致密砂岩气层压裂产能及等级预测方法

致密砂岩储层孔隙度小、渗透率低、含气饱和度低,基本上没有自然产能,需要进行压裂,因此进行压裂产能的预测很有必要。笔者研究了鄂尔多斯盆地苏里格东部地区盒8段致密砂岩气层的压裂产能及等级预测。利用反映储层流动性质的测井参数(电阻率、自然伽马、声波时差、中子、密度)与反应压裂施工情况的压裂施工参数(单位厚度砂体积、砂比、砂质量浓度、单位厚度排量、单位厚度入井总液量),选择数学统计方法神经网络法进行致密砂岩气层压裂产能等级预测。分析比较Elman神经网络、支持向量回归(SVR)、广义回归神经网络(GRNN)3个神经网络预测致密砂岩气层压裂产能模型的网络结构参数、回判及预测精度以及运行耗费时间。结果表明:3个模型中,GRNN网络参数只有1个,回判和预测精度最高,运行时间小于1 s。因此,选择GRNN模型预测致密砂岩气层压裂产能,并用于苏里格东部地区致密砂岩气层压裂产能的等级预测。等级预测准确率达到90%。

Research on Gas Yield and Level Prediction for Post-Frac Tight Sandstone Reservoirs

Tight sandstone reservoirs are always with the characterized by the low porosity,low permeability, and low gas saturation, hardly have any natural capacity,needing fracturing for productivity,fracturing capacity prediction is necessary .We mainly research the method of the gas productivity and level prediction for post-frac tight sandstone reservoirs in the eastern of Sulige region of the Ordos basin.The prediction model is established by neural networks with the logging parameters (R T ,GR,AC,CNL,DEN )and fracturing parameters(the amount of sard per meter,the sand ratio,the sand concentration,delivery capacity per meter,total amount of fluid injection the well per meter).The neural networks are Elman neural network,support vector machine (SVR)neural network,and GRNN neural network.We compared the results of three models,and picked up the best model GRNN model to predict the level of the gas productivity.