Isolation Transformer for "Dirty" Power

Status
Not open for further replies.
E

elecshop

Member
Location
FL, USA
Occupation
Electrical Engineer
I am working on a hospital project which has "dirty" power issue i.e. several small brownouts that affect the equipment and HVAC. The electrician installed a meter between the main switchgear and the two ATS's serving the equipment branches. Voltage sags were noticed continuously on the B leg without corresponding current swells, and the measurements indicate that the B leg is 1.7 degrees out of phase. The thought is that this could be an issue from the substation side, probably from the capacitor banks or switching?
Would isolation transformer be able to resolve the issue by avoiding those sags? I know they improve power quality, but I cannot see how those sags coming from the utility would not be reflected on the load side.
Any other recommendations to resolve such an issue?
 
elecshop said:
I am working on a hospital project which has "dirty" power issue i.e. several small brownouts that affect the equipment and HVAC. The electrician installed a meter between the main switchgear and the two ATS's serving the equipment branches. Voltage sags were noticed continuously on the B leg without corresponding current swells, and the measurements indicate that the B leg is 1.7 degrees out of phase. The thought is that this could be an issue from the substation side, probably from the capacitor banks or switching?
Would isolation transformer be able to resolve the issue by avoiding those sags? I know they improve power quality, but I cannot see how those sags coming from the utility would not be reflected on the load side.
Any other recommendations to resolve such an issue?
Click to expand...
IMHO.... I would call utility and work with them to solve this issue. A transformer is a fixed ratio magnetically coupled circuit integrator. Having poor voltage on the primary side, equates to poor voltage issues on the secondary side. :) That is my $.02. :)
 
You would need a constant voltage transformer (CVT), not a simple noise isolation transformer.

The last CVT I was involved with was probably 20 years ago, so I don't know if they are still made.
 
jim dungar said:
You would need a constant voltage transformer (CVT), not a simple noise isolation transformer.

The last CVT I was involved with was probably 20 years ago, so I don't know if they are still made.
Click to expand...
They are still made, but I don't think they were ever very large. A quick look at the one brand I recall from old installations shows the maximum currently available size at 7.5 kVA.
 
don_resqcapt19 said:
They are still made, but I don't think they were ever very large. A quick look at the one brand I recall from old installations shows the maximum currently available size at 7.5 kVA.
Click to expand...

Yes. And CVT's (aka ferroresonant transformers) are not used as much as they were before because there are alternatives such as double-conversion UPS's that are much more efficient. CVT's intentionally operate their magnetic core in saturation so that the peak magnetic flux developed is relatively constant during variations of the input RMS voltage (at least over a limited range). As a result, the output voltage on the secondary winding remains relatively constant (by Faraday's law of induction). To reduce the amount of harmonics produced by saturation there is a resonant tank circuit using auxilliary windings and a capacitor. But operating in saturation inherently results in low efficiencies.

Certainly the power company should be contacted first to see if they can correct the problem on their end.
 
What is a "small brownout" (aka sag) in this case? A few volts should be a concern to any reasonable equipment. OTOH if thhat phase is dropping 10-15 volts, that's a problem.

B-leg.... I'd expect the current phase to lag voltage (power factor). Or is it the voltage is lagging where it should be? If that's the case first I'd distrust my measurement, then get the PoCo involved to see if they get the same data you do.
 
zbang said:
What is a "small brownout" (aka sag) in this case? A few volts should be a concern to any reasonable equipment. OTOH if thhat phase is dropping 10-15 volts, that's a problem.

B-leg.... I'd expect the current phase to lag voltage (power factor). Or is it the voltage is lagging where it should be? If that's the case first I'd distrust my measurement, then get the PoCo involved to see if they get the same data you do.
Click to expand...
The B leg at the main switchgear had 17 dips, minimum value reached around 0V, which lasted around 50 minutes.
And by phase lag I meant: the B leg phasor is off by 1.7 degrees (voltage, not current). So it is at -238.3 degrees instead of -240 degrees.
The voltage sags on the B leg do not correspond to current peaks in the hospital loads.
 
Is the service point from your utility at low voltage (e.g., 480V, 208V), or is it at medium voltage and you own the step down transformer?

elecshop said:
The B leg at the main switchgear had 17 dips, minimum value reached around 0V, which lasted around 50 minutes.
And by phase lag I meant: the B leg phasor is off by 1.7 degrees (voltage, not current). So it is at -238.3 degrees instead of -240 degrees.
Click to expand...

At this point I would forget about the B leg being off by 1.7 degrees when it's dropping all the way to 0V for as much as 50 minutes when the other 2 legs are OK. Or are you saying that each of these 17 dips was of short duration, but they all occurred during a particular 50 minute time interval?

If the MV to LV transformer is delta-wye and one of the input phases was open or otherwise dead, then the output L-N voltages would be single-phase (0 or 180 degrees relative to each other). And so in this case the problem you mentioned would not be coming from the MV side.
 
Last edited:
There are large systems, called Dynamic Sag Corrector (DySC), that will correct sags of short duration. The original manufacturer, SST, sold a version called a "MegaDySC" that goes up to 2400A at 480V. They are now part of Rockwell.
 
synchro said:
Is the service point from your utility at low voltage (e.g., 480V, 208V), or is it at medium voltage and you own the step down transformer?



At this point I would forget about the B leg being off by 1.7 degrees when it's dropping all the way to 0V for as much as 50 minutes when the other 2 legs are OK. Or are you saying that each of these 17 dips was of short duration, but they all occurred during a particular 50 minute time interval?

If the MV to LV transformer is delta-wye and one of the input phases was open or otherwise dead, then the output L-N voltages would be single-phase (0 or 180 degrees relative to each other). And so in this case the problem you mentioned would not be coming from the MV side.
Click to expand...
The 17 dips are shorter durations but the particular 50min dip among them is the longest one.
I was thinking maybe there could be a problem with switching/contacting occurrence in the utility substation. I forgot to mention that those dips occur almost at the same time each day i.e., at around 2:20pm. Could it be that a non-zero voltage at the MV side during this event is causing this phase lag? Maybe capacitor banks become unsymmetrical/unbalanced when seen from the LV side? I'm not an expert in power on the utility side, but we can be almost sure that it is not an issue from the hospital.
 
Another possibility is that there's an open phase on the input to a wye-wye transfomer with a three-leg core, and where center connection in both the primary and second wye's is grounded (Yg-Yg). This configuration will regenerate the normal voltage on the open phase, as is mentioned on pg. 2 of the document at the link below (pg. 3 in the pdf). The figures in the document show that the missing voltage is fully regenerated at no load or a light load, and its voltage drops about 15% relative to the other phases when the transformer is 50% loaded. I suspect that the voltage may drop closer to zero if it's heavily loaded. It would not be surprising that there could be a slight phase shift in a the regenerated voltage on the B-phase, particularly if it's in the lagging direction.

https://selinc.com/api/download/124320
 
synchro said:
Yes. And CVT's (aka ferroresonant transformers) are not used as much as they were before because there are alternatives such as double-conversion UPS's that are much more efficient.
Click to expand...

The largest FR transformers I have ever seen were in 3-phase UPS's made by Lorain Products Corp. It was about 60 KVA. Their reason for that approach was they didn't need protection circuity for their SCR's.

FR xfmrs are short-circuit proof; short the output and the input draws 100% rated current, and stays there. But you gotta pay the piper. At no-load, the FR xfmr draws ~35% of full load. That's a lot of heat.

The QC test was to take say a 10KVA unit, put a much larger 3-phase motor on it, and push the button. The UPS goes to 100% and stays there. The motor doesn't spin. But if you look closely, you can see it c...r...e...e...p; and over time, it's creeping more and more. After tens of minutes, it's finally spinning. And eventually, it gets up to speed. The UPS load drops down. It passed.
 
Last edited:
elecshop said:
I am working on a hospital project which has "dirty" power issue i.e. several small brownouts that affect the equipment and HVAC. The electrician installed a meter between the main switchgear and the two ATS's serving the equipment branches. Voltage sags were noticed continuously on the B leg without corresponding current swells, and the measurements indicate that the B leg is 1.7 degrees out of phase. The thought is that this could be an issue from the substation side, probably from the capacitor banks or switching?
Would isolation transformer be able to resolve the issue by avoiding those sags? I know they improve power quality, but I cannot see how those sags coming from the utility would not be reflected on the load side.
Any other recommendations to resolve such an issue?
Click to expand...
By “voltage sags” do you mean voltage drops? Installing capacitor banks whether on line feeder or bus side or prior to loads on load side of circuit each have their advantage and can increase line voltage if too low while lowering ampere draw on conductors.

Capacitors on line side require their own fuses disconnect with conductors to them at 130% minimum of the capacitors CKVAR Rating

Capacitors installed directly before load can be protected by feeder breaker - according to my book (but may be out of date in regard to current NEC?)
 
When I worked as an apprentice years ago I heard someone referring to a residential solar power system as the power being “dirty.” Does dirty mean signal/ voltage / ampere waveform abnormalities?
 
elecshop said:
The 17 dips are shorter durations but the particular 50min dip among them is the longest one.
I was thinking maybe there could be a problem with switching/contacting occurrence in the utility substation. I forgot to mention that those dips occur almost at the same time each day i.e., at around 2:20pm. Could it be that a non-zero voltage at the MV side during this event is causing this phase lag? Maybe capacitor banks become unsymmetrical/unbalanced when seen from the LV side? I'm not an expert in power on the utility side, but we can be almost sure that it is not an issue from the hospital.
Click to expand...
I'm wondering at this point if you have some kind of connection issue. Dropping down to 0V for that long is insane. This absolutely warrants a call to utility. The 1.7 degree phase shift isn't that big of a deal at the moment. You really need to find out why it's dropping out. This has a poor connection written all over it. Time to take some thermal images of stuff and investigate hot spots to clear stuff on your end, but I have a hunch this might be a utility problem.

You did say it happens around the same time each day, so that's not nothing, but it's not something either.
 
Jpflex said:
When I worked as an apprentice years ago I heard someone referring to a residential solar power system as the power being “dirty.” Does dirty mean signal/ voltage / ampere waveform abnormalities?
Click to expand...
Dirty means pretty much that. It means that there is some kind of issue with the overall signal causing the rest of the system to suffer. It can be voltage / line imbalances, it can be distortions from non-linear loads, any number of things.
 
Status
Not open for further replies.
Top