"Currently, the CowaTech technology is at technology readiness level (TRL) 2-3."
Which, for the uninitiated, means it is starting feasibility testing and are just barely starting development. I would bet they haven't solved that in any significant scale, that sort of issue wouldn't have come up yet at their development level.
Intuitively, I'm not a specialist, if water desalinization with there compound was more energy efficient than the current process, they would be on course to demonstrate it. They would not be working on "water bottles [...] for outdoor activities". Since they estimate a 50% efficiency gain would make it viable, it probably consumes about 125% energy compared to the current process.
Correcting my own mistake. If a 50% improvement is required to be more efficient than the current process, it means their process consumes up to twice, 200%, the energy.
Having worked on many R&D projects, I am nearly certain that any claims they make about efficiency are about as reliable as a startup telling early employees that their stock will be worth something. Sure, it could be true.
Frankly I doubt that they've done a detailed enough analysis to know these kinds of things at TRL 2-3, at best they've done a back of the envelope calculation based on lab results at small scale to show it's not obviously stupid, or if they're really on the ball they've got a license for Aspen and have some incredibly oversimplified model that shows it's not obviously stupid.
Really, at this development level, "not obviously stupid" is a pretty positive thing. If they're doing the work right, they are trying to demonstrate that it's stupid every day because that's how you avoid discovering it's stupid after five years of R&D and $50M.
Maybe you are in your area of expertise. I admit I am not.
In computer science, such efficiency gains are rather common. In physics, and thermodynamics in particular, a single 5% energy efficiency gain is huge.
Yes, this is very close to my area of expertise, my previous job title was director of chemical process development but I've got my degrees in materials science and physics so I think I mostly understand what they're doing in a broad sense if not the details at small scale. But yes, a 5% energy efficiency gain would be an enormous deal for some chemicals like ammonia. A 1% efficiency gain there is last I checked 2 power plants per year that could be taken offline in response to the lost demand.
But for most chemicals the energy efficiency is important while material efficiency is more important. Usually there are tradeoffs. For instance I have made a chemical to 99.8% purity without a ton of purification. If it were to 99% then I'd have to spend a lot more energy to reach 99.8% purity (which happened to be required to be useful as a product). But the process to reach 99% is simpler and uses less energy than the process to reach 99.8%. So really it's just an optimization problem where you definitely cannot assume that improving the yield or efficiency of one part of the system will result in an overall efficiency improvement unless you model it all.
The problem is that it is premature to talk about energy efficiency in any remotely meaningful way here because the "big box" you need to draw around your system is vastly bigger when you're talking about a complicated plant doing by my count at least a dozen (including heating, cooling, separations, etc) processes simultaneously.
I'm not saying that their idea wouldn't work in that context, I am simply saying that at TRL 2-3, if this were me doing the work, we would have noted the amine issue and deferred it until a later TRL level because it's theoretically a non-core technology to remove it. It doesn't tell you much about your throughput of your membrane or what salt gradients it works with or how to power it -- that's what I would expect them to be focusing on.
Removing the amine would be like TRL 3-4 maybe, if they're on the ball, but they have to attack the problems most likely to result in failure first. Amine contamination is way lower on a priority list than, say, the membrane rupturing every seventeen days. That's literally just more where I would guess they are at development-wise, regardless of whatever claims they make.
I just cannot trust claims made at that TRL level, at best they're extremely optimistic guesses extrapolated from a 1L benchtop apparatus they are trying to model a whole plant using data from. It's a good thing to do. It won't give a correct answer, but it will give you an idea of what's important at least.
Which, for the uninitiated, means it is starting feasibility testing and are just barely starting development. I would bet they haven't solved that in any significant scale, that sort of issue wouldn't have come up yet at their development level.