In recent years, developing a clean and low-carbon energy system has become an increasing challenge due to climate change [1]. Renewable energy sources are playing an increasingly important role in the global energy system [2]. According to the policy, China will build wind and photovoltaic power generation bases with a total installed capacity of about 455 GW in desert areas by 2030 [3,4].

However, solar and wind resources are intermittent and unstable [5,6]. The uncontrollable photovoltaic (PV) and wind power leads to increasing demands on the grid regulation capacity, and can even lead to security and stability issues with the rapid growth of installed renewable energy sources. In such cases, renewable power over power demand must be discarded, while coal-fired power plants are often required to provide baseload and regulation capacity, severely limiting the deployment of large-scale renewable energy systems and resulting in a low share of low-carbon power in the electricity supply. As another promising renewable technology, Concentrating Solar Power (CSP) plants can provide stable and dispatchable electricity when combined with a thermal energy storage (TES) system [7,8]. On the other hand, the stand-alone CSP plants are hampered by high investment costs [9]. The combination of CSP with PV and the wind is then considered given their respective advantages in building a low-carbon, reliable and economic power system [10,11].

Researchers typically consider the integration of PV and CSP plants to achieve dispatchable solar power output [12,13]. To identify the advantages and limitations of the integrated PV-CSP system, scholars have simulated the typical operating scenarios of the integrated system based on several simulation platforms. Then the operating characteristics of the integrated PV-CSP system are compared with those of CSP alone and PV alone under the same conditions to demonstrate the potential and feasibility of the integrated system. To preliminarily assess the dispatchability of a PV-CSP system, Petrollese et al. [14] used a simplified power demand curve to observe the performance of the hybrid system based on linear Fresnel CSP technology, an Organic Rankine Cycle (ORC) power plant and PV panels. Results show that the PV + battery combination performs better in meeting the constant power demand within 8 h, while the hybrid PV-CSP system is a better solution when the power demand lasts longer than 16 h. Ghirardi et al. [15] used the TRANSYS platform to investigate different combinations of PV panels, parabolic trough collectors and central receiver technology by varying the installed capacity of each component. As a result, the CSP plant can meet the residue demand effectively on the basis of a PV plant, achieving a capacity factor of 60 %–85 %. Pan et al. [16] investigated the possibility of combining a CSP central receiver plant with a PV power plant to provide 100 MW baseload capacity to the grid. Simulation of different combinations show that by combining with a PV power plant, a smaller capacity CSP plant can achieve a similar or even higher capacity with a capacity factor of up to 90 %, with a levelized cost of electricity (LCOE) of 0.133 $/kWh-0.157 $/kWh. Behar et al. [17] examined 12 cases of integrating solar power technology (PV, CSP, hybrid PV-CSP) into copper concentrate plants (CCP), providing a practical solution for delivering dispatchable clean energy to the copper mining industry.

Considering the influences of the combination schemes on the technical-economic performance of the system, scholars have explored the optimal configuration of the integrated PV-CSP system through parametric analysis and optimisation algorithms. Starke et al. [18] performed a parametric analysis of a hybrid system based on the TRNSYS platform to obtain the optimal storage and power block size to achieve the minimum LCOE under the varying PV capacity. Bousselamti et al. [19] investigated the effects of design parameters such as tilt angle, solar multiple, TES capacity and hybridisation fraction on the cost and power generation of PV-CSP systems based on a typical meteorological year (TMY) condition and constant baseload curve. Then a multi-objective optimisation is conducted [20] using a genetic algorithm to minimise the LCOE, maximise the capacity factor and minimise the dumped energy to meet the baseload demand, where PV capacity, CSP capacity, solar multiple and thermal storage size are selected as decision variables. The results indicated that the selection of dispatch strategy is strongly related to the optimal PV-CSP configuration, and it is also assumed that an electrical storage system can be configured to further minimise the dumped energy from the PV system. Operational strategies are also focused on investigating the prioritisation of component outputs in different scenarios. By conducting an 8760-h operational simulation based on a 100 MW solar tower power plant using the Smart Dispatch program, Green et al. [21] found the integrated system can achieve a high capacity factor of up to 90 % under the dispatching strategy where CSP is dispatched based on PV generation. A similar conclusion was obtained by Cocco et al. [22], in which the operation of the PV-CSP system under two different strategies is compared based on the design data of the Ottana solar plant and a constant power demand curve.

It should be noted that there is no energy exchange between the PV plant and the CSP plant in the above integration concept, although the two systems work in cooperation to meet the power demand. To further improve the dispatchability of a hybrid system and reduce wasted solar energy, Zhai and Liu et al. [23,24] have previously proposed to configure electric heaters (EH) in the PV-CSP system to convert excess uncontrollable PV power into thermal energy for CSP power generation instead of wasting it. The proposed system is shown to further reduce the power drop and improve the performance at a lower cost, with the annual power output increased by 6.52 % and the capacity factor increased by 4.85 % compared to the conventional PV-CSP system to achieve a constant power curve.

On this basis, scholars have conducted a wide range of studies on PV-CSP systems configured with electric heaters. To further demonstrate the improvement of the proposed system, Sumayli et al. [25] compared the parabolic trough (PT) solar thermal power plant, combined PT-PV with a stand-alone operation strategy, and combined PT-PV with electric heaters. The hybrid concept with electric heaters obtains a 7–18 % reduction in LCOE while maintaining a capacity factor of 79 % in two cities in Saudi Arabia. Giaconia et al. [26] performed an annual operation simulation of a PV-CSP system with electric heaters. The results show that the system can meet up to 70 % of the power demand in cases with large seasonal variations in solar irradiation and demand, and more than 90 % of the power demand in industrial areas with flatter irradiation and power demand. Guccione et al. [27] also indicated that the hybridisation of PV and CSP plants with electric heaters leads to a 14 % reduction in LCOE and that the technical-economic performance is most affected in small scale and low DNI/GHI conditions.

In aspects of design parameter optimisation, Chennaif et al. [28] use the algorithm based on Power Pinch Analysis and Modified Electric System Cascade Analysis to investigate the optimal combination consisting of photovoltaic panels, wind turbine generators and a concentrated solar power system. The case study for an electrical demand of 50 MW shows that the hybrid system can achieve 0 % power shortage probability with an LCOE of $0.18/kWh. Pilotti et al. [29] established a Mixed Integer Linear Program (MILP) model to investigate the optimal design of an integrated CSP-PV power plant. By comparing CPS + thermal storage, PV + battery, hybrid CSP-PV solutions with and without electric heaters, the CSP-PV plant configured with electric heaters is shown to achieve a 3.6–10 % lower LCOE. Based on the proposal to share the TES system between CSP and other power plants, Liu et al. [30] investigated the influence of cost reduction of battery and TES in an integrated PV-CSP system. The results show that when the battery cost is reduced to 160$/kWh, the battery can be scaled up incrementally in an optimal configuration. Yang et al. [31,32] optimised the design of various renewable energy power plants and found that PV alone is the most cost-effective solution for power generation accompanied with lower reliability. The integration of batteries can effectively improve the reliability of the power system within certain limits and the integration of CSP plants is an effective way to further improve the reliability of the system. In terms of operation optimisation, Richter et al. [33] used model predictive control and mixed integer programming to optimise the energy storage strategy and, on this basis, to optimise the scale of the system design. Considering future energy tariffs and meteorological conditions, this real-time predictive control strategy resulted in a 14.2 % increase in annual revenue compared to the non-predictive strategy. In addition, the optimal system configuration was found to be strongly influenced by the energy storage control strategy.

These studies have demonstrated that the integration with CSP and the sharing of TES through electric heaters contribute to the operational performance of the hybrid systems, providing valuable references on aspects of configuration and operation. However, there are still some problems: (1) There is a discrepancy between the operating conditions studied and the practical applications. A single application scenario is studied and the variation of the target is not considered in most studies, although several operating strategies could be performed. In addition, a simplified or constant power demand is usually used instead of the real demand curves. (2) Most of the studies have been conducted under given conditions and the influence of uncertainties such as power demand and meteorological conditions on the applicability of the system has not been discussed.

In this paper, three different operation modes are proposed and compared according to the practical applications of the hybrid system consisting of one or more CSP power plant components configured with TES, PV panels, wind turbines, batteries and electric heaters. Then, the capacity parameters of the subsystems, including the size of the PV plant, the wind farm, the capacity of the corresponding batteries and the TES, are optimised in three operating modes under two typical meteorological conditions. Finally, optimisation based on imperfectly predicted meteorological conditions and different operating conditions including power demand and transmission line capacity is performed and discussed. The main research contributions and scientific values of this study are: (1) Three different operation modes, including local consumption mode, sales mode and virtual power plant (VPP) mode, are proposed to combine the configuration and operation with practical applications, considering the conditions of real power demand and electricity prices. (2) A comprehensive understanding of the optimal combination and coordination between the CSP plant with shared TES, PV panels and wind turbines is obtained, considering the imperfectly predicted meteorological conditions and various operating conditions, which provides a reference for the planning and design of large-scale hybrid solar and wind power plants under specific meteorological conditions.