A modified fuzzy inference system for estimation of the static rock elastic properties: A case study from the Kangan and Dalan gas reservoirs, South Pars gas field, the Persian Gulf

Abstract

Rock deformation parameters including Young's modulus, Poisson's ratio, shear modulus and bulk modulus (E, υ, K and G, respectively) are important geomechanical properties that have many applications in wellbore stability studies. In this study, rock mechanical properties were predicted from well log data by using an improved Fuzzy Inference System (FIS) in the Kangan and Dalan supergiant gas reservoirs, South Pars Gas Field. Using the results of laboratory tests on 33 core samples (static value) and formation wave velocity outcomes (dynamic value) empirical correlations were developed between them. Static rock deformation parameters were calculated using the empirical correlations in all logged intervals. In the first stage of estimating rock deformation modules in un-cored wells, four individual fuzzy models were generated. The correlation coefficients between predicted and calculated E, υ, K and G in the test data were 0.89, 0.49, 0.94 and 0.82, respectively. In an effort to achieve further prediction accuracy, a larger fuzzy model which can simultaneously estimate all parameters was generated. The new model is actually an integration of four previous models in which the outputs of each individual fuzzy model besides well log data are considered as its inputs. Using such an improved fuzzy model the correlation coefficients between predicted and calculated K, E, υ and G data in the testing samples increased to 0.941, 0.993, 0.983 and 0.999, respectively. The results of this study show that there is a good agreement between real and fuzzy-logic derived rock deformation parameters. Fuzzy logic is successful in estimation of static mechanical properties from well logs in un-cored wells where laboratory experiments and shear wave velocity data are missing.