Analysis of pore system model and physical property of coal reservoir in the Qinshui Basin

ARTICLESChinese Science Bulletin 2005 Vol 50 Supp 52-58cropores and transition pores are relatively few:(iii) theAnalysis of pore system model ble pore system model preferably summarizes the poreand physical property of coalsystem characteristics of middle and low rank coal, and itbecomes the foundation for the numerical simulation ofreservoir in the qinshui Basin coalbed methane reservoirs at present. However, dtterentcountries have different geologic conditions, so theories,XU Hao, ZHANG Shanghu, LENG Xueexperiences and techniques from the United States are notTANG Dazhen& WANG Mingshouapplicable to other countries completelyIn the Qinshui Basin of Shanxi province, China, the2. Shanghai Offshore Oil Gas Branch, Sinopec, Shanghai 20012, pore system of high rank coal reservoir is extraordinarily1. China University of Geosciences, Beijing 100083, China;intricate, so it is difficult to be generalized by a uniformChinaCorrespondence should be addressed to Xu Hao (email: XImodel. Consequently, this study did not aim to provide ahao600@163.com)uniform geometric model for the high rank coal pore sys-tem, but aimed to analyze the universal characteristics andAbstract The Qinshui Basin in China is a major area for the diversities of pore system in the high rank coal reserexploration and development of high rank coalbed methane. voirs from the point of view of structures and physicalDue to the high rank coal and complicated pore system, no properties. By taking into account the horizontal and versubstantial breakthrough in the exploration and develop- tical diversities of coal rank and coal facies, samples havement of coalbed methane has been made until now. Many been taken systemically in the Qinshui Basin. The purposesome features as follows: the porosity is relatively low; the of this study was ( to generalize the characteristics ofpore system is dominated by micropores and transition pores; pore system though multi-testing of samples, (i) to clas-mesopores take the second place, and macropores are nearly sify samples using cluster analysis method, (iii)to selectabsent, which is exceedingly adverse for production of coal- the typical samples to establish the corresponding porebed methane. However, testing data also revealed the differ- models, (iv) to analyze the diversities of reservoir physicalential development for the pore of high rank coal reservoirs properties, (v)to contribute to provide a direction for thein the Qinshui Basin, which necessarily led to the different exploration and development of coalbed methane in thephysical properties of desorption, diffusion and permeabilit y. Qinshui Basinmethod and selects the typical samples to establish four pore 1 Sampling and methodystem models, analyzes the differences of reservoir physicalproperty, and provides a guidance for the exploration and samplingdevelopment of coalbed methane in the Qinshui Basin.In order to reflect the universal properties and diyKeywords: Qinshui Basin, coal reservoirs, differential development, ties of pore system in high rank coal reservoirs, approxipore system model, reservoir physical property.mately 40 samples were obtained from 13 mining area ofDOI:10.1360/98zk00046 towns, including Yang Quan-Shouyang, QinyuanChangzhi, and Jincheng-Gaoping in the north, middle andThe pore system ofcoal reservoirs and the migration south of the Basin ( Table 1), in which coalbed weremechanism of monophasic and polyphasic fluid are con- elaborately described and samples were systematicallysiderably complicated. It is not very easy to establish a collected in the No I mine of Yangquan, the Qinxin mineconcise and comprehensive model. The double pore sys- of Qinyuan, the Jingfang mine of Changzhi and the Sihetem model of coal reservoirs was built up first in the mine of JinchengUnited States by corroboration of exploration and produc1. 2 Methodtion. This model was an approximative generalization ofreal pore system, proposed by Warrenh et al., the coalbed Through sample collection, vitrinite reflection, poremethane geologists of the United States, who considered structure and cryogenic nitrogen isothermal adsorptionthat porosity system in coal reservoirs consisted of matrix characteristics were analyzed on 39 samples. A micro-porosity and natural cleat system. Accordingly, they sum- scope photometer was used to test the vitrinite reflectionmarized the migration process of coalbed methane in res- of samples, mercury porosimeter was used to obtain theervoirs into desorption, diffusion and permeability. The data of porosity in sections of different diameters to ana-United States has advantaged geologic conditions for coal lyzemethane, and targets of exploration and development in an atL凵中国煤化工mthe United States are almost middle and low rank coal ticularC N MH Garacters of reservoirsreservoirs, in which the pore system has some features This paper mainly discusses the data of pore structure,as follows:(i) the porosity is relatively high; (ii) the pore while the data of cryogenic nitrogen adsorption and roaresystem is dominated by mesopores and macropores, minot listed completely in this paperChinese Science Bulletin Vol 50 Supp. December 2005ARTICLESTable I Mercury penetration parameters of samplesThroatrial Porosity diameteExpulsion Extrusion GF Uniformitynumber number (%)pressure efficiencyMMPa(%) coefficient0010.1-0.63063-25umum /umyqyk15-13.404.1054.360.360.6294.12.22.7yqyk15-210.5950480040.7474610.111.2rgyk15-32345674.1073440.1204991.14.73.3No, 1 of yoyk15-43.60064000.100.73Yangquan yqyk15-51.200.1079.520.33330.4594.747.510.050.5886.26.648.370060.7381.570yaykI5-911.5549.160.050.7781.666280.14049914482yqxj3-268010.2504893445460.09082883644.yaxj3-468010.2589.14.5yaxj3-552040.010.7275410.010yky3-153.770.100.60ayuan syky9-10123456784.2865720.06Shidianyqsd15-11.100039273.820.060.5188075790080598926.934hm03028.5460030070.676779.7Zuoquanhmzk0102 192.508.9l62380.010.797909.5gyqx2-l4.1075.530.040.574.8qyqx2-288.65.5gyqx2-31.305.3179050.1105589.658234562.80cucc3-15.1056.160050.798259.27.1MiddlecZcc3-3Changcun49.310.150.778994.7c7cc3-449.560.080.5786.5542810.5965010160.7czcc3-6400786.57.17.1856990.100.7886963.730.10wangzhuang czwz3-2 324.1049360.090.9952.3711.126l,222,1l1.5csh3-21.101877010.100.7089.6581.500.0563.570.110.778846.075058.120.110.798785.149Xiaoxian gpxx3-I5.546l.560.07829Guohexhgh3-1 3868.380080.7784.30.5966953.32 Analysis of pore system model and reservoirs The pore content with the radius ranging from 0 to 0. Iumphysical propertyvaried from 61.2% for the smallest to 94.5% for theFrom the capillary pressure data of mercury injection, it ranging from 0. 1 to 0.63 Hm varied from 2.2% to 22.1%may be known that micropores and transition pores aredominant, and mesopores and macropores are lacking in ture in high rank coal reservoir, which consequentiallythe pore system ot high rank coal reservoirs in the Qinshui brings中国煤化工 physical propertyze the differences inrank coal reservoirs. On the other hand, the pore structure measuHCNMHGnodel to analyze theof high rank coal reservoirs is not homologous completely. physical proChinese Science Bulletin Vol 50 Supp. December 2005ARTICLES2. 1 Cluster analysis and the selection of typical samples and permeability of the gas accumulated in the pore sysThe cluster analysis is a multi-statistical analysis tem could be estimated qualitatively. The other is themethod for grouping the research objects into respective porosity, pore diameter mean, expulsion pressure, ejectionategories according to their similarities in property. Due efficiency, grading factor and uniformity coefficient andto the different objects, the cluster analysis could be di- so on, which reflect the permeability of pore system. Thisvided into Q-model and R-model cluster analysis. The study chose two kinds of parameters for the Q-modelformer is used to categorize samples, and the latter is used cluster analysis. As shown in Fig. 1, 39 samples are di-to categorize the variables of samples. In this study the vided into four groups, which preferably indicate theQ-model cluster analysis and the sum of squares of devia- properties of samples from the aspects of adsorption andtions method (i.e. Ward's methoddopted. TheWards method is distinct from all other methods because permeability The property mean of every cluster of sam-ples is thought to be the concentrative reflection of itsit uses an analysis of variance approach to evaluate thedistances between clusters. In short, this method mini. characteristics and is thought to be the model or approxi-mizes the sum of squares(SS)of any two(hypotheticamate model of every cluster of samples (Table 2),solusters that can be formed at each step. In general, this could be investigated as the common characteristics ofmethod is regarded as a very efficient method. Howeveramples. However, there must be some differences bebefore cluster analysis, the summation calibration method tween the ideal results and the actual properties of samplesis adopted to correct raw measured dataSo the samples close to the property mean of samplesIn this study, two kinds of data about pore structure should be chosen for the model examples. After compara-were obtained from the mercury injection capillary pres- tive analysis samples, jcsh3-1, yayk15-9, czjf3-1 andsure of coal samples. One is the data of pore content of qyax2-3 which represented I,Il, Ill and Iv clusters redifferent pore radii, by which the extent of desorption spectively were chosen for the model analysis (Table 3)2732614438243023293536283911262511620122139105172234318195372137833Fig 1. The pedigree chart of Q-model cluster analysisTable 2 Pore content mean of samples in cluster analysisClusterAmount of satPore content mean(%)EvaluationnIcropore-transition pores mesoporemacropore22.08best for desorption and permeabilityn and permeabilit中国煤化工125.31CNMHGand permeabilityChinese Science Bulletin Vol 50 Supp. December 2005ARTICLESTable 3 Pore content mean of typical samples in cluster analysClusterTypical sample numberPore content mean(%)micropore-transition poresmesoporemacropore61.1722.08l149IⅢⅣyqyk15-98.03868689642.2 Analysis of pore system model(iii) Model 3. The representative sample for Model 3() Model 1. The representative sample for Model 1s czjf3-l, collected from the No. 3 main coalbed in thejcsh3-1, collected from the No. 3 main coalbed in the Sihe Jingfang mine, Changzhi. Its vitrinite reflectance is 1.79%,mine, Jincheng. Its vitrinite reflectance was 3.43%0; its and its mercury injection curve and cryogenic nitrogenmercury injection curve and cryogenic nitrogen absorption. absorption isothermal curve are shown in Figs. 2 and 3(c)isothermal curve are showed in Figs. 2 and 3(a). Its poros- Its porosity is basically the same as that of Model 2, andtion is the highest in all samples, up to 45%. In this sam- than that of Model 2. In this samps ximately 7% lowerity is relatively high up to 7.60%o, and its intrusion satura- its intrusion saturation is 18%, approxation pores areple, the pore system is dominated by mesopores, transition the most, approximately 15% higher than that of Model 2,pores take the second place, the third is micropores, and and the contents of mesopore and macropores are about8% lower than that of Model 2. The hysteresis loop ofameter mean is about 0.5 um, one order larger than any mercury injection curve is small, the volume differenceother samples, suggesting that coarse throat pores are con- (pressure difference) between intrusion and extrusion issiderable. The hysteresis loop of mercury injection curve very little, indicating that closed pores predominate in theis large and the volume difference(pressure difference) pore system, and its connectivity is moderate. The cryo-between intrusion and extrusion is great, showing that the genic nitrogen adsorption isotherm curve has an absorp.open spaces predominate in pore system, and its connec- tion regression line and an inflection point, but it is lesstivity is very good. The cryogenic nitrogen adsorption obvious than Model 1 and Model 2, suggesting that theisotherm curve has an obvious absorption regression line pore system is a transition from open pores to closed poresand a inflection point, indicating that the pore system is Compared with Model 1 and Model 2, this pore structuredominated by open pores, including cylindrical pores open was not advantageous for desorption, diffusion and perat both ends and parallel plate pore open four sides. The meability of coalbed methane, and the reservoirs reprepore structure in this type of sample was most advanta- sented was medium or relatively bad for the explorationgeous for desorption, diffusion and permeability, and the and development of coalbed methanereservoirs represented was the favored reservoirs for theexploration and development of coalbed methane.(iv) Model 4. The representative sample for Model 4 isqyqx2-3, obtained from No. 2 main coalbed of Qinxin(i)Model 2. The representative sample for Model 2 is mine in Qinyuan. Its vitrinite reflectance is 1.66%.Theyqyk15-9, which is obtained from the No 15 main coalbed mercury injection curve and isothermal curve of cryogenicin the No I mine of Yang Quan. Its vitrinite reflectance is nitrogen absorption are showN in Figs. 2 and 3(d). Its po2.92%, and its mercury injection curve and cryogenic ni- rosity and intrusion saturation are only 17% and 1.3%trogen absorption isothermal curve are shown in Figs. 2 respectively. In this sample, transition pores are the most,and 3(b). Its porosity is lower than Model I and its intru- accounting for more than 70%, mesopores take the secondsion saturation is 25%, approximately 20% lower than place, micropores and macropores are nearly absent. TheModel 1, suggesting that the development of pores is very hysteresis loop of mercury injection curve is small, andbad compared to Model 1. On the aspect of pore structures, the volume difference (pressure difference) between intruthe content of transition pores is 20% higher than that of sion and extrusion is also very little, suggesting thatModel 1, but the content of mesopores is 20% lower than semi-closed pores whose connectivity is quite bad prethat of Model 1. Mercury injection curve and cryogenic dominate in pore system. The cryogenic nitrogen adsorp-nitrogen absorption isothermal curve are similar to that of tion isotherm curve has almost no absorption regressionModel 1, suggesting that the pore system is also domi- line, also indicating that pore system is mainly constitutednated by open pores. By comprehensive estimation, the by airtight pores, closed at end. The pore structure in thispore structure in this type of samples is most advanta- typegeous for desorption, diffusion and permeability of coal- tion,中国煤化工 he reservoirsbed methane, and the reservoirs represented is fit for sentedCN MH Ge exploration and deexploration and development of coalbed methanevelopment of coalbed methaneChinese Science Bulletin Vol 50 Supp. December 2005ARTICLES0.001Pressure(MPa)he intrusion curve of model IThe intrusion curve of model 2The intrusion curve of model 3The intrusion curve of mode 4-+ The extrusion curve of model 1Hit-The extrusion curve of model 2-A- The extrusion curve of model 3-+- The extrusion curve of model 4Fig. 2. The instrusion and extrusion curves of typical samples in the Qinshui Basin2.3 Discussion on reservoir physical property of the tion, accumulation and reservoir-forming of coalbedcorresponding pore modelmethane. Without the peculiar pore structures in high rankOne of the principal reasons that the high rank coal areacoal reservoirs. the later intense tectonic movementsin the Qin Shui Basin of Shanxi province became the main may very likely deplete the coalbed methane in reser-target for the exploration and exploitation of coalbed voirs". The pore structure of high rank coal contributesmethane is that the high rank coal produced much gas much to its storage ability, and the coal reservoirs in theduring the geologic history. If the preservation situations Qinshui Basin will become the preferable and significantcoordinated well, gas content in reservoirs would achieve reservoirs with high gas content and gas saturationideal target. However, in China many coalbed methane The processes and stages of methane production fromreservoirs are characterized by low permeability, low coalbeds are closely relative to the pore size of reservoirs,pressure and low gas saturation, so the zones with rela- and each process requires that the pores with relevant poretively high gas content become the principal exploration diameter be enriched to some extent and the pore shape bearea,while one of the key elements judging gas content is suitable for the gas migration, which demands a reasonne structure of pore system in reservoirs. In coalbeds, able configuration of pore diameter structure. If the poresmore than 80% of gas exists on the porous surface in ad- of some diameter develop extremely little, seepage will besorption state, and a minority of the macropores or cracks blocked, known as the"bottle neck"phenomenon ofexist in dissolved and free states. The contribution of mi- coalbed methane penetration in high rank coal reservoirscropore and transition pores to the intermal surtace area Because of the large internal surface area and the powerfulpredominates absolutely in Models 1-4(Fig. 4), and under adsorption in high rank coal reservoirs,difficult forthe same reservoir pressure, the adsorption ability gets the gas to be desorbed from the internal surface of poresenhanced remarkably. On the other hand, the less macropores and mesopores and the closed cleat result in the low Howe中国煤化工 the great differencespermeability of high rank coal in the Qinshui Basin, so the of poal reservoirs(Fig. 5)tectonic movements and the change of reservoir pressure DitterCNMHGhe possibility and theinduced by it cannot lead to the massive escape of coalbed quantity of gas desorption from coal internal surface andTherefore, it is favorable for continuous genera- the diffusion process afterwardChinese Science Bulletin Vol 50 Supp. December 2005ARTICLES176a080(b)00.64§0640.320.320.60.00.2040.60.81.00002040.60.81.060[(d)0960.64000.20.40.60.81.00.00.2040.60.81.0P/PFig 3. The nitrogen adsorption isothermal curve (a)Pore model 1; (b)pore model 2; (c)pore model 3, (d) pore model 410604010H」uPore modela micropore■ transitionB micropore transition pore D mesoporeFig. 4. The contribution of various grade pores for surface to volumeFig. 5. Various grade pore contents in pore modelsl-4ratio in pore models 1-4.exist. This type of pore structures is totally adverse to the3 Conclusions and discussionsproduction of coalbed methane. However, pore systemsAccording to the comparative analysis of data attained differ greatly from high rank coal reservoirs in the Qinshuifrom mercury penetration and cryogenic nitrogen test of Basin, This paper addresses the different characteristics ofabundant samples, the pore system of high rank coal res中国煤化工 servoirs. in the Qinshuiervoirs in the Qinshui Basin ordinarily have the common Bof geologic models forcharacteristics as follows: porosity is relatively low, pore theCNMHGrepresent the favorablesystem is dominated by micropores and transition pores, reservoirs with advantageous pore structures and relamesopores rank the second, and macropores nearly do not tively preferable physical properties; Model 3 representsChinese Science Bulletin Vol 50 Supp. December 2005ARTICLESthe reservoir with relatively adverse pore system and We think that the microcracks of coal reservoirs inmoderate or bad physical properties; while Model 4 Yangquan and Shouyang mines, north of the Qinshui basin,represents the reservoir with the most disadvantageous are better than those in middle and south, but the devel-pore structures and the worst physical properties. About opment of microcracks have a high heterogeneity whichthe high rank coal reservoirs in the Qinshui Basin, micro- may be influenced by partial tectonic movements, macer-pores and transition pores contribute much to surface-to als and mechanical property. At the same time, the stressvolume-ratio in overwhelming samples, making the ad- in situ controls the opening and closure of cracks. Thus,sorption ability of reservoirs extremely strong, and play. during selecting and evaluating a new block in high ranking an important part in adsorption, accumulation, preser- coal areas, it is necessary to make a deep-going study ofvation and pool-forming of coalbed methane during geo- the tectonic, microcracks and stress in situ of reservoirslogic history. On the other hand, it also brings some trou- and to give an objective evaluation for the exploration ofbles to the desorption and diffusion of coalbed methane. coalbed methaneThe productivity data of a single well is in Qinshui BasinAcknowledgements We thank Prof Liu Dameng for his constructiveis much less than that of America's which is always at comments on this manuscript. This work was supported by the Nationalributed to the low permeability. We consider that one of Project 973 of China( Grant No. 2002CB11702)the important reasons may be the limited desorption action Referencesof the coal internal surface, which results in the deficiencyof desorption gas and the high content of residual gas, so I. Lo, H B,R. Wilkins, W.T. Ellacott, M. V et al, Assessing thethe preferable permeability of cleat(crack) has no practi-maturity of coals and other rocks from North America using thecal meaningTherefore, the study on the complete desorpfluorescence alteration of multiple macerals(FAMM) technique,tion will be essential. Because of low content of macro-Intermational Journal of Coal Geology, 1997, 33(1): 61-7pore, the high rank coals become the"bottle neck"of gas 2. Alpern, B, Lemos, de Sousa, M. T, Flores, D, A progress reportpenetration. That is why some geologists took an increduclassification, International Joumal of Coal Ge-lous attitude to the recoverability of high rank coalbedology,1989,13(/4):1-19methane. Another factor improving the permeability of3. Zhang, S. H, Tang, D. Z, Wang, M.S., Main control factor af-coalbed is microcracks which communicates pore systemwith macrofractures. In Michigan basin, America, thefecting pore differential development of the coal reservoirs in Qin-shui Basin, Natural Gas Industry(in Chinese), 2005, 25(1): 37-40mechanism of genesis, occurrence, output and recovery 4. Huo, Y Z, Research on the features of gas desorption in coal res-technique of Antlem shale gas are the same as coalbedmethane's, and whose pore system is similar to the poreervoir, Natural Gas Industry(in Chinese), 2004, 24(5): 24-26system in high rank coal reservoirs. The reason of its suc5. Fu, X. H Qin, Y, Jiang, B et al., Physical and numerical simulacessful exploitation is that exogenetic microfractures andmacrofractures with high density and changeable directionof the Qinshui Basin, Shanxi, Chinese Journal of Geology(in Chi-are much enriched, which greatly improves the permeabilnese),2003,38(2):221l-22ity of reservoirs. The production model of coalbed gas 6. Fu, XH Qin, Y, Li, G.Z., An analysis on the principal controlproposed by Australian scholar Gamson, especiallyfactor of coal reservoir permeability in Central and Southern Qinstressed the effects of microcracks and mineral filling onshui Basin, Coal Geology Exploration(in Chinese), 2001, 29(3)the recoverability of coalbed methane. He indicated that 16-19the dimension, density, continuity and communication of 7. Ayer, W.B., Coalbed gas systems, resources, and production andmicrocracks made significant contributions to the totalare view of contrasting cases from the San Juan and Powder Riverpermeability of coalbed methane reservoirs. This theorybasins,AAPG2002;86(11853-1890gives an important and profitable inspiration to the explo- 8. Gamson, P, Beamish, B, Johnson, D, Coal microstructure andration and development of high rank coalbed methanethe Qinshui Basin Coal reservoirs of the Qinshui Basinmineralization: their effect on methane recovery in Coalbed Methexperienced some intense tectonic events and formedane and Coal Geology(eds, Gayer, R, Harme, 1. ) Geology Societymany secondary tectonic units, some of which should posSpecil Publication, 1996, 165-179.sess the possibility to form the suitable fracture systemReceived March 15, 2005; accepted July 1, 200中国煤化工CNMHGChinese Science Bulletin Vol 50 Supp. December 2005