COMO (REINECKER RIDGE)
7/16/2011 to 9/26/2012
LOCATION DETAILS |
Latitude: |
N 39° 16.336’ or N 39° 16’ 20.16" |
Longitude: |
W 105° 54.656’ or W 105° 54’ 39.36" |
Survey Meridian: |
Colorado, Sixth Principal Meridian |
Township: |
9 S |
Range: |
76 W |
Section: |
17 |
Elevation: |
10,289 feet (3,136 m) |
Datum: |
WGS 84 |
Tower Type: |
Existing ham radio tower attached to house |
Tower Height: |
33.5 ft. (10.2 m) |
Vane Offset (deg): |
+142° |
Direction Basis: |
True North |
Mag. Declination: |
9° 20' E, changing by 7' W/yr |
Wind Explorer S/N: |
0750 |
Site No.: |
3905 |
CSU ALP Install Team (from left):Sarah Bass, Ben Ebersole, Scott Little, Ryan Sellden, Mark Goudreault, and Mike Kostrzewa (taking picture).
DATA DETAILS
July 16, 2011 to September 26, 2012:
The anemometer tower was installed on July 16, 2011 and removed on September 26, 2012. The site was located near the northern extent at the top of Reinecker Ridge SW of Como in Park County. The original plan was to install a 20m tower in an opening on the side of a hill near to the lessee's cabin. However, suitable anchorage could not be found in the 2-3 feet of soil that existing over the granite bedrock. The lessee had a 21-1/2 foot tall 2" galvanized steel pole mounted on the deck of the cabin that was previously used for a ham radio installation so the anemometer and a wind vane were mounted on top on the pole, which itself was 12 feet above the ground. The pole intersected the pitched roof line at a height of 10 feet, 8 inches above the deck and was horizontally 7 feet from the roof peak. The peak of the roof was 14 feet, 8 inches above the deck, so the sensors were mounted about 7 feet above the roof peak. This location was considered to be a great site measure the turbulence associated with a near-building turbine installation.
All data was collected using an NRG #40 Calibrated Anemometer and NRG #200 Wind Vane. The certification for the anemometer is as follows:
NRG #40C Calibrated Anemometer |
Model No. |
1900 |
Serial No. |
1795-00087805 |
Calibration Date |
10/31/08 4:27 p.m. |
Slope |
0.761 m/s per Hz |
Offset |
0.35 m/s |
In addition, an NRG #110 uncalibrated temperature sensor with a radiation shield was installed about 5 feet above the deck in the pole. This equipment fed into an NRG Wind Explorer data logger. All data plugs were sent to the Colorado ALP at Colorado State University for analysis. The data plug files and text versions of these files are given below.
It is important to note that these are the raw files without any compensation for offset.
Note: When onsite to replace the plug and after examining the wind direction data from the data plug, wind vane appeared to be wired incorrectly - the wind direction was just too consistent. The wind vane was rewired at the datalogger and appears to be recording correctly. Because the reliability of the wind direction cannot be considered reliable, the wind vane data was removed from 7/16/2011 at 12:50 through 9/7/2011 at 15:00.
Using this data, an analysis of the wind resource report was developed using Windographer 2.4.6. For this data an offset of +142° was applied to the wind vane data. This data was flagged for icing in two ways:
- Any wind speed data where the wind speed was less than 0.5 m/s at a temperature less than 0°C for 3 hours or more was flagged and not included in the wind resource analysis calculations
- Any wind direction data where the wind direction varied by less than 3 degrees at a temperature less than 0°C for 3 hours or more was flagged and not included in the wind resource analysis calculations
A summary report, the combined data files (with and without the validation analysis), and the Windographer files (with and without the validation analysis) are given below::
Final Wind Resource Summary
Highlights of the final wind resource analysis are shown below:
Data Properties |
Variable |
Data Set Starts: |
7/16/2011 12:50 MST |
Height above ground (m) |
10.2 |
Data Set Ends: |
9/26/2012 19:00 |
10-min. mean wind speed (m/s) |
2.459 |
Data Set Duration: |
14 months |
10-min median wind speed (m/s) |
1.900 |
Length of Time Step: |
10 minutes |
10-min min. wind speed (m/s) |
0.350 |
Elevation: |
10,289 ft (3,136 m) |
10-min max wind speed (m/s) |
15.90 |
Mean air density (kg/m³): |
0.872 |
10-min standard deviation (m/s) |
1.952 |
Wind Power Coefficients |
Weibull k |
1.321 |
Power Density at 50m: |
81 W/m² |
Weibull c (m/s) |
2.674 |
Wind Power Class: |
1 (Poor) |
Mean power density (W/m²) |
29 |
Wind Shear Coefficients |
Mean energy content (kWh/m²/yr) |
251 |
Power Law Exponent: |
0.237 |
Mean turbulence intensity |
0.436 |
Surface Roughness: |
0.25 m |
Energy pattern factor |
3.701 |
Roughness Class: |
2.76 |
Total data elements |
252,436 |
Roughness Description: |
|
Flagged wind speed data elements |
211 |
Flagged direction data elements |
1,551 |
Missing data elements |
7,665 |
Data recovery rate (%) |
97.0% |
Note: The wind power density and wind power class at 50m are projections of the data from 20m. A surface roughness of 0.25 meters was assumed for this projection. This is the surface roughness for an area with many trees. This value was then used this to calculate the roughness class and the power law exponent shown above. |
Vertical Wind Shear, Height (m) vs Mean Wind Speed (m/s)
|
Wind Frequency Rose at 10.2 meters
|
Wind Energy Rose at 10.2 meters
|
Daily Wind Speed Profile, Hourly Mean Wind Speed (m/s) vs. Hour of the Day
|
Seasonal Wind Speed Profile, Monthly Mean Wind Speed (m/s) vs. Month
|
Probability Distribution Function at 10.2 m: Frequency (%) vs. Wind Speed (m/s) |
Windographer was used to match up the wind at this site with the performance curves of some common turbines of various sizes and various heights. The table below shows the results. For the larger turbines, the tower height was increased to account for the larger turbine blades - the wind resource was extrapolated to these higher heights. Keep in mind that the larger and the higher the turbine, the better the wind and the greater the output. But of course, as the tower heights and turbine sizes increase so does the cost.
Keep in mind too that listing a particular turbine doesn't imply an endorsement - not does it imply that installing a particular turbine model is feasible or recommended for a particular site. For consistency, the larger turbines are included even at sites that where they may not be practical so that one can compare the relative production of different sites.
Turbine |
Rotor
Diameter
meters |
Rotor
Power
kW |
Hub
Height
meters |
Hub
Height
Wind
Speed
m/s |
Time
At
Zero
Output
percent |
Time
At
Rated
Output
percent |
Average
Net
Power
Output
kW |
Average
Net
Energy
Output
kWh/yr |
Average
Net
Capacity
Factor
% |
Bergey Excel-R |
|
|
10.2 |
2.63 |
72.4 |
0.2 |
0.3 |
2,300 |
3.5 |
Bergey Excel-S |
|
|
10.2 |
2.63 |
54.7 |
0.1 |
0.3 |
2,700 |
3.0 |
Bergey XL.1 |
|
|
10.2 |
2.63 |
27.1 |
0.2 |
0.0 |
400 |
4.4 |
Southwest AIR X |
|
|
10.2 |
2.63 |
74.2 |
0.0 |
0.0 |
44 |
1.3 |
Southwest Skystream 3.7 |
|
|
10.2 |
2.63 |
67.7 |
0.0 |
0.1 |
700 |
4.7 |
Southwest Whisper 500 |
|
|
10.2 |
2.63 |
72.3 |
0.2 |
0.2 |
1,300 |
5.0 |
Northern Power NW 100/21 |
|
|
37 |
3.57 |
53.9 |
0.0 |
7.0 |
61,000 |
7.0 |
GE 1.5s |
|
|
64.7 |
4.07 |
60.3 |
1.1 |
108.4 |
949,400 |
7.2 |
GE 2.5xl |
|
|
75 |
4.21 |
51.3 |
1.3 |
266.4 |
2,333,600 |
10.7 |
GE 3.0s |
|
|
70 |
4.15 |
61.2 |
0.4 |
185.8 |
1,627,500 |
6.2 |
Vestas V90 - 1.8 MW |
|
|
80 |
4.28 |
50.7 |
1.5 |
224.0 |
1,962,000 |
12.4 |
Vestas V90 - 2.0 MW |
|
|
80 |
4.28 |
50.8 |
1.4 |
234.2 |
2,051,200 |
11.7 |
Vestas V90 - 3.0 MW 109.4 dB(A) |
|
|
80 |
4.28 |
47.0 |
0.2 |
262.8 |
2,302,400 |
8.8 |
Vestas V100 - 1.8 MW |
|
|
80 |
4.28 |
47.0 |
2.0 |
274.5 |
2,404,300 |
15.2 |
Vestas V100 - 2.0 MW |
|
|
80 |
4.28 |
49.0 |
0.5 |
274.4 |
2,403,700 |
13.7 |
Vestas V100 - 2.6 MW |
|
|
75 |
4.21 |
49.7 |
0.6 |
287.3 |
2,516,800 |
11.1 |
Vestas V112 - 3.0 MW |
|
|
84 |
4.33 |
49.8 |
1.5 |
391.1 |
3,426,500 |
12.7 |
IMPORTANT: No turbine losses are included in the power, energy, and capacity factor values in the table. Typically, turbine losses can be 5-20% to account for maintenance downtime, icing/soiling and losses from other turbines in a wind farm. Users wanting to be conservative in the performance projections should multiply the power, energy, and capacity values by (1- % losses) to account for these losses.
|