【作者】 薛家锋；

【导师】 贾振岐；

【作者基本信息】 大庆石油学院 ， 油气田开发工程， 2004， 博士

【摘要】 针对杏北开发区三角洲前缘相储层沉积特点，将油层精细划分为5种岩相单元，即水下分流河道砂、表内主体席状砂、表内非主体席状砂、表外层、泥质岩，然后引入随机建模理论，充分对比各种随机建模方法的优缺点，优选出模拟效率高、适用范围广的指示主成分模型。实际应用结果表明，该模型预测砂体分布具有符合程度高、计算速度快的特点；并且总结出砂体预测精度与随机模拟次数成正比、与各种岩相单元随机建模预测结果的偏离量成反比，沉积单元平面非均质性越强所需随机模拟次数越大等规律性认识，为储层三维地质建模工作奠定了坚实的理论基础。 在完成杏1-3区储层三维地质建模的任务后，开展了表外储层剩余油分布状况研究。由于非均质多油层油藏注水开发是一个具有随机过程的复杂动态系统，剩余油分布的影响因素多，且带有不确定性。系统多因素和多过程相互作用的相关性、可变性及协同性，显示出系统的非线性机制。根据油藏动态系统的上述特点，在精细地质描述研究成果的基础上，以井点资料为基础，采用神经网络模式法判别出薄差层的水淹程度和剩余油分布；依据地质、开发因素及相关性，利用多次模糊综合评判法和专家神经网络量化处理了系统的不确定性，绘制出剩余油的平面分布。这种多学科和多方位的综合研究及相互验证，确保了剩余油预测的高精度。 杏1-3区表外储层剩余油研究结果表明，该区块虽然经过二次加密和局部三次加密，但目前注水开发方式下表外储层动用程度仍然较差，存在大量剩余油。主要体现在水井注入压力高，油井产液量低，含水上升速度快，压裂、酸化等措施效果不明显。为此本文提出了注表面活性剂提高表外层开发效果的技术方法，该方面研究主要分三部分内容，表面活性剂驱油室内实验，数值模拟与注入方案设计，矿场试验。通过室内实验，深入分析阐述了低渗透岩心表面活性剂驱提高驱油效率和采收率的机理，测定了不同表面活性剂浓度时的相对渗透率曲线和表面活性剂在岩心中的吸附量，为数值模拟和注入方案设计提供了基础参数。根据活性剂驱油渗流特点，建立了三维两相三组分表面活性剂驱油数学模型。考虑了对流扩散和表面活性剂在油层中的吸附滞留，给出了相对渗透率曲线的处理方法，提高了数学模型与生产实际的符合程度。采用隐压、显饱、隐浓差分方法求解数学模型，并编制了相应的数值模拟软件，对杏1-3区表面活性剂驱试验区进行了实际计算。分析了表面活性剂浓度、段塞长度、注入量对开发指标的影响，并优选出了合理的注入方案。该方案在试验区实施后，取得了良好的开发效果和经济效果，为高效开发表外储层提高了技术手段和理论依据。更多还原

【Abstract】 According to the sedimentation features of delta frontal reservoir in Xing-north region, the reservoir was refined into five lithofacies units, as follow: underwater diffluence channel sand-body, tabulated principal part sheet sand, tabulated non-principal sheet sand, untabulated reservior, mudstone. Adopting to the theory of building model at random and fully contrasting advantages and disadvantages of different building model at random, to find out the best-indicated chief component model, with high simulation efficiency, and a large using scope. As the appliance result show: Not only the model was with the high quality to size up and simulate fast, but also summarize the direct proportion relation between prediction precision and the stochastic simulation times, and the opposite proportion relation between the former and the departure value of the prediction result in each lithofacies units’ stochastic simulation, the stronger in plant heterogeneity, the more the times stochastic simulation cost. All these offer a stability foundation for building 3D reservoir geologic model.After the reservoir geologic model for Xing-north 1-3 region was built, the following emphasis was pay on the institute of remaining oil distribution about untabulated layers. For the waterflooding in the heterogeneous multiplayer reservoirs is a complex and stochastic dynamic system, there are manifold features to generate remaining oil, and all these with uncertainty. Relativity, changeability and cooperateability, showed by the interaction between system multi-factors and system multi-course, reflect the non-linear mechanism of the system. According to the above characters in reservoir dynamic system, fine geologic description, and in terms of well point data, water out degree and remaining oil distribution of thinner and worse layer were distinguished through NN (neural network) pattern. According to geology, development factors, as well as relativity, the uncertainty of system was quantized using fuzzy comprehensive judgement method and expert NN, thus, the remaining oil distribution was plotted. This comprehensive research with multi-subject and multi-directions and interactive validation ensure the high precision during the predication of remaining oil.The remaining oil research results of Xing 1-3 region showed: although secondary and part tertiary infilling wells were drilled, the recovery percent of reserves of the untabulated reservoir can’t reach anticipative index. Main characters: the pressure in injection well is so high, and the liquid rate of production wells is low, water cut increases speedy, acidification and stimulation didn’t get anticipative achievements. Therefore the chapter gives out a method to enhance theoil--surfactant flooding indoor experiment, this research includes three sections:experiment in-door, numerical simulation and inject pattern design, pilot site experiment. By the experiment in-door, and enhancing the comprehension on the mechanism of surfactant enhancing displacement rate and oil recovery, measure relative permeability curve of different surfactant concentration and adsorbing in core, these offer a basic parameters for numerical simulation, and the injection plan design. According to seepage flow features of surfactant oil displacement, 3D3P3-component surfactant oil displacement mathematical model was built.Considering convection diffusion and surfactant’s adsorption retention in oil layer, giving the relative permeability curve’s treatment method, the match degree of mathematical model between production practice was enhanced. Using the difference method of IMPES, implicit concentration to solve mathematical equation, and compile the corresponding numerical simulation software, and the practical application was implemented in the surfactant flooding pilot site- Xing 1-3. Analyzing the impacts, generated by the density of the surfactant, the length of slug and the volume of injection, to development indexes, and reasonable injection plan is optimized. The good development effect and更多还原