Rate Transient Analysis Of Dual Lateral Wells In Naturally Fractured Reservoirs Via Artificial Intelligence

Open Access
Lu, Jia
Graduate Program:
Energy and Mineral Engineering
Master of Science
Document Type:
Master Thesis
Date of Defense:
April 10, 2015
Committee Members:
  • Turgay Ertekin, Thesis Advisor
  • Artificial Neural Network
  • ANN
  • Dual-Lateral
  • Double Porosity
  • Jia
  • Lu
  • Ertekin
In naturally fractured reservoirs, reservoir characterization is critical to the production of hydrocarbon, including but not limited to porosity, permeability, pay zone thickness and fracture spacing. Laboratory measurements, well-logging technique, and mathematical models are three major characterization approaches that are widely used to determine and analyze the reservoir characterization and production profiles. Amongst these approaches, mathematical models are commonly used as estimation tools. The purpose of this thesis is to develop a mathematical model as a reservoir estimation tool for naturally fractured reservoirs with dual lateral well configurations. The tool proposed in this study includes a forward artificial neural network (ANN) with the ability to predict production data via known reservoir and well design parameters. The proposed tool also includes an inverse ANN component that can be used to predict the permeability and porosity of matrix and fracture, as well as fracture spacing and reservoir thickness. By means of the proposed tool, the user would be able to analyze instantaneously predicted reservoir or production data with less cost and time. The software involved in developing the tool were MATLAB, EXCEL, and a commercial modeling software1. The procedures are introduced and discussed in the following chapters including training data generation, selecting training data sets, training forward and inverse ANN models. Moreover, a graphical user interface (GUI) is developed and assembled for each ANN, which allows the user to view results in numerical and graphical formats.