Pumped thermal energy storage (PTES) and liquid air energy storage (LAES) are two technologies that use mechanically-driven thermodynamic cycles to store electricity in the form of high-grade thermal energy, employing abundant materials that are kept in large insulated tanks. Both technologies are free from geographic constraints, providing a significant advantage over competing methods such as pumped hydro or compressed air energy storage. The focus of this thesis is on the analysis, modelling and development of these technologies.

A number of PTES systems have been proposed based on different thermodynamic cycles. A variant based on the Joule-Brayton cycle employing liquid storage media is studied here. An analytical study is presented that reveals how the performance of the cycle varies along a range of operating conditions. Generally, the same strategies that minimise compression/expansion losses also maximise heat exchanger losses, which results in optimal points at certain operating conditions. A numerical model is developed to find these optima while accounting for real fluid properties. Employing a regenerative heat exchanger is found useful to adapt the cycle to the operating temperature ranges of the storage liquids and to increase the performance of the cycle.

A new combined cycle that integrates PTES and LAES is presented. The fundamental advantage is that the cold thermal reservoirs that would be required by the separate cycles are replaced by a single heat exchanger that acts between them, thereby saving significant amounts of storage media per unit of energy stored. Several configurations are possible and these are studied and optimised. The most advanced configuration reaches a round-trip efficiency of 71 % under nominal conditions, compared to 65 % for stand-alone PTES and 61 % for LAES. A further adaptation of the combined cycle is presented which only employs water and liquid air as storage media, dramatically reducing the cost of energy capacity.

The performance of the heat exchangers is found to have a significant impact on the overall performance of the various cycles. For this reason, an optimisation procedure is developed to obtain heat exchanger designs that minimise entropy generation for a given amount of material. These designs are used when estimating the costs of energy capacity and power capacity of each cycle. Results indicate that the best cycle configurations would be competitive with reported costs for pumped hydro and compressed air energy storage.