Brown outs

[Gman1969]

Scenario: 3 phase 120/208 from pole which has transformers provided by utility company. Brown outs occur after heavy rain, building is in flood zone and at sea level. building is 55 years old. all underground conduits are rigid with no grounds and sharing neutrals. provided new 200 amp overhead service and 2 new load centers. provided 2 new ground rods and eventually added a 3rd.
building works fine on gen set which was brought in roughly 3 months ago as every time we go back on utility power brown outs occur again and in fact knock down traffic lights and a drug store across the road a block away. the traffic lights are on single phase portion of the transformers but the drug store has its own transformer from what I've noticed. during the brown out I noticed A phase was reading 60v B and C phases were around 170 volts. then A and C phases swapped and then C was 60 v. Utility company was called out to site and at the transformers they read proper voltages however, 15 ft down the line they were reading what I was the 60/170/170. Also was reading 15 amps on my ground. So I went ahead and re-fed 90 percent of the overhead lights and some other problematic outlets .
It definitely helped clear up my grounds. Also in doing so gave some of the groupings their own neutrals but still no ground wire. the brown outs dont happen all the time they are usually at higher load draws(low tide seems to happening as well) and then the gen set gets hooked up again so the facility can function (government building so I can't leave on utility for too long because they have to stay open.) We did ohm reading on ground rods my new ones were at 1 ohm utility is at 15 ohms, all within range of the 25 ohm rule.

All along the lights would drop for a split second when a/c unit kicks on. Yesterday I went up and took amp readings on the phases (3 phase 120/208 no neutral) I read ... A 45 amps/B38 Amps/C37 amps.
My question is this after all this work that's been done is it possible for the a/c unit to have been causing the issue all along. I find it hard to believe that an unbalanced load could cause brown out and push back on utility power. Oh and the reason I brought up the sea level thought was I kept reading at different times of the day (amps on ground, load amps) and it seemed to coincide with low tides.

 

[mikeames]

There is an issue with the neutral. You answered it in your post. " Utility company was called out to site and at the transformers they read proper voltages however, 15 ft down the line they were reading what I was the 60/170/170 "

Problem solved. Did they just measure that and walk away?
Or you have a phase faulted.

 

[gar]

210608-1044 EST

Gman1969:

I believe your answer is in your post. I also believe you do not understand electrical circuit theory.

You say that power company reads correct voltages at their transformers, and that these values are about 208 V line to line for all combinations, and 120 V from each line to neutral, and that these voltages do not change when you see voltage errors at various customers.

Note: 120^2 - (208/2)^2 = 14400 - 10816 = 3584, and the sq-root of this is 59.86 which is 1/2 of 120 as expected. For exactly 120 V line to neutral a more accurate value for the 208 V voltage is 103.928 X 2 = 207.846 . But you can expect the power company readings were not real precision values, and line voltage readings jump around a little.

But here is the important point. The power company readings were about constant and near the values indicated and expected. However, at your customer when the readings at the transformers were good the customer readings were very bad.

Now lets look at those readings. For example 60, 170, 170. Apparently line to neutral. I don't believe you read line to line voltages. Let us assume that line to line voltages were all about the same at about 208, a reasonable assumption based on your information. The vector squared from your neutral at the load to the midpoint of 208 for the two 170 V vectors is 170^2 -104^2 = 18084. The sq-root of this is 134.5 . Add 60 to this and the result is 196.5 . Using 208 the calculated value for this vector is 180.1 V. Reasonably close to 196 V for the quality of measurements being made. We do not know if the neutral is completely open, and how accurate the 60, 170, 170 values are. But clearly you have at least a partially open neutral.

Then you went on to say that 15 ft from the transformer the voltages were similar to close to the load.

Clearly your problem is at least in the neutral and at a point close to the transformers. Why did you do all the other things mentioned when it was clear that somewhere close to the transformers there was a neutral problem? That needs to be fixed first, then look for other problems.

.

 

[Gman1969]

There is an issue with the neutral. You answered it in your post. " Utility company was called out to site and at the transformers they read proper voltages however, 15 ft down the line they were reading what I was the 60/170/170 "

Problem solved. Did they just measure that and walk away?
Or you have a phase faulted.

They measured and walked away. Their ground rod measured at 15 ohms yesterday. Mine at customer side is 1 ohm on both rods

 

[Gman1969]

210608-1044 EST

Gman1969:

I believe your answer is in your post. I also believe you do not understand electrical circuit theory.

You say that power company reads correct voltages at their transformers, and that these values are about 208 V line to line for all combinations, and 120 V from each line to neutral, and that these voltages do not change when you see voltage errors at various customers.

Note: 120^2 - (208/2)^2 = 14400 - 10816 = 3584, and the sq-root of this is 59.86 which is 1/2 of 120 as expected. For exactly 120 V line to neutral a more accurate value for the 208 V voltage is 103.928 X 2 = 207.846 . But you can expect the power company readings were not real precision values, and line voltage readings jump around a little.

But here is the important point. The power company readings were about constant and near the values indicated and expected. However, at your customer when the readings at the transformers were good the customer readings were very bad.

Now lets look at those readings. For example 60, 170, 170. Apparently line to neutral. I don't believe you read line to line voltages. Let us assume that line to line voltages were all about the same at about 208, a reasonable assumption based on your information. The vector squared from your neutral at the load to the midpoint of 208 for the two 170 V vectors is 170^2 -104^2 = 18084. The sq-root of this is 134.5 . Add 60 to this and the result is 196.5 . Using 208 the calculated value for this vector is 180.1 V. Reasonably close to 196 V for the quality of measurements being made. We do not know if the neutral is completely open, and how accurate the 60, 170, 170 values are. But clearly you have at least a partially open neutral.

Then you went on to say that 15 ft from the transformer the voltages were similar to close to the load.

Clearly your problem is at least in the neutral and at a point close to the transformers. Why did you do all the other things mentioned when it was clear that somewhere close to the transformers there was a neutral problem? That needs to be fixed first, then look for other problems.

.

I did in fact write down the voltages however misplace that piece of paper. those were rough numbers. the point is is that voltages were way out of normal. Power company basically refusing to come and do due diligence. Those other things I did were necessary.
Line to line from what I remember (kaos -was happening) was 208v)

 

[mikeames]

They measured and walked away. Their ground rod measured at 15 ohms yesterday. Mine at customer side is 1 ohm on both rods

The resistance of the ground rod is irrelevant. Either your terminology is not accurate or I can't understand the whole situation. If the utility measured correct voltage at the transformers but not 15 feet down the wire then that's where the problem lies. What is located 15 feet down the line?

As gar stated those other things you did are unrelated to your issue. Your voltages are wack because there is no neutral or a compromised neutral which is also why it's affected by tide and rain. That changed the conductivity of the soil which will change the current on the faulted neutral or phase. The problem is clear. It not a mystery. The question is where.

 

[Gman1969]

The resistance of the ground rod is irrelevant. Either your terminology is not accurate or I can't understand the whole situation. If the utility measured correct voltage at the transformers but not 15 feet down the wire then that's where the problem lies. What is located 15 feet down the line?

a splice in the line. How is it then the the traffic lights on same system are working ok then? Not enough load?

 

[Gman1969]

The resistance of the ground rod is irrelevant. Either your terminology is not accurate or I can't understand the whole situation. If the utility measured correct voltage at the transformers but not 15 feet down the wire then that's where the problem lies. What is located 15 feet down the line?

I and my other electricians have felt bad neutral all along. power company basically refusing to do anything until I can prove my case. hence the reading fro ground rods as this is where their neutral is derived for the transformers on the pole

 

[mikeames]

a splice in the line. How is it then the the traffic lights on same system are working ok then? Not enough load?

I don't have all the info but if there is a splice there, then that splice is bad or the wire feeding the splice. The traffic lights are not working correctly they are affected as well since you said you knocked out the traffic lights, so yes, the issue is not as apparent with a smaller load. This is yet another symptom of a compromised neutral. You have a neutral but its compromised. Thus why the lights work but the AC wont. Also why the AC will take out the lights.

A neutral is NOT DERIVED from a ground rod. The centerpoint of a wye is the neutral which is the transformers and you said the voltages there measured good at the transformers. Forget about the resistance of the ground rods at this point that's irrelevant. Current from the buildings goes back to the transformers not the ground rod. If your reading current on your ground rod thats proof the neutral is compromised. Additionally if you measured different voltages at the splice thats proof. How much more proof do they need. This is not a difficult situation to troubleshoot. This sounds ( I may be wrong) very straightforward.

 

[gar]

210608-2101 EDT

Gman1969:

I want to talk about how to find or prove were your problem may be.

I want you to measure the current from the transformer grounding wire. If these are pole mounted transformers, then a conductor will come down the pole to a ground rod at the pole. There may be substantial step potentials in various spots. You need insulated gloves, and shoes, and be careful how you make measurements.

At least some equipment you need is:
A high impedance AC voltmeter with resolution to 0.1 millivolts. A Fluke 27 meets this requirement.
A clamp-on AC ammeter, separate from the voltmeter.
One or two 12" long screwdrivers.
Some wire and battery clips. You may need a wire long enough to go from the transformers to the main panel.
An AC magnetic field sensor possibly would be useful. I doubt you would find something useful, and at this point I would have to run some experiments to determine how this might help search for your problem.
Said protective equipment mentioned above.

An AC current probe that is clamped around the transformer grounding wire will allow you to see how much current is flowing into the earth. If you disconnect all loads on your main panel, to protect the loads, and apply a single large load to one phase, two 120 V 1500 W space heaters in series to force a measurable current on one hot wire, and the neutral which may actually be thru earth. then check the voltage across the two heaters. This voltage should be 120 V or less. If so. then use just one heater as the load. Now you search for how much curren is flowing in each path, earth, neutral wire, and loaded hot wire.

Report back. The results will determine what voltage measurements to make. This needs to be done at a time when there is an occurring a problem.

.

 

[LarryFine]

Much ado about nothing. Or, in this case, about the obvious. Who owns the splice, POCO or customer?

 

[hillbilly1]

As Larry pointed out, who owns the cable? If it is the customer, that may be why they don’t want to fix it. Or they may be evidence someone dug in that area, didn’t get a locate and damaged it, and the poco is pissed off about it.