Ground Fault Protection in a DELTA DELTA system

Status
Not open for further replies.
D

D!NNy

Senior Member
Location
San Luis Obispo
Here is summary of my application:

Utility is feeding a 500kVA Step down transformer from 16kV to 2.4kV feeding 2.4kV overhead distribution circuit with a Delta Detla Configuration. connected loads are 2.4kV-480/277V step down transformers feeding 480V loads

Goal is to install a ground fault protection device per utility requirement: For the customer service disconnect recloser to be installed on the pole on the 2.4kV system.
prior to the main questions please correct my analysis:
  1. If there is a single phase to ground fault in the system any loads connected on line to ground (which is 2.4kV/1.732=1.38kV) will be under stress of high voltage (which is 2.4kV line to line Voltage). If all the connected loads are rated for the 2.4kV insulation and higher there should be any trouble for single phase faults.
  2. In a Delta Delta system since for the ground fault there is no return path to source there is very minimal fault current and load remains running normal conditions if so why do i need a ground fault protection. Note: Is this because of most single phase faults lead into 2 phase and 3 phase faults. If so my recloser will trip for the 2 phase and 3 phase faults anyway as over current protection will be enabled.
Option1:
Ground Fault Detection: Install the PT's (2.4kV-120V) connect the primary in Wye and secondary in broken Delta. The voltage developed across the broken delta of the PT's will be 3V0 (Zero Sequence Voltage-any unbalance in the Voltage). can some body help me calculate manually how much this phase unbalance will be for a loss of one of the phases? so that i can enable the relays only for a loss of phase instead of the unbalance voltage in the phases. using this voltage i can pickup a contact from a relay and use it in the recloser to trip.
Disadvantage is entire facility will be down.

Option2:
High resistance Ground: This is actually similar to the delta delta system only advantage of HRG is it will tell you when there is ground fault where as in delta delta it don't. there is no interruption in the service. Also no high voltage stress ( Phase to Phase Voltages) appeared in the on the phase to ground connected loads.


Please pull me out of this grounding well.......
Thanks in advance for the input.
 
D!NNy said:
Here is summary of my application:

Utility is feeding a 500kVA Step down transformer from 16kV to 2.4kV feeding 2.4kV overhead distribution circuit with a Delta Detla Configuration. connected loads are 2.4kV-480/277V step down transformers feeding 480V loads

Goal is to install a ground fault protection device per utility requirement: For the customer service disconnect re-closer to be installed on the pole on the 2.4kV system.
prior to the main questions please correct my analysis:
  1. If there is a single phase to ground fault in the system any loads connected on line to ground (which is 2.4kV/1.732=1.38kV) will be under stress of high voltage (which is 2.4kV line to line Voltage). If all the connected loads are rated for the 2.4kV insulation and higher there should be any trouble for single phase faults.
  2. In a Delta Delta system since for the ground fault there is no return path to source there is very minimal fault current and load remains running normal conditions if so why do i need a ground fault protection. Note: Is this because of most single phase faults lead into 2 phase and 3 phase faults. If so my re-closer will trip for the 2 phase and 3 phase faults anyway as over current protection will be enabled.
Option1:
Ground Fault Detection: Install the PT's (2.4kV-120V) connect the primary in Wye and secondary in broken Delta. The voltage developed across the broken delta of the PT's will be 3V0 (Zero Sequence Voltage-any unbalance in the Voltage). can some body help me calculate manually how much this phase unbalance will be for a loss of one of the phases? so that i can enable the relays only for a loss of phase instead of the unbalance voltage in the phases. using this voltage i can pickup a contact from a relay and use it in the re-closer to trip.
Disadvantage is entire facility will be down.

Option2:
High resistance Ground: This is actually similar to the delta delta system only advantage of HRG is it will tell you when there is ground fault where as in delta delta it don't. there is no interruption in the service. Also no high voltage stress ( Phase to Phase Voltages) appeared in the on the phase to ground connected loads.


Please pull me out of this grounding well.......
Thanks in advance for the input.
Click to expand...

Are you sure the utility was a ground fault protection device or do they want a ground fault detection device?

The difference with having an ungrounded system out on poles or underground is if there is no common conductor such as a MGN or EGC which I would just call it an Earth reference conductor since it would not be current carrying run with the ungrounded conductors, without this conductor there is no low impedance path if a fault were to happen at two different point in the system, lets say at pole one you have a conductor lets say phase A goes to ground and being an un grounded system nothing happens, as this conductor now just becomes a conductor that has a Earth reference, now at pole 2 phase B goes to ground, and there just happens to not be enough current to open the closer, but now since phase A has already reference earth you now have a 2400 volt potential around both ground points and someone who unknowingly steps in to these areas could be electrocuted, but if we have a Earth reference conductor between these two points that is also bonded to all the metal of the equipment of both areas, then the first fault to ground does nothing but make a corner grounded system, and the second fault on a different phase now has a low impedance path that will make sure the closer opens.

This is kind of the same way we do in industrial plants with a 480 volt system, even though we have a ungrounded supply, we are still required to run EGC's to each and every piece of equipment and connect it to our earth grounding electrode system so that when when we have the first fault it just turns it into a corner grounded system but when the second fault happens these EGC's will provide a low impedance between the two point of the fault so the OCPD's can open the circuit, Earth can not be depended upon to serve this function.

A ground fault protection device would not serve any purpose on a ungrounded system unless it is set to open the closer if even Earth provides enough current, but then this totally defeats the reason for having an ungrounded system in the first place, so this is why I asked if they meant a ground fault detection device since this in a must so that you would know when a phase faulted to ground so an orderly shut down can take place to repair the fault.

As far as stress on the HV cables, without a reference to Earth I do not see how an ungrounded system will place any more stress on the insulation of the cable as it would in a grounded system, if the cables insulation are rated for the phase to phase voltage or more which I think the NEC will require 5kv rated cables for the 2400 volt system, even if a phase goes to ground all that happens is you now have a grounded system.

I'm not sure what your were trying to figure in item #1
(which is 2.4kV/1.732=1.38kV)
Click to expand...
as that is a WYE calculation for voltage difference between the X/0 and line to line voltage of a WYE you will not have this in a Delta unless you use a zig/zag grounding transformer which you wouldn't be using on a ungrounded system?

remember transformers are isolating so if you use a 2400v to 480v transformer and wish to keep the 480 ungrounded then you have to install a ground detection system for the 480 volt system, it is an unusual installation to have an ungrounded system fed from an ungrounded system since any fault across the 2400/480 volt windings could result in having 2400 volt potential on the lower voltage system, the same common Earth reference conductor run with the primaries I said above, and bonded to the EGCs of the 480 volt system so that there is a low impedance fault path but the secondary loads could be subject to the 2400v if the transformer had a primary to secondary fault, so I'm not sure if that would even be allowed? a 2400/208/120 with the X/0 bonded would protect from this as there would be a fault path since the 208/120 would be bonded to the common EGC of both systems, but not the 480 since it is also an ungrounded system, this transformer might have to be a shielded one so the 2400v would have to fault to this shield before it can fault to the secondary side?

I'm just not sure and maybe some of the utility or HV guys will chime in on this as I don't normally deal with the HV side but do understand the fault current paths that must be maintained for each side of the system and even across two systems to provide a low impedance fault path.
 
One other advantage to option two is that transient overvoltages will be controlled.

And an HRG can also be fitted with equipment to pulse the fault & help locate it.
 
Last edited:
Ground Detection not Fault

Ground Detection not Fault

BTW, when detecting unintentional grounds on an ungrounded system, it is called "ground detection" not ground fault detection.

I have worked on both resistance and inductance grounded systems and ungrounded systems. If ungrounded, then you need the ground detection system like you describe: use 2400V:120V PTs connected grounded wye-broken delta. You want line-line rated PTs so they don't saturate during a ground event. If I remember right, you end up with 120V for a hard corner ground. If you use an LED or high-impedance detector, you want to add a burden (load) across the broken delta to prevent neutral inversion.

A problem with delta-delta systems is that a voltage spike on the high voltage system comes across the transformer inter-winding capacitance and can cause serious overvoltages on the secondary. Make sure to have lots of surge arresters on both sides of the transformers.
 
How long is this feeder? Is it regular open wire? Any URD runs? Does it have to be ungrounded or can you have a solid uni grounded wye? I wouldn't worry to much about over voltages from an arc fault in your set up as long as lighting arresters are applied like with any other modern day system. Keep in mind that most Utilities ran their entire distribution systems at one point with 3 wire ungrounded 2.4kv or 4.8kv delta without any issues. Hold overs still exist, and some utilities in California still operate ungrounded 12kv and 16kv open wire distribution, however most today in California are uni grounded wye.

Now in any case you will need ground fault detection and indication, tripping however is optional if your system is ungrounded for service of continuity reasons. As others have said, ground faults can either be detected with P-P rated voltage transformers or a current transformer at the XO bushing of the supply transformer, XO of a zig zag transformer or in series with a resistor in a broken delta secondary bank. IF the XOs of either the supply transformer or XO of the zig zag transformer are used for detection of ground faults a properly sized resistor will need to be in series if the abilitie to operate a grounded phase or limit fault current is wanted.

There are reclosers that can automatically detect ground faults on ungrounded systems via the current bushings through sensitive earth fault and differential current feeder logic, but those are used on large/long length utility systems, most with many outgoing feeders from the sub.

In your case most likely ground faults will be detected with the ground detector input rather than the recloser bushing CTs.

Before proceeding call the recloser control supplier Cooper, ABB, SEL, GE, Viper ect and see if they have any special recommendations or low cost products/ add ones that can help you with your application. (Programming the recloser control for your application (both over current and ground fault) is a piece of cake btw, you should see selective fuse coordination or auto restoration schemes:jawdrop:).

One recommendation I would make, but its not required by the NESC if your system is being governed by this, would be to add a ground wire from the source substation all the way to the last pole while bonding the ground rods and equipment to this. It will greatly enhance system safety and greatly increase the abilities to detect ground faults as compared to using the the ground rods by themselves. Lightning protection increases 100 fold as well. However since you are in California and the public utilities commission has rules set in place on this you can not use the ground wire as a return neutral for loads as most other states. Even if rules allowed it I still would not recommended it, especially in private owner ship.


Hope this helps
 
hurk27 said:
Are you sure the utility was a ground fault protection device or do they want a ground fault detection device?

The difference with having an ungrounded system out on poles or underground is if there is no common conductor such as a MGN or EGC which I would just call it an Earth reference conductor since it would not be current carrying run with the ungrounded conductors, without this conductor there is no low impedance path if a fault were to happen at two different point in the system, lets say at pole one you have a conductor lets say phase A goes to ground and being an un grounded system nothing happens, as this conductor now just becomes a conductor that has a Earth reference, now at pole 2 phase B goes to ground, and there just happens to not be enough current to open the closer, but now since phase A has already reference earth you now have a 2400 volt potential around both ground points and someone who unknowingly steps in to these areas could be electrocuted, but if we have a Earth reference conductor between these two points that is also bonded to all the metal of the equipment of both areas, then the first fault to ground does nothing but make a corner grounded system, and the second fault on a different phase now has a low impedance path that will make sure the closer opens.

This is kind of the same way we do in industrial plants with a 480 volt system, even though we have a ungrounded supply, we are still required to run EGC's to each and every piece of equipment and connect it to our earth grounding electrode system so that when when we have the first fault it just turns it into a corner grounded system but when the second fault happens these EGC's will provide a low impedance between the two point of the fault so the OCPD's can open the circuit, Earth can not be depended upon to serve this function.

A ground fault protection device would not serve any purpose on a ungrounded system unless it is set to open the closer if even Earth provides enough current, but then this totally defeats the reason for having an ungrounded system in the first place, so this is why I asked if they meant a ground fault detection device since this in a must so that you would know when a phase faulted to ground so an orderly shut down can take place to repair the fault.

As far as stress on the HV cables, without a reference to Earth I do not see how an ungrounded system will place any more stress on the insulation of the cable as it would in a grounded system, if the cables insulation are rated for the phase to phase voltage or more which I think the NEC will require 5kv rated cables for the 2400 volt system, even if a phase goes to ground all that happens is you now have a grounded system.

I'm not sure what your were trying to figure in item #1 as that is a WYE calculation for voltage difference between the X/0 and line to line voltage of a WYE you will not have this in a Delta unless you use a zig/zag grounding transformer which you wouldn't be using on a ungrounded system?

remember transformers are isolating so if you use a 2400v to 480v transformer and wish to keep the 480 ungrounded then you have to install a ground detection system for the 480 volt system, it is an unusual installation to have an ungrounded system fed from an ungrounded system since any fault across the 2400/480 volt windings could result in having 2400 volt potential on the lower voltage system, the same common Earth reference conductor run with the primaries I said above, and bonded to the EGCs of the 480 volt system so that there is a low impedance fault path but the secondary loads could be subject to the 2400v if the transformer had a primary to secondary fault, so I'm not sure if that would even be allowed? a 2400/208/120 with the X/0 bonded would protect from this as there would be a fault path since the 208/120 would be bonded to the common EGC of both systems, but not the 480 since it is also an ungrounded system, this transformer might have to be a shielded one so the 2400v would have to fault to this shield before it can fault to the secondary side?

I'm just not sure and maybe some of the utility or HV guys will chime in on this as I don't normally deal with the HV side but do understand the fault current paths that must be maintained for each side of the system and even across two systems to provide a low impedance fault path.
Click to expand...

Ground fault indication is mandatory at all voltage levels if the system is ungrounded. In this case the OP will need ground detection because the Utility transformers is technically a separately derived source, the secondary is independent from the primary regardless of how both are grounded. You are correct that with 480 volts (and HV as well from the technical side) you need an EGC and electrode system regardless, and both faults in two separate places and different phases will result in touch potential and major currents between anything conductive. If the system is governed by the NEC the ground system is a requirement and using earth as the sole EGC is forbidden, However if the system is governed by the NESC as most utilities are than an ungrounded system or one not needing a neutral does not require a bonding EGC like in the NEC. Ground rods alone may suffice. Most older utility systems where all loads were connected phase to phase (and even a good chunk of newer construction that is like this) only 3 phase wires are run with ground rods bonding equipment. No EGC is present. Not the best practice by all means but as Ive heard before on here and through others "The NEC is concerned with shutting the power off during an emergency, the NESC is concerned with keeping the lights on during one:lol:" Sad but true.

As for Having a primary ungrounded system feeding a secondary one for the most part there is nothing wrong with this and many many install exist both in industrial and utilities. However, most engineers and the NEC recommend shielding between the primary and secondary transformer windings in cases like this, but keep in mind high to low side faults are extremely rare in modern transformers, and when one does happen the outcome is generally not a favorable one regardless of grounding (even when the XOs is solidly grounded, exceptional voltage rises may take place without tripping breakers or tripping them fast enough).


As for voltage stress in small ungrounded systems its a non issue other than a 1.73 increase in voltage on the un-faulted phases. As you mentioned because 5kv cable is being used this will not be an issue. However if the system is very, very large or has an extensive set of shielded cables (anything that increases capacitance to ground) a arcing ground fault will cause voltage spikes over the phase to phase voltage even on systems as low as 480 and 240 volts. But for the OPs short overhead line this will be a non issue.
 
mbrooke said:
Ground fault indication is mandatory at all voltage levels if the system is ungrounded. In this case the OP will need ground detection because the Utility transformers is technically a separately derived source, the secondary is independent from the primary regardless of how both are grounded. You are correct that with 480 volts (and HV as well from the technical side) you need an EGC and electrode system regardless, and both faults in two separate places and different phases will result in touch potential and major currents between anything conductive. If the system is governed by the NEC the ground system is a requirement and using earth as the sole EGC is forbidden, However if the system is governed by the NESC as most utilities are than an ungrounded system or one not needing a neutral does not require a bonding EGC like in the NEC. Ground rods alone may suffice. Most older utility systems where all loads were connected phase to phase (and even a good chunk of newer construction that is like this) only 3 phase wires are run with ground rods bonding equipment. No EGC is present. Not the best practice by all means but as Ive heard before on here and through others "The NEC is concerned with shutting the power off during an emergency, the NESC is concerned with keeping the lights on during one:lol:" Sad but true.

As for Having a primary ungrounded system feeding a secondary one for the most part there is nothing wrong with this and many many install exist both in industrial and utilities. However, most engineers and the NEC recommend shielding between the primary and secondary transformer windings in cases like this, but keep in mind high to low side faults are extremely rare in modern transformers, and when one does happen the outcome is generally not a favorable one regardless of grounding (even when the XOs is solidly grounded, exceptional voltage rises may take place without tripping breakers or tripping them fast enough).


As for voltage stress in small ungrounded systems its a non issue other than a 1.73 increase in voltage on the un-faulted phases. As you mentioned because 5kv cable is being used this will not be an issue. However if the system is very, very large or has an extensive set of shielded cables (anything that increases capacitance to ground) a arcing ground fault will cause voltage spikes over the phase to phase voltage even on systems as low as 480 and 240 volts. But for the OPs short overhead line this will be a non issue.
Click to expand...

My concern is the OP states that the Utility is feeding a 500kVA Step down transformer from 16kV to 2.4kV, if the utility point is at this 16kv transformer and or the primary metering equipment is at the secondary side of this point then the 2.4kv system is the customers and the NEC does apply, if the 2.4kv system is a ungrounded delta as he has stated, then it is a must that ground fault detection and a EGC system has to be used on this 2.4kv system, the first fault on this part of the system would not be a problem as it would just create a corner grounded system but if it happens at the supply end or at any pole between the supply end and the 480 volt transformers then without a common EGC run with the 2.4kv system conductors a second fault could pose serious as well as fatal shock hazards as well as a fire hazard on everything bonded to the 480 volt grounding system if a fault was to happen at one of the 2.4kv/480v transformers to the case of the transformer or to the EGC of the 480 volt system, this would or could put 2.4kv to earth on this 480 volt grounding system if the OCPDs (closer) were not able to clear the fault because the earth connection impedance being too high, I have seen grounded 7200 WYE system that a phase broke loose and landed on the secondary neutral which also fell to the ground that fed two houses that only had 2 ground rod each that didn't cause the OCPD of the 7200 volt system to open until the line mane who responded to the call took the primary line and made contact to a guy wire that was still bonded to the MGN which did clear the closer, so it can happen.

Even if a egc was run all the way to the supply end, the 2.4 could still fault to an earth reference without faulting to the EGC, unless they made sure that any fault that could happen would have to include a connection to this EGC, like making sure the transformer case is bonded to this EGC, even the transformer case of the 500 kva 16kv to 2.4kv of the utility, and maybe the MGN of the utility, this way there would always be a low impedance fault path for the second fault of the 2.4kv system and earth could not become the only path.

I have no problems with ungrounded systems as along as a few rules are followed they can be a vary safe system and in some cases can be safer then a grounded system, but these rules are many times over looked mainly by the linemen because of lack of understanding the fault path and being taught that earth will always clear their closers which is a very dangerous myth, there must always be a low impedance path as if the second fault ever happens on the load end and earth is the only path this load end grounding system could be at the potential of the circuit supplying the fault if the closers do not open fast enough if at all.

I one time fought with a lineman who was supposed to have supplied me with an ungrounded delta but only had 277 volt cans so he proceeded to hook them up in a WYE, which he bonded his center point to the MGN, I had to get his boss a engineer to come out and see why I was making a fuss as he only had the three phase conductors run to my riser for my lift station service I did the wiring on, being that we wanted a ungrounded system because we didn't want the lift station going down from a first fault as they were close to recreational lakes and an over flow could cost them a lots of money to clean up, I had no problem with him using the 277v cans but the WYE point must be left unbonded or the first fault would have put 277 volts on all my grounding and control cabinets to earth, I could not get him to understand how dangerous this was, after his boss showed up, and he drew out the circuit to show him how the first fault would heat up my grounding system did he finally removed the bond and included a EGC that was bonded to the transformers cases and MGN so even a fault at the pole would provide a low impedance path for the second fault, the first faults were detected by a GF detection system and sent by a SCADA to the main monitoring computer which would call the cell phone of the main supervisor over the lift stations for this large subdivision.

Funny thing was that our POCO never use to run a EGC from the MGN at the pole to the service on ungrounded systems because they didn't realize that a first fault at the pole transformer would cause the secondaries to be referanced to Earth and a second fault at the building would heat up the buildings grounding system and everything bonded to it, after this they now include the EGC in their drop and connect it to the MGN so at our end we connect it to our EGC and electrode system.
 
hurk27 said:
Are you sure the utility was a ground fault protection device or do they want a ground fault detection device?

Yes they confirmed that they want the ground fault protection.

The difference with having an ungrounded system out on poles or underground is if there is no common conductor such as a MGN or EGC which I would just call it an Earth reference conductor since it would not be current carrying run with the ungrounded conductors, without this conductor there is no low impedance path if a fault were to happen at two different point in the system, lets say at pole one you have a conductor lets say phase A goes to ground and being an un grounded system nothing happens, as this conductor now just becomes a conductor that has a Earth reference, now at pole 2 phase B goes to ground, and there just happens to not be enough current to open the closer, but now since phase A has already reference earth you now have a 2400 volt potential around both ground points and someone who unknowingly steps in to these areas could be electrocuted, but if we have a Earth reference conductor between these two points that is also bonded to all the metal of the equipment of both areas, then the first fault to ground does nothing but make a corner grounded system, and the second fault on a different phase now has a low impedance path that will make sure the closer opens.

Great example but my 2400V distribution is overhead line so there is no way to run EGC or MGN. or a ground grid to be created to let enough fault current to flow into the recloser to trip.

This is kind of the same way we do in industrial plants with a 480 volt system, even though we have a ungrounded supply, we are still required to run EGC's to each and every piece of equipment and connect it to our earth grounding electrode system so that when when we have the first fault it just turns it into a corner grounded system but when the second fault happens these EGC's will provide a low impedance between the two point of the fault so the OCPD's can open the circuit, Earth can not be depended upon to serve this function.

Lets say if i have a fault on A phase in a conduit and B phase fault in motor and these two are bonded at the termination box so it makes a close loop for the short circuit current n shld be able trip for circuit breaker?

A ground fault protection device would not serve any purpose on a ungrounded system unless it is set to open the closer if even Earth provides enough current, but then this totally defeats the reason for having an ungrounded system in the first place, so this is why I asked if they meant a ground fault detection device since this in a must so that you would know when a phase faulted to ground so an orderly shut down can take place to repair the fault.
Agreed.

As far as stress on the HV cables, without a reference to Earth I do not see how an ungrounded system will place any more stress on the insulation of the cable as it would in a grounded system, if the cables insulation are rated for the phase to phase voltage or more which I think the NEC will require 5kv rated cables for the 2400 volt system, even if a phase goes to ground all that happens is you now have a grounded system.

you are right if there are no Y connected loads on a Delta Delta system there should be any issues with the voltage stresses

I'm not sure what your were trying to figure in item #1 as that is a WYE calculation for voltage difference between the X/0 and line to line voltage of a WYE you will not have this in a Delta unless you use a zig/zag grounding transformer which you wouldn't be using on a ungrounded system?

remember transformers are isolating so if you use a 2400v to 480v transformer and wish to keep the 480 ungrounded then you have to install a ground detection system for the 480 volt system, it is an unusual installation to have an ungrounded system fed from an ungrounded system since any fault across the 2400/480 volt windings could result in having 2400 volt potential on the lower voltage system, the same common Earth reference conductor run with the primaries I said above, and bonded to the EGCs of the 480 volt system so that there is a low impedance fault path but the secondary loads could be subject to the 2400v if the transformer had a primary to secondary fault,
This is an existing system with 16kV/2.4kV feeding 2400V/480V distribution transformers located at different locations on the 2400V over head line.
I don't understand why do we need the grounding on the 480V side?
If we do need it do i have to ground it in multiple locations?


so I'm not sure if that would even be allowed? a 2400/208/120 with the X/0 bonded would protect from this as there would be a fault path since the 208/120 would be bonded to the common EGC of both systems, but not the 480 since it is also an ungrounded system, this transformer might have to be a shielded one so the 2400v would have to fault to this shield before it can fault to the secondary side?



I'm just not sure and maybe some of the utility or HV guys will chime in on this as I don't normally deal with the HV side but do understand the fault current paths that must be maintained for each side of the system and even across two systems to provide a low impedance fault path.
Click to expand...


Thanks for the analysis
 
SG-1 said:
One other advantage to option two is that transient overvoltages will be controlled.

And an HRG can also be fitted with equipment to pulse the fault & help locate it.
Click to expand...

Looks like it can only locate the fault in the above ground conduit. how a pulse technique of a HRG can locate the fault on a overhead system? like insulator failure, tree touching etc?
 
beanland said:
BTW, when detecting unintentional grounds on an ungrounded system, it is called "ground detection" not ground fault detection.

I have worked on both resistance and inductance grounded systems and ungrounded systems. If ungrounded, then you need the ground detection system like you describe: use 2400V:120V PTs connected grounded wye-broken delta. You want line-line rated PTs so they don't saturate during a ground event. If I remember right, you end up with 120V for a hard corner ground. If you use an LED or high-impedance detector, you want to add a burden (load) across the broken delta to prevent neutral inversion.

This is exactly what is referenced in one of the Basler Electric relays BE1-59N can you help me to calculate the burden load across the broken delta (resistor across the broken delta)? and what is neutral inversion?


A problem with delta-delta systems is that a voltage spike on the high voltage system comes across the transformer inter-winding capacitance? and can cause serious overvoltages on the secondary. Make sure to have lots of surge arresters on both sides of the transformers.
Click to expand...

Thanks for the recommendation
 
mbrooke said:
How long is this feeder? its is a 2400v Overhead line runs about 2000Ft long
Is it regular open wire? Yes
Any URD runs? If it is under ground residential distribution answer is NO.
Does it have to be ungrounded or can you have a solid uni grounded wye? Need to check with utility not sure why they are providing the Delta-Delta config transformer. even if it is wye connection we still have an issue to detect the ground faults on a overhead distribution right?
I wouldn't worry to much about over voltages from an arc fault in your set up as long as lighting arresters are applied like with any other modern day system. Keep in mind that most Utilities ran their entire distribution systems at one point with 3 wire ungrounded 2.4kv or 4.8kv delta without any issues. Hold overs still exist, and some utilities in California still operate ungrounded 12kv and 16kv open wire distribution, however most today in California are uni grounded wye.

got it can you tell me what is the advantage on one over other than zero sequence currents?

Now in any case you will need ground fault detection and indication, tripping however is optional if your system is ungrounded for service of continuity reasons. As others have said, ground faults can either be detected with P-P rated voltage transformers or a current transformer at the XO bushing of the supply transformer how to detect using CT on the transformers because i can install CT's on the recloser line side but how? , XO of a zig zag transformer or in series with a resistor in a broken delta secondary bank. IF the XOs of either the supply transformer or XO of the zig zag transformer are used for detection of ground faults a properly sized resistor will need to be in series if the abilitie to operate a grounded phase or limit fault current is wanted.
Installing the HRG or Zig Zag transformer in delta delta are expensive. i would rather use PT's to connect the broken delta config and trip the recloser.

There are reclosers that can automatically detect ground faults on ungrounded systems via the current bushings through sensitive earth fault and differential current feeder logic, but those are used on large/long length utility systems, most with many outgoing feeders from the sub.

In your case most likely ground faults will be detected with the ground detector input rather than the recloser bushing CTs. Agreed but curious to know how it works with CT's

Before proceeding call the recloser control supplier Cooper, ABB, SEL, GE, Viper ect and see if they have any special recommendations or low cost products/ add ones that can help you with your application. (Programming the recloser control for your application (both over current and ground fault) is a piece of cake btw, you should see selective fuse coordination or auto restoration schemes:jawdrop:). i am pretty familiar with settings on the recloser. i called one of the recloser manufacturer they simply said they cant do it.
One recommendation I would make, but its not required by the NESC if your system is being governed by this, would be to add a ground wire from the source substation all the way to the last pole while bonding the ground rods and equipment to this. It will greatly enhance system safety and greatly increase the abilities to detect ground faults as compared to using the the ground rods by themselves. Lightning protection increases 100 fold as well. However since you are in California and the public utilities commission has rules set in place on this you can not use the ground wire as a return neutral for loads as most other states. Even if rules allowed it I still would not recommended it, especially in private owner ship.

This is like running a ground grid from the service transformer to all the way to the loads. Another question can we connect the ground taps of different panels with different voltages to the same ground grid?
like 16kV, 4.16kV, 480V and 120V panels to the same ground grid?

Hope this helps
Click to expand...

Thanks for the analysis
 
Delta Leakage

Delta Leakage

If you have an ungrounded delta source and have a phase to ground fault, the leakage capacitance in the transformer and line will permit a small amount of current to flow. If you have a very sensitive current detector that can sense this very low current, you can use current detection to sense a phase-ground fault. However, in general the leakage current is so low that detection is unreliable. That is why the broken-delta PTs with a voltage relay are a good option.

The only big problem with a ground detection scheme like this is it is not sensitive to location, any ground fault anywhere causes all power to trip. Finding the ground fault takes time (divide and conquer). If you want selective fault isolation, especially in a large system, then a source of ground current is required.
 
Please see the picture for the clarity of the system..

Thanks photo.jpg

Are you saying, In this picture if i had a fault on pole2 phase A on 2.4kV system and fault on phase A or B 480V motor the high voltage will appear on the secondary side of the transformer ie., 480V side. if so how a EGC will help ?
 
Last edited:
Helps

Helps

The 2400V-480V transformers isolate potentials across the transformers; a 2400V line-ground fault has no impact on the 480V and vise-versa. But, an A-phase-ground fault on pole 1 and a C-phase-ground fault on pole 3 will cause an A-C fault through earth. This is the dangerous situation that the recloser ground detector needs to sense and open for.
 
Status
Not open for further replies.
Top