Snow load on existing buildings 2

[donegalcol]

Section 3403.3.1 of IBC 2009 states that if the live load is not changed then the presumed live load at the time the building was constructed may be used in lieu of the new live load per IBC. Does anyone know if this applies to snow loads on roofs.
I have a condition whereby I am adding solar modules on a roof. The roof is adequate is I use the smaller snow load that existed when the building was constructed, however the roof is overstressed if I use the new snow load per IBC 2009.

 

[JAE]

Snow load is not a live load.

 

[msquared48]

Better bite the bullet, mind the new snow load, and reinforce the roof, particularly considering that the panels will cause more snow to collect above the panels.

DO NOT try to be Mr. Nice Guy here and save the client money at the expense of safety. Life is too short and there are way too many humgry lawyers. I have repaired too many older buildings where the roof collapsed due to snow overload.

Mike McCann
MMC Engineering

 

[SkiisAndBikes]

It is very likely that the addition of solar panels on a roof structure will result in increased snow loading due to drifting Combined with the increase in code mandated snow loads, I would be looking at recommending the roof be reinforced. There is a lack of published information on snow drift loading around rooftop PV installations.

 

[paddingtongreen]

I go by, "If you change it, you own it."

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[dcarr82775]

I use Paddington's viewpoint all the time. Strengthen the roof.

 

[kipfoot]

I think you're seeing a consensus. In renovation projects in the northeast we are always adding insulation to the building. This is typically insulation that has never been in the building before and will result in snow sitting on the roof that previously would have melted off. So, I think the chance that the roof will actually realize a greater snow load is real.

 

[MiketheEngineer]

I always thought snow was a live load since I can increase my wood capacity by 15% when that happens??

But I have noticed people seem to call it snow only and seem to have gotten away from calling it a "live" load??

Me - old school!!?? Maybe I should go back??

 

[paddingtongreen]

Mike, it is a live load, it just needs to be separated because it is defined differently from other live loads.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[OHEcoEng]

I wouldn't risk it, not with my license on the line...especially if it's that close. Modules themselves (without racking, wiring, etc) increase dead loading by 2.5 psf. I doubt this is less than a 5% increase.

Skiis, regarding drifting, I have given this much thought and run through many different scenarios with snow. The calculated drift height exceeds the height of the modules in almost all cases, so it typically doesn't present any issues. The solar array "disappears" under the snow at its max...at least up here in Ohio with our 20-25 pg.

 

[MiketheEngineer]

Tend to agree with OHE ---

Usually the increase is in the 1 - 2 psf range - not really significant. BUT check it out and CHARGE the client accordingly!!

 

[woodman88]

Per the UBC (at least back to my 1988 edition) and IBC's snow loads have never been considered a live load. Live loads have always been "Those loads produced by use and occupancy of the building..." in both the UBC and IBC. The duration of wood increases for wood have always been based on the time duration of the load. The 15% is for a two month duration 1) per the UBC "...as for snow" or 2) Per the IBC/NDS "Typical Design Load - Snow".

Garth Dreger PE - AZ Phoenix area
As EOR's we should take the responsibility to design our structures to support the components we allow in our design per that industry standards.

 

[dhengr]

I’m with MiketheEngineer and Paddington, the snow load is a live load, but it must be treaded differently because we know it might not be a uniform live load. But, it seems to me that the question now becomes, can’t the volume under and around the solar modules be filled with snow, a fairly uniform load on the whole roof system; and now the drift depth starts at or near the upper surface of the panels, not at the roof level. And, any snow loading on the panels is now carried by the panels and racking as concentrated loads to a few roof members which will most likely be overloaded. The 5% change really has little meaning here, since it is akin to saying I only changed the load on the roof by 5%, but I put it all at the center span of this one rafter or joist. The problem with parapets, mech. equip. on roofs, screen walls and the like, is they all tend to cause snow to swirl and drop and drift which is part of what I think OHEcoEng is getting at. My only question of him is at what level does he start his drift depth, the actual roof or someplace up on the panels?

 

[JAE]

I disagree completely with snow = live.

Really folks. They are two completely different things.

Yes snow has a short duration but that does not make it a live load.
In the NDS for wood - snow has a different duration factor than live load.

Snow is specifically defined in codes as SNOW and live load is LIVE. They are treated differently in load combinations.

Per the IBC at least (not sure of the British or other codes) you cannot lump snow loads into provisions intended for live loads (which was the question of the original post above).

I agree with Garth above - live loads are loads on floors or roofs that are due to occupational loading (people, furniture, moveable equipment, books, etc.) Snow load is from that nasty white stuff.

 

[OHEcoEng]

dhengr..To clarify on drifting...the drift begins at the top of the roof snow load depth-The drifting load being in addition to the base roof snow load. With top of module at ~1.5 ft and 1-2 ft between rows, the snow fills in there pretty uniformly. I can't remember what depth a 20 psf pg comes out to, but it's well over the top of the module when that happens. Keep in mind that my example deals with individual modules on individual racks in landscape. There are rooftop arrays that reach a height of well over 10 ft with multiple modules "stacked" on the racking. This is not the type of array I am talking about.

As far as the 5%, I agree that the load needs to be looked at locally, not saying that a 100 sf array dead load can be looked at as uniform on a 20,000 sf roof (though I have seen this done pretty commonly).

 

[mtu1972]

There was a time when Snow was considered Live load.

Yes timber treated SL and LL differently, but when doing a steel beam or joist framed roof, we (upper midwest - MI, MN, & WI) could logically decide that if SL > Roof LL, it governed. And yes I am so old that drifting wasn't even considered when I first started.

Where we had Mech and Elec utilities hanging from roof steel, we used DL + LL + SL for those elements. Which was more conservative than how it would be combined now days.

And now we have so many combinations DL + LL, DL + 0.75(LLr + SL), DL + SL, not to mention all the wind direction/combinations, plus diagonal wind loads with torsional components (gag me) that one can't logically tell what will eventually govern.

Is it any wonder that we have developed so many black-box engineers?

gjc

 

[JAE]

mtu1972, I'd be interested to learn what building codes equated live load with snow load. My first code used in business (1973 UBC didn't.

Maybe engineers (in their heads) equated the two - but they aren't the same thing and have extremely different effects on the structure.

 

[msquared48]

Have to agree with JAE here guys (sorry)...

I have the older codes too, and there was never a "live load" reduction applied to snow since it is basically uniform in nature, outside of drifting, unlike a true "live load" that will be spotty in nature. So, it ofalls under different rules than a true live load.

Mike McCann
MMC Engineering

 

[mtu1972]

I don't have any old building Codes at my disposal - but I have all of my old textbooks. Started work in Bay City, MI and then moved to Hibbing, MN amd now WI.

From a 1970 textbook -

"The snow load is an essential element of the live load on roofs in many geographical areas. 20 psf is frequently taken as a minimum, with higher allowances being necessary in localities. A map showng snow load allowances . . ."

From a 1969 textbook; 2nd paragraph in the Live Load section -

"In the colder states, snow and ice loads are often of importance. For roof designs, snow loads of from 10 to 40 psf are used; the magnitude depends promarily on the slope of the roof and to a lesser degree on the character of the roof surface."

So at the least let's agree it is a "subset" of Live Loads.


I never said anything about reducing SL like occupancy LL's. That's why I don't agree with the DL + 0.75(LL + SL): it's a reduction of snow load just because of other live loads in the building. Maybe statistically they shouldn't all be maximum at the same instance - but I've seen roofs with lots of snow on them, for very long times. The roof structure doesn't much care what's happening in the 2nd floor auditorium two levels below it. The roof itself is covered with DL + SL. But this is a reduction of the column loads when in combination with other Live Loads. We are getting as statistically challenged as bridge designers.

My entire life has been in snow country and I witnessed a roof collapse while still in college. When the Timber Design professor was discussing a case he was an expert witness on, I asked if he meant the beer distribution warehouse in my home town. He asked me to stay after class - as he was interested in how the collapse played out. He was never told there were witnesses - even though my brother was the one who broke the news to the building's owner, minutes after the collapse. This was a 1960's era pole barn put into a town of 700 that probably never required plan submittals. He shared that it was a case of a southern WI pole barn company that was totally oblivious to lake effect snow in upper MI.


In the words of Rodney King - "can't we all just get along?"







gjc

 

[JAE]

mtu1972 - OK - a subset of LL in the terminology of some engineers.

But in the view of the original post, snow load should not be considered a live load with regards to code adjustments or handling of live loads in my view. Your textbooks may lump them together in terminology, but building codes really don't.

There are load combinations where full snow is considered so reducing snow for the final design of roof elements shouldn't occur.

Yes collapses occur - but usually due to higher snow loads than mandated by the code, drifting occurring on roof where drifting wasn't considered, or clogged drains creating ponding under the snow on the roof.