Effective DC power rating of PV arrays under challenging operating conditions in desert and tropical regions

The accurate characterization of photovoltaic (PV) system performance is essential for diagnostics, benchmarking, and O&M. Conventional performance ratio (PR) metrics, standardized in IEC 61724–1, are widely used but remain highly sensitive to irradiance variability, thermal dynamics, and curtailment, often generating false alarms in challenging climates. This study extends a recently proposed statistical method for estimating the effective DC power rating (P₀,eff), the array's nominal power corrected to standard test conditions, by testing multiple irradiance thresholds at two contrasting Peruvian sites: the arid desert of Lima and the tropical rainforest of Chachapoyas. Results show that P₀,eff provides a more stable indicator than PR and PR₍₂₅₎, with uncertainties below 3%. High thresholds (>800 W/m²) yielded the lowest variability (≈1%), while intermediate thresholds (>600 W/m²) balanced stability with greater data coverage. In Lima, the method captured capacity losses from dust deposition, whereas in Chachapoyas it proved robust under persistent cloudiness, where PR fluctuated strongly. A monitoring protocol is proposed in which PR serves as the primary indicator and P₀,eff validates alarms when PR falls below a threshold. This combined approach reduces false alarms while retaining sensitivity to genuine performance losses, offering a practical and climate-resilient tool for PV monitoring and O&M optimization.

The accurate characterization of photovoltaic (PV) system performance is essential for diagnostics, benchmarking, and O&M. Conventional performance ratio (PR) metrics, standardized in IEC 61724–1, are widely used but remain highly sensitive to irradiance variability, thermal dynamics, and curtailment, often generating false alarms in challenging climates. This study extends a recently proposed statistical method for estimating the effective DC power rating (P0,eff), the array's nominal power corrected to standard test conditions, by testing multiple irradiance thresholds at two contrasting Peruvian sites: the arid desert of Lima and the tropical rainforest of Chachapoyas. Results show that P0,eff provides a more stable indicator than PR and PR25′, with uncertainties below 3 %. High thresholds (>800 W/m2) yielded the lowest variability (≈1 %), while intermediate thresholds (>600 W/m2) balanced stability with greater data coverage. In Lima, the method captured capacity losses from dust deposition, whereas in Chachapoyas it proved robust under persistent cloudiness, where PR fluctuated strongly. A monitoring protocol is proposed in which PR serves as the primary indicator and P0,eff validates alarms when PR falls below a threshold. This combined approach reduces false alarms while retaining sensitivity to genuine performance losses, offering a practical and climate-resilient tool for PV monitoring and O&M optimization.