The transition to a sustainable energy system necessitates efficient conversion of solar energy into fuels and chemicals. Photoelectrochemical (PEC) and integrated photovoltaic-electrolysis (PV+EC) devices show promise for hydrogen production, but large-scale on-sun demonstrations (>100 W) are lacking. Current solar fuel technologies, including integrated PV+EC, PEC cells, photoparticulate systems, and thermochemical cycles, are mostly limited to small-scale (<100 W) demonstrations. Scaling up these technologies presents challenges related to efficiency, production rates, long-term stability, cost, and sustainability. These challenges vary depending on device design, materials, and experimental setup. For instance, photoparticulate systems may be inexpensive but suffer from low efficiency, while III-V semiconductors and noble metal catalysts offer high efficiency but are costly. Solar concentration emerges as a promising approach to address scaling challenges, enabling the use of more expensive materials while increasing power density and potentially improving efficiency. Thermally integrated PV+EC systems utilizing concentrated solar irradiation offer optimized thermal management, resulting in synergistic benefits such as photoabsorber cooling, reduced recombination, reactant/catalyst heating, and reduced overpotentials. Furthermore, co-generation of fuel, electricity, and heat increases overall system efficiency. However, previous demonstrations have been limited in hydrogen production power (<32 W based on higher heating value (HHV)), with only some tested under real-world solar conditions. Thermal integration is advantageous because it can provide the external heat required for compact electrolyzer systems, thereby removing an additional heater from the balance of plant, which is usually necessary for compact polymer electrolyte membrane (PEM) stacks. Prior work demonstrated a lab-scale integrated PEC device with >15% solar-to-hydrogen (STH) efficiency at 32 W hydrogen power output under simulated conditions, and scaled on-sun systems have also shown promising results, but at a smaller scale. These studies highlight the importance of thermal management, device design, operating conditions, and material limits. Direct electrical and thermal integration introduces coupled effects, previously studied theoretically, where flow-rate control can mitigate degradation and irradiation variations. This work aims to present a scaled prototype of a solar hydrogen and heat co-generation system operating at significantly higher hydrogen production rates, building upon previous lab-scale demonstrations.

The literature extensively covers small-scale demonstrations of solar hydrogen production using various approaches. Integrated photovoltaic (PV) and electrolyzer (EC) systems, photoelectrochemical (PEC) cells, photoparticulate systems, and thermochemical cycles have been explored, achieving varying degrees of success. However, a significant gap exists in scaling these technologies to industrially relevant power levels. Several studies highlight the challenges associated with scaling up, including maintaining high efficiency, achieving high production rates, ensuring long-term stability, managing costs, and ensuring sustainability. The use of solar concentration to achieve higher power densities and improve efficiency has been investigated, showing promise for economically competitive systems. Furthermore, thermally integrated systems, which leverage the waste heat from the PV module to enhance electrolyzer performance, have demonstrated superior efficiency compared to non-integrated systems. Although promising results have been reported at lab-scale, translating these achievements to larger-scale, on-sun demonstrations remains a critical hurdle. Previous work on thermally integrated PV-EC devices, while showing potential, have been limited in terms of hydrogen production capacity and often relied on simulations rather than real-world on-sun testing. This paper builds upon this foundation, tackling the challenge of scaling a thermally integrated PEC device to the kilowatt scale and evaluating its performance under real-world solar conditions.

This study details the design, construction, and operation of a kilowatt-scale solar hydrogen and heat co-generation system. The system utilizes a 7-meter diameter dual-axis tracking parabolic dish concentrator to focus sunlight onto a solar reactor. The reactor integrates a triple-junction III-V PV module and a PEM electrolyzer stack, coupled through a common deionized water stream for efficient thermal integration. A technical illustration of the system shows the parabolic dish, reactor, and ancillary hardware. A close-up view details the reactor's internal components: shield, aperture with flux homogenizer, PV module, and electrolyzer stack. A simplified process and instrumentation diagram illustrates the system's material and energy flows, including PV-generated electricity for electrolysis, heat output from the heat exchanger, and external work for water pumping. Several design challenges were addressed, including decoupling conflicting water flow-rate requirements of the PV and EC using a two-pump design (global and PV recycle) to optimize heat transfer and control water temperature and stoichiometric water ratio in the electrolyzer. An additional water-cooled shield manages excess concentrated light, a design aspect that can be optimized in future iterations. The system was operated for over 13 days under varying environmental conditions, encompassing different ambient temperatures (20°C to 8°C) and meteorological conditions (clear sky to cloudy). Instantaneous and averaged performance metrics were calculated. Hydrogen production rate was calculated from the electrolyzer current, assuming high Faradaic efficiency, and confirmed with gas chromatography. System-level efficiencies (fuel and heat) were calculated based on both enthalpy (HHV) and Gibbs free energy, accounting for auxiliary electricity consumption. A diagnostic device-level efficiency was also calculated, focusing on the ratio of fuel power to solar power incident on the reactor aperture. A zero-dimensional model was developed to simulate system performance, and the model parameters were obtained from optical or individual component performance experiments. The model was validated against experimental data and used for parametric studies to investigate optimization strategies. The effect of variations in water flow rate, number of electrolyzer cells, PV capture area, and light homogeneity were studied. Data was de-noised and filtered for steady-state analysis and used in a Sankey diagram illustrating energy flow through the system. Experimental dynamic response data under fluctuating direct normal irradiance (DNI) conditions and manual flow rate control experiments are analyzed. The impact of pipe thermal losses and opportunities for efficiency improvements are discussed. The paper includes figures illustrating the system overview, temporal system performance, correlations of operating parameters, simulation results, and energy flow analysis.

The kilowatt-scale solar hydrogen production system demonstrated consistently high performance over 13 days. The system achieved a hydrogen production rate up to 0.9 Nm³/hr (49.7 g/hr) with a mean of 0.59 Nm³/hr over the entire operation period, corresponding to an electrolyzer current of ~41.3 A. The peak hydrogen production rate reached 14.0 Nl/min (1.26 g/min) over a 5-minute window, with a total of over 3.2 kg of hydrogen produced. The system produced an average thermal heat output of 10.6 kW at an outlet temperature of 45.1°C; peak thermal output was 14.9 kW, with a total of 679 kWh produced over 13 days. Thermal integration significantly reduced auxiliary electrical demand, estimated at over half (0.6 kW) due to the elimination of an auxiliary heater, and simplified the balance-of-plant. The mean diagnostic device efficiency (based on enthalpy) reached 24.4% ± 2.8%, and 27.2% on the best performing day, which corresponds to 20.3% ± 2.3% and 22.6%, respectively, based on Gibbs free energy. This represents a two-order-of-magnitude increase in solar hydrogen production power (HHV) compared to previous work (32 W vs. >2.0 kW). The average STH device-level efficiency (20.3%) is among the highest reported, comparing favorably with previous thermally integrated PV-EC devices (5-18%). Analysis of process variable correlations showed a relationship between operating current and voltage, and outlet water temperature, indicating that higher temperatures lead to lower overpotentials and voltages. The fuel and heat power showed an approximately linear correlation with DNI, consistent with the model predictions. Deviations at low DNI values were attributed to increased circumsolar radiation during cloudy conditions. Model-based optimization revealed that increasing water flow rate marginally reduces fuel power, increases heat output, and reduces outlet temperature; improvements in PV capture area and light homogeneity significantly enhance fuel power with a minimal decrease in heat power; and increasing the number of electrolyzer cells moderately improves hydrogen production until the PV maximum power point is exceeded. Using realistic improvements in the identified parameters, the system-level STH efficiency is projected to reach nearly 16% (Gibbs) / 19.2% (enthalpy), approaching the experimentally achieved device-level efficiency. Additional theoretical improvements (e.g., using high-reflectivity mirrors and pipe insulation) could increase efficiency further to 19.7% (Gibbs) / 23.8% (enthalpy).

The results demonstrate the successful scale-up of a thermally integrated PEC device for solar hydrogen production, achieving significantly higher hydrogen production rates and efficiencies compared to previous work. The achieved device-level STH efficiency (20.3% based on Gibbs free energy) is remarkable and highlights the benefits of thermal integration and careful design. The system's ability to operate effectively under various meteorological conditions and ambient temperatures further validates its robustness. The model-based optimization identified key parameters for enhancing the system-level performance, providing clear pathways for future improvements. The two-order-of-magnitude increase in hydrogen production capacity showcases the feasibility of this technology for large-scale implementation. The identification of energy losses in the system provides crucial insights for further optimization and points to the feasibility of enhancing overall efficiency. The observed relationship between operating parameters and outlet water temperature underscores the importance of thermal management in such integrated systems. The validated model will serve as a valuable tool for future design and optimization efforts, accelerating progress toward commercially viable solar hydrogen production technologies.

This work successfully demonstrated a large-scale, efficient, solar hydrogen and heat co-generating system. The system overcomes key operational challenges, exhibits fast dynamics, and operates stably under various conditions. The achieved efficiencies and hydrogen production rates represent a significant step towards commercial realization. Future research should focus on advanced control strategies for flexible energy production and integration with energy storage technologies for continuous operation. Exploring further conversion of hydrogen to carbon-based fuels and integration with direct air capture could further enhance the system's value. Long-term stability studies are also crucial to validate system durability and assess the effects of intermittent operation.

The current system design has some limitations that could affect the generalizability of the results. The oversized solar dish relative to the reactor size leads to significant optical losses. The pipe thermal losses contribute to efficiency reduction. The analysis focuses on a specific set of operating conditions and meteorological conditions, and the performance may vary under different circumstances. The model used for optimization is a zero-dimensional model, simplifying several aspects of the system's behavior. Future work will involve addressing these limitations to improve overall system performance and broader applicability.