Improving Ship Operational Efficiency by Statistical Analysis of Voyage Data

Abstract

As a result of increased fuel oil prices the fuel cost accounts for an increasingly larger portion of the total operation costs for a vessel. Traditionally the fuel cost has been low compared to costs related to crewing and management, but today it accounts for over 50% of the total operating cost in some cases. This puts fuel efficiency on the agenda for many ship owners and operators. The attention on fuel efficiency is also driven by the environmental challenges the shipping industry is faced with. To achieve the goal set at the COP15 meeting about avoiding a 2° C increase of the world s average temperature, the emission of greenhouse gases has to be reduced by 50-85 % by the year 2050. Pollution from shipping is a significant contributor to global warming and the industry needs to become more environmentally friendly to take their share of the responsibility towards reducing emissions. The overall aim of this thesis has been to investigate if, and how, existing voyage data can be used to improve operational efficiency for a vessel. This has been done with the use of statistical analysis. Multiple regression analysis has been performed in the statistical software Minitab to evaluate which parameters listed in the post voyage data which is best for predicting fuel efficiency. When a good enough model is achieved this is used as a basis for suggesting fuel saving measures. Then the model is used to calculate the new improved fuel efficiency and the result is compared to the initial operational efficiency. This process is first done as an illustrative case with made up data and then performed on real data provided by Westfal-Larsen Management. With the real data it proved difficult to achieve good enough models for predicting fuel efficiency. There is some doubt related to the quality of the data as there was significant variation in the results. However for the group of voyages between Singapore and the Persian Gulf a model with high goodness-of-fit is achieved. On the basis of this model implementation of trim optimization is considered to be a feasible measure to assess further. All voyages used in this thesis has been done in ballast condition, this gives a better opportunity for trim adjustments. With the trim condition changed to the optimal observed value, the model can predict the new improved fuel efficiency for that particular route. The result is estimated to be around 4 % which does not seem unrealistic compared to other studies. It should however be noted that this estimate is the average value within a certain confidence interval, therefore the potential for fuel saving cannot be estimated with 100 % certainty. But the results still prove that by obtaining a model with high goodness-of-fit valuable information concerning operational efficiency can be extracted from the post voyage data.