Power quality problems.

[Microwatt]

Here is the story. Health care setting. The campus has two generation systems. One system is (4) 2MW gens that parallel up to provide power for the hospital, clinic, and some support buildings. It is primarily used for peak shaving. The other system is (2) 1.2MW gens that are used as the emergency system. Here is the problem. There is a mobile ultrasound machine that will not work in one wing of the hospital when the 4 2MW gens are on-line. The power supply relay just chatters and the unit won't power up. If we transfer that wing to the emergency generators the unit works fine. Also, everything works fine when on utility power. Any suggestions?

 

[bob]

Here is the problem. There is a mobile ultrasound machine that will not work in one wing of the hospital when the 4 2MW gens are on-line. The power supply relay just chatters and the unit won't power up.If we transfer that wing to the emergency generators the unit works fine. Also, everything works fine when on utility power. Any suggestions?

These 2 comments sound contradictory to me. You say the equipment will not work while the gens are on line but the equipment works when the wing is transferred to the gens. Have you checked the voltage at the relay under normal conditions and abnormal conditions? Relays chatter when the voltage is low.

 

[SG-1]

Something is causing this device to chatter. Is it internal to the equipment, say a 27/47 or some other device that is set so that it cannot decide if the incoming power is good or not. Such a device could be de-energizing the coil then allowing it to pull in again. A schematic diagram should show what could interrupt the coil voltage. The coil may have to be disconnected to measure the voltage if it is not stable during chattering. The issue could be external such as miswiring or a loose connection ahead of the unit. More information about the problem would be useful, like did it ever work on that set of Gens ?

Just some stabs in the dark.

 

[gar]

091007-0657 EST

Try putting a Sola constant voltage transformer between the AC supply and the ultrasound unit. Also look at the AC waveform from the generators --- measure voltage, use a scope to see the waveform shape, and do a harmonic analysis.

If it fails when powered from the Sola, then the ultrasound unit has a problem.

.

 

[Microwatt]

091007-0657 EST

Try putting a Sola constant voltage transformer between the AC supply and the ultrasound unit. Also look at the AC waveform from the generators --- measure voltage, use a scope to see the waveform shape, and do a harmonic analysis.

If it fails when powered from the Sola, then the ultrasound unit has a problem.

.

I'll consider the transformer idea. One thing I have tried was putting a true on-line ups ahead of the ultrasound machine and still nothing. The ups accepts the power fine, but I have the same result with machine. I have also taken data recordings with a power quality meter. The sine wave is quite choppy both on utility and generators and harmonics increase a little. The engineers have all that stuff now but they are telling me it is because we are probably outside the IEC standards of what the equuipment was designed to.

What's puzzeling to me is why only one area of the hospital? The equipment works fine everywhere else in the facility no matter what system is providing power.

Also, its not one machine. It is every version of a certain manufacturer's equipment. Other manufacturer's stuff works fine.

 

[Microwatt]

These 2 comments sound contradictory to me. You say the equipment will not work while the gens are on line but the equipment works when the wing is transferred to the gens. Have you checked the voltage at the relay under normal conditions and abnormal conditions? Relays chatter when the voltage is low.

There are two generation systems. The big one has two large 23.9kv ATS's that each power half of an open loop and works to power the entire campus. The little one has 19 480v ATS's that power distribution gear inside the hospital and acts as our emergency system. Basically one generation system over the top of another.

 

[winnie]

Bob, if I understand the OP correctly, they have 2 generator systems, but only have problems when 1 of them is running.

Microwatt, have you tried running the equipment using your 'true on-line ups', but with the UPS unplugged from the wall, so that you are solely on battery power, or tried the equipment with the ups in the wing where the equipment usually works?

The UPS thing might be a red herring- it is possible that the UPS is not sufficient to power the load.

-Jon

 

[Microwatt]

More information about the problem would be useful, like did it ever work on that set of Gens ?

Just some stabs in the dark.

Yes, they did once work. Last summer is when it first became an issue. We determined that 10 new VFD's that were put in earlier that year did not have line reactors installed on them. We thought that may be the cause so we installed line reactors on them. The problem actually went away.

This summer, the first day we were call to peak shave the problem arose again.

 

[Microwatt]

have you tried running the equipment using your 'true on-line ups', but with the UPS unplugged from the wall, so that you are solely on battery power, or tried the equipment with the ups in the wing where the equipment usually works?

The UPS thing might be a red herring- it is possible that the UPS is not sufficient to power the load.

-Jon

Yes, the equipment works soley on battery power.

 

[winnie]

Have you tried measuring the voltage (AC, harmonic AC, and DC) between neutral and local bonded metal? I'm wondering if the equipment is sensing a voltage difference between equipment neutral and 'ground' and somehow kicking in protective devices.

-Jon

 

[gar]

091007-0840 EST

Microwatt:

What is a "true on-line ups"?

I think it is clear that you have noise generating equipment that is affecting certain areas. Different distribution transformers may be providing some filtering.

The Sola constant voltage transformer is a voltage regulator, a low pass filter, and an isolation transformer all in one device. It is a ferro-resonant transformer. Sola has changed the name to power conditioner and this may include additional filtering. Also under the power conditioner name there may be non-ferro-resonant devices. I am not sure whether the original Sola's had a Faraday shield, or even if today's units do, but to some extent the original design without an explicit shield may have had low capacitive coupling from primary to secondary. Connecting one side of the Sola secondary to the incoming EGC may produce no problem if there is not a lot of noise on the EGC.

See "AC voltage stabilizers" at http://en.wikipedia.org/wiki/Voltage_regulator

.

 

[techntrek]

I think winnie is on to something with a facility ground issue.

gar - "online" UPSs convert AC to DC and then back to AC with an always-on inverter so (as the OP thought correctly), it should provide a good power source no matter what genset was operating. The equipment is effectively on its own generator - the inverter in the UPS. So that eliminates power quality as a cause.

Since the equipment still wouldn't run on the UPS while connected to the genset, but would when the UPS was unplugged from the wall, the only common denominator is the ground, IMO. The UPS is still utilizing the facility's ground while plugged in, but (obviously) isn't when its unplugged.

 

[gar]

091007-1152 EST

Thanks techntrek. I assumed that might be the case.

In a hospital situation what can you do and conform to code to reduce a noise problem on an EGC?

A test to see how much of a problem the facility EGC is would be to operate the UPS on backup and connect only the EGC pin of the UPS to the EGC at a wall outlet. Presumably this will reintroduce the problem relative to UPS only and no EGC connection.

Assuming the problem is noise on the EGC, then what is the routing of this EGC thru the building relative to possible noise sources.

It may be possible to investigate the noise source by means of an AC DVM resolving millivolts with one long test lead back to the neutral-ground bus back at the main panel.

Doesn't answer the question of why different power sources have different effects unless the problem source causes current flow on the EGC in question.

.

 

[winnie]

Just to throw another wrench into the discussion, the concept of a 'true online UPS' often gets muddied by the marketing guys, and it is possible that the UPS is not really operating its inverter if the mains is connected.

I believe I recall one system that had some sort of transformer/capacitor system that had a _tiny_ bit of energy storage, enough to maintain the output for a large fraction of a cycle, combined with an inverter that would kick on in a small fraction of a cycle; net result was that the marketing literature made it sound like the UPS was always directly generating the output, rather than bypassing the mains directly to the load...truthful, but sneaky.

-Jon

 

[Microwatt]

091007-1152 EST

Thanks techntrek. I assumed that might be the case.

In a hospital situation what can you do and conform to code to reduce a noise problem on an EGC?

A test to see how much of a problem the facility EGC is would be to operate the UPS on backup and connect only the EGC pin of the UPS to the EGC at a wall outlet. Presumably this will reintroduce the problem relative to UPS only and no EGC connection.

Assuming the problem is noise on the EGC, then what is the routing of this EGC thru the building relative to possible noise sources.

It may be possible to investigate the noise source by means of an AC DVM resolving millivolts with one long test lead back to the neutral-ground bus back at the main panel.

Doesn't answer the question of why different power sources have different effects unless the problem source causes current flow on the EGC in question.

.

I think I will give this a try. Thanks to all of you for your input. It gives me a few directions to head in.

 

[catchtwentytwo]

Microwatt:

What size and quantity of UPS are you dealing with? Sorry if I've missed that in your other posts.

 

[don_resqcapt19]

This doesn't make a lot of sense. If you are operating in parallel with the utility, I would think that the power quality would not be much different than with the utility alone unless the utility is very weak.

 

[techntrek]

Microwatt:

What size and quantity of UPS are you dealing with? Sorry if I've missed that in your other posts.

Brand and model would be useful, to ensure its an online model, but I trust the OP already made sure of that.

 

[Microwatt]

Brand and model would be useful, to ensure its an online model, but I trust the OP already made sure of that.

I can't recall off the top of my head what the brand and model was. But I did verify it was.

 

[catchtwentytwo]

I can't recall off the top of my head what the brand and model was. But I did verify it was.

Module size and system architecture are important to know. For example, if this is a multi-module parallel redundant system, that's a whole different situation than dealing with independent point-of-use or "small" UPS systems (AKA: mini-UPS).