Skip to content

Use Cleaning Analytics

Use Insights → Cleaning Analytics to narrow a portfolio to plants that need review, then inspect each plant before planning cleaning work.

Cleaning Analytics is available when your account includes the feature and you have access to eligible plants. If the menu item is not visible, ask your SolarSENS administrator to confirm your access.

  1. Open Insights → Cleaning Analytics.
  2. Select Overview.
  3. Choose a Portfolio.
  4. Review the number of plants with suggested cleaning in the Next 1 month and Next 3 months.
  5. Compare the portfolio’s Soiling Loss for this week and the current month.
  6. Review the clean, dirty, and very dirty summaries for plants and inverters.
  7. Use Site Overview to rank individual plants.

The seven-week specific-soiling-loss chart helps you see whether the portfolio trend is isolated to one week or persists across several weeks.

These summaries mix time grains — weekly model indicators, the seven-week trend, current-week loss, and month-to-date loss cover different periods, so compare like with like.

ColumnHow to use it
StatusStart with V. Dirty and Dirty plants, then validate the supporting values.
Inverters status (V | D | C)See whether the signal is concentrated in part of the plant or spread across its inverters.
Soiling RateCompare the 1-week and 4-week averages. A higher rate indicates dirtier equipment.
PerformanceCompare the 1-week and 4-week adjusted values and the change from the previous week. Higher is better.
Last CleanedCheck whether the current pattern started before or after the latest recorded cleaning.
Suggested CleaningTreat the modeled timing as a review signal, not a confirmed appointment.
Eval. DateConfirm when the displayed model result was evaluated.
Soil. LossCompare the plant’s estimated energy loss with other candidates.
S. Soil. LossCompare loss normalized by plant capacity.

Select a column heading to sort the table. Use the download action to export the current table as CSV when you need an offline shortlist. Double-click a plant row to open its plant chart.

  1. Select Plants, or double-click a row in Site Overview.
  2. Choose the plant you want to review.
  3. Set Interval to 3 Months, 6 Months, 1 Year, or All.
  4. Keep Relative To → Timeline for the full trend.
  5. Switch to Relative To → Last Clean to evaluate performance from the latest recorded cleaning.
  6. Compare the soiling trend with precipitation and humidity.
  7. Compare normalized performance with irradiance and module temperature.
  8. Note where recorded cleaning events fall on the timeline.

Start with three or six months for recent behavior. Use a longer interval when the trend is noisy or when you need to compare multiple cleaning cycles.

Treat precipitation, humidity, irradiance, and module temperature as weather context. Adjusted performance, soiling rate, suggested cleaning, and estimated loss are modeled results derived from the available plant data. Weather can support or challenge the soiling interpretation, but it does not prove soiling by itself.

Turn the result into an operational decision

Section titled “Turn the result into an operational decision”

Before planning cleaning, confirm that:

  • the one-week signal is supported by the four-week trend
  • declining performance is not better explained by weather or low irradiance
  • estimated loss is meaningful compared with other plants
  • the latest cleaning event is present and its date is plausible
  • the plant can be accessed and cleaned safely under your operating procedures

Cleaning Analytics does not create cleaning work or reserve a crew. Follow your organization’s approved planning and safety process after you decide to proceed.

After cleaning is recorded and new operating data is available, return to the plant view. Compare the post-cleaning performance and soiling trend with the earlier period. If the indicators do not improve, investigate other performance causes instead of assuming the cleaning was ineffective.

A suggested date can change when another weekly observation or corrected input becomes available. Treat it as the latest analytical signal, not a fixed commitment.

If a plant or value is missing:

  1. Confirm that the plant belongs to the selected portfolio and is available to your account.
  2. Try a longer interval.
  3. Check the evaluation date and whether other plant-monitoring data is available for the same period.
  4. Confirm that the latest cleaning event has been recorded in the system your organization uses.
  5. Contact your SolarSENS administrator if the result remains unavailable or appears stale.

A blank suggestion means SolarSENS has no usable result for that field — not that the plant is clean. Check the evaluation date and data coverage before reading a blank or zero as “no soiling”; see Data freshness and corrections.