24

According to Energy Consulting Group's "SpaceX's Starship May Nearly Triple US Natural Gas Demand" page,

Our analysis indicates that a fully configured Starship launch (booster and Starship) will use about a 1000 tonnes of methane in the form of LNG as fuel. This is equivalent to approximately 50 million standard cubic feet (mmscf) of methane. [...] This works out to be about 150 mmscfd per Starship, or about 150 billion cubic feet of methane/natural gas per day (bcfd) for a fleet of 1000 rockets. 150 bcfd of natural gas is roughly equivalent to 25 million barrels a day of oil. For reference, recent (2019) US demand for methane in the form of natural gas averaged about 82 bcfd. In essence, Mr. Musk is suggesting that US demand for natural gas could grow dramatically over the next 10 or so years due to SpaceX activity.

Are these estimates realistic? The quote from Musk citing traffic is

"that the spacecraft is being designed with the plan of flying it for an average of three flights per day, each carrying over 100 tons of payload per flight, for a total of more than 1,000 flights per year, per vehicle"

But is it true that at this level of use a

"fully configured Starship launch (booster and Starship) will use about a 1000 tonnes of methane in the form of LNG as fuel."

And that a fleet of a thousand rockets would require

150 billion cubic feet of methane/natural gas per day (bcfd) for a fleet of 1000 rockets

And finally, is it true that consumption in the USA was only 82 billion cubic feet of methane/natural gas per day (bcfd) (in 2019)?

Evan Carroll
  • 28,401
  • 42
  • 129
  • 239
  • 5
    None of this makes sense even at a glance. There is not enough suitable space on U.S. soil for the thousands of sites you would need. And no individual spaceship will fly thrice a day. Not even once, and surely not at a sustained rate (weather, tech issues). Musk was high from a mixture of sleep deprivation, psychotropics and his latest account balance, and/or was misquoted. – Peter - Reinstate Monica Oct 27 '21 at 16:35
  • @Peter-ReinstateMonica there is no way your claim about usable land is even remotely correct, but [the other point is seems valid](https://space.stackexchange.com/q/55556/38263) – Evan Carroll Oct 27 '21 at 16:54
  • 2
    Hm. How much accessible but unclaimed and unprotected shoreline facing the proper direction do you think there is? The reasons to choose Cape Canaveral in Florida were clear; how many more sites like that could one build in Florida? 2? 10? It is indeed an off-the-cuff remark but I think hundreds Cape Canaveral equivalents on the south shore, facing East, will be hard to find. – Peter - Reinstate Monica Oct 27 '21 at 17:27
  • 3
    @Peter-ReinstateMonica Starship isn't launching from Florida, but rather from SpaceX's own launch site in Boca Chica, Texas. At any rate, the U.S. has a _ridiculously huge_ amount of shoreline, facing pretty much any direction you want. Easily plenty for thousands of launch sites. KSC/Cape alone has dozens of launch sites already. – reirab Oct 27 '21 at 22:57
  • 2
    @Peter-ReinstateMonica To provide a bit more quantification of 'ridiculously huge,' the USA has [12,383 miles of international coastline](https://oceanservice.noaa.gov/facts/shorelength.html), and that's already excluding the almost 5,000 miles from the Great Lakes. (Also, that figure is already accounting for smoothing out the coastline to where the international borders are. The actual coastline itself is [95,471 miles!](https://oceanservice.noaa.gov/facts/shorelength.html)) – reirab Oct 27 '21 at 23:13
  • 1
    I live in Texas. I'm from Houston. For just a point of reference it's a 6-7 hour drive along the coast line to get from Houston to South Padre (near the Mexico border). Other than Corpus Christi I doubt there is a city with more than 200,000 people en route. It's a nice rural drive. – Evan Carroll Oct 27 '21 at 23:50
  • @reirab For the test flights, they launch from Boca Chica because it saves them a 1000 mile ride.; but for actual flights the launch will depend on the desired orbit or trajectory. [Vandenberg is usually used for polar orbits](https://spaceflightnow.com/2020/08/31/spacex-launches-first-polar-orbit-mission-from-florida-in-decades/). And you cannot launch northerly or westerly from Texas. That said, you are right that there is a lot of coastline, and even if the devil is in the details (logistics, preserves, terrain, reservations) one could probably find enough sites. – Peter - Reinstate Monica Oct 28 '21 at 05:44
  • 1
    @Peter-ReinstateMonica Yes, I know that Vanderberg is normally used for polar orbits, but Boca Chica isn't just planned for testing. It's planned to be the primary operational launch site for Starship/Super Heavy (as opposed to Falcon, which mostly launches from KSC/Cape, except for missions that need to be from Vandenburg.) – reirab Oct 28 '21 at 15:03
  • @reirab The U.S. is big, but 3000 starts is a lot. How many starts can one site handle per day? You probably cannot launch in parallel on neighboring pads (SpaceX had that problem in Vandenberg in the early days); the site needs to handle fuel logistics, stack vehicles, handle closures around launch (as you know, the road and beach by Boca Chica is blocked around the take-off) etc. Let's optimistically assume that a launch site with a handful of pads, somewhat spaced out to mitigate cross interference, can handle 3 starts a day. Then you need 1000 sites. That's one site every 12 miles. – Peter - Reinstate Monica Oct 28 '21 at 16:26
  • @reirab Now Cape Canaveral covers about 12 miles of coastline. You see where this leads ;-). OK, CC is big -- maybe we can do with 6 miles. [*That's half the U.S. coastline.*](https://sgp.fas.org/crs/misc/RS21729.pdf) Even at second thought and with an appreciation of the size of the country it's absurd. – Peter - Reinstate Monica Oct 28 '21 at 16:29
  • 1
    @Peter-ReinstateMonica KSC/CC has 40+ launch sites in that 12 miles of coastline, not just one. (Though I do agree that 3,000 daily launches is indeed quite a lot.) – reirab Oct 28 '21 at 16:35

2 Answers2

42

Source Confusion from TechCrunch+

The article by the Energy Consulting Group cites an article by Darrell Etherington at TechCrunch which is not an interview with Musk but instead a badly summarized series of Tweets. The Energy Consulting Group presumes the following

(1 000 rockets) x (3 launches per day) x (1 000 tonnes) = 3 000 000 tonnes/day

This follows from this statement in their TechCrunch source,

total of mor [SIC] than 1,000 flights per year, per vehicle.

This claim in the TechCrunch+ article blurs the 2050 goal and the current goal.

In Tweets from Nov 7, 2019 Musk says

Payload to orbit per year of Starship fleet is most mind-blowing metric, as it’s designed to fly 3X per day, which is ~1000X per year [...] If we build as many Starships as Falcons, so ~100 vehicles & each does 100 tons to orbit, that’s a capacity of 10 million tons of payload to orbit per year

That does NOT say that 1,000 Starships will fly be flying three times a day. That says 100 Starships will fly three times a day, for 300 flights a day, every day where each of the 100 Starships fly 1000 times a year.

You can see this series of Tweets from Jan 16, 2020

Megatons per year to orbit are needed for life to become multiplanetary Starship design goal is 3 flights/day avg rate, so ~1000 flights/year at >100 tons/flight, so every 10 ships yield 1 megaton per year to orbit Building 100 Starships/year gets to 1000 in 10 years or 100 megatons/year or maybe around 100k people per Earth-Mars orbital sync

So where did the actual 1,000 rockets come from? It seems one of two places,

The actual statements of Musk seem like they would incline one to believe that the immediate plan calls for

  • 1,000 Starships that will depart Earth every 26 months to ferry cargo and passengers to Mars,
  • In addition every year there will be 100,000 lifts of cargo on the Starships, performed by 100 rockets lifting off ~3 times a day for a year carrying 100 tons for a yearly capacity of 10 M tons to orbit.

This is a launch cadence of

  • 1,000 launches/2yr (every two years for the main trip) + 100,000 launches/yr (lifting 10M Tons)
  • = 500 launches/yr + 100,000 launches/yr
  • = 100,500 launches/yr

This 100,500 launches/yr will produce drastically different math from 3,000 launches/day (1,095,000 launches/yr) that the Energy Consulting Group assumes.

Launches per day = (100,500 / 365)
                 = 275

btfd = 275 launches * 1000 (tons of LNG) / 1,000,000 (million) * 48.7 (conversion million-metric-tons-LNG to billion-cubic-feet-NG)
btfd = 13.4

So SpaceX will consume 13.4 btfd not the 150 btfd the Energy Consulting Group calculated in their analysis.

As for the 100 MT (1,000 launches/day) goal, this Musk says is the 2050 goal. It's certainly not weighed properly in light of all the other claims Musk has made about the 10 megaton/year goal. In the context of "100k people per Earth-Mars orbital sync" that's no where to be seen (we remain at 0), and SpaceX would have to have a polished Starship production line to produce a Starship every three days (even once never mind for 10 years without stall). Not to mention I don't believe they've reused a single orbital launch craft twice in a day (nevermind a Starship three times in a day). They're no where near achieving either of these goals. They're still making test rockets and the last test was over five months ago.

This claim checks out as being Musk's 30 year "goal". But the more realistic projection is for 100,500 launches/year which he talks far more about, and he's still no where near reaching that. Moreover, if you're going to cite the 2050 goal then the title "SpaceX's Starship May Nearly Triple US Natural Gas Demand" seems to be of poor form as you're comparing the current consumption of LNG to the projected consumption of one company in thirty years.

Evan Carroll
  • 28,401
  • 42
  • 129
  • 239
  • 3
    Based on this answer, it seems your actual question is "does SpaceX have a goal of 3,000 Starship launches per day?" – LShaver Oct 26 '21 at 13:47
  • 2
    Comparing to current consumption doesn't seem *that* unfair, given that ideally consumption will ramp down to reach carbon emissions goals (though it might also switch to hydrogen production via pyrolysis instead of burning the methane) – llama Oct 26 '21 at 17:10
  • 1
    This conversation on whether Mars is even worth the effort has been [moved to chat](https://chat.stackexchange.com/rooms/130883/discussion-on-answer-by-evan-carroll-will-spacexs-fleet-of-rockets-triple-the-e). Feel free to add to it there, and leave comments that try to improve the post here. –  Oct 26 '21 at 19:54
  • Do you have a comparision of the 10 million tons payload per year to orbit to current payload to orbit numbers? I have the impression that is something like 10.000 or 100.000 more than what is achieved today. – quarague Oct 27 '21 at 12:00
  • 1
    It may be worth adding that even at 100, instead of 1,000, the estimates would still be an almost 20% increase in demand. Not 3 fold, but significant. – TCooper Oct 27 '21 at 15:06
12

Will a Starship launch require 1000 tonnes of LNG?

Yes, this is a reasonable estimate. Wikipedia lists 800 tonnes just for the booster, citing an August 2021 interview with Musk. An additional 200 tonnes for the Starship itself doesn't seem unreasonable.

Does a fleet require 150 bcfd?

Yes, this is a reasonable estimate, if we assume 3,000 launches per day is the goal; see Evan's answer for an investigation of that specific claim.

(1 000 rockets) x (3 launches per day) x (1 000 tonnes) = 3 000 000 tonnes/day

3 million tonnes of LNG converts to 146.1 billion cubic feet per day (bcfd). Natural gas is about 95% methane, so rounding up to 150 bcfd is reasonable.

Was U.S. consumption only 82 bcfd in 2019?

Yes, this is fairly accurate. Per the U.S. Energy Information Agency, consumption in 2019 was 85 bcfd. Consumption in 2020 and 2021 is expected to be around 83 bcfd. Production is slightly higher at 93 bcfd, with the difference going to LNG exports.

Obviously, none of this accounts for whether SpaceX could actually scale up to 3,000 Starship launches per day, if there'd actually be demand for that, and if the U.S. could produce that much natural gas. There's also the bottleneck in liquefying natural gas -- current U.S. capacity is only about 10 bcfd. Expanding this capacity costs a lot -- the recently commissioned Freeport facility in Houston cost $13.5 billion dollars, took over five years to build, and only liquefies about 2 bcfd.

LShaver
  • 9,931
  • 6
  • 41
  • 73
  • 5
    Note that SpaceX plans to produce methane chemically from CO2 and water, not extract it from the ground or produce it biologically. They have already installed an air separation unit at Starbase, to separate air into oxygen, nitrogen, and CO2. – Jörg W Mittag Oct 26 '21 at 18:11
  • 11
    @JörgWMittag the irony is that doing that requires electric power, and [most electricity in Texas is generated by natural gas](https://www.eia.gov/state/?sid=TX#tabs-4). – LShaver Oct 26 '21 at 18:18
  • 7
    @LShaver It's not outside the realm of possibility that SpaceX, owned by the same person who runs/owns Tesla/SolarCity, might use a different power source (say, solar panel farms and battery-stored power). Give these are projections for a quarter-century from now, it seems fairly reasonable to assume that as much as it does to assume they'll be using natural gas generated elsewhere in the state. – TylerH Oct 26 '21 at 18:55
  • 6
    @TylerH Realistically though they might just build a nuclear reactor. Neither solar or wind are actually pure "green". Green for no CO2 emissions? Yes. Green for not causing any ecological damage? No. – Nelson Oct 27 '21 at 01:13
  • 5
    @Nelson: at some point, you need to accept ecological damage for using energy. Nuclear reactors are far from ecologically clean, for example, from https://fhr.nuc.berkeley.edu/wp-content/uploads/2014/10/05-001-A_Material_input.pdf _"The construction of existing 1970-vintage U.S. nuclear power plants required 40 metric tons (MT) of steel and 90 cubic meters (m3) of concrete per average megawatt of electricity (MW(ave)) generating capacity"_. Producing concrete and steel are hardly "clean" processes. There is a lot of sun in Texas, and Boca Chica is along the coast (likely good for wind) – Flydog57 Oct 27 '21 at 19:13
  • Assuming you think this tiny little blip is ecological damage. – user2617804 Oct 27 '21 at 22:26
  • 1
    @Flydog57 - not really fair to compare 50+ year old tech to 4th- and 5th-gen reactor designs. Especially in light of things like https://www.radiantnuclear.com – warren Nov 01 '21 at 15:26