Research Progress and Prospects of Data-Driven Reservoir History Matching Methods

Haoxi Shi; Yang Lu; Ziqian Hu; Ming Zhang; Liyang Wang; Botao Liu

doi:10.54097/jzxdmx11

Authors

Haoxi Shi
Yang Lu
Ziqian Hu
Ming Zhang
Liyang Wang
Botao Liu

DOI:

https://doi.org/10.54097/jzxdmx11

Keywords:

History Matching; Reservoir Modeling; Machine Learning; Graph Neural Network (GNN); Transformer; Proxy Model

Abstract

Reservoir history matching is a crucial process in oilfield development, aiming to calibrate geological models, characterize reservoir parameters, and enhance predictive accuracy. Traditional manual history matching relies heavily on human expertise, extensive computational effort, and subjective judgment, which can no longer meet the demands posed by increasingly complex geological structures and massive data volumes. In recent years, data-driven and machine learning approaches have shown great potential in reservoir inversion and production forecasting. Emerging models such as proxy models, deep neural networks, graph neural networks (GNNs), and Transformers provide new paradigms for achieving automatic reservoir history matching. This paper reviews the historical evolution of history matching methods, tracing their progression from early numerical simulation and optimization algorithms to modern data-driven modeling techniques. From the perspective of machine learning, it further investigates the performance of these methods in dynamic reservoir prediction and discusses the application prospects of spatio-temporal fusion models based on GNNs and Transformers. Finally, the paper outlines the current challenges and future directions in this field, aiming to contribute to the advancement of intelligent reservoir modeling and the development of digital twin reservoirs.

References

[1]Lake L W. Fundamentals of Reservoir Engineering [M]. Elsevier, 1988.

[2]Adams Z T, Allen T O. History matching of three-phase flow in a layered reservoir [J]. SPE Journal, 1992, 2 (04): 482-490.

[3]Coats K H. History matching: current status and future trends [J]. SPE Reservoir Engineering, 1989, 4 (04): 482-490.

[4]Al-Mudhafar W. Machine learning applications in petroleum engineering: A review [J]. Journal of Petroleum Science and Engineering, 2018, 167: 687-700.

[5]Bishop C M. Neural networks for pattern recognition [M]. Oxford university press, 1995.

[6] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need [C]//Advances in neural information processing systems. 2017: 5998-6008.

[7]Van Genuchten M T. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils [J]. Soil science society of America journal, 1980, 44 (05): 892-898.

[8]Jacquard A. History matching by regression analysis [J]. Journal of Petroleum Technology, 1965, 17 (8): 1025-1034.

[9]Chavent G., Dupuy M., Lemonnier P. History matching by Use of Optimal Theory. SPE Journal, 1975, 15 (3): 236-245.

[10]Maschio C., et al. Hybrid optimization algorithms for automatic history matching [J]. Journal of Petroleum Science and Engineering, 2019, 172: 345-358.

[11]Holland J H. Adaptation in natural and artificial systems: An introductory analysis with applications to biology, control, and artificial intelligence [M]. MIT press, 1975.

[12]Kennedy J, Eberhart R. Particle swarm optimization [C]//Proceedings of ICNN'95-international conference on neural networks. 1995, 4: 1942-1948.

[13]Kirkpatrick S, Gelatt C D, Vecchi M P. Optimization by simulated annealing [J]. science, 1983, 220 (4598): 671-680.

[14]Yuan H, Zhang J, Liu Y, et al. Genetic algorithm-based history matching for a low-permeability reservoir with fracture networks [J]. Journal of Petroleum Science and Engineering, 2016, 144: 721-730.

[15]Chai S., et al. Closed-loop data assimilation with machine learning models [J]. Fuel, 2021, 305: 121-137.

[16]Price K, Storn R M. Differential evolution–a simple and efficient heuristic for global optimization over continuous spaces [J]. Journal of global optimization, 1997, 11 (04): 341-359.

[17]Shi Y, Eberhart R. A modified particle swarm optimizer [C]//Proceedings of the 1998 IEEE international conference on evolutionary computation proceedings. IEEE world congress on computational intelligence. 1998: 69-73.

[18]Zhang Y, Li G, Liu X, et al. Particle swarm optimization combined with numerical simulation for history matching in a heterogeneous reservoir [J]. Journal of Petroleum Exploration and Production Technology, 2018, 8 (04): 1461-1470.

[19]Caers J. History matching as a statistical inverse problem: A Bayesian approach [J]. Mathematical Geology, 2003, 35 (04): 381-403.

[20]Box G E, Wilson K B. On the experimental attainment of optimum conditions [J]. Journal of the royal statistical society. series B (methodological), 1951: 1-45.

[21]Hardy R L. Multiquadric equations of topography and other irregular surfaces [J]. Journal of geophysical research, 1971, 76 (08): 1905-1915.

[22]Matheron G. Principles of geostatistics [J]. Economic geology, 1963, 58 (08): 1246-1266.

[23]Jin R, Chen W, Sudjianto A. An efficient algorithm for constructing optimal response surface designs [J]. Technometrics, 2001, 43 (04): 439-450.

[24]Fonseca R, Schiozer D J. Artificial neural networks as a proxy for reservoir simulation in history matching [J]. Journal of Petroleum Science and Engineering, 2009, 67 (03-4): 188-196.

[25]Mohaghegh S, Aminian K, Ameri S. Application of machine learning to reservoir characterization and modeling [J]. SPE Reservoir Evaluation & Engineering, 2011, 14 (06): 687-698.

[26]Hinton G E, Osindero S, Teh Y W. A fast learning algorithm for deep belief nets [J]. Neural computation, 2006, 18 (07): 1527-1554.

[27]LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition [J]. Proceedings of the IEEE, 1998, 86 (11): 2278-2324.

[28]Hochreiter S, Schmidhuber J. Long short-term memory [J]. Neural computation, 1997, 9 (08): 1735-1780.

[29]Zhao W, Liu Y, Li Z, et al. Reservoir porosity prediction using 3D convolutional neural network with seismic data [J]. Journal of Petroleum Science and Engineering, 2019, 179: 106133.

[30]Li X, Zhang J, Wang Y, et al. Water cut prediction in oil wells using LSTM neural network [J]. Journal of Petroleum Science and Engineering, 2020, 191: 107230.

[31]Wang H, Zhang L, Yan S, et al. Dual surrogate model with attention mechanism for reservoir history matching [J]. Fuel, 2021, 295: 120568.

[32]Chen X, Ma X, Liu B. Generalization ability of deep learning surrogate models for reservoir history matching [J]. Computers & Geosciences, 2022, 162: 105632.

[33]Zhang Y, Li G, Liu X, et al. Data preprocessing for machine learning-based reservoir history matching [J]. Journal of Petroleum Science and Engineering, 2017, 157: 874-883.

[34]Zhang L., Yan S., et al. Dual surrogate models for reservoir parameter inversion [J]. Petroleum Science, 2023, 20 (2): 455–467.

[35]Peaceman D W. Interpretation of well-block pressures in numerical reservoir simulation with non-square grid blocks and anisotropic permeability [J]. SPE Journal, 1978, 18 (06): 405-412.

[36]Elsheikh A H, Mohaghegh S, Aminian K. Application of support vector regression for predicting bottomhole pressure in oil wells [J]. SPE Journal, 2014, 19 (02): 333-342.

[37]Goodfellow I, Bengio Y, Courville A. Deep learning [M]. MIT press, 2014.

[38]Cybenko G. Approximation by superpositions of a sigmoidal function [J]. Mathematics of control, signals and systems, 1989, 2 (04): 303-314.

[39]Vapnik V N. The nature of statistical learning theory [M]. Springer science & business media, 1995.

[40]Breiman L. Random forests [J]. Machine learning, 2001, 45 (01): 5-32.

[41]Zhou Z H. Ensemble methods: foundations and algorithms [M]. CRC press, 2020.

[42]Kingma D P, Welling M. Auto-encoding variational bayes [C]//2nd International Conference on Learning Representations, ICLR 2014, Banff, AB, Canada, April 14-16, 2014, Conference Track Proceedings. 2014.

[43]Li X, Zhang J, Wang Y, et al. Reservoir parameter inversion using variational autoencoder and differential evolution [J]. Journal of Petroleum Science and Engineering, 2019, 178: 105845.

[44]Belytschko T, Lu Y Y, Gu L. Element-free Galerkin methods [J]. International journal for numerical methods in engineering, 1994, 37 (05): 811-840.

[45]Zhao W, Liu Y, Li Z, et al. Extended connection element model for multiphase flow in reservoirs [J]. Journal of Petroleum Science and Engineering, 2020, 192: 107345.

[46]Liu Y, Zhao W, Li Z, et al. Meshless method combined with machine learning for reservoir history matching [J]. Computers & Geosciences, 2021, 152: 104732.

[47]Zhang Y, Li G, Liu X, et al. Graph structure optimization for connection element model in reservoir modeling [J]. Journal of Petroleum Science and Engineering, 2022, 214: 110432.

[48]Scarselli F, Gori M, Tsoi A C, et al. The graph neural network model [J]. IEEE transactions on neural networks, 2008, 20 (01): 61-80.

[49]Kipf T N, Welling M. Semi-supervised classification with graph convolutional networks [C]//5th International Conference on Learning Representations, ICLR 2017, Toulon, France, April 24-26, 2017, Conference Track Proceedings. 2017.

[50]Parchoudi A, Mohaghegh S, Aminian K. Well connectivity analysis using graph neural networks [J]. SPE Reservoir Evaluation & Engineering, 2020, 23 (06): 1663-1676.

[51]Rajabi M, Mohaghegh S, Aminian K. Pressure distribution prediction in reservoirs using graph neural networks [J]. Journal of Petroleum Science and Engineering, 2022, 213: 110285.

[52]Wang Z., Zeng Y., et al. Graph Neural Network for Reservoir Simulation and Prediction [J]. Computers & Geosciences, 2021, 155: 104-130.

[53]Yang X, Zhang L, Yan S, et al. Physics-informed graph neural network for reservoir history matching [J]. Fuel, 2023, 331: 125845.

[54]Zhang Y, Li G, Liu X, et al. Long-term oil production prediction using transformer neural network [J]. Journal of Petroleum Science and Engineering, 2021, 204: 108645.

[55]Lundberg S M, Lee S I. A unified approach to interpreting model predictions [C]//Advances in neural information processing systems. 2017: 4765-4774.