Portfolio analysis of solar photovoltaics: Quantifying the contributions of locational marginal pricing and power on revenue variability
K. Kumpf, S. Blumsack, G. S. Young, and J. R. Brownson
Solar Energy (September 2015)
DOI: 10.1016/j.solener.2015.06.008
Abstract For firms developing and managing portfolios of PV assets in utility, commercial, and residential markets the financial performance of the portfolio is a highly relevant business decision factor. Informative metrics are required to quantify the revenue variance of a spatially distributed portfolio of PV assets as well as the individual asset. Financial analysis uses a risk measure known as ‘beta’ to describe the movement of assets relative to a broader portfolio. We define a measure termed the ‘solar beta’ that describes the movement of solar PV revenues at a given site with that of a portfolio of sites. The solar beta incorporates correlation between a site and portfolio, and the volatility of a site relative to the portfolio. We also derive and discuss a method to decompose revenue variance of individual PV assets into components representing price, power, and the interplay among diurnal/seasonal cycles and prevailing weather conditions. This decomposition and the solar beta are illustrated using nine modeled sites following a N–S trend within the PJM Interconnection. We find that revenue variance of a PV asset depends more on diurnal, seasonal, and meteorological fluctuations than on price fluctuations at a particular site. Specifically, the contribution of power variance exceeds the contribution of price variance by roughly a factor of five. Changes in mean market price have a larger effect on revenue variance compared to a proportional change in mean power production. The solar beta was found to be near 1.0 for most sites, indicating strong covariance within the portfolio due in large part to high correlation rather than similar volatility ratios. Lower beta values were found for sites at the perimeters of the study region, due to change in climate regime and population-power consumption cycles, implying portfolio risk reduction when these sites are included.
keywords: Photovoltaics; Locational marginal pricing; Portfolio analysis; Beta