Hospital Line Isolation Panel and UPS issue

[Bstir]

Here's a good one. We had our yearly Line isolation panel test in our OR's. The certification test involves a set 4.8 mA voltage leak to the panels and it goes into alarm at the LIM when it hits the set point of 4.6mA I believe. This time it did not pass and passed previous years. What is happing is that it hits the 4.8mA and is in alarm but then drops and goes out of alarm, one theory is the UPS that supply's the OR Line isolation panels is conditioning the power and correcting Leakage if this is possible.
We have 2 locations with the same exact UPS feeding the OR's but only have issue at one facility and wasn't a problem before. We think its something in the UPS but was wondering if anyone was familiar with Line Isolation power and had any thoughts of how this can happen?

 

[synchro]

Do you know if it went in and out of alarm when the leakage test was done on L1 and also on L2, or did it have this issue on just one of the lines?
If the strange behavior was only on one line, then maybe the other line had a lower leakage impedance to ground, and this was just enough to push it over the edge and go into alarm during the transient conditions when the test was first applied. But then after settling to a slightly lower level it went out of alarm. Just a guess at what might be happening.
If it goes into and out of alarm when the leakage test is applied from either line to the isolated ground, then something other than unbalanced leakage current from capacitance, etc. is causing the problem.

It would be helpful if the test leakage could be increased beyond the 4.8mA to see if/when there's a current level where the alarm is sustained. If it still goes into and out of alarm with higher test leakage then the LIM might be defective.

 

[Bstir]

Hello synchro,
We did test both L1 and L2 and same problem
Also did both panels (2 in each room) and all OR with same results
We made a 10mA resistor to create leakage because we don’t have same test set up as guy who game out to certify. Same results. LIM went to 9.9, went into alarm, stayed there for about a minute then slowly dropped off until it went out of alarm. Stopped at .3 then climbed again to 9.9 and started whole process again.
I talked to Eaton the UPS manufacturer and they said the process of rectifier to inverter (AC/DC DC/AC) changes sign waves and in that process is
Conditioning the power. He could not verify though if this could cause that problem.

 

[hillbilly1]

Sounds capacitive, maybe a leaking capacitor in the UPS?

 

[synchro]

Hello synchro,
We did test both L1 and L2 and same problem
Also did both panels (2 in each room) and all OR with same results
We made a 10mA resistor to create leakage because we don’t have same test set up as guy who game out to certify. Same results. LIM went to 9.9, went into alarm, stayed there for about a minute then slowly dropped off until it went out of alarm. Stopped at .3 then climbed again to 9.9 and started whole process again.
I talked to Eaton the UPS manufacturer and they said the process of rectifier to inverter (AC/DC DC/AC) changes sign waves and in that process is
Conditioning the power. He could not verify though if this could cause that problem.

While doing these tests, it would be helpful to measure the L1-G and L2-G voltages simultaneously with two meters if you haven't already done so to check if they're shifting when the current readings change.

Also, can you unplug at least some of the loads when doing these tests to check whether that changes things? Your measurements behave somewhat like a relaxation oscillator so maybe some electronic components involving capacitors and semiconductors in equipment could be contributing to these changing leakage readings.

Are there any circuits that are not on the UPS output that might share raceways with UPS powered circuits? If the UPS output is in close but not fixed synchronism with these other waveforms then the amount of capacitive leakage current between the circuits would change at the difference of the two frequencies. Not the most likely cause but at least something to consider.

 

[mbrooke]

Where is the UPS? If its feeding the primary of the ISO panels I can't picture anything on the secondary having an effect.

 

[Bstir]

Where is the UPS? If its feeding the primary of the ISO panels I can't picture anything on the secondary having an effect.

Just to be more specific 480v 3 phase disconnect to UPS
UPS to 3 Phase panel with 100 amp main
277v to Line isolation panels

 

[Bstir]

While doing these tests, it would be helpful to measure the L1-G and L2-G voltages simultaneously with two meters if you haven't already done so to check if they're shifting when the current readings change.

Also, can you unplug at least some of the loads when doing these tests to check whether that changes things? Your measurements behave somewhat like a relaxation oscillator so maybe some electronic components involving capacitors and semiconductors in equipment could be contributing to these changing leakage readings.

Are there any circuits that are not on the UPS output that might share raceways with UPS powered circuits? If the UPS output is in close but not fixed synchronism with these other waveforms then the amount of capacitive leakage current between the circuits would change at the difference of the two frequencies. Not the most likely cause but at least something to consider.

Sounds capacitive, maybe a leaking capacitor in the UPS?

We put the UPS in bypass and issue went away. Looking more like its the UPS. We have a Tech coming out tomorrow morning early and will mention capacitor. Thanks

 

[texie]

We put the UPS in bypass and issue went away. Looking more like its the UPS. We have a Tech coming out tomorrow morning early and will mention capacitor. Thanks

I'm having trouble understanding how an issue with the UPS supplying the isolated line panel could cause leakage on the output of the panel.

 

[texie]

Also, what is the insulation type of the conductors on the load side of the system?

 

[synchro]

I'm having trouble understanding how an issue with the UPS supplying the isolated line panel could cause leakage on the output of the panel.

I'm in agreement. I don't think the UPS could cause actual leakage at the output side of the isolation transformer.

Perhaps the UPS is faulty and has excess harmonics or other distortion that is causing false readings on the Line Isolation Monitor (LIM). Just a guess. The LIM makes measurements which involve the application of a low-level, typically higher frequency signal to calculate the maximum leakage that would occur if there was a ground fault on either line. So when the LIM displays a given current value, it does not necessarily mean that this amount of leakage is actually present.

 

[Bstir]

I'm having trouble understanding how an issue with the UPS supplying the isolated line panel could cause leakage on the output of the panel.

It’s not causing leakage. It’s seams like the UPS is conditioning the leakage that we are putting on the Line isolation panels for the test purpose.