Multi-disciplinary data integration for inverse hydraulic fracturing analysis: A case study

Hydraulic fracturing served as the principal technique to improve production in low permeability unconventional reservoirs in the last decade. Through core and outcrop studies, advanced logging tools, microseismic mapping and well testing analyses, it has further revealed the complexity of induced fracture network in the presence of natural fractures. Although most natural fractures are cemented by precipitations due to diagenesis, they can be reactivated during fracturing treatments and serve as preferential paths for fracture growth. However, current technologies for post-treatment assessment are incapable of accurately determine fracture geometry or even estimating the distribution of pre-existing natural fractures. Despite significant advances in numerical modelling of the problem, these models require an accurate description of natural fractures, which is often unknown to operators. Moreover, these numerical modeling techniques usually do not incorporate post-treatment data to reflect actual reservoir characteristics. This research proposes an innovative data integration workflow to estimate the characteristics of natural fractures based on formation evaluations, microseismic data, treatment data and production history. Least- square modeling approach is utilized to define possible realizations of natural fractures from selected double-couple microseismic events. Forward modeling incorporating Discrete Fracture Network will subsequently be used for matching treatment data and screening generated fracture realizations. Reservoir simulation tools will also be used thereafter to match the production data to further evaluate the fitness of natural fracture realizations. This workflow is able to integrate data from multiple aspects of the reservoir development process, and results from this workflow will provide both geologist and reservoir engineers an assessment tool for evaluating and modeling naturally fractured reservoirs.