There are several critical challenge areas related to the development of effective CO2 storage Monitoring, Mitigation and Verification (MMV) programs according to the U.S. Department of Energy. The application of high definition borehole seismic techniques address the three check-marked areas (see diagram) related to geologic storage programs. Borehole seismic techniques are also being applied by MMV program managers to develop robust monitoring programs that address the following:
Vertical seismic profiling (VSP) requires that a well is situated in close proximity to the CO2 plume. Surface seismic sources are deployed around the well installation, with sensors deployed down-hole. Conventional VSP with sources close to the wellhead gives quite narrow subsurface coverage around the wellbore. Walkaway VSP where sources are arranged on a radial profile provides 2D subsurface coverage away from the well. Multiple walkaway profiles or sources deployed randomly at multiple azimuths and distances can provide 3D imaging and characterization of the subsurface around the well.
Compared to surface seismic, VSP data offers improved resolution and formation characterization around the well. As with surface seismic, multi-component recording can also be implemented with further potential information gains in terms of pressure/saturation discrimination and anisotropy characterization.
Multi-azimuth Multi-Component VSP data can be used to map fluid pressure changes and anisotropic effects around the well, both in the reservoir and in the overburden, the latter offering the potential for early warning of well leakage. VSPs are principally reflectivity based but analysis of direct arrivals can provide measurements of velocity and attenuation in the overburden, which can be used to improve analysis of surface seismic data. VSP data also offers the potential for providing early warning of migration from the well into the surrounding cap rock. The primary jobs that CO2StorageScan HD is designed to support are:
CO2 Reservoir Object Extraction
The following is an example of SR2020’s advanced Reservoir Object Workflow, interpretive service. After generating high-resolution, true-amplitude images and true-amplitude angle domain gathers we can proceed to extract/calibrate the seismic response and classify each subsurface image point represented by a small volume element according to alterations of the rocks by the presence of CO2. At this point we require full suites of well log measurements and rock physical relationships that describe the subsurface property change in dependence of adding CO2. These rock physical relationships can be obtained in the laboratory or sometimes through in-situ measurements. We then perform a simple pre-stack mapping and progress to depth migrated angle domain gathers as output by the optimized imaging.
The auxiliary information obtained from well log measurements and a–priori rock physics relations allows us to classify and predict the amplitude response in dependence of saturation and in dependence of pressure change. Clustering or other classification will allow us to then flag the subsurface property volumes as to the amount of change caused by the CO2 injection.
The following diagram shows the data flow through the envisioned workflow, where as a first step high-resolution structural images and reservoir objects are extracted from the 3D borehole seismic data and then progressively more auxiliary information is used to determine actual CO2 objects which could be extents, shapes, leakage paths. Part of the work flow relies on manual interpretation, but for reliable monitoring and verification of CO2 sequestration an automatic method is highly desirable.
Automatic CO2 Front Tracking
Having obtained high-resolution and true-amplitude image volumes of the CO2 sequestration we are able to reliably generate various data attributes that can be used to outline the CO2 volume extent as indicated through changes in the reflection and diffraction seismic response. Combinations of these attribute volumes allow us to then to apply automatic surface extraction algorithms as show by (Kadlec, 2008; Dorn 1989, 2002). This automatic extraction uses the entire 3D information and generates digital representations of those surfaces that are consistent with the image attributes. In a time-lapse scenario we will be able to extract the evolving surface geometry. If manual interpretations or modifications are necessary they can be done after generating the initial automatic representation. Using the baseline image volume and its CO2 volume representation, the subsequent analysis and extraction is aided by the previous temporal data set and its representations.
SR2020 has been actively involved in the application of bore-hole seismic imaging to CO2 plume monitoring for many years. Our CO2 plume monitoring project experience includes the following client projects:
Our CO2StorageScan HD service which is based on 3D VSP imaging offers distinctive advantages that make it a powerful technique for time-lapse CO2 plume monitoring. The high-frequency content of VSP data provides imaging detail with vertical and horizontal resolution that is superior to conventional surface seismic surveys. The high signal/noise quality of the data allows us to generate high definition images whose fidelity is sufficient to identify time-lapse effects with confidence and thereby monitor changes in the reservoir. Our previous CO2StorageScan HD project work has shown that the high resolution obtained from our multiple receiver VSPs can track fine and subtle changes due to the injection. These changes in the reservoir can be identified by the high resolution seismic changes in the waveforms (i.e., changes in amplitude) as well as travel time delays associated to velocity changes. There are four primary components to our CO2StorageScan HD service offering:
Time Lapse Modeling Study – We assume that as part of the initial assessment work for a sequestration project, numerical reservoir simulation work is typically done to predict the dynamic reservoir changes over time given various CO2 injection scenarios for the proposed storage area. Study results would provide useful predictions regarding how the target reservoir will respond to CO2 injection. Based on these predicted changes in reservoir rock/fluid properties over time we would be able to predict the seismic response using an advanced seismic modeling application to generate simulated 3D time-lapse images at various stages of CO2 injection. The modeling package SR2020 uses will allow us to modify a variety of rock physical, elastic and geological parameters to best match the predicted changes from the reservoir simulation model. The primary purpose of this study will be to answer questions regarding the effectiveness of time lapse seismic in detecting these reservoir changes at various volumes of injection.
Survey Design & Acquisition Program – the primary objective of this stage is to ensure that the program’s plume monitoring and leak detection objectives are achievable and to optimize the receiver array and seismic source location configurations. SR2020 will obtain various well, log and subsurface information from the client in order to complete a detailed modeling of the anticipated seismic program. The results of this program will be reviewed with the client and form the basis of the detailed technical survey design requirements. Pre-survey modeling with the existing geologic model will determine the source density and maximum offset required to illuminate the target. For example, given a 1600 m TVD of the objective it is expected that source offsets on the order of 1600 m would be sufficient to illuminate an area of 800 m around the well. For the high resolution nature of the survey we suggested using a dense source coverage of up to 1000 shots with spacing between 8 to 16m. Such a proposed time lapse 3D VSP program scenario would provide a very high definition, 3D volume for the baseline and each of the subsequent time lapse surveys. This full volume seismic image would not only allow for much more precise volumetric estimations but also ensure that any CO2 plume migration paths would be imaged no matter which azimuth they may develop along.
SR2020 typically deploys our long arrays of up to 160 levels as deep as possible so the deepest receiver is closest to the target zone. Our long array provides the best combination of image area and resolution available. SR2020’s tubing conveyed acquisition system provides the high-pressure coupling (1200 psi) required to ensure recording of the high-frequency and high signal to noise ratio data, including converted wave energy. Since we will be imaging reservoir scale details of the geology and dynamic reservoir conditions, small differences in the acquisition geometry could disturb the ray paths of one of the data sets that are sampling different geology from the previous survey.
Our patented acquisition system also ensures accurate depth repositioning of the geophones for the repeat survey. Accurately re-occupying depth positions with the geophones and surface positions for the source points ensures that the only differences in the time-lapse recorded wave fields are due to reservoir changes and not differences in acquisition parameters. In addition, having more receivers (80 or 160) in the array for all sources provides more travel time information at wider depth spans all of which allows us to get a higher resolution image.
Seismic Data Processing – The processing of the data for the two surveys typically includes traditional VSP analysis with component orientation, deconvolution and wavefield separation, and prestack depth migration. The use of a depth migration algorithm will require an accurate velocity model that represents the geology. The model built from the borehole seismic data will have enough resolution to track the changes in the CO2 injection. The resulting depth migrated images will then be used to analyze the reflectivity changes resulting from the injection. This analysis of the reflectivity will be conducted after the images have been calibrated to a common baseline reflector not affected by the injection (i.e. a seismic event at shallower depths).
Image Interpretation – In many cases we find that client’s are not familiar or confident interpreting high definition borehole seismic images and therefore engage SR2020 to assist them with interpretation of the acquired data. This can take the form of interpretive processing assistance and/or more detailed reservoir characterization work to generate various structural interpretations. In the case of high definition time lapse imagery, SR2020 is often retained to assist with the interpretation and generation of specific reservoir objects (eg. CO2 plumes) which define changes in the reservoir over time.