Let’s look at what nanotech could do — could be doing now if Feynman’s path had been taken — to make space travel more achievable and affordable — and therefore useful. It’s widely understood how lighter, stronger structures can make rockets more efficient, but that’s of limited use. The rocket equation is still a huge stumbling block.

One way around it is not to use rockets at all. Nanotech would be easily able to build a space pier:

Space Pier

Space Pier

The idea of a space pier is just the same as for a pier for ordinary boats: it bridges the gap between the land and the regime the ships operate. In the case of water ships that’s water of a certain depth; in the case of spaceships it’s vacuum and orbital velocity.

Build a structure 100 kilometers tall and 300 kilometers long. Put a linear induction (or other electromagnetic) motor along the top. An elevator goes straight up 100 kilometers to the starting end. Payloads are then accelerated horizontally into orbit with an acceleration of only 10 G’s (which appropriately cushioned humans can stand for the 80 seconds required). This hybrid approach overcomes the drawbacks of both the typical orbital tower schemes (it’s less than 1% the height of a skyhook) and electrolaunch ones (air resistance at 100 km is a million times less than at sea level).

Like any other launch-track-based scheme, this has the huge advantage that you’re only accelerating the payload, instead of ten times the payload’s mass of fuel and oxidizer, to orbital velocity.

Compared to the skyhook, which is just barely possible with even the theoretical best material properties, a tower 100 km high is easy. Flawless diamond, with a compressive strength of 50 GPa, does not even need a taper at all for a 100 km tower; a 100-km column of diamond weighs 3.5 billion newtons per square meter, but can support 50 billion. Even commercially available polycrystalline synthetic diamond with advertised strengths of 5 GPa would work. Of course in practice columns would be tapered so as not to waste material; and the base of the tower would be broadened to account for transverse forces, such as the jet stream. Only the bottom 15 km (i.e. 15%) of the tower lies in the troposphere and would have to be built taking weather into account.
The electromagnetic accelerator along the top might be fairly heavy. In many designs, coils have iron cores; NASA’s Marshall prototype weighs 100 pounds per foot. If we allow a tonne per meter, the total weight of the accelerator is 300 thousand tonnes. For comparison, the Golden Gate bridge is about 800 thousand tonnes. However, most the weight is (relatively cheap) iron. Note that this entire weight, if it were concentrated in one place, could be supported by a column of currently available polycrystalline diamond less than 80 centimeters on a side.
The overall structure could be openwork like a radio tower, and could have approximately 60 footprints 10 km (30 long x 2 wide) apart on the ground — if we set aside a hectare for each foot, they only occupy 0.02% of the land under the tower. The footprint foundations would each bear the weight of a small office building, no great technical challenge.
One of the interesting challenges in designing the structure is that simple diamond columns only thick enough to support its weight would be too thin to be rigid. Two strategies which can be brought to bear are truss structures and thin shells inflated for stiffness. (Note that since I originally gave a talk about space piers at ISDC in 2005, further technical work has been done on the concept.)  At altitudes below the tropopause, some attention should be paid to reducing the structure’s wind cross section to minimize the effects of hurricanes and the jet stream. It is not inconceivable that the jet stream, or cloud-to-ground voltages, could be used as energy sources.

Another connection to the Feynman Path is that the construction of the tower would be facilitated by tele-operated scale-shifting robots, this time to larger scales, instead of having people in spacesuits crawling around 60 miles up in essentially vacuum.

The reason for the 60-mile (100 km) height is that it’s just the edge of space. The air is about a million times thinner than on the ground, but there’s still enough of it that nothing orbits at that height.  A major problem with a scheme like the skyhook (“space elevator”) is orbital junk; the pier is too low for that. (Another problem with the space elevator is the combination of slow ascent speeds and the Van Allen belts; unprotected humans would be DOA to orbit.)

The tower would be maybe a million tonnes of material in a structure 300 km long. A typical superhighway that long involves 15 million tonnes of material and costs on the order of 1 to 5 billion dollars.  Assuming the land under the tower does not lose too much value (and it might well gain it, especially near the embarkation port), the area occupied by the footprints is trivial.  The major obstacle to construction is likely to be legal hassles, not anything physical.

If the coils and electronics for the accelerator cost $1000 per meter, total for the accelerator is under half a billion. The wildcard is the cost of the diamond (and the ability to fabricate it into structural beams). Diamond is a bit expensive today! If an Apollo-style (and -cost) project could do for diamond what the original one did for electronics, we could build the tower in the next decade or so.  Depending on the level of nanotechnology, diamond might cost $100/kg like graphite for golf clubs and tennis racquets.  Diamond costs much more than that today but there is no mass production — no one is using it as a structural material.  A mature nanotechnology with self-replicating assemblers would push cost toward the cost of inputs; carbon in the form of coal can be had in quantity for as little as $0.02/kg.

Note also that diamond isn’t strictly even necessary; a number of modern materials, ranging from inflated  kevlar-polyethylene tubes to carbon-nanotube-filled aluminum would work.

A single (10-tonne) payload launch would come to about $4300 in electricity (43 cents/kg).   If we estimate maintenance and operating staff costs as similar, operating costs might be about a dollar per kg.  Note that the cost to higher orbits goes up linearly rather than exponentially as with rockets (within certain bounds).
The other major part of the cost of a launch is amortization of the cost of the tower. If we can launch once an hour (and at an interest rate of 8%) we must charge $0.91 per kg per billion dollars of tower cost. If the traffic is there the rate might be increased (and the cost reduced) by a factor of 10 but not 100. (NB: as payload size is increased, power costs go up per launch but not per kg; amortization costs go down per kg; and more kinds of things can be launched in one piece. The main limit is the accelerator. Chances are you could build one to handle 10 tonnes, but that’s a guess.)
Also, your payload needs to be a spacecraft capable of some 330 m/s delta-V to circularize its orbit.  The best strategy is to inject into an orbit whose energy is the same as the desired circular one, and do a correction at the point the orbits cross which only changes direction. This is not a Hohmann transfer orbit, which optimizes total delta-V. With the launch tower, delta-V at correction is considerably more expensive than at launch, so we minimize it preferentially. Even so, given typical chemical Isp’s, propellant for correction is some 10 to 15% of gross vehicle weight. For freight that can be a cheap one-shot strap-on solid rocket. For passengers the vehicle needs to be re-entry capable in case of accidents, and gets expensive.

One way to guage the economics of the space pier is to note that the current rocket-launched cost to orbit is $10,000/kg.  At about a dollar/kg per billion of capital cost, one might claim the tower would be economic at anything less than $10 trillion.  However, this would require a market for launches at that price which does not exist.  So it would be critical to lower the price to the point that there is a market with sufficient volume to keep the tower busy.
Molecular manufacturing, even of a fairly unsophisticated form, could make it economical. Suppose means were found to manufacture diamond and graphite fiber of 5 GPa compressive and tensile strength respectively, in quantity, at $10/kg. If the structure needed half a million tonnes of material it comes to $5 billion. Then the whole business might cost $10 billion. (For comparison, the space station is costing $20 billion and Apollo cost $24 billion in 1960’s dollars.) This gives us a total cost of about $10/kg to orbit. This is nearly cheap enough to make ground-launched powersats feasible, but vacations in orbit still cost $25,000.
With a mature nanotechnology, the cost of diamond comes down and the strength goes up. Likewise the high-tech stuff in the track, and power generation, and so forth. (An example: recent progress in micron-scale vacuum tubes suggests that nanoengineering could make high-powered switching apparatus that was more efficient and cheaper than silicon.) Cost reduction on the order of 100 seems quite feasible. At 10 cents/kg to orbit, the Solar System is our oyster.

The slides for my ISDC talk on the space pier are here.