Comment by adrianmonk | Hacker News Reader

Comment by adrianmonk | Hacker News Reader

> The solar panels convert wasted sunlight into electricity, one of the most useful forms of energy, at the lowest cost ever in human history. The batteries make this power available at night, enabling 100% utilization of the higher capex systems downstream. The RO desal plant uses this power to strip salt out of fresh water

I wonder if you could store the energy as pressure instead of electricity in batteries.

From what I gather[1], the biggest energy draw in reverse osmosis is pressurizing the water to force it through the RO filters. So, while the sun is shining, fill a tank halfway with salty water, then use solar energy to run pumps to pressurize air above the water. When the sun isn't shining, you now have pressurized water you can let through the RO filters.

Compared to a regular RO plant, this requires high-pressure storage tanks and additional pumps. I'm unsure whether that costs less than batteries or not.

Another part of the analysis is that, if you do build the batteries, they can earn extra money with a side gig stabilizing the grid. Occasionally, when grid conditions get very gnarly, you could shut down the RO plant for a few hours and make a quick buck selling your battery power on the market.

Alternatively, you could pump the water uphill during the day and let gravity supply the energy to force it through the RO filters at night. This means you'll have to build a special reservoir for pre-treated salty water somewhere, though, and you need a lot of pressure, so it would need to be at a high elevation.

---

[1] Source (kind of old): https://www.oas.org/dsd/publications/unit/oea59e/ch20.htm

A fundamental challenge with storing pressure is that water is essentially incompressible. This means that you can’t build a 1000 gallon tank and store any useful amount of energy in the form of pressurized water in that tank.

Mathematically, it’s easy to forget that the PV term in enthalpy is a mathematical trick, not a physical thing, and that it mostly does a nice job explaining physical things when P = atmospheric pressure.

If you want to store, say, 100psi water, the energy you care about is P times delta V, so you need some thing that can change its volume easily, cheaply, and reversibly while under pressure. As a practical matter, this means compressed air or maybe springs. So that water “pressure tank” is actually an air pressure tank that happens to have some water in it too.

There’s nothing wrong with storing pressurized air, except that it’s a heck of a lot more dangerous than storing pressurized water due to the fact that it really does have energy stored in it. (There’s a reason that a PVC pipe filled with 80psi water is a common thing in houses and the main danger is that water escapes if it breaks. A pipe full of 80 psi air is quite hazardous.) There are startups that have played with compressed air energy storage. The resulting gizmos are large.

I would expect a RO plant that is optimized for intermittent use without any energy storage at all to end up being a viable alternative to energy storage, but maybe batteries will end up being cheap enough that this isn’t worthwhile.

Batteries have proven extremely resistant to the scaling of solar panels. I would love to be wrong about this, but overwhelmingly people with an agenda just kind of hand-wave past grid scale energy storage problems that we have generations of evidence against being easily or cheaply solvable.

The thing about this plan that I love though, is that who cares if the desal plant runs at night? Desalination can take all of the power you can feed it and then some, and if it shuts off after sundown, who cares? Stop filling your freshwater reservoir until morning, at which point you just run flat out all over again.

It does matter, though. Suppose you need X amount of fresh water output per month. And suppose that for Y dollars, you can build a plant that supplies X if it runs 24 hours per day.

If, instead, you can only run your plant 8 hours per day instead of 24, then you need to build a plant three times as big, which costs 3Y dollars.

This is where batteries (or other energy storage) may become useful. If you can build them for less than 2Y dollars, then you can build the whole system for cheaper.

What I'm saying is that the freshwater reservoir is a totally free grid scale battery for a desal plant. No need to save it, just make fresh water while the sun shines and drink it whenever.

Energy storage would serve two purposes:

(1) Ensure you always have fresh water available since you're continuously producing it.

(2) Reduce the size of desalination plant you must build by allowing you to run it 24x7.

You're right that you can just store the fresh water, so batteries aren't needed for #1. But the article is talking about how to achieve #2.

Yeah, I'm just saying-- a big desalination plant is a solved problem. Grid scale energy storage has been ten years off for almost as long as nuclear fusion. The original author just hand waves their way through what would be the biggest energy breakthrough in a hundred years to save a couple bucks running a desalination plant.

My question is: should it actually cost 3Y? Sure, if you take an off-the-shelf design and triple it, it will cost around three times as much.

But there should be at least some efficiencies that can be gained. For example, energy conversion is not cheap. Rather than building solar inverters that produce 480VAC and using 480VAC pumps or pump drives, what if the pumps ran on DC?

Or, for more fanciness, what if the pump drives were themselves the MPPTs? Have the solar panels arranged so they produce a sufficient voltage for the motor drive bus even at minimum output, and make the drivers smart enough to run the right number of pumps at the right power and to do so while tracking the maximum power point? That’s a lot of power electronics that could be avoided.

Also, if a plant is willing to store a day’s worth of seawater and brine, then the intake and discharge systems only need to be sized for maximum daily average flow, not peak sunshine flow.

Reply on news.ycombinator.com