surge suppression

Status
Not open for further replies.

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
i think this has been asked before but I am going to ask it again because I am not sure anyone actually answered it.

A lot of people stress that a "good" ground is required for a surge suppressor to work, often specifying things like 2 Ohm earth impedance.. I am having a bit of trouble figuring out just how earth is even in the circuit.

Any good tutorials on this?
 
I have never understood this also, as there is no load to protect between the EGC and hot? I can understand that an SPD might live long enough to bleed off a very weak lightning strike to earth, but I doubt it, but for near field, power line hits and surges in general, I don't see the point of N to G or H to G SPD's There is no load to protect??? and the problems they can cause in AFCI and GFCI circuits can be a pain.

There are manufactures out there that agree with this view and produce SPD's that are L-L and L-N devices, but they are hard to find.
 
100517-2142 EST

There is a cloud that develops a large electrical charge distributed over a large area. There is air between the cloud and the earth. The air is primarily an insulator initially. The earth has moderately good conductivity and has a large electrical charge of opposite polarity to the cloud and equal in magnitude. Thus, a big capacitor. When the voltage from the cloud to earth gets sufficiently large the air breaks down somewhere or multiple places and there is a huge electrical arc. Large voltage gradients occur in the earth. Current flow may be 10,000 amperes and voltage differences might be thousands to million volts.

If we have a large copper (conductive) cage around a house, power lines enter in only one place, there are inductive-capacitive filters on the wires entering the house, there are no internal connections to the shield, and the shield is connected to a ground rod, then the potential of the shield may rise to 1000 s to maybe a million volts between the ground rod and some place where the lightning bolt directly hit the earth. Within the shield everything is at the same potential as the shield, thus no damage.

.
 
An externally-generated voltage surge, like the afore-mentioned lightning, might be reduced by a good earth ground, but the typical voltage-surge protector only limits voltage between conductors, and doesn't depend on an earth ground.
 
'Ground' is a typcial engineering scape-goat. That it, many engineers and product designers have been educated and trained under a continuous series of misconceptions, misnomers, and myths on the purpose and performance of 'grounding' of electrical systems.

If something doesn't doesn't work right, it must be bad ground. If something becomes damaged, it must have been bad ground. If all else fails, just ground it. Ground is the answer and the blame for most electrical engineering practices.

Fortunately, the electrical industry is getting better educated and more references, resources, and professionals are becoming more realistic and understanding of grounding and basic principles of electricity.
 
I agree that a good "ground" is often over hyped as a cureall for issues when it should not be.
In the EMC world they refer to "ground" as the sewer system of electronics.

Whether you accect a good ground as a legidimate requirement or not there are many EMC standards that require that you test all combinations of L-L, L-N, L-G, N-G when qualifying equipment against that standard.
If you do not have a protective device from the conductor to ground being tested you are very likely to fail and the device will not be approved.

Thus there are many "typical surge suppessors" with devices from L-G and N-G and these do require a low impedance ground connection in order to perform at the highest levels.

While internally (not induced by lightning) generated transients may be primarily differential rather than common mode a surge protective device from L-G can still provide a parallel path for surge currents (in a bonded system). Depending upon the relative impedance of the neutral return to the ground return this device can be more or less effective.
 
Every time I draw a sketch up of a typical residential electrical system, I see a jumper between neutral and ground at the service point. :)

I don't see how you can reduce the impedance between the neutral and ground bars any by wiring a varistor between them.

It appears to me the earth path and the neutral wire between the pole and the service point are in parallel, so perhaps the earth reduces the impedance of that part of the path somewhat, but that has no effect on the rest of the circuit.
 
This is not the best picture but at least it provides one.
See section 4.1. ( I think they may not have updated all the text on fig. J22 and that it should say Low R2)?

http://www.schneider-electric.com.au/documents/electrical-distribution/en/local/electrical-installation-guide/EIG-J-protection-voltage-surges.pdf

When talking about an external lightning event:
You are not trying to reduce the impedance between neutral and ground with a varistor. The concept is to make the parallel path to ground lower impedance than the path back to the pole ground.
 
This is not the best picture but at least it provides one.
See section 4.1. ( I think they may not have updated all the text on fig. J22 and that it should say Low R2)?

http://www.schneider-electric.com.au/documents/electrical-distribution/en/local/electrical-installation-guide/EIG-J-protection-voltage-surges.pdf

When talking about an external lightning event:
You are not trying to reduce the impedance between neutral and ground with a varistor. The concept is to make the parallel path to ground lower impedance than the path back to the pole ground.

Why is it, people want to think that all current wants to return to Earth like a baby to it's mother????

Almost all current generated in a building will not try to return to earth.

This is one of the biggest myths second only to the ground rod myth that its the holy grail to electrical safety.

Current generated with in a circuit such as transient voltage spikes are only relative to the circuit they are circulating with in, this can include branch circuits back to the panel and out on another branch circuit, but this current is only trying to return back to its source, not Earth.

Only a direct lightning strike will try to flow to earth, but not a near field event such as a near by lightning strike to a tree in your yard, this can induce current into the wiring of a house and cause damage but again this induce current will only be circulating within the affected circuits and will not be trying to return to earth.

think of it this way, look at a transformer, induce a current into the core of this transformer and the voltage out will only return to the transformer, not to earth.

Now for the direct lightning strike? well I haven't seen many SPD's stand up to one yet.
 
Why is it, people want to think that all current wants to return to Earth like a baby to it's mother????

I am not one of those people. I believe there are some kids who do not even like their parents :grin:

Current generated with in a circuit such as transient voltage spikes are only relative to the circuit they are circulating with in, this can include branch circuits back to the panel and out on another branch circuit, but this current is only trying to return back to its source, not Earth.

Yes but in a grounded and bonded system the grounding conductors can be part of the return path.


Only a direct lightning strike will try to flow to earth, but not a near field event such as a near by lightning strike to a tree in your yard, this can induce current into the wiring of a house and cause damage but again this induce current will only be circulating within the affected circuits and will not be trying to return to earth.

think of it this way, look at a transformer, induce a current into the core of this transformer and the voltage out will only return to the transformer, not to earth.

I am not knowledgeable enough on the subject to intelligently discuss all the complexities of a lightning event. I do think it would be an over simplification to compare it to a transformer and limit it to magnetic effects only. I believe there are also electrostatic, capacitive and far field effects involved as well.
We were talking about a grounded secondary and thus the grounding system will be involved in the current path to some extent. When considering that the distribution system is also earthed in multiple places there is also an earth path to some extent.
 
Yes but in a grounded and bonded system the grounding conductors can be part of the return path.

The diagram I drew of a typical residential electrical system does not support that idea to any significant extent.

The only place the GEC comes into play is in the neutral line back to the POCO transformer, in that there is a parallel path of the neutral conductor and earth for that part of the circuit. However the typical earth impedance is going to be much higher than the impedance of the neutral conductor, so it won't make much difference.

I see no way at all for there to be any utility at all to having surge suppressors at the service point wired between the neutral and ground bars since there is already a jumper there that will have about as low an impedance as you can get, and no surge suppressor is going to improve that one iota.
 
The diagram I drew of a typical residential electrical system does not support that idea to any significant extent.

The only place the GEC comes into play is in the neutral line back to the POCO transformer, in that there is a parallel path of the neutral conductor and earth for that part of the circuit. However the typical earth impedance is going to be much higher than the impedance of the neutral conductor, so it won't make much difference.

I see no way at all for there to be any utility at all to having surge suppressors at the service point wired between the neutral and ground bars since there is already a jumper there that will have about as low an impedance as you can get, and no surge suppressor is going to improve that one iota.
The response you quoted was with respect to hurk27s comments about branch circuits.
I agree things are different at the service entrance since the bonding occurs there.
 
To illustrate, i drew a diagram of a typical residential electrical system, but can't figure out how to get it to down load.
You have to get it into one of a select few document formats. PDF is one, JPEG is another. At my office, we have a printer/copier/scanner that can create PDFs. When you hit "post reply," there will be a button located below the place where you type your text, and it is labeled "manage attachments." You can browse through your hard drive, find the PDF of your sketch, and upload it from there.


If you don't have a way to scan it into a PDF, you might try taking a photograph of the sketch. This method is not likely to give us a clean image to review, but it might be good enough to convey your question more clearly.
 
OK. Lets give that a shot. I guess I missed the manage attachments stuff because it is always below the bottom of my screen where I can't see it without scrolling.

It just appears to me that the energy is dissipated by being turned to heat in the wiring, and to some extent in the varistor, but earth is just not a factor to speak of.
 
Last edited:
The only place the GEC comes into play is in the neutral line back to the POCO transformer, in that there is a parallel path of the neutral conductor and earth for that part of the circuit. However the typical earth impedance is going to be much higher than the impedance of the neutral conductor, so it won't make much difference.
Agreed. The earth ground is basically irrelevant except for a bit of rerouting of lightning.

I see no way at all for there to be any utility at all to having surge suppressors at the service point wired between the neutral and ground bars since there is already a jumper there that will have about as low an impedance as you can get, and no surge suppressor is going to improve that one iota.
Agreed again.

Neutral-to-EGC suppression only comes into play downstream, where point-of-use suppression is used, which works in conjunction with service suppression, and helps with locally-induced peaks.
 
Getting a Good Ground

Getting a Good Ground

Getting a good ground is NOT just driving a ground rod. What makes a good ground in one location may not be a good ground somewhere else. When we design a facility like an airport or treatment plant or even something as simple as a pump station, we have to design a GROUNDING SYSTEM for each particular location. We measure the conductivity of the soil, and take into account the time of year tying to estimate how much moisture is in the soil, and what our likely worst case will be. I have yet to see anywhere that got a good ground with a single ground rod. In dry locations, we will require several chemical grounds and lots of copper interconnections between them in the ground.

Grounds are not just for personnel safety; having a good ground return path is essential for the correct operation of ground fault protection systems.

As for suppressors, protection has to be line-to-line and line-to-ground, because the same storm or event may give you either or both kinds of spikes, and it may required different devices to get good protection. Here again, every location has to be considered individually.

As some stated elsewhere in this thread, you can't protect against a direct hit from lightning or sometimes even a really hot near miss, but, luckily, those are few in most areas.

By the way, grounding for the power system is not the same as grounding for lightning protection. The power grounding currents go deep; lightning currents are shallow.
 
Grounds are not just for personnel safety; having a good ground return path is essential for the correct operation of ground fault protection systems.

Can you clarify this, for systems 600 V and under, is a good ground required for the overcurrent protection device to function on a line to ground fault?
 
At the service, a ground is not required due to the main bonding jumper. IE the ground and the neutral are connected together. And you can get listed SPDs for ungrounded electrical systems....
 
Status
Not open for further replies.
Top