What No One Tells You About Light-Induced Degradation (And How It Changed Our Solar ROI Calculation)

When I first started modeling the ROI for our company's solar installation, I assumed the biggest risk was the weather. Cloud cover, seasonal variance, that kind of thing. I had a spreadsheet with 48 columns, five-year cash flow projections, and a pretty confident sense that I'd covered all the bases.

I was wrong. The thing that actually threw my numbers off—by about $12,000 in cumulative energy yield over five years—was something called light-induced degradation (LID). And I almost didn't account for it at all.

Here's what happened, what I learned, and why our eventual choice of First Solar modules came down to a quiet advantage that doesn't show up on spec sheets.

The Assumption That Almost Cost Us

In Q3 2024, I was tasked with evaluating solar panel vendors for our 1.2 MW rooftop project. I'd read all the standard efficiency ratings, temperature coefficients, and warranty terms. I felt ready.

Then I got a quote that was 6% cheaper than the others. My first instinct: “Let's go with Vendor B and save $14,000 upfront.”

Luckily—or maybe out of paranoia from a previous incident with a “free setup” that ended up costing us $450—I decided to dig deeper. I built a total cost of ownership (TCO) model that factored in not just the purchase price, but the expected degradation curve over 25 years.

Standard degradation assumptions:

• Year 1: Most panels lose 2-3% from initial output due to LID.
• Years 2-25: Linear degradation of 0.5% to 0.7% per year.

This is the industry baseline. It's widely accepted. But here's what I missed: not all panels behave the same in Year 1.

The First Solar Difference (and a “Pretty” Shocking Discovery)

I compared three vendors: one using conventional polycrystalline silicon panels, one using mono-PERC with a claim of “anti-LID” treatment, and First Solar's Series 6 CdTe thin-film modules.

In my model, I applied the same Year-1 degradation to all three—about 2.5%, which seemed safe. Then I found a White Papers from NREL and First Solar that showed something surprising: CdTe modules experience minimal light-induced degradation.

Actually, it's not just minimal. For thin-film CdTe, the major initial loss mechanism (boron-oxygen complex formation in silicon) doesn't exist. The NREL study (reference: NREL/PR-5200-68790, “Light-Induced Degradation in CdTe Solar Cells”) reported that well-manufactured CdTe modules can exhibit less than 1% LID in Year 1.

That 1.5% difference in Year-1 output might sound small. Let me tell you: over a 25-year PPA, that's not small.

Using the NREL degradation formula:

• Year 1 output: 100% - LID (2.5% for typical c-Si, maybe 1% for First Solar)
• Years 2-25: Annual linear degradation (say, 0.5% for both)

For a 1.2 MW system generating roughly 1,800 MWh/year in Arizona (where we are):

• C-Si panel with 2.5% LID: Year 1 yield ≈ 1,755 MWh
• First Solar CdTe with 0.8% LID: Year 1 yield ≈ 1,786 MWh

A 31 MWh difference in Year 1 alone. At $0.06/kWh PPA price, that's roughly $1,860 in Year 1 revenue — and the gap narrows but persists over the life of the system.

When I corrected my model, the total projected revenue difference over 25 years between the “cheap” vendor and First Solar was about $12,000 in favor of First Solar, despite the higher upfront cost.

The Procurement Lesson: Discount Rate vs. Data Sheet

I'll be honest: I initially wanted to go with the cheaper vendor. My cost controller brain saw the lower sticker price and the only thing I could think was, “That's $14k back in my budget.”

But after this exercise, I changed our procurement policy for solar projects. We now require:

1. A degradation projection from the manufacturer, not just the marketing spec
2. A TCO calculation that includes Year-1 LID assumptions, not just linear decay
3. Third-party validation (NREL, PVEL, or similar) of degradation claims

Power = Voltage × Current (P = V × I). That's the equation. But the real equation for a solar investment is: Total Energy Harvested = Nameplate Capacity × (1 - Degradation) × Hours of Sunlight × System Efficiency. And the degradation term has a first-year multiplier that matters a lot.

The calculated output for a 400W First Solar module at Standard Test Conditions (STC) is:

• Peak power: 400 W
• Voltage at peak power (Vmp): ~33.6 V
• Current at peak power (Imp): ~11.9 A

This is the real scale: one module, about 2 meters by 1.2 meters, producing roughly the same daily energy as a mid-size air conditioning unit running for 3 hours.

We ended up installing 3,000 of them. The first year output? On track to beat the conservative estimate by about 1.2%, per our monitoring system.

The Bottom Line

Light-induced degradation is a quiet variable. It's not something most procurement checklists catch. But if you're buying solar for a business, especially net zero solar generators, this is a make-or-break assumption.

I can only speak to our specific context: Arizona, a 1.2 MW rooftop install, a 25-year PPA. Your mileage may vary if you're in a cloudier climate or dealing with a different module technology. But the principle holds: the degradation assumption in Year 1 matters more than most people think.

Oh, and one more thing: when you're comparing quotes, ask for the degradation warranty. First Solar's is linear to 86% of the rated power at year 25—above the industry standard. That's not an accident. It's a design choice that shows up in the PPA pricing.

Trust me on this one. I have the spreadsheet to prove it.