Integrating neural network and numerical simulation for production performance prediction of low permeability reservoir
Date
2005
Authors
Yang, Qingjun
Zhang, Shulin
Fei, Qi
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Journal article
Citation
Petroleum Science and Technology, 2005; 23 (5-6):579-590
Statement of Responsibility
Qingjun Yang, Shulin Zhang and Qi Fei
Conference Name
Abstract
It is difficult to predict the production performance of low permeability fractured oil reservoirs. This is because complicated factors such as geological and engineering factors affect well production performance. This paper presents a methodology to predict well production performance in the Hanq oil field, which is a low permeability fractured reservoir. Integration of neural network with numerical simulation is employed. First we study the regularity of fluid flow and oil displacement mechanism by injection well group numerical simulation and analysis of production performance. Then we form the expert knowledge affecting production performance. The neural networks based on expert knowledge are trained using production data. This method will play an important role in future waterflood management and the design of recovery strategy for the Hanq oil field.
School/Discipline
Australian School of Petroleum