3D VSP - a mine of information for mining exploration · 3D VSP - a mine of information for mining...
6
3D VSP - a mine of information for mining exploration Michel DENIS Charles NAVILLE Jean Claude LECOMTE Laurence NICOLETIS CGGVERITAS IFPEN IFPEN IFPEN Massy, FRANCE Rueil, FRANCE Rueil, FRANCE Rueil, FRANCE [email protected] [email protected] [email protected] [email protected] Eric SUAUDEAU Quartus SNYMAN CGGVERITAS Anglo American Massy, FRANCE Johannesburg, RSA [email protected] [email protected] INTRODUCTION Many papers have been written on the processing and acquisition of 3D-VSPs but very few tackle the issue of the interpretation and even less address joint interpretation between surface and borehole data, in order to gain a high confidence in the structural model. The case presented here is one of the first joint surface borehole seismic acquisition. The main driver behind this project was to de-risk a future shaft location by acquiring a high resolution data set to image potential small fractures or shear zones at the bottom of the shaft which could drastically impact the scheduled CAPEX of the shaft due to complication in the shaft bottom infrastructure cementation, shaft pillars set-up, or mechanical instability of the walls. The VSP acquisition and processing has been addressed in Pretorius et al. (2011) and Humphries et al. (2011), but detailed and integrated interpretation of both surface and borehole seismic was never published. METHOD AND RESULTS The survey design was based on the target depth (also the scheduled shaft depth) of 642m and borehole depth of 740m. The target is the platinum reef with different units used for structural interpretation UG1 and UG2. As the overburden except a low velocity layer below surface is nearly a constant velocity media, the geometry of the VSP is straightforward using basic geometrical rules. To image a 200m radius circle at the target level of 645m, a minimum acquisition circle of radius 850m would be enough. In practice we have used a 1200m radius circle filled with an even sampling of receiver points and source points on a 40m regular grid. Binning in either 20m or 5m bins was enabled. Two boreholes were available, one with a 40 level dual hydrophone/geophone tool and the other borehole rig-up with a 12 level slim VSP tool, both tools with 10m depth interval. We will focus on the latest tool and the very interesting results of the hydrophone tool will be published in a future article. Due to the limited number of levels with the slim tool, it was positioned at three different depth ranges ( figure 1 and for each of these positions, about 2900 source positions were acquired, with a high frequency sweep of 30Hz to 250 Hz. . Figure 1: Different VSP tool positions in BH1 borehole and main logs acquired. SUMMARY The 3DVSP technique becomes more popular with the emergence of multilevel 3 component borehole tools. However, the value of the information derived from VSP is not always well understood. In this paper we will present a case history of a joint surface and borehole seismic acquisition, with the goal to de-risk a shaft sinking location on a platinum mine, with a target depth about 650m. On surface a dense grid of receivers and vibrator source points were laid out in a 1.2 km radius circle centred on the well head. A 12 level 3C-VSP digital tool, 110m long, was lowered in the small NQ diameter borehole, in three successive depth positions. The surface 3D cube was processed and interpreted independently from the 3D VSP data. On a near target reflector, the surface data structural interpretation showed mainly a clear E/W fault, and additional sub-seismic lineaments of differing azimuths, difficult to identify in terms of fault. The 3D VSP image limited to a short radius around the borehole confirmed the fault/dyke nature of one such lineaments, separating monocline compartments. As a consequence, the surface data was carefully re- interpreted. On the new set of structural images derived from two surface seismic reflectors and the near surface fault footprint from 3DVSP residual statics, a series of subtle faults were clearly assessed. Last, the few faults intersecting the borehole can be clearly recognized on the cores, the standard logs and the borehole radar images. This case study demonstrates the added value of a joint interpretation of surface and borehole data in a decision making process for shaft sinking. Key words: 3D-VSP, interpretation, fault lineaments, 23 rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 1
Transcript of 3D VSP - a mine of information for mining exploration · 3D VSP - a mine of information for mining...
3D VSP - a mine of information for mining exploration Michel DENIS Charles NAVILLE Jean Claude LECOMTE Laurence NICOLETIS CGGVERITAS IFPEN IFPEN IFPEN Massy, FRANCE Rueil, FRANCE Rueil, FRANCE Rueil, FRANCE [email protected] [email protected] [email protected] [email protected] Eric SUAUDEAU Quartus SNYMAN CGGVERITAS Anglo American Massy, FRANCE Johannesburg, RSA [email protected] [email protected]
INTRODUCTION
Many papers have been written on the processing and acquisition of 3D-VSPs but very few tackle the issue of the interpretation and even less address joint interpretation between surface and borehole data, in order to gain a high confidence in the structural model. The case presented here is one of the first joint surface borehole seismic acquisition. The main driver behind this project was to de-risk a future shaft location by acquiring a high resolution data set to image
potential small fractures or shear zones at the bottom of the shaft which could drastically impact the scheduled CAPEX of the shaft due to complication in the shaft bottom infrastructure cementation, shaft pillars set-up, or mechanical instability of the walls. The VSP acquisition and processing has been addressed in Pretorius et al. (2011) and Humphries et al. (2011), but detailed and integrated interpretation of both surface and borehole seismic was never published.
METHOD AND RESULTS
The survey design was based on the target depth (also the scheduled shaft depth) of 642m and borehole depth of 740m. The target is the platinum reef with different units used for structural interpretation UG1 and UG2. As the overburden except a low velocity layer below surface is nearly a constant velocity media, the geometry of the VSP is straightforward using basic geometrical rules. To image a 200m radius circle at the target level of 645m, a minimum acquisition circle of radius 850m would be enough. In practice we have used a 1200m radius circle filled with an even sampling of receiver points and source points on a 40m regular grid. Binning in either 20m or 5m bins was enabled. Two boreholes were available, one with a 40 level dual hydrophone/geophone tool and the other borehole rig-up with a 12 level slim VSP tool, both tools with 10m depth interval. We will focus on the latest tool and the very interesting results of the hydrophone tool will be published in a future article. Due to the limited number of levels with the slim tool, it was positioned at three different depth ranges ( figure 1 and for each of these positions, about 2900 source positions were acquired, with a high frequency sweep of 30Hz to 250 Hz. .
Figure 1: Different VSP tool positions in BH1 borehole and main logs acquired.
SUMMARY The 3DVSP technique becomes more popular with the emergence of multilevel 3 component borehole tools. However, the value of the information derived from VSP is not always well understood. In this paper we will present a case history of a joint surface and borehole seismic acquisition, with the goal to de-risk a shaft sinking location on a platinum mine, with a target depth about 650m. On surface a dense grid of receivers and vibrator source points were laid out in a 1.2 km radius circle centred on the well head. A 12 level 3C-VSP digital tool, 110m long, was lowered in the small NQ diameter borehole, in three successive depth positions. The surface 3D cube was processed and interpreted independently from the 3D VSP data. On a near target reflector, the surface data structural interpretation showed mainly a clear E/W fault, and additional sub-seismic lineaments of differing azimuths, difficult to identify in terms of fault. The 3D VSP image limited to a short radius around the borehole confirmed the fault/dyke nature of one such lineaments, separating monocline compartments. As a consequence, the surface data was carefully re-interpreted. On the new set of structural images derived from two surface seismic reflectors and the near surface fault footprint from 3DVSP residual statics, a series of subtle faults were clearly assessed. Last, the few faults intersecting the borehole can be clearly recognized on the cores, the standard logs and the borehole radar images. This case study demonstrates the added value of a joint interpretation of surface and borehole data in a decision making process for shaft sinking. Key words: 3D-VSP, interpretation, fault lineaments,
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 1
3VSP – a mine of information for mining exploration denis, naville, lecomte, nicoletis, suaudeau, snyman
The surface 3D data set was first processed and interpreted in a straightforward and timely manner. The 3D-VSP was processed by different contractors and finally by IFPEN as part of a R&D project on VSP pre-processing (old ideas in a new packaging ). The first interpretation of surface data showed an expected EW fault but no other major faults or shear zones were detected with confidence. See Figure 2 for an example of such results and interpretation.
Figure 2: NS section of the surface 3D and dip-map of interpreted UG1 horizon. The EW fault is obvious. The reflection 3D-VSP results from IFPEN evidenced additional faults not clearly seen in the first interpretation of surface seismic. The processing sequence was also focussed on time and amplitude picking of direct arrivals (tricky with three different tool positions), deconvolution, wave field separation, as summarized on figure 3. Figure 3: Simplified 3D-VSP processing flowchart As already mentioned all VSP displays, either stack through VSP-CDP stack, Shot domain stack or Bin migration showed a number of faults undetected by the first interpretation. These faults were organized along NW/SE trends and seen on different kind of VSP images. ( See figure 4)
Figure 4: Initial Shot collection stack, NS line, 3D VSP run2 /415-525m with a clear fault about 120m south of borehole BH1. A new 3D surface interpretation was undertaken, auto-tracking seismic events only where the confidence was high. As the number of in-lines of this particular dataset is limited, the picking was sequentially applied to all the in-lines. A priori lateral variations on seismic traces that could possibly be related to any faint fault presence remained un-interpreted. At the end of this process, unpicked seismic events, or reflector interruptions, showing lateral coherency or line-up, could be attributed to a fault. Note that no interpolation was used during this procedure. A dip map display of the UG1 horizon indicates then a series of subtle faults / lineaments corresponding to some of the subtle faults clearly evidenced by the VSP image as illustrated on figure 5.
Figure 5: Time map of UG1 on the right with dip map on the left showcasing a set of parallel NW/SE lineaments. A series of example of surface seismic with 3D-VSP data spliced on the surface data, clearly link the fault detection on the borehole dataset and the surface 3D data. A series of such images are shown in Figures 6 to 9. On figure 8, the subtle fault located near the borehole on the 3D-VSP image fits nicely with the subtle lineament surface migrated image. It was found that imperfect / non optimized 3D-VSP imaging does not prevent the detection of small faults and their strike.
X = source offset (m)Xr ( UG1) = 0.23 X
100msTWT
South North
UG1
Top UG2
S- diffracted by fault
X = source offset (m)Xr ( UG1) = 0.23 X
100msTWT
South North
UG1
Top UG2
S- diffracted by fault
LABELLING source and receiver geometry, tool depth, etc...• Edition before correlation and Vertical Stack, as needed• Time pick of first break on Direct P wave arrival
using NMO velocity template from ZO-VSP, • Fully Automatic P time pick iteration QC0 time pick• Refining Vrms- direct arrivals 1.07 Vv and computing shot statics• V-Nmo of reflections from VSP interval velocity and direct arrival Vrms
QC1 Static-Velocity• Stack of direct P wavelet, as decon. Signature , ref. ANSTEY [ 1976 ]• S-wave arrivals mute before Signature deconvolution
Mute and Deconvolution tests QC 2 Decon, QC3 Mute
** Processing ( Z component ONLY ) :• Mute of raw data before direct S-wave, • Deconvolution of whole wavefield with signature muted before S arrival • Velocity filter rejection P&S-down, on flattened P-down position • Spherical Divergence recovery ( not applied), replaced by equalization • NMO, Time shift of reflections into twt , rejection of upgoing P- S • Shot collection stacks for fast QC of preprocessing QC 4 SP stacks• VSP- CDP BinStack, with 10m & 5m CDP interval Fault interpretation• Frequency analysis of binstack images using bandpass filters • Confrontation with surface seismic, well tie; estimation of AVO effects• Migration using small aperture so as to avoid smear and fault blurring ,
ref. CLOCHARD [ 1997 ]
Labelling source and receiver geometry, tool depth, etc...Edition before correlation and Vertical Stack, as neededTime pick of first break on Direct P wave arrival
using NMO velocity template from ZO-VSP, Fully Automatic P time pick iteration QC0 time pickRefining Vrms- direct arrivals 1.07 Vv and computing shot staticsV-Nmo of reflections from VSP interval velocity and direct arrival Vrms
QC1 Static-VelocityStack of direct P wavelet, as decon. Signature , ref. ANSTEY [ 1976 ]S-wave arrivals mute before Signature deconvolution
Mute and Deconvolution tests QC 2 Decon, QC3 Mute
** Processing ( Z component ONLY ) :Mute of raw data before direct S-wave, Deconvolution of full wavefield with signature muted before S arrivalVelocity filter rejection P&S-down, on flattened P-down position Spherical Divergence recovery ( not applied), replaced by equalization NMO, Time shift of reflections into twt , rejection of upgoing P- S Shot collection stacks for fast QC of preprocessing QC 4 SP stacksVSP- CDP BinStack, with 10m & 5m CDP interval Fault interpretationFrequency analysis of binstack images using bandpass filters Confrontation with surface seismic, well tie; estimation of AVO effects3DVSP Migration using small aperture so as to avoid smear and fault
blurring , ref. CLOCHARD et al. [ 1997 ]
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 2
3VSP – a mine of information for mining exploration denis, naville, lecomte, nicoletis, suaudeau, snyman
A full image of the UG1 horizon generated by a triangulation of the interpreted horizon ( set of points from time map) and different attributes such as the square root of the normal to each triangle, or the azimuth show a more “disturbed” structure than the first interpretation could have led to believe. Figures 10 and 11 represent such Gocad images of the UG1 reflector, with exaggeration of the vertical scale. The interpretation was further pursued, by studying the picked times of direct arrivals and the derived “residuals” statics to detect potential structure effects on the VSP times. These residuals were obtained by subtraction of an axis-symmetric smooth model to the picked times of the direct arrivals. Then the residuals were averaged for level 6 to 12 of each tool position to build a surface consistent map of statics. This surface consistent term then has been subtracted to the residuals values obtained at different depths. This difference shows some organized features both for PP and SV times (figure 12). An identical approach was done on the direct arrival amplitudes, exhibiting the same structural alignments. Figure 12: Evidence of line-ups at weathered zone level with residual statics These P and S times and amplitude line-ups represent some structural features at weathered zone level, appearing the footprints of the sub-surface faults or dykes, identified on the deep target horizons (UG1). Some of these fault surfaces intersect the borehole, emerge close to the surface and create the observed anomalies in the residual seismic times. Moreover, we can check that at the expected fault position intersecting the borehole hole, the omni-directional borehole radar image shows classic reflection responses linked to fault planes such as vanishing energy or “festoon type” features (figures.7,9). On the following image ( figure 13 ), we have superimposed the sketch maps of the faults at target level (UG1, blue), at intermediate horizon H-top (orange) and at the surface (black). The WE fault and NW-SE faults intersecting the borehole exhibit a geometry coherent with their steep dip ( ~65° ) and respective strike. The NW-SE direction is coherent with the strike of nearby magnetic accident. The above results illustrate the successful integration of the effects detected on the direct arrivals of 3D-VSP (anomalies on residual times ), of the fault confirmation from 3D-VSP reflections of the surface seismic reflection images (fault at target and intermediate level) and of the borehole data ( core, logs, borehole radar). Remark: the surface data and borehole data have been mainly handled separately; innovative joint processing tests of the surface and borehole data sets have been undertaken, in order
to investigate how to further improve the overall imaging results and methods. Figure 13: Display of the interpreted faults and their line-ups derived from the dip maps (left) and from 3D-VSP residual statics (right)
CONCLUSIONS We have shown clearly in this paper the relationship between the information from the surface seismic, the borehole seismic and the conventional borehole data . To alleviate the doubts from one give method, it is important to have similar conclusion in terms of structural images from the other techniques. The borehole radar and other logs add further confidence in the interpretation of the whole core drill vicinity, prior to locate a mining shaft. The integration of the interpretation of seismic data form theses different domains is a key point in the de-risking process and decision making on a future shaft location.
ACKNOWLEDGMENTS Authors would like to thank Anglo Platinum and Anglo American for their constant support during the phases of this project. Additional geophysicists actively contributed to the pre-processing and processing of the presented VSP and surface datasets, namely Patrice Ricarte, Vincent Clochard, Kazem Kazemi and Josette Bruneau of IFPEN, and Fabienne Pradalie of CGGV.
REFERENCES - Anstey, N.A., Seismic delineation of oil and gas reservoirs using borehole geophones, patent GB1569581, 1976, CA1114937, 1977 - Clochard V., Nicoletis L., Svay-Lucas J., Mendes M., Anjos L., 1997. Interest of Ray-Born modeling and imaging for 3D walk-aways, Extended Abstracts of the 59th EAGE Conference in Geneva. - Pretorius, C.C, Gibson, M.A, Snyman, Q, 2011, Development of high resolution 3D vertical seismic profiles, The journal of SAIMM. - Pretorius, C.C, Humphries, M, Trofimczyk, K, Gillot, E, 3D VSP in a mining context, 2011, Istanbul, EAGE borehole geophysics workshop, Abstracts.
S wave static residuals P-wave static residualsS wave static residuals P-wave static residuals
Blue Line-ups at UG1 level 665m
?
Black Line-ups near surface
Orange Line-ups at H-top intermediate level ( 390m)
Blue Line-ups at UG1 level (665m)
Fault dip is about 62°The W-E fault and one of the NW-SE fault intersect both wells
Orange Line-ups at H-top level 390m
W-E Fault
NW-SE Faults/dykes
Blue Line-ups at UG1 level 665m
?
Black Line-ups near surface
Orange Line-ups at H-top intermediate level ( 390m)
Blue Line-ups at UG1 level (665m)
Fault dip is about 62°The W-E fault and one of the NW-SE fault intersect both wells
Orange Line-ups at H-top level 390m
W-E Fault
NW-SE Faults/dykes
?
W-E Fault
?
NW-SE Faults/dykes
W-E Fault
NW-SE Faults/dykes
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 3
3VSP – a mine of information for mining exploration denis, naville, lecomte, nicoletis, suaudeau, snyman
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 4
.
NorthSouth
H-top
UG1
3D VSP section
3D – Surface section
Figure 6: S-N time section: 3D-VSP data spliced on surface PSTM migration with EW red fault evident on both domains (surface and borehole), and blue SE-NW subtle lineament on surface image , fully confirmed as a fault on 3D-VSP image.
Figure 7: S-N section: EW red fault seen by UG1, H-top horizons, the borehole radar image , and emerging at surface (VSP )
NorthSouth
H-top
3D VSP section
3D – Surface section
UG1
145m North of wellhead
EW Fault Dip~61°S
NorthSouth
Intersection @ 255 m.
Loss of penetrationon radar section
H-top
UG1240m South
from well
ZO-VSP
Same EW fault seennear surface
UG1 map
145m North of wellhead
EW Fault Dip~61°S
NorthSouth
Intersection @ 255 m.
Loss of penetrationon radar section
H-top
UG1240m South
from well
ZO-VSP
Same EW fault seennear surface
UG1 map
Same EW fault seennear surface
UG1 map
3VSP – a useful tool denis, naville, lecomte, nicoletis, suaudeau, snyman
Figure 8: SW-NE time section: 3D VSP data spliced on surface PSTM migration with SE-NW blue fault clearly evidenced by the 3D-VSP image, coinciding with a subtle lineament on the UG1 map derived from 3D surface dataset. Figure 9: SW-NE section: SE-NW blue fault detected on UG1, H-top horizons, festoon on the borehole radar image where the fault intersects the hole , and by a track of residual times anomalies of the 3D VSP where the fault emerges at surface
North EastSouth West
Subtle Lineaments on 3D-HR surface patch
3D VSP section
3D – Surface section
250 m
Total extinction of UG1 reflector on 3D-VSP.
North EastSouth West
3D VSP section
3D – Surface section
250 m
Subtle Lineaments on 3D-HR surface patch
Total extinction of UG1 reflector on 3D-VSP.
250 m250 m
150m SW from well
200m NE of wellhead
NW-SE Fault Dip~64°SW
North EastSouth West
Chevron pattern on radar section @385 m
Dip60
°
H-top
UG1
NW-SE fault seen:
ZO-VSP
UG1 map
near surfaceat UG1
Intersection
150m SW from well
200m NE of wellhead
NW-SE Fault Dip~64°SW
North EastSouth West
Chevron pattern on radar section @385 m
Dip60
°
H-top
UG1
NW-SE fault seen:
ZO-VSP
UG1 map
near surfaceat UG1
NW-SE fault seen:
at UG1 near surface
UG1 map Intersection
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 5
3VSP – a useful tool denis, naville, lecomte, nicoletis, suaudeau, snyman
23rd International Geophysical Conference and Exhibition, 11-14 August 2013 - Melbourne, Australia 6
Figure 10: UG1 triangulated surface from seismic times and the square root of normal attribute, evidencing accidents.
Figure 11 : Gocad image of the structure map with the azimuth attribute spliced on a surface seismic 3D cube.
