首页|基于智能算法的气藏水体参数计算

基于智能算法的气藏水体参数计算

扫码查看
气藏水体参数的合理确定是准确计算水侵量的重要因素,为水侵气藏动态分析、开发方案的调整提供有力支撑.为此,以水驱气藏物质平衡方程为基础,建立水体参数模型以形成目标函数,在拟合地层压力与实测地层压力误差最小的前提下计算水体参数,探讨基于智能算法的气藏水体参数计算方法.以某气藏A井区为例,分别采用鲸鱼优化算法和粒子群优化算法对目标函数进行优化,计算水体参数,并作对比分析,研究结果表明:①用鲸鱼优化算法得到最佳的水体参数组合是水体内边界半径(rw)1 101 m,无因次外边界半径(reD)5.10,水侵角度(θ)25.07°,储容比(ω)0.13,窜流系数(λ)0.000 35;用粒子群优化算法得到最佳的水体参数组合是水体内边界半径(rw)1 030 m,无因次外边界半径(reD)4.83,水侵角度(θ)15.68°,储容比(ω)0.21,窜流系数(λ)0.000 16.②两种算法得到拟合地层压力值与实测地层压力值的相对误差均较小,整体不超过 5%;同时,鲸鱼优化算法的平均相对误差 1.36%,较之粒子群算法平均相对误差 2.01%,鲸鱼优化算法的误差相对更小.③鲸鱼优化算法具有收敛速度快、全局搜索能力强和操作简便等优势;粒子群算法作为一种进化算法,具有计算目标函数收敛值更小的特点.两种算法均可行.水体参数计算结果为该气藏水侵规律分析确定了基础参数,算法对比研究结果则增添了新的技术手段,为我国同类气藏的水体参数计算及基于计算结果的气藏高效开发提供技术参考.
Calculating waterbody parameters in gas reservoirs based on intelligent algorithms
Reasonable determination of waterbody parameters in gas reservoirs is not only important to calculate water invasion pre-cisely but a great support for dynamic analysis of water-invaded reservoirs and adjustment of development plan.So,based on the ma-terial balance equation of water drive gas reservoirs,the model of waterbody parameters was developed to form an objective function,and these parameters were calculated under the premise of minimum error between fitting and measured formation pressure.And the calculation method based on intelligent algorithms was discussed.Moreover,taken A well area in certain gas reservoirs as an example,two optimization algorithms of both whale and particle swarm were used respectively to optimize the function.And the obtained pa-rameters were compared.Results show that(i)the optimal waterbody parameters induced from the whale optimization algorithm are:the radius of waterbody's inner broundary(rw)of 1,101 m,the dimensionless radius of outer boudary(reD)of 5.10,the water-invasion angle(θ)of 25.07°,the storativity ratio(ω)of 0.13,and the interporosity flow coefficient(λ)of 0.00035,while those from the particle swarm optimization algorithm are:rw of 1,030 m,reD of 4.83,θ of 15.68°,ω of 0.21,and λ of 0.00016,respectively;(ii)the relative error between the fitting pressure from the two algorithms and the measured pressure is small(<5%generally),and the average rela-tive error is 1.36%in the whale algorithm,which is smaller than that of another algorithm(2.01%);and(iii)the whale enjoys fast con-vergence,strong global search ability and simple operation.And the swarm,as an evolutionary algorithm,is characterized by smaller convergence value of objective function.Both algorithms are feasible.The calculation results lay the foundation for analyzing water invasion in gas reservoirs,and the comparison of two algorithms provides not only an additional technique for calculating waterbody parameters in similar gas reservoirs in China but reference for efficient development of such gas reservoirs.

Waterbody parameter modelWhale optimization algorithmParticle swarm algorithmWater-invasion parameter

宋利红、李成刚、刘波、魏之焯、江良伟

展开 >

中国石化西北油田分公司 新疆乌鲁木齐 830011

成都理工大学能源学院 四川成都 610059

水体参数模型 鲸鱼优化算法 粒子群算法 水侵参数

2024

天然气勘探与开发
中油西南油气田公司勘探开发研究院

天然气勘探与开发

影响因子:0.543
ISSN:1673-3177
年,卷(期):2024.47(2)
  • 18