New Glenn isn't the future

Its the last gasp of a past we should be happy to be leaving behind

I would be very happy if New Glenn was massively successful, because that can only happen if it reduces launch costs significantly over Falcon 9, and lowering launch costs further is the key to massively increasing utilization, exploration and eventual colonization of space and all the abundant resources in the inner solar system.

But it ain’t gonna happen. New Glenn’s best outcome is to offer Falcon Heavy payloads at similar prices, it’s got no hope of matching Falcon 9 prices. And if the first version of this overweight, over-built behemoth can only put 25 tons into orbit as is rumored, that’s the same as an expendable Falcon 9 maybe half its size.

New Glenn’s design is significantly better than the new Ariane 6 and Vulcan because it doesn’t require expensive solid rocket boosters. But it’s significantly worse than Falcon 9 because it depends on two different engines and two different fuels. This increases engine costs, pad costs, and delays.

Engine costs go up because you need two sets of tooling and two production lines. Falcon 9 has a single production line with single tooling investment that gets amortized over more engines. This has been estimated to reduce the cost of each Merlin to under $500k, and possibly as low as $250k. Blue Origin is selling the New Glenn BE-4 engines to ULA for Vulcan at an estimated $7M each, which implies their cost is at least $3M each.

Pad costs go up managing two fuels with significantly different temperature requirements. Delays increase because hydrogen is such a leaky propellant with its tiniest of all molecules. So it’s very unlikely that New Glenn can even be cost competitive with Falcon Heavy, despite the Falcon Heavy’s somewhat jury rigged design.

So if we were to label rocket generations like military types label fighter generations, I’d call Atlas/Delta 2nd generation, Ariane/Vulcan 3rd gen, New Glenn a 3.5 gen design, Falcon 9/Heavy 4th gen, and Starship 5th gen.

When Falcon 9 first flew in 2010, it was the blue print for 4th gen launchers. It didn’t break new ground in rocket design, its advances were where progress had traditionally been slowest, in rocket construction. Its design was basic and cost efficient. It used a single dense fuel, lots of cheaper, smaller engines, and relied on a cheap second stage built using a single first stage engine. These decisions meant its absolute performance broke little ground, in fact the weaker performance of its cheap second stage meant that its high payload capacity to low earth orbit degraded rapidly when directed to higher orbits.

But SpaceX leveraged those mass number of engines to innovate from the industry’s traditional bespoke, hand built, super expensive but quirky individual engines to mass produced, super cheap, consistent, high-quality engines. The fact that it used nine smaller first stage engines also meant a single engine could throttle low enough to propulsively land an empty first stage with a mass not much different than an empty beer can, so that reusing first stages was demonstrated within five years and mastered within seven.

And still with this very public demonstration of the advantages of the fourth generation design in 2010 with the $63M launch cost and final proof of its superiority 2015-2017, only one company since has copied the formula, and that’s another startup new space company, RocketLab. ULA, Arianespace and even twenty year old Blue Origin have ignored the F9s massive success and steadily plodded on honing now obsolete third generation designs.

And history continues to repeat itself as Starship demonstrates step-by-step the enormous advantages of its fifth generation design. A rocket so large even with the payload reduction re-use requires still offers enormous payloads. No longer throwing away a $20M second stage means that its launches should cost less than a Falcon 9 while lifting more than four times the payload. Landing back on the tower means rapid turnaround, and potentially reusing the same hardware daily. If it works, in less than 20 years we will have transitioned from 50 years of ultra expensive bespoke hand-built rockets thrown away every launch to a fully reusable space launch system far closer to commercial airliners in cadence and cost.

Just to put a number to how huge this revolution is and is likely to be, here are rough estimates of the cost (or price) per ton of payload to low earth orbit for commercial launchers of this century. For comparison the Shuttle was about $80M/ton in present day dollars and SLS is $20M - $40M/ton depending on who you believe, NASA or CBO.

Starship should be roughly $100,000/ton, which is why its nearly at zero line in a chart of millions per ton.

Meanwhile the dying former market leaders sit and do nothing, waiting to see if Starship sweeps through the market in the next few years and obsoletes everything they are working on yet again.