Production-Data Analysis—Challenges, Pitfalls, Diagnostics

Summary The analysis of production data to determine reservoir characteristics, completion effectiveness, and hydrocarbons in place has become very popular in recent years. Although production analysis (PA) for reservoir characterization is approaching the popularity of pressure-transient analysis (PTA), there are few consistent diagnostic methods in practice for the analysis of production data. Many of the diagnostic methods for production-data analysis are little more than observation-based approaches—and some are essentially rules of thumb. In this work, we provide guidelines for the analysis of production data, as well as identify common pitfalls and challenges. Although PTA and production-data analyses have the same governing theory (and solutions), we must recognize that pressure transient data are acquired as part of a controlled experiment, performed as a specific event [e.g., a pressure-buildup (PBU) test]. In contrast, production data are generally considered to be surveillance/monitoring data—with little control and considerable variance occurring during the acquisition of the production data. We note that since both PA and PTA have the same governing relations, it is possible "in theory" that the same deliverables of PTA can be obtained using PA. This paper attempts to provide a state-of-the-technology review of current production-data-analysis techniques/tools—particularly tools to diagnose the reservoir model and assess the reservoir condition. The reservoir model is diagnosed mainly by examining the character exhibited by the data [that is the evidence of transient flow (e.g., quarter-slope might indicate a finite-conductivity fracture, or half-slope might indicate radial/pseudoradial flow)]. In addition, one can also assess the reservoir condition by inspecting the character of production data, which can confirm the evidence of boundary-dominated flow such that unit slope may indicate the boundary-dominated-flow regime and, therefore, in-place fluid volume can be estimated. This work also identifies the challenges and pitfalls of PA—and we try to provide guidance toward best practices and best tools. To complement this mission, we use relevant field examples to address specific issues, and we illustrate the value and function of production-data analysis for a wide range of reservoir types and properties. In this work, we propose the use of a sequence of raw and enhanced data plots for the diagnostic analysis of production data. We strongly believe that a comprehensive and systematic approach for production-data diagnosis has significant importance for the analysis and forecast of production performance.