Automatic Source Rock Evaluation From Well Logs for Play Analysis

Source and hydrocarbon migration continues to be one of the major uncertainties when exploring for hydrocar- bons. In regions where high-quality source rock is less prevalent, the automatic evaluation of source rock quality and quantity can serve as a play opener and an important derisking tool. In modern petroleum exploration within conventional reservoirs, charge has been shown to be the primary failure mechanism in almost half of all dry exploration wells on the Norwegian Continental Shelf (NCS) (NPD, 2020). Although advances in quantitative interpretation and geological insight facilitate sand detection and mod- ern imaging techniques simplify trap identification, the risk of failure can be further mitigated by addressing the remaining uncertainty associated with petroleum charge. The precision of assessing whether a reservoir contains hydrocarbons can be significantly improved by more re- liable identification of the source rock that generates the fluids that subsequently migrate to the traps. The automated workflow presented is a deterministic ap- proach, going from raw log data to Total Organic Carbon (TOC) predictions. The approach is particularly advan- tageous for quantifying the organic content in moderate- quality source rocks that cannot be identified by gamma ray logs alone. These moderate quality source rocks, as illustrated in Figure 1, may serve as the primary source of hydrocarbons in areas where more regionally robust source rocks are absent or immature. The methodology has been applied to more than 1000 wells on the NCS. Compared to reliable measurement of TOC, the method performs well throughout the North Sea, Norwegian Sea, and Barents Sea. The result of au- tomatic source rock detection has been used to guide the location of a recently drilled exploration well.