Pore structure and compressibility of coal matrix with ... · structures, including pore size distribution, pore volume and pore connectivity, were significantly changed based on

Pore structure and compressibility of coal matrix withelevated temperatures by mercury intrusion porosimetry

Zhentao Li, Dameng Liu*, Yidong Cai, Yanbin Yao and Hui WangCoal Reservoir Laboratory of National Engineering Research Center of CBM Development

& Utilization, China University of Geosciences, Beijing 100083, China*Author for corresponding. E-mail: [email protected]

(Received 11 February 2015; Accepted 16 August 2015)

AbstractTo gain a better understanding of the effect of heat (e.g., magma intrusion,geothermal fluids and enhanced coal-bed methane recovery process) on coalreservoir properties, the pore structure and compressibility of coal matrix forlow rank coal (0.69% Ro, m) with elevated temperatures were investigated byusing multiple methods, including thermogravimetry-mass spectrometry (TG-MS), scanning electron microscope (SEM), N2 adsorption/desorption at 77 Kand mercury intrusion porosimetry (MIP). The results from TG-MS showed thatmoisture and partial volatiles were removed from the coal matrix, and porestructure almost remained unchanged during the low heat treatment(25~200℃). The micropores and transition pores consisted of more than 80%of the total pore volume based on the MIP. The pore structure was slightlychanged following the temperature increase to 400℃, and the bound moistureand partial organics in the coal were released and decomposed by the increasedheat, respectively. When temperature reached 400℃, organic matterdecomposition of the coal released a large amount of hydrocarbon andmicromolecule gases. The meso- and macropore in the coal were significantlydeveloped, occupying ~35% of the total pore volume. Although there was nolarge change in generated gas composition after 600℃, the pore volume andstructures, including pore size distribution, pore volume and pore connectivity,were significantly changed based on the MIP. The pore structure acquired fromMIP exhibited a deviation when the mercury intruded pressure reached 10 MPa.A fractal model was introduced to correct the MIP data and acquire the porecompressibility of the coal matrix. The results showed that the porecompressibility decreased with increasing pressure and temperature. Thus, thisstudy provides significant implications of the pore structure evolution ofunderground coals that encounter heating.

Keywords: Coal matrix, Pore structure, Compressibility, Elevatedtemperatures, Mercury intrusion porosimetry

ENERGY EXPLORATION & EXPLOITATION · Volume 33 · Number 6 · 2015 pp. 809–826 809

1. INTRODUCTIONThe pore-fracture system is the main accumulation area and migration pathway forcoalbed methane (CBM) in coal reservoirs (Fan et al., 2010; Cai et al., 2014; Leeet al., 2014). The pore structure has significant influence on gas adsorption andseepage capacity in coal reservoirs and therefore affects the enrichment and recoveryof CBM (Clarkson and Bustin, 1999; Yao et al., 2008; Liu et al., 2009; Yu et al., 2012;Li et al., 2013; Cai et al., 2014; Liu et al., 2014). Coal, as a complex porous medium,has a complex pore structure, which includes pore size distribution, pore volume, porespecific surface area and pore surface roughness. Pore size distribution, attributed togas diffusion and flow, can be considered to be imperfectly connected, limiting theaccessibility of methane to some larger pores (Cai et al., 2013; Wei et al., 2013). Manymethods, including scanning electron microscope (SEM), mercury intrusionporosimetry (MIP), low-temperature N2 adsorption/desorption and nuclear magneticresonance (NMR) techniques have been adopted to acquire the information of porestructure. Recently, the pore evolution of coal reservoirs during coalification hasaroused the extensive attention of researchers. A growing interest in the knowledge ofpore structure of coals with elevated temperatures has developed to simulate and gaina better understanding of the coalification process (Feng et al., 2013). Multiplephysical and thermochemical transformations occur during the coalification process(Cai et al., 2014). In the absence of oxygen, coal will be decomposed and generatedifferent gases and tar at elevated temperatures (Yu et al., 2012). Moreover, the porestructure of coal will also change with the gases or tar generation. The mass-change,devolatilization and kinetic characteristics of coals during pyrolysis have been studiedby previous research (Zoller et al., 1999; Sun et al., 2011; Van Krevelen et al., 2013).However, few studies focused on the pore structure of coals (such as porosity, poresurface area, pore volume and pore size distribution) during the heating process(Puente et al., 1998; Cai et al., 2014). Changes of pore structure at elevatedtemperatures may have significant implications to the coalification process, coalreservoir improvement and enhanced CBM recovery.

For the coals that undergo orogenic metamorphism or are affected by magmaintrusion (or underground geothermal fluids), pore compressibility of the coal matrixwill be influenced by the changes of pore structure (Guo et al., 2014). Currently, MIPis still an important and widely used technique for analyzing and characterizing thepore structure of coals (Cai et al., 2013), whereas the pore compressibility has aneffect on MIP results, especially when pressure is greater than 10 MPa. With knownpore compressibility of measured coal, it is also possible to acquire accurate pore sizedistribution in the area where coal compressibility has its effect (Spitzer, 1981). Thepore compressibility could affect the evaluation of CBM adsorption and flow, whichbecomes important for the production of CBM (Zheng, 1993). Detailed study of thepore compressibility of coal matrix by MIP was conducted in previous works thatrevealed that the compressibility values ranged from 0.7 × 10-10 m2/N to 2.3 × 10-10 m2/Nwith varied carbon content (Van Krevelen et al., 1961; Toda and Toyoda, 1972).These values were in agreement with the latest results that show the compressibility tobe 1.55–2.94 × 10-10 m2/N with neglecting of mercury compressibility (0.4 × 10-10 m2/N)(Cai et al., 2013; Li et al., 2013).

810 Pore structure and compressibility of coal matrix with elevated temperatures bymercury intrusion porosimetry

To study the thermochemical reactivity of coal with rising temperature,thermogravimetry coupled with mass spectrometry (TG-MS) were adopted to analyzethe weight loss and generated gases of the low rank coal at elevated temperatures(25–1200℃). Then, the pore structures were investigated at increased temperatures(25℃, 200℃, 400℃ and 600℃) using SEM, low temperature N2

adsorption/desorption (77 K) and MIP. Finally the pore compressibility of the coalmatrix at a high-pressure range (10–200 MPa) was calculated according to the resultsof MIP measurements.

2. SAMPLES AND EXPERIMENTS2.1. Samples collection and coal analysesA low-rank coal sample (L1) was collected from the Wangtian coal mine in the NEOrdos basin, in northern China. The maximum vitrinite reflectance (Ro, m), maceralcomposition and proximate analysis were conducted using the same methods asprevious procedures (Liu et al., 2009). Table 1 shows the results of the Ro, m, maceralcomposition and proximate analysis of the coal sample.

2.2. Experimental proceduresMass loss and generated gases with elevated temperatures were measured using thesame TG-MS system as our previous research (Cai et al., 2014). The coal samplewas placed in a ceramic crucible and heated from 25 to 1200℃ at a heating rate of10℃/min using nitrogen as carrier gas at a constant flow rate of 60 cm3/min. The MSwas scanned over a range of 0 to 100 amu with measurement intervals ofapproximately 19 s. The relationship between the generated gas and the elevatedtemperatures was shown in the mass spectrogram.

Three sets of experiments were conducted, including SEM, low-temperature N2

adsorption/desorption and MIP to characterize the pore morphology and structure ofthe coal sample. The set-I experiment is the SEM analysis. The pore morphology ofcoal sample at different temperatures (25℃, 200℃, 400℃ and 600℃) wereobtained using the SEM. The set-II experiment is the low-temperature N2

adsorption/desorption, which was performed on the Micromeritics ASAP-2000 andfollowed the Chinese Oil and Gas Industry Standard Method SY/T 6154-1995 (Caiet al., 2013). The set-III experiment is the MIP analysis. The block sample wasselected for MIP analysis following the SY/T 5346-2005 standard process andconducted using PoreMasterGT60 (Quantachrome, US). The measurements run up to

ENERGY EXPLORATION & EXPLOITATION · Volume 33 · Number 6 · 2015 811

Table 1. Vitrinite reflectance, maceral composition and proximate analysis of the low rank coal.

Proximate analysis (%) Coal maceral composition (%)Districts Samples No. Ro, m (%) Mad Vdaf Ad V I E MNorth-eastern L1 0.69 2.95 22.44 13.87 68.15 23.70 5.63 2.52Ordos Basin

Mad, moisture; Vdaf, volatile; Ad, ash content; V, vitrinite; I, inertinite; E, exinite; M, mineral.

a pressure of 206 MPa, at which pore throats as small as 4 nm can be penetrated.Mercury intrusion-extrusion curves were obtained and cumulative mercury injectionvolume, pore radius and pore size distribution could be inferred from the curves.

3. RESULTS AND DISCUSSION3.1. Thermal dynamics with elevated temperatures3.1.1. Thermogravimetric analysisFigure 1 presents the thermal gravity (TG) and differential thermal gravity (DTG)curves of the coal at a temperature range of 25 to 1200℃, under the flow ofnitrogen. There are three weight loss peaks at a temperature range of 25 to 1200℃.As shown in the DTG curve, the temperature of the first weight loss peak was~150℃. The main reaction of the low heating process (lower than 200℃) is theremoval of moistures and gases that adsorbed at the surface of the coal matrix (Yanet al., 2003). However, the weight change is not significant in this temperature stage.With the temperature increased, the second weight loss peak is distributed in therange of 300 to 600℃. The maximum weight loss rate reaches 0.19 mg/min whentemperature is ~450℃. The main reaction happens during depolymerization anddecomposition of macromolecular organic matter in coal (Anthony and Howard,1976). Meanwhile, much gases and coal tar are generated in this temperature range.The last weight loss peak appeared at ~710℃, suggesting that the minerals in coal(mainly carbonate minerals) are decomposed under high temperature atmospheres(Huang et al., 1995; Yan et al., 2003).

3.1.2. Gas generation with mass spectrometryThe weight loss of coal is related to organic decomposition and gas production duringpyrolysis. Major gases generated during coal pyrolysis are shown in Figures 2 and 3,including H2O, H2, CH4, CO and CO2. Extensive evolution of gases happened in the


Figure 1. TG and DTG curves for the low rank coal (0.69% Ro, m).


Figure 2. Evolution curves of gaseous products of the coal sample with elevatedtemperatures by MS.

Figure 3. Variable gases composition and cumulative gases from the low rank coalpyrolysis.

temperature range of 300 to 800℃. From Figures 2 and 3, the evolution of CH4 startsat ~300℃, reaching a peak at 400–450℃. The methane generation mainly from CH3

-radicals of hydro aromatics occurs below 600℃ (CH3

- + H+ → CH4) and cleavage ofaromatic heterocyclic structures above 650℃ (Das, 2001; Porada, 2004; Zhao et al.,2011). H2 is originated from the polycondensation of radicals when temperature ishigher than 500℃. The evolution curves of CO and CO2 are similar and appear as aplurality of peaks. During the low temperature stage, CO2 and CO are produced by thedecomposition of the –COOH group and the break of the oxygen-containing –OHgroup. Moreover, CO2 and CO at higher temperatures (~700℃) come from thedecomposition of carbonate minerals (Li et al., 2003; Yu et al., 2008). H2O is relatedto the decomposition of oxygen-containing groups. Previous research found that thethermal stability of oxygen-containing groups in coal was in the order: hydroxyl >carbonyl > carboxyl > methoxyl (Cui et al., 2007). Therefore, the H2O was produced


Figure 4. SEM images showing the pore morphology of the coal sample at elevatedtemperatures.

during pyrolysis and the most obvious peak existed at ~400℃. The gas generation isalso accompanied with the changes the pore structure, which will be discussed insections 3.2.2 and 3.2.3.

3.2. Pore structure with elevated temperaturesThe pore classification by Hodot (1966), micropores (<10 nm), transition pores(10–102 nm), mesopores (102–103 nm) and macropores (>103 nm), was adopted tostudy the pore structure of the coal. SEM measurements were used to acquire theevolution of pore morphology from the perspective of visualization. The lowtemperature N2 adsorption/desorption analysis was used to investigate the micropores(<10 nm) and transition pores (10–102 nm), and the MIP method was used only tofocus on the mesopores (102–103 nm) and macropores (>103 nm).

3.2.1. Pore morphology analysis from SEMFigure 4 shows the evolution of pore morphology of the coal sample at elevatedtemperatures. It can be seen that the pore morphology remained stable during the lowheating process with temperature lower than 200℃ (Figs. 4a–4d). However, the poresurface and volume were enlarged by the removal of volatiles and moistures. Whenthe temperature reached 400℃, the surface morphology of the coal changed to besmooth and appeared in a squamose structure with the pore diameter remarkablyincreased (Figs. 4e–4f), which is confirmed by the MIP. Compared with the sampletreated at 400℃, the surface morphology of the sample treated at 600℃ lookedspongy and contained more gas burst pores (Figs. 4g–4h). Thus the heat treatmentcould improve the pore structure of coal reservoir.

3.2.2. Micro- and transition pores by low temperature N2 adsorption/desorptionThe variation of pore structure with elevated temperatures was shown in Table 2,which is acquired by low-temperature N2 adsorption/desorption (77 K). Results fromlow-temperature N2 adsorption/desorption indicate that the BET pore surface area oflow-rank coal with elevated temperatures ranges from 0.066 to 0.951 m2/g and thetotal pore volume (by BJH model) is in the range of 0.723 × 10−3~3.583 × 10−3 cm3/g.Figure 5 shows the N2 adsorption/desorption isotherms and pore characteristics withelevated temperatures. To gain a better understanding of the gas adsorption/desorption, the previous classification was adopted (Yao et al., 2008a), in which theadsorption/desorption isotherms were divided into four typical types (I, II, III and IV)and an abnormal type (V).

SBET, pore surface area by BET method; VBJH, total pore volume by BJH method;Vmic, volume percentage of micropores <10 nm; Vtran, volume percentage of transitionpores of 10–100 nm; Vmeso, volume percentage of mesopores of 100–300 nm; Tp,average width of pore size; Types, the types of nitrogen adsorption curve.

Figures 5a and 5b show that the isotherms belong to type I. There is a largehysteresis loop when P/P0 is in the range of 0.45 to 0.9. The desorption curveobviously decreases at P/P0 = 0.5. In terms of N2 adsorption/desorption isotherms, themain type for the pore structure is the ink-bottle shape (narrow throat and wide body).The pore surface area and volume of the sample heated at 25℃ were 0.7734 m2/g and3.583 × 10−3 cm3/g, respectively, and for the sample heated at 200℃ were 0.3008 m2/gand 1.701 × 10−3 cm3/g, respectively (Table 2). The average pore diameter of thesample heated at 200℃ is higher than that of the sample heated at 25℃. This indicatesthat the pore morphology remains unchanged with the removal of moistures andpartial volatiles, but the pore diameters can be enlarged by low temperature heattreatment (lower than 200℃).

Figure 5c shows that the N2 adsorption/desorption isotherm is similar to type IV.However, there is an obvious hysteresis loop, and the sharp increase and decreasecurves exist approximately vertical at P/P0 = 0.9–1.0. This indicates that the poreshave gradually changed into narrow slit-shaped or wedge-shaped morphology. Thepore surface area and volume of the sample heated at 400℃ are 0.0660 m2/g and 0.723× 10−3 cm3/g, respectively. And the average pore diameter is extended to 48.86 nm.Previous research (Wang et al., 2010) showed that the coals would be softening, andthe thermal shrinkage would take place after temperatures reached at 400℃ due toorganic decomposition and heat-solid reaction. During this stage, the proportion of


Table 2. Pore surface area, pore volume and pore structure of the low rank coalfrom low temperature N2 adsorption/desorption.

Pores proportions (vol.%)Sample No. SBET (m2/g) VBJH (×10–3 cm3/g) Vmic Vtran Vmeso Tp (nm) TypesL1(25℃) 0.7734 3.583 9.3 63.2 27.5 27.49 IL1(200℃) 0.3008 1.701 5.5 60.6 33.9 35.09 IL1(400℃) 0.0660 0.723 1.3 57.5 41.2 48.86 IVL1(600℃) 0.9508 0.783 0.5 29.9 69.6 84.36 III


Figure 5. Low temperature N2 adsorption/desorption (77 K) and pore characteristicsof the coal sample with elevated temperatures.

micro- and transition pores decreased significantly, which resulted in a vast reductionof pore surface area and volume. However, the opening of the pores will be increasedwith organic matter decomposition.

The N2 adsorption/desorption isotherm of the sample treated at 600℃ belongs totype III (Fig. 5d). The adsorption and desorption curves coincided, and no hysteresisloop appeared, which suggests that the pores are dominated by one end closed pores,with the shape of parallel-plates. As shown in Table 2, the pore surface area and porevolume increased to 0.9508 m2/g and 0.783 × 10−3 cm3/g, respectively. The massivemicro- and transition pores were significantly generated when the temperature reached600℃, which directly resulted in higher pore surface area and volume and provided alarge amount of adsorption sites for gases.

3.2.3. Meso- and macropores by MIPAlthough the MIP method is commonly used to characterize the pore size distributionof coals from a few nanometers to tens of micrometers, the pore compressibility ofcoals is inevitable at high pressures (normally higher than 10 MPa) (Patrick et al.,2004). The pore compressibility at high pressures (higher than 10 MPa, correspondingto 147 nm in diameter) will be detailed in section 3.3. Therefore, the MIP methodcould be used to precisely characterize pores with diameters higher than 100 nm. WithMIP data, the pore volume of meso- and macropores ranges from 0.005 to 0.1696 cm3/g(Table 3). The mercury intrusion/extrusion curves and pore size distribution withelevated temperatures were also acquired (Fig. 6).

Figures 6a and 6b show that the intrusion/extrusion curves are relatively stable andthat mercury injection saturation reaches up to ~65%. The mercury injection saturationranges from 10% to 15% when pressure is lower than 15 MPa. This indicates that themeso- and macropores are not well developed, occupying ~20% of the total volume.The efficiency of mercury withdrawal closes to 80%, suggesting that the pores arewell connected. Moreover, the pore volumes and average pore diameters of the sampletreated at 200℃ are slightly higher than that for the sample treated at 25℃ (Table 3).Thus, the low heat treatment can slightly improve the structures of meso- andmacropores.

Comparing the sample treated at 25℃ with the sample heated at 200℃, the intrusioncurve of the sample treated at 400℃ can be divided into two stages. At ~1 MPa of the


Table 3. Pore volume and pore structure analyzedby MIP.

Sample No. Vpore (×10–3 cm3/g) Sin (%) Eex (%) Tp (μm)L1(25℃) 5.0 67.0 79.5 0.194L1(200℃) 5.2 64.0 79.2 0.226L1(400℃) 11.6 62.3 62.5 0.390L1(600℃) 169.6 77.5 6.5 1.277

Vpore, pore volume of mesopores and macropores; Sin, maximum mercury injectionsaturation; Eex, extrusion efficiency; Tp, average width of pore throat.

mercury intrusion pressure, the mercury injection saturation and the efficiency ofmercury withdrawal are higher than 60% (Fig. 6c). The meso- and macropores weresignificantly developed, occupying ~35% of the total volume, suggesting that hightemperature treatment can effectively improve pore size distribution. However, pore


Figure 6. Mercury intrusion/extrusion and pore size distribution for the coal sampleat elevated temperatures.

connectivity is reduced. This may contribute to the phenomenon of softening andthermal shrinkage of coal. The changes of pore size distribution and connectivity mayhave a significant effect on the gas flow in the coal reservoir.

Figure 6d shows that the mercury injection saturation of the sample heated at 600℃is closed to 80%, but the mercury intrusion saturation reaches up to 70% whenpressure is ~15 MPa. This phenomenon indicates that the meso- and macropores arepredominant, occupying ~90% of total volume. With the gases separated from coalduring the pyrolysis, the original pore diameter was enlarged. Moreover, a largenumber of new pores are generated. However, the efficiency of mercury withdrawal isonly 6.5%, suggesting that the pore connectivity is very poor. This finding may berelated to coal tar blocking the pore throat, which may condense and block the poresat the high temperatures and reduce the pore connectivity. This phenomenon will notbe conducive to gas diffusion and seepage. Therefore, the accompanying tarproduction cannot be ignored when using a heating method to improve the reservoirproperties.

3.3. Pore compressibility of coal matrix by MIPPrevious research (Friesen and Mikula, 1988; Xu et al., 1999) found thatcompression may directly increase pore volume when the intrusion mercurypressure exceeds 10 MPa. Therefore, pore compressibility should be calculated tocorrect the experimental data when the pressure exceeds 10 MPa. The porecompressibility of coal (kp, m2/N) is defined as follows:

(1)

where Vp is the coal pore volume and dVp/dP represents the coal pore volume changeas a function of the pressure. However, kp is different from the effective porecompressibility because P in Eq. (1) is not exactly the same as the pore pressure andthe confining pressure cannot be kept as a constant because of the effect of porepressure variations of the void space volume contained in coal (Li et al., 1999).Previous research (Qu et al., 2010) found that the Eq. (1) can be used to investigatepore compressibility of coal with some corrections to the relationship between Vp andP from MIP measurements.

Based on the Washburn equation (Washburn, 1921), the relationship between poreradius (μm) and mercury intrusion pressure can be defined as:

(2)

where σ (here set to 0.48 J/m2) is the surface tension of mercury and θ is the mercury-solid contact angle (assumed to be 140°). Thus, Eq. (1) can be transformed into:

(3)

It has been demonstrated that pore size distribution (dVp(P)/dr) is related to thesurface fractal dimension (D) by the fractal geometry law (Qu et al., 2010):

kV

dV

dP

1p

p

p=

P /r2 cosσ θ= −

P r0.735/=


(4)

Combined with Eq. (1), Eq. (3) can be written as:

(5)

where Vp(P) denotes the pore volume that can be approximated by the cumulativeintrusion volume (cm3/g), and D represents the fractal dimension. Thus, Figure 7shows the relationship between log (dVm(P)/dP) and log(P) with the MIP data. Thefractal dimension (D) of high-pressure range (P > 10 MPa) varies from 3.30 to 3.94.The fractal dimension is higher than 3, suggesting that the pore compression effectexists in the high-pressure range (Friesen and Ogunsola, 1995; Mahamud et al., 2003).

Previous research (Friesen and Mikula, 1988; Qu et al., 2010) found that due to theuncertainty of the relationship between pore volume Vp(P) and MIP measured volume

dV P

dPD Plog 4 logp ( ) ( )

⎛

⎝⎜⎜

⎞

⎠⎟⎟∝ −

dV

dPr

rDp 2( )−⎡

⎣⎢⎢

⎤

⎦⎥⎥∝ −


Figure 7. Plots of log(dVm(P)/dP) versus log(P) for fractal dimensions analysis.

Vm(P), as well as the occurrence of compression effects, Vp(P) can be replaced byVm(P) in most instances. Thus Eq. (5) can be written as:

(6)

where a and b are constant and can be acquired by fitting the MIP data. As mentionedin Eq. (6), a linear relationship existed between MIP measured volume Vm(P) and PD–3.Therefore, the plots of Vm(P) versus PD-3 and the constants (a and b) can be obtainedas shown in Figure 8.

Based on Eqs. (1) and (6), the pore compressibility of coal can be deduced as:

(7)

The pore compressibility was calculated by Eq. (7) and the pore compressibilityranges from 0.25 – 37.79 × 10-9 m2/N correspond to the pressure range of 10 to 200 MPa

V P a bPDm

3( ) = + −

b D P

a bPk

( 3)p

D

D

4

3= −

+

−

−


Figure 8. Relation between Vm(P) and PD–3 of the coal sample at elevatedtemperatures.

with a temperature range of 25 to 600℃. This shows that pore compressibility of coalsdecreases with increasing pressure and temperature (Figs. 9a–9b). Compared to resultsfrom previous research (Van Krevelen et al., 1961; Cai et al., 2013), the porecompressibility of coal at 25℃ in present study is larger than one order ofmagnitude, which should be related to the base used: The pore volume was used inthis study, whereas the skeleton or sample volume was used in previous publications(Van Krevelen et al., 1961; Toda and Toyoda, 1972; Cai et al., 2013). For samplestreated at 25℃ and 200℃, the pore compressibility does not vastly change when thepressure is higher than 30 MPa, but the pore compressibility of the sample treated at200℃ is slightly lower than that of the sample treated at 25℃. This situation indicatesthat the removal of moisture and partial volatiles have less impact on porecompression of the coal matrix. Due to organic matter decomposition and changes ofthe pore size distribution, the pore compressibility of the sample treated at 400℃significantly decreased. Multiple meso- and macropores were generated, while nomicropores were created in this stage (Figs. 5 and 6), which decreased the stability ofthe pore structure and lowered the pore compressibility in the coal matrix. Whentemperature reached up to 600℃, gases and tar were produced and the pore structureof the coal matrix was significantly altered, which may resulted in extremely low porecompressibility (0.25–2.22 × 10-9 m2/N). As pore evolution occurred, the full scalepores decreased the pore compressibility. Therefore, the change of organic matterdecomposition and pore structure of coal may have an important impact on porecompressibility. However, the process of this change is very complex with elevatedtemperatures. Mechanisms of tar blockage of micro- or transition pores are stillunclear; thus, more work is required to gain a better understanding of thisphenomenon.

4. CONCLUSIONSThe pore structure and compressibility of low rank coal matrix with elevatedtemperatures were investigated by multiple methods, including TG-MS, SEM, N2

adsorption/desorption at 77 K and MIP. The following conclusions can be made:


Figure 9. Relationships between pore compressibility and pressure (a), temperature(b) for coal samples.

1) The removal of moisture and gases occurred in the low temperature stage(< 200℃). Organic matter decomposition generated a large amount of gas inthe temperature range of 400 to 600℃, leading to significant weight loss.

2) During low heat treatment, the main type of pore structure was the ink-bottleshape, and the micro- and transition pores developed. When the temperaturereached 400℃, the meso- and macropores developed, occupying ~35% of thetotal volume. Pore diameter was enlarged, and massive pores were generatedwhen temperature increased to 600℃. The meso- and macropores werepredominant, occupying ~90% of total volume. The pores were poorlyconnected due to generated tar blockage of the pore throat.

3) Based on the MIP data, the pore compressibility of coals with elevatedtemperatures was calculated in the high-pressure range (P > 10 MPa). Thepore compressibility decreased with increasing pressure and temperature.Organic matter decomposition and the change of pore size distribution alsohave important effects on the pore compressibility of the coal matrix.

ACKNOWLEDGMENTSThis research was funded by the National Major Research Program for Science andTechnology of China (grant nos. 2011ZX05034-001 and 2011ZX05062-006), theNational Natural Science Foundation of China (grant no. U1262104), the Program forNew Century Excellent Talents in University (grant no. NCET-11-0721), theFoundation for the Author of National Excellent Doctoral Dissertation of PR China(grant no. 201253) and the Fundamental Research Funds for Central Universities(grant no. 2652013006).

REFERENCESAnthony D.B. and Howard J.B., 1976. Coal devolatilization and hydrogasification.

AIChE Journal 22(4), 625–656.

Clarkson C.R. and Bustin R.M., 1999. The effect of pore structure and gas pressureupon the transport properties of coal: a laboratory and modeling study. 1.Isotherms and pore volume distributions. Fuel 78(11), 1333–1344.

Clarkson C.R. and Bustin R.M., 1999. The effect of pore structure and gas pressureupon the transport properties of coal: a laboratory and modeling study. 2.Adsorption rate modeling. Fuel 78(11), 1345–1362.

Cai Y.D., Liu D.M., Pan Z.J., Yao Y.B., Li J.Q. and Qiu Y.K., 2013. Pore structureand its impact on CH4 adsorption capacity and flow capability of bituminousand subbituminous coals from Northeast China. Fuel 103, 258–268.

Cai Y.D., Pan Z.J., Liu D.M., Zheng G.Q., Tang S.H., Connell L.D., Yao Y.B. andZhou Y.F., 2014. Effects of pressure and temperature on gas diffusion and flowfor primary and enhanced coalbed methane recovery. Energy Exploration &Exploitation 32(4), 601–619.

Cai Y.D., Liu D.M., Yao Y.B., Li J.Q. and Zheng G.Q., 2014. The evolution ofpetrophysical properties of coalbed methane reservoirs under the heattreatment. Earth Science Frontiers 21(1), 240–248.


Cai Y.D., Liu D.M., Yao Y.B., Li Z.T. and Pan Z.J., 2014. Partial coal pyrolysis andits implication to enhance coalbed methane recovery, Part I: An experimentalinvestigation. Fuel 132, 12–19.

Cui Y.P., Qin L.L., Du J. and Chang L.P., 2007. Products distribution and itsinfluencing factors for coal pyrolysis. Coal Chemical Industry 35(4), 10–15 (inChinese with an English abstract).

Das T.K., 2001. Evolution characteristics of gases during pyrolysis of maceralconcentrates of Russian coking coals. Fuel 80(4), 489–500.

Fan J.J., Ju Y.W. and Hou Q.L., 2010. Pore structure characteristics of differentmetamorphic-deformed coal reservoirs and its restriction on recovery ofcoalbed methane. Earth Science Frontiers 17(5), 325–335.

Feng Y.Y., Jiang C.F., Liu D.J. and Chu W., 2013. Experimental investigations onmicrostructure and adsorption property of heat-treated coal chars. Journal ofAnalytical and Applied Pyrolysis 104, 559–566.

Friesen W.I. and Mikula R.J., 1988. Mercury porosimetry of coals: pore volumedistribution and compressibility. Fuel 67(11), 1516–1520.

Friesen W.I. and Ogunsola O.I., 1995. Mercury porosimetry of upgraded westernCanadian coals. Fuel 74(4), 604–609.

Guo X.Q., Yao Y.B. and Liu D.M., 2014. Characteristics of Coal MatrixCompressibility: An Investigation by Mercury Intrusion Porosimetry. Energy& Fuels 28(6), 3673–3678.

Huang H., Wang K., Klein M.T. and Calkins W.H., 1995. Determination of coal rankby thermogravimetric analysis. Preprints of papers-American ChemicalSociety Division of Fuel Chemistry 40(3), 465–465.

Hodot B.B., 1966. Outburst of coal and coalbed gas (Chinese Translation). China CoalIndustry Press, Beijing, pp. 318.

Lee W., Jang H. and Lee J., 2014. Development and application of the artificial neuralnetwork based technical screening guide system to select production methods in acoalbed methane reservoir. Energy Exploration and Exploitation 32(5), 791–804.

Li Y.H., Gao Q.L. and Rudolph V., 1999. Compressibility and fractal dimension offine coal particles in relation to pore structure characterisation using mercuryporosimetry. Particle & Particle Systems Characterization 16(1), 25–31.

Li J.Q., Liu D.M., Yao Y.B., Cai Y.D. and Guo X.Q., 2013. Physical characterizationof the pore-fracture system in coals, Northeastern China. Energy Exploration& Exploitation 31(2), 267–286.

Li S., Tang D.Z., Pan Z.J., Xu H. and Huang W.Q., 2013. Characterization of the stresssensitivity of pores for different rank coals by nuclear magnetic resonance.Fuel 111, 746–754.

Li X., Matuschek G., Herrera M., Wang H. and Kettrup A., 2003. Investigation ofpyrolysis of Chinese coals using thermal analysis/mass spectrometry. Journalof Thermal Analysis and Calorimetry 71(2), 601–612.


Liu D.M., Yao Y.B., Tang D.Z., Tang S.H., Che Y. and Huang W.H., 2009. Coalreservoir characteristics and coalbed methane resource assessment in Huainanand Huaibei coal fields, Southern North China. International Journal of CoalGeology 79(3), 97–112.

Liu S.Q., Sang S.X., Zhu Q.P. and Liu H.H., 2014. Response characteristics andmechanisms of dynamic fluid field for well interference of coal bed methanegroup wells in production block. Energy Exploration and Exploitation 32(5),771–790.

Mahamud M., López Ó., Pis J.J. and Pajares J.A., 2003. Textural characterization ofcoals using fractal analysis. Fuel Processing Technology 81(2), 127–142.

Patrick A.C.G., Cathy J.R., Esa L., Rustem V., Istvan F., Joachim S., Hannu P. andJohn D., 2004. Comparison of NMR cryoporometry, mercury intrusionporosimetry, and DSC thermoporosimetry in characterizing pore sizedistributions of compressed finely ground calcium carbonate structures.Industrial & Engineering Chemistry Research 43(24), 7920–7927.

Porada S., 2004. The reactions of formation of selected gas products during coalpyrolysis. Fuel 83(9), 1191–1196.

Puente G., Iglesias M.J., Fuente E. and Pis J.J., 1998. Changes in the structure of coalsof different rank due to oxidation-effects on pyrolysis behaviour. Journal ofAnalytical and Applied Pyrolysis 47(1), 33–42.

Qu Z.H., Wang G.G.X., Jiang B., Rudolph V., Dou X.Z. and Li M., 2010.Experimental study on the porous structure and compressibility of tectonizedcoals. Energy & Fuels 24(5), 2964–2973.

Spitzer Z., 1981. Mercury porosimetry and its application to the analysis of coal porestructure. Powder Technology 29(1), 177–186.

Sun M., Ma X., Yao Q., Wang R., Ma Y., Feng G., Shang J., Xu L. and Yang Y., 2011.GC-MS and TG-FTIR study of petroleum ether extract and residue from lowtemperature coal tar. Energy & Fuels 25(3), 1140–1145.

Toda Y. and Toyoda S., 1972. Application of mercury porosimetry to coal. Fuel 51(3),199–201.

Van Krevelen D.W., 1961. Coal-typology, chemistry, physics, constitution. ElsevierScience & Technology, Elsevier Publishing Company, Amsterdam.

Van Krevelen D.W., Van Heerden C. and Huntjens F.J., 1951. Physicochemicalaspects of the pyrolysis of coal and related organic compounds. Fuel 30(11),253–259.

Wang Y., Zhao Y.S. and Feng Z.C., 2010. Study of evolution characteristics ofpore structure during flame coal pyrolysis. Chinese Journal of RockMechanics and Engineering 29(9), 1859–1866 (in Chinese with an Englishabstract).

Washburn E.W., 1921. The dynamics of capillary flow. Physical Review 17(3),273–283.


Wei Y.C., Cao D.Y., Yuan Y., Zhu X.S., Zhang X.Y., Zheng Yao and Ji Zhou, 2013.Characteristics of pulverized coal during coalbed methane drainage inHancheng block, Shaanxi Province, China. Energy Exploration & Exploitation31(5), 745–758.

Xu L.J., Liu C.L., Xian X.F. and Zhang D.J., 1999. Compressibility of coal matter andcoal pore. Colloid and Surface A: Physicochemical and Engineering Aspects157(1), 219–222.

Yan J.D., Cui H., Yang J.L. and Liu Z.Y., 2003. Research on pyrolysis behavior ofYanzhou coal using TG/MS. International Journal of Mining Science andTechnology 32(3), 311–315 (in Chinese with an English abstract).

Yao Y.B., Liu D.M., Tang D.Z., Huang W.H., Tang S.H. and Che Y., 2008a. Acomprehensive model for evaluating coalbed methane reservoirs in China. ActaGeologica Sinica (English Edition) 82(6), 1253–1270.

Yao Y.B., Liu D.M., Tang D.Z., Tang S.H. and Huang W.H., 2008. Fractalcharacterization of adsorption-pores of coals from North China: Aninvestigation on CH4 adsorption capacity of coals. International Journal ofCoal Geology 73(1), 27–42.

Yu Y., Liu X.J. and Xu J.J., 2008. TG-MS study on pyrolytic reaction of Huainan coal.Coal Quality Technology 5, 9–15 (in Chinese with an English abstract).

Yu Y.M., Liang W.G., Hu Y.Q. and Meng Q.R., 2012. Study of micro-poresdevelopment in lean coal with temperature. International Journal of RockMechanics and Mining Sciences 51, 91–96.

Zhao Y.P., Hu H.Q., Jin L.J., He X.F. and Wu B., 2011. Pyrolysis behavior of vitriniteand inertinite from Chinese Pingshuo coal by TG-MS and in a fixed bedreactor. Fuel Processing Technology 92(4), 780–786.

Zheng Z., 1993. Compressibility of porous rocks under different stress conditions.International Journal of Rock Mechanics and Mining Sciences 30(7),1181–1184.

Zoller D.L., Johnston M.V., Tomic, J., Wang, X. and Calkins, W.H., 1999.Thermogravimetry-photoionization mass spectrometry of different rank coals.Energy & Fuels 13(5), 1097–1104.


Documents

Pore structure and compressibility of coal matrix with ... · structures, including pore size distribution, pore volume and pore connectivity, were significantly changed based on