Photovoltaic-Assisted Water Electrolysis

At first sight, given that the cost of water is negligible, the economics of the process of hydrogen generation by water electrolysis is driven by the cost of electricity and by the cost of the electrolyzer. However, because the cost of photovoltaic (PV) electricity consumed has been more valuable for decades than the hydrogen produced, this method has not been used. Hydrogen generated by water electrolysis, however, is an ideal way to store intermittent solar electricity generated during the day. The advantages of hydrogen as a storage medium are self-evident: i) high specific energy; ii) low or zero self-discharge rate (H2 can be stored for years, unlike other energy storage media); iii) it is clean, because no pollution is produced.

Solar hydrogen can therefore be used to fuel the power needs of homes, vehicles or boats, thus enabling decentralized energy generation. For example, 11.4 kg (3 gallons) of water, once split into O2 and H2, contains enough energy, when recombined, to satisfy the daily energy needs of a large home in the USA or in the EU.

The idea of using PV energy to crack water molecules into hydrogen and oxygen and then use both gases in a fuel cell to make electricity when the sun is not shining is generally manifested in the form of a closed-loop and an entirely clean energy system, affording water which is captured and used again (Figure 2.8).

Hydrogen production by electrolysis of water using the electricity produced by PV modules started at the beginning of the 1970s. Solar PV technology does not emit any polluting substance during operation, is noise-free and not does involve any moving parts. Furthermore, PV modules are supplied with a striking 25-year power output warranty (reflecting the fact that, at the 25th year, each solar module will still produce a minimum of 80% of their original power output).

In brief, the PV modules are directly connected to the electrolyzer to generate hydrogen and oxygen (Figure 2.9).

Water electrolysis supplied from photovoltaics is limited to low temperature electrolyzers (AWE and PEM technologies). Although AWE is a mature and robust technology, its corrosive liquid electrolyte and less compact designs mean that PEM technology is a more pro­mising WE electrolysis format for direct coupling with renewable elec­trical sources.17


The purity of the O2 and H2 gases produced by an alkaline electro­lyzer is affected by the current density and temperature of the cells.18 In detail, the purities of the hydrogen and oxygen gases are poorer at low current densities (such as when a cloud covers the sun): diffusion of the gases through the liquid electrolyte is a more significant fraction of the

total production at low current densities. The lower flammable limit, 4% for hydrogen impurity in bulk oxygen, is approached at low current densities, and given that there is a greater danger of having hydrogen impurity in the oxygen than the reverse, most advanced electrolyzers used in PV-assisted electrolysis make use of a hydrogen gas purifier (a catalytic converter that recombines any oxygen impurity in the hydro­gen product, and makes water).

Furthermore, the use of intermittent PV electricity results in two shortcomings: i) its activity decreases with time, and ii) shutdown of electrolytic cells provokes Ni dissolution at the cathode because this electrode is driven to more positive potentials by short-circuit with the anode. These shortcomings can be alleviated if the Ni cathodes are activated, i. e., if they are coated with a thin layer of more active and more stable materials (Figure 2.10).

When the said protective materials are not present, even in recent direct coupling experiments using an advanced PEM electrolysis system, stack degradation is clearly observed (Figure 2.11), affording for approximately 60 days over a four-month period an overall solar-to – hydrogen energy conversion efficiency of around 4.7%.19

The connection between the solar generator and the electrolyzer can either be direct, by feeding the electrolyzer with direct current (DC) generated by the modules,20 or, more efficiently, can be mediated by an electronic, instantaneous match between the maximum power point (MPP) of the solar generator and that of the electrolyzer.


electrolysis time, days

Figure 2.10 Variation of overpotential (inversely proportional to activity) for O2 evolution as a function of time for continuous and intermittent electro­lysis. Under the latter conditions, Ni-based cathodes need to be protected by a thin layer of more active and more stable materials.

(Image courtesy of Prof. S. Trasatti.)



Figure 2.11 Voltage-current data and electrolyzer efficiency collected in real time over a period of five days’ operation for a 13-cell PEM electrolyzer stack in the very early stages of the direct coupling experiment with a 2.4 kW PV array. (Reproduced from Ref. 19, with kind permission.)

In other words, to maximize hydrogen generation in a PV-electrolysis system the operating point of the total system must equal the MPP of the solar generator. This is usually realized by an MPP tracker that guarantees the operation of the solar generator at its maximum power point. In addition, a DC/DC converter shifts the power to the char­acteristic of the load, adapting the output of the solar generator to the input of the electrolyzer (P3 in Figure 2.12, where the labeled PV system and electrolyzer curves represent the I, V characteristics of the PV and electrolyzer systems, respectively).

Figure 2.13 shows the hydrogen flow rate measured for a sample day in July in the case of coupling with the MPPT and, for comparison, the hydrogen flow rate in the case of direct coupling and coupling with the MPPT in relation to the variation of solar radiation intensity. Clearly, the MPPT optimizes the system performance, increasing the system current for the same radiation intensity, which leads to greater hydrogen production.

However, Paul and Andrews have recently demonstrated the possi­bility of achieving near maximum power transfer between a directly coupled PV array and a proton exchange membrane (PEM) electrolyzer stack by finding an optimal configuration of the series-parallel con­nection of both the PV modules and the PEM cells.21 This entirely

Chapter 2


Figure 2.12 Characteristics of PV and electrolyzer systems. The line labeled ‘‘locus of MPPT’’ shows the maximum power for a given radiation intensity. (Reproduced from Ref. 19, with kind permission.)

avoids the costs of an electronic coupling system (around $ 500 per kW, as of the end of 2011).

Table 2.3 shows that the application of this procedure to four (75 W) PV modules directly coupled to five 50 W PEM electrolyzer stacks (250 W in total) predicts an energy transfer of over 94% of the theoretical maximum, while experiments indicate an actual energy transfer of around 95%.

The approach has been generalized recently to afford a new method for relative sizing of both components, based on simple modeling of both polarization (for the electrolyzer and the PV array) curves.22 Modeling and simulation is used to extract a cloud of maximum power points under all the radiation and temperature conditions for a nor­malized PV generator. Subsequently, the ideal ratio between the sizes of the components is obtained by fitting a normalized polarization curve for the electrolyzer to this cloud of maximum power points. Direct coupling of an advanced electrolyzer to a matched solar PV source for hydrogen generation and storage, involving minimum interfacing elec­tronics, does lead to substantial cost reduction and enhances the eco­nomic viability of solar-hydrogen systems.

Подпись: Z
Подпись: QE HAm
Подпись: (2.14)

In general, the overall efficiency of the system (zs) is given by Equation 2.14:23

Where Q is the hydrogen flow rate (mL s 1), E is the calorific value for hydrogen (as a net or gross calorific value, 10.8 J mL-1, and 12.7 JmL-1

Подпись: - Merced on 04 June 2012 Published on 24 May 2012 on | doi:10.1039/9781849733175-00040 image073

at 0 °C and 1 atm), H is the solar radiation intensity (W m~2) and Am is the photovoltaic module area (m2). In other words, the efficiency is defined as the ratio of the higher heating value of hydrogen produced in one year to the yearly total solar energy on the PV modules.

Подпись: 56 Chapter 2

Table 2.3 Comparison of experimental and theoretical energy transfer. (Reproduced from Ref. 21, with kind permission.)

Total direct




Maximum total PV energy available (Wh)

Theoretical total energy delivered to the electrolyzer (Wh)

Experimental total energy delivered to the electrolyzer (Wh)

Theoretical overall energy loss DE%

Experimental overall energy loss DE%

Discrepancy between theoretical and experimental energy transfer (% total energy delivered)











Figure 2.14 Overall system efficiency versus solar radiation measured for the direct coupling system and for MPPT.

(Reproduced from Ref. 23, with kind permission.)

The overall typical system efficiencies in the case of direct coupling and with the MPPT are shown in Figure 2.14. Although the electrolyzer efficiency decreases slightly with the MPPT, due to the higher electrolyte temperature and higher ohmic losses, the overall system efficiency increases in comparison with the direct coupling case. The system operates around the maximum power point of the PV module, and the maximum power point tracker increases the system current and accordingly increases the hydrogen flow rate.

In general, with the efficiency of modern photoconverters and elec­trolyzers being about 20% and 80%, respectively, the total efficiency of solar radiant energy transformed to chemical hydrogen energy is nearly 16%.24 Actually, the overall efficiency is of the order of 10%.25 Indeed, Gibson and Kelly demonstrated a total PV-H2 system efficiency of 12.4% by optimizing the choice of the PV module directly coupled to a PEM electrolyzer working at around 31.7 V and 4.7 A at nominal


Now, some argue that PV electricity, with its typical 200 W solar modules measuring 1.2 m2, would require immense consumption of land to cover even a fraction of the power demand of economically advanced countries.

This is simply not the case. On average, for example, covering 0.6% of the European territory by obsolete, 10% efficient PV modules would


Chapter 2

Figure 2.15 Theoretical PV potential: surface of PV modules (10% conversion effi­ciency) mounted at the optimum angle that would be needed to completely satisfy the electricity consumption of some selected European countries, expressed as % of the country’s area. The European average is 0.6%. (Reproduced from Ref. 27, with kind permission.)

theoretically satisfy its entire electricity demand (Figure 2.15).28 This 2007 estimate was largely conservative because the latest PV module technologies have about 18% conversion efficiency, greater than that considered in the cited analysis (10%), so the area covered per kWp is roughly half and will certainly continue to decrease in the future. Fur­thermore, the PV area does not translate directly into land area covered because a large share of PV panels (currently about 90%) are and will normally be placed on rooftops, which would exist regardless of whether solar panels are installed.

Updated: August 15, 2015 — 6:40 am