**1. Introduction**

The share of renewable generation in electricity production is ever increasing with the feasibility of a 100% renewable supply supported by multiple studies [1–3]. The intermittent nature of these sources puts increasing requirements on electricity storage and system flexibility. Lithium batteries are a well-established technology within this field, provide high efficiency (95%, though in real operation, auxiliaries and performance decay by wearing and ageing can notably decrease this value [4,5]) and relatively low cost per unit power (€/kW). For grid scale medium and long duration applications, however, they are economically well fitted to no more than several hours of capacity due to the high cost per unit capacity (€/kWh). Investigation of a hypothetical 100% of renewable scenario for the UK has found, that, apart from the required installation of certain over-generation, it is the medium duration energy storage in the range of multiple hours to days, through which the majority of the stored electricity needs to flow [6,7]. The lifetime of electrochemical batteries, typically below 10 years, furthermore stresses the need to search for other solutions [8].

**Citation:** Novotny, V.; Basta, V.; Smola, P.; Spale, J. Review of Carnot Battery Technology Commercial Development. *Energies* **2022**, *15*, 647. https://doi.org/10.3390/en15020647

Academic Editors: Alon Kuperman and Alessandro Lampasi

Received: 7 December 2021 Accepted: 11 January 2022 Published: 17 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Currently, pumped hydro energy storage (PHES) largely dominates the installed storage capacity in comparison to other solutions, as we can see in Figure 1. Even though a rapid growth is experienced in electrochemical batteries and is also expected across other technologies [9], the logarithmic scale provides an idea of the overall storage power demand when the total installed capacity will need to increase several fold to accommodate the renewable production variation. The values are compiled from the database [10] with further added Carnot battery projects known to the authors, with respect to the year 2020. It might not be exhaustive, especially in case of the electro-chemical batteries, though it provides a good overview of the situation and the scale of storage for future energy systems. One can also observe that some systems classified as Carnot batteries are already operational, though only in the megawatt scale.

**Figure 1.** Overview of global installed grid scale electricity storage systems power rating in 2020.

The PHES is also the most commonly employed large scale storage (>100 MW) for medium to long durations. The PHES has major advantages such as high roundtrip efficiency, fast response time, long duration of operation and low self-discharging effect. However, the PHES suffers from the requirements of a suitable geographical location, impact on the environment and low energy density [11,12]. Compressed air energy storage (CAES) technology utilizes mostly underground caverns for storing large volumes of compressed air. Together with PHES it is therefore dependent on geographically suitable locations [13]. Nowadays, only two commercial large scale CEAS facilities are operating, Huntorf in Germany and McIntosh in Alabama, USA, having installed a capacity of 290 MW and 110 MW and reaching an efficiency of 42% and 54% respectively [14,15]. Both commercial CAES plants come under the first stage of development, but advanced CAES systems have been developed such as adiabatic-CAES (A-CAES), advanced adiabatic-CAES (AA-CAES) and isothermal-CAES (I-CAES), which aim for higher efficiency by being more sophisticated and complex [16]. AA-CAES has been technologically experimentally proven on the MW scale [17] and current development and construction plans are given for up to 2.3 GW and 28 GWh in the coming years [17,18]. Flow electrochemical batteries aim to eliminate some drawbacks of classical batteries, especially in capacity scaling, while retaining their advantages. Owing to notably lower efficiency, low energy density, degradation, still requiring toxic and scarce materials and technical flaws and shortcomings, there is a lot of research needed for actual widespread application [19]. Gravity storage systems are

also either limited by geographical location or capacity. An increasing number of systems are progressing from conceptual to pilot and commercial stages [20,21]. Conversion to hydrogen and other synthetic fuels remains costly with very low efficiency and suitable for rather very long duration to seasonal applications [22].

#### *1.1. Carnot Battery Principles*

Carnot batteries (CB) comprise a set of multiple technologies which have a common underlying principle of converting the electricity to thermal exergy, storing it in thermal energy storage (TES) systems, and in a time of need converting the heat back to electricity. Based on this principle, alternative terms are also used as power to heat to power (P2H2P) or electric thermal (or electro-thermal) energy (electricity) storage (ETES). An excellent review work [23] provides a general overview of CB principles and therefore the reader is referred to this work for details. Prospects of PTES system are then provided in [24]. Here the general aspects will be therefore summarized rather briefly. A general principle of the CB is illustrated in Figure 2.

**Figure 2.** General principle of Carnot battery systems.

Carnot Batteries use surplus electricity as an input of a power to heat (P2H) system to create a temperature gradient (thermal exergy). It can have a form of hot and cold storage systems, or just one of those (hot or cold) with the temperature gradient defined against the environment. During the discharging process, the thermal exergy is converted back to work (electricity) by heat to power (H2P) system, in principle a heat engine. Various concepts of CB can be illustrated regarding the P2H and H2P conversion processes and thermal integration of the heat source in Figure 3.

**Figure 3.** CB concepts regarding thermal integration of heat sources and conversion systems. (**a**) direct heat to power conversion, (**b**) reversible thermodynamic cycle, (**c**) with heat source integration and hot storage or (**d**) cold storage.

The simplest concept is the direct conversion of electricity to heat, which is then stored before its conversion back to power during discharging, typically by a power cycle. Other systems can be considered so called compressed heat energy storage (CHEST) or pumped thermal energy storage (PTES) as they utilize the thermodynamic cycle (in principle a heat pump) for the P2H conversion. The first PTES in Figure 3b works mostly between two distinct temperature levels arbitrarily chosen and (theoretically) independent of the environment, having separate hot and cold storage systems. A specific aspect of CB is the possibility of thermal integration, both on the side of energy input as well as output, providing many possibilities for sector coupling. Regarding the heat input, the heat source can be either upgraded to a higher temperature, which is then used as a heat input of the power cycle or the charging system can prepare cold, which is stored and subsequently used as a heat sink of the power cycle during discharging, increasing the overall temperature gradient of the heat source [25]. A specific case can be defined when the Tsource is identical to the environment. Regarding hot storage, such systems are not really considered due to their low roundtrip efficiency. Regarding cold storage, it could be considered a highly simplified representation of liquid air energy storage, when the air is liquefied and stored at cryogenic temperatures. All real thermodynamic conversions depart from the ideal ones (e.g., minimum temperature differences in heat exchangers, efficiency of compressors and expanders, pressure drop in components). As a result, a portion of the heat needs to be rejected into the environment due to the irreversibilities [26]. The first concept (Figure 3a) minimizes the losses by converting and storing the heat at highest possible temperature, maximizing the power cycle efficiency. The second concept (Figure 3b) then optimizes the charging and discharging cycles to minimize the irreversibilities. In the heat source integrated concepts, efficiency is also a function of the temperature lift of the heat pump. With a very low lift, the roundtrip efficiency (defined below) can theoretically reach values above unity, in case of a zero lift (and work) of the heat pump, even going towards infinity.

As Dumont et al. [23] or Steinmann et al. [27] mention, there are many possible technological variations of CB. Charging can be realized besides direct conversion (joule heating) by any thermodynamic heat pump cycle. Discharging offers a similar range of options with heat engines to those of Brayton, Rankine or Stirling cycle, their combinations and also direct conversion as thermoelectric, thermi-ionic and thermophotovoltaic systems.

The first examples of these technologies can be traced to 1924, when Fritz Marguerre patented his own solution of thermal energy storage [28] or even to 1833 to work of Erricsson [23], but it has not been until the recent decade, when high volatility of electricity production and its mismatch with demand, it attracted wider interest in this technology. It typically provides relatively low efficiency in the range of 30% to 70% but also low cost for medium and long duration electricity storage. The widely increasing interest in CB was also a reason for establishing an IEA Task 36 on Carnot Batteries [29] with an aim of providing and unifying clear definitions, key performance indicators and classification of CB. Results of a short bibliographic study from Google Scholar using the main keywords of CB technologies shown in Figure 4 also confirms the trend of an increased interest in these technologies. At the same time, it shows that the unifying notation of these technologies as a Carnot battery is not yet in mainstream use.

**Figure 4.** Bibliographic study on a yearly number of scientific works with main CB keywords.

CB as any electricity storage system is specified by its roundtrip efficiency (RTE), which is defined as a ratio between electricity produced during discharging and electricity consumed during charging, see Equation (1).

$$\eta\_{RTE} = \frac{\text{electricity discharge}}{\text{electricity charged}} = \frac{\mathcal{W}\_{\text{e,discharge}}}{\mathcal{W}\_{\text{e,charge}}}.\tag{1}$$

Alternatively, useful energy efficiency (also referred to as total efficiency) can be defined by Equation (2) as total useful energy output in the case of sector coupling, especially also providing heat and/or eventually cold. This heat can be used from the CB system in the charging phase, separately drawn from storage, but the highest efficiency is obtained if it is a by-product or rejected heat (possibly also its part) from the discharging phase.

$$
\eta\_{IL} = \frac{W\_{c,\text{discharge}} + Q\_{useful}}{W\_{c,\text{charge}}}.\tag{2}
$$

Note that no heat input is considered even though in some CB concepts it is present. It can be argued that other formulations of efficiency may be also used. From thermodynamic standpoint, exergy efficiency explains the loss of potential and quality of all inputs in the best manner. This work however aims rather at an overview of CB technologies with respect to their prospective application and progress of scientific findings towards much needed commercialization. The heat input is moreover mostly considered as a low or zero cost input, typically in the case of waste heat with no alternative utilization.

#### *1.2. Purpose of the Review*

Throughout the increasingly extensive scientific research of the Carnot Battery technologies, commercial development is present as well. Examples of the main technologies were provided in the CB review [23] along with nine identified prototypes built between 2011 and 2020. Only some companies are publishing their findings and information about the CB systems directly in scientific publications as in [30,31]. A comprehensive summary of the CB technologies under commercial development is however missing. The present work addresses this gap, while the commercially developed technologies and presented parameters are also put into context with the approach and results in scientific publications. This provides both qualitative and quantitative views of this field of technology as a certain approach shared by multiple companies can be identified in some cases, while a certain approach is not, on the other hand, explored in the scientific literature. As a comprehensive technology review, it addresses the needs of both academics and industry practitioners, while it might in some cases bridge between these two, especially when pointing out technologies that are insufficiently addressed in scientific research or, vice versa, highly academically studied concepts with very limited industrial development.

#### *1.3. Classification of CB Technologies*

As was shown above, many classifications of CB concepts, principles and used technologies are available. The method of charging, discharging, thermal energy storage technology or the conversion system used can be named as examples. In this manuscript, classification by the discharging system is used, specifically as the Rankine cycle systems, Brayton cycle systems and other & hybrid systems. This is in accordance with the categories in the IEA Task 36 on Carnot Batteries [29]. Within each of these groups, further distinction is made between the systems with direct P2H conversion and with the PTES systems using a heat pump principle for charging.

Additionally, there are TES systems under commercial development, which are considered specifically for the CB application, mostly for direct P2H conversion and standard power cycle (Rankine or Brayton) technology. These specific technologies are therefore included as a separate section of this paper.

Within the list of the experimental systems, there are several levels reported. The first one is a proof of concept, largely scaled down system for demonstrating technical feasibility, often with poor roundtrip efficiency or other parameters. A demonstrator is then typically at a larger scale and better parameters but works at the manufacturer site with limited benefit of the operation or operated just for purposes of technical tests. A pilot is already installed on site and providing its designed services to its customer, though installation is largely or fully subsidized. Note, however, that this difference does not need to always be very clear in real systems. The last level is finally a fully commercial system.
