Hacker News Reader

Comment by t_mann | Hacker News Reader

Water surfaces are a great place for solar farms (the water provides cooling which improves efficiency), but the real deal are open sea platforms, not lake-based. The engineering challenges seem big, though, I'm aware of only one company deploying those successfully.

But it has promising potential: seas form the majority of the globe's surface, and solar sea platforms seem to be, perhaps surprisingly, biodiversity hotspots. Any solid structure in the ocean will attract dwellers, but those platforms also seem very popular with fish gathering underneath in large numbers.

View on HN

Offshore solar also neutralizes one of solars biggest downsides which is that is uses far more space than traditional generators. Land will only continue to get more expensive so I suspect this will actually become a problem at some point, especially in high density countries.

There are many large arid, desert-like land available in almost every part of the world which can be put to use for Solar. I don't think cost of land is a huge issue.

Well, ideally you want to produce the electricity close to where it's needed, and a significant proportion of the world's population lives in coastal areas (much more than near deserts I suppose).

This. You can't produce all of the worlds electricity in deserts and ship it around the world

Why not? Australia is going to be exporting its solar power to Indonesia through a 2,800 mile direct current cable[1]. Thats approximately the width of the mainland united states. It won't be as efficient as generating it nearby but if its already commercially viable then it will only become more viable as solar becomes cheaper and more abundant.

[1] https://en.wikipedia.org/wiki/Australia-Asia_Power_Link.

COuntries will be looking to become energy independent, especially with the russia fiasco Europe is facing. At the very least they will try to make sure a majority of their energy isn't imported from one country.

Don't tell me what I can't do!

HVDC transmission losses are only around 3.5% per 1000km.

So you could get around 80% of the energy you produce in Mali to Norway. Doesn't seem so bad.

Furthermore, there is strictly negative benefit to siting solar farms in deserts. It makes them run hot, thus less efficiently, and shortens their lifetime.

Siting solar on farmland increases farm yield and cuts water loss.

There's a proposal to use solar power in the desert to create pure Hydrogen gas via electrolysis, then fill autonomous blimps with that hydrogen to ship it to where energy is needed. You wouldn't want to fly over populated areas, but otherwise seems like a feasible idea to transport that solar power.

For that you'd need to use kilotons of water a day, which is usually hard in a desert.

It could work on an arid sea shore though, say, in North or South Africa. But there seem to be closer-by large consumers, and maybe producing freshwater would bring more value than producing hydrogen.

> you'd need to use kilotons of water a day, which is usually hard in a desert

North Africa, India, Australia and other have deserts right next to the ocean.

There is absolutely no need for deserts to site solar farms, so the premise is nonsense.

Solar farms are wholly compatible with agriculture. Look up "agrivoltaics".

There are currently experiments in europe, to combine the too. Turns out alot of plants just need 2 hrs of sun and shut down after too save on evoparation. Meaning 35 % of the sunlight needed and the rest as solar seems feasable.

As a thought experiment I once imagined electric trains running between populated areas and desert solar farms. Maybe they carry "green hydrogen" or perhaps just big batteries. After a few minutes and a good chuckle I moved on.

We have these 'electricity trains' already in a sense, but they are tractor trailers containing fuel oil. It's not as crazy as it sounds, but there are better solutions available.

Usually, the issue with this kind of thing is that

* this is a form of energy transmission that requires labor for the actual transport (the train crew) and so is automatically much more expensive and difficult than a dumb pipeline or power line which requires much less staffing

* it's hard to create energy storage that isn't also a bomb in the wrong conditions, and train tracks pass through populated areas. Fuel already has restrictions on where it can be routed because there have been fuel train explosions.

You very much CAN send electricity long distance and there's been serious discussion of providing electricity for Europe from north Africa.

Producing hydrogen near the ocean and shipping it on a tanker is another realistic alternative.

There were huge projects started to site solar farms in Africa, serving Europe.

They collapsed. The reason they collapsed was that there turns out to be less than zero value in siting solar farms in the desert, and solar panels have got so cheap that you do better posting more of them nearby instead of paying for the long cable.

The reason solar farms in the desert have negative value, vs. siting nearby, is that panels in the desert get hotter, so run less efficiently than over water or plant life, and last many fewer years. Furthermore, panels in farmland improve yield and water demand.

The question was if it's technically possible - and it is.

Projects like https://en.wikipedia.org/wiki/Desertec were based on the cost estimates at the time. Now solar panels are much cheaper and it's a good thing.

The cheaper the panels are, the closer by you want them.

You also can’t produce all the clean energy via ocean based solar and ship it to where humans live.

Offshore cables exist, and if you want to go the route of producing it into some intermediate form, a good majority of heavily populated areas are also ports.

Yet, you really can.

If you put solar in the desert SW, and intend for it to power homes in New England, how much does the transmission lines to move that power cost? And how much do you lose in transit?

There is an efficiency loss, but the problem is actually the cost of building and maintaining the infrastructure.

Theres plenty of resources out there around this idea.

I don't recall where I had seen it, but there was a study at some point on the cost of covering the Sahara in solar panels. The CO2 emissions from the resources for the transmission lines- in steel and concrete especially- meant it would be a net positive CO2 contributor even after shuttering fossil fuel power plants.

This means that we need to learn how to build transmission lines with less steel and concrete. Use more aluminum maybe? More plastics and carbon fiber? More glass?

plastics and carbon fiber are made with hydrocarbons generally. Aluminum might actually be worse, since aluminum requires very high amounts of energy to produce.

Solar energy used to produce aluminum has zero marginal cost. So, unless you assume use of fossil fuels to produce the power used to refine the aluminum, this is nonsense.

Hydrocarbons are not a problem per se. It's burning the hydrocarbons, specifically the carbon part, which has the detrimental effect on climate.

Plastics and carbon fiber effectively keep the carbon from becoming CO₂.

Aluminum takes a lot of energy, but, unlike steel, the process does not release any CO₂, and is fully electric. It can be powered by hydro (and often is), nuclear, or solar energy directly.That's the point.

Another idea I saw on yt a few months back:

Charge some solid state battery (it's not battery it was something better) and then transport it.

Similar on how we ship oil across the planet.

Robert A. Heinlein wrote a book called Friday that had something like this. They would use huge solar arrays to charge proprietary solid state batteries called Shipstones which were used everywhere. There wasn't even a power grid anymore because people would just buy a Shipstone and put it into their house and replace it every so often like coal in a coal bin.

I love Friday and just reread it last weekend. Often I think that Elon Musk believes he is a character in a Heinlein novel. Tesla is Shipstone, the cars are just a mechanism to build better batteries. SpaceX, that's just DD Harriman from the man who sold the moon/sail beyond the sunset. The Boring company is how you build Luna City ala The moon is a harsh mistress. Or maybe a space catapult. Or maybe, I'm just hoping for Heinleinian hero.

Anti-hero, more like. Or, in the instant case, over-funded buffoon.

A big problem with that is energy density. Hydrocarbons are on the order of 50MJ/kg. Li-ion batteries are on the order of .5MJ/kg.

Lithium tech at this exact instant is not the end state of battery development.

So, not a big problem, a speed bump. Check back next year.

I think the problem here is the weight of the batteries and the need for power to move them. Maybe hydrogen by electrolysis would be more favorable, but then you need to pipe in water.

You don't necessarily need to pipe the water to the electricity, you can move the electricity to where the water is (and if the electrolysis works with seawater, you don't need freshwater either). And then you can turn it into something like Methane or Ammonia for longer term storage. And the long term storage is important to level the power production over the course of months.

I think the issue at hand is that long distance transmission of electricity has transmission loss, and is expensive from an infrastructure standpoint. The broader point is that remote solar generation presents some challenges.

Transmission loss matters very little anymore. You just add more panels at the source. Marginal cost of the loss is zero. It is very, very cheap to deliver power by transmission line.

Yes, I made exactly that point above. But the point I'm making is that long distance transmission of water is also expensive from an infrastructure standpoint, possibly even more expensive than power (and also sometimes has transmission losses, depending on how they're moving it)

> But the point I'm making is that long distance transmission of water is also expensive

Oh yeah, in case I wasn't clear that's exactly what I meant also.

The dream of covering a chunk of the Sahara in PV to power Europe and Africa seems like tilting at windmills.

But not for the reason you think.

The right place to put PV for Europe is in Europe. You will want solar farms in the tropics producing ammonia to ship around to wherever the wind flags for a few weeks, and to Finland in winter.

Transmitting power is not, in fact, expensive.

Building an HVDC transmission line has capital cost, but marginal operating cost is near zero.

Desert-like land is terrible for solar panels because the solar panels get too hot (and lose efficiency).

The big, successful desert solar power projects I've seen are giant arrays of mirrors using some kind of weird heat engine, rather than PV-cells.

Cheap, widespread desert solar power would be good to figure out, but with regards to PV-cells, desert environments look like a no-go.

How much efficiency do they lose? Desert land is orders of magnitude cheaper than German farmland, and the latter appears to be economic to cover in PV panels, so if we could get the cost of materials down we could deal with losing 20% or maybe even 90% efficiency.

I don't know the answer to your question, but lets assume 20%.

If PVs are 20% less efficient in the desert, you'll need 25% more PVs. (Ex: Instead of buying 1MW worth of panels, you need 1.25MW of panels to generate only 1MW of power)

If PVs are lol 90% less efficient in the desert, you'll need 900% more PVs (instead of buying 1MW worth of panels, you need 10MW worth of panels to make 1MW)

-----------

Because PVs are the "expensive part" of solar power (and land is really, really cheap, even German Farmland), we're more interested in reducing the number of PV-panels to buy, rather than reducing the cost of land acquisition.

Panels that normally get 20% conversion efficiency are typically 12% in the desert summer. So, a huge loss.

You want your solar panels mixed into farm and pasture land, instead. Which there is quite a lot of.

Land use is absolutely no problem at all for solar.

Solar coexists synergetically with agriculture, operating more efficiently, cutting water demand, and improving yield. Also with parking (keeping cars cooler) and roofs (extending life).

I'd be curious if there are sources showing that solar panels could improve agricultural productivity, but I have doubts, since plants and panels compete for the same resource, sunlight. In terms of cooling, it should be noted that contrary to popular belief, solar panels heat up whatever is underneath them, they don't cool.

You may doubt, but most plants can only use a small amount of light. Once that is exceeded, they endure any more, and benefit from relief.

Some crops lose a bit of yield, shaded, like wheat, but the value of power year-round exceeds the loss.

The cost of land might to be a huge issue with deserts, but the cost of transmission is.

Yet, the cost of transmission is not. A long transmission line might be expensive to build in the first place, but operating expense is near zero.

That is not the reason solar in the desert is a dumb idea. The dumb idea is putting it where it is maximally hot, instead of spread out in farmland where it is actively beneficial and also generates revenue year-round.

Transmission from Arizona or Nevada to California looks reasonable. The cost may be reasonable, too, at least the Nevada part.

The eastern half of Southern California is already desert so you might as well just use that, though one of them had trouble getting of the ground and costs weren't the greatest: https://en.wikipedia.org/wiki/Crescent_Dunes_Solar_Energy_Pr...

Others, including https://en.wikipedia.org/wiki/Ivanpah_Solar_Power_Facility are doing better

Except, desert is a dumb place to put solar.

Great idea but then again we come back to the problem of getting that power back to land. I imagine ocean currents complicate linking these things together or with big cables back to land - not to mention big storms. I do really like the idea of fish gathering below the platform - multiuse stuff is always great. We could probably also do some hydropower while were at it!

Best use case for something like this seems like for powering remote outposts and islands - land is at a premium at these places often plus you probably do not want to be clearing large areas of uniquely biodiverse land for power delivery. Some giant ones off the coast of Fiji could be a cool one for sure!

Actually, you can manufacture ammonia from just water, air, and electricity. And ammonia not only has great uses (right now we burn a huge ammount of NatGas to make it) in agriculture, but it's quite energy/space efficient for shipping. We already ship quite a bit of liquified ammonia (<200 psi at >100F). If you have a lot of "free" energy in the ocean it makes sense to make/store/ship ammonia on the spot.

This of course also enables storage for delivery of power at night.

I think there is some physical-chemical development work left to do for maximally efficient commodity-grade solid-state fuel cells that work with ammonia and air, but it is eminently doable today. In the meantime, you can burn ammonia in combined-cycle turbines where NG is burned now.

Thanks, with solar and wind there do seem many times when there is a surplus of energy. I'm thinking somethings like water cracking to H, Aluminium smelting or Steel Furnaces could be done with such supluses but Ammonia sounds like a good one for the list too.

All of it will happen at the same time, different places.

I would imagine the problem for offshore solar farms wouldn't be bigger as for offshore windmills? That is a relatively expensive, but also not that difficult problem today? What did I miss?

Waves.

Are you saying that waves don't exist around offshore wind farms?

Waves will tend to destroy floating offshore solar farms. Wind farms are less affected because their cross-sectional area near the water surface is small relative to their mass.

But if the panels were in a semi-flexible frame with underwater floats, the waves could be made to mostly run under them.

> ocean currents complicate linking these things together or with big cables back to land

They would be connected by cables to the ocean floor (depends on depth I presume). But the currents shouldn't be an issue. Storms unless there is debris should also be fine (but I don't know what ocean weather is like).

Cables back to land are indeed an issue, especially with high energy. Maybe we'll come up with efficient energy guns, or create fuel on said platforms (hydrogen?).

"Maybe we'll come up with efficient energy guns"

Very likely not. Creating fuel seems more practical, but anchoring floating objects is a known problem as well.

30 ft (10 m) waves on a regular basis would not be unusual

Presumably, that also affects cables for offshore wind farms, and yet we routinely build those in seas that can get rough.

A big enough farm can also be a focused microwave phased array antenna, beaming power to any point overhead. Or even multiple points, given more than one frequency.

> only one company deploying those successfully.

which one? i specifically recall a membrane based design, and the other a floating grid

also, are there estimates of costs ?

Edit: lake floats are ~30% more expensive at the system level. i am asking about ocean systems

https://swimsol.com/

Disclaimer: I know someone who works there. These are meant for ocean deployment, right now I believe they mostly use them in locations with somewhat tempered wave conditions like atolls and bays. It's a floating grid structure that holds the panels a few feet above the water. I don't work there myself, for cost estimates,... you'd have to contact them.

At last report there were 3 GW of floating solar. That number is surely higher already.

Reply on news.ycombinator.com