Drifting Snow from Multi-level roofs

[EngDM]

Hey all,

I'm curious what the typical approach would be for calculating snow loads where you have multiple different levels of source areas, some of which already have drifts.

The situation I've got is a building with balconies/terraces at every floor, but the balconies increase in overhang as you go down. So I'd imagine my 10ft balcony up top could accumulate a snow load, and then this could blow to the 15ft balcony below, so on so fourth. The balconies could re-fill with snow between wind events and just keep blowing down the balconies.

I considered back calculating my Ss value to find what "base snow" of the top balcony drift is, and then using that amplified Ss in the drifting equation as I go down the building for subsequent drifts, but I get an insanely high drift, which if I divide the kPa by my snow density of 3 kN/m3 gives me a height thats larger than my floor-floor height.

Another problem I'm having is that the accumulation factor equation for Ca doesn't account for having source areas from both sides. By this I mean that the component lcs is based on length of longest side to shortest, not based on area. So I'm concerned that the Ca isn't accounting for both source areas for my drift if I calculate two seperate drifts for each of the source areas. I considered taking the overall plan dimension, but you end up taking into account a lot of dead space and get a very high load.

Here is a little sketch to illustrate:



This is a weird situation that I'm not really sure how to approach, but my floor type is hollowcore and I don't think 8" or 10" hollowcore can span 25ft with a 15kPa drift load.

 

[DanKile]

If ASCE 7 is applicable in Canada, this might fall into the open air provisions and allow the balanced (and fetch distance) snow load case to be reduced.

Outside of the drift requirements, I would check a load case of snow being shoveled from a higher balcony to a lower balcony.

 

[EngDM]

If ASCE 7 is applicable in Canada, this might fall into the open air provisions and allow the balanced (and fetch distance) snow load case to be reduced.

Outside of the drift requirements, I would check a load case of snow being shoveled from a higher balcony to a lower balcony.

I'll look into ASCE 7, but I don't think we have something similar.

As far as snow being shoveled, would you take the sum of the snow from balcony 1, and attribute it to balcony 2 but only in the area not covered in overlap? Like if balcony 1 has a 100psf drift, calculate the total lbs of that and spread it over the non-covered portion of balcony 2?

There is talks of having a heated floor and drain snow removal system but I'm not sure if I can rely on that as per the code, in the event that it becomes not operational during service life.

 

[DanKile]

I was thinking of a triangular load with the peak about a foot out from the upper balcony, but it might more similar to sliding snow (and be a uniform load). The total load and area of effect of this would be the balanced + drift over the non-covered area.

 

[skeletron]

Do the balconies have guardrails or parapets? If so, I wonder if this would actually inadvertently contain some of the snow drift or prevent the shoveled snow from Balcony 1 to load Balcony 2. There would need to be some sort of drainage path or mitigation strategy in place so you don't have piling along the sides or the situation you've discussed where a maintenance worker with a snow blower works their way down by firing snow over the guard to the balcony below. Potentially this could overload the bottom balcony if not accounted for.

It appears the balconies are in the inside of a U-shaped building.

1. Develop drift in N/S and then E/W direction separately.
2. Loosely plot out the drift shapes and average them in such a way to make the analysis easier. If that means taking the peak drift for a set distance and then stepping down the value once, so be it. Keep in mind that analysis programs don't allow for varying area loads (to the best of my knowledge), so you're gonna have to do this anyway if you're running a model. If you're keeping it simple, then even more reason to use conservative rectangular loads before sharpening the pencil.
3. I would expect that the height difference to Balcony 2 is deep enough that its source area is not the roof, but the balcony above. You could get snow piling against the building, but I wouldn't expect the roof snow to dump down and create a traditional drift. There would need to be some snow lost to the wind. But your heights should determine whether this assumption is applicable.
4. I would expect that the snow from Balcony 1 could create extra load on Balcony 2. If it's drifting, it's a bit hard to imagine because you would also have a bit of wind shield from the roof above. But it might be shoveled over the edge...which is probably the more practical way to analyze it. Assuming it's just shovel load, assume that all of the balcony snow loads the overhang part of Balcony 2. Again, I'd probably assume it as a average rectangular load unless your balcony is super long.
5. Repeat the process for Balcony 2 to Balcony 1, but the snow on Balcony 2 would represent the maximum case of "drift against the U" or "shoveled over the edge" as your source from above.

I would go through that process just to get to some sort of finish line. Design the slab/floor framing/whatever, then reflect on your comfort level. If you're pushing the 8"/10" typical slab really hot (...the one that your gut is telling you probably doesn't work), then I would go back through the analysis and attack it from a different (possibly more conservative) angle. If you're not pushing the 8"/10" typical slab as much as you thought (...and this goes against your original gut), then go back through the analysis and see where you sit with your original assumptions. I don't think it's reasonable or practical to assume a load equivalent to the floor-to-floor height (12ft?) unless you're in serious mountain country, but then again the situation is kind of unique.

Someone mentioned ASCE 7-22. The method is way more in depth than we are required by the NBCC, which is odd given Canada's snowier climate. I would recommend Michael O'Rourke's book (https://ascelibrary.org/doi/book/10.1061/9780784416136) because he seems to touch on every component considered in the development of the snow provisions. A few examples too. It might help to put some context and confidence into your problem.

 

[lexpatrie]

Well, I mean, Michael writes the provisions for snow load in ASCE 7.....