Rockets of Today

NEW GLENN — USA, 2024?

artist’s conception, with New Shepard

The other American tech-billionaire space startup, usually mentioned in the same breath as SpaceX, is Blue Origin. It’s run by Jeff Bezos, the founder of Amazon who somehow made that company into a wild success story without managing to consistently show a net profit, even in the days when the company was underselling everyone by failing to charge sales taxes. (It still evades a lot of taxes today, via offshore bullshit.) The all time net revenue earned by Amazon is, in fact, far smaller than the billions that Bezos has ended up with in his own pockets. Bezos became the world’s richest man (until Musk surpassed him thanks to extremely overvalued Tesla stock), while Amazon warehouses remained among the toughest places in America to earn $12.50 an hour. With this personal wealth, he has bought such toys as the Washington Post, as well as some neat rockets.

His company developed the suborbital New Shepard for short tourist flights, and finally put people on it (including Bezos) in 2021, just days after Richard Branson took a similar joyride on his SpaceShipTwo. This stubby little one-stage rocket burns hydrogen in a single engine called the BE-3, which they developed in-house. For a long time we had no clear specs on this rocket — not even its exact external dimensions. Its noteworthy feature is that it can do a soft landing for reuse. It doesn’t do these landings anywhere near as smoothly as a Falcon does, but that’s intentional, and even their concept animations for the New Glenn show it hovering and drifting like a manually piloted helicopter just before touching down. This gives them safety margins they wouldn’t have if it were unable to hover. (The Falcon 9 actually can’t hover because its engines can’t throttle low enough, so they were forced to use the “suicide burn” approach. They don’t call it that, of course: their official term is “hover-slam”, in which both words are completely bogus.) As Blue Origin likes to point out, it was the Shepard, not the Falcon, which was the first rocket in the world to return from space to make a successful vertical landing. And one of their Shepard test rockets made five flights long before any Falcon booster could.

That suborbital craft may not be important to us here (unless they add a second stage to it for smallsats, which is something Bezos says they are considering), but the much larger New Glenn is coming along right behind it, and with it they are skipping entirely over the light and medium launch vehicle categories, and most of the heavy category as well, to something that might qualify as a superheavy if it were used expendably. In its basic configuration the Glenn originally was to consist of two stages, both burning methane. Some are calling liquid methane “the fuel of the future” because it’s almost as good as kerosene at sea level and almost as good as hydrogen in vacuum, while being cleaner than the former (kerosene gets soot on everything) and easier to manage than the latter, and cheaper as well. Quite a few new rockets are starting to adopt it.

The first stage uses seven of their big BE-4 engines — quite a contrast to the puny set of two that ULA’s Vulcan would use — and in the original plans, the second would have a single one with an enlarged bell. A third stage would be optional — it would burn hydrogen in a BE-3U (a vacuum version of the Shepard motor), and be capable of escaping from Earth orbit. But they changed their minds, and both the second and third stage will use hydrogen. The two upper stages were originally going to be the same height, but now the second stage will be taller, and have two independently gimballed BE-3U engines to provide sufficient thrust, while the third stage is now apparently going to be small, and ride inside the fairing instead of beneath it — a kick stage satellite platform which will now be dubbed Blue Ring. They say that using hydrogen for the second stage will extend the range of orbits that the rocket can reach without needing a third stage. So they’re going to skip developing a vacuum version of the methane engine.

Actually, the BE-3U differs significantly from its sea-level progenitor: rather than just changing the bell, they are switching the turbopump power cycle from a tap-off system to an open expander cycle, which will require a lot of reworking. In this form, it no longer has the deep throttling capacity of the original BE-3, but it might be easier to restart in space.

The BE-4 methane engine, on the other hand, uses oxygen-rich staged combustion, like an Energomash engine. This makes it more advanced and efficient than most American engines, including SpaceX’s Merlin. By contrast, the BE-3U’s expander cycle design is as old as the hills. Even the venerable Centaur uses a closed expander, and avoids dumping unburned fuel, though the closed cycle does limit it to much lower power and thrust than the “semi-closed” design of the BE-3U, which uses the boiled hydrogen that exhausts from the turbine as a film of cooler gas to protect the inside of the expansion bell.

The upper stage or stages will be expendable, but the first stage will soft-land itself on legs for reuse, just as the Shepard does. Blue Origin points out that with six legs, it will be able to land even if one fails. (This is a dig at SpaceX, which once lost a Falcon booster when a leg failed to latch down.) Only three of the seven engines will gimball, including the center one. They originally hoped to tuck the legs smoothly between the outer engines, but found they had to widen the base of the main stage to about eight meters. At one time they were boasting about setting up the world’s giggest carbon fiber wrapping machine to make stage fuselages with, but now the tanks are made of conventional stir-welded aluminum. The reinforcing grid of interior ridges has to be machined out of a thick sheet — a slow and painstaking process which is apparently a standard practice in the industry. They are looking to switch to a cheaper technique on the expendable upper stage. The only big pieces made of carbon fiber now are the fairing halves and the payload adapter.

They hope to get 25 uses out of each New Glenn booster at first, and 100 later on, with a turnaround of 16 days and no real refurbishment needed. That booster is big — it would be by far the largest nongovernmental rocket ever, if it weren’t for the Starship being developed at the same time. Overall, the New Glenn might amount to about half a Saturn V. The booster appears to be roughly the equivalent of the three boosters of the Falcon Heavy combined into one, and the second stage is many times bigger than the Falcon’s. Despite this size, they do plan to land the booster at sea, just as SpaceX does. The landing boat they originally designed was a converted freighter with a widened deck, rather than the simple barge SpaceX uses. They say they’d be able to land on it while it’s moving, and handle rougher seas than a barge can, thanks to some sort of active stabilization thrusters on it. But then they dropped that and bought a barge.

Note that the sixty ton payload commonly quoted for the Falcon Heavy is for expendable mode, but Blue Origin’s forty-five ton figure for the Glenn is with reuse, so though it may sound like it has less capacity than the Falcon Heavy, it actually has more... if, that is, they actually meet that 45 ton target, which they recently admitted is something they only hope a future enhanced version might do, rather than being what they expect from the initial version. The booster thrust is not extreme, so I figure the capacity advantage over the Falcon Heavy is probably in large part because of the much bigger second stage. Even if it falls short, the New Glenn will be the world’s second most powerful commercial rocket, after Starship, assuming both get off the ground. And for higher orbits, the advantage over the Falcon Heavy just gets bigger — between the larger second stage, the use of hydrogen, and the option of adding a third stage, the capacity of the New Glenn to send stuff to the moon and beyond might be double what the Falcon Heavy can do. Also, the fairing capacity dwarfs that of the Falcon, offering about four times as much room inside. Of course, you’ll also probably pay a lot more for all that expendable stage capacity. And again, this all sounds awesome compared to older rockets, yet the actual performance they achieve might turn out to be substantially less, and even if that isn’t the case, the Starship might make it obsolete.

But Bezos has a plan to fight back. For a later version of the New Glenn, they will try to build a reusable second stage. The plan is apparently to use the same approach pioneered by SpaceX, and make that stage out of stainless steel covered by a thin heat shield. Without reuse, the large size of their expendable upper stage will probably make it impossible to compete with SpaceX on price, but with it, they’ll be in a much better position. This attempt at a reusable second stage is called Project Jarvis, and apparently they have taken serious measures to insulate it from the company’s normal management practices... which gives you some idea of how well-run the rest of the company is.

lunar landers

And speaking of the Moon, in 2019 Bezos announced a lunar lander called “Blue Moon”, which is essentially a rocket platform that any desired payload can be bolted to the top of, with davits to lower rovers and such to the lunar surface. Yes, he wants to offer moon landings as a commercial service! The lander weighs fifteen tons with fuel, and has four legs which will fold up to fit into the seven meter New Glenn fairing. It has exposed spherical tanks... which led Elon Musk to mockingly nickname it “Blue Balls”. The capacity is 3.6 tons of cargo, and they later proposed a stretched version for human landings, raising the capacity to at least 6.5 tons. Any capacity to return to orbit will be left to the payload to implement for itself. The basic lander will use a single engine: a new hydrogen-burner called the BE-7, which is quite a bit smaller than the BE-3U. It runs on a dual expander cycle, is restartable, and supports deep throttling. The human version would have two or three engines — it’s not clear which.

This stretched lander was part of the “National Team” proposal for a human landing system for the Artemis program. This would have carried a crew capsule and ascent stage built by Lockheed, with its interior and electronics based on the Orion capsule. It was regarded as the front runner in the competition, despite how neither of the other proposals it was up against (the Dynetics ALPACA and the Lunar Starship) would leave big expendable stages behind on the lunar surface. But then Congress cut the budget and only SpaceX bid low enough. The other two both lodged formal protests. Congresspeople started arguing to raise the budget again to restore competition, and others (notably Bernie Sanders) decried this as a “Bezos bailout”. In response, Bezos offered to defray up to $2 billion of the cost himself, if his lander was chosen — a controversial move that blurs the line between charity and bribery. Blue Origin also formally protested Nasa’s selection (as did Dynetics, the lead company of the other losing team), but the Government Accounting Office rejected the appeal. Bezos then sued, because he is an asshole. Once that went nowhere, the National Team seemed to break up, with several individual members starting pursuing other designs.

But some of them, particularly Lockheed, got back together for another try, and since they now had a much less hurried schedule, as they were now making a proposal for Artemis V and beyond instead of for Artemis III, they could now come up with something a lot less half-assed than the mostly-expendable National Team proposal. And they did. This time, the crew cabin is at the bottom of the stack, below the fuel tanks and right next to the rocket nozzles, instead of on top of a tall ladder. And this time the whole thing is reusable, as a single piece with no detaching modules. And it won the contract. So officially NASA will only use the Lunar Starship for two flights, but of course we know that once it’s available they’ll find plenty of additional use for it. Having a choice of small or large landers will be a good thing... assuming you’re okay with the idea of the government spending all those billions to attempt to exploit the moon.

Bezos’ plans

But back to the New Glenn. The question that’s begged by the size of it is, of course, who other than NASA will have a use for that much rocket. The Falcon Heavy has found far fewer customers than it was expected to when it was first planned, and the trend lately across the industry has been toward smaller and smaller satellites. Bezos has said that all of their future rockets are going to be even bigger than the New Glenn, but they may have to change their minds about that. Unless NASA revises its current plans and decides it doesn’t want the SLS but does want the Lunar Orbital Platform-Gateway, I don’t know who out there would need such a big rocket. Maybe they’re just hoping that if really big satellites are cheaper to launch, someone will come up with a use for one. We could imagine, for instance, super-powerful communications satellites that enable cell phones to work without towers... but I don’t think that’s likely. Sending things to the lunar surface might be almost the only mission where the New Glenn is the most cost-effective choice.

One way they’ll get some use for that capacity is to stack two or more full-sized sats into a single fairing, like the Ariane V does. The additional juice in the second stage should be able to put them into substantially different orbits, if they’re not huge. This would cut the launch price to half or less for cases where the logistics work out for all customers, and might go a long way toward making it competitive in cost. But for any load that doesn’t make use of that huge capacity (which aside from Starship will be not only the heaviest payload capability commercially available, but by far the largest fairing volume as well), I bet a Falcon will always be able to underprice it because less is thrown away, so I don’t know how the Glenn will manage to get a lot of business, unless they eventually get that reusable second stage working.

Of course they could, like SpaceX, make their own work by putting up satellite swarms. Amazon does indeed have a swarm in the works... and will launch the first nine batches on Atlases or Vulcans as the New Glenn is still far from ready. They have also announced a partnership with Sierra Nevada to launch their proposed “Orbital Reef” space station, which would need a Glenn as each module is huge, but once up it would be mostly serviced by smaller rockets. And in the fall of 2023 they backed away from their Orbital Reef plans, probably to concentrate on finishing the lunar lander for NASA.

They also said they were planning to eventually sell tourist seats on the Glenn. Aside from these options for creating their own markets, there aren’t many customers with payloads over twenty tons. I wouldn’t be surprised if only a modest number of New Glenns end up being built, while cheaper Falcons or mass-produced Starships roll off an assembly line by the dozens... and for the short term, at least, this is explicitly Blue Origin’s plan, because they hope to get so many reuses out of each one. (I was thinking that if this works they could give each one a name, like a ship. And hey, they did paint a name on the first test booster: “So you’re telling me there’s a chance”.) Despite the lack of mass production, I guess it’s possible that the Glenn’s larger scale may eventually make it a cost winner against the Falcon, particularly against the Heavy. So far, though, it sounds like Blue Origin has a long way to go before they can bring costs down that far, and SpaceX seems to be at little risk of losing their lead.

Their factory is at Cape Canaveral, and as of 2021 this factory was still mostly empty and idle. One insider described the whole company as a dysfunctional Potemkin village which puts up a fake appearance of productive industriousness. The company’s performance got so frustrating that Bezos stepped down from running Amazon in order to manage Blue Origin full time. By 2024 the factory was a lot busier. They were supposed to be ready to finally fly the thing that fall, launching a small NASA probe to Mars, but a couple of accidents that damaged some upper stage parts prompted NASA to admit that they were going to miss the window for that (as many suspected was going to be inevitable regardless).

For now Blue Origin is not cutting costs anywhere near as aggressively as SpaceX is, and this is for a very good reason — not just because they’ve embraced a cautious “slow is fast” design philosophy, in contrast to SpaceX’s fail-forward approach. They have a coat of arms which emphasizes this, as it features two tortoises and the motto “Gradatim Ferociter”. Ironically, some other insiders say that the reason Blue Origin stayed stuck in suborbital flight for most of two decades is because Bezos wouldn’t follow his own philosophy, and insisted on trying a large reusable rocket right away instead of solving easier steps first. On the other hand, he did hire old-line aerospace veterans who are accustomed to slowness and caution and minimizing risk, but there ain’t any slow-is-fast in that tradition, only just plain slow.

The bigger reason for not cutting costs aggressively is that while Musk is undercutting the traditional space market for lifting satellites and probes and professional astronauts, Bezos is trying to create a new one for space tourism and space commuting, in an attempt to open up cislunar space as a place for regular people to have jobs. For this game, where their intended market is civilian passengers, safety is paramount, so it makes some sense that Blue Origin is moving at a much more careful pace than SpaceX is. This also fits in with their approach to the BE-4 engine, which is to take an advanced design that’s capable of being pushed to very high performance, but use it at only a medium performance level so it’s nowhere near its limits — this minimizes the safety risk.

But on the other hand, maybe SpaceX is developing safety and reliability in the best way, by practicing lots of real launches. By occasionally losing a Falcon or Starship, they may be gaining more practical safety knowledge than Blue Origin’s low-risk approach can, even if their overall approach can be criticized for sometimes being too hurried to get things right... as shown by for instance the Falcon Heavy, which ended up with four years of delays because their initial plans were naive. Back on the first hand, it’s also possible that SpaceX’s performance goals for their future spacecraft are too aggressive and they are cutting safety margins too thin — a bad habit from their early days when they always had to work exhausting hours and push the deadlines just to stay solvent. It may be that Blue Origin will be able to sustain more safety and reliability over the long term. That’s what Bezos believes.

Neither company has gotten anyone killed by a rocket yet, but inevitably the day will come when one of them does. Rockets are dangerous, period. It’ll be interesting to see to what extent the public is willing to accept and support this from a private venture. If it’s traditional astronauts that die, I suppose things will resume after a year or two, as has happened in the past, but if it’s tourists, that might mean the business goes bust. Even the loss of an unmanned rocket can be very costly: when a Falcon once blew up on the pad, the direct and indirect loss to the company over the following months was something like $700 million, and repair work on the facilities took more than a year. Virgin Galactic and Scaled Composites, which are focused on the suborbital tourist business, have in fact killed a test pilot, and this may have been a factor in the company then leaning away from plans to develop tourism to the orbital level, and instead pivoting toward conventional launch services — a market where the existing work they’d done on suborbital passenger flight was of little use, and they had to start from scratch... which didn’t work out. And Blue Origin had a near miss when a New Shepard booster engine failed. Luckily for them it happened on a crewless flight, and the capsule escape system worked, so it probably wouldn’t have killed anyone.

In any case, Bezos is betting that as the space business expands and gets more affordable, people are going to want to move heavier loads to orbit, and the future will belong to big rockets. He says “rockets want to be big” — that is, achieving efficient performance becomes easier with size — so they only plan to build even larger stuff in the future. If he’s right, the traditional space companies are going to be out of the game for years. If New Glenn can fly at a price competitive with the Falcon Heavy, or turns out to be safer, and works better than the Starship, it could be the dominant space vehicle of the next decade. But if he’s wrong... whoever makes the smallest reusable rocket might be the one who really cleans up. There are a lot of competitors in the small-sat market now, and none of them have figured out reuse... whoever does might end up doing the majority of the world’s launches.

New Armstrong

Where next for Blue Origin? After New Glenn, they say they’re going to try a New Armstrong, to scale up their capacity for lunar and interplanetary missions. But they might need to also scale down — they might find a lot more business below five tons than above fifty tons.

I note that even though the Shepard is flying today, we still don’t have exact specs for it. Blue Origin has in general been much less public about its doings than SpaceX has. But one thing they have said is that their long term plan is to build some kind of base on the lunar surface, in the next decade. Their vision over the even longer term is to enable a future where there might be a million people living and working in orbit or on the moon. Bezos has also shown concept art for what appears to be a luxury space habitat for rich people. That actually looks like a more likely goal for the shorter term. He has not shown any particular interest in going to Mars, which Musk wants to colonize. I guess he figures that people will get there eventually, once it’s no longer so difficult.

So maybe in Bezos’s mind, it doesn’t matter that there won’t be a lot of customers for launching forty ton satellites. Maybe, like SpaceX with their StarLink system, if nobody else wants that much launch capacity, he’ll find a use for it himself. If he envisions thousands of people living and working in orbit, it’ll require hundreds of heavy launches to put the habitats up. Maybe he’s eventually planning on building a private city up there. And in the shorter term, orbital space tourism will probably do a fairly good business... the Glenn could hoist a much larger capsule than any current rivals, able to accomodate rich people in comparatively luxurious conditions, or take twenty or thirty people on a shorter flight.

On a more mundane level, they say that besides providing BE-4 engines to ULA for the Vulcan, they will gladly sell their motors to anyone else who’s interested. They could end up competing with engine builders such as Energomash and Aerojet Rocketdyne as much as they do with ULA and SpaceX.

New Glenn (two stage): mass unstated (1500 t?), diam 7 m (8.5 at base), thrust 17100 kN, imp ~3.4 km/s?, staged combustion (methane), payload hopefully 45 t (3%?) with reuse, cost unknown.