Civil Engineering Reference
In-Depth Information
It is unclear whether the
C
t
factor of 1.3 for such structures provides adequate
coverage for Scenario 3.
The rain-on-snow surcharge associated with rain percolating through
a roof snowpack is composed of two portions. The fi rst, typically small,
portion is the load caused by vertical percolation through the upper unsatu-
rated layer. The second, larger portion is the load caused by rain fl owing
downslope in the saturated layer immediately above the roof.
As expected, the surcharge caused by fl ow in the saturated layer increases
with increases in rainfall intensity, rainstorm duration, and eave-to-ridge dis-
tance, and it decreases as roof slope increases. All other things being equal,
a low-sloped roof with a large eave-to-ridge distance will have the greatest
rain-on-snow surcharge.
Colbeck (1977) developed surcharge-load relations for a given rain-
fall's intensity and duration. Unfortunately it is not possible to determine an
exact 50-year roof surcharge for rain on snow due to the lack of appropriate
weather information. For example, O'Rourke and Downey (2001) present
three estimates of the surcharge for rain percolating through the roof snow-
pack (Scenario 2 or 4). Their results for a ¼-on-12 roof slope (1.19°) and a
100 ft eave-to-ridge distance are presented in
Table G10-2.
In the fi rst estimate, National Oceanic and Atmospheric Administra-
tion (NOAA) ground snow and precipitation data were used to determine the
wintertime maximum daily rain for a 50-year return period. The resulting roof
surcharge is only an estimate since associated rainstorm duration and intensity
had to be assumed. The results in
Table G10-2
are based on an assumed one-
hour duration (i.e., the maximum daily rain was assumed to have fallen over
a 1 h period). For the 19 cities considered by O'Rourke and Downey, the roof
surcharge loads ranged from 0 to 5.68 lb/ft
2
with an average of 2.02 lb/ft
2
.
In the second estimate, rainfall intensity data were used to determine the
two-year MRI rainstorm with a duration of one hour. The resulting roof sur-
charge is an upper-bound estimate since the two-year rain may have occurred
when the roof was free of snow. For the 19 cities considered, the roof sur-
charge loads ranged from 1.03 to 7.61 lb/ft
2
with an average of 4.73 lb/ft
2
.
In the third estimate, the roof surcharge was determined for the 1996-
1997 holiday storm rain-on-snow events in Yakima and Seattle. Fairly detailed
weather information for Yakima Airport allowed O'Rourke and Downey to
quantify the magnitude of the rain-on-snow event. On December 31, 1996,
Yakima had a ground snow load of about 23 lb/ft
2
and a rain event with an
Table G10-2
Estimated Roof Rain-on-Snow Surcharge for ¼-on-12 Roof Slope and
100 ft Eave-to-Ridge Distance
Range
(lb/ft
2
)
Average
(lb/ft
2
)
Method
50-yr (wintertime) daily rain in 19 cities
0 to 5.68
2.02
2-yr (year-round) rain in 19 cities
1.03 to 7.61
4.73
1996-97 winter storm: Seattle and Yakima
—
4.60
Source: Adapted from O'Rourke and Downey (2001).
