Is that this one https://spaceenergyinitiative.org.uk/space-based-solar-power/ ?

I haven't managed to get the geometry of what they propose into my head yet - it feels to me that as we're not at either of the poles, wherever you put the satellites there will be part of the planet in the way at some point of the 24-hours day - or are they thinking of a planet-wide constellation only some of which will be contributing at any point in time. In the latter case, I might suspect conflict with other countries planning similar systems (or maybe co-operation to feed into their system instead?).

They seem to be suggesting 2GW - which is hardly unlimited - UK demand is usually in the mid 30GW range (and we can already generate about 15GW from wind alone - although with more intermittency), but maybe it could be scaled up.

The microwave downlink sounds interesting - 2.9GW at 240W/m² seems to suggest a receiving aerial nearly 2.5 miles in diameter (if my maths is right)  - I'm not sure what's involved with such as device?

I wonder too about the installation, and given the seemingly short lifetime of things in orbit, the eventual de-commissioning and re-cycling of the materials involved.

There also quote 85% efficiency at the satellite end - I wonder what happens to all the waste (heat?) in the vacuum of space? If it's got to radiate out it's probably going to have to be reasonably hot - which maybe isn't good for PV panels or electronics.

Certainly and interesting idea - but probably need the reporters to come up with some more technical details to make a proper judgement though..

   - Andy.

I must say I see at as being harder than the website suggests.

Rectennas (an antenna with a rectification of the RF to DC ) have been demonstrated at moderate efficiencies but only at very power compared to what is proposed here there needs to be enough voltage to forward bias the diodes, but the diodes to rectify AC in GHz frequency range need to be small, low capacitance devices and because of this they are vulnerable to reverse voltage - those RF diodes typically used as mixers in satellite receivers (admittedly nearer 10GHz) have a capacitance of a few hundred fF (femto Farads =1E-15F) and reverse breakdowns of about 6V and forwards voltage drop of 0.4V, Typical max current per junction is a few tens of mA.

You need an awful lot of small diodes like that to make a gigawatt. (This design is typical, where a joint project with Glasgow and Southampton universities after much optimization  generates a few volts of DC into a 20k ohm load from RF at about 1GHz )

Beam tracking

From the satellite side to produce a spot on the earth a few miles across is going some in terms of narrow beam width- most satellite systems illuminate a good chunk of a continent at a time more or less ;-) RF is not opticss ,and I do not know what height the satellites are proposed, but the lowest orbit sensible would be many hundreds of miles, but those would whizz across the sky and need complex tracking to maintain a spot on target. Geostationary orbits are easier as the RF beam angles are fixed, , but at a height of 20 odd thousand miles the antenna array to make a beam that narrow is formidable. And that orbital track is high value and quite full already.

Then there is the generation of that much RF, this is the sort of power needed at CERN at peak pulse. There are RF sources on earth capable of (maybe tens) Megawatt levels not just pulsed, but they are rare and scary beasties, To add even one order of magnitude in power is brave, and to put kit into space where there is no cooling water and so on, would be a tremendous challenge.

And that is before we start to worry about the EMC implications of a gigawatt transmitter at some thousands of mile range - for not all the power will be generated on the right frequency (out of band sprogs, harmonics due to waveform quality, all the usual), and not all the energy at the intended frequency will be landing on the target either.

I Imagine that a much smaller, lower power system will  be demonstrated first, and it may be some decades  or even never that the system as described is actually operational.

Worth looking at certainly but a very hard spec to meet.

Mike.

My back-of-an-envelope calculation suggests that if you're collecting 3.4GW, and converting it at 85% efficiency, that's 510MW of heat that these satellites need to dump back into space.

That amount of waste heat generated in such processes is substantial ! What about harnessing it rather than dumping ? A bit of a challenge with the lack of air or water. Could thermoelectric generators be the solution? The recent coverage by SKY News suggests that this technology (Solar Farm) could be operational by the year 2030.

I'd not be surprised if it (transmission efficiency) ended up  worse than that - I'm not aware of a DC to RF method for the frequencies proposed that manages more than 70% when one includes the RF distribution losses. Even if the satellite is 500miles away, and the beam on earth is 5 miles in diameter, you are looking at something like a 3 degree beam - an antenna that has a beam-width of 3 degrees has a gain of ~35dB  (4000 or so ) in the main beam direction relative to isotropic,  A higher orbit gets proportionally narrower beam angle, and harder.

That degree of pattern control in turn implies a radiating aperture larger than a few hundreds of wavelengths across, filled with up to tens of thousands of in -phase radiating elements. This implies some dBs of  feeder losses, which at full power with central generation of the RF would be glowing, so  perhaps more easily coherent central low power Rf generation distribution with losses at proportionally low levels - after all who cares if we lose half the power in feeder loss, if it is a fraction of a watt at that point?

The low evel RF is then stepped up by an RF power amp to the 100-1000 watt level behind each antenna element or small group of elements, generating the RF  as near as practicable to the antenna sub-element feed-point. This also allows the gain and phase of the signal from each region of the transmit array to be individually adjusted to give control over beam focus and squint.

That is do-able, and spreading the  amplifiers out eases the cooling problems a bit, but creates one heck of a control problem. And all those PAs still need DC power and cooling to somewhere.

Mike.

I think that there was a similar system discussed on here a few years ago. One of the key points then was keeping the beam precisely aimed and what happened if it lost control. I am not sure how hazardous that level of microwave energy is, but i imagaine that the power density will be higher in the centre of the beam than around the edges.

Worth a try, But Newton law of action equal reaction will still apply.  How are you going to anchor the apparatus to stop it flying backwards?

How are you going to anchor the apparatus to stop it flying backwards?

Really!

The gradual change in orbital period and orbit height from the reaction from the radiation pressure will be the least of the worries, as the orbits will need to be tracked to keep the beam more or less focused on the ground station anyway. 

Realise we are talking about 3 newtons (300grams force) per gigawatt intercepted /radiated ( well approximating c= 1/3 *10^9 rather than the more common 3 *10^8 or if you prefer a more accurate answer consider it 3,33N per GW )

The effect on the earth's orbit and day length will be quite a lot less than that already occurring due to the moon pulling the sea about to form tides, and a small fraction of that energy being dissipated  eroding coasts and so on - yes the moon is coming in, but really slowly, not like 'London Calling' , no need to worry.

Cheers.

Mike

PS afterthoughts

Of course there are 3 lots of pressure to consider - the sunlight intercepted, the RF generated and if we radiate heat off the back of our satellite in an attempt at cooling the RF stages, it will to some extent undo this. I suspect that with some cunning we can get the various  forces to cancel to a degree over a time equal to the lowest common multiple of the earths orbit (~ 24Hrs) and the period of the satellite orbit.