Using the Bayesian updating approach to develop time-dependent corrosion wastage model for deck panel of bulk carriers

Abstract

In this paper, a new method to model the time-dependent corrosion wastage for main deck panel in bulk carrier is presented based on empirical data. In this modeling, recorded data of thickness measurements performed on bulk carrier of the Islamic Republic of Iran Shipping Lines from 1983 to 2016, which had the age of 6–28 years at the time of measurement were used. After monitoring and deleting suspicious data, 3814 and 3828 measured points from the deck plate and deck longitudinals elements were used, respectively. Due to non-sufficient information for all the years of ships lifetime, one of the previous models was considered as basic model. The coefficients were updated in modeling using the BAYESIAN method and applying the Markov chain Monte Carlo simulation. The parameters in this method were statistically considered and this method is useful to predict the corrosion wastage thickness of all ship structure elements per year of the ships lifetimes. Even a new data can update the modeling coefficients and provide a better and more accurate model. The resulting models with updated coefficients were used to estimate the corrosion wastage thickness for consecutive years of ship lifetime and its achievements, and important results were summarized.