The BZ 34-1 oilfield is a typical gas cap edge water reservoir in the Bohai oilfield. The main characteristics of the oilfield were multi-phase sand body stacking and the sand body was composed of three parts: gas cap...The BZ 34-1 oilfield is a typical gas cap edge water reservoir in the Bohai oilfield. The main characteristics of the oilfield were multi-phase sand body stacking and the sand body was composed of three parts: gas cap, oil reservoir, and edge water. The actual production site results show that the permeability difference of multi-layer sand bodies has a serious impact on the development effect. This article establishes a typical reservoir model numerical model based on the total recovery degree of the reservoir and the recovery degree of each layer, and analyzes the impact of permeability gradient. As the permeability gradient increases, the total recovery degree of all four well patterns decreases, and the total recovery degree gradually decreases. The recovery degree of low permeability layers gradually decreases, and the recovery degree of high permeability layers gradually increases. As the permeability gradient increases, the degree of recovery gradually decreases under different water contents. As the permeability gradient increases, the reduction rate of remaining oil saturation in low permeability layers is slower, while the reduction rate of remaining oil saturation in high permeability layers was faster. By analyzing the impact of permeability gradient on the development effect of oil fields, we could further deepen our understanding of gas cap edge water reservoirs and guide the development of this type of oil field.展开更多
文摘The BZ 34-1 oilfield is a typical gas cap edge water reservoir in the Bohai oilfield. The main characteristics of the oilfield were multi-phase sand body stacking and the sand body was composed of three parts: gas cap, oil reservoir, and edge water. The actual production site results show that the permeability difference of multi-layer sand bodies has a serious impact on the development effect. This article establishes a typical reservoir model numerical model based on the total recovery degree of the reservoir and the recovery degree of each layer, and analyzes the impact of permeability gradient. As the permeability gradient increases, the total recovery degree of all four well patterns decreases, and the total recovery degree gradually decreases. The recovery degree of low permeability layers gradually decreases, and the recovery degree of high permeability layers gradually increases. As the permeability gradient increases, the degree of recovery gradually decreases under different water contents. As the permeability gradient increases, the reduction rate of remaining oil saturation in low permeability layers is slower, while the reduction rate of remaining oil saturation in high permeability layers was faster. By analyzing the impact of permeability gradient on the development effect of oil fields, we could further deepen our understanding of gas cap edge water reservoirs and guide the development of this type of oil field.
文摘小时天然气负荷预测受外部特征因素与预测方法的影响,为提高其预测精度并解决其他深度学习类模型或组合模型可解释性差、训练时间过长的问题,在引入“小时影响度”这一新特征因素的同时提出一种基于极端梯度提升树(extreme gradient boosting tress,XGBoost)模型与可解释性神经网络模型NBEATSx组合预测的方法;以XGBoost模型作为特征筛选器对特征集数据进行筛选,再将筛选降维后的数据集输入到NBEATSx中训练,提高NBEATSx的训练速度与预测精度;将负荷数据与特征数据经STL(seasonal and trend decomposition using Loess)算法分解为趋势分量、季节分量与残差分量,再分别输入到XGBoost中进行预测,减弱原始数据中的噪音影响;将优化后的NBEATSx与XGBoost模型通过方差倒数法进行组合,得出STL-XGBoost-NBEATSx组合模型的预测结果。结果表明:“小时影响度”这一新特征是小时负荷预测的重要影响因素,STL-XGBoost-NBEATSx模型训练速度有所提高,具有良好的可解释性与更高的预测准确性,模型预测结果的平均绝对百分比误差、均方误差、平均绝对误差分别比其余单一模型平均降低54.20%、63.97%、49.72%,比其余组合模型平均降低24.85%、34.39%、23.41%,模型的决定系数为0.935,能够很好地拟合观测数据。
