Solar Module Degradation & Warranties Explained

Short answer: every solar module carries two separate warranties — a product warranty covering defects in materials and workmanship, and a performance (linear power) warranty guaranteeing a minimum percentage of nameplate output in each year of a 25- or 30-year term. The performance warranty is defined by two numbers: the first-year allowance and the annual degradation rate after that. Everything else — measurement method, labour, freight, transferability, exclusions — sits in the fine print and decides whether a claim is worth anything.

For a buyer, the warranty is not a marketing badge. It is a 30-year financial commitment that sets your energy model's back end and, in project finance, part of your bankability case. This guide explains how to read one properly.

The two warranties on every datasheet

Datasheets usually print something like "15-year product warranty / 30-year linear power warranty". These cover completely different failure modes and often have different remedies.

Two practical consequences follow. First, a long performance warranty behind a short product warranty is weaker than it looks: from year 13 to year 30 you may be covered for gradual power loss but not for a failed junction box. Second, the performance warranty only pays when you can prove underperformance — which is a measurement problem, not a legal one.

How a linear performance warranty actually works

A linear warranty is a straight line with a step at the start:

• Year 1 allowance — the module may lose up to this much in its first year (light-induced effects plus stabilisation).
• Annual degradation rate — the maximum additional loss guaranteed for each year after that.
• End-of-term value — what the two numbers add up to at year 25 or 30.

The arithmetic is simple, and worth doing yourself rather than trusting the headline:

Guaranteed output in year N = 100% − (year-1 allowance) − (annual rate × (N − 1))

So a curve quoted as "1% first year, then 0.4% per year" over 30 years guarantees 100 − 1 − (0.4 × 29) = 87.4% at year 30. A curve that ends at 90% in year 30 with the same 1% first-year step implies an annual slope of about 0.31% — because (100 − 1 − 90) / 29 ≈ 0.31. Any datasheet that quotes an end-of-term figure inconsistent with its own first-year and annual numbers deserves a question.

Why the slope matters more than the end point

Because the loss compounds over the whole term, a modest difference in slope changes the lifetime energy, not just the final year. Take two 30-year curves on identical nameplate:

• Curve A: 1% first year, 0.31%/yr after → about 90% at year 30, averaging roughly 94–95% of nameplate across the term.
• Curve B: 2% first year, 0.55%/yr after → about 82% at year 30, averaging roughly 90% across the term.

That gap — around 4 to 5 percentage points of average lifetime output — is a 4–5% difference in total generated energy from the same installed capacity, with identical mounting, cabling, labour and land. It is one of the few remaining line items where a purchasing decision quietly changes revenue for three decades.

Read the measurement basis

A warranty curve is meaningless without the measurement rules attached to it:

• Nameplate reference. Is the guarantee measured against the label power, or against label power minus the measurement uncertainty allowance (typically around 3%)? The second is much weaker.
• Power tolerance. Modules are usually sold with a positive-only tolerance (for example 0 to +5 W). Confirm that the warranty baseline is the label value, not the top of the bin.
• Test conditions. Claims are normally settled on a flash test at standard test conditions (STC) in an accredited lab, or by an agreed on-site irradiance-corrected measurement. Ask which, and who pays for it.
• Sampling. For a large plant, is a claim assessed module by module, or on a statistical sample of the affected batch?

What actually causes module degradation

Understanding the mechanisms tells you which warranty clauses matter for your site.

• Light-induced degradation (LID). A first-day/first-weeks loss driven, in p-type silicon, largely by the boron–oxygen defect. N-type wafers do not have that defect, which is why first-year allowances shrank as the industry moved to n-type.
• Light and elevated temperature-induced degradation (LeTID). A slower, partly reversible loss seen mainly in certain p-type architectures under heat and current; a specific question for hot-climate projects.
• Potential-induced degradation (PID). Leakage current between cells and the grounded frame at high system voltage — worse in hot, humid conditions. Mitigated by cell/encapsulant design and by correct inverter grounding, and tested for under IEC 62804.
• UV-induced degradation. Slow loss from ultraviolet exposure at the cell surface and in the encapsulant.
• Thermal cycling and mechanical stress. Daily expansion and contraction, wind flutter and snow loading drive micro-cracks in cells and solder-joint fatigue. Most of this risk is set at installation: mishandled modules, over-torqued or wrongly positioned clamps, and walking on glass.
• Encapsulant and backsheet ageing. Yellowing, delamination and backsheet cracking let moisture in. Dual-glass construction with POE encapsulant is more resistant than glass–backsheet designs, which is why 30-year curves are usually offered on dual-glass products.
• Environmental attack. Salt mist near coastlines, ammonia near livestock buildings, and abrasive sand all attack frames, glass and junction boxes. These have their own test standards (IEC 61701 for salt mist, IEC 62716 for ammonia) — check the module is certified to them if your site needs it.

What "low degradation" claims should be backed by

A supplier saying "low LID and PID" should be able to point to the specific certificate or test report — PID testing under IEC 62804, salt mist under IEC 61701, ammonia under IEC 62716, and the base qualification below. If nobody can produce the report number, treat the claim as marketing.

The tests behind the numbers

Two standards underpin almost every commercial module sold today:

• IEC 61215 — design qualification and type approval: thermal cycling, damp heat, humidity freeze, mechanical load, hail impact, hot-spot endurance and more. It is a pass/fail qualification of the design, not a lifetime prediction.
• IEC 61730 — photovoltaic module safety qualification (construction and testing), with UL 61730 as the equivalent route for North American markets.

Two things follow that buyers often miss. First, IEC 61215 is a minimum bar — it demonstrates the design survives a defined stress sequence, not that it will hold a 30-year curve. Second, extended stress testing (longer thermal-cycle counts, sequential or combined stresses, higher damp-heat hours) is where genuine differentiation shows up, so ask whether a product has been through extended testing beyond the certification minimum, and ask for the report rather than a claim.

For utility procurement, also confirm which factory and which bill of materials the certificate covers. Certificates are issued against a specific BOM; a substituted encapsulant, backsheet or junction box can invalidate the assumption behind your reliability case.

The fine print that decides whether a claim pays

This is where nominally similar warranties diverge most.

• Labour, freight and crane costs. Many warranties supply a replacement module but exclude the cost of removing the old one, shipping, and re-installing. On a ground-mount that is inconvenient; on a high roof it can exceed the module value.
• Remedy choice. Repair, replace with an equivalent model, supply additional modules to make up the shortfall, or refund the missing watts at a depreciated value — the supplier usually chooses. Later in the term, "equivalent model" means a physically different module, so confirm mechanical and electrical compatibility rules.
• Transferability. Does the warranty follow the asset if the plant is sold, and does it survive a change of O&M contractor?
• Installation and O&M conditions. Expect requirements on mounting method, clamp zones, torque values, minimum earthing, inverter grounding scheme, cleaning agents and cleaning methods. Failure to keep records of these is a common reason claims stall.
• Claim window and notification period. Some warranties require written notice within a short period after the defect is discovered.
• Force majeure and site events. Hail above the tested impact energy, wind above the design load, flooding, lightning, animal damage and vandalism are typically excluded — these belong to insurance, not warranty.
• Who stands behind it. A warranty is only as good as the entity issuing it. Check the legal entity named in the document, how long it has operated, and whether it is the manufacturer or a trading company. Ask whether the cover is insurance-backed and, if so, by whom and for what scope.

What COTECH publishes

COTECH module warranty and durability terms come from the module datasheets rather than from any general claim:

Supporting durability data published on the same datasheets: certification to IEC 61215, IEC 61730, IEC 61701 (salt mist) and IEC 62716 (ammonia) plus UL 61730; an operating temperature range of −40 °C to +85 °C; mechanical test load of 5400 Pa front / 2400 Pa rear; 1500 V DC maximum system voltage; and resistance to PID with low LID/LeTID. Full electrical, thermal and mechanical tables for each series are in the module datasheets, and the range is summarised on the Solar Modules page.

If you are also deciding between cell platforms, the trade-offs behind these numbers are covered in TOPCon vs HJT vs back-contact and N-type vs p-type PERC.

A ten-point checklist before you sign

• Get both warranty documents, not the datasheet summary — the product and the performance warranty are separate contracts.
• Confirm the first-year allowance, the annual rate and the end-of-term percentage, and check they are arithmetically consistent.
• Confirm the measurement baseline (label power vs label minus uncertainty) and the claim test method.
• Check whether labour, freight and access costs are included in the remedy.
• Check the remedy hierarchy and who chooses it.
• Confirm transferability on sale of the asset.
• List the installation and O&M conditions you must document to keep cover valid — then build them into your commissioning pack.
• Ask for certification reports (IEC 61215/61730, plus salt mist and ammonia if relevant) against the exact BOM you are buying.
• Ask what extended stress testing the product has been through beyond the certification minimum.
• Identify the legal entity issuing the warranty and whether any insurance backing exists.

Modules are one of the few components where the paperwork is genuinely part of the product. Two modules with the same nameplate and the same efficiency can differ by several percent in lifetime energy and by a large margin in whether that difference is ever recoverable. If you want the underlying tables for a specific series, request the datasheets or send your project details and ask for the warranty document itself.

FAQ

What is the difference between a product warranty and a performance warranty on solar modules?

A product warranty covers defects in materials and workmanship — delamination, junction-box failure, frame corrosion, broken interconnects — and typically runs 12 to 15 years. A performance or linear power warranty covers gradual output loss and guarantees a minimum percentage of nameplate power in each year of a 25- or 30-year term. They are separate documents with different triggers and different remedies, so both should be requested before purchase.

How do you calculate the guaranteed output of a solar module in a given year?

Use: guaranteed output = 100% minus the first-year allowance minus (annual degradation rate multiplied by the number of years after the first). For example, a warranty with a 1% first-year allowance and 0.4% per year over 30 years guarantees 100 − 1 − (0.4 × 29) = 87.4% of nameplate at year 30. If a datasheet quotes an end-of-term percentage that does not match its own first-year and annual figures, ask the supplier to reconcile it.

Does a 30-year linear power warranty mean the module lasts 30 years?

No. The performance warranty is a commitment about output levels, not a predicted service life, and it does not cover defects once the shorter product warranty has expired. Reliability comes from the construction and the test evidence — dual-glass laminates, POE encapsulant, IEC 61215 and IEC 61730 qualification, and any extended stress testing beyond the certification minimum — while the warranty defines what happens financially if output falls short.

What causes solar panels to degrade over time?

The main mechanisms are light-induced degradation in the first weeks, potential-induced degradation from leakage current at high system voltage in hot and humid conditions, ultraviolet ageing of the cell surface and encapsulant, thermal cycling and mechanical stress that create micro-cracks and solder fatigue, encapsulant or backsheet ageing that admits moisture, and environmental attack such as salt mist, ammonia or abrasive sand. Careful handling and correct clamping during installation prevent a large share of avoidable loss.

Which warranty exclusions matter most in practice?

The costly ones are the exclusion of labour, freight and access costs from the remedy, installation and maintenance conditions that must be documented to keep cover valid, restrictions on transferring the warranty when a plant changes ownership, and site events such as hail above the tested impact energy or wind above the design load — which belong to insurance rather than to the warranty. Also confirm which legal entity issues the warranty and whether it is insurance-backed.

Do n-type modules degrade less than p-type PERC modules?

Generally yes, and the reason is structural rather than commercial: n-type wafers do not contain the boron–oxygen defect responsible for much of the light-induced degradation seen in p-type cells, so first-year allowances are smaller and annual slopes are typically flatter. That is why 30-year linear curves are now common on n-type dual-glass products. Always confirm the actual figures on the datasheet for the specific series rather than assuming a platform-wide value.