Sizing standby generators for large reactive loads

[mpoulton]

I discovered that the reason the Sirius would go into failure mode was because the single-phase branch supplying the Sirius's control circuits would jump to ~130V (as the generator is trying to start the motor load). Apparently, as the genset struggled to keep up with the load placed upon it by the motor, the voltage would surge and the Sirius would shut down to protect its circuitry.

Also, I just pulled up the spec sheet for that starter and the control circuit is rated for 50-60Hz, but EXACTLY 115V - the minimum and maximum voltages are both listed as 115 with no tolerance specified. Obviously it has some tolerance, but apparently 130V (13% high) is out of range. Maybe it would work properly if you fed it a bit lower voltage so the peak surge isn't quite so high? A 12V buck transformer on the control line would give it more like 108V during normal conditions (6.6% lower than rated) and 118V during motor start (2.6% high). That would be a very easy solution if it works.

 

[Jon456]

It's hard to say what the difference between those two conditions is that's causing the problem, but two things stand out to me: First, a tap-switching conditioner is not a great fit for coping with brief transient power quality problems and it can't do anything about the sagging frequency from the generator which might mess with the starter's operation.

Agreed. It was more of a low-cost proof-of-concept. Still, it did prevent the soft starter from going into fault during pump start, and even once allowed the pump to fully start and run on generator power only.

A UPS would be much better even if it's not double-conversion. As soon as the generator's output gets a little bit wonky, the UPS will switch to battery and the starter should run smoothly. The changeover happens at least as fast as the tap-switching on your conditioner.

There are a number of reasons why I'm resistant to using a UPS. Perhaps the foremost reason is its reliance on a battery which may have a short life in that environment.

Second, the generator's controller settings seem to be contributing to the problem here. It needs to not be shutting itself down due to motor starting transients. Either the regulator needs to be adjusted so it doesn't produce a voltage surge when a sudden load is applied, or the controller needs to be adjusted to tolerate a higher voltage surge without shutting down the unit.

In early tests of the generator, it was shutting down due to exceeding max current. After first consulting with Leroy-Somer (manufacture of the alternator), I was assured that no damage would occur to the genset if the max amperage setting was increased. So I did so. I'm far more hesitant to raise the max voltage setting for fear of damaging electronics in the the control panel for the pumps or our two soft starters.

I still believe that one of the problems is the phase imbalance. It seems that as the genset is trying to deliver the extra current demand on the one leg, it's over-generating on the other two legs which might explain the over-voltage condition.

You have proven absolutely that the generator is capable of starting the motor. The problem is strictly a controls issue. That generator can be made to run these pumps.

Yes, it can be made to start the pump. But it's right on the edge and not very confidence-inspiring. When we previously rented a generator during the storm blackout, it was a 70kW unit. That genset didn't hesitate in the least when starting our pumps (and at that time, our #1 pump's shaft was binding in the bearings and very stiff to turn, which is why we had to send it back to the factory for overhaul).

VFDs are finicky and can create problems, but so can electronic soft-starters... Like the problems your existing one seems to be causing you. A VFD isn't much worse. In this case it would probably fix more problems than it causes. There's no guarantee that whatever specific VFD you choose will play nicely with the generator's poor power quality, but they tend to be pretty tolerant of bad power sources. Many are rated for 50-60Hz input (great for a generator since the frequency is unstable) and wide line voltage ranges. Some will even run with a missing phase! The variable speed would be nice and you could even rig a float switch (or multiple float switches at different levels) to lower the pump speed as the reservoir level drops and raise the speed when the water is higher. The 30HP pump with no VFD provides some assurance that you can still move water even if the electronics fail.

Actually, I was hoping to replace the very old 16" pump (30HP) with a new one and put that on a VFD to be able to reduce flow in low water conditions. A smaller, less powerful pump (and with a deeper intake) running on extended cycle times would work better than trying to throttle back the 18" pump (40HP). Perhaps a VFD for both?

 

[Jon456]

Also, I just pulled up the spec sheet for that starter and the control circuit is rated for 50-60Hz, but EXACTLY 115V - the minimum and maximum voltages are both listed as 115 with no tolerance specified.

Yes, I noticed the exact same thing. Was trying to figure out how to get the voltage down to 115V. I was considering tinkering with the Tripp-Lite to reindex its transformer taps. But it was able to keep the soft starter from faulting during generator tests, so it is successful at keeping the supply power within the Sirius input tolerance range. I still can't understand why the genset overloads with the Tripp-Lite feeding control power to the Sirius, but not when utility power is fed to the controls.

A 12V buck transformer on the control line would give it more like 108V during normal conditions (6.6% lower than rated) and 118V during motor start (2.6% high). That would be a very easy solution if it works.

I'm not sure how a 12V buck transformer is to be applied given that the Sirius require 115V input. Could you elaborate please?

 

[mpoulton]

I'm not sure how a 12V buck transformer is to be applied given that the Sirius require 115V input. Could you elaborate please?

12V buck on a 120V source gives you 108V, which is only 6% low from the 115V rating and will probably work fine. When the input voltage rises to 130V, the voltage at the controller would be 117V, which is barely above the 115V rating. The 10% reduction in line voltage will help keep the voltage at the controller from rising so much that it cuts out.

 

[mpoulton]

Actually, I was hoping to replace the very old 16" pump (30HP) with a new one and put that on a VFD to be able to reduce flow in low water conditions. A smaller, less powerful pump (and with a deeper intake) running on extended cycle times would work better than trying to throttle back the 18" pump (40HP). Perhaps a VFD for both?

Yes, a VFD for both may be the solution. That would reduce the chances of a system failure due to one VFD going down, while eliminating the need for your generator to handle any large motor starting transients at all.

 

[Jon456]

12V buck on a 120V source gives you 108V, which is only 6% low from the 115V rating and will probably work fine. When the input voltage rises to 130V, the voltage at the controller would be 117V, which is barely above the 115V rating. The 10% reduction in line voltage will help keep the voltage at the controller from rising so much that it cuts out.

Oh, I thought you were referring to a 12V device. Stupid me. I understand the math.

Would you happen to know of a particular product that can buck 12V off a 120VAC input?

 

[Jon456]

Yes, a VFD for both may be the solution. That would reduce the chances of a system failure due to one VFD going down, while eliminating the need for your generator to handle any large motor starting transients at all.

When I had previously asked the pump distributor about a VFD, one of his concerns was that (according to him) VFD's generate a lot of transients and noise which backfeeds up the supply line. He said that the utility company doesn't like that and it might cause issues for us too.

I don't think it will be a big issue for the utility, because our pump station is at the end of a long run of distribution lines with no other loads. The transformers at the last pole serve only our pump station.

As for our pump station, the pump controls and starting electronics would be the devices most susceptible to interference for now. At some point, we may replace our "dumb" pump controller with an "intelligent" controller, and we may also add internet-enabled security cameras. We also have one other critical circuit at the pump station: a branch circuit feeding a group of 120V LED lights.

Do you think the VFD's might cause problems?

 

[kwired]

When I had previously asked the pump distributor about a VFD, one of his concerns was that (according to him) VFD's generate a lot of transients and noise which backfeeds up the supply line. He said that the utility company doesn't like that and it might cause issues for us too.

I don't think it will be a big issue for the utility, because our pump station is at the end of a long run of distribution lines with no other loads. The transformers at the last pole serve only our pump station.

As for our pump station, the pump controls and starting electronics would be the devices most susceptible to interference for now. At some point, we may replace our "dumb" pump controller with an "intelligent" controller, and we may also add internet-enabled security cameras. We also have one other critical circuit at the pump station: a branch circuit feeding a group of 120V LED lights.

Do you think the VFD's might cause problems?

No. If they were 2000 HP drives then you may have more concerns, the hamonic distortion from your relatively small drives wouldn't be much trouble.

 

[Jon456]

No. If they were 2000 HP drives then you may have more concerns, the hamonic distortion from your relatively small drives wouldn't be much trouble.

Thanks, that's good to know. My only prior experience with VFD's were for milling machines where the load is 5HP or less. So a 40HP load seems like a large drive to me.

 

[kwired]

Thanks, that's good to know. My only prior experience with VFD's were for milling machines where the load is 5HP or less. So a 40HP load seems like a large drive to me.

Big depends on who you are talking to. Some consider anything over 10 HP big, others maybe it needs to be over 200 HP, then there are a few that think it needs to be over 1000 HP, or at very least 4160 volts or more.

 

[Jon456]

Big depends on who you are talking to. Some consider anything over 10 HP big, others maybe it needs to be over 200 HP, then there are a few that think it needs to be over 1000 HP, or at very least 4160 volts or more.

I understand. It's all about perspective. I'd be very impressed to see a >1000 HP motor!

 

[mpoulton]

When I had previously asked the pump distributor about a VFD, one of his concerns was that (according to him) VFD's generate a lot of transients and noise which backfeeds up the supply line. He said that the utility company doesn't like that and it might cause issues for us too.

I don't think it will be a big issue for the utility, because our pump station is at the end of a long run of distribution lines with no other loads. The transformers at the last pole serve only our pump station.

As for our pump station, the pump controls and starting electronics would be the devices most susceptible to interference for now. At some point, we may replace our "dumb" pump controller with an "intelligent" controller, and we may also add internet-enabled security cameras. We also have one other critical circuit at the pump station: a branch circuit feeding a group of 120V LED lights.

Do you think the VFD's might cause problems?

I'd say that guy didn't know what he was talking about. He's a few decades behind the times. VFDs are very common these days. A large percentage of big motors are on a VFD now. Some of the problems in the past were due to early drives not being very advanced; some of the alleged problems were never very common or serious at all and are more mythological. VFDs are nonlinear loads that do tend to create harmonics and other noise on the utility supply - just like computers, electronic ballasts, and just about every other electronic device these days. The utility can handle it just fine, and your total load of under 100HP is totally insignificant to them. Facilities with huge VFD horsepower sometimes have to take steps to correct for these load characteristics - but nobody would avoid using a VFD because of this. The huge advantages always outweight these small annoyances. It is unlikely that you will have harmful interference with your other systems from the VFD noise, but if it does happen it will be fixable with shielded wiring, better wire routing, and maybe some filtering on the affected devices. No big deal.

 

[Jon456]

VFDs are very common these days. A large percentage of big motors are on a VFD now.

Thanks for the continued advice. I was doing some research and saw that there are significant differences between "inverter-rated" motors and regular motors. So I called the manufacturer of our 40HP motor (US Motors) and they said our motor is not compatible with an inverter because our motor is a "standard efficiency" motor and not a "premium efficiency" motor. She stated that using our motor on an inverter would void its warranty (although I doubt we have any remaining warranty on it anyway). My main concerns would be breakdown of winding insulation and induced shaft currents damaging the bearings.

Thoughts on using a VFD with our existing motor?

I asked if they have an inverter-rated motor that we could purchase to replace our existing motor, but their inverter-rated motors use a frame size of 324TP, whereas our motor has a larger frame size of 364TP. I have not yet contacted our pump manufacturer to see if a 324TP frame motor will mount to our pump, but I doubt upper management will want to spend ~$8,000 for a new motor plus ~$2,000 for a VFD. If it were me, I'd take that money and apply it to a used 70kW generator (like the one we previously rented) and be done with it. The rental company has a few for sale in our area for $15,000 to $18,000.

 

[mpoulton]

Thanks for the continued advice. I was doing some research and saw that there are significant differences between "inverter-rated" motors and regular motors. So I called the manufacturer of our 40HP motor (US Motors) and they said our motor is not compatible with an inverter because our motor is a "standard efficiency" motor and not a "premium efficiency" motor. She stated that using our motor on an inverter would void its warranty (although I doubt we have any remaining warranty on it anyway). My main concerns would be breakdown of winding insulation and induced shaft currents damaging the bearings.

Thoughts on using a VFD with our existing motor?

I asked if they have an inverter-rated motor that we could purchase to replace our existing motor, but their inverter-rated motors use a frame size of 324TP, whereas our motor has a larger frame size of 364TP. I have not yet contacted our pump manufacturer to see if a 324TP frame motor will mount to our pump, but I doubt upper management will want to spend ~$8,000 for a new motor plus ~$2,000 for a VFD. If it were me, I'd take that money and apply it to a used 70kW generator (like the one we previously rented) and be done with it. The rental company has a few for sale in our area for $15,000 to $18,000.

Well, it is possible you might run into a problem. The premium-versus-standard-efficiency issue is not really a big deal here. The premium efficiency motors are better at developing torque at low speeds, which is important for many applications where the load torque is not dependent on speed. But that doesn't really matter in a centrifugal pump application, where torque increases with speed and is very low at low speeds. You will not likely run into the overheating or stalling problems that can happen. However, you do still have the voltage spike situation to deal with. VFDs cause slow long-term damage to insulation in old motors due to corona discharge. The failure that eventually results happens suddenly, but only after the damage has accumulated over a long time. The bearing damage from induced current is a similar long-term accelerated wear issue. I don't get the impression that these pumps operate very often. It seems quite unlikely that it would accumulate enough running time for this to become a problem any time soon. Installing reactors between the drive and motor would be helpful too and may eliminate the potential problem.

The real question here is what the most cost-effective solution is in the long run. Running your existing pumps on VFDs will cost less than replacing the motors or the generator now to assemble a perfectly engineered system. Even if it shortens the life of the motors somewhat and causes the need for earlier replacement in several years (which is far from guaranteed, especially if the pumps run infrequently) it will probably still be cheaper than doing the replacement now.

 

[Ingenieur]

I understand. It's all about perspective. I'd be very impressed to see a >1000 HP motor!

iirc this is a 2500

 

[kwired]

Return on investment is going to be the key here.

How much run time do these motors see?

If they run frequently, running at lower speeds may reduce energy bill, but how much can vary depending on conditions. If peak demand charges come into play they may save a lot.

As mentioned if it is not an inverter duty rated motor the use of line reactors can help with some of the troubles. Also as far as high voltage peaks, in OP it was mentioned that these were operating at 240 volts - they likely can handle those voltage spikes better then if operating voltage were 480, because they likely use same magnet wire but the peaks are half what they are when operating at 480 volts.

 

[petersonra]

Return on investment is going to be the key here.

How much run time do these motors see?

If they run frequently, running at lower speeds may reduce energy bill, but how much can vary depending on conditions. If peak demand charges come into play they may save a lot.

As mentioned if it is not an inverter duty rated motor the use of line reactors can help with some of the troubles. Also as far as high voltage peaks, in OP it was mentioned that these were operating at 240 volts - they likely can handle those voltage spikes better then if operating voltage were 480, because they likely use same magnet wire but the peaks are half what they are when operating at 480 volts.

the amount of energy the pump uses is based on how much water it has to move. it does not matter much what the speed is of a centrifugal pump as far as energy usage goes. it is all about how much work it does.

 

[kwired]

the amount of energy the pump uses is based on how much water it has to move. it does not matter much what the speed is of a centrifugal pump as far as energy usage goes. it is all about how much work it does.

Absolutely, peak demand can be a factor though. In situation like this varying flow doesn't save much of anything, because all the water needs moved regardless. Something like a boiler pump however can be slowed down during low demand periods and can save some energy use, but in that case it is how much heat is moved that is more important then how much water is moved.

 

[Jon456]

Lots of good info to consider, thank you all for assisting. At least two people have asked how often the pumps run. That is difficult to quantify as it depends on the weather. But I shall try to give some kind of guesstimate.

Our ~85 acre property is below mean sea level and is bordered on the north and south by a split tidal creek; most of the property is undeveloped land. The entire property is contained within a perimeter levee. Any rain that falls on the property, plus the constant influx of ground water, must all be pumped off the property into the creek. We have a system of drain pipes and open ditches that carry all water to a single holding pond; our two pumps are located at that pond. If the pumps fail to operate, the property floods. The grade is essentially flat so there is no natural fall: the water moves to the pond by virtue of the pumps drawing the water level down. Because of the distances the water must flow to get to the holding pond, the pumps never operate in a continuous manner. They turn on (alternately), pump the water in the pond down, then turn off while the pond refills.

In the summer, the groundwater is able to evaporate at a greater rate than what flows into the property, so the pumps basically do not run at all during the dry season. They are manually cycled briefly on a periodic basis just to keep them exercised and prevent the shafts from seizing up. During the wet seasons, it depends on the amount of rainfall (and how saturated the ground is), but the pumps can cycle continuously throughout the day & night, running as frequently as 5/5 (5 minutes on, 5 minutes off). I try to minimize the frequency of cycling by raising the "On" float during the wet season. And I've been pushing the property owners to widen/deepen the trenches and replace undersized drain pipe (with some success), to improve the flow of water into the pond. This is also why I originally wanted to explore putting VFDs on our pumps: so we could reduce pump output to more closely match the speed at which water flows into the pond. Running the pumps continuously during heavy rain, or on longer "On" cycles during light rain, should greatly improve their longevity over frequent short on/off cycles.

 

[Barbqranch]

Would they go for adding a third, smaller pump (10 hp?) set to operate before the water gets high enough for the bigger pumps?