Magma Fracking Beneath Active Volcanoes Based on Seismic Data and Hydrothermal Activity Observations

Home , Avachinsky, Kambalny, Koryaksky, Shiveluch, Udina, Ushkovsky

geosciences

Article Magma Fracking Beneath Active Volcanoes Based on Seismic Data and Hydrothermal Activity Observations

Alexey Kiryukhin *, Evgenia Chernykh, Andrey Polyakov and Alexey Solomatin Institute of Volcanology & Seismology FEB RAS, Petropavlovsk-Kamchatsky 683006, Russia; [email protected] (E.C.); [email protected] (A.P.); [email protected] (A.S.) * Correspondence: [email protected]

Received: 28 November 2019; Accepted: 26 January 2020; Published: 29 January 2020

Abstract: Active volcanoes are associated with microearthquake (MEQ) hypocenters that form plane-oriented cluster distributions. These are faults delineating a magma injection system of dykes and sills. The Frac-Digger program was used to track fracking faults in the Kamchatka active volcanic belt and fore-arc region of Russia. In the case of magma laterally injected from volcanoes into adjacent structures, high-temperature hydrothermal systems arise, for example at Mutnovsky and Koryaksky volcanoes. Thermal features adjacent to these active volcanoes respond to magma injection events by degassing CO2 and by transient temperature changes. Geysers created by CO2-gaslift activity in silicic volcanism areas also ﬂag magma and CO2 recharge and redistributions, for example at the Uzon-Geyserny, Kamchatka, Russia and Yellowstone, USA magma hydrothermal systems. Seismogenic faults in the Kamchatka fore-arc region are indicators of geoﬂuid fracking; those faults can be traced down to 250 km depth, which is within the subduction slab below primary magma sources.

Keywords: magma; hydrothermal; fracking; volcanoes; Kamchatka

1. Introduction Volcanoes and crustal magma chambers are the products of magma injection from primary magma chambers at depths of 150–200 km below the Kuril-Kamchatka volcanic zone (Figure1). These magma injections create new hydraulic fractures and reactivate existing faults. The formation of shear cracks in adjacent to the main aseismic opening triggers microseismicity [1]. Earthquake magnitudes (M) for shear deformations with amplitudes of 0.1 mm–1 cm and fracture lengths of a few hundred meters are estimated to range from 1 to 2 (Ks from 3.5 to 5.5), corresponding to the sensitivity of local seismic networks operating in the areas of the Klyuchevskoy, Koryaksky-Avachinsky, and Mutnovsky-Gorely volcano groups. We propose that the planes deﬁned by clusters of microearthquakes correspond to magmatic intrusions. For a search of plane-oriented clusters of earthquakes, we use our original programs Frac-Digger (RU reg. ## 2016616880), and for production feed-zones, we use Frac-Digger2 (RU reg. ## 2017618050) (these applications discrete fracture systems using seismic and geological data). In these programs, the following criteria are used to select clusters of earthquake hypocenters or production feed zones: (1) proximity in time δt (Frac-Digger only); (2) proximity within the horizontal plane δR; (3) proximity to plane orientation δZ (distance between the object and plane); and (4) minimum number of elements in cluster N. This method has already been tested on the example of the 2012 eruption of Tolbachik volcano [2], the 2000–2016 MEQ data of Koryaksky-Avachinsky volcanic cluster [3–5], Paratunsky graben (adjacent to Vilyuchinsky volcano) [6], the Mutnovsky-Gorely volcanic cluster [7], and the Klyuchevkoy group of volcanoes [8]. Note that the methodology of dyke tracking using the Frac-Digger program is presented in detail in Section 3.2.1, “Seismic Data and Method of Plane-Oriented Clusters Identiﬁcation” [5].

Geosciences 2020, 10, 52; doi:10.3390/geosciences10020052 www.mdpi.com/journal/geosciences Geosciences 2020, 10, 52 2 of 16 that the methodology of dyke tracking using the Frac-Digger program is presented in detail in Section 3.2.1, “Seismic Data and Method of Plane-Oriented Clusters Identification” [5]. By coupling magma fracking observations beneath volcanoes with the monitoring of thermal features in adjacent hydrothermal systems, we can develop appropriate geofiltrational and thermal– hydrodynamic TOUGH2-models that allow an assessment of recoverable geothermal resources and explain the mechanisms of the anomalous hydrothermal perturbations. A recently revealed mechanism [9] of geysers activity due to cyclic magmatic CO2-gaslift pointed to the possibility of understanding the mechanisms of relationship between anomalous hydrothermal perturbations, volcano activity, and strong earthquakes. Continuous measurements of CO2 partial pressures in the Koryaksky Narzan thermal mineral springs (northern sector of Koryaksky volcano, 2017–2019) and Geosciences 2020, 10, 52 2 of 16 in the Mutnovsky geothermal Power Plant condenser (2019) provide evidence of this process.

Figure 1. Location map of the volcanoes and hydrothermal systems of Kamchatka. 1—Active Figure 1. Location map of the volcanoes and hydrothermal systems of Kamchatka. 1—Active volcanoes (1, Kambalny; 2, Koshelevsky; 3, Diky Greben; 4, Ilynsky; 5, Zheltovsky; 6, Ksudach; 7, volcanoes (1, Kambalny; 2, Koshelevsky; 3, Diky Greben; 4, Ilynsky; 5, Zheltovsky; 6, Ksudach; 7, Asachinsky; 8, Mutnovsky; 9, Opala; 10, Gorely; 11, Avachinsky; 12, Koryaksky; 13, Zhupanovsky; Asachinsky; 8, Mutnovsky; 9, Opala; 10, Gorely; 11, Avachinsky; 12, Koryaksky; 13, Zhupanovsky; 14, Karymsky; 15, Maly Semyachik; 16, Bolshoy Semyachik; 17, Kihpinych; 18, Taunshits; 19, 14, Karymsky; 15, Maly Semyachik; 16, Bolshoy Semyachik; 17, Kihpinych; 18, Taunshits; 19, Krasheninnikova; 20, Kronotsky; 21, Komarova; 22, Gamchen; 23, Kizimen; 24, Pl. Tolbachik; Krasheninnikova; 20, Kronotsky; 21, Komarova; 22, Gamchen; 23, Kizimen; 24, Pl. Tolbachik; 25, 25, Bezymyanny; 26, Kluchevskoy; 27, Ushkovsky; 28, Shiveluch; 29, Khangar; 30, Ichinsky); Bezymyanny; 26, Kluchevskoy; 27, Ushkovsky; 28, Shiveluch; 29, Khangar; 30, Ichinsky); 2—High- 2—High-temperature hydrothermal systems (1, Koshelevsky; 2, Pauzhetsky; 3, Hodutkinky; 4, temperature hydrothermal systems (1, Koshelevsky; 2, Pauzhetsky; 3, Hodutkinky; 4, Mutnovsky; 5, Mutnovsky; 5, B-Banny; 6, Karymsky; 7, Semyachiksky; 8, Geyserny; 9, Uzonsky; 10, Apapelsky; B-Banny; 6, Karymsky; 7, Semyachiksky; 8, Geyserny; 9, Uzonsky; 10, Apapelsky; 11, Kireunsky; 12, 11, Kireunsky; 12, North-Koryaksky); 3—Hydrothermal systems with temperatures below 150 ◦C; North-Koryaksky); 3—Hydrothermal systems with temperatures below 150 °C; 4—Groups of thermal 4—Groups of thermal springs: (a) temperature from 50 to 100 ◦C, (b) temperature from 20 to 50 ◦C. springs: (a) temperature from 50 to 100 °C, (b) temperature from 20 to 50 °C. Notes: The AB and CD Notes: The AB and CD lines delineates the cross-sections shown below. lines delineates the cross-sections shown below. By coupling magma fracking observations beneath volcanoes with the monitoring of thermal features in adjacent hydrothermal systems, we can develop appropriate geoﬁltrational and thermal–hydrodynamic TOUGH2-models that allow an assessment of recoverable geothermal resources and explain the mechanisms of the anomalous hydrothermal perturbations. A recently revealed mechanism [9] of geysers activity due to cyclic magmatic CO2-gaslift pointed to the possibility of understanding the mechanisms of relationship between anomalous hydrothermal perturbations, volcano activity, and strong earthquakes. Continuous measurements of CO2 partial pressures in the Koryaksky Narzan thermal mineral springs (northern sector of Koryaksky volcano, 2017–2019) and in the Mutnovsky geothermal Power Plant condenser (2019) provide evidence of this process. Geosciences 2020, 10, 52 3 of 16

2. Active Faults and Magma Fracks Based on Seismic Data Analysis

2.1. Koryaksky-Avachinsky Volcanoes Plane-oriented clusters of 5160 seismic events from data of the Kamchatka Branch Federal Research Center United Geophysical Survey Russia Academy of Sciences (KB FRC UGS RAS) of 01.2000–07.2016) below Koryaksky and Avachinsky volcanoes were used to identify magma injection zones [4,5]. Cluster identification was carried out using our Frac-Digger program. The previously defined criteria were used to include a new event in the cluster: δt 1 day, δR 6 km, δZ 0.2 km, N 6. ≤ ≤ ≤ ≥ It was found that 30% of the total number of 5160 earthquakes formed 204 plane-oriented clusters. A careful analysis of the orientations of dykes and sills (using stereograms and histograms) showed the following: (1) the dykes beneath Koryaksky Volcano were mostly emplaced in the depth range from 5000 to 0 m masl; they mostly strike north–south (75% of dykes have strike azimuths between − 320◦ and 40◦) and have dip angles over 50◦ (70% of the dykes), and (2) the dykes beneath Avacha Volcano were mostly in the depth range 1000–2000 m masl. Their strikes are distributed over all possible directions, but there is a local maximum of dykes striking nearly north–south with 25% of dykes having azimuths between 350◦ and 10◦. Steeply dipping dykes are not obviously dominant (61% have dip angles over 45◦). Additionally: (1) A shallow crustal magma chamber, likely as a combination of sills and dykes, lies below the southwestern base of Koryaksky Volcano at depths ranging from 2 to 5 km masl and − − a width of 2.5 km; (2) dyke accumulation in a nearly north–south zone (7.5 by 2.5 km, with absolute depths ranging from 2 to 5 km occurs beneath the northern sector of Koryaksky Volcano; and (3) a − − shallow magma chamber in the cone of Avachinsky Volcano lies at absolute depths ranging from 1 to 2 km and has a width of 1 km. Thus, we propose the geomechanical state according to [10]: (1) extension is recorded beneath Koryaksky Volcano producing normal faulting (NF conditions), the vertical stress Sv is the maximum stress, the horizontal principal stresses are coincident with the north–south (SHmax) and the east–west (Shmin) directions, (2) the cone of Avacha Volcano is found to be under conditions of radial extension. The relationship M = 0.67 log (V) 0.82 between the upper bound of the induced seismic event M − and magmatic injection volume V, suggested by [11], may be used for recorded dykes volume estimates beneath Koryaksky in a range from 220 m3 to 1.988 106 m3, and the total recorded volume of the × injected magma is estimated at 5.862 106 m3 during 2008–2011. Avachinsky volcano characterized × by a range from 111 m3 to 0.064 106 m3, and the total recorded volume of the injected magma is × estimated at 0.169 106 m3 during 2008–2011. × It is also worth noting that seismogenic planes and microearthquake (MEQ) mechanisms planes (19 mechanisms estimated were used) are intersected with the angle of 51◦ (on average) [12]. This may have the following geomechanical interpretation: seismogenic planes mark opening mode fractures (hydrofracking-type fractures) formed orthogonal to the least effective stress during magma injection, whereas MEQ mechanisms planes are mark shear faults, which initiated from the main opening mode fault (dyke) using pre-existing fractures systems. There were six magma injections below the northern foothills of Koryaksky volcano detected in 2011–2019 based on MEQ analysis data using Frac-Digger criteria (δt = 1 day, δR = 6 km, δZ = 0.2 km, N = 6): one dyke on 02.08.2011, one dyke on 28.02.2016, one dyke on 22.08.2017, two dykes on 14.11.2017, and one dyke on 22.04.2019. All of these dykes were injected just below the Koryaksky Narzan and Isotovsky thermal mineral springs area. The depth of magma injections was estimated from seismic data from 7000 to +1000 masl, with M ranges from 0.85 to 2.95 (Figure2, Table1). − Geosciences 2020, 10, 52 4 of 16 Geosciences 2020, 10, 52 4 of 16

Figure 2. Figure 2. MagmaMagma injectionsinjections (dykes) (dykes) below below the the northern northern foothills foothills of Koryakskyof Koryaksky volcano volcano during during the time the period from 2011 to 2019. Koryaksky Narzans are marked as K1, K2, K3, K7, and K8, Isotovsky hot time period from 2011 to 2019. Koryaksky Narzans are marked as K1, K2, K3, K7, and K8, Isotovsky spring—IS, Vodopadny hot spring—VD. Frac-Digger parameters are: δt = 1 day, δR = 6 km, δZ = 0.2 hot spring—IS, Vodopadny hot spring—VD. Frac-Digger parameters are: δt = 1 day, δR = 6 km, δZ = km, and N = 6. Dyke geometries and magnitudes are given in the Table1 below. Note: Dykes patches 0.2 km, and N = 6. Dyke geometries and magnitudes are given in the Table 1 below. Note: Dykes in 3D view are convex polygons, with vertices in the points of projections of plane-oriented clusters of patches in 3D view are convex polygons, with vertices in the points of projections of plane-oriented earthquakes hypocenters on an approximation plane. clusters of earthquakes hypocenters on an approximation plane. Table 1. Dyke geometries and corresponding microearthquake (MEQ) magnitudes of the Koryaksky volcanoTable 1. forDyke the geometries period from and 02.08.2011 corresponding to 22.04.2019. microearthquake (MEQ) magnitudes of the Koryaksky volcano for the period from 02.08.2011 to 22.04.2019. 2 ## Data Dip (◦) Dip Direction (◦) Mmax N Area km ## Data Dip (°) Dip Direction (°) Mmax N Area km2 1 02.08.11 39.3 115.1 2.95 7 1.4 1 02.08.11 39.3 115.1 2.95 7 1.4 2 28.02.16 70.2 64 2.45 7 1.7 2 28.02.16 70.2 64 2.45 7 1.7 33 22.08.17 22.08.17 69 69246.3 246.3 1.30 1.30 6 61.0 1.0 44 14.11.17 14.11.17 78 78291.2 291.2 1.25 1.25 17 1711.5 11.5 55 14.11.17 14.11.17 63.7 63.7 283.2 283.2 0.85 0.85 8 810.0 10.0 66 22.04.19 22.04.19 73 73266.5 266.5 1.95 1.95 6 62.2 2.2 If a less severe Frac-Digger criterion in time, δt ≤ 30 day, is used, the number of detected dykes significantlyIf a less severeincreases. Frac-Digger By using criterionthis approach, in time, it wasδt found30 day, that is used, as a result, the number 91% of of the detected total number dykes ≤ signiﬁcantlyof 5160 earthquakes increases. 452 By plane-oriented using this approach, clusters it were was foundformed. that Figure as a result, 3 shows 91% that of 68 the of total these number features of 5160were earthquakesinterpreted as 452 dykes plane-oriented from 2011 to clusters 2019 beneat wereh formed. Koryaksky Figure volcano.3 shows Most that of 68 them of these were features injected werein a northeast interpreted sector as dykes of the from volcano 2011 struct to 2019ure, beneath where thermal Koryaksky springs volcano. activity Most occurs. of them were injected in a northeast sector of the volcano structure, where thermal springs activity occurs.

Geosciences 2020, 10, 52 5 of 16 Geosciences 2020, 10, 52 5 of 16

Figure 3. Magma injections (dykes) below northern foothills of Koryaksky volcano during the time period from 2011 to 2019. Koryaksky Narzans are are ma markedrked as K1, K2, K3, K7, and K8, Isotovsky hot spring—IS, Vodopadny hothot spring—VD.spring—VD. Frac-DiggerFrac-Digger criteriacriteria are: are: ((δδtt= = 3030 day, δδRR == 6 km, δZ = 0.20.2 km, km, N = 6).6).

2.2. Mutnovsky Mutnovsky and Gorely Volcanoes We already used used plane-oriented plane-oriented earthquake earthquake clusters clusters (KB (KB FRC FRC UGS UGS RAS, RAS, catalogs catalogs 2009–2017) 2009–2017) to trackto track the the dykes dykes injected injected beneath beneath and and around around the the Mutnovsky Mutnovsky volcano volcano by by using using the the Frac-Digger program [5[5].]. Thus,Thus, magmatic magmatic injection injection zones zones (shape-like (shape-like dykes dykes and sills,and [sills,13,14 ])[13,14]) defined defined in this mannerin this weremanner treated were as treated pathways as pathways and heat and sources heat forsources ascending for ascending high-temperature high-temperature upflows upflows that feed that adjacent feed adjacenthydrogeological hydrogeological reservoirs reservoirs and surface and feature surface (hot feature springs (hot and springs fumaroles) and fumaroles) discharge. discharge. Four seismic stations record seismicity in the Mu Mutnovsky–Gorelytnovsky–Gorely volcanic cluster. A total of 1336 earthquakes were recorded by the KB FRC UGS RAS in the edifices edifices and basement of the Mutnovsky and Gorely volcanoes between January 2009 and FebruaryFebruary 2017.2017. Cluster identificationidentification was carried out using our our Frac-Digger Frac-Digger program. program. The The same same criteria criteria as as above above were were used used to toinclude include a new a new event event in thein the cluster. cluster. Our Our treatment treatment of ofthese these data data used used th thee method method previously previously outlined, outlined, which which yielded 973 earthquakes (72.8%) that make up 74 plane-orientedplane-oriented clusters between November 2013 and October 2016 (most of them associated with the Mutnovsky volcano) (Figure4 4).). The analysis of seismic activity at the MutnovskyMutnovsky volcano revealed the following geomechanical features. (1) (1) Most Most of of the the dykes were injected belo beloww and in the northeastern sector of the Mutnovsky volcano in an area 2 km 10 km. (2) Most of the dykes beneath the Mutnovsky volcano have a dip volcano in an area 2 km × ×10 km. (2) Most of the dykes beneath the Mutnovsky volcano have a dip angle angle ranging from 20 to 40 . (3) Most of the dykes were injected at a depth ranging from 4.0 to 2.0 ranging from 20° to 40°.◦ (3) ◦Most of the dykes were injected at a depth ranging from −4.0− to −2.0− km maslkm masl and at and strikes at strikes in a NE–NNE in a NE–NNE (20–50°) (20–50 direction.◦) direction. (4) Six dykes (4) Six were dykes injected were close injected to the close SE boundary to the SE boundary of the Mutnovsky production geothermal reservoir at a depth ranging from 6.0 to 4.0 km of the Mutnovsky production geothermal reservoir at a depth ranging from −6.0 to −4.0− km masl.− (5) Seismicmasl. (5) events Seismic of magnitude events of magnitude M during dyke M during injection dyke ranged injection from ranged 1.0 to 2.8. from (6) 1.0There to 2.8.is a trend (6) There of dyke is a diptrend angle of dyke increase dip anglemoving increase from the moving Mutnovsky from the volcano Mutnovsky to the north volcano (1°/km). to the north (1◦/km).

Geosciences 2020, 10, 52 6 of 16 Geosciences 2020, 10, 52 6 of 16

FigureFigure 4.4. Mutnovsky–GorelyMutnovsky–Gorely area magma fracking geometry vs. known production zone geometry. DykesDykes 21.04.201921.04.2019 andand 21.08.201921.08.2019 injectedinjected closeclose toto wellwell 035035 areare shown,shown, too.too. This ﬁgurefigure also gives a new approachapproach forfor geothermalgeothermal productionproduction targetstargets inin MutnovskyMutnovsky area.area.

Thus,Thus, dykedyke geometrygeometry indicatedindicated reversereverse faultfault (RF)(RF) geomechanicgeomechanic conditionsconditions [[10]10] inin thethe vicinityvicinity ofof thethe MutnovskyMutnovsky volcanovolcano (SHmax(SHmax> > Shmin >> Sv, SHmax strikingstriking to NE), while therethere isis a trend to normal faultfault (NF)(NF) conditionsconditions 1010 kmkm awayaway inin thethe regionregion wherewhere thethe productionproduction geothermal reservoir formed. The relationship M = 0.67 log (V) 0.82 between the upper bound of the induced seismic event M and The relationship M = 0.67 log (V) − 0.82 between the upper bound of the induced seismic event M and magmaticmagmatic injectioninjection volume volume V, V, suggested suggested by by [11 [11],], may may be be used used for for recorded recorded dyke dyke volume volume estimates estimates in a 3 6 3 rangein a range from 520from m 520to 0.252m3 to 0.25210 m × ,10 and6 m the3, and total the recorded total recorded volume ofvolume the injected of the magma injected is magma estimated is 6 3 × at 0.67 10 m from 2009–2016.6 3 estimated× at 0.67 × 10 m from 2009–2016. ThereThere werewere additionaladditional 2222 magmamagma injectionsinjections identifiedidentified duringduring 02.201702.2017 toto 10.201910.2019 inin thethe Mutnovsky–GorelyMutnovsky–Gorely geothermal area. area. Most Most of of the the dykes dykes beneath beneath the the Mutnovsky Mutnovsky volcano volcano have have a dip a dip angle ranging from 20 to 40 and were injected at the depth ranging from 4.0 to 2.0 km masl angle ranging from 20° to ◦40° and◦ were injected at the depth ranging from −4.0− to −2.0− km masl and anddipping dipping in the in theSEE–SE SEE–SE (120°–130°) (120◦–130 direction.◦) direction. Most Most of ofthe the dykes dykes beneath beneath the the Gorely Gorely volcano volcano have aa dip angle ranging from 40 to 70 and were injected at the depth ranging from 2.0 to 1.5 km masl dip angle ranging from 40°◦ to 70°◦ and were injected at the depth ranging from− −2.0 to −−1.5 km masl andand dippingdipping inin thethe NWWNWW (280(280°–320°)◦–320◦) direction. It is worth noting two sub-parallelsub-parallel dykesdykes (dip(dip anglesangles from 39 to 43 , dip azimuth from 149 to 153 , depth range from 2.6 to 1.9 km masl, M from 1.25 to from 39°◦ to 43°,◦ dip azimuth from 149°◦ to 153°,◦ depth range from− −2.6 to− −1.9 km masl, M from 1.25 1.7)to 1.7) injected injected in thein the vicinity vicinity of wellof well 035 035 in 2019in 2019 (Figure (Figure4). 4). 2.3. Northern Group of Volcanoes 2.3. Northern Group of Volcanoes To recover the sequence of magma injections in the area of the Klyuchevskoy Volcanic Cluster, To recover the sequence of magma injections in the area of the Klyuchevskoy Volcanic Cluster, we used a catalog containing data of seismic monitoring at 19 seismic stations operated by the KB we used a catalog containing data of seismic monitoring at 19 seismic stations operated by the KB FRC UGS RAS for the 2000–2017 period (the total number of earthquakes that have been recorded FRC UGS RAS for the 2000–2017 period (the total number of earthquakes that have been recorded between January 1, 2000 and August 23, 2017 is 122,451). The method that we used to identify between January 1, 2000 and August 23, 2017 is 122,451). The method that we used to identify plane- plane-oriented earthquake clusters and fitting planes using the Frac-Digger software was described oriented earthquake clusters and fitting planes using the Frac-Digger software was described in [2– in [2–5]. The relevant calculated plane patches are interpreted as zones of magma emplacement in the 5]. The relevant calculated plane patches are interpreted as zones of magma emplacement in the form form of dykes and sills. The calculations assumed the following parameters for the identification of of dykes and sills. The calculations assumed the following parameters for the identification of plane- plane-oriented clusters: δt 1 month, δR 6 km, δZ 1 km, N 6. The calculations yielded 1788 oriented clusters: δt ≤ 1 month,≤ δR ≤ 6 km,≤ δZ ≤ 1 km,≤ N ≥ 6. The≥ calculations yielded 1788 clusters clustersthat contained that contained 117,412 117,412earthquakes earthquakes (96% of (96%the total of the number total numberof recorded of recorded events). events). AA verticalvertical profileprofile alongalong thethe axisaxis ofof thethe KlyuchevskoyKlyuchevskoy VolcanicVolcanic ClusterCluster thatthat extendsextends fromfrom thethe TolbachikTolbachik volcanoesvolcanoes toto ShiveluchShiveluch (Figure(Figure5 )5) shows shows that that when when extrapolated extrapolated downward, downward, the the dykes dykes beneath Tolbachik and Shiveluch intersect at a point that lies between absolute depths of 165 and beneath Tolbachik and Shiveluch intersect at a point that lies between absolute depths of −−165 and 205 km beneath Klyuchevskoy; it is at this location that the region of primary magma melting is −−205 km beneath Klyuchevskoy; it is at this location that the region of primary magma melting is thoughtthought toto reside,reside, oror there there is is a a primary primary magma magma chamber chamber to to provide provide magma magma for for all all active active volcanoes volcanoes in thein the Klyuchevskoy Klyuchevskoy Cluster Cluster (Klyuchevskoy, (Klyuchevskoy, Plosky Plosky Tolbachik, Tolbachik, Bezymyanny, Bezymyanny, and and Shiveluch). Shiveluch).

GeosciencesGeosciences 20202020, ,1010, ,52 52 77 of of 16 16

Figure 5. Magma fracking system beneath the Northern group of volcanoes. The AB cross-section line Figure 5. Magma fracking system beneath the Northern group of volcanoes. The AB cross-section line was shown in Figure1. was shown in Figure 1. The 2000–2017 magma injections directly beneath Klyuchevskoy were concentrated in the depth rangesThe between 2000–201731 magma and injections28 km masl directly (35%) beneath and between Klyuchevskoy1 and were+2 km concentrated masl (20%), in the where depth we − − − rangeshypothesize between the − existence31 and − of28 akm crustal masl chamber (35%) and (C2) between and a peripheral −1 and +2 chamber km masl (C1), (20%), respectively where we (we hypothesizemean chamber the asexistence a plexus of ofa crustal sills and chamber dykes here).(C2) and The a crustalperipheral magma chamber chamber (C1), (C2) respectively was found (we to mean chamber as a plexus of sills and dykes here). The crustal magma chamber (C2) was found to contain dykes that dip at angles of 70–80◦ (30%) and sills that dip at 15–20◦ (8%); one also finds an containincreasing dykes number that dip of nearly at angles east–west of 70–80° striking (30%) dykes and sills that that supply dip magma at 15–20° to the(8%); nearby one also Bezymyanny finds an increasingand Krestovsky number volcanoes. of nearly east–west Nine episodes striking of dykes intensive that dykesupply emplacement magma to the and nearby seven Bezymyanny episodes of andintensive Krestovsky sill emplacement volcanoes. wereNine recorded episodes in of the inte C2nsive chamber dyke during emplacement the period and of interestseven episodes (2000–2017); of intensivethe dyke episodessill emplacement preceded were the sill recorded episodes in by the an C2 advance chamber time during of one the year period (two cases),of interest the episodes (2000– 2017);occurred the simultaneouslydyke episodes preceded (five cases), the or sill they episodes did not by terminate an advance in sill time emplacement of one year (one(two case), cases), which the episodesprovides occurred evidence simultaneously of a change in the(five geomechanical cases), or they condition did not aroundterminate the in C2 sill chamber emplacement in the range(one case),from which horizontal provides extension evidence to radial of a compression.change in the geomechanical condition around the C2 chamber in theThe range dip from angles horizontal and azimuths extension of dykesto radial and compression. sills in the peripheral chamber (C1) are distributed The dip angles and azimuths of dykes and sills in the peripheral chamber (C1) are distributed rather uniformly; one notes some increase in the number of dykes that dip at angles of 35◦ to 50◦ rather uniformly; one notes some increase in the number of dykes that dip at angles of 35° to 50° and and 65◦ to 80◦. There is also a set of dykes that dip at northeast azimuths (i.e., striking along the 65°Klyuchevskoy-Krestovsky to 80°. There is also a line). set of dykes that dip at northeast azimuths (i.e., striking along the Klyuchevskoy-KrestovskyTolbachik Volcano is characterizedline). by injections of magma in the depth range between 8 and 1 Tolbachik Volcano is characterized by injections of magma in the depth range between− −8 and− km masl, with most dykes dipping more steeply than 60◦ (67%). In addition, we identified a set of −1 km masl, with most dykes dipping more steeply than 60° (67%). In addition, we identified a set of dykes that dip at azimuths of 180–220◦ (26%) and strike along the Ostry Tolbachik to Plosky Tolbachik dykesto the that Bol’shaya dip at azimuths Udina line. of 180–220° (26%) and strike along the Ostry Tolbachik to Plosky Tolbachik to theShiveluch Bol’shaya isUdina characterized line. by magma injections in the depth ranges between 4 and 2 km masl − − (37%)Shiveluch and between is characterized 0 and +2 km by (47%), magma where injections two peripheral in the depth magma ranges chambers between are−4 and hypothesized −2 km masl to (37%) and between 0 and +2 km (47%), where two peripheral magma chambers are hypothesized to reside. The injections mostly occurred in the form of sills that dip at angles below 5◦, with some trend reside.of prevailing The injections western mostly dip azimuths. occurred in the form of sills that dip at angles below 5°, with some trend of prevailingBezymyanny western Volcano dip azimuths. receives magma from the crustal magma chamber beneath Klyuchevskoy VolcanoBezymyanny (C2), which Volcano lies at areceives depth between magma from31 and the crustal28 km masl.magma The chamber injections beneath beneath Klyuchevskoy Bezymyanny − − Volcanooccurred (C2), in the which depth lies range at betweena depth between2 and +2 km−31 (93%),and − where28 km amasl. peripheral The injections magma chamber beneath is Bezymyanny occurred in the depth range− between −2 and +2 km (93%), where a peripheral magma hypothesized to exist. The injections mostly occurred in the form of sills that dip at angles below 15◦ chamber is hypothesized to exist. The injections mostly occurred in the form of sills that dip at angles (27%) and low-angle dykes that dip at 50◦ or less (58%); the dip azimuths are distributed uniformly belowover all 15° directions. (27%) and low-angle dykes that dip at 50° or less (58%); the dip azimuths are distributed uniformlyThe volumesover all directions. of magma injection from magma chambers can be estimated using empirical relationshipsThe volumes that relateof magma the maximum injection magnitude from magma of the chambers triggered can seismicity be estimated to the volume using empirical of injected relationshipsmagma [11]: that M = relate0.67 log(V) the maximum0.82, where magnitude M is the of th maximume triggered magnitude seismicity of to triggered the volume seismic of injected events magma [11]: M = 0.67 log(V) −− 0.82, where M is the maximum magnitude of triggered seismic events

Geosciences 2020, 10, 52 8 of 16 and V is the injection volume. The results from this assessment of dyke and sill injections volumes for the volcanoes in the Klyuchevskoy Cluster during 2000–2017 are Klyuchevskoy (peripheral chamber C1)—0.6 106 m3, Klyuchevskoy (crustal chamber C2) —2.4 106 m3, Bezymyanny—4.7 106 m3, × × × Tolbachik—13.2 106 m3, Shiveluch—22 106 m3. Assuming the enthalpy of magma at a temperature × × 3 of 1200 ◦C to be 1000 kJ/kg and the density of magma to be 2800 kg/m , one can estimate the mass discharge and the corresponding thermal power due to magma injections for the time elapsed from January 2000 to August 2017. The thermal power values for magma injections in the volcanic plumbing systems considered here and the output estimates for the respective volcanoes from [15] yields the following ratios between the discharge of magmas stored beneath volcanoes and that of magma ejected onto the ground surface (volume of intruded magma)/(volume of erupted magma): 0.8% for Klyuchevskoy, 14.9% for Bezymyannyy, 23.2% for Tolbachik, and 72.9% for Shiveluch. Thus, Kluchevskoy volcano magma injections (dykes) took place in “normal fault” (NF) conditions, forming a permeable reservoir down to 35 km msl and two magma chambers within it; Shiveluch − volcano magma injections (sills) occurs in “reverse fault” (RF) conditions at shallow levels from 4 km − masl to 2 km masl, forming sills in an area 15 km across. − 2.4. Kamchatka East Volcanic Belt and Adjacent Shelf Area Critically stressed faults are key players in creation of the productive reservoirs and generation of strong earthquakes [10]. We performed a Frac-Digger analysis of the Kamchatka’s regional seismicity to identify the network of such active faults, which are identiﬁed as plane-oriented clusters of hypocenters in Frac-Digger terms [16]. We used the regional catalog of earthquakes from KB FRC UGS RAS, which includes 5972 earthquakes with M values above 4.25 during the time period from 01.1980 to 02.2016. The following criteria for seismogenic planes selection were used: N 6, δt , δZ 10 km, δR 100 km. These ≥ ≤ ∞ ≤ ≤ criteria correspond to the assumption of the existing network of continuously active regional faults. Based on the above, we found 156 plane-oriented clusters of earthquakes, which are interpreted as active seismogenic faults. Most of them are located beneath Kamchatka’s eastern shelf, between the shore line and ocean trench (Figure6). Seventeen faults are found to be the most active (with more than 100 earthquakes each). Most of the active faults are characterized by dip angles from 50◦ to 70◦ and a dip azimuth NWW from 300◦ to 310◦, striking subparallel to the ocean trench. This points to the extension in the NWW direction and NF geomechanical conditions as a whole. Seismogenic fault #110 includes the hypocenter of earthquake M = 7.3 on 24 Nov 1971, which was the strongest felt in Petropavlovsk-Kamchatsky since 1959. Faults with other directions (especially in shallow conditions at elevations above 10 km masl) are found too, which reﬂects local geomechanical − features. One of such faults is #4 (175 earthquakes included, dip angle 53◦, dip azimuth 217◦) striking in the direction of Petropavlovsk-Kamchatsky. Seismogenic fault #35 includes the hypocenter of earthquake M = 6.9 on 17 Aug. 1983, which was accompanied with the ground deformations described by [17]. Seismogenic fault #35 has a dip angle of 60.3◦, dip azimuth of 281.5◦, and includes 77 hypocenters of earthquakes. Geosciences 2020, 10, 52 9 of 16 Geosciences 2020, 10, 52 9 of 16

FigureFigure 6. 6. KamchatkaKamchatka shelf shelf active active seismogeni seismogenicc faults. faults. Traces Traces of of seismicall seismicallyy active active faults faults (black (black lines lines with numbers at elevation of 5 km masl, red lines at elevations of 20 km masl), circles corresponds with numbers at elevation of −−5 km masl, red lines at elevations of −−20 km masl), circles corresponds to regional earthquakes epicenters in a range of depth above 30 km asl (size of circle proportional to to regional earthquakes epicenters in a range of depth above −−30 km asl (size of circle proportional to earthquakesearthquakes magnitudes). magnitudes). For For other other symb symbols,ols, see see the the legend legend in in Figure Figure 1.1.

SeismogenicSeismogenic faults faults distributions distributions point point to to a a lowe lowerr limit limit of of the the hydrofrack hydrofrack propagation propagation into into the the subduction plate (down to 250 km masl), where this ﬂuid can act as the key ingredient for magma subduction plate (down to −−250 km masl), where this fluid can act as the key ingredient for magma meltingmelting to to recharge recharge the the primary primary chambers chambers of of active active volcanoes volcanoes (Figure (Figure 77).). ItIt is is also worthworth notingnoting that that 98% 98% of of all all earthquakes earthquakes formed formed a plane-oriented a plane-oriented clusters clusters network, network, which whichcharacterizes characterizes the geomechanical the geomechanical conditions conditions of collided of platescollided and plates suggests and hydrofracking suggests hydrofracking mechanisms mechanismsdue to ﬂuids due (possible to fluids phases (possible are: water,phases oil, are: gas) water, generation oil, gas) there. generation there. InIn case case of of water, water, the the empirical empirical relationship relationship M M == 0.670.67 log log (V) (V) ++ 1.421.42 between between the the upper upper bound bound of of thethe induced induced seismic seismic event event M M and and water water injection injection volume volume V, V, suggested suggested by by [11], [11], may may be be used used for for water water volumesvolumes recharge recharge estimates estimates into into the the active active faults faults network network of of the the Kamchatka Kamchatka sh shelf.elf. This This yields yields in in a a range from 0.00007 to 0.3 km3, and the total volume of the recharged water is estimated at 1.87 km3 during the time interval from 1980 to 2016 (over approximately 750 km of arc length).

Geosciences 2020, 10, 52 10 of 16 range from 0.00007 to 0.3 km3, and the total volume of the recharged water is estimated at 1.87 km3 duringGeosciences the time 2020 interval, 10, 52 from 1980 to 2016 (over approximately 750 km of arc length). 10 of 16 In case of gas, the empirical relationship is slightly modiﬁed (supercritical CO2 injection) to M = 0.67 log (V)In case0.30 of [gas,11]; the then, empirical this may relationship be used foris slightly rough estimatesmodified (supercritical of gas volumes CO2 generatedinjection) to from M = deep − hydrocarbon0.67 log (V) sources − 0.30 to [11]; be injectedthen, this into may the be activeused for faults rough network estimates of theof gas Kamchatka volumes generated shelf. This from yields in a rangedeep from hydrocarbon 0.024 to 111sources km3 ,to and be injected the total into volume the active of the faults injected network gas of is the estimated Kamchatka at 691 shelf. km This3 during yields in a range from 0.024 to 111 km3, and the total volume of the injected gas is estimated at 691 the time interval from 1980 to 2016. In relation to this, it is worth noting that methane hydrates and km3 during the time interval from 1980 to 2016. In relation to this, it is worth noting that methane methanehydrates submarine and methane discharges submarine are widely discharges distributed are widely along thedistributed east coast along of Kamchatka. the east coast of Then,Kamchatka. we perform an additional analysis of the 3D distribution of the 200 strongest earthquakes (М> 5.65) fromThen, the above-mentioned we perform an additional catalog analysis of seismic of the events 3D distribution of KB FRC of UGS the 200 RAS. strongest The following earthquakes selection criteria(М were> 5.65) used from in the Frac-Digger2: above-mentionedδZ = catalog4 km, δ ofR se=ismic100 km, events N = of5. KB The FRC output UGS resultsRAS. The show following that 102 of the 200selection strongest criteria earthquakes were used formin Frac-Digger2: 11 plane-oriented δZ = 4 km, clusters δR = 100 of km, hypocenters. N = 5. The output A comparative results show analysis of seismogenicthat 102 of faultsthe 200 orientation strongest withearthquakes the mechanisms form 11 plane-oriented of corresponding clusters earthquakes of hypocenters. (estimated A at comparative analysis of seismogenic faults orientation with the mechanisms of corresponding http://www.globalcmt.org/CMTsearch.html) shows an average intersection angle of 33◦. This may have earthquakes (estimated at http://www.globalcmt.org/CMTsearch.html) shows an average the following geomechanical interpretation: seismogenic planes are marked as opening mode fractures intersection angle of 33°. This may have the following geomechanical interpretation: seismogenic (hydrofractures), which formed orthogonal to the least eﬀective stress during magma injection, while planes are marked as opening mode fractures (hydrofractures), which formed orthogonal to the least the EQ’seffective mechanisms stress during planes magma are injection, marked while as shear the EQ’s faults mechanisms that initiated planes from are marked the main as shear opening faults mode faultthat (dyke) initiated using from pre-existing the main opening fractures mode systems. fault (dyke) using pre-existing fractures systems.

FigureFigure 7. Cross-section 7. Cross-section of of Kamchatka Kamchatka alongalong the CD CD line line (see (see Figure Figure 1),1 ),showing showing traces traces of active of active seismogenicseismogenic faults. faults. The The blue blue color color covers covers the areathe wherearea where plane-oriented plane-oriented seismogenic seismogenic faults faults are possible,are whichpossible, points towhich hydrofrack points to or hydrof brittlerack rock or conditions. brittle rock Numbers conditions. correspond Numbers tocorrespond the fault numbersto the fault shown numbers shown in Figure 6. The star denotes suggested active volcanoes’ primary magma chamber in Figure6. The star denotes suggested active volcanoes’ primary magma chamber positions. positions. 3. Hydrothermal Response to Magma Fracking 3. Hydrothermal Response to Magma Fracking 3.1. Uzon-Geysernaya Caldera Geysers 3.1. Uzon-Geysernaya Caldera Geysers Geysers are examples of cyclically erupted boiling hot springs. It has been shown that the driving mechanism of geysers cycling is gas-lift assisted eruptions, where non-condensable gases (mainly CO2) are important players [9,18] due to CO2 signiﬁcantly drops boiling temperatures.

Geosciences 2020, 10, 52 11 of 16

Geysers are examples of cyclically erupted boiling hot springs. It has been shown that the driving Geosciences 2020, 10, 52 11 of 16 mechanism of geysers cycling is gas-lift assisted eruptions, where non-condensable gases (mainly CO2) are important players [9,18] due to CO2 significantly drops boiling temperatures. There is evidence from gas sampling at the Velikan GeyserGeyser that at the time of full activity before Jan. 3,3, 2014,2014, gasgas composition composition was was dominated dominated by by CO CO2. Gas2. Gas sampling sampling from from September September 2013 2013 showed showed that thethat gas the component gas component of the of hydrothermal the hydrothermal reservoir reservoir that fed that the fed Velikan the Velikan Geyser Geyser was dominated was dominated by carbon by dioxidecarbon dioxide (CО2, 61.5%) (СО2, 61.5%) and nitrogen and nitrogen (N2, 32.1%), (N2, 32.1%), along withalong a with signiﬁcant а significant amount amount of methane of methane (CH4, 5.8%)(CH4, and5.8%) hydrogen and hydrogen (Н2, 0.45%) (Н2, 0.45%) [9]. Gas [9]. composition Gas composition in 2014–2019 in 2014–2019 in the Velikan in the Velikan and Bolshoy and Bolshoy geysers revealsgeysers nitrogenreveals becomingnitrogen becomi a dominantng a gas,dominant while COgas,2 declined while CO to2 less declined than 1% to [19less]. Inthan contrast, 1% [19]. a new In geyser,contrast, Shaman a new geyser, (Mutny), Shaman developed (Mutny), in 2008 developed in a channel in 2008 of in the a formerchannel hot of the springs former of Uzonhot springs caldera, of justUzon 12 caldera, km apart just [20 12]. Thekm gasapart composition [20]. The gas of thecomposition newly formed of the Shaman newly formed Geyser wasShaman characterized Geyser was by COcharacterized2 domination by accordingCO2 domination to our gasaccording sampling to our in2018 gas sampling and 2019. in 2018 and 2019. We interpretinterpret thisthis asas aa redistributionredistribution ofof thethe magmaticmagmatic gasgas rechargerecharge fromfrom magmamagma plumbingplumbing systems of the Uzon-Geysernaya caldera in the followingfollowing way: the Valley of Geysers’ hydrothermal system magmatic CO2 recharge was reduced, while that ofof thethe UzonUzon geothermalgeothermal reservoirreservoir COCO22 recharge increased.increased.

3.2. Koryaksky Narzan Thermal Springs

The Koryaksky Narzan CO2-reach thermal springs (12–14 ◦°CC)) and Isotovsky thermal springs (40–50(40–50 ◦°C)C) are located in a northnorth sectorsector ofof KoryakskyKoryaksky volcano,volcano, wherewhere magmamagma injectionsinjections have taken place sincesince 20082008 [4[4,5].,5]. WeWe have have been been doing doing continuous continuous temperature temperature monitoring monitoring in Isotovskyin Isotovsky spring spring (IS) since(IS) since 2010 2010 and and in Koryaksky in Koryaksky Narzan Nar springszan springs (KN1 (KN1 and and KN2) KN2) since since 2017. 2017. We alreadyalready reportedreported transienttransient temperaturetemperature anomaliesanomalies recorded in IsotovskyIsotovsky in 2012–2013 and in 2015–2016,2015–2016, which were apparently related to dyke injections in adjacent areas that occurredoccurred onon 2.08.2011 and 28.02.2016 correspondingly. The latest dyke injections on 14.11.2017 and 22.04.2019 are also associated with thermal tales inin IsotovskyIsotovsky springspring (Figure(Figure8 8).).

Figure 8. Temperatures recorded in Isotovsky hot spring and suggested times of dyke injections in Figure 8. Temperatures recorded in Isotovsky hot spring and suggested times of dyke injections in adjacent areas. T—observational temperature; Tav—annual maximum monthly temperature (in a adjacent areas. T—observational temperature; Tav—annual maximum monthly temperature (in a referenced year 2010/2011); Tmax—maximum monthly temperature (in a current year). Arrows above referenced year 2010/2011); Tmax—maximum monthly temperature (in a current year). Arrows above correspond to the dyke events shown in Figure2, bars on a plot above correspond to the dykes’ events correspond to the dyke events shown in Figure 2, bars on a plot above correspond to the dykes’ events magnitudes shown in Figure3. magnitudes shown in Figure 3. Koryaksky Narzan-2 (K2 in Figures2 and3) transient temperature records 2017–2019 look very stable with a few temperature drops from 0.1 to 0.4 ◦C (Figure9). The most signiﬁcant drop observed Geosciences 2020, 10, 52 12 of 16

Geosciences 2020, 10, 52 12 of 16 Koryaksky Narzan-2 (K2 in Figures 2 and 3) transient temperature records 2017–2019 look very stable with a few temperature drops from 0.1 to 0.4 °C (Figure 9). The most significant drop observed isis 0.40.4 ◦°C,C, which may bebe causedcaused byby aa HH22O–CO2 boiling temperature drop in a spring pool due to the magmatic CO22 release associated with dykes injections in the adjacent area.

Figure 9. Koryaksky Narzan-2 transient temperature records 2017–2019. The significant temperature Figure 9. Koryaksky Narzan-2 transient temperature records 2017–2019. The significant temperature drop on 0.4 ◦C may be caused by boiling temperature decline due to additional magmatic gas recharge. Thedrop arrows on 0.4 above °C may corresponds be caused to theby dykeboiling events temperature shown in decline Figure2 ,due while to theadditional bars on themagmatic plot above gas correspondrecharge. The to arrows the dyke above event correspon magnitudesds to shown the dyke on Figure events3 .shown in Figure 2, while the bars on the plot above correspond to the dyke event magnitudes shown on Figure 3. 3.3. Mutnovsky Production Reservoir 3.3. MutnovskyThe Mutnovsky Production (Dachny) Reservoir geothermal reservoir includes at least two production faults geometricallyThe Mutnovsky and hydraulically (Dachny) connectedgeothermal to thereservoir magma includes fracking systemat least of two Mutnovsky production volcano faults [7] (seegeometrically also Figure and4). hydraulically connected to the magma fracking system of Mutnovsky volcano [7] (see alsoProduction Figure 4). fault #1 or the Main production zone (dip angle 58◦, dip azimuth 110◦) includes 20 productionProduction feed-zones fault #1 as or follows: the Main wells production O29W, A3-1, zone 4-E, (dip O27, angle O19, 58°, O8, dip O45, azimuth O1, O14, 110°) A2-1, includes A2-2, O16, 20 O13,production 1, A3-2, feed-zones A4, 26, 24, as and follows: 8 and thewells Dachny O29W, thermal A3-1, 4-E, feature. O27, This O19, nearly O8, O45, coincides O1, O14, with A2-1, the “Main A2-2, productionO16, O13, 1, zone” A3-2, defined A4, 26, in24, [ 21and,22 ]8 thatand hasthe aDachny dip angle thermal of 60◦ feature.and a dip This azimuth nearly ofcoincides 106◦. with the “MainProduction production fault zone” #2 or defined the North-East in [21,22] production that has a zonedip angle (dip angleof 60° 57and◦, dipa dip azimuth azimuth 143 of◦) 106°. includes 17 productionProduction feed fault zones #2 or as the follows: North- wellsEast A2-2,production O29W, zone A3-2, (dip A4, angle 26, A2-1, 57°, dip O16, azimuth O13, 1, A3-1,143°) includes O8, O37, O42,17 production O48, O53N, feed and zones O55 (10as follows: of them alsowells belong A2-2, toO29W, production A3-2, zoneA4, 26, #1) A2-1, and the O16, Verkhne-Mutnovsky O13, 1, A3-1, O8, thermalO37, O42, feature. O48, O53N, and O55 (10 of them also belong to production zone #1) and the Verkhne- MutnovskyWe performed thermal 30-day feature. continuous monitoring of the non-condensable gases (most of gas content is COWe2) partial performed pressure 30-day in thecontinuous turbine condensermonitoring of of the the geothermal non-condensa powerble gases plant. (most To estimate of gas content PCO2, weis CO performed2) partial pressure simultaneous in the measurements turbine condenser of the of steamthe geothermal condensate power pressure plant. Pc To and estimate temperature PCO2, Tc;we then,performed PCO2 simultaneouswas calculated measur as a diementsfference of betweenthe steam Pc condensate and saturation pressure pressure Pc and corresponding temperature Tc; to temperaturethen, PCO2 was Tc. calculated as a difference between Pc and saturation pressure corresponding to temperatureFigure 10 Tc. shows the transient PCO2 change during the observational period of time from 25.08.2019 to 25.09.2019.Figure 10 It shows is clearly the seen transient that at leastPCO 142 change maxima during of PCO the2 synchronized observational with period 14 minimums of time offrom Tc, which25.08.2019 detect to non-condensable25.09.2019. It is clearly gas arrivals seen intothat the at turbineleast 14 from maxima the productionof PCO2 synchronized geothermal reservoir. with 14 Someminimums of these of PCOTc, which2 peaks detect may benon-condensable related to magmatic gas gasarrivals recharge into impulses,the turbine followed from the by production the magma frackinggeothermal processes reservoir. described Some of above these (see PCO Section2 peaks 2.2 may of thisbe related paper). to magmatic gas recharge impulses, followed by the magma fracking processes described above (see Section 2.2 of this paper).

Geosciences 2020, 10, 52 13 of 16 Geosciences 2020, 10, 52 13 of 16

Figure 10. Estimated partial pressure of CO2 in the condenser at the outlet of the turbine of the geothermalFigure 10. Estimated power plant. partial pressure of CO2 in the condenser at the outlet of the turbine of the geothermal power plant. 4. Discussion 4. Discussion The key question in this paper is: How can we verify that plane-nested earthquakes are tracking magmaThe injections key question in the in shapesthis paper of dykes is: How or sillscan we beneath verify active that plane-nested volcanoes? The earthquakes answer is are in thetracking form ofmagma a question, injections too: in what the othershapes fluid of dykes except or magma sills beneath can demonstrate active volcanoes? such hydrofracking The answer is capabilities? in the form Superheatedof a question, water too: what and non-condensableother fluid except gases magma (possibly can demonstrate CO2) are other such candidates hydrofracking for such capabilities? working fluidSuperheated duties. water However, and non-co superheatedndensable water gases in shallow(possibly permeable CO2) are other fracture candidates reservoirs for such conditions working is veryfluid sensitiveduties. However, to host rock superheated temperatures water and in shallow is easily permeable converted fracture into high-compressible reservoirs conditions two-phase is very conditions,sensitive to forming host rock geothermal temperatures production and is fields. easily Another converted fluid into isCO high-compressible2, especially magmatic two-phase CO2 havingconditions, less forming compressibility geothermal as compared production to superheatedfields. Another steam, fluid and is CO it may2, especially act as a magmatic workingfluid CO2 couplinghaving less with compressibility magma. If so, as thencompared we should to superheated extend our steam, term ofand magma it may fracking act as a beneathworking active fluid volcanoescoupling with to the magma. term magma If so, +thenCO2 wefracking should beneath extend active our term volcanoes. of magma CO2 frimpactacking is beneath also useful active to explainvolcanoes traces to the of fracks term inmagma+CO a slope of Koryaksky2 fracking beneath volcano (Figureactive volcanoes.3), with no associatedCO2 impact magma is also discharge useful to onexplain the surface tracesat of the fracks same in time. a slope In this of connection, Koryaksky it volcano is worth noting(Figure that 3), magmaticwith no associated CO2 redistribution magma appearsdischarge to on be the keysurface reason at the to explainsame time. the transferIn this connection, of geysers activity it is worth from noting Geysers that Valley magmatic to Uzon CO in2 Kamchatkaredistribution and appears recent (2018)to be the reactivation key reason of Steamboatto explain the Geyser transfer in Yellowstone. of geysers activity from Geysers ValleyAnother to Uzon important in Kamchatka point and is therecent accuracy (2018) reactivation of seismic hypocenter of Steamboat data Geyser and in uniqueness Yellowstone. of the Frac-DiggerAnother method important for plane-oriented point is the accuracy shapes of definition. seismic hypocenter There is no uniqueness,data and uniqueness since we foundof the moreFrac- fracksDigger using method less for severe plane-oriented criteria of selections shapes definition in Frac-Digger. There (see is Sectionno uniqueness, 2.1). Thus, since we we suggest found that more 3D distributionsfracks using less of magma severe frackscriteria be of considered selections in a plausible Frac-Digger scenario (see Section of magma 2.1). fracking Thus, we beneath suggest active that volcanoes,3D distributions but these of magma may include fracks more be considered or less magma a plausible+CO2 dykes, scenario depending of magma on fracking the Frac-Digger beneath selectionactive volcanoes, parameters. but Nevertheless,these may include some more Frac-Digger or less selectionsmagma+CO pointed2 dykes, to 90%–95%depending of on earthquake the Frac- hypocentersDigger selection belonging parameters. to plane-oriented Nevertheless, clusters, some meaning Frac-Digger that fracking selections is apointed dominant to process90%–95% there. of earthquakeHigh-temperature hypocenters (HT) belonging geothermal to plane-oriented system formation clusters, due to meanin magmag fracking that fracking is also wellis a explaineddominant inprocess terms there. of the magma thermal-hydrodynamic modeling. ConceptualHigh-temperature 2D iTOUGH2-EOS1sc (HT) geothermal system thermal formation hydrodynamic due to modelingmagma fracking of the is Mutnovsky also well magmatic–hydrothermalexplained in terms of the systemmagma [ 7thermal-hydrodynamic] reasonably explains its modeling. evolution over the most recent 1500–5000 yearsConceptual in terms of 2D heat iTOUGH2-EOS1sc recharge (supplied thermal by injectedhydrodynamic dykes modeling from the of active the Mutnovsky funnel Mutnovsky-4) magmatic– andhydrothermal mass recharge system (water [7] reasonably injected through explains the its dormant evolution volcanic over the funnels most Mutnovsky-3 recent 1500–5000 and possiblyyears in Mutnovsky-2)terms of heat recharge conditions. (supplied We emphasize by injected that thedykes magmatic from the injection active ratefunnel is approximately Mutnovsky-4) equivalentand mass torecharge the heat (water discharge injected rate through (455 MWt). the dormant This is equivalentvolcanic funnels to approximately Mutnovsky-3 455 and kg possibly/s of magma, Mutnovsky- which is2) inconditions. turn equivalent We emphasize to from that 15.3 the to 25.6magmatic km3 of inject magmaion rate over is 3000–5000approximately years. equivalent This volume to the value heat discharge rate (455 MWt). This is equivalent to approximately 455 kg/s of magma, which is in turn

Geosciences 2020, 10, 52 14 of 16 is comparable to the available space for dyke accommodation under regional horizontal extension conditions [23]. Conceptual TOUGH2 modeling was used to understand and explain the mechanism of the formation of the hydrothermal system beneath northern slope of Koryaksky Volcano [4,5]. For this purpose, the following terms were found to be crucial in this model: (1) heat sources of 20 MW/km3 (340 3 3 MW total in 17 km of reservoir rocks) and gas (CO2) sources of 10 g/s/km acting during 7000 years in the zones of magma injections; and (2) cold water recharge of 580 kg/s through the volcanic funnel to the deep dyke injection area. The modeling results reasonably match the Na–K geotemperature 18 estimates of geothermal reservoirs (300 ◦C), the isotopic values (δD, δ O) of high-elevation meteoric water recharge, the concentrations of magmatic CO2 (up to 4 g/kg) in the hot springs on the northern slope of Koryaksky Volcano, and the thermal reactions to the dyke injections recorded in Isotovsky and Koryaksky Narzan hot springs. This modeling also indicates that a hidden high-temperature geothermal reservoir is present also beneath the southern slope of Koryaksky Volcano (at an elevation of 1 km), which may become a subject of future drilling explorations. − Another interesting related issue is where magma came from to recharge volcanoes for magma fracking and consequent eruptions. The possible answer is that water release from the subduction plate (due to hydro-fracking in the plate itself) forms upﬂows into the mantle edge, which creates melting conditions for primary magma chambers at the depth of 150 km (Figure7). This closed loop of water–magma–water circulation in the subduction zone may be responsible for strong earthquakes, when hydrofracking drives the opening of shear faults. A mini-loop of such water–magma interplay may be Karymsky Volcano, which has been working almost continuously since 1996 due to water recharge fed from Karymsky Lake into dip seismogenic faults #120 and #122 and from there into the magma chambers of the volcano (Figure6).

5. Conclusions (1) Active volcanoes are injectors of magma and water into adjacent structures, which creates high-temperature production reservoirs. (2) Seismic data reveal magma hydrofracking and stress conditions around active volcanoes. (3) Magma fracking reservoirs may reside within production geothermal reservoirs. (4) Seismogenic faults on the Kamchatka shelf are indicators of geoﬂuid generation and water propagation to 250 km depth. (5) Geysers result from CO2-gaslift in active silicic volcanic areas. Continuously performed measurements of CO2 partial pressures in hydrothermal features are a possible key to understanding the mechanisms of the relationship between anomalous hydrothermal perturbations [24], volcanic activity, and strong earthquakes.

Author Contributions: Conceptualization, methodology, software, validation, data curation, writing—original draft preparation: A.K. Koryaksky and Avachinsky volcanoes MEQ’s data curation and analysis: E.C.; Mutnovsky MEQ’s data curation and analysis, field works organization: A.P.; EQ mechanisms analysis: A.S. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by RFBR project # 18-05-00052-19. Acknowledgments: Authors express gratitude to P.A., P.O. Voronin, N.B. Zhuravlev, M.V. Lemzikov, O.O. Usacheva, T.V. Rychkova, M.A. Maguskin, A.I. Kozhurin and Sergeeva A.V. for their help in field survey and data analysis. Special thanks for constructive comments to A.A. Lyubin. Thorough editing of the manuscript performed by J.C. Eichelberger highly appreciated. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Geosciences 2020, 10, 52 15 of 16

References

1. Sigmundsson, F.; Hooper, A.; Hreinsdóttir, S.; Vogfjörd, K.S.; Ófeigsson, B.G.; Heimisson, E.R.; Drouin, V. Segmented lateral dyke growth in a rifting event at Bárðarbunga volcanic system, Iceland. Nature 2015, 517, 191–195. [CrossRef][PubMed] 2. Kiryukhin, A.V.; Fedotov, S.A.; Kiryukhin, P.A. A geomechanical interpretation of the local seismicity related to eruptions and renewed activity on Tolbachik, Koryaksky, and Avacha Volcanoes, Kamchatka, in 2008–2012. J. Volcanol. Seismol. 2016, 10, 275–291. [CrossRef] 3. Kiryukhin, A. Analysis of magma injection beneath an active volcano using a hydromechanical numerical model. Horiz. Wells 2017, 1, 1–5. 4. Kiryukhin, A.V.; Fedotov, S.A.; Chernykh, E.V. Magmatic plumbing systems of the Koryakskii–Avacha Volcanic Cluster as inferred from observations of local seismicity and from the regime of adjacent thermal springs. J. Volcanol. Seismol. 2017, 11, 321–334. [CrossRef] 5. Kiryukhin, A.; Lavrushin, V.; Kiryukhin, P.; Voronin, P. Geofluid systems of Koryaksky-Avachinsky volcanoes (Kamchatka, Russia). Geofluids 2017, 2017, 1–21. [CrossRef] 6. Kiryukhin, A.V.; Vorozheikina, L.A.; Voronin, P.О.; Kiryukhin, P.A. Thermal-Permeability structure and recharge conditions of the low temperature Paratunsky geothermal reservoirs, Kamchatka, Russia. Geothermics 2017, 70, 47–61. [CrossRef] 7. Kiryukhin, A.; Polyakov, A.; Usacheva, O.; Kiryukhin, P. Thermal-permeability structure and recharge conditions of the Mutnovsky high-temperature geothermal field (Kamchatka, Russia). J. Volcanol. Geotherm. Res. 2018, 356, 36–55. [CrossRef] 8. Kiryukhin, A.V.; Fedotov, S.A.; Kiryukhin, P.A. Magmatic systems and the conditions for hydrothermal circulation at depth in the klyuchevskoy volcanic cluster as inferred from observations of local seismicity and thermo-hydrodynamic simulation. J. Volcanol. Seismol. 2018, 12, 231–241. [CrossRef] 9. Kiryukhin, A. Modeling and observations of geyser activity in relation to catastrophic landslides–mudflows (Kronotsky nature reserve, Kamchatka, Russia). J. Volcanol. Geotherm. Res. 2016, 323, 129–147. [CrossRef] 10. Zoback, M.D. Reservoir Geomechanics; Cambridge University Press: Cambridge, UK, 2010; p. 448. 11. Fujii, Y.; Kodama, J.; Fukuda, D. Upper bounds of seismic events in induced seismicity. J. Open Geosci. 2010. submitted for publication. 12. Chernykh, E.V.; Kiryukhin, A.V. Comparison of the geometry of seismogenic plane-sets and the mechanisms of the foci of earthquakes of the Koryaksky Volcano in 2008–2009. In Proceedings of the Geothermal Volcanology Workshop, Petropavlovsk Kamchatsky, Russia, 4–9 September 2019; pp. 58–63. 13. Selyangin, O.B. Mutnovsky Volcano, Kamchatka: New evidence on structure, evolution, and future activity. J. Volcanol. Seismol. 1993, 15, 17–38. 14. Selyangin, O.B. Wonderful World of Mutnovsky and Gorely Volcanoes: Volcanologic and Traveller’s Guide; Novaya Kniga: Petropavlovsk-Kamchatsky, Russia, 2009; p. 108. 15. Polyak, B.G.; Melekestsev, I.V. On the output of volcanoes. J. Volcanol. Seismol. 1981, 5, 22–37. 16. Kiryukhin, A.; Fedotov, S.; Solomatin, A.; Kiryukhin, P. Geomechanical interpretation of seismicity on Kamchatka Shelf: Applications for seismic forecast and hydrocarbon exploration. In Proceedings of the 20th Conference on Oil and Gas Geological Exploration and Development, Gelendzhik, Russia, 10–14 September 2018. 17. Fedotov, S.A.; Maguskin, M.A.; Kirienko, A.P.; Zharinov, N.A. Vertical ground movements in the coast of the Kamchatka Gulf: their specific features in the epicentral zone on August 17, 1983 Earthquake М= 6.9, before and after. Tectonophysics 1992, 202, 157–162. [CrossRef] 18. Hurwitz, S.; Clor, L.E.; McCleskey, R.B.; Nordstrom, D.K.; Hunt, A.G.; Evans, W.C. Dissolved gases in hydrothermal (phreatic) and geyser eruptions at Yellowstone National Park, USA. Geology 2016, 44, 235–238. [CrossRef] 19. Kiryukhin, A.; Sugrobov, V.; Sonnenthal, E. Geysers VALLEY CO2 cycling geological engine (Kamchatka, Russia). Geofluids 2018, 2018, 1–16. [CrossRef] 20. Karpov, G. Evolution of regime and physical–chemical characteristics of the new formed Geyser in Caldera Uzon (Kamchatka). J. Volcanol. Seismol. 2012, 3, 3–14. 21. Kiryukhin, A.V. High temperature fluid flows in the Mutnovsky hydrothermal system, Kamchatka. Geothermics 1993, 23, 49–64. [CrossRef] Geosciences 2020, 10, 52 16 of 16

22. Kiryukhin, A.V.; Takahashi, M.; Poliakov, A.Y.; Lesnykh, M.D.; Bataeva, O.P.Origin of water in the Mutnovsky geothermal ﬁeld an oxygen (δ 18O) and hydrogen (δD) study. J. Volcanol. Seismol. 1999, 20, 441–450. 23. Basmanov, O.L.; Kiryukhin, A.V.; Maguskin, M.A.; Dvigalo, V.N.; Rutqvist, J. Thermo-hydrogeomechanical modeling of vertical ground deformation during the operation of the Mutnovskii Geothermal Field. J. Volcanol. Seismol. 2016, 10, 138–149. [CrossRef] 24. Kiryukhin, A.V.; Korneev, V.A.; Polyakov, A.Y. On possibility of relationship between strong earthquakes and anomalous pressure variations in two-phase geothermal reservoir. J. Volcanol. Seismol. 2006, 6, 3–11.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).