Bus rapid transit (BRT) vs. light rail (LRT)

This is in reply to a post in a newsgroup, however, the text is self-contained and does not require any knowledge of the previous discussion.

> The real competition for most light rail projects in the United States
> comes from Bus Rapid Transit (BRT).  BRT is now being heavily promoted by
> the Federal Transit Administration.

I don't think these are real contenders.  It is the Bush Administration who wants you to believe they are in the same class and thus the cheaper option (BRT) needs to be chosen.  Nevertheless, the fact of life is that any rail mode typically draws more new ridership than BRT.  A bus is a bus, and it is very difficult to make people commuting by car switch to a bus.  On the other hand, history knows multiple examples when a new rail line had two, three or five times the ridership of the bus that existed on the same route before the initiation of the rail.  This fact is even documented in the textbooks on public transportation: general public more willingly selects rail as the mode of choice and almost never selects bus.  Bush Administration promotes BRT because it is cheaper, and NOT because it can make more difference to relieve the congestion.

Technically, the main advantage of the light rail over BRT is that light rail trains can be run with multiple units.  If one articulated 100-feet (30-meter) car is not enough, you couple two of them and run two-car trains.  If two cars are not enough, you couple three of them and run three-car trains.  Nothing of this kind is easily accomplished with BRT.  A single articulated bus with a single driver will never be able to carry as many riders as 300-feet (100-meter) light-rail train with the same single driver.  Thus, by design, BRT has LOWER capacity than light rail, just in principle.  BRT is designed for lower ridership corridors, and for relatively high-ridership corridors is not appropriate at all.  Actually, the "success" of the Orange line in Los Angeles (see below) is an example of the severe limitation of BRT: buses run full and there is no easy way to improve the situation.  The poor decision to replace LRT with BRT is already harming the ridership.  No new growth can be really expected when buses are at capacity.

Another technical issue is that you need a WIDER lane for the bus than for the LRT, because the bus is not guided by rails.  In particular, running buses in the tunnels designed for rail is a nightmare.  One can go to Seattle and see how slowly their buses operate in the center-city tunnel compared to any underground rail lines running in the same-diameter tunnels.  Or one can go to Boston and see how wide their tunnels (designed specifically for buses) are. This is exactly because navigating a bus requires a wider dynamical corridor.

Generally, given the capacity constraints of the bus (no multi-unit trains) and the high construction costs of the tunnels (the highest of all alignments), it is highly impractical and wasteful to build bus-only tunnels.  They are the highest in cost (because they need to be wider than rail ones) but have the lowest capacity (because of the inability to operate multiple-unit sets) of all possible grade-separated right-of-ways.  The tunnel portion of the Boston Silver line provides the best illustration of all the disadvantages of a bus in a tunnel: low speed (15 mph), poor operation reliability, extra expense of dual-mode (diesel-electric) vehicles, and high cost with low capacity (airport buses are running full a year after the opening).

Finally, don't forget, that the comfort of the bus ride is substantially lower than the smooth rider on any modern rail mode.  Any person with back problems can attest that ANY bus (even on a busway) provides lower ride comfort than modern rail, and thus inferior in quality to rail.

> There has been much publicity from the success of the Orange Line BRT
> in Los Angeles, which now has ridership nearly three times the initial
> projection.

As a person who lives near Los Angeles and closely watches its transit developments, I will tell you a bit about the Orange line.  Did you know how bad and unreliable MTA bus system is in general?  Having ridden on virtually all large bus systems in the U.S., Los Angeles MTA is not among my favorite.  Even if buses run jam packed, the MTA often does not bother (they say they don't have money) to run the route more frequently than once in 30-60 minutes.  And even on the routes where buses are officially supposed to run every 10-15 minutes, very often (much more often than in other cities) one has to wait for 30 minutes to witness then two buses coming together.  It is not uncommon to spend 60 minutes on the bus stop of a route that "runs" every 20 minutes.  Even waiting for the most frequent bus route in Los Angeles - Metro Rapid 720 - which officially runs every 6 minutes - one can often spend more than 20 minutes on the bus stop because buses run completely full and just do not stop to pick up more riders.

Now, imagine in this quite bad system you build SOMETHING better, something that runs every 10-15 minutes if scheduled to run every 11-12 minutes (not too bad by Los Angeles standards).  Naturally, this SOMETHING is going to draw a lot of people from the parallel routes, many of which run every 25-60 minutes most of the day.  There are so few frequently operating routes in San Fernando Valley, that the new MTA map "Go Metro without Timetables - 12 minutes wait or less" shows only four (!) frequent bus routes in the entire 2-million-people Valley.  One of these four routes is the Orange line, and another two (750 and 94/394) actually run every 15 minutes and thus do not belong to this map. There is huge demand for service (both potential and existing), and there is very little service actually in place.  ANY solution, ANY mode that is even remotely reliable is going to be successful.

Returning to the projections.  You say ridership is nearly three times the initial projection.  Did you know that the initial projection for BRT was one SIXTH of the one for light rail?  They projected something like 40 thousand a day for light rail, and 20 thousand for BRT.  Then, looking at the Gold line and its poor performance, the reduced initial projection for BRT to 7 thousand a day or so.  Now, the line indeed carries close to 20 thousand per day and is at capacity.  So, looking at official propaganda is not the best way to say how successful a line is.

To summarize: Orange line can hardly be called a "success".  It is at capacity, but still carries half the passengers that could have been carried by the light rail line along the same route.  And the capacity of the light rail line could have been always expanded by running longer trains.  It was a bad decision to replace the LRT along the Orange line corridor with BRT.  As usual, this decision was heavily influenced by politicians, by the way, exactly the same ones who stopped Red line expansion west along Wilshire Boulevard.

Moreover, Los Angeles area has another example of the BRT - Harbor Transitway.  This 6-station monstrosity cost one billion dollars to build. But ridership is so low nowadays, that buses run only once in 30 minutes most of the day.  This compares highly unfavorably to the Blue line that runs two miles east of the Harbor Transit way and has three-car trains running full every 12 minutes most of the day.  This is actually a much better example of how BRT scores compared to the LRT: two parallel lines, Harbor Transitway and the Blue line.  The latter transports probably 10-20 times more passengers per day.  It also demonstrates what riders choose when provided with both options at the same time. (It should be noted that a part of the reason for the low ridership on Harbor Transitway is the very inconsistent routing of various Transitway bus lines in the Downtown area, which is exclusively due to the incompetence of the operating agency.)

> The larger size of light rail trains has interesting consequences.  In Los
> Angeles, the Gold Line LRT and the Orange Line BRT presently carry similar
> passenger loads.  The Gold Line typically operates 2-car trains on 15
> minute headways.  To achieve similar capacity, the Orange Line operates 60'
> articulated buses on 5 minute headways.  The shorter station wait times
> result in greater convenience for the riders of the Orange Line BRT.
> The longer station wait times must be a factor in the Gold Line LRT
> under-performing relative to initial projections.

The facts above are correct.  However, you miss one very important issue.  You look at the CURRENT state of the Gold and Orange lines, and you completely ignore the fact that the system is expanding.  The Orange line cannot be extended substantially, since it is already at capacity and there is no way you can substantially improve the capacity with buses.  Now, for the Gold line, a 2-car train in 15 minutes is good enough NOW.  However, in 2009 we'll get an Eastside extension open, and the ridership will soar.  There are also extensive plans for extending the Gold  line further east from Sierra Madre Villa along the Foothill Freeway (I-210); the line is ultimately planned to reach Montclair in San Bernardino County.  As you can imagine, at that time the line will become the longest on the Metrorail system, and you will need much more than a 2-car train in 15 minutes to carry the loads.

The main point is:  the Gold line can (and will) grow, and the Orange line has no future.

> My own thoughts on the matter were strongly influenced by a recent visit
> to San Jose, where I rode the light rail line from Winchester to Mountain
> View. ...
> Of course, had the San Jose system been BRT rather than light rail, I
> wouldn't have bothered to ride it.  A train is more interesting than a
> bus.  I am not sure that this would be a serious consideration for a daily
> commuter.

Most studies show that this is true for a daily commuter as well.  Rail is generally more attractive for a Joe Doe than just a bus. The fact is that rail draws people from cars, and buses almost don't. By the way, it is also known, that proximity to a rail station substantially increases property values.  Proximity to a bus stop does almost nothing.

Last update: October 15, 2006.