海塘,亦称海堤,是浙江省对修筑于沿海或用以抵御洪潮水的堤防的特定称谓。风险分析和安全评估是海塘工程中一个有难度而富有挑战性的重要研究领域。本文以钱塘江下沙海塘及其保护区为主要研究对象,在系统分析和总结国内外研究现状及发展趋势的基础上,对钱塘江海塘出现的历史险情、海塘破堤原因、破口宽度和溢流堰顶高程,钱塘江水位持续抬升的原因、变异点发生的时间、闸口、仓前站的设计水位,海塘洪（潮）水越顶安全性、海塘整体稳定安全性、海塘渗透稳定安全性、海塘安全评估,海塘破堤后洪泛区灾情、海塘失效经济风险,以及以洪前冲淤、抛石固脚为主要内容的海塘加固技术等方面进行了比较深入的研究和探讨,做到了理论研究与工程实践紧密结合,既评估了下沙海塘的安全性又分析了下沙海塘失效风险,为海塘风险分析和安全评估提供了方法体系。本文的主要研究成果有: （1）提出了钱塘江海塘的主要失效模式。钱塘江海塘以越项、整体失稳和渗透失稳三种破堤形式为主;标准海塘修筑前的1974年13号台风破堤调查资料分析表明,海塘破堤破口宽度服从对数正态分布,破口宽度超越200m的概率为1.22%。标准海塘的修筑较大程度上改善了海塘抵御洪（潮）水的能力,海塘一旦发生破堤,破口宽度可控制在200m以内。 （2）分析了钱塘江高水位持续抬高的原因、水位系列发生变异的年份,建立了海塘越顶安全性分析模型。自20世纪60年代以来,钱塘江河口开展的大规模治江围涂、上游建库是钱塘江河口高水位持续抬升的内因,水情条件是钱塘江发生高水位的外因;运用信息熵理论,通过构造水位变异的比较序列的差异信息相对测度诊断表明,1974年、1997年为水位序列的变异点;仓前站原1%设计水位9.81m,仅相当于现状的2%标准,客观上降低了下沙海塘越顶安全性,采用JC法分析下沙海塘发生越顶的概率为1.80×10-3,敏感性分析表明,钱塘江河口水情、工情条件改变而引起的外江水位特征改变,是钱塘江河口海塘发生越顶的最主要因素。 （3）结合钱塘江河床冲淤变幅大的特点,首次引入塘前滩地高程作为随机变量,建立了海塘整体稳定安全性分析模型。当塘前滩地高程均值为2.0m、标准差为4.38m时,下沙海塘整体稳定可靠指标β为4.91,失效概率为0.46×10-6;敏感性分析表明,对海塘整体稳定性影响最主要的是塘前滩地高程均值、方差和
逆变焊机
Seawall （also called dyke） is the specific name of the dyke built in shore or other place to resist flood and tide in Zhejiang Province. The more rapid development of coastal economy and worse environment has given an impetus to study efforts on engineering of seawall. The risk analysis and safety evaluation of seawall are the most important research areas on seawall engineering with difficulties and challenges. In this paper, some special studies are carried on the Xiasha seawall and its protected region along Qiantang River, which are introduced in the current trends of development and research. The research works include: statistic of historical situation of damaged seawall, the cause of seawall breakage, width of crevasse, the overfall level of weir peak after broken, the reasons of continuously increasing of high water level, occurrence years of its aberrance, and current design water levels of Zhakou and Cangqian in Qiantang River, the safeties on overtopping, overall and seepage instability, and the corresponding safety evaluation by the theory of fault tree, the inundation situation of flood area and correlative economic risk, and practical technique such as erosion before main peak flood and dumped riprap to reinforce the seawall, and the system method provided for risk analysis and safety evaluation of seawall. The major contents are summarized as follows:（1） On the base of historical data, the three major failure states of seawall are given, which are overtopping, overall and seepage instabilities respectively. The width of crevasse, which yields the logarithm normal probabilistic distribution and the probability of the width over 200 meters 1.22% is obtained from the statistic data during typhoon No.13 of the year of 1974. Thereafter, it can also be drawn that the broken width can be controlled within 200 meters after the construction of standard seawall.（2） After analyzing the reasons of continuously increasing of high water level in Qiantang River, and occurrence years of its aberrance in Zhakou and Cangqian, the mathematical model for calculation the probability of seawall overtopping is constructed. Large-scale reclamation in estuary and the carryover storage reservoir of Xinanjiang built in upstream since 1960s are the inherence causes, however the watersituation is the outer cause of continuously increasing high water level occurrence in estuary of Qiantang River. The analysis of differential information comparative measure, which based on the theory of information entropy and the conformation of differential comparative series of high water level, shows that the aberrance of high water level happened in the years of 1974 and 1997. The former design high water level 9.81 meters of 1% in Cangqian, equals to the current value of 2% approximately, that will decrease the safety of Xiasha seawall overtopping accordingly. The probability of the Xiasha seawall overtopping by JC method is 1.80xl03, and the further sensitive analysis demonstrates that the characteristics of water level is the major reason of the occurrence of overtopping.（3） The riverbed is changing greatly and rapidly by the action of flood and tide, so the flat level in front of seawall is firstly considered as a random variable in the model for calculating the safety of overall stability of seawall. It is gained that the reliability index p equals to 4.91, corresponding failure probability 0.46x10’6, when the mean and standard variance of the flat level equal to 2.0 meters and 4.38 meters respectively. Sensitive analyses show that the major factors affecting the overall stable reliability of the seawall are the mean and variance of the flat level in front of seawall and the mean soil friction angle, therefore the flat level in front of seawall should be considered as one of the random variables of great importance in reliability analysis of the overall stability of seawall in a macro-tidal estuary.（4） The model of seepage stability of homogeneous soil dam built on impervious foundation is constructed by Monte-Carlo method. The result shows that the seepage-resistance reliability of Xiasha seawall is a lower value of 0.985, and the sensitive analyses show that major factors affecting the reliability are the ratio of resistance seepage slope of soils and the mean high water level of the river.（5） The fault tree of seawall is the system in series. The conditional probability P（B/A）can be calculated by 1 - 3>（ . P y ） when the functional functions both eventsA and B are linearity functions. It can be found that the failure probability of Xiasha seawall varies within a small range because of the probability in charge of seepage instability, and the probability equals to 1.5% approximately.（6） Calculations of one dimension unsteady flow of river network with the soft package of MIKE series indicate that the areas of disaster, quantified by the water depth over 0.5 meter and the time over 24 hours inundation, can be limited within61.6km in the whole area of 108 km , and pertinent failure economic risk can be calculated as 5474,700 RMB based on the correlativity analysis among historical loss of economy, the area of disaster and GDP per unit area.Finally, the analyses on quantity ratio of water to silt are carried out for the river reach with different lengths by the models constructed of GM（1,1） and one dimension moveable bed. It is brought forward that the erosion chance should be selected during small tide by big discharges and short times before main peak flood, for getting more effect of erosion with the quantity ratio from 50 to 56 of the water to silt. The model for dumped riprap is formed by the theory of kitics of movement of particles in water and Monte-Carlo method. The model is a power utility to dumped riprap since the configuration of rocks under water can be simulated particularity.This research work is partly sponsored by Zhejiang Provincial Department of Science and Technology （ZPDST）. It is a part of the key project of institute of ZPDST, titled as "System study on risk analysis and safety evaluation of dyke" （Grant No. 2004712015）.