OPTIMAL ROUTING OF A WIND-POWERED CARGO VESSEL USING ENSEMBLE WEATHER FORECAST DATA

Abstract

This paper presents a novel optimal routing method using Numerical Weather Prediction wind-wave fields, created in conjunction with the Wind Challenger Project, a Joint Industry Project developing a partially wind powered cargo vessel. The devised algorithm optimizes way-points continuously to form a route that minimizes fuel consumption under a time constraint using the interior point method. The new Ensemble Averaging Method (EAM) takes forecast uncertainty into account using the ECMWF ENS-WAM ensemble data set and Monte-Carlo sampling. The EAM is evaluated against the optimal route found using the control forecast (CFR), considered the most accurate single forecast available. Simulations, run across 2016 on a trans-Pacific route, show that the EAM significantly outperforms the CFR in finding the analysis optimal route. In addition, when comparing using analysis data, both the mean engine power and mean error in arrival time were reduced by 16 BHP and 0.23 hours respectively.