【作者】 樊长江；

【导师】 邓宏文； 卞德智；

【作者基本信息】 中国地质大学（北京） ， 能源地质工程， 2006， 博士

【摘要】 滨里海盆地典型盐下碳酸盐岩油藏由于上部有巨厚盐丘(盐层厚度变化从数十米到3700m不等),盐丘和围岩间存在巨大的速度差异,因此造成盐下地震波场十分复杂,地震资料的分辨率和信噪比低,盐丘侧翼和盐下构造地震成像困难;另外盐下碳酸盐岩储集体非均质性强,油气成藏控制因素复杂,加上盐丘对地震信号的畸变,导致储层识别和预测难度大。本文以滨里海盆地肯基亚克和让纳若尔油田为研究对象,开展了地震成像和储层预测研究。在研究工作中,本文提出了创新链和创新组合的概念,在重视创新点的前提下、更加强调创新链和创新组合的作用,并将这种认识成功应用到盐下碳酸盐岩油藏的地震成像和储层预测过程中。在地震成像中,应用了叠前去噪、面元均化、子波整形、时频域振幅补偿和球面扩散补偿、三维DMO叠加、Stolt偏移等多项关键技术,经过新一轮有针对性的攻关处理,使地震剖面的波组特征更加清楚、分辨率有所提高、各种地质现象更加清晰、局部偏移归位更加合理,为储层预测打下了坚实基础。本文提出了正反演联合储层预测的思路,强调地质正演模型分析的思想,针对不同成因的储层,优选不同的储层预测方法,将地质模型和预测结果有机地结合起来,相互刻度,循环验证,大大减少了地震反演的多解性。建立了该地区高分辨率层序地层格架,将盐下石炭系巴什基尔阶和谢尔普霍夫阶共分为6个准层序组,共16个准层序,层序划分结果为储层预测研究提供格架,从而提高储层预测的精度。综合应用地质、地震、测井、油藏等多方面资料,以建立不同成因储层的地球物理响应特征(测井和地震响应特征)作为储层预测的关键,以地震储层预测方法为主线,综合应用各种地震属性、地震道波形分类、分频信息、相干体属性和三维可视化手段,沿着由岩相寻找有利沉积相带、由物性寻找有利储层、由裂缝寻找高产带的步骤,建立了断裂、岩相、岩溶和裂缝的地质模型,形成了一套比较完整的盐下碳酸盐岩储层预测的配套技术和方法。根据预测结果,提出了5口加密井部署方案,实施后原油平均单井日产量由原来的70吨提高到269吨,其中两口井单井日产量超过500吨,一口井单井日产量突破1000吨,经济效益十分显著。盐下碳酸盐岩油气藏储层预测研究不仅具有重大的理论意义,而且具有非常显著的社会意义。更多还原

【Abstract】 For pre-salt carbonate reservoirs in Pre-Caspian Sea basin, there are massive saltdomes whose thickness varies from several dozen meters to 3700m and velocitydifference is big from that of the wall rock, which makes the seismic wave field onpre-salt domes becoming very complicated, seismic resolution and signal noise ratiobeing very low, and seismic imaging on salt dome’s flanks and pre-salt structuresbecoming very difficult. On the other hand, pre-salt carbonate formations have seriousheterogeneity and the controlling factors to form reservoir are very complicated,which, combined with the seismic events being distorted duo to the effect of saltdomes, enlarges the difficulty in reservoir identification and prediction. This paper,taking the Kenjiyak and Zarnor pre-salt reservoirs of the Pre-Caspian Sea basin as itsresearch objects, presents the study on seismic imaging and reservoir prediction.In the process of study, this paper provides the concepts of the innovation chainand the innovation combination. On the premise of thinking much of innovation spots,we emphasize the function of the innovation chain and the innovation combinationand apply this idea through seismic imaging into reservoir prediction of the carbonatereservoirs successfully.In seismic imaging, a series of techniques for enhancing resolution and signalnoise ratio were adopted, such as pre-stack de-noising, averaging bin size, waveletsform normalization, amplitude compensation in time and frequency domains andspherical spreading compensation, 3D DMO stack as well as Stolt migrationtechniques. Through new round tackling ken problems on seismic processing, seismicprofile quality and seismic resolution were improved, seismic migration locationbecame more reasonable, varied geology bodies became clearer, which then lay downa solid foundation for reservoir prediction.This paper sets forth the united forward and inversion reservoir prediction, whichemphasizes the analysis of geological forward model and screening separately thebest prediction methods for reservoirs of different origins. This approach integratesthe geological forward model with prediction results organically and calibrates andcorrects them interactively to reduce significantly the multi-solutions of seismicinversion.High resolution sequence stratigraphy study was done on this area. Throughstratigraphic correlation, 6 para-sequence group and 16 para-sequences wereidentified in C Formation. Sequence stratigraphy results can provide the frame forreservoir prediction, and so enhance the precision of reservoir prediction.For the features of extraordinarily low porosity and permeability, minor fracturesystem and cave growing in carbonate reservoirs, as well as many other kinds ofinformation such as geology, seismic, logging and reservoir data were combinativelyused to generate the geophysical(log and seismic) response features to differentreservoirs of various origin, which is used as the key prediction factor. Taking seismicprediction as the main thread, and comprehensively using varied seismic attributes ,seismic trace wave classification, spectrum decomposition, coherence property and3D visualization, looking for favorable sediment facies by lithofacies study, searchinggood beds by petrophysical study, and identifying high productive zone by fracturestudy were carried out and then relatively completed workflow and techniques forthe reservoir prediction were formedBased on the result of prediction, a five well in-fill drilling program was set andafter the program was implemented the average daily production rate per well wasraised from 70 to 269 tons, showing very good economic benefit.Pre-salt carbonate reservoir prediction not only has a great theoretical meaning,but produces vast social values.更多还原