PID, LID and LETID: Catch Panel Degradation While the Warranty Still Pays
Every solar panel degrades, and every module warranty allows for it. The problem is the degradation that goes faster than the warranty allows. Potential-induced degradation (PID), light-induced degradation (LID) and light and elevated temperature induced degradation (LETID) each follow a different timeline, and each is easy to mistake for soiling, weather or a drifting sensor. PID can often be reversed if it is caught early, and a warranty claim is only as strong as the evidence behind it. Ellume Vector tracks degradation string by string against sister strings on the same site, and flags PID and early-life degradation while there is still time to act.
A module warranty is a promise about the future: this panel will still produce a stated share of its rated output in year five, year ten and year twenty-five. Most owners never test that promise. They see a plant-level performance ratio that drifts down a little each year, assume it is normal ageing, and move on. Sometimes it is normal ageing. Sometimes it is a degradation mechanism running well ahead of the warranted curve — and by the time it is obvious at plant level, the cheapest window to fix it, or to claim for it, has often passed.
Degradation runs on a clock. So does your warranty.
Who this matters to
Asset managers and owners with plants in their first years of operation, technical advisers reviewing a portfolio before acquisition, and O&M teams who are asked every year whether a site's decline is “within warranty” and have no clear way to answer.
Three mechanisms, three timelines
- •PID: the reversible one, if you are quick. In a long string, modules sit at a high voltage relative to the grounded frame. In p-type modules that stress lets sodium ions migrate towards the cell surface and creates leakage paths, which is why modules nearest the negative end of a string are usually hit hardest. Heat and humidity make it worse. Left alone, PID can take a large share of a string's output. Caught early, it is often at least partly reversible — for example with equipment that applies a reverse voltage overnight, when the plant is not generating. Left for long enough, the damage becomes permanent.
- •LID: the expected one, in the first weeks. Many crystalline silicon modules lose a little output in their first hours to weeks in sunlight, as defects in the silicon activate. In modules without other degradation issues this is typically 1–3%, after which it stabilises. It is expected, and warranties usually allow for it in the first year. The question is not whether it happened, but whether it stopped where it should have.
- •LETID: the slow one, in the hot years. LETID is slower and less predictable. It develops under heat over months to years, and in susceptible modules, particularly some PERC products, industry reports put losses at several per cent — more in some cases. Field studies suggest it can take years to reach its worst point and many more to recover partially. A plant can look fine through commissioning and its first year, then lose output steadily beyond the warranted curve.
| Mechanism | When it shows up | Can it be reversed? |
|---|---|---|
| PID | Months to years; high-voltage strings, worse in heat and humidity | Often partly, if caught early; permanent if left |
| LID | First hours to weeks in sunlight | No, but it should stabilise at a few per cent |
| LETID | Months to years, under sustained heat | Partial recovery over years; not something to wait for |
Why degradation hides in plain sight
Degradation is slow, and slow losses blend into everything else. A plant-level performance ratio averages good strings with bad ones. A dusty month, a hot summer or a drifting irradiance sensor can each produce the same downward trend. Annual reviews compare this year with last year, and a decline of a per cent or two looks like ageing rather than a defect. The signals that separate degradation from everything else are at string level: losses that persist after rain and cleaning, that are not explained by temperature, that grow over time, and that affect some strings more than their sister strings under identical sun. Without string-level data and a way to rule out the other explanations, degradation stays a suspicion rather than a finding.
What a warranty claim actually needs
Module warranties commonly allow a larger loss in the first year and a small, steady decline after that. To claim, you generally need to show measured output below the warranted curve, beyond the stated measurement tolerance — usually with module testing such as flash or I-V curve measurements. Before you pay for testing, you need to know which strings to test, roughly how far below the curve they are, and that the shortfall is not soiling, temperature or a sensor fault. And the manufacturer has to still be trading. A claim delayed by several years can lose to all three: a weaker case, a permanent loss, and nobody left to pay it.
One plant, two years in: an illustrative case
Numbers illustrative. A two-year-old plant in a hot, humid climate is running about 2% below budget, within what the owner considers normal. String-level trends tell a different story. Most strings are declining in line with the warranty. One inverter block is losing output about three times faster than its sister strings, the gap persists after rain, it is not explained by cell temperature, and the irradiance sensor checks out. The pattern is consistent with PID. The owner fits recovery equipment to that block and books targeted module testing, with two years of string-level history already in hand for the warranty discussion. Found in year five, the same block might have been both permanently damaged and much harder to claim for.
How Ellume Vector catches it
- •PID and early-life degradation detected. Vector detects potential-induced degradation, and checks whether new panels are degrading faster than the manufacturer's warranty allows through LID and LETID — while you can still claim.
- •String degradation trends against sister strings. Vector shows which parts of your array are ageing faster than the rest, and compares each against sister equipment on the same site: identical weather, identical age, identical design. It is the fairest benchmark available.
- •The other explanations ruled out first. Soiling is measured as it builds up, rain events and their cleaning effect are recognised, and losses from high cell temperature are quantified. Vector also detects irradiance sensor drift, which can make a healthy plant look degraded. What remains is far more likely to be the panels.
- •Risk assessment and traceable evidence. Vector provides a degradation risk assessment for your panels with mitigation guidance, and every figure can be traced back to the raw readings behind it — the history a warranty discussion depends on.
Three checks to run this year
Find your module warranty and note the warranted curve, the measurement tolerance and the evidence the manufacturer requires. Compare string-level output across sister strings on the same inverter, after a clean or a heavy rain, and look for strings that are falling behind year on year. And check your irradiance sensor before you blame the panels: a drifting sensor makes a whole plant look degraded.
Frequently asked questions
- What is the difference between LID and LETID?
- LID happens in the first hours to weeks of sunlight and usually stabilises at a few per cent. LETID needs heat as well as light, develops over months to years, and can cause larger losses in susceptible modules before partially recovering over a long period.
- Can PID be reversed?
- Often partly, if it is caught early — for example with equipment that applies a reverse voltage to the array overnight. The longer PID runs, the more likely the damage is permanent.
- How do I know a loss is degradation and not soiling?
- Soiling losses recover after rain or cleaning and tend to affect a site broadly. Degradation persists after cleaning, grows over time and often affects some strings more than their neighbours under identical conditions.
- Does Ellume Vector replace module testing for a warranty claim?
- No. Manufacturers usually require module testing. Vector tells you which strings to test and when, rules out the other explanations, and gives you the measured history that supports the claim.
