Snow accumulation losses happen when snow stays on the module surface after a snow event, so covered modules cannot turn the available irradiance into current. Production usually drops sharply after snowfall and comes back gradually as the snow melts or slides off, even while the plant is available and irradiance sensors show good resource. Patchy clearing causes mismatch between clear and covered cells, so current can be uneven across strings and substrings, and the loss may not match the visible covered area. Low sun angles, cold module temperatures, frame geometry, module orientation and tracker position all affect how long snow stays on the array. To confirm, line up timestamped snowfall, snow depth, temperature, wind and irradiance data with SCADA production residuals and with camera, drone or field observations of snow cover. Before blaming snow, filter out cold-weather inverter behavior, clipping, curtailment, frozen tracker stow, sensor icing and albedo effects. Compare measured recovery curves with the snow-loss assumptions in the energy model, and revise the snow-loss factor if the site clears more slowly than modeled. Track winter residuals separately from general soiling, and consider tracker stow, tilt, maintenance or design changes where they pay off.
<!-- review notes
caused_byc1 drafted · lead Heavy snowfall
Loss arises when snow stays on the module surface after a snow event and blocks irradiance.
sec=h4 h=0f32a669be34sec=front h=ef48f3c9a102review note: support check: partial - Passages say snow remaining on modules after a snow event blocks irradiance, but don't name heavy snowfall as a stressor/cause explicitly.
c2 drafted · lead Low sun angle
Low sun angles can extend how long snow stays on the array.
sec=h5 h=7b6f8fb933e9review note: support check: partial - Low sun angle is said to affect how long snow remains, but not stated as a cause of the loss itself.
c3 drafted · lead Cold module temperature
Cold modules can keep snow on the array longer; frame geometry, orientation and tracker position also matter.
sec=h5 h=7b6f8fb933e9affectsc4 drafted · lead Module front glass
Snow covering the module surface stops irradiance from reaching the cells.
sec=h4 h=0f32a669be34sec=front h=ef48f3c9a102presents_asc5 verified · fact Abrupt production drop
Output can fall abruptly right after snowfall.
sec=h4 h=0f32a669be34verified by: joshuahamsaverified on: 2026-10-16
c6 verified · fact Gradual production recovery
Output returns gradually as snow melts or slides off.
sec=h4 h=0f32a669be34verified by: joshuahamsaverified on: 2026-10-16
c7 drafted · lead Low output despite irradiance
The plant is available and irradiance sensors show usable resource, but covered modules produce no current.
sec=h4 h=0f32a669be34c8 drafted · lead Uneven string current
Partial clearing leaves clear and covered cells at different current levels, so loss is not proportional to visible snow area.
sec=h4 h=0f32a669be34sec=h5 h=7b6f8fb933e9c9 drafted · lead Bypassed substring
Partial clearing can produce uneven current across substrings.
sec=h4 h=0f32a669be34detected_byc10 drafted · lead Weather-normalized production comparison
Line up production residuals with snowfall, snow depth, precipitation type, ambient temperature, wind, plane-of-array irradiance and module temperature.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896sec=h6 h=0935b66311acc11 drafted · lead Site camera observation
Camera images confirm whether modules are covered.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896sec=h6 h=0935b66311acc12 drafted · lead Drone inspection
Drone observations of module cover can be matched to SCADA residuals.
sec=h6 h=0935b66311acc13 verified · fact Visual inspection
Field observations of snow cover on the modules.
sec=h6 h=0935b66311acverified by: joshuahamsaverified on: 2026-10-16
c14 verified · fact Loss model comparison
Compare measured recovery curves with regional snow-loss model output and the energy model assumptions.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896sec=h6 h=0935b66311acverified by: joshuahamsaverified on: 2026-10-16
mitigated_byc15 verified · fact Loss factor separation
Track winter residuals separately from generic soiling so snow losses are not hidden in monthly weather variance.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896sec=h5 h=7b6f8fb933e9verified by: joshuahamsaverified on: 2026-10-16review note: reviewer 2026-10-19: target mit:seasonal-residual-tracking merged into mit:loss-factor-separation
c16 drafted · lead Module tilt optimization
Consider tilt changes where they are economic.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896c17 verified · fact Tracker stow strategy
Evaluate tracker stow changes where they are economic.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896verified by: joshuahamsaverified on: 2026-10-16
distinguished_fromc18 drafted · lead Grid Curtailment Yield Losses — told apart by Weather-normalized production comparison
Curtailment can overlap with snow events and should be filtered out using snowfall and cover data before recoverable snow loss is estimated.
sec=h7 h=a5ed6e5ad384review note: support check: partial - Curtailment overlapping with snow events and needing filtering is stated, but not the use of weather-normalized comparison or snowfall/cover data specifically.
c19 verified · fact Soiling Losses — told apart by Weather-normalized production comparison
Snow losses are seasonal and tied to snowfall and recovery intervals; winter residuals should be tracked apart from generic soiling.
sec=front h=ef48f3c9a102sec=h2 h=cf13e86d6896verified by: joshuahamsaverified on: 2026-10-16review note: support check: partial - Tracking winter residuals separately from generic soiling is stated, but not weather-normalized comparison as a distinguishing method nor explicitly seasonal tied to recovery intervals in these passages.
related_to c3related_to c5