pith. sign in

arxiv: 2409.10875 · v3 · submitted 2024-09-17 · 🧮 math.NA · cs.NA

An Adaptive Subdomain Coupling Approach in Domain Decomposition for Multiphase Porous Media Flow

Shizhe Li , Li Zhao , Chen-Song Zhang This is my paper

Pith reviewed 2026-05-23 20:44 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords domain decompositionadaptive couplingmultiphase flowporous medianonlinear solversinitial guessesparallel performancereservoir simulation
0
0 comments X

The pith

An adaptive subdomain coupling framework in domain decomposition handles strong local nonlinearities to improve convergence and scalability of nonlinear solvers for multiphase porous media flow.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes an adaptively coupled subdomain framework based on domain decomposition methods for large-scale multiphase flow simulations in porous media. This framework manages strong local nonlinearities that arise in the fully implicit method by solving subproblems only within dynamically selected coupled regions. It introduces several adaptive coupling strategies along with a novel procedure for computing initial guesses. The goal is to accelerate nonlinear solver convergence and maintain competitive performance on parallel machines, as confirmed by numerical experiments on reservoir models.

Core claim

We present an adaptively coupled subdomain framework based on domain decomposition methods. This framework effectively handles strong local nonlinearities in global problems by solving subproblems within the coupled regions. Furthermore, we propose several adaptive coupling strategies and present a novel method for calculating initial guesses, aimed at improving the convergence and scalability of nonlinear solvers. A series of numerical experiments validate the effectiveness and robustness of the proposed framework. Additionally, large-scale reservoir simulations demonstrate that the proposed method achieves competitive parallel performance.

What carries the argument

The adaptively coupled subdomain framework based on domain decomposition methods, which solves localized subproblems in dynamically chosen coupled regions to address local nonlinearities.

Load-bearing premise

Dynamically choosing coupled subdomains and applying the new initial-guess procedure will reduce total nonlinear iterations enough to offset added communication or setup costs on parallel machines without introducing instability in heterogeneous media.

What would settle it

A large-scale heterogeneous-media simulation in which the adaptive coupling either increases the total number of nonlinear iterations or raises wall-clock time compared with standard domain decomposition.

Figures

Figures reproduced from arXiv: 2409.10875 by Chen-Song Zhang, Li Zhao, Shizhe Li.

Figure 1
Figure 1. Figure 1: Subdomain partitioning and displacement front. [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Illustration of the first subdomain coupling strategy (Strategy A). [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of the second subdomain coupling strategy (Strategy B). [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Illustration of the third subdomain coupling strategy (Strategy C). [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The domain partitioning by 784 processes using ParMetis, with colors corre [PITH_FULL_IMAGE:figures/full_fig_p016_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: shows the field average pressure (FPR), field gas production rate (FGPR), and field water production rate (FWPR), validating the consistency of the computational results across the different solution methods. Fig￾ure 7 presents the cumulative global Newton-Raphson iterations (NRiter), cumulative global linear iterations (LSiter), and total simulation runtime (Runtime) [PITH_FULL_IMAGE:figures/full_fig_p01… view at source ↗
Figure 7
Figure 7. Figure 7: Comparisons of NRiter, LSiter, and Runtime among four methods for Case 1. From [PITH_FULL_IMAGE:figures/full_fig_p017_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison of results for different subdomain coupling strategies: (a)-(d) show [PITH_FULL_IMAGE:figures/full_fig_p020_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The horizontal permeability distribution for grid layers 1 to 2, 3 to 5, and 6 to [PITH_FULL_IMAGE:figures/full_fig_p022_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Comparison of results for subdomain coupling Strategy B: (a)-(d) show the [PITH_FULL_IMAGE:figures/full_fig_p023_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Comparison of the parallel strong scalability results for different methods. The [PITH_FULL_IMAGE:figures/full_fig_p026_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Comparison of the parallel weak scalability results for each method in the [PITH_FULL_IMAGE:figures/full_fig_p029_12.png] view at source ↗
read the original abstract

The numerical simulation of large-scale multiphase flow in porous media is of considerable importance across various application fields, particularly in the petroleum industry. The fully implicit method is preferred in reservoir simulations owing to its superior numerical stability and more relaxed time step constraints. However, this method requires solving a large nonlinear system, which becomes highly nonlinear in complex heterogeneous media with small grid scales, emphasizing the need for efficient and convergent numerical methods to accelerate nonlinear solvers on parallel computing systems. In this paper, we present an adaptively coupled subdomain framework based on domain decomposition methods. This framework effectively handles strong local nonlinearities in global problems by solving subproblems within the coupled regions. Furthermore, we propose several adaptive coupling strategies and present a novel method for calculating initial guesses, aimed at improving the convergence and scalability of nonlinear solvers. A series of numerical experiments validate the effectiveness and robustness of the proposed framework. Additionally, large-scale reservoir simulations demonstrate that the proposed method achieves competitive parallel performance.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The manuscript proposes an adaptively coupled subdomain framework based on domain decomposition for multiphase porous media flow. It claims to handle strong local nonlinearities by solving subproblems in coupled regions, introduces several adaptive coupling strategies plus a novel initial-guess procedure to improve nonlinear solver convergence and scalability, and asserts that numerical experiments plus large-scale reservoir simulations confirm effectiveness, robustness, and competitive parallel performance.

Significance. If the performance claims hold with concrete metrics, the adaptive coupling approach could provide a practical route to mitigating local nonlinearities in heterogeneous media without sacrificing the stability advantages of fully implicit methods, potentially aiding scalability on parallel machines for reservoir simulation.

major comments (1)
  1. [Abstract] Abstract (and implied Results section): the central claim that the adaptive subdomain coupling plus novel initial-guess method reduces total nonlinear iterations enough to offset extra communication/setup cost and deliver competitive parallel performance rests solely on the statements that 'a series of numerical experiments validate the effectiveness' and 'large-scale reservoir simulations demonstrate competitive parallel performance.' No iteration counts, wall-clock times, communication volumes, stability indicators, test-case descriptions, or comparison baselines are supplied, so it is impossible to verify whether the extra costs are offset or whether instability appears in the heterogeneous regimes targeted by the method.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful review and the opportunity to clarify the presentation of our results. We address the major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract (and implied Results section): the central claim that the adaptive subdomain coupling plus novel initial-guess method reduces total nonlinear iterations enough to offset extra communication/setup cost and deliver competitive parallel performance rests solely on the statements that 'a series of numerical experiments validate the effectiveness' and 'large-scale reservoir simulations demonstrate competitive parallel performance.' No iteration counts, wall-clock times, communication volumes, stability indicators, test-case descriptions, or comparison baselines are supplied, so it is impossible to verify whether the extra costs are offset or whether instability appears in the heterogeneous regimes targeted by the method.

    Authors: We agree that the abstract would benefit from including concrete quantitative metrics to make the performance claims immediately verifiable. In the revised manuscript we will expand the abstract to report key results from the numerical experiments, including observed reductions in nonlinear iterations, wall-clock time comparisons, and parallel scalability indicators (e.g., strong and weak scaling efficiency) relative to standard domain-decomposition baselines. The full supporting data—iteration counts, timings, communication volumes, test-case descriptions, and stability indicators—are already present in Sections 4 and 5 with accompanying tables and figures; the revision will simply surface the most salient numbers in the abstract itself. This change directly addresses the concern without altering the technical content or conclusions of the work. revision: yes

Circularity Check

0 steps flagged

No circularity: algorithmic construction with external experimental validation

full rationale

The paper presents an adaptive subdomain coupling framework for domain decomposition in multiphase porous media flow, along with coupling strategies and an initial-guess procedure. These are described as algorithmic proposals whose effectiveness is asserted via separate numerical experiments and large-scale simulations. No equations, fitted parameters, self-citations, or derivation steps appear in the provided text that would reduce any claimed improvement or prediction to a quantity already defined by the method itself. The central claims therefore remain independent of the inputs they are evaluated against.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract supplies no explicit free parameters, mathematical axioms, or newly postulated entities; any coupling thresholds or convergence tolerances would be implementation details not visible here.

pith-pipeline@v0.9.0 · 5698 in / 1216 out tokens · 35452 ms · 2026-05-23T20:44:10.447710+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

  1. [1]

    Cluster Computing 6(3), 215–226 (Jul 2003), https://doi.org/10.1023/A: 1023588520138

    Aavatsmark, I., 2002. Anintroductiontomultipointfluxapproximations for quadrilateral grids. Computat. Geosci. 6, 405–432. doi:10.1023/A: 29 1021291114475

    work page doi:10.1023/a: 2002

  2. [2]

    A com- pact multipoint flux approximation method with improved robustness

    Aavatsmark, I., Eigestad, G., Mallison, B., Nordbotten, J., 2008. A com- pact multipoint flux approximation method with improved robustness. Numer. Meth. Part. D. E. 24, 1329–1360. doi:10.1002/num.20320

    work page doi:10.1002/num.20320 2008

  3. [3]

    Petroleum Reservoir Simulation

    Aziz, K., 1979. Petroleum Reservoir Simulation. Applied Science Pub- lishers

    work page 1979

  4. [4]

    PETSc Web page

    Balay, S., Abhyankar, S., Adams, M.F., et al., 2024. PETSc Web page. https://petsc.org/

    work page 2024

  5. [5]

    Nonlinearly preconditioned inexact New- ton algorithms

    Cai, X.C., Keyes, D.E., 2002. Nonlinearly preconditioned inexact New- ton algorithms. SIAM J. Sci. Comput. 24, 183–200. doi: 10.1137/ S106482750037620X

    work page 2002

  6. [6]

    InexactNewtonmethodswithrestrictedadditive Schwarz based nonlinear elimination for problems with high local non- linearity

    Cai, X.C., Li, X., 2011. InexactNewtonmethodswithrestrictedadditive Schwarz based nonlinear elimination for problems with high local non- linearity. SIAM J. Sci. Comput. 33, 746–762. doi:10.1137/080736272

    work page doi:10.1137/080736272 2011

  7. [7]

    Computational Methods for Multi- phase Flows in Porous Media

    Chen, Z., Huan, G., Ma, Y., 2006. Computational Methods for Multi- phase Flows in Porous Media. SIAM. doi:10.1137/1.9780898718942

    work page doi:10.1137/1.9780898718942 2006

  8. [8]

    IMPES stability: The CFL limit

    Coats, K.H., 2003. IMPES stability: The CFL limit. SPE Journal 8, 291–297. doi:10.2118/85956-PA

    work page doi:10.2118/85956-pa 2003

  9. [9]

    Nonlinear preconditioning: How to use a nonlinear Schwarz method to precondition Newton’s method

    Dolean, V., Gander, M.J., Kheriji, W., Kwok, F., Masson, R., 2016. Nonlinear preconditioning: How to use a nonlinear Schwarz method to precondition Newton’s method. SIAM J. Sci. Comput. 38, A3357– A3380. doi:10.1137/15M102887X

    work page doi:10.1137/15m102887x 2016

  10. [10]

    A method for calculating multi-dimensional immiscible displacement

    Douglas, Jim, J., Peaceman, D., Rachford, H.H., J., 1959. A method for calculating multi-dimensional immiscible displacement. Trans. AIME 216, 297–308. doi:10.2118/1327-G

    work page doi:10.2118/1327-g 1959

  11. [11]

    HYPRE: A library of high per- formance preconditioners, in: Computational Science — ICCS 2002, Springer Berlin Heidelberg, Berlin, Heidelberg

    Falgout, R.D., Yang, U.M., 2002. HYPRE: A library of high per- formance preconditioners, in: Computational Science — ICCS 2002, Springer Berlin Heidelberg, Berlin, Heidelberg. pp. 632–641. doi:10. 1007/3-540-47789-6_66. 30

    work page 2002

  12. [12]

    Application-oriented preconditioning of seepage mechanics

    Feng, C., Li, S., Liu, S., Zhang, C., Zhao, L., 2024. Application-oriented preconditioning of seepage mechanics. Chinese J. Comput. Phys. 41, 98–109. doi:10.19596/j.cnki.1001-246x.8791

    work page doi:10.19596/j.cnki.1001-246x.8791 2024

  13. [13]

    A multi-stage preconditioner for the black oil model and its OpenMP implementation

    Feng, C., Shu, S., Xu, J., Zhang, C., 2014. A multi-stage preconditioner for the black oil model and its OpenMP implementation. Lect. Notes Comput. Sci. Eng. 98, 141–153. doi:10.1007/978-3-319-05789-7_11

    work page doi:10.1007/978-3-319-05789-7_11 2014

  14. [14]

    A class of parallel two-level nonlinear Schwarz preconditioned inexact Newton algorithms

    Hwang, F.N., Cai, X.C., 2007. A class of parallel two-level nonlinear Schwarz preconditioned inexact Newton algorithms. Comput. Methods Appl. Mech. Engrg. 196, 1603–1611. doi:10.1016/j.cma.2006.03.019

    work page doi:10.1016/j.cma.2006.03.019 2007

  15. [15]

    METIS: Unstructured graph partitioning and sparse matrix ordering system

    Karypis, G., Kumar, V., 2009. METIS: Unstructured graph partitioning and sparse matrix ordering system. URL: http://www.cs.umn.edu/ metis

    work page 2009

  16. [16]

    ParMETIS: Parallel graph partitioning and fill-reducing matrix ordering

    Karypis, G., Schloegel, K., Kumar, V., 2020. ParMETIS: Parallel graph partitioning and fill-reducing matrix ordering. URL:https://github. com/KarypisLab/ParMETIS

    work page 2020

  17. [17]

    A nu- merical study of the additive Schwarz preconditioned exact Newton method (ASPEN) as a nonlinear preconditioner for immiscible and compositional porous media flow

    Klemetsdal, Ø., Moncorgé, A., Møyner, O., Lie, K.A., 2022. A nu- merical study of the additive Schwarz preconditioned exact Newton method (ASPEN) as a nonlinear preconditioner for immiscible and compositional porous media flow. Computat. Geosci. 26, 1045–1063. doi:10.1007/s10596-021-10090-x

    work page doi:10.1007/s10596-021-10090-x 2022

  18. [18]

    Study of parallel numerical methods and software for mul- tiphase multicomponent flow in porous media

    Li, S., 2024. Study of parallel numerical methods and software for mul- tiphase multicomponent flow in porous media. Ph.D. thesis. University of Chinese Academy of Sciences

    work page 2024

  19. [19]

    OpenCAEPoro: A parallel simulation frame- work for multiphase and multicomponent porous media flows

    Li, S., Zhang, C.S., 2024. OpenCAEPoro: A parallel simulation frame- work for multiphase and multicomponent porous media flows. Preprint arXiv:2406.10862

    work page arXiv 2024

  20. [20]

    Numer- ical studies of a class of linear solvers for fine-scale petroleum reser- voir simulation

    Li, Z., Wu, S., Zhang, C., Xu, J., Feng, C., Hu, X., 2017. Numer- ical studies of a class of linear solvers for fine-scale petroleum reser- voir simulation. Comput. Visualization Sci. 18, 93–102. doi:10.1007/ s00791-016-0273-3

    work page 2017

  21. [21]

    Raissi, P

    Liu, L., Gao, W., Yu, H., Keyes, D.E., 2024. Overlapping multi- plicative Schwarz preconditioning for linear and nonlinear systems. J. 31 Comput. Phys. 496, 112548. URL:https://doi.org/10.1016/j.jcp. 2023.112548

    work page doi:10.1016/j.jcp 2024

  22. [22]

    Field-split preconditioned inexact Newton algorithms

    Liu, L., Keyes, D.E., 2015. Field-split preconditioned inexact Newton algorithms. SIAM J. Sci. Comput. 37, A1388–A1409. doi: 10.1137/ 140970379

    work page 2015

  23. [23]

    Convergence analysis for the multiplicative Schwarz preconditioned inexact Newton algorithm

    Liu, L., Keyes, D.E., 2016. Convergence analysis for the multiplicative Schwarz preconditioned inexact Newton algorithm. SIAM J. Numer. Anal. 54, 3145–3166. URL:https://doi.org/10.1137/15M1028182

    work page doi:10.1137/15m1028182 2016

  24. [24]

    A note on adaptive nonlinear preconditioning techniques

    Liu, L., Keyes, D.E., Krause, R., 2018. A note on adaptive nonlinear preconditioning techniques. SIAM J. Sci. Comput. 40, A1171–A1186. doi:10.1137/17M1128502

    work page doi:10.1137/17m1128502 2018

  25. [25]

    Nonlinear preconditioning strate- gies for two-phase flows in porous media discretized by a fully implicit discontinuous Galerkin method

    Luo, L., Cai, X.C., Keyes, D.E., 2021. Nonlinear preconditioning strate- gies for two-phase flows in porous media discretized by a fully implicit discontinuous Galerkin method. SIAM J. Sci. Comput. 43, S317–S344. doi:10.1137/20M1344652

    work page doi:10.1137/20m1344652 2021

  26. [26]

    Comparison of solutions to a three-dimensional black- oil reservoir simulation problem

    Odeh, A.S., 1981. Comparison of solutions to a three-dimensional black- oil reservoir simulation problem. J. Pet. Technol. 33, 13–25. doi:10. 2118/9723-PA

    work page 1981

  27. [27]

    Interpretation of well-block pressures in numerical reservoir simulation

    Peaceman, D., 1978. Interpretation of well-block pressures in numerical reservoir simulation. SPE Journal 18, 183–194. doi:10.2118/6893-PA

    work page doi:10.2118/6893-pa 1978

  28. [28]

    Generalpurposecompositionalsimulationformultiphase reactive flow with a fast linear solver

    Qiao, C., 2015. Generalpurposecompositionalsimulationformultiphase reactive flow with a fast linear solver. Ph.D. thesis. The Pennsylvania State University

    work page 2015

  29. [29]

    Saad, Iterative Methods for Sparse Linear Systems, Society for Industrial and Applied Mathematics, 2003

    Saad, Y., 2003. Iterative Methods for Sparse Linear Systems. Second ed., SIAM. doi:10.1137/1.9780898718003

    work page doi:10.1137/1.9780898718003 2003

  30. [30]

    The Boost Graph Library: User Guide and Reference Manual

    Siek, J., Lee, L.Q., Lumsdaine, A., 2002. The Boost Graph Library: User Guide and Reference Manual. Addison-Wesley

    work page 2002

  31. [31]

    Domain de- composition strategies for nonlinear flow problems in porous media

    Skogestad, J.O., Keilegavlen, E., Nordbotten, J.M., 2013. Domain de- composition strategies for nonlinear flow problems in porous media. J. Comput. Phys. 234, 439–451. doi:10.1016/j.jcp.2012.10.001. 32

    work page doi:10.1016/j.jcp.2012.10.001 2013

  32. [32]

    Incomplete Gaussian elimination as a preconditioning for generalized conjugate gradient acceleration

    Wallis, J., 1983. Incomplete Gaussian elimination as a preconditioning for generalized conjugate gradient acceleration. SPE Reservoir Simula- tion Conference SPE-12265. doi:10.2118/12265-MS

    work page doi:10.2118/12265-ms 1983

  33. [33]

    Constrained residual accelera- tion of conjugate residual methods

    Wallis, J., Kendall, R., Little, T., 1985. Constrained residual accelera- tion of conjugate residual methods. SPE Reservoir Simulation Confer- ence SPE-13536. doi:10.2118/13536-MS

    work page doi:10.2118/13536-ms 1985

  34. [34]

    Somenonoverlappingdomaindecompositionmeth- ods

    Xu, J., Zou, J., 1998. Somenonoverlappingdomaindecompositionmeth- ods. SIAM Review 40, 857–914. URL: https://doi.org/10.1137/ S0036144596306800

    work page 1998

  35. [35]

    An adaptive nonlinear elimination pre- conditioned inexact Newton algorithm for highly local nonlinear mul- ticomponent PDE systems

    Yang, H., Hwang, F.N., 2018. An adaptive nonlinear elimination pre- conditioned inexact Newton algorithm for highly local nonlinear mul- ticomponent PDE systems. Appl. Numer. Math. 133, 100–115. URL: https://doi.org/10.1016/j.apnum.2018.01.008

    work page doi:10.1016/j.apnum.2018.01.008 2018

  36. [36]

    A scalable fully implicit frame- work for reservoir simulation on parallel computers

    Yang, H., Sun, S., Li, Y., Yang, C., 2018. A scalable fully implicit frame- work for reservoir simulation on parallel computers. Comput. Methods Appl. Mech. Engrg. 330, 334–350. doi:10.1016/j.cma.2017.10.016

    work page doi:10.1016/j.cma.2017.10.016 2018

  37. [37]

    Masset, R

    Yang, H., Sun, S., Li, Y., Yang, C., 2019. Parallel reservoir simulators for fully implicit complementarity formulation of multicomponent com- pressible flows. Comput. Phys. Commun. 244, 2–12. doi:10.1016/j. cpc.2019.07.011

    work page doi:10.1016/j 2019

  38. [38]

    Active-set reduced-space methods with nonlinear elimination for two-phase flow problems in porous media

    Yang, H., Yang, C., Sun, S., 2016. Active-set reduced-space methods with nonlinear elimination for two-phase flow problems in porous media. SIAM J. Sci. Comput. 38, B593–B618. URL:https://doi.org/10. 1137/15M1041882

    work page 2016

  39. [39]

    Linear solvers for petroleum reservoir simulation

    Zhang, C., 2022. Linear solvers for petroleum reservoir simulation. J. Numer. Methods Comput. Appl. 43, 1–26. doi: 10.12288/szjs. s2021-0813

    work page doi:10.12288/szjs 2022

  40. [40]

    Parallel multilevel domain decomposition preconditioners for monolithic solution of non-isothermal flow in reservoir simulation

    Zhang, M., Yang, H., Wu, S., Sun, S., 2022. Parallel multilevel domain decomposition preconditioners for monolithic solution of non-isothermal flow in reservoir simulation. Comput. Fluids 232, 105183. doi:10.1016/ j.compfluid.2021.105183. 33

    work page arXiv 2022

  41. [41]

    Parallel multi-stage preconditioners with adaptive setup for the black oil model

    Zhao, L., Feng, C., Zhang, C.S., Shu, S., 2022. Parallel multi-stage preconditioners with adaptive setup for the black oil model. Comput. Geosci. 168, 105230. doi:10.1016/j.cageo.2022.105230

    work page doi:10.1016/j.cageo.2022.105230 2022

  42. [42]

    An improved multistage preconditioner on GPUs for compositional reservoir simula- tion

    Zhao, L., Li, S., Zhang, C.S., Feng, C., Shu, S., 2023. An improved multistage preconditioner on GPUs for compositional reservoir simula- tion. CCF Trans. High Perform. Comput. 5, 144–159. doi:10.1007/ s42514-023-00136-0. 34

    work page 2023