RISK-BASED DESIGN OF LARGE-SCALE FLOOD DEFENCE SYSTEMS

Abstract

On the basis of quantitative reliability analysis of flood defence systems, two design methods are developed. The first is reliability-based design, where the optimal geometry of a flood defence system is obtained by minimising the cost of construction under a constraint on the probability of flooding of the protected area. Reliability-based design is an integral part of the second design method, risk-based design. In risk-based design, the appropriate flooding probability of a protected area is obtained by comparing the cost of protection with the risk reduction that is obtained. A case study illustrates the application of both methods. INTRODUCTION The Netherlands is a low-lying country that depends on an extensive system of flood defences for its existence. Clear defined design requirements for the flood defence systems are necessary. Since 1956, the design requirements are given as prescribed exceedance frequencies of the water level in front of the structure (Delta Committee, 1961). At the time, this approach revolutionised the design and safety evaluation of flood defences in the Netherlands. With the statistical analysis of observed water levels, elements of probabilistic methods entered hydraulic engineering in the Netherlands for the first time. Since the Delta Committee, probabilistic methods have been further developed. The fundamentals were developed by Turkstra (1962). Today, a number of tools for reliabilitybased design are now available to the hydraulic engineer (see Oumeraci et al, 2001 for an overview). As a result of the development of probabilistic methods, it is now recognised that: 1 Risk analysis specialist, ARCADIS, Postbus 220, 3800 AE, Amersfoort, The Netherlands, H.G.Voortman@arcadis.nl 2 Assistant professor of probabilistic design, Delft University of Technology, Faculty of Civil Engineering and Geosciences, Stevinweg 1, NL-2628 CN, Delft, The Netherlands, P.H.A.J.M.vanGelder@ct.tudelft.nl 3 Professor of hydraulic structures and probabilistic design, Delft University of Technology, Faculty of Civil Engineering and Geosciences, Stevinweg 1, NL-2628 CN, Delft, The Netherlands, J.K.Vrijling@ct.tudelft.nl