PSV Discharge Header Sizing

[loilfan]

Hello,

What is the correct way to determine if a discharge header has sufficient area for multiple relieving PSVs? This assumes negligible back pressure (i.e. short piping length venting to atmosphere).

Example, two PSVs with NPT 2.5 outlets combine into a single discharge header.

Option 1: Use the ODs
Ar=Required Area = 2*PI/4*2.875^2 = 12.98 in^2
Da=Minimum Diameter Pipe = SQRT(12.98*4/PI) = 4.07 in

Therefore, a NPT 4 pipe is required.

Option 2: Use the IDs (assuming Sch 40)
Ar = 2*PI/4*2.469^2 = 9.58 in^2
Da = SQRT(9.58*4/PI)= 3.49 in

Therefore, a NPT 3.5 pipe is required

Option 3: Use the nominal size
Ar = 2*PI/4*2.5^2 = 9.82 in^2
Da = SQRT(9.82*4/PI)= 3.54 in

Therefore, a NPT 4 pipe is required

I could not provide specific guidance on what option to use in API 520 or 521. This NB paper says "Multiple devices discharging into a discharge manifold or header is a common practice. The discharge manifold or header must be sized so the cross-sectional area is equal to or greater than the sum of the discharge cross-sectional areas of all the devices connected to the discharge manifold or header." which leads me to believe that Option 2 is the correct method.

https://www.nationalboard.org/SiteD...re Relief Device Inspection Guide 1-19-10.pdf

 

[mk3223]

IMO, the pipe sizing or the hydraulic calculation are based on the actual cross section area of the pipe, not by the pipe OD or NPS. According to your example, the option 2 is correct, and options 1 & 3 are acceptable too.

 

[Latexman]

Option 2 is the only choice that has any technical merit, IMO.

The NB document is an inspection guide; not a design guide. It's guidance on discharge manifold is very basic and somewhat dated. API 520 and 521 are the contemporary design guides in use today. As such, they are more rigorous. You should use API 520 and 521 to ensure backpressure requirements are met for the type of relief installed.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[loilfan]

Latexman,

I agree that API 520 and 521 are the best documents to use here, which is why I searched there first. However I couldn't find any guidance on this subject there. It gives guidance on the minimum pipe size for inlet piping but for discharge piping it mainly discusses built up backpressure.

Option 2 makes the most sense to me too, but how is that related to bench testing the PSV? Option 2 considered SCH STD for my example but what if I had installed SCH 160? The discharge area is potentially smaller than what it was tested at.

 

[don1980]

Design guidance for sizing the discharge pipe is indeed found in API 520 Pt II. See paragraph 6.3.1 (6th ed.). As mentioned by latexman, backpressure is the variable used for sizing outlet piping. Whether or not the backpressure exceeds the allowable limits is what counts. The x-sec area is simply a product of this calculation.

 

[Latexman]

Can you send a sketch/drawing of what you are doing? It would help getting on the same page.

Yes, API 520 and 521 speaks to the required pipe diameters by ensuring inlet pressure drop and outlet pressure drop (or backpressure) requirements of the selected relief valves at their capacity meet the requirements for satisfactory operation. Are you using conventional, bellows, or pilot operated reliefs? How long is the manifold? Just wondering if it's long enough to worry about combining the exhausts, or you want only one wall/roof penetration or something?

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[mk3223]

Option 2 considered SCH STD for my example but what if I had installed SCH 160?

The line sizing is based on the pipe cross section area. As the pipe section area reduced due to the wall thickness increased from sch.40 to sch.160, the next larger pipe size may be required to accommodate the PSV discharge flow.

 

[loilfan]

Latexman, I don't have a specific layout to show you. A coworker asked a similar question, so I am asking this more generally.

Envision two conventional PSVs right next to each other. Their discharge pipes have 90s that point towards each other and the two streams combine at a Tee and vent to atmosphere outside a building.

Thanks for your responses mk3223 and don1980, I am going to mull over this some more and re-read some of API 520/521.

 

[Latexman]

Would you expect both PSVs to activate at the same time for some scenario? Like fire in the area?

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[loilfan]

Latexman, in the example my coworker introduced, they were both set at the same set pressure, so I would assume they lift at (for all intents and purposes) the same time.

How would them having different set pressures change things?

If I assume they were set for 100 psig and 150 psig, the 100 psig PSV would only be able to relieve a portion of the required capacity. The pressure would increase until 150 psig where it can relieve the full required capacity. There may be a larger time delay between both PSVs being fully open with different set pressures, but eventually they will both relieve their required flow rates.

 

[Latexman]

Conventional PSVs can tolerate built-up backpressure up to 10% of set pressure. Obviously, having different set pressures may affect the resulting manifold pipe size. It may not. It depends. In your example, the 100 psig PSV sets the pressure drop limit on the outlet manifold to 10 psig. If both PSVs had 150 psig set pressures, the pressure drop limit on the outlet manifold would be 15 psig. This may change the manifold one pipe size.

Also, the built-up backpressure from the 100 psig PSV is now acting as superimposed backpressure on the 150 psig PSV. This may cause the selection of the set pressure for the higher set pressure PSV to change so it actually opens at 150 psig. It won't change by much, but when you get into the detailed design it may change.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[don1980]

There's an important qualification (widely overlooked) to the 10% built-up backpressure limit for conventional valves. The 10% value is actually superseded when the permissible overpressure is more than 10% [ref: API 520 Pt II, 9th ed., 5.3.3.1.3).

The built-up backpressure rule is this: Conventional PRVs are limited to 10% of set pressure, or the amount of overpressure, whichever is greater. Engineers commonly apply just the 10% part of this rule, which is correct for PRVs set at MAWP, without realizing the limit is higher for PRVs set below the MAWP. Note that the latter part of that rule is universally applicable, while the former part (10% limit) is only sometimes applicable. Therefore it's better to avoid referring to this rule as being a 10% limit. Instead, just say that the backpressure is limited to the amount of overpressure - that works in every case, regardless of whether the PRV is set at the MAWP or below the MAWP.

Example: Say you have a vessel (10 barg MAWP) with a conventional PRV set at 5 barg. Let's say this is a non-fire case, which means the allowable accumulation is 10%. So the peak pressure (relieving pressure) is 11 barg for this case. What is the allowable built-up backpressure on this PRV? That answer is 6 barg (120% of set pressure).

 

[Latexman]

My mistake. I was thinking two independent systems, like a 100 psig MAWP and a 150 psig MAWP, with a PSV each and a common discharge header. After re-reading, it does sound more like two PSVs on one system.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.