Replacing Doyle Drive

Past and present solutions
January 1, 2004

Doyle Drive's concrete is crumbling in the salt-moist air of the Golden Gate. Inside its roadway, the rebar rusts. It has been patched and reinforced, and patched again. In places, its lanes are barely wider than one of the 270 buses that travel it each day. Oncoming traffic in the fast lanes is separated by a row of plastic pylons.

Doyle Drive is the one-and-a-half mile southern approach road for the Golden Gate Bridge, from the tollbooths to the Palace of Fine Arts. During an average weekday it carries over 144,000 travelers. But San Franciscans have complained about its deficiencies for decades. In 1970, when a Porsche going more than 100 mph plowed across the pylons, killing ten, the newspapers screamed "blood alley." But Doyle Drive is indispensable. Without a southern approach, the Golden Gate Bridge would have to be closed. The possibility of structural failure was real: in 1993, USA Today reported that the elevated Doyle Drive was the fifth most dangerous bridge in America. Caltrans proposed an eight-lane freeway. San Franciscans refused.

The Loma Prieta earthquake of 1989 heightened concerns, because Doyle Drive's east end is built on the same bay fill that liquefied beneath buildings in the adjacent Marina District. Two years later the city formed a task force of citizens and agencies to seek a consensus design.

Drawing The Line

The camps formed quickly. The traffic engineers and The California Highway Patrol wanted a fast and safe freeway. Environmentalists and the National Park Service, which was due to inherit the Presidio from the Army, wanted no impacts to the land and its historical buildings and spaces. Marina Boulevard residents wanted the Marina exit closed, sending all traffic to Lombard Street. Homeowners on Richardson Avenue, the arterial joining Lombard to Doyle Drive, wanted to send all traffic to the Marina exit. Collectively, neighbors hoped that traffic from the bridge would enter the Presidio, find its way to the Main Post, and disappear down a rabbit hole. Each weekly meeting brought more disagreements. The impasse continued for nearly a year. Some task force members hammered the same speech every week. The threat that kept them meeting was that if they couldn't agree on a design, Caltrans would build an eight-lane freeway.

Fresh Eyes

One day a landscape architect, a spectator at several sessions, asked to speak. Michael Painter years before had transformed the Great Highway along the city's ocean edge from a speedway regularly closed by sandstorms to a seaside arterial lined with recreational trails. He pinned a 20-foot-long, hand-drawn plan to the wall. He proposed a vision of a parkway winding gracefully through what would soon be national parkland. The focus would be great views of the bridge, the bay, and Bernard Maybeck's Palace of Fine Arts. Opposite the tranquil national cemetery and the formal Main Post, the road and its noisy traffic would be hidden in short, cut-and-cover tunnels with landscaped tops that people could walk across. The Presidio and Crissy Field would be visually and physically reunited. The contentious traffic balance between the Marina and Lombard ramps would remain the same--with most traffic encouraged to use commercial Lombard St. and the overflow to residential and recreational Marina Blvd. A new Presidio turnoff would keep park traffic out of the neighborhoods. The presentation took twenty minutes. Everyone loved it (see before-and-after images of Painter's design).

Painter's vision was to see the highway and its setting as one integrated form, and to emphasize the best qualities of both. He was able to turn the task force's sometimes fanciful wishes into design solutions and he would go on to refine his concept for a decade, for free.

United around a good design, the task force took just three months to write and present its report, A Scenic Parkway for the Park (Feb 2, 1993), which the Board of Supervisors adopted as city policy.

To get the project funded, Caltrans had to agree on what would be built. A team of San Francisco negotiators, including representatives of the Mayor's office; heads of the Doyle Drive Task Force; and José-Luis Moscovich, newly hired by the San Francisco County Transportation Authority, which allocates sales tax revenues to local transportation projects; met around a large conference table at Caltrans' district headquarters surrounded by dozens of silently watching engineers, and pushed for the road through the Presidio detailed in its report. Caltrans initially insisted that the project had to be built to eight-lane freeway standards. The San Franciscans were adamant for their scenic, six-lane parkway. "It was a food fight," Moscovich recalls. In the end, Caltrans conceded that traffic studies didn't justify the need for four lanes westbound, and a six-lane design plus an eastbound "auxiliary" lane began creeping up the state's highway priority list.

Finding the Funding

Moscovich had three goals: gain some local control so that the project would be more responsive to local needs, get a consensus road design, and find the design and build money. He got the City's Transportation Authority, instead of Caltrans, designated as the lead agency. But when he approached the Mayor's office for a city budget earmark, he was told that funding was a Caltrans problem. "At that moment, I knew we had an orphan project," he said. "Doyle Drive would be a victim of misunderstandings of how transportation funding had changed over 30 years." In the heyday of freeway building after World War II, the funding came from state and federal taxes. But over the years, as local demands for roads outstripped available funds, projects only moved up the priority lists after showing "local commitment," bureaucratic code for local money.

To counter the claim that Doyle Drive was just a San Francisco road, Moscovich commissioned a study that showed that if Doyle Drive failed, it would devastate Golden Gate Transit's bus service, and major traffic impacts would reverberate from Santa Rosa to San Jose. That made Doyle Drive a regional problem.

Enough money became available in 1999 to engineer the concept and to study its environmental impacts. Over the next three years, starting from Michael Painter's concept, engineering consultants Parsons Brinckerhoff developed four alternative designs, to be commented on by a newly formed citizens committee, and another committee of stakeholder agencies representing park, bridge district, transit, and city traffic interests.

But the engineering alternatives bore only a passing resemblance to Painter's concept. The two short tunnels had been replaced by one more than three times as long and four times as deep. The focus on great views was mostly lost. Two alternatives had 30-foot high bridges crossing a small but precious creek called Tennessee Hollow which was slated for restoration. The wide roadway required removal of many buildings, including historic structures, and affected some sensitive natural areas. The Presidio exit intruded heavily on the eastern end of the national park. Most citizen and agency representatives resigned themselves to deciding which was the least bad design.

What had gone wrong? How had Michael Painter's brilliant concept, endorsed by the original task force and adopted by the San Francisco Board of Supervisors, gotten lost?

Competing Interests

Big highway projects can easily become pork barrels for locally demanded, nice-to-have enhancements. To avoid this, the engineers focused primarily on the road and its geometry. Unfortunately, that made it easy for them to see many of a landscape architect's ideas as mere decoration to be added after the highway was designed. Strike one.

It's also easy for any large organization with a strong track record to believe that it knows the best way to do the job, and thus be less open to outside ideas. And sometimes, big engineers just produce big engineered solutions. Cost estimates for some of the alternatives were over $600 million, and this for a road just over a mile long. Strike two.

Every affected agency saw Doyle Drive through its own filter of laws, regulations, and internal cultural norms. Caltrans and the Federal Highway Administration felt mandated to build the project to national and state highway standards. The Presidio Trust and the National Park Service felt mandated to preserve the Presidio's every architectural, historic, and natural resource. And both sides felt compelled to prove that they had taken every possible step to honor their mandates (or, others felt, to cover their backsides). Case in point: what if Doyle Drive were torn down and not replaced? Voilá! No impacts on the national park! As a corollary, all the Presidio's eastern neighbors would get their cherished hope of no traffic, though at the expense of the city's western neighborhoods. An expensive traffic simulation showed the obvious: all traffic from the bridge would be rerouted down Park Presidio Boulevard and 19th Avenue, requiring both to be widened by six lanes, with up to five additional left turn lanes so traffic could get downtown.

Back in the real world, the engineers struggled to find a route for a road that had swelled to twice the width of the existing one (to meet highway standards), through a minefield of "don't touch that building; don't change that street alignment; don't affect the hydrology within that bluff when you excavate that tunnel," and so on. Strike three.

Back to the Board

By late 2001, Moscovich, now Executive Director of the Transportation Authority, had considerable success in getting the project funded. But from a design perspective, SPUR could see trouble. None of the alternatives seemed likely to create much public enthusiasm, and they were all too expensive to compete effectively for state funds. For a fresh look, we assembled a committee of urban planners, architects, engineers, environmentalists, and transportation experts to review all the designs. And we asked Michael Painter to share his latest ideas.

Through the years Painter had maintained his vision of a parkway integrated with its landscape, even as the engineered alternatives marginalized it. He had also added two big improvements: a redesign of Doyle Drive's interchange with Park Presidio Boulevard which set it back from historic buildings on Crissy Field and Cavalry Valley, and a construction technique that kept most of the existing highway in place while the new one was built. The engineers had rejected his parallel construction plan, but Painter still believed it was feasible. If he were right, the project would save millions on temporary bypass roads, and years of construction time. Painter had also embraced an idea from the Presidio Trust's landscape architect, Michael Boland, to replace the engineers' 30- foot high bridges over Tennessee Hollow with less obtrusive causeways skimming a few feet above the creek. To make that part work, the park's resource guardians agreed to alter the alignment of an historic street so that several other historic structures could be saved. So compromises were possible, if they brought benefits.

SPUR's report to the Transportation Authority was welcomed by Moscovich. The international engineering firm Arup was hired to do a full feasibility study of the parkway alternative. They concluded it could be built, with minor changes. It would have far less impact on the Presidio's resources, require fewer exceptions to highway standards, cost as much as $200 million less than the other alternatives, and cut seven years of construction to four. The stakeholder committees recommended that it be added to the environmental evaluation. Then, in November 2003, both committees overwhelmingly recommended that the other alternatives, with their greater cost and impacts, weren't worth further study.

A new Doyle Drive is closer to reality, but not there yet. Park managers want more studies of possible impacts to Presidio resources. Caltrans has come far in accepting a parkway character for the roadway, but with no national standards for what a parkway is, they are in uncomfortable bureaucratic territory. While they have agreed to narrower lane widths from the standard 12 feet (which permit 90 mph speeds on I-5), they want 10-foot continuous shoulders for emergency vehicles. At the road's eastern end, many members of the citizens advisory committee instead favor emergency turnouts, to make the road seem like a city street, not a freeway.

San Francisco voters approved the transportation sales tax extension (Prop. K) in November 2003, ensuring the City's funding for Doyle Drive. But California, in budget crisis, cancelled all new transportation funding for now. There's enough money to complete the environmental review, but not for final engineering or construction. If San Franciscans and the agencies can agree on what to build, when State money is again available, Doyle Drive will be ready to be rebuilt. Michael Painter, and the rest of us, can hardly wait.Spur logo

About the Authors: 

Michael Alexander was vice-chair of the Doyle Drive Task Force (1991-93)and the author of "A Scenic Parkway for the Park," its report to the Board of Supervisors. He is a long-time SPUR Board Member, a member of SPUR's Advisory Council, the chair of SPUR's Doyle Drive committee, and represents SPUR on the Transportation Authority's Doyle Drive subcommittee.