Alternate Forms of Electric Storage
For six chapters now, we’ve been focusing on electric batteries. The decision to focus most acutely on energy storage through electrochemical reactions was deliberate—for the general public and even small business and industry, we believe that the electric battery offers great promise in both stationary and mobile applications.
That said, there is still an important role to be played by other energy storage technologies in the deployment of low-carbon technologies, both now and in the future. Some of these alternatives to battery storage are existing technologies (like pumped hydropower), some are relatively familiar (like flywheels), and some are longer-shot technologies on the early end of the R&D cycle.
In this chapter, we expand on our otherwise focused consideration of the electric battery to consider some of these alternative storage technologies and what they mean for the future of the electric battery.
For the most part, the alternative forms of energy storage you’ll discover in this chapter are designed for the grid level. The most promising potential for alternative storage is found in larger projects—for grid support and even bulk power management.
Main Applications of Alternative Storage: The Bulk Power System
Beyond batteries, there exists a broad portfolio of technologies that can be deployed as energy storage systems. Some might sound familiar; others may sound like an invention of a science fiction writer. But by the end of this chapter, you’ll know how they all work and how they could best serve the grid and end users.
Because each technology has different characteristics—rooted in the physics and the chemistry of the technology itself—some are better suited to particular applications than others. (This should sound familiar, as the distinct characteristics of different battery technologies make certain batteries better for distributed storage and others better for electric vehicles or for grid storage.) The power and energy relationship of each determines how it best serves the grid or the end user.1 Ultimately, the best use for a particular storage technology boils down to how quickly the stored energy can be discharged. As it happens, those with the highest power ranges (or greatest megawatt capacities) tend to deliver their energy the slowest, making them better at longer-term bulk storage and less apt to be called upon for immediate grid support.
Of these dozens of possible end-use applications, for the sake of simplicity, we’ll group them into a couple of larger buckets: “energy and capacity storage” opportunities and a variety of grid support and balancing opportunities called “ancillary services.” As discussed previously, these services can play a crucial role in bringing intermittent renewable energy sources onto the grid.
Energy and Capacity Storage Opportunities
When there’s a large volume of energy available at a time when it isn’t immediately needed, bulk energy storage is the answer. Traditionally, these systems have been characterized by high energy ranges (measured in megawatt hours) and high power ranges (measured in megawatts). Bulk energy systems routinely have energy capacities of 1,000 or more megawatts and can deliver anywhere from 50 megawatts to 3,000 megawatts of power, depending on the setup.2 As discussed in Chapter 6, in today’s competitive electric markets, these resources largely provide energy and capacity to the centralized market.
Bulk energy storage is by far the most mature field of energy storage services, having been commercialized and deployed for decades now. Pumped hydropower systems, for instance, are widespread in the United States and around the world.3 We describe pumped hydro in detail later in the chapter, but it involves using electricity that isn’t needed at a certain time to pump water from a reservoir up to another reservoir at a higher elevation. Then, the system acts like any conventional hydroelectric dam. When electricity is needed, water is released from the upper pool and it runs down through a turbine, generating electricity.
By various measures, pumped hydro is the most effective and most utilized form of energy storage in the world. Globally, over 95 percent of the total power capacity of all energy storage projects (including batteries) is found in 344 pumped hydro projects.4
Because it can deliver so much power and hold so much energy, bulk energy storage like pumped hydro is incredibly useful in integrating renewables like wind power to the grid. Consider this hypothetical scenario: a wind farm in the Great Plains produces the most energy overnight when demand for electricity is the lowest. This particular wind farm, however, has partnered with a pumped hydro project, and the wind power is used to pump water uphill all night long. When customers start turning on computers and air conditioners during the day, the pumped hydro managers uncork that upper reservoir, sending water cascading down through the turbine, producing electricity.
While pumped hydro is the most widely used form of bulk storage, other technologies, such as compressed air energy storage (CAES), are showing great promise. We’ll take a much closer look at both technologies later in the chapter.
Ancillary Services Storage Opportunities
If bulk energy storage seems straightforward, grid support and balancing is a rather unwieldy catchall for a number of complex grid-based energy storage services that we defined as ancillary services in Chapter 6. To oversimplify things, grid support here refers to the ways in which energy storage can help serve the various operational needs of the grid, from balancing continuously shifting supplies and demands to maintaining voltage levels.
Some alternative storage systems, such as flywheels, supercapacitors, and superconducting magnetic energy storage (SMES), have been commercialized to provide ancillary services for the grid and are showing enough promise that we will examine each in this chapter.
Energy storage for these types of end-use applications typically has lower power ranges and often has faster discharge times. There’s power as soon as you need it. We’ll take a closer look at the flywheels and supercapacitors—and a unique ice energy storage system—that some industries and businesses are trying out.
Energy Storage Systems: Beyond the Battery
Now that we’ve had a bit of an overview of the services that these energy storage systems can provide, let’s take a closer look at the systems themselves. We’re going to start with the biggest—the bulk storage systems that can deliver a lot of power, but aren’t as quick and nimble. Then we’ll work our way down to the least powerful (but no less important!) technologies, which can discharge faster and be used for more immediate grid support and power quality needs.
Much of this book explores the cutting edge of energy technology, with innovations in battery and alternative storage technology rolling out at a breakneck pace.
Pumped hydro isn’t one of these innovative technologies. Rather, this relatively simple system for storing energy was pioneered over a century ago, and—like the hydroelectric dams it is derived from—hasn’t evolved all that much in the decades since.
In 1929, Connecticut Light & Power (CL&P) debuted the country’s first pumped hydro system in the Rocky River plant.5 (Both Italy and Switzerland claim to have birthed the world’s first system in the 1890s.6) The utility operated a couple of dams on the Housatonic River, but struggled to generate enough power when river levels were low. So, when there was more than enough water, CL&P would turn on a pair of 8,100-horsepower pumps to carry water about 230 feet uphill, where it would be stored in Candlewood Lake. (For some perspective on the size of these pumps, a standard basement sump pump offers just 1/3 horsepower7; the “world’s largest flood pumps” in New Orleans run on 5,000-horsepower engines.8) During peak loads, water would be released from the lake, funneling down a 1,000-foot penstock and through turbines, generating electricity.
Popular Science magazine called it at the time, “A Ten-Mile Storage Battery,” and the magazine’s enthusiastic description written more than 85 years ago still holds up as a solid explanation of the technology today.
“How to store reserve power for daily peak loads and seasonal shortages of water was a problem CL&P solved by erecting a unique plant near New Milford, Connecticut—a sort of gigantic electric storage battery. By pumping water uphill and then letting it flow down again through a water turbine and generator, this power station can store more electricity than all the storage batteries of all the automobiles in the United States put together.”9
In the decades since the Rocky River plant came online, more than 22 gigawatts of pumped hydro plants have been built in the United States. During that time, turbines and generators have gotten more efficient, but the general physics of the technology remain the same. Hundreds of pumped hydro systems around the world all use energy to move water to higher elevations where it can, when needed, be released.
For practical purposes, pumped hydro’s sweet spot is in bulk energy shifting for utilities and grid operators, as a utility can use low-cost, off-peak electricity to lift the water and then turn that potential energy into power at a moment’s notice. While the systems take hours to recharge, once that upper pool is loaded, electricity can be generated on demand within minutes.
The majority of pumped hydro projects in operation today are “open-loop” systems, meaning there is a free-flowing natural water source for either the upper or the lower reservoir (see Figure 7.1). (The Rocky River plant was such a system.) Some developers are now building “closed-loop” systems that are completely self-contained and isolated from free-flowing rivers.10 Closed-loop systems are said to have fewer environmental impacts because they eliminate threats to the well-being of fish and aquatic species and don’t cause problems like sediment migration in free-flowing rivers.
Today, the 344 pumped hydro projects worldwide represent a whopping 95 percent of energy storage capacity globally. Systems boast a round-trip efficiency of between 76 and 85 percent, which is in the upper bounds of what most energy storage systems can offer.11
Figure 7.1 The Seneca pumped hydro storage reservoir above the Allegheny River in Pennsylvania holds 2.1 billion gallons of water and covers 100 acres. (From U.S. Army Corps of Engineers.)
In recent decades, construction of new pumped hydro in the United States has slowed considerably “due to the scarcity of further cost-effective and environmentally acceptable sites in the U.S,” according to the Department of Energy.12 Globally, however, the International Energy Agency’s 2012 Technology Roadmap: Hydropower, predicts that by 2050, “Pumped storage hydropower capacities would be multiplied by a factor of 3 to 5.”13
Japan leads the world in pumped storage plants, with total capacity “capable of absorbing and discharging 26 gigawatts of power.”14 And while pumped storage capacity has flattened in the United States and Europe, China’s capacity has more than doubled since 2008.15 Though pumped storage is now being considered as a way to deal with intermittent renewables that can produce power when demand is low, development of the technology was historically driven by nuclear power’s inflexibility. In contrast to intermittent renewables, nuclear power is best suited for producing a constant level of electricity around the clock. Pumped storage can store excess nuclear power produced during low-demand evening hours and utilize it to meet peak-demand periods.16
In addition to being the world leader in capacity, Japan is also a leader in regard to the actual technology of pumped hydro storage. Japanese utilities have chosen to install variable-speed pumps on their facilities, allowing them to “adjust a plant’s charging and discharging to simultaneously balance power supply and demand,” regulating the grid’s frequency.17 While this technology costs more, it reduces the need to inefficiently ramp up and down Japan’s oil-fired generators.18 In post-Fukushima Japan, these facilities increasingly help balance the 10 GW of renewables that have come online since the March 2011 catastrophe.19
For all its potential, pumped hydro storage is facing some new economic risks, ironically caused by the continued development of competitive energy markets that were once seen to benefit pumped storage hydro. The rapid growth of solar in these competitive markets in places like Europe and California is softening the midday demand peaks, thereby reducing the premium payment that pumped storage hydro facilities have relied on for economic viability.20
Case Study: California’s Bison Peak Projects
About 100 miles north of Los Angeles, the Tehachapi Mountains form a stark northern boundary to the Antelope Valley. The valley and the mountains to its north have long been on the forefront of American renewable energy. In the early 1980s, the Tehachapi Pass Wind Farm became one of the first large-scale wind projects in the country. Today, the country’s largest wind farm—the Alta Wind Energy Center—is online in the Tehachapi, supplying more than 1,500 megawatts of renewable energy to California customers.21
Just south in the valley, the Solar Star project is, as of 2015, the largest photovoltaic power plant in the world, with 579 megawatts plugged into the California ISO grid.22 The Antelope Valley as a whole hosts nearly 1,200 megawatts of solar capacity, with more projects in development.
And thanks to the auspicious physical geography of the Tehachapi Mountains, the area is well suited for pumped hydro storage.
The relatively flat mountaintops are surrounded by ravines, dropping down nearly half of a mile vertically to the valley below. It is here that Alton Energy is hoping to build two massive pumped hydro storage projects, called Bison Peak 1 and 2. If built, each of the closed-loop Bison Peak projects could deliver 1,000 megawatts of power for up to nine hours.
According to Alton Energy’s application to FERC for a preliminary permit for the Bison Peak Pumped Storage (or Bison Peak 1), the upper reservoir would be over 45 acres in surface area, sitting at an elevation of 7,800 feet. This “liquid battery” would be contained by a ring dam over one mile in diameter.23 Three alternatives are offered for the location of the lower reservoir, each one between 2,000 and 3,000 vertical feet below the upper pool. Because vertical drop is the most important physical factor in pumped hydro, project consultant Matthew Shapiro has described the topography as “world class.”24
Compared to some other proposed pumped hydro facilities, “The greater elevation differential would allow Bison Peak to supply more power faster with less water,” Alton Energy president Ed Duggan told Utility Dive.25
All told, the Tehachapi region hosts roughly 8,000 megawatts of intermittent solar and wind generation. If stored by pumped hydro, this power could serve the nearby LA Basin load center or could ride the high-voltage transmission Path 26 up to Pacific Gas and Electric’s territory in and around San Francisco. Alton Energy is hoping to bring the Bison Peak Pumped Storage project online by 2019.
Case Study: Iowa Hill Pumped Storage Project
In 2001, directors at the Sacramento Municipal Utility District (SMUD) started evaluating a proposal to build a 400-megawatt pumped storage hydroelectric plant along the Upper American River in California’s El Dorado County. The project gained momentum and by 2010 official plans were drawn up, but in February 2016, SMUD announced that it had canceled plans to proceed with construction.
Though rising costs were given as the main reason the project was scrapped—cost estimates increased from $800 million in 2010 to roughly $1.45 billion—SMUD directors also pointed to a few other important factors. First, increasing amounts of solar power in the utility district were knocking down the daytime load peaks, providing less incentive for stored hydro, and SMUD noted that the industry was moving away from large central station power plants. Other technologies such as batteries were becoming more cost effective.26
What if you took all of the benefits of pumped hydro, but cut out the physical constraints of large reservoirs? Could water be replaced by railcars and the penstock and turbines replaced by railroad tracks and generators?
One company, called Advanced Rail Energy Storage (ARES), is working to prove that a specialized rail system can take the basic physics—potential energy—of pumped hydro and make it easier to site, and more economical.
ARES has built a one-quarter scaled demonstration project of this sort of rail storage in Tehachapi, California. There, on 800 feet of track, railcars loaded with four-ton blocks of concrete are pushed uphill with off-peak electricity. Electric motors within each car turn the wheels. When power is needed, a signal cuts the brakes, and gravity goes to work. As the cars roll downhill, regenerative braking produces power (see Figure 7.2).27
At the demo project, ARES claims a round-trip efficiency of 78.3 percent, which puts rail energy storage among the most efficient energy storage systems out there, including pumped hydro.28 Up to 3,000 megawatts could be delivered by a full-scale commercial system, and with a quick (for bulk storage) response time of 25 seconds to full discharge, rail storage could provide a lot of flexibility for grid support beyond broad supply shifting.29
The first commercial project, a 50-megawatt facility, is planned for the hills west of Las Vegas, where the system could help buffer the grid from the heavy winds and doldrums and sun and clouds that impact the local wind and solar generators. The ARES Nevada project received critical approvals from the state and the Bureau of Land Management in early 2016.30
Though 50 megawatts is large for battery storage technology, ARES wants eventually to develop much bigger projects that take advantage of economies of scale. “We are more efficient as we get larger,” ARES CEO James Kelly told Utility Dive, explaining that the rail storage technology could scale up to 1-gigawatt-sized projects.31
Analysts at Lux Research estimated that the ARES Nevada project—slated to cost about $55 million—would price out at roughly $4,400 per kilowatt-hour and $1,000 per kilowatt, which would be cheaper than an equivalent bank of batteries per unit of energy storage (kilowatt-hour), but more expensive per unit of power (kilowatt).32
Figure 7.2 A one-quarter scale demo of ARES rail energy storage in California. (Courtesy of AES Energy Storage.)
To achieve economic viability, ARES is forgoing price arbitrage strategies (buying electricity cheaply and selling it for higher costs) and instead will be bidding into the California ISO’s (CAISO) regulation market. To balance the California grid’s supply and demand, CAISO, the grid operator, sends new price signals out to regulation service providers every four seconds, and ARES railcar batteries could be charging up the hill or discharging down the slope based on these prices.
The ARES Nevada project should answer a lot of questions about the economic viability of energy storage by rail. In theory, rail storage has much greater geographic potential than pumped hydro, which requires a very specific physical landscape, but it remains to be seen if the capital costs can be earned back quickly enough on the ancillary services markets and if these larger-scale projects will be able to compete with batteries as their prices continue to drop.
As we mentioned, pumped hydro installations have slowed in the United States, largely due to siting difficulties. It’s not the easiest task to find the right physical geography with a ready-made depression for a big, elevated reservoir. Some engineers and entrepreneurs are hoping that compressed air energy storage (CAES) can replicate pumped hydro’s early success.
Here’s how it works. A CAES system uses off-peak electricity to compress air and pump it into underground storage caverns. When the power is needed later, the air is released, spinning a generator and sending electricity into the grid (see Figure 7.3).
Like pumped hydro, compressed air systems can offer bulk energy storage that can ease the integration of variable renewable resources. And like pumped hydro, geography is one of the biggest limiting factors to its deployment. Underground salt caverns aren’t that hard to find in North America, but finding one that’s the right size and near a major transmission line—or even better, near a wind farm or large solar plant—is a much tougher task.
Figure 7.3 A diagram of a compressed air energy storage system. Off-peak electricity is used to compress and pump air into underground storage caverns. (From the U.S. Department of Energy, Pacific Northwest National Laboratory.)
For these reasons, and because CAES has some considerable upfront costs, CAES systems are not nearly as commercially mature as pumped hydro. In fact, there are only three systems in operation in the world. The world’s first compress air system went online in Germany in 1978, with the capacity to supply 290 megawatts of power for four hours. In 1991, the McIntosh facility in Alabama became the first project in the United States, offering 26 hours’ worth of 110 megawatts.33
Another drawback of these first-generation compressed air systems is that on top of the electricity they use to compress the air, they burn natural gas on site to heat the pressurized air before expansion.
For more than 20 years, the projects in Germany and Alabama were the only two CAES systems in the world, but over the past decade there has been a flurry of interest and investment in the technology.34 Still, only one more commercial system has come online, the Gaines, Texas Dispatchable Wind Project. The Gaines project doesn’t use any natural gas and is directly tied to a wind turbine, essentially capturing the energy as it is generated and allowing operators to release it onto the grid on demand. The Gaines project, operated by General Compression, is a modest 2-megawatt project and might foretell a future for compressed air storage that serves less bulk energy storage and more grid support and reliability services.35
A number of pilot projects and demonstrations are currently underway, many of which are addressing CAES’s early limitations. Companies like SustainX and LightSail are working with smaller, more modular systems, storing the compressed air in pipes and steel tanks, respectively.
Case Study: LightSail Distributed Compressed Air Systems
At the age of 20, Danielle Fong left a PhD program at Princeton for Silicon Valley, where she set out to take compressed air energy storage technology out from underground and into the mainstream. Fong founded LightSail Energy with the goal of developing modular-style tanks that could hold up to the intense pressures that CAES demands. Eight years later, LightSail has raised $70 million from Bill Gates, Khosla Ventures, and venture capitalist Peter Thiel, and Fong claims that the company is on a “clear path” to making an economical, commercially viable product.36
Here’s how Fong described the technology to PBS NewsHour: “By the time it gets to the end, it’s at 200 atmospheres. So that’s, in units of pressure, 3,000 pounds per square inch. It’s really a lot. When you want to get the energy back, you have a valve open. As the piston is drawing back, the valve closes, and then the air expands, and it drives the piston, which drives the crankshaft, which drives a generator, which produces AC power.”37
LightSail currently has a half-megawatt prototype that pumps air into a carbon-fiber tank. Because temperature rises considerably when air is pressurized, heat is the biggest obstacle for any compressed air storage system, and LightSail has innovated the concept of injecting a cool water mist as the air is compressed. The heat is captured by the water, and then when the air is released to generate power, warm water is sprayed back into the chamber, converting it back to mechanical energy.38
LightSail’s ultimate goal is to provide a modular storage solution that can be integrated to the grid where needed, and which doesn’t degrade or suffer the efficiency losses that batteries do. In theory, these LightSail tanks—roughly the size of shipping containers—could be stacked and clustered to allow renewable energy producers, industries, or utilities to store excess energy when supply is high and demand is low.
“Replacing batteries with the proven simplicity and durability of engines could bring energy storage costs within reach for mass deployment in support of intermittent wind and solar power,” Fong told Greentech Media.39
Superconducting Magnetic Energy Storage
Batteries store energy with chemical reactions. Pumped hydro stores energy by converting electricity to potential energy, and superconducting magnetic energy storage systems store energy in a magnetic field.
Three simple principles make this technology possible. First, certain materials can convey electric current with no loss. Second, an electric current produces magnetic fields. Third, magnetic fields are actually a type of pure energy that can be stored.40
With that in mind, the heart of a superconducting magnetic energy storage (SMES) system is the superconducting coil (or superconductor). Current will flow in a superconductor even after the voltage is cut off. So once a current is circulated on that coil, the magnetic field is formed, and it can store that energy indefinitely, with no losses, so long as the coil is kept cold enough. In fact, the only energy losses at all in a SMES system come from the quick transformation through an inverter, and the energy requirements of keeping the coil cool. Round-trip efficiencies are clocked at around 95 percent.41
While SMES technology was first developed with bulk energy markets in mind, over the decades researchers realized that the biggest benefit of these systems was their rapid discharge. Because they can discharge energy in less than a second, SMES systems are currently being used strictly for ancillary services such as regulation.42
In theory, this all works great. In practice, however, the costs of the very few materials, all of which are rare, that work as a superconducting coil are extremely high.
Today, the few SMES systems installed are all micro-SMES systems mostly being used by industrial customers for maintaining power quality. Because a relatively small coil and a relatively small cooling system are all that’s needed for these power quality applications, the costs aren’t prohibitive and, actually, for power quality applications that require lower wattage, SMES ranks as one of the cheaper storage systems.43
Champions of the technology are hoping that larger systems for grid stability will become commercially viable, but, so far, all grid-scale projects are in the prototype and demonstration phases.
Not to be confused with Doc Brown’s famous flux capacitor in the film Back to the Future, supercapacitors are real and don’t require “1.21 gigawatts of power.” Supercapacitors (also often called ultracapacitors) are often contrasted with batteries because they are similarly small and modular. Whereas batteries store energy in a chemical reaction, however, supercapacitors (just like their weaker cousins, capacitors) store energy in an actual electrical field. This means that they can charge and discharge almost instantaneously. Because there’s less wear and tear through chemical reactions, they can also endure tens or hundreds of thousands of charge/discharge cycles before degrading, an endurance unmatched by any battery.44 They are also quite efficient, operating at above 95 percent efficiency.
So why are we always talking about (and writing books about) batteries, and supercapacitors are still on the margins? One big factor is that supercapacitors are more expensive per energy unit (kilowatts) than batteries. Another issue is that while supercapacitors can discharge rapidly, they can’t hold a lot of energy in storage.45
As Joel Schindall, a researcher at MIT’s Laboratory for Electromagnetic and Electronic Systems, described to GigaOm: “the ultracapacitor is like a small bucket with a big spout. Water can flow in or out very fast, but there’s not very much of it.”46
Thus, supercapacitors are great when you need a quick burst of power for a short amount of time. The biggest potential for the technology is in electric vehicles, where supercapacitors are already being used to help start and accelerate vehicles before batteries take over.
There’s also strong potential for supercapacitors to help solar and wind “play nice with the rest of the grid,” in the words of Shaw Lynds of Maxwell Technologies, a major developer of the technology.47 Maxwell’s storage systems are optimized to provide up to six or seven minutes of storage capacity, which may not sound like a lot on the face of it, but is more than enough time to help a solar array better connect with the grid.
Consider this: when clouds suddenly pass over an array, the solar power output is thrown into flux. Armed with a bank of supercapacitors and the proper inverters and connections, however, that array can level out the power and deliver a constant, steady stream of electrons to the grid.
In a grant-funded pilot project on the University of California-San Diego’s microgrid, Maxwell’s supercapacitors are teamed up with a photovoltaic solar array to smooth out those kinds of fluctuations to better serve the needs of the grid operators.
The flywheel is a simple technology, and actually one of humanity’s oldest ones. For thousands of years, potters have been sitting in front of flywheels, their feet pumping a pedal that spins the potter’s wheel. Today, some decidedly more refined and efficient flywheels are acting as mechanical batteries, relying on the most basic properties of kinetic energy to store energy.
Flywheel energy storage systems involve spinning a mass, called a rotor, on some kind of near frictionless magnetic bearing. The energy used to get the rotor spinning is converted to kinetic energy, and when power is needed, the flywheel system turns it back into AC power through a generator. The bigger the mass and the faster it spins, the more energy can be stored. Greater storage loads can be achieved by banking together a number of flywheels.
For decades, flywheels for electric storage were made of steel and topped out at just 10,000 rotations per minute. The modern flywheels that make up the innovative storage systems of today are made of mostly carbon fiber and operate in air-tight vacuums to minimize drag on the rotating masses. The most efficient flywheels today are topping 100,000 rotations per minute. A well-designed system can last longer than batteries and, because of advanced ball bearings and centuries of technological refinement, typically require little to no maintenance.48
Another big benefit is the overall environmental impact. Most of the energy storage systems we cover have some troubling environmental or land use issues, but flywheel systems don’t use any hazardous materials, don’t require any significant swaths of land, and produce no emissions. The manufacturing of the carbon fiber and composite components certainly have some resource and climate impacts, but the life cycle of a flywheel module is long enough that those impacts over time are minimal.
Industries have been using flywheels for decades now, both for mechanical efficiencies in their manufacturing processes and—more germane to our focus here—as reliable power backup and for power quality control. Research facilities use them. NASA wants to put them on spacecraft.49 And increasingly, they are selling their services into wholesale ancillary services markets such as regulation.
Like all of these energy storage systems, a flywheel project can easily capture energy from intermittent sources or during off-peak times and then inject a consistent supply of power to the grid when it’s needed. Because they tend to be power rich but energy poor—meaning they can deliver a lot of energy fast, but not for long—flywheel systems are best suited for providing ancillary services to the grid. Nobody is using flywheels to store megawatt hours for weeks on end, but they can provide valuable storage potential for leveling out intermittent supplies.
Today, there are 17 flywheel storage systems on line in the United States and at least 30 around the world.50 Beacon Power operates a couple of 20-megawatt plants in upstate New York and Pennsylvania, both optimized for providing frequency regulation, to the New York Independent System Operator (NYISO) and PJM markets, respectively.
To trace the struggle and resurgence of flywheels as a commercially viable energy storage technology, one can simply follow the story of Beacon Power. For a decade, Beacon worked to develop a modest 3-megawatt plant and, having proved the concept, the company set out to construct a much larger 20-megawatt plant for grid regulation in Stephenstown, New York, near Albany. In 2010, the company received a $43 million loan from the Department of Energy, built the plant in a year, and then promptly went bankrupt.51
The Stephenstown plant worked, effectively keeping the frequency of the grid stable as intermittent loads from upstate wind farms caused sudden changes, but the company wasn’t earning enough revenue from the NYISO regulation market to fund the operation.
Then in 2011, just as the company was filing for bankruptcy, FERC, as discussed in Chapter Six, through Order 755, required regional grid operators to implement new rules (called “Pay for Performance” regulations) that better compensate fast-performing grid regulating projects like Beacon’s flywheels for the services they provide.52
A private equity company purchased Beacon Power, put the Stephenstown plant back online, and built another plant of the same size and capacity in Hazle Township, Pennsylvania.
The 200 flywheels in each of Beacon’s plants now get paid a premium for almost instantly correcting the frequency deviations on the local grid.
Beacon Power’s CEO Barry Brits claims that the company is currently developing two more plants of the same size, and because of lessons learned and dropping prices, they will cost less than half of the Pennsylvania plant to build.
The company emphasizes the flexibility of flywheel systems for serving different needs. “Like building blocks, single flywheel modules fit together with others to build a flywheel energy storage system of any size from 100 kW to multi-MW power plants. The modular configuration minimizes site footprint and enables owners to place the exact amount of stabilizing resource in the exact location needed. The layout of the modules can be configured to maximize use of space. Depending on the specific site, more than 20 MW can be installed per acre. Each module in a flywheel energy storage system is designed to function on a fully independent basis which results in high plant availability and optimizes performance.”53
In fact, Beacon Power is working with the Chugach Electric Association, an electric cooperative in Alaska, to provide a modular 160-kilowatt unit to be integrated with batteries into a hybrid storage system for the utility.54
When you’re pouring a hot cup of coffee out of a thermos or twisting an ice tray to drop a couple of cubes into your glass of water, you probably aren’t thinking about energy. But, in a way, that thermos and that ice cube tray are thermal batteries, storing energy in the form of heat or cold.
Thermal energy storage systems take this concept and put it to productive use. Broadly speaking, thermal energy storage is a technology that stows away thermal energy by heating or cooling something, and then using that stored energy at a later time for heating or cooling buildings, or for actual power generation.
A good quality thermos can store energy as heat for a whole day. A cooler can do the same with cold. A well-designed thermal energy storage system can store heat generated from electricity during the day and use it at night. Or, conversely, it could store energy by chilling cold water or ice when power is cheap or plentiful, and that cold could be put to use when the air conditioner is in demand. On a larger scale, industrial facilities or utilities could use off-peak power to heat up a good storage medium—like water, sand, molten salt, or rocks—and then release the heat to generate steam, spin a turbine, and produce power to be used when needed.
As you may gather, thermal energy storage is a broad category with a number of possible applications at different scales and for different uses. Thermal storage can be used on site at power-generating facilities like concentrated solar farms, or it could be used in a small stand-alone home or business.
Because we’re focusing on electricity storage in this book, we’re going to limit this brief discussion of thermal energy storage to applications that have both direct or closely indirect ties to electricity and the grid. We won’t delve into passive solar storage (systems that rely on solar thermal energy and then release that energy as heat over time) or any other system that doesn’t involve electric systems.
Pumped Heat Electricity Storage
The concept of pumped heat electricity storage (PHES) is rooted in the simple heat pump: a system like those in an air conditioner or refrigerator that transfers heat from one place to another.
A PHES system needs two containers—steel tanks or silos will do—each filled with a fine crushed rock-like gravel. The containers are connected by pipes filled with an inert gas. Argon, the favored gas for pumped heat systems, will reach 500°C when pressurized, and excess electricity can be used to power a gas compressor that heats one tank. Now there are two gravel reservoirs, one hot and one cold.
The heat pump then pumps heat from the cold to the hot tank, and whenever the energy is needed, the system is essentially reversed. The heat pump becomes an engine, as gas heated in the hot tank flows back to the cold one and spins the motor as a generator.55
While systems like this are still in demonstration mode, a leading pumped heat storage company, Isentropic, claims that round-trip efficiencies in the 72–80 percent range are realistic.56
Liquid Air Electricity Storage
When air is cooled, it liquefies. When liquefied air is stored and mixed with ambient air or with some waste heat, it converts back to gas. If electricity is used to cool and liquefy the air, most of the stored energy can be recovered when it turns back to gas by sending it through turbines.
Sometimes referred to as Cryogenic Energy Storage (because it’s so cold), Liquid Air Energy Storage promises to do exactly this, potentially on a big scale. A 5-megawatt plant near Manchester, England that will be grid-tied and provide load balancing services is currently under construction.57 Proponents of the technology say that systems boasting hundreds of megawatts of power potential are realistic, as the only limitations are the sizes of the tanks that hold the liquid air.58
For all the rare-earth materials and intricate chemistries of batteries, it turns out that one of the biggest innovations in energy storage has arrived in the form of plain old salt. Salts, it turns out, are remarkably good at storing heat—they can be heated up until they melt and then stored in tight, insulated containers. Today, molten salt is being used to store solar energy to produce electricity overnight, long after the sun has set.
To understand how molten salt can extend the useful life of the sun’s thermal energy, you have to understand concentrated solar power generation. Whereas photovoltaic panels directly convert sunlight into an electric current, concentrated solar uses the sun’s heat thermal energy to spin a turbine and generate power to pump out into the grid. The concentrating of solar rays to harness the energy is far from a new technology. As far back as the sixth century BCE, the Chinese used mirrors to focus the sun’s rays and start fires. Soon after, the Greeks lit the sacred fire at Delphi using the same “burning mirrors.” Legend has it that around 212 BCE, Archimedes ordered his soldiers to use their shields to focus sunlight on invading Roman warships, setting them ablaze.59
Today’s concentrated solar projects use this same basic technology in a much more controlled manner, on an industrial scale, and for the express purpose of generating electricity. Concentrated solar power plants typically focus hundreds or thousands of mirrors on the top of a tall tower, which is filled with either water or (as we’ll soon discuss) molten salt. The mirrors’ movements are all closely controlled by software that tracks the sun and keeps the rays focused on the “solar furnace” atop the tower. In water-based systems, the concentrated sun’s rays boil the water, which releases steam that is fed through a turbine, producing electricity.60
When you replace the water in the solar furnace with molten salt, you have the ability to store energy. Acting as a solar battery of sorts, molten salt can be heated up to more than 1,000°F, storing the energy for up to 12 hours. The heat can then be released into water reserve, producing the steam that spins the turbines that generates the power.61
What are the drawbacks? Technically speaking, molten salt storage is less efficient than batteries, as only 70 percent or so of the energy used to heat up the salt becomes electricity (compared to batteries that regularly score over 90 percent efficiency).62 However, according to an assessment by the National Renewable Energy Laboratory (NREL), the costs of concentrated solar power plus storage are currently cheaper than photovoltaics plus battery backup.63
And while molten salt is proving its value as a storage medium for concentrated solar in facilities like Crescent Dunes, in theory, it could also be used to store energy from any source of electricity. A company called Halotecnics has developed a system that uses electricity to run a basic heat pump and then captures that heat in molten salt, to be stored and then released to generate electricity when needed.64
Case Study: Crescent Dunes Solar Energy Project
A few hours northwest of Las Vegas, near Tonopah, Nevada, the first utility-scale concentrated solar-plus-storage project has recently come online. The Crescent Dunes Solar Energy Project, developed by storage company Solar Reserve, circulates molten salt through an integrated energy storage system—heating to roughly 1,050°F, cooling to 500°F as heat is taken to produce steam, and then circulating back up to the top of the tower to heat up again.
Crescent Dunes can provide its full load of 110 megawatts for up to 10 hours after the sun sets.65 According to Solar Reserve, this load can power up to 75,000 homes at times of peak demand.66 In all, Crescent Dunes can offer 1.1 gigawatt-hours of electricity, delivered to the grid throughout the night.67
In October 2015, Crescent Dunes first synchronized with the grid, and in February 2016 successfully generated its full nameplate capacity of 110 megawatts. Since then, the facility has been ramping up its actual energy output in accordance with its 25-year power purchase agreement with the Nevada utility NV Energy.68 By the end of 2016, Solar Reserve expects to be producing power at its full capacity and offloading the full energy output to NV Energy.
While the pumped heat and liquid air storage systems just described are still unproven commercially, chilling or freezing water as a means of storing energy is an established storage solution. Ice and chilled water storage is already used in hundreds of commercial buildings around the country and is on the cusp of breaking through into residential homes and apartment buildings through the development of HVAC-integrated and refrigeration-integrated thermal energy storage systems.69 The technology is so tried and true that Leadership in Energy and Environmental Design (LEED) includes it in its criteria for green building certification.70
The process is simple enough. Ice is made (or water is chilled) with low-priced, off-peak electricity, typically at night. During the warmer daytimes, cooling loops connected to the HVAC or refrigeration units run through the ice or cold water tanks, extract the cold, and pump the cool air through the building or harness it for refrigeration.71 Ice energy storage is basically a way to power your air-conditioning or for stores to keep coolers chilled with off-peak electricity.
One company leading the charge on distributed thermal energy storage is Ice Energy, with their flagship Ice Bear product. The Ice Bear is essentially a big tank of water that sits outside a home or on a roof of a commercial building and stores energy as ice. Basically, during off-peak hours, the Ice Bear uses electricity to make ice, and then during peak hours—when electricity is in high demand and (depending on the state and utility) possibly more expensive—that stored ice can provide up to four hours of cooling using only 5 percent of the power that would have been required to run an air conditioner.
To be clear, this sort of ice storage doesn’t ever put electricity back into the grid, but it does reduce electricity consumption by reducing the demand on the electric grid during high-priced periods (demand response.).
The company’s first product, the Ice Bear 30, targeted commercial and industrial users, attaching to the standard commercial rooftop HVAC units. A scaled-down version, the Ice Bear 20, is designed for individual houses and can connect to a ductless mini-split system or to existing ducting within the house.72
When employed by a utility, the Ice Bear system can be deployed in smart grid-enabled, megawatt-scale fleets to participate in utility dynamic pricing programs or regional grid demand response markets. The first such trial started in 2014 when Southern California Edison procured 1,800 Ice Bear units for over 25 megawatts of storage. SCE later ordered hundreds more units to be deployed across West Los Angeles.73
“Essentially what we’re doing is we’re shutting air conditioners off during the day, consuming energy at night and displacing that peak load for the utility company,” said Greg Miller, executive vice president of the company, Ice Energy.74
Conclusion: Ready for Prime Time?
Like the battery, alternative forms of energy storage cover the whole range of the scale of technological maturity. Some, like pumped hydro, have been proven commercially for a century, while others—like superconducting magnetic energy storage—are still straining to move beyond research and development.
With battery costs continuing to decline,75 it remains to be seen how well these alternative storage technologies can compete with banks of lithium-ion or other batteries over the next couple of decades.
