Snow Drift Requirements of Old Building Codes

[sawoods]

I am currently evaluating an existing building structure. There is a small office space, about 30'x70'x12' tall, tacked onto the front of a warehouse space that is 25' high and a roof length of 550' that is dumping snow drift on the office roof. The ground snow load is 20 psf (Ohio). The building, both office and warehouse, were built in 1971. The drawings say the building was designed for 25 psf live load. In evaluating the office roof joists for their ability to carry the drift load, they fail spectacularly, just like seemingly every existing building I run into. We are working for the contractor on this job, who is working for the building owner. I informed the contractor about the structural issue, and that additional joists should be installed. Of course the contractor doesn't want to spend more money, so his answer was that as long as the building met the building code when it was built, and we don't add any load to it, we don't have to reinforce the roof. So my questions are: Why do all the existing buildings I evaluate always fail to carry the snow drifts (according to the calcuations, they work in real life)? Did old building codes not include snow drifts? If so, when did snow drifts become part of the code? Apparently Ohio didn't adopt a national code until the late 70's, and I can't find a copy of the Ohio Building Code from 1971 to see if it included drifts.

 

[ajh1]

The attached file contains one page from the BOCA 1970 code. The snow load section is all of two sentences long and states absolutely nothing regarding snow drifts.

The other two pages of the attached file are from the BOCA/Ohio 1978 code. The special snow loads are shown in an appendix (which didn't exist in 1970). The BOCA snow drift loads of the time were a function of the step height and the ground snow load ONLY. There was no allowance for upper length. That made them rather unconservative for large upper building lengths such as you have and sometimes overly conservative for very large roof step dimensions with small upper roofs.

The 1978 version was the earliest Ohio code I had in my archive.

 

[IsaacStructural]

in addition to Ajh1's comment, I believe some of the older codes didn't require an explicit check of unbalanced load conditions, or the requirements were less strict.

M.S. Structural Engineering
Licensed Structural Engineer and Licensed Professional Engineer (Illinois)

 

[JAE]

Older codes perhaps before 1985, didn't have anything for drifting.

 

[sawoods]

Thanks for the help. This certainly explains why none of the calculations work out when I'm looking at existing buildings.

 

[BAretired]

As a matter of interest, because this would not affect design in any part of the USA, I am looking at the 1970 Supplement No. 4 to the National Building Code of Canada. Both drift loads and sliding snow loads were considered showing triangular distributions adjacent to higher buildings or penthouses and provision for snow accumulation or scouring on sloping or arched roofs.

The historical notes indicate that the 1953 NBC considered using ground snow loads on all flat roofs with reductions permitted only for sloped roofs.

The 1965 NBC adopted a more rational approach to snow loads, allowing for varying influences which could cause accumulations of snow loads on roofs. I do not have the details of that as I foolishly tossed that code when the new one came out...oh, the rashness of youth.

BA

 

[dhengr]

My recollection is about the same as BA’s, we started paying attention to snow sloughing from roofs and drifting on lower roofs in the mid 60's. And, hats off to Canada, they were the early leaders in this research and effort. We didn’t need 20 pages of mathematical gymnastics, but we didn’t ignore it either. I’m sure the saving grace in existing bldgs. is conservative loadings, material properties, safety factors, etc. I’d look for any distress in the roof system from these kinds of overloads, give some consideration for the likelihood of historic snow events in the life of the bldg. proving its strength, and then likely agree with the contractor, unless we were materially altering conditions with the roof structure. Again, don’t be in a position to be accused of ignoring the matter, but what does that drifting usually do? It increases the stl. bar jst. reaction near the high wall, but that reaction or shear doesn’t often control the design; it has a lesser percentage increasing effect on the bending moment and deflection, so they continue to stand despite what the current code says. If the original roof design had been skinned to the bone, to save 2$ a square, this might be another matter.

 

[Brad805]

We have also found that the older tar and gravel roof systems coupled with poor insulation packages lead to a much greater amount of snow melt. We retrofitted a number of local schools in the early 90's after they installed new SBS roof systems with much thicker insulation and found that the roofs started to deflect excessively.

 

[haynewp]

I don't know your scope in doing the evaluation; I would write a letter stating it fails under current Code for drift and check for anything unsafe under that roof (like joist seats showing crushing). I agree with the contractor in what he is saying per IBC.

 

[msquared48]

In the 1979 UBC, unbalanced loading due to snow was mentioned, but not quantified, in Section 2305(d). It was left to the engineer and building engineer to quantify and identify any such conditions.

Same nomenclature in Section 2305(c) of the 1973 UBC.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com