Structuring An Integrative Approach For Field Development Planning Using Artificial Intelligence And Its Application To Tombua Landana Asset In Angola

Open Access
Ketineni, Sarath Pavan
Graduate Program:
Petroleum and Mineral Engineering
Doctor of Philosophy
Document Type:
Date of Defense:
August 12, 2015
Committee Members:
  • Turgay Ertekin, Dissertation Advisor
  • Turgay Ertekin, Committee Chair
  • Eugene C Morgan, Committee Member
  • Sridhar Anandakrishnan, Committee Member
  • Yilin Wang, Committee Member
  • Luis F Ayala H, Committee Member
  • Reservoir Characterization
  • Artificial Neural Networks
  • Field Development Planning
  • Petroleum Engineering
  • Seismic data analysis
Field development studies are at the forefront of common engineering practices in petroleum industry to maximize the returns on a given asset. In early stages of reservoir depletion, it is often a challenging task to accurately determine reservoir properties that are representative of the actual field. Reservoir modeling is the traditional way that engineers performed to develop field development and depletion plans. Due to different scales of data obtained from various sources like seismic data, well logs, cores, and production data, there is a lot of uncertainty in solving the inverse problem of estimating formation rock and fluid properties from the field data. Increase in complexity of formations and scarcity of reservoir data have made reservoir characterization a challenging task. Soft computing techniques have gained popularity in petroleum industry to identify complex patterns that exist between various reservoir data collected from multiple sources and be able to successfully characterize a reservoir. In this work, a work-flow is developed for devising a comprehensive reservoir characterization tool based on artificial neural network. A case study of Chevron’s Tombua Landana Asset is used in demonstrating the tenets of the work-flow. The reservoir under consideration is highly heterogeneous in terms of property distribution and is believed to be highly channelized. The ANN based tool will assist in identifying sweet spots by predicting optimal well location/path/completion parameters and production schedule. The multilayer feed forward back propagation based neural network tool developed is able to capture the correlations that exist amongst seismic data, well logs, completion data, and production data. Well logs are correlated using surface seismic attributes and geometric location of wells with an average testing error of less than 15%. The range of testing errors is in between 1-30%. The tool enables the user to predict the entire well log suite for even a horizontal well of user defined configuration through a graphic user interface. Having correlated seismic data with well logs, synthetic well logs are generated for the entire area of seismic coverage. To predict production data, along with seismic data and well logs, schedule of production and interference factors are incorporated as functional links. Upon analyzing the relevancies of input data, functional links based on geographic location and injection wells are included to make the prediction more reliable and robust. Production performance networks comprising cumulative oil,gas and water production performance prediction modules are developed to forecast performance of wells at undrilled locations. Oil networks indicated an average error of 21% in blind testing cases. Highly variable gas production could also be correlated with the seismic data and well log data within 32% error. Water production networks indicated a high error of 46% on blind testing cases. Oil, gas and water production forecast maps are generated using production performance networks. Maps generated indicate flow paths that exist in the field.