On 30 April 2015, electric vehicle manufacturer Tesla announced that it would begin selling stationary lithium-ion battery systems intended for use in the electric power system. What does Tesla’s announcement tell us about where the grid energy storage industry is going and about the global outlook for battery storage?
Few electricity users realize, as they switch on the lights or the television, what a complex balancing act it takes to keep the electric power grid running. Supply and demand for electricity must be very closely matched in real time or the lights will go out. Recent developments in the power sector—most notably the steep decline in the cost of renewable electricity sources whose output is intermittent—will soon be exacerbating this long-standing operational challenge.
For decades, grid operators have had to manage their balancing act with limited tools, as storage technology was expensive and the system operator had little influence over real-time customer demand. Operators have historically had just one option for satisfying the technical mandate that supply must meet demand, as well as the legal mandate to keep the lights on: building more reserve power plants. Reserve capacity is needed on a short-term basis for what grid operators call “frequency regulation”—essentially the balancing of fluctuations of supply and demand. It is also needed on a longer-term basis to ensure that the grid has a sufficient margin of capacity for supply to satisfy peak demand each year.
The possibility of holding “inventory” that can be drawn down and replenished as needed to meet customer demand and other market requirements is what makes energy storage such an attractive prospect for providers. At present, however, the grid can be managed relatively cost-effectively as it always has been— without a sharp increase in storage capacity—even as intermittent renewable energy sources begin to penetrate the market. Broadly speaking, there are few power systems using renewables where current levels of wind and solar generation cannot be managed with existing grid resources. Even though the dramatic fall in the price of solar energy over the past decade has enabled electricity customers in some places to begin producing power at homes and businesses more cheaply than they can buy it from the grid, they still rely on the grid for backup when the sun is not shining.
With few exceptions the electric power system does not need storage. However, it would certainly benefit from a cost-effective storage solution across the grid. Storage is also set to become more valuable in the decade to come owing to two key trends: an increasing reliance on renewables in established markets; and the increasing electrification of less developed countries and areas with weaker or non-existent centralized grid infrastructure. A crucial factor in how these trends play out will be, of course, the evolution of storage technology, particularly of batteries.
What batteries can offer
Until recently, the only significant sources of electricity storage on the grid have been large hydroelectric plants with pumps designed to reverse the flow of water, effectively storing electrical energy by converting it into potential energy. Currently thermal plants—and in a few cases, pumped hydro storage plants—are the global go-to option for incremental peaking-capacity needs.
However, the best sites have mostly been tapped in established power markets; in the United States, for example, no new “pumped hydro storage” has been installed in over a decade, while in Europe new developments have a long lead time, require huge investment, and face strong environmental and social opposition.
Meanwhile, a combination of policy support and niche economic opportunities has ushered in a wave of technology development focused on alternative storage systems, including electrochemical batteries. Some of the interest in batteries is driven by the fact that even though, as we have noted, existing grid resources are adequate for managing most wind and solar generation at current levels of penetration, grid constraints are already emerging as penetration increases. In several European markets, for example, grid operators have needed to curtail 2-3% of annual wind generation— meaning that they have put the brakes on wind turbines when the wind is blowing—because of the challenge of integrating all the power they generate into the grid; this percentage, moreover, is expected to increase with additional wind capacity.
Batteries as a means of energy storage, then, have a potential role in enabling power systems to continue pushing the boundaries of renewable energy generation and accelerating the electrification of poor, remote areas of the world. Other applications include serving whatever onsite customer storage needs eventually emerge in the marketplace, as well as managing the short- and long-term grid flexibility needs that exist even without wind and solar.
From a technical standpoint, most batteries are an extremely good fit for this last application—indeed, batteries have already become an economically viable option for it in some regions. In the PJM Interconnection in the United States, for example, where a sizable industrial load creates a high demand for frequency regulation, batteries are likely to be providing at least 150 megawatts (MW) for this purpose by the end of this year. In South Korea, the state-owned utility is funding an initiative to provide 500 MW of storage for frequency regulation by 2017.
Beyond these current niche applications, batteries are most valuable as an alternative source of peaking capacity, the standby resource that grid operators need to be prepared for when electricity demand hits its peak. Yet if they are to play a widespread role in established power systems, batteries will have to compete against the current conventional sources of peaking capacity (gas-fired power plants). Furthermore, the evolution of the regulatory framework will determine how batteries for peaking capacity will compete with demand-side management, which is at a relatively early stage of market development in many regions. Such considerations point to the need to understand the economics of battery storage to assess its viability for any of the applications for which it can be used.
The economics of battery storage
Li-ion batteries, like those Tesla is selling, are a family of battery chemistries distinguished by their high efficiency and energy density. These characteristics have made them the battery of choice so far from a performance standpoint, but not necessarily from a cost standpoint. Falling li-ion prices, however—epitomized by Tesla’s recent announcement—are now fuelling intense price competition in the sector and challenging all other nascent energy storage technologies, including: sodium-sulfur batteries, currently the most widely deployed type on the grid; emerging “flow” batteries using a liquid electrolyte; and earlier-stage alternatives such as metal-air batteries and advanced compressed air storage.
Three years ago, li-ion batteries intended for use on the grid were some of the most expensive storage units on the market, with a price over $1,000 per kilowatt- hour (kWh). Yet that cost has fallen rapidly due to cross-sector research and development and economies of scale, and li-ion has become the clear winner for consumer electronics and automotive applications. While Tesla has not yet announced any of the specifications for its $250-per- kWh product, nor any information on its margins, that price demonstrates the progress the industry has made in the past few years alone.
But is this price low enough? From a utility cost-of-service perspective, there are some locations where Tesla’s prices would be low enough to compete with combustion turbines—in particular, densely populated, urban areas, which are a common source of demand for new peaking capacity in the US, and where siting and building new thermal power plants or transmission lines is very challenging and expensive. Outside of those areas, however, Tesla’s prices are not yet low enough to compete.
Similarly, from a consumer perspective, there will likely be a small subset of residential customers willing to pay for a Tesla home-backup unit. However, those who simply want a reliable source of backup power at the lowest cost can still buy a small-scale diesel or natural gas generator for much less.
Despite these challenging economic fundamentals, a variety of government policies and the intricacies of certain electricity customer rate designs in the United States, Europe, and Japan are improving the economics for battery deployment, particularly at larger commercial and industrial customer sites. Both the German and Japanese governments, for example, are now providing incentives for acquisition of small, behind-the-meter batteries. In the United States, the California Public Utilities Commission issued a mandate in October 2013 for the state’s three largest investor-owned utilities to add 1.3 GW to their grids by 2020. California’s storage mandate highlights the potential scale of storage demand in the power sector relative to other major sources of battery demand.
Where the opportunities are
Despite the obstacles facing battery suppliers such as Tesla, Panasonic, Samsung, and LG Chem in becoming widely competitive, the kinds of government policies and electricity rate designs described above are creating near-term opportunities for deployment. Most of these initial opportunities depend on installing batteries “behind-the-meter” in residential and commercial buildings. Li-ion’s high energy density enables the technology to be deployed at a distributed level—behind- the-meter as well as on the distribution network—where it can provide highly localized capacity benefits.
While some “early-adopter” customers in the United States may be motivated to buy Tesla’s residential products, there is little economic rationale for installing energy storage in US homes. However, there is a possible business case for residential energy storage in regions where the price utilities will pay customers for their excess solar generation is much lower than the “retail” price consumers pay for electricity from the grid.
One of the most extreme examples of this phenomenon is Germany, where the solar feed-in tariff currently pays residential system owners €0.12 per kWh versus a retail electricity rate of approximately €0.29 per kWh—potentially saving the customer €0.17 per kWh of solar electricity consumed rather than exported to the grid. Germany also happens to be the world’s largest solar market in terms of total installed capacity. As a result of these market and policy factors, approximately 100 MW of systems similar to Tesla’s Powerwall will have already been installed with small residential PV systems in Germany by the end of this year.
While the business case for this application is heavily reliant on government incentives and a growing desire from customers to reduce the amount of electricity they purchase from utilities after several years of steep price increases, falling battery prices will accelerate the development of this business model and help it become viable in an increasing number of regions. However, this model could be threatened by changes in grid interconnection rules for solar and storage systems—changes that are needed to protect the revenue that pays for the grid, which customers still rely on.
In the United States, Tesla is tapping into a different segment: commercial and industrial (C&I) buildings in California and New York/New Jersey, where electricity rate designs for the large C&I customers include very high “peak demand charges.” These charges are a function of the largest amount of power that the customer draws from the grid in the course of a billing cycle, providing an incentive for customers to smooth peaks in electricity usage. For example, Tesla has already announced plans to provide 4.8 megawatt- hours (MWh) of batteries to an Amazon data center in California, and has submitted over 70 MW of applications for incentives from California utilities for similar projects.
Several early-stage companies (including Stem and Green Charge Networks) as well as leading US residential and commercial solar developers (SolarCity and SunPower) are targeting this application and have developed “grid-ready” battery systems—which include a battery module, an inverter, control systems, interconnection equipment, and specialized dispatch algorithms—that can leverage a relatively small amount of energy storage to clip the peak for customers for whom this is a significant need, thereby saving the customer a substantial amount of money. These companies also offer “no up-front cost” solutions such as leases and shared savings agreements. Tesla’s close relationship with the US’ largest solar installer, SolarCity, which already has significant C&I customer channels, could help it accelerate its push into this market.
Developers have seized upon this business model not just in California but also in New York, where incentives are also available to make the economics more attractive. IHS expects this model to drive approximately 50 MW of commercial behind-the-meter installations in North America by the end of 2015. Falling prices and increased adoption in behind-the-meter sites are also enabling storage project developers to begin pioneering “distributed peaking plant” business models by aggregating many distributed battery units as a dispatchable peaking resource for grid operators.
Given that storage hardware—essentially batteries and inverters—is useless until integrated with the other components that go into creating an intelligent storage system, the downstream segment of the energy storage value chain is also attracting investment from a wide range of players, which shows a selection of companies currently active in the sector and their positions along the value chain. Notable early entrants include established power equipment suppliers such as GE, S&C Electric, ABB, and Eaton, which have leveraged their existing power conversion technologies to develop utility-scale battery systems.
Globally, it is clear that grid battery technology has evolved beyond the demonstration stage to the early-commercialization stage of market maturity. Niche opportunities abound, in frequency regulation and in peak-shaving in very dense, urban locations. It is no surprise that the initial wave of commercial deployments is centered in regions with supportive policy, most notably California, Germany, Japan, and South Korea. If these early opportunities are sufficient to drive continued investment in scale and R&D, storage could emerge as a significant new resource for grid operators around the world. And the adoption of cost-effective storage technology will smooth the way for greater grid penetration of wind and solar energy in decades to come.
Andy Lubershane is senior analyst, North America renewable power, IHS Energy; and Sam Wilkinson is research manager, solar & energy storage, IHS Technology