PRV For Shell & Tube HX

[ekschwab]

I have an application for an ASME Section VIII shell and tube heat exchanger. Unfortunately I do not have the ASME Section VIII Code in my hands yet to review UG-125 etc. I’m working on getting it.

To me, it is a clear cut answer that we need a pressure relief valve to protect the shell. But I am getting some opposition for adding one to the system.

The tube side is process water with a 900 # design pressure and the shell side is cooling water at 150# design pressure. We currently have thermal reliefs on the shell side but no pressure relief. The heat exchanger will be U-stamped.

Until I get the Code in hand, is there a simple answer why we should not consider a PRV in the system on for the shell side?

 

[TBP]

If there are isolation valves on the water side then I'd install a relief valve. You can get the same effect with control valves and backflow preventors. Potentially, the hot side flow can keep coming, while the cold side is bottled-up. Probably, the worst that would happen is a leaking gasket or valve packing, but a relief valve is pretty cheap insurance.

Just an aside, in any plants I've worked in, "PRV" is "pressure reducing valve". Abreviations are not necessarily standardized across industries, or even different plants within the same industry.

 

[devaxrayz]

ekschwab,

I assume that your cooling water system is supplied by pumping operation.

If your pump is centrifugal, the maximum pressure in your cooling water system will be the pump shut-off pressure (+ suct pressure). GENERALLY, the discharge piping and equipment downstream of this pump is rating (design) at pressure higher than the shutoff pressure. If this what exist in your system than you need NO PRV at your HEX shell.

Even if your pump is PD, good practice is that you already have any PRV in your pump discharge so you dont need anymore at your HEX shell.

In other case different from i state above you may think about installing a PRV. What you should consider is where is the place?? at your HEX shell or in pump discharge piping. Better if you put it in your weakest point of the system.

You have to check your entire cooling water system before deciding of install a PRV. And maybe do some hydraulic analysis to justify where to put it.

TRV on the other side is required for in case that you have low flow of cooling water and boiled at your HEX. TRV fuction to relief this excessive pressure caused by vaporization i think.

regards,

-Rayz-

 

[StoneCold]

I think that his point is that the process side is liquid at 900# (design) and his shell side is water at 150# design.
He has violated the 2/3 rule and needs to consider a tube rupture into his relieving scenarios. I would say yes you need a relief valve. You need to look at the blocked in thermal expansion case and the tube rupture case, and possibly some others that I am not thinking of. How hot is the 900# water? Is it going to flash if the tube ruptures? If so then your relief valve will have to be designed for a two phase flow until the exchanger empties and then a steam flow rate after that.


Hope this helps some.
Regards

StoneCold

 

[ekschwab]

Thanks for the responses. Both fluids are pumped with different sets of centrifugal pumps.

There are isolation valves on the water side. But as I said we already have thermal reliefs in the design for the shell side (cooling water). Blocked-in cooling water will be relieved if it heats up and expands.

I was thinking more along the lines of a tube rupture on the 900# side (hot process tube side), potentially exposing the 150# shell (cooling water) and system with high pressure.

Also, thanks for the PRV comment - you are correct. I should be more careful using abbreviations.

 

[ekschwab]

Stonecold,

Yes, it helps. Thank you.

The process fluid side has a design temperature of 400F so you are correct... the relief valve would have to account for 2-phase flow. The normal operating T=<150F, but since we have the 400F design (I don't know the basis for that yet) I have to assume it would flash for now.

I just began looking into the 2/3 rule. I know there are some other posts on this site about that.

To clarify: Nothing is built yet. We are in the system design/equipment spec stage at the moment.

 

[djack77494]

ekschwab,
It sounds as if you'll need a pressure relief device and that it will be sized for the tube rupture case. Relief requirements sound pretty significant as you may have saturated water at 400F suddenly dropping in pressure to the cooling water system pressure. That could release much steam. For the quick response needed, I'd recommend a rupture disk or rupture pin valve. Locate it in the center of the shell if possible.

 

[SmartEngineer]

I think you can do the following for your case:
A) if there is isolation valve in the shell side stream outlet
1)If the HE is designed for 2/3 rule, then you just need a thermal relief on the shell side stream that is CW due to the reasons stated by Mr.StoneCold
2)If the HE is not designed for 2/3, then the PRV shall be designed for both thermal and tube rupture cases whichever is greater (usually the later is the governing design case)
B) If there is no isolation valve on the shell side stream then, no PRV is needed at all as far as open path is foreseen and protection from tube rupture is mitigated by other relief devices as PC or PRV or vents at downstream equipment of the HE in the question

In your case it seems that you should look for case A1 or A2

Cheers
SmartEngineer

 

[ekschwab]

Is the 2/3 rule ASME VIII or API? I don't see anything in ASME about 2/3, but I just started looking.

And yes, there are isolation valves on the shell side system.

Thanks.

 

[panduru]

The 2/3 rule figures in API RP 521 referenced in API 660

pl see also:

1. thread292-104457 ASME (mechanical) Code Issues Forum
"Protecting Shell & Tube Exchanger from a Ruptured Tube"

2.

http://www.cheresources.com/forums/forum5/messages/10001.shtml

"By the way, the 2/3 rule developed from the fact that pipes and equipment are usually hydrotested at 1.5 times their design pressures. However, ASME just recently published new guidelines that allows us to perform the hydraulic testing at (I believe) only 1.25. This means the 2/3 rule disappears and becomes tighter. I believe the math works out to something like 3/4 now. That is, if the high pressure side design pressure is 100 psig, the low pressure side design pressure cannot be less than 75 psig (not 66 psig) if you want to eliminate tube rupture as a scenario. This is not hard and fast. Any company can still stick with the 1.5 factor if they wish (it is more conservative)."