Snow drift on lower roofs. 3

[diegotorres1]

HI everyone, I was wondering if i could get a little guidance as I am looking at the designing a canopy structure next to an existing building that has a 7'-0" parapet, and the height difference between the top of parapet and roof of new canopy is about 10'-6". I'm particularly looking at the potential snow drift, but I wasn't able to see anything along those lines in the ASCE 7. I did however, come across a diagram from the NBCCa 2015, were it has an exact depiction of what i am looking at. Any advice and guidance is well appreciated. I have attached a screenshot of the diagram in the NBCC 2015.

 

[JAE]

What exactly is your question?

 

[diegotorres1]

Sorry for the confused question. I want to know if there is a reference for this type of scenario under the ASCE Snow provisions. Also, is there a Parapet height that will null the snow drift leeward on the canopy structure i am designing?

 

[JAE]

I want to know if there is a reference for this type of scenario under the ASCE Snow provisions.
[blue] Well, simply the overall provisions for drifting on lower roofs.[/blue]

Also, is there a Parapet height that will null the snow drift leeward on the canopy structure i am designing?
[blue] We've always interpreted it this way:
1. The lower roof drifting is due to turbulent snow above the high roof, thrown into the air from the high roof or from the sky.
2. This turbulent snow eventually drops to the lower roof from above, not directly laterally where a parapet might affect it.
3. Thus, the parapet above does not affect the drifting below - other than create a higher wall on which snow can drift.
4. So in some sense, the higher parapet actually could intensify the height of drift, rather than lower it.
[/blue]

 

[SteelPE]

We would typically use the windward and leeward drift requirements as outlined in ASCE 7. In accordance with ASCE 7, the parapet will have no effect in reducing or increasing the amount of snow on your canopy (other than what JAE indicated in regards to the height possibly increasing the depth of the drift for a long roof).

I know I have had arguments proposed where the parapet will block the drift from forming on the canopy..... but I have never really bought into this argument.

 

[canwesteng]

Parapet will certainly block snow from the upper roof from landing on the lower roof, but it won't block snow drifts from the lower roofs piling up against the wall. NBC is certainly quite a bit more sophisticated with respect to snow loading than ASCE 7, but unless you are outside the US you will have to use the provision of ASCE.

 

[r13]

Along the wall-roof junctions and exposed corners, the air flow is restricted due to abrupt changes in geometry, thus more accumulation of snow/dirt than the open areas. This phenomenon is well understood and adequately covered by any codes, I believe. I agree that there will not have any snow slide off the higher roof with parapet wall, but when the wall is facing the wind, there should be some snow drop off, through hitting/bouncing from the wall, in close vicinity of the wall-roof (canopy) junction. As this case, with 10'-6" wall height above canopy, the addition can be significantly higher. I don't know if this is considered in the codes or not.

 

[JAE]

I suspect there is more snow in Canada but I would not ever assume a parapet “blocks” snow. I’ve seen very large drifts below large parapets and also have seen the snow turbulence above high roof parapets. Snow doesn’t always move in a linear, horizontal fashion above the high roof and just drop down nice and neat on the lower roof.

You could perhaps assume that the upper parapet snow drift represents a reduction in available high roof snow but that quantity is generally fairly small compared to the overall available snow.

 

[diegotorres1]

Thank you all for the inputs and suggestions. I have been deeply reading ASCE 7-10 Sow provisions and found a similar discussion in the "Frequently asked Questions" and it shows how some of the snow can be "captured" by the parapet, but only about 30% is done so, the rest is blown over onto the canopy. I have a large fetch distance, and it between that are a couple of parapet walls. Does this change the situation where you can assume some of the snow will be trapped by the parapet walls? i know that windward snow drift having the fetch distance of the canopy will also be considered, but i have been getting larger loads as the leeward towards the canopy. I have attached a sketch of the roof profile, hopefully this can illustrate better my case.

 

[jayrod12]

In the NBCC, if I run your building assuming it's half as big into the page as you show on your cross section I get the following:
1) Standard roof snow load = 36 PSF
2) Build-up accounting for Parapet = 134 PSF extending out 25 ft
3) Build-up ignoring Parapet = 189 PSF extending 40ft

Pretty significant reduction in loads taking this into account. The build-up on the high roof from the parapet would be 84 PSF extending 15ft.

 

[DaveAtkins]

ASCE 7-16 has new provisions for snow buildup on canopies, which is MUCH more realistic than previous versions of ASCE 7.

The angle of repose of snow is 30 degrees, so the snow drift (on top of the balanced snow) on the canopy is limited to a 30 degree angle from the edge of the canopy back toward the building.

DaveAtkins

 

[SWitmer]

I am trying to decipher the changes in ASCE 7-16 section 7.7.1 mentioned by DaveAtkins above as related to a canopy attached to a building.

I'm designing a project with a 5' wide canopy. It has a leeward drift from the upper roof. The calculated drift height would be about 3.6'.

The max drift height is limited to 60% of the canopy width per 7.7.1, or 3' in my case. What is confusing is the sentence near the end of this paragraph. "If the drift width, w, exceeds the width of the lower roof, the drift shall taper linearly to zero at the far end of the lower level roof." To me, this makes the most sense when talking about a leeward drift, as in my case, meaning the 3' drift tapers down to zero at the edge of the canopy. However, this sentence is after several sentences referring to windward drift.

Past threads have had input from engineers who have witnessed large drifts all the way to the edge of canopies, so either my interpretation, or the Code provision if I've interpreted it correctly, seems to be questionable.

Is it me or the Code?

 

[DaveAtkins]

You never get a windward drift on a canopy. Am I misunderstanding your question?

DaveAtkins

 

[SWitmer]

I agree that a narrow canopy like this one would never see a windward drift.

My question is about the sentence in quotation marks regarding drifts on lower roofs tapering to zero. Does this refer to windward or leeward drifts on lower roofs?

 

[jayrod12]

For both drift types it tapers to zero. the windward drift is just more compact of a triangle than the leeward drift. And by zero, they mean the additional snow load due to drifting tapers to zero, there is still the base snow load.

 

[SWitmer]

For typical drifts, I agree that both types just taper to zero.

The code wording that I'm questioning is talking about the special condition where the lower roof width is narrower than the calculated drift width. In my case, the 5' canopy width is narrower than the calculated leeward drift width (no windward drift since it is a narrow canopy). If the code section I highlighted is talking about leeward drifts on lower roofs, it seems to say the drift loading above base snow load should taper to zero at the edge of the 5' canopy.

The only alternative I see is that the drift loading would just truncate at the edge of the 5' wide canopy and be considerably more than zero at that point.

 

[jayrod12]

I've always truncated it at a non-zero (that is also accounting for leeward drifting), however I'm north of the border so can't comment directly on the ASCE provisions. I misread the quotation thinking it meant lower roof width exceeds drift length.

Inherently though, it would make sense that for a windward drift on a low roof, it would have to be zero at the edge of the roof. That would be where it starts to pick up snow to deposit adjacent the high roof.

 

[drbrainsol]

An interesting discussion here. Also worth noting that design codes (at least in the UK but I think this is a general principle in most legal jurisdictions) are considered Best Practice / Guidance documents. If the code provisions are not clear for your situation, you should really err on the side of caution rather than try to use as low a load as possibly allowable by the code.

Based on this logic, I would not consider a reduction of more than 30% in snow drift due to the upper level parapets in this situation. Also worth considering the removal of the higher parapets at some point in the future. Can you guarantee they will still be there in 20 years?

 

[jittles]

I've always truncated it at a non-zero (that is also accounting for leeward drifting), however I'm north of the border so can't comment directly on the ASCE provisions. I misread the quotation thinking it meant lower roof width exceeds drift length.

Previous versions of ASCE 7 would have you truncate the drift if the length of the drift exceeded the length of roof/canopy.

The changes in 7-16 to limit the height of the drift based on canopy projection AND to taper the drift to 0 at the edge have been a huge reduction in snow drift loading on canopies compared with previous versions of the standard.

 

[jayrod12]

I wasn't practicing at the time, however the 1995 NBCC allowed for some reduction on snow loading in general, drifting included. Then in the following years there were a few roof collapse issues in drift areas. Changes were made to take it back up to where it was before, and still is similar today.