Relief load calculations for Thermal Expansion

[ASDFqwer4]

Hi everyone,

Recently I am working on thermal expansion case for relief valve design. System is HX with crude being heated by shell side other HC. Process fluid is crude (on tube side of HX) and PSV set value is 400 psig. I have considered tube rupture - not applicable, blocked out- not applicable, fire relief load-explanation mentioned below.

In fire relief load calculation, I found out that fluid vapor properties after 10% flashing are above critical point. So the liquid is very heavy and not easy to flash, and liquid may be thermally expanding. So I assumed HoV of 40 BTU/LB and completed fire relief load calculations.

That being said, I was also considering thermal expansion as my possible relief scenario. I never came across such situation. In this calculations assuming that tube side of HX is blocked due to human error, then I was trying to find out the relief pressure using heating curve of a PRO-II. In that heating curve, I am looking for calculated pressure (412 PSIA) for maximum possible temperature (design temperature of hot side = 650 F) in order to determine whether thermal expansion is credible scenario or not.

PSV set point 400 PSIG
Relieving pressure according to 10% rule = 440PSIG

Heating Curve results
Maximum possible temperature of cold stream = design temperature of hot side = 650 F.

Calculated possible pressure is 412 PSIA


My question is, should I take the heat exchanged value ( available in HX datasheet) from hot fluid (hot which is on shell) to crude (cold which is on tube)and find the thermal relief load?


What I think about this?
I think that if the cold side fluid is blocked by any means and even if you supply maximum possible heat from hot side, the cold fluid doesn't reach 10% set pressure value. So there will not be any thermal expansion case.

Kindly reply
Thanks

 

[chemebabak]

ASDFqwer4,

Do you mean that phase change occurs at 412 psia and 650°F? Or do you mean that is the pressure calculated by using the cubic expansion coefficient?

In either case, the PSV is required if the pressure reaches the vessel MAWP, not 110% of the set pressure. The the PSV set pressure should be set at the MAWP or less.

So, if you calculate 412 psia (397.3 psig) then you do not need a PSV at 400 psig. Nevertheless, if you wish to be conservative then you should consider a PSV. If it were me, I would definitely ask a highly qualified engineer verify the expansion calculations and sign that the pressure will not meet 400 psig. 397.3 psig is way to close to not consider a PSV.

 

[MortenA]

IMO, if your find that you cant get above design pressure+10% then you dont need to consider this as a relief case. Wrt flow rate: Since the pressure cant get above dP+10% then flow at dP+10% will be zero (by definition). The valve may relieve something though, and if your concerns are with regards to the relief header then there will be a flow. Did you arrive at the max pressure following a similar logic to this:

-Initially you have X kg in a volume V, that is the denisty is X/V, and this should match your calculated density (using e.g. PRO II).

-Without considering relief then the density will ramain the same since X and V remains constant, therefore you solve for density at the temperatre 650ºF?

If yes, then i think you did it correctly. The remaining question could be solved using a step wise approach e.g. by spread sheet considering heat transfer rate and then making an inventory ballance for each time step. I think you only need to do this once, since it will quickly drop of? All in all i think it will be overkill, but you need to convice somebody then i might be worth it.

Remember that the PSV has a re-set pressure below set pressure. Use the above mentioned methode to calculate the temperature where the P=set P then calculate the flow required for keep ing P stable when the T goes up and keep tract of changes to the mass in the HX. It should not be very difficult in a spreadsheet.

Best regards

Morten

Best regards

Morten