Wednesday, September 10, 2014

Solar power from space?


When I published Sustainable Energy - without the hot air, there were quite a few topics I could have included but didn't because they seemed a little too blue-sky. For example I have draft chapters on Osmotic Power and on Kite Power (and if I were to rewrite the book today I think that I would now include Kite Power after all). Another idea I dismissed at the time was "what if we put solar panels in space, in geo-synchronous orbit?" I dismissed that idea on the grounds that "the advantage of space over the deserts of Libya and Nevada as a location for solar panels is only roughly a factor of 4, and surely that's outweighed by the difficulty and cost of getting panels (and associated power-re-transmission systems) into space, compared with just plopping them on the ground in a desert?" However, Keith Henson has for some time been working out the details of a scheme that might prove me wrong.
It involves many clever ideas, and some ambitious ones - such as the idea of powering a Skylon freight-delivery space-craft by space lasers that are powered from the ground with GW-scale microwave transmitters! I encourage people who are interested to read Keith's 'dollar a gallon' post and his follow-up post.
Further reading: Solar energy in the context of energy use, energy transportation, and energy storage - a paper in which I provide data for the power per unit area of real solar farms, and discuss the need for significantly cheaper energy storage if ground-based solar power is ever to contribute a significant fraction of energy consumption.

12 comments:

rks said...

Kirk Sorensen (the LFTR guy at Flibe Energy) worked, when younger, at a research job on solar energy from space. He got the job of doing the financial modelling. He could never get the sums to add up. Finally, as a joke, he set the cost of space deployment to 0. It still didn't add up. So he left. Unfortunately I can't work out which of the many videos this was in.

If you want to get the latest on all the energy miracles that might save us, then the easy way is to subscribe to Brian Wang's "Next Big Future": http://nextbigfuture.com/ .

Keith Henson said...

Robert, Kirk and I have gone around and around and he has never backed up his assertions about power satellites with numbers. Lost the papers or something.

I have had articles on nextbigfuture. The ones there are out of date but are not hard to find.

It's not hard to analyze what you can spend on power satellite and undercut fossil fuels. If you run the levelized cost of power calculations, the maximum capital cost is about 80,000 times the per kWh cost.

So, for example, for 3 cents per kWh, you can't spend more than $2400 per kW. Given cost of the rectenna, the cost of the parts, the mass per kW and the cost of lift cost to GEO, that's where the $200/kg comes from. It's not zero.

However, I don't knock LFTR. In fact, I think it's probably the next best thing to power satellites and if for some reason political or technical, we can have them then I would support LFTR.

Benjol said...

I somehow doubt that lasers (or high-powered microwave beams) in/from space is going to be any more politically acceptable than nuclear power (in space or elsewhere).

Keith Henson said...

Ben, microwave beams to get the power down are not dangerous. Physics limits the intensity of them to around 1/4 of sunlight.

Using conventional Skylon to LEO and microwaves to move cargo from LEO to GEO is not going to generate much opposition.

Propulsion lasers *can* be used as a weapon, a game changing weapon. But if the Chinese or the EU decided to solve their energy problems by hauling power satellite parts up with lasers, what could anyone do about it? Start a nuclear war?

One laser in GEO can only target about 1/3 of the world and the lasers can be put under a distributed control.

Anonymous said...

It seems to me that, although space-based solar avoids most of the intermittency of ground based PV, it still has much of the same problems of density - you have a large physical footprint for your power, and with that, long distances to transmit it over. Instead of one compact reactor near a city, you have a huge array, miles out in space. People grumble about the visual effect of wind farms and solar farms; as a keen stargazer, I think it would be hard to do worse than the unavoidable light pollution from hundreds of square kms of shiny stuff, always overhead.
John ONeill

Anonymous said...

So, orbital PV solves the problem of intermittency of solar power, however this problem can also be addressed by energy storage (which also has the advantage of addressing power demand fluctuations).

The orbital PV also produces more energy per panel but it seems unlikely that getting a panel to orbit will ever be cheaper than simply making 5 times more panels, even if we include the cost of storage that ground panels require.

So, the only point at which I can see orbital PV being economical if the cost of space the panels occupy on the ground becomes too high. As of this writing, cheap desert land remains in ample supply.

David MacKay FRS said...

Hang on, "anonymous" - you say "you can solve the problem by storage". Yes, but hang on - what is the cost of that storage going to be? Think about the whole system, cost it out, and then see if you can beat that cost with the solar-from-space invention - that is what Keith Henson is saying. If you want to see some possibly-helpful thoughts about the role of storage and current costs and target costs for storage, please take a look at the paper I cite at the end of my blog post, on "Solar power in the context of...".

Anonymous said...

Thank you for your reply. You are right, I didn't check the numbers, I just assumed storage would be cheaper.

Looking at Keith's article again, he assumes the cost of 900$/kW for the power sat, 200$ of which is for the rectenna, which leaves 700$/kW for the photovoltaic panels. He included the 50% rectenna losses which would not be present in a ground based installation, but on the other hand a ground based installation would produce about 5 times less energy, so the cost per kWh would still be 2.5 times greater: 700x2.5 = 1750, which is already above the target cost of 1600, not even taking into account the cost of storage. So, with these numbers, even if storage were free, orbital panels would still win on account of their higher energy production.

This assumes that the cost of panels is the same whether they are built for and assembled in orbit or built for and assembled on ground. Lacking any expertise in the field, this was the only way I could keep the comparison fair.

In practice I imagine the cost of ground panels might be higher since they need additional components to be protected from the weather, while the cost of installing them might be lower since it doesn't have to be done by astronouts/robots. On the other hand, orbital panels would need to be as light as possible to reduce delivery costs so that might increase manufacture costs. Whether these factors cancel each other out or whether one dominates is a question that I am not qualified to answer.

Of course, as long as orbital delivery does not reach 500$/kW, the target cost for a ground based system to be competitive with an orbital one will be higher than 1600$.

Just for comparison, if I am reading the graphs in your article correctly, a current photovoltaic installation (as much as 2011 can still be considered current) with pumped storage would cost between 30,000$ and 50,000$ per kW of steady power depending on how many days of storage you required.



Another problem to consider is that Keith assumes orbital delivery cost reduction partially through economies of scale - however, what could require such scale besides power sats? We are in a catch 22 situation where we need a reduced cost of orbital delivery to make the power sat industry viable, but we also need the power sat industry to provide the demand necessary to reduce the cost of orbital delivery.

It's not an insurmountable obstacle, for example a similar problem exists with regard to battery costs and electric vehicles and Tesla is trying to overcome it with its gigafactory.

energie vergelijken said...

als het veel problemen oplost en de wereld/natuur/klimaat er vooruit op gaat is het wel de moeite waard toch ?

hangar door said...

I think this is one of the most important think in humans history to take part in that we must to use space solar power

Anonymous said...

I know this is an old post, but I stumbled upon your blog when looking up information in the online version of your book. Is there any chance you could put up your draft chapters? Obviously if you think marking it as a draft isn't sufficient then mark in some way if calculations haven't been thoroughly checked or if you need more references ([citation needed]?). Since you've already embraced the creative commons approach to publishing (part of the reason I bought a physical copy of the book) is it such a stretch to share your unpolished work too?

Catprog said...

Using 300 W/kg ( https://en.wikipedia.org/wiki/Solar_panels_on_spacecraft ) and a space elevator($220/kg) I get a cost of $733/KW for space solar.

A falcon nine heavy is $5,333/Kw. (for LEO)
Assuming $3,000 for the 1KW solar,transmitter,receiver,grid etc itself.
80% efficiency for beaming.

$8,333/KW for space solar. ($10,416.25 / KW delivered)


$6,499 for a 24kWh Nissan leaf battery.
$12,000 for a 6kw solar system. (quick google search)

$19,000 for a 1KW for 24 hours system. (And you can store up to 24Kw of unused power.)
$10,416 for a 1KW LEO space solar. (any power not used is wasted, and we have time in the shadow of the earth and time when we are not in sight of the receiver. )
$20,416 for a 1KW GSO space solar. (and any power not used is wasted)