VFD centrifugal pump power ridethrough

[Billgas]

I have a building cooling water loop pump system is controlled with a VFD. The pump is 480V, 40 HP. Power is through a ATS (main/diesel). When the main fails, the VFD go down on loss of AC, the pump stops, the loop pressure drops, the diesel starts, the ATS transfers, the VFD may then restart.

The problem is that certain large HVAC equipment shuts down an loss of head pressure, and requires manual resets.

One method to solve this is to keep the vfd/pump running during the say 15 seconds while waiting for the generator to come online.

How can I keep the VFD in power?

 

[RobWard]

UPS for the VFD?

"I love deadlines. I love the whooshing noise they make as they go past." Douglas Adams

 

[waross]

I think that your A/C may be shutting down on low flow. The next stage of shutdown will be high head pressure, not low head pressure.
You may be able to add a 20 second delay to the low flow circuit. Many machines will run without cooling for more than 20 seconds before head pressures become excessive. If the head pressure is high, then the machine must shut down.
Depending on the age of the machine and possible warranty issues you may want to check with the manufacturer for recomendations.
If you go with a back-up system for the VFD, there are others here who have more expertise with VFDs than I.
respectfully

 

[CJCPE]

It is possible to back up a VFD, but that does not often turn out to be the most attractive alternative. It is usually preferable to provide all of the necessary automatic resets.

For a VFD back-up system, the necessary energy storage can be provided by a capacitor bank, a flywheel or a battery bank.

For any back-up system, you would want to determine the minimum operating speed that needs to be sustained during the back-up time and the horsepower that the motor needs to deliver at that speed. When power is lost, the VFD should be allowed to drop to that speed as quickly as possible. Some VFDs have a built-in power loss ride-thru system that uses that strategy. That alone will prevent shut down under some conditions, but not for 15 seconds unless a huge flywheel is installed between the motor and pump. Such a flywheel might interfere with normal flow or pressure regulation by speed modulation.

As stored energy is recovered from capacitors, the voltage drops. If the VFD can tolerate only a 10% voltage drop, only about 20% of the stored energy can be recovered. A capacitor bank requires a means for limiting the charging current including recharging after a line voltage dip. At a minimum, the capacitor bank needs to have protection that will prevent the consequences of a catastrophic failure from extending beyond the enclosure.

A battery bank requires a charging system, a periodic maintenance program, an appropriate enclosure and appropriate protection. The total battery voltage would need to be about 600 volts for connection to the DC bus of a 480 volt VFD.

Energy storage can be provided by a flywheel between the motor and pump as mentioned above. Another method would be to supply the VFD from a continuously operating motor generator set with a flywheel. Since VFDs are not very particular about input power frequency, an induction motor and induction generator might be sufficient. Any type of MG set would require some control components.

 

[aolalde]

You need a lot of energy to provide 40 HP for 15 seconds to the pump. The solution is very expensive, an Uninterruptible Power Supply (UPS) to feed the drive-motor-pump. You will have several types but all of them are a major investment.

http://www.toshiba.com/ind/group_display.jsp?id1=14&source=mv

 

[jraef]

...probably a lot more expensive that modifying the building controls to allow auto-reset in the case of a momentary power loss.

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[ozmosis]

If fitting a UPS to a VFD as back-up, we would normally advise an increase of kVA rating by 40~50% on the UPS to cope with the peak (non-sinusoidal) currents. This makes the cost of a UPS very expensive. I would agree with jraef and others, investigate the control sequence of your BMS to allow better control.

 

[LionelHutz]

When a VFD is decelerated it will take the mechanical energy in the motor and load and put it back into the buss capacitors. If you find the right deceleration rate you can sustain the VFD buss voltage to self-power the VFD and ride through a short power interruption. When power comes back on, the buss voltage is still about right and the VFD can then be re-accelerated. However, I would expect the flywheel required to have enough inertia to keep the pump rotating for 15 seconds would be huge.

As a side note, after a power loss you can also decelerate quickly and "brake" the motor by dumping the excess energy into a resistor using a VFD. This is where I have applied this before - when a power outage occurs give the VFD a stop command and the VFD will quickly ramp down the motor and then power-off.

The VFD has to be able to ignore the loss of input voltage for this to work.

I really would not consider a 3-phase UPS to just back-up power a VFD - it doesn't make sense. Provide a back-up source for the DC buss voltage to "UPS" a VFD.

 

[Billgas]

The building management opportunities have been investigated and are not do-able. I have been told that this inability to change has to do with computer room cooling units that would require major changes.

I have a 480V, 300kVA/240kW rated UPS loaded to 35kW available.

What I am "dreaming up" is a system as follows:

Install a solid state transfer switch. Connect source A of this switch to normal/emergency power. Connect source B of this switch to UPS power. Connect load of this switch to the VFD controller.

When N/E power goes down, the SSTS would sense loss of preferred source A and tranfer over to source B within 6mS. Once the diesel generator comes online the automatic transfer switch re-energizes the source A with N/E power, then the SSTS would transfer back to source A.

I am thinking that the VFD will just stay running during the SSTS transfer time, the motor would continue to run the pump, and all would be well.

The motor would already be spinning and any inrush would not present.

The UPS would only need to supply power for the time until the diesel picks up (about 10 seconds).

There is a VFD "bypass" feature that would be wired out of the control while on UPS.

The UPS and the VFD are fed from the same sources, be it utilty or diesel.

On return to normal I would delay the SSTS a few seconds until the UPS was repowered from the utiltiy, to limit phasing differences.


The pump is rated 767 GPM at 132 ft. Measures values show is running at about 60% speed with a suction P of 24psi and a discharge P of 48psi. I have the pump and system curve data. I am not a pump savvy and I am having trouble with figuring out how to read the flow based on the given operating condition. (I think the sppeed curve and system curve intersection would read across for the head and drop down for the flow, but these values then are different from the measured values.

 

[waross]

If the UPS and the VFD are compatible, forget the fancy switching and feed the VFD permanently from the UPS. You appear to have ample capacity.
respectfully