Analysis of CNW: sensitivity-thresholds
 

Hi folks,

Recently, someone posted an open question about the energy needed to make a hybrid electric. Quickly, several of us posted a pointer to the Carnegie Melon study but something kept bothering me:

As I thought about the Carnegie Mellon study it occurred to me that their summary data was enough to do a "what if" analysis of a 1990 Ford Tarus, the study vehicle, to a simulated Toyota Camry and Hummer-like vehicle. The Carnegie Mellon summary data could be put in a spreadsheet with "CNW adjustment factors" for:

(1) Manufacturing cost
(2) Vehicle life

This allows us to test "CNW fudge factors" to see if we could figure out what was needed for their claims. This is important because it takes work to 'adjust' real data to give the illusion of 'truthiness' (see Wikipedia.)

To give the same energy cost per mile as the Ford, I found:

A) Holding manufacturing costs constant, requires the hybrid life to go down by ~64% while Hummer-like lifetime must go up 170%

Folks were quick to spot the mileage discrepancy in the "Dust-to-Dust" report but the worst was the "Land Cruiser" versus every other car. It had nearly twice the 'lifetime' of any other car to the point of being way too obvious:
http://hiwaay.net/%7Ebzwilson/prius/life_time_cost.jpg

Within the same or similar models, we always found the hybrid with a shorter life even though it was more economical to operate. CNW had to invent the paradox that the same model hybrid would not be used as much as the gas-only vehicle, even if owned by the same family by the false claim that driving a hybrid is done to make a "style" statement.

B) Holding vehicle life constant, requires the hybrid manufacturing costs to go up by 450% and even if built for free, the Hummer-like vehicle would never come close

Folks also quickly noticed the inflated "design" costs as well as "recycling" costs. But the problem is even if you build a Hummer-like vehicle for free, you still have to make it Superman indestructible to compete.

To no surprise, there is evidence of both "adjustment" factors: cost of manufacture and vehicle life to get the targeted cost per mile to where it would have "truthiness." (a term borrowed from Stephen Colbert describing how to act like someone who knows something while being totally full of fecal material.)

So where did CNW Marketing get the detailed spreadsheets referenced in the "Dust-to-Dust" report? I suspect a copy of an actual life-time study but with the full data tables. Upon figuring out what numbers had to change by what 'adjustment' factors, the massive spreadsheets were released as backup to the "Dust-to-Dust" conclusions.

When I reviewed the "Dust-to-Dust" spreadsheets, I was always bothered that some of the 'calculations' were omitted. This is not how a real engineering paper works!

Now if I were a University professor concerned about sources or the author of a legitimate life-time study, I would compare the row elements of the CNW "Dust-to-Dust" report with published papers on vehicle life-time studies. Upon finding the rows looking a lot alike, I would then see if the data in the columns looked 'familiar'.

All I've done is take the summary data from a legitimate lifetime study, the Carnegie Mellon report, and test the hypothesis that two simple 'adjustments' to their data could deliver "Dust-to-Dust" conclusions. Near as I can tell, that hypothesis remains viable.

Bob Wilson

Re: Analysis of CNW: sensitivity-thresholds
 

Here's my analysis of CNW: He's full of crap :D

Re: Analysis of CNW: sensitivity-thresholds
 

Bob,

What does "CNW" stand for?

Thanks in advance.
Chris

Re: Analysis of CNW: sensitivity-thresholds
 

Hi Chris,
"CNW Marketing" is a small (one man?) operation in Oregon that published a widely shared but disputed report partially titled "Dust-to-Dust." It claims to be a life-cycle analysis of hybrids versus other gas cars and comes to the conclusion that Hummers have a lower total energy cost per mile than a Prius. Google them up if you want to read about them.

We quickly noticed several problems:

1. abnormally low hybrid life-time miles, always lower than a gas equivalent
2. abnormally high, large format, SUV life-time miles, beyond believable
3. abnormally high development energy costs to the point that existing vehicles have none
4. absence of peer review, the gold standard for any credible study
5. claims that given a hybrid and non-hybrid cars, identical models, the non-hybrid would get more miles per year
6. claims that hybrids get less than 7,000 miles per year (Doug recently shot that one down using Ebay mileage records!)
7. claims that a hybrid is a "style," not a practical vehicle (many of us are putting more than 15,000 miles per year on our hybrids!)

What I was pointing out is the availability of credible, peer review studies that actually shared their data. These studies universally report the actual, life-time energy cost is 70-80% from vehicle operation and the rest comes from manufacturing and disposal.

My speculation is CNW collected the data from these peer reports and then 'adjusted' the numbers. This would give the illusion of research with credible categories but the numbers would be illusionary. But it is easier to simply state the CNW Marketing is full of it.

Bob Wilson