31 replies to this topic

[Justin P]

Query to those with operational experience:

Do any of you have any experience with or awareness of the practice of removing relief devices (PSVs) for maintenance while the protected system is still in service (and not having any other means of fallback protection)?

 

The concept of doing this is completely absurd to me, but a discussion with a colleague revealed that a large owner/operator does this.

 

I've always assumed that nobody in their right mind would pull a valve while the equipment is in service--feels very dangerous.

 

thanks,

Justin

[Bobby Strain]

Justin,

    API recommended practices recognize that there is sometimes a need to remove a relief device while the equipment is in service. You should take a look there to find answers.

 

Bobby

Edited by Bobby Strain, 14 August 2013 - 08:21 PM.

[fallah]

Justin,

 

I haven't seen any statement regarding your query, named gaging of a PSV, in relevant API standards...

 

It normally hasn't been allowed but ASME Sec.VIII Div.1 (Appendix M) just allows gagging or taking a PSV out of service for maintenance/testing by closing relevant block valve if the valve is permanently attended 'manned' by an Operator. It is assumed that the Operator is in communication with the Control Room, or can see a pressure gauge that will hint him/her any overpressure that must immediately be relieved. In fact the Operator plays the role of a Pressure Safety Relief 'Device' in this situation means that Operator cannot walk away from that valve for any reason, till the gag is removed...

[gegio1960]

Justin,

the procedure described by Naser is the one adopted in refineries.

[quiet.life]

Or if there is a bypass around the PSV, that can be left cracked open so that the over pressure situation doesn't arise. Some gas will be continuously venting to flare till the PSV is put back into service again & the bypass is closed.

[fallah]

Or if there is a bypass around the PSV, that can be left cracked open so that the over pressure situation doesn't arise. Some gas will be continuously venting to flare till the PSV is put back into service again & the bypass is closed.

 

Hi,

 

The procedure you did mention above isn't almost practical based on following reasons:

 

1) The by pass line size, if there would be, is normally 2" in most cases regardless of the PSV size itself; because is to be used just for maintenance case...

 

2) Then for cases in which PSV/inlet line/outlet line set up have capability of relieving more pressure than a 2" line,  by pass line cannot relieve the pressure (in an overpressure case due to a process upset) such that to be able to prevent the relevant system to go in non allowed overpressure situation..(more than 10%, 16%, 21%,..., based on the case)...

 

3) Left open situation of the by pass line while the system is in operation, will lead to huge loss of inventory...

[quiet.life]

Hi Fallah

Your points are valid, but this may be the only practical solution available in this scenario. I saw this procedure being followed by the client in an oil refinery during commissioning when I was there from the LSTK side. I dont know if this is allowed as per codes or not.

This was done for a PSV at column top where the PSV required some maintainence and had to be removed for the same for some time. The pressure in that column was monitored regulary.

I understand that this procedure may not protect the equipment against major contingencies like fire case, blocked outlet etc.

Regards

[Justin P]

Everyone,

 

Thank you for your input. Fallah pointed me to the reference that answers my questions.

 

Justin

[Technocrat]

Is there any spare installed to that PSV which is being removed?

[fallah]

Is there any spare installed to that PSV which is being removed?

 

Technocrat,

 

The discussion has mostly focussed on the cases with a single PSV...PSV with installed spare has no problem in maintenance standpoint...

[Chilaous]

Hi Fallah,

Thanks for your comment but I didn’t understand how an operator can relieve the pressure when the relief valve is removed and is not in place. Do you mean an operator must be there to open the block valve and relieve the gas into the atmosphere? And what if there is a toxic gas in the system?

[fallah]

Do you mean an operator must be there to open the block valve and relieve the gas into the atmosphere? And what if there is a toxic gas in the system?

 

Chilaous,

 

Yes, the operator should be ready there to open the block valve and relieve the gas into wherever it was previously specified to go...might be flare or might be atmosphere...

[Chilaous]

Fallah,

It was previously specified to go to the flare (closed system) but now that the PSV is removed it has to go to atmosphere (not safe) because as you said the 2” bypass hasn’t got enough capacity.

[fallah]

It was previously specified to go to the flare (closed system) but now that the PSV is removed it has to go to atmosphere (not safe) because as you said the 2” bypass hasn’t got enough capacity.

 

Chilaous,

 

Please see attached shows a PSV system connected to flare...Then please let's know: How can the relieving fluid go to atmosphere if PSV upstream/downstream block valves are closed, PSV is removed and the manual by pass valve is opened?!!!

Attached Files

•  PSV.pdf   143.66KB   116 downloads

Edited by fallah, 20 August 2013 - 06:05 AM.

[Chilaous]

Fallah,

The manual bypass valve can’t be open all the time (loss of inventory). So it only needs to be open if there is an overpressure scenario in the system. Based on what you’ve attached, the PSV is 6Q8. That means the 2” bypass can’t handle the relief load of the PSV. So the operator must open the upstream bypass valve to prevent / mitigate overpressure in the system and that means relieving the gas into the surrounding atmosphere which could be unsafe!

[fallah]

Chilaous,

 

Please read all posts of this thread carefully. Then you will see that using manual by pass valve for relieving the pressure isn't a logical and normal procedure but as OP requested to know about possibility of removing PSV while the protected system is still in service i referred to Appendix M of ASME Sec. VIII Div. 1 included the conditions based on which one can do so.

 

The sample i attached is a typical one and yes, a 2" by pass valve might not be able to relieve the over pressure same as 6Q8 PSV and i myself referred to this point in post No. 6, but you should consider that the code allowed gaging in general provided that in any case shall not lead to passing maximum accumulated pressure of the system to be protected. Then if PSV is removed in a short time, let say, e.g. for installation of the spare one; operator might be able to manage any overpressure using by pass valve, otherwise a pipe spool (should already be prepared) can be fitted in place of removed PSV to be able to relieve the over pressure toward flare by opening upstream/downstream valves of the PSV.

Edited by fallah, 21 August 2013 - 11:26 AM.

[Chilaous]

Fallah,

Ok now it is correct. When we recommend something we should mention it completely. However when a vessel has a single installed PSV and there might be other overpressure scenario than fire (shouldn’t be normally like this but I’ve seen some cases other than fire and only one installed PSV), replacing a PSV can take one hour while overpressure can be a matter of a few second. So special attention must be paid to a vessel which is connected to a high pressure source and has only one installed and active PSV (weird but possible).

[curious_cat]

hen if PSV is removed in a short time, let say, e.g. for installation of the spare one; operator might be able to manage any overpressure using by pass valve, otherwise a pipe spool (should already be prepared) can be fitted in place of removed PSV to be able to relieve the over pressure toward flare by opening upstream/downstream valves of the PSV.

 

The procedure sounds dangerous to me in cases where a flare discharge is needed. How long would you say it takes to block both valves, remove PSV, add in the spool? Even under best conditions. I am curious from an operational perspective. 

 

During that period the vessel is essentially unrelieved, right? Besides multiply the risk by two since same occurs in reverse while fitting PSV back in. 

 

For atmospheric discharge this seems ok, but for the rest I wonder if this workaround isn't too risky. 

Edited by curious_cat, 05 September 2013 - 01:05 AM.

[fallah]

 

The procedure sounds dangerous to me in cases where a flare discharge is needed. How long would you say it takes to block both valves, remove PSV, add in the spool? Even under best conditions. I am curious from an operational perspective. 

 

During that period the vessel is essentially unrelieved, right? Besides multiply the risk by two since same occurs in reverse while fitting PSV back in. 

 

For atmospheric discharge this seems ok, but for the rest I wonder if this workaround isn't too risky. 

 

 

Hi,

 

Suppose we did focus on "silgle PSV+2" parallel by pass" configuration terminated to flare network... 

 

Yes, the procedure appears to be dangerous at the first glance, but to rely on code allowance, on one hand don't appear there would be a better alternative and on the other hand as per the following reasons no incident is expected to be occured:

 

1) The single PSV is normally be applied for only fire case...

2) In fire case it normally isn't expected the PSV would protect the vessel from failure...Then in many cases a small PSV might be applied just for code satisfaction...

3) During the PSV removing till putting new one the by pass valve can be left open to prevent worrying about any overpressure within mentioned period...

[curious_cat]

What about using a larger line as the bypass line instead of 2"? Won't that be safer at very little marginal cost?

 

Say making the bypass line the same size as the PSV line? Then even in manual mode the vessel will always be protected. No temporary spool needed as well. The operator just stands by a different valve as his manual bypass mode?

 

A removable orifice plate beyond the valve can be added if slow pressure reduction via bypass is desirable at vessel shutdown. 

[fallah]

Hi

 

In "only fire case" situation and as per my explanations in previous post about the fire case, no need to enlarge the by pass line from 2" size. For the situations other than "only fire case" with a single PSV (rarely occured), using by pass line with almost the same size of the PSV orifice area could be studied in economical and safety standpoints...

[Babu Prasad]

With reference Mr. Justin question, generally, in the Petrochemical Industries, most of the equipment will be equipped with standby PSV to release main PSV for overhauling and test purpose. Single PSV installed where the protected equipment got standby one (typically filters, pumps). Hence related equipment can be isolated and depressurized prior to removal of the single PSV irrespective of whether PSV designed for fire/thermal relief purpose.  

In Petrochemical Industries, during normal operation time even though most of the PSV (designed for fire service ) does not come into picture however any normal operating condition can become uncontrolled situation in the petrochemical plant where operator cannot relieve the pressure by opening the bypass of PSV to control the pressure in case of fire or ay other operation abnormality. Hence each single PSV needs HAZOP studies by operating personnel and proper procedure needs to be established for maintenance job according to its service. Sometime Single PSV used in different purpose in the petrochemical plant and it can be released without isolating the equipment.

Example, thermal PSV is provided in the propane or ethane liquid pump suction line to protect pump from over pressurization due to thermal expansion when pump kept isolated with liquid. In that case during outage of PSV pump suction & discharge valve will be kept locked opened to protect from over pressurization. Otherwise Pumps isolation valves must be spaded at immediate flange to remove this single PSV which leads to pump outage and additional spading jobs. So it is purely based on operation/ technical department to decide every PSV according to HAZOP studies and guidelines to release the single PSV.

Edited by Babu Prasad, 14 September 2013 - 07:12 PM.

[seeker_of_knowledge]

Hi,

 

With regards to this topic, I have couple of questions. If there is no spare PSV (say on equipment A) installed nor there is any bypass around PSV which is designed to handle full relief load, then is it acceptable to rely on a PSV located on adjacent piece of equipment (say equipment if the relief line to equipment B  PSV is locked open while the PSV (on equipment A) is taken out of service? If yes, then I think I have to determine overpressure which  makes me ask next question. In order to determine overpressure, shall I use (PSV set pressure +pressure loss in the system) OR (PSV relieving pressure+pressure loss in the system)?

 

Thanks

 

Fez

[fallah]

Hi,

 

It might be possible and depends on design pressure/MAWP of equipment A and B and set pressures of two PSVs,...

 

Please provide such info and along with a simple sketch of the system...

[seeker_of_knowledge]

Thanks Fallah. Please find attached simple sketch. Hot oil is on the shell side (lower pressure side). Exchangers are 2x50%. PSV set pressure is 295 psig, MAWP shell side is 295 psig. MAWP tube side is 650 psig. My Questions are:  For tube rupture case, If PSV A is removed from shell side for maintenance and operations rely on PSV B as a fall back protection while both exchangers are in service, then is it acceptable? Pressure loss I calculated is: 12 psi in the shell of exchanger & total Line losses (upto PSV B inlet) are 2 psi. Corrected hydrotest pressure for shell is 440psig. Piping is ANSI 300#.

 

Hope I have made my point clear and if not let me know please,

 

Regards,

 

Fez

 

 

Attached Files

•  Sketch.xlsx   14.19KB   23 downloads