Generator Cannot Charge UPS

[gepman]

We have a 60kW 120/208Y 3 phase generator connected to a UPS through a 4 pole ATS (see attached simplified one-line). The normal power to the UPS is through a 75kVA delta/wye transformer. The UPS is an Eaton (Cutler-Hammer) Powerware Blade 48kw unit with batteries and inverter. The UPS works fine on utility power and charges the batteries and supplies the critical load. When utility power is interrupted, the UPS powers the load from the batteries, the generator starts, and after about 4 minutes the UPS rejects the standby power source (generator) due to low frequency, which it says is between 48 and 55 Hz.

Measurements on the generator output show that both the voltage (206 to 217) and frequency (59.8 to 60.1 Hz after initial connection of load) are fine.

This is a new installation for the UPS. The generator worked fine with an APC UPS however the UPS was changed in order to increase its capacity.

Is it possible that there is some incompatibility between the generator and the UPS? The UPS manufacturer puts the blame on the generator.

Any other suggestions will be appreciated.

 

[edison123]

From what you have posted, the generator looks to be fine. May be instrumentation error on the ups side ?

 

[itsmoked]

Possibly the rectification pulses are interacting with the generator and then being misinterpreted by the UPS.

A lot of frequency circuits have issues with this.

I have a DMM that has a frequency display. Often I must select other unexpected voltages before it outputs the correct frequency. I have to know what frequency I am looking for then 'hunt for it'.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

I hope that you are not trusting the frequency meter on the Genset. The old vibrating reed meters were good, but the newer analogue frequency meters are bought from the lowest bidder.
I consider an analogue frequency meter on a genset to be a rough indication. If I have a frequency issue I get a good frequency meter and check it on the mains before using it.
Your generator should normally run at 61.8Hz at no load and droop to 60Hz at full load.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[DanEE]

Have you called Powerware tech support on the specific unit you have regarding settings for generator operation?

I don't know the specific parameters on the blade systems vs the Powerware 9170s, FERRUPS and Unity Systems I'm familiar with, but there can be other parameters other than frequency window and voltage range that must be changed to run reliably on generator power.

For example on some Powerware systems there are settings for slew rate, phase lock count, freq G1 count, Line G1 count, line delta that must be changed for generator operation. These in essence tune how the UPS samples line power purity; even if the frequency is in range, without these changes they frequently won't transfer back to line power.

Other Powerware systems have an interface that needs a contact closure signal from the generator transfer switch to let it know when it is on generator power. When the UPS sees this signal, the unit will change the above related parameters automatically.

 

[sibeen]

gepman, you state that the generator is a 60 kW model, what is this in kVA. Is it 75 kVA?

I haven't worked with the powerblade style of UPS so don't know what sort of circuit topology it uses, but from the published data I could find on it and a reported effeciency of >97% I can only assume it is some form of off-line style, also as it is stated that the input inrush current is load dependent.

The input power factor is stated as <5% in the brochure. Again I would assume that this is best case (a marketing department wouldn't lie, would they?) and would be with the batteries fully charged and an ideal load. The input power factor is stated as >99% (again load dependent). Given this, during battery recharge you should be drawing approximately 55kW if the UPS was fully loaded.

There used to be a general 'rule of thumb' when sizing generators to supply a UPS load. Depending upon the input rectifier configuration of the UPS (6 or 12 pulse) the generator was sized between 1.5 and 1.8 times the size of the UPS. With a 60 kW generator you are only slightly above the maximum draw of the UPS.

Have a look at the input current distortion when the UPS is charging the batteries and being supplied by the generator. A reasonable high THD with a relatively high load would probably indicate that the generator is under-sized for the application.

 

[itsmoked]

Hey wait a minute.. If the UPS is 48kW doesn't this imply the load is around 40+kW? You start a generator and it has to carry that same load plus the added load of recharging the batteries. This certainly fits with the 1.5~1.8x sibeen has stated.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[gepman]

First I would like to say that I should have stated "We have a client who has a ...", not "We have a ...". I had just gotten a call to go out there the next day and I was in a hurry to post this to give time for answers before the next morning and also have time to call Eaton Tech Support before they closed on the East Coast. The information given in the post was from the client via telephone and email, which as most people know, may or may not be accurate.

Second I would like to thank everyone for their responses.

Third I would like to let everyone know the resolution.

To answer waross, the frequency measurements were made with a Fluke scopemeter.

To answer DanEE I did call Tech Support immediately after this post. I wanted to get the forum's opinion in addition to Tech Support. My experience with Tech Support is highly variable, it depends both on the company and who you get in contact with at the company. This UPS is so efficient because it basically bypasses the rectifier and inverter when the power is clean. When it senses "dirty" power it goes into "double conversion" mode, which uses the rectifier and inverter to clean up the power. Battery charging load is minimal with fully charged batteries. Tech support was very helpful and suggested that the problem could be due to power grounding or a poor neutral. I asked about phase reversal and they said that phase reversal would cause the unit not to be able to measure the power properties (including frequency) properly due to the software used to calculate the them. This is apparently due to the sudden reversal of phases.

To answer sibeen, I have heard of that general rule before. For this system it would depend on the charging rate of the batteries, any starting loads, and maximum demand. When I got to the site they had already connected a rental 220kW generator and had the same problem so generator capacity probably wasn't the issue. The permanent generator is 60kW and 75kVA.

It turns out as you might have surmised that the generator phase sequence was opposite the utility phase sequence. Both the permanent and rental generator had been connected ABC while the "normal" side was connected ACB. This was determined using a phase sequence meter that I purchased for the occasion since Tech Support had told me that phase reversal could cause the symptoms that were occuring.

My only suggestion to Eaton is that the Installation and Operation manual should mention that phase sequence MUST be consistent between normal and standby sources (although I will agree with anyone who says that it should be checked during installation). You would also think that the UPS unit would alarm "Phase Reversal" instead of just not being able to measure the power correctly. The I&OM is very basic and does not give any circuit diagrams so I cannot tell how they are measuring the power and power quality.

This got me to thinking about standard utility phase sequence so I read the post "How to Recognize Phase A" and I made a comment there.

 

[edison123]

If the phase sequence was reversed, didn't someone notice the other non-ups fed motors changing the rotation direction ? Or does this genny feed only thru' ups ?

 

[waross]

Thanks for the feedback, gepman. It does make trouble shooting challenging when the error message is one of the errors.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[gepman]

edison123, the UPS is for a small data room only but there were two three phase loads, both air conditioners for the data room.

The client's comment was that "this explains why the air conditioners never came on when we were on standby power". Apparently they never had a long power outage, at least long enough for the rooms to heat up.

 

[waross]

He was lucky to have reverse power protection on his A/Cs. When I have to change connections around a three phase A/C I find a small three phase motor to hook in parallel. The power doesn't go back on to the A/C until the test motor runs the right way.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[edison123]

And why would an UPS care what is the incoming phase sequence ? After all, it is a generic product supposed to be used anywhere, whose incoming may be ABC or ACB. The UPS supplier cannot expect the client to change their phase sequence to please the UPS.

I sure don't understand the logic. May be some UPS expert will clarify that ?

 

[ScottyUK]

Hi Edison,

The UPS output is usually synchronised to the bypass source. Unless the UPS can reverse the output phase sequence automatically it will require the correct phase rotation at the input in order to synchronise. The rectifier gate drives will likely be designed for a specific phase sequence too.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[edison123]

Scotty

"Unless the UPS can reverse the output phase sequence automatically it will require the correct phase rotation at the input in order to synchronise"

Doesn't the ups invert from DC ? If so, why would its input phase sequence matter ?

I agree that ups output has to match the original utility supply sequence but I still don't get the input sequence requirement.

 

[FreddyNurk]

Perhaps I'm speculating here, but a few things may be apparent in all of this, at least in terms of operation of a UPS. Clearly there are multiple types (topologies, perhaps) of UPS, though some of them only appear at the smaller end of the market, such as a standby only UPS. If in the case that was presented, that the generator was the opposite phase sequence to the grid, and the UPS decided that it wasn't going to work anymore, and run in bypass mode, clearly there'd be an issue, as ScottyUK had stated.

It shouldn't, however, have only one defined phase sequence for the input. Interesting that the UPS's refusal to transfer (along with incorrect errors) is what gave rise to finding the actual problem.

 

[DanEE]

I can't imagine any company designing a UPS to handle a misswired phase reversal (other than self protection), given the dramatic down stream impact to any rotating 3 phase loads.. I'm sure their attorneys would hang the designers.

Even on single UPS systems (running on 208vm or 2 phase as many call it), the installation process has detailed checks to make sure the phases aren't turned over.

The system did what it should do, i.e. protect itself from damage and it's amazing that the static bypass was not blown up.

btw, Powerware has a written policy that only their employees can install their 3 phase systems, otherwise the warrany is null and void. The Blade system due to it's unique packaging design to make the 3 phase much more safe to install, is the first exception to this policy.

 

[ScottyUK]

"Doesn't the ups invert from DC ? If so, why would its input phase sequence matter ?"

Yes, in a double conversion type it does. Most designs I am familiar with are designed to produce an A-B-C phase rotation and have the ability to alter frequency and phase within certain limits to match up with the bypass supply. Outside those limits the UPS rejects the bypass supply. They simply aren't designed to change the rotation because there is no need to do so when installed correctly.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[sibeen]

Frankly I'm amazed that the powerblade doesn't have an alarm status indicating that the incoming phase rotation is incorrect. Doesn't give me a warm and fuzzy feeling about the rest of the box.

In large UPS systems which use a controlled Thyristor bridge as the rectifier, the phase sequence is definitely required to be A-B-C (U-V-W, R-Y-B) as the thyristors need to fire in a defined sequence.

 

[waross]

I am still amazed that the unit annunciates a phase reversal as under frequency. Does the development team participate in the special olympics?

Bill
--------------------
"Why not the best?"
Jimmy Carter