公开(公告)号：US20230323772A1

公开(公告)日：2023-10-12

申请号：US17719022

申请日：2022-04-12

Applicant: CHEVRON U.S.A. INC.

Inventor： Changdong YANG ,  Jincong HE ,  Tsubasa ONISHI ,  Ronglei ZHANG ,  Yanbin ZHANG ,  Song DU ,  Zhenzhen WANG ,  Xiaoyue GUAN ,  Jianping CHEN ,  Xian-Huan WEN ,  Babafemi Anthony OGUNYOMI

IPC: E21B49/08 ,  G06F30/27

CPC classification number: E21B49/087 ,  E21B2200/20 ,  G06F30/27

Abstract: Embodiments for generating a reservoir performance forecast are provided. The embodiments may be executed by a computer system. In one embodiment, a method includes obtaining inflow performance relationship data generated from a physics-based subsurface-surface coupled simulation model having a surface, a subsurface, and one or more wells fluidly connecting the subsurface to the surface. The inflow performance relationship data comprises performance data for at least one phase of fluid for each well. The method also includes generating a performance forecast for the reservoir using a subsurface simulator and a surface simulator. The subsurface simulator uses the inflow performance relationship data to represent the subsurface during generation of the performance forecast, and the performance forecast satisfies constraints solved by the surface simulator. In one embodiment, a method does not utilize a surface simulator.

公开(公告)号：US20220164657A1

公开(公告)日：2022-05-26

申请号：US17534076

申请日：2021-11-23

Applicant: Chevron U.S.A. Inc.

Inventor： Jincong HE ,  Xian-Huan WEN ,  Meng TANG ,  Chaoshun HU ,  Shusei TANAKA ,  Kainan WANG

IPC: G06N3/08 ,  G06N3/04

Abstract: Embodiments of generating a field development plan for a hydrocarbon field development are provided herein. One embodiment comprises generating a plurality of training reservoir models of varying values of input channels of a reservoir template; normalizing the varying values of the input channels to generate normalized values of the input channels; constructing a policy neural network and a value neural network that project a state represented by the normalized values of the input channels to a field development action and a value of the state respectively; and training the policy neural network and the value neural network using deep reinforcement learning on the plurality of training reservoir models with a reservoir simulator as an environment such that the policy neural network generates a field development plan. A field development plan may be generated for a target reservoir on the reservoir template using the trained policy network and the reservoir simulator.