Open Delta - Open Wye Fault Question 2

[rhatcher]

I am looking at a system fed by an open-wye/open-delta transformer pair. The primary is 14.4kV and the secondary is 240V, three phase, four wire with a grounded neutral connected to a center tap on one transformer. The nominal single phase voltages are high leg 120V/120V/190V. The loads are primarily single phase lighting and 'appliance' type loads with 2 three-phase motors used to operate large chillers.

The two single phase utility transformers are pad mounted with underground primary and secondary feeds and underground jumpers making the connections to form the open-wye/open-delta configuration. This is a rural distibution system in an area that is almost exclusively single phase loads.

The fault on the utility side was described as a 'failed underground jumper.' The actual location of the jumper in the circuit is unknown.

The result on the secondary side was failure of the 2 three-phase motors that burned up on confirmed single phase winding failures. The other obvious fault symptom is that the safety grounding jumpers on both equipment cabinets were burned free of the connection point. No single phase equipment was damaged. However, to my knowledge the only single phase equipment is lighing and appliance loads that were mostly off at the time.

The fault occurred at night and was discovered early the next morning because the (single-phase) lighing was dark/dim. Subsequent voltage checks revealed that "the voltages were all wrong." Sorry but there is no record of what the measured voltages were. Subsequently the damage to the chillers and the the failure of the underground jumper were discovered.

The big question is that phase loss monitors were installed on both three phase motor loads at the time of the failure. Obviously they did not activate during the failure event. I am also told that although they appeared physically undamaged that when the power was restored that the phase loss monitors failed to power up. Somehow they were internally damaged by the 'event.'

This would seem to be pretty simple case to explain if it were not for the presence of the the phase loss monitors, the fact they failed to act, and the fact that they were somehow damaged and rendered inoperative by the 'event.'

Although the failure of any primary or secondary jumper on the transformer pair would cause single-phasing of down stream three-phase loads, I have a gut feeling that my culprit is the primary side neutral conductor. This is because I am looking for a failure mode that would be accompanied by voltage spikes and/or transients that would take the phase loss monitors out of service (ie, fry the phase loss monitors).

I am suspecting that if the primary side neutral connection failed (burned loose) that the resulting effect of the transformer voltages and currents trying to 'instantaneously' change phase from 120 degree phase shift to 180 degree phase shift would cause some interesting and perhaps dramatic transients.

I would also think that this is would be especially true if the failure mode was a slow burn with a sputtering arc (ie. back and forth transients) instead of a fast and simple 'blow-out.'

Does this make sense to anyone else?

I have reviewed past posts on this topic and I see that waross seems pretty knowledgeable on this type of transformer arrangement so I am hoping that he has some thoughts on this.

 

[waross]

I'm going to ignore the issue of the secondary grounding jumper burned off for now.
First the primary jumper. From the description I will assume that the transformers remained connected but the connection to the primary neutral was lost.
If a chiller(s) was running, it would act as an induction generator and would force the transformer bank to maintain close to the normal phase angles.
The phase monitor may not detect this condition.
The motor(s) would be drawing much more than normal current.

That the motors burned out indicates that they may not have had proper overload protection or current unbalance protection. The phase monitors may have had voltage unbalance protection but on an open delta supply the voltage unbalance feature may have been disabled or set to an unrealistic value.

When the motors burned out, one or both may have gone to ground. The resulting ground fault current may have burned off the ground connection. There may have been some poor workmanship involved. A loose connection is much more likely to fail under fault current.

When the motors burned out, the transformers would revert to a single phase circuit with the two transformers in series across one primary phase. There would be more back EMF from the motor's to force the vectors into close to the normal open delta position.

Now there would be two transformers in series across one primary line to line phase. These would be 8.3 kV transformers in series across 14.4 kV. If the loading was equal the secondary voltages would be 69.2V, 69.2V and 138.3 V rather than 120V, 120V and 240V

But the loading is not equal.
Case one:
One transformer is loaded and one may be shorted by burned out motors. If the motors are shorted the normal 120v, 120V, 240V single phase voltages will try to rise to 208V, 208V, 416V. I say try because before the voltages rise that high, the transformer will saturate and limit the voltages to about 120% or 144V, 144V, and 188 Volts. Although this may have happened and probably did, it would have been very short lived or the primary protection would have disconnected one or both transformers.
It may have occurred momentarily as the motors failed and then burned free.
Case two:
Two transformers are in series. One is unloaded as the motors have burned free but the other is loaded with single phase loads. If these were resistors, one a low value (loaded) and the other a very high value (unloaded), the voltage across the load would drop to close to zero and the voltage on the other side would rise to 416 Volts. But these are not resistors they are transformers and saturation will limit the over voltage to about 288V on the unloaded transformer and about 64V, 64V, and 128V to the sinle phase loads rather than 120V, 120V, and 240V.
This seems to fit the information.
Somewhere in there, the monitors saw something that they couldn't cope with and died. Depending on which phase they were on and the saturation voltage relative to the transformers the magnetic coils in the starters may have shorted. If that was the case, when the monitors were re-powered and tried to switch shorted coils they may have failed then. One flash and it's all over.
a couple of notes:
1> I am using the voltages of 120:120:240 as the single phase voltages and the 240 volt delta voltage. I have ignored the "wild leg" voltage to make things a little easier (for me).
2> When the neutral connection is lost and the circuit transforms from a three phase circuit to a single phase circuit, there is a factor of 1/1.73 or 2/1.73 to consider.
3> Anyone who wishes to propose 115% rather than 120% as the saturation point is welcome to and is also welcome to redo the math.
I hope you have been able to follow my ramblings.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rhatcher]

Thanks for the quick repsonse waross. You have given me a lot to digest. I am thinking about your response will reply when I have it sorted out in my mind.

 

[waross]

I hope that I haven't made any silly typos.
Thanks for the kind words.
Bill

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rhatcher]

Hey waross, are you making fun of me??

Thanks for the quick repsonse waross.

I hope that I haven't made any silly typos like you did, hehe.

Ok, I did add the part in italics and I know that you weren't making fun of me (I think).

Anyway, I am still mentally digesting your last post. It was a good one. I came to check the forum and see if anyone else had responded and there are none.

So, do you think that no one else is responding because I 'asked' for you in my original post, is it because of the ungrounded-wye/ungrounded-delta topic, or, is it because of my silly typos??? Haha...

 

[waross]

Any number of things could have started the failure. One motor may have shorted out and the resulting current may have caused the failure of the primary neutral jumper. A secondary fault, either ground fault or phase to phase can cause heavy current in the primary neutral jumper.
I originally presented one possible scenario starting with the failure of the neutral jumper to keep things as simple as possible and make an explanation simpler. But, the primary jumper may have failed due to corrosion or poor workmanship and things went downhill from there, OR a secondary fault could have led to the failure of the primary jumper.

Hey waross, are you making fun of me??

Emphatically no. I have read your postings for years and consider you to be a competent professional. I have no reason or inclination to denigrate you in any way. grin.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rhatcher]

waross, I know that were not actually making fun of me....It was just ironic that your post expressing concern about making a typo was responding to my post that had a typo in the first line.

I couldn't resist the opportunity to make fun of that irony, even if it was at my own expense. In either case, a good irony and a good laugh are both too good to pass up.

And thank you for the nice complement. Fortunately, I am at least slightly better at my job as an engineer than I would be if my job were a typist.

About the engineering topic at hand, your responses in the first and second posts are on topic and relevant. Thanks so far. I will respond seriously on this topic soon. I am still gathering my thoughts (with a butterfly net).

 

[rhatcher]

waross, I agree with the following.

I will assume that the transformers remained connected but the connection to the primary neutral was lost.....When the motors burned out, one or both may have gone to ground. The resulting ground fault current may have burned off the ground connection... If the loading was equal the secondary voltages would be 69.2V, 69.2V and 138.3 V rather than 120V, 120V and 240V...Somewhere in there, the monitors saw something that they couldn't cope with and died.

Where you lose me is:

If a chiller(s) was running, it would act as an induction generator and would force the transformer bank to maintain close to the normal phase angles....

I would think that the primary transformers would force the phase angle to 180 degrees, single phase. The secondary transformers would have the same phase angle as the primaries, and the motors would have no choice but to receive single phase voltage. I cannot see how the single phased motor could become an induction generator and force the secondary and primary transformers to hold a 120 degree phase angle once the primary neutral connection was lost.

Also: you lost me with everything beyond:

But the loading is not equal.

This is not because I think you may be wrong, it is just because I am having trouble visualizing the effect you are talking about. As an example, a single phase transformer with a split, 240/120V secondary would therefore have unbalanced secondary voltage if the attached load was unbalanced. This is not the case.

However, I see where you are coming from with this comment so I am still trying to figure this out....

Again I will say that I agree with your conclusions about the burned out single phased motors and the burned grounding jumpers. What I am looking for is a mechanism that would explain a transient high voltage surge through the secondary circuit that would fry the phase monitors and take them out of service before they could act to prevent the motors from being single phased.

 

[waross]

Hi let's start where we lost it.
Many induction motors but probably all induction motors suitable for chillers will act as induction generators. NO question. If they are three phase motors they will, by virtue of the 120 electrical degree spacing of the phase windings generate three phase power. This is how a rotary phase converter works. A three phase induction motor is started on single phase power with the aid of capacitors connected to the third motor lead.
Now in the case of an open delta with a missing primary connection to the neutral, I am still trying myself to visualize the currents in the various transformer and motor windings, but I have no doubt that the motor will have a good chance to continue running on single phase and will supply the missing phases to hold the open delta common point in close to the correct vector position.
Let me say that a different way.
The motor WILL act as a generator and will back feed into the transformer windings and will force the approximately correct phase angles.
With the phase angles held in the proper position the open side of the delta will energize the motor in a single phase condition. The motor will continue to rotate and will continue to act as an induction generator. This is not something that may be started from rest but if it is running on three phase power and one phase is lost, it will continue to run.
Please note. Current will be excessive and it will not run long before failure, but we are analysing a failure, the motors did fail.

"This is not because I think you may be wrong, it is just because I am having trouble visualizing the effect you are talking about. As an example, a single phase transformer with a split, 240/120V secondary would therefore have unbalanced secondary voltage if the attached load was unbalanced. This is not the case. "
You are correct there, the unbalanced voltages that I meant were the outputs of the two transformers. You would expect that when two 240 Volt transformers end up in series across 1.73 times the normal voltage the average voltage will drop to 1.73/2 but the actual voltages will be inversely proportional to the loading until one transformer approaches saturation. One transformer may develop 200 Volts while the other develops 215 Volts. The voltages on the two 120 Volt windings will still be equal but will probably no longer be 120 Volts. They will both be 50% of whatever voltage the center tapped transformer is developing.
Back to your original statment
"This is not because I think you may be wrong, it is just because I am having trouble visualizing the effect you are talking about. As an example, a single phase transformer with a split, 240/120V secondary would therefore have unbalanced secondary voltage if the attached load was unbalanced. This is not the case." Think "open neutral".
As for the phase monitors, I have been saved by them a few times. I have also been let down a time or two.
A fascinating problem rhatcher. Any number of failures are possible with the information available. The chillers may have failed in such a way as to have shorted out the teaser transformer but burned free on the main transformer. This may have caused 173% voltage on the lighting transformer, (limited by saturation).

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rhatcher]

"Any number of failures are possible with the information available"....This is very true

Since there were no reported failures on single phase equipment I am focusing on the three phase aspects of the problem.

In particular, I am trying to come up with possible causes of high voltage transients (ie. a few cycles) or possible causes of sustained high voltage that would internally burn-out two separate phase monitors on two separate pieces of equipment. The burn-out of the two separate phase monitors is the only aspect of this failure that I cannot explain.

If the primary neutral burned free, what transients would be created? Before the fault, voltage and current in the primary have 120 degrees phase separation. After the fault, from a few cycles up to a few seconds later (?), the primary voltages and phase will have 180 degree phase separation. What happens during the transition? Specifically, the voltages and currents in the primary will not, cannot, instantaneously change phase and amplitude. There must be some transient condition.

As an example of what I am trying to say, consider the assymetrical short circuit current created during a three phase fault. The DC offset of the current is created because the current cannot change amplitude instantaneously to respond to the fault demand.

In the case we are considering, there is no short circuit, there is an instantaneous(?)disconnect from the primary neutral connection. The voltages and currents will attempt to change phase angle instantaneously but theoretically cannot. Will this cause any strange transients that would produce high phase voltages?

"Any number of failures are possible with the information available"....

I do not know which jumper failed and I am trying to get this information. It could have been any jumper on the high side or the low side. I have been favoring the idea of the primary neutral jumper because it seems most likely to cause high voltage transients. However, anything is possible. One thought that recenly came to mind is the possibility that the failure resulted in a autotransformer effect. Keep in mind that the primary neutral and secondary neutral have a common grounding point. Do you see this as a possiility?? I'm still trying to decide for myself....

Finally, about the analogy between the rotary phase converter and the single-phased compressor motor. I'm sorry to be stubborn about this point but I do not get it. My understanding is that the rotary phase converter requires capacitors to provide the excitation necessary to generate the missing phase. In the case of the single phased induction motor, there are no capacitors and therefore no excitation.

If what you are saying is true, then any three phase induction motor could be used on a single phse system as a rotary phase converter without requiring the capacitors and other gizmos that you find on a rotary phase converter.

I'm sorry I do not understand some of your points. Thanks for your patience.

 

[ccjersey]

A 3 phase motor used strictly as a phase converter needs the capacitor(s) only to get it started. If it can be spun up with a cord wound around a pulley for example, it doesn't have to have capacitors to regenerate the 3rd phase.

A 3 phase motor used on single phase supply can do more work if it has a run capacitor between one single phase line and the 3rd phase.

 

[waross]

As ccjersey stated. The capacitors are not needed once the motor is turning but they help and they also increase the capacity of the phase converter.
I am wondering whether the control circuit saw some part of the fault current rather than a transient. That may take out the load switching circuit, possibly with an arc which damages other internal circuits.
Is it possible that a remote fault on the distribution system sent some transients down the line and that started the destruction, taking out first the phase monitors and leaving the motors unprotected?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ccjersey]

In my limited experience, phase monitor units are not very robust. The ones I have used frequently fail when nothing else has been damaged. I would not put much emphasis on what caused them to fail since it seems to take very little.

I do think it's unlikely that they failed before the compressor motors since that would take the compressors off-line unless the contacts welded closed instead of opening when the monitor control circuit failed from a voltage transient etc.

There are supposedly monitors that are better at detecting running motors which have been single phased. I don't know any specifics, just have been reading the advertisements.

The integrated overload/phase loss controls do seem to be more sensitive than the stand-alone phase monitors I have used. One I have on a grain aeration fan is almost a nuisance since it will not reset itself once the 3phase event has passed and as a result it usually runs only a few hours at at time before needing to be reset.

 

[rhatcher]

Is it possible that a remote fault on the distribution system sent some transients down the line and that started the destruction, taking out first the phase monitors and leaving the motors unprotected?

I do think it's unlikely that they (phase monitors) failed before the compressor motors since that would take the compressors off-line unless the contacts welded closed instead of opening when the monitor control circuit failed from a voltage transient etc.

Waross and ccjersey are correct in their thinking; these are the exact conflicting points that I would like to be able to explain. Waross's suggestion about voltage transients taking out the phase monitors seems likely.

But, ccjersey is right to say the failure of the phase monitors should have opened the output contacts and stopped the motor before it failed. The exception would be if the output contacts became welded closed.

Perhaps it's would be sufficient to assume that if there were a voltage transient that would take out the phase monitor electronics that it also could (would?) weld the output contacts closed(?).

 

[ramiro1]

Your description of the two transformers providing single phase and three phase capability suggest that the setup is a Scott T distribution used in many old distribution systems.
The Scott system interconnects the secondary windings of the transformers such that one side of the teaser transformer provides 86% of the voltage where the other side provide 50%. This set up allows to have 240v three-phase to run the three-phase induction motors.
Since the sencondary windings of the transformers are interconnected,the interruption of power to the primary of one tranformer will reverse the polarity (180 out of phase ) of the inerconnected section of the tranformer. The induction three phase motor will be supplied with 240 volts and with 106 volts reverse polarity to complete the three phase. The normal motor control scheme will not detect the fault since it operates with single phase power. The result is that the induction motor will vibrate and short circuit the rotor and sator windings. i.e. burn the motor. The effect is almost instantaneusly. For this to occur the motors must be running. Sconcha

 

[busbar]

An aside—from practical experience, voltage-mode imbalance [ANSI device 47] protection should only be applied to inhibit 3-phase motor starting—where current-mode relays [ANSI device 46] are needed to reliably interrupt motor running.

 

[davidbeach]

Hear Hear. Voltage imbalance, particularly if phase angle is not considered, is a poor way of determining service loss of phase if there is a significant amount of motor load or if there are transformers capable of creating a phantom phase. Current imbalance/negative sequence current works great.

 

[waross]

Not a Scott connection.
The open delta connection is common in some areas of North America and never seen in some other areas. It depends on the standards of the utility in the area.
Early in the last century when Scott transformation was used to convert three phase to two phase and vice versa, the open delta connection was also common. The open delta has long outlasted the Scott transformation.
There have been T connected three phase to three phase transformers built that resemble a truncated Scott connection. They were at one time common for 480 V to 120/208 V transformation at about 25 KVA.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[lowend15]

I apologize for being slightly off topic, but I wanted to post this in a recent thread about the Open Delta connection.

It seems that every time the Open Wye Open Delta is spoken of, nobody truly understands it (not saying that about anybody here, but just in general...there is some good discussion going on in this particular thread.) Somehow, the mystery of three phases from two phases baffles even the most clever lineman, electricians and engineers. They know it works, and that they will need orange tape, and there is a "phantom" or "ghost" or "high" or "stinger" or (insert other colloquial term here) leg, but why exactly?

A utility I used to work for used this connection exclusively in their substations for station service, and I felt I needed to understand it for myself. Not that this is even the question of this thread, per se, but hopefully the attached diagram can shed some light on how this connection truly works...as I have read, unless you draw this you'll never understand it, at least not to a point you can teach somebody else how it works! I drew this up a while back, and feel, in light of how much I have learned through volunteers on the internet, it was time to give back a little . Hopefully this helps somebody that may do a search for this.

Warning, I remember finding a typo, and don't know if I fixed it or not, so if you find it, just let me know. It may or may not be there, just wanted to throw it out there just in case.

Robert