Differences between a 3 dimensional probabilistic method of berthing structure design and the traditional method of a berthing structure design

Abstract

Jetties are designed by the guidance of the deterministic standards. These standards are based on the standardized values of safety factors (semi-probabilistic) and use the load and the capacity of every element as a standard value. The interaction of every separate element as a part of a whole is not taken in to account. The use of deterministic standards can result in a design that is not economically optimal. A probabilistic approach gives a better insight in the occurrence of an unwanted events and leads to more insight of the optimizing of the structure. The “Lyondell jetty” in the Europort in Rotterdam is used as reference structure. This jetty is constructed in 1997 and is in use by a manufacturer of chemical derivatives for all kinds of synthetic products. It concerns a continues berthing structure founded on piles and is covered with fender wood, with a double deck jetty, that offers space to berth two large sea vessel and two smaller barges. The failure of the berthing structure is a conditional chance for functioning of the jetty. The berthing structure is submitted to further examination by means of a Monte Carlo simulation. Using the Monte Carlo simulation a measure of failure of an event is expressed in strength and resistance. In this study this event consists of reaching the yielding stress in the structure or the exceeding of a determined boundary of deformation during the berthing of a vessel. The energy loading is introduced with the equation of Saurin. The soil is modelled as a spring and meets the requirements of the Winkler model. In analysing the deformations of the structure the soil needs to have elasto-plastic characteristics (p-y curves that were developed by Reese). The actual structure that was designed with the Blum method, the design was recalculated, using help of p-y curves. This becomes the Kool model that is used for comparison with the results of the Monte Carlo simulation. In this thesis only the berthing of the mooring point 2 (large sea vessel) is considered. First the question if the developed 3 dimensional simulation technique is applicable answered and second whether the application of the model leads to differences in capacity. The soil model developed in Scia-engineer has the same behavior as in M-pile and the structure shows the expected behavior according the rules of mechanics. In the model relating to the energy load the variables of length, width, depth, mass, angle of berthing and the coordinates of impact are described as a stochastic. In the resistance of the structure the volume weight, internal angle of friction, cohesion, yielding stress and wall thickness of the tubular segments are described as a stochastic. From the results it appears that the loads occurs only very locally. Consequently there appears an over capacity in the reference design relating to the length of all the piles and the diameter of the piles. The five most loaded piles in the structure are checked due to the energy loading of the bow at Mooring Point 2. It appeared that the diameter of these piles can have a smaller fitted diameter. In length a reduction of 9% is found. A trend can be distinguished that indicates a large reduction of capacity in length over the whole structure. The critical variable in the design is the velocity of the vessel. The critical element in the structure is the berthing beam self. The sensitivity analysis of strength indicates that the chance of failure in strength can be diminished by the strengthening of the berthing beam and reducing the thickest wall thickness’s of the piles.