That's all wrong. You only need to get the materials up to construct the batteries. After that the batteries never leave space and the only other thing you need to send up now and again is fuel.
The Batteries will dock with an orbital transmitter which will beam the energy down to the surface of the Earth via Microwaves and then leave back to the powerplants.
Then skip the "battery" step and plug the output of your powerplant directly into a microwave relay. Otherwise, the cost of sending up fuel "now and again" remains around $5,000/kg, which leaves you with several thousand dollars per kilowatt hour.
Of course, the trouble with this is microwave power stations--the most efficient designs I've seen, anyway, none of which have ever been used in practice--have efficiencies of around 25%. If you combine this with a powerplant with 60% efficiency, this means 15% of the energy produced is being shipped back to Earth. The power station is therefore NOT competitive with less expensive and more efficient Earth-based powerplants.
The only advantage to building in space is the extra radiation from the sun without the atmosphere to filter it. Unless the customers receiving that energy are already IN space, that advantage cannot be tapped, since you still loose most of that extra energy in shipping it through the atmosphere and even then that energy cannot be delivered at competitive costs.
The only thing that will cost anything in regards to the batteries is the fuel to get it to and from Earth orbit.
And how do you propose you get the fuel INTO orbit? By osmosis?
If you want to send fuel into orbit, you need a rocket to carry it there. Even cheap Russian launchers will still cost around $5,000/kg of payload. If the orbital transfer vehicles consume 10kg of fuel every time they move a battery, then that's $50,000 a trip. That would end up being MORE expensive than simply dropping the batteries out of the atmosphere since not only do you have to send a launch vehicle into orbit, you also have to send 10kg of extra fuel you wouldn't otherwise need.
But I keep saying it and I'll continue to say it till I'm blue in the fingers. These powerplants could be used for space based colonies.
And nothing else, although even in this case the need to keep shipping fuel from Earth makes it an extremely dubious proposition.
But even then a photovoltaic powerplant would end up being more efficient anyway, since that system would be scalable and increasing power demands could be met by either 1) installing more of them or 2) replacing old ones with more advanced ones as they become available.