Skip to content

Spectral Correction

Spectral correction accounts for how atmospheric scattering and absorption alter the solar spectrum, impacting PV module performance differently from clear sky spectrum. These variations depend on atmospheric water content, aerosols, and path length, quantified by Air Mass (AM). PVsyst implements several models to describe spectral correction:

  1. The CREST model that is used for amorphous silicon modules. This correction is applied automatically when spectral correction is activated.
  2. The spectral correction for PV modules in the Sandia database. This correction is applied automatically when spectral correction is activated.
  3. The FirstSolar spectral correction model. Disabled by default; user must enable it manually in 'Detailed losses'.

First Solar Spectral correction model management

The FirstSolar spectral correction model is switched on in the dialogue 'Detailed losses':

SpectCorrDialogue

Choice of coefficient set

When spectral correction is enabled, the coefficient set matching the PV module technology is used by default during simulation. You can override this by unchecking the corresponding box and selecting any available coefficient set. Once overridden, the selected coefficient set remains fixed regardless of changes to the PV module's technology in system definitions. Users must ensure the chosen set is appropriate for the specific simulation.

Prerequisites in weather data

The FirstSolar spectral correction model uses Air Mass and Precipitable Water Column as input variables. Air mass is calculated from altitude and sun position, while precipitable water must either be available in the weather data file or be estimated from relative humidity. Therefore, this correction can only be applied during simulation if either precipitable water or relative humidity is present in the weather data. For synthetic hourly weather data files created with PVsyst 6.7.4 or later, relative humidity is always present.

Corretion factor graph

The plot displays the correction factor as computed by the model for the selected coefficient set in a range from 0 cm to 10 cm of precipitable water and an air mass of 1 up to 5. The color code is chosen such that a factor of 1 corresponds to green. Factors larger than one will get shades of turquise and blue, whil factors smaller than one will become yellow to red. The plot also marks a distinct point as 'STC'. It is calculated for a precipitabl water column of 1.416 cm and an air mass of 1.5. These values correspond to the conditions of the AM1.5 spectrum used at Standard Test Conditions (STC). Normally this point should lie close to the iso-line for the values of 1. To get an idea on how the spectral correction model impacts the simulation results the user can check the option 'Overlay simulation points'. This will display the pairs of precipitable water and air mass for all the data points in the weather data.

Output on report

If spectral correction is used in the simulation, it will be noted on the final report. The system summary will list the coefficient set used, along with the graph of the correction factor. The loss diagram at the end of the report will include a 'spectral losses' contribution.

Custom coefficients

It is possible to customize the six coefficients of the FirstSolar spectral correction model. To enable the editing of the coefficients, select one of the predefined coefficient sets, and press the button 'Use as custom coefficients'. The choice of the coefficient set will change to 'Custom', and the six coefficient values will become editable. If you select directly 'Custom' in the drop-down options, the default values of the coefficients will all be zero except for C0, which is set to one. When using a custom set of coefficients, PVsyst enforces bounds on the spectral correction factor to ensure physically reasonable results:

  1. Validation Range: Within the typical range of 0 to 10 cm PWc and AM 1 to 5, the correction factor must fall between 0.6 and 1.2. If it falls outside this range, a warning is displayed.
  2. STC Check: The model also computes the correction value for standard test conditions (1.416 cm PWc at AM 1.5), which should approximate 1.0. If this value strays outside the narrower tolerance of 0.99 to 1.25, a warning is issued.