Zeroing in on Freeway Ramps

New Vision Zero efforts to make streets safer focus on an oft-neglected part of SoMa’s infrastructure.

The intersection of Fifth and Bryant has faded crosswalks, making it more dangerous for pedestrians crossing the freeway on-ramp. | Photo by Nuala Sawyer

It was shortly after 1 a.m. on June 15, 2014, and Wilbert Williams was asleep in his tent. He’d set it up where many unhoused people do: next to a freeway ramp in SoMa, a spot seldom disturbed by disgruntled neighbors calling the cops. On Fifth Street between Bryant and Harrison sits a small, curved cement park, sandwiched between the two highway ramps and surrounded by a low, three-foot wall. It was this wall that Jaime Juarez flew over in his 2001 Toyota SUV that night. He crashed into Williams’ tent, immediately killing the 62-year-old as he slept.

Four years later, the ramps off Fifth Street are largely unchanged. Safe-streets advocates have for years asked for improvements to the on-and-off ramps to San Francisco’s major highways, most of which are in SoMa. Now, thanks to a new grant, 10 of those are under review for some serious safety improvements.

Freeway ramps are far less sexy projects to tackle than other troublesome streets in San Francisco, yet every one of the 10 under study sits on a high-injury corridor, or the 13 percent of city streets that account for 70 percent of serious or fatal traffic collisions. Each has its own dangers: faded or nonexistent pedestrian or bicycle infrastructure, yield signs instead of stop lights, long crossing distances for pedestrians, or steep, short ramps that don’t give much time for reduced speed — all of which can be compounded by the gloomy darkness of overhead freeways.

It’s clear when looking at most of these ramps that they were designed with a vehicle-first mentality, despite the fact that they exist in densely packed neighborhoods. But even there, they often fail. Stand at the Fifth and Bryant intersection during rush hour and watch how many cars accidentally find themselves in the wrong lane.

There are many reasons freeway ramps have been left on the back burner, not least of which is that any changes made must be done as a collaboration between Caltrans, which supports our state’s highways, the San Francisco County Transportation Authority, and the San Francisco Municipal Transportation Agency. If you think the SFMTA is slow, try getting three transit agencies together in one room.

But change is imminent, with in-depth studies on solutions already underway. Leading much of these efforts is Colin Dentel-Post, a senior transportation planner with the SFCTA. The agency has already conducted one local freeway ramp study, and thanks to a Caltrans grant, is now focused on a second. To select the intersections, Dentel-Post and his team had the not-so-fun task of perusing police reports of collisions to carefully map out exactly what happened, where, and why. More than 800 people took a freeway survey, and the SFCTA met with community groups, sent mailers to neighbors, and tabled to catch passers-by. Through all that time-consuming information gathering, a plan is beginning to take shape.

“It’s a fun project to engage people on because they’re pretty passionate about their intersections,” Dentel-Post says. “We got some good ideas from people, like flashing beacons at off ramps where there isn’t a traffic light, like at 13th and Mission, or on Seventh and Eighth where there are yield signs.”

One of those groups collaborating with the SFCTA to determine pedestrian safety improvements is Walk SF. Community Organizer Natasha Opfell brings a critical walking-focused approach to the designs, which she says are “an amazing start.”

The 10 intersections under review by the SFCTA.

Incorporated into all of the design drafts so far, she notes, are data-backed improvements. There’s a 47-percent reduction in collisions at zebra-striped crosswalks over non-striped ones, for example. When there’s a good, solid cement refuge island in the middle of a crosswalk, that can reduce collisions by 56 percent.

Science aside, some other existing design flaws are just common sense. At the Eighth and Bryant on-ramp, only three sides of the intersection have crosswalks, meaning that if someone walks along the north side of the street they need to navigate through three full light cycles to continue their journey.

“It’s so unintuitive. You have to make the streets work for how people use them,” Opfell says, noting that people who are in a rush to work or have to catch a bus will be tempted to dash across the street, even without a crosswalk. While that might seem like a user error, “The onus is on the designers.”

When looking at all of these intersections with freeways and neighborhoods it’s difficult to ignore the populations disproportionately affected by the dangerous designs. The Division Circle Navigation Center, at South Van Ness Avenue and 13th Street, has 126 beds adjacent to a freeway on-ramp. Sidewalks are missing around much of the property, which was originally designed for cars and not people, and the crosswalks across 13th Street nearby are wide and dark. A new Navigation Center is set to open at Fifth and Bryant — a busy freeway off-ramp — next year.

Williams was far from the only unhoused person to be killed near freeway ramps in the past few years. In 2018, at least three marginally housed or homeless people — Russell Franklin, Gregory Blackman, and Modesto Fergurdo — were killed by drivers. In 2017, ten percent of the people killed in traffic collisions were homeless, despite being only one percent of the city’s population. With the old adage that most traffic collisions occur close to home, it makes sense that our homeless population — continually swept underneath freeways and out of sight — are at high risk of being injured or killed near our freeway ramps.

But change is on the way. Dentel-Post and his team are making their community-suggested revisions to the designs, to be finalized by early next year. As for a timeline, everyone is aware that speed is of the essence. 

“We made recommendations targeted at things that could be done in the near-term for safety,” he says. “Ideally, five years or less.”

View Comments