Using one multi-port microinverter to support multiple separate arrays + Rapid Shutdown

[zanydroid]

I currently have a rooftop install with 10 HM-800 2-port microinverters, and I'm looking to expand with 2- and 4-port microinverters. These are pretty cost competitive with string inverters and optimizers / rapid shutdown only strings.

I'm trying to understand the viability of using one of the 2- or 4-port microinverters for panels installed on small roof planes that can only fit 1 or 2 panel.

EDIT: To clarify. The situation is 4 separate arrays of one panel each. So the DC cables have to stretch across arrays. DC conductor will exceed 30V due to Voc as well as typical operating voltage.

This would require running DC wire from the panel to the microinverter, at a length that will mandate the use of rapid shutdown devices. That will increase the cost of this system.

Wireless RSD MLPE seems like the easiest way to implement this. However, I haven't found any wireless ones yet. If you know of one, please let me know.

As for Wired RSD MLPE, I would like to understand the theory of operation of implementations such as TIGO, APSystems, and SMA. My understanding is that they use a core encircling a DC conductor to modulate a PLC signal onto that conductor.

• Is the PLC signal a keep-alive that is sent if and only if there is power to the transmitter? So to initiate rapid shutdown, power is removed from the transmitter
• How does the PLC signal propagate through multiple MLPE? Does it only open one polarity across all MLPE on the same string, leaving continuity on the other polarity for the PLC to propagate through?
• I believe they are somewhat interoperable since they (I think) all implement SunSpec

Code questions

• Is the listing of the microinverters appropriate for this use?
• How do you install the transmitter in a code compliant way indoors? Are you allowed to use a PVC box, or does the box have to be metal? 690.31(D) suggests that the DC cable is only allowed in metal enclosures. In NEC2020 there is an exception for PV hazard control systems but I can't parse the English.
• Can the transmitter share the same circuit as the microinverter, so that it shuts down at the same time that power is cut to the microinverter branch circuit.

Cost analysis

• Conclusions
  • Using multi-port microinverters with RSD devices can definitely save cost over using single microinverters. However, the labor will be higher.
  • Likewise, they can save cost vs the smallest string inverter sizes

Note: does not count the cost of wires, including proprietary trunk cables

Single port microinverters: HM300 / IQ8+ are around $170-190
Two port microinverter (700W) $260
Four port microinverter (1500W) $365 * Note: 2 MPPT / 4 port, so not comparable to two two-port inverters

TIGO RSD are:

• $30 single port / $50 dual port (note these must be located in one place)
  • This is comparable to a trunk cable in cost
• Transmitter is $50, needs a $15-$20 12V DIN power supply + metal enclosure and DIN rail. Probably $100 total
• Transmitter can control 10 strings for 150A total current, 20 strings with two cores.

It is not at all cost effective to use these with a 2-port microinverter. The cost of a single port is identical to just getting a single port microinverter. Maybe if there is a cost-reduced transmitter that is integrated with a junction box, a power supply, and smaller coil, the cost would be a little more optimal.

2x HM700: $520 (4 MPPTs)
1x HM1500: $365, difference of $155 (2 MPPTs)
4 single port microinverters = $680. (4 MPPTs)

If you have 3 panels remote from an HM1500 (there is a 1 panel local array sitting above the HM1500), you can use a single transmitter.
$365 + $100 + 3 * $30 = $555
If the remote panels are on different roof planes then the fact that this has 2 MPPTs can degrade efficiency.

Suppose you have 6 panels scattered all over the place, and you have a 2 panel array sitting above 2 HM1500s.
2 * $365 + $100 + 6 * $30 = $1010
vs 4 single port microinverters = $1360

Note this approach has similar AC capacity to using a 3-3.8kW string inverter. The string inverter has much greater overpaneling flexibility than microinverters. The microinverters may end up having more MPPT channels.

Compared to TIGO optimizers, HM micros are obviously more capable, since TIGO optimizers can only buck within a limited voltage range to match the current of two panels connected in series. While I am not sure about the power topology of HM1500, which have 2 MPPT with 2 ports each, I think it is safe to assume that they are wired in parallel.

 

[timselectric]

Is the PLC signal a keep-alive that is sent if and only if there is power to the transmitter?

Yes
It's a radio signal that is using the circuit wiring as an antenna.
Shut down happens when signal is not seen. (Transmitter loss of power)

 

[Shimmy]

Sorry, not getting why you need RSD for a multi-port microinverter. The panels are never tied in series (DC), so once the AC power is cut the panel voltage is under the threshold, no?

 

[zanydroid]

Sorry, not getting why you need RSD for a multi-port microinverter. The panels are never tied in series (DC), so once the AC power is cut the panel voltage is under the threshold, no?

I edited the thread title and OP for clarity

Title: Using one multi-port microinverter to support multiple separate arrays + Rapid Shutdown

To clarify. The situation is 4 separate arrays of one panel each. So the DC conductors have to stretch across array boundary. This means RSD is needed because the DC conductor will exceed 30V due to Voc as well as typical operating voltage.

Multi-port microinverters are designed to source power from solar panels in the SAME array. This allows the inverter to satisfy RSD code by simply not having DC exit the array. Moreover the operating voltage on each cable within the array is <80 VOC.

 

[zanydroid]

Yes
It's a radio signal that is using the circuit wiring as an antenna.
Shut down happens when signal is not seen. (Transmitter loss of power)

Do you know why there is a current limit on the bundle of conductors that are used to transmit the PLC? Is it due to induced current in the coil on startup / shutdown of the system?

 

[timselectric]

Do you know why there is a current limit on the bundle of conductors that are used to transmit the PLC? Is it due to induced current in the coil on startup / shutdown of the system?

I assume that it is to avoid signal degradation over multiple conductors/antennas. And interference from the current flowing on the conductors.