Some Thoughts on Cost Overruns

Over the last several days, I've been thinking about the interview with the head of NASA's science program, Ed Weiler, posted by the journal Science. (See here for a summary of the accompanying article.)

Some of his thoughts have stuck in my mind: "Most "cost estimates" that go to the [National Research Council]-and I'm speaking from 31 years of experience with decadals-come from NASA centers or contractors; they don't come from NASA headquarters, and they don't get reviewed by NASA headquarters... Scientists cannot do cost estimation, I'm sorry, they're not trained for that... Anybody who compares a decadal cost-survey estimate to anything in reality, well, I have no comment on that kind of person... And if you think we don't get mean with contractors, go talk to some of the CEOs who've sat here and got an earful from me."

I have known CEOs of several of companies, including some on the Fortune 100 list. Their egos are quite big enough to take a tongue lashing from Weiler. And what is NASA going to do to punish these companies? Tell JPL that they can't build any more planetary spacecraft? Who else has expertise in a number of critical areas? Or tell Lockheed-Martin or Boeing that they can't bid on any more NASA contracts? There are only 2 - 3 companies big enough to support these missions.

Another part of the Weiler interview also got me to thinking. He said, "A billion dollars doesn't buy much any more, it buys you 3500 major aerospace engineers for a year."

The planetary community is starting another Decadal Survey to prioritize the next decade's missions. (They are planning on significantly improving cost estimates.) My best guess at NASA's purchasing power for the next decade -- extrapolating from current budget and guessing that there may be years of frozen budgets as there were the last decade -- is that NASA will have $6-8 billion to spend in 2011 - 2020 on new missions in today's dollars.

Here are some costs for missions that I have gathered over the last year stated approximately in today's dollars (figures in millions of dollars):

Jupiter Europa Orbiter	$2,700
Minimum Mars Science Orbiter	500
Mars Network	1000
Mars Mid-range rover	1400
Comet sample return	1000
New Frontiers (each)	650
Discovery (each)	375

In addition to these missions, NASA is likely to spend $200M+ in this timeframe developing the Mars Maven orbiter (2013 launch) and perhaps $200 to support ESA's ExoMars mission.

Here's what one line up of missions might look like:

		cumulative
MAVEN & ExoMars	400	400
2 Discovery	750	1150
2 New Frontiers	1300	2450
Minimum MSO	500	2950
Mars mid-range rover	1400	4350
Jupiter Europa Orbiter	2700	7050

If the mission costs are approximately right and there are no major cost overruns, then this is a rich mission set. Mars receives several new missions, Jupiter and Europa finally get their return visit, and there are some smaller missions to provide balance. On the other hand, there's not a lot of room for error here and still maintain a balanced program. If the mid-range rover and the Jupiter Europa Orbiter, for example, overrun by 50% each, the Discovery and New Frontiers missions become unaffordable.