节点文献
高阶煤层气储层测井评价方法及其关键问题研究
Research on Log Evaluation of High Rank Coalbed Methane Reservoirs and Some Key Issues
【作者】 黄兆辉;
【作者基本信息】 中国地质大学(北京) , 地球探测与信息技术, 2014, 博士
【摘要】 论文针对我国高阶煤层气储层测井现阶段需要解决的有关问题,以沁水盆地南部某研究区的煤岩实验分析资料和常规测井资料为基础,围绕煤层气储层测井评价技术,分别从地质影响因素、测井响应特征及井环境影响、测井解释模型等方面开展系列研究。通过煤层气的成因及其储层地质特征的总结,分析了有关煤层气储层物质组成、孔裂隙结构、渗透性特征、含气量的控制因素等内容,为煤层气测井的解释模型方法及参数选取提供有关地质理论依据。针对高阶煤层气储层的特点,通过数理统计方法,利用研究区现场和实验数据,分析了煤岩工业分析、显微组分、粘土矿物、镜质体反射率、含气量及各种常规测井响应值的分布特征,并根据煤岩工业分析组分、煤岩介质孔隙结构、煤层气赋存状态进行煤层气储层测井响应机理的探索。为了获取高质量的测井数据,针对煤层气储层扩径情况,采用蒙特卡罗数值模拟方法,模拟了小井径的煤层气井结构环境,完成了不同密度煤岩和不同扩径情况下密度测井响应的模拟计算,得到长、短源距计数率的变化特征,制作了煤层气储层扩径情况下的密度测井响应“脊肋图”,解释了部分高阶煤层气储层段密度低于1.1g/cm3的看似不合理值,为煤层气储层密度的扩径校正提供了图版法的解决方案。针对煤层气勘探开发过程中需要解决的有关煤质分析、煤层储集性能、天然气储量及煤岩顶底板评价等问题,利用地球物理测井资料,建立了工业组分、含气量、孔隙度、岩石强度等重要参数的定量解释模型。研究归纳了一些适应研究区高阶煤层气储层的改进预测方法。提出在工业组分计算当中,在样品数据充足情况下,优选基于统计模型的回归分析方法;对于小样品数据,采用BP神经网络算法模型;并在体积模型的基础上提出了适于高阶煤层气储层快速预测的约束最优化算法。提出以煤层平均数据为研究对象,采用多元逐步回归分析法,得到基于对数变换形式的工业组分和测井资料预测含气量方程。通过常规测井的声波、密度资料计算煤层及围岩的力学参数,并拟合了纵横波速度比预测抗压强度的回归公式。实例表明,上述方法在高阶煤层气储层测井预测中效果良好。
【Abstract】 Based on the coal bed experimental analysis data and conventional logging dataof a high rank coalbed methane study area in Southern Qinshui Basin, and takingcoalbed methane reservoir evaluation technique, the geological factors, loggingresponse and well environment as well logging interpreting model are systematicallystudied.In this thesis, the formation of coalbed methane and its reservoir geologicalcharacters are concluded, the reservoir material composition of coal bed, crackstructure, permeability characteristics, and controlling factors of gas contents arestudied, and then yield geological theoretical supports for the model interpretingmethods and parameter selection for coalbed methane logging. For high rank coalbedmethane reservoir, mathematical statistics are introduced to analyze the data fromfield data and experimental data, then deduced the distributions of all kinds of generalanalysis, like proximate analysis, microscopic components, clay minerals, vitrinitereflectance and gas contents, then explored the log response mechanism of coalbedmethane reservoir based on the proximate analysis, pore structure of coalbed, andstorage status of coalbed methane.In order to get high quality logging data with the situation of expanding ofcoalbed methane reservoir, Mont Carlo method is employed to simulate the structureenvironment of small diameter well of coalbed methane, and complete the density logsimulation of different coal bed densities and density logging response in differentexpanding situation, and then the count rate variation in long and short distance arededuced, a “spine-rib-plot” of density logging in different expanding, also solve theparadox of partial high rank coal bed reservoir density is less than1.1g/cc, whichprovides a solution to density expanding correction of coalbed methane reservoir.For issues should be solved during coalbed methane exploration like coal qualityanalysis, coal reservoir properties, natural gas reserves and roof and floor evaluationof coal bed, a quantitative interpretation models about important parameters likeproximate analysis, porosity, gas content, vitrinite reflectance, rock strength are constructed in this thesis. And sum up some developed interpreting method in highrank coalbed methane reservoir. That can be concluded as when calculate theproximate analysis, regression analysis method based on statistical models should beapplied with adequate sample data, meanwhile, if the sample data is not enough, BPneural network algorithm should be used. Finally, constrained optimization algorithmfor high rank coal bed methane quick predicting is proposed based on volume model.Propose stepwise multiple regression analysis method which taking coal bed averagedata as study objects to get the gas contents functions, which can be used to deducethe gas contents with logarithm of proximate components and logarithm of loggingdata respectively. Mechanical parameters of coal bed and surrounding rocks can becalculated via regular logging information like sonic and density data. Also,regression formula of predicting compressive strength via the ratio of compressionalto shear wave velocity is fitted. At last the case study showed that the above methodsworked well in logging evaluation of high rank coalbed methane reservoirs.
【Key words】 Coalbed Methane Reservoir; Logging Evaluation; Proximate Analysis; Gas Content; High Rank Coal;
- 【网络出版投稿人】 中国地质大学(北京) 【网络出版年期】2014年 08期
- 【分类号】P618.13;P631.81
- 【被引频次】3
- 【下载频次】849
- 攻读期成果