基于数据驱动的低渗透油藏优势流道识别及调堵动态优化

Water plugging is a significant adjustment measure for water drive development in low permeability reservoirs. Due to natural fractures, artificial fracturing and long-term water injection, the dominant flow channels can not be identified clearly in low-permeability reservoirs and the success rate of plugging measures is low. The conventional mathematical model methods are not mature yet so that the optimization for profile control is hard to realized. Therefore, this project focuses on the following research points: 1) Targeting at identifying dominant flow channels, considering the process of profile control and water plugging, establishing one data-driven model which could predict water-oil dynamic production and profile control and water plugging in low permeability reservoirs, combining inversion solution by history matching, realizing the connectivity identification and prediction of profile control and water plugging in low permeability reservoirs. 2) Based on the production optimization theory, the optimization method, which established by using data-driven model for production dynamics and the quantity of plugging, achieving the fast optimization on injection-production scheme and plugging adjustments which was verified by indoor tests. This project intends to use daily production data to establish a rapid optimization method for plugging adjustment measures for low-permeability reservoirs without relying on complex geological modeling to achieve rapid decision-making on the adjustment of plugging wells, injection timing and usage.

调剖堵水是低渗油藏水驱开发的重要调整措施。由于天然裂缝、人工压裂改造及长期注入水的作用，现场对低渗油藏优势流道分辨不明，调堵措施的成功率较低，而且针对调堵措施的传统数模方法尚不成熟，难以实现对调堵方案的优化设计。因此，本项目重点进行以下研究：1）以识别优势流道为目标，考虑调堵模拟过程，建立一种适用于低渗油藏的油水动态及调堵预测数据驱动模型，结合历史拟合反演求解，实现对低渗油藏的连通性识别及调堵动态预测；2）基于生产优化理论，建立基于数据驱动模型的生产动态、调堵用量优化方法，实现对注采制度、调堵方案的快速优化，并通过室内实验加以验证。本项目拟主要利用日常动态数据建立一种不依赖于复杂地质建模的低渗油藏调堵措施快速优化方法，实现对调堵井、注入时机及用量的快速决策。

低渗透油藏是我国石油増储上产的重要资源基础，调剖堵水已成为我国低渗透油藏注水开发中不可缺少的一类工艺改造措施。目前针对低渗透油藏调剖堵水的动态模拟预测比较困难，加之对地层连通规律认识不清，油藏优势渗流通道识别不准确，造成现场调剖堵水措施见效差、失效快，需要开展针对该类油藏优势流道识别及调堵动态优化研究。在国家自然科学基金的资助下，针对以上问题的研究，取得的主要成果包括：1）基于连通性思想，考虑低渗油藏裂缝、压裂水平井特点，建立了刻画油水两相流动动态的数据驱动模型，实现了加入调堵剂后的油水动态计算；2）建立了计算机辅助历史拟合方法，实现了生产动态的快速拟合和模型参数准确反演，结合注水劈分系数、洛伦兹系数，实现了井间优势窜流通道识别；3）建立了调堵措施井优选方法和考虑调堵的油藏生产优化控制模型，利用优化算法实现了注采参数和调堵用量的快速优化；4）开展了室内物理模拟实验，建立堵剂用量与传导率变化的关系模型，将模型计算结果与实验数据对比，验证了数据驱动模型的正确性。本项目形成了一套适用于低渗油藏，集数据驱动模型建模、计算机历史拟合反演、注采参数及调堵模拟与优化为一体的方法体系，可以科学指导低渗油藏的注水开发。