Tuesday, July 3, 2012

Solar power: come rain or come night

With the Rio+20 eco/enviro summit newly ended, this seems like an appropriate time to post about solar power.

A solar garden
Some would have it that solar energy can meet a large fraction of the world's energy requirements. The International Energy Agency, in fact, would have it that "Solar power could produce 25% of global electricity by 2050." (Not to be outdone, Greenpeace claims "Wind Power Can Produce One Third of World's Electricity by 2050." Wind power is, of course, merely another way to leverage solar energy.)

Of course the wind doesn't always blow, or the sun shine. The more we come to depend on intermittent power sources like these, the more we will also need to store power for later use (if part of the time we can generate a surplus). Logical sites for solar farms (deserts) and wind farms (on open plains and high hills; off the coasts) may be remote from where power is needed -- say, Minneapolis in the winter. Read the preceding as: lots of new infrastructure (with attendant costs) for power generation, power storage, and long-range power distribution. A more subtle point is that a power grid reliant upon many intermittent supplies will also need improvements to maintain stability while generators and storage sites ramp up and down, come on and offline.

And land ... lots of land. Energy sprawl, some call it. Wind and sunlight are diffuse sources of energy.

Upon due reflection ...
The sunlight that reaches the top of Earth's atmosphere isn't all available for exploitation on the ground. Some sunlight is scattered in the atmosphere: that's why the sky appears blue. Sunlight that's not scattered may still not reach the ground: clouds can reflect back into space upwards of 80% of incident light. You'll want to put your solar farm somewhere without clouds. Daylight will limit your power generation, on average, to twelve hours out of twenty-four. Adverse weather will further reduce solar-power production. 

Beam it down, Scotty
How about siting the solar farm ... above the atmosphere? Put your solar-power generator in a geosynchronous orbit and you have unscattered solar input 24/7 (with a ~75-minute exception twice a year, at the equinoxes). Beam power to the ground as microwaves (to which the atmosphere is largely transparent) then rectify the microwaves at the receiving station to recover electrical power. And beam the microwaves to the receiving station convenient to where power is most useful: say Minneapolis in the winter and Atlanta in the summer.

To summarize: more energy input than ground-based solar, virtually all the time (the exceptions known in advance with astronomical precision), with the resulting power delivered close to where it's wanted.

So what's the catch? 

Upfront costs. Putting stuff into space is pricey. The coming age of commercialized near-Earth space operations (Go, SpaceX! Go, Dragon!) will surely reduce launch costs somewhat, but lofting entire power plants will never be cheap. The answer -- and a premise of my upcoming novel Energized -- is capturing an asteroid into Earth orbit, and building Up There with those materials.

With Planetary Resources newly announced (and backed by billionaires), maybe mega-resources will be available in Earth orbit sooner rather than later.


Unknown said...

Mr Lerner,

Another consideration rarely mentioned in the push for large scale renewable is the impact on the grid. Our grid is divided into four major regions: Western, Eastern, ERCOTT and Quebec. In each of these regions the grid is subdivided into smaller interconnects (FRCC, MRO, NPCC, RFC, SERC, SPP, TRE, and WECC). The concept when the grid was unified and regulated was simple: unify the continent’s electric grid so everyone is on the same frequency and all generators are synched but divide it so each regional interconnect can operate independently of other regional interconnects should the need arise. Transmission between interconnects was capable but only done in emergencies. The interconnects act as firewalls so when a disruption occurs at one plant or substation the local grid operator can isolate that damaged section and if that cannot be done the outage will be contained only to the regional interconnect. Much of this had a cold war rationale: isolate damage to critical infrastructure from military attacks. Deregulation allowed the routine transmission of power from one interconnect to another and allowed areas like the North East to close many power plants and rely on generators outside their interconnect to supply them with electricity. I’m kind of on the fence on deregulation: on the one hand the grid operates more economically but at a loss to reliability. There’s an optimal balance to be sure, but I couldn’t tell you where that point is.

As renewables are typically concentrated by geography new transmission lines and increased transmission across the interconnects will be needed if wind farms in Wyoming or solar farms in Arizona are going to power street lights and HVAC loads in Atlanta (as the renewable advocates are pushing for). Aside from the cost to build new and upgrade existing transmission infrastructure (which will be astronomical), the reliability of our gird will be severely compromised … there’s no way around it. A severe regional weather event, human error, or software/hardware failure could compromise the entire continents grid. That’s too big a risk.

The possibility of several smaller grids failing simultaneously is practically nil, however the possibility of one single integrated grid failing is a metaphysical certainty. When it comes to our grid smaller and more local is the way to go.

Edward M. Lerner said...

Mr. Hanson,

I quite agree ... optimizing for efficiency often (almost invariably) means sub-optimizing with respect to other important attributes -- like reliability. Efficiency means, at the least, "reduced safety margin."

And that's scary.

- Ed

Enviko said...

Solar energy is very much a debated topic. Not in regards to using it, but instead what is the most efficient way. The use of outer space is a very interesting concept and although I do not know a substantial amount about how it will work, the idea seems possible. However, like you stated, this won't happen unless it gets large financial backing.