@cwhanse thanks very much for your thoughtful feedback and taking the time to review my post. I appreciate your perspective on capacity testing (effective nominal power) and the suggestion to consider integrating this method into pvanalytics.

Would you recommend that I open a new Issue in the pvanalytics repository to propose this addition? Or are there any preliminary steps or guidelines you suggest I follow before submitting a proposal?

Thanks again for your time and willingness to engage with this idea

@JRAnguloPUCP an issue referencing this discussion would be a good starting point. I wouldn't expect a lot of discussion on pvanalytics, it doesn't get the traffic and attention that pvlib-python does.

To be candid implementing the method you describe is better to be done in multiple, smaller pull requests as we identify each function in the process and an appropriate location, which may be in pvlib-python as @adriesse comments.

@markcampanelli IEC TS61724-2 leaves the choice of a power prediction model to the interested parties. I think that complicates comparison with ASTM 2848 and similar capacity methods, which specify a prediction equation.

I’m not sure I agree. The ASTM-E2848-13 “specific regression equation” is first fit to an arbitrary prediction model like PVsyst or SolarFarmer. Also, the equation used in practice does not always match the standard, especially since it doesn’t clarify what is E_poa or how to deal with backside irradiance or bifaciality

The ASTM-E2848-13 “specific regression equation” is first fit to an arbitrary prediction model like PVsyst or SolarFarmer.

I'm confused by this statement. Are you referring to some kind of procedure for relating a measured capacity to a contract proforma capacity? It is my understanding that the ASTM-E2848 equation (which is the PVUSA model) is not fit to a proforma model, but rather to the measurements at the site (after filtering). Then that fitted equation is evaluated at the reference condition to determine its measured capacity.

You make good points about the model (and its inputs) being less than completely specified for all situations.

Thank you very much for your observation and for the opportunity to clarify this point.
IEC TS 61724-2:2016 includes an equation (Annex B, Eq. B.2) that describes the power output of a photovoltaic generator based on irradiance and module temperature.

This is the similar discussed in the paper Martinez et al 2012. However, B.2 equation is intended to obtain the predicted power and compare it with the measured. In both Martinez et al 2012 and in our proposal, the objective is to determine the effective nominal power under outdoor conditions.
Regarding the definition of “effective nominal power,” this parameter represents the power under conditions approximating STC (1000 W/m², 25 °C, and AM1.5), but it considers cable losses, module mismatch, and other inefficiencies inherent to field operation—especially since the measurement is taken prior to the PV inverter. This distinguishes it from the manufacturer’s nominal power, which is based on the sum up of each module’s power at STC and does not always accurately reflect real operating conditions.
By calculating the effective nominal power, a more realistic representation of the system’s overall performance is obtained, which is invaluable for evaluating profitability, establishing warranties, and making reliable comparisons between the installed and the expected power.
While Martinez et al 2012 clearly defines how to calculate this parameter in the field, its method is limited by environmental conditions (only clear sky condition), abrupt changes in irradiance, and/or module temperature fluctuations. In our proposal, we apply non-parametric statistical methods to process the data under non-ideal condition, thereby achieving the same value as would be obtained under ideal clear-sky conditions. This method was initially validated on a 104 kW plant in Granada, Spain, and subsequently tested more rigorously on systems installed in desert areas, such as in Lima, as well as in regions with high cloud cover, like the tropical Amazon near Chachapoyas Peru, where it would be impossible to have ideal conditions MatER-Group
For example, after four years of operation of a photovoltaic generator, one can compute the effective nominal power with outdoor data, which can then be used to model the predicted power with new values of irradiance and module temperature.

Finally, and to comment on the ASTM E2848-13 standard, it defines the Power Test Condition capacity rating, the test condition could be STC and thus obtain the effective power. To do this, the coefficients of:

must first be calculated through a linear regression example is this, according to the ASTM standard, can only be done under ideal conditions.