**1. Introduction**

The interest in electricity storage has significantly increased with higher shares of intermittent renewable energy sources in the grid. In particular, grid-scale electricity storage with low costs are considered suitable to integrate renewable electricity generation and introduce flexibility to the power grid. Cryogenics-based energy storage (CES), frequently referred to as liquid air energy storage (LAES), is the only energy storage technology so far, which is capable to store large quantities of electricity without geographical limitations or a substantial negative environmental impact.

The thermo-electric energy storage technology stores electricity in the form of a liquefied gas (air) at a cryogenic temperature. The integrated methods of operation (charge, storage, discharge) are displayed in Figure 1. An energy-intensive liquefaction process forms the charging process of CES. The liquefied gas (cryogen) is stored in a site-independent insulated storage tank at approximately ambient pressure and a cryogenic temperature (e.g., −194 ◦C). The compression process of the liquefaction is presented separately, as in the adiabatic CES the heat of compression is recovered and stored to be used in the discharge process. In the discharge process, the liquefied gas is pumped to supercritical pressure in a cryogenic pump, evaporated and superheated, with thermal energy provided by the heat storage, and supplied to a series of expanders regaining a part of the electricity charged to the system.

**Figure 1.** Illustration of cryogenic energy storage steps of operation (charge, storage, discharge), heat and cold recovery and storage.

The cold exergy rejected during the evaporation process is stored in order to increase the efficiency of the liquefaction process (charge). The CES system is composed of well-known components from the industrial gas and liquefied natural gas (LNG) supply chain.

As CES systems are based on mature technology, developers expect comparatively fast progress towards commercialization, competitive costs and efficiency enhancement. CES exergy densities are by approximately two orders of magnitudes higher than of competing technologies such as pumped hydro and compressed air energy storage reaching values higher than 430 kJ/kg. A detailed comparison of CES characteristics to other energy storages can be found in [1]. Moreover, long cycle life, low storage costs, the economy of scale and the independent sizing of charge and discharge unit speak for economic viability. Yet, the adiabatic CES systems upper limit to efficiencies is 45–50%. The thermal integration at the system level is crucial to its performance, which is the reason why the integration of cold storage into the liquefaction process is the subject of this paper.

#### *State of the Art*

Both, cryogenic energy storage and air liquefaction, are no new concepts. Large-scale air liquefaction for industrial purposes became commercial in the 1940s [2] and the first conception of storing electricity in liquid air dates back to the year 1977 [3].

Nowadays CES is rated as a pre-commercial technology being evaluated with a technology readiness level (TRL) of about 8 [4]. The CES concept was confirmed viable in testing, after Mitsubishi extended an existing air liquefier with the first pilot cryogenic power recovery unit (2.6 MW) [5]. The second pilot plant was the first integrated CES plant (350 kW/2.5 MWh) which was the result of joint research between the University of Leeds and Highview Power Storage Ltd. (London, UK) in the year 2011. The results were published in 2015: the CES economic viability was confirmed and a positive outlook on performance and costs was given [6]. A demonstration plant of 5 MW/15 MWh started operation in 2018, demonstrating a number of balancing services [7].

The significance of the liquefaction process to the CES's performance was addressed by [8] as "the key part" of the system as the discharge unit is relatively simple and efficient. The liquefaction was found to account for more than 70% of the overall exergy destruction (MW) of the CES system by the authors in a comparative exergy analysis of two 10 MW CES systems [9].

One of the key findings from the testing of the first pilot plant was the significant increase of system efficiency by cold recovery and storage [6]. The effect of cold recycle was firstly quantified by the authors as the introduction of cold storage doubled the liquid yield of the liquefaction process of the analyzed system [9].

Large-scale air liquefaction has been commercial for several decades. A number of processes exist. The simplest (no moving parts) and first-industrialized configuration is the Linde process, where purified compressed air is cooled and undergoes isenthalpic (free) expansion in a throttling valve, thus brought to its due point by the Joule-Thomson effect [10]. Gas liquefaction is nowadays performed in more complex configurations [11].

Recently, a number of publications have discussed the thermodynamic performance of CES. In the reviewed literature, CES systems with different liquefaction processes, pressures and cold storage configurations are presented in Table 1. Two kinds of a cold storage configuration are presented: (1) quartzite gravel based packed bed store with dry air as secondary working fluid, and (2) a two-tank fluid storage with methanol and propane (or R218) as secondary working fluids and storage media on two different temperature levels.

The liquid yield *γ*, the ratio between the mass flow of the air liquefied in the liquefaction process and the mass flow of the compressed air, is an indication of the charging-unit performance. The liquid yield varies strongly from one publication to the other. The liquid yield increases with liquefaction pressure. Yet, with increased pressures, the power consumption of the compression process increases as well. This is why the liquid yield cannot be considered as the sole indicator for the performance of the liquefaction process.

In general, different assumptions are made in the different references, e.g., ideal dry air was assumed, heat and pressure losses in most components as well as heat losses in the cold box were neglected [12], or assumed lower than 8% [6] which is why comparing the various configurations is problematic.

Three comparative evaluations of air liquefaction processes in CES systems were presented in [8,13,14]. Borri et al. [13] compared three air liquefaction processes (Linde-Hampson, Claude, Collins) for application in a micro-scale CES. The Claude process was identified as the most suitable air liquefaction process. The Linde-Hampson process (with a Joule-Thomson valve only) was found to be inferior and the second cold expander used in the Collins process was claimed to be economically not feasible. Yet, the integration of cold recovery and storage was not considered. Li [8] came to the same conclusion, that the throttle-valve-based Linde-Hampson system is not applicable for CES. Therefore, only the integration of a cold expander instead of a throttling valve in the Linde process and an expander process, employing a refrigeration process with Helium as working fluid, are compared in [8].



1 calculated from: 12 h charging, 3:1 (charge-to-discharge ratio), . *mchar* = 34.1 kg/s. 2 with cold expander/throttling valve.

Abdo et al. [14] compared the by Chen et al [21] patented CES system design based on a simple Linde-Hamson liquefaction process to two alternative systems based on the Claude and the Collins process. The heat of compression was taken into account but cold storage was not comprised. The Claude and Collins process showed similar thermodynamic performance with greater RTE that the Linde based system. Despite the Linde-Hampson having the lowest specific costs, the Claude-based

system was evaluated the best option. The present paper aims to compare a number of air liquefaction process configurations with integrated cold storage in order to identify the most suitable process for implementation in CES systems.
