基于流固耦合的油井极限产液量预测新方法

油井提液是油田进入高含水期后的一项重要稳产措施，常规极限产液水平研究的关键点仅局限于能量供给水平、极限生产压差的判断，对极限提液水平主要以达西线性方法预测，其中均假定油藏中流体流过的多孔介质是刚性的，储层渗透率恒定不变。渤海湾的海上油田广泛分布高孔隙度、高渗透率、胶结程度差的疏松砂岩油藏，储层的特点和提液方式造成了流体流过的多孔介质会发生弹性与塑性变化、骨架变形与出砂，导致渗流条件发生根本性转变，使用常规极限产液水平预测方法误差大。以剖析常规极限产液水平研究为出发点，理论分析多孔介质变化影响因素、规律，与油田生产实践资料结合，将其变化规律融入渗流理论中，进而推导了考虑流固耦合下产液水平预测方法。将此预测方法应用于渤海C油田，与油田实际数据对比，误差率仅为5%，比常规方法提高40%~50%，具有一定推广意义。

A new method for predicting the limit liquid production rate of oil well based on fluid-solid coupling

The liquid increasing of oil well is one important measure to stabilize the production when an oilfield steps into the period of high water cut. Speaking of the conventional limit liquid production capacity study, its key point is only confined to the discrimination of energy supply capacity and limit production pressure difference, and its prediction on the limit liquid increasing capacity is mainly by means of the Darcy linear method. It is assumed that the porous medium where the fluid flows in the reservoir is rigid and the reservoir permeability is constant. Unconsolidated sandstone reservoirs of high porosity, high permeability and poor consolidation are widely distributed in the offshore oilfields of the Bohai Bay. These reservoir characteristics and liquid increasing method lead to elastic and plastic deformation, skeleton deformation and sand production of the porous medium where the fluid flows, which result in the essential transformation of the percolation conditions, so the error of the conventional limit liquid production capacity prediction method is large. After firstly dissecting the conventional limit liquid production capacity study, this paper theoretically analyzed the influence factors and change laws of porous medium. Then, combined with the oilfield production practice data, the change laws were introduced into the percolation theory to infer the liquid production capacity prediction method based on fluid-solid coupling. This prediction method is applied in Bohai C Oilfield. And compared with the actual oilfield data, the error rate of the fitting result is only 5% and the fitting rate is 40%-50% higher than that of conventional method. Therefore, this prediction method is worth popularizing.