Tandem pumps: How to prevent overpressure of second pump?
Tandem pumps: How to prevent overpressure of second pump?
(OP)
How to best protect the second pump (P2 here) from failing due to being over pressurized?
Please see attached pdf for sketch. P1 pumps water through GAC1 via level control of T1 Contents. Under some circumstances a large DP is possible across GAC1; P2 is being added to provide pressure boost for flowrate maintenance (GAC2 can also see a large DP). P2 paces P1 flowrate via pressure control in the suction line. That is the fundamental operational and control intent. My concern is that under some conditions (operator error, possibly control excursions, and perhaps startup) it would be possible to expose the suction side of the P2 casing to nearly 100 psi from P1. Some casing areas could be exposed to 200 psi under worst case conditions of both pumps on and dead-headed, I believe. The fluid is ambient temperature water, relieving it to the deck is fine.
My reaction is to simply add a PSV (the one set at 30psi), but having never done this before I want to ask: Is that a perfectly acceptable solution, or would best practice in this circumstance dictate use of a pump rated to the highest pressure it might see?
Centrifugal pumps, GAC1&2 are pressurized.
Thank you for any guidance you can provide.
Please see attached pdf for sketch. P1 pumps water through GAC1 via level control of T1 Contents. Under some circumstances a large DP is possible across GAC1; P2 is being added to provide pressure boost for flowrate maintenance (GAC2 can also see a large DP). P2 paces P1 flowrate via pressure control in the suction line. That is the fundamental operational and control intent. My concern is that under some conditions (operator error, possibly control excursions, and perhaps startup) it would be possible to expose the suction side of the P2 casing to nearly 100 psi from P1. Some casing areas could be exposed to 200 psi under worst case conditions of both pumps on and dead-headed, I believe. The fluid is ambient temperature water, relieving it to the deck is fine.
My reaction is to simply add a PSV (the one set at 30psi), but having never done this before I want to ask: Is that a perfectly acceptable solution, or would best practice in this circumstance dictate use of a pump rated to the highest pressure it might see?
Centrifugal pumps, GAC1&2 are pressurized.
Thank you for any guidance you can provide.





RE: Tandem pumps: How to prevent overpressure of second pump?
Simply ensure that the pump vendor knows that the max possible suction pressure is 100 psi, and they will select an appropriate pump.
RE: Tandem pumps: How to prevent overpressure of second pump?
It looks like each pump has a BEP dP of 100 psi. I say that because a 120 shutoff pressure would be about right for a 100 psig pump. If I make a typical curve using that data, you'd get 30 psi at around 1500 gpm.
But it makes no sense to have a PSV set at 30, for a pump with discharge pressures running between 100 to 120. Even worse, why set PSV1 at 120 when you have one behind it set at 30??? The 30 will blow first. I think you might need to do it because the pumps are only designed for 120 psi and you're going to blow up the second one, if it ever gets above 150 or so. And make sure that the suction of #2 is designed for that 120 psi, or is that what the 30 psv is trying to prevent.
Use the right equipment to do the job properly, or don't do it at all. If it's possible it's expensive. If it's not, then it becomes dangerous.
From "BigInch's Extremely simple theory of everything."
RE: Tandem pumps: How to prevent overpressure of second pump?
BEP for GAC1 and 2: Unknown, subject to enormous spread in conjecture and we have to live with that. We will not get reliable flow/loss numbers and can only say that the DP for 750gpm could range from, say, 10psi to 100psi. Neither vessel may be pressurized beyond 130 psi, so we can't ditch P2 and select P1 for higher pressure. If the GAC units flow 750 at low DP then P2 will get by-passed, but intermediate or higher DP's will require both pumps.
PSVs on top of GAC units are there to protect the vessels alone and since there are several valves between the discharge of each GAC and the next pump or effluent point I want the dedicated 120psi reliefs.
The "set 30" PSV is exists only to ensure that the suction side of P2 won't see damaging pressures (though it also limits the discharge pressure of P2 to acceptable levels). That's really my question: If I have a system such that this P2 might (in event of error or huge excursion) be exposed to 100ish psi on the suction side, is it fully reasonable/acceptable/SOP to prevent that with a PSV on the suction line, or does good practice dictate a pump that can withstand that pressure?
RE: Tandem pumps: How to prevent overpressure of second pump?
Just spell this out to your pump vendor, this sounds like a non-issue to me.
RE: Tandem pumps: How to prevent overpressure of second pump?
From "BigInch's Extremely simple theory of everything."
RE: Tandem pumps: How to prevent overpressure of second pump?
What is the Code of Construction? You might be able to limit the intermediate pressure without a relief valve. I am not as concerned about the overpressure protection as I am about the control philosophy. On first glance I think the answer lies therein. If I was to elaborate on this, I'd need a P&ID. But my assumption is that the control philosophy has already been thought through and established by experienced and intelligent engineers to a level beyond what I have gleaned from the sketch. Otherwise, my initial thought is to control the upstream pump P-1 on level just as you have shown and then use a flow transmitter downstream of P-1 that sends the corresponding flow signal to the VFD for P-2. One of three situations will exist: (1) the flows match and the end result is achieved; (2) P-2 flow is not as high as P-1 flow, resulting in system pressure rise between P-1 and P-2 to some equilibrium condition; (3) P-2 flow starts to run-away from P-1 flow, resulting in system pressure decline between P-1 and P-2 to some equilibrium condition. In any case, you seem to have this covered by pressure control on the VFD for P-2.
I am wondering if you really need the PSV (30 psig) at all? Depending on your Code of Construction you might get away with a pressure override on the controls you already have in place. The trouble that I see with the intermediate PSV is that you probably never want it to open; if it ever did, the dynamics of sudden low flow to P-2 might harm the system more than the high suction pressure. With ~ 1000 USGPM (how far to run-out?) @ 100 psi and a pump efficiency, say, around 0.5, I arrive at an approximate pump size of 100+ hp. You probably want to be as easy on those pumps as you can be, especially if there is already a concern about these levels of pressure on the suction side.
Just my thoughts...let me know if I am way off base.
Regards,
SNORGY.
RE: Tandem pumps: How to prevent overpressure of second pump?
I must be a boring person with no life...
Anyway, maybe what one might try is to move the the level control out to the downstream pump P-2 and the pressure control to the upstream pump P-1. The net system outflow is then set by P-2 while P-1 maintains adequate pressure profile through GAC-1. This might be better than the feed-forward flow control I thought of earlier. If you have enough pump at P-1, the overall system might line out acceptably. You would probably still need the pressure override on P-2 to slow it down on decaying pressure downstream of GAC-1 for the case where the GAC-1 pressure drop is rising and, consequently, shifting P-1 leftward on its characteristic. Presumably, you would ultimately go down on high level at T-1.
You might need to be able to lock out the whole control scheme during start-up so that you could start P-1 before P-2.
A gas-padded buffer tank / vessel between the two pumps might be among the better things to consider. In any event, due to the transient effects that might be caused by the intermediate PSV lifting, I would personally be looking for a way to delete it.
Regards,
SNORGY.