Protecting Shell & Tube Exchanger from a Ruptured Tube

[apatjr]

Need assistance in calculating how to size a relief valve on the water size of a sheel & tube heat exchanger for a sudden rupture of one of the pressurized tubes on a 3-stage compressor cooling system.

The internal pressure is 3000psig and the closed loop water pressure is 70 psig. A water relief valve is in the water line that opens at 150 psig. The compressor flow rate is 130 scfm so once the sudden surge is displaced there will no problem relieving the steady state flow. However, how can the sudden expansion rate be calculated. Which in turn can be used to calculate the rate at which the water will be pushed out of the way to see if the water relief valve can flow sufficiently to protect the shell.

Any references or ideas on how to attack this problem would be appreciated.

 

[metengr]

apatjr;
One option you can consider is to install a rupture disc on the shell in conjuntion with the current safety relief valve. This way, any sudden overpressure condition caused by one or more burst tubes could be dissipated rapidly.

 

[codeeng]

API 660 makes reference to 2 standards:

API RP 521 Guide for pressure-relieving and depressuring systems, 4th edition.

The Institute of Petroleum, Guidelines for the design and safe operation of shell-and-tube heat exchangers to withstand the impact of tube failure.

You should consider the 2/3 rule per API RP 521 (modified to 1/1.3 in keeping with the 1.3 Code test pressure).

 

[EGT01]

Apatjr,

You should review the discussions in API RP-521 on
2.3.13 Heat Exchanger Tube Failure
2.3.14 Transient Pressure Surges
3.18 Heat-Transfer Equipment Failure

Basically, you need to determine the volumetric rate of gas that results from the tube failure but based on your low pressure side relieving conditions. For example, you could determine the mass flow rate of gas and simply divide by the gas density at the low pressure side relieving conditions. The required liquid relief rate would be a volumetric rate of liquid equal to that volumetric rate of gas.

You should especially note the discussions in the above API references about transient pressure effects. Typically a relief valve is not considered as having a fast enough response time to react to a sudden tube failure with high pressure differentials involved and a rupture disc will be used such as Metengr suggests. API further cautions that transient effects could cause significant overpressure even when protected by a relief device.