Viability of orbital refueling

Question

One of more successful tactics in Kerbal Space Program in order to get big missions with a lot of delta-V going is to launch the big interplanetary craft (with landers, rovers, service modules and everything needed on the mission) on top of a relatively weak disposable first stage(s). The craft reaches the orbit almost completely depleting its fuel supply, and then consecutive launches deliver fuel, refueling the craft in orbit.

Of course real life is not KSP, but on top of prohibitive cost of such a mission in general (even if launched "traditionally" it would be vastly more expensive) what are the factors that would make this approach impractical?

(which I assume it is, since smarter minds than us decided not to go with it.)

Answer 1

This is a short answer but in KSP pumping fuel is totally costless, whereas in reality pumping fuel between stages is difficult even in the case where the stages are mated together on the ground. This is especially true when the propellants are cryogenic and therefore more akin to a thick slush than easily-pumped liquid.

On the other hand, I have found a NASA study which seems to validate the possibility of (some degree of) orbital re-fueling:

"Cryogenic Propellant Storage and Transfer Technology Demonstration for Long Duration In-Space Missions"

Answer 2

This is basically the approach being considered for getting all of the hardware to Mars for a crewed mission. However instead of refueling an upper stage that was used to get you to orbit, it is considered much easier with almost the same benefit to launch an entire fueled stage with engines instead of just the tanks with propellant. The only penalty is that you have the mass and cost of some engines attached to the bottom of the tanks that you were going to transfer propellant from, which is a relatively small fraction of the mass. The benefit is that a) you don't have to transfer propellant between tanks -- you just dock with the stage, and b) you can now design the stage for its interplanetary mission instead of having to modify an upper stage for two different missions (that and getting to orbit). For example you may want to keep cryogens from boiling off in a long cruise, which is never a concern for a launch vehicle.

In any case, there is research in both preventing cryogen boil off and in the transfer of propellants in space.

The International Space Station is regularly refueled in orbit. Though it's not so that it can depart Earth orbit.

Answer 3

If you use a VASIMR you wouldn't have to refuel because the amount you could possibly carry(weight to thrust) is much much much greater than the amount you'd burn on the launch and trip. Using one you could make the trip to Mars in 39 Days instead of six months or more with a conventional rocket.

Sometime this year NASA will install a VASIMR on the ISS for testing. The ISS boosts are currently provided by spacecraft with conventional thrusters, which consume about 7.5 tonnes of propellant per year. A VASIMR would cut this amount down to 0.3 tonnes.