HomeMy WebLinkAbout2. C RED Nome_WIND DATA_Summary_Report July 10 20082
Harvesting the Wind Resource in
Nome, Alaska
Data Analysis Summary
July 10,2008
FOR
www.westerncommunityenergy.com
BY
SANDRA CARDON, BSME, MSATSC, EIT
2792 DESERT WIND RD
OASIS, IDAHO 83647-5020
208-908-2500
www.clever-ideas.com
3
1.1 Introduction
This report is a summary of the data analysis which led to the final turbine array layout on
Banner Peak near Nome, Alaska. The Snake River anemometer tower was crucial to the final
analysis.Originally,the assumption for the wind project was that the location would be near the
anemometer on the New Glacier Creek Road adjacent to the new powerline to the mine.
Ultimately the data proved that this would not be justified and a methodical process of evaluating
every possible location around Nome was instigated. The resulting analysis involved detailed
array analysis on Anvil and Bonanza before settling on Banner as the best overall location. No
doubt the increases particularly in utility powerline expense and site access roads at this location
are substantial.Despite the exceeded original budget, which assumed the turbines would be
located in the Snake River Valley, the strong winds and increased energy output for the
community will more than compensate the difference in the long-term.
1.2 Data Analysis Resources
Although the Snake River seemed promising as a potential build site for the year-round
accessibility, the detailed analysis on the Snake River tower data and comparison to the detailed
Wind Map provided by the Alaska Energy Authority and forced us to explore other options.
According to the wind map in Figure 1, the Snake River Valley is a Class 2 -3 wind site. The
detailed analysis confirmed that the estimated energy production in the Snake River Valley was
too low to support a commercial wind farm.
Figure 1. Wind Resource Map for the area near Nome, AK provided by the Alaska Energy Authority
4
Martina Dabo at the Alaska Energy Authority was very supportive in getting the complete data
set as quickly as possible as well as providing the AEA detailed wind map for comparisons.
1.3 Explored Locations
The wind map also indicates, however,Class 5 on Bonanza Hill ,Class 6 on the majority of
Banner Peak, and even some Class 7 on Anvil Mountain .The data available from the Anvil
Mountain tower were cross-referenced with the Snake River data to verify both locations with
the wind map and move forward in the data analysis. The Snake River data was then modeled to
provide estimated wind flow patterns over the varying terrain in the area and three alternate
locations were identified and explored.It is important to keep in mind that the Snake River data
does not fully reflect the winds over the exposed ridges since the anemometer is located in the
valley and is somewhat sheltered as a result. The prevailing wind direction in the Snake River
Valley is roughly N-NE, but a large amou nt of energy comes from the E -SE, see Figure 2.
Since the strongest winds occur on Anvil Mountain, it was explored as the first alternate site.
The anemometer on Anvil was broken, but the comparisons to the data sets that existed between
the towers were extremely helpful for both direction comparisons and speed comparisons.The
extreme weather conditions, however, would increase icing and loading of the turbine which
would result in more frequent and costly repairs and maintenance. Limited access to the site
could also prolong turbine down time, if a repair was necessary and the site was inaccessible.
Bonanza Hill and Banner Peak were then explored and compared. Although Bonanza has
favorable site accessibility and buildability, Banner has the higher wind speeds. According to
the data analysis, Banner is expected to produce 15% more energy than Bonanza Hill .
The following estimates are for a W eibull distribution using some generalizations with the
Entegrity Wind Turbine.They are based on the wind classes shown on the map.
Banner
Class 6: 7.4-8.2 m/s Annual Production average 184,223 kWh one machine
3,316,014 kWh OR 3,316 MWh annually for the wind farm
Bonanza
Class 5: 7.0-7.4 m/s Annual Production average 160,141 kWh one machine
2,882,538 kWh OR 2,882 MWh annually for the wind farm
While the foundations might have been and the roads would have been cheaper on Bonanza, the
power line extension would have been comparable in price between the two .The final turbine
wiring and utility line are shown in Figure 6 and the new access roads are shown in Figure 3.
Banner was finally chosen because of the high energy production output and comparatively
minimal construction sacrifice.The increased energy output on Banner more than justifies the
higher capital costs.
5
Figure 2. Monthly channel 1 and 8 wind roses for the Snake River tower 2007
6
Figure 3. GPS Mapping Analysis
Extensive GPS roads, trails and possible wind location analys es were performed around Nome to
compare and contrast wind turbine construction options. Ultimately the target areas were
reduced from Anvil to Bonanza, and then finally to Banner after each comparison.
Figure 4. Google Earth View of Bonanza, Banner, and Anvil at bottom Right
7
Figure 5. Original Banner Array Layout and Design
The original layout and design for the Banner Ridge was created from the topo map and the wind
data to minimize the interaction and impacts of the turbines but also keep the site wiring and
layout organized and control construction expenses.
The following Figures show how this array layout design was modified after the extensive back
and forth analysis from dealing with the actual site conditions. The topography and rocky terrain
was not conducive to the original layout and in fact had to be modified extensively and then re-
checked through the software analysis to arrange the turbines.
Six turbines were moved to the lower ridge line to the south of the Banner Peak by quite a
distance which allowed the other twelve to be fit fairly linearly along the ridge itself. Due to the
wind variability in direction, and the fact that the ridge line is in close proximity to prevailing
strong winds from the North, the turbines were kept approximately 4-10 diameters apart from
each other in most cases. Along the 60 degree line, some turbines are lined up with tighter
spacing.
In the end the final array layout was able to achieve less than 5% wake interference losses
between the turbines in total based on the data. The array layout was conducive to a linear
arrangement of the 25kV utility line which still maintains a construction efficiency, however, the
total wind farm from North to South now encompasses approximately one mile of the ridge
which is a dramatically extended project layout.
8
Figure 6. 480V turbine wiring (left) and 25kV utility line (right) on Banner Peak
Figure 3. New road access to Banner Peak
Figure 7. Existing Private Property trail access to South and two new road spurs for wind turbine access
9
1.4 Final Array Design
Creating a wind farm array on Banner was a very difficult task. The terrain, topography, road
access, and wind directions resulted in a higher capital cost for the project than planned. The
original project budget revolved around a wind farm located in Snake River Valley with much
lower construction costs, but also much lower wind speeds.
The varied wind rose along with the orientation of the ridge resulted in a balance between
capturing the wind and minimizing wake losses, while being constrained within the topography
of the hill, as shown in Figure 8.The analysis predicts that the total energy production from the
turbines range from a few as low as 140,000 kWh per year to a few as high as 170-175,000 kWh
per year with an average of 163,000 kWh per year.It estimates the total annual wind farm
output at 2,936,000 kWh.This is net after about 4.4% wake losses are taken out.Normally we
would assume a 95-97% availability and less than 5% total electrical losses depending on the
distribution grid. In this case we feel that it is difficult to estimate the actual availability,
however, with local maintenance and support this number has been consistently above 97%
across Entegrity’s fleet of installations around the world. The turbines come with a five year
complete O&M and Warranty package that requires internet connection and communications.
The analysis from the Snake River is undoubtedly missing some of the wind components that
Banner Ridge will see and some of the direction indications are no doubt biased by the local
terrain. Again, Banner should have a higher output than the data may be indicating. Also, there
is a significant amount of vertical separation between the turbines that should reduce wake
effects and interference between turbines or at least help to minimize it. In general the
predominant wind energy is from N-NE and E-SE, but the roses show enough variation that all
directions are affected. The least interference is at approximately 60 degrees so the general array
lines that are barely apparent show up on the North Array 5 (top three turbines 5-2, 5-3, 5-4),
Turbines 4-3 and 4-4 across the top of Banner Ridge, and Array 3 (Turbines 3 -1, 3-2, 3-4 {2,3,4
on this diagram}). {ALSO NOTE: that the previous 3-1 location on this diagram was moved
closer to the ridge based on foundation excavation where bedrock was too deep.}
One thing to note is that the topography from the field verification trip and the geotech
exploration required a complete re -analysis and design of the original layout.
10
Figure 8. Array Layout after Field Verification and Detailed modeling production estimates and evaluations
After final evaluation, the energy output expected from Banner Peak is more than 30%greater
than what is expected in the Snake River Valley.As a result, the increased capital costs for
siting the wind farm on Banner Peak will result in dramatically increased project output.
1.5 Conclusions
Although the Snake River Valley was not selected for the final turbine layout, the data from the
Alaska Energy Authority tower led to the final location, design, and implementation of the wind
farm on Banner Peak.
The actual data confirmed the wind map classifications and helps increase confidence in the
regional assessments. For smaller village projects with fewer turbines it may be possible to use
wind map energy predictions for minimum output estimates as is occurring in Oregon and Texas
projects with the Entegrity turbines. This can avoid the delays and expenses associated with
such detailed energy analysis and anemometer expenses and time requirements for long term
data collection.
After taking the best data available and modeling several locations, we are confident that the
chosen location is optimal for this wind farm.In fact, we are confident that this is the best
location with a proximity to Nome.Although the capital costs were higher than planned, the
output should be as well resulting in increased green energy production for the community of
Nome, Alaska.
11
12
13