Abstract
Due to the severe thermal gradients that develop on coke drums as they are filled with hot oil during coking and water during quenching, accurate modeling of the heat transfer from the fluid to the vessel is of critical importance in the analysis of coke drums. To this end, transient thermo-mechanical simulations typically utilize point measurements of thermal profiles extracted from coke drums in operation, but oftentimes use these point measurements as uniform thermal inputs across the entire inner surface of the drum. In reality, non-uniform thermal fields develop along the height of the vessel depending on fluid fill rates during operation, the effects of which have also been considered in a number of studies to date. The current study focuses on comparing these two temperature methodologies by estimating fatigue life at the skirt-to-vessel attachment weld (a prevalent location of fatigue damage) for two common skirt designs, to assess the feasibility of the former, simpler approach. Although the results indicate similarities in the overall trends, differences in fatigue-life estimates of up to 20% are calculated between the two approaches. Given these differences, the methodology presented in this study should be considered when a higher level of accuracy is required in estimating fatigue life. However, the alternative, simpler methodology provides conservative estimates of the fatigue life when directional insight or expediency is paramount.