Load Monitors Reliability For Mag Drive Pumps

[PKEngineer]

On our LPG/LNG facility we use sealless pumps for LPG, caustic soda and ammonia services ranging from 1-40 HP. LPG mag drive pumps (40 HP) were originally supplied with power monitors for pump protection against no/low/ high flows plus some other inherent mag pumps troubles.

Our I&E department refused to use these monitors as does not have previous proven experience and adamant to depend on overamp protection available for any underload/overload conditions. One objection is that these monitors usually take current readings from only one phase and does not consider imbalance in all three phases.

Some of these pumps have low and high pressure alarm/shutdowns on suction and discharge side and some with flow meters on delivery side.

I would appreciate for passing past experiences about using power monitors for these sealless pumps.

Some of the queries are:

How are they reliable to protect mag pumps?
Any nuisance troubles like false trips etc. Can they be mitigated by using this accessory?
Current versus power monitors comparison.(We understand power number than current readings).
Do these monitors protect against dry run conditions.

I do apologize for making the statement longer and will provide any missing info if needed.

Thanks
Have a blessed year 2011 to everyone.

 

[Clvet]

First: my apologies for the length of this post.

To me, for centrifugal pumps, they work exactly as intended( don't have a lot of experience with using them on mixers and other equipment). The only time I've ever seen them not work is if they're not wired correctly, not programmed correctly, or the operators continually override the monitor when it keeps tripping out.

As far as nuisance trips, it all depends on how it's programmed. It can mean you have the extremes/upset conditions covered and your normal operation runs dangerously close to those values. It can also mean that your limits are set excessively tight or that your response delay is too short and transient effects are causing a high/low main alarm.

Current is fairly flat until about 40-60% of nameplate and then goes up in sort of an x^2 relationship. By contrast, the motor's power is continually increasing as you go across a pump's performance curve. What this means is that given a pump curve, and a power reading, you can tell where the pump is running on the curve, even without a pressure gauge. This can be especially important for finer control with flatter pump curves.

Every pump is different but in general, I program the initial motor parameters. Dead head the pump, take a reading, turn the pump off, set the lower limit a little bit above the dead head power. To make sure it will work. I start the pump up again, dead head it, and make sure it kicks off. This should protect against dead head, dry running, and decoupled operation. Since dead head would be the highest energy using of the three.

For the high trip points, if the motor isn't sized for run-out, set the high trip to be nameplate horsepower. For cavitation, you need to look where that power is going to max(where cavitation will begin to cause a drop in performance) and make sure you set the trip point somewhere below that so that it should never begin to cavitate.

Areas where I would not use power monitors:
1. Low flow pumps, if the flow is low enough, there won't be enough of a power difference between normal operation and dead head. Setting the limit would be a difficult balance between protection an nuisance trips. A flow meter/sensor might be better.
2. Variable frequency drives - if the motor speed changes, the whole pump curve sort of gets thrown out the window. Most stand-alone power monitors will not update power values on the fly, it would be better to get a VFD with built-in protection features.

 

[PKEngineer]

Clvet
Thanks for providing your insight and experience with this product.

One thing I couldn't catch is to use the deadhead band for protection against dry runnig. In case of no flow(like when someone closes the suction valve while pump is running)I am not sure this monitor will shutdown the pump or we need other protection like level or flow switch.

In case of pump and/or process stream overheat, does this monitor provide any protection (except protection against cavitation)?

Thanks

 

[ccfowler]

I don't like using current as the only means of protecting a critical pump, and even though it is a more accurate parameter to use, I still would not want to use just a power monitor alone either. I would always want to monitor flow rate, pressures (suction and discharge), and temperature (suction, discharge, and especially the differential temperature from suction to discharge).

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

 

[Clvet]

@PKEngineer: best thing to do would set a small, 1-2 second delay when programming and to make sure that it works, dead head the pump and make sure that motor trips off. Dead heading will use more power than dry running so if your lower limit excludes dead head, it will also exclude(prevent) dry running. If it doesn't trip then, it probably isn't quite set right and you're not protected. Does that make sense?

You could also set a much higher trip point(example, 50% of design flow but well above minimum flow). It would certainly protect against upset conditions that would damage the pump but it could also lead to increased nuisance trips if the flow fluctuates during operation.

@CCFowler: Current can work but only under certain performance envelopes. For a mag drive pump, I would figure that a power monitor would cost at most 25% of the initial cost. Repair of mag-drive failures can vary but I would think that if it saves you only once or twice it has paid for itself. Flow meters and gauges(pressure and temperature) are crucial for finely controlled processes but without any sort of centralized control system that they feed into, they really don't protect against anything. If you have the system already in place, the cost of the gauges can easily be justified with the increased control/monitoring of the system but if it's a small operation or a basic application like loading/transfer application? It would be nearly impossible to justify the added expense.