Sónia Liléo, Managing Director / Kjeller Vindteknikk AB

Development of an Icing Atlas for Finland
Øyvind Byrkjedal
Sónia Liléo
[email protected]
[email protected]
Vaasa, 2015-03-17
Outline
 Short presentation of KVT
 Presentation of KVTs Wind and Icing maps for
Finland
 How have the maps been generated?
 Applications of our icing model, including model
validation
Kjeller Vindteknikk, KVT

Established 1998

Ownership:
Employees:
75%
Board members: 25%

25 employees

Main office at the Kjeller Campus,
20 km from Oslo. Office also in Stavanger

Subsidiary in Sweden with offices in
Stockholm and Gothenburg

Main markets are Finland, Sweden and
Norway
Wind measurements
Wind mapping
Independent provider of wind
measurements and resource assessment
Long experience and high expertise from
cold climate sites
Energy assessment
Main services
Measurements
 Wind measurements
 Icing measurements
Analysis
• Wind power
o Pre-construction
o Post-construction
• Wind Engineering
Research and Development
• Participation in national and
international research projects
Icing Map for Finland
 Long-term (20 y) average
number of annual icing
hours at 140 m above
ground
 Based on mesoscale
simulations using the WRF
model
 Horizontal resolution of the
map is 50 m x 50 m
 Map available for free in pdf
format
Wind Map for Finland
 Expected long-term average
(20 y) wind speed at 140 m
height
 Based on mesoscale
simulations using the WRF
model
 Horizontal resolution of
1 km x 1 km
 Map available for free in pdf
format
Methodology - Mesoscale model

Based on the mesoscale model WRF – Weather Research and Forecasting
 The model performs calculations in the time domain giving as output a
time series with 1 hour temporal resolution
 Improved microphysics (Thompson scheme) that
gives a state-of-the-art description of ice formation

Model resolution of 1 km x 1 km

Height adjustment of the modeled parameters
based on topography data with 50 m x 50 m
resolution, in order to more accurately
describe the terrain elevation
Meteorological icing and instrumental icing
IEA Wind R,D&D Task 19
Purpose of the project: Gather and provide information about wind
energy in cold climates
”Best practices for wind energy in cold climates – Resource
assessment and site classification” http://arcticwind.vtt.fi/
Calculation of meteorological icing
 ISO 12494 standard describes the formation of
atmospheric icing on structures
 Ice loads are important in the design of masts,
power lines, and other structures in cold climate
Ålvikfjellet, Hardanger, Norway, January 2014
Photo: Ole Gustav Berg, Statnett
Calculation of meteorological icing
 Icing intensity calculated
according to ISO 12494:
dM
  1 2 3  w  A  V
dt
α1- collision efficiency, α1=f(V,d,D)
α2- sticking efficiency, α2 ≈ 1
α3- accretion efficiency, α3= f(V,d,w,T,e,D,α1)
w – cloud liquid water content
A – collision area, perpendicular to flow
V – Wind speed
Calculation of meteorological icing
The parameter shown
in the icing map is
the “annual number
of icing hours”
Annual number of
hours with
dM/dt > 10 g/hr
dM
  1 2 3  w  A  V
dt
Height dependence of icing
 For each grid point we calculate the number of icing hours.
 The number of icing
hours are strongly
dependent on the
height above sea level
 The figure shows:
 icing hours from
the model
 Icing hours
calculated from
METAR (airports)
observations
Applied height adjustment
We do a height correction of the 1 km x 1 km modeled parameters
based on topography data with 50 m x 50 m resolution, in order to
more accurately describe the terrain elevation, and avoid
underestimation of the icing on hills
Snapshot of icingrate: [g/hr]
Applied height adjustment
The terrain elevation is smoothed in the coarse model resolution
Terrain elevation with
1 km x 1km resolution
Terrain elevation with
50 m x 50 m resolution
14
Applied height adjustment
Forces the lifting of the air masses from the coarse
smoothed terrain to the height given in the 50 m x 50 m
terrain model
WRF model:
1 km x 1km resolution
WRF model:
With height adjustment
50 m x 50 m
Icing Map for Finland
 Freely available in pdf
format. The icing maps KVT
developed for Norway and
Sweden are also freely
available in pdf format
 Maps available for purchase
in GIS format
 Long-term time series of
wind, temperature, icing
and other parameters are
also available
Wind Map for Finland
 Freely available in pdf
format
 Maps available for purchase
in GIS format
 Long-term time series of
wind speed and direction
can be used in the long-term
correction of wind
measurements
Applications of the icing model
Structure design – Loads modeling
Design of met and telecommunication masts
Design of power lines
IceLoss - Estimate of the expected icing
losses of a wind farm


Icing will change the aerodynamic
properties of the turbine blade
Result in reduced power output
Foto: Kent Larsson, abvee
Icing influence on turbine blades
May 2010
November 2009
60% production loss due to icing
during this specific month
Estimate of the expected icing losses of a wind
farm
IEA wind task 19 has defined 5 icing classes*
Translates meteorological icing into production loss
*Best practices for wind energy in cold
climates – Resource assessment and site
classification - http://arcticwind.vtt.fi/
icing map
Power [% of rated]
IceLoss – Model developed by KVT to estimate
the expected icing losses
IceLoss validation – wind farm A
Weekly production losses:
Observed (black)
Modeled (red)
IceLoss validation – wind farm B
Weekly production losses:
Observed (black)
Modeled (red)
IceLoss validation
Modeled icing loss using IceLoss
30,0 %
Modeled vs Observed Annual Icing Loss
25,0 %
20,0 %
15,0 %
10,0 %
5,0 %
0,0 %
0,0 %
5,0 %
10,0 %
15,0 %
20,0 %
Observed icing loss
25,0 %
30,0 %
IceRisk - Estimate of risk zones for ice
throw from wind turbines
Foto: Kent Larsson, abvee
IceRisk
Estimate of risk zones for ice throw from turbines
 IceRisk gives a map showing
the occurrence probability
of ice throw at different
distances from the turbines
 IceRisk is based on local
time series of icing from the
model calculations
 Calculations with a
trajectory model
 IceRisk takes into account
the local topography
IceForecast – Short-term forecast of
expected energy production, including icing
loss
Foto: Kent Larsson, abvee
IceForecast
Production forecast including forecast of icing losses
 On-going icing forecast for several wind farms in Sweden
Summary
 A new wind and icing maps for Finland have been
developed




140 m above ground level
Based on 1 km x 1 km mesoscale calculations
Icing atlas presented at 50 m x 50 m horizontal resolution
Uses a state-of-the-art description of the microphysics associated to icing
 Model validation based on experiences from Sweden
 Icing calculations have been verified against icing observations
 The model describes the timing of icing quite well
 Good agreement between modeled and observed icing loss for sites with
low/moderate icing losses
 Applications of the icing model




Structure design – Ice loads modeling
IceLoss - A tool to estimate the production losses in a wind farm
IceRisk – A tool to estimate the risk for ice throw
IceForecast – A tool for short-term energy forecasting incl icing losses
Thank you for your attention
Sónia Liléo
[email protected]
Øyvind Byrkjedal
[email protected]