Not Another BFR Analysis...

So I got bored at work and started playing with numbers. This is what engineers do when they’re bored.

In the months following Elon’s presentation in Adelaide, I've seen a lot of analysis of the BFR’s financials when operating to Mars or operating as a point-to-point transport system. Despite all of this however, I haven't seen that much analsyis of the financials of using the BFR to launch satellites, regular old GEO communication satellites, and based on what Elon has said, I think this is the market that the BFR will address initially.

Now for this analysis, I want to see if the BFR can be profitable and pay back a several billion dollar development cost at current launch rates within current launch markets. Things will definitely change once this beast is flying regularly, but for the first few years, we’re not going to see any of the amazing rocketpunk this machine will allow. Essentially, can the BFR Business Case close with no new sci-fi fun starting?

 Assumptions – Making an A#$ of U and Me!

So, assumptions then! Remember I’m going to try and make this as hard for SpaceX as possible, my motto here is ‘harsh but fair’:

1. No significant changes to the Launch Market over the next 10 years (~5 years of BFR Flights). Why? This is a worst-case scenario, and nothing will probably change until the BFR has flown at least a few times, but any new missions will take several years of work/planning to bring to launch. Additionally, no one seems to be in a hurry to embrace reusability beyond Blue Origin, and while the market does seem to be shifting, it is unclear how much and how quickly it will change. 
2.  Launch Rates for SpaceX will essentially ‘max out’ at 30 launches per year. Why? Well firstly, see above. Secondly if global launch rates stay constant at about 60 to 70 launches globally per year, I think there is a maximum amount SpaceX can take, if only because the rest are government launches for other countries. I think it would be difficult for SpaceX to rise above about half of the global macro satellite launch rate. This also makes it harder to find an initial market for the BFR, making this a worst case scenario. Additionally, I'm ignoring Starlink, again to make it harder for SpaceX, but also because those will be internal launches, most likely done at cost. 
3. Launches will be almost exclusively for GEO Satellites of the flavour that SpaceX mainly launches now. Hence the yearly launch rate of the BFR won’t go far above ~20 for the first 5 years of its life. Why? NASA and the DoD have all sorts of restrictions for their launches. While SpaceX might do cargo runs to the ISS with BFR, I don’t see them doing crew on it for a while, even if it has the capability. Plus, it makes this an even worse case scenario. I can see satellite provides jumping on board quickly if it can do the job and costs less, but Government organisations can afford to be more conservative. Finally, see Point 2 regarding Starlink.
4. No one else develops a rocket that can meaningfully challenge the BFR in its opening years. Why? Well simply put, who can? Who has the reusable technology, is developing the engines AND has the drive to do so? People have only just started to react to the EXPENDABLE Falcon 9, and are only just beginning to think about the Falcon 9R. The only one who comes close is Blue Origins, and while I think they could adapt what they have to a similar rocket…a BlueFR if you will… they need to get to orbit first. 
5.  Costs scale off the original ITS from IAC2016. Why? Because I don’t have any better numbers…

Equations - The Tyranny of Maths!

Just some extra knowledge for anyone not familiar with it, or as a quick refresher, this is the Rocket Equation:

It comes curtesy of this guy:

Konstantin Eduardovich Tsiolkovsky - The OG Rocket Scientist

When it comes to rockets, this is basically the One Equation to Rule Them All!. It means you can work out pretty much anything to do with a rocket if you know about its mass and engines, and Elon was kind enough to provide both! Thanks Elon!

Side Note: Just remember that if your M2 is less than the dry mass of your rocket, you can’t do the mission.

Figure 1: IAC2017 Presentation "Making Life Multi-planetary" 22:43 - Lookin' under the hood!

Payloads – I think I can! I think I can! I think I can!

So the BFR is big, I mean, it’s kind of in the name.

And a lot of people say it should be able to take DOZENS of Satellites to GEO and come back, but has anyone checked? Let’s do it now!

So per Elon Musk “you can get 150 tonnes to Low Earth Orbit…but then you can’t go anywhere”. Thanks again Elon! That’s all I need (see above). So with a dry mass of 85 tonnes and 150 tonnes to LEO, that means any BFR going to GEO will have a mass of 235 tonnes. Ship, fuel and satellite all!

Table 1: I had to have a table

So let’s go to GEO! Typically it takes 3,800 m/s of delta V to get to GEO from LEO, so with an initial mass of 235 tonnes, that mean we get to GEO with a mass of:

83.64 tonnes

That’s less than the dry mass of the BFR Upper stage…

Damn…

And this is backed up by Elon’s own slides.

Figure 2: Exhibit A - Evidence we're not going to GEO today... (Rhyming!)

So basically, not only can we not take satellites to GEO with the BFR…we can’t even GO to GEO with the BFR! At least not with a single launch. Now we could refuel the Whale in LEO, but I feel that would introduce too much operational complexity for most satellite operators, at least for now. If something goes wrong with the fueling, your satellite is stuck in LEO. Fortunately, we have options.

Now let’s try GTO and see if we can do that. Most GTO missions need 2,500m/s of delta V from LEO. So with the mass, the engine details and the delta V, we can see how much mass is going to GTO.

119.11 tonnes

Phew! We can make that one with some room to spare! Now if the space ship wasn’t coming back that would mean our satellite could be almost 35 tonne… but we would like our Spaceship back please! (It’s kind of expensive). So, let’s roll with a standard GEO communication satellite with a mass of less than 7.5 tonnes. Let’s say 10 tonnes for good measure. DEPLOY! And now our ship is:

109.11 tonne

Now fortunately, it doesn’t take much to come back down from GTO. 50 m/s of delta V delivered at apogee (the highest point of the spaceships orbit) to slow the Spaceship down will be more than enough to lower the perigee (the lowest part of the spaceships orbit) to less than Earth’s radius. So after that second burn, our Spaceship now has a mass of:

107.64 tonne

At this point, most of the leg work of slowing down will be done by the atmosphere, so this will give us 22.64 tonnes of fuel for landing. That means we can be traveling up to 764.37 m/s when we’re coming into land, AKA Mach 2.25.

So… not exactly flying dozens of GEO birds around the place, but it can definitely do the mission for pretty much anything commercial satellite providers fly today.

So why does such a Big F@#$ing Rocket take such a Small F@#$ing Payload to GTO when it’s supposed to be able to colonise Mars? The reason is just that, it’s a massive rocket designed to colonise Mars. The upper stage weighs about 85 tonnes, while the upper stage of a Falcon 9 only weighs about 4.5 tonnes, AND it doesn’t have to come back. So once the BFR is in space, its size seriously works against it, BUT it does the job! So how much would a mission cost SpaceX to run, and how much could they charge their customers?

Figure 3: Orbits (to scale!....ish)

Launch Costs – That Rocket Money!

Figure 4: Launch Costings from IAC2016 - Important things are always circled in red...

So now we know the BFR can do a straight forwards GTO mission in a single launch. How much will it likely cost SpaceX to do? We don’t know the costs currently, but we can guess. I think it's reasonable to assume the cost will scale with the mass from the design of the ITS of 2016, given their similarity.

So, we know the ITS Booster was projected to cost $230,000,000 dollars to build and the Spaceship was projected to cost $200,000,000. If we assume the cost scales roughly with the dry mass, the new Spaceship will cost $133,333,333 and the Booster will cost… we don’t have a mass!

But if we assume the dry mass of the Booster scales roughly with the dry mass of the Spaceship, the new booster will cost approximately $169,761,901. Let’s call it $135,000,000 for the Spaceship and $170,000,000 for the booster to make the numbers nicer. Great! We have the base cost of the rocket, $305,000,000 in total to build.

Next, reusability is key here. I think it’s reasonable to assume that they can reuse the booster and spaceship as many times as they stated in 2016, I.E. the spaceship 100 times and the booster 1000 times, however, given that I am assuming only 20 launches per year, I think it would be reasonable to assume the booster would be charged at a rate of 100 reuses instead. If not it would take 50 years to pay off the booster. Again, worst case scenario.

Therefore, for a single launch the cost of the rocket would be $3,050,000. Add to this the fuel cost, approximately $700,000, maintenance costs scaled again based on the mass, approximately $485,000 and the launch site costs, kept the same as the ITS prediction so approximately $200,000. Add this all together and include a 25% margin for insurance, administration and extras and you have a predicted launch cost of:

$5,543,750

...Bargain

This is inline with what Elon said in Adelaide, implying that the BFR would cost less than the Falcon 1 to launch, so at most $8,500,000.00.

This doesn’t include a profit margin, and that’s where things get REALLY interesting!

Step 3 - Profit!

So here’s the meat of these musings, how much money can SpaceX get from launching a regular GEO communication sat on the BFR? Let’s have a look at where the market is likely to be in 2022.

Ariane 6 will launch for between €75,000,000 and €90,000,000, approximately $90,000,000 to $105,000,000. This is for 4.5 to 12 tonnes to GTO respectively. At this point the Ariane 6 would have been flying for about 2 years.

Though there aren’t any hard numbers yet, ULA is targeting a price point for the basic, bare-bones Vulcan of $99,000,000, with a maximum payload of 15.1 tonnes to GTO launching on a Vulcan 561 probably going for somewhere in the neighborhood of $125,000,000+. Again, Vulcan would have been flying for 2-3 years. 

Blue Origin’s New Glenn will be flying as well, but there aren’t any numbers for it yet beyond it's payload capability of 45 tonnes to LEO and 13 tonnes to GTO. With its colossal 7m fairing it would be big enough to throw 2 GEO communication satellites at once. One would assume they would at least compete with where the rest of the market is at, however. 

Also Proton… maybe.

Interestingly though, none of that matters, except for New Glenn, but like I said, I can’t find numbers for it.

This is because Elon Musk and SpaceX will, at the very least, have to compete with Elon Musk and SpaceX. In this time frame they will be flying their partially reusable rockets, namely the Falcon 9 and Falcon Heavy, from anywhere between $40,000,000 to $90,000,000 depending on what people want, and with the recent launch of the Block 5, I think this is a reasonable expectation. Both configurations of the Falcon will be able to do the job of the of the BFR to GTO, so Elon is going to have to compete with Elon at the very least. Possibly worse depending on the New Glenn.

Figure 5: FIGHT!

So, in order for the BFR to be attractive, it’ll have to launch for less than a Falcon 9, again in a worst case scenario. Let’s say going for $40,000,000.00. So what does that mean?

SpaceX could launch the BFR with a profit margin of $34,456,250.

That’s what I'd like to call an Obscene Level of Profit.

At that rate, with 20 launches per year, SpaceX could make a yearly profit of at least $689,125,000! And that doesn’t include profits from other launches on the Falcon’s over the same period of time we’re talking about.

That means that over the 5 years from when the BFR débuts to when I believe markets would catch up with the new paradigm, SpaceX could net $3,445,625,000 in profit. No flights to Mars, or the Moon or even putting Comcast out of business required.

But how much would the BFR cost to develop?

In his book, The Case for Mars, Robert 'Bob' Zubrin estimates that any new, high end aerospace project roughly costs $10,000 per kilogram to develop. This means that the BFR would cost roughly $2,450,000,000, with an extra $1,000,000,000 for engine development. Finally, let’s throw in the cost of building three full BFR’s. One for development, one for the first 5 years of flights and one back up. That’s going to be another $915,000,000. Therefore, if the total cost of development is $4,365,000,000, then it would take just over 6 years to pay for the full development with no change to the existing market.

I think I know why SpaceX is going all in.

What’s the Catch?

Now of course, this doesn’t take into account changes in the market or competitors, which quite frankly could go either way. The wildest card in this whole thing is Blue Origin and the New Glenn. SpaceX NEEDS the average launch price for the market to be in the tens of millions so that it can make this Obscene Level of Profit, and although Blue Origin hasn’t announced prices, it is likely it could do large payloads to GTO. Combine that with the 7m fairing, they conceivably could do multiple satellite launches, which means they could rival a cost of ~$40-30,000,000 to GTO. This would cut dangerously into the Obscene Level of Profit, which SpaceX NEEDS to pay off the BFR quickly.

On the other hand, with people already anticipating lower prices, plans are starting to develop on the horizon to capitalize on low cost access to space, see Planetary Resources and Bigelow Aerospace. While NASA may be cagey about the BFR even getting close to the ISS, private space stations may have no such qualms. With a demonstration of in-space refueling, the Moon and Mars open up, and of course there is the point-to-point transport and the satellite network.

Additionally, I think my estimates on how much the BFR will cost to develop is the weakest part of my analysis above, mainly because it doesn't include the cost of building new infrastructure beyond the launch pad costs per flight, the cost of transporting the stages to the East Coast of the States, or the cost of development flights. On the other hand, SpaceX doesn't have to develop everything from scratch. They already know a lot about rockets and doing rockets cheaply. That would make the BFR a lot cheaper to develop than, say, the SLS. 

The point is, if the market doesn’t move or only moves slowly, SpaceX may not need any new futuristic applications, like point-to-point transport, trips to the Moon or trips to Mars to pay for developing the BFR. They don’t even need it to take over all their launches, just most of them, avoiding the risk of losing longer-term customers. All they need is for launch providers to not mind the look of a Saturn V launching a regular satellite to GTO. In my mind it’d look like the equivalent of using a 747 to deliver a bundle of air mail, but if it means a discount of between $10,000,000 and a $100,000,000, they’d get over it.

This is why I believe SpaceX is doing this now, and doing it as fast as they reliably can, beyond just 'Elon is impatient for Mars', although that too. They’ve let the genie out of the bottle (or are close to it), and they only have a brief window to make these huge levels of profits. Once the next guy comes along, they won’t be able to make a $30,000,000 profit on every launch anymore, and the next guy’s name is probably Jeff. 

Cheers,
A Bored Engineer

PS:
Did anyone else notice that the first 'trip' after the 'New York to Shanghai' depiction for the point-to-point transport concept was from Adelaide, Australia, and looked like it was going towards LA at 1:25 into the video? Well played Elon...well played...

Figure 6: A Land Down Under

PPS:
While I was writing this, I got a bunch of other ideas in my head. What should I do with my time at work next? Instead of actual work of course:

1. Landing Site Selection for the first BFR's to Mars.
2. Adapting BFR's to Bob Zubrin's "Mars Direct".
3. BlueFR.
4. Martian Economics