compared to the pre-industrial level, implied by the Paris agreement [1], economies should reach net-zero greenhouse gas emissions by 2050 [2]. Recent studies have shown the technical feasibility and paved pathways for achieving net-zero greenhouse gas emissions goal in this time frame [3], [4], [5], [6].

Photovoltaic (PV) solar cells are the most widespread means to harness solar energy, due to their low cost [7], and are expected to remain an important factor in the transition to a low-carbon economy. PV cells convert solar radiation directly to electricity and thus are extremely valuable for supplying demand when solar radiation is sufficient. However, since PV electricity generation and demand temporal profiles do not match well, overgeneration from PV cells may be significant in power systems with high solar penetration levels, depending on the grid’s flexibility to absorb this surplus electricity. One of the most common methods to overcome the overgeneration obstacle when PV penetration is low is by curtailment, i.e., dumping energy that cannot be consumed, which may become unacceptably expensive in power systems with high solar penetration levels. Energy storage systems such as battery energy storage systems (BESS) provide another method to overcome overgeneration. As utility-scale BESS prices had plummeted significantly in recent years [8], such systems are now used as flexible means to maintain power balance within power grids and to provide other grid services [9]. However, enabling high solar penetration levels using energy storage systems is still an expensive solution [10], [11].

In addition to solar cells, Concentrated Solar Power (CSP) plants, such as parabolic troughs and solar power tower plants, may be used to harness solar energy [12]. In contrast to PV cells, these technologies convert solar radiation to heat, which is used to generate electricity by a power block. The heat can be stored efficiently and used when needed in thermal energy storage (TES) tanks. Between these two technologies, solar power tower technology has the advantage of high working temperature as higher solar concentration can be achieved, which enables a higher efficiency Rankine cycle [13], [14]. In these plants, solar energy is concentrated using a field of mirrors (heliostats) to a receiver which is held on top of a central tower. The high energy concentration (500–1000 suns) heats a heat transfer fluid to a high temperature up to 565

, which is used by a steam turbine to generate electricity. To further increase efficiency, recent studies have suggested boosting working temperatures even higher with particle-driven CSP receivers [12], [15], [16]. High-efficiency TES may allow 15 hours or more of full-capacity operation, which enables 24 hours of energy generation [17].

Numerous hybrid approaches of CSP and PV have been suggested to leverage the advantages of each, namely the low cost of PV cells with the dispatchability that is offered by CSP plants [18]. Particularly, a configuration in which electric heaters are used to store PV surplus electricity in TES has been suggested as a readily available and feasible solution [19]. One of the core advantages of such technology is that the PV cells and TES can be operated independently in different locations, meaning that already existing PV cells can be used.

In addition, recent works on solar-aided hydrogen production give rise to the production of hydrogen with CSP heat and electricity [20], [21], [22], [23]. This is a promising opportunity to support high levels of solar penetration levels by storing energy for months [24].

One of the main barriers to CSP deployment is its high costs [25], which are commonly analyzed using the levelized cost of electricity (LCOE) merit. Although CSP LCOE is much higher than for PV cells, CSP with TES should be compared to PV cells combined with electricity storage to compare similar functionalities. As such, recent studies have compared CSP with TES to PV with BESS or TES, and inferred that the LCOE of CSP with TES is lower than PV plus BESS for storage duration greater than 4–6 hours [26], [27], [28]. This implies that for power systems with high PV penetration levels CSP with TES would be more economically viable than PV cells with BESS.

Nevertheless, the LCOE calculation captures the total produced electricity, and thus, does not reflect any electricity curtailment. Since high curtailment rates have been shown to be economic in power systems with high solar penetration levels [29], the LCOE merit is not suitable to determine the most cost-effective solar-based solution in such scenarios. To increase solar penetration level, it may be cheaper to incorporate some curtailment instead of installing expensive storage solutions, particularly for PV with BESS technology, in which PV cells are much cheaper than BESS per delivered power. On the other hand, when solar penetration level is high and excess electricity is abundant, energy storage may become the dominant factor in the collective cost of electricity generation and storage favoring the use of low-cost TES. These factors are not well represented by the conventional LCOE merit. Furthermore, these factors change for any given level of solar penetration; thus, there is no fixed value for a wide range of solar penetration. Showing the possible error of the LCOE measure, a recent study [30] evaluated the LCOE sensitivity to the capacity factor, a measure of the annual number of hours a technology has been used, showing that this sensitivity is high. For example, the corrected LCOE when considering the actual capacity factor value for a coal power plant during 2020 is

4 times higher than the anticipated conventional LCOE.

To overcome the flaws of the LCOE measure, other metrics have been suggested, including levelized avoided cost of electricity (LACE) [31], system profitability [32], cost of valued energy (COVE) [33], and system LCOE [34], [35], [36], [37]. In addition to a more thorough analysis of the total costs, these metrics incorporate expected revenues, which are dependent on the generation mix, fuel prices, and energy policies, and therefore, are complex and case-specific [32]. In contrast, another metric, the net-LCOE, similar to the conventional LCOE does not include revenues. It introduces a simple improvement compared to conventional LCOE by considering the curtailment caused by overgeneration and is calculated by net-LCOE