Discharge Piping Diameter Reduction

[maddocks]

I have always believed that we shouldn't reduce the diameter of the PSV discharge line. We now have a client questioning my belief (can you imagine that!) - now I can't seem to find any reference that specifically forbids me from reducing my 2.5" line down to a 2" line. Sure, it's not good practice and you're all going to say that - so "Show me the paper". Can anyone find this reference? B31.3? API ?

 

[LittleInch]

Standards are rarely that prescriptive and concentrate on performance requirements compared to strict rules as every system is different. What is required for a simple short atmospheric vent is very different ot a refinery vent system.

What you need to do is demonstrate / calcualte the effect of the reduced diameter on the pressure at the relieving point.

In B 31.3 section K322.6.3 ( a simple search found this) the relief system which includes all the downstream piping needs to be sized such that at your reliving flowrate, the presusre in the thing you're protecting doesnt exceed 110% of the deisgn presusre (116% for multiple reliefs). for a long high flow rate system, 2 1/2" might not be big enough (you need to check) or a short low flow rate, 2" might be good enough. This is called design, not rule following and is why the codes normally don't get too prescriptive, but provide boundaries and limits on things that actually matter like pressure, temperature and flow, not line size.

As an aside, whilst I'm sure 2 1/2" piping exists, I've never come accross it and would prefer 3" if you're going to go for a bigger size.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[maddocks]

I totally agree with you and we're doing the pressure drop calc's right now but didn't want to go down the path if there was some obscure code reference that said "don't ever do that". The selection of 2.5" piping is not our choice, but the PSV is 1.5" x 2.5" so we're forced to either go up (not practical) or go down (preferred) if we can prove that capacity and pressure drop restrictions are within the plan. Thanks for your help.

 

[LittleInch]

Not as far as I know and given the sort of code 31.3 is I doubt if it would say that sort of thing. Apologies if I insulted your intelligence, but there are a lot of less experienced people out there...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[maddocks]

You know, there's some huge gaps in the industry on things that I take for granted. It seems bizarre, because I've been doing this for so long, that (like you) I've lost touch with what I consider to be a learned skill and what I consider to be inherently engineering/common sense. It's never the huge things, but it seems like hundreds of small things - everyday issues like level switches, how long should my LG be, why does my PSV pop before set pressure, how does my control valve fail 100% open, etc. We've been discussing this in house here and we're not sure how to manage the gaps in "basic process". I'd like to send my junior engineers to the field for a few months..

I think we need some sort of a basics prep course on process equipment.

 

[Bill3752]

See

http://www.eng-tips.com/viewthread.cfm?qid=195989.

My opinion is that good engineering practice (RAGAGEP) means use the 2.5" piping.

 

[TD2K]

API 520 explicitly states that the inlet line must not be smaller than the opening of the PSV or words to that effect so even if your hydraulics are okay, you can't run a 4" line up to a 6" PSV inlet flange. However, in previous versions it's silent on the outlet other than to say that the backpressure must be within the type of PSV you have selected. I ran it by one of our other engineers a few years ago who concurred that API 520 didn't explicitly forbid downsizing the outlet line but pointed out for most of my clients, I'd spend more time fighting that battle that I'd save in piping.

I'd have to check the latest version of API 520 to see if that's been changed, I thought I had seen something that mirrored the inlet piping, the outlet piping shall not be less than the outlet of the PSV but I'm not certain.

I'd echo the comment about everything not being in the codes. If that was the case, lawyers would do the engineering.

 

[TD2K]

API 520, Pt 1 discusses inlet and outlet line sizing. The only issued raised for outlet line size is that the backpressure for all relief valves must be within their allowable limits.

 

[DSB123]

maddocks,
I agree it seems logical and a requirement to always increase the discharge pipework since as the gas expands it would need a larger pipe to accommodate/maintain the mass flowrate unless you allow the velocity to increase. However I remember years ago working on a project where the owner insisted on a decrease in pipe size local to the end of the discharge pipe. (i.e. the tailpipe was 3in and a 3in x 2in concenric reducer was welded on the end) The reason being to ensure high velocity at the tailpipe end to disperse the contents. Obviously the backpressure on the RV was checked for compliance.

 

[jistre]

API 521, 7.2.2

...

"Each pressure-relief device that vents directly to the atmosphere should normally have an individual vent pipe sized for a relatively high exit velocity; however, the outlet piping should not be smaller than the pressure-relief-device outlet."

...

From the language, it appears that this applies only to atmospheric relief, not to relief going to a flare, treatment, or collection system.

 

[jistre]

Also, that section applies to single devices.

 

[Latexman]

The above confirms my understanding.

Discharge piping from a single safety valve must not be smaller than the safety valve nominal outlet flange size and must meet pressure drop constraints.

For valves discharging into a header system, the size of discharge piping must ensure that piping discharge pressure drop under any expected condition/scenario will not adversely affect set pressure or capacity of any safety valve discharging into the header system.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[SNORGY]

For what it is worth...

My experience has been (in Western Canada, specifically Alberta and Saskatchewan) that the Regulators indeed have not accepted installations in which the outlet piping from the valve to the header is a smaller size than the outlet connection on the valve, and that further, for several PSV outlet pipes tying into one header, they have enforced the interpretation that the summation of cross-sectional areas for the individual pipes shall be less than or equal to the cross sectional area of the header.

I think I came across a written Information Letter or interpretation about this a couple of years ago. If I can find it, I will post a link.