HomeMy WebLinkAboutLevelock Wind Reconnaissance Study Wind Resource Assessment Report - Jan 2017 - REF Grant 7060911Levelock, Alaska Wind Resource
Assessment Report
Levelock met tower, photo by Douglas Vaught
January 25, 2017
Douglas Vaught, P.E.
V3 Energy, LLC
Anchorage, Alaska
Levelock, Alaska Wind Resource Assessment Report P a g e | 1
Contents
Summary.......................................................................................................................................................2
Levelock Met Tower......................................................................................................................................2
Test Site Location......................................................................................................................................2
Met Tower Sensors...................................................................................................................................5
Data Quality Control .................................................................................................................................6
Wind Speed...............................................................................................................................................7
Time Series................................................................................................................................................7
Long-term Wind Speed Comparison.....................................................................................................8
Probability Distribution Function..............................................................................................................9
Wind Shear and Roughness....................................................................................................................10
Extreme Winds........................................................................................................................................11
Periodic Maxima .................................................................................................................................11
Method of Independent Storms.........................................................................................................11
EWTS II ................................................................................................................................................12
Summary.............................................................................................................................................12
Temperature, Density, and Relative Humidity........................................................................................13
Wind Direction........................................................................................................................................14
Turbulence..............................................................................................................................................14
WAsP Wind Flow Model .............................................................................................................................15
Wind Turbine Energy Production............................................................................................................17
Recommendations......................................................................................................................................17
Levelock, Alaska Wind Resource Assessment Report P a g e | 2
Summary
This report documents measurement of the wind resource in the community of Levelock in the Bristol
Bay Region of Alaska. V3 Energy, LLC was contracted by Lake and Peninsular Borough to assist with site
selection, install a 34 meter meteorological (met) test tower, review data periodically, compare the met
tower data to a nearby reference station, create a wind flow model of surrounding area, and
recommend a wind turbine for estimate energy production. This report is a summary and compilation
of those objectives.
Levelock has a moderate Class 3 wind resource which offers potential for wind power development in a
battery storage configuration with a grid-forming converter. Note that although the more distant
Levelock Hill may be windier, from the perspective of development costs, a near-village site is likely the
best option for wind power in the community.
Levelock Met Tower
A 34-meter high met tower was installed on the south side of the village near a GCI communication
tower in July, 2014. The met tower was equipped with an NRG Symphonie data logger, three
anemometers, a wind vane, and temperature sensor. To enhance visibility from the air, the met tower
was equipped with orange high-visibility marker balls near the top and was painted alternating bands of
red and white.
Data dates 7/22/2014 to 3/16/2016 (20 months with 3 months missing,
9/21/2015 to 12/16/2015)
Wind speed, mean, 34 m, annual 5.65 m/s (12.6 mph)
Wind power density, mean, 34 m 246 W/m
2
Max. 10-minute average wind speed 25.8 m/s (57.7 mph)
Maximum 2-second wind gust 37.9 m/s (84.8 mph), December, 2015
Weibull distribution parameters k = 1.91, c = 6.48 m/s
Wind shear power law exponent 0.398
Surface roughness 2.11 m
IEC 1 61400-1, 3rd ed. classification Class III-B
Turbulence intensity, mean 0.13 (at 15 m/s)
Test Site Location
The met tower site is immediately south of the southeast-northwest trending road on the southern edge
of the community. This road serves a very high GCI cellular communications tower and the community
landfill. The site itself is just off an ATV trail about 450 ft. due south of the GCI tower. The brushy
nature of the site is not ideal for wind measurement and this was not the first choice of location for the
met tower. A preferable location was open ground nearer the river, about 950 ft. to the southeast. FAA
1 International Electrotechnical Commission, a Swiss-based organization that prepares and publishes international
standards for electrical, electronic, and related technologies.
Levelock, Alaska Wind Resource Assessment Report P a g e | 3
though would have required obstruction lighting of the met tower, which was not within project budget.
A disadvantage of the near-river site, however, is its lower elevation.
Site number (logger ID) 7271
Latitude/longitude N 61° 15’ 30.9”, W 149° 31’ 53.7”
Time offset -9 hours from UTC (Yukon/Alaska time zone)
Site elevation 1,173 meters (3,850 ft.)
Datalogger type NRG SymphoniePLUS3, 10-minute averaging time step
Tower type Tubular, ~15 cm (6 in.) diameter, 10-meter height
Tower base, brushy site vegetation Data logger home screen
Met tower and site Data logger wiring panel
Levelock, Alaska Wind Resource Assessment Report P a g e | 4
Levelock, Alaska Wind Resource Assessment Report P a g e | 5
Met Tower Sensors
The Levelock met tower was equipped with a standard sensor package supplied by Renewable NRG
Systems, Inc. of Hinesburg, Vermont, USA. Heated sensors and sensors requiring supply power were not
included. This enabled nearly maintenance-free operations, except for periodic battery changes and
data card swaps.
Logger
Channel Sensor type
Sensor
Designation
Serial
Number
Height
AGL Multiplier Offset 2 Orientation 3
1 NRG #40C
anemometer
34 m A 231806 34.2 m 0.762 0.33 045 T
2 NRG #40C
anemometer
34 m B 231807 33.7 m 0.755 0.39 175 T
3 NRG #40C
anemometer
20 m 231808 20.9 m 0.760 0.33 050 T
7 NRG #200P wind
vane
Direction n/a 9.0 m 0.351 000 000 T
9 NRG #110S Temp C Temp n/a 2 m 0.136 -86.38 000 T
North side East side
South side West side
2 Multiplier and offset are variables of the straight-line transfer function to calculate sensor value from its raw data
inputs with the equation y = mx + b where y is sensor value, m is multiplier and b is offset.
3 In true degrees (reference pt. is Earth’s geographic North Pole)
Levelock, Alaska Wind Resource Assessment Report P a g e | 6
Data Quality Control
Data was filtered to remove presumed icing events that yield false zero wind speed data and non-variant
wind direction data. Typically, data that meets the criteria listed below is automatically filtered:
Anemometer icing – data filtered if temperature < 1°C, speed SD = 0, and speed changes < 0.25
m/s for minimum 2 hours
Vane icing – data filtered if temperature < 1°C and vane SD = 0 for minimum of 2 hours
Tower shading of 34-meter A and B paired anemometers – data filtered when winds from ± 15°
of behind tower
In addition to automatic data filtering, icing data was manually filtered in situations where the automatic
filtering failed to detect obvious icing conditions, or where automatic filtering flagged data as icing when
it did not appear to be so. A relative humidity sensor would have aided the detection of icing
conditions, but the met tower was not equipped with one.
Time Steps Flagged As
Data Column
Possible
Records Missing
<Unflagged
data>Icing Invalid
Tower
shading
Speed 34 m A 86,817 12,612 67,425 2,768 75 4,156
Speed 34 m B 86,817 12,612 58,841 2,788 75 13,012
Speed 20 m 86,817 12,612 71,421 2,851 75 0
Direction 34 m 86,817 12,612 69,107 5,098 72 0
Temperature 86,817 12,612 74,205 0 72 0
0%
5%
10%
15%
20%
25%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Met Tower Sensor Icing Loss
Average of 34 m A Average of 34 m B Average of 20 m Average of Vane
Levelock, Alaska Wind Resource Assessment Report P a g e | 7
Wind Speed
Anemometer data obtained from the Levelock met tower, from the perspectives of both mean wind
speed and mean wind power density, indicate a moderate wind resource. The table below includes
wind filtered (for icing and tower shadow) and raw wind data.
Speed 34 m A Speed 34 m B Speed 20 m
filtered raw filtered raw filtered raw
Mean measured wind speed (m/s) 5.76 5.51 5.60 5.37 4.70 4.54
Mean annual wind speed (m/s) 5.66 5.41 5.63 5.26 4.64 4.45
Mean annual wind speed (mph) 12.7 12.1 12.6 11.8 10.4 10.0
Max 10-min wind speed (m/s) 25.9 25.9 25.7 25.7 22.3 22.3
Max gust wind speed (m/s) 37.9 36.0 33.2
Max gust wind speed (mph) 84.8 80.5 74.3
Weibull k 1.92 1.70 1.81 1.70 1.82 1.63
Weibull c (m/s) 6.48 6.12 6.29 5.98 5.27 5.03
Mean annual power density (W/m²) 235 220 246 207 136 129
Mean annual energy content (kWh/m²/yr) 2,059 1,924 2,151 1,815 1,192 1,126
Energy pattern factor 4 2.02 2.17 2.17 2.21 2.12 2.27
Frequency of calms (%) (< 4 m/s speed) 31.2 34.4 34.5 35.8 45.0 47.1
Time Series
Time series calculations indicate higher wind speeds during the winter months compared to the summer
months, which is an expected seasonal variation. The daily wind profile (annual basis) indicates
relatively even wind speeds throughout the day with slightly higher wind speeds during late afternoon
hours, which is an expected diurnal variation.
Mean Max Gust
Std.
Dev.
Weibull
k Weibull c
Month (m/s)(m/s)(m/s)(m/s)(-)(m/s)
Jan 7.37 21.6 29.3 3.32 2.27 8.25
Feb 6.80 21.6 29.3 3.41 2.05 7.65
Mar 6.06 20.0 26.9 3.15 1.94 6.79
Apr 5.64 17.3 23.7 2.64 2.25 6.37
May 4.23 13.5 19.1 2.41 1.81 4.75
Jun 4.58 11.7 15.5 2.04 2.38 5.16
Jul 4.25 15.6 20.0 2.24 1.95 4.78
Aug 4.81 13.1 17.6 2.26 2.23 5.42
Sep 4.71 13.3 18.3 2.35 2.08 5.30
4 With an assumption of constant air density, energy pattern factor is the ratio of the actual mean wind power density
to the wind power density calculated from only the mean wind speed. Definition obtained from Windographer 4.0
Help file.
Levelock, Alaska Wind Resource Assessment Report P a g e | 8
Mean Max Gust
Std.
Dev.
Weibull
k Weibull c
Month (m/s)(m/s)(m/s)(m/s)(-)(m/s)
Oct 5.38 14.0 20.0 2.39 2.35 6.04
Nov 7.14 20.2 28.1 3.77 1.97 8.04
Dec 6.88 25.9 37.9 3.67 1.96 7.76
Annual 5.63 25.9 37.9 3.09 1.91 6.42
Long-term Wind Speed Comparison
Comparing the 17 months of measured wind speed data at the Levelock met tower to a long-term
reference source is difficult in that the Levelock airport is not equipped with an automatic surface
observing system (ASOS) or automatic weather observing system (AWOS). Nearest stations are King
Levelock, Alaska Wind Resource Assessment Report P a g e | 9
Salmon, Dillingham, and Igiugig, all of which are rather distant. Nevertheless, regional wind patterns
have broad impact and hence comparing Levelock to a reference station is valuable.
The King Salmon Airport weather station indicates declining wind speeds from 1990 through 2016 but
possibly a view earlier in time may indicate that 1990 to 1994 was windier than normal, in which case
the declining trend would be less pronounced or not exist. Note that the operational years of the
Levelock met tower – 2014 and 2015 mostly – were windier than the preceding 20 years.
Probability Distribution Function
The probability distribution function (PDF), or histogram, of the Levelock met tower site wind speed
indicates a shape curve reflective of a standard wind profile, known as the Rayleigh distribution
5 where
Weibull k = 2.0. The Rayleigh distribution is defined as the default wind probability distribution for wind
power analysis.
Weibull Proportion Power R
Weibull c Mean Above Density Squared
Algorithm k (m/s)(m/s)5.803 m/s (W/m2)(-)
Maximum likelihood 1.92 6.48 5.75 0.450 232 0.979
Least squares 1.90 6.54 5.81 0.456 241 0.976
WAsP 1.88 6.47 5.74 0.448 237 0.975
Actual data 5.80 0.389 397
5 One form of the Weibull distribution is the Rayleigh, which occurs when k = 2.0. This is equivalent to a standard
deviation of 52% of the mean wind speed. Wind turbine manufacturers provide standard performance data based
on a wind resource with a Rayleigh distribution. Definition from www.wind-power-program.com and
www.windpower.org.
0
1
2
3
4
5
6
King Salmon Airport
Levelock, Alaska Wind Resource Assessment Report P a g e | 10
Wind Shear and Roughness
Wind shear 6 and roughness at the Levelock met tower site is quite high with a calculated power law
exponent (shear) value of 0.398, using data from the northeast-facing 34m A and 20m anemometers.
The calculated surface roughness is 2.15 meters. Both the power law exponent and the surface
roughness length are too high; a 0.20 to 0.25 power law exponent and 0.2 to 0.5 m surface roughness
length are more reasonable. The high value can be explained by only two anemometer measurement
heights, which can in locations with heavy brush lead to excessive variation in wind speed
measurements between the two heights. Brush dramatically slows surface winds compared to higher
wind speeds above. This results in a wind shear calculation that is not representative of higher
elevations above the top-most anemometer and/or more open and less brushy locations.
6 Change in wind velocity with height above ground level.
Levelock, Alaska Wind Resource Assessment Report P a g e | 11
Extreme Winds
International Electrotechnical Commission (IEC) 61400-1, 3rd edition extreme wind probability
classification is one criteria – with turbulence the other – that describes a site with respect to suitability
for wind turbine models. Extreme wind is described by the 50 year (10-minute average) Vref; in other
words, the 10-minute average wind speed predicted to occur once every 50 years.
IEC 61400-1, 3rd ed.
Class Vref, m/s
I 50
II 42.5
III 37.5
S designer-
specified
Periodic Maxima
One method to estimate Vref is a Gumbel distribution analysis modified for monthly maximum winds
versus annual maximum winds, which are typically used for this calculation. Seventeen months of data
though are minimal for this method. Nevertheless, with data available the predicted Vref in a 50-year
return period (in other words, predicted to occur once every 50 years) by this method is 32.9 m/s at 34
meters above ground level. This result meets IEC 3
rd edition Class III criteria, the lowest defined.
Method of Independent Storms
A second technique, Method of Independent Storms, yields a Vref estimate of 33.2 m/s, near that
predicted by the periodic maxima method.
Levelock, Alaska Wind Resource Assessment Report P a g e | 12
EWTS II
A third method, known as EWTS II (European Wind Turbine Standards II), ignores recorded peak wind
speeds and calculates Vref from the Weibull k factor. There are three variations of this method – Exact,
Gumbel and Davenport – which yields a Vref between 25.6 and 28.1 m/s. This is lower than the other
methods and well within IEC 3
rd edition Class III extreme wind criteria.
Summary
The calculated Vref wind speeds by the three methods described above all meet IEC 61400-1, 3rd edition
criteria for Class III wind classification, which has a Vref limit of 37.5 m/s. The practical importance is that
any wind turbine on the market is suitable for Levelock, based on extreme wind behavior.
Vref (50 yr)
Method (m/s)
Periodic Maxima 32.9
Method of Independent Storms 33.2
EWTS II (Exact)25.6
EWTS II (Gumbel)25.9
EWTS II (Davenport)28.1
Levelock, Alaska Wind Resource Assessment Report P a g e | 13
Temperature,Density,and Relative Humidity
Levelock experiences relatively warm summers and not especially cold winters by interior Alaska
standards, although note that the two winters of the met tower measurement period were warmer than
typical. Calculated mean air density during the met tower test period exceeds the 1.225 kg/m
3 standard
air density for a sea level elevation by 2.9 percent. This is advantageous in wind power operations as
wind turbines produce more power at low temperatures (high air density) than at standard temperature
and density.
Month
Temp (°C)Temp (°F)Mean
DensityMeanMinMaxMeanMinMax
(°C)(°C)(°C)(°F)(°F)(°F)(kg/m3)
Jan -4.6 -24.6 6.6 23.7 -12.3 43.9 1.311
Feb -2.3 -22.7 12.3 27.9 -8.9 54.1 1.299
Mar -1.6 -24.2 11.5 29.1 -11.6 52.7 1.294
Apr 1.7 -8.8 15.3 35.1 16.2 59.5 1.279
May 8.2 0.2 22.3 46.8 32.4 72.1 1.249
Jun 13.5 3.6 31.6 56.3 38.5 88.9 1.225
Jul 13.8 3.0 27.2 56.8 37.4 81.0 1.223
Aug 13.1 0.1 26.5 55.6 32.2 79.7 1.227
Sep 9.0 -4.1 20.3 48.2 24.6 68.5 1.241
Oct -0.7 -13.2 13.2 30.7 8.2 55.8 1.255
Nov 0.8 -16.0 8.9 33.4 3.2 48.0 1.251
Dec -2.9 -19.0 6.7 26.8 -2.2 44.1 1.281
Annual 4.0 -24.6 31.6 39.2 -12.3 88.9 1.261
Levelock, Alaska Wind Resource Assessment Report P a g e | 14
Wind Direction
Wind rose data indicates that winds in Levelock are primarily northerly, northeasterly and to a lesser
extent southwesterly. The energy component of the wind rose indicates that the power-producing
winds are strongly northeasterly to easterly, and less often, northerly.
Turbulence
The turbulence intensity (TI) at the Levelock met tower site at 34 meters above ground level is
somewhat high with a mean turbulence intensity of 0.134 and a representative turbulence intensity of
0.161 at 15 m/s wind speed, indicating moderately rough air for wind turbine operations. This classifies
as International Electrotechnical Commission (IEC) 61400-1, 3
rd Edition (2005) turbulence category B
(Category A is most turbulent; C is least). Note however that the brush surrounding the met tower site
has a significantly deleterious impact on turbulence. It is very likely that turbulence in a more open
location, such as tundra or along the river, or appreciably higher than 34 meters at the met tower site,
would be much lower.
Levelock, Alaska Wind Resource Assessment Report P a g e | 15
IEC 61400-1, 3rd ed.
Category Iref, m/s
A 0.16
B 0.14
C 0.12
S designer-specified
WAsP Wind Flow Model
Wind flow modeling was accomplished with WAsP (Wind Atlas Analysis and Application Program), a
Danish PC-based software for predicting wind climates, wind resources and power production from
wind turbines and wind farms and can be used to predict wind turbine performance. WAsP is the most
widely used wind power analysis software in the world. Modelling begins with a digital elevation map
IEC 3 ed.
Height Data Mean SD Repres.Peak Data Mean SD Repres.Turbulence
Wind Speed Sensor (m) Points TI of TI TI TI Points TI of TI TI Category
Speed 34 m A 34.2 67,425 0.17 0.10 0.30 1.61 348 0.13 0.02 0.16 B
Speed 34 m B 33.7 58,841 0.18 0.11 0.31 1.27 316 0.13 0.02 0.16 B
Speed 20 m 20.9 71,421 0.23 0.12 0.38 1.49 148 0.17 0.02 0.19 S
All Speed Bins 15 m/s Speed Bin
Levelock, Alaska Wind Resource Assessment Report P a g e | 16
(DEM) of the wind farm site and surrounding area and conversion of coordinates to Universal Transverse
Mercator (UTM).7
A wind data (or wind atlas) reference point is added to the digital elevation map, wind turbine locations
identified, and a wind turbine selected to perform annual energy production calculations. WAsP
considers the orographic (terrain) effects on the wind, plus surface roughness and obstacles, and
calculates wind velocity increase or decrease at all nodes of the map. The mathematical model has
several limitations, including the assumption that the overall wind regime of the turbine site is same as
the met tower reference site, prevailing weather conditions are stable over time, and surrounding
terrain at the wind data reference point and turbine sites is sufficiently gentle and smooth to ensure
laminar, attached wind flow.
For this analysis, a simplified WAsP model without a roughness map was employed. This will result in
moderate error when projecting wind speed from the brushy terrain where the met tower had been
located to more open terrain where wind shear and roughness length would be less. But, the purpose
here is to illustrate possible site options should alternate sites be considered. One such alternate site is
an exposed plateau-like, tundra-covered location referred to as Levelock Hill and/or Levelock Site 2 in
the site investigation phase of the project. Levelock Hill is approximately 2.1 miles northwest of the
7 UTM is a geographic coordinate system that uses a two-dimensional Cartesian coordinate system to identify
locations on the surface of Earth. UTM coordinates reference the meridian of its zone (60 longitudinal zones are
further subdivided by 20 latitude bands) for the easting coordinate, and distance from the equator for the northing
coordinate. Units are meters.
8 Projected 37 meter level, 5.5 (blue color) to 6.4 m/s (red color) mean annual
Levelock, Alaska Wind Resource Assessment Report P a g e | 17
north threshold of Levelock Airport, but nearer 2.7 miles by ATV trail. The site is on the winter trail to
New Stuyahok, is well exposed, and is quite suitable for a wind power site except for its distance from
existing power distribution infrastructure. This distance would be a severe challenge as distribution line
construction across undeveloped permafrost terrain is very expensive.
Wind Turbine Energy Production
Met tower data was projected to 37 meters – the hub height of the Northern Power NPS100C-24 wind
turbine – for creation of the preceding map. The NPS100C-24 is a possible wind turbine option for
Levelock. It is manufactured in Barre, Vermont and is in wide use in Alaska for village wind projects.
At the met tower site, WAsP software predicts that at 100% annual energy production (AEP), the
NPS100C-24 turbine would generate 271 MWh/year of electricity. This equates to a capacity factor
9 of
32.5%. At a more realistic 80% net AEP (assuming 20% production loss due to electrical, icing,
maintenance, curtailment, wake, etc.), a NPS100C-24 would generate 217 MWh/year of energy at a
26.0% capacity factor.
At Levelock Hill (or Site 2), with the NPS100C-24 wind turbine at 37 meters hub height, WAsP predicts
297 MWh/year energy production (35.6% capacity factor) at 100% AEP, or 238 MWh/year (28.4%
capacity factor) at 80% AEP.
Recommendations
Although Levelock’s wind resource is moderate, there are opportunities to develop wind power in the
community. The challenge though is to avoid the standard configuration where wind turbine(s) operate
in parallel with the diesel generator(s). Levelock’s electric load is too low for this type configuration and
a design of this nature would be difficult to control and would not be efficient. A preferable approach is
to consider a battery storage option with a grid-forming converter. In this configuration, the converter
controls grid voltage and frequency as the prime mover and the diesel generator operates in droop
mode. Wind turbines and possibly other renewable generation such as photovoltaic panels operate in
parallel when available with excess energy directed to the battery and/or a remote thermal node such
as the school’s hydronic heat system.
Evaluation of the design aspects and economic merits of this approach are beyond the scope of this
report, but recent developments in lithium ion batteries and inverter technology are promising and
warrant further investigation.
9 Capacity factor is percent annual energy generated, or predicted, compared to maximum possible (100% power
output 100% of the time)