In the last few years, the demand for testing extra-long cables, such as submarine cables has grown rapidly. The existing testing methods have been complemented by a new testing technology called DRT (Differential Resonance Technology). This testing method enables testing of extra-long cables by comparably small and light-weight equipment using a low frequency for the test voltage, e.g. 0.1 Hz up to 5 Hz. This leads to a significant decrease of the required power of the test source (P. Mohaupt and A. Bergman in CIGRE 2010). In a resonant circuit only the losses of the generator’s individual components, specifically the high voltage reactor, have to be covered by the mains. The testing power itself remains fully compensated. Typical ratios between the testing power and the input power of resonant test systems start at 50 and go up to 100, depending on the load. Unfortunately, voltage generation based on inductive generation principles such as resonant circuits cannot economically be used for frequencies below 10 Hz due to the massive iron cores needed for such a low frequency. The DRT method for the generation of low frequency high voltage is based on a high frequency voltage whose amplitude is modulated by the desired low frequency. Using a resonator, which is tuned to the high frequency, and a demodulator, the desired low frequency high voltage can be generated (P. Mohaupt and A. Bergman in CIGRE 2010; P. Mohaupt and T. Mehl in Jicable 2011). The input power required—and in direct relation to this the size and weight of the equipment—is significantly smaller than for other existing methods. In order to optimize the operation performance of the DRT system, this paper describes mathematical methods and algorithms, which have already been implemented and tested in a DRT test set. The basis for these algorithms is a mathematical description of the system based on an envelope model. Using this mathematical description of the nonlinear system behavior, a systematic analysis of the performance and the limits of the system can be given. The theoretical approach was experimentally proven by measuring the output voltage and the input power of a prototype unit ultimately designed to produce 200 kV rms. A first test was performed at SP Technical Research Institute of Sweden, using their reference measurement system for very low frequency (VLF) S. Bergman and A. Bergman (Proc. CPEM Conf. Dig., pp. 682–683, 2010; IEEE Trans. Instrum. Meas. 60:2422–2426, 2011) to measure the high VLF voltage. The reference measurement system provides a traceable uncertainty of down to 0.04 % over a voltage range up to 200 kV rms. The frequency range of the reference system is from 0.1 Hz up to 50 Hz. This system permits acquisition of complete wave-forms that can be analysed for harmonic content and/or THD (Total Harmonic Distortion). Further tests are planned, where the connected load will be increased to the specified maximum 1 μF at 200 kV, and the characteristics will be explored both as regards to output voltage quality, input power requirements and distortion on the input current.
Read full abstract