Friday, October 5, 2012

Knox Demonstration Road Diet, Before & After

As you may know, Knox Street was identified on the Dallas Complete Streets plan as a road suitable for being, ahem, "completed."  The gist of a complete street is that it isn't just for cars, but the infrastructure is essentially balanced for all forms of transportation.

Though Knox is not well-served by mass transit, to complete it without significant changes to the public transportation system, means a road diet.  In this case, the road diet took a four lane road and reduced it to two travel lanes in each direction and a shared center turn lane.  The leftover space from the removal of the fourth lane became a two-way cycle track, effectively extending the Katy Trail onto Knox as "an urban detour."

Throughout the country, similar road diets such as this one have proved to effectively calm traffic, improve safety, and actually move MORE vehicles with no evidence of any negative economic impact (usually there is economic improvement, but these numbers are harder to come by.  This is why I often point out that traffic isn't necessarily a bad thing.  Places need energy.  But to create a center of gravity that energy must be "condensed into a slow vibration," to paraphrase from physics).  Though this seems counter-intuitive, there are two primary reasons:  1) the place is more desirable when it is safer, so more people are attracted to it, and yes, drive there, and 2) the slowed traffic reduces headways between vehicles thereby increasing its overall capacity as more cars can fit on the road.  As people drive slower, drivers feel more comfortable being within 2-3 car lengths of other vehicles.  There is adequate stopping time and even if something were to happen, at such low speeds, damage would be minimal.

Because I knew this was happening, I decided to take speed readings before the installation and after.  Here was my methodology:

- data readings before and during the road diet demonstration were taken at similar times from 12-2pm on a weekday.
- data readings taken shortly after installation was complete, but before any of the programmed festivities so as to avoid distortions to the data.
- measurements were taken using a handheld radar gun from behind vehicles so as to avoid detection and subsequent unnatural slowing thinking I might be 5.0
- vehicular speed measurements were ONLY recorded when there were no break lights present and a green light ahead so as to record only "full speed."



The ranges shown in black at each segment of road show the full range of speeds on those particular street segments.
No vehicle moved along these segments at speeds greater or lower than these ranges with two exceptions. There were only two exceptions. One was a porsche heading towards Highland Park at 27mph and the other was an SUV heading towards 75 at 24 mph.
These are shown in brackets next to the segment where it was recorded.

In red parentheses is the reduction in speed on average per segment.
Predictably, traffic slowed all along Knox where the road diet was installed.

Interestingly, the vehicle speeds as cars entered Knox was the same as before, but then slowed noticeably as they approached McKinney Ave., where the start of the road diet was visible.

- The vast majority of vehicles actually moved slower than above, as most approached red lights or used their brakes. As noted, these were not recorded.
- During the time of recording, traffic generally moved at a steady flow, generally 1-3 car lengths apart. Stacking at red lights rarely over a few cars at a time.
- Choke points/backups/conflicts occurred not at intersections but at mid-block alleys, curbcuts, and crossing traffic into on-street parking.
- There were more pedestrians present at this time than before, though not in overwhelming numbers. Also, it was 92 degrees during, and 98 degrees when the Before measurements were taken.
- The right turn only lane from Knox to McKinney was critical in reducing conflicts.

How “Tethered” is the street?
A street can be considered knitted together based on the degree to which pedestrians are willing to cross a road for various reasons including cross-shopping, an intricate component to synergy of place. How knitted a street is, is a measure of inherent pedestrian comfort and confidence in their safety.

 - A road where pedestrians do not cross at all can be considered to have 0 knitted value.

 - Where they cross only at control points, such as crosswalks and intersections is a 1.

 - Where pedestrians cross at desire lines, a road can be considered to have a ‘knitted’ value of 2.

Lemmon Ave, a zero for integration

Greenville Ave, lose the zero, get with the one?

Main Street, lose the zero and the one and get with the hero, a 2.

Based on observation, it is believed that Knox as it is currently designed and operating generally functions as a 1 or slightly below. When pedestrians cross, they will use the crosswalks at signalized intersections. However, there seems to be little evidence of cross-shopping, partially due to tenant mix/location, and partially due to road design.

During the demonstration project however, pedestrian crossing at intersections was rare as most crossed where they pleased, at desire lines between origin and destination. Though pedestrian activity at the time was relatively low (compared to Main Street as shown above, though higher than normal), it operated closer to a 2.  As traffic speeds were suitably reduced by the demonstration road diet to below 20 mph, pedestrians felt more comfortable to cross the street without using crosswalks or waiting for the crossing signal.