# 3 phase unbalanced lighting system

I am working at oil pump station plant as electrical technician, I have 4 years experience in the field, and I face a problem which you can help me with. I appreciate any kind of information that will help me.

First of all, I was assigned to repair our plant fence lights. We have huge area of fence lights and at the switchgear room, we have 4 circuit breakers 480 VAC from main substation feeding 9 loads. It's our fence light divided in several loops including some area lighting.

My job is to clear a short circuit on isolated fence lighting loop. After finding shorted connection terminal in lighting pole, I clear this issue and return that loop in service, but when I was checking the voltage Phase to phase I got 460 to 470 VAC, and I got 270 phase to ground which is good, because I don’t have neutral in the circuit. So I started to energize second breaker while I am checking voltage at breaker number 1 I found (unbalanced voltage) phase to ground I got
A-G = 418
B-G = 170
C-G = 200

Then I switch off second breaker, and I got normal voltage, then I start third one, found again unplanned voltage
A-G = 600
B-G = 190
C-G = 70

I started all 4 breaker one by one and I check them. Each one working alone is balanced, but when any 2 of 4 breakers working together, I got unbalanced voltage.

Note : all breaker running together (phase to phase voltage readings are 480 VAC) which is good and our transformer in each lighting pool needed that amount of voltage.

The breakers are not tripped and all lights working fine because it gets 2 phase 480 power. But I still don’t understand if this issue will affect our lighting systems or what?

Also the current reading for each phase is different in all 4 breakers. I am really thankful if you reach this point in my long explanation.

I hope you can help me.

If you have no neutral in the circuit, not even at the station transformer —true triangle circuit—, the hot-to-ground voltages are arbitrary. The whole circuit is floating in respect to neutral/protective ground.

I expect the voltages you measure depend on the light's EMI circuit's Y capacitors which connect hot1/hot2 to protective ground. This also explains why they change with hot1/hot2 being connected to the station transformer or not.

To avoid this floating, one could tie a leg of the triangle to ground. But that's something the design engineer had to decide, not something to be changed later. It requires some thought about the protective equipment used.

• I am really thankful for such information, I am not sure if the main transformer has a nutral or not, because I am allowed only to work at 220- up to 480VAC, but I am sure in all 3 phace breakers we only recive 3 phace not another cables, also we are always having balanced loud to ground at all 3 phace systems in our plant even on motors we check each phace to ground found it balances. Sep 28 '18 at 1:00
• It's the secondary that is interesting here. You told you had been sent to clear a short circuit in a pole. I think the lights were still running despite of that, and you simply got a reading from a ground fault detector? If yes, that's why you have these things – the lights should not fail at single ground faults – the ground fault detector tells you something is wrong instead. Sep 28 '18 at 1:02
• Actually we don’t have a ground falut detector, we know when breaker is trip, so now all 4 breakers are working fine without any problem and lighting is good, because each light receiving 480 volt phace to phace, the transformer and ballast is need 480 , 2 phases. So 3 phace are distributed randomly to the lighting poles. Sep 28 '18 at 1:07
• but I am sure in all 3 phace breakers we only recive 3 phace not another cables, also we are always having balanced loud to ground at all 3 phace systems in our plant even on motors we check each phace to ground found it balances. But even one lighting breaker of the 4 working alone is balanced, but if 2 breakers energize together or all the 4 we get unbalanced Sep 28 '18 at 1:09

In building 3 phase surge suppressors (15 years) the ones for South Korea had to be rated for phase to phase voltage due to all the floating delta feeds they had. If a phase shorted to ground loads would see 480 VAC instead of 277 +/- 20%. It was blowing the fuses in our surge suppressors as well.

Over time we came up with fixes at both ends. We built our South Korean units for 480 VAC phase to phase and phase to ground.

The Koreans could not just ground a phase as this would have required an overhaul of the entire network. There solution, though not perfect, was to install high-impedance capacitor-resistor ground on each phase every few kilometers and after each transformer. They checked all connections for the entire grid to get rid of possible phase to ground shorts, which would be a problem in any country.

After all these changes we no longer heard from them about severe imbalance in the phases or exploding equipment/lights or surge suppressors. It took about 2 years to make all these changes due to the size and poor quality of their grid.

• I am sure that we should not have any connection to ground in our plant, so any unbalanced phace to ground should be corrected, is it the AMPS defense between phase Causing unbalanced voltage? Sep 29 '18 at 0:02
• The current was flowing to ground may have not exceeded the breaker rating. yet enough to offset the centre of a balanced Y measurement normally floating to ground in a Delta transformer Sep 29 '18 at 1:51
• @TonyEErocketscientist. That was the key to hunting down "almost shorts" that would drag down one phase and make other phases rise, sometimes damaging lighting systems and surge suppressors, but not blow a fuse. It took years to track down problems and install load balancers and surge suppressors that could tolerate some imbalance.
– user105652
Sep 29 '18 at 2:14
• yes Sparky I got that.. good stuff+.. I remember 24kV bushings outside my apartment in San Diego zzzz every morning with dew and rapid PD or quiet corona thru the surface morning condensation to earth grounded transformer tank.. Just a nuicance but not much power Sep 29 '18 at 2:16
• Wow, now after 4 days of founding thsi issue breaker still not trpped and lighting are working fine because phace to phace 480 and each light has a 480 transformer and ballast with capastor, almost 400 lighting pole working fine but unbalanced voltage appeared on phace to ground Sep 29 '18 at 4:28

An imbalanced fault load to earth ground on a DELTA 3 phase supply will always result in an offset apparent impedance with respect to ground.

From your description, I might request that a Power Eng. what can be done to improve the lightning arrestor to be self-extinguishing on holding the current, such as a PTC series puck.

The imbalanced corona load to earth will always offset delta voltages when the read as WYE or Y to ground as the DELTA is, in theory, floating except for distributed reactance and leakage to ground from cables and lamps

• So maybe we have a simi ground falut? Or some leaking voltage to ground? But why the brekaers not trip? Sep 29 '18 at 0:00
• If your impedance is high, it is not that significant. Using a plug made a very light Y load (3x 10W bulbs) with earth bond for neutral the shift in balance reflects the amount of leakage to ground perhaps from a tiny corona Sep 29 '18 at 2:02