We need to store the huge quantities of excess electricity generated by variable renewables. But what’s the best way? Currently, over 99% of large-scale electricity storage uses pumped hydro dams. But geography severely limits where you can build one. And the growth of grid-scale batteries is limited by raw material costs and short lifecycles. Antoine Koen and Pau Farres Antunez at Cambridge University review an important alternative, Pumped Thermal Electricity Storage. It’s still in the development stage, but the technologies it uses – heat exchangers, compressors, turbines, electrical generators – are commonplace and proven in the power and processing industries. Units can be installed anywhere. They are compact: 1kg of water stored at 100°C can release ten times more electricity than 1kg stored at a height of 500 metres by a pumped hydro plant. The components last for decades, unlike batteries. The heat is stored in cheap and plentiful materials: gravel, molten salt, water. The downside is their energy conversion efficiency. At 50-70%, its well below 80-90% for lithium-ion batteries or 70-85% for pumped hydro storage. But who cares if, as expected, costs decline as the prototypes move to commercialisation, and costs continue to decline for variable wind and solar.
The effect that fossil fuels are having on the climate emergency is driving an international push to use low-carbon sources of energy. At the moment, the best options for producing low-carbon energy on a large scale are wind and solar power. But despite improvements over the last few years to both their performance and cost, a significant problem remains: the wind doesn’t always blow, and the sun doesn’t always shine. A power grid that relies on these fluctuating sources struggles to constantly match supply and demand, and so renewable energy sometimes goes to waste because it’s not produced when needed.
One of the main solutions to this problem is large-scale electricity storage technologies. These work by accumulating electricity when supply exceeds demand, then releasing it when the opposite happens. However, one issue with this method is that it involves enormous quantities of electricity.
Cost limits batteries. Geography limits pumped hydro
Existing storage technologies like batteries wouldn’t be good for this kind of process, due to their high cost per unit energy. Currently, over 99% of large-scale electricity storage is handled by pumped hydro dams, which move water between two reservoirs through a pump or turbine to store or produce power. However, there are limits to how much more pumped hydro can be built due to its geographical requirements.
One promising storage option is pumped thermal electricity storage. This relatively new technology has been around for about ten years, and is currently being tested in pilot plants.

The conversion of electricity to heat happens in the central circuit, then stored in hot and cold tanks. / image by Pau Farres Antunez
Pumped thermal electricity storage works by turning electricity into heat using a large-scale heat pump. This heat is then stored in a hot material, such as water or gravel, inside an insulated tank. When needed, the heat is then turned back into electricity using a heat engine. These energy conversions are done with thermodynamic cycles, the same physical principles used to run refrigerators, car engines or thermal power plants.
Known technology
Pumped thermal electricity storage has many advantages. The conversion processes mostly rely on conventional technology and components (such as heat exchangers, compressors, turbines, and electrical generators) that are already widely used in the power and processing industries. This will shorten the time required to design and build pumped thermal electricity storage, even on a large scale.
Cheap, clean materials: gravel, molten salt, water
The storage tanks can be filled with abundant and inexpensive materials such as gravel, molten salts or water. And, unlike batteries, these materials pose no threat to the environment.
Large molten salt tanks have been successfully used for many years in concentrated solar power plants, which is a renewable energy technology that has seen rapid growth during the last decade. Concentrated solar power and pumped thermal electricity storage share many similarities, but while concentrated solar power plants produce energy by storing sunlight as heat (and then converting it to electricity), pumped thermal electricity storage plants store electricity that may come from any source – solar, wind or even nuclear energy, among others.

A concentrated solar power plant. / National Renewable Energy Lab, CC BY-NC-ND
Easy to deploy, compact
Pumped thermal electricity storage plants can be installed anywhere, regardless of geography. They can also easily be scaled up to meet the grid’s storage needs. Other forms of bulk energy storage are limited by where they can be installed. For example, pumped hydro storage requires mountains and valleys where substantial water reservoirs can be built. Compressed air energy storage relies on large subterranean caverns.
Pumped thermal electricity storage has a higher energy density than pumped hydro dams (it can store more energy in a given volume). For example, ten times more electricity can be recovered from 1kg of water stored at 100°C, compared to 1kg of water stored at a height of 500 metres in a pumped hydro plant. This means that less space is required for a given amount of energy stored, so the environmental footprint of the plant is smaller.

Molten salt tanks for thermal energy storage in a concentrate solar power plant. / Abengoa
Long life, decades
The components of pumped thermal electricity storage typically last for decades. Batteries, on the other hand, degrade over time and need to be replaced every few years – most electric car batteries are typically only guaranteed for about five to eight years.
Conversion efficiency is lower than batteries, pumped hydro
However, even though there are many things that make pumped thermal electricity storage well-suited for large-scale storage of renewable energy, it does have its downsides. Possibly the biggest disadvantage is its relatively modest efficiency – meaning how much electricity is returned during discharge, compared to how much was put in during charge. Most pumped thermal electricity storage systems aim for 50-70% efficiency, compared to 80-90% for lithium-ion batteries or 70-85% for pumped hydro storage.
…but who cares, if costs decline enough
But what arguably matters most is cost: the lower it is, the faster society can move towards a low carbon future. Pumped thermal electricity storage is expected to be competitive with other storage technologies – though this won’t be known for certain until the technology matures and is fully commercialised. As it stands, several organisations already have working, real-world prototypes. The sooner we test and start deploying pumped thermal electricity storage, the sooner we can use it to help transition to a low-carbon energy system.
***
Antoine Koen is a PhD Candidate in Pumped Thermal Energy Storage, University of Cambridge
Pau Farres Antunez is a postdoctoral researcher in Energy Storage, University of Cambridge
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Many heat-engine possibilities are being pursued. The Siemens version with electrical dissipation into hot rocks is subject to Carnot losses but is simple, and the waste heat could be useful. Cryogenic storage is being tested, perhaps with optional heat stores. A real heat pump prototype at Newcastle University using reversible piston engines with argon gas and a gravel hot/cold store has been built; this is perhaps the most elegant device but seems to have been a very complicated development project.
As a non-expert I have no idea of the future economic prospects – it must depend a great deal on the frequency and length of the gaps in renewables generation, and I wonder whether there has been some serious modelling of how and at what scale these would fit in to a wind/solar energy economy, and what cost levels must be achieved.
Any research done how this could be integrated with district heating networks? I would assume that could increase the efficiency.
We encourage commenters to contact the authors directly if they need specific answers (they are always listed at the end of the article). They won’t all be replying to comments here.
99 per cent of storage is not pumped hydro.
There is a lot of storage in ordinary (non-pumped) hydro for example in Norway and Sweden (together about 50 TWh) but have no pumped hydro at all. For that reason electricity prices do not differ a lot between day and night. When demand is high, more water runs through the turbines. When demand is low, water levels rise due to rain and melting snow. It is used that way for all timescales from seconds to years.
This storage resource is far from fully used.
LAES has dropped the cost of energy storage using Li-ion batteries by 94%. Surely this is the end of the ‘intermittency problem’.
Wind power for July 2018 fell 3.4 TWh below the monthly average for the year. To raise that figure up to the monthly average using Li-ion batteries would have cost £5,272 billion, but LAES, it would only have cost £300 billion:
https://bwrx-300-nuclear-uk.blogspot.com/2019/10/the-intermittency-problem-will-energy.html
The technology is worth while commercialising. How ever, the impact will be significant in countries were utility scale wind and solar is predominant.