PSV outlet piping - am I understanding this correctly?

[cswosu]

After perusing this forum and debating with my coworkers, I think I understand the approach to discharge pipe sizing...maybe. Can you guys let me know if this is right?

For discussion's sake, let's assume:
1) PSV is 6M8
2) Flow medium is air
3) PSV rated flow is 264,000 lb/hr
4) Back pressure will not be high enough to hinder PSV from reaching rated flow
5) Air temp at valve outlet is 95 F

I'm using Hysys to obtain my results, FYI.

Let's say I use a 10' section of 8" Sch. 40 as my outlet piping. In order to achieve the rated flow and not exceed Mach 1 at the pipe outlet, I need 43 psia at the PSV outlet/pipe inlet. The pipe dP in this case is 5.2 psi, and Hysys gives my pipe outlet pressure as 37.8 psia. Therefore, I have a shock at the end of my pipe where the gas transitions to atmospheric pressure.

Now, if I care to avoid reaching Mach 1 anywhere in the outlet piping, I would have to go up to 14" pipe. I obtained this result by specifying 95 F and 14.7 psia at the end of the pipe and increasing the pipe size until the maximum velocity in the pipe stays under Mach 1.

Questions:
1) Are both of those scenarios correct?
2) 14" outlet piping seems awfully large to keep the flow from a 6x8 valve under Mach 1. I've typically never seen much more than an increase to the next size (10") in practice, but this won't keep me from reaching sonic velocity in this case (and maybe it's just a function of my assumptions). In any case, should I always strive to avoid reaching Mach 1 in the outlet piping even if a large increase in pipe size is required?

Thanks in advance for your help. This has been confusing me for a while.

 

[Latexman]

1) Both look about right to me. It would have helped if you specified some details like set pressure, % overpressure, flow coefficient, etc. Guessing 800 psi at 21% OP got me close to your numbers.

2) The reason you've never seen a 3 pipe size increase is no ones willing to pay that price just to avoid reaching Mach 1. 8" will work, so why pay for larger pipe?

Good luck,
Latexman

 

[JoeWong88]

cswosu,

Designing PSV outlet with Mach 1 is possible however there are associated conditions such as mechanical integrity of the tail pipe, structure integrity of pipe support, capacity bottleneck of the PSV due to mach 1 (secondary choke), noise problem, potential PSV chattering due to high backpressure, etc. You may read more in Is PSV tail pipe & lateral at CHOKED (Mach no = 1) Accpetable ? .

Many company taking a conservative approach e.g. SHELL, EXXON, etc as they can not afford to any failure and cost associated to failure...

 

[insult2injury]

For open vent stacks, the vent stack size is often significantly larger than the SRV discharge to avoid blowback conditions (especially at high elevations).

I2I

 

[cswosu]

Thanks everyone. I appreciate the sanity check.

Thanks for the link, Joe. Those guidelines are useful.

 

[Latexman]

2) Flow medium is air

This fact had the strongest influence on my prior response. Would I invest 1 cent more of my company's money and resources to avoid Mach 1 on a PSV that releases air to the air. No. Of course, all the proper analysis must be done to make it a safe operation if it ever activates.

Good luck,
Latexman

 

[JoeWong88]

Latexman,
Good eye...

1) PSV is 6M8
.
3) PSV rated flow is 264,000 lb/hr

The risk associate to large PSV and flow has lead me to be cautious about the design the downstream piping...

In addition, workmanship, welder qualification, construction OC procedures, plant location, environment i.e. air acidity which lead to corrosion issue, material being used for the piping, routing of discharge piping... all these factors increase the risk...

First few factors are hard to control and manage and associated risk is hard to quantify

If i have selected Mach 1 as my criteria, i would request Qualitative Risk Analysis to be conducted to qualify the approach.

 

[Latexman]

I agree that whatever analysis the Engineer deems necessary should be done. Where I have trouble on gas flow is, I don't see where there can be much reduction in analysis by decreasing gas flow from Mach 1 to Mach 0.8. Therefore, if about the same analysis has to be done anyway, just accept the Mach 1 flow. Air is one of the most studied compressible gases ever, so I don't understand the need for additional safety factor in this case. Now if we were discussing two phase flow or a gas that decomposes at the relief temperature, I would be much more conservative.

Good luck,
Latexman

 

[4Pipes]

Just to stir up up the process engineers.

If the layout is fairly simple, say a few dozen feet long and direct to atmosphere, go for the following.
a) Sling hisys in the bin.
b) Get a text book out and do a fanno line calc. M>0.8 only exists near the end of the pipe even if the pipe is miles long. Do a velocity profile along the pipe. Its quite surprising.
c) If the original set pressure pressure is reasonably high, your may find that reducers give you a higher average flow velocity in a longer section of the pipe than you are trying to avoid.
d) You could even find that your reducers are choking or near choked. In which case have a look at the acoustic energy (and associated vibration) they are creating. Again, you might find it higher than without. A bit like a old fashioned megaphone.

I always find it comical that reducers are added to avoid choked flow at the end of the pipe. If the flow will choke the end of the pipe, chances are that the reducer will be choked. Which is worse? Secondly, a reducer looks a bit like a the back end of a rocket motor. Depending on the geometry, you could actually be creating M>1 plus a whole mess of shock waves in the pipe. Again, which is worse.

I would only use reducers to control the back pressure. However, you need to be using the right design method. Systems designed with Hisys always appear to end up with reducers (expanders) even if the tail pipe is only a couple of foot long.

 

[cswosu]

This fact had the strongest influence on my prior response. Would I invest 1 cent more of my company's money and resources to avoid Mach 1 on a PSV that releases air to the air. No.

You're reading too much into my example. I just used air so it would be easy for anyone to reproduce my results. I typically deal with PSVs in gas processing plants. It was easier to use air than to provide a gas analysis.

The purpose of my post was to confirm how the flow develops in the outlet piping - the PSV flows at its rated capacity, and a shock may develop at the end of the pipe. I have a colleague who was claiming that as flow started to develop in the outlet piping, the flow would increase through the PSV until Mach 1 was reached at the pipe outlet, and then the pipe and PSV were both choked. Using same size piping on the outlet usually meant this flow was less than the rated capacity of the valve. This seemed odd to me as I would expect a shock at the PSV nozzle to set the flow at its rated capacity (assuming the back pressure isn't too high).

Sorry for the confusion. Thanks again to all who responded.

 

[Latexman]

Thanks for the feedback. It goes to show how critical background information is to a post. This reminds me of something my Dad said a lot, "if you want the right answer, you have to ask the right question".

Good luck,
Latexman

 

[Latexman]

Btw, your understanding, "I would expect a shock at the PSV nozzle to set the flow at its rated capacity (assuming the back pressure isn't too high), is correct for the case in your post. If you go to bigger and bigger PSVs, you can encounter calculations for a tail pipe of the same size as the outlet flange that say the built-up back pressure exceeds the allowable backpressure limit even if the tailpipe length is reduced to zero! This is commonly known as "body choke" or "body flow limitation". You can read more on this in "Guidelines For Pressure Relief and Effluent Handling Systems", Center For Chemical Process Safety of the American Institute Of Chemical Engineers, 1998, pages 38-39.


Good luck,
Latexman