Comparative Study of Earthquake Clustering in Relation to Hydraulic Activities at Geothermal Fields in California

Abstract

Earthquake clustering properties are investigated in relation to fluid balance H(t) (the difference of fluid injection and production rates) using about nine years of data from The Geysers (both the entire field and a local subset), Coso, and Salton Sea geothermal fields in California. Individual earthquake clusters are identified and classified using the nearest-neighbor approach of Zaliapin and Ben-Zion (2013a, https://doi. org/10.1785/0120150211, 2013b, https://doi.org/10.1093/gji/ggw300). These are used to calculate nine complementary cluster statistics as time series with a step of about one month. Three alternative techniques (moving window correlation, analysis of variance, and regression) are employed to assess the relations between (possibly nonstationary) time series of cluster statistics and H(t). A total of 108 pairwise relations between cluster statistics and H(t) are analyzed to clarify effects of fluid activities on seismicity in different places. The seismic clustering response to the fluid balance differs among the examined fields. The Geysers and Salton Sea areas display the highest and lowest clustering responses, respectively. The proportion of clusters consisting of a single event with no offspring (singles) is correlated significantly with H(t) at all examined data sets, with a lower proportion of singles during periods of high fluid balance. This may reflect increased susceptibility to earthquake triggering in time intervals with high injection rates. The background seismicity rates significantly increase with H(t) at the Geysers and Coso, while an opposite relation holds at the Salton Sea. This could be related to the high structural and tectonic complexity at the Salton Sea compared to the other two geothermal fields.Motivated by these challenges, Zaliapin and Ben-Zion (2013a, 2013b) systematically examined properties of earthquake clusters in Southern California with a novel nearest-neighbor approach (see section 3.1). This method allows to robustly identify and analyze seismicity clusters covering a broad range of sizes, energies, and spatio-temporal extents in the presence of varying levels of completeness and uncertainties. Their results highlight the existence of three different basic types of clusters: burst-like clusters (reminiscent of classical

Clustering characteristics of gas-extraction induced seismicity in the Groningen gas field

“…Several studies have derived a possible correlation between injection/extraction rates and background seismicity rates of humaninduced seismicity (Llenos & Michael 2013;Johnson et al 2016;Trugman et al 2016;Martínez-Garzón et al 2018). A similar correlation between the rates of induced seismicity and the rates of gas extraction have been observed for the Groningen gas field (Muntendam-Bos & De Waal 2013;De Waal et al 2015;Nepveu et al 2016;Muntendam-Bos et al 2017;Sijacic et al 2017).…”

Earthquake Declustering Using the Nearest‐Neighbor Approach in Space‐Time‐Magnitude Domain

“…Swarms occur generally in areas with elevated fluid‐temperature conditions and low effective viscosity, whereas classical mainshock‐aftershock sequences tend to occur in relatively cold high viscosity regions (Ben‐Zion & Lyakhovsky, ; Zaliapin & Ben‐Zion, ). Oil and gas production, exploration of geothermal energy, and other anthropogenic activities can also lead to and modify observed forms of earthquake clustering (e.g., Ellsworth, ; Goebel & Brodsky, ; Martínez‐Garzón et al, ; Schoenball et al, ).…”

“…The problem of detecting earthquake clusters (which is different from although closely related to declustering) was shown recently to be effectively addressed by a nearest‐neighbor analysis in space‐time‐magnitude domain (Zaliapin et al, ; Zaliapin and Ben‐Zion, , , , , ). This cluster identification methodology is effective in diverse settings, including tectonic seismicity (Gentili et al, , ; Gu et al, ; Kossobokov & Nekrasova, ; Moradpour et al, ; Peresan & Gentili, ; Reverso et al, ; Ruhl et al, ; Trugman et al, ; Zaliapin et al, ; Zaliapin & Ben‐Zion, , , ), induced earthquakes (Goebel et al, ; Martínez‐Garzón et al, ; Schoenball et al, ; Schoenball & Ellsworth, ; Teng & Baker, ; Vasylkivska & Huerta, ; Zaliapin & Ben‐Zion, ), synthetic seismicity in Epidemic Type Aftershock Sequence (ETAS) models (Gu et al, ; Zaliapin et al, ; Zaliapin & Ben‐Zion, ), and laboratory rock fracture experiments (Davidsen et al, ).…”

Analysis of Microseismicity Framing ML > 2.5 Earthquakes at The Geysers Geothermal Field, California

“…Although seismicity at The Geysers strongly correlate with injection rate and volume, we did not see a strong dependence between the seismic productivity of a main shock and the injected volume around its time of occurrence here (Figure ). Similarly, Martínez‐Garzón et al () using cluster analysis found that at The Geysers the number of aftershocks only partly correlates with the injection rates of the field, with the relation varying significantly with time and no continuous trend found.…”