Modeling Wellbore and Reservoir CO2 Flow for the Heletz ... Belinkov 2012_EGU.pdf · the experiment...
1
Phase states along the pipe Bottom depth [m] 1632.8 Final shut-in pressure [PSI] 2140 Final shut-in pressure [Pa] ~1.47E+07 TDS [mg/l] 53570 Model salinity mass fraction [-] 0.04988 Temperature [°C] 67 Field carbon dioxide injection experiments are necessary for demonstration, increasing the scientific understanding and quantification of the relevant processes occurring during geological storage in deep saline aquifers. As part of the large scale EU-FP7 project MUSTANG, a carbon dioxide injection experiment is to be carried out at the Heletz site, Israel. Estimating the well head conditions is an important part of planning the experiment and an approach is presented here for determining wellhead conditions needed to ensure that at least a specified flow rate is provided to the formation, given target layer conditions, while still respecting pressure buildup constraints. The main part of the study combines the multiphase flow in the target layer using the well known TOUGH2/ECO2N model, with the flow in the injection pipe solving the 1D steady, real gas, augmented Euler system. The Matching is carried on in a two-stage process. The second part consists of a transient simulation of the combined well-reservoir flow using the new T2WELL software for the same parameters. Preliminary conclusions of the comparison of the two strategies are derived. Sensitivity analyses were carried out with respect to target layer properties and to pipe model assumptions M.Kitron Belinkov* (1) , K.Rasmusson (2) , M.Rasmusson (2) , F.Fagerlund (2) , A.Niemi (2) , J.Bensabat (3) , J. Bear (1) Modeling Wellbore and Reservoir CO 2 Flow for the Heletz Experiment *(1) Technion – I.I.T., Department of Civil, Environmental and Water Resources engineering, Haifa 32000, Israel, corresponding author:[email protected] (2) Uppsala University – Department of earth sciences, Villavagen 16, S-75236 Uppsala, Sweden. (3) Environmental & Water Resources Engineering Ltd., PO Box 6770, Haifa 31067, Israel, [email protected] EGU2012- 7823 XL23 ERE2.2 1. Reservoir Info Conclusions *Formation permeability has a significant effect on the pressure at the injection point. * The effect of the geology (here simulated as the type of boundary condition) is crucial not only on the size but on the direction the change of the reservoir press. * The reliability of the result depends on the quality of the representation of the actual site by the characteristic functions. Objective 1D Steady Single phase, single component, compressible Heat Transfer: (lateral only) Friction force: , Real gas: Benedict-Webb-Rubin EOS Boundary conditions: known & given at injecting end. Mass balance 0 v dz d g R v f dz dv v dz dp 4 2 g R v z Q dz dv v dz dh 2 ) ( dz dT T p dz d p dz dp T dz dT dz dp T T dz dh c p p ] ) 1 ( 1 [ Momentum balance Energy balance Diff. form of any EOS: p=p(ρ,T) 5. Well flow: Assumptions & model * R v f 4 2 )) ( ) ( ( 2 ) ( z T z T HTC R z Q WB 01 . 0 Re 32 . 0 25 . 0 f 6. Well Flow results Acknowledgement This research was funded by the European Community's Seventh Framework Program FP7/2007-2013 under grant agreement n° 227286, as part of the MUSTANG project. Enthalpy dependence on p and T Figure showing the location of faults (black), the injection well (blue x), top of caprock and the bottom of the caprock, target layers and interbedded shale layers at the Heletz site. Pipe Depth[m] Pressure[Pa] Pipe Depth[m] Pipe Depth[m] Pipe Depth[m] Pipe Depth[m] Density[Kg/m3] Enthalpy[Pa] Velocity[m/s] Temperature[ 0 K] Legend (Well Head conditions): 70 Bar, 25 0 C 80 Bar, 25 0 C 70 Bar, 35 0 C 80 Bar, 35 0 C * Gravity plays an important effect on the flow – on well bottom pressure. * Friction plays a negligible effect for the presented range of parameters. * Heat transfer coefficient has a significant effects on WB temperature & density. * Heat transfer coefficient - a weak effect on well bottom pressure. * The size of the heat transfer coefficient for Heletz is not determined. * CO 2 changes smoothly from liquid or gas to supercritical state – single phase * The matching procedure guarantees the prescribed injection rate at least and predicts the reservoir pressure for the prescribed injection rate. * Preliminary comparison with T2Well modeling show differences in WB pressures and CO 2 behavior along the well. The models differ in many aspects. 9. Well Flow and final Conclusions A tool which enables adaptation of the CO 2 well head conditions to the reservoir data and mass flow rate is established. 140 160 180 200 220 240 260 60 70 80 90 100 110 120 130 Bottom well pressure[bar] Well head pressure [bar] Q=1 ton/h, gas Q=1 ton/h, liquid Q=3 tons/h, gas Q=3 tons/h, Liquid Q=5 tons/h, gas Q=5 tons/h, liquid Q=7 tons/h, gas Q=7 tons/h, liquid 7. Matching well and reservoir simulations 2. Reservoir models The table below shows the radial symmetrical reservoir models created for different combinations of radius, boundary conditions and permeability values. A: One sandstone layer (10.6 m) and one shale layer, total depth 18 m. B: Multi layered, sandstone and shale, total depth 18m. Model Radius [m] Boundary Conditions Permeability cases [mD] A 500 Dirichlet (static pressure profile) – open boundary 10, 50, 100, 250 and 500 A 500 Estimated site volume – closed boundary 10, 50, 100, 250 and 500 A 500 Neumann “no flux” – closed boundary 10, 50, 100, 250 and 500 A 250 Neumann “no flux” – closed boundary* 10, 50, 100, 250 and 500 B 500 Dirichlet (static pressure profile) – open boundary 10, 50, 100, 250 and 500 B 500 Estimated site volume – closed boundary 10, 50, 100, 250 and 500 B 500 Neumann “no flux” – closed boundary 10, 50, 100, 250 and 500 B 250 Neumann “no flux” – closed boundary* 10, 50, 100, 250 and 500 Characteristic function: k=5.0x10 -13 m 2 , ϕ=17%, λ =0.762, =7.5 kPa and S wr =0.3 (Dana & Skoczylas, 2002). For the different sandstone permeability cases the function was scaled and a porosity of 0.143 used. The shale was assumed to be impermeable. 3. Injection scenarios Non-isothermal injection simulations of ~1000 tons of CO 2 was run with the multiphase flow simulator TOUGH2/ECO2N (Pruess, 2005) for 1, 3, 5 and 7 tons/h injection rates. 4. Results of reservoir simulations Figure showing pressure in the bottom well element during non-isothermal simulations of CO 2 injection into the formation models A and B, for four different boundary conditions and k=100 mD. The initial pressure was 1.47x10 7 Pa. Figure showing maximum pressure in the injection element during non-isothermal simulations of CO 2 injection into model B. No results are shown for the Q=1ton/h, k=500 mD case. References Dana, E., Skoczylas, F. (2002). Experimental study of two-phase flow in three sandstones. II. Capillary pressure curve measurement and relative permeability pore space capillary models. International Journal of Multiphase Flow 28 (11), 1965-1981. Lu, M., Connell, L.D. (2008). Non-isothermal flow of carbon dioxide in injection wells during geological storage. Int. J. Green. Gas Control, 2, 248-258. Pruess, K. (2005). ECO2N: A TOUGH2 fluid property module for mixtures of water, NaCl, and CO2. LBNL-57952. Pan, L. et al. (2011). T2Well/ECO2N: Model for coupled well-reservoir Flow of CO2, LBNL 4291-E. Matching procedure Preliminary well simulation: Input - Well head Mass Flow Rate, Temperature, Pressure Output - Well bottom Temperature, Pressure Reservoir simulation: Input – Mass flow rate, Temperature from the well sim. Output – Time dependent Reservoir Pressure Matching: Planning well head Pressure so that the Well bottom Pressure surpasses time dependent Reservoir Pressure In the Low WH Pressure values (65-90 bar) there are significant differences between the bottom well pressure values (BHP) of the liquid and the gaseous phases. Differences decrease with increasing WHP. When WHP are low it is impossible to calculate in the lower range flow rates in the gaseous phase. See the phase space graph in section 6. The part of the graph which seems feasible physically & numerically, for initial pressure values higher than 100 bar which result in well bottom values larger than 220 bar, is inapplicable from an engineering point of view: expensive to maintain highly pressurized CO 2 on the ground, and exceeds instrumentation bounds at well bottom. Density[Kg/m3] Temperature[ 0 K] Pipe Depth[m] Pipe Depth[m] Pipe Depth[m] Pipe Depth[m] Velocity[m/s] Pressure[Pa] 8. Well Flow – preliminary Comparison with T2Well Parameters Well Head conditions: P=76 bar, T=305 0 K(32 0 C) Diam. 6cm Legend Lu& Connell T2Well
Transcript of Modeling Wellbore and Reservoir CO2 Flow for the Heletz ... Belinkov 2012_EGU.pdf · the experiment...
Phase states along the pipe
Bottom depth [m] 1632.8
Final shut-in pressure [PSI] 2140
Final shut-in pressure [Pa] ~1.47E+07
TDS [mg/l] 53570 Model salinity mass fraction [-] 0.04988
Temperature [°C] 67
Field carbon dioxide injection experiments are necessary for demonstration, increasing the scientific understanding and quantification of the relevant processes occurring during geological storage in deep saline aquifers. As part of the large scale EU-FP7 project MUSTANG, a carbon dioxide injection experiment is to be carried out at the Heletz site, Israel. Estimating the well head conditions is an important part of planning the experiment and an approach is presented here for determining wellhead conditions needed to ensure that at least a specified flow rate is provided to the formation, given target layer conditions, while still respecting pressure buildup constraints. The main part of the study combines the multiphase flow in the target layer using the well known TOUGH2/ECO2N model, with the flow in the injection pipe solving the 1D steady, real gas, augmented Euler system. The Matching is carried on in a two-stage process. The second part consists of a transient simulation of the combined well-reservoir flow using the new T2WELL software for the same parameters. Preliminary conclusions of the comparison of the two strategies are derived. Sensitivity analyses were carried out with respect to target layer properties and to pipe model assumptions
M.Kitron Belinkov*(1), K.Rasmusson (2), M.Rasmusson(2), F.Fagerlund(2), A.Niemi(2), J.Bensabat (3), J. Bear(1)
Modeling Wellbore and Reservoir CO2 Flow for the Heletz Experiment
*(1) Technion – I.I.T., Department of Civil, Environmental and Water Resources engineering, Haifa 32000, Israel, corresponding author:[email protected]
(2) Uppsala University – Department of earth sciences, Villavagen 16, S-75236 Uppsala, Sweden. (3) Environmental & Water Resources Engineering Ltd., PO Box 6770, Haifa 31067, Israel, [email protected]
EGU2012- 7823 XL23 ERE2.2
1. Reservoir Info
Conclusions *Formation permeability has a significant effect on the pressure at the injection point.
* The effect of the geology (here simulated as the type of boundary condition) is crucial not only on the size but on the direction the change of the reservoir press.
* The reliability of the result depends on the quality of the representation of the actual site by the characteristic functions.
Objective
1D
Steady
Single phase, single component, compressible
Heat Transfer:
(lateral only)
Friction force: ,
Real gas: Benedict-Webb-Rubin EOS
Boundary conditions: known & given at
injecting end.
Mass balance 0vdz
d
gR
vf
dz
dvv
dz
dp
4
2
gRv
zQ
dz
dvv
dz
dh
2
)(
dz
dT
T
p
dz
dp
dz
dp
T
dz
dT
dz
dp
TT
dz
dhcpp
])
1
(1
[
Momentum balance
Energy balance
Diff. form of any EOS: p=p(ρ,T)
5. Well flow: Assumptions & model *
R
vf
4
2
))()((2)( zTzTHTCRzQ WB
01.0Re32.0 25.0 f
6. Well Flow results
Acknowledgement This research was funded by the European Community's Seventh Framework Program FP7/2007-2013 under grant agreement n° 227286, as part of the MUSTANG project.
Enthalpy dependence on p and T
Figure showing the location of faults (black), the injection well (blue x), top of caprock and the bottom of the caprock, target layers and interbedded shale layers at the Heletz site.
Pipe Depth[m]
Pre
ssu
re[P
a]
Pipe Depth[m] Pipe Depth[m] Pipe Depth[m]
Pipe Depth[m]
De
nsi
ty[K
g/m
3]
Enth
alp
y[P
a]
Ve
loci
ty[m
/s]
Tem
pe
ratu
re[0
K]
Legend (Well Head conditions):
70 Bar, 250 C 80 Bar, 250C 70 Bar, 350C 80 Bar, 350C
* Gravity plays an important effect on the flow – on well bottom pressure.
* Friction plays a negligible effect for the presented range of parameters.
* Heat transfer coefficient has a significant effects on WB temperature & density.
* Heat transfer coefficient - a weak effect on well bottom pressure.
* The size of the heat transfer coefficient for Heletz is not determined.
* CO2 changes smoothly from liquid or gas to supercritical state – single phase
* The matching procedure guarantees the prescribed injection rate at least and
predicts the reservoir pressure for the prescribed injection rate.
* Preliminary comparison with T2Well modeling show differences in WB pressures
and CO2 behavior along the well. The models differ in many aspects.
9. Well Flow and final Conclusions A tool which enables adaptation of the CO2 well head conditions to the reservoir data and mass flow rate is established.
140
160
180
200
220
240
260
60 70 80 90 100 110 120 130
Bott
om
wel
l p
ress
ure
[bar]
Well head pressure [bar]
Q=1 ton/h, gas
Q=1 ton/h, liquid
Q=3 tons/h, gas
Q=3 tons/h, Liquid
Q=5 tons/h, gas
Q=5 tons/h, liquid
Q=7 tons/h, gas
Q=7 tons/h, liquid
7. Matching well and reservoir simulations
2. Reservoir models The table below shows the radial symmetrical reservoir models created for different combinations of radius, boundary conditions and permeability values. A: One sandstone layer (10.6 m) and one shale layer, total depth 18 m. B: Multi layered, sandstone and shale, total depth 18m.
Model Radius [m] Boundary Conditions Permeability cases [mD]
A 500 Dirichlet (static pressure profile) – open boundary
10, 50, 100, 250 and 500
A 500 Estimated site volume – closed boundary 10, 50, 100, 250 and 500
A 500 Neumann “no flux” – closed boundary 10, 50, 100, 250 and 500
A 250 Neumann “no flux” – closed boundary* 10, 50, 100, 250 and 500
B 500 Dirichlet (static pressure profile) – open boundary
10, 50, 100, 250 and 500
B 500 Estimated site volume – closed boundary 10, 50, 100, 250 and 500
B 500 Neumann “no flux” – closed boundary 10, 50, 100, 250 and 500
B 250 Neumann “no flux” – closed boundary* 10, 50, 100, 250 and 500
Characteristic function: k=5.0x10-13 m2, ϕ=17%, λ =0.762, =7.5 kPa and Swr=0.3 (Dana & Skoczylas, 2002). For the different sandstone permeability cases the function was scaled and a porosity of 0.143 used. The shale was assumed to be impermeable.
3. Injection scenarios Non-isothermal injection simulations of ~1000 tons of CO2 was run with the multiphase flow simulator TOUGH2/ECO2N (Pruess, 2005) for 1, 3, 5 and 7 tons/h injection rates.
4. Results of reservoir simulations
Figure showing pressure in the bottom well element during non-isothermal simulations of CO2 injection into the formation models A and B, for four different boundary conditions and k=100 mD. The initial pressure was 1.47x107 Pa.
Figure showing maximum pressure in the injection element during non-isothermal simulations of CO2 injection into model B. No results are shown for the Q=1ton/h, k=500 mD case.
References Dana, E., Skoczylas, F. (2002). Experimental study of two-phase flow in three sandstones. II. Capillary pressure curve measurement and relative permeability pore space capillary models. International Journal of Multiphase Flow 28 (11), 1965-1981. Lu, M., Connell, L.D. (2008). Non-isothermal flow of carbon dioxide in injection wells during geological storage. Int. J. Green. Gas Control, 2, 248-258. Pruess, K. (2005). ECO2N: A TOUGH2 fluid property module for mixtures of water, NaCl, and CO2. LBNL-57952. Pan, L. et al. (2011). T2Well/ECO2N: Model for coupled well-reservoir Flow of CO2, LBNL 4291-E.
Matching procedure Preliminary well simulation: Input - Well head Mass Flow Rate, Temperature, Pressure Output - Well bottom Temperature, Pressure Reservoir simulation: Input – Mass flow rate, Temperature from the well sim. Output – Time dependent Reservoir Pressure Matching: Planning well head Pressure so that the Well bottom Pressure surpasses time dependent Reservoir Pressure
In the Low WH Pressure values (65-90 bar) there are significant differences between the bottom well pressure values (BHP) of the liquid and the gaseous phases. Differences decrease with increasing WHP.
When WHP are low it is impossible to calculate in the lower range flow rates in the gaseous phase. See the phase space graph in section 6. The part of the graph which seems feasible physically & numerically, for initial pressure values higher than 100 bar which result in well bottom values larger than 220 bar, is inapplicable from an engineering point of view: expensive to maintain highly pressurized CO2 on the ground, and exceeds instrumentation bounds at well bottom.
De
nsi
ty[K
g/m
3]
Tem
pe
ratu
re[0
K]
Pipe Depth[m] Pipe Depth[m]
Pipe Depth[m] Pipe Depth[m]
Ve
loci
ty[m
/s]
Pre
ssu
re[P
a]
8. Well Flow – preliminary Comparison with T2Well
Parameters Well Head conditions: P=76 bar, T=3050K(320C) Diam. 6cm
Legend Lu& Connell T2Well
