Development of A Machine Learning Based Framework for Well Spacing Optimization in Unconventional Oil and Gas Production
Abstract
Efficient and cost-effective recovery of unconventional oil and gas (UOG) resources depends critically on the knowledge of primary factors controlling the reservoir producing behaviors, as well as on well completion strategies and well spacing. The conventional workflow entails fracture network simulation, followed by coupled reservoir and geomechanics simulation, which is computationally demanding. The primary goal of this study is to identify geological factors and well completion strategies important to production using Design of Experiment (DoE) methodologies, and then train a data-driven, machine-learning (ML) proxy model to expedite optimization of well infilling, with a special emphasis on predicting and minimizing the adverse impact of parent-child well interactions, referring to the fracture interference between existing production wells (parent wells) and infill wells (child wells). We will demonstrate the efficacy and accuracy of the developed framework using a realistic field dataset. Our base model is a multi-stage UOG well history-matched using production data. Coupled reservoir and geomechanical simulations are performed to generate training and testing data for developing the ML proxy model.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H15O1222Y