Hydrogen sulfide (H2S) poses a perennial risk in the oil and gas industry. To better tackle potential H2S related issues, a systematic geochemical sampling, analysis, and modeling program has been implemented at ConocoPhillips’ Montney asset since the beginning of development. The program entails delineating reduced sulfur species, probing their sources, and developing predictive capabilities to enhance risk assessment and mitigation.
Core, cuttings and mud gas samples were collected during drilling. Produced fluid samples were collected over time once each well came online. Rock samples were extracted either with methylene chloride or carbon disulfide (CS2) for better preservation of the light end hydrocarbons. Gas samples were analyzed for bulk hydrocarbon and non-hydrocarbon composition as well as carbon isotopic composition. Total sulfur content, whole oil GC and GCMS analyses were performed on produced oil samples and rock extracts. A small subset of oil and gas samples were analyzed for sulfur isotopes. Machine learning models were developed to predict the total sulfur content in oil and the H2S content in gas, with data obtained from geochemical analyses. A kinetics model was derived to model the co-generation of sulfur bearing species from kerogen with hydrocarbons and the subsequent evolution during secondary cracking, from which the total sulfur content in oil and the H2S content in gas at different maturities were predicted.
Across ConocoPhillips’ Montney acreage, a strong positive correlation was observed between the H2S content in produced gas and the total sulfur content in produced oil. Leveraging this correlation, produced gas H2S content could be predicted based on produced oil total sulfur content. Machine learning models were developed to predict produced oil total sulfur content with data from geochemical analyses of cuttings collected during drilling, whose output was then used to predict the H2S content in produced gas accurately. The kinetics modeling was able to adequately capture the Upper Montney oil total sulfur content changes with changing thermal maturity, but significantly overpredicted the H2S content in gas produced from the Upper Montney. As per the machine learning model, kinetically modeled total sulfur content in oil was used to predict the H2S content in gas, enabling regional predrill H2S risk assessment at pad scale resolution. Current geochemical analysis and modeling results indicate that the Upper Montney itself is the primary source of H2S produced from it. Although a potent source rock for both hydrocarbons and H2S, the overlying Doig Formation is unlikely to have contributed substantial amounts of H2S to the Upper Montney in this study area.
A multifaceted geochemistry program has been successfully developed and implemented to delineate and predict H2S risk for an unconventional asset. When predrill H2S risk prediction is not readily attainable, prediction while-drilling, based on geochemical characteristics of fluids and rock samples, is a viable alternative, which delivers reliable estimates of reduced sulfur content of production streams prior to flowback and in some cases production facility installation.
