Maryland’s Governor-Elect Larry Hogan has promised to cancel the Purple Line, another low-capacity rail boondoggle in suburban Washington DC that would cost taxpayers at least $2.4 billion to build and much more to operate and maintain. The initial projections for the line were that it would carry so few passengers that the Federal Transit Administration wouldn’t even fund it under the rules then in place. Obama has since changed those rules, but not to take any chances, Maryland’s current governor, Martin O’Malley, hired Parsons Brinckerhoff with the explicit goal of boosting ridership estimates to make it a fundable project.

I first looked at the Purple Line in April 2013, when the draft EIS (written by a team led by Parsons Brinckerhoff) was out projecting the line would carry more than 36,000 trips each weekday in 2030. This is far more than the 23,000 trips per weekday carried by the average light-rail line in the country in 2012. Despite this optimistic projection, the DEIS revealed that the rail project would both increase congestion and use more energy than all the cars it took off the road (though to find the congestion result you had to read the accompanying traffic analysis technical report, pp. 4–1 and 4–2).

A few months after I made these points in a blog post and various public presentations, Maryland published Parsons Brinckerhoff’s final EIS, which made an even more optimistic ridership projection: 46,000 riders per day in 2030, 28 percent more than in the draft. If measured by trips per station or mile of rail line, only the light-rail systems in Boston and Los Angeles carry more riders than the FEIS projected for the purple line.

Considering the huge demographic differences between Boston, Los Angeles, and Montgomery County, Maryland, it isn’t credible to think that the Purple Line’s performance will approach Boston and L.A. rail lines. First, urban Suffolk County (Boston) has 12,600 people per square mile and urban Los Angeles County has 6,900 people per square mile, both far more than urban Montgomery County’s 3,500 people per square mile.

However, it is not population densities but job densities that really make transit successful. Boston’s downtown, the destination of most of its light-rail (Green Line) trips, has 243,000 jobs. Los Angeles’s downtown, which is at the end of all but one of its light-rail lines, has 137,000 downtown jobs. LA’s Green Line doesn’t go downtown, but it serves LA Airport, which has and is surrounded by 135,000 jobs.

Montgomery County, where the Purple Line will go, really no major job centers. The closest is the University of Maryland which has about 46,000 jobs and students, a small fraction of the LA and Boston job centers. Though the university is on the proposed Purple Line, the campus covers 1,250 acres, which means many students and employees will not work or have classes within easy walking distance of the rail stations. Thus, the ridership projections for the Purple Line are not credible.

In terms of distribution of jobs and people, Montgomery County is more like San Jose than Boston or Los Angeles. San Jose has three light-rail lines, all of which together carry fewer than 35,000 riders per day, less than was projected by the DEIS for the Purple line.

Given the FEIS’s higher ridership numbers, it’s not surprising that it reported that the line will save energy and reduce congestion, the opposite of the DEIS findings. However, a close look reveals that, even at the higher ridership numbers, these conclusions are suspect.

The traffic analysis for the DEIS estimated the average speeds of auto traffic in 2030 with and without the Purple Line. Without the line, speeds would average 24.5 mph; with they line, they would average 24.4 mph. Multiplied by the large number of travelers in the area and this meant the line would waste 13 million hours of people’s time per year.

The traffic analysis for the FEIS made no attempt to estimate average speeds. Instead, it focused on looking at the level of service (LOS)–a letter grade from A to F–at various intersections affected by the rail line. Without the line, by 2040 15 intersections in the morning and 16 in the afternoon would be degraded to LOS F. With the line, only 8 in the morning and 15 in the afternoon would be LOS F (p. 30). So that makes it appear that the rail line is reducing congestion.

A careful reading reveals this isn’t true. For the no-build alternative, planners assumed that absolutely nothing would be done to relieve congestion. For the rail alternative, planners assumed that various mitigation measures would be applied “to allow the intersections to operate in the most efficient conditions.” It seems likely that these mitigation measures, not the rail line, are the reasons why the preferred alternative has fewer intersections at LOS F.

Meanwhile, the energy analysis contains two serious flaws. First, it assumes that cars in 2040 will use the same energy per mile as cars in 2010. In fact, given the latest fuel-economy standards, the average car on the road in 2040 will use less than half the energy of the average car in 2010.

Even more serious, the final EIS assumed that each kilowatt hour of electricity needed to power the rail line required 3,412 BTUs of energy (calculated by dividing BTUs by KWhs in table 4–41 on page 4–142). While one KWh is equal to 3,412 BTUs, due to energy losses in generation and transmission, it takes 10,339 BTUs of energy to generate and transmit that KWh to the railhead (see page A‑18 of the Department of Energy’s Transportation Energy Data Book). This is such a rookie mistake that Parsons Brinckerhoff’s experts would have had to work hard looking the other way for it to slip through. In any case, after correcting both these errors, the rail line ends up using more energy than the cars it take off the road, just as the DEIS found.

In short, Maryland’s ridership projections for the Purple Line are extremely optimistic, but even if they turned out to be correct, the Purple Line would still increase both traffic congestion and energy consumption. There is no valid reason for funding this turkey, and Governor-elect Hogan should chop off its head.