1921
Design Guidelines Based on Thermal Management for Integrated Hydrogen Producing Devices Working with Concentrated Irradiation

Tuesday, 31 May 2016: 15:50
Sapphire Ballroom I (Hilton San Diego Bayfront)
S. Tembhurne and S. Haussener (Lab. of Renewable Energy Science & Engg. (LRESE), EPFL)
The direct conversion of solar energy and water into a storable fuel via integrated photoelectrochemical (PEC) devices is investigated. Particularly, the proposed device uses concentrated solar irradiation to reduce amount of rare and expensive components such as light absorbers and catalysts [2]. We propose a novel integrated design, shown in Fig. 1, combining EC and PV. Detailed device design has been discussed elsewhere in [1].

We developed a 2D coupled multi-physics model using finite element and finite volume methods to predict the performance of integrated PEC device. The model accounts for charge generation and transport in triple/dual junction solar cell and the components of the integrated electrolyzer (polymeric electrolyte and solid electrode), electrochemical reaction at catalytic sites, fluid flow and species transport in the channels delivering the reactant (water) and removing the products (hydrogen and oxygen), and radiation absorption and heat transfer in all components. The various heat source/sink term calculations were treated in detail to ensure accurate energy calculations and to allow for subsequent effective thermal management. The model and its simulation flow was automatized employing multiple interactions between MATLAB Inc. and COMSOL. Efficient computational power saving techniques have been rigorously employed making our model one of the most detailed yet computationally economical.

The investigated triple junction thin film aSi-ucSi-ucSi photovoltaic (PV) cell and dual junction III-V based Ga0.51In0.49P-GaAs cell, both show a decreasing trend in STH efficiency with increasing irradiance concentration (C), see Fig. 2(a)-(b). The low fill factor of aSi-ucSi-ucSi cell made it dependent on both the EC and PV J-V curves for operating at its maximum feasible C; on the other hand the Ga0.51In0.49P- GaAs cell was limited only by the EC’s saturation current. For producing increasing amounts of H2 it is proposed to use high irradiance concentrations which also helps in bringing the device cost down. However due to the limiting saturation current of electrolyser (EC) and fill factor of PV, there exists an optimum feasible concentration above which further concentration will significantly reduce the performance ; however this can be compensated by changing the EC to PV area [2]. The water channel on top of PV if provided with sufficient mass flow rate can effectively cool down the device even at high concentrations.

The J-V curves of the EC, with varying mass flow rates, intersect in a small region of voltage scale giving rise to two distinctive regions of operation- I and II (Fig. 2(c)). Region I is characterized by temperature effects leading to increasing slope of EC J-V curve i.e. changing ohmic losses, whereas region II is characterized by mass transport limitations which lead to increase in EC's saturation current with increasing mass flow rates. Formation of these two characteristic regions of operation leads to trend reversals in the objective functions of STH efficiency and H2 production. Both H2 production and STH efficiency, irrespective of C, are almost independent of mass flow rate after a mean flow velocity of 0.2m/s. The maximum hydrogen production can be achieved at C=700-1000 and mean flow velocity of 0.1-2m/s; on contrary the maximum STH efficiency is observed at C=1-30 and doesn’t depend on mass flow rate. Tradeoff between STH efficiency and H2 production forms a pareto front, shown in Fig. 2(d), consisting mainly of the device configurations with highest mean flow velocity.

Similar analysis have been performed for other parameters like exchange current density, active specific surface area (ASSA), membrane thickness, catalyst and GDL thickness etc. and it has found that the heat sources vary significantly for catalyst properties like exchange current density & ASSA and for dimensional properties like membrane & catalyst thickness; however the heat sources and hence temperature is found relatively less sensitive to GDL porosity and thickness. Even for significant heat source increments it was observed that the temperature changes were minimal and this was due to the choice of optimized mean flow velocity of water (i.e 0.2 m/s); this signifies better thermal management.

The model developed is the most detailed yet computationally economical model of an integrated PEC reported. Model proves to be a valuable tool for design of integrated PEC cells working with concentrated irradiation at elevated temperatures and highlights that the smart thermal management can help in achieving low cost production of solar fuel at fast rates.

References:

[1] S. Y. TembhurneM. Dumortier and S. Haussener.  Heat transfer modeling in integrated photoelectrochemical hydrogen generators using concentrated irradiation. 15th IHTC, Kyoto, Japan, 2014.

[2] M. Dumortier, S. Y. Tembhurne and S. Haussener.  Holistic design guidelines for solar hydrogen production by photo-electrochemical routes. EES, vol. 8, p. 3614-3628, 2015.