Framework for a risk informed maintenance strategy for deteriorating hydraulic structures

Civil structures, especially hydraulic structures, are subject to influences from the environment around the structure. As a result of these influences the structure degrades, which lowers the strength of the structure. Several actions are available to lower the speed of degradation and/or restore the structure to a certain condition. Within the operations and maintenance of a structure, the question is whether regularly small reparations or once a large replacement needs to be performed. In order to make a founded decision a quantification is needed. One of these is using risks as part of the decision making process. Risk is defined as the probability of occurrence times the consequences. Since this can be expressed as costs, it can be compared with the costs of the reparations or replacement to find the value of this action. In this thesis, a framework is presented to aid in the development of a one replacement-multiple reparations-strategy. The method describes the steps needed to find all information to make a stochastic based deterioration model. Also the steps to calculate the costs for each strategy, where the moment of replacement and reparation interval are variable, are shown. The framework ends with an optimisation, resulting in the most optimal strategy within the set boundaries and assumptions. Testing of the framework has been performed with a case study of a fictional lock. After collecting all needed information about the degradation models, distribution and parameters, several calculations have been performed. With both a level II as a level III method these calculations have been made. The framework proved to be a good guideline for the determination of sought-after strategy. Since degradation models can have a more complex (than adding and multiplying parameters) shape, the level II method became significantly more complex than what it is used for (an indicative calculation). Even though level III method (the Monte Carlo method was used) requires more processing time (on a commercial laptop) than the level II method, the time it took to program was equal. A level II method is therefor only of use when a comparison with an analytical method is wanted.Hydraulic EngineeringStructural EngineeringCivil Engineering and Geoscience