Monday, July 12, 2010

Astronomical Sums ...

Rocketpunk era orbital scene
In the previous post I considered whether we could swindle our way to the stars by way of an economic bubble. In the discussion that followed, commenter Ian Wright put a set of numbers to the idea, at the early side of the midfuture, the sort of thing that might happen by midcentury. I started to play with my own numbers in a comment, then decided to do it here as a front page post.

I'll start with a comparison point. The global airline industry is about a $500 billion business, and roughly 16,000 jetliners are in service - so each jet accounts for some $30 million in tickets and airfreight annually.

Boeing's planes (which account for about 75 percent of the total, according to Boeing) range in price from $50-60 million for 737s to $235-305 million for 747s. More small and midsize jetliners are in service than big ones, so the 'average' jet probably costs about $100 million, roughly three times the revenue it is expected to earn each year. Keep that in mind as a general rule of thumb; it happily matches the guess I made in a commentary on interstellar trade at my old website.

Now, suppose that TSTO shuttles cost 10 times as much to operate as jetliners, thus $300 million per year in revenue for each shuttle in service. (Shuttle price at the factory gate would thus be roughly $1 billion.) Let each shuttle carry 25 tons to LEO, and be good for a mission each month - say, a one week mission, from rollout to the launch pad to roll-in to the service hanger, plus a week for postflight checkout, a week for pre-mission prepping and setup, and a week for Murphy. So it costs $300 million to orbit 300 tons, a nice even $1 million per ton.

This is a thoroughly conservative estimate, by space geek standards, and yields a launch cost 'just' one order of magnitude less than current practice. If passengers can go up in airline style seating the cost of a round trip to orbit could fall as low as $250,000, a mere pittance, not much more than Virgin Galactic plans to charge for a mere suborbital pop-up.

Remarkably, the commercial satellite industry, taken in all, is a $160 billion business, a third the size of the airlines. But only a small fraction of this money goes into space: about $10 billion per year to build satellites and $2-3 billion per year to launch them. (A clean apples to apples comparison is hard to come by, but these data are at least indicative.)

Sooo ... putting this together, what might we get?

Suppose that, at the peak of a boom - or bubble - the space launch industry is a fifth the size of the current world airline industry, $100 billion per year. This is an impressively big industry, considering that the airlines go to places where people already live.

It supports an operational fleet of some 300 shuttles - more or less the fleet size at which you can spread out development cost across a decent sized production run, and get some real production efficiencies going. Together they can loft 90,000 tons into orbit each year; call it an even 100,000 tons per year, for luck. In the course of a decade, a million tons, at total cost of a trillion dollars.

(Those of you with suitable keyboards can substitute the Euro symbol; at this level of approximation the exchange rate makes no matter.)

But in the satellite industry today, space launches account for less than 2 percent of revenue. If you scale our hypothetical space-boom industry accordingly, it is about a $5 trillion industry, or about 8 percent of current world GDP.

Bzzzzzz! No sale. This does not pass the smell test, short of alien invasion scenarios. We will have to goose these numbers downward.

Commercial satellites today cost about five times as much to build as they do to launch, but they are packed with very expensive electronics, while our much heavier payloads are mostly (relatively) cheaper structural fabrications, supplies, fuel, and the like. Let these cost about as much at the factory gate as launching them does, $100 billion for 100,000 tons of hardware and supplies - an average cost of $1 million per ton, close to what an airliner costs.

The direct cost of our space effort - building stuff and launching it Out There - is now $200 billion per year. The satellite industry spends about 12 dollars on earthbound goods and services - everything from surface stations to home dish installers to marketing - for every dollar it spends on space launches and hardware. Let our space industry spend only twice as much on what stays down as what goes up, thus $400 billion. Thus the total operating cost of the industry is now $600 billion, one percent of current world GDP. This makes it four times the size of the current satellite industry and rather larger than the current world airline industry.

In the course of a decade we rack up $6 trillion in expenses, money on which the investors expect a handsome return, say $400 billion per year on top of operating costs. (Which makes it a trillion dollar industry, if it expects to return a rather modest 7 percent clear return on investment.)

For this we get, over a decade, a million tons lofted into space, including structures and equipment, supplies, propellant, everything. Suppose that 20 percent of that is station or hab facilities, thus 200,000 tons. The ISS has a mass of 300 tons and a crew of six, one per 50 tons, so let us say that we have 4000 people in space.

Getting all of this ginned up and ready to go - for example, developing and testing those shuttles - might take another decade, so we're looking at 20 years, give or take, from inception to the point where we have a million tons and 4000 people in space. But we don't charge extra for this; the development cost is folded into the overall cost as given above.

Ian Wright in his comments starts with pretty similar numbers and ends up with a lot more people in space - a million - which merely shows that small differences in input assumptions can yield big differences in end results. But even 4000 people is a meaningful down payment, the germ of a space-living community.

(As another interesting reference point, Robert Heinlein back in 1949 gave $30 per pound as the Earth-Moon shipping charge, equivalent to $67,200 per ton. Allowing for inflation between 1949 and 2009 that is equivalent to a shade under $600,000 in current dollars. That's for getting to the Moon, not just LEO, but LEO is the toughest part of the trip, accounting for perhaps half of the total cost, $300,000 per ton, similar to Ian's estimate and about a third of mine.)

To get this, or something on roughly the same order, all you need to do is to persuade the world to pony up about one percent of its income for a decade or so, and build an industry comparable in size to the airline, automotive, or tech industries. Possible? Yes! Challenging? Very!

Related link: My old website post on Interstellar Trade.

The image is, naturally, swiped from Atomic Rockets, just above the Heinlein passage cited above.


Albert said...

You say that the costs of satellite industry are mostly due to expensive payloads.

But if you are sending up something cheap like water, hydrogen/oxygen, simple construction materials, and so on, you can build much cheaper and smaller launch rockets, that have 1/3 or more of failure rate (after all, if they go boom you don't lose a lot) but this big cut in reliability means they are far cheaper.

This is what makes me argue against using a human-rated thing to boost up cheap expendables (or even expensive hardware for that matter).

There should be three classes: -Dirt-Cheap for dirt-cheap bulk-freight payloads like water, propellant, structural beams and whatever.
-Reasonably Reliable for high-cost hardware (what we have now for satellites)
-Human Rated, that lifts up Pink Humies in safety but would be uneconomical to use for bulk freight, but less so for hardware delivery.

A recent Dirt-Cheap SSTO concept that should achieve 600'000$ per ton per a hundred launches per year:

A good read:
A Rocket A Day Keeps the Costs Away.


Unknown said...

Keep in mind that my numbers were all over the place - I think I dropped three orders of magnitude somewhere, and added another three orders of magnitude somewhere else. But either analysis goes to show that a bit of irrational exuberance goes a long way. The original consortiums involved in the bubble, be they government-backed or corporate, will probably not survive the bust. But their example of how to/how not to build a management structure will be invaluable in the recovery following the bust.

Expect the recovery to have a brief period of pull-back, possibly as shoddy hardware is abandoned or allowed to burn up in atmosphere, followed by a long period of marginal or no growth but steady profits for a few players. Then the margins start to creep up again, and a new generation enters the business. This generation doesn't really remember the bubble, except as a brief lecture in their MBA program. But they are absolutely certain that their new cost-control and risk analysis models have managed away the risks involved in their business...

When I'm not at work I'm going to take another look at the numbers involved. I wonder what the oil boom actually worked out to as a percentage of GWP.

M. D. Van Norman said...

One percent? No can do, Rick. We have wars to fight and old people to feed.

Thucydides said...

The fact that space enterprise is currently a monosopy and there are very high costs of entry (to date) coupled with very low volume of traffic, you get a very liited ability to generate a decent ROI on space investment.

Even the various launch companies who have laboured over several decades to build cheap launch platforms have to figure out how to amortize the R&D costs and the start-up costs over as few as 3-4 launches a year (the Government will certainly use its monosopy power to reward favoured clients, and even if there were some mechanism like drawing straws, the total number of launches per year is still very low).

I still am inclined to believe that space tourism will have the potential to break the monosopy, and provide the amortization to support larger numbers of manufacturers and operators. A boom and bust will probably happen during the startup rush when SpaceShip Three shows its chops and Scaled Composite decides to sell them to all comers; everyone will say "$200,000/ticket? I'm going for that market", while Boeing, Airbus Industrie, Lockheed Martin, Mikoyan-Gurevich and everyone else with an aerospace division will also be swarming aboard to cash in on the new market.

Companies which really can produce spacecraft will have a certain edge, and finally have a market to recoup some of their costs (SpaceX is probably going to be a winer under these circumstances), followed by companies which produce plain, low cost craft (offshore companies in India, China Brazil etc. will become big winers here).

While this would seem to cover ground from past posts, I am thinking of how to generate the cash flow to support "astronomical sums".

Given that the current economic situation is in unstable equilibrium; deficit spending over the past several decades has created a huge inflationary bubble. Either it pops on its own, resulting in a depression and deflation [the "long depression" of the 1870's was a deflation of a bubble caused by speculation in frontier settlement followed by inflationary spending during the Civil War; while the "Great Depression" was caused by the vast inflationary bubble generated by the warring powers as they tried to finance the conflict.

If more "Stimulus" spending and vast bailout schemes to prop up debtors like the PIIGS is attempted, a runaway inflationary bubble or stagflation will be created, with negative effects for all.

After a decade of sorting through the financial rubble, investors will flock to "new" and potentially profitable opportunities to make up their losses, and space will be a beacon for them. Money will flow from mattresses and municipal bonds and every other "safe" but low ROI investment, generating the astronomical cashflow required.

Rick said...

A few tourism stats. The global cruise ship industry is roughly a $25 billion industry, while 'luxury travel' is $100 billion.

Looking at Xtreme vacations, recreational scuba diving is about a $1 billion industry. The US skydiving industry is about $25 million - the world industry might be $50-100 million, at a guess.

You can draw a lot of different conclusions from these numbers, but I'll toss them out there.

Separately, I'd argue that the relationship between inflation, per se, and speculative bubbles seems fairly tenuous. The great inflation of the 1940-1980 era dwarfs anything before the Great Depression, and doesn't seem much tied to the current Great Recession. But I'll avoid belaboring the issue further!

Gridley said...

Full disclosure: I work for Boeing. I have access to proprietary information on this general topic that I am going to try VERY hard not to utilize here; ie everything here is open-source data.

I'm puzzled why people assume that a space-launch industry must make money. Historically the commercial airlines don't, and they're still around. With the notable exception of Southwest, IIRC every single US airline has suffered a net loss over the course of its history. Quite how the airlines survive as a whole I'll leave to people who understand the slight-of-hand that is modern economics. The fact is that they do, so if we use them as a model space-lines need not be profitable in and of themselves.

I'm also very confused by the jump you make at the start of paragraph four; no problem with 10 times the cost to operate, but how does that boost ticket and production costs 10 times?

Noting that airlines as a whole are a money-loosing section of the economy (even in good economic times). This is NOT to say the aviation industry is, BTW, just the airlines. So your 'average' jet costs more to buy and operate than it makes in revenue.

Boeing engineers our planes for a 20 year first line service life. Your average plane thus has a total revenue of $600 million. It costs $100 million. Throw in another $100 million for financing (using my home mortgage as a guideline), and 100 employees (nice round number) at $100k/year each for another $200 million. Another $100 million for fuel and parts. Then landing fees, in-flight food service, taxes, and everything else for another $100 million plus.

So 1/3 of the money goes to buy the plane, half of which goes to a bank. Fully 1/3 goes for people's salaries.

Why will people demand 10 times as much money to work in space? I don't think they will, and most of them will be located ground-side anyway. So making the bird ten times as expensive doesn't mean you need to raise fees by a factor of 10.

Unknown said...

@ Gridley: "I'm puzzled why people assume that a space-launch industry must make money. Historically the commercial airlines don't, and they're still around. With the notable exception of Southwest, IIRC every single US airline has suffered a net loss over the course of its history. Quite how the airlines survive as a whole I'll leave to people who understand the slight-of-hand that is modern economics."

That part in bold explains why I assume the space-launch industry must make money. For one, future economies will do away with our modern sleights in favour of their own, and who knows what those will be (What would an accounting trick look like in a distributionist economy with an energy-surplus-based currency?). For two, if we do get into a discussion of the modern sleights of hand, then these comments will quickly degenerate into a political 'discussiion', and I for one don't want to go there.

Anonymous said...

Ok, space launches don't need to make money; they just need to be cheap enough to allow the related sectors of the industry to be able to make money by increasing the rate of access to space. If you make money from comsats, then having the biggest, most, and most capable in orbit, then so long as the method you use(rockets), to get those comsats into orbit is less than what you make from the comsat sevices, then you simply fold the negative revenue into your overhead costs, and go on from there.

The lower the cost of space launches, either from 'natural' or 'artificial' causes, then more customers will buy launch space (both passengers and cargo), and even if the space launch companies loss money, the industry that utillizes what they put up into orbit makes money; and subsidizes the supporting industry that losses money but is vital to the profit making industry. Passenger airlines might loss money, but I bet that other industies that depend on air travel also subsidize that money lossing industry.
Both space launch and passenger airlines may loss money, but neither operates in a vacuum (pun intended); both are part of a larger web of interrelated industries(a sector of the overall economy) and must be viewed in that larger context. If profibility goes up, significantly, in one industry of the space sector of the economy, then that will fuel the expansion of the entire space sector. Whether that is reduced launch costs, profitible PowerSats, He3 mines on Luna, a rush to find micro-aliens under the surface of icy moons, or whatever the next big thing is in space, so long as the increase in one industry is more than enough to offset the loss in any one (or even all the rest) of the other industries in that sector, then the overall effect will be an increase in that sector as a whole. What happens with one part affect all the other parts.

And now my knowledge of economics is exhasted...


Rick said...

(Belated) welcome to another new commenter!

Gridley is quite correct that the airlines, historically, have not been profitable. A couple of points:

From broad economic perspective, profit is a cost of doing business - what you pay investors for the capital you need. There's also a saying that on advanced accounting exams the true correct answer to 'What is the firm's profit?' is (within fairly loose constraints)'What profit do you want to report?'

My concern in space economics is not really 'profit' as such, but whether space operations more or less pay for themselves, without needing subsidies that might be withdrawn at any time.

But note that since the Apollo boom and bust, NASA's budget has been relatively stable for 40 years. The US political process, through a mix of 'special interests' and a moderate public enthusiasm for Cool Space Stuff, is willing to pony up roughly 1 percent of the federal budget for space travel. This might continue for generations to come, even if no purely economic reason for doing it emerges.

As for my cost figures, I simply assumed that the cost of operating a 'typical' shuttle for a year will be 10x the cost of operating a typical jetliner, and that the ratio of purchase price to annual operating cost will be similar.

My estimated ticket price is much more than 10x intercontinental airfare, because (in my model) shuttles are much less productive - flying one mission per month, whereas jetliners can fly roughly one full-range trip per day.

Probably, per Albert's observation, there will be a considerable difference between human-rated shuttles and unmanned cargo types.

I'm semi-skeptical of the 'Rocket a Day' argument - the general principle is probably right, but the V-2 was a much less demanding vehicle even when allowance is made for 1940 vintage tech. The basic challenge of getting to orbit is minimizing vehicle mass, which forces very extreme designs, thus expensive to build and tricky (and thus expensive) to launch.

Damien Sullivan said...

NSF budget is $6 billion. NIH, looks like $30 billion. I've got $70 billion stuck in my head, maybe total US research spending? Quarter to half a percent of US GDP (14,000 billion), over very many projects, most aimed at protecting our health or defense (DARPA, weapons).

1% for for sending a few people into space for no clear purpose?

Tangentially, it occurred to me that wishful space economics tend to be really dependent on economic inequality. Altruistic billionaires to fund stuff, multi-millionaires tourists to take ridiculously expensive sub-orbital hops. Instead of working hard (and having good genes) to be a NASA astronaut, you go to space by being filthy rich. What an inspiring dream.

Rick said...

My 1 percent figure was from a different benchmark, the federal budget, not GDP. It is also just plain wrong.

It was a fair approximation from the mid 70s to the turn of the century, when NASA's budget ran between 0.8 and 1.0 percent - since then, NASA's budget has increased, but the federal budget has increased faster, and NASA now only gets a shade over 0.5 percent.

Just for the record, Clinton's last budget was $1.9 trillion (and running a surplus). Bush's last budget was $3.1 trillion; Obama's 2011 budget is $3.8 trillion. I don't want to belabor all this, but can't resist observing that Republicans suddenly discovered the evils of deficits only when Dems were in charge.

Back to the point, ahem, constant dollar (2007) spending on NASA has been in the $10-20 billion range since Apollo, a fair metric of what the American public is willing to spend on space through the political process.

I think they've gotten many times their money's worth, since without planetary exploration we wouldn't have a clue about greenhouse gases and climate change. How we act on it is yet to be determined, but knowing about it is a crucial first step.

Interesting observation about 'wishful space economics' and its implicit assumptions. I also find those assumptions a bit problematic.

But there's a deal with the devil element here. Given the existence of multimillionaires (which even my most leftward impulses would not abolish, merely constrain somewhat), them blowing their money on space tourism may at least yield some technical and operational streamlining. Whereas blowing their money on real estate just yields McMansions.

Oddly enough, if there does come to be a substantial human presence in space, the culture will probably be far from rugged individualism. Space habs will tend to be a very 'urban' environment, and with a very strong sense of mutual interdependence.

Albert said...

Instead of working hard (and having good genes) to be a NASA astronaut, you go to space by being filthy rich. What an inspiring dream.

Well, that money must come from somewhere. They may have worked hard to earn it. Or less hard to steal it. But have worked nontheless.

And if they just eredited your fortunes, you just got lucky as the astronaut with the good genes.


Anonymous said...

Nice site, nice and easy on the eyes and great content too.