Analyzing the Effects of Temporal Wind Patterns on the Value ofWind-Generated Electricity at Different Sites in California and theNorthwest

Fripp, Matthias ; Wiser, Ryan

Lawrence Berkeley National Laboratory

关键词: Anemometers;    Capacity;    Imports;    Energy Policy;    Wind Turbine Arrays;   

DOI  :  10.2172/883800
RP-ID  :  LBNL--60152
RP-ID  :  DE-AC02-05CH11231
RP-ID  :  883800

美国|英语

来源: UNT Digital Library

【 摘 要 】

Wind power production varies on a diurnal and seasonal basis. In this report, we use wind speed data modeled by TrueWind Solutions, LLC (now AWS Truewind) to assess the effects of wind timing on the value of electric power from potential wind farm locations in California and the Northwest. (Data from this dataset are referred to as ''TrueWind data'' throughout this report.) The intra-annual wind speed variations reported in the TrueWind datasets have not previously been used in published work, however, so we also compare them to a collection of anemometer wind speed measurements and to a limited set of actual wind farm production data. The research reported in this paper seeks to answer three specific questions: (1) How large of an effect can the temporal variation of wind power have on the value of wind in different wind resource areas? (2) Which locations are affected most positively or negatively by the seasonal and diurnal timing of wind speeds? (3) How compatible are wind resources in the Northwest and California with wholesale power prices and loads in either region? The latter question is motivated by the fact that wind power projects in the Northwest could sell their output into California (and vice versa), and that California has an aggressive renewable energy policy that may ultimately yield such imports. Based on our research, we reach three key conclusions. (1) Temporal patterns have a moderate impact on the wholesale market value of wind power and a larger impact on the capacity factor during peak hours. The best-timed wind power sites have a wholesale market value that is up to 4 percent higher than the average market price, while the worst-timed sites have a market value that is up to 11 percent below the average market price. The best-timed wind sites could produce as much as 30-40 percent more power during peak hours than they do on average during the year, while the worst timed sites may produce 30-60 percent less power during peak hours. (2) Northwestern markets appear to be well served by Northwestern wind and poorly served by California wind; results are less clear for California markets. Both the modeled TrueWind data and the anemometer data indicate that many Northwestern wind sites are reasonably well-matched to the Northwest's historically winter-peaking wholesale electricity prices and loads, while most California sites are poorly matched to these prices and loads. However, the TrueWind data indicate that most California and Northwestern wind sites are poorly matched to California's summer-afternoon-peaking prices and loads, while the anemometer data suggest that many of these same sites are well matched to California's wholesale prices and loads. (3) TrueWind and anemometer data agree about wind speeds in most times and places, but disagree about California's summer afternoon wind speeds: The TrueWind data indicate that wind speeds at sites in California's coastal mountains and some Northwestern locations dip deeply during summer days and stay low through much of the afternoon. In contrast, the anemometer data indicate that winds at these sites begin to rise during the afternoon and are relatively strong when power is needed most. At other times and locations, the two datasets show good agreement. This disagreement may be due in part to time-varying wind shear between the anemometer heights (20-25m) and the TrueWind reference height (50m or 70m), but may also be due to modeling errors or data collection inconsistencies.