Research Themes in Stimulation Geomechanics How do stress, fractures and rock properties affect the success of stimulation? How do pressure and stress (and formation properties) evolve during stimulation? What factors affect seismic and aseismic deformation mechanisms and how do these affect the reservoir? Can we accurately model pore pressure and stress in the reservoir before, during, and after stimulation? How do we optimize slickwater frac ing?
Current Efforts New Barnett Data Set (ConocoPhillips) Piceance Tight Gas (ExxonMobil) Horn River (Apache and Encana) Bakken (Hess) Qusaiba Shale (Aramco) Pending BP, EOG
The Standard Paradigm S Hmax Shear Slip on Pre-existing Fractures and Faults Enhances Permeability Of Shale and Stimulates Production Dan Moos et al. SPE 145849 Huron Fm Horizontal Drilling and Multi-Stage Slick-Water Hydraulic Fracturing Induces Microearthquakes (M ~ -1 to M~ -3) To Create a Permeable Fracture Network Well Microseismic Events Hydraulic Fractures
Earthquake/Fault Scaling Relationships Earthquake Magnitude 8 7 6 5 4 3 2 1 0-1 - 2-3 Major: can cause serious damage over large areas. Strong: can be destructive in populated areas Moderate: can cause damage to poorly constructed buildings Noticeable shaking but damage is unlikely Minor: felt but does not cause damage slip on fault EQ stress drop largest Guy, AR event largest Blackpool, UK event typical microseismic event during stimulation 10 0 10 1 10 2 10 3 10 4 10 5 Fault Patch Size (m) 10 20 10 18 10 16 10 14 10 12 10 10 10 8 10 6 Earthquake Moment (Nm)
Production Does Not Correlate with Microseismicity Dan Moos et al. SPE 145849 5
Microearthquakes and Spectragrams Vr ~ 2 km/s tr ~ 0.5 ms aaaa Das and Zoback, The Leading Edge (July 2011)
Long Period Long Duration Seismic Events Long period (below 80 Hz) long duramon (10-100 s) events detected in the spectrograms LPLD waveforms aqer band- pass filtering from 10-80 Hz 100 Hz Stage 7 Stage 8 No seismic recording 10 4 Psi 0 100 3000 s 5000 s 3000 s 0 350 s Tectonic tremor waveforms from Vancouver Island, Central Range (Taiwan) and the SAF (Peng and Gomberg, 2010) Hz 0 60 s 350 s Das and Zoback (2011) 350 s 8
Slow Slip on Cross-Cutting Faults 3 2 1 aaaa aaaa
How Big Are Slow Slipping Faults? Spectral Division 80s ω -3 Corner Freq. ~ 30 Hz 1s 1 sec window ~M -1 ~ 60 sec LPLD ~M 0
Earthquake/Fault Scaling Relationships Earthquake Magnitude 8 7 6 5 4 3 2 1 0-1 Major: can cause serious damage over large areas. Strong: can be destructive in populated areas Moderate: can cause damage to poorly constructed buildings Noticeable shaking but damage is unlikely Minor: felt but does not cause damage largest Guy, AR event largest Blackpool, UK event LPLD event 10 20 10 18 10 16 10 14 10 12 10 10 10 8 Earthquake Moment (Nm) - 2 10 6-3 10 0 10 1 10 2 10 3 10 4 10 5 Fault Patch Size (m)
LPLD events Associated with Region of Greatest Pore Pressure Perturbation Instantaneous Shut- In Pressure (ISIP) values per stage for well A- B Minimum pressure perturbamon in the reservoir during hydraulic fracturing of the five wells 12
Organic Rich Shales Sample group Clay Carbonate QFP TOC (wt%) Barnett-dark 29-43 0-6 48-59 4.1-5.8 Barnett-light 2-7 37-81 16-53 0.4-1.3 Haynesville-dark 36-39 20-23 31-35 3.7-4.1 Haynesville-light 20-22 49-53 23-24 1.7-1.8 Fort St. John 32-39 3-5 54-60 1.6-2.2 Eagle Ford-dark 12-21 46-54 22-29 4.4-5.7 Eagle Ford-light 6-14 63-78 11-18 1.9-2.5 Bedding plane and sample cylinder axis are either: parallel (horizontal samples) or perpendicular (vertical samples) 3-10 % porosity All room dry, room temperature experiments In-situ (and lab) effective stress between 15-30 MPa 13
Rate and State Friction
The Case for Slow Slip on Faults Coefficient of Friction 1 0.8 0.6 0.4 0.2 Unstable Barnett Light Eagleford Light Haynesville Light Eagleford Dark Haynesville Dark Barnett Dark Stable x 10 3 10 5 0 (a - b) 0 0 10 20 30 40 50 60 5 Clay + Organic Content (wt%)
An Alternative to the Standard Paradigm Evidence for slowly slipping faults (not seen in microseismic surveys) during multi-stage hydraulic fracturing of shale gas reservoirs 1 <Now seen in 4/4 data sets> Why appreciable slow slip on faults during stimulation is expected 2 Implications of slowly slipping faults for optimization of the stimulation process 2
Slow Slip on Mis-Oriented Faults Prior to Stimulation During Stimulation
Induced Slip on Mis-Oriented Faults Seismic Slip Aseismic Slip No Slip
Slip is Always Slow on Mis-Oriented Faults High Pressure Reduces Normal Stress and Induces Slip But Slip can Only Occur as Pressure Propagates Along the Fault
Concentrate Stimulation in Parts of the Reservoir with Pre-Existing Faults Devonian Shale Northeastern U.S. Micro seismicity Natural Fracture / Joint Density Correlation of Flow with Fractures not Meqs Ports Packers Dan Moos et al. SPE 145849 Production from PLT
Creating a Fault Network
Building A Quantitative DFN Model All Fractures Stimulated Fractures Connected to Frac and Wellbore
Fracture Extension in Shear
Building Quantitative DFN Models Stimulated Fractures Connected to Frac and Wellbore New Fractures All Fractures Connected to Frac and Wellbore
Q = α t q = 1 2 t α α A k m