Wednesday, June 25, 2014

Bad Intersections of Dallas: What Flow/Friction Mean and Why They Matter

Sometimes I start things and get to busy I can't keep up with them.  One is a series (of 1) about the bad intersections of Dallas (and DFW), of which there are too many to count.  Since I've recently moved from downtown to the hipster paradise of third world Amsterdam we call North Oak Cliff (oh noez, I'm the yuppy pushing the cool factor out - Brooklyn is so over anyway), I've come into much greater contact with an intersection everyone knows is terrible: Zang and Beckley at the point of the Zang triangle.

Embedded image permalink

Today, I'm going to explain why it's so terrible, why it matters, and why everyone complains about high taxes and bad roads.  We've designed our city that way and it's time to change course.

First, the basics.  Zang and Beckley is a five-way intersection including the residential street Eldorado.  The intersection is very close to a major (but underutilized) park (Lake Cliff - which is beautiful and full of potential), a major employer (Methodist Hospital), an elementary school (James Hogg - which, if anything, perhaps rather than liquor laws near schools we ought to have 'safe zones' where traffic throughput isn't prioritized over say, kids getting to school in one piece), and a booming housing market since so many people want to be part of the interesting character of North Oak Cliff (what I've always referred to as the Trastevere of Dallas (ok fine or Brooklyn or Left Bank).  All of those things are inherent advantages ripe for placemaking, public spaces for social and economic exchange, however squelched for the narrow minded focus of traffic throughput, at high speed.

There are three primary problems with this intersection and I'll touch on each in more detail afterwards:

1) Each of the primary arterials entering the intersection are individually over-built.

2) The intersection itself, once you exceed four-way, must be designed as a special case, otherwise both traffic flow and socio-economic exchange suffer.  Either of which could be prioritized and defended, but having neither is catastrophic.

3) Concrete medians.  Oh, those awful concrete medians, occasionally only 2' wide stretching into the intersections.  I'm going to spend the remainder of this post on the first two, since this one is mostly a trivial annoyance in proportion.  However, it's these medians that encourage higher speed flow while not doing anything a well designed median can and should do, provide pedestrian refuge for crossing or shade trees/vegetation.  Furthermore, they stretch so far into the intersection that they prevent decent tactical tests in examination of better ways to design the street.  Therefore, let's just (at least hypothetically) rip the entire thing apart and start again.

First, let's get to the scale of the roads and the traffic counts.  Below is Colorado by Methodist Hospital.  It isn't part of the intersection but is useful to the case study because Beckley counts drop by a quarter whence crossing Colorado, presumably and partially due to Methodist circulation and access (and the residential behind).  What his means is that Beckley can change character after crossing Colorado southbound.

Embedded image permalink
Here is Zang shorly after passing through.  Sure lives up to the lofty name 'boulevard' does it not?

Embedded image permalink

Below are the rough traffic counts and approximate percentages of the capacity they're built for (amazing how many arterials in Dallas are in the 30-40% range.  It's almost like that's a standard or something, cough cough.  Sure, the traffic flow is great, but there are implications (you mean besides the original widening took private property and destroyed buildings?) worth considering when you acknowledge that roads, our public infrastructure should not be about 'moving cars' but facilitating social and economic exchange.  Moving traffic (of any form) is a means, not an end.  However, the dinosaurs treat the movement of traffic as an end itself.  Therefore, these roads get a grade 'A' from them, but an 'F' from economists and citizens alike.  We must design better and to design better we must think better.

I use a red-yellow-green system, but in a counter-intuitive way.  Red is very low traffic (we'll get into that).  Green is a good traffic balance appropriate for the purpose of place.  Above, Beckley south of Zang is green despite only 44% because it becomes more residential in scale and nature at this point.  Yellow (not shown here) is for higher traffic areas approaching capacity.  Red, semiotically has bad connotations (STOP), and thus represents the worst case scenario.  Here's why:

Think of flow as infrastructural burden.  Widen the roads traffic flow moves better but it increases tax burden to maintain.  It also often erodes investment and economic activity along it, ie tax base.  Economic activity = friction that can interrupt flow or flow can erode, depending upon how the space is designed.

In the Red cases above where traffic is too low for the properties along it, we have the following condition:

Infrastructural Burden (cost) >>> Tax Base (benefit)

A better situation is either:

Infrastructural Burden = Tax Base, or even...

Infrastructural Burden <<< Tax Base.

In the scenario, traffic flow may be terrible.  However, the high value of increased tax base means the ability to afford other amenities such as increased transit to maintain mobility.  Think of Stroget in Copenhagen.  Ain't no cars getting through (except morning deliveries), but amenity and value is high.  Excuse the colloquialism, I'm trying to fit in.

Zang east of Beckley only moves 7-8000 cars per day despite (formerly) being a 6 lane arterial.  That's our nightmare scenario.  A two lane road can handle that traffic.  Lucky for us the new Oak Cliff streetcar will be eating up some dedicated lanes.  However, because the tracks were laid centrally (if memory serves) riders would have to cross two lanes of traffic to board.  We can narrow this segment of street down to one travel lane and one dedicated street car lane in each direction.  That gives us plenty of more space to play with for other investments and/or opportunities such as on-street parking or bike lanes or whatever we want.

The busier part of this road network is where Beckley bends into Zang as it transitions through the intersection.  Through here, the traffic is still only in the middling teens of thousands of cars per day.  14k, 16k, 18k varying on date and location of count.  For comparison, these are the typical counts for road diets around the country when 4 lanes are converted to 3 (one lane each way with a center turn lane).  We're dealing with 6(!) plus a center turn lane.  Thus, the road is built for 44,000 cars per day.  That's Broadway.  That's also anything but.  See, my red yellow green flow-to-friction gradient.  These roads can easily lose a lane on both sides.  That's critical if we want to redesign the intersection

Prioritizing FLOW on the streets as they approach the intersection then has negative implications once we hit the intersection.  I mentioned earlier that with anything more than 4 roads meeting at an intersection (this is 5-way), we have to make special design precautions.  The reason, with the way signalized intersections work, is that with 10 directional movements, only 4 can ever be moving at one time.  Whereas a typical intersection is 4 of 8.  This creates an artificial choke point where traffic flow is actually reduced.

Because it's designed entirely for the car and all space is given over to it, there is little to no economic activity happening either.  At least, much less than could be.  The only pedestrians are those traversing from parked cars to get Spiral Diner's ridiculously good blueberry pancakes.  You're willing to take some risks to get those pancakes.

Somebody even had enough of a sense of humor to paint some crosswalks.  As if.

Therefore, it's at least defensible to have the goal of FLOW or FRICTION, however when neither performs then we must rethink how the intersection is designed.  Choke points can only fit so many cars through them, but they make for some of the greatest public spaces in the world when they're redesigned.  See Times Square or Campus Martius in Detroit.  However here, the effort to prioritize flow undermines both flow and the friction of activity.  There is a way to do both better.

Below is roughly how the junction looks today with some approximated linework for travel lanes.  You can see that there are several other roads that funnel closely into the various five roads forming the intersection.  We can either squeeze all of these travel lanes through this confined space poorly, or we can try to move the same amount of traffic through, albeit in a calmer fashion, in order to 1) carve space out for people and 2) make the space more comfortable and appealing for people.  Nobody wants to inhale traffic fumes nor sit at a cafe if the road is so busy and loud as to not be able to hear.

So what if we were to appropriately size the streets entering the junction, thus improving some of the geometries and radii which makes for loads of extraneous paving and underutilized square footage.  Then, we create a roundabout of sorts, a slow-moving two-lane rotary, with cross walks, within the confines of existing buildings allow.  Because of the odd geometries, we have some existing (and interesting) buildings in some odd places.  What if we were to incorporate the little wedge building within a 'triangle-about'.  Sure, we're taking some private land in the area, but we're also turning over more public space into either developable private land or from road paving to people space.  The crossing pedestrian crossing points are all foreshortened as well.

Green above doesn't necessarily represent green space, but rather repositioned land.  This could be green space, or plaza space, or cafe space, or some on-street parking or bike lanes (in existing right of ways), or widened sidewalks, or street trees, or even to leverage land use intensification and development.  All of that could still be determined.  The key point, however is that we can actually improve the traffic flow (as I mentioned the choke point inherently under-performs) while also flipping the equation so that there is pedestrian activity, thus friction, and in turn, increased socio-economic exchange occuring.  The very purpose of cities, and thus the purpose for our infrastructure as a subset of cities.