Overview of mass transfer correlations in electrolyzers
Numerous experimental studies with the aim of establishing mass transfer correlations have been reported. [1-5,12-14,18-37] Table 4 and figure 8 list a selection of these correlations. Generally, they are of the form \(Sh=a\ Re^{b}Sc^{c}Le^{-d}\). For these, the Schmidt power is typically set to 0.33 which is in accordance with the Chilton-Colburn analogy. In rare cases, the relative length of the electrode\(Le\ =\ L/d_{H}\) is taken into consideration, denoted by a non-zero d coefficient. Typically, this is done when a dependency on the electrode length could be confirmed. However, most studies do not vary the electrode length which could be a reason for not including this parameter in their investigation.
[TABLE 4]
[FIGURE 8]
Generally, literature correlations show Reynolds powers between 0.6 and 0.8, which is an indication of a turbulent regime. The correlations established by Picket et al. [13] for developed turbulent flow also follow this trend: for short electrodes the power is 0.66, for long electrodes it is 0.8. Interestingly, most electrolyzers appear turbulent long before the expected transition point of Re = 2000. This could be due to several reasons, such as surface roughness, turbulence caused by the inlet or even imperfections in the cell itself.
The rate of mass transfer varies significantly depending on the electrolyzer type that is used. In for instance the UA16.15 a rate of mass transfer is found that is nearly 10 times higher than the correlations established by Picket et al. In other electrolyzers, increased mass transfer is also observed, though the effect is less extreme. There are several possible explanations for this difference. Firstly, most electrolyzers do not use calming sections and are therefore not operating in developed flow conditions, unlike the one used in the work of Pickett et al. [13] Secondly, significant turbulence is generated by the inlets into the channel, which varies between designs. Thirdly, the design of the channel could be of some importance: square, rectangular and even circular parallel plates have been used in literature. [23,28,35] Finally, within the same electrolyzer variations in performance may exist as a result of small differences in the way the cell is reassembled after maintenance. For instance, with a filter-press design, the layers of gaskets spacers and electrodes may be stacked in slight misalignment. This in turn would result in more mass transfer than expected due to extra protrusions of these misaligned components into the channel. In turbulent flow the surface roughness is also of importance. [29]