Fractures and their parameters are considered as being the most important features in evaluation of the fractured reservoirs. This may be of more importance in the heterogeneous reservoirs, when primary porosity cannot explain the high production capacity. Cores and image logs are two direct methods for fractures evaluation. However, these methods have high cost drawback. So, petrophysical logs are useful tools for fractures detection due to their low cost and accessibility in all wells. This paper proposes a reliable and inexpensive method for the evaluation of fracture parameters in the carbonate heterogeneous reservoir using the preprocessed petrophysical logs and ANFIS (Adaptive Neuro-Fuzzy Inference System). This study indicates that there are two vital limitations when using petrophysical logs for fractures evaluation; firstly, fracture distribution is a complicated process to be predicted by classical methods and none of intelligent systems can be considered as a magic for fracture parameters estimation from raw conventional logs. Secondly, fractures determination is not valuable alone, unless they are used for evaluation of porosity and permeability systems. To solve these problems, a log preprocessing method and a new statistical equation are implemented on the raw conventional logs. Also, ANFIS is used as a powerful method for fracture parameters estimation using preprocessed logs. It is shown that sonic and porosity logs are the best tools for fracture studies. Also, resistivity and gamma ray family logs may have a good evaluation of fractures in the fractured zones. Results confirm that conventional logs could be more useful tools for fractures evaluation if they preprocessed by some statistical methods and correlated by image logs or core data. Proposed method may also be usable for estimation of fractures aperture in some image tools which they are not traditionally able to measure fractures aperture (e.g., sonic image tools). Undoubtedly, fracture aperture is the most important parameter for determination of fractures effect on the porosity and permeability systems. Due to high correlation between petrophysical logs and images/cores results (R2 ≈ 0.8), the results are dependable and extensible for other carbonate and naturally fracture reservoirs.

Keywords