ABSTRACT
Well Test engineering is a complex function requiring both technical knowledge (facts) and experience (beliefs and heuristics). A large number of computer programs have been developed by E&P companies, service companies, consultants and software vendors to assist engineers and technicians in performing some of the tasks which are considered essential to successful well test engineering problem-solving. The distinguishing feature of a majority of these programs is their emphasis on technical features, rather than overall problem-solving capabilities. In general, as the level of program technical sophistication increases, the problem-solving efficiency of the users decreases, thus presenting a software selection dilemma to users and their managers. One solution to this dilemma is to combine the desired level of technical sophistication with problem-solving assistance in the form of procedures and constraints recommended by "experts" in the theory and application of well test engineering technology. This approach effectively places an "applications" expert over the shoulder of the program user, who is the real expert in the "behavior" of the wells and reservoirs being tested. The solution described in the paper evolves from the emerging science of artificial intelligence and is referred to as "expert systems" technology. The paper describes expert systems technology and its application to a well test engineering software package being developed through the combined efforts of an experienced well test engineer and a computer programming specialist. The concepts introduced enhance the utility of well testing by making expert well test engineering problem-solving capability available to non-expert engineers and technicians.
