Abstract
A novel workflow is presented for integrating fiber optic Distributed Temperature Sensor (DTS) data in numerical simulation model for the Cyclic Steam Stimulation (CSS) process, using an intelligent optimization routine that automatically learns and improves from experience. As the steam-oil relationship is the main driver for forecasting and decision-making in thermal recovery operations, knowledge of downhole steam distribution across the well over time can optimize injection and production. This study uses actual field data from a CSS operation in a heavy oil field in California, and the value of integrating DTS in the history matching process is illustrated as it allows the steam distribution to be accurately estimated along the entire length of the well. The workflow enables the simultaneous history match of water, oil, and temperature profiles, while capturing the reservoir heterogeneity and the actual physics of the injection process, and ultimately reducing the uncertainty in the predictive models. A novel stepwise grid-refinement approach coupled with an evolutionary optimization algorithm was implemented to improve computational efficiency and predictive accuracy. DTS surveillance also made it possible to detect a thermal communication event due to steam channeling in real-time, and even assess the effectiveness of the remedial workover to resolve it, demonstrating the value of continuous fiber optic monitoring.