Utility and Generator Phase Rotation

[kevinware]

I need help understanding something. I am in Kentucky and over the last few days we had a really bad ice storm. I work in a healthcare environment and we have small server rooms in the hospitals that are protected by the Powerware Blade UPS system. This UPS system requires an input clockwise phase rotation (ABC). During the ice storm one of our hospitals lost utility and the UPS went to battery and waited for the generator to come on-line. The generator started and came on-line but the UPS did not transfer from battery to generator, it could not verify the correct phase rotation. Thus causing the batteries in the UPS system to fully discharge and bring the server room down. What I am not understanding is this. The powerware technician is stating the phase rotation from the utility source and the phase rotation from the generator were different. To my knowledge a new ATS was installed at the same time this new ups system was installed. At the start-up for this ups system everything went well and the ABC phase rotation was verified, but this was on utility feed. Looking at a one-line diagram for this section of the electrical system the utility and the generator are obviously feeding the same ATS. So if the phase rotation for both sources were different wouldn’t that cause really bad things to happen on the load side of this ATS? My guess as of now is when the new ATS was installed the two phase rotations, Utility and Generator was not matched at the ATS does that sound right? Wouldn’t it be common practice to match these two phase rotations to a single ATS? Any comments are welcome.

 

[qcroanoke]

Phase rotation should have been verified at start-up of the generator to make sure that utility and generator were in rotation with each other. Obviously it was not.
Also the UPS and Genset should have been tested together at UPS start-up for correct rotation.
Bad things did happen, your servers crashed from power loss. And any motors or compressors that were supplied by the genset ran backwards while on generator power.
That's why the UPS would not go to generator power, phase rotation was wrong.
UPS's alway look for A,B,C rotation.

 

[jimday]

Has there been a power outage since the new ATS was installed? Was the new ATS tested with load? Sounds like a phase conductor was switched when the new switch was installed. Have you checked phase rotation on both sources since this happened? You may not notice issues with loads upon transfer to generator since load side voltage has had time to collapse before connection to the generator, but on re-transfer you might have an issue depending on the speed of the ATS. Does the new ATS have a synching function for re-transfer? I 've had experience with a manual ATS where we had to wait to for motors to spin down before we switched sources since the two sources were not synched.

 

[Jraef]

Common practice to match rotation? Absolutely!
Does it sometimes not happen? Yes, unfortunately.
Would it cause a disaster at the ATS? No, not necessarily. The transition is open; neither source is connected at the same time. BUT, any 3 phase motors connected when the ATS switched to the generator source would run backwards. If there were no 3 phase motors on-line when they tested, it may be that nobody noticed. Had they tested the UPS on the generator load, they would have seen the problem.

I'm also a little puzzled that the UPS would be phase rotation sensitive. Many IT server UPS systems are what are called "double conversion", meaning the incoming line is just rectified (converted) to DC, then re-inverted to AC on the output side. That type usually does not care what the input phasing is because all it is doing is rectifying it to DC. It must be that this one had what is called an Overload Bypass, a feature usually found in UPS systems that are feeding transformers and motors. Because of the high inrush current, they have a bypass switch that automatically shunts line power around the UPS briefly when it is on-line (not on batteries) to avoid voltage drops or having the inverter shut down to protect itself. Since servers rarely have that issue and can suffer from the spikes that this can cause, that feature is rarely specified for server feed UPS systems.

But still, bottom line, whomever was responsible for the ATS and generator installation was sloppy. If there were project specs, I would look into whether or not there was a requirement for testing and even certification of compatibility with the loads. If so, someone screwed up royal. In a hospital, that could have caused loss of life, luckily it was only a server. If there were no project specs, the hospital tried cutting corners by letting vendors specify equipment. This is the kind of thing that can happen when they do that.

 

[qcroanoke]

I'm also a little puzzled that the UPS would be phase rotation sensitive. Many IT server UPS systems are what are called "double conversion", meaning the incoming line is just rectified (converted) to DC, then re-inverted to AC on the output side. That type usually does not care what the input phasing is because all it is doing is rectifying it to DC.

Jraef, I've never seen one that wasn't....

 

[zbang]

I'm also a little puzzled that the UPS would be phase rotation sensitive. Many IT server UPS systems are what are called "double conversion", meaning the incoming line is just rectified (converted) to DC, then re-inverted to AC on the output side. That type usually does not care what the input phasing is because all it is doing is rectifying it to DC.

Some 3-phase input battery charger have one set of controls for all three SCR modules, and definitely expect ABC rotation. I probably still have the prints for one of those around here somewhere (an old Deltec).

 

[quogueelectric]

I would get someone in there to verify whether or not the phase rotation was not the same on the generator side as this is a very large installation mistake imo.
Check that any equipment has not been damaged also. You cannot trust a vendors rep to be telling you what the problem is he could just be covering his butt.
It sounds like this is the problem though .
It is an easy fix at the generator connection points. A shutdown at the ats would be very complicated in a hospital if it was done to osha standards.
If any hvac equipment was reverse phased the compressors will compress in either direction but any 3 phase fans will reverse direction.
Compromising fire detection airflow backwards on duct detectors and possibly disabling cooling for the servers .
Call the installing electrician he will probably run down and fix it imediately for free to try and cover his butt from lawsuits. Take lots of dated pictures now.

 

[qcroanoke]

If any hvac equipment was reverse phased the compressors will compress in either direction.

If they are scroll compressors they will run but they make a lot of racket and let you know they don't like it. Bard 3 phase units have a rotation sensor that will not let the compressor start if rotation is not A,B,C on scroll compressors.

 

[27hillcrest]

I have seen this happen first hand! it was not pretty! There were 3 phase pumps that unscrewed their impeller and crack. I would have maintenance do a complete test on all equipment on the installing EC's dime!!

 

[chaterpilar]

There is another reason why the UPS goes into the "Bypass" mode..and that is the frequency. If the frequency of the standby power generator is not within the UPS set limits, the UPS will go to bypass.

Check the phase rotation and the UPS frequency bypass limits also,.my experience is keeping the generator in isochronous mode and increasing the UPS frequency band helps solving this problem.

cheers.

 

[wasasparky]

Did the ATS transfer power to the UPS?

The UPS will operate with either rotation-but will output ABC no matter what.

It will however inhibit internal bypass if the bypass source is Not ABC. (Could also defeat an external wrap-around bypass...depends on the design...)

 

[gbudke]

LG&E provides CBA, most gen sets are ABC some ups can be set as to what phase rotation they are supose to be at (CBA, BCA, etc..) so the contractor that installed the UPS, may have set CBA not even knowing what the correct phase rotation of the gen set. (sounds dumb but I found an Master that did not know what phase rotation was and just hooked it up BOY and had a phase problem and did not even know it)

 

[ATSman]

Basics

Basics

I need help understanding something. I am in Kentucky and over the last few days we had a really bad ice storm. I work in a healthcare environment and we have small server rooms in the hospitals that are protected by the Powerware Blade UPS system. This UPS system requires an input clockwise phase rotation (ABC). During the ice storm one of our hospitals lost utility and the UPS went to battery and waited for the generator to come on-line. The generator started and came on-line but the UPS did not transfer from battery to generator, it could not verify the correct phase rotation. Thus causing the batteries in the UPS system to fully discharge and bring the server room down. What I am not understanding is this. The powerware technician is stating the phase rotation from the utility source and the phase rotation from the generator were different. To my knowledge a new ATS was installed at the same time this new ups system was installed. At the start-up for this ups system everything went well and the ABC phase rotation was verified, but this was on utility feed. Looking at a one-line diagram for this section of the electrical system the utility and the generator are obviously feeding the same ATS. So if the phase rotation for both sources were different wouldn?t that cause really bad things to happen on the load side of this ATS? My guess as of now is when the new ATS was installed the two phase rotations, Utility and Generator was not matched at the ATS does that sound right? Wouldn?t it be common practice to match these two phase rotations to a single ATS? Any comments are welcome.

When you state " ...The generator started and came on line...." I am assuming that the ATS transferred the load (UPS) to the generator and due to the wrong rotation (a different rotation that the UPS was expecting) the UPS did not turn on (charge batteries) and the load stayed powered from the batteries and died. All of the UPS's I have been involved with have been phase rotation sensitive and will indicate an alarm and not turn on if it is reversed rotation.
Now, if the senario was such that the ATS did not transfer the UPS to the generator then that may have been because most newer ATSs have an in-phase monitor (synchronizing) feature built into the relay logic or microprocessor that will only allow transfer to the generator when the two source voltages are in the synch window. Two sources cannot be synchronized unless they have the same rotation so in a sense, synch check is proving rotation as well as synchronization.
The following procedure is used when I do startups on ATSs:
With no load connected to the ATS and the normal source energized I have the gen vendor manually start the generator. I have the elctrician measure the phase rotation on both sources. If the generator is different then we shut down and flop 2 cables(s) at the most convienient connection in the system. Since you state that the UPS runs fine on utility then you need to match the rotation of the generator to the utility source.

 

[Cold Fusion]

Now, if the senario was such that the ATS did not transfer the UPS to the generator then that may have been because most newer ATSs have an in-phase monitor (synchronizing) feature built into the relay logic or microprocessor that will only allow transfer to the generator when the two source voltages are in the synch window. .

tony -
I'm not clear on which two sources you mean. In the OPs case, the utility is dead, so the two sources re the gen and ....?

cf

 

[ATSman]

tony -
I'm not clear on which two sources you mean. In the OPs case, the utility is dead, so the two sources re the gen and ....?

cf

CF
Thanks for correcting me. There is no synching on a power failure (dead bus- hot bus transfer) I was referring to hot bus-hot bus as in a retransfer to untiliy or if the test switch is used to transfer the ATS. So in the OP's case it sounds like the ATS did transfer the UPS and it was the wrong generator phase rotation at the UPS input that
prevented normal UPS operation.

 

[weressl]

Common practice to match rotation? Absolutely!
Does it sometimes not happen? Yes, unfortunately.
Would it cause a disaster at the ATS? No, not necessarily. The transition is open; neither source is connected at the same time. BUT, any 3 phase motors connected when the ATS switched to the generator source would run backwards. If there were no 3 phase motors on-line when they tested, it may be that nobody noticed. Had they tested the UPS on the generator load, they would have seen the problem.

I'm also a little puzzled that the UPS would be phase rotation sensitive. Many IT server UPS systems are what are called "double conversion", meaning the incoming line is just rectified (converted) to DC, then re-inverted to AC on the output side. That type usually does not care what the input phasing is because all it is doing is rectifying it to DC. It must be that this one had what is called an Overload Bypass, a feature usually found in UPS systems that are feeding transformers and motors. Because of the high inrush current, they have a bypass switch that automatically shunts line power around the UPS briefly when it is on-line (not on batteries) to avoid voltage drops or having the inverter shut down to protect itself. Since servers rarely have that issue and can suffer from the spikes that this can cause, that feature is rarely specified for server feed UPS systems.

But still, bottom line, whomever was responsible for the ATS and generator installation was sloppy. If there were project specs, I would look into whether or not there was a requirement for testing and even certification of compatibility with the loads. If so, someone screwed up royal. In a hospital, that could have caused loss of life, luckily it was only a server. If there were no project specs, the hospital tried cutting corners by letting vendors specify equipment. This is the kind of thing that can happen when they do that.

UPS's ARE phase sensitive and always require ABC rotation. When the UPS is 'islanding' eg. loss both input and bypass voltage it synthetizes the three phase voltage and frequency based on internal program parameters and internal clock. BTW UPS output will always synchronized to the bypass frequency regardless of the normal input.

Some ATS's ARE rotation sensitive and will not transfer to reversed phase roatation source, be it the 'normal' or the 'standby' source.