This is probably the most substantive Launch HN thread (I mean of the official YC startup launches at https://news.ycombinator.com/launches) that we've yet seen. And it was very little work to put together.
Rule of thumb for electrolytic water splitting is 50 kWh per kg of hydrogen. That would mean about 0.29 as much H2 as estimated in the blog post. If you scale all salable products by 0.29 (not sure if that's sensible since I haven't identified the root error) it would still yield $142 of products from $70 of electricity. And good news there is that industrial scale electricity can be had for well under $0.13/kWh.
They don't need to do it all alone. If this idea is solid, a lot of different companies and nations could be doing this.
I'm excited for China to have carbon-neutral cement. If they do, it will probably be of their own making and not tied to this company growth.
Rau, G.H., Willauer, H.D. & Ren, Z.J. The global potential for converting renewable electricity to negative-CO2-emissions hydrogen. Nature Clim Change 8, 621–625 (2018).
https://www.nature.com/articles/s41558-018-0203-0?WT.feed_na...
I'd also like to clarify up front that I'm limiting my analysis to the proposed electrochemical mechanism, and have no information on the specifics of Heimdal's proposed implementation - which may vary significantly and materially:
The theoretically minimum electrical energy input (often) winds up being set by one of the reaction intermediates, rather than the overall reaction enthalpy. Typically these analyses are done via determining the corresponding half cell potentials, then counting electrons and computing power via P = IV.[0] So we only need to consider H2O (and its dissociation -0.83[1]) and Cl- to Cl2 at 1.36). That sums to approx 2.19V for the cell, and 2 electrons to do 2H2O + 2Cl- -> H2 + 2OH- + Cl2. That the formed OH-'s pair with Ca2+, and/or CO2 is immaterial to the theoretical electrical efficiency of the cell. Accordingly, the cell energy requirement is on a molar basis identical to the chloralkali process. The only difference is the presence of 2Na+ vs. Ca2+.
Less concretely the lower reactant concentrations have two specific negative effects: Cell potential is adversely affected (per the Nernst equation) and Cell current can be adversely affected if/when depletion occurs. At 0.01M of Ca2+, vs. 6+M Na+[2], current densities could be 1-3 orders of magnitude lower. Cell count (CAPEX) is inversely proportional to current density.
Ultimately making CaCO3 this way (and CaO) winds up substituting a 400$/tonne product in NaOH for an approx 40$/tonne product in CaCO3. It is a very technically feasible approach for turning $$$ into sequestered carbon via non emission from natural limestone.
The analyzed approach reminds me a lot of Calera, who had an apparently similar electrochemical approach to CaCO3.
[0]Not that it can't be done from Gibb's energies, indeed the standard potentials for a reaction can be computed from the delta Gibbs, but the specific species the electrons are being pulled from/pushed into matters. Phrased a different way: electrical energy and overall reaction enthalpy are not necessarily fungible. e.g. (at a simple level) the reaction of CO2 with Ca(OH)2 doesn't affect the electrical energy requirement, nor does CaCO3 --> CaO because neither reaction involves electrons. Any exotherm just winds up 'wasted' as heat, instead of lowering the electrical demand.
[1]Per 2H2O + 2e- --> H2(g) + 2OH-(aq) #6.8.11 https://chem.libretexts.org/Courses/Mount_Royal_University/C...
[2] Saturated near room temp, don't have a better source handy.
You're pointing out the crux of the issue: "Ultimately making CaCO3 this way (and CaO) winds up substituting a 400$/tonne product in NaOH for an approx 40$/tonne product in CaCO3."
I calculated that Heimdal's process has 7x reality factor overhead, whereas chloralkali has 2x reality factor. This is seemingly in opposition to your statement. I think the diff comes from considering all the other energetics - precipitation of CaO from the hydroxide is thermodynamically quite favorable, reducing the overall theoretical cost by 65% relative to the chloralkali process. So to the extent that Heimdal can usefully harness that energy gradient, that's the difference.
I guess the crux of my argument is that it is impossible[0] to capture the energy from CaO precipitation. E.g.[1] one has to spend 2J of electrical to create the hydroxide in Ca(OH)2, but the 1.5J get released on precipitation of CaO comes as heat[2], rather than a reduction in energy input.
So, in summary, Electrochemical CaCO3 would have [approximately] equal theoretical electrical demand to chloralkali and, to borrow your term, a higher reality factor overhead[3], to make a product an order of magnitude cheaper than NaOH.
[0]For a loose definition of impossible, to be fair reality's never quite so black and white.
[1]Made up numbers.
[2]And heat in sub-boiling water is not useful for doing work.
[3]I broadly agree with your position on reality factor overhead, just made less of an emphasis on it trying for 'equivalent but worse and creates less value'
Sorry I don't have the chops to work out if that's economically feasible.
We need zero CO2.
I'm aware of that. There are two main sources of CO2 in cement manufacture: the calcium carbonate raw material, and the fossil fuel required to cook the raw materials into clinker. Heimdal only deals with one. The questions I'm interested in are which source contributes more CO2 and which can be most economically eliminated.
> We need zero CO2.
Agree, but Heimdal won't get us there. We'll either have to find a replacement for cement, or devise a practical method to convert CO2 waste to something benign, like an improved Bosch process.
https://www.moderndescartes.com/essays/carbon_neutral_concre...
TL;DR: ignoring all sources of overhead/inefficiency and purely on consideration of thermodynamic costs, $70 of electricity can generate $100 of lime, $300 of chlorine gas, and $75 of hydrogen gas.
Much more details on the chemistry and economics in the blog post...