Does 290.64(B)(2) apply to feeder frm rooftop PV inverters to bsmt service?

Status
Not open for further replies.

yeknom

Member
We are installing solar panels on the roof of an apartment building. Each of three inverters on the roof generates 34amp, 208v, 1-phase power which then feeds via three 50a 2-pole breakers into a combining panel with a 90 amp 3-phase main breaker. From this 90 amp breaker feeders descend 200' to a 100amp switch (fused at 90amp), mounted in an existing main distribution panel (MDP) in the basement electrical room.
The question arises: How to determine the size of the feeder from the basement to the roof? The max current generated by the PV panels is 68 amps in each of the three phases. Considering this alone, a # 4 THHN (95a rated) wire is appropriate (with 1.25 factor for continuos use), even considering voltage drop in the 200' run.

However, some have interpreted NEC 2005 290.64(B)(2) as applying to this feeder which is supplied with power via two over-current devices (90a on the roof and 90a in the bsmt) and that therefore the rating of the conductor must be at least 180 amps (90 + 90) (2/0 THHN). The suggestion is that someone could tap into this riser to feed loads which could cause the current in the riser (fed by BOTH the PV sources on the roof and the utility power in the basement) to exceed its 95 amp rating without activating either OCPD.

My interpretation of the code is that it applies only to busbars and conductors at the "point of connection", not to such a riser. In this view it makes sense to protect busbars & conductors at the building service in the basement where additional loads might be applied, not to protect against some fool indiscriminately tapping into risers dedicated to supply power to the building from a rooftop PV system.

What is the correct interpretation?
 
First, it appears your current calculation is incorrect. When you have three 34A 208V 1? sources, that I assume are synchronized for the 3? combining, the current on each line after combining is 34A ? 1.732 = 59A.

The feeder and its PV-end ocpd must be sized based on the output rating of the inverters. Is that 34A each? Regardless, the ocpd and feeder are sized to 125% of the inverter output rating... so if it is 34A, your ocpd and feeder minimum rating would calculate as 34A ? 1.732 ? 1.25 = 74A

BTW, I believe you meant 690.64(B)(2)... which does not apply to this feeder. First off, you say it could be tapped in the future... well that'll be the problem of whomever does the tapping... but in the meantime, this section does not apply to the feeder as discussed.
 
Check the voltage drop for 200'

Check the voltage drop for 200'

The serious question you need to address is the voltage drop from inverter to utility. This results in the inverter operating at a higher voltage than the utility and the upper voltage limit of the inverter may be reached. The usual specification is to allow no more than a 1.5% drop (1% is better). Most inverters have installation manuals with charts of wire size vs distance, or consult the manufacturer.

I calculate that a 12 KW 3-ph inverter on a 120/208 would need 1/0 cable for a 200' run for 1% drop. You appear to have three 6,000 or 7,000-watt inverters and would need even heavier cable.

You may want to consider DC feeds to an inverter that is closer to the main panel. Then voltage drop is mostly an economic decision in which the power lost can be overcome at lower cost by adding PV modules. The inverters accept a wide input range.
 
Status
Not open for further replies.
Top