Tube rupture case 2

[Pfluid]

Hi All experts
Need your help and advise, I'm designing a shell and tube heat exchanger. The shell side and tube side design pressure is 100 psig, there is hot oil on the tube side and cooling glycol on the shell side. The operating pressure on both sides are nearly the same so either fluid can leak into each other. I have installed a PSV on the shell side to protect against thermal expansion case during cooling glyol shut off. The fire is not a credible scenario. There is a hot oil pump upstream of the heat exchanger, the pump shut off head is less than 100 psig. So the questions are:
1) It is necessary to install a PSV on the tube side?
2) It is necessary to consider a tube rupture case? The shell side and tube side design pressures are equal
3) Is it necessary to consider a glycol leaking into hot oil side causing quick pressure build-up due to vaporization?
4) If there are no credible scenario then is it still required to install a PSV on the tube side?

Thanks in advance for your help
Pfluid .

 

[RVAmeche]

1) You have to look at the various scenarios for that process. If it's liquid filled, thermal expansion at a minimum will be credible. Process pumps, regulator failure, etc are also potentially credible cases.
2) See API 521. If the MAWP of the two sides are the same the risk of a tube rupture is decreased. Typically pin-hole leaks (~1/8" OD) are always considered.
3) Is that a consequence of a tube rupture/pinhole leak or some other sort of leak? If it's a credible case (ie, it could physically happen), then yes. API 521 also discusses this type of scenario.
4) ASME requires vessels to protected from overpressurization. While "safe by design", which says that there are no process conditions that could produce overpressure are possible, is allowed it's rare and you need to be very clear when trying to use that. In the past I've had inspectors say I don't care, put a relief vale on it anyway. As mentioned before, if you have liquid streams on both sides it's likely that thermal expansion is at least credible.

 

[Latexman]

4) If you go the "overpressure protection by system design" route, you need to follow UG 140 in ASME Sec VIII (Thanks, don1980!). It lists conditions where, if applicable, a pressure relief device would not be required. IMHO, following UG 140 to the letter is quite onerous. A small PSV may be more economical.

Good luck,
Latexman

Engineers helping Engineers

 

[Sawsan311]

Thermal Expansion PSV is recommended to be installed on low pressure side if a blocked in heating condition can result in vapor pressure exceeding the design pressure. Forced administrative controls can be put in place such as key mechanical interlock system which prevents closing the cold side low pressure shell side before closing the inlet of the hot side tube side.

 

[Ehzin]

So this is just for information as I don't see this applying but if the design temperature of the tube side is somehow lower than the shell, then during a tube rupture you could exceed the tube side stress limit before the relief device on the shell side lifts properly. I understand this very likely doesn't apply since the tube side is hot oil it's just information. It would more apply if this was because you re-purposed an existing exchanger.

Regarding item 3, that could be fairly complicated to calculate. If the tube side is operating at a lower pressure at the time of rupture then glycol could migrate to the hot oil side. Based on differential pressure say this is 5 drops a second or .02 in^3. When the glycol boils say it expands to 32 in^3 drastically building pressure. This volume of vapor that now tries to relieve back into the shell side can only make its way back through the same sized whole and can do it at a much slower lb/hr rate than the liquid that went came from the glyocl/shell side. This could lead to a spike in pressure. Trying to quantify this could be incredibly complicated and I'm not sure if glycol would even boil at the temperatures you're discussing. This was intended to be food for thought I suppose. If you'd like to brainstorm to quantify this let me know. The safe route is to size a relief device on the tube side for the vaporization rate of the shell side in the event of a tube rupture.

Thanks,
Ehzin