ULA wins contract to launch 2020 NASA Mars rover

Please consider donating to Behind the Black, by giving either a one-time contribution or a regular subscription, as outlined in the tip jar to the right. Your support will allow me to continue covering science and culture as I have for the past twenty years, independent and free from any outside influence.

NASA today awarded ULA the contract to launch in 2020 its next Martian rover.

The contract is for $243 million, which isn’t cheap, but I think NASA decided to pay the extra money because they used an Atlas 5 to launch Curiosity, and they have been attempting to simplify the 2020 mission by duplicating Curiosity as much as possible.


  • alex

    Mr. Zimmerman: Yes, it is not cheap, but this (old-style, old-space) payload is extreme costly (2 billion?). In this regard, the launch cost fits to payload cost. We need also a paradigm shift in regard to such scientific payloads. Away from large, flag-ship style, extreme expensive single robotic probes to a large number of small scientific, “cheap” spacecraft, which are linked together and where the loss of single does not count. In this respective we need also low launch cost of a few million bucks.

  • Localfluff

    Indeed, cheaper launchers won’t compete with safer launchers above some payload value. ULA made its hundredths successful launch in a row last year. Falcon 9 is at the 95% success level, with a small sample so that it might turn out to be worse on average the next couple of years, who knows, they keep evolving their launchers, potentially introducing new unproven risks. If the launch price difference is around 10% of the payload (replacement) value, that statistics should matter big league. And as for the JWST, launch cost just doesn’t matter significantly for the budget any more. Like paying taxi fare for going to the emergency hospital with a thrombus, instead of taking the bus.

  • Tom Billings

    “Indeed, cheaper launchers won’t compete with safer launchers above some payload value.”

    The real changes will come when people accept that the fifth launch of a flight-proven stage *is*safer* than the first launch of a stage crafted as ULA does. The entirety of NASA, and its contractor club, has been beholden to the agency cost of the agents at the top of their hierarchy, the appropriation sub-committee chairs in Congress for 57 years, by now. That has warped NASA/contractor assumptions about what makes for safety, and what makes for good excuses for appropriators, who want the flow of new rockets employing old voters to build them to continue, when the universe bites back.

  • Dick Eagleson

    This decision was pretty much a lock even before its formal announcement.

    1) The Mars 2020 Rover is a Class 3 NASA mission: expensive, critical, one-of-a-kind and nuclear-powered. ULA has the only rockets certified to carry such payloads.

    2) The SpaceX Falcon 9, even if upgraded to a NASA Class 3 certification from its current Class 2 status, probably can’t handle a Mars-bound payload as big as the Mars 2020 Rover.

    3) The SpaceX Falcon Heavy, which could handle the payload, hasn’t flown yet and will require NASA certification separate from that for Falcon 9. That may or may not happen by 2020, but the decision needed to be made now.


    As to reliability, SpaceX was operating Falcon 9 at a 95% reliability rate as of its return-to-flight mission last December. It has since launched eight additional times, bringing its reliability up to 96.4%. SpaceX has nine more Falcon 9 missions scheduled through the end of this year. If it makes all these launches successfully, its reliability will stand at 97.3%.

    The statistical base of Falcon 9 flights is not that small. Falcon 9, in all versions, has flown 28 times with 27 successes. ULA’s claim of perfect reliability over 110 launches combines the records of three different vehicles: Delta II, Delta IV and Atlas V. Delta IV and Atlas V each first flew in 2002, four years before the creation of ULA. Even allowing all pre-ULA launches of each into the reliability figures, Delta IV, in all versions, has launched 33 times and Atlas V, 64 times. Thus, the statistical base for Falcon 9 already nearly equals that for Delta IV and is closing in on half of that for Atlas V. Falcon 9’s launch reliability is not based on a “small sample.”

    As noted, there are nine more Falcon 9 launches scheduled through year’s end 2016. There are also 6 more Atlas V’s and one Delta IV scheduled to fly. If all these missions are accomplished successfully, Falcon 9’s mission total will pass that of Delta IV at 37 vs. 34. It will also pass the halfway mark anent Atlas V at 37 vs. 70.

    Atlas V has launched nine times in each of the past two years. It is scheduled for 10 launches in 2016. ULA may well advance its launch cadence by one or two per year in 2017 and 2018 prior to the Vulcan’s putative debut in 2019. Even after Vulcan starts flying, there will still be Atlas V missions, just fewer each year until it is finally retired. Given NASA’s decision anent the Mars 2020 Lander mission, that retirement won’t take place until at least 2020, possibly later.

    For the sake of argument, assume an average annual flight rate of 12 Atlas V’s per year in the interval 2017 – 2019. That would take total Atlas V missions up over the century mark to 106 by year’s end 2019. If SpaceX does no better than equaling its planned 2016 performance – 17 Falcon missions – in each of the years 2017 – 2019, it will still finish 2019 with a total of 88 missions flown.

    It is far likelier, though, that SpaceX will exceed its 2016 launch cadence in each of the next three years, probably by quite a bit. Falcon Heavy will enter service and the launch cadence of Falcon 9 will continue to rise as Commercial Crew flights will be added to the ongoing CRS freight missions and what is likely to be an even faster cadence of comsat and other missions. LC-39A and Boca Chica launch sites should also be entering service over the 2017 – 2019 interval so a considerably higher overall launch cadence will be possible without straining any particular launch facility overmuch. I think it likely SpaceX will launch at least two dozen Falcons next year, more in 2018 and more yet in 2019. By year-end 2019, the Falcons will almost certainly have tied or surpassed Atlas V’s overall mission count. This could even happen by year-end 2018.

    Nor is there any particular reason to think SpaceX will have any more failures over the next three years. Certainly not due to continuing improvements to the Falcon 9 launch vehicle. The one failure Falcon 9 has had was due to misfeasance by a subcontractor and occurred toward the end of that particular model of Falcon 9’s run of service, not at the beginning.

    It’s also worth noting that ULA’s “perfect” mission success record includes a few sub-optimal outcomes including the RD-180 early shutdown anomaly on the second Cygnus mission it flew for Orbital-ATK. Had that same anomaly occurred during launch of a heavy Mars probe mission, the Centaur would not have had sufficient propellant to make up the deficit as it was, fortunately, able to do in the case of the Cygnus mission. ULA is good, but it has also had a bit of luck on its side.

    The best endorsement of SpaceX’s reliability comes from the insurers who cover comsat launches. These are people who have actual skin in the game of launch vehicle reliability. Falcon 9 missions get insured at the same rates charged for the missions of Ariane 5, another famously reliable rocket that has an even longer string of unbroken successes than does Atlas V.

  • Edward

    alex wrote: “We need also a paradigm shift in regard to such scientific payloads. Away from large, flag-ship style, extreme expensive single robotic probes to a large number of small scientific, “cheap” spacecraft, which are linked together and where the loss of single does not count.”

    Excellent point, but this has already been thought of and done. After Mars Observer was lost, NASA did exactly this, calling it: “faster, better, cheaper,” but after a couple of failures of these cheaper spacecraft and landers, it became clear that the political fallout was the same as for more expensive lost spacecraft. Thus, since fewer expensive spacecraft are launched, there is more time between failures, so the “faster, better, cheaper” paradigm is not much talked about, anymore.

    Dick Eagleson wrote: “I think it likely SpaceX will launch at least two dozen Falcons next year”

    I’m not quite that optimistic, but I agree that –barring another failed launch — they will ramp up to more launches next year than this. My expectation is that as they add their own private launch pad, they will be able to ramp up to a very rapid cadence. I suspect that their own pad is being built for a rapid cadence, but heritage pads were built more for monthly, not weekly, launch rates.

    SpaceX’s (Space Explorations’s) commitments to reusability, rapid launch cadence, and the exploration of space are three of the reasons why they are such an exciting space company. They are challenging the rest of the industry to improve in ways unimaginable, only ten years ago. Those companies that rise to the challenge will survive, but those that don’t will become museum exhibits of early space pioneers.

  • Steve Earle

    The “faster, better, cheaper” paradigm might be gone, but I would like to see a return to the old mindset of:

    “Why buy one, when you can have two at twice the price” (can anyone name that movie quote?)

    Two Pioneers (10 and 11), two Voyagers, two Vikings, two Mars Rovers, etc. etc.

    When they both work you get twice the results, when one fails you still have a mission.

  • Alex

    Edward and others: I am proposing as next step the application of a “massive” number of very small interplanetary S/C (number: 50-100?, single S/C mass < 50 kg?), only hundreds of thousands miles separated each other, which are interconnected and linked each other to build a flying network.

    In 100 years realized laser driven interstellar missions need also a continuous string of thousands of launched extreme very small S/C (traveling at 0.2 c) to bridge huge interstellar distances to nearby stars.

  • Localfluff

    When SpaceX gets their reusability working and multiply their launch frequency, they will also become the most reliable launcher, that’s true. Because of the routine, the flight proven hardware, and the detailed understanding of what is actually happening to their landed hardware during launches. But they’ve been doing so very well thus far that the contrarian inside of me comes out and warns that they might’ve been lucky, and that there are a couple of bumps ahead for them to deal with. They are still evolving quickly and that introduces risks. But it’ll just be a matter of time, if it takes two or four years. If they need to add a few tons of launch mass to improve refurbishments they have the margin to do that

    Dick Eagleson,
    On your point 2), SpaceX recently upgraded to Falcon 9 Full Thrust, which is able to launch about 20% more than the Atlas V 551 to LEO (22 versus 18 tons). Even if all of those “mass to LEO”-figures are fuzzy, F9FT should be as capable as Atlas to Mars. But it failed last year, so I understand why NASA chose ULA (and I’m glad they did!) Btw, The Ariane 5 will launch the JWST. I think that ESA pays for the launch cost to help their scientists be part of the mission.

  • Edward

    Alex wrote: “I am proposing … a continuous string of thousands of launched extreme very small S/C (traveling at 0.2 c) to bridge huge interstellar distances to nearby stars.”

    What a good idea for relaying data from and commands to interstellar probes. It allows for lower power transmitters and smaller antennas. The relays may not need to travel quite so fast, but the concept seems good.

  • Localfluff

    How will these probes be powered? I could imagine hibernation and then battery power during the short flyby and data transmission. But within 3 gramme payload? If the whole chain of probes timed its battery activation maybe, and all but one spent it all on radio. And every link in the chain still works several decades after launch.

  • wayne

    “When they both work you get twice the results, when one fails you still have a mission.”
    –that is an excellent point.
    –Space hardware is akin to Pharmaceuticals– costs a fortune to crank out the first unit, but the marginal-cost for each additional unit, drops dramatically.
    “Why buy one, when you can have two at twice the price.”
    — That’s either “2010” or “Contact?” (or maybe “Godfather” 1, 2, or 3?)

    Alex– I like your idea! Start writing your Patent application!
    ( Remember Arthur C. Clark! > http://lakdiva.org/clarke/1945ww/ )

  • Alex

    Hello Wayne, the idea of geostationary orbit is much older as 1945.

    I state: “However it was Hermann Oberth and Hermann Potocnik who wrote about orbiting stations at an altitude of 35 900 km above the Earth that had a rotational period of 24 hours making it appear to hover over a fixed point on the equator. ”

    I remind myself also to a book from Hermann Oberth, which date back to 1929, in which he described the concept in detail. I learned now that Oberth even complied 1923 for the German Rocket Society about of the geostationary orbit concept. Other German space travel pioneers (as Hermann Noordung, Willy Ley) detailed it later out (following Oberth) several times, many before Clarke. Not to mention Tsiolkovksy.
    Many years before Clarke wrote about it, it was already a quite common idea in space oriented circles.



  • Dick Eagleson


    It would be nice if SpaceX could do a launch a week from Bolsa Chica, but the deal it has with the locals is for 12 missions per year only two of which can be Heavies. This is due to the necessity of closing the nearby state beach on launch days. Once the facility is actually built and operating at permitted capacity, the question of launch cadence can probably be revisited. If the net effect on third-party local businesses of SpaceX’s presence has been positive – e.g., launches as additional tourism attractions, regular purchases from local vendors, etc. – I expect SpaceX would get a favorable response despite the need for more beach closings on an increased number of launch days.

    As for the legacy pads, SpaceX’s SLC-40 has already supported launch cadences appreciably higher than one per month on several occasions including consecutive turnarounds of as little as two weeks. I see no reason to suppose LC-39A will not be equally capable once it goes live. These may be legacy pads, but they’ve been pretty thoroughly reworked by SpaceX. SpaceX’s default of horizontal integration and transporter-erector technology makes prepping the rockets and getting them out to the pads a quick process compared to that used by many legacy launch vehicles.

    My projection of two dozen missions next year is based on:

    1) at least as many comsat and CRS missions as last year from Florida. The CRS missions may start going up from LC-39A. If so, that frees up SLC-40 to do even more comsat and other types of missions.

    2) Three or four additional missions from LC-39A related to Commercial Crew.

    3) Two or three additional missions from LC-39A for Falcon Heavy.

    4) At least seven missions from Vandenberg, well up from the four they intend to launch from there before the end of this year.

    Barring another failure of Falcon 9, I think 24 missions for 2017 is a fairly conservative estimate. I wouldn’t be at all shocked if SpaceX manages to do 30.

    Payload readiness is likely to be more of a limitation than either manufacturing capacity or ability to maintain a rapid operations tempo, especially as 1st-stage reusability becomes routine practice. If SpaceX manages to do two missions this year on reused 1st stages, I see no reason to suppose they won’t find takers for at least a half dozen such missions next year. There could easily be more than that.

    Next year’s missions out of Vandy are all relatively low-energy missions so the 1st stages for these shouldn’t be stressed any harder than on a CRS run. They may even all be recoverable on land which would speed the reuse cycle.


    I’d say SpaceX has been more skillful than lucky. If anything, they’ve been unlucky. The CRS-7 failure resulted from use of a particularly out-of-spec part from an external supplier, most of whose products were later found to be below standard.

    Changes can introduce unanticipated risks, true, but at least some changes are also made to provide even better margins against suffering from anticipated risks and/or to improve performance. Thus far, SpaceX’s changes appear to have improved, rather than compromised, both performance and reliability. The only engine failure SpaceX ever had on a Falcon 9 was on an early flight of the v1.0 model. And it wasn’t the F9 FT that suffered the CRS-7 failure, by the way, but the previous version of F9. The FT has a perfect success record to-date.

    Perhaps the current version of Falcon 9 could handle the Mars 2020 Rover mission, but, I suspect, only if launched as an expendable. Falcon Heavy would likely be mostly reusable after boosting the same mission.

    You’re correct that use of Ariane 5 for JWST is an in-kind contribution from the ESA.

    As for powering Starwisp-type interstellar probes, the power would come from the same laser beams that push them outward from Sol.


    Your swarm probes idea is likely to be implemented some time in the not-too-distant future. But I don’t expect that it will be NASA doing so, at least not in a pioneering way. It seems far likelier that such a mission design would appeal more to one of the asteroid mining outfits as a way to get multiple prospector craft out to multiple NEOs or even to multiple Belt asteroids. This would be a good mission for Falcon Heavy.

  • Edward

    Dick Eagleson wrote: “It would be nice if SpaceX could do a launch a week from Bolsa Chica, but the deal it has with the locals is for 12 missions per year only two of which can be Heavies.”

    Well, that’s a bummer.

  • wayne

    Dick Eagleson
    — good stuff! I enjoy your insight!

    — thanks for the links. That looks like an interesting book! (“Science Fact and Science Fiction: An Encyclopedia” von Brian M. Stableford)

Leave a Reply

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