Skip to main content

M2D2021 – effects of climate change on the vibratory behavior of an arch dam

By 1 juillet 2021octobre 18th, 2021No Comments

Effects of climate change on the vibratory behavior of an arch dam – continuous and cables-free monitoring project


Continuous vibration monitoring of civil engineering structures is a necessary and important step in understanding the impact of climate change on structures. The surveillance systems deployed today are essentially composed of several sensors. We can see this at the M2D2021 SHOW, time synchronization between sensors is normally required by conventional modal identification techniques. To do this, sensors are linked to a GPS and an acquisition system.. These means are very expensive and quite complicated to install in some cases. The use of cables to ensure the synchronization and the continuous transmission of data represents the major limitation of these systems. This
work deals with a new continuous monitoring and cable-free system. This system was installed on the Ribou arch dam (France) and has been monitoring it since March 2018. In order to deal with the influence of external parameters on the measurements, the HST (hydrostatic-season-time) model has been developed. A slight deviation from the HST model
has been observed in the last 6 months. This deviation is probably due to the irreversible effect of climate change.


Climate change is a phenomenon that is growing very rapidly. It corresponds to a lasting of the statistical parameters of the global climate of the earth. These changes may be due to processes intrinsic to the earth or to human activities. This phenomenon particularly affects civil engineering structures involving the safety of goods and people [Singhal, 2019 ; Nasr, 2019]. Consequently, it is necessary to face this predicted impact and anticipate risk by monitoring the existing structures such as bridges or dams and by analysing to what extent these structures resist attacks from the natural environment. A solution lies in continuous vibratory auscultation. This vibratory study consists of identifying “health indicators”, such as Eigen frequency or damping ratios, and monitoring them over time [Frigui, 2018]. A vibration Monitoring System (VMS) generally consists of sensors, GPS, data acquisition system, data transmission system and several meters of cable. “Health indicators” are identified, in a deported way, using Operational Modal Analysis algorithms (OMA) [Chen, 2017]. In this traditional architecture, two major issues are identified; the first one is the sensor wiring, meaning the use of long cables and data synchronization through GPS. The second issue deals with data transmission that requires the installation of expensive and cumbersome means to ensure continuous data flow. For some civil engineering structures, such as historical monuments or a multi-floors building, this conventional system can become very expensive.
In light of these difficulties, significant advantages, in terms of flexibility and economy, can be obtained if indicators would be performed using cable-free VMS.

In this article a new surveillance system is presented briefly. This system is based on the use of vibration sensors, embedded signal processing allowing data selection and compression and an IoT type transmission system. This new system allows natural frequencies to be monitored over time [Judenherc, 2018]. It has been monitoring, since March 2018 the Ribou arch dam in Cholet (France).


 fig 1


M2D2021 AD-SIGNUM fig 2

The Ribou dam was initially equipped with a conventional monitoring system since March 2018. This system has allowed the development and validation of the algorithms put in place for the new monitoring system. In September 2020, the new system was installed in 14 minutes


The natural frequencies identified and monitored have presented, since the first months, a rather particular evolution, meaning a sinusoidal tendency with annual period (fig3)


Without exceeding the alert threshold, this trend is mainly related to the internal temperature
of the concrete. Indeed, the implementation of a HST modeling, the aim of which is to link the observed evolutions to reversible phenomena and to irreversible phenomena and predict the frequency variation of the structure [Salazar, 2016], we observed a remarkable correlation between the model prediction and the real measurements. The daily monitoring makes it possible to highlight the effect of the change in the dam rigidity, driven by the change in temperature, on the variation in natural frequency (fig.3). This observation is validated by the HST prediction model (fig.4). The periodic variation has shown a very slight drift since July 2020 which may be linked to the irreversible contribution of climate change (fig.4).

M2D2021 AD-SIGNUM fig 3
M2D2021 fig 4