Doug Plata Positively Reviews SpaceX Super Heavy Starship

32 thoughts on “Doug Plata Positively Reviews SpaceX Super Heavy Starship”

1. So, if I’m understanding you, you’re saying that the dry mass penalty associated with SEP over chemical is large enough that it doesn’t make sense below some threshold demand for delta-v. That sounds right to me. And it also sounds right that that threshold often requires so much prop that the low thrust puts the time-on-station so far in the future that it’s not economically feasible.

2. I accept that. Your numbers are also good. 
   I checked.

   And you’re right about “taking long at lower power”. My basic contention was that that ratio: of available power and available reaction mass kind of balances out to where the high ISP thing doesn’t make sense when the power generating capacity is very low. Even if the amount of reaction mass is high (indicating a really good ΔV potential), the amount of time to exhaust it becomes egregiously long.

   GoatGuy

3. Terrific job, Doug, and thanks for the update below.

4. THAT is a fair point. They’d get there, but not in a hurry.

5. That’s a good point: Low thrust, high ISP options, like ion thrusters, only start to show a clear advantage once the mission has enough delta V requirement to make up for their higher overhead mass.

   However, several points:

   1. In this case the satellite has to have the propulsion system anyway, so you can’t count the full mass of the propulsion system against this option, a good deal of it counts as payload. I estimate that a 500kg satellite only needs about 170kg added mass in propellant and tankage.
   2. Even if it were a wash, it would establish that using the satellite’s own propulsion system to make up Radical Moderate’s azimuth deficit is a viable option.
   3. But, most important, what we’re talking about here is staging. The BFR brings itself and the satellites to LEO, then only the satellites themselves have to proceed to the designated orbits

   That last makes a big difference.

6. I’m pretty sure the only limits on how much xenon or argon you can haul around are how big a tank you want. Solar panel weight is an issue, but only if you want to increase thrust.

   The real problem is that, as the tank gets bigger, the burn time to exhaust everything in the tank gets longer and longer. T = (mass flow) * (exhaust speed). For an 80 mN thruster (about 1500 W power requirement) and Isp=1600 s, that means that mass flow is 5.1E-6 kg/sec. So even 100 kg of prop takes 7.6 months to “burn”.

   So “choose the high Isp, low thrust, when you can get away with it” is pretty good advice. It’s just that “when you can get away with it” often requires waiting for several years to get your payload where you want it.

7. About this tipping point idea: It’s such a massive threat to the entire launch industry beyond SpaceX that I think it’s too dangerous for NASA to entertain. NASA can’t afford for ULA and BO and the international partners see no future beyond BFR. If they all take their bat and go home, the funding for everything will dry up.

   The proper thing to do is to admit publicly that distributed launch is the way to go, right now. (I’m pretty sure that the architecture guys have already admitted this privately, or they wouldn’t be considering 3-stage lunar lander architectures.) Once that’s done we can go about the important work of getting the TRL up on:

   1) Long-dwell (at least two weeks) cryogenic storage.
   2) Orbital rendezvous and docking/berthing of cryogenic prop.
   3) Cryogenic prop transfer.
   4) Routine, low-risk assembly of separately launched spacecraft components.
   5) Lunar precision landing. (You can’t preposition prop on the surface if the crew lands 20 miles away.)

   Note that BFR/BFS fits perfectly well into a distributed launch architecture, especially since it’s a lot better at delivering big payloads to LEO than it is at landing on the Moon.

8. They could make oversized prop tanks on them, but there are some problems:

   1) You’re going to run into fairing volume constraints at some point.

   2) IIRC, argon tanks are pressurized. Big pressurized tanks=heavy tanks. Also, big pressurized tanks=more debris when they get breached.

   3) Heavy birds have higher ballistic coefficients, which lengthens their worst-case disposal time.

   4) The birds are probably going to have a power budget of <1500 W. I wouldn’t count on more than 80 mN of thrust. If we use 500 kg as the mass of what has to be delivered to orbit, we need 137 kg of argon for the plane change. It’ll take more than 10 months to burn that much. But if we only get 75% sunlight in LEO, it’s closer to 14 months. That limits the amount of time that BFR can make a dent in the problem pretty substantially.

   BTW, I tried refueling the BFS on-orbit, which effectively halves the per-launch capacity to something like 70-ish. That’s not a lot more than an FH could do with a Cat C fairing extension (about 60 by my reckoning).

9. Hmmm… Tsiolovsky’s Rocket Equation shows something kind of interesting.

   The thrust from a chemical rocket, ISP=450 (hydrogen, vacuum motor cone), burning 90% of its rest mass is about 4,400 m/s delta-V.

   Contrarily, a ion engine, powered by big old solar panels, using mercury or lithium or virtually any other ion, having ISP=1600, but burning (ionizing) its reduced tank capacity (of if you will, the higher overhead of propulsion motor mass and energy capture array) of 50%, delivers almost exactly the same delta-V of 4,400 m/s.

   So, no matter how attractive the sentiment, the tradeoff isn’t just as easy as “choose the high ISP, low thrust, when you can get away with it” metric.

   Just saying,
   GoatGuy

10. Yeah, but the tipping point occurs when the decision-makers recognize that an alternate is sufficiently capable and a better deal from a cost standpoint. I believe that even a BFR/SHS that retrieves its first stage and can put up 100 tonnes would meet the criteria for the tipping point. Certainly we hope that he recovers the tanker stage but even if not, it’s still a significant improvement from a cost standpoint. With retrieval of the first stage and without all of the bureaucratic burden, I would guess that a partially-reusable BFR/SHS would be about 1/10th the price per tonne to LEO. Still a game-changer.

11. … So I don’t believe that the Starship has to return from orbit and land in order to reach the tipping point. …

    Yet, without that, the most central core of the Musk-o-verse promise is washed away. Basically, he has always been promoting not just new-and-different from conventional play, but also critically the idea of much cheaper due to re-use.

    Thing is, that re-use apparently takes somewhere between 25% and 40% of the ultimate payload. Per kilogram of fuel and oxidizer.

    The BFR AKA SHS really needs the recycling-and-fast-recertification gambit in order that the “under $100 per pound, to LEO” is achieved.

    Just saying, GoatGuy

12. The Hall effect thruster has an ISP of 1600. The BFR has an ISP of 330, or 380 if they run vacuum engines.

    From a mass budget standpoint, doesn’t it simply make sense to put the satellites into the lowest, least energetic feasible orbit, and do as much of the maneuvering with the higher ISP engines as possible? Especially since this simplifies their flight logistics, they’re doing the same flight for a hundred or more satellites at the same time.

    Save chemical propulsion for maneuvers that have to be high G. Do everything you can with the low acceleration, high ISP propulsion system.

    Remember, SpaceX is designing these satellites themselves. THEY designate how much delta V they are capable of, not some third party. If they want an extra 110kg of propellant capacity, they just design it in. It’s worth doing if it saves them fuel on the front end.

13. To change inclination from 26 degrees (Boca Chica) to 53 degrees (Starlink group 1 inclination) at a 550 kilometer altitude requires 3805 m/s of delta-v. For a Hall thruster with Isp=1600 s and the published dry mass of 386 kg for the Starlink, that would be 106 kg of propellant. (It’s probably more than that, because SEP burns aren’t impulsive.)

    My guess is that’s not only all of the stationkeeping prop, but it’s also more prop than the bus can carry.

14. Thinner air, and the highest wind speeds seen in Martian dust storms have been 60mph, so practically no force to speak of compared to an ordinary wind on Earth.

15. “3) BFR from Boca Chica is too azimuth-constrained to hit the inclinations needed for Starlink, so it probably doesn’t help unless they build a pad at the Cape.”

    Depends on how much station keeping Delta V the Starlink satellites have; I understand the plan is to just dump them into LEO, and then position them in the correct orbits using the high ISP Hall effect thrusters. Presumably this could involve taking them to the necessary azimuth, too.

16. Elon can beat this timeline.  I would bet somewhere between Elon’s schedule and this one.

17. Mat … Need more infastructure before sending people to mars. Strategic reflective material on phobos. Lots of water from the moon if not on mars. Cutting down or eleviating dust storms. More plausible habitats. Buried in craters or if on surface paint them dark colors so they absorb not reflect light.

18. No, it ran out of the fluid used to re-light the engines and it splashed into the ocean beside ‘Just Read the Instructions.’ The intent was to recover all 3 boosters, but the center core was lost.

19. 99% less – ie. absolutely impossible – it would have to be teetering like a coin standing on its edge for anything to happen

20. I hate the new name – ‘Starship’ – it sounds too corny

21. I hope they’re successful too. I’m confident that they will be, eventually. But I’m going to list all the things that make me skeptical/worried:

    1) The extreme range capacity problems SpaceX will have getting enough Starlink launches to make the “use it or lose it” FCC license deadlines in 3/2024 and 11/2024.

    2) BFR could easily be starved for financing by Starlink deadlines.

    3) BFR from Boca Chica is too azimuth-constrained to hit the inclinations needed for Starlink, so it probably doesn’t help unless they build a pad at the Cape.

    4) I agree that I’d tack about 2-4 years onto Elon time for the BFB/BFS cargo version.

    5) I don’t think a crewed BFB/BFS launch will ever happen. Even with no payload, BFS T/W is barely 2.2, to say nothing of the 6 or so you need for launch escape. The Raptors can’t spin up fast enough for abort purposes, and the interstage can’t vent without the pneumatic separation. I expect a radically different BFS variant for launching crew to LEO. BFS is fine to carry crew BEO, but getting them launched will be way too scary for NASA, and likely too scary for the FAA.

    6) Neither Bezos nor Bruno is an idiot. I worry that SpaceX could fail to engage in the ongoing architecture wars, only to discover that cis-lunar is populated with ACES and Blue Moon systems by the time they’re ready. They’re gonna have to get dirty on the SLS debate sooner than they’d like.

    7) I’m pretty sure that the SSTO thing for BFS is nonsense. SSTSO is doable, though.

22. Of course he will, he stated it in his last 5 years plan tweet.

23. Wasn’t the FH center stage a voluntary non RTLS/RTDS?

24. https://twitter.com/elonmusk/status/1015648140341403648

25. If they won’t tip over on Earth, what makes you think they’ll tip over on Mars with a 90% less atmospheric pressure at the surface?

26. Looking at that artist’s concept, is there any chance Martian winds can tip one of those over, even considering how thin the atmosphere is?

27. The landing attempts was an gradual improvement by testing, first it was the added systems to control first stage after separation and controlled reentry at one point they found they was experienced enough that they could try to land.
    Then they first managed to land it was one fail in addition to the FH core I think.

28. Thanks. I missed it then. Was watching it late last night.

29. I think that I did mention it but the F9 development had, IIRC, nine or was it eleven setbacks. But you are right that the FH was successful on the first launch. Yet the FH was (to some extent) three F9s. So, a whole lot of heritage there. What I did was to take into account the learning that SpaceX has and so said that my educated guess was that there would be only two setbacks with the BFR and not another 9 to 11 setbacks.

30. Watched the video Doug.

    I may have missed it, but when you talked about all the failures of SpaceX with the landings… it seems you did not consider Falcon Heavy basically worked the first time. Showing that SpaceX has been collecting data and each new launch got closer and closer to solving all problems.

    Ok, BFS will be a really NEW booster/ship, unlike Falcon Heavy.

    But how do you see past SpaceX learning and step improvements in relation to building the whole new rocket architecture?

31. In my presentation, I make the arguement that the 2022 deadline is unlikely to be met not just because Elon has a history of missing deadlines but because it is more likely than not that there will be at least one setback (e.g. crash) which will delay the BFR/SHS. But, in the final analysis, even a delayed BFS is a total game-changer so it hardly matters.

    Since my presentation, two things have happened. The BFR has been changed to the Super Heavy Starship and we have been told that substantive changes to the BFR consideration will be made. I would point out that the ITS/BFR/SHS design has been scaling down so that now, the SHS is only 57% more capable as the Falcon Heavy. As the SHS becomes more modest, I believe that it becomes more feasible.

    One additional change to my thinking since my presentation, I made the case that the BFS would need to reach orbit, return, and land in order to reach the tipping point for the transition of support from SLS to the SHS. That may (or may not) require the development of the Big Falcon Booster (SHB). But for the SHS to be more cost-effective than the SLS, really, it only has to deliver its payload to LEO. So I don’t believe that the Starship has to return from orbit and land in order to reach the tipping point.

    Best of luck to SpaceX. Here’s hoping that they are successful.

32. Commercial Falcon Heavy to lunar surface might still be first if you really want to reach that arbitrary 2022 date, only 3 years away. If Europe/US/India/China all continue to make the right noises and could be massaged into acting as a first customer for surface services to a robotic/human village, there will be an entrepreneur (probably a data provider like Surrey/O2) spotting an opportunity. This entrepreneur
    might be able to convince his investors to take the 150M launch and 200M medium tech infrastructure gamble. Maybe Elon’s Japanese artist client who is paying for ~8 passengers (and billionaire friends) could be persuaded to pay for an art installation on the Pole(-s) of the Moon (e.g. a mythology evoking wooden tree of the world)….with a bulldozer, sublimator, habitat attached to it…by 2022

Comments are closed.