Modelling Elemental Mercury Partitioning and Transport: A Case Study of Fenris: A New Offshore High Pressure, High Temperature Gas-Condensate Field

Abstract

Mercury is highly toxic and corrosive to certain metals and is therefore a highly undesirable contaminant in produced hydrocarbons. Its concentration in reservoir fluids differs by over four orders of magnitude globally, which means its operational consequences can differ enormously. We present a case study of the Fenris field: a new, offshore HPHT gas-condensate tieback to Valhall (North Sea) in which mercury has been detected but there remains considerable uncertainty about its abundance in the fluids. Predictions from a cubic Equation of State (EoS) model are strongly dependent on the parameter set chosen and there is not common agreement within the industry on the most suitable. Prior to field simulations, the suitability of commercially available models was evaluated by comparing outputs with literature data. Once a suitable EoS parameter set was selected, partitioning of Hg0 over all possible phases (gas, condensate, MEG-water and liquid Hg0) was evaluated for a variety of Hg0 concentrations (due to the uncertainty thereof), as well as the influence of conditions both in the subsea flowline and in the facilities. A plausibly conservative base case was selected for the Hg0 concentration in the reservoir fluids. This allowed partitioning and transport of Hg0 to be evaluated in terms of both the quantity and concentration of Hg0 in each produced fluid stream. Specifications for mercury-removal units (MRUs) were initially set using these values. A set of simulations performed using a higher Hg0 concentration allowed for evaluation of the suitability of these values under worst-case conditions. Considerable seasonal variation was anticipated, with the fluid arrival temperature at the facilities expected to fluctuate between 0 and 10 °C with related changes in the Hg partitioning. It was identified that the greatest quantity of liquid Hg0 was expected to form in the flowline and facilities at around Year 3 following First Gas, consistent with the maximum gas rate expected over field life and winter conditions. Of particular interest is the influence of the condensation and agglomeration kinetics of liquid Hg0, which may not only change the locations where the liquid accumulates but can also affect Hg0 partitioning into the other produced phases and can therefore affect the sizing of any MRUs to achieve product specification for this contaminant. This work describes the challenges in predicting the consequences of mercury production at FEED when its expected Hg0 concentration is significant-to-high but substantially uncertain. A conservative approach was taken in modelling quantities at various locations to ensure risk is suitably managed without adopting design specifications that unduly increase capital expenditure. The paper describes the predicted risks associated with Hg0 in this new development and the steps identified to manage risks during the upcoming production stage.