Assessment of Reservoir-Souring Potential of a Sour Middle Eastern Field Using a Quantitative and Predictive Approach

Abstract Souring potential of fields during depletion and waterflood developments needs to be investigated ahead to enable making sensible decisions on injection-water source, well material selection, and facility design. This paper presents the application of a novel reservoir-souring approach to assess the souring potential of a Middle Eastern field (B) that comprises two reservoirs (the "U" reservoir with a strong aquifer pressure support, and the "M" reservoir with negligible aquifer pressure support. Both reservoirs produced sour fluids before the start of water injection, although the H2S levels in the U reservoir are much higher than those in the M reservoir. Because there is no sulfur-isotope data to determine the origin of the produced H2S, four possible hypotheses were proposed: (I) the reservoirs were initially sweet and bacterial-sulfate reduction (BSR) was triggered with drilling fluids; (II) the reservoirs were initially sour because of non-microbial geological sources and BSR has occurred since the start of waterflooding; (III) the same as hypothesis II but BSR has happened since drilling, and (V) the reservoirs were originally sour because of bacterially mediated sulfate reduction during early diagenesis. Each of these plausible hypotheses were modeled to history match the historical H2S and to predict future souring levels. A comparison between the simulated and historical H2S concentrations indicates that these two out of the four proposed hypotheses are more likely to explain the origin of the produced H2S, viz., (I) the reservoirs were initially sweet, but BSR was triggered with drilling and continued with production given that all the reservoir and environmental parameters were in favor of microbial activities, or else (II) the reservoirs were originally sour because of a non-microbial source, and then BSR has been occurring with water injection. However, because the H2S-abstraction capacity only plays a role in hypothesis (I), sensitivity analysis was performed on this hypothesis. The different H2S levels in the U and M reservoirs as well as in different wells can be attributed to different H2S abstraction capacities of the reservoir rock. Sensitivity runs reveal that for the U reservoir that waterflooding is not effective, the nutrients content of the formation water and the H2S-abstraction capacity have significant impacts on the max H2S concentration. However, for the M reservoir with waterflooding, the sulfate content of the injection water and the injection rate in addition to the aforementioned parameters have significant effects on the max H2S concentration.