Advanced Battery Technologies for Stationary Energy Storage Applications
Inside this Article
At the grid level, lead-acid batteries account for just a fraction of a percentage of new energy storage deployments. So, what battery chemistries are developers and utilities using today, and why?
A radical energy transformation is under way today, one that we will likely fully appreciate only in hindsight. Auto manufacturers are transitioning to electric vehicles, which will enable new transportation paradigms and vehicle-to-grid services. Utility regulators and operators are beginning to rebuild the bulk power system to make it more resilient and better able to accommodate high penetration levels of variable renewable generation. The prime mover in these transitions is the rapid advancement in electrochemical energy storage technologies.
In this article, I briefly review grid applications for energy storage solutions, both in front of and behind the customer meter. I then provide a compendium of advanced energy storage solutions for stationary energy storage applications, looking at representative technologies, vendors and field deployments. Because there is much to cover here, this is less of a design guide for applications engineers than it is a snapshot of a very dynamic and exciting space, one that solar professionals would do well to keep tabs on.
Grid Applications for Energy Storage
It is with good reason that industry stakeholders, researchers and analysts often describe energy storage as the missing piece of the puzzle or the great enabler in the renewable energy revolution. Use cases for energy storage systems exist in front of the customer meter, at both the transmission and distribution level of the bulk power system, as well as behind the customer meter, in commercial and even residential applications. Depending on how and where developers deploy energy storage systems, benefits might accrue to a utility, reliability coordinator, balancing authority, third-party system operator, community, commercial or residential customer, society at large or some combination of these. As shown in Figure 1, a 2015 report published by the Rocky Mountain Institute (see Resources) identifies 13 different services that energy storage systems can provide to three general stakeholder groups.
Use cases for energy storage are generally built around opportunities to avoid incurring costs or opportunities to generate income. Examples of the former include energy storage deployments that allow utilities to defer transmission or distribution system upgrades, or that enable commercial and industrial customers to avoid demand charges. Examples of the latter include energy storage systems that participate in markets for ancillary services, such as frequency regulation, voltage support or demand response. Energy storage systems can also improve grid resiliency, provide generation capacity or facilitate the integration of more wind and solar. Some of these applications have broad societal benefits, such as disaster preparedness or greenhouse gas reductions, that are not easy to quantify in economic terms—at least not given today’s market structures.
Though there are many opportunities to deploy grid-interactive energy storage systems, it is important to recognize that different applications are not created equally. Frequency regulation is a relatively short-duration service, measured in seconds or minutes, intended to reconcile momentary differences in the generation-to-load balance; as such, it favors a fast response time but does not necessarily require large amounts of energy, because the battery is alternately discharging and charging in rapid succession. Applications that shift electric energy in time, over a period of hours or even days, are comparatively more energy intensive but may have more modest power requirements. Other applications are potentially both energy and power intensive. In demand management applications, for example, batteries store off-peak energy for a period of hours, then discharge stored energy during on-peak pricing periods as needed to offset the demands associated with heavy loads.
These different application characteristics underscore the need for different batteries and battery technologies. Some chemistries or technologies are better suited for short-duration power applications, whereas others are better suited for long-duration energy applications. Since deploying an energy-optimized battery in a power application or vice versa can degrade system performance in the long term, some hybrid utility-scale applications actually utilize both power- and energy-type batteries. While it is tempting to describe grid-interactive energy storage systems in general as a Swiss Army knife, no one battery is the ideal tool for all applications.
Advanced Battery Technologies
Here I look specifically at those alternatives to conventional lead-acid energy storage technologies that are commercialized in stationary grid applications or deployed in pilot projects. Though lead-acid batteries have been commercialized for nearly 160 years, they have ceded market share in recent decades to next-generation secondary (rechargeable) battery technologies. This is perhaps most apparent in portable tools and consumer electronics, but is equally true in automotive traction and stationary grid applications. Lead-acid batteries have a long history in off-grid applications, but are seldom the technology of choice in emerging grid-interactive energy storage applications because of limitations associated with round-trip efficiency, energy density, depth of discharge and cycle life.
According to the most recent edition of the US Energy Storage Monitor (see Resources), lithium-ion (Li-ion) batteries are by far the dominant energy storage technology in today’s grid-interactive applications. Continuing a trend that dates back to Q4 2014, Li-ion deployments accounted for 94.2% of the energy storage market in Q2 2017. The leading manufacturers in this space are generally subsidiaries or divisions of well-known multinationals or specialty battery vendors.
Lithium has several characteristics that make it ideal for use in batteries. For one, it is the third chemical element on the periodic table, after hydrogen and helium, making it the lightest of all elemental metals. Additionally, it has the highest electrochemical potential (-3.02 volts) of any metal and is highly reactive. Inherent advantages and disadvantages are associated with these chemical properties. On the one hand, Li-ion batteries provide excellent energy and power density; on the other, they present potential issues with chemical and thermal stability.
All Li-ion batteries have three basic parts: a positive electrode (cathode), a negative electrode (anode) and a chemical compound (electrolyte) that allows the movement of ions between the electrodes. The cathode material is often a metal oxide, and the anode is usually porous graphite. When a Li-ion battery is charging, lithium ions migrate from the cathode to the anode; on discharge, the anode loses electrons and the cathode gains electrodes.
Manufacturers package Li-ion cells individually, often in pouch or cylinder form, and integrate multiple cells into a battery module with a battery management system to keep each cell balanced. To scale systems up, companies integrate multiple battery modules into a battery rack. Container-scale solutions incorporate multiple battery racks. Li-ion energy storage systems at this scale have multiple layers of battery management and control—at module level, rack level and system level—as well as multiple layers of safety and protection such as fuses, software, containment, climate control and so forth.
Li-ion batteries in general have relatively high intrinsic cell voltage and low self-discharge rates, and respond quickly when charging or discharging. Important differences exist between specific Li-ion chemistries, however, as each has different performance characteristics and a unique value proposition. In most cases, the name given to various Li-ion technologies refers to the chemistry of the cathode. Common Li-ion cathode chemistries in grid applications are lithium iron phosphate (LFP) and lithium nickel manganese cobalt oxide (NMC).
LFP-type Li-ion. In 1996, a research team led by John Goodenough at the University of Texas first described lithium ferro phosphate (LiFePO4) as a cathode material for rechargeable Li-ion batteries. Superior chemical and thermal stability is one of the most compelling features of LFP-type Li-ion batteries. Phosphate-based cathode materials release little heat and oxygen gas when exposed to high-temperature or overvoltage conditions, which means they are not susceptible to thermal runaway. In addition to being chemically stable, LFP cells are not combustible, which further improves their safety relative to Li-ion batteries with metal oxide cathodes. LFP batteries also offer an extended cycle life and lifespan compared to competing technologies. The tradeoff is that while LFP batteries can support high load currents and retain their power capabilities at a low state of charge, they have a relatively lower energy density and a higher self-discharge rate.
Notable grid-interactive LFP battery vendors include BYD, a Chinese manufacturer of automobiles and rechargeable batteries, and Murata. LFP-type batteries are in use in a wide variety of stationary applications. At the grid scale, project developers or EPCs have deployed BYD’s LFP-based energy storage systems at multiple locations across the US and Canada in both microgrid and frequency regulation applications. For example, EDF Store & Forecast commissioned a 19.8 MW/7.9 MWh battery storage project outside Chicago in January 2016 that provides the grid operator, PJM, with ancillary services, including autonomous frequency regulation and dynamic power reserves. At the other end of the application spectrum, Blue Planet Energy uses Murata’s LFP-type batteries in its residential energy storage platform to support nanogrid, backup power and self-supply applications.
NMC-type Li-ion. In 2001, Zhonga Lu, a staff scientist for 3M, and Jeff Dahn, a physics and chemistry professor at Dalhousie University, filed international patents for lithium nickel manganese cobalt oxide (LiNiMnCoO2) as a cathode material. As an industrial research chair for the Natural Sciences and Engineering Research Council of Canada, Dahn specializes in materials for advanced batteries. After completing a 20-year research agreement with 3M in 2016, Dahn’s group began a 5-year research partnership with Tesla.
Cobalt is a common ingredient in many Li-ion battery cathodes because it provides a high energy density. The downside is that cobalt is expensive, and the cobalt-based chemistries used for portable electronics have issues with thermal stability and capacity fade. Adding nickel and magnesium to the mix not only reduces costs, because nickel is less expensive than cobalt, but also improves thermal stability and cycle life. Li-ion batteries with NMC-type cathodes are increasingly popular in the market because they offer good all-around performance—in terms of cost, safety and lifespan—and can be tailored for energy or power applications. In its 2017 report, “Status of the Rechargeable Li-ion Battery Industry,” Yole Développement predicts that NMC materials will account for more than half of the global cathode market by 2022.
Notable grid-interactive NMC battery vendors include LG Chem; Kokam, a South Korean company that specializes in rechargeable Li-ion polymer batteries; Panasonic, which is Tesla’s primary business partner in its much-anticipated Gigafactory; and Samsung. Kokam’s Li-ion rack system, which is integral to its containerized energy storage systems, provides a good example of the flexibility of NMC batteries. With two battery racks in parallel, Kokam’s high power–type NMC battery has a usable energy capacity of 211 kWh and a power rating of 888 kW when charging or discharging. By contrast, the same configuration of high energy–type NMC batteries provides 17% more usable energy (253 kWh), but only one-third as much power (266 kW). As a rule of thumb, high power–type NMC batteries are intended for short-duration (<1 hour) applications that require the rapid dispatch of large amounts of power; energy-type NMC batteries are intended for longer-duration (>1 hour) applications with more-continuous loads.
Project developers, utilities, EPCs and system integrators are deploying NMC-type batteries at every level of the electric power system, from grid-scale to customer-sited applications. For example, Tesla deployed a massive 20 MW/80 MWh energy storage system, consisting of roughly 400 Powerpack 2s, at the Southern California Edison (SCE) Mira Loma substation. SCE will use the project to store low-cost off-peak energy for dispatch as a means of reducing its reliance on natural gas peaker plants. At the other end of the size spectrum, Tesla and Green Mountain Power are providing Tesla’s Powerwall 2 to residential customers in Vermont for a low monthly lease or up-front purchase price. Tesla is working with Green Mountain Power to aggregate these distributed residential energy storage resources and bundle them with Powerpack deployments located on utility-owned property for dispatch as a virtual power plant that can provide a variety of grid services.
According to GTM Research, flow batteries accounted for just 5% of the US energy storage market in Q2 2017. As the technology matures, flow batteries could capture more market share in niche applications. While most analysts believe it will be difficult to displace Li-ion batteries, at least over the short term, in applications with a sub–4-hour discharge duration, many think that an opportunity exists for new technologies in applications where discharge duration is on the order of 4–6 hours or more. Examples of applications that favor a long-discharge battery include electric energy time shift or energy arbitrage, transmission and distribution upgrade deferral, and microgrids with variable renewable resources.
Flow batteries are constructed and scale up very differently from Li-ion batteries. Most flow batteries consist of two liquid electrolyte tanks and a cell stack area, like the vanadium-based flow battery in Figure 2. During operation, pumps circulate the electrolyte materials through porous electrodes in the cell stacks, which an ion-specific membrane separates to allow electron exchange between the positively charged catholyte and the negatively charged anolyte. The cell stack structure of this battery type is similar to that of a fuel cell, except that a secondary flow battery can re-energize and reuse the electrolyte.
Generally speaking, flow batteries are categorized as either true redox flow batteries or hybrid redox flow batteries. The term redox is a contraction of reduction and oxidation (see “Energy Storage Glossary”), which describes the ion exchange at the core of the flow battery. In a true redox flow battery, the electrolyte chemicals remain dissolved and in solution at all times; in a hybrid redox flow battery, some of the chemicals that store electrochemical energy are plated as a solid.
One of the unique aspects of the true redox battery configuration is that it effectively decouples the battery’s energy rating from its power rating. The battery stores energy in the electrolyte itself, meaning that its energy rating increases or decreases in relation to tank volume. The power rating, meanwhile, is a function of cell stack area, meaning that it increases or decreases in relation to the number of cell stacks. By varying tank volume and the number of cell stacks, suppliers can tailor the energy and power ratings of a true redox flow battery to application-specific requirements.
While flow batteries do not provide the energy density of Li-ion devices, they have a much longer lifespan and present fewer safety concerns. Because the reactants are in the electrolyte, and the anode and cathode do not really participate in the chemical reaction, charge-discharge cycles do not age the electrodes, and battery capacity does not degrade. Because only a fraction of the electrolyte volume is in the battery cell stack at any one time, the short-circuit potential of a flow battery is negligible, typically posing no danger to equipment or personnel. Flow batteries also do not present a thermal runaway hazard, and the electrolyte is generally not flammable. Some flow batteries do not even require auxiliary cooling systems because the liquid electrolyte itself regulates temperatures inside the battery cell stack.
Though the hardware that makes up a flow battery is capital intensive, the product itself is relatively simple. Instead of having thousands of battery cells and cell-level management systems, the flow battery is more monolithic and has a lower parts count. Of course, moving parts such as pumps need replacing over the life of the system. In addition, even if the electrolyte is designed to last the service life of the battery, periodic electrolyte maintenance is required to reestablish the proper chemical balance and optimal fluid characteristics.
As is the case with Li-ion batteries, different flow battery chemistries have different profiles in terms of performance, toxicity, cost and so forth. Examples of flow battery chemistries deployed in grid applications include all-iron hybrid redox, vanadium redox, zinc-bromine hybrid redox and zinc-iron hybrid redox.
All-iron hybrid redox. In October 2012, Portland, Oregon–based Energy Storage Systems (ESS) received a nearly $3M ARPA-E award to commercialize a 10 kW/80 kWh all-iron hybrid redox flow battery. Part of the value proposition of an iron-based flow battery is that it uses abundant, low-cost materials that are environmentally benign. When charging, ferrous ions plate out as solid iron on the negative electrode; on discharge, the solid iron dissolves and releases two electrons to the positive electrode. The benefit of using the same element on both sides of the battery is that this eliminates degradation issues associated with cross-contamination of the electrolyte materials. An inherent challenge of an all-iron design is finding ways to improve power density. ESS has fielded its all-iron flow batteries in a variety of grid applications, including a 60 kW/225 kWh microgrid demonstration project at Fort Leonard Wood in Missouri.
Vanadium redox. Among the various flow battery chemistries, vanadium redox is the current market leader. The vanadium redox flow battery is a true redox flow device that uses vanadium-based electrolytes on both the positive and negative sides of the battery. During the discharge cycle, vanadium2+ ions oxidize in the negative electrode to form vanadium3+, allowing an electron to migrate to the positive electrode and reduce vanadium5+ ions to vanadium4+ ions, as shown in Figure 2 (p. 27). While vanadium is more expensive than iron, vanadium electrolytes provide a relatively higher cell voltage, which improves power and energy density. The downside of vanadium’s energetic nature is that the sulfuric acid–based electrolytes are corrosive, which exposes battery subcomponents to chemical stresses.
In the late 1980s, the University of New South Wales fielded the first vanadium redox battery. Companies have deployed the technology at scale globally in demonstration projects and in commercial applications for roughly a decade. Active vanadium redox flow battery companies include Sumitomo Electric, a large Japanese conglomerate; Vionx Energy, a Massachusetts-based start-up founded in 2015; and UniEnergy Technologies (UET), a Seattle-based company founded in 2012. In partnership with the US Department of Energy and a number of national laboratories, EPB, a municipally owned utility serving the greater Chattanooga, Tennessee, area, recently deployed a 100 kW/400 kWh vanadium redox flow battery from UET as part of a smart grid demonstration project. EPB will use the battery for renewables integration, voltage regulation, backup power and advanced microgrid operations and energy management.
Zinc-bromine hybrid redox. Exxon developed the zinc-bromine hybrid redox flow battery in the early 1970s. During a charge, zinc is plated as a solid metal on the positive electrode in the battery cell stack; upon discharge, the zinc metal releases two electrons and oxidizes to form zinc2+. This chemistry provides not only a high cell voltage, but also a very high energy density. Since 2012, Australia-based Redflow has deployed zinc-bromine hybrid redox flow batteries in a variety of interactive applications, from the grid scale down to the residential scale. In the US, California-based Primus Power recently released the EnergyPod 2, a second-generation zinc-bromine flow battery that is scalable in 25 kW/125 kWh increments up to 25 MW. The architecture that Primus Power uses is unique in that it features a single tank, a single pump and a single flow loop, eliminating components and costs relative to other flow batteries. The resulting battery has a 5-hour discharge duration and a small footprint.
Zinc-iron hybrid redox. Lockheed Martin pioneered the alkaline-based zinc-iron hybrid redox flow battery in the 1980s. In this battery, the catholyte is food-grade iron salt dissolved in an alkaline solution, and the anolyte is battery-grade zinc oxide suspended in an alkaline solution. During charging, zinc plates out of the anolyte onto the anode. While these electrolytes are caustic, they are nontoxic and do not require any exotic materials. Because the electrolyte is not acid based, the battery does not require complex, corrosion-resistant materials for the subcomponents. The resulting battery is less energy dense than an acid-based hybrid redox flow battery, but is also safer and costs less.
The highest-profile zinc-iron flow battery supplier is Austin, Texas–based ViZn Energy Systems. Founded in 2009, the company started deploying its batteries in pilot projects 6 years later. The company has since developed a suite of containerized solutions for commercial, industrial and utility applications that scale up from 100 kW to more than 100 MW and has delivered solutions to customers in the US, Canada, Central America, Europe and India. In March 2017, the company announced that it was supplying a 200 kW/800 kWh battery to a microgrid project at a luxury resort in Nicaragua that will include diesel backup and an 800 kW solar array.
Though Li-ion and flow batteries account for the majority of the market in stationary applications, other technologies have some track record in the field or are beginning to come to market. Sodium-based chemistries, for example, had some market traction in the emerging grid-scale battery sector before the ascendance of Li-ion. In recent years, zinc-air batteries with aqueous electrolyte have made progress toward commercialization.
Sodium-sulfur batteries. The sodium-sulfur (NaS) battery uses molten sodium as the negative electrode and molten sulfur as the cathode. During discharge, the sodium donates an electron. Advantages of the NaS battery include high energy density, excellent cycle life, affordable materials, high efficiencies and low self-discharge. Disadvantages are that NaS batteries require high internal temperatures to keep electrolytes in a molten state and are not well suited for power applications. This technology is best for applications that require a long (>6 hour) discharge duration.
Tokyo-based NGK Insulators is a century-old Japanese ceramics company known for its insulators and NaS batteries. The company began developing molten-salt NaS batteries in 1984 and successfully commercialized large-scale products by 2002. NGK has deployed its NaS batteries at nearly 200 locations globally to provide a cumulative installation base of 530 MW/3,700 MWh for load leveling, renewables integration, transmission and distribution network management, and microgrid and ancillary services.
Zinc-air batteries. Nonrechargeable zinc-air batteries have a long commercial history, dating back to the 1930s. In the 1970s, companies started building small button-type zinc-air batteries to power hearing aids and other medical devices. Because atmospheric air serves as one of the reactants, these batteries offer excellent performance in terms of energy density and specific energy. Miro Zoric, a Slovenian inventor, produced the first rechargeable zinc-air battery in 1996 and began mass production in Singapore for traction applications the following year. Within the last 5 years, rechargeable zinc-air batteries have begun to make inroads in stationary grid applications. While zinc-air batteries cannot match Li-ion batteries for power delivery, they have the potential to be cost competitive and may be able to match the lifespan of flow batteries.
The companies pioneering grid applications for zinc-air batteries are generally small start-ups. For example, Eos, an Upstate New York–based supplier founded in 2008, has developed a 1 MW/4 MWh zinc-air battery system for grid applications. The Eos Aurora 1000|4000 uses a zinc-hybrid cathode battery with an aqueous electrolyte. The company designed the product to provide 5,000 100% discharge cycles, which equates to a 15-year calendar life, and claims that it can undercut the per-kWh cost of Li-ion by as much as 50%. Scottsdale, Arizona–based Fluidic Energy, meanwhile, has been deploying commercial-scale zinc-air batteries in long-duration applications in the developing world as an alternative to lead-acid batteries. Bloomberg New Energy Finance selected Fluidic Energy as one of its ten 2017 New Energy Pioneers.
The parallels between the nascent energy storage market and the early days of the solar market are striking. Promising and potentially disruptive technologies are moving from research and development to pilot projects and commercial applications. Venture capital–backed start-ups championing new technologies are going toe-to-toe with deep-pocketed multinational corporations heavily invested in somewhat more proven technologies. In spite of steep cost declines, viable business cases are few and far between. Though fielding projects is capital intensive, it is difficult to prove bankability and find financing. A few forward-looking states and utilities are developing and implementing goals, policies and incentives intended to jump-start the market. California and a handful of other states with high energy prices are leading the way in terms of field deployments.
For solar industry veterans, it is déjà vu all over again. The energy storage industry is changing so quickly that technologies and vendors in the ascendance last year could be out of the game next year. As was true in the Wild West days of solar, some pioneers of storage will take arrows and many will fail in order for a few to succeed. The stakes are high, however, as those settlers who succeed in staking a claim will transform the energy industry for decades to come.
David Brearley / SolarPro / Ashland, OR / solarprofessional.com
Fitzgerald, Garret, et al, The Economics of Battery Energy Storage: How Multi-Use, Customer-Sited Batteries Deliver the Most Services and Value to Customers and the Grid, Rocky Mountain Institute, October 2015
GTM Research and the Energy Storage Association, US Energy Storage Monitor: Q3 2017, September 2017