Ice Accretion in Freezing Rain
KATHLEEN F. JONES
data with their proprietary freezing rain, snow,
1. INTRODUCTION
and rime ice accretion model. The Southeast Re-
Freezing rain occurs throughout the United
gional Climate Center (1993) determined the aver-
States. It may cause hazardous conditions for pe-
age number of freezing-rain events per year and
destrians and cars as the rain freezes to sidewalks
the average event duration in Virginia, North and
and streets. The rain may also freeze to structures.
South Carolina, Georgia, Alabama, and Florida
Open structures that are not designed to with-
using archived weather information.
stand the accreted ice load and the increased wind
Hourly weather information from hundreds of
load, due to the greater projected area of the struc-
weather stations across the country has been com-
ture, may fail. Recent severe freezing rain storms
piled at the National Climate Data Center (NCDC)
in the Midwest and the South that caused struc-
and its military counterpart, the Environmental
tural failures initiated by ice and wind are de-
Technical Applications Center (ETAC). Weather
scribed in Storm Data (NOAA 19591995: Mar.
data are available in magnetic tape format, begin-
1990; Oct., Nov. 1991; Feb., Mar. 1994). Both trans-
ning in the 1940s for some stations. These hourly
mission-line design guidelines (IEC 1990, ASCE
measurements of the weather conditions, includ-
1991) and tower standards (EIA/TIA 1991) recom-
ing precipitation rate, air temperature, dew-point
mend designing for accreted ice, but provide little
temperature, wind speed, atmospheric pressure,
information on the ice load that should be used in
and solar radiation, can be used with an ice accre-
design.
tion model to hindcast the amount of ice accreted
There have been a few nationwide and regional
in past freezing-rain storms.
studies to establish the severity of freezing rain in
This report describes two freezing-rain models
this country. The only systematic measurements
I developed to use the historical hourly weather
of accreted ice thicknesses in the United States are
data. The first, described in sections 4 and 5, is a
described by Bennett (1959), who presents data
detailed heat-balance model that determines the
from nine years of measurements in the 1920s and
amount of the impinging rain that freezes either
1930s. Tattelman and Gringorten (1973) used
directly to a structure or as icicles. The second is a
Bennett's data and information from Storm Data
simple flux model that requires only precipitation
and its predecessors to determine ice thicknesses
and wind-speed measurements to determine the
from freezing rain storms in the United States over
amount of accreted ice, assuming that all the im-
a 50-year period. They then calculated extreme ice
pinging water freezes. This simple model, pre-
loads for return periods of up to 100 years in each
sented in section 6, gives physical insight into the
of seven regions of the country. Vilcans and
factors governing the ice load on structural ele-
Burnham (1989) mapped the number of hours
ments with different cross sections. Both models
with freezing rain in the 24 years from 1965 to 1988
are incorporated in the Microsoft FORTRAN77
at 110 airport weather stations in the eastern half of
program ZRAIN.* The weather parameters used
the U.S. using archived weather data. MRI (1977)
in the two models are described in section 2. Ac-
estimated ice loads on transmission lines for 25-,
50-, and 100-year return periods in Washington,
*Available by request from
Oregon, Idaho, and part of Montana using weather
kjones@crrel.usace.army.mil.
Previous Page
