Safety Relief Valves

[x926e]

I have a safety relief valve on a water pipe tested to 200 psi being heated by steam at 525 F, 300 psi, and 6000 pph. I would like to determine the capacity and the pressure setting required for the valve. The safety valve will protect against thermal expansion of the water if the flow in the water pipe is stopped but is still being heated by the steam. Does anyone have a suggestion on how best to approach this problem.

 

[TBP]

x926e - Is this a heat exchanger application where you need to size a liquid relief valve for the cold side? Do you really have superheated steam heating this water, or is it just a typo? 300 PSIG steam will be 422*F.

 

[kyong]

Depending on heat transfer efficiency and probable heat loss from the pipe to ambient or somthing else, maximum temperature reachable of the pipe can be determined. We are interested in the case where pipe flow is blocked and steam is still flowing, of cource. Maximum reachable temperature of pipe should be determined, but the calculation will not be simple or the result will not be very exact. If pipe is strong enough to withstand the same pressure and same temperature as the steam, you wouldn't need this calculation, you wouldn't need a safety valve.

 

[MortenA]

You should know the design duty of the heater (or use design max inlet outlet temperature and flow to calculate the duty).

Use the common formulas for thermal expansion of water to size valve.

Thermal expansion is one thing - but what happen when the water starts to boil? Find latent heat at the SP (remember that the code allows a 21% overpressure for a single valve installation) and calculate steam flow at this P by use of your calculated duty and the latent heat. Redo the calculation for water but for the steam volume (the steam might accumulate somewhere and "push" water through the PSV.

Then there is the question of two phase flow... Since its a valve on a pipe this cannot be ruled out but im not familiar with two phase steam sizing sorry.

I think my methodology is somewhat on the conservative side -The duty when the water has stopped flowing will be lower. Usually in a "steam-heats-water" exchanger the major part of the thermal resistance will be on the water side. BUT i suggest you do the calculations - and if it dosnt mean a new (larger) valve then its not a problem and you _are_ on the conservative side. If it does mean a new valve go into more details.


Best Regards

Morten

 

[Drazen]

I think contributions are very good, taking into account some very important points that are basic in engineer's heat transfer calculation; but when we take into account two-phase flow and heat transfer, things get very complicated. I'm not sure do we really need it; it can be justified only if accuracy gives you some tangible benefit.

I will try to approch from total practical side. This problem is safety problem. Safety is the most important, all other things come after that.
1. Not only thermal expansion of water, but water pressure. When water expand thermally in this case its pressure rises because it is confined by pipe. The water pressure is what we need to control!
2. You must have some data for working pressure of water pipe. In this case you simply set safety relief valve to 1,1-1,15 x working pressure and that's it! Pressure must not come beyond it anyhow. But this is SECOND safety measure!
3. FIRST safety measure is: you must block steam inflow when water side pressure reaches some point that is above working pressure but bellow relief valve set pressure (let us say 1, 05 x water working pressure). This way real purpose of safety relief valve is to protect water pipe only from accumulated heat. It means when FIRST safety measure block the source of trouble (steam), there will be further water heating and pressure rise for a while because of thermal accumulation. This can be one moment or "little more", depending on water and steam volumes and contact area, but this moment is critical for safety because pressure still rises although you switch off everything
4. Not earlier then now thermal expansion comes into play. You calculate critical thermal expansion to find minimal opening area for relief valve. This is where you must consult standards. In my little country there is mandatory standard for this (it is called "Safety Relief Valve Calculation"; there must be some equivalent ANSI stuff).
Regards

 

[MortenA]

Drazen: I dont think the question was with regards to SP but HOW to size the valve including how to determine the releif volume.

Usually you dont consider double jepordy - but i dont think that assuming that your safetysystem that should cut steam supply could fail when you have closed of the pipe with water is double jepordy. So you HAVE to consider the case where steam supply continue even though it SHOULD stop.

Best Regards

Morten

 

[butelja]

MortenA,

I believe the code only allows the 121% of MAWP during a FIRE event. This would not apply, as there is no fire. The correct flowing pressure during relieving would be no more than 110% of MAWP. Of course, the set pressure of the PSV must be at or below 100% of MAWP.

 

[krb]

The pressure setting needs to be set to protect the pipe or vessel from rupture. Determine the MAWP of the pipe and that should be your pressure setting. Be sure to include any safety factors and allow for higher temperatures when calculating MAWP. As Butelja said, the set pressure should be at or below the MAWP.

As for relieving capacity, I believe the ASME code has some information on how to perform these calculations, but I would contact the valve manufacturer and get their input. They designed the valves and given your parameters, they should be able to help you size you valve correctly.

 

[Drazen]

Morten,
You are absolutely right when you say that failure of steam blockade must be considered. But this is what critical expansion calculation must take into account. Water evaporization must not start before relief valve begin opening (if that happens, THAT is the failure). What you want to avoid is explosion! If steam countinue to flow all the time (the worst case), only thing what can happen in any case is pipe wall overburn (which still must hold against rupture for the certain amount of time) because of decreased amount of water. This can be mitigated with adequte sizing of water volume ( as large as possible in critical area of effect).
One more thing: when that happens (steam blocade failure, which is not really often if double control system is mandatory like in my area) there must be an alarm as well which will warn people to escape as soon as they can.

 

[MortenA]

Butelja

You are 100% correct 21% overpressure is ONLY for fire. Please note that there is an error in my first posting. its only 110%

Best Regards

Morten

 

[MortenA]

Drazen Im sorry i dont follow you: Why is it bad that steam starts to form and how will you prevent that when the SP of the valve is 200 spig and the steam on the hot side is 300 psig. Without looking in a steam table that tells me that at the same temperature but at 200 psig there will also be steam?

Im sorry i brough fire into the talk because its not about fire but about failure to close steam (on the hot side) when cold flow stops - a very likely situation.

I dont think steam formation in it self is a failure/catastrophe (at leat you got to elaborate on that) but the PSV must be sized to handle the three cases i described.

Best Regards

Morten

 

[Drazen]

You are right (I'm really not familiar with psi units)! If given values are corect (although steam temperature seems strange). That involves completely different calculation of minimal pipe volume. But than! In that case should we talk about thermal expansion? Pressure remains constant as long as we have two phases, isn't it??

 

[billbill20002000]

Hi everyone,

from Consolidated's catalog, section Thermal Expansion/API fire sizing.

GPM = BH / (500*GC)

where

GPM=Flow rate, in US gallons per minute at the flowing temperature

B=Cubical expansion coefficient per degree F for the liquid at the expected temperature differential.

water B=0.0001

H= Total heat transfer rate in BTU/hr.This should be taken as the maximum exchanger duty during operation.

G= Specific gravity referred to water=1.00 at 60F compressibility factor of liquid is usually ignored

C=Specific heat in BTU/lb?F of the trapped fluid.

Takeb from API RP 520 part 1

hope this will help

 

[MortenA]

Drazen

When you have a PSV its something like this: Temperature remains constant bescause pressure remains constant (because "boil-of" is vented). For a HC system in theory temperature will rise because the composition changes but i wouldnt consider this.

If you didnt vent and kept a constant volume and kept adding energy (from the steam on the hot side) then the pressure would increase with the temperature. If the cold side is only water i assume that in the end it would evaporate when T_cold=T_hot because on the hot side its steam?

Best Regards

Morten

 

Just would like to add one thing--make sure you use a steam trim (not liquid trim) on the PSV. EVEN if you don't think the line will boil, you determine if flashing on discharge occurs as a result of the escaping fluid from the psv dropping low enough in pressure (and is hot enough). The reason fo the concern is that if you use a liquid trim and flahing occurs, the psv will operate like a machine gun, opening and closing rapidly. THis leads to rapid failure and possibly seizing in open position.

 

[Fco]

Does anyone knows any criteria for calculating Thermal Relief Valve Set Pressure?

Thanks

 

[DeltaCascade]

Thermalbob hit the nail on the head? Sizing for worst case, fluid escape as vapour? If slugs of steam and water exit, or just liquid water, so be it?

FCO, you may want to start a new thread with your question, or better yet, search for many excellent previous threads on this topic in the forums on this site. Check API 520 and 521 recommended practices, although you may have to buy these.

Good luck.