Cheapest Method for Storing Energy to be Lithium-ion Technology by 2050

January 10, 2019 by Scott McMahan

By 2050, lithium ion-based batteries will be the least expensive way to store energy from power generation like solar or wind farms, according to a new study by researchers at the Imperial College of London. The new research determines the cost of storing energy with various technologies, such as pumped-storage hydroelectricity and large-scale batteries, and forecasts those costs into the future.

Based on these predictions, lithium-ion-based battery technologies are expected to be the most economical way of storing energy for the majority of applications, including ensuring that energy grids do not suffer large fluctuations and letting consumers manage the bills.These applications are especially critical for energy sources that may be intermittent including solar power or wind.

The new study, published in Joule, could help industry and policymakers choose where to invest in research and industry to make the best use of energy storage technologies. While previous studies of energy storage cost mainly focused on the investment costs, the new study determines the ‘levelized cost of storage,' which includes the full costs of storing energy with the investment, operation, and charging expenses, as well as the efficiency, the technology lifetime, and performance degradation.

"Our model is the first to project full energy storage costs into the future, allowing predictions of which technology will be most competitive in a particular application at a particular time," said lead researcher Oliver Schmidt, from the Grantham Institute and the Centre for Environmental Policy at Imperial College.

The model reveals that presently, the cheapest energy storage method is pumped-storage hydroelectricity, where spare energy is used to pump water to a higher elevation. It is then released to collect the energy when demanded. However, the study suggests that as time progresses, pumped-storage hydroelectricity costs do not decline, whereas lithium-ion battery costs drop, making them the cheapest options for most applications from 2030, according to the study.

Flywheel and Hydrogen storage technologies also become the least expensive for particular applications, including when the stored energy has to be released over a long time period, or when it must be discharged very frequently. Despite this, lithium-ion technologies were found to be the cheapest for the bulk of applications.

Schmidt explained, "Our projections show that lithium-ion technologies will see a rapid decline in costs over the next couple of decades. This is driven mainly by the fact that lithium-ion is manufactured at scale. The resulting reduction in initial investment costs is more significant than for other newer technologies such as flow batteries and flywheels, potentially outcompeting any performance advantages of these newer technologies."

Economics of Lithium-Ion Battery Technology has Same Pattern as Crystalline Silicon Solar Panels

Dr Iain Staffell, senior author on the article from the Center from Environmental Policy, said, "We have found that lithium-ion batteries are following in the footsteps of crystalline silicon solar panels. First-generation solar cells were high performance but very expensive, so cheaper second- and third-generation designs were developed to supersede them. However, sheer economies of scale mean these first-generation panels now cannot be beaten on price.

"Similarly, lithium-ion batteries were once expensive and only suited to niche applications, but they are now being manufactured in such volumes that their costs are coming down much faster than the competing storage technologies."

Schmidt added, "This doesn't mean that these other technologies should be abandoned, but they have to perhaps focus on performance and efficiency, making them the best they can be before being deployed at a larger scale."

The team made their model open access, which lets academics, consumers, as well as industry, and policymakers run their own simulations with the data for particular applications.