Optimal capacity allocation and scheduling strategy for CSP+PV hybrid standalone power plants

Most islands are located in remote areas. Due to inconvenient transportation, lack of energy infrastructure, and small economic scale, island regions often find it difficult to obtain stable and sustainable energy supplies, leading to multiple challenges in energy provision. On the other hand, many islands are located in tropical or subtropical regions, where there is abundant sunlight throughout the year and high solar radiation intensity, making them suitable for solar power generation. Additionally, islands are typically situated in marine environments with strong and stable winds, favorable for wind energy utilization. Therefore, renewable energy generation presents a promising solution to the problems above. Recently, reports have indicated that the newly installed capacity of renewable energy in the electricity sector has surpassed that of traditional energy sources. According to statistics, the global cumulative installed capacity of solar power has reached 407 GW.^1,2 Among them, off-grid solar power generation is expected to achieve self-generation and self-consumption of electricity.^3–6 Solar power generation can not only mitigate the environmental impact caused by the burning of fossil fuels but also increase the independence of energy supply in remote regions,^7,8 for example, in remote regions such as pelagic islands and deserts.

Solar power generation mainly includes photovoltaic (PV), concentrating solar power (CSP), Concentrated Photovoltaic Thermal (CPVT), and High Concentration Photovoltaic Thermal (HCPVT). Currently, CPVT and HCPVT technologies are relatively new in the market. They have lower market acceptance and maturity. These systems involve complex design and installation processes, and their installation and maintenance costs are higher. Their application scenarios are relatively limited, primarily focusing on combined heat and power systems for buildings. They are less suitable for large-scale power generation projects and are not well-suited for island environments.

Photovoltaic (PV) and concentrating solar power (CSP) are two solar power generation technologies with relatively high maturity levels. PV generation absorbs visible light through solar panels, utilizing the photovoltaic effect to convert solar radiation into electricity. It has such advantages as low cost and high commercialization.⁹ However, individual PV power plants are subject to intermittent changes in sunlight, resulting in significant fluctuations in power output.^10,11 This can lead to a mismatch between supply and demand in the island power grid, thereby affecting the utilization rate of PV power plants CSP generation utilizes large-scale mirror collection of solar thermal energy, heating steam through heat exchange devices, and driving traditional steam turbines for power generation.¹² Due to the presence of heat exchange processes in CSP plants, they inherently have the conditions for integration with thermal storage systems.^13,14 Therefore, by using thermal storage units, their output characteristics can be improved, and power output can be smoothed. In this case, CSP plants become dispatchable. However, the technique of CSP generation emerges relatively late and is still during the early stages of commercialization. The investment cost per kilowatt is 2–3 times that of photovoltaics, leading to a lower level of popularity for standalone CSP power plants.^15,16

At present, solutions to address the limitations of individual solar power generation primarily fall into two categories.^17,18 One is the development of hybrid renewable energy systems that integrate multiple types of generation equipment. For example, (1) wind–solar complementary power generation system, which includes PV (Photovoltaic) and wind power generation. Although wind and solar energy complement each other in time (strong winds at night, strong sunlight during the day), which can extend the generation hours and ensure stable grid operation, maximizing economic benefits.^19,20 However, the entire system still exhibits strong intermittency and large fluctuations, leading to grid supply-demand imbalance and increased peak shaving requirements, which are not conducive to the economic operation of the grid. (2). Power-to-gas (P2G) technology, which converts electricity into gas fuel. Although P2G technology can convert intermittent renewable energy sources (such as wind and solar energy) into stable gas fuels, the overall energy conversion efficiency is low. Particularly in the process of converting electricity into hydrogen and then into methane, there are significant energy losses. Additionally, storage and transportation of hydrogen and methane require specialized infrastructure, which is lacking in island regions, limiting the widespread adoption of P2G technology on islands.

Another approach is to equip renewable energy power generation with energy storage systems, using storage batteries to perform real-time charging and discharging to enhance the output characteristics of PV stations.^21,22 However, energy storage batteries are costly, have a short lifecycle, and incur high recycling and disposal costs, posing environmental pollution risks. These factors similarly limit the widespread application of this approach.

The CSP+PV hybrid solar power station, which integrates both CSP and PV technologies, can be seen as a combination of the two aforementioned approaches. It incorporates both renewable energy generation units, thereby enhancing energy utilization efficiency, and utilizes the CSP's thermal storage system instead of energy storage batteries to smooth out fluctuations and stabilize power output.^23,24 During the coordinated operation of CSP and PV, PV generation is primarily utilized during the day, with CSP reinforcement during periods of insufficient PV generation. At night, CSP converts the stored thermal energy into electricity, ensuring the stable operation of the hybrid station. This approach fully leverages the dispatchability of CSP and the low cost of PV, not only mitigating the intermittency of renewable energy generation and promoting effective supply-demand matching^25,26 but also further reducing the levelized cost of electricity of solar power stations (Levelized Cost of Energy, LCOE).

Reference 27 suggests that the combination of CSP and PV can meet the load demands over some time, thereby enhancing power supply reliability. Therefore, numerous scholars have researched the hybrid configuration schemes of CSP and PV. Reference 28 analyzed the economics of CSP+PV hybrid power plants, and the results indicate a significant reduction in LCOE for CSP+PV systems, which is lower than that of standalone CSP power plants. In the CSP+PV systems, one of the most important cost factors of a solar tracker is the amount of material used for the support structure. Reference 29 proposes a new method for geometric trackers based on linear optimization, aimed at improving the economic efficiency of solar power plants. This approach reduces the costs of CPV and PV trackers, as well as heliostat support structures, making it a powerful tool for optimizing and executing preliminary designs of CSP and PV systems. The CPS+PV system designed in Ref. 30 has been simulated in Australia and Spain, yielding relatively ideal results. The system has a peak power of 10 MW CSP and 15 MW PV, and it can provide more than 60% (62% and 68%, respectively) of the actual power load with minimal curtailment. The system also innovatively designed a CSP fluidized bed receiver capable of directly collecting and storing thermal particles and established a CSP dual-tank storage system that can generate electricity at night. This system possesses stable and controllable energy collection and conversion capabilities. For islands with relatively low power load, Ref. 31 indicates that CSP+PV systems are more suitable for small-scale power generation. The results show that hybridizing PV with state-of-the-art CSP can lead up to a 22% reduction in the Levelized Cost of Electricity (LCOE) compared to standalone CSP systems. A hybrid CSP+PV system costs approximately 25% less than a CSP-only system of the same scale. Additionally, the proportion of CSP and PV in hybrid solar power plants can be optimized further with the objectives of maximizing LCOE and minimizing the energy deficit rate. The above studies focus on commercial hybrid solar power plants tailored to meet the constant base load demand of large power grids, without considering the impact of CSP+PV plants on the operation of other units within the grid. Regarding configuring CSP+PV hybrid power plants for isolated grids, Ref. 32 designed a CSP+PV hybrid power plant to meet the actual load demand of the island power grid but did not utilize thermal power/diesel generators as backup energy sources. As a result, although the thermal storage system of CSP enables the hybrid solar power plant to be dispatchable, relying solely on renewable energy for grid supply results in lower reliability and economic viability. On the one hand, this may lead to redundant configurations of hybrid solar power plants, increasing investment costs and prolonging payback periods.³³ On the other hand, in the case of prolonged cloudy weather, the generation capacity of CSP+PV hybrid power plants may not be sufficient to meet load demands or may even lead to intermittent operation. Therefore, fossil fuel-based backup units remain indispensable for independent power grids at the current stage.

Recent practices have shown that the participation of renewable energy generation in grid dispatch and their coordination with traditional thermal power/diesel generator units has become the trend of new energy development.^34–36 To establish a stable and economically viable infrastructure for CSP+PV stations operating in independent power grids, this paper focuses on analyzing the operational characteristics of CSP+PV and establishes a hybrid solar power station model combining CSP and PV technologies. Given the realistic possibility of integrated planning and scheduling coordination in island power grids, to achieve the unity of long-term and short-term benefits of the island power grid in a refined manner, a capacity configuration method for CSP+PV hybrid solar power station coordinated with thermal power units based on island power grid dispatching is proposed. Under the premise of meeting the requirements of the island power grid, the optimal configuration of the CSP+PV hybrid solar power station is obtained by traversing to minimize the system's equivalent annual cost. Simulation results validate the feasibility and effectiveness of the proposed configuration method.

The main structure of the CSP+PV hybrid solar power station operation is shown in Fig. 1. The hybrid solar power station consists of two types of power generation equipment: CSP and PV. The CSP part comprises three subsystems: the solar field (SF), the thermal energy storage system (TES), and the power block (PB). In this paper, the solar field (SF) consists of tower-type solar collectors and molten salt is used as the heat-transferring fluid. Since the heat-transferring fluid is consistent with the thermal storage medium (both are molten salt), the thermal energy storage (TES) part adopts a direct two-tank thermal storage system. Molten salt at 288 °C from the cold storage tank is pumped to the solar field, heated to 565 °C, and then stored in the hot storage tank. The power block (PB) consists of a thermal cycle system and a turbine generator unit. The heat from the thermal storage system is transferred through the cycle system to generate steam, which drives the turbine generator unit for electricity generation. The PV part mainly comprises solar cell arrays and inverters, which convert solar radiation into direct current (DC) and then into alternating current (AC).

Compared to standalone PV power plants, CSP+PV hybrid solar power plants are not only structurally more complicated, but the scheduling strategy also requires further exploration. It is necessary to consider not only the scheduling between CSP and PV within the plant but also the coordinated operation between CSP+PV plants and other power generation units within the island power grid. It should be noted that, taking into account the relatively high construction costs in CSP+PV plants, planning and configuring CSP+PV plants based on coordinated scheduling strategies are particularly important.

The existing planning and configuration of CSP+PV hybrid power stations mainly focus on the coordinated optimization between CSP and PV within the power station. The goal is to achieve low LCOE and high capacity factor while meeting the base load demand. Due to the presence of thermal energy storage systems in CSP plants, they have strong schedulable characteristics. Therefore, to guarantee the constant output power from the perspective of the hybrid plant, the generating characteristic of CSP needs to respond in real time to the output of PV. As shown in Fig. 2, during peak periods of PV generation, CSP will reduce its output accordingly. Conversely, during periods of low PV generation, CSP will increase its output or even reach full capacity to compensate the gap between PV generation and the base load demand. The coordination between CSP and PV enables the hybrid solar power plant to achieve a constant power output of 1.0p.u. during continuous load periods.

The above strategy only involves the coordinated scheduling between CSP and PV within the hybrid power plant, without considering the impact of load curve variations on the plant's output. It is characterized by its simplicity, stable output, and high penetration of renewable energy sources. However, for island power grids with significant fluctuations in load throughout the day, the constant output of CSP+PV stations during their generation periods can substantially alter the equivalent load of the microgrid, resulting in a reduced equivalent load that often falls below the minimum threshold required for the economic operation of thermal power plants. This leads to thermal power plants continuously operating in uneconomical regions, resulting in higher peak shaving cost.³⁷ Therefore, while this strategy simplifies the optimization scheduling of the hybrid power plant itself, it is not conducive to the economical operation of other units within the island power grid. Instead, it may increase the overall operational costs of the system. To address the aforementioned issues and determine the optimal configuration of CSP+PV power plants, this paper first establishes a power model for CSP+PV plants. Ensuring the economical operation of thermal power units, the coordinated optimization scheduling of CSP+PV plants and thermal power units is conducted. Furthermore, a method for the capacity configuration of CSP+PV plants is proposed.

1. The real-time power balance of CSP is given by

   $PPBt−ηdPth,dt+Pth,ct/ηc=Psolart.$

   (1)

   Among which, $PPBt$ represents the thermal power flowing into the PB; $Pth,ct$ and $Pth,dt$ represent the charging and discharging thermal power of the TES, respectively; $ηc$ and $ηd$ represent the charging and discharging efficiency of the TES, respectively; $Psolart$ represents the thermal power from the SF.

2. SF Part

   The thermal power received by SF is equal to the product of $DNIt$⁠, $ηSF$⁠, the photothermal conversion efficiency of the SF, and $ACSP$ the SF area, as given by

   $Psolart=DNIt·ηSF·ACSP−PCSP,Sur,$

   (2)

   where $PCSP,Sur$ represents the reduction in thermal power of the SF.

3. TES Component

   The energy stored $Etht$ at time t of TES system is determined by the energy stored at the previous time $Etht−1$ and the charging/releasing power and is constrained by the capacity of TES and the minimum energy stored (to prevent the solidification of molten salt), as shown in expressions of below equations:

   $Etht=(1−γ)Etht−1+(Pth,ct−1−Pth,dt−1)Δt,$

   (3)

   $Eminth≤Etht≤Emaxth,$

   (4)

   where $Emaxth$ and $Eminth$ are the maximum and minimum energy storage of the TES system, respectively. $γ$ is the dissipation coefficient of TES, and $Δt$ is the interval.

   Among them, the charging/heat release power is limited by the maximum charging/heat release power of the TES system, and the charging/heat release cannot be carried out simultaneously:

   $0≤Pth,ct≤bctPmaxth,c,$

   (5)

   $0≤Pth,dt≤bdtPmaxth,d,$

   (6)

   $bct+bdt≤1.$

   (7)

   Among which, $bct$⁠, $bdt$ are the charging and releasing states of TES system, respectively, and 1 indicates that they are in the charging state.

4. PB Component

   $Pet/ηe+ uCSPtPSUth=PPBt.$

   (8)

   Among which, $Pet$ is PB generation power; $ηe$ is the thermoelectric conversion efficiency of PB; $uCSPt$ is the on/off variable of PB, and 1 indicates that the unit starts at time t.

5. PV Component

   $PPVt=min{ηinv(GHIt·ηPV·APV−PPV,sur),PPVmax}.$

   (9)

   Among which, $ηinv$ is the AC–DC conversion efficiency of the converter, and $ηPV$ is the photoelectric conversion efficiency of the PV module. $GHIt$ is the total solar radiation; $APV$ is the PV module area; $PPV,sur$ represents the power reduction of PV; and $PPVmax$ indicates the rated power of PV.

The optimization variables of CSP+PV power station and thermal power unit coordination scheduling include the on–off state of thermal power unit, the real-time power of the unit, the state variable of TES, the charging/heat release power of TES, the generating power and heat production power of CSP, and the generating power of PV.

Among these, $γ$ is the collection of thermal power units, $f(Put)$ is the operating cost of thermal power units, which is usually expressed by a quadratic function. To reduce the complexity of the solution, this function is replaced by a piecewise linear function. $Eut$ is the start-stop cost of thermal power units.

Among which, $PLt$ is the electrical load on the user side; $xit$ is the working state of the Class i unit, and 1 denotes the operating state. $TMinoni$ and $TMinoffi$ are the minimum operating and outage times of Class i units, respectively. $Pmaxi$ and $Pmini$ are the maximum/minimum output of Class i units, respectively. $RDi$ and $RUi$ are the maximum ramping-up and ramping-down capacities of Class i units, respectively.

Since both the objective function and constraint conditions of the above-integrated energy system coordinated scheduling model have been linearized, the mixed integer linear programming method is adopted in this paper to solve the above problems, and the objective function is used as the return value for the calculation of the capacity configuration of CSP+PV power station in Sec. III D.

The optimal capacity configuration of a CSP+PV power station can be regarded as the optimal set of the following variables—SF area, TES capacity, PB-rated power, and PV-rated power. At the same time, to facilitate the understanding and definition of the characteristics of CSP+PV power station, the SF area and TES capacity are represented by solar multiple (SM) pairs and heat storage time (StH), respectively. Where, SM is the ratio of SF output thermal power to PB rated thermal power, and StH is the multiple of heat storage capacity relative to PB rated thermal power.

In order to obtain the optimal capacity configuration of the CSP+PV power station, the lowest annual operating cost of thermal power units under each configuration (different numerical combinations of SM, StH, PB-rated power, and PV-rated power) is obtained through the coordinated scheduling model in Sec. III C, and the equivalent annual investment cost is further converted according to the initial investment of each configuration. The annual operating cost of thermal power units and the equivalent annual investment cost of CSP+PV power stations are added to obtain the equivalent annual cost of regional energy systems under different configurations. Finally, the combination of design variables with the lowest equivalent annual cost is selected as the optimal configuration of the CSP+PV power station.

At present, there are two categories of methods for the combinatorial optimization design of variables: Category 1, utilizing genetic algorithm and other appropriate intelligent algorithms to select the most suitable combination; Category 2, the limited number of design variable combinations are calculated one by one with ergodic method. The first category of the methods takes the risk of getting a local minimum value, while the second category of the methods is affected by the step size of the design variable, and too many variable combinations will lead to a large increase in calculation time. In addition, most of the existing CSP and PV-rated power are fixed integer values, the exact value obtained by method 1 is difficult to use directly. Therefore, to obtain the achievable design value of a CSP+PV power station and avoid the local minimum value, this paper adopts the ergodic method to calculate different configurations one by one and aims to obtain the optimal configuration of CSP+PV power station with the lowest equivalent annual cost of the system.

Based on the light data of a large-scale island grid, this paper will verify the effectiveness of the proposed capacity allocation method for CSP+PV power stations and further analyze the operation performance of CSP+PV hybrid solar power stations. The island power grid contains two thermal power units, and the parameters are shown in Table I. To achieve clean and efficient use of energy, a new solar power station is planned.

Since solar lighting fluctuates strongly with the season and time of day, the configuration of CSP+PV power stations needs to be based on local lighting data. However, for the CSP+PV power station capacity configuration method proposed in this paper, if all year of lighting data are used, the CSP+PV power station and thermal power unit under different configurations need to be coordinated for the whole year, which will greatly increase the calculation burden. Therefore, this paper selects a typical day in each of the 12 months to represent the lighting of the whole month. The direct solar radiation of the region is shown in Fig. 3, and the total solar radiation under the optimal PV installation Angle is shown in Fig. 4.

Similarly, due to the influence of seasonal temperature and sunshine duration, the power load of typical days of each month also varies greatly. According to the analysis of the power load characteristics of this region in Ref. 38, the power load of each typical day is shown in Fig. 5.

According to the relevant data from the International Energy Agency, the economic parameters of each component are shown in Table II. To verify the superiority of the CSP+PV power station and the effectiveness of the configuration method presented in this paper, according to the economic parameters of each component, the optimal design results of the following four schemes are obtained by ergodically searching all feasible schemes taking into account all combination of variables in Table III under the typical daily scenario of above 12 months.

• Scheme 1: Considering the coordinated scheduling of the PV power station and thermal power unit (with PB power, SM, and StH remaining at zero during the traversal), the optimal configuration of the PV power station is searched with the goal of the lowest total system operating cost.

• Scheme 2: Consider the coordinated dispatch of PV power plants, electrochemical battery power plants (EB), and thermal power units (with PB power, SM, and StH remaining at zero during the traversal) to find the optimal configuration of EB+PV power plants, aiming to minimize the total system operating cost.

• Scheme 3: Considering the coordinated scheduling of the CSP power station and thermal power unit (with PV power remaining at zero during the traversal), the optimal configuration of the CSP power station is searched with the goal of the lowest total operating cost of the system.

• Scheme 4: Considering the cooperative optimization between CSP and PV in the CSP+PV power station (keeping the output power constant for a certain period), the optimal configuration of the CSP+PV power station is searched with the lowest LCOE as the goal.

• Scheme 5: Considering the coordinated scheduling of the CSP+PV power station and thermal power unit, the optimal configuration of the CSP+PV power station is searched with the lowest total operating cost of the system as the goal.

In addition, to ensure that the five schemes are comparable, the total installed capacity of the four schemes is set to be equal. That is, the capacities of Schemes 1–5 are restricted to be equal to the sum of the optimal PV and CSP capacities obtained in Scheme 5.

Table IV shows the optimal configuration results and corresponding operating costs of the five schemes. Among them, Scheme 1 uses only PV power plants. This scheme has the lowest initial investment cost and the lowest LCOE, at only 0.06 USD/kWh. However, because PV power generation is not dispatchable, it has a greater impact on other power generation units in the island power grid, and the annual operating cost of thermal power units is correspondingly the highest. Scheme 3 only takes CSP power station into account, and the CSP with heat storage is dispatchable. The annual operating cost of thermal power units is greatly reduced, which is only 85.8% of Scheme 1, but the investment cost is 2.25 times of Scheme 1, resulting in the LCOE of Scheme 3 being the highest among the four configuration schemes. It can be seen that a single PV power station and a single CSP power station are difficult to achieve the balance between their own benefit and the overall benefit of the grid. For scheme 4 and Scheme 5 involving a CSP+PV hybrid solar power station, the LCOE of the CSP+PV hybrid configuration is significantly reduced compared with that of a single CSP power station, and the operating cost of thermal power units in the island power grid is also reduced compared with that of a single PV power station. It shows that CSP+PV power station can combine the schedulability of CSP with the low cost of PV to some extent. However, the configuration results of Scheme 4 and Scheme 5 are quite different, and the impact on the island power grid is different.

Specifically, in scheme 4, the rated power of the battery (PB) and photovoltaic (PV) systems are similar. This is because scheme 4 requires maintaining constant output over a certain period, and excessively low or high PV power would impose higher demands on the scheduling characteristics of the CSP component, leading to further increases in the solar field area and storage capacity. Nevertheless, despite this, compared to Scheme 5, the values of SM (solar field area) and StH (storage capacity) in scheme 4 remain relatively high, resulting in two main effects on the CSP+PV power plant. On the other hand, the high-rate light field area and long-term heat storage can increase the annual power generation of CSP+PV power stations to 162 000 MWH and ultimately reduce the LCOE of solar power stations to 0.10 USD/kWh. If only from the benefit of CSP+PV power station, scheme 4 has great advantages, that is, higher annual power generation and lower LCOE, which is also in line with the design purpose of scheme 4 in this paper, that is, finding the configuration of CSP+PV power station with the lowest LCOE.

Because scheme 5 needs to coordinate and optimize the scheduling of CSP+PV power stations and thermal power units, in order to expand the operation adjustment space of other units, the PV capacity configuration that cannot participate in coordinated scheduling is low, only 25% of the CSP capacity, and the optical field area and heat storage time are higher than that of Scheme 3, so the initial investment cost of scheme 5 is slightly higher than that of scheme 3. However, compared with scheme 4, 1.4 times the optical field area and 7 h of heat storage time still significantly reduce the initial investment cost of CSP+PV power station in scheme 5, and the equivalent annual investment cost calculated by the investment cost also decreases accordingly. In addition, because the configuration scheme optimizes the operation mode of other units, the annual operating cost of thermal power units is greatly reduced, which is only 91% of scheme 4. Therefore, in terms of the equivalent annual cost of island power grid derived from the addition of the annual operating cost of thermal power units and the equivalent annual investment cost of the CSP+PV power station, the equivalent annual cost of scheme 5 is the lowest, and the annual saving of scheme 4 is 482.08 × 10⁶ USD. Obviously, considering the economic operation of thermal power units, Scheme 5 is obviously better than Scheme 4, which is conducive to the optimal scheduling of island power grids.

To further analyze the operation performance of the CSP+PV hybrid power station under the two CSP+PV power station configuration schemes and their impact on the island power grid, taking August and January as examples, the output of CSP+PV power station and thermal power unit under the optimal configuration of the two schemes in winter and summer were compared and analyzed.

As can be seen from Fig. 6, in summer with strong sunlight, the CSP+PV power station in Scheme 4 can maintain a constant output of 30 MW from 8:00 to 24:00, providing stable and reliable power supply for 17 consecutive hours. In winter, the light is weak, but the CSP+PV power station can still provide 10 h of uninterrupted power supply between 10:00 and 20:00, which effectively solves the problem of large fluctuation and strong intermittence of PV power generation. However, the output constraint of this scheme on the CSP+PV power station is too high, which is not conducive to the CSP+PV power station's own scheduling and the economic operation of the island power grid. The specific performance is as follows: The phenomenon of light abandonment and heat abandonment of CSP+PV power stations mainly occurs in the summer when the light is strong. For example, good lighting conditions before 16:00 in summer make the heat storage system in CSP accumulate a large amount of heat. However, due to the limited capacity of the heat storage system from 16:00 to 18:00, the excess heat absorbed by the light field cannot be stored in the heat storage system. This part of the heat is directly used for CSP power generation, resulting in an increase in CSP power generation and a corresponding decrease in PV power generation, resulting in the phenomenon of light abandonment. Thermal power units of the uneconomic operation, summer 14:00–16:00 and winter 11:00–16:00 user load in the “valley period,” CSP+PV power station 30 MW constant output compression thermal power unit operating interval, thermal power units to maintain at the lower limit of economic operation and generate excess electricity, further leading to thermal power unit operating costs rise.

In contrast, Fig. 7 shows that the PV power generation of Scheme 5 is consistent with its maximum available power generation regardless of winter or summer, that is, there is no light abandonment phenomenon. CSP power generation is more dispersed, CSP is kept in the power generation state within 24 h and adjusts the power generation according to the load curve, such as 13:00–18:00 in summer and 14:00–17:00 in winter, CSP stores most of the heat collected by the light field in the heat storage system, and a small part is used for the power generation of the turbine unit. In addition, in the summer when the light is strong, the small and medium-sized thermal power unit 2 of the island power grid does not participate in the operation, but in the winter when the light is weak, the user still needs the thermal power unit 2 to participate in the operation at the peak of the evening consumption (19:00–20:00). It can be seen that the CSP+PV power station in Scheme 5 can replace thermal power units in months with better lighting conditions and reduce the operating costs of island power grids.

Given the current problems of strong randomness of single PV power generation and poor economy of single CSP power generation, this paper proposes to replace a single solar power station with a CSP+PV hybrid solar power station and establish a CSP+PV hybrid solar power station model, considering the coordinated scheduling of CSP+PV power station and thermal power unit, The capacity allocation method of CSP+PV power station based on island power grid economic dispatching is proposed. The performance superiority of CSP+PV power station and the effectiveness of the proposed configuration method are verified by the simulation examples. The main conclusions are as follows:

1. CSP+PV hybrid solar power station can simultaneously demonstrate the schedulability of CSP and the low-cost advantage of PV. In the solar power station of the same capacity, the LCOE of a CSP+PV hybrid solar power station is lower than that of a single CSP power station, and the operating cost of thermal power units in the island power grid is also lower than that of a single PV power station.

2. The capacity allocation method of CSP+PV power station based on economic dispatch of the island power grid in this paper can provide sufficient adjustment interval for optimizing the operation mode of other units, reduce the equivalent annual cost of the island power grid by reducing the overall operation cost, and finally maximize the benefits of the entire life cycle of the island power grid.

This study was supported by the National Key Research and Development Program of China under Grant No. 2022YFE0120400.

The authors have no conflicts to disclose.

Xiangning Lin: Data curation (equal); Software (equal); Supervision (equal); Validation (equal); Writing – original draft (equal); Writing – review & editing (equal). Shu Chen: Data curation (equal); Software (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.