The frequent occurrence of hydrogen sulfide (H2S) in oil-producing wells is becoming increasingly problematic for Bakken petroleum system (Bakken) operators and midstream service providers. Understanding the mechanisms of H2S generation might reduce the risk of souring in new wells and enable more effective mitigation and management strategies in the wells already producing sour gas. Multiple known mechanisms can potentially lead to H2S occurrence in the oil production stream of Bakken wells, including 1) thermochemical sulfate reduction (TSR), 2) bacterial sulfate reduction (BSR), 3) source rock generation, 4) migration from deeper or shallower H2S-containing reservoirs, and 5) frac fluid chemical reactivity. Determining the exact causes of souring in the Bakken is challenging because of the apparent randomness of occurrence, unclear geographical and temporal trends, and potential linkages to recent well completion practices. Crude oil produced from the Middle Bakken (MB) and Three Forks (TF) reservoirs of the Williston Basin has historically been light (API gravity of 40°-50°) with low sulfur content (<0.2 wt%), typically exhibiting nonexistent to low (<0.01%) concentrations of H2S in the production stream. Unlike the formations overlying and underlying the Bakken, such as the Madison, Interlake, and Red River, neither the MB nor TF reservoirs have public records indicating the presence or shows of H2S during drill stem tests (DSTs). However, the frequency of H2S occurrence in production (souring) from MB and TF wells has increased since 2011. While the observed increases correlate with changing completion practices in the Bakken toward larger, multistage completions, the exact mechanisms responsible for the increase are an open research question. A 2023 Energy & Environmental Research Center study showed that the δ34S values of Mission Canyon source rocks closely match some but not all H2S of Bakken-produced samples. The study suggested that Mission Canyon organic-rich intervals could be a source of H2S for a limited number of Bakken producing wells via communication between reservoirs caused by out-of-zone stimulation work.
The present work aimed to provide insight into potential mechanisms using a combination of laboratory pyrolysis and exposure experiments, sulfur isotope analysis, and wellsite measurements. Unlike previous local efforts, the present work covered the broader Bakken and focused on five possible mechanisms of H2S generation.
Samples of H2S collected from wells demonstrated relatively wide variations of δ34S. However, the samples all have positive δ34S values ranging from +1.7 to +22.5%o, supporting nonbacterial sources. The bacterial mechanism of souring was refuted because of the harsh reservoir conditions, which are less favorable for sulfate-reducing bacteria growth, and the absence of sulfate-reducing bacteria in the produced brine indicated by screening-level biological activity reaction test (BART). The data also did not support the TSR process because of the different sulfur isotope signatures in the TF anhydrite − a potential source of sulfate needed for the TSR reaction and well-produced H2S. In this study, multiple laboratory experiments simulating TSR in the presence of TF anhydrite, Bakken oil, brine, HCl acid, and frac fluid did not generate detectable concentrations of H2S. The negative results of laboratory simulation might not be surprising because of the slow kinetics of the TSR reaction. Unlike BSR, the TSR process is relatively slow and may take thousands of years. The presence of HCl acid − a component well stimulation treatment in the system − might accelerate the reaction; however, it has yet to be proven by laboratory experiments and field observations.
The δ34S in the H2S from some producing wells matched the δ34S of the H2S generated by the pyrolysis experiments of the Mission Canyon source rocks of the Madison Formation, which overlies the Bakken reservoir. The sour nature of the Madison petroleum system is confirmed directly by the H2S and indirectly in DSTs reported by the North Dakota Geological Survey and higher sulfur content in oils produced from the Madison reservoir. These results suggest that the Madison Formation can hypothetically be a source of H2S in Bakken wells. Bakken production was sweet at the time of its discovery and initial production. However, applying high-energy multistage well completions in the Bakken reservoir could result in out-of-zone completion and the creation of artificial communication between the Bakken and sour reservoirs above or/and below the target zone. The source of H2S below the Bakken reservoir has not been identified. However, several petroleum systems located stratigraphically below the Bakken could be a source of H2S migrated upward, including the Devonian Birdbear and Duperow, Silurian Interlake, and Ordovician Red River. There is evidence indicating that the deeper Interlake (Silurian) and Red River (Ordovician) reservoirs are sour and produce high concentrations of H2S at a level of hundreds and thousands of parts per million. In conclusion, this study suggests that the sources of H2S are likely outside the Bakken reservoir, stratigraphically above or below the target producing zone.