The importance of sub-hourly input data in PV systems simulation using satellite-based solar model data

Open presentation

Overview

Accurate PV system simulation and performance analysis increasingly depends on the temporal resolution of input data. This study, presented at the European PVPMC Workshop 2024, explores the impact of using sub-hourly (15-minute and 1-minute) versus hourly satellite-based solar resource datasets in PV performance simulations. It demonstrates how sub-hourly data uncovers key information about PV power variability, clipping losses, and grid impacts—critical aspects that are often hidden when relying solely on hourly datasets.

The findings are based on Solargis in-house PV simulator, applied to ground-mounted, tracker-based PV systems at sites representing different climate categories. The analysis includes both single-year and multi-year (with degradation) scenarios and addresses the practical implications for system design, performance evaluation, and grid integration.

Key methodology

• Simulations used Solargis satellite-based datasets at 60-minute, 15-minute, and synthetically generated 1-minute resolutions.

• PV system configuration was standardized: ground-mounted with N-S trackers, bifacial half-cut cells, and one inverter, across four global sites representing tropical (Malaysia), dry (Chile), temperate (Italy), and continental (Sweden) climates.

• The approach included multi-year analyses (up to 5 years) with and without performance degradation.

• DC to AC ratios between 1.1 and 1.7 were tested to assess the effect on results.

• The study specifically compared PV output (PVout) and clipping losses (CL) at varying timescales.

Main findings

• Sub-hourly data reveals hidden system variability: Using only hourly data masks short-term fluctuations in irradiance and output, particularly in regions with intermittent clouds or high irradiance variability.

• Quantification of clipping losses: Sub-hourly data provides a more accurate estimate of clipping losses. Over yearly or multi-year horizons, sub-hourly timescales capture non-linear reductions in clipping due to PV system degradation, which hourly data frequently underestimates.

• Differences in PV output calculation: The gap between PV outputs calculated using hourly versus sub-hourly data is most pronounced in tropical climates and for systems with higher DC/AC ratios. For instance, in Kuching (Malaysia), overestimation of annual PV output using hourly data exceeded 1% (DC/AC ratio of 1.3); in dry climates like Calama (Chile), the difference was smaller but still significant.

• Grid and battery system design: Sub-hourly data enables accurate modelling of rapid PV power changes, essential for external grid evaluation and planning for battery storage.

• Optimal data resolution: 1-minute data is the most accurate, but 15-minute data offers a practical compromise between computational cost and modelling accuracy.

Conclusions

Sub-hourly time series provide essential insights missing from hourly simulations—particularly in quantifying clipping losses and accurately evaluating PV system performance under real-world variability and degradation. The benefits of sub-hourly analysis increase with higher DC/AC ratios and in more variable climates. While hourly data is sufficient for general performance estimation, detailed, high-resolution datasets are required for rigorous system evaluation, energy storage system configuration, grid impact studies, and bankable design—highlighting the need for careful selection of input data resolution in PV simulations.

Further reading

• “The importance of sub-hourly input data in PV systems simulation using satellite-based solar model data” by Jozef Rusnak, Branislav Schnierer, Marcel Suri.