So, eight hours of breathable air. Was this still doable?

Artemis was three kilometers away. The trip had a lot of rough terrain but it was also downhill. Call it two hours.

My original plan had been to wait until night (clock-night I mean, not actual lunar night) and then sneak in when everyone was asleep. But I didn’t have enough air to wait that long. I’d have to enter in the middle of the day.

New plan: the ISRO airlock. It led into Space Agency Row in Armstrong Bubble. There’d be a few confused nerds and someone might say “um…” but I’d just keep walking. With the sun visor down, no one would see my face. And, unlike the Conrad airlock, it wouldn’t be littered with EVA masters.

Okay, problem sorta solved. That meant I had six hours before I had to leave the collection area. Ninety minutes per harvester. Time to hustle.

I got as comfortable as I could in my hammock and assembled the welding gear. I laid the acetylene and oxygen tanks between my legs to keep them stable. On the harvester’s undercarriage, I eyeballed ten centimeters from the coolant valve and scratched a three-centimeter circle there with a screwdriver. That’s where I had to cut.

I flipped down my helmet’s sun visor. I’d duct taped a welding lens shade to the middle. I cranked the acetylene valve, set the torch mixture to ignition mode, sparked it, and—

…it didn’t start.

Um.

I tried again. Nothing. Not even sparks.

I checked the acetylene tank. No flow problems. What the hell?

I flipped up the visor and inspected the sparker. Dad taught me to use a flint sparker because an electric one is “another thing to break.” It was just a piece of flint and steel grooves attached to a springy handle. Nothing complicated about it. This was thousand-year-old technology we’re talking about here. Why wasn’t it working?

Oh.

Right.

When flint strikes steel, it knocks microscopic flecks of metal into the air. The metal burns because of some complicated crap related to surface area and oxidization rates. Basically, it rusts so fast that the reaction heat makes fire.

Fun fact: Oxidizing requires oxygen. Flint and steel won’t work in a vacuum. All right. No need to panic. A welding flame is just acetylene and oxygen on fire. I adjusted the valves and set the mixture to be a trickle of acetylene amongst a torrent of oxygen. Then I scraped the sparker right in front of the nozzle.

Sparks! Boy did they ever fly! That oxygen made the metal flecks go apeshit. But I’d got too far. There wasn’t enough acetylene to ignite the flame itself. I added a bit more to the mix and tried again.

This time, the shower of sparks managed to light a sputtering, inconsistent flame. I spun the valves back to a normal mix and the flame settled into a familiar, stable shape.

I breathed a sigh of relief and flipped my visor down. I held the torch steady despite the clunky EVA suit. Pain in the ass. But at least I didn’t have to deal with molten metal. This was a cut, not a join. When you cut, you aren’t melting metal. You actually turn it into an oxidized gas. Yeah, it’s that hot.

The actual cutting was a lot easier than I expected. It took less than a minute. The little three-centimeter circle of steel plopped down on my chest, followed by a blob of molten wax. The wax bubbled and re-hardened almost instantly.

My positioning was perfect. I’d cut into the wax reservoir without nicking the coolant lines nearby. I didn’t care about the health of the coolant system, but I didn’t want the harvester to call home about a coolant leak. The small daub of wax that fell on me wouldn’t be enough loss to worry the harvester. At least, I hoped not.

I pulled a pressure valve from my duffel. I’d bought six of them from Tranquility Bay Hardware the day before (one per harvester and two spares). Standard pressure connector on one side, three centimeters raw pipe on the other. I jammed the connector into the hole. I’d done well on my cut—it was a snug fit. I fired up the torch again (with the same oxygen-crazy ignition mix as last time) and grabbed a rod of stock aluminum. I needed a strong, airtight seal around the valve.

I’d done a million valve installations with Dad as a kid. But never in an EVA suit. And unlike the cut, this time I was melting stock metal to make a seal.

If I screwed up, a blob of molten metal would fall on me and bore a hole straight through my suit. Holes in EVA suits are bad.

I got as far to the side as I could—if I screwed up, maybe the Aluminum Droplet of Doom would miss me. I got to work and watched the aluminum puddle grow. The droplet trembled along the weld site, then finally seeped upward into the crack above it. My heartbeat returned to somewhere near normal. Thank God for surface tension and capillary action.

I was careful, and took my time. I worked around the valve slowly, trying to keep my body from being directly underneath. Finally, I finished the deed.

I’d installed a pressure valve into the wax reservoir. Now it was time for the dastardly part of my plan.

I attached the line from my welding oxygen tank to the valve and cranked the flow to full.

Sure, the reservoir was full of wax, but there were gaps. And believe me, when you blow fifty atmospheres of air into a pressure vessel, it finds the gaps. Once the tank equalized with the compartment, I very carefully closed the valve and disconnected the tank line.

I slid out from under the harvester. I watched it for a second to make sure the damn thing wasn’t about to move. I don’t like making the same mistake twice.