I produce CO2 constantly. It’s one of my powers. Possibly we could recruit astonauts with similar abilities.
Me neither 
But I’m thinking that the simplicity of the gas light chemistry, coupled with our ability to test it extensively on earth outweighs the mining of water from soil which we can only superficially test here. Especially if we’re proposing using this system to not only build out water and oxygen for crew use but for creating the return fuel. It would be a complete failure to land, and discover that the potential water to be harvested was off by 50% due to reflection or scattering or sublimation.
Excellent! How long do you need to make 100 tons of it?
The atmosphere of Mars is mainly CO2. No need for astronauts with PMF - pronounced male flatulence - syndrome.
While there may actually be ice just under the Martian soil, I think I’d want to actually see proof it is there and can be made into fuel before I’d head out to Mars without enough return fuel.
I can deliver it in 294 years, give or take. You wanted metric tons, right? Cause if it’s English units I can turn it over in less than 270 years.
The answer no doubt has a lot to do with the fact that this is quite challenging and would have added substantial cost to any space mission yet undertaken. That it looks to be necessary for most of the Mars voyage schemes so far proposed suggests it may eventually be developed.
Some people are awfully picky.
That’s a bit misleading. The ISS is refueled from Progress spacecraft. Also, a DARPA project called Orbital Express demonstrated automated docking and automated fuel transfer between two unmanned satellites.
Actually, liquid hydrogen isn’t necessarily that great of a fuel. Its low molecular weight gives it a high exhaust velocity for a given pressure, but the low specific gravity (density) of hydrogen requires more tankage (and thus, give a poorer mass fraction, or ratio of dead weight to overall weight) and requires a larger throat to get the same mass flow. It’s also kind of dangerous to handle, owing to its chemical reactivity to many steel and aluminum alloys, easy combustability, and the invisibility of the resulting flame. LH[sub]2[/sub]/LOX is a suitable fuel for first stage disposable boosters (or, in the case of the STS “Shuttle” the External Tank) where the weight prohibition isn’t as significant, but it is mostly used because it is cheaply manufactured and nearly pollution free. Only a handful of upper stages have used LH[sub]2[/sub]/LOX as the propellants (the Centaur Upper Stage, Delta 8000 & 9000 Series Stage 2, Saturn S-IVB), and of those, only the Delta is a “modern” design.
For long term use in space (stationkeeping, orbital maneuvering, space probes) some form of hydrozine (H[sub]2[/sub]H[sub]4[/sub], UDMH, MMH) is used for chemical propulsion systems, either as a monopropellant or with a storable oxidizer like RFNA. RP-1 (highly refined kerosene) and other petroleum distillates have a tendency to break down and leave a gummy residue with temperature cycling, and cryogenic propellants have obvious problems. Liquid methane is a possibility for higher specific impulse (with RFNA or LOX as the oxidizer) but currently there are no commercial rockets using methane as the fuel.
As for the OP, there is no absolute reason that such a system can’t work; however, you have to account for the pretty significant difficulty of handling liquid propellants in free orbit. Because they don’t settle due to gravity, you either have to have some kind or pressurization system or ullage thrust to force the propellants out of your storage tank and into the vehicle tankage. Then you need to assure that the transfer system works, can be done by astronauts in bulky pressure suits or is completely automatic, and has a negligible chance of fouling up the propellant system or vehicle in the process. Instead of trying to transfer fuel from one vessel to another, it’s generally considered easier to simply launch a “departure stage” motor with fuel already incorporated, which just requires in-orbit docking and connection. This was the plan for the Constellation system–having the Command Service Module link up with an Earth Departure Stage to take the vessel into a Lunar intercept trajectory–and also for the early Soyuz A-B-V proposal for the abortive Soviet Moonshot program.
The advantage of this is being able to increase the fuel available for a given mission profile without having to develop yet a larger heavy lift rocket. However, not only does it add increased complexity (multiple launches, docking in orbit), it doesn’t address the fundamental problems of a manned Mars mission; that is, it won’t significantly reduce the transit time (about 9 months each way, and a 13 month layover on Mars for a Hohmann trajectory), and so the radiation and long duration freefall effects are still highly problematic.
First of all, there is a big difference in orbital energy between the low orbit of the ISS and geostationary and Molniya orbits that communications satellites reside in. This would require regular fuel supply, which is nearly as costly (from a launch operations standpoint) as launching a new satellite. Second, the production cost of communications satellites has being going down steadily as communications technology becomes more sophisticated and robust, whereas there is a limited amount of maintenance or repair that an astronaut can perform. Servicing satellites–which are often rotating slightly for stability–is a hazardous activity. And finally, comsats, because they are regularly exposed to radiation, will have a limited lifespan regardless of maintenance. While not infeasible, servicing missions just don’t make a lot of fiscal sense (as NASA discovered with the Shuttle).
However, the concept of an orbital space tug is not novel; Grumman and NASA studied this concept extensively during the Apollo Applications Program proposals using the Lunar Module as a basis for a modular space tug vehicle. (With its docking capability, restartable engine, and sophisticated avionics control it was a natural vehicle for this type of work.) The Soviets also played with this concept with in their ALMAZ military space station program using the TKS spacecraft, and the USAF with their (conceptualized) Blue Gemini and Manned Orbiting Laboratory. Ultimately, there was and currently is no real demand for such a vehicle.
Stranger
Your MTBF is too small.