【作者】 张玉荣；

【导师】 艾池；

【作者基本信息】 东北石油大学 ， 石油与天然气工程， 2011， 博士

【摘要】 在油藏注水开发过程中,由于注水井的不断注水冲洗,不同开发时期的储层物性参数有不同的变化特征,并随开发时间的延长呈现出更显著变化。分层注水时由于纵向上层间相互作用会产生层间干扰现象,且在注水量较高的中低渗透油层表现得尤为明显,这些问题都会为注水开发后期开发方案设计及配注方案编制等带来影响。以往研究建立的储层参数变化模型没有考虑时间的概念,因而无法对储层参数未来的变化规律做出预测,也不能将储层物性参数变化引起的层间干扰引入,而正是这些储层物性参数变化加剧了层间干扰现象,使得注水井分层配注量发生变化,仍按照设计分层注水量配注已不能满足实际注水开发的需求,也无法准确预测合理分层配注量,影响了注水井开发效果和原油采收率进一步提高。本文将考虑岩石骨架变形引起的体积应变作为表征时间变量的耦合参数作为变量引入模型,建立起考虑空间和时间变化的储层物性参数模型。通过储层物性参数时变模型预测孔隙度、渗透率的变化规律,分析分层注水中层间f扰规律,运用BP神经网络方法计算分层配注量及预测调整配注时间,优化分层注水调配方案。本文主要完成以下几方面研究内容：(1)分析注入水对储层物性参数的影响：注入水对储层孔隙度的影响体现在油田注水开发过程中储层内粘土矿物和其它微粒的机械搬运及聚散作用,导致孔喉半径增加,流体的流动能力增强,储层孔隙度也随之增大：注入水对于中高渗储层,储层渗透率在水洗过程中有一个先降低后增大的现象；注水开发以后,受注入水的影响,层间、层内流体性质矛盾突出,导致流体性质出现沿空间分布的非均质性,导致原油粘度在开发过程中增加。(2)考虑岩石骨架变形对于注水储层岩石孔隙变化的影响,将考虑骨架变形的岩石体积应变作为变量建立孔隙度、渗透率等储层物性参数动态变化模型；分别建立了以岩石体积应变为耦合变量的岩石骨架变形数学模型和储层流固耦合渗流模型；通过有限差分法和有限元法迭代求解储层流固耦合数学模型,得到不同时间不同空间下的储层孔隙压力分布和饱和度分布,进而得到相对应的岩石体积应变值,从而能够通过储层物性参数动态变化模型求得不同时间下的体积应变引起的物性参数变化值。(3)分析分层注水中单层吸水量的影响因素,引入BP神经网络的方法分析影响因素,建立新的计算小层吸水量的计算方法,并分析油层改造系数,渗透率,有效厚度,含油饱和度和孔隙度等重要影响因素之间的相互影响,得出考虑储层参数随时间变化的分层注水调配参数动态变化规律。(4)建立新的层段配水量优化调配方案,优化分层注水量、预测注水井配水调整时间：编制分层注水参数优化调配软件,针对提出的配注水量优化调配方案,对注水井注水量进行优化调配。应用此软件对某油田区块分层注水井进行计算,并与实际配注情况进行对比,结果较为精确；最后应用此软件对分层注水配水方案进行优化,编制出相应配注方案。综上可知,通过本文研究进一步明确注水后储层参数变化机理,建立储层物性参数的时变模型,提出分层配注的动态调配方法,对实现油田分层注水方案设计、调整、跟踪的一体化精细配注具有理论指导意义。更多还原

【Abstract】 During the process of reservoir water flooding development, because of the input well was continuously flooded and flushed, the physical property parameters of reservoir in different development phase have significant change. Interlayer interference would be happened in which layers interact longitudinal during layered water injection, obviously in mid-low permeability layers, and this problem could affect the design of development program and the project of injection allocation in late stage of water flooding development. In the past, the variation model of reservoir parameter was not considered the concept of time, so the variation law of reservoir parameter during a period of time could not be predicted and also could not be considered interlayer interference which caused by parameter variation meanwhile interlayer inference was increased and injection flow rate changed by this variation, so the designed layered water injection rate could not be satisfied actual requirement of water injection development, rational layered injection rate also could not be predicted precisely, development effect of injection well and oil recovery factor were affectedIn this paper, considering the cubic strain caused by deformation of rock skeleton as the couple parameter characterizing time variable to be called in as the pattern variation, the model of reservoir physical property parameters which considered variation of time and space was established. According to this model, variation law of porosity and permeability could be predicted and interlayer interference law of layered water injection was analyzed, the rate of layered injection allocation was calculated and time of injection allocation was predicted and program of layered water injection allocation was optimized by utilizing the method of BP neural network, The main research and achievements in this paper were finished as follows:(1) The affects of injected water on reservoir physical parameters were analyzed:The affect of injected water on reservoir porosity could be reflected in mechanical transportation and vergent action of clay mineral and other articles which could increase radius of pore throat, flowability and reservoir porosity also increased in development of water injection; To mid-high permeability layer, the affect of injected water could made reservoir permeability decreased at first stage and subsequently increased; After water injection, affected by injecting water, the contradiction of fluid properties between interlayer or in layer was prominent, which made distribution of fluid properties in space were inhomogeneous and viscosity was increased in water flood development.(2) Considered affect of rock deformation on variation of reservoir porosity with water injection reservoir. The dynamic model of reservoir physical parameters including porosity and permeability which considered rock volume strain as variable was established. With rock volume strain as coupling variable, solid deformation mathematical model and reservoir fluid-solid coupling permeability model were established; Through the finite difference method and finite element method, the reservoir fluid-solid coupling permeability model was solved, and reservoir pore pressure distribution and saturation distribution under different time and different space was got, and the corresponding rock volume strain value was also got. So physical parameters value caused by volume strain under different time through the reservoir physical property parameter dynamic model could be got..(3) Analyzed affects of individual water injection rate with BP neural network method in layered water injection, a new individual water injection rate calculation method was established. And the mutual affect among reservoir permeability coefficient, transformation analysis, effective thickness, oil saturation and porosity, and other important factors were analyzed with the new method. So variation law of layered water injection allocation parameters which considered changing parameters was got.(4) A new optimization allocation program of layered water injection rate was established. Using this new program, optimized the layered injection rate, predicted injection adjusted time. Layered water injection parameters optimization allocation software was work out. According to the proposed water injection allocation optimization program, water injection rate was optimized. Using this software in a particular injection, the calculated result compared with actual injection was accurately. Also water injection allocation program could be optimized and corresponding water injection allocation program could be worked out using with this software.Finally through the study of this thesis, the variation mechanism of reservoir parameters after water injection could be identified, and the dynamic model of reservoir physical parameters was established, dynamic allocation method of layered injection was proposed which have theoretical significance for delicate layered injection allocation to realize the integration of design, regulation and tracking.更多还原