Mitigation
Initial degradation in yield model
- ID
mit:yield-model-initial-loss
- IRI
https://solarlattice.org/kb/mit/yield-model-initial-loss
- Status
- active
This action means building the expected initial light-induced power loss into the energy yield model, so that early stabilization losses are not mistaken for faults and do not inflate production forecasts. It addresses Light Induced Degradation (LID), which appears as a drop in STC power, short-circuit current and open-circuit voltage early in system life, with no visible defect. If initial degradation is under-estimated, energy yield predictions and investor expectations are directly affected. Some yield assessors add a separate initial LID term to the long-term yield prediction, sized by module technology and site location. The size of the allowance must match the technology: crystalline LID is a few percent in the first days, amorphous silicon can lose much more over the first months, and some CIS modules actually gain power. LeTID is harder to model because it develops over years, depends on module temperature and site, and may not recover within the module lifetime. There is no consensus on which degradation values to use, so pair this action with low-LID module selection, and check unexpected early losses with IV and flash testing.
<!-- review notes
- sentence 1: support check: partial - Mitigation says account for stabilization in yield assumptions, but the 'not mistaken for faults/inflate forecasts' rationale is only loosely implied.
- sentence 6: support check: partial - Passage supports years-long development, temperature/site dependence, possible non-recovery and difficulty predicting trend, but doesn't say LeTID is harder to model than LID in yield models explicitly.
-->
Claims
typical_power_loss
c1 verified · fact
0.3–1 PERCENT
Conditions: Initial LID term some yield assessors add in long-term yield predictions; depends on module technology and location
verified by: joshuahamsaverified on: 2026-10-20review note: reviewer: the compiler-flagged conflict is a different scenario or technology, not disagreement
c2 verified · fact
1–2 PERCENT
Conditions: First-year drop commonly assumed by PVPS Task 13 survey participants in long-term yield predictions
verified by: joshuahamsaverified on: 2026-10-20review note: reviewer: the compiler-flagged conflict is a different scenario or technology, not disagreement
c3 verified · fact
1–3 PERCENT
Conditions: Boron-oxygen LID in p-type crystalline silicon modules, within the first days of sun exposure
verified by: joshuahamsaverified on: 2026-10-20review note: reviewer: the compiler-flagged conflict is a different scenario or technology, not disagreement
c4 verified · fact
≤ 10 PERCENT
Conditions: LeTID in susceptible modules (first seen with PERC), adding to the LID loss over years
verified by: joshuahamsaverified on: 2026-10-20
c5 verified · fact
10–30 PERCENT
Conditions: Amorphous silicon modules within the first months of outdoor exposure (Staebler-Wronski effect)
verified by: joshuahamsaverified on: 2026-10-20review note: reviewer: the compiler-flagged conflict is a different scenario or technology, not disagreement
Referenced by
Data