Unconventional resources play an important role in meeting the energy demand across the world, and their inclusion as a major asset class in Chevron's portfolio reflects this global reality. Another reality is that horizontal well development coupled with Intensive hydraulic fracturing is required to achieve commercial production from these ultra-low permeability shale formations. Moreover, the production performance of these wells is largely a function of the quality, properties, and interactions of the induced hydraulic fractures. We introduce an innovative and patent-pending digital workflow for measuring an operationally induced phenomena colloquially referred to as "stress shadow effect." We call this innovation the Fracture Recognition from ISIP in a Zone (FRITZ). Stress shadowing has been attributed to adverse operational and productivity results, especially in rapidly developing shale plays. Our new approach to measuring a proxy for stress shadow effect for an unconventional reservoir leverages a preponderance of horizontal well completion stage records. The FRITZ tool projects a regression-based trend between instantaneous shut-in pressure (ISIP) interpretations for each completion stage along a horizontal well bore. FRITZ then maps the slope coefficients for these ISIP by stage regressions in geographic space after numerous quality control checks. The coefficients represent the reservoir's (rock) ability to build stress from completion activity. Mapping these regional slope features makes a geographic representation for a proxy for stress shadow effect and how it changes regionally across a basin. The grid format output allows for utilization of this proxy data in numerous subsequent subsurface workflows including simulation and earth modeling and statistical modeling approaches, such as Chevron's ALICE machine learning workflow for recovery forecasts. We demonstrate FRITZ with a Permian Basin example that this stress shadow proxy is amongst the top predictors of well productivity when included in subsurface machine learning.
In 2020, fossil fuels supplied 81% of primary energy globally, and 12% of that fossil fuel came from unconventional reservoirs (US-IEA, 2021a). The total unconventional oil resource on Earth is approximately (2,200–9,300) × 108 t, which is up to 1.9 times that of conventional hydrocarbon systems (US-IEA, 2021a). This tremendous resource potential from unconventionals has made the asset class a primary focus for future development and investment. Unconventional plays feature reservoirs with hydrocarbons distributed throughout wide expanses of organic-rich, but low permeability geologic formations, requiring advanced extraction techniques. Geo-steered horizontal well designs, multi-stage well hydraulic fracture completions, and careful attention to spacing of infills is the epitome for unconventional reservoir development, and all are relatively costly endeavors for these generally low yield wells. Profitable production increasingly is dependent on continued optimization. Volatile markets since the early 2000s have forced producers to learn how to optimize the complex and expensive extraction techniques required for unconventional reservoirs to be competitive. Simply put, modern unconventional oil producers must quickly learn to adapt quickly to survive.
