Study on three-dimensional permeability model of gravel sand control layer based on simulated annealing algorithm.

In the process of natural gas hydrate exploitation, the secondary hydrate formation may occur in both the reservoir and the completion layer. The secondary hydrate formed during the exploitation process and the sand particles falling from the reservoir will block the flow channel and reduce the gas production. In this paper, the permeability change of sand filling layer under the influence of sand and hydrate blockage is studied from a three-dimensional perspective, and the permeability change characteristics of sand mixing and secondary hydrate formation in filling layer caused by pore space blockage are summarized. A three-dimensional model for predicting the permeability of gravel layer is developed. In previous studies, the mathematical model parameters derived from the empirical formula are always not unique, which will cause inaccurate permeability prediction results. In this study, the simulated annealing algorithm is used to optimize the parameters such as boundary conditions and cooling tables. The undetermined parameters in the existing mathematical prediction model of gravel layer permeability are optimized into a determined maximum value to solve the optimization problem of the undetermined parameters of the permeability model. The algorithm in this paper takes the actual value of gravel layer permeability with different sand content and hydrate concentration in the experiment as the standard, and takes the mean square error of the predicted value and the experimental value of the prediction model as the objective function. The simulated annealing algorithm is used to estimate and optimize the undetermined parameters of the model. The algorithm results show that the mean square error between the predicted value and the experimental value is minimized with n = 22.7 and τ = 1.014(n and τ are the parameters to be optimized in the permeability model of this paper). The fitting degree between the optimized permeability prediction model and the experimental value is higher than that before optimization, which verifies the correctness and rationality of the algorithm in this paper. • Simulated annealing algorithm is used to optimize the permeability model. • The simulated annealing algorithm is optimized to have faster solution speed. • The NMSE error calculation method is used to make the predicted permeability model better fit the actual value. • The optimized model is combined with the capacity model. [ABSTRACT FROM AUTHOR]