Bizarre partial power failure exprienced...

| So at around 9PM, here in Miami, the power blinked for about three | seconds. No biggie, I have a UPS. Then I got some a report that a power | blink ocurred at work, but only certain circuits were affected. I | confirmed that both failures happened at the same time. | | Is it possible for one phase feeding a substation to open up, resulting | in a power failure for only a certain fraction of the single-phase loads | fed from that substation? And yes, I'm certain that the loads that | remained powered at work are not on any sort of emergency generator or | battery backup system.

Yes it is possible for one phase to open up. As I understand it, if it happens, it is supposed to be detected, and if not corrected, open the other two phases to avoid problems.

Exactly why certain circuits did not experience this short outage, or were not affected by it, is harder to determine. But if one phase going into a delta primary transformer were to open up, one of the windings would still have full voltage, and one of the phases on the wye secondary would still have full voltage. The other two would have half voltage since the two windings that have the failed phase in common are now just two windings in series with the remaining phases. Down the line, you could have additional delta-wye transformers which would translate that back into a dead phase (two legs of a wye secondart flatten out and have

0 volts between them).

| Either that or it was just a coincidence that both power failures | happened at once, and that there was a malfunction on the power grid at | work.

Unlikely coincidence.

Usually the first time a fault occurs, it will be reclosed within a few seconds, and 3 seconds seems plausible (I've seen other time frames). If the fault cleared in that time, power's back on.

But I have also seen 3 incidents of a single phase outage with lots of service getting half voltage.

In the US electric utililities generally provide no protection from "single-phasing" and state so in the terms of service in the tariff. It is the customers responsibility to protect their facility from a loss of one phase. Most customers don't seem to realize this even though it is right there in the service document.

Charles Perry P.E.

Why is it that the Electrical Codes do not normally require this protection then? I know the standard answer is that the electrical codes are concerned with safety and not convenience. But, this would seem to be a safety issue...

Case in point. I was once vice-president of a condo association that sustained $24,000 in damages when all the elevator motors burned out due to a loss of one of the three phases when there was a problem at the fire pump. (The utility fuse at the pole had blown for one phase only).

These were critical life-safety elevators serving elderly people in wheelchairs. Everyone assumed that the building developer and his electrical contractor had done their job properly and, of course, by bare minimum standards, they claim that they did. It seemed to be an afterthought that a few inexpensive devices costing ($100 - $150) could have fully protected us, yet they were not installed because they were not required. It seems to me that if fuses or circuit breakers are required to protect your wiring, an protection from single phasing should be required to protect your motors.

Beachcomber

Don't ask me. I once did a power quality investigation at a brand new, very expensive, cash processing center. They had a gold plated electrical system...it was all way over built. They had 4 times the onsite generation that they needed and had 2.5 times the UPS capacity. But.....they tried to save $15k and left of the loss of phase protection. It cost them about 3 times that is repairs to elevator equipment. The person responsible for cutting out the protection (a VP) was "released" I believe.

Charles Perry P.E.

|> Yes it is possible for one phase to open up. As I understand it, if it |> happens, it is supposed to be detected, and if not corrected, open the |> other two phases to avoid problems. |>

| | In the US electric utililities generally provide no protection from | "single-phasing" and state so in the terms of service in the tariff. It is | the customers responsibility to protect their facility from a loss of one | phase. Most customers don't seem to realize this even though it is right | there in the service document.

However, the problem exists with single phase customers getting half voltage. That is not allowed by the tariffs, at least in this state, based on successful lawsuits against the utilities. Running 120 volt applicances on 60 volts for long periods of time can cause widespread damage.

I'm talking about this taking place on a city-wide basis, not about single three phase customer where one transformer blows a fuse.

Obviously the power company can be exempted when things are beyond their control. But choosing to not open the remaining phases when one opens up for some reason, is within their control, at least after some reasonable period of time, short of substation damage.

| Why is it that the Electrical Codes do not normally require this | protection then? I know the standard answer is that the electrical | codes are concerned with safety and not convenience. But, this would | seem to be a safety issue... | | Case in point. I was once vice-president of a condo association that | sustained $24,000 in damages when all the elevator motors burned out | due to a loss of one of the three phases when there was a problem at | the fire pump. (The utility fuse at the pole had blown for one phase | only). | | These were critical life-safety elevators serving elderly people in | wheelchairs. Everyone assumed that the building developer and his | electrical contractor had done their job properly and, of course, by | bare minimum standards, they claim that they did. It seemed to be an | afterthought that a few inexpensive devices costing ($100 - $150) | could have fully protected us, yet they were not installed because | they were not required. It seems to me that if fuses or circuit | breakers are required to protect your wiring, an protection from | single phasing should be required to protect your motors.

The NEC doesn't cover safety issues like the proper use of railing on stairways, or the proper glass to install in windows. It only covers the safety issues of fire (it is published by NFPA) and electrocution. Just because the issue here is electrical based, does not mean NEC is the authority for it. One could argue that electrical failure of a refrigerator could cause food to spoil resulting in illness or death if eaten. But that's not an NEC issue.

I see the word "assumed". It should have been the responsibility of the developer to understand the nature of the building use, if that was known at the time, or for those who re-purpose the building to upgrade it as appropriate when the change is made.

City-wide? Not a distribution problem then, has to be transmission.

Not sure what state you are in, but if the utility is using line to ground connected transformers (high side connection) then a single phase of the primary opening means you are either totally out or totally in power. This is true if it is the high side fuse of the transformer or a main line fuse or single phase recloser on the main line.

Charles Perry P.E.

Code are only for human safety. Codes don't care whether equipment is damaged or not. It is manufacturer of equipment (ie hospital equipment) to design any internal protection that may be required to protect equipment or human. Equipment must assume standard failures such as loss of any one phase. If loss of one phase damages equipment, then design of equipment was defective.

Your elevator is a classic example. Loss of phase was not the problem. Bad elevator design was reason for failure. This would not be covered by electrical codes. However you make a good point about equipment installed to meet fire safety codes. Fire safety codes (may have) failed to foresee a danger - or elevator was not in compliance with those building fire safety codes - not to be confused with electrical codes. This elevator failure is not the domain of electrical codes.

Electrical codes do not def> Why is it that the Electrical Codes do not normally require this

It could be upstream of a transformer bank in the distribution network. Say there's a delta-wye bank wired between a newer portion of the network (at a higher voltage) and an older one (lower voltage) and a phase is lost. Downstream system has two phases at half voltage, and any houses fed by transformers on those phases get half voltage (assuming no protection gear).

I often see power 'hits' here at work. I often see only a flicker but I hear an AC hum for a second, and maybe 1/3 or 2/3 of the people lose their office systems. The 1 second hit sounds like a recloser in action. The big systems in the lab crash (no UPS, and I think there is protection gear that cuts the power if anything funny happens).

Weirdest power problem that I've seen was one night I was awakened by an AC-powered carbon monoxide detector going off. Except it sounded rather sick. When I turned on the light it glowed a very dull orange. Checked out everything to see if it was a problem on my end, found nothing. AC voltmeter plugged into an outlet read something like 45 volts. Unplugged the refrigerator, shut off the furnace emergency switch and went back to bed.

Because of Starling migration we used to suffer with line line shorting due to their weight suddenly landing on the lines at 4 o'clock every day during the winter - we installed phase loss detection which coped quite well except when the problem wa reduced voltage as that caused chatter in the contactors which burnt them out so we had to go to voltage sensing with dropout & manual reset - luckily the Starlings have now moved on

Wow, that was pretty strange. Did you hear anything about it in the news or from friends/coworkers the next morning? Like people complaining that equipment was damaged? Or was it very localized and no one cared?

I had something similar happen once here; when hurricane Andrew hit us back in 1992, the power flickered (little blinks) for the first 20 minutes or so of the storm, then suddenly all the lights got very dim, much as you described, and the refrigerator sounded like it was lurching... This condition lasted for about 10 seconds or so and then it failed completely and didn't come back for six days... I didn't have enough time to get a voltmeter; from the sound of it you did, so it must have persisted longer (or you had the voltmeter very handy)

So apparently now I'm enlightened; it was probably a phase going out somewhere.

-Z

I think it was fairly localized, didn't hear about it from anyone else or the news. It lasted long enough for me to be rudely awakened, to figure out what the heck was going on, check for something burning or misbehaving electrically, realize it was probably very low voltage, look at the street light and the neighbor's houses for power (not that a dark house meant anything at 3:00AM but the light was meaningful) get out and use the fairly handy meter, yank the power and go back to bed. I remember weird noises, I think from the fridge. I guess I was up at least 10 minutes, everything was OK when I got up later.

| City-wide? Not a distribution problem then, has to be transmission.

In one case where it occurred in a city of 60,000 near where I lived, I heard that it was a fuse on the primary of the transformer feeding most of that city. That in itself is not the issue. The issue is they left it that way for nearly 3 hours before cutting power.

|> Obviously the power company can be exempted when things are beyond their |> control. But choosing to not open the remaining phases when one opens up |> for some reason, is within their control, at least after some reasonable |> period of time, short of substation damage. |>

| Not sure what state you are in, but if the utility is using line to ground | connected transformers (high side connection) then a single phase of the | primary opening means you are either totally out or totally in power. This | is true if it is the high side fuse of the transformer or a main line fuse | or single phase recloser on the main line.

Single phase lost on a delta primary where the wye secondary feeds street level distribution where lots of single phase line to ground transformers feed residential areas. If the street primary was 12470/7200, then what you'd have is 3600 on 2 of the 3 hot phases, and 120/60 coming out of those transformers on those 2 low phases.

I've seen this in Illinois, Ohio, and West Virginia. Many years ago my grandfather successfully sued his electric company for losses due to the low voltage persisting for half a day (he had to replace two freezers and an air conditioning unit).

If it was a blown fuse, there is no way to automatically detect and open the other three phases. Someone has to drive out and do it.

Wrong. The single phase transformer is connected line to ground. If the voltage on the primary is 1/2 of normal then the secondary voltage will be

60-N-60 not 120-N-60. A single phase transformer connected line to ground has no reference to the other two phases and is unaffected by the voltage on them.

Charles Perry P.E.

One idea that I haven't seen mentioned yet in this thread is a single-phase fault on a third feeder.

Imagine you and your work are fed from a three-phase substation. And consider further another load, unrelated to you also fed from that substation.

Now, the third load develops a large phase-ground or single phase-phase fault. If it is severe, it can cause the substation secondary voltage to be pulled down near zero. So you and work see a loss of one phase. Then the fault is cleared by any number of protection schemes and the 'lost' phase voltage restores to the remaining loads (you and work).

With a severe fault, this can happen in the blink of an eye. Medium faults may take a second or two to be cleared depending on how they are protected (fuse, inverse-time-delay relay, mho-relay, etc...). Or if primary protection failed and a backup trip had to take place to clear an outgoing line from the substation.

Just speculation.

daestrom

Did the managment of the utility company not have any common sense?

Assuming they were in control and could shut things down completely, did they not realize that running perhaps hundreds or thousands of their customers refrigerators at 60 volts for several hours was going to create liability issues?

Beachcomber

Elevators are defined in standard ICC A117.1. Don't know details or even if this defines internal electrical protection

- how elevator should resp> Why is it that the Electrical Codes do not normally require this

|> | City-wide? Not a distribution problem then, has to be transmission. |>

|> In one case where it occurred in a city of 60,000 near where I lived, I | heard |> that it was a fuse on the primary of the transformer feeding most of that |> city. That in itself is not the issue. The issue is they left it that | way |> for nearly 3 hours before cutting power. | | If it was a blown fuse, there is no way to automatically detect and open the | other three phases. Someone has to drive out and do it.

If they have no automatic switch that can be programmed to open when the phase loss is detected (it wouldn't be that hard to detect), or at least opened when someone at the control center discovers the problem, then yes, someone will have to drive out. 3 hours? OK, maybe. In another event much further back in time I know about, it was for at least 15 hours.

| |> Single phase lost on a delta primary where the wye secondary feeds street |> level distribution where lots of single phase line to ground transformers |> feed residential areas. If the street primary was 12470/7200, then what |> you'd have is 3600 on 2 of the 3 hot phases, and 120/60 coming out of | those |> transformers on those 2 low phases. | | Wrong. The single phase transformer is connected line to ground. If the | voltage on the primary is 1/2 of normal then the secondary voltage will be | 60-N-60 not 120-N-60. A single phase transformer connected line to ground | has no reference to the other two phases and is unaffected by the voltage on | them.

Nothing wrong about what I said. You just misread it. The term "240/120" commonly referrs to a center tapped 240 volt winding where the tap is the grounded neutral. When the primary is 1/2 voltage, the whole winding has only 120 volts, and each hot leg has 60 volts. What "120/60" means is you get 120 volts and 60 volts based on hot-to-hot and hot-to-neutral. It's not 120-N-60 and I never said it was.

since:

240/120 generally refers to 120-N-120 then: 120/60 refers to 60-N-60

If you doubt these terms I suggest you look at lots of literature, including transformer catalogs from companies like Square-D, Cutler-Hammer, Acme, and others, as well as utility tariffs and service guides.