Evaluating vessel technical performance index using physics-based and data-driven approach

Abstract

Ship's performance and its emissions are affected by several factors like operational state, weather, and maintenance condition. Rating ships with performance indices, such as AER, is not really helpful as these factors are affected by the ship's technical and operational condition, obscuring the benefits of energy-saving technologies and improved maintenance condition. To promote energy-efficient ships and reduce emissions, DNV proposed a simple and intuitive Vessel Technical Index (VTI) which is expected to represent the change in technical performance of ships over time. This paper presents two different schemes to estimate VTI, namely, physics-based and data-driven approaches, to validate their applicability as well as compare their performance. The results are based on the measured in-service data from a bulk carrier, equipped with advanced sensors for speed and wave measurements. Although the VTI values from the data-driven approach are slightly overestimated, the mean trend of VTI over time, representing the change in performance of the ship, is quite comparable. The uncertainty in the results is reduced almost by half using the measured wave spectrum for physics-based VTI, whereas the data-driven approach creates an opportunity for ship performance monitoring without the need for speed-through-water measurements, which are considered unreliable for most ships.