Since 2012 many wind turbines have been installed on the Frankonian Jura and a number of them also in the vicinity of stations of the Gräfenberg array (GRF), consisting of 13 broadband stations within an area of about 50x100 km. It has been shown that these turbines take a significant effect on the noise level at many of the GRF station sites (Stammler & Ceranna, 2016, reference below). The array as a whole suffers from a deterioration of its sensitivity to teleseismic events of more than 0.1 magnitude units at wind speeds above 3.5 m/s (in 10m height). At individual station sites the noise signatures at frequencies above 2 Hz can be attributed to close-by wind turbines observing an approximate power decay law with increasing distance to the recording site. At a frequency of about 1.1 Hz, however, at most stations the strongest influence is visible, but the relation between measured PSD amplitudes and turbine distances does not support a simple decay law when taking into account only the closest wind turbine locations. This suggests that for this frequency turbines at larger distances play a role. This investigation tries to model the propagation of the turbine induced noise and to explain the observed PSD values at the GRF stations. As a result the contributing turbines can be identified as well as average propagation properties for the noise waves determined.

Influence of Wind Turbines on Seismic Records of the Gräfenberg Array, Klaus Stammler and Lars Ceranna, Seism.Res.Lett. (2016) 87(5): 1075-1081, https://doi.org/10.1785/0220160049

Seismologist noticed are worsening of station quality after the installation of wind turbines (WTs) close to seismological stations. Since WTs and seismological stations are installed mostly in areas with low population density, both are looking for solutions to diminish this conflict.

For this, we tested different denoising techniques at affected seismological stations to reduce or to eliminate the disturbing WT signal from the seismological data. Usually, spectral filtering is used to suppress noise in seismic data processing. However, this approach is not effective when noise and signal have overlapping frequency bands which is the case for WT noise. First, we applied a nonlinear thresholding function on our data. This method leads to good results when the event can be already seen in the raw data but it fails when the event is fully covered by noise. As a second method, we used a denoising autoencoder (DAE), which learns a sparse representation from time-frequency coefficients, and maps from there to output masking functions for signal and noise. The DAE is more time consuming in comparison to the nonlinear thresholding function but when the convolutional neural network is trained, using an adequate training dataset, it can also be applied instantly on the raw data. The DAE distinguishes between signal and noise and we are able to correct the seismograms from most of the disturbing noise signals.

Aerodynamic infrasonic signals generated by wind turbines can be detected by highly sensitive micro-barometers showing spectral peaks at the blade passing harmonics, which are above the background noise level. As infrasound is one of the four verification technologies for the compliances with the Comprehensive Nuclear-Test-Ban Treaty (CTBT), decreases in detection capability for dedicated infrasound arrays have to be avoided. Therefore, preventing such decrease is particularly important for the two German infrasound stations IS26 in the Bavarian Forest and IS27 in Antarctica, which are both part of CTBT’s International Monitoring System and have to meet stringent specifications with respect to their infrasonic ambient (natural and anthropogenic) noise levels.

In 2004, micro-pressure variations along a profile starting at a single horizontal-axis wind turbine were measured during a field experiment with mobile micro-barometer stations. As one of the results, a minimum distance to wind turbines for undisturbed recording conditions at infrasound array IS26 was estimated based on numerical modelling, validated with this dataset. Both observations and modelling were in agreement with the literature, where infrasonic signatures of wind turbines are reported at distance ranges up of tens of kilometres. Nevertheless, for broadening the dataset further infrasound measurements at two wind parks with modern large wind turbines have recently been carried out in Lower Saxony and Saxony-Anhalt, respectively. Here various instruments (micro-barometers, microphones, pressure sensors) have been deployed in a comparative manner. We will give an overview of these campaigns, followed by first results of our analysis and interpretation.

Repeated seismic activity can cause progressive failure of rock masses due to material fatigue [1]. To simulate induced seismic loading, an iron ore with alternating quartz- and magnetite-rich layers from the Sydvaranger mine (Finnmark/Norway) was subjected to laboratory uniaxial compressional cyclic loading at low stresses in the range of elastic deformation (about 6 MPa static ±3 MPa dynamic pressure) and frequencies related to induced seismicity (10 – 100 Hz) [2]. Some of the experiments were performed until material failure occurred at up to 2 million cycles. Changes in magnetic behaviour were identified by measurements of magnetic susceptibility from Verwey transition in magnetite (about 120 K) to room temperature after intervals of 150 000 loading cycles. Magnetic domains were analysed with magnetic force microscopy. Deformation-related surface structures on magnetite and quartz were identified by reflected light microscopy and high-resolution scanning electron microscopy and compared to results from Raman spectroscopy.

Dynamic mechanical analysis (DMA) was used for cyclic loading, which is a common method for determination of viscoelastic behaviour, especially of soft materials with pronounced viscoelasticity like polymers. Suitability of DMA for determination of relative changes of viscoelastic parameters during cyclic loading will be discussed under consideration of the relatively high Young’s modulus and almost ideal elastic behaviour of the iron ore under the chosen experimental conditions.

[1] Gischig et al. (2016) Rock Mech Rock Eng, 49, 2457-2478.

[2] https://www.geosig.com (2009), Seismic Signals and Sensors.

Traffic light systems (TLSs) are used in different energy technologies for limiting the strength of induced seismicity. The TLS concept is based on real-time monitoring of induced seismicity in combination with operational mitigation measures that are triggered if the intensity of the seismicity exceeds certain threshold values. We use observations, conceptual and numerical models for investigating efficiency and limitations of TLSs. Most notably, TLS efficiency can be limited by trailing effects caused by post-injection pressure diffusion and stress concentrations at the periphery of previous seismic activity. The latter ‘stress memory’ can be the cause for seismicity occurring a long time after reservoir activities stopped. We speculate that a similar type of stress concentration could have been the nucleus of the M_w=5.5 earthquake at the Pohang geothermal site.

Wind turbines are massive tall buildings which swing considerably above the ground, especially when they are in operation. These oscillations are composed of the eigenmodes of the whole building and the interaction of the passing blades with the tower. The oscillations are transferred into the ground by interaction of the moving foundation with the ground (soil / rock). The radiated wavefield is composed of elastic waves, conformable to seismic waves. Mainly surface waves are excited whose amplitudes decay with distance due to the geometric amplitude decay, the anelastic damping and the wave scattering, depending on the elastic properties of the rock along the propagation path. The related ground shaking is hardly felt by humans, however, sensitive high-tech instruments (electron microprobes, seismometers etc.) can be disturbed even at long distances. Thus, the knowledge about the source and propagation properties is vital to predict ground motion emissions and plan counter measurements. To enhance to development of wind energy as contribution to the exit from nuclear and fossil-fuel energy, expertise is needed to cope with the ground motion emissions of wind turbines.