The Bay Bridge Bolts That Broke the Rules They Followed

In September 2013, the new eastern span of the San Francisco-Oakland Bay Bridge opened to traffic after years of construction and a $6.5 billion price tag. It was an engineering achievement — a self-anchored suspension bridge designed to withstand a major earthquake, carrying over 240,000 vehicles a day across one of the most seismically active corridors in the United States.

Six months earlier, 32 of its anchor bolts had already snapped.

The bolts didn't fail because they were cheap. They didn't fail because someone substituted a lower grade. They met every spec they were given. They failed because nobody accounted for what the environment would do to them once they were under load — and by the time it became clear, the bridge was already open.

What the Bolts Were Supposed to Do

The new Bay Bridge span was designed with a system of shear keys — concrete blocks installed at piers to dampen seismic energy and reduce the amplification of motion through the bridge deck during an earthquake. Each shear key was anchored by large-diameter rods, some up to four inches across, made from ASTM A354 Grade BD steel — a high-strength, heat-treated alloy rated to a minimum tensile strength of 140,000 psi.

These were serious bolts for a serious application. They were hot-dip galvanized for corrosion protection, and they were manufactured to meet the mechanical property requirements of their ASTM specification. On paper, the selection was correct.

The rods were installed into the shear keys in 2008, but couldn't be fully tensioned until the superstructure of the bridge was in place. That happened in March 2013. When engineers tensioned the rods to 70% of their minimum specified ultimate tensile strength — a standard pre-tension level for structural anchors — 32 of the 96 rods fractured within two weeks. All failures occurred at or near the threaded section.

When the pre-tension was reduced to 40%, the failures stopped.

Source: Thomas Langill, Ph.D., "Lessons Learned from the Bay Bridge Bolt Failure," Structure Magazine, February 2017

What Went Wrong

Three independent investigators conducted metallurgical analysis on the failed rods and landed on the same conclusion: hydrogen embrittlement.

Hydrogen embrittlement is what happens when hydrogen atoms diffuse into high-strength steel under load and attack the grain structure from the inside. The steel becomes brittle — not visibly, not immediately, but progressively. Microscopic cracks initiate at the thread roots, grow slowly under sustained tension, and then fracture suddenly when they reach a critical size. The bolt doesn't bend. It doesn't stretch. It snaps.

The investigation revealed that the 2008 batch of rods had microstructural inhomogeneity — uneven grain structure from the fabrication and heat treatment process — that left them with lower-than-expected toughness. Their Charpy V-notch impact test values came in at 13-18, well below the expected range of 25-35. They were technically within the A354BD mechanical property spec, but at the susceptible end of the material's range.

The environment did the rest. The top sections of the rod sleeve assemblies were exposed to the Bay environment for nearly five years before tensioning — saltwater air, moisture intrusion, the works. When the rods were finally tensioned to 70% Fu, the combination of sustained high load, susceptible microstructure, and environmental hydrogen diffusion was enough to initiate cracking.

The Townsend Test — a controlled lab test that slowly loads rods while immersed in saltwater — later confirmed it: the 2008 rods failed at exactly the same load (0.70 Fu) that caused failure on the bridge. When the same test was conducted in air, without saltwater, the rods did not fail.

Without saltwater exposure, they would have held.

The Fastener Lesson

What makes the Bay Bridge failure so instructive is that the bolts were not obviously wrong. They met spec. They were the specified material. They were hot-dip galvanized. And they still failed — because the interaction between high strength, high sustained load, and a corrosive environment wasn't fully accounted for in the selection and installation process.

High-strength fasteners and hydrogen embrittlement have a well-understood relationship in materials engineering. The higher the tensile strength, the more susceptible the steel becomes to hydrogen diffusion under load. For fasteners above roughly 150,000 psi tensile strength — which A354BD approaches — environmental controls aren't optional. They're part of the specification.

Three things went wrong that are directly applicable to any fastener application involving high-strength steel in a wet or corrosive environment:

Exposure time before tensioning was not treated as a risk factor. The rods sat in a marine environment for nearly five years before being loaded. That's five years of potential hydrogen absorption in a susceptible material. Protecting high-strength fasteners from environmental exposure between installation and tensioning is not just good practice — in corrosive environments, it's a requirement.

Microstructure variability within spec is still variability. The rods complied with A354BD. But compliance with a minimum spec doesn't mean uniform performance. The 2008 batch was at the low end of toughness. For critical structural applications in aggressive environments, specifying to the minimum is not the same as specifying for the environment.

The environment is part of the fastener selection. A354BD is appropriate for many structural applications. But high-strength anchor rods under sustained tension in a saltwater environment require supplemental corrosion protection, controlled installation sequencing, and careful attention to the interaction between load and exposure. The spec tells you what the bolt can do in isolation. The application tells you what it will face in the real world.

The remaining A354BD rods throughout the bridge were tested and confirmed safe — their hydrogen embrittlement thresholds exceeded their design loads. Supplemental protections including dehumidification systems, additional paint systems, and grout were added. A replacement anchoring system was designed and installed for the failed shear key locations.

The fix worked. But the cost — in engineering hours, investigation, redesign, and public scrutiny of a brand-new bridge — was significant.

The Takeaway

The Bay Bridge bolt failure is a reminder that fastener selection doesn't end with the grade designation. Strength grade tells you what a bolt can carry. Environment tells you whether it will still be carrying it in ten years.

High-strength steel, sustained load, and moisture are a known combination that demands attention. If your application involves any two of those three, the third one needs to be part of your selection criteria — not an afterthought.

The bolts on the Bay Bridge did everything they were specified to do. The specification just didn't account for everything the bolts would face.