论文部分内容阅读
对TNK-BP公司负责的西西伯利亚下白垩纪的一个油田的生产数据分析表明,用他们目前所考虑的一种水驱开发方案开发油田,其最终采收率不会超过15%。很明显这个方案还有改进的余地。然而,由于从8个含油层获得的监测数据有限,用有限的数据推断油田目前的动用状况,对于探测剩余储量来说其结果必定有极大的不确定性。这促使我们检测油藏和油井特征,将油藏潜力的不确定性量化,并由此确保未来的提高采收率战略具有可行性。为了明确预测中的不确定性,有必要建立多个与观测到的矿场数据相匹配的预测模型。如果是人工拟合,那么这种拟合过程极其耗费时间。因此尝试将自动历史拟合算法连接到动态模拟器上,形成一个程序,用该程序加速拟合过程。
Analysis of production data from a Lower Cretaceous West Siberia oilfield to TNK-BP indicated that the final oil recovery will not exceed 15% with the development of a waterflood development scheme currently under consideration. There is clearly room for improvement in this program. However, due to the limited monitoring data available from the eight oil-bearing formations, the limited availability of data to infer current utilization of the field is bound to have great uncertainties in detecting residual reserves. This prompted us to test reservoir and well characteristics, quantify the uncertainty of reservoir potential and thereby ensure future future oil recovery strategies are viable. In order to clarify the uncertainty in the forecast, it is necessary to establish several forecast models that match the observed mine data. If it is artificial fitting, then this fitting process is extremely time-consuming. So try connecting the automatic history fitting algorithm to the dynamic simulator to form a program that accelerates the fitting process.