Temporal Resolution of Solar Resource Data and Reliability of PV–Battery Systems: A Five-Year Multi-Climate Chronological Simulation

Authors

  • Morgan D. Phill PhD, Department of Renewable Energy Engineering, Oregon University, Oregon, USA Author

Keywords:

Photovoltaic-Battery System, Temporal Resolution, Nasa Power, Loss of Power Supply Probability, Battery Sizing, Solar Resource

Abstract

This study quantifies how temporal treatment of solar-resource data changes the simulated reliability and storage requirements of a photovoltaic (PV)–battery system. Hourly NASA POWER data for global horizontal irradiance (GHI), 2-m air temperature, and 10-m wind speed were analyzed for Phoenix, Tucson, Atlanta, London, Munich, and Rovaniemi over 1 January 2019 to 31 December 2023. The native hourly series (H) was compared with two daily-to-hourly reconstructions: a 24-h flat reconstruction (D_flat) and a daylight-weighted sinusoidal reconstruction (D_sin). The baseline system comprised a 10-Wp PV array, a constant 2-W DC load, and a 50-Wh nominal battery, with 80% maximum depth of discharge, 85% round-trip efficiency, and 100% initial state of charge. PV output was simulated using tilted-plane irradiance, module temperature, and a PVWatts-type relation. Reliability was measured as the fraction of load-hours with unmet demand (LPSP), and nominal battery capacity was searched for 1% and 0.1% LPSP targets. Annual PV yields under H ranged from 8.55 to 20.24 kWh yr−1, while D_flat reduced yield by 12.1–38.8% relative to H. With the 50-Wh battery, LPSP under H ranged from 6.4% in Phoenix to 62.1% in Rovaniemi, whereas D_flat increased LPSP by 55–349%. In contrast, the relative LPSP difference under D_sin ranged from −5.1% to +4.2% compared with H. Friedman’s test showed a significant overall treatment effect on LPSP (χ²=9.33, p=0.0094, Kendall’s W=0.78). At the 1% target, the required nominal battery capacities in Tucson and Phoenix were 592–728 Wh under H/D_sin, compared with 2.45–3.82 kWh under D_flat. The results show that preserving daily solar energy does not necessarily preserve chronological PV–battery reliability and that daylight-constrained reconstruction is substantially less biased than uniform 24-h redistribution.

Downloads

Published

2026-07-02

Issue

Section

Articles