深部煤层气地质–工程一体化甜点区预测

Prediction of geological and engineering integrated sweet spots of deep coalbed methane

  • 摘要:
    背景 随着深部煤层气成为煤层气资源开发的热点领域,其埋深大、应力环境复杂、储层非均质性强等特征,严重制约了大规模开发中甜点区预测及井位精准部署。
    方法 以鄂尔多斯盆地东北缘某区块为研究对象,采用声波时差、密度、井径测井建立深部煤体结构指数M模型;基于深部煤层的煤体结构差异,引入煤体结构指数,结合上覆地层压力、构造应力、孔隙压力等因素建立深部煤层适应性水平地应力差模型;通过岩石强度参数、井径扩大率、岩石断裂韧性构建了天然裂缝指数模型;结合煤体结构、地应力差和天然裂缝指数模型,基于地质甜点和工程甜点2大类6项指标,采用支持向量机建立了深部煤层气地质–工程一体化甜点智能预测模型。
    结果 地质–工程一体化智能甜点预测精度为88.2%,研究区可划分为一类、二类和三类甜点区,一类甜点区面积为117.4 km2,占比14.0%,二类甜点区面积为258.4 km2,占比30.8%,三类甜点区面积为463.1 km2,占比55.2%,预测平均产能分别为6 478.6、5 076.7、4 022.0 m3/d。
    结论 研究区深部煤层气布井工作应重点立足一类甜点区、积极探索二类甜点区、主动规避三类甜点区,精细化地质–工程一体化甜点区预测对鄂尔多斯盆地东北缘深部煤层气增储上产具有重要指导价值。

     

    Abstract:
    Objective Deep coalbed methane (CBM) has emerged as a hot topic in CBM resource development. However, deep CBM has characteristics such as great burial depths, complex stress environments, and strong reservoir heterogeneity, which seriously restrict sweet spot prediction and accurate well location deployment in its large-scale exploitation.
    Methods This study investigated a deep CBM field along the eastern margin of the Ordos Basin. Using sonic, density, and caliper logging, this study developed a coal structure index model for deep coals. By introducing the coal structure index based on the coal structure differences in deep coal seams and combining factors including overburden formation pressure, tectonic stress, and pore pressure, this study established an adaptive horizontal in-situ stress difference model for deep coal seams. Based on the rock strength parameter, the enlargement rate of wellbore diameter, and the fracture toughness of rocks, a natural fissure index model was constructed. By integrating these three models, as well as the six indices of geological and engineering sweet spots, this study developed an intelligent prediction model of geological and engineering integrated sweet spots of deep CBM using support vector machine (SVM).
    Results The results indicate that the intelligent prediction model of geological and engineering integrated sweet spots yielded a prediction accuracy of 88.2%. Classes I, II, and III sweet spots were identified in the study area, with areas of 117.4 km2 (14.0%), 258.4 km2 (30.8%), and 463.1 km2 (55.2%), respectively, and average predicted production of 6478.6 m3/d, 5076.7 m3/d, and 4022 m3/d, respectively.
    Conclusions Based on the results of this study, it is recommended to focus on Class I sweet spots, actively explore Class II sweet spots, and proactively avoid Class III sweet spots in the well location deployment for deep CBM in the study area. The fine-scale prediction of geological and engineering integrated sweet spots can provide valuable guidance for reserve growth and production addition of deep CBM along the eastern margin of the Ordos Basin.

     

/

返回文章
返回