# Weight vs Density of Snow On Your Roof

Boat Hill, where I live since two years back, is a 70s tract-housing estate where roofs are almost flat. Snow thus tends to build up on them. Of course, pile enough snow onto any structure and it will collapse. But I’ve come across a curious notion here. Several neighbours have told me to beware wet snow “because it’s so heavy”.

They’re not talking about snow that becomes secondarily soaked by rain that adds to its weight. They believe that if I have a tonne of powdery snow at -10 Celsius on my roof, I’m OK, but if that tonne approaches 0 Celsius and compacts down into a thinner, less fluffy layer, it will break my roof.

As far as I understand, they’re confusing weight and density. A shovel of powdery snow is lighter than a shovel of slush. But when you shovel slush, you need to shift fewer shovels of the stuff to get your yard (or roof) free of snow.

1. #1 TheBrummell
January 7, 2011

In my experience, people who have never been trained in quantitative things have a shockingly poor grasp of basic physical concepts like weight and density.

I worked with a woman not long ago, an elementary school teacher, who thought ice was heavier than water – not more dense, but that a beaker full of water would gain weight if placed in a freezer. I do not think her attitude is rare.

I can’t say I’m too surprised by your conversations with your neighbours.

2. #2 Martin R
January 7, 2011

Water is so strange, being densest at 4 degrees C. Lucky for us — and for the fish!

3. #3 johan
January 7, 2011

Jag trodde det hängde ihop med luftfuktigheten…

“Blir snön tyngre när det töar?”
“Ja, om det samtidigt är fuktig väderlek, vilket är det vanligaste. Den från början torra snön drar då åt sig luftens fuktighet och blir tyngre.”

http://www.nyteknik.se/popular_teknik/teknikfragan/article18328.ece

4. #4 Sparky Clarkson
January 7, 2011

Alternatively, and presuming that your neighbors decide whether they need to shovel snow off their roofs primarily based on a visual assessment, what they’re saying is sensible. One should be wary of wet snow because, due to its greater density, an amount that looks innocuous may in fact be a danger to structural integrity. Without hearing how it was specifically phrased by your neighbors, it’s difficult to say that this is what was meant.

5. #5 Martin R
January 7, 2011

Johan offers a link suggesting that dry snow sponges up air humidity about 0 centigrade, whoch would add to its weight. But does that effect continue to operate once the snow has become wet through its own melting?

6. #6 kai
January 7, 2011

I think johan’s point is important enough to be translated: at warmer temperatures snow absorbs moisture from the air and becomes heavier with no apparent change in volume.

7. #7 Geoff Carter
January 7, 2011

“Heavy snow” is a term often used to describe snow’s water/volume ratio;

“The water content of snow may range from 3% for
very dry snow to 33% for a wet, heavy snow, to
nearly 100% for ice”
http://www.aragriculture.org/disaster/ice_snow/ice_snow_accumulation.pdf

I have covered this in my blog as it applies to archaeological buildings, with a pitched roof, snow is an important factor in the design of a built environment.
[flat rooves are archaeologically untested, and are probably best avoided for a few centuries].

8. #8 Eric Lund
January 7, 2011

One potential issue is that depending on whether you get direct sunlight and how well insulated your ceiling is, you may have portions of the roof with temperatures above the melting point, from which some of the snow will melt and run off toward other parts of the roof. The risk with shallow roof pitches is that an ice dam will develop which prevents such melt water from running off the roof, instead allowing it to re-freeze right behind or on top of the ice dam, or even worse, to find a path through the roof and into the house. That, rather than snow load, may be the issue for you. I presume Sweden has standards for the amount of pressure (force, in the form of the snow’s weight, divided by the area of the roof) due to snow loading the roof has to withstand, as places in the northern US generally do. In the part of the US (New Hampshire) where I live, preventing ice dams from growing is usually a more urgent issue than preventing roof collapse, the exceptions coming when there has been a particularly heavy snowfall (even then, my roof has withstood a series of snowstorms that dumped over a meter of snow in the space of a week, with 80 cm of that coming in a single storm).

9. #9 Martin R
January 7, 2011

The site that Johan links to describes an experiment where a bucket of snow held at 1 Celsius gained 2% in weight in one day, and then lost 1% the second. On the third day all of it had melted, and the resulting water weighed 103% of the original. Hardly significant changes.

10. #10 Martin R
January 7, 2011

I’m concerned about what my body weight, boot heels and snow shovel might do to the roof paper if I went up there.

11. #11 Brandon
January 7, 2011

I come from Colorado and heard variants on “beware wet snow because it is heavy” expression often. I don’t think the expression is necessarily incorrect: for a given volume, wetter, denser, snow weighs more than dry snow. That was always the meaning I ascribed to it.

12. #12 codero
January 8, 2011

Though your roof won’t mind either way, weight is properly measured in newtons.

13. #13 jozef
January 8, 2011

When it rains on snow the rainwater cannot run off but is collected in the snow. And whatever the measure its heavy.

14. #14 Phinneus
January 8, 2011

The ‘holier than thine neighbor’approach taken by Dr. Martin,the skeptic and apparently ‘knower of all things’omits the common sense approach taken by his ‘cretin’ neighbors. Should you have 6 inches of light powdery snow on the roof of your hovel, and 6 inches of heavy slush on that same roof, the SLUSHY SNOW will weigh more, and be more likey to cave in that yurt of yours!! Smugly, Dr. Martin ignore these local Neanderthals, and watch your roof cave in on you and your arrogant profile.

15. #15 Phinneus
January 8, 2011

The Brummel… Ask yourself why ice floats on water. Tell me then which is heavier…the same VOLUME of ice or of water? Then tell me then what the Earth would be like if water did not expand as it freezes. Perhaps contracted like every other substance when it freezes or cools! Why do you think that happens…expansion rather than contraction? Would it perhaps sink? What would happen then to any and all bodies of water as they froze. Would all bodies of water freeze if those were the conditions..water contracting as it freezes.
Have you ever given that any thought?
Have you been trained in ‘quantitative things’?

16. #16 Aki
January 8, 2011

Let’s measure and calculate the snow load (kg/m2).

http://www.ymparisto.fi/default.asp?node=15820&lan=fi

17. #17 flub a dub
January 8, 2011

dmab; I thought you were going to make a sincere effort to stop dropping acid this new year..oh well…try again!

18. #18 Phinneus
January 8, 2011

Actually Martin, in snowy areas there are roof ‘shovels’ designed to remove roof snow from the ground. They have extensions and will easily do the building you picture.Your local hardware store will have them. Be sure not to stand under that part of the roof when the snow slides off…especially if it is that elusive ‘heavy snow’. Perhaps one of your ‘notsobrite’ neighbors could show you how to use one.

19. #19 Magnus Sjödin
January 8, 2011

I live a few houses down the road from Martin in an identical house. Apparently, according to the insurance people, the greater risk of the snow and the melting of the snow is not a cave in of the roof due to the weight of the snow regardless of water content. The greatest risk, from a general insurance claim statistic point of view, is leaks in the roof allowing the water to enter the roof and damaging the property. Please don’t take that as a guarantee that your roof won’t cave in, but I personally think it is less likely. Having said that I know for a fact that roofs have already caved in over the side entrance passages in some houses. But they are as you know not constructed in the same way as the roof om the rest of your house.

20. #20 eleanora.
January 8, 2011

Martin: “I’m concerned about what my body weight, boot heels and snow shovel might do to the roof paper if I went up there.”
Then don’t wear your stilettos up on the roof.

More seriously, the roof SHOULD have been designed to withstand both the dead and live loads it would normally be subjected to, such as wind, snow, people doing maintenance, etc. Australian building codes have quite a large safety margin factored in, but I have no idea what Swedish codes of the 70’s had.

21. #21 Martin R
January 9, 2011

Magnus, my side passage has no roof. I’ve been shoveling a lot of snow there lately to be able to reach the compost Dalek. (-;

Eleanora, Swedish building codes in the 70s were pretty strict.

22. #22 Tor
January 10, 2011

Has anyone here ever heard of the roof of a private residential house — as opposed to, say, a giant mall or sports arena — caving in under too much snow?

23. #23 TheBrummell
January 10, 2011

Phinneus, check your reading comprehension. I didn’t say what you seem to think I said. Neither did Dr. Rundkvist. At what point did either of us discuss solid ice floating on liquid water? How is that property of H2O relevant to the current discussion?

The water-absorbing properties of snow are something new to me. I suppose I shouldn’t be surprised, but I’m curious about the temperature effect – does snow absorb atmospheric water more at higher temperatures because of some structural difference in the snow (shape / size of flakes and pores, for example), or because warmer air carries more water, all else being equal?

Tor, I seem to remember cases of residential buildings (houses and apartment buildings) having damaged roofs (not necessarily caving in or totally collapsing) in Victoria, B.C., Canada, during a very severe storm near New Year’s day 1996/1997. Very large amounts (more than 30 cm) of very-high-water-content snow were deposited in something like 24 hours in a city that in a normal year receives very little snow (I think the long-term average is something less than 10 cm / year).

24. #24 Nate
January 21, 2011

Hey, Im a new homeowner in MN and I have a house that was built in the 70’s and looks very similar to the one pictured above. I havent gotten a chance to remove snow from the roof and we have had alot of snow here. There is deffinately an ice damn up there and I would say about 3-5 inches of snow on the roof. Im getting pretty concerned about all the ice in the gutter and on the roof and im hearing what sounds like snow crunching comming from the roof. Its especcialy cold tonight (below 10 degrees f) and I dont really want my roof to fall on my head. What do you suppose the chances of that actually happening are? I am trying to work with some folks to help me out but Im afraid time is running short. Any advice would be greatly appreciated. Thanks.

25. #25 Martin R
January 21, 2011

I think you can relax. An ice dam (if you do have one) is only a concern when the temperature is above the freezing point of water. And 4 inches of snow is not much. The noises you hear are probably from the snow doing things on its own, not bits of your roof cracking.

If you’re concerned, get up there tomorrow, tie yourself to the chimney with a rope and shovel the snow off the roof.

26. #26 Bill
January 28, 2011

A snow load of 2′ 6″ works out to be the equivalent of 17 lbs. 8 ozs. per square foot on your roof. Even old stick trusses are designed to carry a load on your roof of about 40 lbs. per square foot. Current day engineered trusses are 80 to 90 lbs. per square foot, there is plenty of room left for snow load.

On pitched roofs, the load is actually carried by the walls, not the trusses.

You can calculate the water content of the snow wherever you want on your house or any building. Pick a snowy spot. Use a measuring tape and measure the snow depth. Mark a foot square area. Shovel the full depth of snow/ice into one or more containers. Melt the snow and weigh the water, or measure the volume.

An imperial cup of water weighs 10 ounces or 10/16 ths. of a pound. Make sure you use the full depth of the snow since there are a number of layers, each having different water content.

If you want to shovel, do so equally, if you don’t, you can
change the load factor on the roof; potentially big problems.

Also, when you are shovelling your are walking around on frozen shingles. You should leave 4″ to 5″ of snow on the roof (do not take it down to the wood). Moreover, you’re putting pressure on those frozen shingles, they are very brittle and likely causing damage which will likely show-up months or years later.

Here are some basic calculations for those who might be interested:

1″ of liquid water weighs 5 lbs. 3 ozs. per square foot. A rough calculation is that 1″ of rain precipitation is 1/10th of snow precipitation or 10″ of snow. That is the same as a water content of 10%. A very dry snow, the kind that falls on very cold days, can have a water content of less than half of that. Wet, heavy snow, the kind that falls at around the freezing point, can have a water content of about 30%.

Add rain to dry snow, two things happen, there is more water content since the rain is typically absorbed by the snow. The snow also starts to collapse and the depth is reduced. An inch of precipitation within a 24 hour period is unusual at any time of year. So that would add 5 lbs. 3 ozs. of load. The other thing that happens is that there is a heat loss from all roofs.

Those that have well ventilated attic spaces, which is rare even when standards are followed, will have less heat loss at the roofing surface. Since snow starts to melt with heat, a deep snow on the roof will eventually reduce in depth due to heat loss as well as sublimation, direct change from snow to water vapour. A similar thing occurs with Sunlight on the exposed layers.

Look at a few properties with houses and free standing sheds. Since the sheds are likely not heated, the snow depth on the sheds is typically quite a bit deeper than that on the house. Everything being equal though, the two buildings should have similar weights of snow on them. The house may
have less since it is higher and thus, on windy days, more snow tends to be blown off than on a lower building.

Let’s for good measure, assume that the snow on the roof of an average house at the moment is 12″ deep and that there is a layer immediately adjacent to the roof surface which has a high water content of 50% and is for good measure, 1″ thick. That gives it as weight of say 3 lbs. Then let’s give the rest of the 12 inches at 20% water content an average weight of 12 lbs. per square foot for good measure.

That gives us a snow load of 15 lbs. per square foot. Older trusses were designed for 40 lbs. per square foot.

27. #27 j
April 2, 2012

MY GARAGE TOTALLY COLLAPSED LAST NIGHT DUE TO SNOW ON ITS ROOF. IT HAS NOT SNOWED FOR OVER TWO WEEKS. NO RAIN HAS FALLEN. IT HAS BEEN THAWING MOST DAYS AND FREEZING AT NIGHT FOR THE LAST WEEK OR SO. SO IF I UNDERSTAND IT DID NOT GET HEAVIER FROM THE DAILY MELTING, WHY DID IT COLLAPSE TWO WEEKS AFTER THE LAST SNOW WEIGHT WAS ADDED? i AM GUESSING THE STRESS OVER TIME DID IT.

SHOULD i BE FRIGHTENED TO SLEEP IN THE HOUSE TONIGHT?

i AM GETTING PARANOID BUT IT IS NEARLY MIDNIGHT. HOPE I WILL STILL BE HERE IN THE MORNING.

28. #28 Martin R
April 2, 2012

Maybe you should get up with a rope and a shovel, tie yourself to the chimney and shovel some snow off your roof to ease your mind?