On One-Way Loops

Detroitpeoplemovermap On Transport has a nice post on the Detroit People Mover, a loop that connects a number of major employment and activity centers in downtown Detroit.

In a recent post I argued that downtown shuttles aren’t of much use unless they’re extremely frequent.  The Detroit People Mover doesn’t have that problem; it runs every 3-5 minutes.  It has the other common problem of downtown shuttles: it’s a big one-way loop.

(Detroit’s loop is clockwise, having reversed direction in 2008.  Some one-way loops, like the one at the heart of Melbourne’s train system, also reverse direction in the middle of the day.)

In a one-way loop, the the way you go from A to B is completely different from the way you go from B to A.  It’s likely to be much longer or shorter.  In fact, the more direct the service from A to B, the more circuitous it’s likely to be if you want to come back.

Whenever someone proposes a one-way loop as the solution to their transit problem, especially downtown, I feel the need to take a deep breath and offer — in my most calming and supportive voice, as though speaking to someone standing on a ledge — this crucial bit of wisdom that it took me years of study to acquire:

Very few people actually want to travel in circles.

Sometimes, of course, a loop serves a rational non-transit agenda.  Managers of a city’s tourism industry, for example, don’t particularly want tourists to get where they’re going. They want instead to create an experience that will show them other places that visitors might not have intended to go, and that might even be used to tour the city and come back to where you began.

But even where this isn’t the purpose, some people are just comforted by loops.  Transport planners describe our travel demands in terms of “desire lines,” straight lines from where people are to where they want to be, but some people seem to have “desire loops” instead.  When community leaders are asked, in a meeting, to talk about their transit needs, it’s not uncommon for one of them to say, usually with circular hand gestures, that they need some kind of loop. (The same people may use the word linear to mean narrow-minded or conceptually trapped.)  Straight lines can seem so aggressive, while loops offer a sense of closure or embrace.  And as a model of the fundamental nature of being, there’s a lot to be said for the loop.  See, for example, the medieval notion of the wheel of fortune, or the “cycle of death and rebirth” that underlies Hinduism and Buddhism.

For whatever reason, there are a lot of one-way loops out there.  In transit, one-way loops do have their legitimate uses, but they’re very specialized:

  • When you’re spreading a very small quantity of service over the largest possible area, with no concern for travel time, one-way loops are the answer.  Bus systems in small cities or low-density suburbs, where the goal is exclusively to provide lifeline access to the most transit-dependent persons, often use one-way loops for this purpose.  (As these systems grow, undoing these loops becomes a crucial restructuring step in the maturation of the system.)
  • In a very small circulator system, with very high frequencies, loops can be logical.  There’s nothing wrong with a one-way loop inside an airport, for example, where there are just three or four stops and the whole cycle is done in just a few minutes.  There’s no significant penalty to going the long way around the loop.
  • Bus lines and some rail services need to make one way loops at the end of the line to turn the vehicle around.  (Most trains and some streetcars can be operated from either end, so they can reverse direction in place without looping.)  The best practice is to make these turnaround loops as small as possible to minimize the number of people affected and potentially confused by them. In the best case, a station or interchange provides the turnaround capability so that there’s no need to loop on streets at all.

Two-way loops, of course, are a totally different topic (and can often be a ground of compromise between linear and circular modes of thought).  On a two-way loop, any portion of the loop can be experienced as an ordinary two-way route; some are even described as two or more linear routes that happen to be connected at the ends.

But be careful with one-way loops.  If you connect all the important dots in your downtown, and you call that your ideal route, you’ll probably find that you have some kind of loop.  Tourists will ride it, including locals visiting your downtown for pleasure, but it will have very little relevance to anyone else.  Most of our transit desires, I’m sorry to say, are linear.

9 Responses to On One-Way Loops

  1. Michael Druker July 17, 2009 at 7:20 pm #

    Yep. Sure, I understand the motivation — to “connect” all those places. But not only does it take a relatively long time to get to a station that’s close but in the wrong direction, it’s also plain humiliating to be in such a situation. (Note: My opinion is not based on experience of any such systems.)

  2. Alon Levy July 17, 2009 at 9:08 pm #

    Two-way loops are often invaluable parts of a city’s transit network, even when they’re not really two linear lines connected at the ends. Examples include Tokyo, London, and Moscow, whose circular lines provide crucial transfers.

  3. PointSpecial July 17, 2009 at 9:39 pm #

    What about Chicago’s Loop? It has spurs radiating out from it (thus the Loop is the destination in many respects, with different nodes), but it is circular and it is one-directional.

  4. Cap'n Transit July 17, 2009 at 10:19 pm #

    No, the Chicago Loop is not one-directional.

  5. Matthew July 17, 2009 at 10:36 pm #

    As a Chicagoan, I understand the desire to defend the Loop, but you’ve glossed over some important points.
    First of all, the Loop is not one-directional: the Pink and Orange lines run clockwise, but the Brown line runs counter. Plus, the Green line sort of runs around the northeast side, trying to avoid slowing the other three lines down too much.
    Secondly, the Loop is an historical artifact. The L system was originally many independently owned and operated lines which all converged on the central business district. Once the lines reached downtown, the trains had to be turned around some way, so a circuit track was built which allowed the trains to do a loop. The Loop was never intended as a one-way loop as Jarrett describes, i.e. it was never a route in and of itself. Rather, it was a pragmatic solution to a problem caused originally by a bunch of streetcar routes converging on a dense area with no room for turntables.
    Besides, Jarrett’s point still holds: do you know anyone in their right mind who would take the Brown line from LaSalle and Van Buren to Clark and Lake, and then back? Of course not, only a tourist would do a circuit of the Loop like that. Most people would just walk the 8 blocks.

  6. Dave in KY July 19, 2009 at 8:38 pm #

    While I agree that loops stink, lines also stink. Lines can be okay if their endpoints generate a lot of trips – say an industrial park or airport – but most lines I’m familiar with push out to the hinterland. When it comes time to turn around, who wants to go back to where they just were? Nobody, otherwise they would have gotten off. So you see buses with 55 seats and 0-1 passengers for a mile or five. That’s a disaster, it brings down the efficiency of the whole line.
    I think loops need to be looked at more closely. If the loop cuts across densitites in an eliptical orbit, it will also have lower ridership in the hinterland, but probably not zero. If the loop runs across constant density, then ridership might be about constant.
    Of course all your points about not wanting to run 330 degrees around the loop to go back one stop are coming into play here.
    Concentric rings operating in opposite directions?

  7. Jarrett at HumanTransit.org July 20, 2009 at 9:10 am #

    When you talk about “lines” I think you’re talking about unanchored lines — i.e. lines that end in a low-density area. Better network design (and better land use) leads to lines that always end at a major destination, so that there is strong demand all the way to the end.
    What you’re describing in the latter half of your comment is a two-way loop. As Alon Levy mentions, there are lots of great examples of those, and they work well because they can be treated as two-way lines at any point.

  8. EngineerScotty July 20, 2009 at 9:56 am #

    Tri-Met has many bus lines that go nowhere are are used by nobody, but seem to be necessary to maintain the taxing district. And many lines that go into the weeds may be ninety-percent empty most of the time, but still provide a valuable service to somebody…

  9. Ari August 31, 2009 at 8:18 am #

    Re: Melbourne’s City Loop:
    The City Loop in Melbourne is not a one-way loop, but much more of a distributor loop like Chicago’s L. About a dozen suburban rail lines (most with 15-30 minute off-peak headways, however) converge on the downtown, and loop through five stations in the central business district (much like the Chicago Loop). The reason the City Loop shuts down at midday is that, ostensibly, certain stations have more patronage and the goal is to minimise the trip length for those stations (although I’m really not sure; it seems to me that the switch is confusing and not necessary). In any case, at any time of the day, trains run in multiple directions.
    That said, the City Loop is not meant for looped travel, as not services make a full loop. Trains enter the loop, drop off and pick up passengers, and then leave. Cross-CBD travel is facilitated by one of several tram lines, although these services are generally a lot slower since they operate at street-level. In fact, there are several parts of the day where it is impossible to get from Melbourne Central to Flinders Street by train, although there are trams that make the trip every minute or two. Since Melbourne’s ticketing system covers trams, trains and buses, the fact that the City Loop does not allow for circular travel is not a huge deal.

Leave a Reply