Author:
Rostami Alireza,Kordavani Ali,Parchekhari Shahin,Hemmati-Sarapardeh Abdolhossein,Helalizadeh Abbas
Abstract
AbstractThe need to determine permeability at different stages of evaluation, completion, optimization of Enhanced Oil Recovery (EOR) operations, and reservoir modeling and management is reflected. Therefore, various methods with distinct efficiency for the evaluation of permeability have been proposed by engineers and petroleum geologists. The oil industry uses acoustic and Nuclear Magnetic Resonance (NMR) loggings extensively to determine permeability quantitatively. However, because the number of available NMR logs is not enough and there is a significant difficulty in their interpreting and evaluation, the use of acoustic logs to determine the permeability has become very important. Direct, continuous, and in-reservoir condition estimation of permeability is a unique feature of the Stoneley waves analysis as an acoustic technique. In this study, five intelligent mathematical methods, including Adaptive Network-Based Fuzzy Inference System (ANFIS), Least-Square Support Vector Machine (LSSVM), Radial Basis Function Neural Network (RBFNN), Multi-Layer Perceptron Neural Network (MLPNN), and Committee Machine Intelligent System (CMIS), have been performed for calculating permeability in terms of Stoneley and shear waves travel-time, effective porosity, bulk density and lithological data in one of the naturally-fractured and low-porosity carbonate reservoirs located in the Southwest of Iran. Intelligent models have been improved with three popular optimization algorithms, including Coupled Simulated Annealing (CSA), Particle Swarm Optimization (PSO), and Genetic Algorithm (GA). Among the developed models, the CMIS is the most accurate intelligent model for permeability forecast as compared to the core permeability data with a determination coefficient (R2) of 0.87 and an average absolute deviation (AAD) of 3.7. Comparing the CMIS method with the NMR techniques (i.e., Timur-Coates and Schlumberger-Doll-Research (SDR)), the superiority of the Stoneley method is demonstrated. With this model, diverse types of fractures in carbonate formations can be easily identified. As a result, it can be claimed that the models presented in this study are of great value to petrophysicists and petroleum engineers working on reservoir simulation and well completion.
Publisher
Springer Science and Business Media LLC
Reference94 articles.
1. Brie, A., Endo, T., Johnson, D. & Pampuri, F. Quantitative formation permeability evaluation from Stoneley waves. SPE Reservoir Eval. Eng. 3, 109–117 (2000).
2. Hornby, B., Johnson, D., Winkler, K. & Plumb, R. Fracture evaluation using reflected Stoneley-wave arrivals. Geophysics 54, 1274–1288 (1989).
3. Brie, A., Hsu, K. & Eckersley, C. in SPWLA 29th Annual Logging Symposium. (Society of Petrophysicists and Well-Log Analysts).
4. Endo, T., Ito, H., Brie, A., Badri, M. & El Sheikh, M. in SPWLA 38th Annual Logging Symposium. (Society of Petrophysicists and Well-Log Analysts).
5. Tang, X. & Cheng, C. Fast inversion of formation permeability from Stoneley wave logs using a simplified Biot-Rosenbaum model. Geophysics 61, 639–645 (1996).
Cited by
8 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献