INDUCED SEISMICITY: Where are we now, and what challenges do we face?

This is a series of webinars organized by TCS AH Consortium.

If you would like to be added to the participant's list to get regular information about the webinars you can sign up

Reminder: To watch past webinars (for which presenters agreed to be recorded), please visit the TCS-AH YouTube Channel

Webinar schedule for 2024 (4:00 to 5:00 pm CET)

• 19 December 2024

  speaker: Ade Nascimento (Universidade Federal do Rio Grande do Norte, Brazil)

  topic: Integration of reflection seismics and earthquake monitoring in a salt solution mining seismicity case in Brazil

• 31 October 2024

  speaker: Adam Lurka (Central Mining Institute, Poland)

  topic: How to find a group of clusters of induced seismicity in space and time? - Hierarchical cluster analysis example

  Abstract

  Seismicity induced by mining exploitation is one of the most dangerous physical phenomena in underground mines and often leads to the destruction of mine workings. Local seismic systems installed in Polish underground coal mines are used routinely for monitoring seismic events with seismic energy range between 102J and 109J. These systems are able to locate accurately both low and high energy seismic tremors providing detailed information of seismicity distribution in space and time. The SOS - Seismic Observation System developed by Laboratory of Mining Geophysics at Central Mining Institute has also been installed in Marcel coal mine in the area of safety pillar.

  Seismic events that occurred in the area of this safety pillar between 2009-2021 were located and analyzed within the EPOS PL+ project to prepare seismic data catalogue as part of its research infrastructure. The recorded mining induced seismicity can be grouped to form spatial clusters of seismic events related to the geological and mining structures in mines, such as faults and pillars. However, typical seismicity in coal mines is mainly related to the current mining process and there is no distinct spatial or temporal group of seismic events, but rather they form a continuous cloud of events related to mining process.

  Therefore, for the clustering of such event clouds, a new time-space metric was introduced along with the Ward's minimal variance method, and a hierarchical induced seismicity cluster analysis was performed. The introduced time-space metric allowed for the simultaneous assessment of the seismic hazard in time and space, linking individual seismic clusters with a group of cracks dynamically formed during mining operations. The applied method of hierarchical clustering was used to study the seismicity induced in the Marcel Coal Mine with high seismic hazard in the area of the safety pillar. As a result, 20 separate clusters of seismic events correlated with the mining process were obtained.

  In each cluster, the energy distribution of seismic activity was calculated, showing, in some cases, a significant deviation from the Gutenberg-Richter distribution, which may be important in the seismic hazard analysis. As a measure of the accuracy of the location of each seismic cluster specially designed bootstrap algorithm was introduced. In this way, it was shown that errors in the location of seismic phenomena affect the location of the center of each cluster. The presented analysis of induced seismicity as groups of clusters relates them with groups of seismic sources and areas of high stress in the rock mass, which may potentially allow for determination of seismic and rock burst hazards.

  Watch the Webinar on YouTube

• 30 September 2024

  speaker: Ian Stimpson & Glenda Jones (Keele University, UK)

  topic: EPISODES Platform use cases for teaching and outreach

  Abstract

  One of the key challenges in induced seismicity is the escalating shortage of geophysicists needed to undertake research in this area. To help address this problem teaching of the subject needs to be embedded across multiple levels of education alongside outreach to build a greater societal awareness (and interest!) in earthquakes triggered by anthropogenic activities. In this webinar we provide examples of how induced seismicity using the Episodes research platform has been embedded into higher education teaching and share details on how outreach incorporating the platform has been developed and delivered. Free open access online resources that can be used in educational outreach in a range of settings from primary school to postgraduate and even public events will be shared and encouraged!

  Watch the Webinar on YouTube

• 24 June 2024

  speaker: Francesca De Santis (Ineris, France) & Julie Maury (BRGM, France)

  topic: Good practices guide for managing seismicity induced by deep geothermal operations

  Abstract

  Ineris and BRGM published a good practice guide for managing induced seismicity in the context of deep geothermal energy exploitation. The guide, commissioned by the French Ministries of the ecological transition and the energy transition, is addressed primarily to operators and French administration involved in deep geothermal energy.

  In this webinar, we present the approach followed in this guide to develop a methodology for assessing the seismic hazard. Firstly, a worldwide review of deep geothermal projects has been realized to identify key factors triggering induced seismicity. Projects selected in the review are characterized by different levels of induced seismicity and have been chosen to be representative of different types of geothermal systems and operating conditions. Based on this case study review, we propose a methodology to assess induced seismic hazard along the whole lifecycle of a geothermal project (from exploration to shutdown), for estimating to what extent induced seismicity is of concern. Hazard assessment is based on a decision tree approach, involving specific criteria for each project phase, which allows to obtain hazard levels ranging from 0 (low concern) to 3 (high concern). Per each of these hazard levels, recommendations are, then, given in terms of data to be acquired, seismic monitoring, as well as Traffic Light Protocols (TLS) to be applied.

  Watch the Webinar on YouTube

• 24 Apr 2024

  speaker: Francesco Grigoli (Università di Pisa, Italy)

  topic: Offshore microseismic monitoring (with application to the Castor case, Spain)

  Abstract

  Analyzing offshore seismic sequences is challenging for the lack of optimal seismic monitoring networks. These conditions limit the application of standard data analysis procedures, leading to low accuracy results that may affect monitoring and early identification of the source of seismicity, particularly when potentially related to offshore industrial operations. Here we apply a recently developed location procedure for offshore seismic sequences capable of achieving high location accuracy with suboptimal monitoring conditions. We focus on the analysis of the Magnitude 4.2 earthquake sequence in the offshore of Valencia Gulf (Spain) occurred in September 2013. The earthquake sequence lasted for about one month and consisted of about 1000 seismic events. Due to the sparsity of the local seismic network, composed of only two stations within 15 km from the epicentral area of the seismicity, the use of standard location methods did not allow to identify the structure associated with the seismic sequence.

  In this study, we locate the seismic sequence with more advanced location techniques allowing us to achieve a higher location accuracy, even when dealing with suboptimal monitoring conditions. We analyze this dataset using waveform-based location techniques and a recently developed location technique based on Distance Geometry Solvers (DGS). This last approach uses inter-event distances between earthquake pairs estimated at one or two seismic stations to get high-resolution locations of seismicity clusters. The application of such techniques led to different improvements in locating the seismic sequence, which is more clustered and clearly shows trends compatible with the geological setting of the area. Although, this approach has been already successfully applied the 2019 Mw 6.9 Ridgecrest (California, US) Earthquake sequence this work clearly shows how the application of advanced methos for seismicity characterization lead to enhanced results even when dealing with extremely sparse seismic networks.

  Watch the Webinar on YouTube

• 22 Mar 2024

  speaker: Grzegorz Kwiatek (GFZ, Germany)

  topic: Tracking deviation from stable seismic energy release during hydraulic stimulations using physics-informed features

  authors: Grzegorz Kwiatek, Sadegh Karimpouli, Patricia Martinez-Garzon and Georg Dresen

  Abstract

  Near-realtime high-frequency seismic monitoring of fluid injection allowed mitigating induced seismicity during two hydraulic stimulations performed in 2018 and 2020 in a deep geothermal well near Helsinki, Finland. Using near-realtime information on the evolution of seismic and hydraulic energy, pumping was either stopped or varied, following the theoretical predictions from a physics-based model of maximum magnitude, avoiding occurrence of project stopping large earthquakes. In this talk we present factors contributing to the project success:

  1. adaptive stimulation strategy,

  2. structural inventory of the reservoir,

  3. limited stress transfer

  and discuss the pool of physics-informed seismo-mechanical and statistical parameters that can be used to characterize the propensity of the geothermal reservoir to display an unstable (runaway) behavior.

  We then discuss the development of unsupervised machine learning framework to characterize the stability of the complex fault system under natural or anthropogenic loading and test it on data from a series of laboratory triaxial stick-slip experiments. Physics informed features reveal a transition from stable deformation in the fault zone to an intermittent criticality state allowing the occurrence of runaway events. In this stage, numerous confined slips superimpose and interact, collectively preparing the fault zone for a runaway event by progressive smoothing of the longer length scales of the stress field. We find that runaway earthquake occurring at this stage is a statistical event that cannot be predicted deterministically. However, the combination of physics-informed parameters empowered with unsupervised ML techniques allows identification of onset time when a fault system enters a critical stage, potentially allowing to trigger relevant mitigation actions.

  Watch the Webinar on YouTube