June 12, 2002
Saturn Screwup

My dad used to work on the Apollo space program. He has any number of stories, all of them funny. Some of them may actually be true. Here's one of them:

Being responsible for the mobile launch platform (not the crawler, as previously reported... the MLP was the thing that sat on top of the crawler and actually held the rocket), my dad was also in charge of some of the various fuel systems that supplied the Saturn V, and the Apollo space capsule it carried, with power. One of these systems was the fuel cells.

Fuel cells create electricity by forcing hydrogen and oxygen past a platinum-coated membrane. The various electrochemical reactions this creates produce electricity. The problem was impurities. If you got impurities, especially in the hydrogen, it would cause tears in the membrane, potentially ruining the fuel cell. And, hydrogen being what it is, it was actually pretty darned hard to keep impurities out for any length of time.

The nice thing about hydrogen is that since it's by definition the lightest thing in the universe, any impurities will float at the top. In theory, all one needs to do is purge a little of the hydrogen out of the tank and the impurities will go out with it. Vent it out and all is well. Put an ignition flame at the end of a vent pipe, and you'll be sure not to end up with any really dangerous buildups.

That was the theory. In practice purging was a real pain. The valves used for venting had poorly designed seats, and the seals in them would blow completely out whenever you tried opening one. This meant you had to completely drain the tank down, pull the valve assembly apart, rebuild it, replace it, and carefully fill the thing back up again. It took hours, and it happened all the time.

(Why didn't they just redesign the valve? Well, I'm not sure. You have to remember everyone was in a titanic rush, and probably nobody had the time or inclination to fight through the government red tape to even convince NASA there was a problem. Note this is not to convince them of a fix, but to convince them there was a problem that needed fixing. You just couldn't afford the wait.)

So, the problem is that the hydrogen gas, being held in at no less than 6,000 psi (413685 millibars) mind you, goes out with such force that it blows out the seals in the valve. However, there were eight (I think) valves scattered around the tank. It stood to reason that perhaps if you let all the valves open at the exact same time, perhaps the pressure at each valve would be low enough not to blow the seals. So a plan was hatched.

A man was stationed at each valve, waiting for the signal to purge. At that signal, each man was to turn his valve as fast as he could. They were even told to stand a certain way to get maximum torque and synchronicity as they turned the valve. The signal was given, and true to form everyone did exactly what they were supposed to.

It was at this point that physics and engineering tolerances began to cause -- problems -- with the plan. The ignition source at the end of this particular vent dump happened to be a series of ignition flames arranged at the edges of the box-like vent structure. They were designed to function with the assumption that only a certain volume of gas would be flowing past them at any particular time.

Also, hydrogen all by itself isn't actually flammable. Combustion by definition needs oxygen to occur. A flame, properly introduced, will actually go out in the presence of completely pure hydrogen. Fortunately for NASA hydrogen quite readily mixes with oxygen, but, as with all good dance partners, they must be properly introduced.

Unfortunately the volume of gas actually flowing through the vent stack was probably eight times greater than the tolerances it was designed to work with. As near as anyone could tell, once that much hydrogen hit the vent stack at 6,000 or so psi, most of the flames just went out. The ones that didn't go out actually got bent out of true by the force of the gas, making them even less likely to properly ignite the stuff. Worst of all, the hydrogen went out with such force that it completely displaced the oxygen all around it, and started gushing into the sky like some titanic ghostly geyser.

Now, if the hydrogen had had the simple decency to keep exiting at 6000 psi, it probably would've just dissipated far beyond the vent stack's ignition points. Unfortunately that didn't happen. Like a balloon running out of air the pressure began to fall a bit, and that was all that was needed.

The gas escaped with such force that it was quite visible next to the Saturn V on the pad. A giant turbulent column of gas blasting into the sky is actually kind of hard to miss. Because of the lighting, you could see the turbulence on the edges of the column. That was the air (30% oxygen) that had got carried by the hydrogen high into the sky, gently making its way back down to earth, like leaves falling through a dust devil. This was what my dad termed the "ohshit" moment.

Because you see not all of the ignition points had gone out. Like tiny piranhas they waited, flaming mouths open, for the right conditions to set off a frenzy.

Almost in slow motion my dad picked out a particular curly-que of air as it tumbled down the column. Once it reached the stack, by now quite thoroughly mixed with the hydrogen around it, the entire thing ignited with a roar. However, because the center of the stack was still venting at a huge rate, a whirlpool-like sheet of flame started up the column like a brilliant cellophane wrapper around a piece of candy.

Even this would've just been particularly bright fireworks, if not for the top of the column, which was just sort of floating around about 500 feet in the air merrily mixing with the oxygenated air around it. Once the flame raced up to that point, it all went off at once with an almighty bang and a display that could be seen all over the cape.

Pretty much every alarm over the entire cape went off at once. Capcom started yelling on the air trying to figure out what the hell had happened, and every fire truck in the entire world rushed to the scene. Fortunately hydrogen isn't all that dense a combustible, so there really wasn't any damage done to anything.

Dad got his redesigned valve shortly afterward.

Posted by scott at June 12, 2002 10:54 PM

eMail this entry!

Thank you bery, bery much. I love your Saturn essays. I assume this one was from your dad's recent visit?

Posted by: Pat on June 12, 2002 10:19 PM

BTW I am sure your dad had a perfectly logical explanation for Capcon,all the firemen answering the panic alarm and everyone else on the Cape.

Posted by: Pat on June 12, 2002 11:26 PM

Love to hear your most exceptional essays on
The Mighty Saturn, Joe

Posted by: Joe on June 14, 2002 11:07 PM

Heh. Unintentional fuel-air explosions sure can get your attention... at Rotary Rocket in 1998, we were using Silane gas, SiH4, to ignite the 5000 lb thrust rocket engines on the test stand.

It worked fine, although the techs hated it because it ignites spontaneously on contact with air or oxygen. Leaks weren't hard to find.

One evening, though, the inert nitrogen purge managed to dilute the silane enough that it didn't ignite immediately, instead allowing a cloud of silane, oxygen, and nitrogen to accumulate in the air outside the engine nozzle... then it ignited inside the chamber and the flame propagated outward.

WHAM! The noise was so loud the test operators shut it down right away, and the video playback showed lights and cables knocked down from the overhead, a moth fluttering about madly, and the entire weather enclosure shed for the stand was *bulged*.

After that, we used hydrogen gas plus a pyrotechnic (Estes D motor) for ignition... that experience was a large part of the reason we focused on reliable igniters as our first task at XCOR. 1496 runs later, we haven't had any such drama.

Posted by: Doug Jones on September 2, 2002 02:17 PM
Post a comment

Email Address:



Remember info?