How concerned are you about lightning?

northerner

I am looking at the best way to protect ones solar system from surges induced by nearby lightning strikes.

Lightning has always been a concern for me with regards to not only personal safety, but in protecting ones significant investment in a solar PV system. But I'm even more concerned after experiencing a particular storm last summer. In fact, I have not ever seen so much electrical activity in my entire life, as I had with this one storm. Phase one of the storm lasted for close to 3 hours and there were static discharges on average every 3 to 4 seconds. One bolt which consistently went from cloud to ground (approaching from a particular direction) was getting closer and closer and appeared to be coming straight in line with my home. There were at least 2 strikes that were very, very close. I think one hit a nearby small substation, and there was a bright green glow that persisted after the strike. I saw that green glow from another strike during the second phase of the storm, which persisted for another 1-2 hours. It was actually the first time I had ever seen that green glow, and had me wondering if it was produced from a powerful burst, or caused secondary arcing with the electrical grid.

I know that there is little one can do about a direct strike, but with storms like this, there is a high likely hood of surges induced into a solar PV system from nearby strikes. I know that your best defense is to put in a ground that makes very good contact with the earth and locate it as close to the PV array as possible. Also, good to run a good sized conductor, with as few sharp bends as possible, down from the frames of the solar array, to this ground. Also, important to tie this earth ground into another earth ground that one may be using for equipment, etc... if you have 2 or more earth grounds in use. And of course the use of surge arrestors, both for the DC and both AC sides if applicable(ie generator input and inverter output). My system is off grid and will apply in my case. And I've read that surge capacitors are fast reacting for the spikes, but may not be necessary if using the midnite surge arrestors?

Other things that can be done is run the PV wiring in a metal conduit located a distance away from the other equipment. Put in disconnects to both the negative and positive leads for both the PV array and the generator input, which can be turned off in the event of a storm (and someone happens to be there at the time). That had me thinking that if you shut off both inputs ie from the PV array, wouldn't you want to disconnect the ground bond from the array as well (ie the tie that comes into the home to tie into the equipment ground). In other words, completely isolate it. I know that would not be code, and maybe is a little extreme, but I'm also sure that would help keep out unwanted surges in the event of a violent storm.

Any other thoughts or suggestions?

northerner

Re: How concerned are you about lightning?

I think if someone was really concerned about protecting their equipment from lightning (aka paranoid);), they could install an atmospheric electric field monitor, and rig it up to temporarily isolate their PV array, etc... and instead turn it into a giant lightning rod!:D

http://www.boltek.com/efm100.html

BB.

Re: How concerned are you about lightning?

I would start here:

BB. wrote: »

A couple threads about Lightning:

Off Grid Grounding Technique?
Another Question, this time about Lightning

Note, the above are discussions, not a do A, B, and C--and you will be "safe". There probably is no such thing with lightning. Several different techniques are discussed--and a few of those posters even have experience with lightning. :cool:

And our host's consolidated FAQ page:

www.windsun.com
Lightning Protection for PV Systems

From other past posts here, Windsun (admin/owner of NAWS), he said that most of lighting induced failures he saw were in the Inverters' AC output section.

Towards the end of this thread is a very nice discussion of proper generator grounding.

-Bill

northerner

Re: How concerned are you about lightning?

BB. wrote: »

I would start here:

-Bill

Thanks for those links Bill. Some really interesting reading!

Obtaining a good ground should not be a problem where I live. There is plenty of moisture below the ground as we are living on top of a large aquifer and there is plenty of moisture in the soils below the top soil here. I may eventually look at either disconnecting the outside lines and/or connecting them to ground temporarily, in the event of a storm.

westom

Re: How concerned are you about lightning?

northerner wrote: »

Obtaining a good ground should not be a problem where I live.

Getting a good ground is an art. For example, we thought the incoming AC wires were properly earthed via a 'whole house' protector. We did not know about a vein of graphite in earth behind the house. The best path from a cloud to distant earthborne charges was into the house ignoring our good ground. Through appliances. Then to a better earth ground behind the building. Those appliances (on or off) were a best and destructive connection.

Eventually, we had to surround the house with a buried wire. Only then was the service entrance ground good enough.

The discussion should make obvious one important fact. Protection means the connection to earth (to charges located maybe miles distant) is by far the best connection. If that wire has sharp bends, splices, or is inside metallic conduit, then that grounding is compromised.

If any wire enters the building without connecting to that same ground, then protection is compromised. Earth grounding is an art. Conductivity (what you posted) is only one of many important aspects.

NorthGuy

Re: How concerned are you about lightning?

BB. wrote: »

I would start here:

Interestting reading.

For some reason, if the PV array is 60-70ft away or more they recommend a separate ground. Why?

To my understanding, there could be a lightning strike nearby, which creates a dissipation gradient. The ground around an array may happen to be thousand volts above (or below) the ground around a house. The separate grounding puts this thousand volts potential between panel framing and cells. The distance between those is very short and easy to arc, thus destroying the system. Wouldn't it be better to ground everything to a single rod?

northerner

Re: How concerned are you about lightning?

westom wrote: »

If any wire enters the building without connecting to that same ground, then protection is compromised. Earth grounding is an art. Conductivity (what you posted) is only one of many important aspects.

Yes, I'm aware of some of the other factors westom, but I'm certainly not an expert on the subject either. But I have learned that the your best protection is making that good earth bond to start. Then it is a matter of fine tuning to direct any potential static build up to that ground, and at the same time try to keep it out of the house and away from occupants and equipment inside.

How do you go about measuring resistance to ground? I do have an equipment ground in place currently (two 3 meter ground rods spaced about 2-3 meters apart). Can one simply just measure resistance between the rods with an ohmmeter, or is there some other technique involved?

inetdog

Re: How concerned are you about lightning?

NorthGuy wrote: »

Interestting reading.

For some reason, if the PV array is 60-70ft away or more they recommend a separate ground. Why?

To my understanding, there could be a lightning strike nearby, which creates a dissipation gradient. The ground around an array may happen to be thousand volts above (or below) the ground around a house. The separate grounding puts this thousand volts potential between panel framing and cells. The distance between those is very short and easy to arc, thus destroying the system. Wouldn't it be better to ground everything to a single rod?

I think the idea is that if you tried to get primary strike lightning current, or even secondary induced current to follow a long grounding electrode conductor to the real earth ground there would be too great a chance of the current going elsewhere and damaging something else along the way.
Since the lighting strike is very high voltage as well as high current, the 10-25 ohm resistance of the ground electrode to earth connection is nothing compared to the inductance of the 70+ foot wire.
But for AC or DC faults, you still need the low ohm wire connection back to the primary grounding system to carry enough current to trip a protective device.
In some ways like the difference in the way grounds behave and are used in the case of primary high voltage circuits belonging to the POCO, compared to the low voltage circuits on your side of the transformer.

inetdog

Re: How concerned are you about lightning?

northerner wrote: »

How do you go about measuring resistance to ground? I do have an equipment ground in place currently (two 3 meter ground rods spaced about 2-3 meters apart). Can one simply just measure resistance between the rods with an ohmmeter, or is there some other technique involved?

An ohmmeter just will not do it, among other things because there is no place to connect the other lead of the ohmmeter to earth when it is carrying the testing current.

You could do it by injecting a known current between two of the electrodes and then measuring the voltages between each rod and a simple screwdriver in the earth at least 3 or 4 meters off the line between the ground electrodes. Since your meter is now measuring voltages, it will not be thrown off the effect of the current that the meter lead in the earth is carrying.
That will give you a good estimate of the resistance of each of the electrodes with one test.

You have to disconnect the wire joining the two ground rods first, of course.

BB.

Re: How concerned are you about lightning?

I think Inetdog stated the issue very clearly...

We are talking about two different sets of "energy" here.

The first is lightning which is really "RF" (radio frequency) with a maximum effective frequency around 7.5 kHz (at least one source). So all of the wiring/ground conduction, etc.) really depends on the physical nature of the conductors (impedance, not just DC resistance).

So, it turns out that a wide/flat wire with lots of conductive surfaces (i.e., a braided flat cable used for grounding in computer systems) has less "impedance" to high frequencies that the same amount of copper in one solid round wire.

A standard "round piece of wire" is useful for carrying lightning energy a few tens of feet before the "impedance" of the cable forces the lightning to take other paths to dissipate.

The other energy is the 50/60 Hz and DC current. There, a 8-6 awg cable can carry the current using Ohm's Law of I=V/R and is "of use" for 100's of feet (in engineering/safety business, the resistance of the cable run does matter. We have to make sure the cable has low enough resistance to trip a circuit breaker if there is a short circuit).

-Bill

northerner

Re: How concerned are you about lightning?

inetdog wrote: »

You could do it by injecting a known current between two of the electrodes and then measuring the voltages between each rod and a simple screwdriver in the earth at least 3 or 4 meters off the line between the ground electrodes. Since your meter is now measuring voltages, it will not be thrown off the effect of the current that the meter lead in the earth is carrying.
That will give you a good estimate of the resistance of each of the electrodes with one test.

Thanks for that inetdog. I see that there are various types of specialty meters designed to measure earth resistance. Of course the resistance can vary considerably, not only location, but with seasonal changes as well. Here's an interesting link about earth resistance.
http://www.weschler.com/_upload/sitepdfs/techref/gettingdowntoearth.pdf

westom

Re: How concerned are you about lightning?

northerner wrote: »

How do you go about measuring resistance to ground? I do have an equipment ground in place currently (two 3 meter ground rods spaced about 2-3 meters apart).

If the resistance is lower, then the conductivity during a surge will also be lower. Those ohm numbers will be different. But lower DC or near DC (60 hz) resistance also means better conductivity for a surge.

inetdog has proposed simple means of measuring (comparing) the conductivity of two soils. Expensive conductivity tester involve three or more electrodes.

Your concern involves two different concept: conductivity and equipotential. Above was about conductivity. We try to make the earth ground as conductive as possible. But sufficient conductivity can never be achieved. So also make earth beneath the building as equipotential as possible. However that also can never be sufficient. So we do both.

One demonstrated a solution on a rocky mountaintop: http://scott-inc.com/html/ufer.htm

Superior conductivity (concrete) and equipotential (wire mesh) create something equivalent to an Ufer ground.

If panels are separate from the house, then both must be treated as if separate structures - each have its own single point earth ground.

But again, it is an art. Some basic rules were introduced. One is that every incoming wire must first connect low impedance (ie 'less than 3 meters') to a common earthing electrode. That connection is made directly (a hardwire) or via a protector. From every wire inside a cable to earth.

If an earth connection is longer, then conductivity (impedance) is too high. Yes, a ground wire 20 meters long or inside a metallic conduit is, essentially, a poor conductor. For reasons that both BB and inetdog have summarized. You are not just concerned with electrode conductivity. Also worry about the connection.

More can be said if geology (soil) is defined.

NorthGuy

Re: How concerned are you about lightning?

BB. wrote: »

I think Inetdog stated the issue very clearly...

So, the idea is that the earth itself, even if it has relatively low DC resistance, is not inductive, but the long wire is much more inductive. So, for the lightning the path through the earth is better than through the wire (opposite to DC). And as a result long wire is useless for lightning protection. right?

Wouldn't then be a good idea to ground panel frames to a separate rod, but then run a wire from this rod to the house, to make sure that both grounds are at the same voltage level?

inetdog

Re: How concerned are you about lightning?

NorthGuy wrote: »

So, the idea is that the earth itself, even if it has relatively low DC resistance, is not inductive, but the long wire is much more inductive. So, for the lightning the path through the earth is better than through the wire (opposite to DC). And as a result long wire is useless for lightning protection. right?

Wouldn't then be a good idea to ground panel frames to a separate rod, but then run a wire from this rod to the house, to make sure that both grounds are at the same voltage level?

That is exactly what some experts in the field, such as John Wiles, recommend, if my memory of published papers is correct.
There is nothing that prevents you from adding as many ground electrodes as you want as long as you bond them together with an appropriately sized wire and that you do not connect the grounded conductor (neutral) in your wiring to the Equipment Grounding Conductor or Ground Electrode Conductor in more than exactly one location.

NorthGuy

Re: How concerned are you about lightning?

inetdog wrote: »

That is exactly what some experts in the field, such as John Wiles, recommend, if my memory of published papers is correct.
There is nothing that prevents you from adding as many ground electrodes as you want as long as you bond them together with an appropriately sized wire and that you do not connect the grounded conductor (neutral) in your wiring to the Equipment Grounding Conductor or Ground Electrode Conductor in more than exactly one location.

How do you size this wire appropriately?

BB.

Re: How concerned are you about lightning?

NorthGuy wrote: »

So, the idea is that the earth itself, even if it has relatively low DC resistance, is not inductive, but the long wire is much more inductive. So, for the lightning the path through the earth is better than through the wire (opposite to DC). And as a result long wire is useless for lightning protection. right?

Wouldn't then be a good idea to ground panel frames to a separate rod, but then run a wire from this rod to the house, to make sure that both grounds are at the same voltage level?

That is what I would recommend (not that I am any expert in lightning).

The "earth" has realitively high DC resistance but, relative to a standard 6 awg wire, has lower "impedance" (complex resistance) or is "less inductive" than the 6 awg wire and better able to conduct the (relatively) high frequency of the lighting induced current over longer distances.

What I try to suggest is that you don't (for example) run a DC ground from the array to the negative terminal of the battery bank (or even the neutral/ground bus of the AC Main Panel), but run the ground wire to the main ground rod for the home (which should be just outside of the main AC/power panel).

From a 60Hz/DC point of view, either grounding method is equivalent. From an RF (radio frequency) point of view, the path of the grounding is important. The main ground rod should have a "lower impedance" path to ground and will (hopefully) dissipate most of the lighting energy before it enters the home.

-Bill

westom

Re: How concerned are you about lightning?

NorthGuy wrote: »

Wouldn't then be a good idea to ground panel frames to a separate rod, but then run a wire from this rod to the house, to make sure that both grounds are at the same voltage level?

Ignore voltage. Voltage is only a symptom. Your concern is current. Anything that might stop, block, or absorb that current causes an increased voltage. Voltage increases as necessary to blow through any obstruction (including 3 kilometers of an insulator - air).

Ground may be 10 ohms resistance. And may be more conductive during a surge. IOW all those numbers are only for comparison.

Concrete is a good conductor. But wire (or rebar) in concrete makes it an every better conductor (Ufer ground). Same applies to dirt. Some of the best earth grounds are a wire looping the entire building. Wire sized by electrical safety codes. That means both better conductivity and better equipotential.

For earth, be concerned both for conductivity and equipotential. For wire to the single point ground, impedance (wire length and other factors).

Grounding the case only makes the case another possible path to earth. Or sometimes a path into the case. The case probably must be grounded for human safety. But for surges, the current must always stay outside.

Another solution is to suspend a catenary above the panels. The a best path to earth is through that catenary to earth. Panels do not conduct a current; are protected.

BB.

Re: How concerned are you about lightning?

Looping a ground wire around the base of the building is the beginning of a Faraday Cage.

Faraday cage - Wikipedia, the free encyclopedia

Add lightning cables going up the four corners of the home (and more spread around the home if it is a larger building) helps complete the Faraday cage.

In the links I provided above, as I recall, there is some information to help calculate the spacing of the "grid of cables" to help prevent lightning energy from entering the home.

The idea is that everything is at the same voltage potential at all times so that you don't have induced current/arcs inside the home.

Choosing the correct type of cable is also important. In theory, a round solid wire has the worst impedance. Something like this is much better:

Attachment not found.
An interesting sets of FAQs from the same website:

http://www.tlpinc.com/lightning-safety/faq-old-wives-tales.html
http://www.tlpinc.com/lightning-safety.html

-Bill

northerner

Re: How concerned are you about lightning?

westom wrote: »

Concrete is a good conductor. But wire (or rebar) in concrete makes it an every better conductor (Ufer ground). Same applies to dirt. Some of the best earth grounds are a wire looping the entire building. Wire sized by electrical safety codes. That means both better conductivity and better equipotential.

For earth, be concerned both for conductivity and equipotential. For wire to the single point ground, impedance (wire length and other factors).

So another way to optimize both conductivity and equipotential would be to drive ground rods all the way around your house and then connect them with a ground wire and only connect the equiment to one of the rods. That may be a little extreme in practice but very effective?

westom

Re: How concerned are you about lightning?

northerner wrote: »

So another way to optimize both conductivity and equipotential would be to drive ground rods all the way around your house and then connect them with a ground wire ...

A professional's application note (Need for Coordinated Protection) demonstrates. One structure is an antenna. For you, it might be panels. Both structures (house and panels) have their own single point earth ground. The loop also has driven electrodes. Any wire that enters either structure must first connect to single point earth ground. Every wire inside every cable. A connection made either by a protector or a direct connection.
http://www.erico.com/public/library/fep/technotes/tncr002.pdf

Connect to an earth ground best at a single point - the service entrance. A utility demonstrates good, bad, and ugly (preferred, wrong, and right) solutions.
http://www.duke-energy.com/indiana-business/products/power-quality/tech-tip-08.asp

BB.

Re: How concerned are you about lightning?

Again, I am not the expert--But you want the conductors to be "evenly" dispersed around the structure and as far apart as possible.

I.e., you want the conductors on the four corners of the building rather than two running a few feet apart down one wall.

Basically, is a version of the "skin effect" that I posted earlier.

When you have current flowing in one direction, it creates a magnetic field. This field wants to push the current conductors (i.e., the electrons) as far apart as possible (to the surface/skin of the wire, or to the outside perimeter of the home/building). You don't, for example, want the down lead from a lighting rod (or a TV/Radio antenna) on a roof peak going down the middle of the inside of the building. The lighting energy/current will naturally want to spread out from the center conductor to any low impedance conductors nearby (plumbing, AC wiring, phone wiring, etc.).

That is why some of us have an issue with the NEC that suggests the lightning/electrical ground follow the wiring down to the inverter/service panel. The solar panel wiring may enter the attic and go down the middle of the home and eventually end up at the inverter/AC main panel. If the solar array ground were to follow the same path, your are bringing lighting into the center of the home. Where is can easily find alternate paths to earth ground.

-Bill

northerner

Re: How concerned are you about lightning?

More can be said if geology (soil) is defined.

Soil in this area (and much of northern Alberta and BC) is glacial till consisting of a mix of clay and silt. There are also pockets of gravel in the area, as we are situated on a large aquifer (about 50 km's/30 miles) across

northerner

Re: How concerned are you about lightning?

BB. wrote: »

That is why some of us have an issue with the NEC that suggests the lightning/electrical ground follow the wiring down to the inverter/service panel. The solar panel wiring may enter the attic and go down the middle of the home and eventually end up at the inverter/AC main panel. If the solar array ground were to follow the same path, your are bringing lighting into the center of the home. Where is can easily find alternate paths to earth ground.

-Bill

Yes, that makes sense Bill. That is why I may look into disconnecting external lines that come into the house temporarily, when there is a storm.

westom

Re: How concerned are you about lightning?

northerner wrote: »

Soil in this area (and much of northern Alberta and BC) is glacial till consisting of a mix of clay and silt.

Clay usually makes a conductive soil. However ground that is frozen is not conductive and can become less compacted - less conductive when not frozen. To be effective, ground rods and the wire loop must be below the frost line.

inetdog

Re: How concerned are you about lightning?

westom wrote: »

Clay usually makes a conductive soil. However ground that is frozen is not conductive and can become less compacted - less conductive when not frozen. To be effective, ground rods and the wire loop must be below the frost line.

A practical matter to do this, although the NEC does not require anything of the sort. Although as long as the frost line is less than 8 feet deep, the ground rod will reach to it. :-)

northerner

Re: How concerned are you about lightning?

inetdog wrote: »

A practical matter to do this, although the NEC does not require anything of the sort. Although as long as the frost line is less than 8 feet deep, the ground rod will reach to it. :-)

Frost line here is 8 feet. I've heard that it's 10 feet about 100 km to the north and 12 feet in the High Level area. I'm not concerned about the frost line as it won't be a factor in the summer when electrical storms are prevalent, and it usually does not freeze that deep in the vicinity of a house (with a heated crawl space or basement). Code here is only 4 feet down for a footing (if heated).

NorthGuy

Re: How concerned are you about lightning?

BB. wrote: »

That is why some of us have an issue with the NEC that suggests the lightning/electrical ground follow the wiring down to the inverter/service panel. The solar panel wiring may enter the attic and go down the middle of the home and eventually end up at the inverter/AC main panel. If the solar array ground were to follow the same path, your are bringing lighting into the center of the home. Where is can easily find alternate paths to earth ground.

This is certainly helps in case of direct strike. But if the strike is nearby, the separate ground rod creates an extra path for the lightning - from ground rod for frames, arc to cells, solar wiring, negative battery bus, ground rod for the house. This path may happen to be more favorable than path through the ground from one ground rod to another. You sort of invite a lightning in the house, but through different route. Isn't that a danger too?

May be it's a good idea not to ground frames at all. When they're grounded, they sort of an antenna for lightning (just as high tower is), and may attract direct lightning. Do you think that's a good idea?

BB.

Re: How concerned are you about lightning?

There has been talk of insulated/floating solar panel frames to reduce the attraction of, and effects from lighting

If you have a building inspector--they probably will not go for it (Solar Guppy has done this, as I recall, in Florida and swears by this method).

From an engineering point of view, the solar panels and electrical wiring/components should have to meet the same insulation test requirements (1,800 VAC highpot test as I recall for 600 VAC rated components/wiring). So--Even if you insulate/isolate the solar array frames, the wiring in the conduit/routed down the metal array support hardware still should be about as susceptible to lightning energies.

If you have "floating" solar frames/mounting rails--That does raise the possibility of static charging of the metal (electric fields from thunder storms, and even the normal earth's electric fields of ~100+ Volts per Meter) or from failed electrical wiring energizing the floating frame work--So from a safety point of view, warnings and locked fencing would normally be required.

No easy/ideal solutions when working with lighting.

-Bill

NorthGuy

Re: How concerned are you about lightning?

westom wrote: »

For earth, be concerned both for conductivity and equipotential. For wire to the single point ground, impedance (wire length and other factors).

If I have a solar array that is 300 ft away, how do I make it equipotential? Or if you have several buildings hundereds feet away connected to the same electric system, how do you make them equipotential?

northerner

Re: How concerned are you about lightning?

NorthGuy wrote: »

This is certainly helps in case of direct strike. But if the strike is nearby, the separate ground rod creates an extra path for the lightning - from ground rod for frames, arc to cells, solar wiring, negative battery bus, ground rod for the house. This path may happen to be more favorable than path through the ground from one ground rod to another. You sort of invite a lightning in the house, but through different route. Isn't that a danger too?

May be it's a good idea not to ground frames at all. When they're grounded, they sort of an antenna for lightning (just as high tower is), and may attract direct lightning. Do you think that's a good idea?

I think it is a good idea. Many of the issues with surges are from nearby strikes or sudden discharges of an electric field build up, which could be picked up by the array, and should be sent to ground. Grounding the array could increase the chance of direct strike marginally, but would depend on what else exists in the immediate area. Around me, there's plenty of street lamps, power poles, trees, etc... that are all higher than my array. I also plan to do a temporary total disconnect from the solar array (power, and ground lines) in the event of a storm. This could be handled automatically by detecting an electric field, and then activating relays to do the disconnect.

Also will help to put in a large direct ground line (braided) from the solar array to earth ground. If the other lines into the house use solid wire and if you make the path indirect, (ie with sharp bends) then it will offer greater resistance to larger surges. Your surge arrestor should also have a good direct path to ground and may take care of the smaller surges.

NorthGuy

Re: How concerned are you about lightning?

BB. wrote: »

From an engineering point of view, the solar panels and electrical wiring/components should have to meet the same insulation test requirements (1,800 VAC highpot test as I recall for 600 VAC rated components/wiring). So--Even if you insulate/isolate the solar array frames, the wiring in the conduit/routed down the metal array support hardware still should be about as susceptible to lightning energies.

My panels are mounted on a wooden structure, so they're isolated from earth relatively good. Nearby trees, being 3 times taller, seem to be a better target for lightning.

I didn't have time to ground the frames because to much snow fell and it wasn't safe any more to climb up. The combiner boxes (with lightning arrestors on them) are grounded to the ground bus in DC disconnect. Still can't decide the best way for frames. Seems like no matter what you do, there are always weak points.

Then I think about installing another array, this time 300 ft away. That's even more puzzling.

BB. wrote: »

If you have "floating" solar frames/mounting rails--That does raise the possibility of static charging of the metal (electric fields from thunder storms, and even the normal earth's electric fields of ~100+ Volts per Meter) or from failed electrical wiring energizing the floating frame work--So from a safety point of view, warnings and locked fencing would normally be required.

No easy/ideal solutions when working with lighting.

So true. The more I think about it, the more I get confused.