IP Library Granted Patent US 12,724,946
Granted Patent B2
US 12,724,946 · App. 17/905,707 · Granted Sep 1, 2026

Fast front tracking in EOR flooding simulation on coarse grids

Inventors: Barys Andreevich Samson (Piddington, GB); Sophie Nazik Godefroy (Abingdon, GB); Soham Sheth (Oxford, GB); Christopher Marooney (Abingdon, GB)
Assignee: Schlumberger Technology Corporation
G06F30/27E21B43/162E21B2200/20
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,724,946
App. No.
17/905,707
Granted
Sep 1, 2026
Kind
B2
Abstract

The present disclosure provides a workflow for modelling EOR flooding operations performed on a reservoir by separating front tracking from the reservoir simulation process, so that the front's position and topology evolves in parallel with the coarse grid simulation, through modifications using machine-learning-trained correlations.

Claims (66)

1 . A method of modeling or simulating enhanced-oil recovery (EOR) flooding operations performed on a reservoir, comprising:

collecting training data based on simulation of EOR flooding operations using a first grid having a relatively high resolution;

using the training data to configure and train a machine learning system;

performing simulation of EOR flooding operations using a second grid having a relatively low resolution than the relatively high resolution of the first grid; and

using output data generated by the simulation using the second grid as input to the machine learning system to predict at least one parameter characterizing a displacing agent front in the second grid for a time step in the simulation using the second grid, wherein the at least one parameter characterizing the displacing agent front comprises a front speed for the time step and the front speed is used during execution of the simulation using the second grid to adjust reservoir simulation parameters associated with at least one cell of the second grid, wherein the reservoir simulation parameters are used to compute a subsequent time step of the simulation using the second grid.

2 . A method according to claim 1 , wherein:

the first grid has a relatively fine resolution and the second grid has a relatively coarse resolution.

3 . A method according to claim 1 , wherein:

the machine learning system is trained to predict the front speed for the time step given at least one localized static property of the reservoir, at least one localized dynamic property of the reservoir at the time step provided by the simulation using the second grid, and at least one localized dynamic property of the reservoir at an earlier time step provided by the simulation using the second grid.

4 . A method according to claim 3 , wherein:

the at least one localized static property of the reservoir is selected from the group consisting of porosity and anisotropic permeabilities of the reservoir at a cell of the second grid.

5 . A method according to claim 3 , wherein:

the at least one localized dynamic property of the reservoir at the time step is selected from the group consisting of anisotropic pressure gradients, fluid saturations and displacing agent concentration at a cell of the second grid at the time step; and

the at least one localized dynamic property of the reservoir at the earlier time step is selected from the group consisting of anisotropic pressure gradients, fluid saturations and displacing agent concentration at the cell of the second grid at the earlier time step.

6 . A method according to claim 3 , wherein:

the machine learning system is trained to predict the front speed for the time step given front parameter data from an earlier time step.

7 . A method according to claim 6 , wherein:

the front parameter data represents at least one of position of the displacing agent front relative to center position of the group of cells, orientation of the front, or curvature of the front at the earlier time step.

8 . A method according to claim 1 , wherein:

the machine learning system is trained using label data representing front speed based on front parameter data derived from the simulation of EOR flooding operations using the first grid having a relatively high resolution.

9 . A method according to claim 1 , further comprising:

adjusting front position in the second grid based on recursive properties refinement to account for fine-scale geological features of the reservoir.

10 . A method according to claim 1 , further comprising:

combining predictions of front position for multiple cells of the second grid to produce a continuous front line.

11 . A method according to claim 1 , further comprising:

extracting coarse grid front speed for the time step from the simulation using the second grid; and

using the front speed for the time step as predicted by the machine learning system and the coarse grid front speed to update at least one localized static property of the reservoir.

12 . A method according to claim 11 , wherein:

the at least one localized static property of the reservoir comprises anisotropic permeability of the reservoir at one or more cells of the second grid.

13 . A method according to claim 1 , wherein:

the machine learning system is trained using a random forest algorithm.

14 . A method according to claim 1 , wherein:

the training data is based on at least one localized static property of the reservoir derived by upscaling data representing the at least one static property for a group of adjacent cells of the first grid, wherein the group of adjacent cells of the first grid is defined by a predefined stencil.

15 . A method according to claim 14 , wherein:

the at least one localized static property is selected from the group consisting of porosity and anisotropic permeabilities of the reservoir.

16 . A method according to claim 1 , wherein:

the training data is based on at least one localized dynamic property of the reservoir derived by upscaling data representing the at least one dynamic property for a group of adjacent cells of the first grid for a respective time steps of a pair of time steps, wherein the group of adjacent cells of the first grid is defined by a predefined stencil.

17 . A method according to claim 16 , wherein:

the at least one localized dynamic property is selected from the group consisting of pressure gradients, fluid saturations, and displacing agent concentration.

18 . A method according to claim 1 , which is performed by at least one processor.

19 . A method according to claim 1 , wherein:

the training data is based on output data of a reservoir simulator that is configured to simulate EOR flooding operations using the first grid having a relatively high resolution; and

the simulation using the second grid is performed by a reservoir simulator configured to simulate EOR flooding operations using the second grid having a relatively low resolution.

20 . A system for modeling enhanced-oil recovery (EOR) flooding operations performed on a reservoir, comprising:

at least one processor configured to:

perform simulation of enhanced-oil recovery (EOR) flooding operations using a first grid having a relatively low resolution; and

use output data generated by the simulation as input to a machine learning system to predict at least one parameter characterizing a displacing agent front in the first grid for a time step in the simulation, wherein the at least one parameter characterizing the displacing agent front comprises a front speed for the time step and the front speed is used during execution of the simulation to adjust reservoir simulation parameters associated with at least one cell of the first grid, wherein the reservoir simulation parameters are used to compute a subsequent time step of the simulation,

wherein the machine learning system is trained on training data collected from simulation of EOR flooding operations using a second grid having a relatively high resolution.

21 . A system according to claim 20 , wherein:

the machine learning system is trained to predict the front speed for the time step given at least one localized static property of the reservoir, localized dynamic properties of the reservoir at the time step provided by the simulation using the first grid, and localized dynamic properties of the reservoir at an earlier time step provided by the simulation using the first grid.

22 . A system according to claim 21 , wherein:

the machine learning system is trained to predict the front speed for the time step given front parameter data from an earlier time step.

23 . A system according to claim 22 , wherein:

the front parameter data represents at least one of position of the front relative to center position of the group of cells, orientation of the front, or curvature of the front at the earlier time step.

24 . A system according to claim 20 , further comprising:

extracting coarse grid front speed for the time step from the simulation using the first grid; and

using the front speed for the time step as predicted by the machine learning system and the coarse grid front speed to update at least one localized static property of the reservoir.

25 . A system according to claim 24 , wherein:

the at least one localized static property of the reservoir comprises anisotropic permeability of the reservoir at one or more cells of the grid.

26 . A system according to claim 20 , wherein:

the simulation using the first grid is performed by a reservoir simulator configured to simulate EOR flooding operations using the first grid having a relatively low resolution.

27 . A method of modeling or simulating enhanced-oil recovery (EOR) flooding operations performed on a reservoir, comprising:

collecting training data based on simulation of EOR flooding operations using a first grid having a relatively high resolution;

using the training data to configure and train a machine learning system;

performing simulation of EOR flooding operations using a second grid having a relatively low resolution than the relatively high resolution of the first grid, wherein a resolution of the second grid is maintained during execution of the simulation using the second grid; and

using output data generated by the simulation using the second grid as input to the machine learning system to predict at least one parameter characterizing a displacing agent front in the second grid for a time step in the simulation using the second grid, wherein the at least one parameter characterizing the displacing agent front comprises a front speed for the time step and the front speed is used during execution of the simulation using the second grid to adjust reservoir simulation parameters associated with at least one cell of the second grid, wherein the reservoir simulation parameters are used to compute a subsequent time step of the simulation using the second grid.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 14, 2022
From: SAMSON, BARYS ANDREEVICH; GODEFROY, SOPHIE NAZIK; SHETH, SOHAM; MAROONEY, CHRISTOPHER
To: SCHLUMBERGER TECHNOLOGY CORPORATION
Reel/Frame 064105/0816 →
Continuity (2)
Provisional Application 62986872 · Mar 9, 2020
Related Publication 20230104036A1 · Apr 6, 2023
References Cited (44)
US 8301429B2 · Hajibeygi · 2012 [cited by examiner]
US 8725477B2 · Zhang · 2014 [cited by examiner]
US 8798977B2 · Hajibeygi · 2014 [cited by examiner]
US 9187984B2 · Usadi · 2015 [cited by examiner]
US 10060245B2 · Cullick · 2018 [cited by examiner]
US 10208577B2 · Hoteit · 2019 [cited by examiner]
US 10209401B2 · Ronot · 2019 [cited by examiner]
US 10519766B2 · Hu · 2019 [cited by examiner]
US 11680465B2 · Albinali · 2023 [cited by examiner]
US 11754745B2 · Kayum · 2023 [cited by examiner]
US 11781416B2 · Bakulin · 2023 [cited by examiner]
US 20080046223A1 · Noetinger · 2008 [cited by examiner]
US 20100132450A1 · Pomerantz · 2010 [cited by examiner]
US 20100138196A1 · Hui · 2010 [cited by examiner]
US 20100179797A1 · Cullick · 2010 [cited by examiner]
US 20100185428A1 · Vink · 2010 [cited by examiner]
US 20120158378A1 · Enchery · 2012 [cited by examiner]
US 20120179443A1 · Van Batenburg · 2012 [cited by examiner]
US 20120232859A1 · Pomerantz · 2012 [cited by examiner]
US 20130118736A1 · Usadi · 2013 [cited by examiner]
US 20130226540A1 · Pita · 2013 [cited by examiner]
US 20130245952A1 · Lin · 2013 [cited by examiner]
US 20140257577A1 · Prange · 2014 [cited by examiner]
US 20150100293A1 · Hoteit · 2015 [cited by examiner]
US 20190040724A1 · Dykstra · 2019 [cited by examiner]
US 20190094199A1 · Shi · 2019 [cited by examiner]
US 20190095792A1 · Kashinath · 2019 [cited by examiner]
US 20200202056A1 · Samson · 2020 [cited by examiner]
EP 3671290A1 · 2020 [cited by applicant]
WO 2012015516A1 · 2012 [cited by applicant]
Nilsson et al. (A Novel Adaptive Anisotropic Grid framework for Efficient Reservoir Simulation, 2005, SPE, pp. 1-23) (Year: 2005). [cited by examiner]
International Preliminary Report on Patentability of International Patent Application No. PCT/US2021/020610 issued Sep. 22, 2022, 7 pages. [cited by applicant]
Christensen et al., “Applications of Dynamic Gridding to Thermal Simulations”, Society of Petroleum Engineers, SPE 86969-MS, 2004, 10 pages. [cited by applicant]
Sammon, “Dynamic Grid Refinement and Amalgamation for Compositional Simulation”, Society of Petroleum Engineers, SPE 79683, 2003, 11 pages. [cited by applicant]
Berger and Colella, “Local Adaptive Mesh Refinement for Shock Hydrodynamics”, Journal of Computational Physics, vol. 82, pp. 64-84, 1989. [cited by applicant]
Hornung and Trangenstein, “Adaptive Mesh Refinement and Multilevel Iteration for Flow in Porous Media”, Journal of Computational Physics, vol. 136, pp. 522-545, 1997. [cited by applicant]
Cusini et al., “Algebraic dynamic multilevel (ADM) method for fully implicit simulations of multiphase flow in porous media”, Journal of Computational Physics, vol. 314, pp. 60-79, 2016. [cited by applicant]
Mostaghimi et a., “A Dynamic Mesh Approach for Simulation of Immiscible Viscous Fingering”, Society of Petroleum Engineers, SPE-173281-MS, 2015, 12 pages. [cited by applicant]
Van Batenburg et al., “Application of Dynamic Gridding Techniques to IOR/EOR-Processes”, Society of Petroleum Engineers, SPE 141711, 2011, 16 pages. [cited by applicant]
Gonzalez, “Adaptive Grid Refinement Improves Gas Injection Modeling”, Society of Petroleum Engineers, Spe 184487-STU, 2016, 18 pages. [cited by applicant]
Wikipedia, “Random forest”, 12 pages, accessed Sep. 2, 2022, available at: https://en.wikipedia.org/wiki/Random_forest. [cited by applicant]
Search Report and Written Opinion of International Patent Application No. PCT/US2021/020610 issued on Jun. 8, 2021; 11 pages. [cited by applicant]
Extended Search Report issued in European Patent Application No. 21767183.3 dated Feb. 19, 2024, 7 pages. [cited by applicant]
Samson, B. et al., “Machine Learning for Fast EOR Flooding Simulation”, 17th Annual European Conference on the Mathematics of Oil Recovery, 2020, 13 pages. [cited by applicant]