What is it about?

Our study addressed the challenge of detecting small earthquakes caused by injecting fluids into the ground. We identified 196 low-magnitude earthquakes belonging to the fluid-injection induced seismicity cluster located in the High Agri Valley, Southern Italy, using a template-matching algorithm on the continuous seismic data streams acquired by the INSIEME network. These findings provided insights into the damage zone of a SW dipping fault, characterized by a variety of fractures critically stressed in the dip range between ~45° and ~75°. Notably, the earthquakes occurred in clusters and were found to be closely linked to the volume and pressure of the injected fluids.

Featured Image

Why is it important?

The main findings of this study are: 1. The single-station template matching algorithm allowed us to detect weak events with local magnitudes in the range − 1.2 ≤ Ml ≤ 1.2 and with a completeness magnitude Mc = − 0.5, much lower than the completeness magnitude obtained for this seismicity cluster in previous studies. 2. The seismicity is characterized by a swarm time-dependent earthquake occurrence with no identifiable mainshock and a b-value of 1.37 ± 0.05, inferred from the frequency-magnitude distribution of seismicity (Fig. 3). Furthermore, the seismicity well correlates with injection operational parameters (Fig. 5 and Table 1) with only 8 events occurred with injection rates equal or less than 100 m3/day. These observations suggest the reduction of the frictional fault strength due to pore fluids pressure. 3. The distribution of hypocenters obtained from relative locations depicts the damage zone of a SW dipping fault (Fig. 4). It may constitute a hydraulic connection path between the bottom-hole of the CM2 well and the NE dipping back-thrust which was originally reactivated by injection operations in the period 2006–2012 (Fig. 6). 4. The maximum injection pressure at the bottom-hole of 6 MPa operated by the oil company may critically stress fractures dipping in the range from ~ 45° to ~ 75° (Fig. 7), thus justifying the observed variety in the earthquake waveforms (Supplementary Figs. S3 and S4). 5. A strong increase in the number of seismic events is observed after acidification operations and for injection rates greater or equal than 1900 m3/day, but never exceeding the magnitude threshold (M = 1.5) that activates the attention level introduced in the Italian guidelines. When acid treatment is not executed, injection rates of about 2000 m3/day do not significantly affect the earthquake production.


A step forward is the implementation of the proposed analysis into fully automatized procedures, where the selection of event templates, the discrimination between true and false detections, and the automatic picking of P- and S-waves could be based on deep neural network algorithms.

Dr Tony Alfredo Stabile
Consiglio Nazionale delle Ricerche

Read the Original

This page is a summary of: Analysis of the 2016–2018 fluid-injection induced seismicity in the High Agri Valley (Southern Italy) from improved detections using template matching, Scientific Reports, October 2021, Springer Science + Business Media,
DOI: 10.1038/s41598-021-00047-6.
You can read the full text:




The following have contributed to this page