Generator and UPS

[Low Roller]

I'd like to know why a backup generator would seem to play nice with more expensive UPS systems, but not with typical consumer-grade UPS systems that are passive or have AVR.

A little background:

I'm getting a 13kW Generac standby generator installed at my condo in a few weeks. While my condo is a nice upgrade from my little downtown apartment I used to live in, the power reliability in my current area just plain sucks, specifically the circuit I'm on.

Anyway, I've been reading some of the documentation on my generator, and one thing I noticed was UPS recommendations for picking up the power during the interval between utility power cut and generator hand-off.

Generac seems to recommend Minuteman UPS's, and states that total UPS loading for Minuteman single-phase UPS's installed can be up to 50% of the generator's output capacity.(page 10) While I don't need that much, Generac also states that if using cheaper AVR UPS systems(like the two Cyberpower units I have), not to exceed 15%-20% of generator capacity for UPS backed-up equipment.

Between my two desktops, server, 2-3 large LCD monitors, and other home office stuff, as well as 58" plasma TV, receiver, and other home entertainment stuff in the living room, I'm pretty sure I'd exceed my generator's max UPS compatibility using cheap UPS systems.

I've read a lot on this topic in the last couple days to get a better understanding of the dynamics in play, but I'm pretty much on information overload at this point. If anyone has a plain explanation of what I need to look out for, I'm all ears.

 

[keenan]

You'd have to ask the generator manufacturer exactly what they mean to know for sure, but I suspect that their recommendations are due to power factor. It really depends on the specifics of the UPS units in question, but many don't include PFC. Though this would mostly apply to double-conversion units; I doubt the battery charger in your cheap line interactive units is going to create a load anywhere close to the rating on the UPS, even without PFC. In lieu of speaking about things like power factor that the consumer isn't going to know about, they just make a specific recommendation they know will work well.

I think I'd still rather have a double conversion UPS behind a standby generator though, just to feel more safe about my gear, and to ride out transients better, so if you go that route be sure to get one with PFC.

 

[Low Roller]

Thanks for the info. There's quite a lot of information to digest that I hadn't previously considered.

The generator documentation suggests double-conversion units are the most compatible - up to 25% of generator capacity with single phase products, or 50% capacity for Minuteman single-phase, double conversion units. Three-phase models with harmonic filters from any manufacturer can be used up to 80% of this generator's capacity. (wayyyyy beyond what I need, however)

 

[Low Roller]

I found some really good information on this topic here.

Uninterruptible Power Supply (UPS)

A UPS is not an acceptable load for many small generators for a variety of reasons such as harmonics and power factor. If you contemplate this approach, get assurances from the generator and UPSmanufacturers that their products will work with each other. Only an online UPS (the most expensive type) might be able to deliver any sort of correction for the poor power from a small generator. Then again, the UPS will expect to have a constant supply of power to charge the batteries, and if the generator is used infrequently, the batteries may be drained when the generator is started. A constant deep drain and recharge of the batteries significantly shortens the battery life resulting in frequent battery replacement.

This seems to match up with some of the other explanations I read(that were very dense and pretty much over my head), and it does so in a much more concise manner. It seems compatibility issues stem more from line harmonics than total load. The generator's voltage regulator does not interact well with a typical UPS, which are known to introduce quite a bit of THD into the line upstream.

 

[Surly73]

I found some really good information on this topic here.



This seems to match up with some of the other explanations I read(that were very dense and pretty much over my head), and it does so in a much more concise manner. It seems compatibility issues stem more from line harmonics than total load. The generator's voltage regulator does not interact well with a typical UPS, which are known to introduce quite a bit of THD into the line upstream.

If the UPS is not online and it's just charging the battery when the AC power is on, why are lots of harmonics being introduced upstream? In a line interactive UPS the only thing actually on the circuit is the battery charger and the load is passed straight through via relay. During switchovers the matter may be entirely different, of course.

 

[keenan]

Because battery charger implies AC->DC converter, which implies rectifier, and rectifiers produce a lot of high-frequency crap on the power lines (in addition to having a pretty bad power factor).

Though I don't think noise is the problem, I suspect the reason they make these recommendations is solely power factor. A device with low power factor will cause a lot more current to flow in the lines than it is actually consuming, so everything has to be de-rated as a consequence.

 

[Low Roller]

A device with low power factor will cause a lot more current to flow in the lines than it is actually consuming, so everything has to be de-rated as a consequence.

Keenan I think you just hit the nail on the head!!!

The main concern with a generator-UPS system seems to be the UPS tricking the generator's voltage regulator to send far more power into the house than it is actually consuming, and, well, 'pop' goes the weasel. At least that's the jist of it from all the documentation I've been pouring through. And I'd bet you are correct that PF is the biggest concern.

I need to learn a lot more about power factor, but for the sake of my setup, I think I have enough of a basic understanding of what's going on, and why its vital for me to use double-conversion, high power factor UPS's.(I'll be going with Generac's recommendations and use Minuteman PSU's that meet the criteria. Not an inexpensive proposition, though.)

Thanks for your input, keenan. It was very helpful.

 

[punisher]

Any small generator (non inverter) will introduce noise into the output. The waveforms certainly are not as pretty as what you normally get out of the wall. There have been attempts to fix this (skewed rotors, better electronics) etc but you will never get as pretty a waveform with a backup/portable. Cheap 2 pole brushless rotor generators are the worst (+10% THD sometimes), while the 4 pole 1800 RPM brushed alternators generally have the cleanest power. It's sobering to put an oscilliscope on your generator's output and check the waveform.

I run a power conditioner with my main entertainment center/HTPC. Everything else in the house works fine with the dirty power I get from my generator, including my current UPS'. I had an older "no-name" UPS that would not come off of battery when the generator was running. Every UPS is different and you won't find out whether it will switch properly until you get everything hooked up, throw the transfer switch, and run those load tests on the new genny for the first time.

You have to be careful nowadays as everthing has some sort of electronics in it. I was on backup power for 4 hours one afternoon and I worried endlessly about $2K worth of refrigerator blowing a mainboard. Luckily, it held up fine.

 

[Surly73]

Because battery charger implies AC->DC converter, which implies rectifier, and rectifiers produce a lot of high-frequency crap on the power lines (in addition to having a pretty bad power factor).

Though I don't think noise is the problem, I suspect the reason they make these recommendations is solely power factor. A device with low power factor will cause a lot more current to flow in the lines than it is actually consuming, so everything has to be de-rated as a consequence.

But a rectifier is hardly specific to a UPS. Everything except the fridge and washing machine - (check that, they're all electronic now too) will have rectifiers in them. What makes a line interactive UPS charger so special? Furthermore, an online UPS (which they say is OK) is still a rectifier followed by an inverter.

Perhaps the recommendations aren't to be taken too literally. An online UPS may very well have PFC, hiding the poor PF of "typical" computing loads. A line interactive UPS can't PFC for downstream devices so it won't protect the gene from poor PF devices downstream. On the other hand if your PC power supplies etc... have PFC then there shouldn't be a problem. I just fail to see how having net inductive loads downstream of a LI UPS is worse than having them directly connected.

 

[keenan]

But a rectifier is hardly specific to a UPS. Everything except the fridge and washing machine - (check that, they're all electronic now too) will have rectifiers in them. What makes a line interactive UPS charger so special? Furthermore, an online UPS (which they say is OK) is still a rectifier followed by an inverter.

Nothing, except that it (and what it powers) is likely the only high-power DC device in the average household. Most other electronics consume power in the 10s of watts at most, inconsequential. A UPS doing a quick charge, or a double-conversion one powering a lot of equipment could be pulling a kW or more, where PF becomes a significant extra burden.

I wonder if the only reason the warning is there is because people were buying them as backups for their small business 'data centres' and overloading them.

Just trying to reason through their recommendations, not suggesting that they're good ones...

 

[Surly73]

I wonder if the only reason the warning is there is because people were buying them as backups for their small business 'data centres' and overloading them.

Just trying to reason through their recommendations, not suggesting that they're good ones...

You may have hit the nail on the head. This is why I said that perhaps the warnings don't need to be taken so literally.

If you were buying the genset to do a small "data center" on the cheap you had better heed their warning unless you have PFC everywhere. If you're buying a 13kVA unit as a residential backup generator there may be nothing to worry about.

 

[Low Roller]

Just a follow up:

If you're buying a 13kVA unit as a residential backup generator there may be nothing to worry about.

This is exactly what I did about 4 months ago. Had a few power outages already, including Black Friday and Superbowl Sunday.

I've got a couple PC's and most of my entertainment center on consumer-grade UPS's by CyberPower. They play well with my little stand-by generator for the most part. Only thing that seems to bother then is when I'm using the oven. That causes the voltage in my lines to fluctuate a bit and the UPS's kick over to battery if lots of other stuff is drawing powered as well.

Other than that, the generator and the UPS's play nice with each other. When the utility power cuts out, the UPS's keep my PC's and entertainment center powered. After 10 seconds, the generator kicks on and takes over, sending the UPS's back into their normal pass-through state. All in all I really like having the extra security.

 

[Stupid_Newbie]

Voltage regulation is one reason. Two is if the UPS has line filtering or not. And this PFC i laugh when I see this. the PF should be 90%. It is caused to drop from having inductive loads, such as motors. The fans and coils in a PSU is so little, I see no reason its PF is not close to 90%. It can also climb to a leading PF by use of many many caps. But they offset eachother. And it takes a lot more capacitance then inductance, in real world testing.