PV Output Circuit Fuse Sizing Confusion

[Andrew445]

Hello, I seem to be confusing myself when trying to figure out fuse sizing.

I have a Isc x 1.25 = 207.9A. Isc x 1.56 = 260A.

The inverter I am using has fuse sizes 250A, 315A, etc. Note there is no 300A available.

I have gone through the entire sizing exercise and determined that I need to go with a 315A OCPD and 500kcmil aluminum conductors. I am trying to see if there are any code violations in using a 250A OCPD instead.

Per NEC 690.8(B)(1)(a), the OCPD is required "To carry not less than 125 percent of the maximum currents . . ." I understand that this refers to the 1.56 x Isc number, but why is carry the verb that is chosen? Further down in (d) it specifies the rating or setting is in accordance with 240.4. 240.4 essentially requires that the derated ampacity of the conductor be less than the next higher standard OCPD rating. To me, 240.4 is the only thing that matters when it comes to OCPD rating, and it is setting a maximum, not a minimum. What is 690.8(B)(1)(a) getting at with "To carry" ?


I am probably reading too much into this, but I appreciate anyone setting me straight.

 

[jaggedben]

I am trying to see if there are any code violations in using a 250A OCPD instead.

I believe that would be a violation of 690.8(1)(a). (Unless the device meets the exception.)

240.4 essentially requires that the derated ampacity of the conductor be less than the next higher standard OCPD rating.

Well, I wouldn't quite agree with that. 240.4 (B) does permit it to be higher. What's weird for you is that your 315A fuse isn't a 'standard' rating. I don't know quite what to do with that. Argue to the AHJ that 'available' = 'standard'? Does your derate factor bring the ampacity after conditions of use below 300A, or not? It makes a difference.

To me, 240.4 is the only thing that matters when it comes to OCPD rating, and it is setting a maximum, not a minimum. What is 690.8(B)(1)(a) getting at with "To carry" ?

I think it's setting a minimum for PV circuits. Chapters 5-7 "modify" Chapters 1-4, and thus take precedence. See 90.3.

 

[ggunn]

I believe that would be a violation of 690.8(1)(a). (Unless the device meets the exception.)



Well, I wouldn't quite agree with that. 240.4 (B) does permit it to be higher. What's weird for you is that your 315A fuse isn't a 'standard' rating. I don't know quite what to do with that. Argue to the AHJ that 'available' = 'standard'? Does your derate factor bring the ampacity after conditions of use below 300A, or not? It makes a difference.



I think it's setting a minimum for PV circuits. Chapters 5-7 "modify" Chapters 1-4, and thus take precedence. See 90.3.

I haven't been awake long this morning, but this question occurs to me: Once you have fuses in place at the string level and the conductors from the combiner to the inverter are sized appropriately to handle the total Isc X 1.56 (or Isc X 1.25 derated for conditions, whichever is larger) of all the modules, why would any fusing on the PV output circuit be necessary? The modules are incapable of producing enough current to damage the conductors and the inverter cannot backfeed DC into its inputs. Unless the inverter has multiple inputs, what would fusing at that point be protecting against?

 

[Andrew445]

Yes, the inverter has multiple inputs. There are (9) similarly fused fused inputs.

I think I'll just stick with the 1.56 minimum. I'm still relatively new at this, so I appreciate the clarification. Dropping to 250A OCPD would have allowed for a significant decrease in cable size.

 

[ggunn]

Yes, the inverter has multiple inputs. There are (9) similarly fused fused inputs.

I think I'll just stick with the 1.56 minimum. I'm still relatively new at this, so I appreciate the clarification. Dropping to 250A OCPD would have allowed for a significant decrease in cable size.

The rule is that your conductors must be sized for combined Isc X 1.56 or combined Isc X 1.25 derated for conditions, whichever is larger. And then there's voltage drop.

 

[jaggedben]

I haven't been awake long this morning, but this question occurs to me: Once you have fuses in place at the string level and the conductors from the combiner to the inverter are sized appropriately to handle the total Isc X 1.56 (or Isc X 1.25 derated for conditions, whichever is larger) of all the modules, why would any fusing on the PV output circuit be necessary?

The rules for PV output circuits are the same as for PV source circuits. (690.9) Those are the rules which, generally speaking but not always, do not require OCPD unless there are more than two circuits. With combined outputs (rather than strings) that rule of thumb isn't quite as hard and fast because it's more likely that two combined outputs will have different Isc. If there is only one output circuit then no OCPD is required by the code.

 

[ggunn]

The rules for PV output circuits are the same as for PV source circuits. (690.9) Those are the rules which, generally speaking but not always, do not require OCPD unless there are more than two circuits. With combined outputs (rather than strings) that rule of thumb isn't quite as hard and fast because it's more likely that two combined outputs will have different Isc. If there is only one output circuit then no OCPD is required by the code.

Interesting. I thought about that, and it seemed to me that the two vs. more than two rule should be different for output circuits.

At the string level the weak link is the module; strings are usually wired with #10 conductors which have much higher ampacity than string Isc. A fault on a string when there are only two strings could have 2 X Isc fault current, but #10 copper can handle it.

With PV output circuits, though, the conductors are sized to 1.56 X combined Isc or 1.5 X combined Isc derated for conditions (whichever is higher), while an unfused fault could have 2 X combined Isc fault current when there are two of them. Isn't that a problem?

 

[jaggedben]

Interesting. I thought about that, and it seemed to me that the two vs. more than two rule should be different for output circuits.

At the string level the weak link is the module; strings are usually wired with #10 conductors which have much higher ampacity than string Isc. A fault on a string when there are only two strings could have 2 X Isc fault current, but #10 copper can handle it.

With PV output circuits, though, the conductors are sized to 1.56 X combined Isc or 1.5 X combined Isc derated for conditions (whichever is higher), while an unfused fault could have 2 X combined Isc fault current when there are two of them. Isn't that a problem?

Yes, I think that's right. Note that the 'two vs. more than two rule' isn't actually the rule, it's just the usual consequence of the rules when dealing with strings.

It may be that the usual consequence when dealing with combiner outputs is 'one vs. more than one', and that actually is one of the rules. What the other rule says is that if conductors are large enough to handle the Isc from all parelled sources, then OCP is not required. I suppose in most cases it will be more economical to install OCP than to upsize conductors, but there might be exceptions.

 

[Andrew445]

The rule is that your conductors must be sized for combined Isc X 1.56 or combined Isc X 1.25 derated for conditions, whichever is larger. And then there's voltage drop.

Thanks for all the help! This is definitely a confusing code section. I still find it odd that you apply the derate factor to the available current value, when typically you apply it to a given conductor ampacity. But I understand the reasoning. So in this case, the second value would be Isc x 1.25 / 0.87. Typically the first value is the larger unless your combined derates are less then 0.80.

Voltage drop is a whole other can of worms, not to mention thermal ampacity analysis with the correct soil rho value and ambient temperature.



If anyone is still reading this thread I have another small question that popped into my mind:

If I locate all DC overcurrent protection in a single enclosure located adjacent to the inverter, and do not use any internal inverter fuses, is this compliant? The OC protection will be breakers that can be manually operated as disconnects. To me this sounds like a "tap rule" thing, but from reading over that section, I'm not sure it strictly applies. Just making sure there wouldn't be any additional ampacity rules I need to employ for the short conductor length between the enclosure and inverter.

 

[shortcircuit2]

If anyone is still reading this thread I have another small question that popped into my mind:

If I locate all DC overcurrent protection in a single enclosure located adjacent to the inverter, and do not use any internal inverter fuses, is this compliant? The OC protection will be breakers that can be manually operated as disconnects. To me this sounds like a "tap rule" thing, but from reading over that section, I'm not sure it strictly applies. Just making sure there wouldn't be any additional ampacity rules I need to employ for the short conductor length between the enclosure and inverter.

Not sure on this question. The inverters instructions may require fusing at the outputs. 690.9(A) requires OC protection in accordance with article 240.

240.21 requires conductors to be protected where they receive their supply. I have wondered how come PV Output Circuit conductors have the fusing opposite where they receive their supply. Maybe it has to do with the reverse currents during shortcircuit of the outputs.

There is a change to wording in 690.9 for 2014...

(A) Circuits and Equipment. PV source circuit, PV output
circuit, inverter output circuit, and storage battery circuit
conductors and equipment shall be protected in accordance
with the requirements of Article 240. Protection
devices for PV source circuits and PV output circuits shall
be listed for use in PV systems. Circuits, either ac or dc,
connected to current limited supplies (e.g. PV modules, ac
output of utility-interactive inverters) and also connected to
sources having significantly higher current availability (e.g.
parallel strings of modules, utility power) shall be protected
at the source from overcurrent.

SHALL BE PROTECTED AT THE SOURCE is added...

Anyone else want to chime in with input on this?

 

[ggunn]

Not sure on this question. The inverters instructions may require fusing at the outputs. 690.9(A) requires OC protection in accordance with article 240.

240.21 requires conductors to be protected where they receive their supply. I have wondered how come PV Output Circuit conductors have the fusing opposite where they receive their supply. Maybe it has to do with the reverse currents during shortcircuit of the outputs.

There is a change to wording in 690.9 for 2014...

(A) Circuits and Equipment. PV source circuit, PV output
circuit, inverter output circuit, and storage battery circuit
conductors and equipment shall be protected in accordance
with the requirements of Article 240. Protection
devices for PV source circuits and PV output circuits shall
be listed for use in PV systems. Circuits, either ac or dc,
connected to current limited supplies (e.g. PV modules, ac
output of utility-interactive inverters) and also connected to
sources having significantly higher current availability (e.g.
parallel strings of modules, utility power) shall be protected
at the source from overcurrent.

SHALL BE PROTECTED AT THE SOURCE is added...

Anyone else want to chime in with input on this?

Fusing on the DC side is there to protect the modules and conductors in a string in the event of a fault from backfeed from the other strings. In that case, the source of the current is the point where all these conductors are joined and the OCPDs are placed as close to that point as possible, which is the opposite end from the array. Fusing is NOT there to protect the conductors from the PV array in the absence of a fault; assuming that the conductors are appropriately sized, the modules are incapable of putting out sufficient current to endanger the conductors.

 

[SolarPro]

If I locate all DC overcurrent protection in a single enclosure located adjacent to the inverter, and do not use any internal inverter fuses, is this compliant? The OC protection will be breakers that can be manually operated as disconnects.

Sure. Benetek Solar offers a listed solution like this:

http://www.bentek.com/commercial-safety-systems-bentek.html#cbss

I've also seen solutions like this from Eaton.

The fuse servicing requirements in 690.16 will likely steer a lot of designs in this direction, since circuit breakers provide both OCP and a disconnecting means.

 

[Andrew445]

Sure. Benetek Solar offers a listed solution like this:

http://www.bentek.com/commercial-safety-systems-bentek.html#cbss

I've also seen solutions like this from Eaton.

The fuse servicing requirements in 690.16 will likely steer a lot of designs in this direction, since circuit breakers provide both OCP and a disconnecting means.

This product is exactly what I had in mind when I posed the question! It also helps because the inverter does not have a wide variety of ratings for 1000V listed fuses.

I was just wondering how the unprotected conductor between this enclosure and the inverter is to be treated, or if there is a code section that covers this. I plan on specifying RGS conduit to ensure full mechanical protection.

 

[SolarPro]

You may need to defer to the product manufacturer and the product listing. If the CBSS were manufacturer by the inverter manufacturer, then the connection between the two busses would be considered internal wiring. Technically, that's not the case, but the Code doesn't seem to provide much clarity in terms of how to treat this. If the inverter is incapable of backfeeding from the ac side to the dc side, then the current on the inverter input circuit will always be lower than the rating of any OCPD you might install. The PV generator is inherently current-limited.

It used to be (2008 NEC and earlier) that the Exception to 690.9(A) was written in a general enough way to apply to PV output circuits. That is no longer the case the way this exception reads as of the 2011 cycle. The 2014 Code will require that the the OCP used in PV source and output circuits is listed for use in PV systems. That suggests to me that you need to use the CBSS per the manufacturers instructions, since the CBSS is designed and listed for this particular application. (Just be sure to consult both Benetek and the inverter manufacturer, per the informational note in 690.9.)