PSV - Snap Acting - Instantaneous flow greater than blocked flow? 1

[SlavaU336]

Hello everyone,

I used Korf to simulate a PSV opening at 9300 kPag to estimate the back pressure in the flare line.
This PSV protects a separator against blocked flow.

The instantaneous flow through the orifice is close to 150 MMSCFD. Yet the blocked flow case for this PSV is only 70MMSCFD, it seems like the PSV will release at a rate of 150 MMSCFD for a brief moment as it opens.

Does this make sense to you? and does this mean the entire downstream system has to be designed with this maximum release? even though the normal blocked flow is only half of that?
Or is it possible that the volume of piping in the flare line may act as a buffer against this instantaneous flow rate and actual back pressure be less?

Thank you for your help!

 

[Latexman]

It makes perfect sense. A PSV is either wide open, or closed (zero flow). You need to study how a PSV functions, especially the “huddling chamber”.

Good Luck,
Latexman

 

[georgeverghese]

At 50% of the rated flow, PSV flow should be stable, since it is greater than 25% of rated.
For a conventional type PSV, the PSV will open partially at this 50% flow at some pressure between set pressure and 110% of set pressure.
API rules require the tail pipe be sized for the rated flow (at some coincident backpressure I dont recall now), while the main vent/ flare headers be sized for the process design flow at the max coincident design case flow relevant to each of the headers.

 

[shvet]

What is your question? It is not clear. Are you curious which flow to be used for flare piping? In this case answer depends on PSV's design - see table 8 API 521. If this PSV is a conventional spring-loaded valve than lateral pipe to be calculated based on rated capacity while main flare header - on required one.

I do not know what instantaneous flow stands for. It is not recommended to use uncommon in-house terms for overpressure design as this leads to failure and therefore is not able to called an inherently safer design.

Note:
- if this PSV has unbalanced design (e.g. conventional spring-loaded) then actual flowrate will be equal rated capacity, not required one
- if rated capacity will be much more then required one then PSV will cycling

https://www.eng-tips.com/viewthread.cfm?qid=504687

 

[Snickster]

Seems like you would only get a higher instantaneous flowrate if the relief valve is oversized so that the flow will stabilize to equal the flow into the vessel, or maybe it will just chatter.

 

[The Obturator]

Looks to me that your simulation has calculated the RATED flow of the valve (using its fixed API-526 orifice size) against the REQUIRED flow of the blocked flow case.

The rated flow should be used for any downstream calculations.

If the rated flow is a problem, you may want to consider having a restricted flow/lift option for the valve should the design allow it.

Your post title 'PSV - Snap Acting...' suggests you have a pop (snap) action pilot opertated Pressure-relief Valve - Is that correct? If so, such a design has instanteous lift at set pressure without requiring system overpressure (though ASME requires flow certification at 10 % overpressure).


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

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

 

[SlavaU336]

Hello Everyone,

Thank you all for spending the time to answer my question, greatly appreciate it.

I should maybe have used the term "rated" flow instead of "instantaneous". Thanks for correcting me.

It seems like there are different opinions, some mentioned that the tail pipe of the PSV should be sized for the RATED flow and the flare header for the REQUIRED flow (blocked flow), while others mentioned that both the tail pipe and flare header should be sized for the RATED flow?

Right now my rated flow is a problem in the flare header and would over pressure the flare knock out drum.

As @theobturator mentioned, I should be looking at options to restrict the lift of the PSV to minimize my rated flow. I'm looking into this now.

This is not a conventional PSV but a pilot operated Snap acting one.
Does this mean that even though it has instantaneous lift at set pressure without requiring system overpressure, ASME still requires to calculate the rated flow based on a 10% over pressure?

If this PSV was of the modulating design, would this change anything in calculating the rated flow?

Also, my KORF simulation is calculating a rated flow and associated built up back pressure based on a steady state. In reality, is it possible that the system might never see the built up back pressure associated to the rated flow? I mean if the PSV opens for only half a second, releases the rated flow but the volume of the piping downstream acting like a "buffer", maybe the actual built up back pressure is not as bad?
I'm wondering if using AFTx would allow me to calculate the actual built up back pressure?

Thanks again for everyone helping me better understand the world of PSVs

 

[shvet]

You will find a guide in para. 5.4.1 API 521. Note that as per API 521 main header to be designed based on required flow regardless.

Or is it possible that the volume of piping in the flare line may act as a buffer against this instantaneous flow rate and actual back pressure be less? ... I mean if the PSV opens for only half a second, releases the rated flow but the volume of the piping downstream acting like a "buffer", maybe the actual built up back pressure is not as bad?

The emergency pressure relief technology ... applies to emergencies during which the pressure rises uniformly and at a measurable rate throughout the protected equipment. The technology does not apply to situations in which the pressure-producing process propagates through the equipment from a point of initiation (such as the ignition and burning of flammable vapors and runaway reactions that propagate through liquids. The calculated vent flow area for effective pressure relief can be large even for nonpropagating events. This problem occurs if the pressure increase is too rapid at the lowest feasible relieving pressures.

Transient conditions are not applicable to overpressure protection and treated in another way. For inlet/outlet piping it is assumed that flowrate is established all the way from overpressure source to relief discharge while a buffer you have described works another way - upstream flow is more than that downstream as a consequence of relief accumulation.

Balance - you are describing a system that is imbalanced while overpressure relief deals with balanced one (mass&heat both).

Right now my rated flow is a problem in the flare header and would over pressure the flare knock out drum.

A flare tip&stack have enough capacity to accommodate the relief while a knock-out drum located next to it has not and is overpressurized, right? Post here calculations as those seem doubtful. More details = better comments.

PS
Are you sure that an anonym forum is a proper source of instructions in process safety?

3.1.1 Required Background
The application of the methods ... requires college-level proficiency in both hand and computerized engineering calculations. The user also needs at least some familiarity with pressure relief technology and access to current editions of the following principal documents and handbooks:
...
Newcomers to this field should attend appropriate seminars or short courses on pressure relief fundamentals and practices as the need arises ...

 

[Snickster]

In my opinion if the relief valve is oversized and the rated flow is much greater than the design flow, which appears to be the case, then you would design the inlet and outlet tail pipe for the rated flow and the flare main, knockout drum, and flare for the actual design flow. However, in the first place I would size the relief valve for the actual design flows as to not get into these types of hypothetical and unnknown things that may or may not be happening physiclally during the relief of an oversized relief valve.

My reasonining for this is that the immediate inlet and outlet pipe if undersized may effect the relief valve operation if too much pressure drop occurs, so could effect the opening and closing. On the other hand I believe that the downstream larger main, knockout drum and flare will really only see design flow in any case. This is because even if you get a blast of rated flow or even if pulsations of rated flow if vavle chatters, in the end those pulsations seen by the discharge pipe will smooth out to really only equal the flow into the system or design flow. In other words the average flow rate out as seen by the downstream pipe can only be equal to the maximum flow rate into the vessel to be protected - that is, the design flow rate.

 

[SlavaU336]

Thank you for your replies. Helps a lot.

I'll be looking for webinars and books to increase my knowledge on PRVs.
@Shvet seems to be using "CCP's Guidelines for Pressure Relief and Effluent Handling Systems 2nd ed"

Do you guys recommend any other materials or online courses/ webinars that are reasonably priced?

Thank you and have a great weekend!

 

[Snickster]

API 520 and API 521 have good information on relief valves and systems.

 

[Snickster]

Also if you go to the large manufacturers of relief valves websites they have a lot of free references on relief valves and relief system design.

 

[The Obturator]

All you need to know about Pressure-relief Valves

https://www.emerson.com/documents/a...-valve-engineering-handbook-en-gb-4244934.pdf

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

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

 

[shvet]

API 520+521+526+537
Proven manufacturers' handbook, e.g. Emerson, Leser
CCPS's Guideline for Pressure Relief and Effluent Handling Systems
Malek's Pressure Safety Devices
Note that low pressure vessels and storage tanks comprise a separate world and require a separate list.

There is no a comprehensive source/set/list and many issues require a particular study, e.g. noise/dispersion modelling, local pressure vessels/piping codes, hydraulics calculations etc. An experienced mentor/tutor is most recommended as overpressure relief design is some kind of exam of one's skills&knowledge. Stuff like this is related to issues called Inherently Safer Design and Engineering/Safety Culture.

Ordinary/small/cheap does not mean simple or inconsiderable. Often an ordinary device, e.g. a spring-loaded valve, requires a huge amount of works&background like a feasibility study of possible options. For example - are you sure that a pilot-operated PSV is indeed required in your case? Why a conventional spring-loaded valve is not the best option?

 

[LittleInch]

I suspect this is a mixture between the way the analysis program is coded and some element of very short term burst of flow until the steady state emerges.

You would need a pretty good transient analysis program to tell you what the impact is on all your downstream systems as if this is a pop action, then flow is a function of orifice size and differential pressure. Then if the pressure in your vessel falls to ?? below the set pressure does the valve "pop" shut again? The issue will be whether the "rated" flow" continues for long enough to cause an issue. That depends on the volume of the PV and what the feed rate is for the vapour / gas.

What is the difference between your rated flow and required flow for long term flow from the relief?
Why the pop action pilot?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[SlavaU336]

Thank you all for the input, its greatly appreciated.

I agree, I'm not sure a snap acting pilot valve is the best option in my case.
I'm reevaluating everything and learning to use AFT to see assess the transient.

Seems like I would benefit from a mentor on PRVs. If you're interested or if you know someone who would mentor me, feel free to private message me and let me know your rate/availability.
Might not be able to accept the offer right away but in the near future.

Have a great weekend!

 

[shvet]

feel free to private message me

For info - this forum has no such feature.