I've been following a discussion among professional ecologists concerning shifting from older cars to newer cars. The general argument favoring new cars is that they are more efficient in nearly every sense and pollute less than older cars.

I think that's a good first approximation solution to both energy conservation and pollution reduction. But then I started wondering about a second approximation. In the second approximation I was thinking about the energy used to build a new car from ore extraction to refining for all parts (metal, plastic, leather, paint, etc), to parts production including energy for transportation and finally, energy for assembly and distribution. Then I realized you'd have to work in constant dollars (my new Datsun 710 wagon in 1974 cost about $3,000, a similar car now would cost $20,000). Also, you'd have to depreciate it. I chose a method that took into account the scrap value, initial value, and takes the first derivative as the depreciation at any given time.

Then you'd have to look at the fuel efficiency of both cars and take the difference over time to determine the intersection point of both independent curves.

Then do the same for carbon production and NOx. (And SOx if diesel).

Does that sound right? has it been done? Damned if I want to reinvent the wheel (again!).