煤矿底板采动变形及带压开采突水评判方法研究.pdf

博士学位论文 煤矿底板采动变形及带压开采突水评判 方法研究 Study on Mining Deation of Floor and uation of Water Inrush Mining above Confined Aquifer 江苏省普通高校研究生科研创新计划资助项目（CXZZ11-0306） 作 者段宏飞 导 师姜振泉 教授 中国矿业大学 二○一二年十二月 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） 。 作者签名段宏飞 导师签名姜振泉 2012 年 12 月 15 日 2012 年 12 月 15 日 中图分类号 TD12 学校代码 10290 UDC 密 级 公开 中国矿业大学 博士学位论文 煤矿底板采动变形及带压开采突水评判 方法研究 Study on Mining Deation of Floor and uation of Water Inrush Mining above Confined Aquifer 作 者 段宏飞 导 师 姜振泉 教授 申请学位 工学博士 培养单位 资源与地球科学学院 学科专业 地质资源与地质工程 研究方向 水文地质与工程地质 答辩委员会主席 魏久传 评 阅 人 盲审 二○一二年十二月 论文审阅认定书论文审阅认定书 研究生 段宏飞 在规定的学习年限内，按照研究生培养方案的要求， 完成了研究生课程的学习，成绩合格；在我的指导下完成本学位论文,经审 阅，论文中的观点、数据、表述和结构为我所认同，论文撰写格式符合学 校的相关规定，同意将本论文作为学位申请论文送专家评审。 导师签字姜振泉 2012 年 12 月 15 日 致致 谢谢 本文是在导师姜振泉教授的精心指导下完成的。在攻读博士学位期间，导师在论文 选题、研究思路及论文的成稿整个过程都倾注了大量的心血，给予了悉心的指导。导师 深厚而又渊博的学术造诣、开阔而又敏捷的思维模式、严谨而又创新的治学态度、踏实 而又求实的工作作风和兢兢业业而又孜孜不倦的敬业精神， 无时无刻不在激励和鞭笞着 我，给我营造了极好的学习氛围，使我在科研上受益匪浅。在这三年里，学生一点一滴 的成长无一不渗透着导师的心血，导师在思想、生活上对学生的无私关怀和照顾，学生 会一直铭记在心。值此论文完成之际，借这张薄薄的纸向老师表达最崇高的敬意和最真 诚的感谢。 衷心感谢我的硕士导师吴圣林副教授一直以来对作者学业和生活上的关心和帮助， 是吴老师启蒙我对所学专业领域产生了兴趣，并在学业上给予了我不断的鼓励和指导 衷心感谢我的大学班主任董青红教授、 王档良副教授一直以来对作者学业和生活上的关 心和帮助，每逢遇到困难，二位班主任都倾力帮忙，给予了作者很多援助师恩难忘 在论文选题过程中，得到了隋旺华教授、李文平教授、孙亚军教授、曹丽文教授、 董青红教授和孙强副教授等的大力帮助，他们对作者的选题、研究思路和文献查阅及论 文的撰写等提出了许多有益的建议和帮助，同时也得到了他们在学习和生活上的帮助， 在此作者表示衷心的感谢 在论文写作过程中，矿业学院邹喜正教授在采场支承压力模型、模型参数的选取方 面方面给予了指导； 资源学院郑世书教授在底板突水灾害问题、 突水评判模型构建方面， 多次给予了悉心指导；朱术云副教授在支承压力模型计算、底板破坏深度实测及破坏深 度预测模型的构建上给予了很多思路； 安徽皖西学院雷能忠教授在非线性预测模型方面 给予了很多思路；他们严谨的工作态度和乐于助人的精神使我深受感动，在此对他们表 示衷心的感谢 感谢本校资源与地球科学学院各位老师在我九年矿大学习生涯中， 在三尺讲台上的 倾囊相授和所给予的指导和帮助。 感谢各位师兄弟和姐妹们尤其是与姜春露博士、胡巍博士、钱自卫博士、张蕊博 士、王海硕士、安宝剑硕士、徐瑞朋硕士、任申硕士、刘芙蓉硕士、常刚硕士、黄震硕 士、陈凯硕士、王一栋硕士、王震硕士、李帅硕士、史继彪硕士等之间的坦诚学术交流， 使我获益颇深。 感谢兖州矿业集团杨村煤矿、兴隆庄煤矿、东滩煤矿、鲍店煤矿、赵楼煤矿的领导 与技术人员在作者底板破坏深度现场测试中的给予的关照与帮助 特别感谢我的家人多年来的默默支持、鼓励和理解，给了我强有力的精神支柱和勇 敢前行的勇气。 最后，感谢各位专家在百忙之中对论文评审及提出宝贵意见 I 摘摘 要要 煤矿底板采动变形问题的研究，不仅对于承压水上带压安全开采具有科学价值，而 且可为采区巷道围岩变形控制维护提供关键依据。 本文对采动底板应力分布特征、 底板应力及其塑性区分布以及底板破坏深度等变形 破坏规律进行了系统研究，研究发现在前人对支承压力分布规律研究的基础上构建的 采场完整支承压力作用下的底板应力分布模型能够得到底板内任一位置的应力分布解 析解，通过杨村煤矿 4602 工作面的底板应力解析计算说明解析得出的底板下的应力分 布规律具有实用性；从底板采动变形的角度厘定了底板破坏深度的概念，并采用 FLAC3D 数值模拟软件开展了底板破坏深度斜长、顶底板岩性组合、采深、采高、倾角 的六因素五水平正交数值模拟试验，构建了首次考虑顶板岩性组合这一因素的斜长-顶 底板岩性组合-采深-采高-倾角的底板破坏深度预测模型，通过 10 个煤矿相应的工作面 底板破坏深度实测实例进行分析验证了该预测模型精度较高，可以满足工程使用。 通过现场底板变形破坏综合实测深刻揭示了底板矿压显现过程及其分区特点、 底板 破坏深度分区特征以及底板变形与矿压显现的关联规律，具体表现为采动矿压对底板 的影响具有较远距离的采前“超前”显现和采后“滞后”延续的特点，且这种“超前” “滞后”影响具有分区特征；从采动底板变形与采动矿压的关联效应角度将底板所受采 动矿压的扰动作用分为“超前聚压扰动”和“采后卸压扰动”两种类型，为合理解释底 板采动变形破坏的力学机制提供了力学依据； 底板破坏经历超前聚压破坏和采后卸压破 坏两个过程，其破坏机制均为剪切破坏，前期为受压状态下的剪切破坏，后期为受拉状 态下的剪切破坏，二者具有累进关系，即在采前聚压破坏基础上采后卸压破坏可导致破 坏程度进一步加剧，但对于采动破坏深度不具有延伸效果。 在 70 组底板破坏深度实测资料统计的基础上，探索了底板破坏深度与其影响因素 的规律， 构建了底板破坏深度预测的遗传-改进遗传算法优化 BP 神经网络模型 （BP-GA、 BP-GA-MOD）和 PSO 优化 SVM 模型（PSO-SVM） 。基于 MATLAB 软件平台，编制了 底板破坏深度非线性预测模型系统，能够实现底板破坏深度快速准确的预测。 最后，首次明确建立了煤矿底板突水的三级评判模型，以此实现对煤矿底板突水 由粗到细、 由经验判别到力学分析的多级分布筛选递进评价预测。 三级评判分别包括 煤矿底板突水初判、煤矿底板突水详判与煤矿底板突水精判。底板突水初判判据为 P0.0025M2-0.0865M-16.8534/M2.2440 临界方程， 该方程是通过对华北七个矿区以及 湖南涟邵矿区 354 个工作面突水点、202 个巷道突水点以及 318 个安全回采工作面的 调查分析，应用数学统计回归的方法所得。该临界方程形式与考虑动水压力作用下的 底板均质裂隙弱板模型 P-M 临界方程形式一致；煤矿底板突水的详判，是在初判的基 础上对可能发生突水的区域进行的进一步突水判别，综合指标法考虑了影响底板突水 II 的诸多因素，以此判别将更为全面，考虑模型的简化，构建了基于膨胀界限抗渗强度 的底板突水评判模型。煤矿底板突水精判，需要对底板岩体特性、含水层特性以及所 赋存的地质条件进行更为详细的勘查，掌握突水评判区域非常精细准确的第一手资 料，在此基础上，对底板进行力学稳定性分析，确定底板在采动矿压、水压作用下的 潜在稳定性。 该论文有图 73 幅，表 31 个，参考文献 207 篇。 关键词关键词底板；采动变形；矿压显现；底板破坏深度；突水；三级评判方法 III Abstract The study of mining deation of floor is not merely of great scientific value for safety mining of coal mining above confined aquifer, but also provides crucial basis for control and maintenance of surrounding rock in mining roadway. This dissertation aims to make systematic studies involving such aspects as stress distribution of mining floor, floor stress and plastic zone distribution as well as deation failure law such as floor failure depth. The research results show that stress distribution model of floor rock under the complete bearing pressure in the face based on the previous studies on distribution regularities of bearing pressure enables to obtain the analytical solution of stress distribution in any position inside the floor rock. Meanwhile, the stress distribution regularities computed by the stress of floor rock of mining face 4602 in Yangcun is of great practical applicability. And, the notion of floor rock failure depth is apportioned from the perspective of mining floor rock deation. By employing numerical modeling software FLAC3D, six-factor and five-level orthogonal numerical simulation test of floor failure depth concerning working face length, lithological association of roof rock and floor rock, mining depth, mining height and dip angle was launched. Furthermore, the prediction model of floor rock failure depth where the lithological association of roof rock is considered for the first time which includes working face length, lithological association of roof rock and floor rock, mining depth, mining height and dip angle. On the basis of actual measurement and instances on floor rock failure depth of corresponding mining faces in ten coal mines, it can be concluded that this prediction model involves high precision which can meet the requirement of project. By on-site comprehensive actual measurement of floor rock deation failure, the process and partition features of mine pressure behavior on floor rock, partition features of floor rock failure depth as well as relevance law between floor rock deation and underground pressure behavior were expounded in great details. In other words, by comprehensive actual measurement of floor rock deation, the influence of mining underground pressure on floor rock is characterized by manifestation of pre-mining “leading” at a long-distance and continuity of post-mining “lagging”. In addition, both “leading” and “lagging” impacts have partition features. Based on the linkage effects between mining floor rock deation and mining underground pressure, the perturbation action of mining underground pressure on floor rock can be classified into “leading pressure-gathering perturbation” and “post-mining pressure-relieving perturbation”, which provided the IV mechanical basis for explaining the mechanism of mining floor rock deation failure. What’s more, floor rock undergoes two processes of leading pressure-gathering breakage and post-mining pressure-relieving breakage. And, their breakage mechanisms are shear breakdowns which are under the pressing condition in the early stage and is under the tension condition is the late stage. They are related in progression. That’s to say, post-mining pressure-relieving breakage made after pre-mining pressure-gathering breakage can lead to further aggravation of failure degree. However, it involves no extension effects as to mining failure depth. On the basis of actual measurement data on 70 groups of floor rock failure depth, floor rock depth and the regularities of its influential factors were explored. Besides, the predicted genetic-improved genetic algorithm optimization BP neural network model and PSO optimization SVM model by floor rock failure depth were constructed. Based on MATLAB software plat, non-linear predicted model system of floor rock failure depth was compiled, which could achieve fast and accurate prediction of floor rock failure depth. Finally, three-level uation model of floor rock water inrush in mines was established in order to achieve multistage distribution and screening progressive uation prediction from thickness to thinness and from empirical judgment to mechanical analysis on floor rock water inrush in mines. To be specific, three-level uation model refers to primary judgment, detailed judgment and accurate judgment of floor rock water inrush in mines. Firstly, the criterion of primary judgment is critical equation P0.0025M2-0.0865M-16.8534/M2.2440 which is obtained by investigating and analyzing water inrush points of 354 mining faces in seven mining areas in North China and Lianshao mining area in Hunan Province, 202 roadway water inrush points and 318 safety actual mining faces, and then adopting the of mathematical statistical regression. The of this critical equation is in accordance with that of weak plate of homogeneous crack model P-M in terms of the consideration of dynamic water pressure. As for the detailed judgment of floor rock water inrush in mines, it is a further judgment on the possible water inrush occurrence areas based on the primary judgment. What’s more, aggregative indicator has taken several influential factors on floor rock water inrush into account. Due to this judgment, it is much more comprehensive. Considering model simplification, the uation model of floor rock water was built up based on inpervious intensity of expanding boundary. When it comes to accurate judgment, it entails more specific prospecting on floor rock characteristics, aquifer characteristics and occurring geological conditions. In this way, the meticulous and accurate first-hand data of water inrush uation areas needs to be mastered. Based on the above, mechanical stability analysis of floor rock was analyzed and then potential stability of floor V rock under the influence of mining underground pressure and water pressure was determined. There are 73 Figures, 31 Tables, and 207 references in the dissertation. Keywords floor; mining deation; underground pressure behavior; floor failure depth; water inrush; three-level uation VI Extended Abstract The study of mining deation of floor is not merely of great scientific value for safety mining of coal mining above confined aquifer, but also provides crucial basis for control and maintenance of surrounding rock in mining roadway. This dissertation aims to make systematic studies involving such aspects as stress distribution of mining floor, floor stress and plastic zone distribution as well as deation failure law such as floor failure depth. The research results show that stress distribution model of floor rock under the complete bearing pressure in the face based on the previous studies on distribution regularities of bearing pressure enables to obtain the analytical solution of stress distribution in any position inside the floor rock. Meanwhile, the stress distribution regularities computed by the stress of floor rock of mining face 4602 in Yangcun is of great practical applicability. And, the notion of floor rock failure depth is apportioned from the perspective of mining floor rock deation. By employing numerical modeling software FLAC3D, six-factor and five-level orthogonal numerical simulation test of floor failure depth concerning working face length, lithological association of roof rock and floor rock, mining depth, mining height and dip angle was launched. Furthermore, the prediction model of floor rock failure depth where the lithological association of roof rock is considered for the first time which includes working face length, lithological association of roof rock and floor rock, mining depth, mining height and dip angle. On the basis of actual measurement and instances on floor rock failure depth of corresponding mining faces in ten coal mines, it can be concluded that this prediction model involves high precision which can meet the requirement of project. By on-site comprehensive actual measurement of floor rock deation failure, the process and partition features of mine pressure behavior on floor rock, partition features of floor rock failure depth as well as relevance law between floor rock deation and underground pressure behavior were expounded in great details. In other words, by comprehensive actual measurement of floor rock deation, the influence of mining underground pressure on floor rock is characterized by manifestation of pre-mining “leading” at a long-distance and continuity of post-mining “lagging”. In addition, both “leading” and “lagging” impacts have partition features. Based on the linkage effects between mining floor rock deation and mining underground pressure, the perturbation action of mining underground pressure on floor rock can be classified into “leading pressure-gathering perturbation” and “post-mining pressure-relieving perturbation”, which VII provided the mechanical basis for explaining the mechanism of mining floor rock deation failure. What’s more, floor rock undergoes two processes of leading pressure-gathering breakage and post-mining pressure-relieving breakage. And, their breakage mechanisms are shear breakdowns which are under the pressing condition in the early stage and is under the tension condition is the late stage. They are related in progression. That’s to say, post-mining pressure-relieving breakage made after pre-mining pressure-gathering breakage can lead to further aggravation of failure degree. However, it involves no extension effects as to mining failure depth. On the basis of actual measurement data on 70 groups of floor rock failure depth, floor rock depth and the regularities of its influential factors were explored. Besides, the predicted genetic-improved genetic algorithm optimization BP neural network model and PSO optimization SVM model by floor rock failure depth were constructed. The findings indicate that the predicted precision of PSO-SVM model is a little higher than that of BP-GA-MOD model, and the predicted precision of BP-GA-MOD model is a little higher than that of SVM model, and the predicted precision of SVM model is a little higher than that of BP-GA. However, there is minute difference among the predicted precision of four models, which can achieve the needs of project. Also, the precision of four models is far higher than the empirical ula in “Mining under Three Regulations”. Yet, it is basically equal to the precision of floor rock failure depth on predicted model which involves working face length, lithological association of roof rock and floor rock, mining depth, mining height and dip angle. Besides, the precision of predicted model is relatively lower, which cannot meet the requirements of project. Based on MATLAB software plat, non-linear predicted model system of floor rock failure depth was compiled, which could achieve fast and accurate prediction of floor rock failure depth. F