What to Look for in a Solar Panel Supplier: A Procurement Manager's TCO Checklist

Short answer: if the supplier can't walk you through TCO, walk away

If you take one thing from this piece: ranking solar panel suppliers by price-per-watt is where good procurement goes to die. The number that actually matters is total cost of ownership — landed module cost plus everything that happens after the pallet hits your warehouse. In my experience, a quote that looks 3–4 cents cheaper per watt can end up costing more by the time the project reaches commissioning.

Let me put a number on it. In Q2 2024, I reviewed nine bids for a utility-scale module order. The cheapest came in around $0.10/W. The third-most-expensive came in around $0.13–0.14/W. I built out a 25-year landed cost model — freight, insurance, port delay probability, warranty claim friction, replacement logistics — and the "expensive" one came out roughly 3% cheaper over the modeled horizon. Not a rounding error. Not at scale.

That's the whole game. Everything below is just supporting evidence.

Why I'm telling you this

I'm a procurement manager at a ~900-person renewables developer. I own our annual solar and wind equipment spend — somewhere around $40M — and I've held that responsibility for seven years. I've probably reviewed 200-plus purchase orders over that span; 180 sounds more realistic, I'd have to pull the log. Every one of them is in our internal cost-tracking system. I've been tracking line-item variance since my second year on the job.

I'm not an engineer. I can't tell you whether PERC or TOPCon is better for a specific irradiance profile — I have engineers for that. What I can tell you is where money gets lost. And it's almost never on the spec sheet. It's in the footnotes.

With pure-play developers like Adani Green Energy publicly targeting 50 GW of renewable capacity by 2030, the volume of utility-scale module demand keeps compressing margins on the supply side. That makes TCO discipline more important, not less — because when per-watt prices converge, the differentiator moves to everything around the panel.

The four things that actually drive solar supplier TCO

1. Warranty — and what sits behind it

A 25-year linear output warranty sounds solid. It only means something if it holds up in year two. I got burned on this exactly once — I trusted a sales rep's verbal assurance, didn't get the replacement RMA process documented in the contract. First batch of modules arrived with junction box defects. Six weeks of back-and-forth. Roughly $12,000 in rework and storage costs, not counting the interconnection delay interest.

This is now in our standard terms: warranty isn't about the years, it's about the year-one response SLA and the dispute path. Put another way — a 25-year warranty from a vendor who won't answer emails in month one isn't a warranty at all.

2. Certification isn't a checkbox

Everyone lists IEC 61215, IEC 61730, IEC 62804 on the slide deck. Fine. Ask which certification body issued them, when, and — most importantly — which specific product variant the certificate covers. I've seen bids where the certificate was for a Japanese fab line while the actual shipment came from a contract-manufactured Malaysian batch (which, honestly, had different degradation curves). The cert looked clean. The product didn't match.

3. Freight and yield rate

Modules are heavy and fragile. Freight as a percent of TCO is routinely underestimated. But the bigger variable I've learned to weight is yield rate — the percentage of units that arrive undamaged, correctly labeled, and ready to install. Some suppliers run 3%+ damage rates. That cost comes back to you, either as claim-processing overhead or as direct write-off. Neither shows up in the price-per-watt.

4. Delivery schedule risk — this is not a project-team problem

Procurement teams often hand this off to project management. Don't. A supplier who slips delivery by six weeks against an interconnection window doesn't just annoy your PM — that's IRR. It doesn't appear on any quote. It's real.

People assume rush orders cost more because they're harder to produce. The reality is they cost more because they're unpredictable and disrupt the supplier's planned workflow. The premium isn't labor. It's scheduling chaos.

Three things to check before you look at any quote

• Do you have a TCO model? Build it before you open the first bid. Otherwise price-per-watt will anchor you whether you intend it or not.
• Does your contract spell out response time and replacement authority? If not, you don't have a warranty — you have a promise.
• Do you have at least three technically comparable suppliers with different cost structures? I only made this mandatory after a single-source emergency buy cost us roughly $12,000 over what a competitive process would have. Everyone told me to always get three quotes. I only believed it after ignoring that advice once.

On the wind side, the same discipline applies — blade logistics, gearbox warranty terms, and turbine delivery sequencing follow the same TCO logic as modules. The unit economics differ; the evaluation framework doesn't.

When this approach doesn't apply

If you're procuring for a small distributed project — rooftop, sub-1 MW — price-per-watt probably is your primary driver. Volume is low enough that freight variance and warranty friction get diluted. Running a 25-year TCO model on a 200 kW order is overkill.

Same caveat if you don't have real historical data. If you've never worked with a given supplier and can't get reference data from their existing customers, your TCO model is fiction. In that scenario, asking three peer buyers for actual damage rates and claim outcomes beats whatever the sales deck gives you.

One more thing — scale benefits aren't linear. A TCO framework built for a 400 MW pipeline doesn't automatically translate to a 40 MW one. Scale is the most consistently ignored variable in TCO analysis. If I remember correctly, the deviation between our modeled and actual landed cost was around 4% at 400 MW and closer to 11% at a sub-100 MW project. I don't have the exact logs in front of me, so treat those as approximate.

Bottom line: the framework travels. The specific numbers don't. Build your own model, pressure-test it with real supplier data, and check what's not in the quote before you sign anything.