Geochemical characteristics and origin of natural gas from Wufeng-Longmaxi shales of the Fuling gas field, Sichuan Basin (China)

Accepted Manuscript Geochemical characteristics and origin of natural gas from Wufeng-Longmaxi shales of the Fuling gas field, Sichuan Basin (China)

Rui Yang, Sheng He, Qinhong Hu, Dongfeng Hu, Jizheng Yi PII: DOI: Reference:

S0166-5162(16)30465-7 doi: 10.1016/j.coal.2016.12.003 COGEL 2762

To appear in:

International Journal of Coal Geology

Received date: Revised date: Accepted date:

12 August 2016 16 December 2016 16 December 2016

Please cite this article as: Rui Yang, Sheng He, Qinhong Hu, Dongfeng Hu, Jizheng Yi , Geochemical characteristics and origin of natural gas from Wufeng-Longmaxi shales of the Fuling gas field, Sichuan Basin (China). The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Cogel(2016), doi: 10.1016/j.coal.2016.12.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Geochemical characteristics and origin of natural gas from Wufeng-Longmaxi shales of the Fuling gas field, Sichuan Basin (China) Rui Yang a, b, Sheng He a, Qinhong Hu b, Dongfeng Hu c, Jizheng Yi d Key Laboratory of Tectonics and Petroleum Resources, Ministry of Education, China University

PT

a

of Geosciences, Wuhan 430074, China.

Department of Earth and Environmental Sciences, the University of Texas at Arlington,

Arlington, TX 76019, USA.

RI

b

Exploration Company, Sinopec, Chengdu 610064, China.

d

Petroleum Exploration and Development, Jianghan Oilfield Branch Company, Sinopec, Wuhan

SC

c

NU

430223, China.

Abstract

MA

As the first giant shale gas field in China, the Fuling gas field has recently been regarded as one of the most important regions for natural gas exploration and

D

production in the Sichuan Basin. However, the origin of natural gas from Upper

PT E

Ordovician Wufeng and the bottom of Lower Silurian Longmaxi (WL) shales in the Fuling gas field is poorly understood to limit a comprehensive understanding of gas generation, accumulation and exploration.

CE

In this work, based on molecular and stable carbon isotopic composition of a total of 24 gas samples from five shale gas wells in the Fuling field, we analyzed the

AC

geochemical characteristics and gas origin, and discussed the cause for the geochemical anomalies (carbon isotopic reversals). Molecular composition results show that gases from the Fuling gas field are dry and mainly composed of methane (97.9 - 98.9 %), with a very low level of ethane (C2H6), propane (C3H8) and non-hydrocarbon gases (mainly CO2 and N2). These dry gases are classified as oil-associated gas and mainly derived from secondary cracking. Due to the lack of gas samples across a maturation gradient from immature to late mature, the WL gases in the Fuling field show an unclear evolution trend between the δ13C2 and wetness values; however, all these samples are located in the isotopic reversal zone. Carbon isotopes

ACCEPTED MANUSCRIPT of gaseous alkanes clearly display full isotopic reversals (δ13C1 > δ13C2 > δ13C3), which is indicative of a relatively high thermal maturity and consistent with the measured vitrinite reflectance and modeled values (~ 3.0 % Ro). The observed complete carbon isotopic reversals in the WL gases are caused by a combination of several mechanisms, in which isotope exchange at high temperature is the primary

PT

controlling factor. Other secondary factors include Rayleigh-type fractionation of C2H6 and C3H8, secondary cracking and gas diffusion mixing of gases at different

RI

thermal maturity levels.

Key words: Sichuan Basin; Wufeng and Longmaxi shales; Gas origin; Gas

NU

1. Introduction

SC

components; Carbon isotopic reversal

As a benefit from advanced development of innovative technologies (e.g.

MA

horizontal drilling and hydraulic fracturing technologies), gas production from black shale formations has been significantly increased in the United States (US Energy

D

Information Administration, 2015), and unconventional resources has gradually

PT E

become the focus of hydrocarbon exploration around the world (Chalmers and Bustin, 2007; Chen et al., 2011; Curtis, 2002; Guo, 2015; Klaver et al., 2012; Mathia et al., 2016). The recent accelerated exploration of shale gas also provides an opportunity to

CE

understand the geochemical characterization of gases by organic and gas isotopic geochemistry, especially stable carbon isotopic composition of alkanes, which are an

AC

effective tool to investigating the genetic characteristics of natural gases (Dai et al., 2004; 2010; Hao and Zou, 2013; Schoell, 1980; Tang and Jenden, 1995; Whiticar, 1990).

Studies of gas composition and isotopic geochemistry from shale formation in typical gas-producing sedimentary basins have been recently reported in the literature (Dai et al., 2014b; Hao and Zou, 2013; Rodriguez and Philp, 2010; Strąpoć et al., 2010; Tilley et al., 2011; Wei et al., 2016; Zumberge et al., 2012; Zumberge et al., 2009). Methane is the main molecular composition of natural gas in typical US and Chinese shale gas plays, followed by a very low content of ethane and propane, as

ACCEPTED MANUSCRIPT well as little non-hydrocarbon gases (Burruss and Laughrey, 2010; Cao et al., 2015; Zumberge et al., 2012). The molecular composition and content of the alkane gases are often closely related to the thermal maturity (Dai et al., 2014b; Zumberge et al., 2012). Especially, when the shale formation has an extremely high thermal maturity, the content of methane can exceed 98% as the dry gas (Cao et al., 2015; Dai et al.,

PT

2014b; Wei et al., 2016; Zumberge et al., 2012). Generally, the stable carbon isotopic composition will become more positive

RI

with an increasing carbon number from methane δ13C (δ13C1), ethane δ13C (δ13C2), propane δ13C (δ13C3) to butane δ13C (δ13C4) at an identical maturity, which is often

SC

called as positive carbon isotope series (i.e., δ13C1 < δ13C2 < δ13C3 < δ13C4). However, an abnormal sequence of carbon isotopic compositions, such as full (δ13C1 >δ13C2 >

NU

δ13C3 > δ13C4) or partial (e.g. δ13C1 < δ13C2 > δ13C3 < δ13C4, or δ13C1 > δ13C2 < δ13C3 < δ13C4) isotopic reversals, have also been observed in some shale gas plays and

MA

conventional gas reservoirs (Dai et al., 2004; Dai et al., 2014b; Wei et al., 2016; Xia, 2014; Xia et al., 2013; Zumberge et al., 2012). Even though a full or partial carbon

D

isotopic reversal pattern has been reported in some shale gas plays with high maturity

PT E

values, no consensus has been reached regarding this enigmatic phenomenon (Dai et al., 2016; Hao and Zou, 2013; Jenden et al., 1993; Rodriguez and Philp, 2010). More importantly, a close relationship between gases with an abnormal carbon isotopic

CE

reversal pattern with high production rates was observed in shale gas fields, such as the Barnett Shale in the Fort Worth Basin (Gao et al., 2014; Hao and Zou, 2013; Slatt

AC

and Rodriguez, 2012; Tilley and Muehlenbachs, 2013; Zumberge et al., 2012); these authors suggest that a carbon isotopic reversal of the gas could be an effective indicator for a favorable production area to some extent. Similar phenomenon was first observed in other areas to draw attention, partially due to its probable relation to gas productivity (Brown, 2010; Gao et al., 2014; Tilley and Muehlenbachs, 2013; Zumberge et al., 2012). Therefore, the cause of carbon isotopic reversal, and its implication for shale gas exploration and production, is an important issue to be studied for the origin of gas and the sustainable shale resource development. Completed with horizontal drilling and multi-stage hydraulic fracturing

ACCEPTED MANUSCRIPT technologies, a large amount of natural gases has been produced from many wells within the Upper Ordovician Wufeng and the bottom of Lower Silurian Longmaxi shales (termed as WL shales in this work) in the Fuling gas field, the first giant shale gas field in China (Guo, 2013; Guo, 2015; Guo and Zhang, 2014; Yang et al., 2016d). It is reported that the annual shale gas production in the Fuling gas field has reached

PT

12×108 m3 (or 0.043 tcf) in 2014, and is expected to reach 3.2×109 m3 (or 0.113 tcf) in 2015 (Dai et al., 2016; Guo, 2015; Yang et al., 2016a). Carbon isotopic reversal of

RI

gases from Wufeng and Longmaxi shales was also reported in the work of Guo (2015), Guo and Zhang (2014) and Liu (2015); however, no consensus has been reached and

SC

different causes were proposed to explain this abnormal phenomenon. Even though large amounts of commercial shale gas has been produced from WL shales in the

NU

Fuling gas field, detailed studies about the geochemical characteristics and origin of natural gas from these black shales are poorly documented. In this work, we modeled

MA

the burial and thermal history of WL shales, studied the geochemical characteristics, determined the gas origin, and discussed the cause of carbon isotopic reversal of gas

PT E

2. Geological setting

D

samples from WL shales in the Fuling gas field.

Discovered in 2012, the Fuling gas field is located in the Fuling District of Chongqing City, Southeast Sichuan Basin, which is one of the most stable and

CE

gas-productive sedimentary basins in China (Fig. 1a). Since the development of basement (Yangtze craton), the Fuling area has experienced five tectonic cycles, Caledonian

AC

including

Indosinian

(Triassic),

(late

Sinian-Silurian),

Yanshanian

Hercynian

(Jurassic-Cretaceous),

(Devonian-Permian), and

Himalayan

(Tertiary-Quaternary) (Li et al., 2015; Tong, 1992; Yang et al., 2016b; Yang et al., 2016c; Yang et al., 2016d). The Caledonian, Hercynian, Yanshanian and Himalayan movements caused uplift and erosion, resulting in the relatively complex geological settings (Tong, 1992). Overall, our study area is a gentle anticline and relatively stable with few faults in the middle part (Figs. 1b-c). Fig. 1c is a southeast cross-section of Well A (Line A-A‘ in Fig. 1b), showing that the stratigraphic intervals are uniformly distributed in the middle part and some reverse faults with large dip angles are well

ACCEPTED MANUSCRIPT developed at the wings of the anticline. Affected by the Caledonian movement, Devonian sediments are eroded and most of the Carboniferous sediments are absent with only a thin set of Huanglong (C2h) Formation remaining in the Fuling area. Three large-scale tectonic uplifts occurred since Yanshanian and Himalayan movements, resulting in the absence of Cretaceous,

PT

Paleogene and Neogene sediments (Figs. 2-3). Due to the confined paleogeography and rapid rise of global sea level, the Upper Ordovician Wufeng and the bottom part

RI

of Lower Silurian Longmaxi shale formations (marked as WL shales in Fig. 2) were successively deposited in a deep shelf environment. Several lines of evidence from

SC

the sedimentary record originally suggested the WL shales were deposited in an anoxic depositional environment, and evolved to a dysoxic-oxic environment for the

NU

middle and upper units in the Longmaxi Formation (Guo and Zhang, 2014; Yang et al., 2016c). The WL shales are underlain conformably by Upper Ordovician Jiancaogou

MA

Limestone and are lithostratigraphically dominated by black carbonaceous shale, siliceous shale and carbonaceous mudstone with a large amount of fossils, such as

D

radiolarian and graptolite (Yang et al., 2016c). In contrast, the middle and top part of

PT E

Longmaxi Formation deposited in slope facies and shallow-water continental shelf, respectively, and their lithologies mainly consist of gray siltstone, dark gray argillaceous siltstone and gray mudstone with fewer fossils. The total thickness and

CE

burial depth of the WL shales in the Fuling area ranges from about 60 m to 150 m, and from less than 2000 m in northeast region to about 3000 m in the southwest (Figs.

AC

1a-b; Guo, 2015).

It is reported that the WL shales are characterized by appropriate burial depth, large thickness, high total organic carbon (TOC) content (averaged 3.56 %), high thermal maturity (Ro >2.5 % to be in dry gas window), appropriate rock brittleness index [defined as 100 % × (quartz + feldspar + carbonate) / (quartz + feldspar + carbonate + clay); 55 - 65 %], desirable over-pressure, high gas content (averaged 4.3 m3/t) and strong gas generation capacity (Dai et al., 2016; Guo, 2013; 2015; Guo and Zhang, 2014; Liu, 2015; Yang et al., 2016d). Generally, δ13C of kerogen is less intensively affected by thermal maturation and can be used to classify organic matter

ACCEPTED MANUSCRIPT types (Hao and Chen, 1992; Hao et al., 2013). According to the work of Guo (2013) and Guo and Liu (2013), the kerogen δ13C values from the WL shales are between -29.2 ‰ and -30.5 ‰, suggesting that the WL shales are dominated by type Ⅰ with some type Ⅱ. This is also supported by the kerogen maceral analysis that kerogen from the WL shales is dominated by sapropel components with a content from 92.8 % to 100 % (Guo and Liu, 2013). Measured porosities range from 1.17 % to 7.22 %,

PT

with an averaged value of 4.52 %, and matrix permeability values are from 1.1×10-6

RI

μm2 to 0.1×10-3 μm2; both are comparative with those of Barnett Shale in U.S. (Chalmers et al., 2012; Gao and Hu, 2016; Guo, 2015; Loucks et al., 2009; Pan et al.,

favorable for gas generation and exploration.

NU

3. Samples and methods

SC

2015). Overall, these geological characteristics indicate that the WL shales are quite

Using stainless steel cylinders with two valves, a total of 24 gas samples from

MA

five shale gas wells in the Fuling gas field were collected, and analyzed for molecular composition and stable carbon isotope at Petroleum Exploration and Development in

D

Jianghan Oilfield Branch Company of Sinopec. In addition, geochemical data of other

PT E

typical gas fields reported in the literature are summarized to compare with our results.

Using BasinMod 1-D software from Platte River Associates, Inc., Well A was

CE

selected to model the burial and thermal history of the WL shales. The key input data, e.g., stratigraphic information, present-day, geothermal data, vitrinite reflectance,

AC

measured temperature, were used in the modeling. Our previous study on thermal maturity of bitumen reflectance shows that the WL shales in the Fuling field have equivalent vitrinite reflectance values from 2.87 % to 3.66 % with an average of 3.21 %, suggesting that the WL shales are over-mature (Yang et al., 2016d). Because there is not enough vitrinite to measure a representative number of vitrinite reflectance, the equivalent vitrinite reflectance values can generally reflect the high maturities of the WL shales. In this study, the modeling process was calibrated by the measured temperature and equivalent vitrinite reflectance values. Paleo-heat flow values are quoted from the work of Cao et al. (2015) and Zhu et al. (2013), and the

ACCEPTED MANUSCRIPT present-day heat-flow was obtained by transient heat flow model. Maturity was calculated based on Easy% Ro method proposed by Sweeney and Burnham (1990). The kerogen type and TOC content were adopted according to actual geochemical results of the WL shales. Following the Chinese Oil and Gas Industry Standard (GB/T) 13610-2003, the

PT

gas composition was determined using an Agilent 6890N gas chromatograph (GC) with helium as a carrier gas, which is equipped with a flame ionization detector (FID)

RI

and a thermal conductivity detector,. The GC oven temperature was initially set at 50℃ for 5 minutes, and rise to 105℃ at a rate of 10℃ per minute for 3 minutes, and

SC

finally ramped to 180℃ at a rate of 10℃ per minute for 10 minutes. The relative content for a certain gas component was determined by the ratio of the corresponding

NU

peak area value to integral value of the all peak areas, and the absolute contents for the gas compositions were corrected for oxygen and nitrogen in the atmosphere. In

MA

addition, wetness, defined as Σ (C2-C5) / Σ (C1-C5) × 100 %, was used to evaluate the variation of gas composition.

D

The stable carbon isotopic values of gas were determined using a Thermo

PT E

Scientific Delta V mass spectrometer interfaced with a Trace GC Ultra gas chromatograph, following the Chinese Oil and Gas Industry Standard (GB/T) 18340.2-2010. The hydrocarbon gas components (C1-C4) and carbon dioxide (CO2)

CE

were first separated using a fused silica capillary column on a GC, then converted to CO2 in a combustion interface, and injected into the mass spectrometer for

AC

quantification. The temperature of the GC oven was ramped from 35℃ to 80℃ at 8℃ per minute, then increased to 260℃ at 5℃ per minute and maintained for 10 minutes. Helium was used as the carrier gas. Gas samples were analyzed in triplicate, and the stable carbon isotopic values are expressed in the δ notation as per mil (‰) relative to the Vienna Pee Dee Belemnite (VPDB) standard. Two-point calibrations were conducted with international measurements standards for carbon isotope ratios, and the repeated analytical precision is of ± 0.5 ‰ for individual hydrocarbon gas compounds of the VPDB standard.

ACCEPTED MANUSCRIPT 4. Results 4.1 Burial and thermal history of the WL shales The modeling results of burial and thermal history for WL shales in Well A are shown in Figs. 3a-b. Overall, the measured temperature and equivalent vitrinite reflectance values show an excellent match with the modeled values (Fig. 3b). The

PT

WL shales experienced a long-term subsidence before they reached their maximum burial depths, as well as two minor uplifts during late Caledonian (late Silurian) and

RI

middle Hercynian movements (late Carboniferous). The WL shales began to generate oil in late Silurian (Ro 0.5-0.7 %) and with generation reaching a peak between late

SC

Permian and middle Triassic (250 ~ 230 Ma; Ro 0.7-1.0 %). Main gas generation began from the early Jurassic (~ 190 Ma) to the late Cretaceous (~ 100 Ma). In

NU

particular, the bottom of the WL shales reached 150℃ and 190℃ at about 198 Ma and 150 Ma, respectively (Fig. 3a), and increased to a maximum burial depth of

MA

approximately 6000 m with a temperature of ~ 210℃ (in dry gas window with corresponding Ro % of around 3.0 %) in late Yanshanian (~ 80 Ma). After the WL

D

shales reached the maximum burial depth, they undergone three tectonic uplift and

PT E

gradually elevated to the present depth of 2500-3000 m with a present-day temperature of about 80℃. The long and complex thermal and burial histories of the WL shales are similar with that of the Barnett and New Albany shales in U.S.A.

CE

(Curtis, 2002; Strąpoć et al., 2010). Our modeling results from the WL shales, as well as the Barnett and New Albany shales from the North American in the literature,

AC

confirm that gas-producing shales often have first experienced deep burial with an increasing temperature and pressure. Then these mature or high-maturity shales further undergone intensive uplift and erosion of overburden rocks, forming the suitable present-day depths for commercial gas production (Fig. 3a; Curtis, 2002; Strąpoć et al., 2010). 4.2 Chemical composition The natural gases from the Fuling gas field are mainly composed of methane (CH4), followed by ethane (C2H6) and propane (C3H8), and a very low content of non-hydrocarbon gases (mainly CO2 and N2). No hydrogen sulfide gas is detected.

ACCEPTED MANUSCRIPT The methane contents range from 97.9 % to 98.9 %, with an average of 98.4 % (Table 1). The high contents of methane from the Fuling gas field are similar to other gas fields in the Sichuan Basin (Cao et al., 2015; Dai et al., 2016; Liao et al., 2014; Wang et al., 2016), and are also comparable to the typical shale (Barnett, Fayetteville, New Albany, Marcellus and Utica) plays in U.S. (Burruss and Laughrey, 2010; Jenden et

PT

al., 1993; Rodriguez and Philp, 2010; Tilley and Muehlenbachs, 2013; Zumberge et al., 2012). Ethane contents vary from 0.4 % to 0.9 %, and propane contents are

RI

generally less than 0.1 %. The contents of butane are below the detection in most of these gas samples. Wetness values for these gas samples are small with a range from

SC

0.43 % to 0.97 %, suggesting typical dry gases. The contents of CO2 from the Fuling gas field are low (0.1 - 0.5 %; Table 1), and are comparable to the Longmaxi shale gas

NU

in the Weiyuan (0.02 - 1.1 %; Dai et al., 2016; Wang et al., 2016) and Changning fields (0 - 0.9 %; Dai et al., 2016; Wang et al., 2016). N2 contents in the Fuling area

MA

are slightly higher (0.5 - 1.0 %) than the Longmaxi shale gas from the Changning gas field (0.03 - 0.4%), but lower than those results from the Weiyuan gas field (0.01 -

D

3.0 %; Dai et al., 2016; Gao, 2015; Wang et al., 2016).

PT E

4.3 Carbon isotopes of gaseous alkanes The results of stable carbon isotopic composition of gaseous alkanes are presented in Table 1. The δ13C1 values of these gas samples range from -30.7 ‰ to

CE

-28.4 ‰, with an averaged value of -29.8 ‰, which are consistent with the work of Wang et al. (2016) and Wei et al. (2016). Our data show that the δ13C1 values are

AC

clearly heavier than other typical shale gas fields, such as the gases from Longmaxi shale in the Weiyuan field (-36.8 ‰ ~ -34.5 ‰; Cao et al., 2015; Dai et al., 2016; Wang et al., 2016), Barnett (Ro = 1.3-2.1 %) in the Fort Worth Basin (-49.7 ‰ ~ -35.7 ‰; Rodriguez and Philp, 2010; Zumberge et al., 2012), New Albany (Ro = 0.5-1.5 %) in the Illinois Basin (-56.3 ‰ ~ -52.1 ‰; Strąpoć et al., 2010) and Fayetteville (Ro = 2.5-3.0 %) in the Arkoma Basin (-41.9 ‰ ~ -35.4 ‰; Zumberge et al., 2012), but lighter than Longmaxi gases in the Changning field (-31.3 ‰ ~ -26.7 ‰; Cao et al., 2015; Dai et al., 2016; Dai et al., 2014b; Wang et al., 2016). The values of δ13C2 in the Fuling gas field range from -35.7 ‰ to -34.1 ‰, with

ACCEPTED MANUSCRIPT an average of -34.6 ‰ (Table 1). The δ13C3 values are much lighter (between -38.7 ‰ and -35.0 ‰) than δ13C1 and δ13C2 values, with an averaged value of -36.7 ‰. Compared with the reported δ13C3 values of Longmaxi shale gases in the work of Wang et al. (2016), our results are much heavier than some of their results (Fig. 4). Obvious differences in the δ13C3 values were observed between our WL shale gas

PT

samples in this study with previous samples from the same Fuling gas field; unfortunately, currently we are unable to completely explain this phenomenon. Even

RI

so, the carbon isotopes of gaseous alkanes from all the WL shales in the Fuling gas field display obviously full carbon isotopic reversals, as well as the shale gases from

SC

the Changning and Weiyuan gas fields (Fig. 4, δ13C1 > δ13C2 > δ13C3). Fig. 5 shows the plot of δ13C1 versus δ13C2 for shale gases from typical

NU

shale-producing basins in both China and North American. The dashed line in Fig. 5 represents that the δ13C1 values are equal to δ13C2 values; the area above this dashed

MA

line, characterized by δ13C1 < δ13C2, means that the gas samples are in a thermally low mature / mature stage; while the area under the dashed line suggests a thermally

D

highly mature / over-mature stage which are characterized by δ13C1 > δ13C2. It is

PT E

obviously that gas samples from the Barnett Shale in the Fort Worth Basin, Yanchang Formation in the Ordos Basin and New Albany in the Illinois Basin are mainly located in the non-reversed zone, while gases from the WL shales and Fayetteville in the

CE

Arkoma Basin have plotted in the reversed zone, suggesting that the WL shale gases in the Fuling gas field have the highest thermal maturity values. In addition, a

AC

n-shaped relationship is observed with an evolutionary path from thermally low mature stage, mature stage, highly mature and finally to over-mature stage (WL shale gases). This is consistent with the measured vitrinite reflectance values (Burruss and Laughrey, 2010; Dai et al., 2016; Guo, 2015; Jenden et al., 1993; Rodriguez and Philp, 2010; Tilley and Muehlenbachs, 2013; Yang et al., 2016a; Zumberge et al., 2012). 5. Discussion 5.1 Genetic characterization of shale gases from the WL shales As a new type of unconventional natural gas, the origin of shale gas is important to understand the mechanisms for shale gas enrichment and storage, which has

ACCEPTED MANUSCRIPT attracted an extensive attention in different shale gas-producing basins (Hao et al., 2013; Liu et al., 2012; Schoell, 1983; Tilley and Muehlenbachs, 2013). Shale formation is a self-sourcing hydrocarbon reservoir (Curtis, 2002), and shale gas generally represents the gas molecular composition at the location of gas generation (Hao and Zou, 2013; Price and Schoell, 1995); therefore they can provide

PT

representative samples in-situ to investigate the complex physicochemical processes at varying pressures and temperatures (Price and Schoell, 1995). In particular, the

RI

molecular composition and carbon isotopes of gaseous alkanes often contain abundant information for the gas source precursor type, thermal maturity and gas generation

SC

process (Cao et al., 2015; Dai et al., 2004; Dai et al., 2014b; Liu et al., 2014; Schoell, 1980; Wu et al., 2015; Xia et al., 2013). Dai et al. (2013) and Li et al. (2016) also

NU

reported that the values of δ13C2 might be ineffective to determining gas origin, especially when carbon isotopic composition of gaseous alkanes exhibits full isotopic

MA

reversals, just like the phenomenon presented in Fig. 4. Therefore, it maybe not enough to use only δ13C2 values to obtain information about gases.

D

Bernard diagram, based on the relationship between the carbon isotope of

PT E

gaseous alkanes and gas molecular compositions, is commonly used to determine natural gas origin (Bernard et al., 1978), especially for gases with a high thermal maturity (Hao et al., 2008). Because the Bernard diagram takes into consideration the

CE

effects of thermal maturity and kerogen type, in this study we applied this method for the determination of biogenic, thermogenic and mixed gases. Fig. 6 shows a

AC

cross-plot of gas molecular composition and carbon isotope of gaseous alkanes from the WL shale gases. For comparison purpose, geochemical data of some typical gas fields reported in the literature are also plotted in this diagram. Similar to the gases from other typical gas fields in North American, gases from the Fuling field have a thermogenic origin (Fig. 6). All gas samples from the Fuling field, as well as the Changning field (2.8-3.3 % Ro; Feng et al., 2016) and part of the gases in Appalachian Basin, plot in the area between type Ⅱ and type Ⅲ zone, indicating that gases from the WL shales may be derived from a mixture of type Ⅱ and type Ⅲ kerogen. In contrast, gases from the Weiyuan field, Barnett in the Fort Worth Basin, Fayetteville

ACCEPTED MANUSCRIPT in the Arkoma Basin and some in the Appalachian Basin, fall into the type Ⅱ zone (Fig. 6). With the increasing maturity values, there is a clearly evolution trend for the Barnett gas samples (Fig. 6). Previous studies show that both the WL shales from the Weiyuan and Fuling fields are dominated by type Ⅰ kerogen with few type Ⅱ (Cao et al., 2015; Guo, 2013; 2015), so it is quite strange that the data from these two

PT

fields have plotted in quite different zones (Fig. 6). This inconsistent phenomenon may be related to the different thermal maturity history and tectonic activity between

RI

the Weiyuan field (2.0-2.2 % Ro; Feng et al., 2016) and Fuling filed (~3.0 % Ro; Yang et al., 2016d; Cao et al., 2015; Prinzhofer and Huc, 1995; Prinzhofer et al., 2000).

SC

After the WL shales reached their maximum burial depth in late Yanshanian, diverse uplift and denudation in the Weiyuan and Fuling fields resulted in a shallower present

NU

burial depth for Weiyuan field (~ 2000 m; Dai et al., 2014b). Because carbon isotope of gaseous alkanes and gas molecular compositions are not only closely related to the

MA

kerogen types, but also are susceptible to the effects from the thermal evolution of the source rock, pressure and gas migration process (Prinzhofer and Huc, 1995;

D

Prinzhofer et al., 2000; Xia et al., 2013).

PT E

Two genetic types of hydrocarbon gas (oil-associated gas and coal-derived gas) can be identified based on the stable carbon isotopes of gaseous alkanes (Dai et al., 2012; Liu et al., 2012; Schoell, 1980). Dai (2011) and Dai et al. (2003) systematically

CE

analyzed the amount of carbon isotopes of alkane gases from a total of 48 large gas fields in China, and found δ13C2 values can be used to classify these two types of

AC

natural gases. Usually, δ13C2 values of oil-associated gases are lighter than -28.5‰, while coal-derived gas is mostly heavier than -28.0‰ (Dai, 2011) or -29.0‰ (Gang et al., 1997). The gases with δ13C2 values between the threshold values indicate a mixed oil-associated gas and/or coal-derived gas (Dai, 2011). In this study, the δ13C2 values between -35.7 ‰ and -34.1‰ (Table 1 and Fig. 7) for the WL shale gases in the Fuling field suggest oil-associated gases, which is similar to the gases from other typical gas fields (Fig. 7). Modeling of burial and thermal history shows that WL shales have reached highly over-mature stages with vitrinite reflectance values around 3.0 %, and the production is mainly dry gas (Table 1 and Fig. 3). Due to the lack of

ACCEPTED MANUSCRIPT gas samples across a maturation gradient from immature to late mature, the WL shale gases in the Fuling field show an unclear evolution trend between the δ13C2 and wetness values (Figs. 6-7). In contrast, the Barnett Shale from the Fort Worth Basin with a maturity from 1.3 % to 2.1 % shows an isotopic reversal for values of δ13C2 within a wetness range of 0.8 % - 22 % (Fig. 7; Rodriguez and Philp, 2010; Zumberge

PT

et al., 2012). When the wetness is greater than 5 %, a normal maturity trend was observed for the Barnett shale gas. When the wetness is below 5 %, a reversal of

RI

ethane isotopic composition occurs and the δ13C2 values become more isotopically lighter (Fig. 6; Rodriguez and Philp, 2010; Zumberge et al., 2012). Hao and Zou

SC

(2013) found a concurrent reversed stable carbon isotopic distribution between δ13C2 and δ13C3, as well as iC4/nC4 ratios, suggesting the inflection point reflect the

NU

beginning of secondary cracking (Dai et al., 2014b; Xia et al., 2013). The WL shale gases in the Fuling field, as well as the gas data from the Weiyuan field, Changning

MA

field, plot in the carbon isotopic reversal zone (Figs. 6-7), indicating gases in the Fuling field are likely to be secondary cracking gases.

D

Based on experimental simulation of both kerogen and oil cracking in a confined

PT E

system, Behar et al. (1992) attempted to distinguish kerogen-cracking gas (primary cracking gas) from oil-cracking gas (secondary cracking gas). Prinzhofer and Huc (1995) discovered that gases from kerogen cracking are characterized by a rapid

CE

increase in ln(C1/C2) values, and relatively stable (or decrease slightly) C2/C3 values. On the contrary, gases from secondary cracking of hydrocarbon are characterized by a

AC

relatively stable ln(C1/C2) values and obvious increase of ln(C2/C3) (Prinzhofer and Huc, 1995). The plot of ln(C1/C2) versus ln(C2/C3), shown in Fig. 8, indicates that most of the WL gas samples from the Fuling area exhibit a narrow range of ln(C1/C2) with values from 4.7 to 5.4, and a wide range of ln(C2/C3) values (0.9 - 4.7). These variations suggest that the gases from the Fuling field are derived from secondary cracking. In contrast, most of the gases from Barnett Shale originate from primary cracking, while gas samples from the New Albany, Fayetteville, the Appalachian Basin and some from the Barnett Shale are derived from a mix of primary cracking and second cracking gases (Fig. 8). This is consistent with the work of Hill et al.

ACCEPTED MANUSCRIPT (2007) that gases within the Barnett Shale were cogenerated with oil and some gas has oil-cracking origin. With an increase of temperature, crude oil will crack and convert to lighter hydrocarbons and ultimately to methane (Tissot, 1984). Research on the oil-cracking temperatures has been carried out and discussed intensely. Such thermal cracking is

PT

generally initiated at temperatures of 120-140℃, and crude oil completely cracks to methane at temperatures of 160-190℃ (Barker, 1990; Barker and Pawlewicz, 1994;

RI

Mackenzie and Quigley, 1988; Tian et al., 2008). The WL shales have experienced continuous subsidence before the reached their maximum depth with a temperature of

SC

around 210℃ (Fig. 3). Therefore, oil cracking or wet-gas cracking have occurred within the WL shales during the deep burial stage between about 200 Ma and 80 Ma

NU

(Fig. 3).

5.2 Cause of the carbon isotopic reversal in the Fuling gas field

MA

Carbon isotopic reversal has been observed in conventional/unconventional reservoirs in many gas-producing basins from both China and North American, such

D

as Barnett Shale in the Fort Worth Basin (Xia et al., 2013; Zumberge et al., 2012;

PT E

Zumberge et al., 2009), Fayetteville Shale in the Arkoma Basin (Zumberge et al., 2012), fractured carbonate and tight sandstone reservoirs in the Appalachian Basin (Burruss and Laughrey, 2010), mixed siliciclastic carbonate units in the Western

CE

Canada Sedimentary Basin (Tilley et al., 2011), Shanxi Formation from the Sulige gas field in the Ordos Basin (Dai et al., 2004; Yu et al., 2013).

AC

Based on the different distribution patterns of alkanes (mainly CH4, C2H6 and C3H8) in various gas fields, Dai et al. (2004) summarized three types of stable carbon isotopic pattern. The most common type (typeⅠ) has positive carbon isotope series and the stable carbon isotopic composition will become more positive with an increasing carbon number (i.e. δ13C1 < δ13C2 < δ13C3), which is one of the typical characteristics for the thermogenic gases from primary cracking of kerogen (Dai et al., 2014a). Type Ⅱ, called partial reversed isotopic distribution (i.e., δ13C1 < δ13C2 > δ13C3, or δ13C1 > δ13C2 < δ13C3), is observed in a considerable number of gas samples.

ACCEPTED MANUSCRIPT Type Ⅲ is a completely or fully reversed stable carbon isotopic distribution pattern, namely δ13C1 >δ13C2 > δ13C3, which is often characterized by abiogenic gas (Dai et al., 2005b; Des Marais et al., 1981; Hosgörmez, 2007). Recent studies show that a completely or fully reversed stable carbon isotopic distribution was also observed in some shale gas producing plays with high thermal maturity (Dai et al., 2016; Xia et al.,

PT

2013; Zumberge et al., 2012; Zumberge et al., 2009). Previous studies summarized that reversed isotopic distribution would likely be

RI

caused by the following origins: (1) mixing of biogenic and abiogenic alkanes, or mixing of coal-associated gases and oil-derived gases (Dai et al., 2004; 2005a; Jenden

SC

et al., 1993), or mixing of water-soluble gas (Qin, 2012); (2) mixing of gases from two source rock intervals of same kerogen type at different levels of thermal maturity,

NU

or from the same source rock unit at various thermal maturities (Dai et al., 2004; Jenden et al., 1993; Xia et al., 2013); (3) bacterial oxidation of some alkane

MA

components (Dai et al., 2004; James and Burns, 1984); (4) thermochemical sulfate reduction (TSR) (Hao et al., 2008; Krouse et al., 1988); (5) Rayleigh-type

D

fractionation (Burruss and Laughrey, 2010; Pan et al., 2006); (6) water-kerogen redox

PT E

reactions (Burruss and Laughrey, 2010; Lewan, 1997; Price, 2001); (7) Carbon exchange at high temperature (Dai et al., 2016; Vinogradov and Galimov, 1970). Usually, it is difficult to determine what mechanism(s) is/are responsible for partial

systems

CE

and/or full isotopic reversals, as it can be extremely different between hydrocarbon with

various

geological

settings.

In

addition,

for

a

given

AC

hydrocarbon-producing system or basin, carbon isotopic reversals may be caused by the combination of different mechanisms, resulting in multiple origins of isotopic anomalies.

The WL shales have experienced from low mature, mature, into highly over-mature stages (the present Ro is about 3.0 %) over geologic times (Fig. 3). Oil cracking or wet-gas cracking occurred during the WL shales reached their maximum burials in late Yanshanian. With the increasing content of oil cracking or wet-gas cracking gases, the early mature wet gases within WL shales (heavier δ13C values) mixed with the dry gases (lighter δ13C values) generated in high levels of thermal

ACCEPTED MANUSCRIPT maturity (Qin et al., 2016). The δ13C2 values of these cracking gases will become lighter, resulting in the carbon isotopic reversal between the δ13C1 and δ13C2 values (δ13C1 > δ13C2; Dai et al., 2004; Jenden et al., 1993; Qin et al., 2016). Similar phenomenon will occur between the δ13C2 and δ13C3 values and lead to the heavier δ13C2 compared with δ13C3 (δ13C2 > δ13C3). Thus, this multi-charge and mixing of gases (early mature wet gas and post-mature dry gas) from the same WL shales at

PT

different levels of thermal maturity may result in the full isotopic reversals (Type Ⅲ;

RI

Fig. 9) in the Fuling gas field, which also illustrate the isotopic reversals for natural gases in the Appalachian Basin (Jenden et al., 1993). It is also noted that the thermal

SC

maturity values of the Barnett Shale are between 1.3 % and 2.1 %, suggesting the oil cracking or wet-gas cracking gases may also occur in geological times. However,

NU

most of the Barnett gases display a positive carbon isotopic distribution pattern (Type Ⅰ), and only few samples with wetness between 1.0 % and 1.75 % exhibit partial

MA

reversed isotopic distribution (Type Ⅱ; Fig. 9). No complete carbon isotopic reversals were observed for the Barnett shale gas, which is quite different with the WL shales in

D

the Fuling field. This may suggest that, even though carbon isotopic reversals can be

PT E

caused by the mixture of the early mature wet gas and the dry gases at different thermal maturity levels (Dai et al., 2004; Jenden et al., 1993; Xia et al., 2013), the mixing mechanism can result only in partially isotopic reversal (Dai el al., 2004) and

CE

is merely an accessory factor for the WL shale gases. Thus, there are other secondary alterations to the primitive wet gases have resulted in the fully reversed stable carbon

AC

isotopic distribution pattern for the WL shale gases in the Fuling field. In addition, multiple lines of evidence from carbon, oxygen and strontium isotope values shows an absence of cross-formations flow and pore fluid activities were restricted within the WL shales (Guo and Zhang, 2014). Thus, gases from the WL shales in the Fuling area have not undergone a mixing from other source rocks to lead to geochemical anomalies. The alkane gases from the WL shales in the Fuling area are mainly thermogenic and oil-associated (Figs. 6-7), and no abiogenic alkane and coal-derived gases are discovered, suggesting the extremely low possibility of mixing of abiogenic alkanes

ACCEPTED MANUSCRIPT and mixing of coal-associated gases and oil-derived gases. The gas compositions show a non-detection of hydrogen sulfide (H2S) in the WL shale gases, indicating that TSR can also be excluded as the main cause for geochemical anomalies. Due to different diffusion coefficients, molecule with light carbon isotopic composition will diffuse first and δ13C1 become heavier than δ13C2 and δ13C3 in the

PT

remaining gas, leading to the formation of carbon isotopic reversal pattern (Dai et al., 2016). The Fuling gas field is characterized by overpressure with a pressure

RI

coefficient (the ratio of the actual fluid pressure versus the corresponding normal hydrostatic pressure at the same depth) of 1.55 at the bottom of the WL shales,

SC

suggesting a relatively confined fluid system during the post-generation evolution (Guo, 2015). Due to the relatively confined system, the retained gases (mainly

NU

adsorbed and free gases) were well preserved (Guo, 2016; Guo and Zhang, 2014) and substantial gas diffusion and loss are not obvious. Thus, diffusion process may

MA

isotopically alter the gas isotopic composition, but it cannot be the main reason for fully reversed stable carbon isotopic distribution in the Fuling field.

D

All the WL shale gases in the Fuling gas field are typically dry with wetness

PT E

values less than 1.0 % and show a fully reversed stable carbon isotopic distribution (Fig. 9). In contrast, the gases from the New Albany Shale, Yanchang Shale and most of the Barnett Shale with a wetness larger than 1.75 % are characterized by a positive

CE

carbon isotopic distribution (Fig. 9; Dai et al., 2016; Rodriguez and Philp, 2010; Strąpoć et al., 2010; Zumberge et al., 2012); while most of the Fayetteville gases and

AC

some Barnett gases with wetness between 1.0 % and 1.75 % show partial reversed isotopic distribution (Fig. 9; Rodriguez and Philp, 2010; Strąpoć et al., 2010; Zumberge et al., 2012). Thus, the wetness values are closely related with the carbon isotopic distribution for shale gases, which are ultimately affected by the different levels of thermal maturity. Experimental investigation shows that carbon isotopic exchange between alkane gases is closely related to temperatures being experienced (Vinogradov and Galimov, 1970). In particular, when the temperature is greater than 150℃, the carbon isotopic reversal between the δ13C1 and δ13C2 values (δ13C1 > δ13C2 ) will occurs. When the

ACCEPTED MANUSCRIPT temperature increases further and reaches over 200℃, the positive carbon isotopic distribution will be replaced by a fully reversed stable carbon isotopic distribution. Take the WL shale gases for example, they have been deeply buried and reached the maximum temperature of about 210℃. Before the WL shales were uplifted to the present burial depth (Fig. 3), they have experienced high temperatures (>150℃) for

PT

about 120 Ma (between about 200 Ma and 80 Ma), which can trigger carbon isotope exchange among alkane gases and finally result in the fully reversed stable carbon

RI

isotopic distribution in the Fuling field. In contrast, the maximum ancient temperature of the Barnett shale gases (Ro = 1.3-2.1 %) is about 300°F (~149 ℃; Curtis, 2002),

SC

which are obviously smaller than that of the WL shales (about 210℃). Thus, the different thermal maturity levels can explain the different stable carbon isotopic

NU

distributions between the Barnett and WL shale gases, as well as the positive carbon isotopic distribution for the gases in the Yanchang Formation and a partial reversed

MA

isotopic distribution for the Fayetteville shale gases.

Burruss and Laughrey (2010) pointed out that the gas mixing and Rayleigh-type

D

fractionation of C2 and C3 were the causes for the reversed isotopic distribution of the

PT E

natural gases in the Appalachian Basin. When the individual components of mixed gases plot on hyperbolic curves as a function of the inverse of concentration (the red line in Fig. 10), such as the mixed gases from the Appalachian Basin, it indicates that

CE

the mixing mechanism is not obvious in WL shales. Even though our WL gas samples form the Fuling area are relatively concentrated in the Fig. 10, with no evidence of

AC

Rayleigh-type fractionation trends. By combining all shale gases from the Fuling field, Weiyuan field and Changning field, these data points are plotted on/around a straight line (the green one in Fig. 10), suggesting that the carbon isotope reversal of shale gases in these gas fields are likely affected by the Rayleigh-type fractionation of C2 and C3 to some extent. In addition, the homogenization temperatures for the fluid inclusions hosted in the Wufeng and Longmaxi shales in Well A are between 215.4℃ and 223.1℃ (Guo and Zhang, 2014), which can lead to the beginning of the process of Raleigh-type fractionation. Meanwhile, the WL shales are at high thermal maturity levels (Ro = 3.0 %), thus, the content of ethane and propane are decreasing to a low

ACCEPTED MANUSCRIPT level in the system with very limited replenish from hydrocarbon generation. In this case, the process of Rayleigh-type fractionation of C2 and C3 can result in a 60-70 % destruction of the isotopic compositions from gas‘ initial concentration (Burruss and Laughrey, 2010), which may also explain the carbon isotope reversal of the WL shale gases in the Fuling field.

PT

6. Conclusions Based on our results, the following conclusions can be drawn about the gases in

RI

WL shales from the Fuling gas field:

(1). Natural gases are typically dry and mainly composed of CH4 (97.9-98.9 %),

SC

and include a low content of C2H6, C3H8, and non-hydrocarbon gases. δ13C1 values range from -30.7 ‰ to -28.4 ‰, and are obviously heavier than those of typical shale

NU

gas fields in U.S.A. Different from Barnett shale gas, all the carbon isotopes of gaseous alkanes from the Fuling gas field display obviously full isotopic reversals,

MA

suggesting a relatively higher thermal maturity, which is confirmed by the reported measured vitrinite reflectance and the modeled values (Ro ~ 3.0 %) in this study.

D

(2). Gases are thermogenic in origin (mainly derived from secondary cracking)

PT E

and are totally oil-associated gas. The observed complete carbon isotopic reversals in the WL shales are caused by a combination of several mechanisms with isotope exchange at high temperature as the primary controlling factor. Other factors,

CE

including Rayleigh-type fractionation of C2 and C3, secondary cracking, gas diffusion mixing of gases at different thermal maturity levels, can lead to a carbon isotopic

AC

reversal to some extent, but they are not the main causes. Acknowledgments We acknowledge China Geological Survey (No. 12120114046901), Introducing Talents of Discipline to Universities (No. B14031), the National Natural Science Foundation of China (No. 41672139); China National Science and Technology Major Project (No. 2016ZX05005-001), China National Science and Technology Major Project (No. 2016ZX05025002-003) and Open Funds of State Key Laboratory of Oil

ACCEPTED MANUSCRIPT and Gas Reservoir Geology and Exploitation at Chengdu University of Technology (No. PLC-201602) for financial assistance to this research. Many thanks go to China Scholarship Council for the financial support to 1st author. We thank both SINPEC Exploration Company and SINPEC Jianghan Oilfield Branch Company for providing

PT

the shale gas samples for this study. Our special thanks are extended to

RI

Editor-in-Chief Dr. C. Özgen Karacan, as well as two anonymous reviewers, for many

SC

critical and constructive comments. References

NU

Barker, C., 1990. Calculated volume and pressure changes during the thermal cracking of oil to gas in reservoirs. AAPG Bulletin, 74(8): 1254-1261.

Barker, C.E., Pawlewicz, M.J., 1994. Calculation of vitrinite reflectance from thermal histories and

MA

peak temperatures-a comparison of methods. Vitrinite reflectance as a maturity parameter, 570: 216-229.

Behar, F., Kressmann, S., Rudkiewicz, J.L., Vandenbroucke, M., 1992. Experimental simulation in a confined system and kinetic modelling of kerogen and oil cracking. Organic Geochemistry,

D

19(1): 173-189.

Bernard, B.B., Brooks, J.M., Sackett, W.M., 1978. Light hydrocarbons in recent Texas continental shelf

PT E

and slope sediments. Journal of Geophysical Research: Oceans, 83(C8): 4053-4061. Brown, D., 2010. Research getting unconventional boost. AAPG Explorer, 31: 8-10. Burruss, R.C., Laughrey, C.D., 2010. Carbon and hydrogen isotopic reversals in deep basin gas: Evidence for limits to the stability of hydrocarbons. Organic Geochemistry, 41(12):

CE

1285-1296.

Cao, C., Zhang, M., Tang, Q., Lü, Z., Wang, Y., Du, L., Li, Z., 2015. Geochemical characteristics and implications of shale gas in Longmaxi Formation, Sichuan Basin, China. Natural Gas

AC

Geoscience, 26(8): 1604-1612 (in Chinese with English abstract). Cao, H., Zhu, C., Qiu, N., 2015. Thermal evolution of Lower Silurian Longmaxi Formation in the Eastern Sichuan Basin. Journal of Earth Sciences and Environment, 37(6): 22-32 (in Chinese with English abstract). Chalmers, G.R., Bustin, R.M., 2007. The organic matter distribution and methane capacity of the Lower Cretaceous strata of Northeastern British Columbia, Canada. International Journal of Coal Geology, 70(1): 223-239. Chalmers, G.R., Bustin, R.M., Power, I.M., 2012. Characterization of gas shale pore systems by porosimetry,

pycnometry,

surface

area,

and

field

emission

scanning

electron

microscopy/transmission electron microscopy image analyses: Examples from the Barnett, Woodford, Haynesville, Marcellus, and Doig units. AAPG Bulletin, 96(6): 1099-1119. Chen, L., Lu, Y., Jiang, S., Li, J., Guo, T., Luo, C., 2015. Heterogeneity of the Lower Silurian

ACCEPTED MANUSCRIPT Longmaxi marine shale in the southeast Sichuan Basin of China. Marine and Petroleum Geology, 65(8): 232-246. Chen, S., Zhu, Y., Wang, H., Liu, H., Wei, W., Fang, J., 2011. Shale gas reservoir characterisation: A typical case in the southern Sichuan Basin of China. Energy, 36(11): 6609-6616. Curtis, J.B., 2002. Fractured shale-gas systems. AAPG Bulletin, 86(11): 1921-1938. Dai, J., 2011. Significance of the study on carbon isotopes of alkane gases. Natural Gas Industry, 31(12): 1-6 (in Chinese with English abstract). Dai, J., Chen, J., Zhong, N., Pang, X., Qin, S., 2003. Large gas fields in China and their gas sources. Beijing: Scientific Press.

PT

Dai, J., Li, J., Luo, X., Zhang, W., Hu, G., Ma, C., Guo, J., Ge, S., 2005a. Stable carbon isotope compositions and source rock geochemistry of the giant gas accumulations in the Ordos Basin,

RI

China. Organic Geochemistry, 36(12): 1617-1635.

Dai, J., Liao, F., Ni, Y., 2013. Discussions on the gas source of the Triassic Xujiahe Formation tight

SC

sandstone gas reservoirs in Yuanba and Tongnanba,Sichuan Basin: An answer to Yinfeng et al. Petroleum Exploration & Development, 40(2): 268-275.

Dai, J., Ni, Y., Hu, G., Huang, S., Liao, F., Yu, C., Gong, D., Wu, W., 2014a. Stable carbon and

NU

hydrogen isotopes of gases from the large tight gas fields in China. Science China Earth Sciences, 57(1): 88-103.

Dai, J., Ni, Y., Huang, S., 2010. Discussion on the carbon isotopic reversal of alkane gases from the Chinese with English abstract).

MA

Huanglong Formation in the Sichuan Basin, China. Acta Petrolei Sinica, 31(5): 710-717(in Dai, J., Ni, Y., Zou, C., 2012. Stable carbon and hydrogen isotopes of natural gases sourced from the Xujiahe Formation in the Sichuan Basin, China. Organic Geochemistry, 43(2): 103-111.

D

Dai, J., Xia, X., Qin, S., Zhao, J., 2004. Origins of partially reversed alkane δ 13C values for biogenic gases in China. Organic Geochemistry, 35(4): 405-411.

PT E

Dai, J., Yang, S., Chen, H., Shen, X., 2005b. Geochemistry and occurrence of inorganic gas accumulations in Chinese sedimentary basins. Organic Geochemistry, 36(12): 1664-1688. Dai, J., Zou, C., Dong, D., Ni, Y., Wu, W., Gong, D., Wang, Y., Huang, S., Huang, J., Fang, C., Liu, D., 2016. Geochemical characteristics of marine and terrestrial shale gas in China. Marine and

CE

Petroleum Geology, 76(9): 444-463. Dai, J., Zou, C., Liao, S., Dong, D., Ni, Y., Huang, J., Wu, W., Gong, D., Huang, S., Hu, G., 2014b. Geochemistry of the extremely high thermal maturity Longmaxi shale gas, southern Sichuan

AC

Basin. Organic Geochemistry, 74(9): 3-12. Des Marais, D., Donchin, J., Nehring, N., Truesdell, A., 1981. Molecular carbon isotopic evidence for the origin of geothermal hydrocarbons. Nature, 292(8): 826-828. Feng, Z., Liu, D., Huang, S., Wu, W., Dong D., Peng, W., Han, W., 2016. Carbon isotopic composition of shale gas in the Silurian Longmaxi Formation of the Changning area, Sichuan Basin. Petroleum Exploration & Development, 43(5):769-777 (in Chinese with English abstract). Gang, W., Gao, G., Hao, S., Zhu, L., 1997. Carbon isotope of ethane applied in the analyses of genetic types of natural gas. Experimental Petroleum Geology, 19(2): 164-167 (in Chinese with English abstract). Gao, B., 2015. Geochemical characteristics of shale gas from Lower Silurian Longmaxi Formation in the Sichuan Basin and its geological significance. Natural Gas Geoscience, 26(6): 1173-1182 (in Chinese with English abstract).

ACCEPTED MANUSCRIPT Gao, L., Schimmelmann, A., Tang, Y., Mastalerz, M., 2014. Isotope rollover in shale gas observed in laboratory pyrolysis experiments: Insight to the role of water in thermogenesis of mature gas. Organic Geochemistry, 68(3): 95-106. Gao, Z., Hu, Q., 2016. Wettability of Mississippian Barnett Shale samples at different depths: Investigations from directional spontaneous imbibition. AAPG Bulletin, 100(1): 101-114. Guo, C., Xu, J., Wu, K., Wei, M., Liu, S., 2015. Study on gas flow through nano pores of shale gas reservoirs. Fuel, 143(3): 107-117. Guo, T., 2013. Evaluation of highly thermally mature shale-gas reservoirs in complex structural parts of the Sichuan Basin. Journal of Earth Science, 24(6): 863-873.

PT

Guo, T., 2015. The Fuling Shale Gas Field — A highly productive Silurian gas shale with high thermal maturity and complex evolution history, southeastern Sichuan Basin, China. Interpretation,

RI

3(2): SJ25-SJ34.

Guo, T., 2016. Discovery and characteristics of the Fuling shale gas field and its enlightenment and

SC

thinking. Earth Science Frontiers, 23(1): 29-43 (in Chinese with English abstract). Guo, T., Liu, R., 2013. Implications from marine shale gas exploration breakthrough in complicated structural area at high thermal stage: Taking Longmaxi Formation in Well JY1 as an example.

NU

Natural Gas Geoscience, 24(4): 643-651 (in Chinese with English abstract). Guo, T., Zhang, H., 2014. Formation and enrichment mode of Jiaoshiba shale gas field, Sichuan Basin. Petroleum Exploration and Development, 41(1): 31-40.

MA

Hao, F., Chen, J., 1992. The cause and mechanism of vitrinite reflectance anomalies. Journal of Petroleum Geology, 15(4):419-434.

Hao, F., Guo, T., Zhu, Y., Cai, X., Zou, H., Li, P., 2008. Evidence for multiple stages of oil cracking and thermochemical sulfate reduction in the Puguang gas field, Sichuan Basin, China. AAPG

D

Bulletin, 92(5): 611-637.

Hao, F., Zou, H., 2013. Cause of shale gas geochemical anomalies and mechanisms for gas enrichment

PT E

and depletion in high-maturity shales. Marine and Petroleum Geology, 44(6): 1-12. Hao, F., Zou, H., Lu, Y., 2013. Mechanisms of shale gas storage: Implications for shale gas exploration in China. AAPG Bulletin, 97(8): 1325-1346. Hill, R.J., Jarvie, D.M., Zumberge, J., Henry, M., Pollastro, R.M., 2007. Oil and gas geochemistry and

CE

petroleum systems of the Fort Worth Basin. AAPG Bulletin, 91(4): 445-473. Hosgörmez, H., 2007. Origin of the natural gas seep of Cirali (Chimera), Turkey: Site of the first Olympic fire. Journal of Asian Earth Sciences, 30(1): 131-141.

AC

James, A., Burns, B., 1984. Microbial alteration of subsurface natural gas accumulations. AAPG Bulletin, 68(8): 957-960. Jenden, P., Drazan, D., Kaplan, I., 1993. Mixing of thermogenic natural gases in northern Appalachian Basin. AAPG Bulletin, 77(6): 980-998. Klaver, J., Desbois, G., Urai, J.L., Littke, R., 2012. BIB-SEM study of the pore space morphology in early mature Posidonia Shale from the Hils area, Germany. International Journal of Coal Geology, 103(12): 12-25. Krouse, H.R., Viau, C.A., Eliuk, L.S., Ueda, A., Halas, S., 1988. Chemical and isotopic evidence of thermochemical sulphate reduction by light hydrocarbon gases in deep carbonate reservoirs. Nature, 333: 415-419. Lewan, M., 1997. Experiments on the role of water in petroleum formation. Geochimica et Cosmochimica Acta, 61(17): 3691-3723.

ACCEPTED MANUSCRIPT Li, P., Hao, F., Guo, X., Zou, H., Yu, X., Wang, G., 2015. Processes involved in the origin and accumulation of hydrocarbon gases in the Yuanba gas field, Sichuan Basin, southwest China. Marine and Petroleum Geology, 59(1): 150-165. Li, Y., Sun, Y., Lan, L., 2016. Discussion on the recognition of gas origin by using ethane carbon isotope. Natural Gas Geoscience, 27(4): 654-664 (in Chinese with English abstract). Liao, F., Yu, C., Wu, W., Liu, D., 2014. Stable carbon and hydrogen isotopes of natural gas from the Zhongba Gasfield in the Sichuan Basin and implication for gas-source correlation. Natural Gas Geoscience, 24(1): 79-86 (in Chinese with English abstract). Liu, Q., Jin, Z., Li, J., Hu, A., Bi, C., 2012. Origin of marine sour natural gas and gas-filling model for

PT

the Wolonghe Gas Field, Sichuan Basin, China. Journal of Asian Earth Sciences, 58(9): 24-37. Liu, Q., Jin, Z., Wu, X., Lin, J., Gao, B., Zhang, D., Xu, M., 2014. Origin and filling model of natural

RI

gas in Jiannan Gas Field, Sichuan Basin, China. Energy, Exploration & Exploitation, 32(3): 569-590. 1488-1498 (in Chinese with English abstract).

SC

Liu, R., 2015. Typical features of the first giant shale gas field in China. Natural Gas Geoscience, 26(8): Loucks, R.G., Reed, R.M., Ruppel, S.C., Jarvie, D.M., 2009. Morphology, Genesis, and Distribution of Sedimentary Research, 79(12): 848-861.

NU

Nanometer-Scale Pores in Siliceous Mudstones of the Mississippian Barnett Shale. Journal of Mackenzie, A.S., Quigley, T.M., 1988. Principles of geochemical prospect appraisal. AAPG Bulletin,

MA

72(4): 399-415.

Mathia, E.J., Bowen, L., Thomas, K.M., Aplin, A.C., 2016. Evolution of porosity and pore types in organic-rich, calcareous, Lower Toarcian Posidonia Shale. Marine and Petroleum Geology, 75(8): 117-139.

D

Pan, C., Yu, L., Liu, J., Fu, J., 2006. Chemical and carbon isotopic fractionations of gaseous hydrocarbons during abiogenic oxidation. Earth and Planetary Science Letters, 246(1): 70-89.

PT E

Pan, S., Zou, C., Yang, Z., Dong, D., Wang, Y., Wang, S., Wu, S., Huang, J., Liu, Q., Wang, D., Wang, Z., 2015. Methods for shale gas play assessment: A comparison between Silurian Longmaxi Shale and Mississippian Barnett Shale. Journal of Earth Science, 26(2): 285-294. Price, L.C., 2001. A possible deep-basin-high-rank gas machine via water-organic-matter redox

CE

reactions. US Geological Survey, Digital Data Series. Chapter H, 67. Price, L.C., Schoell, M., 1995. Constraints on the origins of hydrocarbon gas from compositions of gases at their site of origin. Nature, 378(6555): 368-71.

AC

Prinzhofer, A.A., Huc, A.Y., 1995. Genetic and post-genetic molecular and isotopic fractionations in natural gases. Chemical Geology, 126(3-4): 281-290. Prinzhofer, A., Mello, M.R., Takaki, T., 2000. Geochemical characterization of natural gas: A physical multivariable approach and its applications in maturity and migration estimates. AAPG Bulletin, 84(8), 1152-1172. Qin, H., Fan, X., Liu, M., Hao, J., Liang, B., 2016. Geochemical characteristics and geological significance of desorbed shale gas in Longmaxi Formation, Jiaoshiba area. Acta Petrolei Sinica, 37(7): 846-854 (in Chinese with English abstract). Qin, S., 2012. Carbon isotopic composition of water-soluble gases and their geological significance in the Sichuan Basin. Petroleum Exploration and Development, 39(3): 335-342. Rodriguez, N.D., Philp, R.P., 2010. Geochemical characterization of gases from the Mississippian Barnett shale, Fort Worth basin, Texas. AAPG Bulletin, 94(11): 1641-1656.

ACCEPTED MANUSCRIPT Schoell, M., 1980. The hydrogen and carbon isotopic composition of methane from natural gases of various origins. Geochimica et Cosmochimica Acta, 44(5): 649-661. Schoell, M., 1983. Genetic characterization of natural gases. AAPG Bulletin, 67(12): 2225-2238. Slatt, R.M., Rodriguez, N.D., 2012. Comparative sequence stratigraphy and organic geochemistry of gas shales: commonality or coincidence? Journal of Natural Gas Science and Engineering, 8(8): 68-84. Strąpoć, D., Mastalerz, M., Schimmelmann, A., Drobniak, A., Hasenmueller, N.R., 2010. Geochemical constraints

on

the

origin

and

volume

of

gas

in

the

New

Albany

Shale

(Devonian–Mississippian), eastern Illinois Basin. AAPG Bulletin, 94(11): 1713-1740.

PT

Sweeney, J.J., Burnham, A.K., 1990. Evaluation of a simple model of vitrinite reflectance based on chemical kinetics. AAPG Bulletin, 74(10): 1559-1570.

RI

Tang, Y., Jenden, P., 1995. Theoretical modeling of carbon and hydrogen isotope fractionations in natural gas. In Organic Geochemistry, Developments and Applications to Energy, Climate and

SC

Human History (eds. J. O. Grimalt and C. Dorronsoro), European Association of Organic Geochemists: 1067-1069. A.I.G.O.A.

Tang, Y., Xia, X., Ferworn, K., Zumberge, J., 2012. Interpretation of gas isotope reversal in natural gas

NU

from high mature source rocks, in: Isotopic reversal (‗rollover‘) in shale gases produced from the Mississippian Barnett and Fayetteville Formations: 43-52. Tian, H., Xiao, X., Wilkins, R.W., Tang, Y., 2008. New insights into the volume and pressure changes

MA

during the thermal cracking of oil to gas in reservoirs: Implications for the in-situ accumulation of gas cracked from oils. AAPG Bulletin, 92(2): 181-200. Tilley, B., McLellan, S., Hiebert, S., Quartero, B., Veilleux, B., Muehlenbachs, K., 2011. Gas isotope reversals in fractured gas reservoirs of the western Canadian Foothills: Mature shale gases in

D

disguise. AAPG Bulletin, 95(8): 1399-1422.

Tilley, B., Muehlenbachs, K., 2013. Isotope reversals and universal stages and trends of gas maturation

PT E

in sealed, self-contained petroleum systems. Chemical Geology, 339(2): 194-204. Tissot, B., 1984. Recent advances in petroleum geochemistry applied to hydrocarbon exploration. AAPG Bulletin, 68(5): 545-563.

Tong, C., 1992. Structure evolution and petroleum accumulation of Sichuan Basin. Geology Press, US

Energy

CE

Beijing, 21-24 pp.

Information

Administration,

2015,

Annual

energy

outlook

2015,

https://www.eia.gov/pressroom/presentations/sieminski_05042015.pdf.

AC

Vinogradov, A. P., Galimov, E. M., 1970. Isotopism of carbon and the problem of oil origin. Geochemistry, 3: 275-297. Wang, Z., Li, X., Zhang, J., Xu, H., Wang, G., Zhou, B., Yang, J., Wang, Y., 2016. Longmaxi formation shale gas geochemical features comparison between different blocks in Sichuan Basin. Coal Geology of China, 28(2): 22-27 (in Chinese with English abstract). Wei, X., Guo, T., Liu, R., 2016. Geochemical features of shale gas and their genesisin Jiaoshiba block of Fuling Shale Gasfield,Chongqing. Natural Gas Geoscience, 27(3): 539-548 (in Chinese with English abstract). Whiticar, M.J., 1990. A geochemial perspective of natural gas and atmospheric methane. Organic Geochemistry, 16(1-3): 531-547. Wu, W., Fang, C., Dong, D., Liu, D., 2015. Shale gas geochemical anomalies and gas source identification. Acta Petrolei Sinica, 36(11): 1332-1340, 1366 (in Chinese with English

ACCEPTED MANUSCRIPT abstract). Xia, X., 2014. Kinetics of gaseous hydrocarbon generation with constraints of natural gas composition from the Barnett Shale. Organic Geochemistry, 74(9): 143-149. Xia, X., Chen, J., Braun, R., Tang, Y., 2013. Isotopic reversals with respect to maturity trends due to mixing of primary and secondary products in source rocks. Chemical Geology, 339(2): 205-212. Yang, R., He, S., Hu, Q., Hu, D., Zhang, S., Yi, J., 2016a. Pore characterization and methane sorption capacity of over-mature organic-rich Wufeng and Longmaxi shales in the southeast Sichuan Basin, China. Marine and Petroleum Geology, 77(11): 247-261.

PT

Yang, R., He, S., Li, T., Yang, X., Hu, Q., 2016b. Origin of over-pressure in clastic rocks in Yuanba area, northeast Sichuan Basin, China. Journal of Natural Gas Science and Engineering, 30(3):

RI

90-105.

Yang, R., He, S., Wang, X., Hu, Q., Hu, D., Yi, J., 2016c. Paleo-ocean redox environments of the Upper

SC

Ordovician Wufeng and the first member in lower Silurian Longmaxi formations in the Jiaoshiba area, Sichuan Basin. Canadian Journal of Earth Sciences, 53(4): 426-440. Yang, R., He, S., Yi, J., Hu, Q., 2016d. Nano-scale pore structure and fractal dimension of organic-rich

NU

Wufeng-Longmaxi shale from Jiaoshiba area, Sichuan Basin: Investigations using FE-SEM, gas adsorption and helium pycnometry. Marine and Petroleum Geology, 70(2): 27-45. Yu, C., Huang, S., Gong, D., Liao, F., Li, J., Sun, Q., 2013. Partial reversal cause of carbon and

MA

hydrogen isotope compositions of natural gas: A case study in Sulige gas field, Ordos Basin. Acta Geologica Sinica, 34(8): 92-101 (in Chinese with English abstract). Zhu, C., Rao, S., Yuan, Y., Wang, Q., Qiu, N., Hu, S., 2013. Thermal evolution of the main Paleozoic shale rocks in the Southeastern Sichuan Basin. Journal of China Coal Society, 38(5): 834-838

D

(in Chinese with English abstract).

Zumberge, J., Ferworn, K., Brown, S., 2012. Isotopic reversal (‗rollover ‘) in shale gases produced from

PT E

the Mississippian Barnett and Fayetteville formations. Marine and Petroleum Geology, 31(1): 43-52.

Zumberge, J., Ferworn, K., Curtis, J., 2009. Gas character anomalies found in highly productive shale

AC

CE

gas wells. Geochimica et Cosmochimica Acta, 73(13): A1539.

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

Fig. 1(a) Location map of the study area and the total thickness contour line of the Upper Ordovician Wufeng and the bottom part of Lower Silurian Longmaxi shale formations in the Sichuan Basin (modified after Chen et al., 2015; Guo and Zhang, 2014; Yang et al., 2016a); (b) The bottom depth map of Wufeng Formation in the study area; (c) A southeast cross section of Well A showing the main stratigraphic intervals.

AC

CE

PT E

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Fig. 2. Simplified stratigraphic units in the Fuling area and the Upper Ordovician Wufeng and the bottom of Lower Silurian Longmaxi Formations (referred as WL shales in this paper). The wavy lines represent unconformities. Sym. = Symbol.

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

Fig. 3. Burial and thermal history of WL shales for Well A in the Fuling area (its location is shown in Fig. 1b). The equivalent vitrinite reflectance (Ro %) values are from the work of Yang et al. (2016d).

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

Fig. 4. Distribution of carbon isotope of gaseous alkanes (CH4, C2H6 and C3H8) in Fuling gas field, showing a full isotopic reversal (δ13C1 > δ13C2 > δ13C3); similar phenomenon was also observed in Weiyuan and Changning gas fields.

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

Fig. 5. Plot of δ13C1 versus δ13C2 for gas samples in this study and gases from the following basins or fields in the literature: (1) the Fuling gas field (Wang et al., 2016; Dai et al., 2016); (2) the Changning gas field (Wang et al., 2016; Cao et al., 2015); (3) the Weiyuan gas field (Wang et al., 2016; Cao et al., 2015); (4) Yangchang Formation in the Ordos Basin (Dai et al., 2016), Fayetteville in the Arkoma Basin (Zumberge et al., 2012); Barnett in the Fort Worth Basin (Rodriguez and Philp, 2010; Zumberge et al., 2012); the Appalachian Basin (Burruss and Laughrey, 2010); and New Albany in the Illinois Basin (Strąpoć et al., 2010), showing a trend of ―N‖ shape. All gas samples from the Fuling gas field are located in the reversed region.

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

Fig. 6. Plot of δ13C1 versus C1H4/(C2H6+C3H8) (modified from Bernard et al., 1978) in gases from the Fuling gas field (red diamonds). Data points from other typical shale gas-producing basins are also plotted for comparison (see the caption of Fig. 5 for data source).

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

Fig. 7. Relationships between wetness and δ13C2 (ethane) from the WL shale gases in the Fuling gas field (red diamonds) and other typical shale gas-producing basins.

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

Fig. 8. Plot of ln(C1/C2) versus and ln(C2/C3) in gases from the WL shales in the Fuling area and other gas-producing areas.

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

Fig. 9. The distribution of wetness contents for three types of stable carbon isotopic pattern for gases from the typical shale gas-producing areas.

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

Fig. 10. Relationship between ethane (C2) content and δ13C2 values for gas samples from the WL shales in the Fuling field.

ACCEPTED MANUSCRIPT Table 1 Molecular components and stable carbon isotope of shale gases from the WL shales in the Fuling gas field, Sichuan Basin. ND = no data.

2395-240 8 2410 2410 2778 2790 2870

3560-377 2 4010 4010 3158 3158 4152

8

N2

(%)

Carbon isotope (‰ VPDB) δ13C1

δ13C2

δ13C3

-29.9 8 -28.3 5 -30.3 3 -30.5 1 -29.5 5 -29.5 5 -30.3 1 -29.6 0 -29.3 0 -29.5 7 -28.3 6 -30.5 1 -29.5 5 -29.5 5 -29.0 3 -30.0 1 -30.5 1 -30.7 1 -30.0 2

-35.5 2 -34.1 8 -34.3 4 -34.1 0 -34.6 8 -34.6 8 -34.2 5 -34.6 0 -34.1 0 -34.5 9 -34.1 8

-38.4 2 -36.7 1 -37.5 6

-34.1

ND

-34.6 8 -34.6 8 -34.4 7 -35.7 3 -34.4 7 -34.3 7 -35.4 7

-35.0 3 -35.0 3 -37.0 5 -38.4 5

δ13CCO

0.17

0.75

0.59

0.63

0.02

0.20

0.80

0.66

0.52

0.02

0.27

0.78

0.55

0.60

0.02

0.32

0.75

0.63

0.84

0.02

0.24

ND

0.62

ND

0.22

0.48

0.62

0.02

0.22

0.70

0.02

0.22

0.48

0.73

0.68

0.02

ND

ND

0.71

0.68

0.02

N D

N D

0.71

0.68

0.02

0.1

0.84

0.70

0.66

0.02

0.12

0.81

0.69

0.84

0.02

ND

ND

0.86

0.72

0.03

ND

ND

0.76

0.94

0.02

0.18

0.96

0.97

0.51

0.02

0.13

0.75

0.53

0.51

ND

0.33

0.84

0.52

0.50

0.02

0.36

0.94

0.53

0.53

0.02

0.21

0.73

0.56

0.67

36

RI

0.02

SC

0.56

PT

2

0.86 0.63

NU

2395

6

CO2

MA

2395

C3H

D

2385-240 0 2385-240 5 2385-240 5 2410-241 5

98.5 0 98.3 5 98.4 1 98.3 1 98.9 0 98.4 7 98.4 7 98.5 8 98.2 6 98.4 1 98.3 4 98.3 4 98.9 0 98.2 3 97.9 0 98.5 9 98.3 2 98.1 8 98.5 1

C 2H

PT E

2380

CH4

CE

Well A-01 Well A-02 Well A-03 Well A-04 Well A-05 Well A-06 Well A-07 Well A-08 Well A-09 Well A-10 Well B-01 Well B-02 Well B-03 Well B-04 Well C-01 Well C-02 Well C-03 Well C-04 Well D-01

Depth (m)

AC

Sampl e ID

Wetnes s

Main component (vol. %)

0.65

ND -35.0 3 -35.0 3 -36.4 4 -36.1 0 ND -36.1 2 -36.7 2

-6.14 ND ND ND -9.50 -11.74 -11.14 -11.70 ND ND ND ND -9.50 ND ND -5.38

ND

ND

ND

ND

-37.5 8

-3.75

ACCEPTED MANUSCRIPT

4152 4152 2778

0.54

0.02

0.25

0.82

0.57

0.65

0.02

0.27

0.73

0.68

0.43

ND

0.18

0.74

0.43

0.52

0.02

0.22

0.68

0.54

0.67

0.02

0.46

0.70

0.70

-29.0 7 -30.1 2 -30.4 1 -29.9 9 -30.2 0

-34.3 4 -34.3 1 -34.3 3 -35.4 4 -34.6 0

CE

PT E

D

MA

NU

SC

RI

2778

98.3 7 98.3 3 98.6 5 98.5 6 98.1 5

PT

4152

AC

Well D-02 Well D-03 Well D-04 Well E-01 Well E-02

37

-37.1 4

ND

ND

ND

-36.0 9 -38.7 1 ND

ND -3.30 ND

ACCEPTED MANUSCRIPT Highlights 1. A total of 24 gas samples from Wufeng-Longmaxi shales of Fuling field are analyzed for compositions and carbon isotopes 2. Gases are dry with an average methane content of 98.4 %

PT

3. Gases are classified as oil-associated gas, and mainly derived from secondary cracking

RI

4. Gases display full alkane carbon isotopic reversals (δ13C1 > δ13C2 > δ13C3), mainly caused by

AC

CE

PT E

D

MA

NU

SC

isotope exchange at high temperature

38