Snow Shed - What slope should it be at?

[EngDM]

If I want to build a snow shed to avoid causing a snow drift on a high low condition, does my snow shed slope need to match the slope of the drifting snow pack? I want to avoid inducing any sliding, so my first thought is if my drift is a 1:5 slope, then my shed should match that. I'm not familiar if NBCC has any mention to snow sheds in Part 9 maybe. I usually work in Part 4.

 

[ANE91]

Sketch? Sounds like you’re describing a capture wall.

Sliding has to do with slipperiness of the roof covering as well as slope and whether the building is heated or not, if I recall correctly. Drifts are different; the slopes can vary as the drift forms.

 

[EngDM]

Sketch? Sounds like you’re describing a capture wall.

Sliding has to do with slipperiness of the roof covering as well as slope and whether the building is heated or not, if I recall correctly. Drifts are different; the slopes can vary as the drift forms.

Not sure I know what a capture wall is, but here's quick sketch of the cross section. Yellow line is the drift, red line is the snow shed would be framed with pony walls and joists between the walls. Basically prevents the high-low drift from forming by elimnating the "step" portion of the high-low, but I don't know if I need to match the same slope as the drift, or if I could use a steeper angle (whatever the max angle is before I start getting sliding is. I think NBCC has this at 30°).

 

[ANE91]

Ok got it.

Yep, this config eliminates from consideration the roof-step drift, assuming 2ft is short enough that h_c/h_b < 0.2. Slope doesn’t matter.

However, a different type of drift will still form, regardless; we would call this particular drift an unbalanced snow load in the US. I see this as functionally no different than a sawtooth roof with wind blowing right-to-left. You can play with the variables to minimize this condition.

You may be able to eliminate sliding by demonstrating that snow on the right blocks the snow on the left from sliding down, per the second paragraph of 7.9 of ASCE 7-22. Or just make the thing shallower than 2:12.

 

[EngDM]

Ok got it.

Yep, this config eliminates from consideration the roof-step drift, assuming 2ft is short enough that h_c/h_b < 0.2. Slope doesn’t matter.

However, a different type of drift will still form, regardless; we would call this particular drift an unbalanced snow load in the US. I see this as functionally no different than a sawtooth roof with wind blowing right-to-left. You can play with the variables to minimize this condition.

You may be able to eliminate sliding by demonstrating that snow on the right blocks the snow on the left from sliding down, per the second paragraph of 7.9 of ASCE 7-22. Or just make the thing shallower than 2:12.

Gotcha. Went with 3:12 since that still didn't induce any new drifts per NBCC, and checked sliding as a sawtooth. Makes more sense in my head now.

 

[ANE91]

Thanks for teaching me about snow sheds. Creative. Never thought about it before.

 

[dold]

I've designed "snow canopies" before. Where the entire snow drift from the higher roof is supported (captured) by a canopy which is suspended off of the new structure's wall. Interesting workaround. In this particular case it was built above the lower roof (same owner/property) of the older building.

 

[EngDM]

Thanks for teaching me about snow sheds. Creative. Never thought about it before.

I had a drift of nearly 125psf without it. Needless to say the glulam below was not having that lol.