Failure not only IS an option, it’s inevitable

Seth Jarvis

First, a disclaimer: ATK Space Systems is a financial supporter of Clark Planetarium. They sponsor our IMAX Theatre through annual cash and in-kind donations. That said, I would have written this post even if ATK Space Systems never gave the planetarium a dime.

I have to weigh-in on the last Ares booster ground test that was scrubbed at T-minus 20 seconds on Thursday, August 27.

I was there, along with a handful of other planetarium staff and members of our board of directors.  It was hot, crowded and exciting as we jostled for position so we could get a view of the first-ever test firing of an Ares booster (a longer and more powerful version of the Space Shuttle’s Solid Rocket Booster) from a safe distance of a just over a mile.

The Ares 1 First Stage is the long white tube lying on its side. It is 156 feet long and will burn its million pounds of solid fuel in 122 seconds, while generating 3.6 million pounds of thrust.

The Ares 1 First Stage is the long white tube lying on its side. It is 156 feet long and will burn its 1.38 million pounds of solid fuel in 122 seconds.

T-minus 25! 24! 23! 22! 21! 20! Then the clock stopped. The voice over the loudspeaker announced that there was a “hold.”

It was so exciting!  And then...

It was so exciting! And then...

Ten minutes passed.  Carloads of engineers could be seen driving purposefully up the road toward the rocket.  For a moment I entertained myself with thoughts of a burly, red-faced engineer angrily striding up to the rocket, giving it a sharp kick and then shouting into a radio, “OK, try it now!” or dozens of men in lab coats swarming around the back end of the rocket as they frantically patted their pockets, yelling to one another, “Who’s got the matches?”

Several more minutes passed, and the voice on the loudspeaker announced that the test had been scrubbed for the day.


ATK has identified the problem as a failure of a valve to supply fuel to one of the Auxiliary Power Units that powers a small but powerful turbine that powers a small but powerful pump that creates the hydraulic pressure that allows the rocket nozzle to be tilted a few degrees in any direction while the rocket is firing (Everyone clear on that?).

Directing the thrust from a rocket engine is how you “steer” as you roar into space, which is obviously a critically important thing for a rocket to be able to do.  In fact, it is so critical that the rockets are designed with redundant systems so that if one APU fails it has a separate backup system up and running that can do the job by itself.

When astronaut lives and a billion dollars’ worth of hardware are at stake you make sure that valve failures are vanishingly rare events and that critical systems always have a backup standing by.

When a valve that is this critical to the mission (to say nothing of astronaut lives) fails during a test you don’t write it off as “just one of those things,” you stop and get to the bottom of the problem. These valves aren’t supposed to fail.

Over this last weekend I had a gas station conversation with an acquaintance from my old neighborhood.  Knowing that I was affiliated with the planetarium and had an interest in things spacey, he asked, “So whatcha think of that big rocket failure?”  Ugh.  Before I could finish my answer, (“Well… ‘failure’ is a loaded word…”) his next question was, “I wonder how many millions they wasted?”

Double-ugh.  How to even begin addressing such a perspective?

Both our gas tanks were filled and “Well, see ya laters” were quickly exchanged without so much as an “It’s not that simple…” from me.

Rocket science is hard.  If it were easy, everyone would be doing it.  There is a reason the U.S. is the world leader in spaceflight – we have been at it longer, we have worked harder on the difficult science and technologies involved and as a result we have been more successful at it than any other nation on earth.

Always keep in mind that omniscience is not a human trait.  Failure – whether it is a failure of a component, a failure of a design or a failure of an entire design philosophy, is inevitable. The only choice any of us have in the matter is how we prepare for the inevitability of failure, what we learn when we encounter it and what we do with what we learn from the experience.

An axiom that has guided me over the years is, “Thought is born of failure.”

To think one thing and then later discover that you were wrong is a natural part of life.  Some would even say that making mistakes is necessary to progress.  If you are never wrong then you never have a reason to learn anything genuinely new.

Where do you want to discover that there is a problem with a critical fuel valve when major scientific advancement, major economic development and international prestige depend on that valve’s function?  On the test bed, or on the launch pad?

The Ares 1 test is scheduled to take place this Thursday, September 10, and the odds are very good that when the test is concluded ATK engineers will have truckloads of data that will be encouraging, scientifically valuable, and that will reveal new challenges that will need to be overcome before astronauts ride an Ares booster into space.

There is a wonderful latin phrase that sums up everything I’m trying to say here: Per aspera ad astra. (“To the stars through difficulties.”)

“Through difficulties.” This is the way we got to the Moon forty years ago, and it is the way we will go back to the Moon and beyond.

Tags: , , , ,

2 thoughts on “Failure not only IS an option, it’s inevitable

  1. Yes – it’s an issue with a failed server, a shift to a new web host, followed by the corruption of backed-up data. Meltdown. In other words, I don’t think we’ll ever be able to restore the images from many of our older blog posts.

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>