Most bridge news is about building something new or explaining why something failed.
The Golden Gate Bridge offers a third story, and arguably a more useful one for asset owners: what it actually costs to keep a landmark standing.
The scope of the commitment
The seismic retrofit of the Golden Gate Bridge has been running for nearly 30 years.
The final stage, Phase 3B, begins in early 2026 and is scheduled to conclude in 2036, a further decade of work.
The total project cost now exceeds $1.8 billion, with roughly $1 billion in additional funding approved to offset rising construction costs and inflation.
Phase 3B is split into two construction manager/general contractor contracts.
The first, worth $864 million, went to Halmar International, with a $41.3 million contingency fund attached and a $141.7 million budget increase drawn from district reserves.
The second subphase is expected to cost around $900 million and will begin immediately after the first, adding five years to the six-year opening contract.
For context, the previous stage, Phase 3A, cost $125 million and ran from 2008 to 2014 on the north anchorage housing and pylon N1.
The escalation between those figures tells its own story about infrastructure cost inflation.
Why the work is necessary
The bridge sits approximately six miles from the San Andreas Fault.
After the 1989 Loma Prieta earthquake damaged the structure, the district determined that a retrofit was needed.
Built between 1933 and 1937, the bridge measures 1.7 miles between abutments, with the main span and side spans covering 1.2 miles.
It is the second-longest suspension bridge in the Americas, behind New York's Verrazzano-Narrows.
Its current maximum downward deflection is 10.8 feet, with a maximum upward deflection of 5.8 feet.
What Phase 3B actually involves
The scope reads as a catalogue of modern retrofit practice.
Foundation work includes strengthening, the installation of micropiles and rock bolts, and the construction of reinforced concrete shearwalls.
The main span between the two towers and both side spans will be strengthened, along with both towers, the tower trusses and the south tower pier.
The tower bases receive new steel plates.
Expansion joints where the roadway meets the towers will be replaced, as will the pylons, specifically to allow three-dimensional movements during earthquakes.
Crews will reinforce lateral bracing, retrofit 255 floor beams, and install 38 energy-dissipation devices.
The housing roof and roadway deck will be replaced with a pre-cast concrete slab-on-steel stringer deck system.
The south tower will be repainted above the roadway.
The constraint that shapes everything
The bridge remains open throughout, with the deck replacement and other structural modifications carried out under nighttime lane closures.
That single condition governs the entire construction methodology.
It dictates the choice of a pre-cast deck system, the sequencing of joint replacement, and the productivity assumptions behind a ten-year programme.
Any engineer who has worked on a live structure will recognise the trade: the owner protects revenue and public access, and pays for it in schedule.
The questions worth asking
Three points make this project worth watching beyond San Francisco.
First, a retrofit programme running from the 1990s into the 2030s raises the question of whether staged, funding-dependent retrofits are genuinely cheaper than decisive intervention, or simply spread the same cost across more decades and more inflation.
Second, the energy-dissipation devices and three-dimensional expansion joints reflect a shift from strength-based to performance-based seismic design, which many agencies with 1930s-era assets have yet to make.
Third, and most uncomfortably, the district secured funding for only about half the total cost at the point Phase 3B was announced.
For owners of ageing landmark structures anywhere, that gap between what the engineering requires and what the budget covers is the recurring problem.
← All briefs