Rednight

You are indeed correct on the ammonia being way more efficient for hydrogen storage (urea is a good candidate too). I think the refinery should have defined cell volumes (for mass flow) and densities (for part mass) for all the processes (some processes can use the same cells) so you can balance all of them properly (I didn't dive into the code yet so maybe you already have this convention). This way the mass flow from the different reactions can be precisly calculated. And as inline refinery can already do alumina electrolysis wich has high energy requierement I see no problem to do the ammonia one. (For information: industrial alumina electrolysis need a cell temperature above 900°C and strong currents, this is a lot of energy - source: http://www.balcoindia.com/operation/pdf/aluminium-production-process.pdf) Another thing is most of the numbers are from current electrode geometries (plain rod-like or parallelepiped-like), but you can find in the litterature meso/nano-porous electrodes that have way more specific surface areas thus cells with higher yields for a given volume. It could be an upgrade feature for the refineries (like the radiators) and a good excuse to have high yield cells to fuel engines (but the calculation must be done first ^^). Maybe Northstar can solve the issue with a simpler formalism.

Hi, Liquide ammonia electrolysis requires 7.4 MJ per kg of H2 (gaz) using a standard voltage of 7,7 mV assuming the eletrolysing cell is 100% efficient (a standard steel/pt/ceramic cell is 71% efficient IIRC) To compare Water electrolysis is 130 MJ/H2 kg (gaz) assuming a 100% efficiency and need higher voltage (> 1.23 V) So ammonia electrolysis is approx 94% more efficient than water electrolysis, but the energy needed to get liquid hydrogen should not be forgotten (I'll come back and edit this post with the correct value) As far as I know, ammonia electrolysis is good to make direct H2 feeding for fuel cells, however I'm not sure it'll be enough of a fuel flow to use it in direct engine feeding (calculations must be done to be sure of that) Cheers. Source: http://www.rsc.org/suppdata/cc/c0/c0cc01982h/c0cc01982h.pdf