论文部分内容阅读
综合沉积学研究、层序地层学分析、地质统计模拟、生产数据分析以及流体流动模拟,对Chaunov油田的储层进行了模拟。该储层与法国巴黎盆地中三叠统冲积扇体系的远源相对应,且具有强烈非均质性和分层的特点。该储层主要由夹有河漫滩和湖相泥岩的小型带状河道沉积物组成,河道合并速度随周期性的湖面升降变化,从而直接控制着储层的几何形态。在一个基准面周朝内,河道在低可容空间期间进行合并,形成高度非均质性的砂席。随着可容空间的增加,河漫滩泥岩范围内河道逐渐变得孤立。最终,湖侵在油田范围内沉积了湖相泥岩,并形成了虽薄却分布广泛的垂直渗透隔层。可容空间开始减少时,成壤作用的变更频繁发生,这一点可由白云石质壳结层和地下水成因的白云石证实。识别出构成储层分层格架的5个旋回。采用随机截断高斯函数法,可对这些储层单元内的岩类分布进行地质统计模拟。利用井资料可计算水平和垂直岩类比例曲线和方差图。由于井距宽,所以无法确定水平试验方差图的范围。在高分辨率网格内,对3种岩类现实进行模拟,以便对短、中、长相关长度进行比较。在把岩石物理性质赋予岩类并进行放大后,对这3种现实进行了流体流动模拟。然后将这3种流体流动模拟结果与油田10年的生产史进行比较。除了油田北部,若不考虑方差图范围,模拟结果相当吻合,说明该区储层分层中存在问题。流体模拟结果不受相关长度的影响,一是因为合并的河道储层单元内净产层含量高,一是因为约束井数量大;但用最长的相关长度进行的流体模拟结果与生产史的拟合非常好。
Comprehensive sedimentology studies, sequence stratigraphy analysis, geostatistical modeling, production data analysis, and fluid flow modeling have been used to simulate reservoirs in the Chaunov field. The reservoir corresponds to the distal source of the Middle Triassic alluvial fan system in the Paris Basin, France, and is characterized by strong heterogeneity and stratification. The reservoir is mainly composed of small zonal sediments with floodplains and lacustrine mudstones. The river merging speed changes periodically with the rise and fall of the lake, which directly controls the reservoir geometry. Within a datum perimeter, the channels merge during periods of low accommodation, creating a highly anisotropic sand mat. With the increase of accommodation space, the river courses in mudflats of river floodplains gradually become isolated. In the end, lakes invaded lacustrine mudstones within the field and formed a thin but widely distributed vertical infiltration compartment. Changes in pedogenesis occur frequently as the vapors begin to decrease, as evidenced by the dolostone-type crust and groundwater-forming dolomites. Identify the 5 cycles that make up the reservoir stratification grid. Stochastic truncated Gaussian function method can be used to conduct geostatistical simulation of the distribution of rock types in these reservoir units. Well data can be used to calculate horizontal and vertical rock mass ratio curves and variograms. Due to the wide well spacing, it is not possible to determine the range of the horizontal test variance map. Within the high-resolution grid, three rock-like realities were simulated to compare short, medium and long correlation lengths. After rock physical properties were given to rocks and amplified, the three kinds of realities were simulated by fluid flow. The results of these three fluid flow simulations are then compared with the 10-year production history of the field. Except for the northern part of the oilfield, the simulation results are in good agreement without considering the range of the variance plot, indicating that there are problems in the reservoir stratification in this area. The results of the fluid simulations are independent of the length of the correlation. The first is that the net payoffs in the combined fluvial reservoirs are high because of the large number of confinement wells; however, the fluid simulations performed with the longest correlation lengths are not related to production history Fitting is very good.