A group of 35 wind turbines on the Castello Ranch in Altamont Pass, California, was investigated to quantify array wake effects (losses in energy production due to operation of upwind turbines) and the factors influencing them. Approximately 65 hours of field measurements were made in summer 1986, with turbine energy production and wind velocity data recorded for various scenarios of array operation. Customized software and hardware were developed and installed by Fayette to facilitate these measurements. The existence of wake effects was fairly well established. Relative energy-production losses averaged 6% at the second row, when the first row was operating, and 7 to 8% at the third row, when the first two were operating. Apparently, then, the impact of the first row on the third (at a 21-rotor-diameter distance) was minimal. Ambient wind speed did not appear to affect the relative wind speed pattern within the array due to wakes, but because of the shape of the performance curve, it did affect relative energy production losses (particularly for the low-RPM mode of machine operation). The influences of ambient atmospheric conditions, such as stability, turbulence, and shear on the array wakes, were also investigated by testing over a range of themore » conditions available during a typical 24-hour day at the test site. None of these variables showed any significant effect on the degree of wake-induced energy losses. While the results of this study apply only to this specific array and type of wind turbine, the methodology could be applied to study wake effects at other wind farms. 6 refs., 7 figs., 20 tabs.« less