This paper describes three corrosion case histories caused by overconfidence and lack of thorough understanding of the corroding systems. The first case history is exchanger tube failure from excessive high temperature sulfidation caused by changes of operation condition, run length and cleaning interval. The second case is about presumably hydrogen induced cracking(HIC) and high temperature hydrogen attack(HTHA) during weld repair and heat treatment in wet sour service. Inappropriate mixing of process streams was the cause of failure discussed in the third case history.
Inspection history is very important in inspection and maintenance work. Prediction of future damages and consequently optimizing inspection and maintenance works are possible through careful review of inspection histories. Inspection personnel should well aware of the histories and should utilize them.
Additionally, knowing the equipment history should always be accompanied with thorough understanding of the corrosion/degradation mechanisms of the equipment in service environments. Without understanding of the mechanisms, subtle change of service conditions can be overlooked and results in unexpected problems.
Engineers should always keep not having overconfidence from tracking well of inspection histories. Overconfidence can sometimes make to underestimate small changes, while small changes can lead to large differences. Here are several expensive lessons learned from the results of overconfidence.
Two months before scheduled maintenance period, leakage was detected in the splitter reboiler of a crude distillation unit (CDU). Giving up maximum efficiency, the unit could continue operation until the planned turnaround. When maintenance period was reached, exchanger bundle was hydrotested and tube leak was confirmed.
Table 1 is general information of the reboiler. Heavy Gas Oil (HGO) is used as heating medium in the tube side. It was first installed more than 40 years ago and had voluminous inspection files. Brief summary of maintenance history is reported below. Before 2003, corrosion rate was moderate and average life of exchanger bundle was around eight years with considerable remaining thickness even when retired.
(Table in full paper)
The bundle was pulled out and was inspected. Figure 1 is a schematic drawing of exchanger with indication of flow direction and marking of severely corroded area. Corrosion occurred at the same area as reported in 2003 and 2006 inspection reports. Very severe fouling on the tube outside was also observed. Tubes of 1st pass were twisted by irregular thermal expansion. Several tubes were sampled and were cut open to examine.
Figure 2 is corrosion morphology of sampled tube. The tube was corroded away relatively uniformly on both the inside and outside surfaces. Undoubtedly, corrosion was caused by high temperature sulfidation. Outside surface inserted in the tube sheet did not contact corrosive process stream and was left intact. Thickness reduction was calculated compared to the intact surface, assuming original thickness was the same as nominal thickness, namely 2.77mm. Thickness reduction inside of the tube was 0.7~0.8mm, while that of the outside was 2.0~2.1mm, three times that of the inside value.