BitsNotWatts

Fig. 1: space solar power satellite

Fig. 2: ground rectenna, converter, power line

Fig. 3: In 1974, GEO orbit was almost devoid of assets, and computers were big and slow and power hungry. Much has changed in four decades, but Space Based Solar Power is still wedded to the Glaser model, and still a speculation frozen in time, based on the assumption that we would grow ever bigger boosters evolving out of the Saturn V.

Instead, space is as expensive to reach as it was four decades ago, we have nothing as big as a Saturn V, and GEO is full of satellites whose mission would be compromised by high-power microwave emitters located in similar orbits.

Both ends of the comsat-earthdish link, as well as airport radars and radio astronomy dishes, use unfiltered Low Noise Amplifier front ends with fW/m² sensitivities that become nonlinear and create intermodulation products in the presence of high power (μW/m²) interferers. Filters to remove out-of-band noise on SSPS transmitters add mass and heat, and subtract efficiency. Rectennas are highly nonlinear, and will reflect or rebroadcast many harmonics, scattered all over the sky but concentrating in megawatt grating lobe beams.

Perhaps it is time to revisit our assumptions, and think about what pioneering really means.

Fig. 4: When American farmers settled west of the Allegheny mountains, 600 kg of rye required three pack animals to transport to eastern markets, and sold for $6. The same rye could be distilled into 30 liters of whiskey worth $16. But the pioneers did not ship whiskey east, either.

Fig 5: The pioneers traded a quart of whiskey for an Indian beaver fur. 30 liters of whiskey became a stack of beaver furs worth $240 in New York. Successful pioneers aren't stupid, and don't do things the hard way. They move valuables, not commodities.

Fig. 6: This is the Google data center in The Dalles, Oregon. It consumes megawatts of electric power. All that electricity ends up in these coolers on the roof, after passing through the machines inside.

Fig. 7: A pioneering company like Google is very profitable because they make a LOT of money with that electricity. Google sells 20 thousand dollars worth of services for every megawatt hour they consume.

Fig. 8: Let's rescale the last graph so that we can see 20,000 on it. The markup ranges from 70 to 17 hundred, enough to attract most smart pioneers.







In the rich developed world, we have a vast and expensive fiber infrastructure. When Netflix and Youtube aren't saturating the backbone, Google can reach almost a billion fiber connected customers, and billions more with DSL and modems.

However, like the pioneers, Google struggles to deliver their product to their worldwide customers.

Fig. 9: After the data leaves the data center, it travels over land and under water on fiber optic cables. Every 50 kilometers, the signal is cleaned up with repeaters, consuming more high reliability electrical power.

Finally, the internet reaches the end user. In the developed world, there are 750 million broadband connections to endpoints with reliable electricity.

Fig. 10: But what about the rest of the world, especially the three billion without grid electricity or wired telephones and cable? We can reach them with 70 GHz microwaves - at least we can if we use space energy and spectrum for high value computation and communication, not low value grid power.

3 petabytes of information, at 1 electron volt per bit, is 4 millijoules of energy. Even if we move that with 1e-9 efficiency, that is one kilowatt hour of energy, worth less than 10 cents. A downloadable blueray movie is perhaps 45GB for $15 from Amazon in 2013, and 3PB is 67,000 such movies, a million dollars worth of information, sent for a dime.

We can do even better if we start small and enable lower-bandwidth, life-changing e-services to remote and isolated areas, such as remittences, remote medicine, remote employment, etc. These will be far more valuable to poor people than movies. Server sky can be profitable with 25 arrays, 1 metric tonne, in orbit, and grow exponentially from there.

Systems that convert space solar power into computation and transmit the results to earth, bypassing energy and capital intensive fiber networks, may prove cheaper to operate and to deploy, especially in regions without established high bandwidth infrastructure.

Fig. 11: 183 GHz microwaves are completely attenuated in the moist troposphere. The stratosphere is very dry, fractions of a ppm of moisture rather than many percent, so a rectenna in the stratosphere will efficently collect a 183 GHz microwave beam from space, while less than 10 parts per billion of the sidelobe power will reach the ground.

The radius of the orbiting transmitter multiplied by the radius of the rectenna is approximately equal to the distance times the wavelength, and the area is the square of the radius; changing from 2.45 GHz to 183 GHz reduces both antenna sizes by an area factor of 75.

Note that the air density is 7% of sea level, so the path losses are 0.3% at latitude 60° with an elevation above the horizon of 22°, much less than 2.6% for < 6 GHz microwaves to the ground. The platforms can be directly over cities, with the power cables going straight down.

At 20 kilometers altitude, a one kilometer platform casts a 650 meter dark shadow with a 1400 meter penumbra, like a high dark cloud, with the shadow moving like a sundial across the earth. Also like an isolated cloud, atmospheric scattering will still cast plenty of light into the dark shadow.

Rectenna power density will be cooling limited; rectennas may need to be larger and lower power density to dissipate waste heat by thermal radiation. The ambient air temperature is -18C, though at 0.07kg/m³, it is too diffuse to move much power by convection. Black body radiation: assuming IR-black, optically-white surfaces with an IR emissivity of 0.8, then a rectenna running at 125°C (402K) will dissipate 1185 W/m² into deep space on the top, and 955 W/m² towards the 270K stratosphere/earth below - 2140 W/m² total. If the rectenna is 70% efficient, it can sent 5 kW/m² to the ground.

Fig. 12: This suggests an intriguing possibility - if there are 6 rectenna panels arranged in an annulus, as shown, we can sweep 15 kW/m² peak illumination of the panels two or three at a time, turning 60 times per second, directly generating the plus and minus power voltages for three phase AC line power.

If the squared power sum is always constant, and we use opposite panels to generate plus and minus voltages, we can transmit constant power from the SSPS, circling and shaping the beam with SSPS emitter phasing, and obviate the need for power conversion and grid phasing electronics after the rectenna. This can help manage regional grid phasing.