Hacker News Reader

Comment by jessriedel | Hacker News Reader

Can anyone point me toward Musk's reasoning about why solar will beat out wind power in the long term? I know he owns a solar company, but why did he choose that over wind?

(I'm aware of the basic pros and cons of both. I'm really just looking for Musk's thinking.)

View on HN

There's a recent talk here where Musk discusses what he thinks of the future of energy, though I don't think he goes into wind vs. solar specifically (he does explain in the Q&A at the end why he doesn't think biofuels make sense).

http://www.treehugger.com/green-investments/tesla-ceo-elon-m...

I could not find him comapring solar and wind, either.

On the risk that you already considered all of these, here are some reasons I can think of:

http://www.withouthotair.com/c4/page_32.shtml gives 2W/m^2 for wind and http://www.withouthotair.com/c6/page_39.shtml gives 22W/m^2 for photovoltaic.

Yes, there are places where you can get much more energy from wind per m^2, but you can get that photovoltaic output almost everywhere.

It also is easier to start a solar business than one based on wind power. With solar, you can start small and scale up. With wind, you almost have to start with huge structures, as they have better ROI (higher = faster wind = way more energy). Building such larger structures means that you no longer operate in the B2C market. He may not like that.

Also, he may not like having to spend effort and time getting building permission for those structures.

He may also think photovoltaics will get cheaper faster than wind.

Finally, it might be just a personal preference. To me, wind has more of a blue collar feel than photovoltaics (someone will get his hands greasy operating a wind farm). That does not make me dislike it, but might have swung the choice for him.

> 2W/m^2 for wind and ... 22W/m^2 for photovoltaic.

This is a misleading number for my point that wind is denser than solar. For solar, the 22W/m^2 is spread out evenly an area, and you must put something (either photovoltaic or a mirror) to collect it. For wind, you only need a single turbine to effectively cover a relatively large area. The density number comes from the fact that, by virtue of how winds are generated and blow, you can't just stick a second turbine right next to the first one. From a cost per kWh perspective, the important number is more like the number of watts generated by single turbine with a 1 meter x 1 meter footprint.

Misleading in one sense, but revealing in another.

That 10MW wind turbine on your 10,000m^2 of land will only get to that output by using wind blowing into it from other land.

So, if your neighbor builds another such turbine on his land, your output will drop.

So, you will have to pay your neighbors for not building such a turbine on their land. In that sense, you really need that area to get that output.

Having said that, ROI of wind turbines seems to be way better than that for photovoltaics, in almost every aspect. https://sites.google.com/site/anatomyofglobalclimatechangevj..., for example, gives you about 6 months to recover the energy investment of building a wind mill and 3 years for photovoltaics.

> So, if your neighbor builds another such turbine on his land, your output will drop.

This is not an important effect. The optimal spacing for wind turbines is 5-15 times the rotor diameter, and the largest turbines are 75ft in diameter, giving a largest spacing of roughly ~ 1/4 mi. In the rare cases that this is an issue, property liens can handle the conflict.

The largest turbines are more like 300ft in diameter (http://en.wikipedia.org/wiki/Wind_turbine_design#Turbine_siz...)

Also, the effect is important. At 1/4 mile distance, you can only place about 20 or so on a square mile of land. Looking at http://en.wikipedia.org/wiki/Wind_turbine#Horizontal_axis_wi..., that gives you less than 18MW peak power.

That same square mile has 2M plus m^2. So, you would need only 10W per m^2 peak solar power to beat that wind turbine.

Yes, there is lots of hand waiving here. For example, I am assuming equal duty cycle; that will vary with location. The crux still is that there is way less wind power than solar per m^2.

Is he actually rooting for photovoltaics? I personally have a number of issues specifically with photovoltaics that I do not have with many other types of solar. IMO any kind of solar "plant" shouldn't be using photovoltaics.

Could you explain why?

The alternatives seem simpler, specifically solar thermal collectors[1] where they can be used in conjunction with steam turbines and molten salt heat storage.

The most obvious disadvantage is that energy collection ability drops sharply on clouded days and it relies entirely on clear skies.

[1] http://en.wikipedia.org/wiki/Solar_heating

I don't understand how a solar thermal collector is in any way simpler than a photovoltaic array. Actually, it seems more complex to me. Again, could you explain?

They are simpler than photovoltaics in the sense that a mechanic can repair them. I like them because they:

- Use simple materials, not a bunch of rare earth metals that the USA doesn't have

- Will last as long as you maintain them (to my knowledge) unlike solar panels

- Have better efficiency compared to solar panels

You are right that they are not as "simple" as slapping a panel on a flat surface and wiring it up. Clearly, they are not suited for installation on houses. However, power plants have always used more sophisticated systems because they can handle the tradeoffs to reap the rewards.

This is a bit like saying a vacuum tube TV is simpler than an LCD one because a mechanic can repair it. It's a myopic view.

Are you sure a thermal solar installation will last as long as photovoltaics? Any plant with the efficiency gains you describe has gimballing mirrors, highly thermally cycled metals, steam circuits, and freakin' turbines among other mechanical contraptions all over the place. I see a whole lot of disadvantages compared to a completely solid-state, arbitrarily scalable panel that lasts (conservatively) 50 years.

I realize some places might get really neat solutions out of thermal solar, but I just don't see how it's anything more than a niche. By contrast, photovoltaics have the prerequisites to take over the entire market from the bottom up. They are our future.

Solar panels are nice in that if one fails, you just replace it. That's also why they are not nice; all the upsides and downsides of a disposable unit-based system.

I'm curious where you are getting this "conservative" estimate of 50 years. I typically hear 20 years, 40 years tops.

photovoltaics might be the future, but they aren't the today as best I can tell. They are inefficient, and much like the Prius batteries people like to talk about, are pretty environment-unfriendly to manufacture.

I would speculate (but don't know) that concentration of the heat energy to power a more conventional steam generation plant would be more efficient, at least beyond a certain scale. The best photovoltaic panels are about 20% efficient under optimal conditions.

They are also closer to the consumer, negating a big chunk of the transmission losses.

Also, while thermal electric plants may be more efficient at large scale, I'm not sure how that works at small scale, or if there is a difference in how much capacity they lose in cloudy weather compared to photovoltaics. Actually, I wonder what the ratio of installed photovoltaics to thermals is per region and generator scale.

Thanks much.

I believe I remember him pointing out, of all the things that need energy on earth, most of them use the sun as their source either directly or indirectly. Plants are obvious, and animals rely on plants. While products of evolution aren't perfect, they can often be a useful guide.

Wind is really an indirect result of solar energy as well -- solar energy creating different pressures in the atmosphere, which produce wind.

You are incorrect in stating that most things use the sun for energy. In fact, everything uses solar energy in one way or another.

Bacteria that live on nutrients emitted from deep sea hydrothermal vents do not use solar energy. Some theorize this is where life on Earth originated.

http://en.wikipedia.org/wiki/Hydrothermal_vent#Biological_co...

Doesn't that energy ultimately come from tidal forces in the Earth's interior that wouldn't exist if there weren't a Sun to orbit?

Earth's spin is due to asymmetry in the accretion of matter during planet formation. The dust and gas that would form the planet was orbiting the center of the nebula before either the sun or planets formed out of it; interstellar nebula begin rotating and flattening before they form a proto-star. If you could pop the sun out of existence, hydrothermal vents would continue feeding bacteria with absolutely no solar energy until the planet eventually shed too much its internal heat (from its initial formation, also not from the sun).

There is nuclear energy transformations inside planets that creates lots of energy:

http://physicsworld.com/cws/article/news/2011/jul/19/radioac...

Geothermal and nuclear are basically the only forms of energy that are not ultimately solar-derived.

Edit: I guess you could throw tidal in there as well.

Well, it depends on how far back you want to go. All elements come from nuclear reactions in suns/stars which occur partly due to gravity. Some elements are unstable making them radioactive, some of which we use for nuclear fission. Big chunks of stars form planets and the associated geothermal energy. So, solar reactions and gravity account for all energy either directly or indirectly. Gravity is also the cause of tidal energy.

Those don't count as we can't access them directly. The idea with solar is we can access it directly, but we are using indirect methods.

Nuclear.

Well, you could say the supernova that created our radioactive heavy elements billions of years ago were once "suns". But that power didn't come from our sun.

360 by 360 degrees of solar energy spilling off into space, near enough.

He also owns a rocket company. Coincidence?

You just blew my goddamn mind.

He's thinking so long term, he's planning to corner the market on space-based solar after he's converted the world to run on electric power.

Kardashev 2 or bust.

He isn't in the electric car business. He's on the empire business.

Besides Musk's viewpoint, I'd argue some simple math. There is vastly more potential power being emanated from the sun every second than there is in the prevailing winds on Earth's surface. The atmosphere being in differing states of high and low pressure is caused by multiple factors, but heating from the sun is one of them. In the long term, almost all of our power is derived from the sun (with exceptions like geothermal, even hydro is dependent on the water cycle being driven by solar heating) anyway - might as well go directly to the source.

I just easily see the entire planet being powered by some terrestrial satellite grid of solar panels in a few centuries. Capture it in space, transport it to surface via a space elevator or something, maximum efficiency. Would probably also slow global warming since significantly less heat by that point would be reaching the surface (given the size of an artificial ring belt of solar panels we would need to run everything by then!)

As I said elsewhere, I believe there is much more wind and solar energy each than humanity's power demands. It's not a question of total energy, it's a question of cost. And my intuition says that wind would be cheaper because it is of higher density. And in fact wind is significantly cheaper in the current situation where renewable sector is so small that variability and storage isn't an issue.

Wind technology seems quite a bit closer to the best possible implementation than solar, so I would expect more improvement in solar pricing than wind pricing.

If solar improves enough, the capital increments become people going to home depot to buy a 1 kW panel.

If you are capturing energy in space, and channeling it to the ground somehow, how have you reduced net energy reaching the surface of the earth?

The energy is in a different form which won't necessarily raise the temperature of the earth as much as the sunlight hitting the earth would have. Not to mention the hydrocarbon based energy sources replaced by the clean electricity wouldn't be contributing further damage to the environment.

When that energy which has been beamed down from space is used, it WILL give off heat.

Long term, climate management will be necessary regardless.

Wind is far too variable. Even if you have a very well sited wind-farm it could potentially be weeks or months when the effective power output is zero. This is ruinously impractical as a base load power system, whereas solar is far, far more predictable (even with weather variation). The big problem that solar has is storage, which hasn't been worked out yet and probably raises the cost by anywhere from a factor of 2 up to an order of magnitude.

Seems like you could easily diversify away most of the wind risk by building wind turbines in many locations and just moving the power (with loss). Also, whatever the cost of solving the storage issue for solar also solves the wind issue, unless storage is only possible over some intermediate time scale. My understanding is that most of the plausible storage mechanisms (pressure, gravitational potential) can store essentially indefinitely, although maybe total capacity is an issue.

Without more info, I really don't think it's clear cut.

You'd think so but that's not the case. Sometimes there isn't enough wind power in an entire continent to provide power, and this could potentially last for days, even weeks. It's simply not practical to use wind power as base load power.

If you look at the Netherlands, for example, you'll see that the way they've always used wind power is for pumping water (draining the polders). And there it doesn't much matter what the instantaneous amount of wind power available is, or even if the windmills are stagnant for a while. All that matters is that over a long period of time you end up with around the same amount of average power produced, and that's what wind power is good at. Potentially it would make sense to use wind turbines for powering specific sorts of activities that didn't rely on low latency of power production and could be automated (perhaps water desalination, for example). But using it as base power is a far more troublesome prospect than using solar.

> Sometimes there isn't enough wind power in an entire continent to provide power, and this could potentially last for days, even weeks.

Do you have a reference you could link me to?

Also, any reason to think this can't be solved by pumped-storage hydroelectricity?

http://en.wikipedia.org/wiki/Pumped-storage_hydroelectricity

The world uses ~2,000 GW and has ~100 GW of PSH capacity. Scaling up by a factor of 20 would be expensive because the low hanging fruit has been taken, but it really doesn't seem that bad if it means our base generators are half (or less) as expensive. See, for example, this take in Scientific American:

http://www.scientificamerican.com/article.cfm?id=here-is-the...

Hell, even storing it in batteries isn't that expensive:

http://www.bloomberg.com/news/2013-01-23/duke-completes-bigg...

($44 million for 36 MW, which is added to the typical costs of wind turbines of ~$2 million per MW.)

There is much, much more solar energy than wind energy for us to harvest.

This. The energy output of the Sun is legit ridonculous.

Something along the lines of 1 or 2 all the energy ever consumed by humans since forevers per second.

Is that the total output of the sun or just the energy from the Sun that reaches Earth?

I believe there is much more wind and solar energy each than humanity's power demands. It's not a question of total energy, it's a question of cost.

Reply on news.ycombinator.com