Supermarket Roof Collapse - Burnaby, B. C., Canada

James N. Wilson, BAE/MAE, Penn State, 2009


On April 23, 1988 the rooftop parking deck of a supermarket collapsed at Station Square in Burnaby, British Columbia, Canada. Five minuts after the opening ceremony, four bays of the roof structure came plummeting to the ground. Despite there being an estimated 600 customers and 370 employees in the building when signs of failure started to occur, no one was killed, but 21 people were injured (Closkey 1988, pp.1).

The government of British Columbia established a Commissioner Inquiry to find the causes of collapse and to make recommendations that could prevent similar occurrences. Their report attributes the cause of failure to an undersized steel beam in a Cantilever-Suspended Span arrangement and inadequate buckling resistance of the beam compression flange. In addition to the technical errors were a number of procedural deficiencies in the project delivery system. The following provides a brief summary of the collapse including background information, causes of failure, recommendations for future prevention, similar failures, and lessons learned. In order to produce this Wiki, information was mainly retrieved from the report of the Commissioner Inquiry and was supplemented by information reviewed in books, research papers and news articles.

Key Words: Steel Beams, Station Square, Save-on-Foods, Cantilevered-Suspended Span, Gerber System

Figure 1: Overhead Photo of Roof Collapse. Image Courtesy of Commissioner Inquiry



Figure 2: Roof Cross Section. Image based on report prepared by Sandwell Swan and Wooster
The rooftop collapsed over a one story Save-on-Foods store, which was part of a regional community shopping center that included a hotel, apartments, retail space, theaters and offices. The 9000 square foot supermarket was a 400ft by 225ft (122m by 69m) rectangular building. Located on the rooftop was parking for customers, accessed by ramps that led up to the roof deck (Closkey 1988).

The supermarket structure consisted of composite steel roof decking supported by steel joists, beams and columns. The composite steel roof decking was topped with a water proofing membrane, light styrofoam insulation and a top-slab wearing surface (figure 2). Supporting the deck were 42 inch (1067mm) deep open web steel joists spaced at approximately six feet (1.82m). A Cantilever Suspended Span system, also known as a Gerber system, was used for the configuration of steel framing that supported the steel joists. As illustrated in figure 3, the beams do not simply span between the columns, instead, certain beams cantilever beyond the columns, while others are suspended between the cantilevers. When designed correctly this system can provide a more economical design due to reduction in steel section sizes, allowed because of lower stresses induced in the beams (Essa et. al. 1995, pp.1667). The beams used in the Save-on-Foods store were 24 inch (610mm) deep wide flange sections. The columns were 12x12 inch (305x305 mm) hollow steel sections supported on concrete spread footings (Closkey 1988, pp. 13).

Figure 3: Gerber System or Cantilever Suspended System. Image based on report prepared by the Commissioner Inquiry


Figure 4: Trevor Burkitt Photo
On the day of the failure the store had a special preview opening for neighborhood senior citizens. Shortly after the store had opened there were signs of the roof structure starting to fail. Witnesses reported noises coming from the roof structure. Some described them as "Popping and cracking noises" while others said "it sounded like thunder"(Closkey 1988, Spokesman Review 1988). A beam was observed to be rotating above an interior column location and caused an overhead pipe to burst and spray water over the bulk-foods area (Closkey 1988 pp.1). An amateur photographer, Trevor Burkitt, was able to take a photo of the roof before it collapsed (figure 4). Luckily, the occupants acted quickly and started evacuating the store. About 5 minuts after the first signs of failure had been observed, four bays of the roof along with 20 cars collapsed into the produce department.

What Failed?

Photographs identified that the collapse was attributed to a failure in a beam and beam-column assembly at the location indicated on figure 3. At such a location, the beam experiences negative bending. Figure 4 shows how Sandwell Swan and Wooster describe the failure at different stages of the collapse. The cross sections in figure 4 shows the bottom flange of the beam rotating to the side, trying to escape compressive loads, indicating lateral torsional buckling. According to the report of the Commissioner Inquiry, the beam moved about 21 inches (533mm) to the side, and was also the cause a ruptured water line that gave warning to staff and shoppers below. Why the roof did not collapse immediately was not rigorously explained in the report of the Commissioner Inquiry. However, it was believed that the deck, joists, and beams acted together as a membrane in tension to delay the collapse (Jones et. al 1990, pp.159).
Figure 5: Cross Section Through Column and Beam During Collapse. Image based on report prepared by Sandwell Swan and Wooster

Causes of Failure

There were a number of technical and procedural errors or that led to the collapse of the roof structure. The Commissioner's report states that the principle causes of failure were an undersized steel beam and inadequate buckling resistance of the beam compression flange. Other procedural contributors were unclear assignments and responsibilities of members involved, incomplete peer review and the competitive bidding process used for selecting design services.

Undersized Steel Beam

The report states that a beam of inadequate strength was shown on the design drawings due to a miscalculation in the office of the structural engineer. Summarized in the report are a number of non-conservative design decisions, assumptions, judgments or errors made by the design engineer that reduced the margin of safety.

The report discusses three dead load calculation errors made by the design engineer, that when combined, resulted in a dead load 55% higher than that designed for (Closkey 1988, pp. 23). Firstly, the thickness of the top-slab was initially assumed to be 2 inches (51mm), however, it was later increased to 3 inches (76mm); the beam design did not account for this increase in concrete. Secondly, during construction, extra concrete was placed in order to keep a plane-sloping top-surface under the deflection of the supporting elements. Finally, a walkway was widened using concrete instead of foam-type insulation, as specified in the drawings. Therefore, there was an additional three foot - ten inches of solid concrete of dead load on the beams.

Other errors included miscalculating live loads, unconservative selection of plane at which the critical moment in the beam was calculated and incorrectly assuming an upgrade in yield strength of the steel members based on mill certificates (Closkey 1988, pp.23-25).

On top of all this, for reasons not fully known, the beam that failed was reduced from a W24x104 to a W24x76. Throughout the design process, a number of revisions were made, during which the beam was reduced in size. The Commissioners' report states that they were unable to obtain the supporting calculations made by the design engineer (Closkey, pp.24). However, it was likely due to a design or drafting mistake (Jones et. al, 1990 pp.159)

Inadequate Buckling Resistance

The Commissioner's report also attributed the collapse to a second mode of failure which was inadequate buckling resistance of the compression flange. The design engineers had not at any point analyzed stability of the beam column assembly (Jones et. al. pp.159). Research conducted by Essa and Kennedy revisit the failure analysis using distortional buckling finite element software. Findings from this research reveal that "even with improved restraint details at the critical beam-column location, the beam would have been inadequate to support the factored load".

Independent Review

After large deflections were noticed, a third party contractor was hired to review the structure. The contractor identified the beam as suspect to critical but remedial work was not performed. The Commissioners Report identifies that the contractor made numerous miscalculations, many of them similar to those made by the design engineer. Summarized in the report are seven factors, decisions, assumptions or errors that the contractor made that resulted in a reduced factor of safety (Closkey 1988, pp. 29).

Competitive Bidding

The Structural Engineer for the Save-on-Foods store was selected through competitive bidding. This drove the drove the fee levels so far down that it raised questions about the quality of professional services provided. The bid was won for a fee of $5.4 million, which was subsequently negotiated down by the building's development manager (ACEC 2009, pp.2). The Commissioners Report discusses how such a low fee made it challenging to perform the provided services to the degree of precision that is required (Closkey 1988, pp. 42,49).

Unclear Assignments and Responsibilities

The Commissioner Inquiry studied the responsibilities and obligations of the different members involved in the design and construction for the Save-on-Foods store. A trend that brought concern at the time of construction, which is one that continues today, is the fragmentation of practices. With a construction industry that involves many different contractors, coordination between all of them can brings great challenges. Also, varying organization models in the construction industry make it difficult to delegate responsibility and liability.


The report of the Commissioner Inquiry made a number of recommendations to prevent similar occurrences from happening again in the future. The following bullets summarize the recommendations made (Jones et. al 1990 pp153-158). These recommendations continue to influence important to standards of practice in Candada (Feld 1996, pp.193).
  • Implementation:
    • Develop a task force to oversee the implementation of all recommendations

  • Technical recommendations to the steel industry:
    • The Canadian Institute of Steel Construction (CISC) should consider adding a guide on the Gerber system to the Handbook of Steel Construction

  • Building Regulations and Inspectors:
    • The Code should clearly define the function of the building inspector
    • Require the design professional to submit a letter certifying that the design meets code requirements

  • Construction Industry Procedures:
    • A manual of construction practices and procedures be developed

  • Engineering Standards:
    • Higher standards for consulting engineers
    • Structural calculations be submitted with the drawings and specifications when applying for a building permit
    • Provincial standards of practice should be established for the preparation and submission of all building design drawings and calculations
    • Create a Review Panel that at random or for reason review 10% of all submitted design projects
    • Companies be registered by the APEBC and be subject to registration under unethical, unprofessional or incompetent practice
    • Structural engineers be required to carry a specified amount of professional liability insurance

Similar Failures

Hartford Civic Center, 1979

A similar torsional instability failure to the Save-on-Foods store occurred in the collapse of the Hartford Civic Center , where heavy snow loads caused the roof structure to collapse on January 18th, 1979. In this case, the top chords of the space frame roof structure failed under compression loads due to lack of lateral bracing. Another similarity between these failures is that even when large deflections were reported under construction the error went overseen.

Magic Mart Store, Bolivia, Tennessee, 1983

Failure in a building with similar constriction details to that of the Save-on-Foods store occurred in a regional shopping center in Boliviar, Tennessee on July 2, 1983 (Feld, 1996 pp. 197). During a rain storm, the ponding of rain-water on the roof caused the entire roof structure to collapse. The failure was similarly attributed to lack of lateral support in the bottom flange of a beam in a cantilever-suspended-span system.

Conclusion / Lessons Learned

If it weren't for the 4.5 min delay in the roof collapse of the Save-on-Foods store, that allowed for occupants to evacuate the building, the outcome of this failure would have been catastrophic on a whole other level. It is important to ensure that similar failures will not occur in the future. There are many technical and procedural lessons that can be drawn from this failure, many of which are presented in the Report of the Commissioner's Inquiry. The following summarizes the lessons learned by the author of this case study.

A major lesson learned, is the importance of performing design-checks for torsional instability in steel beams. In simply supported beams, maximum positive bending is experienced at midspan, where lateral resistance is often provided by the decking (Feld 1996). However, in some framing configurations, including overhanging, cantilever and continuous beams there is compression in the bottom flange of the beam, where lateral support may not be provided. In these cases lateral support is a critical design consideration which has proven to be overlooked in the the preceding failure discussions.

Another lesson learned is that a competitive market can create rushed work that isn't carried out with the precision required. Peer-review and checking procedures need to be conducted with care to mitigate mistakes. When larger deflections than expected are encountered during construction, a thorough review must be undertaken. If this would have been done in the Save-on-Foods store or the Hartford Civic Center, collapse may have been avoided.

The underlying lesson learned by the author of this case study is the ethical responsibility that the construction industry has toward ensuring the safety of the public.


Report of the commissioner inquiry - Station Square development, (1988). Government of British Columbia, Victoria, British Columbia, Canada.

  • Report: The government of British Columbia established a Commissioner Inquiry to find the causes of collapse and to make recommendations that could prevent similar occurrences.

The Spokesman Review (April 25, 1988) "Authorities Investigate Roof Collapse", Spokane Washington, pp. B8

  • News paper Article: The article was released shortly after the Save-on-Foods store roof collapse and summarizes some of the key events. It also provides insight into the dramatic experiences of witnesses.

Jones, C.P. and Nathan, N.D., (1990). "Supermarket Roof Collapse in Burnaby, British Columbia, Canada," Journal of Performance of Constructed Facilities, ASCE, Vol.4, No.3, pp.142-160.

  • Journal Article: The authors provide a case study on the supermarket roof collapse, including background information on the failure as well as a discussion on the investigation and lessons learned. In short, it is a condensed summary of the report prepared by the Commissioner Inquiry.

Bear, B.R., (1992). "Discussion: Supermarket Roof Collapse in Burnaby, British Columbia, Canada," Journal of Performance of Constructed Facilities, ASCE, Vol.6, No.1, pp.67-68.

  • Journal Article: Bear provides a response to the journal article of Jones and Nathan (Jones et. al. 1990). He poses a series of questions with respect to the journal information or lack there of.

Jones, C.P. and Nathan, N.D., (1992). "Closure: Supermarket Roof Collapse in Burnaby, British Columbia, Canada," Journal of Performance of Constructed Facilities, ASCE, Vol.6, No.1, pp.67-68.

  • Journal Article: Jones and Nathan reply to the questions posed by Bear (Bear 1990).

Moore, D.B. and Weller, A.D., (1990). "Lessons from a Canadian Carpark Roof Collapse", The Structural Engineer, Vol.68 Issue 12, pp 229,230

  • Journal Article: The Building Research Establishment, assisted by Bradshaw, Buckon & Tonge study the report prepared the Inquiry Commssioner (Closkey, 1988) with an intention to determine if there are any lessons that can be learned for the recommendations in BS 5950: Part 1

Essa, H. and Kennedy, D.J. (1994). "Station Square Revisited: Distortional Buckling Collapse" Canadian Journal of Civil Engineering, Vol. 21, no. 3 pp. 377-381

  • Journal Article: Essa and Kennedy revisit the failure analysis of the Save-on Foods store using distortional buckling finite element software. Findings from this research reveal that even with improved restraint details at the critical beam-column location, the beam would have been inadequate to support the factored load.

Essa, H. and Kennedy, D.J. (1995). "Design of Steel Beams in Cantilever-Suspended-Span Construction" Journal of Structural Engineering, ASCE, Vol. 121, Issue 11 pp. 1667-1673

  • Journal Article: Essa and Kennedy address the inadequacy of current design procedures of joist-beam connections in cantilever-suspended-span systems. In their article they provide a simplified design procedure for these beams, determined through finite-element modeling and tests.

Shepherd, R. and Frost, D.J., (1995). Failures in Civil Engineering: Foundation and Geoenvironmental Case Studies, ASCE, New York, N.Y.

  • Book: Shepherd and Frost provide a brief case study on the Save-on-Foods store roof collapse.

Feld, J. and Carper, L.K., (1996). Construction Failure, 2nd Ed., John Wiley & Sons, New York, N.Y.

  • Book: Feld and Carper discuss case studies on stability failures in steel structures and some of the lessons that can be learned. The two case studies they focus on are the roof collapses of the Hartfrod Civic Center and the Save-on-Foods store.