Structural Health Monitoring Using Multi-Parameter Information: Case of the Kurpsai Dam in the Kyrgyz Republic

Session: Innovative Seismo-Acoustic Applications to Forensics and Novel Monitoring Problems [Poster]
Type: Poster
Date: 4/29/2020
Time: 08:00 AM
Room: Ballroom
Description: 

Given the plans to construct hydroelectric dams in the Kyrgyz Republic, and the need to ensure the integrity of existing structures, techniques for gaining rapid and robust information about the dam structures and the surrounding slopes are required. Such techniques would be used to establish monitoring systems that will support the decision-making process in the event of an emergency. The MI-DAM project (MultI-parameter monitoring and risk assessment of hydro-electric DAMs in the Kyrgyz Republic) has installed and is testing a robust, cost-effective and flexible monitoring system at the Kurpsai Hydropower Station in Western Kyrgyzstan. The monitoring concept distinguishes between two time scales: the long-term monitoring of static deformations over days, months and years and the short-term monitoring of structural response to earthquake shocks and extreme operational regimes. While the long-term monitoring includes a combination of measurements of absolute static displacements by GPS sensors and fibre optical strain sensors as well as long-term deformations measured by Interferometric Synthetic Aperture Radar (InSAR), the short-term changes in the dam and the surrounding hillsides are monitored by means of multi-parameter sensors placed at selected sites on the structure and surroundings based on a fully decentralised approach. Such an approach allows the critical parameters for monitoring (and, correspondingly, the fragility curves) to be directly integrated into the on-site calculations, allowing some degree of decision-making without the necessity of a remote centre. Strong motion recordings show that the dam exhibits a transient non-stationary behaviour as its fundamental frequency changes during each strong motion, then returning to the starting value after each event. While a solid structural model can serve as a prerequisite for reliable monitoring of engineering structures, results further illustrate severe challenges in the creation of a digital twin. This is essential for the successful physical interpretation of the structural behaviour.

Presenting Author: Marco Pilz

Authors