陕北-陇东地区侏罗系煤层采动破坏及顶板突水危险性评价方法研究.pdf

国家自然科学基金重点项目41430643 国家自然科学基金与神华集团有限责任公司联合资助项目U1261202 国家自然科学基金面上项目41774128, 41374140 硕士学位论文 陕北-陇东地区侏罗系煤层采动破坏及顶板 突水危险性评价方法研究 Study on Risk Assessment of Jurassic Coal Seam Mining Failure and Roof Water Inrush in Shanbei-Longdong Area 作 者韦 瑜 导 师陈同俊 教授 中国矿业大学 二○一八年五月 学位论文使用授权声明学位论文使用授权声明 本人完全了解中国矿业大学有关保留、使用学位论文的规定，同意本人所撰 写的学位论文的使用授权按照学校的管理规定处理 作为申请学位的条件之一， 学位论文著作权拥有者须授权所在学校拥有学位 论文的部分使用权，即①学校档案馆和图书馆有权保留学位论文的纸质版和电 子版，可以使用影印、缩印或扫描等复制手段保存和汇编学位论文；②为教学和 科研目的，学校档案馆和图书馆可以将公开的学位论文作为资料在档案馆、图书 馆等场所或在校园网上供校内师生阅读、浏览。另外，根据有关法规，同意中国 国家图书馆保存研究生学位论文。 （保密的学位论文在解密后适用本授权书） 。 作者签名 导师签名 年 月 日 年 月 日 中图分类号 P631.4 学校代码 10290 UDC 550 密 级 公开 中国矿业大学 硕士学位论文 陕北-陇东地区侏罗系煤层采动破坏及顶板 突水危险性评价方法研究 Study on Risk Assessment of Jurassic Coal Seam Mining Failure and Roof Water Inrush in Shanbei-Longdong Area 作 者 韦 瑜 导 师 陈同俊 教授 申请学位 工学硕士 培养单位 资源与地球科学学院 学科专业 地球探测与信息技术 研究方向 应用地球物理 答辩委员会主席 潘冬明 教授 评 阅 人 二○一八年五月 致谢致谢 硕士三年，如白驹过隙，倏忽而逝。在即将过去的硕士生涯中，无论在学习 上还是生活上，老师、同学、亲人和朋友都给予了我极大的关怀与帮助，值此毕 业之际，特向他们致以最诚挚的谢意 本论文是在陈同俊教授的悉心指导下完成的。 从论文选题到论文撰写编排的 每个环节都渗透着陈老师的精心指导。在研究过程中，陈老师针对我所遇到的问 题，为我指点迷津、开拓思路，正是这细致入微的指导和关怀才使得本论文得以 顺利完成。本论文的顺利完成还离不开崔若飞教授的关心和指导。从研究过程到 论文撰写，崔老师给予了我无私的帮助和大量的建设性意见。陈老师和崔老师高 屋建瓴的研究思路、渊博的专业知识、严谨缜密的治学态度，都为我树立了光辉 典范，而他们的谆谆教诲也使我终生受益。在此，谨向我的导师---陈同俊教授以 及时刻关心指导我的崔若飞教授致以崇高的敬意和由衷的感谢。 同时，还要感谢资源学院应用地球物理研究所的各位老师。多年来，他们默 默地传授知识，并为我提供了许多学习和科研的机会。我深深地感到我的每一点 成长都与他们的无私奉献是分不开的，在此向他们表示深深的感谢。 另外，我要感谢我的师兄彭刘亚、张亚兵、赵立明、代琦、管永伟、张明川， 我的同学马国栋，以及我的师弟宋雄和金俊俊。他们在平时的研究学习中，为我 提供了很多帮助，也为实验室营造出了的良好研究氛围。他们的支持、帮助和鼓 励使我愉快充实地度过了硕士三年的学习生活。 此外，还要特别感谢江晓雨同学，在我平时的学习中和完成论文期间对我的 无私帮助。 最后，我还要感谢我的父母、亲戚和朋友。无论在什么时候，他们始终都在 背后支持我，鼓励我，给我坚持不懈的勇气。 I 摘摘 要要 针对陕北地区侏罗系浅埋煤层采动引起浅表层水资源流失和陇东地区煤矿 开采引发顶板突水事故问题，以水文地质理论为指导，利用测井、三维地震、 Gassmann 流体替换、波阻抗反演、概率神经网络反演等多种手段，系统研究了 侏罗系浅埋煤层采动破坏和顶板突水的各种影响因素。 将层次分析方法应用到陕 北和陇东地区，分别建立了不同的评价模型，成功预测了 N2红土层采动破坏及 煤层顶板突水危险性，预测结果与实际揭露的地质资料高度吻合。 对于侏罗系浅埋煤层采动破坏的预测， 以陕北地区小保当二号井为研究目标 区，对其 1 号煤层之上，浅表层水资源下直接隔水层（新近系上新统保德组红粘 土层，即 N2红土层）的采动破坏进行预测。由于影响红土层采动破坏的因素包 括 （1）红土层厚度、强度和孔隙度等内部因素； （2）红土层与 1 号煤层间距等 外部因素。预测时，以区内 17 个钻孔揭露值为输入，利用 Kriging 插值计算得 到红土层厚度；以反演的波阻抗值为输入，依据红土层强度、杨氏模量与波阻抗 之间的正相关关系，计算了红土层强度；以提取的多个地震属性为输入，利用概 率神经网络反演得出了红土层孔隙度； 以三维地震解释的层位时间和波阻抗反演 的层速度为输入，计算了红土层与 1 煤层的间距。最后，利用层次分析法计算出 各影响因素的权重系数，预测了研究区红土层的采动破坏。 对于煤层顶板突水危险性的预测，以陇东地区崔木煤矿为研究区，对 3 煤层 顶板突水危险性进行预测。 由于研究区影响突水危险性的因素主要包括主要含水 层富水性、导水层导水性、隔水层隔水性和隔水层厚度等四类。其中，含水层富 水性受其厚度、 含水性和孔隙度等三个因素影响； 导水层导水性受其孔隙度影响； 隔水层隔水性受其砂质含量影响。预测时，以多个地震属性为输入，利用概率神 经网络反演得出了主要含水层和导水层的孔隙度； 以三维地震解释层位和波阻抗 反演的速度为输入，计算了主要含水层和隔水层厚度；以反演的波阻抗数据为输 入，根据隔水层隔水性与波阻抗的负相关关系，计算了隔水层的隔水性；以反演 的波阻抗数据和密度数据为输入，依据 Gassmann 流体替换理论，计算了主要含 水层的含水性。最后，利用层次分析法计算出各主要影响因素的权重系数，预测 研究区 3 煤层顶板突水危险性。通过将预测结果与实际突水点对比，二者吻合较 好，验证了本次预测方法的可行性与准确性。同时，利用 GA-BP 神经网络模型， 以煤层埋深、煤层倾角、覆岩层段波阻抗值、工作面斜长和采厚等五个因素为输 入，预测了全区连续分布的导水裂隙带发育高度。 本文以侏罗系浅埋煤层采动破坏及顶板突水为切入点， 提出了利用测井和三 维地震等多源数据的综合预测方法， 预测了生态脆弱地区的煤层采动破坏和顶板 II 水害， 预测成果为西部生态地质环境脆弱区侏罗系煤层开采提供了可靠的地质依 据。 该论文有图 69 幅，表 18 个，参考文献 131 篇。 关键词关键词N2红土；采动破坏；顶板水害；三维地震；Gassmann 流体替换；层次 分析法；神经网络 III Abstract Aiming at the problems of shallow water-resource losses and roof water-inrush accidents induced by the mining of shallow buried coal, I systematically study the prediction s and technologies considering the influence factors of mining destruction and roof water inrush under the guidance of hydrogeology theory. The used data and technologies include logging data, 3D seismic data, gassmann fluid substitution theory, well-log constrained impedance inversion, PNN inversion, genetic algorithm optimized BP neural network and analytic hierarchy process. Applying related s and technologies to Shanbei and Longdong areaes, I found that the proposed can predict the mining failure of N2 laterite in Shanbei area and the water inrush risk of coal roof in Longdong area. The prediction results are highly consistent with true uncovered geological data. For the prediction of mining induced failure of shallow buried coal seam, I use Xiaobaodang No.2 mine in Shanbei area as the research area, where the predicted aquifer The red clay layer of Neogene System Pliocene series Baode ation, N2 laterite is sandwitched between shallow water resources and the No.1 coal seam. The factors affecting the mining failure include 1 internal factors such as thickness, strength and porosity of laterite; 2 external factors such as the interval between laterite and No. 1 coal seam. During prediction, the thickness of N2 laterite is calculated with Kriging interpolation using true 17 well data as s; the strength of N2 laterite is calculated with inversed acoustic impedance and the positive correlation association between the Youngs modulus and impedance; the porosity of N2 laterite is calculated with PNN inversion from multiple seismic attributes; the interval distance between N2 laterite and No.1 coal seam is calculated with 3D seismic interpreted horizons and seismic inversed interval velocity. Finally, the weight coefficient of each factor is calculated with analytic hierarchy process. In this way, I predicted the mining failure to N2 laterite in the study area. For the prediction of water inrush risk of coal roof, I use the roof of No.3 coal seam in Cuimu coal mine in Longdong area as the research target. The factors affecting the risk of water inrush mainly include four types, including water enrichment of main aquifer, water conductivity of water conducting stratum, water resistance of aquifuge and bed thickness of aquifuge. The water enrichment of main aquifer is affected by its thickness, water bearing property and porosity. The water IV conductivity of water conducting stratum is affected by its porosity. And the water resistance of aquifuge is affected by its sandy content. The porosity of main aquifers and water conducting stratum is calculated with PNN inversion using multiple seismic attributes as s. The thickness of main aquifer and aquifuge is calculatedwith 3D seismic interpreted horizons and seismic inversed interval velocity. The water resistance of aquifuge is calculated with seismic inversed acoustic impedanceand negative correlation association between water resistance and aquifuge impedance. The water bearing property of main aquifer is calculated with seismic inversed acoustic impedance and density and Gassmann fluid substitution theory. Finally, the weight coefficient of each factor is calculated with analytic hierarchy process. In this way, I predict the risk of water inrush from coal roof in the study area. The predicted results is in accordance with the actual observed water inrush points, verifing the accuracy and feasibility of the prediction . Meanwhile, I use GA-BP neural network model ted with the factors such as depth of coal seam, dip angle of coal seam, impedance of the overlying strata, length ofmining panel and mining thickness to predict the height of water conducting fissure zone in whole area. Keywords N2 laterite; Mining failure; roof water hazards; 3D seismic; Gassmann fluid substitution; analytic hierarchy process; neural network V 目 录 目 录 摘摘 要要 ........................................................................................................................... I 目 录目 录 .......................................................................................................................... V 图清单图清单 ........................................................................................................................ IX 表清单表清单 ..................................................................................................................... XIII 1 绪论绪论 ........................................................................................................................... 1 1.1 选题目的及意义...................................................................................................... 1 1.2 国内外研究现状...................................................................................................... 2 1.3 主要研究内容与技术路线...................................................................................... 9 1.4 主要创新点............................................................................................................ 10 2 水文地质基础水文地质基础 ......................................................................................................... 12 2.1 水害类型............................................................................................................... 12 2.2 含水层富水性....................................................................................................... 13 2.3 覆岩隔水性........................................................................................................... 14 2.4 导水裂隙带发育高度及有效隔水层厚度计算................................................... 15 3 岩性解释技术理论基础岩性解释技术理论基础 ......................................................................................... 17 3.1 波阻抗反演............................................................................................................ 17 3.2 概率神经网络反演............................................................................................... 20 3.3 基于 Gassman 方程的流体替换 .......................................................................... 25 3.4 遗传算法与 BP 神经网络 .................................................................................... 26 3.5 层次分析法........................................................................................................... 29 4 陕北地区陕北地区 N2红土层采动破坏预测红土层采动破坏预测 ....................................................................... 32 4.1 研究区概况............................................................................................................ 32 4.2 评价模型建立........................................................................................................ 34 4.3 基于波阻抗反演的红土层岩性解释.................................................................... 36 4.4 基于概率神经网络反演的红土层岩性解释........................................................ 39 4.5 评价指标及权重确定............................................................................................ 41 4.6 综合评价................................................................................................................ 43 4.7 小结........................................................................................................................ 44 5 陇东地区煤层顶板突水危险性预测陇东地区煤层顶板突水危险性预测 ..................................................................... 46 VI 5.1 研究区概况............................................................................................................ 46 5.2 评价模型建立........................................................................................................ 48 5.3 基于 Gassmann 方程的主要含水层含水性研究 ................................................. 52 5.4 基于波阻抗反演的岩性解释................................................................................ 58 5.5 基于概率神经网络反演的孔隙度研究................................................................ 65 5.6 评价指标及权重确定............................................................................................ 68 5.7 综合评价................................................................................................................ 70 5.8 基于 GA-BP 模型的导水裂隙带高度预测.......................................................... 71 5.9 小结....................................................................................................................... 75 6 结论结论 .......................................................................................................................... 77 6.1 主要研究成果........................................................................................................ 77 6.2 存在不足与展望.................................................................................................... 78 参考文献参考文献 ..................................................................................................................... 80 作者简历作者简历 ..................................................................................................................... 87 学位论文原创性声明学位论文原创性声明 ................................................................................................. 88 学位论文数据集学位论文数据集 ......................................................................................................... 89 VII Contents Abstract ...................................................................................................................... III Contents .................................................................................................................... VII List of Figures ............................................................................................................ IX List of Tables ........................................................................................................... XIII 1 Introduction ............................................................................................................... 1 1.1 Purpose and Significance of the Subject .................................................................. 1 1.2 Research Reviews .................................................................................................... 2 1.3 Main Research Contents .......................................................................................... 9 1.4 Innovations ............................................................................................................. 10 2 Hydrogeological basis ............................................................................................. 12 2.1 The type of water disasters .................................................................................... 12 2.2 The watery characteristics of aquifer ..................................................................... 13 2.3 Water resistance of overlying strata ....................................................................... 14 2.4 Calculation of development height of water conducting fissure zone and thickness of effective aquifuge .................................................................................................... 15 3 Theories of Lithology Interpretation Technologies .............................................. 17 3.1 Impedance Inversion .............................................................................................. 17 3.2 Probabilistic Neural Network Technique ............................................................... 20 3.3 Fluid substitution based on Gassman equation ...................................................... 25 3.4 Genetic algorithm and BP neural network ............................................................