Using an expert system to identify the well test interpretation model

A new approach is presented to solve the problem of identifying the well test interpretation model that describes the behavior of a reservoir. The approach is based on the emerging technology of Artificial Intelligence. The authors have developed a Knowledge-Based System (Expert System) of carefully extracted rules and facts for buildup and drawdown test analysis. In this approach, the Knowledge-Based System is designed to use a description of the shapes of different graphs of test data, particularly the derivative plot. When needed, the system seeks information in addition to test data from all relevant sources such as production statistics, geological interpretations, well logs, fluid analyses and completion data. The approach presented is implemented on a microcomputer environment which consists of three major parts: an interactive well test analysis module, a knowledge base, and an automatic history-matching module.

Using an expert system to identify the well test interpretation model Conference Paper