Sensitivity Analysis of Key Parameters of Shale Oil Production Based on XGBoost Algorithm

Shale oil, a significant unconventional oil and gas resource, entails numerous intricate geological and technical parameters in the extraction process. The sensitivity of these parameters is inextricably linked to the extraction efficiency and economic returns. Initially, this paper delineates the backdrop of shale oil extraction and the significance of parameter sensitivity analysis. Subsequently, the sensitivity analysis approach grounded in the XGBoost algorithm is comprehensively expounded. The dynamic alterations of shale oil reservoirs are constantly monitored. In accordance with the laws of production decline and formation pressure variations, the sensitivity of crucial parameters during extraction is analyzed to enhance the ultimate recovery rate of shale oil. This paper emulates the shale oil extraction environment, sets the learning rate at 0.1, and conducts tests and analyzes the alterations in the characteristic mutation parameters and local sensitive parameters of shale oil during extraction. The experimental outcomes demonstrate that the sensitivity of the fracturing fluid volume is high. Within a certain scope, the XGBoost algorithm attains more favorable production enhancement by augmenting the fracturing fluid volume and improving porosity compared to traditional extraction methods. Thus, in the actual extraction process, the parameters of the XGBoost algorithm can be rationally adjusted within the sensitive range of the parameters to further boost production.