How Bridges Are Built to Survive Earthquakes: Engineering Guide 2025

On October 17, 1989, at 5:04 PM, the Loma Prieta earthquake struck Northern California. On the San Francisco-Oakland Bay Bridge, a 50-foot section of the upper deck collapsed onto the lower deck, killing one person and closing this vital link for a month. Just 40 miles south, dozens of other bridges suffered catastrophic damage. The Cypress Street Viaduct—a double-deck freeway structure—pancaked completely, crushing 42 people in their vehicles. Yet remarkably, many bridges in the same earthquake, experiencing similar ground shaking, emerged essentially unscathed.

What separated survival from collapse? Engineering. Specifically, seismic engineering principles that determine whether a bridge rides out violent shaking or fails catastrophically. The difference between a bridge designed to resist earthquakes and one that isn't can be measured in lives saved, economic continuity maintained, and disaster response enabled—or in deaths, economic paralysis, and communities cut off from aid.

Bridges present unique earthquake engineering challenges. Unlike buildings, which sit on foundations at a single location, bridges span gaps—sometimes vast distances—connecting points that move independently during earthquakes. A bridge over a river or bay may experience different ground motions at each end. Long bridges can see earthquake waves traveling along their length, creating complex motion patterns. And because bridges are critical infrastructure—lifelines for transportation, commerce, and emergency response—their failure has cascading consequences far beyond the structure itself.

This comprehensive guide explores how modern bridges are engineered to survive earthquakes, why older bridges fail in patterns engineers can now predict, the evolution of seismic bridge design through deadly disasters, specific engineering solutions including seismic isolation bearings and ductile column design, the challenge of retrofitting vulnerable existing bridges, how different bridge types (suspension, cable-stayed, arch, beam) respond to earthquakes, ground failure risks including liquefaction and fault rupture, inspection and monitoring technologies, and real-world performance data from major earthquakes proving which approaches work.

Why Bridges Fail in Earthquakes: Historical Lessons

The Patterns of Bridge Collapse

Span collapse (complete failure):

• Bridge deck falls off its supports
• Most catastrophic failure mode
• Often caused by inadequate connection between deck and piers
• Deck literally slides or bounces off its supports during shaking
• Results in complete bridge loss and often fatalities

Column failure:

• Vertical support columns crack, crush, or topple
• Brittle failure—sudden collapse without warning
• Common in older bridges with inadequate reinforcement
• Can lead to progressive collapse as remaining columns overload

Bearing failure:

• Connection devices between deck and substructure fail
• Allows excessive movement or displacement
• Can progress to span collapse if unchecked

Foundation failure:

• Liquefaction causes foundation to sink or tilt
• Lateral spreading moves bridge piers
• Scour from water flow during/after earthquake undermines foundations

1971 San Fernando Earthquake: Wake-Up Call for Bridge Engineering

The disaster:

• Magnitude 6.6 earthquake in Los Angeles area
• Multiple freeway overpass collapses
• 64 people killed, many in collapsed structures
• Revealed systemic vulnerabilities in bridge design

Specific failures:

Foothill Freeway (I-210) overpasses:

• Four overpasses collapsed completely
• Precast concrete spans fell from their supports
• Problem: Inadequate seat width (deck rested on narrow ledge)
• During shaking, deck shifted beyond support and fell
• Simple support = no redundancy—once unseated, nothing prevented fall

What engineers learned:

• Seat width matters enormously—wide seats prevent unseating
• Need positive connection between deck and substructure (not just gravity)
• Column ductility essential—brittle columns fail catastrophically
• Multi-span continuous bridges more vulnerable than thought

Code changes: California completely revised bridge design standards. Introduced requirements for ductile column design, adequate seat widths, and restrainer cables.

1989 Loma Prieta Earthquake: Modern Bridges Tested

The disaster:

• Magnitude 6.9 in San Francisco Bay Area
• 63 deaths total, many in bridge/freeway collapses
• $6 billion in damage
• Exposed continuing vulnerabilities despite 1971 code improvements

Cypress Street Viaduct (I-880):

• Double-deck freeway structure collapsed for 1.25 miles
• Upper deck pancaked onto lower deck
• 42 people crushed in vehicles
• Why it failed: Built 1957, pre-seismic design era. Inadequate column reinforcement, poor connections between deck sections, built on soft soil that amplified shaking. Columns failed in brittle shear—no ductility.
• Impact: Structure had been flagged as vulnerable but retrofit hadn't occurred. Demonstrated need for proactive retrofit programs.

San Francisco-Oakland Bay Bridge:

• 50-foot section of upper deck collapsed onto lower deck
• One fatality
• Bridge closed for 30 days—major economic impact
• Why it failed: Connection failure at expansion joint. Bolts sheared, allowing section to drop. Built 1936, pre-seismic standards.
• Impact: Led to major retrofit of entire bridge and eventually complete replacement of eastern span (completed 2013 with modern seismic design)

Bridges that survived:

• Golden Gate Bridge: Minor damage only, suspension design provided flexibility
• Newer bridges with post-1971 design: Generally performed well
• Validated importance of ductile design and proper connections

1994 Northridge Earthquake: Unexpected Failures

The shock:

• Magnitude 6.7 in Los Angeles area
• Seven major freeway bridges collapsed
• Shocking because many were "modern" bridges built post-1971 with supposedly adequate seismic design
• Revealed that even improved codes had gaps

I-5/SR-14 Interchange collapse:

• Connector ramps collapsed despite being post-1971 construction
• Column failures—inadequate confinement of concrete
• Problem: Columns were "ductile" on paper but under extreme shaking, confinement steel (spiral reinforcement) wasn't adequate
• Revealed need for better understanding of column performance under strong shaking

What this taught engineers:

• "Code-compliant" doesn't guarantee earthquake-proof
• Real earthquake forces can exceed design assumptions
• Column confinement requirements needed strengthening
• Near-fault ground motions create unique challenges
• Continuous improvement mindset essential—codes must evolve

1995 Kobe Earthquake: Japan's Bridge Failures

The disaster:

• Magnitude 6.9 in heavily developed area
• 6,434 deaths total
• Elevated highways collapsed, killing hundreds
• Shocking for a country known for earthquake engineering excellence

Hanshin Expressway collapse:

• Multiple sections of elevated highway fell over sideways
• Not pancaking but complete structural collapse
• Columns toppled
• Why: Pre-1980 construction with inadequate reinforcement. Columns lacked sufficient ties (lateral reinforcement). Under strong shaking, columns failed and structures toppled.

Lessons for Japan:

• Complete revision of bridge design codes
• Massive retrofit program for existing bridges
• Recognition that building stock vulnerability remains despite new construction standards

Modern Bridge Seismic Design Principles

The Hierarchy of Seismic Design

Performance objectives (what engineers aim for):

Small earthquakes (frequent, M4-5):

• Bridge remains fully operational
• No damage or cosmetic damage only
• No traffic disruption

Moderate earthquakes (occasional, M5.5-6.5):

• Bridge remains operational with possible speed restrictions
• Repairable damage acceptable (cracked concrete, minor spalling)
• Inspection required but no extended closure

Large earthquakes (rare, M6.5-7.5):

• Bridge remains standing and passable for emergency vehicles
• Significant damage acceptable but no collapse
• May require closure for repairs but emergency access maintained
• Columns may have substantial cracking and concrete damage

Maximum credible earthquake (very rare, M7.5+):

• Prevent collapse—life safety paramount
• Severe damage acceptable as long as bridge doesn't fall
• Bridge may need to be demolished and rebuilt after
• But occupants and people beneath bridge survive

Load Path Design: How Forces Flow Through Bridge

Understanding force transmission:

1. Ground moves: Earthquake shaking moves bridge foundations
2. Piers respond: Columns or towers flex and transmit forces upward
3. Bearings mediate: Connection devices between pier and deck control force transmission
4. Deck moves: Bridge deck responds to forces transmitted through bearings
5. Forces distribute: Continuous deck distributes forces along length

Design challenge: Each component must have capacity to handle its role in load path. Weak link causes failure.

Critical load path elements:

• Foundations: Must remain stable (resist liquefaction, lateral spreading, scour)
• Columns: Must resist lateral forces without brittle failure
• Bearings: Must accommodate movement without failing or unseating deck
• Deck connections: Must keep deck sections connected
• Expansion joints: Must allow movement without failure

Seismic Isolation Bearings: Protecting Bridges

How Bridge Isolation Differs from Building Isolation

Similarities:

• Basic principle same: Decouple superstructure from ground motion
• Uses similar bearing technologies (lead-rubber, friction pendulum)
• Dramatically reduces forces transmitted to superstructure

Differences:

• Movement requirements: Bridges need larger displacement capacity (thermal expansion, live loads, plus earthquake)
• Load characteristics: Bridges have moving loads (vehicles), buildings don't
• Redundancy: Bridges typically have many bearings (one per pier-deck connection), buildings have fewer
• Inspection access: Bridge bearings more accessible for inspection than building isolators

Types of Bridge Seismic Bearings

Lead-rubber bearings (LRB):

• Same technology as building isolation
• Layers of rubber bonded to steel plates with lead core
• Rubber provides flexibility, lead provides damping
• Typical bearing: 600-1200mm diameter, 200-400mm tall
• Can support 1000-5000 kN vertical load per bearing
• Allows 300-600mm horizontal displacement
• Cost: $5,000-20,000 per bearing depending on size

Friction pendulum bearings (FP):

• Articulated slider on curved stainless steel surface
• Low friction (0.05-0.15 coefficient)
• Self-centering through gravity
• Very long displacement capacity (up to 1 meter possible)
• Excellent for long-span bridges requiring large movements
• Temperature-independent performance
• Cost: $8,000-30,000 per bearing (more expensive than LRB)

Sliding bearings with restoring mechanisms:

• Simple sliding surface (PTFE on stainless steel typically)
• Separate devices provide restoring force (springs, rubber elements)
• Very large displacement capacity
• Used for special applications (movable bridges, very long spans)

Bridge Isolation Design Considerations

Bearing placement:

• Typically one bearing assembly per pier-deck connection point
• For multi-column bent, may have bearing per column or one large bearing assembly
• Must accommodate thermal expansion and contraction (can be meters over long bridge)
• Must handle live loads (moving vehicles)
• Must survive millions of load cycles over decades

Movement accommodation:

• Bridge decks expand/contract with temperature (can be significant for long bridges)
• Bearings must allow this normal movement
• Plus accommodate earthquake displacement
• Restrainer systems prevent excessive displacement while allowing designed movement

Inspection and maintenance:

• Bearings inspected regularly (every 2-5 years typically)
• Check for damage, wear, proper alignment
• Elastomeric bearings checked for rubber cracking, hardening
• FP bearings checked for slider surface condition
• Replacement possible but expensive and complex (requires jacking bridge deck)

Performance of Isolated Bridges in Earthquakes

1989 Loma Prieta—Sierra Point Overhead:

• One of first seismically isolated bridges in California
• Lead-rubber bearings installed
• Experienced strong shaking from earthquake
• Performance: Bearings functioned perfectly. Bridge undamaged. Bearings displaced as designed, then returned to center. Validated isolation concept for bridges.

1994 Northridge—Isolated bridges:

• Several retrofitted bridges with seismic isolation
• Minimal damage compared to non-isolated bridges
• Further validation of technology

2011 Christchurch, New Zealand:

• South Rangitikei Rail Bridge: Base-isolated
• Survived multiple strong earthquakes
• Remained operational throughout disaster
• Critical for post-earthquake transportation

Ductile Column Design: Bending Without Breaking

The Concept of Ductility

Brittle vs. ductile behavior:

Brittle failure:

• Material breaks suddenly without warning
• Little deformation before failure
• Like glass breaking
• Catastrophic—once failure starts, immediate collapse
• Examples: Unreinforced concrete, poorly detailed reinforced concrete

Ductile behavior:

• Material deforms extensively before failure
• Can bend and flex without breaking
• Like bending paper clip—takes force but doesn't snap immediately
• Provides warning—visible damage before collapse
• Absorbs energy through plastic deformation
• Examples: Properly detailed reinforced concrete, structural steel

Why ductility matters for bridges:

• Earthquake forces can exceed design assumptions
• Ductile structures survive unexpected forces by deforming
• Damage visible after earthquake allows inspection and repair
• Prevents catastrophic collapse even in extreme shaking

Creating Ductile Concrete Columns

The challenge: Plain concrete is brittle. Needs steel reinforcement, but simply adding steel isn't enough. The detailing determines whether behavior is brittle or ductile.

Key design elements:

1. Confinement steel (transverse reinforcement):

• What it is: Spiral or circular hoops of steel wrapped around column
• Purpose: Confines concrete, preventing it from exploding outward under compression
• Spacing: Very close together in plastic hinge zones (where most bending occurs)—typically 2-4 inches apart
• Effect: Confined concrete can deform to 5-10 times the strain of unconfined concrete before failure
• Modern requirement: Detailed requirements in codes for amount and spacing of confinement steel

2. Longitudinal reinforcement:

• What it is: Vertical steel bars running along column length
• Purpose: Resist bending moments, provide tensile strength
• Amount: Typically 1-4% of concrete cross-section area
• Detailing: Must be properly spliced and anchored into foundation

3. Plastic hinge design:

• Concept: Designate specific location where ductile yielding will occur
• Typical location: Column base (where it meets foundation)
• Design approach: Extra confinement steel in plastic hinge zone, ensuring this is weakest point that yields first
• Effect: Controlled damage—yielding occurs where designed, protecting other elements

4. Shear capacity:

• Critical requirement: Column must not fail in shear (brittle mode) before flexural yielding (ductile mode)
• Design principle: "Capacity design"—ensure shear capacity exceeds maximum shear that can develop from flexural yielding
• Practical result: Heavy transverse reinforcement throughout column

Column Design Evolution

Pre-1971 columns:

• Widely-spaced ties (12-18 inches)
• Minimal confinement
• Brittle failure common
• Example: Cypress Viaduct collapse (1989 Loma Prieta)

1971-1990 columns:

• Introduction of ductility concept
• Improved tie spacing (6-8 inches typical)
• Better but not optimal
• Some failures in 1994 Northridge revealed continued issues

Modern columns (post-1995):

• Very closely-spaced confinement steel (2-4 inches in plastic hinge zone)
• Capacity design ensuring flexural yielding before shear failure
• Detailed requirements for reinforcement lap splices
• Excellent earthquake performance documented

Retrofitting Vulnerable Existing Bridges

The Challenge of Legacy Bridges

The problem:

• Thousands of older bridges built before modern seismic codes
• Many are critical infrastructure—can't simply close them
• Complete replacement expensive (millions to billions per bridge)
• But existing bridges highly vulnerable to earthquakes
• Retrofit needed but must maintain traffic during construction

Cost considerations:

• Major bridge retrofit: $5-50 million typical
• Complete replacement: $50-500 million or more
• Retrofit typically 20-40% of replacement cost
• But retrofit extends bridge life 30-50 years and dramatically improves safety

Common Retrofit Strategies

Column jacketing (concrete or steel):

• What it is: Wrapping columns in steel plates or additional concrete with heavy reinforcement
• Purpose: Provide confinement and shear strength old columns lack
• Process: Steel jacket welded around column, gap between steel and concrete filled with grout
• Effect: Transforms brittle column into ductile one
• Cost: $10,000-50,000 per column depending on size
• Effectiveness: Very high—one of most effective retrofits

Seismic isolation retrofit:

• What it is: Installing isolation bearings between existing deck and piers
• Challenge: Must jack up deck to install bearings
• Benefit: Reduces forces on existing vulnerable columns
• Cost: $2-10 million typical for medium bridge
• Used when: Column retrofit too difficult or existing columns acceptable if forces reduced

Restrainer cables and seat width extension:

• What they are: Steel cables connecting deck to pier, preventing unseating. Extending width of ledge deck rests on.
• Purpose: Prevent span collapse from deck sliding off supports
• Relatively inexpensive: $50,000-500,000 per bridge
• Quick to install: Minimal traffic disruption
• High effectiveness for this specific failure mode

Foundation strengthening:

• Deep foundations added if existing shallow foundations inadequate
• Soil improvement (grouting, compaction) if liquefaction risk
• Very expensive and disruptive
• Sometimes unavoidable for bridges on poor soils

Major Retrofit Programs

California bridge retrofit program (post-1989 Loma Prieta):

• Identified 1,039 bridges needing seismic retrofit
• $1.3 billion program over 15 years
• Prioritized by traffic volume and structural vulnerability
• Program largely complete by mid-2000s
• Dramatically improved bridge safety statewide

San Francisco-Oakland Bay Bridge replacement:

• After 1989 damage, eastern span deemed unrepairable to modern standards
• Complete replacement: 1997-2013
• Cost: $6.5 billion (original estimate $1.3 billion)
• New span: World-class seismic design with suspension span and self-anchored suspension configuration
• Can withstand M8.0 earthquake

Different Bridge Types and Seismic Vulnerabilities

Suspension Bridges

Design characteristics:

• Main cables suspended between towers, deck hangs from cables
• Naturally flexible structure
• Long span capability (Golden Gate: 4,200 feet main span)

Seismic advantages:

• Flexibility allows large deformations without damage
• Self-stabilizing—cable geometry provides restoring forces
• Towers can sway significantly without structural distress
• Generally perform well in earthquakes

Seismic concerns:

• Tower-deck connection requires special attention
• Anchorages (where cables attach to ground) critical—must resist enormous forces
• Long period of vibration—can resonate with long-period earthquake waves
• Differential ground motion at towers if earthquake waves travel along bridge length

Example: Golden Gate Bridge

• Survived 1989 Loma Prieta with minimal damage
• After retrofit (completed 2012), rated for M8.3
• Flexible design is actually advantageous for earthquake resistance

Cable-Stayed Bridges

Design characteristics:

• Cables run directly from towers to deck (no suspending cables)
• Stiffer than suspension bridges
• Popular for medium spans (500-2,000 feet)

Seismic advantages:

• Redundancy—many cables, failure of one doesn't cause collapse
• Clear load path through cables
• Towers and deck move more as unit

Seismic concerns:

• Less flexible than suspension bridges—higher seismic forces
• Tower-deck connection critical
• Cable anchorages must be designed for earthquake loads
• Aerodynamic instability during shaking possible

Modern cable-stayed bridges: Incorporate seismic isolation bearings, ductile tower design, and redundant load paths. Generally excellent seismic performance.

Arch Bridges

Design characteristics:

• Load carried through arch action (compression)
• Very strong and durable
• Beautiful aesthetics

Seismic advantages:

• Compressive strength—arches naturally strong in compression
• Massive structures—high inertia resists movement
• Historic arches often survived earthquakes well

Seismic concerns:

• Thrust forces—arch pushes outward on foundations, earthquake increases these forces
• Unreinforced masonry arches (historic) vulnerable to cracking
• Deck-to-arch connection requires proper design
• Foundation stability critical—any settlement catastrophic

Beam (Girder) Bridges

Design characteristics:

• Most common bridge type
• Simple spans or continuous spans over piers
• Concrete or steel beams support deck

Seismic vulnerabilities:

• Highest vulnerability of common bridge types
• Spans can unseat from piers
• Piers (columns) vulnerable if not properly designed
• Most catastrophic bridge failures have been beam bridges

Why so vulnerable:

• Simple bearing connections—deck just rests on pier
• Limited seat width in older designs
• Column failures common in pre-code bridges
• Multiple spans create complex dynamics

Modern improvements:

• Wide seat widths preventing unseating
• Restrainer cables as backup
• Ductile column design
• Seismic isolation bearings
• When designed properly, beam bridges can be very safe

Ground Failure: When the Foundation Fails

Liquefaction

What it is: Saturated sandy soil loses strength during earthquake shaking, behaving like liquid.

How it happens:

1. Loose, saturated sand below water table
2. Earthquake shaking rearranges sand grains
3. Water pressure between grains increases dramatically
4. Grains lose contact with each other
5. Soil acts like liquid—no strength

Effects on bridges:

• Bridge foundations sink or tilt
• Differential settlement—one pier sinks more than others
• Bridge deck develops kinks or breaks
• Can cause collapse even if structure itself undamaged

Example: 1964 Niigata, Japan:

• Showa Bridge collapsed due to liquefaction
• Foundations sank into liquefied soil
• Bridge tilted and fell into river
• Structure itself was undamaged—foundation failure caused collapse

Mitigation strategies:

• Deep foundations: Piles or drilled shafts extending through liquefiable soil to stable layers
• Soil improvement: Densification, grouting, or replacement of liquefiable soil
• Structural flexibility: Design to tolerate some settlement without collapse
• Avoidance: Site selection avoiding high liquefaction zones (not always possible)

Lateral Spreading

What it is: Ground flows laterally toward free face (river bank, shoreline) during or after earthquake.

How it happens:

• Liquefied soil layer below surface
• Sloping ground or nearby water body
• Liquefied layer flows downslope or toward water
• Overlying ground cracks and moves with it
• Can move several feet to tens of feet

Effects on bridges:

• Bridge approaches spread apart—gap opens between bridge and road
• Piers pushed laterally—creates huge forces on foundations
• Can break piers or move them off foundations
• Very destructive failure mode

Mitigation:

• Deep foundations resisting lateral loads
• Ground improvement (soil densification)
• Structural design accounting for lateral spreading forces
• Sometimes unavoidable—site constraints may require accepting risk

Fault Rupture

The challenge: If earthquake fault ruptures ground surface AND bridge crosses fault, bridge must accommodate sudden permanent ground offset.

Ground displacements:

• Moderate earthquakes: 1-3 feet of offset
• Large earthquakes: 10-30+ feet of offset
• Can be vertical (thrust faults) or horizontal (strike-slip faults)

Design approach:

• Avoidance preferred: Don't build bridge across active fault if possible
• If unavoidable: Special design allowing large deformations
• Articulated deck that can accommodate offset
• Sacrificial elements designed to fail in controlled way
• Expectation that bridge will be severely damaged but not collapse

The Future of Bridge Seismic Engineering

Emerging Technologies

Shape memory alloys:

• Materials that return to original shape after deformation
• Can be used in connections that "self-repair"
• Reduces residual displacement after earthquake
• Still expensive but promising for critical bridges

Advanced damping systems:

• Viscous dampers, friction dampers integrated into bridges
• Dissipate earthquake energy, reducing structural response
• Increasingly used in long-span bridges

Real-time monitoring and structural health assessment:

• Sensor networks measuring bridge response during earthquakes
• Automated damage assessment
• Helps prioritize inspection and repairs
• Informs decisions about bridge reopening

Challenges Ahead

Aging infrastructure:

• Many bridges built 50-100+ years ago
• Pre-seismic design era
• Retrofit expensive, replacement more expensive
• Funding challenges

Climate change effects:

• Rising sea levels affect foundation design
• Increased scour risk from more intense storms
• Temperature extremes affect bearing performance

Cascadia Subduction Zone (Pacific Northwest):

• Threat of M9.0+ megaquake
• Most bridges built before seismic codes
• Extensive vulnerability
• Retrofit program ongoing but will take decades

Conclusion: Bridge Seismic Engineering Saves Lives

The contrast between bridge failures and bridge survival in earthquakes tells a clear story: engineering works. Modern seismic design principles—ductility, isolation, energy dissipation, redundancy—protect bridges that would have collapsed under older design approaches.

Key takeaways:

• Bridge age matters enormously. Pre-1970 bridges are vulnerable. Post-1995 bridges have excellent seismic design.
• Seismic isolation is transformative. Reduces forces by 70-80%, protecting structure and contents.
• Ductility prevents catastrophic collapse. Even if damaged, ductile bridges don't suddenly fall.
• Retrofit programs work. Billions invested in California retrofits have dramatically improved safety.
• Different bridge types have different vulnerabilities. Suspension bridges generally excellent. Beam bridges most vulnerable without proper design.
• Ground failure can doom even well-designed bridges. Foundation design as critical as superstructure.

Bridges are lifelines. When earthquakes strike, functioning bridges enable emergency response, maintain economic activity, and connect communities. Failed bridges isolate communities, strand emergency services, and multiply disaster impacts.

The next major earthquake will test our bridges. Those built to modern standards will survive. Those that haven't been retrofitted remain vulnerable. Understanding how bridge seismic engineering works helps appreciate the invisible protection built into infrastructure we use daily—and highlights the continued need for retrofit investment in aging bridge stock.

For more earthquake engineering insights, explore our guides on building seismic design, base isolation technology, and why older structures fail. Monitor seismic activity on our real-time earthquake map.