‘The Streetcar’ – 2 – Errors of Geometry: split one-ways are dumb
This is the second post in a rather deep series on The Streetcar. I’ll be critically analysing the City of Cincinnati’s proposed streetcar, known as “The Cincinnati Streetcar” in the local vernacular. I think it’s a poorly developed plan, but not one that deserves the abuse the local right has heaped on it, nor even half the praise the urban liberals have been lofting skyward. Like most things, it’s hideously complicated, and like some things, I’ll try to make it’s twists and turns illustrative of broader concepts. So roll your bicycle onto this level-boarding train of thought and get comfy! Here’s quick lay of the series:
- Conceptual Flaws: civic boosters lead the charge
- Errors of Geometry: split one-ways are dumb
- The role of transportation in economic development
- Symbolic Transit
- On the back of branding
- Separate and Unequal: on therails of division
- Chicken or Egg: Shaping the future or following the past?
- A New and Highly Sarcastic Plan for Economic Development
I said in the last post in this series that the drive to hit a number of culturally and economically important ‘destinations’ with the route of the streetcar led it to be less than ideal as actual transportation. Now it’s time to back that up!
It is a little difficult to actually assess the efficacy of the streetcar as transporation because it’s not totally clear what it’s trying to do. And of course because the range of possible alternatives to it is almost infinite. The vague notion of “connectivity” or “circulation” isn’t sufficient here; we must actually understand just what exactly it’s connecting and how well it’s doing it. But for the sake of argument, let’s say that each of the two extreme stops are in generally the right location so we can simply judge the effectiveness of movement between them.
One very important principle to consider is frequency, or how often a vehicle is scheduled to come by. By knowing how often we can expect a trip to occur, we can know our maximum possible wait time and the average wait too. If a vehicle comes by every ten minutes all day, the most anyone could possibly end up waiting is ten minutes. The average wait then, assuming a steady stream of accumulating passengers, is half of that–only five minutes. If the frequency doubles, each of those numbers is halved. When multiple distinct lines such as the 4 sub-routes of the #17 and the two sub-routes of the #19 completely overlap like they do between Downtown and Northside, they can create a really high frequency transit corridor with about 10-15 minutes between vehicles. For many riders, that’s little enough of an average wait(5 to 7.5 minutes) that they won’t need to check a schedule.
This is why redundancy is absolutely critical, particularly for such a short trip as the streetcar’s. The streetcar is planned to operate at 12 to 20 minute headways1, meaning an average wait, as we’ve calculated it of between 6 and 10 minutes. Anything that can be done to reduce that wait time would increase the likeliness that transit will be a useful option.
So let’s not forget that a lot of other transit lines are and will be operating in Downtown and Over-the-Rhine and that they present opportunities for redundancy and thus increased frequency just as the #17 and #19 already compliment each other. Most lines can be safely ignored since they’re not going in the same general direction as the streetcar. Of the ones that are and are fairly frequent, we have the #6, #17, #19, #43, #46, #49, #64 and #78 as possibilities. Let’s see how they line up:
They don’t really line well up at all. Actually, none of these really line up with any other through Downtown/OTR except for the #17 and #19. That’s a lot less than ideal, particularly for short trips where a long waiting time will disproportionately give other modes of transportation a strong competitive advantage. If your trip is only going to take five minutes anyway, an extra five minutes spent waiting doubles your total travel time and makes walking look a whole lot better.
A better route would be one that was redundant with at least one other transit line going the same general direction with a relatively high frequency. If the streetcar lined up with the #17 & #19 for example, then instead of 12-20 minute headways and 6-10 minute average waits, we could effectively2 experience 5.53 to 8.64 minute headways and a 2.7 to 4.3 minute average wait for either line. That’s a pretty big difference! Instead, by not sharing a route, they’ll each operate with a significantly lower effective frequency for the trip between Downtown and the north end of OTR. Redundancy with the #6 and #48 would be even better5, as they all go up Vine St. That would extend the high frequency corridor even further.
Another important principle to consider when laying out a route like this is walking distance to stops. Like frequency, the importance of this consideration is magnified by the very short length of the initial route. Obviously, at a certain distance, one may have to walk more than the length of the route itself to get to a stop, at which point the line will have likely ceased being a useful mode of transportation6. One would simply better ride a bike or drive the whole way. A short route places a low upper limit on the distance people will reasonably be willing to walk. On the contrary, people will come from all over for an airport because the trip is generally quite long.
So the placement of stops is important if we want to be within a reasonable walking distance of as many people and things as possible. Streetcar advocates have been quick to show us maps of all the stops with buffers around them covering most of OTR. This is pretty misleading though and I personally added to the confusion when I posted this picture on facebook thinking it was simply quite pretty:
Someone commented, “So about 75-80% of OTR will be served, along with about 50% of downtown. Excellent.”
This is indicative of the erroneous thinking that leads these kinds of maps to be popular among advocates of less generally useful coverage-type transit services. As Jarrett Walker has well explained, routes that split two directions of travel onto two one way streets, look like they cover more territory while actually covering significantly less. You might be within walking distance of a stop going in one direction, but in order for the entire service to be properly within range you also need to be close enough to a stop going the other direction. Afterall, if you want to get home at the end of the day, you’re likely going to need to come back the way you came!
So let’s take a look at what that same map would look like if I took the time to complete my rather crude7 analysis before I shared it with the overexcited world:
Since the area inside that green caterpillar is small, flat, pleasant, and well-sidewalked, we should also compare it to such an obvious choice as walking. The average speed is expected to be about 6.7mph, roughly equivalent to a light jog. That’s actually not so bad once you’re on it, but we have to consider a few other factors here. Wait time is a big one. As I’ve said, with a 12-20 minute headway, the average wait will be between 6 and 10 minutes. Let’s settle on 8 for the moment. If we’re walking at 3.1mph, a typical speed, we’ll cover 0.41 miles8 in eight minutes. The whole route would be about a 1.7 mile walk, leaving just 1.29 miles that could be more quickly traversed by the average user starting from the very last stop and going all the way to the other end. How much more quickly? About 12 minutes. Now, most people won’t start from the very last stop though, and go all the way to the far side of the line. Most people don’t live just right above either of those last stops. People starting from or going to places not at the end of, but along the line will find it significantly less useful. You can imagine not many people will wait at any stop to go only one or two stops down. Maybe most won’t wait to go even three stops. But three stops is a pretty big portion of the whole length on a route this short. …This topic really deserves some more space though, so I’ll do a more complex and further reaching analysis some other time to see just what kind of trips anywhere would actually be faster than walking.)
The streetcar’s route also falls a bit short of expectations in what I call conceptual efficiency. One of the big reasons people don’t use transit is because it can be just plain confusing:
A big part of what I’m trying to do with this blog is about simplifying the use of transit by cutting out all the clutter. I think people should be able to use transit without thinking too hard about it. But the streetcar, like most of SORTA’s lines through downtown, looks like it may end up being somewhat more complicated and confusing than it needs to be.
Why, really does it need to take that little three block jog over to the northwest? And what’s with all the one-ways? And then that tiny little side-step toward Vine? If it just went straight up and down Vine Street can you imagine how much simpler it would be to give directions to tourists and other newcomers? “How do I get to UC?” “Vine Street!” rings out your bold reply, “Vine Street!” Instead, in it’s very first phase, the streetcar manages to use not less than 11 streets for it’s route. So much for simple directions.
Even without changing the direction of any one-way streets, a route that simply piggy-backed on the much simpler #78 would be a big step up not only for clarity, but for frequency too. And if it went two blocks further south it wouldn’t significantly change the area the streetcar would actually cover.
It could also potentially use both directions of Vine until it turns into a one-way and split off only through Downtown. Adding further confusion, once you start including all of the proposals for later phases, you get something which isn’t very clear at all in how it might actually work. Can you picture how the vehicles will actually flow through this?
To the extent that something is too complicated for people to bother understanding, they won’t use it. And since this topic is rather more ethereal I hope you’ll forgive me for slipping into a looser tongue. Perception is a slippery business.
In my opinion, all of this ultimately adds up to a new and quite hyped transit line that seems to do fairly little to actually improve transportation in Downtown and OTR. It might be a 12 minute improvement for people who live at the Banks and work at Findlay Market, but there really aren’t many of them and they probably have bikes anyway. For the rest of us, the benefits are a lot less clear. Any way you slice it, the route could have taken better advantage of redundant transit lines, two-way streets, and the potential clarity provided by one of the region’s few large regular street grids.
- “Headway” is another term for what we’re talking about. It’s the temporal distance between one vehicle and the next, such that if a driver tells you “The next bus is coming in ten minutes,” you know the next bus is operating with a ten minute headway, or if that is consistent for all following buses, a frequency of about ten minutes. ↩
- Not exactly, but effectively, because schedules wouldn’t ever line up such that vehicles would be perfectly spaced out. It would be fairly close though, and even random delays and accelerations would work toward an even distribution if you can discount the effect of bunching. ↩
- 60/((60 minutes/10 minute headway)+(60 minutes/12 minute headway)) ↩
- 60 minutes/((60minutes/15 minute headway)+(60 minutes/20 minute headway)) ↩
- If they were even redundant with each other…what’s that about SORTA? Two lines going the exact same places(bottom of Vine St to Government Square, and they don’t take the same route?) ↩
- Transportation planners generally call the area within which people will be willing to walk to something it’s “walkshed” ↩
- To really do it properly, I ought to have measured the distance along streets rather than straight-line distance in a circle around the stop. But that plugin is super complicated and I’m still trying to figure it out. You’ll see more on that topic once I’ve figured it out! ↩
- 3.1x(8÷60) ↩