The “complete street” movement has picked up steam as urban planners and city leaders have relearned that public rights-of-way need to accommodate all modes of transportation. With the growing interest in this urban design strategy I tried my hand at it. This proposal reconfigures 2.4 miles of 15th Avenue NE in northeast Seattle in order to accommodate travel by pedestrians, bicyclists, transit, and cars. The result is a practical multi-modal corridor that can be replicated across the rest of the city.
Why 15th Avenue? It’s one of several access routes to the University of Washington and connects to other major streets. There are few commercial destinations along it and many people traveling in this corridor are going from one end to the other. Most of the street is four lanes wide, but from my own observations and calculations this is excessive. A road diet is implemented, which reduces the number and width of vehicle lanes. The goals for this project became the following:
- Create a safe and comfortable bicycle route
- Ensure transit is efficient and easy to access
- Provide a safe and pleasant walking experience
- Minimize the potential for vehicular collisions
- Maintain driveway access to residences, businesses, and institutions
- Preserve on-street parking where possible
- Preserve or add new street trees and vegetation
The design area is from from Pacific Street and the Burke-Gilman Trail in the south to Lake City Way in the north, a 2.4 mile stretch. Based on right-of-way widths, the lower segment is defined as Pacific Street to 45th Street, the middle segment as 45th Street to 62nd Street, and the upper segment as 62nd Street to Lake City Way. 15th Avenue’s right-of-way ranges from 55 to 80 feet wide, the cross streets range from 55 to 70 feet wide, and the blocks range from 160 to 700 feet long. So, almost every intersection and block has a unique design, but there is consistency to ensure smooth travel. These measurements were acquired by measuring the distance between property lines in a parcel file from Seattle’s GIS (geographic information system). Each segment has a typical configuration (click the graphic below), and cross street layouts are mostly left as they are.
However, the standard configurations were modified based on changing conditions. For two blocks around Campus Parkway, for instance, there are three major bus stops that see high volumes of passengers throughout most of the day; four travel lanes were kept in this area to accommodate heavy bus traffic. In some areas this becomes five lanes wide because of a left turn pocket, which reduces the sidewalk and buffer widths; the solution reroutes the bike path behind the sidewalk in these areas. If these blocks were lined with private residences or businesses that wouldn’t be feasible, but fortunately there is ample pedestrian space on the UW’s property around bus stops. Also, the cross street Cowen Place has a 55 feet wide roadway for four lanes and a tiny pedestrian island as a bus stop. The bus stop is moved one block south and the intersection is filled in; this safety measure acknowledges the adjacent Cowen and Ravenna Parks, which are popular with families with children.
The corridor is fairly well used by cyclists, but sharing lanes with vehicles is beyond most peoples’ comfort level. Parts of this corridor have been designated for improvements by Seattle’s Bike Master Plan, but this design goes further than what is recommended. The Urban Street Design Guide, published by the National Association of City Transportation Officials, was referenced for ideas on lane widths, intersection designs, and other features.
Travel Lanes and Parking
Cars are given the lowest priority because the point of a complete street is to facilitate all travel options; the corridor also carries a low amount of traffic to begin with. However, by preserving the current on-street parking restrictions at driveways there will no impact on vehicles’ ability to enter private properties. The few loading zones in the lower segment can be similarly retained. Because they have little use, the northbound left turn pockets at 41st, 42nd, and 43rd Street are eliminated, which will also streamline that segment. Southbound left turn pockets are maintained at 41st and 40th Streets because these are both major entries to the UW campus.
The entire corridor is reduced to one travel lane in each direction (with peak hour exceptions noted below). I opted not to go with a center turn lane, as is common in road diets for four lane roads, because there are very few commercial driveways on the corridor. Most travel lanes are 11 feet wide and parking lanes are 8 feet wide. The outside travel lanes in the lower segment are 10 feet wide, which is just enough for buses.
In the lower segment, below 45th Street, weekday parking will be prohibited southbound during the morning peak (7-9 AM) and northbound during the afternoon peak (4-6 PM) to provide two travel lanes for heavy bus and car commuter traffic to and from the UW. This is based on the corridor’s current parking setup, as noted above, which unnecessarily has the same restrictions on the entire length.
Two permanent parking lanes are created in the upper segment. There is only one parking lane in the middle segment because there is not enough room; there were two parking lanes in an earlier design with 10 feet wide travel lanes, which is too narrow for opposing traffic. This will actually not affect parking capacity in the middle segment area; during the current peak period restrictions owners already move their cars onto side streets and private lots, so there is obviously excess parking capacity nearby.
No effect on travel times or congestion is projected due to these changes. According to the Highway Capacity Manual, even at the lowest service level (E) for “principal arterials” (see Seattle street classifications) the four-lane lower segment of 15th Avenue NE can carry a maximum of 3,080 vehicles per hour. In 2012, the latest year for which data is available, this segment averaged only 612 vehicles per hour. And more importantly, the traffic volume is on a downward trend over the last decade (see chart above). The northern sections are similar; 45th to 65th Street has a two-lane capacity of 1,520 vehicles per hour but averages only 408 vehicles per hour. (Of course this does not consider that traffic volume varies throughout the day, but streets should not be designed for the few hours that peak times represent.) Clearly there is excess capacity and travel lanes can be reduced without adverse impacts. Multiple studies of road diets confirm similar findings, such as this one in Calgary. This project may actually reduce traffic by drawing people to bikes and buses, but a comprehensive citywide network of complete streets is needed to continue progress on reducing single-occupant-vehicle use.
If some drivers coming from the north are hampered by reduced travel speeds, they can easily use two nearby alternate routes. 25th Avenue NE also provides access to parking on the UW campus, and 11th Avenue and Roosevelt Avenue are both speedy one-way arterials that connect to Lake City Way and north Seattle.
Seattle’s current system of protected bike lanes and cycle tracks is a confusing patchwork that has contributed to collisions. The U-District has also had its share of bicycle accidents over the last few years. So, the new design has 4 to 5 feet wide bike lanes next to the curb in each direction for the entire route. I opted against a two-way cycle track on one side of the street in order to keep consistent directionality and ensure predictability for all users. Cyclists will be protected from traveling vehicles with a parking lane, a 6 to 12 inch wide curb, or a planting strip up to 6 feet wide. Missing is the 2-3 feet door zone common on complete streets because I assume passenger side doors are used less. There is also no conflict between cyclists and drivers maneuvering their cars in and out of the parking lane.
At intersections, cyclists will ideally use the pedestrian crosswalks but can use the roadway where it is safe to do so. As noted below, in most places where pedestrians cross the bike path there are 15 feet long ramps to meet the level of the sidewalk. This avoids two successive curb cuts.
The entire corridor has bus routes on it, though none run continuously from one end to the other. On-time transit is vital for the many UW students who ride in this area, so the design ensures that buses can move quickly between stops by eliminating the need to merge in and out of traffic. This is achieved with “bus-bulbs”, which are extensions of the sidewalk that allow the bus to stop in the travel lane to pick up and drop off passengers. In most areas, with only one lane per direction these bus-bulbs will cause stopped buses to stop traffic. But by creating a gap in traffic, this allows the bus to quickly move forward to the next stop. Allowing cars to pass buses (with a yellow dashed center line) may be feasible in lower traffic areas.
Most bulbs are 8 feet wide, but in the middle segment a few are only 6 feet wide due to space constraints. The bulbs can also host shelters, seating, lighting, and plantings. All of the bulbs are 60 feet long to accommodate a full-length bendy bus, and a path across the adjacent bike lane aligns with the rear doors. At some intersections the bike path is also raised at the front, and a lean bar along the bike path channels pedestrians into the crossings.
Excluding bus stops with significant existing infrastructure, most stops are moved to the “near side” of intersections. This is the side of the intersection that is encountered by a vehicle before traveling through. This is better than “far side” stops because it eliminates double stops for buses after red lights and prevents cars from backing up into intersections with only one travel lane. At uncontrolled intersections, having both stops on the near sides allows waiting passengers to quickly cross the street to the opposite bus stop if needed.
Where needed, sidewalk widths are increased and marked crosswalks are at every intersection. New mid-block crossings are also created where the block is 500 feet or longer. Sidewalk “bulb-outs”, which lessen the roadway distance pedestrians must cross, are at all intersections and mid-block crossings. The bulb-outs are 30 feet long at intersections and 20 feet long at mid-blocks in order to preserve sight lines for drivers. They also cap the ends of parking lanes, which emphasizes that parking lanes are not for travel, and provide space for new street trees and vegetation. Wide curb cuts ensure accessibility for wheelchairs, strollers, and other wheeled devices. At intersections the sidewalks have curved corners with only a 2 feet radius, which forces vehicles to slow down when turning and lessens the width of roadway that pedestrians and bicyclists are exposed to traffic.
Stormwater systems were not explicitly included in this plan, but it it would be feasible to use the wider planting strips as rain gardens for controlling and filtering stormwater into the ground. Drainage and piping infrastructure is already in place, so it would be less of a cost-saving than an environmental measure. Another consideration is permeable concrete and asphalt, which could be used everywhere and would also allow rainwater to filter into the ground. In addition to plantings, the buffer areas could host small public spaces with seating or play areas.
There are some limitations to the design, but cost is not one of them. A similar project on Dexter Avenue in Seattle only cost $1.8 million/mile. I estimate this probably would be significantly less, and could perhaps be done at $1 million/mile or less. And in fact, complete streets can save money by reducing maintenance costs of vehicle paths and possibly allow removal of traffic signals (I would suggest this at the 47th Street intersection). Complete streets are typically built in commercial areas and the project area is mostly residential, but the U-District is being planned for intensive growth over the coming decades with more apartments and possibly taller office buildings.
Curb barriers for crosswalks at intersections are probably too long for most vehicles to make a right-hand turn without hitting them. I haven’t shown how driveways would remain accessible, but they would simply have curb cuts across any adjacent raised areas. The actual CAD drawing is not terribly detailed and does not show signage, drains, grading, street poles, or trees, and likely doesn’t meet engineering standards (but, again, I’m not being paid for this). If you can view .dwg files and are interested in seeing it, drop me a line.
I believe this project could be a catalyst for change. Encouraging alternative transportation reduces the external costs of car travel, including air pollution, congestion, and risk to life and limb. Seattle and its residents must push for a city-wide system of complete streets and bike paths in order to make this a worthwhile investment. Together we can make Seattle a friendly place for all modes of travel.