| This dissertation is concerned with several aspects of radio frequency noise caused by corona discharges on electric power lines. First, a concise history of the problem is presented spanning a time period of more than sixty years. Next, an improved method of calculating the fields for corridors with multiple circuits is presented. The method treats each three phase circuit in a power line corridor separately in the noise field calculations. The results compare to the more cumbersome method of collectively solving a multiple circuit corridor to within 1 dB for typical corridors and measurement distances.; Following is a comparison of empirically determined radio noise generation functions (used with a common field calculation formula) to long term radio noise data. Each generation function is modified by an additive constant to minimize the rms error between measured and calculated data. The BPA generation function, when modified and used with the field calculation formula, provides the best agreement with the long term data, having an rms error of 1.7 dB for the experiments analyzed.; The problem of applying high and medium frequency field calculation methods is presented next. The methods are applied to the problem of determining the induced open circuit voltage on an arbitrarily oriented biconical dipole antenna.; Finally, a detailed evaluation of corona streamers is offered. Using assumed charge distributions for models, multipole expansions are performed to validate the commonly used dipole approximation for a corona streamer. The models provide a theoretical explanation as to why positive streamers induce more high frequency current in wires than do negative streamers. The models are further used to predict induced current waveforms that agree with experimental data in related work. |
