Fault and non-fault areas detection based on seismic data through min/max autocorrelation factors and fuzzy classification

Accurate detection of faulted and non-faulted areas is significant step in oil and gas exploration and production. Different methods such as Discrete Fracture Network Detection, seismic attribute study, ant tracking, and meta-attributes are used in fault detection in seismic data. The method proposed in this research is a geostatistical approach which is based on combination of Minimum and maximum Autocorrelation Factor (MAF) and fuzzy logic applied on a set of seismic attributes. It is common in advanced seismic data interpretation to have multi variables of interest (seismic attributes), those which are spatially correlated. MAF approach is a geostatistical technique to obtain uncorrelated attributes by modifying the coordination axis in two steps and reducing the dimensions without losing the information. In order to develop fuzzy logic method to predict the faulted and non-faulted areas based on combination of normalized factors, the low value of faults and fractures in the seismic attributes is eliminated and attributes which have common point on the faulted or non-faulted areas, are superimposed on a fuzzy system and introduced as faults. According to the size of membership degree between symptoms and causations to detect and eliminate faults of gas turbine. The results have shown this fuzzy mathematics method has reliable and suitable detection of faults containing gas and oil in actual and complex environment.