Hi All

David's figures are utterly unrealistic. Logical error: Strawman argument.

his words [my comments]

Consider the boost cost alone. That would cost more than the sale of power from the satellite for its entire lifetime.

[1000 MW station over 30 years - 263,000 hours - selling power @ $50/MWhr will earn $13.15 billion [$13.15 G.] Keep the costs under that figure and it pays for itself. Factoring in interest etc. and it needs to cost <$10 G or so.]

It is an idea like the space elevator: Sounds good in principle, but totally unworkable in practice. Just forget it.

[unproven. Make your case first...]

Suppose that the satellite has a mass of 10 kg per square meter, which is rather light for an extended structure.

[The old NASA studies figured a mere 1.25 - 0.8 kg/m^2 was realistic. With flexible thin-films the PVs are more like 0.1 kg/m^2 and inflatable structures a mere percentage of that. Your figures are grossly inflated... ;-) ]

It costs well over $10,000 per kg to put something into geosynchronous orbit.

[granted. However, cost can be reduced by deploying sub-units of the array at LEO to power electric thrusters for the GEO transfer. Reduces the LEO mass to a mere 30% of chemical propulsion. If SpaceX can deliver to LEO for just $2,200/kg, and the electric system is another ~30% mass, then GEO delivery is just ~$3,000/kg. Even less because the bulk delivery drives launcher costs down with a guaranteed market. Costs approach ~ $1000/kg or less. ]

Suppose that the PV cells are 50% efficient over their entire lifetime. That gives about 650 watts per square meter. Now the antenna array must be pointed at the Earth while the PV array points at the Sun, so that they can not be mounted on the same structure.

[that's a real problem - but placing the SPS at the Earth/Sun L1 or L2 eliminates this hassle, and with higher microwave frequencies the array mass goes down. Energy cost to L1/2 isn't much higher than to GEO either. Don't give me the line that L1/2 are unstable because the station-keeping cost is very low with SEP thrusters even over 30 years.]

Suppose that the antenna array has the same area as the PV array

[bad assumption]

and has a mass of 5 kg per square meter.

[another strawman assumption]

Now consider conversion efficiencies and losses. For each watt produced by the PV array, only about 50% (if we are lucky) will go through the wires, through the conversion to microwaves, through the atmsophere to the receiver arrays, through the conversion back into electricity.

[you got that point right, but it is susceptible to amelioration by technological advancement.]

Even of the cost of the satellite were ZERO, which it will not be, it would STILL be absurdly expensive.

[cost it realistically and we will see.]

Add the cost of the satellite hardware: (PV array, antenna array, conversion equipment, housing for workers, supplies)

[bad assumption. Will need to self-deploy.]

, the cost of assembly in orbit, the cost of maintenance,

[on-orbit resupply via teleoperation should be fractional cost]

the cost of the receiving array,

[simple, identical units should be cheap per unit]

the cost of conversion equipment on the receiving end.

[same as DC-AC costs for ground solar, but no need for storage. Or direct DC can electrolyse hydrogen at +90% efficiency for later use.]

If it was expensive before, it is even more expensive now. Look how much the International Space Station (in low orbit, not synchronous) costs. Multiply it by 10,000 and you get the cost of a single orbiting PV station.

[mass wise, even with your figures that's just silly factoring. Your SPS weighs in at ~ 30,800 tons/GW and the ISS ~ 370 tons. Hence x83. Besides the ISS was done by NASA... says it all really]

Realistically (given the usual cost overruns), figure $1,000,000 per watt.

[more like ~ $2,500/W using the ISS (~$30 G cost) for comparison. Over 30 years a coal plant costs ~ $2.65/W - though factor in growth on coal prices and it's more like ~ $9/W. Hard to compete against but worth it in the long-run if costs can come down... and they CAN...]

Phooey! Why not just put hamsters into spinning cages? It would be MUCH cheaper.

[unproven. Your strawman SPS weighs ~ 20-200 times more than it has to. And costs ~ x3 to deploy. And ISS high-tech prices are even more inflated. Of course it's nuts... ;-) ... but it's not a real SPS by any measure.]

[However what is silly is the O'Neil style concept of Space Colonies selling SPS to finance themselves.]

Adam