Location of PSV on Shell & Tube Heat Exchanger 3

[ghamsa]

We have a shell & tube heat exchanger which uses sea water for cooling the hot process stream.
we need to install a PSV (pressure relieve valve)on sea water cooling stream based on tube faluir case. The question is that where should we install the PSV (on inlet side or outlet side) of the sea water cooling stream and why? or it makes no difference either way.

Thanks for your usual support

 

[Latexman]

Code-wise, it doesn't matter. I'd select the side that had the lowest corrosion and fouling potential or, if there is no difference, the side that is more convenient to maintenance.

Good luck,
Latexman

 

[pleckner]

I agree with Latexman. However, you should seriously consider one of these two ideas:

1. Put a highly corrosive resistant rupture disk ahead of the PSV so you can purchase a less expensive PSV (lower grade of material of construction).

2. Use a rupture disk instead of the PSV altogether.

 

[CJKruger]

ghamsa,

Firtly, my preference would be to increase the design pressure so that you only have a thermal relief PSV.

Secondly, for cooling water exchangers the PSV is typically on the water outlet to ensure some flow/cooling continue even if the PSV lifts. As Latexman mention, other things may come into play for seawater.

 

[djack77494]

I'd put the PSV at the seawater outlet for two reasons. First, if it pops, you'll have some flow thru the exchanger which would probably be a good thing if the conditions were such that it would happen. And, secondly, the PSV would be up higher and therefore better positioned for piping.
Doug

 

[ghamsa]

I agree this PSV should be located on the cooling stream outlet side.

In this case, do I need to install a second or spare PSV whenever the main PSV is taken for maintanence or one PSV is enough?

thanks all for your support

 

[Latexman]

If you cannot take the time to shut down the heat exchanger and drain and vent both sides (remove the chance for all scenarios), you need two PSVs.

Here's another thought. Is it possible to have no valves in the sea water outlet line all the way to the end where it probably discharges to atmosphere, or "lock" ALL the valves open administratively? If so AND the outlet line will relieve your worst case scenario, then you can call your outlet line the relief.

Good luck,
Latexman

 

[ghamsa]

Latexman

you mentioned in your first replay that "Code-wise, it doesn't matter" which you a PSV to be installed on either side of the cooling suppy stream.

My question: why the code accepts locating the PSV on inlet side of the cooling stream? eventhough the outlet side is the logical location for the PSV per our previous discusions since the reason for having theis PSV is to prevent any chances for blocking the cooling water stream so that the heat exchanger will always receive cooling water through it.

Is there another reason for installing this PSV on the cooling supply stream other than blocked cooling stream scenario?

Thank you very much.

 

[Latexman]

What may be logical to you for your pressure vessel may not be logical to someone else and their pressure vessel. Notice I did not say heat exchanger. The Code applies to all applications of all pressure vessels. This is one area the Code lets us use our engineering judgement. Would you have them take that away?

Good luck,
Latexman

 

[ghamsa]

No, pls. keep it. Now, I agree we need to use our engineering judgment.

Many, many thanks "latexman" for your support.

 

[gr2vessels]

ghamsa,
It appears to me that the shell side is the hot process side and the operating pressure is much higher than the cooling water pressure. If the tube rupture is the cause of the overpressure and the process fluid is gas, the PSV should be installed on the pipe connected to the top channel nozzle, not on the exchanger, to allow venting of the excghanger after closing the isolation valves. If the process fluid is liquid, the PSV shoud be installed on the pipe connected to the bottom nozzle of the channel, to allow full drain of the mixed fluids. Conversely, if the shell side is the cooling water side, the location is determined by the process fluid (gas flows up and liquid flows down).
However, normally you would avoid the installation of the PSV nozzles on the exchanger shell or channel, unless the installation on the piping sections is not possible or not practical.
Cheers
gr2vessels

 

[ghamsa]

The shell side is liquid with higher pressure than the cooling water.

Thanks gr2vessels.

 

[Latexman]

Venting and draining are usually handled independently from the location of the PSV.

Put the PSV where it'll work and the long term cost of ownership is minimized.

Put vents and drains where you need them.

Good luck,
Latexman

 

[jcaiken]

Just a follow-up to gr2vessels response.

If the high pressure fluid (tube-side) is a gas or a cold, light liquid that would be vapour at the (shell-side) LP conditions, then a PSV may not react fast enough to give protection against the pressure transients created by tube failure. A bursting-disk or rupture-pin system may be needed to have a suitably rapid response for this case.

As the transient forces in such a case include the acceleration and flow of liquid to and through the relief system ahead of the gas/vapour, the location of the nozzles may become even more critical. To minimise this part of the transient I would want to:
i) use multiple nozzles located near likely failure points - tubesheet, bends, baffles
ii) locate nozzles on the top of the shell and orientate to give side-to-side flow paths
iii) minimise piping distance from the nozzle to the bursting disk.

Any comments?

 

[frogcurry]

What jcaiken says is correct as I understand it. the Institute of Petroleum has some guidance published on this. For the relief backpressure, the line/ disk sizing methodology is based on the backpressure produced by the slug of seawater pushed ahead of the expanding gas, as per API 520/521.