Demonstrating the service performance of gas-insulated HVDC equipment

In the recent years, new designs of gas-insulated HVDC equipment up to ±550 kV DC have been developed for future grid extensions. Common engineering practice is to demonstrate their service operation within a pilot installation. At large and complex infrastructure projects – or rather high voltage levels – the integration of pilots is challenging. This results mainly from the unknown reliability of the new equipment, for which reason the estimated risk may prevent its implementation, especially at large investment costs. Up to now, the few installations up to ±250 kV DC have collected only little experience for gas-insulated HVDC equipment. Additionally, critical effects at long-term operation with this type of equipment are known, namely the charging of the gas-solid interfaces and the motion of mobile particles at DC voltage. Within this context CIGRE Technical Brochure 842 proposes a prototype installation test, in which the operation of the gas-insulated HVDC equipment is proven at typical DC grid stresses within a one-year long-term test. Such approach is a safer solution to prove the readiness for grid installations of newly developed gas-insulated HVDC equipment rather than directly installing pilots in the grid. One particular development in the last decade is the DC gas-insulated line (GIL) for rated DC voltages up to ±550 kV. Its readiness was tested within a research project, which the presented dissertation is based on. The recommendations for the CIGRE prototype installation test are used for this purpose. The resulting device under test had a loop length of 100 m, which is challenging to test in the laboratory. A modified test approach is presented to be able to execute such and future long-term tests more efficiently than with existing test methods and equipment. The first optimization is the application of conductor heating by DC current rather than the conventional conductor heating by AC current. The second optimization is the use of oscillating rather than standard lightning and switching impulse voltages superimposed to the DC voltage. The developed concept and equipment are put to the test during the executed long-term test on the DC GIL. The feasibility and benefits of the modified test approach as well as the constructed test equipment is thereby proven. The test results observed at the investigated DC GIL are also a new reference for the comparatively new gas-insulated HVDC technology. Gained knowledge in combination with existing knowledge from literature are summarized in this dissertation.