Calculating back pressure with a scrubber fan discharge?

[YungPlantEng]

Have a set of PRVs off a common header that discharge to a scrubber. The scrubber has a fan that blows some portion of fumes to atmosphere and recycles the remainder.

This is an old and poorly designed system with no relief documentation. I’m going to make some assumptions and say that only 3 of the relief valves are discharging… they were sized to 2000 scfm with no assumed back pressure because the max scfm of the scrubber (3000 scfm fan on it) is 6000…

This is a PSM system and a coded pressure vessel with an existing rupture disc. My basic questions are:

1. How can I estimate backpressure through the system when relieving? I can estimate pipe losses and the scrubber has a maximum pressure inlet of 10” W.C or so but from my understanding the constant backpressure is essentially nothing since the system is run as a “vacuum”. I’m used to systems that just outbreathe to atmosphere so I’m a bit out of my element.
2. How can I estimate the impact of piping additions preceding the PSV? This is a very small amount (maybe a spool piece?) but I’m not sure how exaggerated that impact would be on the system.

Sorry for the silly questions - feel free to just point out some texts if these questions aren’t useful and I need to do some reading up first.

 

[Latexman]

1. I start at the discharge to atmosphere and work backwards. You'll need the fan curve and pressure drops to create a pressure profile.

2. Its not perfectly clear to me what this means. Can you explain it with a before/after sketch please.

Good Luck,
Latexman

 

[YungPlantEng]

Essentially the length of piping from the tank nozzle to the prv would increase to remove condensation pool-up in the dead leg. Concern here is that pressure loss might be greater than 3% from my understanding…. Which if it’s 40 psi off a 4’ OD pipe that might very well be the case.

 

[Latexman]

2. Calculate the pressure drop of an entrance and 7’ of pipe at the rated capacity of your PSV at 10% overpressure.

Good Luck,
Latexman

 

[YungPlantEng]

Thanks Latexman.

Not sure how I’m struggling so much with this… is there some way to estimate pressure drop through a PRV? I can’t seem to easily estimate backpressure but I feel like I don’t understand it correctly. Is your backpressure just your PRV minus the pressure drop across it in a fully relieving case? In that case wouldn’t all PRVs in a common header need bellows?

 

[Latexman]

Have you looked at API-520 Parts 1 and 2 and API-521? Also, you can Google your vendor's or other PSV vendor's documentation on how to size reliefs. There's some good instructions out there, and software too.

So, step by step:
[ol 1]
[li]You size the PSV. Determine rated capacity (10% overpressure and ASME flow coefficient). This is all ignoring inlet and outlet.[/li]
[li]Check inlet dP at rated capacity. 3% limit. Change design to accomodate.[/li]
[li]Check outlet dP at rated capacity flow. Estimate backpressure at PSV outlet flange. 10% limit for conventional PSV, 30-50% limit for bellows PSV (check your PSV), up to 90% limit for remote Pilot operated PSVs. Change design to accomodate to the type of PSV selected.[/li]
[/ol]

You really don't need the dP of your PSV, but you can calculate it if you want.

Good Luck,
Latexman

 

[YungPlantEng]

Does inlet differential pressure just depend on pressure frictional losses? We have a tank nozzle that is piped out a few feet before reducing to a smaller diameter to fit the PSV. Presumably this would increase the pressure following Bernoullis law…. Maybe I need to brush up on my fluid flow…

 

[Latexman]

For inlet pipe and fittings [Δ]P for a PSV you want only frictional losses.

Good Luck,
Latexman