(Invited) Electric Energy Storage Technologies for Grid-Scale Applications

Abstract

Large-scale electric-energy-storage technologies (EESTs) have the potential to transform the electric grid by significantly improving the operating capabilities and mitigating infrastructure investments[1]. EESTs are also critical to enabling a predominance of carbon-free electricity-generation sources, such as intermittent solar and wind power[2]. In 2015, the U.S. DOE identified four key barriers that had historically limited the widespread deployment of EESTs: 1) lack of cost-competitive systems, 2) need for validated performance and safety, 3) development of equitable regulatory environments, and 4) widespread industry acceptance[1]. In the past few years, deployment of EESTs has been growing rapidly. In 2018, the annual deployment worldwide was > 3GWh, which is 4.4X times larger than the 0.7 GWh deployed in 2015; however, the large majority of these deployments were lithium-ion batteries (88% of EEST batteries in 2016)[3]. The key topic of this symposia and the focus of this talk is: Which stationary energy-storage applications can be potentially be best addressed by non-lithium-ion-battery solutions?There are many types of grid-scale storage applications, with a wide variety of requirements. For example, some grid applications require short response times and relatively short discharge durations at rated power (e.g., some Ancillary services, such as area regulation, typically require discharge durations of ≤ 2 h)[4]. Lithium-ion batteries are well suited for these applications and are a relatively mature EEST. Since these applications also have relatively high economic benefits on a power basis (i.e., $/kW), it is not surprising that these applications have dominated recent battery-based EEST deployments. At these low energy/power ratios, lithium-ion batteries can be lower cost than the historical EEST baseline, which is pumped-storage hydropower (PSH)[5]. However, since the cost of most conventional batteries scales linearly with energy capacity, the energy-storage cost ($/kWh per cycle) is not lower for applications that require longer discharge cycles. For these long-duration energy-storage (LDES) applications, EESTs with independent power and energy components (e.g., PSH, redox flow batteries[6], reversible fuel cells[7], and electrofuels[8]), will have a distinct cost advantage, since one can readily increase the energy/power ratio by selectively adding more energy capacity, without the cost of additional power components. Even if one considers only LDES applications, there are still a variety of requirements, and different types of EESTs may be best suited for certain applications. From a cost perspective, one must consider both capital and operational cost. This is typically done by calculating a levelized cost of storage (LCOS)[5], and LCOS strongly depends on the energy-storage application. For example, operational costs (e.g., cost of input electricity) is far more important for daily-cycle applications than for seasonal-storage applications, where capital costs will dominate.Another key criterion is safety. Risks are typically assessed based on the likelihood of occurrence and the potential impact of each occurrence. With large-scale systems, the impact of a single incident can have substantially larger consequences than is the case with much smaller units. Even though lithium-ion batteries have an impressive safety record (based on the total number of cells), if the potential impact is sufficiently high, then even a small likelihood may be deemed unacceptable. The result of some failure modes of GWh-scale battery systems can be catastrophic, especially with batteries that use flammable or toxic components. However, although batteries with aqueous electrolytes are inherently safer than batteries with organic electrolytes, even aqueous systems have safety concerns that must be adequately addressed[9].In addition to assessing what EESTs are best suited for various applications, this talk shall also consider what future technology improvements have the greatest potential to make these EESTs substantially more compelling. We shall also consider how anticipated market forces (e.g., projected costs of renewable electricity) may impact the key requirements and thereby increasingly favor certain EEST attributes for the future grid. Acknowledgements Many thanks to my flow-battery collaborators at Vionx Energy and at Raytheon Technologies Research Center (RTRC, formerly UTRC), as well as many other outstanding colleagues on three different UTRC-led ARPA-E Projects on flow-battery technologies.