【摘要】 油藏数值模拟的历史拟合过程中模型参数多、可调整自由度大 ,目前普遍采用实验校正法进行拟合 ,主观随意性大且费时多。引入最优化理论建立目标函数 ,以渗流方程和油藏的边界条件、初始条件为约束条件 ,将历史拟合转化为非线性规划中求解定义在n维空间的函数极值问题 ,用改进的共轭梯度法求解 ,基本解决了收敛问题。拟合过程中调整参数的原则为 :不确定优先 ,敏感优先 ,先全局后个别。为减少调整参数 ,用解析函数表达相对渗透率与饱和度的关系 ,将模拟区域按油藏地质和井网条件划分为若干小区 ,分区拟合。对某封闭小断块油藏的区块与单井综合含水和地层压力进行自动历史拟合 ,调整的主要参数是相对渗透率和渗透率 ,计算了 12 6次 ,拟合结果与实际动态相符。图 2参 8更多还原

【Abstract】 A non linear optimum problem defined in a n dimensional space is solved with the creation of an objective function by introducing optimum theory, restraining initial and boundary reservoir conditions, and applying historical matching methodology. The principles for adjusting parameters during historical matching process include:①priority is not predetermined; ②sensitivity is given priority; and ③local adjustment follows overall consideration. To reduce the adjusted parameters, an analytical function is used to express the relationship between relative permeability and saturation. The studied area was divided into several sections based on reservoir geology and well distribution, and each section was simulated separately. An automatic historical matching was then conducted on individual wells in fault confined blocks to collect composite water cut and formation pressure data. The major adjustable parameters are relative and absolute permeability. With 126 experiments, the matching results are consistent with historical production data.更多还原