New Wind Energy Resource Maps of California

Publication Number: 500-02-055F
Publication Date: November 2002

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

This report describes a wind-mapping project conducted by TrueWind Solutions for the California Energy Commission. The purpose of this contract is to develop more accurate and reliable wind resource maps for California using state-of-the-art numerical modeling techniques and site validation data. This effort not only updates the existing annual wind resource map for California produced in the late 1970’s, but includes several enhancements, including the incorporation of new meteorological, geographical and terrain data that have been collected but were unavailable when the original map was produced. Validation of map results also has been performed in conjunction with the modeling effort. These new maps will help to better define wind corridors as well as identify new potential sites for wind energy integration.

Objectives

The specific technical objectives were as follows:

• Access state-of-the-art numerical modeling techniques to predict wind speed and power at various heights and to refine those predictions with validation data gathered from various meteorological towers throughout the state.
• Create updated, high-resolution wind resource maps for California, including maps of wind speed and wind power at varying heights.
• Create maps depicting the seasonal variability of the wind resource.
• Create computer files of wind resource data that are ready for immediate integration into State cartography system (GIS format).

Approach

The MesoMap system consists of an integrated set of atmospheric simulation models, databases, and computers and storage systems. At the core of MesoMap is MASS (Mesoscale Atmospheric Simulation System), a numerical weather model that simulates the complete physics of the atmosphere. MASS is coupled to a simpler wind flow model, WindMap, which is used to refine the spatial resolution of MASS to account for the local effects of terrain and surface roughness. MASS simulates weather conditions over the region for 366 historical days randomly selected from a 15-year period. When the runs are finished, the results are input into WindMap for the final mapping stage. In this project, the MASS model was run on a grid spacing of 2 kilometer (KM), and WindMap on a grid spacing of 200 meters (M).

The preliminary wind maps produced by MesoMap were thoroughly validated by TrueWind Solutions in collaboration with NREL and independent consultants. The validation process used data for 262 stations from a wide variety of sources, including airports, ocean buoys, and towers instrumented specifically for wind resource assessment.

The validation concluded that the initial wind speed estimates at 50 M height, before any adjustments, were accurate to within a standard error of 0.4-0.6 meters per second (M/S), or 6% to 8%. Qualitatively, the preliminary maps presented an accurate overall picture of the wind resource, but tended to underestimate winds in certain well-known wind corridors and to overestimate winds on mountaintops. We believe that the most important source of error is the finite grid scale of the MASS simulations, a consequence of the size of the state and limitations of budget and schedule, which resulted in an inability to fully resolve passes through mountains or the blocking of low-level winds by mountain ranges.

Following the validation, the wind maps were adjusted to improve the agreement with the data, and the revised maps were reviewed once more. We avoided adjusting the maps for specific points, but rather attempted to correct for clear patterns of error occurring over sizable regions. The speed adjustment ranged from a decrease of up to 15% to an increase of up to 25%. Most adjustments were around 5-10% in either direction.

Outcomes

Using our MesoMap system, which was developed over four years ago, we have produced new maps of California’s mean wind speed and power for a range of heights above ground on a 200 M grid. We have also produced data files of the predicted frequency, mean speed, and energy by direction, as well as the seasonal characteristics of the resource. (The data files are provided separately on a CD-ROM.) The validation process provided a mechanism for objectively comparing the wind maps against data from a wide variety of sources, for estimating the map errors, and for independent review of the maps by leading wind energy consultants and government researchers. The final, published wind maps have been adjusted to reflect the validation findings and, consequently, represents the best current estimate of California’s wind resources, at a very high resolution.

Conclusions, Recommendations, and Benefits to California

The preliminary map estimates correlated well with data obtained for 266 towers and extrapolated to a height of 50 M, indicating that the method overall is sound.

Aside from confirming the existence of several well-known wind resource areas, the maps point to a number of other promising sites, some already known to wind energy experts and others perhaps previously unsuspected.

The report concludes with some recommendations for further research. The main recommendations are: (1) High-resolution mapping of promising areas to better resolve mountain blocking and channeling effects and consequently to improve the accuracy of the wind resource estimates; (2) research to improve methods of simulating the stable nighttime boundary layer and its effects on wind speeds at the hub height of turbines; (3) the development of an improved data base of land cover and surface roughness throughout the state; and (4) a new program of measurement of winds at or near the hub height of large turbines using tall towers, sodar, and other tools.

The project’s potential benefits to California include the following:

• Improving energy cost/value of California's electricity by accelerating the initial stages of wind project development with reliable wind maps that substantially reduce risk and siting barriers for new developers.
• Improving electricity reliability/quality/sufficiency of California's electrical system by providing the most current and reliable information on wind resource data for the State.
• Providing data for identifying new potential sites for wind energy integration.
• Providing high-resolution wind data useful for forecasting and optimizing wind resource management.
• Strengthening the California economy by encouraging development of new wind sites and job opportunities.
• Providing greater choices for California consumers by supporting the expansion of clean energy resources and by providing data to make the resources more manageable.
• Improving the environment, public health and safety by providing the most reliable and updated data for basing decisions and integrating with existing infrastructure (transmission) and planning strategies.

---

Abstract

The MesoMap system has been used to produce new wind energy resource maps and databases for the State of California on a 200 M grid. The wind resource maps confirm the locations of several major wind resource areas and also point to the existence of new areas that may not be widely known. An objective validation process, carried out using data from over 260 sites throughout the state and advice from independent consultants, concluded that the preliminary wind resource estimates were accurate to a standard error of 0.4-0.6 M/S (6-8%). Adjustments to the maps were subsequently made to improve the match to the data. The adjustments included increases in the predicted resource within some known wind resource areas, and reductions along some mountaintops and in some coastal areas. The finite grid scale of the model is suspected of being the main cause of the observed errors. The project has resulted in the most accurate current assessment of wind resources in California at a scale suitable for identifying promising sites for wind energy projects. Several topics for further research are suggested to help improve the accuracy of the maps in promising resource areas.

---

Table of Contents

Preface

Executive Summary

Abstract

1.0 Introduction

2.0 Approach

3.0 Outcomes

4.0 Conclusions

Appendices

Appendix I: Guidelines for Use of Maps

Appendix II: The Data CD-ROM

Appendix III: Maps

Tables

Table 1. Land Cover Classifications and Surface Roughness

Table 2. Validation Summary

Table 3. Land area in each wind speed or power band, in square miles

Figures

Figure 1. Scatter plot of predicted and measured/extrapolated mean wind speeds.

Figure 2. The relationship between the estimated model error and the uncertainty in wind shear.

Figure 3.Effect of MASS grid scale on simulated winds through San Gorgonio Pass.

Figure 4. Effect of MASS grid scale on simulated winds through Pacheco Pass.

Figure 5. Wind speed adjustment factor.

Figure 6. Adjustment in map power needed to eliminate the discrepancy with observations.