A Production Performance Prediction and Field Development Design Tool for Coalbed Methane Reservoirs

Open Access
Rajput, Vaibhav Hiralal
Graduate Program:
Energy and Mineral Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
May 10, 2012
Committee Members:
  • Turgay Ertekin, Thesis Advisor
  • Coalbed Methane
  • Reservoir Simulation
  • Artificial Neural Networks
  • Field Development Optimization
ABSTRACT Ever growing energy demand has forced oil and gas exploration and production (E&P) companies to search and economically produce unconventional energy sources such as coalbed methane (CBM), shale gas, shale oil, tight sands etc. To produce these reservoirs in a more efficient manner, employment of horizontal wells seems to be one of the feasible options. However, the additional economic benefits due to increased production and additional drilling/operational costs need to be carefully calculated to access the overall benefits of horizontal wells in CBM reservoirs. Reservoir simulators offer a powerful tool to reservoir engineers to predict the production performance of CBM reservoirs in an accurate manner. However, full scale simulation runs take hours, or even days, to complete. Hence, it becomes practically cumbersome to explore a full suite of different design scenarios and operational behavior of a given reservoir within a reasonable period of time. It is here where Artificial Neural Networks (ANN) are used to develop expert systems which can accurately mimic the production performance of a given reservoir within fraction of seconds. Once properly trained and validated, the expert system is capable of not only predicting the production performance under a given set of design conditions, but also provide accurate information on design specifications in an inverse manner, thus going beyond the reservoir simulator’s capability. In this study, an additional inverse formulation was implemented in which reservoir characteristics were obtained once production data and design specifications were provided to the expert system. Finally, the expert system is then used for field development planning where design parameters such as drainage area, sandface pressure, stimulation factor (skin) and horizontal well length are predicted.