Rocket Lab goes big with Neutron

[Note added engine discussion and market fallout at end of this article 3/19].

The launch vehicle global scene got a lot more interesting with recent SPAC's. First Astra decided to fund through public markets it's "portable" microlaunch vehicle, and now Rocket Lab (RL) likewise goes big with an Antares clone done on a shoestring budget/schedule ($200 million in 3 years). Welcome to the new normal.

In some ways this isn't much of a surprise that we'd see moves like these as the "glut" of launch capacity and the tumble of launch pricing and futures. But still when it hits, especially with already successful firms like RL , it's a shocker as to how fast and hard this change will hit.

While successful, RL's Electron hasn't been much of a "moneymaker" - they only made $40 million in 2020 revenue. Largely this is due to not being able to get to weekly or better launches like they planned. They even increased to three pads at two locations, and recovered a booster so as to find ways to up launch cadence, but apparently this isn't enough.

Electron is squeezed from the bottom by Astra and Virgin Orbit, and from the top by SpaceX's Transporter bulk launches plus polar from Florida launches, and eventually Momentus allowing orbital bulk launch to hit a larger plurality of orbits. There are many attempting to "clone" Falcon 9 on the horizon in Europe and Asia too, with StarShip SuperHeavy (SS/SH) also threatening orbit this year. And a dozen others also in the pipeline.

Note that Neutron is a close match for Antares, whose launch facilities at Wallops are right next to Electron launch facilities (note - Peter Beck since confirmed the intended shared use of Northrop Grumman's facilities). Size, payload and propellants are the same. The two longest lead time items for launch vehicle development are engine development and launch facilities. RL has already demonstrated fast, low-cost, and effective novel LRE development. Using Northrop Grumman's highly underused facilities (1-2 per year) might allow them to achieve their goal of human spaceflight vehicles to the ISS in that time frame, with an avenue to someone who holds an open ISS cargo contract to the ISS already.

Antares only launches to the ISS. It was funded alongside Falcon 9 as a means to supplant the obsolete Delta II launch vehicle, which is exactly what Version 1.0 of the Falcon 9 actually did - to the ISS and many other spacecraft/destinations as well. It later was up-scaled and made reusable to out-compete Proton, Ariane 5, Atlas V, and likely will soon put Delta IV-H out of business. Which is why only a lucrative ISS contract keeps it in business.

Perhaps RL can grow that business broadly, while offering a launch cost less than the expended Russian RD-181 engines of the Antares? Even go "eyeball-to-eyeball" with Falcon 9 ISS/DOD contracts (which Antares can't due to sourcing), before even SS/SH can muscle in on human rated spaceflight for less? Because even if SS/SH can launch more cheaply, ISS, NASA, and DOD "needs" still have a while to go before they can grow to encompass that grand of use that SS provides.

Noticeably absent from Neutron discussion is much involving the Moon and Mars, two prime candidates for high mass payloads that SS/SH is near term focused on. And where Falcon, Neutron, and SS/SH all share a focus is on large satellite constellation deployment. Perhaps an RL-developed electric pumped kerosene LRE could escape coking that limits Falcon cadence, and get close enough to SS/SH economics to attract non-Starlink constellations as a viable alternative?

How does this change the launch landscape? It might not just undercut Falcon 9 but all other global "clones" under development like those in Europe and Asia by beating them to market and denying them competitive access to customers. But the real impact is to arsenal system legacy launch providers who have been struggling with Falcon 9 as well, by limiting them to institutional launches that are insufficient to maintain their industrial base. Vega, Soyuz, and Angara are already on hard times. Late to market Ariane 6 was expected to be short lived, perhaps it's follow-on also might face the same, to the misfortune of European's footing multiple huge launch vehicle and pad development expenses. Even India's PSLV and GSLV will be strongly affected by this. Unflown KSLV-2 has been on the edge of cancellation with Falcon 9 economics causing them to rethink program costs, might this also be a victim? And other marginal launch systems might also face the crunch as they still are yet to deliver on the difficult challenge of making it to orbit. The overfunding of launch systems is about to reach a moment of truth, as the glut of capacity makes many financial projections fly out the window.

Vulcan and New Glenn are least of these to be affected, because of their respective unique status. However it is likely that Vulcan's scope of payloads will be greatly reduced and narrowly specialized, making for a lean business future. New Glenn will slow (it's already glacial) as it's scope of competition will be narrowed to uncommon payloads that will require longer time on pad and more careful attention to complexity for attracting a necessary launch success string.

So its understandable that Peter Beck's pivot of RL to survive is sensibly timed for them. As well as untimely for all others. Except, of course SpaceX. Who has even more motivation to field SS/SH, as the only fully reusable vehicle allows them to dominate where launch futures will lead.

SpaceX looks to finally get its first real competitor for Falcon 9 with RL. Who ever would have guessed it would come from New Zealand?

Added:

Engine development seems to be much discussed. Here's some thoughts to spur that discussion.

Electron's novelty is in two things - a carbon fiber tankage / thrust structure, and it's lithium ion battery powered propellant pumps in the Rutherford engine. Like SpaceX, RL is passing on the carbon fiber exterior, perhaps due to reuse of the thermal cycles of extreme hot/cold.

But the real difficultly isn't the materials but the engine. Rutherford was relatively quick to develop because of its unique nature, and to field a rocket rapidly and cost effectively, you'd want something like it to follow in the non-traditional footsteps of Rocket Lab engines. The electric pump in Rutherford takes the place of the gas generator in open cycle rocket engines.

This isn't for free, and as Elon Musk points out, that an electrically pumped large scale engine would have too great an energy demand to simply scale up Rutherford. Perhaps there might be a better way than that?

Closed cycle rocket engines have a "rocket engine in a rocket engine", using a preburner to drive a turbine pump to leverage the effectiveness of the preburner's combustion, which itself is used to create a more effective gas-gas injection to the main combustion chamber of the engine.

What if 1-2 derivatives of Rutherford ... were to be the preburner(s) of such a closed cycle (ORSC or FFSC) engine? Then the battery demand is a fraction, and the coking problem faced with reusable kerosene rocket engines like SpaceX's Merlin 1D might not be present.

Elon Musk talks about the cleaning of the Merlin gas generator as limiting flight rate. What if RL ... doesn't use a gas generator? Perhaps ... higher flight rate?

This approach is not without development challenges - you'd need high temperature alloys/coatings to stand up to extreme highly reactive hot oxygen, high pressure environments, as well as turbopumps that could stand Rutherford engine output. SpaceX FFSC Raptor still has teething problems in landing StarShip. And even with Blue Origin's massive capital, getting the ORSC BE-4 engine to full thrust, full duration test firings for ULA's Vulcan have gone half a decade longer than planned.

But at least this would be an approach that both leverages RL's experience and resources, rather than starting from scratch, and also allows enhancements to Rutherford development to improve RL's primary revenue stream, which could use some shoring up.

Startups always have to think like this, or they go out of business. And the capital demands of your second product can always dominate, especially in development. RL doesn't need a "second system syndrome" right now.

Added:

Asked "how quickly to develop".

Gas generator or tap-off "open cycle" engines are fairly quick to develop, because you build one part, and add more to it til you get to a useful system. First Merlin took 4-5 years, and first launch was a year following.

Closed cycle engines take at least a decade. The increased difficultly is that everything needs to work in a system from the beginning, because of nested interdependence. So you have a lot of expensive and time-consuming "booms"and unstable operation, much in the same way squishing a water balloon in your hands tends to escape through your fingers.

Now, here's another advantage of using a electrically pumped engine as a ORSC preburner. You might be able to save time by varying the combustion parameters with software altering the preburners function such that you can handle start-up and shutdown before the test engine goes out of the operational "box" of parameters, because the propellants for the preburner can be quenched fast (motors stop).

It's not perfect, because the large volume turbopump the preburner is spinning, which is the "oxygen rich" part of ORSC, still lags due to mass flow momentum and is where the damage gets done. However, in FFSC, you increase complexity but half the flow/pressure by having two less critical sides of fuel and oxidizer, while still increasing combustion chamber pressure above what a closed cycle engine can do, which is what SpaceX's Raptor does.

There's also a nonlinear effect as the scale of the engine grows. In scaling up BE-4 for ULA, Blue Origin extended development time significantly over it's initial plans - more "booms". One can compensate with "hardware rich" if one has either Jeff Bezos wealth or Elon Musk's ability to throw together an impromptu launch vehicle assembly line in South Texas, but not so much for frugal RL. Whose objective is about a half scale Falcon 9 (or Falcon 5) that's even more pinch penny than SpaceX.

So it's in the trades of complexity, risk, and scale that the choices here get made. Which is likely where RL is spending its time at the moment.

Added:

Asked about market landscape in 2024.

As mentioned at the beginning of this article, the "glut" of capacity grows to become unsupportable for some. Why? Because the cynicism becomes mere fatalism, where you really should take seriously the "fatal" part.

Frankly all the usual launch providers ceded the future around 2011, because they did not accept inconvenient changes. And mostly kept with that for a decade. Now we get to see the consequences of those decisions.

The biggest consequences come from the smallest launch providers (not the most revenue though). Astra still doesn't have a successful orbital launch under the belt, and have a unpredictable management team trying to fill the ranks with talent which perhaps they might hold onto with SPAC monies. The threat is launch from anywhere with four containerized cargo containers and six people, meaning no costly unscalable fixed GSE/pads that RL depends on. That and much lower cost vehicles that look like 500x .22 caliber bullets potentially undercuts RL's Electron to the degree it may become unprofitable by stealing launches, which heads off reuse as a launch cadence "improver" because you can't make up the difference on volume.

What most miss is that RL has done this to other vehicles globally as well. And, in the near term RL's business will ramp deceptively because many customers will want to gain experience in advance of the perceived eventual "microlaunch glut" of Astra and others to come. Also, because such launches are almost as cheap as access to effective space simulators, many will use smallsats to evaluate subsystems/sensors/quickturn missions rather than aggregate/postpone, "flying first". Perhaps this is "spacecraft agile development"?

It robs from the medium capacity in the same way geosat's got to be rare, and LEO constellations have begun to dominate. If everyone goes the smallsat first route, anything bigger waits longer and needs to be cheaper because there are many starved LV "mouths to feed".

Which brings us to all those expensive national launchers who use commercial launch to fill out manifests. While they can discount/aggregate, at what point are they still too expensive/tardy for the now lowered "low end" to starve them completely?

That's Neutron's primary manifest.

If they were to undercut/overcadence Russian/Chinese/Indian/other launch vehicles that absorb that, the national launchers have no cadence but their own institutional. China can afford this indefinitely, as well as eventually building something to compete with SS/SH, but their agenda is becoming quite clogged, because of the penchant for micromanaging in order to not lose face on mission/business/market/development failures, so they'll be outmaneuvered perpetually (nothing like a cranky Chinese aerospace manager).

Russia and Europe are too arrogant to relax their grip on the past to adapt to a new future, and both have overspent on the past to have the capital to allocate to the "more than one rocket but five families" needed. Note that no more Soyuz 2 LV's are planned, and Angara's automation is still unsuccessful. Ariane 6 stretches further into the future, Ariane NG joins its predecessor in not making it to first launch before becoming obsolete - wheres Ariane NNG when you need it? China adds more and more LV developments with fewer launches spread over more of them, with less oversight. ISRO also has too broad a range of vehicles and slows down as well for same. Many others face this as the cost for indigenous development doesn't keep pace with global competition.

Not everyone can launch on SS/SH. So who do you turn to when you can't? Many are halfheartedly attempting a "Falcon clone" but are taking "baby steps" at reuse. The only one other than SpaceX to recover a potentially reusable orbital launch vehicle is RL, but they aren't rushing like SpaceX did to do so. Perhaps because a small composite vehicle isn't ideal for that?

But perhaps one that is smaller than SS/SH might be the ideal "little rocket" for "everything else"? Because no one, not even Blue Origin, dares a SS/SH in the foreseeable future - Blue can't even field a New Glenn structural test vehicle for fit checks. So beat the "Falcon clones" to market with an idealized "big enough" (meaning traditional payload size or Delta II/Soyuz 2/Antares/CZ-4/GSLV one). Falcon 9/Heavy payloads were to compete with EELV/NSSLV sizing - that top end well eventually go to Vulcan/New Glenn/SSSH and be unreachable. That's the market opportunity - big enough to matter but not in the "big guy" space.

China will still compete (slowly) with slow Blue Origin, but SpaceX will dominate. National launch will compromise with selective, minimal "secure" payloads that enable larger "unsecure" payloads launched more economically (constellations, bulky subsystems, and propellant). The sideshow of "point to point" suborbital logistics will take a long time to work out, as will uneven reliance on large scale payloads of +7M and/or +30T. HSF is too early to call, and too irrational a market to predict, although its likely that too infrequently flown vehicles won't be able to maintain enough reliability to support HSF use.

This will likely mean that Atlas V will continue to fly alongside Vulcan, which isn't HSF rated. Could Neutron inherit the "non SpaceX" HSF business, simply due to marginalization of ULA as a launch provider? Who eventually succumbs to the "death of 1,000 cuts" simply because of rarefied manifest? Perhaps among national launch systems it may be the "last rat standing"? (James Bond in "Skyfall"). Too soon to know. But cost growth on national launchers is already underway.

None of this is really new. Providers knew this. It's just this change took them from their "safe space" into what RL is currently experiencing, and they didn't want to go there.

Thanks for the requests to extend this article. Feel free to request more.