A predictive model to estimate ice accumulation on ship and offshore rig

Ice accumulation on ships and offshore rigs creates unsafe working conditions and may damage critical equipment. Several approaches have been developed in the past to predict ice accumulation; these include analytical models, experimental investigations, computational fluid dynamics simulations, empirical and statistical models. This work proposes a probabilistic causal relationship-based model to predict ice accumulation on ships or offshore rigs. The model uses a Bayesian probabilistic approach to establish the relationships among the factors affecting icing. The model is successfully tested on an experimental set-up designed to simulate the spray icing condition observed on a seagoing vessel in the subzero environment. The results of the experimental tests were compared with the outputs from the predictive model. It was observed that the predicted values gave a reasonably good match with the observed values. The proposed model considered a range of environmental and process parameters that affect ice accumulation. The model has the flexibility to include more parameters affecting icing, based on location and system. The model can be used for dynamically changing conditions with minimal computational load and time.