors are, however, subject to radiolysis leading to the formation of nitrous acid and NOx species in equilibrium with nitric acid and water. The redox chemistry of irradiated reprocessing plant liquors is therefore complex, depending on a large number of factors including acidity, nitrate ion concentration, temperature, pressure, radiation dose rate and the nature/concentration of dissolved species. High acidities, high temperatures and low dose rates favour the oxidation of species such as Ce(III). For example, when Ce(IV) forms, the corrosion rate of stainless steel is strongly increased since the reduction of Ce(IV) forms a kinetically-favoured path way. Furthermore, the presence of nitrous acid (which is radiolytically formed from nitrate/nitric acid) can act to reduce potential corrosion accelerators (e.g. Ce(IV)) to their non-oxidising valency states. These dependencies are only semi-quantitatively understood at present, hampering useful prediction of actual effects when conditions are changed. The research presented within this thesis is divided between two experimental campaigns which are interrelated by their applicability to highly active storage tank conditions; I. An investigation into the conditions which effect the radiolytic production of nitrous acid in nitric acid based solutions was undertaken. This included the quantitative measurement of the steady state concentration of nitrous acid experienced under different conditions. The conditions investigated include temperature, dose rate, gaseous headspace and liquor composition in order to elucidate which factors are of importance in estimating the concentration of nitrous acid which can be expected at the base of a highly active storage tank. The major result of this campaign was that nitrous acid data collected could be used to formulate a g-value of nitrous acid formation (which was found to be 0.71) and this value was used to calculate the nitrous acid production rate expected within a highly active storage tank which is a pre-requisite of underpinning the corrosion chemistry within. II. Investigation into the potential formation of in situ corrosion accelerators in a reprocessing liquor simulant was undertaken. For this, a bespoke experimental rig has been designed, built and operated in order to identify the valency of potential corrosion accelerators at high temperatures while closely representing the conditions expected at the base of a highly active storage tank. This included the simulation of nitric acid radiolysis by means of an appropriate nitrite addition underpinned by the radiolysis studies described above in (I). It was found that none of the conditions investigated were oxidising enough to promote the generation of Ce(IV), which is contrary to the current understanding and should be favourable with regards to the storage tank remnant life expectations. Results reported in this thesis provide insight into the corrosivity of reprocessing liquors under representative storage tank conditions at various temperatures (up to the local liquor saturation temperature) and this knowledge will help improve remnant life predictions of the highly active storage tanks on the Sellafield site.