They provided further details privately. The conclusion was the pricing of the Hoymiles HSRD-2C devices was not materially different than the commonly available and seemingly equivalent Tigo TS4-A-2F devices. I dislike the RSD requirements, so anything I can do to reduce the cost impact of them is worth looking into. The UL 3741 change will basically eliminate them for the most part, but timing is such that code and vendors have not caught up to that yet, sadly.
For new installs, it is fair to compare DC versus AC. From my perspective, DC coupled is going to be more efficient since PV to battery has no unnecessary mid path conversion to and from AC produced by micro inverters. So your energy in the battery gets there with much higher efficiency. The battery to load/grid will be the same for both systems, so overall PV to battery to load is better on DC coupled. This is not news.
An AC system with battery means you buy two inverters, the array of micro inverters on the panels, and then another one that is battery to load/grid. The Enphase battery system even uses plugged in micro inverters to make that explicit. A DC coupled system only has one inverter, PV and/or battery to load/grid.
Their micro inverters due seem generously rated. Hopefully this doesn't impact their efficiency at partial loads too much.
When I pencil out the systems I am considering, DC coupled string inverters win on cost and efficiency. Long term, a string inverter inside the building seems like it will last longer than the extreme temperature cycling of the micro inverters on the roof. The go from the cold of night to the baking heat of full sun under a hot solar panel and that temperature cycling wears out electronics.
Mike C.
All good information. Thank you. A few things I'd like to point out about having a single string inverter vs microinverters + AC coupled inverter.
1. You're right. Whatever efficiency the battery charger has will be 3% to 3.5% lower when switching from PV to AC to DC. However, this configuration with Hoymiles' inverters has maximum redundancy with minimum components. I like redundancy! I come from the UPS Market, where critical computer loads require 99.999% uptime, guaranteed.
2. A single-string inverter can fail ("cough" SolarEdge), and replacing it can take days, weeks, or more. If it's down for more than 10 days, then that 3% efficiency gain is lost for the whole year.
3. With these microinverters, if one fails (None have failed yet for any of my customers!), the rest keep right on working, and if the AC-coupled battery inverter fails, the micros don't care and keep right on producing.
4. Since there are no PV connections to the AC-coupled inverter, if it does fail, it's less laborious and less dangerous to replace.
5. The system does not contain HVDC, making it great for DIYers who are still learning about electricity, and is much less prone to ground faults starting fires.
6. Installing microinverters allows the cash-strapped DIYer to do the project in phases. Go grid-tied first because it costs less, and then add the AC-coupled inverter and batteries later without going back on the roof.
I used to design giant utility-scale systems, where every fraction of a percent of efficiency mattered. On residential-scale systems, IMO, the difference of a few percent is not a big deal if it means greater reliability and less downtime. Downtime due to a single central or string inverter failure is always unexpected and costly.