Mar 31, 2004 - Resolving Difficult Issues of Wind Power Micrositing in Complex Terrain. Russell G. Derickson a. Michael McDiarmid b. Brad C. Cochran c.
RESOLVING DIFFICULT ISSUES OF WIND POWER MICROSITING IN COMPLEX TERRAIN
SESSION 11B RESOURCE ASSESSMENT AND MICROSITING STRATEGIES AND EXPERIENCE WEDNESDAY, MARCH 31, 2004 1:45 P.M. - 3:15 P.M. RUSSELL G. DERICKSON A MICHAEL McDIARMID B BRAD C. COCHRAN C JON A. PETERKA D A,C
ASSOCIATE, CPP INC., FORT COLLINS, COLORADO 80524 USA MECHANICAL ENGINEER, STATE OF NEW MEXICO, SANTA FE, NEW MEXICO 87505 D VICE PRESIDENT, CPP INC., FORT COLLINS, COLORADO 80524 USA B
1
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois Resolving Difficult Issues of Wind Power Micrositing in Complex Terrain Russell G. Derickson a Michael McDiarmid b Brad C. Cochran c Jon A. Peterka d a,c
Associate, CPP Inc., Fort Collins, Colorado 80524 USA b Mechanical Engineer, State of New Mexico, Santa Fe, New Mexico 87505 USA d Vice President, CPP Inc., Fort Collins, Colorado 80524 USA
wind power output or introduces severe, damaging turbulence. Three scenarios leading to poor performance are explored:
ABSTRACT Micrositing of wind turbines in complex terrain is tricky game, and many of the current siting methods and tools, while useful and improving, remain inadequate to the task in extreme terrain. As a consequence, there are numerous wind turbine installations that are buffeted by damaging turbulence or are faced with suboptimal wind energy performance.
1) It may not be a good strategy to place turbines near the leading edge of steep escarpments or mesas, where flow separation occurs, turbulence is high, a large vertical wind component exists, and turbines encounter extreme wind shear. Several commercial projects have faced these unfavorable conditions. Other locations on a mesa may not be appropriate, either, depending on terrain character and atmospheric thermal stability.
Flow separation at mountain peaks and the leading edge of cliffs and escarpments is a primary culprit that can lead to poor performance or damaging turbulence. Many practitioners poorly understand these phenomena, but the blame can also be laid at the feet of assessment tools that are not suitable for complex, separating flow conditions.
2) Assessing high points or ridge tops for wind power siting requires analysis of any upstream terrain features that may produce flow separation regions that encompass the downwind sites, undermining performance. Subtle, hard-to-quantify effects can occur. Thus, initial qualitative site assessments utilizing simple rule-of-thumb methods for characterizing a site may inadvertently disqualify a good site or accept a bad one. It appears that many numerical simulation tools in existence may not accurately account for the downwind influence of flow separation, either. Lastly, it may be difficult or expensive to fully account for disruptive upstream separation effects with field measurements alone.
Terrain geometries can do strange things to flow patterns. Geophysical phenomena such as thermal stratification and Earth’s rotation can add to the complexity. Thus, mountain peaks, terrain saddle points, steep escarpments, and other intricate terrain shapes can play havoc with flow conditions. A combined approach utilizing wind tunnel testing, a “refined” mesoscale numerical model, and “effective” field testing is shown to be a viable tool for the accurate micrositing of wind turbines in these complex terrain settings. INTRODUCTION Micrositing for wind power in complex terrain remains a formidable task subject to potentially significant error. Two effects may be at play: misconceptions about basic wind phenomena, and assessment tools that degrade in extreme terrain. Consequently, many commercial projects have fallen victim to suboptimal siting that undermines
3) Flow in complex terrain varies remarkably, depending on daily and seasonal variations in the thermal stability of flowing air masses. Consequently, peaks and ridges may be good candidates under some thermal conditions but not others. In some cases, locations within
2
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois valleys or at their openings to plains may be ideal sites. It is important to assess an area for seasonal variations in thermal stratification in choosing a site for overall wind power viability.
FLOW SEPARATION Fig. 1 displays the multifaceted and complicated phenomena of flow separation and reattachment over mountainous terrain. The picture is of flow visualization of a scale model in a boundary-layer wind tunnel. Flow “separates” within a range of various points either upwind or downwind of the peak, depending on precise terrain geometry and upwind flow turbulence. “Reattachment” occurs at some point on the lee side of the peak. The processes of separation and reattachment are highly transient and erratic (and flip-flop in nature), as is the character of turbulent flow in general. At the leading edge of an escarpment, flow separation and reattachment are similar, yet distinct from that over a peak. Exact flow behavior is a critical function of detailed peak or escarpment features, and it is difficult to generalize flow response. Thus, each case is unique. Each candidate site therefore would have to be explored independently. Fig. 2 shows a simplified schematic of flow over two types of steep escarpment.
A hybrid tool, incorporating mesoscale numerical modeling, boundary-layer wind tunnel testing, and field measurements, provides a powerful means for improving micrositing in the complex terrain scenarios described above.
reverse flow in separation region (recirculation)
Wind flow separation occurs at peak
1
flow reattachment on lee of slope (no recirculation)
1
Fig. 2. Schematic of two types of steep escarpment showing flow separation recirculation bubbles and flow reattachment. DESCRIPTION OF MICROSITING TOOLS Fig.1. Capturing the transient nature of flow separation and reattachment in the wind tunnel at a peak. The peak corresponds to Location 1 in the scale model.
A hybrid tool incorporating boundary layer wind tunnel testing and mesoscale modeling (i.e., a numerical simulation software program for atmospheric analysis and study) for micrositing in
3
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois complex terrain was detailed in Derickson and Peterka [1]. The potential pitfalls and complexities of field measurements were also presented in that reference. In this paper, we summarize key points of our previous efforts and move on to new information based in more recent wind tunnel results with escarpments of various leading edge slopes.
large scale) atmospheric forcing and all intermediate scales down to the micro-scale (i.e., small scale) through a process of nesting the numerical grids.
Boundary-Layer Wind Tunnels A boundary layer wind tunnel is well suited to capture the physics of flow separation and reattachment around bluff bodies of extreme geometry. These include highly articulated terrain features such as cliffs and escarpments, as well as the intricate building shapes and arrays of buildings found in city environments. Current numerical models cannot deal accurately with city geometries and are limited in the level of complexity treated accurately in complex terrain. On the other hand, wind tunnels are limited in certain geophysical realities such as thermally stratified atmospheric conditions and effects of Earth’s rotation. Numerical models can address those geophysical effects. Hence, the combined use of the wind tunnel and a comprehensive mesoscale model represents an impressive tool.
Open-Circuit Wind Tunnel
Fig. 3. Pictorial of boundary layer wind Our efforts with ARPS entailed further developments and software refinements that enhanced its ability to simulate flow separation in complex terrain. The process of refinement was enabled by synergistic applications with the wind tunnel. Further explanation of this procedure and the hybridization of the wind tunnel and ARPS are presented in Derickson and Peterka [1].
A pictorial of two types of wind tunnel are shown in Fig. 3. Scale physical models of terrain are placed in the tunnel and tested with flows that correctly simulate the characteristics of the atmospherics boundary layer (i.e., the mean wind structure and turbulence characteristics that vary with elevation as influenced by surrounding terrain features and various geophysical forcing mechanisms. Fig. 1 suggests the power of the wind tunnel to simulate these phenomena and capture the essential character of transient flow behavior.
Field Measurements Ultimately, field site measurements and comprehensive atmospheric observations are indispensable for micrositing and for validation of micrositing tools such as a wind tunnel, a mesoscale numerical model, or their hybrid use. However, field measurements are limited by expense, time, and scale. Most critically, the vast variability of nature and its myriad of interactive parameters (weather systems, terrain and surface features, solar insolation, thermal stratification, Coriolis forces, etc.) create a formidable impediment to the effective comparison of numerical modeling output or wind tunnel results
Atmospheric Numerical Model In our numerical simulations of flow over complex terrain, we employ the Advanced Regional Prediction System (ARPS), which was developed at the Center for the Analysis and Prediction of Storms (CAPS) at the University of Oklahoma [2,3,4]. ARPS is a comprehensive, multi-scale model, originally developed for the mesoscale (i.e., intermediate scale), particularly to forecast severe storms. It can simulate synoptic scale (i.e.,
4
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois to field data. Correlation between neighboring measurements and determination of actual speedup effects of terrain sites can break down, depending on diurnal meteorological conditions as discussed by Meroney [5]. Hence, ultimate validation is a tricky, elusive process.
1 2
3 4
5
6
7
wind
300 m
For example, uncertainties and seasonal variations in surface roughness produce equal uncertainty and variation in wind speeds, particularly in the atmospheric surface layer where wind turbines are located. It is difficult to quantify a precise value for zo in the field at a specific location, let alone its spatial variation for upwind fetches that influence overall flow characteristics. If surface roughness changes with season as vegetation either grows or changes by various processes, the task of identifying proper values for zo becomes increasingly convoluted. The consequences on wind power assessment are profound for turbines near the surface and not insignificant for larger scale devices with higher hub heights.
190 m 3600 m
Fig. 4. Schematic of escarpment with windward slopes of 30º, 45º, 60º, and 90º. Wind profiles are indicated at Locations 1-7. In this study, we explore the details between locations 1 & 2, as depicted in the subsequent figures.
Our previous study [1] examined Locations 1-7, looking at horizontal wind speed only and turbulent gusts. Results are presented in Figs. 5-10. With steep approach slopes (60 and 90 degrees), the vertical velocity component remains large compared to the horizontal component for distances up to 20 to 30 meters from the edge. Even at 40 meters, the vertical component is large. This would induce negative consequences on the aerodynamics of a wind turbine experiencing this condition. Notice also that the point of maximum vertical wind speed is displaced upward as the distance from the edge increases. At 30 degree and 45 degree slopes, the vertical wind component is much less.
We recognize, however, the crucial need for comparisons to field measurements as a final step in the validation process, and have planned for that activity. However, we emphasize that any comparative study between ARPS (or any numerical model) and the field is replete with complex issues that may be difficult to resolve. FLOW OVER LEADING EDGE OF AN ESCARPMENT Recent wind tunnel studies have complemented our previous explorations [1] of flow over escarpments and mountainous terrain in general. Due to unanswered questions from the previous study, we chose to look in greater spatial detail at flow in the vicinity of the leading edge of an escarpment. These questions included uncertainty in the magnitudes of vertical wind components near the leading edge. We used a refined measurement devise, a 5-hole probe (briefly described in [1]) to help answer our questions.
At Location 2, the vertical wind speed has decreased substantially for the 60 and 90 degree cases, but the horizontal wind speed manifests a large vertical gradient that would induce strong asymmetrical loading on a wind turbine. Thus for steep escarpment slopes, one is faced with either large upward incursions of wind flow or strong asymmetrical turbine loading, depending on distance from the edge. In cases with shallow slopes, there is horizontal wind speed-up relative to an approach flow upwind of the escarpment, as shown in the figures. With the steep slopes, the situation is more complicated with regard to speed-up at various heights.
Fig. 4. is a schematic of a scale model of an escarpment used in our wind tunnel study. We focused on the region between Locations 1 & 2 in the figure (shown as a detail), and measured both horizontal and vertical wind velocity components.
5
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois
Wind Components: Location 1 (leading edge) vertical wind
Elevation, m
200
horizontal wind approach profile
150
2
1
wind
100 900
50 0
30
0 0
2
600
900
450
4
6
8
600
10
450
300
12
14
16
Speed, Fig. 5. Flow parameters at Location Wind 1: at leading edgem/s of escarpment.
Wind Components: Location A (10m from edge) vertical wind
Elevation, m
200
horizontal wind 1
approach profile
150
2
A
wind
100 900
450
50
300
900
0
450
60
0 0
2
4
300
6
8
10
12
0
60
14
16
Wind Speed, m/s Fig.6. Flow parameters at Location A: 10 m downwind of escarpment leading edge.
The lesson is clear, locations in the vicinity of the leading edge of a steep escarpment present undesirable wind conditions for essentially all hub heights. On the other hand, escarpments with shallow slopes that produce little or no flow separation are excellent candidates for wind power siting. Thus, a solid recommendation would be to
select those sites for turbine siting and seriously consider avoiding locations that produce high levels of flow separation.
6
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois
Wind Components: Location B (20m from edge) vertical wind
Elevation, m
200
horizontal wind approach profile
150
B
wind
100
450 0
30
50
2
1
900
900
0
60
450
0
600
0
2
4
6
8
300
10
12
14
16
Wind Speed, m/s Fig.7. Flow parameters at Location B: 20 m downwind of escarpment leading edge.
Wind Components: Location C (30m from edge) vertical wind
Elevation, m
200
horizontal wind approach profile
150
C
wind 100
900 0
50
0
45
300
60
900
0
450
600
-2
0
2
4
6
8
300
10
12
14
16
Wind Speed, m/s Fig. 8. Flow parameters at Location C: 30 m downwind of escarpment leading edge.
7
2
1
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois
Wind Components: Location D (40m from edge) vertical wind
Elevation, m
200
horizontal wind approach profile
150
0
note : no 60 case 2
1
100
300
450
900
50
D
wind
900
450
0
300
-2
0
2
4
6
8
10
12
14
16
Wind Speed, m/s Fig.9. Flow parameters at Location D: 40 m downwind of escarpment leading edge.
Wind Components: Location 2 (190m from edge) vertical wind
Elevation, m
200
horizontal wind
300
150
approach profile
900
450
1
wind
100 900
600
50 0
600
-2
450
0
2
4
6
300
8
10
12
14
16
Wind Speed, m/s Fig. 10. Flow parameters at Location 2: 190 m downwind of escarpment leading edge.
8
2
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois
In our previous study [1], we explored wind behavior on the full extent of the top of escarpments of varying approach slopes. That study revealed the high levels of turbulence near the leading edge of steep escarpments. It also showed that the lee regions of escarpments might induce a slight speedup, if the approach slope is not steep. This has been confirmed by a wind tunnel study performed by Bowen and Lindley [6]. Figs. 11 & 12 are pictures of installed wind turbines in the vicinity of an escarpment edge. Without knowing the details of the site or any analysis that was performed in site selection, it is uncertain as to what sort of total wind environment these turbines face. However, the results and discussion presented above indicate the possibility that the turbines may be subjected to adverse wind conditions.
Fig 12. Wind turbines near an escarpment edge. (Photo by Michael McDiarmid).
2
1
3
Fig. 11. Wind turbines near escarpment edge.
MULTIPLE PEAKS AND VALLEYS IN MOUNTAINOUS TERRAIN A scenario of mountainous terrain with multiple peaks and valleys is shown in Fig. 13. This is actually a scale model that was employed in a series of wind tunnel studies (Fig. 1 shows an isolated peak, denoted as Location 1, of the entire model.). For a given wind condition, any peak can be either upwind or downwind of another peak, complicating the matter of which peak influences other downwind peaks or valleys. A detailed study would be necessary to
Fig.13. Scale model of mountainous terrain (Lantau Island) with multiple peaks and valleys. For wind tunnel study.
9
AWEA Global WINDPOWER 2004 Conference -- March 28-31, 2004 -- Chicago, Illinois evaluate specific peak sites or valley locations for effective wind power utilization. This would require knowledge of the regional wind climatology and all other relevant meteorological conditions. Many existing micrositing tools can address certain of the issues surrounding such a siting evaluation, but most tools would be inadequate in exploring the effects of flow separation and reattachment in the regions with extreme terrain geometry.
Liu, Y., and Wang, D., 2001, “The Advanced Regional Prediction System (ARPS)-A MultiScale Nonhydrostatic Atmospheric Simulation and Prediction Model. Part II: Model Physics and Applications,” Meteorol. Atmos. Phys., 76, 143-165. [5] Meroney, R.N., 1990, "Fluid Dynamics of Flow over Hills/Mountains – Insights Obtained through Physical Modeling", Chapter 7 in Atmospheric Processes Over Complex Terrain, ed. William Blumen, American Meteorological Society, 145-171.
SUMMARY AND CONCLUSIONS Micrositing in complex terrain is a tricky affair, particularly in various escarpment scenarios and mountainous areas with multiple peaks and intervening valleys. The ability to accurately assess the effects of flow separation and reattachment is crucial to the task. However, most current tools are inadequate when it comes to extreme, detailed terrain geometry. Field measurements, while crucial, are replete with their own limitations and difficulties. The combined use of wind tunnel testing and mesoscale numerical modeling represents a powerful hybrid tool for wind power site assessment in highly complex terrain.
[6] Bowen. A.J. and Lindley, D., 1976, “A Wind-Tunnel Investigation of the Wind Speed and Turbulence Characteristics Close to the Ground Over Various Escarpment Shapes’, Boundary-Layer Meteorology, 12, 259-271. ACKNOWLEDGEMENTS We wish to extend our appreciation to the CPP crew responsible for wind tunnel testing, data reduction, and computer administration. Particular thanks go to Bill Roeder, Aki Hosoya, Morgan Downing, Steve Mike, Kurt Fleckenstein, Mike Schoonover, Adrian Manzanares, Brian James, Tim Vice, and Matt Petersen. We also thank our colleagues Daryl Boggs, Leighton Cochran, Gary Larrew, and Dave Banks for their helpful discussions and technical insights during the study leading to this paper.
REFERENCES 1] Derickson, R.G., and Peterka. Jon A. (2004), “Development of a Powerful Hybrid Tool for Evaluating Wind Power in Complex Terrain: Atmospheric Numerical Models and Wind Tunnels”, Proceedings of the 23rd ASME Wind Energy Symposium, Reno, Nevada (in press). 2] Xue, M., Wang, D., Gao J., Brewster, K, and Droegemeir, K. K., 2003, “The Advanced Regional Prediction System (ARPS), StormScale Numerical Weather Prediction and Data Assimilation,” Meteorol. Atmos. Phys., 82, 139170. [3] Xue, M., Droegemeir, K. K., and Wong, V., 2000, “The Advanced Regional Prediction System (ARPS)-A Multi-Scale Nonhydrostatic Atmospheric Simulation and Prediction Model. Part I: Model Dynamics and Verification,” Meteorol. Atmos. Phys., 75, 161-193. [4] Xue, M., Droegemeir, K. K., and Wong, V., Shapiro, A., Brewster, K., Carr, F., Weber, D.,
10
