Pump MAWP and shut off Pressure Criteria 2

[Sawsan311]

Dear All,

For pump fed from a suction vessel, do you agree that the estimated shut off pressure shall consider suction PSV set pressure with maximum static height at high liquid level.
Some Engineering designs considers the combination of PSV set pressure along with pump reaching shut off head at zero flow - is highly not possible and hence the below is proposed:
- maximum normal suction pressure+ liquid static high high level head+ shut off head OR
- PSV set pressure on suction vessel + liquid static high high level head+ normal rated head

Do you agree of taking the higher one of the above? or you still recommend to always design for suction system design pressure + maximum static head+ shut off head at zero flow.

Additionally, API 610 section 6.3.5 states that in estimating MAWP, 10% margin to be considered on maximum discharge pressure. Do you conclude that:
- maximum discharge pressure referred to by API 610 means the discharge pressure corresponding to the suction system design pressure, or this is up to the use to decide on the same? Does API refer to this maximum discharge pressure as the one corresponding to shut off conditions or it is the rated maximum discharge pressure?
- such 10 % margin emerges from the tolerances in table 16 to account for the excursion increases in shut off head during testing. Or it is not related?
- Do we still need to apply 5% margin on the estimated shut off conditions in accordance with API 610 section 6.1.4 to account for slight variations in impeller hydraulic design or variable speed capability?

Appreciate your views

 

[Keit]

Hello Sawsan311,

I think your question can be divided into two queries (design pressure at pump discharge, and API pump head tolerance).
I would like to reply your first query as below.

It is true that some engineering standards consider the combination of PSV set pressure at suction drum along with pump shut-off head. My proposals are:

[ol 1]
[li]If there is any specific requirement in the project specification or Company standard, we need to follow it unless any deviation is raised.[/li]

[li]As far as we can ensure no simultaneous occurrence of pump shut-off condition during suction drum PRV relieving event, I agree to set design pressure of discharge line as higher value of the following a) and b), as you mentioned. In this case, I recommend putting such justification in somewhere of the calculation sheet or design report, for traceability.[/li]
[/ol]

a) maximum normal suction pressure+ liquid static HH level head+ shut off head
b) PSV set pressure on suction vessel + liquid static HH level head+ normal rated head

(Example 1) We assume the suction drum is pressurized by inert gas, and overpressure of the suction drum occurs when the inert gas supply valve is spurious open. If mechanical auto-recirculation valve is installed at pump discharge and no other overpressure scenario is associated with the suction drum, we can consider the higher value of a) above or PSV set pressure on suction vessel + liquid static HH level head+ min. flow pump head (instead of shut-off head). This is because the spurious open of inert supply valve and mechanical failure of the auto-recirculation valve are not related with each other, so based on single-failure criteria, we don’t consider such simultaneous failure.

(Example 2) We assume the pump transfer the liquid to another drum at discharge, and the differential head is enough high. And we assume only overpressure scenario of the suction drum is reverse flow from the other drum at discharge, due to the pump trip. In this case we don’t need to consider PSV set pressure on suction vessel + liquid static HH level head+ shut-off pump head, because the pump already trips when the suction drum PRV popping.

Remark)
We should consider maximum liquid density to estimate the liquid static head if the fluid density may vary.

I will reply to your second query (API pump head tolerance) later.

 

[keeplrng]

Additionally, API 610 section 6.3.5 states that in estimating MAWP, 10% margin to be considered on maximum discharge pressure. Do you conclude that:
- maximum discharge pressure referred to by API 610 means the discharge pressure corresponding to the suction system design pressure, or this is up to the use to decide on the same? Does API refer to this maximum discharge pressure as the one corresponding to shut off conditions or it is the rated maximum discharge pressure?

See definition of Max. discharge pressure in API 610. So max. discharge pressure will be max. suction Pr. + max differential considering max. sg.
max. diff. is at shut off so need to consider that.

- such 10 % margin emerges from the tolerances in table 16 to account for the excursion increases in shut off head during testing. Or it is not related?
I suppose this 10% takes care of excursion during testing. I am not sure if that is intent to mention 10% margin on MAWP.
- Do we still need to apply 5% margin on the estimated shut off conditions in accordance with API 610 section 6.1.4 to account for slight variations in impeller hydraulic design or variable speed capability?
Process would typically specify their requirements of flow and head requirements as per system design. process normally doesnt consider this 5% margin to be considered in system design. during offer evaluation, mechanical engineer ensure that this 5% margin is available during pump selection based on operating point.

 

[Sawsan311]

thank you very much Mr. keeplrng for your great support. Answering your first reply in examples 1 & 2, shall we consider the credit of the auto-recirculation valve since a power failure scenario can be a common mode which can cause the blanketing PRV and the minimum flow recycle controller to go to opposite to its failure position.

I have another question, if the pumping system feeds a large network of sinks, would it be a credible scenario to consider pump shut off conditions in settling the design pressure of the pipeline. Some argues that considering the network as an open sink with low probability of having simultaneous blocked outlet condition driving the pump towards shut off. I agree with this argument. However, I am also thinking of the blocked suction scenario where the pump may go into starvation mode and consequently reaching shut off. in this case there will be no prevention of pipeline overpressure unless a SIL-3 SIF is present to trip the pump immediately at such conditions. Rating the pipeline system to the shut off head in conjunction with suction system design pressure (although there is no suction vessel feeding the pump) can be extremely economically challenging.

Hence, would it be safe to rate the piping system and pump casing MAWP to full shut off conditions YET, underrating the pipeline following maximum normal suction pressure in conjunction with maximum normal pump differential head (at maximum speed) & NOT at the shut off conditions.

Thanks again for your kind support.

Regards,