Multifractal analysis of the heterogeneity of nanopores in tight reservoirs based on boosting machine learning algorithms

Exploring the geological factors that affect fluid flow has always been a hot topic. For tight reservoirs, pore structure and characteristics of different lithofacies reveal storage status fluids in reservoir environments. The size, connectivity, distribution fillers sedimentary environments have posed challenge studying microscopic heterogeneity. In this paper, six logging curves (gamma-ray, density, acoustic, compensated neutron, shallow resistivity, deep resistivity) two marker wells, namely, J1 J2, Permian Lucaogou Formation Jimsar Basin are tested by using four reinforcement learning algorithms: LogitBoost, GBM, XGBoost, KNN. total percent correct training well J2 is 96%, validation 75%, 68%, 72%, respectively. Based on classification obtained algorithm, micropores, mesopores, macropores comprehensively described high-pressure mercury injection low-pressure nitrogen gas adsorption tests. multifractal theory servers for quantitative characterization heterogeneity regarding samples, as observed, higher probability measure area generalized fractal spectrum affects local interval mesopores estuary dam. micropore mesopore, evaporation lake showed large variation due to influence area, mesopore macropore, was controlled lower area. According correlation analysis, single-fractal dimension related parameters, individual fitting degree reaches up 99%, which can serve characterizing size uniformity. combination boosting machine help better characterize micro-heterogeneity under ranges, helpful exploration development oil fields.