Wire size and fusing for 3P array

[fafrd]

I’m toying with the idea of combining 2 or 3 panels before feeding into a high-capacity Microinverter and have some newbie questions about wire size and fusing.

I know that 3P generally requires a fuse on each incoming branch/panel while 2P does not and I just want to confirm the logic for that again.

The concern is a short in the panel itself, not the panels lead-in wires, correct?

A short in the panel can draw 2 x Isc from the other 2 panels + Isc from the shorted panel so total current at the point at the location of the short can be as much as 3x.Isc.

Am I correct that that is the concern?

Or is it more generally that the panel output wires are only rated for 1x Isc which can never be exceeded by a short within the panel at 2P but can draw 2 x Isc through those wires at 3P?

And my second question has to do with PV wire sizing and chassis wiring limits versus power transmission limits.

For a free-standing wire from a battery to an inverter, chassis wiring limits apply, so wouldn’t that be the case for a short free-standing wire within a PV array such as an MC4 ‘Y’ cable?

The Y cables I am considering are 12AWG PV wire and state that they are rated for up to 35A maximum on the common output.

12AWG is limited to 9.3A for power transmission but 41A for chassis wiring: https://www.powerstream.com/Wire_Size.htm

So is 41A the correct limit to use for a short length of 12AWG wire freestanding within a solar array (appropriately de-rated)?

If I’m understanding all of this correctly, the MC4 inline fuses are expensive enough that 2P strings may be the only option that makes sense…

 

[FilterGuy]

I’m toying with the idea of combining 2 or 3 panels before feeding into a high-capacity Microinverter and have some newbie questions about wire size and fusing.

I know that 3P generally requires a fuse on each incoming branch/panel while 2P does not and I just want to confirm the logic for that again.

The concern is a short in the panel itself, not the panels lead-in wires, correct?

A short in the panel can draw 2 x Isc from the other 2 panels + Isc from the shorted panel so total current at the point at the location of the short can be as much as 3x.Isc.

Am I correct that that is the concern?

Or is it more generally that the panel output wires are only rated for 1x Isc which can never be exceeded by a short within the panel at 2P but can draw 2 x Isc through those wires at 3P?

And my second question has to do with PV wire sizing and chassis wiring limits versus power transmission limits.

For a free-standing wire from a battery to an inverter, chassis wiring limits apply, so wouldn’t that be the case for a short free-standing wire within a PV array such as an MC4 ‘Y’ cable?

The Y cables I am considering are 12AWG PV wire and state that they are rated for up to 35A maximum on the common output.

12AWG is limited to 9.3A for power transmission but 41A for chassis wiring: https://www.powerstream.com/Wire_Size.htm

So is 41A the correct limit to use for a short length of 12AWG wire freestanding within a solar array (appropriately de-rated)?

If I’m understanding all of this correctly, the MC4 inline fuses are expensive enough that 2P strings may be the only option that makes sense…

You may want to review this. It covers your questions.

Fusing & Wire Sizing guidelines for solar panels

To get the paper, click on the orange'' button at the top of the screen. This resource is intended to help the user understand when they need to add Over Current Protection Devices (OCPDs - Fuses or Breakers) to the solar panel array and what...

diysolarforum.com

 

[fafrd]

You may want to review this. It covers your questions.

Fusing & Wire Sizing guidelines for solar panels

To get the paper, click on the orange'' button at the top of the screen. This resource is intended to help the user understand when they need to add Over Current Protection Devices (OCPDs - Fuses or Breakers) to the solar panel array and what...

diysolarforum.com

That is an absolutely fantastic tutorial - thanks for putting it together.

When I add in the costs of the fusing and the more pensive 3-way Y connectors, I’m coming to the conclusion that a 3P combine isn’t worth it.

Your guide is fantastic and comprehensive but does not address one of my questions.

Your 3P example needs wire sized for 45A but you state this means at least 8 AWG without indicating what ampacity table you are using.

The guidelines for adjustments for number of conductors in conduit all refer to conductor ratings in conduit, but my understanding is that free single conductors in air can use chassis wiring ratings - do you agree or is there an issue with that understanding?

Chassis wiring rating for 10AWG is 55A, so is there a reason 45A of worst-case Isc could not be carried by a short length of 10AWG PV wire freestanding in air such as a 10AWG ‘Y’ connector?

Thanks again.

 

[FilterGuy]

Your 3P example needs wire sized for 45A but you state this means at least 8 AWG without indicating what ampacity table you are using.

Sorry 'bout that.

This is the chart I like to use because it gives ampacity for all the standard insulation temperature ratings.

https://2ryul01momzwnfgn01s3t1kg-wpengine.netdna-ssl.com/wp-content/uploads/ABYC-Unbundled-Current-Rating-Chart.jpg

 

[fafrd]

Sorry 'bout that.

This is the chart I like to use because it gives ampacity for all the standard insulation temperature ratings.

https://2ryul01momzwnfgn01s3t1kg-wpengine.netdna-ssl.com/wp-content/uploads/ABYC-Unbundled-Current-Rating-Chart.jpg

Nice, thanks. So based on that, you’d agree that a single 10AWG 90C wire could handle 44A, correct?

These 3Y cables are rated for 30A per input and 40A of output, but also up to 222F: https://www.amazon.com/BougeRV-Conn...=9032080&hvtargid=pla-359698458424&psc=1&th=1

Your table has 105F which would allow 12AWG to be rated up to 45A, so I’m trying to u set stand what wire guage these are using and make sure their current ratings are correct….

 

[FilterGuy]

Nice, thanks. So based on that, you’d agree that a single 10AWG 90C wire could handle 44A, correct?

Yes but....... That indicates it is *safe* but does not necessarily indicate it is efficient. Consequently, you must also consider voltage drop. If the run is very long, you may want to bump to 8AWG.

Also, at least for the NEC, if the wire is connecting to something like a switch or breaker with a lower temp rating, the wire must be treated as if it has the lower temp rating.

 

[fafrd]

Yes but....... That indicates it is *safe* but does not necessarily indicate it is efficient. Consequently, you must also consider voltage drop. If the run is very long, you may want to bump to 8AWG.

Also, at least for the NEC, if the wire is connecting to something like a switch or breaker with a lower temp rating, the wire must be treated as if it has the lower temp rating.

Yes, understand. I need to check on the temperature ratings of the MC4 connectors, but PV wire itself seems to be rated to 150C (302F).

The 125F column in your table would mean 12AWG can be rated for up to 50A (assuming all of the MC4 connectors and inline fuses are also rated to at least 125C).

 

[fafrd]

Yes, understand. I need to check on the temperature ratings of the MC4 connectors, but PV wire itself seems to be rated to 150C (302F).

The 125F column in your table would mean 12AWG can be rated for up to 50A (assuming all of the MC4 connectors and inline fuses are also rated to at least 125C).

The BourgRV MC4 connectors are only rated to 105C: https://www.bougerv.com/products/mc4-male-female-solar-panel

So back to a maximum rating of 45A for 12AWG (assuming the MC4 fuses are also rated to at least 105C).

 

[fafrd]

The BourgRV MC4 connectors are only rated to 105C: https://www.bougerv.com/products/mc4-male-female-solar-panel

So back to a maximum rating of 45A for 12AWG (assuming the MC4 fuses are also rated to at least 105C).

The BourgRV 15A MC4 fuses are only rated to 90F: https://www.bougerv.com/products/15a-solar-fuse-holder

But since the fuse is protecting the entire assembly, isn’t it the case that it just needs to be rated at 125% of the rated current and at 80% of the wires it’s protecting?

So Iscwc of 9.29A (my case) needs a fuse of at least 11.6125A (rounded up to 15A) flowing in wires rated for at least 14.515A and the outbound worst-case current of 3 x Iscwc = 27.87A needs wires and MC4 connectors rated for at least 43.55A (but no additional fuse).

Do I have that right?

 

[fafrd]

Yes but....... That indicates it is *safe* but does not necessarily indicate it is efficient. Consequently, you must also consider voltage drop. If the run is very long, you may want to bump to 8AWG.

Lengths are very short (< 7”)

Also, at least for the NEC, if the wire is connecting to something like a switch or breaker with a lower temp rating, the wire must be treated as if it has the lower temp rating.

This is the key point. If the fuse/breaker is only rated to 90C, doesn’t that 90C rating only need to be applied to 125% of the current it is protecting, not 156% (which is the wire it is protecting)?

Fuse size needs to be at least 125% of worst-case current.

Wire rating needs to be at least 125% of fuse size.

So if it is only the fuse itself only rated to 105C, why should that limit the overall temperature rating of the full connection - at worst it means a nuisance trip but no safety issue, right?

124% of Isc flowing = 11.53A (my case) c 3 = 34.56A.

12AWG PV wire and BourgRV MC4 connectors have no problem handling 125% of that current (43.2A) up to 105C.

If temps pass 90C so that a 15A fuse blows with only 77% of trip rating, it’s a nuisance trip that could have been avoided, but only in the case that ambient temperature passed 90C / 194F.

Or is the point that above 90C the fuse may not blow even if there is a short and current surpasses 15A on one branch?

1.56 x 3 x Iscwc = 43.2A which is below the 45A rating of 12AWG wire @ 90C, so looks like I should be OK in any case…

 

[FilterGuy]

This is the key point. If the fuse/breaker is only rated to 90C, doesn’t that 90C rating only need to be applied to 125% of the current it is protecting, not 156% (which is the wire it is protecting)?

I don't think the NEC carves it out that way.

 

[fafrd]

I don't think the NEC carves it out that way.

Sorry, I edited my above post while you were typing.

To repeat:

My worst-case Iscwc = 9.29A.
My fuse is only rated to 90C.

Each incoming leg/branch needs to be protected by a fuse sized at least 125% x Iscwc = 11.6125A rounded up to 15A and to be carried by a wire rated for at least 125% x 15A = 18.75A @ 90C.

3 times Iscwc = 27.87A so I need the outbound wires and MC4 connectors to be rated for at least 1.5625 x 27.87A = 43.55A @ 90C.

At 90C, 12AWG PV wire freestanding in air is rated for 45A which is more than 156.25% above worst-case current of 3 x Iscwc of 43.55A.

So short length of 12AWG PV wire will be OK (barely, correct?

 

[FilterGuy]

Sorry, I edited my above post while you were typing.

To repeat:

My worst-case Iscwc = 9.29A.
My fuse is only rated to 90C.

Each incoming leg/branch needs to be protected by a fuse sized at least 125% x Iscwc = 11.6125A rounded up to 15A and to be carried by a wire rated for at least 125% x 15A = 18.75A @ 90C.

3 times Iscwc = 27.87A so I need the outbound wires and MC4 connectors to be rated for at least 1.5625 x 27.87A = 43.55A @ 90C.

At 90C, 12AWG PV wire freestanding in air is rated for 45A which is more than 156.25% above worst-case current of 3 x Iscwc of 43.55A.

So short length of 12AWG PV wire will be OK (barely, correct?

The following was edited for clarity....


For 3 or more Parallel Strings of 1 or more series panels the basic rules are:
• OCPDs (Fuses or breakers) required
• The voltage rating of the OCPDs must be Ns x Vmax or greater. (Ns = number of serial panels)
• Each Source circuit must have OCPD rated for at least 156% of Isc and less than the Series Fuse rating of the panels.
• Source circuit cables must be rated for at least 156% of Isc
• Source circuit cables must be rated for more current than the OCPD
• Output circuit cables rated for at least Np x 156% of Isc (Np = number of parallel strings

The source circuits are the wires that hook the panels in series on a single string
The output circuits are the wires after the source circuits are combined in parallel.

=======================

If the Isc is 9.29A per string and there are 3 or more strings then the NEC wants a fuse that is ~ 1.55 x Isc = 1.55 x 9.29A=14.4A. Round up to 15A fuse. The Fuse must also be *less* than the fuse rating of the panel.

If there are 3 strings, the output conductors must be able to handle 3 x 15A = 35A
There are no requirements for an over-current protection device on the output conductors, but there are requirements for a disconnect that breaks both lines on the output conductors. (Some people use breakers as disconnects)

If the connectors on the breakers, fuses, or disconnects are rated at a temp below the temp rating of the source or output conductors, the wire must be sized as if it's temp rating was the same as the breaker or fuse. The reason for this is that the wire can safely get all the way up to it's temperature rating and still be safe. If that wire is tied to a device with a lower temp rating, it can over-heat that device. By treating the wire as if it has the lower temp rating, it will ensure the device does not overheat.

There are additional wire de-rating requirements for the following:
- if there are more than two current carrying conductors in a conduit.
- If the conductors are in a conduit that is run across the roof. (The derating varies depending on how close to the roof the conduit is.)

 

[FilterGuy]

Duplicate post removed

 

[fafrd]

The following was edited for clarity....


For 3 or more Parallel Strings of 1 or more series panels the basic rules are:
• OCPDs (Fuses or breakers) required
• The voltage rating of the OCPDs must be Ns x Vmax or greater. (Ns = number of serial panels)
• Each Source circuit must have OCPD rated for at least 156% of Isc and less than the Series Fuse rating of the panels.

I thought the OCPD had to be at least 125% of Iscmax (to avoid nuisance trips) - what is the logic to use 125% x 126%?

• Source circuit cables must be rated for at least 156% of Isc

Agreed, which I thought meant a cable rating of SDR least 125% of the OCPD sized at 125% of max current.

• Source circuit cables must be rated for more current than the OCPD

I thought it had to be sized at 125% of the OCPD - it can be greater than OCBD by any small amount?

• Output circuit cables rated for at least Np x 156% of Isc (Np = number of parallel strings

Agreed.

The source circuits are the wires that hook the panels in series on a single string
The output circuits are the wires after the source circuits are combined in parallel.

=======================

If the Isc is 9.29A per string and there are 3 or more strings then the NEC wants a fuse that is ~ 1.55 x Isc = 1.55 x 9.29A=14.4A. Round up to 15A fuse. The Fuse must also be *less* than the fuse rating of the panel.

We’re coming to the same fuse rating of 15A but you are getting there by rounding up from 156% and I was getting there by rounding up from 125%…

If there are 3 strings, the output conductors must be able to handle 3 x 15A = 35A

3 x 15A = 45A where I come from .

But back to the central question, since short lengths of 12AWG at 90C freestanding in air are rated for 45A, they would suffice in my case, correct?

There are no requirements for an over-current protection device on the output conductors, but there are requirements for a disconnect that breaks both lines on the output conductors. (Some people use breakers as disconnects)

Panels connected directly to Microinverters do not require any disconnect (the Microinverter itself is considered a disconnect). So why would a 2P or 3P subarray connected to a Microinverter be any different?

If the connectors on the breakers, fuses, or disconnects are rated at a temp below the temp rating of the source or output conductors, the wire must be sized as if it's temp rating was the same as the breaker or fuse. The reason for this is that the wire can safely get all the way up to it's temperature rating and still be safe. If that wire is tied to a device with a lower temp rating, it can over-heat that device. By treating the wire as if it has the lower temp rating, it will ensure the device does not overheat.

Yes, understand. But even with fuses rated at only 90C, I believe a 12AWG wire freestanding in air has the rating to carry my 3 x 9.29A protected by 3 x 15A fuses and would appreciate to understand whether you agree.

There are additional wire de-rating requirements for the following:
- if there are more than two current carrying conductors in a conduit.
- If the conductors are in a conduit that is run across the roof. (The derating varies depending on how close to the roof the conduit is.)

Understand about derating for conduit and proximity to roof as well but since my 12AWG conductor will be freestanding in air and above roof at racking height of ~4”, don’t believe either derating applies to my case.

 

[fafrd]

I’m going to be cautious about going down this ‘Y’ connector parallel subarray route but am taking the opportunity to bone up on my understanding.

I’ve done a bit more sleuthing and the BourgRV Y cables I am considering use 12AWG on the incoming branches and 10AWG on the output wire.

They state a maximum of 30A per input and a maximum of 40A on the output but also state that the maximum panel rating of all panels connected to the inputs cannot exceed 600W. And the temperature rating is 90C.

12AWG @ 90C is rated for 40A freestanding in air and has a standard ampacity in conduit of 30A, so either they are being conservative or the ‘Y’ bond itself does not have the same rating as 12AWG wire.

10AWG @ 90C is rated for 50A freestanding in air and has a standard ampacity in conduit of 40A, so again, they are either using conduit ratings to be conservative or the Y bond itself is the limiting factor even when freestanding in air.

So I get all of that and I have no problem with being conservative, but what confuses me is the ‘total of 600W’ maximum.

That means no 2 panels greater than 300W or no 3 panels greater than 200W.

If we take a worst-case Isc like mine of 9.29A protected by 15A fuses. a standard 2-input Y would need to have an output rated for over 30A (your definition) or over 37.5A (my 125% safety margin definition).

In either case, 10AWG at the conservative rating of 40A (in conduit) can handle that level of output current.

But my panels are 335W panels and 2 of them are 12% above their 600W maximum rating.

These two-input Y cables are not connected: the positive only sees the voltage drop of the positive voltage across it’s Y splice and the negative only sees the voltage drop across it’s Y splice.

So I’m not understanding any reason why the string voltage or the total Watts of the panels driving the current is a factor.

The cables are rated for 1000VDC, so nothing to do with arcing or any of that.

The 600W maximum does not appear anywhere in the formal specs but only in the Q&A section, so I’m suspecting this is just a guideline to steer naive customers towards acceptable applications and to avoid uneducated customers getting into trouble.

With 10AWG output wires, I’m guessing that as long as 156% x 2 x Isc is below 40A, these dual Y connectors should work fine regardless of string voltage (within limits) or wattage.

The tripe-input Y connector is a bit dicier. Inputs need to be protected by fuses (15A in my case) and according to your understanding, that means output wire needs to be able to handle 45A (above their rating of 40A).

10AWG @ 90C freestanding in air is rated to 50A, so ‘greater than 3 x fuse rating of 15A by your definition, but since nothing is known about the actual internal Y cable bond other than their ratings, assuming it is also rated to more than 45A is risky.

My takeaway from all of this is that dual-panel splices in parallel using these Y connectors should be fine, but triple-panel splices in parallel are probably not worth the added cost and risk.

 

[FilterGuy]

I thought the OCPD had to be at least 125% of Iscmax (to avoid nuisance trips) - what is the logic to use 125% x 126%?

The problem with panels and OCPD is that the 'normal' current is so close to the 'max' current that it becomes nearly impossible to fuse a panel (or array) in a way that will work reliably *and* not have nuisance trips. Consequently, they require the wires to be large enough to handle all the current from the panel or array without causing a fire. The NEC does this by defining Iscmax as 1.25 x ISC. This is to cover all possibilities, including current variations due to temperature. It then requires an additional 1.25 margin for the string fuses and wiring in order to ensure safety. Since Iscmax is 1.25 x Isc, the wiring and fusing must handle 1.25 x (1.25 x Isc) = 1.56 x Isc.

Also, note that the string fuses on the source circuits are there to prevent a fire if multiple panels are dumping current into a shorted panel. The string fuses do not protect the output circuit. (The source circuits are the wires that connect the string of panels in series. The output circuit is the set of wires *after* the strings are put in parallel)

3 x 15A = 45A where I come from .

Oops. I'll correct that.

Panels connected directly to Microinverters do not require any disconnect (the Microinverter itself is considered a disconnect). So why would a 2P or 3P subarray connected to a Microinverter be any different?

You may be correct on that... I have not tried to tease out the requirements around micro-inverters. The disconnect requirements in the NEC are confusing to start with and micro-inverters may have some carve-outs.

Yes, understand. But even with fuses rated at only 90C, I believe a 12AWG wire freestanding in air has the rating to carry my 3 x 9.29A protected by 3 x 15A fuses and would appreciate to understand whether you agree.

In answer to your question, any wire with an 80c Insulation (or higher) can handle 3 x 9.29 = 27.87A. But I am not sure that is relevant to anything.

 

[FilterGuy]

I’m going to be cautious about going down this ‘Y’ connector parallel subarray route

It is unusual to have parallel panels on a micro-inverter, so caution is a good idea. Since I am not aware of others that have done this, I don't know what the best practices might be. However, I would certainly use the PV wiring and fusing requirements that apply to parallel panels on a 'regular' inverter.

 

[fafrd]

The problem with panels and OCPD is that the 'normal' current is so close to the 'max' current that it becomes nearly impossible to fuse a panel (or array) in a way that will work reliably *and* not have nuisance trips. Consequently, they require the wires to be large enough to handle all the current from the panel or array without causing a fire. The NEC does this by defining Iscmax as 1.25 x ISC. This is to cover all possibilities, including current variations due to temperature. It then requires an additional 1.25 margin for the string fuses and wiring in order to ensure safety. Since Iscmax is 1.25 x Isc, the wiring and fusing must handle 1.25 x (1.25 x Isc) = 1.56 x Isc.

Got it. Thanks for the explanation.

Also, note that the string fuses on the source circuits are there to prevent a fire if multiple panels are dumping current into a shorted panel. The string fuses do not protect the output circuit. (The source circuits are the wires that connect the string of panels in series. The output circuit is the set of wires *after* the strings are put in parallel)

Yes, that much I had understood.

Oops. I'll correct that.

You may be correct on that... I have not tried to tease out the requirements around micro-inverters. The disconnect requirements in the NEC are confusing to start with and micro-inverters may have some carve-outs.

Pretty certain that a Microinverter offers the same level of protection as an RSD (and thus is one reason that the new rules requiring module-level RSD pretty much close the gap between the cost of Microinverters versus String Inverters…).

In answer to your question, any wire with an 80c Insulation (or higher) can handle 3 x 9.29 = 27.87A. But I am not sure that is relevant to anything.

I think you meant 90C, correct?

But my question was related to 3 15A fuses.

Now I understand that the NEC considers 1.25x Iscmax to be the required current to fuse above by 125% (so 14.5A in my case).

So with 3 incoming 15A fuses rated at 90C, the outbound wire needs to be rated for at least 45A, correct?

You are correct that now that I know the outbound wire of the Y connectors I’m considering are 10AWG and not 12AWG, the question of outbound wire rating is moot.

But the question of overall triple Y-connector rating remains: if rated for 45A it would be safe protected by 3 15A fuses and if only rated for 40A, it would not (at least according to NEC), correct?

 

[fafrd]

It is unusual to have parallel panels on a micro-inverter, so caution is a good idea. Since I am not aware of others that have done this, I don't know what the best practices might be. However, I would certainly use the PV wiring and fusing requirements that apply to parallel panels on a 'regular' inverter.

I’ve spoken to the Microinverter manufacturer. They had never considered this possibility either, but primarily because they are focused on ‘new’ installs using modern panels.

Their newest 800W dual-Microinverter they recommend to be connected to 2 600W panels (jumbos).

When I pointed out to them that a 600W half-cut panel is essentially 2 300W half-panels in parallel and that splicing together two older ~300W panels with Y-connectors essentially amounted to using more real estate to deliver the same power, they said they would check with engineering and then came back the next day to say that that should be fine.

It’s unlikely to ever be attractive to new solar customers but for existing solar customers with older panels either upgrading their Microinverters to the new, more powerful offerings (my case) or for string inverter customers deciding to upgrade to Microinverters, it’s an attractive option to reduce the number of Microinverters required by half…