An advisory system for the development of unconventional gas reservoirs.

摘要

With the rapidly increasing demand for energy and the increasing prices for oil and gas, the role of unconventional gas reservoirs (UGRs) as energy sources is becoming more important throughout the world. Because of high risks and uncertainties associated with UGRs, their profitable development requires experts to be involved in the most critical development stages, such as drilling, completion, stimulation, and production. However, many companies operating UGRs lack this expertise. The advisory system we developed will help them make efficient decisions by providing insight from analogous basins that can be applied to the wells drilled in target basins.;In North America, UGRs have been in development for more than 50 years. The petroleum literature has thousands of papers describing best practices in management of these resources. If we can define the characteristics of the target basin anywhere in the world and find an analogous basin in North America, we should be able to study the best practices in the analogous basin or formation and provide the best practices to the operators.;In this research, we have built an advisory system that we call the Unconventional Gas Reservoir (UGR) Advisor. UGR Advisor incorporates three major modules: BASIN, PRISE and Drilling & Completion (D&C) Advisor. BASIN is used to identify the reference basin and formations in North America that are the best analogs to the target basin or formation. With these data, PRISE is used to estimate the technically recoverable gas volume in the target basin. Finally, by analogy with data from the reference formation, we use D&C Advisor to find the best practice for drilling and producing the target reservoir.;To create this module, we reviewed the literature and interviewed experts to gather the information required to determine best completion and stimulation practices as a function of reservoir properties. We used these best practices to build decision trees that allow the user to take an elementary data set and end up with a decision that honors the best practices. From the decision trees, we developed simple computer algorithms that streamline the process.