Dynamic Operation and Optimization Analysis of an Innovative Distributed Energy System Based on Full-Spectrum Solar Energy Cascade Utilization

Abstract

The cascade utilization of spectral beam splitting represents an effective method for enhancing the efficiency of solar energy utilization. However, most research has been conducted under stable conditions, and the impacts across different climatic zones have not been taken into account. Therefore, this paper investigates an innovative distributed energy system utilizing full-spectrum solar cascade in three different climate zones. A full-spectrum solar model is established in MATLAB 2023, and the corresponding photovoltaic model files are invoked in the TRNSYS 18. After operation, the performance of full-spectrum frequency division solar energy is obtained. The equivalent carbon emissions (ECE), self-sufficiency ratio ($SS$), self-consumption ratio ($SC$) and levelized cost of energy (LCOE) are adopted as indicators to assess the environmental, energy and economic benefits of each system. The results show that the net present value (NPV) in Chengdu is the highest (186,674.19 USD), while that in Beijing is the lowest (171,458.75 USD), and that in Guangzhou is in the middle (180,650.23 USD). After optimization, Beijing achieves the lowest LCOE and the highest $SS$, Guangzhou achieves the highest $SC$ and the lowest ECE, Chengdu achieves a balanced configuration where moderate on-site generation meets nearly half of the total demand.

1. Introduction

In recent years, fossil fuels have accounted for 84.3% of global energy consumption, while non-fossil energy sources represent less than 16%. This fossil fuel-dominated energy consumption pattern results in substantial greenhouse gas emissions [1], contributing significantly to global climate change. Renewable energy includes a wide range of sources, including biomass, solar, wind, ocean, hydro, tidal and geothermal energy. Among these, solar energy stands out as a clean and abundant resource [2]. Currently, the two primary methods for harnessing solar energy are solar thermal utilization and photovoltaic power generation [3]. The former stores solar energy in the form of heat, while the latter converts it into electricity. However, the low flux density of energy and intermittent nature of solar radiation often lead to inefficiencies and operational instability in solar energy systems, thereby hindering their large-scale adoption. Therefore, the key challenges lie in maximizing solar resource utilization and reducing the reliance of solar energy systems on the electrical grid.

Recent research on the optimal utilization of solar resources has concentrated on the graded utilization of full-spectrum solar energy. As the term suggests, this approach involves separately harnessing energy of different qualities (wavelengths) within the solar spectrum. By employing appropriate conversion pathways, energy of varying grades can be matched with suitable conversion units, thereby minimizing energy loss. This strategy not only cools PV cells effectively, reducing their operating temperature, but also recovers waste heat for domestic hot water, thereby improving overall energy utilization efficiency.

Current full-spectrum solar graded utilization systems can be categorized by system output into three main forms: hydrogen energy, power/hydrogen energy and cooling/heating/power/hydrogen energy based on full-spectrum solar energy.

1.1. Hydrogen Energy Based on Full-Spectrum Solar Energy

Li et al. [4] introduced a novel approach to full-spectrum solar hydrogen generation by integrating photothermal synergistic reactions with photovoltaic power generation to facilitate water electrolysis. A mathematical model was developed to analyze and deliberate upon this method. Liu et al. [5] introduced a solar photovoltaic-thermal hydrogen production system that utilized the full spectrum of sunlight. The concentrated sunlight was separated into two components based on wavelength. The longer-wavelength sunlight was converted into thermal energy and directly supplied to the endothermic water electrolysis reaction. The shorter-wavelength sunlight was guided onto the photovoltaic cell to generate electrical energy for the reaction. This study presented a comprehensive overview of the mathematical model and equations, while also identifying the key parameters necessary for system simulation. In comparison to conventional photovoltaic water electrolysis systems, this approach enhanced efficiency. Song et al. [6] introduced a novel hydrogen production system that integrated a two-step thermochemical cycle (TC) with photovoltaic water electrolysis, marking the first instance of such a combination and enabling the efficient harnessing of the entire solar spectrum. Unreacted high-temperature steam from the thermochemical cycle underwent direct electrolysis in a solid oxide electrolyzer (SOEC). Additionally, the waste heat from the photovoltaic cell was utilized to preheat water, thereby minimizing energy losses, facilitating zero-carbon hydrogen production, and promoting energy cascade utilization. Through numerical simulations, the system’s performance in terms of solar-to-hydrogen (STH) efficiency and exergy efficiency was evaluated across various parameters, with a detailed examination of material and energy flows. Huang et al. [7] integrated an Ag@SiO₂-Au hybrid nanofluid filter into a concentrating photovoltaic/thermal (CPV/T) system. The nanofluid filter aimed to break down incoming sunlight to power both the PEM and MSR systems efficiently. This synergistic design continuously harnessed the entire solar spectrum, enabling solar energy to be stored in the form of hydrogen. By developing models for the CPV/T, PEM, and MSR systems, a comprehensive analysis of the integrated system’s performance was conducted. Liu et al. [8] proposed a solid oxide electrolysis cell (SOEC) hydrogen production system that integrated a thermal storage module, thereby maximizing the utilization of full-spectrum solar energy. The system employed spectral splitting, whereby the short-wave spectrum was converted into electrical energy via photovoltaic cells, while the long-wave spectrum was transformed into thermal energy within the reactor. It operated in two modes: during the day, the system harnessed all available solar energy for electricity generation and a portion of solar thermal energy for hydrogen production; at night, it relied on grid power and stored solar thermal energy to produce hydrogen. This innovative system showed substantial improvements in both efficiency and stability of solar hydrogen production, presenting a viable solution for sustainable and efficient hydrogen generation (

Table 1. Hydrogen energy based on full-spectrum solar energy.

Researcher	System	Software	Parameter	Location
Li [4]	A full-spectrum solar energy system that integrates photothermal synergistic reactions with photovoltaic water electrolysis	-	T0 = 300 K DNI is fixed	-
Liu [5]	A solar photovoltaic–thermal hydrogen production system that employs comprehensive spectrum utilization, harnessing both thermal and electrical energy in a synergistic approach to generate hydrogen	-	T0 = 298.15 K DNI is fixed	-
Song [6]	A novel hydrogen production system combining thermochemical water splitting, photovoltaic cells, and SOEC water electrolysis	-	T0 = 298.15 K, DNI is fixed	-
Huang [7]	A full-spectrum solar-driven hydrogen production system that couples a concentrated photovoltaic/thermal system with a proton exchange membrane electrolyzer and a methanol steam reforming reactor	MATLAB 2023	T0 = 298.15 K, DNI is fixed	-
Liu [8]	SOEC hydrogen production system with thermal storage module based on full-spectrum solar energy utilization	SMARTS	T0 = 298.15 K DNI = 200–1000 W/m2	Beijing, China

).

1.2. Power/Hydrogen Energy Based on Full-Spectrum Solar Energy

Li et al. [9] combined spectral-splitting photovoltaic/thermal (PV/T) technology with methane steam reforming. In this configuration, short-wavelength solar radiation was directed to the photovoltaic cells for power generation, while long-wavelength solar radiation was utilized by the methane steam reforming device through a solar-to-chemical-to-thermal energy conversion pathway. Zhu et al. [10] established a thermodynamic model coupling the full solar spectrum with a two-step thermochemical cycle and theoretically examined the cycle performance based on this model. Tunable cut-off wavelength spectral splitting was employed to flexibly allocate the solar spectrum to meet the thermal and electrical demands of the cycle. The introduction of spectral splitting reduced the levelized cost of electricity and significantly enhanced system efficiency. Wu et al. [11] conducted an analysis of a photoelectric/thermochemical synergistic conversion system from a thermodynamic perspective. By establishing a thermodynamic model and comparing it with other spectral-splitting systems, they employed the golden section algorithm to determine the thermodynamic limits of the PV/T synergistic conversion system. The study examined the influence of various parameters, including cut-off wavelength, concentration ratio, and internal irreversibility, on system performance. Key performance metrics assessed included output voltage, available energy per absorbed photon for photocatalysis, conversion temperature, and selective absorption cut-off wavelength. This research established a thermodynamic foundation for the advancement and design of photoelectric/thermochemical synergistic conversion systems. Zhu et al. [12] integrated photovoltaic and thermochemical conversion through spectral-splitting technology, facilitating the concurrent generation of electricity and solar fuels from the entire solar spectrum. They investigated daily operational performance under actual solar irradiation conditions using both simulation and experimental approaches (

Table 2. Power and hydrogen energy based on full-spectrum solar energy.

Researcher	System	Software	Parameter	Location
Li [9]	A photovoltaic/thermal–methane steam reforming hybrid system for spectral frequency splitting	Aspen Plus	${\mathrm{T}}_0$ = 25 °C DNI is fixed	-
Zhu [10]	A thermodynamic model coupling the full solar spectrum with a two-step thermoelectric cycle	Lab experiment	${\mathrm{T}}_0$ is changing DNI is fixed	-
Wu [11]	The thermodynamic model system of the PV-PTC system from the perspectives of thermodynamic limits	MATLAB	${\mathrm{T}}_0$ = 300 K DNI is fixed	-
Zhu [12]	A spectral frequency splitting technology to couple photovoltaic and thermochemical conversion, simultaneously generating electricity and solar fuels	Lab experiment	${\mathrm{T}}_0$ is changing DNI is changing	Nanjing, Jiangsu Province, China

).

1.3. Cooling/Heating/Power/Hydrogen Based on Full-Spectrum Solar Energy

Wang et al. [13] proposed a combined cooling, heating, and power system that employed a full-spectrum hybrid solar energy device that integrated a molecular solar thermal system with a solar water heating system. This innovative approach comprehensively harnessed solar energy, fuel chemical energy, and waste heat from methanol decomposition, thereby achieving full-spectrum utilization of solar energy and improving the operational stability of the system. Han et al. [14] proposed a full-spectrum solar-driven trigeneration system that incorporated an organic Rankine cycle. In the DeST software, meteorological parameters (outdoor temperature and DNI) and indoor design parameters of regional buildings in Beijing were considered. An engineering equation solver (EES) was employed to establish the thermodynamic model of the system. The capacity was determined according to the maximum load of the building, while the building’s electricity demand was met by the grid. The system was optimized by using the heat distribution ratio as the decision variable to enhance its environmental benefits. The annual performance of the system was subsequently analyzed under optimal conditions. Hao et al. [15] developed an innovative solar spectral split cogeneration system for cooling, heating, and power, based on the energy levels and operational characteristics of the refrigeration cycle. This system divided the solar spectrum into a high-frequency component for power generation and a low-frequency component for heat collection. Photovoltaic waste heat was recovered for fluid heating, while high-temperature thermal energy was used to drive an ejector refrigeration system to provide cooling capacity. Additionally, they determined the parameter distribution of energy load and efficiency. Fang et al. [16] proposed a photovoltaic electrolysis green hydrogen and thermochemical reforming gray hydrogen co-production system based on spectral splitting technology, combining electrochemistry and thermochemistry to maximize the utilization of full-spectrum solar energy. Zhong et al. [17] proposed a novel hybrid system combining concentrated photovoltaic (CPV) and ion thermal energy conversion (iTEC) and conducted a comprehensive evaluation of its feasibility. A three-dimensional numerical model was proposed and validated through initial experiments. The impact of channel height and width on iTEC was thoroughly investigated, along with the influence of concentration ratio and flow rate of the redox electrolyte on the CPV-iTEC hybrid system’s performance to determine the optimal parameters. Advanced redox electrolyte/electrode combinations were utilized to boost the CPV-iTEC system’s efficiency (

Table 3. Cooling/heating/power/hydrogen based on full-spectrum solar energy.

Researcher	System	Software	Parameter	Location
Wang [13]	A combined cooling, heating, and power system utilizing a full-spectrum hybrid solar energy device that integrates a molecular solar thermal system and a solar water heating system	Aspen Plus + EES	T0 = 25 °C DNI = 1000 W/m2	-
Han [14]	A full-spectrum solar-driven tri-generation system incorporating an organic Rankine cycle	EES	${\mathrm{T}}_0$ is changing DNI is changing	Beijing, China
Hao [15]	A new solar spectral split system integrating cooling, heating, and power generation by recovering waste heat from photovoltaics for fluid heating. Additionally, high-temperature thermal energy is used to power an ejector refrigeration system, thus enhancing cooling capacity.	EES	T0 = 42 °C DNI is fixed	-
Fang [16]	A photovoltaic electrolysis green hydrogen thermochemical reforming ash hydrogen co-production system based on spectral splitting technology	Aspen Plus	T0 = 298.15 K, DNI = 200–1000 W/m2	Lanzhou, China
Zhong [17]	A new hybrid system combining iTEC and CPV for comprehensive solar cascade utilization	COMSOL 5.4 Multiphysics	${\mathrm{T}}_0$ = 300 K DNI is fixed	-

).

Overall, a significant amount of research has been conducted on spectral frequency-splitting photovoltaic/thermal systems both domestically and internationally. However, most studies focus on the spectral splitting of full-spectrum solar energy and the analysis of its products under design conditions, without considering the system’s performance under varying irradiance and temperature. Moreover, solar radiation and outdoor temperature vary across different climatic zones, which directly impacts system configuration and building loads [18]. Therefore, this article considers the effects of different irradiance and temperature and discusses the configuration and operation of various components of the system under different climatic conditions.

This research proposes a novel distributed energy system based on full-spectrum solar energy cascade utilization, whose innovation lies in:

(1)

A dynamic simulation method based on full-spectrum solar energy cascade utilization is proposed. The electrical power generation and heat collection based on spectral-splitting utilization are calculated using MATLAB 2023, and the results are input into TRNSYS 18 to simulate system operation, realizing the dynamic characteristics of energy flow and efficiency of the system.

(2)

The impact of different climatic zones (hot summers and cold winters zone, severe cold zone, and hot summers with mild winters zone) on system performance has been taken into account. This is crucial for analyzing the compatibility between systems and building loads, thereby ensuring the model’s feasibility across various regional conditions at the source-load coordination level.

(3)

The NPV is taken as the objective to optimize system performance, evaluating the system from technical, economic, and environmental perspectives.

2. System Introduction

This section introduces the concept and composition, and elaborates on the models necessary for analyzing system frequency splitting, power generation, heat collection, and energy storage of an innovative distributed energy system that utilizes full-spectrum solar energy through a cascade approach.

2.1. Mathematical Modeling in MATLAB

This study employs a full-spectrum photovoltaic mathematical model implemented in MATLAB. The absorption range of conventional silicon photovoltaic cells is limited, with an upper limit of approximately 1100 nm. The portion of solar energy not absorbed during the photovoltaic process is converted into heat, which increases the module temperature and reduces power generation efficiency. Spectral beam splitting (SBS) is considered a feasible solution to this limitation. SBS divides the solar spectrum into two parts based on wavelength: high-quality short-wavelength solar energy is utilized for power generation, while low-quality long-wavelength solar energy is converted into thermal energy [15].

2.1.1. Frequency Divider

The design principle of the frequency-splitting concentrating photovoltaic and thermal receiver primarily involves selecting suitable frequency-splitting liquids for different solar cells. The liquid frequency splitter is positioned on the light source side, first concentrating the light through a linear Fresnel lens [19] and then splitting it into two parts using spectral splitting technology. One part consists of high-grade solar energy, which is converted into electrical energy by monocrystalline silicon solar cells. The other part, which cannot be utilized by photovoltaic cells, is converted into thermal energy. Additionally, the frequency-splitting liquid can be used as a cooling medium to cool the solar cells, thereby reducing their temperature and allowing for the recovery and reuse of waste heat from the solar cells [20]. The design principle of the liquid frequency-splitting concentrating photovoltaic and thermal receiver is to collect and concentrate solar energy.

$$\mathrm{W}\mathrm{h}\mathrm{e}\mathrm{n}\lambda \left(\mathrm{nm}\right)\ge 1240/{\mathrm{E}}_{\mathrm{g}}\left(\mathrm{e}\mathrm{V}\right),{\tau }_{\lambda }=0;\mathrm{w}\mathrm{h}\mathrm{e}\mathrm{n}\lambda \left(\mathrm{nm}\right)<1240/{\mathrm{E}}_{\mathrm{g}}\left(\mathrm{e}\mathrm{V}\right),{\tau }_{\lambda }=1$$

Total transmittance can be calculated as

$${\eta }_{\mathrm{SBS}}={\displaystyle \frac{{\int }_{280\mathrm{n}\mathrm{m}}^{2500\mathrm{n}\mathrm{m}}{\tau }_{\lambda }\cdot R_{\lambda }d\lambda }{{\int }_{280\mathrm{n}\mathrm{m}}^{2500\mathrm{n}\mathrm{m}}{R}_{\lambda }d\lambda }}$$

(1)

where R_λ represents the energy carried by the wavelength (W/(nm·m²)) and λ represents the wavelength (nm).

2.1.2. PV/T

Silicon photovoltaic cells are extensively utilized in commercial applications due to their high efficiency, environmental compatibility, reliability, and long lifespan [21]. Therefore, this study chooses silicon photovoltaic cells as the photoelectric conversion medium.

The electrical power output from PV/T can be calculated as [19]

$${\mathrm{E}}_{\mathrm{PV}}={\mathrm{J}}_{\mathrm{SC}}\cdot {\mathrm{V}}_{\mathrm{O}\mathrm{C}}\cdot \mathrm{F}\mathrm{F}$$

(2)

where V_OC stands for open-circuit voltage (V), J_SC stands for short-circuit current density (A), and FF stands for fill factor.

The short-circuit current density can be calculated as

$${\mathrm{J}}_{\mathrm{SC}}={\int }_{280\mathrm{n}\mathrm{m}}^{{\lambda }_g}q\cdot {\displaystyle \frac{{\alpha }_{\lambda }\cdot \mathrm{E}\mathrm{Q}{\mathrm{E}}_{\lambda }\cdot {\mathrm{R}}_{\lambda }}{h_p·c}}\cdot \mathrm{d}\lambda \cdot {\displaystyle \frac{\mathrm{D}\mathrm{N}\mathrm{I}}{{\int }_{280\mathrm{n}\mathrm{m}}^{4000\mathrm{n}\mathrm{m}}R\left(\lambda \right)d\lambda }}+{\eta }_{\mathrm{P}\mathrm{V}}\cdot \left({\mathrm{T}}_{\mathrm{P}\mathrm{V}}-298.15\right)$$

(3)

where DNI represents direct normal irradiance (W/m²); λg represents the wavelength range at the PV bandgap (nm); ${\alpha }_{\lambda }$ is the correction factor, set to 0.93; and ${\eta }_{\mathrm{P}\mathrm{V}}$ is the photovoltaic temperature correction coefficient.

The open-circuit voltage can be calculated as

$${\mathrm{V}}_{\mathrm{OC}}={\displaystyle \frac{\alpha \cdot {\mathrm{k}}_{\mathrm{B}}\cdot {\mathrm{T}}_{\mathrm{P}\mathrm{V}}}{\mathrm{q}}}\cdot \ln \left({\displaystyle \frac{{\mathrm{J}}_{\mathrm{S}\mathrm{C}}}{{\mathrm{J}}_0}}+1\right)$$

(4)

The saturation current density can be calculated as

$${\mathrm{J}}_0=1.5\times 10^{-5}\cdot {\mathrm{e}}^{-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{g}}}{\omega \cdot {\mathrm{k}}_{\mathrm{B}}\cdot {\mathrm{T}}_{\mathrm{P}\mathrm{V}}}}}$$

(5)

The fill factor can be calculated as

$$\mathrm{F}\mathrm{F}={\displaystyle \frac{\frac{V_{oc}\cdot q}{\omega \cdot k_B\cdot T_{PV}}-\ln \left(\frac{V_{oc}\cdot q}{\omega \cdot k_B\cdot T_{PV}}+0.72\right)}{\frac{V_{oc}\cdot q}{\omega \cdot k_B\cdot T_{PV}}+1}}$$

(6)

The total solar radiation used for heating in full spectrum can be calculated as

$${\mathrm{Q}}_{\mathrm{S}}={\mathrm{S}}_{\mathrm{s}}\cdot {\mathrm{C}}_{\mathrm{s}}\cdot {\int }_{{\lambda }_{\mathrm{g}}}^{4000\mathrm{n}\mathrm{m}}{\mathrm{R}}_{\lambda }\mathrm{d}\lambda$$

(7)

where C_s is the concentration ratio, set to be 6; S_s denotes the collector area (m²).

The photovoltaic heat loss can be calculated as [16]

$${\mathrm{Q}}_{\mathrm{PV}-\mathrm{heat},\mathrm{loss}}={\mathrm{h}}_{\mathrm{P}\mathrm{V}}\cdot \left({\mathrm{T}}_{\mathrm{P}\mathrm{V}}-{\mathrm{T}}_0\right)\cdot {\mathrm{A}}_{\mathrm{P}\mathrm{V}}+{\epsilon }_{\mathrm{P}\mathrm{V}}\cdot \sigma \cdot \left({\mathrm{T}}_{\mathrm{P}\mathrm{V}}^4-{\mathrm{T}}_{\mathrm{s}\mathrm{k}\mathrm{y}}^4\right)\cdot {\mathrm{A}}_{\mathrm{PV}}$$

(8)

where T₀ is the ambient temperature (K), T_pv represents the operating temperature of the photovoltaic cell (K), T_sky represents the effective temperature (K), σ represents the Stefan-Boltzmann constant, ε_PV represents the emissivity of the photovoltaic cell, h_PV stands for the convection coefficient, and A_pv signifies the photovoltaic panel area, set to be 50 m².

2.2. System Modelling in TRNSYS

The system design and optimization were carried out using TRNSYS, in combination with MATLAB to manage the inputs and outputs of the photovoltaic/thermal (PV/T) system. TRNSYS is a dynamic simulation software for modeling and analyzing renewable energy systems; it enables the simulation and integration of various components to model system performance under dynamic conditions. TRNSYS 18 also provides external interfaces with MATLAB 2023 and Genopt, 3.1.1 making it easy to call and perform simulations. Owing to their complementary strengths, the interaction between MATLAB and TRNSYS provides significant advantages. TRNSYS reads weather data (direct radiation and ambient temperature) and inputs it into MATLAB. MATLAB then processes this data by applying Formulas (1)–(8) to perform precise calculations to obtain power generation and thermal output, after which the results are sent back to TRNSYS, which then initiates a new simulation.

Figure 1 is a system schematic diagram, mainly including a photovoltaic power generation subsystem, a full-spectrum utilization heat collection subsystem, a thermoelectric complementary drive refrigeration subsystem, a heating subsystem, and an alkaline fuel cell subsystem. The comparison of different outputs obtained by the present model and the experimental data are shown in

Table 4. Comparison of different outputs obtained by the present model and the experimental data.

Variable	The Present Study	The Experimental Data [22]	Unit	Error (%)
DNI	1000	1000	W/m2	-
Atmospheric temperature	25	25	°C	-
Joc	32.65	30.17	mA/cm2	8.22
Voc	0.85	0.79	V	7.59
Electrical efficiency	30.23%	28.8%	-	4.96

, the simulation verification of the alkaline fuel cell is presented in

Table 5. Comparison of simulation results of alkaline fuel cells with those reported in literature [23].

Variable	The Present Study	Literature [23]	Unit	Error (%)
Modules per stack in series	16	16	-	-
Modules per stack in parallel	1	1	-	-
The heat generated by an alkaline fuel cell during the operation	8775	8700	W	0.86
The electricity output of an alkaline fuel cell	6774	6700	W	1.10

, and the component modules used in TRNSYS are listed in

Table 6. Main equipment modules and their codes.

Model	Type	Component Description	Model	Type	Component Description
PV/T	Type155	The module connecting TRNSYS and MATLAB	Absorption chiller	Type909	Harnessing heat to drive cooling systems for buildings
Water tank	Type4c	Hot water storage	Electric boiler	Type138	Heat supply
Evaporative refrigeration	Type655	Utilize electricity to provide cooling for building	Electrolyzer	Type160	Electrolyze surplus electricity into hydrogen
Alkaline fuel cell	Type164b	Transform hydrogen into electrical energy	Multi-stage compressor	Type167	Compress hydrogen for storage

.

2.3. Operation Strategy

2.3.1. PV/T System

The operation of photovoltaic panels is determined by solar radiation. When solar radiation is abundant during the day, the photovoltaic panels start to work. However, when there is no solar radiation at night, the photovoltaic panels do not work.

Schematic diagram of an innovative distributed energy system is shown as Figure 2. The schematic diagram of the SBS-PV/T hybrid system based on nanofluid introduced in this study mainly consists of three parts: a Fresnel lens concentrator, a nanofluid frequency splitter, and photovoltaic cells [15]. In this system, solar radiation first passes through a Fresnel lens, then traverses a nanofluid frequency divider, and ultimately impacts the surface of photovoltaic modules. The nanofluid primarily absorbs solar radiation outside the spectral response band of photovoltaic cells, thereby achieving efficient utilization of full-spectrum solar energy [24]. The nanofluid serves as both a spectral divider and a heat collector, effectively reducing energy loss during heat transfer (Figure 3).

2.3.2. Heating System

The main heat sources in this subsystem are the PV/T system and an electric boiler. When the heat provided by the PV/T system and the waste heat from the AFC is insufficient to supply DHW, the system provides electricity to the electric boiler. The hot water heated by the electric boiler is supplied to users through a circulation pump.

When the heating load is greater than 0, the system supplies power to the electric boiler, then the heated water is delivered to the user through the heating water pump.

2.3.3. Refrigeration System

The core of the refrigeration cycle consists of absorption refrigeration units and evaporative chillers. When the heat supply is insufficient, the chilled water produced by the refrigeration unit is delivered to the room through the water pipe for heat exchange. After absorbing heat, the chilled water returns to the refrigeration unit.

2.3.4. Alkaline Fuel Cell System

In the solar energy–hydrogen storage–alkaline fuel cell system, excess photovoltaic power drives the alkaline electrolyzer to produce hydrogen, which is then compressed by a hydrogen compressor and stored in a dedicated tank. When solar power generation cannot meet the demand, the alkaline fuel cell utilizes the stored hydrogen as fuel to produce electricity, thereby maintaining the system’s operational stability (Figure 4).

3. Optimization Variable

To ensure the system operates at its optimal state, this study utilizes Genopt software 3.1.1 to perform optimization calculations by selecting appropriate optimization variables and objective functions [25]. Genopt is a program designed to find the minimum value of an optimization function. Users can specify multiple optimization variables and perform interpolation calculations on the optimization function by selecting the optimization algorithm provided by Genopt. The Hooke–Jeeves algorithm in Genopt is utilized to optimize the selected set of variables for the hot water tank volume (V_tank1), the storage hot water tank volume (V_tank2), the electric boiler capacity (E_h), the capacity of evaporative refrigeration (E_d) and the capacity (E_x) of the absorption refrigeration unit. Main equipment modules and prices are shown as

Table 7. Main equipment modules and prices [25].

Model	Price	Model	Price
PV/T	982 USD/m2	Absorption chiller	216 USD/kW
Water tank	84.3 USD/m3	Evaporative refrigeration	28.10 USD/kW
Electric chiller	136.28 USD/kW	Alkaline Fuel Cell	1100 USD/kw
Hydrogen storage tank	2100 USD	Multi-stage compressor	15,000 USD

.

The constraint condition is key to ensuring that the calculation results of the optimization model are reasonable and available. The constraints of the optimization model in this study are mainly parameter inequality constraints. The value range of the decision variable is the result range of the optimization model. Therefore, for the operation of each system, the range of decision variable is as follows:

$$0\le {\mathrm{V}}_{\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{k}}\le {\mathrm{Q}}_{DHW}^{max}$$

(9)

$${0\le {\mathrm{E}}_h\le \mathrm{E}}_{heating}$$

(10)

$${0\le {\mathrm{E}}_d+{\mathrm{E}}_x\le \mathrm{E}}_{cooling}$$

(11)

This study aims to maximize the NPV; the values of final variables of each system accordingly are as follows (

Table 8. Optimization variables of each system.

City	Parameters
Chengdu	Vtank1 = 10 m3, Vtank2 = 10 m3, Eh = 50 kW, Ed = 351.67 kW, Ex = 70 kW
Beijing	Vtank1 = 10 m3, Vtank2 = 10 m3, Eh = 200 kW, Ed = 131 kW, Ex = 85 kW
Guangzhou	Vtank1 = 10 m3, Vtank2 = 10 m3, Eh = 50 kW, Ed = 280 kW, Ex = 70 kW

):

The optimized decision variables are shown in the Table 8. The NPV of Chengdu is the highest (USD 186,674.19), and the NPV of Beijing is the lowest (USD 171,458.75), the NPV of Guangzhou is in the middle (USD 180,650.23).

4. Evaluation Indicators

Based on economy-oriented selection optimization, a comprehensive comparison of solar energy system performance is conducted across three cities. The technical, economic, and environmental indicators are defined as follows:

4.1. Technical Indicators

Self-sufficiency and self-consumption ($SS$ and $SC$) are commonly employed to assess the technical performance of grid-connected residential energy systems [26]. Self-sufficiency is defined as the ratio of residential load satisfied by self-generated energy, thereby reflecting the system’s autonomy and resilience. Self-consumption indicates the proportion of residential load supplied by self-generated electricity, showing whether the installed solar technology is oversized. These two indicators are expressed as follows:

$$SS=1-{\displaystyle \frac{{\sum }_{t=1}^{8760}{E}_{grid}\left(t\right)}{{\sum }_{t=1}^{8760}{E}_{load}\left(t\right)}}$$

(12)

$$SC=1-{\displaystyle \frac{{\sum }_{t=1}^{8760}\left[E_{sc}\left(t\right)+Q_{sc}\left(t\right)\right]}{{\sum }_{t=1}^{8760}\left[E_{PV}\left(t\right)+Q_S\left(t\right)\right]}}$$

(13)

where $E_{grid}\left(t\right)$ represents the grid input power, $E_{load}\left(t\right)$ represents the total hotel load, $E_{sc}\left(t\right)$ and $Q_{sc}\left(t\right)$ represent the solar power and heat for self-consumption, and $E_{PV}\left(t\right)$ and $Q_S\left(t\right)$ represent the total electrical and thermal output from the solar energy system.

4.2. Economic Indicator

$${NPV}_{total}=C_i+u·C_i·{\displaystyle \frac{{\left(1+i\right)}^m·i}{{\left(1+i\right)}^m-1}}+C_g·{\sum }_{t=1}^{8760}{E}_{grid}\left(t\right)$$

(14)

$$LCOE={\displaystyle \frac{{NPV}_{total}}{E_{load}·}}·{\displaystyle \frac{{\left(1+i\right)}^m·i}{{\left(1+i\right)}^m-1}}$$

(15)

where LCOE is based on a core financial concept: the time value of money. LCOE refers to the present value of total costs over the life cycle divided by the present value of total electricity generation over the life cycle [27]; ${NPV}_{total}$ refers to the net present value of the system; $E_{load}$ represents the building load; $i$ is the discount rate (taken as 5%); ${\mathrm{C}}_g$ is the price for purchasing 1 kWh of electricity; and ${\mathrm{C}}_i$ is the total system purchase cost.

4.3. Environmental Indicator

Solar energy systems significantly reduce electricity costs and mitigate emissions by replacing high-carbon grid power, providing both economic and environmental benefits. Therefore, the ECE index [28] is employed to assess its environmental performance:

$$ECE={\delta }_{carbon}·{\sum }_{t=1}^{8760}{E}_{grid}\left(t\right)$$

(16)

where ${\delta }_{carbon}$ represents the carbon emission coefficient of the utility grid, indicating the specific carbon emissions associated with generating 1 kWh of electricity.

5. System Analysis

This section analyzes and discusses the results of the proposed system of simulation and optimization.

5.1. Location

According to the code for design of civil buildings in China, with the average temperature of the coldest month and the hottest month as the main index, the country is divided into five regions: severe cold region, cold region, cold winter and hot summer region, hot summer and warm winter region and mild region. This paper selects the following cities as the research sites: Chengdu (hot summer and cold winter region, Figure 5), Guangzhou (hot summer and warm winter region, Figure 6) and Beijing (cold region, Figure 7).

5.2. Building Characteristics

In order to conduct load simulation, the hotel model created in SketchUp 2018 (Figure 8) is integrated into the TRNBuild environment. Type 56 is used to calculate hourly heating, cooling, and electrical loads, and the building’s hourly heating, cooling and electrical loads are imported into TRNSYS via Type 9e. The meteorological inputs are sourced from a Typical Meteorological Year (TMY2) file, implemented in TRNSYS via Type 15-6.

$$Q_{\mathrm{DHW}}=4.19\cdot {\displaystyle \frac{{1000m}_{dhw}\cdot {(t}_{hot}-t_{cold})}{3600}}$$

(17)

where $Q_{\mathrm{DHW}}$ is the heat consumption of domestic hot water (kw); $m_{dhw}$ is the water quota for hotel rooms in the “Code for Design of Building Water Supply and Drainage” GB50015 [29], which is 120–160 L/bed; $t_{hot}$ is the temperature of hot water; and $t_{cold}$ is the temperature of cold water. During operational periods, the temperature in guest rooms is maintained at 22 °C (heating) and 26 °C (cooling). The corridor temperature is set at 16 °C (heating) and 26 °C (cooling).

This study is based on a five-story hotel.

Table 9. Building envelope structures in different regions.

Building Envelope	Heat Transfer Coefficient (W/m2·K)
	Chengdu	Beijing	Guangzhou
external wall	0.97	0.478	1.7
exterior window	2.4	1.1	2.43
roof	0.67	0.44	1.06
internal wall	0.99	1.718	1.86

lists the specifications of the building envelope at different locations within the building. The hotel area is 2700 m². The daily occupancy rate is 80%, and the daily domestic hot water consumption is 120 L per bed. Figure 9, Figure 10 and Figure 11 show the cooling, heating, and electricity demand curves for the three cities.

5.3. Annual Operation Analysis

For the full-spectrum solar energy system in Beijing, in terms of electricity supply, the photovoltaic power generation, alkaline fuel cell power generation, and electricity purchased from the grid account for 48.4%, 12.9%, and 7.4% of the total energy supply, respectively, with 592,187.316 kWh, 157,890.00 kWh, and 90,594.832 kWh. Among them, hydrogen production from electrolytic cells accounts for 69%, building electricity load accounts for 11%, the heating subsystem accounts for 10%, the cooling subsystem accounts for 9%, and electricity loss accounts for 1%.

In terms of heat supply, the full-spectrum solar energy system has a heat collection capacity of 350,737.192 kWh, accounting for approximately 28.7% of the total energy supply. The waste heat recovery during alkaline fuel cell operation is approximately 30,803.333 kWh, accounting for approximately 2.5% of the total energy supply. In terms of heat consumption, heating accounts for 50%, domestic hot water consumption accounts for 23% throughout the year, heat supply to absorption refrigeration units accounts for 24%, and heat loss accounts for 3% (Figure 12).

For Chengdu’s full-spectrum solar energy system, in terms of electricity supply, photovoltaic power generation, alkaline fuel cell power generation, and electricity purchased from the grid account for 42%, 19.1%, and 12.9% of the total energy supply, respectively, with values of 347,991.853 kWh, 158,000 kWh, and 107,807.605 kWh. Among them, hydrogen production from electrolytic cells accounts for 51%, building electricity load accounts for 26%, the heating subsystem accounts for 13%, the cooling subsystem accounts for 9%, and electricity loss accounts for 1%.

In terms of heat supply, the full-spectrum solar energy system has a heat collection capacity of 208,084.692 kW, accounting for about 20.1% of the total energy supply. The waste heat recovery during alkaline fuel cell operation is about 37,896.675 kWh. Of this, the annual consumption of domestic hot water is 85,775 kWh, accounting for 40%. The heat supplied to absorption refrigeration machines accounts for 59%, and the heat loss accounts for 2% (Figure 13).

For Guangzhou’s full-spectrum solar energy system, the electricity supply consists of photovoltaic power generation, alkaline fuel cell power generation, and electricity purchased from the grid account for 51.4%, 13.9%, and 2.2% of the total energy supply, respectively, with 656,000.000 kWh, 177,000 kWh, and 27,500.000 kWh. Among them, hydrogen production from electrolytic cells accounts for 75.4%, building electricity load accounts for 10%, the heating subsystem accounts for 6%, the cooling subsystem accounts for 8%, and electricity loss accounts for 0.6%.

In terms of heat supply, the full-spectrum solar energy system has a heat collection capacity of 381,000.000 kWh, accounting for approximately 29.9% of the total energy supply. The waste heat recovery during alkaline fuel cell operation is approximately 34,444.444 kWh, accounting for approximately 2.7% of the total energy supply. In terms of heat consumption, heating accounts for 33%, domestic hot water consumption accounts for 21% throughout the year, heat supply to absorption refrigeration units accounts for 44%, and heat loss accounts for 2% (Figure 14).

These results indicate the characteristics of the proposed full-spectrum solar energy system under different climates: Guangzhou has the largest proportion of hydrogen storage and low dependence on the power grid. Chengdu relies more on fuel cells and the power grid, while directly supplying a higher proportion to building load consumption. Beijing shows a balanced feature, making full use of solar energy for heating in winter and cooling in summer.

5.4. Typical Day Analysis

5.4.1. Analysis of Typical Winter Days

Figure 15 illustrates the hourly operation strategy of the proposed full-spectrum solar energy system in Beijing on a typical winter day. The time from 8:00 to 12:00 is when the photovoltaic subsystem collects solar energy, which is about 5 h, with peak output occurring around 12:00. When the photovoltaic subsystem fails to generate enough electricity, the power required for the heating subsystem is first provided by alkaline fuel cells and then by the municipal power grid. Electric boilers primarily supply the necessary heat for heating, whereas the storage water tank of the solar energy collection subsystem provides domestic hot water.

Figure 16 shows the operation results of Chengdu’s full-spectrum solar energy system on typical winter days. The photovoltaic subsystem begins generating power at 10:00, reaches a peak at 13:00, and ceases generation after 17:00, resulting in a total effective generation period of approximately 7 h. From 10:00 to 17:00, the electricity required by the electric boiler and the building can be fully supplied by the photovoltaic subsystem, and the remaining photovoltaic electricity can be converted into hydrogen for storage. When the photovoltaic subsystem fails to generate enough electricity, the power required for the heating subsystem is first provided by alkaline fuel cells and then by the municipal power grid. The heat required for heating load is mainly provided by electric boilers, while the heat required for domestic hot water and a small portion of heating is provided by solar energy collection subsystems.

Figure 17 shows the operation results of Guangzhou’s full-spectrum solar energy system on typical winter days, with only trace amounts of solar energy from 11:00 to 13:00 on this day. When the photovoltaic subsystem fails to generate enough electricity, the power required for the heating subsystem is first provided by alkaline fuel cells and then by the municipal power grid. The heat required for domestic hot water and heating is mainly provided by electric boilers; unlike in colder climates, the heating demand in Guangzhou is relatively modest.

5.4.2. Analysis of Typical Summer Days

The operation results for typical summer day in Beijing are shown in Figure 18. From 6:00 to 17:00, the full-spectrum solar energy system continuously provides and stores electrical and thermal energy. During this period, the cooling load and electrical load can be fully provided by the photovoltaic subsystem, which operates for about 13 h. Firstly, the heat provided by the system is supplied to domestic hot water and absorption refrigeration units. Excess hot water is stored in a hot water storage tank. When there is no solar energy, the heat required for domestic hot water and absorption refrigeration units is provided by the hot water storage tank. After the heat is consumed, the photovoltaic subsystem directly supplies power to the electric refrigeration unit, while the remaining photovoltaic power is converted into hydrogen for storage. When the photovoltaic subsystem cannot generate electricity at night, the required power is first provided by alkaline fuel cells and then by the municipal power grid.

The operation results for a typical summer day in Chengdu are shown in Figure 19. From 8:00 to 6:00, the full-spectrum solar energy system begins to provide and store electrical and thermal energy. During this period, the cooling and electrical loads can be fully provided by the photovoltaic subsystem, which operates for about 10 h. Firstly, the heat provided by the system is supplied to the water tank, while the heat required for domestic hot water and absorption refrigeration units is provided by the water tank. After the heat is consumed, the photovoltaic subsystem directly supplies power to the electric refrigeration unit, while the remaining photovoltaic power is converted into hydrogen for storage. When the photovoltaic subsystem is not working at night, the required power is first provided by alkaline fuel cells and then by the municipal power grid.

The operation results for a typical summer day in Guangzhou are shown in Figure 20. From 8:00 to 16:00, the full-spectrum solar system continuously provides and stores electric energy and heat energy, and the output peaks around 13:00. During this period, the cooling load and electric load can be fully provided by the photovoltaic subsystem, which operates for about 9 h. The heat provided by the system is supplied to domestic hot water first, and the surplus hot water is supplied to the absorption refrigeration unit. When cooling is still needed after heat consumption, the photovoltaic subsystem directly supplies power to the electric refrigerator, and the remaining photovoltaic power is transferred to the electrolytic cell for hydrogen storage. System operation at night follows a defined power hierarchy: the alkaline fuel cell is dispatched first to meet the load, with the municipal power grid supplying residual demand.

These results demonstrate that under summer conditions with abundant solar radiation, the system achieves effective full-spectrum utilization through coordinated thermal storage, electric cooling, and hydrogen conversion, maintaining near-zero grid dependency during daytime hours.

5.4.3. Analysis of Typical Days in Transitional Season

The operation results for a typical day in the transition season in Beijing are shown in Figure 21. There are both cooling and heating loads on this day. The heating load is required from 4:00 to 10:00, and the cooling load is required for the rest of the time. However, the photovoltaic subsystem works only for a short time in the afternoon, so the power required is first provided by the alkaline fuel cell and then by the municipal power grid. Part of the heat required for domestic hot water can be provided by the photovoltaic subsystem. If it is insufficient, it will be provided by the electric boiler, and if it is surplus, it will be provided to the absorption refrigeration unit.

The operation results for a typical day in the transition season in Chengdu are shown in Figure 22. The photovoltaic subsystem has only a small amount of work in the afternoon, so the power required is first provided by the alkaline fuel cell and then by the municipal power grid. The heat required for domestic hot water is provided by the electric boiler.

The operation results for a typical day in the transition season in Guangzhou are shown in Figure 23. On this day, there is only cooling load and refrigeration is required throughout the day. The photovoltaic subsystem works from 11:00 to 5:00. When photovoltaic power generation is insufficient, the power required is first provided by the alkaline fuel cell and then by the municipal power grid. The heat required for domestic hot water is mainly provided by the photovoltaic subsystem. When it is insufficient, it is provided by the electric boiler, and when it is surplus, it is provided to the absorption refrigeration unit.

These results demonstrate that during transition seasons with moderate but limited solar availability, the system relies on a balanced mix of photovoltaic generation, hydrogen storage, and grid backup to maintain reliable operation, with fuel cells playing a particularly important role during non-solar hours.

5.5. Comprehensive Performance Evaluation

Taking $SS$, $SC$, LCOE and ECE as indicators, the technical feasibility and the economic and environmental benefits of the system are evaluated.

5.5.1. $SS$, $SC$

Figure 24 shows the $SS$ and $SC$ of the systems in Chengdu, Beijing, and Guangzhou. Beijing’s annual heating load is 1,335,818 kWh, and its cooling load is 321,860 kWh. The high winter heating demand results in a high $SS$, but the seasonal mismatch between peak summer generation and peak winter demand forces a large amount of summer electricity to be exported, thereby reducing $SC$. Chengdu exhibits balanced characteristics, with a moderate $SS$ and a relatively high $SC$, reflecting its nearly equal heating (602,668 kWh) and cooling (464,746 kWh) loads, which enable on-site utilization for each season. Guangzhou’s heating load is 203,914 kWh, and its cooling load is 962,284 kWh, typical of a cooling-dominated climate with abundant solar resources. However, the massive cooling demand exceeds on-site generation, resulting in a low $SS$.

5.5.2. LCOE

Figure 25 compares the levelized cost of energy (LCOE) for the three systems. Beijing achieves the lowest LCOE, benefiting from its minimal grid electricity purchases. Despite moderate solar resources, the large heating load ensures high equipment utilization, reducing the unit energy cost. Guangzhou exhibits a moderate LCOE. Although solar resources are abundant, the system requires substantial investment in electrolysis and hydrogen storage, thereby increasing capital costs. Chengdu has the highest LCOE, primarily due to greater grid dependence.

5.5.3. ECE

As the core form of a distributed renewable energy supply scheme for buildings, the application value of solar energy systems is not only reflected in the economic optimization of building energy consumption, but is also in line with the core goal of building low-carbon transformation. Guangzhou has achieved the lowest annual carbon emissions, which can be attributed to minimal reliance on the power grid and a high proportion of photovoltaic and concentrated solar power generation. Beijing’s emissions are moderate. Despite a high self-sufficiency rate, winter heating demands require support from the power grid, resulting in substantial carbon emissions during this period. Chengdu has the highest emissions; frequent overcast days increase reliance on the power grid, contributing to higher carbon emissions.

6. Conclusions

To assess the viability of the proposed integrated system and its optimization, this study selected Guangzhou, Beijing and Chengdu as the research sites, taking a hotel as the research case, and also analyzed the annual and typical daily performance dynamics of the three systems. The following conclusions are drawn:

(1)

After optimization, the system in Beijing demonstrates the most favorable economic performance, achieving the lowest levelized cost of energy. This cost advantage stems from the system’s high SS, which substantially reduces grid electricity purchases despite the significant winter heating load. The large heating demand ensures high utilization of the electric boiler and thermal storage equipment, effectively amortizing capital costs. The system in Guangzhou achieves a balanced economic outcome with a moderate LCOE. The excellent temporal matching between its dominant cooling load and abundant solar generation enables the highest SC and lowest ECE. The system in Chengdu exhibits the highest LCOE among the three cities. Although its balanced heating and cooling loads allow for stable year-round operation, its greater reliance on grid electricity during cloudy periods makes its overall economic efficiency lower than that of the other configurations.

(2)

Due to the role of SBS, the power generation efficiency of the new system is significantly improved. The low-frequency spectrum is directly collected and converted into heat energy, which avoids the efficiency reduction caused by photovoltaic temperature rise. At the same time, the recovery of heat energy improves energy utilization efficiency and significantly reduces dependence on the grid. The solar energy system can not only reduce electricity costs but also indirectly reduce carbon emissions.

(3)

Compared with the traditional photovoltaic system, the proposed system can realize the storage of electric energy and thermal energy. After meeting the hot water supply needs, the excess heat can be supplied to the absorption chiller and used to preheat water in the electric heating unit, providing a reference for the full-spectrum cascade utilization of solar energy and distributed energy systems in different regions.

Other limitations of this study should be mentioned. First, the analysis only considers the impact of temperature and DNI on the system, without considering the climate elements such as wind speed. Second, the optimization objective is considered solely from the economic perspective, and future research should consider multi-objective optimization to obtain the optimal solution.