基于分阶段拟合的多重介质试井解释方法

A Stage Wise Fitting-Based Well-Test Interpretation Method for Multiple Media

摘要: 由于缝洞型碳酸盐岩油藏显著的非均质性与多重渗流机理的并存，使得常规模型化的试井解释模板一直面临着试井解释模板有限的局限性与拟合失配等问题。针对此问题，本文提出了一种基于分阶段拟合的多重介质试井解释方法。该方法首先在理论上建立了包含基质、裂缝与洞穴多相介质以及多类边界条件的数学模型，并根据数学模型绘制出相应的特征响应曲线，建立响应曲线库；随后对现场试井曲线按流动阶段进行分段划分，并对各流动阶段分别与响应曲线库中相对应的理论模板进行参数化拟合；最终通过组合优化将拟合后的模板按时间序列重构为完整试井响应曲线，并在此基础上开展储层判断与参数反演。根据数值算例表明，该方法在曲线拟合精度和局部响应的判识能力方面均优于基于整体固定模板的传统方法。该方法可有效降低固有模板选择带来的局限性，提高对缝洞型碳酸盐岩复杂渗流行为的识别与定量估计能力。

Abstract: The pronounced heterogeneity of fracture-vug carbonate reservoirs and the coexistence of multiple flow mechanisms have long limited conventional template-based well-test interpretation, causing template constraints and poor fits. To address this problem, this paper proposes a staged-fitting, multi-medium well-test interpretation method. First, a theoretical mathematical model is established that incorporates matrix, fracture and cavity media and multiple types of boundary conditions, and the corresponding characteristic response curves are generated to form a response-curve library. Field well-test curves are then segmented according to flow stages, and each segment is parametrically fitted to the corresponding theoretical template from the library. Finally, by means of combinatorial optimization the fitted templates are reconstructed in time sequence to produce a complete well-test response curve, on the basis of which reservoir classification and parameter inversion are carried out. Numerical examples show that this method outperforms traditional fixed-template approaches in both curve-fitting accuracy and the ability to identify local responses. The proposed approach effectively reduces the limitations associated with fixed template selection and improves the recognition and quantitative estimation of the complex flow behaviours in fracture-vug carbonate reservoirs.