【作者】 王树华；

【导师】 刘怀山；

【作者基本信息】 中国海洋大学 ， 海洋地球物理学， 2009， 博士

【摘要】 准噶尔盆地永进油田为储量规模上亿吨的整装油田。钻探证实,其油藏类型以白垩系“底砾岩”、侏罗系西山窑组煤下砂岩和侏罗系西山窑组煤上地层-岩性等隐蔽油藏类型为主。目的层埋藏深,砂体厚度薄,岩石物性差,储层预测和圈闭评价难,是该区目前面临着艰巨的油气勘探开发难题。针对永进地区复杂的地质构造特征,本文在区域地质背景和隐蔽油气藏成藏特征系统分析的基础上,利用高精度层序地层学分析、储层地震预测和地层不整合研究等方法技术,分别对岩性、地层-岩性等隐蔽圈闭的主控要素及匹配关系进行了详细研究。利用储层物性分析、孔隙度和叠前属性含油性预测、地层超、剥带的识别与确定和地层-岩性圈闭综合分析技术,对隐蔽圈闭进行了分析评价,并指出了有利的勘探目标。本文采用了地质、物探、测井相结合的综合研究手段,依据单井层序分析和地震剖面横向追踪,建立了永进地区侏罗系-白垩系底部层序地层时空格架,划分出侏罗系-白垩系底部8个三级层序界面。开展沉积体系和沉积相的精细研究,确定了四大沉积相类型(辫状河流、曲流河、三角洲、湖泊)。研究了层序地层框架中隐蔽油气藏形成条件与控制因素,建立起相应的层序地层成藏模式。本文较系统分析了隐蔽油气藏地震识别与描述的方法技术系列,包括圈闭识别技术、储层预测技术和油气预测技术等。通过地震解释和储层预测技术,落实构造形态和断裂系统分布,进行储层空间展布预测研究。在有利储层发育的沉积相带、沉积微相识别与分布研究的基础上,利用GR曲线开展储层特征反演,落实各个砂组在平面上的分布。利用属性分析、频谱分解等技术研究研究储层的物性,并进行分类。变速成图技术可以较好的解决速度横向变化较大、埋藏深的条件下的时深转换问题,有效识别了低幅度构造。相干分析、倾角方位角扫描、分频振幅谱较好的识别了地层尖灭带和准确描述了断层的展布。在常规地震技术很难区分砂泥岩的情况下,储层特征反演有效区分了砂泥岩。反演约束下的属性分析技术可预测储层的物性及储层的发育情况。油气检测技术可以做为隐蔽圈闭的含油气性预测的一种辅助手段。在白垩系-侏罗系储层预测描述研究、不整合结构体分析的基础上,通过储层物性分析、孔隙度和叠前属性含油性预测、地层超、剥带的识别与确定和地层-岩性圈闭的综合分析技术,进行了隐蔽圈的综合评价,优选钻探目标,为进一步勘探开发提供了比较可靠的依据。通过攻关和研究主要取得了以下3项创新成果:1.永1井区超深薄砂体储层空间准确归位技术首次采用变速成图技术较好的解决了在速度横向变化较大、埋藏深的条件下的时深转换问题,有效的识别低幅度构造。利用相干分析、倾角方位角扫描、分频振幅谱有效的描述了储层、确定了地层尖灭带和断层展布。2.永1井区少井超深薄砂体储层特征属性反演技术首次在目的储层和非储层的波阻抗发生重叠、而导致难以单纯使用波阻抗进行储层准确预测的条件下,利用测井曲线,结合录井资料,在井点构建能反映储层物性空间变化的特征属性,并利用神经网络映射技术反演此特征属性数据体,比较准确地预测了储层的展布。3.永1井区储层预测评价技术首次从不同角度、不同层次综合多种储层预测评价技术有效的分析预测了储层的发育特征。首先利用伪波阻抗曲线和自然伽玛曲线交会构建储层特征反演技术预测砂体;,其次利用常规属性对储层物性定性分析和多属性储层物性定量预测,最后,通过建立φ-AC孔隙度模型预测孔隙度,利用序贯高斯协模拟方法预测含油饱和度,利用分频技术和叠前地震AVO分析进行含油性预测,形成了一套储层评价预测序列方法。更多还原

【Abstract】 The Yongjin Oilfield had been explored to be a large oilfield with in-place volume of hundreds millions ton in Junggar Basin. It had been proved by drilling data that its oil reservoirs were mainly the basal conglomerate reservoir entrapping in Cretaceous, sandstone reservoir below the coal bed and stratigraphic-lithologic reservoirs above the coal bed entrapping in Jurassic Xishanyao Group. It was difficult to predict and evaluate the reservoirs due to deep buried target, thin sandstone developed and poor reservoir physical property, which made the hydrocarbon exploration difficult at present.The factors controlling lithologic or stratigraphic-lithologic reservoirs and its corresponding matching relationships had been studied through using high accuracy sequence stratigraphy, reservoir seismic prediction techniques and strata unconformity analysis basing on regional geologic background and subtle oil reservoir characters analysis. Finally the subtle traps had been analyzed and evaluated through analyzing reservoir physical property, porocity and prestack property oilness prediction, comprehensive unconformity and stratigraphic-lithologic traps analysis, and then the benefit reservoir development zone and two benefit targets had been predicted for drilling.The sequence stratigraphic time and space framework for Jurassic and Cretaceous bottom had been built according to individual well sequence analysis and seismic data lateral tracing combining with geologic data, geophysical exploration data and logging data, as a result eight third order sequence interfaces depositing at Jurassic and Cretaceous bottom had been classified. Four types deposit facies(braided stream, meandering stream, delta and lakes) had been determined through studying depositional system and depositional facies finely. The conditions for subtle traps configuration and their controlling factors had been studied to build corresponding sequence stratigraphic model for hydrocarbon entrapment.The seismic identification and description methods applied to analyzing subtle traps had been studied systematically in order to determine structure geometry and fractures distribution and predict reservoirs distribution in space, which were consisted by trap identification technique, reservoir prediction and hydrocarbon prediction techniques. The individual sandstone distribution in plain had been identified through seismic inversion with Gr curve for the benefit depositional facies zone. The reservoir physical property had been analyzed and classified through properties analysis and spectrum decomposition method. The variable velocity mapping was used to resolve questions of velocity great change in lateral and time and depth transformation under conditions of deep buried dept. it could be used to identify low amplitude structure. The coherence analysis, dip and azimuth scanning and frequency division amplitude were all applied to identify strata pinchout boundary and describing correctly faults distribution. The inversion technique used could differ sandstone and mudstone each other when they could not separated from each other using conventional method. The property analysis basing on constrain inversion was applied to predict reservoirs physical property and their development. The hydrocarbon could help predicting subtle traps bearing hydrocarbon. The subtle traps had been evaluated comprehensively to optimize the targets for drilling and provide reliable evidence after the reservoirs developed at Cretaceous-Jurassic had been predicted and described through analyzing unconformities structure and their physical property, porosity and pre-stacked extracted property, identifying strata onlapping and denudation zone As a result of study three new techniques had been developed. They are as following:①The developed technique for identifying correctly thin sandstones developed in deep formation in Yong1 well areaThe variable velocity mapping was used to resolve questions of velocity great change in lateral and time and depth transformation under conditions of deep buried dept. it could be used to identify low amplitude structure. The coherence analysis, dip and azimuth scanning and amplitude spectrum obtained by frequency division were all applied to identify strata pinchout boundary and describing correctly faults distribution.②The developed inversion technique with less wells drilled for thin sandstone characters prediction in Yong1 well area.It was difficult to use only the impedance to predict correctly reservoir distribution due to impedance overlapping in the area of benefit reservoir and non reservoir. So it was first developed that combining logging data to reconstruct characters property reflecting the change of reservoir in space at well drilled, and then applying neural network technique to inversing the characters cube, by with finally the reservoir distribution in space could be predicted correctly.③The developed technique for evaluating reservoirs in Yong1 well areaThe reservoir developed characters had been predicted combining variable prediction and evaluation techniques from different respects. Firstly it used inversion technique constrained with pseudo acoustic impedance and Gr curves to predict sandstones. Secondly the reservoirs physical property had been calculated in quantity. Finally aφ-AC model had been built to predict porosity, the oil saturation had been predicted with sequential gaussian simulation modeling, the reservoir oilness had been predicted using spectrum decomposition and prestack AVO analysis, as a result an effective prediction and evaluation for reservoir had been developed.更多还原