New Three-Phase Modeling Capabilities in Eclipse
Transcript of New Three-Phase Modeling Capabilities in Eclipse
New Three-Phase ModelingCapabilities in Eclipse
PhD meeting
NTNU
7th February, 2008
by
Odd Steve Hustad
StatoilHydro R&D, Trondheim
Gas Based Recovery Methods
Gas
Water Oil
Water flooding
Gas flooding
2
Outline
• Extensions to Stones first model (SPE 24116)
– Available in Eclipse version 2008_1
• The ODD3P model option
– Available in Eclipse version 2007_1, immiscible option
– Available in Eclipse version 2008_1, miscible option
3
Modifications to Stone’s first model Introducing the n-exponent
SPE 24116
n = 1 n = 4.5 n = 25
4
Exponent’s impact on saturation distribution during flow• Low exponent value keeps oil between gas and water
• High exponent value causes gas to bypass oil
• This exponent parameter will appear in the 2008 version of Eclipse as a gridblock and directional dependent parameter
n = 1 n = 4.5 n = 25
5
• Exponent value of 4 was sufficient to match oil recovery curve
• Oil iso-curves are more linear in compliance with Oak et al. (JPT, Aug. 1990)
SPE 24116
Modifications to Stone’s first model sufficient for modeling equilibrium gas injection
Note
Sgmax ~ 62%
6
Tertiary gas displacement experiments
• Vertical 1.2 m long Bentheimer Core, water-wet condition
• Gas injection from top of core after water injection from below
1) Equilibrium gas
2) Dry separator gas
• Oil’s bubble point at 313.5 bar and 91.9 °C
– Bo=1.62
– GOR=200.9
– μo=0.43 cP
– σgo=1.2 mN/m
SPE 24116
WaterWater,
Oil & Gas
Oil & Water
Equilibrium Gas or
Separator Gas
7
Some conclusions from experiments
• Significant phase behavior effects from vaporization
• Long after drainage, oil still draining at end of experiments
• Modeling flash process from reservoir to ambient conditions important for proper fluid description
• Capillary pressure and relative permeability are important parameters when modeling experiment
SPE 24116
8
The flash process
• Sufficient number of pseudo-components are required for proper molecular weight representation (composition) in modeling gas injection with an Equation-of-state (EOS)
SPE 24116
Flash Procedure Saturation Pressure• If too few pseudo-components are used in an EOS model, significant errors in modeling saturation pressure may occur
9
Preliminary simulations using Stone’s models
Equilibrium Gas Injection
•Modification required to Stone model
•Water recovery profile proved most difficult to history match
10
SPE 24116
Modifications to Stone’s first model insufficient for modeling dry gas injection• Water recovery not properly modeled for dry gas injection experiment
• Vaporization impacts capillary forces which are not properly represented
• Correct modeling of maximum gas saturation remain a challenge
11
ODD3P three-phase model option• Developed to extend Stone’s water-wet three-phase models
• The ODD3P model extends the capability to handle other wetting conditions through each three-phase property being dependent on two two-phase properties
• The ODD3P model is an extension to the IKU3P model option in compositional Eclipse
– Relevant papers: SPE 74705, SPE 75138 and RUTH Program Summary book
– Three-phase properties are now consistent at all two-phase boundaries
– Couples capillary pressure and relative permeability through saturation
•
– Primary, secondary and tertiary process data may be applied
– Dynamic gridblock dependent end-point saturations for all processes
– Hysteresis capability (for three sets of two-phase data)
– miscibility capability
( )[ ]ciri PSfk =
12
ODD3P model concept
( ) ( )( ) ( )( ) ( )
Rock Type (Table) data
;
;
;
Gridblock Data, , , ,
w w cow g g cgo
row row w rgo rgo g
rwo rwo w rog rog g
r r r r r rgro grw org orw wrg wro
S S P S S P
k k S k k S
k k S k k S
S S S S S and S
= =
= =
= =
( ( ( (
( ( ( (( (
( ( ( (( (
k r g
r w
k r g
k r g
r o k r
g , o , w
w rS o r
S g o
w rS o r
S r
w
k g
o
o
k wo w
( S S S )
S gg
r
S ow
g w
100%Gas
100%Water
100%Oil
Model Input – Two-Phase Data
SPE 74705
13
Eclipse keywords in PROPS section• ODD3P
– Specifies to use model
• EPSODD3P
– Specifies end-point saturation scaling preference, reference and threshold values for interfacial tension and capillary number
• PCODD3P, PCODD3PW & PCODD3PG
– Specifies the capillary pressure and saturation calculation control parameters
• PSORG, PSGRO, PSORW, PSWRO, PSGRW & PSWRG
– Primary end-point saturations for gridblocks
• HSORG, HSGRO, HSORW, HSWRO, HSGRW & HSWRG
– Hysteresis end-point saturations for gridblock
14
Eclipse keywords in REGIONS section• PSTNUM, ISTNUM & DSTNUM
– Primary, increasing (water) and decreasing saturation table numbers
• SDROG, SDRGO, SDROW, SDRWO, SDRGW & SDRWG
– Initial gridblock direction saturation indicators
15
Eclipse keywords in SOLUTION section• DBGODD3P
– Debugging information
16
Eclipse keywords in SUMMARY section• BSDROG, BSDRGO, BSDROW, BSDRWO, BSDRGW & BSDRWG
– Gridblock phase saturation direction indicators
• BSGNRM, BSONRM & BSWNRM
– Gridblock normalized saturations
• BSOGTN, BSGOTN, BSOWTN, BSWOTN, BSGWTN & BSWGTN
– Gridblock turning point saturations
• BSOGNH, BSGONH, BSOWNH, BSWONH, BSGWNH & BSWGNH
– Gridblock hysteresis normalized saturations
• BSOGNE, BSGONE, BSOWNE, BSWONE, BSGWNE & BSWGNE
– Gridblock equivalent opposite direction normalized saturations
• BSORGB, BSGROB, BSORWB, BSWROB, BSGRWB & BSWRGB
– Residual saturations to calculate normalized saturations
17
Eclipse keywords in SUMMARY section cont.
• BSORGP, BSGROP, BSORWP, BSWROP, BSGRWP & BSWRGP
– Process dependent residual saturations
• BIFTGO, BIFTWO & BIFTGW
– Gridblock interfacial tension values
• BKROGN, BKRGON, BKROWN, BKRWON, BKRGWN & BKRWGN
– Gridblock representative relative permeability
• BKROGH, BKRGOH, BKROWH, BKRWOH, BKRGWH & BKRWGH
– Gridblock turning point relative permeability
• BKROGE, BKRGOE, BKROWE, BKRWOE, BKRGWE & BKRWGE
– Gridblock equivalent relative permeability
• BKROGT, BKRGOT, BKROWT, BKRWOT, BKRGWT & BKRWGT
– Gridblock opposite direction turning point relative permeability
18
Tertiary gas injection recovery, varying wettability
Water-wet Mixed-wet
Oil-wet
Book:
RUTH Program SummaryNPD, Stavanger, 1997
19
Pore scale modeling (eCore)Thin section
Network model
flowsimulation
3D pore space
pore scale physicswettabilty measurementscryo ESEM
wogs
0,0 0,2 0,4 0,6 0,8 1,0So
1E-4
1E-3
1E-2
1E-1
1E+0
measuredpredictedaverage
0 0,2 0,4 0,6 0,8 1
Sw
1E-4
1E-3
1E-2
1E-1
1E+0
krw (exp)krw (sim)krw (avg)kro (exp)kro (sim)kro (avg)
Core-scale multi-phase flow parameters
k rw
or k
row
two-
phas
e m
odel
k r
ogth
ree-
phas
e m
odel
Technology offered by Numerical Rocks AS
SPE 52052
20
Modeling of mixed-wet core flooding experiments• Improved model for saturation functions is established – ODD3P – in reservoir
simulation software
• Numerical generated two-phase saturation functions can be established through the eCore technology
• Can Core flood modeling be done better?
• Old core flooding experiments with mixed wettability revisited:
– Establishing “irreducible” water saturation
– Water displacement
– Gas displacement
21
Displacement experimentsComposite core and numerical model
• Vertical composite core of six core plugs
• Three gridblocks representing each core
• Pore volume of 143.57 cm3
• Two gridblocks were added to represent the end-pieces having zero capillary pressure
• Each end gridblock has a pore volume of 1 cm3
0.26
0.27
0.28
0.29
0.3
0.31
0.32
0.33
0.34
1.183
33.5
500
5.916
78.4
003
11.00
0313
.6003
15.98
3318
.1503
20.31
7322
.7503
25.45
0328
.1503
30.68
3333
.0503
35.41
7337
.1333
38.20
0339
.2673
Center depths from top of core (cm)
Poro
sity
(fra
ctio
n)
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
Gas
per
mea
bilit
y (m
d)
Porosity (fraction)
Gas Permeability (md)
22
Oil-water Relative permeability• Observe the crossing hysteresis relative permeability curves to oil from eCore• Oil-water relative permeability is the same for all the cores• Sorw is 6.8% and Swro is 23.82%
0.0
0.2
0.4
0.6
0.8
1.0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Water saturation
Oil-
wat
er re
lativ
e pe
rmea
bilit
y
RPWORPWOIRPWODRPOWRPOWIRPOWDHistory matched krwHistory matched krow
0.000000001
0.00000001
0.0000001
0.000001
0.00001
0.0001
0.001
0.01
0.1
1
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Water saturation
Oil-
wat
er re
lativ
e pe
rmea
bilit
y
RPWORPWOIRPWODRPOWRPOWIRPOWDHistory matched krwHistory matched krow
23
Gas-oil relative permeability• Secondary drainage (network) curves have been estimated• Note difference in estimated data through history matching• The gas-oil relative permeability curves are the same for all the cores, and these are at Swro
• Sorg is 5.8% and Sgro is 31%, both at Swro of 23.82%
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Gas saturation (fraction)
Rel
ativ
e pe
rmea
bilit
y
RPGOI - estimatedRPOGI - estimatedRPGODRPOGDRPGOPRPOGPHistory matched krogHistory matched krg
0.00000001
0.0000001
0.000001
0.00001
0.0001
0.001
0.01
0.1
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Gas saturation (fraction)
Rel
ativ
e pe
rmea
bilit
y
RPGOI - estimatedRPOGI - estimatedRPGODRPOGDRPGOPRPOGPHistory matched krogHistory matched krg
24
Capillary pressures• The cores are assumed of fine grained sandstone and gas-oil data are at Swro
• Capillary pressure curves are scaled according to Leverett J-function formulation from core to core. Curves for one core are illustrated here
• History matched data are significantly different
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Water saturation
Oil-
wat
er c
apill
ary
pres
sure
(bar
)
SWI
SWD
SW
History matched Pcow
0.001
0.01
0.1
1
10
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Gas saturation (fraction)
Cap
illar
y pr
essu
re (b
ar)
SGD
SGI - added
SGP
History matched Pcgo
25
Comparing gas-oil saturation functions• The presence of irreducible water does not impact the relative permeability data,
except near the end-point saturations.• Maximum gas saturation is lower in the presence of water• Slightly higher gas-oil J-function values are obtained at low gas saturation (<20%)
and different near the maximum gas saturation.• Residual gas saturation is slightly lower in the presence of water (0.8% points lower).
0.00000001
0.0000001
0.000001
0.00001
0.0001
0.001
0.01
0.1
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Gas saturation (fraction)
Gas
-oil
rela
tive
perm
eabi
lity
krgoi krogi krgodkrogd krgop krogpkrog(Sgd) @ Swir krgo(Sgd) @ Swir krog(Sgp) @ Swirkrgo(Sgp) @ Swir
0.01
0.1
1
10
100
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Gas saturation (fraction)
Gas
-oil
J-fu
nctio
n
Sgi
Sgd
Sgp
Sgd @ Swir
Sgp @ Swir
26
Primary Drainage• Initial composite core water saturation was 45% and 44.8% for the water flood and
gas flood experiment• With direct use of network model data a match was achieved. No history matching
or alterations performed!• Initial water saturation of 45% and 44.8% corresponds to 2.5 and 4.5 pore volumes
of oil injected
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0 1 2 3 4 5 6 7 8
Pore Volumes Injected
Cor
e av
erag
e w
ater
sat
urat
ion
Ave. Sw
Ave Sw= 0.45 at 2.46 PVI
Ave Sw= 0.448 at 4.5 PVI
0.3
0.35
0.4
0.45
0.5
0.55
0.6
0 5 10 15 20 25 30 35 40 45
Distance from top of core (cm)
Wat
er s
atur
atio
n (fr
actio
n)
Ave. Sw=0.4614 at 1.0659 PVI
Ave. Sw=0.4518 at 2.041 PVI
Ave. Sw=0.4503 at 2.4589 PVI
Ave. Sw=0.4477 at 6.9652 PVI
27
Water flooding• Oil-water Pc-curve had to be
modified/scaled by a factor 0.626• Application of Leverett J-function may
be questioned
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0 5 10 15 20 25 30 35 40 45
Distance from top of core (cm)
Wat
er s
atur
atio
n (fr
actio
n)
Ave. So=0.4499 at 0.1 PVI
Ave. So=0.379 at 0.2 PVI
Ave. So=0.3547 at 0.4 PVI
Ave. So=0.347 at 3 PVI
• The Amott wettability index of 0.1 may not be representative
• Good match of recovery curve was achieved with minor adjustments
• Zero capillary pressure will result in optimistic recovery
Water flooding LFR medium core
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 0.5 1 1.5 2 2.5 3 3.5
Pore volume injected
Oil
reco
very
(fra
ctio
n PV
)
Fig. 11Fig. 6No Pcow scaling, Sorw = 0.068Pcow scaled by 0.001, Pcow~0Pcow scaled by 0.626
28
Gas Flooding
• Omitting capillary pressure in Stone’s first model results in optimistic recovery• Too low oil recovery with eCore data and ODD3P model with no adjustments• Reducing oil-water capillary pressure gave more correct oil recovery level and
adjusting gas-oil capillary pressure had little impact • A fair match was achieved with ODD3P when the capillary pressure was turned off
in the solution of the pressure equation. This example show significant CPU time improvement
• Simulated scenarios used Stone’s first model and ODD3P model–Only primary data applied in Stone’s first model (no hysteresis)
–All eCore data applied using ODD3P model
• Stone’s first model “matched” the oil recovery, but the input gas-oil capillary pressure had to be adjusted by a factor 4!
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Pore volume injected
Oil
reco
very
(fra
ctio
n PV
)
Fig. 22Primary data only: Pcow*0.626, Pcgo*4.0, Stone 1Primary data only: Pcgo=Pcow=0, Stone 1Pc used as is with no adjustments, Swir=0, odd3pPcow*0.537, No Pcgo adjustment, Swir=0, odd3pPc turned off in pressure eq'n, Swir=0, odd3p
29
Gas Flooding cont.
• The simulated final gas saturation show large gradient with the composite core
• The initial water saturation is redistributed when gas enters the core with less end effects
• The gas displaces mainly the oil volume and not the water volume during the gas displacement.
• The gas injection process does not leave primary curve due to breakthrough (short-circuiting flow) and high end-point hysteresis saturation (Srg>30%)
• The laboratory boundary conditions seem to have little impact on the saturation end-effects (i.e. large gradients near the outlet) due to the wetting conditions
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1.2 3.6 5.9 8.4 11.0 13.6 16.0 18.2 20.3 22.8 25.5 28.2 30.7 33.1 35.4 37.1 38.2 39.3
Distance from top inlet (cm)
Sat
urat
ion
(%)
Final Sg
Final So
Final Sw
0
0.1
0.2
0.3
0.4
0.5
0.6
0 5 10 15 20 25 30 35 40
Distance from top of core (cm)
Sat
urat
ion
(frac
tion)
Initial Sw Final Sw
Final So Final Sg
30
Concluding remarks• Reservoir simulation technology improvements by StatoilHydro innovation
• StatoilHydro is approaching predictive capability when modeling core displacement experiments
• Work is ongoing to validate eCore technology for modeling core flooding experiments using ODD3P
• Further research still required
– Need more data containing multi-phase flow
• Lack good experimental data with different wettability, miscibility, hysteresis, etc.
– Improve the modeling capability and understanding the application of experimental data
• Better relationships between the parameters for multi-phase flow
31
The End
Thank you for listening
Questions?