Commentary: Who wants to be in charge?
21 November 2017
The authors of this commentary are IEA Energy Technology Analysts Simon Bennett and Luis Munuera
Innovation and investment in battery technologies are imperatives for a faster energy transition that cut across different parts of the energy system. The cost and performance of batteries will not only shape how fast electric vehicles (EVs) get into the hands of price-sensitive consumers, but also how easy it is to integrate renewable electricity into the grid. How quickly energy access is expanded and how many smart, connected devices we will be using to manage our energy demand will also be determined by batteries. These days you are just as likely to encounter stories about lithium, cobalt and nickel resources, all essential battery ingredients, in the energy press as stories about coal, oil and hydro.
Recently, statements have been made by the European Commission, and governments of China, India and Japan, among others, about the development of their respective battery production and electric vehicle industries, stressing the importance of manufacturing in these sectors to future prosperity.
Prices of lithium-ion batteries have fallen dramatically in recent years. For each doubling of output, lithium ion battery prices have been getting 19% cheaper. For batteries with sufficient power and energy density for electric vehicles and electricity storage on the grid or in buildings, this “learning rate” translates into a price drop of around three and a half times since 2009.
Two factors are at play: better technology and mass manufacturing to meet expectations of growing demand.
Manufacturing capacity for these types of batteries has soared upwards in the last seven years. Based on our database of factories around the world, we expect the total manufacturing capacity to reach around 200 GWh of EV and energy storage lithium ion batteries this year. If factories open as announced in 2017, then they will represent the largest amount of investment in new battery manufacturing capacity in any single year, at almost USD 8 billion. In total, there are over 200 such factories in operation around the world and, since 2015, most of the capacity is in China. However, the field remains diverse: among the top manufacturers are automakers, such as Tesla and BYD, a major electronics firm (Panasonic), a chemical company (LG Chem) and newcomer battery specialists like CATL and Guoxuan High-Tech.
Note: Database may not include all relevant battery factories in China, especially for 2022. Some media reports estimate the total capacity to be higher.
More is being invested per year in new battery production capacity despite falling manufacturing costs. The size of the facilities has grown tenfold from an average of 500 MWh per year in 2006 to 2010, to an average of around 5 GWh for plants announced to be constructed after 2016. Through larger scales, maturing supply chains and better production techniques – robotics and standardisation – the announced capital costs of the plants per unit capacity have fallen fivefold since 2010.
Note: this neither directly reflects battery production costs nor battery installation costs, which are dependent on non-hardware costs of preparing and implementing electricity storage projects.
As a sign of just how fast battery manufacturing capacity is growing, it continues to be around double the annual additional installed capacity of batteries in vehicles, buildings and grids despite rapid demand growth. Traditional markets are still growing – at around 6% per year for cell phones – but EVs have reshaped demand in just a few short years, moving from 14% of the battery market in 2013 to almost half in 2017; over 40% annual demand growth. It is likely that 2018 will be the first year when EVs represent more than 50% of the lithium-ion battery market.
This does not mean that plants are on average running at 50% utilisation levels, because many of these plants are also capable of producing batteries that can be used in consumer appliances, low-speed EVs and e-bikes. What it suggests is that battery production capacity will not be a limiting factor in EV expansion in the near future and market tightness will not be a major pressure on battery prices. Government policy, especially in China, has driven up capacity on the supply side. On the demand side, policies have played an equally important role. Demand continues to be boosted, but also limited, by policies including national, state and city-level support for EVs and renewables. The complexity of generating revenues from utility-scale storage projects is also a policy-related issue that constrains demand growth.
Note: We assume that all-electric buses today have an average battery capacity of 150 kWh.
The unfolding story of battery costs is more complex than the high-level numbers imply. Not all lithium ion battery applications use the same chemistry or battery designs, so the aggregate manufacturing capacity cannot all serve consumer needs. Grid-scale energy storage applications have more relaxed weight and space requirements than cars and homes. EV makers have tailored their vehicles to different battery types, and vice versa; plug-in hybrid EVs need different batteries to pure battery EVs. Partly for this reason, no single type of battery has a share greater than around one third of the market. While nickel-cobalt-aluminium (NCA) cathodes are popular in the United States and Japan, nickel-manganese-cobalt (NMC) remains the global market leader. In fact, since China changed its EV subsidies in 2017 to make NMC batteries eligible, NMC has dramatically displaced the cheaper lithium-ion phosphate (LFP) in China and pushed the manufacturing trend strongly towards nickel-based cathodes. This shows that factories can sometimes be rapidly reconfigured to a new type of battery, and that the demand outlook for metals like nickel and cobalt can be transformed by changes in policy and performance needs.
Even within cathode types, battery designs do not serve identical functions. For example, designers for hot countries are identifying different design solutions from those that are most efficient in colder climates like Europe, despite both being based on the same NMC type. Battery research is therefore pushing the frontiers in several directions at once. These efforts are focusing on different needs for energy density, power output, charge-discharge deterioration, cost, safety and temperature. Today’s batteries will be superseded in the next decade by configurations that perform better against these metrics. It is unclear whether the result will be multiple variants of, say, NMC, or whether newer technologies – like lithium sulphur, solid state, lithium-air, vanadium flow – will break through. To date, it has taken around a decade to bring a new battery technology to market.
The differentiation of the market between battery types makes cost predictions hard. Currently, batteries are following the type of learning curve of solar photovoltaic (PV) or Moore’s Law for semiconductors, but the fundamentals are not the same as for these other technologies and the rate of future improvement is not guaranteed. Mass manufacturing drove cost reductions in PV panels because the product has become relatively standardised and we can expect the capital costs of battery factories to follow a similar pattern. However, the material inputs to PV are not subject to the same commodity cycles as lithium, cobalt and nickel, which are not widely available in all regions. Cobalt prices have doubled since late 2016 and there have been several recent high-profile mining deals. While a fourfold increase in costs of these inputs would only raise their share of marginal battery costs from around 4% to 15%, cyclical prices for these commodities will certainly affect battery prices and thus propel innovation to reduce reliance on them. The future of the historical 19% battery price learning rate will be closely fought between mass production, metal commodity cycles and better chemistry and design.
For batteries it seems there is an alternative to the mass production-driven “winner takes all” scenario. The alternative “race to the top” scenario opens up opportunities for countries that do not necessarily have the edge in mass production but have other competitive edges in technology or value chain integration. So-called battery “gigafactories” are under construction or in planning in a wide range of countries for varying reasons. In Australia there are lithium deposits and a growing demand for batteries to support solar PV; in Sweden the support for innovation is high; in the United States there are integrated EV value chains; in Japan there are battery companies with long-established industrial ecosystems; and there are diverse other reasons in India, Thailand, Poland, Hungary, Korea and the Germany. New plants and companies around the world could earn an edge and win contracts with carmakers, electricity utilities and households.
On the other hand, if overcapacity in one type of adequate, durable chemistry drops market prices then automakers may design cars to take advantage of cheap, available batteries. In this PV-like scenario, mass manufacturing of a standardised design could be the biggest factor, crowding out other variants. Importing regions will nevertheless be able to find value in associated electronics equipment, software, consumer services and high-tech refurbishment of degrading cells.
Governmental industrial strategies could play a key role in developing competitive advantage and building value chains, including for battery recycling and used batteries (for example, EV batteries make great home storage devices even after their performance is too low for use in cars). But, at the same time, it will be international trade and multi-lateral collaboration that will reduce costs and improve performance fastest. Furthermore, the environmental, social and geopolitical footprint of batteries will need to be carefully managed if their potential is to be fully realised.