1 

CHAPTER C10 ICE LOADS - ATMOSPHERIC ICING

2 

3  C10.1 GENERAL

4  In most of the contiguous United States, freezing rain is considered the cause of the most severe

at
5  ice loads. Maps of ice thickness from freezing rain are in the standard. Values for ice thicknesses

e LY
6  caused by in-cloud icing and snow Maps of ice thickness from freezing rain are in the Provisions.

rm
7  In thisthe 2022 revision, ice load maps for in-cloud icing are included in Section C10.9. Those

lin N
8  maps may be used as part of a site-specific study. Values for snow loads suitable for inclusion in

Fo
er n O
9  this standard are not currently available.

10  Very few sources of direct information or observations of naturally occurring ice accretions (of
any type) are available. Bennett (1959) presents the geographical distribution of the occurrence of
nd o
11 
U ati
12  ice on utility wires from data compiled by various railroad, electric power, and telephone
13  associations, in the nine-year period from the winter of 1928–1929 to the winter of 1936–1937.
t / rm

14  The data include measurements of all forms of ice accretion on wires, including glaze ice, rime
15  ice, and accreted snow, but do not differentiate among them. Ice thicknesses were measured on
ou fo

16  wires of various diameters, heights above ground, and exposures. No standardized technique was
17  used in measuring the thickness. The maximum ice thickness observed during the nine-year period
e In

18  in each of 975 squares, 60 mi (97 km) on a side, in a grid covering the contiguous United States is
reported. In every state except Florida, thickness measurements of accretions with unknown
r

19 
Fo

20  densities of approximately one radial inch were reported. Information on the geographical
21  distribution of the number of storms in this nine-year period, with ice accretions greater than
22  specified thicknesses, is also included.
rik

23  Tattelman and Gringorten (1973) reviewed ice load data, storm descriptions, and damage estimates
24  in several meteorological publications to estimate the maximum ice thicknesses with a 50-year
St

25  mean recurrence interval (MRI) in each of seven regions in the United States. Storm Data (NOAA
26  1959–Present) is a monthly publication that describes damage from storms of all sorts throughout
27  the United States. The compilation of this qualitative information on storms causing damaging ice
28  accretions in a particular region can be used to estimate the severity of ice and wind-on-ice loads.

1 
 
 1  The Electric Power Research Institute has compiled a database of icing events from the reports in
2  Storm Data (Shan and Marr 1996). Damage) and prepared damage severity maps were also
3  prepared.

4  Bernstein and Brown (1997) and Robbins and Cortinas (1996) provide information on freezing
5  rain climatology for the 48 contiguous states based on recent meteorological data. Konrad (1998)
6  discusses spatial patterns of freezing rain in the Appalachians.

at
e LY
7  C10.1.1 Site-Specific Studies.

rm
8  Freezing rain. In determining equivalent radial ice thicknesses from freezing rain from historical

lin N
Fo
9  weather data, the quality, completeness, and accuracy of the data should be considered along with

er n O
10  the robustness of the ice accretion algorithm. Weather stations may be closed by ice storms because
11  of power outages, anemometers may be iced over, and hourly precipitation data recorded only after
12  the storm when the ice in the rain gauge melts. These problems are likely to be more severe at
nd o
13  automatic weather stations where observers are not available to estimate the weather parameters
U ati
14  during power outages or correct erroneous readings. Note also that (1) air temperatures are
t / rm

15  recorded only to the nearest 1 °F (0.6 oC), at best, and may vary significantly from the recorded
16  value in the region around the weather station; (2) the wind speed during freezing rain has a
ou fo

17  significant effect on the accreted ice load on objects oriented perpendicular to the wind direction;
18  (3) wind speed and direction vary with terrain and exposure; (4) enhanced precipitation may occur
e In

19  on the windward side of mountainous terrain; and (5) ice may remain on the structure for days or
20  weeks after freezing rain ends, subjecting the iced structure to wind speeds that may be
r
Fo

21  significantly higher than those that accompanied the freezing rain. These factors should be
22  considered both in estimating the accreted ice thickness at a weather station in past freezing rain
23  events and in extrapolating those thicknesses to a specific site.
rik

24  Bernstein and Brown (1997) and Robbins and Cortinas (1996) provide information on freezing
25  rain climatology for the 48 contiguous states, based on recent meteorological data. Konrad (1998)
St

26  discusses spatial patterns of freezing rain in the Appalachians.

27  In-cloud icing. In-cloud icing may cause significant loadings on ice-sensitive structures in
28  mountainous regions and for very tall structures in other areas wherever supercooled clouds/fog

2 
 
 1  exist, such as in mountainous regions, where supercooled clouds often occur with elevated winds,
2  and in basins that trap freezing fog for extended periods of time under light wind conditions.
3  Mulherin (1996) reports that of 120 communications tower failures in the United States caused by
4  atmospheric icing, 38 were caused by in-cloud icing, and in-cloud icing combined with freezing
5  rain caused an additional 26 failures. In-cloud ice accretion is very sensitive to the degree of
6  exposure to moisture-laden clouds, which is related to terrain, elevation, and wind direction, and
7  velocity. Large differences in accretion sizeice load can occur over a few hundred feet and can

at
e LY
8  cause severe load unbalances in overhead wire systems. Advice from a meteorologist familiar with

rm
9  the area is particularly valuable in these circumstances. Significant loads may also accrete on very
tall structures anywhere in the country when supercooled clouds are advected past some portion

lin N
10 

Fo
11  of the structure.

er n O
12  Mulherin (1996) reports that of 120 communications tower failures in the United States caused by
13  atmospheric icing, 38 were caused by in-cloud icing, and in-cloud icing combined with freezing
nd o
14  rain caused an additional 26 failures. In Arizona, New Mexico, and the panhandles of Texas and
U ati
15  Oklahoma, the United States Forest Service specifies ice loads caused by in-cloud icing for towers
16  constructed at in specific mountaintop sitesregions (U.S. Forest Service 1994). Severe in-cloud
t / rm

17  icing has been observed in southern California (Mallory and Leavengood 1983a, 1983b), eastern
18  Colorado (NOAA Feb. 1978), the Pacific Northwest (Winkleman 1974; Richmond et al. 1977;
ou fo

19  Sinclair and Thorkildson 1980, Thorkildson 2019), Alaska (Ryerson and Claffey 1991), and the
e In

20  Appalachians (Ryerson 1987, 1988a, 1988b, 1990; Govoni 1990).

Snow. Snow accretions also can also result in severe structural loads and may occur anywhere
r

21 
Fo

22  snow falls, even in localities that may experience only one or two snow events per year. Some
23  examples of locations where snow accretion events resulted in significant damage to structures are
24  Nebraska (NPPD 1976), Maryland (Mozer and West 1983), Pennsylvania (Goodwin et al. 1983),
rik

25  Georgia and North Carolina (Lott 1993), Colorado (McCormick and Pohlman 1993), Alaska
26  (Peabody and Wyman 2005), and the Pacific Northwest (Hall 1977; Richmond et al. 1977).).
St

27  Alaska. For Alaska, available information indicates that moderate to severe snow and in-cloud
28  icing can be expected. The measurements made by Golden Valley Electric Association (Jones et
29  al. 2002Peabody and Wyman 2005) and Alaska Energy Authority (Peabody 1996) are consistent
30  in magnitude with visual observations across a broad area of central Alaska (Peabody 1993).

3 
 
 1  Several eteorological studies using an ice accretion model to estimate ice loads have been
2  performed for high-voltage transmission lines in Alaska (Gouze and Richmond 1982a, 1982b;
3  Richmond 1985, 1991, 1992; Peterka et al. 1996). Estimated 50-year mean recurrence interval
4  accretion thicknesses from snow range from 1.0 to 5.5 in. (25 to 140 mm), and in-cloud ice
5  accretions range from 0.5 to 6.0 in. (12 to 150 mm). The assumed accretion densities for snow and
6  in-cloud ice accretions, respectively, were 5 to 31 lb / ft 3 (80 to 500 kg / m3 ) and 25 lb / ft 3 (

400 kg / m3 ). These loads are valid only for the particular regions studied and are highly

at
7 

e LY
8  dependent on the elevation and local terrain features.Peterka et al. (1996) used an ice accretion

rm
9  model to estimate ice loads for high-voltage transmission lines in Alaska.

lin N
Fo
10  Hawaii. In Hawaii, for areas whereIce loads from freezing rain, in-cloud icing, or snow provided

er n O
11  in reports by Richmond, some of which were referenced in previous editions of ASCE 7, may be
12  excessively high because of choices made in the application of the ice accretion model that are
inconsistent with the available weather data (Jones 2019).
nd o
13 
U ati
14  In Hawaii, freezing rain (Wylie 1958), snow, and in-cloud icing are known to occur at higher
15  elevations, site-specific meteorological investigations are needed.
t / rm

16  .
ou fo

17  Local analyses. Local records and experience should be considered when establishing the design
e In

18  ice thickness or load, concurrent wind speed, and concurrent temperature., and ice density.
19  Information on ice-related damage and weather conditions in the area of interest may be available
r

20  from the local electric utilities; tree companies that are called on to remove fallen ice-covered trees
Fo

21  from roads and distribution lines; and state and county road departments. These sources would
22  supplement damage descriptions in Storm Data (1959-Present) and can help establish the
23  frequency of icing events along with any typical spatial variations in the icing severity in the
rik

24  area.In determining equivalent radial ice thicknesses from freezing rain from historical weather
25  data, the quality, completeness, and accuracy of the data should be considered along with the
St

26  robustness of the ice accretion algorithm. Meteorological stations may be closed by ice storms
27  because of power outages, anemometers may be iced over, and hourly precipitation data recorded
28  only after the storm when the ice in the rain gauge melts. These problems are likely to be more
29  severe at automatic weather stations where observers are not available to estimate the weather

4 
 
 1  parameters or correct erroneous readings. Note also that (1) air temperatures are recorded only to
2  the nearest 1°F, (0.6oC), at best, and may vary significantly from the recorded value in the region
3  around the weather station; (2) the wind speed during freezing rain has a significant effect on the
4  accreted ice load on objects oriented perpendicular to the wind direction; (3) wind speed and
5  direction vary with terrain and exposure; (4) enhanced precipitation may occur on the windward
6  side of mountainous terrain; and (5) ice may remain on the structure for days or weeks after
7  freezing rain ends, subjecting the iced structure to wind speeds that may be significantly higher

at
e LY
8  than those that accompanied the freezing rain. These factors should be considered both in

rm
9  estimating the accreted ice thickness at a weather station in past stormsfreezing rain events and in
extrapolating those thicknesses to a specific site.

lin N
10 

Fo
er n O
11  In using local data to determine extremes, it must also be emphasized that sampling errors can lead
12  to large uncertainties in the specification of the 500-year ice thicknessthicknesses or loads for long
13  mean recurrence intervals (MRI). Sampling errors are the errorsthose associated with the limited
nd o
14  size of the climatological data samples (years of record) number of cases in the sample of extremes,
U ati
15  related to the duration of the period of record and the frequency of icing events. When local records
16  of limited extent are used to determine extremes ice thicknesses, care should be exercised in their
t / rm

17  use.
ou fo

18  Ice loads from freezing rain, in-cloud icing, or snow provided in reports by Richmond, some of
19  which were referenced in previous editions of ASCE 7, may be excessively high because of choices
e In

20  made in the application of the ice accretion model that are inconsistent with the available weather
data (Jones 2019).
r

21 
Fo

22 

23  A robust ice accretion algorithm is not sensitive to small changes in input variables. For example,
rik

24  because temperatures are normally recorded in whole degrees, the calculated amount of ice
25  accreted should not be sensitive to temperature changesdifferences of fractions of a degree.
St

26  Because cloud water content is not measured at meteorological stations, ice loads from in-cloud
27  icing cannot be determined using local weather data.

28  C10.1.2 Dynamic Loads.

5 
 
 1  While design for dynamic loads is not specifically addressed in this edition of the standard, the
2  effects of dynamic loads are an important consideration for some ice-sensitive structures and
3  should be considered in the design when they are anticipated to be significant. For example, large-
4  amplitude galloping (Rawlins 1979; Section 6.2 of Simiu and Scanlan 1996) of guys and overhead
5  cable systems occurs in many areas. The motion of the cables can cause damage because of direct
6  impact of the cables on other cables or structures and can also cause damage because of wear and
7  fatigue of the cables and other components of the structure (White 1999). Ice shedding from the

at
e LY
8  guys on guyed masts can cause substantial dynamic loads in the mast.

rm
9  C10.1.3 Exclusions.

lin N
Fo
er n O
10  Current revisions of the consensus standards and other documents referenced in this document are
11  ASCE (2020), TIA (2019), and IEEE (202217). Additional guidance is available in Committee on
12  Electrical Transmission Structures (1982) and ). In Canada, the standards CSA (1987, 1994).2018,
nd o
13  2019) standards apply.
U ati
14 
t / rm

15  C10.2 DEFINITIONS

ou fo

16  FREEZING RAIN: Freezing rain occurs when warm, moist air is forced over a layer of
e In

17  subfreezing air at the eEarth’s surface. The precipitation usually begins as snow that melts as it
18  falls through the layer of warm air aloft. The drops then cool as they fall through the cold surface
r

19  air layer and may then freeze on contact with structures or the ground. Upper air data indicate that
Fo

20  the cold surface air layer is typically between 1,000 and 3,900 ft (300 and 1,200 m) thick (Young
21  1978), averaging 1,600 ft (500 m) (Bocchieri 1980). The warm air layer aloft averages 5,000 ft
22  (1,500 m) thick in freezing rain, but in freezing drizzle, the entire temperature profile may be below
rik

23  32 °F (0 °C) (Bocchieri 1980).

St

24  Precipitation rates and wind speeds are typically low to moderate in freezing rainstorms. Induring
25  freezing rain, he. The water impingement rate is often greater than the freezing rate. The excess
26  water drips off and may freeze as icicles, resulting in a variety of accretion shapes that range from
27  a smooth cylindrical sheath, through a crescent on the windward side with icicles hanging on the

6 
 
1  bottom, to large irregular protuberances, see Figure. C10.2-1. The shape of an accretion depends
2  on a combination of varying meteorological factors and the cross-sectional shape of the structural
3  member, its spatial orientation, and flexibility.

at
e LY
rm
lin N
Fo
er n O
4 

5 
nd o
U ati
t / rm
ou fo
e In
r
Fo

6 
rik
St

7 
 
 3 
2 
1 

 
St Fo
rik r
e In
ou fo
t / rm
U ati
nd o
er n O
lin N
e LY
Fo
rm

8 
at
 at
e LY
1 

rm
2 

lin N
FIGURE C10.2-1. Glaze iIce aAccretion cCaused by fFreezing rRain.

Fo
3 

er n O
4  Source:

Note that the theoretical maximum density of ice ( 917 kg / m3 or 57 lb / ft 3 )(57 lb/ft3917 kg/m3 or
nd o
5 
U ati
6  (917 kg/m357 lb/ft3)) is never reached in naturally formed accretions because of the presence of
7  air bubbles.
t / rm

8  HOARFROST: Hoarfrost, which is often confused with rime, forms by a completely different
9  process. Hoarfrost is an accumulation of ice crystals formed by direct deposition of water vapor
ou fo

10  from the air on an exposed object. Because it forms on objects with surface temperatures that have
e In

11  fallen below the frost point (a dew point temperature below freezing) of the surrounding air, due
12  to because of strong radiational cooling, hoarfrost is often found early in the morning after a clear,
r

13  cold night. It is feathery in appearance and typically accretes up to about 1 in. (25 mm) in
Fo

14  thickness, and has with very little weight. Hoarfrost does not constitute a significant loading
15  problem; however, it is a very good collector of supercooled fog droplets. In light winds, a
16  hoarfrost-coated wire may accrete rime faster than a bare wire (Power 1983).
rik

17  ICE-SENSITIVE STRUCTURES: Ice-sensitive structures are structures for which the load
St

18  effects from atmospheric icing control the design of part, or all, of the structural system. Many
19  open structures are efficient ice collectors, so ice accretions can have a significant load effect. The
20  sensitivity of an open structure to ice loads depends on the size and number of structural members,
21  components, and appurtenances and also on the other loads for which the structure is designed.

9 
 
 1  For example, the additional weight of ice that may accrete on a heavy wide-flange member is
2  smaller in proportion to the dead load than the same ice thickness on a light angle member. Also,
3  the percentage increase in projected area for wind loads is smaller for the wide-flange member
4  than for the angle member. For some open structures, other design loads, for example, snow loads
5  and live loads on a catwalk floor, may be larger than the design ice load.

6  IN-CLOUD ICING: This icing condition occurs when a cloud or fog (consisting of supercooled

at
7  water droplets, 100 microns (100 mm) or less in diameter), blown by the wind encounters a surface

e LY
8  that is at, or below, freezing temperature. It occurs in mountainous areas where adiabatic cooling

rm
9  causes saturation of the atmosphere to occur at temperatures below freezing, in free air in

lin N
10  supercooled clouds, and in supercooled fogs produced by a stable air mass with a strong

Fo
er n O
11  temperature inversion. In-cloud ice accretions can reach thicknesses of 1 ft (0.30 m) or more
12  because the icing conditions can include high winds and typically, persist or recur episodically
13  during long periods of subfreezing temperatures. Large concentrations of supercooled droplets are
nd o
14  not common at air temperatures below about 0°F (  18  C ).-22 oF (-30 oC).
U ati
15  In-cloud ice accretions have densities ranging from that of low-density rime to glaze. When
t / rm

16  convective and evaporative cooling removes the heat of fusion as fast as it is released by the
17  freezing droplets, the drops freeze on impact. When the cooling rate is lower, the droplets do not
ou fo

18  completely freeze on impact. The unfrozen water then spreads out on the object and may flow
19  completely around it and even drip off to form icicles. The degree to which the droplets spread as
e In

20  they collide with the structure and freeze governs how much air is incorporated in the accretion,
and thus, its density. The density of ice accretions caused by in-cloud icing varies over a wide
r

21 
Fo

22  range from 5 to 56 pcflb/ft3 (80 to 900 kg / m3 ) 900 kg/m3) (Macklin 1962; Jones 1990). The
23  resulting accretion can be either white or clear, possibly with attached icicles; see Fig.ure C10.2-
24  2.
rik
St

10 
 
 3 
2 
1 

 
St Fo
rik r
e In
ou fo
t / rm
U ati
nd o
er n O
lin N
e LY
Fo
rm

11 
at
 1  FIGURE C10.2-2. Rime iIce aAccretion cCaused by iIn-cCloud iIcing.

2  Source:

3  The amount of ice accreted during in-cloud icing depends not only on the cloud properties, but
4  also on the size of the accreting object, the duration of the icing condition, and the wind speed. If,
5  as often occurs, wind speed increases and air temperature decreases with height above ground,
6  larger amounts of ice accrete on taller structures. The accretion shape depends on the flexibility of

at
e LY
7  the structural member, component, or appurtenance. If it is free to rotate, such as a long guy or a

rm
8  long span of a single conductor or wire, the ice accretes with a roughly circular cross section. On
more rigid structural members, components, and appurtenances, the ice forms in irregular pennant

lin N
9 

Fo
10  shapes extending into the wind.

er n O
11  SNOW: Under certain conditions, snow falling on objects may adhere because of capillary forces,
12  interparticle freezing (Colbeck and Ackley 1982), and/or sintering (Kuroiwa 1962). On objects
nd o
13  with circular cross section, such as a wire, cable, conductor, or guy, sliding, deformation, and/or
U ati
14  torsional rotation of the underlying cable may occur, resulting in the formation of a cylindrical
15  sleeve, even around bundled conductors and wires; see Figure. C10.2-3. BecauseWhen accreting
t / rm

16  snow is often accompanied by high winds, the density of accretions may be much higher than the
17  density of the same snowfall on the ground.
ou fo

18 
e In
r
Fo
St
rik

19 

20  FIGURE C10.2-3. Snow aAccretion on wWires.

12 
 
 1  Damaging snow accretions have been observed at surface air temperatures ranging from about 23
2  to 36 °F ( 5(-5 to 2 °C). Snow with a high moisture content appears to stick more readily than drier
3  snow. Snow falling at a surface air temperature above freezing may accrete even at wind speeds
4  above 25 mi/ h ( 10 m / s ),25 mi/h (10 m/s), producing dense 37 to 50 pcf (600 to 800 kg / m3 )800
5  kg/m3) accretions. Snow with a lower moisture content is not as sticky, blowing off the structure
6  in high winds. These accreted snow densities are typically between 2.5 and 16 pcflb/ft3 (40 and
250 kg / m3 )and 250 kg/m3) (Kuroiwa 1965).

at
7 

e LY
rm
8  Even apparently dry snow can accrete on structures (Gland and Admirat 1986). The cohesive
strength of the dry snow is initially supplied by the interlocking of the flakes and ultimately by

lin N
9 

Fo
10  sintering, as molecular diffusion increases the bond area between adjacent snowflakes. These dry

er n O
11  snow accretions appear to form only in very low winds and have densities estimated at between 5
12  and 10 pcflb/ft3 (80 and 150 kg / m 3 )150 kg/m3) (Sakamoto et al. 1990; Peabody 1993).
nd o
13  Recent references oninvolving measuring and modeling snow loads on ice sensitive structures in
U ati
14  Europe and Japan include Eliasson et al. (2013, 2015), Nygaard et al. (2013a), Lacavalla et al.
(2015, 2019), Nikolov et al. (2015), Ueno et al. (2015), Faggian et al. (2019), Iversen et al. (2019),
t / rm

15 
16  and Matsumiya et al. (2019). iIn Nygaard et al. (2013b), d.Design ice loads for snow and in-cloud
icing are mapped for Great Britain. in Nygaard et al. (2013b).
ou fo

17 
e In

18  C10.4 ICE LOADS CAUSED BY FREEZING RAIN

r

19 
Fo

20  C10.4.1 Ice WeightLoad.

21  The ice thicknesses shown in Figuress. 10.4-2 throughand 10.4-635 were determined for a
rik

22  horizontal cylinder oriented perpendicular to the wind. These ice thicknesses cannot be applied
23  directly to cross sections that are not round, such as channels and angles. However, the ice area
St

24  from Eq.uation (10.4-1) is the same for all shapes for which the circumscribed circles have equal
25  diameters (Peabody and Jones 2002; Jones and Peabody 2006). It is assumed that the maximum
26  dimension of the cross section is perpendicular to the trajectory of the raindrops. Similarly, the ice
27  volume in Equation. (10.4-2) is for a flat plate perpendicular to the trajectory of the raindrops. The

13 
 
 1  constant π in Equation . (10.4-2) corrects the thickness from that on a cylinder to the thickness on
2  a flat plate. For vertical cylinders and horizontal cylinders parallel to the wind direction, the ice
3  area given by Equation . (10.4-1) is conservative.

4  C10.4.2 Nominal Ice Thickness.

5  The ASCE 7-22 Atmospheric Icing Subcommittee generated the nominal ice thickness maps
shown in Figuress. 10.4-2 (lower 48 states), and 10.4-3 (Alaska), 10.4-4 (Columbia Gorge

at
6 

e LY
7  region), and 10.4-5 (Lake Superior region) for Risk Categories I, II, III, and IV, using an ice

rm
8  accretion model and local data. Location-specific nominal ice thicknesses may be determined

lin N
9  using the ASCE Atmospheric Ice Geodatabase, which can be accessed at the ASCE 7 Hazard

Fo
er n O
10  Tool (https://asce7hazardtool.online/). This geodatabase provides nominal ice thickness,
11  concurrent wind and concurrent temperature data based on a defined location using either
12  latitude/longitude or an address. The website results access the same data used to develop the
nd o
13  paper maps currently in the standard.
U ati
14  The MRI for the four rRisk cCategories come directly from matching the ASCE 7-2016
t / rm

15  Importance Factors for ice in Table 1.5-2, with the ASCE 7-2016 Table C10.4-1 multipliers on
16  500-year ice thickness as a function of MRI, as shown in the first three columns of Table C10.4-
ou fo

17  1 below. As the generalized Pareto distribution is used to characterize extreme ice thicknesses
18  from freezing rain, the tail shape parameter, k, is determined by the sample of extremes for each
e In

19  superstation. The tail is finite for k>0, exponential for k=0, and increasingly heavy for more
20  negative k. Thus, the values in cColumn 2 represented the mean ice thickness multiplier for the
r
Fo

21  MRI. Columns 4 through 6 show the mean, minimum and maximum values for the multiplier for
22  the 2022 extreme value analysis. The variation in tail weight among the superstations is reflected
23  in the contours in Figs. 10.4-2 andthrough 10.4-35.
rik

24  Table C10.4-1.: Risk Category Importance Factors and MRI from ASCE7-16 and Statistics
25  of Ice Thickness Multipliers for the 2022 Extreme Value Analysis.
St

ASCE 7-2016 Statistics of ice thickness

multipliers for ASCE 7-2022

14 
 
 Risk Ice MRI Mean Minimum Maximum
Category thickness (Table multiplier multiplier multiplier
multiplier C10.4-
(Table 1.5- 1)
2)
I 0.80 250 0.85 0.70 1.00
II 1 500 1 1 1

at
e LY
III 1.15 1000 1.17 1.00 1.43

rm
IV 1.25 1400 1.26 1.00 1.70

lin N
1 

Fo
er n O
2  The 500-year mean recurrence interval ice thicknesses shown in Figs. 10.4-2 toand 10.4-63 are
3  based on studies using an ice accretion model and local data.
nd o
4  Historical weather data from 540 National Weather Service (NWS), military, Federal Aviation
U ati
5  Administration (FAA), and Environment Canada weather stations were used, with the U.S. Army’s
6  Cold Regions Research and Engineering Laboratory (CRREL) and simple ice accretion models
t / rm

7  (Jones 1996, 1998), to estimate uniform radial glaze ice thicknesses in past freezing rainstormsrain
8  events. The models and algorithms have been applied to additional stations in Canada along the
ou fo

9  border of the lower 48 states and Alaska. The station locations are shown in Fig.ure C10.4-1 for
the 48 contiguous states and in Fig.ure 10.4-63 for Alaska. The period of record of the
e In

10 
11  meteorological data at any station is typically 20 to 50 years. The ice accretion models use weather
r

12  and precipitation data to simulate the accretion of ice on cylinders 33 ft (10 m) above the ground,
Fo

13  oriented perpendicular to the wind direction in freezing rainstormsrain. Accreted ice is assumed to
14  remain on the cylinder until after freezing rain ceases and the air temperature increases to at least
15  33 °F (0.6 °C). At each station, the maximum ice thickness and the maximum wind-on-ice load
rik

16  were determined for each stormevent. Severe stormsevents, those with significant ice or wind-on-
17  ice loads at one or more weather stations, were researched in Storm Data (NOAA 1959–Present),
St

18  newspapers, and utility reports to obtain corroborating qualitative information on the extent of and
19  damage from the storm. Yet, very little corroborating information was obtained about damaging
20  freezing rainstormsrain events in Alaska, perhaps because of the low population density and
21  relatively sparse newspaper coverage in the state.

15 
 
 at
e LY
rm
lin N
1 

Fo
er n O
2  FIGURE C10.4-1. Locations of wWeather sStations uUsed in pPreparation of
3  FiguresfFigure 10.4-2 through 10.4-65.
nd o
4  Extreme ice thicknesses were determined from an extreme value analysis using the peaks-over-
U ati
5  threshold methodL-moments and the generalized Pareto distribution (HoskingsHosking and
6  Wallis 1987, 1997; Wang 1991; Abild et al. 1992). To reduce sampling error, weather stations
t / rm

7  were grouped into superstations (Peterka 1992) based on the incidence of severe storms, the
8  frequency of freezing rainstormsrain events, latitude, proximity to large bodies of water, elevation,
ou fo

9  and terrain., and similarity in the occurrence rate of ice thicknesses of at least 1 mm (0.04 in.ch)
(1 mm). Concurrent wind-on-ice speeds were back-calculated from the extreme wind-on-ice load
e In

10 
11  and the extreme ice thickness. The analysis of the weather data and the calculation of extreme ice
r

12  thicknesses are described in more detail in Jones et al. (2002). The maps in Figuress. 10.4-2 toand
Fo

13  through 10.4--56 represents3 represent the most consistent and best available nationwide map for
14  nominal design ice thicknesses and wind-on-ice speeds.. The icing model used to produce the
15  mapmaps has not, however, been verified with a large set of collocated measurements of
rik

16  meteorological data and measured equivalent uniform radial ice thicknesses. Furthermore, the
17  weather stations used to develop this these maps are almost all at airports. Structures in more
St

18  exposed locations at higher elevations or in valleys or gorges, for example, Signal and Lookout
19  Mountains in Tennessee, the Pontotoc Ridge and the edge of the Yazoo Basin in Mississippi, the
20  Shenandoah Valley and Poor Mountain in Virginia, Mount Washington in New Hampshire, and
21  Buffalo Ridge in Minnesota and South Dakota, Stampede Pass in Washington, Sexton Summit in

16 
 
 1  Oregon, Blue Canyon in California, and Flagstaff in Arizona, may be subject to larger ice
2  thicknesses and higher concurrent wind speeds. However, structures in more sheltered locations,
3  for example, along the north shore of Lake Superior within 300 vertical ft (90 m) of the lake, may
4  be subject to smaller ice thicknesses and lower concurrent wind speeds.. Where complex terrain
5  traps cold air near the surface, freezing rain may last longer than at nearby weather stations,
6  resulting in larger local ice thicknesses than arethose shown on the maps. Complex terrain causing
7  wind speed-up or precipitation enhancement during freezing rain will also cause larger ice

at
e LY
8  thicknesses. These complex terrain features include, but are not limited to, hills, mountains, ridges,

rm
9  valleys, gorges and basins. Loads from snow or in-cloud icing may be more severe than those
from freezing rain (see Section C10.1.1).

lin N
10 

Fo
er n O
11  Special Icing Regions. Special icing regions are identified by shading on the mapmaps in Figures
12  10.4-2 andthrough 10.4-35. As described previously, freezing rain occurs only under special
13  conditions when a cold, relatively shallow layer of air at the surface is overrun by warm, moist air
nd o
14  aloft. For this reason, severe freezing rainstormsrain at high elevations in mountainous terrain
U ati
15  typically dodoes not occur in the same weather systems that cause severe freezing rainstormsrain
16  at the nearest airport with a weather station. In the Cascades of Oregon and Washington, ice
t / rm

17  thicknesses may exceed the mapped values in foothills and passes. However, at elevations above
18  5,000 ft, freezing rain is unlikely. Furthermore,Shaded elevations above 2,100 ft (640 m) in these
ou fo

19  regions, ice thicknessesthe east, 6,000 ft (1,829 m) in the west, and wind-on-ice loads may vary
e In

20  significantly over short distances because of local variations in elevation, topography, and
21  exposure.1,600 ft (488 m) in Alaska are not well represented by weather data. In these mountainous
r

22  regions, the values given in Figuress. 10.4-12 andthrough 10.4-35 should be adjusted, based
Fo

23  onlocal on a combination of local historical records and experience, to account for possibly higher
24  ice loads from both freezing rainreanalysis data (e.g. Mesinger et al 2006, Gelaro et al 2017), and
25  in-cloud icing (see Section C10.1.1).numerical weather prediction systems (e.g. Powers et al
rik

26  2017). .
St

27  C10.4.4 Importance Factors.3 Height factor

28  The importance factors for ice and concurrentRaindrops can evaporate as they fall, sometimes
29  resulting in virga, which is rain that does not reach the ground. Thus, within the cold air layer at
30  the ground surface in freezing rain, there is more water in the air further from the ground. This

17 
 
 1  enhances the height factor associated with the increase in wind wind adjust the nominalspeed with
2  height above ground (Jones and Eylander 2017). This study suggests that the increase in ice
3  thickness ice thickness with height above ground should be applied along the full structure height,
4  rather than limited to the lowest 900 ft (275 m).

5  C10.4.5 Design Ice Thickness for Freezing Rain.

The design load on the structure is a product of the nominal design load and the load factors

at
6 

e LY
7  specified in Chapter 2. The load factors for load and resistance factor design (LRFD) for

rm
8  atmospheric icing are 1.0. and

lin N
Fo
9  The studies of ice accretion, on which the maps are based, indicate that the concurrent wind gust

er n O
10  pressure for Risk Category I structures from a 500-yearspeed on ice does not increase with mean
11  recurrence interval to a 250-year. The lateral wind-on-ice load does, however, increase with mean
12  recurrence interval. For Risk Category III and IV structures, they are adjusted to 1,000-year and
nd o
13  1,400-year mean recurrence intervals, respectively. The importance factor is multiplied times the
U ati
14  ice thickness rather than the ice load because the ice load from Eq. (10.4-1) depends on the
t / rm

15  diameter of the circumscribing cylinder as well as the design ice thicknessthickness increases. The
16  concurrent wind gust speed used with the nominal ice thickness is based on both the winds that
ou fo

17  occur during the freezing rainstormrain and those that occur between the time the freezing rain
18  stops and the time the temperature rises to above freezing. When the temperature rises above
e In

19  freezing, it is assumed that the ice melts enough to fall from the structure. In the colder northern
20  regions, the ice generally stays on structures for a longer period of time after the end of a storm
r
Fo

21  resulting in higher concurrent wind speeds. The results of the extreme value analysis show that the
22  concurrent wind speed does not change significantly with mean recurrence interval. The lateral
23  wind-on-ice load does, however, increase with mean recurrence interval because the ice thickness
increases. The importance factors differ from those used for both the wind loads in Chapter 6 and
rik

24 
25  the snow loads in Chapter 7 because the extreme value distribution used for the ice thickness is
St

26  different from the distributions used to determine the extreme wind speeds in Chapter 6 and snow
27  loads in Chapter 7. See also Table C10.4-1 and the discussion under Section C10.4.6.freezing rain
28  resulting in higher concurrent windgust speeds.

29  Table C10.4-1 Mean Recurrence Interval Factors

18 
 
 Mean Recurrence Interval Multiplier on Ice Thickness Multiplier on Wind Pressure

25 0.40 1.0

50 0.50 1.0

100 0.625 1.0

at
200 0.75 1.0

e LY
rm
250 0.80 1.0

lin N
Fo
300 0.85 1.0

er n O
400 0.90 1.0
nd o
500 1.00 1.0
U ati

1,000 1.15 1.0

t / rm

1,400 1.25 1.0

ou fo

1 
e In

2  C10.4.6 Design Ice Thickness for Freezing Rain.

r
Fo

3  The design load on the structure is a product of the nominal design load and the load factors
4  specified in Chapter 2. The load factors for load and resistance factor design (LRFD) for
5  atmospheric icing are 1.0. Table C10.4-1 shows the multipliers on the 500-year mean recurrence
rik

6  interval ice thickness and concurrent wind speed used to adjust to other mean recurrence intervals.

7  The studies of ice accretion on which the maps are based indicate that the concurrent wind speed
St

8  on ice does not increase with mean recurrence interval (see Section C10.4.4).

9  The 2002 eEdition of ASCE 7 was the first edition to include atmospheric icing maps. At that time,
10  Ffifty-year mean recurrence interval (MRI) maps were provided at that time in order to match the

19 
 
 1  approach used for the wind and snow load maps. An ice-thickness multiplier equal to 2 was
2  included (see Equation. (10.4-5) in ASCE 7-02 to ASCE 7-10) to adjust the mapped values to a
3  500-year MRI for design. Thus, the adjusted 500-year MRI value would be appropriate for use
4  with the LRFD load factor of 1.0 shown for ice loads in Chapter 2. At that time, aA 500-year MRI
5  load was consistent at that time with those historically used for seismic and wind loads. The 2016
6  edition of ASCE 7 includesincluded atmospheric icing maps based on a 500-year MRI load with
7  no ice-thickness multiplier in Equation. (10.4-5). The 2016 maps have beenwere redrawn directly

at
e LY
8  from the original extreme value analysis. Design load changes from the 2010 edition to the 2016

rm
9  edition awere caused by the maps being redrawn, not caused by changing the map MRI.by
changing the map MRI. For thise 2022 Eedition, the weather stations are grouped into larger

lin N
10 

Fo
11  superstations to generate longer periods of record. Ice thickness extremes are determined for MRI

er n O
12  from 250 to 1,400 years, using L-moments to calculate all three generalized Pareto distribution
13  parameters, rather than setting a threshold just below the smallest value in the sample of extremes
and calculating the other two parameters. The maps in this editionASCE 7-22 differ from those in
nd o
14 
U ati
15  all the previous editions in that they are contour maps rather than zone maps. Columns 4 through
16  6 in Table C10.4-1 show the mean, minimum and maximum values for the multiplier for the 2022
t / rm

17  extreme value analysis. The variation in tail weight among the superstations is reflected in the
18  contours in Figuress. 10.4-2 through 10.4-5.
ou fo

19  Factors to multiply the 500-year MRI ice thicknesses from freezing rain to estimate thicknesses
e In

20  for shorter mean recurrence intervalsMRI are shown in Table C10.4-12.
r

21  Table C10.4-12. Mean Recurrence Interval Factors.

Fo

MRI Ice thickness

(years) multiplier
rik

25 0.4
50 0.5
St

100 0.625
500 1.0
22 

20 
 
 1  When the reliability of a system of structures or one interconnected structure of large extent is
2  important, spatial effects should also be considered. All of the cellular telephone antenna structures
3  that serve a state or a metropolitan area could be considered to be a system of structures. Long
4  overhead electric transmission lines and communicationscommunication lines are examples of
5  large, interconnected structures. Figuress. 10.4-2 throughand 10.4-63 are for ice loadsthicknesses
6  appropriate for a single structure of small areal extent. Large, interconnected structures and
7  systems of structures are hit by icing storms more frequently than a single structure. The frequency

at
e LY
8  of occurrence increases with the area encompassed or the linear extent. To obtain equal risks of

rm
9  exceeding the design load in the same icing climate, the individual structures forming the system,
or the large, interconnected structure, should be designed for a larger ice load than a single structure

lin N
10 

Fo
11  (CEATI 2003, 2005; Chouinard and Erfani 2006; Golikova 1982; Golikova et al. 1982; Jones

er n O
12  2010).

13  C10.5 WIND ON ICE-COVERED STRUCTURES

nd o
U ati
14  The ASCE 7 Atmospheric Icing Subcommittee generated the concurrent 3-s gust speed maps
15  shown in Figuress. 10.5-1 (lower 48 states) and 10.5-2 (Alaska), using an ice accretion model and
t / rm

16  local data. The calculation of concurrent wind speeds is described in Sections C10.4.2 and C10.4.5.
ou fo

17  Ice accretions on structures change the structure’s wind drag coefficients. The ice accretions tend
18  to round sharp edges, reducing the drag coefficient for such members as angles and bars. Natural
e In

19  ice accretions can be irregular in shape with an uneven distribution of ice around the object on
20  which the ice has accreted. The shape varies from storm to storm and from place to place within a
r
Fo

21  storm. The actual projected area of a glaze ice accretion may be larger than that obtained by
22  assuming a uniform ice thickness.

23  C10.5.5 Wind on Ice-Covered Guys and Cables.

rik

24  There are practically no published experimental. Wind tunnel data giving theon force coefficients
St

25  for ice- covered guys and cables. are compiled and analyzed in CEATI (2009), using measurements
26  from references dating from 1950 to 2005. Force coefficients vary with the shape of the accreted
27  ice and the angle of attack of the wind relative to the maximum dimension of the shape. There
28  have also been many studies of the force coefficient for cylinders without ice. The force coefficient

21 
 
 1  varies with the surface roughness and the Reynolds number. At subcritical Reynolds numbers,
2  both smooth and rough cylinders have force coefficients of approximately 1.2, as do square
3  sections with rounded edges (Figure. 4.5.5 in Simiu and Scanlan 1996). For a wide variety of
4  stranded electrical transmission cables, the supercritical force coefficients are approximately 1.0
5  with subcritical values as high as 1.3 (Fig.ure 5-2 in Shan 1997). The transition from subcritical to
6  supercritical depends on the surface characteristics and takes place over a wide range of Reynolds
7  numbers. For the stranded cables described in Shan (1997), the range is from approximately 25,000

at
e LY
8  to 150,000. For a square section with rounded edges, the transition takes place at a Reynolds

rm
9  number of approximately 800,000 (White 1999). The concurrent 3-s gust wind speed in
Figsure. 10.4-2 through 10.4-55-1 for the contiguous 48 states varies from 30 to 60 mi/ h 60 mi/h

lin N
10 

Fo
(13.4 to 26.8 m/s ), 27 m/s), with speeds in Figure. 10.4-65-2 for Alaska up to 80 mi/ h ( 35.8 m/s

er n O
11 

12  ).80 mi/h (36 m/s). Table C10.5-1 shows the Reynolds numbers (using U.S. standardthe kinematic
13  viscosity of air at 5 oF (-15 oC) at 1 atmosphere) for a range of iced guys and cables. In practice,
nd o
14  the Reynolds numbers range from subcritical through critical to supercritical depending on the
U ati
15  roughness of the ice accretion. Considering that the shape of ice accretions is highly variable from
16  relatively smooth cylindrical shapes to accretions with long icicles with projected areas greater
t / rm

17  than the equivalent radial thickness used in the maps, a single force coefficient of 1.2 has been
18  chosen.
ou fo

19  Table C10.5-1. Typical Reynolds Numbers for Iced Guys and Cables.
e In

       
r

Guy or
Fo

Cable Ice Design Ice Iced Concurrent 3-s

Diameter Thickness Importance Thickness t d Diameter Gust Wind Reynolds
(in.) t (in.) Factor Iw (in.) (in.) Speed ( mi / h ) Number
rik

Contiguous
St

48 States

0.250 0.25 0.80 0.20 0.650 30 15,200

22 
 
 0.375 0.25 0.80 0.20 0.775 30 18,100

0.375 1.25 1.25 1.563 3.500 60 163,000

1.000 0.25 0.80 0.20 1.400 30 32,700

1.000 1.25 1.25 1.563 4.125 60 192,000

at
2.000 1.25 1.25 1.563 5.125 60 239,000

e LY
rm
Alaska

lin N
Fo
0.250 0.25 0.80 0.20 0.650 50 27,000

er n O
2.000 0.50 1.25 0.625 3.250 80 202,000
nd o
1 
U ati
guy or  design 
t / rm

cable  ice  iced  concurrent 

diameter  thickness  diamete gust speed  Reynolds 
ou fo

(in.)  td(in.)  r (in.)  (mi/h)  number 

e In

0.3 0.25  0.8  30 22,000 

2.0 0.5  3  30 84,000 
r

4.0 0.25  4.5  40 169,000 

Fo

6.0 1.5  9  40 337,000 

1.0 3.5  8  40 300,000 
3.0 0.5  4  50 187,000 
rik

5.0 1  7  50 328,000 
St

4.0 0.75  5.5  60 309,000 

2.0 2  6  60 337,000 
1.0 0.5  2  80 150,000 

23 
 
 1  Note: To convert in. to mm, multiply by 25.4. To convert mi / h mi/h to k m / h ,m/s, multiply by
2  1.60.45. 

3  C10.6 DESIGN TEMPERATURES FOR FREEZING RAIN

4  Some ice-sensitive structures, particularly those using overhead cable systems, are also sensitive
5  to changes in temperature. In some cases, the maximum load effect occurs around the melting

at
6  point of ice (32 °F or 0 °C) and in others at the lowest temperature that occurs while the structure

e LY
7  is loaded with ice. Figuress. 10.6-1 and 10.6-2 show the low temperatures to be used for design in

rm
8  addition to the melting temperature of ice.

lin N
Fo
9  The freezing rain model described in Section C10.4.2 tracked the temperature during each modeled

er n O
10  icing event. For each event, the minimum temperature that occurred with the maximum ice
11  thickness was recorded. The minimum temperatures for all the freezing rain events used in the
extreme value analysis of ice thickness were analyzed to determine the 10th percentile temperature
nd o
12 
at each superstation (i.e., the temperature that was exceeded during 90% of the extreme icing
U ati
13 
14  events). These temperatures were used to make the maps shown in Figuress. 10.6-1 and 10.6-2. In
t / rm

15  areas where the temperature contours were close to the wind or ice thickness contours, they were
16  moved to coincide with, first, the concurrent wind boundaries, and, second, the ice zone
ou fo

17  boundaries.
e In

18  C10.7 PARTIAL LOADING

r

19  Variations in ice thickness caused by freezing rain on objects at a given elevation are small over
Fo

20  distances of about 1,000 ft (300 m). Therefore, partial loading of a structure from freezing rain is
21  usually not significant (Cluts and Angelos 1977).

In-cloud icing is more strongly affected by wind speed, thus partial loading caused by differences
rik

22 
23  in exposure to in-cloud icing may be significant. Differences in ice thickness over several
St

24  structures or components of a single structure are associated with differences in the exposure. The
25  exposure is a function of shielding by other parts of the structure and by the upwind terrain.

26  Partial loading associated with ice shedding may be significant for snow or in-cloud ice accretions
27  and for guyed structures when ice is shed from some guys before others. 

24 
 
 1  REFERENCES

2  Abild, J., Andersen, E. Y., and Rosbjerg, L. (1992). “The climate of extreme winds at the Great
3  Belt, Denmark.” J. Wind Engrg. Indust. Aerodyn. 41–44, 521–532.

4  ASCE. (2020). Guidelines for Eelectrical Ttransmission Lline Sstructural Lloading, 4th edition,
5  Manual of Practice 74, edited by Frank Agnew, Reston, VA.

at
6  Bennett, I. (1959). “Glaze: Its meteorology and climatology, geographical distribution and

e LY
7  economic effects.” Environmental Protection Research Division Technical Report EP-105, U.S.

rm
8  Army Quartermaster, Research and Engineering Center, Natick, MA.

lin N
Fo
9  Bernstein, B. C., and Brown, B. G. (1997). “A climat ologyclimatology of supercooled large drop

er n O
10  conditions based upon surface observations and pilot reports of icing.” Proc., 7th Conf. on
11  Aviation, Range and Aerospace Meteorology, Feb. 2–7, Long Beach, CA.
nd o
12  Bocchieri, J. R. (1980). “The objective use of upper air soundings to specify precipitation type.”
U ati
13  Mon. Weather Rev. 108, 596–603.
t / rm

14  Canadian Standards Association (CSA). (19872019). “Overhead transmission lines.” CAN/—
15  Design criteria.” CSA- C22.3 No. 1-M87. Rexdale60826. Toronto, ON.
ou fo

16  CSA. (19942018). “Antennas, towers and antenna-supporting structures.” CSA-S37-9418. CSA,
e In

17  RexdaleToronto, ON.

CEATI (Centre for Energy Advancement through Technological Innovation). (CEATI). (2003).
r

18 
Fo

19  “Spatial factors for extreme ice and extreme wind: Task 1 literature review on the determination
20  of spatial factors.” T033700-3316B. Montreal.

21  , QC.
rik

22  CEATI. (2005). “Spatial factors for extreme ice and extreme wind: Task 2 calculation of spatial
St

23  factors from ice and wind data.” T033700-3316B-2. Montreal, QC.

24  CEATI (2009). “Technology Watch for the Drag Coefficient of Ice-Covered Conductors”
25  T083700-3360, Montreal, QC.

25 
 
 1  Chouinard, L., and Erfani, R. (2006). “Spatial modeling of atmospheric icing hazards.” Intl. Forum
2  on Engineering Decision Making, April 26–29, Lake Louise, Canada.

3  Cluts, S., and Angelos, A. (1977). “Unbalanced forces on tangent transmission structures.” IEEE
4  Winter Power Meeting, Paper No.A77-220-7. IEEE, Los Alamitos, CA.

5  Colbeck, S. C., and Ackley, S. F. (1982). “Mechanisms for ice bonding in wet snow accretions on
6  power lines.” Proc., 1st Intl. Workshop on Atmospheric Icing of Structures, L. D. Minsk, ed. U.S.

at
e LY
7  Army CRREL Special Report, 25–30, 83–17. Hanover, NH, 25-30.

rm
8  Committee on Electrical Transmission Structures. (1982). “Loadings for electrical transmission

lin N
9  structures by the committee on electrical transmission structures.” J. Struct. Div., 108(5), 1088–

Fo
er n O
10  1105.

11  CSA (Canadian Standards Association). (2019). “Overhead transmission lines—Design criteria.”
CSA C22.3 No. 60826. Toronto, ON.
nd o
12 
U ati
13  CSA. (2018). “Antennas, towers and antenna-supporting structures.” CSA-S37-18. CSA, Toronto,
14  ON.
t / rm

15  Danielson, J. J. and Gesch, D. B. (2011). Global multi-resolution terrain elevation data
ou fo

16  (GMTED2010); U.S. Geological Survey Open-File Report 2011–1073, 26 p.

e In

17  Druez, J., Louchez, S., and McComber, P. (1995). “Ice shedding from cables”, Cold Regions
18  Science and Technology, 23(4), 377-388.
r
Fo

19  Elíasson, A. J., Ísaksson, S. P., Ágústsson, H., and Thorsteins, E. (2015). “Wet snow icing--
20  Comparing simulated accretion with observational experience”, Proc., 16th Intl. Workshop on
21  Atmospheric Icing of Structures, 28 June-3 July, Upsalla, Sweden.
rik

22  Elíasson, A. J., Ágústsson, H., Hannesson, G. M., and Thorsteins E. (2013). “Modeling wet-snow
St

23  accretion; Comparison of cylindrical model to field measurements”, Proc., 15th Intl. Workshop on
24  Atmospheric Icing of Structures, 8-11 September, St. Johns, Canada.

25  Engineering Toolbox (2003) Altitude above Sea Level and Air Pressure. [online] Available at:
26  https//www.engineeringtoolbox.com/air-altitude-pressure-d_462.html [Accessed 17 Sep 2020].

26 
 
 1  Faggian, P., Bonanno, R., and Pirovano, G. (2019). “Research activities to cope with wet snow
2  impacts on overhead power lines in future climate over Italy”, Proc., 18th Intl. Workshop on
3  Atmospheric Icing of Structures, 23-26 June, Reykjavik, Iceland.

4  Finstad, K. J., Lozowski E. P., and Gates, E. M. (1988). “A computational investigation of water
5  droplet trajectories”, J. of Atmospheric and Oceanic Technology, 5, 160-170.

Gelaro, R., McCarty, W., Suárez, M.J., Todling, R., Molod, A., Takacs, L., et al. (2017). “The

at
6 

e LY
7  Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2)”, J.

rm
8  Clim., doi: 10.1175/JCLI-D-16-0758.1

lin N
Fo
9  Gland, H., and Admirat, P. (1986). “Meteorological conditions for wet snow occurrence in

er n O
10  France—Calculated and measured results in a recent case study on 5 March 1985.” Proc., 3rd Intl.
11  Workshop on Atmospheric Icing of Structures, L. E. Welsh and D. J. Armstrong, eds., Canadian
12  Climate Program, Vancouver, 91–96.
nd o
U ati
13  Golikova, T. N. (1982). “Probability of increased ice loads on overhead lines depending on their
14  length.” Soviet Power No. 10, Ralph McElroy Co., 888–894.
t / rm

15  Golikova, T. N., Golikov, B. F., and Savvaitov, D. S. (1982). “Methods of calculating icing loads
ou fo

16  on overhead lines as spatial constructions.” Proc., 1st Intl. Workshop on Atmospheric Icing of
17  Structures, L. D. Minsk, ed. U.S. Army CRREL Special Report No. 83-17, Cold Regions Research
e In

18  and Engineering Laboratory, Hanover, NH, 341–345.

r

19  Goodwin, E. J., Mozer, J. D., DiGioia, A. M., Jr., and Power, B. A. (1983). “Predicting ice and
Fo

20  snow loads for transmission line design.” Proc., 3rd Intl. Workshop on Atmospheric Icing of
21  Structures, L. E. Welsh, and D. J. Armstrong, eds., Canadian Climate Program, 1991, Vancouver,
22  267–275.
rik

23  Gouze, Sand Richmond, M. C. (1982a). Meteorological evaluation of the proposed Alaska
St

24  transmission line routes, Meteorology Research, Altadena, CA.

25  Gouze, S. C., and Richmond,M. C. (1982b). Meteorological evaluation of the proposed Palmer to
26  Glennallen transmission line route, Meteorology Research, Inc., Altadena, CA.

27 
 
 1  Govoni, J. W. (1990). “A comparative study of icing rates in the White Mountains of New
2  Hampshire, Paper A1-9.” Proc., 5th Intl. Workshop on Atmospheric Icing of Structures, Tokyo.

3  Hall, E. K. (1977). “Ice and wind loading analysis of Bonneville Power Administration’s
4  transmission lines and test spans.” IEEE Power Engineering Society Summer Meeting, July 20,
5  Mexico City.

6  HoskingsHosking, J. R. M., and Wallis, J. R. (1987). “Parameter and quantile estimation for the

at
e LY
7  generalized Pareto distribution.” Technometrics, 29(3), 339–349.

rm
8  Hosking, J. R. M. and Wallis J. R. (1997). Regional frequency analysis: An approach based on

lin N
9  L-moments, Cambridge University Press, New York.

Fo
er n O
10  IEEE. (202217). National Electrical Safety Code, C2-202217, The Institute of Electrical and
11  Electronics Engineers, Inc., New York, NY.
nd o
12  Iversen, E.C. Thompson, G., Nygaard, B. E., and Lacavalla, M. (2019). “Improved prediction of
U ati
13  wet snow”, Proc., 18th Intl. Workshop on Atmospheric Icing of Structures, 23-26 June, Reykjavik,
14  Iceland.
t / rm

15  Jones, K. (1998). “A simple model for freezing rain loads.” Atmos. Res., 46, 87–97.
ou fo

16  Jones, K. F. (1990). “The density of natural ice accretions related to nondimensional icing
e In

17  parameters.” Q. J. Royal Meteorolog. Soc., 116, 477–496.

r

18  Jones, K. F. (1996). “Ice accretion in freezing rain” U.S. Army CRREL Report No. 96-2. Cold
Fo

19  Regions Research and Engineering Laboratory, Hanover, NH.

20  Jones, K. F. (2010). Evaluation of extreme ice loads from freezing rain for Newfoundland and
Labrador Hydro, Final Report, Muskrat Falls Project Exhibit 96. Terrestrial and Cryospheric
rik

21 
22  Sciences Branch, Cold Regions Research and Engineering Laboratory, Hanover, NH.
St

23  Jones, K. F. and Peabody, A. B. (2006). “The application of a uniform radial ice thickness to
24  structural sections.” Cold Reg. Sci. Tech., 44(2), 145–148.

28 
 
 1  Jones, K. F., Thorkildson, R., and Lott, J. N. (2002). “The development of the map of extreme ice
2  loads for ASCE Manual 74.” Electrical Transmission in a New Age. D. E. Jackman, ed., ASCE,
3  Reston, VA, 9–31.

4  Jones, K. F. (2019). “How was the MRI model applied?”. Proc., 18th Intl. Workshop on
5  Atmospheric Icing of Structures, 23-28 June, Reykjavik, Iceland.,

6  Jones, K. F. and Eylander, J. B. (2017). “Vertical variation of ice loads from freezing rain” Cold

at
e LY
7  Reg. Sci. Tech., 143, 126–136.

rm
8  Jones, K. F., Thorkildson, R., and Eylander J. B. (2017). “Median volume drop diameter of in-

lin N
9  cloud icing simulations”, Proc., 17th Intl. Workshop on Atmospheric Icing of Structures, 25-29

Fo
er n O
10  Sep, Chongqing, China, .

11  Konrad, C. E. II (1998). “An empirical approach for delineating fine scaled spatial patterns of
freezing rain in the Appalachian region of the USA”, Climate Research, 10, 217-227.
nd o
12 
U ati
13  Kuroiwa, D. (1962). “A study of ice sintering.” U.S. Army CRREL, Research Report No. 86,
14  Hanover, NH.
t / rm

15  Kuroiwa, D. (1965). “Icing and snow accretion on electric wires.” U.S. Army CRREL Research
ou fo

16  Paper 123, Hanover, NH.

e In

17  Lacavalla, M., Marcacci, P., and Freddo, A. (2015). “Wet snow research activity in Italy”, Proc.,
18  16th Intl. Workshop on Atmospheric Icing of Structures, 28 June-3 July, Uppsala, Sweden.
r
Fo

19  Lacavalla, M., Sperati, S., Bonanno, R., and Marcacci, P. (2019). “A revision of wet snow load
20  map for the Italian power lines with a new high resolution reanalysis dataset”, Proc., 18th Intl.
21  Workshop on Atmospheric Icing of Structures, 23-26 June, Reykjavik, Iceland.
rik

22  Lott, N. (1993). NCDC Technical Report Nos. 93-01 and 93-03, National Climatic Data Center,
St

23  Asheville, NC.

24  Macklin, W. C. (1962). “The density and structure of ice formed by accretion.” Q. J. Royal
25  Meteorolog. Soc., 88, 30–50.

29 
 
 1  Mallory, J. H., and Leavengood, D. C. (1983a). “Extreme glaze and rime ice loads in Southern
2  California: Part I—Rime.” Proc., 1st Intl. Workshop on Atmospheric Icing of Structures, L. D.
3  Minsk, ed., U.S. Army CRREL Special Report No. 83-17. Hanover, NH, 299–308.

4  Mallory, J. H., and Leavengood, D. C. (1983b). “Extreme glaze and rime ice loads in Southern
5  California: Part II—Glaze.” Proc., 1st Intl. Workshop on Atmospheric Icing of Structures, L. D.
6  Minsk, ed., U.S. Army CRREL Special Report No. 83-17. Hanover, NH, 309–318.

at
e LY
7  Matsumiya, H. I., Aso, T., Shugo, M., Nishihara, T., Shimizu, M., and Sugimot, S. (2019). “Field

rm
8  Observation of Wet Snow Accretion and Galloping on a Single Conductor Transmission Line”,

lin N
9  Proc., 18th Intl. Workshop on Atmospheric Icing of Structures, 23-26 June, Reykjavik, Iceland.

Fo
er n O
10  McCormick, T., and Pohlman, J. C. (1993). “Study of compact 220 kV line system indicates need
11  for micro-scale meteorological information.” Proc., 6th Intl. Workshop on Atmospheric Icing of
12  Structures. Budapest, Hungary. 155–159.
nd o
U ati
13  Mesinger, F., Dimego, G., Kalnay, E., Mitchell, K., Shafran, P.C., Ebisuzaki, W., et al. (2006).
14  “North American Regional Reanalysis”, Bulletin of the American Meteorological Society, 87(3),
t / rm

15  343-360, http://dx.doi.org/10.1175/BAMS-87-3-343.

ou fo

16  Mozer, J. D., and West, R. J. (1983). “Analysis of 500 kV tower failures.” Meeting of the
17  Pennsylvania Electric Association, Harrisburg, PA.
e In

18  Mulherin, N. D. (1996). “Atmospheric icing and tower collapse in the United States.” 7th Intl.
r

19  Workshop on Atmospheric Icing of Structures. M. Farzaneh and J. Laflamme, eds., June 3–6,
Fo

20  Chicoutimi, Quebec.

21  National Centers for Environmental Prediction/National Weather Service/NOAA/U.S.

Department of Commerce. 2005, updated monthly. NCEP North American Regional Reanalysis
rik

22 
23  (NARR). Research Data Archive at the National Center for Atmospheric Research, Computational
St

24  and Information Systems Laboratory. http://rda.ucar.edu/datasets/ds608.0/. Accessed 23 Nov

25  2018.

26  National Oceanic and Atmospheric Administration (NOAA). (1959–Present). Storm data,
27  National Oceanic and Atmospheric Administration, Washington, DC.

30 
 
 1  Nebraska Public Power District (NPPD). (1976). The storm of March 29, 1976, Public Relations
2  Department, Columbus, NE.

3  Nikolov, D., and Makkonen, L. (2015). “Testing six wet snow models by 30 years of observations
4  in Bulgaria”, Proc., 16th Intl. Workshop on Atmospheric Icing of Structures, 28 June-3 July,
5  Uppsala, Sweden.

NOAA (National Oceanic and Atmospheric Administration. (1959–Present). Storm data, National

at
6 

e LY
7  Oceanic and Atmospheric Administration, Washington, DC.

rm
lin N
8  Nygaard, B. E., Ágústsson, H., and Somfalvi-Tóth, K. (2013). “A study on the sticking efficiency

Fo
9  of wet snow using 50 years of observations”, Proc., 15th Intl. Workshop on Atmospheric Icing of

er n O
10  Structures, 8-11 Sep, St. Johns, Canada.

11  Nygaard, B. E., Seierstad, I. A., Fikke, S. M., Horsman, D., and Wareing, J. B. (2013). “The
nd o
12  Development of New Maps for Design Ice Loads for Great Britain”, Proc., 15th Intl. Workshop on
U ati
13  Atmospheric Icing of Structures, 8-11 Sep, St. Johns, Canada.
t / rm

14  Peabody, A. B. (1993). “Snow loads on transmission and distribution lines in Alaska.” Proc., 6th
15  Intl. Workshop on Atmospheric Icing of Structures. Budapest, Hungary. 201–205.
ou fo

16  Peabody, A. B. (1996). “Measurements of ice and snow loads on the Tyee Lake 138 kV line”,
e In

17  Proc., 7th Intl. Workshop on Atmospheric Icing of Structures, June 3-7, Chicoutimi, Canada.
r

18  Peabody, A. B., and Jones, K. F. (2002). “Effect of wind on the variation of ice thickness from
Fo

19  freezing rain.” Proc., 10th Intl. Workshop on Atmospheric Icing of Structures. June 17–20, Brno,
20  Czech Republic.

Peabody, A. B., and Wyman, G. (2005). “Atmospheric icing measurements in Fairbanks, Alaska.”
rik

21 
22  Proc., 11th Intl. Workshop on Atmospheric Icing of Structures, Masoud Farzaneh and Anand P.
St

23  Goel, eds., Oct. 9–13, Yokohama, Japan.

24  Peterka, J. A. (1992). “Improved extreme wind prediction for the United States.” J. Wind Engrg.
25  Indust. Aerodyn., 41–44, 533–541.

31 
 
 1  Peterka, J. A., Finstad, K., and Pandy, A. K. (1996). Snow and wind loads for Tyee transmission
2  line. Cermak Peterka Petersen, Fort Collins, CO.

3  Power, B. A. (1983). “Estimation of climatic loads for transmission line design.” CEA No. ST 198.
4  Canadian Electric Association, Montreal.

5  Powers, J. G., Klemp, J.B., Skamarock, W.C., Davis, C. A., Dudhia, J., Gill, D. O., et al. (2017).
6  “The Weather Research and Forecasting Model: Overview, System Efforts, and Future

at
e LY
7  Directions”, Bull. Am. Met. Soc., 1717-1737.

rm
8  Rawlins, C. B. (1979). “Galloping conductors.” Transmission line reference book, wind-induced

lin N
9  conductor motion, prepared by Gilbert/Commonwealth. Electric Power Research Institute, Palo

Fo
er n O
10  Alto, CA, 113–168.

11  Richmond, M. C. (1985). Meteorological evaluation of Bradley Lake hydroelectric project 115kV
transmission line route, Richmond Meteorological Consultant, Torrance, CA.
nd o
12 
U ati
13  Richmond,C. (1991). Meteorological evaluation of Tyee Lake hydroelectric project transmission
14  line route, Wrangell to Petersburg, Richmond Meteorological Consulting, Torrance, CA.
t / rm

15  Richmond, M. C. (1992). Meteorological evaluation of Tyee Lake hydroelectric project

ou fo

16  transmission line route, Tyee power plant to Wrangell, Richmond Meteorological Consulting,
17  Torrance, CA.
e In

18  Richmond, M. C., Gouze, S. C., and Anderson, R. S. (1977). Pacific Northwest icing study,
r

19  Meteorology Research, Altadena, CA.

Fo

20  Robbins, C. C., and Cortinas, J. V., Jr. (1996). “A climatology of freezing rain in the contiguous
21  United States: Preliminary results.” Preprints, 15th AMS Conference on Weather Analysis and
rik

22  Forecasting, Norfolk, VA, Aug. 19–23.

23  Ryerson, C. (1987). “Rime meteorology in the Green Mountains.” U.S. Army CRREL Report No.
St

24  87-1. Cold Regions Research and Engineering Laboratory, Hanover, NH.

25  Ryerson, C. (1988a). “Atmospheric icing climatologies of two New England mountains.” J. Appl.
26  Meteorol., 27(11), 1261–1281.

32 
 
 1  Ryerson, C. (1988b). “New England mountain icing climatology.” CRREL Report No. 88-12. Cold
2  Regions Research and Engineering Laboratory, Hanover, NH.

3  Ryerson, C. (1990). “Atmospheric icing rates with elevation on northern New England mountains,
4  U.S.A.” Arctic Alpine Res., 22(1), 90–97.

5  Ryerson, C., and Claffey, K. (1991). “High latitude, West Coast mountaintop icing.” Proc.,
6  Eastern Snow Conference, Guelph, ON, 221–232.

at
e LY
7  Ryerson, C. and Kenyon, P. (1996). “Ice ablation processes on Mt. Equinox, Vermont, USA”,

rm
8  Proc., 7th Intl. Workshop on Atmospheric Icing of Structures, Chicoutimi, Canada, June 3-7, 277-

lin N
9  281.

Fo
er n O
10  Sakamoto, Y., Mizushima, K., and Kawanishi, S. (1990). “Dry snow type accretion on overhead
11  wires: Growing mechanism, meteorological conditions under which it occurs and effect on power
lines.” Proc., 5th InternationalIntl. Workshop on Atmospheric Icing of Structures, Tokyo, Paper
nd o
12 
5–9.
U ati
13 

14  Shan, L. (1997). “Wind tunnel study of drag coefficients of single and bundled conductors.” EPRI
t / rm

15  TR-108969. Electric Power Research Institute, Palo Alto, CA.

ou fo

16  Shan, L., and Marr, L. (1996). Ice storm data base and ice severity maps, Electric Power Research
17  Institute, Palo Alto, CA.
e In

18  Simiu, E., and Scanlan, R. H. (1996). Wind effects on structures: Fundamentals and applications
r

19  to design. John Wiley & Sons, New York.

Fo

20  Sinclair, R. E., and Thorkildson, R. M. (1980). “In-cloud moisture droplet impingement angles
21  and track clearances at the Moro UHV test site.” BPA 1200 kV Project Report No. ME-80-7.
rik

22  Bonneville Power Administration, Portland, OR.

23  Tattelman, P., and Gringorten, I. (1973). “Estimated glaze ice and wind loads at the earth’s surface
St

24  for the contiguous United States.” Report AFCRLTR-73-0646. U.S. Air Force Cambridge
25  Research Laboratories, Bedford, MA.

33 
 
 1  Thorkildson, R. M., Jones, K. F. and Emery, M. K. (2009). “In-cloud icing in the Columbia Basin”,
2  Monthly Weather Review, 137, 4369-4381.

3  Thorkildson, R. M. (2019). “Rime Ice Occurrences from Radiation Fog that Impact Overhead
4  Transmission Lines in Central Washington State”. Proc. 18th Intl. Workshop on Atmospheric Icing
5  of Structures, 23-26 June, Reykjavik, Iceland.

6  TIA. (2019) Structural Standard for Antenna Supporting Structures, Antennas and Small Wind

at
e LY
7  Turbine Support Structures, TIA 222-H-1, Arlington VA.

rm
8  Ueno, K., Eguchi, Y., Nishihara, T., Sugimoto, S., and Matsumiya, H. (2015). Development of

lin N
9  snow accretion simulation method for electric wires in consideration of snow melting and

Fo
er n O
10  shedding, Proc., 16th Intl. Workshop on Atmospheric Icing of Structures, 28 June-3 July, Uppsala,
11  Sweden. nd o
12  USFS (U.S. Forest Service (USFS). (1994). Forest Service handbook FSH6609.14,
U ati
13  Telecommunications handbook, R3 Supplement 6609.14-94-2. USFS, Washington, DC.
t / rm

14  Wang, Q. J. (1991). “The POT model described by the generalized Pareto distribution with Poisson
15  arrival rate.” J. Hydrol., 129, 263–280.
ou fo

16  White, H. B. (1999). “Galloping of ice covered wires.” Proc., 10th International Conf. on Cold
e In

17  Regions Engineering, Hanover, NH, 799–804.

18  Winkleman, P. F. (1974). Investigation of ice and wind loads: Galloping, vibrations and
r
Fo

19  subconductor oscillations, Bonneville Power Administration, Portland, OR.

20  Wylie, W. G. (1958). “Tropical ice storms—Winter invades Hawaii.” Weatherwise (June), 84–90.

Young, W. R. (1978). “Freezing precipitation in the Southeastern United States.” M.S. Thesis,
rik

21 
22  Texas A&M University, College Station, TX.
St

34