no-ground whole home surge protector

[rjacobs]

Looking at doing a "whole home" surge protector in my metal shop. Have Square-D homeline panel.

Im aware I am not protecting my building from a direct lightning strike. We have a lot of brown outs, second or two power interruptions, etc...

I can get a 50ka unit that simply installs in the panel in 2 breaker spaces... but it has no ground.

The 80ka unit is installed in a knock out and has a ground and you wire to a 30a double pole breaker.

Im trying to wrap my brain around how the 50ka unit works if it has no path to ground. Does it simply sacrifice itself if the surge is to great where as the 80ka would likely still function?

 

[mm08822]

It may be a plug-on neutral type. What's the part #?

 

[rjacobs]

It may be a plug-on neutral type. What's the part #?

it is plug on neutral... HOM250PSPD

But that doesnt mean its grounded...

My neutral and ground are not bonded, at least not in the panel. I dont think they are at the disconnect nor at the main panel in my house we are pulling power from.

 

[mm08822]

it is plug on neutral... HOM250PSPD

But that doesnt mean its grounded...

My neutral and ground are not bonded, at least not in the panel. I dont think they are at the disconnect nor at the main panel in my house we are pulling power from.

Without having internal wiring diagrams of each it is hard to say.
Typically, grd is not floating as it would serve no purpose to trip a cb otherwise.
The first point of disconnect for the service, typically the main panel, should have the neutral bonded to grd in it. There should also be a connection to a grounding electrode(s).

With the first surge suppressor's neutral and ground disconnected, see if there is continuity between them. If so, one of those conductors would appear to be redundant.

 

[rjacobs]

Without having internal wiring diagrams of each it is hard to say.
Typically, grd is not floating as it would serve no purpose to trip a cb otherwise.
The first point of disconnect for the service, typically the main panel, should have the neutral bonded to grd in it. There should also be a connection to a grounding electrode(s).

With the first surge suppressor's neutral and ground disconnected, see if there is continuity between them. If so, one of those conductors would appear to be redundant.

I dont have any surge suppressors... im trying to buy one.
I will have to open my main panel and see how the electrician wired the 100a breaker and other wires , then ill have to open the disconnect
and see how thats wired. My "sub" at the shop definitely doesnt have the neutral connected/bonded directly to the ground bus.

But none of this even answers my question I guess... What good is a surge suppressor that potentially isnt connected to the ground bus??

 

[mike93lx]

I dont have any surge suppressors... im trying to buy one.
I will have to open my main panel and see how the electrician wired the 100a breaker and other wires , then ill have to open the disconnect
and see how thats wired. My "sub" at the shop definitely doesnt have the neutral connected/bonded directly to the ground bus.

But none of this even answers my question I guess... What good is a surge suppressor that potentially isnt connected to the ground bus??

The surge suppressor is there to kill itself in protection of everything else connected and it doesn't need a ground to do that

 

[rjacobs]

The surge suppressor is there to kill itself in protection of everything else connected and it doesn't need a ground to do that

I dont disagree, in general, but what I was reading that the units are supposed to be designed to take a certain number of hits at their full rating. To be able to do that, in my mind, they have to be able to dissipate that to ground and not just absorb it all.

 

[mike93lx]

I dont disagree, in general, but what I was reading that the units are supposed to be designed to take a certain number of hits at their full rating. To be able to do that, in my mind, they have to be able to dissipate that to ground and not just absorb it all.

I dunno. I'd trust that it's designed properly, though

 

[MovingAlong]

I dont have any surge suppressors... im trying to buy one.
I will have to open my main panel and see how the electrician wired the 100a breaker and other wires , then ill have to open the disconnect
and see how thats wired. My "sub" at the shop definitely doesnt have the neutral connected/bonded directly to the ground bus.

But none of this even answers my question I guess... What good is a surge suppressor that potentially isnt connected to the ground bus??

Neutral and ground typically bond at the main panel, they should not be bonded downstream anywhere.

Could you post a link to the suppressors you're looking at? I'm sure the manufacturer offers some insights as to "What good" their suppressors are..

 

[mm08822]

I dont have any surge suppressors... im trying to buy one.
I will have to open my main panel and see how the electrician wired the 100a breaker and other wires , then ill have to open the disconnect
and see how thats wired. My "sub" at the shop definitely doesnt have the neutral connected/bonded directly to the ground bus.

But none of this even answers my question I guess... What good is a surge suppressor that potentially isnt connected to the ground bus??

I thought you had the 80ka. So you can't do that test I suggested.

A sub panel fed from a 4-wire feeder should NOT have a bond screw in it.
The neutral and ground combine back at the bonding screw, so effectively they are at the same potential.

The surge suppressor has at least 3 MOV's in it. L1 - L2, L1 - N, L2 - N. Each MOV clamps the voltage across it to a specified value. The mov conducts and absorbs energy above that threshold. If the surge is too high or too long, it can burn out and cease further protection.

 

[rjacobs]

I thought you had the 80ka. So you can't do that test I suggested.

A sub panel fed from a 4-wire feeder should NOT have a bond screw in it.
The neutral and ground combine back at the bonding screw, so effectively they are at the same potential.

I dont have the 80ka unit, but I am looking at and the 50ka version.

I am 100% positive my sub-panel is not bonded. We removed the screw.

I need to open up my panel we are pulling from to see how its setup.

 

[rjacobs]

I dunno. I'd trust that it's designed properly, though

Thats the thing, I dont think that they would sell something that didnt work. From what I can tell NEC 2020 and newer require these on every panel now. I would "assume" that if thats the case, then something that "doesnt work" wouldnt be allowed. But my brain has a hard time wrapping around not having a direct ground hookup.

 

[mm08822]

Thats the thing, I dont think that they would sell something that didnt work. From what I can tell NEC 2020 and newer require these on every panel now. I would "assume" that if thats the case, then something that "doesnt work" wouldnt be allowed. But my brain has a hard time wrapping around not having a direct ground hookup.

The MOVs within the SP are protecting the devices wired across the circuit conductors.

Earth connections are used to dissipate static electricity between the air and earth surface. In the case where your building is in between statically charged air and earth, the ground conductors provides a low resistance path to dissipate the energy through the thing (building) in between. This is where the grounding electrode(s) come into play.

 

[mike93lx]

Thats the thing, I dont think that they would sell something that didnt work. From what I can tell NEC 2020 and newer require these on every panel now. I would "assume" that if thats the case, then something that "doesnt work" wouldnt be allowed. But my brain has a hard time wrapping around not having a direct ground hookup.

Tells me that the ground connection isn't required. I bet the external ones have a ground only to satisfy a requirement for one due to being connected to a breaker, or something like that

 

[rlitman]

The surge suppressor is there to kill itself in protection of everything else connected and it doesn't need a ground to do that

Not that simple. In the case of a PoN panel, the neutral is a better ground than ground.

I dont disagree, in general, but what I was reading that the units are supposed to be designed to take a certain number of hits at their full rating. To be able to do that, in my mind, they have to be able to dissipate that to ground and not just absorb it all.

Kind of. The MOV just shorts across itself (each individual MOV has only two wires), absorbing whatever energy is released while that happens. If you have an MOV bridging line and ground, it will short line to ground. If it is bridging line and neutral, then it shorts them. and if it is between line and line, you get the point.

What it is doing, is clamping the voltage between its two tails above a certain point (around 400V). Once the voltage rises above that, it shorts and takes a hit. Your three wire PoN SPD example will have MOVs for L-L and L-N, and it is assuming the impedance between neutral and earth is low enough to just connect to neutral. In a PoN panel, that's a reasonable assumption, since the neutral wire (which is directly bonded to ground somewhere upstream) will be sized the same as the line wires, while the ground wire is generally much smaller, and we want to use the lowest impedance path to ground.

It would take a very nearby lightning strike to see something over a 400V potential between neutral and ground (considering they're bonded relatively nearby), where you could even use L-G protection. If you had a UL96 lightning protection system and took a direct hit, you might experience that kind of ground swell voltage.

 

[rjacobs]

Not that simple. In the case of a PoN panel, the neutral is a better ground than ground.

Kind of. The MOV just shorts across itself (each individual MOV has only two wires), absorbing whatever energy is released while that happens. If you have an MOV bridging line and ground, it will short line to ground. If it is bridging line and neutral, then it shorts them. and if it is between line and line, you get the point.

What it is doing, is clamping the voltage between its two tails above a certain point (around 400V). Once the voltage rises above that, it shorts and takes a hit. Your three wire PoN SPD example will have MOVs for L-L and L-N, and it is assuming the impedance between neutral and earth is low enough to just connect to neutral. In a PoN panel, that's a reasonable assumption, since the neutral wire (which is directly bonded to ground somewhere upstream) will be sized the same as the line wires, while the ground wire is generally much smaller, and we want to use the lowest impedance path to ground.

It would take a very nearby lightning strike to see something over a 400V potential between neutral and ground (considering they're bonded relatively nearby), where you could even use L-G protection. If you had a UL96 lightning protection system and took a direct hit, you might experience that kind of ground swell voltage.

So it sounds like your opinion is the 50ka that plugs straight into the PON box with no actual ground wire is likely good enough for my setup? My shop doesnt have a ton of stuff and most isnt plugged in constantly. My Unifi switch is plugged in constantly as well as my Lutron Caseta hub

Would I be "better" off just getting a good quality single outlet surge protector? Ive got one of these on my home theater projector outlet in the house and I feel it works pretty well. It seems like it will trip the gfci breaker if there is any surge... Not sure if thats by design of the surge protector or if the GFCI doesnt like it.

https://www.amazon.com/Tripp-Lite-Protector-INSURANCE-ISOBLOK2-0/dp/B00006B81D/?tag=atomicindus08-20

That is rated at 1410 joules. The 50ka panel mount is only rated at 720 joules. So "in theory" the single outlet one is "better"... but hard to believe the rating numbers.

 

[kngelv]

The one with the plug on neutral is designed for a service entrance panel not for sub-panels. Your SEP should have the ground and neutral bars bonded together. This is not the case in a sub-panel as it's a code violation. I just went through this same thing. I put the 80kA one in my SEP on a two pole breaker in the first two slots. I then ran the neutral and the ground to their respective bars. The breaker hanging in space is the one that feeds my sub-panel. I did this because having it in the SEP protects both the house and the garage. Yours will still work but I just liked having a direct ground attachment which is why I went the way I did. I'd suggest doing it that way.

James

 

[wyliesdiesels]

Im aware I am not protecting my building from a direct lightning strike. We have a lot of brown outs, second or two power interruptions, etc...

not sure what this has to do with surge arrestors as they will not help with these issues

 

[wyliesdiesels]

My neutral and ground are not bonded, at least not in the panel. I dont think they are at the disconnect nor at the main panel in my house we are pulling power from.

neutral is required to be bonded in the first means of disconnect or main service panel.... whichever comes right after the meter.

without this bond, breakers will NOT trip when there is a ground fault creating a dangerous potential for shock. I have seen what happens when plumbing isnt bonded. the entire plumbing in the house (including gas lines and appliances) becomes energized and shocks users.

 

[wyliesdiesels]

It seems like it will trip the gfci breaker if there is any surge... Not sure if thats by design of the surge protector or if the GFCI doesnt like it.

A GFCI is not a surge protector. GFCI only cares about a ground fault where more than 5ma is leaking outside the circuit (line to neutral and line to line). likewise, a breaker will not trip either if there is a surge. breakers trip on short circuits and overcurrents, both dealing with amperage/current NOT voltage. a breaker isnt there to check voltage or trip if it gets too high...