When I lived in Seattle from 1992 to 2002, I was fascinated by the Ship Canal Bridge that carries I-5 across the canal connecting Lake Washington to Lake Union and from there to Puget Sound.
On my bike, I often wished there were a bike lane somehow suspended from the bridge, so that when I was traveling between the University District and Capitol Hill, I wouldn't have to go all the way down to water level, cross the canal on University Bridge, then make my way back up the other side.
(Well, not
all the way down to the water; it's not like I had to swim. And I didn't try any
Evel Knieval stunts going over it when it was open like this.)
But that notion of a bike lane in the sky was just a wistful dream. What really fascinated me about the bridge was the way it illustrated the interface between economics and engineering.
On each shore of the canal, the approach to the bridge is made out of ferro-concrete, and is supported by relatively closely spaced piers.
For the next section, the structure becomes steel, the spans become longer, and the piers become taller, as the land falls away toward the water and the bridge rises slightly on its way to the necessary high clearance over the canal.
Finally, in the middle of the bridge you get the longest, highest span--it
has to be high to let sailboats under it, and it
has to be wide to keep the bridge piers from getting in the way of the navigable waters.
The steel has a better strength-to-weight ratio, so it can reach across longer spans than the ferroconcrete. But concrete is cheaper than steel, so the ferroconcrete by itself is cheaper than the steel.
For the middle of the bridge where you need a really long span, you use the steel. For the approaches, where the ground isn't that far below you and you can use short piers, you use ferroconcrete.
But the place where you really see the economics is in the part between the approaches and the central span. You don't
have to have your piers far apart, because you're still over land. But the piers are getting so tall that they're getting expensive. When you look at the bridge as a whole, you can see the point where the savings from having fewer tall piers outweighs the cost of building the spans out of a more expensive material.
Of course, this is probably oversimplifying. The other advantage of fewer piers is that you don't have to sacrifice as much stuff on the ground. There's a lot of space on the ground between those high piers to carry on normal activity (unless of course you're a plant, in which case the reduction in sunlight and rainfall is a bit of a problem).
But I suspect that the cost tradeoff of fewer tall piers vs. more expensive material played a role in the bridge's design.