It's not difficult. It's just energetically uphill. Undoing entropy costs energy. Second Law of Thermodynamics.
A thousand liters takes about 3kwh. It's not really that expensive. If you run a very inefficient house in the US, that's actually what you'd need per day. You might consider some cost/water saving solutions if that worries you. But, either way, we're talking cents per day per household basically.
Not nothing. But cheap enough that it is a common solution to get water in places that have average incomes far below those common in places like the US where desalination is mostly science fiction.
A large desalination plant means large patches in the sea where life is not sustainable.
BTW, if an average US house really needs 1m^3 per day, that's appalling. These past years, my house has used less than 10m^3 per year and per person. 30× less. I'm afraid most US homes will keep wasting drinkable water and pressure society on building desalination plants, rather than halve (at least!) their water usage and protect the environment.
Molar mass of salt is 58g/mol, and the average sea water salinity is around 3.6%
So a cubic meter of sea water will have 1000*0.036/0.058=620 moles of salt, and it'll require 2.4MJ of energy to remove the salt in a perfect desalinator.
In more common units, 2.4MJ is about 0.75 kWh. Around here electricity is ~10 cents per kWh, so the absolutely lowest price of one cubic meter of desalinated water would be around 8 cents.
> A thousand liters takes about 3kwh.
So, if those numbers are right, desalination is currently at about 25% of theoretical energy efficiency. Is that correct?
But otherwise it's correct, we're at about 20% of the theoretical maximum. The best RO systems are right now working towards 2kWh per cubic meter: https://uh.edu/uh-energy/educational-programs/tieep/content/...
> desalinating 35 g L–1 seawater at 50% water recovery has a theoretical minimum energy requirement of 1.1 kWh m–3 and a practical minimum of 1.6 kWh m–3.
SOTA is apparently ~3.7 kWh m-3. That's not a huge factor
For 90% salt removal with 50% waste-water, they say the limit is 1.09kWh per cubic meter (3.924 MJ)
NB: It is not 100% clear to me if the result is independent of the type of technology, but they do claim:
> We first derive the general expression of the thermodynamic minimum energy of separation determined by the Gibbs free energy, which is independent of the method of desalination
Their result is independent of technology, it's derived from fundamental thermodynamic principles.
Carnot cycle, technically, doesn't apply to all energy sources directly.
For example, solar panels have their "hot side" at around 6000K, so Carnot efficiency would be close to 100%. Real solar panels have other limiting factors, and I believe the absolute achievable theoretical maximum is around 80%.
On the other side of the spectrum, wind turbines have very lousy Carnot efficiency because they're exploiting a temperature difference of just a few degrees. However, the "Carnot tax" is not paid by us directly, so we don't really care about it.
Most raw material processing involves some separation process which requires energy. First, there's gathering something that contains some of what you want. Then there's a phase that often involves breaking big stuff into little stuff and some mechanical separation of easily removed crud. Then there's some chemical step, such as smelting, leaching or distillation, which takes energy and feedstocks to pull the good stuff out of the bad stuff. Then there's getting rid of the bad stuff, which is the source of most industrial pollution. Now you finally have something that's mostly what you want, and go on from there. From desalinization to iron making to fertilizer to oil production, the front end looks like that.
All of those processes are energetically uphill, and all are routinely done on huge scales.