Neural networks models bring a significant dimension to the capability of computer aided approaches for well test analysis. This study shows, however, that when dealing with complex reservoirs, these models need augmentation from an expert system containing two types of information; independent field data and tables containing frequency of occurence of non-related models. A Monte Carlo Simulation method is proposed to provide the augmented statistical informations required for the expert system.

You can access this article if you purchase or spend a download.