Engineering (2017)
A Piece is a Piece, No Matter How Small: How the smallest parts led to the 1989 collapse of the Oakland Bay Bridge, Matthew Moore
The Oakland Bay Bridge’s rebar reinforcements in section E9 buckled under the stress of lateral seismic activity, and completely buckled on one of the bridge’s piers
Inspectors and welders had been rushed through their work prior to the disaster, resulting in an unstable structure
The bridge was built with “hot rivets” rather than earthquake resistant box beams or seismic dampening beams
In the early evening hours of the seventeenth of October 1989, a 7.1 magnitude earthquake known as the Loma Prieta Earthquake rocked the San Francisco, California Bay Area. This intense seismic event was responsible for sixty-three deaths and 3,757 injuries. The double decker Oakland Bay Bridge, on the northern side of the San Francisco Bay, suffered a partial collapse and structural failure. A fifty-foot top section of the eastern truss collapsed down onto the bottom deck over bridge pier E9.1 The collapse was responsible for one death and over a month of closure. While the Loma Prieta was rated as a severe and violent earthquake, the building and structures of San Francisco were designed to hold up to it. Most structures performed well, including Candlestick Park, where the 1989 World Series was being played at the time, but infrastructures, such as the Bay Bridge did not. This was due to the failure of not large, but very small parts. Although San Francisco was hit with intense seismic activity during the Loma Prieta Earthquake of 1989, the Oakland Bay Bridge suffered a partial collapse solely for faulty engineering reasons, including the lack of earthquake resistant riveting and rebar-supported concrete pillars.
Construction of the Oakland Bay Bridge began in 1933 and opened November twelfth, 1936, six months before the Golden Gate Bridge.2 It was designed as a double level bridge that incorporated both passenger car and train decks across its span of almost 10,000 feet. The middle point on the bridge rests on Yerba Buena Island, making the bridge a two span suspension type. At the time of construction, the bridge was one of the most advanced of its type. With 120 million pounds of reinforced steel and 450,000 cubic feet of concrete, the sheer size of the bridge and its construction made for a complicated engineering process and the chance for things to go wrong at many stages.3
To complicate things further, over its nearly fifty-six year history leading up to 1989 and the earthquake, no major retrofits or renovations had been done on the bridge. Granted, the bridge was indeed holding up well and proved to be sturdy, but engineers knew that the earthquake resistant properties of the bridge were degrading as it aged.4 Engineers from Caltrans, the Bay Area’s transportation division, reported cracks and chips in the concrete support pillars built on the bridge piers. The collapsed section of the bridge sat directly over pier E9, pointing to stress from the concrete pillars. Those concrete pillars were reinforced with rebar, a type of metal pole, but were incorrectly positioned inside the concrete to handle lateral seismic activity. The rebar concrete was superior at supporting vertical stresses, such as from the weight of the double decks on the bridge above. However, when the intense and sudden lateral movements of the earthquake began, the pillar’s strength was greatly reduced. Take a stack of Jenga blocks, for instance, and imagine stacking a pile of books directly on top of it. The wood blocks of the tower would hold up impeccably under the stress of the weight directly above. But if the books or tower were to be jostled from side to side, the tower would either collapse or become extremely unstable in no time at all. This same sort of thing happened on the Bay Bridge the evening of October seventeenth, which weakened the support structure to the point of bringing down the top deck into the lower deck. While the improper rebar concrete was a major factor to blame, it is only one piece to the puzzle that created this disaster.
In the fashion of cause-and-effect, the faulty rebar concrete led way to another important piece of the bridge to fail. Up on the decks were the suspension beam structures that gave the bridge its iconic look and created a skeleton like tunnel that encased the traffic decks. Supporting and holding together these beams were millions of rivets. These eliminated the need for welding every piece together as smaller pieces could be riveted instead. While rivets are strong, they do not match the integrity of welds, so more rivets had to be used over each beam. The resulting structure resembled a lattice of metal strips connecting together beams. Again, this type of load bearing product was excellent at supporting vertical loads, but any lateral motion would risk compromising its integrity. A good example of how the lattices were methodized is to think of a scissor lift. The lift does well going up and down, but suddenly bumping one from the side could tip it over. The crisscrossing supports up the middle of the bridge’s trestles acted in this way but remained stationary, as they were riveted into place. The tactic of welding in the 1930’s was to perform hot rivets. The fast and efficient way to install the millions of rivets needed used heat to make the rivets extra malleable before being punched into the metal. While this was a quick method, it left one huge safety concern. Unlike the hardened steel being used as tresses and trestles, the rivets could not be hardened, for they needed to be soft during the hot riveting process. This led to weaker pieces of the bridge trying to support hardened steel beams. While it was an acceptable engineering decision for daily operations, the earthquake resistivity of a hot riveted bridge degraded sharply over time.
On that fateful day in October, the combination of old rivets and concrete pillars that were not earthquake resistant, led to the collapse of the Bay Bridge section at pier E9. The inspectors of the bridge had made notes prior to the disaster outlining the potential structural faults they found. However, multiple accounts depict inspections being rushed or abbreviated as the bridge underwent routine maintenance.5 The cracks and chips that had been found in the concrete were declared normal due to settling of the bridge’s pylons on the ocean floor of the Bay. On the contrary, those cracks should have never passed the inspections, as the chips that accompanied them pointed to a weakness in lateral movement. In addition, on the pillar sections closer down to the water of the Bay, the cracks exposed access for salt water into the metal rebar reinforcements. The salt progressed rust of the supports, but this was never cited as a concern to the structure.6 When the Loma Preita earthquake hit, the concrete pillar’s weight bearing capabilities were sharply reduced, putting immense stress on the metal rivets. This abrupt shift of weight initiated what became the collapse at section E9.
While this collapse and engineering disaster on the Bay Bridge was abhorrent, the fact that only section E9 collapsed is worth recognition. The rest of the bridge remained almost unfazed and structurally sound during the earthquake. Citing the fifty six year old structure’s age, the hardened steel and settled concrete supports could have easily turned into a pile of rubble. When bridges are first built, they have not settled yet, leaving cushion room for earth vibrations to dissipate in the concrete and steel structure of the bridge.7 More than half a century of settling had occurred to the Oakland Bay Bridge prior to the 7.1 magnitude earthquake, making hardly any natural dampening existent. Just off the end of the bridge, on land, was a double decker section of freeway known as the Cyprus Street Viaduct that completely collapsed under its own weight, killing sixty four people.8 The Cyprus Street Viaduct suffered the same kind of rebar concrete weakening as the Bay Bridge. However, this weakening caused the entire Viaduct to violently collapse when the rebar and concrete sheered. If the Bay Bridge had encountered such fate, there would have been countless more loss of life. The resilience of the bridge in light of the collapse at section E9 was certainly shown on that October evening.
Engineers quickly realized that upgrades to the Bay Bridge were long overdue and began drafting solutions during the month closure following the earthquake. Initially, sealing cracks and replacing rivets was the course of action. Engineers and Caltrans workers spend countless hours replacing the millions of rivets with devices known as bolted box beams. These are steel rods that are installed into a steel beam that is installed in place of the riveted sections. The steel beam acts as a buffer with the rods supporting it. Since the rods are positioned laterally, the bridge maintained its vertical integrity with the added benefit of now being laterally stable as well. The rods allowed the bridge to sway slightly from side to side if needed, thereby reducing any lateral stress that used to be placed on the rivets had the bridge been moved from side to side. A further improvement that was also made on the Bay Bridge, especially at section E9, was the installation of seismic dampeners. Essentially giant shock absorbers, these dampeners would act as a primary defense against any unwanted shaking or movement. The combination of the bolted box beams and seismic dampeners also reduced the effects of uneven weight loads the bridge could experience. One of the problems any bridge can face is inadequate seat length, especially during a seismic event.9 Seat length is measures where the sections of the bridge decks sit on the support pillars or structures. The bolted box beams helped the Bay Bridge’s decks all have similar seat lengths, reducing a heightened level of stress in any one area. The seismic dampeners served to maintain this similar seat length, should a seismic event occur and movement were to threaten the seat lengths.
The Oakland Bay Bridge has had an impressive eighty-three year history. Over 240,000 vehicles traverse its length daily. It has supported cars, trucks, and even trains at one point or another. The sheer volume and impact this one bridge has on a city indicates its importance. Engineering disasters occur, but some of the problems revealed in the 1989 earthquake collapse could have easily been prevented. For how significant the Bay Bridge, trivial problems and rushed inspections should have never been a problem. Human error and poor decisions caused those errors to arise, costing lives and billions of dollars.10 Unfortunately, these problems still exist today. The entire east side of the bridge, which contained section E9, has been entirely replaced in recent years due to safety concerns. However, the new section of the bridge is already appearing to suffer from similar problems as its older counterpart. Concrete cracks have already been noted and support rods, similar to the box beams of the old bridge, are in danger of sheering. Current problems are indicative of past issues, potentially leading to a dangerous and expensive cycle of fixing and re-fixing issues. If anything can be learned from the damage sustained in the 1989 Loma Prieta Earthquake, it should be to install the safest and best hardware from the beginning.
Works Cited
California Department of Transportation. "Bay Bridge Info." BAY BRIDGE HISTORY
TIMELINE. State of California, n.d. Web. 24 Feb. 2016
United States. US Geological Survey. Department of the Interior. Scenarios for Historic San Francisco Bay Region
Earthquakes. By W. H. Bakun. Menlo Park, CA: U.S. Dept. of the Interior, U.S. Geological Survey, 1998.
United States. Department of Transportation. Federal Highway Administration. Framework for Improving
Resilience of Bridge Design. By Brandon W. Chavel and John M. Yadlosky. Washington, D.C.: U.S. Dept.
of Transportation, Federal Highway Administration, 2011.
Orozco, Greg L., and Scott A. Ashford. Effects of Large Velocity Pulses on Reinforced Concrete Bridge Columns.
Report no. EEC-9701568.
Tucker, Catherine, and Luis Ibarra. "Effects of Partial-Design-Strength Concrete on the Seismic Performance of
Concrete-Filled Tube Columns in Accelerated Bridge Construction." J. Bridge Eng. Journal of Bridge Engineering, 2016, 40. Accessed February 23, 2016.
WEINTRAUB, DANIEL M. "BAY AREA QUAKE : Inspectors Had Rated Nimitz Sound : Safety: Regular Checks
Had Turned up Cracks and Chips in the Concrete Pillars, but a Caltrans Aide Says They Did Not Contribute to the Freeway's Collapse." Los Angeles Times. October 20, 1989. Accessed February 24, 2016. http://articles.latimes.com/1989-10-20/news/mn-142_1_freeway-collapse.
Bibliography
Anthony, Laura. "Cost of Fixing Faulty Bay Bridge Bolts Undetermined." ABC7 San Francisco.
ABC7 San Francisco, 27 Mar. 2013. Web. 24 Feb. 2016.
California Department of Transportation. "SAN FRANCISCO-OAKLAND BAY
BRIDGE." SAN FRANCISCO-OAKLAND BAY BRIDGE. State of California, n.d. Web. 24 Feb. 2016.
The Engineer. "Cypress Street Viaducts ENGINEERING.com." Cypress Street Viaducts
ENGINEERING.com. N.p., 13 Oct. 2006. Web. 23 Feb. 2016.
Tweney, Dylan. "Gallery: How to Build an Earthquake-Resistant Bridge." Wired.com. Conde
Nast Digital, 29 July 2010. Web. 23 Feb. 2016.