Pressure Buildup Data Collection and Analysis from the Frio Brine Pilot GCCC Digital Publication Series #05-03a S. M. Benson Cited as : Benson, S.M., Pressure buildup data collection and analysis from the Frio Brine Pilot (poster): presented at the American Geophysical Union Fall Meeting, San Francisco, California, December 5–9, 2005, paper GC13A-1218, unpaginated [1 p.]. GCCC Digital Publication #05-03a, pp. 1. Keywords : Pressure Transient Testing Program, Water Flooding, Relative Permeability, CO2 Fractional Flow
Transcript
Pressure Buildup Data Collection and Analysis from the Frio Brine Pilot
GCCC Digital Publication Series #05-03a
S. M. Benson
Cited as: Benson, S.M., Pressure buildup data collection and analysis from the Frio Brine Pilot (poster): presented at the American Geophysical Union Fall Meeting, San Francisco, California, December 5–9, 2005, paper GC13A-1218, unpaginated [1 p.]. GCCC Digital Publication #05-03a, pp. 1.
Keywords: Pressure Transient Testing Program, Water Flooding, Relative Permeability, CO2 Fractional Flow
Pressure Buildup Data Collection and Analysis from the Frio Brine Pilotpaper GC13A-1218
Sally M. Benson, Lawrence Berkeley National Laboratory, Earth Sciences Division, Berkeley, California
Pressure transient data from the Frio Brine Pilot
Downhole pre s s ure a nd te mpe ra ture da ta c olle c tion
P re s s urea nd T e mpe ra tureT ra ns duc e r (1483 m)
P a cke r
T ubingE le c tric C a ble
T o Da ta Ac quis itionS ys te m
1540 m
1546 m
P e rfora te dC a s ing
Inje ction We llPre-CO2 injection pumping test 24 hoursRecirculation water pumping test 14 daysCO2 injection test 10 days with 4 shut-in periods
Water pumping and injection tests forformation characterization
Injection Well Pressure Response
146
147148
149
150151
152
9/4 9/6 9/8 9/10 9/12 9/14 9/16 9/18 9/20 9/22
Pre
ssure
(bar)
Drawdown at the Injection Well
0
0.2
0.4
0.6
0.8
0 20000 40000 60000 80000Time (s)
� p (b
ars)
MeasuredCalculated
k = 2.1x10-12 m2
φch=1.1x10-8 m/Pa
Pumping Well Buildup at Start of Injection
00.10.20.30.40.50.6
0 30000 60000 90000 120000
time(s)
Π p (b
ars)
Measured
Calculated
k = 2.1x10-12 m2
φch=1.1x10-8 m/Pas = 9
Injection Well Pressure Buildup
00.5
11.5
22.5
33.5
44.5
0 20000 40000 60000 80000
Time (s)
� p
(bar
)
MeasuredCalculated
Formation Characterization
s=12Multi-rate, multi-well analysis
1.1x10-82.1x10-12FalloffInjectionInjection
Multi-rate, multi-well analysis
1.1x10-82.1x10-12BuildupPumpingInjection
s = 9Multi-rate, multi-well analysis
1.1x10-82.1x10-12BuildupInjectionInjection
Multi-rate analysis
1.1x10-82.2x10-12DrawdownInjectionPumping
Commentsφch(m/Pa)k (m2)Test TypeMonitoring Well
Active Well
CO2 injection test pressure transient data
Injection and observation well data
Interpretation of multi-phasepressure transient data
Comparison Between Pressure Buildup at the Injection and Observation Wells
Immiscible displacement of water by CO2
Relative permeability effects Capillary pressure effects Adverse Mobility ratio
Pressure and tempurature dependent CO2
viscosity and densityPartitioning of CO2 into the water phasePartitioning of water into the CO2 phase
Buckley-Leverett type displacementVertical equilibriumHorizontal reservoirHomogeneous reservoirNeglect capillary pressureSlightly compressible fluid
rw rf
qC O2
hk,ct,φ
∞rw rf
qC O2
hk,ct,φ
∞
A new kind of inversion
Injection Well Observation Well
∆p
∆p
time time
∆p
time
?
∆pi(t) – ∆po(t)Injection Well Observation Well
∆p
∆p
time time
∆p
time
?
∆pi(t) – ∆po(t) rw rwLog r (m)
∆p
t2t1 t3 t4∆p1
∆p4
C O2 F r o n t
rw rwLog r (m)
∆p
t2t1 t3 t4∆p1
∆p4
C O2 F r o n t
AbstractAs part of the Frio Brine Pilot, downhole pressure measurements were obtained from both the injection well and observation well throughout the injection and recovery phases of the test. In addition, a pre-injectioninterference test was used to obtain accurate information about the permeability and compressibility of the formation. By comparing the twodata sets it is possible to obtain information on field-scale relative permeability during CO2 injection. Both numerical and analytical techniqueshave been used to interpret the pressure transient data. Comparison between pressure transients at the injection well and observation well allowinference about changes in near-well permeability in the vicinity of theinjection well. This is the first time a data set such as this has beencollected for CO2 injection into a saline formation, and the results provideimportant insights into the multi-phase flow behavior of CO2. They alsoprovide important information for predicting the injectivity of saline formations.
Interpretation of water pumping and injection test
Processed pressure data
ObjectivesPressure transient data from the Frio Brine Pilot was collected tocharacterize the single and multi-phase hydrological properties of theinjection interval.
Overview of the Pressure Transient Testing Program
A two-well interference test was carried out to obtain the permeability andstorativity of the formation prior to CO2 injection. During CO2 injection, continuous pressure transient measurements were made in the injectionand observation wells. The configuration of the pressure transient monitoring system is shown in Figure 1. Because the pressure transducerwas located about 50 m above the perforated interval of the injection well, corrections to the baseline pressure were needed once CO2 injection began(e.g. the borehole was now filled with a fluid having a density about 0.64 times that of water). Similar corrections were needed in the observationwell after CO2 breakthrough occurred.
Figure 1. Schematic of downhole pressure monitoring system.
The pumping rate for the test is shown in Figure 2. Pressure drawdown andbuildup data from the observation/injection well are shown in Figure 3.
Figure 2. Water pumping and injection rate. Figure 3. Injection well pressure response.
Pumping and Injection Rate Injection Well Pressure Response
Characterization of the Single Phase Hydrological Properties of the FrioFormation
The hydrological properties of the formation were determined using conventional single and multi-rate well test analysis methods. Figures 4, 5 and 6 illustrate the history matches for the pressure drawdown and buildup analysis. The Theis solution matched the data well, indicating that overthe test period the formation behaved as an infinite, homogeneous formation of constant thickness. The low value for the storativity of the formationsuggested that no free-phase gas was present before CO2 injection began. Asummary of the hydrological parameters determined are provided in Table 1.
Figure 6. Match between measured and calculated pressure buildup at the pumping well.
Figure 5. Match between measured and
calculated pressure buildup.
Table 1. Summary of hydrologic parameters.
Pressure Buildup During CO2 Injection
The pressure buildup during CO2 injection is shown in Figures 7 and 8. The testtook place over 12 days, with 4 separate injection periods. The injection rate wasabout 3.2 l/s. As discussed above, because the pressure transducers were locatedabove the perforated interval, when CO2 displaced the water in the wellbore, the baseline pressure changed. For the injection well, this happened shortly afterinjection began, as illustrated by the marked increase in pressure that occurredshortly after this. The appropriate pressure baselines for the CO2 filled boreholesare shown in Figures 7 and 8.
Figure 7. Pressure buildup data during CO2 injection.
Figure 8. Pressure buildup at the observation well during CO2 injection.
Figure 4. Match between measured and calculated pressure drawdown.
Figure 9 shows the complete pressure transient record for both the injectionand observation wells after correcting for the changed pressure baseline. Interestingly, the pressure buildup in the injection well reaches a maximumof 2.5 bars shortly after injection begins and continues to decline thereafterto a nearly steady value of 1.5 bars. (The proposed explanation for theunusual observation is provided shortly.) The pressure buildup in the observation well reaches a maximum of 1.2 bars.
Figure 9. Processed pressure transient data corrected for changing baseline.
Interpretation of pressure transient data during CO2 injection must consider the factors listed in Table 2. These data were interpreted based on an approximateanalytical solution developed by Benson (1984, 1987) that was adapted for multi-phase flow. The assumptions used in this solution are provided in Table 3.A schematic showing the underlying conceptual model is shown in Figure 10.
µwater >>µCO2
Table 2. Considerations for pressure transient analysis of CO2 injection into brine formations.
Table 3. Assumptions used in the development of the approximate analytic solution
1. Estimate fg
g
g sdsdf
by the breakthrough time of CO2 a t the obse rva tion
we ll, where fg
g
g sdsdf
is the de riva tive of the fractiona l gas flow curve a t the
CO2 front.
2. Estimate gfrg
g skf
by history matching the pressure buildup data, where
gfrg
g skf
is the ratio between the fractional flow of gas and the relative
permeability to gas at the CO2 front.
3. Knowing fg
g
g sdsdf
and gfrg
g skf
, it is possible to determine rls and rgs if the
relative permeability to gas is assumed to be of the form
grlr
lrl
rg
sssss
wheressk
−−−
=
−−=
1*
*)1)(*1( 22
Procedure for interpreting data
Figure 10. Conceptual model.
The match between the calculated and measured pressure for the injectionwell is shown in Figure 13. Based on the parameters obtained from the injection well analysis, the match between the observed and calculated response at the observation well is shown in Figure 14. The excellent matchsuggests that the parameters obtained from the analysis of the injection well provide reliable estimates of the multi-phase flow parameters.
The relative permeability to gas obtained by using this procedure is shownin Figure 11. In addition, the fractional flow curve, including the CO2 saturationat the front and average saturation behind the front, is shown in Figure 12.
Figure 13. Match between the measured and calculated pressure buildup at the injection well.
A New Kind of Inversion Scheme for Pressure Transient Analysis
In addition to the analysis method provided above, a new kind of inversion forinterpreting multi-phase pressure transient interference test is proposed. Theinversion is based on the idea that the time rate of change of the pressuredifference between the buildup at the injection and observation well can revealimportant information about multiphase flow. Figures 15 and 16 illustrates this concept. As time proceeds, the CO2 front migrates between the injectionand observation wells. Because of the different hydrologic properties inside and outside of the front, the slope of the distance versus drawdown data will differin the two regions. The consequence of this is that over time, unike under single phase conditions, the pressure difference between the two wells will varyover time, yielding information about the rate of advance and hydrologic properties of the CO2 flooded region.
This is illustrated in Figure 17. Four curves are shown: the measured data;the calculated data for pure water and pure CO2 systems; and the calculated response for a multi-phase flow system using the parameters obtained fromthe analysis provided above. For an all water system, the pressure drop shouldbe constant at about 1.5 bars. For an all CO2 system, the pressure drop would beabout 0.15 bar. The measured data falls between these two extremes. Also, asshown, after the first injection period there is good agreement between the measured data and predicted response using the multi-phase flow parametersprovided above. This provides aditional suport for the validity of the parametersestimated using the procedure described above. It can also be used as anindependent method to estimate flow parameters.
Figure 17 also provides an indication of the pressure buildup that would be associated with the skin factor of 9 that was determined from the water injection test. Thispressure buildup accounts for most of the difference between the measured andcalculated pressure difference. Based on the gradual convergence between themeasured and predicted pressure differences, we hypothesize that the skin factordecreased during CO2 injection. This was not confirmed by follow-up testing.
Figure 15. Conceptual model for new inversion.
Figure 14. Match between the measured and calculated pressure buildup at the observation well.
Figure 16. Pressure profiles.
Figure 17. Comparison between measured and predicted pressure differencebetween the injection and observation wells.
A high quality pressure transient data set was collected from the Frio Brine Pilot.From this, it was possible to use an approximate analytical solution to estimatemulti-phase flow parameters. Follow-up work will include numerical simulation andinversion of these data to provide an independent assessment of the multi-phaseflow parameters. In the future, the pressure transducers should be located within the perforated interval to avoid the need for correcting the pressure datato a changing baseline.
References
S.M. Benson (1984) "Analysis of Injection Tests in Liquid-Dominated Geothermal Reservoirs," M. S. Thesis, University of California, Berkeley; Report No. LBL-17953, Lawrence Berkeley National Laboratory, Berkeley, California.
S.M. Benson, J.S. Daggett, E. Iglesias, V. Arellano, and J. Ortiz-Ramirez (1987) "Analysis of Thermally Induced Permeability Enhancement in Geothermal Injection Wells," Proceedings of the 12th Annual Workshop, Geothermal Reservoir Engineering, January 20-22, 1987, Stanford, California; Report No. LBL-23022, Lawrence Berkeley National Laboratory, Berkeley, California.
Acknowledgements
The author would like to thank Sandia Technologies for collecting excellent pressure transient data, Susan Hovorka for leading the Frio Brine Pilot, and Larry Myer for coordinating LBNL's efforts. This work was part of the GEO-SEQ project supported by DOE's Office of Fossil Energy through the National Energy Technology Laboratory. Thanks are due to Charlie Byrer, Karen Cohen and Scott Klara for continued support of the GEO-SEQ program. This work was funded under DOE Contract No. DE-AC03-76SF00098.
