相对渗透率曲线是研究多相渗流的基础,在计算气井产量、分析气井产水规律等方面具有重要意义。由于常规的相对渗透率曲线一般通过岩心实验获取,耗时较长,成本高。为此,结合经验公式、油气藏数值模拟与机器学习的方法,提出一种采用生产动态数据计算气水相对渗透率曲线的方法。以四川盆地某气藏的一研究区块为例,通过气藏数值模拟计算的产气量、产水量、地层压力组成的样本集作为模型的输入,Brooks-Corey模型的参数作为模型输出,对比SCG、BR、LM神经网络学习算法,优选LM算法建立训练模型,进一步讨论了样本集的大小对预测结果的影响,并提出了相应的优化策略。研究结果表明：①由于不同的隐含层设置对神经网络的训练效果不同,当训练算法为LM、网络隐含层数为2层、节点数分别为41、32时对相渗曲线预测效果最好;②样本数的大小对网络训练速度和模型预测效果有重要影响,适当减少样本数可以改善模型预测效果,但同时也会增加网络训练误差;减少输入变量时间段,会增加网络训练误差,降低预测模型最终预测效果。实际应用表明,预测相渗曲线与岩心相渗曲线之间差异较小,并且产水量与压力拟合精度较高,因此该方法可以快速、准确地计算气水相对渗透率曲线,为气田开发生产提供有力的支持和指导。
Relative permeability curve is fundamental for investigating multiphase seepage, and it is of great significance in calculating the gas well production and analyzing the water-producing law. Conventional relative permeability curves are generally obtained through core experiments, which are time-consuming and costly. Therefore, this paper proposes a method for calculating gas-water relative permeability curves using production performance data, based on integrating empirical formula, reservoir numerical simulation, and machine learning. Taking a study block of a gas reservoir in the Sichuan Basin as an example, a sample set consisting of gas production, water production and formation pressure calculated by numerical simulation was utilized as the model input and the parameters of the Brooks-Corey model as the output. By comparing the learning algorithms of the Scaled Conjugate Gradient (SCG), Bayesian Regularization (BR), and Levenberg-Marquardt (LM) neural networks, a training model was established by the selected LM algorithm. Furthermore, the influence of the size of the sample set on prediction results was discussed and the corresponding optimization strategy proposed. The following results are obtained. (i) Different hidden layer settings can cause different training effects on the neural network, the best prediction of relative permeability curve is achieved when the LM algorithm is adopted with 2 hidden layers which have 41 and 32 nodes respectively. (ii) The sample size has great influences on the network training speed and model prediction effect. Appropriate reduction of the samples can improve the model prediction effect, but will increase the network training error. Reduction in the input variable time period will increase the network training error and reduce the final prediction effect of model. Practical application demonstrates that the difference between the predicted relative permeability curves and the core relative permeability curves is small, while the fitting accuracy of water production and pressure is high. This method is proved to be efficient and accurate for calculating gas-water relative permeability curves, providing a strong support and guidance for the development of gas fields.
[an error occurred while processing this directive]
[1] 何更生, 唐海. 油层物理(第二版)[M]. 北京: 石油工业出版社, 2011: 314-335.
HE Gengsheng, TANG Hai. Reservoir Physics (Second Edition)[M]. Beijing: Petroleum Industry Press, 2011: 314-335.
[2] LAI F.P., LI Z. P., WEI H. X., et al. Characterization of the generalized permeability jail in tight reservoirs by analyzing relative-permeability curves and numerical simulation[J]. Petroleum Science, 2023, 20(5): 2939-2950.
[3] MACHADO C.G., REYNOLDS A. C. Approximate semi-analytical solution for injection-falloff-production well test: An analytical tool for the in situ estimation of relative permeability curves[J]. Transport in Porous Media, 2018, 121(1): 207-231.
[4] LIU C., ZHOU W.S., JIANG J. Z. Dynamic calculation of water sweep efficiency and relative permeability curve on water drive reservoir[J]. Frontiers in Energy Research, 2022, 10: 922435.
[5] YANG Y., ZHANG X., ZHOU X.F., et al. Experimental analysis of the pore structure, relative permeability, and water flooding characteristics of the Yan’an Formation sandstone, southwestern Ordos Basin[J]. Energy Geoscience, 2023, 4(3): 100184.
[6] GONG Q.S., LIU Z. G., ZHU C., et al. Heterogeneity of a sandy conglomerate reservoir in Qie12 block, Qaidam Basin, Northwest China and its influence on remaining oil distribution[J]. Energies, 2023, 16(7): 2972.
[7] TANG B.C., REN K., LU H. T., et al. Study on residual oil distribution law during the depletion production and water flooding stages in the fault-karst carbonate reservoirs[J]. Processes, 2023, 11(7): 2147.
[8] LI Y.Y., WANG S. L., KANG Z. H., et al. Research on the correction method of the capillary end effect of the relative permeability curve of the steady state[J]. Energies, 2021, 14(15): 4528.
[9] BORAZJANI S., HEMMATI N., BEHR A., et al.Determining water-oil relative permeability and capillary pressure from steady-state coreflood tests[J]. Journal of Petroleum Science and Engineering, 2021, 205: 108810.
[10] ZHANG W.L., HOU J., LIU Y. G., et al. Determination of relative permeability curve under combined effect of polymer and surfactant[J]. Journal of Petroleum Science and Engineering, 2022, 215: 110588.
[11] WANG X. Y., WANG X. Q., WANG J. F., et al. Derivation of relative permeability curves from capillary pressure curves for tight sandstone reservoir based on fractal theory[C]//SPE/AAPG/SEG Unconventional Resources Technology Conference, 1-3 August2016, San Antonio, Texas, USA. DOI: https://doi.org/10.15530/URTEC-2016-2451467.
[12] CHEN S., LI G.M., PERES A., et al. A well test for in-situ determination of relative permeability curves[J]. SPE Reservoir Evaluation & Engineering, 2008, 11(1): 95-107.
[13] 周克明, 刘婷芝, 袁小玲, 等. 川西地区二叠系火山岩储层岩石敏感性及气水两相渗流特征[J]. 天然气勘探与开发, 2023, 46(1): 77-84.
ZHOU Keming, LIU Tingzhi, YUAN Xiaoling, et al.Rock sensitivity and gas-water two-phase seepage characteristics of Permian volcanic reservoirs in western Sichuan Basin[J]. Natural Gas Exploration and Development, 2023, 46(1): 77-84.
[14] CZERNIA B., BARRUFET M. A mechanistic approach for calculating oil-gas relative permeability curves in unconventional reservoirs[C]//SPE Annual Technical Conference and Exhibition, 30 September - 2 October2019, Calgary, Alberta,Canada. DOI: https://doi.org/10.2118/196051-MS.
[15] 王东琪, 殷代印. 水驱油藏相对渗透率曲线经验公式研究[J]. 岩性油气藏, 2017, 29(3): 159-164.
WANG Dongqi, YIN Daiyin.Empirical formulas of relative permeability curve of water drive reservoirs[J]. Lithologic Reservoirs, 2017, 29(3): 159-164.
[16] LIU S.Z., ZOLFAGHARI A., SATTARIN S., et al. Application of neural networks in multiphase flow through porous media: Predicting capillary pressure and relative permeability curves[J]. Journal of Petroleum Science and Engineering, 2019, 180: 445-455.
[17] SPADA N.S., CARNEIRO C. C., GIORIA R. S. Adjustment of relative permeability curves parameters by supervised artificial neural networks[C]. Rio de Janeiro: Rio Oil & Gas Expo and Conference, 2020.
[18] 侯贤沐, 王付勇, 宰芸, 等. 基于机器学习和测井数据的碳酸盐岩孔隙度与渗透率预测[J]. 吉林大学学报(地球科学版), 2022, 52(2): 644-653.
HOU Xianmu, WANG Fuyong, ZAI Yun, et al.Prediction of carbonate porosity and permeability based on machine learning and logging data[J]. Journal of Jilin University (Earth Science Edition), 2022, 52(2): 644-653.
[19] 冯曦, 彭先, 李隆新, 等. 人工智能技术在气藏工程专业的应用前景展望[J]. 天然气勘探与开发, 2023, 46(1): 65-76.
FENG Xi, PENG Xian, LI Longxin, et al.Artificial intelligence techniques and their application to gas reservoir engineering[J]. Natural Gas Exploration and Development, 2023, 46(1): 65-76.
[20] COREY A.T. The interrelation between gas and oil relative Permeabilities[J]. Producers Monthly, 1954, 19(1): 38-4l.
[21] BROOKS R.H., COREY, A. T. Hydraulic Properties of Porous Media[M]. Fort Collins: Colorado State University, 1964: 3-27.
[22] PURCELL W.R. Capillary pressures–Their measurement using mercury and the calculation of permeability therefrom[J]. Journal of Petroleum Technology, 1949, 1(2): 39-48.
[23] BURDINE N.T. Relative permeability calculations from pore size distribution data[J]. Journal of Petroleum Technology, 1953, 5(3): 71-78.
[24] 唐永强, 吕成远, 侯吉瑞. 用毛细管压力曲线计算相对渗透率曲线的方法综述[J]. 科学技术与工程, 2015, 15(22): 89-98.
TANG Yongqiang, LÜ Chengyuan, HOU Jirui.A review of methods to calculate relative permeability curve by using capillary pressure curve[J]. Science Technology and Engineering, 2015, 15(22): 89-98.
[25] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016: 97-117.
ZHOU Zhihua.Machine Learning[M]. Beijing: Tsinghua University Press, 2016: 97-117.
[26] 焦斌, 叶明星. BP神经网络隐层单元数确定方法[J]. 上海电机学院学报, 2013, 16(3): 113-116.
JIAO Bin, YE Mingxing.Determination of hidden unit number in a BP neural network[J]. Journal of Shanghai Dianji University, 2013, 16(3): 113-116.