Optimum location for panel surge protection device

Status
Not open for further replies.

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
Just got a square d qo surge protection device, to install in my main 200A panel.

Separately, we have been doing some detailed, high resolution studies looking at current balance of parallel DC sources with modest esr differences, and found that there is a position dependence on the bus bar. Apples to oranges, perhaps, but still in the same spirit...

so it got me thinking, shouldn't the surge protector be in the first two spots on the panel to avoid a modest but present resistance in the panel bus bar? Or is the reality that the line resistance (and inductance perhaps more importantly) such a driving factor that the protection will always bleed a surge first?

Perhaps that's academic, but is there a best practice?

thanks!

Manual: http://static.schneider-electric.us...es/Plug-in SPDs for Loadcenters/8226-0003.pdf
 

victor.cherkashi

Senior Member
Location
NYC, NY
From study I remember that wave go to the end of the cable and go back, on the end of the line you can get higher voltage that on surge protector

Sent from my Nexus 5 using Tapatalk
 

mgookin

Senior Member
Location
Fort Myers, FL
Page 3
NOTE:
The SPD requires two adjacent mounting spaces and must be installed

as close to neutral assembly and main circuit breaker or main lugs as possible.

Keep wire lengths as short as possible with no sharp bends.
 

petersonra

Senior Member
Location
Northern illinois
Occupation
engineer
does it make any real difference? I wonder if anyone has done any actual testing of such things. I suspect sharp bends (or maybe even not so sharp bends) have more influence on how well they work than wire length. They only come with short leads anyway. It's not like you can put them halfway across the room.

It would seem to me that the ideal place would be right at the POCO transformer if wire length is a major issue, yet that is not done.
 

kwired

Electron manager
Location
NE Nebraska
I have to ask if it makes a real difference as well.

Does the "surge" know enough to look for a protective device and to only take that path?

I realize there may be different resistance/impedance in different possible paths that will effect how much goes where. Once clamping voltage is reached the SPD does become lower resistance then when below clamping voltage as well, so more of the current associated with the surge should go through the SPD then other paths.

Surge can come from utility side or premises side as well.
 

don_resqcapt19

Moderator
Staff member
Location
Illinois
Occupation
retired electrician
does it make any real difference? I wonder if anyone has done any actual testing of such things. I suspect sharp bends (or maybe even not so sharp bends) have more influence on how well they work than wire length. They only come with short leads anyway. It's not like you can put them halfway across the room.

It would seem to me that the ideal place would be right at the POCO transformer if wire length is a major issue, yet that is not done.
There is at least one company that makes a surge protective device that plugs into the meter socket and then the meter plugs into it. Not sure if any utilities will let you install one, but some utilities will provide and install one ....for a monthly fee.
 

Tony S

Senior Member
I’m just looking in to this at the moment, I’ve decided on “as near as possible” to the incoming supply.

My reasoning is the DNO (PoCo) supply is overhead and the most vulnerable part to a transient spike. If their transformer blows up, it’s their problem. From the transformer everything is either underground or enclosed in the building. There’s nothing I can do about a direct strike on the building.

Obviously our supply is different at 433/250V so I’m going for three 275V 40KA varistors. Worth having a look http://docs-europe.electrocomponents.com/webdocs/13c0/0900766b813c0d82.pdf
I could use proprietary units but there is a hell of a price difference.

godconfidence_zps0581862e.jpg
 

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
Page 3
NOTE:
The SPD requires two adjacent mounting spaces and must be installed

as close to neutral assembly and main circuit breaker or main lugs as possible.

Keep wire lengths as short as possible with no sharp bends.

Yeah but what determines as close as possible? "Possible" implies lack of surety that it matters.

So if the panel is already populated but at the bottom, do we have to move all breakers? Or does the "as possible" part mean it doesn't matter if inconvenient? Is there a practical difference because there is a minor esr difference in a large fault current, or is the fact that the fault current is high mean it doesn't much matter? Does L di/dt apply to the inductance of the bus bar, and the length to the SPD changes inductance much? Does the entire electrical,distribution downstream turn into one big inductor if it's not bled off in the first position?

This isn't a "read the manual" question. That's exactly why I put it in the electrical calculations and engineering section, so the theory could be discussed instead...
 

Smart $

Esteemed Member
Location
Ohio
...
Re: Doesn't a church steeple with a lightning rod show a lack of confidence?

IMO no. For one, natural phenomena and events are often and quite errantly called acts of God. For example, electricity is a natural phenomena that we manipulate all the time... and we are not God.
 

kwired

Electron manager
Location
NE Nebraska
IMO no. For one, natural phenomena and events are often and quite errantly called acts of God. For example, electricity is a natural phenomena that we manipulate all the time... and we are not God.
Apparently that one went well over your head;)

Even my wife and some non electrician friends got it.
 

Smart $

Esteemed Member
Location
Ohio
Apparently that one went well over your head;)

Even my wife and some non electrician friends got it.
You know, to tell someone that something went over there head without explaining how it supposedly went over there head, is IMO mean. Then again, are you sure it went over my head or did my retort go over yours?
 
Last edited:

kwired

Electron manager
Location
NE Nebraska
You know, to tell someone that something went over there head without explaining how it supposedly went over there head, is IMO mean. Then again, are you sure it went over my head or did my retort go over yours?
If you didn't find even a hint of humor in that, then I'm sorry, AFAIK that is all it was supposed to be.
 

Smart $

Esteemed Member
Location
Ohio
If you didn't find even a hint of humor in that, then I'm sorry, AFAIK that is all it was supposed to be.
Hint of humor in your post or the pic?

If the pic, yes I see what is intended to be ironically humorous... but I more so see the error in the premise (according to my beliefs).
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
back to technology

back to technology

PSpice for transient voltage at a TV or computer

100 feet of service entrance from lightning strike, 24 inch panel bus bars, big epcos block varistor for arrestor, very short and straight leads form the arrestor to grounds.
<600 V to zap TV set for arrestor at top of panel bus bar, 8 KILOVOLTs if TV fed from top slot breakers and arrestor installed in bottom slots.
20,000 amp 1 us rise, 240 us fall time lightning strike.

Another way to interpret this from the 'doesn't matter opinions, is that if there are ADDITIONAL arrestors at the loads that can handle 3us, 100A pulses and keep voltage under 600 V, then it really does NOT make much difference. However, those secondary arrestors only need to handle < 10A vs. 100 A if the main arrestor is in the top slots.

RED voltage curve is bottom space, Green is top space. Pretty much overlap after 3 us.

Another 'however' -- if the secondary arrestors are in a plastic outlet strip vs metal, secondary arrestor inductance comes into play and voltage rises.

mov.jpg
 

kwired

Electron manager
Location
NE Nebraska
Some of us were maybe asking if it matters more so then stating that it doesn't matter.

Thanks for the information indicating that it may matter, but that there are still things to consider when designing an installation.
 

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
PSpice for transient voltage at a TV or computer

100 feet of service entrance from lightning strike, 24 inch panel bus bars, big epcos block varistor for arrestor, very short and straight leads form the arrestor to grounds.
<600 V to zap TV set for arrestor at top of panel bus bar, 8 KILOVOLTs if TV fed from top slot breakers and arrestor installed in bottom slots.
20,000 amp 1 us rise, 240 us fall time lightning strike.

Another way to interpret this from the 'doesn't matter opinions, is that if there are ADDITIONAL arrestors at the loads that can handle 3us, 100A pulses and keep voltage under 600 V, then it really does NOT make much difference. However, those secondary arrestors only need to handle < 10A vs. 100 A if the main arrestor is in the top slots.

RED voltage curve is bottom space, Green is top space. Pretty much overlap after 3 us.

Another 'however' -- if the secondary arrestors are in a plastic outlet strip vs metal, secondary arrestor inductance comes into play and voltage rises.

View attachment 14189

Wow, actual analysis of the situation, that is awesome!

curious though... When you mention stuff like short and straight leads from the arrestor to ground, the question becomes what a difference a few inches of lead or a specific curvature vs another en route to the bus at actually makes.

Im sure we could find that everything makes a difference in theory. Can spice account for the geometric factors of install somehow, other than conductor lengths and effective resistance and capacitance values?

Thanks, this sort of analysis is wonderful!
 

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
what a difference a few inches of lead or a specific curvature vs another en route to the bus at actually makes.


Notice on the analysis schematic used 22 nH for the arrestor leads, that is about inch long leads to the enclosure case - about as short as one could get.

rule of thumb that electricians could use is 1 uH per meter inductance or 25 nH per inch, straight wire.

For very detailed analysis, such as for EMP or nuclear lightning protection of missile silos, the exact wire routing is controlled and the inductance (taking into account any bends, which are very bad) is calculated using a finite element analysis program (FEA) to determine the value.

In the example given, here is another set of curves, showing the difference for arrestors in the top slot with 1" wires to the enclosure case (22nH) vs. 3 " straight wires to the ground bar and then thru the screw to the enclosure (105 nH).
One can see the large difference in the initial voltage spike from just a couple of inches of extra wire. Red is the 3" wires, blue is the short 1" wires, same 20kA lightning as the original example.

BTW, this size analysis can be done by anyone, there are 'student' versions online from Orcad (cadence) and Linear Technologies. The LTSpice from Linear Technologies is probably the easiest to learn and allows bigger circuits than the Orcad student version. Do a search, I don't have a link handy.

wf3.jpg
 
Status
Not open for further replies.
Top