City and County of San Francisco Wind Resource Assessment Project

Publication Number: 500-04-066
Publication Date: October 2004

The executive summary, abstract and table of contents for this report are available below. This publication is available as an Adobe Acrobat Portable Document Format Files. In order to download, read and print PDF files, you will need a copy of the free Acrobat Reader software installed in and configured for your computer. The software can be downloaded from Adobe Systems Incorporated's website.

---

Executive Summary

Introduction

In November 2001, San Francisco voters approved two bond measures earmarking $100 million for installation of wind energy generation, solar energy generation, and energy efficiency technologies on city-owned property. In the case of wind energy generation, optimum allocation of resources will hinge on availability of wind energy resource information. To date little information has been developed concerning wind energy resource variability within the City of San Francisco. The California Energy Commissionís (Commission) Public Interest Energy Researchís (PIER) Renewable Energy program undertook this urban wind resource assessment project to help the City and County of San Francisco optimize its future investment in distributed wind energy generation capacity.

To determine whether there may be economically feasible wind resources available at several select sites within the City/County geographic boundaries an urban wind resource assessment research project was conducted. Project objectives were to:

• Measure wind speeds at five sites within the City/County of San Francisco,


• Identify relationships between the measured wind speed data at the five monitored sites and wind speed data for the San Francisco International Airport (SFO), a nearby weather station for which a long-term historical record exists,


• Develop 8,760-hour wind speed datasets reflecting typical, or "normal" wind energy resources for each of the five monitored sites, and


• Estimate annual energy production for a generic 10 kW wind turbine installed on a 98-foot tower at each of the five monitored sites.

The information gathered and developed as a result of this project will be used by the City/County of San Francisco staff in ascertaining the feasibility of developing urban wind projects. The scope of this project was limited to quantifying the wind resources at select urban locations within the City/County of San Francisco.

Project Approach

Before visiting any prospective monitoring sites the various factors influencing urban wind energy resource were compiled. During subsequent site selection meetings and site visits these criteria were used to assess the suitability of particular sites. Two parallel paths were pursued to accomplish monitoring system selection and installation. First, a local meteorological consulting firm with many yearsí experience in wind resource assessment both in California and elsewhere was invited to join the project team. Data for three sites were collected via this path. Data for two additional sites were obtained from secondary sources and incorporated into the analysis.

Analysis of wind speed data included development of typical 8,760-hour wind speed datasets for a "normal" wind energy resource year, and estimation of annual energy production for hypothetical generic 10 kW wind turbines installed on 98-foot towers at each of the 5 monitored sites. The normalization process entailed three fundamental steps.

1. Define a normal 8,760-hour SFO wind speed dataset based on long-term historical wind speed data,


2. Identify relationships between SFO wind speeds and coincident wind speeds at each monitored site, and


3. Estimate 8,760-hour normal wind speed datasets for each of the monitored sites.

Depending on seasonal availability of measured data for both SFO and a monitored site one of two different analytic methodologies was used to identify a relationship between coincident wind speeds at the two locations. In 80% of cases sufficient data were available to enable use of a whole-day substitution approach akin to the method used to define the normal 8,760-hour SFO wind speed dataset. Days of measured data for the monitored sites were selected for inclusion in the normal wind speed datasets based on daily wind energy totals for SFO. A principal advantage of this approach is that it retains all of the site-specific diurnal wind speed characteristics in the measured data.

When less than two months of data were available for a season a single-hour substitution approach was used to avoid estimation of normal wind speed data with unsatisfactory diurnal characteristics. This approach entailed using measured data to determine distributions of monitored-site wind speeds corresponding to particular SFO wind speeds. For individual hours in the normal wind speed dataset a monitored-site wind speed was probabilistically selected from a distribution of wind speeds corresponding to the SFO wind speed for that hour.

Wind speed data used in the analysis were collected at a variety of heights ranging from 24 feet to 60 feet. Initial results of the normalization analysis were adjusted such that final normal wind speed estimates are based on a sensor height of 33 feet. Finally, assumed performance characteristics for a generic wind turbine were combined with normal wind speed datasets in a calculation of annual energy production. For this analysis the 33-foot normal wind speeds were adjusted upward to estimate wind speeds for 98-foot tall towers.

Project Outcomes

Five sites were selected for inclusion in the wind resource assessment project. The characteristics of the monitored sites are summarized in Table 1. Data were obtained from several sources, and the monitoring period extended from July 2002 through November 2003. Data availability varied from site to site. The large quantity of SFO data was required for the normalization analysis.

Annual average wind speeds corresponding to each of the normal 33-foot 8,760-hour wind speed datasets developed for this project are presented in Table 2 alongside an estimate of the corresponding wind power class and generic turbine annual energy production. Annual energy production was estimated for a generic 10 kW wind machine installed on a 98-foot tower. The annual energy production estimates range from 7,371 kWh/yr to 15,632 kWh/yr.

Conclusions and Recommendations

With the exception of Twin Peaks the wind energy resources at the monitored sites appear to be quite modest relative to levels customarily associated with wind energy generation development. From the standpoint of economic feasibility of prospective wind energy generation facilities a wind power class equal to 1 is generally considered 'very poor' or 'poor'. In the case of small turbines, under the right circumstances (e.g., valuation of generated electricity at a high retail rate) a wind power class of 2 may be sufficient to justify development on financial grounds.

City and County of San Francisco staff responsible for making decisions concerning investment in wind energy generation equipment at the five monitored sites should incorporate results of this urban wind resource assessment project into their technical and financial analyses. It is conceivable that now or at some time in the future installation of wind energy generation systems on the roofs of high-rise buildings may be technically feasible. In this eventuality the Commission or others may want to augment the present research with an urban high-rise rooftop wind resource assessment.

This project developed site-specific wind energy resource information for five prospective wind energy generation facility sites within the City of San Francisco. This information will help the City and County of San Francisco make better decisions regarding their future investments in renewable energy distributed generation, thereby benefiting California.

Benefits to California

Economic, political, electric transmission system loading, Renewable Portfolio Standards (RPS), fuel diversity, and other factors have combined to create an atmosphere of increased interest in developing wind energy generation facilities in California and elsewhere throughout the US and Europe. However, economic feasibility of particular, prospective wind energy projects is highly site specific. In the absence of satisfactory information concerning wind energy resource there is an elevated risk of investing in wind energy generation facilities that are located sub-optimally. This project developed site-specific wind energy resource information for five prospective wind energy generation sites within the City of San Francisco. This information will help the City and County of San Francisco to make better decisions regarding their future investments in renewable distributed generation, thereby benefiting California.

---

Abstract

The California Energy Commissionís (Commission) Public Interest Energy Research (PIER) Renewables program element undertook an urban wind resource assessment project to help the City and County of San Francisco optimize its future investment in distributed wind energy generation.

Five prospective wind energy generation sites within the City and County were selected. Wind speed data were collected for periods ranging from 5 to 14 months. Sites included in the analysis are Twin Peaks, Treasure Island, Hunters Point, S.F. Zoo, and Pier 39. Wind data over an 11 year period from the S.F. International Airport (SFO) weather station were used in a normalization analysis for each of the monitored sites. For SFO and for each of the monitored sites, an 8,760-hour wind speed dataset was developed to represent typical, or "normal" wind energy resource. For each site annual energy production estimates were calculated for a generic 10 kW horizontal axis wind turbine on a 98-foot tower. Annual energy production estimates ranged from 7,371 kWh/yr for Hunters Point to 15,632 kWh/yr for Twin Peaks. With the exception of Twin Peaks the wind energy resource at the monitored sites appear to be quite modest relative to levels customarily associated with wind energy generation development. Stakeholders responsible for making decisions concerning investment in wind energy generation equipment at the five monitored sites should consider limitations of normalization analyses relying on short-term datasets when incorporating the results of this wind resource assessment into their turbine-specific performance, economic and financial analyses.

The findings of this project will help the City and County of San Francisco make better decisions regarding their future investments in small wind-based renewable energy distributed generation, thereby benefiting California.

Keywords:

• California,
• Urban,
• Wind,
• Data,
• Analysis,
• Energy,
• San Francisco,
• Small turbine,
• Renewable

---

Table of Contents

Abstract

Executive Summary

1.0 Introduction

2.0 Project Approach

3.0 Project Outcomes

4.0 Conclusions and Recommendations

Appendices

Appendix A: Wind Speed Data

Appendix B: Task 2.1 Site tour and Selection Report and Photographs

Appendix C: Site Installation Worksheets and Photographs

List of Figures

Figure 1 Normalized 33-foot Monthly Average Wind Speeds

Figure 2 Summary of 11-Year SFO Wind Speed Data Record

Figure 3 Summary of 11-Year SFO Wind Power Data Record

Figure 4 Example of Hourly Wind Speed Correspondence Distribution (Pier 39 Wind Speeds Coincident with 10 MPH SFO Wind Speed)

Figure 5 Generic Wind Turbine Power Versus Wind Speed Curve

Figure 6 Distribution of Instantaneous Wind Speeds around 20 MPH Average - Minimum, Median, and Maximum Observed Wind Speed Standard Deviation

Figure 7 Calculated Energy Factors Based on Measured Data

Figure 8 Energy Output Fractions Based on Observed and Assumed Hourly Wind Speed Standard Deviations

Figure 9 Observed Monthly Average Wind Speeds

Figure 10 Observed Diurnal Wind Speed Patterns - SFO

Figure 11 Observed Diurnal Wind Speed Patterns - Hunters Point

Figure 12 Observed Seasonal Diurnal Wind Speed Patterns - S.F. Zoo

Figure 13 Observed Seasonal Diurnal Wind Speed Patterns - Twin Peaks

Figure 14 Observed Seasonal Diurnal Wind Speed Patterns - Treasure Island

Figure 15 Observed Seasonal Diurnal Wind Speed Patterns - Pier 39

Figure 16 Normalized 33-foot Diurnal Wind Speed Patterns - SFO

Figure 17 Normalized 33-foot Monthly Average Wind Speeds

Figure 18 Normalized 33-foot Diurnal Wind Speed Patterns - Hunters Point

Figure 19 Normalized 33-foot Diurnal Wind Speed Patterns - S.F. Zoo

Figure 20 Normalized 33-foot Diurnal Wind Speed Patterns - Twin Peaks

Figure 21 Normalized 33-foot Diurnal Wind Speed Patterns - Treasure Island

Figure 22 Normalized 33-foot Diurnal Wind Speed Patterns - Pier 39

List of Tables

Table 1 Characteristics of Five Monitored Sites (& SFO)

Table 2 Annual Average Wind Speeds and Wind Power Classes for Monitored Sites (& SFO)

Table 3 Constituent Year-Months of Normal SFO Wind Speed Dataset

Table 4: Basis of Season Assignments

Table 5: General Form of Whole-Day Substitution Lists

Table 6: Summary of Normalization Analytic Methodologies

Table 7: Sensor Heights at Monitored Sites

Table 8: Characteristics of Five Monitored Sites (& SFO)

Table 9: Source and Quantity of Wind Speed Data

Table 10: Normal 33-Foot Wind Speed - Annual Averages

Table 11: Annual Energy Production for Generic Turbines on 98-foot Towers

Table 12: Wind Power Class Definitions (33-Foot Sensor Height)

Table 13: Annual Average Wind Speeds and Wind Power Classes for Monitored Sites (& SFO)