Where is a Geophysicists Best Source of Seismic? Downhole.

Last week’s trip down memory lane, looking at some of the history of downhole source development, seemed to spark a lot of interest. This technology does that, it seems like such a great idea, why wouldn’t we try it? 400Hz seismic, high resolution velocity variations, 4D, what’s not to like! Well, this week I will look at examples of how this technology has been used, where and why and the types of answers it has provided, and also areas where it could be used more! Sure, this is niche, but I would argue for lots of niches in several domains, and for many applications. Consequently, it is a technology that we all need to be aware of for when that niche problem, which we are all increasingly faced with, arises for us.

It seems to me there are four broad types of borehole seismic surveys where a downhole source is used.

• Crosswell seismic tomography

• Reverse VSP

• Single well imaging

• Microseismic velocity model calibration

These surveys can be utilized across the full range of domains we may operate in,

• CO2 sequestration

• EOR (CO2, steam)

• Thermal (SAGD)

• Engineering

• Mining

• Structural

• Geothermal

The reverse VSP is the least used survey type using a downhole source, which is not surprising given that it does nothing more (or less) than a regular VSP could achieve given source-receiver reciprocity. One of the only situations where is might be useful would be if there were restrictions on using a seismic source on the surface, perhaps in an urban environment. 

Downhole source applications

For the purposes of this discussion I will define a Single Well Imaging (SWI) system as one which has an active source which is separated from the receivers in the same well by wireline or fiber. It therefore excludes more conventional acoustic logging tools oriented to acquire near wellbore reflection data and receivers deployed to ‘listen’ for elevated wellbore noise. There was a very interesting LinkedIn post recently about just this (https://www.linkedin.com/posts/rj-wang_technical-program-utc-5-activity-6726659016224845824-cCAB), discussing using a hydrophone array for leak detection. This reminded me of the wonderfully named Seismic Hydrophone Array Tool which was used in conjunction with a surface source some years ago for detecting wellbore fractures. It was around the time that it was popular to include the word ‘Imager’ into tool names, good job they didn’t with the SHAT tool, that wouldn’t have been good.

Of course, with the increasing adoption of DAS, one wonders how relevant the concept of a single well imager is. Surely a downhole source run on a hybrid fiber heptacable would provide the same solution? Or using an existing permanent fiber in the well? Not quite, at least not yet. The major advantage of geophones in this scenario is that they are clamped to the wellbore to significantly improve coupling and they are three component sensors (even perhaps dual or quad sensors for each component in some instances) as opposed to the fiber which responds to inline strain only (or a component of it).

An important application of SWI is for salt flank imaging, particularly offshore. Li et al (2016) describe how offset VSPs are used for imaging given different source-receiver configurations. Resulting errors are large and dependent on well constrained velocity models and accurate ray tracing due to long and complex ray paths from the surface.

Borehole seismic survey geometry of a) Salt Proximity, b) Sediment Proximity, and c) Reflection Salt Proximity (RSP) with seismic rays. (Li et al, 2016)

An alternative would be a SWI deployed where the ray paths would be significantly shorter and less complex. Subsequent errors in salt boundary distances would be greatly reduced.

In an early example of such a survey, Yu et al (2001) showed the need for a well coupled, powerful and repeatable source to ensure sufficient energy is recorded at the receivers.

Other applications of such a system would include high resolution (sub surface seismic) imaging for cap rock integrity studies. The SWI could be deployed in the lateral (pumped or tractored) and reflections from the upper cap rock would image the continuity of the interface. Whilst this type of application currently exists using traditional sonic logging arrays those typically do not provide sufficient energy to give the required resolution in terms of frequency and signal to noise.

Single well imager deployed for salt flank imaging

Single well imager deployed for cap rock integrity

The main advantages of the SWI system are reduced and therefore less complex ray paths and the resultant improvement in resolution. The last application to touch on here is in mapping hydraulic fractures. Obviously, this application has been dominated in the last fifteen years or so by the passive mapping of microseismic events, however mapping reflectivity and velocity changes induced by the hydraulic fracture is also possible and provides another measure of the hydraulically stimulated volume. An example or using a SWI approach to this is given by Daley et al (2003). Using a Piezoelectric source and hydrophone array at the Lost Hills test site in California, they were able to map the reflectivity associated with a gas filled hydraulic fracture.

Single well imager (left) and common offset gathers before and after predictive deconvolution (right), Daley et al (2003).

Aside from SWI, downhole sources have other applications in hydraulic fracture mapping. Locating microseismic events requires a velocity model to accurately predict travel times, this is built initially from dipole sonic logs (including anisotropy). It is (or should be) QC’d, and calibrated as required, using perforation shots or the like (including sliding sleeves, ball drops and string shots in some instances). The limitation of these data are that they may be too noisy to be useful or in many cases not ‘loud’ enough to be seen above the noise floor. Using a downhole seismic source is an attractive alternative.

Hogarth et al (2017) discuss experiments comparing velocity model calibration results using perforations and downhole seismic sources. They show that there are obvious benefits of the latter such as knowing time zero and the ability to position the source anywhere in the treatment well, for example above the shallowest perforation depth.

Results from a vertical monitor well (MW) and treatment well (TW) shows improvement in modeled perforation shot locations using a model calibrated with crosswell shots (green) over locations based on a model with no calibration (purple). Hogarth et al (2017)

Of course, the obvious operational constraint to this approach as described, is convincing the completions engineer to allow you to run an active seismic source into the well prior to completion! Perhaps the more practical approach would be to utilize, if possible, adjacent offset wells straddling the treatment well.

 Traditional crosswell seismic tomography can provide another novel application for a downhole source in microseismic monitoring. The process of hydraulic fracturing a well, results, it is hoped, in creating an open fracture network around, and away from, the treatment well. The impact of this network is to change the elastic properties of the rock mass, causing a reduction in both compressional and shear velocities. This change can be measured as a time lapse effect using crosswell tomography. Zorn (2016) describes an example where a before and after crosswell seismic experiment was carried out in the Marcellus shale. The vertical source and receiver wells straddled the completion well but unfortunately did not extend as deep. The reduction in shear and compressional velocities show interesting correlations with the microseismic events and give a unique insight into the geomechanics of the completion.

Crosswell tomographic P-wave (left) and S-wave (right) velocity difference section. Microseismic events generally do not cross into the lower P-wave or S-wave velocity anomaly. The 0.8 % decrease in P-wave velocity (left) straddles the proposed location of a low offset fault controlling microseismicity.  The 3 % decrease in S-wave velocity coincides with the proposed location of the same fault. (Zorn, 2016).

By far the most common use of downhole seismic sources is for crosswell seismic tomography which provides a high resolution (about ten times that of the equivalent surface seismic) 2D seismic or velocity image between two co-planar wells. This has been applied in many environments over the years.

An excellent example of monitoring EOR steam injection is given by Bair et al (1999). Cyclical steam injection in the San Joaquin Valley of California is mapped over a period of a year. The resulting decrease in P-wave velocity is evident and clearly indicates the steam propagation is inhibited by vertical fractures. It is worth noting that the impact of temperature (steam) on P-wave velocity is sufficiently larger (up to 25% reduction) to be able to map in static images such as those shown. Difference images could also have been produced.

Velocity tomogram snapshots taken over a year of steam injection (Bair et al 1999)

EOR using CO2 injection has also been successfully mapped. Lazaratos et al (1997) show an example from the Grayburg formation in the McElroy field in West Texas. The image below shows the change in seismic velocity after nine months of CO2 injection. The operator was hoping for a uniform CO2 sweep from left to right, however and open vertical fracture allowed migration of CO2 into a zone above the reservoir.

Time lapse P-wave velocity tomogram showing difference in velocity after nine months of CO2 injection (Lazaratos et al 1997)

A great example of the resolving power of crosswell seismic is given by Syarif et al (2013). The Bunyu field in Indonesia is characterized by poor surface seismic due to the presence of near surface coals and complex structure. A crosswell program was undertaken to provide enhanced reservoir characterization and to develop new infill drilling locations. The image below shows the vastly improved resolution of the surface seismic, even after it has been reprocessed using velocity information derived from the crosswell.

Comparison of surface seismic (2005), re-processed seismic (2010) & Crosswell Seismic profile XR2-X17 (2011).

Another environment where the improved resolving power of crosswell has provided significant value is thermal plays such as Canadian Steam Assisted Gravity Drainage (SAGD). Successful SAGD wells require uninhibited steam propagation between wells, mudstones and clay inhibit this but are beneath the resolving power of surface seismic in the area. Zhang et al (2005) describe a project at Christina Lake in Alberta, Canada, where the improved resolution of the crosswell highlights potential barriers to steam chamber development.

Crosswell Seismic and 3D Seismic Acquired in 2004. Crosswell Up-Going P Wave Reflection Image between Well A and C (Left) and Surface 3D Seismic for The Same Location (Right). (Zhang et al, 2005)

Engineering applications of crosswell include near surface evaluation for potential solution mining features for large scale engineering projects such as Boone et al (2008). They describe a project as part of the subsurface evaluation for the Canadian end of the Detroit River International Crossing Bridge.

Interpreted crosswell reflection profile between wells showing salt collapse feature edges (blue), associated faulting (green) and shallow salt layers (orange). Boone et al (2008).

Carbon capture sequestration (CCS) is an area where crosswell seismic has not been extensively used. A case study from the Encana Weyburn project in Canada (Guoping et al, 2002) describes a project where the source and receiver wells were horizontal. The resulting velocity tomogram, a horizontal depth slice through the two wells, shows low velocity areas consistent with surface seismic impedance variations at the same depth.

Integration of velocity tomogram inverted from crosswell direct P-wave traveltimes (left)

with impedance map from surface 3-D seismic inversion (right). The highlighted area denotes the

crosswell coverage.

So, there you have it, a few examples of how using a downhole source can add a unique dimension to our geophysical arsenal. I am convinced that this technology has an important role to play in the future of borehole seismic. More on that next week, but for now, happy Halloween!

References

J. F. Bair, S. J. Johnson, D. R. Julander, R. T. Langan, J. S. Meyer, and J. K. Washbourne, (1999), "Time‐lapse imaging of steam and heat movement in the Cymric 36W Cyclic Steam Pilot using crosswell seismology," SEG Technical Program Expanded Abstracts: 1643-1646.

S. Boone, M. Monier‐Williams, R. Turperning, T. Morgan, K. Tandon, and B. Bryans, (2008), "Identification and interpretation of solution mining features in bedded salt deposits on a crosswell reflection profile," SEG Technical Program Expanded Abstracts: 1407-1410.

Daley, Thomas & Gritto, Roland & Majer, E.. (2003). Single well seismic imaging of a gas-filled hydrofracture. 10.2172/821042.

Syarif, A., Irwanzah, Z., Handayani, T., & Dogra, S. (2013). Crosswell Seismic Guided 3D Seismic Interpretation Results in Successful Infill Well Location in Bunyu Field (Indonesia). Society of Petroleum Engineers. doi:10.2118/166490-MS

Leah J. Hogarth, Conrad M. Kolb, and Joël H. Le Calvez, (2017), "Controlled-source velocity calibration for real-time downhole microseismic monitoring," The Leading Edge 36: 172–178.

Li et al, 2016. Reflection Salt Proximity. First break volume 34, October 2016

Spyros K. Lazaratos and Bruce P. Marion (1997). Crosswell seismic imaging of reservoir changes caused by CO2 injection. The Leading Edge, 16(9), 1300-1308. 

Downhole orbital vibrator source and single well imaging Gang Yu, Larry Walter, Bill Chmela, Leif Jahren, John O'Brien, and Justin C. Lime SEG Technical Program Expanded Abstracts 2001. January 2001, 396-399 

Zorn, Erich (2016) Integrated analysis and interpretation of microseismic monitoring of hydraulic fracturing in the Marcellus shale. Doctoral Dissertation, University of Pittsburgh.

Weimin Zhang, Sung Youn and Quang Doan, EnCana Corp. Understanding Reservoir Architectures and Steam Chamber Growth at Christina Lake, Alberta by Using 4D Seismic & Crosswell Seismic Imaging. SPE/PS-CIM/CHOA 97808PS2005-XXX

Guoping Li and Ernest Majer. High-Resolution Crosswell Seismic Imaging Between Horizontal Wells. CSEG Recorder Nov. 2002, Vol. 27 No. 09