de la Concorde Overpass Collapse (September 30, 2006)
Laval, Quebec, Canada
Gea Johnson, BAE/MAE, The Pennsylvania State University 2012

Key Words

Box girders, abutment, seat, corbel, prestressing, pre-tensioning, post-tensioning, elastomeric bearing pad, redundancy, fatigue


Introduction

At 12:30 PM on September 30, 2006, three eastbound lanes and a pedestrian walk-way of the de la Concorde overpass in Laval, Quebec, Canada collapsed in the midst of vehicles traveling on the overpass and beneath the overpass on Highway 19 (Gazette, 2006), as can be seen in Figure 1. As a result, three vehicles and a motorcycle fell about 15 meters (49 feet) onto Highway 19 and the overpass collapsed on top of two vehicles that were driving on Highway 19. Six people were severely injured and five people were killed due to the concrete road crushing the two vehicles driving beneath the overpass (Gazette, 2006).
Aerial view of the de la Concorde overpass partial collapse onto Highway 19 GJJ.jpg
Figure 1. Aerial view of the de la Concorde overpass partial collapse onto Highway 19
(Photo credit: Original Commissioner’s investigative report and the Ohio
Department of Transportation)

The overpass, described as "young and modern," was built in 1970 and was to have a 70 year life span (Gazette, 2006). Months before the collapse, local residents who lived near the de la Concorde overpass and Highway 19 had noticed debris falling off of the overpass and an hour before the collapse large chunks of debris measuring 15 cm x 36 cm (5.9 in x 14.2 in) were found near the overpass (Gazette, 2006). A Transport Quebec crew was dispatched to remove the debris, but they didn't deem it necessary to shut down the overpass. Less than an hour before the collapse, a Transport Department patroller inspected the overpass by using a visual and auditory inspection and concluded that there was no indication the overpass was unstable (Gazette, 2006). Soon after the collapse, former premier Pierre Marc Johnson was given the task to determine what caused the de la Concorde Blvd. collapse and it was later concluded that improper rebar placement was the fatal flaw that led to the Concorde overpass collapse (Versace, 2007).















De la Concorde Overpass, As Designed and As Built

The de la Concorde overpass enabled the unobstructed crossing of two very busy traffic arteries in Laval, Quebec: de la Concorde Boulevard, which passed on the overpass' surface, and Highway 19, that passed below the overpass. The structure carried three westbound and three eastbound lanes. A plan and an elevation of the overpass can be seen in Figures 2 and 3. As can be seen in the plan view in Figure 2, the structure had a skew, which influenced the stress distribution and concentrated more force towards the southeast and northwest corners of the abutments. In addition, a 3-D perspective of the de la Concorde overpass can be seen in Figure 4.

A plan of the de la Concorde Overpass GJJ.jpg
Figure 2. A plan of the de la Concorde overpass
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

An elevation of the de la Concorde Overpass GJJ.jpg
Figure 3. An elevation of the de la Concorde overpass
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

3-D perspective of the de la overpass and Highway 19 located below the overpass GJJ.jpg
Figure 4. 3-D perspective of de la Concorde Blvd. and Highway 19 located below the overpass
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

In addition, there were several differences described on the structure's as-designed drawings and the as-built drawings (Johnson, 2007, 25). The main differences between these two set of drawings were the following:
  • "The bearing pad details at the ends of the box girders," (Johnson, 2007, 127)
  • "The 2 1/2" thick shoulder on each side of the expansion joint" (Johnson, 2007, 127)
  • "The 2 1/2" by 3/8" bent steel plates welded to the steel angles on both sides of the expansion joint," (Johnson, 2007, 127)
  • "The U-shaped hanger details," (Johnson, 2007, 127)
  • "The bar marks identifying the bending types," (Johnson, 2007, 127) and
  • "The details of the upper and lower reinforcement layers in the deck slab," (Johnson, 2007, 127).

Abutments and Reinforcement, As Designed

The overpass had two abutments, which were supporting structures located at the ends of the overpass. "Each abutment was composed of an inclined wall at the front and four longitudinal retaining walls. These retaining walls supported a thick concrete slab, part of which extended over Highway 19 forming a cantilever, a structure with only one supported end," (Johnson, 2007, 29).

The three stages in constructing the overpass structure were as follows:

Stage 1: Bedrock supported the footings. The front wall, which was inclined towards the freeway, was supported by four triangular retaining walls anchored in the rear to the bedrock by tie backs (Johnson, 2007, 29).

Stage 2: The shorings and formwork were prepared for construction of the thick slabs after the caissons, which were formed by the walls, were backfilled. The reinforcing bars for each slab was laid and the concrete slab was cast (Johnson, 2007, 29).

Stage 3: A cantilever over the freeway was formed following the completion of the thick slabs (Johnson, 2007, 29).

The eastern cantilever abutment, as can be seen in Figure 5, was divided into two sections: "a chair bearing support located at the free end, which supported the central span box girders, and a thick slab, which attached the chair bearing support to the rest of the abutment. The equilibrium of the abutment was achieved by the tie backs into the bedrock at the opposite end of the cantilever. The chair bearing support (or corbel) was an extension of the cantilever on which the box girders forming the central span of the structure were supported," (Johnson, 2007, 31).

The east abutment principal dimensions were as follows:
  • The east and west cantilevers measured 13 ft. from the wall of the abutment to the center of the seat (Johnson, 2007, 31).
  • The total abutment width, including the overhanging sidewalks on each side, was 90 ft. or 82 ft. without the sidewalks (Johnson, 2007, 31).
  • The slab thickness varied. The slab thickness, including the thickness of the asphalt pavement, was roughly 4'-5 3/8" at the origin of the cantilever at the face of the wall and decreased to 4 ft. towards the chair bearing support (Johnson, 2007, 31).


Eastern abutment principal dimensions GJJ.jpg
Figure 5. Diaphragm and principal dimensions of the eastern abutment
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

The as-designed reinforcing bar layout in the de la Concorde overpass east abutment cantilever excerpted from the original drawings can be seen in Figure 6.

Original beam detail without transverse reinforcement GJJ.jpg
Figure 6. Original beam detail
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

In addition, an as-designed illustration of the reinforcing layout can be seen in Figure 7, where each bar is shown in its intended location by replacing the graphic conventions in Figure 6 with the components' colors and textures.

A description of the reinforcing bars found in the cantilever sections were as follows:
  • No.14 main bars (in red) were placed longitudinally in the top part of the abutment and were spaced 6" apart (Johnson, 2007, 33).
  • No. 7 transverse bars (in blue) were placed under the No.14 bars and were spaced 12" apart (Johnson, 2007, 33).
  • No.8 longitudinal bars (in purple) were placed in the bottom part of the abutment and were spaced 10" apart (Johnson, 2007, 33).
  • No.7 transverse bars (in pale blue) were placed above the No.8 bars and were spaced 12" apart (Johnson, 2007, 33).

The reinforcing bars in the cantilever's chair bearing support region included:
  • The No.8 bars (in green), U-shaped hanger bars or ties, were spaced 10" apart. To engage the No.14 bars, these structural hangers lifted the load applied on the bearing seat support towards the top of the cantilever (Johnson, 2007, 33).
  • The No.6 diagonal reinforcing bars (in yellow), "prevented cracks from opening in the corner of the chair bearing support. These bars also served to transmit the loads from the chair bearing seat to the top of the cantilever and to intercept the tensile stresses in the concrete near the support," (Johnson, 2007, 33).
  • The No.10 pin-shaped bars (in pink) formed the chair bearing support's (or corbel) principal reinforcement and were spaced 5" apart. "They transmitted the loads exerted on the corbel towards the interior of the thick cantilever slab," (Johnson, 2007, 33).

Reinforcing bars layout and perspective in the East Abutment cantilever as specified on the As Designed Drawings GJJ.jpg
Figure 7. Reinforcing bars layout and perspective in the east abutment cantilever as specified on the as-designed drawings
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

Lastly, in the de la Concorde overpass, the cantilever's thick slab did not contain stirrups or any kind of shear reinforcement in the regular zone, as can be seen in Figures 5 through 7. As a result, the concrete alone resisted the shear stress (Johnson, 2007, 35). During the time of the design and construction of the overpass, there were no minimum shear reinforcement requirements for slabs in the code (Johnson, 20007, 119).

Abutments and reinforcement, As Built

The Commission investigation, which will be discussed later in this report, revealed that certain reinforcing bars were not installed in accordance with the plans, as can be seen in Figure 8. Figure 8 shows that the U-shaped No.8 hanger bars and the diagonal reinforcing bars were placed under the No.14 bars instead of at the top in the same plane as the No.14 bars, as can be seen in the as-designed drawings. In addition, the contractor added additional bars, including No.6 vertical bars, installed every 4' by 4' and a few horizontal bars to support the bars of the upper layers. The plans and specifications did not have a provision for the installation of the No.6 bars (Johnson, 2007, 35).

Rebar placement as designed (left) and as constructed (right) GJJ.jpg
Figure 8. Chair bearing support rebar placement as-designed (left) and as-constructed (right)
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

Box Girders and Central Span

Two series of ten prefabricated, prestressed concrete box girders, which were supported on the chair bearing supports of the east and west abutments, formed the central span of the overpass that carried the vehicular traffic in the east-west direction (Johnson, 2007, 39), as can be seen in Figure 4. These girders were hollow which reduced the weight of the central span. The geometry and section of the box girders can be seen in Figure 9. Each girder measured 3'-6" high, 4' wide, and 90' from center-to-center of the seats. Lastly, there was a 1'-5 1/4" spacing at the center of the overpass between the two series of ten girders (Johnson, 2007, 41).

Transverse section of the de la Concorde overpass GJJ.jpg
Figure 9. Transverse section of the de la Concorde overpass
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

In addition, the deck weight and traffic loads that were applied on the bridge created tensile stresses in certain parts of the box girders. As a result, the concrete box girders were prestressed to counter the tensile stresses. To prestress the girders, "high axial compression forces were applied to the concrete by means of steel cables (strands), tensioned to a force of 28,900 lbs, incorporated into the concrete," (Johnson, 2007, 41). The steel cables were pre-tensioned during the manufacturing of the box girders to apply this longitudinal prestressing (Johnson, 2007, 41).

In constructing the central part of the deck, "the two series of ten girders were held together in the transverse direction by high strength steel strands inserted in sleeves provided in the three diaphragms of the box girders. These strands were inserted and tensioned after the girders were in place on the abutments," (Johnson, 2007, 41). As a result, the transverse prestressing is referred to as post-tensioning. A shear key between the box girder elements were formed when the longitudinal grooves on the box girders were filled with cement grout. Finally, a 3 1/2" concrete slab was installed over the box girders and a waterproofing membrane was intended to protect the slab and be covered with an asphalt pavement layer with a 2 1/2" nominal thickness (Johnson, 2007, 41).

Lastly, "each girder rested on an elastomeric bearing pad sitting on the chair bearing supports. These 6"x14"x1" neoprene pads were reinforced by a 3/32" thick steel plate, cast into the pad," (Johnson, 2007, 43). To enable water draining, the surface of the seat was slightly inclined. The center of the west-side bearing pads, which locked the position of the box girder in place on that side, was inserted with 3/4" diameter pins. Variations in temperature induced expansion movements of the box girders, therefore these movements were transferred to the east abutment expansion joint (Johnson, 2007, 43).

Waterproofing Membrane

The specifications for the de la Concorde overpass required a Type 1 membrane composed of an asphalt mastic layer, which represented the best practice at that time, be applied to the deck to waterproof the slab. The 1992 rehabilitation, which will be discussed later in this report, revealed that the membrane was never installed because there was no trace of it on the existing structure. During the 1992 repairs, the specifications called for the installation of a Type 3 (prefabricated adhesive) membrane. During the collapse investigation, the demolition work and expert studies on various concrete blocks revealed a Type 3 membrane was never installed, but that only a liquid membrane or liquid asphalt barrier was used in place of the Type 3 membrane (Johnson, 2007, 43). The liquid membrane applied was inadequate in waterproofing the thick slab.

Special Characteristics of the de la Concorde Overpass

At the time of its design in North America, the de la Concorde overpass was considered innovative. Prestressed box girders enabled the overpass to cross Highway 19 with a single span without any intermediate support. The box girders were placed side by side and resulted in a uniform surface beneath the central span, "which rested on a beam seat continuous over the entire width of the bridge," (Johnson, 2007, 2). The excavation depth required for the open-cut construction of the freeway was minimized due to the superstructure's slenderness (Johnson, 2007, 2).

"The box girders forming the central part of the deck rested on beam seats located at the ends of the cantilevers, directly under the expansion joints, as can be seen in Figure 10. The end of the cantilever was a complex load transfer area. The expansion joints were highly exposed parts which lost their ability to seal off water when damaged, contributing to the accumulation of water, road salts and debris on the chair bearing support. This vulnerability was even greater because the seats couldn't be inspected and maintained without lifting the deck," (Johnson, 2007, 2). In order to inspect the continuous chair bearing supports, the vehicular traffic on both de la Concorde Blvd. and Highway 19 would have had to be interrupted. The expansion joints and the ends of the cantilevers on this structure were therefore critical zones that required special attention during inspections and maintenance work (Johnson, 2007, 2).

Un-inspectable bearing seat detail GJJ.jpg
Figure 10. Un-inspectable bearing seat detail
(Photo credit: Original Commissioner’s investigative report and the OhioDepartment of Transportation)

Lastly, a structure exposed to freeze-thaw cycles and de-icing requires the installation of a waterproofing membrane to prevent infiltration of salt-laden water, which can deteriorate the concrete. Even though this practice was not common during the time the overpass was constructed, waterproofing membrane installation became current in 1992 when the de la Concorde overpass underwent major repairs (Johnson, 2007, 3).

In summary, the de la Concorde overpass was a very unique and vulnerable structure. Because continuous beam seats along spans are impossible to inspect, they have not been built for the last 30 years and are not allowed under current codes (Johnson, 2007, 4).

Events Leading Up to the Collapse

Months before the collapse, local citizens that lived near the overpass and Highway 19 noticed pieces of debris falling off the overpass and large chunks of concrete that appeared to had fallen from the structure. Between 10:30 AM and 11:00 AM on the day of the collapse, a local citizen driving south in the right lane of Highway 19 had hit a concrete chunk with his car that was estimated to be 18 feet north of the overpass. The piece of concrete was estimated to be 5 inches by 3.5 inches. Further, around 11:20 AM on the day of the collapse, another local citizen saw a concrete chunk measuring about 3 feet long by 1.5 feet wide separate from the overpass deck on the southeast side of the overpass before crashing down on the shoulder (Johnson, 2007, 46).

At 11:26 AM on the day of the collapse, a Transport Department patroller was called and informed that pieces of concrete were found near the de la Concorde overpass. After arriving to the scene at 11:45 AM, the patroller removed a piece of concrete measuring 18 inches long by 7 inches wide and 3 inches thick, as can be seen in Figure 11, as well as removed twenty concrete fragments roughly the same size as golf balls (Johnson, 2007, 49). While on the scene, the patroller inspected the overpass by performing both visual and auditory inspections (Gazette, 2006), as can be seen in Figure 12. During the 15 minute inspection, he did not notice any particular jumping on the overpass, slowing down of traffic, noise or vibration, and no other concrete chunk fell (Johnson, 2007, 50). As a result, the Transport patroller concluded that there was no indication the overpass was unstable and deemed it unnecessary to shut down the overpass (Gazette, 2006), but he did fill out an anomaly report requesting the overpass be inspected as soon as possible stating it was a definite emergency (Johnson, 2007, 51).
Large piece of concrete that fell off of the overpass on the morning before the collapse occurred GJJ.jpg
Photo taken by a transport department patroller during his inspection an hour before the collapse occurred GJJ.jpg
Figure 11. Large Piece of Concrete that fell off of the overpass on the morning before the collapse occurred
(Photo credit: Original Commissioner’s investigative report and the Ohio
Department of Transportation)
Figure 12. Photo of a hole and crack in the east abutment taken by a transport department patroller during his inspection an hour before the collapse occurred
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)


Collapse

At approximately 12:30 PM on September 30, 2006, the de la Concorde overpass partially collapsed, with three eastbound lanes crashing onto Highway 19 below. As a result, three vehicles and a motorcycle fell about 15 meters (49 feet) onto Highway 19 and the overpass collapsed on top of two vehicles that were driving on Highway 19 (Gazette, 2006), as can be seen in Figure 13. Soon after the collapse, emergency crews appeared on the scene to remove sections of the collapsed overpass in an effort to rescue survivors and to retrieve those who were killed in the collapse (CBC News, 2006), as can be seen in Figure 14. Six people were severely injured and five people were killed due to the concrete road crushing the two vehicles driving beneath the overpass (Gazette, 2006).

The collapsed portion of the de la Concorde Overpass’ east abutment GJJ.jpg
Firefighters on the scene in the process of saving those who were injured and retrieving those who were killed in the collapse GJJ.jpg
Figure 13. The collapsed portion of the de la Concorde Overpass' east abutment
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)
Figure 14. Firefighters on the scene in the process of saving those injured and retrieving those who were killed in the collapse
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)

As soon as the overpass' central span lost its support on the abutment, which had fractured, the deck of the de la Concorde overpass fell to the ground in a fraction of a second. "The expert reports reveal that the deck hit the ground on the east side first, leaving a mark in the roadway of the freeway," (Johnson, 2007, 53). The fracture stopped at the center of the east abutment and the bearing support was twisted at this location, which allowed the south deck to collapse. Therefore, the northern half of the deck did not fall after the southern half collapsed. The south deck fractured at midspan after falling in a single block and it hit the concrete guardrail separating the north and south lanes of the freeway (Johnson, 2007, 53).

Investigation

Soon after the collapse, former premier Pierre Marc Johnson was handed the task to lead the Commission and to determine what caused the de la Concorde Blvd. overpass collapse (Gazette, 2006). On October 3, 2006, the Quebec Government established the Commission to investigate the circumstances and causes of the collapse and to provide recommendations to prevent the reoccurrence of this type of collapse. In the investigation, the Commission focused on protecting and preserving evidence necessary for the investigation; using scientific investigations to determine the causes of the collapse; and compiling and analyzing all available documentation to reconstruct the life of the overpass, ranging from its design through the time of the collapse. During the overpass dismantlement, core samples and concrete pieces were collected for testing, the box girders were examined, the fracture plane was carefully gathered, a site survey was completed and radar measurements were taken. In addition, the Commission identified and met with individuals and organizations that were involved in the design, construction, and maintenance of the structure and the witnesses of the collapse. During the public hearing, the Commission heard 58 witnesses and expert testimony. The Commission also consulted individuals and organizations who were able to provide information on various aspects of the bridge management systems (Johnson, 2007, 1).

Design, Construction, and Maintenance of the de la Concorde Overpass

The ministère des transports du Québec (MTQ) was the owner of the de la Concorde overpass and they hired Desjardins Sauriol & Associés (DSA), a consulting engineering firm. DSA designed the concepts and prepared the drawings, specifications, and documents for the bridge overpass. DSA was also responsible for material quality control and for supervising the construction work (Johnson, 2007, 61).

In addition, Inter State Paving Inc. (ISP), a contractor, was hired to construct the overpass in accordance to the drawings and specifications. Despite having built an overpass above Highway 19 in Montreal, ISP had little bridge construction experience. For the de la Concorde overpass, ISP retained the services of several subcontractors for the prestressed box girders, formwork, steel rebar and rebar placement, and concrete (Johnson, 61, 62).

During the design phase of the structure, the elevation of the rock and curving of the road presented design difficulties regarding the overpass. A solution suggested was to use prefabricated beams for the de la Concorde overpass. One of the lead design engineers explained to the Commission "the design chosen for the de la Concorde overpass was a concrete bridge with a drop-span supported by a beam seat at the end of a cantilever," (Johnson, 2007, 63). The many advantages to this design included:
  • A thin superstructure with a better exterior finish and a flat underside without the use of separate beams (Johnson, 2007, 63).
  • "Elimination of a central pier," (Johnson, 2007, 63).
  • "Reduced expropriation costs since the right-of-way would be narrower," (Johnson, 2007, 63).
  • Simplified construction of the overpass (Johnson, 2007, 63).

Further, with excavation being expensive, a design that did not require deep excavation was preferred. Anchors could easily be added to stabilize the abutments and improved driver visibility was achieved with the elimination of the central pier.

Moreover, the lead engineers indicated in their testimony to the Commission that numerous unanticipated problems occurred after the structure was put in service. These problems included:
  • Drainage - an accumulation of water formed between the sidewalk and the pavement due to the absence of drains on the overpass and the small slope of the pavement (Johnson, 2007, 65).
  • Expansion joint - the preliminary studies called for the installation of a watertight expansion joint because these joint types must be watertight. In the de la Concorde overpass, the joints were located directly above the beam seats, which led to both water and salt penetration through the joints onto the beam seats. In the design, it was assumed that the water would run off the beam seat and that the accumulated salt would be removed by a pressure wash (Johnson, 2007, 65).
  • Inspection - The beam seat inspection was very difficult due to limited access. The structure section located under the expansion joint was inaccessible and it was impossible to inspect the interior of the box girders. During the design, it was assumed that the inaccessible beam seat wouldn't be an issue since this has not been a proven problem for structures designed similarly in France (Johnson, 2007, 65).
  • Maintenance - Since the concrete, steel, and neoprene pads were located under the expansion joint and with the inaccessibility of the beam seat, maintenance work would require bridge lanes to be closed and the freeway to be closed for major repairs (Johnson, 2007, 65).

Lastly, MTQ was responsible for maintaining and repairing the de la Concorde overpass based on inspection reports. The inspections were to be used for keeping track of the evolution of the structure over time and for identifying the problems and defects in the structure. Each inspection was to be followed up by a report containing the observations, photographs taken during the inspection, measurements, and note defects and weaknesses (Johnson, 2007, 4).

Findings of the Commission

During the investigation, evidence showed that the specifications were confusing and did not meet the 1966 or current CSA standards and therefore resulted in the use of low quality concrete (Russell, 2008). The concrete did not have the required properties to resist deterioration caused by freeze-thaw cycles (Johnson, 2007, 39, 130). In addition, the construction and management of the overpass consisted of faulty installations, unfulfilled obligations, and negligent inspections and interventions performed by the overpass' management (Johnson, 2007, 5).

Further, the most obvious weakness of the construction of the overpass was the general lack of accountability for the quality control of the work and materials. DSA failed to comply with their work supervision responsibilities and ISP and its contractors failed to fulfill their legal and contractual obligations regarding quality of work and compliance with drawings and specifications by passing their responsibilities on to workers and to the consulting engineers (Johnson, 2007, 5).

In addition, based on the inspection report and testimonies heard by the Commission it was clear that over the last 30 years the Ministry's staff was aware of the special characteristics of the de la Concorde overpass, a structure that was built according to an unusual concept that had serious inspection problems. During the time in which MTQ was responsible for the overpass, the bridge was never subjected to an inspection and maintenance program that took into account its special characteristics, particularly, the critical beam seats at the ends of the cantilevers. This was demonstrated through the fact the scheduled maintenance activities were delayed (Johnson, 2007, 5). There were several inadequacies regarding MTQ's inspections and maintenance of the de la Concorde overpass, which included:
  • The inspection records were missing the "real as-built drawings, the maintenance file and the specified and in-place material properties," (Johnson, 2007, 136).
  • Some of the MTQ inspection manuals' requirements were not entirely met regarding:
    • "The values assigned to some of the ratings," (Johnson, 2007, 136)
    • "The detailed content of the inspection reports," (Johnson, 2007, 136) and
    • "The frequency intervals prescribed by the manual for maintenance activities," (Johnson, 2007, 136).
  • Absence of a complete file accessible to the MTQ inspectors was a key factor that contributed to the lack of follow-up on the progressive deterioration of the overpass," (Johnson, 2007, 136).
  • The documents relating to the maintenance of the structure, particularly the 1992 repairs, and a formal report on the special inspection performed in 2004 were missing (Johnson, 2007, 138).
  • The files kept did not include warnings pertaining to the particular character of the structure or the need to in-depth inspections (Johnson, 2007, 139).
  • Inspection reports filed by MTQ's personnel had significant deficiencies and were not compliant with MTQ's structure inspection manuals (Johnson, 2007, 140).
  • The inspection reports weren't maintained or precise (Johnson, 2007, 140).
  • The Inspection report deficiencies and "inadequate record keeping made it difficult to accurately monitor the evolution of the structure's condition," (Johnson, 2007, 140).

The two missed opportunities to conduct a detailed evaluation of the overpass were in 1992 and 2004. The initial scope of the 1992 repairs was to replace the expansion joints, but the final scope of the repairs far exceeded the initial scope. In the 1992 rehabilitation the asphalt wearing surface was replaced, the expansion joints were replaced, and there were significant concrete repairs (Russell, 2008). Despite the evidence of serious concrete deterioration and improper rebar placement, the repairs were performed without re-evaluating the structure's condition and without installing the specified protective waterproofing membrane (Johnson, 2007). The record plan documents from the 1992 repairs were not maintained (Russell, 2008). The damage induced and the oversights in the 1992 repair work included the following:
  • Large quantities of concrete were removed from the structure that resulted in the exposure of several reinforcing bars near the seat of the cantilever (Johnson, 2007, 135, 173).
  • MTQ engineers failed to notice reinforcing bar placement was not in accordance to the drawings and that the U-shaped hangers were not connected to the No.14 bars (Johnson, 2007, 135, 173).
  • Failure to lead to an in-depth detailed investigation of the structure and characterization of the concrete to determine the cause of its deterioration (Johnson, 2007, 135, 173).

Lastly, the 2004 special inspection was intended to reassure the regional engineer who expressed concerns of a specific issue pertaining to the overpass by getting a second opinion from an engineer in the structure's office that had "additional overall knowledge of Quebec's entire roadway structure inventory," (Johnson, 2007, 105). In his request for assistance from the Structure's Office (In Quebec), a Regional Office (in Montreal) engineer noted in his inspection the major deterioration of the bearing seats and the presence of wide shear cracks on the abutment cantilevers (Russell, 2008). The Structure's Office relied on original record plans, which lacked the reinforcing steel schedule and did not include documents on the 1992 repairs, which included reports of misplaced reinforcement (Russell, 2008). The Structure's Office engineers reviewed the beam seat (corbel) capacity, but did not review the slab capacity and ignored the cracks and efflorescence on the fascias (Russell, 2008), as can be seen in Figure 16. They focused on the beam seats and found that the seats' design capacity was adequate and that the deterioration appeared localized at the fascias; the Structures Office engineers failed to address the cracking at the fascias (Russell, 2008). Lastly, they concluded a detailed inspection of the bearing seats was not immediately required, but recommended careful monitoring of the beams (Russell, 2008). The 2004 special inspection should have led to "a condition evaluation of the structure including an evaluation of the load carrying capacity and an evaluation of the condition of the materials." The problems of the overpass could have been detected and appropriate remedial actions could have been taken if a condition evaluation was conducted (Johnson, 2007, 144).

Missed detailed evaluation opportunity in 1992 GJJ.jpg
Missed detailed evaluation opportunity in 2004 GJJ.jpg
Figure 15. Concrete degradation and exposed rebar (circled in red) on the south lateral face of the east abutment in 1992
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)
Figure 16. Cracks in the bearing support area of the east cantilever in 2004
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)

Causes of the Collapse

The Commission concluded the collapse resulted from a chain of events and that no single entity or individual was responsible for the collapse. None of the defects or omissions identified in itself caused the collapse. The collapse resulted from an accumulation of shortcomings that included the design codes applicable at the time the overpass was built, which are considered inadequate today; the bridge's design; the construction work; and the management of the structure (Johnson, 2007, 5).

Further, the de la Concorde overpass collapse resulted from a chain of physical causes. The participants' and Commission's experts all agreed on the main physical causes, but their opinions differed on the secondary causes. The Commission believed that some of the secondary causes and the human interventions that allowed the physical circumstances of the collapse to develop were significant (Johnson, 2007, 5).

Lastly, the de la Concorde overpass collapse occurred instantly, but was caused by the accumulation of gradual deterioration that occurred throughout the bridge's life. The organizational and human causes included failure to fulfill obligations and to comply with procedures, incomplete files, lack of teamwork, missed detail evaluation opportunities, and failure to take in to account the bridge's special characteristics. On September 30, 2006, the bridge reached an advanced deteriorated state that caused it to collapse under its own weight (Johnson, 2007, 6).

Principal Physical and Contributing (Secondary) Physical Causes of the Collapse

During the investigation, experts agreed the overpass collapsed as a result of a shear failure in the south-east cantilever. A view and detail of the failed abutment can be seen in Figures 17 and 18. The deterioration behind the concrete and not of the rebar was behind the collapse and the development and growth of a crack in a weakness zone located under the upper bars starting from the beam seat area caused the collapse. During the bridge's life, the freeze-thaw cycles and the de-icing salts caused the concrete to deteriorate in the beam seat area. As a result, the deterioration caused a cracking plane to spread inside the thick slab.

View of failed abutment GJJ.jpg
Detailed view of the failed abutment GJJ.jpg
view of failed abutment (2) GJJ.jpg
Detail of failed abutment (2) GJJ.jpg
Figure 17. View of the failed east abutment
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)
Figure 18. Detail of the failed east abutment
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)

The exact source of the cracking is unknown, but the main physical causes of the collapse were found to be the following:
  • Improper rebar detailing during design
In the as-design structure, a plane of weakness, where horizontal cracking can occur, was created by the concentration of numerous bars on the same plane in the upper part of the abutment. In addition, shear reinforcement was absent from the details and the top No.14 bars were not anchored at the end. The code used at the time of the overpass' design, the CSA-S6-1966 code, did not require minimum shear reinforcement. During the investigation, it was found that the slab was insufficient to resist the dead and live loads based on the requirements in the CSA-S6-2006 code. In addition, current detailing standards, such as the CSA-S6-2006, would require the No.8 U shaped hanger bars to be hooked around the No.14 bars as well as require the use of shear reinforcement to strengthen the slab section. The shear reinforcement in the thick slab would had intercepted the weakness zone and controlled the internal cracking. The collapse would have been prevented or the collapse would have occurred gradually accompanied by noticeable deformations (Johnson, 2007, 6, 122).

  • Improper rebar installation at the time of construction
A potential horizontal weakness plane was created due to the high concentration of rebars at the top of the beam seat. A much larger weakness zone extending deeper inside the thick slab was created by the incorrect placement of U-shaped hangers and diagonal bars (Johnson, 2007, 6, 122).

  • Low quality concrete used for the abutments
The concrete was too porous and the air bubble network was deficient. Therefore, the concrete abutments did not have characteristics to resist deterioration caused by freeze-thaw cycles in the presence of de-icing salts (Johnson, 2007, 6, 122).

The possible causes of the cracks were as follows:
  • The high bond stress between the No.14 bars and the concrete in the bearing support area (Johnson, 2007, 7)
  • A weakness zone above the U-shaped hanger bars (Johnson, 2007, 7)
  • Concrete deterioration caused by successive freeze-thaw cycles in the presence of de-icing salts (Johnson, 2007, 7)
  • Concrete shrinkage at the level of the longitudinal bars (Johnson, 2007, 7)
  • Thermal effects created by hydration of concrete, solar radiation, and hot asphalt placement (Johnson, 2007, 7)
  • Repeated impact from vehicles and traffic passing on the expansion joint (Johnson, 2007, 7)
  • Corrosion of No.8 and No.14 bars (Johnson, 2007, 7)

Additional contributing physical causes were the lack of a watertight protective layer for the thick slab surface and damage caused during the 1992 rehabilitation. All of the investigation experts, except the MTQ experts, agreed that the 1992 repairs created permanent damage (Johnson, 2007, 173). The absence of a protective waterproofing layer exacerbated the deterioration of the concrete, which was one of the main factors that led to the collapse. Also, extensive damage was noticed during the 1992 repairs and more concrete than anticipated had to be removed, which exposed the U-shaped hanger bars and No.14 bars over a considerable distance. As a result, the 1992 repair work contributed in accelerating the growth of the critical crack that was already present in the mass of the cantilever (Johnson, 2007, 7). The general condition of the un-collapsed portion of the overpass with exposed rebar can be seen in Figure 19.

General condition of non-collapsed portion of the bridge GJJ.jpg
Figure 19. General condition of the un-collapsed portion of the de la Concorde overpass with exposed rebar
(Photo credit: Original Commissioner’s investigative report and the Ohio Department of Transportation)

Conclusions with Respect to Persons, Firms and Organizations Involved

The Commission found DSA, its supervising engineers and its managers responsible for failing to fulfill their contractual obligations of executing full-time construction supervision of the overpass and therefore "not preventing the faulty installation of the steel reinforcement that resulted in a structure not in accordance with the drawings and specifications," (Johnson, 2007, 8).

In addition, the Commission found ISP and its managers responsible for failing to satisfy their legal and contractual obligation for the construction site. ISP was found negligent in adequately controlling the quality of work on the site by passing on their responsibilities to the workers and consulting engineer. This lack of quality control resulted in the faulty installation of steel reinforcement, which was one of the main physical causes of the collapse (Johnson, 2007, 8).

Furthermore, it is the Commission's opinion that MTQ did not adequately take into account de la Concorde's vulnerabilities due to it being a unique structure that was difficult to inspect. Despite numerous signs of deterioration, MTQ did not effectively execute all the means to properly evaluate the condition of the overpass. MTQ also failed to maintain adequate records, which prevented its inspectors and maintenance workers from having better guidance when evaluating the structure (Johnson, 2007, 8).

Moreover, the Commission found that the overpass inspections were deficient, lacked adequate quantification of the deterioration, were incomplete because not enough time was devoted to the inspections and not thorough because the inspectors failed to look for the causes of the deterioration (Johnson, 2007, 8).

Finally, The Commission found that the Ministry missed at least two detailed evaluation opportunities of the structure in the 1992 rehabilitation and the 2004 special inspection. However, despite these two missed opportunities, the Commission found MTQ mostly to blame for "tolerating ambiguous accountability, for its lack of rigorous record keeping, and for not developing an adequate inspection and maintenance program despite knowing about the special characteristics of the de la Concorde overpass," (Johnson, 2007, 8).

Prevention

After the de la Concorde overpass collapse and investigation, the Commission suggested several recommendations to improve bridge design, maintenance, and repair. Among these recommendations were to update codes, standards, and manuals; for the Government to review the legal framework for the design, construction of structures, and construction supervision for new structures and major rehabilitation work; to improve record keeping practices for the condition of existing structures; clarify accountability with respect to the system used to manage the bridges in the municipal road network (MUNRN); and the need for a broad-based bridge rehabilitation program (Johnson, 2007, 8-15).

Lessons To Learn

Many lessons have been learned from the de la Concorde overpass collapse relating to design, construction, bridge management, rehabilitation, special inspections, and agencies involved in bridge projects.

Design

As was seen in the de la Concorde overpass, the bridge failure was brittle due to the absence of shear reinforcement in the thick slab, misplaced reinforcement and the lack of structural redundancy. As a result, structures need to have ductile failure mechanisms; reinforcement detail requirements need to be taken into account in designing reinforcement details, and need the presence of redundancy in the event of a failure. Further, the continuous bearing seat was un-inspectable, which made it difficult to inspect, maintain, and repair. This shows that connection details need to be inspectable to enable them to be properly inspected and accessed for maintenance and repair work. Lastly, the incorrect concrete type was used because the specifications that called for the type of concrete to be used were confusing, which can lead to a catastrophic failure in the structure's life. Therefore, specifications need to be clear and direct to ensure the correct material types are used in the structure to prevent future problems and failure.

Construction

During the construction of de la Concorde overpass, DSA failed to meet their legal contractual obligations in supervising the construction site and work quality. Therefore, construction operation and the material supply chain need to be staffed, supervised, and inspected by competent personnel and construction records and true as-built drawings need to be stored safely and kept organized to enable access to them for future reference.

Management and Inspection

When managing and inspecting the de la Concorde overpass, MTQ failed to keep records of their inspection documents and repair work; to include warnings in the files pertaining to the structure's condition or the need for in-depth inspections; to determine the cause of the deterioration; and to take necessary actions to restore the bridge. Therefore, bridge inventory management and inspections need to conform to standards and best practices, document findings, reply on records, and be used to identify defects and weaknesses in the structure and to ensure proper steps are taken towards restoring the structure after these discoveries. In addition, inspection records need to be stored safely and kept organized.

Rehabilitation

After the 1992 rehabilitation, MTQ failed to keep records of the 1992 repair work, such as the document plans and reports of misplaced rebar. Therefore, rehabilitation projects need to rely on records; thoroughly document the work; including as-built drawings; and provide warnings to the parties involved pertaining to the structure's defects and weaknesses. Rehabilitation records also need to be safely stored and kept organized (Russell, 2008).

Special Inspections

MTQ failed to keep record of a formal report on the special inspection performed on the overpass in 2004 and failed to detect the problems in the overpass during the special inspection. Therefore, special inspections must rely on records; require thorough field work; be inquisitive and not dismissive of their findings pertaining to structural defects and weaknesses; and consist of thorough documented results. They must also be stored safely to be retrievable for future reference and kept organized (Russell, 2008).

Agencies

Agencies need to expect and encourage professionalism of its staff, consultants and contractors; enforce conformance of their staff, consultants, and contractors to its standards and best practices; agency personnel need to admit and be prepared to take the blame for any oversights and mistakes made on a project (Russell, 2008).

Updates in the Industry Codes and Practices

After the de la Concorde overpass was constructed, improvements and updates have been made to the codes. Some of these updates include:
  • Continuous beam seats along spans are not allowed under current codes because they are inaccessible for inspection and maintenance (Johnson, 2007, 4).
  • Minimum shear reinforcement requirements were added to the code. It was found that as the thickness of the slab increases, the shear resistance of the slab decreases (Johnson, 2007, 122). As a result, shear reinforcement is required to meet strength requirements if the concrete slab by itself is found inadequate to resist the shear stresses.
  • The requirement for the No.8 U-shaped bars have to be anchored (hooked) around the No.14 main bars to prevent creating a zone of weakness where reinforcement is not present between the hooks and No.14 bars (Johnson, 2007, 72).
  • A minimum drainage slope is required to evacuate runoff water off the bridge deck (Johnson, 2007, 126).

Similar Cases

Similarly to the de la Concorde overpass collapse, several lessons are to be learned from the Kinzua Bridge collapse, Oakland Bay Bridge eye bar failure, Turcot Interchange failure, and Ville Marie Tunnel Collapse relating to maintenance, structural design details, material properties, material fatigue, redundancy, drainage, misplaced reinforcement, and the use of permeable concrete.

Kinzua Bridge

In the Kinzua Bridge redesign, the design detail of the collar coupling linking the existing anchor bolts to the new anchor should have been more thorough. Further, since the load bearing structural elements were exposed to the natural elements, the bridge should have been frequently inspected for the condition of the steel members, connections and other deficiencies. Lastly, if the bridge had structural redundancy, the failure of the bridge could have been less catastrophic.

Oakland Bay Bridge Eyebar Failure

Two lessons learned from the Oakland Bay Bridge Eyebar failure are that material properties represent a significant role in a structure's behavior and aging bridges need to be inspected and maintained more frequently. To begin, the bridge's repair connection failed due to material fatigue, which displays the importance of knowing about the material properties of structural elements and connections used in a structure. Lastly, During the 70 years the bridge was in operation, the bridge experienced considerable stresses due to vehicular traffic, temperature differentials, and the environment. As a result, it is important to have an inspection program in place for assessing both new and old bridge structures.

Turcot Interchange Failure and Ville Marie Tunnel Collapse

The Turcot Interchange lacked drainage and the overpasses were built with permeable concrete. As a result, the interchange's structural condition became poor while it was in operation and pieces of concrete slabs, measuring up to one square meter, fell from the overpasses. The additional problems associated with the interchange were the presence of concrete cracks and misplaced reinforcing bars. Furthermore, the interchange was designed to support 50,000 to 60,000 vehicles per day, but in 2000 it was found that more than 300,000 vehicles used the interchange on a daily basis, which is far more load than what the interchange was designed to support.

In addition, in 2007 working crews for Transport Quebec found major cracks in a support pillar of the Ville Marie Tunnel and as result several lanes of the expressway were closed. On July 31, 2011 part of the roof of the Ville Marie Tunnel entrance collapsed, resulting in large pieces of concrete falling onto the road below the tunnel. At the time of the collapse, repair and maintenance work was being performed on the tunnel. At the time this report was written, the cause of the Ville Marie Tunnel collapse was still under investigation, but transportation officials have speculated that the cause of the collapse was due to the construction activities being performed at the time of the collapse and not the quality of the infrastructure inspections.

The two cases discussed above, along with the de la Concorde overpass collapse, show a pattern in faulty design and construction of bridges. They highlight the importance of:
  • Using the correct concrete type for bridge construction.
  • Placing the reinforcing rebars in accordance to the design drawings during construction to prevent future failure.
  • Using drains and sloped pavement for drainage to prevent ponding water from accumulating on the bridge.
  • Using a waterproofing membrane to protect the concrete slab.

Conclusion

The de la Concorde overpass collapse serves as a reminder of the need to execute rigor and discipline when designing, building, and maintaining bridges. It highlights the importance of having a proper framework with standards, manuals, and strictly implemented programs to help inspectors and maintenance workers. It also stresses the importance of encouraging them to be aware and inquiring when they encounter problems on bridges under their responsibility. Lastly, inspections are extremely important and have a direct impact on a structure's lifespan. They allow an assessment of the structure and the determination of the interventions required to maintain the structure.

Annotated Bibliography


CBC News (September 30, 2006). "Hope fades for victims trapped in overpass collapse." CBC News. <http://www.cbc.ca/news/canada/story/2006/09/30/overpass-collapse.html> (accessed October 1, 2012).
  • This news article written on the day of the collapse describes the rescue efforts performed by the Laval firefighters to save those who were injured and to retrieve those who were killed during the collapse.

"Chain of Causes Contributed to Laval Overpass Disaster: Report." CBC News. October 18, 2007. <http://www.cbc.ca/news/canada/montreal/story/2007/10/18/overpass-report.html> (accessed October 1, 2012).
  • The news article summarizes the chain of events that led to the de la Concorde Overpass Collapse as well as includes several recommendations for bride funding and repair and improving bridge maintenance. The information provided in this news article are facts provided by Pierre M. Johnson in the original investigative report and his statements made to a news conference.

Johnson, Pierre M (October 12, 2007). Commission of Inquiry into the Collapse of a Portion of de la Concorde Overpass. Quebec: <http://www.cevc.gouv.qc.ca/UserFiles/File/Rapport/report_eng.pdf>.
  • The original commission’s 200-page investigative report.

Journal of Commerce (JOC) (January, 1, 2007). "Commission of Inquiry examines construction of de la Concorde overpass." Journal of Commerce: <http://www.journalofcommerce.com/article/id24934>.
  • This journal article summarizes some of the key points made in the original investigative report, including listing the advantages of the de la Concorde overpass design; describing the main structural load bearing support elements, such as the prestressed girders and the concrete abutments; and discussing the repair work and the damage induced on the structure by the bridge's 1992 rehabilitation.


Russell, Michael A (October 2008). "de la Concorde Overpass Collapse: Presentation." Ohio Department of Transportation. <http://www.dot.state.oh.us/engineering/OTEC/2008%20OTEC%20Presentations/51B-Russell.pdf> (accessed October 1, 2012).
  • A presentation given to the Ohio Transportation Engineering Conference by Senior Project Structural Engineer Michael A. Russell. The presentation includes facts provided by the Quebec Commission of Inquiry and coverage by the Montreal Gazette.

The Gazette (October 2, 2006). "Dust Settles, Questions Arise." <http://www.canada.com/montrealgazette/news/story.html?id=4aba69d2-0e1f-401d-b93c-1d41fe86c01f&k=64917> (accessed October 1, 2012).
  • This news article written two days after the collapse describes how local residents noticed the bridge was crumbling months before the collapse by noticing pieces of debris falling off the bridge. On the day of the collapse, residents called 911 to inform them that large 5”x14” pieces of concrete were on the ground near the overpass. Even after the discovery, the transportation department didn’t deem it necessary to shut down the highway. An hour after the bridge inspection, “three eastbound lanes and the pedestrian walkway came crashing down on the highway below.”

Versace, Vince (January 1, 2007). "Bridge Engineers Identify Fatal Flaw in Concorde overpass." Journal of Commerce: Western Canada's Construction Newspaper: <http://www.joconl.com/article/id24931>.
  • The journal article describes the main failure cause of the de la Concorde overpass collapse was the misplacement of reinforcement.

Versace, Vince (January 1, 2007). "Concorde commission finds lax inspection regime allowed decay to spread." Journal of Commerce: Western Canada's Construction Newspaper: <http://www.journalofcommerce.com/article/id24929>.
  • This journal article describes the poor maintenance and infrequent bridge inspection of the overpass allowed its structural deficiencies to expand, which contributed to the bridge failure. If the deficiencies were noticed sooner with increased bridge inspections, the de la Concorde overpass collapse could have been prevented. The inspections between 1993-2004 “noted that the expansion joints continued to leak and the concrete slab required an assessment, but the assessment was never performed.” In addition the bridge’s rating in 2002 was dropped from good to acceptable, but in 2004 it was given a rating of good despite the fact no work was performed to improve the bride’s condition.

Additional Resources


General Books LLC (2010). 2006 Road Accidents: De la Concorde Overpass Collapse. General Books LLC.

General Books LLC (2010). Bridge Disasters in Canad: Quebec Autoroute 15, de la Concorde Overpass Collapse, Quebec Bridge, Ironworkers Memorial Second Narrows Crossing. General Books LLC.
MacAlevey, F. Niall (August, 10, 2010). Structural Engineering Failures: Lessons for Design. CreateSpace Independent Publishing Platform.