- Email: [email protected]
Block scale investigation on gas content of coalbed methane reservoirs in southern Qinshui basin with statistical model and visual map
Journal of Petroleum Science and Engineering 114 (2014) 1–14
Contents lists available at ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Block scale investigation on gas content of coalbed methane reservoirs in southern Qinshui basin with statistical model and visual map Huihu Liu a, Shuxun Sang b,n, Geoff G.X. Wang c, Mengxi Li d, Hongjie Xu a, Shiqi Liu b, Junjun Li e, Bo Ren f,g, Zhigen Zhao a, Yan Xie a a
School of Earth and Environment, Anhui University of Science & Technology, Huainan, Anhui 232001, PR China School of Resources and Earth Science, China University of Mining and Technology (CUMT), Key Laboratory of CBM Resources and Reservoir Formation Process, State Ministry of Education, Xuzhou, Jiangsu 221008, PR China c School of Chemical Engineering, The University of Queensland, Brisbane, St. Lucia Qld 4072, Australia d Shanxi Coalbed Methane Branch of Huabei Oilfield Company, Jincheng, Shanxi 048000, PR China e Production & Technology Department, Shanxi Lanyan Coalbed Methane Group Co., Ltd., Jincheng, Shanxi 048204, PR China f State Key Laboratory of Deep Coal Mining & Environment Protection, Huainan Mining Industry (Group) Co., Ltd., Huainan, Anhui 232001, PR China g National Engineering Research Center for Coal Gas Control, Anhui 232001, PR China, Huainan, Anhui 232001, PR China b
art ic l e i nf o
a b s t r a c t
Article history: Received 21 February 2013 Accepted 26 August 2013 Available online 10 December 2013
This study performs a block scale investigation on gas content of a coal reservoir in Zhengzhuang Block of the southern Qinshui basin in China. The gas content of Coal Seam No. 3 in this coal reservoir was measured in field and laboratory in conjunction with tests on coal properties such as adsorption isotherm, maximum vitrinite reflectance, coal composition and maceral component etc. Total 36 coal cores collected from 3 adjacent coalmines in the southern Qinshui basin were investigated, including analysis of logging data from the drilling wells. The investigations provided experimental data for block scale modeling and visualization analyses on the correlation between gas content and the key factors such as coal properties and geological conditions of the coal reservoir. Data obtained by field and lab tests were analyzed by statistical models in order to correlate gas content and individual type of coal properties and geological variables. The statistical model was then used to map the gas content of the target coal seam in the studied area, resulting in a flood map of gas content at a 1:50000 scale. The flood map was further visualized with other variables in terms of the properties of coal and coal reservoir and its geological conditions. These visualized maps provide useful geological interpretation for block scale investigation of the comprehensive relationships between the gas content and the coal properties and regional structure in the given coal reservoir. The results show that gas content has little correlation with coal rank, maceral composition, coal thickness, cap and bottom lithology, while it is highly related to the structural properties such as burial depth and effective cover thickness. A stagnant hydrodynamic condition is favorable to the higher gas content on the whole but does not contribute to gas lateral and local variation. Canonical correlation and principal component analysis on the statistical model reveal the key factors that control the gas content are burial depth, effective thickness of overlying strata, groundwater level and moisture content in coal seam. & 2013 Elsevier B.V. All rights reserved.
Keywords: coalbed methane gas content reservoir evaluation southern Qinshui basin statistical model visual map
1. Introduction Gas content of coalbed methane (CBM) reservoirs is one of most important indices for exploration and exploitation of CBM resource deposited in coal seams. The methane adsorption isotherm has been widely used to measure the gas content of coals (AQSIO and SAC, 2004a, 2008a). However measurement of the adsorption isotherm on coals cannot accurately represent the gas content of CBM reservoirs since the ideal laboratory conditions are different from the in-situ circumstance of the CBM reservoirs. To
n
Corresponding author. E-mail address: [email protected] (S. Sang).
0920-4105/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.petrol.2013.08.039
determine the gas content of CBM reservoirs, standardized field tests have to be carried out, consisting of estimating the lost gas content during sampling and measuring desorbed and residual gas contents based on the adsorption isotherm method (AQSIO and SAC, 2004b,2008b; Wang et al., 2011). The field experiments are integrated with other measurements of coal and reservoir properties, such as measurements of adsorption isotherm and gas component, the maceral composition, proximate and ultimate analyses, and tests for vitrinite reflectance and porosity. These measurements provide the experimental data for block scale investigation on the gas content of CBM reservoirs. Qinshui basin is the largest basin for CBM bearing resource as well as the most active area for exploration and development of coalbed methane in China. It is located in Shanxi province of
2
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
northwest China, characterized by one of the deposited most abundant anthracite coal reservoirs in the world. Currently, there are about 3000 CBM wells completed in China and most of them are located in the southern Qinshui basin, making up over 95% of the total CBM wells in the country. It can be predictable that there will be increasing research activities and investments on projects for CBM recovery in this basin. However, the studies on block scale evaluation of gas content in the whole Qinshui basin are still limited (Qin et al., 2005; Song et al., 2005; Zhao et al., 2005; Cai et al., 2011; Song et al., 2012), in particular for the southern part of Qinshui basin (Zhao et al., 2005, Cai et al., 2011; Song et al., 2012). Among these studies, only a few attempted to investigate the effect of various geological factors on gas content (Cai et al., 2011). These studies investigated the control mechanism on gas content from different perspectives, mainly based on large-scale CBM basins. The distribution of gas content in the field, which is essential for optimizing well pattern and minimizing investment risk for CBM recovery, has not been fully understood and needs to be investigated further. As well known, gas content of CBM reservoirs is generally controlled by comprehensive geological conditions, including tectonic setting, depositional environments (Ayers, 2002; Song et al., 2005; Song et al., 2012), hydrological features (Kaiser et al., 1994; Scott, 2002; Pashin, 2007), reservoir pressure (Liu et al., 2005), burial depth (Su et al., 2005; Liu et al., 2012; Song et al., 2012), coal quality and rank (Langenberg, et al., 2006), and coal petrology (John et al., 1989; Levine, 1992; Scott, 2002). Despite many attempts which have been made to measure gas content of various coals, and to discuss a general relationship between the gas content of coal and the coal geological conditions in basin
scale, the quantitative relationship between gas content and the geological factors has not been clear particularly in the developing CBM field. This uncertainty makes it difficult to evaluate the CBM potential capacity of CBM reservoirs in a block scale, and hence increases the risk of investments for the exploitation of a CBM field. This paper presents a block scale investigation on gas content of the CBM reservoir in Zhengzhuang Block of the southern Qinshui basin by systematic sampling of coal samples and comprehensive analyses of gas content data measured in field and laboratory.
2. Geological setting Fig. 1 shows the studied area with the primary geological information such as faults and well locations for sampling. Coalbearing strata of the studied area consists of Benxi, Taiyuan, Shangshihezi and Shiqianfeng formations of the PermoCarboniferous from bottom to top, and Coal Seam No. 3 in the studied area is developed in Shanxi formation, as indicated in Fig. 2. Shanxi formation comprises mudstone, sandstone, sandy mudstone and coal deposited in deltaic environments, such as mouth bar, interdistributary estuary and delta plain, of Paleozoic North China Epicontinental Sea. Its thickness is 50 m in average, ranging from 39 to 79 m. The coal-bearing strata in the area mainly experienced four periods of tectonic movement: Hercynian period, Indosinian period, Yanshanian period and Himalyan orogenic period. During the Hercynian–Indosinian period, the area experienced a long
Fig. 1. Tectonic setting of Zhengzhuang Block in southern Qinshui basin.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
3
(Song et al., 2005; Zhao et al., 2005). These tectonic movements directly resulted in the current Qinshui morphology, which forms a fundamental impact on CBM preservation. Correspondingly, during the Indosinian period, folds with E–W strike were formed and extensional tectonics such as normal faults with NE–SW strike were developed under the early compressive tectonic stress parallel to the normal faults strike. In the Yanshanian period, a synclinore striking NE, the basic tectonic shape of the whole basin was formed under NW–SE compression stress, which was accompanied by a series of folds striking NNE and joints striking NW or EW. The Himalyan period was featured by a NNE–SSW compression stress, which led to the folds striking NW combined with early folds with NE strike. Finally, during the Quaternary period of the late Himalyan period, a few of Epigenetic minor folds with NW strike were formed due to approximately horizontal and compressive tectonic stress (Fig. 1).
3. Methodology 3.1. Prospecting data
coal seam NO.3
Data such as coal bed thickness, burial depth, coal roof lithology, cap thickness, and coal bed roof elevation was obtained by drilling tests with the aid of well logging measurements. The well logging tests mainly detected resistivity, spontaneous potential and Gamma ray with the logging intervals including Shangshihezi formation, Xiashihezi formation, Shanxi formation, and Taiyuan formation from top to bottom of Permo-Carboniferous system. The logging tests provided the well logging interpretation results such as the thickness, the location and the lithology of formation, coal quality such as ash content, moisture content, volatile content, porosity, gas content, and rock mechanics parameters such as Poisson' s ratio, shear modulus, pressure gradient, and ground stress. 3.2. Sampling and experiments
coal seam No.15
Fig. 2. Stratigraphic column showing coal-bearing formations in southrn Qinshin basin.
Fig. 3. In-situ temperature of coal samples in Coal Seam No. 3 in southern Qinshui basin.
burial history and a sustainable sedimentation process. In the early time of the middle Yanshanian period, the area experienced a fluctuant burial process. From the late Yanshanian period to the Himalyan period, the area experienced a continuous uplifting
There are in total 36 coal samples of Shanxi formation collected from different well cores of Zhengzhuang Block and 3 underground coalmines named Sihe, Tangan, Chengzhuang in the southern Qinshui basin. Sampling sites are uniformly scattered in the studied area (see Fig. 1). All the coal samples were tested with well-established measurement methods including measurements of adsorption isotherm and gas component, the maceral, proximate and ultimate analyses, and tests for vitrinite reflectance and porosity etc. Canister desorption method was used for measurement of desorbed gas from coal core samples in field tests by means of a standard desorption device. The measurement was conducted through the procedures as described by the National Standard GB/T 19559-2008 in China. According to the standard, prior to putting a coal sample into the canister, the canister was flushed with argon to remove the air. During the natural desorption, desorption capacity was gauged in given time intervals. The natural desorption measurement was stopped when the average desorption capacity was less than 10 cm3 per day within 7 days. Sometimes the measurement needed to be carried out for about 3 months because of the slow diffusion/desorption of gas in coal. After measuring the natural desorption, a part of the coal samples were chosen to grind in a ball mill for 2–4 h, then they were put into an indoor canister. Subsequently, the canister was put into a thermostat and its temperature adjusted to the same as the reservoir coal. The residual gas capacity was measured. The escaped gas content was calculated according to the direct method proposed by US Bureau of Mine (USBM) (McLennan et al., 1995). It
4
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Table 1 Properties of interest for coal samples used. Sample number
Moisture (ad%) Ash (d%) Volatiles (ad%) Critical desorption pressure (MPa) Langumiur volume (cm3/g, ad) Langumiur pressure (MPa, ad) Gas content (cm3/g, ad)
No. 1
No. 2
No. 3
No. 4
No. 5
No. 6
No. 7
No. 8
…
No. 36
1.11 11.75 6.10 8.73 36.44 3.55 26.99
1.62 9.23 5.94 4.81 41.44 3.04 25.81
1.39 12.36 7.14 7.74 37.56 3.2 26.69
1.21 10.59 6.40 3.82 37.54 3.05 22.51
1.33 12.50 6.72 3.03 2.2 37.03 21.48
1.36 10.96 6.39 3.28 39.5 2.8 21.4
1.23 9.79 6.35 6.17 40.97 3.69 25.74
1.91 13.63 6.74 6.26 31.02 2.78 22.23
… … … … … … …
1.28 13.27 7.13 1.15 40.7 3.41 13.44
y = 0.0124x + 13.456 2 R = 0.7006
30
3
Gas content (m /t)
35
25 20 15 10 400
Fig. 4. Typical adsorption isotherms of selected coal samples (referring to Fig. 1 for sampling points) based on air dry at 30 1C.
600
800 1000 Burial depth (m)
1200
1400
30
3
Gas content (m /t)
35 is usually determined by the positive relationship between desorption volume of coalbed methane and time. Gas content was determined by the sum of the escaped gas content, the natural desorption and the residual capacity, generally expressed on the basis of ash, water free. After measurements of the gas content, each sample remainder was split into several aliquots (using the standard coning and quartering technique) and prepared for analyses such as proximate analysis, coal petrology (maceral composition and vitrinite reflectance), specific surface area and pore structure. Proximate and maceral composition analyses were conducted according to Chinese National Standard GB/T 212-2001 and GB/T 8899-1998, respectively. Pore structure and specific surface area analyses abided by Chinese Petroleum and Natural Gas Industry Standards SY/T 5346-2005 and SY/T6154-1995, respectively. AXIO Imager M1m Microspectrophotometer, ASAP 2000 and Microstructure Analyzer 9310 were used for measurements of maceral composition, specific surface area and pore structure. The collected coal samples were analyzed for gas composition using gas chromatograph (GC) in coordination with the National Standard procedures of GB/T 13610-2003. The adsorption isotherm was measured by referring to Chinese National Standard GB/T 19560-2004, using the Isotherm Adsorption/Desorption System IS-100. Coal samples for isotherm measurements were prepared by crushing and screening fresh air-dried grains to the particle size range of 0.18–0.25 mm. Since in-situ reservoir coal in the target coal seam was deposited under a condition of temperature at around 30 1C (Fig. 3) and water content would be saturated in coal, the isotherm measurements were performed at 30 1C under the moisture equilibrium condition.
25 20 15 10 400
600 800 1000 1200 Overlying formation thickness (m)
1400
Fig. 5. Relationship of gas content with (a) burial depth; and (b) effective thickness of overlying strata in Zhengzhuang Block.
investigations as described above were used for modeling by the statistical method to correlate gas content with individual type of geological factors in order to predict gas content for CBM exploration and exploitation. 3.4. Visual analysis on gas content and the key factors The field and laboratory investigations show that the gas content can be correlated statistically to the properties of coal and coal reservoir and its geological conditions, which can usually be expressed as variables of a given coordinate plane. Thus the correlations shown in the predicted results from the statistical model and the experimental results of gas content can be used to map the gas content of the target coal seam in the studied area.
3.3. Statistical model for gas content 4. Results Data from experiment test and prospecting were conducted to correlation analysis on gas content and the different factors. Therefore, data obtained by well logging and laboratory
Table 1 summarizes some properties of interest for representative coal samples from a total of 36 samples for simplicity. Note
35
35
30
30
Gas content (m /t)
25
3
3
Gas content (m /t)
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
20 15 10
9
12
15 Aad (%)
18
21
30
30 3
25 20 15 10
0.5
1
1.5 Mad (%)
2
70 80 Vitrinite (%)
90
100
10
20 30 Inertinite (%)
40
50
25 20 15 10
0
2.5
0
Fig. 8. Variation of gas content with (a) vitrinite and (b) inertinite macerals in Zhengzhuang Block.
35 Gas content (m /t)
30 3
25 20 15 10 5
30 25 20 15 10 5
0 3.3
3.5
3.7 Ro,max (%)
3.9
0
4.1
35
30
30
3
Gas content (m /t)
35
25 20 15 10 5 0
60
5 0
35
3
10
35
Fig. 6. Variation of gas content with (a) ash content (Aad); and (b) moisture content (Mad) in Zhengzhuang Block.
Gas content (m /t)
15
35
0
3
20
0 50
24
Gas content (m /t)
3
Gas content (m /t)
6
5
Gas content (m /t)
25
5
5 0
5
7
8
9
Vdaf (%) Fig. 7. Variation of gas content with (a) maximum vitrinite reflectance (Ro,max) and (b) volatile content (Vdaf) in Zhengzhuang Block.
that adsorption properties (i.e. Langmuir volume and pressure) were measured using the adsorption isotherm as shown in Fig. 4. As can be seen, Langmuir volumes of these coal samples vary from 32.94 m3/t to 50.55 m3/t with 43.11 m3/t on average and Langmuir pressures are 2.21–3.93 MPa with the average of 2.96 MPa. Gas content in the studied area remarkably changes in ranges of
5 10 Immediate roof thickness (m)
15
0
5 10 Immediate bottom thickness (m)
15
25 20 15 10 5 0
6
0
Fig. 9. Relationships of gas content with thickness of (a) immediate roof and (b) immediate bottom in Zhengzhuang Block.
8.03–31.44 cm3/g with an average of 20.82 m3/t (air dried basis). The variation of gas content obviously reflects the heterogeneity of gas-bearing in the coal reservoir as well as the complex effect of multiple geological factors on gas enrichment. The correlation analysis is often used to determine the relationship gas content and other influenced factors (Gentzis et al., 2006).
6
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
and bottom determines the sealing condition of CBM, which should be considered in filed scale investigation on gas content of coal seam. According to linear regression analyses, a simple statistical model was expressed in vector form as
Figs. 5–9 show the results of coal core analyses, correlating the gas content to the various factors that may affect the gas content of the coal seam. The relationships between gas content and burial depth, effective thickness of overlying strata in the studied area are shown in Fig. 5a and b. As can been seen, gas content has an obvious correlation with both burial depth and effective thickness of overlying strata. Ash content (Ad) in the studied area varies between 9.23% and 21.63% with the average of 12.91%. Relationship of gas content and ash content (Aad) is shown in Fig. 6a, implying that gas content has a relatively obvious negative correlation with Aad. Moisture content (Mad) in the studied area varies between 0.71% and 1.91% with the average of 1.36%. Likewise the relationship between gas content and Mad shows a negative correlation (Fig. 6b). Fig. 7 shows the variation of gas content with maximum vitrinite reflectance (Ro,max) and volatile matter (Vdaf) content. As can been seen, the correlation between gas content and coal maturity is not obvious (Fig. 7a) whilst the gas content is generally decreased with volatile matter content (Fig. 7b). The relationship of gas content with maceral composition is not obvious too, but the gas content seems to be increased with the increase of vitrinite and decreased with the increase of inertinite (Fig. 8). Finally, the relationships between gas content and thicknesses of the immediate roof and the immediate bottom were analyzed, as indicated in Fig. 9. It shows that there are no obvious correlations between them. However the immediate roof
y ¼ Xβ þ ε
ð1Þ
where y is a vector consisting of measured gas content yi (i ¼1,2,…, n) (m3/t) from n coal samples; X represents the defined matrix comprised from the input variables xnm as follows: 0 1 x11 ⋯ x1m B x21 … x2m C B C ð2Þ X¼B C ⋱ ⋮ A @ ⋮ xn1 ⋯ xnm with m numbers of the measurable values of the key variables that may affect the gas content; β is a m-dimensional parameter vector consisting of regression coefficients; and ε is a vector formed by the error term, or so-called disturbance term.β and ε can be calculated based on the data shown in Table 2 with the least square method. In this study, a two-step regression approach was used to formulate the statistical model above. Firstly the 12 key factors were chosen for the modeling, i.e. m ¼ 12, including immediate roof thickness (xi1), immediate bottom thickness (xi2), coalbed
Table 2 Data used for statistical analysis. Sample no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ⋮ 36
Vitrinite content
Inertinite content
Ground water level
wt%
Ash content (Aad) wt%
wt%
wt%
m
1.11 1.62 1.39 1.21 1.33 1.36 1.23 1.91 1.62 1.86 1.37 1.28 1.04 1.28 0.85 0.81 1.30 1.33 1.16 ⋮ 1.46
11.75 9.23 12.36 10.59 12.5 10.96 9.79 13.63 14.34 11.52 9.68 10.51 10.81 11.44 10.62 10.36 12.84 15.60 10.78 ⋮ 10.39
75.40 73.80 71.30 69.80 67.50 79.30 89.20 63.20 61.50 60.90 72.80 69.00 87.90 92.30 83.40 68.00 69.30 86.30 81.61 ⋮ 66.40
24.60 25.60 28.70 30.20 32.50 20.70 10.80 36.80 38.50 39.10 27.20 31.00 12.10 7.70 16.60 32.00 30.70 13.80 18.39 ⋮ 33.60
503 498 535 496 502 501 510 514 524 542 539 587 557 586 590 603 598 540 535 ⋮ 582
Gas Immediate content roof
Immediate bottom
Coalbed thickness
Coal rank
Burial depth
Overlying strata
Volatile Moisture content (Vdaf) content (Mad)
m3/t
m
m
m
%
m
m
vol%
26.99 25.81 26.69 22.51 21.48 21.4 25.74 22.33 20.28 21.7 23.81 26.65 28.5 25.03 30.04 31.44 28.45 20.94 22.8 ⋮ 22.02
1.92 2.51 9.09 4.16 4.20 5.06 1.05 0.18 11.9 4.00 4.00 3.05 2.54 5.89 11.15 14.31 10.02 6.48 5.90 ⋮ 2.99
1.06 1.55 10.94 11.56 6.29 0.51 1.07 0.24 10.37 2.91 11.81 4.04 1.03 11.91 10.74 2.24 6.10 1.14 9.51 ⋮ 2.80
5.50 4.90 5.80 5.30 6.00 5.35 5.20 5.30 5.60 5.60 4.70 5.70 5.75 6.00 6.10 6.20 5.90 6.30 6.90 ⋮ 6.10
3.85 3.88 3.47 3.83 3.91 3.89 3.88 3.81 3.61 3.62 3.67 3.53 3.49 3.62 3.77 3.68 3.44 3.50 3.50 ⋮ 3.53
980.4 638.5 960.3 753.7 652.6 659.95 961.5 900.6 746.8 751.0 888.9 1267.9 1038.5 1114.5 1059.2 1156.5 1336.9 513.6 626.4 ⋮ 1117.1
906.7 619.6 921.4 717.7 629.5 634.3 922.7 890.6 718.8 748.4 860.1 1249.9 1010.5 1109.9 1038.2 1111.5 1299.1 473.0 596.4 ⋮ 1030.5
6.10 5.94 7.14 6.40 6.72 6.39 6.35 6.74 7.48 6.81 6.44 7.11 6.63 6.69 6.74 8.16 7.95 7.52 7.17 ⋮ 6.58
Table 3 Correlation coefficients of gas content with different factors based on canonical correlation and principal component analysis. Analysis method
Immediate roof thickness
Immediate bottom thickness
Coalbed thickness
Coal rank
Burial depth
Effective thickness of overlying strata
Volatile content
Moisture content
Ash content
Vitrinite content
Inertinite content
Ground water level
Canonical relation
0.335
0.008
0.091
0.087
0.779n
0.768n
0.078
0.625n
0.469
0.159
0.159
0.633n
Principal component
0.335
0.008
0.091
0.087
0.779n
0.768n
0.078
0.625n
0.469
0.159
0.159
0.633n
n
Means the factors have an obvious correlation with gas content at 0.01 confidence level during significance test.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
7
Fig. 10. Flood map of gas content predicted for Coal Seam No. 3 of Zhengzhuang Block in southern Qinshui basin.
Fig. 11. Visual relationship between gas content and coal thickness of Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: m for coal thickness and m3/t for gas content with different color areas.)
8
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
thickness (xi3), coal rank (i.e. Ro,max) (xi4), burial depth (xi5), effective thickness of overlying strata (xi6), volatile content (xi7), moisture content (xi8), ash content (xi9), vitrinite content (xi10), inertinite content (xi11) and groundwater level (xi12). As a result of applying the data in Table 2 into Eqs. (1) and (2), the first-step regression eqution can be determined, giving 12
Y ¼ ∑ Aj X j þ B j¼1
ð3Þ
where Y represents gas content (m3/t); Aj (j¼1,2,.., 12) and B are the regressed coefficients. In order to identify the significance of these correlations to predict gas content, the canonical correlation analysis (CCA) and principal component analysis (PCA) were further performed for the first-step regression eqution above. The details can be found in Table 3. It can be seen from these statistical analyses that gas content shows little correlation with coal thickness, coal rank and coal maceral composition. It has a relative strong correlation with other factors in small-scaled CBM field. As shown in Table 3, the correlation coefficients between gas content and burial depth, effective thickness of overlying strata, groundwater level and moisture content are above 0.5, showing the strong effects of these four geological factors on the gas content. While the first three factors have an obvious positive correlation with gas content, the fourth factor, i.e. moisture content obviously exhibits a negative effect on gas content. To pursue quantitative recognition
for coupling control of these dominant factors on gas content, the second-step regression was conducted by taking these four factors into account. Thus the resultant statistical model was expressed in a simplified format of Eq. (3), giving Y ¼ 0:085D þ0:0007T 4:7708M ad þ 0:0015L þ23:675
ð4Þ
where D is burial depth (m); T is effective thickness of overlying strata (m); L is groundwater level (m) and Mad is moisture content (wt%). The fitting degree of the statistical model was conducted. The results show that the fitting degree increases with the amount of the factors adopted in the model. The estimated residual standard deviations of Y are 1.82% and 1.45%, the fitting degrees are 75.12% and 92.63%, when 4 key factors and 12 key factors were considered respectively. Significant inspection F-statistics of the models are 11.32 and 7.34, which are above the critical values (F0.05 (4,15) ¼ 5.86, F0.05 (12,7) ¼2.91), whilst the p-values of the models are 0.0002 and 0.0069, which are far below 0.05, when 4 key factors and 12 key factors were considered respectively, Those demonstrate the models achieve a significant level. The geological conditions are related to the location, which can usually be expressed as variables of a given coordinate plane. Thus the correlations shown in the predicted results from Eqs. (3) and (4) in statistical model can be used to map the gas content of the target coal seam in the studied area. For this purpose, the resultant gas content was firstly plotted in the flood map (Fig. 10) at a 1:50000 scales, defined by the coordinates of sampling boreholes
Fig. 12. Visual relationship between gas content and burial depth of Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: m for burial depth of Coal Seam No. 3 and m3/t for gas content with different color areas.)
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
in the studied area (Fig. 1). Then different contour maps were drawn as Figs. 11–17 in order to visualize the correlations between gas content and various variables of interest such as thickness and burial depth of coal seam, effective thickness of overlying strata, moisture and ash content of coal, etc. These visualized maps provide useful information for block scale evaluation of the comprehensive relationships between distribution of the gas content and the coal reservoir properties and regional structure, which will be discussed in details as follows.
5. Discussion 5.1. Gas content vs. reservoir properties 5.1.1. Coal thickness The well tests show that coal thickness is about 4.5–6.5 m in the studied area (Fig. 11). The coal seam gas is unevenly distributed in the target coal seam in the studied area, more in the north compared with the south. As can be seen, the distribution of gas content is apparently inconsistent with the distribution of the coal thickness. This implies that coal thickness has less control effect on gas content. Because stratigraphy and geologic structure determine the continuity of coal (Pashin, 1998), coal thickness trends and depositonal fabric strongly influence the distribution of gas content (Scott, 2002). Variation of coal thickness is relatively small
9
in this developing field and the effect of coal thickness on gas content is mainly dependent on depositional environment that has no obvious difference in the whole coalfield.
5.1.2. Tectonic effect The thermal evolution of coal seam in the studied area experienced two hydrocarbon generations in the late Triassic and late Jurassic–early Cretaceous. The coalification degree and gas generation reached the maximum value due to the abnormal thermal event in the Yanshanian period (Song et al., 2005; Zhao et al., 2005). Tectonic uplift has determined the preservation degree of coalbed methane since the late Yanshanian period. The present gas content is mainly dependent on effective thickness of overlying strata on coal seam during the key geological period (Zhao et al., 2005). In the studied area, different types of folds and faults were well developed due to tectonic movement. EN–WS trending F2 Fault was identified (Fig. 1). The gentle and approximate parallel folds were developed widely in the developing area, trending towards NNE and SN. The analysis shows that the gas content of coal seam located in syncline is relatively high, and gas content located in anticline and fault zone is relatively low (Fig. 10). In other words, multi-period tectonic stress field determines the development of folds and faults, and tectonic style, in terms of folds and faults, controls enrichment and dissipation of coal seam gas.
Fig. 13. Visual relationship between gas content and distribution of ash content in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: wt% for ash content and m3/t for gas content with different color areas.)
10
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Fig. 14. Visual relationship between gas content and distribution of moisture content in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: wt.% for moisture content and m3/t for gas content with different color areas.)
5.1.3. Burial depth Many case studies revealed that burial depth is an important factor that affects gas potential of coal seams (Su et al., 2005; Liu et al., 2012; Song et al., 2012). The gas content was usually lower if the coal bearing had ever suffered serious erosion when coal bed was in a shallow buried stage. In this case, coal seam gas was easily migrated because of the strata uplift in geological history, even if coal bed has a deep burial depth later. Therefore the effective thickness of overlying strata would play a more important role in controlling the gas content (Ulery and Molinda, 1984). An effective thickness of overlying strata is defined by the bottom thickness from coal bed to the first unconformity where CBM was largely generated. It reflects tectonic uplift and erosion on the CBM preservation. Theoretically speaking, it can provide the superior geological conditions for CBM enrichment compared with overlying bedrock thickness and burial depth. As a result, higher gas content usually appears in areas with thicker overlying strata. Burial depth of coal seam in the developing field of the studied area varies from 500 m to 1200 m. The correlation between burial depth and gas content is described in Fig. 12, consisting of the flood map of gas content together with the contours of burial depth. As can be seen, distribution of gas content in this studied area is generally consistent with the contours of burial depth.
5.1.4. Overlying strata The analysis of strata thickness and residual strata thickness shows that, in most part of the studied area, Triassic and
Quaternary strata is missing due to tectonic uplifts, and Triassic strata is retained the form of Shiqianfeng strata in partial area of the field. Quaternary strata and residual Shiqianfeng strata are relatively shallow by contrast. In general, the effective thickness of overlying strata is a little different from burial depth because of some small uplifts and deposition of the field in geological history. Therefore strata thickness controls the gas content almost in the similar way as the burial depth. However, the correlation of the effective thickness of overlying strata with gas content in the studied area seams weaker than that between the burial depth and gas content, as shown in Figs. 5 and 12. The results suggest that the denudation and deposition caused by the tectonic uplifts and sedimentation has a certain influence on gas content. As a result, the gas content in the studied area almost remained the same until the decline stage, which would be mainly controlled by burial depth of the coal seam. 5.2. Gas content vs. coal properties 5.2.1. Coal quality Generally speaking, if the ash content in coal is low, there will be more gas generation. This is because the gas production in coalification would be higher due to the less inorganic substance in the source of coal during evolution. The effect of ash content on gas content is illustrated in Fig. 6a. As can be seen, gas content has a relatively obvious negative correlation with ash content, in line with the fact that total surface area of pores and adsorption
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
11
Fig. 15. Visual relationship between gas content and distribution of Ro,max in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: % for Ro,max and m3/t for gas content with different color areas.)
capacity declines with the increasing of gas content. That is, the higher mineral particles such as clay minerals, pyrite brought by syngenetic sedimentation, epigenetic sedimentation or magmatic intrusion are filled in pores, and hence decrease the adsorption and migration capacity of coal seam gas. Meanwhile, the relationship of gas content and ash content reflects the depositional effect and tectonic synergistic effect. Thus ash content is a relatively important factor that affects the gas content. The effect of coal quality on distribution of gas content in the target coal seam can be further visualized by plotting the flood map of gas content with the contours of ash content, as shown in Fig. 13. It can be seen that the distribution of gas content in the studied area is correlated with the distribution of ash content to a certain extent. The variation of gas content from the north to the south in the studied area is highly consistent with the variation of ash content in the same direction and trend. Moisture content represents another important index of coal quality that significantly affects the gas content, as shown in Fig. 6b. This is because water molecule has a polarity that can be preferentially absorbed on the pore surface of coal, and can partially replace adsorption sites of methane (Clarkson and Bustin, 2000). Therefore, high moisture content in coal can lower adsorption capacity of coal, which can generally lead to lower gas content. The visual correlation between moisture and gas content illustrated in Fig. 14 shows that the distribution of gas content is
largely negatively correlated with distribution of moisture content. The variation of gas content from the north to the south in the studied area is also highly consistent with the variation of moisture content in the same direction and trend.
5.2.2. Coal rank Coal rank is one of the key factors that affect adsorption capacity of coal (Laxminarayana and Crosdale, 2002). Many researchers have shown that total adsorbable methane capacity increases with coal rank, which has been considered the main factor influencing the methane adsorption capacity (Ryan, 1992; Crosdale et al., 1998; Scott, 2002; Yao et al., 2009). As shown in Fig. 7a, the maximum vitrinite reflectance (Ro,max) of coal in the studied area varies from 3.2% to 4.0%; and the volatile matter (Vdaf) content varies between 5.38% and 8.41%. The correlation of gas content and coal maturity is not obvious, which is not consistent with the evolution of coal rank proposed by Qin et al. (1999), and demonstrates the relatively small difference of coal maturity and effects from other factors in the developing field; and the effect of coal maturity on gas content is relatively weak. However, the relationship between gas content and volatile matter content shows an obvious negative correlation (Fig. 7b). This indicates that gas content increases with the decrease of volatile matter (Vdaf) and probably the increase of coal rank. The visual relationship between the distribution of gas content and coal rank in
12
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Fig. 16. Visual relationships between gas content with the distribution of immediate roof thickness in Coal Seam No. 3 in Zhengzhuang Block. (Unit of contour lines: m for immediate roof thickness of Coal Seam No. 3 and m3/t for gas content with different color areas.)
terms of the maximum vitrinite reflectance (Ro,max) of coal is illustrated in Fig. 16. As can be seen, the distribution of gas content seems to have no correlation with coal rank in the block scale.
5.2.3. Coal maceral composition The gas generation potential is related to the maceral composition directly (Scott, 2002). For the relationship between the maceral composition and adsorption capacity, coal maceral composition and coal maturity determine surface area, pore size, and distribution of pore volume in coal (John et al., 1989). The coal in the studied area is rich in vitrinite, vitrinite varies between 53.9% and 92.7%, inertinite between 7.3% and 46.15%. The relationship between gas content and maceral composition indicates that gas content increases with the increase of vitrinite, and decreases with the increasing of inertinite in general (Fig. 8). Meanwhile, the correlations between gas content and the maceral content indicate that adsorption capacity of coal depends on the content of vitrinite, i.e. the vitrinite-rich coal has strong adsorption capacity and hence leads to high gas content. However, the inconspicuous correlation between gas content and the maceral composition implies that the influence of the maceral composition on gas content is so complicated that it is difficult to evaluate without more detailed information about the comprehensive effects of coal rank, moisture content, ash content, tectonic setting, hydrological features, and depositional setting (or coal facies) (Pashin, 1998;
Crosdale et al., 1998; Pashin, 2007; Faiz et al., 2007; Yao et al., 2009; Song et al., 2012). This needs to be further studied. 5.3. Gas content under different geological circumstances 5.3.1. Sealing condition The lithology and thickness of surrounding rocks of coal directly affect the sealing of CBM as well. The immediate roof and the immediate bottom of coal mainly consist of mudstone and silty mudstone, siltstone and fine-grained sandstone in the studied area. The mudstone thickness of roof or bottom of Coal Seam No. 3 is above 10 m (Fig. 2). Mudstone is quite tight and no fracture was observed. It can prevent CBM from migrating through diffusion, seepage and water-soluble flow. Therefore the sealing condition of the developing field is believed very well. The field measurements reveal no obvious correlation between gas content with immediate roof thickness and immediate bottom thickness (Fig. 9). This implies that sealing condition of CBM in terms of immediate roof thickness and immediate bottom thickness seems to have no influence on gas content as the coal seam in the studied area is well sealed. However the visual correlation between immediate roof thickness and gas content illustrated in Fig. 16 shows that the distribution of immediate roof thickness is consistent with gas content to a certain extent. The effect of sealing condition on gas content might reflect the depositional control because depositional setting determines overburden thickness and coal floor in general.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
13
Fig. 17. Visual relationship between gas content of Coal Seam No. 3 and hydrodynamic condition in Zhengzhuang Block. (Unit of contour lines: m for groundwater level and m3/t gas content with different color areas.)
5.3.2. Hydrodynamic conditions Basin hydrodynamic conditions have a strong influence on reservoir properties and play a vital role in environmental issues (Scott, 2002; Pashin, 2007). The south of Qinshui basin is a catchment area. Overlying strata of Coal Seam No. 15 is mainly thick limestone of 2–15 m (see Fig. 2), and there is no hydrodynamic connection between Coal Seam Nos. 3 and 15. These coal seams are separated by multiple impermeable layers. Multiple hydrodynamic systems exist in the basin, controlled by the watershed, the process of groundwater recharge, migration, and discharge happened mainly at Qinshui basin margin. Hydrodynamic condition of the flow intersectional region was formed in the southern basin. This made the southern basin stagnant, and thus formed rich gas region. As shown in the visual relationship between gas content of Coal Seam No. 3 and hydrodynamic condition (Fig. 17), the developing field in the studied area is just located in the stagnant zone. As a result, a stagnant hydrodynamic condition is favorable to the higher CBM content on the whole but has no contribution to gas lateral and local variation (Su et al., 2005; Cai et al., 2011).
analyses at a block scale. Statistical analyses reveal that burial depth, effective thickness of overlying strata, groundwater level and moisture content in coal seam are dominant factors that control gas content in block scale. The results show that gas content has little or no correlation with coal rank, maceral composition, coal thickness, and immediate bottom thickness. The factors such as coal thickness, coal rank and sealing condition do not significantly affect the distribution of gas content in the studied area. While the gas content has a relatively high negative correlation with coal quality in terms of ash and moisture contents, the distribution of gas content is highly consistent with the variation of ash content, moisture content in the coal seam. Moreover, gas content has an obvious correlation with burial depth, and effective thickness of overlying strata. The effects of tectonic setting, burial depth, and effective thickness of overlying strata on gas content are dominant. A stagnant hydrodynamic condition is very favorable to CBM preservation, but has minor effect on the distribution of gas content in the developing field.
Acknowledgments 6. Conclusions A comprehensive investigation on gas content of Coal Seam No. 3 in Zhengzhuang Block of the southern Qinshui basin, China, was conducted by means of the statistical models and visualization
The research was supported by National Natural Science Fund of China (Grant nos. 41302129, 41330638, 41272154, 4117238 and 41272171), the PhD Fund of Anhui University of Science and Technology (Grant no. 11224), Project 2012012008 of Shanxi Province United Research Funding of Coalbed Methane and the
14
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Scientific Research Program of Huabei Oilfield Company. Geoff Wang acknowledges Australian Research Council (ARC) Discovery project for financial support on this collaborative research between the China University of Mining and Technology and the University of Queensland, Australia. References AQSIO, S.A.C., 2004a. GB/T 19560-2004 Experimental Method of High-Pressure Adsorption Isothermal to Coal-Capacity Method. Standards Press of China, Beijing. AQSIO, S.A.C., 2004b. GB/T 19559-2004 Method of Determining Coalbed Methane. Standards Press of China, Beijing, Beijing. AQSIO, S.A.C., 2008a. GB/T 19560-2008 Experimental Method of High-Pressure Isothermal Adsorption to Coal. Standards Press of China, Beijing, Beijing. AQSIO, S.A.C., 2008b. GB/T 19559-2008 Method of Determining Coalbed Gas Content. Standards Press of China, Beijing, Beijing. Ayers, W.B., 2002. Coalbed gas system, resources, and production and review of contrasting cases from the San Juan and Powder River basin. AAPG Bull. 86, 1853–1890. Cai, Y.D, Liu, D.M, Yao, Y.B, Li, J.Q, Qiu, Y.K, 2011. Geological controls on prediction of coalbed methane of No. 3 coal seam in Southern Qinshui Basin, North China. Int. J. Coal Geol. 88, 101–112. Clarkson, C.R., Bustin, R.M., 2000. Binary gas adsorption desorption isotherms: effect of moisture and coal composition upon carbon dioxide selectivity over methane. Int. J. Coal Geol. 42, 241–271. Crosdale, P.J., Beamish, B.B., Valix, M., 1998. Coalbed methane sorption related to coal composition. Int. J. Coal Geol. 35, 147–158. Faiz, M., Saghaf, A., Sherwood, N., Wang, N., 2007. The influence of petrological properties and buiral history on coal seam methane reservoir characterization, Sydney Basin, Australia. Int. J. Coal Geol. 70, 193–208. Gentzis, T., Schoderbek, D., Pollock, S., 2006. Evaluating the coalbed methane potential of ghe Gething coals in NE British Columbia, Canada: an example from the Highhat area, Peach River coalfield. Int. J. Coal Geol. 68, 135–150. John, F.U., Colin, S.F., Leslie, F.J., 1989. Moisture in coal: 2. Maceral effects on pore structure. Fuel 68, 18–26. Kaiser, W.R., Hamilton, D., Scott, A.R., Tyler, R.J., 1994. Geological and hydrological controls on the producibility of CBM. J. Geol. Soc. London 51, 341–372. Langenberg, C.W., Beaton, A., Berhane, H., 2006. Regional evaluation of the coalbedmethane potential of the Foothills/Mountains of Alberta, Canada. Int. J. Coal Geol. 65, 114–128. Laxminarayana, C., Crosdale, P.J., 2002. Controls on methane sorption capacity of Indian coals. AAPG Bull. 86, 201–212. Levine, J.R., 1992. Petrologic controls on reservoir characteristics of coal beds. In: Ayers, W.B., Kaiser, W.R., Levine, J.R. (Eds.), Depositional, Hydrologic and
Petrologic Controls on Reservoir Characteristics of Coal Beds. Am. Assoc. Pet. Geol, 1992 Annual Convention, Short Course. (pp. 58–160). Liu, H.H., Sang, S.X., Wang, G.X, Li, Y.M., Li, M.X., Liu, S.Q., 2012. Evaluation of the synergetic gas-enrichment and higher-permeability regions for coalbed methane recovery with a fuzzy model. Energy 39, 426–439. Liu, S.B., Song, Y., Zhao, M.J., 2005. Influence of overpressure on CBM reservoir in south Qinshui basin. Chin. Sci. Bull. 50, 124–129. McLennan, J.D., Schafer, P.S., Pratt, T.J., 1995. A Guide to Determining Coalbed Gas Content. Gas Research Institute, Chicago, Illinois. (GRI-94/0396). Pashin, J.C., 1998. Stratigraphy and structural of coalbed methane reservoir in the United States: an overview. Int. J. Coal Geol. 35, 209–240. Pashin, J.C., 2007. Hydrodynamics of CBM reservoirs in the Black Warrior Basin: key to understanding reservoir performance and environmental issues. Appl. Geochem. 22, 2257–2272. Qin, S.F., Song, Y., Tang, X.Y., Fu, G.Y., 2005. The mechanism of the flowing ground water impacting on coalbed gas content. Chin. Sci. Bull. 50, 118–123. Qin, Y., Fu, X.H., Ye, J.P., Lin, D.Y., Tang, S.H., Li, G.Z., 1999. Geological controls and their mechanisms of coal-reservoir petrography and physics of coalbed methane occurrence in China. J. China Univ. Min. Technol. 28 (1), 14–19. (in Chinese with English abstract). Ryan, B., 1992. An Equation for Estimation of Maximum Coalbed-Methane Resource Potential: Geological Fieldwork 1991, Paper 1992-1. British Columbia Geological Survey, Columbia. Scott, A.R., 2002. Hydrogeologic factors affecting gas content distribution in coal beds. Int. J. Coal Geol. 50, 363–387. Song, Y., Zhao, M.J., Liu, S.B., Wang, H.Y., Chen, Z.H., 2005. The influence of tectonic evolution on the accumulation and enrichment of CBM. Chin. Sci. Bull. 50, 1–6. Song, Y., Liu, H.L., Hong, F., Qin, S.F., Liu, S., Li, G.Z., Zhao, M.J., 2012. Syncline reservoir pooling as a general model for coalbed methane (CBM) accumulations: mechanisms and case studies. J. Pet. Sci. Eng. 88–89, 5–12. Su, X.B., Lin, X.Y., Liu, S.B., Zhao, M.J., Song, Y., 2005. Geology of coalbed methane reservoir in the Southeast Qinshui Basin of China. Int. J. Coal Geol. 62, 197–210. Ulery, J.P., Molinda, G.M., 1984. Influence of Overlying Strata on Methane Emission in a Northern West Virgina Coal Mine. U.S. Dept. of the Interior, Bureau of Mines, Washington, DC. Wang, G.X., Zhang, X.D., Wei, X.R., Fu, X.H., Jiang, B., Qin, Y., 2011. A review on transport of coal seam gas and its impact on coalbed methane recovery. Front. Chem. Sci. Eng. 5, 139–161. Yao, Y.B., Liu, D.M., Tang, D.Z., Tang, S.H., Che, Y., Huang, W.H., 2009. Preliminary evaluation of the coalbed methane production potential and its geological controls in the Weibei Coalfield, Southeastern Ordos Basin, China. Int. J. Coal Geol. 78, 1–15. Zhao, M.J., Song, Y., Su, X.B., Liu, S.B., Qin, S.F., Hong, F., Lin, X.Y., 2005. The key stage and moment of coalbed gas reservoir evolution in the Qinshui Basin. China Chin. Sci. Bull. 50, 92–98.
Contents lists available at ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Block scale investigation on gas content of coalbed methane reservoirs in southern Qinshui basin with statistical model and visual map Huihu Liu a, Shuxun Sang b,n, Geoff G.X. Wang c, Mengxi Li d, Hongjie Xu a, Shiqi Liu b, Junjun Li e, Bo Ren f,g, Zhigen Zhao a, Yan Xie a a
School of Earth and Environment, Anhui University of Science & Technology, Huainan, Anhui 232001, PR China School of Resources and Earth Science, China University of Mining and Technology (CUMT), Key Laboratory of CBM Resources and Reservoir Formation Process, State Ministry of Education, Xuzhou, Jiangsu 221008, PR China c School of Chemical Engineering, The University of Queensland, Brisbane, St. Lucia Qld 4072, Australia d Shanxi Coalbed Methane Branch of Huabei Oilfield Company, Jincheng, Shanxi 048000, PR China e Production & Technology Department, Shanxi Lanyan Coalbed Methane Group Co., Ltd., Jincheng, Shanxi 048204, PR China f State Key Laboratory of Deep Coal Mining & Environment Protection, Huainan Mining Industry (Group) Co., Ltd., Huainan, Anhui 232001, PR China g National Engineering Research Center for Coal Gas Control, Anhui 232001, PR China, Huainan, Anhui 232001, PR China b
art ic l e i nf o
a b s t r a c t
Article history: Received 21 February 2013 Accepted 26 August 2013 Available online 10 December 2013
This study performs a block scale investigation on gas content of a coal reservoir in Zhengzhuang Block of the southern Qinshui basin in China. The gas content of Coal Seam No. 3 in this coal reservoir was measured in field and laboratory in conjunction with tests on coal properties such as adsorption isotherm, maximum vitrinite reflectance, coal composition and maceral component etc. Total 36 coal cores collected from 3 adjacent coalmines in the southern Qinshui basin were investigated, including analysis of logging data from the drilling wells. The investigations provided experimental data for block scale modeling and visualization analyses on the correlation between gas content and the key factors such as coal properties and geological conditions of the coal reservoir. Data obtained by field and lab tests were analyzed by statistical models in order to correlate gas content and individual type of coal properties and geological variables. The statistical model was then used to map the gas content of the target coal seam in the studied area, resulting in a flood map of gas content at a 1:50000 scale. The flood map was further visualized with other variables in terms of the properties of coal and coal reservoir and its geological conditions. These visualized maps provide useful geological interpretation for block scale investigation of the comprehensive relationships between the gas content and the coal properties and regional structure in the given coal reservoir. The results show that gas content has little correlation with coal rank, maceral composition, coal thickness, cap and bottom lithology, while it is highly related to the structural properties such as burial depth and effective cover thickness. A stagnant hydrodynamic condition is favorable to the higher gas content on the whole but does not contribute to gas lateral and local variation. Canonical correlation and principal component analysis on the statistical model reveal the key factors that control the gas content are burial depth, effective thickness of overlying strata, groundwater level and moisture content in coal seam. & 2013 Elsevier B.V. All rights reserved.
Keywords: coalbed methane gas content reservoir evaluation southern Qinshui basin statistical model visual map
1. Introduction Gas content of coalbed methane (CBM) reservoirs is one of most important indices for exploration and exploitation of CBM resource deposited in coal seams. The methane adsorption isotherm has been widely used to measure the gas content of coals (AQSIO and SAC, 2004a, 2008a). However measurement of the adsorption isotherm on coals cannot accurately represent the gas content of CBM reservoirs since the ideal laboratory conditions are different from the in-situ circumstance of the CBM reservoirs. To
n
Corresponding author. E-mail address: [email protected] (S. Sang).
0920-4105/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.petrol.2013.08.039
determine the gas content of CBM reservoirs, standardized field tests have to be carried out, consisting of estimating the lost gas content during sampling and measuring desorbed and residual gas contents based on the adsorption isotherm method (AQSIO and SAC, 2004b,2008b; Wang et al., 2011). The field experiments are integrated with other measurements of coal and reservoir properties, such as measurements of adsorption isotherm and gas component, the maceral composition, proximate and ultimate analyses, and tests for vitrinite reflectance and porosity. These measurements provide the experimental data for block scale investigation on the gas content of CBM reservoirs. Qinshui basin is the largest basin for CBM bearing resource as well as the most active area for exploration and development of coalbed methane in China. It is located in Shanxi province of
2
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
northwest China, characterized by one of the deposited most abundant anthracite coal reservoirs in the world. Currently, there are about 3000 CBM wells completed in China and most of them are located in the southern Qinshui basin, making up over 95% of the total CBM wells in the country. It can be predictable that there will be increasing research activities and investments on projects for CBM recovery in this basin. However, the studies on block scale evaluation of gas content in the whole Qinshui basin are still limited (Qin et al., 2005; Song et al., 2005; Zhao et al., 2005; Cai et al., 2011; Song et al., 2012), in particular for the southern part of Qinshui basin (Zhao et al., 2005, Cai et al., 2011; Song et al., 2012). Among these studies, only a few attempted to investigate the effect of various geological factors on gas content (Cai et al., 2011). These studies investigated the control mechanism on gas content from different perspectives, mainly based on large-scale CBM basins. The distribution of gas content in the field, which is essential for optimizing well pattern and minimizing investment risk for CBM recovery, has not been fully understood and needs to be investigated further. As well known, gas content of CBM reservoirs is generally controlled by comprehensive geological conditions, including tectonic setting, depositional environments (Ayers, 2002; Song et al., 2005; Song et al., 2012), hydrological features (Kaiser et al., 1994; Scott, 2002; Pashin, 2007), reservoir pressure (Liu et al., 2005), burial depth (Su et al., 2005; Liu et al., 2012; Song et al., 2012), coal quality and rank (Langenberg, et al., 2006), and coal petrology (John et al., 1989; Levine, 1992; Scott, 2002). Despite many attempts which have been made to measure gas content of various coals, and to discuss a general relationship between the gas content of coal and the coal geological conditions in basin
scale, the quantitative relationship between gas content and the geological factors has not been clear particularly in the developing CBM field. This uncertainty makes it difficult to evaluate the CBM potential capacity of CBM reservoirs in a block scale, and hence increases the risk of investments for the exploitation of a CBM field. This paper presents a block scale investigation on gas content of the CBM reservoir in Zhengzhuang Block of the southern Qinshui basin by systematic sampling of coal samples and comprehensive analyses of gas content data measured in field and laboratory.
2. Geological setting Fig. 1 shows the studied area with the primary geological information such as faults and well locations for sampling. Coalbearing strata of the studied area consists of Benxi, Taiyuan, Shangshihezi and Shiqianfeng formations of the PermoCarboniferous from bottom to top, and Coal Seam No. 3 in the studied area is developed in Shanxi formation, as indicated in Fig. 2. Shanxi formation comprises mudstone, sandstone, sandy mudstone and coal deposited in deltaic environments, such as mouth bar, interdistributary estuary and delta plain, of Paleozoic North China Epicontinental Sea. Its thickness is 50 m in average, ranging from 39 to 79 m. The coal-bearing strata in the area mainly experienced four periods of tectonic movement: Hercynian period, Indosinian period, Yanshanian period and Himalyan orogenic period. During the Hercynian–Indosinian period, the area experienced a long
Fig. 1. Tectonic setting of Zhengzhuang Block in southern Qinshui basin.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
3
(Song et al., 2005; Zhao et al., 2005). These tectonic movements directly resulted in the current Qinshui morphology, which forms a fundamental impact on CBM preservation. Correspondingly, during the Indosinian period, folds with E–W strike were formed and extensional tectonics such as normal faults with NE–SW strike were developed under the early compressive tectonic stress parallel to the normal faults strike. In the Yanshanian period, a synclinore striking NE, the basic tectonic shape of the whole basin was formed under NW–SE compression stress, which was accompanied by a series of folds striking NNE and joints striking NW or EW. The Himalyan period was featured by a NNE–SSW compression stress, which led to the folds striking NW combined with early folds with NE strike. Finally, during the Quaternary period of the late Himalyan period, a few of Epigenetic minor folds with NW strike were formed due to approximately horizontal and compressive tectonic stress (Fig. 1).
3. Methodology 3.1. Prospecting data
coal seam NO.3
Data such as coal bed thickness, burial depth, coal roof lithology, cap thickness, and coal bed roof elevation was obtained by drilling tests with the aid of well logging measurements. The well logging tests mainly detected resistivity, spontaneous potential and Gamma ray with the logging intervals including Shangshihezi formation, Xiashihezi formation, Shanxi formation, and Taiyuan formation from top to bottom of Permo-Carboniferous system. The logging tests provided the well logging interpretation results such as the thickness, the location and the lithology of formation, coal quality such as ash content, moisture content, volatile content, porosity, gas content, and rock mechanics parameters such as Poisson' s ratio, shear modulus, pressure gradient, and ground stress. 3.2. Sampling and experiments
coal seam No.15
Fig. 2. Stratigraphic column showing coal-bearing formations in southrn Qinshin basin.
Fig. 3. In-situ temperature of coal samples in Coal Seam No. 3 in southern Qinshui basin.
burial history and a sustainable sedimentation process. In the early time of the middle Yanshanian period, the area experienced a fluctuant burial process. From the late Yanshanian period to the Himalyan period, the area experienced a continuous uplifting
There are in total 36 coal samples of Shanxi formation collected from different well cores of Zhengzhuang Block and 3 underground coalmines named Sihe, Tangan, Chengzhuang in the southern Qinshui basin. Sampling sites are uniformly scattered in the studied area (see Fig. 1). All the coal samples were tested with well-established measurement methods including measurements of adsorption isotherm and gas component, the maceral, proximate and ultimate analyses, and tests for vitrinite reflectance and porosity etc. Canister desorption method was used for measurement of desorbed gas from coal core samples in field tests by means of a standard desorption device. The measurement was conducted through the procedures as described by the National Standard GB/T 19559-2008 in China. According to the standard, prior to putting a coal sample into the canister, the canister was flushed with argon to remove the air. During the natural desorption, desorption capacity was gauged in given time intervals. The natural desorption measurement was stopped when the average desorption capacity was less than 10 cm3 per day within 7 days. Sometimes the measurement needed to be carried out for about 3 months because of the slow diffusion/desorption of gas in coal. After measuring the natural desorption, a part of the coal samples were chosen to grind in a ball mill for 2–4 h, then they were put into an indoor canister. Subsequently, the canister was put into a thermostat and its temperature adjusted to the same as the reservoir coal. The residual gas capacity was measured. The escaped gas content was calculated according to the direct method proposed by US Bureau of Mine (USBM) (McLennan et al., 1995). It
4
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Table 1 Properties of interest for coal samples used. Sample number
Moisture (ad%) Ash (d%) Volatiles (ad%) Critical desorption pressure (MPa) Langumiur volume (cm3/g, ad) Langumiur pressure (MPa, ad) Gas content (cm3/g, ad)
No. 1
No. 2
No. 3
No. 4
No. 5
No. 6
No. 7
No. 8
…
No. 36
1.11 11.75 6.10 8.73 36.44 3.55 26.99
1.62 9.23 5.94 4.81 41.44 3.04 25.81
1.39 12.36 7.14 7.74 37.56 3.2 26.69
1.21 10.59 6.40 3.82 37.54 3.05 22.51
1.33 12.50 6.72 3.03 2.2 37.03 21.48
1.36 10.96 6.39 3.28 39.5 2.8 21.4
1.23 9.79 6.35 6.17 40.97 3.69 25.74
1.91 13.63 6.74 6.26 31.02 2.78 22.23
… … … … … … …
1.28 13.27 7.13 1.15 40.7 3.41 13.44
y = 0.0124x + 13.456 2 R = 0.7006
30
3
Gas content (m /t)
35
25 20 15 10 400
Fig. 4. Typical adsorption isotherms of selected coal samples (referring to Fig. 1 for sampling points) based on air dry at 30 1C.
600
800 1000 Burial depth (m)
1200
1400
30
3
Gas content (m /t)
35 is usually determined by the positive relationship between desorption volume of coalbed methane and time. Gas content was determined by the sum of the escaped gas content, the natural desorption and the residual capacity, generally expressed on the basis of ash, water free. After measurements of the gas content, each sample remainder was split into several aliquots (using the standard coning and quartering technique) and prepared for analyses such as proximate analysis, coal petrology (maceral composition and vitrinite reflectance), specific surface area and pore structure. Proximate and maceral composition analyses were conducted according to Chinese National Standard GB/T 212-2001 and GB/T 8899-1998, respectively. Pore structure and specific surface area analyses abided by Chinese Petroleum and Natural Gas Industry Standards SY/T 5346-2005 and SY/T6154-1995, respectively. AXIO Imager M1m Microspectrophotometer, ASAP 2000 and Microstructure Analyzer 9310 were used for measurements of maceral composition, specific surface area and pore structure. The collected coal samples were analyzed for gas composition using gas chromatograph (GC) in coordination with the National Standard procedures of GB/T 13610-2003. The adsorption isotherm was measured by referring to Chinese National Standard GB/T 19560-2004, using the Isotherm Adsorption/Desorption System IS-100. Coal samples for isotherm measurements were prepared by crushing and screening fresh air-dried grains to the particle size range of 0.18–0.25 mm. Since in-situ reservoir coal in the target coal seam was deposited under a condition of temperature at around 30 1C (Fig. 3) and water content would be saturated in coal, the isotherm measurements were performed at 30 1C under the moisture equilibrium condition.
25 20 15 10 400
600 800 1000 1200 Overlying formation thickness (m)
1400
Fig. 5. Relationship of gas content with (a) burial depth; and (b) effective thickness of overlying strata in Zhengzhuang Block.
investigations as described above were used for modeling by the statistical method to correlate gas content with individual type of geological factors in order to predict gas content for CBM exploration and exploitation. 3.4. Visual analysis on gas content and the key factors The field and laboratory investigations show that the gas content can be correlated statistically to the properties of coal and coal reservoir and its geological conditions, which can usually be expressed as variables of a given coordinate plane. Thus the correlations shown in the predicted results from the statistical model and the experimental results of gas content can be used to map the gas content of the target coal seam in the studied area.
3.3. Statistical model for gas content 4. Results Data from experiment test and prospecting were conducted to correlation analysis on gas content and the different factors. Therefore, data obtained by well logging and laboratory
Table 1 summarizes some properties of interest for representative coal samples from a total of 36 samples for simplicity. Note
35
35
30
30
Gas content (m /t)
25
3
3
Gas content (m /t)
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
20 15 10
9
12
15 Aad (%)
18
21
30
30 3
25 20 15 10
0.5
1
1.5 Mad (%)
2
70 80 Vitrinite (%)
90
100
10
20 30 Inertinite (%)
40
50
25 20 15 10
0
2.5
0
Fig. 8. Variation of gas content with (a) vitrinite and (b) inertinite macerals in Zhengzhuang Block.
35 Gas content (m /t)
30 3
25 20 15 10 5
30 25 20 15 10 5
0 3.3
3.5
3.7 Ro,max (%)
3.9
0
4.1
35
30
30
3
Gas content (m /t)
35
25 20 15 10 5 0
60
5 0
35
3
10
35
Fig. 6. Variation of gas content with (a) ash content (Aad); and (b) moisture content (Mad) in Zhengzhuang Block.
Gas content (m /t)
15
35
0
3
20
0 50
24
Gas content (m /t)
3
Gas content (m /t)
6
5
Gas content (m /t)
25
5
5 0
5
7
8
9
Vdaf (%) Fig. 7. Variation of gas content with (a) maximum vitrinite reflectance (Ro,max) and (b) volatile content (Vdaf) in Zhengzhuang Block.
that adsorption properties (i.e. Langmuir volume and pressure) were measured using the adsorption isotherm as shown in Fig. 4. As can be seen, Langmuir volumes of these coal samples vary from 32.94 m3/t to 50.55 m3/t with 43.11 m3/t on average and Langmuir pressures are 2.21–3.93 MPa with the average of 2.96 MPa. Gas content in the studied area remarkably changes in ranges of
5 10 Immediate roof thickness (m)
15
0
5 10 Immediate bottom thickness (m)
15
25 20 15 10 5 0
6
0
Fig. 9. Relationships of gas content with thickness of (a) immediate roof and (b) immediate bottom in Zhengzhuang Block.
8.03–31.44 cm3/g with an average of 20.82 m3/t (air dried basis). The variation of gas content obviously reflects the heterogeneity of gas-bearing in the coal reservoir as well as the complex effect of multiple geological factors on gas enrichment. The correlation analysis is often used to determine the relationship gas content and other influenced factors (Gentzis et al., 2006).
6
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
and bottom determines the sealing condition of CBM, which should be considered in filed scale investigation on gas content of coal seam. According to linear regression analyses, a simple statistical model was expressed in vector form as
Figs. 5–9 show the results of coal core analyses, correlating the gas content to the various factors that may affect the gas content of the coal seam. The relationships between gas content and burial depth, effective thickness of overlying strata in the studied area are shown in Fig. 5a and b. As can been seen, gas content has an obvious correlation with both burial depth and effective thickness of overlying strata. Ash content (Ad) in the studied area varies between 9.23% and 21.63% with the average of 12.91%. Relationship of gas content and ash content (Aad) is shown in Fig. 6a, implying that gas content has a relatively obvious negative correlation with Aad. Moisture content (Mad) in the studied area varies between 0.71% and 1.91% with the average of 1.36%. Likewise the relationship between gas content and Mad shows a negative correlation (Fig. 6b). Fig. 7 shows the variation of gas content with maximum vitrinite reflectance (Ro,max) and volatile matter (Vdaf) content. As can been seen, the correlation between gas content and coal maturity is not obvious (Fig. 7a) whilst the gas content is generally decreased with volatile matter content (Fig. 7b). The relationship of gas content with maceral composition is not obvious too, but the gas content seems to be increased with the increase of vitrinite and decreased with the increase of inertinite (Fig. 8). Finally, the relationships between gas content and thicknesses of the immediate roof and the immediate bottom were analyzed, as indicated in Fig. 9. It shows that there are no obvious correlations between them. However the immediate roof
y ¼ Xβ þ ε
ð1Þ
where y is a vector consisting of measured gas content yi (i ¼1,2,…, n) (m3/t) from n coal samples; X represents the defined matrix comprised from the input variables xnm as follows: 0 1 x11 ⋯ x1m B x21 … x2m C B C ð2Þ X¼B C ⋱ ⋮ A @ ⋮ xn1 ⋯ xnm with m numbers of the measurable values of the key variables that may affect the gas content; β is a m-dimensional parameter vector consisting of regression coefficients; and ε is a vector formed by the error term, or so-called disturbance term.β and ε can be calculated based on the data shown in Table 2 with the least square method. In this study, a two-step regression approach was used to formulate the statistical model above. Firstly the 12 key factors were chosen for the modeling, i.e. m ¼ 12, including immediate roof thickness (xi1), immediate bottom thickness (xi2), coalbed
Table 2 Data used for statistical analysis. Sample no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 ⋮ 36
Vitrinite content
Inertinite content
Ground water level
wt%
Ash content (Aad) wt%
wt%
wt%
m
1.11 1.62 1.39 1.21 1.33 1.36 1.23 1.91 1.62 1.86 1.37 1.28 1.04 1.28 0.85 0.81 1.30 1.33 1.16 ⋮ 1.46
11.75 9.23 12.36 10.59 12.5 10.96 9.79 13.63 14.34 11.52 9.68 10.51 10.81 11.44 10.62 10.36 12.84 15.60 10.78 ⋮ 10.39
75.40 73.80 71.30 69.80 67.50 79.30 89.20 63.20 61.50 60.90 72.80 69.00 87.90 92.30 83.40 68.00 69.30 86.30 81.61 ⋮ 66.40
24.60 25.60 28.70 30.20 32.50 20.70 10.80 36.80 38.50 39.10 27.20 31.00 12.10 7.70 16.60 32.00 30.70 13.80 18.39 ⋮ 33.60
503 498 535 496 502 501 510 514 524 542 539 587 557 586 590 603 598 540 535 ⋮ 582
Gas Immediate content roof
Immediate bottom
Coalbed thickness
Coal rank
Burial depth
Overlying strata
Volatile Moisture content (Vdaf) content (Mad)
m3/t
m
m
m
%
m
m
vol%
26.99 25.81 26.69 22.51 21.48 21.4 25.74 22.33 20.28 21.7 23.81 26.65 28.5 25.03 30.04 31.44 28.45 20.94 22.8 ⋮ 22.02
1.92 2.51 9.09 4.16 4.20 5.06 1.05 0.18 11.9 4.00 4.00 3.05 2.54 5.89 11.15 14.31 10.02 6.48 5.90 ⋮ 2.99
1.06 1.55 10.94 11.56 6.29 0.51 1.07 0.24 10.37 2.91 11.81 4.04 1.03 11.91 10.74 2.24 6.10 1.14 9.51 ⋮ 2.80
5.50 4.90 5.80 5.30 6.00 5.35 5.20 5.30 5.60 5.60 4.70 5.70 5.75 6.00 6.10 6.20 5.90 6.30 6.90 ⋮ 6.10
3.85 3.88 3.47 3.83 3.91 3.89 3.88 3.81 3.61 3.62 3.67 3.53 3.49 3.62 3.77 3.68 3.44 3.50 3.50 ⋮ 3.53
980.4 638.5 960.3 753.7 652.6 659.95 961.5 900.6 746.8 751.0 888.9 1267.9 1038.5 1114.5 1059.2 1156.5 1336.9 513.6 626.4 ⋮ 1117.1
906.7 619.6 921.4 717.7 629.5 634.3 922.7 890.6 718.8 748.4 860.1 1249.9 1010.5 1109.9 1038.2 1111.5 1299.1 473.0 596.4 ⋮ 1030.5
6.10 5.94 7.14 6.40 6.72 6.39 6.35 6.74 7.48 6.81 6.44 7.11 6.63 6.69 6.74 8.16 7.95 7.52 7.17 ⋮ 6.58
Table 3 Correlation coefficients of gas content with different factors based on canonical correlation and principal component analysis. Analysis method
Immediate roof thickness
Immediate bottom thickness
Coalbed thickness
Coal rank
Burial depth
Effective thickness of overlying strata
Volatile content
Moisture content
Ash content
Vitrinite content
Inertinite content
Ground water level
Canonical relation
0.335
0.008
0.091
0.087
0.779n
0.768n
0.078
0.625n
0.469
0.159
0.159
0.633n
Principal component
0.335
0.008
0.091
0.087
0.779n
0.768n
0.078
0.625n
0.469
0.159
0.159
0.633n
n
Means the factors have an obvious correlation with gas content at 0.01 confidence level during significance test.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
7
Fig. 10. Flood map of gas content predicted for Coal Seam No. 3 of Zhengzhuang Block in southern Qinshui basin.
Fig. 11. Visual relationship between gas content and coal thickness of Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: m for coal thickness and m3/t for gas content with different color areas.)
8
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
thickness (xi3), coal rank (i.e. Ro,max) (xi4), burial depth (xi5), effective thickness of overlying strata (xi6), volatile content (xi7), moisture content (xi8), ash content (xi9), vitrinite content (xi10), inertinite content (xi11) and groundwater level (xi12). As a result of applying the data in Table 2 into Eqs. (1) and (2), the first-step regression eqution can be determined, giving 12
Y ¼ ∑ Aj X j þ B j¼1
ð3Þ
where Y represents gas content (m3/t); Aj (j¼1,2,.., 12) and B are the regressed coefficients. In order to identify the significance of these correlations to predict gas content, the canonical correlation analysis (CCA) and principal component analysis (PCA) were further performed for the first-step regression eqution above. The details can be found in Table 3. It can be seen from these statistical analyses that gas content shows little correlation with coal thickness, coal rank and coal maceral composition. It has a relative strong correlation with other factors in small-scaled CBM field. As shown in Table 3, the correlation coefficients between gas content and burial depth, effective thickness of overlying strata, groundwater level and moisture content are above 0.5, showing the strong effects of these four geological factors on the gas content. While the first three factors have an obvious positive correlation with gas content, the fourth factor, i.e. moisture content obviously exhibits a negative effect on gas content. To pursue quantitative recognition
for coupling control of these dominant factors on gas content, the second-step regression was conducted by taking these four factors into account. Thus the resultant statistical model was expressed in a simplified format of Eq. (3), giving Y ¼ 0:085D þ0:0007T 4:7708M ad þ 0:0015L þ23:675
ð4Þ
where D is burial depth (m); T is effective thickness of overlying strata (m); L is groundwater level (m) and Mad is moisture content (wt%). The fitting degree of the statistical model was conducted. The results show that the fitting degree increases with the amount of the factors adopted in the model. The estimated residual standard deviations of Y are 1.82% and 1.45%, the fitting degrees are 75.12% and 92.63%, when 4 key factors and 12 key factors were considered respectively. Significant inspection F-statistics of the models are 11.32 and 7.34, which are above the critical values (F0.05 (4,15) ¼ 5.86, F0.05 (12,7) ¼2.91), whilst the p-values of the models are 0.0002 and 0.0069, which are far below 0.05, when 4 key factors and 12 key factors were considered respectively, Those demonstrate the models achieve a significant level. The geological conditions are related to the location, which can usually be expressed as variables of a given coordinate plane. Thus the correlations shown in the predicted results from Eqs. (3) and (4) in statistical model can be used to map the gas content of the target coal seam in the studied area. For this purpose, the resultant gas content was firstly plotted in the flood map (Fig. 10) at a 1:50000 scales, defined by the coordinates of sampling boreholes
Fig. 12. Visual relationship between gas content and burial depth of Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: m for burial depth of Coal Seam No. 3 and m3/t for gas content with different color areas.)
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
in the studied area (Fig. 1). Then different contour maps were drawn as Figs. 11–17 in order to visualize the correlations between gas content and various variables of interest such as thickness and burial depth of coal seam, effective thickness of overlying strata, moisture and ash content of coal, etc. These visualized maps provide useful information for block scale evaluation of the comprehensive relationships between distribution of the gas content and the coal reservoir properties and regional structure, which will be discussed in details as follows.
5. Discussion 5.1. Gas content vs. reservoir properties 5.1.1. Coal thickness The well tests show that coal thickness is about 4.5–6.5 m in the studied area (Fig. 11). The coal seam gas is unevenly distributed in the target coal seam in the studied area, more in the north compared with the south. As can be seen, the distribution of gas content is apparently inconsistent with the distribution of the coal thickness. This implies that coal thickness has less control effect on gas content. Because stratigraphy and geologic structure determine the continuity of coal (Pashin, 1998), coal thickness trends and depositonal fabric strongly influence the distribution of gas content (Scott, 2002). Variation of coal thickness is relatively small
9
in this developing field and the effect of coal thickness on gas content is mainly dependent on depositional environment that has no obvious difference in the whole coalfield.
5.1.2. Tectonic effect The thermal evolution of coal seam in the studied area experienced two hydrocarbon generations in the late Triassic and late Jurassic–early Cretaceous. The coalification degree and gas generation reached the maximum value due to the abnormal thermal event in the Yanshanian period (Song et al., 2005; Zhao et al., 2005). Tectonic uplift has determined the preservation degree of coalbed methane since the late Yanshanian period. The present gas content is mainly dependent on effective thickness of overlying strata on coal seam during the key geological period (Zhao et al., 2005). In the studied area, different types of folds and faults were well developed due to tectonic movement. EN–WS trending F2 Fault was identified (Fig. 1). The gentle and approximate parallel folds were developed widely in the developing area, trending towards NNE and SN. The analysis shows that the gas content of coal seam located in syncline is relatively high, and gas content located in anticline and fault zone is relatively low (Fig. 10). In other words, multi-period tectonic stress field determines the development of folds and faults, and tectonic style, in terms of folds and faults, controls enrichment and dissipation of coal seam gas.
Fig. 13. Visual relationship between gas content and distribution of ash content in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: wt% for ash content and m3/t for gas content with different color areas.)
10
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Fig. 14. Visual relationship between gas content and distribution of moisture content in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: wt.% for moisture content and m3/t for gas content with different color areas.)
5.1.3. Burial depth Many case studies revealed that burial depth is an important factor that affects gas potential of coal seams (Su et al., 2005; Liu et al., 2012; Song et al., 2012). The gas content was usually lower if the coal bearing had ever suffered serious erosion when coal bed was in a shallow buried stage. In this case, coal seam gas was easily migrated because of the strata uplift in geological history, even if coal bed has a deep burial depth later. Therefore the effective thickness of overlying strata would play a more important role in controlling the gas content (Ulery and Molinda, 1984). An effective thickness of overlying strata is defined by the bottom thickness from coal bed to the first unconformity where CBM was largely generated. It reflects tectonic uplift and erosion on the CBM preservation. Theoretically speaking, it can provide the superior geological conditions for CBM enrichment compared with overlying bedrock thickness and burial depth. As a result, higher gas content usually appears in areas with thicker overlying strata. Burial depth of coal seam in the developing field of the studied area varies from 500 m to 1200 m. The correlation between burial depth and gas content is described in Fig. 12, consisting of the flood map of gas content together with the contours of burial depth. As can be seen, distribution of gas content in this studied area is generally consistent with the contours of burial depth.
5.1.4. Overlying strata The analysis of strata thickness and residual strata thickness shows that, in most part of the studied area, Triassic and
Quaternary strata is missing due to tectonic uplifts, and Triassic strata is retained the form of Shiqianfeng strata in partial area of the field. Quaternary strata and residual Shiqianfeng strata are relatively shallow by contrast. In general, the effective thickness of overlying strata is a little different from burial depth because of some small uplifts and deposition of the field in geological history. Therefore strata thickness controls the gas content almost in the similar way as the burial depth. However, the correlation of the effective thickness of overlying strata with gas content in the studied area seams weaker than that between the burial depth and gas content, as shown in Figs. 5 and 12. The results suggest that the denudation and deposition caused by the tectonic uplifts and sedimentation has a certain influence on gas content. As a result, the gas content in the studied area almost remained the same until the decline stage, which would be mainly controlled by burial depth of the coal seam. 5.2. Gas content vs. coal properties 5.2.1. Coal quality Generally speaking, if the ash content in coal is low, there will be more gas generation. This is because the gas production in coalification would be higher due to the less inorganic substance in the source of coal during evolution. The effect of ash content on gas content is illustrated in Fig. 6a. As can be seen, gas content has a relatively obvious negative correlation with ash content, in line with the fact that total surface area of pores and adsorption
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
11
Fig. 15. Visual relationship between gas content and distribution of Ro,max in Coal Seam No. 3 in Zhengzhuang Block. (Units of contour lines: % for Ro,max and m3/t for gas content with different color areas.)
capacity declines with the increasing of gas content. That is, the higher mineral particles such as clay minerals, pyrite brought by syngenetic sedimentation, epigenetic sedimentation or magmatic intrusion are filled in pores, and hence decrease the adsorption and migration capacity of coal seam gas. Meanwhile, the relationship of gas content and ash content reflects the depositional effect and tectonic synergistic effect. Thus ash content is a relatively important factor that affects the gas content. The effect of coal quality on distribution of gas content in the target coal seam can be further visualized by plotting the flood map of gas content with the contours of ash content, as shown in Fig. 13. It can be seen that the distribution of gas content in the studied area is correlated with the distribution of ash content to a certain extent. The variation of gas content from the north to the south in the studied area is highly consistent with the variation of ash content in the same direction and trend. Moisture content represents another important index of coal quality that significantly affects the gas content, as shown in Fig. 6b. This is because water molecule has a polarity that can be preferentially absorbed on the pore surface of coal, and can partially replace adsorption sites of methane (Clarkson and Bustin, 2000). Therefore, high moisture content in coal can lower adsorption capacity of coal, which can generally lead to lower gas content. The visual correlation between moisture and gas content illustrated in Fig. 14 shows that the distribution of gas content is
largely negatively correlated with distribution of moisture content. The variation of gas content from the north to the south in the studied area is also highly consistent with the variation of moisture content in the same direction and trend.
5.2.2. Coal rank Coal rank is one of the key factors that affect adsorption capacity of coal (Laxminarayana and Crosdale, 2002). Many researchers have shown that total adsorbable methane capacity increases with coal rank, which has been considered the main factor influencing the methane adsorption capacity (Ryan, 1992; Crosdale et al., 1998; Scott, 2002; Yao et al., 2009). As shown in Fig. 7a, the maximum vitrinite reflectance (Ro,max) of coal in the studied area varies from 3.2% to 4.0%; and the volatile matter (Vdaf) content varies between 5.38% and 8.41%. The correlation of gas content and coal maturity is not obvious, which is not consistent with the evolution of coal rank proposed by Qin et al. (1999), and demonstrates the relatively small difference of coal maturity and effects from other factors in the developing field; and the effect of coal maturity on gas content is relatively weak. However, the relationship between gas content and volatile matter content shows an obvious negative correlation (Fig. 7b). This indicates that gas content increases with the decrease of volatile matter (Vdaf) and probably the increase of coal rank. The visual relationship between the distribution of gas content and coal rank in
12
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Fig. 16. Visual relationships between gas content with the distribution of immediate roof thickness in Coal Seam No. 3 in Zhengzhuang Block. (Unit of contour lines: m for immediate roof thickness of Coal Seam No. 3 and m3/t for gas content with different color areas.)
terms of the maximum vitrinite reflectance (Ro,max) of coal is illustrated in Fig. 16. As can be seen, the distribution of gas content seems to have no correlation with coal rank in the block scale.
5.2.3. Coal maceral composition The gas generation potential is related to the maceral composition directly (Scott, 2002). For the relationship between the maceral composition and adsorption capacity, coal maceral composition and coal maturity determine surface area, pore size, and distribution of pore volume in coal (John et al., 1989). The coal in the studied area is rich in vitrinite, vitrinite varies between 53.9% and 92.7%, inertinite between 7.3% and 46.15%. The relationship between gas content and maceral composition indicates that gas content increases with the increase of vitrinite, and decreases with the increasing of inertinite in general (Fig. 8). Meanwhile, the correlations between gas content and the maceral content indicate that adsorption capacity of coal depends on the content of vitrinite, i.e. the vitrinite-rich coal has strong adsorption capacity and hence leads to high gas content. However, the inconspicuous correlation between gas content and the maceral composition implies that the influence of the maceral composition on gas content is so complicated that it is difficult to evaluate without more detailed information about the comprehensive effects of coal rank, moisture content, ash content, tectonic setting, hydrological features, and depositional setting (or coal facies) (Pashin, 1998;
Crosdale et al., 1998; Pashin, 2007; Faiz et al., 2007; Yao et al., 2009; Song et al., 2012). This needs to be further studied. 5.3. Gas content under different geological circumstances 5.3.1. Sealing condition The lithology and thickness of surrounding rocks of coal directly affect the sealing of CBM as well. The immediate roof and the immediate bottom of coal mainly consist of mudstone and silty mudstone, siltstone and fine-grained sandstone in the studied area. The mudstone thickness of roof or bottom of Coal Seam No. 3 is above 10 m (Fig. 2). Mudstone is quite tight and no fracture was observed. It can prevent CBM from migrating through diffusion, seepage and water-soluble flow. Therefore the sealing condition of the developing field is believed very well. The field measurements reveal no obvious correlation between gas content with immediate roof thickness and immediate bottom thickness (Fig. 9). This implies that sealing condition of CBM in terms of immediate roof thickness and immediate bottom thickness seems to have no influence on gas content as the coal seam in the studied area is well sealed. However the visual correlation between immediate roof thickness and gas content illustrated in Fig. 16 shows that the distribution of immediate roof thickness is consistent with gas content to a certain extent. The effect of sealing condition on gas content might reflect the depositional control because depositional setting determines overburden thickness and coal floor in general.
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
13
Fig. 17. Visual relationship between gas content of Coal Seam No. 3 and hydrodynamic condition in Zhengzhuang Block. (Unit of contour lines: m for groundwater level and m3/t gas content with different color areas.)
5.3.2. Hydrodynamic conditions Basin hydrodynamic conditions have a strong influence on reservoir properties and play a vital role in environmental issues (Scott, 2002; Pashin, 2007). The south of Qinshui basin is a catchment area. Overlying strata of Coal Seam No. 15 is mainly thick limestone of 2–15 m (see Fig. 2), and there is no hydrodynamic connection between Coal Seam Nos. 3 and 15. These coal seams are separated by multiple impermeable layers. Multiple hydrodynamic systems exist in the basin, controlled by the watershed, the process of groundwater recharge, migration, and discharge happened mainly at Qinshui basin margin. Hydrodynamic condition of the flow intersectional region was formed in the southern basin. This made the southern basin stagnant, and thus formed rich gas region. As shown in the visual relationship between gas content of Coal Seam No. 3 and hydrodynamic condition (Fig. 17), the developing field in the studied area is just located in the stagnant zone. As a result, a stagnant hydrodynamic condition is favorable to the higher CBM content on the whole but has no contribution to gas lateral and local variation (Su et al., 2005; Cai et al., 2011).
analyses at a block scale. Statistical analyses reveal that burial depth, effective thickness of overlying strata, groundwater level and moisture content in coal seam are dominant factors that control gas content in block scale. The results show that gas content has little or no correlation with coal rank, maceral composition, coal thickness, and immediate bottom thickness. The factors such as coal thickness, coal rank and sealing condition do not significantly affect the distribution of gas content in the studied area. While the gas content has a relatively high negative correlation with coal quality in terms of ash and moisture contents, the distribution of gas content is highly consistent with the variation of ash content, moisture content in the coal seam. Moreover, gas content has an obvious correlation with burial depth, and effective thickness of overlying strata. The effects of tectonic setting, burial depth, and effective thickness of overlying strata on gas content are dominant. A stagnant hydrodynamic condition is very favorable to CBM preservation, but has minor effect on the distribution of gas content in the developing field.
Acknowledgments 6. Conclusions A comprehensive investigation on gas content of Coal Seam No. 3 in Zhengzhuang Block of the southern Qinshui basin, China, was conducted by means of the statistical models and visualization
The research was supported by National Natural Science Fund of China (Grant nos. 41302129, 41330638, 41272154, 4117238 and 41272171), the PhD Fund of Anhui University of Science and Technology (Grant no. 11224), Project 2012012008 of Shanxi Province United Research Funding of Coalbed Methane and the
14
H. Liu et al. / Journal of Petroleum Science and Engineering 114 (2014) 1–14
Scientific Research Program of Huabei Oilfield Company. Geoff Wang acknowledges Australian Research Council (ARC) Discovery project for financial support on this collaborative research between the China University of Mining and Technology and the University of Queensland, Australia. References AQSIO, S.A.C., 2004a. GB/T 19560-2004 Experimental Method of High-Pressure Adsorption Isothermal to Coal-Capacity Method. Standards Press of China, Beijing. AQSIO, S.A.C., 2004b. GB/T 19559-2004 Method of Determining Coalbed Methane. Standards Press of China, Beijing, Beijing. AQSIO, S.A.C., 2008a. GB/T 19560-2008 Experimental Method of High-Pressure Isothermal Adsorption to Coal. Standards Press of China, Beijing, Beijing. AQSIO, S.A.C., 2008b. GB/T 19559-2008 Method of Determining Coalbed Gas Content. Standards Press of China, Beijing, Beijing. Ayers, W.B., 2002. Coalbed gas system, resources, and production and review of contrasting cases from the San Juan and Powder River basin. AAPG Bull. 86, 1853–1890. Cai, Y.D, Liu, D.M, Yao, Y.B, Li, J.Q, Qiu, Y.K, 2011. Geological controls on prediction of coalbed methane of No. 3 coal seam in Southern Qinshui Basin, North China. Int. J. Coal Geol. 88, 101–112. Clarkson, C.R., Bustin, R.M., 2000. Binary gas adsorption desorption isotherms: effect of moisture and coal composition upon carbon dioxide selectivity over methane. Int. J. Coal Geol. 42, 241–271. Crosdale, P.J., Beamish, B.B., Valix, M., 1998. Coalbed methane sorption related to coal composition. Int. J. Coal Geol. 35, 147–158. Faiz, M., Saghaf, A., Sherwood, N., Wang, N., 2007. The influence of petrological properties and buiral history on coal seam methane reservoir characterization, Sydney Basin, Australia. Int. J. Coal Geol. 70, 193–208. Gentzis, T., Schoderbek, D., Pollock, S., 2006. Evaluating the coalbed methane potential of ghe Gething coals in NE British Columbia, Canada: an example from the Highhat area, Peach River coalfield. Int. J. Coal Geol. 68, 135–150. John, F.U., Colin, S.F., Leslie, F.J., 1989. Moisture in coal: 2. Maceral effects on pore structure. Fuel 68, 18–26. Kaiser, W.R., Hamilton, D., Scott, A.R., Tyler, R.J., 1994. Geological and hydrological controls on the producibility of CBM. J. Geol. Soc. London 51, 341–372. Langenberg, C.W., Beaton, A., Berhane, H., 2006. Regional evaluation of the coalbedmethane potential of the Foothills/Mountains of Alberta, Canada. Int. J. Coal Geol. 65, 114–128. Laxminarayana, C., Crosdale, P.J., 2002. Controls on methane sorption capacity of Indian coals. AAPG Bull. 86, 201–212. Levine, J.R., 1992. Petrologic controls on reservoir characteristics of coal beds. In: Ayers, W.B., Kaiser, W.R., Levine, J.R. (Eds.), Depositional, Hydrologic and
Petrologic Controls on Reservoir Characteristics of Coal Beds. Am. Assoc. Pet. Geol, 1992 Annual Convention, Short Course. (pp. 58–160). Liu, H.H., Sang, S.X., Wang, G.X, Li, Y.M., Li, M.X., Liu, S.Q., 2012. Evaluation of the synergetic gas-enrichment and higher-permeability regions for coalbed methane recovery with a fuzzy model. Energy 39, 426–439. Liu, S.B., Song, Y., Zhao, M.J., 2005. Influence of overpressure on CBM reservoir in south Qinshui basin. Chin. Sci. Bull. 50, 124–129. McLennan, J.D., Schafer, P.S., Pratt, T.J., 1995. A Guide to Determining Coalbed Gas Content. Gas Research Institute, Chicago, Illinois. (GRI-94/0396). Pashin, J.C., 1998. Stratigraphy and structural of coalbed methane reservoir in the United States: an overview. Int. J. Coal Geol. 35, 209–240. Pashin, J.C., 2007. Hydrodynamics of CBM reservoirs in the Black Warrior Basin: key to understanding reservoir performance and environmental issues. Appl. Geochem. 22, 2257–2272. Qin, S.F., Song, Y., Tang, X.Y., Fu, G.Y., 2005. The mechanism of the flowing ground water impacting on coalbed gas content. Chin. Sci. Bull. 50, 118–123. Qin, Y., Fu, X.H., Ye, J.P., Lin, D.Y., Tang, S.H., Li, G.Z., 1999. Geological controls and their mechanisms of coal-reservoir petrography and physics of coalbed methane occurrence in China. J. China Univ. Min. Technol. 28 (1), 14–19. (in Chinese with English abstract). Ryan, B., 1992. An Equation for Estimation of Maximum Coalbed-Methane Resource Potential: Geological Fieldwork 1991, Paper 1992-1. British Columbia Geological Survey, Columbia. Scott, A.R., 2002. Hydrogeologic factors affecting gas content distribution in coal beds. Int. J. Coal Geol. 50, 363–387. Song, Y., Zhao, M.J., Liu, S.B., Wang, H.Y., Chen, Z.H., 2005. The influence of tectonic evolution on the accumulation and enrichment of CBM. Chin. Sci. Bull. 50, 1–6. Song, Y., Liu, H.L., Hong, F., Qin, S.F., Liu, S., Li, G.Z., Zhao, M.J., 2012. Syncline reservoir pooling as a general model for coalbed methane (CBM) accumulations: mechanisms and case studies. J. Pet. Sci. Eng. 88–89, 5–12. Su, X.B., Lin, X.Y., Liu, S.B., Zhao, M.J., Song, Y., 2005. Geology of coalbed methane reservoir in the Southeast Qinshui Basin of China. Int. J. Coal Geol. 62, 197–210. Ulery, J.P., Molinda, G.M., 1984. Influence of Overlying Strata on Methane Emission in a Northern West Virgina Coal Mine. U.S. Dept. of the Interior, Bureau of Mines, Washington, DC. Wang, G.X., Zhang, X.D., Wei, X.R., Fu, X.H., Jiang, B., Qin, Y., 2011. A review on transport of coal seam gas and its impact on coalbed methane recovery. Front. Chem. Sci. Eng. 5, 139–161. Yao, Y.B., Liu, D.M., Tang, D.Z., Tang, S.H., Che, Y., Huang, W.H., 2009. Preliminary evaluation of the coalbed methane production potential and its geological controls in the Weibei Coalfield, Southeastern Ordos Basin, China. Int. J. Coal Geol. 78, 1–15. Zhao, M.J., Song, Y., Su, X.B., Liu, S.B., Qin, S.F., Hong, F., Lin, X.Y., 2005. The key stage and moment of coalbed gas reservoir evolution in the Qinshui Basin. China Chin. Sci. Bull. 50, 92–98.