关键词: 测井, 神经网络, 储集层, 产能预测, 应用

Abstract: The factors that will affect petroleum productivity can be divided into two categories,namely the artificial factors and reservoir factors. For a specific oil zone,under the roughly similar artificial operating conditions,its productivity will mainly depend upon reservoir property. Based on Darcy's two-dimensional productivity formula, this paper studied the theoretical equation of productivity, using the function relation between relative permeability and water saturation as a linkage,and derived multivariant relational expression between the effective porosity,permeability and resistivity. Then, combined with theory of well logging, taking it as a parameter for comprehensive evaluation, adopting the artificial nervous network technique,this method has been applied to productivity prediction of Xiaowa region's Dongying formation of Liaohe oilfield. The results are good, and verified the usefulness of the method.

Key words: well logging, nervous network, reservoir, productivity prediction, application