The Economics of Variable Renewables and Electricity Storage

Abstract

The transformation of the electricity sector is a main element of the transition to a decarbonized economy. Conventional generators powered by fossil fuels have to be replaced by variable renewable energy (VRE) sources in combination with electricity storage and other options for providing temporal flexibility. We discuss the market dynamics of increasing VRE penetration and their integration in the electricity system. We describe the merit-order effect (decline of wholesale electricity prices as VRE penetration increases) and the cannibalization effect (decline of VRE value as their penetration increases). We further review the role of electricity storage and other flexibility options for integrating variable renewables, and how storage can contribute to mitigating the two mentioned effects. We also use a stylized open-source model to provide some graphical intuition on this. While relatively high shares of VRE are achievable with moderate amounts of electricity storage, the role of long-term storage increases as the VRE share approaches 100%.

Renewable Energy Targets and Unintended Storage Cycling: Implications for Energy Modeling

“…It is a stylized version of the larger Dispatch and Investment Evaluation Tool with Endogenous Renewables (DIETER) capacity expansion model [17]. A similar model set-up has been previously used for analyzing electricity storage needs for renewable energy integration [20], for reflections on the changing role of electricity storage in the renewable energy transition [51], and for an illustration of the economics of renewables and electricity storage [52]. The implementation used here is integrated in a Python-GAMS interface, which enables Python-based data pre-and post-processing, scenario analysis, and visualization [53].…”

“…Both the electricity demand and the output of many generation technologies are fluctuating in nature. Especially the value of VRE depends on the time when they generate, i.e., their variable availability profile [52], and the location of generation in the geographical context. Consequently, the electricity price fluctuates over time and space, too.…”

“…A standard RLDC sorts all hourly residual load values of a full year in descending order (solid lines in Figure 7). With increasing VRE penetration, the RLDC shifts downwards especially on the right-hand side [52]. Augmented versions of residual load duration curves can be calculated that also consider optimal curtailment and electricity storage use (Figures 6 and 7).…”

“…Figure 8 also shows that unintended storage cycling occurs only in low-price periods, i.e., in hours with renewable curtailment. Note that prices may become negative in models using constraint family (1), which can be interpreted as the consequence of an energy-based renewable support mechanism [49,52]. Price formations of models based on constraint families (2) to (4) are characterized by the same variation in the distribution across SLCR levels, but a constant level shift may apply for each constraint family.…”