Do my panels need Double Pole breakers? If so, why?

VanHausen

So for some years now, I've had all my parallel panels on my van, protected against a short circuit dump by running the positive of each through their own 10A breaker, just before the controller

Today, I had someone inform me however, that this is a No No - all floating panels (ie negative not grounded) in strings of 3 or more must have Double Pole breakers (ie on + and - ).

When I pressed him on why, got the usual "Er....um...safety...best practice...regulations say so", but nothing in terms of a technical explanation as to WHY you need both lines cut?

I've been hunting all day and can't find anything conclusive, jut a lot of equally vague references about how this is the done thing with 'ungrounded' panels. 

Can anyone explain the actual hazard mechanism behind having only a single breaker per string, on the positive line? 

Thanks!

BB.

I will try to make this short and sweet--Please feel free to ask more questions if you want. Grounding is one of the most complex set of questions that we get around here.

If your solar battery bank is grounded to the van frame, then you do not need to have a double pole breaker on the solar array (or the battery output). Just single pole breakers and fuses are fine.

Why? When you negative ground your battery bank, there are no, or almost no, solar charge controllers that have isolation from solar array to the battery bank. So, negative grounding, solar panels are not floating, but ground referenced.

When you ground reference a power system (DC power, North American Ground bonded white/neutral conductor), the ground referenced conductor cannot get above "zero volts" with respect to frame/safety/green wire ground. So, there is no reason to put a fuse/breaker on the ground referenced conductors.

The reason, for example, we use double pole breakers on 240 VAC split phase North American circuits, is that both L1 and L2 (Black and Red) conductors are 120 VAC with respect to ground (and 240 VAC from L1 to L2). The pole transformer is a center tap configuration (L1-Neutral center tap-L3) connection. And the Neutral is bonded (typically in the main panel of the home), so only L1 and/or L2 need circuit breakers.

For "floating power system". In theory, they are very save. If you short one of the output conductors (whether 12 VDC or 120 VAC), since there is no connection to the frame/earth ground elsewhere, there is no electrical return path for you to get shocked (unless you touch bother output wires).

The reason you are supposed to (and few people do) have to put a breaker on both +/- (if DC) output for a floating supply is if you get two short circuits to ground. Say your first ground is a 20 AWG gave wire in an LED ceiling lamp where the negative wire gets cut and shorted to frame ground on the fan.

Now, you get a second short circuit from Battery positive bus (and you have a 100 Amp breaker/fuse only on the positive bus and no breaker/fuse on the "return or negative wiring anywhere). The short circuit path goes from Battery + (100 amp breaker) to the van sheet metal, to the return/negative 20 AWG wire in the LED fixture, and follows the 20 AWG wire back to the negative battery bus and back to the battery.

Your 20 AWG wire is now your "fuse" in the circuit (and a very real threat of fire). If you had a floating system with double pole fuses or breakers, then you would have one on each wire that leaves the battery + and - buses. The overcurrent protection device would be rated to protect the "down stream" wiring from over rated current (say a 2 amp fuse on 20 AWG +/- wire for your LED fixture).

That is why two fuses/breakers on "floating power systems" and only one breaker/fuse is required for ground referenced systems.

There are floating power systems out there... A common one is to use an isolated power transformer during manufacturing of a 120 VAC system so that if there is a short somewhere during manufacturing (and before final hipot test), the assembly person is not electrocuted. And the isolation transformer is tested once a shift or once a month type thing to make sure it is OK.

For large ships, they have "isolated" AC power systems (floating) so that the hull does not become a conductor during a short circuit (safety). A simple method (probably not used anymore, very old) is to connect a small 120 VAC lamp from L1 to ground and a second lamp from L2 to ground. If the AC circuit is OK, each lamp glows at 1/2 voltage. If L1 or L2 become shorted to ground, the shorted leg "lamp" goes out, the the second leg lamp goes bright.

So, for small portable floating power supplies (small solar panels and battery), generally, people, at best, have a single fuse/breaker on the + output. There is no way for a single short circuit to create a path where the unground return wire can get more current than it is designed for (i.e., 15 amp fuse, 14 AWG wire + out to load and - return back to battery).

For larger floating power systems, where you may have large conductors (50+ amps going to a solar charge controller, 150 amps to an AC inverter, 2 amps to an LED lamp), you should have properly rated double pole breakers/fuses for each pair of cables that leave the battery bus.

There are other issues why and how grounding is done in AC and DC power systems (lightning, static discharge, reducing the possibility of electrocution, cathodic protection systems for buried pipe/boats/telecom/natural gas piping/etc., fluorescent lighting, spark ignition for propane/natural gas burners, etc.)...

And there are reasons to (possibly) put fuse/breaker per solar panel on your van (in positive lead only required for DC ground bonded systems)--Typically with 3 or more panels in series (usually 100+ Watt per panel size) that help prevent solar wiring overheating if one panel gets shorted and the other 2+ panels feed excessive current to the shorted panel). But this may not even be an issue for a small solar power system.

-Bill

BB.

Van,

Just keep everything in one thread for your questions... It will make things less confusing:

So - apologies for this being a double post. I put it up on "advanced technical" forum but then realized this is probably the more appropriate place...maybe? Anyway I'll leave it to the mods to put it wherever they feel best.

So for some years now, I've had all my parallel panels on my van, protected against a short circuit dump by running the positive of each through their own 10A breaker, just before the controller

Today, I had someone inform me however, that this is a No No - all floating panels (ie negative not grounded) in strings of 3 or more must have Double Pole breakers (ie on + and - ).

When I pressed him on why, got the usual "Er....um...safety...best practice...regulations say so", but nothing in terms of a technical explanation as to WHY you need both lines cut?

I've been hunting all day and can't find anything conclusive, jut a lot of equally vague references about how this is the done thing with 'ungrounded' panels. 

Can anyone explain the actual hazard mechanism behind having only a single breaker per string, on the positive line? 

I think I have answered your basic question(s)... Feel free to ask more.

-Bill

VanHausen

https://forum.solar-electric.com/discussion/comment/396500#Comment_396500

Bill

Wow. Thaks muchly for such a quality reply. Appreciate the time it took to write that all out.

I understand the point of two-pole breakers with multi-phase supplies, and in the DC double fault scenario you describe. What I'm having difficulty seeing the purpose of them in the PV context.

With my current setup, I have X panel strings in parallel, with each string line set coming across the roof and into the cab, where each individual string positive enters a breaker, and then collects to one line, and feeds immediately into the controller's pos port, all inside a few inches of insulated box. The negative lines simply collect separately, then also enter the controller's neg port. Very simple parallel circuit, zero prospect of accidental short to anything.  

Now lets take the rough equivalent of your fault scenario - one of the panel lines develops a positive short to chassis/ground, in the only place it can: between the breaker and panel, ie along the roof somewhere. Chassis is now hot.

Any negative short now, be it on the same string or a different one, would complete the back-feed circuit now. But that back-feed current MUST pass through the strings +ive CB to get to the the chassis short and complete the circuit. Thus tripping the positive breaker.

 In short (hah), with the breaker installed at the controller, any fault-to-ground scenario on a +ive string line MUST trip that lines breaker...right? 

Basically, only way I can see a hazard eventuating needing a breaker on the negative line, is if you have some sort of short BEFORE the positive OCPD ie the positive fuse is effectively bypassed. In your scenario, that was on the bus, before the appliance fuses. In the case of PV arrays, that would have to be a short BEFORE the individual string breakers.

But then how would that happen? The breakers are, as a matter of procedure, always installed close to the controller, just like your main fuse sits right at the battery terminal? Or at least that's what I was always taught - OCPD as close to the source or the junction as possible, specifically so it can't be bypassed.

Or am I missing something?

BB.

If you are talking about Grid Tied Solar inverters... The old inverters had isolation transformers between the PV input and the 240 VAC output. And did not have a ~1 amp fuse between solar array ground and the green wire safety ground (one form of NEC approved "arc fault" prevention--If the fuse was blown, it signaled to the GT controller to shutdown until the short was cleared and the fuse replaced).

The fuse based "arc fault" detection is also used for larger/modern installations of off grid power systems too. Because of the use of the 1 amp fuse (as a detection circuit) or a 1 amp breaker ganged with two pole 63 amp (or whatever amp rating) breaker for the array as a mechanical disconnect to stop the current flow from the array to the charge controller input (1 amp fault current between negative array bus and earth safety ground, trips 1 amp breaker which gang trips the other two solar breakers and cuts off the DC power, and in theory, cuts the current that would flow through a DC short to ground/which could turn into an arcing condition).

I have a huge issue with the (slapped together using UL approved fuses/breakers) NEC arc fault prevention system to stop the real problem of arc faults. Because many of the systems use the 1 amp fuse/detection breaker between the array negative and safety ground, it turns a ground referenced system into floating power system--Which does need two pole breakers for safety.

It is so bad, I wrote a white paper to address the issue (but it was ignored by the powers that be). If you want to see my take on why DC Ground Fault detection via the NEC directive is highly unsafe, see here:

https://forum.solar-electric.com/discussion/9345/system-grounding/p1

http://midniteftp.com/forum/index.php?topic=142.0 (note that many of the old links to the Midnite NEC DC GFI discussion are "bad" (board reorganized). This is the "good link"

There are now electronic Arc Fault detection now (basically a receiver that listens to the RF frequencies on the electric circuit and if it "hears" an arc fault, the system shuts down--Which can have its own issues--universal "brushed" motors like small drills and vacuum cleaners make lots of "arch fault like" noise and can false trip a system).

A simple example of DC vs AC current and DC power to sustain arcs "very nicely" (and why you need to "respect" the DC rating of switches breakers, and fuses on DC circuits, and why AC only rated devices on DC can be dangerous--And why DC rated devices are usually bigger than similar rated AC devices. And why on AC+DC rated devices, the DC ratings for voltage and current are usually much lower than the AC version):

https://www.youtube.com/watch?v=Zez2r1RPpWY

And, to finish the discussion--There are newer GT Inverter technologies that do not use isolation transformers between solar array and AC mains (transformers are big, heavy, lots of costly copper, and many are hand wound toroid shaped with lots of labor).

That, and adding the 1 amp return to green wire detection fuse in many systems, one bus (tyically negative, but years ago one solar panel mfg. had a requirement for positive grounded arrays for operational reasons), the newer (present production) GT solar power inverters make the array "floating" if the 1 amp fuse blows, or both +/- legs "hot with respect to ground" because of the direct connection (no longer transformer isolated) to 240 VAC power (in North America, AC mains are 120/240 VAC Split Phase power--which both L1 and L2 240 VAC legs are hot at 120 VAC each to earth ground), those arrays are now "floating" and would need double pole breakers for safety and full shutdown.

For central GT inverter systems, there are now also roof top disconnects required for firefighter safety (shut down the array so that firefighters do not cut "hot" large high voltage/high current solar power connection--The rooftop disconnect makes the array "safer") so that addresses another set of issues (I have not looked in detail at the fault scenarios of roof top disconnects--After the NEC DC ARC FAULT DISCONNECT safety fiasco--It just was not another subject that I wanted to delve into--Whether it was "safe" or "not").

So, this migrated from a small 3 panel van mounted off grid power system into kWatt and larger grid tied solar arrays mounted on home roofs...

There is a lot here... I am not quite sure where you want to go with this. The threads (and paper) I wrote on NEC DC Ground Fault Disconnect systems--There is a bunch of details there too.

Note that NEC (and all other codes) require hard bolting (or even welding) of chassis grounds to green wire safety grounds/ground rods.

And even on simple metal chassis power supplies and such, the green wire safety wire has its own ground stud to the case or is "double nutted (the safety ground is the last nut to remove from the chassis ground). And why the 1 amp detection fuse/breaker really frosted me.

Solar array fires (and arc fault versions of the faults) are a big problem. Working with multiple power sources that are solar powered makes safe detection and shutdown difficult (in general, N+1 redundant power systems in computers are a pain in computers to large utility systems that have hundreds or thousands of generators distributed around multiple states).

There are lots of videos out there that show DC power on an AC rated circuit breakers, and even the failed wiring (like a connection that went "open") just sitting there smoking and spitting flames while an arc just sits there happily arcing (and why DC arc welders are so useful).

Grounding and safety are highly interrelated and a complex subject with conscious and unconscious tradeoffs.

-Bill

DZL

BB. said:

I will try to make this short and sweet--Please feel free to ask more questions if you want. Grounding is one of the most complex set of questions that we get around here.

If your solar battery bank is grounded to the van frame, then you do not need to have a double pole breaker on the solar array (or the battery output). Just single pole breakers and fuses are fine.

Why? When you negative ground your battery bank, there are no, or almost no, solar charge controllers that have isolation from solar array to the battery bank. So, negative grounding, solar panels are not floating, but ground referenced.

When you ground reference a power system (DC power, North American Ground bonded white/neutral conductor), the ground referenced conductor cannot get above "zero volts" with respect to frame/safety/green wire ground. So, there is no reason to put a fuse/breaker on the ground referenced conductors.

The reason, for example, we use double pole breakers on 240 VAC split phase North American circuits, is that both L1 and L2 (Black and Red) conductors are 120 VAC with respect to ground (and 240 VAC from L1 to L2). The pole transformer is a center tap configuration (L1-Neutral center tap-L3) connection. And the Neutral is bonded (typically in the main panel of the home), so only L1 and/or L2 need circuit breakers.

For "floating power system". In theory, they are very save. If you short one of the output conductors (whether 12 VDC or 120 VAC), since there is no connection to the frame/earth ground elsewhere, there is no electrical return path for you to get shocked (unless you touch bother output wires).

The reason you are supposed to (and few people do) have to put a breaker on both +/- (if DC) output for a floating supply is if you get two short circuits to ground. Say your first ground is a 20 AWG gave wire in an LED ceiling lamp where the negative wire gets cut and shorted to frame ground on the fan.

Now, you get a second short circuit from Battery positive bus (and you have a 100 Amp breaker/fuse only on the positive bus and no breaker/fuse on the "return or negative wiring anywhere). The short circuit path goes from Battery + (100 amp breaker) to the van sheet metal, to the return/negative 20 AWG wire in the LED fixture, and follows the 20 AWG wire back to the negative battery bus and back to the battery.

Your 20 AWG wire is now your "fuse" in the circuit (and a very real threat of fire). If you had a floating system with double pole fuses or breakers, then you would have one on each wire that leaves the battery + and - buses. The overcurrent protection device would be rated to protect the "down stream" wiring from over rated current (say a 2 amp fuse on 20 AWG +/- wire for your LED fixture).

That is why two fuses/breakers on "floating power systems" and only one breaker/fuse is required for ground referenced systems.

There are floating power systems out there... A common one is to use an isolated power transformer during manufacturing of a 120 VAC system so that if there is a short somewhere during manufacturing (and before final hipot test), the assembly person is not electrocuted. And the isolation transformer is tested once a shift or once a month type thing to make sure it is OK.

For large ships, they have "isolated" AC power systems (floating) so that the hull does not become a conductor during a short circuit (safety). A simple method (probably not used anymore, very old) is to connect a small 120 VAC lamp from L1 to ground and a second lamp from L2 to ground. If the AC circuit is OK, each lamp glows at 1/2 voltage. If L1 or L2 become shorted to ground, the shorted leg "lamp" goes out, the the second leg lamp goes bright.

So, for small portable floating power supplies (small solar panels and battery), generally, people, at best, have a single fuse/breaker on the + output. There is no way for a single short circuit to create a path where the unground return wire can get more current than it is designed for (i.e., 15 amp fuse, 14 AWG wire + out to load and - return back to battery).

For larger floating power systems, where you may have large conductors (50+ amps going to a solar charge controller, 150 amps to an AC inverter, 2 amps to an LED lamp), you should have properly rated double pole breakers/fuses for each pair of cables that leave the battery bus.

There are other issues why and how grounding is done in AC and DC power systems (lightning, static discharge, reducing the possibility of electrocution, cathodic protection systems for buried pipe/boats/telecom/natural gas piping/etc., fluorescent lighting, spark ignition for propane/natural gas burners, etc.)...

And there are reasons to (possibly) put fuse/breaker per solar panel on your van (in positive lead only required for DC ground bonded systems)--Typically with 3 or more panels in series (usually 100+ Watt per panel size) that help prevent solar wiring overheating if one panel gets shorted and the other 2+ panels feed excessive current to the shorted panel). But this may not even be an issue for a small solar power system.

-Bill

Thank you so much for this explanation. This is a question I have pondered and been asking for at least 6 months now and like Van, I found many people that recommended dual pole breakers, but none who could articulate why or when to do so (either technically or based on code).