Wednesday, January 4, 2012

MAX is slow. And other findings from a comparison of 15 rail transit lines.

One of my major takeaways from Public Transportation class in grad school was this: A successful transit system provides a careful balance of accessibility, speed, frequency, reliability and quality. Which factors are most important depends on who you ask. For someone with mobility challenges, having a short distance to a transit stop is obviously critical. For a harried downtown worker, super-quick travel times might be just the thing. For a free-wheeling teenager allergic to planning ahead, it may be the ability to hop on frequent transit without checking a schedule.



Satisfying all of these requirements is a difficult planning exercise, and some well-intentioned transit agencies emphasize one or two of these factors while sacrificing others. I’ve found that Portland’s TriMet, for example, has a tendency to bend over backward for accessibility - placing transit stops very close together and creating meandering routes that cover lots of real estate - at the expense of travel time. Try taking a MAX train from Goose Hollow to Lloyd Center - that’s a 2.8 mile trip that will take you past 14 MAX stations and occupy 25 minutes of your time. That’s an average speed of 6 mph - like running a ten-minute mile. Other agencies may do the opposite, emphasizing speed, but skipping a lot of important places. My personal preference is for fast and frequent service, even if I have to walk a bit to the nearest stop. Walking is good for you.

The end of 2011 got me thinking about the approaches to that different transit agencies take. I’ve been fortunate to revisit several North American cities and regions this year, and to use their transit offerings - Portland, Seattle, Vancouver, San Francisco, Los Angeles, Philadelphia and New Jersey among them. While I would enjoy performing an exhaustive comparison of how their transit agencies approach the five service factors mentioned above, I will spare you (for now). I’m most interested in comparisons in travel time, or more specifically, average speed. In my experiences this year, I couldn’t help but notice the stark contrast between Portland’s crawling MAX train and the speedy SkyTrain in Vancouver, or the accurately-named High Speed Line in South Jersey (which I first rode when I was half my height). So I looked into these and other transit lines during the ample down time offered by unemployment.

I compared 15 rail transit lines in the US and Canada, 10 of which I have ridden. Some are “light rail” lines, which operate both within street rights-of-way as well as on separated tracks, and are usually powered by overhead catenary wires. Others are “rapid transit” lines that have complete separation from automobile traffic via underground, elevated or fenced off trackways, and are usually powered by the famed “third rail.” (Initially this may seem like comparing apples to oranges, but it turns out there is some overlap on how these two types of rail systems perform.) All 15 are rail lines that connect a region’s central business district with outlying suburbs or secondary cities. “Heavy rail” commuter rail lines, which often have several miles between stations, are not included, nor are rapid transit lines that stay mostly within city limits, like the New York subway or Chicago El. While my list includes transit lines from coast to coast and from Canada to near Mexico, it is not comprehensive. It skews westward for selfish reasons, and leaves out remarkable systems in Boston, Atlanta and Miami, among other places.

The basis for comparison of the 15 rail transit lines is the following question: What is the train’s average (mean) speed from the most centrally-located downtown transit stop to the center of an outlying suburb near the end of the line, starting at 5:00pm on a weekday? Put another way: How far away is that beer in your fridge?

As I’m sure you know, average speed is total distance divided by total time. Times were readily available on train schedules from the 15 transit agencies. Distance was a bit more labor intensive - I used gmap-pedometer.com to click mileages along the transit lines. The latter can get tricky where tunnels are involved, but I managed to be pretty accurate. I also took note of the number of station stops between downtown and the chosen suburb, as well as how many stops are in the downtown area. (Having lots of downtown stops slows things considerably, as the trains cannot get up to speed, especially at grade on city streets).

My results confirmed some suspicions, but also brought some surprises. Below are the 15 rail transit lines listed in order from fastest to slowest, with the station pairs studied, transit type, average speed, distance covered, total travel time, number of station stops after boarding, and some caveats and/or further comments. Also keep in mind that while these speeds may sound slow, they are averages that include the time slowing, stopping and accelerating at numerous stations. Most of these trains reach speeds of 60-80 mph.

1. BART Yellow Line: Montgomery station (San Francisco, CA) to Walnut Creek station
Type: Rapid transit
Average speed: 40.0 mph
Distance: 22.7 miles
Travel time: 34 minutes
Station stops: 9 (1 in central city)
Comments: The winning line’s time advantages include the lengthy Trans-Bay Tube, where the train reaches 80 mph underneath San Francisco Bay, another tunnel through the Oakland hills, and a small number of stops relative to its over-20-mile length.


2. PATCO High Speed Line: 15th/16th St station (Philadelphia, PA) to Lindenwold station (NJ)
Type: Rapid transit
Average speed: 33.0 mph
Distance: 14.3 miles
Travel time: 26 minutes
Station stops: 12 (3 in central city)
Comments: Like BART, the High Speed Line also benefits from crossing a large body of water: the Delaware River. But it’s also quick on land, sailing over and under towns on elevated and trenched tracks.


3. DART Red Line: Akard station (Dallas, TX) to Downtown Plano station
Type: Light rail
Average speed: 29.4 mph
Distance: 18.1 miles
Travel time: 37 minutes
Station stops: 14 (2 in central city)
Comments: Per my earlier apples and oranges statement, here is a case of a light rail line competing with our nation’s best rapid transit lines from a speed standpoint. I guess Texans like it fast. While computing distances, I noticed that the stations are further apart than many other light rail lines - a key influencing factor.

4. TransLink SkyTrain Expo Line: Granville station (Vancouver, BC) to Surrey Central station
Type: Rapid transit
Average speed: 28.8 mph
Distance: 16.3 miles
Travel time: 34 minutes
Station stops: 16 (2 in central city)
Comments: This mostly-elevated line is automated - no driver required. For a region of just over 2 million people, Vancouver has the transit of a first-tier city. It’s land use policies have also resulted in very dense transit-oriented development along the line.


5. UTA TRAX Blue Line: City Center station (Salt Lake City, UT) to Historic Sandy station
Type: Light rail
Average speed: 25.0 mph
Distance: 12.9 miles
Travel time: 31 minutes
Station stops: 13 (2 in central city)
Comments: Here’s another surprising performer in the light rail division. TRAX’s speed may be courtesy of flat, straight trackage.

6. RTD Green Line: 16th & Stout station (Denver, CO) to Littleton Downtown station
Type: Light rail
Average speed: 24.5 mph
Distance: 10.2 miles
Travel time: 25 minutes
Station stops: 9 (2 in central city)
Comments: Yet another light rail line beating out rapid transit lines. A low number of stations, ample distance between some of those stations, and long straightaways may be contributing factors.

7 (Tie). Sound Transit Central Link: University station (Seattle, WA) to Tukwila Park & Ride
Type: Light rail
Average speed: 24.0 mph
Distance: 13.2 miles
Travel time: 33 minutes
Station stops: 10 (4 in central city)
Comments: This line - our fourth and final light rail line to beat out the rapids - was completed in 2009. Downtown, the line travels through Seattle’s unique transit tunnel, mixed with buses. South of here, it is on elevated track, in the Beacon Hill tunnel, or in the median of MLK boulevard.


7 (Tie). PATH: World Trade Center station (New York, NY) to Newark Penn Station (NJ)
Type: Rapid transit
Average speed: 24.0 mph
Distance: 8.8 miles
Travel time: 22 minutes
Station stops: 5
Comments: I am surprised this line is not faster, given its small number of stops and long uninterrupted stretches under the Hudson and through the Meadowlands. Perhaps it suffers from old age.

9. WMATA Metro Red Line: Judiciary Square station (Washington, DC) to Rockville station (MD)
Type: Rapid transit
Average speed: 22.8 mph
Distance: 12.9 miles
Travel time: 34 minutes
Station stops: 15 (3 in central city)
Comments: This is the slowest rapid transit line in our selection, though it still gets you from the heart of DC to beyond the beltway in a half hour (try that in a car at rush hour). There are more destinations than some of the other lines, and also quite a bit of turns and topography that may slow the train down. My friend Terra also tells me this is the oldest DC Metro line, and that there are water infiltration problems during rainstorms. That will slow you down.


10. MetroTransit Hiawatha Line: Nicollet Mall station (Minneapolis, MN) to Bloomington Central station
Type: Light rail
Average speed: 22.1 mph
Distance: 10.7 miles
Travel time: 29 minutes
Station stops: 14 (2 in central city).
Comments: This line runs at grade alongside the namesake Hiawatha Avenue through neighborhoods, preventing the train from traveling at highway speeds.


11. SDMTS San Diego Trolley Orange Line: Civic Center station (San Diego, CA) to El Cajon station
Type: Light rail
Average speed: 21.7 mph
Distance: 16.6 miles
Travel time: 46 minutes
Station stops: 16 (4 in central city)
Comments: With regard to speed and length, this line is comparable to Portland’s MAX to Hillsboro. Also like MAX, San Diego Trolley has a lot of downtown stations that slow the line overall.

12. TriMet MAX Blue Line - Hillsboro: Pioneer Courthouse Square station (Portland, OR) to Hillsboro Central station
Type: Light rail.
Average speed: 21.3 mph
Distance: 17.4 miles
Travel time: 49 minutes
Station stops: 20 (4 in central city)
Comments: MAX stops at three ridiculously close stations in Goose Hollow, but then speeds through the West Hills tunnel. Then it simply has a lot of places to stop in Beaverton and Hillsboro. If only there were an “express train” option.


13. TriMet MAX Blue Line - Gresham: Pioneer Courthouse Square station (Portland, OR) to Gresham Central station
Type: Light rail
Average speed: 17.3 mph
Distance: 14.4 miles
Travel time: 50 minutes
Station stops: 24 (9 in central city)
Comments: I feel that, once you get into the teens in average speed, there is a problem. Probably the single biggest problem facing MAX to Gresham is that it stops 9 times in the Central City (which includes Lloyd District), taking 17 minutes just to reach open track at NE 16th Avenue. Some of MAX’s downtown stations are just 400 feet apart. MAX to Gresham goes the same distance as PATCO’s High Speed Line to Lindenwold, NJ, but takes twice as long. OK, I’m done bitching about my local transit line.


14. Port Authority T Blue Line: Steel Plaza station (Pittsburgh, PA) to Library station
Type: Light rail
Average speed: 16.8 mph
Distance: 12.6 miles
Travel time: 45 minutes
Station stops: 31 (2 in central city)
Comments: 31 is a lot of stations, and many of them are spaced tightly in old streetcar suburbs like Bethel Park, where there are five stops within one mile.

15. METRO Rail: Main Street Square station (Houston, TX) to Fannin South station
Type: Light rail
Average speed: 14.7 mph
Distance: 6.6 miles
Travel time: 27 minutes
Station stops: 13 (5 in central city)
Comments: It would be easy to blame Texas oil, but Dallas’s DART came in third. Houston’s light rail stays almost entirely within street right-of-way, keeping it at urban speed limits. It also passes through continuous urban fabric, with stations roughly every half mile. Better luck next time, Houston, if there is one!

So, what did I learn by nerding out on this issue? Probably the biggest surprise was that expensive, grade-separated rapid transit lines are not necessarily faster or better than light rail. Four light rail lines - in Dallas, Salt Lake, Denver and Seattle - outperformed the DC Metro and PATH. It's also apparent that the geographic layout of a region has a major influence on speed - such as the wide expanse of San Francicso Bay speeding up trains, or the uninterrupted neighborhoods of south Houston slowing them down. Topography also plays a role - not so much because of elevation, but because of the winding routes that hills force you to make. Pittsburgh T and DC Metro must navigate rolling hills, while Salt Lake’s TRAX and Dallas’s DART speed straight as an arrow through flatlands.

But in the end, the amount of time you sit in a chair on a train is based largely on decisions made by planners - especially station quantity and spacing. The grueling journey from Portland to Gresham, or from Pittsburgh to Library, could be made noticeably quicker by culling five or ten stations. But it’s hard to get rid of them once people have been using them for years. To be fair, MAX was one of the first modern light rail systems, built in the early 1980’s. Since then, other transit agencies have studied MAX’s pros and cons, and built new, arguably better systems based on what they learned. Seattle is a good example of this.

Planners (and their political leaders) also decide whether a rail transit line is in a mixed street environment (where it cannot move much faster than 15 mph), or in a separated, protected right-of-way, or a mix of both. While subways are expensive and elevated tracks are sometimes ugly, some places have gotten it right with both - Vancouver comes to mind.

In any case, I hope that the US and Canada continue to have the political will and financial means to invest in high-quality rail systems and other transit innovations in our cities. Thanks to a century of growth and experimentation in urban transit, we have impressive knowledge of the “best practices” and are now capable of building truly impressive networks to help people get around our cities efficiently.

4 comments:

  1. Very nice analysis, Steve. Certainly politics plays a big part in how many stations each line has, perhaps at the expense of efficiency.

    ReplyDelete
  2. I'm not done bitching about it. I want to lose my shit every time I have to ride all the way through downtown. You actually did Trimet a favor by stopping both trips at pioneer square. They should be ranked lower.

    They need to get rid of the PGE park and pioneer place stops, move the streetcar/library stop to the other side of 12th consolidate the three 1st avenue stops into two and get rid of at least one of the inner east side stops.

    At least they did better on the Green and yellow line. Although I still say running one of the lines down the inner east side where the streetcar is going and then over the new bridge so that you have counter-rotating circles around both sides of downtown makes way more sense than the current plan.

    ReplyDelete
  3. Agreed. I think the PGE Park stop is good, but ditch the King's Hill/Salmon stop that's literally a stone's throw away. Then ditch Pioneer Place, Oak/SW 1st, Old Town/Chinatown, Convention Center, and NE 7th. Yellow/green line is better, until it dawdles and meanders by Union Station. And they're all screwed by the Steel Bridge! Maybe an earthquake will take care of that. Just kidding. Sort of.

    ReplyDelete
  4. We need a better measure than speed over the ground. What really counts is how many dwellings, workplaces, and stores are passed per minute. From this perspective all the time those commuter oriented trains spend speeding through suburbia from one park-and-ride to the next doesn't count for much. The Portland Street Car is covering more valuable distance in only a few minutes as it ambles between Powells and PSU. With GIS we could calculate something like quality sources and destinations per minute of travel.

    Scott

    ReplyDelete