Another snow drift thread

[mijowe]

I am trying to reduce a new snow drift load on an existing roof created by a building addition.

My question is, if I build several new parapets perpendicular to the wind direction spaced at a frequency so that the new snow drifts are small enough so that the existing structure works, can I eliminate the drift problem?. This would assume that I reduce the fetching length to the spacing of the new parapets. If I was analyzing a penthouse using this length would be a legitimate argument. However, the high roof with parapet argument has been frequently discouraged here.

On a side note ASCE 7 7.12 Existing Roofs reads: "Existing roofs shall be evaluated for increased snow loads caused by additions or alterations. Owners or agents for owners of an existing lower roof shall be advised of the potential for increased snow loads where a higher roof is constructed within 20 ft."

Does this release the owner of the new structure of any responsibility other than notifying the adjacent owner?

 

[SteelPE]

There are a number of "sketchy" ways to mitigate snow drift on lower structures.

I remember reading somewhere that if you were to try and prevent snow from drifting on another part of the structure, you would need to create a parapet that was tall enough to accumulate all of the balanced snow against the side of the parapet as a drift load (basically the volume of the drift load must equal the volume of balanced snow). If I remember correctly, the paper was written by the AISC and they came to the conclusion that it was an uneconomical way to deal with drifting snow.

Other ways discussed in the paper.... create a shed roof (4 to 6 pitch) which will prevent the "sudden" change in elevation and therefore prevent the buildup of drifting snow. I have also heard of geofoam being used for the same purpose. I have seen people create snowdrift roofs above existing roofs (this can be costly). I seen people create snowdrift roofs withing their new structures. All unique and a little "sketchy".... but allowed from what I can tell.

 

[dcarr82775]

If your series of little parapets disturbs the air flow, then it would theoretically deposit snow in all those various places instead of one big drift. It seems like a reasonable approach to get rid of the single large drift load, but instead I think you end up with something akin to the unbalanced snow loads on a gable roof. You still have that same contributing area of roof the snow gets moved from, but I think you can get it to distribute differently on the downwind side. Of course the unbalanced snow loads in ASCE7 keep changing significantly so don't sharpen the pencil too much.

 

[JAE]

I don't believe, based on the current code and general knowledge of drifting, that you can use parapets to spread out the drifting.

Wind turbulence, and drifting in general, doesn't follow that type of logical path but is probably way more random and complex than that.

I would think that parapets might help a bit - but how to quantify the degree of help would be impossible I think.

 

[DST148]

One way of preventing leeward drift from a high roof to a low roof is to build a screen wall at the
edge of high roof, high enough to arrest snow from falling on to the lower roof. However, the low roof
still needs to be checked for windward drift build-up against the wall of the higher structure / addition.
Screen wall may not be economical, but sometimes may be explored if reinforcing members of the lower roof is
not an option.
Sometimes the length of the roof downwind of the drift (lower roof) could be such that the windward snow build-up may
not exceed the height of the parapet of the lower roof.