Hacker News new | comments | ask | show | jobs | submit login

> Desalination used to be an expensive energy hog, but the kind of advanced technologies being employed at Sorek have been a game changer. Water produced by desalination costs just a third of what it did in the 1990s.

How much of that price drop is due to lower energy prices? How sustainable is the power source?

Even with significant efficiency improvements (3X), shifting from local freshwater sources to desalination would have a massive energy requirements.[1] For such a bold claim ("the Desalination Era is Here"), this article has a rather lightweight examination of the environmental impacts.

[1] https://en.wikipedia.org/wiki/Desalination#Energy_consumptio...

They're quoting 58 cents for a thousand liters. Even if we pretend that is based on free power, we can add the cost of a solar plant and produce each thousand liters for under a dollar. So desalination is completely sustainable.

That's like saying all high-energy industries are sustainable, if we assume sufficient solar build-out. That's a big if. We can't just assume this project is going to be powered sustainably. It's something the article should have explored.

> That's like saying all high-energy industries are sustainable, if we assume sufficient solar build-out.

Sure. Just about all high-energy industries that don't use up limited resources are sustainable.

The desalination itself is sustainable. That's a different question from where the plant buys power, or where it buys parts, or whether the employees use gasoline to get to work and back. There's nothing tying it to buying sustainable power or unsustainable power; that's a significantly different question and while it could fit in the article, it's by no means necessary for the article.

That's verging on overt fallacy.

Every multi-factor process has some limiting factor. Your overall capability is defined by that limiting factor (constraint). Ecology knows this as Leibig's Law of the Minimum.

There's a finite energy flux available on Earth. Excluding nuclear, there are solar, geothermal, and tidal fluxes. Nuclear introduces fission (uranium, plutonium, breeder), or fusion (yet unproven: hydrogen). The only reasonably unconstrained energy sources might be breeder or fusion, and I'm not sure either is long-term viable for other reasons (too long to go into now, though waste / "hygiene factors" play large).

Otherwise, solar is what we've got. 1 kW/m^2 surface flux, modulo capacity factor (0.3), spacing factor (~0.5), conversion efficiency (presently <0.2, single-layer max 0.47, infinite layer max, ~0.85).

The present roughly 17 TW of power consumed by humans would requite a nothin-but-solar region 800 km on a side.

Factor in increased energy for the 6 billion people not living in the US or EU, and increased energy demands of the 8 bilion people (7 presently, plus one more) who'd like to live at US-standard-of-living levels, and you're increasing demand significantly. 85 TW just for population and standard of living would require 3.1 million km^2 -- about a third of the land area of the United States, Australia, or the Sahara Desert. With nothing in it but solar plant. Wall to wall.

Oh, and 1/5 of that needs replacing. Every. Single. Year.

Solar panels, for numerous systemic reasons, degrade with time, and have about a 20 year effective life.

(Research on extending panel life strikes me as more useful than increasing efficiency or massive cost reductions.)

Keep in mind that there are other creatures on this planet and they've got their own demands for land use. We may rely on them in complex ways.

If you're building more power plant capacity than you can sustain, any industry relying on that power won't be sustainable.

1. The roofs of houses and solar cells over roads would cover the need for everything but (maybe) really high population cities. The area to collect sun isn't really a problem, especially not when they get more efficient over time.

2. "and 1/5 of [solar cells?] needs replacing. Every. Single. Year." Huh? The technical life length of solar cells are much higher, as you note yourself.

And certainly, solar cells are not that useful in a Scandinavian winter, because there is just not enough sun. But as it seems now, unless General Fusion or a similar fusion project works out, solar will in twenty++ years supply electricity to most people's normal energy use on the planet. (Few live up in the north.)

Residential power is probably easy with solar. A mill near here uses more power than the 40,000 person county (and I think also uses the waste heat from the power plant, so the nominal electric output of the plant doesn't capture the energy required to run the mill)).

Still, I think the US standard of living is unnecessarily energy intensive (cheap buildings heated/cooled more than necessary, lots of beef, vanity carburetors, it goes on and on and on so all the little things actually add up). It's pretty likely that the US can substantially reduce consumer energy use without really having any impact on standard of living.

Sorry. Five percent, not 20%.

Five percent of a large number remains a fairly large number, especially if yo've got to keep it up forever.

The only comparable infrastructure by size now are highways, though technically they're far less complex. Full replacement of a concrete roadbed happens every 20-50 years or so.

Estimates of rooftop coverage only largely account for electrical use only, and don't account for storage or fuel synthesis, e.g., Jacobson & Delucci (Stanford / UC Davis researchers). I suspect they're optimistic.

David Macay's Without the Hot Air, looking at the UK, is somewhat more subdued.

Electricity is generally about 33% of energy end-use. Transport is another third, and various uses, mostly industrial and heating, the remainder. Generation would have to increase significantly to cover this. Storage would have further efficiency losses. And then there are the capital requirements.

> Just about all high-energy industries that don't use up limited resources are sustainable.

This is an odd statement. Don't most high energy industries use fossile fuels? Such fuels are non sustainable.

We are not yet at the point of such plentiful sustainable energy deployments that we can separate the notion of power utilization from source. High levels of utilization, in absence of specific plans to create sustainable energy supply, must be assumed to exploit non sustainable sources.

> We are not yet at the point of such plentiful sustainable energy deployments that we can separate the notion of power utilization from source.

Sure we are, if we exclude processes that are only barely profitable and also locked in price. Anything else can pay the small premium to get non-fossil power. For a technology to be sustainable doesn't require that specific implementations buy green energy. If it's viable to do so, that's good enough.

Fresh water is very easy to store for prolonged amounts of time, that makes it easier to run a desal plant on intermittent (solar/wind) power than many other high-energy industries.

(However, in practice the capital costs of a plant are such that it usually makes sense to use non-renewable power to let them run 24/7 instead of having them idle when there is no renewable power)

Well, desalination is usually of most urgent practical application in areas which happen to have a lot of deserts (Israel, for example, is 60% desert by land area). Which is exactly the kind of places that are perfect for a large-scale solar build-out from economic perspective.

And I would imagine that desalination plants don't have to work around the clock. If they can be wound down and restarted relatively fast, they can track the sun, and only produce (and store) freshwater while that solar power is flowing in.

So, which part here is a "big if"?

You have situations like at the Ivanpah Solar Power Facility, which is in the Mojave Desert, where the project had to be scaled down to preserve some desert tortoise habitat.

If you're placing solar panels in a sand desert, you have to deal with sandstorms and shifting dunes.

The more land you use for solar panels or heliostats, the more difficult it becomes to find additional usable land.

> "Concerns about the impacts of the Ivanpah Solar thermal project led the developers to hire some 100 biologists and spend US$22 million caring for the tortoises on or near the site during construction."

One of my favorite takes on this topic is Clarke and Dawe's take on the Victoria desalination project[1] (the two are comedians).


It's a bit sad, though - the Victorian government was in a no-win situation with that one. A nine-year drought with the water reserves to a city of 4M people down to 15%? Something had to be done, and the common refrain of "drought will break next year" had been heard every year for the previous half-decade. The plant was over-expensive and should have been bought in a progressive manner rather than full-capacity-at-once, but Melbourne was actually in danger of running out of water. They joke in the video "why not build a dam where it's actually raining?", but the point was, there'd been barely any rain for a decade. There wasn't anywhere that fit that description (and Melbourne has tons of water reserves outside of long drought conditions).

Of course, because Fate is an arsehole, they got the construction underway, and the next year the drought broke. If the drought hadn't broken, people wouldn't be so quick to condemn the desal plant...

> How sustainable is the power source?

Oil (and such and such) will last a 100+ years and obviously won't destroy the planet in that time.

I'd call that sustainable.

And given it's just a "power source", at any time we could swap it out for X, Y, Z.

Plus they specifically talk about efficiencies in the article?

Not sure your point at all?

I haven't seen huge drops in energy cost across the world? Electricity seems the same or more than the 1990s?

"Oil (and such and such) will last a 100+ years and obviously won't destroy the planet in that time."

It is not at all obvious that it won't.

Two words: climate change

Yep. Three words:

Black flag weather

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact