I don’t know what the costs are, but I suspect part of it is a matter of maintenance, and partly a matter of established tech. Say the flow rate is 400 gph. That’s somewhere around 100x what our systems produce, and also run continuously or at least 24/5 to 24/6. That’s a lot of membranes and carbon blocks to deal with. It may also be that the output is not 0 TDS, but merely removes the items that cause issues in the specific process. This system ran nearly lights out for years other than an occasional prefilter change and replacing the drums of acid at a very predictable rate.
After several years sitting idle we decided to bring the ano line back to life. I remember asking if a high throughput RO system would be more cost effective. The answer was somewhere between “RO, what’s that?” and “This is already here, paid for, and we know how to use it.”
I remember looking at a 2,000 GPD RODI system a couple years ago. It was more cost effective in the sense that everything was well tuned out, but with an effort that only made sense at scale.
For example, my RO unit does things like auto purge at startup, occasionally backflush the membrane with product water, etc. Those extra features have an awfully long payback period if you only make 250 gallons/ year of water. At that rate the up front cost alone is years from breaking even with buying water at the LFS.
A better example might be big vs small tanks. On a big tank you can save money on water changes by putting on a skimmer, a denitrator, and dosing pumps. This makes it cheaper than linearly scaling water changes, but not a ton cheaper.
It was not the same as say one of our parts where the prototype would cost $500 and the real thing would cost $2, but only if you bought a million of them per year.
