snow collapse ashrae fronapfel sbsa

Copyright 2015 SBSA 1

STRUCTURAL ENGINEERING www.callSBSA.com

The March 2003 Snow Event and Resulting Damages

Edward L. Fronapfel,

PE, MSCE, CFCC, CBIE, CBCPForensic Engineer

Copyright SBSA 2015

National Weather Service (Source: National Weather Service)

Copyright SBSA 2015

National Weather Service (Source: National Weather Service)

Copyright SBSA 2015

Is it true that there is one inch of water in every ten inches of snow that falls?

The water content of snow is more variable than most people realize.

While many snows that fall at temperatures close to 32oF and snows accompanied by strong winds do contain approximately one inch of water per ten inches of snowfall, the ratio is not generally accurate.

Ten inches of fresh snow can contain as little as 0.10 inches of water up to 4 inches depending on crystal structure, wind speed, temperature, and other factors.

The majority of U.S. snows fall with a water-to-snow ratio of between 0.04 and 0.10.

1971 Snow Load Design Data: National Weather Service with 70 years of record, Converted to Weight of Snow w=0.9x^1.2 “The equation was designed to be slightly on the heavy or conservative side of the average.”Summary of Snow Survey Measurement for Colorado and New Mexico, published in 1967 by CSUMeasurement of Snow Depth and Water Content

After maximum weight was determined multiplied by 0.8 to adjust from ground load to roof loadSecond factor was to adjust the recorded maximum to what would be expected maximum in a 30 year reoccurrence.

Equation created to adjust for elevation (20psf) minimumIce Loads of 100 plf should be used at the edges of sloped roofs in conjuction with the full snow load on on the overhang and no snow load on the interior span.

Copyright SBSA 2015


Based on the National Weather Service Data:

The weight per inch of snow was found to be around 2 pounds per inch.

Reviewing the depth records: Arvada had 50 inches at elevation 6,000 feet

S = 32/(K+12)*(A – K/2)^2

K = 5 or 6A = 6

S = 16 psf to 23 psf

Per Map Depth = 50Snow Load = 50 * 2 = 100psf or over 4 times the 30 year design.

If snow has 1 inch per 10 inches Snow Load = 5 * 5.2 = 26 psf

Using equation for weight W = 0.9*x^1.20.9*(50)^1.2 = 98 psf

Copyright SBSA 2015

Zurich Insurance

Copyright SBSA 2015

Does snow always get fluffier as temperatures get colder?

A Number of Studies in the Rocky Mountains have shown that the fluffiest, lowest density (0.01 -0.05) snows typically fall with light winds and temperatures near 15°F. At colder temperatures, the crystal structure and size change. At very cold temperatures (near and below 0°F) crystals tend to be smaller so that they pack more closely together as they accumulate producing snow that may have a density (water-to-snow ratio) of 0.10 or more.

Copyright SBSA 2015

RECORD EVENT STATEMENTNATIONAL WEATHER SERVICE DENVER COLORADO 157 PM MST FRI MAR 21 2003

SINCE OUR LATEST WINTER STORM...MARCH 17-20TH 2003...QUESTIONS HAVESURFACED ABOUT MOST SNOWIEST MARCH NOT TO BE CONFUSED WITH DENVER'SMOST SNOWIEST MONTH. THESE TOTALS ARE BASED ON ACTUAL SNOWFALL AMOUNTS

DENVER SNOWIEST DENVER SNOWIEST MARCH (INCHES) MONTH (INCHES)-------------- ---------------------33.2 2003 *(SO FAR THRU 3/21) 57.4 DEC 1913 32.5 1944 42.6 NOV 1946 30.5 1983 33.8 APR 1933 29.2 1961 33.2 MAR 2003 * 26.8 1959 32.5 MAR 1944

32.0 APR 1885 31.3 MAR 1991 31.2 OCT 1969 30.8 DEC 1973 30.5 MAR 1983

Copyright SBSA 2015


Copyright SBSA 2015

The accumulation of 30 to 40 inches of snow loads between 26 pounds per square foot and 35 pounds per square foot, based on approximately 1” of water per every 6” of snow. Roof structures are required to support a minimum snow design load based on building department requirements, including drift load, valley load and other considerations.

Also see ASCE 7-02 Section 7.0

Copyright SBSA 2015

“We found the water density of snow initially following the snowstorm to be consistent. If 16 inches (406 mm) of snow weighed 20 pounds per square foot (98 kg/m²), areas with 24 inches (610 mm) of snow weighed 30 pounds per square foot (146 kg/m²) and areas of snow with 32 inches (813 mm) of snow weighed 40 pounds per square foot (195 kg/m²). If the moisture density is consistent, the measurements will verify conditions. It was our experience that during melting periods after the storm, water density within the snow varied greatly. Near drainage areas, snow and water density increased significantly (as much as two times). Snow density may vary, but water density (the weight of water) is consistent.

An NRCA contractor member learns some business lessons from Mother Nature by Conrad A. Kawulok

Copyright SBSA 2015

• CS = 50 year mean recurrence interval or a 2 percent chance of occurrence (annual probability)

• Flat Roof

• Sloped Roof

• Partial Load

• Unbalanced Load

Copyright SBSA 2015


• Drifts formed in the Wind Shadow of same or adjacent structure

• Projections, parapet walls, mechanical walls

• Sliding Snow

• Rain on Snow surcharge (20 psf or less)

• Ponding Instability

• Existing roofs (Owners shall be advised of the potential for increased snow if building constructed within 20 feet)

Copyright SBSA 2015

Snow Density by Equation

• Weight = 0.14 pg + 14 but no more than 30 pcf (at Pg = 30 pcf, weight = 18.2 pounds)

• Some rough math

– 1 inch of water in 6 inches of snow = 10 pcf

– 4 inch of water in 6 inches of snow = 40 pcf

• 1 inch of water is 5.2 pounds

Copyright SBSA 2015

Blizzard Snow Loads And Building Codes, Contact: Blaine [email protected] University News Service

ITHACA, N.Y. -- For building engineers and climatologists, the memory of the Blizzard of '96 refuses to melt away. Well over a year since the heaviest snow of the century fell over the Northeast, climatological studies now show that had it notbeen for structures built "better-than-code," more roofs could have collapsed under the weight, researchers say. "In places where roofs should have collapsed, they didn't collapse," said Arthur T. DeGaetano, a climatologist with the Northeast Regional Climate Center at Cornell University. "It showed that most buildings exceeded the minimum building code requirements." The Blizzard of '96 (Jan. 8-10, 1996) was a whopper, but with snow resting on rooftops, another smaller storm followed and contributed to even greater snow weight on roofs. "It snowed, stayed cold, and we got more snow," DeGaetano said. "It was theaccumulation of snow from the two storms that posed the snow weight problem." Enough snow fell throughout the 1995-96 season, that Boston, New York City, Philadelphia, Baltimore, and the greater Washington area smashed their snowfall records. But, during the Blizzard of '96, news reports of roof failures throughout the Northeast corridor -- Boston to Washington --prompted DeGaetano; Thomas W. Schmidlin, associate professor of geography, Kent State University; and Daniel S. Wilks, Cornell associate professor of meteorology, to author a peer-reviewed paper, "Evaluation of East Coast Snow Loads Following the January 1996 Storms," in the Journal of Performance of Constructed Facilities (May 1997), a publication of the American Society of Civil Engineers, New York City. The study was funded by the National Oceanic and Atmospheric Administration. DeGaetano and the other scientists examined the weight of snow during the period Jan. 6-15, 1996. Once the snow weights were determined based on water content within the snow, they corroborated the climatological information with minimum building codes for given areas. For example, some of the hardest hit areas were northern and central New Jersey. For northern New Jersey, building-code minimums generally require roofs to withstand snow weights of 21 pounds per square foot -- the equivalent of a storm that occurs once every 50 years. More infrequent storms bring snow weights of 26 pounds-per-square-foot in 100-year events, and 31 pounds-per-square-foot for 200-year values. The Blizzard of '96 and the subsequent storm amounted to the equivalent snow weight of a 125-year snowstorm.

Copyright SBSA 2015

Coastal Maine has different code requirements, as their 50-year snow weight level is 62 pounds-per-square-foot, their 100-year snow-weight level is 72 pounds-per-square-foot, and their 200-year mark is 83 pounds-per-square-foot. Snow weights here did not exceed the code.

In areas south of Boston, the researchers found that snow weights exceeded the minimum building requirements.

The Philadelphia building inspector's office reported 55 collapses, mainly from porches and rowhouse roofs. In Camden, N.J., there were 16 reported porch and rowhouse roof collapses. These types of structures, however, are often not covered by building code requirements, or pre-date the establishment of these standards.

Staggered news reports during the Blizzard of '96 and the subsequent snow storms provided the scientists with anecdotal information as well, lending insight to the type of structures and geographic locations to pay attention to. For example, a barn collapsed in Woodsboro, Md., killing 150 cows inside; a supermarket roof collapsed in Massapequa, N.Y., injuring 10 people; a 2,500-square-foot section of shopping mall collapsed in Tewksbury, Mass.; a building that housed a printing company in Bethesda, Md., completely collapsed; and a lawn and garden shop in Berks County, Pa., collapsed under the snow weight, killing one person.

"Climatology affects a lot of different fields, someone has to show how much snow can fall and what an engineered-structure can handle," DeGaetano said. "You have to ask what's the maximum snow weight that a roof can hold, what's the frost depth for underground utility lines, what's the max winds in an area so that roofs can withstand the winds, what's the maximum rainfall for an area, so that sewers can handle rainfall. That's what climatology is all about."

Copyright SBSA 2015


University of Minnesota Extension Service

Snow Loads on RoofsMoist Minnesota winters create frustrating dilemmas for homeowners. Many and prolonged snowfalls create concern about the roof's ability to support tons of ice and snow for long periods of time. What should you do? If you have a newer structurally sound house, the roof can probably support two feet of snow or one inch of ice, plus one foot of snow, over an extended time period. Should you leave the ice and snow and risk roof collapse, or should you remove the snow and risk damage to shingles and yourself? Because it is dangerous, it is best to hire professionals to remove snow. There is no simple solution to this dilemma. When listening to weather forecasts, pay attention to the type and amount of snow that has recently fallen. Depth and moisture content of the snow in your locale dictate how concerned you need to be. But you still need to have some idea about how much weight your roof is capable of supporting, how much dead weight is already on the roof (layers of shingles) and how heavy the ice/snow accumulation is on your roof. Local building and housing inspectors, structural engineers, carpenters, architects and lumber managers may be a source of help in determining the existing roof conditions and potential support it will give. You need to know, however, how many layers of shingles are now on the roof. Most building codes restrict the layers to two. Most homes in Minnesota carry 10 pounds of roofing materials per square foot per one layer of shingles. The more layers already on the roof, the less weight of snow and ice the roof will be able to support. You also need to measure the weight of ice and snow. It is not the depth of snow but the moisture content that is the critical factor. Very dry snow contains 3% water and very heavy wet snow is about 20% water content. For more information on measuring snow weight and safe snow removal, contact your local Extension office and ask for publication, FO-6891-GO, Roof Snow Loads.

Copyright SBSA 2015

Some Facts:

Some roofs have endured many winters of heavy snow did not fail but that does not mean the roofs can last indefinitely.

Roofs of most older homes were built with little or no insulation, so snow melted fairly quickly and roofs carried snow loads for very brief periods.

If insulation has been added to the attic, less heat escapes, so snow and ice do not melt as rapidly and snow loads accumulate to new and perhaps excessive weight levels. This reduces ice damn loading, but increases weight of the captured snow.

Furthermore, the roof deck may not have been properly ventilated when insulation was added or due to the indadequate truss size (heel). In that case, ice-damming and subsequent backup of snowmelt under shingles, as well as excessive condensation due to moisture transfer from the interior of your home into the attic and rafter cavities may occur. This creates favorable conditions for deterioration of roof structural members and a decrease in fastener holding ability.

Copyright SBSA 2015

Roof structural failure often begins with rafter deflection or rafter spread. Deflection results when horizontal loads (such as snow and ice buildup) cause wood fibers to bend.

Eventually, deflection causes rafters to rupture in the center third of the span or at such weak points, as notches or knots on the top or bottom edge of the rafter. Rafter spread results from the failure of mechanical ties, such as nails, to hold ceiling joists, top plates, and studs together or, occasionally, failure in the ceiling joist itself.

Rafter deflection and spread represent initial stages of structural failure. These problems should be corrected or stabilized with the assistance of a knowledgeable contractor, engineer, or architect.

Copyright SBSA 2015

Colorado doesn’t generally experience large losses from winter-related storms:1. $93.3 million in insured damage resulting from heavy snow and ice on March 18-19, 2003. 2. $10.5 million in insured damage resulting from heavy snow and ice on Oct. 24-26, 1997. 3. $6.4 million in insured damage occurred from Sept. 20, 1995.4. $4.9 million in insured damage occurred on Dec. 22-26, 1982.Colorado’s most costly catastrophe in state history was a $625 million hailstorm that battered the Denver-metro area on July 11, 1990.

Rocky Mountain Insurance Information

Claims are still expected to come in as people observe damage and the settlement process continues, but the pace has slowed significantly from the days following the storm. The insured damages are based on the approximately 28,257 claims filed so far.More than 90% of the storm damage is based on homeowners insurance claims. Less than 10% of the insured damage is due to auto insurance claims, with the bulk of that from vehicles being crushed rather than weather-related accidents.

"Results for the first quarter of 2003 were negatively affected by a severe snow storm in Colorado which effected office performance for most of the week of March 17, including the closing of a significant number of offices for three days. We estimate the company lost about $300,000 in

group practice revenue during that week which would have generated approximately an incremental $90,000 in net income for total net income of $367,000. Without the storm, we believe earnings per share for the quarter ended March 31 would have been approximately 24 cents or a 60

percent increase from the same quarter in the prior year," said Fred Birner, CEO, in a statement.

Denver Business Journal

Copyright SBSA 2015


Copyright SBSA 2015 Copyright SBSA 2015




Snow GalleryParadise Inn, Mount Rainier, Washington. During the winter of 1916-1917, 789.5 inches of snow fell at Paradise Inn. When this photo was taken, in March 1917, the snow was 27 feet deep. (Source: NOAA/Department of Commerce. Courtesy of the Historic National Weather Service Collection. Photograph originally published in "Monthly Weather Review," July 1918, p. 330.)



NOAA Photo Library




Copyright SBSA 2015

Due to roof collapse, BOB STEMPER’S SUPER BUFFET

will re-open in July 2003.

Copyright SBSA 2015

snow collapse ashrae fronapfel sbsa

Engineering