Wednesday, September 28, 2016

SpaceX: The Mars Plan

I've had a thousand thoughts bouncing around in my head since Elon Musk's Mars keynote yesterday. It's a lot to take in, and I was trying to organize my thoughts. However Jonathan Goff at Selenian Boondocks and Eric Berger at Ars Techica have offered some good takes already. You should read them.

Selenian Bookdocks: Jon's First Take
Ars Technica: Musk's Mars Moment

Like Eric and Jonathan I'm excited by the possibilities and I'm a lot more enthusiastic about this architecture than anything NASA is working on, but there's also quite a few things that worry me.

Part 1: Non-Mars Stuff

Skip this if you only want to talk about Mars.

Falcon 9: Even More Thrust?
Just sort of off-hand, Elon mentioned that the bulk of SpaceX engineering right now is focused on a "final" version of the Falcon 9 due next year. I previously had no idea there would be any more versions of the Falcon 9. I thought the Full Thrust model was going to be the last one. What could they possibly be adding? I really doubt they'll move the Falcon 9 to the Raptor engine (and methane fuel), as that would require a redesign from the ground up. That would be a whole new rocket. So what are they adding to Falcon 9? I have no idea.

Part 2: Concerns

As Jon summarized, SpaceX is pursuing both world-beating performance per launch and reusability at the same time. Just to give you an idea of how much performance we are talking about, here are the thrust/weight ratios of some well known rocket engines:

Engine    Vehicle   T/w ratio
SSME      Shuttle      73
RD-180    Atlas V      78
F-1       Saturn V     94
NK-33     Soyuz       136
Merlin 1D Falcon      180

As you can see, the SpaceX Merlin 1D is already the best rocket engine in the world. Now Elon wants to triple that? At the same time, adding the material reinforcement necessary to withstand 1,000 launches (as they are hoping their BFR will reach) worth of wear and tear only makes your thrust/weight worse. Is tripling what's already the best-in-class number even believable? It's a lot to swallow.

Interplanetary Transport Ship (ITS)
I think the ITS can be thought of (and probably is thought of within SpaceX) as a really, really, really, really big Dragon 3. The ITS tries to be a lot of things. It's a second-stage rocket boosting out of Earth gravity with near-SSTO performance even without its booster. It's an in-space habitat for 100 people (20x the ISS crew complement) for 80 days. It's also a lifting-body decelerator (like the Space Shuttle) that ends up landing on its butt on Mars (or any other solid body in the solar system).

Any one of these things would be amazing. All three of them in the same vehicle reminds me of the F-35 Joint Strike Fighter. A cruise ship that's space-rated and flies?

The only reason to do all these things with one vehicle is because you lack a real infrastructure of more specialized vehicles. As Jonathan also mentioned, a lot of the cost of SpaceX's architecture is that the ITS can only be used once every couple years when the planets align. It's dead weight (economically) the rest of the time. If you broke up the architecture into things that could be used during the other 22 months of the bi-annual cycle, you'd spread your costs out over a lot more missions.

For example, imagine instead of the 100 Mars colonists launching up in the same vehicle they fly to Mars in, you built instead the equivalent of a space-rated 737 sitting on top of a Falcon 9. That could take the colonists to a colonial transport ship which stays in orbit all the time. You board up when it's time to go to Mars, but the rest of the time the same 737-like vehicle could fly passengers to space stations in Cislunar space or maybe even Lunar destinations.

Similarly, Robert Bigelow is building large inflatable habitats for space already. They can't land on and take off from Mars like the ITS, but they're cheaper and more effective as habitats because of that. The BFR proposed by SpaceX can launch 300 tons to orbit per reusable launch, which is big enough to put (I'm estimating here) maybe a 50-60 person Bigelow module in orbit. Attach some solar panels and a couple vacuum-rated Raptor engines to that and you've got your colonial transport. And during the months when Mars isn't in the right place for a colonial mission, you could rent out the space in the Bigelow module to commercial or scientific missions.

If you built something like an ITS for transporting crew and cargo from Mars orbit to the Mars surface, but it stayed at Mars all the time, you could use it all the time for point-to-point travel on Mars, moving cargo or fuel between Mars' surface and orbit, or even accessing Phobos and Deimos. It would also be designed just for that, and wouldn't have to have all the hardware necessary to keep 100 humans alive for months at a time.

To reason by analogy, there's a reason people drive to the airport in cars, fly to Tampa in a plane, and then sail around the Caribbean in a cruise ship - rather than have a cruise ship pick them up at their house in Cleveland and fly the thing door-to-door to the B&B in Nassau.

Solar Power
Elon's keynote and the animations only showed one power source for the system - solar panel arrays. I imagine you could also hide some methane fuel cells on board and run them off the methane in the fuel tanks, but they would be a backup power source.

Solar power is not a bad way to go for small probes and satellites. But the size of those solar arrays, rated for 200kw (2x the ISS) indicate that the IST will need a lot of power. Moreover, Elon tried selling the ITS as a "go anywhere in the solar system" ship, but solar power doesn't work much beyond the asteroid belt.

Simply put, ships of this size, and missions beyond the asteroid belt, will need nuclear. Elon mentioned the need for "public cooperation" to get nuclear power on Mars, and what he really means is that the Department of Defense will have to get involved to build nuclear power plants. They have decades of experience building them for carriers and submarines after all.

Fuel Depots
SpaceX is moving to an architecture where the colonial ship is launched with dry fuel tanks, and then it is topped off. That's good, because it massively improves performance. The BFR (assuming it flies as designed) will launch 300 tons per launch, but the ITS says it will take 450 tons to Mars. The difference is the fuel that will be loaded into it once it's already in orbit.

However you'll note the immediate problem which is that the ITS cannot be loaded with fuel until it's already in orbit around Earth - with all its colonists already on board. How long will it take to launch multiple tanker flights to it?

A much better architecture would be to have a large, passively-cooled fuel depot in orbit. The tankers could spend months or years (whenever there is a free launch window) loading up the depot, or the depot could even accept fuel from in-space sources such as the Moon or asteroids if that ever comes online. Then when the ITS launches it would top off its tanks from the depot, rather than wait in orbit for multiple tanker rendezvouses.

Alternatively, if the ITS launched without colonists, it could be the fuel depot itself for a couple months, and then the colonists come up in a 737-equivalent when the Mars window opened.

Biggest Concern: Too Much, Too Soon
My overall concern, which is reflected in all my other concerns, is that it's too much, too soon, and the over-complication of the ITS is really driven by the fact that there isn't an infrastructure in place to support this yet.

For instance, Elon say he's trying to be the rail road to Mars. That makes sense, to an extent, but he's trying to build a railroad to a place with no humans to build the railroad, no oil or coal to fuel the engines, or anything else. The ITS is less like a railroad to California and more like an all-in-one expedition to Antarctica.

I think the companies that are trying to gradually build up in-space resources by first, e.g., building a business to refuel the satellites already in orbit, then start getting fuel from near-earth asteroids, then go to the Moon, etc. are going to be more successful in the long term.

Elon's project is the ultimate example of a Moon Shot. And most Moon Shots fail.

Funding & Partnerships
This is where I think Elon said the least but implied the most. SpaceX cannot afford all the up-front costs associated with this, and they probably cannot sell a significant number of "tickets to Mars for one" decades in advance. Their launch business and eventual satellite business will produce a lot of cash flow, and Elon would probably be willing to sell off his Tesla stake at some point to pay for this, but that really only covers the R&D and early prototype work. Getting from there to "hundreds of ships in orbit all leaving for Mars at once" isn't within SpaceX's capital budget under any scenario.

And to be absolutely clear, the $200,000 (or less) ticket price only works (if it ever works) once the system is fully at scale. It may genuinely only cost $150,000 to land the 10,000th person to on Mars, just like it genuinely costs only $50 to fly from L.A. to Las Vegas, but that's only because all the planes and airports already exist. Landing the first person on Mars will cost $10-20 billion.

What I think is happening is that Elon is trying to play NASA off against some other national governments. Maybe not China, but possibly Europe or maybe some rich Middle Eastern country. The question that every national space agency is going to ask themselves is "Are we going to let [that other country] land the first man on Mars?"

America was the first and is still the only country to land men on the Moon. That's a huge part of our national pride and self-conception. Do you think Congress would be happy if the first men on Mars were a couple of Saudi Princes who paid $1 billion each for the privilege? Or even a joint EU-Japan mission?

The challenge that Elon needs to overcome for NASA's support though are the contractors who currently suck down $$billions each year building NASA's Space Launch System and Orion craft. There's no room in NASA's budget for both SLS and the BFR. It's one or the other, and the Senators and House Reps in the pocket of Boeing and Lockheed will fight like hell to keep the SLS gravy train going.

Elon has a powerful card to play though. Humanity only lands on Mars for the first time once, and the people who do so will probably be remembered for as long as written records from this era persist into the future. Just as we all know that Queen Isabella sponsored Christopher Columbus, who will history remember as the sponsor for taking humanity to Mars? Is it going to be Saudi Arabia? Will history say that Bill Gates had to pay for it because Congress was too captured by Boeing? You better believe that SpaceX lobbyists will employ this argument at some point.

Part 3: Open Questions

In Situ Resource Utilization
The transport system outlined by SpaceX had four major hardware parts - the booster, the colonial ship, the tanker, and the ISRU system for making fuel on Mars. But we saw exactly zero information about the ISRU plant. Has SpaceX even started designing what it would look like? How much fuel can it produce? At what rate? Can it be shipped to anywhere in the solar system and assembled on site? Does it need people to operate or is it 100% robotic? We have no idea.

The entire proposal (both Mars and destinations beyond) rests on the ability to produce fuel at the other end of each trip. There's no trip back, no reusability, without ISRU.

Martian Gravity
We know that zero-g is terrible for you. We have no idea whether Martian gravity is strong enough for long-term human health. We have no idea whether Martian gravity is strong enough to allow a woman to carry an infant to term and bear a healthy baby, or what effects the gravity will have on a growing child. Frankly the whole discussion of making humanity multi-planetary is entirely premature until we answer this question.

Free-space & Asteroids
Elon always talks about planets. He never talks about all the resources in space itself. His focus for humanity is always planetary, never in-space habitats as detailed by Gerald O'Neil. The resources of the Solar System in space are many times greater than the resources of any planet. Solar power is more abundant, the asteroids can be mined in their entirety (as opposed to a thin surface layer like on a planet; you can't mine a planet's mantle or core), and rotating 1g habitats could be mass-produced to create internal surface areas sufficient to accommodate any level of colonization. But Elon never talks about this. Why? Would the BFR be available to people who are willing to try?

Part 4: The Good Stuff

Despite my concerns, I'm still very excited and wish SpaceX the best of luck. Here's a short summary of why.

The Vision
NASA's vision of "Journey to Mars" is basically the same idea as Apollo. Spend a lot of money, send a few missions of a few people to Mars, and then stop. Flags & footprints are nice, and they're certainly a nice feather in the cap of the United States, but they don't change the picture for humanity. Even the Norsemen who tried to settle North America in the 10th century had a bigger vision than NASA.

Will Elon's particular mission architecture work? I have no idea. I have many concerns. But I'm glad as hell he's trying to do it. The American people can look at what SpaceX is trying to do and turn to NASA and say "What the hell, man? Why is SpaceX trying to build a whole fleet of ships while you're still dicking around building one Orion capsule?"

Elon has the potential to commercialize deep space at "flat" (not cost-plus) prices, just as he has already commercialized launch to LEO and GEO. This has implications far beyond Mars. If NASA can be convinced to switch to this model, we'll be in a world where NASA's mission planning office is a guy with a telephone asking for quotes. "We want to send 100 tons to Europa. Please RFQ."

Bare Launch Costs
Although he glossed over it quickly, the projected launch costs per ton to LEO for the BFR are about $150/kg. That's a great number. Cheap commodity transport is what allows globalization to work. Even if the BFR never sends a colonist to Mars, sending paid cargo to Cis-lunar space at that price would be awesome.

ISRU/Refueling on Orbit
There is no opening up the solar system (or even the Low Earth Orbit) as long as we need to take all the fuel for the return trip in the initial launch. Real space ships refuel in orbit and at their destination. A real space-faring civilization has fuel and supply depots throughout the solar system. NASA has never even tried to build this. If SpaceX succeeds in this, humanity's future in space is bright.

Sunday, September 18, 2016

Blue Origin: New Glenn, New Armstrong

New Glenn

Jeff Bezos announced to the media via an email (which in its own way is an unusual choice - why not a blog post on the company website) that their new rocket they were building is going to be called New Glenn, and it's a big sucker. Here's a picture:

As you can see, Blue Origin will not be messing around with small rockets like their New Shepard for much longer. They're going straight past the small orbital rocket stage, even past the medium-sized rocket stage, and going right for heavy lift.

The only numbers we know about New Glenn are-

  • 23 feet diameter
  • 270-313 feet tall
  • 3.85 million lbs force at sea level
  • 7 BE-4 engines

The interesting number here is the force at sea level. 3.85 million lbs is a lot. That's 60% more force than the fully optimized Falcon 9 or Delta IV Heavy, and 72% more force than the Falcon 9 was rated at when it first entered service. It's about 30% less than the Falcon Heavy is aiming for and half of the Saturn V. In other words, Blue Origin is skipping right past the current state of the art in heavy lift and is going for super-heavy lift. It's ambitious, to say the least.

However the two most important numbers are missing - price and payload mass to orbit. A rocket could have infinite thrust, but it only takes off to the extent its thrust exceeds its own mass * force the gravity. And it's only useful to the extent that its thrust is greater than mass*gravity such that there's mass left over for cargo. And it's only economically relevant to the extent it can do that for a reasonable amount of money. After all, NASA's SLS will have plenty of thrust too - it's the $2 billion per flight price tag that's the issue.

I am optimistic that Blue Origin's numbers will be good, but that's mostly just because I trust Jeff Bezos. I can't look at this pretty graphic up above and tell you what mass fraction New Glenn will be able to reserve for cargo. I'm just sort of assuming that Jeff Bezos knows a lot about shipping products from point A to point B and has given his engineers clear goals about what sort of payload sizes he expects. I'm also assuming that New Glenn will be fully reusable (like New Shepard is proving to be) and that whatever construction costs Blue Origin faces, it will be amortized over multiple flights and will therefore already be cheaper than any provider that isn't SpaceX.

Anyway, the real takeaway here is that this is interesting but "watch this space for further details".

New Armstrong

At the end of the letter announcing New Glenn, Bezos teased (without providing any further details) that Blue Origin's next rocket will be named New Armstrong. If it's not clear to you already, let me spell this out - Blue Origin is naming their rockets after the first Americans to achieve certain milestones.

Blue Origin's first liquid fueled test vehicle was named Goddard, after Robert H. Goddard, the inventor of the liquid fueled rocket.

Blue Origin's first sub-orbital rocket that could reach space was named New Shepard, after Alan Shepard, the first American to fly in a sub-orbital Mercury rocket to space.

Blue Origin is now naming their first orbital rocket New Glenn, after John Glenn, the first American to orbit the Earth.

New Armstrong must then of course be named after Neil Armstrong, the first man to walk on the Moon. The assumption then is that Blue Origin's New Armstrong is being designed to go beyond Earth orbit to the Moon and beyond.

Bezos has stated that he wants to help build the infrastructure that opens up the entire solar system to economic development and colonization. He's specifically mentioned manufacturing on the Moon and developing the resources available from asteroids. I suspect that New Armstrong will be the rocket that starts us going down that path. To that end, I would expect the following-

  • It's going to be really, really big. Obviously.
  • All stages will be reusable. Not just the first stage.
  • It's primary mission will be to launch mining equipment, true space ships, and space habitats - infrastructure that never comes back to Earth, but spends the rest of its existence in space. These habitats will allow workers (who fly up on a New Glenn maybe) to service commercial and scientific activity in LEO, GEO, and Lunar Lagrange points.

As I mentioned in the New Glenn section, I don't have enough numbers here to really get excited, but I'm starting to get excited by Jeff Bezos. His long term goals are ones I share, and thanks to running Amazon he knows a thing or two about shipping products from point A to point B, operating at large scales, and delivering value at low marginal cost. I think this is really happening, and Elon Musk finally has some real competition for being the man that opens space.