So let's see, the Falcon 1 was the first rocket we built. The Falcon 1 was really developed from a clean sheet to on the launch pad in three years, and that includes the entire vehicle. It's the fastest launch vehicle development in history, including war time. The entire vehicle was designed, build and tested at SpaceX, almost, there were a few key pieces that were procured outside, but the big stuff was developed from scratch. We kind of had to do that, because if we were to cobble together stuff from existing quasi-official components, then we would have been unable to reduce the cost. Because to the degree that you inherit the legacy components, while you may inherit their heritage of course, you also inherit their cost. I don't want to paint all aerospace suppliers with the same negative brush. I think there are definitely some good ones out there, but generally we found that if you want something cheap, fast, and that's probably going to work, then you should use a regular commercial supplier. If you want something that's expensive, takes a long time, and might work, use an aerospace supplier. So of necessity we were forced to make the major items, like the engines, and the stages, and the avionics, and the launch ops, and all that from scratch.
The idea behind the Falcon 1 was that it was built as a scale model, so we could test out the technologies. When we made mistakes they were made at a smaller scale, rather than jump immediately to a large rocket and make mistakes that cost ten times as much.
Two years after starting the company we had the qualification article of Falcon 1 on the launch pad at Vandenberg, and then about six months later we did the static fire. We initially set up at Vandenberg Air Force Base, which is about two hours away from Santa Barbara. The first launch was going to be from the Space Launch Complex at Vandenberg as soon as the Titan IV departed. Unfortunately Titan IV got significantly delayed, so we were forced to move from Vandenberg to Kwajalein in the Marshall Islands. So we had two launch sites and control centers. We expected to actually have two rockets - one at Vandenberg and one at Kwajalein, possibly on the pad at the same time. We had our own little island there called Omelek. It sounds a bit like a Bond villain launching rockets from a remote tropical island with Dr. Evil or something.
From May of 2005 to November of 2005 we were able to set up a launch facility at Kwajalein, which is quite difficult because the island we were given in the Kwajalein Atoll was - just had nothing on it, it was just jungle. We're not currently using the Marshall Islands launch site. The logistics are just too difficult, getting out there. It's like Waterworld out there, it's miles from anywhere. It's convenient in some ways but then inconvenient from a logistics standpoint. We had to bring in power, water, RP - rocket propellant (kerosene), all the pressurants, offices, that sort of thing. We had many challenges, liquid oxygen in particular. Kwajalein is 5000 miles away from California and over 2000 miles away from Hawaii, which is the nearest source of liquid oxygen.
We managed to have our first countdown right on Thanksgiving 2005, had turkey on the island.
It took us four countdowns to get to the first attempt at launch, which was in March of 2006, which failed. The first rocket didn't go very far, went about a minute up, and then there was an engine fire and that was it. The engine shut off about 30 seconds into flight, it continued ballistically for another 30 seconds, and then landed like an anti-tank weapon maybe a couple hundred yards away from the launch site, in tiny fragments. In fact I spend the day picking up bits of rocket pieces off the reef, which sucks. The telemetry actually showed that there was a kerosene leak at the turbo pump inlet pressure transducer, which started about 400 seconds prior to liftoff. You couldn't see it because the wind was blowing and kerosene is actually very difficult to see. When the wind's blowing you can't actually see that it's leaking. We didn't know why the leak had arisen, and it was a big shock to us.The failure review board, which was actually CO-chaired by Pete Worden of NASA/Ames, concluded that it was due to corrosion - stress corrosion cracking of the aluminum 'B' nut on the engine. That leak ignited a few seconds prior to start, and the fire basically burned through the entire powered flight. About 25 seconds into the flight it burned through a helium pneumatic line, resulting in losses in helium pressurant and that caused the pump pre-valves to shut, and essentially turning off the engine. Other than that, everything looked good. The vehicle was proceeding along its designed trajectory within 0.2 degrees. All first stage systems were nominal, and all avionics were nominal. The sad thing is that the problem was a corrosion issue due to the Kwaj climate. It's a problem that would not have occurred at the launch site at Vandenberg.
We took a bunch of corrective actions. We improved vehicle robustness by eliminating as many fittings as possible and going to orbital tube welds. There were a number of other changes. We also added more detailed procedures, more personnel per process. The biggest single change was, we had messed up software monitoring launch and automation, we were monitoring approximately 30 variables. We went to monitoring 800, including both the vehicle and the ground support equipment. We would have caught the fuel leak if we had this system in place. The countdown was now also fully automated, which reduces the potential for human error and allowed us to review the data. It also allowed us to take some number of personnel out of the countdown process.
We actually had two more failures after that one. The second and third flights arguably got to space, but they did not reach full orbital velocity. Demo Flight 2, which took place in March 2007, made it almost all the way to orbit. The post-flight review showed late in the second burn a roll control anomaly. The only orbit critical issue was the lack of slosh baffles in the second stage LOX tank, which caused a coupling of the controller slosh modes. Exposed the propellant line going into the engine, it sucked in helium gas and flamed up the engine.
Flight three also didn't get all the way to orbit. With the stage separation, the residual thrust on the first stage basically send it back and gently collided with the second stage, but it meant that the second stage did not exit the interstage, what joins the first and second stage. The interstage is kind of a sleeve that joins the first and second stage. So when the second stage engine ignited, it ignited in that interstage. The impact of the first stage on the second stage didn’t cause any damage, but the problem was that the second stage ignited in the interstage which it was not designed to do. You had this huge plasma blowback because it was igniting kinda within a closed space, that fried the second stage.
I was really devastated, that was awful, and so was the rest of the company. It was tough, tough going. At some point we'll release the blooper reel, but I think we'll wait a few years before we do that. That was a lot of pain.
We learned with each successive flight, and were able to eventually with the fourth flight in late 2008 to reach orbit. I was so stressed out at the launch, I didn't even actually feel elation, I just felt relief. I think there's a pretty powerful fear response ingrained because of the images of those rocket failures kind of going through my mind as I'm seeing the rocket launch. It's extremely nerve-racking. The thing with rocket launch is that all your work is distilled into these few minutes, particularly the first several seconds around the lift off. Because the worst thing that can happen with a rocket is if you have an engine failure or some huge failure right above the launch pad. The whole thing can come down with about 1 million pounds of TNT equivalent and destroy the whole launch pad. That’s what's going through my mind in case you're wondering. When it clears the lighting towers, and it's gotten further enough away from not actually destroying the launchpad, then you sort of go down a notch. Then after the first stage separation that's another one, and when the second stage lights up. You sort of go down in intensity as the rocket is going up. After the rocket lifts off, 9 minutes later it's either in orbit or it's exploded. That's a nervous 9 minutes.
That fourth launch was a very very close call, because I'd run out of money, that was with the last bit of money we had. We were really down to our last pennies. In fact, I only thought I had enough money for three launches, but we were able to scrape together enough to just barely make it into a fourth launch. So, that was a bit of a nail biter — thank goodness — that launch succeeded, If it hadn't SpaceX wouldn't be around. Thank goodness that happened. I think the saying is fourth time is the charm?
There was really no ability to raise outside money in a meaningful way in 2008 because of the financial crisis. There weren't a lot of people who were keen on funding a rocket company. I think if we'd said: “yes, our fourth launch wasn't successful, but the fifth one's the charm” that would not have gone down well. You can imagine trying to go to raise money and saying, well yes, we've just had four failures and the world is in financial ruin, but would you like to give us some money? It would be a definite no.
I think perhaps if I had been more knowledgeable we would have gotten to orbit sooner than flight 4. I think we had a critical mass of technical talent, and just enough money, and a design that was sensible, those were probably the three ingredients that resulted in success eventually.
The thing about a rocket is that the passing grade is 100%, and you don't actually get to test the real environment that the rocket is going to be in. At least with a car you can do a recall or a software update. It's not going to happen with the rockets. It's like passing grade is 100%, which induces anxiety. I think the best analogy for rocket engineering is a software analogy. It would be like if you had to write a whole bunch of complicated software modules. You can never run them together as an integrated whole, and you could not run them on the target computer, so when you're testing them you have to test them individually and not in the actual computer that they are going to run on. Then you put all the modules together in a completely different computer, but the first time you run it, it has to run with no bugs. That's basically the essence of it.
Thankfully, the fourth launch did work, that I think gave customers of ours, NASA and others enough confidence to award us additional launch contracts, and for additional private investment to come in and help fund the company besides myself. I think that was very helpful that coming out of PayPal I had a bunch of capital that I could spend developing rocket technology, even though I had no experience in rockets at all. If I'd tried to get funding from a venture capitalist, they would have been angry that I met with them, probably. Even in the best of circumstances, space is outside the comfort zone of most venture capitalists. Although, in a few years - I think five years after starting the company was when the first venture capital came in, and I would like to thank those investors for having faith there at an early stage to invest in a rocket company, we raised some good partners there.
Then about five or six years after the start of the company we started getting support from NASA. Then we also developed the Dragon spacecraft, because somewhat opportunistically NASA announced they were going to retire the Space Shuttle. They didn't have the budget to develop a cargo transport capability to the Space Station via the normal large government way, and so they put it out to bid to commercial industry, for the first time in NASA history. It was quite a big step, and we were lucky enough to win one of those contracts. Then the other company wasn't able to execute, so they got cut, and so we ended up being the primary means of transporting cargo to and from the Space Station.
NASA's certainly been a key customer of ours for a few years. The first five years or six years of the company nobody would talk to us on the government side, NASA or the military. We got a few sort of fringe customers on the commercial side, that was it. We didn't have any government anything for the first half decade. Then NASA nibbled a little bit and we were able to get a small contract, then we were able to get a much larger contract.
I just want to say that I'm incredibly grateful to NASA for supporting SpaceX, despite the fact that our rocket crashed. It was awesome, — so thank you very much to the people that had the faith to do that. I’m a big fan of NASA. In fact at one point my password was “ILOVENASA” literally that was my password. Hopefully I don’t have some old email account. I'm NASA's biggest fan, and SpaceX would not be where it is without the help of NASA, both historically the great things that NASA has done, and currently with the business that NASA gives us, and the expert advice and everything, so I should make sure to very strongly credit NASA in this arena in terms of how helpful they've been. We have a number of NASA personnel working at SpaceX. In fact we also have NASA personnel who are permanently resident at SpaceX.
So we go from 2002, where we were basically clueless, and then to Falcon 1, the smallest useful orbital rocket that we could think of, which would deliver half a ton to orbit. That rocket ended up costing around $6 million compared to other rockets in that class, which were about say $25 million. Yeah, like a quarter. We were able to go from the Falcon 1 to begin designing the Falcon 9, which is an order of magnitude larger vehicle and in fact, has more than 20 times the payload. It's got a payload to orbit of over ten tons. From those first days, where myself and the team were picking up bits of rocket off the reef, things have come a long way thank goodness.