Enhanced Coalbed Methane Recovery and CO 2 Storage in Coal Seams

Transcription

1 Enhanced Coalbed Methane Recovery and CO 2 Storage in Coal Seams Sevket Durucan Department of Earth Science and Engineering Imperial College London mining AND ENVIRONMENTAL ENGINEERING RESEARCH GROUP OUTLINE Background Coalbeds as both the reservoir and source rock Retention and release of gas in coal Coal permeability, dynamic permeability modelling in coalbed reservoir simulation Reservoir simulation of enhanced Coalbed Methane recovery and CO 2 storage in coal seams Conclusions

2 World Carbon Dioxide Storage Options Deep Ocean Saline Aquifers Unminable Coal Seams > Gt CO 2 Depleted Oil Reservoirs Depleted Gas Reservoirs Source: USGS Major Coal Basins 2

3 Major Coal Basins Major Coal Basins and Methane Resources Continent Country Coal Resources Methane Resources x 9 tonnes x 2 m 3 Europe and Belgium.7 the Russian France.6 Federation Germany Hungary.8 Poland Russia 6, 7-3 Ukraine 4.7 UK 9.7 North America Canada 7,.7-76 USA 3,97 Asia China 4, 3-3 India 6.8 Indonesia 6 Kazakhstan 7.3 Australia, Africa.8 World Totals ~2, ~ Source: ARI, 992 3

4 Methane Extraction from Coal Seams: Well Technology Underground Methane Drainage Practice Coalbed Methane Technology Coal as a Reservoir Rock: Structure Uniform and orthogonal fracture (CLEAT) structure microcleat 4

5 Coal as a Reservoir Rock: Structure butt cleat f face cleat l t microcleat Coal Matrix containing pores adsorbed gas Macropores > nm M Mesopores 2 nm free gas Micropores < 2 nm Cleat system (2mm - 2 mm) CH4 Pore surface area 2 2 m2 Coal as a Reservoir Rock Gas Retention butt cleat face cleat adsorbed gas free gas CH4 6 Gas s Content, m3/t Coal Matrix containing micropores CH4 adsorption isotherm Pressure, MPa 6 7 8

6 Coal as a Reservoir Rock: CH 4 / CO 2 Retention in Coal - Adsorption Isotherm, Gas content (m 3 /t t) CO 2 7%CO 2 2%CH 4 CH Pressure (MPa) Gas Content (scft/tonne) Langmuir equation: V = VL p P + p L ,, 2 Pressure (psi) Enhanced Coalbed Methane Recovery (ECBM) two principal methods of ECBM, namely N 2 and CO 2 injection (inert gas stripping and displacement sorption respectively) injection of nitrogen reduces the partial pressure of methane in the reservoir, thus promotes methane desorption without t lowering the total t reservoir pressure coal can adsorb approximately two to six times as much CO 2 by volume as methane, therefore, the assumption has been that the CO 2 injection stores 2-6 moles of CO 2 for every mole of CH 4 desorbed. 2 Gas content (m 3 /t) 2 CH 4 7% moist. coal CO 2 7%CO 2 2%CH 4 N 2 7% moist. coal 3 /t) Gas content (m 3 CH Pressure (MPa) Pressure (MPa) 6

7 San Juan Basin Simon Pilot Colorado New Mexico McElmo Dome CO 2 Field Coal Site COLORADO NEW MEXICO Durango Florida River Plant Simon Pilot Tiffany Unit Allison Unit Farmington Sweet Spot Overpressured: Sw =. Scale, miles 2 Underpressured: Sw <. Underpressured: Sw =. Simon Pilot Amoco Simon N 2 Injection Pilot,6 Total Gas Gas/Water Rate Mscf/D or STB/D,2 8 4 Natural Gas Nitrogen Start Nitrogen Injection End Nitrogen Injection Natural Water Gas Source: Wong, Gunter, Law and Mavor, 2 7

8 San Juan Basin Tiffany Unit Colorado New Mexico McElmo Dome CO 2 Field Coal Site COLORADO NEW MEXICO Durango Florida River Plant Simon Pilot Tiffany Unit Allison Unit Farmington Sweet Spot Overpressured: Sw =. Scale, miles 2 Underpressured: Sw <. Underpressured: Sw =. Tiffany Unit BP Tiffany Unit N 2 Injection (Full Scale Commercial Pilot) 9 Years of primary production, 2 N 2 injector wells, 34 CH 4 production wells, N 2 injection started in January 998, 4-6 fold increase in production early N 2 breakthrough Source: Reeves and Peckot, ARI, USCBM Symp, 2 Gas Production Water Production Nitrogen Injection Nitrogen Injection Suspended Rate, Mcfd or Bpd Water Measurement Discrepency May-9 Sep-9 Jan-93 Jun-94 Oct-9 Mar-97 Jul-98 Dec-99 Apr- Sep-2 8

9 Tiffany Unit BP Tiffany Unit N 2 Injection (Full Scale Commercial Pilot) Gas Production Water Production Nitrogen Injection Nitrogen Injection Suspended Rate, Mcfd or Bpd Water Measurement Discrepency May-9 Sep-9 Jan-93 Jun-94 Oct-9 Mar-97 Jul-98 Dec-99 Apr- Sep-2 Source: Reeves and Pecot, ARI, USCBM Symp, 2 San Juan Basin Allison Unit Colorado New Mexico McElmo Dome CO 2 Field Coal Site COLORADO NEW MEXICO Durango Florida River Plant Simon Pilot Tiffany Unit Allison Unit Farmington Sweet Spot Overpressured: Sw =. Scale, miles 2 Underpressured: Sw <. Underpressured: Sw =. 9

10 Burlington Resources Allison Unit CO 2 Injection McElmo Dome CO 2 Field Allison Unit 6 Years of primary production (988/89 99), 4 CO 2 injector wells, 9CH 4 production wells, CO 2 injection started in May 99, reduced CO 2 injectivity with time No significant CO 2 breakthrough Source: Reeves and Pecot, ARI, USCBM Symp, 2 Alberta Research Council, Fenn Big Valley CO 2 Injection Field micro-pilot testing started in 997 coal swelling and reduced permeability observed Source: Wong and Gunther, 999 Law, Van der Meer, Mavor and Gunter 2 Burlington Resources Allison Unit CO 2 Injection McElmo Dome CO 2 Field Allison Unit Source: Reeves, ARI, Coal-Seq Forum, 22

11 Coal Permeability and Gas Flow contrary to what is usually supposed, solid coal is extremely airtight, and lets very little air or gas through, even with a driving pressure of a whole atmosphere. Ivor GRAHAM, 96.that the rate of gas flow through the coal is a function of the difference in partial pressure of methane along the flow path. Therefore, the emission of methane from a lump of coal is not dependent on the total external pressure, but upon the partial pressure of the methane in the atmosphere and the pressure of the gas in coal. Ivor GRAHAM, 99 Strength, Elastic and Flow Properties of Coal Coal structure is highly elastic Young s Modulus, E (GPa) Coal (Various) Weak Reservoir Sandstone Sandstone Limestone Shale - 7 fractures matrix Coal permeability is Anisotropic Highly stress dependent Stress (MPa) Axial S 2 2 CAYDAMAR BARNSLEY COCSHEAD BANBURY DUNSIL DEEP HARD Axial Strain (millistrain)

12 Stress Effects and Permeability Permeability (m md) Intact coal Durucan Puri et al Somerton Effective Stress (psi) Permeability (k), - m Fractured coal Three Yard Great Row Cannel Row Confining Stress (MPa) Field Experience: San Juan Basin Field Permeability Behaviour k / k(p = 8 psi) Pressure dependent permeability multiplier Reservoir pressure (psia) (after McGovern, 24) 2

13 Pore Pressure Effects on Permeability: Matrix Shrinkage and Swelling fractures matrix tric strain ostrain) Volumet (micro Lorraine CO 2 CH 4 He Cell pressure (MPa) α V p ε = s s L P + L p Permeability (md).. CH 4 permeability (Lorraine) CO 2 permeability (Schwalbach) CO 2 permeability (Lorraine) Pore pressure (MPa) Primary and Enhanced Coalbed Methane Recovery Permeability Model 3

14 Model for Changes in Coalbed Permeability During Primary Recovery k = k e 3c f ( σ σ ) 3 b k = 2 a Bundled matchstick model b Flow c f cleat volume compressibility a σ σ - changes in effective horizontal stress (after Seidle et al., 992) σ σ = ν ν ( p p compaction term ) + EαS ( V V ) 3( ν) shrinkage term α S shrinkage coefficient V adsorbed gas volume E Young s modulus ν Poisson s ratio Schematic Permeability Behaviour σ σ p > p rb A Eα S ( V V ) ν σ σ = ( p p) 3( ν) ν p rc p rb p p ent (SCF/ton) Gas conte V L = 448 SCF/ton /b = 2 psi p rb Sorption pressure (psia) B A σ σ p < p rb p p rb p rc B p 4

15 Model Validation: US San Juan Basin History Match Valencia Canyon Wells: Steady increase in permeability Fairway B boomer Well: Strong permeability rebound k/k 2 2 VC 29-4 well test VC 32- (Mavor and Vaughn) VC 32-4 VC p/p k/k 2.. history matched permeability response curve at fairway B (Palmer and Mansoori) Reservoir pressure (psi) k/k 2 2 VC 32-4 VC 32- VC 29-4 well test present METSIMstudy Mavor and Vaughn p/p k/k 2 p = psi. φ =.8 Present model METSIM P & M model Reservoir pressure (psi) Permeability Model For Enhanced Recovery k = k e 3c ( σ f σ ) Cleat permeability Primary recovery σ σ = ν ν ( p p ) + EαS ( V V ) 3( ν) Shrinkage/swelling term Enhanced recovery σ σ ν = ( p p ν E ) + 3( ν) ( V Sj j j j= a Sj shrinkage/swelling coefficient for gas component j n α V )

16 Imperial College in-house ECBM Simulator METSIM2 3D two-phase Darcy flow in cleats, Fickian diffusion in matrix (allows for bidisperse diffusion) multi-component gas mixture (CH 2, CO 2, N 2 ) matrix shrinkage/swelling effects one permeability mixed gas sorption and diffusion extended Langmuir model Yubari Field Pilot, Hokkaido, Japan (after Fujioka, 26) In a mountainous national forest area Bounded by four major faults 6

17 Yubari Field Test Three stages CO 2 injection huff-puff test (well IW-, 7. tons CO 2 ) Multi-well CO 2 injection tests 24 (6 days, 3.7 tons CO 2 ) 2 (42 days,.4 tons CO 2 ) N 2 flooding test 26 N 2 -flooding (9 days, 32 tons N 2 ) 26 Pre- and Post-N 2 flooding CO 2 injection Well tests prior to the test and after 24 CO 2 injection ->.8 md CO 2 Injection and N 2 Flooding Test Field Injection and Gas Production Rates 4 Multi-well test: Production and injection rates 2 24: days injection tones CO 2 /day ( ton = 6 std. m 3 ) 2: 4 days injection.7 3. tones/day Low production rates: < 4 m 3 /day CO2 & N2 injected (ton) / Gas production rate (m 3 /d) CO 2 injection rate production rate Aug -Sep 3-Sep 2-Sep 7-Oct 9-Oct 3-Oct 2 CO2 & N2 injected and gas & water production rates 7 8 CO2 (ton) N2 (ton) 6 Gas (m3) Water (m3) 7 water production rate (m 3 /d) /d) Gas production rate (m 3 /d) CO2 injection rate (m 3 -Apr 8-Apr 2-Apr 2-May 9-May 6-May 23-May 3-May 6-Jun 26 7

18 2 Reservoir Simulation of the 26 N 2 Flooding Test 6 N 2 flooding: downhole tubing pressure Simulated N 2/CO 2 injection wellblock permeability and BHP / BHP (MPa) Iwellblock / Simulated injection wellblock pressure & permeability end of N 2 flooding start of CO 2 injection pressure permeability BHP rmeability (md) Wellblock pe ssure (MPa) Tubing pres model BHP N 2 injection time (day) Simulated vs. field N 2 injection BHP Elapsed time from the start of N 2 flooding (day) Model Prediction vs Field Data: Pre- and Post- N 2 Flooding CO 2 Injection Rates Pre-N 2 flooding Recorded daily amount injected (m 3 )/ CO2 injection rate (m 3 /d) model field (injected amount) field (24-hour rate) Apr -Apr 2-Apr 2-Apr 3-Apr -May -May 26 CO2 injection rate (m 3 /d) field Post N 2 flooding CO 2 injection rate model Post-N 2 flooding Injection time (day) N 2 flooding temporarily improved CO 2 injectivity, which declined quickly back to the pre-flooding level (~ 3 tones/ day) after two days. 8

19 Concluding remarks Coalbed reservoirs have unique characteristics: storage, transport and production mechanisms, permeability behaviour. Considerable advances in the reservoir simulation of ECBM, especially in permeability modelling have been made. While matrix shrinkage is desirable during primary recovery, CO 2 matrix swelling can have a severe impact on coalbed permeability and well injectivity. Long term fate of injected CO 2 in coalbeds is uncertain, as CO 2, especially at supercritical conditions, reacts with the reservoir rock and fluids. 9