Geomechanical aspects of coalbed methane (CBM) production: Flow model formulation François B ERTRAND & Frédéric C OLLIN I NTRODUCTION Coalbed methane (CBM): miner’s curse ⇒ valuable fuel. From [1]. Coals = naturally fracturated reservoirs Blocks delimited by two sets of orthogonal fractures (fractures = cleats). Coal deposits = (generally) aquifers → methane maintained adsorbed within the coal matrix by the hydrostatic pressure. ↓ CBM production = generate a pressure drop by dewatering the cleats. → Gas molecules diffuse in the matrix to reach the cleats which are preferential pathways (higher permeability). From [2]. Two distinct phenomena affecting permeability: 1. Pressure depletion → Reservoir compaction → Cleat permeability & 2. Gas desorption → Coal matrix shrinkage → Cleat permeability % H YDRAULIC MODEL Flow model • Matrix: diffusive gas flow → Fick’s law (Continuum modelling) J g i = -D ∂C g ∂x i (1) Direct modelling Continuum modelling • Cleats: advective flow – Macroscopic approach: Darcy’s law (Continuum modelling) – Microscopic approach: solve Navier-Stokes between two parallel plates (Direct modelling) h x x 2 1 v (x ) 1 2 q = - h 2 12 · 1 μ dp dx 1 (2) → Equivalent to Darcy with k = h 2 12 For unsaturated conditions: k r (S r ) accounts for the reduction in permeability. q i = - k r · k ij μ · ∂p ∂x j (3) Gas storage → in the matrix Pressure Adsorbed volume V L P L V L 2 V g,Ad [m 3 /kg ]= V L · p P L + p (4) P L and V L : Langmuir’s parameters [3] → Adsorbed gas density: ρ g,Ad [kg/m 3 ]= ρ coal · ρ g,std · V g,Ad (5) Mass balance equations Microscopic approach • Matrix (Continuum modelling) – Gas ∂ ∂t (ρ g,Ad )+ ∂ ∂x i (J g i )= E Matrix→Cleats (6) • Cleats (Direct modelling) – Gas ∂ ∂t (ρ g,f (1 - S r,w )) + ∂ ∂x i (ρ g,f q g,f i ) | {z } Gas phase + ∂ ∂t (Hρ g,f S r,w )+ ∂ ∂x i (Hρ g,f q w i + S r,w J g i ) | {z } Dissolved gas in water phase = E Cleats→Matrix (7) – Water ∂ ∂t (ρ w S r,w )+ ∂ ∂x i (ρ w q w i ) =0 (8) Gas desorption Matrix Cleat Gas diffusion Gas flow E Water flow C ONTACT I NFORMATIONS [email protected] [email protected] R EFERENCES [1] S. Laubach, R. Marrett, J. Olson, and A. Scott, “Characteristics and origins of coal cleat: a review,” International Journal of Coal Geology, vol. 35, no. 1, pp. 175–207, 1998. [2] A. Al-Jubori, S. Johnston, C. Boyer, S. W. Lambert, O. A. Bustos, J. C. Pashin, and A. Wray, “Coalbed methane: clean energy for the world,” Oilfield Review, vol. 21, no. 2, pp. 4–13, 2009. [3] I. Langmuir, “The adsorption of gases on plane surfaces of glass, mica and platinum.,” Journal of the American Chemical society, vol. 40, no. 9, pp. 1361–1403, 1918. C ONCLUSION Changes in reservoir properties = crucial issue for CBM recovery. But sorption- and stress-induced coal permeability alteration are improperly simplified by classical macroscopic modelling approaches! → It is preferable to use a "Microscopic" approach because the discretization is made at the scale of the cleats and matrix. However, the computational cost is too expensive at the scale of a reservoir. → The microscopic model will be the basis for a multi-scale approach. A CKNOWLEDGMENT Ces recherches sont soutenues par le FRIA (Fonds pour la Recherche dans l’Industrie et l’Agriculture).
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Geomechanical aspects of coalbed methane (CBM) production:Flow model formulation
François BERTRAND & Frédéric COLLIN
INTRODUCTION
Coalbed methane (CBM): miner’s curse⇒ valuable fuel.
From [1].
Coals = naturally fracturated reservoirs
Blocks delimited by two sets of orthogonal fractures(fractures = cleats).
Coal deposits = (generally) aquifers
→methane maintained adsorbed within the coalmatrix by the hydrostatic pressure.
↓CBM production = generate a pressure drop by
dewatering the cleats.
→ Gas molecules diffuse in the matrix to reach the cleats which arepreferential pathways (higher permeability). From [2].
Two distinct phenomena affecting permeability:
1. Pressure depletion → Reservoir compaction → Cleat permeability↘2. Gas desorption → Coal matrix shrinkage → Cleat permeability↗
HYDRAULIC MODEL
Flow model
• Matrix: diffusive gas flow→ Fick’s law (Continuum modelling)
Jgi = −D∂Cg
∂xi(1)
Direct modelling Continuum modelling
• Cleats: advective flow
– Macroscopic approach: Darcy’s law (Continuum modelling)
– Microscopic approach: solve Navier-Stokes between two parallel plates (Direct modelling)
hx
x2
1
v (x )1 2
q = −h2
12· 1µ
dp
dx1(2)
→ Equivalent to Darcy with k = h2
12
For unsaturated conditions: kr(Sr) accountsfor the reduction in permeability.
qi = −kr · kijµ
· ∂p∂xj
(3)
Gas storage → in the matrix
Pressure
Ad
so
rbe
d v
olu
me
VL
PL
VL
2
Vg,Ad [m3/kg] =VL · pPL + p
(4)
PL and VL: Langmuir’s parameters [3]
→ Adsorbed gas density:
ρg,Ad [kg/m3] = ρcoal · ρg,std · Vg,Ad (5)
Mass balance equationsMicroscopic approach
• Matrix (Continuum modelling)
– Gas ∂
∂t(ρg,Ad) +
∂
∂xi(Jgi) = EMatrix→Cleats (6)
• Cleats (Direct modelling)
– Gas
∂
∂t(ρg,f (1− Sr,w)) +
∂
∂xi(ρg,f qg,f i)︸ ︷︷ ︸
Gas phase
+∂
∂t(Hρg,f Sr,w) +
∂
∂xi(Hρg,f qwi + Sr,w Jgi)︸ ︷︷ ︸
Dissolved gas in water phase
= ECleats→Matrix
(7)
– Water ∂
∂t(ρwSr,w) +
∂
∂xi(ρw qwi) = 0 (8)
Gas desorption
Matrix
Cleat
Gas diffusion
Gas flow
E
Water flow
CONTACT INFORMATIONS
[email protected]@ulg.ac.be
REFERENCES
[1] S. Laubach, R. Marrett, J. Olson, and A. Scott, “Characteristics and origins of coal cleat: a review,” International Journalof Coal Geology, vol. 35, no. 1, pp. 175–207, 1998.
[2] A. Al-Jubori, S. Johnston, C. Boyer, S. W. Lambert, O. A. Bustos, J. C. Pashin, and A. Wray, “Coalbed methane: cleanenergy for the world,” Oilfield Review, vol. 21, no. 2, pp. 4–13, 2009.
[3] I. Langmuir, “The adsorption of gases on plane surfaces of glass, mica and platinum.,” Journal of the American Chemicalsociety, vol. 40, no. 9, pp. 1361–1403, 1918.
CONCLUSION
Changes in reservoir properties = crucial issue for CBM recovery.
But sorption- and stress-induced coal permeability alteration are improperly simplified by classical macroscopic modelling approaches!
→ It is preferable to use a "Microscopic" approach because the discretization is made at the scale of the cleats and matrix.
However, the computational cost is too expensive at the scale of a reservoir.
→ The microscopic model will be the basis for a multi-scale approach.
ACKNOWLEDGMENT
Ces recherches sont soutenues par le FRIA (Fonds pour laRecherche dans l’Industrie et l’Agriculture).
Related Documents
