According to Matt Savinar, at http://www.lifeaftertheoilcrash.net/Index.html:

"The construction of an average car consumes the energy equivalent of approximately 20 barrels of oil, which equates to 840 gallons, of oil. Ultimately, the construction of a car will consume an amount of fossil fuels equivalent to twice the car's final weight."

So, to replace the 225 million or so automobiles in the U.S., it would require approximately 225M*20bbl = 4.5 billion barrels of oil to replace all the existing cars, or about a half-year's-worth of consumption in the U.S.

Of course, to replace all the cars in the world, it would take even more oil...

And I'm sure I'm missing some energy inputs in this calculation.

At $20,000 a piece, all these new cars would cost people in the US about $4.5 trillion. I'm sure the auto manufacturers would be into this plan...

The more batteries in a hybrid would greatly increase the amount of oil used to build a hybrid car compared to non-hybrid.
There was a link on TOD or some other forum that indicated that hybrids used, IIRC, 154% more energy then a regular car.  Sorry, I don't have a link at my finger tips to support this.  In any case, it isn't a 1:1 situation.
I'm sorry you only heard the (marketing) survey, and not the rebuttal!!!

Those hybrid critics stacked the deck. They claimed, quite arbitrarily I thought, that a "car" lasts 100,000 miles while a "truck" lasts 250,000 miles. Those convenient assumptions lead to calculations showing a lower per-mile energy costs for a Hummer H3 than for a Toyota Prius.

Someone happened to report the real numbers:

The improvements are helping cars' longevity. In 1977, half of all U.S. passenger cars lasted until they were 10.5 years old, according to National Highway Traffic Safety Administration estimates. Their travel lifetime was 107,000 miles. By 2001 -- the latest year tallied -- median longevity was 13 years for passenger cars and their travel lifetime was up to 152,000 miles.

For light trucks, the mileage rose from 128,000 to 180,000, reports NHTSA, but longevity remained 14 years, largely because more trucks were being used like cars.

For more on the way they worked that number, look here.

To name just one other funny numeric business:

The study includes the energy put into research and development, which Art said is much higher for the hybrid than it is for the ICE. I'd like to see these numbers though. There is still research and development work going into the ICE.

So, you've got a few hundred million conventional cars on the road, their R&D all amortized ... what happens when you force an R&D accounting on any new technology?  Fewer units to divide by, and higher "costs" ... even if they'd really be paid over time by higher production.

If we weren't avoiding bad words today ....

When we've only got a 4% per year auto retirement rate, I don't think we need to worry about anybody putting forward a serious plan to increase that to 100%.  I agree that no one is going to expand the auto production by 20x, just to abandon it a year later.

So the 4% is good news and bad news.  It makes "electric cars" (as some fraction of the replacement fleet) more possible, but it also makes them one of the "silver bbs" and not the "silver bullet"

4% per year retirement rate??????????

Does this mean cars on U.S. roads average 25 years before going to the great death assembleges of vehicles known as junkyards?

Sorry, I should have given a link the first time:

http://www.greencarcongress.com/2006/02/us_vehicle_flee.html

Just means cars are lasting longer than ever. They don't make them like they used to (thank goodness).
4% retirement rate for the cars??? That would mean over 20 years for a car's average lifespan! I sure don't think so. You will be awful lucky to get 20 years of a lifespan out of a car. Even ships and planes are pushing the envelope at 20 years old. ValuJet bought up old Boeing 737s which were 30 years old. And we all know about what crocodiles in Florida call "airline food". Not good!

Given how I like to affect an Aussie accent, I liked this variation of the old ValuJet joke:

Q: "A ValuJet plane tried to make a mission to Sydney (Australia) but they missed! What did one croc say to the other?"

A: "Put another Yank on the barbie, mate!"

I don't know man, follow the link and see what you thing.

I do know that the older a car is, the fewer miles it is driven.  There is an inverse age/VMT correlation.  I suppose old cars tend to sit there in a multi-car family, or in the garage of retirees.  They become the "extra cars" but aren't scrapped as long as they keep up registration.

My buddy Walter Hiatt died in that crash.  He was a damned fine songwriter.
Probably the best thing that we can hope for here in the U.S. for the time being is to do what is possible to encourage those who drive large, fuel inefficient vehicles is to a) drive them less, and/or b) increase the number of people or amount of cargo in the vehicle when it is being driven.

My recollection of the '70s is that the number of passengers per vehicle increased with more carpooling, ride sharing and hitchiking.  However, my perspective could be warped by the fact that I was a student throughout that period.

Seriously, though, a medium-sized SUV usually sits 4 comfortably.  With one person driving, the vehicle may get something like 18 mpg and 18 passenger mpg.  Put 4 people in the vehicle, and you have 72 passenger mpg or about the same as two people each driving his or her own Corolla.

If we couls utilize our present SUV  and mini-van fleet similarly, we'd save quite a bit of embodied energy.  However, I think that gasoline prices would have to go much higher before too many folks would be all that interested.

I've commented on this before, but it's good to do again.  Think about the likelyhood of finding 4 people going to the same place at the same time.  Then think about the likelyhood of finding 2 people going to the same place at the same time.  It's probably exponentially more likely to find 2 people going in the same direction at the same time than it is for 4 people.
Note that it's the energy equivalent of approximately 20 bareels of oil. A lot of that energy will come from coal used in smelting the metals used in the car and to produce the electricity used to drive the manufacturing process. It's not all oil.
I would even suggest that most of it is not oil. I can imagine some amounts of oil going into transporting the vehicle and the parts, but too far from the 4 barrels mention.

That part of the Matt Savinars representation of PO I found a little bit biased.

Dear all
During the last 10 years, a number of Life cycle analysis (LCA)"cradle to grave" have been made on different aspects on transport. The best of these take into account most of the objections I have seen on the Drum. So no need to guess- but rather critizise.
One of the better LCA's is this , made by the VW. It is in german- but the numbers are self explaining.
http://www.volkswagen-umwelt.de/_download/sachbilanz_golf_a4_deutsch.pdf

The Functional unit (the  is 1 car driven 150.000 km (93.300 miles) The Primary energy ( energy at the source- oil well, coal mine, iron extraction- including end- of life- that is scrapping and recycling)-  cost for 150.000 km is between 70 to 150 MWh. This value is cradle to grave- that is raw material extractiuon and production, production of the vehicle, use + maintenance of the vehichle 150.000 km and scrapping and recycling. As a rule of thumb 1 kilo car cost presently 4 kilo +/- 0.5 kilo oil to produce.