Electrical Properties of Snow
M. Farzaneh, I. Fofana and H. Hemmatjou
NSERC / Hydro-Quebec / UQAC Industrial Chair on Atmospheric Icing of Power Network Equipment (CIGELE)
and Canada Research Chair on Engineering of Power Network Atmospheric Icing (INGIVRE) at Université du
Québec à Chicoutimi, Quebec, Canada, G7H 2B1
Abstract: This paper reviews the parameters that affect
the electrical properties of snow. All the parameters
listed are important, but some have a special merit for
the characterization of the electrical properties of snow.
The most important are volume conductivity and
density, followed by the liquid water content and
conductivity of water melted from snow. The dc
conductivity and liquid water content as a function of
snow density and temperature were measured and the
relationships between them were determined.
When designing insulation coordination of transmission
lines and substations, emphasis is placed on the
environmental and meteorological conditions of the
routes or sites. The insulation design for a contaminated
area is based on the contamination withstand-voltage
characteristics under the operating voltage. In Nordic
regions, yearly snowfall is between 70 and 300 inches
[1-3]. Besides mechanical damages, the accretion of ice
and wet snow on outdoor insulators causes a
considerable reduction in their electrical performance
[1-3]. Especially in heavy snowfall areas, sometimes
wet snow and superimposed contamination cause
flashover faults on insulators [2, 4-7]. Actually,
flashover faults of snow-covered insulator strings, due
to switching surges or ac operating voltage, have been
experienced in Japan, Canada, and other cold-climate
countries [2, 4-7]. Such problems generally appear on
tension insulator assemblies, and rarely on suspension
assemblies, since tension assemblies are more prone to
large amounts of snow accumulation . In order to
understand the flashover processes of snow-covered
insulators, and to establish design criteria for outdoor
insulators, the electrical characteristics of natural snow
should first be investigated.
Although the determination of those parameters is
amongst the most important, to the best of our
knowledge, very few studies have been carried out on
this aspect. Therefore, this paper reviews the major
parameters affecting the electrical properties of snow,
such as temperature, contamination, volume
conductivity and density, liquid water content, and
conductivity of water melted from snow . The liquid
water content and dc conductivity were measured as a
function of volume density and temperature for a large
number of snow samples, in order to establish whether
any relation exists between them.
Review of parameters affecting electrical
properties of snow
Dry snow behaves like an insulating material made of
air and ice , that is to say, it is a two-component
system. Considerable air is trapped in miniscule pockets
among the individual ice crystals comprising the snow
cover. When wet, the snow becomes a three-component
system composed of air, ice, and water. Wet snow is
treated as a three-phase mixture, considering the ice and
water particles as inclusions embedded in air, which is
the background material.
There is a fundamental difference between wet and dry
snow since liquid water causes major reconfigurations
of both grains and bonds [8, 9]. Within the wet and dry
snow categories there are also two important divisions:
wet snow at low and high liquid contents, and dry snow
at low and high growth rates. Wet snow is less cohesive
and slushy at high liquid contents because the grain
boundaries are unstable against pressure melting .
However, wet snow is well-bonded at low liquid
contents, where ice-bonded clusters form. A transitional
form of snow, melt–freeze grains, can be either wet or
dry. These amorphous, multi crystalline particles arise
from melt–freeze cycles. They are solid within and well-
bonded to their neighbors.
The electrical properties of snow are altered according
to the content of ice within it, that is, according to its
density. The parameters characterising the electrical
properties of snow are its volume conductivity,
dielectric constant, salt content, water content, volume
density, crystal structure and size, impurities, as well as
the electric field and frequency to which it is submitted
. These properties also depend on geometric factors
and contact of the electrodes with snow. However, it has
been acknowledged that among these, the volume
conductivity is the major parameter which affects the
insulation characteristics of snow . The volume
conductivity is largely affected by its volume, density,
and salt content , and thus, chemical analysis of snow
collected from various places has shown that the
conductivity of water melted from snow is due mainly
to NaCl . The salt content of snow is expressed by
the conductivity of water melted from snow, which is
corrected to the value at 20°C.
Results and discussions
For all the measurements carried out and reported here,
samples of natural snow stored in a freezer at –20°C
after collection were used.
Liquid water content: The free water content is
defined as the ratio of the amount of water contained to
the total weight of the wet snow. The free water content,
LWC, is determined by separating the water from the
snow, by putting the wet snow into a graduated test-tube
(Figure 1) in a motor-driven centrifuge with a rotation
speed of 1,700 rpm. The time of rotation was half a
minute, to prevent the snow from melting during the
Two test-tubes containing filters were especially
designed for this purpose. These filters separate dry
snow from the liquid water contained in the snow
sample, as depicted in Fig. 1. Three wires are attached
to the filter, which makes it possible to remove it from
the tubes after each test. The amount water separated
from the snow is then measured using the graduated
Figure 1: Test-tube before and after the utilisation of the centrifuge.
After each water-percentage test, the test-tubes are
cleaned and dried, in order to avoid any deposit that
could influence further measurements. After drying, the
test-tubes are kept in a cold chamber at the same
temperature as the snow sample in order to avoid further
melting during centrifuge measurements. The
temperature inside each snow sample is measured
before and after the test using thermocouples, to
determine an average value.
The weight of the water (MW) extracted from the snow
sample is then calculated:
where ρ is the density of water (in g/cm3) separated
from the snow sample, and υ indicates the water volume
contained in the snow. Water content in the snow
sample is then expressed in percentage of the total mass
of MT, ie. the mass of ice crystals and water:
Figure 2 depicts the results of some tests performed
using snow sample having a density equal to 0.36 kg/m3
with the measured conductivity of water melted from
snow about to 83 µS/cm.
Figure 2: liquid water content according to the temperature.
Clearly, the liquid water content (LWC), determined
either during heating or cooling of snow samples,
increases with temperature. A mathematical approach to
data suggests the existence of some kind of relationship
between the LWC and the snow sample temperature
(TS). Indeed, the LWC can be described as a
mathematical function of the temperature TS represented
by the solid-line curves, fitting the experimental data
plotted in Fig. 2:
, for the heating process (3)
and 5.170T662.0LWC S
for cooling, (4)
where TS is expressed in Kelvin (K).
DC conductivity of Snow: For this purpose, a snow
sample was packed into a capacitive cell with guard
rings so as to obtain a desired volume density.
Since the liquid water content is defined as the ratio of
the amount of water contained in snow to the total
weight of the snow sample, LWC suffers no alteration
from packing. The capacitive cell used had a volume of
about 190 cm3. Figure 3 shows the experimental set-up
used for measuring snow sample conductivity.
For the measurements, the capacitive cell was placed
inside a small cold chamber, type ‘Envirotronics EH40-
Wet snow having a
total mass of WT Dry
Volume of water content in
the snow sample (υ)
After using the
2-3’ that had been pre-cooled to the desired temperature
TS. The air temperature accuracy, ± 1.1°C, was ensured
by a microprocessor-based temperature-humidity
programmer controller. Temperatures inside both the
cold room and the snow sample were also recorded
during the experiment.
Figure 3: Experimental set-up used to determine the dc conductivity
of the natural snow samples.
The temperature setting of the capacitive cell was
changed every 15 minutes in both the heating and
cooling processes, and the conductivity was measured
just before the temperature changed to a new value.
The snow sample conductivity σS was calculated from
the direct current strength flowing through a sample 2.4
cm in length and 78.5 cm2 in section, and submitted to
100 Vdc (average field 41.25 V/cm). It should be noted
that snow volume conductivity is independent of the
applied electric field strength . The current is
sampled at a rate of 1,200/s, transferred to a data buffer
and stored. The main components of the data acquisition
system are a National Instrument DAQ plug-in board in
a PC, and LabVIEWTM application software. The data is
stored as both ASCII text files and binary files. This
ensures that the LabVIEWTM data could also be
analyzed further using other software applications like
The resistance, RS, of the snow sample in this circuit is
calculated using the following basic relation:
and so to speak the conductivity of the snow sample.
In addition to the conductivity, snow sample densities
Also, the chemical impurities of the snow sample,
including NaCl, were quantified by measuring the
conductivity of the water melted from this snow sample.
Snow conductivity is well known to depend not only on
snow temperature, but also on the density of the sample.
Some of the results, as a function of snow temperature
for different densities of natural snow, collected in
various places, for both the cooling and heating
processes, are shown in Figures 4 and 5. The results
show that the higher the density, the higher the
conductivity of the snow sample.
Figure 4: Effect of the density on the dc conductivity of snow in the
Figure 5: Effect of the density on the dc conductivity of snow in the
By compressing snow, one breaks its structure, and
increases the density and the surface of contacts
between crystals. The resistance of the contacts to the
passage of current influences the sample conductivity.
Indeed, compression generates an increase in activation
energy and, it follows, an increase in conductivity .
The conductivity values from -12°C to around the
temperature corresponding to its peak value can be
explained with data in Figure 3. Indeed, liquid water
content increases as temperature of the snow sample
increases. The presence of water in liquid phase, which
increases with temperature, causes the electric
conductivity to increase. Pure water is a poor conductor
of electricity. It is the impurities in water, such as
dissolved salts, that enable water to conduct electricity.
The higher is the amount of impurities, the higher is the
electrical conductivity. Thus, as the quantity of
conductive liquid water increases, so does conductivity.
Figures 4 and 5 show that conductivity increases as the
temperature increases, from -12°C to around the
temperature corresponding to its peak value, and then
decreases as the temperature increases from peak
conductivity to the melting temperature. This is a
curious behaviour because dc conductivity for ice
generally increases with increasing temperature .
This apparently curious behaviour already observed by
Takei et al. for low frequency  seems to indicate
important changes of the texture of snow near the
Peak conductivities around -2°C and -4°C are
respectively observed in the heating and cooling
processes. This indicates that conduction mechanisms
during heating and cooling processes may be different.
Figure 6 shows the temperature dependence of the dc
conductivity for a snow sample having a density of 0.42
g/m3 and the measured conductivity of water melted
from snow equal to 41.2 µS/cm which undergone
alternatively heating and cooling processes.
Figure 6: Temperature dependence of the dc conductivity for snow in
the heating and cooling process sequences.
During the first heating process from -12°C to 0°C
(sequence 1), part of the snow sample melts and
transforms to ice when cooled down to -12°C (sequence
2). When heated again from -12°C to 0°C (sequence 3),
a reduction in the dc conductivity is noticed. During the
fourth sequence, the conductivity is also found to be
lower than values obtained during the second sequence.
Clearly, reorganisation of water molecules changes the
microstructure of the snow sample, affecting
considerably the properties of the latter.
The parameters that affect the electrical properties of
natural snow have been considered. The liquid water
content and dc conductivity were measured as a
function of density and temperature for a large number
of snow samples, and relationships were established
between them. An apparently curious behaviour was
observed in the temperature dependence of the dc
conductivity of snow near the melting temperature. One
of the most important consequences of such a curious
behaviour for outdoor insulation coordination could
indicate that temperatures corresponding to peak values
of snow conductivity in both cooling and heating
processes constitute a major factor determining the
flashover performance of snow-covered insulators. The
authors are engaged in a study to further investigate this
curious behaviour, and they expect to publish their
results in the near future.
 CIGRE Task Force 33.04.09, "Influence of ice and snow on the
flashover performance of outdoor insulators, part I: Effects of
Ice", Electra, No. 187, pp. 91-111, Dec. 1999.
 CIGRE task force 33.04.09,"influence of ice and snow on the
flashover performance of outdoor insulators", part II: effects of
snow, Electra No. 188 February 2000.
 M. Farzaneh, "Ice accretion on high voltage conductors and
insulators and related phenomena", Phil. Trans. of the Royal Soc.
vol. 358, No. 1776, pp. 2971-3005, 2000.
 Y. Higashiyama, T. Sugimoto, K. Asano, M. Johsho and S.
Tachizaki, "Electrical breakdown of heavily polluted capped
snow on insulators strings", The 8th Int. Workshop on
Atmospheric Icing of Structures (IWAIS), pp. 199-203, 1998.
 M. Yasui, K. Naito, Y. Hasegawa, "AC withstand voltage
characteristics of insulator strings covered with snow", IEEE
Trans. on Power Delivery, Vol. 3, No. 2, pp. 828-838, 1988.
 Y. Watanabe, "Flashover tests of insulators covered with ice or
snow", IEEE Trans. on Power Apparatus and Systems, Vol.
PAS-97, No. 5, pp. 1788-1794, 1978.
 H. Matsuda, H. Komuro and K. Takasu, "Withstand voltage
characteristics of insulator string covered with snow or ice",
IEEE Trans. on Power Delivery, Vol. 6, No. 3, pp. 1243-1250,
 A. N. Arslan, H. Wang, J. Pulliainen and M. Hallikainen,
"Effective permittivity of wet snow using strong fluctuation
theory", Progress In Electromagnetics Research, PIER 31,
 Samuel C. Colbeck, "A Review of Sintering in Seasonal Snow",
CRREL report 97-10, December 1997.
 P.M. Kopp, "Conductivité électrique de la neige en courant
continu", Zeitschrift fuer angewandete Mathematik und Physik,
Vol. 13, pp. 431-441, 1962.
 Z. Yoshida, H. Oura, D. Kuroiwa, T. Huzioka, K. Kojima and S.
Kinosita, "Physical studies on deposited snow: dielectric
properties", Contrib. Inst. Low Temp. Sci., Vol. 14, pp. 1-33,
 I. Takei and N. Maeno, "The low frequency conductivity of
snow near the melting temperature", Annals of glaciology, Vol.
32, pp. 14-18, 2001.
Author address: Prof. M. Farzaneh, NSERC/Hydro-Québec/UQAC
Industrial Chair on Atmospheric Icing of Power Network Equipment
and Canada Research Chair on Engineering of Power Network
Atmospheric Icing (INGIVRE), Université du Québec à Chicoutimi
555, Boulevard de l’Université, Chicoutimi, Québec, Canada, G7H
2B1, E-mail: email@example.com