RheEnergise demonstrated that hydropower storage using a high-density liquid can operate outside the laboratory after building upper and lower reservoirs at the Cornwood site in Devon and installing pumps, pipes, and generation equipment, then pumping a proprietary liquid uphill and returning it through a turbine. The demonstration reached the design capacity of 500 kilowatts, but it was not connected to the grid and provided only about 15 minutes of discharge at full power.
This assessment is based on an analysis published by Michael Barnard and on the final government report on the British demonstration. The main conclusion is not that the technology failed, but that proving the system can operate is different from proving that it can scale economically to multi-hour or multi-gigawatt storage projects.
What Did the Cornwood Project Prove?
The system was primarily designed to operate for four hours at a power of 500 kilowatts, but difficulties producing a sufficient quantity of liquid to the required specifications reduced the available inventory. The integrated system also recorded a round-trip efficiency of 59% before accounting for parasitic loads, which had not yet been measured. By contrast, RheEnergise assumes an efficiency of approximately 80% for commercial systems, with larger equipment potentially improving performance, although the gap between the assumption and the current measurement remains a point requiring practical validation.
The demonstration also revealed operational challenges, including the sensitivity of barite to moisture, unexpectedly slow mixing, differences in the chemistry of alternative materials, and the need for months of troubleshooting and reformulation of the liquid-production process. These results are natural in a first-of-its-kind project, but they mean that the liquid is not merely an inert substitute for water; it is a complex part of the storage plant itself.
Density Does Not Come for Free
The liquid used has a density of approximately 2.5 times that of water. From an engineering perspective, this makes it possible to store the same gravitational energy using a smaller liquid volume or a lower height, potentially opening sites unsuitable for conventional pumped-storage plants. But this advantage requires manufacturing a mineral-loaded liquid while maintaining its homogeneity, chemical properties, and behavior during pumping and discharge.
The problem increases as storage duration rises. With power and height held constant, each additional hour requires roughly a proportionate quantity of liquid and therefore a larger quantity of weighting mineral. The analysis uses an example with a capacity of 600 megawatts and a duration of 18 hours, estimating that a barite-rich formulation, according to what RheEnergise’s patent reveals, could require approximately 8.8 million tonnes of barite for the active liquid alone. This excludes inactive inventory, processing losses, and reserves.
The Duration and Competition Dilemma
Storage duration places the technology between different competitors. If RheEnergise targets shorter systems, such as four, six, or eight hours, securing minerals becomes relatively easier, but it enters a field where lithium-ion batteries have a strong presence and declining costs. The analysis notes that BloombergNEF’s 2025 survey estimated the price of stationary-storage battery packs at approximately $70 per kilowatt-hour, while cautioning that the pack price is not the same as the full installed cost of an energy-storage system.
The company’s early commercial focus is also centered on units ranging from 10 to 20 megawatts, with a storage duration of approximately four to eight hours. Moving to 12, 18, or 24 hours reduces the strength of the comparison with batteries, but raises mineral requirements approximately linearly with the stored energy while simultaneously strengthening the advantage of conventional pumped storage, which uses water—a cheap and stable medium that does not require the manufacture of a specialized material.
Where Might the Technology Work in Practice?
The analysis suggests that the most realistic opportunity may lie in specific sites near mines or quarries. A single site could combine land previously affected by industrial activity, suitable terrain, limited grid connectivity, and a large supply of suitable material or waste that could be used. But this model does not eliminate the site constraint; instead, it adds new conditions involving material density and chemistry, recovery rates, particle behavior, and the economics of processing it.
In practical terms, Cornwood proved that pumping the dense liquid and recovering electricity from it are possible. But it has not yet proved that this system can provide long-duration storage at a cost and scale competitive with alternatives. Shortening the duration brings it closer to batteries, while extending it magnifies its dependence on minerals and once again highlights the advantage of water in conventional pumped-storage plants. Therefore, based on the current evidence, the technology appears better suited to specific application areas than to serving as a general solution for the long-duration storage market.