Estimation of uncertainty in ship performance predictions

Both for a shipyard, and its customers, it is important that a newly delivered ship meets the contracted performance criteria. To test whether the contractual requirements are met, the actual ship performance is measured during sea trials and compared against the contracted values. If the contracted performance is not met, this can lead to penalties for the shipbuilder and dissatisfaction for the customer In extreme cases this can even lead to non-acceptance by the customer On the other hand, an under predicted performance can lead to non-competitive design and a lower income for the shipbuilder Contracted performance values are based on predictions and experience. Unfortunately accurate predictions of performance are often difficult due to the uncertainties that are involved in the design and build of the ship and its (propulsion) installation. Furthermore, uncertainty is introduced by the full scale trials during which the various contracted performance aspects are to be demonstrated. This paper describes two methods to estimate the uncertainty in performance predictions. In particular a ship speed prediction and a bollard pull prediction are discussed. Both methods are explained, compared and strengths and weaknesses are discussed.