We recommend that you upgrade your browser. The following is a list of popular options:
For the last year, I've been working on data from two seismic experiments acquired as part of the Frio CO2 sequestration pilot project managed by TBEG [the project was run by Susan Hovorka]. The Frio Project, initiated to help understand the behavior of supercritical CO2 within a saline aquifer, consisted of two CO2 injections, one in Oct. of 2004 (Frio I) and a second in Sep. of 2006 (Frio II). Both Frio I and Frio II had crosswell monitoring components (acquired by Tom Daley, LBNL - see right) designed to characterize the dynamic seismic signature of CO2. The two experiments were on the opposite extremes of time/space survey design and provide two very different views of fluid movement (see below). I spent a fun (if tiring) week in the field (Dayton TX) during the Frio II injection.
The Frio I seismic experiment was a traditional time-lapse crosswell survey acquired with an orbital vibrator source (deployed on wireline) and a large multi-level 3C VSP array. Measurements were made before and after the CO2 injection yielding the prototypical timelapse dataset with fine spatial resolution and coarse temporal resolution (only 2 time samples!). I've been working on applying various types of novel tomographic algorithms to this dataset, including differential compactness constraints, temporal regularization operators, and waveform inversion. To the left is one of our time-lapse P-wave tomograms generated by applying compactness constraints to differenced traveltimes.
For more information on the Frio I experiment, check out the summary paper, "Measuring permanence of CO2 storage in saline formations : the Frio experiment" S.D. Hovorka, S.M. Benson, C. Doughty, B.M. Freifeld, S. Sakurai, T.M. Daley, Y.K. Kharaka, M.H. Holtz, R.C. Trautz, S.H. Nance, L.R. Myer and K.G. Knauss Environmental Geosciences, Vol.13, No.2, 2006, p.1-17
The Frio II seismic experiment was a non-traditional time-lapse experiment; a single permanently installed piezo source and a fixed 24 channel hydrophone array were used to continously monitor fluid movement. In contrast to Frio I, the spatial sampling was too coarse to allow traditional imaging (just one shot location) but the temporal resolution was excellent with data acquired every 2 minutes for large portions of the injection process. For this dataset I've been working on characterizing temporal variations in waveform character using some fun signal processing approaches. I'm also investigating techniques for extracting spatial information on fluid movement from this minimal array - as you might guess, this requires some very strong assumptions about geometry but can still yield some interesting constraints on plume movement. To the left is a very confusing plot which shows the temporal response of each hydrophone as a color image with axis of seismic time (x) and injection time (y). More interesting stuff soon!
As part of my graduate research, I acquired a significant volume of crosswell seismic, crosswell radar, continuous core, and logging data at the Pinellas DOE site located near Tampa, Florida, in concert with LBNL. The target of these investigations were several regions of DNAPL contamination (dense non-aqueous phase liquids), in particular zones of with high levels of the chlorinated solvent TCE. In total, 26 short offset crosswell seismic profiles, 11 crosswell radar profiles, several seismic VSP surveys, and a suite of well logs (gamma, conductivity) were obtained. I am in the process of developing a dedicated page with information on the Pinellas site and selected datasets available for download.