The research employed the Engerer2 decomposition model with two distinct clear-sky schemes. The first utilized Threlkeld-Jordan, used in Engerer2’s original calibration, while the second applied ARGPv2, an empirical model specifically developed for high-altitude sites in northwestern Argentina’s Salta and Jujuy provinces and recalibrated using the McClear clear-sky model. The resulting irradiance components fed into a Liu-Jordan transposition model to calculate radiation on module planes, followed by standard test conditions-based PV production modeling across full matrices of tilt and orientation combinations.
Without direct DNI and DHI measurements at either site, researchers compared decomposition results against Solargis satellite-derived data. This comparison revealed systematic divergence between approaches that varies with altitude, particularly significant for Altiplano and elevated Andean projects where atmospheric conditions deviate from those underlying conventional model calibrations.
The research highlights limitations in GHI-exclusive solar resource assessment methodologies. While enabling long time-series reconstruction and PV production calculations at sites lacking comprehensive radiation measurements, uncertainty in separating direct and diffuse components becomes systematic at high altitudes. The limited availability of direct DNI and DHI measurements represents a primary challenge for solar resource studies across Andean regions, forcing reliance on models that introduce additional uncertainty when only GHI data exists.
The findings indicate that models calibrated for lower-altitude conditions may systematically misestimate irradiance components in high-elevation environments, potentially propagating errors through project development workflows. This has direct implications for generation forecasts and project sizing in Peru’s expanding highland solar sector, where atmospheric transparency and reduced air mass differ substantially from coastal or lower-elevation reference conditions used in standard model development.
This article was curated and published as part of our South American energy market coverage.
