CROSS Safety Report
Partial collapse of a shopping mall roof under drifting snow
Partial collapse of a shopping mall roof under drifting snow caused by lack of building code provisions for drifting snow and poor fabrication of open web steel joists.
Key Learning Outcomes
- Consider periodic structural reviews of older buildings
- Note that satisfactory performance over 20 or more years is not a guarantee that all is well
- Be aware that abnormal loads, from snow or other environmental effects, may be higher in future than have previously been experienced
For designers and structural engineers
- When assessing older buildings be mindful of structures designed for previous codes that may not be relevant to current circumstances
- If there have been roof failures due to snow loads in the region, then consider whether there may be any general issues with structural members such as roof joists
- In the event of a failure always ask why did it happen now?
Find out more about the Full Report
Our secure and confidential safety reporting system gives professionals the opportunity to share their experiences to help others. If you would like to know more, please visit the reporting to CROSS-US page.
In 1985 a portion of a roof of a shopping mall in central Ohio collapsed during a winter storm, which produced deep drifts of snow in some areas of the roof. The one-story structure was built around 1966. In the area of the collapse the structure was steel framed with 40 ft x 20 ft bays, consisting of 1-1/2 in. metal deck, 24 in. deep open web steel joists spaced at 6’-8” and spanning the 40 ft bay dimension, and 12 in. deep wide flange beams cantilever-framed in the 20 ft bay dimension. Adjacent to the collapse area there was a 4’-6” upward step in roof elevation, causing 50 to 60 in. deep drifts of snow, 24 to 36 ft long, to accumulate on the lower roof section that collapsed. The open web steel joists were perpendicular to the change in roof elevation and framed into a beam that ran along the terminus of the low roof.
Thorough investigations by two experienced investigating engineering firms showed that the collapse initiated by failure of the open web steel joists on the low roof. The modes of failure were buckling of the first compression diagonals of the joists, brittle fracture of the end-tension diagonals, or both. The investigations included (1) careful measurements of snow drift depth, length, and density and (2) laboratory load testing to failure of several joists extracted from areas of the structure near the collapse area.
The investigations concluded that several factors caused the failure, as summarized below:
- Design snow load: The building was designed under an older code that did not include provisions for drifting snow. The actual drifting snow produced end-shears in the open web steel joists that were 38% higher than required by the code.
- Selection of roof open web steel joists: The joists shown on the design drawings had 32% greater design strength than required by code, but the joists used in the actual construction had only 82% of design rating of the joists shown on the design drawings. The result was that the joists used in the actual construction had an 8% (1.32 x 0.82 = 1.08) greater design rating than required by code.
- Capacity of roof joists relative to expectation: The load tests of joists removed from the structure showed they had only 73% of their expected strength. (Rather than an expected strength of 1.65 times rated design capacity, the joists had an ultimate strength of only 1.20 times the rated design capacity.) The joists were produced by a manufacturer no longer in business. Their members were cold-formed steel, hat-shaped for the top and bottom chords and V-shaped for the web members. The fabrication resulted in large member eccentricities at the joints, non-uniform stress flow in members, and embrittlement of the cold-formed steel. These fabrication deficiencies caused premature buckling of the compression diagonals and premature, brittle fracture in the end-tension diagonals.
The result of the above is that rather than an expected minimum factor of safety against failure of 1.65, the structure had the following factor of safety at the time of failure:
FS = (1/degree to which the loads exceeded code) x (the excess “design strength” of the open web steel joists over code) x (the measured factor of safety of the joists relative to design rated capacity (cf. 1.65))
FS = (1/1.38) x (1.08) x (1.20) = 0.94
Failure is expected at 1.0 or less.
Why did the structure fail almost twenty years after construction and not before? The summer/fall before the winter of the failure there was a non-structural renovation of the mall. In that renovation, several non-loadbearing steel-stud-and-gypsum board partitions that had been built up tight to the underside of joists that failed were removed. Those areas of the structure that experienced the same drifting snow but for which the partitions were not removed did not collapse.
Based on the findings of the investigation, approximately 270,000 SF of the mall roof structure employing the defective joists was reinforced.
Submit a report
Your report will make a difference. It will help to create positive change and improve safety.
Our secure and confidential safety reporting system gives professionals the opportunity to share their experiences to help others.
No feedback has yet been published for this page.
Expert Panel Comments
An Expert Panel comment on the reports we receive. They use their experience to help you understand what can be learned from the reports. If you would like to know more, please visit the CROSS-US Expert Panel page.
This is a report of an historic failure from which lessons can be learned today. In 1966, few codes even had very specific values for snow loads and even those that did were not very specific by today's standards. Because of the age of the collapse we do not know all the details that might be considered relevant now. However the IBC has been revised and upgraded several times over the last twenty years to more adequately address snow loads, and especially drifting snow loads.
As with many failures, this one had multiple contributing factors and may not have been just a building code issue. Manufacturing and quality control issues with the joists may also have been a factor, if not in this case, then in other similar buildings. Another factor may have been additional dead load due to re-roofing and/or additional insulation which may have led to a greater build-up of snow than had been experienced previously. Were there any changes to the adjacent terrain / buildings that would have resulted in a changed drift pattern from past years? That said, the biggest cause of snow failures is drift and those structures built before drift was included in the codes and standards are the most at risk.
Owners are not generally expected to re-check their structures against the current code provisions at specified intervals and upgrade/update as necessary, so this kind of failure may continue to happen as we keep pushing the limit of materials and technology. However the fact that the structure collapsed after 20 years of apparently satisfactory performance is a worry.
Other structures that haven’t considered drifting snow are more at risk, but are not necessarily in imminent danger of collapse. However once there is a failure from a storm, especially if more than one structure has had a problem, other similar buildings, of a similar vintage, in the area should have their roofs checked for damage or load capacity.
The quality of the joists is questioned by the reporter and of course there can be a risk with off the shelf products but the situation is much better today with SJI and AISC member certification. That said, production problems still occur, and CROSS-US knows of a number of retail and storage buildings that have experienced quality problems in more recent years and required remediation. Of course, the SEOR is always responsible for the adequacy of any products, such as open web steel joists, that are designed by third parties and incorporated into the SEOR’s design. Not all products are equal, and an SEOR should specify that joists must be provided from a SJI certified company with complete shop drawings and PE sealed calculations.
When an older structure collapses we must always ask; why did it happen now? Just because it has stood for 40 years doesn’t mean it can’t or won’t fail.