Analyzing Potential Road Diet Candidates in St Paul

stp-road-diet-map-davison

The final product of my analysis based off Average Annual Daily Traffic (AADT) volumes and the number of through lanes per direction of traffic. Based off data from MnDOT. Source: Self-created

Larger version of map (24×36 PDF)

Preface

Since Bill did a phone interview with me about this map and wrote about it last Monday, I figured I should explain this map in greater detail, such as my reasoning behind creating it. This article contains my motives, the process of creating the map, the limitations of the data, along with my data sources. This article contains my personal opinions and are not official nor endorsed statements from my employer (MnDOT).

My Motive Behind the Analysis

With the recent news on pedestrian fatalities in Saint Paul, I became curious and started brainstorming possible solutions the city could use to improve safety for pedestrians (in addition to cyclists). I work in Saint Paul at MnDOT’s Central Office building (also known as the Transportation Building) near the State Capitol. I currently am commuting by car by myself from Little Canada (about 5-6 miles one-way), but I usually walk during my lunch break around the Capitol Complex. I also sometimes walk over to downtown or near the Cathedral along Summit and Selby. During my walks, I tend to notice how many motorists are quite inconsiderate of pedestrians and bicyclists. Many don’t use their turn signals, and many do not follow the state statutes that grant pedestrians and bicyclists the right of way among all crosswalks and intersections. I even saw a person stopped on the light rail tracks at the Cedar and 12th intersection earlier this month. I also did see a bicyclist run through a red light, and distracted pedestrians due to cell phones; ultimately drivers breaking roadway laws was the most common in my experience. There will always be impatient or inattentive people no matter what mode of transportation they are using. I’d rather have them be walking than behind the wheel though if they are glued to their phone.

I try to make it my best effort to keep an eye out for all people, but I have admittedly failed at times. There is no excuse for not seeing pedestrians or cyclists, yet many of us drivers do it on a daily basis multiple times throughout the day. I’ve documented myself doing it twice recently, which is disappointing. It is both a rude, inconsiderate gesture, in addition to it being very dangerous. It is also a very expensive mistake to make if a police officer notices your inaction to yield, as you risk getting a citation. What is even worse is if you injure or kill a pedestrian or cyclist, the financial cost is severe (estimated to be around $10.6 million per fatality) and the emotional toll is priceless. We must do better, and I must do better. Being patient and attentive is better than having your impatience or inattentiveness harm or kill an innocent person.

Creating the Map

stp-road-diet-map-davison-KEY

Map Key

Since I work in MnDOT’s Transportation Data and Analysis Office, I know that our traffic count data is publicly available through our webpage for the Traffic Forecasting & Analysis section. In addition, we are in the process of transferring our roadway data into a new system called LRS (Linear Referencing System). The new LRS data has a layer with the number of through lanes (per direction of traffic) for all routes. Through lanes are also known as general purpose lanes, which are traffic lanes that are not designated turn or auxiliary lanes. Our office “froze” our data back in January 2014 since we are still in the process of migrating and updating all of our data into LRS. This means that the data available is outdated by over two years. Given that this was a personal project, I ended up updating the through lane data for the routes within Saint Paul myself on my own time to do this analysis. I ended up comparing aerial photos on Google Earth with the existing data I had. The aerials for the city are quite recent (March 11th of this year), so I was able to make sure that my edits were done using recent aerials. I also took a look at the city’s website of road projects to see if there any upcoming changes to the roadway network. With this data, I was able to see which routes could be involved in a “road diet“. Mainly road diets are regarding 4-to-3 lane conversions, which have been proved to improve overall safety for everyone in multiple cases. I also factored in possible lane narrowing (such as downsizing to 10-11 ft lanes instead of the 12 ft standard lane width) and lane reductions (such as a 3 to 2 lane reduction (per direction), such as southbound Rice St near Sears and my workplace). I then took the AADT (Average Annual Daily Traffic; referenced as vehicles per day (vpd) in this article at times), and classified them into four options (Great, Good, Fair, and Poor Candidate; see ‘Map Key’ graphic), which was put over another layer that classifies the number of through lanes within that section of the route. I only looked at sections of routes that had two to four through lanes per direction of traffic.

The numbers I used to decide the suitability was based on the FHWA’s Road Diet Informational Guide and various case studies done by various transportation agencies. I was rather liberal and broad with my definition of suitability, so my version of a “good candidate” may be a poor one based from another perspective. Page 17 of the Information Guide discusses the metrics agencies have used as a cutoff for road diet feasibility. I used the higher cutoff rate of Seattle’s road diet feasibility limit of 25,000 vpd as my cutoff on road diets. The FHWA states multiple studies on road diets that provide a wide assortment of values on what the daily traffic volume limit is when a 4-to-3 road diet would become unfeasible. A Kentucky study from 2011 showed road diets can work up to 23,000 vpd, while a more conservative limit of 15-17,000 vehicles per day was suggested in a 2006 study by UW-Madison (sponsored by MnDOT and the LRRB). The FHWA suggests routes of under 20,000 vpd for being a candidate for a road diet. Given this variation, my metrics for determining potential candidacy need to be taken with a grain of salt. The number of vehicles per day cannot be the only metric used in the final decision on determining whether a road diet can be done. Routes with under 15,000 vpd seem most likely to still be candidates when including other factors, and routes with 15-20,000 vpd have a good chance based off the studies showing that. Routes that have between 20-25,000 vpd might be more controversial to convert. Personally, I don’t think impact of level-of-service for vehicular traffic shouldn’t be the sole major reason for rejecting a road diet, but I am not the master builder of the city and have limited knowledge in traffic engineering. I’m sure at least one traffic engineer in Ramsey County or even MnDOT would shake their head at my map if they saw it.

Regarding the design of the map, cartography isn’t my strong suit, but I tried to make the map look more aesthetically pleasing and easier for end users to interpret. Transportation planning has a lot of esoteric information tied with acronyms (ex: AADT vs ADT, VMT, ESAL, etc.) and algorithms that I can’t keep track of even after being at MnDOT for over a year, making it difficult to properly display spatial data to the public in a concise and clear way. I also transferred the layers of the map to Google Map Maker and added data from the Saint Paul Police Department regarding crashes involving pedestrians and/or cyclists, listed here (this online map is still being worked on and is subject to change). An interactive map allows users to see aerial (satellite) photos of the possible candidate routes and features like Google Street View allow citizens to identify trouble spots along routes (such as unsafe pedestrian crosswalks, overly wide traffic lanes, etc.).

Thoughts on the Outcome

RiceSt_at_MilfordSt

Rice Street at the corner of Milford Street in the North End neighborhood of Saint Paul. It already acts as a two-lane road along some portions due to parked cars in the right lane. Source: Google Streetview, August 2014

One street in particular that is bad for all modes of transport is Rice Street. Rice Street is a reliever/alternate route for Interstate 35E, that sees between 10,600 to 15,700 vpd within the city. Most of the route is under a layout of 4 narrow lanes (2 per direction) between Wheelock Parkway and its terminus near Interstate 94. There are no proper bike facilities within or nearby, and the sidewalk infrastructure varies in quality depending on which part you are on (I did my senior thesis on that). Rice Street could be considered the main commercial thoroughfare for the North End, given it does have various clusters of small businesses along it. Parking is allowed in the right lane at some parts, usually depending on what time it is (such as no parking from 4 to 6 pm on weekdays if you driving northbound, though this is not too strictly enforced based off my experience). There is also a lack of designated left turn lanes. This compels drivers to constantly lane change lanes especially near major intersections such as Maryland. Given the narrow lanes, parking, and lack of left turn lanes, it makes for a very unsafe route for everyone. I believe Rice typically has a ROW of about 45′ (ballpark estimate), which would allow a rather easy 4-to-3 road diet, or it could even be cut down to two lanes with left turn lanes at intersections with a high enough volume of left turning traffic. I would be fine with a center-turn lane through the entire stretch though, given Rice is that way near my home in Little Canada and it carries a similar amount of traffic (14,500 vpd). Rice Street is currently under a three-lane layout between Wheelock Parkway and County Road B, where near Highway 36 it is very wide at four lanes, and then it goes back down to three. Vehicle queuing does occur during rush hour periods, although traffic still moves relatively quickly. During periods of construction of nearby highways does cause worse traffic along the route, but there are enough alternative routes to spread out traffic. During the recent reconstruction of Interstate 35E last year, I usually just used Dale Street instead of Rice during rush hour.

Bike lanes could be installed (hopefully wider than the existing bike lanes on Rice near Highway 36); a layout similar to Riverside Ave in Minneapolis should be considered. If bike lanes aren’t able to be built, the good news is that the city has a future bike boulevard planned in their Bicycle Plan nearby on Park Street, which is only one block to the east of Rice. The bad news about Park is while it already has bike lanes along a short part near the Capitol, for it to be a true route for cyclists would require connecting existing gaps of Park between the Capitol and Arlington or Wheelock Parkway. Currently there is no access to Park St from multiple spots due to railroad tracks and Pennsylvania Ave. If the city wants to compromise for Park as the designated bike route, then they will have to find a way to provide a direct route for cyclists, otherwise it won’t be of much use to them. Jackson Street is only a couple more blocks over to the east with bike lanes along much of the route, although it is impractical for those who want to access businesses on Rice. Crosstown bike routes near Maryland and Arlington Avenues would also have to be implemented to provide a safe city-wide cycling network rather than one fixated on Downtown or Capitol area destinations. Cyclists aren’t going anywhere, and I would rather see expansion towards cycling facilities versus expanding existing routes. Expanding transportation infrastructure usually induces demand; increasing bicycling demand versus adding more drivers in the city seems more beneficial for the city in the long run.

Limitations of the Data

Given the data freeze limiting myself to using AADT counts from 2008-2014, the traffic counts may be different in the present day. Extended routes such as Cayuga St and it’s soon to be completely open interchange with Interstate 35E doesn’t have traffic counts yet, and the completion of the Cayuga and MnPass construction project on 35E probably will change traffic volumes along streets and roads near the freeway. Also, since it was self-edited, human error is likely as I may have missed or incorrectly classified some routes. This was a basic analysis and not a professional one, and is not an official analysis done by MnDOT. MnDOT is not responsible for the integrity of this data given it was edited from their official record.

AADT counts have their flaws as well no matter how recent the counts were made, since they are calculated through a total yearly sum of traffic volume divided by the number of days in a year (365). There are some adjustment factors that improve the precision of the values, but most counts are done through 48-hour short counts and that results in a possible margin of error. An AADT of 10,000 vpd could mean the roadway is at or over capacity during rush hour peri0ds, but has low traffic volumes outside of that timeframe. Other roadways might be more consistently busy throughout the day. The true daily traffic volume of the roadway could be anywhere from 8,000 vpd on a weekend to 14,000 vpd on a weekday. An example could be when a large event occurs such as a Wild game at the Xcel Energy Center. Nearby roadways are usually packed before and after events. It could contested that routes shouldn’t be built for its peak traffic if the capacity for the majority of the day makes the route underutilized. If a route is severely congested mainly due to major events, then other options such as improved transit should be considered to quell the demand for vehicular traffic.

Kellogg Blvd at Mulberry St

Even if a route has too high of an AADT (over 25K vpd), that doesn’t mean it can’t be improved. Kellogg Blvd for example, could use pedestrian refuge islands at this crosswalk where a mother and daughter were hit by a driver who illegally passed another driver that adhered to the right of way law (violation of MN Statute 169.21 Subd. 2 (b)). The mother sadly did not survive the crash. Source: Google Maps, October 2015

And lastly, I am not an official transportation planner nor an engineer. I majored in Urban Studies as an undergraduate, but my knowledge on traffic engineering is rather limited and is mostly self-taught. Therefore this data could easily be contested by a professional analysis that factors in the spacing of signalized interchanges (leading to higher risk of queuing, leaving the possibility of interchanges being blocked), the number of vehicles per hour), along with other criteria. In order for these routes to actually go though a road diet, they will need to be studied further by the city and will need public support.

Despite my admittance of myself being very naive when it comes to traffic engineering, I think a basic map such as this one can help to spark more conversation on potential roadway improvements. It seemed like the time to do so given the city keeps seeing pedestrians killed trying to cross the street, which is saddening. The metropolitan area recently just became one of four winners of the USDOT’s Every Place Design Challenge. This leaves a perfect opportunity for finding solutions for retrofitting our transportation network to make it into a safer environment for everyone. So lets fix our past mistakes, and work towards making Saint Paul the most livable city in America (otherwise they better change their slogan).

This was crossposted to (sub)urban studies


Streets.mn is a non-profit and is volunteer run. We rely on your support to keep the servers running. If you value what you read, please consider becoming a member.

, , , ,

2 Responses to Analyzing Potential Road Diet Candidates in St Paul

  1. Adam Froehlig
    Adam Froehlig June 30, 2016 at 11:18 am #

    Nice explanation of the process, though I would’ve like to see more about which GIS software you used and how you manipulated it (personally, I use ArcGIS for all of my maps as I have their Home Use license).

    One caveat: despite the “data freeze” you mentioned, 2014 would be the most recent year anyway for most traffic data, as MnDOT has long had a two-year update cycle (even year data collection, with data released in the following odd year), while local MSAS and county route counts are generally on a 4-year cycle spread out unless the county or jurisdiction does it themselves more often.

    From what I can tell, St. Paul (and Ramsey County) were due for an update in 2016…most recent local/county counts were in 2012.

    • Al Davison
      Al Davison June 30, 2016 at 5:47 pm #

      I used a mixture of ArcGIS and QGIS. QGIS deals with large datasets better (at least from personal experience, so I used that to query out roadway segments outside of Saint Paul’s city limits. ArcGIS likes to lag when I have layers with data on a statewide scale. QGIS usually doesn’t, though it does occasionally crash on me at times when handling the data. I have both programs installed on my SSD, so do they run faster than on my work PC (we stick to HDDs at least for now). I’m still not used to QGIS enough, so I ended up making the maps in ArcGIS and then exporting them to PNGs or PDFs. I’m trying to rely more on QGIS since I want to learn more open source software in general.

      Regarding the counts, I saw data years listed from all the way back in 2008 in my table, so I added that note to list a potential fault in my data analysis. Even the most recent counts from 2012-14 might be quite different from AADT counts in 2016 (Cayuga especially) and upcoming years due to large construction projects such as along 35E and US 52 affecting the nearby roadways.