【作者】 付雅琴；

【导师】 王乘； 张秋文；

【作者基本信息】 华中科技大学 ， 系统分析与集成， 2009， 博士

【摘要】 流域梯级水电开发能够高效利用水能资源，自“六五”开始，中国一直将大力推进流域梯级水电开发作为重要的能源战略。然而近二十年来，水电站引发的负面生态环境影响日益受到关注，与单个水电站相比，一条河流上的多个梯级水电站所引发的生态环境影响，其成因更复杂、范围更广泛。如何全面、整体、系统地分析和评价这些生态环境影响，进而强化其中的正面因素，缓解、控制甚至避免负面因素，对于提升中国流域水电梯级开发的可持续性至关重要。梯级水电站群体通常包含一个或多个具有较强调节能力的大型水电站，将一条河流分割成为数个河段，并且可以进行联合调度。与单个水电工程相比，在一条河流上的各梯级水电站作为一个整体，其生态环境影响既根源于单个水电站的生态环境影响，又因为电站之间的系统性和关联性而可能增强或削弱单个水电站的生态环境影响，甚至在流域内产生新的累积影响——各级水电站与所在流域的各类生态环境要素相互作用进而构成了一个复杂系统，即“梯级复合生态环境系统”。分析和评价梯级水电开发的生态环境影响，不仅需要在各生态环境要素所属的专业学科领域内分门别类地开展研究，更需要基于系统科学，尤其是复杂系统科学，多学科交叉地构建更为全面的生态环境影响评价理论与方法框架。这样的理论与方法框架能够增强各专业领域研究成果的内在联系，并有助于获取针对梯级水电开发生态环境影响的整体认识。然而，我国目前有关梯级水电开发生态环境影响评价的研究主要集中于专业学科领域，十分缺乏基于系统视角的整体评价模型和评价方法。因此，探讨如何基于(复杂)系统科学，多学科交叉地构建梯级水电开发生态环境影响的整体分析与评价模型，对于理论研究和实际应用都有重要意义，既能够体现梯级水电开发生态环境影响评价的研究趋势，也能够迎合我国流域水电可持续发展的实际需求。本文就是在系统科学的指导下，构建梯级水电开发生态环境影响分析与评价的整体模型，并探讨各类适于该模型的生态环境影响系统分析和评价方法。论文不仅注重增强研究内容的学科交叉性、综合性、普适性与全面性，使研究成果能够对更多的(梯级)水电工程生态环境影响评价工作起到参考作用；而且注重提升系统分析与评价方法的结构化，使评价过程便于维护、易于计算机化。论文的主要研究结果如下：(1)归纳综合国内外梯级水电开发实例，类比分析梯级水电开发与单项水电开发在各类生态环境影响方面的相同点与不同点，进而提取出梯级水电开发生态环境影响的基本特征。研究表明：梯级水电开发的基本模式可以分为3大类：Ⅰ型(龙头-径流复合式)、Ⅱ型(连续调节式)、Ⅲ型(连续径流式)；梯级水电开发对生态环境造成的负面影响，在河流破碎化、河流生境改变、水体自净能力变化以及流域景观格局等方面比单项水电工程要显著的多，但与此同时，梯级水电开发具备更多可以调控这些生态环境影响的环节；梯级水电开发生态环境影响的基本特征为整体性、累积性、可控性；给出梯级水电开发的主要生态环境影响因素清单，在其中选取“突出生态环境效益”以及“敏感生态环境问题”作为后续的重点研究内容。(2)定义梯级复合生态环境系统，并基于系统动力学(System Dynamics，SD，一种复杂系统分析方法)的因果反馈分析理论，对梯级水电开发生态环境影响的产生过程进行机理分析。研究表明：梯级复合生态环境系统是复杂系统，也是梯级水电开发生态环境影响分析与评价的研究对象；该系统在5个典型场景下的机理分析结果一方面描述了各种累积影响的影响路径，另一方面证明了梯级水电开发生态环境影响的“可控性”与“生态调度”策略之间的密切关系。(3)基于复杂系统静态结构建模方法，构建梯级水电开发生态环境影响回顾评价模型。研究成果包括：回顾评价的概念模型和评价流程，回顾评价指标体系，回顾评价方法，以及基于UML的“水体富营养化”回顾评价的逻辑模型。结果表明：以模糊数学为基础的影响程度评判过程对于目前的数据积累水平而言，是可行而有效的；基于UML构建面向对象的逻辑模型，该模型与面向过程的逻辑模型相比，在可重用性、可维护性等方面具有明显优势，能够提升回顾评价过程计算机化的可行性。(4)在机理分析的研究成果的基础上，选择复杂系统动态行为建模方法，构建梯级水电开发生态环境影响预测评价模型。研究成果包括：将SD方法与基于Agent的建模(Agent-based Modelling，ABM)方法相结合构建预测评价的整体框架模型，以及基于SD方法的“水电工程移民相关影响”预测评价实例。结果表明：复杂系统的动态行为难以用数学模型或统计分析模型来表述，而系统仿真方法是复杂系统动态行为建模的有效方法；SD仿真方法与ABM仿真方法各有优势，都拥有较为成熟的仿真平台，适用于通用领域的复杂系统行为预测，因此应该利用它们的优势互补，将二者结合起来构建预测评价的整体模型；基于SD与ABM的预测评价整体框架模型不仅具有面向对象的灵活结构，而且保留了耦合其他专业领域模块、以及根据知识的增长不宽扩展和修正的空间；Vensim是较为成熟的SD仿真软件，能够处理一定范围的梯级水电开发生态环境影响预测评价问题。总之，基于(复杂)系统科学理论开展梯级水电工程的生态环境影响分析与评价研究，构建分析与评价的整体框架模型，并提出适于该模型的结构化分析与评价方法，有助于改善我国目前缺乏梯级水电开发生态环境影响整体评价模型的现状，兼具理论研究价值与实际应用价值。本文的研究结果表明，复杂系统分析与建模方法能够提升梯级水电开发生态环境影响分析与评价过程的全面性、普适性、整体性和结构化，促进各评价环节的集成和计算机化，从而为相关研究提供具有参照意义的过程与模式。更多还原

【Abstract】 Cascade Hydropower Development (CHD) in a river basin is an effective pattern to explore and make use of the water energy resources, and promoting CHD continuously has been an important energy strategy in China since 1981. However, over the past two decades, increasing concern have been attracted on the negative eco-environmental impacts (EIs) caused by hydropower projects. Compared with the EIs caused by a single station, the EIs caused by the cascade stations of a same river are more complex and broader. How to analyze and assess these EIs comprehensively, holistically and systematically, so that to strengthen the positive factors, control and even avoid the negative factors, will be the key to promote the sustainability of CHD in China.CHD in a river basin is a series of stations from the upstream to the downstream of the river, which could enjoy a joint operating regime as a group. Among these stations, one or more large-scale reservoirs will be settled down in order to deeply increase the regulating ability of the whole cascade. Hence, large reservoirs with high dams divide the continuous river into several discontinuous sections, so that compared with the single hydropower projects (SHP), the EIs caused by CHD are going to be more cumulative in temporal and spatial scale. In other words, the relevance among cascade stations of a same river may strengthen or weaken the EIs of a single cascade station, even a new "cumulative impact" would arise because of the interactions among these stations: cascade stations of the river, eco-environmental elements of the river basin, and the interactions among them have compounded into a complex system named "Cascade-compound Eco-environmental System".Compared with the respective researches with specific discipline, the EIs Analysis and assessment of CHD can be accomplished more effectively by establishing a more comprehensive and multi-disciplinary assessing theory and methodology framework on the basis of system science, especially the complex systems science. This theory and methodology framework can enhance the inner links among different research fields help to acquire the overall awareness on the EIs of CHD. However, the outcome of current study about the assessment of EIs for CHD in China mostly focused on specific disciplines, lacking of the overall analyzing and assessing model from system perspective. Therefore, there are both theoretical and practical values on the exploring of constructing overall analyzing and assessing model for EIs of CHD, it can not only reflect the trends of EIs assessment for the CHD, but also meet the need of sustainable development of CHD in China.This dissertation proposes an overall analysis and assessment model for EIs of CHD under the guidance of system science, and to explore the various types of system analysis and assessment methodologies suitable for this model. For one hand, more interdisciplinary, comprehensive and universal content could be included into the researching scope, so that the results of the dissertation can play a reference role in EIs evaluation of CHD. For the other hand, the EIs analyzing and assessing procedures of CHD are tend to be repeatable, easily maintained and easily computerized. The major jobs of this dissertation are as follows:(1) Large amount of national and abroad CHD cases have been summarized, the CHD has been compared with SHP in various types of eco-environmental elements to reveal the similarities and differences between them, hereafter, basic features of EIs of CHD have been extracted. Results show: 3 modes could be generalized to represent common strategy of CHD, they are mode of type I (leader - runoff compound), mode of type II (continuous reservoirs) and mode of type III (Low Dam-runoff). The negative EIs of CHD are more significant in several aspects, such as the fragmentation in the rivers, the river habitat alteration, the changes in self-purification capacity of water and the changes of watershed landscape pattern; but at the same time, the EIs of CHD can be more controllable. 3 basic features of the EIs of CHD are extracted, they are systematic, cumulative and controllability. A set of eco-environmental elements which could be influenced by CHD are presented, with the highlight of eco-environmental benefits and sensitive ecological problems.(2) A Cascade-compound Eco-environmental System (CCES) has been defined as the research object to implement the structured analysis and assessment of EIs of CHD, several typical scenarios which in accordance with the 3 basic modes of CHD have been designed, and for each scenario, a mechanism analysis has been accomplished to reveal the generating pathways of EIs of CHD. Results show: the CCES is a complex system; 5 typical scenarios are designed to represent the 3 basic modes of CHD, and for each scenario, its relevant mechanism analysis can depict the cumulative pathways of EIs of CHD; furthermore, it is necessary to promote the study of environment-restoration regime based on the controllability of EIs of CHD.(3) On the basis of mechanism analysis of typical scenarios, a structured retrospective assessment model for EIs of CHD has been established. First of all, a conceptual model and a evaluation process were proposed, secondly, an index system was developed, thirdly, a series of specific evaluation methods were proposed, and finally, "eutrophication" was chosen to be the evaluation target to construct the logical model of the structured retrospective assessment. Results show: choosing fuzzy mathematics to be the basic theory to decide the influence extent of the CCES is feasible and effective considering the current level of relevance data accumulation; the logic model based on UML is object-oriented, which having the obvious advantages of reusability and maintainability, so that it can be converted into a computer software conveniently.(4) On the basis of mechanism analysis of typical scenarios, a structured predicting assessment model for EIs of CHD has been established. An overall framework of this predicting assessment model was developed based on the combining of two system simulation methodologies, they are: the System Dynamics (SD) and the Agent-based Modeling (ABM). Finally, a predicting assessing case was displayed to solve the influence forecasting of immigration that originating from dams construction. Results show: mathematical models or statistical models can not describe the structure of a complex system effectively, whereas system simulation methodology can be an appropriate way to modeling the complex system; the structured predicting assessment model should be constructed through taking advantages of both SD and ABM, for both of them are general prediction and system simulation tools with mature simulation software; the overall framework model based on SD and ABM is an object-oriented model, which is flexible, and easy to be expanded and maintained; and finally, being a mature simulation software for SD, Vensim can solve a range of prediction requirements on EIs of CHD.In all, researching the analysis and assessment of EIs for CHD based on the (complex) system science, constructing the overall framework model for the EIs assessment and proposing the suitable and structured analyzing and assessing methodologies for this model, will help to improve the current lack of overall assessment model for CHD in China, and be valuable both in theoretical research and practical application. The results of this dissertation show: the theories and methodologies of complex system analysis and assessment can enhance the interdisciplinary, synthesis, comprehension and structuring of analysis and assessment of EIs for CHD, promote the integration and computerization of various aspects of the assessment, so as to provide reference for relevant research.更多还原