I listened in a bit on yesterday’s Augustine Committee discussion. The debate at the end between Bo and the others reminded me of a point I’ve made a bunch of times on this blog–that when you’re talking about exploration missions, ascent reliability is only a small component of the overall risk.
Since I’ve been doing a lot of rocket plumbing, let me use an analogy. Imagine you have a small rocket engine, like our igniter. It’s got a tiny fuel orifice that’s only a tiny fraction of an inch in diameter. Now, if the solenoid valve upstream of that orifice is small enough, it can meaningfully decrease the overall flow. But after you reach a certain valve size, the valve size ceases to be relevant to the overall flow rate. You could put a 1″ full port ball valve leading up to that orifice, and the flow difference between that an a solenoid valve with a 1/32″ port is going to be round-off error. Once the size difference between the most constricting component and the rest of the components gets past a certain point, you can pretty much ignore them.
There are tons of other analogies from electronics, manufacturing, structures, etc. Basically, in almost any system, you can improve one component only so far before you hit rapidly diminishing returns. A wise engineer will spend his resources in a way to maximize the overall system reliability, not just one small sub-component.
Unfortunately, that’s exactly the mistake that the CxP guys have been making with Ares-I. Even if, in spite of all evidence so far, and in spite of all historical precedent, Ares-I really is as reliable as their Probibalistic Risk Assesments suggest, it still isn’t a wise investment of capital when you look at the overall exploration mission.
Let’s just go back to the math again.
According to ESAS, the predicted probability of losing a crew on a lunar mission was something like 1.6% (or about 1/60). Of that, only about 1/2000 (or .05%) came from ascent risk–or about 3% of the overall crew risk for the entire mission. A mission that used the worst numbers they came up with for existing EELVs with an LAS attached estimated a 1/600 probability of losing a crew (or about .16% chance). That would increase the probability of losing a crew to 1.7% or about 1/59…
Investing tens of billions of dollars to reduce the probability of losing a crew from 1.7% to 1.6% only makes sense if there are no better safety investments out there. With a 1/600 ascent safety rating, about 90% of the danger to the crew is coming from other phases of the mission–which strongly suggests that the best safety return on investment is not in overoptimizing the ascent reliability.
Now, to be fair, Shuttle has a safety rating estimated to be around 1/100. At that reliability rating for crew launch, it would be one of the dominant crew safety risks in a lunar mission. Spending money to get it up past the 1/500 range is a good investment. But after a point, if your goal is to improve the overall probability of getting a crew safely to the moon and back, you’re best off finding other areas to invest your money.
On a related note, the self-righteous attitude that many CxP people like Griffin and Hanley take that the Augustine Committee is ignoring crew safety is kind of farsically hypocritical at best. Blowing so much of your budget on ascent risks, when they aren’t the dominant risk is actually making the overall mission less safe, not more. Wisely spending that money instead on mitigating the risks in lunar landing, ascent, surface ops, and earth return would result in a higher probability of not killing astronauts on a given mission. If astronaut safety is really so important to CxP, why do they only seem to care about the first 5-10 minutes of the mission, when it’s the rest of the mission that accounts for 90-97% of the danger?