hummer-prius.jpgA study came out recently claiming to prove a Hummer has lower lifecycle energy use than a Prius. Because the result was so obviously bogus — and in sharp contradiction with every other major lifecycle analysis ever done — I didn’t spend time debunking it.

But it made it into the comments of my blog and continues to echo around the internet, and the authors keep updating and defending it. A couple of good debunking studies — by the Pacific Institute (PDF) and by Rocky Mountain Institute (PDF) — haven’t gotten much attention, according to Technorati, so let me throw in my two cents.

The study’s title is revealing: Dust to Dust: The Energy Cost of New Vehicles From Concept to Disposal, The non-technical report, from CNW Marketing Research, Inc. Yes, although lifecycle energy use is probably the most complicated kind of energy analysis you can do, this 458-page report is “non-technical” and by a market research company to boot.

Their website says the report “does not include issues of gigajuelles [sic!], kW hours or other unfriendly (to consumers) terms. Perhaps, in time, we will release our data in such technical terms. First, however, we will only look at the energy consumption cost.”

Grist thanks its sponsors. Become one.

Wouldn’t want to confuse consumers with unfriendly technical stuff like kilowatt-hours, like those annoying electric utilities do every month. No, let’s put everything in dollar terms so no one can reproduce our results. When you misspell gigajoules on your website — and have for a long time (try googling “gigajuelles”) … you aren’t the most technical bunch.

I am mocking this report because it is the most contrived and mistake-filled study I have ever seen — by far (and that’s saying a lot, since I worked for the federal government for five years). I am not certain there is an accurate calculation in the entire report. I say this without fear of contradiction, because this is also the most opaque study I have ever seen — by far. I defy anyone to figure out their methodology.

Grist thanks its sponsors. Become one.

In this post I’m just going to highlight the most inane claims — and again, they can only be treated as claims because the report omits all the underlying calculations.

Let me first give one rule of thumb. U.S. energy costs have been about 7% to 8% of GDP for most of the last two decades, but were a bit higher during the energy shocks of the ’70s, as well as in the last couple of years. The Energy Information Administration projects energy costs will be 5.3% of GDP in 2030. As a rough estimate, then, you can figure the dollar value of energy embedded in most products as 5% to 10% of their cost — and that includes all the energy consumed in the product lifecycle, such as manufacturing and shipping.

The CNW study makes the astonishing claim that the Prius has lifecycle energy costs of $3.25 per mile (of which its on-road gasoline consumption is only $0.075 [yes, 7.5 cents] , whereas a Hummer H3 has an energy cost per mile of $1.95 (of which $0.187 is direct fuel consumption). I kid you not.

This is in direct contradiction with essentially every other lifecycle study ever done, in which direct fuel consumption is typically 70% to 90% of a vehicle’s lifecycle energy use (see here). CNW finds it comprises about 2% to 10%!

This range of energy costs per mile leads to one obvious question about CNW’s results, which many people have raised. Let me quote one Roy W. Spencer — yes, the global warming denyer — from the Report’s Appendix:

Question: I was wondering, how can a car that costs the consumer, say, $20,000 new and uses around $15,000 in fuel over a 100,000 mile lifetime end up having a total energy cost of, say, $250,000 ($2.50 per mile)? (Since this is way more than the consumer has paid, … which is more like $35,000.) If $250,000 really was the true energy cost, wouldn’t a car be much more expensive to the buyer than it is now?

Duh! Or maybe, D’oh! Indeed, since, as noted, embedded lifecycle energy costs are typically 5% to 10% of a product’s total lifecycle costs, CNW’s study implies that the total ownership cost of a family car is, roughly, $2.5 to $5.0 million — or about $400,000 to $800,000 a year! You never knew you were so rich.

How could CNW possibly come up with such absurd results? I can’t say for sure since the report omits any methodology, but CNW’s unbelievable answer to Spencer contains some clues:

Roy … Excellent question and point.

If an automobile lived in a capsule, if there were no other energy requirements for supporting the infrastructure of automotive driving, you are correct. A consumer would be asked to pay literally 10s of thousands of dollars for a vehicle.

But cars live in an infrastructure including support services (oil changes, for example) and disposal industries.

That added cost per mile is borne by other industries and generate profits for those industries. For example, recycling (many) of the parts of a vehicle is highly energy intense. Fortunately, those costs are borne by secondary industries because they are willing to pay in excess of the cost for the resulting components or recycled material.

Honest to goodness that is their entire answer (see pages 335 and 336 [PDF]). I’ll bet you never knew those oil changes were so expensive and so energy-intensive — next time, get the service plan! Do they really mean to suggest that the energy consumed by servicing your car is comparable to the energy consumed running it? Yes. Seriously. In fact, they say the lifetime gasoline consumption for a Prius comes to around $8,000 but the energy cost from repair and maintenance is $22,000!

(You may wonder how CNW makes such an obvious blunder. I think the mistake they made is assuming the energy cost is equal to the entire lifetime cost of repair and maintenance, rather than, say, 5% of it. But don’t quote me on that — their methodology is unfathomable.)

The recycling claim is even more stunning. Most people would say that in a lifecycle analysis, you should reward a company for recyclable parts. That is, you should probably subtract the embedded energy in recyclable parts, since such parts avoid the need to manufacture new parts in another car. The study even notes, “Current hybrids have components that are capable of being recycled in a higher proportion of their total social energy costs than non-hybrid models.”

So does a hybrid in fact benefit from being more recyclable? Of course, not, says CNW:

In all, while the industry as a whole the cost of recycling [sic] is about $119,000 per vehicle, hybrids cost more than $140,000 per vehicle to recycle.

I know you must by now be thinking I’m making this up. No one on the face of this earth — not even the most benighted analyst desperately trying to make hybrids look bad — could possibly believe this. Let me quote further (page 235):

How can a vehicle costing $30,000 generate $140,000 in recyclables?

Remember that we are discussing energy usage, not the cost of the vehicle. Over time, for instance, the vehicle will sell on average of five times in its lifetime, each time at a portion of its original cost but generally bringing the lifetime expenditure for the vehicle into the two to four times original cost range depending on desirability and demand.

Seriously. It appears — and again, I’m trying to take this nonsense at face value — somehow the resale cost of a used car is being counted fully as part of its lifecycle energy cost under the recycling category. Even though the resale cost has no connection to energy use at all. What is scary is CNW really seems to believe this claptrap. They continue:

We are also discussing energy consumption, not costs. That $140,000 in recyclable energy costs will generate $160,000 to $220,000 in net revenue to recyclers. Additionally, the support industries to recyclers expend significant energy for the production and maintenance of necessary recycling equipment. Government agencies and those who remanufacture recycled material into other products similarly expend significant quantities of energy in support of the recycling of a single car.

The reason we include all of this into the vehicle’s social energy cost is simple: As with research and development, we cannot be certain that recycled material will ever get into secondary products.

Has your head exploded yet? What could they possibly mean, “$140,000 in recyclable energy costs will generate $160,000 to $220,000 in net revenue to recyclers”? Does the $140,000 come from the used car buyers or just the recyclers? How could a $30,000 car generate $160,000 to $220,000 in net revenue to recyclers? Who could believe such a thing?

Even their own internal logic is illogical. They are including this phantasmagorical $140,000 because “we cannot be certain that recycled material will ever get into secondary products.” This implies someone is paying that money for no reason whatsoever. They are spending $140,000 in recyclable energy costs, but then aren’t recycling. Who are these recyclers? Donald Trump?

“Government agencies … similarly expend significant quantities of energy in support of the recycling of a single car.” I feel like John McEnroe: “You cannot be serious!”

If you want to see how a highly credible lifecycle energy model — Argonne National Laboratory’s widely used GREET model — analyzes a Prius versus a Hummer, read the short Rocky Mountain Institute study (PDF).

As for CNW, the only explanation I can come up for this absurdity piled on top of absurdity is that the marketing firm is putting on an elaborate hoax, seeing how many reputable news organizations repeat these laughable numbers without bothering to check the original study to see that they have no basis whatsoever in fact. After all, no one could believe all this. Could they?