Monitoring of CO 2 injected at Ketzin using 3D time-lapse seismic data Alexandra Ivanova 1,2, Ursula Lengler 1, Stefan Lüth 1 and Christopher Juhlin 2 (1)GFZ German Research Centre for Geosciences, Germany (2) Uppsala University, Sweden EGU , Vienna, Austria
injection well Ktzi 201 observation well Ktzi 200 observation well Ktzi 202 injection facility CO 2 Pilot Site
Geology of Ketzin Structural model V 0 Caprock V 1 Aquifer V 0 Auitard V2V2
Geology of Ketzin Structural model V 0 Caprock V 1 Aquifer V 0 Auitard V2V2
3D Repeat Survey 3D Baseline 2005: 41 Templates ~14km 2 3D Repeat 2009: 20 Templates (after ~20 kilotons CO 2 injected) 1 Template: 5 lines, 240 geophones 12 lines, 180 shotpoints Start of acquisition 2009: 25 Sep 2009 Last shotpoint: 6 Nov 2009 Ketzin
Repetability 2005 baseline 2009 repeat
Step: Original parameters: Read raw SEGD data Vertical diversity stack Bulk static shift to compensate for source delay: 6 ms Extract and apply geometry Trace edit and polarity reversal Pick first breaks: offset range 300–500 m Remove 50-Hz noise on selected receiver locations Spherical divergence correction: v 2 t Band-pass filter: Butterworth 7–14–150–250 Hz Surface consistent deconvolution: filter 120 ms, gap 16 ms, white noise 0.1% Ground roll mute Spectral equalization 20–40–90–120 Hz Band-pass filter: 0–300 ms:15–30–85–125 Hz 350–570 ms:14–28–80–120 Hz 620–1000 ms: 2–25–70–105 Hz Zero-phase filter Refraction statics: datum 30 m, replacement velocity 1800 m /s, v m/s Trace balance using data window Velocity analysis: every 20th CDP in the inline and crossline direction Residual statics Normal moveout correction: 50% stretch mute Stack Trace balance: 0–1000 ms FX-Decon: Inline and crossline directions Trace balance: 0–1000 ms Migration: 3D FD using smoothed stacking velocities Depth conversion: using smoothed stacking velocities New parameters: Bandpass filter: Butterworth Hz Spectral equalization Hz Band-pass filter: ms: Hz ms: Hz ms: Hz Processing steps applied to 3D repeat compared with the original flow
Amplitude change anomaly over the CO 2 injection site low fold area for anomaly 1.max change in amplitude ~ amplitude of K2 reflection 2.anomaly ~ centered on injection well 3.positive polarity
VOLUMETRICS Volumetric estimation of mass of the injected CO 2 visible in seismic data will be performed using - average values of CO 2 saturation in reservoir obtained for every well of the Ketzin site with a) Pulsed Neutron-Gamma Well Logging; b) Multiphase Flow Modeling; - average CO 2 density calculated with a) measured temperature and pressure in reservoir at the injection site (in the injection well) to the time of the repeat survey; b) from Multiphase Flow Modeling for every well of the Ketzin site. Results show that we can see approximately the same quantity of the injected CO 2 in the time-lapse seismic data : a) Pulsed Neutron-Gamma Well Logging (~ 94%); b) Multiphase Flow Modeling (~91%).
Parameters for volumetric estimation PARAMETERSOURCE Time delayVelocity push-down P-Velocities/ CO 2 and brine saturation Petrophysics on samples from reservoir saturated with CO 2 and brine CO 2 saturation in reservoir Seismic time-lapse interpretation and: a)PNG Logging in 3 wells b)Multiphase Flow Modelling CO 2 density in reservoir a)Temperature and pressure in reservoir on the injection site b)Multiphase Flow Modelling PorosityCore analyses and logging
Time Difference for CDP bin Push-Down Effect: dt = 2h * (1/ v 2 – 1/ v 1 ) => thickness h of CO 2 plume v2v2 v1v1
Parameters for CO 2 mass estimation per CDP bin
Distribution of CO 2 from 3D time-lapse observaons: 20.5 kTons23 kTons Mass of injected CO 2 : kTons 20 kTons22.1 kTons 20.5 kTons 23 kTons Min Max.... a) PNG Well Logging Mass of injected CO 2 : Repeat Survey Beginning : 22.1 kTons Repeat Survey End: 24.2 kTons b) Multiphase Flow Modelling Min Max kTons22.1 kTons 23 kTons20.5 kTons 20 kTons Min
1.Results of seismic interpretation and petrophysical data allowed a satisfactory CO 2 mass estimation (91%-94% mass of injected CO 2 to the time of the repeat survey) using values of CO 2 saturation derived by a) Pulsed Neutron-Gamma Logging; b)Multiphase-Flow Modeling; and average CO 2 density calculated with a)measured temperature and pressure in reservoir at the injected site (in the injected well) to the time of the repeat survey; b)Multiphase Flow Modeling for every well of the Ketzin site. 2. Consequences for seismic data of CO 2 effect on seismic velocities in reservoir rock (petrophysics) will be illustrated on synthetic seismic models. 3. Predicted distributions will be used to generate synthetic seismic data that will be compared directly to seismic field data. Summary and Outlook
Volumetric estimation based on velocity push-down Structural model V 0 Caprock V 1 Aquifer V 0 Auitard V2V2
Parameter Value Receiver line spacing / number 96 m / 5 Receiver station spacing / channels 24 m /48 Source line spacing / number 48 m / 12 Source point spacing 24 m or 72 m CDP bin size 12 m x 12 m Nominal fold 25 Geophones 28 Hz single Sampling rate 1 ms Record length 3 s Source 240 kg accelerated weight drop, 8 hits per source point 3D Acquisition Parameters
Seismic Interpretation STACK MIGRATION
Volumetric estimation based on velocity push-down Structural model V 0 Caprock V 1 Aquifer V 0 Aquitard Top aquifer Bottom aquifer Bottom aquitard
Volumetric estimation based on velocity push-down Structural model V 0 Caprock V 1 Aquifer V 0 Aquitard Reflectors in time migrated image Top aquifer Bottom aquifer Bottom aquitard ΔTΔT V2V2
CO 2 Storage at Ketzin Start of injection: 30th June, 2008 Food grade CO 2 (99.9%) ~ 47,000 tons Saline aquifer (Stuttgart Formation) ~ 620 m – 650 m Juni 2008