SIZING OF PSV FOR A TANK WHEN THE SCENARIO IS EXCHANGER SPLIT TUBE RUPTURE

[nancymarilyn1]

THE SCENARIO FOR SIZING A PSV TO PROTECT A TANK WAS A HEAT EXCHANGER SPLIT TUBE RUPTURE. THE TUBE SIDE HAS MEDIUM PRESSURE STEAM WHICH WILL BE RELEASED TO A TANK FULL OF LIQUID SOLVENT. THE PSV TO PROTECT THE TANK WAS SIZED USING API 520 GAS/VAPOR EQUATION (6) SECTION 5.6.3.1 FOR CRITICAL FLOW. THE PROPERTIES USED WERE THAT OF STEAM. MY QUESTION IS : THE PSV SHOULD HAVE BEEN SIZED USING THE API 520 EQUATION (25) SECTION FOR SIZING STEAM RELEIF AT CRITICAL FLOW CONDITIONS EQUATION (25)SECTION 5.7.1.

 

[Latexman]

Wow! First question in 17 years! (You don't have to yell. )

I believe you are correct. Use Equation 25 in Section 5.7.1. If the proper properties were used, I'd expect similar results from both equations, but Equation 25 is more specific to the sizing.

Good Luck,
Latexman

 

[georgeverghese]

The rupture flow from the HX tube may be critical if the steam pressure is high enough. But I dont see how the tank PSV could be operating at critical flow, since tanks have low design pressures, and flow through the PSV would most likely be subcritical. What is the set pressure at the PSV on this "tank" ? Or this a high pressure API620 tank with PSV set at the upper design pressure limit of 15psig?

 

[Snickster]

I believe that the equation for sizing for steam in critical flow could be just a simplified form of the equation for sizing gas or vapor, since with steam all of the parameters, such as "C" and "M" factors, in the gas/vapor equation can be combined into one factor because they are known because you know it is steam (water vapor). In other words API 520 provides a more simplified quicker equation to use. I would check both equations to see if they produce answers very different from each other. Saturated steam may act even less than an ideal gas due to the intermollecular forces between the water polar molecules when very close together. So with steam there may be more than a "Z" compressibility correction factor as in the gas/vapor sizing equation. I never worked much with steam systems and relief valves though.

I would check both equations out of curiosity and make sure that the relief valve is at least good for the expected relieving gas. It also could be that as the steam exits the tubes into the liquid solvent then the solvent will evaporate and tend to condense some or all of the steam. In this case the relieving vapor will be mostly solvent vapor with a relatively high molecular weight "M" and lower temperature.

So you may have all steam, all liquid/solvent vapor, or a mixture of both flowing through the relief valve. The worst case should be used to size the valve.

 

[Latexman]

nancymarilyn1,

george has a good point. What will happen in the tank? Will the steam condense and heat/boil the solvent, such that solvent properties become important to size the tank's PSV? You said "PSV" which makes me think the tank MAWP > 15 psig, but you said "tank" and I think the tank MAWP < 2.5 psig. If you supply more details, we can help you better. Tank MAWP? Tank PSV set point? What is the solvent? What temperature is the solvent normally in the tank? Steam supply pressure?

Good Luck,
Latexman

 

[LittleInch]

Also depends a lot on the specifics. I've often seen for a low pressure tank or PV that you can't get the speed of response you need from a valve for a tube rupture but might instead need a rupture disk for that eventuality. Of course if this is in fact a huge tank with a large vapour space and a 1" steam line running through it then maybe not.

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

 

[nancymarilyn1]

Hi everyone,

Thanks very much for all of your responses.

Latexman here are the answers to your questions plus additional information:
PSV set point: 17 psig to avoid decomposition of solvent
Tank MAWP: 32 psig
Solvent tank design pressure: 25 psig
Solvent tank design temperature: 232 C
Solvent tank operating temperature: 130 C
Steam supply pressure: 151.939 psig
Steam supply temperature : 366.8 F
SG of solvent: 1.2

Regards,

Nancy

 

[Latexman]

I get the feeling the solvent boils at a higher temperature than water at the same pressure. Is that correct?

With a PSV set point of 17 psig and a tank MAWP of 32 psig and assuming an atmospheric discharge, it is likely the tank PSV will be choked at the flow nozzle exit. So, if H2O boiling point < solvent boiling point, IMO, you are likely looking at the right equation (25). If there is appreciable solvent in the relief stream, you might use mixture properties with equation (6) and compare to results from equation (25).

To dig into LittleInch's concern over the speed at which this relief device has to act, how long would it take for the tank (normal level) to pressure up from normal pressure to 32 psig (assuming no relief) after the tube rupture occurred? Is it milliseconds, seconds, minutes, hours? I'm looking for ball park numbers for now.

Good Luck,
Latexman

 

[TiCl4]

The other main question is more along the lines of mixing. It looks like your solvent is a higher boiler than water. However, since you have a tube rupture scenario, I assume you will have intimate mixing before the mixture dumps back into the tank.

In that case your release fluid will be a mixture of water and solvent. You will need to do the VLE or VLLE calcs to determine composition.

 

[Latexman]

And you should screen for two phase flow.


Good Luck,
Latexman

 

[nancymarilyn1]

the solvent in the tank does not boil because boil point is higher than the steam. Steam vapor will be released. Thanks for all you inputs.

 

[nancymarilyn1]

I have another question which is bothering me. I am planning to size my valve for the MAWP ( 32 psig) of the vessel which will give me a "L" orifice 3 X 4. But I want to set this valve at 17 psig to avoid decomposition of one of my solvent. Can I use the L valve ? or do I have to purchase a PSV for the set point of 17 psig?

 

[Latexman]

MAWP + 10% ~ 35 psig sizing pressure (actually 35.2).

For the 17 psig set pressure, are you going to use the same 35 psig sizing pressure, or are you going to use 20 psig sizing pressure (10% or 3 psi whichever is greater)?

Good Luck,
Latexman

 

[georgeverghese]

If you have a penultimate high pressure instrumented trip (tank vapor space high pressure trip at about 15psig or less) to cut out steam supply to the HX, then likelihood of this PSV activation will be very low (low risk of solvent decomposition for small leaks which may raise tank pressure >17psig). Would then suggest setting the PSV at 17psig, with a full relieving pressure of 35psig (overpressure = 107% of set press). But inlet line to the PSV would have to sized for a pressure drop of no more than 3% of MAWP of 32psig as required by API.

What would be max normal tank vapor space pressure when you are using this solvent that corresponds to this decomposition pressure of 17psig ? Is it less than 15psig ? Can we see a P&I D of tank that also shows vapor space pressure controls ?

 

[nancymarilyn1]

Hi Latexman,
I am planning to use the 35 psig sizing pressure because I will get a smaller orifice L 3 x 4 than if I size using the 17 psig (17*1.1=18.7 psig) where I will get an orifice of N 4 X 6. The tank has a nitrogen blanket.

Thanks,

(PS: Did you keep a tale of engineers who ask questions , lol)

Nancy

 

[Latexman]

Then, you can use the L PSV with a set pressure of 17 psig and sizing pressure of 35 psig.

Good Luck,
Latexman

 

[nancymarilyn1]

Thanks very much y'all responses.

Nancy