特高含水油田常规开发动态指标预测方法分析

ANALYSIS ON PREDICTING METHOD OF CONVENTIONAL DEVELOPMENT DYNAMIC INDEXES IN OILFIELD WITH ULTRA-HIGH WATER CUT

因为油田有相当一部分剩余可采储量将在特高含水期采出，所以如何将油田的剩余可采储量高效地在此阶段采出，是目前我国各油田急需解决的问题。解决此问题的基础就是对该阶段油田的开发动态指标进行准确的预测。针对特高含水油田的开发特点和面临的问题，结合油田常规开发动态指标预测方法的特征和实用性分析，对常规动态分析预测方法从机理和缺陷方面进行了评价，提出了在不考虑预测可靠性的前提下，适合该阶段的常规动态分析预测方法依次是油藏数值模拟法、预测模型法、递减规律法、联解法和水驱特征曲线法。指出了研究新的预测方法的必要，并指明了研究特高含水油田开发动态指标预测新方法的方向：支持向量机预测、基于时变系统的功能模拟预测、组合预测和智能预测。

An obstacle for oil fields development in China to conquer nowadays is to effectively produce the remained recoverable reserves during ultra-high water cut stage, for during this stage, a large quantity of remained recoverable reserve can indeed be exploited. A premise to settle this problem is the accurate prediction of development dynamic indexes during this stage. Regarding to the development characteristics and the confronting problems of oil fields with ultra-high water cut, as well as combining the features and validity of conventional dynamic indexes prediction method, the mechanism and deficiency of the conventional prediction method have been analyzed and evaluated. Approaches like oil reservoir numerical simulation, prediction model, declination laws, and resultslinked method and water-flooded characteristic curves have been proposed and assumed as conventional prediction methods that are fit for the development stage of the oil fields, which are given not considering the reliability of the prediction results. The necessity of a new prediction method has been indicated and the new methods for dynamic indexes predictions such as SVM （ support vector machine ） prediction, function simulation prediction based on time-varied system, combination prediction and intelligent prediction have been proposed.