Technical note

Solar Sourcing FAQ: SMA Solar Data, Module Manufacturer Vetting, and the TCO Math Most Buyers Skip

2026-09-24 · Yuna Park

What this actually answers

I've been buying PV hardware for a mid-sized EPC and distribution outfit for six years now. Roughly $2.4M in cumulative spend across inverters, modules, and microinverter kits, and I've logged every PO in the same procurement system since 2019. Same questions keep coming up from teammates, from newer buyers, and from the occasional distributor who calls to ask how we decide. So here's the short version.

  • How do I evaluate a photovoltaic module manufacturer before I sign anything?
  • What do SMA Solar's shipment and revenue figures actually tell a buyer?
  • Microinverter specs — which ones change the cost, which ones don't?
  • Why the cheapest quote so often loses
  • What belongs in a TCO sheet for PV hardware
  • How often to re-check a vendor's financial health
  • The one question I now ask every single vendor

How do I evaluate a photovoltaic module manufacturer before I sign anything?

Four things, in this order: certifications, warranty terms in writing, third-party audit trail, and bankability. That last one is the part most buyers skip.

On certifications, you want to see IEC 61215 (design qualification) and IEC 61730 (safety) on the actual model you're buying — not on the manufacturer's marketing page. Model-level certs, not company-level. I learned that the hard way in 2022 when a distributor quoted us a "Tier 1" panel and the certification on file covered a different frame series entirely.

Warranty terms: read the exclusions before you read the coverage. A 25-year performance warranty means nothing if the replacement labor isn't covered and the shipping is on you. On a 40-panel residential array, that's roughly $1,800-$2,400 in labor and freight per claim that the buyer absorbs.

Bankability is where SMA Solar's annual report comes in — more on that below.

What do SMA Solar's shipment and revenue figures actually tell a buyer?

They tell you whether the company will still be honoring a 20-year warranty when you need to file a claim. That's the whole point.

Every spring I pull the SMA Solar annual report (the 2023 report, published in 2024, is the one I reference most often). I'm not going to quote the exact GW figure or EUR revenue off the top of my head — verify it at the source, because the number gets revised in quarterly updates and I don't want to hand you a stale figure. What I will tell you is what to look for once you're in there:

  • Reported shipments in GW — direction and stability matter more than the absolute number.
  • Revenue split by region — if 80% of revenue comes from one market, your warranty exposure is tied to that market's health.
  • R&D spend as a % of revenue — a vendor cutting R&D to hit margin is a vendor whose replacement parts will get discontinued early.
  • Warranty provision on the balance sheet — this is the line item almost nobody reads. It's an estimate of future claims. Compare it to shipment volume year over year.

When I compared two inverter vendors' disclosures side by side — same nominal warranty length, very different warranty provisions — I finally understood why the "equal" warranties weren't equal at all.

Microinverter specifications — which ones change the cost?

Most of the spec sheet is noise for procurement purposes. Three numbers move your total cost: output power (VA), MPPT channel count, and warranty length. CEC efficiency is real but the delta between a 96% and 97% unit is usually under $40 in lifetime energy on a residential system — not the deciding factor it's often made out to be.

Output power and channel count determine how many units you need. A 4-channel unit at 350VA per channel replaces two 2-channel units, which cuts your per-unit installation labor roughly in half. On a 20-panel job, that's the difference between 20 mounting points and 10. At our labor rates that's about $600 in the TCO.

I went back and forth between string and micro for three weeks on a 180kW commercial job. String won on capex, micro won on per-panel monitoring and rapid shutdown simplicity. We ended up hybrid — string on the main array, micro on the shaded east-facing row. Not elegant, but the TCO calculator said it was right and the client signed it.

Why does the cheapest quote so often lose?

Because "cheapest" is the starting number, not the ending one. Everyone told me to build a TCO sheet before comparing inverter quotes. I only believed it after skipping that step once and eating an $800 mistake on a small commercial install.

That job: Vendor A quoted $1,240 all-in. Vendor B quoted $980. I nearly went with B until I actually opened the fine print. B charged $55 per unit for commissioning support, $180 for extended monitoring access past year one, and $40 per unit for the mounting brackets that A bundled. On 8 inverters, that's $440 + $180 + $320 = $940 in add-ons. B's real number was $1,920. A was cheaper by 35%.

Now I don't compare quotes. I compare TCO sheets. The quote is just the top line of the sheet.

What belongs in a TCO sheet for PV hardware?

The line items I track, based on roughly 400 POs over six years:

  • Unit price (obviously)
  • Freight and import duties — and whether it's DDP or FOB, because that changes everything
  • Commissioning and setup support fees
  • Monitoring subscription cost past year one
  • Spare parts availability and per-unit cost
  • Warranty replacement labor allowance — who pays the electrician
  • Downtime cost per incident (this one is always underestimated — I use $0.18/kWh foregone on commercial, $0.14 on residential against local rates)
  • Currency exposure if the invoice is in EUR or CNY against a USD contract
  • Payment terms cost — a 2% early-pay discount is worth more than a 1.5% cheaper unit price on 60-day terms

Items 5 through 9 caused roughly 62% of our "budget overruns" between 2019 and 2023. We now require quotes on all nine lines before a PO goes out. Overruns dropped by half.

How often should I re-check a vendor's financial health?

Annually at minimum, and again before any PO above $50k. I pull the annual report, the most recent quarterly, and any credit rating change. Fifteen minutes of reading beats a 20-year warranty from a company that ships its last spare part in year three.

One thing to watch: "reviewed" and "approved" are not the same word. I said we'd "review" a distributor's updated financials. Our ops lead heard "approved the distributor." Two weeks later we'd issued a $68k PO. Nobody did anything wrong — we were using the same word with different meanings. Now the policy is written: financial review sign-off requires a dated email with the words "approved for PO issuance" in the subject line. Boring. Works.

The one question I now ask every vendor

"Walk me through what your standard warranty actually excludes, in your own words."

Not the spec sheet version. The spoken version. That's where you find out whether the distributor knows the product or is just relaying manufacturer documents. The good ones answer in under 90 seconds and mention at least two exclusions you didn't know about. The bad ones start with "well, the standard warranty covers..." and never get to the exclusions.

Ask it before you ask for price. Ask it before you build the TCO sheet. The answer changes how you weight every other line item on the list.

Previous: I Stopped Buying Inverters on GW Numbers — Including SMA Solar's 2023 FiguresNext: SMA Solar Shipments 2023 GW: The Research Rabbit Hole That Cost Me a Bid