Optical Soiling Measurement: From Cleanliness to PV Loss

Optical soiling measurement describes how contamination changes an instrument's optical response. It can support photovoltaic cleaning decisions, but a local cleanliness value is not, by itself, a measured percentage loss of plant energy. The NiuBoL NBL-W-PSS provides two cleanliness channels for observing surface contamination. Using those readings well requires a clear channel definition, representative exposure and a separate method for relating the optical trend to the electrical behavior of the modules.

Separate the optical reading from the electrical soiling ratio

Electrical soiling assessment compares an appropriate dirty PV reference with a clean reference under comparable operating conditions. An optical instrument observes light interacting with contamination on its sensing surface. The methods can be related, but they do not measure exactly the same quantity. Keep the instrument's cleanliness label and unit intact until the project's conversion method has been established.

The PSS communication example expresses a cleanliness value of 965 as 96.5%. The difference between that reading and a nominal 100% clean reference is 3.5 percentage points. That is a difference on the indicated optical scale. Calling it 3.5% lost plant energy adds assumptions about module response, exposure, time and operating conditions that the number alone does not establish.

This distinction matters when several instruments use the abbreviation SR. A project may use it for a cleanliness-related sensor tag, a transmittance estimate or an electrical soiling ratio. Put the full quantity name in the data dictionary. Store the original PSS channels alongside any derived loss estimate so an engineer can inspect what was actually observed rather than only the final dashboard calculation.

Consider the spectrum and distribution of the contamination

The PSS uses blue-light optical sensing. Different contaminants can affect different wavelengths differently, while PV materials and incident sunlight also have spectral characteristics. Research on wavelength selection for optical soiling detection explains why a single optical measurement cannot automatically account for every combination of solar spectrum and PV spectral response. This is a reason to validate a site-specific relationship, not a reason to discard optical monitoring.

A small sensing area also does not reproduce every deposit on a full module. A uniform dust film, a lower-edge band and an isolated opaque patch can produce different electrical effects even when a nearby optical point looks similar. Photographs and inspection notes help identify these patterns. Do not assume that a module-frame instrument directly resolves cell mismatch, bypass-diode behavior or the fraction of the array affected by an isolated deposit.

Define a comparison that can test the proposed conversion

If the project needs an electrical loss estimate, agree a comparison method before enabling that tag. Select a representative module type and an appropriate clean/dirty reference arrangement. Record the optical values, irradiance on the relevant plane, module temperature and electrical measurements over matching intervals. The comparison should include the kinds of deposits and seasonal conditions that matter at the site.

Do not validate a loss conversion with raw inverter power immediately before and after washing. Irradiance may change, modules may cool, clipping may limit the inverter, and cleaning may not cover the whole electrical block. Use the project's accepted normalization and exclusion method. Preserve observations that do not meet those conditions as flagged records rather than silently forcing them to fit a curve.

For example, assume an optical trend suggests a 4% loss while a suitable electrical comparison indicates 2% for the same period. The difference calls for investigation of exposure, deposit pattern, reference condition and conversion assumptions. It does not justify multiplying all later optical readings by one-half after a single test. Repeat the comparison across the useful operating range and retain the limits of the resulting relationship.

Specify the PSS without assigning it extra functions

The NBL-W-PSS soiling sensor is the local optical measurement device in this arrangement. A data logger or gateway reads its channels; an application can then apply an agreed analysis. Cleaning forecasts, work orders and robot control are separate functions and need their own implementation and acceptance checks.

Request the measurement specification and applicable conditions for the ordered unit when defining acceptance tolerances. A percentage accuracy statement needs an identified measurand and range; it is not automatically the uncertainty of a derived energy-loss model. Do not assign an ingress rating, temperature envelope or cable material that has not been confirmed for that configuration.

Keep surface interventions visible in the optical record

Mount the instrument with the exposure intended for the surrounding modules, using a fixing arrangement accepted for the module frame. Clean both optical probes when cleaning the associated modules. Do not permanently maintain one PSS probe as a separate clean electrical reference unless that operating arrangement is expressly specified for the installed instrument. Two reported cleanliness channels are not proof of a conventional clean/dirty PV reference pair.

Record rain, washing, dew-related observations and any baseline adjustment as different events. Water on an optical surface can affect a reading without establishing that dry contamination has been removed. Compare the stable condition after the event as well as the immediate response. If the instrument is cleaned while adjacent modules remain dirty, mark that interval so it is not interpreted as a sudden recovery across the array.

Keep the original baseline before any recalibration. Changing the baseline changes the meaning of later readings; it does not document how much energy was recovered. Where commissioning calibration is required, use the specified procedure and conditions for the supplied instrument. Routine maintenance and a deliberate baseline change should remain distinguishable in the station history.

Use the optical trend at the scale it can support

At a site with consistent deposits and a validated relationship, optical soiling measurement can help identify periods for closer inspection or cleaning assessment. At a site with patchy contamination or different module orientations, one local value may not describe the relevant electrical block. Define which modules the sensor represents before aggregating readings into a plant-level indicator. A PV cleaning schedule then needs its own operational and economic assessment.

For a NiuBoL application enquiry, provide module technology, glass or coating information where available, array orientation, typical deposits and the intended output: a cleanliness trend, an electrical loss estimate or evidence for a cleaning study. Include the comparison method and acquisition system. This makes the sensor and validation scope explicit without promising a universal conversion from one optical percentage to revenue.

Soiling Measurement FAQ

Q1. Does 96.5% cleanliness mean the plant has lost 3.5% of its energy?

No. It indicates a 3.5-percentage-point deficit from a nominal 100% reference on that optical scale. Electrical and plant-energy losses require a suitable relationship and operating data for the represented modules and period.

Q2. Are the two PSS outputs automatically a clean and dirty electrical reference pair?

No. They are two cleanliness channels. Preserve their labels and use the operating method specified for the instrument; do not infer a separate electrical reference system from the presence of two outputs.

Q3. Can the same optical conversion be used at every site?

Not without evidence. Deposit composition, spatial distribution, module response and exposure can differ. Validate the conversion under the conditions for which the project will use it.

Q4. Why can the sensor improve after rain while plant performance does not?

The instrument and modules may have different wetting or cleaning outcomes, while plant output also depends on irradiance, temperature and electrical constraints. Compare stable post-event conditions rather than attributing an immediate change solely to cleaning.

Q5. What should an acceptance test for a loss model include?

It should identify the optical inputs, independent electrical comparison, matching time windows, exclusions and the range of conditions tested. A successful Modbus read is an interface test, not validation of the conversion model.

Summary

Optical soiling measurement is useful when its local cleanliness indication remains distinct from the electrical conclusions drawn from it. NiuBoL PSS readings can support a documented comparison and inspection process. Preserve the raw channels and intervention history, then apply only the loss relationship that the site evidence supports.

NBL-W-PSS Soiling Sensor Photovoltaic Dust Monitoring Instrument Data Sheet.pdf