Space Based Solar Power II

Ground-based and space-based solar power, in spite of certain misconceptions, shouldn’t need to compete because they’re optimal for two different energy markets. Consider ground-based solar – its peak output is during the hottest part of the day, when the Sun is bright and high in the sky. But it’s at its worst when low in the sky (and at night, of course) and during the colder seasons. Space-based is good all-day round, all-year round… BUT it’s hard to up the output. That’s especially needed when the Sun is high in the sky and the ground-based solar is working hardest.

Couldn’t ground-based cover both by more collectors being on the ground to make up for low-light and night conditions? Let’s look at that proposition in some detail. How is power stored? If we assume batteries then we’re faced with some difficulties – firstly a top lead-acid battery, with a supercapacitor to help, is only able to store 75 W.hr/kg. Say a household needs 3,000 W averaged over its whole day – and we’re assuming the car is being charged by the power system too. That’s 40 kg of battery per hour of supply, some 960 kg of battery for the 24 hours in the day. And batteries have a discharge efficiency of 80-85% – meaning we need a total of 1,200 kg of batteries to store the day’s energy needs.

But how much day have we got to collect the energy from the Sun in? At best it’s about 0.25 of the day – thus 3,000 W continuous supply needs to be fed by 15,000 W of solar collectors (remember that battery inefficiency.) Thus at $0.5/W our “Total Power” solar-power system costs ~$7500 of solar array plus all those batteries (~100 total, costing ~$15,000, plus the power handling system ~$5,000.) Thus our system costs $27,500 on the optimistic side of costing, since there aren’t any $0.5/W solar collectors on sale. The system would recover costs in ~5.5 years at ~$0.2/kW.hr from the energy retailers. Assuming its output didn’t decline over time that is…

But what about space based solar? Firstly it doesn’t need storage – it can supply all the time. Of course not everything is being used “all the time” but with a decent amount of homes with some level of power storage and the excess could charge those, drawing back when it’s peak time. Honest retailers might even reduce the “consumers'” bills in exchange for the storage service. So no storage. No battery costs. However the collectors need to be ~1/6th of what’s needed for the same continuous output on Earth. And concentrators, rather than flat panels, can be used all the time. Expensive, ultra-efficient cells can get 500-1000 times the sunlight of their terrestrial counterparts and convert at 40-65% efficiency…

Towards nanowire solar cells with a 65-percent efficiency

Published on: 16 June, 2010

TU/e researchers want to develop solar cells with an efficiency of over 65 percent by means of nanotechnology. In Southern Europe and North Africa these new solar cells can generate a substantial portion of the European demand for electricity. The Dutch government reserves EUR 1.2 million for the research.

…thus immense savings in materials and array mass. My rough BoTE computations suggest a 1 GW SPS can mass 1,400-720 tons, with huge savings in launch costs. Assuming we’re launching via Falcon 9 Heavies that’s 72-36 launches. Elon Musk would love us!

So instead of spending ~$7,500 on 15 kW of array per household, only ~$1,250 is needed. And no batteries. Thus, potentially, a BIG saving overall. But some extra panels for peak power usage – in the bright, hot part of day – and there’s a neat synergy between the two power supply sources. Of course that’s all evened out over ~1 GW/3000 W households (~333,000) as individual supply isn’t so easy to do with a SPS.

However initial demonstration SPS units might only provide a few MW, enough for small communities. In theory there’s no reason why a bunch of smaller sub-units can’t eventually be ganged together in a common structure and modern phase arrays used for sending out power-beams to several different rectennas on the ground. If 720 tons is 1 GW of supply, then a single launch demo SPS massing ~20 tons at GEO might supply ~30 MW to the ground. That’s enough for a 10,000 household township in a remote area. Worth considering.