Application and design of fuzzy intelligent evaluation software for sand production and steam channeling prediction of steam injection well

Abstract

Fuzzy intelligent evaluation software is designed to predict situation of sand production and steam channeling for steam injection well, which can improve the development effect of heavy oil reservoir. Based on expert knowledge, the software transfers the qualitative knowledge into quantitative index by the use of fuzzy evaluation and grey relational analysis. Combined with actual data of oil field as sample set, the software establishes an intelligent optimization analysis model through neural network and numerical analysis methods. Through multi-model regression analysis, optimum model and its parameters can be chosen to predict the production effect. This software adopts three-tier model, with which the structure has the following advantages: reasonable design, reliable performance, convenient operation, high portability and good expansibility. According to the calculated results of actual data of oil field, the software can predict sand production and steam channeling more effectively, and guide further formulation and improvement of development program. © 2011 Published by Elsevier Ltd.