zaterdag 5 oktober 2013

SLS: Some more about costs

Some weeks ago I wrote a blog about SLS costs. Now, it was more meant to provide a rough idea as to what SLS would cost per launch, not a full on cost analysis; the methods I used for this were obviously flawed in several ways. In this one I'll look a NASA budget availability study from 2011 (I won't do my own analysis), as well as some more things about budget that I often see confusion about.

What exactly is "SLS"?
This is a dumb sounding question that is actually a lot more complicated than you might think. When people talk about "SLS" they often mean the complete system, not just the rocket which is what I refer to as "SLS". The current Exploration Systems Development programs contains three different main programs: SLS, MPCV, and 21st Century Ground Systems. They all have their own costs and their own purposes, with the current budget looking like roughly $1.4 billion for SLS, $1 billion for the MPCV and about $400 million for 21st CGS, totaling at 2.8 billion dollars. Now, when you want to compare the costs of SLS with other systems, it's very important with what you actually compare. If you compare SLS to a complete system like the Space Shuttle, you need to factor in complete costs. However, if you compare it with just a launch vehicle, something like Saturn V or Falcon Heavy, like is often done, you need to look only at SLS itself.

"According to NASA, SLS would cost $10 Billion per launch." No, it wouldn't.
I've seen this one far too often. The study this refers to does indeed claim the entire system would cost $41 billion through to FY2025. However, taking this as "SLS cost $10 billion" is completely wrong. First of all, it includes the entire system, not SLS. The costs for SLS itself are "only" 20.245 billion dollars. In the second case scenario, SLS can launch five times for 21.524 billion dollars. That's a 1.279 billion dollar difference for a single SLS. Not quite 10 billion per launch. Anyway, this study takes development costs into account as well as the first few flights, and with a projected total development cost of $18 billion dollars and with a low initial flight rate of 1 every two years, I'm hardly surprised cost would run up that high.

How much would it really cost according to that study? For SLS alone, if you take case #2, which is president budget+escalation, it's estimated SLS can launch once a year from 2023 on at an annual budget of $1.73 billion dollars. I should say, however, that these are estimated 2023 dollars. These numbers are estimated with an inflation rate of 2.5% per year, which means that translating it to 2013 dollars gives a modern day flight cost of $1.35 billion, for one flight per year. That includes on going development of SLS Block 2. If you take SLS as a complete system, it would cost about $2.8 billion, of which ~1.4 is for SLS, ~0.8 is for Orion and ~0.4 is for 21st CGS. This budget fits roughly within the current budget. So, SLS would, according to the study, be capable of one flight per year even under current budgets when adjusted for inflation.

Final words on cost discussion
To be honest, I think it's better to not argue too much about the cost of SLS. There are simply way too many unknowns at this point. The NASA study I talked about assumes SLS as a Space Shuttle External tank with SSME engines and Ares 1 boosters and upper stage engines. However, many design changes have been made that would affect costs in a number of ways. The core, for example, is pretty different from the STS ET. It's a lot bigger and uses more engines, for example, but on the other hand it uses easier to work with materials, new more automated tooling and new "mass produced" engines that all significantly reduce the amount of people needed. The boosters will be replaced after 2023 if the Block 1A approach is taken, and while these boosters are much more capable than the current ones, they also offer a number of ways to reduce costs, be it either cheaper and easier materials (ATK's AB), liquid kerosene propulsion (Aerojet, Dynetics) or simply using cheaper, existing engines (RS-68 boosters are still being considered). If the Block 1B approach is taken, SLS will use a Dual Use Upper Stage, which is a lot lower on development costs than the J-2X powered Large Upper Stage currently baselined.  And then there's the fact that it's just all estimates. SLS cost could end up much lower or much higher than what is currently estimated at about 1.4 billion (or the $500 million cost figure that makes no sense and is very unclear in every way). Until Block 1A or 1B has flown we simply can't know for sure what the costs for the rocket will be. The only answer for SLS costs you can give right now is "somewhere between 500 million and 3 billion."