Electricity storage creates economic value by providing four fundamental services:
Power Quality: Keeping frequency and voltage within permissible limits
Power Reliability: Providing electricity in case of supply reduction or interruption
Increased utilisation: Optimising use of existing assets in the power system
Arbitrage: Exploiting temporal price differentials
Markets vary substantially around the world, with heterogeneity in their structure and ownership, level of vertical integration, regulations, and the actual services and products that are offered. In order to identify insights that are common across major electricity markets, Figure 1 compares the economic value (i.e. revenue potential) of storage across electricity markets in the United States, Great Britain, Germany, and Australia. These markets are chosen as they have been widely explored, and so this figure combines 176 values found for the major applications, drawn from recent studies and market analyses.
Figure 1 – The value of different electricity storage applications across four major markets. Data are taken from 176 individual valuation studies and published market transactions as compiled by Balducci for the US and Housden for Great Britain, Germany, and Australia. The specifications of each application vary across the individual studies and are not necessarily aligned. The scope of each study also varies in terms of the timeframe and market considered (as the US and Australia have multiple electricity markets). All values are presented in USD/kW-year even for applications which are remunerated by energy discharged rather than power capacity, as this then incorporates device utilisation. A value of 100 USD/kW-year can be interpreted as a 1 MW system receiving USD 100,000 annual revenue from service provision.
Three key features evident from this figure are:
Variability within the same application and market (compare the points within a single column). For example, valuations of distribution network deferral within US markets range from 9 to 177 USD/kW-year.
Variability within the same application across markets (compare the four columns of points within a given box). For example, ramping reserve averages just 19 USD/kW-year in the US versus 131 USD/kW-year in Germany.
Variability across applications within the same market (compare columns of the same colour in different boxes). For example, black start is worth 3 USD/kW-year in Great Britain, whereas congestion relief is worth 50 USD/kW-year.
Figure 2 summarizes the values provided in Figure 1, revealing some general trends despite the wide variability. The highest revenues, generally above 100 USD/kW-year, are available in all markets for providing frequency regulation and customer services (managing time-of-use charges). The latter is especially important in Germany and Australia, due to high residential electricity prices. These services are more valuable as they have fewer viable competitors than, for example, arbitrage. The value of arbitrage is broadly in the region of 40–80 USD/kW-year, which is tied to the incremental cost of conventional generation technologies. Transmission deferral is another service with similar and consistently high value across American, European, and Australian markets, due to the high economic cost and non-monetary barriers to building or upgrading transmission lines. Black start is also comparable between regions as the least valuable service, as it can be provided by many existing power stations, and the cost of providing it is minimal due to few cycles and short duration.
Some notable differences are visible. The value of peak capacity is low in Great Britain relative to other markets (7 versus 49–106 USD/kW-year), due to power generation overcapacity in the market during the timeframe studied. Congestion relief is more valuable in the US and Great Britain than in Germany, as grid congestion is a greater problem due to limited grid interconnection. The US average of 71 USD/kW-year is skewed by one especially high value in ERCOT (Texas) which has particularly low interconnection to neighbouring markets. This explains the US mean being above its 75th percentile, and without this value the US mean falls to 26 USD/kW-year, in line with Germany. Australian markets also have limited interconnection in the south-western states, but it is difficult to comment on the value provided as it only comes from a single study.
Figure 2 – Summary of electricity storage valuation studies and transactions in four major markets, giving the mean, 25th and 75th percentiles across all studies for each market and service. All values in USD/kW-year.
Assessing the economic value of electricity storage requires transparency around the variation of this value along application requirements. Figure 3 matches the potential revenues for storage in different applications to the respective discharge duration and cycle frequency requirements in the US markets.
While power reliability applications have up to six hours discharge duration but less than 100 full discharges per year, power quality applications are characterized by less than one-hour discharge duration at various discharge frequencies. Applications that deliver increased asset utilisation have between one and eight hours, and up to 500 discharges. For arbitrage, there are two types: discharge duration below one hour at up to 350 cycles and discharge duration up to six hours at below 250 annual cycles.
There seems to be a positive relationship between economic value and increasing discharge and frequency requirements (i.e., increasing number of running hours). Applications with up to eight hours discharge duration or 10,000 cycles are valued at around 125 USD/kW-year respectively, and those with a moderate mix of discharge duration and cycle frequency are valued at around 105 USD/kW-year. The values reduce with a reduction in discharge duration and annual cycle frequency, down to 8 USD/kW-year for black start at only one hour and 10 to 20 cycles. The residential arbitrage application time-of-use bill management represents an outlier with only 65 (7-130) USD/kW-year at up to six hours discharge and 250 annual cycles. However, a detailed review of the respective studies reveals that those valuing storage at the lower end of this range assume discharge durations below four hours, more in line with the identified value-requirement relationship.
Figure 3 – Mean economic value for different electricity storage applications in US markets, measured in USD/kW-year (from Figure 1 and 2) plotted along respective discharge duration and frequency application requirements. 25th and 75th percentile economic values shown in small font.
Figure 4 expands this value analysis to the entire spectrum of possible discharge duration and frequency requirements from 1 to 1,024 hours and 1 to 10,000 cycles. It also aims to generalize this analysis on revenue potential for energy storage by forming the average across the four studied major markets. It confirms the positive relationship identified between revenue potential and increasing discharge and frequency requirements. Some exceptions to this trend also shine through: potential ‘sweet spots’ identified for selected markets and timeframes. A Monte-Carlo analysis accounts for the ranges in economic value per application and market (see Figure 1), and the ranges in discharge duration and cycling frequency for each application (see Figure 3). Each discharge-frequency combination on the spectrum is assigned an economic value (in terms of USD per kW of power capacity per year) by interpolating the Monte-Carlo results.
While this analysis reveals high-level insights that support the general understanding of energy storage revenue potential, investment decisions should be based on detailed, market-specific data. This requires a bespoke assessment of detailed revenue data for a specific market, which would allow these landscape figures to be reproduced with greater coverage and validity.
Figure 4 – Graphic representation of trends in revenue potential for energy storage across the landscape of cycling frequency and discharge duration requirements.