Part 1. How 400 meters becomes 100 meters
When I lived on West North Avenue (the namesake of this blog), I could walk out my front door, hear an “L” train approaching from behind the apartment, dash across the street and around the block, and catch said train at the station a block away. When I recently asked Chicago friends about how much time it took for them to travel through their “L” stations, the responses were quizzical: “Uh, seconds?” “Less than a minute on either end. Perhaps you should be measuring in seconds.”
The Washington Metro might have record-smashing escalators and awe-inspiring cathedral ceilings in place of the L’s humdrum wooden platforms, but the sheer size of its stations hurts its usability for short trips, writes Ian Rasmussen:
“You’ll almost never hear about how long it takes to get from the street to the platform when people are telling you how long a transit trip is going to take… [I]n the context of systems designed to attract longer trips (30, 40 minutes), it hardly matters. But in the case of shorter trips, such as those in the urban core where the system is intended to act as a circulator, the issue cripples the system… just think of how you feel waiting to get off an airplane when you are about to miss your connecting flight.”
When combined with shameful 20-minute headways, the two or three minutes* it takes to descend to, or emerge from, a Metrorail platform add up to a substantial fixed time penalty on short trips within the core. It isn’t just the flowing mezzanines and interminable escalators, either: even in dense residential neighborhoods, stations often empty into meaningless plazas rather than seamlessly meeting the neighborhood. Over time, buildings will grow towards the station (as at Columbia Heights), but this process takes decades and has often been stymied by poor planning decisions.
This is not to absolve Chicago: it arguably invented the expressway median transit line and thus spawned places like Rosemont — which Yonah Freemark called “the Land of Missed Opportunity” for its uniquely awful transit-adjacent development pattern. The town of Rosemont** obviously understands that its “L” access gives it a valuable advantage over more distant suburbs. However, its station area pedestrian experience is just monumentally bad, with an uncharacteristically lengthy “L” station emptying out into a bus parking lot in the middle of an interchange. The net result: absolutely nothing, besides said bus terminal, is within the five-minute walk shed (below). Rosemont has attempted to compensate by subsidizing all-day circulator shuttles to feed its new retail/entertainment hub, but no shuttle can match the spontaneity of a quick lunchtime walk.
Expressway median stations suffer from a triple whammy of poor geometry:
1. The geometries of the surrounding environment are often defined by the 70 MPH cars swirling around them (particularly since the busy streets that make sense for station entrances also make sense for land-gobbling interchanges), rather than the 3 MPH pedestrians within;
2. Much of the walk shed is wasted crossing the freeway itself, much less interchanges;
3. The adjacent land uses either want to shy away from the freeway’s noise and smoke, or surround themselves with moats of parking and limited access routes, or both.
The same geometric problem is hardly intrinsic to rail. Bus rapid transit, which essentially is the interface between a highway for heavy buses and pedestrians, faces exactly the same problem. Here’s the award-winning system in Guangzhou:
The service had better be really fast, and really frequent, to be worth braving all that just to get to the bus stop. This sort of grade separation (also seen in Ottawa) is unusual; cost containment usually leaves pedestrians running in front of buses at grade, as in Cleveland:
Many of the inexpensive freight-rail alignments used for recent light-rail projects suffer from a similar (although less extreme) distance from the urban fabric. The north end of Baltimore’s light rail line runs in the former Baltimore & Susquehanna (B&S) Railroad ROW north to Timonium, alongside a stream valley, separated from adjacent development by buffers, grade changes, woods, and (to the east) the Jones Falls Expressway.
The LRT corridor passes many major activity centers on the north side of Baltimore, including two campuses of Johns Hopkins University, parks encompassing the stream valleys and adjacent hills, the Woodberry area of redeveloped mills, prosperous neighborhoods like Hampden and Roland Park, and at the northern end the backs of retail and business complexes facing York Road in the northern suburbs. However, historically development of residential and retail uses focused on the hills above the stream valley and freight railroad; the only uses directly fronting onto the railroad today are station access uses and some renovated mills. Jeff Wood notes that Minneapolis is about to embark on a similar mistake with its Southwest LRT project.
In short, to genuinely intertwine transit with city life, it has to be as close & convenient as physically possible. Don’t cheap out on an inexpensive but inconvenient alignment, don’t over-engineer stations, and seek the smallest possible station footprints that will do the job. These principles should seem obvious, but too many new transit projects still don’t get this interface right. I’ll explore some more examples in two future posts.
* Times measured at Rosslyn and Court House.
** For those unfamiliar with Chicago, Rosemont has leveraged its unique location surrounded by the city’s transport links (airport, freeway, beltway, transit) to suck “profitable” airport-adjacent offices & hotels from a city that warehouses its low-wage workforce. Therefore, it’s the quintessential parasitic “affluent job center.” [OK, slightly strange link, but I couldn’t find any other summary of Myron Orfield’s Metropolitics suburban-town typology that was in HTML rather than PDF.]