Five 230-kV disconnect switches were selected for testing and evaluation. Switches of this rating were known to be vulnerable to the effects of earthquake shaking. The class of the switches selected included two horizontal-break and three vertical-break switches. Switches with both porcelain and composite polymer insulators, and switches with both cast aluminum and welded steel base hardware were tested. First, an individual pole (out of three) of each switch was mounted directly on the simulator platform. Static and resonance search tests were conducted to determine the dynamic properties of the posts and to assist in preparing simple analytical models of the posts. Earthquake tests were then conducted to see if the poles were robust. Following these tests, three-pole (phase) switches were mounted on a stiff, low-profile frame, similar to a frame proposed by PG&E for new construction, and tested. Static and resonance search tests were conducted to determine the dynamic properties of the switch posts.
Triaxial earthquake tests were used to qualify the switches on the frame. Two sets of part spectrum-compatible ground motion records were derived from the near-field motions recorded during the 1978 Tabas, Iran, earthquake for the earthquake-simulator studies. Neither of the horizontal-break switches was qualified to the High Level on the stiff frame. The vertical-break switches were qualified to the High Level on a stiff frame provided that welded steel spacers were used at the base of the insulators. The most vulnerable components in the switches were the cast aluminum spacers at the base of the switch posts, the welded post-blade connection, and the bolted connections at the base of the posts.
Single-degree-of-freedom models of the sw itch insulator posts were developed using experimental data. These models predicted reasonably well the displacement response of the switches mounted on the test frame. These mode ls were then used to evaluate the likely amplification of response of switches when mounted on frames of different heights and stiffnesses. Four frames were selected for study: two in-service (one tall and one short), and two proposed for new construction (one tall and one short). The short braced frame proposed for new construction was stiff and did not amplify the response of the switches significantly. For the other three frames, amplification factors of greater than 2 were recorded. The tall braced frame proposed for new construction produced the largest amplification factors, and consideration should be given to stiffening this frame.
- Errata (posted June 7, 2019)
Full List of PEER Reports: click here.