Logging identification and characteristic analysis of marine–continental transitional organic-rich shale in the Carboniferous-Permian strata, Bohai Bay Basin

Marine and Petroleum Geology 70 (2016) 273e293

Contents lists available at ScienceDirect

Marine and Petroleum Geology journal homepage: www.elsevier.com/locate/marpetgeo

Research paper

Logging identification and characteristic analysis of marineecontinental transitional organic-rich shale in the Carboniferous-Permian strata, Bohai Bay Basin Jianhua He a, b, c, Wenlong Ding a, b, c, *, Jinchuan Zhang a, b, c, Ang Li a, Wei Zhao a, Peng Dai a a

School of Energy Resources, China University of Geosciences, Beijing 100083, China Key Laboratory for Marine Reservoir Evolution and Hydrocarbon Abundance Mechanism, Ministry of Education, China University of Geosciences, Beijing 100083, China c Key Laboratory for Shale Gas Exploration and Assessment, Ministry of Land and Resources, China University of Geosciences, Beijing 100083, China b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 May 2015 Received in revised form 22 October 2015 Accepted 11 December 2015 Available online 15 December 2015

Currently, global shale gas exploration and exploitation are focused on marine shale. Recently, major shale gas-oil breakthroughs have been made within continental and marineecontinental transitional shale in China. This study will show how transitional shale is of great importance. Based on geological field surveys, core observations, thin section analysis, organic geochemistry and X-ray diffraction, we systematically studied the basic geological characteristics (including lithology, mineral composition, and organic geochemistry) of this transitional shale. By comparative analysis of well logging data from 260 wells in the TaiyuaneShanxi shale, we will show that these methods are effective for identifying organicrich shale from conventional well logs and determining its thickness distribution in the Carboniferous ePermian strata of the TaiyuaneShanxi transitional coal-bearing formation. The results indicate that the TaiyuaneShanxi shale has a high TOC (most 2%e4%) and that the lithology is primarily carbonaceous shale with type Ⅱ2-Ⅲ kerogens. The high thermal maturation (Ro
1.1%) favors the generation of gas. The mineral components primarily include clay minerals, quartz, and plagioclase with a moderate brittle mineral content (47 wt.%) and high clay mineral content (51 wt.%) dominated by kaolinite (43%) and mixed-layer illite-smectite (31%). The transitional organic-rich shale on conventional log curves is generally characterized by higher gamma ray (GR), neutron porosity (CNL), acoustic travel time (AC), resistivity (Rt), potassium (K), and uranium (U) readings and a lower density (DEN), photoelectric absorption index (PE) and thorium-uranium ratio (TH/U). After analyzing the log response characteristics of the organic-rich shale, the most sensitive logging curves (such as CNL, AC, DEN, PE and U) were optimized to conduct logging overlays and to construct cross-plots to qualitatively identify organic-rich shale. The identified organic-rich shale amalgamates in the middle-upper member of the Taiyuan Formation and the lower member of the Shanxi Formation consistent with the results of the TOC analysis and practical gas logging. Based on the qualitative evaluation methods of the modified △LogR and a multivariate linear regression model, we calculated the TOC of shale wells in the TaiyuaneShanxi formation. From this we calculated the characteristic values of organic-rich shale thickness. The results indicate that organic-rich shale in the Taiyuan formation is thicker than that in the Shanxi formation. Additionally, the thickness of organic-rich shale within lagoons and deep reed swamp facies are much thicker (25e35 m and 40e80 m) than other structural profile types, whereas their lateral distribution is less than that of marine shale. The relatively small continuous thickness of the single shale layer and high clay content may have negative effects for developing the shale gas potential. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Bohai Bay Basin CarboniferousePermian Marineecontinental transitional facies Shale gas Geological characteristics Logging identification Gas shale potential

1. Introduction * Corresponding author. School of Energy Resources, China University of Geosciences, Beijing 100083, China. E-mail address: [email protected] (W. Ding). http://dx.doi.org/10.1016/j.marpetgeo.2015.12.006 0264-8172/© 2015 Elsevier Ltd. All rights reserved.

Shale gas is a type of unconventional continuous natural gas

274

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

accumulation that is derived from the biodegradation and/or the thermal maturation of organic-rich shale. Most shale gas is stored in micro-fractures or nanoscale pores in a free state, although some is adsorbed onto kerogen, clay grains and pore surfaces. A small part is stored in kerogen and silt interlayers in a dissolved state (Schoell, 1980; Curtis, 2002; Martini et al., 2003; Zhang et al., 2004; Hill et al., 2007). Because shale gas is an important unconventional natural gas resource, it plays a key role in meeting American and global energy demands. According to statistics from EIA on 7 May 2014, the production of shale gas reached 10.95 TCF, accounting for 60% of American total natural gas production, and production is still increasing (Conti et al., 2014). China is one of the countries that pioneered shale oil and gas research and has made several breakthroughs and discoveries, such as the occurrence of the shale gas in the Longmaxi shale formation in Southern China (e.g., the areas of Jiaoshiba, Changning, Weiyuan, and Zhaotong) (Jiang et al., 2014; Guo and Zhang, 2014). During the polycyclic structural evolution process, three types of organic-containing shale, including marine, marineecontinental transitional (hereafter referred to as “transitional”) and lacustrine shale have accumulated in China's sedimentary basin. Additionally, CarboniferousePermian transitional shale was widely deposited in northern China (Ma et al., 2004). The organic-rich shale layers in the CarboniferousePermian strata are well-developed and are significant potential targets for shale gas prospecting. Compared to marine and lacustrine shale, transitional shale has unique geological characteristics. Recently, gas was detected in the CarboniferousePermian shale during drilling, fracturing and field desorption experiment (Field desorption experiment can quickly and effectively evaluate the gas content in the shale cores recently collected from the bottom hole). Examples are found in the Taiyuan and the Shanxi formation shale, including the Mouye 1 well and the Weican 1 well in the southern North China Basin, the Eye 1 well (3.8  104 m3/d production after fracturing) in the Ordos areas and the T2905 well in the eastern salient of the Liaohe Depression (Ge et al., 2012; Bao et al., 2014). These well results indicate that transitional shale has a large shale gas exploration potential and may be as important as marine shale. Lower economic cost, higher vertical resolution, better continuity, and geologically abundant information have led to the wide application of logging technology to the exploration and production of marine shale gas in North America and southern China. Productive achievements and good application effects have been achieved by evaluating logging technology and simultaneously considering the geochemistry (Schmoker, 1979, 1981; Passey et al., 1990; Zhao et al., 2007; Jacobi et al., 2008; Pemper et al., 2009; Bowman, 2010), the mineral composition (Flower, 1983; Luffel et al., 1992; Pollastro et al., 2007; Ross et al., 2008; Montgomery et al., 2005), the reservoir characteristics (Soeder, 1984; Luffel, 1989; Ross and Bustin, 2008; LeCompte, 2010; Wang et al., 2015), the gas content (Kim, 1977; Lewis et al., 2004; Pan et al., 2011), and the fracturing feasibility (Gatens et al., 1990; Rickman et al., 2008) of marine organic-rich shale. However, logging technology was seldom used to evaluate transitional organic-rich shale reservoirs. Moreover, effective methods to systematically identify and evaluate transitional organic-rich shale reservoirs are lacking. Thus, there is a shortage of shale gas reservoir evaluation by logging technology

275

and this method requires improvement. We focus on the basic geological characteristics of CarboniferousePermian shale by core observations, microscopic analysis of thin sections, and field geological surveying, combined with data from organic geochemistry and X-ray diffraction. Log response characteristics of organic-rich shale and the detailed comparative analysis of logging curves in shale layers are used to qualitatively identify and quantitatively evaluate transitional organic-rich shale. The characteristics of the organic-rich shale's thickness are also used to provide guidance in the exploration of shale gas.

2. Geological setting The Bohai Bay Basin is one of the major petroliferous basins in the Bohai Sea area and the coast of eastern China. It covers approximately 2  105 km2 and includes the northern part of the North China Plain, the marine Bohai Bay and the Xialiaohe Plain. The area reaches the Yanshan Mountains in the north, the Western Shandong Hills in the south, the Taihang Mountains in the west and the Eastern Jiaozhou Peninsula in the east (Fig. 1). According to the distribution of regional faults, the Bohai Bay Basin can be subdivided into eight independent depressions (-blank areas in Fig. 1) and five uplifts (-orange areas in Fig. 1). These include the Liaohe Depression and the Liaodongwan Depression in the north; the Jizhong Depression, the Cangxian Uplift, the Huanghua Depression, the Chenning Uplift, the Jiyang Depression, the Bozhong Depression, and the Changwei Depression in the central region; and the Linqing Depression and the Neihuang Uplift in the south. In the Neopaleozoic, the Bohai Bay Basin was a “cratonic basin” within an Archaeozoic- Paleopr oterozoic crystalline basement setting, which was part of the North China Platform. It was wedged between the Siberia Plate in the north and the Yangtze Plate in the south. The tectonic setting of the Bohai Bay Basin reflects the characteristics of the North China Platform and its dynamic processes. The sedimentological evolution of the Bohai Bay Basin comprises four major stages: (1) a resistant marine carbonate platform in the Early Paleozoic, (2) a marine and continental alternation in the Late Paleozoic, (3) an intracontinental lake basin depression that was filled with coarse clastics in the Early Mesozoic, and (4) rifting and post-rifting in the Late Mesozoic and the Cenozoic (Qi and Yang, 2009). Throughout the CarboniferousePermian of the Late Paleozoic, continuous regression with multiple small-scale transgressions occurred and formed vertically thin and laterally wide sea-flooding layers (Fig. 2). Furthermore, the Late Carboniferous-Early Permian was a crucial period during which the Bohai Bay Basin environment transitioned from an epicontinental sea to an intracontinental lake (Fig. 2). During this phase, the Taiyuan and Shanxi Formation shales were widely deposited in the basin. The Taiyuan Formation is dominated by tidal flat-lagoon sediments, harboring thin-bedded sandstone, limestone and coal seams, and has a relatively thick layer of black shale, approximately 20e120 m thick (Fig. 2). The Shanxi Formation features delta front-coastal marsh sediments, with middle-bedded sandstone and thin-bedded coal seams, and has produced two sets of black shale with a thickness of 10e100 m (Zhang et al., 2009). Apart from a shortage of shale deposition in the uplift succession,

Fig. 1. a. Location of the study area; Maps show the structural features and structural divisions of the Bohai Bay Basin. The names of the tectonic units are as follows: JZ.D for Jizhong Depression, CX.U for Cangxian Uplift, HH.D for Huanghua Depression, LQ.D for Linqin Depression, XH.U for Xunheng Uplift, DP.S for Dongpu Sag, NH.U for Neihuang Uplift, TY.G for Tangyin Graben, XLH.D for Xialiaohe Depression, LDW.D for Liaodongwan Depression, HZ.U for Haizhong Uplift, CB.D for Central Bohai Depression, EB.S for E. Bohaig Sag, CN.U for Chenning Uplift, JY.D for Jiyang Depression, CW.D for Changwei Depression. The names of the master boundary faults (MBFs) are as follows: ① is the MBF of Jizhong Depression, ② is the MBF of Tangyin Graben, ③ is the MBF of Huanghua Depression, ④ is the MBF of Linqin Depression, ⑤ is the MBF of Dongpu Sag, ⑥ is the MBF of Jiyang Depression, ⑦ is the MBF of East Bohai Sag, ⑧ is the MBF of Central Bohai Sag. ⑨ is the MBF of Liaodongwan Depression, ⑩ is the MBF of Xialiaohe Depression; b. Structural cross sections through the Bohai Bay Basin: (a) Cross-section AeA0 across the Liaohe Depression. (b) Cross-section BeB0 across the Jizhong Depression. (c) Cross-section CeC0 across the Dongpu Sag.

276

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 2. Stratigraphic column and target layers (elements are shown in red) of the Upper Paleozoic transitional strata.

dark shale is broadly distributed in the eastern salient of the Liaohe Depression, the Suqiao-Wenan Slope of the Jizhong Depression, and the southern regions of the Huanghua, the Linqin and the

Dongpu Depressions. Additionally, there are some scattered outcrops of the Taiyuan and the Shanxi Formation shale in the peripheral zones of the basin (Fig. 1a, Fig. 6aec).

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

3. Databases and methods 3.1. Sample sources In the study area, there are hundreds of wells that were drilled previously with the goal of discovering conventional gas-bearing layers in the CarboniferousePermian strata. The T2905 well in the eastern salient of the Liaohe Depression was drilled with the specific aim of characterizing an unconventional shale gas reservoir. In this study, core descriptions, experimentally analyzed samples, and thin-section photographs were obtained from the Taiyuan and the Shanxi Formation shale cores from 41 cored wells in four depressions (the Jizhong, the Jiyang, the Liaohe, and the Dongpu Depressions) of the Bohai Bay Basin, with a total length of 603.04 m, and three outcrops in the adjacent areas, totaling 58 m (Fig. 1a).

277

entitled “Quantitative Analysis of Total Contents of Clay Minerals and Common Non-clay Minerals in Sedimentary Rocks by X-ray Diffraction”. The lithofacies descriptions, microscopic structure, and structural features were determined using an OLYMPUS Polarizing Microscope at the China University of Geosciences (Beijing) from 16 polished thin sections of shale that were cut from cores and three outcrops. All of these sections were ground to 2 cm  2 cm before analysis. The conventional well logs, abnormal gas logging, drilling and coring materials of 260 wells from five main depressions in the Bohai Bay Basin were sorted to analyze the log response characteristics of the organic-rich shale in the Taiyuan and the Shanxi Formations from the Upper Paleozoic.

3.2. Experimental analyses

3.3. Logging identification methods of organic-rich shale

Organic geochemistry data (TOC, vitrinite reflectance, kerogen maceral composition and rock-eval pyrolysis) were provided by the Geochemistry Laboratory of Yangtze University. A Leco CS-200 carbon-sulfur analyzer was used to measure the total organic carbon content (TOC) of 209 samples. Crushed samples (approximately 100 mg of 120-mesh size) were heated to 1200  C in an induction furnace after removing the carbonates with hydrochloric acid (HCl). The experiments were conducted according to Chinese Standard GB/T19145-2003, “Determination of Total Organic Carbon in Sedimentary Rock”, at a temperature of 25  C and a relative humidity (RH) of 60%. Twenty-two shale samples were analyzed with a vitrinite reflectance test using microscope photometer instruments. This detailed experimental process is based on the Chinese Standard SY/T5124-1995, “Determination of Vitrinite Reflectance in Sedimentary Rock”. According to Chinese Standard SY/T5125-1995, “Identification and Classifications of Kerogen Maceral Composition under Reflected and Transmitted Light”, we first separated the kerogen from the shale with HCL or HF to identify the kerogen maceral composition. Next, the kerogen sample was thin-sectioned with polyvinyl, and its characteristics were viewed with a microscope. Additionally, 151 shale samples were analyzed using a CoreLab OGE-Ⅱ Rock-Eval instrument. Crushed samples (approximately 100 mg of 100-mesh size) were heated at a programmed rate. Then, the hydrogen and carbon dioxide emitted from the heated organic matters in the rock was quantitatively detected by a flame ionization detector (FID) and a thermal conductivity detector (TCD). The measured parameters include TOC content, free oil or volatile hydrocarbon content expressed as mg HC/g rock (S1) vaporized at a temperature of 300  C, the remaining hydrocarbon generation potential as mg HC/g rock (S2) between the temperature of 300 and 600  C, and the temperature of maximum pyrolysis yield (Tmax). The testing methods were from the Chinese Standard GB/T191452003, “Rock-Eval Analysis”. X-ray diffraction (XRD) was used on 40 samples to quantitatively analyze the mineral composition at the CNNC Beijing Research Institute of Uranium Geology. All of the shale samples were ground into a fine powder (<40 microns) and were then analyzed with a Panalytical X0 PertPRO MPD X-ray diffraction with Cu Karadiation (40 kV, 30 mA) at a scanning speed of 2 /min and a testing angle range of 3 e30 . Computer analyses of the diffraction patterns showed the relative abundances of various mineral phases to be determined and a semi-quantitative assessment was performed. The experimental standards and testing methods employed here are in accordance with SY/T6201-1996, the oil and natural gas industry standard of the People's Republic of China

Organic matter content detection and evaluation is an important objective for shale gas-oil exploration and exploitation. Many effective methods have been successfully applied to identify and evaluate marine organic-rich shale in North America and South China, such as the DlogR technique (Passey et al., 1990), GR-DEN (Schmoker, 1981), elemental capture spectroscopy (ECS) and conversion factors of kerogen (Lewis et al., 2004), spectral gamma ray and pulsed neutron (Pemper et al., 2009), and artificial neural networks (Khoshnoodkia et al., 2011), and DEN-ECS-NML (nuclear magnetism log) (Jacobi et al., 2008). Logging technology has rarely been used to identify and evaluate transitional organic-rich shale. Because of the difference in the sedimentary environment between marine and transitional organic-rich shale, the geological characteristics and logging responses of transitional organic-rich shale are different from that of marine organic-rich shale. Some methods are effective for identifying marine organic-rich shale but are not applicable to transitional organic-rich shale (e.g., DEN-GR). By analyzing the log response characteristics of transitional organic-rich shale, some logging parameters sensitive to the variation of TOC were optimized to conduct multiple overlays and construct cross-plots to qualitatively or semi-quantitatively identify organic-rich shale. Then, a quantitative evaluation of the TOC in the significant shale intervals for the shale gas resource potential was performed. Continuous direct experimental measurements of TOC on well-cores are accurate, although expensive, and cannot be systematically conducted for every well. In contrast, the logging method used to calculate the TOC of shale in this situation is more advantageous. Considering the limitations and assumptions of logging data and experimental TOC data, two types of logging methods to calculate the TOC were selected: simple or multivariate linear regression models (equation (1) and (2)) and a modified DlogR technique to compute the TOC (equation (3)). TOC ¼ a$GR þ b,

(1)

TOC ¼ a$GR þ b$CNL þ c$AC þ d$DEN þ e$RT þ f,

(2)

TOC ¼ a$logRt þ b$AC þ c,

(3)

where a, b, c, d, e, and f are undetermined coefficients obtained from the relationship between the TOC data and the logging value; GR in Equation (1) is the natural gamma ray logging value, which can be replaced by the CNL, the AC, the DEN, or the RT value; Equation (2) is a combination model of the whole logging parameters. Equation (3) is a modification of the DlogR technique from Passey et al. (1990).

278

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 3. Histogram of total organic carbon content in the C3t (a) and the P1s (b) shale.

4. Results and discussion 4.1. The geological characteristics of transitional shale 4.1.1. Geochemical characteristics A high TOC is a prerequisite for dark shale to be a high-quality gas resource rock and has some influence on the adsorbed gas content of the shale (Ross and Bustin, 2009). According to data from 209 samples of C3t and P1s shale, the TOC of the TaiyuaneShanxi Formation shale is high and ranges widely from 0.1% to 21.1%, with most values (55%) reaching the standard for organic-rich shale (TOC
2.0%). The C3t shale has a total organic carbon content of 0.10%e21.11%, mainly 2%e4%, and 80% of the samples have a total organic carbon content of more than 2% (Fig. 3a). Other than the carbonaceous shale samples with a higher TOC (TOC > 10%), the TOC average value of the C3t shale is 2.93%. However, the TOC of the P1s shale is lower than that of the C3t shale, ranging from 0.10% to 18.84%, with most values in the range of 1%e3%, and 59% of the samples have a TOC of more than 2%. Except for the carbonaceous shale samples with a higher TOC (TOC > 10%), the TOC average value of the P1s shale is 1.92% (Fig. 3b). There is a difference in the TOCs from different types of shale lithologies. The TOC of carbonaceous shale is much higher than that of ordinary dark shale (Fig. 3). Although the carbonaceous shale comprises only a small portion of the shale samples (10%), it contributed greatly to the quantity of hydrocarbon that was generated. A total of 151 shale samples from the Upper Paleozoic, in which the S1þS2 and the TOC

Fig. 4. Plot of the hydrogen index (S1 þ S2) versus the TOC, outlining the source potential of different lithotypes and the relationship between the TOC and S1þS2 from the C3t and the P1s shale.

(S1: volatile hydrocarbon content; S2: the remaining hydrocarbon generation potential) could be tested were used to plot the correlation on a log10 scale for the S1þS2 and the TOC (Fig. 4). A doublelogarithmic relationship exists between the S1þS2 and the TOC. The curve fit equation is log (S1 þ S2) ¼ 1.29 log (TOC) 0.25. The correlation coefficient (R2) is 0.90, indicating that the log (S1 þ S2) and

Fig. 5. The organic matter type of the C3t and the P1s shale. (a) Cross-plot between hydrogen index and Tmax showing the kerogen-type and source potential of the C3t and the P1s shale (North American marine shale data provided by Hui and Stephen (2013), Travis et al. (2008), and Dariusz et al. (2010). (b) Ternary chart of kerogen maceral of the C3t and the P1s shale.

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

the log (TOC) are well-correlated. Generally, when S1 þ S2 is 1.3 mg/ g or more (TOC
2.0%), the shale might be organic-rich shale. The hydrocarbon potential of ordinary dark shale mainly ranges from 0.1 mg/g to 5 mg/g, whereas the hydrocarbon potential of carbonaceous shale is generally more than 10 mg/g (Fig. 4). Therefore, the carbonaceous shale layer may be the most important organic-rich shale layer. The type of organic matter not only determines whether hydrocarbons are generated in shale but also influences the richness of organic pore development (Loucks et al., 2009). The organic matter type of the C3t- P1s shale was dominated by Type Ⅲ kerogen (68.7%) and rarely contained TypeⅡ1eⅡ2 kerogen (23.8%), as shown in the cross-plot of Tmax (temperature of maximum pyrolysis yield) vs. HI (hydrogen index) in Fig. 5a. However, American marine shale that has been commercially exploited (Barnett, Bakken, and New Albany shale) is dominated by Type Ⅱ1eⅡ2 kerogen, and the HI (mostly
150 mg/g) is much larger than that of the C3t- P1s shale (50 mg/g). Additionally, according to the composition determination of kerogen maceral from 25 samples, most of the C3t- P1s shale samples plotted in the left corner of the ternary cross-plot, which indicates vitrinite þ interrinite > 40%, sapropelinite < 30%, and exinite < 30%. This result indicates that the organic matter originated from higher plants. Type Ⅱ2 eⅢ kerogen dominates the organic matter of the C3t- P1s shale, which indicates that the type of hydrocarbon generated will be mainly gas.

279

The degree of thermal evolution of organics is the primary factor in determining the generation of oil and/or gas from organic-rich shale and also affects the hydrocarbon generation potential and the adsorption capability (Nie et al., 2009; Hui and Stephen, 2013). The organic matter must be buried at depths with temperatures suitable for conversion to hydrocarbons. The Rock-Eval analysis precisely simulated this hydrocarbon-generating process by cracking the kerogen with heat. According to the Rock-Eval parameter (Tmax) from 151 shale samples, the temperature of maximum development of the S2 peak (Tmax) ranges from 450  C to 530  C, indicating a moderate-high mature evolution stage (Fig. 5a). In contrast, the Tmax of the main gas-bearing shales in American (Barnett, New Albany and Lewis shale) ranges from 430  C to 530  C, indicating that the organic matter was mature and was seldom higher than at the mature stage. Furthermore, the vitrinite reflectance test results of the C3t-P1s shale from different areas are dramatically different and range widely from 0.65% to 1.80% (on average 1.12%), indicating high maturity with a strong capacity for shale gas generation (Table 1).

4.1.2. Lithology and mineralogy characteristics Detailed observation of cores, outcrop samples and thin section analysis identified three main lithofacies in the C3t- P1s shale: silty shale, carbonaceous shale, and clay-rich shale. Silty shale is darkgray and laminate. The light laminas are mainly quartz and

Table 1 The mineral composition (wt.%), TOC (wt.%) and Ro (%) of the C3t and the P1s shale. (In the table, K ¼ kaolinite, C ¼ chlorite, I ¼ illite, I/S ¼ Illite smectite mixed layer indicating the proportion in the total clay mineral. PF ¼ potash feldspar, PL ¼ plagioclase, Cal ¼ calcite, Dol ¼ dolomite, Py ¼ pyrite and Sid ¼ siderite). Well

Depth/m

K/%

C/%

I/%

I/S/%

Clay/wt.%

Quartz/wt.%

PF/wt.%

PL/wt.%

CaL/wt.%

Dol/wt.%

Py/wt.%

Sid/wt.%

TOC/wt.%

Ro/%

LG2 LG2 T2905 T2905 T2905 T2905 T2905 T2905 T2905 T2905 T2905 T2905 T3 T3 G5 G5 G5 G5 S50 S50 S50 S50 S50 S24 S24 S24 S24 S24 S24 S24 S24 S24 S24 outcrop1-1 outcrop1-2 outcrop1-3 outcrop1-4 outcrop2-1 outcrop2-2 outcrop2-3

957.1 969.8 1290.00 1364.00 1376.00 1358.00 1380.00 1370.00 1402.00 1438.00 1450.00 1455.00 2124.60 2127.00 2934.20 2935.70 2940.70 2992.73 4309.9 4753.65 4755.25 4756.85 4758.45 2327.04 2329.64 2330.64 2331.64 2332.64 2333.13 2333.88 2334.41 2350.64 2351.14

/ / / / / / 13.2 19.0 / / / / / / 66 62 72 12 42 37 48 43 31 36 28 35 38 42 59 44 58 26 31 48 35 97 50 40 31 56

/ / / / / / 8.2 9.9 / / / / / / 24 12 20 5 12 17 22 18 15 / 10 6 18 24 20 18 17 11 15 10 16 3 19 13 5 14

/ / / / / / 22.9 32.2 / / / / / / 5 / / 15 3 15 16 6 16 26 33 7 / 2 / 2 2 5 13 19 15 0 14 20 45 3

/ / / / / / 55.7 38.9 / / / / / / 5 26 8 68 43 31 14 33 38 38 29 53 44 32 21 35 23 58 41 23 34 0 17 27 19 27

37.8 44.4 20.7 34.6 32.0 40.2 25.5 29.7 31.9 35.0 30.1 42.0 41.2 35.7 76.6 44.4 85.1 39.8 71 71.2 74.0 72.3 59.3 41.4 68.4 47.7 67.7 69.2 66.4 68 59.4 55.8 54 76.07 44.15 45.39 67.28 77.95 51.60 50.49

51.1 52.4 71.3 62.9 35.8 45.9 48.1 68.4 52.3 55.2 55.0 48.6 49.2 46.8 23.4 9.4 14.9 34.9 28.2 24.9 23.7 23.1 28.9 34.8 28.2 21 32.3 30.8 33.6 32 40.6 37.1 41.6 17.62 48.73 45.48 30.83 13.60 38.98 42.38

/ / / / / / / / / / / / / / / / / / / / 2.3 3 8 11.7 0.5 / / / / / / 5.8 2 3.77 4.64 0.00 0.00 1.54 1.56 1.30

7.1 / 5.0 / / / 19.2 / 8.8 4.0 4.4 3.0 3.0 3.4 / / / 6 0.8 3.9 / 1.6 3.8 7.6 1 2.9 / / / / / 1.3 2.4 0.00 1.74 7.99 0.00 0.00 1.27 2.01

2.0 / / / / / / / / / / / / / / / / / / / / / / / / 7.7 / / / / / / / / / / / / / /

/ 1.6 / / / 4.6 5.2

/ / / / / 9.3 2.0 1.9 / 1.8 10.5 2.5 / 8.0 / / / / / / / / / / / / / / / / / / / 2.55 0.00 0.00 1.89 6.19 3.46 0.00

2.0 1.6 3.0 2.5 32.3 / / / / 4.0 / 3.9 4.6 6.0 / 46.2 / / / / / / / 4.5 1.9 / / / / / / / / 0.00 0.73 1.14 0.00 0.73 3.13 3.82

1.75 0.17 / / / 2.96 / 1.82 / / / / 2.30 1.91 3.26 1.82 2.09 4.10 0.60 3.03 0.50 0.38 1.08 / / / / / / / / / / 11.76 0.65 1.26 3.92 12.5 2.79 2.19

1.60 1.25 / / / 1.80 / 1.75 / / / / 1.66 1.71 0.70 0.65 0.73 0.66 0.80 0.76 0.81 0.76 0.91 / / / / / / / / / / 0.85 1.38 1.42 1.31 1.27 0.84 0.87

7.0 / / / / / / / / 19.3 / / / / / / / 11.9 / / / / / / / / / / / / / /

280

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 6. Typical lithology pictures of dark C3t and P1s shale in the study area. (a) Dark -gray silty shale showing vertical fractures without filling material from the Shanxi Formation in Xiayunling (XYL), Mentougou District. (b) Grayish black carbonaceous shale from the Taiyuan Formation in Yuzhou (YZ), Henan District. (c) Back clay shale showing X-type conjugated fractures from the Shanxi Formation in Xiayunling (XYL), Mentougou District. (d) The S50 well in the Jizhong Depression, 4578 m, Taiyuan Formation, gray silty shale, showing high-angle fractures without filling minerals. (e) The T2905 well in the Liaohe Depression, 1377 m, the Shanxi Formation, grayish black carbonaceous shale showing the isolated plant carbon cuttings. (f) The T3 well in the Liaohe Depression, 2173.73 m, Taiyuan Formation, gray-black clay shale, showing slip fractures that had a smooth fracture plane and displayed mirror properties. (g) Gray silty shale with thin organic matter lamina, 2 magnification, showing microfractures that developed in organic matter lamina from the Shanxi Formation in Xiayunling (XYL), Mentougou District. (h) Gray carbonaceous shale with small silt, 25 magnifications, showing two sets of fractures formed in different periods that cut through one another with filling quartz, from the Taiyuan Formation in Quyang (QY), the Shanxi District. (i) Gray-black clay shale, 10 magnifications, showing small and partly disseminated (speckles) quartz particles from the Shanxi Formation in Xiayunling (XYL), Mentougou District.

feldspar, whereas the dark laminas are mainly clays that are rich in organic matter. Debris particles account for 25%e45% of the particles floating among the clay minerals, which have a lower TOC (TOC < 2.0%) (Fig. 6a, d, g). Carbonaceous shale is dark and laminate or massive. This lithofacies has a very high TOC (TOC > 6.0%). Banded or conferva carbon cutting is floating among clay minerals, indicating that this lithoface formed in a peat swamp (Fig. 6b, e, h). The clay shale is grayish black and laminate or massive with a high TOC (TOC > 2.0%). Clay minerals predominate, with few particles of clay/silt-size terrigenous quartz and feldspar floating among the clay minerals (Fig. 6i). The clay shale was deposited in a deep-water lagoon with weak hydrodynamic conditions and limited terrigenous importation (Fig. 6c, f). Therefore, the dark carbonaceous shale and grayish black clay shale are the most important organicrich shale lithologies in the study area. Compared to the organicrich shale lithology of the marine Longmaxi Formation (Wang et al., 2014) in southern China, the lithology type of the transitional shale in the study area is not as diverse and complex. The X-ray diffraction patterns of 40 shale samples of cores from six wells and two outcrops indicate that the C3t-P1s shale is dominated by clay minerals (51%) and clastic components (mainly quartz and feldspar, 47%), as well as subordinate plagioclase, potassium feldspar, calcite, dolomite and pyrite (Table 1). The brittle mineral (e.g., quartz, feldspar and pyrite) content of the C3t shale is 14.9%e69.9%, with an average of 45.4%. Among the brittle minerals,

the quartz content (9.2%e55.4%, with an average of 34.6%) is higher than the content of the other minerals (10%) and lower than that of the Barnett and Woodford shales in North America (Fig. 7a). In contrast, the brittle mineral content of the P1s shale is much higher and mainly ranges from 21.0% to 79.3%, with an average of 48.9%, among which the quartz content is 21.0%e71.3%, with an average of 42.7%. Although the brittle mineral content of the C3t and the P1s shale has met the minimum standards of fracturing feasibility (40%) (Rickman et al., 2008), the clay mineral content in both is relatively high, averaging 52.8% and 51.0%, respectively, which is much higher than that in the North American marine exploitable shale (Fig. 7a, Table 1). Clay minerals are mainly illite-smectite mixed layer mineral (on average 31%) and kaolinite (on average 42%) with a certain amount of illite (15%) and fewer chlorites (13%), which is different from the clay mineral composition of the Barnett shale (Fig. 7b). Additionally, the clay mineral composition distinguishes the C3t shale from the P1s shale. The kaolinite mineral content of the C3t shale (on average 45.8%) is much higher than that of the P1s shale (on average 40.7%), whereas the illite-smectite mixed layer and illite mineral content (on average 31.8% and 14.7%, respectively) of the P1s shale is higher than that of the C3t shale (on average 26.1% and13.2%, respectively). Additionally, the chlorite mineral content in both shales is lower than 13% (Fig. 7a, Table 1). Therefore, compared to the marine gas-bearing shales in North America (including the Barnett, Marcellus, and Eagleford shales), the brittle

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 7. Ternary plot of the composition of X-ray minerals (a) and clay minerals (b) in the C3t and the P1s shale (North American marine shale data provided by Ross and Bustin (2008), Gupta et al. (2013), Li and Shen, (2011)).

mineral content of the C3t and P1s shales is relatively low, with a general shortage of carbonate minerals and more clay minerals (Fig. 7a and Table 1).

4.2. Log response characteristics and identification of organic-rich shale 4.2.1. Log response characteristics of organic-rich shale and its lithology identification The analysis of the characteristic log responses in shale that is rich in organic matter (OM) is the most reliable way of qualitatively identifying and quantitatively evaluating organic-rich shale. Organic-rich shale mainly consists of three parts: rock skeleton (including clastics and clay minerals), organic matter (including kerogen, bitumen, pyrobitumen, and immovable carbon), and pore fluid (including oil, gas and water). The physical characteristics of kerogen are different from those of other minerals in the rock and include non-conductivity, lower sonic transit capacity, lower density and higher adsorption capacity. Therefore, when considering organic-rich shale, especially relatively mature gas-bearing shale, the log response characteristics will also differ from those of ordinary shale and non-shale lithologies. Generally, the higher the OM content is, the more obvious this abnormal log response is. The response characteristics of organic-rich shale from conventional log well curves are as follows:

281

4.2.1.1. Lithology indicator logging suite. This logging suite consists of the spontaneous potential (SP), the borehole diameter (CAL), natural gamma rays (GR), and the photoelectric absorption index (PE). The SP log reacts to the variation in salinity formation and the wellbore. Based on the analysis of its relative variation, we can effectively differentiate the sandstone from the shale layers, and calculate the shaliness of the formation. However, the analysis has a weak capacity for distinguishing other lithologies (such as organicrich shale, coal and limestone). Considering marineecontinental transitional shale (hereafter referred to as “transitional shale”), compared to marine shale, the lithology variation is much more complex. In a vertical profile, the organic-rich shale is usually intercalated with coal beds, limestone and tight sandstone (Fig. 8). Thus, as a reference for dividing the organic-rich shale layers, a high OM content corresponds to high shaliness intensity, whereas CAL is non-sensitive to OM content variation. By contrast, GR and DEN are more effective for distinguishing lithologies and dividing organic-rich shale layers (Fig. 9a). There are large differences in the log values between the different types of lithologies. Usually, the GR intensity value of mudstone shale (>100 API) is much higher than that of other lithologies and will increase with an increase in the OM content, with a maximum value of 200 API. Additionally, the GR intensity value is relatively low compared to the marine shales of North America and southern China marine shales. The GR intensity value of the sandstone and the coal layers (on average 75.8 API and 56.2 API, respectively) is much lower, and the GR intensity value of limestone is the lowest (on average 34.7 API). The photoelectric absorption index of the rocks obtained from lithology-density logs can better discriminate the typical lithology because the fluid properties and the content in the rocks have little influence on the PE value. A DEN-PE plot ideally identifies lithology in conjunction with other logs or plots (Fig. 9b). Additionally, PE logging is very sensitive to the TOC variation in shale. Generally, as the TOC decreases, the PE logging value of shale increases sharply. As shown in Fig. 8, the TOC of the Shanxi formation shale decreases from 7.76% to 0.78% in the interval from 2342 m to 2345 m of the S2 well in the Wenan Slope, Jizhong Depression, whereas the PE logging value increases from 0.98 b/e to 2.99 b/e. 4.2.1.2. Tri-porosity logging suite. This suite includes acoustic (AC), neutron (CNL), and density (DEN) logs. The tri-porosity log response is mainly influenced by the lithology and pore fluid. Therefore, these three types of logging curves, separately or combined with other logging curves, can be cross-plotted to identify a typical lithology (Fig. 9c, d). Generally, the acoustic logging value will sharply increase with an increase in the TOC because of the high sonic transit time (approximately 524.9 ms/m) of the OM. Therefore, the acoustic logging value of a coal bed with abundant OM is the highest (on average 412.3 ms/m), followed by the acoustic logging values of carbonaceous shale (on average 289.2 ms/m) and the ordinary dark shale (on average 241.4 ms/m), which is slightly higher than the acoustic logging values of the marine Longmaxi Formation shale in southern China (Yan et al., 2014). CNL logs have a similar response as the AC logs when the OM content is increased. However, the DEN log response to an increase in OM content is opposite to the responses of the CNL and the AC logs. Because of the lower density (1.1e1.4 g/cm3) of OM compared to quartz (2.65 g/ cm3), calcite (2.71 g/cm3) and clay (approximately 2.77 g/cm3), the effects on the corresponding logs from dark shale are evident. In particular, as the organic-rich shale matures, some of the solid organic matter is transformed to liquid or gaseous hydrocarbons that move into the pore space, displacing the formation water in response to the DEN log. Therefore, organic-rich shale can be easily identified by the DEN log, even in a favorable gas-bearing shale reservoir.

282

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 8. Characteristic log response of transitional organic-rich shale in C3t and P1s from the S2 well, Jizhong Depression.

4.2.1.3. Resistivity logging. OM is an electrically resistive substance (near 105e109 U m) and increases the resistance of the host rock compared to an identical rock that is devoid of OM. In a particular dark shale formation, an increase in resistance (with depth for instance) can be interpreted as an increase in OM. Organic-rich shale resistance depends not only on its kerogen (solid insoluble HC) content but also on its HC content in the pore space, the hydrothermal alteration and the development degree of fractures in shale. Therefore, the resistivity log must be combined with other log suites (such as the AC log, known as “DlogR”) to effectively differentiate organic-rich shale layers. 4.2.1.4. Radioactivity logging. The main natural contributors to

radioactivity in rocks are potassium (K), uranium (U) and thorium (TH), which can be measured by a spectral gamma ray (SGR). An SGR can provide more detailed information than a total GR log. Generally, clay largely contributes to the radiation properties. Moreover, different clay minerals have different contribution ratios because of the variation of cation exchange capacity, surface and radioactivity in different types of clay minerals. The montmorillonite contributes the most to the shale's radiation content, followed by illite, whereas chlorite and kaolinite have only a slight contribution to radioactivity. Therefore, from the content analysis and ratios of the three radioactive elements (U, Th and K), we can qualitatively and quantitatively evaluate the types and relative content of clay minerals in shale formations. The radioactivity

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

283

Fig. 9. Cross-plots for typical lithology recognition in C3t and P1s from the S2 well.

logging value of Th and K in the C3t shale (2430 me2436 m) and the P1s shale (2340 me2356 m) intervals of well S2 are used to plot a correlation with a linear scale for Th and K (Fig. 10). As shown in Fig. 10, mixed illite-smectite and kaolinite mainly dominate the types of clay minerals, and only a trace amount of smectite appears in the TheK cross-plot (Fig. 10a). Additionally, the kaolinite mineral content of the C3t shale is much higher than the P1s shale (Fig. 10b), consistent with the results of the X-ray diffraction results of 40 C3tP1s shale samples (Fig. 7b and Table 1). In addition to the clay minerals having strong absorptivity for radioactive elements, these radioactive components are also often linked to OM in sedimentary rocks. Concerning the radioactivity logging response to organicrich shale, there are some differences between these three radioactive element logs. With an increase in the TOC, the U content increases significantly, and the K content increases slightly. However, the Th content decreases in the C3t shale (2450 me2460 m) and the P1s shale (2340 me2347 m) of the S2 well. Overall, the logging response characteristics of transitional organic-rich shale from conventional well logs are similar to marine organic-rich shale. The characteristics feature a typical log response with high GR, high AC, relatively high RT, high CNL, high U and K, low DEN, low PE and a low Th/U ratio. The absolute values of the general logging responses are different from those of marine organic-rich shale. Additionally, the gas logging value (total hydrocarbon) of the organic-rich shale layers is often higher than the ordinary dark shale (Fig. 8).

4.2.2. Logging identification results of organic-rich shale The CNL, DEN, AC, GR and radioactivity logging suite is sensitive to TOC variation. Therefore, this logging parameter suite can be

optimized and combined to identify organic-rich shale layers and to ascertain the thickness distribution of transitional organic-rich shale in C3t and P1s. The organic-rich shale interval can be recognized by multiple overlays of the CNL-GR, DEN-CNL, DEN-PE, UeTh, and AC-RT logs at well depths ranging from 2556 m to 2644 m (Fig. 11). For example, with the DEN-PE overlay logs, the DEN log is more sensitive to the TOC variation and can be selected as the base curve. If the PE log decreases to the left of the baseline (“positive variance”), the larger spacing amplitude of “positive variance” intervals will be interpreted as organic-rich shale layers (TOC >2.0%). Additionally, the larger the overlay area of DEN-PE is, the higher the TOC of the shale layer will appear. If the PE log increases to the right of the baseline (“negative variance”), this interval is defined as organic-lean shale or a non-shale layer. Combined with the results of other overlay logs, the transitional shale profile in the entire interval of 2556 me2644 m can be differentiated into specific organic-rich shale zones. As shown in Fig. 11, the vertical anisotropy of the transitional shale is much stronger when the layers alternate between organic-rich shale and organic-lean shale. The identified organic-rich shale amalgamates in the lower member of the Shanxi formation, with a thickness of 28 m. The logs described above that are sensitive to organic-rich shale can also be used for cross-plot analyses. The logging data corresponding to the TOC experimental data of the C3t- P1s shale intervals in the S1 and the S2 wells can be used to conduct cross-plot analysis for the semi-quantitative determination of the logging response values of organic-rich shale (TOC > 2.0%). As shown in Fig. 12, the radioactivity and DEN logs can best distinguish organicrich shale from organic-lean shale (Fig. 12c and Fig. 12d), followed

284

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 10. Recognition cross-plots (a) and quantitative calculation plot (b) for clay mineral types in the C3t and the P1s shale intervals.

by the CNL and the PE logs (Fig. 12c and Fig. 12d), whereas the shale can't be fully distinguished by the total GR log (Fig. 12a). The logging response threshold of organic-rich shale from a conventional well log series is different: GR
120 API; U
10 mg/g; CNL
22%; DEN2.4 g/cm3; PE  2.3 b/e; TH/U  1. Moreover, the GR and U logging thresholds of transitional organic-rich shale are lower than that of the Barnett shale in North America and the Longmaxi Formation marine shale in southern China, which have characteristic values of GR
150 API, U
14 mg/g, and DEN  2.6 g/cm3 (Lewis et al., 2004; Yan et al., 2014), and the DEN threshold is much higher. This result is consistent with the unique logging response

characteristics of transitional organic-rich shale. Additionally, the log suite parameters from the C3t and the P1s shale intervals in well LG2 of the Liaohe Depression and wells S2 and S1of the Jizhong Depression, corresponding to the 39 samples with measured TOC data, were used to fit the three types of functions (Equations (1)e(3)). As shown in Fig. 13a, there is a strong correlation between laboratory- and log-derived TOC from the DEN log, followed by the AC and the RT logs. However, the correlation between the measured TOC and the GR and the CNL logs is relatively poor. Additionally, multivariate linear regression (MLR), which was used to determine the TOC, was more effective than the

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 11. Multi-parameter logging overlay for organic-rich shale recognition in C3t and P1s from the S1 well, the Jizhong Depression.

285

286

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 12. Cross-plots for transitional organic-rich shale recognition in C3t and P1s from the S1 and the S2 wells.

modified DlogR technique and simple linear fitting. Taking the logcalculated TOC results from 800 m to 1200 m in well LG2, as an example (Fig. 14), the DlogR technique is not as useful to determine the TOC compared to the MLR method, whereas it has a greater advantage for calculating the TOC in a coal bed. The logRT and AC logging values can be optimized to make cross-plots to quantitatively evaluate the shale TOC (Fig. 13b and AeB shale intervals in Fig. 14). Therefore, it is better to combine the MLR method with a modified DlogR technique to quantitatively evaluate the TOC. The heterogeneity of TOC variation is large; the identified organic-rich shale converges in the middle-upper member of the Taiyuan Formation and the lower member of the Shanxi Formation, consistent with the TOC test results and practical gas logging. 4.3. Distribution characteristics of organic-rich shale 4.3.1. Vertical distribution of organic-rich shale Using the described qualitative and quantitative evaluation methods of organic-rich shale, we can identify the organic-rich shale intervals of C3t- P1s and calculate the characteristic values of the organic-rich shale thickness in the selected key wells to analyze the vertical profile and distribution characteristics of transitional organic-rich shale. The transitional organic-rich shale in the vertical profile comprises four types of structure. 4.3.1.1. Continuous shale profile (Type I). This type of shale interval in the C3t is mainly beneath the lagoon facies. The lithology is dominated by dark gray or black mudstone shale, with a mudstone ratio that is greater than or equal to 0.95. The shape of the GR logging curve is a rounded peak, with values of 120 API to 150 API. The AC becomes larger and the curves increase to the left of the

baseline, whereas the RT curve is stable. The overlay area of DlogR in the shale interval is much larger. The log-derived TOC ranges from 1.5% to 3.5%. In this profile structure, the single layer of organic-rich shale is the thickest (30 me50 m) and has the best continuity (Fig. 15a). 4.3.1.2. Para-continuous shale profile (Type II). This type of shale layer is mainly formed in deep reed swamp facies, which develop in the C3t and the lower member of the P1s. The layer is mainly characterized by gray-black mudstone or carbonaceous shale, intercalated with thin-bed, fine-grained sandstone and thick coal beds. The mudstone ratio of this profile ranges from 0.75 to 0.95, and has a smooth, box-shaped curve in the GR log, with values ranging from 130 API to 160 API. The RT values increase and the curves increase to the right of the baseline. In contrast, an AC increase is not obvious. The overlay area of DlogR in the shale interval is larger. The log-derived TOC ranges from 1.0% to 5.5%. Additionally, in this profile, the single layer thickness of the organic-rich shale is usually 15 m-25 m; however, the continuous cumulative thickness of the organic-rich shale is much larger, with a thickness of 30 me80 m (Fig. 15b). 4.3.1.3. Intermediate shale profile (Type Ⅲ). This type of shale interval occurs mainly beneath the tidal flat facies and is distributed on the C3t. The shale is characterized by gray black mudstone shale, intercalated with some sets of thin-bed fine-grained stone, siltstone or coal beds, with a mudstone ratio ranging from 0.6 to 0.75. The shape of the GR curve is a tooth-like, box-shape with log values varying between 90 API and 150 API. The AC increases and the curves increase to the left of baseline, whereas an increase in RT is not obvious. The overlay area of DlogR in the shale interval is smaller than those of the above two types of shale profiles. The log-

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

287

thickness of a single organic-rich shale interval (indicating that the single shale layer (TOC > 2.0%)), the much larger interbed thickness relative to other adjacent lithology interbeds (thickness of <3 m), the total thickness of interbeds (accounting for less than 40% of shale intervals) and the continuous cumulative thickness of organic-rich shale (indicating that a single organic-rich shale interval thickness can only be added to cumulative thickness when it is more than 15 m) have shown that the total thickness of dark shale is much larger (
50 m) and that the continuous thickness of a single organic-rich shale interval is smaller (generally 25 m), with a larger continuous cumulative thickness(generally
35 m). Additionally, the continuous thickness of a single organic-rich shale interval in the P1s ranges from 1.2 m to 34.4 m, and the continuous cumulative thickness varies between 15 m and 35 m, with a maximum thickness of 83.7 m (Fig. 16a, Table 2). In contrast, the continuous thickness of a single organic-rich shale interval in the C3t is much larger than that of the P1s, which ranges from 1.5 m to 54.9 m, whereas the continuous cumulative thickness is 25 me45 m, with a maximum thickness of 85.3 m (Fig. 16b, Table 2). With respect to the organic-rich shale thickness, C3t may be more favorable for shale gas exploration than P1s with an opportune burial depth (<3500 m). However, both are thinner than the marine organic-rich shale in North America and southern China (Fig. 16c). Additionally, the transitional shale in the vertical profile, intercalated with thin-bed coal beds and sandstone (<3 m), can form coal-bearing strata with a thickness of >30 m so that shale gas, coal bed methane and tight sandstone gas reservoirs are adjacent, which cannot be neglected when considering areas for favorable gas exploration and development potential.

Fig. 13. (a) Correlation between log- and laboratory-derived TOC from different interpretation techniques. (b) logRt-AC plot for quantitatively evaluating the TOC of C3t shale in the LG2 well, eastern salient of the Liaohe Depression.

derived TOC ranges from 1.0% to 5.5%. Therefore, in this profile structure, the single layer thickness of organic-rich shale is smaller (10 me20 m); the continuous cumulative thickness of organic-rich shale can reach 30 me40 m (Fig. 15b). 4.3.1.4. Interbedded shale profile (Type Ⅳ). This type of shale layer in the P1s occurs mainly under the delta front-coastal marsh facies. The shale is characterized by bedded gray-black mudstone and carbonaceous shale, frequently interbedded with thin-bed, finegrained stone, muddy siltstone or coal beds, with a mudstone ratio of 0.4e0.6. It has a tooth-like dumbbell-shaped curve in the GR log, the values of which range from 50 API to120 API. The RT stabilizes, and an AC increase is not obvious. Therefore, the overlay area of DlogR in the shale interval is the smallest among these types of profile structures. The log-derived TOC ranges from 0.5% to 3.5%. Furthermore, in this profile structure, the single layer thickness of organic-rich shale is the smallest (only 10 me15 m); the continuous cumulative thickness of organic-rich shale is 15 me35 m (Fig. 15b). Among these four types of shale profile structures, the single organic-rich shale layer thickness and the continuous cumulative thickness of types Ⅰand Ⅱ are relatively large, with better continuity in the vertical profile, which may be the most favorable structural element type for shale gas exploration. Statistical studies on the characteristic values of organic-rich shale thickness in the C3t and the P1s from 16 wells in five depressions in the Bohai Bay Basin (Table 2), including the total thickness of dark shale, the continuous

4.3.2. Lateral distribution of organic-rich shale The sedimentary environment of transitional organic-rich shale is less stable than that of marine organic-rich shale. Therefore, in the lateral view, the continuity is much poorer with significant variation. Based on the above identification methods for dividing several organic-rich shale intervals and introducing coal seams as a marked bed, we can develop an organic-rich shale cross-well profile. The profile is oriented from south to north, from well D7 to G2 in the Dacheng Region, Jizhong Depression, showing a facies stacking pattern, lateral trends, thickness variation of the coal bed and organic-rich shale. As shown in Fig. 17, in the cross-well profile, the center of the organic-rich shale is much thinner than the edges, indicating that the closer it is to the Cangxian Uplift, the thinner the organic-rich shale is. Additionally, the continuous cumulative thickness of the organic-rich shale reaches a maximum in the D9 and SH1 wells, with respective thicknesses of 98 m and 87 m. Generally, the organic-rich shale of the C3t is much thicker than that of the P1s. Additionally, the middle-upper member of the Taiyuan Formation and the lower member of the Shanxi Formation have a better continuity in lateral profile. In contrast, both are less stable and less continuous than the distribution of marine organicrich shale. 5. Conclusions (1) The organic-rich shale lithology of the Upper Paleozoic Shanxi Formation and the Taiyuan Formation is dominated by gray-black laminate clay shale and carbonaceous shale beneath deep-water lagoon and deep reed, swamp facies. The minerals are dominated by clay, quartz and feldspar. The brittle mineral content of the P1s shale (48.9%) is much higher than that of the C3t shale (45.4%). However, both have relatively high clay mineral contents, averaging 52.8% and 51.0%, respectively, with the clay type overwhelmingly kaolinite and mixed-layer I/S. The C3t- P1s shale has a high

288

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 14. Quantitative interpretation of the TOC of the C3t and P1s in the LG2 well of Liaohe Depression.

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Fig. 15. The types of shale profile structures in the C3t and P1s, Bohai Bay Basin.

289

290

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

Table 2 Statistical results of characteristic values of organic-rich shale in C3t and P1s from five depressions, Bohai Bay Basin (1900e2178: initial depth-ending depth; 2e23.36: minimumemaximum value). Depression name

Well

TaiyuaneShanxi formation Formation Total thickness of burial depth/m dark shale/m

Liaohe T3 1900e2178 Depression T2905 1256.34e1470 LG2

840e1154

WC1

1265.4e1482.2

Jizhong DC1 Depression S1

1093e1332.5 2126e2387

S18

3016.5e3194.5

G2

3128e3429.5

Huanghua WG1 3990e4318 Depression K71 2100e2390 KG6

2390e2670

KG4

3440e3735

Jiyang QG3 Depression Dongpu Sag MG5 QG2

3980e4280 3158e3360 3995e4320

WG2 4320e4460

P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s C3t P1 s P1 s C3t P1 s C3t

59.33 75 41 35 75 88 50.8 45.3 48.4 35 70.5 87.5 55.8 52.5 69 101.5 110.8 83.7 41.5 100.5 39.9 110.5 39.9 105.6 50.5 98.2 85.4 76.2 112.3 46.1 47

TOC (most 2%e4%), with type Ⅱ2-Ⅲ kerogens and high thermal maturation (Ro
1.1%), which is favorable for generating shale gas. (2) Transitional organic-rich shale on conventional well log curves has distinct log response characteristics, similar to those of marine organic-rich shale. Additionally, there is a typical log response with high GR, high AC, relatively high RT,

Continuous thickness of single organic- Continuous cumulative thickness of rich shale interval/m organic-rich shale (>15 m)/m 2e23.36 2e35.1 8.83e27 6.07e21.7 6.0e26.4 5.0e20.8 4.9e17.4 9.1e26.2 2.0e15.7 2.5e15.5 3.5e24 7.1e17 3.5e18 3.0e25 2.5e18 1.5e25.4 3.0e32.9 4.0e22.9 2.7e17 1.6e54.9 2.5e19.5 4e38 1.2e18.5 2.0e26.5 1.8e34.4 7e29.6 3.1e20 3.8e21.5 5.3e33.1 1.7e23.5 3.5e27

23.36 35.1 27 21.7 44.39 37.5 17.4 26.2 30 15.5 48.5 33.5 18 40 18 44.5 83.7 61.8 17 72 19.5 76 18.5 66.8 34.4 45.9 40 40 85.3 23.5 27

high CNL, high U and K, low DEN, low PE and low Th/U ratio, and the logging response value differs from that of marine organic-rich shale. (3) The logging parameters that are the most sensitive to TOC variation in shale (such as radioactivity, tri-porosity, and PE logs) can be optimized to construct multi-parameter overlay logs for differentiating organic-rich shale and organic-lean

Fig. 16. Histogram of organic-rich shale thickness values ((a), (b)) and comparison with the main marine shale in North America (c).

Fig. 17. Cross-section of organic-rich shale of the C3t and P1s from wells D7 to G2 along a north/south direction in the Dacheng Region of the Jizhong Depression.

292

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293

shale in the vertical profile. Additionally, these logging parameters can be selected to construct tri-porosity-GR, DENPE, and Th/UeU plots to identify shale lithologies and organic-rich shale. Moreover, by combining modified DlogR technology with multivariate linear regression, we can quantitatively evaluate each well for TOC in the C3t- P1s shale. The heterogeneity of the TOC variation in the vertical profile is very strong, and the identified organic-rich shale intervals mainly converge in the middle-upper member of the Taiyuan Formation and the lower member of the Shanxi Formation, consistent with the TOC and practical gas logging results. (4) The characteristic values of organic-rich shale thickness indicate that the continuous thickness of single organic-rich shale intervals in the C3t and the P1s is smaller (generally 25 m) and the continuous cumulative thickness is larger (generally
35 m). Additionally, the organic-rich shale thickness of C3t is much greater than that of P1s, which may be favorable for shale gas exploration. The single shale interval thickness and continuous cumulative organic-rich shale thickness of para-continuous or continuous shale profiles beneath the lagoon and the deep reed swamp facies are much greater than other structural profile types. However, in the lateral profile, the distribution of transitional shale is relatively less stable and less continuous than that of marine shale. Acknowledgments The study was sponsored jointly by the National Natural Science Foundation Project (41372139, 41072098, and 41002072), Major Special Project for National Science and Technology “Evaluation on Shale Gas Resources in Bohai Bay Basin” (2011ZX05018-001-002) and “Geological characteristics and evaluation of reservoir in new oil and gas exploration field ” (2011ZX05009-002-205, 2016ZX050406-003 and 2011ZX05033-004). We are grateful to the Geochemistry Laboratory of Yangtze University and the CNNC Beijing Research Institute of Uranium Geology, which helped to test and analyze samples. We are also grateful to the anonymous reviews and editors, whose comments improved the quality of my manuscript. References Bao, S.J., Ren, S.M., Gao, B.J., et al., 2014. The progress and challenges of shale gas exploration and development in China. Shale Gas. Trends 20 (26), 5e8. Bowman, T., 2010. Direct method for determining organic shale potential from porosity and resistivity logs to identify possible resource play. In: Genesis of Shale Gas-Physicochemical and Geochemical Constrains Affecting Methane Adsorption and Desorption: AAPG Annual Convention, New Orleans, LA, April 11e14. Conti, J.J., Holbeg, P.D., Diefenderfer, J.R., 2014. Annual Energy Outlook 2014 with Projections to 2040. Energy Information Administration, Washington, pp. 10e15. Curtis, J.B., 2002. Fractured shale-gas systems. AAPG Bull. 86, 1921e1938. Dariusz, S., Maria, M., Arndt, S., et al., 2010. Geochemical constraints on the origin and volume of gas in the New Albany Shale (Devonian-Mississippian), eastern Illinois Basin. AAPG Bull. 94, 1713e1746. Flower, J.G., 1983. Use of sonic-shear-wave/resistivity overlay as a quick-look method for identifying potential pay zones in the Onio(Devonian) Shale: paper SPE-10368. J. Pet. Technol. 638e642. Gatens, J.I., Harrison, C.W., Lancaster, D.E., Guidry, F.K., 1990. In-situ stress tests and acoustic logs determine mechanical properties and stress profiles in the Devonian shale: paper SPE-18523-PA. SPE Form. Eval. 5, 248e254. Ge, M.N., Zhang, J.C., Mao, J.L., et al., 2012. Evaluation on Neopaleozoic shale gas resource potential in the eastern salient of the Liaohe depression. Nat. Gas. Ind. 32 (9), 28e32. Guo, T.L., Zhang, H.R., 2014. Formation and enrichment mode of Jiaoshiba shale gas field, Sichuan Basin. Pet. Explor. Dev. 41 (1), 28e36. Gupta, N., Rai, C.S., Sondergeld, C.H., 2013. Integrated petrophysical characterization of the Woodford shale in Oklahama. Petrophysics 54 (4), 1e35.

Hill, R.J., Jarvie, D.M., Zumberge, J., Henry, M., Pollastro, R.M., 2007. Oil and gas geochemistry and petroleum systems of the Fort Worth Basin. AAPG Bull. 91, 445e473. Hui, J., Stephen, A.S., 2013. Characterization for source rock potential of the Bakken Shales in the Williston Basin, North Dakota and Montana. In: Unconventional Resources Technology Conference Held in Denver, Colorado, USA, 12e14 August 2013. SPE 168788, pp. 1e10. Jacobi, D.J., Breig, J.J., LeCompte, B., et al., 2008. Effective geochemical and geomechanical characterization of shale gas reservoirs from the well bore environment: Caney and the Woodford Shale. In: 2009 SPE Annual Technical Conference and Exhibition Held in New Orleans, Louisiana, USA, 4e7 October 2009. SPE 124231, pp. 1e20. Jiang, T.G., Liu, S.B., Sun, X., 2014. Analysis on shale gas accumulation conditions of Longmaxi formation in the Northeastern Yunnan Province. Special Oil Gas Reserv. 21 (6), 23e27. Khoshnoodkia, M., Mohseni, H., Rahmani, O., et al., 2011. TOC determination of Gadvan Formation in South Pars Gas field, using artificial intelligent systems and geochemical data. J. Pet. Sci. Eng. 78 (1), 119e130. Kim, A.G., 1977. Estimating Methane Content of Bituminous Coalbeds from Adsorption Data. U. S: Dept. of the Interior, Bureau of Mines, pp. 1e11. LeCompte, B., 2010. Comprehensive Resource Play Evaluation for Well Completion Decisions- Mineralogy, Acoustic, and NMR Integration. Murphy Oil Inc, EI Dorado Arkansas. Lewis, R., Ingraham, D., Pearcy, M., et al., 2004. In: New Evaluation Techniques for Gas Shale Reservoirs. Reservoir Symposium 2004. Schlumberger, pp. 1e11. Li, Z., Shen, Z.M., 2011. Exploration Potential Analyses of the Shale Gas in Dongying Depression. The Master Degree Paper of Chengdu University, pp. 72e73. Loucks, R.G., Reed, R.M., Ruppel, S.C., et al., 2009. Morphology, genesis, and distribution of nanometer-scale pores in siliceous mudstones of the Mississippian Barnett shale. J Sediment. Res 79 (12), 848e861. Luffel, D.L., Guidry, F.K., 1989. Reservoir rock properties of Devonian shale from core and log analysis. In: Soc Core Analysis 3rd Annual Technical Conference, p. 13. Luffel, D.L., Guidry, F.K., Curtis, J.B., 1992. Evaluation of Devonian Shale with new core and log analysis method: paper SPE-21297. J. Pet. Technol. 44, 1192e1197. Ma, L., Chen, H.J., Gan, K.W., et al., 2004. Geostructure and Marine Facies Hydrocarbon Geology of South China. Geological Publishing House, Beijing. Martini, A.M., Walter, L.M., Ku, T.C.W., et al., 2003. Microbial production and modification of gases in sedimentary basins: a geochemical case study from a Devonian shale gas play, Michigan Basin. AAPG Bull. 87, 1355e1375. Montgomery, S.L., Jarvie, D.M., Bowker, K.A., et al., 2005. Mississippian Barnett Shale, Fort Worth basin, north-central Texas: Gas-shale play with multi-trillion cubic foot potential. AAPG Bull. 89, 155e175. Nie, H.K., Tang, X., Bian, R.K., 2009. Controlling factors for shale gas accumulation and prediction of potential development area in shale gas reservoir of South China. Acta Pet. Sin 30 (4), 484e491. Pan, R.F., Chen, L., Liu, P.C., 2011. Study on the evaluation of shale gas resource by a classified method. J. Oil Gas Technol. 33 (5), 172e174. Passey, Q.R., Creaney, S., Kulla, J.B., et al., 1990. A practical model for organic richness from porosity and resistivity logs. AAPG Bull. 74, 1777e1794. Pemper, R.R., Han, X.G., Mendez, F.E., et al., 2009. The direct measurement of carbon in wells containing oil and natural gas using a pulsed neutron mineralogy tool. In: 2009 SPE Annual Technical Conference and Exhibition Held in New Orleans, Louisiana, USA, 4e7 October 2009, SPE 124234, pp. 1e14. Pollastro, R.M., 2007. Total petroleum system assessment of undiscovered resources in the giant Barnett Shale continuous (unconventional) gas accumulation, Fort Worth Basin, Texas. AAPG Bull. 91, 551e578. Qi, J.F., Yang, Q., 2009. Cenozoic structural deformation and dynamic processes of the Bohai Bay basin province, China. Mar. Pet. Geol. 27 (4), 757e771. Rickman, R., Mullen, M.J., Petre, J.E., et al., 2008. A practical use of shale petrophysics for stimulation design optimization: all shale plays are not clones of the Barnett Shale. In: 2008 SPE Annual Technical Conference and Exhibition Held in Denver, Colorado, USA, 21e24 September 2008. SPE 115258, pp. 1e11. Ross, D.J., Bustin, R.M., 2008. Characterizing the shale gas re-source potential of Devonian-Mississippian strata in the Western Canada sedimentary basin: application of an inte-grated formation evaluation. AAPG Bull. 92, 87e125. Ross, D.J., Bustin, R.M., 2009. The importance of shale composition and pore structure upon gas storage potential of shale gas reservoirs. Mar. Pet. Geol. 26 (6), 916e927. Schmoker, J.W., 1979. Determination of organic content of Appalachian Devonian shales from formation-density logs. AAPG Bull. 63, 1504e1537. Schmoker, J.W., 1981. Determination of organic-matter content of Appalachian Devonian shales from gamma-ray logs. AAPG Bull. 65, 1285e1298. Schoell, M., 1980. The hydrogen and carbon isotopic composition of methane from natural gases of various origins. Geochim. Cosmochim. Acta 44 (5), 649e661. Soeder, D.J., 1986. Porosity and permeability of eastern Devonian gas shale: paper SPE-15213. Unconventional Gas Technology Symposium, Louisville, KY, SPE, pp. 75e81. Travis, J.K., Lance, W.C., John, A.B., et al., 2008. Hydrocarbon potential of the Barnett Shale (Mississippian), Delaware Basin, west Texas and southeastern New Mexico. AAPG Bull. 92, 967e991. Wang, Z.F., Zhang, Y.F., Liang, X.L., et al., 2014. Characteristics of shale lithofacies formed under different hydrodynamic conditions in the Wufeng-Longmaxi Formation, Sichuan Baisn. Acta Pet. Sin. 35 (4), 623e632. Wang, R.Y., Ding, W.L., Wang, Z., et al., 2015. Progress of geophysical well logging in shale gas reservoir evaluation. Prog. Geophys. Chin. 30 (1), 0228e0241.

J. He et al. / Marine and Petroleum Geology 70 (2016) 273e293 Yan, W., Wang, J.B., Liu, S., et al., 2014. Logging identification for the Longmaxi mud shale reservoir in the Jiaoshiba area, Sichuan Basin. Nat. Gas Ind 34 (6), 30e36. Zhang, J.C., Jin, Z.J., Yuan, M.S., 2004. Reservoiring mechanism of shale gas and its distribution. Nat. Gas. Ind. 24 (7), 15e18. Zhang, S.W., Sui, F.G., Lin, H.X., et al., 2009. Petroleum Geology and Perspective

293

Evaluation on Pre-Paleogene System in Bohai Bay Basin. Geological Publishing House, Beijing. Zhao, H., Givens, N.B., Curtis, B., 2007. Thermal maturity of the barnett shale determined from well-log analysis. AAPG Bull. 91, 535e549.