Challenges for accurately predicting the solubilities of Hg0 in various solvents and their impact on modelling the distribution of Hg0 throughout oil and gas infrastructure

Abstract

Oil and gas operators need to be able to understand how elemental mercury (Hg⁰) is distributed throughout their infrastructure to be able to minimise the cycling of Hg⁰ through their process systems. This is essential to be able to avoid Hg-induced corrosion, embrittlement and to guarantee the safety of their workers. However, determining Hg concentrations directly at high temperatures or pressures is difficult. An alternative is to model the mobility of Hg⁰ by first experimentally determining equilibrium solubilities in relevant solvent mixtures under manageable conditions. The van’t Hoff approximation has often been applied to monitor the quality of produced experimental data. This is possible by plotting the equilibrium solubility against temperature producing an exponential relationship. If the fit is poor, it may suggest an error occurred in the experimental protocol, otherwise the data can be used to predict Hg solubilities at extreme temperatures. Here we present Hg solubility data in various solvents, including methanol, ethylene glycol (MEG), triethylene glycol (TEG), hexane, methylcyclopentane and water, with the aim of highlighting issues extrapolating to high temperatures based on laboratory measurements. To do this we compared the van’t Hoff plot (exponential fit) with a second approach (quadratic fit). This showed that when extrapolating to industrially relevant temperatures, for solvents such as MEG, the Hg⁰ solubility can be 7.8 times higher when applying an exponential fit compared to the quadratic fit. Showcasing the potential for significant error when attempting to model the persistence of Hg across petrochemical infrastructure.