Log Evaluation Methods For Shale Brittleness And Rock Properties Estimation Using Mechanical Characterization And Advanced Statistical Analysis

Verifiable and accurate prediction of reservoir rock properties from geophysical logging data in a well with no previous exposure of the core data is the bottom line for any technique that claims the rock properties estimation capabilities.Despite a large amount of information available from core measurements technique,it is not commonly used because of its high cost.Core analysis data,however,is only available from limited wells in a field,while majority of the wells are logged.This dissertation propose a logging based methods for shale brittleness and fundamental rock properties estimation in different reservoirs.The main focus of this dissertation is to quantify brittleness based on the combination of fundamental rock properties and mechanical characterization.This dissertation also looks into the role and limitations of empirical models,multiple variable regression,and artificial neural networks technique for the estimation of shear wave velocity,permeability,and pore size classification in a highly heterogeneous reservoir and describes an improved methodology for accurate prediction of reservoir rock properties.Understanding brittle and ductile behavior in low permeability shale formation is essential for optimizing the completion and stimulation treatment in shale play.For the quantification of brittleness in shale,several methods for estimating brittleness index?BI?have been proposed in the literature.However,uncertainties still exist in the interpretation of fracability to decide where to place perforation clusters,what magnitude of fracability index mean to hydraulic fracture initiation and propagation,and how brittleness is best defined.This study aims to clarify these unresolved issues by introducing an improved fracturing index?FI?model after modifying Yuan et al.?2017?fracability-evaluation model(F_rac).Comparing with F_rac,the improved FI accurately predicts the brittle and ductile regions and provides a comparatively better idea about favorable fracturing sweet spots at TOC rich zones.In this study,I designed brittleness templates based on the correlation of fundamental rock properties and mechanical characterization for differentiating the brittle and ductile region in shale gas reservoir using specialized logging tools.The effectiveness of designed brittleness templates was verified through improved FI evaluation model and pre-existing fractures.The obtained result revealed that the points of high FI fall exactly into the predicted brittle regions and the points of low FI fall exactly into the predicted ductile regions.The applicability of the designed brittleness templates was further verified through the data from offset Well-2,laboratory testing of several outcrop samples,and shale sample from different origin with varying composition.In this study,I also propose a technique for the identification and characterization of naturally fractured shale gas reservoir using conventional and specialized logging tools.The potential and reliability of the developed models were verified through improved FI model,mechanical properties,FMI?Formation Micro Image?logging,and elastic constant.Permeability,porosity,and pore size classification are fundamental characteristics of reservoir rock that are naturally distributed in a non-linear and varying behavior.Sawan Gas Field is one of the most promising Gas Field in Pakistan with a cumulative production of850BCF?billion cubic feet?.The repetition of coarse sand,medium sand,and sand shale intercalation in the production zone make extreme heterogeneous,consequently permeability vary enormously?from 0.01 mD to more than1000 mD?.As a result,verifiable and accurate prediction of permeability,porosity,pore size classification,and shear wave velocity in the production zone with no previous exposure of laboratory-measured data is considered a challenging task.In the second half of this dissertation,I explore a methodology for improving the estimation of shear wave velocity,permeability,and pore size classification based on the combination of clustering,classification,and regression using logs data.This methodology work in two steps,First,I compute shear wave velocity,permeability,and pore size classification using more sophisticated methods proposed in the literature for comparing the reliability of classification model.Second,the clustering algorithm was employed over the data to train classification model for the identification of electrofacies?EF?and hydraulic flow units?HFU?and regression was employed from each distinct EF and HFU class zone for qualitative estimation of shear wave velocity,permeability,and pore size classification.To validate the effectiveness of this classification model,the proposed framework along with sophisticated methods were employed to the offset wells and the results were compared with actual data.The final results revealed that the proposed approach,which combines data mining task of clustering,classification,and regression,led to more uniform and accurate predictions of shear wave velocity,permeability,and pore size classification in comparison with the use of stand-alone rock physics model,virtual intelligence,and statistical regression.