Electric power systems in many parts of the world are undergoing a transformation from relying almost exclusively on dispatchable power (e.g., fossil, nuclear, and large hydropower) toward incorporating more variable nondispatchable generation (e.g., wind and solar PV). We show for the first time that solar generation can decrease some aspects of variability in the peak residual load in power systems. The electric load minus generation from nondispatchable resources is known as the “residual load.” The maximum or peak residual load provides an estimate of the quantity of dispatchable generation capacity required to supply electric load during all hours. We study the peak residual load as a function of increasing wind and solar generation for three power systems in the U.S.: the PJM system in the Mid-Atlantic, the ERCOT system in Texas, and the NYISO system in New York. We analyze more than a decade of historical data for each region. The introduction of variable renewable power is often thought to increase the variability of most characteristics of power systems. Contrary to this idea, we show the inter-annual variability in peak residual load decreases for all three systems as a function of increasing solar generation. We attribute this effect to correlations between solar generation and peak electric load values. Peak electric load values for all three systems occur during summer heat waves, when air conditioning is used. We find that as solar generation increases, the quantity of dispatchable generation capacity needed to supply the residual load becomes more similar year-to-year. Therefore, in some systems, expansion of variable solar generation can increase predictability of the peak residual load. Thus, an increase in solar generation could ease achievement of certain system reliability targets.
Keywords residual load, net demand, peak load, resource variability, inter-annual variability