But is there not a very mundane explanation here?
1. Stress release in the cathode as H ions is coming out would raise the temperature
2. Recombination to H2 as hydrogen ions come out of the cathode would increase the temperature
3. Possible recombination to H2O would raise the temperature (even If stainless is not a catalyst)...
An increase of the far IR band emissivity of the plates, modified by electrolysis effects somehow, could possibly explain these results. (Surface roughness, reflectivity changes, etc.)
It is not mentioned whether the heat balance between IR input and radiant power from the affected plates is expected to be anomalous, or if the affected plate(s) merely become hotter than other plates while being bathed in IR. Do the IR-hotter plates also cool quicker to ambient temperature when they are raised (together with the controls/others) to a higher-than-ambient temperature by conduction and then the input heat shut off?
See pages 12-13 in the linked report for an example of extreme emissivity modification of 316 stainless in a different environment. (Total emissivity changes from about 0.17 to 0.75)
For example, If the emissivity of the plate changed to only halfway to 0.75 from 0.17 , (new emissivity of 0.46), such a plate could absorb nearly 3 times the IR energy as the control in the same band.