14 replies to this topic

[ryn376]

Is there any part in the ASME code or API standards which disallows chattering of relief valves or is it just not optimal as the relief valve may be damaged/destroyed?

[curious_cat]

What's the reason you want to allow them to chatter on purpose? 

[ryn376]

The only reason anybody would ever want that - money. Either the valve is already installed in service or multiple relief devices would have to installed on an existing vessel without additional nozzles (meaning more involved calculations [more money], more piping [more money], more valves [more money]). If the only issue is the device being destroyed then that is an acceptable risk because the only cost is money. If the vessel ruptures, then that is not, as people could be injured.

Edited by ryn376, 10 September 2015 - 08:16 AM.

[fallah]

Is there any part in the ASME code or API standards which disallows chattering of relief valves or is it just not optimal as the relief valve may be damaged/destroyed?

 

ryn376,

 

Although there might be no direct referring to PSV chattering avoidance in API/ASME, all limitations such as maximum %3 inlet loss, maximum %10 percent back pressure, minimum %25 ratio for lowest relief load over highest one,...which included in mentioned code/standard are to be established mostly in order to prevention of PSV chattering...
 

[ryn376]

In that case, if the overpressure scenario is very infrequent (e.g. fire) I see no need to replace a valve just because it could potentially chatter if it does not endanger people.

[fallah]

In that case, if the overpressure scenario is very infrequent (e.g. fire) I see no need to replace a valve just because it could potentially chatter if it does not endanger people.

 

ryn376,

 

Yes, provided that the fire case is the only overpressure scenario...

[oscarsender]

Hi,

 

The link below may help.

 

http://www.hydrocarb...rcent-rule.html

[Art Montemayor]

The subject of generating detailed relief valve calculations and storing them in a reliable, available database has always meant one thing to process owners:  Money invested.  As such, without any hard and fast rules from those existing agencies that are presumed to protect the health and welfare of human beings, some owners have previously turned a deaf ear to the common sense reality that a chattering relief valve is the onset of potential physical valve damage and failure during operation.  The potential physical damage is obvious: no relief valve has ever been designed to withstand the continuous and rapid-fire slamming of the valve's seat and plug/stem.  Sooner or later this abuse of the design is destined to failure - or inability to operate as designed.

 

Any plant or process engineer who is exposed to this situation on a daily basis and is aware of the potential consequences would never allow it to occur.  So why would the same engineer(s) who work in a designer's office and never are exposed to the potential consequences allow it as a possibility?   During my prior years as a plant process engineer I would never think of asking an operator to do a task that I was not capable or willing to do myself if placed in the same position.   In my opinion, Safety Relief Valves are never intended nor designed to operate during normal, expected operations.  They are designed and installed for the purpose of protection of humans and property in the event of an unexpected, hazardous situation.  They are expected to operate timely and efficiently in the event of a hazardous situation.  If they do not operate as designed, something is obviously wrong with the design, specification, installation, or settings of the device and this should be corrected immediately for the safeguard of all personnel and property involved.

 

I don't believe the ASME code allows or disallows the implementation of chattering safety relief valves - that decision has been left to the sound, professional judgment of the engineers involved in the design, purchase, installation, and calibration of the instrument.  That's our job, and to knownly allow an installed safety relief valve to continue to operate while chattering is, in my opinion, folly or a potentially hazardous position for any engineer to take.

 

The information found in oscarsender's submitted URL is good to know and hope is expected to find a way to generate good, accurate, and sound engineering design for installed PSVs.  More discussion in this important topic from our Forum is welcomed and awaited.

[ryn376]

In my mind failure, a relief device failure from chattering means that it will still relieve the required capacity as the valve will now be permanently open... is this correct? The real question is not, "will the valve (possibly) have to be replaced", but "is the vessel going to rupture"? Otherwise, my current client will not replace the valve.

[curious_cat]

Isn't there always potential for piping damage? Possibly catastrophic?

 

Also, in case the valve does fail open isn't there an economic cost associated with the lost of inventory / downtime / process upset etc.? Do those not outweigh the cost of resizing the valve? 

 

For some valves the noise of a chattering SRV might itself be rather annoying and perhaps even outright injurious. 

Edited by curious_cat, 10 September 2015 - 12:56 PM.

[latexman]

On post #9, I think that is wishful thinking.  Where's the data on failure mode of chattering PSV's?

[Art Montemayor]

I can recall prior Hazops where it was resolved as credible that a chattering PSV that self-destroys due to the constant pounding of the valve plug + seat would result in valve debris plugging the valves outlet and/or discharge piping and, consequently, preventing design discharge flow rate from taking place.

 

Result: pressure vessel failure due to inability to relieve as designed.

Solution: Design, install, and calibrate the PSV correctly and accurately in the first place.

[Bobby Strain]

I don't recall ever hearing or reading that a relief valve failure due to "chattering" ever resulted in a serious incident. My personal experience in operations is that the valve fails to reseat and continues to vent to some extent. And, quite often, large valves fail to reseat after relieving. Relief valve opening is more likely to occur during startup or shutdown when the design flow is not achieved, causing the valve to "chatter". There is little to be done about this. But prudence suggests that valves and their installation follow good practice. And, when existing valves are determined not to follow good practice, then some corrective action is imperative. Anything else, too, that might be problematic should be corrected, not ignored.

 

Bobby

[fallah]

PSV chattering is a consequence of an incomplete overpressure scenarios study (specifying all credible scenarios and determination relevant relief loads...)  or inappropriate sizing of PSV inlet/outlet lines sizing and, as far as i know, its possibility is normally not to be evaluated in HAZOP study especially because it's to be considered as final safeguard.

PSV failure, due to whatever reason, is normally evaluated in quantitaive risk assesment and if there would be a concern for PSV plugging due to viscous or contaminated fluid, it might a rupture disc to be provided at PSV upstream to prevent...

Edited by fallah, 11 September 2015 - 12:22 AM.

[milind kamat]

Hello, I agree with Mr. Montemayor. Please understand that safety is first and cannot be compromised at any cost. The chattering can cause the valve connection to go off and release the contents,or as Mr. Montemayor explained, plugging may occur preventing safety release the next time the overpressure occurs. Let's not speculate on something trying to save on cost and risking the lives. Please follow the API guidelines; if we try to waive certain clauses in the name of judgment, then this is where API and the authorities would want us to stick to the letter - at least currently they allow us to use SOUND judgment.

 

Best regards,

Milind Kamat