Finding the direction of AC current flow

11bgrunt

Pragmatist
Location
TEXAS
Occupation
Electric Utility Reliability Coordinator
In the scenario where the current flows over the jumper from the Utility neutral to the fence, the question would be which one is the source and which one is the load. It's easy to assume the Utility neutral is at a higher voltage relative to the fence. But in the scientific method, both assumptions are valid (or would have to be listed as possibly true). Both would be tested, but only one would be tested and proven true.
I expect bunny trails away from the question asked. Many different trade experienced people from all over the world have the opportunity to respond.
Dan said it better. One side or the other has to be the source. Using harmonic distortion ranging from 4-8%, a phase contact is suspected. The fence being the source in this N-E connection has been seen before. The complaint is from a pool owner, so 10V is not going to be the acceptable threshold.
In the early morning, with a lightly loaded POCO line, there is a N-E of 5V. In most stray voltage investigations a loaded single phase POCO line or pipeline rectifiers are usually the first suspects to rule out. I think this has been done.
 

Xptpcrewx

Senior Member
Location
Las Vegas, Nevada, USA
Occupation
Power System Engineer
Can you explain more why line to neutral primaries have better voltage regulation? I'm not clear on this.
Think about the imaginary neutral on the remote/receiving end of the line in a delta connected bank. In a perfect world this is right in the center (perfectly balanced). With unbalanced loading the imaginary neutral point can shifts away from the center making unbalanced line-to-neutral voltages. Also with the delta connection, unbalanced loads on one phase have the effect of changing the voltage phase angles (no longer a 60/60/60 triangle). Having a neutral conductor can be thought of adding yet another voltage at the remote/receiving end of the line to better anchor the bank in place when connecting it in wye. From an lightning protection perspective, it’s best to have multi-grounded-neutral since you wouldn’t want to have the lightning strike directed all the way back to the substation to ground.


Sent from my iPhone using Tapatalk
 

gar

Senior Member
210413-1109 EDT

In my area, neighborhood, and others close by, the primary is 3 phase un-grounded delta. Probably fed from a 3 phase wye at the substation 3/4 to 1 mile away.

My center tapped pole transformer secondary is grounded by a rod at the pole. The pole transformer is 50 kVA, supplies one neighbor and me. Both my neighbor and I are grounded only at the main panel via connection to our water lines. My water line and his are probably not connected together closer than a mile away. My water meter is close to my main panel, probably 8 ft of grounding wire from the panel to my copper water line. My water line is about 125 to 150 ft of 1 1/4 copper tubing. This is a far better ground rod than any normal driven rod.

Neutral voltage drop from pole ground rod to my main panel is consisted with the drop on either hot line. Obviously load dependent.

Within my backyard the voltage drop between two screwdrivers spaced 12 ft apart, and inserted into the earth is usually less than 200 millivolts.

An 8 ft driven ground rod with 120 V applied is something in the range of 6 to 8 A of current flow, which less than 25 ohms.

.
 

__dan

Senior Member
Think about the imaginary neutral on the remote/receiving end of the line in a delta connected bank. In a perfect world this is right in the center (perfectly balanced). With unbalanced loading the imaginary neutral point can shifts away from the center making unbalanced line-to-neutral voltages. Also with the delta connection, unbalanced loads on one phase have the effect of changing the voltage phase angles (no longer a 60/60/60 triangle). Having a neutral conductor can be thought of adding yet another voltage at the remote/receiving end of the line to better anchor the bank in place when connecting it in wye. From an lightning protection perspective, it’s best to have multi-grounded-neutral since you wouldn’t want to have the lightning strike directed all the way back to the substation to ground.


Sent from my iPhone using Tapatalk
At the last transformer stepdown to the customer utilization voltage, you can either float the primary or float the secondary, referring to the bonding junper to the transformer can. I can understand wanting to bond the can to the primary neutral, for primary to can fault detection and clearing.

That would mean it's the secondary that floats, with a new reference to ground. Faulting to the earth is not going to pass (much) current with a floating secondary, but faulting line to line, line to neutral, line to EGC bonded to the neutral, will (enough to trip and clear the fault as intended).

The customer premises wiring system, if that is bonded to the Utility neutral, why that would be done so it not clear at this point. Single phase imbalanced loading within the kVA rating of the equipment should be mostly ignored up and down stream with the transformer maintaining voltage according to its turns ratio.
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
I expect bunny trails away from the question asked. Many different trade experienced people from all over the world have the opportunity to respond.
Dan said it better. One side or the other has to be the source. Using harmonic distortion ranging from 4-8%, a phase contact is suspected. The fence being the source in this N-E connection has been seen before. The complaint is from a pool owner, so 10V is not going to be the acceptable threshold.
In the early morning, with a lightly loaded POCO line, there is a N-E of 5V. In most stray voltage investigations a loaded single phase POCO line or pipeline rectifiers are usually the first suspects to rule out. I think this has been done.
I assume that the fence is situated between the POCO poles and the customer's property?

A suggestion is to drive a short ground rod on the customer side of the fence, where this rod is located between the fence and the pool. Then make voltage measurements between:
1. the POCO pole ground wire and the fence as you have done before. Call this V1.
2. the fence and the short ground rod. Call this V2.
3. the POCO pole ground wire and the short ground rod. Call this V3

One scenario:
V3 is the highest. Perhaps V1 + V2 ≅ V3. In this case the fence would likely be acting mainly as an equipotential conductor connected to the earth. In this case the source of the stray current is likely to be the POCO ground rod, with the fence establishing an equipotential voltage along its length. The voltage gradient would likely be consistently in the same direction as we move away from the POCO pole toward the fence and beyond it.

Other results may indicate that the fence is taking an active role in supplying current, perhaps from an inadvertent connection some distance away.
With your Fluke 345 it could be helpful to put the current clamp around the POCO pole ground wire, not just to measure the amount of current but also to provide a synchonization reference for getting waveforms of the above voltages. Then you can see which of these voltages are in phase, out of phase, or otherwise phase shifted relative to each other. Also to check if one or more voltages is distorted. There can be ambiguities if you're measuring RMS voltages without any phase information.
 

Xptpcrewx

Senior Member
Location
Las Vegas, Nevada, USA
Occupation
Power System Engineer
At the last transformer stepdown to the customer utilization voltage, you can either float the primary or float the secondary, referring to the bonding junper to the transformer can. I can understand wanting to bond the can to the primary neutral, for primary to can fault detection and clearing.

That would mean it's the secondary that floats, with a new reference to ground. Faulting to the earth is not going to pass (much) current with a floating secondary, but faulting line to line, line to neutral, line to EGC bonded to the neutral, will (enough to trip and clear the fault as intended).

The customer premises wiring system, if that is bonded to the Utility neutral, why that would be done so it not clear at this point. Single phase imbalanced loading within the kVA rating of the equipment should be mostly ignored up and down stream with the transformer maintaining voltage according to its turns ratio.
The main reason to bond primary and secondary neutral is to actually effectively transfer potential from one system to another in a controlled manner. You mention primary to tank fault, but how about primary to secondary crossover faults? You wouldn’t want primary voltage to transfer into the premises wiring, so the need to ground the secondary neutral (outdoors before entering into any buildings) is required. It would be a bad idea to have two grounding electrodes (one for the primary system and another for secondary system) not bonded together due to the potential gradients this would create, so both H0 and X0 are bonded at the transformer tank for overhead applications. Another mechanism to consider is voltage division from the primary to secondary to ground capacitance (associated with floating the secondary), which induces some amount of charge/overvoltage onto the secondary system via coupling between the inter-winding capacitance. By grounding the secondary winding, the floating capacitance (of the secondary winding to ground) is shunted and any overvoltage to ground disappears. With grounded secondaries, the potential difference between primary and secondary systems only appears between primary and secondary windings.

Also, I agree that this problem is irrespective of the transformer KVA ratings, but primary unbalanced load on the system actually creates voltage drops, unbalance and slight phase shifting along the line which are responsible for the NEV you are seeing. The neutral conductor from the substation to each service will be imposing different voltage magnitudes on the earth. Current will flow wherever a difference of potential exists between these points.


Sent from my iPhone using Tapatalk
 

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
The main reason to bond primary and secondary neutral is to actually effectively transfer potential from one system to another in a controlled manner.
Which can be more dangerous depending on the system impedance and clearing time of its devices.

You mention primary to tank fault, but how about primary to secondary crossover faults? You wouldn’t want primary voltage to transfer into the premises wiring, so the need to ground the secondary neutral (outdoors before entering into any buildings) is required.
So why does the NEC and IEEE still allow for ungrounded secondary services?

It would be a bad idea to have two grounding electrodes (one for the primary system and another for secondary system) not bonded together due to the potential gradients this would create, so both H0 and X0 are bonded at the transformer tank for overhead applications.
Well, does the NESC not require that where the primary is 3 wire ungroudned, the tank, LAs, ect be connected to one ground rod while the secondary is conenctced to another rod X feet away from the pole?

Another mechanism to consider is voltage division from the primary to secondary to ground capacitance (associated with floating the secondary), which induces some amount of charge/overvoltage overvoltage to ground disappears. With grounded secondaries, the potential difference between primary and secondary systems only appears between primary and secondary windings.
Right, but again this does not stop ungrounded secondary services. In theory shielding would reduce this greatly while reducing the likelyhood of primary to secondary contact.

Also, I agree that this problem is irrespective of the transformer KVA ratings, but primary unbalanced load on the system actually creates voltage
onto the secondary system via coupling between the inter-winding capacitance. By grounding the secondary winding, the floating capacitance (of the secondary winding to ground) is shunted and any drops, unbalance and slight phase shifting along the line which are responsible for the NEV you are seeing. The neutral conductor from the substation to each service will be imposing different voltage magnitudes on the earth. Current will flow wherever a difference of potential exists between these points.


Sent from my iPhone using Tapatalk

Agree, secondary imbalance create primary imbalance.

Forgive the swamping of your quotes, sent from a device which sucks.
 

11bgrunt

Pragmatist
Location
TEXAS
Occupation
Electric Utility Reliability Coordinator
I assume that the fence is situated between the POCO poles and the customer's property?

A suggestion is to drive a short ground rod on the customer side of the fence, where this rod is located between the fence and the pool. Then make voltage measurements between:
1. the POCO pole ground wire and the fence as you have done before. Call this V1.
2. the fence and the short ground rod. Call this V2.
3. the POCO pole ground wire and the short ground rod. Call this V3

One scenario:
V3 is the highest. Perhaps V1 + V2 ≅ V3. In this case the fence would likely be acting mainly as an equipotential conductor connected to the earth. In this case the source of the stray current is likely to be the POCO ground rod, with the fence establishing an equipotential voltage along its length. The voltage gradient would likely be consistently in the same direction as we move away from the POCO pole toward the fence and beyond it.

Other results may indicate that the fence is taking an active role in supplying current, perhaps from an inadvertent connection some distance away.
With your Fluke 345 it could be helpful to put the current clamp around the POCO pole ground wire, not just to measure the amount of current but also to provide a synchonization reference for getting waveforms of the above voltages. Then you can see which of these voltages are in phase, out of phase, or otherwise phase shifted relative to each other. Also to check if one or more voltages is distorted. There can be ambiguities if you're measuring RMS voltages without any phase information.
Testing continues.
The POCO single phase main line is fed from a three phase wye. The single phase line is over 5 miles long, #4 ACSR overhead primary and neutral. Today the line was cut in half, the half farther away from source was de-energized. The test site was checked before and during the clearance. There was no change in the voltage readings. The area affected is not only at the pool complaint location.
At each transformer location in this area, the N-E voltage is checked and if it is above 2V+or-, the transformer or tap is de-energized with the test meter connected.
The Fluke 345 is my favorite meter for this work. During each test, the voltage is checked, then shunted. The waveform is captured. The coil of 15 turns of loosely wound #10 is used to connect the pole ground, now the N reference, to the fence that becomes the E reference. The voltage reading falls because the N&E are at the same voltage when the jumper is connected. Amps are measured and the data/waveform screens captured. The amp scale starts at 50A so the 345 shows a low current wave.
The single phase amp spread between the POCO primary wire and neutral are far apart. A common measurement at the start of this line will show 15 on the primary and 4 on the neutral. Down the line the spread is similar. The closest so far has been 8 on the primary with 3 on the neutral.
Other long single phase lines in the area were checked, with similar numbers. The affected area is dry sand with ohm readings well above 100.
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
Testing continues.
The POCO single phase main line is fed from a three phase wye.
...
The single phase amp spread between the POCO primary wire and neutral are far apart. A common measurement at the start of this line will show 15 on the primary and 4 on the neutral. Down the line the spread is similar. The closest so far has been 8 on the primary with 3 on the neutral.
Other long single phase lines in the area were checked, with similar numbers.
At the start of the line you can't attribute a low neutral current to voltage drop along the MGN conductor (and the resulting current through the earth in parallel with the neutral) like you might further down the line.
Perhaps the deficit on neutral current at the start of the line is from currents driven into the earth by the MGNs of the single phase lines connected to the other two phases of the wye. These currents from the other phases could flow through the earth and partially balance out the neutral current present at the start of the line under consideration.
 

Xptpcrewx

Senior Member
Location
Las Vegas, Nevada, USA
Occupation
Power System Engineer
Which can be more dangerous depending on the system impedance and clearing time of its devices.



So why does the NEC and IEEE still allow for ungrounded secondary services?



Well, does the NESC not require that where the primary is 3 wire ungroudned, the tank, LAs, ect be connected to one ground rod while the secondary is conenctced to another rod X feet away from the pole?



Right, but again this does not stop ungrounded secondary services. In theory shielding would reduce this greatly while reducing the likelyhood of primary to secondary contact.




Agree, secondary imbalance create primary imbalance.

Forgive the swamping of your quotes, sent from a device which sucks.
No problem. Interesting discussion...

System impedance and clearing time don’t really have much to do with transfer potential hazards. Bonding of the neutrals is to keep the supply system and premises grounding systems as equipotential as possible during faults and lightning surges.

Nobody is recommending ungrounded power systems. It’s allowed by the NEC mostly because some existing facilities are still utilizing ungrounded systems. Also in rare circumstances (under engineering supervision and where qualified personnel are ensured to service the installation) the benefits may be viewed as outweighing the hazards. Service continuity is the main benefit cited, but ungrounded systems are plagued by safety issues, serviceability/troubleshooting difficulties and transient overvoltage.

The IEEE does not have the authority to “allow” or “prohibit” anything per se. IEEE standards are recommended practices and if you are familiar with the color book series standards, you’ll know the consensus is they do not recommend ungrounded systems (instead they suggest resistance grounded systems as an alternative which provides some benefits of both grounded and ungrounded systems).

Regarding the NESC requirement you stated, I understand rule 097’s purpose is to have dedicated grounding electrodes and grounding electrode conductors due to the possibility of one being accidentally disconnected (which appears to be a greater hazard here). Keeping these conductors isolated/apart is supposedly required because during surge arrestor operation, the ungrounded system would be momentarily grounded at that location alone and may impose dangerous voltages not otherwise present on nearby unaffected equipment. The only exception for interconnection between these grounding electrodes is when multi-grounded bus or low resistance connections (as with ground grid system in substations for example) can be ensured. Interconnection via spark-gap is also allowed. Otherwise NESC prefers dedicated grounding electrodes as obtaining low resistance connections to earth may not be guaranteed. Note: this history of rule 097 is kinda of all over the place and I am not sure I buy off on the logic as they have flip flopped on this issue in the past.


Sent from my iPhone using Tapatalk
 

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
No problem. Interesting discussion...

System impedance and clearing time don’t really have much to do with transfer potential hazards. Bonding of the neutrals is to keep the supply system and premises grounding systems as equipotential as possible during faults and lightning surges.

Nobody is recommending ungrounded power systems. It’s allowed by the NEC mostly because some existing facilities are still utilizing ungrounded systems. Also in rare circumstances (under engineering supervision and where qualified personnel are ensured to service the installation) the benefits may be viewed as outweighing the hazards. Service continuity is the main benefit cited, but ungrounded systems are plagued by safety issues, serviceability/troubleshooting difficulties and transient overvoltage.

The IEEE does not have the authority to “allow” or “prohibit” anything per se. IEEE standards are recommended practices and if you are familiar with the color book series standards, you’ll know the consensus is they do not recommend ungrounded systems (instead they suggest resistance grounded systems as an alternative which provides some benefits of both grounded and ungrounded systems).

Regarding the NESC requirement you stated, I understand rule 097’s purpose is to have dedicated grounding electrodes and grounding electrode conductors due to the possibility of one being accidentally disconnected (which appears to be a greater hazard here). Keeping these conductors isolated/apart is supposedly required because during surge arrestor operation, the ungrounded system would be momentarily grounded at that location alone and may impose dangerous voltages not otherwise present on nearby unaffected equipment. The only exception for interconnection between these grounding electrodes is when multi-grounded bus or low resistance connections (as with ground grid system in substations for example) can be ensured. Interconnection via spark-gap is also allowed. Otherwise NESC prefers dedicated grounding electrodes as obtaining low resistance connections to earth may not be guaranteed. Note: this history of rule 097 is kinda of all over the place and I am not sure I buy off on the logic as they have flip flopped on this issue in the past.


Sent from my iPhone using Tapatalk

Interesting indeed :)

Bonding the HV neutral to the LV neutral brings any voltage rise on the primary MGN into the building. System impedance and clearing time has everything to do with the severity and danger of touch potential there after in relation to remote earth. While everything in a building may be bonded such that there is no remote earth within the building (although code does not require gas pipe bonding), exterior objects like grills, power tools, spas, ect are another matter in that the person touching them is referenced to remote earth.

In theory, by have having two separate ground rods a fault to the can of the pole pigs or lightning arrestor conduction (shunting) does not transfer the surge over into the secondary neutral.

The issue with rule 097 is trying come to one size fits all, when in reality a number of system parameters determine whether or not there is greater risk in interconnecting the secondary grounding system with the primary grounding system.
 

Xptpcrewx

Senior Member
Location
Las Vegas, Nevada, USA
Occupation
Power System Engineer
A couple of points here worth clarifying...

Bonding the HV neutral to the LV neutral brings any voltage rise on the primary MGN into the building.
Same can be true for not bonding the HV neutral to the LV neutral. Both systems are grounded and are relatively in the vicinity of each other and depending on the direction(s) of current flow and where a facility is physically situated with respect to the substation, the pole, distribution equipment, communication lines, number of electrodes on the premises, and a fault, isolation can be ineffective. Thinking about things in terms of a fault going backwards to the source is an oversimplification. Reality tends to be more complex when spacial dimensions are considered and when the facility involved happens to be in the path of an external/remote fault. Also, the use of a spark-gap will impose a voltage on the secondary system neutral regardless (if it is used).

Note: Voltage rise (transfer potential) into a building is not a problem if everything is free to be raised with it.

The main difference between the MGN approach and neutral isolation is that anytime bonding is used, it is intended as a "guiding" technique directing electricity down a particular pathway, whereas isolation is a "blocking" technique to obstruct that pathway. The MGN approach is intended to create additional equipotential points throughout the ground system, but you cannot say the same for the isolated neutral. A MGN system requires good grounding design and practices for it to be effective.

My opinion is unless you actively direct the electricity where you want it (via bonding), it will go wherever it wants, which is usually unpredictable and dangerous.

System impedance and clearing time has everything to do with the severity and danger of touch potential there after in relation to remote earth.
I agree duration of shock has everything to do with lethality, but you said "system impedance" and "clearing time". How exactly does one apply clearing time to lightning surge currents, surge arrestor operation, low magnitude persistent faults (which would never clear btw), induced currents or stray NEV. While clearing time does apply to fault currents seen by protective devices, I assume the context of this discussion is not restricted to transfer potential as a result of faults alone, but rather transfer potential from any mechanism in which it may arise in general. In your defense, clearing time does play a role in the selection of protective safety grounds, but this is merely to ensure the protective safety grounds can withstand the duration/magnitude of the fault until protective devices operate. Also, consider that a facility's grounding system could be designed to sustain a fault indefinitely while only creating a negligible step/touch/mesh/transfer potential. This is why I prefer to leave clearing time out of the discussion, since no-one uses that parameter as a basis for determining whether or not to bond the primary and secondary neutrals.

While everything in a building may be bonded such that there is no remote earth within the building (although code does not require gas pipe bonding), exterior objects like grills, power tools, spas, ect are another matter in that the person touching them is referenced to remote earth.
Like the infinite bus concept,"remote earth" is only a theoretical analytical tool. Remote earth represents a mass with infinite conductance and is characterized by the inability to raise/lower its potential. In other words, it is always at zero volts. Remote earth is inaccessible so it's not possible for anyone to be in contact with it ever.

In theory, by have having two separate ground rods a fault to the can of the pole pigs or lightning arrestor conduction (shunting) does not transfer the surge over into the secondary neutral.
Both in theory and in practice this is not really true. Anything can be true or false if you oversimplify it. Keep in mind that a surge that gets clamped by a surge arrestor only limits the overvoltage to the MCOV rating, so there will still be a traveling wave being passed to rest of the system with a clipped waveform with magnitude equal to the MCOV rating. This can get transferred to the secondary of a system via transformer or capacitive action.

The issue with rule 097 is trying come to one size fits all, when in reality a number of system parameters determine whether or not there is greater risk in interconnecting the secondary grounding system with the primary grounding system.
Sure. Good grounding design, understanding and practice is always a plus.
 
Last edited:

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
A couple of points here worth clarifying...



Same can be true for not bonding the HV neutral to the LV neutral. Both systems are grounded and are relatively in the vicinity of each other and depending on the direction(s) of current flow and where a facility is physically situated with respect to the substation, the pole, distribution equipment, communication lines, number of electrodes on the premises, and a fault, isolation can be ineffective.
Sure, if both are to close.

However, the further an LV ground rod is driven from the primary HV electrode, the less potential will be transferred to the LV system during an HV fault. 6 feet is enough to make a difference in reducing transfer voltage by thousands of volts.

Direction of current does not matter.

Thinking about things in terms of a fault going backwards to the source is an oversimplification. Reality tends to be more complex when spacial dimensions are considered and when the facility involved happens to be in the path of an external/remote fault. Also, the use of a spark-gap will impose a voltage on the secondary system neutral regardless (if it is used).
Right, though in the end it boils down to potential dividers and kirchoff's voltage law.

Note: Voltage rise (transfer potential) into a building is not a problem if everything is free to be raised with it.
Inside the building itself, with everything bonded including foundation rebar and gas piping, correct.

The main difference between the MGN approach and neutral isolation is that anytime bonding is used, it is intended as a "guiding" technique directing electricity down a particular pathway, whereas isolation is a "blocking" technique to obstruct that pathway. The MGN approach is intended to create additional equipotential points throughout the ground system, but you cannot say the same for the isolated neutral. A MGN system requires good grounding design and practices for it to be effective.
Right, but at the same time when you extend an equal potential system (thus making it larger) you create a bigger sphere of energized metal outside it. Bonding while beneficial in some ways is also a double edge sword merely kicking the can down the road so to speak.

My opinion is unless you actively direct the electricity where you want it (via bonding), it will go wherever it wants, which is usually unpredictable and dangerous.
Bonding isn't so much about directing electricity but bringing everything within it to the same potential.


I agree duration of shock has everything to do with lethality, but you said "system impedance" and "clearing time". How exactly does one apply clearing time to lightning surge currents, surge arrestor operation, low magnitude persistent faults (which would never clear btw), induced currents or stray NEV.
By separating the HV and LV grounding system, you either reduce or eliminate primary over voltages from transferring into the LV system.

But in any case these events are either sub-cycle or low magnitude ie a high impedance fault will have the overwhelming voltage drop across the fault itself and not the phase & MGN themselves.

On the other hand low impedance faults involving the MGN in any way or insulators flash-overs create the most prolonged over voltages on the MGN relative to remote earth making system impedance and clearing times the central focus to reducing the probability of lethality and fire.

While clearing time does apply to fault currents seen by protective devices, I assume the context of this discussion is not restricted to transfer potential as a result of faults alone, but rather transfer potential from any mechanism in which it may arise in general. In your defense, clearing time does play a role in the selection of protective safety grounds, but this is merely to ensure the protective safety grounds can withstand the duration/magnitude of the fault until protective devices opensrate. Also, coider that a facility's grounding system could be designed to sustain a fault indefinitely while only creating a negligible step/touch/mesh/transfer potential.
Right, until you step outside of the building itself.


This is why I prefer to leave clearing time out of the discussion, since no-one uses that parameter as a basis for determining whether or not to bond the primary and secondary neutrals.
Most people do end up using clearing time as a basis of whether or not to bond the primary to secondary neutral, even the secondary system grounding type/design since primary clearing time play both an essential and profound role in human safety and protection of property and the secondary system itself.


Just one factor influencing whether the two should be connected:

1618488319640.png


Like the infinite bus concept,"remote earth" is only a theoretical analytical tool. Remote earth represents a mass with infinite conductance and is characterized by the inability to raise/lower its potential. In other words, it is always at zero volts. Remote earth is inaccessible so it's not possible for anyone to be in contact with it ever.
Well, same can be said about ohms law or the imaginary part of an impedance equation. It doesn't exist in reality. However, that doesn't stop us from using ohms law because it gives us really close numbers to the actual value- so much so we can ignore the heating coefficient of most material like nichrome (for example) allowing us to skip equations with log, feedback, ect. A value by a few %s will not make a meaningful difference sizing a circuit for a water heater, stove, baseboard ect at 208 volts.

With that said the earthing electrodes at a building will not make any significance in terms of elevating the voltage under a person's feet standing in his back yard while using an electric BBQ or touching single circuit from his or her remote shed/garage.




Both in theory and in practice this is not really true. Anything can be true or false if you oversimplify it. Keep in mind that a surge that gets clamped by a surge arrestor only limits the overvoltage to the MCOV rating, so there will still be a traveling wave being passed to rest of the system with a clipped waveform with magnitude equal to the MCOV rating. This can get transferred to the secondary of a system via transformer or capacitive action.



Sure. Good grounding design, understanding and practice is always a plus.

Right, however a capacitor will have a higher Z than a solid copper or AL bond wire/strap.[/QUOTE]
 

Xptpcrewx

Senior Member
Location
Las Vegas, Nevada, USA
Occupation
Power System Engineer
Sure, if both are to close.

However, the further an LV ground rod is driven from the primary HV electrode, the less potential will be transferred to the LV system during an HV fault. 6 feet is enough to make a difference in reducing transfer voltage by thousands of volts.

Direction of current does not matter.
The context for direction in my previous comment is the physical path and not current polarity as you might be thinking. The direction the current is traveling makes a difference if you or the isolated secondary system happens to be in the fault current path back to the other grounding electrode.

Right, though in the end it boils down to potential dividers and kirchoff's voltage law.
Only if you make simplifications to lumped elements and and make a leap of faith to circuit analysis. My main point is about spacial dimensions and potential gradients, so the concepts of potential dividers and KVL used in circuit analysis are somewhat irrelevant. All I'm saying here is you shouldn't look at 3D ground currents only as a 2D problem, since resistance elements aren't limited to finite series/parallel relationships.

Inside the building itself, with everything bonded including foundation rebar and gas piping, correct.
Good design requires it. The alternative would be to control the rate of change and magnitude of the potential gradient.

Right, but at the same time when you extend an equal potential system (thus making it larger) you create a bigger sphere of energized metal outside it. Bonding while beneficial in some ways is also a double edge sword merely kicking the can down the road so to speak.
The idea is to kick the can down the road far enough to the point where it's not a problem. As mentioned, the alternative is to control the rate of change and magnitude of the potential gradient.

Bonding isn't so much about directing electricity but bringing everything within it to the same potential.
This is only one reason for bonding, but really there are actually more. In the context of grounded systems, the single most important reason for bonding is to provide an effective ground fault current path to clear protective devices. This establishes a path to "direct" electricity to the source instead of allowing stray voltage/current to persist and take whatever unintended path it can through persons/property.

By separating the HV and LV grounding system, you either reduce or eliminate primary over voltages from transferring into the LV system.

But in any case these events are either sub-cycle or low magnitude ie a high impedance fault will have the overwhelming voltage drop across the fault itself and not the phase & MGN themselves.

On the other hand low impedance faults involving the MGN in any way or insulators flash-overs create the most prolonged over voltages on the MGN relative to remote earth making system impedance and clearing times the central focus to reducing the probability of lethality and fire.
Im still not seeing the correlation between clearing time and lightning surge currents, surge arrestor operation, low magnitude persistent faults (which would never clear), induced currents or stray NEV.

Also, not to get too tied up in semantics, but when someone says "system impedance", a power system engineer like myself tends to think Thevenin equivalent impedance of a system and not about the grounding system impedance.

Right, until you step outside of the building itself.
See comments 3 and 4 above.

Most people do end up using clearing time as a basis of whether or not to bond the primary to secondary neutral, even the secondary system grounding type/design since primary clearing time play both an essential and profound role in human safety and protection of property and the secondary system itself.
Please provide references. The only factors I am aware of for bonding the primary neutral to the secondary neutral are those outlined in the NESC. This is the governing standard.

Just one factor influencing whether the two should be connected:

View attachment 2556195
First of all this looks like an IEC standard so its not applicable. Second of all, this table looks like it's about the allowable power frequency stress voltage on insulation for equipment. Third of all, there is nothing mentioned here about using any of this as criteria to bond the primary neutral to the secondary neutral.

Well, same can be said about ohms law or the imaginary part of an impedance equation. It doesn't exist in reality. However, that doesn't stop us from using ohms law because it gives us really close numbers to the actual value- so much so we can ignore the heating coefficient of most material like nichrome (for example) allowing us to skip equations with log, feedback, ect. A value by a few %s will not make a meaningful difference sizing a circuit for a water heater, stove, baseboard ect at 208 volts.
Sorry but no. Ohms law is very real as it was derived empirically. Also, the imaginary part of impedance is not make believe (despite its name), it actually represents quadrature conduction for reactive elements in AC time- and frequency-domain analysis.

Not sure what you are referring to about with heating coefficients, log equations or feedback... None of these have anything to do directly with ohms law or complex impedance but I understand the point you are trying to make. At the end of the day, all math is theoretical, the difference is whether or not these analytical tools have sound applications and reasonable results/assumptions.

Going back to the point: Persons/property are not in contact with remote earth, but they may be thought of as being connected to it by some large equivalent impedance.

With that said the earthing electrodes at a building will not make any significance in terms of elevating the voltage under a person's feet standing in his back yard while using an electric BBQ or touching single circuit from his or her remote shed/garage.
Again, this highly depends on the physical direction the current is traveling and whether or not you or the isolated secondary system happens to be in the fault current path back to the other grounding electrode.

Right, however a capacitor will have a higher Z than a solid copper or AL bond wire/strap.
Im not talking about capacitors here. I am referring to the interwinding capacitance between primary and secondary windings. Capacitive and transformer action will still transfer potentials from the primary system to the secondary system regardless.
 

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
Im still not seeing the correlation between clearing time and lightning surge currents, surge arrestor operation, low magnitude persistent faults (which would never clear), induced currents or stray NEV.
A low magnitude (high impedance) phase to neutral fault out on the line would never need to clear from a electrocution perspective since the bulk voltage drop will be along the fault itself and not across the phase and MGN as with a low impedance fault. In such a case current across the MGN will typically be less than a few hundred amps magnitude and as such the resulting voltage drop across the MGN will only produce 10s of volts to remote earth.

{Or, another way to view it is that if a fault does not pickup ground settings in a recloser or feeder breaker, or is less than a lateral fuse, it can be likened to worse case load imbalance}


Assuming 1000 ohms hand to foot, a voltage 50 volts or less does not necessitate immediate removal:


1619103028304.png



Also, not to get too tied up in semantics, but when someone says "system impedance", a power system engineer like myself tends to think Thevenin equivalent impedance of a system and not about the grounding system impedance.
I think this is the root of our debate with any inadvertent befuddlement.

Thevenin equilvalent impedance along with the impedance of each individual component is at least half the picture when dealing with system grounding- often playing a much greater role than grounding system impedance itself.

{In simple/tangible terms is a substation with an 80ka fault current (New Jersey's 230kv system) going to have the same ground grid as a substation with 10ka short circuit current?}

While not limited to just two, these play a major roll:

1) The internal voltage drop of the substation transformer along with its primary source positive sequence impedance.

A fault nearby the substation will result large current flow causing a significant dip in voltage as measured between the X0 and X1 terminals resulting in less voltage to remote earth at the fault. Where as a fault miles out from the substation will result in less current to flow and thus relatively unchanged voltage between X0 and X1. (assume A phase fault in both these scenarios)

2) The voltage divider which forms between the phase conductor and the MGN conductor including all of its parallel paths. A 50/50 impedance will result in half the voltage to remote earth- 3,600 volts to remote earth assuming an infinite supply transformer. 40/60 = 4320 volts to remote earth and 70/30 = 2160 volts to remote earth.

These two factors primarily determine voltage to remote earth.



See comments 3 and 4 above.
Comments 3 and 4 ignore-

1) the potential divider which exists in every fault loop

2) Nothing requires that a ground grid be installed in someone's back yard. Me using an electric grill has me at remote earth.

Please provide references.
Both interconnection and none interconnection is used around the globe:

1619121089975.png

The only factors I am aware of for bonding the primary neutral to the secondary neutral are those outlined in the NESC. This is the governing standard.

First of all this looks like an IEC standard so its not applicable.
Until one realizes that the NESC and NFPA-70 hail from the IEC. The International Electrotechnical Commission controls, outlines, harmonizes and governs every single code and electrical standard on earth.

Second the physics do not change regardless of what flavor of standard or code is followed.

I'm well aware NESC has decided to forgo table 44.A1 ignoring Uf connecting the MV neutral with the LV neutral shamelessly disregarding RE, RB and RA or the disconnection time of feeder, midline or tap device.

This however does not change Uf, or the physics or outcome of a one size fits all approach.


1619120490813.png





Second of all, this table looks like it's about the allowable power frequency stress voltage on insulation for equipment.
You asked for an example of where feeder disconnection times may result in separation of MV and LV and I gave one:

1619121990861.png



Third of all, there is nothing mentioned here about using any of this as criteria to bond the primary neutral to the secondary neutral.

A dozen factors can dictate connection vs separation.

Here is another example- With RE and RB connected, consideration must be given for Uf.


1619120683153.png


1619121528825.png



And of course 442.2.3 mentions separation as a means to fulfill requirements.


Sorry but no. Ohms law is very real as it was derived empirically. Also, the imaginary part of impedance is not make believe (despite its name), it actually represents quadrature conduction for reactive elements in AC time- and frequency-domain analysis.

Not sure what you are referring to about with heating coefficients, log equations or feedback
Ohms law doesn't exist in nature. It is purely man made, only created as a comprise to simplify the extraordinarily complex

You really think a 2,400 watt heater drawing 10 amps at 240 volts will draw exactly 8.66 amps at 208 volts? Or a 100 light bulb will measure the same resistance when its cold vs when its lit? If heating coefficients did not matter I would get 144 ohms on a cold light bulb- however we both know that not to be true.

... None of these have anything to do directly with ohms law or complex impedance but I understand the point you are trying to make. At the end of the day, all math is theoretical, the difference is whether or not these analytical tools have sound applications and reasonable results/assumptions.
Right, hence why it is necessary to simplify down to lumped impedance like RE, RB, RA, ect.

Going back to the point: Persons/property are not in contact with remote earth, but they may be thought of as being connected to it by some large equivalent impedance.
Right, which does not bring the person to the same potential as the grounding system in the home.



Again, this highly depends on the physical direction the current is traveling and whether or not you or the isolated secondary system happens to be in the fault current path back to the other grounding electrode.

Duration and magnitude as seen in Table A. Interconnecting RE and RB places voltage into the premises wiring due to the potential divider formed between the hot and MGN. 3,600 volts is brought out to the grill, across a 1000 ohm body and in theory a near zero ohms between the person's shoes and the 138kv-12.47kv substation ground grid.

Im not talking about capacitors here. I am referring to the interwinding capacitance between primary and secondary windings. Capacitive and transformer action will still transfer potentials from the primary system to the secondary system regardless.
Right, but this is brief, where as a power system fault can produce current lasting many cycles.
 

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
The context for direction in my previous comment is the physical path and not current polarity as you might be thinking. The direction the current is traveling makes a difference if you or the isolated secondary system happens to be in the fault current path back to the other grounding electrode.
I'm not sure direction would be the right word then, rather I think describing the physical path itself





e series/parallel relationships.
Only if you make simplifications to lumped elements and and make a leap of faith to circuit analysis. My main point is about spacial dimensions and potential gradients, so the concepts of potential dividers and KVL used in circuit analysis are somewhat irrelevant. All I'm saying here is you shouldn't look at 3D ground currents only as a 2D problem, since resistance elements aren't limited to finit
Right, but sometimes you need to start of with a 2D problem before you can go 3D.


Good design requires it. The alternative would be to control the rate of change and magnitude of the potential gradient.
Right, not bringing voltage into the building and/or speeding up clearing of the devices.

How many buildings have their gas piping bonded? Foundation rebar? Code even lets you disconnect grounding electrodes, and if you have two they seem less concerned vs hitting all of them.

The idea is to kick the can down the road far enough to the point where it's not a problem. As mentioned, the alternative is to control the rate of change and magnitude of the potential gradient.
Well- All back yards need a ground grid with a grading ramp or sign do not use line voltage equipment outside this perimeter.



This is only one reason for bonding, but really there are actually more. In the context of grounded systems, the single most important reason for bonding is to provide an effective ground fault current path to clear protective devices. This establishes a path to "direct" electricity to the source instead of allowing stray voltage/current to persist and take whatever unintended path it can through persons/property.

Right, or as I call it source conduction (automatic disconnection of supply). However, even with a low impedance path back to the source it is impossible to achieve zero ohms between the fault point and X0, thus as a result of voltage division there will always be stray voltage and touch potential during a fault before the OCPD clears.

It is for this reason I like to make a distinction between source conduction and bonding, because while while bonding is a double edged sword sometimes making things worse either by extending the potential outward to a greater concentric circle or creating potential from its own unavoidable impedance, it is source conduction that ultimately ends the threat.
I know source conduction isn't an industry term- it is my own term- but I believe it necessary as opening an OCPD is really a 3rd entity concept that while related stands as its own objective.
 

mbrooke

Batteries Not Included
Location
United States
Occupation
Electricity
Sorry for the double reply, and the swapped order but the forum will not let me post more than 10,000 characters.
 
Top