by Yelim Kim
In the last eight months, I have been diving deep into the energy sector for the power grid, looking into the existing technologies and challenges in the space. In this journey, I had many points of confusion and uncertainty, which encouraged me to create a resource for anyone who would be interested in working in or learning more about energy storage. I am not an expert in this space and, of course, still have many gaps of knowledge to fill up. However, I think this article could give you a solid foundation for starting your own journey to build a more in-depth insight into the innovations in this space.
In this article, you will learn:
- What is energy storage?
- What energy storage technologies exist?
- Why is energy storage crucial for decarbonizing the power grid?
- What are the ways in which energy storage is used in the power grid?
- What is currently being worked on in the energy storage field?
Introduction to energy storage
If you are familiar with how energy storage works and the various energy storage technologies that exist, you can skip this section.
What is energy storage?
Energy storage is a technology that can efficiently store energy (usually in the form of electricity) and release (discharge) it when the user needs it.
Several key properties characterize energy storage devices (all of them are important):
- Energy capacity
- Power capacity
- Duration
- Round-trip efficiency
- Cost
- Safety/environmental impact
- Energy density
- Cycling life
- Self-discharge life
- Operating conditions
- Abundance and cost of materials
Energy capacity
The maximum amount of energy that the device can store. It is usually measured in kWh (kilowatt-hour) or MWh (megawatt-hour).
Power capacity
The maximum rate at which the device can discharge energy. We measure power in units of kW (kilowatt). There is an inherent limit to the rate of energy discharge, just like how a waterering can can discharge only so much water at a time, depending on the size of its holes.
Duration
How long the device can continuously discharge energy at maximum power, starting from a fully charged state. The duration can be calculated as [Energy capacity] / [Power capacity]. Realize how the unit of this quotient is the hour: kWh / kW = h (hour).
Round-trip efficiency
The percentage of the initially stored energy that the device discharges back out. This value can also be expressed as [energy dischraged] / [energy inputted]. All energy storage devices lose a part of the energy that they initially store due to side reactions and internal resistance.
Cost
The total cost of an energy storage device depends on many factors. If we’re just talking about the upfront capital cost of building the battery (aside from the costs for installation, interconnection, and operation/maintenance), this mainly consists of the power cost ($/kW) and the energy cost ($/kWh). The power cost results from the cost of the components that determine the power capacity while the energy cost is mainly caused by the cost of the material/substance that physically holds/stores the energy.
Safety/environmental impact
Risks of fire or explosion from short-circuiting; toxicity to the environment/humans when chemicals leak from the device.
Energy Density
The amount of energy per volume (kWh/L) or per mass (kWh/kg) of the device. The energy density influences the geographic footprint, transportability, and mass, as well as the overall capital cost (a larger energy storage device requires more materials, which increases cost).
Cycling life
The number of charging and discharging cycles that the device can undergo before its energy capacity drops below a certain threshold (ex. 80% of its initial capacity). Cycling life determines the number of years that the device can be used, also dependent on how often the device is cycled. Across many charging and discharge cycles, the device undergoes permanent damage/wear that ultimately reduces its energy capacity.
Self-discharge life
The period of time that it takes for a fully-charged device to lose all of its energy when left undisturbed (without being charged or discharged in the middle).
Operating conditions
The set of environmental conditions, including temperature, humidity, and pressure, that the energy storage technology can operate reliably under.
Abundance of materials
The geographical availability and abundance of the materials that make up the bulk of the storage device. This directly affects the cost and scalability of the technology.
Note
An overarching trend for energy storage technologies is that certain properties go against each other: often, improving one characteristic leads to the exacerbation of another conflicting characteristic (ex. low power costs vs low energy costs). Therefore, developing an energy storage system is a matter of carefully balancing its properties to best cater to the specific needs of the application it will be used for.
Energy storage technologies divide into several categories
- Chemical energy storage
- Electrochemical energy storage
- Thermal energy storage
- Mechanical energy storage
Chemical energy storage
How it works: A chemical fuel (usually H2) is fed into a device, called a fuel cell, where it is combusted to discharge energy. The energy is stored in the form of chemical bonds in the H2 molecules. The H2 gas can be stored in a tank or an underground cavern.
Challenges/limitations: The cost of producing H2 from electricity without generating carbon emissions (in a process called electrolysis) is very high, making H2 expensive.
Advantages: H2 has a very high energy density. It is also very versatile and can be used in other industries, such as in transportation and steel/chemical manufacturing.
Electrochemical energy storage (a.k.a. batteries)
How it works: At the most basic level, most electrochemical energy storage technologies generally consist of two chemicals — one that naturally likes to accept electrons (oxidative chemical) and one that naturally likes to give them away (reductive chemical). The electron-receiving chemical makes up a part of the battery called the cathode, and the electron-donating chemical makes up the anode. Energy is discharged when the anode and cathode are linked by an electrical pathway (i.e. via a wire) that allows electrons to spontaneously flow from the anode to the cathode, and thereby release electrical energy. The transfer of electrons from one chemical species to another is called a redox reaction. At the same time, the transfer of electrons is countered by the movement of ions between the anode and cathode to offset the charge imbalance. Energy is stored when voltage is applied to shuttle the electrons and ions back to their original places, allowing the same discharge reaction to happen all over again.
There are multiple types of batteries, each with a different architecture, mechanism, and anode and cathode composition. Here are some of the most prominent battery technologies:
- Lithium-ion batteries —Energy is charged and discharged as electrons and lithium ions shuttle between shelves of graphite layers (anode) and a multi-oxide crystal (cathode). The anode and cathode are separated by a layer of inorganic solvent. Challenges: Lithium and other metals used in Li-ion batteries are not very abundant, which is exacerbated by the heavy use of Li-ion batteries in the electric vehicle industry. Therefore, in the long term, alternatives to Li-ion batteries that use more abundant materials must be developed for the power grid. Advantages: Li-ion batteries are the most advanced battery technology currently standing (maybe alongside lead-acid batteries) and have very good performance aside from their high energy cost. Thus, for now, Li-ion is one of the most promising options for both electric vehicles and the power grid.
- Redox flow batteries (RFBs)—Unlike lithium-ion batteries, where the anode and cathode are both solid, flow batteries consist of reductive and oxidative species dissolved in liquid solutions. Most of these solutions are stored in two large external tanks, while a small portion of each solution is stored in a reactor (a box where the redox reaction happens). When the solutions in the reactor are fully used up (either fully charged or discharged), “new” solutions are pumped into the reactor from the external tanks while the existing solutions are pumped back into those tanks. Challenges: redox flow battery technology is a nascent area and much more development must be done to reach cost competitiveness. Advantages: they are safe (because they are based on water), can tolerate high depths of charge, can be made from cheap chemicals (ex. Iron, Zinc, etc), and allow for power and energy capacity to be scaled independently due to the separation of the tanks from the reactor.
- Metal-air batteries — A metal substance in solid or liquid solution form (ex. Aluminum, Zinc, Iron) acts as the anode while air or oxygen acts as the cathode. The charging and discharging reactions work like any other battery. Challenges: metal-air batteries have significant technical barriers that must be overcome in able to reach reasonable cost and performance. These challenges include the low efficiency of redox reactions involving oxygen, the high cost of catalysts used to accelerate those reactions, the degradation of the battery’s components from reacting with oxygen, and the formation of dendrites (spiky metal structures that deposit on the anode over multiple cycles). Advantages: because the cathode is made from air, they have the potential to reach very high energy density (per mass) and low energy costs (even cheaper than RFBs).
Thermal energy storage (a.k.a. thermal batteries)
How it works: Electricity is used to heat up a material (molten salt, carbon blocks, etc) to a very high temperature. The hot material is enclosed in a heat-insulating case (basically a giant Thermos). The heat can be discharged back into electricity in several ways, including by using the heat to boil water into steam that turns turbines or capturing the electromagnetic radiation (light) emitted from the glowing-hot material using a photovoltaic cell (the latter technology is being developed by Antora Energy)
Challenges/limitations: thermal energy storage is still relatively new, especially with the emergence of new ways of storing and discharging heat in the last few years. Not many thermal battery projects have been implemented so far.
Advantages: Very cheap materials can be used to store the energy, giving it the potential to reach low energy costs. Also, energy can be charged and discharged in the form of both electricity and heat, which is useful in the industry sector that commonly uses heat to directly power its systems.
Mechanical energy storage
How it works: Some form of mechanical force is involved in the process of storing and discharging energy. Mechanical storage systems can vary very widely depending on which mechanical force they are based on, but they share some underlying challenges.
Challenges: Mechanical energy storage systems have a very low energy density and require a large storage infrastructure, which limits where they can be geographically installed. As a result, these systems require high upfront capital costs that can make them high-risk investments for communities/governments. This is also why I personally think electrochemical storage technologies have a higher potential of having a wider impact overall.
- Pumped Hydro Storage: Pumped hydro systems consist of two reservoirs, one located at a higher altitude than the other one. Energy is released when a large body of water falls from the higher reservoir to the lower reservoir. The falling water passes through turbines that convert their kinetic energy into electrical energy. To recharge the system, electricity is used to pump the water back up to the higher reservoir. Advantages: pumped hydro storage has been around for a very long time and has shown promising performance and low energy costs (at least in the long term). Fun fact: 99% of the current energy storage capacity in the world consists of pumped hydro storage. Limitations: pumped hydro storage sites are limited by high capital costs. Although there are enough pumped hydro-suitable geographical sites to fulfill all of the energy storage needs in the US, basically all contracts and plans for new pumped hydro sites in the US from the last few decades have been canceled due to high up-front costs.
- Compressed-air energy storage: Compressed air is stored in a tank or an underground reservoir (like for hydrogen), and when energy needs to be discharged, the air is decompressed and passed through a turbine to generate electricity. Limitations: the infrastructure used for compressing and discharging the air is costly, and finding geographic locations to store the compressed air (particularly for underground storage, which is more economical than above-ground) is difficult, especially since compressed air would likely compete with Hydrogen for storage locations. Advantages: could provide energy for long durations.
Note
There are many other energy storage technologies that I have not mentioned here, as I tried to focus on the technologies that I personally deemed were the most promising and relevant to the power grid. Specifically, for mechanical energy storage, some technologies I did not mention include flywheel — because they are intended for extremely short-duration energy storage (which are not very relevant to the particular applications in the power grid that I discuss in this article) — and gravitational energy storage, which entails very high capital costs, high geographic footprint, and may not be as low cost as pumped hydro (however, I am not completely rejecting this technology; gravitational energy storage may be useful depending on the situation and location of the application).
The role of energy storage in decarbonization
Energy storage devices can be used to address climate change in two major ways: electric vehicles and the power grid.
Energy storage in electric vehicles
Electric vehicles (EVs) require a lightweight, high-density, and high-efficiency battery to supply electrical energy to the vehicle as a replacement for the gas combustion engine used in conventional cars. Right now, lithium-ion batteries are dominating the energy storage industry for electric vehicles, and most of the current research in Li-ion batteries (such as increasing energy density and lowering costs) is focused on this application. EV batteries are crucial for allowing the transportation sector to transition to electricity from renewable sources from fossil fuels.
In this article, however, I will devote the rest of the discussion to the usage of energy storage in the power grid.
Energy storage for decarbonizing the power grid
The power grid is a network of energy generators, transmission and distribution systems, and consumers, in which electricity is transported every second from the places where it is produced to where it is used (homes, buildings, hospitals, factories, etc).
In the last few decades, the implementation of renewable energy systems— particularly, solar photovoltaics (PVs) and wind turbines[*]—in place of traditional fossil fuel generators has been increasing. Several key factors contributed to this change: a significant drop in price of renewables since the 1970s (right now, solar and wind energy are cheaper/competitive with fossil fuels), government policies pushing for the transition to renewable energy, and, of course, a rising public urgency around climate change (the electricity sector accounts for about 25% of global carbon emissions).
[*]There are other types of renewable energy out there, like geothermal and biomass, but solar PV and wind turbines are the only ones that have reached cost competitiveness with fossil fuels, at least in North America and Europe.
Despite this important progress, renewable energy contributes to only 20% of the energy generation in the US. Why aren’t we installing more solar and wind plants? A key challenge in solar and wind energy is that the amount of energy produced varies depending on the weather (this is why solar and wind energy are collectively called variable renewable energy). During some hours, the amount of renewable energy generated might be higher than the customer demand at the moment, and if the excess energy isn’t stored right away, it would be wasted or thrown away. At other times, the renewable energy systems might fail to meet the electricity demands, and without a backup energy supply (i.e. discharging stored energy or firing up gas turbines), this would result in blackouts. When used by itself, variable renewable energy is highly unstable and fails to efficiently meet the demands of the customers.
Energy storage is the key technology that would fill up this gap in renewable energy systems by soaking up excess renewable energy and discharging them when needed. Therefore, energy storage is crucial for ensuring that the electricity demands in the grid are met at all times regardless of the weather, especially when renewable energy makes up a high percentage of our energy supply in the future.
Using energy storage to “time shift” excess energy to times when electricity production doesn’t meet the demand is called energy arbitrage. Energy arbitrage is the largest way for energy storage to enable the decarbonization of the power grid, and I will be focusing on this application for the rest of this article. However, energy storage is also very useful for other purposes in the power grid, particularly in ensuring that a smooth flow of electricity is maintained in the grid in the presence of momentary fluctuations in demand and supply. In this case, the energy storage device outputs electricity in short bursts lasting from seconds to minutes.
Various types of energy arbitrage and the corresponding energy storage technologies
There are several questions we must answer to further narrow down the energy storage applications in the power grid:
- Exactly in what situations is energy storage used for energy arbitrage? (short answer: there are three different energy arbitrage applications)
- What types of energy storage technologies are optimized for each arbitrage application?
Short, medium, and long-duration storage
Remember, the main purpose of energy storage is to supply previously saved power when not enough renewable energy is produced to meet the demand at the moment.
There are different time scales at which energy storage is needed to make up for these “gaps”, and we define each time scale in terms of the minimum duration that the energy storage must be able to discharge for:
- Short-duration: less than 4–6 hours of discharge
- Medium-duration: 6–12 hours of discharge
- Long-duration: 12+ hours, including several days or a whole week without recharging in the middle.
Short and medium duration are collectively called diurnal storage because they’re designed to make up for the daily fluctuations in energy. Likely, short-duration energy storage would be charged and discharged daily, and during times/in geographical locations where energy discharge is needed for more than 6 hours, medium-duration energy storage would be used.
Long-duration energy storage is used for much larger-scale fluctuations in renewable energy. For example, some long-duration storage systems might save up excess solar energy during the summer and discharge when the electricity supply is lower in the winter. Or, some storage systems could serve as backup for emergency situations, such as blackouts from extreme weather events that last for several days. A backup for blackouts is especially crucial for institutions like hospitals, airports, military bases, data centers, etc, where a continuous supply of electricity at all times is paramount. A key thing here to note is that, aside from applications in emergency situations like blackouts, the need for long-duration storage is lower than the need for diurnal storage (medium-duration storage is less necessary than short-duration). Right now, as mentioned above, renewable energy doesn’t make up a high enough proportion of our energy generation system, and therefore our current grid doesn’t require mitigation of long periods of low renewable energy with long-duration energy storage [*]. However, as we reach a higher penetration of renewable energy in our grid system, longer-duration energy storage is needed to replace some of the buffering roles that on-demand gas plants have been previously fulfilling.
[*]In fact, the current low demand for long-duration energy storage is hampering a lot of progress for long-duration storage technologies (see image below).
Classifying energy storage technologies into the three buckets of duration
Each of the short, medium, and long-duration applications is suited for energy storage technologies that prioritize a different set of properties.
One notable difference between the three applications is the relative costs of power and energy capacity. Remember from the beginning of the article that the discharge duration is equal to [energy capacity] / [power capacity]. As a result, devices serving short-duration discharge would have a higher power capacity relative to energy capacity than long-duration applications, for which the opposite would be true (high energy capacity, lower power capacity). Thus, to minimize the total cost of storage, short-duration applications would prefer energy storage technologies with a low power cost, even if that means the energy cost is not very low (as long as it is within a reasonable cost threshold) [*]. On the other hand, long-duration energy storage technologies should first focus on achieving low energy costs, and only then work form there to reduce the power cost. Medium-duration applications would be somewhere in between the requirements of short-duration and long-duration technologies. However, generally, medium-duration energy storage tends to be more similar to short-duration technologies than to long-duration energy storage.
[*] This prioritization is very important because, as mentioned at the beginning of the article, storage technologies with a low energy cost often have a high power cost and vice versa.
There are some other ways that the technical demands of short, long, and medium-duration applications differ.
Short-duration energy storage: high cycle life and high roundtrip efficiency. Short-duration storage devices would be cycled frequently, around once or twice a day, and thus require a high cycle life to last 15–20 years [*]. For the same reason, short-duration technologies should have a high roundtrip efficiency (the efficiency of li-ion batteries is around 85%) to minimize economic losses from the low efficiency.
[*] Large-scale energy storage facilities should plan to last about 20 years, which is the project life for energy generation plants (to maximize the usage of your assets, you wouldn’t want to get a brand-new energy storage facility before the commission period for your renewable energy plant ends). Using this number, here’s a rough calculation for the minimum cycle life for a short-duration energy storage device expected to be cycled once per day: 1 cycle/day * 365 days/year * 20 years = 7300 cycles. (Note: this number could vary greatly depending on the expected commission time and frequency of use)
Long-duration energy storage (LDES): high abundance of materials. LDES systems are designed to be used seasonally or during special occurrences happening multiple times a year, and thus would be cycled much less frequently than short and medium-duration technologies [*]. As a result, long-duration technologies can last the same number of years as diurnal storage even with a lower cycle life. For this same reason, long-duration storage applications are tolerant with lower roundtrip efficiencies (for example, metal-air batteries, a technology suited for long-duration energy storage, have around 40% efficiency). Compared to short-duration storage, LDES is used for applications that use high total amouts of energy and thus should use cheap materials that are available in large amounts to fulfill high energy capacity needs.
[*] Because long-duration energy storage devices are used so infrequently, there are a relatively small number of “opportunities” for these devices to provide economic benefit to the users. As a result, it takes a long time for these cost benefits to pay up for the total capital and operational costs for the energy storage facilities (which I will explain more about later). This and the overall low penetration of renewable energy (and hence low demand for long-duration storage) are some of the major challenges in the long-duration energy storage market.
How do existing technologies fit into these criteria?
Short-duration storage: Lithium-ion batteries (key characteristics: relatively low power cost, high efficiency, high cycle life)
Medium-duration storage: Redox flow batteries (RFB) (key characteristics: can be made from abundant metals/chemicals, good round-trip efficiency, energy cost decreases with higher energy capacity due to economics of scale; yet, higher power cost and energy cost than short-duration and long-duration storage technologies, respectively)
Long-duration storage: Pumped hydro storage, thermal batteries, metal-air batteries
- Pumped hydro storage key characteristics: low energy cost in the long term (main energy-storing material is water), large storage capacity; yet, extremely high upfront capital costs.
- Thermal battery key characteristics: uses abundant, low-cost materials and is inherently suitable for high storage capacity (which leads to low surface area-to-volume ratios and thus minimizes heat dissipation).
- Metal-air batteries: can be made from cheap/abundant materials + half of the energy-storing “chemical” is air, resulting in low energy costs; but high power costs due to expensive components in the battery cell.
What drives innovations in the energy storage space: reaching cost competitiveness
Across all types of arbitrage, economic competitiveness (against the costs of electricity from fossil fuels) is the key factor that would ultimately enable the large-scale deployment of energy storage technologies.
So, how do you measure economic competitiveness? Essentially, we can calculate a value called the levelized cost of storage (LCOS). The LCOS is the total capital and operational cost per kilo/megawatt-hour of an energy storage facility divided by the number of years that it is expected to be used. It’s the minimum amount of “revenue” (money saved by using energy storage) that the facility must generate each year to make up for the money invested to build and use it. LCOS allows you to compare the economic viability of various energy storage technologies with each other and with the costs of other methods for mitigating the variability in renewable energy, such as gas combustion turbines. This LCOS is combined with the LCOE (levelized cost of energy) of renewable energy and compared with the LCOE for conventional gas generation plants [*]. Economic competitiveness is reached when the energy storage + renewable energy combination has a lower cost than the LCOE of gas-based electricity.
[*] Traditional generation plants that operate on gas are dispatchable resources, meaning that the amount of energy being produced in the generator can be directly controlled based on customer demand. Thus, they don’t need energy-balancing devices like energy storage.
Researchers have published multiple papers that break down the economically competitive value for LCOS into a specific cost target for the capital cost of energy storage devices (here is one paper that does this). Generally, this cost target varies by the application that energy storage would be used for. For example, if the energy storage facility is used around the clock to meet the baseload demand, it would be competing directly with low-cost gas generation plants and thus would have low (stringent) cost targets. However, if the energy storage facility is intended for being used for just a few hours a week, a higher energy cost would be tolerable. In the paper I linked above, the authors found that the capital cost of energy storage per unit of energy capacity (not power capacity) must reach about $20/kWh [*] to be economically competitive in a baseload demand application.
[*] Although this number gives a general idea of what the cost target should be, the exact numerical value of this target can vary depending on factors like the geographical location, the projected level of renewable energy penetration, and demand-side management, where the customers intentionally reduce their energy demand at certain times of day when the production is low/demand is too high.
Now, the final question is, how exactly do we reach this $20/kWh goal?
Here’s the short answer: the capital cost of an energy storage technology is dependent on the combined cost of all of its components, but we can point to several of these components that take up the largest portions of the total capital cost and try to reduce their cost while not compromising the overall performance of the technology. If you’d like to learn more about specific advancements that are being made right now to reduce cost and improve performance, look out for my next article, where I will discuss the major hotspots of research happening in the electrochemical energy storage field.
Summary
- An energy storage device must fulfill a set of requirements before it can be deployed in the real world. All of these requirements must be met, but depending on the application, some of these requirements are more stringent/prioritized than others.
- Energy storage is used to smooth out the weather-dependent fluctuation of electricity generation in renewable energy systems. Energy storage is crucial for achieving a highly reliable grid when the vast majority of our energy supply is met by renewable energy in the future.
- Energy storage devices in the power grid can divide into short, medium, and long discharge durations. Short-duration applications prioritize a low power cost and an acceptable energy cost, high roundtrip efficiency, and high cycling life. Long-duration applications prioritize low energy costs and must use highly abundant materials.
- The economic viability of energy storage + renewable energy systems compared to gas-based electricity is the most important determinant of achieving high penetration of renewable energy in the grid. Therefore, the current advancements in the grid battery space are focused on lowering the cost per unit of energy to ultimately meet the capital cost targets.
