- Email: [email protected]
Modeling and performance prediction for water production in CBM wells of an Eastern India coalfield
Journal of Petroleum Science and Engineering 103 (2013) 115–120
Contents lists available at SciVerse ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Modeling and performance prediction for water production in CBM wells of an Eastern India coalfield A. Agarwal, A. Mandal, B. Karmakar, K. Ojha n Department of Petroleum Engineering, Indian School of Mines, Dhanbad, India
a r t i c l e i n f o
abstract
Article history: Received 23 September 2011 Accepted 19 February 2013 Available online 14 March 2013
Dewatering of coal bed methane (CBM) reservoirs is a very important part of methane production. Efficient production depends very much on the proper designing of the wells. In the present study, a comprehensive testing is conducted on 17 wells of a particular block in Eastern India and a general reservoir flow equation is modeled. Prediction of the water flow potential of a particular well using the derived flow equation helps in monitoring the variables of the artificial lift facility. The outcome of work can be used comprehensively to predict the future water and gas flow rates of simulated wells under the designed test. & 2013 Elsevier B.V. All rights reserved.
Keywords: CBM dewatering water influx model flow potential well testing
1. Introduction Faster depletion of conventional resources and increasing demand for clean energy force India to hunt for alternatives to conventional energy resources. Coalbed methane (CBM) and shale gas are two of the most promising and potential resources of natural gas. India has approximately 4.6 trillion cubic meters (DGH Report, 2009–10) of reserve which may fulfill the country’s future growing energy demand to a large extent (Singh, 2002). However, production of gas from coalbed is very much different from conventional gas reservoir and requires utmost care and precautions in developing the production strategy. Coal is a complex heterogeneous system with methane gas remaining adsorbed on coal surface by lithostatic and hydrostatic pressure. Production of gas is controlled by depletion in pressure of the reservoir. Flow mechanism follows a three step process: (a) desorption of gas from the coal matrix, (b) diffusion to the cleat system, and (c) flow through fractures (Thimons and Kissell, 1973). Mostly, coal reservoirs are water saturated, and water provides the reservoir pressure to hold gas in the adsorbed state. As a result, water saturation in CBM reservoir is often nearly 100%, and this water must be produced to lower the reservoir pressure below the saturation pressure of methane to get desorbed and then produced. CBM formations are often in communication with an aquifer. As a consequence, it is likely that gas production from CBM reservoirs will result in encroachment of
n
Corresponding author. Tel.: þ91 9431125577; fax: þ 91 3262296632. E-mail address: [email protected] (K. Ojha).
0920-4105/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.petrol.2013.02.006
water from the associated aquifer. This additional water must subsequently be pumped to the surface along with the desired methane. A successful production strategy that reduces water production and increases the methane production will depend on a variety of factors including cleat spacing, aquifer strength, efficient dewatering technique and sorption characteristics (Sawyer et al., 1987). The initial stage of the CBM production is the dewatering process. However, inefficient production may cause the damage and change in coal seams properties, which in turn will reduce the methane production. Thus, water production becomes one of the key factors to optimize the methane production from CBM reservoirs. Designing of water withdrawal rate thus becomes an important criterion for efficient recovery of gas. In this investigation, the dewatering is modeled as a function of water level which will help in optimizing the rate and designing of artificial lift that is required in the future to produce methane at the depletion stage. Most of the promising CBM fields in India are lying in the Gondowana basin, specifically Jharia and Raniganj coal fields (Geological Survey of India, 1994); the present study area belongs to this field. Though, a number of investigations have been reported on the geological aspects of these (Sastry et al., 1977), detailed study on reservoir characterization or production aspect is yet to be carried out. The present paper discussed the dewatering technique of the CBM wells and optimization of water production rate for efficient and effective production of methane from coal beds, which not only optimizes the methane production but can also predict the performance of the wells of the basin. In general, water productivity of CBM wells is determined using type curve analysis
116
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
Nomenclature P.I. Q Pws Pwf
productivity index liquid flow rate, m3/day if not mentioned otherwise static bottom hole pressure flowing bottom hole pressure, psi
(Aminian et al., 2005), which is time consuming and required a number of data. Hence, simplified mathematical correlations have been proposed and developed for the wells to predict water production rate from the known water level or vice-versa.
2. Methods 2.1. Geology of the Raniganj basin, the study area Most of the CBM reserve in India is confined mainly to Raniganj and Jharia coalfields. So the study area, Raniganj, is chosen from this region. The region is under the Gandowana basin. Origin and deposition history of the coal are discussed here because these parameters play important roles in gas content and other properties of coal (Chandra, 1992). The Raniganj Basin has a semielliptical, elongated shape, and covers an area of 3000 km2 between the Damodar and Ajoy rivers (Ghosh, 2002). Raniganj Coalfield (Fig. 1) in West Bengal is the largest coalfield in India and comprises a total area of nearly 1260 km2 (Coal Map, 1993). Raniganj basin is the most prolific for CBM reserve. The evolution of the Gondwana basin was initiated almost at the beginning time period of Gondwana sedimentation, with the tectonic impulses generated by Permo-Triassic (Hercynian) orogeny (Pareek, 2004). Sediments in the Raniganj basin commenced with the deposition of glacial, peri-glacial sediments of Talchir formation. The lithostratigraphic sequences that are demonstrated in Fig. 1 can provide the idea of the different formations present in the sub-surface. This is also helpful in identification of the coal seams present in the subsurface. A subsequent idea of the gas and other properties of coal can also be generated with the idea of the formations. The Raniganj coalfield is elongated in an east–west direction following the trend of Damodar valley basins. The coal seams of the Raniganj
water flow rate at the surface conditions, m3/day Qw Pdynamic bottom hole flowing water pressure, psi shut in water pressure, psi Pstatic WLdyanmic, Ldyanmic dynamic water level in the well, m WLstatic static water level, m C, C1, n, n1 constant derived from the test plots
Formation in the Raniganj Coalfield area have a unique development pattern over a wide stretch. Synthesis of the large volume of data generated from the spurt in exploration activities during the post-nationalization period suggests complex multi-directional splitting and merging tendencies of the coal seam. The Raniganj Basin is one of the few coal fields of peninsular India where both the Lower Gondwana (Permian) and Upper Gondwana (Triassic–lower Cretaceous) formations are present. The generalized stratigraphy is given in Table 1 (Chandra, 1992; Ghosh et al., 1996). 2.2. Designing of well-testing in CBM wells Performance prediction tests are run routinely to measure oil, gas and water produced by a particular well under normal producing conditions. From the standpoint of well and reservoir operation, they provide periodic physical evidence of well conditions and unexpected changes if there is any. It is worth mentioning here that coal beds are very much fragile; rubbles and fines are formed during drilling as well as fracturing. These fines generally block the paths of water and gas to flow to the well if proper precaution is not taken. This can drastically reduce the water and gas production rate. Specially designed drilling fluid and fracturing fluids are used to prevent formation damage. Generally underbalanced drilling technology is adopted for drilling of coal bed. Application of light weight foam fluid which can take out the fines from fractures solves this problem. 2.2.1. Conventional wells For oil wells, results are usually reported as oil production rate. The test equipments consist of (i) a gas oil separator, (ii) a stock tank, with appropriate measuring devices such as (iii) an orifice meter for gas (iv) a hand tape for oil and water, and (v) down hole
Fig. 1. Geological map of Raniganj coal field.
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
117
Table 1 Generalized stratigraphic succession of the Raniganj Basin (after Chandra, 1992; Ghosh et al., 1996). Age
Formation
Litho-type
Maximum thickness (m)
Jurassic or tertiary Lawer Jurassic Upper Permian Middles Permian Lower Permian
–
800 730 1250
Upper Carboniferous
Talchir
Archean
Metamorphics
Dolerite dyke Mica lamprophyre and sills Fine grained feldspathic sandstones, shales with coal seams Buff-color sandstones, shales and carbonaceous shales Buff colored coarse and medium grained feldspathic sandstones, grits, Shales, carbonaceous shales and coal seams Greenish shale and fine grained sandstone Unconformity Granite, granite gneissess, quartzites, mica schists and amphibolites
Raniganj Barren measures Barakar
245 –
pressure gauges which are run through a coiled tubing unit. The tests must be conducted under stabilized production condition since change in rate often influences the relative quantities of oil, gas and water. For gas wells routine production tests are less common, since gas production is usually metered continuously from an individual well. The first problem confronting the petroleum engineer after the completion of an oil or gas well is to determine its capacity (productivity, potential). There are various other physical and mechanical parameters which may affect the delivery rate. These factors are different for conventional wells and CBM wells; thus the testing techniques are changed accordingly. For a conventional oil well, there are a number of well-defined tests performed in the field for the estimation of flow potential. To determine the productivity index, the well is produced at various rates and the flowing pressures are recorded. The productivity index and flow potential are then determined from the formulae P:I: ¼
Q Pws Pwf
ð1Þ Fig. 2. Test designed for water influx of the CBM well.
Flow potential ¼ P:I: P ws Subsequently many other tests are designed for oil wells because the above test alone may be unable to predict all the conditions of the well properly. Similarly, for a gas well, the test is called the back pressure test. The relation between the production rate of a well and the respective flowing pressure may be expressed empirically by the formula (Beggs, 2008) as given below: Q ¼ C ðP 2ws P 2wf Þn
where c ¼
703kh uTZ lnðr e =r w Þ
ð2Þ
where ‘Q’ is in standard cubic feet/day, ‘k’ in Darcy, ‘h’ in ft, m in cP, ‘T’ in R, ‘P’ in psi and r in ft. ‘z’ is the gas compressibility factor. 2.2.2. CBM wells However, CBM reservoir behaves differently from the conventional oil and gas reservoir. In CBM reservoir, dewatering is essential to facilitate gas production. It is very important that the dewatering is done with immense precaution so that the fragile coal seams will not be damaged and the gas will be produced efficiently. In the present investigation the above guideline has been properly considered and the results were carefully analyzed to generate the model for the evaluation of water rate. The test steps are designed (as shown in Fig. 2) as follows:
(i) First, the well is made to flow at a particular flow rate.
(ii) At stabilized dynamic water level in the annulus, at different bottom hole pressures, water flow rates are noted. (iii) The tests are performed at various rpm of the pump. (iv) Flowing bottom hole pressure (FBHP) vs. water flow rate are noted. The following assumptions were made during the test: (i) the shut-in pressure gradient on each well was considered to be according to the geological study conducted in this area and found to be 1.41 psi/m. This gives the static reservoir pressure as the wells have just started the dewatering phase. (ii) The fluid flow is considered single phase as in most of the cases flowing pressures are above critical desorption pressure of 754.2 psi, and the gas produced at surface is negligible. 2.3. Model to evaluate the productivity for a CBM well The CBM water productivity analysis is normally carried out using type curves but these methods are very time consuming. So in the present investigation, a model has been proposed which can estimate the water flow potential of the similar regions in the block from the dynamic water level only. The tests discussed above were conducted on a number of wells of the field and the results were carefully studied. Various empirical relations were made which could describe the flow property of the water in the initial dewatering phase. Finally a successful equation was derived which satisfied the flow conditions at the bottom hole. This model is very similar to the model used for the gas wells to evaluate the absolute open hole flow potential of gas and different from the conventional
118
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
Darcy equation. In general we use Darcy’s equation to describe fluid flow through a porous media with constant permeability. However, the situation is different for the coalbeds. Here, absolute permeability of coal cleats is stress dependent. Assuming the vertical lithostatic load is unchanged during depletion of pressure, the change in stress can be represented in terms of change in the reservoir pressure. As per the Palmer–Monsoori model (Palmer and Mansoori, 1998) and the IMC model (Shi and Durucan, 2003), there is a continuous change in porosity and permeability of coalbed with change in pressure. According to the IMC model, the permeability ratio (ratio of permeability at any pressure ‘P’ to that at unstressed condition) increases exponentially with depletion in pressure. Incorporating the permeability variation with pressure, the flow equation is expressed as Q ¼C
ðP 2static P 2dynamic Þn
Q w ¼ C ðP 2static P2dynamic Þn The above equation may be expressed in terms of the water level assuming constant water density. Hence Q w ¼ C1ðWL2static WL2dynamic Þn1
ð4Þ
As the static flow water level is generally known or constant, the production rate ‘Q’ could be expressed as a function of dynamic water level only. Once a number of data are available, the constants can be estimated and hence the water flow rate can be modified to optimize the dewatering phase of the CBM production well. Once the flow equation is known the well’s current flow water potential (FWP) can also be estimated using the formula given below. FWP ¼ C2 ðPdynamic Þ2n2 :
ð5Þ
The above equation may be expressed in terms of the water level as the density of formation water remains almost constant. Hence FWP ¼ C3 ðL2dynamic Þn3 :
ð6Þ
Using the above equation water flow potential can be determined easily with known dynamic water level. Hence, the pump rpm can be carefully monitored to change the water flow potential (water influx) of a well for optimizing the well conditions. Subsequent tests can also give an idea of the economic aspect of the well i.e. duration of dewatering period with total 5 well 6 well 5 well 4 well 3 well 2 well 1
Flow rate (m3/hr)
4
3
2
1
0 30
Time Fig. 4. Water flow potential vs. time of the CBM well.
ð3Þ
Analyses of these data also show that the water flow rate varies exponentially with the dynamic water level. The equation derived from the tests on CBM well was finally concluded to
0
Water Flow Potential
60
90
120
150
Dynamic Water level (m) Fig. 3. Variation of flow rate with water level.
180
disposed water volume and starting of gas production. In Fig. 3, the water production rate (m3/h) is plotted against the dynamic water level (m). Now using the production decline curve the water potential can be plotted to understand the dewatering profile and also the performance behavior of the well. Fig. 4 shows the way the water potential can be used to model a well and also to find the well behavior and the corresponding surface facility which will be required in the future. This figure is a general decline curve, and no field data are plotted here.
3. Results and discussions Petrophysical properties and coal characteristics are very important for analyzing the performance of the CBM reservoir. The important properties of the reservoir under study are described in Table 2. Well testing was conducted on a number of wells for characterization of the reservoir. The input data for the model equation were obtained from the various tests conducted on the 17 wells of the blocks as mentioned earlier. Fig. 5 shows the bottom hole pressures of different wells at the starting of the test. It could be observed clearly that pressure of the reservoir is not unique and the initial bottom-hole pressure varies from 723 psi to 1390 psi depending on the position of the wells. As bottom hole pressures of most of the wells are near or above the critical desorption pressure of the reservoir, these are initially under single phase flow conditions. The gas production starts when pressure falls below the critical desorption pressure as well as when the gas saturation attains a minimum value of critical saturation. For the present study, it was observed (Fig. 6) that the number of days required to start gas flow varies almost linearly with bottom hole flowing pressure. This is expected because as the days of water production increase, the reservoir pressure is being reduced. As soon as the free gas saturation reaches critical gas saturation, it starts to flow. However, as the gas flow rate was very less compared to the water flow rate, single phase flow was assumed in developing the model equation. Variation in water flow rate as a function of potential difference (P2ws P2wf) is plotted in Fig. 7 for different wells. Straight line is fitted to the data from each and every well. Average slope and intercept are determined using the least squares method. Slope and intercept of the plot of Log (Q) vs. Log (P2ws P2wf) are 5.1067 and 1.0296 respectively. Coal matrix in the CBM reservoir is very much heterogeneous and stress dependent as mentioned earlier. Permeability of coal matrix varies largely from well to well (space to space); even in the same well it may vary depending on pressure changes. The constant ‘C’ depends on the viscosity and density of water, permeability of
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
7.5
Table 2 Reservoir properties and coal properties of the study area. Parameter
Value
Average porosity (%) Average permeability (mD) Density of the coal (g/cc) Gas content (scf/ton) Langmuir volume (scf/ton) Langmuir pressure (psi) Critical desorption pressure (psi) Initial water saturation (%) Maximum depth (m) Maximum seam thickness (m) Sorption time constant (d)s
2.84 2.4 1.42 492 801 353.6 754.2 100 980 4.5 0.5181
119
Q =1.28×105 (Pws2 -Pwf2)1.0296 6.5
Well 1 Well 2 Well 3
5.5 Log(Q)
Well 5 Well 6 Well 7
4.5
Well 9 Well 12 Fit
3.5
2.5 -0.5
0
0.5
1
Log (Pws2-Pwf2) Fig. 7. Variation of water production rate with pressure.
Fig. 5. Static bottom hole pressure of different wells.
Fig. 8. Variation of water production rate with dynamic water level.
6.0 2
Log(Qmodel)
5.7
R =0.7346
Well 2 Well 5 Well 4
5.4
5.1
4.8 4.8
5.4
5.7
6.0
Log (Qtest)
Fig. 6. Time required for two phase flow as a function of time.
matrix at unstressed condition, and cleat porosity (Saulsberry et al., 1996). This may cause varied deviation of field data of different wells from the model. However, the standard deviation was determined using ‘63’ data from the formula sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P ðQ model Q f ield Þ2 S¼ n1
5.1
ð7Þ
Fig. 9. Validation of model equation (predicted vs. test production rate, m3/day).
The calculated standard deviation, S ¼0.28849. Fig. 8 shows the results of the test conducted on a number of wells. It may be observed from the figure that the water production rate increases with increase in the dynamic water level for each well. The relation between the water production rate and dynamic water level follows a power law model (Q a L2dynamic),
120
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
which is mentioned earlier. The values of constant ‘C3’ and ‘n3’ are determined from the equation of trend line as 1.8259 and 0.4799 respectively using data from seven wells. The equation becomes Q ¼ 1:825ðL2 Þ0:4799 :
Ranchi,, and the Department of Petroleum Engineering, Indian School of Mines, Dhanbad, India. Thanks are also extended to all individuals associated with the project.
ð8Þ
Eq. (8) is now validated with the data from three other wells as shown in Fig. 9. Production rates obtained from the model equation are plotted against the actual production rates of well nos. 2, 4 and 5 for the same dynamic water levels. From the results it could be observed that except for one datum, deviation of the test data from the model results is within the tolerable limit with a regression coefficient of 0.7346.
4. Conclusions Using field data from a number of wells, model equations have been derived to correlate water flow potential and production rate as functions of dynamic water level and pressure, respectively. Developed models can be used directly in the region to find the water potential of the pay zones and thus it can be used to design the artificial lift facility. The correlation is very simple and provides more accurate results from direct calculation compared to that of the conventional methods of CBM well test analysis. The water influx at a particular time can be formulated from the above mentioned method using the correlation of ‘Q’ with dynamic water level, ‘L’, at any time and thus the capacity of the wells (water influx) can be compared for different wells at equal time interval. Water flow potential vs. time plot may give the general decline of water influx of the well; it can also be extrapolated to find the water influx at the end of a particular time period. It can be utilized for determining the total dewatering period also by fixing the minimum water production rate. Thus, the developed model is a helpful mathematical tool in predicting the water production rate, dewatering period and designing of artificial lift for efficient production of methane from coal bed.
Acknowledgment We gratefully acknowledge the financial assistance provided by the Ministry of Coal, Govt. of India (CE/29), through CMPDI Ltd.,
References Aminian, K., Ameri, S., Bhavsar, A.B., Lakshminarayanan, S., 2005. Type curves for production and evaluation of coalbed methane reservoirs. SPE Reservoir Eval. Eng. 97957, 14–16. Beggs, H.D., 2008. Production Optimization Using Nodal Analysis, Tulsa OK. Oil & Gas Consultants International, Vol-II. Chandra. D., 1992, Geological Society of India, p. 13. Coal Map, 1993, Coal Atlas of India, Coal India Ltd. Geological Survey of India (GSI), 1994. Recent advances in the study of Gondwanas of Peninsular India. In: Proceedings of the IXth International Gondwana Symposium. Spec. Publ. G.S.I., Hyderabad, India, Calcutta, 30. Ghosh, S.C., 2002. The Raniganj Coal Basin: an example of an Indian Gondwana rift. Sediment. Geol. 147, 155–176. Ghosh, S.C., Nandi, A., Ahmed, G., Roy, D.K., 1996. Study of Permo–Triassic boundary in Gondwana sequence of Raniganj Basin. In: Proceedings of the IXth International Gondwana Symposium. Oxford and IBH Publication, New Delhi, Calcutta, pp. 195–206. Palmer, I., Mansoori, J., 1998. How permeability depends on stress and pore pressure in coalbeds: a new model. SPE 52607. SPE Reservoir Eval. Eng. 1 (6), 539–544 1998. Pareek., H.S., 2004. Progress of coal petrology in India, GSI. Memoir 57, 161. Petroleum Exploration and Production Activity Report, 2009–10. Directorate General of Hydrocarbon, Govt. of India. Sastry, M.V.A., Acharyya, S.K., Shah, S.C., Satsangi, P.P., Ghosh, S.C., Raha, P.K., Singh, G., Ghosh, R.N., 1977. Stratigraphic lexicon of Gondwana formations of India. Geol. Surv. India, Misc. Publ. 36, 1–170. Saulsberry, J.L., Schafer, P.S., Schraufnagel, R.A. (Eds.), 1996. A Guide to Coalbed Methane Reservoir Engineering. Gas Research Institute, USA. Sawyer, W.K., Zuber, M.D., Kuuskraat, V.A., Horner, D.H., 1987. Using reservoir simulation and field data to define mechanisms controlling coalbed methane production. In: Proceedings of the 1987 Methane Symposium, November 16– 19. Tuscaloosa, Alabama, pp. 295–307. Shi, J.Q., Durucan, S., 2003. Changes in permeability of coalbeds during primary recovery-Part-1: model formulation and analysis. In: Paper 0341 in International Coalbed Methane Symposium, University of Alabama, Tuscaloosa, Alabama, USA. Singh, U.P., 2002. Progress of Coalbed Methane in India. North American Coalbed Methane Forum, November. Thimons, B., Kissell, F.N., 1973. Diffusion of methane through coal. Fuel 52, 274–280.
Contents lists available at SciVerse ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Modeling and performance prediction for water production in CBM wells of an Eastern India coalfield A. Agarwal, A. Mandal, B. Karmakar, K. Ojha n Department of Petroleum Engineering, Indian School of Mines, Dhanbad, India
a r t i c l e i n f o
abstract
Article history: Received 23 September 2011 Accepted 19 February 2013 Available online 14 March 2013
Dewatering of coal bed methane (CBM) reservoirs is a very important part of methane production. Efficient production depends very much on the proper designing of the wells. In the present study, a comprehensive testing is conducted on 17 wells of a particular block in Eastern India and a general reservoir flow equation is modeled. Prediction of the water flow potential of a particular well using the derived flow equation helps in monitoring the variables of the artificial lift facility. The outcome of work can be used comprehensively to predict the future water and gas flow rates of simulated wells under the designed test. & 2013 Elsevier B.V. All rights reserved.
Keywords: CBM dewatering water influx model flow potential well testing
1. Introduction Faster depletion of conventional resources and increasing demand for clean energy force India to hunt for alternatives to conventional energy resources. Coalbed methane (CBM) and shale gas are two of the most promising and potential resources of natural gas. India has approximately 4.6 trillion cubic meters (DGH Report, 2009–10) of reserve which may fulfill the country’s future growing energy demand to a large extent (Singh, 2002). However, production of gas from coalbed is very much different from conventional gas reservoir and requires utmost care and precautions in developing the production strategy. Coal is a complex heterogeneous system with methane gas remaining adsorbed on coal surface by lithostatic and hydrostatic pressure. Production of gas is controlled by depletion in pressure of the reservoir. Flow mechanism follows a three step process: (a) desorption of gas from the coal matrix, (b) diffusion to the cleat system, and (c) flow through fractures (Thimons and Kissell, 1973). Mostly, coal reservoirs are water saturated, and water provides the reservoir pressure to hold gas in the adsorbed state. As a result, water saturation in CBM reservoir is often nearly 100%, and this water must be produced to lower the reservoir pressure below the saturation pressure of methane to get desorbed and then produced. CBM formations are often in communication with an aquifer. As a consequence, it is likely that gas production from CBM reservoirs will result in encroachment of
n
Corresponding author. Tel.: þ91 9431125577; fax: þ 91 3262296632. E-mail address: [email protected] (K. Ojha).
0920-4105/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.petrol.2013.02.006
water from the associated aquifer. This additional water must subsequently be pumped to the surface along with the desired methane. A successful production strategy that reduces water production and increases the methane production will depend on a variety of factors including cleat spacing, aquifer strength, efficient dewatering technique and sorption characteristics (Sawyer et al., 1987). The initial stage of the CBM production is the dewatering process. However, inefficient production may cause the damage and change in coal seams properties, which in turn will reduce the methane production. Thus, water production becomes one of the key factors to optimize the methane production from CBM reservoirs. Designing of water withdrawal rate thus becomes an important criterion for efficient recovery of gas. In this investigation, the dewatering is modeled as a function of water level which will help in optimizing the rate and designing of artificial lift that is required in the future to produce methane at the depletion stage. Most of the promising CBM fields in India are lying in the Gondowana basin, specifically Jharia and Raniganj coal fields (Geological Survey of India, 1994); the present study area belongs to this field. Though, a number of investigations have been reported on the geological aspects of these (Sastry et al., 1977), detailed study on reservoir characterization or production aspect is yet to be carried out. The present paper discussed the dewatering technique of the CBM wells and optimization of water production rate for efficient and effective production of methane from coal beds, which not only optimizes the methane production but can also predict the performance of the wells of the basin. In general, water productivity of CBM wells is determined using type curve analysis
116
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
Nomenclature P.I. Q Pws Pwf
productivity index liquid flow rate, m3/day if not mentioned otherwise static bottom hole pressure flowing bottom hole pressure, psi
(Aminian et al., 2005), which is time consuming and required a number of data. Hence, simplified mathematical correlations have been proposed and developed for the wells to predict water production rate from the known water level or vice-versa.
2. Methods 2.1. Geology of the Raniganj basin, the study area Most of the CBM reserve in India is confined mainly to Raniganj and Jharia coalfields. So the study area, Raniganj, is chosen from this region. The region is under the Gandowana basin. Origin and deposition history of the coal are discussed here because these parameters play important roles in gas content and other properties of coal (Chandra, 1992). The Raniganj Basin has a semielliptical, elongated shape, and covers an area of 3000 km2 between the Damodar and Ajoy rivers (Ghosh, 2002). Raniganj Coalfield (Fig. 1) in West Bengal is the largest coalfield in India and comprises a total area of nearly 1260 km2 (Coal Map, 1993). Raniganj basin is the most prolific for CBM reserve. The evolution of the Gondwana basin was initiated almost at the beginning time period of Gondwana sedimentation, with the tectonic impulses generated by Permo-Triassic (Hercynian) orogeny (Pareek, 2004). Sediments in the Raniganj basin commenced with the deposition of glacial, peri-glacial sediments of Talchir formation. The lithostratigraphic sequences that are demonstrated in Fig. 1 can provide the idea of the different formations present in the sub-surface. This is also helpful in identification of the coal seams present in the subsurface. A subsequent idea of the gas and other properties of coal can also be generated with the idea of the formations. The Raniganj coalfield is elongated in an east–west direction following the trend of Damodar valley basins. The coal seams of the Raniganj
water flow rate at the surface conditions, m3/day Qw Pdynamic bottom hole flowing water pressure, psi shut in water pressure, psi Pstatic WLdyanmic, Ldyanmic dynamic water level in the well, m WLstatic static water level, m C, C1, n, n1 constant derived from the test plots
Formation in the Raniganj Coalfield area have a unique development pattern over a wide stretch. Synthesis of the large volume of data generated from the spurt in exploration activities during the post-nationalization period suggests complex multi-directional splitting and merging tendencies of the coal seam. The Raniganj Basin is one of the few coal fields of peninsular India where both the Lower Gondwana (Permian) and Upper Gondwana (Triassic–lower Cretaceous) formations are present. The generalized stratigraphy is given in Table 1 (Chandra, 1992; Ghosh et al., 1996). 2.2. Designing of well-testing in CBM wells Performance prediction tests are run routinely to measure oil, gas and water produced by a particular well under normal producing conditions. From the standpoint of well and reservoir operation, they provide periodic physical evidence of well conditions and unexpected changes if there is any. It is worth mentioning here that coal beds are very much fragile; rubbles and fines are formed during drilling as well as fracturing. These fines generally block the paths of water and gas to flow to the well if proper precaution is not taken. This can drastically reduce the water and gas production rate. Specially designed drilling fluid and fracturing fluids are used to prevent formation damage. Generally underbalanced drilling technology is adopted for drilling of coal bed. Application of light weight foam fluid which can take out the fines from fractures solves this problem. 2.2.1. Conventional wells For oil wells, results are usually reported as oil production rate. The test equipments consist of (i) a gas oil separator, (ii) a stock tank, with appropriate measuring devices such as (iii) an orifice meter for gas (iv) a hand tape for oil and water, and (v) down hole
Fig. 1. Geological map of Raniganj coal field.
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
117
Table 1 Generalized stratigraphic succession of the Raniganj Basin (after Chandra, 1992; Ghosh et al., 1996). Age
Formation
Litho-type
Maximum thickness (m)
Jurassic or tertiary Lawer Jurassic Upper Permian Middles Permian Lower Permian
–
800 730 1250
Upper Carboniferous
Talchir
Archean
Metamorphics
Dolerite dyke Mica lamprophyre and sills Fine grained feldspathic sandstones, shales with coal seams Buff-color sandstones, shales and carbonaceous shales Buff colored coarse and medium grained feldspathic sandstones, grits, Shales, carbonaceous shales and coal seams Greenish shale and fine grained sandstone Unconformity Granite, granite gneissess, quartzites, mica schists and amphibolites
Raniganj Barren measures Barakar
245 –
pressure gauges which are run through a coiled tubing unit. The tests must be conducted under stabilized production condition since change in rate often influences the relative quantities of oil, gas and water. For gas wells routine production tests are less common, since gas production is usually metered continuously from an individual well. The first problem confronting the petroleum engineer after the completion of an oil or gas well is to determine its capacity (productivity, potential). There are various other physical and mechanical parameters which may affect the delivery rate. These factors are different for conventional wells and CBM wells; thus the testing techniques are changed accordingly. For a conventional oil well, there are a number of well-defined tests performed in the field for the estimation of flow potential. To determine the productivity index, the well is produced at various rates and the flowing pressures are recorded. The productivity index and flow potential are then determined from the formulae P:I: ¼
Q Pws Pwf
ð1Þ Fig. 2. Test designed for water influx of the CBM well.
Flow potential ¼ P:I: P ws Subsequently many other tests are designed for oil wells because the above test alone may be unable to predict all the conditions of the well properly. Similarly, for a gas well, the test is called the back pressure test. The relation between the production rate of a well and the respective flowing pressure may be expressed empirically by the formula (Beggs, 2008) as given below: Q ¼ C ðP 2ws P 2wf Þn
where c ¼
703kh uTZ lnðr e =r w Þ
ð2Þ
where ‘Q’ is in standard cubic feet/day, ‘k’ in Darcy, ‘h’ in ft, m in cP, ‘T’ in R, ‘P’ in psi and r in ft. ‘z’ is the gas compressibility factor. 2.2.2. CBM wells However, CBM reservoir behaves differently from the conventional oil and gas reservoir. In CBM reservoir, dewatering is essential to facilitate gas production. It is very important that the dewatering is done with immense precaution so that the fragile coal seams will not be damaged and the gas will be produced efficiently. In the present investigation the above guideline has been properly considered and the results were carefully analyzed to generate the model for the evaluation of water rate. The test steps are designed (as shown in Fig. 2) as follows:
(i) First, the well is made to flow at a particular flow rate.
(ii) At stabilized dynamic water level in the annulus, at different bottom hole pressures, water flow rates are noted. (iii) The tests are performed at various rpm of the pump. (iv) Flowing bottom hole pressure (FBHP) vs. water flow rate are noted. The following assumptions were made during the test: (i) the shut-in pressure gradient on each well was considered to be according to the geological study conducted in this area and found to be 1.41 psi/m. This gives the static reservoir pressure as the wells have just started the dewatering phase. (ii) The fluid flow is considered single phase as in most of the cases flowing pressures are above critical desorption pressure of 754.2 psi, and the gas produced at surface is negligible. 2.3. Model to evaluate the productivity for a CBM well The CBM water productivity analysis is normally carried out using type curves but these methods are very time consuming. So in the present investigation, a model has been proposed which can estimate the water flow potential of the similar regions in the block from the dynamic water level only. The tests discussed above were conducted on a number of wells of the field and the results were carefully studied. Various empirical relations were made which could describe the flow property of the water in the initial dewatering phase. Finally a successful equation was derived which satisfied the flow conditions at the bottom hole. This model is very similar to the model used for the gas wells to evaluate the absolute open hole flow potential of gas and different from the conventional
118
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
Darcy equation. In general we use Darcy’s equation to describe fluid flow through a porous media with constant permeability. However, the situation is different for the coalbeds. Here, absolute permeability of coal cleats is stress dependent. Assuming the vertical lithostatic load is unchanged during depletion of pressure, the change in stress can be represented in terms of change in the reservoir pressure. As per the Palmer–Monsoori model (Palmer and Mansoori, 1998) and the IMC model (Shi and Durucan, 2003), there is a continuous change in porosity and permeability of coalbed with change in pressure. According to the IMC model, the permeability ratio (ratio of permeability at any pressure ‘P’ to that at unstressed condition) increases exponentially with depletion in pressure. Incorporating the permeability variation with pressure, the flow equation is expressed as Q ¼C
ðP 2static P 2dynamic Þn
Q w ¼ C ðP 2static P2dynamic Þn The above equation may be expressed in terms of the water level assuming constant water density. Hence Q w ¼ C1ðWL2static WL2dynamic Þn1
ð4Þ
As the static flow water level is generally known or constant, the production rate ‘Q’ could be expressed as a function of dynamic water level only. Once a number of data are available, the constants can be estimated and hence the water flow rate can be modified to optimize the dewatering phase of the CBM production well. Once the flow equation is known the well’s current flow water potential (FWP) can also be estimated using the formula given below. FWP ¼ C2 ðPdynamic Þ2n2 :
ð5Þ
The above equation may be expressed in terms of the water level as the density of formation water remains almost constant. Hence FWP ¼ C3 ðL2dynamic Þn3 :
ð6Þ
Using the above equation water flow potential can be determined easily with known dynamic water level. Hence, the pump rpm can be carefully monitored to change the water flow potential (water influx) of a well for optimizing the well conditions. Subsequent tests can also give an idea of the economic aspect of the well i.e. duration of dewatering period with total 5 well 6 well 5 well 4 well 3 well 2 well 1
Flow rate (m3/hr)
4
3
2
1
0 30
Time Fig. 4. Water flow potential vs. time of the CBM well.
ð3Þ
Analyses of these data also show that the water flow rate varies exponentially with the dynamic water level. The equation derived from the tests on CBM well was finally concluded to
0
Water Flow Potential
60
90
120
150
Dynamic Water level (m) Fig. 3. Variation of flow rate with water level.
180
disposed water volume and starting of gas production. In Fig. 3, the water production rate (m3/h) is plotted against the dynamic water level (m). Now using the production decline curve the water potential can be plotted to understand the dewatering profile and also the performance behavior of the well. Fig. 4 shows the way the water potential can be used to model a well and also to find the well behavior and the corresponding surface facility which will be required in the future. This figure is a general decline curve, and no field data are plotted here.
3. Results and discussions Petrophysical properties and coal characteristics are very important for analyzing the performance of the CBM reservoir. The important properties of the reservoir under study are described in Table 2. Well testing was conducted on a number of wells for characterization of the reservoir. The input data for the model equation were obtained from the various tests conducted on the 17 wells of the blocks as mentioned earlier. Fig. 5 shows the bottom hole pressures of different wells at the starting of the test. It could be observed clearly that pressure of the reservoir is not unique and the initial bottom-hole pressure varies from 723 psi to 1390 psi depending on the position of the wells. As bottom hole pressures of most of the wells are near or above the critical desorption pressure of the reservoir, these are initially under single phase flow conditions. The gas production starts when pressure falls below the critical desorption pressure as well as when the gas saturation attains a minimum value of critical saturation. For the present study, it was observed (Fig. 6) that the number of days required to start gas flow varies almost linearly with bottom hole flowing pressure. This is expected because as the days of water production increase, the reservoir pressure is being reduced. As soon as the free gas saturation reaches critical gas saturation, it starts to flow. However, as the gas flow rate was very less compared to the water flow rate, single phase flow was assumed in developing the model equation. Variation in water flow rate as a function of potential difference (P2ws P2wf) is plotted in Fig. 7 for different wells. Straight line is fitted to the data from each and every well. Average slope and intercept are determined using the least squares method. Slope and intercept of the plot of Log (Q) vs. Log (P2ws P2wf) are 5.1067 and 1.0296 respectively. Coal matrix in the CBM reservoir is very much heterogeneous and stress dependent as mentioned earlier. Permeability of coal matrix varies largely from well to well (space to space); even in the same well it may vary depending on pressure changes. The constant ‘C’ depends on the viscosity and density of water, permeability of
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
7.5
Table 2 Reservoir properties and coal properties of the study area. Parameter
Value
Average porosity (%) Average permeability (mD) Density of the coal (g/cc) Gas content (scf/ton) Langmuir volume (scf/ton) Langmuir pressure (psi) Critical desorption pressure (psi) Initial water saturation (%) Maximum depth (m) Maximum seam thickness (m) Sorption time constant (d)s
2.84 2.4 1.42 492 801 353.6 754.2 100 980 4.5 0.5181
119
Q =1.28×105 (Pws2 -Pwf2)1.0296 6.5
Well 1 Well 2 Well 3
5.5 Log(Q)
Well 5 Well 6 Well 7
4.5
Well 9 Well 12 Fit
3.5
2.5 -0.5
0
0.5
1
Log (Pws2-Pwf2) Fig. 7. Variation of water production rate with pressure.
Fig. 5. Static bottom hole pressure of different wells.
Fig. 8. Variation of water production rate with dynamic water level.
6.0 2
Log(Qmodel)
5.7
R =0.7346
Well 2 Well 5 Well 4
5.4
5.1
4.8 4.8
5.4
5.7
6.0
Log (Qtest)
Fig. 6. Time required for two phase flow as a function of time.
matrix at unstressed condition, and cleat porosity (Saulsberry et al., 1996). This may cause varied deviation of field data of different wells from the model. However, the standard deviation was determined using ‘63’ data from the formula sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P ðQ model Q f ield Þ2 S¼ n1
5.1
ð7Þ
Fig. 9. Validation of model equation (predicted vs. test production rate, m3/day).
The calculated standard deviation, S ¼0.28849. Fig. 8 shows the results of the test conducted on a number of wells. It may be observed from the figure that the water production rate increases with increase in the dynamic water level for each well. The relation between the water production rate and dynamic water level follows a power law model (Q a L2dynamic),
120
A. Agarwal et al. / Journal of Petroleum Science and Engineering 103 (2013) 115–120
which is mentioned earlier. The values of constant ‘C3’ and ‘n3’ are determined from the equation of trend line as 1.8259 and 0.4799 respectively using data from seven wells. The equation becomes Q ¼ 1:825ðL2 Þ0:4799 :
Ranchi,, and the Department of Petroleum Engineering, Indian School of Mines, Dhanbad, India. Thanks are also extended to all individuals associated with the project.
ð8Þ
Eq. (8) is now validated with the data from three other wells as shown in Fig. 9. Production rates obtained from the model equation are plotted against the actual production rates of well nos. 2, 4 and 5 for the same dynamic water levels. From the results it could be observed that except for one datum, deviation of the test data from the model results is within the tolerable limit with a regression coefficient of 0.7346.
4. Conclusions Using field data from a number of wells, model equations have been derived to correlate water flow potential and production rate as functions of dynamic water level and pressure, respectively. Developed models can be used directly in the region to find the water potential of the pay zones and thus it can be used to design the artificial lift facility. The correlation is very simple and provides more accurate results from direct calculation compared to that of the conventional methods of CBM well test analysis. The water influx at a particular time can be formulated from the above mentioned method using the correlation of ‘Q’ with dynamic water level, ‘L’, at any time and thus the capacity of the wells (water influx) can be compared for different wells at equal time interval. Water flow potential vs. time plot may give the general decline of water influx of the well; it can also be extrapolated to find the water influx at the end of a particular time period. It can be utilized for determining the total dewatering period also by fixing the minimum water production rate. Thus, the developed model is a helpful mathematical tool in predicting the water production rate, dewatering period and designing of artificial lift for efficient production of methane from coal bed.
Acknowledgment We gratefully acknowledge the financial assistance provided by the Ministry of Coal, Govt. of India (CE/29), through CMPDI Ltd.,
References Aminian, K., Ameri, S., Bhavsar, A.B., Lakshminarayanan, S., 2005. Type curves for production and evaluation of coalbed methane reservoirs. SPE Reservoir Eval. Eng. 97957, 14–16. Beggs, H.D., 2008. Production Optimization Using Nodal Analysis, Tulsa OK. Oil & Gas Consultants International, Vol-II. Chandra. D., 1992, Geological Society of India, p. 13. Coal Map, 1993, Coal Atlas of India, Coal India Ltd. Geological Survey of India (GSI), 1994. Recent advances in the study of Gondwanas of Peninsular India. In: Proceedings of the IXth International Gondwana Symposium. Spec. Publ. G.S.I., Hyderabad, India, Calcutta, 30. Ghosh, S.C., 2002. The Raniganj Coal Basin: an example of an Indian Gondwana rift. Sediment. Geol. 147, 155–176. Ghosh, S.C., Nandi, A., Ahmed, G., Roy, D.K., 1996. Study of Permo–Triassic boundary in Gondwana sequence of Raniganj Basin. In: Proceedings of the IXth International Gondwana Symposium. Oxford and IBH Publication, New Delhi, Calcutta, pp. 195–206. Palmer, I., Mansoori, J., 1998. How permeability depends on stress and pore pressure in coalbeds: a new model. SPE 52607. SPE Reservoir Eval. Eng. 1 (6), 539–544 1998. Pareek., H.S., 2004. Progress of coal petrology in India, GSI. Memoir 57, 161. Petroleum Exploration and Production Activity Report, 2009–10. Directorate General of Hydrocarbon, Govt. of India. Sastry, M.V.A., Acharyya, S.K., Shah, S.C., Satsangi, P.P., Ghosh, S.C., Raha, P.K., Singh, G., Ghosh, R.N., 1977. Stratigraphic lexicon of Gondwana formations of India. Geol. Surv. India, Misc. Publ. 36, 1–170. Saulsberry, J.L., Schafer, P.S., Schraufnagel, R.A. (Eds.), 1996. A Guide to Coalbed Methane Reservoir Engineering. Gas Research Institute, USA. Sawyer, W.K., Zuber, M.D., Kuuskraat, V.A., Horner, D.H., 1987. Using reservoir simulation and field data to define mechanisms controlling coalbed methane production. In: Proceedings of the 1987 Methane Symposium, November 16– 19. Tuscaloosa, Alabama, pp. 295–307. Shi, J.Q., Durucan, S., 2003. Changes in permeability of coalbeds during primary recovery-Part-1: model formulation and analysis. In: Paper 0341 in International Coalbed Methane Symposium, University of Alabama, Tuscaloosa, Alabama, USA. Singh, U.P., 2002. Progress of Coalbed Methane in India. North American Coalbed Methane Forum, November. Thimons, B., Kissell, F.N., 1973. Diffusion of methane through coal. Fuel 52, 274–280.