12. The role of geophysics in shale gas exploitation

12. The role of geophysics in shale gas exploitation - GPHY...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: GPHY 3423 Petroleum Geology and Geophysics Kurt J. Marfurt and Roderick Perez (The University of Oklahoma) The role of geophysics in shale gas exploitation 12-1 Learner Objectives After this section, you will be able to: By hand, locate a microseismic event in 3D from a suite of P- and Sevents and a background velocity model. Predict which way hydraulic fractures may propagate given the present-day stress regime and maps of open or healed paleo fracture zones of weakness Use seismic anisotropy measures to predict fracture intensity and azimuth 12-2 USA gas production curves 12 10 8 6 4 2 0 1012 ft3 (TCF) Onshore unconventional Onshore conventional Offshore Alaska 1990 2000 2010 2020 2030 (EIA AEO, 2008) 12-3 Marfurt born here 12-4 Shale plays in the USA (EIA, 2009) 12-5 (Engelder, 2008) 12-6 (Engelder, 2008) 12-7 (Engelder, 2008) Natural fractures in the Woodford Shale (Wyche shale pit, OK) 12-8 (Portas, 2009) Something Marfurt remembers from thin plate elasticity Ve r ti ca ld is p Ve r ti la ce m en t ca q w= D 2 2 at Pl igi er dit y l lo ad Routinely used in 2D interpolation/data fitting Solutions to the biharmonic equation are periodic... Does loading thin rigid strata result in periodic fractures? 12-9 Hydraulic Fracturing and Location of Microseismic Events An increasing number of reservoirs require fracture stimulation to be economically viable. Monitoring of fracture geometry is essential in optimizing stimulation of reservoirs and ultimately oilfield production. Measuring the location and propagation of microseismic events allows to map the direction and intensity of individual fractures and/or fracture swarms. 12-10 Overview of microseismic monitoring of hydrofractures Treatment Well Monitor Well Velocity survey Acoustic event of known depth for geophone orientation With treatment, rock fractures generate microseisms 12-11 Released energy is recorded by geophones and located in 3D space. (Courtesy of Schlumberger) Schematic Cross sectional view of borehole acquisition in the Barnett Shale. 12-12 (Awakessian, 2005) Step 1: Build a background velocity model from P-wave and S-wave logs Upper Barnett Average Velocities Vp = 15,800 ft/s Vs1 = 8,200 ft/s Vs2 = 8,100 ft/s Lower Barnett Average Velocities Vp = 17,400 ft/s Vs1 = 9,300 ft/s Vs2 = 9,200 ft/s 12-13 (Courtesy Xavier Refunjol) Step 2: Fire perforation gun and pick P and S arrivals on each 3-component microseismic event 12-14 (Awakessien, 2009) Step 2 1/2: Modify the velocity model Find velocity model that best fits known perforation shot location Blocking 20 30 50 500 Smoothi ng 100 12-15 (Courtesy Xavier Refunjol, OU) Step 2 1/2: Modify the velocity model Find velocity model that best fits known perforation shot location 12-16 (Courtesy Xavier Refunjol, OU) Step 3: Calibrate the geophone orientation: Determine orientation of each 3-component geophone by assuming that the first arrival is a P-wave. Modify shear velocity to account for anisotropy Marble Falls Limestone Upper Barnett Limestone Upper Barnett Shale P Lower Barnett Shale P Viola Limestone 12-17 Step 4: Microseism Detection Use polarity of arrivals to determine direction of propagation 12-18 Macroseismic event location 12-19 (www.iris.edu) Time of event using Wadati diagram 60 40 TS-TP (s) VP/VS=1.769 20 0 7220 Time of event 7240 7260 7280 TP (s) 7300 7320 12-20 (redalyc.uaemex.mx) Macroseismic event location Standard earth travel time tables tS-tP 12-21 Location of event Step 5. Map time and location of each microseismic events 1495 events! 12-22 (Awakessien, 2009) Microseismic event location i. ii. iii. iv. v. Determine if an event has occurred (thresholding) Pick and compute tS-tP for each multicomponent phone Compute time of event and distance from tS-tP and velocity chart Compute azimuth from polarization of P event Compute elevation from moveout on vertical phone array Marble Falls Limestone Upper Barnett Limestone Upper Barnett Shale P Lower Barnett Shale P Viola Limestone 12-23 (Awakessien, 2009) Representative hydrofracture wells and microseismic location Pad 1 Pad 2 12-24 (Courtesy of Schlumberger) Types of induced fractures 12-25 (Fischer et al., 2005) Outline We know where the shales are. What can surface seismic add? Identify potential drilling "hazards" With microseismic data, explain variations in EUR With image logs, predict the presence of natural fractures With P-wave and S-wave sonic logs, (and ideally core) predict "fracability" With induced fractures seen on image logs, predict intensity and strike of maximum horizontal stress 12-26 Expected Ultimate Recovery (EUR) in hydraulically-fractured shale reservoirs 12-27 (Thompson, 2010) Idealized Production Scheme 12-28 (Thompson, 2010) Mathematically `Ideal' Production Scheme 12-29 (Thompson, 2010) More Realistic Production Scheme 1 1 12-30 (Thompson, 2010) Actual Production Scheme R EU =$ 5M M EUR=$1MM U= ER M 0M $ .7 Bashing B e ss a y-p ! ay dp EU R M EUR= 3M . $2.5MM $1 = Tickling Ba 12-31 ! ing sh (Thompson, 2010) Impact of bashing (Fort Worth Basin) Negative long-term impact No long-term impact Positive long-term impact 12-32 (Thompson, 2010) Micro-seismic View from the side Vertical from above Vertical well completion: microseismic events `out of zone' vertically and do not extend equally around the wellbore 12-33 4 ge a St 3 ge 2 a St tage S 1 ge a St Horizontal well completion: microseismic events `out of zone' vertically and may not connect back through every perforation. (Thompson, 2010) Barnett Shale Stratigraphic Column NE Pennsylvania n A to k a ( u n d iv id e d ) SW Frac Barrier M a r b le F a lls F o r m a t io n U p p e r L im e U p p e r S h a le Black Shale Interbedded Conglomerates, Clastics & Shales Tight Limestone Porous Limestone Porous Dolomite B a rn e tt F o r m a tio n Mississippi an F o r e s tb u r g L im e L im e W a s h L o w e r S h a le C h a p p e l R e e fs m . V io la F io n F o rm a t ro u p rg e r G im p s o n S E lle n b e iv id e d ) (U n d Frac Barrier Ordovicia n Cambrian C a m b r ia n ( U n d iv id e d ) B a r n e tt U n d iv id e d 12-34 (Thompson, 2010) Seismic data showing the Barnett Shale Marble Falls Forestburg Viola Ellenburge r 12-35 (Thompson, 2010) Correlation of microseismic events to paleostructural deformation Core area Fort Worth Basin, TX 1 Regional stress is NE-SW 1 160 horizontal wells drilled NW-SE. 300 vertical wells. Seismic survey acquired after hydraulic fracturing Hydraulic fracturing changes the anisotropy Ridges in this area appear to form hydraulic fracture barriers 12-36 (Thompson, 2010) Correlation of microseismic events to paleostructural deformation Seismic survey acquired after hydraulic fracturing Microseismic clusters fall within `bowl' shaped structure, even though the well is drilled to the NW. The effect of hydraulic fracturing is to reduce velocity anisotropy. 12-37 (Thompson, 2010) Gridded EUR (177 m by 177 m grid) 1 1 175 horizontal wells 260 vertical/deviated wells scaling: 0 EUR 10 12-38 (Thompson, 2010) Gridded EUR (177 m by 177 m grid) co-rendered with most-positive curvature 12-39 (Thompson, 2010) Major shale reservoirs in the USA Barnett 12-40 (http://www.slb.com/media/services/solutions/reservoir/shale_gas.pdf) An example from the Ft Worth Basin Well A Well C Well B 12-41 (Refunjol et al. 2010) Curvature Positive 0 Correlation of microseismic event locations with curvature Well A Negative Time slices through most-positive principal curvature k1 Well A 12-42 Well B Well C (Refunjol et al., 2010) Impedance inversion 12-43 (Refunjol et al. 2010) 8 44 Microseismic events color-coded to correspond to values of impedance inversion 12-44 (Refunjol et al. 2010) Impedance inversion ZS microseismic `window' ZS microseismic `window' 12-45 (Refunjol et al. 2010) 8 46 Crossplot showing `fracable' rock properties 12-46 (Refunjol et al. 2010) 3 Wichita-Arbuckle High tal on iz lt Fau ells lW era Min 3 3P t en es r um im x ma ess y st da 3 3 r ho 3 n ectio S 2 Llano Uplift 12-47 General tectonic features and present day regional stress field Fort Worth Basin, TX 1 Geologic cross section of Fort Worth Basin, TX 12-48 (Roth and Thompson, 2009) Karsting 12-49 (Jennings, 1985;Perez , 2009) Coherence at the approximate Barnett Shale level 12-50 k2 curvature at the approximate Barnett Shale level 12-51 12-52 k2 curvature with 50% opacity co-rendered with coherence approximate Barnett Shale level 12-53 Collapse features seen on vertical seismic and coherence time slice Vertical slice through seismic co-rendered bump map of coherence Time slice through coherence 12-54 (Roth and Thompson, 2009) Collapse features seen on vertical seismic and coherence time slice Acoustic impedance Seismic amplitude Time structure image of the Ellenberger horizon 12-55 (Roth and Thompson, 2009) Multi-attribute expression of collapse features we ll Low most-negative curvature anomalies in red Low coherence anomalies in black Dip-azimuth indicated by arrows 12-56 (Roth and Thompson, 2009) Microseismic patterns heel stage middle stage toe stage 12-57 (Roth and Thompson, 2009) Collapse chimney geobodies, interpreted horizons, and microseismic event locations 12-58 (Roth and Thompson, 2009) Anisotropy (Example of Shear-wave splitting) Me llamen el simpatico 59 12-59 Azimuthal AVO: ~ 75 Square miles 16 receiver lines, 98 channels each, 21,750 SPs (290 / sq mi) 29,100 Rcvr Stns (388 /sq mi) 30 fold Acquisition parameters 6 sectors 250 ft offset classes sector fold of 20 110 by 110 ft bins 13% empty bins 9900 ft 6 SP's 110 ft interval 880 ft line interval 10,670 ft Spider plots 12-60 (Roende et al., 2008) Stacked azimuth sector gathers Anisotropy indicators CMP no. 0.5 Time (s) 0.6 CMP CMP CMP 398 399 400 0.7 0.8 12-61 Data aligned Data Unaligned (Roende et al., 2008) Common azimuth images (Note differences in amplitude and focusing) 0.5 Time (s) 1.0 0.5 Time (s) 1.0 s 0.5 Time (s) 1.0 12-62 1 mile (Roende et al., 2008) AVO as a function of azimuth (AVOZ) 0.6 ~8 ms Time (s) 0.7 0.8 12-63 ~3500 ft (Roende et al., 2008) AVAz analysis Amplitude Azimuth 12-64 (Roende et al., 2008) AVOZ Resulting maps AVOZ 220 ft Low High 12-65 (Roende et al., 2008) Induced fractures versus expected ultimate recovery (E.U.R.) 2000 ft RGB1 Average E.U.R. RGB5 Very High E.U.R. Injection wells Observation wells RGB6 Average E.U.R. RGB7 Average E.U.R. Micro-Seismic studies suggest that large E.U.R. depends on creation (by hydrofractures) of large network of multi-azimuth vertical fractures 12-66 (Simon, 2005) Acoustic log Resistivity log Azimuth Frequency Diagram Drilling-induced fractures show that the main present-day stress field is N45E. 0 Most pre-existing natural fractures are oriented N50W. 25 % ral natu induced natural natural 12-67 (Simon, 2005) Azimuthal velocity anisotropy vs. induced fractures (Fort Worth Basin, Texas, USA) vfast-vslow (m/s) 1000 0.5 km 500 0 Poor well (fractures parallel) 12-68 (Simon, 2005) Azimuthal velocity anisotropy vs. induced fractures (Fort Worth Basin, Texas, USA) vfast-vslow (m/s) 1000 0.5 km 500 0 Good well (orthogonal fracture sets) 12-69 (Simon, 2005) Major shale reservoirs in the USA Woodford 12-70 (http://www.slb.com/media/services/solutions/reservoir/shale_gas.pdf) Deposited environment: Deep marine, highly anoxic conditions Geological time: Late Devonian to early Mississippian Depositional depth: 6000~12000ft (1830~3660m) Layer thickness: 120~280ft (36~85m) Woodford shale: organic-rich, fissile black shale Survey in the Arkoma Basin 12-71 Major geologic provinces of Oklahoma (After Northcutt and Campbell, 1995) Woodford Shale Photo Lidar Image 12-72 (Portas , 2009) The Woodford Shale serves as both source and reservoir 12-73 (Guo et al., 2010) Comparison of mineral composition Woodford shale 10% 14% Carbonate Others Barnett shale 11% Others 22% Quartz 34% Carbonate 28% Clay 48% Quartz 33% Clay Others=Orthoclase Feldspar, Oligoclase Feldspar, Albite, Anhydrite, Pyrite, Apatite (FTIR result from IC3 lab at OU) 1. The high organic content of the Woodford shale make it an excellent petroleum source rock The major quartz component allows it to be easily fractured, making it a good reservoir rock. (Guo et al., 2010) 2. 12-74 Time-structure 5 km Time (s) 1.4 1.6 1.8 2.0 2.2 2.4 fault N Woodford Horizon, Arkoma Basin, OK 12-75 (Guo et al., 2010; Seismic data courtesy of CGG-Veritas) Coherence maps lateral discontinuities in seismic waveforms 5 km Coh High ? unton e in H ps Colla ult fa fault Horizon slice along Woodford, Arkoma Basin, OK 12-76 (Guo et al., 2010; Seismic data courtesy of CGG-Veritas) Channel? Low N k2-most negative principal curvature highlights valleys and bowls (Moderate wavelength) Curv Pos 5 km 0 Channel? ? unton e in H ps Colla Neg ult fa fault N Horizon slice along Woodford, Arkoma Basin, OK 12-77 (Guo et al., 2010; Seismic data courtesy of CGG-Veritas) 1 0 78 Orientation of valley-shaped structural lineaments - displayed as volumetric rose diagrams Rose 90 0 -90 12-78 (Guo et al., 2010) Seismic derived inversion Well logs VP VS Seismic traces Synthetic Measured 12-79 The cross correlation coefficient between synthetic and real seismic is 0.932. Correlation between well location, gas/oil production and seismic attributes gas oil gas oil 6.4km 12-80 (Guo et al., 2010) Preliminary multiattribute clustering Anomalously low impedance lineament Structural valleys F re actu r s? Better production 12-81 (Guo et al., 2010) Microseismic between top Woodford and top Hunton 12-82 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Correlation between microseismic and paleostructure Azim +90 0 -90 12-83 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Correlation between microseismic and paleostructure Azim +90 0 -90 12-84 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Most-negative principal curvature, k2 ford p Wood To dford se Woo Ba 12-85 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Most-negative principal curvature, k2 12-86 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Most-negative principal curvature, k2, on Woodford (Positive values set to be transparent) Fractures concentrated in ridges and domes ? 12-87 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Microseismic between top Woodford and top Hunton 12-88 (Guo et al., 2010, microseismic data courtesy Pablo LLC) Attributes and Hydraulic fracturing of Tight Shale Gas Reservoirs In Summary: Gas-age shales are the source rock, reservoir rock, seal, and migration pathway (there is none) In general, gas shales have zero permeability and the natural fractures are healed Production from tight shale requires hydraulic fracturing, typically with water, and most commonly using horizontal wells (often 2-4 parallel wells comprising the same hydraulic fracturing operation) Induced fractures preferentially follow the direction of maximum horizontal stress Induced fractures will `pop open' previously healed fractures formed during paleotectonic deformation Azimuthal anisotropy and break-out (drilling-induced) features in image logs are good measures of present-day stress Volumetric curvature are often correlated to paleo (healed) fractures that can be reactivated during hydraulic fracturing Deposition over previous collapse features can be non-uniform providing locally thicker pay 12-89 ...
View Full Document

Ask a homework question - tutors are online