shell and tube relief sizing 1

[compo]

With regard to the sizing of bursting discs for a burst tube scenario within a shell and tube exchanger where the tubes have high pressure gas and the shell low pressure cooling water. Note in this case the shell is not designed for 77% of the tube side design pressure.

When sizing the relief device is it common to assume that the water will be displaced as relief via the relief device bursting disc in the first instance or should it be assumed that the exchanger water volume will be initially displaced via the cooler outlet water line hence only sizing for gas?

 

[chrisflower]

The conservative approach is to assume that there is a sudden tube rupture and in this case the relief device would have to pass the cooling water at a volume rate equal to the gas rate flowing from the ruptured tube.

However an alternative approach is to assume that the leak develops slowly which will push the liquid out of the exchanger and the relief device can be sized for the gas rate only. This relaxation has been borne out by experience but use it with discretion because it may be questionable for exchangers with large bore tubes. The down liquid lines may need to be designed on the assumption that liquid will need to be displaced through them during pressure relief.

In the limit, when the liquid lines and the inlet/exit nozzles on the exchanger shell are sufficiently larger than the exchanger tubes and there are no restrictions in the liquid lines, then it may be reasonable to assume that the gas flow resulting from a tube failure could always pass along the liquid lines so that no relief device need be provided.

As ever with these things there is no general answer and is specific to the application, but I hope this helps.

 

[EGT01]

Although you haven't mentioned your exchanger's pressure values, as an additional word of caution, API RP-521 also suggests considering a dynamic analysis of the relief event when there is a wide difference in pressures between the exchanger sides and especially when the low side is liquid full and the high side contains gas or liquid which could flash.

"Modeling has shown that under these circumstances transient conditions may produce significant overpressure, even when protected by a pressure relief device."