14 Figure 12: Folsom and Main East and West Buildings, Northwest, West‐Northwest and Southwest Wind Directions, Shown Left‐to‐Right, for the Existing Setting Winds Blow from Top to Botto
Introduction
Background on Urban Wind Energy Converters
Interest in wind energy has led to increased technological advancement, however most wind energy production occurs in rural areas, where energy transmission can be difficult and costly
Deploying wind energy converters (WECs) at demand sites, including urban areas, could mitigate some issues, but urban deployments may also create new challenges such as noise, visual impacts, electromagnetic interference, and safety concerns Because urban-specific studies are limited, researchers can analyze findings from rural-area studies and relate them to potential urban impacts to inform planning and mitigation strategies Ongoing evaluation in urban contexts is needed to fully understand these effects and to guide the integration of urban wind energy into city environments.
One effective approach to siting potential wind energy conversion systems (WECs) in cities is to conduct wind tunnel surveys of urban sites Wind tunnels enable rapid data collection on urban wind flows around buildings and streets, supporting quick, data-driven assessments of site suitability for WECs This method also provides a forward-looking view of how ongoing urban development will alter wind conditions, helping planners anticipate future changes and optimize wind energy deployment in evolving cityscapes.
Horizontal-axis wind turbines (HAWTs) are the predominant form of wind energy conversion systems (WECs) in the United States, with most installations in rural areas Their environmental impacts include visual intrusion, seismic concerns, electromagnetic interference, noise, and disturbance to avian life Placing large HAWTs away from population centers tends to reduce these effects, while siting them near urban areas may require substantial mitigation Although this study did not directly evaluate these issues, it recommends that they be acknowledged and investigated more thoroughly before WECs are located in urban environments.
Figure 1: Wind Farm in California
Urban Wind Energy Converter Survey
Urban wind energy converters (WECs) remain relatively scarce, but rapid technological progress is bringing new concepts to the market and, in some cases, into deployment worldwide Notable examples include the Aeolian Roof Wind Energy System TM (Tyler 2002), Aeolian Towers TM (Tyler 2002), the Vawtex (ASHRAE 2003), and the Architectural Wind Rotor (AeroVironment).
2004) are a few examples of developing or currently used technologies to capture wind energy in urban environments
Tyler proposes integrating rooftops with an Aeolian Roof Wind Energy System™ (Figure 2), a concept that depends on a building being properly oriented and a roof engineered to exploit rooftop aerodynamics Cross-flow turbines, while acting as static concentrators, can capture a wide range of wind angles, increasing the system's versatility.
This type of a system will limit visual appearance, and the small diameter turbine does not require a gearbox, reducing noise (Tyler 2002) and design simplicity may lower maintenance and overall cost This system’s design also provides a simple way to integrate solar and wind power (Tyler 2002) The Aeolian Roof Concept appears to work best with long, relatively low and narrow buildings and will probably not work in an urban city with a tall skyline
The Aeolian Towers concept uses a device mounted on a tower, and a corner attachment provides acoustic insulation, helping to mitigate noise issues (Tyler 2002; Figure 3) Both designs are suitable for deployment in areas with little or no power transmission lines (Tyler 2002).
Two notable examples of developing urban wind energy converters are the Vawtex, short for Vertical Axis Wind Turbine Extractor, and the Architectural Wind R system (Figure 4) The Vawtex, designed by the Harare-based engineering firm Ove Arup, uses wind to cool buildings, illustrating how urban wind power can enhance energy efficiency in city architecture The Architectural Wind R system offers another innovative approach to integrating wind energy within the built environment.
In Zimbabwe, vertical-axis wind turbines like the Vawtex are touted for capturing wind power as directions shift, and their construction from local materials makes them environmentally friendly and potentially viable for poorer communities (ASHRAE 2003) The Architectural Wind R, designed by AeroVironment, can generate up to 2.4 kW in a footprint of about 9.3 square meters (roughly 100 square feet), yielding an average power density of 240 W per square meter (AeroVironment 2005).
Figure 4: Vawtex and Architectural Wind R Pictures
Another company, Aerotecture International, Inc., is producing an urban WEC, the
Aeroturbine, and currently claims to have sold several of these devices The website for this company gives a power curve for the turbine, with a conceptual drawing (left) and photo
Figure 5: Concept Drawing of an Aerotecture Aeroturbine and Vertical Mounting on a Rooftop
Among wind energy converters (WECs) designed for urban use and with substantial data reported, the Aeroturbine was the sole model to provide a power curve and actual sales figures, which is why it was selected for the comparative power production analysis in this study.
During the 2005 CWEC forum in San Diego, additional urban WECs were discussed, highlighting ongoing developments in urban energy and water collaborations For comprehensive details on the forum proceedings and the WECs mentioned, see the proceedings page at http://cwec.ucdavis.edu/forum2005/proceedings.
Methods
Wind‐Tunnel Testing
Wind‐tunnel testing was the method used to gather data for this study The Atmospheric
All testing was conducted in a Boundary Layer Wind Tunnel due to the extensive prior testing of pedestrian-level winds in San Francisco The wind tunnel itself, along with the methods for setup of the tests, is described in detail in the next few sections Validation of wind-tunnel based results has been conducted on numerous occasions and is not a part of this study; thus it will only be recapped in this section A more thorough investigation into the validation of wind-tunnel studies is conducted in Appendices C and D (modified from White 2001).
To achieve flow similarity between wind-tunnel and full-scale flows, flow similarity parameters are defined by nondimensionalizing the governing equations for turbulent flow Starting from the conservation of mass, momentum and energy and applying the Boussinesq density approximation, non-dimensional quantities are introduced and substituted into these equations to yield the dimensionless continuity, momentum and turbulent energy equations (White 2001) From these equations, the non-dimensional parameters that describe flow similarity are observed (White 2001).
Key parameters in fluid dynamics include U, the fluid speed; L, the characteristic length; Ω, the angular rotation; g, gravity; T, the temperature in Kelvin; ρ, the fluid density; and cp, the specific heat capacity of the fluid Together, these variables govern how momentum, heat, and mass are transported and how buoyancy, rotation, and thermal storage shape the flow For example, U and L set the Reynolds number and the scale of motion, Ω and g introduce rotational and gravitational forcing that drive circulation and possible instabilities, while temperature T controls density through thermal expansion and drives heat transfer The specific heat cp determines how much energy is needed to raise the fluid temperature, influencing transient thermal responses during heating or cooling A complete model couples the Navier–Stokes equations with an energy equation, using these parameters to predict convection, diffusion, and rotational effects in the system.
The Rossby number quantifies the strength of the Coriolis effect in fluid flows, and for small-scale modeling where the horizontal domain is under about 5 kilometers or measurements are confined to the boundary layer, the Coriolis effect is negligible and the Rossby number can be ignored, a finding supported by White (2001) and applicable to this study.
The densimetric Froude number is the ratio of inertial forces to buoyant forces in the flow In a wind tunnel that simulates neutrally stable atmospheric conditions, buoyancy is negligible, causing the Froude number to tend toward infinity and allowing it to be neglected in this analysis.
Since air is the same working fluid in both the wind tunnel and full-scale tests, the Prandtl number is matched between measurements The Eckert number, which relates to compressible flow, is negligible at the low speeds characteristic of both the wind tunnel and full-scale conditions, so compressibility effects can be ignored.
Reynolds number is a central factor in this wind-tunnel investigation, guiding the interpretation of model-scale results for real-world applications Because the model is geometrically downsized, it is impractical to replicate the full-scale Reynolds number in the wind-tunnel tests Instead of forcing a full-scale Reynolds match, the study demonstrates Reynolds-number independence, ensuring that the essential aerodynamic trends observed at model scale translate to full-scale conditions.
Sutton (1949), if the roughness Reynolds number, ν 0
To achieve Reynolds-number independence and faithfully simulate the large-scale turbulence of full-scale flows in the wind tunnel, the tests were designed to satisfy a criterion based on the product of the friction speed u* and the roughness height z0, kept at or below 2.5 With a free-stream velocity of about 3.8 m/s, this yields a friction speed u* ≈ 0.24 m/s and a roughness height z0 ≈ 0.0025 m, confirming that the wind-tunnel tests satisfy the condition (White 2001).
To ensure accurate wind-tunnel simulations, several conditions must be satisfied: matching the power-law and Jensen’s length-scale criteria, and proper matching of H/δ when H is the measurement height in the wind tunnel; if H/δ exceeds 0.20, H/δ must be matched, whereas for the lower 20 percent of the boundary layer when H/δ is less than 0.20, matching is only required for that region, and the test-section blockage by the model should be limited to 5–15 percent of the total cross-sectional area to prevent any influence from streamwise pressure gradients, a constraint met by using a sufficiently small model as in this study (White 2001).
U , where α is the power‐law exponent, U is the velocity at height H,
U∞, the mean‑free velocity of the wind above the boundary layer of height δ, with z representing the roughness height (White 2001), is reproduced in the wind tunnel by configuring roughness elements These elements are 2 inch by 4 inch wooden blocks, 12 inches long, laid out on the floor upwind of the test section in a pattern previously determined to produce that velocity value.
Jensen’s length-scale criterion requires that the ratio of roughness height to building height, z0/H, be the same in wind-tunnel and full-scale simulations (White 2001) In this study, the model and the roughness height were geometrically scaled, ensuring that the z0/H ratio is preserved and the criterion is satisfied.
Under the law of the wall, keeping H/δ below 0.2 for the lower 20% of the boundary layer means that if the full-scale H/δ is under 0.2, the full-scale value does not need to be matched in wind tunnel simulations; the wind tunnel only needs H/δ < 0.2 (White 2001) Since the boundary layer height in the atmospheric boundary layer wind tunnel (ABLWT) is about 1 meter, measurement heights should be limited to no more than 0.2 meters unless H/δ is matched at higher elevations Fortunately, the tall buildings’ obstruction of the Ekman spiral makes it possible to obtain good data for measurement heights above 0.2 meters.
2.1.1 The Atmospheric Boundary Layer Wind Tunnel
Located at the University of California, Davis, the Atmospheric Boundary Layer Wind Tunnel (ABLTWT) reproduces airflow within the Earth's turbulent boundary layer The facility has three main sections—the flow development, the test, and the diffuser—through which air enters the inlet, passes flow straighteners and spires, and then traverses a long fetch of roughness elements formed by wood blocks no taller than two inches, arranged in a pattern to shape the proper flow profile in the flow development section By the time the flow reaches the test section, it exhibits the required turbulence characteristics, and Plexiglas windows on both sides provide visibility After testing, the flow exits through the diffuser, the flow straighteners, and the fan More detailed specifications of the wind tunnel are available in Appendix A.
Figure 6 shows the schematic of UC Davis’s Atmospheric Boundary Layer Wind Tunnel (White 2001), with the tunnel length corresponding to fifty feet (15.24 meters) in full scale The model is built from modular blocks that do not have to match actual city blocks, and these blocks are assembled like a puzzle to recreate the downtown area depicted in Figure 7.
After selecting a wind direction for testing, the test building’s model is centered in the wind tunnel’s test section, and surrounding model blocks are arranged to fill the entire section Blocks are placed far enough upwind—about 600 meters or 1,970 feet at full scale—to faithfully simulate how nearby buildings influence the wind as it approaches the test building, just as in the city Any portions that wouldn’t fit in the wind tunnel are represented with wooden blocks of approximate size to stand in for missing buildings When testing for the wind direction is finished, the model assembly can be rotated to simulate a new wind direction, with some blocks rotating out of the test section and others filling in the created gaps By rotating blocks or swapping in different model blocks, virtually any area of the city and wind direction can be tested, enabling many wind-direction scenarios to be explored in less time than full-scale testing and data collection.
Wind Data
San Francisco was chosen for this wind study due to its relatively high wind levels, with meteorological data collected from an anemometer installed at the old Federal Building, 50 U.N Plaza The instrument is positioned at a height of 40.2 meters (132 feet) above ground level, and the data cover 1945 through 1947, reported as percentages of wind occurrence per year This wind dataset was selected for its completeness and because these years are representative of typical San Francisco wind conditions (White).
2006) All wind speed data is originally from both the National Climatic Center and the
Data from the Department of Water Resources (California) were used to construct tables, with the author compiling these tables from the dataset Originally, the data were broken down into 3-hour increments per month (for example, January 12am–2am, January 3am–5am, December 12pm–3pm, etc.) The wind measurements at this anemometer location were not free of local building effects, so correction factors must be applied before wind energy analyses can be completed The uncorrected meteorological data are shown in Tables 6 through 9, and the corrections to this data are described in detail in Section 2.4.2.
Table 6 presents wind data digitized in a Microsoft Excel spreadsheet and expressed as percentages of time per year The left column lists wind directions, while the top headers define wind speed bins in knots, miles per hour (mph), and meters per second (m/s) For each direction, Table 6 shows the percentage of time the wind blows within each speed bin, the total percentage of time from that direction, and the average wind speed for that direction The overall average wind speed over the period is 11 mph (about 5 m/s or 10 knots) A wind rose is included in Table 6 to visualize the data, with concentric circles representing 1% of annual time and speeds shown in knots.
Certain calculations require that wind data be organized into a percent-exceeded wind speed table, where the wind speed is the value exceeded for a specified percent of time during a typical year For example, a 10-percent exceeded wind speed, written as U 10%, is the wind speed exceeded 10 percent of the time These values can be broken out by wind direction as well, such as U 10% SW, which represents the wind speed exceeded 10 percent of the time when winds come from the southwest When a directional reference is included, the percent time is still calculated relative to all occurrences in a typical year.
To construct the percent exceedance wind speed table, begin with the highest recorded wind speed and sum the occurrences (or yearly percentages) of all events at or above that speed, producing a data point that pairs wind speed with the percent of time exceeded The next data point is formed by adding the next lower wind speed’s percent occurrence to the cumulative percent above, using that wind speed as the threshold for the cumulative exceedance This stepwise calculation continues down to zero speed, yielding a complete cumulative exceedance curve and showing that winds exceed zero speed for a nonzero portion of time.
Wind-tunnel testing focused on four wind directions—northwest, west-northwest, west, and southwest—rather than analyzing 16 directions The first three directions were selected because they have the highest annual occurrences among wind directions: northwest (207 hours per year), west-northwest (244 hours per year), and west (131 hours per year) The southwest direction was included to evaluate wind effects on buildings south of Market Street in San Francisco, where the street grid aligns with the southwest and wind-tunnel effects may drive higher wind speeds along those streets.
Because wind conditions do not change significantly with small shifts in wind angle, wind directions are grouped when calculating the percent exceedance of wind speed over time Ideally, half of each neighboring wind's data is included with the tested direction’s data For example, the Northwest group includes half of the North-Northwest information, though no West-Northwest data are included since West-Northwest is analyzed separately Similarly, West-Northwest does not borrow data from other directions because Northwest and West winds are analyzed individually The West direction includes half of the West-Southwest data The Southwest direction includes half of the West-Southwest data as well as half of the South-Southwest data All remaining wind directions are combined into an “Other” category, summed across wind bins not used by the four specified directions The term “All” refers to the cumulative effects of all wind directions and equals the sum of the data from the four primary directions plus the data from the “Other” category.
Data points for each wind direction were connected with a smoothed line in Excel, providing a realistic interpolation between observations Because the wind bins in Table 6 were relatively large, spanning at least 3 knots in each bin, the data points were comparatively sparse, so the smoothed line offered a plausible interpolation between points The analysis selected the percentages of time exceeded and manually catalogued the corresponding velocities from the graph Time bins of 5 percent were used, creating twenty data points for the percent wind speed exceeded table, which yielded more data points than the original data set.
Wind-speed exceedance was calculated for all directions at a given exceedance percentage of time For a full year, the exceedance value of wind speed is the same across directions, but the times at which that speed is exceeded differ by direction In other words, the percent of time that the selected exceedance wind speed occurs must sum across directions to the overall percentage of time the wind speed is exceeded, by the definition of directional occurrence percentages The results are shown in Table 7, where the left column lists wind directions, the top rows list wind speeds, and each cell reports the percentage of time that a given wind speed is exceeded in a typical year.
*One circle equals one percent of time of occurrence per year.
Min Speed (mps) 0 1 2 4 6 9 11 14 17 21 25 29 Hours Mean Mean Mean
Max Speed (mps) 0 2 3 5 8 11 14 17 21 24 28 % time per Wind Speed Wind Speed Wind Speed
Ave Speed (mps) 0 1 3 4 7 10 13 16 19 23 26 29 Year (knots) (MPH) (mps)
Ave Speed (knots) 0.00 1.75 2.75 3.50 4.50 5.50 7.25 8.50 9.50 10.50 11.50 12.25 13.25 14.50 15.50 16.75 18.00 19.50 21.50 24.50 Ave Speed (MPH) 0.00 2.01 3.16 4.03 5.18 6.33 8.34 9.78 10.93 12.08 13.23 14.10 15.25 16.69 17.84 19.28 20.71 22.44 24.74 28.19 Ave Speed (mps) 0.00 0.90 1.41 1.80 2.32 2.83 3.73 4.37 4.89 5.40 5.92 6.30 6.82 7.46 7.97 8.62 9.26 10.03 11.06 12.60 Direction
Table 7: San Francisco Wind Data in Percent Exceeded Wind Speeds, Calculated Manually
% Time Exceeded vs Speed (knots)
Figure 19: Percent Exceeded Time versus Wind Speed (Knots as Shown in Table 6)
2.2.1 San Francisco Winds from 6am to 8pm
Although the prior analysis treated winds over the full day, there are hours when energy use in San Francisco is higher than average The city’s 1985 wind ordinances restrict high wind speeds during development projects between 7am and 6pm, signaling activity in that window and making it a favorable period for wind analyses (Arens 1989) Because the wind data from 1945–1947 were recorded as three-hour averages, the subsequent analysis uses data from 6am to 8pm (White 1992) Methods for calculating annual wind data and percent exceedances within 6am–8pm are equivalent to the all-day data shown in Tables 6 and 7 Table 8 presents the 6am–8pm wind distributions for a typical year and includes a wind rose; each circle marks one percent of time within that window per year, with speeds in knots The 6am–8pm window contains 5475 hours per year, and the average wind speed for this dataset is approximately 6 meters per second (13 mph or 11 knots).
Figure 20 illustrates the percent exceeded times versus wind speeds for the hours of 6am to 8pm for all wind directions, including a breakdown by the four wind-tunnel tested wind directions, and Table 9 presents the hand-tabulated percent exceeded winds broken into time bins of 5 percent within the 6am–8pm window, where the upper rows correspond to the exceeded wind speeds, the left column indicates wind direction, and the data are presented in a matrix form.
*One circle equals one percent of time of occurrence per year.
Min Speed (mps) 0 1 2 4 6 9 11 14 17 21 25 29 Hours % time Mean Mean Mean
Max Speed (mps) 0 2 3 5 8 11 14 17 21 24 28 per per Wind Speed Wind Speed Wind Speed
Ave Speed (mps) 0 1 3 4 7 10 13 16 19 23 26 29 Year Year (knots) (MPH) (mps)
Ave Speed (knots) 0.00 2.25 3.25 4.25 6.25 7.50 8.30 10.00 11.00 12.00 13.25 14.25 15.25 16.25 17.25 18.25 19.50 21.00 22.75 25.25 Ave Speed (MPH) 0.00 2.59 3.74 4.89 7.19 8.63 9.55 11.51 12.66 13.81 15.25 16.40 17.55 18.70 19.85 21.00 22.44 24.17 26.18 29.06 Ave Speed (mps) 0.00 1.16 1.67 2.19 3.22 3.86 4.27 5.14 5.66 6.17 6.82 7.33 7.85 8.36 8.87 9.39 10.03 10.80 11.70 12.99 Direction
Table 9: San Francisco Wind Data in Percent Exceeded Wind Speeds from 6am to 8pm, Calculated Manually
Figure 20: Percent Exceeded Time versus Wind Speed (Knots as Given in Table 8) from 6am to
There are various stability conditions that occur within the atmospheric boundary layer Those stability conditions are illustrated in Table 10 (modified from Gifford 1976) Since no heating or cooling elements are employed in this study, and due to the scale of the boundary layer simulated in the wind tunnel, the wind tunnel simulated only neutrally stable flow conditions Fortunately, Pasquill (1971) suggests that in strong winds, thermal stratification effects in the lower portion of the boundary layer are negligible as shown in the table In addition, the tall building structures in the urban areas of San Francisco further add to the mixing within the boundary layer due to the turbulent wakes shedding from the upwind structures Taking this into consideration, the neutrally stable flow conditions in the wind‐tunnel study are realistic
Table 10: Stability Criteria: Meteorological Conditions Defining Pasquill Turbulence Type
Data Collection
Wind‐tunnel measurements of the mean velocity, R‐values and turbulence intensity were performed using hotwire anemometry The hotwire used in this study was a standard Thermo Systems Inc (TSI) single hotwire sensor, model 1210‐60 The sensor was placed at the end of a
A 50 cm TSI probe support, model 1150 (White 2001), was mounted on the platform of a three‑dimensional positioning system above the test section of the ABLWT The probe was connected via a 10 m shielded tri‑axial cable that ran from the end of the probe support to a TSI model IFA‑100 Intelligent Flow Analyzer, a constant‑temperature thermal‑anemometry flow analyzer with an integrated signal conditioner (White 2001) Each hotwire probe used in this study was calibrated with the ABLWT facility and equipment prior to testing.
The IFA 100 was run by a LabVIEW software virtual instrument (VI), which initialized and configured the analog‐to‐digital data acquisition board by United Electronics Inc., linked to a
Measurements were made by recording raw voltages at a sampling rate of 1,000 Hz, with 30,000 total samples, to satisfy the Nyquist sampling theorem for a wind-tunnel turbulence signal with an average frequency of 300 Hz (White 2001) Additional information and specifications are provided in Appendix B.
Data Reduction and Analysis
Data reduction and analysis were performed in several steps: first, raw data were reduced; next, estimated full-scale speeds were calculated for each measurement location (also referred to as receptor locations or points); finally, wind-speed information for each point was used to estimate power densities and to predict annual energy outputs for multiple WECs.
LabVIEW yielded a single data file with 30,000 voltage readings per measurement A Quick Basic script, commonly used in the ABLWT laboratory, processes each voltage through the hot-wire calibration to derive the wind speed at the measurement point, and then averages these speeds to obtain the mean wind speed The script then computes the wind-speed ratio and the turbulence intensity for each measurement The wind-speed ratio (R-value) is defined as the wind tunnel velocity at the measurement location divided by the reference velocity at 0.70 meters (2.3 feet), while the turbulence intensity is defined as the root-mean-square of the instantaneous deviations from the mean velocity, divided by the mean velocity (Arens 1989).
2.4.2 Estimated Full-Scale Speed Calculations
Raw data are reduced to R-values, and the percent exceedance wind speeds are determined for each wind direction at full scale; the estimated full-scale speeds that occur during a typical year are then calculated, with all calculations performed in an Excel spreadsheet The definitions below describe the variables used in the following equations.
• U % = the percent exceeded wind speed; for example, U 10% is the wind speed exceeded 10 percent of the time in a typical year
• t %direction = the percentage of time U % is exceeded for the specified wind direction; for example, t 75%SW is the percentage of time winds from the southwest exceed U 75%
• R direction = the R‐value of one point for the specified wind direction; for example, R WNW is the R‐value of a single point for the west‐northwest wind direction
• CF direction = the correction factor for a specified wind direction; for example, CF W is the correction factor to be applied west winds
• U point = the wind speed at the point
• U ref = the reference wind speed
• z point = the height corresponding to U point
• δ = the boundary layer height, which is 402.3 meters (1320 feet) for San Francisco (White
• α = the power‐law exponent, which is 0.3 for San Francisco (White 1992)
• The subscript “direction” shall denote that the value is for one wind direction The actual wind direction in consideration may also be used instead of the word “direction”
• The subscript “Wind Tunnel” refers to wind tunnel data
• The subscript “Full Scale” refers to full‐scale values
The power‐law is used to show the relationship of full‐scale wind speeds to measured wind speeds in the wind tunnel (White 1992): α
⎛ ref point Tunnel ref Wind point Scale ref Full point
Rearranging the variables and multiplying and dividing by U ∞ yields (White 1992):
Wind point Scale point Full
Scale ref Full Tunnel point Wind Scale point Full
By definition, (U_point / U_ref) is the R-value However, wind-tunnel data are not accurate at the level of U∞ due to the Coriolis effect under full-scale conditions (White 1992) Consequently, a unified relationship among these variables is sought Wind-tunnel tests on a model of the old Federal Building indicate that (U∞ / Uref) ≈ 2 With the boundary-layer height, the height of the reference velocity, and the San Francisco power-law exponent, the power-law yields the following relation (White 1992): δ^0.5 z / U.
Equation (3) aligns well with the wind-tunnel results described earlier Substituting this finding into the preceding equations yields the relationship among the reference wind speed, the R-value, and the full-scale speed at a specific point for wind from a single direction (White, 1992).
(U point ) Full Scale =2⋅R⋅(U ref ) Full Scale (4)
R-values were obtained only for the four wind directions that were tested; to estimate the R-values for the remaining directions, a weighted average was calculated using the measured values together with data from other wind directions This approach provides a consistent estimate of thermal resistance across all wind directions by integrating information from tested directions with supplementary data.
Wind data for San Francisco from 1945 to 1947, collected by an anemometer atop the Federal Building, were influenced by surrounding structures, requiring correction for building interference To quantify this effect, the model of the old Federal Building area was tested in a wind tunnel twice: once with the measurement point at the anemometer location and once with the probe positioned away from local building interferences (White 1992) The tests yielded direction-specific correction factors (CF) to adjust the reference speeds: 1.02 for northwest, 1.00 for west-northwest, 0.96 for west, 0.85 for southwest, and 0.96 for all other wind directions By applying these CFs to the reduced data shown in Tables 6 through 9, the full-scale wind speed at a given point is determined by the updated equation, providing corrected wind measurements for 1945–1947 (White 1992).
(Upoint direction ) Full Scale =2⋅Rdirection⋅CFdirection ⋅ ( Uref ) Full Scale (6)
U input can be the regular wind speed from the San Francisco wind data or the U percent value representing the exceeded wind speed at a point For this study, percent exceedance was determined for each direction—northwest, west-northwest, west, southwest, and other—and for each point, with exceedance levels ranging from 5% to 100% in 5% increments Once these values were obtained, a weighted average was calculated to estimate the average full-scale (EFS) percent exceeded wind speed across all wind directions for every point, and this weighted-average calculation was performed for each point.
At this stage of the calculations, only the percent-exceeded wind speeds are known To determine the wind speeds at each point through a specific wind energy converter (WEC) power curve or to obtain power densities, a wind speed histogram must be constructed for every location Therefore, the average wind speed within each bin is used to represent that bin; i.e., the bin’s representative value is its mean wind speed, which then feeds the WEC curve to yield power outputs or power densities.
Within the scope of this study, it was deemed reasonable to average the two exceeded speeds rather than fit a curve through each data point to obtain an analytical solution This approach simplifies the analysis while preserving the essential information about speed behavior, providing a clear and interpretable result suitable for reporting and discussion.
To create a histogram of percent-exceeded wind speed data, the values are substituted into the defined equation with the percent-exceeded itself used as the index There are 20 bins in the histogram, spanning from 0% exceeded to 100% exceeded, providing a complete view of the exceedance distribution across the full range.
The method partitions annual time into five-percent exceedance bins, so each bin represents 5 percent of the year with evenly spaced time durations Because calculations do not extend beyond the 5% exceedance level, the zero-exceedance wind speed is defined by adding two meters per second to the 5% exceedance speed Review of the wind data shows no measured speeds more than two meters per second above the 5% exceedance value As a result, the analysis yields a histogram of mean wind speeds for every 5% of time in a year—twenty distinct wind speeds, each occurring for 5% of the year, for every measurement point in the wind tunnel.
Because the San Francisco meteorological wind data used in this study were organized into large wind bins (Tables 6 and 8), the data were manually rebinned into smaller intervals to enable analysis A single representative wind speed must be assigned to each bin, typically the bin’s average (midpoint) wind speed, a choice this study employs The total variation in possible wind-speed values within a bin is the difference between the bin’s maximum and minimum speeds, Ui+1 − Ui Accordingly, the error in any wind power calculation—when converting percent-exceeded wind speeds to the wind speed used in the power curves—can be estimated by summing, over 20 occurrences, the ratio (Ui+1 − Ui)^3 divided by the cube of the bin’s average wind speed, [(Ui+Ui+1)/2]^3.
The estimated error in power due to the wind speed selection for each point tested was less than
10 percent for the 24 hour day scenario, and less than 10.7 percent for the 15 hour day scenario Also, it is generally accepted that hotwire measurements made close to a surface are within ±5
Utilizing quantifiable standards is essential in wind energy resource assessment In wind power generation, calculating wind power density and average wind power density provides a clear measure of the resource at a specific location Wind power density is the amount of available power in the wind per unit area perpendicular to the wind flow, and if a wind energy converter operated at 100 percent efficiency, this would equal the device’s potential power output per unit area perpendicular to the flow; the fundamental relation is P = ρ U^3, as described by Manwell (2003).
Wind power captured by a unit cross‑section is given by P = 1/2 ρ A U^3, where P is power, A is the cross‑sectional area through which air moves, ρ is the air density at standard temperature and pressure, and U is the wind speed perpendicular to the area; the total annual energy density per square meter can be calculated as E_year = (1/2) ρ U^3 × 8760 hours, which converts to kilowatt‑hours per square meter per year as E_kWh/m^2/year = 4.38 ρ U^3.
Results
Fox Plaza Results
The “good” wind resource points’ results of the wind‐tunnel testing for Fox Plaza are shown in Table 11, and the “great” wind resource points’ results are in Table 12
Average wind power densities were highest near or above the roof level, with the peak average reaching 1,629 W/m^2 at point 105125 in the existing setting and 1,488 W/m^2 in the cumulative setting The northern face of the building is identified as a strong wind resource because it shows high average wind power density across points 3 through 8 Among all points, point 7 experienced the greatest increase in wind power density due to local development, rising by 36%, while point 102375 had the largest decrease, dropping by 26%.
The highest turbulence intensity is observed at the current site with good to great wind, suggesting that elevated turbulence could reduce the wind energy yield Therefore, if the city chooses to develop in this area, the building’s overall power production could decline from projected levels This finding should be incorporated into development planning and energy modeling to set realistic expectations for performance.
Fox Plaza wind data for the 6am–8pm window are presented in the same format as Tables 11 and 12 The wind-tunnel test results for the “good” wind resource points at Fox Plaza appear in Table 13, while the results for the “great” wind resource points are in Table 14 Tables 15a and 15b show the ratios of the average wind speed densities for the 6am–8pm case relative to the all-hours case for each tested point.
During the 15-hour day case (6am–8pm) for the existing setting, point 105125 recorded the highest average wind power density at 2,067 W/m^2, and the same point also had the highest value in the cumulative setting at 1,778 W/m^2 Across all points, the average wind power density increased compared with the 24-hour day case, indicating stronger winds during business hours.
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Good" Wind Resource Locations - Fox Plaza
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Fox Plaza
Table 11: Results for “Good” Points at Fox Plaza (Shown Top), Using 24-Hour Wind Data
Table 12: Results for “Great” Points at Fox Plaza (Shown Bottom), Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Fox Plaza - 6am - 8pm
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Fox Plaza - 6am - 8pm
Table 13: Results for “Good” Points at Fox Plaza from 6am to 8pm (Shown Top)
Table 14: Results for “Great” Points at Fox Plaza from 6am to 8pm (Shown Bottom)
Fox Plaza 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 15a: Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for Fox Plaza
Fox Plaza 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day) (continued)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 15b Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for Fox Plaza (Continued from Table 15a)
CSAA Building Results
The “good” wind resource points’ results of the wind‐tunnel testing for the CSAA Building are shown in Table 16, and the “great” wind resource points’ results are in Table 17
Average wind power densities were highest near roof level, with the peak mean density reaching 2,476 W/m^2 at point 105125 under the existing setting and 2,181 W/m^2 at the same point under the cumulative setting In the existing setting, the northeastern and southwestern corners offer good wind resources, showing high average densities at points 44–49 and 63–69, respectively, while in the cumulative setting the southeastern corner provides a great wind resource, with high densities at points 63–69 The most pronounced change due to local development occurred at point 65, with a 103% increase, while point 102125 experienced the largest decrease, at 27%.
Within the current analysis of a good to great wind resource, the highest turbulence intensity occurs at point 815 (60%), while the highest per-square-meter cumulative value in the cumulative setting is at point 88 This suggests there could be an overall increase in power production at this building if the city chooses to develop in this area.
The reduced, full‑scale wind data for Fox Plaza for winds from 6 am to 8 pm are presented in the same format as Tables 16 and 17 The wind‑tunnel testing results for the CSAA Building at the “good” wind resource points are shown in Table 18, while the results for the “great” wind resource points are in Table 19 Tables 20a and 20b display the ratio of the average wind speed densities for the 6 am–8 pm case to the all‑hours case for each tested point.
During the 15-hour day (6 am–8 pm), the existing setting peaked at point 104125 with an average wind power density of 2,699 W/m^2, while the cumulative setting peaked at point 105125 with 2,748 W/m^2 Across all points, the average wind power density rose relative to the 24-hour day case, indicating higher winds during business hours.
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Good" Wind Resource Locations - CSAA Building
Table 16: Results for “Good” Points at the CSAA Building, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - CSAA Building
Table 17: Results for “Great” Points at the CSAA Building, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Good" Wind Resource Locations - CSAA Building - 6am - 8pm
Table 18: Results for “Good” Points at the CSAA Building from 6am to 8pm
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - CSAA Building - 6am - 8pm
Table 19: Results for “Great” Points at the CSAA Building from 6am to 8pm
CSAA Building 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 20a Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the CSAA Building
CSAA Building 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day) (continued)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 20b Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the CSAA Building (Continued from Table 20a)
Bank of America Building Results
Wind-tunnel testing results for the Bank of America Building are organized by wind resource quality: Table 21 presents the results for the good wind resource points, while Table 22 presents the results for the great wind resource points.
Average wind power densities peak near or above roof level The highest average wind power density recorded was 2,085 W/m^2 at point 81 under the existing setting, with 1,911 W/m^2 at point 101000 under the cumulative setting In the existing setting, the southwestern building face and the northern corner are identified as "good" wind resources, with high average wind power densities observed from points 33–36 and 43–46, respectively Under the cumulative setting, the southwestern face is a "great" wind resource, driven by high average wind power densities at point 37 The point showing the largest increase due to local development was point 47, with a 105% rise, while the largest decrease occurred at point 63, down 59%.
The highest turbulence intensity for the existing setting of a point with “good” or “great” wind resource was point 86 with 64 percent; and point 88 held the highest value for cumulative
With a cumulative setting of 794 watts per square meter, the site demonstrates strong solar potential If the city proceeds with development in this area, the building could see a significant increase in power production, boosting energy efficiency and supporting sustainable urban growth.
Reduced, full-scale wind data for Fox Plaza for the 6 am to 8 pm period are presented in the same format used for Tables 21 and 22 The wind-tunnel testing results for the Bank of America Building are categorized into “good” wind resource points shown in Table 23 and “great” wind resource points shown in Table 24 Tables 25a and 25b provide the ratios of the average wind speed density for the 6 am–8 pm case to the all-hours case for each tested point.
During the 6am–8pm window (the 15-hour day case) for the existing setting, point 875 exhibited the highest average wind power density at 2,781 W/m², and the same point also led in the cumulative setting with 2,650 W/m² All measurement points showed increased average wind power density compared with the 24-hour day case, indicating stronger winds during business hours.
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Good" Wind Resource Locations - Bank of America Building
Table 21: Results for “Good” Points at the Bank of America Building, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Bank of America Building
Table 22: Results for “Great” Points at the Bank of America Building, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Bank of America Building - 6am - 8pm
Table 23: Results for “Good” Points at the Bank of America Building from 6am to 8pm
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Ratio of Cumulative Setting to Existing Setting Average Wind Power Density
"Great" Wind Resource Locations - Bank of America Building - 6am - 8pm
Table 24 Results for “Great” Points at the Bank of America Building from 6am to 8pm
Bank of America Building 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 25a Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Bank of America Building
Bank of America Building 15-hour Vs 24-hour Day Analysis (6am - 8pm vs all day) (continued)
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 15-hour Day Average Wind Power Density [W/m 2 ]
Cumulative 24-hour Day Average Wind Power Density [W/m 2 ]
Table 25b Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Bank of America Building (Continued from Table 25a)
Folsom and Main Street Buildings’ Results
3.2.1 Folsom and Main East Results
Results from wind-tunnel testing identify two levels of wind resource points for the Folsom and Main East buildings: the "good" points, with results shown in Table 26, and the "great" points, with results shown in the following table.
27 Average wind power densities were highest near or above the roof level The highest average wind power density was 750 Watts per square meter at point 105500 The only “great” wind resource sites are located at or above the rooftop level of the building, and the only
“good” wind resource site not on or above roof level is point 48 with an average wind power density of 409 Watts per square meter
Among locations with good or great wind resource, the highest turbulence intensity was recorded at point 86, at 60% Across the measurement sites in Folsom and Main East, the average wind power density was 235 W/m^2.
The reduced, full‐scale data for Fox Plaza for winds from 6am to 8pm are displayed in the same manner as Tables 26 and 27 The “good” wind resource points’ results of the wind‐tunnel testing for the CSAA Building are shown in Table 28, and the “great” wind resource points’
The point with the highest average wind power density during the hours of 6am to 8pm (the 15‐ hour day case) was point 105500 which had a value of 942 Watts per square meter All points showed an increase in average wind power density from the 24‐hour day case, demonstrating that the winds are higher during business hours
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Good" Wind Resource Locations - Folsom and Main East
Table 26: Results for “Good” Points at Folsom and Main East, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Great" Wind Resource Locations - Folsom and Main East
Table 27: Results for “Great” Points at Folsom and Main East, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Good" Wind Resource Locations - Folsom and Main East - 6am - 8pm
Table 28: Results for “Good” Points at the Folsom and Main East Building from 6am to 8pm
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Great" Wind Resource Locations - Folsom and Main East - 6am - 8pm
Table 29: Results for “Great” Points at the Folsom and Main East Building from 6am to 8pm
Existing 15-hour Day Average Wind Power Density [W/m 2 ]
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Table 30a; Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Folsom and Main East Building
Folsom and Main East (continued)
Existing 15-hour Day Average Wind Power Density [W/m 2 ]
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Table 30b: Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Folsom and Main East Building (Continued from Table 30a)
3.2.2 Folsom and Main West Results
Wind-tunnel testing for the Folsom and Main East building identifies two categories of wind resource points The results for the "good" wind resource points are shown in Table 31, while the results for the "great" wind resource points are in the accompanying table.
Average wind power densities peak near or above roof level, with the highest value—756 W/m²—recorded at point 105125 The analysis shows that only sites at or above the rooftop level qualify as “great” wind resources, while the only “good” resources outside this zone are point 40 and point 49, with average wind power densities of 485 and 451 W/m², respectively.
The highest turbulence intensity for a point with “good” or “great” wind resource was point
104000 with 77 percent The average of the measurement locations’ average wind power density for Folsom and Main West was 233 Watts per square meter
Reduced, full-scale data for Fox Plaza for winds from 6 a.m to 8 p.m are presented in the same manner as Tables 31 and 32 The results for the “good” wind resource points from the wind-tunnel testing of the CSAA Building are shown in Table 33, and the results for the “great” wind resource points are presented in the subsequent table.
During the 15-hour day case (6am–8pm), the highest average wind power density occurred at point 105125, with a value of 947 W/m^2 All points showed an increase in average wind power density compared with the 24-hour day case, indicating that winds are higher during business hours.
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Good" Wind Resource Locations - Folsom and Main West
Table 31: Results for “Good” Points at Folsom and Main West, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Great" Wind Resource Locations - Folsom and Main West
Table 32: Results for “Great” Points at Folsom and Main West, Using 24-Hour Wind Data
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Good" Wind Resource Locations - Folsom and Main West - 6am - 8pm
Table 33: Results for “Good” Points at the Folsom and Main West Building from 6am to 8pm
Average 1kW Wind Turbine Annual Energy Production [kW-hr/year]
Aerotecture WEC Annual Energy Production [kW-hr/year]
Estimated Error in Ave Wind Power Density [%]
"Great" Wind Resource Locations - Folsom and Main West - 6am - 8pm
Table 34: Results for “Great” Points at the Folsom and Main West Building from 6am to 8pm
Existing 15-hour Day Average Wind Power Density [W/m 2 ]
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Table 35a: Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Folsom and Main West Building
Folsom and Main West (continued)
Existing 15-hour Day Average Wind Power Density [W/m 2 ]
Existing 24-hour Day Average Wind Power Density [W/m 2 ]
Table 35b: Ratio of Average Wind Power Densities of the 6am to 8pm Case to the 24-Hours per
Day Case for the Folsom and Main West Building (Continued from Table 35a)
Results in Graphical Form
Data presented in the sections that follow are shown graphically to complement the preceding tables of average wind power density ratios Photos from the wind-tunnel tests, including the actual models and the local areas used, are overlaid with color-coded markers that indicate performance categories: yellow dots with red text and outlines denote “great” locations, green dots with black text and outlines denote “good” locations, and other markers indicate “poor” locations per the tables.
In this system, a "good" location is distinguished from a "poor" location, with the latter indicated by a white dot featuring black text and an outline The numbers shown inside the dot correspond to the point number of that measurement location, and these numbers also match the point numbers listed in the preceding tables.
Point placements are approximate and not to scale; for better visualization, some point positions in the photos were slightly adjusted, so the analysis is qualitative rather than strictly quantitative A more detailed and precise description of the point locations appears in Tables 1 through 5 The actual graphical figures can be found in Section 3.3.6.
For the Fox Plaza Building, the results from Tables 15a and 15b are presented graphically in Figures 22–25; Figure 22 shows the results for the existing setting and Figure 23 shows the results for the cumulative setting, with both figures assuming the WECs run continuously all day and night, while Figure 24 provides the graphical results for the existing setting and Figure 25 provides the results for the cumulative setting, with both of these figures assuming the WECs run only from 6 a.m to 8 p.m.
Fox Plaza features a slender protruding element on its north and south facades and the roof, designed to promote wind flow acceleration when not obstructed by surrounding structures Figures 22–25 illustrate how this feature can yield higher annual average wind power density on the building’s north or south facades or near its corners.
Figure 22 indicates that the best locations to place a wind energy converter (WEC) are above the roof level and on the building’s north face Because the prevailing winds come from the northwest, the CSAA building would be upwind for many tested wind directions, likely creating a region of accelerated flow along its north facade.
Installing wind energy converters on the building’s south face is not advisable due to low annual average wind power density, a consequence of the wind being redirected around the building toward the north face The north face presents superior wind potential, driven by the building-induced flow dynamics Other areas also show local flow accelerations that increase wind power density, with the upper southwest corner emerging as a particularly favorable location.
Figure 23 shows that annual average wind power densities decrease slightly from the existing to the cumulative settings due to area development, with an upwind building at One Polk located between the CSAA building and Fox Plaza, and an addition to Fox Plaza that may block some wind from accelerating around the building A large development on 10th Street and Market Street next to Fox Plaza could be expected to create a wind-tunnel effect along Market Street, yet it appears to disturb the flow in a way that restricts wind acceleration near the surface of Fox Plaza.
Figure 24 resembles Figure 22 but uses wind data limited to 6 am to 8 pm; because higher winds occur during this period, the values are slightly higher than those in Figure 22, though the overall trends remain the same Likewise, Figure 25 presents the cumulative setting for the same 15-hour day, and its trends are similar to those observed in Figure 23.
Figures 26–29 present the results from Tables 20a and 20b for the CSAA Building in graphical form Figure 26 shows the results for the existing setting, while Figure 27 shows the results for the cumulative setting; in both figures the Wind Energy Converters (WECs) are assumed to operate from 6 AM to 8 PM.
The CSAA building's rooftop penthouse elevates measurement point 105 above the other points, and because there are relatively few upwind structures of comparable height for San Francisco's most frequently occurring winds, point 105 exhibits some of the highest annual average wind power densities This distinctive rooftop feature can also induce local flow accelerations at nearby rooftops.
Figure 26 indicates that the most favorable locations for wind energy converters (WECs) are at rooftop level or higher, with the northeast and southwest corners also offering viable siting options The northeast corner may experience local flow acceleration as wind interacts with the smaller structure attached to the CSAA building In contrast, the southwest corner provides relatively unobstructed exposure from tall upwind structures for the most common wind directions.
Figure 27 illustrates how annual average wind power densities respond to potential local developments in the area Overall, these developments increase the available wind power, with higher densities observed in the southwest corner of the building and at several rooftop locations The only upwind development is a small building located across the corner of the intersection at 77 Van Ness, but this structure—and the building at One Polk—together with other downwind developments, contribute to more favorable wind conditions at this location, even though they produce less favorable conditions at Fox Plaza.
Figure 28 is similar to Figure 26, but the analysis uses wind data only from 6am to 8pm Because winds are typically higher during this window, the values are slightly elevated compared with Figure 26, yet the overall trends remain the same Figure 29, which shows the cumulative setting for the same 15-hour day, also exhibits trends that align with Figure 26.
3.3.3 Bank of America Building Graphical Results
Results from Tables 25a and 25b are presented graphically in Figures 30–33 for the Bank of America Building Figure 30 shows the results for the existing setting, while Figures 31–33 illustrate the outcomes under alternate configurations, enabling a direct comparison of performance across scenarios The graphical presentation provides a concise visual summary that supports the accompanying analysis and highlights how changes in conditions affect the building's performance.