摘要
介绍了一种可用于计算机自动与递减曲线拟合分析的新方法。Arps的指数、双曲以及调和曲线的经验递减方程只能用于各种变量间是纯线性关系的单一关系中。即q、N_p和qt呈线性关系。然后对这些变量进行多元线性回归,以便得到Arps方程中q_i、d_i和b之间的最佳相关系数。一旦q_i,D_i和b的值确定下来,就可确定出未来产量值,因此,就不必要再进行适配典型曲线了。该方法对分析预测油气井的生产状况很有用处。
This paper presents a new approach which can be used for the computer-automated curve fitting for decline curve analysis.Arps' empirical-decline-curve equations for the exponential,hyperbolic and harmonic solutions are manipulated to produce a single relation which is a linear combination of variables:q,np and qt.A linear multiple regression is performed for these variables to obtain optimal coefficients of the relation which are related to q_i,D_i and b in Arps^1 equations.Once q_i,D_i and b are known,the future production rate can be determined.Thus,the type-curve matching is not needed. The proposed approach is,therefore,useful in application of the computer-aided analysis for the production forecast for oil and gas wells.
