Quantifying photovoltaic fluctuation with 5 kHz data: Implications for energy loss via maximum power point trackers

Abstract

We aim to systematically quantify photovoltaic (PV) variability contained within different frequency bands, primarily for applications in PV maximum power point tracking (MPPT) design. We first discuss the usefulness in quantifying energy capture for various maximum power point (MPP) update rates from nearly 500 days of 5 kHz photovoltaic recordings. Next, we justify the methods used to convert MPP sweep data to single-point, usable, current, voltage, and power values. We discuss fitting methods that yield the MPP under calm irradiance dynamics and explore the approach used during periods of more stochastic changes. This is followed by analysis of raw, high-frequency content and a proposed method to calculate associated energy capture reduction. The conclusion finds an absolute upper bound on solar data variability for a given MPP update rate in terms of energy capture. Finally, we use the previous results and demonstrate an economic analysis that can aid in designing future MPP tracker specifications.