June 2022, 7:40 a.m. — the phone call
My phone rang before I'd finished my first coffee. It was Marco, our lead installer, standing on a flat warehouse roof somewhere between Linz and Wels. He'd commissioned eight string inverters the day before. Seven of them were throwing DC overvoltage faults before the sun was fully up.
I've been sourcing solar components for commercial EPC projects for about 7 years now — roughly 40 jobs a year through my desk. My first thought was firmware. Second thought was a wiring error. Third thought, which was the right one and should've been the first one: we'd sized the wrong inverter for the modules we actually installed.
What actually went wrong
The job was a 118 kW retrofit — 214 solar modules on a warehouse roof, four strings per MPPT, standard layout.
About three weeks before we placed the inverter order, the customer switched panel suppliers. Better price per watt, slightly higher output. Good news, on paper.
I approved the change without re-running the string sizing. That was the whole mistake.
The new solar panels had a different open-circuit voltage — about 4 volts higher per module at cold temperature than what I'd based the original design on. On a 22-module string, at 5°C in the morning, Voc climbs past what the inverter can take. Protection kicks in. Every module on that roof is fine. Every inverter in that row is faulted.
We had to pull strings down, re-wire them into shorter runs, add two more inverters — which we paid for, because the original hardware had already shipped — and re-commission everything.
Total damage: $4,380 in parts, labor, and expedited freight. Plus 9 days off a customer deadline I'd already put in writing.
Nobody got hurt. No hardware burned. But that's a version of "fine" I don't want again.
What I got wrong about "support"
Here's the part I actually want to write about.
In February 2023 I was on a different job — another commercial rooftop, this one with a customer who'd specified a Fronius solar inverter setup from day one. GEN24 Plus, specifically, because they wanted the hybrid and battery path available later even if they weren't buying storage that year.
My assumption going in: Fronius solar support would work like support works everywhere else. You email, you wait, someone sends you a datasheet you already have. I'd worked with 6 or 7 inverter brands by then and support was mostly reactive — you call when something breaks.
What their technical rep actually did: asked for the full solar module spec sheet before production, including temperature coefficients, and came back two days later with a one-page compatibility note.
Two things flagged that I hadn't caught. Our string length was fine on max voltage, but the top of the MPPT operating window was a problem. And the DC isolator we'd spec'd was the wrong current class for that configuration.
Cost of those two catches if they'd shipped? Probably in the $1,500 range, plus a schedule slip. I don't have a precise number because we never had to find out.
Maybe that rep was just having a good week. But the convention I'd been running on — that compatibility is 100% the buyer's job and the vendor exists only after something fails — turned out to be wrong. At least in this case.
Where microinverter OEM vs private label actually gets tricky
Same year, we quoted two parallel sub-projects for a mixed portfolio client. One was a straight microinverter job, and the client asked us to price both an OEM microinverter family and a private-label version of the same unit out of the same factory.
On unit cost, private label won. Usually does. 12–18% cheaper per unit isn't unusual, and on a 60-unit job that's real money.
But here's what most people don't realize when they compare microinverter OEM vs private label: the certification paperwork is on the unit, but supportability isn't. UL 1741 SB and IEC 62109-1/-2 apply to the hardware. Firmware update channels, RMA handling, spare stock, and — critically — who picks up the phone in year four when a unit fails on a roof — those are commercial decisions the manufacturer makes, and they change depending on which version you bought.
We went OEM on that job. Not because private label was bad, but because the warranty handling on the private-label version routed through a distributor we'd never worked with, and I wasn't going to explain that to a client in 2027.
The 12-point check I actually use now
I'm not going to pretend this is an industry-standard document. It's mine. It came out of that June 2022 morning and I've edited it maybe 40 times since.
- Module temperature coefficient (Voc, %/°C) — read from the actual spec sheet, not a summary.
- Voc at site record low temperature, per string.
- Vmp at highest expected cell temperature, per string.
- MPPT operating window — both ends, not just the middle.
- Inverter max DC input voltage, with margin.
- Inverter max input current per MPPT.
- DC isolator current class and voltage rating.
- Market certification with the file number (UL 1741 SB for US deployments; IEC 62109-1/-2 for safety).
- Who reviews the sizing before production?
- Who answers the phone on a Saturday during commissioning?
- Warranty terms and RMA turnaround on a failed unit.
- If hybrid or storage — actual battery and EV charger roadmap, not brochure talk.
Points 9 and 10 are the ones that changed my behavior. Everything else is just careful reading.
I want to say this checklist has caught 47 potential issues in the last 18 months — I'd have to dig through the log for the exact number, but that's roughly right. The most concrete saving we can point to is about $8,000 in rework that we didn't have to do.
What I actually took away
Five minutes of verification beats five days of re-stringing modules on a roof you already billed.
That's it. It's not clever. It doesn't make a good conference slide.
The conventional wisdom in solar procurement is that the inverter is the technical decision and everything else is logistics. My experience across roughly 280 commercial jobs suggests the opposite — the inverter choice is usually the easiest part, and the compatibility check between the solar module and the inverter is where the money actually hides.
One last thing, because it comes up: if you're publishing anything about system output or "clean energy" claims, FTC Green Guides (16 CFR Part 260) apply to solar marketing the same way they apply to recyclable packaging. Unsubstantiated performance claims are a marketing problem, not just a legal one.
The checklist is boring. I think that's kind of the point.
