How do you apply 2023 NEC 690.8(A)(1)(a)(2) to calculate PV source circuit maximum current

[pv_n00b]

I've seen people use 690.8(A)(1)(a)(2) and software like SAM to calculate PV source circuit maximum current in two ways:

• Find the three maximum hourly currents and average them. The three data points do not need to be consecutive hours.
• Find the highest three hour average of any three consecutive data points.

NEC 690.8(A)(1)(a)(2) is not very specific about how to get the three hour average so either of these will comply. I'm just curious if one or the other is more commonly used. I've made a poll.

 

[wwhitney]

The term "3-hour average" implies a consecutive 3 hour window, so the first option is unnecessarily conservative. Unspecified is whether the 3 hour window should be sliding, and if so how frequently it should be recalculated. But if you have simulated hourly data, then recalculating it every hour is reasonable.

Cheers, Wayne

 

[wwhitney]

BTW (off topic), it seems to me that the definition of "continuous load" could be improved similar to the language in this section. I.e. in addition to the usual rating, assign every load a continuous rating, which would be the maximum average current for 3 hours. Then for usual OCPD, a continuous load would be one whose continuous rating exceeds 80% of its usual rating, and for which OCPD sizing would therefore be based on 125% of the continuous rating.

Cheers, Wayne

 

[pv_n00b]

The term "3-hour average" implies a consecutive 3 hour window, so the first option is unnecessarily conservative. Unspecified is whether the 3 hour window should be sliding, and if so how frequently it should be recalculated. But if you have simulated hourly data, then recalculating it every hour is reasonable.

Cheers, Wayne

I agree with you, but I see the first option being used all the time, and it's the method used at my current company.
I noticed in the first draft of the 2026 NEC the 100kW minimum limit has been removed for the use of 690.8(A)(1)(a)(2).