Geochemical characteristics of the shallow soil above the Muli gas hydrate reservoir in the permafrost region of the Qilian Mountains, China

Journal of Geochemical Exploration 139 (2014) 160–169

Contents lists available at ScienceDirect

Journal of Geochemical Exploration journal homepage: www.elsevier.com/locate/jgeoexp

Geochemical characteristics of the shallow soil above the Muli gas hydrate reservoir in the permafrost region of the Qilian Mountains, China Zhongjun Sun a,b,⁎, Zhibin Yang a,b, Hai Mei c, Aihua Qin a,b, Fugui Zhang a,b, Yalong Zhou a,b, Shunyao Zhang a,b, Bowen Mei c a Key Laboratory of Geochemical Cycling of Carbon and Mercury in the Earth's Critical Zone, Institute of Geophysical & Geochemical Exploration, Chinese Academy of Geological Sciences, Langfang 065000, China b Geochemical Research Center of Soil Quality, China Geological Survey, China c AE&E Geomicrobial Technologies (Beijing) Inc., Beijing 102200, China

a r t i c l e

i n f o

Article history: Received 2 January 2013 Accepted 10 October 2013 Available online 19 October 2013 Keywords: Gas hydrate Mid-latitude permafrost Qilian Mountains Geochemical indicator Gas hydrate accumulation model

a b s t r a c t In this study, the Muli gas hydrate reservoir in the Qilian Mountains was chosen as a test area for the geochemical exploration of gas hydrates in mid-latitude regions. Soil headspace gases, acid-extracted hydrocarbons and stable carbon isotopes of methane, and soil magnetic susceptibility as well as microbes were tested. The results show that the distribution of geochemical anomalies can be well correlated with the underlying gas hydrate reservoir. Acid-extracted hydrocarbons, soil headspace gases, and the stable carbon isotopes of methane can be considered as major indicators for geochemical exploration of gas hydrates, whereas magnetic susceptibility and microbes served as complementary indicators. The stable carbon isotopes of methane and the hydrocarbon composition of the surface geochemical anomalies indicated a thermogenic origin, which shows that the gas source of the potential gas hydrate reservoir in this area may be contributed by deep oil and coal-formed gases. An accumulation model for the gas hydrate reservoirs was also developed and an integrated exploration project for gas hydrate, oil and coal bed methane is also proposed for the Muli area. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Known natural gas hydrates are mainly distributed in marine sediments and permafrost regions. Gas hydrate reservoirs were discovered for the first time in polar permafrost region in 1967 by a Russian scientist (Sloan, 1990). Later, eight documented discoveries occurred in high-latitude areas, including USA, Russia, Canada and Norway (Collett, 1993; Collett and Dallimore, 1999, 2000; Dallimore and Collett, 1999, 2005; Sloan and Koh, 2008). On November 5th, 2008, with the drilling of Well DK-1, gas hydrates in the mid-latitude region were discovered for the first time in the Qilian Mountains permafrost region in China (Zhang et al., 2007). The gas hydrates in this area mainly comprise methane, with a small amount of ethane, propane and butane, and it appears to be a type II gas hydrate (Zhu et al., 2010). Methods used in gas hydrate exploration in polar permafrost regions include the high-resolution seismic imaging and integrated well logging. These methods appear to be effective in both Alaska and Mackenzie Delta (Brent et al., 2005; Schmitt et al., 2005). However, gas hydrates in the Qilian Mountains mainly occur in fine-grained clastic siltstones, mud stones and dunnet shales of middle Jurassic Jiangchang ⁎ Corresponding author at: Institute of Geophysical and Geochemical Exploration CAGS, Langfang 065000, China. E-mail addresses: [email protected] (Z. Sun), [email protected] (B. Mei). 0375-6742/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gexplo.2013.10.006

formation and appear in the form of “fracture filling” and “pore filling” (Wang et al., 2011). These geophysical premises made the generation of BSR (bottom simulating reflections) difficult. Although scientists have conducted several geochemical survey lines aiming at gas hydrates in the Qinghai–Tibet Plateau (Lu et al., 2010; Wu et al., 2006; Zhang et al., 2008), gas hydrate exploration in the mid-latitude permafrost region in general remains at its early stage. An effective geochemical method for gas hydrate exploration in this region is yet to be developed. A pilot geochemical exploration project in the Qilian Mountains was conducted to examine the effectiveness of geochemical methods for gas hydrate exploration in the mid-latitude region. Soil headspace gases, acid-extracted hydrocarbons, stable carbon isotopes of methane, soil magnetic susceptibility, and microbes were tested in the project. The geochemical characteristics of the gas hydrate area discovered were interpreted, and an effective combination of geochemical indicators was proposed. The gas hydrate accumulation model in the Muli area was also developed for the geochemical exploration of gas hydrates in mid-latitude areas. 2. Geological characteristics of the study area The Qilian Mountains is located in the north of the Qinghai–Tibetan Plateau. The tectonic units of these mountains belong to the North Qilian tectonic belt, the Middle Qilian land block, and the South Qilian tectonic belt. These three tectonic units are divided by four fractures,

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

namely the northern edges of the North Qilian and Middle Qilian fractures, the southern edge of the Middle Qilian fracture and the Tuergendaban Mountains–Zongwunong Mountains–Qinghai Lake fracture. The study area is located in Juhugeng mining area of Muli coalfield in the Qilian Mountains (Fig. 1). The mining area is generally an anticline structure. Sanjingtian, Erjingtian, and Yilutian are distributed in the northern syncline, whereas Sijingtian, Yijingtian, Sanlutian, and Erlutian are distributed in the southern syncline. The gas hydrate reservoir is located in Sanlutian (Fu and Zhou, 1998). The outcropped strata in the mining area are Upper Triassic (T3), Middle Jurassic (J2), Upper Jurassic (J3) and Quaternary (Q) (Fig. 1). The Upper Triassic layer widely occurs in the northern and southern parts of the mining area, where the anticlinal axis occurs. The Upper Triassic mainly consists of black siltstones, mudstones and thin coal seams. The Middle Jurassic contains Muli (J2m) and Jiangchang groups (J2J). The lower Muli group (J2m1) is a braided river alluvial plain deposit. It consists of middle-coarse clastic rocks and occasionally a thin layer of carbonaceous mudstones or thin coals. The upper Muli group (J2m2) is a lake–swamp deposit, and its lithology mainly comprises dark-gray siltstones, fine sandstones, gray-fine-grained sandstones and coarse-grained sandstones, among which two main coal seams are sandwiched. The lower Jiangcang group (J2J1) mainly consists of gray fine-grained sandstones, medium-grained sandstones, dark-gray mudstones, siltstones, in which two to six layers of coal occur. The upper Jiangcang group (J2J2) mainly comprises fine-grained clastic mudstones and siltstones. The Upper Jurassic (J3) is a set of red clastic rocks formed in semi-arid and arid climate. Gas hydrates can be accumulated in three forms: pore filling, fracture filling, and agglomerate forms (Yakushev, 1989). The natural gas hydrates in the Qilian Mountains is hosted in the Jiangcang group of Middle Jurassic. The natural gas hydrates are accumulated in two forms, of which fracture filling form is of a thin-bedded, flaky, lumpy

161

shape and is mainly located in the fractured surface of siltstones, mudstones and oil shales. The pore filling form is of a disseminated shape and is mainly located in the pores of siltstones and sandstones (Lu et al., 2007; Wang et al., 2011). There are three layers of natural gas hydrates in wells DK-1 to DK-3, embedded between 133.0 m and 396.0 m (Zhu et al., 2010). The natural gas hydrates discovered there contains 54% to 76% methane, 8% to 15% ethane, 4% to 21% propane and a limited amount of butane and pentane. Its CO2 content is generally 1% to 7% but could be as high as 15% to 17%. The curve of Raman spectrum that is in the range of 100–4000cm−1 for natural gas hydrates from the permafrost in the Qilian Mountains is similar to that of the Mexico submarine hydrates. Thus the gas hydrates in the Qilian Mountains is considered to be type II gas hydrate (Zhu et al., 2010). Part of the Qilian Mountains is within the permafrost region. This permafrost area covers up to 10 × 104 km2. The annual average ground temperature for the bottom side of the seasonal melting layer is from −1.2 °C to −3.6 °C. The thickness of the permafrost layer is between 50 and 139 m (Zhu et al., 2006). The study area is located between the Tuolainan Mountains and the Datong Mountains, with an altitude of approximately 4000 m to 4300 m. This area located in the central part of the Qilian Mountains permafrost region and is covered with swamp and alpine grassland soils.

3. Sampling and analytical methods 3.1. Sampling The study area covers 50 km2. The sampling interval is generally 500 m × 500 m, while some areas were sampled in a denser grid spaced at 250 m (Fig. 2). Total 229 soil and headspace gas samples as well as 32 microbial soil samples were collected in the area.

Fig. 1. Geological sketch map of coalfield, Juhugeng mining area.

162

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

Fig. 2. Sample location map of geochemical survey.

Soil samples were collected at the depth of 60 cm. Each sample weighed approximately 1 kg. Headspace gas samples were collected using headspace bottles filled with saturated saline solution. The purpose of filling the headspace bottles with saturated saline solution is to reduce the solubility of light hydrocarbon in aqueous solution and force most of light hydrocarbon to be gathered in headspace. Soil samples were placed in bottles with 200 ml saturated saline solution until the total volume of the soil and solution reached to 400 ml. The screws and caps of the bottles were tightened, and the bottles were placed upside down in an indoor environment. Microbial soil samples were collected at the depth of 20 cm below the ground surface. Each sample weighed approximately 300 g and was packed in two layers in the dedicated soil sample bags. All samples were kept frozen on site.

equipped with a flame ionization detector and an HP-AL/S column (50 m × 0.530 mm × 15.0 μm). Values were calculated using an external standard method, and results were reported in μl/kg. 50 g soil samples of particles ≤0.42 mm (with 40 mesh sieve) were extracted by acid (the same method with acidolysis hydrocarbon), and the prolapsed gas was injected in chromatography–isotope mass spectrometry system (GC–IRMS MAT 253,Thermo Fisher Scientific inc., U.S.), after chromatographic separation, CH4 is oxidized to CO2 in the combustion furnace, then the methane carbon isotope values were examined (Li et al., 2008). The carbon isotopic composition is reported in per mil (‰) relative to the PeeDee Belemnite (PDB). The δ13C is calculated using the following equation: 13

3.2. Analytical methods The soil samples collected in the field were dried in a cool and ventilated lab (Debnam, 1969), crushed by hand, sieved through a 0.42 mm pore sieve (40 mesh), then each sample is mixed and split into several piles of 50 g. Then 50 g soil samples were placed in a special apparatus, evacuated to about −0.1 MPa in a 40 °C water bath, then exposed to 1:6 hydrochloric acid (V/V) and decomposed (Tedesco, 1995). The gases released from the samples were absorbed by 300 g/L potassium hydroxide solution to remove CO2, and then the light hydrocarbon components (C1–C5) of the desorbed gases were measured using a 7890A gas chromatography (Agilent Technologies, U.S.) which was equipped with a flame ionization detector and an HP-AL/S column (50 m × 0.530 mm × 15.0 μm). Values were calculated using an external standard method, and results were given in μl/kg. The headspace gas samples collected in the field were placed indoors for 72 h and then the headspace gases were used for the analysis of the light hydrocarbon components (C1–C5) (Hunt and Whelan, 1979) with 7890A gas chromatography (Agilent Technologies, U.S.) which is

δ C¼

n


13


  12 13 12 C= C sample= C= C PDB −1g  1000

ð1Þ

Then 8 g soil samples (weighed by an electronic balance) were placed in sample holders with nonmagnetic plastic sample boxes and tested for its susceptibility value. The real and virtual knobs of the magnetic susceptibility probe were adjusted until the head pointer of the instrument reached the minimum limit. The magnetic susceptibility of the batch of samples was measured at ×1file. The instrument noise was less than 10−7 CGSM, and magnetic susceptibility was about 10−6 SI (Saunders et al., 1991). Methane-oxidizing and butane-oxidizing bacteria have been examined by Geo-Microbial Technologies Inc. in the United States. Soil samples were blended into sterile mineral medium. The soil suspension PH value was then adjusted to 7 while the suspension was being agitated. A serial dilution of 1 to 100 and 1 to 1000 was made. Triplicate cultures were set up by incorporating 1 ml of aliquot portion of the desired dilution suspension into 25 ml agar medium in a Petri dish. Different kinds of alcohol were added for selective cultivation of methane-oxidizing and butane-oxidizing bacteria, respectively. The

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169 Table 1 The detection limit or sensitivity of main methods. Methods

Detection limit or sensitivity

Headspace gas Acid extracted hydrocarbon Microbe Soil magnetism

0.02 × 10−6 0.05 × 10−6 4 cfu/g 10−6 SI

Cfu: Colony-forming units.

dishes were then placed into an incubator at a constant temperature of 37 °C for 6 days. The colonies developed were counted, and the average number was calculated as MV (Microbe Value) for later interpretations. Details of these methods can be found in U.S. patent 3880142 (Hitzman, 1959). Table 1 lists the detection limits (sensitivity) of analytical methods used. Over 10years of experience in oil and gas geochemical exploration have indicated that these methods can meet the technical requirements of gas hydrate geochemical exploration. 4. Results 4.1. Characteristics of soil geochemical indicators As those shown in Table 2, the characteristics of subsurface soil geochemical indicators in the Muli area can be summarized as follows. (1) Acid-extracted hydrocarbon compositions are present in the soils, with a characteristic of C1 N C2 N C3 N C4 N C5. However, the content and standard deviation for methane are relatively high. Headspace gases have a characteristic of C1 N C3 N C2, which differs from that in oil and gas fields. (2) The standard deviation of methane from acid-extracted hydrocarbons and head space gases, and of soil magnetism are greater than 1. The standard deviation for methane from headspace gases is 151.67, indicating that the hydrocarbon gas migration is significantly influenced by the fault structure and soil microbes. (3) The content of methane-oxidizing bacteria is significantly higher than that of butaneoxidizing bacteria, which indicates that the soil gas components in this area are mainly methane, which corresponds with the fact that gas hydrate possesses high methane content. 4.2. Distribution characteristics of geochemical parameters Three anomaly areas were identified in the study area by circling headspace gas methane contour lines of 260, 650 and 1000 μl/L. The western anomaly area is located in the northwest coalfield of the test area. The apical anomaly value is medium, and the area is approximately 2.2 km2. The central anomaly area is located in the junction of northeast blatt flaws and northwest abnormal fault, corresponding to gas hydrates discovered by wells DK-1, DK-2, and DK-3. The intensity of the central anomaly is close to that of the western area. The eastern anomaly area

163

is located in the southeast side of the research area (Erlutian and Sanlutian), in which the highest anomaly value reaches a maximum value of 12,214 μl/L. The area of the eastern anomaly area is more than 3.8 km2 (Fig. 3). Drilling results have proven that the headspace gas anomaly in central anomaly area is caused by the gas hydrates. Three hydrate discovery wells (DK-1, DK-2, and DK-3) are located in the southern high-concentration center of this anomaly area. Three layers of hydrate were found in three drilling wells, respectively, in which the buried depth is from 133.0 m to 396.0 m, and the distribution of hydrates is controlled by the faulted structure (Wang et al., 2011). NE-trending blatt flaws were formed in a later period and disconnected by the NWW regional reverse fault (Fig. 1). The near-surface geochemical anomalies are distributed along the NE-trending structure, indicating that the NE-trending structure controls not only the accumulation of hydrates but also the areal distribution of hydrate geochemical anomalies. The accumulation of methane can be identified the logging data of coal well 5-33 located in the western anomaly area. Drilling results indicate that a gas hydrate reservoir layer may be present in the well. Therefore, the western anomaly area is assumed to be associated with gas hydrate reservoir. The logging data of coal well 7-10 show that its methane content is extremely low, laying in the low value area of headspace gas. The eastern anomaly area has the highest anomaly value and the most developed fault structures. Two NWW-striking regional abnormal faults and an NWW-striking displacement fault pass through this area. The unusual intensity and distribution of the geochemical anomalies can be attributed to the impact of the faults. Hydrate well and coal well results suggest that headspace methane is an effective indicator for finding gas hydrates in mid-latitude region. The headspace gas dry coefficient C1/(C2 + C3) indicates that light hydrocarbon composition is re-differentiated by geological processes. Three anomaly areas were identified in the area (Fig. 4). The locations are similar to the anomaly areas circled by headspace methane. The dry coefficient anomaly of the known gas hydrate deposits is evidently concentrated in a regular shape and at the medium-strength level. This anomaly is controlled by the NE-striking fracture, similar to headspace methane anomaly. The characteristics of the western and eastern anomaly areas are similar to those found in the known gas hydrate deposits. This finding indicates that the light hydrocarbon compositions of the two anomalies are similarly re-differentiated as what happened to gas hydrate deposits. The apical anomaly of the headspace gas dry coefficient that appeared on top of gas hydrate deposits could be explained in two aspects. Firstly, the material exchanges between hydrate deposit and free gas resulted in light hydrocarbon differentiation. Secondly, the upward migration of light hydrocarbon gas through the frozen layer caused the composition differentiation. From the anomaly characteristics observed on top of the known hydrate deposits, the above two factors play significant roles in geochemical differentiation.

Table 2 Eigenvalue of soil geochemical indicators. Methods

Parameters

No. of samples

Maximum value

Minimum value

Mean

Standard deviation

Headspace gas (μl/L)

CH4 C2H6 C3H8 C1/(C2 + C3) CH4 C2H6 C3H8 n-C4H10 i-C4H10 C1/(C2 + C3) κ MMV BMV

213 213 213 213 213 213 213 213 213 213 213 32 32

12,214 2.72 6.66 151,446 1253.66 307.70 193.20 397.10 334.10 43.82 16.13 300 191

1.62 0.00 0.00 2.74 0.49 0.00 0.00 0.00 0.00 2.17 0.57 1 0

60.08 0.04 0.18 31.31 13.20 0.65 0.28 0.09 0.04 12.93 4.63 115.59 57.19

151.67 0.03 0.15 18.92 11.04 0.61 0.23 0.11 0.05 5.48 1.71 103.03 47.65

Acid extracted hydrocarbon (μl/kg)

Soil Magnetism (10−3 SI/kg) Microbe

164

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

Fig. 3. Contour map of headspace methane.

Fig. 4. Contour map of headspace gas aridity coefficient.

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

Therefore, the headspace gas dry coefficient could be used as an indicator of gas hydrate deposits. Three distinct anomaly areas were also identified using the anomaly contour lines of methane of acid-extracted soil hydrocarbon at 47.86, 159.9 and 396.6 μl/kg (Fig. 5). The western anomaly area is located in Yijingtian at the northwest region of the area. This anomaly area is at least of 3.6 km2, and its anomaly intensity is high with a maximum value of 1500 μl/kg. The central anomaly area is related to hydrate discovery wells DK-1, DK-2 and DK-3. This anomaly area has an area of approximately 1.6 km2, and its anomaly intensity is relatively weak. The eastern anomaly area is located in the southeast of the survey area, where the anomaly intensity is significantly lower than that of the western anomaly area. The acid-extracted soil hydrocarbon anomaly contains C1, C2, C3, iC4, nC4, iC5, and nC5 (Fig. 6). This anomaly exhibits the same pattern and reveals three anomaly areas. The area and intensity of the western anomaly area are greater than those of the central and eastern anomaly areas, which differs from the headspace result. The main reason for this difference is that acid-extracted soil hydrocarbon is an accumulative index, whereas western coal mine is an open-pit coal mine, and the coal-formed gases accumulated near the subsurface contributed to the formation of western anomaly area as well. Soil magnetic susceptibility is a non-seismic oil and gas exploration “chimney effect” indicator on top of oil and gas fields. This indicator reflects the secondary substance variations of oil and gas vertical migration (Donovan et al., 1979; Foote, 1992; Machel and Burton, 1991; Saunders et al., 1991). After a thorough study on the central anomaly area of the Muli area, the soil magnetic susceptibility of the known hydrate deposits has been demonstrated as a negative anomaly similar to that of some small oil and gas fields. This phenomenon indicates that the vertical migration of gas hydrate hydrocarbons can also form a “chimney” effect. The “Chimney” effect of gas hydrates is a particular cylindrical object that described the vertical migration

165

process and hydrocarbon-induced alteration of soils and sediments, including the mineral alteration (Pang et al., 2011), physical field variation (Schmitt et al., 2005) and chemical composition anomaly. The key features include: (1) different forms of light hydrocarbon anomaly (acid-extracted hydrocarbon, headspace gas), trace gas anomaly and inert gas anomaly; (2) seismic anomaly (BSR, high velocity zone, etc.) (in both Alaska and Mckenzie Delta); (3) calcium carbonates, pyrites and other secondary mineral anomalies (Pang et al., 2011); (4) microbe community (Han et al., 2011); and (5) magnetic susceptibility anomaly. The negative abnormal distribution range of soil magnetic susceptibility in the western and central survey areas is similar to that of acid-extracted hydrocarbons. The soil magnetic susceptibility anomaly area located in the eastern region is smaller than that of acidextracted soil hydrocarbon and headspace anomalies. The positive anomaly of the magnetic susceptibility in the southern part of the survey area is primarily controlled by a regional fracture (Fig. 7). Acidextracted hydrocarbon and soil magnetic susceptibility anomalies both demonstrate that western area's anomaly is greater in area and intensity. To understand the acid-extracted methane anomaly from the perspective of the magnetic ratio, the hydrocarbon migration intensity in the western side has been proven to be greater than that in the eastern side. This observation could also be verified by the strong methane anomaly discovered inside coal well 5-33. Therefore, more attention could be focused on the western anomaly area. Six microbial survey lines were conducted in the study area (Fig. 2). The main survey line F and auxiliary survey line C intersects at the known Well DK-1, DK-2 and DK-3. On the top of the known gas hydrate deposits, the methane microbial value (MMV) and the butane microbial value (BMV) presented apical anomaly characteristics. MMV is 193, while BMV is 82. MMV is evidently higher than BMV. This result can be correlated with the fact that the gas hydrate in the Muli area mainly contains methane and trace amounts of heavy hydrocarbons. Further

Fig. 5. Contour map of acid extracted methane.

166

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

Fig. 6. Contour map of acid extracted propane.

Fig. 7. Contour map of soil magnetic susceptibility.

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

studies found that MMV and BMV have high anomaly intensity in the northeast region, and their anomaly areas are similar (Figs. 8–9). The abundant development of faults and fractures in the area provides good migration pathways for hydrocarbons. The evolution of hydrocarbon source rocks and the thermal metamorphism of the Jurassic coal layer can generate hydrocarbon gas with methane as the major component. Gas supplies thermogenic C1 to C4 hydrocarbon components to the gas hydrates formed in the permafrost region when it migrates along the fractures and unconformity surface. Survey line C is located in the blatt flaws in the NNE direction. The fracture system can keep the gas hydrate accumulation from moving forward, thus contributing to the differentiation between the western and eastern sides of the survey line. The western side contains more background values, whereas the eastern side has more anomaly values. This finding indicates that the hydrocarbon microseepage in the east is stronger than that in the west, thus further indicating that the eastern region is a better accumulation site for gas hydrate deposits. This conclusion is similar to that of headspace anomalous characteristics. 5. Discussion 5.1. Genetic characteristics of soil geochemical anomalies The origin of gas hydrates in permafrost region can be divided into three major types based on the different formation mechanisms: thermogenic, biogenic and mixed formations. Gas hydrates in Messoyakha, Alaska, and Mackenzie Delta are of thermogenic origin. A few places were reported to have both biogenic and mixed formation characteristics, including Yamburg gas hydrate in west Siberia (Kvenvolden, 1995). Microbial activities may generate methane and induce strong isotope fractionation in methane. Microbial-altered hydrocarbons

167

contain a high concentration of methane and a low δ13C1 value, whereas thermogenic hydrocarbons have a higher δ13C1 value, and its isotopic composition is closer to that of organic matter in the sediments. Thus, the C1/(C2 + C3) value and δ13C1 of methane can be used to determine the origin of the gas in hydrates. Previous studies proved that methane with a C1/(C2 + C3) value greater than 1000 and δ13C1 value lower than −60‰ should be regarded as biogenic formation. Methane with C1/ (C2 + C3) value less than 100 and δ13C1 value higher than −50‰ is of thermogenic origin. Methane with C1/(C2 + C3) value and δ13C1 value between the above two cases is regarded as mixed formation, with both thermogenic and biogenic origin (Kvenvolden, 1995). Two methane samples were collected from the gas hydrate layer of the discovery well DK-1, their δ13C1 values were −39.5‰ and −50.5‰, which shows an apparent thermogenic origin of the source gas (Huang et al., 2011). Nine subsurface soil samples were collected from the three geochemical anomaly areas, and their acid-extracted hydrocarbons were analyzed. The δ13C1 values of the samples are between −34.5‰ and −48.0‰, and their C1/(C2 + C3) values are all less than 35. All samples indicated the thermogenic gas origin (Fig. 10). The consistency between acid-extracted soil hydrocarbons and gas hydrates indicates that the isotopic characteristics of acid-extracted soil hydrocarbons can be used to determine the genetic type of the underlying gas hydrates. A previous study revealed that a huge amount of hydrocarbon gas was released from Jurassic coal layers in the Muli area. This type of gas may be captured and accumulated in the permafrost region and may serve as a source for gas hydrate formation. Several sets of oil source rock, including carbonaceous dark mudstones, lower Permian dark limestones, upper Triassic dark mudstones and Jurassic dark shales, were developed in the Muli depression. All oil source rocks appeared to possess mature or high-over mature characteristics and have high potential to generate oil and gas.

Fig. 8. Content distributions of methane-oxidizing bacteria in the study area.

168

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

Fig. 9. Content distributions of butane-oxidizing bacteria in the study area.

5.2. Accumulation model of gas hydrates in the Qilian Mountains Previous studies on the accumulation model of gas hydrates remained at the theoretical level and mainly focused on marine gas hydrates. The accumulation model of permafrost gas hydrates is poorly understood. Several Russian scientists studied the accumulation model of gas hydrates in polar permafrost region (Yakushev and Chuvilin, 2000). They assumed that hydrocarbon accumulates at the depth of hydrate stabilization layer,

Fig. 10. Plot of aridity coefficient against isotopic composition to distinguish biogenic and thermogenic gas by geochemical abnormal samples in the Muli (Reproduced Claypool and Kvenvolden, 1983).

whereas pores in the region are filled with ice and the pressure within the region increases gradually. When pressure exceeds the thermodynamic critical value of gas hydrate formation, the accumulated gas may be converted into gas hydrates. This finding suggests that gas hydrate and free gas are in a non-equilibrium state, which is consistent with the phenomenon that both gas hydrate and free gas were discovered in the drilling core. An accumulation model based on the geological and geochemical characteristics of the gas hydrates in the permafrost region in the Qilian Mountains is proposed in this study. Hydrocarbons that form gas hydrates may come from an oil gas or a coal gas. A hydrocarbon gas accumulated in gas hydrate stabilization layers, then it went through the processes of geochemical differentiation and crystal fractionation and finally formed gas hydrates in the permafrost region. This formation may occur in the fractures or pores of the reservoir rocks. The gas hydrates formed remains in a non-equilibrium phase and may be destabilized back into a free gas due to constantly changing thermodynamics conditions. Given that the overlying strata comprise a well-developed micro-fracture system, a hydrocarbon gas formed by the destabilization of a gas hydrate may migrate upward to form the gas hydrate chimney. This accumulation model explains the following geological phenomena: (1) The genetic characteristics of gas hydrates in the Muli area can be determined by analyzing either a gas sample from the drilling core or an acid-extracted hydrocarbon sample and its methane isotope signature from a subsurface sample; (2) both hydrocarbon anomalies and the chimney indicator appear in response to the vertical migration of light hydrocarbons in the surface soil; (3) the geochemical anomalies of gas hydrates are more intense in the fractured area; (4) the gas hydrate anomalous area is consistent with the areal distribution of the gas hydrate reservoir; and (5) subsurface geochemical anomalies could reflect the compositional characteristics of the underlying gas hydrates.

Z. Sun et al. / Journal of Geochemical Exploration 139 (2014) 160–169

6. Conclusions 1. The geochemical survey of gas hydrates in the permafrost region of the Qilian Mountains revealed that geochemical anomalies exist in the surface area of gas hydrate reservoir in the mid-latitude region. The anomaly data may be used either to determine the genetic type of gas hydrate or to describe the areal distribution of gas hydrates. 2. Evident geochemical anomalies were discovered on top of the gas hydrate deposits in the Muli area of the Qilian Mountains. Among the geochemical results, acid-extracted hydrocarbons and headspace gases appear to be apical anomalies in response to the underlying gas hydrates, whereas magnetic susceptibility seems to form negative anomalies. Integrated with other research findings in the midlatitude region, the optimized geochemical parameters for gas hydrate exploration in the current study are acid-extracted hydrocarbons, soil headspace gases and stable carbon isotopes of methane. Magnetic susceptibility and light hydrocarbon-oxidizing bacteria values serve as secondary indicators. 3. An accumulation model based on the integrated interpretation was proposed in this study. 4. The aridity coefficient and isotopic composition indicate that the source gases of gas hydrates in the Muli area include petroleum and coal gases. This finding shows that an integrated exploration of gas hydrates, oils and coal beds should be conducted in future survey projects. Acknowledgments This work is jointly financed by Public-good Research Fund (201111019) , the Geological Survey and Mineral Resources Assessment Project (1212011120974) and National Special Foundation for Gas Hydrate Exploration and Test-production (GZHL20110303). We also thank three anonymous reviewers and Prof. Changjiang Li (guest editor) for their comments and suggestions. References Brent, T.A., Riedel, M., Caddel, M., Clement, M., Collett, T.S., Dallimore, S.R., 2005. Initial geophysical and geological assessment of an industry 3-D seismic survey covering the JAPEX/JNOC/GSC et al. Mallik 5L-38 gas hydrate production research well. In: Dallimore, S.R., Collett, T.S. (Eds.), Scientific Results from the Mallik 2002 Gas Hydrate Production Research Well Program, Mackenzie Delta, Northwest Territories, Canada. Geological Survey of Canada Bulletin, 585, pp. 1–14. Claypool, G.E., Kvenvolden, K.A., 1983. Methane and other hydrocarbon gases in marine sediment. Annu. Rev. Earth Planet. Sci. 11, 299. Collett, T.S., 1993. Permafrost-associated gas hydrate accumulations. In: Sloan, E.D., Happer, J., Hnatow, M. (Eds.), International Conference on Natural Gas Hydrates. Annals of the New York Academy of Science, 715, pp. 247–269. Collett, T.S., Dallimore, S.R., 1999. Hydrocarbon gases associated with permafrost in the Mackenzie Delta, Northwest territories, Canada. Appl. Geochem. 14, 607–620. Collett, T.S., Dallimore, S.R., 2000. Permafrost-related natural gas hydrate. In: Max, M.D. (Ed.), Natural Gas Hydrate in Oceanic and Permafrost Environments. Kluwer Academic Publishers, The Netherlands, pp. 43–60. Dallimore, S.R., Collett, T.S., 1999. Regional gas hydrate occurrences, permafrost conditions, and Cenozoic geology, Mackenzie Delta area. In: Dallimore, S.R., Uchida, T., Collett, T.S. (Eds.), JAPEX/JNOC/GSC Mallik 2L-38 Gas Hydrate Research Well, Mackenzie Delta, Northwest Territories, Canada, Bulletin, 544, pp. 31–43. Dallimore, S.R., Collett, T.S., 2005. Scientific results from the Mallik 2002 gas hydrate production research well program, Mackenize Delta, Northwest Territories, Canada. Geol. Surv. Can. Bull. 585, 1–140.

169

Debnam, A.H., 1969. Geochemical prospecting for petroleum and natural gas in Canada. Geol. Surv. Can. Bull. 177. Donovan, T.J., Forgey, R.J., Roberts, A.A., 1979. Aeromagnetic detection of diagenetic magnetite over oil fields. AAPG Bull. 63, 245–248. Foote, R.S., 1992. Use of magnetic field aids oil search. Oil Gas J. 4, 137–142. Fu, J., Zhou, L., 1998. Carboniferous–Jurassic stratigraphic provinces of the southern Qilian basin and their petro-geological features. Northwest Geosci. 19 (2), 47–54 (in Chinese). Han, L., Wu, S., Li, J., Lü, J., Zhu, Y., 2011. Microbial community in DK-2 gas hydrate borehole, Qilian Mountain permafrost. Geol. Bull. China 30 (12), 1874–1882 (in Chinese with English abstract). Hitzman, D.O., 1959. Prospecting for petroleum deposits (detecting hydrocarbon consuming bacteria colonies by artificial hydrocarbon nutrient culturing). U.S. Patent, 3,880,142, assigned to Phillips Petroleum Co. Huang, X., Zhu, Y., Wang, P., Guo, X., 2011. Hydrocarbon gas composition and origin of core gas from the gas hydrate reservoir in Qilian Mountain permafrost. Geol. Bull. China 30 (12), 1851–1856 (in Chinese with English abstract). Hunt, J.M., Whelan, J.K., 1979. Volatile organic compounds in Quaternary sediments. Org. Geochem. 1, 219–224. Kvenvolden, K.A., 1995. A review of the geochemistry of methane in natural gas hydrate. Org. Geochem. 23 (11/12), 997–1008. Li, Z., Li, X., Ma, C., Li, J., Zhang, Z., Gao, X., Zhang, W., Xu, C., 2008. China Oil and Gas Industry Standard (SY/T 5238-2008)-Analysis Method for Carbon and Oxygen Isotopes in Organic Matter and Carbonate. Petroleum Industry Press, Beijing 1–7 (in Chinese). Lu, Z., Wu, B., Rao, Z., 2007. Geological and geochemical anomalies of gas hydrate in permafrost zones along the Qinghai–Tibet Railway, China. Geol. Bull. China 26 (8), 1029–1040 (in Chinese with English abstract). Lu, Z., Zhu, Y., Zhang, Y., Wen, H., Li, Y., Jia, Z., Liu, C., Wang, P., Li, Q., 2010. Basic geological characteristics of gas hydrates in Qilian Mountain permafrost area, Qinghai province. Mineral Deposits 29 (1), 182–191 (in Chinese with English abstract). Machel, H.G., Burton, E.A., 1991. Chemical and microbial processes causing anomalous magnetization in environments affected by hydrocarbon seepage. Geophysics 56, 598–605. Pang, S., Su, X., Yang, X., Wang, P., Li, Y., Guo, X., Li, Q., 2011. Sedimentological features of Middle Jurassic strata revealed by scientific drilling boreholes of natural gas hydrate in Qilian Mountain permafrost. Geol. Bull. China 30 (12), 1829–1838 (in Chinese with English abstract). Saunders, D.F., Burson, K.R., Thompson, C.K., 1991. Observed relation of soil magnetic susceptibility and soil gas hydrocarbon analyses to subsurface hydrocarbon accumulations. AAPG Bull. 75, 389–408. Schmitt, D.R., Welz, M., Rokosh, C.D., 2005. High-resolution seismic imaging over thick permafrost at the 2002 Mallik drill site. Geol. Surv. Can. Bull. 585, 1–13. Sloan, E.D., 1990. Clathrate Hydrates of Natural Gases. Marcel Dekker, Inc., New York. Sloan, E.D., Koh, C.A., 2008. Clathrate Hydrates of Natural Gases, third ed. CRC Press, New York, Taylor and Francis Group, Publishers, pp. 554–555. Tedesco, S.A., 1995. Surface Geochemistry in Petroleum Exploration. Chapman & Hall, New York 43–44. Wang, P., Zhu, Y., Lu, Z., Guo, X., Huang, X., 2011. Gas hydrate in the Qilian Mountain permafrost and its distribution characteristics. Geol. Bull. China 30 (12), 1839–1850 (in Chinese with English abstract). Wu, Z., Lu, X., Wang, Z., 2006. Formation and geochemical exploration model of gas hydrate in the perennial frozen soil area, Qinghai–Tibet plateau. Geol. Sci. Technol. Inf. 25 (4), 9–14 (in Chinese). Yakushev, S.E.M., 1989. Gas hydrate in cryolithoregion. Sov. Geol. Geophy. 11, 100–105 (Russia). Yakushev, V.S., Chuvilin, E.M., 2000. Natural gas and gas hydrate accumulations within permafrost in Russia. Cold Reg. Sci. Technol. 31, 189–197. Zhang, H., Zhang, H., Zhu, Y., 2007. Gas hydrate investigation and research in China: present status and progress. Geol. China 34 (6), 953–961 (in Chinese with English abstract). Zhang, Z., Zhu, Y., Su, X., 2008. Characteristics of thermoluminescence from sediment and its implication in the Qiangtang Basin, Qinghai–Tibet plateau. Geoscience 22 (3), 452–456 (in Chinese with English abstract). Zhu, Y., Liu, Y., Zhang, Y., 2006. Formation conditions of gas hydrates in permafrost of the Qilian Mountains, Northwest China. Geol. Bull. China 25 (1–2), 58–63 (in Chinese with English abstract). Zhu, Y., Zhang, Y., Wen, H., Liu, Z., Jia, Z., Li, Y., Li, Q., Liu, C., Wang, P., Guo, X., 2010. Gas hydrates in the Qilian Mountain Permafrost, Qinghai, Northwest China. Acta Geol. Sin. (Engl. Ed.) 84 (1), 1–10.