Worst-Case Photovoltaic Generation and Power Change Distribution Under Dense Cloud Cover

Abstract

Photovoltaic (PV) power generation units located in tropical or equatorial sites frequently experience low-lying, dense cloud cover; and are thereby exposed to rapid real time changes in irradiance. Commonly referred to as cloud transients, the changes lead to short duration power variations and substantial reduction of output from individual units. The paper proposes pessimistic (or worst-case) analytical assessment of power variation statistics at a PV generation unit, as well as the consequent drop in output power. Statistical estimates are introduced for worst-case short duration mean output power, its temporal variability, and the associated power change distribution. The metrics are verified against ensemble estimates from field studies conducted in Brazil and India. Statistical estimates have the advantage of exclusive parametric dependence on optical air mass, which makes them applicable across diverse geography and climate. They can therefore be used for quick evaluation of short duration performance as well as comparative studies, both at existing as well as prospective PV generating stations.