Veröffentlichung MeteoSchweiz Nr.

75

COST 727: Atmospheric Icing on Structures
Measurements and data collection on icing:
State of the Art

S. Fikke, G. Ronsten, A. Heimo, S. Kunz, M. Ostrozlik, P.-E. Persson, J. Sabata, B. Wareing, B. Wichura, J. Chum,
T. Laakso, K. Säntti, L. Makkonen








COST is
supported by
the EU RTD
Framework
Programme
ESF provides
the COST
Office through
an EC contract

Herausgeber
Bundesamt für Meteorologie und Klimatologie, MeteoSchweiz, © 2007

MeteoSchweiz
Krähbühlstrasse 58
CH-8044 Zürich
T +41 44 256 91 11
www.meteoschweiz.ch
Weitere Standorte
CH-8058 Zürich-Flughafen
CH-6605 Locarno Monti
CH-1211 Genève 2
CH-1530 Payerne


Veröffentlichung MeteoSchweiz Nr. 75
ISSN: 1422-1381

COST 727: Atmospheric Icing on Structures
Measurements and data collection on icing:
State of the Art


S. Fikke (Chair, Norway)
G. Ronsten (Sweden)
A. Heimo (Switzerland)
S. Kunz (Switzerland)
M. Ostrozlik (Slovakia)
P.-E. Persson (Sweden)
J. Sabata (Czech Rep.)
B. Wareing (United Kingdom)
B. Wichura (Germany)
J. Chum (Czech Rep.)
T. Laakso (Finland)
K. Säntti (Finland)
L. Makkonen (Finland)





Bitte zitieren Sie diese Veröffentlichung folgendermassen:
Please quote this publication in the following way:

COST-727, Atmospheric Icing on Structures: 2006,
Measurements and data collection on icing: State of the Art
Publication of MeteoSwiss, 75, 110 pp.




2/110
3/110
Foreword
COST Action 727 “Measuring and forecasting atmospheric icing on structures” was estab-
lished in April 2004 and comprises 12 signatory countries: Austria, Bulgaria, the Czech Re-
public, Finland, Germany, Hungary, Norway, Slovakia, Spain, Sweden, Switzerland and the
United Kingdom. Following the “Memorandum of Understanding” (MoU), three working
groups were established, WG1 “Icing modelling”, WG2 “Measurements and data collection
on icing” and WG3 “Mapping and forecasting of atmospheric icing”.
The present report covers the work of WG2 during Phase 1 of the Action. The main scope of
this phase was to create an inventory of earlier and current activities on icing measurements,
data resources and instrument testing. The emphasis is on activities within the signatory coun-
tries, however some additional information from other countries like Russia and Canada is
included as well.
It is important to notice that COST does not support project activities. Therefore all contribu-
tions concerning individual countries are provided according to available time and engage-
ments of the participants. Hence the structure and details of each contribution will vary, and
the reader will not necessarily find the same information for all countries. A lot of references
are given, however, and the reader will find links to institutions where further information can
be retrieved.
It is the intention of Phase 2 to structure and update information from existing test sites and
open data sources in a more systematic way than was possible in this report. Phase 2 will also
include instrument comparisons from test sites, and also elaborate recommendations for
WMO observations and permanent data bases for icing in Europe.
COST Action 727 acknowledges Dr Wiel M. F. Wauben, the Royal Netherlands Meteorologi-
cal Institute (KNMI) for reviewing this report and MeteoSwiss for their generous offer to
print the report as part of their series of internal reports.
Svein M. Fikke
Chairman of Working Group 2
4/110
Table of contents
Foreword............................................................................................................................. 3
1 Management summary ....................................................................................................... 7
2 Introduction ........................................................................................................................ 9
2.1 Memorandum of Understanding ................................................................................ 9
2.2 Interface with WG1 and WG3 ................................................................................. 10
2.3 Past and present activities ........................................................................................ 10
3 Definitions and meteorological conditions ...................................................................... 13
3.1 Generic definition..................................................................................................... 13
3.2 Icing types (extracts from ISO-12494)..................................................................... 13
4 Ice measurements as described in standards .................................................................... 17
4.1 International Standardization Organization (ISO) ................................................... 17
4.2 International Electrotechnical Commission (IEC) ................................................... 17
4.3 World Meteorological Organization (WMO) .......................................................... 18
4.4 Icing and wind turbines ............................................................................................ 19
5 Meteorological measurements under icing conditions..................................................... 20
5.1 Introduction .............................................................................................................. 20
5.2 WMO/CIMO Recommendations ............................................................................. 20
5.3 Definitions................................................................................................................ 21
5.4 Site effects ................................................................................................................ 23
5.5 Site Icing Index ........................................................................................................ 23
5.6 Measurements under icing conditions...................................................................... 24
6 Examples of existing icing data ....................................................................................... 28
6.1 Finland...................................................................................................................... 28
6.2 Germany................................................................................................................... 28
6.3 Slovak Republic ....................................................................................................... 28
6.4 Norway..................................................................................................................... 29
6.5 Czech Republic ........................................................................................................ 29
6.6 UK............................................................................................................................ 30
6.7 Sweden ..................................................................................................................... 30
6.8 Bulgaria .................................................................................................................... 30
6.9 Hungary.................................................................................................................... 30
6.10 Russia ....................................................................................................................... 30
6.11 Canada...................................................................................................................... 30
6.12 WMO/CIMO inter-comparisons of wind instruments under harsh conditions....... 31
6.13 EUMETNET/SWS II project ................................................................................... 31
7 Requirements for ice detectors......................................................................................... 33
7.1 Concepts ................................................................................................................... 33
7.2 Siting of icing sensors .............................................................................................. 34
7.3 Guidance for selecting ice detectors......................................................................... 36
8 Availability, verification and requirements of ice detectors ............................................ 39
8.1 Available ice detectors ............................................................................................. 39
8.2 Data requirements for icing models ......................................................................... 40
8.3 Verifications of data................................................................................................. 40
9 Experiences with automatic instruments for ice measurements....................................... 41
9.1 ICEmeter .................................................................................................................. 41
9.2 Labko ice detectors................................................................................................... 41
9.3 Rosemount/BFGoodrich, model 0872J .................................................................... 41
5/110
9.4 EAG 200................................................................................................................... 42
9.5 Gerber....................................................................................................................... 42
9.6 METEO device......................................................................................................... 42
9.7 IceMonitor................................................................................................................ 42
9.8 T20-series Ice Detectors........................................................................................... 43
9.9 Instrumar IM101 ...................................................................................................... 43
10 Long term recommendations for ice measurements in Europe.................................... 44
10.1 Regional variability.................................................................................................. 44
10.2 Requirements for measuring sites ............................................................................ 44
10.3 Permanent forum for monitoring icing in Europe.................................................... 45
Annexes
1 Measurements in Finland ............................................................................................. 46
2 Measurements in Germany........................................................................................... 49
3 Measurements in Slovak Republic............................................................................... 51
4 Measurements in Norway............................................................................................. 52
5 Measurements in Czech Republic................................................................................ 55
6 Measurements in UK.................................................................................................... 59
7 Measurements in Sweden............................................................................................. 64
8 Measurements in Bulgaria............................................................................................ 66
9 Measurements in Hungary ........................................................................................... 69
10 Measurements in Russia............................................................................................... 71
11 Measurements in Canada and USA.............................................................................. 72
12 Icemeter (Czech Republic)........................................................................................... 73
13 Labko Ice detector (Finland) ........................................................................................ 77
14 Rosemont, BF Goodrich, model 0872J (Finland) ........................................................ 79
15 EAG 200 (Germany) .................................................................................................... 80
16 Gerber (USA) ............................................................................................................... 83
17 METEO device............................................................................................................. 89
18 IceMonitor .................................................................................................................... 92
19 T20-series Ice Detector (Sweden) ................................................................................ 96
20 Instrumar IM101 .......................................................................................................... 99
21 Wind tunnel calibration.............................................................................................. 100
22 References .................................................................................................................. 106
6/110
7/110
1 Management summary
The COST-727 Action ”Measuring and forecasting atmospheric icing on structures” was es-
tablished in April 2004. It is divided in 3 working groups dealing with modelling, measure-
ments and forecasting of icing.
Phase 1 of the action is dedicated to gathering available information for comprehensive state-
of-the-art reports with the following deliverables:
- Reports on the state-of-the-art
- Inventory of users' needs based on analyses
- Working plan for the Second Phase of the Action
Phase 2 of the Action is dedicated to R&D and will concentrate on research on in-cloud icing,
measurement on atmospheric icing, modelling of icing processes, improved forecasting sys-
tems, verification of existing icing sensors and mapping of icing occurrences and potentials in
Europe. The following deliverables will be expected:
- Scientific and technical publications on measurements and predictions of in-cloud ic-
ing
- Publications on verification of icing forecasts
- European icing map
- Recommendations for WMO observations and further work
.
The present paper deals with the result of WG2 concerning measurements of icing as well as
measurements performed under icing conditions. It contains information on:
Definition of icing: WG 2 recommends adopting the ISO12949 standard.
Past and present activities: International projects such as WMO/CIMO Instrument Inter-
comparison, EUMETNET SWS I and II projects, EU/WECO and NEW ICETOOLS projects
as well as entities such IEC/CENELEC, ISO, IWAIS are shortly presented.
Standards: Prevailing standards in use (ISO, IEC and WMO) dedicated to icing on structures
and icing measurements are shortly presented.
Measurements under icing conditions: As the WMO has presently no specific recommen-
dations for measurements performed in harsh conditions, e.g. icing, a set of recommendations
is presented concerning classification of sites and classification of sensors depending on se-
verity of icing and the site climatic environment.
Requirements and availability of ice detectors: It is shown that requirements on ice detec-
tors are dependent of the user’s requirement (wind energy, power lines, meteorology etc.) and
on the application. Installation procedures are presented, depending on users requirements
together with validation and verification processes.
Examples of existing data and experiences with existing ice detectors: A number of avail-
able long term experiments are presented concerning icing measurements and characterization
of icing sensors. These activities have taken place in numerous countries like Finland, Ger-
8/110
many, Slovak Rep., Norway, Czech Rep., UK, and indirectly from France, Switzerland, Swe-
den, Bulgaria, Russia, Canada, etc.
Recommendations for future activities: The establishment of test centres within the COST-
727 Action (Phase 2) have a temporary character. It is recommended that long-term interna-
tional calibration stations are established with a sufficient financial support for continuous
operation. These calibration centres are to be recognized for delivering approved certificates
for icing detectors and ice-free sensors.
9/110
2 Introduction
2.1 Memorandum of Understanding
COST Action 727 "MEASURING AND FORECASTING ATMOSPHERIC ICING ON
STRUCTURES" was established in April 2004 according to a Memorandum of Understand-
ing (MoU) [1]. The present report was prepared by Working Group 2 WG2 “MEASURE-
MENTS AND DATA COLLECTION ON ICING” and was given the following objectives:
“Measurements over a specific period of time on ice accretion and testing of icing sensors
will be based on existing test sites in the far north (Luosto/Finland) and in the Alpine region
(Guetsch/Switzerland). Additional experimental data from other ongoing activities will be
used for this Action.
WG2's activities will be dedicated to the following activities:
a) create an inventory and collect available experimental data on icing as well as ancil-
lary data
b) review and assess existing ice detectors and their performance
c) review and assess existing verification data from different sources
d) contribute to the set up of icing measurements at different locations in Europe and to
the development of existing test sites
e) set up a data quality control scheme for measured icing data
f) establish a basic data set for icing modeling and verification
g) provide recommendations to set up a long-term icing measuring network and data base
(to be submitted to WMO)
h) establish an icing monitoring core group for collecting and maintaining data on icing
during and especially after the course of the Action
i) develop the scientific and technical bases of specifications of ice detectors
j) set up recommendations for testing/approving ice detectors and ice/free sensors.”
The present report summarizes the information and material WG2 has collected for the Pre-
paratory Phase (Phase 1) of the Action, where the focus is to establish the “state of the art” in
the field of icing and to indicate the data available for icing in Europe. Some information from
Japan, Canada and Russia is also presented. The MoU focuses mainly on in-cloud icing, but
wet snow and freezing rain are included when appropriate.
This report is based on input from WG members according to their current activities and re-
lated references. Furthermore it includes information on measuring activities in Europe re-
lated to both wind turbines and electric overhead power lines.
10/110
2.2 Interface with WG1 and WG3
During Phase 1 the Action WG2 had a close collaboration with WG1 and WG3 which were
entitled:
- WG1 “ICING MODELLING”,
- WG3 “MAPPING AND FORECASTING OF ATMOSPHERIC ICING”.
These WGs are merged for Phase 2 and renamed WG1.
In particular, WG1 requires knowledge of what kind of icing data are available and can be
provided for the purpose of validating and calibrating icing models as well as meteorological
data that are unaffected by icing. WG3 needs similar data for mapping icing climates in
Europe, as well as information on measurement networks that can be incorporated in forecast-
ing routines by National Weather Services.
2.3 Past and present activities
This chapter is based on the input from different countries and covers both experimental work
and administrative activities.
2.3.1 WMO/CIMO Wind Instrument Intercomparison
Mt. Aigoual, France: 1992-1993
A documented experiment has been conducted at the Mt. Aigoual station, France (within a
joint venture between France and Switzerland) in order to analyze the performances of a
number of ice-free anemometers under extreme meteorological conditions. [2]. See section
6.10 for more details regarding the available data.
2.3.2 EUMETNET / SWS I&II
SWS I: 1997-1998
The EUMETNET launched a study of severe weather sensors (SWS) to summarize the ex-
periences concerning icing effect on sensors, knowledge in handling the ice affected data by
the meteorological services, requirements of ice free sensors and direct measurements of ic-
ing, to make a market survey in ice free gauges available, and to give a proposal of specifica-
tion of improved measurements under cold climate and ice affected sites. [3]
SWS II: 2000-2002
A documented experiment [4] has been conducted at three sites in Finland, France and Swit-
zerland in the period 2000-2002 in order to analyze the performances of ice-free instruments
under extreme meteorological conditions.
See section 6.11 for more details regarding the available data of the SWS projects.
11/110
2.3.3 WECO: Wind Energy in cold climates
“Wind Energy Production in COld climates” WECO (JOR3-CT95-0014) 1996-1998, which
was partially supported by the European Commission DG XII Non Nuclear Energy Pro-
gramme aims at the investigation of wind turbines under cold climate operation.
It is shown experimentally and by numerical simulations that icing of rotor blades or other
components lead to decreased production due to ice accretion or safety demands. The icing
effect is directly related to the climate of the site of the wind turbine, and varies strongly from
region to region in Europe. Extreme low air temperature again set new demands for design
parameters. Icing of anemometers and other wind gauges typically lead to wrong estimation
of wind power potential and operational problems of wind turbines [5].
2.3.4 NEW ICETOOLS
“Wind Turbines in Icing Environment: Improvement of Tools for Siting, Certification and
Operation” NEW ICETOOLS NNE5-2001-259, 2002-2004, was partially supported by the
European Commission Energy, Environment and Sustainable Development Programme. The
aim was to produce tools and information to improve safety, availability and reliability of
wind turbines and their components and thus improve the economics of wind power produc-
tion in icing environments [6].
2.3.5 IEA WIND R&D Annex XIX
IEA R&D Wind is an agreement between 19 countries and the European Commission to fol-
low international development on wind energy deployment and to stimulate co-operative re-
search and development of wind technology. In 2001, International Energy Agency (IEA)
R&D Wind started Annex XIX; “Wind Energy in Cold Climates”. Since the start-up, the par-
ticipants of Annex XIX have been collecting operational experiences from selected sites that
experience frequent atmospheric icing or low temperatures. Collected data include informa-
tion on performance of standard wind turbines as well as performance of adapted wind turbine
technology specifically developed for cold climate sites. The aim of the work is to reduce the
risk that originate from cold climate and thereby reduce the cost of wind electricity produced
in cold climates [7, 8]. A second 3-year period of Annex XIX started late 2005 with Italy and
Germany as additional members.
2.3.6 CIGRE
The „Conseil International de Grands Réseaux Electriques“ is a non-profit NGO dealing with
all types of electrical component: production, transmission, distribution of electric energy. It
is research oriented and organized in study committees. Study Committee B2 „Overhead
lines“ deals also with meteorological aspects such as icing on overhead lines (WGB2.16 „Me-
teorology for overhead lines“) (www.cigre.org).
2.3.7 IEC/CENELEC
The International Electrotechnical Commission is the standardization body for all electrical
components in parallel to ISO (see below). IEC prepares standards for the design of overhead
12/110
lines taking into account meteorological parameters such as icing. CENELEC is the European
counterpart of IEC (www.iec.ch).
2.3.8 ISO
The International Standard Organization has issued the ISO-12494 [9] recommendation which
represents today the most widely used reference for icing on structures in general, but not for
overhead lines. The standard describes the ISO standard instrument to measure icing (see Sec-
tion 4.1).
2.3.9 IWAIS
The International Workshop on Atmospheric Icing of Structures is an informal institution
supported by research institutions and utilities. IWAIS Workshops that have been organized
every 2-3 years since 1982 are meant to be the main international gathering of researchers in
icing and icing related problems and assess the state of the art in icing research [10]. The
eleven published proceeding volumes of IWAIS contain approximately 5000 pages of infor-
mation on icing related issues.
13/110
3 Definitions and meteorological conditions
3.1 Generic definition
According to the ISO-12494 standard [9], ice accretion can be defined as any process of ice
build up and snow accretion on the surface of an object exposed to the atmosphere.
WG2 recommends adopting as standard the ISO-12494 and in particular the definitions pre-
sented in the following section.
3.2 Icing types (extracts from ISO-12494)
Atmospheric icing is traditionally classified according to two different formation processes:
a) Precipitation icing (including freezing precipitation and wet snow).
b) In-cloud icing (also called rime/glaze, including fog)
c) Hoar forst (not considered here?).
However, a classification may be based on other parameters, see table 1 and 2.
The physical properties and the appearance of the accreted ice will vary widely according to
the variations of the meteorological conditions during the ice growth.
Besides the properties mentioned in table 1, other parameters, such as compressive strength
(yield and crushing), shear strength, etc., may be used to describe the nature of accreted ice.
The maximum amount of accreted ice will depend on several factors, the most important be-
ing humidity, temperature and the duration of the ice accretion.
A main precondition for significant ice accretion is the dimensions of the object exposed and
its orientation to the direction of the icing wind. This is explained in more detail in chapter 7
of the ISO document.
Table 2 gives a schematic outline of the major meteorological parameters controlling ice ac-
cretion.
A cloud or fog consists of small water droplets or ice crystals. Even if the temperature is be-
low the freezing point of water, the droplets may remain in the liquid state. Such super cooled
droplets freeze immediately on impact with objects in the airflow.
When the flux of water droplets towards the object is less than the freezing rate, each droplet
freezes before the next droplet impinges on the same spot, and the ice growth is said to be dry.
When the water flux increases, the ice growth will tend to be wet, because the droplets do not
have the necessary time to freeze, before the next one impinges.
In general, dry icing results in different types of rime (containing air bubbles), while wet icing
always forms glaze (solid and clear).
14/110
Figure 1 gives an indication of the parameters controlling the major types of ice formation.
The density of accreted ice varies widely from low (soft rime) over medium (hard rime) to
high (glaze).
Table 1: Typical properties of accreted atmospheric ice
1
Type of
ice
Density Adhesion & General Appearance
|kg/m
3
| Cohesion Colour Shape
Glaze 900 strong transparent evenly distributed/ icicles
Wet snow 300-600 weak (forming)
strong (frozen)
white evenly distributed/ eccen-
tric
Hard rime 600-900 strong opaque eccentric, pointing wind-
ward
Soft rime 200-600 low to medium white eccentric pointing wind-
ward
Table 2: Meteorological parameters, controlling atmospheric ice accretion
Type of ice Air tempe-
rature |°C|
Wind
speed
|m/s|
Droplet
size
Water
content
in air
Typical event
duration
Precipitation icing
Glaze (freez-
ing rain or
drizzle)
-10 < t
a
<0 any large medium hours
Wet snow 0 < t
a
< +3 any flakes very
high
hours
In-cloud icing
Glaze see fig. 1 see fig. 1 medium high hours
Hard rime see fig. 1 see fig. 1 medium medium days
Soft rime see fig. 1 see fig. 1 small low days
1
In practice, accretions formed of layers of different types of ice (mentioned in table 1) may also occur, but from
an engineering point of view, the types of ice do not need to be described in more detail.
15/110
0
5
10
15
20
25
-20 -15 -10 -5 0
Air temperature ( degree Celcius)
Soft rime
Hard rime Glaze
Wind speed (m/s)
Figure 1: Type of accreted ice as a function of wind speed and air temperature
2
.
3.2.1 Glaze
Glaze is the type of precipitation ice having the highest density. Glaze is caused by freezing
rain, freezing drizzle or wet in-cloud icing and normally causes smooth evenly distributed ice
accretion.
Glaze may result also in formation of icicles, and in this case the resulting shape can be rather
asymmetric.
Glaze can be accreted on objects anywhere, when rain or drizzle occurs at temperatures below
freezing point
3
.
The surface temperature of accreting ice is near freezing point, and therefore liquid water, due
to wind and gravity, may flow around the object and freeze also on the leeward side.
The accretion rate for glaze mainly varies with:
- Rate of precipitation
- Wind speed
- Air temperature
3.2.2 Wet snow
Wet snow is, because of the occurrence of free water in the partly melted snow crystals able
to adhere to the surface of an object. Wet snow accretion therefore occurs when the air tem-
perature is just above the freezing point.
2
The curves In figure 1 shift to the left with increasing liquid water content and with decreasing object size
3
Freezing rain or drizzle occurs when warm air aloft melts snow crystals and forms rain drops, which afterwards
fall through a freezing air layer near the ground. Such temperature inversions may occur in connection with warm
fronts or in valleys, where cold air may be trapped below warmer air aloft.
16/110
The snow will freeze when wet snow accretion is followed by a temperature decrease. The
density and adhesive strength vary widely with, among other things, the fraction of melted
water and the wind speed.
3.2.3 Rime
Rime is the most common type of in-cloud icing and often forms vanes on the windward side
of linear, non-rotary objects, i.e. objects which will not rotate around the longitudinal axis due
to eccentrically loading by ice.
During significant icing on small, linear objects the cross section of the rime vane is nearby
triangular with the top angle pointing windward, but as the width (diameter) of the object in-
creases, the ice vane changes its form, see chapter 7 of the ISO document.
Evenly distributed ice may also be formed by in-cloud icing when the object is a (near) hori-
zontal "string" (linear shape) which is rotary around its axis. The accreted ice on the wind-
ward side of the "string" will force it to rotate when the weight of ice is sufficient. This
mechanism may continue as long as the ice accretion is going on
4
. It results in an ice accretion
more or less cylindrical around the string.
The most severe rime icing occurs at freely exposed mountains (coastal or inland), or where
mountain valleys force moist air through passes, and consequently both lifts the air and in-
creases wind speed over the pass.
The accretion rate for rime mainly varies with:
- Dimensions of the object exposed
- Wind speed
- Liquid water content in the air
- Drop size distribution
- Air temperature
3.2.4 Other types of ice
Hoarfrost, which is due to direct phase transition from water vapour into ice, is common at
low temperatures. Hoarfrost is of low density and strength, and normally does not result in
significant load on structures
5
.
4
The liquid water content of the air becomes so small at temperatures below about -20°C that practically no in-
cloud icing occurs.
5
Comments to ISO definitions: Icing types on wind turbine blades depend on the velocity i.e. the radial position.
Glaze ice may occur on the tip while rime occurs near the root. Figure 1 should be extended to 80 m/s.
17/110
4 Ice measurements as described in standards
Atmospheric icing affects all kinds of installations in susceptible areas. The following stan-
dards have been found depending on the field of application.
4.1 International Standardization Organization (ISO)
The ISO issued in 2001 a standard [9] for ice accretion on all kinds of structures, except for
electric overhead line conductors. In this recommendation, a standard ice-measuring device is
described as:
- A smooth cylinder with a diameter of 30 mm placed with the axis vertical and rotat-
ing around the axis. The cylinder length should be a minimum length of 0.5 m, but, if
heavy ice accretion is expected, the length should be 1 m.
- The cylinder is placed 10 m above terrain
6
.
- Recordings of ice weight may be performed automatically.
However, it is important to be aware of the different properties of various icing types, espe-
cially the wet and dry growth of freezing rain, but also the possibly weaker adhesion of wet
snow. If the cylinder cannot rotate freely due to wind drag, it may be provided with a motor to
force the rotation. The speed of rotation of the vertical collector is not critical.
The vertical cylinder is not fully appropriate for freezing rain in the wet growth stage
7
. For
this purpose it is preferred to use sets of horizontal collectors (rods) which are oriented or-
thogonal, like the Soviet standard ice collector [11] or the Canadian Passive Ice Meter (PIM)
as described in [12].
4.2 International Electrotechnical Commission (IEC)
4.2.1 Overhead lines
For electric overhead power lines, icing is often the most significant design parameter in eco-
nomic terms.
The IEC 60826 [12] gives rules for design of overhead lines in order to make them function
reliably under icing conditions. However, IEC does not give numeric values for ice loads
which are to be taken into account in various countries, as this aspect is considered to be the
responsibility of each country. The only requirement used in [12] is that the “reference ice
load” should be related to a “horizontal, circular conductor of 30 mm in diameter”.
6
Consideration must be given to the maximum snow depth during the winter. The cylinder should preferably be
placed in an area, where snow is blown away. For practical reasons, different erection heights above terrain are
accepted, as long as the results correspond to those for 10-m height.
7
For this purpose it is preferred to use sets of horizontal collectors (rods) which are oriented orthogonal, like the
Soviet standard ice collector [11] or the Canadian Passive Ice Meter (PIM) as described in [12].
18/110
IEC has also issued a Technical Report on measurement of ice loadings for overhead lines
[13]. It includes a description of historical test spans in Europe as well as in countries outside
Europe.
4.2.2 Wind turbines
Extracts from IEC 61400-1, ed 2 [14]:
”Environmental (climatic) conditions other than wind can affect the integrity and safety of the
Wind Turbine Generator Systems WTGS, by thermal, photochemical, corrosive, mechanical,
electrical or other physical action. Moreover, combinations of the climatic parameters given
may increase their effect. The following other environmental conditions shall be at least taken
into account and the action taken stated in the design documentation:
- temperature;
- humidity;
- air density;
- solar radiation;
- rain, hail, snow and ice
Other extreme environmental conditions that shall be considered for WTGS design are wind
speed, lightning and earthquakes. No minimum ice requirements are given for the standard
WTGS classes.
Other loads such as wave loads, wake loads, impact loads, ice loads, etc. may occur and shall
be included where appropriate.
During installation, environmental limits specified by the manufacturer shall be observed.
Items such as the following should be considered:
- wind speed, and
- snow and ice”
The standards issued by CENELEC include for Europe the same principles of overhead line
design as described in [12].
Furthermore, the international standard IEC 61774 [13] covers meteorological data with re-
spect to icing observation and measurement concerning electric overhead lines. This standard
has been designed on the basis of previous experiences of icing observation and has been
adopted in such a way that the information acquired is unified. These procedures are useful
because data conversion from different systems is very difficult.
4.3 World Meteorological Organization (WMO)
There are presently limited standards defined by the WMO and its Commission for Instru-
ments and Methods of Observation (CIMO) either for measurements performed under harsh
conditions (e.g. icing) or for the measurements of icing itself at Automatic Weather Stations
(AWS).
19/110
However, taking into account the recommendations of the final report of the recent EUMET-
NET (Network of European Meteorological Services) / Severe Weather Sensor experiment
SWS II [15], the WMO/CIMO Expert Team of Surface Measurements has recently decided to
include in the next version of the CIMO Guide – or in a later version - a chapter concerning
meteorological measurements performed under harsh conditions such as artic/mountain cli-
mate (icing), desert, urban, tropical, ocean, etc.. This topic should be further handled during
the CIMO XIV session.
4.4 Icing and wind turbines
Icing of wind turbines is briefly described in the standards [14, 16]. However, the results from
the EU-project NEW ICETOOLS [6] indicate that these standards seem to underestimate:
- the actual amount of ice,
- the influence on fatigue loads from extended periods of frost and
- the time period of ice accretion.
In the following, extracts concerning icing from present wind turbine design standards are
presented.
Extract from GLRP3.0-1998 [16]:
For non-rotating parts an ice formation with a thickness of 30 mm on all sides is to be as-
sumed on all exposed surfaces. The ice density is to be taken as 700 kg/m
3
. For the Wind En-
ergy Converter (WECs) at standstill also the rotor-blades have to be analyzed with this ice
cover.
For the rotating machine the situations "all blades are iced over" and "all but one blade iced
over" have to be analyzed. The mass distribution (mass per unit length) is to be assumed on
the leading edge. The mass distribution increases linearly from zero in the rotating axis to the
value µ
E
at half radius and stays constant to the blade tip. The value µ
E
is given by:
) (
min max min
c c kc
E E
÷ = µ µ
where
tip at the lenght chord c
lenght chord maximum c
1m R
radius rotor R
) 0.32R/R 0.3exp( 0.00675 k
700kg/m : ice the of density ȡ
min
max
1
1
3
E
=
=
=
=
÷ + =
=
20/110
5 Meteorological measurements under icing conditions
5.1 Introduction
The accuracy of the surface measurements of various meteorological variables is essential for
meteorological services, researchers in climatology (e.g. climate change
8
), aeronautical mete-
orology, etc. It is therefore essential to characterize the effects of ice accretion on the sensors
and, when possible, to prevent it.
The WMO Guide for meteorological measurements [17] does not define the temporal reliabil-
ity of sensors, e.g. the required availability of data per year or per month, so that most mete-
orological services have specified their own targets for availability of data. Similar targets are
defined also for other applications. Furthermore, the WMO Guide does not separately con-
sider severe weather conditions like icing, even if low temperature is specified in the require-
ments. In the same way, the manufacturers typically specify their instruments’ performance
for severe weather conditions by taking into account low temperature (for instance operating
range: -40°C… + 50°C), but not icing. Presently, icing events are defined as periods of time
where the temperature is below 0°C and the relative humidity is above 95%, a very simplified
approach. Usually low air temperature is not a major problem for meteorological observa-
tions: for many sensors, this is taken into account e.g. by using shaft heating for anemometers
with rotating parts (at small and/or mobile automatic stations, the power supply may not be
sufficient even for shaft heating).
5.2 WMO/CIMO Recommendations
The following recommendations are stated by WMO/CIMO:
- Improve the quality of meteorological measurements under cold climate conditions,
- Provide manufacturers data for design of ice- free sensors,
- Provide users and providers of meteorological information better bases for selection of
suitable sensors for their purposes.
To improve the general knowledge on icing and icing climatology, the following recommen-
dations for further activities are given:
- Improve the design of the instrument (mechanical) and heating system to optimize the
required heating power
- Promote the development of icing observation instruments
- Promote the results of past, present and future experiments
- Promote national " icing maps"
- Promote a classification for "meteorological" sensors taking into account accuracy,
climatic conditions and reliability of data required for different applications
8
The anticipated increase in air temperature will inherently lead to higher contents of water (vapour) in the lower
atmosphere. In mountains, and especially in northern latitudes, there is therefore an increased frequency of tem-
peratures near freezing, and together with more water available this could lead to higher frequencies of icing as
well as higher icing intensities and ice loads on structures. An anticipated increase in extreme wind speeds will
contribute likewise and indeed lead to significantly higher static and dynamic loads on exposed infrastructure.
[Svein Fikke, private communication]
21/110
- Promote the improvement of the WMO/CIMO Guide 8 for measurements in severe ic-
ing conditions
- Promote WMO-approved test sites for ice- free sensors, preferably combined with the
use of icing wind tunnels for testing of sensors, including anemometers.
It must be noted that most activities on developing requirements for instruments in harsh cli-
matic environments focus on icing conditions only (i.e. in extremely cold mountainous/Arctic
climates). Therefore, equipment for dusty and dry deserts, humid and hot tropics and oceans
with a harsh climate need further investigation. For these climates only very limited guidance
material is available on the implementation and maintenance of automatic observing systems
and, therefore, further studies are necessary. Moreover, not only the performance and mainte-
nance issues of a system are a point of concern but also destruction of instruments caused by
extreme weather should be considered (e.g. tropical cyclones reaching 300 km/h or more.)
In line with past developments and published material on this matter, documentation on the
requirements for instruments for observations in harsh climatic conditions has to be generated.
In particular the IOM report, as announced at CIMO XIII should be finalized and published.
Moreover, as stated in the recommendations in the SWS II report (see above) more attention
to this topic should be given in a future revision of the CIMO Guide [17].
Taking these recommendations into account, and in line with CIMO XIII, the CIMO man-
agement group has decided to continue the work on the provision of guidance material on
implementation of instruments in harsh climatic environments as requested by Congress and
Technical Commissions. To realize this, further study on already existing test reports should
be carried out. Moreover the already announced IOM report has to be finalized and published.
A decision should be made on how to implement requirements on the instruments capable of
measurements in a harsh environment in a new revision of the CIMO Guide and it should be
considered to carry out inter-comparisons of dedicated sensors in such environments. Preced-
ing such inter-comparison an inquiry should be carried out among the Members facing short-
comings of today's equipment due to harsh and extreme weather.
5.3 Definitions
5.3.1 Meteorological icing M
icing
Meteorological icing M
icing
is defined as the duration of a meteorological event or per-
turbation which causes icing [unit: time].
Meteorological icing can be characterized by:
a) the duration of the icing event, and/or
b) the meteorological conditions,
and possibly with additional information such as:
c) the total amount of ice accreted on a standard (reference) object during the icing event,
d) the average and maximum accretion rate.
22/110
Automatic certified reference sensors are lacking for the determination of items a) and d),
whilst the items b) and c) can be more or less achieved with presently available technology.
However, it must be noted that meteorological icing is not easy to define. It is today widely
accepted that it depends on the following factors:
a) the shape of the object,
b) the wind speed,
c) the air temperature,
d) the liquid water content LWC,
e) the droplet size distribution,
the latter two being difficult to measure in operational mode. Tentative developments have
been achieved, such as the Rotating Multicylinder RMC. Unfortunately, these cannot be oper-
ated in an automatic way and cannot therefore be implemented at automatic stations. New
developments may improve this situation [18].
Today, ice accretion can be measured directly by instruments measuring:
a) changes of a vibrating frequency (Rosemount, Vibrometer, Wavin-Labko),
b) changes in electrical properties (Instrumar, Labko)
c) the load of ice (ISO 12494)
d) the growth rate of ice by yielding a yes/no output (at regular intervals) by a heating
cycle
e) optically (obstruction of light path, IR or reflection technologies: Infralytics, HoloOp-
tics).
In addition, icing can be measured indirectly by measuring the variables that cause icing (see
Chapter 4.2) or variables that correlate with the occurrence of icing, such as cloud height and
visibility.
5.3.2 Instrument icing I
icing
Instrument icing I
icing
is defined as the duration of the technical perturbation of the in-
strument due to icing [unit: time].
Instrument icing is the effect of icing on the quality (e.g. degradation) of the measurements,
depending on icing conditions as well as the design of the instrument. It can be today only
recorded by analyses of video recordings, and/or regular visual observation, or by comparison
of the measurements with a reference that is kept ice free.
This definition is valid for all objects or structures. It can be easily generalized to “structural
icing”.
23/110
5.4 Site effects
In the preceding section meteorological icing has been shown to be different from instrumen-
tal icing, the latter being the consequence of the former, but with different effects depending
on the characteristics of the meteorological conditions and of the instrument design.
Instruments (or structures) will behave differently depending on the location of their installa-
tion. A sensor operated in northern countries may get frozen at the beginning of the winter
and remain as such due to the low temperatures and the lack of sunshine. On the contrary, this
instrument may be installed further south and work more or less undisturbed under milder,
sunnier conditions. Therefore, the instruments characteristics must be evaluated as a function
of the site of installation in terms of local icing conditions.
5.5 Site Icing Index
To be able to express the maximum expected amount of accreted ice at a certain site, the term
ICE CLASS (IC) is introduced in the ISO 12494 standards, for design purposes. For the pur-
pose of meteorological instruments a site icing index, Sn, is introduced by using icing fre-
quency, duration and intensity:
Sn is the parameter to be used by the meteorological community to determine how severe ice
accretion is expected at a particular site, in regard to meteorological instruments. Maximum
loads are not included in this definition.
The climatologists may provide information about Sn, which (in general terms) tells how
much icing can be expected at a given location. Measurements and/or model studies are nec-
essary to obtain the information needed for a specific site, unless experience can supply the
same information.
The station class may vary within rather short distances of a specific area. Measurements
should be carried out either where ice accretion is expected to be most severe, or at the precise
station site, or both.
Therefore, it is recommended that a classification of sites, e.g. Automatic Weather Station
(AWS) is introduced indicating the degree of severity of local icing conditions. The following
table was set up during the EUMETNET/SWS II experiment and describes tentatively the
framework of such a classification.
Table 3: Classification of sites according to the severity of icing (from EUMETNET/SWS II
Report).
Site icing
index
Days with me-
teorological
icing / year
Duration of
meteorological
icing %/year
Intensity of
icing g/100
cm
2
/h (typical)
Icing sever-
ity
S5 > 60 > 20 > 50 Heavy
S4 31-60 10-20 25 Strong
S3 11-30 5-10 10 Moderate
S2 3-10 < 5 5 Light
S1 0-2 0-0.5 0-5 Occasional
24/110
5.6 Measurements under icing conditions
In the preceding section, it was possible to set up definitions which are used to describe the
behaviour of meteorological instruments under harsh conditions and to select new adequate
sites, or specify existing ones for Automatic Weather Stations (AWS). In the following, these
definitions are used to classify sensors accordingly.
5.6.1 Ice-free meteorological sensors
Icing environments set special requirements for sensors. Accurate meteorological measure-
ments under cold and icing conditions are required for various applications such as aviation,
transportation, emergency services, tourism industry, meteorological observations, wind en-
ergy production, agriculture, etc. Instruments performing measurements (such as humidity,
temperature, wind speed, wind direction, precipitation, radiation, etc.) in cold climate envi-
ronments have to be properly heated to maintain their accuracy under icing conditions. Today,
there are several types of ice-free sensors available on the market. Some of them will not fulfil
the requirements of the World Meteorological Organization (WMO) for accuracy and avail-
ability when operated under icing conditions, especially at mountainous sites, but also at sites
like airports, road stations e.g. in northern Europe. The practical requirements on accurate
operation of sensors are even more demanding for many applications other than for meteoro-
logical synoptic purposes.
During the EUMETNET/SWS II project, a simple classification for instrument icing charac-
teristics was introduced for analyzing the behaviour of the different available - defined as ice-
free - sensors under icing conditions. This classification made it possible to study statistically
the effects of different amounts of ice upon sensors, and the magnitude of the resulting errors.
The state of the instruments was classified at all stations with a value between 0 and 3:
- Class 0: totally free of ice
- Class 1: light ice accretion, without obvious effect on the sensitive part but which
could influence the wind field.
- Class 2: medium ice accretion, probably disturbing the measurements – Sensitive ele-
ments seem free of ice and wind field is obviously disturbed.
- Class 3: totally covered with ice – sensitive elements are covered with ice.
A tentative classification of reliability of different types of sensors under icing conditions was
performed in such a way. The results were summarized with respect to general requirements
for a meteorological synoptic station. The sensor got one star when not adequate to provide
reliable data at the climatic conditions met at sites with more than 60 icing days per year, but
could be used reliably at some other less demanding sites. Three stars indicate that the sensor
is close to 100 % ice-free but has other significant errors, and sensors with five stars can be
strongly recommended for measurements at sites with harshest icing conditions.
It must be kept in mind that for practical measurements the requirements and the choice of
sensor depend significantly on the goals of the application and on the location of weather sta-
tion (icing climate). This leads to the following classification proposal.
25/110
5.6.2 Performance Index
During a meteorological icing event, the relationship between meteorological and instrument
icing can be expressed in the following way:
An instrument which remains free of ice during a meteorological icing period (good heating,
good coating, etc.) may be considered as well adapted for the station’s climatology. On the
other side, an instrument that becomes frozen during a meteorological icing period and re-
mains in that state after the meteorological icing period must be classified as poorly adapted
to the site’s environmental conditions. This leads to the following definition:
The Performance Index (PI) is the ratio of the instrument icing to the meteorological
icing, both expressed in the same time unit.
icing
icing
M
I PI =
The PI can be used for the selection of the instrument as function of some station’s classifica-
tion. A value of PI near 0 reflects a good performance of the instrument in terms of icing (e.g.
good heating). Values of PI between 0 and 1 may be acceptable as long as ice detector infor-
mation is available to “flag” dubious periods of measurements. Values of PI higher than 1
indicate a sensor which is sensitive to icing (e.g. poor heating) for a time period (much)
longer than the meteorological icing.
Further useful definitions:
Incubation time: delay between the beginning of the meteorological icing and of the instru-
mental icing.
Recovery time: delay between the end of the meteorological icing and the full recovery of the
performance of the instrument
Instrument icing can be smaller, equal or longer than meteorological icing. The incubation
time indicates how quickly the instrument responds to icing while the recovery time may be
much longer that the meteorological icing, especially in northern countries with low solar
irradiance in winter.
In summary, the following definitions are needed to characterize the properties of an instru-
ment under icing conditions:
The meteorological icing M
icing
is the duration of the icing event (see §5.3.1)
The instrument icing I
icing
describes the effect of M
icing
the on the instrument (see § 5.3.2).
The Performance Index PI characterizes the behaviour of the instrument under icing condi-
tions.
26/110
5.6.3 Instrument Class Index
It is evident that a classification for meteorological sensors will be difficult to achieve taking
into account accuracy and required reliability of data combined with climatic conditions.
Therefore, the goal is now to build a common indicator by combining PI and Sn (see § 5.5).
A potential indicator may be given by ICIn (n=1..5), the Instrument Class Index, which corre-
sponds to PI values ranging from 0 -> ’ depending on the different site icing indices S1 to S5
as indicated in Table 4. The range of this classification extends from ICI5 (PI = 0; availability
= 100 % Æ perfect icing non-sensitive instruments) to ICI1 (PI = very high values; availabil-
ity ”40 % Æ instruments which could remain frozen for a very long period after the meteoro-
logical icing period, e.g. long recovery time for high latitude stations without sun during
whole seasons).
Table 4: Classification of instruments in terms of mean performance depending on the sta-
tion’s site icing index (the availability values displayed in italic are purely hypothetical and
will have to be specified in future).
Instrument
Class Index
PI for
S1 … S5
Mean availability in
% for S1 … S5
Remarks
ICI5 0 100 % Excellent instrument not
sensitive to icing
ICI4 0 …1 99 ... 90 % Good instrument, little
sensitivity to icing
ICI3 1 ... 5 89 ... 70 % Instrument moderately
sensitive to icing.
ICI2 5 ... 20 69 ... 40 % Instrument to be used
only with separate icing
detection
ICI1 20 ... ’ 39 ... 0 % Instrument not recom-
mended for such appli-
cations
The interpretation of the ICIn index is strongly linked to the station’s site icing index (Sn,
n=1-5) and the effect of icing on the results’ quality as described by the PI defined above.
This leads to the following graphical representation (Figure 2) where the user could select the
class of instruments needed to fulfil his requirements depending on the location (e.g. classifi-
cation) of his station and requirement.
27/110
Instrument classification
0
20
40
60
80
100
S1 S2 S3 S4 S5
Site Icing Index

A
v
a
i
l
a
b
i
l
i
t
y

[
%
]
ICI1
ICI2
ICI3
ICI4
ICI5
Figure 2: Graphical display of the instruments’ classification presented in Table 4 (hypotheti-
cal sensors) reflecting the availability of an ICIn sensor installed at a Sn site. In reality, the
ICI will have to be determined for each Sn. The availability will then be computed consider-
ing the severity and duration of icing periods.
28/110
6 Examples of existing icing data
Numerous activities in the field of measurement of icing have already taken place in different
countries. The following gives an overview of what has been achieved to date. More details
may be found in the Annexes, as indicated.
The contributions in this chapter are based on information provided by members of the COST
Action 727. Other data may be available in some countries.
6.1 Finland
VTT and Digita (former Finnish Broadcasting Co., Distribution Dept.) have made measure-
ments of icing on tower structures and measurements of drop size and liquid water content of
clouds and comparisons of meteorological instruments on hilltops in severe icing conditions
in Finland since 1986. An operating 128m tall TV tower and a 7.5 m test tower at Ylläs (700
m asl) have both been equipped with load cells, so that the ice load on them could be continu-
ously measured [19,20]. Ice detectors were also tested [21], and VTT has also later performed
ice detector tests in four locations during the period 1998 – 2005.
The Luosto test station (500 m asl) in northern Finland was set up during the winter
2000/2001 by Finnish Meteorological Institute. The main purpose of the Luosto test station is
to measure icing as well as the behaviour of meteorological instruments.
More details are given in Annex 1.
6.2 Germany
Icing measurements were carried out at altogether 40 locations in the eastern part of Germany
during 1965 – 1990, up to 35 locations were operated simultaneously. Since 1991 five stations
are still in operation. A standard observation pole has been used for all stations. On the
Deutscher Wetterdienst’s meteorological observatory Lindenberg, ice measurements are cur-
rently made at 10, 50 and 90 m above ground.
More details can be found in Annex 2
6.3 Slovak Republic
The Slovak Hydrometeorological Institute has data on icing from 13 stations, ranging from
115 to 2 634 m asl. The measurements are both visual and by instruments. The oldest data are
from 1957.
More details can be found in Annex 3.
29/110
6.4 Norway
A measuring station for ice monitoring has been operated at a coastal mountain of about 800
m asl in central western Norway with the support of the Norwegian Research Council and two
Norwegian energy companies. The station was equipped with ice-free wind sensors, tempera-
ture sensors and a web-camera. Ice accumulation was derived from web-camera pictures of
wires. Measurements took place during two winters. The monitoring results have been com-
pared with an ice model and the results are reported in Harstveit [22] and Harstveit et al. [23].
Accumulated ice at 10 m up to a maximum value of 20-25 kg/m where reported. As a part of
the development of wind farms in the costal mountains of Norway new ice monitoring pro-
grams have recently been set up. Different ice detection equipment, such as rotating multi-
cylinder, ice indicators and web-cameras are utilized. Unfortunately, no data are yet available,
but hopefully useful information will become available from these programs in the future.
During the period 1978-2000 the Norwegian Power Grid Company, Statnett SF, has operated
more than 20 sets of racks for ice measurements in 16 locations in mountainous areas for
power line design purposes. Each set consists mainly of two perpendicular racks, where one
leg is perpendicular to the main icing wind direction. Some sets were established to study the
effect of local topography. Most of these racks are in coastal mountains in the range of 600 –
1 200 m asl.
More details can be found in Annex 4.
6.5 Czech Republic
Two institutes have performed icing studies in Czech Republic, EGU Brno and Institute of
Atmospheric Physics (IPA, Prague). EGU Brno has operated a test site on Studnice (800 m
asl) continuously since 1940. Ice loads were measured on a rack with orthogonal rods 2 m
above ground. The annual maxima of loads on this rack for the period 1940/41 – 1998/99 are
presented in Figure 3. This unique time series is outstanding, since it is the only series of this
kind in the world covering such a long time period.
Figure 3. The Studnice site has also a test installation for power lines consisting of a central
observation tower and 2 spans of about 250 m length on each side.
0
5
10
15
20
25
40/41 50/51 60/61 70/71 80/81 90/91
t (years)
Q
[
k
g
.
m
-
1
] Studnice 800 m, n=59
30/110
Furthermore, EGU Brno has developed the instrument “METEO” which is installed at 14
locations in the country. The measuring probe is a vertical rod of 30 mm diameter.
The IPA has a similar instrument, IceMeter, installed in two locations, Milesovka (837 m asl)
and Nová Ves.
More details can be found in Annex 5.
6.6 UK
Test data on icing has been available on rotating rigs and test spans since 1988. Many of
these sites lasted only a few years before being closed down for financial reasons. The long-
est running site, at Deadwater Fell in Northern England, was established in 1991 and is still
currently open, although it has been ‘mothballed’ occasionally so there is no continuous ice
measurement over this period. It currently monitors wind speed and direction, temperature,
ice loads (by time lapse video cameras and also load cells), precipitation and relative humid-
ity. It has operated the Gerber instruments on loan from the UK Meteorological Office.
Measurements have been made on conductors from 16mm² to 800mm² of the copper, alumin-
ium and covered variety as well as fiber optic systems such as Optical Pipe Ground Wire
(OPGW), fiber-wrap and All Dielectric Self-Supporting (ADSS).
More details can be found in Annex 6.
6.7 Sweden
Three different sensors (IceMonitor, HoloOptics and Segerstrom) are currently being tested
under field conditions in Sweden and Norway. Meteorological parameters are measured,
along with icing data, in Ritsem, Åre and Drammen (Norway).
More details can be found in Annex 7.
6.8 Bulgaria
More details can be found in Annex 8.
6.9 Hungary
More details can be found in Annex 9.
6.10 Russia
Some information on measurements in Russia is found in Annex 10.
6.11 Canada
Some information on measurements in Canada is found in Annex 11.
31/110
6.12 WMO/CIMO inter-comparisons of wind instruments under
harsh conditions
The WMO Wind Instrument inter-comparison was organized following the recommendation
of the Commission for Instruments and Methods of Observation CIMO [2]. It was carried out
under the aegis of the WMO by both Meteo-France and the Swiss Meteorological Institute
(presently MeteoSwiss) at the Mt. Aigoual (France) from July 1992 to October 1993.
The objectives of the inter-comparison were the following:
- To derive performance characteristics on the operational use of wind sensors based on
the detailed record of their measurement values and a record of the prevailing atmos-
pheric conditions.
- To determine the suitability of these instruments for long-term unattended operation
especially in a mountainous environment
- To make proposals for further improvements to the WMO regulatory materials con-
cerning the measurements of wind
- To evaluate and publish the results of this analysis in a WMO publication
Icing phenomena have been studied from visual sensor control by the observer and by study-
ing the data. It was noted that:
- The importance of the icing phenomenon could not be characterized from the ice de-
tectors.
- The formation of ice made almost all the calculated parameters incoherent.
In the conclusions of the report, the difficulties of performing measurements under harsh icing
conditions are reported in the following way:
- The only sensors having supported severe icing events without noticeable measure-
ment errors are the Pitot sensors and one vertical axis anemometer. These sensors re-
quire a high amount of energy for heating (300 to 500W). The meteorological per-
formance of these sensors is not perfect and does not meet the WMO accuracy rec-
ommendations. It appears difficult to be both “accurate” and rugged for severe icing.
- Manufacturers commercializing measurement instruments should put at the users’ dis-
posal a detailed and complete documentation including a detailed installation and
maintenance book and the exact metrological specifications of the sensors.
6.13 EUMETNET/SWS II project
The EUMETNET "Severe Weather Sensors II" project (SWS II) tested 15 wind sensors, 6
temperature and humidity measurement systems with different types of shields and 4 solar
radiation sensors equipped with heating. During the project also different methods of meas-
urement of atmospheric icing were used and tested. The three test sites were located in north-
ern Finland, in the Swiss Alps and close to the Mediterranean in the French mountains, all
with more than 60 days/year of atmospheric icing.
From the results presented here, it can be seen that heating power is required especially for
wind measurements, but the power consumption can be relatively low if the sensors are prop-
erly designed. The tests and verifications showed that wind speed, wind direction and air tem-
32/110
perature could be measured with high accuracy and high reliability at cold climate sites under
most severe icing conditions even at automatic weather stations. For temperature and humid-
ity sensors, some of the shields provide significant improvement in comparison with meas-
urements performed with other systems in use at different national meteorological services.
However, under harsh conditions, the reliability of temperature and humidity measurements
does not yet reach the level available for wind measurements. Concerning test measurements
on the heating/ventilation systems for solar radiation measurements, results show that strong
icing conditions may dramatically disturb the measurements. None of the tested systems were
able to fully withstand the harsh climatic conditions prevailing at such sites.
It was not possible to study the performance of the various sensors versus intensity of ice ac-
cretion due to the lack of dedicated sensors to measure icing [4].
33/110
7 Requirements for ice detectors
7.1 Concepts
7.1.1 Purposes of measurements
At the present time, many sensors that are designed and labelled as ice detectors are available.
Some of the instruments measure icing rate, some measure the weight of ice (persistence and
maximum loads) and some indicate if an icing event is ongoing. Therefore, the purpose for
using ice detectors needs to be defined. Some detectors are clearly designed to indicate incipi-
ent icing only whereas others have been designed to measure the total mass of ice accumula-
tion.
Requirements regarding time resolution, measuring range, threshold values as well as re-
sponse time of sensors depend on the purpose of individual measurements, and are therefore
not further specified in these generic descriptions.
7.1.2 Range of use
The range of use varies between different ice detectors. For example, some sensors have been
designed for aviation purposes and perform well on airplanes, but may not be very well
adapted for meteorological purposes due to different environmental conditions as indicated
below.
The parameters that have an effect on the operation of an ice detector are air velocity, size of
water droplets, amount of liquid water present and the physical size of the probe i.e. all the
parameters that have an effect on the amount of water ending up on a detection surface. The
range of use should be defined by means of these parameters and should be validated in con-
trolled circumstances (see below). Furthermore, it must be kept in mind that in certain cir-
cumstances the operation of the ice detector may be affected by other phenomena, such as ice
sublimation due to dynamic heating in high velocity airflows and Ludlam’s limit, which sets
the upper LWC limit of use for some ice detectors [24].
Heated sensors may be iced up during non-icing periods due to melting of dry snow.
7.1.3 Icing types
All icing types that adhere on static or moving structures can be harmful and need to be iden-
tified.
7.1.4 Verification of performance
Considerable deviations between the results of ice detectors of the same type and even similar
ice detectors can be found [25,26].
Definition of the range of use and some calibration scheme might improve the current situa-
tion. Range of use and data verification could possibly be carried out in icing wind tunnels,
where the icing condition can be regulated and monitored. Kanagawa Institute of Technology
(KAIT) has conducted wind tunnel test for investigation of icing events on airfoil models and
anemometers as described in Annex 21. Wind tunnel tests include various kinds of testing
carried out in an atmosphere of low and moderate temperatures. The primary objectives were
to quantitatively find out the effect of icing on wind speed measurements and to evaluate the
effectiveness of measures to prevent ice or snow accretion on specific objects.
34/110
A further possibility lies in the development and long term operation of “icing test centres”
similar to (or included in) the Regional Instruments Centres (RICs) of the WMO where mar-
ket available and future instruments could be tested under different climatic conditions (e.g.
Scandinavia, Alps, Pyrenees, etc.).
7.2 Siting of icing sensors
7.2.1 Micrositing
Ice sensors should be placed so that the detection surface of the device faces up wind and free
air flow is granted. In addition all ice detection devices should be placed above tree tops and
other possible obstacles. Appropriate locations for ice detectors are support structures of
overhead power lines, wind turbines, link masts of mobile phone networks and in general high
structures that provide free air flow around the ice detector. ISO 12949 [9] recommends 10m
measurement height above ground. However, as icing measurements are dependent on the
different application types, ice sensors can be installed at different heights. Automated
weather stations are not generally appropriate as they are located close to ground level and
seldom provide a correct representation of those icing conditions that prevail at a higher level
e.g. wind turbine rotors.
Details such as the mounting orientation and height detectors will have to be analysed directly
at the test centre sites.
7.2.2 Standard Reference and procedures
Ice accretion on structures is not only a function of environmental parameters, but is also de-
pendent on the properties of the accreting object itself, e.g.:
a) size (diameter, width etc.)
b) shape (flat, sharp edges, cylindrical, spherical etc.)
c) flexibility (rigid/flexible member in bending/torsion etc.)
d) orientation relative to wind direction (angle of incidence)
and to some extent:
a) surface structure (paint, steel, concrete etc.)
b) material (wood, steel, plastics etc.)
Measurements of ice accretions therefore have to be specified with respect to devices, proce-
dures, arrangements on site etc. The set-up must be designed in a way that causes the lowest
possible influence on the accretion process itself:
A standard reference device should always be part of the measurements, giving the trace-
ability to standard measurements of ice accretion. Other parts of the set-up may help to estab-
lish the connections between “standard accretions” and the most important structural parame-
ters as described above (size, shape, etc.). These extended measurements should only be exe-
cuted at special selected sites, and collected data should be analysed and used, generally to-
gether with the standard measurements. Frequency of observations may be adjusted to the
local conditions.
35/110
On sites where melting or shedding are likely to occur shortly after the accretion period, ob-
servations must be carried out before this happens for example by making use of camera sys-
tems.
When automatic recordings are performed, it is important to add also visual observations dur-
ing and/or after the accretion period, because only these types of observation can give suffi-
cient information on such complex load situations. These visual observations have to be
logged, and documented with appropriate digital camera pictures. Remote reading (including
camera observations) makes it possible to get online information about an icing event so that
the site may be visited in proper time.
7.2.3 Macrositing
The following table 5 describes the information needed concerning ice types that are re-
quested for the different fields of application.
Table 6 displays the density of meteorological networks equipped with ice detection systems
for the different fields of application. For example a developer of a wind energy project would
need relatively dense measurement network due to the considerable influence of local land-
scape to the icing conditions.
Table 5: Ice parameters required
Requested information
Application Icing rate Ice load Icing time Persistency
Wind turbine operation x x x x
Wind project planning x x x
Power line design x
Power line operation x x x x
Aviation x x
Telecommunication masts x x
Suspension bridges x x
Transport (roads, railways) x x
Meteorology and climatology x x x x
36/110
Table 6: Location of ice detection needed by the different users.
Minimum distance to closest ice detection point
Application
On the site of
interest
Less than 50km
from the site of
interest
More than 50
km from the
site of interest
Wind turbine operation x
Wind project planning x x
Power line design x
Power line operation x
Aviation x
Telecommunication masts x
Suspension bridges x
Transport (roads, railways) x x
Meteorology and climatology x x
7.3 Guidance for selecting ice detectors
7.3.1 General
Appropriate ice detectors should be chosen with respect to the purpose of their use. There are
presently two systems of ice detectors:
- with status icing / no icing
- with recording of the whole icing cycle (ice mass, ice accretion rate).
The size of the detector probe has a significant effect on performance of an ice detector. When
icing detectors shall be selected, the purpose of the measurements has to be considered care-
fully. For example, smaller droplets in low speed airflow pass large objects more efficiently
due to their low inertia and the fact that large objects deflect the airflow upstream from the
object (collision efficiency). Therefore, no single ice detector can provide data that are di-
rectly applicable to all types of structures and conditions [27].
Ice detectors should be used only in conditions for which the devices were designed. For ex-
ample, in order to design overhead lines, it is necessary to know the ice mass during an icing
cycle. Measurement systems recording the whole icing cycle must be used in such a case
without ice shedding or heating of a sensor during an icing cycle.
37/110
7.3.2 Applications
Wind turbines
Ice detectors are needed for planning and operation of wind turbines. Results from the plan-
ning phase will influence the selection of type of turbine and equipment. During operation, ice
detectors are needed to detect incipient icing as quickly as possible for controlling heating
systems, turbines and to give a warning about possible ice shedding in populated environ-
ments. Another type of ice detector is needed to indicate whether there is ice on some surface
or not. The primary output would be the duration of the period that accumulated ice adheres
on a detection surface without any heating or external forces. The most important parameter is
time in both applications.
Icing is closely related to the speed of air flow and so the ice detector for wind turbine appli-
cations should preferably be attached to the outer part of the turbine blade.
Power lines
Power line companies are mainly interested in local wind and ice loads and wind-on-ice fac-
tors. Ice detection is needed to determine whether the power supply is likely to be secure or
whether an emergency response may be required. It could also enable pro-active measures. In
terms of line design the use of ice sensors to provide historical data for probability purposes is
important.
Because power line conductors have low torsional rigidity and thus rotate along the span dur-
ing icing events, the instrument that measures icing rate or ice load for this application should
be elongated horizontally and free to rotate [11].
Road safety
Mountain roads and bridges are commonly provided with wind and ice sensors for traffic
safety information, for combination with weather forecasts for pro-active road treatment (grit-
ting etc) as well as falling ice from bridges. Incorrect timing due to poor or badly interpreted
data can mean that road treatment is less effective and can lead to increased accident rate.
Airports
Preventive treatment of runways as well as de-icing of airplanes before take-off may represent
a important security problem as well as a heavy financial load for airport authorities. Early ice
accretion warnings should represent a safety increase, pollution decrease and financial sav-
ings. As airports are usually located at low terrain, the emphasis is to detect freezing precipita-
tion. Icing may be a hazard on an aeroplane wing even at temperatures above freezing due to
cold fuel stored in the wing tanks.
Transport facilities
Railways, cable cars, ski lifts, etc. require icing information so as to allow pro-active meas-
ures to be put in place (e.g. heating of switches) and also to enable staff/de-icing trains to be
on standby to combat icing. The ice can occur on power (3
rd
rail) and overhead power lines
38/110
causing pantograph damage. It is also serious when accumulations on traction rails reduce the
adhesion of train wheels.
Telecommunications
Telecommunication masts and towers are subject to icing on structures, antennas and/or guy
wires, with potential effects on the quality of the transmissions and on the structural loads. At
least 150 telecommunication towers have collapsed due to ice loads in the last 40 years. Wind
loads are an important contributing factor, so that a combined ice and wind load needs to be
estimated for the design.
39/110
8 Availability, verification and requirements of ice detec-
tors
8.1 Available ice detectors
There are presently few available instruments on the market. However, there are some proto-
type instruments which seem very promising and may lead to interesting products after thor-
ough testing and certification. These instruments are based on different working principles:
a) Vibrating rods: the vibrating frequency depends on the state of the rod (yes/no infor-
mation)
b) Direct infrared beam backscatter: light is reflected as soon as the sensor’s surface is
covered with a film of ice.
c) Infrared beam reflected on surface: the reflection characteristics change when the
„mirror“ is covered with ice.
d) Measurement of the weight of ice.
e) Measurements of LWC and droplet size distribution.
f) Detection of the attenuation of ultrasonic signal on ice detector structure due to ice.
g) Detection of changes in the electrical impedance on the surface of the probe.
h) Obstruction of light path.
The following table displays the available and prototype instruments, to the best knowledge of
the COST-727 / WG2 participants.
Table 7: List of available and prototypes of ice detectors on the market
Item Instrument Manufacturer
a Rosemount 0872J / 0871LH1 Goodrich (USA)
Rosemount 872C2 (ASOS-USA) Goodrich (USA)
SYGIVRE (Icing Rate Meter (IRM)) Hydro Quebec – Transénergie (CA)
Vibrometer (Prototype) Boschung (CH)
b Infralytic IR detector (Prototype) Infralytic (D), MeteoSwiss (CH)
c T21, T23 and T26 HoloOptics (SE)
d ICEmeter IAP (CZ)
METEO device EGU (CZ)
IceMonitor Combitech (SE)
ICECylinder (Prototype) FMI (FI)
EAG 200 (D) No longer manufactured
e Rotating Multicylinder (Prototype) VTT (FI), STATNETT (NO)
Gerber Gerber Scientific Inc. (USA)
f Labko LID-3210C Wavin-Labko (FIN)
g Instrumar IM101 V2.4 Instrumar Inc. (CA)
h
i
Jokkmokk
IceMeister
Segerström (SE)
www.newavionics.com
The ISO 12494 standard ice collector mentioned in the preceding chapters has been built in
one version in Sweden (Combitech: automatic weighting, free rotation) and two in Finland
(Digita: automatic weighing, forced rotation [28], FMI: manual weighting, forced rotation). A
further development is presently being designed in Switzerland to yield ISO compatible sen-
40/110
sors with automatic weighting and forced rotation (Markasub) within the framework of a na-
tional project linked to the COST-727 Action. Devices similar to the ISO ice collector have
been used at some locations in the past as well [29].
For the detection of the meteorological icing (see definition above) M
icing
, there are a few sys-
tems which are either available on the market (Rosemount Goodrich), or available as proto-
types (HoloOptics, Infralytic, Vibrometer/Boschung, etc.)
8.2 Data requirements for icing models
Icing models can be run for sites where no specific in-situ icing data is available, but where
basic meteorological information is provided by Automatic Weather Stations. The following
list summarizes the parameters which must be provided in order to characterize the start/end
of all types of icing events as well as the total mass of ice accreted:
- Wind
- Temperature
- Humidity
- Precipitation
- Visibility
- Cloud base
- Radiation
- Present weather (precipitation type)
Ideally, icing models require information concerning LWC and median volume diameter
(MVD) of the droplet size distribution. Unfortunately, these parameters are measured only at
research test centres, if at all. However, it may be possible to use small-grid weather forecast
models to perform an approximation of these values (see WG1 report). Such models also in-
clude information about vertical air stability which influences the LWC and MVD.
8.3 Verification of data
Assessment and verification of existing data is scarce, unless visual observations or camera
information or both are available at the station. Unfortunately, measurements of atmospheric
icing are presently not included in the existing meteorological WMO standard observations.
Some information may be found in the annexes related to section 6.
41/110
9 Experiences with automatic instruments for ice meas-
urements
A number of instruments and resulting data are presented in the following and references are
made to more detailed descriptions in the annexes.
9.1 ICEmeter
Manufacturer: Institute of Atmospheric Physics, Prague, CZ.
Experiences made with the automatic measurement of the mass of icing accumulated on the
collector of icing sensor developed in the Institute of Atmospheric Physics, Prague, Czech
Republic. First, the measuring device and its basic specifications are briefly described. Then
a short description of sites where the measurements were performed is given and the selected
data presented. Next the results of the measurements and experience with the sensor are dis-
cussed along with outlines for its possible improvement.
See Annex 12 for detailed explanations.
9.2 Labko ice detectors
Wavin-Labko Oy has one ice detector model LID-3210C to offer, LID (Labko Ice Detector).
Ice detection of Labko ice detectors is based on longitudinal wire waves [30]. Usability of the
device has been found to be good. It is also reasonably easy to get acquainted with the device
and to adjust the parameters of the device to correspond with different icing climates. How-
ever, Labko's different versions have suffered from snow-induced icing indications and inabil-
ity to melt all the ice. It seems to have a lower correlation with ice detections by humidity data
than the Instrumar sensor [21].
See Annex 13 for detailed explanations
9.3 Rosemount/BFGoodrich, model 0872J
The Rosemount 0872J Ice Detector
9
(prototype) is designed for various applications from
meteorological measurements to radio tower de-icing programs [31]. The Rosemount ice de-
tector uses an ultrasonically axially vibrating probe to detect the presence of icing conditions.
The sensing probe is a nickel alloy tube mounted in the strut at its midpoint with one inch
exposed to the air-stream. As the ice detector enters an icing environment, ice collects on the
sensing probe. The added mass of accreted ice causes the frequency of the sensing probe to
decrease in accordance with laws of classical mechanics. Ice detector software monitors probe
frequency and detect this decrease. The ice signal activates at 0.52 mm ice accretion. At the
same point the internal probe heater power is applied until the frequency rises to a predeter-
mined set point. The probe is then heated to melt the ice. Once de-iced, the sensing probe
9
This instrument is different from the Rosemount Model 872C3 Sensor, known as a freezing rain sensor and
used in the ASOS network in the USA [31].
42/110
cools within a few seconds and is ready to sense ice formation again. The ice detector outputs
include ice detection indication and fault status indication.
See Annex 14 for detailed explanations
9.4 EAG 200
The ice load sensor EAG 200 is the automatic icing measurement device that is used at
DWD‘s measuring sites at present (see Annex 2, Figure A.2). It measures the weight of ice
accumulated on a vertical pole by the use of an electro-mechanical scale system. The com-
parison of EAG 200 results with those of manually operated poles show the reliability of
EAG200‘s data. Experiences from continuous long term measurements show, that the system
operates well even for short icing periods and for small amounts of ice accretion.
See Annex 15 for detailed explanations.
9.5 Gerber
The UK Meteorological Office sourced two instruments (Gerbers) capable of measuring the
LWC of the air. Two Gerber PVM-100 instruments were installed at the EA Technology se-
vere weather test site at Deadwater Fell test site on the English/Scottish border in the UK.
The instruments are laser devices which monitor the scatter and transmission of a laser beam
over a distance of around 450 mm. They are particulate volume monitors and have a 0 – 10
VDC output which give a measurement range 0.002 – 10 g/m3. One PVM-100 was mounted
and aligned north/south and the other aligned east/west as shown in Annex 16.
See Annex 16 for detailed explanations.
9.6 METEO device
EGU Brno has developed an instrument for the measurement of temperature, ice mass and the
velocity and direction of wind
The measuring device METEO is made of stainless steel and it has no moving parts requiring
frequent maintenance. It consists of a body with sensors and of a measuring rod fixed verti-
cally downwards.
See Annex 17 for detailed explanations.
9.7 IceMonitor
The IceMonitor is an instrument aimed for automatic weighing ice deposit on a vertical steel
pipe, and it has been designed according to the recommendations in ISO 12494. The steel pipe
has 30 mm diameter, it is 0.5 m long, and it can rotate freely to allow ice build-up to form
cylindrically.
The instrument has been installed in two sites in Norway - the first in 2003 close to Drammen
(west of Oslo) and the second in the far north (in Finmark 2005) - and in two sites in Sweden
2005. The Åre site is located in the middle of Sweden at a height of 1300 m above see level,
and at this site both weather - air temperature, relative humidity and wind speed/direction -
43/110
and ice load data is measured. Measuring data from the Swedish site in Åre includes a number
of heavy icing events during winter 2005-2006.
Manufacturer: SAAB Technologies, Sweden
See Annex 18 for detailed explanations.
9.8 T20-series Ice Detectors
All T20-series ice detectors are based on a patented digital optronic ice-indicator that indi-
cates the presence of any type of atmospheric ice including clear ice. It comprises a head with
an IR emitter and a photo detector and a probe.
The T20-series indicators:
• come in single-direction (±45° upwind) or omni-direction versions
• indicating either all icing types including clear ice or rime ice/wet snow growth only.
An indication of ice is made if more than 95 % of the probe is covered with a 50 μm thick
layer of clear ice or a 90 μm thick layer of other types of ice. Testing of T20-series sensors
have been carried out at locations noted in Annex 19.
Manufacturer: HoloOptics, Sweden
See Annex 19 for detailed explanations.
9.9 Instrumar IM101
Instrumar IM101, V.2.4 is an ice detector based on measuring the electrical impedance and
surface temperature [32]. Its long-term use has shown it to be quite durable and it seems to
detect icing in reasonable agreement with humidity indications and video recordings.
44/110
10 Long term recommendations for ice measurements in
Europe
10.1 Regional variability
As icing conditions and icing climate vary significantly within Europe it is important to per-
form the measurements at different parts of Europe, noting the different climatic aspects:
- Northern European mountains with long icing periods under wide temperature and
humidity range and lack of solar radiation (typically rime ice)
- Alpine regions with icing strongly depending on the altitude (typically rime ice and
wet snow)
- Central and Southern European mountainous areas with icing and strong sunshine pe-
riods causing numerous melting and freezing consecutive events (typically glaze and
rime ice)
- Maritime regions in Western Europe (typically wet snow)
As discussed above, meteorological instruments will respond differently depending on the
location of their installation. At cold climate sites or at sites with temperatures below 0 ºC,
meteorological sensors are subject to different types, durations and formation of icing. As no
measurement data on atmospheric icing are available at the European Meteorological Services
only a rough estimation of classification on severity of icing can be implemented at this stage.
On the basis on the previous work on the classification of icing climates made mainly for
wind energy purposes and the recordings made within the EUMETNET project the classifica-
tion shown in § 5.6 was proposed. For example, it is typical that the duration of ice loads upon
structures in class S5 is longer than duration of meteorological icing events.
10.2 Requirements for measuring sites
Consequently, a number of test centres should be established and operated in Europe (and in
the world). There are however two points of view which will have to be combined in order to
fulfil the requirements of the different communities.
10.2.1 Specifications of users
There are a number of potential users either for ice detectors or for icing data in different
fields. Ice measuring centres will have to fulfil the requirements regarding:
- Modelling
- Electrical power lines
- Wind turbines
- Towers and masts
- Road/railway safety networks
- Cable cars and others equipments for tourism purposes
- Airport safety
- Weather forecasters: development and control procedures for models
- Climatologists: maps, long-term representativity and climate change
- Others
45/110
10.2.2 Specification of purpose
Ideally, instrumentation at all weather stations located in cold climates should be extended
with icing recording systems, which is presently not the case. Therefore, the goal of establish-
ing icing test centres is to provide an adequate infrastructure to fulfil the requirements of the
above-mentioned potential users. In a general way, the following aspects will have to be dealt
with:
- 3-6 test sites in Europe covering the different climatic environments (e.g. Luosto,
Finland; Guetsch, Switzerland; Mt Aigoual, France; Studnice, Czech Republic; Spain;
Germany; UK; etc.)
- 2-3 „reference“ instruments common to all sites: standardized testing and certification
procedures and standardized data format
- Flexible infrastructures for the installation of different test beds (e.g. for wind turbine
and power line testing)
- Common monitoring and quality control procedures (for future certification)
- Complete high quality data sets for forecasting and climatology
10.3 Permanent forum for monitoring icing in Europe
In relationship with the establishment of icing test centres, a permanent forum for monitoring
icing in Europe needs to be established. The proposed way to achieve this goal is to establish
projects on the European level.
Appropriate task specifications will be handled within Phase 2 of the present COST-727 Ac-
tion in relationship with the establishment of the test centres. There are two complimentary
activities which will have to be further analyzed:
- Launching of a new EUMETNET and/or EU project for the establishment of long-
term icing test and observation sites in Europe
- Integration of icing measurements in meteorological networks under the umbrella of
WMO/CIMO
46/110
Annex 1 Measurements in Finland
Technical Research Centre of Finland
Ice detector data of Technical Research Centre of Finland (VTT) is presented in Table A. 1.
Data consists of status parameter, which tells whether ice was accumulating or not. Data was
collected at 1Hz. Ten-minute averages, min, max and standard deviations were calculated and
stored. Amplitude levels of wire wave signals (see §6) have also been recorded because all the
performed icing measurements have been closely connected to development of the used ice
detectors. The development of ice detectors has been driven by the need for development of
blade heating systems for wind turbines.
Table A. 1 Ice detector data of Technical Research Centre of Finland.
Site LAT LON Ice detector Start End Parameters
Olostunturi 67.55 23.48 Labko Ice detector 3200 12/1999 - Status (icing/no icing)
Pori 61.37 21.30 Labko Ice detector 3500 9/1999 - Status (icing/no icing)
Lammasoaivi 68.47 21.20 Labko Ice detector 3500 2/1999 10/1999 Status (icing/no icing)
Pyhätunturi 67.01 27.13 Labko Ice detector 3500 11/1998 7/2000 Status (icing/no icing)
Ylläs 67.4 24.15 Instrumar IM101 10/1991 - Status (icing rate)
VTT has within the COST 727 Action also studied humidity measurements in icing condi-
tions. These studies [33] showed that
- The conventional humidity instruments measure incorrectly in icing conditions
- A heated Vaisala HMP243 (and HMT337) measures correctly in icing conditions
- It seems to be possible to detect rime icing by correct humidity measurements
FMI test station: Luosto
The Luosto test station is located in northern Finland, on the top of Luosto fell (500 m asl, N
67 08’, E 26 54’). The station is operated by Finnish Meteorological Institute (FMI). The test
station was established for the EUMETNET SWSII project, which studied the effect of icing
on meteorological instruments and produced specifications for ice-free sensors. The platform
for the instruments at Luosto was set up in the winter 2000/01. Since then icing measurements
and meteorological measurements with ice-free sensors have been made at the site. Experi-
ence from the first winter was used to develop the measurement systems and the infrastruc-
ture. Performance of the instruments is monitored continuously with video cameras.
The icing climate of the site has been characterized as site class A (see §5): an elevated site
inland in northern Europe with harsh and frequent icing climate.
The Labko LID-3503 and the Rosemount 0872J were installed at the site in the winter
2001/02. Both ice detectors indicate the presence of icing conditions. According to the meas-
urements performed with the LID-3503 ice detector in winter 2001/2002, it was noted that the
sensor is inadequate for icing measurements in extreme conditions. Measurements with this
sensor were not continued.
Four video cameras are monitoring operation of the instruments. The cameras also give in-
formation on visibility (fog and clouds). Digital pictures are sent every 10 minutes to the FMI
via the network.
47/110
The Vaisala’s FD12P weather sensor is used for visibility observations. The FD12P measures
meteorological optical range (MOR) from 10 m to 50 km.
Also temperature, dew point temperature and relative humidity measurements are analysed to
get more accurate data on events and duration of icing.
The ICEcylinder has been built by FMI according to the International Standard ISO 12494.
The cylinder with a diameter of 30 mm and length of 0.5 m is installed with the axis vertical
and forced rotating around the axis. Accumulated ice load is weigh manually.
A device used for icing rate measurements is manufactured by Institute of Atmospheric Phys-
ics, Academy of Science of the Czech Republic. The ICEMETER measures the mass of ice
accumulated on the sensor surface.
Table A. 2 List of instruments installed for icing measurements at the Luosto fell. Technical
availability: the sensors are in operation and data (erroneous or correct) are received by the
data acquisition system.
Instrument Technical data availability Parameter
LID-3503 ice detector 01.10.01-30.4.02 Icing occurrence
Rosemount 0872J ice detector 01.10.01 -> Icing occurrence
ICEMETER CZ - IAP 09.03.06 -> Icing rate
ICEcylinder 01.10.01 -> Icing rate
2-4 video cameras 01.10.01 -> Icing occurrence
FD12P weather sensor 01.10.01 -> Visibility and rain
Figure A. 1 The test platform at Luosto with installed severe weather sensors.
48/110
DIGITA / VTT CAMPAIGN
Digita Oy (former Finnish Broadcasting Co., Distribution Dept.) and Technical Research
Centre of Finland (VTT) made measurements of icing on tower structures, tests on ice detec-
tors, measurements of drop size and liquid water content of clouds and comparisons of mete-
orological instruments on hilltops in severe icing conditions in Finland in a four year national
project in 1986-1990 and later. In this project both an operating 128 m tall TV tower and a 7.5
m test tower were equipped with load cells, so that the ice load on them could be continuously
measured. The rotating multi-cylinder method was developed to a stage that would allow fur-
ther development to automatic measurements of drop size and liquid water content during
icing. The results of these activities have been reported in a number of publications
[19,20,34,18,35,21,33].
49/110
Annex 2 Measurements in Germany
German Meteorological Service (DWD)
Icing measurements were carried out at altogether 40 locations in the eastern part of Germany
during 1965-1990 [36,37], see red points in the left part of Figure A. 2). The number of loca-
tions has changed slightly during the years. However, up to 35 locations were operated simul-
taneously.
The measurements were performed at all locations by the use of manually operated vertical
icing poles with a diameter of 35 mm. A standard measurement procedure defined an exposi-
tion period of 24 hours: The pole was exposed every day at 8:30 a.m. at a height of 2 m a.g.l..
In the case of ice accumulation the pole was exchanged after 24 hours. The icing mass was
determined and additional information (e.g. icing diameter and direction, ice vane dimension,
icing type(s)) was compiled. By using the standard measurement procedure the ice accretion
was interrupted at least every 24 hours. For ice accretion periods exceeding 24 hours or on
occasions of multiple icing events the standard procedure underestimated the maximum ice
masses.
Time Period: 1965-1990,
Red points: 24-hour measurements,
Blue circles: Icing cycle measurements.
Time Period: 1990 (partly) - Present,
Continuous icing measurements at all loca-
tions.
Figure A. 2 Locations of Icing Measurements in Germany, depicted in topographic maps for
different time periods and resolutions of measurements.
Therefore the measurement procedure was modified at 11 locations in 1978 (blue circles in
the left part of Figure A. 2) so that a second icing pole was exposed at those locations. In the
event of ice accumulation all the measurements of the standard procedure (ice mass, addi-
tional information) were carried out after 24 hours. Afterwards the pole was re-exposed for
another 24 hour period. By the application of this procedure it was possible to improve the
Arkona
Falkenberg
Zinnwald
Chemnitz
Kahler
Asten
Hohenpeißenberg
50/110
knowledge about whole icing cycles (accumulation and loss of ice), so they were called icing
cycle measurements. One of the main results of icing cycle measurements was ‘real’ (in dis-
tinction to the standard procedure) maximum ice mass.
The number of locations with icing measurements has been reduced to a total of five since
1991. The icing sensors are implemented at 5 locations (for details see the right part of Figure
A. 2 and Table A. 3. They are operated continuously at a single height level. Furthermore, at 3
locations the icing measurements are still carried out by the use of manually operated icing
poles. The latest advancement during the ongoing winter period are icing measurements at
three heights (10 m, 50 m and 90 m above ground) at DWD’s Meteorological Observatory
Lindenberg.
Table A. 3 Overview of present locations with icing measurements in Germany
Location
Height
above sea
level (m)
Measurement
height (m)
Measurement device
Arkona 42 2, 5
Chemnitz 418 2, 5
Zinnwald 877 2, 5
EAG 200 (Annex 15), continuous, accumu-
lation and Icing pole, intermittent, manual
Kahler Asten 839 10
Hohenpeißenberg 977 10
Lindenberg 73 5, 50, 90
EAG 200 (Annex 15), continuous, accumu-
lation
51/110
Annex 3 Measurements in Slovak Republic
Extensive experimental data from the measurements of atmospheric icing can be obtained
from the Slovak Hydrometeorological Institute (SHMI). This Institute is an administrator of
the meteorological networks in Slovak territory. There are both visual observations and meas-
urements of the amount of atmospheric icing. Visual observations of atmospheric icing enable
an analysis of the lengths and annual occurrences of atmospheric icing events to be made in
individual seasons. There are a large number of regular weather stations where atmospheric
icing observations are carried out. Measurements are performed on a horizontal pair of or-
thogonal (N-S and E-W) wooden rods. The diameter of the rods is 32 mm and length is 1 m.
Data are acquired 3 times a day at 7, 14, 21 hrs. The sites where the atmospheric icing meas-
urements are made are fewer and are given in Table A. 4. [38]
Table A. 4 List of sites in the Slovak territory, where the parameters atmospheric icing are
measured.
Location Latitude Longitude Altitude Recording Since
Bratislava-
Koliba
48
o
10’ N 17
o
06’ E 289 m asl Climat. term Dec. 1974
Hurbanovo 47
o
52’ N 18
o
12’ E 115 m asl Climat. term Dec. 1971
Chopok 48
o
56’ N 19
o
35’ E 2004 m asl Climat. term Jan. 1957
Jaslovské
Bohunice
48
o
30’ N 17
o
50’ E 177 m asl Climat. term Dec. 1971
Kamenica
nad Cirochou
48
o
56’ N 22
o
00’ E 178 m asl Climat. term Jan. 1972
Košice 48
o
42’ N 21
o
16’ E 206 m asl Climat. term Dec. 1972
Lomnický
štít
49
o
12’ N 20
o
13’ E 2634 m asl Climat. term Jan. 1957
Luþenec 48
o
20’ N 19
o
40’ E 187 m asl Climat. term Dec. 1971
Nitra 48
o
19’ N 18
o
05’ E 145 m asl Climat. term Jan. 1973
Poprad 49
o
04’ N 20
o
15’ E 703 m asl Climat. term Jan. 1973
Sliaþ 48
o
38’ N 19
o
09’ E 312 m asl Climat. term Dec. 1976
Stropkov 49
o
12’ N 21
o
39’ E 209 m asl Climat. term Dec. 1971
Telgárt 48
o
51’ N 20
o
11’ E 901 m asl Climat. term Jan. 1972
52/110
Annex 4 Measurements in Norway
Icing on overhead lines varies strongly with height above sea level, exposure to air of mari-
time origin and local topography [39]. Therefore lines of 100 – 200 km length have a variety
of local exposures in mountainous terrains. In Norway 100% of the electricity production is
from hydropower. As the bulk production is on the western coast of the country, where the
precipitation rates are highest, and the main consumption is in the south-eastern region around
Oslo, there are many high voltage lines crossing the mountains and thereby also exposed to
severe ice loads, as indicated in Figure A. 3 and Figure A. 4.
Figure A. 3 A cross section of southern Norway
Figure A. 4 420 kV line crossing the mountain divide (1 100 m asl) between western and east-
ern Norway (Photo: S.M. Fikke)
Probably the world’s largest ice load on an electric power line was observed in Norway in
1961. Figure A. 5 shows an example of the accretion that measured 1,4 m x 0,95 m and
weighed 305 kg/m. This is probably the clearest example on how topography and exposure
Rain and icing
Clouds
North Atlantic Ocean Central mountain range
W E
0
O
C
53/110
influence the icing conditions on a power line. This line was feeding a radio and TV transmit-
ter 1 412 m above sea level. It appeared to be impossible to maintain this line no matter how
short the spans would be, how strong the (wood) poles were made and how strong the steel
conductors were. It was built on the top of a mountain ridge more or less parallel to the coast,
and therefore maximum exposed to the humid south-westerly winds from the Atlantic Ocean.
Only when a new line was built up from the leeward side of the mountains the power supply
to this radio and TV transmitter became stable.
Figure A. 5 In-cloud icing on a 22 kV line feeding a TV tower 1 400 m above sea level in
south-western Norway. (Photo: O. Wist)
The topography influences the icing differently depending icing type. Freezing rain occurs
mainly in basins and depressions where cold air can be trapped while warm air with precipita-
tion may intrude the air aloft (temperature inversion). In-could icing occur only above cloud
base, but the cloud base varies significantly with topography. A mountain (ridge) only 50 –
100 m higher in the upwind direction may be sufficient to reduce this icing to a minimum.
Wet snow however may occur in all altitudes and also on the leeward side of mountains and
ridges.
The great dependence on topography may result in a likewise great variety of expected ice
loads along a line passing through zones with different exposure and climatic conditions from
one sub-station to another. As a result of this variability all transmission lines belonging to the
main grid of Norway are designed span by span according to the expected ice and wind loads
for the particular span. For lower voltage lines in the distribution system the loads are given
regionally according to the National code.
Since the late 1970s the Norwegian Power Grid Company, Statnett, has operated more than
20 ice racks of the type shown in Figure A. 6 and Figure A. 7 in remote and exposed locations
in mountainous terrain in Norway. Figure A. 8 shows the distribution of the locations of the
measuring stations and Figure A. 9 shows an example on extreme value calculations.
54/110
Figure A. 6 Norwegian rack for measuring
ice loadings in remote areas
Figure A. 7 Details of dynamometer suspen-
sion in the rack
Figure A. 8 Location of the 16 ice measuring
sites in Norway
Figure A. 9 Example of extreme value calcu-
lations of ice loadings from the ice racks
Oslo
Stavanger
Trondheim
Tromsø
Bodø
58 N
60 N
62 N
64 N
66 N
68 N
70 N
12
13
14
11
8 9
10
15
1
2
4
3
5
6
7
16
17
km 100 200 300 400
55/110
Annex 5 Measurements in the Czech Republic
EGU Brno carries out measuring of icing on racks at test site Studnice thus continuing the
long time-series of icing data [40]. It is located about 60 km northwest from Brno at 800 m
above see level. The whole site was built in the late 70’s and has been in operation since
1980. There are 2 spans (ca 250 m each).
Table A. 5 Location of racks at Studnice
Location Latitude Longitude Altitude
Studnice 49˚36’27” N 16˚05’7” E 800 m asl
Figure A. 10 Test site at Studnice (Photo: J. Sabata)
Various measurements are performed here:
- Ice measurement on conductors and measuring rods of various diameters
- The measurements on samples of stranded isolated conductors
56/110
- Observation of icing with respect to altitude gradient.
For measurement of ice accumulation we use different instruments with horizontally and ver-
tically oriented rods (with diameter of 30 mm).
Table A. 6 Location of measurement devices on racks at Studnice
Measuring device Measurement height (m) Output Since
Vertical rods (length of 0.5 m) 5, 10, 20, 30, 40, 50 Digital 1997
Horizontal rods (length of 1 m) 10, 30 Digital 1997
Temperature 5, 10, 20, 30, 40, 50 Digital 1997
Meteo device (length of 0.5 m) 10 Digital 2001
Figure A. 11 Vertical, freely rotating rod with sensor, at test site Studnice, 17.12.2004
(Photo: J.Sabata)
57/110
In addition to measurement at Studnice station fourteen Meteo devices are installed at loca-
tions supplied by two regional utilities. Two devices are located in the area of the regional
utility VCE and the others in the area of regional utility JME. The Meteo device locations on
the territory of JME cover the whole area susceptible to regular icing.
Table A. 7 Location of Meteo devices
Location Latitude Longitude Altitude Output Parameters* Since
Cotkytle Digital IM, T, WS, WD 1999
Novy Hradek Digital IM, T, WS, WD 2003
Beranov 49˚24’ N 15˚39’ E 554 m asl Digital IM, T, WS, WD 2001
Buchlov 49˚06’ N 17˚18’ E 477 m asl Digital IM, T, WS, WD 2001
Hlína 49˚06’ N 16˚25’ E 441 m asl Digital IM, T, WS, WD 2001
Kasarna 48˚53’ N 16˚00’ E 379 m asl Digital IM, T, WS, WD 2001
Klucov 49˚10’ N 15˚55’ E 570 m asl Digital IM, T, WS, WD 2001
Kralovec 49˚08’ N 18˚02’ E 634 m asl Digital IM, T, WS, WD 2001
Nyklovice 49˚36’ N 16˚20’ E 722 m asl Digital IM, T, WS, WD 2001
Kralovec 49˚08’ N 18˚02’ E 634 m asl Digital IM, T, WS, WD 2001
Protivanov 49˚28’ N 16˚50’ E 685 m asl Digital IM, T, WS, WD 2001
Predin 49˚12’ N 15˚40’ E 640 m asl Digital IM, T, WS, WD 2001
Ruda 49˚19’ N 16˚07’ E 580 m asl Digital IM, T, WS, WD 2001
Vyskovec 48˚56’ N 17˚49’ E 761 m asl Digital IM, T, WS, WD 2001
*) Measured values: IM (Ice mass), T (temperature), WS (wind speed), WD (wind direction).
Icemeter sites (for results see section 6)
Milesovka: Most of the measurements were carried out at the top of “Milešovka” mountain,
which is the highest peak of the tertiary volcanic range of “ýeské StĜedohoĜí”. A synoptic
meteorological station, now belonging to our Institute, was built on Milesovka’s summit
(837m asl). Milesovka has a shape of isolated forested cone, which exceeds the surrounding
terrain by approximately 300 m. The steepness of slopes ranges from 20° to 30°. Concerning
temperature, the long time average is 5.1 °C, absolute minimum was -28.3 °C and absolute
maximum reached 34.7 °C. The average of annual precipitation is 564 mm. The mean wind
speed is 7.7 m/s with most frequent winds from northwest, west and southwest. From the win-
ter 2003/2004 two Icemeters, situated at different height above the ground (see Figure A. 12),
have been operated.
Nová Ves: Since 2005, one Icemeter has been situated close to the wind turbine in Krušné
hory (Ore mountains), near to the village of Nová Ves. Krušné hory is the region most suit-
able for wind energy production in the Czech Republic. Unfortunately that region is also af-
fected by severe icing. It is probably most exposed region in the whole country, concerning
the severe ice episodes.
58/110
Figure A. 12 left – Observatory at the Milesovka peak with marked position of Icemeters;
right – location of Milesovka mountain and Nova Ves in the Czech Republic
Other sites:
Two Icemeters have been in operation in Austria (Obersthralbach, Sternstein), both of them
situated close to wind power station. One piece was tested in Luosto, Finland during the win-
ter 2004/2005. The Icemeter was also located temporarily in other sites in the Czech Repub-
lic. Due to the short time of location at these sites, results are not available.
59/110
Annex 6 Measurements in UK
Deadwater Fell
Deadwater Fell is one of EA Technology’s field stations at sites located to encompass a full
range of weather conditions: wet and dry snow, hard and soft rime ice and glaze icing (freez-
ing rain). The sites are situated in exposed locations at heights between 550m and 750m and
range from hilltop to open moorland. Overhead line conductors are erected in test spans of
between 90m and 200m. The data presented here comes from this and two other sites located
on Susseter Hill in Shetland (an island off the north coast of Scotland) and south west Scot-
land (Green Lowther). These two sites have since been closed down and all work is now con-
centrated on the Deadwater Fell site on the English/Scottish border.
The site consists of a 200m test span with terminal H-poles supported by 14 stay wires each.
The test spans are orientated North-South and suffer from severe winds as well as ice inci-
dents and blizzards. Figure A. 13 shows the full 200 m span viewed from below the summit
ridge and Figure A. 14 shows the rotating rig. Provision is made for shorter span lengths when
appropriate. The site has the capacity to provide full meteorological measurements on-site.
Load cells are used to measure loads on the conductors.
Figure A. 13 The full 200m span at Deadwater Fell
60/110
All the data is collected and stored at the site. It is downloaded automatically every 24 hours
via a mobile telephone, for analysis at EA Technology at Capenhurst, near Chester, UK. Pro-
vision can be made for close up video coverage to identify ice shape. Conductors of different
sizes and of different span lengths can also be installed to provide direct measurement of the
total force on the conductors at their connection point to the supports. This force is made up
of:
a) Conductor weight
b) Ice load
c) Wind-on-ice load.
Item a) is known and items b) and c) are measured together. This system can be used to
evaluate the ice load by calculation of the wind load. The latter can be calibrated in above
0°C incidents with no ice present. However, the main concern with overhead lines is the total
wind and ice load, so from a practical point of view it may not be necessary to separate these
components. The site also has a rotating rig to test conductor samples (Figure A.14). This rig
is designed so that the samples are always facing normal to the prevailing wind.
Test Spans
The H-poles are designed to withstand impulsive forces from the galloping of large conduc-
tors and at the same time, to withstand blizzard conditions. Intermediate poles are installed as
required for shorter spans. In the test spans, each conductor is fitted with a load cell and, if
required, a vibration monitor. Video surveillance can detect any galloping, rotation or general
conductor movement under wind and ice loads. The spans are monitored 24 hours a day
throughout the year by time lapse video cameras with low light level sensitivities down to 0.1
lux. The surveillance also allows short periods of real time coverage every 30 minutes. The
cameras are mounted within specially adapted housings with insulation, internal heating and
externally wound heating tape to reduce ice growth and are used to give close-up and long
distance views using, if required, barely visible, environmentally friendly infra red floodlight-
ing.
61/110
Figure A. 14 The rotating rig with snow accretion.
Conductor Data
Ice densities
Data from the test spans at Deadwater Fell was investigated for direct comparison of two bare
conductors at a span length of 200 m. A series of load comparisons between Upas (diameter
24.7 mm) and Lynx (19.5 mm) conductors was extracted at various ice loads, but at wind
speeds around 20 knots. The load cells will, of course, measure the total wind and ice load.
The measured ice loads measured were between 1.6 and 8.8 kg/m. Ice load density was meas-
ured on occasions when the normally un-manned site was visited. The accretion density for
rime ice gave an average value of 570 kg/m³ and for wet snow 825 kg/m³. This is in line with
field data published for Iceland [41].
Ice loads
Shackleton et al [42], under an EATL contract, calculated the rime ice and wet snow loads for
a range of conductors. The initial wet snow accretion rate (Model 1) is:
Mass = [ A + B.D + C.D² ] . 10
-3
kg/m
where A = 0.449
B = 1.794
C = -0.0003
and D = Conductor Diameter (mm)
62/110
The longer time version is called the Large Relaxation Time (LRT) Model, which gives ice
loads according to:
Mass = [0.243 + 1.792.D + 0.0034.D² ] . 10
-3
kg/m
The factors A, B and C vary with time with A reducing and C increasing.
The EATL field data is given in Figure A.15 with a trend line polynomial in which:
where A = 0.013
B = 0.097
C = 0.1409
The formulae imply that over time the conductor diameter has an increasing effect on the
modeled snow load.
Snow Loads against conductor diameter
y = 0.0013x
2
+ 0.0971x + 0.1409
0
1
2
3
4
5
6
0 5 10 15 20 25 30 35 40
Conductor diameter (mm)
S
n
o
w

l
o
a
d

(
n
o
r
m
a
l
i
s
e
d
)
Figure A. 15 Field data with polynomial trend line
Sufficient information has been extracted from the database to give comparative values for 28
conductors. In order to eliminate the different field conditions, all the data is related back to
the original ‘control’ conductor, 32 mm² Hard Drawn Copper (8.0mm diameter). This was
used as the ‘control’ at the Green Lowther and Shetland sites. At Deadwater, the control con-
ductor was changed to the Hazel conductor (9.9mm diameter), which is a bare AAAC 60 mm²
conductor. Hazel was used extensively at Green Lowther to establish a relationship between
this and the 32 mm² Copper conductor and so maintain the validity of the data from each site.
It can be seen in Figure A. 16 that assuming a constant accretion thickness for any conductor
size (as used in current UK wind/ice maps in ET111) gives a reasonable fit for conductors up
to Oak (14 mm diameter AAAC). For larger conductors up to 40 mm in diameter, the con-
stant thickness assumption underestimates loads as compared with the EATL field data.
63/110
Field data against contant accretion thickness (ET111)
0
1
2
3
4
5
6
0 5 10 15 20 25 30 35 40
Conductor diameter (mm)
S
n
o
w

l
o
a
d

(
n
o
r
m
a
l
i
s
e
d
)
Constant accretion thickness (ET111)
Field data
Figure A. 16 Snow load field data plotted against constant accretion thickness scenario for
the same conductors.
Figure A. 17 Typical accreted ice loads on different conductors after a blizzard at Deadwater
Fell
64/110
Annex 7 Measurements in Sweden
Three different sensors (IceMonitor, HoloOptics and Jokkmokk) are currently being tested
under field conditions in Sweden and Norway.
The IceMonitor sensor is presently connected to monitoring stations located in
- Åre (SE)
- Ritsem (SE)
- Drammen (NO)
At these stations measurements of weather parameters, such as air temperature, relative hu-
midity, wind speed and direction, and in some cases also precipitation, are being carried out.
Measurements are normally made once every half hour (in Drammen once every 10 minutes).
Data is retrieved to a server in Östersund via radio/radiolink or in some case via telephone
modem.
A camera is monitoring the IceMonitor sensor at the site in Åre. Icing of this camera has been
a problem during the winter 2005/2006 – and the installation and de-icing of the camera will
be modified before the next icing season. The monitoring equipment was originally designed
for use in road weather information systems and it is capable of handling a number of differ-
ent sensors, including camera – which stores files in jpg format.
At the site in Åre, several icing events occurred during the winter 2005 /2006. A couple of the
icing events have been severe with ice loads of more than 40 kg. (see figure below.) During
the same period, no significant icing occurred at the site in Ritsem (Sweden) or at the test site
in Drammen (Norway).
Figure A. 18 Icing measured at Åre during winter of 2005/2006 (Ice load signal error oc-
curred by end of Feb.
65/110
The HoloOptics T23 Clear-Ice Indicator and the T26 Icing-Rate Sensor beta versions have in
periods been tested at the following locations during 2003-2006:
- The Suorva wind power plant, Sweden. 150 km north of the polar circle. In co-operation
with FOI (The Swedish Defence Research Agency), see figure A.19.
- KTH (The Royal Institute of Technology), Stockholm, Sweden
- Bromma Airport, Stockholm, Sweden. In co-operation with the airport authorities
- Åre Ski Resort, ski lift protection. Cooperation with Combitech
Sourva Icing Rate
0
50
100
150
200
495 500 505 510 515 520
hour
i
c
i
n
g

r
a
t
e

g
/
m
2
h
Icing Rate
1h average
Figure A. 19 Icing rate measured at Suorva, Northern Sweden
66/110
Annex 8 Measurements in Bulgaria
The initial icing observations in Bulgaria were started in the late 50-s and in the beginning of
the 60-s of the 20th century in some stations of the meteorological institute. The stations were
chosen to cover the territory of the whole country. The ice measurement device was a couple
of perpendicular conductors with diameter of 5mm located in the directions N-S and E-W.
Data about the meteorological conditions during icing events and the final ice depositions
have been collected. A list of all stations with initiated icing observations is presented in the
table below.
The collected icing data in some of the stations are very short and/or not reliable due to the
bad organization of observations, non motivated work as well simply to the fact that icing had
been observed rarely in these regions for the whole period of measurements. However, other
stations have collected long time series of icing events or short but very detailed measure-
ments. These stations are highlighted with yellow in the table bellow (some of them still re-
port icing data). They represent the regions where icing occurs most often and usually is very
severe.
We have also additional information from the National Electric Company with data about
icing depositions in cases of damages of the power lines in the period 1962 –1990. All these
data have been used to characterize the typical icing conditions of the territory of the country.
As example of these conditions the change of the number of icing events in the mountain re-
gions of Bulgaria is depicted on the figure.
Our intention now is to continue with icing measurements on the places with good historical
data using further visual observations as well as some ISO12494 icing sensors (rotated and
non rotated).
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70
number of cases
a
l
t
i
t
u
d
e
,

m
Figure A. 20. The change of the mean number of cases with the altitude in the mountain re-
gions
67/110
Table A. 8: Meteorological stations with icing observation
Station
Altitude,
m
Latitude Longitude
Part of coun-
try, mountain
Start/End of
icing observa-
tion
Data
quality
Kozloduy 34 43.47 23.44 NW 1969-1970 not good
Vidin 31 43.59 22.51 NW 1992-1993
relatively
good
Orjahovo 29 43.43 23.58 NW 1955-1964
relatively
good
Gramada 257 43.50 22.40
NW, Stara
planina
1955-1986 very good
Petrohan 1400 43.25 23.14
NW, Stara
planina
1968-1983 good
Peak Murgash 1678 42.50 23.40
NW, Stara
planina
1980-1994 very good
Vakerel 851 42.55 23.17
W, Lozenska
planina
1959-1989 very good
BAC station,
Vitosha
1485 42.41 23.10 W, Vitosha 1968-1974 not good
Borovec 1244 42.15 23.36 SW, Rila 1960-1975
relatively
good
Smoljan 1180 41.34 24.12
SW, The
Rhodopes
1968-1978
relatively
good
Pamporovo 1599 41.39 24.41
SW, The
Rhodopes
1976 not good
peak Snejanka 1925 41.40 24.41
SW, The
Rhodopes
1974-1999 very good
peak Rozen 1750 41.53 24.44
SW, The
Rhodopes
1999-2006
Elhovo 136 42.11 26.35 SE 1959-1969 not good
Krumovgrad 235 41.28 25.39 S 1956-1975
relatively
good
peak Botev 2376 42.43 24.55
Central Stara
planina
1966-1994 very good
Chirpan 178 42.18 25.17
Central Bul-
garia
1971 not good
Ljuljakovo 217 42.53 27.05 E 1957-1974 not good
Kaliakra 63 43.21 28.27 NE 1969-2006 very good
Gen. Toshevo 236 43.39 28.01 NE 1956-1986 very good
Dobrich 200 43.35 27.50 NE 1987-2006 very good
68/110
Razgrad 346 43.33 26.30 NE 1978-2006 very good
Gara Samuil 474 43.30 26.44 NE 1955-1979 very good
69/110
Annex 9 Measurements in Hungary
Visual observations of atmospheric icing enable from 1970 till nowadays. The measuring in-
strument system was established by the Hungarian meteorologist Mihaly Csomor. Icing
measurements were carried out by Hungarian Meteorological Service (HMS) and Hungarian
Defence Forces (HDF). The number of stations were 34 (1970-1992), but now we have only
16 (2005). The sites with the atmospheric icing measurements are given in Table A. 9. For
measurement of ice accumulation we use horizontally oriented wires (diameter 31 mm).
Table A. 9 Record of the sites in Hungary, where the parameters of atmospheric icing are
measured
Location Latitude Longitude Altitude Monitoring Output
Bekescsaba 46.41 17.09 156 HMS Manual
Budapest/Lorinc 47.26 19.11 138 HMS Manual
Debrecen 47.29 21.36 108 HMS Manual
Eger 47.54 20.23 220 HMS Manual
Gyor 47.43 17.41 116 HMS Manual
Kecskemet 46.54 19.44 113 HDF Manual
Kekesteto 47.52 20.01 1010 HMS Manual
Miskolc 48.05 20.46 233 HMS Manual
Mosonmagyarovar 47.53 17.16 121 HMS Manual
Nagykanizsa 46.27 16.58 139 HMS Manual
Paks 46.37 18.5 97 HMS Manual
Papa 47.21 17.29 146 HDF Manual
Pecs 46 18.13 202 HMS Manual
Siofok 46.55 18.02 108 HMS Manual
Szeged 46.15 20.06 82 HMS Manual
Szolnok 47.07 20.12 89 HDF Manual
The observers determine the types of accretion (soft rime, hard rime, icy-hard rime, freezing
rain, frozen dew, wet snow) and the duration. In addition the observers measure the thickness
and water content of ice from the last 24 hours, and also the accumulated ice thickness at 06
UTC every day.
70/110
Figure A. 21 Hungarian instrument for manual icing measurement on samples of bare con-
ductors.
71/110
Annex 10 RUSSIA
During the Soviet era, icing observations were included as part of routine meteorological ob-
servations beginning 1931 [43,44]. By 1970's the number of weather stations in the European
part of the USSR that made regular manual icing observations was 220 [45]. In 1984 meas-
urement based estimates of the design ice loads were reported to have been made for 700
weather stations in the USSR [46]. The current situation in Russia is unknown.
The ice collector device used in the Russian measurement network is a stand consisting of two
rigidly clamped horizontal wires with a diameter of 5 mm. The wires are oriented at South to
North and West to East and are at 2 m height from the ground. The measurement is done
manually by weighing on a daily basis.
The main application of the Russian data has been in mapping the design ice loads on power
lines. Because of that, considerable attention was paid in the USSR to determine the relation-
ship between the ice load measured on the ice collector device and that on a real overhead
power line conductor [47]. To that end, special measurement campaigns have been run at se-
lected sites using wires with different diameter, torsional rigidity and height above terrain [48,
49]. Based on these studies, a measurement stand better suitable for the purpose of power line
icing has also been developed and tested at many mountainous sites [50]. This device includes
wires of both 5 mm and 10 mm diameter which are free to rotate around their axes. Another
proposed ice collector includes a vane which adjusts the wires so that they are perpendicular
to the wind direction [51].
72/110
Annex 11 CANADA AND USA
Hydro Québec has run a measurement network for glaze ice caused by freezing rain in the St. Law-
rence Valley since 1974 [52, 53, 54]. Initially there were 35 observation sites and later observations
have been made at 180 sites altogether. The present observation network includes 150 stations and
has a grid dimension of about 50 km.
The ice collector used in the Hydro Québec observation network is the Passive Ice Meter (PIM). It
includes four vertical flat faces, a 25 cm x 25 cm horizontal surface and two groups of horizontally
oriented fixed cylinders with diameters of 1 and 2.5 cm [52]. The observations are made manually
twice a day or, at synoptic weather stations, every three hours during freezing precipitation.
Other Canadian power utilities, e.g. Ontario Hydro, Newfoundland and Labrador Hydro and B.C.
Hydro have also made icing tests, but only at some sites. Mount Washington Laboratory in USA
continuously observes icing and was the first to make rotating multicylinder measurements.
The automatic weather observing system AWOS/ASOS (www.weather.gov/asos) covers almost
1000 sites in North America. The sites are mostly airports and the data are aimed at operational use.
The AWOS/ASOS sites include an automated detection of freezing precipitation by the Rosemount
872/C3 ice detector (Figure A. 22). At these sites this is called “Freezing rain sensor”. The system
reports on freezing precipitation based on an algorithm which takes into account the change in the
resonant frequency of the Rosemount probe, the measured ambient temperature and precipitation.
Figure A. 22 The Rosemount freezing rain sensor mounted at an AWOS site
73/110
Annex 12 Icemeter (Czech Republic)
Description of the instrument
The icing sensor (or “Icemeter”) developed in our Institute of Atmospheric Physics, Prague,
Czech Republic measures the mass of icing accumulated on the surface of its collector. The
first prototype was described in [55], together with a short review of previous methods of ice
measurements that were applied in the Czech Republic. Primarily, it was considered for the
investigation of the most favourable meteorological conditions for in cloud icing growth and
the investigation of accumulation of in-cloud icing and its chemical analysis. The cylinder is
orientated vertically in order to eliminate the detection of wet snow as much as possible. Nev-
ertheless, the “Icemeter” can find its application also in the monitoring system of power lines,
wind power station, and traffic roads; the vertical orientation of the collector can be changed
to a horizontal one, if required.
The mass of accumulated ice is measured by means of a tensometric bridge (strain gauge load
sensor), the output of which is tied to the precise AD converter. The digital signal is preproc-
essed by a micro-controller, which assigns the time and stores the data into the device mem-
ory.
In order to prevent the freezing of the horizontal rod, which couples the cylindrical collector
to the tensometer, which is located together with the electronics in the housing. The passage
through the housing may be heated depending on the passage temperature. A test electro-
mechanical impulse is applied each hour to verify the free force transition to the tensometer,
and thus to check whether the acquired data are reliable or not.
The „Icemeter” can operate autonomously as data logger (it has sufficient memory for ap-
proximately 50 days of operation with 10 min sampling intervals) or can pass the data on re-
quest to the PC. In the case of power supply failure, the Real Time Clock circuit is powered
from the backup rechargeable battery, so the information is not lost. The basic technical
specification of the instrument is summarized in Table A. 10.
Table A. 10 Basic technical specification of Icemeter
Measuring range 0...10 kg
Resolution 1 g
Accuracy ±5 g
Surface area of the sensor 0,05 m
2
Mass of the device 4,25 kg
Interface RS232 (optionally RS485, RS422)
Capacity of the memory ~7500 data points incl. date and time
Voltage Supply 12-15 V
Electricity consumption ~50 … 230 mA depending on heating
Operating temperature range -30°C …50 °C
A description of the sites is given in §6. Results of the measurements with details of the ex-
perience with the Icemeter operation are given below:
74/110
Examples of measurement
The Icemeter has been operated on the Milesovka peak from 2000. The maximum icing load
was detected in the end of 2002, when more than 2.5kg accumulated on the collector of length
48 cm with surface area 0.05 m
2
(see Figure A. 23), thus the accumulated icing was about
1.26 kg/m
2
. At that time several masts were broken due to heavy icing in the Czech Republic.
Figure A. 23 The severe icing recorded at Milesovka peak in the end of 2002
Figure A. 24 presents the example of measurements from different heights above the ground
at the Milesovka observatory. We can see that significantly higher mass of icing accumulates
at the higher position. The shape of the curve – record is similar, but not exactly the same
during the presented period (January/February 2005). Interesting is the comparison with the
measurement in Nova Ves, which is situated ~30km west from Milesovka at about the same
altitude. Here, the icing lasted quite longer. At Milesovka the icing lasted just about two days
from the evening 16.2.2005 to 18.2005, whereas in Nova Ves it lasted more than one week up
to 1.3.2005 (see Figure A. 25). The accumulated mass was however comparable. Note that the
records are not from the same time period. They overlap only partially.
75/110
Figure A. 24 Example of icing measurement at different heights on Milesovka peak.
Figure A. 25 Example of recently (February 2005) installed icing measurement in Nova Ves.
76/110
Problems of operation
Although the Icemeters installed in the Czech Republic measured correctly most of the time,
there were also time periods, when the instruments gave obviously wrong negative values.
These times can usually be associated with the periods when the horizontal rod coupling the
tensometer with the vertical collector became icebound to the instrument housing. Thus there
was no free force transition. We could identify that from the fact that we didn’t see the proper
electromechanical pulse (see section 2) in the technological data. These periods of freezing –
malfunctioning can be seen in Figure A. 25 as data gaps. The instrument gave the wrong
negative values. Note that the location in Nova Ves has more severe cold and windy condi-
tions than Milesovka. This freezing problem has been very rare on Milesovka.
Discussion, conclusions, future plans
Although we have reasonably good data from most of the time of operation in the Czech Re-
public, it seems that we should enhance the available heating power in order that the Icemeter
can operate in cold climate conditions. We consider it should be no problem, provided there is
no limitation to power consumption. Additionally, improvements are needed to the wind
shielding of the passage of the coupling rod through the housing. Currently, the heating power
is only ~ 2 W maximum. Since we haven’t had many problems with freezing on Milesovka,
where we have got most of our experience, we have maintained the heating power as low as
possible to be sure that the heat doesn’t prevent the ice growth.
Except for the enhancement of available heating power, we will try to follow future recom-
mendation for ice sensors that may occur. For example, we will consider the possibility to
build an instrument with a rotating collector. Probably we will mainly continue in focusing on
sensors that measure accumulated icing. We had preliminary talks about the possible place-
ment of the Icemeter in the test site in Switzerland, and in a site in Bulgaria.
Regarding the investigation of conditions favourable for icing growth in the Czech Republic,
we will make the comparison of icing measurement with the measurement of liquid water
content on Milešovka peak. Such comparison should be possible at Milesovka beginning from
the winter 2005/2006. We believe that such measurements could improve our knowledge of
icing. The recent attempts to simulate icing measurements by using temperature, wind and
humidity records haven’t shown sufficiently good results.
77/110
Annex 13 Labko Ice Detector (Finland)
Technical features of LID3210C
The main functional hardware parts of the ice detection system are a sensor probe and a con-
trol unit. The sensor probe consists of a 350 mm long electrical heating element fastened to a
protective cylindrical housing. The ice sensing wire, which is an ultra sonic sensor, is heli-
cally wound around the cylinder as near as. When ice accretion has been detected, the cylin-
der is heated to make the sensor wire ice free again. Heating power, cut off temperature and
ice alarm amplitude can be controlled. So, the ice alarm level and restart delays can be well
set to correspond to changing conditions and purposes.
Operational experiences
Mechanically the Labko 3200-ice detector that has been developed for wind turbine use has
been good. The older model 3500, which was originally developed for meteorological pur-
poses, has been found to indicate ice more accurately compared to the LID 3200 series but it
has also been more fragile to vibrations due to its physical shape [25].
Performances of ice detectors have been monitored with video cameras in order to solve the
noted problems with delayed ice detection. Delayed ice detection is harmful because quick
and reliable indication of the beginning of icing is especially important in wind turbine appli-
cations. This is due to the fact that ice detectors are used to control anti- and de-icing systems
as well as to control operation of turbines in populated environment where ice throw may
pose a safety risk [31].
One such case where an ice detector was not able to detect incipient icing is presented in
Figure A. 26.
0
100
200
300
400
500
600
700
0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00
P
o
w
e
r

[
k
W
]
Measured Power
Power curve
Icing on video
Heating / Ice detector
Figure A. 26 Wind turbine’s performance during icing event 22.2.2002 in northern Finland.
78/110
The performance of the wind turbine deteriorated due to delayed ice detection. Delayed ice
detection also increases the heating power demand of anti-icing systems due to the increased
heat transfer as a consequence of non-laminar (=turbulent) airflow over a rough iced blade
surface.
79/110
Annex 14 Rosemount, BFGoodrich, 0872J (Finland)
Finnish Meteorological Institute (FMI) has performed field tests to verify the performance of
ice detectors on an arctic mountain. Measurements on icing events and duration of icing were
carried out during three winter using Rosemount model 0872J (prototype), at the FMI’s test
site. The test station is located in northern Finland on the top of Luosto fell (500 m asl, N 67
08’, E 26 54’). The Luosto test sites represent an elevated site inland with harsh and frequent
icing climate. Information and documented experimental data on meteorological conditions
during icing events and performance of ice detectors was collected.
The performance of the two ice detectors was monitored with two video cameras Also, data
measured with present weather sensors (visibility/fog), temperature, humidity, dew point and
wind sensors was evaluated.
According to the results, the two automatic instruments used for ice detection were not en-
tirely reliable and differences were found between the performances of ice detectors. The ice
detectors are to some extent insensitive to icing under heavy icing conditions. It was possible
to record more or less accurately the start and ending of icing periods but not the accretion
rate or type of the icing.
The Rosemount sensor has yielded fairly good measurements at Luosto and detected the pres-
ence of icing conditions. In soft icing conditions (snow-like formation composed mainly of
thin ice needles or flakes of ice), ice accretion may exist on the sensor probe during short pe-
riods of time especially in the beginning of the icing event but the sensor does not detect ice.
This sensor is fairly adequate for icing measurements and it operated better than the other
tested instruments. Nevertheless, it cannot guarantee accurate measurements in all icing con-
ditions especially in soft icing conditions.
80/110
Annex 15 EAG 200 (Germany)
Figure A. 27 shows an example of the ice load sensor EAG 200. The measurement pole of the
ice load sensor EAG 200 has diameter of 0.032 m, a length of 0.5 m and consists of PVC. The
measurements are carried out by an electro-mechanical sensor in the lower part of the instru-
ment. It scales the mass of ice that is accumulated at the pole. The measured mass is con-
verted to a frequency signal in a range from 100 Hz to 10 kHz.
The EAG 200 has a measurement range from 0 g to 4000 g (note: type EAG 210 has a capac-
ity of 10kg) and a resolution of 1 g. The measurement error at 0°C is s±0,5% of the measured
value with an accuracy s±10 g. The measurements of the ice load sensor EAG 200 show a
slight dependency on temperature with a temperature coefficient of s±15 g/10 K [56].
Icing pole Ice load sensor EAG 200
Figure A. 27 Icing measurement instruments used by German Weather Service ([37], pictures
were taken at Zinnwald, February 2005)
Before 1991 the icing measurements were performed by the use of manually operated icing
poles at all locations the eastern part of Germany (see chapter 4 for a detailed description).
The icing poles have a diameter of 0.035 m, a length of 1 m and consist of PVC. Figure A. 27
shows an example of icing measurement poles, operated at the station Zinnwald of the Ger-
man Weather Service.
The icing poles are still used for measurements at three locations at present (see 6.2) in order
to compare their results with simultaneous automatic icing measurements by the use of
EAG 200. The measurement results of both instruments were evaluated for the maximum ice
masses that were measured during whole icing cycles (accumulation and loss of ice). The
findings are displayed in Figure A. 28.
81/110
0
1000
2000
3000
4000
5000
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000
EAG 200, Ice mass (g/m)
I
c
i
n
g

P
o
l
e
,

I
c
e

m
a
s
s

(
g
/
m
)
"Pole to heavy,
no measurement possible"
"End of icing cycle before icing
pole measurement"
"Loss of ice before icing pole
measurement"
Figure A. 28 Comparison of maximum ice masses, that were measured in icing cycles by the
use of a manually operated icing pole and automatic icing measurements (EAG 200) during
the period 1996-2004 at the station Zinnwald of the German Weather Service (remarks are
from the measurement protocol of icing pole measurements).
Figure A. 28 illustrates the limitations of manually operated icing pole measurements: For
very large ice masses either no measurements are carried out, because the impossibility of
pole-handling, or the measurements are erroneous. Furthermore the icing cycle may have
been finished before the measurement was carried out.
Regression analysis of both measurement techniques shows acceptable results if the unreliable
data points are excluded (see Figure A. 29). This illustrates that reliable results can be
achieved by the use of an automatic instrument. Nevertheless, additional information about
the icing types or about the icing geometry (e.g. icing diameter, icing vanes) are missed.
y = 0.72x + 129.48
R
2
= 0.70
0
1000
2000
3000
4000
5000
0 1000 2000 3000 4000 5000
EAG 200, Ice mass (g/m)
I
c
i
n
g

P
o
l
e
,

I
c
e

m
a
s
s

(
g
/
m
)
Figure A. 29 As Figure A. 28, but without confirmed unreliable data points
82/110
For the illustration of the EAG 200 system operation and performance a selected result of an
icing event is presented in Figure A. 30. It was measured at the 100 m mast of the Falkenberg
measurement site of DWD’s Meteorological Observatory Lindenberg. Icing measurements at
three heights (see Table “Overview of present locations with icing measurements in Ger-
many”) have been carried out at this location since winter 2004/2005. The example shows,
that the system operates well even for short icing periods and for small amounts of ice accre-
tion. Furthermore, it shows a minor zero offset for all ice load sensors as well as a zero offset
drift, at least for the instrument in 5 m height. These offsets and their drift for the measure-
ment system at the Falkenberg measurement site are mainly due to a temperature dependence
of a signal converter, that had to be used to adapt the ice load sensors frequency data output to
the acquisition system of the tower. It will be reduced by appropriate temperature compensa-
tion in the future.
-20
0
20
40
60
80
100
120
0
2
.0
1
.0
5
1
8
:0
0
0
3
.0
1
.0
5
0
0
:0
0
0
3
.0
1
.0
5
0
6
:0
0
0
3
.0
1
.0
5
1
2
:0
0
Time
I
c
e

M
a
s
s

[
g
/
m
]
-20
0
20
40
60
80
100
120
I
c
e

L
o
a
d

[
g
/
m
]
90 m a.g.l.
50 m a.g.l.
5 m a.g.l.
Figure A. 30 Vertical profile of ice masses for an icing event, measured at the 100 m high
tower of DWD’s Meteorological Observatory Lindenberg
83/110
Annex 16 Gerber (USA)
Gerber Instruments
The UK meteorological office sourced two instruments (Gerbers) capable of measuring LWC
of the air. Two Gerber PVM-100 instruments were provided by the UK Meteorological office
and were installed at the EA Technology severe weather test site at Deadwater Fell test site on
the English/Scottish border in the UK. Data was collected over the winter period December
2003 to April 2004. The aim was to see whether output from these instruments could be re-
lated to the conditions under which overhead line conductors suffer ice loads.
Test Spans
The test span is 200m long and orientated North-South. H-pole supports are used, the south-
ern H-pole having a platform attached for working and sensor attachment. The Gerbers are
mounted on this southern H-pole platform (Figure A. 32 and Figure A. 33).
Figure A. 31 The site on a clear day after winter snowfall.
Monitoring of the Gerbers started on 11
th
December 2003. The site is fitted with load cells
and turnbuckles to adjust the conductor tensions. Figure A. 32 shows the southern dead end
platform and rotating rig.
84/110
Figure A. 32 The Southern dead end platform and rotating rig at Deadwater Fell.
Figure A. 33 Gerber instruments on southern dead end platform.
Performance
The Gerber particulate volume monitors have a 0 - 10VDC output which give a measurement
range 0.002 – 10g/m3. Figure A. 34 to Figure A. 36 show some typical weeks data from the
test site.
85/110
One PVM-100 was mounted and aligned north/south on the southern platform and the other
aligned east/west as shown in Figure A. 33.
Deadwater Fell Weather Data 26th January - 1st February 2004
-10
-5
0
5
10
15
20
25
30
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
Date
T
e
m
p
e
r
a
t
u
r
e

i
n

D
e
g
r
e
e
s

C

/

W
i
n
d

S
p
e
e
d

i
n

m
/
s
0
60
120
180
240
300
360
D
i
r
e
c
t
i
o
n

i
n

D
e
g
r
e
e
s
Temperature
Direction (10% Ave)
Wind Speed (10% Ave)
Figure A. 34 Weather data 26 January – 1 February 2004
Deadwater Fell 26th January - 1st February 2004
0.000
1.000
2.000
3.000
4.000
5.000
6.000
7.000
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
6
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
7
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
8
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
2
9
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
0
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
3
1
-
J
a
n
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
0
1
-
F
e
b
-
0
4
Date
G
e
r
b
e
r

I
n
s
t
r
u
m
e
n
t

g
/
M
3

/

T
e
n
s
i
o
n

i
n

k
N
-10
-5
0
5
10
15
20
T
e
m
p
e
r
a
t
u
r
e

D
e
g
r
e
e
s

C
Gerber
North/South
Gerber
East/West
Hazel
Temperature
Figure A. 35 Load cell and Gerber data for conditions in Figure A.34
86/110
The site span is orientated north-south so winds around 0° or 180° should cause very little ice
load accretion on the conductors according to the meteorological models even if icing condi-
tions are met. Winds around 90° and 270° will be normal to the span and icing on the con-
ductor should occur if the meteorological conditions are right. However, sub-zero tempera-
tures and appropriate winds should not cause icing if there is no liquid water content i.e. the
Gerbers do not indicate the presence of water particles. In Figure A. 34 and Figure A. 35,
there are sub-zero temperatures up to 29 January with occasional winds normal to the span.
However, the Gerbers do not indicate the presence of moisture and there is also no indication
of ice load on the Hazel conductor. Late on 29 January, the temperatures are rising slightly
but still sub-zero and both Gerbers are starting to indicate the presence of moisture. The wind
direction is normal to the span and ice accretes on the Hazel as indicated by the increasing
load in a steady wind speed. Later on 30 January the temperature rises above zero and there is
no icing even though the Gerbers indicate moisture. So this week indicates correct perform-
ance.
Deadwater Fell Weather Data 22nd - 29th February 2004
-10
-5
0
5
10
15
20
25
30
2
3
-
F
e
b
-
0
4
2
3
-
F
e
b
-
0
4
2
3
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
Date
T
e
m
p
e
r
a
t
u
r
e

i
n

D
e
g
r
e
e
s

C

/

W
i
n
d

S
p
e
e
d

i
n

m
/
s
0
60
120
180
240
300
360
D
i
r
e
c
t
i
o
n

i
n

D
e
g
r
e
e
s
Temperature
Direction (10% Ave)
Wind Speed (10% Ave)
Figure A. 36 Weather data from Deadwater Fell 23-29 February
87/110
Deadwater Fell 22nd - 29th February 2004
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
2
3
-
F
e
b
-
0
4
2
3
-
F
e
b
-
0
4
2
3
-
F
e
b
-
0
4
2
3
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
4
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
5
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
6
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
7
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
8
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
2
9
-
F
e
b
-
0
4
Date
G
e
r
b
e
r

I
n
s
t
r
u
m
e
n
t

g
/
M
3

/

T
e
n
s
i
o
n

i
n

k
N
-10
-5
0
5
10
15
20
T
e
m
p
e
r
a
t
u
r
e

D
e
g
r
e
e
s

C
Gerber
North/South
Gerber
East/West
Hazel
Temperature
Figure A. 37 Data output from Gerbers and Hazel load cell for period of Figure A. 34
Figure A. 36 and Figure A. 37 show a period in late February. Sub-zero temperatures were
present for almost the whole period and there were indications of moisture but no ice loads as
the winds were always along the span. Figure A. 38 and Figure A. 39 show a week in March
when the winds were almost always in a direction giving a major component across the span.
These were associated with sub-zero temperatures for most of the period.
Deadwater Fell Weather Data 8th - 14th March 2004
-10
-5
0
5
10
15
20
25
30
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
Date
T
e
m
p
e
r
a
t
u
r
e

i
n

D
e
g
r
e
e
s

C

/

W
i
n
d

S
p
e
e
d

i
n

m
/
s
0
60
120
180
240
300
360
D
i
r
e
c
t
i
o
n

i
n

D
e
g
r
e
e
s
Temperature
Direction (10% Ave)
Wind Speed (10% Ave)
Figure A. 38 Weather data for 8-14 March
88/110
Deadwater Fell 8th - 14th March 2004
0.000
0.500
1.000
1.500
2.000
2.500
3.000
3.500
4.000
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
8
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
0
9
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
0
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
1
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
2
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
3
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
1
4
-
M
a
r
-
0
4
Date
G
e
r
b
e
r

I
n
s
t
r
u
m
e
n
t

g
/
M
3

/

T
e
n
s
i
o
n

i
n

k
N
-10
-5
0
5
10
15
20
T
e
m
p
e
r
a
t
u
r
e

D
e
g
r
e
e
s

C
Gerber
North/South
Gerber
East/West
Hazel
Temperature
Figure A. 39 Gerber and load cell data for period covered in Figure A. 38
On 9 March a minor indication of moisture from the Gerbers combined with a wind normal to
the span and sub-zero temperatures indicated icing conditions which was confirmed by a mi-
nor load increase on the Hazel. Late on 10 March the Gerbers indicated more severe icing
conditions with a wind at an angle of 30° to the span. The Hazel suffered a significant in-
crease in ice load which stopped when the Gerbers indicated no further liquid water content in
the air. The ice load stayed for a while as the temperatures were still below zero. A less sig-
nificant incident occurred on 12/13 March when icing conditions showed up with a further ice
load on the Hazel. Gerber indications of moisture but at above-zero temperatures on 13/14
March did not indicate icing conditions and no ice-load was measured. In the final day of the
period load variations measured were due to an increase in wind speed and violent conductor
movement.
A full 17 weeks data has been supplied to the UK Meteorological office. This includes logged
tension data from the bare Hazel conductor strung over the 200m span with an intermediate
pole at 100m. A camera was also set up to monitor the ice load characteristics near the H-
pole. This would identify rime and wet snow icing events.
89/110
Annex 17 METEO device (Czech Republic)
During long-term development at the stand at Studnice, EGU Brno designed a new automated
monitoring equipment METEO which, in connection with communication means, enables a
continuous measurement of ice mass built up by an automated system including, wind veloc-
ity and direction and temperature without the requirement for the presence of attending per-
sonnel [57,58,59,60].
Device Meteo measures following quantities:
- ice mass up to 20 kg with accuracy 0.01 kg on the measuring rod with length of 0.5 m
and diameter of 30 mm. The value must be multiplied by two in order to obtain the
value corresponding to 1 m
- temperature with accuracy 0.1°C
- range of wind velocity from 0 to 10 m/s with accuracy ± 2 m/s, range from 10 to 40
m/s with accuracy ± 1 m/s
- wind direction – angle from 0 to 355 degrees in steps of 5 degrees.
Measured values are stored in the flash memory, velocity and direction of wind are recorded
every minute and temperature and ice mass every ten minutes. The size of the daily file is 6
192 B and the capacity of the disc is sufficient for 40 or 80 days (depending on size of the
memory used). When the flash memory is filled up the oldest daily file will be overwritten by
the latest one. For eliminating the growing together of the icing on the measuring rod with the
body of the Meteo device heating can be activated (heating is set up in dependence on tem-
perature and on ice mass) [61,62].
Figure A. 40 Device Meteo on a concrete pole of LV power line, location Predin (Photo:
J.Sabata)
90/110
The concept of its solution assumes to build up a network of monitoring equipments by which
the measurement will be performed and the data on air temperature, ice mass and wind veloc-
ity and direction processed.
Means such as radio networks or GPRS may be used for data transmission and for communi-
cation between the centre and individual monitoring equipments.
The operation itself of the automated icing monitoring system is realized in two regimes:
A. Standard regime
The monitoring equipment measures the current quantities, processes the measured values and
downloads them into the local database for each equipment. By the command of the server
from the centre or on demand of the dispatcher the newly measured data will be transmitted
and downloaded into the central data base.
B. Warning regime
Each monitoring equipment may be set up with regard to the expected local conditions in such
a way so that it may, by itself, send a warning message into the centre when the set up values
have been exceeded or when another abnormal event arised.
The warning messages may indicate:
- exceeding the set up ice mass,
- exceeding the set up steepness of ice growing,
- exceeding the value of wind velocity,
- outage of supply and its restoration,
- foreign intervention into the monitoring equipment.
Actual state
At present 14 devices are in the operation, two of them on the territory of regional utility
VCE, 12 on the territory of regional utility JME. The first one was put into operation in 1999,
the others in 2001.
The automated monitoring equipment are installed on the towers of LV, MV a HV overhead
lines.
Data transmission
The data and alarms sent by each monitoring device to the central dispatch office are transmit-
ted by the radio network via retransmission points.
Each hour current data from each METEO device are sent into the dispatch system and are
displayed in the dispatch office. The dispatcher can also make query any time to get immedi-
ate values.
Daily files with measured values, recorded in the flash memory, are sent into SCADA system
at the time of low radio network load (usually at the night). All data are archived in this sys-
tem [63].
91/110
Figure A. 41 Daily record of temperature and ice mass, 18th December 2004 –
location Nyklovice; the end if icing event
92/110
Annex 18 IceMonitor (Sweden)
The IceMonitor was originally developed for the use in a power line surveillance system in-
stalled in Norway 2003, for the Norwegian Power Grid Company, Statnett, and their research
test site in the mountains west of Oslo. The prototype was designed for a maximum ice load
of 100 kg, and later it was modified for lower maximum loads (10, 25 or 50 kg). The ice that
accretes on the vertical, freely rotating sensor (steel pipe with a surface area of 5 dm
2
) is
weighed by a load cell, as the pipe is supported by a rod which is resting on the load cell. To
avoid ice in the area for the bearing of the rod there is electrical heating of the bearing that is
controlled with a thermostat. The output of the load cell is connected to a precision amplifier
and converted into a standardized output current loop – 4 to 20 mA. To be able to perform
testing of the instrument remotely there is a test relay included that will activate an electrical
unbalancing of the load cell – at which the output signal will increase with 8 mA to indicate
that acquired data are reliable.
To log data any kind of data logger with standardized current input (4 – 20 mA) can be used.
At the first site in Norway (Drammen) and at the Swedish sites: in Åre and in the far north
(Ritsem) the IceMonitor is connected to a monitoring station designed by SAAB Technolo-
gies (previously AerotechTelub). These stations also perform measurements of weather pa-
rameters, such as air temperature, relative humidity, wind speed and direction, and in some
cases also precipitation is detected – type and intensity. Measurements are made once every
half hour normally, but in Drammen once every 10 minutes. Data is retrieved to a server in
Östersund via radio/radiolink or in some case via telephone modem. A camera is monitoring
the IceMonitor sensor at the site in Åre. Icing of this camera has been a problem during the
winter 2005/2006 – and the installation and de-icing of the camera will be modified before the
next icing season. The monitoring equipment was originally designed for use in road weather
information systems and it is capable to handle a number of different sensors, including cam-
era – which stores files in jpg format.
At the site in Åre, several icing events occurred during the winter 2005 /2006. A couple of the
icing events have been severe with ice loads of more than 40 kg. (see figures below.) During
the same period, no significant icing occurred at the site in site in Ritsem (Sweden) at the test
site in Drammen (Norway).
93/110
Figure A. 42 Ice build-up developed during the two severe icing events during 2005/2006.
Details are shown in the figures below.
Figure A. 43 First significant icing event at Åre during winter of 2005/2006
94/110
Figure A. 44 Second significant icing event at Åre, winter 2005/2006
95/110
Figure A. 45 Pictures of the site installation in Åre on top of the roof of a ski lift house, with a
camera mounted on the wind sensor support. The site is also prepared for remotely controlled
de-icing of the sensor using
96/110
Annex 19 HoloOptics T20-series Ice Detectors (Sweden)
• The T21 Ice-Duration Indicator™ and Ice-Guard™ indicate the duration of ice on non-
heated structures and can be used for controlling anti-icing or defroster equipment. T21 has
no probe heating so if used as controller the probe is to be heated by the anti-icing or defroster
system. The sensitivity is 50 g/m
2
• The T23 Clear-Ice Indicator™ indicates the presence of atmospheric icing. The probe heat-
ing is internally controlled and the sensitivity is • 50 g/m
2
.
• The T26 Icing-Rate Sensor™ measures the atmospheric icing rate using internally con-
trolled high power probe heating. The sensitivity range is 50 g/m
2
h – 18 kg/m
2
h
Common features of the T20-series ice detectors
The T20-series indicators
- single-direction (±45° upwind) or omni-direction versions
- indicates either all icing types including clear ice or rime ice/wet snow growth only.
The probe is mounted on a cylinder ș=30 mm L=500 mm. A single-direction T20-series indi-
cator is sufficient if there is only one wind direction of interest and it is known (e.g. on a wind
turbine nacelle). In all other cases the T20-series omni version is recommended. The sensors
are equipped with a health-test which provides an early warning if the performance of a sen-
sor is degraded.
All T20-series ice detectors are based on a patented digital optronic ice-indicator that indi-
cates the presence of any type of atmospheric ice including clear ice. It comprises of a head
with an IR emitter and a photo detector and a probe. If not indicated otherwise, reference in
this Annex is made to the indicator sensitive to any type of ice including clear ice.
Figure A. 46 T26 Icing Rate Sensor¹ Single and omni direction versions
97/110
Description of operation – T26
Icing is detected if more than 95 % of the probe is covered with a 50 µm thick layer of clear
ice or 90 µm thick layer of other types of ice. As ice is detected, the internally controlled
probe heating
10
is turned on without time delay. After a short period of time the ice has melted
and the water has fallen off the probe. The indication of icing will stop when the probe heat-
ing is turned off. The time interval between two indications may be as short as ten seconds.
Sourva Indications
495 500 505 510 515 520
hour
I
n
d
i
c
a
t
i
o
n
s
o
f
f











o
n
Figure A. 47 Typical icing indications measured at Suorva, Northern Sweden
The time it takes to melt the ice is dependent on many factors (e.g. icing rate, air and surface
temperature, melting power, wind speed and type of ice). Ice may start to build up on the
probe again as the probe cools down. This cycle is repeated for as long as ice is created on the
probe surface. If sufficient heating is applied, the time it takes for the probe to be covered
with ice after it has been de-iced, is mainly dependent of the icing rate. The time between ic-
ing indications is used to calculate the icing rate using the T26 Icing-Rate Sensor¹.
Sourva Icing Rate
0
50
100
150
200
495 500 505 510 515 520
hour
i
c
i
n
g

r
a
t
e

g
/
m
2
h
Icing Rate
1h average
Figure A. 48 Icing rate measured at Suorva, Northern Sweden
10
External control of the probe heating is available
98/110
Measurement campaigns 2003-2006
The 23 Clear-Ice Indicator and the T26 Icing-Rate Sensor beta versions have been tested at
the following locations:
- The Suorva wind power plant, Sweden. 150 km north of the polar circle. In co-operation
with FOI (The Swedish Defence Research Agency)
- KTH (The Royal Institute of Technology), Stockholm, Sweden
- Bromma Airport, Stockholm, Sweden. In co-operation with the airport authorities
- Greenville Wind power plant, Sc, USA. In cooperation with General Electric
- Sandhaugen test felt, Tromsö, Norway, 400 km North of the polar circle. Cooperation
with Norsk Miljøkraft
- University of Narvik, Norway. 300 km North of the polar circle. Cooperation with the
University of Narvik.
- Keller test field, Norway. Power line protection. In close cooperation with Norsk Hydro
- Åre Ski Resort, ski lift protection. Cooperation with Combitech.
- The T21 Ice-Guard is, in cooperation with KTH (The Royal Institute of Technology),
tested as a defroster controller for freezers
99/110
Annex 20 Instrumar IM101
Instrumar IM101, V.2.4 is an ice detector based on measuring the electrical impedance and
surface temperature [29]. Its long-term use has shown it to be quite durable and it seems to
detect icing in reasonable agreement with humidity indications and video recordings [21]. No
further information is available on this instrument.
100/110
Annex 21 Wind tunnel calibration
This annex refers to Kanagawa Institute of Technology, Japan, as a collaborating member of
COST Action 727. No references are made to wind tunnel studies in other countries in this
report.
Verification of an ice detector
To know the duration of an icing event exactly pointing out its beginning and end is essential
in taking efficient countermeasures for the prevention of icing. Note that the terms of begin-
ning and end mean the onset and completion of an icing event, not of icing conditions. For
that purpose an ice detector is usually employed. The Finnish Meteorological Institute has
been carrying out the field tests to verify the performance of ice detectors in the arctic region.
KAIT conducted the icing wind tunnel test for the German ice detector Infralytic developed
for wind turbines, The principle of ice detection is based on the reflection of infrared light
within an ice deposit emitted from the tip of the optical fibre connected to the sensor. It turned
out that the ice detector can detect ice on the test model. At that moment, the thickness of ice
could not be measured. [64]
Icing on anemometer
In the first place, KAIT has conducted the icing wind tunnel test using a cup anemometer to
know the manner how ice accretes on a cup anemometer and the effect of ice accretion on
measurements. Icing wind tunnel test was done according to the ordinary test scheme as de-
picted in Figure A. 49. The most important parameters of LWC and MVD were obtained by
the rotating multi-cylinder method through some calculation. [65,66,67]
Figure A. 50 shows the growth of ice accretion on a cup anemometer at the different tempera-
tures. Dry-icing and Wet-icing mean the conditions in which rime and glaze grow respec-
tively. As can be seen in the figure, rime affects significantly the performance of the cup
anemometer. As ice grows, the measured wind speed dramatically decreases. The reduction
rate in the figure was defined by the following equation:
where V
clean
and V
*
denote the wind speed measured by a clean anemometer and by one with
iced cups respectively.
The additional wind tunnel test (shown in Figure A. 51) for the acquisition of the aerody-
namic characteristics of an iced cup-shaped body was carried out to infer the aforementioned
reduction of measured wind speed by the iced cup anemometer. As shown in Figure A. 52, ice
grown under the conditions of rime formation exerts the significant influences on the aerody-
namic properties of the cup particularly in the range of the angle of incident higher than 90
degrees. Ordinarily in this range, there is a negative thrust force generated which reduces the
cup revolution. The greater the negative thrust becomes, the weaker the torque around the
shaft tends to be. After all, it can be implied, as easily expected, that the decrease of the
measured wind speed by an iced cup anemometer is attributed to a decline of driving force of
cups with an ice deposit. With using the acquired data, the numerical approach was taken to
evaluate the effect of icing on a cup anemometer. The behaviour of cups was calculated by
solving the equation of motion with respect to a cup part of a cup anemometer considering the
force acting on each cup and stem, and at bearings supporting a rotating part. The results are
shown in Figure A. 53
(1)
clean
clean
V
V V
RR
*
÷
=
101/110
Snow accumulation of anemometers
How snow accumulates on an anemometer and how a snow deposit affects measurements
have been examined by the wind tunnel test where snow flakes of dendritic shape were fed
into the wind tunnel test section by a snow fall device as shown in Figure A. 54. The conse-
quence revealed that wet snow accretes dramatically only on a cup anemometer and the sur-
face of the snow deposit is asymmetric with respect to the axis through the centre of a cup
(Figure A. 55). The most intriguing feature is that the snow deposit grows no further beyond
some certain height, which was approximately 50% in the centre in those tests. These findings
led us to undertake additional tests using cup anemometers with cups covered by a thin mem-
brane in the cup open face, and with cups with snow deposit having the leveled surface in
order to quantitatively evaluate the effect of change of the shape of snow accumulation inside
a cup in terms of measured wind speed as well. All the results are shown in Figure A. 56. This
indicates that snow accumulation inside cups reduces the measured wind speed by a maxi-
mum of 13% and the reduction of the measured wind speed by the cup anemometer is merely
20%, even if the cup open face is covered completely. [68,69]
Based on the test results mentioned earlier, the wind tunnel test (Figure A. 57) was conducted
in order to investigate the aerodynamic effect of the cup covering of a cup anemometer. The
results show that the skilfully fabricated covering exerts negative influences on the aerody-
namic characteristics of a cup-shaped body only when the air stream flows into the covered
side of a cup (Figure A. 58). Figure A. 59 shows the linear relation between the measured
wind speed and the cup revolutions of the covered cup anemometer slightly changing the co-
efficients of the linear transfer function provided by the manufacturer.
Field measurements using a heated covered-cup anemometer and a heated ultra-sonic ane-
mometer were implemented for verification of the effectiveness of cup-covering in icing-
endangered mountains in Japan. Wind speed, together with temperature, humidity, solar radia-
tion and precipitation was successively measured for 186 days starting from November. Dur-
ing the period of campaign, precipitation over a period of 754 hours was identified. The cov-
ered cup anemometer stopped its measurement for only 1 hour, whilst the ultra-sonic ane-
mometer transmitted meaningless signals for 56 hours, probably due to ice or snow accretion
on the emitter/receiver(s). The correlation between the covered and ultra-sonic anemometer
was depicted in Figure A. 60 where the wind speed measured by the covered cup anemometer
was calibrated by using the transfer function that can be seen in the figure. It turned out that
the correlation is so satisfactory that the cup covering of a cup anemometer would be effective
when used in icing prone areas.
Coatings to reduce the adhesive strength of ice
In order to seek the possibilities of coating as de- and/or anti-icing measures, the performance
of commercially available coatings was examined by load test. An ice deposit accreted on
specimens was created in two different manners: dynamic and static ice formation. The for-
mer was the formation of ice in an icing wind tunnel test letting super-cooled minute water
droplets collide onto an airfoil model with a specific velocity where the complicated thermo-
dynamic process always takes place. The latter was actualized in the low temperature cham-
ber where water was poured into a steel ring placed on a specimen and then frozen at rest. In
those tests, the force was applied parallel to the adhesion surface until ice was removed. The
typical results are shown in Figure A. 63 and Figure A. 63. Although there may be a problem
that has to be solved in terms of durability, the test results shows that one type of coating can
significantly reduce the ice adhesion regardless of the manner of formation of ice [70,71].
102/110
Figure A. 49 Schematic view of icing wind tunnel
Figure A. 50 Effect of ice accretion on measurements by a cup
Figure A. 51 Wind tunnel test for acquisition of aerodynamic characteristics
Airstream
Sprayer
Test model
0 50 100 150
0
10
20
30
40
50
60
Elapsed Time (min.)
R
e
d
u
c
t
i
o
n

R
a
t
e

(
%
)
Dry-Icing
Wet-Icing
t30
(b). Imitated model
(c). Actually iced cup
(a). Test model in the test sec-
i
103/110
Figure A. 52 Aerodynamic characteristics of an iced cup-shaped body
0 1 2
0
0.2
0.4
0.6
0.8
1
Elapsed time (s)
R
P
M
,

n
o
r
m
a
l
i
z
e
d

b
y

t
h
e

m
a
x
i
m
u
m

r
p
m

o
f

t
h
e

c
l
e
a
n

c
u
p

v
a
l
u
e
Clean cup
20min.
60min.
120min.
180min.
63.2%
0 1 2 3
0
0.2
0.4
0.6
0.8
1
Time (s)
R
P
M
,

n
/
n
m
a
x
63.2% value
Clean cups
Ice thickness from 0.2mm to 1.0mm
at 0.2mm intervals
Figure A. 53 Calculations for cup behaviour
Figure A. 54 Schematic view of the wind tunnel test for snow accumulation on anemometer
Figure A. 55 Comparison of snow deposit between wind tunnel test and field test
Airstream
Sprayer
Snow fall device
Test model
Dendritic snow flakes
0 30 60 90 120 150 180
-2
-1.5
-1
-0.5
0
0.5
1
1.5
Angle of incidence (deg.)
N
o
r
m
a
l

f
o
r
c
e

c
o
e
f
f
i
c
i
e
n
t
,

C
n
D-10
D-15
D-25
Re=3.1e5
D-35
D-45
0 30 60 90 120 150 180
-1
-0.5
0
0.5
1
1.5
2
Angle of incidence (deg.)
N
o
r
m
a
l

f
o
r
c
e

c
o
e
f
f
i
c
i
e
n
t
,

C
n Clean
W-10
W-15
Re=1.1e5
(a). Wet-Icing (b). Dry-icing
104/110
Figure A. 56 Effect of snow accumulation and cop-covering on measurements
Figure A. 57 Test model in wind tunnel
Figure A. 58 Aerodynamic characteristics of a covered cup
Figure A. 59 Transfer function of a covered cup anemometer
Figure A. 60 Field measurements by a covered cup anemometer and an ultra-sonic anemome-
ter
0 25 50 75 100
0
10
20
Dimensionless depth of snow (%)
R
e
d
u
c
t
i
o
n

R
a
t
e

(
%
)
Naturally accreted in WT
V=5m/s
Artificially accreted
V=5.0m/s
V=7.5m/s
V=10.0m/s
Fully covered with felt
Clean Cup
Cup-A
Cup-B
0 30 60 90 120 150 180
-0.5
0
0.5
1
1.5
2
0 30 60 90 120 150 180
-0.5
0
0.5
1
1.5
2
0 30 60 90 120 150 180
-0.5
0
0.5
1
1.5
2
.) (deg , o angle Incident
Cup-A Cup-B
3879 . 0 02732 . 0 + × = RPM V
0 200 400 600 800 1000
0
5
10
15
20
25
30
Eq. from VAISALA
Wind tunnel tests & curve fitting
for the covered cup anemometer
Revolutions (RPM)
W
i
n
d

s
p
e
e
d

(
m
/
s
)
105/110
Figure A. 61 Ice detector tested
Figure A. 62 Signals from the ice detector in ice detection
Figure A. 63 Comparison of the adhesive strength of ice formed in the wind tunnel test and in
a chamber
0
0.5
1
1.5
0 100 200 300 400 500 600 700
No. of acquired data
I
n
t
e
n
s
i
t
y
Ch1 W/E 2mm
Onset of icing event End of icing event
0
¯0
100
1¯0
?00
?¯0
l | k S " k |
Coatings
A
d
h
e
s
i
b
e

S
t
r
e
n
g
t
h

(
k
P
a
)
0
¯0
100
1¯0
?00
?¯0
l | k S " k |
Coatings
A
d
h
e
s
i
b
e

S
t
r
e
n
g
t
h

(
k
P
a
) Static test
WT Test
0
100
200
300
400
500
1 ? 3 ´ ¯ b ¯ 8 9 10
No. of iteration
A
d
h
e
s
i
v
e

s
t
r
e
n
g
t
h

(
k
P
a
)
|··huu·
S" k·¦·!!.
106/110
Annex 22 References
1 COST-727 Action: Memorandum of Understanding
2 WMO Wind Instrument Intercomparison, Mont Aigoual, France 1992-1993: Final re-
port by P. Gregoire and G. Oualid. Instruments and observing methods report; no. 62
Technical document (WMO); no. 859
3 Tammelin, B., Joss, J. and Haapalainen, J., 1998: The Final report on the EUMETNET
project “Specification of Severe Weather Sensors”. Finnish Meteorological Institute,
Helsinki.
4 B. Tammelin and al Improvement of Severe Weather Measurements and Sensors –
EUMETNET SWS II Project. FMI reports 2004:3
5 Tammelin, B., Cavaliere, M., Holttinen, H., Morgan, C., Seifert, H. and Säntti, K.,
2001: Wind Energy Production in Cold Climate, Meteorological Publications No. 41,
Finnish Meteorolgical Institute, Helsinki 2000.
6 Tammelin, B., Säntti, K., Dobesch, H., Durstewich, M., Ganander, H., Kury, G., La-
akso, T., Peltola, E. and Ronsten, G., 2005: Wind turbines in icing environment: im-
provement of tools for siting, certification and operation – NEW ICETOOLS. Reports
2005:6, Finnish meteorological Institute.127 p
7. State-of-the-art of wind energy in cold climate, T. Laakso, H. Holttinen, G. Ronsten,
L.Tallhaug, R.Horbaty, I. Baring-Gould, A. Lacroix, E. Peltola, B. Tammelin, 2005: IEA
Wind Annex XIX, 53 p. http:\arcticwind.vtt.fi, date of access 20.3.2006
8. Wind Energy Projects in cold Climates, T. Laakso, G. Ronsten, L.Tallhaug, R.Horbaty,
I. Baring-Gould, A. Lacroix, E. Peltola, 2005: IEA Wind Annex XIX, 37p, 2005,
http:\arcticwind.vtt.fi, date of access 20.3.2006.
9 ISO 12494: Atmospheric icing of structures. ISO/TC 98/SC 3, 2000-07-20
10 11th Workshop on Atmospheric Icing of Structures (IWAIS 2005). Montreal, Canada,
12 - 16 June 2005 (2005)
11 Popov, N.I., Holodov, V.V., 1978: On the applicability of icing measurements. Trudy
GGO 408, 44-48. (in Russian)
12 IEC 60826: Design criteria of overhead transmission lines, 2003.
13 IEC 61744: “Overhead lines – Meteorological data for assessing climatic loads”, First
Edition, August 1997.
14 IEC 61400-1: "Wind turbine generator systems - Part1: Safety requirements", Second
Edition, 1999, ref. number IEC 61400-1:1999(E)
15 Tammelin, B., Heimo, A., Leroy, M., Rast, J., Säntti, K., Bellevaux, C., Dal Cin, B.,
Musa, M. and Peltomaa, A., 2004: Improvement of severe weather measurements and
107/110
senors – EUMETNET SWS project, Final Report. Reports 2004:3, Finnish Meteoro-
logical Institute, Helsinki. 101 p.
16 Germanischer Lloyd Rules and Regulations IV - Non Marine Technology Part 1 - Wind
Energy, (GLRP3.0-1998).
17 World Meteorological Organization, 1996: Guide to Meteorological Instruments and
Methods of Observations. WMO-No. 8. Secretariat of the World Meteorological Or-
ganization, Geneva, Switzerland.
18 Makkonen, L., 1992: Analysis of rotating multicylinder data in measuring cloud droplet
size and liquid water content. Journal of Atmospheric and Oceanic Technology, 9 (3):
258 - 263.
19 Lehtonen, P., Ahti, K. & Makkonen, L., 1986: The growth and disappearance of ice
loads of a tall mast. Proceedings, Third International Workshop on Atmospheric Icing
of Structures, 6 - 8 May 1986, Vancouver, 363 - 368.
20 Lehtonen, P., Laiho, J., Makkonen, L., Riisiö, P. & Miettinen, H., 1988: Studies on at-
mospheric icing in Finland. Proceedings, 4th International Conference on Atmospheric
Icing of Structures, 5 - 7 September 1988, Paris, 338 - 343.
21 Makkonen, L., 2003: Evaluations of ice-free anemometers and ice detectors. BOREAS VI
Meeting, 9-11 April 2003, Pyhätunturi, Finland, 12 p. ISBN 951-697-584-4.
22 Harstveit, K., 2003: Calculating duration of ice weights at reference objects along the
Norwegian coast originated with in-cloud rime. Proc. Boreas VI, Wind energy produc-
tion in cold climates, Pyhätunturi, Finland, 2003.
23 Harstveit, K, L.Tallhaug and A. Fidje, 2005: Ice accumulation observed by use of web
camera and modelled from meteorological parameters. Proc. Boreas VII, Wind energy
production in cold climates, Saariselkä, Finland, 2005.
24 Mazin, I. P., Korolev, A. V., Heymsfield, A., Isaac, G. A., Cober, S. G., 2001: Thermo-
dynamics of Icing Cylinder for Measurements of Liquid Water Content in Supercooled
Clouds, Journal of Atmospheric and Oceanic Technology, Vol 18, 543-558.
25 Laakso, T., 2001: The Follow-up of Olostunturi Arctic Wind Farm, VTT Energy reports
24/2001, in Finnish.
26 Peltola, E. Laakso, T. Tammelin, B., Säntti, K., Ronsten, G., 2002:Observation Tools of
Icing Events, Public report of EU-sponsored project New Icetools, NNE5/2001/259,
2002.
27 Makkonen, L., 1992: Can ice load measurements be standardized? International Semi-
nar on iceload measurements, Kristiansand, Norway, 18-20 May 1992.
28 Lehtonen, P., 2000: Experience on the ISO Reference Collector. 8th IWAIS, Reykjavik,
Iceland, 357.
108/110
29 Rothig, H., 1973: A device for continuous measuring and recording of ice accretion.
Abhandlungen Meteorologische Dienst DDR 107, 26-30. (in German)
30 Internet page of Wavin-Labko Oy, http://www.labko.fi, Date of reference 20.3.2006
31 Cober et Al, Assessing the Rosemount Icing Detector with In Situ Measurements, Jour-
nal of Atmospheric and Oceanic Technology, Vol 18, 515-528.
32 Inkpen, S., Nolan, C. Oleskiw, M.M., 1990: Development of a sensor for the detection
of atmospheric ice. 5th IWAIS, Tokyo, Japan, Paper A1-5.
33 Makkonen, L. & Laakso, T., 2005: Humidity measurements in cold and humid envi-
ronments. Boundary-Layer Meteorology, 116, 131-147.
34 Makkonen, L., 1992: Structure-independent measurement of icing. Proceedings, Inter-
national Expert's Meeting on Wind Power in Icing Conditions (BOREAS), 10 - 13 Feb-
ruary 1992, Hetta: 43 - 49.
35 Makkonen, L., Lehtonen, P. & Helle, L., 2001: Anemometry in icing conditions. Jour-
nal of Atmospheric and Oceanic Technology, 18(9): 1457-1469.
36 Teucher, F., 1992: Iceload measurements in East Germany. International Seminar on
Iceload Measurements, Kristiansand, Norway, 18-20 May 1992, 7 p.
37 Wichura, B., 2005: A survey of icing measurements in Germany. In: Tammelin et al.
(Editor). BOREAS VI. Finnish Meteorological Institute. Saariselkä, CD-ROM, ISBN
951-697-643-3.
38 Ostrožlik M., Atmospheric icing at Lomnický štít, Geophysics and Geodesy Vol.35/3,
2005.
39 K. Savadjiev, M. Farzaneh, S. Fikke and A. Goel, 2001: Extreme Value Analysis of Ice
Accretion Data from Norwegian Measurement Rack Network, Proceedings ISOPE
2001, Stavanger, Norway.
40 Popolansky, F., 2002: Historical survey of ice investigation on the territory of the Czech
Republic. 10th IWAIS, Brno 2002, 6 pp.
41 Eliasson, A J, Thorsteins, E,, 1993Data analysis of icing measurements in Iceland, Proc.
6th IWAIS, Budapest, 1993.
42 Shackleton L et al, 1993: Comparison of growth rates on groups of stranded conductors’
EATL Contract, 1993
43 Nikiforov, E.P., 1982: Icing related problems, effect on line design and ice load map-
ping. First IWAIS, U.S. Army CRREL Special Report 83-17, 239-245.
44 Dranevic, E.P., 1971: Glaze and Rime. Leningrad Gidrometeorologitsheskoe, 227 p.
45 Guljajev, J.N., 1979: On the duration of icing. Trudy GGO 425, 80-87.
109/110
46 Zaharov, S.G., 1984: Distribution of ice loads in the USSR. Trudy GGO 485, 87-93.
47 Anon, 1966: Instructions on Determining Glaze-ice Loads. Standard SN-318-65,
Moskow, 19 pp.
48 Holodov, V.U., Popov, N.I., 1976: The effect of structural parameters of electric power
lines on ice loads on cables. Trudy GGO 379, 30-36. (in Russian)
49 Zavarina, M.V., Arnautov, P.K., 1973: Correction factors for ice load measurements.
Trudy GGO 303, 154-159. (in Russian)
50 Popov, N.I., Holodov, V.V., 1978: On the applicability of icing measurements. Trudy
GGO 408, 44-48. (in Russian)
51 Podrezov, O.A., Popov, N.I., Naumov, A.O., 1987: Intense icing in mountain regions
and ways to improve ice deposit measurements. Meteorologiya i Gidrologiya 4, 59-65.
52 Felin, B., 1976: The observations of rime and glaze deposits in Quebec. Canadian Elec-
trical Association Spring Meeting, 22-24 March 1976, Toronto, 53 p
53 Laflamme, J., 1992: Space and time estimates of ice loads from two interactive net-
works of ice accretion measurement sites. International Seminar on Ice Load Measure-
ments, Kristiansand, Norway, 18-20 May 1992, 13 p.
54 Laflamme, J.N., Periard, G. 1998: The climate of freezing rain over the province of
Quebec in Canada: A preliminary analysis. Atmospheric Research 46, 99-111.
55 Fišák. J., Chum J., Vojta J., TesaĜ M.: Instrument for measurement of the amount of the
solid precipitation deposit - ice meter, J. Hydrol. Hydrolmech 49, 3
56 Academy of Science of GDR, Center for Construction of Scientific Equipment, 1986.
Eisablagerungsgeber EAG 200. Beschreibung und Bedienungsanleitung (in German).
Berlin, 15 pp.
57 Lehky, P., Zalesak, Z., Kvacova, H., 1998: Automated system for icing monitoring –
supply area of VýE (in Czech). EGÚ Brno, a.s. – 020, 1998
58 Lehky, P., Zalesak, Z., Kvacova, H., 1999: Automated system for icing monitoring –
supply area of JME (in Czech). EGÚ Brno, a.s. – 020, 1999
59 Lehky, P., Zalesak, Z., Kvacova, H., 2000: Automated system for icing monitoring –
supply area of SME (in Czech). EGÚ Brno, a.s. – 020, 2000
60 Lehky, P., Maderová, P., Kvacova, H., 2001: Automated system for icing monitoring.
Project documentation for supply area of JME (in Czech). EGÚ Brno, a.s. – 020, 2001
61 Lehky, P., Sabata J., 2000: Automated icing monitoring system, IWAIS 2000, Brno.
Czech Republic
110/110
62 Lehky, P., Sabata J.: Automated icing monitoring system, BOREAS VII 2005,
Saariselkä, Finland
63 Vaculik P.,Rampl M.: Automated icing monitoring system on the territory supplied by
JME (in Czech), CK CIRED 2005, Tabor, Czech Republic
64 Kimura, Shigeo, Takeshi Sato, Holger Freitag & Bengt Tammelin, 2005: Icing wind
tunnel test for a new ice detector, BOREAS VII, Finnish Meteorological Institute
65 Kimura, S., B. Tammelin, A. Peltomaa & K. Tsuboi: Icing effect on a cup anemometer,
Meteorological publication, Report number 44, Finnish Meteorological Institute,
29pages
66 Kimura Shigeo, Bengt Tammelin & Kazuhiro Tsuboi: How much does ice affect wind
speed measurement? , BOREAS V, Finnish Meteorological Institute, 2000, 11 pages
67 Kimura, S., K. Abe, B. Tammelin, K. Tsuboi & K. Suzuki: Aerodynamic characteristics
of an iced cup-shaped body, Journal of Cold Regions Science and Technology, Vol.33,
Issue1, 2001, pp.45-58
68 Kimura, S., T. Sato, K. Kosugi, et al., 2004: Why do the cups of a cup anemometer have
an open face? Proc. of The 7th International Symposium on Cold Region Development,
CD-ROM, No.62, 10-pages
69 Kimura, Shigeo, Takeshi Sato, Kenji Kosugi, & Hiroshi Morikawa: Wind Tunnel Tests
for Wet-Snow Accretion on Anemometers, (Submitted to Journal of Cold Regions Sci-
ence and Technology)
70 Kimura Shigeo, Takeshi Sato & Kenji Kosugi, 2003: The effect of anti-icing paint on
the adhesion force of ice accreted on a wind turbine blade, Proc. of The International
Meeting on Wind Power in Cold Climates BOREAS-VI, Finnish Meteorological Insti-
tute
71 Kimura, S., K. Furumi, T. Sato & K. Tsuboi, 2004: Evaluation of anti-icing coatings on
the surface of wind turbine blades for the prevention of ice accretion, Proc. of The 7th
International Symposium on Cold Region Development, CD-ROM, No.63, 10-pages








Veröffentlichungen der MeteoSchweiz

74 Schmutz C, Müller P, Barodte B: 2006, Potenzialabklärung für Public Private Partnership (PPP)
bei MeteoSchweiz und armasuisse Immobilien, 82pp, CHF 76.-
73 Scherrer SC: 2006, Interannual climate variability in the European and Alpine region, 132pp,
CHF 86.-
72 Mathis H: 2005, Impact of Realistic Greenhouse Gas Forcing on Seasonal Forecast
Performance, 80pp, CHF 75.
71 Leuenberger D: 2005, High-Resolution Radar Rainfall Assimilation: Exploratory Studies with
Latent Heat Nudging, 103pp, CHF 81.-
70 Müller G und Viatte P: 2005, The Swiss Contribution to the Global Atmosphere Watch
Programme – Achievements of the First Decade and Future Prospects, 112pp, CHF 83.-
69 Müller WA: 2004, Analysis and Prediction of the European Winter Climate, 115pp, CHF 34.
68 Bader S: 2004, Das Schweizer Klima im Trend: Temperatur- und Niederschlagsentwicklung
seit 1864, 48pp, CHF 18.-
67 Begert M, Seiz G, Schlegel T, Musa M, Baudraz G und Moesch M: 2003, Homogenisierung
von Klimamessreihen der Schweiz und Bestimmung der Normwerte 1961-1990, Schluss-
bericht des Projektes NORM90, 170pp, CHF 40.-
66 Schär Christoph, Binder Peter, Richner Hans (Eds.): 2003, International Conference on Alpine
Meteorology and MAP Meeting 2003, Extended Abstracts volumes A and B, 580pp, CHF 100.
65 Stübi R: 2002, SONDEX / OZEX campaigns of dual ozone sondes flights: Report on the data
analysis, 78pp, CHF 27.-
64 Bolliger M: 2002, On the characteristics of heavy precipitation systems observed by Meteosat-
6 during the MAP-SOP, 116pp, CHF 36.-
63 Favaro G, Jeannet P, Stübi R: 2002, Re-evaluation and trend analysis of the Payerne ozone
sounding, 99pp, CHF 33.-
62 Bettems JM: 2001, EUCOS impact study using the limited-area non-hydrostatic NWP model in
operational use at MeteoSwiss, 17pp, CHF 12.-
61 Richner H, et al.: 1999, Grundlagen aerologischer Messungen speziell mittels der Schweizer
Sonde SRS 400, 140pp, CHF 42.-
60 Gisler O: 1999, Zu r Methodik einer Beschreibung der Entwicklung des linearen Trends der
Lufttemperatur über der Schweiz im Zeitabschnitt von 1864 bis 1990, 125pp, CHF 36.-
59 Bettems J-M: 1999, The impact of hypothetical wind profiler networks on numerical weather
prediction in the Alpine region, 65pp, CHF 25.-
58 Baudenbacher, M: 1997, Homogenisierung langer Klimareihen, dargelegt am Beispiel der
Lufttemperatur, 181pp, CHF 50.-
57 Bosshard, W: 1996, Homogenisierung klimatologischer Zeitreihen, dargelegt am Beispiel der
relativen Sonnenscheindauer, 136pp, CHF 38.-
56 Schraff, C: 1996, Data Assimilation and Mesoscale Weather Prediction: A Study with a
Forecast Model for the Alpine Region, 138pp, CHF 38.-
55 Wolfensberger, H: 1994, Chronik der Totalisatoren, Handbuch zu den Niederschlags-
Totalisatoren, 390pp, CHF 78.-








Kürzlich erschienen
Frühere Veröffentlichungen und Arbeitsberichte finden sich unter
www.meteoschweiz.ch » Forschung » Publikationen

Arbeitsberichte der MeteoSchweiz

214 Schmucki D., Weigel A., 2006, Saisonale Vorhersage in Tradition und Moderne: Vergleich der
"Sommerprognose" des Zürcher Bööggs mit einem dynamischen Klimamodell, 214, 46pp, CHF
68.-
213 Frei C: 2006, Eine Länder übergreifende Niederschlagsanalyse zum August Hochwasser 2005.
Ergänzung zu Arbeitsbericht 211, 10pp, CHF 59.-
212 Z’graggen, L: 2006, Die Maximaltemperaturen im Hitzesommer 2003 und Vergleich zu früheren
Extremtemperaturen, 74pp, CHF 75.-
211 MeteoSchweiz: 2006, Starkniederschlagsereignis August 2005, 63pp., CHF 72.-
210 Buss S, Jäger E and Schmutz C: 2005: Evaluation of turbulence forecasts with the aLMo,
58pp, CHF 70.–
209 Schmutz C, Schmuki D, Duding O, Rohling S: 2004, Aeronautical Climatological Information
Sion LSGS, 77pp, CHF 25.–
208 Schmuki D, Schmutz C, Rohling S: 2004, Aeronautical Climatological Information Grenchen
LSZG, 73pp, CHF 24.–
207 Moesch M, Zelenka A: 2004, Globalstrahlungsmessungen 1981-2000 im ANETZ, 83pp, CHF
26.–
206 Schmutz C, Schmuki D, Rohling S: 2004, Aeronautical Climatological Information St.Gallen
LSZR, 78pp, CHF 25.–
205 Schmutz C, Schmuki D, Ambrosetti P, Gaia M, Rohling S: 2004, Aeronautical Climatological
Information Lugano LSZA, 81pp, CHF 26.–
204 Schmuki D, Schmutz C, Rohling S: 2004, Aeronautical Climatological Information Bern LSZB,
80pp, CHF 25.–
203 Duding O, Schmuki D, Schmutz C, Rohling S: 2004, Aeronautical Climatological Information
Geneva LSGG, 104pp, CHF 31.–
202 Bader S: 2004, Tropische Wirbelstürme – Hurricanes –Typhoons – Cyclones, 40pp, CHF 16.-
201 Schmutz C, Schmuki D, Rohling S: 2004, Aeronautical Climatological Information Zurich
LSZH, 110pp, CHF 34.-
200 Bader S: 2004, Die extreme Sommerhitze im aussergewöhnlichen Witterungsjahr 2003, 25pp,
CHF 14.-
199 Frei T, Dössegger R, Galli G, Ruffieux D: 2002, Konzept Messsysteme 2010 von
MeteoSchweiz, 100pp, CHF 32.-
198 Kaufmann P: 2002, Swiss Model Simulations for Extreme Rainfall Events on the South Side of
the Alps, 40pp, CHF 20.-
197 WRC Davos (Ed): 2001, IPC - IX, 25.9. - 13.10.2000, Davos, Switzerland, 100pp, CHF 32.-
196 Hächler P et al.: 1999, Der Föhnfall vom April 1993, 139pp, CHF 40.-
195 Urfer Ch, Vogt R, 1999, Die Niederschlagsverhältnisse in Basel 1964-1998, 43pp, CHF 40.-
194 Courvoisier HW: 1998, Statistik der 24-stündigen Starkniederschläge in der Schweiz 1901-
1996, 20pp, CHF 11.-
193 Defila C, Vonderach G: 1998, Todesfälle und Wetterlagen in Schaffhausen, 72pp, CHF 25.-









































COST is supported by the EU RTD Framework Programme

ESF provides the COST Office through an EC contract

Veröffentlichung MeteoSchweiz Nr. 75

ISSN: 1422-1381

COST 727: Atmospheric Icing on Structures Measurements and data collection on icing: State of the Art
S. Fikke (Chair, Norway) G. Ronsten (Sweden) A. Heimo (Switzerland) S. Kunz (Switzerland) M. Ostrozlik (Slovakia) P.-E. Persson (Sweden) J. Sabata (Czech Rep.) B. Wareing (United Kingdom) B. Wichura (Germany) J. Chum (Czech Rep.) T. Laakso (Finland) K. Säntti (Finland) L. Makkonen (Finland)

Bitte zitieren Sie diese Veröffentlichung folgendermassen:

Please quote this publication in the following way:
COST-727, Atmospheric Icing on Structures: 2006, Measurements and data collection on icing: State of the Art Publication of MeteoSwiss, 75, 110 pp.

Herausgeber Bundesamt für Meteorologie und Klimatologie, MeteoSchweiz, © 2007
MeteoSchweiz Krähbühlstrasse 58 CH-8044 Zürich T +41 44 256 91 11 www.meteoschweiz.ch Weitere Standorte CH-8058 Zürich-Flughafen CH-6605 Locarno Monti CH-1211 Genève 2 CH-1530 Payerne

2/110 .

Therefore all contributions concerning individual countries are provided according to available time and engagements of the participants. Germany. Bulgaria. COST Action 727 acknowledges Dr Wiel M. data resources and instrument testing. Wauben. Sweden. WG1 “Icing modelling”. Hungary.Foreword COST Action 727 “Measuring and forecasting atmospheric icing on structures” was established in April 2004 and comprises 12 signatory countries: Austria. It is the intention of Phase 2 to structure and update information from existing test sites and open data sources in a more systematic way than was possible in this report. however. three working groups were established. It is important to notice that COST does not support project activities. Hence the structure and details of each contribution will vary. Phase 2 will also include instrument comparisons from test sites. The main scope of this phase was to create an inventory of earlier and current activities on icing measurements. Following the “Memorandum of Understanding” (MoU). Finland. Svein M. and also elaborate recommendations for WMO observations and permanent data bases for icing in Europe. The present report covers the work of WG2 during Phase 1 of the Action. WG2 “Measurements and data collection on icing” and WG3 “Mapping and forecasting of atmospheric icing”. A lot of references are given. Switzerland and the United Kingdom. Fikke Chairman of Working Group 2 3/110 . F. The emphasis is on activities within the signatory countries. the Royal Netherlands Meteorological Institute (KNMI) for reviewing this report and MeteoSwiss for their generous offer to print the report as part of their series of internal reports. and the reader will find links to institutions where further information can be retrieved. the Czech Republic. and the reader will not necessarily find the same information for all countries. Norway. Slovakia. however some additional information from other countries like Russia and Canada is included as well. Spain.

..............................................11 Canada................ 10 Definitions and meteorological conditions ....... 34 7.......1 International Standardization Organization (ISO) .............3 World Meteorological Organization (WMO) ..................................................................................................................3 Verifications of data ........................... 28 6............1 Finland..............9 Hungary ................................................ 9 2.................................3 Slovak Republic ...................................................5 Site Icing Index ............................................. 33 7........................3 Past and present activities ....................................................................... 21 5......................................................................................1 Generic definition................................................................................................. 24 Examples of existing icing data ................................................... 36 Availability... 9 2....................................... 31 6........................................................... 13 3......................................................................... verification and requirements of ice detectors ......................................................................................................................................... 19 Meteorological measurements under icing conditions..................................... 17 4............................................2 Interface with WG1 and WG3 ....................... 41 9...............................................5 Czech Republic . 10 2...............2 Labko ice detectors.................................................... 39 8..................................... 17 4........................................................................................ 29 6..................................................................................................................................... 7 Introduction .....4 Icing and wind turbines ............................................................2 Data requirements for icing models ...................................................................................................... model 0872J ..............3 Definitions........... 23 5......................................................................... 41 9.................8 Bulgaria ............................................................................................................................. 41 9...............3 Guidance for selecting ice detectors........ 28 6............1 Available ice detectors ....................................................................................................................... 40 8.. 33 7..............................................................2 Siting of icing sensors .4 Norway ............................................................... 28 6.........................................................................................................................................................................................4 Site effects ..... 20 5...................1 Memorandum of Understanding ............................. 41 4/110 3 4 5 6 7 8 9 ........................13 EUMETNET/SWS II project .......... 30 6................. 28 6........................... 29 6..................................................... 18 4................. 20 5.......... 3 Management summary .................................................. 40 Experiences with automatic instruments for ice measurements...10 Russia ............................................................................................................................ 30 6..................................................................... 39 8.................................... 20 5..........................................................................................................................................1 Concepts .................... 31 Requirements for ice detectors .................................................................................................................................................. 17 4.......12 WMO/CIMO inter-comparisons of wind instruments under harsh conditions......... 30 6...............2 International Electrotechnical Commission (IEC) ................Table of contents 1 2 Foreword........................................7 Sweden ............. 30 6........................................................................................................................................................................................................... 30 6................................................................................ 13 3.............................................6 Measurements under icing conditions..........................................2 Icing types (extracts from ISO-12494).........................1 ICEmeter ...2 WMO/CIMO Recommendations ...................................................................... 30 6............................................... 23 5...................................1 Introduction ........ 13 Ice measurements as described in standards .............................3 Rosemount/BFGoodrich......................................................................2 Germany ...................................................................................................................................................................................................................................................................................6 UK ........................

........................... 51 Measurements in Norway............................................................................................................... 106 5/110 ........................................ 55 Measurements in UK..................................... 42 9............................................................................... 73 Labko Ice detector (Finland) ... 59 Measurements in Sweden........... 83 METEO device..............................1 Regional variability ......................................................................... 46 Measurements in Germany......................................5 Gerber.. 44 10............................................. 89 IceMonitor ............................................................................................................................................................................................................ 42 9........9 Instrumar IM101 .......................8 T20-series Ice Detectors............................................................................................... 79 EAG 200 (Germany) ......................................... 44 10............................................................................................................... 42 9. 49 Measurements in Slovak Republic .......................................................................................................................................................................4 EAG 200........................................................................................................................................................................................................... 43 9.... 72 Icemeter (Czech Republic).. 77 Rosemont................................................................................................................................................................... 99 Wind tunnel calibration ............... 44 10............................................................... 64 Measurements in Bulgaria................ 96 Instrumar IM101 ........................................................................................................................................... model 0872J (Finland) ............. 43 10 Long term recommendations for ice measurements in Europe.. 66 Measurements in Hungary ................................................................ 69 Measurements in Russia............................................... 100 References ...... 80 Gerber (USA) ............ 92 T20-series Ice Detector (Sweden) ............................................................................................................3 Permanent forum for monitoring icing in Europe ......................................... 45 Annexes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Measurements in Finland .. BF Goodrich..................................................................................................................................................................................................................................................................... 52 Measurements in Czech Republic ....................................................................................... 42 9...............................................................................................................................................................................................................................................................7 IceMonitor......................................................................................................................................................................................9.2 Requirements for measuring sites ....................................................................................6 METEO device...... 71 Measurements in Canada and USA..............................................

6/110 .

It contains information on: Definition of icing: WG 2 recommends adopting the ISO12949 standard. ISO. verification of existing icing sensors and mapping of icing occurrences and potentials in Europe. depending on users requirements together with validation and verification processes. a set of recommendations is presented concerning classification of sites and classification of sensors depending on severity of icing and the site climatic environment. Requirements and availability of ice detectors: It is shown that requirements on ice detectors are dependent of the user’s requirement (wind energy. Installation procedures are presented. Measurements under icing conditions: As the WMO has presently no specific recommendations for measurements performed in harsh conditions. Examples of existing data and experiences with existing ice detectors: A number of available long term experiments are presented concerning icing measurements and characterization of icing sensors. icing. meteorology etc. EUMETNET SWS I and II projects. power lines.g. improved forecasting systems. The present paper deals with the result of WG2 concerning measurements of icing as well as measurements performed under icing conditions. measurement on atmospheric icing.1 Management summary The COST-727 Action ”Measuring and forecasting atmospheric icing on structures” was established in April 2004. The following deliverables will be expected: Scientific and technical publications on measurements and predictions of in-cloud icing Publications on verification of icing forecasts European icing map Recommendations for WMO observations and further work . It is divided in 3 working groups dealing with modelling.) and on the application. EU/WECO and NEW ICETOOLS projects as well as entities such IEC/CENELEC. measurements and forecasting of icing. e. IEC and WMO) dedicated to icing on structures and icing measurements are shortly presented. Ger- 7/110 . Phase 1 of the action is dedicated to gathering available information for comprehensive stateof-the-art reports with the following deliverables: Reports on the state-of-the-art Inventory of users' needs based on analyses Working plan for the Second Phase of the Action Phase 2 of the Action is dedicated to R&D and will concentrate on research on in-cloud icing. These activities have taken place in numerous countries like Finland. modelling of icing processes. IWAIS are shortly presented. Past and present activities: International projects such as WMO/CIMO Instrument Intercomparison. Standards: Prevailing standards in use (ISO.

Switzerland.. etc. Russia. Slovak Rep. Bulgaria. 8/110 . These calibration centres are to be recognized for delivering approved certificates for icing detectors and ice-free sensors. Sweden.many.. It is recommended that long-term international calibration stations are established with a sufficient financial support for continuous operation. and indirectly from France. UK. Canada. Czech Rep. Recommendations for future activities: The establishment of test centres within the COST727 Action (Phase 2) have a temporary character. Norway.

WG2's activities will be dedicated to the following activities: a) create an inventory and collect available experimental data on icing as well as ancillary data b) review and assess existing ice detectors and their performance c) review and assess existing verification data from different sources d) contribute to the set up of icing measurements at different locations in Europe and to the development of existing test sites e) set up a data quality control scheme for measured icing data f) establish a basic data set for icing modeling and verification g) provide recommendations to set up a long-term icing measuring network and data base (to be submitted to WMO) h) establish an icing monitoring core group for collecting and maintaining data on icing during and especially after the course of the Action i) develop the scientific and technical bases of specifications of ice detectors j) set up recommendations for testing/approving ice detectors and ice/free sensors. where the focus is to establish the “state of the art” in the field of icing and to indicate the data available for icing in Europe. The present report was prepared by Working Group 2 WG2 “MEASUREMENTS AND DATA COLLECTION ON ICING” and was given the following objectives: “Measurements over a specific period of time on ice accretion and testing of icing sensors will be based on existing test sites in the far north (Luosto/Finland) and in the Alpine region (Guetsch/Switzerland). The MoU focuses mainly on in-cloud icing.2 Introduction 2. Additional experimental data from other ongoing activities will be used for this Action. Canada and Russia is also presented. Some information from Japan. This report is based on input from WG members according to their current activities and related references. Furthermore it includes information on measuring activities in Europe related to both wind turbines and electric overhead power lines.1 Memorandum of Understanding COST Action 727 "MEASURING AND FORECASTING ATMOSPHERIC ICING ON STRUCTURES" was established in April 2004 according to a Memorandum of Understanding (MoU) [1]. but wet snow and freezing rain are included when appropriate.” The present report summarizes the information and material WG2 has collected for the Preparatory Phase (Phase 1) of the Action. 9/110 .

These WGs are merged for Phase 2 and renamed WG1. Aigoual station. and to give a proposal of specification of improved measurements under cold climate and ice affected sites.1 WMO/CIMO Wind Instrument Intercomparison Mt.2 Interface with WG1 and WG3 During Phase 1 the Action WG2 had a close collaboration with WG1 and WG3 which were entitled: WG1 “ICING MODELLING”.2. WG3 needs similar data for mapping icing climates in Europe. [2].2 EUMETNET / SWS I&II SWS I: 1997-1998 The EUMETNET launched a study of severe weather sensors (SWS) to summarize the experiences concerning icing effect on sensors. In particular. to make a market survey in ice free gauges available. Aigoual.10 for more details regarding the available data. See section 6.3. WG3 “MAPPING AND FORECASTING OF ATMOSPHERIC ICING”.3.3 Past and present activities This chapter is based on the input from different countries and covers both experimental work and administrative activities. France and Switzerland in the period 2000-2002 in order to analyze the performances of ice-free instruments under extreme meteorological conditions. knowledge in handling the ice affected data by the meteorological services. 2.11 for more details regarding the available data of the SWS projects. 2. [3] SWS II: 2000-2002 A documented experiment [4] has been conducted at three sites in Finland. 2. as well as information on measurement networks that can be incorporated in forecasting routines by National Weather Services. WG1 requires knowledge of what kind of icing data are available and can be provided for the purpose of validating and calibrating icing models as well as meteorological data that are unaffected by icing. France: 1992-1993 A documented experiment has been conducted at the Mt. 10/110 . requirements of ice free sensors and direct measurements of icing. See section 6. France (within a joint venture between France and Switzerland) in order to analyze the performances of a number of ice-free anemometers under extreme meteorological conditions.

2. It is shown experimentally and by numerical simulations that icing of rotor blades or other components lead to decreased production due to ice accretion or safety demands. 2. which was partially supported by the European Commission DG XII Non Nuclear Energy Programme aims at the investigation of wind turbines under cold climate operation. 2.cigre.5 IEA WIND R&D Annex XIX IEA R&D Wind is an agreement between 19 countries and the European Commission to follow international development on wind energy deployment and to stimulate co-operative research and development of wind technology. The aim of the work is to reduce the risk that originate from cold climate and thereby reduce the cost of wind electricity produced in cold climates [7. Environment and Sustainable Development Programme. The icing effect is directly related to the climate of the site of the wind turbine.7 IEC/CENELEC The International Electrotechnical Commission is the standardization body for all electrical components in parallel to ISO (see below). 2. Since the start-up. International Energy Agency (IEA) R&D Wind started Annex XIX. 8].3. IEC prepares standards for the design of overhead 11/110 .3. Study Committee B2 „Overhead lines“ deals also with meteorological aspects such as icing on overhead lines (WGB2. Certification and Operation” NEW ICETOOLS NNE5-2001-259. Extreme low air temperature again set new demands for design parameters. and varies strongly from region to region in Europe. 2002-2004. was partially supported by the European Commission Energy. The aim was to produce tools and information to improve safety. In 2001. distribution of electric energy.6 CIGRE The „Conseil International de Grands Réseaux Electriques“ is a non-profit NGO dealing with all types of electrical component: production. It is research oriented and organized in study committees.16 „Meteorology for overhead lines“) (www. the participants of Annex XIX have been collecting operational experiences from selected sites that experience frequent atmospheric icing or low temperatures.3. A second 3-year period of Annex XIX started late 2005 with Italy and Germany as additional members. 2.org).4 NEW ICETOOLS “Wind Turbines in Icing Environment: Improvement of Tools for Siting.3 WECO: Wind Energy in cold climates “Wind Energy Production in COld climates” WECO (JOR3-CT95-0014) 1996-1998.3. Icing of anemometers and other wind gauges typically lead to wrong estimation of wind power potential and operational problems of wind turbines [5]. Collected data include information on performance of standard wind turbines as well as performance of adapted wind turbine technology specifically developed for cold climate sites. availability and reliability of wind turbines and their components and thus improve the economics of wind power production in icing environments [6]. transmission. “Wind Energy in Cold Climates”.3.

2. IWAIS Workshops that have been organized every 2-3 years since 1982 are meant to be the main international gathering of researchers in icing and icing related problems and assess the state of the art in icing research [10]. CENELEC is the European counterpart of IEC (www. The eleven published proceeding volumes of IWAIS contain approximately 5000 pages of information on icing related issues.9 IWAIS The International Workshop on Atmospheric Icing of Structures is an informal institution supported by research institutions and utilities. The standard describes the ISO standard instrument to measure icing (see Section 4. but not for overhead lines.ch).3. 2.8 ISO The International Standard Organization has issued the ISO-12494 [9] recommendation which represents today the most widely used reference for icing on structures in general.1). 12/110 .iec.3.lines taking into account meteorological parameters such as icing.

When the flux of water droplets towards the object is less than the freezing rate. In general. A main precondition for significant ice accretion is the dimensions of the object exposed and its orientation to the direction of the icing wind. The maximum amount of accreted ice will depend on several factors. the ice growth will tend to be wet.2 Icing types (extracts from ISO-12494) Atmospheric icing is traditionally classified according to two different formation processes: a) b) c) Precipitation icing (including freezing precipitation and wet snow). In-cloud icing (also called rime/glaze. Even if the temperature is below the freezing point of water. WG2 recommends adopting as standard the ISO-12494 and in particular the definitions presented in the following section. Table 2 gives a schematic outline of the major meteorological parameters controlling ice accretion. etc.. because the droplets do not have the necessary time to freeze. When the water flux increases. shear strength.3 Definitions and meteorological conditions 3. The physical properties and the appearance of the accreted ice will vary widely according to the variations of the meteorological conditions during the ice growth. and the ice growth is said to be dry. 3. 13/110 . the droplets may remain in the liquid state. Besides the properties mentioned in table 1. A cloud or fog consists of small water droplets or ice crystals. a classification may be based on other parameters. may be used to describe the nature of accreted ice. such as compressive strength (yield and crushing). However. ice accretion can be defined as any process of ice build up and snow accretion on the surface of an object exposed to the atmosphere. dry icing results in different types of rime (containing air bubbles). Such super cooled droplets freeze immediately on impact with objects in the airflow. before the next one impinges. the most important being humidity. see table 1 and 2. including fog) Hoar forst (not considered here?). each droplet freezes before the next droplet impinges on the same spot. temperature and the duration of the ice accretion. while wet icing always forms glaze (solid and clear).1 Generic definition According to the ISO-12494 standard [9]. This is explained in more detail in chapter 7 of the ISO document. other parameters.

the types of ice do not need to be described in more detail. but from an engineering point of view. 1 large hours any flakes very high high medium low hours see fig. 1 Soft rime see fig. The density of accreted ice varies widely from low (soft rime) over medium (hard rime) to high (glaze). 1 medium medium small hours days days In practice. Table 1: Typical properties of accreted atmospheric ice1 Type of ice Glaze Wet snow Hard rime Soft rime Density kg/m3 900 300-600 600-900 200-600 Adhesion & Cohesion strong weak (forming) strong (frozen) strong low to medium General Appearance Colour transparent white opaque white Shape evenly distributed/ icicles evenly distributed/ eccentric eccentric. 1 Hard rime see fig. 1 see fig. 1 14/110 . controlling atmospheric ice accretion Type of ice Air temperature C Wind speed m/s any Droplet size Water content in air medium Typical event duration Precipitation icing Glaze (freez. accretions formed of layers of different types of ice (mentioned in table 1) may also occur.Figure 1 gives an indication of the parameters controlling the major types of ice formation. pointing windward eccentric pointing windward Table 2: Meteorological parameters. 1 see fig.-10 < ta <0 ing rain or drizzle) Wet snow 0 < ta < +3 In-cloud icing Glaze see fig.

Glaze can be accreted on objects anywhere. where cold air may be trapped below warmer air aloft. Glaze may result also in formation of icicles. The surface temperature of accreting ice is near freezing point. Glaze is caused by freezing rain. because of the occurrence of free water in the partly melted snow crystals able to adhere to the surface of an object. which afterwards fall through a freezing air layer near the ground. due to wind and gravity. may flow around the object and freeze also on the leeward side. The accretion rate for glaze mainly varies with: Rate of precipitation Wind speed Air temperature 3. and in this case the resulting shape can be rather asymmetric. 3. Such temperature inversions may occur in connection with warm fronts or in valleys. 3 2 15/110 .2 Wet snow Wet snow is.Wind speed (m/s) 25 20 15 10 5 0 -20 -15 -10 Soft rime Hard rime Glaze -5 0 Air temperature ( degree Celcius) Figure 1: Type of accreted ice as a function of wind speed and air temperature2.2.1 Glaze Glaze is the type of precipitation ice having the highest density. and therefore liquid water. Wet snow accretion therefore occurs when the air temperature is just above the freezing point. freezing drizzle or wet in-cloud icing and normally causes smooth evenly distributed ice accretion. The curves In figure 1 shift to the left with increasing liquid water content and with decreasing object size Freezing rain or drizzle occurs when warm air aloft melts snow crystals and forms rain drops.2. when rain or drizzle occurs at temperatures below freezing point3.

the radial position. is common at low temperatures. i. It results in an ice accretion more or less cylindrical around the string. 16/110 . Glaze ice may occur on the tip while rime occurs near the root. but as the width (diameter) of the object increases. 3. During significant icing on small. the fraction of melted water and the wind speed.2. objects which will not rotate around the longitudinal axis due to eccentrically loading by ice. 4 The liquid water content of the air becomes so small at temperatures below about -20°C that practically no in- cloud icing occurs. among other things. The density and adhesive strength vary widely with. This mechanism may continue as long as the ice accretion is going on4. Figure 1 should be extended to 80 m/s.e.e. 5 Comments to ISO definitions: Icing types on wind turbine blades depend on the velocity i. Evenly distributed ice may also be formed by in-cloud icing when the object is a (near) horizontal "string" (linear shape) which is rotary around its axis. which is due to direct phase transition from water vapour into ice.2. and normally does not result in significant load on structures5. Hoarfrost is of low density and strength.4 Other types of ice Hoarfrost.The snow will freeze when wet snow accretion is followed by a temperature decrease.3 Rime Rime is the most common type of in-cloud icing and often forms vanes on the windward side of linear. The most severe rime icing occurs at freely exposed mountains (coastal or inland). the ice vane changes its form. or where mountain valleys force moist air through passes. linear objects the cross section of the rime vane is nearby triangular with the top angle pointing windward. non-rotary objects. The accretion rate for rime mainly varies with: Dimensions of the object exposed Wind speed Liquid water content in the air Drop size distribution Air temperature 3. and consequently both lifts the air and increases wind speed over the pass. see chapter 7 of the ISO document. The accreted ice on the windward side of the "string" will force it to rotate when the weight of ice is sufficient.

the length should be 1 m. The speed of rotation of the vertical collector is not critical. 7 For this purpose it is preferred to use sets of horizontal collectors (rods) which are oriented orthogonal. In this recommendation. The only requirement used in [12] is that the “reference ice load” should be related to a “horizontal. The cylinder should preferably be placed in an area. but also the possibly weaker adhesion of wet snow. The IEC 60826 [12] gives rules for design of overhead lines in order to make them function reliably under icing conditions. 4. as long as the results correspond to those for 10-m height. The cylinder is placed 10 m above terrain6. different erection heights above terrain are accepted. a standard ice-measuring device is described as: A smooth cylinder with a diameter of 30 mm placed with the axis vertical and rotating around the axis. For practical reasons. However.5 m. if heavy ice accretion is expected. as this aspect is considered to be the responsibility of each country. except for electric overhead line conductors. Consideration must be given to the maximum snow depth during the winter. Recordings of ice weight may be performed automatically. it may be provided with a motor to force the rotation.2. IEC does not give numeric values for ice loads which are to be taken into account in various countries. especially the wet and dry growth of freezing rain. icing is often the most significant design parameter in economic terms. like the Soviet standard ice collector [11] or the Canadian Passive Ice Meter (PIM) as described in [12]. where snow is blown away. If the cylinder cannot rotate freely due to wind drag. The vertical cylinder is not fully appropriate for freezing rain in the wet growth stage7. 6 17/110 .4 Ice measurements as described in standards Atmospheric icing affects all kinds of installations in susceptible areas.1 International Electrotechnical Commission (IEC) Overhead lines For electric overhead power lines.1 International Standardization Organization (ISO) The ISO issued in 2001 a standard [9] for ice accretion on all kinds of structures. The following standards have been found depending on the field of application. The cylinder length should be a minimum length of 0. However.2 4. For this purpose it is preferred to use sets of horizontal collectors (rods) which are oriented orthogonal. it is important to be aware of the different properties of various icing types. 4. like the Soviet standard ice collector [11] or the Canadian Passive Ice Meter (PIM) as described in [12]. but. circular conductor of 30 mm in diameter”.

Furthermore. humidity. may occur and shall be included where appropriate. wake loads. impact loads. The following other environmental conditions shall be at least taken into account and the action taken stated in the design documentation: temperature. etc. hail.IEC has also issued a Technical Report on measurement of ice loadings for overhead lines [13].3 World Meteorological Organization (WMO) There are presently limited standards defined by the WMO and its Commission for Instruments and Methods of Observation (CIMO) either for measurements performed under harsh conditions (e. snow and ice Other extreme environmental conditions that shall be considered for WTGS design are wind speed. Moreover. environmental limits specified by the manufacturer shall be observed. It includes a description of historical test spans in Europe as well as in countries outside Europe. rain. corrosive. photochemical. the international standard IEC 61774 [13] covers meteorological data with respect to icing observation and measurement concerning electric overhead lines. air density. ice loads. 4. These procedures are useful because data conversion from different systems is very difficult. electrical or other physical action. ed 2 [14]: ”Environmental (climatic) conditions other than wind can affect the integrity and safety of the Wind Turbine Generator Systems WTGS.2. lightning and earthquakes. solar radiation. icing) or for the measurements of icing itself at Automatic Weather Stations (AWS). This standard has been designed on the basis of previous experiences of icing observation and has been adopted in such a way that the information acquired is unified.g. During installation. mechanical. Other loads such as wave loads.2 Wind turbines Extracts from IEC 61400-1. by thermal. and snow and ice” The standards issued by CENELEC include for Europe the same principles of overhead line design as described in [12]. 4. Items such as the following should be considered: wind speed. 18/110 . No minimum ice requirements are given for the standard WTGS classes. combinations of the climatic parameters given may increase their effect.

The mass distribution increases linearly from zero in the rotating axis to the value µE at half radius and stays constant to the blade tip. The ice density is to be taken as 700 kg/m3. 4.3exp( 0. the influence on fatigue loads from extended periods of frost and the time period of ice accretion. etc. Extract from GLRP3. urban. ocean. taking into account the recommendations of the final report of the recent EUMETNET (Network of European Meteorological Services) / Severe Weather Sensor experiment SWS II [15]. The mass distribution (mass per unit length) is to be assumed on the leading edge. The value µE is given by: E E kc min (c max c min ) where density of the ice : 700kg/m 3 E k 0. tropical. For the rotating machine the situations "all blades are iced over" and "all but one blade iced over" have to be analyzed. 16]. This topic should be further handled during the CIMO XIV session.0-1998 [16]: For non-rotating parts an ice formation with a thickness of 30 mm on all sides is to be assumed on all exposed surfaces.However.4 Icing and wind turbines Icing of wind turbines is briefly described in the standards [14.00675 0. desert. extracts concerning icing from present wind turbine design standards are presented.32R/R 1 ) R rotor radius R 1 1m c max c min maximum chord lenght chord lenght at the tip 19/110 . In the following. the results from the EU-project NEW ICETOOLS [6] indicate that these standards seem to underestimate: the actual amount of ice. However. the WMO/CIMO Expert Team of Surface Measurements has recently decided to include in the next version of the CIMO Guide – or in a later version .a chapter concerning meteorological measurements performed under harsh conditions such as artic/mountain climate (icing).. For the Wind Energy Converter (WECs) at standstill also the rotor-blades have to be analyzed with this ice cover.

5 Meteorological measurements under icing conditions
5.1 Introduction

The accuracy of the surface measurements of various meteorological variables is essential for meteorological services, researchers in climatology (e.g. climate change8), aeronautical meteorology, etc. It is therefore essential to characterize the effects of ice accretion on the sensors and, when possible, to prevent it. The WMO Guide for meteorological measurements [17] does not define the temporal reliability of sensors, e.g. the required availability of data per year or per month, so that most meteorological services have specified their own targets for availability of data. Similar targets are defined also for other applications. Furthermore, the WMO Guide does not separately consider severe weather conditions like icing, even if low temperature is specified in the requirements. In the same way, the manufacturers typically specify their instruments’ performance for severe weather conditions by taking into account low temperature (for instance operating range: -40 C… + 50 C), but not icing. Presently, icing events are defined as periods of time where the temperature is below 0°C and the relative humidity is above 95%, a very simplified approach. Usually low air temperature is not a major problem for meteorological observations: for many sensors, this is taken into account e.g. by using shaft heating for anemometers with rotating parts (at small and/or mobile automatic stations, the power supply may not be sufficient even for shaft heating).

5.2

WMO/CIMO Recommendations

The following recommendations are stated by WMO/CIMO: Improve the quality of meteorological measurements under cold climate conditions, Provide manufacturers data for design of ice- free sensors, Provide users and providers of meteorological information better bases for selection of suitable sensors for their purposes. To improve the general knowledge on icing and icing climatology, the following recommendations for further activities are given: Improve the design of the instrument (mechanical) and heating system to optimize the required heating power Promote the development of icing observation instruments Promote the results of past, present and future experiments Promote national " icing maps" Promote a classification for "meteorological" sensors taking into account accuracy, climatic conditions and reliability of data required for different applications
The anticipated increase in air temperature will inherently lead to higher contents of water (vapour) in the lower atmosphere. In mountains, and especially in northern latitudes, there is therefore an increased frequency of temperatures near freezing, and together with more water available this could lead to higher frequencies of icing as well as higher icing intensities and ice loads on structures. An anticipated increase in extreme wind speeds will contribute likewise and indeed lead to significantly higher static and dynamic loads on exposed infrastructure. [Svein Fikke, private communication]
8

20/110

Promote the improvement of the WMO/CIMO Guide 8 for measurements in severe icing conditions Promote WMO-approved test sites for ice- free sensors, preferably combined with the use of icing wind tunnels for testing of sensors, including anemometers. It must be noted that most activities on developing requirements for instruments in harsh climatic environments focus on icing conditions only (i.e. in extremely cold mountainous/Arctic climates). Therefore, equipment for dusty and dry deserts, humid and hot tropics and oceans with a harsh climate need further investigation. For these climates only very limited guidance material is available on the implementation and maintenance of automatic observing systems and, therefore, further studies are necessary. Moreover, not only the performance and maintenance issues of a system are a point of concern but also destruction of instruments caused by extreme weather should be considered (e.g. tropical cyclones reaching 300 km/h or more.) In line with past developments and published material on this matter, documentation on the requirements for instruments for observations in harsh climatic conditions has to be generated. In particular the IOM report, as announced at CIMO XIII should be finalized and published. Moreover, as stated in the recommendations in the SWS II report (see above) more attention to this topic should be given in a future revision of the CIMO Guide [17]. Taking these recommendations into account, and in line with CIMO XIII, the CIMO management group has decided to continue the work on the provision of guidance material on implementation of instruments in harsh climatic environments as requested by Congress and Technical Commissions. To realize this, further study on already existing test reports should be carried out. Moreover the already announced IOM report has to be finalized and published. A decision should be made on how to implement requirements on the instruments capable of measurements in a harsh environment in a new revision of the CIMO Guide and it should be considered to carry out inter-comparisons of dedicated sensors in such environments. Preceding such inter-comparison an inquiry should be carried out among the Members facing shortcomings of today's equipment due to harsh and extreme weather.

5.3
5.3.1

Definitions
Meteorological icing Micing

Meteorological icing Micing is defined as the duration of a meteorological event or perturbation which causes icing [unit: time].

Meteorological icing can be characterized by: a) the duration of the icing event, and/or b) the meteorological conditions, and possibly with additional information such as: c) the total amount of ice accreted on a standard (reference) object during the icing event, d) the average and maximum accretion rate.

21/110

Automatic certified reference sensors are lacking for the determination of items a) and d), whilst the items b) and c) can be more or less achieved with presently available technology. However, it must be noted that meteorological icing is not easy to define. It is today widely accepted that it depends on the following factors: a) b) c) d) e) the shape of the object, the wind speed, the air temperature, the liquid water content LWC, the droplet size distribution,

the latter two being difficult to measure in operational mode. Tentative developments have been achieved, such as the Rotating Multicylinder RMC. Unfortunately, these cannot be operated in an automatic way and cannot therefore be implemented at automatic stations. New developments may improve this situation [18]. Today, ice accretion can be measured directly by instruments measuring: a) b) c) d) changes of a vibrating frequency (Rosemount, Vibrometer, Wavin-Labko), changes in electrical properties (Instrumar, Labko) the load of ice (ISO 12494) the growth rate of ice by yielding a yes/no output (at regular intervals) by a heating cycle e) optically (obstruction of light path, IR or reflection technologies: Infralytics, HoloOptics). In addition, icing can be measured indirectly by measuring the variables that cause icing (see Chapter 4.2) or variables that correlate with the occurrence of icing, such as cloud height and visibility.

5.3.2

Instrument icing Iicing

Instrument icing Iicing is defined as the duration of the technical perturbation of the instrument due to icing [unit: time].

Instrument icing is the effect of icing on the quality (e.g. degradation) of the measurements, depending on icing conditions as well as the design of the instrument. It can be today only recorded by analyses of video recordings, and/or regular visual observation, or by comparison of the measurements with a reference that is kept ice free. This definition is valid for all objects or structures. It can be easily generalized to “structural icing”.

22/110

On the contrary.5 Site Icing Index To be able to express the maximum expected amount of accreted ice at a certain site. The climatologists may provide information about Sn. Automatic Weather Station (AWS) is introduced indicating the degree of severity of local icing conditions.5. is introduced by using icing frequency. Instruments (or structures) will behave differently depending on the location of their installation. A sensor operated in northern countries may get frozen at the beginning of the winter and remain as such due to the low temperatures and the lack of sunshine.5 Intensity of icing g/100 cm2/h (typical) > 50 25 10 5 0-5 Icing severity Heavy Strong Moderate Light Occasional 23/110 .g. Table 3: Classification of sites according to the severity of icing (from EUMETNET/SWS II Report).4 Site effects In the preceding section meteorological icing has been shown to be different from instrumental icing. The following table was set up during the EUMETNET/SWS II experiment and describes tentatively the framework of such a classification. Site icing Days with meindex teorological icing / year S5 > 60 S4 31-60 S3 11-30 S2 3-10 S1 0-2 Duration of meteorological icing %/year > 20 10-20 5-10 <5 0-0. Sn. this instrument may be installed further south and work more or less undisturbed under milder. Maximum loads are not included in this definition. For the purpose of meteorological instruments a site icing index. the term ICE CLASS (IC) is introduced in the ISO 12494 standards. or at the precise station site. sunnier conditions. or both. it is recommended that a classification of sites. unless experience can supply the same information. in regard to meteorological instruments. Measurements should be carried out either where ice accretion is expected to be most severe. Measurements and/or model studies are necessary to obtain the information needed for a specific site. which (in general terms) tells how much icing can be expected at a given location. The station class may vary within rather short distances of a specific area. Therefore. the latter being the consequence of the former. the instruments characteristics must be evaluated as a function of the site of installation in terms of local icing conditions. but with different effects depending on the characteristics of the meteorological conditions and of the instrument design. e. 5. duration and intensity: Sn is the parameter to be used by the meteorological community to determine how severe ice accretion is expected at a particular site. Therefore. for design purposes.

a simple classification for instrument icing characteristics was introduced for analyzing the behaviour of the different available . especially at mountainous sites. During the EUMETNET/SWS II project. or specify existing ones for Automatic Weather Stations (AWS). In the following. A tentative classification of reliability of different types of sensors under icing conditions was performed in such a way.) in cold climate environments have to be properly heated to maintain their accuracy under icing conditions. Instruments performing measurements (such as humidity.defined as icefree . road stations e. precipitation. wind direction. wind energy production.g. The state of the instruments was classified at all stations with a value between 0 and 3: Class 0: totally free of ice Class 1: light ice accretion. meteorological observations. but also at sites like airports. It must be kept in mind that for practical measurements the requirements and the choice of sensor depend significantly on the goals of the application and on the location of weather station (icing climate). in northern Europe. Accurate meteorological measurements under cold and icing conditions are required for various applications such as aviation. and the magnitude of the resulting errors. This leads to the following classification proposal.5. but could be used reliably at some other less demanding sites. Some of them will not fulfil the requirements of the World Meteorological Organization (WMO) for accuracy and availability when operated under icing conditions. temperature. there are several types of ice-free sensors available on the market. Class 3: totally covered with ice – sensitive elements are covered with ice. The practical requirements on accurate operation of sensors are even more demanding for many applications other than for meteorological synoptic purposes. Class 2: medium ice accretion. The results were summarized with respect to general requirements for a meteorological synoptic station. emergency services.sensors under icing conditions. agriculture. these definitions are used to classify sensors accordingly. radiation.6. 5. wind speed. without obvious effect on the sensitive part but which could influence the wind field.1 Ice-free meteorological sensors Icing environments set special requirements for sensors. 24/110 .6 Measurements under icing conditions In the preceding section. This classification made it possible to study statistically the effects of different amounts of ice upon sensors. Three stars indicate that the sensor is close to 100 % ice-free but has other significant errors. etc. etc. transportation. probably disturbing the measurements – Sensitive elements seem free of ice and wind field is obviously disturbed. it was possible to set up definitions which are used to describe the behaviour of meteorological instruments under harsh conditions and to select new adequate sites. tourism industry. and sensors with five stars can be strongly recommended for measurements at sites with harshest icing conditions. Today. The sensor got one star when not adequate to provide reliable data at the climatic conditions met at sites with more than 60 icing days per year.

A value of PI near 0 reflects a good performance of the instrument in terms of icing (e. The Performance Index PI characterizes the behaviour of the instrument under icing conditions. the following definitions are needed to characterize the properties of an instrument under icing conditions: The meteorological icing Micing is the duration of the icing event (see §5. Recovery time: delay between the end of the meteorological icing and the full recovery of the performance of the instrument Instrument icing can be smaller. good heating).2 Performance Index During a meteorological icing event. both expressed in the same time unit. poor heating) for a time period (much) longer than the meteorological icing.1) The instrument icing Iicing describes the effect of Micing the on the instrument (see § 5.g.) may be considered as well adapted for the station’s climatology. On the other side. good coating. especially in northern countries with low solar irradiance in winter. etc.6. equal or longer than meteorological icing. the relationship between meteorological and instrument icing can be expressed in the following way: An instrument which remains free of ice during a meteorological icing period (good heating.3. PI I icing M icing The PI can be used for the selection of the instrument as function of some station’s classification. 25/110 . Further useful definitions: Incubation time: delay between the beginning of the meteorological icing and of the instrumental icing. The incubation time indicates how quickly the instrument responds to icing while the recovery time may be much longer that the meteorological icing.3.2). Values of PI between 0 and 1 may be acceptable as long as ice detector information is available to “flag” dubious periods of measurements. In summary. an instrument that becomes frozen during a meteorological icing period and remains in that state after the meteorological icing period must be classified as poorly adapted to the site’s environmental conditions.g. Values of PI higher than 1 indicate a sensor which is sensitive to icing (e. This leads to the following definition: The Performance Index (PI) is the ratio of the instrument icing to the meteorological icing.5.

. The range of this classification extends from ICI5 (PI = 0. the Instrument Class Index. little 99 .. the goal is now to build a common indicator by combining PI and Sn (see § 5. n=1-5) and the effect of icing on the results’ quality as described by the PI defined above.g.5). Instrument to be used 69 .... classification) of his station and requirement.. Instrument Class Index ICI5 ICI4 ICI3 ICI2 PI for S1 … S5 0 0 …1 1 . 26/110 .g. 0 % mended for such applications ICI1 20 .. e. 70 % sensitive to icing. which corresponds to PI values ranging from 0 -> depending on the different site icing indices S1 to S5 as indicated in Table 4. long recovery time for high latitude stations without sun during whole seasons). This leads to the following graphical representation (Figure 2) where the user could select the class of instruments needed to fulfil his requirements depending on the location (e...5).. The interpretation of the ICIn index is strongly linked to the station’s site icing index (Sn. 20 Mean availability in Remarks % for S1 … S5 Excellent instrument not 100 % sensitive to icing Good instrument.5. A potential indicator may be given by ICIn (n=1... 40 % only with separate icing detection Instrument not recom39 .6. 90 % sensitivity to icing Instrument moderately 89 . availability = 100 % perfect icing non-sensitive instruments) to ICI1 (PI = very high values.. 5 5 ... availability 40 % instruments which could remain frozen for a very long period after the meteorological icing period. Table 4: Classification of instruments in terms of mean performance depending on the station’s site icing index (the availability values displayed in italic are purely hypothetical and will have to be specified in future). Therefore.3 Instrument Class Index It is evident that a classification for meteorological sensors will be difficult to achieve taking into account accuracy and required reliability of data combined with climatic conditions.

The availability will then be computed considering the severity and duration of icing periods. 27/110 .Instrument classification 100 Availability [%] 80 60 40 20 0 S1 S2 S3 S4 S5 Site Icing Index ICI1 ICI2 ICI3 ICI4 ICI5 Figure 2: Graphical display of the instruments’ classification presented in Table 4 (hypothetical sensors) reflecting the availability of an ICIn sensor installed at a Sn site. In reality. the ICI will have to be determined for each Sn.

so that the ice load on them could be continuously measured [19. The oldest data are from 1957. On the Deutscher Wetterdienst’s meteorological observatory Lindenberg. More details are given in Annex 1.3 Slovak Republic The Slovak Hydrometeorological Institute has data on icing from 13 stations. 50 and 90 m above ground. The main purpose of the Luosto test station is to measure icing as well as the behaviour of meteorological instruments. The contributions in this chapter are based on information provided by members of the COST Action 727. More details can be found in Annex 2 6. as indicated. ice measurements are currently made at 10.1 Finland VTT and Digita (former Finnish Broadcasting Co. The Luosto test station (500 m asl) in northern Finland was set up during the winter 2000/2001 by Finnish Meteorological Institute. Other data may be available in some countries.2 Germany Icing measurements were carried out at altogether 40 locations in the eastern part of Germany during 1965 – 1990. More details may be found in the Annexes. More details can be found in Annex 3. and VTT has also later performed ice detector tests in four locations during the period 1998 – 2005. A standard observation pole has been used for all stations. ranging from 115 to 2 634 m asl. 6. Since 1991 five stations are still in operation.) have made measurements of icing on tower structures and measurements of drop size and liquid water content of clouds and comparisons of meteorological instruments on hilltops in severe icing conditions in Finland since 1986. Distribution Dept. Ice detectors were also tested [21].20].. 6.6 Examples of existing icing data Numerous activities in the field of measurement of icing have already taken place in different countries. An operating 128m tall TV tower and a 7. The following gives an overview of what has been achieved to date. 28/110 . up to 35 locations were operated simultaneously.5 m test tower at Ylläs (700 m asl) have both been equipped with load cells. The measurements are both visual and by instruments.

n=59 20 15 10 5 0 40/41 50/51 60/61 70/71 80/81 90/91 t (years) Figure 3. has operated more than 20 sets of racks for ice measurements in 16 locations in mountainous areas for power line design purposes. where one leg is perpendicular to the main icing wind direction. 6. [23]. such as rotating multicylinder. Each set consists mainly of two perpendicular racks. Ice accumulation was derived from web-camera pictures of wires.m ] -1 Studnice 800 m. EGU Brno has operated a test site on Studnice (800 m asl) continuously since 1940. As a part of the development of wind farms in the costal mountains of Norway new ice monitoring programs have recently been set up. Most of these racks are in coastal mountains in the range of 600 – 1 200 m asl. The monitoring results have been compared with an ice model and the results are reported in Harstveit [22] and Harstveit et al. but hopefully useful information will become available from these programs in the future. Statnett SF. The station was equipped with ice-free wind sensors. Ice loads were measured on a rack with orthogonal rods 2 m above ground. ice indicators and web-cameras are utilized. The annual maxima of loads on this rack for the period 1940/41 – 1998/99 are presented in Figure 3.5 Czech Republic Two institutes have performed icing studies in Czech Republic. Unfortunately.4 Norway A measuring station for ice monitoring has been operated at a coastal mountain of about 800 m asl in central western Norway with the support of the Norwegian Research Council and two Norwegian energy companies.6. This unique time series is outstanding. During the period 1978-2000 the Norwegian Power Grid Company. Accumulated ice at 10 m up to a maximum value of 20-25 kg/m where reported. EGU Brno and Institute of Atmospheric Physics (IPA. temperature sensors and a web-camera. since it is the only series of this kind in the world covering such a long time period. More details can be found in Annex 4. Some sets were established to study the effect of local topography. no data are yet available. 25 Q [kg. Different ice detection equipment. The Studnice site has also a test installation for power lines consisting of a central observation tower and 2 spans of about 250 m length on each side. Measurements took place during two winters. 29/110 . Prague).

6. 6. The longest running site. in Ritsem. More details can be found in Annex 6. Many of these sites lasted only a few years before being closed down for financial reasons.11 Canada Some information on measurements in Canada is found in Annex 11. The IPA has a similar instrument. although it has been ‘mothballed’ occasionally so there is no continuous ice measurement over this period. aluminium and covered variety as well as fiber optic systems such as Optical Pipe Ground Wire (OPGW).7 Sweden Three different sensors (IceMonitor.8 Bulgaria More details can be found in Annex 8. More details can be found in Annex 5. Milesovka (837 m asl) and Nová Ves.6 UK Test data on icing has been available on rotating rigs and test spans since 1988.Furthermore. It currently monitors wind speed and direction. It has operated the Gerber instruments on loan from the UK Meteorological Office. along with icing data. EGU Brno has developed the instrument “METEO” which is installed at 14 locations in the country. Åre and Drammen (Norway). was established in 1991 and is still currently open.9 Hungary More details can be found in Annex 9.10 Russia Some information on measurements in Russia is found in Annex 10. IceMeter. 6. The measuring probe is a vertical rod of 30 mm diameter. More details can be found in Annex 7. 6. 6. Measurements have been made on conductors from 16mm² to 800mm² of the copper. 30/110 . ice loads (by time lapse video cameras and also load cells). at Deadwater Fell in Northern England. HoloOptics and Segerstrom) are currently being tested under field conditions in Sweden and Norway. installed in two locations. temperature. fiber-wrap and All Dielectric Self-Supporting (ADSS). Meteorological parameters are measured. precipitation and relative humidity. 6.

13 EUMETNET/SWS II project The EUMETNET "Severe Weather Sensors II" project (SWS II) tested 15 wind sensors. The three test sites were located in northern Finland.To evaluate and publish the results of this analysis in a WMO publication Icing phenomena have been studied from visual sensor control by the observer and by studying the data.To determine the suitability of these instruments for long-term unattended operation especially in a mountainous environment .6. It was noted that: The importance of the icing phenomenon could not be characterized from the ice detectors. the difficulties of performing measurements under harsh icing conditions are reported in the following way: The only sensors having supported severe icing events without noticeable measurement errors are the Pitot sensors and one vertical axis anemometer. . Manufacturers commercializing measurement instruments should put at the users’ disposal a detailed and complete documentation including a detailed installation and maintenance book and the exact metrological specifications of the sensors. The objectives of the inter-comparison were the following: . in the Swiss Alps and close to the Mediterranean in the French mountains.12 WMO/CIMO inter-comparisons of wind instruments under harsh conditions The WMO Wind Instrument inter-comparison was organized following the recommendation of the Commission for Instruments and Methods of Observation CIMO [2]. wind direction and air tem31/110 . but the power consumption can be relatively low if the sensors are properly designed. It appears difficult to be both “accurate” and rugged for severe icing. it can be seen that heating power is required especially for wind measurements.To make proposals for further improvements to the WMO regulatory materials concerning the measurements of wind . all with more than 60 days/year of atmospheric icing. It was carried out under the aegis of the WMO by both Meteo-France and the Swiss Meteorological Institute (presently MeteoSwiss) at the Mt. In the conclusions of the report. - 6. The meteorological performance of these sensors is not perfect and does not meet the WMO accuracy recommendations.To derive performance characteristics on the operational use of wind sensors based on the detailed record of their measurement values and a record of the prevailing atmospheric conditions. These sensors require a high amount of energy for heating (300 to 500W). 6 temperature and humidity measurement systems with different types of shields and 4 solar radiation sensors equipped with heating. From the results presented here. The tests and verifications showed that wind speed. Aigoual (France) from July 1992 to October 1993. During the project also different methods of measurement of atmospheric icing were used and tested. The formation of ice made almost all the calculated parameters incoherent.

results show that strong icing conditions may dramatically disturb the measurements. 32/110 . However. None of the tested systems were able to fully withstand the harsh climatic conditions prevailing at such sites. the reliability of temperature and humidity measurements does not yet reach the level available for wind measurements.perature could be measured with high accuracy and high reliability at cold climate sites under most severe icing conditions even at automatic weather stations. Concerning test measurements on the heating/ventilation systems for solar radiation measurements. For temperature and humidity sensors. under harsh conditions. some of the shields provide significant improvement in comparison with measurements performed with other systems in use at different national meteorological services. It was not possible to study the performance of the various sensors versus intensity of ice accretion due to the lack of dedicated sensors to measure icing [4].

Some of the instruments measure icing rate.3 Icing types All icing types that adhere on static or moving structures can be harmful and need to be identified. some measure the weight of ice (persistence and maximum loads) and some indicate if an icing event is ongoing. Range of use and data verification could possibly be carried out in icing wind tunnels. all the parameters that have an effect on the amount of water ending up on a detection surface. Requirements regarding time resolution. size of water droplets.26]. such as ice sublimation due to dynamic heating in high velocity airflows and Ludlam’s limit. some sensors have been designed for aviation purposes and perform well on airplanes.1. 7. 33/110 .1 Purposes of measurements At the present time.e. the purpose for using ice detectors needs to be defined. Therefore. The parameters that have an effect on the operation of an ice detector are air velocity.1. 7. Some detectors are clearly designed to indicate incipient icing only whereas others have been designed to measure the total mass of ice accumulation. which sets the upper LWC limit of use for some ice detectors [24]. For example. Heated sensors may be iced up during non-icing periods due to melting of dry snow.4 Verification of performance Considerable deviations between the results of ice detectors of the same type and even similar ice detectors can be found [25. Wind tunnel tests include various kinds of testing carried out in an atmosphere of low and moderate temperatures. and are therefore not further specified in these generic descriptions. The primary objectives were to quantitatively find out the effect of icing on wind speed measurements and to evaluate the effectiveness of measures to prevent ice or snow accretion on specific objects.7 Requirements for ice detectors 7. amount of liquid water present and the physical size of the probe i. many sensors that are designed and labelled as ice detectors are available. The range of use should be defined by means of these parameters and should be validated in controlled circumstances (see below). it must be kept in mind that in certain circumstances the operation of the ice detector may be affected by other phenomena.1 Concepts 7. 7. Definition of the range of use and some calibration scheme might improve the current situation. Kanagawa Institute of Technology (KAIT) has conducted wind tunnel test for investigation of icing events on airfoil models and anemometers as described in Annex 21.2 Range of use The range of use varies between different ice detectors.1. threshold values as well as response time of sensors depend on the purpose of individual measurements. measuring range. where the icing condition can be regulated and monitored.1. but may not be very well adapted for meteorological purposes due to different environmental conditions as indicated below. Furthermore.

A further possibility lies in the development and long term operation of “icing test centres” similar to (or included in) the Regional Instruments Centres (RICs) of the WMO where market available and future instruments could be tested under different climatic conditions (e.g. Scandinavia, Alps, Pyrenees, etc.).

7.2

Siting of icing sensors

7.2.1 Micrositing Ice sensors should be placed so that the detection surface of the device faces up wind and free air flow is granted. In addition all ice detection devices should be placed above tree tops and other possible obstacles. Appropriate locations for ice detectors are support structures of overhead power lines, wind turbines, link masts of mobile phone networks and in general high structures that provide free air flow around the ice detector. ISO 12949 [9] recommends 10m measurement height above ground. However, as icing measurements are dependent on the different application types, ice sensors can be installed at different heights. Automated weather stations are not generally appropriate as they are located close to ground level and seldom provide a correct representation of those icing conditions that prevail at a higher level e.g. wind turbine rotors.

Details such as the mounting orientation and height detectors will have to be analysed directly at the test centre sites.
7.2.2 Standard Reference and procedures Ice accretion on structures is not only a function of environmental parameters, but is also dependent on the properties of the accreting object itself, e.g.:

a) b) c) d)

size (diameter, width etc.) shape (flat, sharp edges, cylindrical, spherical etc.) flexibility (rigid/flexible member in bending/torsion etc.) orientation relative to wind direction (angle of incidence)

and to some extent: a) surface structure (paint, steel, concrete etc.) b) material (wood, steel, plastics etc.) Measurements of ice accretions therefore have to be specified with respect to devices, procedures, arrangements on site etc. The set-up must be designed in a way that causes the lowest possible influence on the accretion process itself: A standard reference device should always be part of the measurements, giving the traceability to standard measurements of ice accretion. Other parts of the set-up may help to establish the connections between “standard accretions” and the most important structural parameters as described above (size, shape, etc.). These extended measurements should only be executed at special selected sites, and collected data should be analysed and used, generally together with the standard measurements. Frequency of observations may be adjusted to the local conditions.

34/110

On sites where melting or shedding are likely to occur shortly after the accretion period, observations must be carried out before this happens for example by making use of camera systems. When automatic recordings are performed, it is important to add also visual observations during and/or after the accretion period, because only these types of observation can give sufficient information on such complex load situations. These visual observations have to be logged, and documented with appropriate digital camera pictures. Remote reading (including camera observations) makes it possible to get online information about an icing event so that the site may be visited in proper time.
7.2.3 Macrositing The following table 5 describes the information needed concerning ice types that are requested for the different fields of application.

Table 6 displays the density of meteorological networks equipped with ice detection systems for the different fields of application. For example a developer of a wind energy project would need relatively dense measurement network due to the considerable influence of local landscape to the icing conditions. Table 5: Ice parameters required
Requested information Application Wind turbine operation Wind project planning Power line design Power line operation Aviation Telecommunication masts Suspension bridges Transport (roads, railways) Meteorology and climatology x x x x x x x x x x Icing rate x Ice load x x x x x x x x x Icing time x x Persistency x x

35/110

Table 6: Location of ice detection needed by the different users.
Minimum distance to closest ice detection point Application Wind turbine operation Wind project planning Power line design Power line operation Aviation Telecommunication masts Suspension bridges Transport (roads, railways) Meteorology and climatology Less than 50km More than 50 On the site of from the site of km from the interest site of interest interest x x x x x x x x x x x x

7.3

Guidance for selecting ice detectors

7.3.1 General Appropriate ice detectors should be chosen with respect to the purpose of their use. There are presently two systems of ice detectors:

with status icing / no icing with recording of the whole icing cycle (ice mass, ice accretion rate). The size of the detector probe has a significant effect on performance of an ice detector. When icing detectors shall be selected, the purpose of the measurements has to be considered carefully. For example, smaller droplets in low speed airflow pass large objects more efficiently due to their low inertia and the fact that large objects deflect the airflow upstream from the object (collision efficiency). Therefore, no single ice detector can provide data that are directly applicable to all types of structures and conditions [27]. Ice detectors should be used only in conditions for which the devices were designed. For example, in order to design overhead lines, it is necessary to know the ice mass during an icing cycle. Measurement systems recording the whole icing cycle must be used in such a case without ice shedding or heating of a sensor during an icing cycle.

36/110

Early ice accretion warnings should represent a safety increase. Icing may be a hazard on an aeroplane wing even at temperatures above freezing due to cold fuel stored in the wing tanks. Road safety Mountain roads and bridges are commonly provided with wind and ice sensors for traffic safety information. etc. for combination with weather forecasts for pro-active road treatment (gritting etc) as well as falling ice from bridges. As airports are usually located at low terrain. require icing information so as to allow pro-active measures to be put in place (e.7. In terms of line design the use of ice sensors to provide historical data for probability purposes is important. Because power line conductors have low torsional rigidity and thus rotate along the span during icing events. pollution decrease and financial savings. heating of switches) and also to enable staff/de-icing trains to be on standby to combat icing.g. During operation. The ice can occur on power (3rd rail) and overhead power lines 37/110 .2 Applications Wind turbines Ice detectors are needed for planning and operation of wind turbines. Another type of ice detector is needed to indicate whether there is ice on some surface or not. The primary output would be the duration of the period that accumulated ice adheres on a detection surface without any heating or external forces. Ice detection is needed to determine whether the power supply is likely to be secure or whether an emergency response may be required. ice detectors are needed to detect incipient icing as quickly as possible for controlling heating systems. ski lifts. the emphasis is to detect freezing precipitation. Power lines Power line companies are mainly interested in local wind and ice loads and wind-on-ice factors. Transport facilities Railways. Icing is closely related to the speed of air flow and so the ice detector for wind turbine applications should preferably be attached to the outer part of the turbine blade. It could also enable pro-active measures. Results from the planning phase will influence the selection of type of turbine and equipment. The most important parameter is time in both applications. Airports Preventive treatment of runways as well as de-icing of airplanes before take-off may represent a important security problem as well as a heavy financial load for airport authorities. the instrument that measures icing rate or ice load for this application should be elongated horizontally and free to rotate [11]. Incorrect timing due to poor or badly interpreted data can mean that road treatment is less effective and can lead to increased accident rate. turbines and to give a warning about possible ice shedding in populated environments.3. cable cars.

antennas and/or guy wires.causing pantograph damage. At least 150 telecommunication towers have collapsed due to ice loads in the last 40 years. 38/110 . so that a combined ice and wind load needs to be estimated for the design. with potential effects on the quality of the transmissions and on the structural loads. It is also serious when accumulations on traction rails reduce the adhesion of train wheels. Wind loads are an important contributing factor. Telecommunications Telecommunication masts and towers are subject to icing on structures.

forced rotation [28]. to the best knowledge of the COST-727 / WG2 participants. f) Detection of the attenuation of ultrasonic signal on ice detector structure due to ice.newavionics. A further development is presently being designed in Switzerland to yield ISO compatible sen- 39/110 . e) Measurements of LWC and droplet size distribution. verification and requirements of ice detectors 8. h) Obstruction of light path. g) Detection of changes in the electrical impedance on the surface of the probe.com The ISO 12494 standard ice collector mentioned in the preceding chapters has been built in one version in Sweden (Combitech: automatic weighting. c) Infrared beam reflected on surface: the reflection characteristics change when the „mirror“ is covered with ice.1 Available ice detectors There are presently few available instruments on the market. free rotation) and two in Finland (Digita: automatic weighing. MeteoSwiss (CH) HoloOptics (SE) IAP (CZ) EGU (CZ) Combitech (SE) FMI (FI) (D) No longer manufactured VTT (FI). The following table displays the available and prototype instruments. (CA) Segerström (SE) www. there are some prototype instruments which seem very promising and may lead to interesting products after thorough testing and certification. However.8 Availability. These instruments are based on different working principles: a) Vibrating rods: the vibrating frequency depends on the state of the rod (yes/no information) b) Direct infrared beam backscatter: light is reflected as soon as the sensor’s surface is covered with a film of ice. STATNETT (NO) Gerber Scientific Inc. FMI: manual weighting. d) Measurement of the weight of ice.4 Jokkmokk IceMeister Manufacturer Goodrich (USA) Goodrich (USA) Hydro Quebec – Transénergie (CA) Boschung (CH) Infralytic (D). Table 7: List of available and prototypes of ice detectors on the market Item a b c d e f g h i Instrument Rosemount 0872J / 0871LH1 Rosemount 872C2 (ASOS-USA) SYGIVRE (Icing Rate Meter (IRM)) Vibrometer (Prototype) Infralytic IR detector (Prototype) T21. forced rotation). T23 and T26 ICEmeter METEO device IceMonitor ICECylinder (Prototype) EAG 200 Rotating Multicylinder (Prototype) Gerber Labko LID-3210C Instrumar IM101 V2. (USA) Wavin-Labko (FIN) Instrumar Inc.

However. Unfortunately. Unfortunately. 40/110 . For the detection of the meteorological icing (see definition above) Micing. if at all. it may be possible to use small-grid weather forecast models to perform an approximation of these values (see WG1 report).3 Verification of data Assessment and verification of existing data is scarce. unless visual observations or camera information or both are available at the station. Infralytic.sors with automatic weighting and forced rotation (Markasub) within the framework of a national project linked to the COST-727 Action. Devices similar to the ISO ice collector have been used at some locations in the past as well [29]. icing models require information concerning LWC and median volume diameter (MVD) of the droplet size distribution. there are a few systems which are either available on the market (Rosemount Goodrich). Vibrometer/Boschung. Some information may be found in the annexes related to section 6. Such models also include information about vertical air stability which influences the LWC and MVD. these parameters are measured only at research test centres. 8. or available as prototypes (HoloOptics. The following list summarizes the parameters which must be provided in order to characterize the start/end of all types of icing events as well as the total mass of ice accreted: Wind Temperature Humidity Precipitation Visibility Cloud base Radiation Present weather (precipitation type) Ideally. measurements of atmospheric icing are presently not included in the existing meteorological WMO standard observations.2 Data requirements for icing models Icing models can be run for sites where no specific in-situ icing data is available.) 8. etc. but where basic meteorological information is provided by Automatic Weather Stations.

See Annex 13 for detailed explanations 9.52 mm ice accretion. The ice signal activates at 0. the measuring device and its basic specifications are briefly described. model 0872J The Rosemount 0872J Ice Detector9 (prototype) is designed for various applications from meteorological measurements to radio tower de-icing programs [31]. Once de-iced.1 ICEmeter Manufacturer: Institute of Atmospheric Physics. As the ice detector enters an icing environment. Experiences made with the automatic measurement of the mass of icing accumulated on the collector of icing sensor developed in the Institute of Atmospheric Physics. It is also reasonably easy to get acquainted with the device and to adjust the parameters of the device to correspond with different icing climates. The added mass of accreted ice causes the frequency of the sensing probe to decrease in accordance with laws of classical mechanics. CZ. The sensing probe is a nickel alloy tube mounted in the strut at its midpoint with one inch exposed to the air-stream. ice collects on the sensing probe. The Rosemount ice detector uses an ultrasonically axially vibrating probe to detect the presence of icing conditions. Prague. Prague.9 Experiences with automatic instruments for ice measurements A number of instruments and resulting data are presented in the following and references are made to more detailed descriptions in the annexes. Then a short description of sites where the measurements were performed is given and the selected data presented.3 Rosemount/BFGoodrich. Ice detection of Labko ice detectors is based on longitudinal wire waves [30]. 41/110 . Czech Republic. Ice detector software monitors probe frequency and detect this decrease. LID (Labko Ice Detector). known as a freezing rain sensor and used in the ASOS network in the USA [31]. Usability of the device has been found to be good. 9. However. It seems to have a lower correlation with ice detections by humidity data than the Instrumar sensor [21]. First. See Annex 12 for detailed explanations. The probe is then heated to melt the ice.2 Labko ice detectors Wavin-Labko Oy has one ice detector model LID-3210C to offer. Labko's different versions have suffered from snow-induced icing indications and inability to melt all the ice. the sensing probe 9 This instrument is different from the Rosemount Model 872C3 Sensor. Next the results of the measurements and experience with the sensor are discussed along with outlines for its possible improvement. At the same point the internal probe heater power is applied until the frequency rises to a predetermined set point. 9.

4 EAG 200 The ice load sensor EAG 200 is the automatic icing measurement device that is used at DWD‘s measuring sites at present (see Annex 2.6 METEO device EGU Brno has developed an instrument for the measurement of temperature. The instruments are laser devices which monitor the scatter and transmission of a laser beam over a distance of around 450 mm. and at this site both weather . The Åre site is located in the middle of Sweden at a height of 1300 m above see level. See Annex 17 for detailed explanations. They are particulate volume monitors and have a 0 – 10 VDC output which give a measurement range 0. See Annex 16 for detailed explanations. See Annex 14 for detailed explanations 9. Experiences from continuous long term measurements show. It consists of a body with sensors and of a measuring rod fixed vertically downwards.5 Gerber The UK Meteorological Office sourced two instruments (Gerbers) capable of measuring the LWC of the air. 9. and it has been designed according to the recommendations in ISO 12494. One PVM-100 was mounted and aligned north/south and the other aligned east/west as shown in Annex 16. Two Gerber PVM-100 instruments were installed at the EA Technology severe weather test site at Deadwater Fell test site on the English/Scottish border in the UK. and it can rotate freely to allow ice build-up to form cylindrically.the first in 2003 close to Drammen (west of Oslo) and the second in the far north (in Finmark 2005) . The instrument has been installed in two sites in Norway .and in two sites in Sweden 2005.7 IceMonitor The IceMonitor is an instrument aimed for automatic weighing ice deposit on a vertical steel pipe. Figure A. 9. 9. The ice detector outputs include ice detection indication and fault status indication.cools within a few seconds and is ready to sense ice formation again. It measures the weight of ice accumulated on a vertical pole by the use of an electro-mechanical scale system. See Annex 15 for detailed explanations. The steel pipe has 30 mm diameter. The comparison of EAG 200 results with those of manually operated poles show the reliability of EAG200‘s data. ice mass and the velocity and direction of wind The measuring device METEO is made of stainless steel and it has no moving parts requiring frequent maintenance.5 m long.2).air temperature. relative humidity and wind speed/direction - 42/110 .002 – 10 g/m3. it is 0. that the system operates well even for short icing periods and for small amounts of ice accretion.

9. V.2. Its long-term use has shown it to be quite durable and it seems to detect icing in reasonable agreement with humidity indications and video recordings. 43/110 .8 T20-series Ice Detectors All T20-series ice detectors are based on a patented digital optronic ice-indicator that indicates the presence of any type of atmospheric ice including clear ice. Sweden See Annex 19 for detailed explanations.9 Instrumar IM101 Instrumar IM101. 9. Manufacturer: HoloOptics. The T20-series indicators: • come in single-direction (±45° upwind) or omni-direction versions • indicating either all icing types including clear ice or rime ice/wet snow growth only. An indication of ice is made if more than 95 % of the probe is covered with a 50 μm thick layer of clear ice or a 90 μm thick layer of other types of ice. It comprises a head with an IR emitter and a photo detector and a probe. Sweden See Annex 18 for detailed explanations.4 is an ice detector based on measuring the electrical impedance and surface temperature [32].and ice load data is measured. Testing of T20-series sensors have been carried out at locations noted in Annex 19. Manufacturer: SAAB Technologies. Measuring data from the Swedish site in Åre includes a number of heavy icing events during winter 2005-2006.

1 Regional variability As icing conditions and icing climate vary significantly within Europe it is important to perform the measurements at different parts of Europe. it is typical that the duration of ice loads upon structures in class S5 is longer than duration of meteorological icing events. At cold climate sites or at sites with temperatures below 0 ºC. long-term representativity and climate change Others 44/110 . There are however two points of view which will have to be combined in order to fulfil the requirements of the different communities. On the basis on the previous work on the classification of icing climates made mainly for wind energy purposes and the recordings made within the EUMETNET project the classification shown in § 5. For example.10 Long term recommendations for ice measurements in Europe 10.1 Specifications of users There are a number of potential users either for ice detectors or for icing data in different fields.2. meteorological instruments will respond differently depending on the location of their installation. meteorological sensors are subject to different types.2 Requirements for measuring sites Consequently. 10. durations and formation of icing. As no measurement data on atmospheric icing are available at the European Meteorological Services only a rough estimation of classification on severity of icing can be implemented at this stage.6 was proposed. noting the different climatic aspects: Northern European mountains with long icing periods under wide temperature and humidity range and lack of solar radiation (typically rime ice) Alpine regions with icing strongly depending on the altitude (typically rime ice and wet snow) Central and Southern European mountainous areas with icing and strong sunshine periods causing numerous melting and freezing consecutive events (typically glaze and rime ice) Maritime regions in Western Europe (typically wet snow) As discussed above. a number of test centres should be established and operated in Europe (and in the world). Ice measuring centres will have to fulfil the requirements regarding: Modelling Electrical power lines Wind turbines Towers and masts Road/railway safety networks Cable cars and others equipments for tourism purposes Airport safety Weather forecasters: development and control procedures for models Climatologists: maps. 10.

g. Finland. a permanent forum for monitoring icing in Europe needs to be established.g. Luosto. Appropriate task specifications will be handled within Phase 2 of the present COST-727 Action in relationship with the establishment of the test centres. the goal of establishing icing test centres is to provide an adequate infrastructure to fulfil the requirements of the above-mentioned potential users.3 Permanent forum for monitoring icing in Europe In relationship with the establishment of icing test centres. In a general way.2. etc. France.) 2-3 „reference“ instruments common to all sites: standardized testing and certification procedures and standardized data format Flexible infrastructures for the installation of different test beds (e. which is presently not the case. The proposed way to achieve this goal is to establish projects on the European level.10. UK. Czech Republic. the following aspects will have to be dealt with: 3-6 test sites in Europe covering the different climatic environments (e. for wind turbine and power line testing) Common monitoring and quality control procedures (for future certification) Complete high quality data sets for forecasting and climatology 10. Spain. There are two complimentary activities which will have to be further analyzed: Launching of a new EUMETNET and/or EU project for the establishment of longterm icing test and observation sites in Europe Integration of icing measurements in meteorological networks under the umbrella of WMO/CIMO 45/110 . Mt Aigoual. Switzerland. instrumentation at all weather stations located in cold climates should be extended with icing recording systems. Guetsch. Studnice.2 Specification of purpose Ideally. Therefore. Germany.

which studied the effect of icing on meteorological instruments and produced specifications for ice-free sensors. Since then icing measurements and meteorological measurements with ice-free sensors have been made at the site.13 24. 46/110 . The cameras also give information on visibility (fog and clouds). 1 Ice detector data of Technical Research Centre of Finland. it was noted that the sensor is inadequate for icing measurements in extreme conditions. Measurements with this sensor were not continued. Ten-minute averages. The platform for the instruments at Luosto was set up in the winter 2000/01.15 Ice detector Labko Ice detector 3200 Labko Ice detector 3500 Labko Ice detector 3500 Labko Ice detector 3500 Instrumar IM101 Start 12/1999 9/1999 2/1999 11/1998 10/1991 End 10/1999 7/2000 Parameters Status (icing/no icing) Status (icing/no icing) Status (icing/no icing) Status (icing/no icing) Status (icing rate) VTT has within the COST 727 Action also studied humidity measurements in icing conditions. Four video cameras are monitoring operation of the instruments. These studies [33] showed that The conventional humidity instruments measure incorrectly in icing conditions A heated Vaisala HMP243 (and HMT337) measures correctly in icing conditions It seems to be possible to detect rime icing by correct humidity measurements FMI test station: Luosto The Luosto test station is located in northern Finland. Amplitude levels of wire wave signals (see §6) have also been recorded because all the performed icing measurements have been closely connected to development of the used ice detectors.30 21. on the top of Luosto fell (500 m asl.55 61. The Labko LID-3503 and the Rosemount 0872J were installed at the site in the winter 2001/02. The test station was established for the EUMETNET SWSII project. Performance of the instruments is monitored continuously with video cameras. min. Data consists of status parameter. The station is operated by Finnish Meteorological Institute (FMI). E 26 54’). which tells whether ice was accumulating or not. Table A.01 67. The icing climate of the site has been characterized as site class A (see §5): an elevated site inland in northern Europe with harsh and frequent icing climate. Site Olostunturi Pori Lammasoaivi Pyhätunturi Ylläs LAT 67.48 21.47 67. 1. max and standard deviations were calculated and stored. Experience from the first winter was used to develop the measurement systems and the infrastructure. Digital pictures are sent every 10 minutes to the FMI via the network.37 68.4 LON 23. N 67 08’.20 27. Both ice detectors indicate the presence of icing conditions. The development of ice detectors has been driven by the need for development of blade heating systems for wind turbines. Data was collected at 1Hz.Annex 1 Measurements in Finland Technical Research Centre of Finland Ice detector data of Technical Research Centre of Finland (VTT) is presented in Table A. According to the measurements performed with the LID-3503 ice detector in winter 2001/2002.

The cylinder with a diameter of 30 mm and length of 0.10.10.10. Table A. Academy of Science of the Czech Republic.The Vaisala’s FD12P weather sensor is used for visibility observations.06 -> 01.4. 1 The test platform at Luosto with installed severe weather sensors.01 -> 01. Instrument LID-3503 ice detector Rosemount 0872J ice detector ICEMETER CZ .03. The FD12P measures meteorological optical range (MOR) from 10 m to 50 km.5 m is installed with the axis vertical and forced rotating around the axis. 2 List of instruments installed for icing measurements at the Luosto fell.01 -> 09. A device used for icing rate measurements is manufactured by Institute of Atmospheric Physics. Also temperature. Technical availability: the sensors are in operation and data (erroneous or correct) are received by the data acquisition system.01-30.01 -> 01.10. 47/110 .01 -> Parameter Icing occurrence Icing occurrence Icing rate Icing rate Icing occurrence Visibility and rain Figure A. dew point temperature and relative humidity measurements are analysed to get more accurate data on events and duration of icing.02 01. The ICEcylinder has been built by FMI according to the International Standard ISO 12494. Accumulated ice load is weigh manually.IAP ICEcylinder 2-4 video cameras FD12P weather sensor Technical data availability 01.10. The ICEMETER measures the mass of ice accumulated on the sensor surface.

) and Technical Research Centre of Finland (VTT) made measurements of icing on tower structures. 48/110 . The results of these activities have been reported in a number of publications [19. measurements of drop size and liquid water content of clouds and comparisons of meteorological instruments on hilltops in severe icing conditions in Finland in a four year national project in 1986-1990 and later.35.. tests on ice detectors. In this project both an operating 128 m tall TV tower and a 7.33]. so that the ice load on them could be continuously measured.5 m test tower were equipped with load cells.34. Distribution Dept.18.DIGITA / VTT CAMPAIGN Digita Oy (former Finnish Broadcasting Co.21.20. The rotating multi-cylinder method was developed to a stage that would allow further development to automatic measurements of drop size and liquid water content during icing.

In the event of ice accumulation all the measurements of the standard procedure (ice mass. up to 35 locations were operated simultaneously. icing type(s)) was compiled. The number of locations has changed slightly during the years. Blue circles: Icing cycle measurements. Figure A. A standard measurement procedure defined an exposition period of 24 hours: The pole was exposed every day at 8:30 a.g. The icing mass was determined and additional information (e.g. depicted in topographic maps for different time periods and resolutions of measurements. 2) so that a second icing pole was exposed at those locations. Afterwards the pole was re-exposed for another 24 hour period. Therefore the measurement procedure was modified at 11 locations in 1978 (blue circles in the left part of Figure A. However.Present.Annex 2 Measurements in Germany German Meteorological Service (DWD) Icing measurements were carried out at altogether 40 locations in the eastern part of Germany during 1965-1990 [36. Red points: 24-hour measurements. By using the standard measurement procedure the ice accretion was interrupted at least every 24 hours. 2)..l. The measurements were performed at all locations by the use of manually operated vertical icing poles with a diameter of 35 mm. Continuous icing measurements at all locations. ice vane dimension. By the application of this procedure it was possible to improve the 49/110 . see red points in the left part of Figure A. For ice accretion periods exceeding 24 hours or on occasions of multiple icing events the standard procedure underestimated the maximum ice masses. at a height of 2 m a.m. 2 Locations of Icing Measurements in Germany. In the case of ice accumulation the pole was exchanged after 24 hours. Time Period: 1990 (partly) . Arkona Falkenberg Kahler Asten Chemnitz Zinnwald Hohenpeißenberg Time Period: 1965-1990. icing diameter and direction. additional information) were carried out after 24 hours.37].

The icing sensors are implemented at 5 locations (for details see the right part of Figure A. Table A. 3. 50 m and 90 m above ground) at DWD’s Meteorological Observatory Lindenberg. The number of locations with icing measurements has been reduced to a total of five since 1991. 50. intermittent. manual EAG 200 (Annex 15). so they were called icing cycle measurements. Furthermore. 3 Overview of present locations with icing measurements in Germany Height Location above sea level (m) Arkona 42 Chemnitz 418 Zinnwald 877 Kahler Asten 839 Hohenpeißenberg 977 Lindenberg 73 Measurement height (m) 2. accumulation and Icing pole. 5 2. accumulation 50/110 . They are operated continuously at a single height level. 5 10 10 5. One of the main results of icing cycle measurements was ‘real’ (in distinction to the standard procedure) maximum ice mass. 2 and Table A. at 3 locations the icing measurements are still carried out by the use of manually operated icing poles. 5 2. continuous.knowledge about whole icing cycles (accumulation and loss of ice). continuous. The latest advancement during the ongoing winter period are icing measurements at three heights (10 m. 90 Measurement device EAG 200 (Annex 15).

term Climat. 1957 Dec. term Climat. The sites where the atmospheric icing measurements are made are fewer and are given in Table A. This Institute is an administrator of the meteorological networks in Slovak territory. There are a large number of regular weather stations where atmospheric icing observations are carried out. term 2634 m asl Climat. term Since Dec. term Climat. 1972 Jan. where the parameters atmospheric icing are measured. term 177 m asl 178 m asl 206 m asl Climat. 14. Data are acquired 3 times a day at 7. 1971 Jan. term Climat. term Climat. term Climat. 1973 Dec. [38] Table A. term 187 m asl 145 m asl 703 m asl 312 m asl 209 m asl 901 m asl Climat. 1971 Jan. 1971 Jan. 1974 Dec. There are both visual observations and measurements of the amount of atmospheric icing.Annex 3 Measurements in Slovak Republic Extensive experimental data from the measurements of atmospheric icing can be obtained from the Slovak Hydrometeorological Institute (SHMI). 1972 2004 m asl Climat. 1957 Dec. term 51/110 . 1971 Jan. The diameter of the rods is 32 mm and length is 1 m. term Climat. 1973 Jan. term Climat. Visual observations of atmospheric icing enable an analysis of the lengths and annual occurrences of atmospheric icing events to be made in individual seasons. 4. 1972 Dec. 1976 Dec. Measurements are performed on a horizontal pair of orthogonal (N-S and E-W) wooden rods. 21 hrs. 4 List of sites in the Slovak territory. Location BratislavaKoliba Hurbanovo Chopok Jaslovské Bohunice Kamenica nad Cirochou Košice Lomnický štít Lu enec Nitra Poprad Slia Stropkov Telgárt Latitude 48o 10’ N 47o 52’ N 48 56’ N 48o 30’ N 48o 56’ N 48o 42’ N 49o 12’ N 48o 20’ N 48 19’ N 49 04’ N 48o 38’ N 49 12’ N 48 51’ N o o o o o Longitude 17o 06’ E 18o 12’ E 19 35’ E 17o 50’ E 22o 00’ E 21o 16’ E 20o 13’ E 19o 40’ E 18 05’ E 20 15’ E 19o 09’ E 21 39’ E 20 11’ E o o o o o Altitude 289 m asl 115 m asl Recording Climat.

4 420 kV line crossing the mountain divide (1 100 m asl) between western and eastern Norway (Photo: S. exposure to air of maritime origin and local topography [39]. As the bulk production is on the western coast of the country.95 m and weighed 305 kg/m. as indicated in Figure A.4 m x 0.M. there are many high voltage lines crossing the mountains and thereby also exposed to severe ice loads. W E Rain and icing Clouds 0O C North Atlantic Ocean Central mountain range Figure A. and the main consumption is in the south-eastern region around Oslo.Annex 4 Measurements in Norway Icing on overhead lines varies strongly with height above sea level. This is probably the clearest example on how topography and exposure 52/110 . Figure A. 3 A cross section of southern Norway Figure A. 4. where the precipitation rates are highest. In Norway 100% of the electricity production is from hydropower. Therefore lines of 100 – 200 km length have a variety of local exposures in mountainous terrains. 3 and Figure A. 5 shows an example of the accretion that measured 1. Fikke) Probably the world’s largest ice load on an electric power line was observed in Norway in 1961.

Statnett. Figure A. A mountain (ridge) only 50 – 100 m higher in the upwind direction may be sufficient to reduce this icing to a minimum. As a result of this variability all transmission lines belonging to the main grid of Norway are designed span by span according to the expected ice and wind loads for the particular span. 53/110 . For lower voltage lines in the distribution system the loads are given regionally according to the National code. but the cloud base varies significantly with topography. It was built on the top of a mountain ridge more or less parallel to the coast. Wet snow however may occur in all altitudes and also on the leeward side of mountains and ridges. Since the late 1970s the Norwegian Power Grid Company. In-could icing occur only above cloud base. (Photo: O. Wist) The topography influences the icing differently depending icing type. Only when a new line was built up from the leeward side of the mountains the power supply to this radio and TV transmitter became stable. and therefore maximum exposed to the humid south-westerly winds from the Atlantic Ocean. Figure A. This line was feeding a radio and TV transmitter 1 412 m above sea level. The great dependence on topography may result in a likewise great variety of expected ice loads along a line passing through zones with different exposure and climatic conditions from one sub-station to another. It appeared to be impossible to maintain this line no matter how short the spans would be. 7 in remote and exposed locations in mountainous terrain in Norway. has operated more than 20 ice racks of the type shown in Figure A. how strong the (wood) poles were made and how strong the steel conductors were. 9 shows an example on extreme value calculations. Freezing rain occurs mainly in basins and depressions where cold air can be trapped while warm air with precipitation may intrude the air aloft (temperature inversion).influence the icing conditions on a power line. 5 In-cloud icing on a 22 kV line feeding a TV tower 1 400 m above sea level in south-western Norway. 6 and Figure A. 8 shows the distribution of the locations of the measuring stations and Figure A.

9 Example of extreme value calcusites in Norway lations of ice loadings from the ice racks 54/110 .Figure A. 8 Location of the 16 ice measuring Figure A. 7 Details of dynamometer suspension in the rack 70 N Tromsø 1 68 N 2 3 4 Bodø 66 N 5 6 64 N Trondheim 62 N 7 9 8 15 km 100 200 300 400 60 N 17 10 11 16 Oslo 13 14 Stavanger 12 58 N Figure A. 6 Norwegian rack for measuring ice loadings in remote areas Figure A.

10 Test site at Studnice (Photo: J. 5 Location of racks at Studnice Location Latitude Studnice 49˚36’27” N Longitude 16˚05’7” E Altitude 800 m asl Figure A.Annex 5 Measurements in the Czech Republic EGU Brno carries out measuring of icing on racks at test site Studnice thus continuing the long time-series of icing data [40]. There are 2 spans (ca 250 m each). Sabata) Various measurements are performed here: Ice measurement on conductors and measuring rods of various diameters The measurements on samples of stranded isolated conductors 55/110 . Table A. The whole site was built in the late 70’s and has been in operation since 1980. It is located about 60 km northwest from Brno at 800 m above see level.

20.12. 10. For measurement of ice accumulation we use different instruments with horizontally and vertically oriented rods (with diameter of 30 mm). 6 Location of measurement devices on racks at Studnice Measuring device Measurement height (m) Output Since Vertical rods (length of 0.Sabata) 56/110 . freely rotating rod with sensor.5 m) 5. 30. at test site Studnice.Observation of icing with respect to altitude gradient. Table A. 20. 50 Digital Digital Digital Digital 1997 1997 1997 2001 Horizontal rods (length of 1 m) 10. 30. 17. 30 Temperature 5. 40. 10.5 m) 10 Figure A. 11 Vertical.2004 (Photo: J. 50 Meteo device (length of 0. 40.

T. WS. T. west and southwest. WS. which exceeds the surrounding terrain by approximately 300 m. T. WS. It is probably most exposed region in the whole country. which is the highest peak of the tertiary volcanic range of “ eské St edoho í”. T. the long time average is 5. WS. T. one Icemeter has been situated close to the wind turbine in Krušné hory (Ore mountains). WD IM. Concerning temperature. WD IM. WD Since 1999 2003 2001 2001 2001 2001 2001 2001 2001 2001 2001 2001 2001 49˚24’ N 15˚39’ E 49˚06’ N 17˚18’ E 49˚06’ N 16˚25’ E 48˚53’ N 16˚00’ E 49˚10’ N 15˚55’ E 49˚08’ N 18˚02’ E 49˚36’ N 16˚20’ E 49˚08’ N 18˚02’ E 49˚28’ N 16˚50’ E 49˚12’ N 15˚40’ E 49˚19’ N 16˚07’ E 554 m asl 477 m asl 441 m asl 379 m asl 570 m asl 634 m asl 722 m asl 634 m asl 685 m asl 640 m asl 580 m asl Digital Digital Digital Digital Digital Digital Digital Digital Digital Digital Digital Vyskovec 48˚56’ N 17˚49’ E 761 m asl Digital IM. WD IM. WD (wind direction). WD IM. Table A. Milesovka has a shape of isolated forested cone. T. WD IM. WS. The Meteo device locations on the territory of JME cover the whole area susceptible to regular icing. WD IM. WS. The steepness of slopes ranges from 20° to 30°. WD IM. WS. 12). Krušné hory is the region most suitable for wind energy production in the Czech Republic. Two devices are located in the area of the regional utility VCE and the others in the area of regional utility JME. T. T. T (temperature). Unfortunately that region is also affected by severe icing. T. absolute minimum was -28. WD IM. WD 2001 *) Measured values: IM (Ice mass). was built on Milesovka’s summit (837m asl). WS. WD IM. A synoptic meteorological station. 7 Location of Meteo devices Location Cotkytle Novy Hradek Beranov Buchlov Hlína Kasarna Klucov Kralovec Nyklovice Kralovec Protivanov Predin Ruda Latitude Longitude Altitude Output Digital Digital Parameters* IM. WD IM. situated at different height above the ground (see Figure A. WS. WS. WS. WS. near to the village of Nová Ves. concerning the severe ice episodes. WD IM. T. Nová Ves: Since 2005.3 °C and absolute maximum reached 34.7 °C. Icemeter sites (for results see section 6) Milesovka: Most of the measurements were carried out at the top of “Milešovka” mountain.1 °C. now belonging to our Institute. WS.In addition to measurement at Studnice station fourteen Meteo devices are installed at locations supplied by two regional utilities. T. From the winter 2003/2004 two Icemeters. T. The mean wind speed is 7. WS. have been operated. The average of annual precipitation is 564 mm. T. 57/110 .7 m/s with most frequent winds from northwest. T. WD IM. WS (wind speed).

Sternstein). The Icemeter was also located temporarily in other sites in the Czech Republic.Figure A. both of them situated close to wind power station. Due to the short time of location at these sites. 58/110 . One piece was tested in Luosto. 12 left – Observatory at the Milesovka peak with marked position of Icemeters. results are not available. right – location of Milesovka mountain and Nova Ves in the Czech Republic Other sites: Two Icemeters have been in operation in Austria (Obersthralbach. Finland during the winter 2004/2005.

The site has the capacity to provide full meteorological measurements on-site. The site consists of a 200m test span with terminal H-poles supported by 14 stay wires each. The sites are situated in exposed locations at heights between 550m and 750m and range from hilltop to open moorland. hard and soft rime ice and glaze icing (freezing rain). These two sites have since been closed down and all work is now concentrated on the Deadwater Fell site on the English/Scottish border. Load cells are used to measure loads on the conductors. Provision is made for shorter span lengths when appropriate. Figure A.Annex 6 Deadwater Fell Measurements in UK Deadwater Fell is one of EA Technology’s field stations at sites located to encompass a full range of weather conditions: wet and dry snow. Figure A. The data presented here comes from this and two other sites located on Susseter Hill in Shetland (an island off the north coast of Scotland) and south west Scotland (Green Lowther). 14 shows the rotating rig. 13 shows the full 200 m span viewed from below the summit ridge and Figure A. The test spans are orientated North-South and suffer from severe winds as well as ice incidents and blizzards. Overhead line conductors are erected in test spans of between 90m and 200m. 13 The full 200m span at Deadwater Fell 59/110 .

Item a) is known and items b) and c) are measured together.1 lux. rotation or general conductor movement under wind and ice loads.14). if required. 60/110 . the main concern with overhead lines is the total wind and ice load. However. to withstand blizzard conditions. near Chester. for analysis at EA Technology at Capenhurst. UK. The surveillance also allows short periods of real time coverage every 30 minutes. In the test spans. Test Spans The H-poles are designed to withstand impulsive forces from the galloping of large conductors and at the same time. This force is made up of: a) Conductor weight b) Ice load c) Wind-on-ice load. each conductor is fitted with a load cell and. The cameras are mounted within specially adapted housings with insulation. a vibration monitor. Video surveillance can detect any galloping. Provision can be made for close up video coverage to identify ice shape. The spans are monitored 24 hours a day throughout the year by time lapse video cameras with low light level sensitivities down to 0. The site also has a rotating rig to test conductor samples (Figure A. It is downloaded automatically every 24 hours via a mobile telephone. if required. This rig is designed so that the samples are always facing normal to the prevailing wind. Intermediate poles are installed as required for shorter spans. barely visible. The latter can be calibrated in above 0°C incidents with no ice present. internal heating and externally wound heating tape to reduce ice growth and are used to give close-up and long distance views using. environmentally friendly infra red floodlighting. Conductors of different sizes and of different span lengths can also be installed to provide direct measurement of the total force on the conductors at their connection point to the supports.All the data is collected and stored at the site. This system can be used to evaluate the ice load by calculation of the wind load. so from a practical point of view it may not be necessary to separate these components.

14 The rotating rig with snow accretion.Figure A.0003 and D = Conductor Diameter (mm) 61/110 . The measured ice loads measured were between 1.8 kg/m.794 C = -0. The initial wet snow accretion rate (Model 1) is: Mass = [ A + B.6 and 8. A series of load comparisons between Upas (diameter 24. 10-3 kg/m where A = 0. Ice load density was measured on occasions when the normally un-manned site was visited. measure the total wind and ice load.449 B = 1. The accretion density for rime ice gave an average value of 570 kg/m³ and for wet snow 825 kg/m³. The load cells will.D + C. calculated the rime ice and wet snow loads for a range of conductors. Conductor Data Ice densities Data from the test spans at Deadwater Fell was investigated for direct comparison of two bare conductors at a span length of 200 m. Ice loads Shackleton et al [42].D² ] . under an EATL contract. of course.5 mm) conductors was extracted at various ice loads. This is in line with field data published for Iceland [41].7 mm) and Lynx (19. but at wind speeds around 20 knots.

In order to eliminate the different field conditions. 16 that assuming a constant accretion thickness for any conductor size (as used in current UK wind/ice maps in ET111) gives a reasonable fit for conductors up to Oak (14 mm diameter AAAC). 32 mm² Hard Drawn Copper (8.9mm diameter).D + 0.0034.792.D² ] .The longer time version is called the Large Relaxation Time (LRT) Model. B and C vary with time with A reducing and C increasing. all the data is related back to the original ‘control’ conductor.15 with a trend line polynomial in which: where A = 0. Snow Loads against conductor diameter 6 y = 0.0013x + 0. Hazel was used extensively at Green Lowther to establish a relationship between this and the 32 mm² Copper conductor and so maintain the validity of the data from each site. It can be seen in Figure A.0971x + 0. The EATL field data is given in Figure A. 15 Field data with polynomial trend line Sufficient information has been extracted from the database to give comparative values for 28 conductors. At Deadwater. the control conductor was changed to the Hazel conductor (9. which gives ice loads according to: Mass = [0.0mm diameter). This was used as the ‘control’ at the Green Lowther and Shetland sites. For larger conductors up to 40 mm in diameter.013 B = 0.1409 5 2 Snow load (normalised) 4 3 2 1 0 0 5 10 15 20 25 30 35 40 Conductor diameter (mm) Figure A. the constant thickness assumption underestimates loads as compared with the EATL field data. 10-3 kg/m The factors A.097 C = 0.1409 The formulae imply that over time the conductor diameter has an increasing effect on the modeled snow load.243 + 1. which is a bare AAAC 60 mm² conductor. 62/110 .

16 Snow load field data plotted against constant accretion thickness scenario for the same conductors.Field data against contant accretion thickness (ET111) 6 5 Snow load (normalised) Field data 4 3 Constant accretion thickness (ET111) 2 1 0 0 5 10 15 20 25 30 35 40 Conductor diameter (mm) Figure A. 17 Typical accreted ice loads on different conductors after a blizzard at Deadwater Fell 63/110 . Figure A.

At the site in Åre. Data is retrieved to a server in Östersund via radio/radiolink or in some case via telephone modem. A camera is monitoring the IceMonitor sensor at the site in Åre. Icing of this camera has been a problem during the winter 2005/2006 – and the installation and de-icing of the camera will be modified before the next icing season. (see figure below. such as air temperature. The monitoring equipment was originally designed for use in road weather information systems and it is capable of handling a number of different sensors. wind speed and direction. and in some cases also precipitation. 64/110 . HoloOptics and Jokkmokk) are currently being tested under field conditions in Sweden and Norway. several icing events occurred during the winter 2005 /2006. The IceMonitor sensor is presently connected to monitoring stations located in Åre (SE) Ritsem (SE) Drammen (NO) At these stations measurements of weather parameters. are being carried out.) During the same period. including camera – which stores files in jpg format. Figure A.Annex 7 Measurements in Sweden Three different sensors (IceMonitor. A couple of the icing events have been severe with ice loads of more than 40 kg. 18 Icing measured at Åre during winter of 2005/2006 (Ice load signal error occurred by end of Feb. relative humidity. no significant icing occurred at the site in Ritsem (Sweden) or at the test site in Drammen (Norway). Measurements are normally made once every half hour (in Drammen once every 10 minutes).

Sweden. 150 km north of the polar circle. see figure A. 19 Icing rate measured at Suorva. ski lift protection. Northern Sweden Icing Rate 1h average 510 515 520 65/110 . Sweden. Sweden Bromma Airport. KTH (The Royal Institute of Technology). Cooperation with Combitech Sourva Icing Rate icing rate g/m2h 200 150 100 50 0 495 500 505 hour Figure A.19. Stockholm. In co-operation with the airport authorities Åre Ski Resort.The HoloOptics T23 Clear-Ice Indicator and the T26 Icing-Rate Sensor beta versions have in periods been tested at the following locations during 2003-2006: The Suorva wind power plant. Stockholm. In co-operation with FOI (The Swedish Defence Research Agency).

We have also additional information from the National Electric Company with data about icing depositions in cases of damages of the power lines in the period 1962 –1990. The stations were chosen to cover the territory of the whole country. These stations are highlighted with yellow in the table bellow (some of them still report icing data). All these data have been used to characterize the typical icing conditions of the territory of the country. A list of all stations with initiated icing observations is presented in the table below. Data about the meteorological conditions during icing events and the final ice depositions have been collected. They represent the regions where icing occurs most often and usually is very severe. However. The ice measurement device was a couple of perpendicular conductors with diameter of 5mm located in the directions N-S and E-W. other stations have collected long time series of icing events or short but very detailed measurements. 20. The collected icing data in some of the stations are very short and/or not reliable due to the bad organization of observations. m 2000 1500 1000 500 0 10 20 30 40 50 60 70 num of cases ber Figure A.Annex 8 Measurements in Bulgaria The initial icing observations in Bulgaria were started in the late 50-s and in the beginning of the 60-s of the 20th century in some stations of the meteorological institute. Our intention now is to continue with icing measurements on the places with good historical data using further visual observations as well as some ISO12494 icing sensors (rotated and non rotated). As example of these conditions the change of the number of icing events in the mountain regions of Bulgaria is depicted on the figure. 3000 2500 altitude. non motivated work as well simply to the fact that icing had been observed rarely in these regions for the whole period of measurements. The change of the mean number of cases with the altitude in the mountain regions 66/110 .

53 43.43 43.27 28.55 42.50 43. Vitosha SW.17 27.40 23. Toshevo Dobrich 1244 1180 1599 1925 1750 136 235 2376 178 217 63 236 200 67/110 . Rila SW.10 23.44 26.01 27.40 23. Rhodopes SE S Central planina Central garia E NE NE NE Stara BulThe The The The 1969-1970 1992-1993 1955-1964 1955-1986 1968-1983 1980-1994 1959-1989 1968-1974 1960-1975 1968-1978 1976 1974-1999 1999-2006 1959-1969 1956-1975 1966-1994 1971 1957-1974 1969-2006 1956-1986 1987-2006 not good relatively good very good not good not good very good very good very good not good relatively good relatively good very good good very good very good not good relatively good relatively good not good very good BAC station. Rhodopes SW. m 34 31 29 257 1400 1678 851 Latitude 43. Stara planina W.17 23.05 28.35 Longitude 23. Rhodopes SW.25 42. Lozenska planina W. Stara planina NW.39 41.58 22.34 41.21 43. 1485 Vitosha Borovec Smoljan Pamporovo peak Snejanka peak Rozen Elhovo Krumovgrad peak Botev Chirpan Ljuljakovo Kaliakra Gen.35 25. Rhodopes SW.50 Start/End of Part of counData icing observatry. 8: Meteorological stations with icing observation Station Kozloduy Vidin Orjahovo Gramada Petrohan Peak Murgash Vakerel Altitude.12 24.43 42.14 23.50 42.41 24.41 24.39 43.11 41.47 43.28 42.41 42.55 25.44 22.40 41.15 41.Table A.51 23.39 24.59 43. mountain quality tion NW NW NW NW.36 24.53 42. Stara planina NW.18 42.

30 26.33 43.44 NE NE 1978-2006 1955-1979 very good very good 68/110 .Razgrad Gara Samuil 346 474 43.30 26.

46 17. The sites with the atmospheric icing measurements are given in Table A. For measurement of ice accumulation we use horizontally oriented wires (diameter 31 mm). hard rime.5 17.43 46.27 46. Table A.29 18. 9 Record of the sites in Hungary.26 47.37 47. 69/110 .15 47.12 Altitude Monitoring Output 156 HMS Manual 138 HMS Manual 108 HMS Manual 220 HMS Manual 116 HMS Manual 113 HDF Manual 1010 HMS Manual 233 HMS Manual 121 HMS Manual 139 HMS Manual 97 HMS Manual 146 HDF Manual 202 HMS Manual 108 HMS Manual 82 HMS Manual 89 HDF Manual The observers determine the types of accretion (soft rime. wet snow) and the duration.23 17. 9.07 Longitude 17.21 46 46. The number of stations were 34 (1970-1992). and also the accumulated ice thickness at 06 UTC every day.02 20.06 20. Icing measurements were carried out by Hungarian Meteorological Service (HMS) and Hungarian Defence Forces (HDF).01 20. icy-hard rime. The measuring instrument system was established by the Hungarian meteorologist Mihaly Csomor.29 47.52 48.44 20.13 18.53 46. but now we have only 16 (2005).36 20.54 47.16 16.54 47.41 47.41 19.Annex 9 Measurements in Hungary Visual observations of atmospheric icing enable from 1970 till nowadays.55 46.09 19.05 47. where the parameters of atmospheric icing are measured Location Bekescsaba Budapest/Lorinc Debrecen Eger Gyor Kecskemet Kekesteto Miskolc Mosonmagyarovar Nagykanizsa Paks Papa Pecs Siofok Szeged Szolnok Latitude 46.11 21. freezing rain. In addition the observers measure the thickness and water content of ice from the last 24 hours.58 18. frozen dew.

21 Hungarian instrument for manual icing measurement on samples of bare conductors.Figure A. 70/110 .

49]. The wires are oriented at South to North and West to East and are at 2 m height from the ground. Another proposed ice collector includes a vane which adjusts the wires so that they are perpendicular to the wind direction [51]. By 1970's the number of weather stations in the European part of the USSR that made regular manual icing observations was 220 [45]. This device includes wires of both 5 mm and 10 mm diameter which are free to rotate around their axes. considerable attention was paid in the USSR to determine the relationship between the ice load measured on the ice collector device and that on a real overhead power line conductor [47]. torsional rigidity and height above terrain [48. Because of that. a measurement stand better suitable for the purpose of power line icing has also been developed and tested at many mountainous sites [50]. The measurement is done manually by weighing on a daily basis. In 1984 measurement based estimates of the design ice loads were reported to have been made for 700 weather stations in the USSR [46]. The ice collector device used in the Russian measurement network is a stand consisting of two rigidly clamped horizontal wires with a diameter of 5 mm. icing observations were included as part of routine meteorological observations beginning 1931 [43. special measurement campaigns have been run at selected sites using wires with different diameter. The current situation in Russia is unknown. To that end. The main application of the Russian data has been in mapping the design ice loads on power lines.Annex 10 RUSSIA During the Soviet era. 71/110 .44]. Based on these studies.

It includes four vertical flat faces. every three hours during freezing precipitation. The ice collector used in the Hydro Québec observation network is the Passive Ice Meter (PIM). the measured ambient temperature and precipitation.5 cm [52]. Initially there were 35 observation sites and later observations have been made at 180 sites altogether. 53. 22 The Rosemount freezing rain sensor mounted at an AWOS site 72/110 . Hydro have also made icing tests. The automatic weather observing system AWOS/ASOS (www. At these sites this is called “Freezing rain sensor”. but only at some sites. Ontario Hydro. The AWOS/ASOS sites include an automated detection of freezing precipitation by the Rosemount 872/C3 ice detector (Figure A.weather.C. Other Canadian power utilities. e.Annex 11 CANADA AND USA Hydro Québec has run a measurement network for glaze ice caused by freezing rain in the St.gov/asos) covers almost 1000 sites in North America. The sites are mostly airports and the data are aimed at operational use. The observations are made manually twice a day or. The system reports on freezing precipitation based on an algorithm which takes into account the change in the resonant frequency of the Rosemount probe. 54]. 22). Mount Washington Laboratory in USA continuously observes icing and was the first to make rotating multicylinder measurements. Figure A. The present observation network includes 150 stations and has a grid dimension of about 50 km. Lawrence Valley since 1974 [52. at synoptic weather stations.g. a 25 cm x 25 cm horizontal surface and two groups of horizontally oriented fixed cylinders with diameters of 1 and 2. Newfoundland and Labrador Hydro and B.

10. The cylinder is orientated vertically in order to eliminate the detection of wet snow as much as possible. The mass of accumulated ice is measured by means of a tensometric bridge (strain gauge load sensor). wind power station.05 m2 4.25 kg RS232 (optionally RS485. the vertical orientation of the collector can be changed to a horizontal one. which assigns the time and stores the data into the device memory. 10 Basic technical specification of Icemeter Measuring range Resolution Accuracy Surface area of the sensor Mass of the device Interface Capacity of the memory Voltage Supply Electricity consumption Operating temperature range 0. The digital signal is preprocessed by a micro-controller. Prague. RS422) ~7500 data points incl. and traffic roads. the “Icemeter” can find its application also in the monitoring system of power lines. Czech Republic measures the mass of icing accumulated on the surface of its collector.. date and time 12-15 V ~50 … 230 mA depending on heating -30°C …50 °C A description of the sites is given in §6. The first prototype was described in [55]. which couples the cylindrical collector to the tensometer. the Real Time Clock circuit is powered from the backup rechargeable battery. Results of the measurements with details of the experience with the Icemeter operation are given below: 73/110 . In the case of power supply failure. The passage through the housing may be heated depending on the passage temperature. it was considered for the investigation of the most favourable meteorological conditions for in cloud icing growth and the investigation of accumulation of in-cloud icing and its chemical analysis..Annex 12 Icemeter (Czech Republic) Description of the instrument The icing sensor (or “Icemeter”) developed in our Institute of Atmospheric Physics. if required. Primarily.10 kg 1g ±5 g 0. Table A. A test electromechanical impulse is applied each hour to verify the free force transition to the tensometer. so the information is not lost. The basic technical specification of the instrument is summarized in Table A. Nevertheless. which is located together with the electronics in the housing. the output of which is tied to the precise AD converter. and thus to check whether the acquired data are reliable or not. In order to prevent the freezing of the horizontal rod. together with a short review of previous methods of ice measurements that were applied in the Czech Republic. The „Icemeter” can operate autonomously as data logger (it has sufficient memory for approximately 50 days of operation with 10 min sampling intervals) or can pass the data on request to the PC.

Here.3.26 kg/m2. whereas in Nova Ves it lasted more than one week up to 1. Figure A. 23 The severe icing recorded at Milesovka peak in the end of 2002 Figure A. 25). The shape of the curve – record is similar. 23). thus the accumulated icing was about 1. 24 presents the example of measurements from different heights above the ground at the Milesovka observatory. At that time several masts were broken due to heavy icing in the Czech Republic.2005 to 18. the icing lasted quite longer. We can see that significantly higher mass of icing accumulates at the higher position. Note that the records are not from the same time period. when more than 2. At Milesovka the icing lasted just about two days from the evening 16. They overlap only partially.2005 (see Figure A.Examples of measurement The Icemeter has been operated on the Milesovka peak from 2000. but not exactly the same during the presented period (January/February 2005).2. Interesting is the comparison with the measurement in Nova Ves. The maximum icing load was detected in the end of 2002. which is situated ~30km west from Milesovka at about the same altitude. The accumulated mass was however comparable. 74/110 .5kg accumulated on the collector of length 48 cm with surface area 0.2005.05 m2 (see Figure A.

75/110 . 25 Example of recently (February 2005) installed icing measurement in Nova Ves. Figure A. 24 Example of icing measurement at different heights on Milesovka peak.Figure A.

Discussion. we have maintained the heating power as low as possible to be sure that the heat doesn’t prevent the ice growth. there were also time periods. We had preliminary talks about the possible placement of the Icemeter in the test site in Switzerland. when the instruments gave obviously wrong negative values. 76/110 . We could identify that from the fact that we didn’t see the proper electromechanical pulse (see section 2) in the technological data. wind and humidity records haven’t shown sufficiently good results. Currently. This freezing problem has been very rare on Milesovka. For example. These periods of freezing – malfunctioning can be seen in Figure A. where we have got most of our experience. Note that the location in Nova Ves has more severe cold and windy conditions than Milesovka. Regarding the investigation of conditions favourable for icing growth in the Czech Republic. the heating power is only ~ 2 W maximum. Probably we will mainly continue in focusing on sensors that measure accumulated icing. provided there is no limitation to power consumption.Problems of operation Although the Icemeters installed in the Czech Republic measured correctly most of the time. We believe that such measurements could improve our knowledge of icing. improvements are needed to the wind shielding of the passage of the coupling rod through the housing. we will make the comparison of icing measurement with the measurement of liquid water content on Milešovka peak. These times can usually be associated with the periods when the horizontal rod coupling the tensometer with the vertical collector became icebound to the instrument housing. We consider it should be no problem. future plans Although we have reasonably good data from most of the time of operation in the Czech Republic. Additionally. we will try to follow future recommendation for ice sensors that may occur. Such comparison should be possible at Milesovka beginning from the winter 2005/2006. Thus there was no free force transition. 25 as data gaps. The recent attempts to simulate icing measurements by using temperature. conclusions. Except for the enhancement of available heating power. it seems that we should enhance the available heating power in order that the Icemeter can operate in cold climate conditions. and in a site in Bulgaria. The instrument gave the wrong negative values. we will consider the possibility to build an instrument with a rotating collector. Since we haven’t had many problems with freezing on Milesovka.

the cylinder is heated to make the sensor wire ice free again. 700 600 500 Power [kW] 400 300 Measured Power 200 100 0 0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 8:00 9:00 10:00 Power curve Icing on video Heating / Ice detector 11:00 12:00 13:00 14:00 Figure A. The older model 3500. which was originally developed for meteorological purposes. One such case where an ice detector was not able to detect incipient icing is presented in Figure A. When ice accretion has been detected. Delayed ice detection is harmful because quick and reliable indication of the beginning of icing is especially important in wind turbine applications. This is due to the fact that ice detectors are used to control anti. Heating power. Operational experiences Mechanically the Labko 3200-ice detector that has been developed for wind turbine use has been good. 26. So. which is an ultra sonic sensor. The ice sensing wire. 77/110 .and de-icing systems as well as to control operation of turbines in populated environment where ice throw may pose a safety risk [31]. cut off temperature and ice alarm amplitude can be controlled. has been found to indicate ice more accurately compared to the LID 3200 series but it has also been more fragile to vibrations due to its physical shape [25]. 26 Wind turbine’s performance during icing event 22. is helically wound around the cylinder as near as. The sensor probe consists of a 350 mm long electrical heating element fastened to a protective cylindrical housing.2002 in northern Finland.Annex 13 Labko Ice Detector (Finland) Technical features of LID3210C The main functional hardware parts of the ice detection system are a sensor probe and a control unit.2. Performances of ice detectors have been monitored with video cameras in order to solve the noted problems with delayed ice detection. the ice alarm level and restart delays can be well set to correspond to changing conditions and purposes.

The performance of the wind turbine deteriorated due to delayed ice detection. Delayed ice detection also increases the heating power demand of anti-icing systems due to the increased heat transfer as a consequence of non-laminar (=turbulent) airflow over a rough iced blade surface. 78/110 .

79/110 . The Luosto test sites represent an elevated site inland with harsh and frequent icing climate. humidity. temperature. E 26 54’). dew point and wind sensors was evaluated. This sensor is fairly adequate for icing measurements and it operated better than the other tested instruments. it cannot guarantee accurate measurements in all icing conditions especially in soft icing conditions. The test station is located in northern Finland on the top of Luosto fell (500 m asl. The performance of the two ice detectors was monitored with two video cameras Also. ice accretion may exist on the sensor probe during short periods of time especially in the beginning of the icing event but the sensor does not detect ice. the two automatic instruments used for ice detection were not entirely reliable and differences were found between the performances of ice detectors. Measurements on icing events and duration of icing were carried out during three winter using Rosemount model 0872J (prototype). BFGoodrich. Nevertheless. at the FMI’s test site.Annex 14 Rosemount. Information and documented experimental data on meteorological conditions during icing events and performance of ice detectors was collected. The ice detectors are to some extent insensitive to icing under heavy icing conditions. It was possible to record more or less accurately the start and ending of icing periods but not the accretion rate or type of the icing. N 67 08’. The Rosemount sensor has yielded fairly good measurements at Luosto and detected the presence of icing conditions. data measured with present weather sensors (visibility/fog). In soft icing conditions (snow-like formation composed mainly of thin ice needles or flakes of ice). 0872J (Finland) Finnish Meteorological Institute (FMI) has performed field tests to verify the performance of ice detectors on an arctic mountain. According to the results.

27 shows an example of the ice load sensor EAG 200. pictures were taken at Zinnwald.5 m and consists of PVC. 27 shows an example of icing measurement poles.032 m. a length of 1 m and consist of PVC. It scales the mass of ice that is accumulated at the pole.Annex 15 EAG 200 (Germany) Figure A. 28. Figure A. operated at the station Zinnwald of the German Weather Service. The icing poles are still used for measurements at three locations at present (see 6. The measured mass is converted to a frequency signal in a range from 100 Hz to 10 kHz. The measurement error at 0°C is 0. February 2005) Before 1991 the icing measurements were performed by the use of manually operated icing poles at all locations the eastern part of Germany (see chapter 4 for a detailed description). 27 Icing measurement instruments used by German Weather Service ([37]. The measurement pole of the ice load sensor EAG 200 has diameter of 0.5% of the measured value with an accuracy 10 g. 80/110 . Icing pole Ice load sensor EAG 200 Figure A. The measurements are carried out by an electro-mechanical sensor in the lower part of the instrument. The findings are displayed in Figure A. The icing poles have a diameter of 0. The measurement results of both instruments were evaluated for the maximum ice masses that were measured during whole icing cycles (accumulation and loss of ice). a length of 0.035 m. The measurements of the ice load sensor EAG 200 show a slight dependency on temperature with a temperature coefficient of 15 g/10 K [56].2) in order to compare their results with simultaneous automatic icing measurements by the use of EAG 200. The EAG 200 has a measurement range from 0 g to 4000 g (note: type EAG 210 has a capacity of 10kg) and a resolution of 1 g.

or the measurements are erroneous.48 R2 = 0. 28 illustrates the limitations of manually operated icing pole measurements: For very large ice masses either no measurements are carried out. Nevertheless. 28. Ice mass (g/m) Figure A. Regression analysis of both measurement techniques shows acceptable results if the unreliable data points are excluded (see Figure A. This illustrates that reliable results can be achieved by the use of an automatic instrument. Ice mass (g/m) y = 0. but without confirmed unreliable data points 81/110 .5000 4000 Icing Pole. Figure A. no measurement possible" "End of icing cycle before icing pole measurement" 0 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000 14000 15000 EAG 200. because the impossibility of pole-handling. 5000 4000 Icing Pole. icing vanes) are missed. 28 Comparison of maximum ice masses. 29). Ice mass (g/m) Figure A. icing diameter. that were measured in icing cycles by the use of a manually operated icing pole and automatic icing measurements (EAG 200) during the period 1996-2004 at the station Zinnwald of the German Weather Service (remarks are from the measurement protocol of icing pole measurements).70 3000 2000 1000 0 0 1000 2000 3000 4000 5000 EAG 200. additional information about the icing types or about the icing geometry (e.g.72x + 129. Furthermore the icing cycle may have been finished before the measurement was carried out. 29 As Figure A. Ice mass (g/m) "Loss of ice before icing pole measurement" 3000 2000 1000 "Pole to heavy.

measured at the 100 m high tower of DWD’s Meteorological Observatory Lindenberg Ice Load [g/m] 82/110 . The example shows. 03. It will be reduced by appropriate temperature compensation in the future. 100 5 m a. Furthermore. 02. 120 90 m a.l.g. These offsets and their drift for the measurement system at the Falkenberg measurement site are mainly due to a temperature dependence of a signal converter. 03. 00 06: 0 01. 03.g. 00 18: 0 01. that the system operates well even for short icing periods and for small amounts of ice accretion. 30 Vertical profile of ice masses for an icing event.For the illustration of the EAG 200 system operation and performance a selected result of an icing event is presented in Figure A. It was measured at the 100 m mast of the Falkenberg measurement site of DWD’s Meteorological Observatory Lindenberg.l. it shows a minor zero offset for all ice load sensors as well as a zero offset drift. 0 2:0 51 -20 Time Figure A. that had to be used to adapt the ice load sensors frequency data output to the acquisition system of the tower. 0 0:0 50 05 01. 50 m a. 30. Icing measurements at three heights (see Table “Overview of present locations with icing measurements in Germany”) have been carried out at this location since winter 2004/2005. 100 120 80 80 Ice Mass [g/m] 60 60 40 40 20 20 0 0 -20 05 01. at least for the instrument in 5 m height.g.l.

31 The site on a clear day after winter snowfall. 33). H-pole supports are used. The Gerbers are mounted on this southern H-pole platform (Figure A. 32 shows the southern dead end platform and rotating rig. the southern H-pole having a platform attached for working and sensor attachment. Two Gerber PVM-100 instruments were provided by the UK Meteorological office and were installed at the EA Technology severe weather test site at Deadwater Fell test site on the English/Scottish border in the UK.Annex 16 Gerber Instruments Gerber (USA) The UK meteorological office sourced two instruments (Gerbers) capable of measuring LWC of the air. 83/110 . The site is fitted with load cells and turnbuckles to adjust the conductor tensions. Figure A. Test Spans The test span is 200m long and orientated North-South. Figure A. The aim was to see whether output from these instruments could be related to the conditions under which overhead line conductors suffer ice loads. Monitoring of the Gerbers started on 11th December 2003. 32 and Figure A. Data was collected over the winter period December 2003 to April 2004.

84/110 . Figure A. 34 to Figure A.10VDC output which give a measurement range 0. 33 Gerber instruments on southern dead end platform. Performance The Gerber particulate volume monitors have a 0 . 36 show some typical weeks data from the test site. 32 The Southern dead end platform and rotating rig at Deadwater Fell. Figure A.002 – 10g/m3.Figure A.

1st February 2004 Deadwater Fell Weather Data 26th January . 33.000 3.000 7.000 2.000 5.1st February 2004 Figure A.Gerber Instrument g/M3 / Tension in kN 4.000 Temperature in Degrees C / Wind Speed in m/s -10 10 15 20 25 30 -5 0 5 26-Jan-04 26-Jan-04 26-Jan-04 26-Jan-04 26-Jan-04 27-Jan-04 27-Jan-04 27-Jan-04 27-Jan-04 28-Jan-04 28-Jan-04 28-Jan-04 28-Jan-04 29-Jan-04 29-Jan-04 29-Jan-04 29-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 31-Jan-04 31-Jan-04 31-Jan-04 31-Jan-04 01-Feb-04 01-Feb-04 01-Feb-04 01-Feb-04 0 60 Date 26-Jan-04 26-Jan-04 26-Jan-04 26-Jan-04 26-Jan-04 27-Jan-04 27-Jan-04 27-Jan-04 27-Jan-04 27-Jan-04 28-Jan-04 28-Jan-04 28-Jan-04 28-Jan-04 29-Jan-04 29-Jan-04 Deadwater Fell 26th January .000 0.34 Direction in Degrees Gerber East/West Gerber North/South Temperature Direction (10% Ave) One PVM-100 was mounted and aligned north/south on the southern platform and the other aligned east/west as shown in Figure A. -10 Temperature Degrees C Hazel Temperature Wind Speed (10% Ave) 85/110 . 34 Weather data 26 January – 1 February 2004 Date 5 29-Jan-04 29-Jan-04 29-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 30-Jan-04 31-Jan-04 31-Jan-04 31-Jan-04 31-Jan-04 01-Feb-04 01-Feb-04 01-Feb-04 01-Feb-04 01-Feb-04 10 15 20 0 120 180 240 300 360 -5 Figure A.000 6. 35 Load cell and Gerber data for conditions in Figure A.000 1.

35. the temperatures are rising slightly but still sub-zero and both Gerbers are starting to indicate the presence of moisture. 36 Weather data from Deadwater Fell 23-29 February 86/110 . The wind direction is normal to the span and ice accretes on the Hazel as indicated by the increasing load in a steady wind speed. Deadwater Fell Weather Data 22nd . sub-zero temperatures and appropriate winds should not cause icing if there is no liquid water content i. the Gerbers do not indicate the presence of water particles.e.The site span is orientated north-south so winds around 0° or 180° should cause very little ice load accretion on the conductors according to the meteorological models even if icing conditions are met.29th February 2004 30 Temperature in Degrees C / Wind Speed in m/s 360 25 300 20 Direction in Degrees 240 15 Temperature Direction (10% Ave) Wind Speed (10% Ave) 10 5 180 120 0 60 -5 -10 23-Feb-04 23-Feb-04 23-Feb-04 24-Feb-04 24-Feb-04 24-Feb-04 25-Feb-04 25-Feb-04 25-Feb-04 26-Feb-04 26-Feb-04 26-Feb-04 26-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 0 Date Figure A. However. So this week indicates correct performance. 34 and Figure A. there are sub-zero temperatures up to 29 January with occasional winds normal to the span. the Gerbers do not indicate the presence of moisture and there is also no indication of ice load on the Hazel conductor. Winds around 90° and 270° will be normal to the span and icing on the conductor should occur if the meteorological conditions are right. Late on 29 January. However. In Figure A. Later on 30 January the temperature rises above zero and there is no icing even though the Gerbers indicate moisture.

50 0.00 10 2. 36 and Figure A. 37 show a period in late February. 34 Figure A. Deadwater Fell Weather Data 8th . These were associated with sub-zero temperatures for most of the period.50 0 1.00 20 3.00 5 1.50 Temperature Degrees C Gerber North/South Gerber East/West Hazel Temperature 2.50 Gerber Instrument g/M3 / Tension in kN 15 3. Sub-zero temperatures were present for almost the whole period and there were indications of moisture but no ice loads as the winds were always along the span. 37 Data output from Gerbers and Hazel load cell for period of Figure A. Figure A. 38 and Figure A.00 23-Feb-04 23-Feb-04 23-Feb-04 23-Feb-04 24-Feb-04 24-Feb-04 24-Feb-04 25-Feb-04 25-Feb-04 25-Feb-04 26-Feb-04 26-Feb-04 26-Feb-04 26-Feb-04 26-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 27-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 28-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 29-Feb-04 Direction in Degrees -10 Date Figure A.Deadwater Fell 22nd .00 -5 0. 38 Weather data for 8-14 March 87/110 . 39 show a week in March when the winds were almost always in a direction giving a major component across the span.14th March 2004 30 Temperature in Degrees C / Wind Speed in m/s 360 25 300 20 240 15 Temperature Direction (10% Ave) Wind Speed (10% Ave) 10 5 180 120 0 60 -5 -10 08-Mar-04 08-Mar-04 08-Mar-04 08-Mar-04 08-Mar-04 09-Mar-04 09-Mar-04 09-Mar-04 09-Mar-04 10-Mar-04 10-Mar-04 10-Mar-04 10-Mar-04 11-Mar-04 11-Mar-04 11-Mar-04 11-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 13-Mar-04 13-Mar-04 13-Mar-04 13-Mar-04 14-Mar-04 14-Mar-04 14-Mar-04 14-Mar-04 0 Date Figure A.29th February 2004 4.

000 20 3.000 10 2. This includes logged tension data from the bare Hazel conductor strung over the 200m span with an intermediate pole at 100m. Gerber indications of moisture but at above-zero temperatures on 13/14 March did not indicate icing conditions and no ice-load was measured. 38 On 9 March a minor indication of moisture from the Gerbers combined with a wind normal to the span and sub-zero temperatures indicated icing conditions which was confirmed by a minor load increase on the Hazel. A camera was also set up to monitor the ice load characteristics near the Hpole.500 0.000 5 1. Late on 10 March the Gerbers indicated more severe icing conditions with a wind at an angle of 30° to the span. 39 Gerber and load cell data for period covered in Figure A. A full 17 weeks data has been supplied to the UK Meteorological office.500 0 1.500 Gerber Instrument g/M3 / Tension in kN 15 3. 88/110 . The ice load stayed for a while as the temperatures were still below zero.Deadwater Fell 8th .500 Temperature Degrees C Gerber North/South Gerber East/West Hazel Temperature 2. A less significant incident occurred on 12/13 March when icing conditions showed up with a further ice load on the Hazel.000 -5 0. This would identify rime and wet snow icing events.000 08-Mar-04 08-Mar-04 08-Mar-04 08-Mar-04 08-Mar-04 09-Mar-04 09-Mar-04 09-Mar-04 09-Mar-04 09-Mar-04 10-Mar-04 10-Mar-04 10-Mar-04 10-Mar-04 11-Mar-04 11-Mar-04 11-Mar-04 11-Mar-04 11-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 12-Mar-04 13-Mar-04 13-Mar-04 13-Mar-04 13-Mar-04 14-Mar-04 14-Mar-04 14-Mar-04 14-Mar-04 14-Mar-04 -10 Date Figure A.14th March 2004 4. The Hazel suffered a significant increase in ice load which stopped when the Gerbers indicated no further liquid water content in the air. In the final day of the period load variations measured were due to an increase in wind speed and violent conductor movement.

enables a continuous measurement of ice mass built up by an automated system including.5 m and diameter of 30 mm.59. in connection with communication means. 40 Device Meteo on a concrete pole of LV power line.58. When the flash memory is filled up the oldest daily file will be overwritten by the latest one. Device Meteo measures following quantities: ice mass up to 20 kg with accuracy 0. location Predin (Photo: J.01 kg on the measuring rod with length of 0.Annex 17 METEO device (Czech Republic) During long-term development at the stand at Studnice. For eliminating the growing together of the icing on the measuring rod with the body of the Meteo device heating can be activated (heating is set up in dependence on temperature and on ice mass) [61. The value must be multiplied by two in order to obtain the value corresponding to 1 m temperature with accuracy 0. EGU Brno designed a new automated monitoring equipment METEO which. wind velocity and direction and temperature without the requirement for the presence of attending personnel [57.62]. range from 10 to 40 m/s with accuracy ± 1 m/s wind direction – angle from 0 to 355 degrees in steps of 5 degrees. velocity and direction of wind are recorded every minute and temperature and ice mass every ten minutes.60]. Figure A. The size of the daily file is 6 192 B and the capacity of the disc is sufficient for 40 or 80 days (depending on size of the memory used).1°C range of wind velocity from 0 to 10 m/s with accuracy ± 2 m/s. Measured values are stored in the flash memory.Sabata) 89/110 .

recorded in the flash memory. Daily files with measured values. Standard regime The monitoring equipment measures the current quantities. two of them on the territory of regional utility VCE. outage of supply and its restoration. by itself. are sent into SCADA system at the time of low radio network load (usually at the night). The automated monitoring equipment are installed on the towers of LV.The concept of its solution assumes to build up a network of monitoring equipments by which the measurement will be performed and the data on air temperature. B. The warning messages may indicate: exceeding the set up ice mass. The first one was put into operation in 1999. All data are archived in this system [63]. 12 on the territory of regional utility JME. The operation itself of the automated icing monitoring system is realized in two regimes: A. By the command of the server from the centre or on demand of the dispatcher the newly measured data will be transmitted and downloaded into the central data base. exceeding the value of wind velocity. the others in 2001. Actual state At present 14 devices are in the operation. Warning regime Each monitoring equipment may be set up with regard to the expected local conditions in such a way so that it may. processes the measured values and downloads them into the local database for each equipment. MV a HV overhead lines. ice mass and wind velocity and direction processed. Each hour current data from each METEO device are sent into the dispatch system and are displayed in the dispatch office. exceeding the set up steepness of ice growing. foreign intervention into the monitoring equipment. Means such as radio networks or GPRS may be used for data transmission and for communication between the centre and individual monitoring equipments. Data transmission The data and alarms sent by each monitoring device to the central dispatch office are transmitted by the radio network via retransmission points. 90/110 . send a warning message into the centre when the set up values have been exceeded or when another abnormal event arised. The dispatcher can also make query any time to get immediate values.

18th December 2004 – location Nyklovice. 41 Daily record of temperature and ice mass. the end if icing event 91/110 .Figure A.

several icing events occurred during the winter 2005 /2006. and in some cases also precipitation is detected – type and intensity. including camera – which stores files in jpg format. At the site in Åre. no significant icing occurred at the site in site in Ritsem (Sweden) at the test site in Drammen (Norway). 92/110 . The prototype was designed for a maximum ice load of 100 kg. as the pipe is supported by a rod which is resting on the load cell. but in Drammen once every 10 minutes. Statnett. and later it was modified for lower maximum loads (10.) During the same period. for the Norwegian Power Grid Company. (see figures below. 25 or 50 kg). At the first site in Norway (Drammen) and at the Swedish sites: in Åre and in the far north (Ritsem) the IceMonitor is connected to a monitoring station designed by SAAB Technologies (previously AerotechTelub). The monitoring equipment was originally designed for use in road weather information systems and it is capable to handle a number of different sensors. These stations also perform measurements of weather parameters. Icing of this camera has been a problem during the winter 2005/2006 – and the installation and de-icing of the camera will be modified before the next icing season. wind speed and direction. Measurements are made once every half hour normally. relative humidity. To log data any kind of data logger with standardized current input (4 – 20 mA) can be used. Data is retrieved to a server in Östersund via radio/radiolink or in some case via telephone modem. and their research test site in the mountains west of Oslo. The output of the load cell is connected to a precision amplifier and converted into a standardized output current loop – 4 to 20 mA. To avoid ice in the area for the bearing of the rod there is electrical heating of the bearing that is controlled with a thermostat. A camera is monitoring the IceMonitor sensor at the site in Åre. freely rotating sensor (steel pipe with a surface area of 5 dm2) is weighed by a load cell. The ice that accretes on the vertical. A couple of the icing events have been severe with ice loads of more than 40 kg. such as air temperature. To be able to perform testing of the instrument remotely there is a test relay included that will activate an electrical unbalancing of the load cell – at which the output signal will increase with 8 mA to indicate that acquired data are reliable.Annex 18 IceMonitor (Sweden) The IceMonitor was originally developed for the use in a power line surveillance system installed in Norway 2003.

43 First significant icing event at Åre during winter of 2005/2006 93/110 . 42 Ice build-up developed during the two severe icing events during 2005/2006.Figure A. Figure A. Details are shown in the figures below.

44 Second significant icing event at Åre. winter 2005/2006 94/110 .Figure A.

Figure A. The site is also prepared for remotely controlled de-icing of the sensor using 95/110 . with a camera mounted on the wind sensor support. 45 Pictures of the site installation in Åre on top of the roof of a ski lift house.

It comprises of a head with an IR emitter and a photo detector and a probe. The sensors are equipped with a health-test which provides an early warning if the performance of a sensor is degraded. The sensitivity is 50 g/m2 • The T23 Clear-Ice Indicator™ indicates the presence of atmospheric icing. The probe heating is internally controlled and the sensitivity is 50 g/m2. The probe is mounted on a cylinder =30 mm L=500 mm.Annex 19 HoloOptics T20-series Ice Detectors (Sweden) • The T21 Ice-Duration Indicator™ and Ice-Guard™ indicate the duration of ice on nonheated structures and can be used for controlling anti-icing or defroster equipment. A single-direction T20-series indicator is sufficient if there is only one wind direction of interest and it is known (e. 46 T26 Icing Rate Sensor Single and omni direction versions 96/110 . In all other cases the T20-series omni version is recommended. on a wind turbine nacelle). If not indicated otherwise. Figure A. The sensitivity range is 50 g/m2h – 18 kg/m2h Common features of the T20-series ice detectors The T20-series indicators single-direction (±45° upwind) or omni-direction versions indicates either all icing types including clear ice or rime ice/wet snow growth only. T21 has no probe heating so if used as controller the probe is to be heated by the anti-icing or defroster system. reference in this Annex is made to the indicator sensitive to any type of ice including clear ice.g. • The T26 Icing-Rate Sensor™ measures the atmospheric icing rate using internally controlled high power probe heating. All T20-series ice detectors are based on a patented digital optronic ice-indicator that indicates the presence of any type of atmospheric ice including clear ice.

Ice may start to build up on the probe again as the probe cools down. Sourva Icing Rate icing rate g/m2h 200 150 100 50 0 495 500 505 hour Figure A. The indication of icing will stop when the probe heating is turned off. air and surface temperature. the time it takes for the probe to be covered with ice after it has been de-iced. 47 Typical icing indications measured at Suorva. Sourva Indications Indications off on 495 500 505 hour 510 515 520 Figure A. Northern Sweden Icing Rate 1h average 510 515 520 10 External control of the probe heating is available 97/110 . The time interval between two indications may be as short as ten seconds. This cycle is repeated for as long as ice is created on the probe surface. After a short period of time the ice has melted and the water has fallen off the probe. the internally controlled probe heating10 is turned on without time delay. As ice is detected. Northern Sweden The time it takes to melt the ice is dependent on many factors (e.Description of operation – T26 Icing is detected if more than 95 % of the probe is covered with a 50 m thick layer of clear ice or 90 m thick layer of other types of ice. icing rate. The time between icing indications is used to calculate the icing rate using the T26 Icing-Rate Sensor . wind speed and type of ice). 48 Icing rate measured at Suorva. If sufficient heating is applied. melting power. is mainly dependent of the icing rate.g.

Stockholm. In co-operation with the airport authorities Greenville Wind power plant. Norway.Measurement campaigns 2003-2006 The 23 Clear-Ice Indicator and the T26 Icing-Rate Sensor beta versions have been tested at the following locations: The Suorva wind power plant. In close cooperation with Norsk Hydro Åre Ski Resort. In co-operation with FOI (The Swedish Defence Research Agency) KTH (The Royal Institute of Technology). In cooperation with General Electric Sandhaugen test felt. Sc. Cooperation with Norsk Miljøkraft University of Narvik. tested as a defroster controller for freezers 98/110 . 150 km north of the polar circle. Sweden. Cooperation with the University of Narvik. 400 km North of the polar circle. Keller test field. in cooperation with KTH (The Royal Institute of Technology). Norway. Cooperation with Combitech. USA. Norway. Sweden Bromma Airport. The T21 Ice-Guard is. ski lift protection. Stockholm. 300 km North of the polar circle. Tromsö. Power line protection. Sweden.

Its long-term use has shown it to be quite durable and it seems to detect icing in reasonable agreement with humidity indications and video recordings [21]. 99/110 . V. No further information is available on this instrument.Annex 20 Instrumar IM101 Instrumar IM101.2.4 is an ice detector based on measuring the electrical impedance and surface temperature [29].

66. [65. No references are made to wind tunnel studies in other countries in this report. ice grown under the conditions of rime formation exerts the significant influences on the aerodynamic properties of the cup particularly in the range of the angle of incident higher than 90 degrees. it can be implied. 53 100/110 . that the decrease of the measured wind speed by an iced cup anemometer is attributed to a decline of driving force of cups with an ice deposit. The most important parameters of LWC and MVD were obtained by the rotating multi-cylinder method through some calculation. The reduction rate in the figure was defined by the following equation: Vclean V * (1) RR Vclean where Vclean and V* denote the wind speed measured by a clean anemometer and by one with iced cups respectively. Dry-icing and Wet-icing mean the conditions in which rime and glaze grow respectively. the weaker the torque around the shaft tends to be. [64] Icing on anemometer In the first place.Annex 21 Wind tunnel calibration This annex refers to Kanagawa Institute of Technology. 52. The behaviour of cups was calculated by solving the equation of motion with respect to a cup part of a cup anemometer considering the force acting on each cup and stem. After all. The principle of ice detection is based on the reflection of infrared light within an ice deposit emitted from the tip of the optical fibre connected to the sensor. Verification of an ice detector To know the duration of an icing event exactly pointing out its beginning and end is essential in taking efficient countermeasures for the prevention of icing. the thickness of ice could not be measured. At that moment. KAIT conducted the icing wind tunnel test for the German ice detector Infralytic developed for wind turbines. 50 shows the growth of ice accretion on a cup anemometer at the different temperatures. As can be seen in the figure. For that purpose an ice detector is usually employed.67] Figure A. 49. KAIT has conducted the icing wind tunnel test using a cup anemometer to know the manner how ice accretes on a cup anemometer and the effect of ice accretion on measurements. 51) for the acquisition of the aerodynamic characteristics of an iced cup-shaped body was carried out to infer the aforementioned reduction of measured wind speed by the iced cup anemometer. The Finnish Meteorological Institute has been carrying out the field tests to verify the performance of ice detectors in the arctic region. It turned out that the ice detector can detect ice on the test model. rime affects significantly the performance of the cup anemometer. and at bearings supporting a rotating part. As shown in Figure A. The results are shown in Figure A. there is a negative thrust force generated which reduces the cup revolution. the measured wind speed dramatically decreases. Ordinarily in this range. Icing wind tunnel test was done according to the ordinary test scheme as depicted in Figure A. as a collaborating member of COST Action 727. as easily expected. As ice grows. Japan. The additional wind tunnel test (shown in Figure A. The greater the negative thrust becomes. Note that the terms of beginning and end mean the onset and completion of an icing event. the numerical approach was taken to evaluate the effect of icing on a cup anemometer. not of icing conditions. With using the acquired data.

[68. even if the cup open face is covered completely. 58). 63. whilst the ultra-sonic anemometer transmitted meaningless signals for 56 hours.69] Based on the test results mentioned earlier. The former was the formation of ice in an icing wind tunnel test letting super-cooled minute water droplets collide onto an airfoil model with a specific velocity where the complicated thermodynamic process always takes place.and/or anti-icing measures. the wind tunnel test (Figure A. An ice deposit accreted on specimens was created in two different manners: dynamic and static ice formation. 55).Snow accumulation of anemometers How snow accumulates on an anemometer and how a snow deposit affects measurements have been examined by the wind tunnel test where snow flakes of dendritic shape were fed into the wind tunnel test section by a snow fall device as shown in Figure A. humidity. The results show that the skilfully fabricated covering exerts negative influences on the aerodynamic characteristics of a cup-shaped body only when the air stream flows into the covered side of a cup (Figure A. All the results are shown in Figure A. Coatings to reduce the adhesive strength of ice In order to seek the possibilities of coating as de. 57) was conducted in order to investigate the aerodynamic effect of the cup covering of a cup anemometer. together with temperature. In those tests. 60 where the wind speed measured by the covered cup anemometer was calibrated by using the transfer function that can be seen in the figure. The correlation between the covered and ultra-sonic anemometer was depicted in Figure A. solar radiation and precipitation was successively measured for 186 days starting from November. and with cups with snow deposit having the leveled surface in order to quantitatively evaluate the effect of change of the shape of snow accumulation inside a cup in terms of measured wind speed as well. Field measurements using a heated covered-cup anemometer and a heated ultra-sonic anemometer were implemented for verification of the effectiveness of cup-covering in icingendangered mountains in Japan. the test results shows that one type of coating can significantly reduce the ice adhesion regardless of the manner of formation of ice [70. The most intriguing feature is that the snow deposit grows no further beyond some certain height. Figure A.71]. The typical results are shown in Figure A. The covered cup anemometer stopped its measurement for only 1 hour. 63 and Figure A. It turned out that the correlation is so satisfactory that the cup covering of a cup anemometer would be effective when used in icing prone areas. which was approximately 50% in the centre in those tests. During the period of campaign. 59 shows the linear relation between the measured wind speed and the cup revolutions of the covered cup anemometer slightly changing the coefficients of the linear transfer function provided by the manufacturer. the performance of commercially available coatings was examined by load test. probably due to ice or snow accretion on the emitter/receiver(s). the force was applied parallel to the adhesion surface until ice was removed. Wind speed. 101/110 . The consequence revealed that wet snow accretes dramatically only on a cup anemometer and the surface of the snow deposit is asymmetric with respect to the axis through the centre of a cup (Figure A. These findings led us to undertake additional tests using cup anemometers with cups covered by a thin membrane in the cup open face. precipitation over a period of 754 hours was identified. 56. 54. The latter was actualized in the low temperature chamber where water was poured into a steel ring placed on a specimen and then frozen at rest. Although there may be a problem that has to be solved in terms of durability. This indicates that snow accumulation inside cups reduces the measured wind speed by a maximum of 13% and the reduction of the measured wind speed by the cup anemometer is merely 20%.

51 Wind tunnel test for acquisition of aerodynamic characteristics 102/110 . Imitated model (c).Test model Airstream Sprayer Figure A. Actually iced cup (a). Test model in the test seci Figure A.) Figure A. 50 Effect of ice accretion on measurements by a cup (b). 49 Schematic view of icing wind tunnel 60 50 Reduction Rate (%) 40 t30 30 20 10 Wet-Icing 0 0 50 100 150 Dry-Icing Elapsed Time (min.

6 0.8 0. 60min.2% 0.2mm to 1.2% value 63.5 0 D-10 D-15 D-25 D-35 D-45 Normal force coefficient.2mm intervals 120min.) 180 (a).8 0.2 1. 54 Schematic view of the wind tunnel test for snow accumulation on anemometer Figure A. n/nmax Ice thickness from 0. 52 Aerodynamic characteristics of an iced cup-shaped body 1 RPM.6 0. Dry-icing Figure A.5 0 -0.) 180 Clean W-10 W-15 Normal force coefficient.4 RPM. 63.4 0.5 -1 0 30 60 90 120 150 Angle of incidence (deg.2 0 0 1 Time (s) 2 3 0 1 Elapsed time (s) 2 Figure A.5 -2 0 30 60 90 120 150 Angle of incidence (deg. 180min.1e5 1.5 -1 -1. Cn Re=3. normalized by the maximum rpm of the clean cup value 1 0.2 0 Clean cups Clean cup 20min. 53 Calculations for cup behaviour Snow fall device Dendritic snow flakes Sprayer Test model Airstream Figure A. Wet-Icing (b).5 1 0. 55 Comparison of snow deposit between wind tunnel test and field test 103/110 . 0. Cn Re=1.1e5 -0.5 1 0.0mm at 0.

0m/s V=7. 02732 RPM 0 .5m/s 10 V=10. 60 Field measurements by a covered cup anemometer and an ultra-sonic anemometer 104/110 . 30 60 90 120 (deg . 3879 Eq.Reduction Rate (%) 20 Naturally accreted in WT V=5m/s Artificially accreted V=5. 58 Aerodynamic characteristics of a covered cup 30 25 Wind speed (m/s) Wind tunnel tests & curve fitting for the covered cup anemometer 20 15 10 5 0 0 V 0 .) 150 180 Figure A. from VAISALA 200 400 600 800 1000 Revolutions (RPM) Figure A.5 1 0. 57 Test model in wind tunnel Figure A. 59 Transfer function of a covered cup anemometer Figure A.5 Cup-A Cup-B Cup-A Cup-B 0 Incident angle.0m/s Fully covered with felt 0 0 25 50 75 100 Dimensionless depth of snow (%) Figure A. 56 Effect of snow accumulation and cop-covering on measurements 2 Clean Cup 1.5 0 -0.

of acquired data End of icing event Ch1 W /E 2mm 500 600 700 Figure A.5 Intensity 1 0. 62 Signals from the ice detector in ice detection Adhesive strength (kPa) 500 400 300 200 100 0 Coatings Adhesibe Strength (kPa) Static test WT Test No.5 0 0 100 200 300 400 No. 63 Comparison of the adhesive strength of ice formed in the wind tunnel test and in a chamber 105/110 . 61 Ice detector tested Onset of icing event 1.Figure A. of iteration Figure A.

.Horbaty. (in Russian) IEC 60826: Design criteria of overhead transmission lines. H. and Säntti. Joss. E. Dobesch.Horbaty. Baring-Gould. Tammelin. 2000-07-20 11th Workshop on Atmospheric Icing of Structures (IWAIS 2005). A.fi. T. 12 . G. Mont Aigoual. B. 2004: Improvement of severe weather measurements and 3 4 5 6 7.Part1: Safety requirements". L. 8.. Säntti. J. Holodov... Durstewich. Tammelin. Finnish Meteorological Institute.3. Bellevaux. http:\arcticwind. number IEC 61400-1:1999(E) Tammelin.. 2005: IEA Wind Annex XIX. Instruments and observing methods report. A. First Edition. and Ronsten. 1998: The Final report on the EUMETNET project “Specification of Severe Weather Sensors”. H. B. Holttinen.. A. A.Tallhaug.. Peltola. Rast. K. Musa. and Peltomaa.. 2003.V. C.. 37p. M. IEC 61400-1: "Wind turbine generator systems . H. no. T. H. 44-48. Leroy.. 2005.. T. Trudy GGO 408. Heimo..vtt. Peltola. 9 10 11 12 13 14 15 106/110 . B. http:\arcticwind. Laakso. ref. Kury. B. Ganander.2006. Meteorological Publications No. 2005: IEA Wind Annex XIX.3. Finnish Meteorolgical Institute. N.fi.127 p State-of-the-art of wind energy in cold climate.Tallhaug. Dal Cin. 41. C. B. Canada.. 62 Technical document (WMO). August 1997. ISO/TC 98/SC 3. M. Morgan. 1999. FMI reports 2004:3 Tammelin. R. date of access 20.Annex 22 References 1 2 COST-727 Action: Memorandum of Understanding WMO Wind Instrument Intercomparison. V. E. date of access 20. K.. Säntti. Reports 2005:6.16 June 2005 (2005) Popov. IEC 61744: “Overhead lines – Meteorological data for assessing climatic loads”. J. G. 859 Tammelin. M. I. Ronsten.. Tammelin and al Improvement of Severe Weather Measurements and Sensors – EUMETNET SWS II Project. Finnish meteorological Institute.. Helsinki 2000. R. ISO 12494: Atmospheric icing of structures. Laakso. Lacroix. I.. E. Seifert. Laakso. 2005: Wind turbines in icing environment: improvement of tools for siting.. Oualid. B. Cavaliere. Baring-Gould. G. Gregoire and G. B. Montreal. Ronsten. no. 1978: On the applicability of icing measurements.. 53 p.. certification and operation – NEW ICETOOLS. Second Edition. Peltola. L. J. Holttinen.2006 Wind Energy Projects in cold Climates.. K. France 1992-1993: Final report by P. G. H.. 2001: Wind Energy Production in Cold Climate. M.. Helsinki.vtt. Lacroix. and Haapalainen..I..

in Finnish. 4th International Conference on Atmospheric Icing of Structures. Iceland. World Meteorological Organization. Wind energy production in cold climates.. Lehtonen. 2001: The Follow-up of Olostunturi Arctic Wind Farm. Journal of Atmospheric and Oceanic Technology. 2002. S. Lehtonen. Switzerland. 543-558. 9-11 April 2003. Proceedings. L. K. Third International Workshop on Atmospheric Icing of Structures. Finnish Meteorological Institute.. P.. Paris. 101 p. 2002:Observation Tools of Icing Events. L. Reykjavik. BOREAS VI Meeting.343. Laiho.. Proc. Vol 18. Laakso. E. A. 8th IWAIS. Korolev. L. Harstveit. & Makkonen. Makkonen. Secretariat of the World Meteorological Organization. ISBN 951-697-584-4.0-1998). Boreas VI..Tallhaug and A. Norway. T. Riisiö. Finland. Makkonen.. Mazin. Harstveit. 18-20 May 1992. 2005: Ice accumulation observed by use of web camera and modelled from meteorological parameters. B. 2000: Experience on the ISO Reference Collector. L. I. Wind energy production in cold climates.. Public report of EU-sponsored project New Icetools.368. Proceedings. 1992: Can ice load measurements be standardized? International Seminar on iceload measurements. P. VTT Energy reports 24/2001. 9 (3): 258 .... Reports 2004:3. Tammelin. Vancouver.7 September 1988. J. G. 5 .263.8 May 1986. 2001: Thermodynamics of Icing Cylinder for Measurements of Liquid Water Content in Supercooled Clouds. Final Report.. Heymsfield. NNE5/2001/259. T. & Miettinen. G. Cober. 16 17 Germanischer Lloyd Rules and Regulations IV . P. 338 . Journal of Atmospheric and Oceanic Technology. K. Pyhätunturi.. 1988: Studies on atmospheric icing in Finland. A. 6 . Lehtonen.. K.Non Marine Technology Part 1 . P. Makkonen. Laakso. Ahti. Proc. Finland.. Finland. WMO-No. Helsinki. A.. Isaac.Wind Energy. Makkonen. 1992: Analysis of rotating multicylinder data in measuring cloud droplet size and liquid water content. 2005. L.. 1986: The growth and disappearance of ice loads of a tall mast. Geneva.. K. 363 .. 2003: Calculating duration of ice weights at reference objects along the Norwegian coast originated with in-cloud rime.. Saariselkä. 18 19 20 21 22 23 24 25 26 27 28 107/110 . L.. Boreas VII. 12 p. Peltola. 8. 1996: Guide to Meteorological Instruments and Methods of Observations. 2003: Evaluations of ice-free anemometers and ice detectors. Kristiansand. V. H. Fidje. G. Pyhätunturi. Ronsten. 357. Säntti. 2003. (GLRP3. P.senors – EUMETNET SWS project.

In: Tammelin et al. Tokyo. B. Budapest. E.3. 1971: Glaze and Rime.. Dranevic. A J. Proc. Fikke and A. International Expert's Meeting on Wind Power in Icing Conditions (BOREAS). Finnish Meteorological Institute. L. Date of reference 20. CD-ROM. International Seminar on Iceload Measurements. Trudy GGO 425. 239-245. 7 p. Wichura. 10 . S. C.. ISBN 951-697-643-3. L. M. Saariselkä. Guljajev. L. Brno 2002... Atmospheric icing at Lomnický štít.fi. 227 p. First IWAIS.M. Journal of Atmospheric and Oceanic Technology. 26-30. Leningrad Gidrometeorologitsheskoe.49. L. U. Ostrožlik M. 2001: Extreme Value Analysis of Ice Accretion Data from Norwegian Measurement Rack Network. Teucher.. 1982: Icing related problems. Proceedings ISOPE 2001. 1979: On the duration of icing.P. Oleskiw. Makkonen. 515-528... Assessing the Rosemount Icing Detector with In Situ Measurements. effect on line design and ice load mapping. & Helle.labko. (in German) Internet page of Wavin-Labko Oy. 1973: A device for continuous measuring and recording of ice accretion. Nolan. M. & Laakso. 2001: Anemometry in icing conditions. 1992: Structure-independent measurement of icing. 2005: A survey of icing measurements in Germany.35/3. 2005: Humidity measurements in cold and humid environments. E. 131-147. Farzaneh.2006 Cober et Al.. 10th IWAIS. Goel.. Hetta: 43 . 18(9): 1457-1469. Paper A1-5. 2002: Historical survey of ice investigation on the territory of the Czech Republic. Journal of Atmospheric and Oceanic Technology. 80-87.. 1993: Comparison of growth rates on groups of stranded conductors’ EATL Contract..13 February 1992. Abhandlungen Meteorologische Dienst DDR 107. Popolansky. 1990: Development of a sensor for the detection of atmospheric ice. E. Japan. H. K. 35 36 37 38 39 40 41 42 43 44 45 108/110 . F. Army CRREL Special Report 83-17. Thorsteins. Vol 18. P. 1993. Norway. Kristiansand. 1992: Iceload measurements in East Germany. BOREAS VI. 1993 Nikiforov. J.. Makkonen. T. 116.. 5th IWAIS. Savadjiev.29 30 31 32 33 34 Rothig. Lehtonen. Proceedings. 18-20 May 1992. S. Norway. Geophysics and Geodesy Vol. 1993Data analysis of icing measurements in Iceland.. Makkonen. Boundary-Layer Meteorology. Eliasson.S. 6th IWAIS. Shackleton L et al. 6 pp..N. http://www. Stavanger.P. F. 2005. Inkpen. (Editor).

P. 1998: The climate of freezing rain over the province of Quebec in Canada: A preliminary analysis. a. Tesa M. Norway. 1998: Automated system for icing monitoring – supply area of V E (in Czech). Popov. Periard. Brno. Laflamme. Kvacova. Atmospheric Research 46. B.O. P. H. Trudy GGO 379. Maderová. – 020.s. Center for Construction of Scientific Equipment. Zalesak. 154-159. Czech Republic 54 55 56 57 58 59 60 61 109/110 . J. N. (in Russian) Podrezov. 1987: Intense icing in mountain regions and ways to improve ice deposit measurements. N. 87-93.. Beschreibung und Bedienungsanleitung (in German). 2000: Automated icing monitoring system.s. N. – 020. P..N.46 47 48 49 50 51 52 53 Zaharov. Zalesak... 3 Academy of Science of GDR..I. Fišák. Holodov. 1984: Distribution of ice loads in the USSR. International Seminar on Ice Load Measurements. IWAIS 2000. 19 pp. 99-111.. Standard SN-318-65. J. EGÚ Brno.. – 020. EGÚ Brno. a.. 59-65. 1986. Hydrol. 1976: The effect of structural parameters of electric power lines on ice loads on cables. EGÚ Brno. 30-36...K. Z. Trudy GGO 303..V. 1978: On the applicability of icing measurements. Trudy GGO 485. Berlin. A. Vojta J.I. J. Naumov. Z. O. 2001: Automated system for icing monitoring. Kvacova. 44-48. Toronto. 1973: Correction factors for ice load measurements.. S. 1999: Automated system for icing monitoring – supply area of JME (in Czech). Zalesak.. Kvacova.. P. Holodov. 15 pp. P. a. M.U.. 2000: Automated system for icing monitoring – supply area of SME (in Czech). Popov. G.I. 22-24 March 1976. H. 53 p Laflamme. Project documentation for supply area of JME (in Czech). P. (in Russian) Zavarina. 13 p. Lehky.G. 1976: The observations of rime and glaze deposits in Quebec.s. Felin. V.V..ice meter.A. 1992: Space and time estimates of ice loads from two interactive networks of ice accretion measurement sites. J.. Canadian Electrical Association Spring Meeting. 2000 Lehky.: Instrument for measurement of the amount of the solid precipitation deposit . (in Russian) Popov. a.... V.. 1966: Instructions on Determining Glaze-ice Loads. Arnautov. Kvacova. 18-20 May 1992. Trudy GGO 408. 2001 Lehky.. 1998 Lehky. Eisablagerungsgeber EAG 200.. Anon.s. EGÚ Brno. P. Moskow. H. Kristiansand. H.. Sabata J... Z. Chum J. Meteorologiya i Gidrologiya 4. – 020.... Hydrolmech 49. 1999 Lehky..

. Finnish Meteorological Institute Kimura. Abe.62. 10-pages 66 67 68 69 70 71 110/110 . No..63. CD-ROM.62 63 64 65 Lehky. 2005: Icing wind tunnel test for a new ice detector. Tsuboi & K. Takeshi Sato. Holger Freitag & Bengt Tammelin. Sabata J. S.Rampl M. Czech Republic Kimura. S.45-58 Kimura.33. Proc. Sato. Takeshi Sato. S.. CD-ROM. Shigeo. Bengt Tammelin & Kazuhiro Tsuboi: How much does ice affect wind speed measurement? . 10-pages Kimura. Proc.. K. Furumi. CK CIRED 2005. T. of The 7th International Symposium on Cold Region Development. Kenji Kosugi. Tsuboi. Kosugi. 11 pages Kimura. Finnish Meteorological Institute. Sato & K. BOREAS VII. Finnish Meteorological Institute. T. BOREAS V. of The 7th International Symposium on Cold Region Development. & Hiroshi Morikawa: Wind Tunnel Tests for Wet-Snow Accretion on Anemometers. of The International Meeting on Wind Power in Cold Climates BOREAS-VI. K. Saariselkä.. Tammelin. Report number 44. Suzuki: Aerodynamic characteristics of an iced cup-shaped body. BOREAS VII 2005. (Submitted to Journal of Cold Regions Science and Technology) Kimura Shigeo. 2004: Why do the cups of a cup anemometer have an open face? Proc. Issue1. No. 2004: Evaluation of anti-icing coatings on the surface of wind turbine blades for the prevention of ice accretion. 2003: The effect of anti-icing paint on the adhesion force of ice accreted on a wind turbine blade. 2001. Meteorological publication. Peltomaa & K.. K. A. pp. Tsuboi: Icing effect on a cup anemometer. Tabor. B. S.: Automated icing monitoring system on the territory supplied by JME (in Czech). 2000. Takeshi Sato & Kenji Kosugi.. Finnish Meteorological Institute Kimura. K. Journal of Cold Regions Science and Technology. B. 29pages Kimura Shigeo.: Automated icing monitoring system. et al. Finland Vaculik P. Shigeo. P. Tammelin. Vol.

CHF 75. CHF 78. 390pp. Grundlagen aerologischer Messungen speziell mittels der Schweizer Sonde SRS 400. 125pp. CHF 18. Musa M. 181pp. CHF 86. CHF 12. CHF 100. CHF 42. Leuenberger D: 2005. International Conference on Alpine Meteorology and MAP Meeting 2003. Das Schweizer Klima im Trend: Temperatur. W: 1996. Interannual climate variability in the European and Alpine region. CHF 36. CHF 25. M: 1997.Veröffentlichungen der MeteoSchweiz 74 73 72 71 70 69 68 67 Schmutz C. Bader S: 2004.Bolliger M: 2002. 80pp. CHF 36. Handbuch zu den NiederschlagsTotalisatoren. et al. CHF 83. 99pp. The Swiss Contribution to the Global Atmosphere Watch Programme – Achievements of the First Decade and Future Prospects. Stübi R: 2002. CHF 50. 103pp. CHF 27.Müller G und Viatte P: 2005.und Niederschlagsentwicklung seit 1864. 170pp. CHF 33.Begert M. Extended Abstracts volumes A and B. 48pp. CHF 34. Homogenisierung klimatologischer Zeitreihen. H: 1994. Barodte B: 2006. Baudraz G und Moesch M: 2003.): 2003. Analysis and Prediction of the European Winter Climate. 132pp. SONDEX / OZEX campaigns of dual ozone sondes flights: Report on the data analysis. 115pp. Schlegel T. Potenzialabklärung für Public Private Partnership (PPP) bei MeteoSchweiz und armasuisse Immobilien. Seiz G. CHF 76.Baudenbacher. Jeannet P. CHF 38. High-Resolution Radar Rainfall Assimilation: Exploratory Studies with Latent Heat Nudging. dargelegt am Beispiel der Lufttemperatur. EUCOS impact study using the limited-area non-hydrostatic NWP model in operational use at MeteoSwiss.Richner H. 136pp. Zu r Methodik einer Beschreibung der Entwicklung des linearen Trends der Lufttemperatur über der Schweiz im Zeitabschnitt von 1864 bis 1990.Schraff.Bettems J-M: 1999. Data Assimilation and Mesoscale Weather Prediction: A Study with a Forecast Model for the Alpine Region.Müller WA: 2004. Richner Hans (Eds. 138pp.Schär Christoph. 82pp.Gisler O: 1999.Mathis H: 2005. dargelegt am Beispiel der relativen Sonnenscheindauer. The impact of hypothetical wind profiler networks on numerical weather prediction in the Alpine region. CHF 81. On the characteristics of heavy precipitation systems observed by Meteosat6 during the MAP-SOP.Wolfensberger. Homogenisierung von Klimamessreihen der Schweiz und Bestimmung der Normwerte 1961-1990. Schlussbericht des Projektes NORM90. 116pp.Bosshard.- 66 65 64 63 62 61 60 59 58 57 56 55 .Scherrer SC: 2006.Bettems JM: 2001.Favaro G. Homogenisierung langer Klimareihen. Müller P. C: 1996. 112pp. Stübi R: 2002. Binder Peter. Chronik der Totalisatoren. Impact of Realistic Greenhouse Gas Forcing on Seasonal Forecast Performance. 580pp. 140pp. 17pp.: 1999. 78pp. Re-evaluation and trend analysis of the Payerne ozone sounding. CHF 38. 65pp. CHF 40.

Gaia M. CHF 59. 81pp. 2006. 74pp. Rohling S: 2004.– Schmuki D. Swiss Model Simulations for Extreme Rainfall Events on the South Side of the Alps. 80pp. Switzerland. 73pp.9.Buss S. CHF 72. Rohling S: 2004. Starkniederschlagsereignis August 2005. Vonderach G: 1998. Aeronautical Climatological Information Grenchen LSZG. 58pp. CHF 25. CHF 34.– Schmuki D. Tropische Wirbelstürme – Hurricanes –Typhoons – Cyclones.– Schmutz C. CHF 26. Ruffieux D: 2002.MeteoSchweiz: 2006. Die Maximaltemperaturen im Hitzesommer 2003 und Vergleich zu früheren Extremtemperaturen. 10pp.– Duding O. Aeronautical Climatological Information Bern LSZB.– Schmutz C. Vogt R.Kaufmann P: 2002. Aeronautical Climatological Information Sion LSGS. 100pp. L: 2006. CHF 16.– Moesch M. 214.Frei C: 2006.13. Aeronautical Climatological Information St. Globalstrahlungsmessungen 1981-2000 im ANETZ.WRC Davos (Ed): 2001. Aeronautical Climatological Information Zurich LSZH. Schmutz C. Aeronautical Climatological Information Lugano LSZA. CHF 14. 72pp. Saisonale Vorhersage in Tradition und Moderne: Vergleich der "Sommerprognose" des Zürcher Bööggs mit einem dynamischen Klimamodell.. Galli G. Die Niederschlagsverhältnisse in Basel 1964-1998.. IPC . Rohling S: 2004. CHF 24. CHF 20.meteoschweiz. 78pp.Gallen LSZR. CHF 26. Der Föhnfall vom April 1993. Rohling S: 2004. Eine Länder übergreifende Niederschlagsanalyse zum August Hochwasser 2005. 83pp.ch » Forschung » Publikationen . Schmuki D. CHF 70. 100pp. 25.Bader S: 2004.2000.Z’graggen.– Bader S: 2004.. 40pp. CHF 40. 110pp. Schmuki D. Schmuki D.– Schmutz C. CHF 75. Jäger E and Schmutz C: 2005: Evaluation of turbulence forecasts with the aLMo. Zelenka A: 2004.Schmutz C. Ergänzung zu Arbeitsbericht 211. CHF 11. Schmuki D. 63pp. CHF 32.Defila C. CHF 25. Dössegger R.Kürzlich erschienen Arbeitsberichte der MeteoSchweiz 214 Schmucki D. 20pp. Schmutz C. 77pp. CHF 25. 43pp. Schmutz C. 46pp.: 1999.Frei T. Aeronautical Climatological Information Geneva LSGG. CHF 31. Statistik der 24-stündigen Starkniederschläge in der Schweiz 19011996. Rohling S: 2004. Ambrosetti P. Todesfälle und Wetterlagen in Schaffhausen. 25pp. CHF 32. 40pp. Weigel A.Urfer Ch. Rohling S: 2004. 139pp. Konzept Messsysteme 2010 von MeteoSchweiz. Davos. 104pp. Schmuki D. Die extreme Sommerhitze im aussergewöhnlichen Witterungsjahr 2003.IX.10.Hächler P et al. CHF 68. Rohling S: 2004. Duding O.- 213 212 211 210 209 208 207 206 205 204 203 202 201 200 199 198 197 196 195 194 193 Frühere Veröffentlichungen und Arbeitsberichte finden sich unter www. CHF 40. CHF 25. 1999.Courvoisier HW: 1998. .

Sign up to vote on this title
UsefulNot useful