Department of Energy Gets Basic Math Wrong in its Rail Analysis

When it comes to the carbon consumption of cars, trains, and buses, the U.S. Department of Energy’s (DoE) Transportation Energy Data Book [PDF] is an indispensable resource. But this year’s Data Book contains an eyebrow-raising error in its analysis of rail’s energy use.

Edition28.jpg(Image: DoE)

Page 66 of the Data Book, reprinted on the DoE’s website on Inauguration Day, contains a table ranking the energy intensity of various light rail systems across the country.

The DoE lists the "average" energy efficiency of all light rail systems as 7,605 Btus per passenger mile, while the average for cars was 3,514 Btus per passenger mile.

Those numbers were enough to spark inflammatory headlines about the energy consumption of light rail. The only problem: The rail data is wrong.

An eagle-eyed Streetsblog Capitol Hill reader discovered that the DoE used simple averaging to obtain its light rail number, without weighting each city’s light rail network based on how many passengers it carries.

So Kenosha’s streetcars, which carry a bit more than 60,000 passengers annually, were treated the same as Seattle’s light rail, where ridership is exceeding 60,000 every week.

Even famously anti-transit Randal O’Toole recognized the DoE’s error and pointed out the actual average energy efficiency for light rail is 3,642 Btus per passenger mile — comparable with the numbers for cars, which don’t fully account for the choice of auto driven.

The same averaging error is made on page 67 of the Data Book, which states that the "average" energy efficiency of heavy rail is more than 3,600 Btus per passenger mile. That average put Cleveland’s energy-chugging system, which carry about 30,000 passengers on an average weekday, on equal footing with the New York City subway, where the average weekday ridership tops 7 million.

When the Streetsblog reader contacted the DoE to inform them of the error, he got a quick acknowledgement and a promise to correct the data as soon as possible. The incorrect averaging should never have been used, the DoE said.

One wonders how many misleading commentaries transit critics can publish using the false data before the government corrects it.

  • garyg

    Elana is confused. The relevant average depends on the context. The average energy use per passenger-mile for all light rail systems combined is dominated by the efficiencies of a small number of systems in big cities with high ridership (Boston, Los Angeles, Philadelphia, etc.). It is therefore highly misleading to cite this average in support of an expansion of light rail in smaller cities, where ridership is likely to be much lower than it is in big cities. The energy efficiency of light rail systems in smaller cities is more likely to resemble that of the average calculated by system, not the average calculated by aggregating all the systems together.

    And as Elana herself points out, even the aggregate average is no better or virtually no better than passenger cars.

  • Graham

    So what is going on with Little Rock, Kenosha, Memphis and Galveston Island? These systems use way more energy on a per passenger-mile basis than the other systems – something must be going on.

    And indeed something is: These massively inefficient “systems” are all heritage streetcar lines that run early 20th century cars (some made of wood!) on short (2 to 7 mile) loops. They are designed to circulate pedestrians and tourists in historical areas, and are more tourist attraction than light-rail transit.

    Real Light rail transit systems – designed to get people to and from work – are significantly more efficient, with little correlation between city size and the efficiency of the system: Portland, Salt Lake and Minneapolis all have systems whose efficiency matches those of larger cities. Looking at this list we see that energy efficiency correlates roughly with the size of the system. (There are outliers – poor Baltimore! – and this speaks volumes to the importance of design and planning.)

    Locally, we can take heart in the fact that the new 16 mile Purple Line – with an estimated 68,000 daily riders – will be as energy efficient as the best of them.

    (And garyg: Elana is not advocating anything here; just reporting an error. She is certainly not confused. You tire with your ad hominem swipes.)

  • garyg

    Looking at this list we see that energy efficiency correlates roughly with the size of the system.

    Yes, and not surprisingly, the big systems tend to be in big cities. That’s why I said the aggregate efficiency is dominated by a small number of systems in big cities with high ridership. But you’ve run out of big cities in which to build new systems. Phoenix was really the last big city to get light rail. Expanding light rail now means either building new systems in smaller cities, or adding routes to existing systems. In both cases, the efficiency of the new routes is likely to be lower than the aggregate efficiency of existing routes. The low-hanging fruit has already been picked.

    And garyg: Elana is not advocating anything here; just reporting an error. She is certainly not confused.

    You’re both confused. She’s not reporting an error. She’s making a false claim of error. It is not an “error” to calculate the average efficiency of light rail by system. For the reasons I have explained, aggregating all the systems together, as if they were one big system, is not a useful way of estimating the efficiency of any expansion of light rail.

    And as usual when transit proponents discuss the supposed environmental benefits of light rail, you miss the big picture. Even if light rail were much more energy than motor vehicles, it’s simply too small a component of our total transportation system to have any meaningful effect on energy consumption or pollution. After 40 years of growth, light rail accounts for only about 4% of transit trips, and a negligible share of total motorized trips. It’s simply too small to matter. And no conceivable expansion of light rail in coming decades will change that.

  • Graham

    Re: “She’s making a false claim of error.”

    Note: “The incorrect averaging should never have been used, the DoE said.”

  • garyg

    You mean the unidentified Streetsblog reader told Elana the DoE said that. It would be interesting see an exact quote of what the DoE actually said in regard to the average reported in the Energy Data Book.

ALSO ON STREETSBLOG

Americans Still Use a Lot of Gas

|
The release of the Department of Energy’s Transportation Energy Data Book is a transportation stat geek’s dream — 300-plus pages of numbers detailing the way the country burns this or that moving people and freight from city to city. Of course, not everyone gets a thrill from poring through data tables for hours at a […]

Separating Myth from Fact on “Cash for Clunkers”

|
As debate rages on in the capital over whether to keep assisting the auto industry by giving out more "cash for clunkers" rebates, two assertions are becoming commonplace: the program is helping diminish U.S. oil consumption, and the program is not paid for with new money. (Photo: AP) The first argument was reiterated on Friday […]

Ed Glaeser’s Rail Fail

|
The story so far: Ed Glaeser recently began an effort to assess the costs and benefits of constructing high-speed rail lines at the New York Times’ Economix blog. Last week, he posted his first substantive take on the issue, an attempt to estimate direct costs and benefits from a hypothetical line between Houston and Dallas. […]

Seattle Cyclists Find Safety in Numbers

|
Here’s what’s happening around the Network today: Safer Cycling in Seattle, and More Of It: Cycling is increasing in Seattle, but traffic collisions involving cyclists are not. That’s according to census data complied by a local newspaper and reported by Network blog Seattle Bike Blog. In the last three years, the cycling rate has grown […]