How to Assess Solar Module LID and LeTID: Two Mechanisms on Two Clocks (Boron–Oxygen LID Typically Within Hours, Possibly Several Percent, Sometimes a Slow Recovery; Hydrogen-Linked LeTID Over the First Years, Field Losses Sometimes Over 10%, Then Weather-Dependent Regeneration), What the Test Can and Cannot Tell You (IEC TS 63342:2022 Current-Injection Detection Reveals Sensitivity but Gives No Exact Measure of Field Degradation; IEC 61215 Results Not a Lifetime Prediction), Why the Industry Says the Problem Is Mostly Solved (Gallium Doping, Thinner Wafers, Firing Profiles; TOPCon Modules Through a LeTID Test the Report Gives as 322 h and as 2 × 162 h, With Minor Loss) and Where the Residual Risk Sits (a Commercial Lab Reporting a Lower Percentage of LID+LETID Top Performer BOMs in 2026; First-Year Warranty Steps of 1.0%, 2.0% and 2.5% Stated Without Naming the Mechanism)

Short answer. LID and LeTID are two different losses that happen to share three letters. Boron–oxygen LID is fast and modest: the IEA PVPS Task 13 report says its maximum typically occurs within the first few hours of light exposure, may amount to several percent of initial power, and is sometimes followed by a slow recovery that is hard to see in the field. LeTID is slow and potentially large: hydrogen-linked defects activated by combined light and heat, with degradation then regeneration over the first years of operation and field yield losses that have 'sometimes reached values over 10%'. The detection test, IEC TS 63342:2022, injects current at elevated temperature and, in its own words, 'can reveal sample sensitivity … but does not provide an exact measure of field observable degradation'; it does not address B–O or Fe–B LID at all. The same report says both effects 'lost importance' in current products because gallium replaced boron, wafers got thinner and firing profiles were adapted, and that a large set of commercial TOPCon modules showed minor loss after a LeTID test the report gives as 322 hours (2 × 162 h). Yet Kiwa PVEL's 2026 Scorecard reports that the percentage of Top Performer bills of materials fell for LID+LETID while other tests improved, and warranties still step first-year degradation at 1.0%, 2.0% or 2.5% by product type without naming a mechanism. The assessment therefore is: which cell technology, which test result, and what the first-year warranty figure actually covers.

1. Two mechanisms

CNWattHub reading: the two clocks matter for how a buyer would ever see the loss. A LID of a few percent that typically completes within hours of first exposure has largely happened before any field measurement taken after that; a LeTID loss that builds over years and then partly regenerates will look like a high early degradation rate followed by an apparent improvement, and a 25-year linear warranty line is a poor instrument for catching it. NREL's review of long-term rates (median 0.5%/year) states that its rates exclude short-term LID and concedes it could not completely exclude its influence, which is a reminder that published degradation statistics are not LID statistics.

2. What the tests detect — and do not

• IEC TS 63342:2022 uses current injection at elevated temperature; only that approach is covered; B–O and Fe–B LID are out of scope; the result reveals sensitivity but, in the specification's words, gives no 'exact measure of field observable degradation'.
• IEC 61215-2:2021 — the abstract lists its new tests (dynamic mechanical load MQT 20, PID MQT 21, bifacial and flexible methods, MQT 22 bending, NMOT removed); a LID or LeTID test is not among the significant changes listed there.
• IEC 61215-1:2021 says test results 'are not construed as a quantitative prediction of module lifetime'.
• In the Task 13 report the authors refer to 'IEC 61215-2:2021 gate 1' as LID preconditioning and to 'IEC 61215 MQT 23' as a LeTID test, and recommend running light soaking (MQT 01) and the LeTID test before UV-degradation testing, with module temperature kept below 60 °C in UV tests so that UVID and LeTID/LID effects can be separated.

CNWattHub reading: the IEC 61215 abstracts read here do not mention LeTID among the edition's listed changes and say nothing about what a certificate records, so this guide cannot say from them what an IEC 61215 certificate shows about LeTID; the Task 13 report's references to gate 1 and MQT 23 are the only clause-level pointers this guide has, and the test report is the document to ask for. A TS 63342 result is a sensitivity screen: a module that degrades in it is sensitive; a module that passes has not been shown to be free of field LeTID, because the specification itself says it gives no exact measure of field-observable degradation. The test-sequence order also matters: a UV test run on an unstabilised module can mix three mechanisms into one number.

3. Why the industry says it is mostly solved

The Task 13 report's position is unusually definite: LID/LeTID 'has been well understood and solved' by the gallium switch, adjusted cell processes with thin wafers and fewer wafer impurities; both mechanisms 'lost importance within current (2023) PV module products'; LID relevance is 'low for nearly all current PV modules' though still possible from boron impurities even in n-type; TOPCon modules are 'substantially less susceptible' than early PERC, and a large set of commercial TOPCon modules showed only minor loss after TS 63342 testing the report gives as 322 h and as 2 × 162 h at ISC − IMPP (the two figures differ by 2 h) following gate-1 LID preconditioning, so the authors expect no detectable LeTID at module or system level in 'current TOPCon modules based on Ga doped p-type wafers' — a phrase that sits oddly beside the same section's statement about n-type TOPCon cells, and is quoted as written. The conclusion is that LeTID's relevance 'practically vanished' in TOPCon and SHJ types and may fall further as wafers thin.

CNWattHub reading: these statements are technology-specific. They hold for Ga-doped p-type and n-type products made after the transition; they say nothing about an older p-type PERC fleet or about a low-cost product whose wafer and firing process are unknown. The experiment quoted is a single reported test series on commercial TOPCon modules, shown in a figure, not a field census, and the report's own description of its duration and the wafer type is not fully consistent.

4. Where the residual risk shows

Kiwa PVEL's 2026 Scorecard (12th edition) includes LID+LETID among its eight PQP tests and reports that, while PID and UVID losses fell and MSS/HSS losses stayed minimal, 'the percentage of Top Performer BOMs decreased for LID+LETID'; only five of 246 Top Performer model types were Top Performers across all reliability tests, and 87% of manufacturers had at least one PQP failure.

CNWattHub reading: the page gives a direction, not a number — a lower percentage of tested bills of materials cleared the LID+LETID threshold this year (the page does not give the denominator, so whether the count fell is unknown) — and it is a commercial lab's own programme, with its own criteria. Set beside the Task 13 report it suggests that 'solved in principle' and 'solved in every BOM' are different claims, and that the gap is a bill-of-materials question.

5. What the first-year warranty figure covers

Trina's table is for its back-sheet products (glass-glass products have their own schedule, not quoted), measured at STC with measurement-system uncertainty to be included in the actual power figure.

CNWattHub reading: the first-year step is a guaranteed limit on measured power decline in year one; neither warranty text names LID or LeTID, so the step is not a statement about mechanism, though it is smaller for n-type than for p-type, which is consistent with the mechanism story. Because a LeTID loss can peak after year one and then regenerate, a module could sit inside a 2.0%-then-0.55% line in year one and outside it in year three, and back inside later; whether a claim can be made at the trough depends on how the warranty measures and when. How measurement uncertainty is 'included' in the measured figure is not spelled out in the text read here, and it affects the margin.

6. An assessment sequence

1. Technology. Wafer doping (B or Ga), cell type (PERC, TOPCon, SHJ), wafer thickness — the three levers the report names.
2. Test evidence. A TS 63342 result with hours, current and temperature stated; the stabilisation steps actually run in the IEC 61215 report; the order of UV and LeTID tests.
3. Reading the result. Sensitivity, with no exact measure of field loss; qualification, not lifetime.
4. Independent lab data. Whether the specific BOM, not the brand, cleared a LID+LETID threshold.
5. Warranty. First-year step, measurement at STC with uncertainty, and whether a mid-term trough is claimable.
6. Monitoring. Separate early-year PR changes from long-term rate statistics, which are stated to exclude short-term LID (though not perfectly).

Scope limits

This guide reads the IEC TS 63342:2022 and IEC 61215-1/-2:2021 webstore abstracts (no clause text), the IEA PVPS Task 13 2025 report, NREL's 2012 degradation-rate review, the Trina and JinkoSolar warranty texts and the Kiwa PVEL 2026 Scorecard home page. It quotes no Chinese standard, datasheet LID tolerance, lab test report, field LID/LeTID measurement from a Chinese plant or numeric Scorecard result; the LeTID experiment (reported as 322 h and as 2 × 162 h, which do not agree) is a single reported test series on commercial TOPCon modules. Not engineering, procurement or investment advice.