Recently, we started up our Delayed Coking unit and were able to draw only less than 20% of our normal light gas oil volume from the fractionator. A scan of the tower showed that the trays from just above the LGO draw tray to the HGO cold reflux (Trays 6-15) are holding very little liquid. Upon further investigation we found a pressure spike from 25 psig to 58 psig during our heat-up of the system. We were following our normal procedure, but it appears that we had a large water accumulation that flashed at once in the tower. Our current practice is to initiate the boiler feed water to the coke heater and preheater during this warm-up period. The top of the tower was not hot enough for the water to make it up at this point during the heat-up. We also added steam to the valves on the coke drum decks which makes it way to the tower through the recirculation line around the same time as the pressure spike. We are assuming that we vaporized a slug of water which damaged the tower, but we can’t see why this time would have been any different than before. To add to the confusion, this is the first start-up that we’ve had DCS so we now have much more data than we’ve ever before. However, we don’t have the historical data to compare with past heat-up periods.
Has anyone had similar tray damage during the warm-up / start-up of their unit that might shed some light on the root cause of our problem?
