Outlet of PSV on shell-and-tube heat exchanger

[Orsiz]

Hi all,

I'm currently looking into the design of some relief valves in our plant. During a field visit, I noticed something I don't understand regarding the outlet of a PSV located on the tube side of a shell-and-tube heat exchanger. In the heat exchanger, water is heated to 55°C by 2 barg steam. The set-pressure of the PSV is 14 barg. In my opinion, the relief case is thermal expansion of the water in the fully filled tubes when the heat exchanger is blocked in.

The PSV has a 1" inlet and 2" outlet. According to API 521 STD, a 3/4" x 1" should already be large enough as only a very small amount of liquid has to be relieved in order to reduce the pressure back below the set pressure in case of thermal expansion. The strange thing is that the 2" outlet goes up vertically to leave the building. The piping is also insulated. This looks like the PSV is designed to relief steam.

I think only a very small amount of liquid water will be relieved upon thermal expansion and no steam because a small temperature increase will already increase the pressure to the set pressure. What is your opinion? The PSV is installed in the late 70's.

Orsiz

 

[don1980]

The API 521 reference you mention (defaulting to a 3/4x1 PSV for thermal expansion) doesn't apply here because it is specifically for thermal expansion due to atmospheric heating.

 

[georgeverghese]

Agree 2 barg steam with a sat temp of approx 275degF cannot boil water at 14barg, for which boiling temp is just over 380degF. So this would only be a liquid only thermal expansion case.

However, external firecase relief cannot be excluded even in the case of the tubeside in most plants, since the HX channel / bonnet can still be exposed to fire case heat flux. Note heat exchangers are classified as pressure vessels and not piping.

If you can confidently say that flammable liquids with a flash point temp exceeding minimum ambient cannot accumulate beneath this HX ( from within the common kerbed area), then you could justify asking for a dispensation from having to accomodate firecase at this HX - check your national codes also in case there are some rules re firecase relief on heat exchangers also.

 

[georgeverghese]

At the OpCo I used to work in, company standards do not require tubeside relief for firecase if the only reason the tubeside could get blocked in during a fire is if maintenance isolation valves (on the tubeside) at this HX could be closed. For this contingency, the Company requires that the tubeside be drained and depressurised before Operations hands over the HX to Maintenance.

Dont know if your Company has such rules imposed / established, else my previous comment would be still applicable.

 

[KoachCSR]

Can you provide a P&ID or sketch of the system?

I agree that 1" x 2" sounds too large for a thermal expansion scenario. I'm curious if it's actually intended to provide relief for a tube-rupture scenario.

 

[Orsiz]

Thank you for your insights.

No flammable liquids are present near/below the HX. The only surrounding fluids are hot air, low pressure steam and water. Is a tube rupture scenario possible hear as a scenario? The tube (water) side is the HP side, the shell (steam) side is the LP side.

 

[Latexman]

Orsiz,

You should be able to retrieve and study the sizing documentation that led to the selection of the installed PSV, and why it is the size it is. Pressure relief device documentation should be kept and maintained for the life of the facility. Auditors ask for ours from time to time.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Orsiz]

Latexman,

These documents should be indeed available but they aren't for our plant (I looked in the original design documents of the plant, but no success). I could only find a datasheet with specified set pressure, relief area and construction materials.

 

[Latexman]

I hate when records are not set-up or maintained properly. I'm redoing PRD records for a plant now that has done poorly on recent audits. It's a very common problem!

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[georgeverghese]

If you can confidently exclude even a remote risk of a flamamble pool beneath this HX (and this would include pressurised flammable liquids jetting into this kerbed area from adjacent areas), then you could exclude having to accomodate firecase at this HX, assuming there are no other rules that dictate otherwise in the pressure vessel codes applicable to your location.
From how you describe this HX, does look like tube rupture is not applicable to this tubeside PSV.

 

[mk3223]

Is it possible that the equipment installed in 70's was changed the operating condition but keeping the original design?
IMO, it seems that the PSV of 14barg set pressure could be used to protect the water tube from overpressure due to the discharge blockage.

 

[LittleInch]

What is the design pressure / operating pressure of the water system.

In your blocked in case surely you could have superheated water at atmospheric pressure flashing of to steam as it leaks out?

Could the HX be fed with water greater than 14 bar? Without a fuller description of e system is difficult to say much more.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Orsiz]

The water flow through the HX is provided by a centrifugal pump. The max discharge pressure at zero flow condition is 8,7 barg (system design pressure is 10 barg). The normal operating pressure is around 7,9 barg.

Regarding the flashing of to steam. How can you know that the water temperature will be above 100°C at the set pressure of 14 barg? Only then, you will have flashing I assume?

 

[LittleInch]

Initially it won't flash but if left lung enough it will as the steam is more than 100.

Even if it's just v ghot water you don't want it in the building.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Orsiz]

We indeed don't want any hot water in the building but the problem in my case is that the outlet of the PSV immediately goes upward vertically. If the PSV should open, the releasing hot water can't escape the outlet piping.

 

[LittleInch]

Agree it sounds like it was designed for steam not liquid.

Is there an alternative closed drain system?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Ehzin]

See if there's a bleed hole or whatever you'd like to call it in the bottom of the outlet piping. If not, then the liquid level built up could limit its ability to lift at set pressure. Unsure how many feet/meters up the outlet pipe goes.

Also, what's the gpm or m^3/s, whatever units work, of the pump normally? Is there a control valve directly downstream of the exchanger? Does the steam work off a simple steam trap or level control or does it have some other form of temperature control? If you have a thermal data sheet for the exchanger, what's the designed flow of both the steam/condensate and water along with temps in/out? And lastly, how much piping is there between the pump to the exchanger and then from the exchanger to the control valve (Assuming there is a valve)? Oh, and line size?

May be a bit of information to provide but I believe there exists a scenario where the exchanger could continuously exhaust out flash steam. Depends on the design of the exchanger. If that's the case then the pipe routing makes more sense and the only issue to correct would be possibly routing of a drain line.

Thanks,
Ehzin

 

[Orsiz]

Ehzin,

The normal water flow is 26 m3/h. There is a control valve upstream of the HX which regulates the steam flow depending on the water temperature. There is no control valve on the water side. The steam works off a simpele steam trap.

Steam (130°C) design flow: 1500 kg/h
Water (15°C -> 55°C, 4" piping, approximately 30 m from pump to HX inlet): 26 m3/h

 

[Ehzin]

So this is a bit late but I think there's two possible issues to consider. It seems it was originally thought that maybe continuous steam relief was possible but I only believe that's likely if you blocked in the discharge of the exchanger and keep it lined up to the pump. Instead of demo'ing pipe or redoing original design you could tap a low point into the RV discharge and route a drain line from there outside. Considering it'll be for liquid it shouldn't require a large size as compared to the main outlet. The point being you don't want to build a liquid column in your vent as it could lead to piping issues and also affect set pressure depending on the type of RV. That and you don't negatively effect the original design if there's some aspect that's been missed so you handle liquid and vapor, sort of "separately" I suppose.

If you're curious on what scenario could probably cause continuous flash steam then we could go in detail on that. Depends on if you're interested.

Thanks,
Ehzin

 

[Orsiz]

Ehzin,

I still have to check the presence of a bleed hole in the outlet piping of the PSV. If this would be the case, then I indeed believe the original designer assumed liquid relief in the beginning. After some time (HX discharge still blocked), steam would be relieved. Do you agree?

Orsiz