Where to place snow guards?
Part of my retrofit included a standing-seam metal roof, the type
with snap-together panels. I’ve read some of the folks’ prior
experiences with metal roofs and snow, and spent some of my first
winter watching the big cornices slowly form over the edge and
then thud down — in the back of the house, onto the basement
bulkhead and outer edge of the little matching rooflet over the
heat pump [and boy, am I glad I constructed *that* to be strong],
and shoveled most of what fell farther away from the house so it
wouldn’t form a huge mound and then melt toward the basement wall
later. It was kind of annoying and probably not very good for
the basement bulkhead. Now I’m thinking it would have been much
easier to let all that melt and run off through the drainage
system.
The front has a 12/12 pitch and no critical infrastructure under
it, so I’m not going to worry about that side for now. The back
faces south, and has a 4.6/12 pitch or so on the shed dormer that
forms most of the roof’s width and that’s where I want to put snow
guards. From what I read it’s in the “critical region” of pitch
where snow guards make the most sense. The roof panels there
have a run length of about 16 feet. Both front and back have a
run of gutter, so snow sliding down and slowly curling over had
to negotiate over the leaf-guards and around the outer edge of the
gutter as well and while it managed to bear the load this time, it
was clear that it was under some amount of strain at times.
So to mitigate all this I’m looking at the S-5 “colorgard” system
and have already done the load math for the assumed “two-plus feet
of wet crap” worst-case they would have to hold back. The rear
overhang is two feet, built out over the “beer cooler” of 4″ polyiso
on roof and walls so there’s a pretty generous soffit “cave” under
the lower two feet of roof run. With the insulation, there shouldn’t
be any heat-leakage problem from underneath — I noticed no evidence
of anything resembling ice dams, just the whole “glacier” slowly
sliding down and big chunks of it dropping off during sunny days.
So the question is, with this particular roof structure, does
anyone have opinions on the best placement of the snow guards
relative to the exterior wall? Does it matter much in this case?
I was figuring on a foot or two up from the bottom edge of the
roof like most of them get placed, leaving only a little snow
below them to fall off, but wanted to draw on any prior experience
where better-insulated structures are concerned.
Even while thinking ahead to next winter, I’m also discovering
one other downside with metal roofing: wasps love them, and have
already been trying to move into the spaces under where the shed
dormer splits from the main roofline. There are a few small
unavoidable holes where the soffit and roofing couldn’t
completely mesh, which I predicted up front in the writeup
would be an attractive place for insects. For now I’ve stuffed
some fiberglass into the visible holes in a way that won’t be
in the path of any water, but if anyone’s got better ideas…
_H*
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
From a quick read, it sounds like you are planning on having the snow on one side of your roof, and no snow on the other. If that is correct, that is very dangerous. Roofs are generally not designed to have a load on one side only, and could (and do) collapse. As for the placement of the avalanche clips, I am not an engineer but I see no reason to not let the roof stay fully loaded until it melts off, if it was designed for that load. In other words, put the clips near the edge of the roof; that worked for me. That said, you may want to see an engineer to see if you need another row up higher. A bunch of snow on a steep roof generates a fair load along the roof. Falling snow is extremely dangerous. If you have not been caught in a roof load that has slid off, I would advise remaining inexperienced thereof. The snow packs around you quite tightly, just like in an avalanche. Been there, done that. I hope my humble opinions help.
You need to put more than one set in if you're expecting a fair amount of snow. Typically, snow builds up on a metal roof and then lets go once it reaches a certain weight. If you only have the guards at the bottom, the snow can have a lot of momentum. I've seen set ups like you're proposing get torn off.
You don't need a full snow guard at the higher spots - some clips like these - http://www.snojax.com/store/index.php?l=product_detail&p=11 - would be fine.
And we've done snow guards on one side only without ill effect, but it certainly wouldn't hurt to run it by en engineer.
Just be careful with unbalanced loads. When we had extreme snow one year, a few roofs collapsed because people were shoveling off one side at a time. A quick search revealed this sites of interest. Unbalanced loads can be problematic, and I would sure never design a system that was going to have them for sure. Cheers.
http://www.structuremag.org/article.aspx?articleid=796
http://voices.yahoo.com/balanced-unbalanced-snow-loads-gable-roof-3501177.html?cat=15
http://www.eng.auburn.edu/users/staylor/snow_loads_basics.PDF
Okay, I've been through the various references, and the only thing
I can determine hazardous about "unbalanced loads" is the fact that
the heavier portion exceeds the capacity of the structure immediately
underneath.
One of this past winter's major storms came with high winds from a
fairly uniform direction, and most of the houses ended up with a big
pile of drifted snow on the south sides of their roofs and almost
nothing on the north side. This would have happened whether anybody
had snow-guards or not, and on all the asphalt-shingle roofs, so we
really have no control over that. Besides, my roof is asymmetric
to begin with, 12/12 on the front and 4.x/12 over most of the back
so snow loads are always going to be different across those parts
independent of roofing material or melt/slide rate. So I'm not
getting the bit about snow-guards causing any additional problems
that wouldn't already exist. Are you-all trying to tell me that
if my roof has a uniform snow load front and back and the portion
on the front suddenly slides off, the back now has some different
and more hazardous loading condition with the same amount of snow
that was already there? I'm not buyin' that.
This roof survived the "ice dam hell" we all experienced in New
England two winters ago, which was a combination of a lot of snow
over a short period of time with intervening rainfall which gave
us a thick layer of heavy wet crap. Quite a few commercial buildings
didn't make it. I was up in my attic gingerly examining the rafters
and mid-span bracing under the shed dormer, but even though nothing
looked amiss or excessively stressed I went up top and shoveled off
most of that side anyway. All that was way pre-retrofit, of course,
and now if anything the structure up there is quite a bit stronger
than it used to be.
So I'm not too worried about structural capacity here, I just want
most of the snow in the back to not thud down onto my infrastructure
below. The original question is how much of it to leave non-retained,
e.g. how far up the guards should go. I figure if the last foot's
worth of snow falls off directly, that's far less an issue than if
all sixteen feet worth wants to cascade down and bury my heat pump
when the sun comes out. And as I said, I've done the load math,
anticipating worst-case panel-parallel loads of 250 pounds per
attachment on the back shed-dormer side. The reason I'm not
bothering with the front is that the run length is longer, the pitch
is 12/12, and the per-panel longitudinal loading would be quite a
bit more, enough to worry about the ability of the screws at the
top fixity point to hold it. So that stuff can just slide off
into the unoccupied front yard as it sees fit.
For the most part, snow didn't really "avalanche" off either side
over my first winter with this roof. One or two sections may have
come down a little faster than others, but for the most part the
slight lip of the gutter moderated slide rate. I realize that's
always going to be heavily dependent on the type of snow we get,
daytime vs nighttime temps, melting conditions over subsequent days,
etc. But holding the stuff on the back and letting it melt down
into the drainage pits just seems like a better idea than going
out and shoveling compacted accumulations away from the back wall.
_H*