Tailor-Made Workflow to Understand and Forecast H2S Production Risks in Thermal Projects

Abstract

Abstract In-situ extraction of heavy sulfured oil based on steam injection comes with a high level of risk in terms of H2S production, resulting from aquathermolysis reactions. This could lead to on-site living being casualties, environment damage, surface facilities and wells corrosion. Also, there is a strong need to understand aquathermolysis reactions and to forecast acid gases generation in such context. For that aim a tailor-made workflow was developed to estimate H2S concentration at the well-head. To meet these challenges, a 3-steps approach combining laboratory studies and numerical predictions has been developed. It firstly consists in a fast preliminary assessment of the highest H2S risk areas, based on the measurements of sulfur characteristics of reservoir core samples using our Rock-Eval Sulfur set-up. Then aquathermolysis experiments from the previously selected core samples are conducted in order to calibrate the compositional chemical model of the reservoir simulator. The latter is eventually used to carry out thermal compositional reactive simulations at field scale. The reservoir simulator allows simulations of SAGD processes and handles H2S distribution over oil/water/gas phases and its migration in these phases. Simulation results show that acid gases are generated within the steam chamber, before they accumulate at the chamber edges where they dissolve in the water and oil phases. This contributes to reduce viscosity, allowing the oil to flow along the steam chamber edges before it has reached the steam temperature. Therefore the H2S produced at surface is mainly carried towards the wells by water and to a lesser extent oil. This flowing oil has not reacted with the steam: its composition is close to the initial reservoir oil but enriched with dissolved gases. The steam chamber shape, the temperature distribution and the H2S produced at surface are strongly modified when heterogeneities are introduced in the reservoir model. Synthetic cases allow a deeper understanding of the effects of heterogeneities. Vertical permeability is thus found to be a key factor of H2S production variations. When steam reaches a lower permeability lithology, a delayed rise in H2S production at wellhead is observed as aquathermolysis reactions rates increase. Finally Foam Assisted-SAGD has been considered. While the foam improves the Steam-Oil ratio, no clear improvement was observed regarding the H2S production.