BeaverRat,

You don't want to mix up power density, current density and power dissipation. They are all different things and are not interchangeable. Power density is a volumetric variable that is measured in watts per cubic inch, centimeter, liter or whatever volume metric you want to use and has nothing to do with surface area. There is also weight based power density, but that is something else. Surface area is only used to calculate current density and power dissipation.

Unless you want to try and predict how hot your reactor will get, all you need to be concerned with is current density using only amps per square inch and the active surface area of one face of one plate. 0.25 AMPS (not watts) per square inch is the recommended operating parameter to limit plate degradation and 0.5 amps per square inch is the recommended absolute limit.

For our purposes, heat is not a real factor in plate degradation and it can get very complicated trying to figure in all the other necessary variables required to determine how hot your reactor will get such as ambient temperatures, air flow, thermal condutivity of all the materials and the percentages each material will contribute to the whole heat dissipation picture. You don't even want to try and go down that road.

Looking at your opening post, your predicted 50 amps will be divided among 4 stacks with 12.5 amps going to each stack and you will have a surface area of 30.25 square inches...so 12.5 amps divided by 30.25 square inches gives you a current density of 0.41 amps per square inch. That's below the absolute limit, but well above the recommended safe limit. With your configuration you should shoot for a total amp draw of 30 amps or so to play it safe.