基于开发流体势的蚁群算法在优势渗流通道预测中应用

优势渗流通道的形成会影响油藏注水开发的效果，如何判别与预测优势渗流通道成为研究重点。本文提出了基于开发流体势场的蚁群算法预测优势通道的方法，认为开发流体势场包括位能、压能、动能及界面能，流体在油藏内部分布的差异，决定了油藏内部流体势能大小的差异。将蚁群算法理论引入到油藏开发中，提出了以流体势场为基础的蚁群算法模型。在东河1油田进行运用，在建立油藏流体势模型的基础上，对过井流体势剖面进行了蚁群算法模拟。在模拟初期，受层内流体势分布差异性影响，不同韵律小层内部高浓度信息素的分布存在差异，即优势通道分布存在差异。在模拟后期，受层间流体势分布差异性影响，高浓度信息素主要集中在CIII2-2、3-1和3-2小层，即在这些小层中优势通道较发育，运用动态资料进行验证，效果较好。

The Application of Ant Colony Algorithm Based on the Theory of Development Fluid Potential in Predicting Preponderance Flow Path

The formation of preponderance flow path will affect the result of water injection in a reservoir. How to recognize and predictit becomes the focus of this research. In this paper, a method has been proposed to predict dominant flow path using ant colonyalgorithm based on the theory of development fluid potential. It is believed that a development fluid potential includes gravitationalpotential energy, pressure energy, kinetic energy and interfacial energy. The difference in fluid distribution in a reservoir determinesthe differences of fluid potential in the reservoir. Ant colony algorithm is introduced into the reservoir development, and it is consideredthat the formation of both the ant migration path and the preponderance flow path in the reservoir have the characteristics of regularity,stochastic and optimality. The model of ant colony algorithm based on fluid potential field is proposed. The new methods have beenused in Donghe 1 oilfield, after establishing the fluid potential model in a reservoir, ant colony algorithm simulation is conducted influid potential profile crossing wells. In initial simulation stage, the distribution of high density pheromones is different among differentrhythm layers controlled by development fluid potential distribution differences in inner layer, which means that the distributions ofpreponderance flow path are different in different layers. In later simulation stage, controlled by fluid potential distribution differencesinterlayer, high concentration pheromones are mainly distributed in the 2-2, 3-1 and 3-2 layers namely, the preponderance flow path iswell developed in these layers. Dynamic data have been used to verify it and the results are in good agreement.