Body Bowl choking 1

[Justice100]

Hello all,

I am looking at some PSV discharge piping calculations using Hysys for choked vapour flow. One thing that I am finding is that it appears that there is choking in the outlet nozzle of the valve. Having read around this is I see it is called body bowl choking but is more of a concern for large relief valves with high set pressures. The relief valves I am looking at have set pressures of around 20 barg at F and G sizes.

For example I have a 1.5F2 conventional relief valve with a set pressure of 20 barg relieving steam. If I model it with a 2"x3" reducer located directly on the PSV outlet nozzle then all is well and velocities/backpressure requirements are easily met with a short discharge line to atmosphere. However, if I put in a very short length of 2" pipe to represent the outlet nozzle of the PSV then it shows that the flow chokes in this 2" section. The only way I can see to remove this choke is to increase the downstream pressure (by adding some sort of restriction downstream), which increases the pressure/density at the 2" section. I feel I am probably missing something here... is this a concern? Would it be reasonable to ignore the valve outlet flange? I suppose ulitimately if I check that there is no choke at a back pressure of 10% of set pressure then it should be ok. It seems to be a problem with lots of my relief valves in hydrocarbon service and they are only F and G sizes.

 

[Latexman]

faq1203-1960

Thanks to don1980!

Good Luck,
Latexman

 

[The Obturator]

You mention that you are using a program to determine discharge flows. I don't know how familiar you are with Pressure-relief Valves or whether the FAQ posted in the last comment is any help to you, but you need to understand the basic issues with flow through a PRV.

Taking the 1.1/2" x F x 2" size as an example.

Flow will enter the PRV through the 1.1/2" NB Inlet. This is usually a (full) nozzle threaded into the inlet of the valve body.

When the PRV is fully open at Set Pressure + Overpressure (normally 10%) the nozzle will be fully flowing the fluid. BUT It is flowing out of the nozzle area, which is considerably less than the PRV's outlet 2" size you mention. The chart in don1980's referenced FAQ provides this information, except that the nozzle ACTUAL area should be used not the 'Effective' area stated.

A typical actual area for a given PRV with F orifice is approx 10 % larger than effective (each manufacturer has differing actual areas).

The actual area of an F nozzle is roughly 0.34 in2 compared to the outlet 2" area of 3.14 in2 - so basically X 10 the nozzle area. That is what makes the pressure drop.

Of course any backpressure acting against the PRV will also influence the flow.

Have you considered the nozzle area in the calculation ?

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[Latexman]

Justice100,

On your 1.5F2 PSV, what actual nozzle flow area and flow coefficient is in the HYSYS calculation?

I ran my software with it's default values, which uses API values of an "F" nozzle of 0.307 in[sup]2[/sup] and a 0.95 flow coefficient. And, yes, outlet flow in a 2" Sch. 40 pipe is very close to Mach 1. I know some PSV vendors offer a 1.5F2.5 model. Is your PSV existing, or are you specifing a new PSV? Is a restricted lift, see FAQ in post above, an alternative? If the sizing flow is very close to the rated flow, how about upsizing to a 1.5G2.5 or 2G3 with a restricted lift?

Putting a restriction in a PSV tailpipe is NOT recommended.


Good Luck,
Latexman

 

[Justice100]

Thanks Latexman. I have got there now.

The obturator - Thanks for responding. I am calculating the DP based on the rated capacity of the PRV from the vendor. I am considering the 2" outlet flange as the start of the DP calculation (3.14in2 area).

If this is of any value to anyone I think I have solved it. I have separated the pipe segments so I have a very small 2" section of pipe followed by a dummy valve (to model the discontinuity) and then a second pipe segment that contains the rest of the downstream piping in this case 3". The mass flow is fixed (at the PRV rated capacity) and the simulation will calculate what the back pressure is backwards from the end of the tail pipe to downstream of the dummy valve. I can then increase the pressure upstream of the dummy valve until the 2" section is at sonic velocity to find the pressure at the discharge of the PRV and confirm if it is less than 10% or not.

 

[Latexman]

API 526 Part 1 says

3.3
backpressure
The pressure that exists at the outlet of a pressure-relief device as a result of the pressure in the discharge system.

As such, I've always considered the tailpipe flange that bolts to the PSV outlet flange to be the start* of the DP calculation. In fact, the tailpipe flange does not really enter into the DP calculation. Whatever is attached to the tailpipe flange is the starting point. In this case, it's 2" pipe. Since [≤] 10% of set pressure is desired (~ 2 barg), it seems to me that 2" Sch. 40 (3.355 in[sup]2[/sup]) is more than sufficient. The 3.14 in[sup]2[/sup] PSV outlet flange flow area does not even enter my calculations.

*The calculation starts at the exit and goes backwards to the PSV, so nitpickers may say it's the end of the DP calculation.

Good Luck,
Latexman

 

[Justice100]

Hi Latexman – not sure what has happened as my reply post was to your original post but for some reason comes up after your second post. Confusing. Anyway…

Nozzle orifice is 185mm2. Kd is 0.975.

My mistake I have taken 2” sch 40 pipe as the starting point (i.e. the pipe that bolts on to the PSV flange), which immediately swages up to 3”. Having spoken to a guru, he is not concerned with a choke on the tiny section of pipe/swage immediately downstream of the PSV as long as the back pressure is less than 10% and obvs no choking downstream where the lateral meets the header.

 

[The Obturator]

That nozzle area does not look right. Is that what the PRV vendor advised ? It's more like the minimum (calculated) area.

Actual area of an F nozzle is approx 0.34 in2 (219 mm2) as opposed to API effective area of 0.307 in2 (198 mm2).Then again you have used the vendor rated capacity which should be OK.

I'm not expert on Hysys, but shouldn't actual, rather than effective (API) nozzle sizes be used?

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[Justice100]

Sorry should be 254mm2 actual. (The rated capacity supplied by the vendor has a 0.9 factor applied.). Ive not been using nozzle sizes at all just vendor supplied rated capacity.

 

[Latexman]

Yes, weird, I was wondering why no reply to my later post.

So, what is the vendor supplied rated capacity? About 6596 lb/hr?

If you'd characterise the pipe and fitting take-off of the tailpipe to the header, I can check it out in just a few minutes. Something like, a inches 2" Sch. 40, 2x3 commercial expander, b feet 3" Sch. 40, c 90[sup]o[/sup] elbows, and a dx3xd branch run reducing tee. I assume DP in the header is very, very low, right? If not, characterise that too.

Good Luck,
Latexman

 

[apetri]

a chocked flow condition at PSV discharge should not be considered a problem except for additional vibrations and noise,
reading your posts it seems that you are more concerned about the warnings and results produced by your simulator,
I am not familiar with Aspen, with a different software (Prode Properties) for a compressible flow I can specify the outlet pressure and the software calculates the pressure profile (up to PSV discharge) keeping in account the speed of sound / mach number in the different sections.
You can simulate the system in different ways (see your second post) as far as you can verify that it satisfies all process / standards requirements.