1887

Abstract

Summary

The paper presents a novel approach that reduces the execution time for a closed-loop reservoir management workflow significantly. The speed-up comes from using coarse models that are calibrated to data using an improved history-matching workflow.

A reservoir simulation study consists of two main steps. First, a history matching step where the model is calibrated to historical well and seismic data, then a second step where the calibrated model is used to simulate future reservoir responses that can be used to optimize well locations and well controls. For robust decisions, ensemble methods using hundreds of simulation models are typically used to incorporate uncertainty in the predictions. This motivates fast simulation approaches.

The simulation time scales directly with the number of cells in the model, using coarser models is thus tempting to reduce the simulation time, but balancing accuracy and efficiency is challenging, and too coarse models may lead to large errors in the prediction. To remedy this, we use a recently developed history-matching setup where the shape of the relative permeability curves is adjusted in flow regions pre-computed from the drainage and flooding regions around the wells. The new history-matching workflow gives a good match to the data both for the training and the validation period even on significantly coarser grids.

The workflow is demonstrated on the Drogon model, which is a full reservoir simulation model created and shared by Equinor for testing closed-loop reservoir simulation workflows. It has a historical period including data for history matching, and a prediction setup. Our results show that we can match the historical data very well, even on a significantly coarser grid, if the main geological structures (faults, layers, oil-water contact, etc.) are preserved. The accuracy of the prediction, however, deteriorates if the grid is coarsened too much. Still, results shows that the optimized controls computed from a coarsened model, with 10 times speed-up in total simulation time compared to the original model, give significant improvements to the net-present-value of the original model.

© EAGE Publications BV
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202437071
2024-09-02
2026-04-15

  • From This Site

    /content/papers/10.3997/2214-4609.202437071
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. Baker, L. [1988] Three-phase relative permeability correlations. In: SPE Enhanced Oil Recovery Symposium. SPE, 17369.
     [Google Scholar]
  2. Chang, Y., Lorentzen, R.J., Nævdal, G. and Feng, T. [2019] OLYMPUS optimization under geological uncertainty.Computational Geosciences, 24(6), 2027–2042.
     [Google Scholar]
  3. Chen, Y. and Oliver, D.S. [2013] Levenberg–Marquardt forms of the iterative ensemble smoother for efficient history matching and uncertainty quantification.Computational Geosciences, 17(4), 689–703.
     [Google Scholar]
  4. Christie, M.A. [1996] Upscaling for reservoir simulation.Journal of petroleum technology, 48(11), 1004–1010.
     [Google Scholar]
  5. Equinor [2020] Drogon conceptual description.https://webviz-subsurface-example.azurewebsites.net/drogon-conceptual-description.
     [Google Scholar]
  6. Goel, G., Abidoye, L.K., Chahar, B.R. and Das, D.B. [2016] Scale dependency of dynamic relative permeability–satuartion curves in relation with fluid viscosity and dynamic capillary pressure effect.Environmental Fluid Mechanics, 16, 945–963.
     [Google Scholar]
  7. Kiærr, A., Lødøen, O., De Bruin, W., Barros, E. and Leeuwenburgh, O. [2020] Evaluation of a data-driven flow network model (FlowNet) for reservoir prediction and optimization. In: ECMOR XVII, 1. European Association of Geoscientists & Engineers, 1–18.
     [Google Scholar]
  8. Leeuwenburgh, O., Egberts, PJ., Barros, E.G., Turchan, L.P., Dilib, F., Lødøen, O.P. and de Bruin, W.J. [2024] A Hybrid Data-Physics Framework for Reservoir Performance Prediction with Application to H2S Production.SPE Journal, 1–17.
     [Google Scholar]
  9. Lie, K.A. and Krogstad, S. [2023] Data-driven modelling with coarse-grid network models.Computational Geosciences, 1–15.
     [Google Scholar]
  10. Lomeland, F. [2018] Overview of the LET family of versatile correlations for flow functions. In: the International Symposium of the Society of Core Analysts (SCA) held in Trondheim, Norway, 2018, SCA2018-056. https://www.jgmaas.com/SCA/2018/SCA2018-056.pdf.
     [Google Scholar]
  11. Møyner, O., Krogstad, S. and Lie, K.A. [2015] The application of flow diagnostics for reservoir management.SPE Journal, 20(02), 306–323.
     [Google Scholar]
  12. Oliver, D.S., Reynolds, A.C. and Liu, N. [2008] Inverse Theory for Petroleum Reservoir Characterization and History Matching. Cambridge University Press.
     [Google Scholar]
  13. OPM [2022] The Drogon Reservoir Model.https://github.com/OPM/opm-tests/blob/master/drogon/model/DROGON_HIST.DATA.
     [Google Scholar]
  14. OPM [2023a] OPM Flow Manual.https://opm-project.org/wp-content/uploads/2023/06/OPM_Flow_Reference_Manual_2023-04_Rev-0_Reduced.pdf.
     [Google Scholar]
  15. OPM [2023b] Utilities for Developing and Testing Flow Diagnostics Compuational Kernels.https://github.com/OPM/opm-flowdiagnostics-applications.
     [Google Scholar]
  16. PET [2023] Python Ensemble Toolbox.https://github.com/Python-Ensemble-Toolbox/PET. NORCE Energy, Data assimilation and optimization group.
     [Google Scholar]
  17. Ponting, D.K. [1989] Corner point geometry in reservoir simulation. In: ECMOR I-1st European conference on the mathematics of oil recovery. European Association of Geoscientists & Engineers, cp–234.
     [Google Scholar]
  18. Rasmussen, A.F., Sandve, T.H., Bao, K., Lauser, A., Hove, J., Skaflestad, B., Klöfkorn, R., Blatt, M., Rustad, A.B., Sævareid, O. et al. [2021] The open porous media flow reservoir simulator.Computers & Mathematics with Applications, 81, 159–185.
     [Google Scholar]
  19. Ren, G., He, J., Wang, Z., Younis, R.M. and Wen, X.H. [2019] Implementation of physics-based data-driven models with a commercial simulator. In: SPE Reservoir Simulation Conference? SPE, D010S017R010.
     [Google Scholar]
  20. Ren, G., Wang, Z., Lin, Y., Onishi, T., Guan, X. and Wen, X.H. [2023] A Fast History Matching and Optimization Tool and its Application to a Full Field with More than 1,000 Wells. In: SPE Reservoir Simulation Conference? SPE, D011S002R004.
     [Google Scholar]
  21. Sandve, T., Sævareid, O., Lomeland, F and Lorenzen, R. [2022] History Matching Field Scale Model Using LET Based Relative Permeability. In: ECMOR 2022, 1. European Association of Geoscientists & Engineers, 1–9.
     [Google Scholar]
  22. Stordal, A.S., Szklarz, S.P. and Leeuwenburgh, O. [2015] A Theoretical look at Ensemble-Based Optimization in Reservoir Management.Mathematical Geosciences, 48(4), 399–417.
     [Google Scholar]
  23. Wang, L., Yao, Y., Luo, X., Daniel Adenutsi, C., Zhao, G. and Lai, F. [2023] A critical review on intelligent optimization algorithms and surrogate models for conventional and unconventional reservoir production optimization.Fuel, 350, 128826.
     [Google Scholar]
  24. Zhou, X.H., Wang, H., McClure, J., Chen, C. and Xiao, H. [2023] Inference of relative permeability curves in reservoir rocks with ensemble Kalman method.The European Physical Journal E, 46(6), 44.
     [Google Scholar]
/content/papers/10.3997/2214-4609.202437071
Loading
Closed-Loop Reservoir Management Using Fast Data-Calibrated Coarse Models
Conference Proceedings 2024, 1 (2024); https://doi.org/10.3997/2214-4609.202437071
/content/papers/10.3997/2214-4609.202437071
/content/papers/10.3997/2214-4609.202437071
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error