An Mw 5.5 earthquake occurred in Pohang, South Korea on November 15, 2017, resulting in a great impact on society. Despite a lot of controversy about the cause of the earthquake in relation to the enhanced geothermal system (EGS), the location of earthquake-related active faults is poorly known. Here, we first report the results of the geochemical and isotopic analyses of dissolved gases in groundwater in the Heunghae, Yeonil, and Sinkwang areas. According to the N2-Ar-He relationship, samples from the Heunghae and Yeonil areas are contributed to the mantle, except for the Sinkwang area, where all samples are atmospheric. The Pohang samples consist mainly of N2 and CO2, and some samples of the Heunghae and Yeonil areas contain substantial CH4. Stable isotope compositions of N2 (), CO2 (), and CH4 () indicate that these components are derived from organic substances in sedimentary layer of Pohang Basin. On the other hand, helium isotope ratios (3He/4He, up to 3.83 Ra) represent the significant mantle contribution in the Heunghae and Yeonil areas. Through the distribution of high 3He/4He ratios, we propose that the Heunghae, Namsong, and Jamyeong faults are the passage of mantle-derived fluids. Computed 3He fluxes of the Heunghae (120 to 3,000 atoms cm-2 sec-1), Namsong (52 to 1,300 atoms cm-2 sec-1), and Jamyeong (83 to 2,100 atoms cm-2 sec-1) faults are comparable to other major active faults around the world, reflecting either high porosity or high helium flow rates. Therefore, our results demonstrate that there are active faults near the EGS facilities, which can provide the basis for future studies.

1. Introduction

Enhanced geothermal system (EGS) is a type of heat exchanger designed to improve the efficiency of geothermal energy plants. EGS is configured to enable convective production or to improve heat production. One of the main goals of EGS is to increase the permeability of reservoir rocks with high temperatures but low permeability. For this purpose, hydraulic fracturing, fluid injection (and/or extraction), and acidification can be used [1].

The correlation between EGS and seismic activity has been proposed for decades [1]. Geysers (USA), Cooper Basin (Australia), Berlin (El Salvador), Soultz-Sous-Forêts (France), and Basel (Switzerland) are well-known examples of EGS-related earthquake activities. In addition to these cases at the EGS site, for other cases such as wastewater injection, carbon capture and storage (CCS), or hydrocarbon (e.g., shale gas and oil) extraction, fluid injection, and hydraulic fracturing are often proposed as triggers of earthquakes (e.g., Keranen et al. [2]). Two mechanisms for triggering earthquakes associated with fluid injection and/or extraction are described by McGarr et al. [3]; (1) direct fluid pressure effects on injection and (2) changes in solid stress due to fluid extraction and/or injection. As we can infer from earthquake-inducing mechanisms of fluid injection, knowing locations of faults around the EGS site is important for society. For both cases, it is necessary to identify not only faults beneath the EGS site but also any potentially unknown faults. In the case of the 2019 Ridgecrest earthquake, for example, it shows how the earthquake swarm can propagate to interlocked faults [4]. This refers to the possibilities that the small induced earthquake can trigger much larger seismicity than expected, which amplifies the seismic hazards and risks around the EGS site.

Since mantle-derived fluids have been identified and quantified from the San Andreas Fault system [5] in the nonvolcanic region, noble gas studies have been conducted in several active fault zones to understand the fluid behavior related to seismicity (e.g., Sano et al. [6]). Even in some cases, helium isotope ratio distribution could detect concealed fault zones [7]. Noble gases and their isotope compositions can be used as natural tracers. They are chemically inert, retaining their properties through the water-rock systems. Therefore, the contents and isotopic compositions of noble gases allow us to trace the fluid sources into mantle, crust, and atmosphere. It is also possible to quantify the contribution of each source [8]. Mantle-derived helium and CO2 degassing through faults in the southeastern Korean peninsula has recently been reported [9].

Here, we report new chemical and isotopic compositions of dissolved gases in groundwater, which are rarely been documented in the study area. Then, we will first discuss the general characteristics of gas compositions in this area. Based on the results, the perceptual impact on the composition of the noble gas is assessed to suggest that there are active fault zones near the EGS site, reaching the upper mantle through continental crust. In addition, helium flux through faults was calculated and compared with the characteristics of major fault zones around the world.

1.1. Geological Setting

The Pohang region consists of three subdivisions: Heunghae, Yeonil, and Sinkwang areas (Figure 1). The Heunghae and Yeonil areas are located in Pohang Basin, the Miocene sedimentary basin, and the Sinkwang area is near the Yangsan fault (Figure 1). Pohang Basin is one of the largest sedimentary basins in the Korean Peninsula [10]. The border faults in Pohang Basin are the Yeonil tectonic line, the Ocheon fault system, and segmented faults bounding to the west (Figure 1). These western boundary segment faults are almost parallel to the Yangsan fault, 2-7 km away from the west. During the early Miocene period, Pohang Basin was formed while the East Sea (Sea of Japan) which is a back-arc basin was opened, and the opening was ceased at ~15 Ma [11, 12]. The sedimentation in the Pohang Basin lasted from 17 to 10 Ma [11, 12]. The basin is filled with Paleogene volcanic rocks and granite, followed by Neogene conglomerates, sandstone, and mudstone, with a total thickness of less than 500 m [13, 14]. The basin has been cut by normal faults with the NNE-SSW strike and eastern dip. These normal faults have been formed after 15 Ma, blocked by other normal faults with the ENE-WSW strike [14]. The Sinkwang area is spread over Cretaceous biotite granite cut by the Yangsan fault covered with quaternary sediments (Figure 1, [14]). The Yangsan fault, located at the current Cretaceous sedimentary area, is a strike-slip fault formed in the early Cretaceous period as a result of tension due to the subduction of the Izanagi plate [15, 16]. After the subduction of the Pacific plate in the late Cretaceous period, compressive stresses affected the Yangsan fault. The direction of compressive stress was initially in the NW-SE direction at the end of the Cretaceous period, and the direction of subduction at the end of the Paleogene changed and moved in the NE-SW direction [15, 16].

1.2. 2017 Pohang Earthquake

The Pohang enhanced geothermal system (EGS) project was launched in November 2010 to produce 160°C geothermal water and 1.2 MW geothermal energy in a nonvolcanic area [17]. To construct the EGS facility, two boreholes (PX-1 and PX-2) were drilled through the sedimentary basin into the granodiorite basement rock. The measuring depths of the two boreholes (MD, measured along the borehole) are 4,362 m and 4,382 m, respectively. Hydraulic fracturing and fluid injection were performed from January 2016 to September 2017 to increase geothermal productivity [17]. During this period, five hydraulic stimuli were performed through PX-1 and PX-2. First, the third and fifth hydraulic stimuli were performed on the PX-2 with maximum well-head pressures of 89.2 MPa, 88.8 MPa, and 84.6 MPa, respectively. The second and fourth hydraulic stimuli were performed on the PX-1, and the maximum well-head pressures were 27.71 MPa and 25.16 MPa, respectively.

After the third hydraulic stimulation, an earthquake occurred with Mw 3.2. About two months after the cessation of the fifth hydraulic stimulus, on November 15, 2017, an earthquake with Mw 5.5 occurred (Figure 1), followed by more than one hundred aftershocks (≥ Mw 2.0). The earthquake was the second-largest earthquake in the Korean Peninsula since modern earthquake observation has begun in 1978, resulting in physical and economic damage to local residents (135 people were injured and more than 1,700 people were in emergency housing; directly USD 75 M and total economic impact USD 300 M) [17]. Due to this great impact on Korean society, it is necessary to study active faults related to potential seismic crises in this area.

2. Methods

We collected groundwater samples from groundwater wells in the Heunghae, Yeonil, and Sinkwang areas, Pohang, Republic of Korea (Table 1, Figure 1). Ranges of water temperatures and pH are 14.3 to 20.5°C and 6.0 to 9.0, respectively, and well depth ranges between 30 and 230 m from the topographic surface. The samples were stored in copper tubes and sealed with clamps, except for PH-3 that was collected in a preevacuated Giggenbach bottle. Dissolved gases were extracted from water samples by the high vacuum system and analyzed in the Atmosphere and Ocean Research Institute (AORI), the University of Tokyo. Concentrations (consist of CO2, N2, O2, CH4, Ar, and He) of dissolved gases were measured by a Pfeiffer QMS 200 quadrupole mass spectrometer (QMS). Stable isotope compositions of nitrogen for N2 and carbon for CO2 and CH4 were measured by an isotope ratio mass spectrometer (Isoprime 100 by Elementar). For samples with high CH4 concentration (>4%), coexisting CO2 and CH4 were separated before measurement by liquid nitrogen. To measure 3He/4He and 4He/20Ne ratios, dissolved gas samples were purified by titanium getters at 400°C and charcoal traps at liquid nitrogen temperature. Neon was trapped by the cryogenic pump at 40 K after measuring 4He/20Ne ratios via online QMS (Pfeiffer Prisma 80). Then, purified helium was injected into a noble gas mass spectrometer (Helix SFT by ThermoFisher) to measure 3He/4He ratios. Calibration of He isotope ratios was conducted by using the internal He standard of Japan (HESJ) [18]. Measured 3He/4He ratios were corrected for atmospheric helium by using measured 4He/20Ne ratios, since 20Ne is assumed to be mostly atmospheric [19]. From Sano et al. [20]: where is the corrected 3He/4He ratio, and is the 4He/20Ne ratio of air-saturated water (ASW). Analytical errors for and 3He/4He and 4He/20Ne ratios are about 1% and 5% (1σ), respectively.

3. Results

The measured gas compositions are reported in Table 1. N2 is the most abundant gas component for most samples with the range of 32.5 to 94.0 vol.%. CO2 is also observed in all samples ranging from 0.1 to 45.4 vol.%. CH4 is also one of the main components for samples in the Heunghae and Yeonil areas (for example, P-14, PH19-12, PH19-07, PH19-08, and PH-3) with the range from 18.7 to 42.9 vol.%. Ar and He range from 0.6 to 2.1 vol.% and from 7 to 2744 ppm, respectively. Except for PH-3 (), almost no O2 is observed in most samples, indicating minimum air contamination during sampling and analysis. Isotopic compositions of N, C, and He are summarized in Table 2. The nitrogen isotope compositions of N2 (δ15N-N2) for all samples range from 0.19 to 3.56‰, all heavier than air (0‰). Carbon isotope values of CO2 (δ13C-CO2) for all samples range from -27.33 to -16.01‰. There is no significant regional difference for δ15N-N2 and δ13C-CO2. The carbon isotope composition of CH4 (δ13C-CH4) is from -76.05 to -70.04‰, implying similar CH4 sources for both Heunghae and Yeonil areas. In these areas, most samples show 4He/20Ne ratios higher than ASW (, [21]), with 3He/4He ratios ranging from 0.18 to 3.83 Ra, where Ra is the 3He/4He ratio of air (, [22]). The 3He/4He ratios (0.90 to 1.22 Ra) of the Sinkwang groundwater samples are atmospheric because 4He/20Ne ratios of the samples are similar to air.

4. Discussion

4.1. Gas Geochemistry

Based on the N2-Ar-He ternary diagram (Figure 2), nonreactive gases (N2, Ar, and He) in the Heunghae and Yeonil areas display a two-component mixing relationship between the mantle and atmospheric end members. Lee et al. [9] have shown that fault gases in the southeastern Korean peninsula are continental gases rather than subduction zone gases. Dissolved gases in groundwater from the Sinkwang area are atmospheric with low He/Ar ratios (0.003 to 0.001). Lee et al. [9] addressed that some gas samples which are severely contaminated by air in southeastern Korea are due to shallow well depths (130 to 296 m). However, in the Heunghae and Yeonil areas, it is unlikely that well depths (Table 1) and contribution from deeply derived gases are relevant. Instead, we suggest that distance to the permeable fault zones where gases released from deep sources are transported is important. In general, groundwater wells in the Heunghae and Yeonil areas contain more dissolved helium (Table 2), implying there are more permeable areas than the Sinkwang area.

CO2 concentrations are negatively correlated with pH (), indicating that CO2 is likely to be trapped in high pH water to be present as HCO3- and CO32- (Figure S1a, [23, 24]). To verify this, we show that CO2/N2 () and CO2/CH4 () ratios also have negative trends with pH (Figure S1b, c). It is attributable that CO2 removal increased those ratios because N2 and CH4 are not responsive to pH changes. Also, N2 and/or CH4-rich gases are found in alkaline springs [23, 25, 26]. Moreover, CO2/CH4 ratios and helium concentrations show a negative correlation (, Figure S1d). Although Lee et al. [9] suggested both CO2 and helium are derived from the mantle source in southeastern Korea, it is plausible that their origins are decoupled in the Pohang region.

4.2. Origins of Nitrogen, Carbon Dioxide, and Methane

δ15N-N2 values of all samples with the range of 0.19 to 3.56‰ are between the air (0‰) and sediment (7‰) end-members. By plotting δ15N with N2/3He ratios (Figure 3(a)), we can identify contributions of the air, sediment, and mantle end-members. To quantify the contribution of each end-member, we adopted the three-component mixing model from Sano et al. [27]: where obs is the observed value; , , and are the contributions of the mantle, sediments, and air; δ15Nmantle, δ15Nsediment, and δ15Nair are , , and 0‰; N2/3Hemantle, N2/3Hesediment, and N2/3Heair are , , and , respectively [21, 27, 28, 29 and references therein]. The results are summarized in Table S1. Air is the most dominant source for N2 in the Pohang region with the fA range of 49.1 to 97.0%. Also, sediment is another main source for N2 with the fS range of 2.8 to 50.9%. The mantle contribution is very minor (), indicating N2 is the primarily sedimentary origin and is contributed by air at shallow depths. Moreover, the sediment-derived N2 in the Pohang region is of shallow origin rather than the recycled nitrogen through subduction as discussed in section 4.1.

The Pohang region has no mantle-derived CO2 which has been reported in Gyeongju and Ulsan areas, southeastern Korea (Lee et al., 2019). δ13C values (-27.33 to -16.01‰) of CO2 in all samples are lighter than the MORB value (, [30]) and lie approximately between the mean δ13C-CO2 values of C3 (-27‰) and C4 (-13‰) plants [31] (Figure 3(b)). The results are similar to those of most fault gases previously reported in southeastern Korea () as well as δ13C-CO2 values (−11.9 to −24.0‰) in global fault zones without magma activity, such as San Andreas Fault and North Anatolian Fault [9 and references therein]. On the δ13C-CO2 vs CO2/3He plot (Figure S2), a majority of samples are outside the mixing curve between the biogenic and mantle end-members due to their low CO2/3He ratios. As discussed in section 4.1, the decrease in CO2/3He ratios can be attributed to CO2 loss in this area under the influence of pH. Moreover, the negative correlation () between 4He concentrations and CO2/3He ratios is displayed well (Figure S1e). Gilfillan et al. [32] have shown the same trend for natural gas fields in North America. They argued that noble gases are unlikely involved in increasing or decreasing CO2/3He ratios. However, the trend between 3He/4He and CO2/3He ratios (Figure S1f) is also negatively correlated (). Thus, in the Pohang region, we argue that not only CO2 loss but also external helium was introduced to reduce CO2/3He ratios in the local groundwater layer (see section 4.3-4.5). This further supports that CO2 source derived from shallower sediments than the mantle.

The origin of CH4 in Pohang is relatively uniform. In the measured samples (Table 2), δ13C values of CH4 range from -76.05 to -70.04‰, indicating a typical microbial origin [33]. Considering δ13C of CO2, the mechanism of production of CH4 is resulted by carbonate reduction with slight oxidation after the process of methanogenesis (Figure S2b, [34]). Therefore, the isotope separation factors (εCCO2-CH4) which are approximately from 54.36 to 58.12 allow us to estimate the growth temperature of CH4 at ~40°C [34].

4.3. Helium Isotope Geochemistry in the Pohang Area

In the Pohang region, higher 3He/4He ratios (up to 3.83 Ra) than ASW/air are found in the Heunghae and Yeonil areas. Although elevated 3He/4He ratios can be resulted by the 3H-derived 3He, it is unlikely because the samples with high 3He/4He ratios also have high 4He/20Ne ratios (up to 145.29). Helium is also remobilized from old igneous rocks [35]. It is known that Pohang Basin is filled by sediments on the granodiorite basement rocks with some Tertiary basaltic rocks (Daljeon basalt) which erupted at 13.8 Ma [14 and references therein]. However, the basaltic rocks show a limited distribution (<1 km2) in the Yeonil area [14 and references therein]. Considering the typical 4He contents in inclusion bearing olivine (10-8 to 10-9 ccSTP/g, [8, 35]) and the 4He concentrations of high 3He/4He ratio samples from the Yeonil area ( to ccSTP/g), the amounts of trapped 3He in the Daljeon basalt is insufficient to be the source of 3He in the Yeonil area. Furthermore, in consideration of general Li concentration of igneous and sedimentary rock, radiogenic 3He by decay reaction 6Li(n, α)3H(β-)3He cannot affect 3He/4He ratio of groundwater [35 and references therein]. Therefore, the excess 3He in the Pohang region originates from the mantle as well as 4He [23, 36].

The range of 3He/4He ratios (0.18 to 3.83 Ra) in the Heunghae area is wide, indicating that both mantle and crustal helium sources are present in a short range, up to 10 km (Figures 1 and 4). Even though some samples (PH19-09 and PH-3) in the Yeonil area are atmospheric (Figures 1 and 4), two samples (PH19-07 and PH19-08) have high 3He/4He (3.09 Ra and 2.32 Ra) and 4He/20Ne (145.29 and 61.86) ratios, implying the presence of mantle-derived fluids. All samples in the Sinkwang area show atmospheric 3He/4He (0.90-1.34 Ra) and 4He/20Ne (0.27-0.36) ratios (Figures 1 and 4). A 3He/4He ratio of up to 5.69 Ra was reported for the Yangsan fault zones [9]. The absence of mantle signatures for all samples in the Sinkwang area can be explained by the distance from the main fault line (Figure 1), which means the influx of external helium is quite low.

The highest 3He/4He ratio (3.83 Ra) represents about 50% of the mantle contribution to the fluid (Figure 4). It is known that mantle-derived helium can be actively released to the surface through magmatism [37, 38]. However, the Pohang region is located hundreds of kilometers away from the active volcanoes of the Japanese arc (Figure 1). A low-velocity zone beneath Ulleungdo has been proposed by Chen et al. [39] and references therein, which is also 200 km away from Pohang. Moreover, magma activity in this area has been ceased after 9,300~6,300 BP [40 and references therein].

The appropriate model for the occurrence of mantle-derived helium in this region is that there are permeable fault zones like the release of mantle fluids in the San Andreas Fault zones [5]. According to Song [14], there are some faults in the Pohang region, such as Heunghae, Gokgang, Hyeongsan, and Ocheon faults (Figure 1). Also, Westaway and Burnside [41] named the new fault as the Namsong fault on the basis of the aftershock distribution of the Pohang earthquake (Figure 1) and proposed that the fault has been already critically stressed before the EGS project [41]. In addition, according to the Korea Meteorological Administration (KMA), the depths of the 2017 Pohang earthquake and aftershocks are less than 16 km, which is shallower than the Moho depth (~28 km) of the region [42]. Kennedy and Van Soest [37] suggested that the mantle fluids of the San Andreas Fault penetrate the brittle-ductile boundary based on helium isotope ratios and strain rates measured by GPS. To explain the mantle-derived helium in the Gyeongju and Ulsan areas, southeastern Korea, Lee et al. [9] also proposed that mantle helium migrates along the ductile shear zone underneath the brittle regime.

From the above information, in the Pohang region, we suggest that there have been already tectonically active areas that have developed from the ductile shear zone to the brittle fault zone. This condition enabled the inflow of the mantle-derived fluids through the lower crust into the permeable faults. Furthermore, in this area, relatively high temperatures, heat contents, and heat flows have been reported [43 and references therein], supporting that 3He came from the mantle through active faults (e.g., Umeda et al. [44]).

4.4. Distribution of Active Faults

As discussed in section 4.3, we identified the existence of permeable faults in the Heunghae and Yeonil areas. The locations of the faults can be constrained based on the geographical distribution of helium isotope ratios [7]. Since the latitude variation in the 3He/4He ratios is the most prominent to specify fault locations, we can show the relationship between the 3He/4He ratios and latitude (Figure S3). Although it is known that the location of the Heunghae fault is still ambiguous, we found that the distribution of higher 3He/4He ratios are well consistent with the fault striking EW at 36.126°N (Figure 1). Also, there is a relationship between higher 3He/4He ratios and distance from the fault line (Figure 5(a)), which has been observed in the San Andreas, North Anatolian, and Karakoram faults [5, 8, 37, 45]. In Figure 5(a), we could observe a sample (PH19-12) with a high 3He/4He ratio (2.25 Ra), which is about 3 km away from the Heunghae fault to the south (Figures 1, S3). The sampling location of PH19-12 is still in the area where aftershocks have frequently occurred (Figure 1), suggesting there can be another 3He discharge in the Heunghae area. To confirm the pathway of mantle-derived helium, we calculated the distance of all samples from the Namsong fault (Figure 5(b)). Based on the mainshock strike (N34°E) and dip (51°NW), we assumed that the easternmost boundary of the aftershock occurrences (from 2017 Annual report of Earthquake) with the N34°E strike is the uppermost line of the fault. By using this uppermost line, the distance from the closest fault and 3He/4He ratios are displayed (Figure 5(c)), showing a better correlation with exponential distribution () than Figure 5(a). To validate the relationship between the distance from faults and 3He/4He ratios, we selected six samples at latitudes higher than P12 (Figure 5(a), red circles). These samples are well correlated with distance from the Heunghae fault exponentially (), validating the model of the Heunghae-Namsong fault system as shown in Figure 5(c). Samples with lower 3He/4He ratios than 1 Ra indicate that the inflow of 3He into the aquifer is less than crustal or atmospheric contributions (Figure 4). Although it is not well known about the exact fault locations in the Yeonil area, previous studies have reported the presence of faults [14 and references therein]. Based on some samples with high 3He/4He ratios (Figure 1), we suggest that there can be a highly permeable fault zone. In this study, we propose to name the Jamyeong fault considering the name of the village called Jamyeong-ri in this area.

4.5. Helium Flux from the Faults

To compare fluid dynamics with other fault systems in the world, helium flow rates and 3He flux were estimated. We calculated helium flow rates from the most reliable mantle helium source from each fault (P-12 for Heunghae fault, PH19-07 for Jamyeong fault, PH19-12 for Namsong fault), following Menzies et al. [46]: where is the helium flow rate (cm/yr); is the crust thickness; and are the density of the crust and the fluid, respectively; is the current production rate of 4He from crust; is the concentration of helium in the original mantle fluid; , , and represent helium isotopic ratios (3He/4He) of the sample, crust, and mantle, respectively. The crust below the Pohang region was assumed to be a double layer with thicknesses for the upper and middle crust of 12.3 and 15.6 km, respectively [42, 47]. We used the average density values of the middle crust (2.72 g/cm3) and the upper crust (2.65 g/cm3) as and the water density (1 g/cm3) as [48 and references therein]. From Menzies et al. [46], P(He) (ccSTP g-1 yr-1) can be obtained as: where 4He production rates of U and Th are ccSTP μg-1 yr-1 and ccSTP μg-1 yr-1, respectively; and are concentrations for U and Th in ppm; is the porosity of the material. By using the average concentration of each layer for U (middle crust: 0.7 ppm and upper crust: 2.7 ppm) and Th (middle crust: 0.63 ppm and upper crust: 10.5 ppm) [48 and references therein], Hcρc P(He) for each crust layer (in equation (3)) was calculated and were added all up. The [He]F.m for each sample was calculated using Rs (0.02 Ra), Rm (8 Ra), and helium concentration of each sample.

Each variable is measured or obtained except , resulting in that the helium flow rate is a function of porosity (). The porosity of Yeongnam Massif granodiorite underneath the study area is 0.48% [49]. The porosity measured by wireless logging along PX2 at depth is 5.2% [49], which includes the void volume of fractures. The porosity of the fault zone itself should be higher than that of the fractured basement rock. Therefore, we assumed four different porosity conditions with which are about 0.2, 1, 2, and 4 times the porosity of the basement rock.

Calculated helium flow rates (qHe) for a given porosity range from 26.95 cm/yr to 128 cm/yr (Heunghae fault), 5.34 cm/yr to 25.37 cm/yr (Namsong fault), and 4.00 cm/yr to 19.02 cm/yr (Jamyeong fault, Table S2). Compared with other global fault zones (1.7 to 12.7 cm/yr of San Andreas Fault; 0.87 cm/yr of North Anatolian Fault; 55 cm/yr of Alpine Fault, New Zealand) [46, 50], faults in the Pohang region show relatively high helium flow rates. To assume that faults in the Pohang region have helium flows of the approximately same magnitude, the porosity of the faults need to be higher than other fault systems. Thus, in the fault zones of the Pohang region, it is believed that the porosity is high, or the helium flow rate is high. With the calculated helium flow rates and measured helium concentration of each sample, we were able to calculate the 3He flux for each sample site by using measured 3He/4He ratios. The 3He flux per unit area () can be calculated as: where is measured 4He concentration of each sample; is helium isotopic composition (3He/4He). The calculated 3He flux values are 120 to 3,000 atoms cm-2 sec-1 (Heunghae fault), 52 to 1,300 atoms cm-2 sec-1 (Namsong fault), and 83 to 2,100 atoms cm-2 sec-1 (Jamyeong fault, Table S2).

In Figure 6, samples related to the Heunghae fault show the correlation between 3He flux and fault. Considering the size of the study area (<10 km from the fault), this trend can be compared with the results near the Futagawa fault (<40 km from the fault) in the Kumamoto area, Kyushu, Japan, where an earthquake of magnitude 7.3 occurred on April 16, 2016 (Sano et al. [6], Table S3, Figure 7). The Heunghae fault zone shows the sharper pattern than the Futagawa fault zone because samples from this area are collected in a smaller area. The maximum flux of the Futagawa fault is 5,600 atoms cm-2 sec-1, which is higher than that of the Heunghae area because the sample is close to the fault as well as Mt Aso to supply helium from the nearby magma [6].

Like the helium flow rate, this 3He flux is relatively high compared to other major fault systems, such as 1.7 to 34 atoms cm-2 sec-1 of the San Andreas Fault, 75 atoms cm-2 sec-1 of the North Anatolian Fault, and 170 atoms cm-2 sec-1 of the Alpine Fault, New Zealand [46, 50], showing high porosity in the fault zone or high 3He flux per unit area (Figure 7). Also, this 3He flux is several orders of magnitude higher than the continental 3He flux in steady-state (3.9 to 7.2 atoms cm-2 sec-1, from Sano et al. [51]). Therefore, the helium flux results suggest that the Pohang region may have faults comparable to other active fault zones around the world (Figures 1, 6).

5. Conclusions

We first analyzed dissolved gases in groundwater in the Pohang region, South Korea, where the Mw 5.5 earthquake occurred on November 15, 2017. The N2-Ar-He relationship shows that there is the contribution of the mantle component in the Heunghae and Yeonil area samples, which is similar to that previously reported in fault zones of the southeastern Korean peninsula [9]. However, the dissolved gases in the Sinkwang area are mostly close to atmospheric components. N2 (32. 5 to 94.0 vol.%) and CO2 (0.1 to 45.4 vol.%) are present in all areas of the Pohang region, and CH4 (18.7 to 42.9 vol.%) is observed as a major component in some samples of the Heunghae and Yeonil areas. The results of the stable isotope analysis indicate that N2 (), CO2 (), and CH4 () in the Pohang region are derived from organic material sources at shallow depths. Helium isotope ratios (3He/4He) with mantle signatures (up to 3.83 Ra) are observed in the Heunghae and Yeonil areas except in the Sinkwang area, where atmospheric 3He/4He ratios are mainly observed. Helium originates from the mantle, but the Pohang region, a sedimentary basin formed during the Miocene period, is believed to contain a large amount of organic matter that can be the source for N2, CO2, and CH4. The distribution of the helium isotope ratio seems to be related to the locations of faults which are permeable passage. Based on the observation, we suggest that the Heunghae, Namsong, and Jamyeong faults in the Pohang region are active faults that release the mantle fluids. Although the Heunghae and Namsong faults are close to the EGS facilities, considering the depths (<10 km) of the 2017 earthquake and aftershocks in the area, the Moho depth (~28 km) is far below, which is similar to the Gyeongju and Ulsan areas [9]. Thus, we propose that there were already active faults extending into the ductile shear zone to release the mantle helium. In order to show that the faults in this area are active, we computed 3He flux () for the Heunghae (120 to 3,000 atoms cm-2 sec-1), Namsong (52 to 1,300 atoms cm-2 sec-1), and Jamyeong (83 to 2,100 atoms cm-2 sec-1) faults. These values are comparable to those in the regions known as active faults around the world, which may be due to either high porosity or high helium flow rates. Therefore, our results demonstrate that there are active faults in Pohang, especially around the EGS facilities, and will provide important information for future research.

Data Availability

All the data in this study is contained in the tables of both the main manuscript and the Supplementary Materials.

Conflicts of Interest

The authors declare that there is no conflict of interest regarding the publication of this paper.


This study was financially supported by the fundamental research project of Korea Institute of Geoscience and Mineral resources (KIGAM) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2019R1G1A1002297). We are grateful to Myungok Jung and Kwangtae Kim for their support on the sampling.

Supplementary Materials

Figure S1: (a) relative concentration of CO2 and pH plot. (b) CO2/N2 and pH plot. (c) CO2/CH4 and pH plot. (d) CO2/CH4 and concentration of 4He plot. (e) CO2/3He and concentration of 4He plot. (f) CO2/3He and 3He/4He ratio plot. Figure S2: (a) CO2/3He and δ13C plot. Dashed arrow indicates CO2/3He and δ13C decrease trend due to CO2 trap, which described in Section 4.2. (b) δ13C of CO2 and δ13C of CH4 plot (modified from Wfhiticar [34]). Figure S3: helium samples from the Heunghae area (diamond symbol) and the helium isotope and latitude plot. Each symbol is colored by its helium isotope ratio (3He/4He). The locations of Heunghae (HF) and Namsong (NF) fault lines (dashed line) are described in Section 4.4. The gray triangle indicates the location of the EGS site, and the brown star indicates the location of the Mw 5.5 earthquake. Table S1: δ15N, N2/3He, and the contribution of three nitrogen endmembers on the Pohang samples: the mantle, sediment, and the air. The δ15N and N2/3He of each endmember and the mixing model are described in Section 4.2. Table S2: the corrected Helium isotope ratio, the concentration of helium in original mantle fluid, the helium flow rate, and the 3He flux of each fault system. Table S3: the 4He concentration, corrected 3He/4He ratio, and 3He flux of each sample and their distance from related faults, respectively. The data of the Futagawa fault is from Sano et al. [6]. (Supplementary Materials)