1

I've been thinking about the #8 wire between a sub-panel, on a 50 amp breaker, and a 10kw solar inverter.

Between the 50 amp breaker and inside the inverter, in the event of a fault, the wires would be driven on two sides, potentially at upwards of 90 amps continuously. The size of the wire between the inverter and sub-panel was sized by the 50 amp breaker - #8, and is contained within PVC conduit. Inside the inverter the wire is both #8 and ~#16. The #16 wires goes to a WattNode power meter that consumes ~1 watt.

How is that #8 is allowed between the appliance and the breaker, and within the appliance, ~#16?

This is within the US and is for a SolarEdge 10k inveter. Sub-panel has 100 amp bus bar, w/utility at top and solar at bottom. inverter internal wiring

close-up of #16 and #8 wire connected

  • Where are you on this planet, and what make and model is the inverter in question? – ThreePhaseEel May 11 at 1:00
  • So you have a subpanel that's supplied in one direction by the solar panels, and supplied in the other direction by utility? – Harper May 11 at 5:14
  • @Harper - yes - supplied by utility on 1 side, and inverter on the other. As is standard, the utility enters 1 end of the panel, the inverter at the bottom. 125% rule applies. [its been 2.5 years since I reviewed this if I have the term wrong.] – pathfinder May 12 at 3:04
  • @ThreePhaseEel - why are these things important? I did use the word "allowed' - but maybe I mean "how does this make sense". In the US, and its a SolarEdge 10k something. – pathfinder May 12 at 3:07
  • I used a WattNode in a project and recall from its instructions that connection to its voltage inputs should be properly protected. (Eg #14 -> 15 amp breaker). – pathfinder May 12 at 3:09
1

The NEC has no bearing on factory wiring within listed equipment

The gauge of factory wiring in listed equipment is governed by the applicable UL standards and by the testing UL conducts in accordance to those standards (usually various temperature rise tests) as part of the listing process, not by the NEC. This is reflected in NEC 310.1:

310.1 Scope. This article covers general requirements for conductors and their type designations, insulations, markings, mechanical strengths, ampacity ratings, and uses. These requirements do not apply to conductors that form an integral part of equipment, such as motors, motor controllers, and similar equipment, or to conductors specifically provided for elsewhere in this Code.

As a result of this, we must look outside the NEC for an explanation of this. While I have not seen explicit provisions for these sorts of potential or control taps in what I have been able to read of UL 1741, the UL 508A standard for industrial control panels contains a provision in 41.1 Exception 1 that can be used for an argument by analogy in this case:

41.1 The conductors of a control circuit tapped off the load side of the branch circuit protective device shall have overcurrent protection sized in accordance with the ampacity of the control circuit conductor as specified in Table 28.1 and Table 38.1.

Exception No. 1: When the control circuit is tapped off a motor branch circuit protective device and the control wires do not leave the industrial control panel enclosure (such as when a start-stop button is provided on the enclosure cover) the motor branch circuit protective device provides the required overcurrent protection when its rating does not exceed that specified in Table 41.1.

Given that Table 41.1 in UL 508A-2014 permits a 14AWG control tap on a motor branch circuit of up to 100A in size and a 16AWG tap on a circuit up to 40A when the conditions of Exception 1 are met, I would say that the safety case for the potential tap wiring you see in your inverter works analogously to the logic behind the above quoted exception.

  • Motor branch - very interesting. How about between the inverter and panel? I'll mention that those #8s are in a PVC conduit end to end. – pathfinder May 12 at 15:42
  • @pathfinder -- the #8s in the branch circuit can use the 75degC (50A) line provided that the terminations are rated at 75degC (breaker/loadcenter lugs definitely are, and the inverter's termination's may be rated for that, check the installation manual for the inverter to be sure) – ThreePhaseEel May 12 at 16:33
0

Wiring inside the appliance is subject to the rules of UL, not NEC.

I'm kind-of surprised by the #8 wire on a 50A breaker, as #6 is required for a 50A breaker normally. (you are only allowed to use the 60C current rating on branch circuits and feeder <100A). They may just be saying that it's rather unlikely that the #8 wire will be loaded beyond 40A, because reaching the full 10kw would require perfect conditions and panel sizing. But if #8/50A breaker is in the labeling and instructions, then it was part of the UL listing, and UL tested the failure modes there.

I see what you mean about a wire fault mid-run causing 90A of flow, but that would only be at the fault - the wire from the inverter would only be 41.6A tops, and the wire from the panel is breaker-restricted at 50A.

  • I'll verify the size of the wire today. Yes - like a bus bar, the current on either side of the load is under the limit. So maybe this is why its okay. – pathfinder May 12 at 15:54
  • Are you concerned about load inside the subpanel, or a hypothetical hot-neutral or hot-hot arc fault in the wiring itself? – Harper May 12 at 15:59
  • @Harper -- UL may be allowing the #8 wire for 50A due to all terminations involved being 75degC rated – ThreePhaseEel May 12 at 16:32
  • @Harper - Concerned that I don't understand why it allowed and not a hazard. As you pointout, #8 only has the 90amps at the fault - but the #16 potentially has 90 amps. I plan to add another no-load controller similar to the WattNode and based on previous weeks conversations see conflict. Can't really add more #8 in the box! – pathfinder May 14 at 13:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.