In this work, a green hydrogen generation system driven by solar-wind hybrid power with the water electrolysis technology is developed. The system will stabilize the output of hydrogen as the main goal, but also to meet the requirements of large-scale green hydrogen production throughout the year. It consists of wind turbines, photovoltaic arrays, alkaline electrolyzers, energy storage batteries and hydrogen tanks, as shown in Figure 1. The wind turbine and photovoltaic systems are employed as the primary power generation equipment to supply eco-friendly energy for electrolyzing hydrogen production. Concurrently, to mitigate the impact of fluctuations on hydrogen production, battery and hydrogen storage tanks are utilized as coordination equipment for power and hydrogen transmission, effectively enhancing system stability. Furthermore, the incorporation of the power grid enables the absorption of surplus wind and solar power, thereby optimizing the utilization of these renewable sources while also furnishing additional power to the electrolytic cell, further bolstering the system’s stability.

Due to the instability of the wind speed and solar radiation throughout the year, in order to further realize the demand of generating stable hydrogen load, the basic operation strategy of grid-connected solar-wind hybrid hydrogen production system is developed. Firstly, the hydrogen output load is predetermined, and the system utilizes this load as a stable output, with wind and solar power generation serving as the primary power supply sources. Based on the hydrogen load and available wind power, two modes of operation can be identified:

(1) Wind-solar power exceeds the power required for hydrogen output load: In this scenario, the hydrogen production rate of the electrolytic cell surpasses the hydrogen output load, leading to an excess of hydrogen, which is then stored in the hydrogen storage tank. Any surplus power generated is directed towards charging the battery. If the battery capacity reaches its upper limit, the surplus power can be transmitted to the grid for external use;

Compared with the off-grid type system, by considering the power grid connection scenario, the grid-connected hydrogen production system allows for the maintenance of a minimum operating interval for the electrolyzer, thus minimizing start-stop cycles, and also enhancing the hydrogen production capacity and the operational lifespan of the electrolyzer. This operational strategy ensures the fulfillment of stable hydrogen load requirements while guaranteeing the safe and stable operation of each equipment component.

The implementing the operation strategy of solar-wind hybrid hydrogen production system contributes to effectively achieving the goal of stable hydrogen production, meanwhile, in order to make full use of the renewable energy, a grid-connected solar-wind hybrid hydrogen production system is established. The operational dynamics and capacity configuration of this system significantly influence the comprehensive benefits throughout the life cycle, subsequently impacting the expansion and investment in solar-wind hydrogen production. As a result, it plays a vital role in the development of related investments and requires comprehensive performance evaluation methods for assessing its overall benefits, as depicted in Figure 2.

While the evaluation encompasses economic performance evaluation, quantifying the economic benefits throughout the system’s life cycle; environmental performance evaluation, assessing the environmental friendliness during the system’s life cycle; and energy efficiency evaluation, measuring the energy utilization during system operation.

Initially, the key indicators corresponding to each evaluation criterion are defined as the foundation for optimizing the capacity configuration. Specifically, a multi-objective optimization approach is employed to optimize the capacity configuration of wind turbines, photovoltaic arrays, alkaline electrolyzers, energy storage batteries, hydrogen storage tanks, and other components in the solar-wind hybrid hydrogen production system. The NSGA-III algorithm is applied to achieve an optimal configuration that maximizes the overall performance. Subsequently, the optimized scheme is thoroughly analyzed using comprehensive performance indicators, resulting in the development of a comprehensive evaluation methodology for solar-wind hybrid hydrogen production system.

The performance evaluation model comprises three categories of performance indicators, establishing a comprehensive framework for evaluating the solar-wind hybrid hydrogen production system. This model is utilized to simulate and assess the system’s performance by evaluating its economic performance, environmental performance, and energy efficiency.

With the target annual hydrogen output capacity of 20,000 tons, the solar-wind hybrid hydrogen production system defines the optimization objective based on the comprehensive performance evaluation criteria. The following contents are taken as the optimization objectives: (1) minimizing the levelized cost of hydrogen (LCOH) to improve the economy of the system; (2) maximizing the carbon emission reduction (CER) to reduce environmental impact and increase hydrogen production; (3) minimizing the abandoned energy power rate (AEPR) of wind and solar energy to improve the utilization of renewable energy. These objectives are formulated as the objective functions for optimization, while various constraints are employed to ensure system stability. The optimization algorithm, specifically the NSGA-III algorithm, is employed to solve the comprehensive optimization of the system. Additionally, to comply with the requirements of the NSGA-III algorithm, the maximization of CER is transformed into the minimization of the reciprocal of CER. Therefore, the optimization goal is expressed by Eq. 17 min ⁡ f u n 1 = L C O H max ⁡ f u n 2 = C E R W T _ P V + C E R H 2 min ⁡ f u n 3 = A E P R → min ⁡ f u n 1 = L C O H min ⁡ f u n 2 = 1 / C E R W T _ P V + C E R H 2 min ⁡ f u n 3 = A E P R (17)

According to the objective function, the rated capacity of wind turbines, photovoltaic arrays, electrolyzers, batteries, and hydrogen storage tanks are selected as decision optimization variables, and expressed by Eq. 18 X = E W T , E P V , E A E , E B A , Q H T _ s c (18) where E _WT, E _PV and E _AE are the construction scale capacity of wind turbine, photovoltaic array and alkaline electrolyzer cell, respectively.

In order to improve the system reliability, the following constraints should be satisfied. During operation, it is imperative for the system to uphold law of conservation of energy, ensuring that power input and output adhere to the following balance constraints: P W T t + P P V t + P B A t = P A E t + P g r i d t + P a b a n t (19) where P _aban is the abandoned power of solar-wind power generation.

Energy storage state constraints: ensuring that the pressure in the hydrogen storage tank remains within operational requirements. Battery charge and discharge constraints: maintaining the state of charge within the desired range. These constraints are expressed as Eq. 20 J H T _ m i n ≤ J H T t ≤ J H T _ m a x S O C e s s _ m i n ≤ S O C e s s t ≤ S O C e s s _ m a x (20) where J _{HT_min} and J _{HT_max} mean the upper and lower pressure constraint values of hydrogen storage tank. SOC _{ess_min} and SOC _{ess_max} are the upper and lower limits of the battery state of charge.

Power operation constraints: during the system operation, it is necessary to ensure the service life of the electrolyzer and the safe transmission of the power grid. Therefore, the power operation constraints of the electrolyzer and the transmission power constraints of the power grid line can be expressed as Eq. 21 P A E _ m i n ≤ P A E t ≤ P A E _ m a x P g r i d _ m i n ≤ P g r i d t ≤ P g r i d _ m a x (21) where P _{AE_min} and P _{AE_max} are the upper and lower constraint values of the power of the electrolytic cell, respectively. P _{grid_min}, P _{grid_max} are the transmission power limits of power down and power up.

The capacity configuration optimization of solar-wind hybrid hydrogen production system is a multi-objective and multi-constrained optimization problem. Therefore, the Non-dominated Sorting Genetic Algorithm (NSGA) III algorithm is adopted. Compared with the NSGA-II algorithm, NSGA-III uses widely distributed reference points to maintain the diversity of decision variables, which reduces the time complexity of the algorithm and improves the effect on high-dimensional problems (Sharma et al., 2023). The specific algorithm optimization process is shown in Figure 3.

(1) Initialize the scale of wind turbines, photovoltaic arrays, alkaline electrolyzers, batteries, hydrogen storage tanks and other equipment to form a population of N, form initial variables and select reference points on the hyperplane;

The economic parameters of the solar-wind green hybrid hydrogen production system in this study including the construction costs, operation costs, and other relevant economic factors associated with the main equipment, are presented in Table 1. In addition, in order to ensure the rationality of the system simulation and capacity configuration optimization process, the technical and operational parameters of each individual device are kept within the design range, as shown in Table 2. Among them, the maximum pressure of one hydrogen tank volume is 5 MPa, and the maximum SOC of one battery is 0.9.

The meteorological data from a specific region in Taonan, Jilin Province of China is utilized as input for the configuration optimization. The wind speed and solar irradiation data for the area are sampled at hourly intervals over the course of 1 year. Wind speed and solar irradiation are crucial parameters for assessing wind power and solar power potential. The variation trends of these parameters are depicted in Figure 4. It reveals that there is a certain complementarity between wind and insolation, which is better between summer and autumn of resources. Further calculation and analysis of cases are performed based on this weather data.

Aiming at the grid-connected solar-wind hybrid hydrogen production system, the NSGA-III algorithm is employed to address the capacity optimization configuration problem. The optimization objectives include the levelized cost of hydrogen, the reciprocal of carbon dioxide emission reduction, and the rate of wind and solar curtailment. Considering the diversity and operation of the solution set, the population size is set to 500, and the number of iterations is 100 generations. The resulting optimal solution set is illustrated in Figure 5, the Pareto surface exhibits clear patterns, and the distribution of the target solutions appears wide and uniform, indicating that the distribution of the solution set has diversity.

Based on the diversity scheme resulting from the capacity configuration optimization, a final optimization scheme can be selected through analysis. As for the side-view projection of optimization results, it is a curve depicting the reciprocal of carbon emission reduction and the rate of wind and solar curtailment, showing a clear negative correlation trend. Specifically, there is a positive correlation between carbon emission reduction and the rate of wind and solar curtailment. As the rate of wind and solar curtailment decreases, the carbon emission reduction also decreases. This trend indicates that although reducing the rate of solar and wind curtailment allows for increased utilization of renewable energy, it results in a decrease in the overall carbon emission reduction, reflecting a reduction in the overall level of comprehensive accessibility and carbon emission reduction.

Based on the face projection of optimization results and the top view projection of optimization results, it can be seen that among the schemes with lower LCOH, there are variations in the wind and solar curtailment rate and carbon emissions, but the three optimization objectives exhibit a non-linear relationship with each other. Consequently, a single scheme cannot simultaneously achieve the optimal solution for all three objectives. Therefore, a weighting method is employed to select the scheme. Firstly, the scheme is preliminarily selected based on the requirement of the solar-wind hybrid hydrogen production system, that is, f ≤ [25, 1.0 × 10⁻⁶, 5]. Then, the solution of the Pareto solution sets are normalized. Finally, the weighted method is used to get the final scheme. Given the significance of economy in the design process, a weight ω = [0.6, 0.2, 0.2] is applied to sort and select normalized schemes. By calculating the fitness of the solution in the Pareto solution set, the scheme with the minimum fitness is selected as the design scheme. Finally, the corresponding objective function values are f = [16.69, 9.76 × 10⁻⁷, 3.32], and the configuration of each device is shown in Table 3. Through the NSGA-III algorithm, the capacity configuration scheme for the solar-wind hybrid hydrogen production system is determined based on these three categories of indicators.

Through the application of the multi-objective optimization algorithm, an optimized scheme for solar-wind hybrid hydrogen production has been obtained. Based on the capacity optimization configuration results, a comprehensive analysis of the scheme is conducted using various performance indicators. The calculation results are presented in Table 4. Notably, the LCOH is determined to be 16.69 CNY/kgH₂, and Initial cost is 5.60×10⁹ CNY. To ensure economic viability, a recommended sales price of 27.80 CNY/kgH₂ is proposed, along with a payback period of 12.18 years and a total investment profit rate of 10.60%.

Otherwise, under this specific capacity configuration scale, the system demonstrates notable environmental performance indicators, with a CER of 1.02×10⁶ tCO₂. A predominant contribution of renewable energy sources and the grid primarily serves the role of load regulation, resulting in a high proportion of 98.51% for green hydrogen ratio, further enhancing its environmental friendliness. Concerning energy utilization parameters, the wind and solar energy curtailment rate is recorded at 3.32%, indicating satisfactory utilization of scenic power resources. Moreover, the average load factor of the electrolyzer stands at 64.77%, falling within the conventional range for electrolyzers. Thus, it is evident that the proposed capacity configuration method yields a comprehensive and well-performing solution for a solar-wind hybrid hydrogen production system, meanwhile to further enhance the economic benefits, as the cost of related equipment decreases. Therefore, it is crucial to emphasize the carbon reduction advantages of green hydrogen production and integrate them into the overall economic evaluation of the project through measures like carbon taxation and carbon trading. This would ultimately contribute to the reduction of green hydrogen production costs.

In the solar-wind hybrid hydrogen production design scheme, the initial investment cost plays a significant role, as depicted in Figure 6. To ensure the stability of hydrogen production in the electrolytic cell and mitigate the fluctuations of wind and solar energy, an ample amount of energy storage equipment is employed. Among these, the battery construction cost constitutes 27.34% of the total cost, while the hydrogen storage equipment exhibits high capacity, the unit equipment cost is low, accounting for only 0.33% of the total construction cost. The construction cost distribution reveals that the photovoltaic, wind turbine, and electrolyzer components contribute 36.99%, 12.66% and 22.68%, respectively. The electricity utilized for hydrogen production is predominantly sourced from wind power, photovoltaic, and the grid. Specifically, wind power and photovoltaic sources contribute 23.66% and 73.02%, while the grid occupies a marginal portion of 3.32%. This distribution underscores the substantial reliance on solar and wind energy sources in providing energy for the system solution, resulting in predominantly green hydrogen production.

Due to the variations of solar-wind resources throughout the year, the monthly solar-wind power generation and hydrogen production levels differ. Therefore, Figure 7 illustrates the calculated monthly solar-wind power generation and the hydrogen production load of the electrolytic cell. Compared with the two electricity generation, solar-wind power is the main power supply of the system. Concurrently, the power generation of photovoltaic and wind power is the highest in April, which is 1.17 × 10⁵ MWh and 5.76 × 10⁴ MWh respectively, while the power generation is the lowest in November and December. In contrast, the monthly load of the electrolytic cell remained basically stable in the range of 9.06×10⁴–1.03×10⁵ MWh. In order to maintain the hydrogen production and solar-wind utilization rate of the electrolytic cell, as for the transmission power, when the power generation is sufficient, the excess power is connected to the grid, and the maximum power on the grid in April is 6.35×10⁴ MWh. Concurrently, it supplies power to the electrolytic cell when it is insufficient, and the maximum power supply in December is 8.65×10³ MWh.

Further comparing the hydrogen production for each month, as shown in Figure 8, the hydrogen production is consistent with the trend of the load power of the electrolytic cell. The highest hydrogen production in March is 1,713 tons, and the lowest hydrogen production in February is 1,536 tons.

Moreover, the annual economic cost recovery relies primarily on the output of electrolytic hydrogen production and grid-connected solar-wind hybrid power generation. Thus, a detailed analysis of daily electrolytic hydrogen production and solar-wind hybrid power generation throughout the year is conducted, as illustrated in Figure 9. The average daily hydrogen production is 54.76 tons of H₂, the amount of hydrogen produced per day exhibits lower fluctuations around the average. Simultaneously, as the solar-wind hybrid power generation is capable of fulfilling the hydrogen production capacity, the surplus power is integrated into the grid. According to the power integration, the average daily online power is 1,007 MWh, accounting for 8.92% of the total solar and wind hybrid power generation, thereby further enhancing the utilization rate of solar and wind power. However, seasonal variations have an impact, the daily grid-connected power generation fluctuates significantly.

For the solar-wind hybrid hydrogen production system with the selected capacity configuration scheme, dynamic operation analysis is performed on typical weeks representing each month by the seasons.

The power generation profiles of wind power and photovoltaic systems during these weeks, highlighting their complementary characteristics. The dynamic operation analysis of the system is carried out for the typical weeks of the specific months of season. The dynamic behaviors observed during three typical weeks in spring, and three typical weeks in summer, as shown in Figure 10, exhibit similar patterns. Notably, there is a favorable complementarity between PV and wind power generation. SOC of the energy storage device remains relatively stable, while SOH experiences significant fluctuations. This is attributed to the ample availability of solar and wind power generation, causing the battery to reach its upper limit of energy storage. Consequently, the primary objective of ensuring a stable hydrogen supply is primarily achieved through the utilization of hydrogen storage equipment. This operational strategy prioritizes the adjustment of the hydrogen output rate and the smoothing of hydrogen production fluctuations through hydrogen storage equipment. Additionally, the charging and discharging of the battery are employed to regulate the electrolytic hydrogen production rate, further enhancing the stability of the solar and wind complementary hydrogen production system.

In the autumn typical weeks and the winter typical weeks shown in Figure 10, both wind and PV power generation are comparatively lower, with less noticeable complementarity between the two. The generation capacity of wind and solar power is significantly lower during the winter typical weeks. In contrast to typical weeks in spring and summer, batteries and hydrogen storage devices are now adjusted more frequently. Additionally, there are instances where the steady hydrogen output criteria cannot be met by energy storage and wind and solar power generation. In these situations, the electrolyzer receives power from the power grid to meet the standard. In this system, the grid can effectively control the load power of the electrolyzer, ensure a consistent output of hydrogen, and absorb solar power generation, increasing the rate at which solar power is utilized.

However, due to the region’s superior photovoltaic resources compared to wind resources, the former generates more electricity than the latter. The calculations reveal that during typical weeks in spring and summer, the average power generation of photovoltaic is 24,208 MWh and 26,589 MWh, respectively, which is much higher than the average power generation of photovoltaic typical weeks in autumn and summer (17,135 MWh and 17,022 MWh), while during the typical weeks in summer, the average power generation of wind power is 10,425 MWh, which is much higher than the average power generation of typical weeks in other seasons (3,451 MWh in spring, 4,298 MWh in autumn and 4,575 MWh in winter). Despite notable variability in the power generation capacity of power throughout different seasons, the electrolytic hydrogen production system can ensure basic operation through the regulation of energy storage devices and grid infrastructure. Its typical weekly load power remains stable, reaching 22,766 MWh, 24,208 MWh, 23,171 MWh and 22,123 MWh across all seasons, thereby guaranteeing the consistent hydrogen production.

As mentioned above, in order to achieve stable hydrogen output, according to the dynamic operation strategy, the smoothing ability of low-cost hydrogen storage equipment in wind and solar output scenarios is first fully utilized. This reduces the use of the battery and the frequency of charge and discharge, thereby improving its operating life. At the same time, combined with the power grid to adjust, it also ensures the system power balance and meets the load demand of electrolytic water hydrogen production. Consequently, this solar-wind hybrid hydrogen production strategy is well-suited to leverage its advantages in large-scale hydrogen production scenarios.