We’re turning water into rocket fuel. It’s easier than you’d think.

Yesterday we finished the last bit of modifications on the exterior of the MAV, when I very carefully installed the pony thruster systems on top of the holes left by removing the secondary maneuvering thruster systems. The last bit of undesired interior equipment, the tertiary communications system, is getting removed by Dragonfly and Fireball today. After that all that’s left is installing the three remaining pony flight couches, then the two RTGs- and that last item won’t happen until the day before launch. We need the heat from at least one RTG for sleeping at night, and the MAV makes the Whinnybago look like a blimp hangar by comparison. So with the major work done, we’re moving on to secondary tasks like extra fuel.

All you need to separate oxygen and hydrogen from water is electricity and two electrodes. The problem lies in collecting the hydrogen. One early design for the MAV actually had an electrolysis plant included as part of the fuel plant. The MAV would launch with a fuckload of water to split apart to get hydrogen, which it could then use to make rocket fuel. It would get the carbon for the fuel from carbon dioxide, which is mostly what Mars’s athmosphere is made of. The oxygen would be stored as oxidizer for the fuel, with excess oxygen being vented into the Martian air.

Unfortunately for me, a clever boffin found a way to apply the oxygenator tech to replace the old system of converting water and carbon dioxide into fuel and oxygen. Instead the fuel plant splits apart carbon dioxide and takes stored hydrogen to make methane. The CO2 provides enough oxygen for the fuel oxidizer and no more, which means no wasted mass being sent to Mars. Even with the super-insulated storage bottle for the molecular hydrogen and the inevitable leaks and losses, the new system ended up less than half the mass of the old, simply because about 91% of the reactant weight at launch had been eliminated.

The upshot of all this is, the MAV fuel plant can’t accept water directly. It has no electrolysis unit and no ability to separate oxygen from hydrogen. All I can do is swap out empty hydrogen tanks with full ones, feed it electricity, and watch it work.

Fucking mass-conservation obsessed NASA engineers. How could they not have known that some dumb schmuck would be stranded for years on Mars and suddenly need to make enough fuel to top off the fuel tanks in order to make a lunatic lunge at an escape-velocity rendezvous with a rescue ship? It’s such an obvious scenario that could happen to anybody, after all. No excuse.

But since the fuel plant can’t isolate hydrogen for us, we have to do it the complicated way. Fortunately, we have a small air tank compressor built into Rover 2. It’s what vents the airlock when it cycles, so that we don’t waste precious oxygen and nitrogen on Mars, ungrateful world that it is. It’s a simple trick to get that pump to suck up the entire contents of Rover 2’s interior into its tank instead of just the airlock.

But doing that gets us a bunch of mixed gases. That’s no good. And since we have no equipment that can un-mix gases, naturally we turn to magic for the solution. In fact, we’re using almost the same spell that we used to purge the methane from the cave farm. (Except, of course, we’re being much more careful, since we can’t afford to rip this airlock out from its frame.)

Here’s how it works. We isolate Rover 2 from the trailer and depressurize it. Then we use that tanked air to return the rover interior to about a quarter of an atmosphere. We need some air pressure, or else the water will just boil off before we can electrolyze it. Then we take a plastic box full of water (filled from the pony life support spigot), drop a couple of electrodes in it, and let it run. At the end we have an atmosphere that’s roughly ten percent hydrogen by weight, most of the rest of it being molecular oxygen. We can read that on the rover’s atmospheric analyzer.

If you’re thinking that this basically turns the inside of Rover 2 into a bomb, you’re right. And if you’re wondering where Starlight Glimmer and I are standing for all this, well, wonder no more- we’re at Ground Zero. Then Starlight casts her little spell, creating a force field that isolates the atmosphere sensors from the rest of the rover. The force field stops anything as heavy as molecular nitrogen, but molecular hydrogen just flits right through. A little bit of force field flexing later, and the hydrogen just flows into the bubble. The analyzer tells us when we’ve got it going.

Once we have the bubble of 100% hydrogen started, Starlight moves it from the analyzer to the airlock, and we wait until the analyzer shows the hydrogen content of the air in the rest of the rover as less than 1%. Then I activate the airlock air compressor and pump our harvested hydrogen into the tank.

Each box full of water contains about fifty liters, weighing fifty kilograms. That water contains about five and a half kilograms of hydrogen. Each five and a half kilos of hydrogen, fed into the fuel plant, makes about seventy-two kilos of rocket fuel. If we do this for twenty-five sols, once per sol, we get a bit more than 1,400 kilograms of fuel.

Why are we bothering with this, you ask? Hasn’t the MAV fuel plant already filled its tanks? Well, as it happens, the answer is no.

The MAV fuel tanks, first and second stage combined, have a bit of excess volume for both fuel and oxidizer to compensate for the hydrogen the MAV might or might not lose to leaks from leaving Earth until the stored supply is exhausted. The MAV is rated to achieve normal orbit, with normal cargo, on about eighteen tons of fuel, and the fuel plant is calculated to make over nineteen tons to provide a safety margin… but an absolutely full load for the MAV would be twenty-one tons of fuel plus adequate oxidizer to go with it.

Thanks to the ponies, who are being very understanding about us drinking their planet dry, we have an infinite supply of water to electrolyze. With the solar cells on the Whinnybago, we have enough power to supercharge the fuel plant systems to make it run much faster than normal. Normally the fuel plant only makes about forty to forty-five kilos of fuel per day. Extra power lets us double that. We could go a lot higher without risking a burnout or accident, but we don’t need to. One electrolysis session gives us enough hydrogen for seventy-two kilos, and we have enough time for that to add up to all the remaining space in the tanks.

That’s our only safety margin. With all the stuff we’ve ripped off the MAV, almost anything that breaks from now on means we’re fucked. But we’ll have ten percent more delta-V (not counting the magic boosters) than we otherwise would, which means if we lose a rocket engine or magic booster, we can make up the difference with a longer burn.

In one way it’s a shame we have all the water we’d ever need. I could just as easily electrolyze urine, which is mostly water. Then, for the rest of my life, I could brag about being so tough I pissed rocket fuel. But then, I think my guests are just as happy we don’t have to do that. It smells bad enough in the Whinnybago as it is without adding the smell of boiled pee.

Author's Notes:

In real life hydrogen tanks leak pretty much constantly... and the emptier the tank, the faster it leaks. Molecular hydrogen just doesn't want to be contained, and it will flit, molecule by molecule, through the solid metal walls of anything you put it in.

That's the single biggest critique of Andy Weir's take on harvesting fuel on Mars for the return trip. Yes, it makes lots of sense on paper to send up only a few hundred kilograms of hydrogen, rather than enough nice, stable, electrolyzable water that weighs enough to make the weight savings of the process trivial. But with six months in transit and fifteen months on the surface cooking up rocket fuel, it just doesn't make sense that all the hydrogen's going to be there.

Also, in the book the idea of adding fuel to the already full-up MAV is considered such a trivial issue that it simply isn't addressed. "The tanks are full!" "Add more!" "Okay." This is logic that, on the face of it, doesn't work outside of infinitely large hotels.

But consider this: let's say that hydrogen storage tech improves enough to make this fuel plant feasible. Thing is, it's never going to be 100%. And although you can predict an average amount of hydrogen losses, you can't really predict an exact amount. Conditions will vary. So a smart engineer, in this situation, would give himself a large fudge factor to compensate for both less hydrogen than expected (send a lot more hydrogen than the minimum launch requirement) and more hydrogen than expected (extra capacity in the fuel tanks).

In the book Mark had a finite water supply and could only make 780 kilograms of fuel. He started later and didn't electrolyze as often. Here he has more time, more water, and thus more extra fuel.

We'll see if it makes a damn come Launch Day.