Post on 17-May-2020
Pressure-Anomaly Based Leakage Detection
JP Nicot, Mehdi Zeidouni,Alex Sun, and K.-Won Chang
Bureau of Economic GeologyJackson School of GeosciencesThe University of Texas at Austin
Progress Review of STAR Grant Researchon Carbon Geosequestration
Washington, D.C. – January 7, 2013
2
Bureau of Economic Geology
“Expert-Based Development of a Standard in CO2 Sequestration Monitoring Technology”
•STAR R834384 addresses the requirement for a GS project to develop a monitoring plan (UIC Class VI)
• focuses on 5 well-known fields:– Soil gas approaches– Groundwater monitoring and tracers– Pressure and temperature monitoring– Well-based techniques– Seismic techniques
•Above-Zone Monitoring Intervals (AZMI) measurements at Cranfield, MS and other fields
• In-Zone IZ monitoring
3
Bureau of Economic Geology
surface
Surface casing
Cemented in
Cement to isolate
Injection zone
AZMI Above Zone Monitoring IntervalTime
Pre
ssur
e
Injection zone
AMZI
Confining = No fluid communication
Using AZMI pressure to assess storage permanence
Courtesy Sue Hovorka
4
Bureau of Economic Geology
Pressure Monitoring
CO2 Injection Zone
Above-Zone Monitoring Interval (AZMI) –leakage detection
Within Injection Zone (IZ) reservoir management
Daily injection rate2000
1000
30 m
Metric ton/day
300
310
AZMI
bars
300
350
400
IZ
bars
Tao, Bryant, Meckel, IJGGC accepted ms, available online
5
Bureau of Economic Geology
Pressure monitoring subtask
•Presenting results from on-going modeling work:
•Set the stage: how is pressure attenuated, how is it transmitted from IZ to AZMI? (KWC)
•How big is the pressure signal, can we detect it? (MZ)
•Can we locate the leak if there is one? (AS, MZ)
•Optimize location of monitoring wells (AS)
6
Bureau of Economic Geology
Role of the ambient rock during CO2injection
– Pressure propagation is governed by ratios of mudrock/sandstone permeability and storativity and relative thickness of formations
– Permeable and compressible surrounding rock reduces pressure propagation within a reservoir
Decrease mudrock permeability
Increase mudrock storativity
Mudrock = ambient rock
sandstone = reservoir
From Chang, Hesse, Nicotin review WRR, 2013
7
Bureau of Economic Geology
Pressure diffusion: pumping / CO2 injection
From Chang, Hesse, Nicotin review WRR, 2013
8
Bureau of Economic Geology
Development of a leaky well analytical model: Pressure signal magnitude
hm
2rl
hs
hl
AZMI
InjectionZone
kl2rw
r
R
q
r’ r
L
Injection Well Leaky WellAbandoned Well
Observation Well
Zeidouni, JPSE in review
9
Bureau of Economic Geology
Analytical solution (single-phaseflow)
0
32
1
00
11
( ) 1( )
D
lD
lDDlD
lD lDD lD lD
D D
K sR
s K sq
sK rK sr
r sK sr s sT r K r
0
1
( )lD DD
mD
D lD lDD D
sq s KP
s sT r K r
Zeidouni, JPSE in review
10
Bureau of Economic Geology
Asymptotic solution
2 21 1 ln ln
1 2 2 4 2lDD D
mDD D D D
qRPT t T
21 1 ln1 1/lD D D
D
q C t RT
2 22 3
2 3 4
2 (3)1 6 22 ln 4 2 ln 4 2 ln 4 2 ln 4
DD D D D
C tt t t t
11
2
2 ln1
DTD D
D lD
TT r
Zeidouni, JPSE in review
11
Bureau of Economic Geology
Asymptotic solution
2 21 1 ln ln
1 2 2 4 2lDD D
mDD D D D
qRPT t T
21 1 ln1 1/lD D D
D
q C t RT
2 22 3
2 3 4
2 (3)1 6 22 ln 4 2 ln 4 2 ln 4 2 ln 4
DD D D D
C tt t t t
11
2
2 ln1
DTD D
D lD
TT r
Zeidouni, JPSE in review
12
Bureau of Economic Geology
Detection type-curves
13
Bureau of Economic Geology
Detection type-curves
14
Bureau of Economic Geology
Cranfield Example
• D=0.5 (diffusivity ratio), TD=0.28 (transmissivity ratio)• t = 1 yr equivalent to tD=1010 (dimless time)
• P = 0.15 psi ~to PD = 0.0005 (dim pressure threshold)
• = 6 x 10-6 (leakage coefficient)
• and assuming RD = D = 2500
15
Bureau of Economic Geology
Detection type-curves
> 6 x 10-6 is detectable
kl > 400 mD is detectable
16
Bureau of Economic Geology
Detection type-curves
RD and D < 2500 are detectable
R and < 250 m are detectable
17
Bureau of Economic Geology
Similar work on faults:Leaky fault analytical model
Zeidouni, WRR., 2012
18
Bureau of Economic Geology
Similar work on faults:Leaky fault analytical model
Zeidouni, WRR., 2012
Multilayer leakage attenuation
Validation ofthe solution
19
Bureau of Economic Geology
Inversion of pressure anomaly data:“remediation”
•We observed “pressure anomaly”: can we detect rate and location of source?
Simple homogeneous system
Given observation heads, ordinary least-square method can recover leakage rate history
Sun and Nicot, Adv.W.R., 2012
20
Bureau of Economic Geology
Inversion of pressure anomaly data:“remediation”
•We observed “pressure anomaly”: can we detect rate and location of source?
Simple homogeneous system
Given observation heads, ordinary least-square method can recover leakage rate history
Sun and Nicot, Adv.W.R., 2012
21
Bureau of Economic Geology
Inversion of pressure anomaly data:“remediation”
•We observed “pressure anomaly”: can we detect rate and location of source?
Simple homogeneous system
Effect of observation errors
OLS TSVD
Sun and Nicot, Adv.W.R., 2012
22
Bureau of Economic Geology
Inversion of pressureanomaly data: “remediation”
Multiple leaky wells
Sun and Nicot, Adv.W.R., 2012
23
Bureau of Economic Geology
Sun and Nicot, Adv.W.R., 2012
Inversion of pressureanomaly data: “remediation”
Heterogeneous; multiple leaky wells; locations known but rates unknown (could be 0)
Now, heterogeneous K field
24
Bureau of Economic Geology
Inversion of pressure anomaly data:“remediation”
ONE leaky well but location not known: simulated annealing
O: Well locations
X: Initial guess
: Actual location
Conclusions:Can detect leakage rates and locations Try different solvers
Sun and Nicot, 2012, Adv. WR
25
Bureau of Economic Geology
Leakage source detection:design perspective
Sun, Zeidouni, Nicot et al., accepted ms, Adv.W.R.
0.1 psi = 0.07 m
Earlier signalsLonger tails
Effect of AZMI lnKvariance on SNR
26
Bureau of Economic Geology
Leakage source detection:design perspective
An observation point placed closer to potentially leaky
locations may not necessarily always
yield the most reliable signals when aquifer
properties are uncertain.
Sun, Zeidouni, Nicot et al., accepted ms, Adv.W.R.
0.1 psi = 0.07 m
Earlier signalsLonger tails
Effect of AZMI lnKvariance on SNR
27
Bureau of Economic Geology
Pressure and temperature
• AZMI temperature changes because:• (1) Leak temperature; (2) Pressure changes (JT cooling); (3)
Dissolution and vaporization (exothermic and endothermic processes)
• CMG-GEM simulations
Zeidouni et al., SPE, ms in progress
28
Bureau of Economic Geology
Above-zone temperature pulse
Zeidouni et al., SPE, ms in progressTemperature detection limit = 0.15°C; Pressure detection limit = 0.15 psi).
29
Bureau of Economic Geology
• Chang, K.-W. M. A. Hesse, and J.-P. Nicot, 2013, Reduction of lateral pressure propagation due to dissipation into ambient mudrocks during geological carbon dioxide storage, in review WRR
• Sun, A. Y., and Nicot, J. -P., 2012, Inversion of pressure anomaly data for detecting leakage at geologic carbon sequestration sites: Advances in Water Resources, v. 44, p. 20‒29.
• Sun, A.Y., Zeidouni, M., Nicot, J.-P., Lu, Z., and Zhang, D., Assessing leakage detectability at geologic CO2 sequestration sites using the probabilistic collocation method : Advances in Water Resources, accepted
• Sun, A.Y. and others, Optimal CO2 leakage monitoring network design under uncertainty, in preparation
• Zeidouni, M., 2012, Analytical model of leakage through fault to overlying formations, Water Resour. Res., 48, W00N02, doi:10.1029/2012WR012582.
• Zeidouni M. and Pooladi-Darvish M., Inter-aquifer flow through leakage conduits: analytical solutions for leakage rate and pressure change evaluation, in review for Journal of Petroleum Science and Engineering.
• Zeidouni M., Hosseini S.A., and Nicot J.-P., Above-zone temperature variations due to CO2 leakage from the storage aquifer, in preparation for SPE 2013
• Zeidouni M., Hosseini S.A., and Nicot J.-P., Leakage identification through above-zone pressure monitoring, in preparation