9 replies to this topic

[AndyK]

I have a system with a centrifugal pump, were the shut-in pressure is equal to 105% of the discharge pipework’s design pressure. Conventional wisdom would be if the design pressure can be exceeded a relief valve should be installed, however if I install a relief valve I would set it at the design pressure and taking account of the 10% allowable accumulation, then during relief during relief the system would still see 105% of design pressure. In other words with a without a relief valve, during a shut-in case the system would be overpressurised by the same amount of 5%. In this case I can’t see any justification for installing a relief valve, however the client is arguing that as the design pressure can be exceeded a relief valve should be installed. Would it not be fairer to say” A relief valve should be installed if the system design pressure can be exceed by more than 10%”. Does any one have any reference source which states that a PSV is not required in this case.

[Art Montemayor]

Andy:

In order to clarify and fully understand what you are presenting, lets go back to basics.

If you have an installed centrifugal pump delivering fluid to a process, the associated discharge piping for the pump should be designed to withstand the MAXIMUM process pressures imposed on it. This is a basic design requirement. You don't tell us if the pump's dead head pressure is a recent result - or if it has always existed. It really doesn't matter for considering what the situation should be: SAFE. If you have a dead head pump pressure exceeding the Maximum Allowable Working Pressure (or Design Pressure) of the piping AND you require that high pressure, then you had better change the piping to a higher rating.

If you do not require the high dead headed pressure to operate the process (in other words, the dead heading is just a possible process excursion), then you should do one of two things to protect the piping (assuming the piping is the lowest pressure rated item in the system):

1) Install a PSV set at the MAWP (or design pressure) of the piping, on the piping; or
2) Trim the size of the impeller used in the centrifugal pump so that the dead head reached as a maximum is below the MAWP of the piping.

I would do the latter. It is simpler, safer, and more reliable.

[AndyK]

The lowdown
- Existing pumps are been replaced with higher flow, higher head pumps to meet revised process conditions
- Shut-in pressure of new pumps will be 320 barg
- Design presssure of existing discharge pipework is 300 barg
- Discharge pipework is considerable so changing out existing pipework is not a feasible option
- If I install a PSV it would be set at the pipework design pressure of 300 barg. If the pump deadheads and high pressure trip fails the PSV would have a fully open pressure of 10% above design pressure, i.e 330 barg. As pump can only generate 320 barg, the pipework would only see 320 barg, therefore whether I install the PSV or not the pipework would see 320 barg during a shut-in case, so what's purpose does the PSV serve.

[Art Montemayor]

Andy:

Your client expanded an existing process to get more production or efficiency. The pumps were changed for bigger pumps and the piping was not changed to match the new pump conditions. The MAWP of the existing discharge piping is 300 barg. Your new pumps will exceed that if dead headed. That is what I understand of your present situation.

My position, as a project or process engineer on that project, would be to protect the personnel and the equipment - in that order of importance. In order to ensure that, I either install an adequate, safe PSV on the pump's discharge or I replace the entire discharge piping out with properly, designed piping that can easily support the excessive dead head (this should have been done at the very outset of the modifications since it should have been foreseen by the design team).

Popping a PSV at 320 barg is not going to be fun or simple. I would have insisted that the piping be changed from the very outset, but that is now history and clients are getting cheaper and tighter.

[AndyK]

Lets assume for the moment, that replacing the discharge pipework is not feasible, so you would install a PSV. In which case I presume you would set it at 300 barg and therfore due to the 10% allowable acccumulation during relief the pressure in the 300 barg pipework would rise to 330 barg, assuming the system was capable of generating 330 barg. In this case the pump can only generate 320 barg, so the maximum pressure during relief would be 320 barg. So I still ask in this case, what purpose does the PSV serve. With or without the PSV the pipework will see 320 barg during a shut-in case. The PSV will not limit the pressure to 300 barg.

[skearse]

I would have to agree with Art here. While you are correct, your system would not get to, or exceed, the 10% overpressure, you are still required to provide a relief device that BEGINS to open at the design pressure: 300 barg. All the accumulation/overpressure allowed means, in terms of PSV operation, is that the PSV must be fully opened at that accumulation. You still must protect both personnel and equipment from pressures that exceed the DESIGN pressure, NOT the design pressure plus accumulation.

[djack77494]

The ASTM/ASME piping standards, or at least 31.3, allow short term excursions above the design pressure for piping. There are two separate scenarios provided that vary for frequency and duration of excursion events. Since, I suspect, these pressure excursions are relatively rare events, it may well be permissible to allow for short term conditions where the pressure in your piping exceeds the design pressure. Please visit the applicable codes for verification if you choose to go this route.

Another alternative that your company or client may accept is to instrument your system with highly reliably instrumentation which will act to prevent such excursions. In my experience this approach is often used in "tight situations" such as your own where no good alternatives exist. If you go this route, please be aware that the requirements for your instrumentation may be quite severe.

[Qalander (Chem)]

The ASTM/ASME piping standards, or at least 31.3, allow short term excursions above the design pressure for piping. There are two separate scenarios provided that vary for frequency and duration of excursion events. Since, I suspect, these pressure excursions are relatively rare events, it may well be permissible to allow for short term conditions where the pressure in your piping exceeds the design pressure. Please visit the applicable codes for verification if you choose to go this route.

Another alternative that your company or client may accept is to instrument your system with highly reliably instrumentation which will act to prevent such excursions. In my experience this approach is often used in "tight situations" such as your own where no good alternatives exist. If you go this route, please be aware that the requirements for your instrumentation may be quite severe.

Dears,
Further to what is rightly pointed out by 'Doug' in his above post's 2nd paragraph;

May I add that we were once in similar tight situatuation with our previous employer's and initially manual 'Discharge valve throttling' was resorted to.

However susequently we went on to

1)Install a good quality Spill-back control valve on dscharge adjacent to the NRV
2)Availed change of service Possibility for one of the discharge side pipelines incidentally.

This provided sufficent outflow X-sectional area and avoided subjecting the discharge system to higher than rated pressure on sustainable basis.

Hope this proves helpful/ guiding indeed!

[Qalander (Chem)]

The ASTM/ASME piping standards, or at least 31.3, allow short term excursions above the design pressure for piping. There are two separate scenarios provided that vary for frequency and duration of excursion events. Since, I suspect, these pressure excursions are relatively rare events, it may well be permissible to allow for short term conditions where the pressure in your piping exceeds the design pressure. Please visit the applicable codes for verification if you choose to go this route.

Another alternative that your company or client may accept is to instrument your system with highly reliably instrumentation which will act to prevent such excursions. In my experience this approach is often used in "tight situations" such as your own where no good alternatives exist. If you go this route, please be aware that the requirements for your instrumentation may be quite severe.

Dears,
Further to what is rightly pointed out by 'Doug' in his above post's 2nd paragraph;

May I add that we were once in similar tight situatuation with our previous employer's and initially manual 'Discharge valve throttling' was resorted to.

However susequently we went on to

1)Install a good quality Spill-back control valve on dscharge adjacent to the NRV
2)Availed change of service Possibility for one of the discharge side pipelines incidentally.

This provided sufficent outflow X-sectional area and avoided subjecting the discharge system to higher than rated pressure on sustainable basis.

Hope this proves helpful/ guiding indeed!

[fallah]

Lets assume for the moment, that replacing the discharge pipework is not feasible, so you would install a PSV. In which case I presume you would set it at 300 barg and therfore due to the 10% allowable acccumulation during relief the pressure in the 300 barg pipework would rise to 330 barg, assuming the system was capable of generating 330 barg. In this case the pump can only generate 320 barg, so the maximum pressure during relief would be 320 barg. So I still ask in this case, what purpose does the PSV serve. With or without the PSV the pipework will see 320 barg during a shut-in case. The PSV will not limit the pressure to 300 barg.

What is operating pressure of the system?

If there is considerable difference between operating pressure and shut-in pressure (let say 30%),you can set the PSV on value such that after adding 10% allowable accumulation the pressure of discharge line would approach to around 300 barg.

Anyway,as per ASME B 31.3 there in no concern about short term exposure to pressure above design pressure.