Cell cracks are breaks in the brittle silicon of the solar cells and are usually invisible to the naked eye. They start in manufacturing, transport and installation, for example when a module is dropped or stepped on, and grow under wind, snow, hail and temperature changes. Most cracks cost little power, but cracks that isolate part of a cell reduce current, can switch on the bypass diode and can create hotspots and burn marks. In the field, look for snail tracks or bleached lines along cracks, which make small cracks visible. Confirm with electroluminescence imaging, where cracks show as dark lines; UV fluorescence also works, and infrared thermography shows cracked cells that are heating. Use IV curves to judge the electrical impact. Handle modules carefully during transport and installation, monitor cracked modules for progression, and replace those with severe inactive areas or hotspot risk.
caused_byc1 drafted · lead Manufacturing defect
Wafer slicing, cell production, stringing (soldering residual stress) and embedding can start cracks; repetitive crack patterns point to production faults.
sec=h4 h=9ddd690f2930sec=6.2.2 h=17668c52ed6fc2 drafted · lead Transport damage
Poor packaging, vibration and reloading during transport induce cracks.
sec=h4 h=9ddd690f2930sec=6.2.2 h=17668c52ed6fc3 drafted · lead Installation error
Dropping modules or stepping on them during installation.
sec=h4 h=9ddd690f2930sec=6.2.2 h=17668c52ed6fc4 drafted · lead Mechanical stress
Hail, wind and snow loads; dendritic crack patterns indicate heavy mechanical load after lamination.
sec=h4 h=9ddd690f2930sec=6.2.2 h=17668c52ed6fc5 drafted · lead Thermal cycling
Temperature variations and uneven thermal expansion grow short cracks into longer, wider ones.
sec=h4 h=9ddd690f2930sec=6.2.2 h=17668c52ed6faffectsc6 verified · fact Solar cell
sec=6.2.2 h=17668c52ed6ftbl=7.r1 h=790734be7823verified by: joshuahamsaverified on: 2026-10-16
presents_asc7 verified · fact Dark lines in EL image
An abrupt EL intensity change at the dark grey line confirms a crack rather than a crystal defect.
sec=5.4 h=1fb2586fa36bverified by: joshuahamsaverified on: 2026-10-16
c8 drafted · lead Snail tracks
Discoloured silver gridlines along cracks, typically 3 months to 1 year after installation.
sec=h7 h=8b69da4daaactbl=7.r1 h=790734be7823tbl=10 h=f95eb9235f8dc9 drafted · lead Encapsulant photobleaching
Bleached lines in yellowed encapsulant along cracks and cell borders.
sec=h7 h=8b69da4daaactbl=7.r1 h=790734be7823tbl=10 h=f95eb9235f8dreview note: support check: partial - Photobleaching along cracks and cell borders is stated, but 'bleached lines' appearance is only loosely implied.
c10 drafted · lead Hotspot in IR image
When the cracked cell is driven into reverse bias.
sec=h8 h=ec36b7bfbdb3sec=6.2.2 h=17668c52ed6fc11 drafted · lead Burn marks
tbl=7.r5 h=0a21fb8dc566sec=6.2.2 h=17668c52ed6fc12 drafted · lead Reduced short-circuit current
Isolated cell area no longer generates current; current reduction is the main loss contributor.
sec=h6 h=020d995c9f1fc13 verified · fact Increased series resistance
sec=h6 h=020d995c9f1fsec=2.3.3 h=b6c82133fe8cverified by: joshuahamsaverified on: 2026-10-16
c14 drafted · lead Reduced fill factor
sec=2.3.3 h=b6c82133fe8cc15 verified · fact Bypassed substring
Large inactive areas activate the bypass diode of the affected substring.
sec=h6 h=020d995c9f1fsec=6.2.2 h=17668c52ed6fverified by: joshuahamsaverified on: 2026-10-16
detected_byc16 drafted · lead Electroluminescence imaging
sec=front h=a37d82d92ad0sec=h8 h=ec36b7bfbdb3sec=5.4 h=1fb2586fa36btbl=7.r1 h=790734be7823c17 drafted · lead UV fluorescence imaging
sec=h8 h=ec36b7bfbdb3tbl=7.r1 h=790734be7823c18 drafted · lead Photoluminescence imaging
sec=h8 h=ec36b7bfbdb3c19 verified · fact Infrared thermography
Shows cracked cells only when they heat up under reverse bias.
sec=front h=a37d82d92ad0sec=h8 h=ec36b7bfbdb3verified by: joshuahamsaverified on: 2026-10-16
c20 drafted · lead IV curve tracing
sec=front h=a37d82d92ad0sec=h8 h=ec36b7bfbdb3c21 drafted · lead Visual inspection
Usually only works once snail tracks, photobleaching or delamination appear, or for large cracks.
sec=h8 h=ec36b7bfbdb3tbl=7.r1 h=790734be7823mitigated_byc22 verified · fact Proper installation
Avoid mechanical stress during transport and installation.
sec=front h=a37d82d92ad0verified by: joshuahamsaverified on: 2026-10-16
c23 drafted · lead Close monitoring
sec=front h=a37d82d92ad0c24 verified · fact Module replacement
For confirmed severe inactive area or hotspot risk.
sec=front h=a37d82d92ad0verified by: joshuahamsaverified on: 2026-10-16
c25 drafted · lead Multi-busbar cell design
sec=2.3.3 h=b6c82133fe8ctbl=7.r5 h=0a21fb8dc566leads_toc26 drafted · lead Module Hotspot from Cell Damage
sec=h6 h=020d995c9f1fsec=h7 h=8b69da4daaacsec=6.2.2 h=17668c52ed6fc27 drafted · lead Delamination
sec=h7 h=8b69da4daaactbl=7.r1 h=790734be7823typical_power_lossc28 verified · fact ≤ 2.5 PERCENT
Conditions: Every cell in a 60-cell module cracked but with no electrically separated cell area, at STC
sec=2.3.3 h=b6c82133fe8ctbl=7.r5 h=0a21fb8dc566verified by: joshuahamsaverified on: 2026-10-19review note: support check: partial - Passages say 'typically below 2.5%' (an upper bound), not a typical value of 2.5%.; possible conflict with c34 (same source; check the conditions); reviewer 2026-10-17: the compiler-flagged conflicts are different scenarios (STC full-crack test, per-half-cell value, inactive-area simulations, one field survey), not disagreement
c29 verified · fact ≤ 2.5 PERCENT
Conditions: Loss from increased recombination current caused by cracks
sec=h6 h=020d995c9f1fverified by: joshuahamsaverified on: 2026-10-16review note: support check: partial - Recombination effect 'remains under 2.5%' is an upper bound, not a typical 2.5% value.; possible conflict with c32 (same source; check the conditions); possible conflict with c33 (same source; check the conditions); reviewer 2026-10-16: the compiler-flagged conflicts are different scenarios, not disagreement
c30 verified · fact nominal 0.2 PERCENT
Conditions: Per dendritic-like cracked half-cell in multi-wire PERC modules
tbl=7.r5 h=0a21fb8dc566verified by: joshuahamsaverified on: 2026-10-19review note: reviewer 2026-10-17: the compiler-flagged conflicts are different scenarios (STC full-crack test, per-half-cell value, inactive-area simulations, one field survey), not disagreement
c31 verified · fact nominal 0.2 PERCENT
Conditions: Simulation, one cell with 5% inactive area
sec=h6 h=020d995c9f1fverified by: joshuahamsaverified on: 2026-10-19review note: possible conflict with c32 (same source; check the conditions); possible conflict with c33 (same source; check the conditions); reviewer 2026-10-17: the compiler-flagged conflicts are different scenarios (STC full-crack test, per-half-cell value, inactive-area simulations, one field survey), not disagreement
c32 verified · fact nominal 5 PERCENT
Conditions: Simulation, one cell with 15% inactive area
sec=h6 h=020d995c9f1fverified by: joshuahamsaverified on: 2026-10-16review note: possible conflict with c29 (same source; check the conditions); possible conflict with c31 (same source; check the conditions); possible conflict with c33 (same source; check the conditions); reviewer 2026-10-16: the compiler-flagged conflicts are different scenarios, not disagreement
c33 verified · fact nominal 35 PERCENT
Conditions: Simulation, one cell with 45% inactive area, bypass diode of substring activated
sec=h6 h=020d995c9f1fverified by: joshuahamsaverified on: 2026-10-19review note: possible conflict with c29 (same source; check the conditions); possible conflict with c31 (same source; check the conditions); possible conflict with c32 (same source; check the conditions); reviewer 2026-10-17: the compiler-flagged conflicts are different scenarios (STC full-crack test, per-half-cell value, inactive-area simulations, one field survey), not disagreement
c34 verified · fact nominal 12 PERCENT
Conditions: Simulated mean string loss after complete isolation of all cracked cell parts, medium crack occurrence
sec=2.3.3 h=b6c82133fe8cverified by: joshuahamsaverified on: 2026-10-16review note: possible conflict with c28 (same source; check the conditions); reviewer 2026-10-16: the compiler-flagged conflicts are different scenarios, not disagreement
c35 verified · fact ≥ 10 PERCENT
Conditions: Nearly half of 159 modules in one solar park after 6 years of operation
sec=6.2.2 h=17668c52ed6fverified by: joshuahamsaverified on: 2026-10-19review note: reviewer 2026-10-17: the compiler-flagged conflicts are different scenarios (STC full-crack test, per-half-cell value, inactive-area simulations, one field survey), not disagreement
degradation_ratec36 drafted · lead ≤ 3 PERCENT-PER-YR
Conditions: All climate zones except cold and snowy
sec=h6 h=020d995c9f1freview note: support check: partial - Source says below 3% per year, an upper bound, not a rate of 3%.; possible conflict with c37 (same source; check the conditions)
c37 verified · fact nominal 8 PERCENT-PER-YR
Conditions: Affected part in cold and snowy climate zones
sec=h6 h=020d995c9f1fverified by: joshuahamsaverified on: 2026-10-19review note: possible conflict with c36 (same source; check the conditions); reviewer 2026-10-17: 8 %/yr is the rate of the affected part, up to 7 %/yr is a mean over full-cell modules; different statistics from the same T13 work, not disagreement
c38 verified · fact ≤ 7 PERCENT-PER-YR
Conditions: Mean rate for full-cell modules in cold and snow climate zones
tbl=7.r5 h=0a21fb8dc566verified by: joshuahamsaverified on: 2026-10-19review note: support check: partial - Source says 'up to 7%/y', a high bound on mean rates, not a typical 7%.; reviewer 2026-10-17: 8 %/yr is the rate of the affected part, up to 7 %/yr is a mean over full-cell modules; different statistics from the same T13 work, not disagreement
time_to_onsetConditions: Cell cracks dominate early failures in the first two years after installation
sec=h5 h=451648d48890review note: support check: partial - Cell cracks dominate early failures in years one and two, but the range is stated as years rather than an onset time.
related_to c8related_to c2related_to c4leads_to c25caused_by c4leads_to c13caused_by c1leads_to c21leads_to c15