Adiabatic Compressed Air Energy Storage: Fundamental efficiency limits based on exergy analysis, key technology outstanding challenges and application oriented axial-flow compressor design and performance prediction

Adiabatic Compressed Air Energy Storage (ACAES) is a theoretically appraised solution that populates most of the technical literature on large-scale energy storage. However, major experimental projects and commercial ventures have so far failed to yield any viable prototypes. This apparent contradiction between the projected success and lack of tangible results indicates the presence of challenges, overlooked in technical literature, hindering widespread ACAES implementation. In the first part of this thesis, the underlying reasons behind ACAES failures so far are investigated. An extensive literature review is undertaken, covering well-established and recently published sources. Three potentially problematic trends are identified: (i) A general shortage of full-system experimental setups, especially on mid-scale or larger (⩾ 10’s kW). (ii) Unverifiable performance claims based on best case estimates of performance with simple models. (iii) In most ACAES designs the component operation significantly differs from conventional turbomachinery and heat exchanger operation. Subsequently, it is arbitrated that the combination of these reasons may have led to over-optimistic development in commercial ventures too early in the research, development, and technology demonstration processes, leading to developer bankruptcy or project abandonment.

Following the literature review, an analytical model of three-stage reversible compression, three-stage reversible expansion, fully effective heat exchangers ACAES design is modelled, and fundamental losses even with idealised components are identified – something overlooked in previous works. Through exergy analysis, it is found the conventional 3-stage ACAES fully ideal system can only reach efficiencies up to ≈ 93%, with isochoric air storage. Furthermore, there exists a strong coupling between the charging and discharging processes which imposes reversibility challenges. When real-component performance is considered, losses further increase and additional, practical challenges were identified. Firstly, while compressors tend to stack inter-coolers in between stages to reduce work consumption, turbines (especially gas) operate more efficiently at higher temperatures. Since in ACAES the compressor discharge and turbine inlet temperatures are strongly coupled through the TES, an apparent contradiction arises, in whether this limiting temperature should be kept lower (benefiting the compressor) or higher, to ensure turbine performance is satisfactory. Secondly, in isochoric ACAES, the charging process invariably occurs with variable pressure on the compressors, which impacts the compressor operating point, changing it mass flow rate, isentropic efficiency and rotational frequency. 

Finally, because of the above, an axial compressor on- and off-design model is developed based on an ideal-gas, inviscid adiabatic through flow method, in which mass continuity, energy balance and Navier-Stokes’ momentum equation are used to calculate the performance and define the compressor geometry given a desired final outlet total pressure and mass flow rate. Due to time constraints, the turbine and heat exchangers are not designed in detail. The compressor performance results were implemented into the ACAES model developed and tested for a case study, which consisted of a grid-scale, two spool inter- and after-cooled axial compressor, capable of delivering air between 5.5 and 7.7 MPa to charge a 300,000 m3 reservoir. The compressor design mass flow and power consumption are 100 kg s−1 and 60 MW, respectively, with an on-design isentropic efficiency of ≈ 92%.The turbines are designed to run constantly at 90 kg s−1 at a throttled pressure of 5.4MPa and isentropic efficiency of 90%. Heat exchangers are assumed isobaric, operating with variable mass flow rate to compensate for changes to the air heat capacity. The system calculated round trip efficiency under continuous cycling is 72%, with the full charge and discharge periods of 15 hours each. To the best of the author knowledge, such extensive ACAES compressor design and the resulting determination of its on- and off-design performance is missing from the available CAES literature. Future work is required to extend this highly detailed, application-oriented approach into heat exchangers, Thermal Energy Storage and turbines. Still, the product of this thesis consists of a well-defined guideline on what such requirements are and a modular compressor routine that can not only be easily integrated to these ACAES oriented components but also implemented into already existing "black-box" thermodynamic models.