Vibration Threshold Exceedances in the Groningen Building Vibration Monitoring Network

Abstract

Vibration assessment criteria for cosmetic building damage are specified in terms of peak velocity thresholds. For masonry buildings, these thresholds are relatively low compared to vibration levels generated by various sources in the built environment. Therefore they will be exceeded numerous times over the lifetime of most buildings. These exceedances are important when one tries to establish or disprove a causal relation between building damage and a vibration source. Despite the relevance of this information, little is known about the number of times velocity thresholds for cosmetic building damage are exceeded in the built environment. The vibration data collected over a period of more than 5 years in 326 buildings, mostly constructed in masonry and randomly distributed over the Province of Groningen, allowed for a unique opportunity to gain insight. First of all, to determine the number of times these thresholds were exceeded, and secondly to make a comparison between shallow earthquakes and other sources. The number of exceedances and the return period were determined for a set of thresholds (i.e. 0.3, 1, 3, 5, 10, and 20 mm/s), which align with peak velocity thresholds typically encountered in vibration assessment guidelines and standards. The results obtained over a 1 year period with very low seismic activity (only earthquakes with M < 2.0) show that the velocity thresholds for cosmetic damage to masonry buildings (i.e. 1, 3, and 5 mm/s) were exceeded on average less than once a month for 80% of the building sensors. For buildings with a large number of exceedances of these thresholds, the most important sources appear to be traffic, construction work, and hits near the sensor. A subset of 59 building sensors was selected, for which the full operational period of about 5 years was analysed to determine the contribution of earthquakes to the threshold exceedances. Less than 1% of the 1 mm/s exceedances measured by the 59 sensors over their full operational period were caused by earthquakes. Only for building with few exceedances (less than 8), the relative influence of induced earthquakes on the number of exceedances is significant (between 30 and 50%).