Ungrounded Delta service transformers in industrial environment

[francis.key19]

Hi,

I am hopeful someone can help me understand some of the questions I have concerning Arc Flash Hazards and our power systems on site.

My plant is about 60 years old, one structure of about 120,000ft2. There are several ground electrodes installed on the outside perimeters and bonded to building steels at various place using large wires. There are four transformers installed at various time during the last 60 years by the utility company. Three of these were installed about 20 years ago, the fourth one was installed about 5 years ago. Each transformer is 2500KVA with 480VAC on secondary. The first three were installed as ungrounded delta. The newer fourth one is a Wye with neutral, but the neutral is not carried from the transformer to the switchboard inside the building (ie. only phas conductors are carried inside). All transformers are connected to ground electrodes and thus building steel. The earthing system seems to be in good shape and all equipment inside the plant have earth ground connected to them.

The plant over the years had seen many cases where one leg of the delta would be grounded, but typically the people would not know of it until a second leg is grounded or they would detected because process heaters or process parameters are out of range. Then they would go an look for it. This seems to be normal mode of operation for us here for many years. There are no ground fault detector installed at any feeders on the three transformers. Last few months, we had a company came in to monitor surges on these feeders and saw several surges (> 20 in a course of an hour of large magnitude).

I do not have any experience working with ungrounded delta system and hope that the forum can help me with basic understanding of the followings. I also read several pdfs from schneider and eaton but still not very clearly understood.

1. What is meant by ground capacitive coupling per phase? This is in the context of if the ungrounded phase are properly balance, in normal situation, one would read close to line voltage at each phase to ground (ie. 277v or thereabout for a 480V phase) eventhough there is no solidly ground, but by capcitive coupling?

2. If one leg is grounded unintentionally, the voltage at that leg to ground would be zero? This would leave the other two phases at 480VAC ?

3. If situation in number 2 to occur, I read that the current to ground would be small, would this still not pose a hazards to an operator who may have had his/her hands on the equipment that was faulted? What would the transient voltage level be at that point ?

4. What is meant by "arcing ground transients"? Is this the second leg going to ground after the firs leg gone to ground?

5. How would you find the first leg fault if the bus has several motors, starters, pumps..etc..? These systems were grandfathered in, thus no ground fault detectors were installed, any idea what the cost for the detectors would be per system? Even that, how would you find where the fault is along the bus?

6. I read the effect of leaving the first fault and the second fault come along would cause a phase to phase fault in which case the OCPD should work to clear this. Would this not cause an arcing event somewhere else along the bus? If the first fault is left alone, what is the impact on the motors life that are connected to this bus?

7. Is an arc flash hazard worse for the ungrounded system vs. a grounded system like a WYE? That is, is it more likely to have more arc flash occurences with larger incident energy per event? Would operators be more at risks with these systems?

8. Lastly, since we also have 480VAC Wye feeder and 480VAC Delta, troubleshooting techniques would be different, how would we labeled at the device to make sure people aware of the differences? Can one system cause problem for the other if conductors are inadvertently connected.?

Any suggestions to improve the safety aspects of the ungrounded systems would be most appreciated. I am not a plant electrical engineer but am very concern about the electrical safety here. As I started to read more and more, I get a lost. Thus I am hoping the forum would be kind to help me understand these issues.

I did a search on this forum for "ungrounded delta" and found 8 pages of information, while they are quite helpful (troubleshooting), I still have some questions.

Please help if you would. Any basic resources you may have (links, pdfs..etc..) would be greatly appreciated as well.

Thank you in advance.

 

[texie]

Hi,

I am hopeful someone can help me understand some of the questions I have concerning Arc Flash Hazards and our power systems on site.

My plant is about 60 years old, one structure of about 120,000ft2. There are several ground electrodes installed on the outside perimeters and bonded to building steels at various place using large wires. There are four transformers installed at various time during the last 60 years by the utility company. Three of these were installed about 20 years ago, the fourth one was installed about 5 years ago. Each transformer is 2500KVA with 480VAC on secondary. The first three were installed as ungrounded delta. The newer fourth one is a Wye with neutral, but the neutral is not carried from the transformer to the switchboard inside the building (ie. only phas conductors are carried inside). All transformers are connected to ground electrodes and thus building steel. The earthing system seems to be in good shape and all equipment inside the plant have earth ground connected to them.

The plant over the years had seen many cases where one leg of the delta would be grounded, but typically the people would not know of it until a second leg is grounded or they would detected because process heaters or process parameters are out of range. Then they would go an look for it. This seems to be normal mode of operation for us here for many years. There are no ground fault detector installed at any feeders on the three transformers. Last few months, we had a company came in to monitor surges on these feeders and saw several surges (> 20 in a course of an hour of large magnitude).

I do not have any experience working with ungrounded delta system and hope that the forum can help me with basic understanding of the followings. I also read several pdfs from schneider and eaton but still not very clearly understood.

1. What is meant by ground capacitive coupling per phase? This is in the context of if the ungrounded phase are properly balance, in normal situation, one would read close to line voltage at each phase to ground (ie. 277v or thereabout for a 480V phase) eventhough there is no solidly ground, but by capcitive coupling?

2. If one leg is grounded unintentionally, the voltage at that leg to ground would be zero? This would leave the other two phases at 480VAC ?

3. If situation in number 2 to occur, I read that the current to ground would be small, would this still not pose a hazards to an operator who may have had his/her hands on the equipment that was faulted? What would the transient voltage level be at that point ?

4. What is meant by "arcing ground transients"? Is this the second leg going to ground after the firs leg gone to ground?

5. How would you find the first leg fault if the bus has several motors, starters, pumps..etc..? These systems were grandfathered in, thus no ground fault detectors were installed, any idea what the cost for the detectors would be per system? Even that, how would you find where the fault is along the bus?

6. I read the effect of leaving the first fault and the second fault come along would cause a phase to phase fault in which case the OCPD should work to clear this. Would this not cause an arcing event somewhere else along the bus? If the first fault is left alone, what is the impact on the motors life that are connected to this bus?

7. Is an arc flash hazard worse for the ungrounded system vs. a grounded system like a WYE? That is, is it more likely to have more arc flash occurences with larger incident energy per event? Would operators be more at risks with these systems?

8. Lastly, since we also have 480VAC Wye feeder and 480VAC Delta, troubleshooting techniques would be different, how would we labeled at the device to make sure people aware of the differences? Can one system cause problem for the other if conductors are inadvertently connected.?

Any suggestions to improve the safety aspects of the ungrounded systems would be most appreciated. I am not a plant electrical engineer but am very concern about the electrical safety here. As I started to read more and more, I get a lost. Thus I am hoping the forum would be kind to help me understand these issues.

I did a search on this forum for "ungrounded delta" and found 8 pages of information, while they are quite helpful (troubleshooting), I still have some questions.

Please help if you would. Any basic resources you may have (links, pdfs..etc..) would be greatly appreciated as well.

Thank you in advance.

While I have not thought through all of your good questions, and others may be more qualified to answer some of them, one thing jumps out to me. You mention that the POCO has replaced one of the transformers with a 480Y. Are you certain that they have not grounded the X0 at their end? If they have, and as you mentioned, you have no grounded conductor brought to the service disconnect and bonded, this is a serious NEC violation as you no longer have an ungrounded supply system and this could present a serious threat to life.

 

[hurk27]

Ungrounded systems have been around for many years, the problem is that not many truly understand them and are afraid of them, my personal view is for arc flash they are a much safer system when the system is monitored by a ground detection system, just because that it would take two phases to go to ground before an arc flash would occur, sure if you cause a phase to phase fault its still there, but most faults are phase to ground.

Finding a grounded phase is just a matter of elimination, by turning off circuits till the fault is gone, then trouble shooting that circuit to find the fault really narrows it down.

The whole idea of an ungrounded system is when a ground fault happens the whole plant is not shut down, instead it should sound an alarm that gives you time to do the repairs at a time that wont affect production as bad, some processes don't like to be interrupted and can cause allot of equipment damage if the production line was suddenly shut down, or there could be dangers to workers if it happens.

As far as having a phase go to ground on an ungrounded system being a shock hazard, its no more of a shock hazard then a corner grounded system which is what it turns into anyways, as long as there is a common grounding path through out the plant and to any other buildings that might also be fed from this system because without this common low impedance path these out buildings would or could get energized if one phase wend to ground in the main building and a different phase went to ground in the remote building and earth was the only path you would have up to 480 volts on all the grounding to earth in this remote building because earth will not have a low enough impedance to open any OCPD's.

Also you mention that the utility had installed a newer 480/277 WYE transformer and didn't bring the X0 (neutral) into the building? If there are no grounding path between this transformer and the buildings grounding system, and the X0 of this transformer has been bonded to its case and or the primary MGN then if a fault should happen inside of the building on this service there would not be a fault clearing path and again like the remote building the grounding of this service to earth could rise to 277 volts.

We had this happen once because the line men didn't know any better when they changed out a bad set of delta tanks to a wye they bonded the WYE connection of the tanks hanging on a pole to the poles MGN but since the plant was an ungrounded system that used to have a delta, and they only had 277 volt tanks with them, they had left the ungrounded drop to the building with no neutral, luckily I caught it when I saw I had a pretty solid 277 volts to the buildings grounding until I put a heavy load on it, we then measured almost 277 volts to earth with this heavy load pulling the buildings grounding up to the phase voltage level, we had to get the POCO to disconnect the Wye point to the MGN which was a hassle because we had to get their engineer involved to understand the problem.

Over all I think an ungrounded system could be a somewhat safer system if properly installed and maintained along with installing ground detection system, also just because its not grounded you still have to treat it like it was, because the shock hazards are still there, but like I said you don't have the danger of an arc on the first fault to ground.

 

[don_resqcapt19]

...
1. What is meant by ground capacitive coupling per phase? This is in the context of if the ungrounded phase are properly balance, in normal situation, one would read close to line voltage at each phase to ground (ie. 277v or thereabout for a 480V phase) eventhough there is no solidly ground, but by capcitive coupling?

Yes, if everything is equal, the voltages will read much like they do on a solidly grounded wye system. Most of the time everything is not equal, and the voltages are all over the place.

2. If one leg is grounded unintentionally, the voltage at that leg to ground would be zero? This would leave the other two phases at 480VAC ?

Yes, you have created a corner grounded system with one conductor having no voltage to ground and the other to having 480 to ground. You still have 480 phase to phase for all combinations.

3. If situation in number 2 to occur, I read that the current to ground would be small, would this still not pose a hazards to an operator who may have had his/her hands on the equipment that was faulted? What would the transient voltage level be at that point ?

Assuming a solid ground fault there is no hazard and a solid ground fault should not create transient voltage spikes.

4. What is meant by "arcing ground transients"? Is this the second leg going to ground after the firs leg gone to ground?

My understanding of this, is that it is a single phase "arcing" to ground. The capacitive current is enough to create an arc. This condition can create large voltage transients in the system to the point that equipment is damaged.

5. How would you find the first leg fault if the bus has several motors, starters, pumps..etc..? These systems were grandfathered in, thus no ground fault detectors were installed, any idea what the cost for the detectors would be per system? Even that, how would you find where the fault is along the bus?

If there is no ground detector you have to use a meter to find the ground fault. You would turn off the equipment one at a time and see when the grounded leg becomes ungrounded. I don't know how expensive the installation would be.

6. I read the effect of leaving the first fault and the second fault come along would cause a phase to phase fault in which case the OCPD should work to clear this. Would this not cause an arcing event somewhere else along the bus? If the first fault is left alone, what is the impact on the motors life that are connected to this bus?

Yes, if you have a ground fault on one leg, and then a second ground fault on another leg, you have a phase to phase fault with the current flowing on the EGC and other parallel paths. Assuming solid faults, this should quickly trip the OCPD. If one or both of the faults are arcing, this will limit the current, then OCPD may not trip and there will be extensive damage at the point(s) of the arcing.

7. Is an arc flash hazard worse for the ungrounded system vs. a grounded system like a WYE? That is, is it more likely to have more arc flash occurences with larger incident energy per event? Would operators be more at risks with these systems?

I do not know. I am sure someone else can address this.

8. Lastly, since we also have 480VAC Wye feeder and 480VAC Delta, troubleshooting techniques would be different, how would we labeled at the device to make sure people aware of the differences? Can one system cause problem for the other if conductors are inadvertently connected.?

Yes, the trouble shooting will be different for the two systems. Training and labeling will help. Not sure what would happen if the systems became interconnected.

 

[francis.key19]

Thank you so much for all your feedback.

If we convert these services to Wye, are there any changes needed to be done downstream of the services if the loads are all 3 phase, 480VAC as they are today? or it is better just to add GFD? or corner grounded one leg?

I am just curious to see what make sense economically to enhance the reliability and safety of our systems. We are in a very low margin business and company would want to get the most benefits for the dollars.

With much regards,

Francis.

 

[don_resqcapt19]

If you convert to a wye system, you will have to install a grounded conductor from the transformer to the service disconnect and install a main bonding jumper in the service disconnect.

 

[pfalcon]

I'm also in a very old plant. We still have large areas that are ungrounded delta. We also have solidly grounded wye and high resistance wye areas.

In any case the building steel is used as the grounding electrode system. Only the 480 conductors run through the buss. I've never seen the wandering values that Don speaks of. All our measurements are very close to 277 no matter where we are. The distinction may be the large number of machine transformers that we have installed. Each of those have a secondary bonded to the building steel. I believe the secondaries are coupling backward through those transformers to maintain the 277 readings on the 480.