Goals

The following table is a comparison of the underlying models used in different software's respective model-chains. The goal of the table is to provide readers an easy reference guide to compare and contrast different photovoltaic performance modeling software. The table also includes information about whether or not the implementation in the software matches the implementation in PVLlib. Not every single model can be found in the table. Some sub-models are omitted such as the calculation of delta T in the solar position algorithm in order to save space. If the performance modeling software has multiple models for a given model type, only the most commonly used models are included.

Column Descriptions

• Category: A grouping of models that relate to a specific folder in the repository structure
• Model_Type: The type of model being described in the table
• Model: The name of the model used by a given performance modeling software
• Metric: Accuracy is presented in terms of the following metrics depending on what is stated in the cited paper:
  • mean bias difference (MBD),
  • root mean squared difference (RMSD),
  • maximum difference (MxD),
  • standard deviation (SD),
  • R squared (R2)
  • Uncertainty (U)
  • Model is precise (-)
  • If the cell in the table is blank, then the accuracy is unknown. If you know the accuracy of a given model, please see the "Contributing" section which can be found after the table.
• Accuracy:
  • If an accuracy value is present in the table, then the accompanying publication is included in the repository folder structure.
  • Published accuracy must be from a third party and reputably published in order to be included in this table.
  • If many different published accuracy values exist such as accuracy values at different time-steps, then only the most relevant values are included in the table.
  • Discerning readers may refer to the reference for additional information.
• PVLIB: The model implemented in the software is identical or very nearly identical to the implementation in pvlib.

Table

		PlantPredict				PVSyst
Category	Model_Type	Model	Metric	Accuracy	PVLIB	Model	Metric	Accuracy	PVLIB
Solar_Position_Algorithm	Solar_Position_Algorithm	NREL	U	0.0003°	☑	Proprietary	U	0.08°
Time Series	Minimum Time Step	1 minute	MxD	~< 0.4%		1 hour	MxD	~< 1.8%
Fixed Racking	Fixed Tilt	Yes	-	-		Yes	-	-
	East West	No	-	-		Yes	-	-
	Seasonal Tilt	Yes	-	-		Unknown	-	-
Tracking	Astronomical Tracking	Marion & Dobos	-	-	☑	Proprietary	-	-
	Slope Aware Backtracking	Anderson & Mikofski	-	-	☑	Proprietary	-	-
	Terrain Aware Backtracking	None				None
	Custom Backtracking	Yes	-	-		No	-	-
	Diffuse Optimization	3 Parameter				1 Parameter
	Wind Stow	None				Fixed Threshold
Irradiance	Decomposition	Dirint	MBD RMSD	-3 W/m2 74 W/m2	☑	Erbs	MBD RMSD	-17 W/m2 88 W/m2	☑
	Transposition	Perez w/ coefficients	MBD RMSD	-1% 8%	☑	Perez w/out coefficients	MBD RMSD	-1% 8%	☑
	Retro-Transposition	GTI-Dirint	MBD RMSD	0.0 W/m2 6.4 W/m2	☑	Reverse Hay	MBD RMSD	Unknown Unknown
	Reflection on Front Side	Yes				Yes
Optical	IAM	Physical			☑	Fresnel
	Soiling	Time-Step Level				Monthly
	Horizon Shading (Far Shading)	Sub-TimeStep				1 hour
	3D Near Shading	Polygon Clipping	-	-		Polygon Clipping	-	-
	Bifacial	NREL BifacialVF	U	<2%	☑	Proprietary VF	U	<2%
Spectral	Relative_Humidity from Dew_Point	August-Roche-Magnus				None		N/A
	Precipitable_Water from Relative_Humidity	Gueymard94	MxD	<20%	☑	Gueymard94	MxD	<20%	☑
	c-Si	First Solar v2.0			☑	Crest
	a-Si	First Solar v2.0			☑	Sandia			☑
	Cd-Te	First Solar v2.0			☑	First Solar v2.0			☑
DC	Cell Thermal Model	Modified Faiman				Modified Faiman
	Diode Model	Single Diode				Single Diode
	Mismatch	Simple Statistical				Detailed
	Wiring	% at STC				Ohmic
	Degradation	%				%
AC	Inverter Off MPP	Proprietary				Proprietary
	Inverter Efficiency	Sandia			☑	Quadratic Interpolation
	Air Pressure for Altitude Correction of Inverters	Yes	-	-		No	-	-	☐
	Aux Losses	% loss	-	-		W loss	-	-
	Transformer	McCann Lawrence				Proprietary
	Wiring	% at STC				Ohmic

Contributing

The table is open source and any user or software company can submit a pull request to include their information in this list.

• Rules for contribution:
  • If a free reference exists for the underlying model, please include the PDF in a folder that matches the model category.
  • If no reference exists (especially for proprietary software algorithms) then the best known reference is provided in the "Notes" markdown file instead.

Contributors

• Kurt Rhee
• Renn Darawali
• James Alfi
• Paul Wolfersdorff

References

![[pvlib_logo.png]] William F. Holmgren, Clifford W. Hansen, and Mark A. Mikofski. “pvlib python: a python package for modeling solar energy systems.” Journal of Open Source Software, 3(29), 884, (2018). https://doi.org/10.21105/joss.00884