Geomechanical Effects of Waterflooding II INTERNATIONAL SEMINAR ON OILFIELD WATER MANAGEMENT OCTOBER 16 19, 2007 Dale Walters A. (Tony) Settari Taurus Reservoir Solutions
General Outline Characteristics of waterflood/pwri/cri fracturing Poro- and thermo-elastic elastic stress changes Geomehcanical Effects Below Fracturing Pressure Stress-dependent permeability and porosity Fault instability Caprock integrity Geomehcanical Effects Above Fracturing Pressure Fracturing pressures Fracture growth Fracture containment Formation plugging effects Integrated Reservoir Management
Waterflood,, PWRI, CRI - Unconventional fracturing Features different from conventional fracturing: Large leak-off (1-D D leak-off approximation not adequate) Large poroelastic and thermal effects on stresses and permeability Large time scale (avg( reservoir pressure and stresses can change) Low viscosity fluids K c and stress containment more dominant for fracture geometry Formation damage Conventional tools inadequate, generally overestimate frac growth Large economic consequences water and waste disposal is a major cost
Useful Case from Poro- and Thermoelasticity: Uniaxial Strain Free Surface Ground Level Reservoir Reservoir Material Properties: E, ν, α, a L Loads: p, T Assumptions: σ ε x x σ = σ = ε z = 0 y y = σ = ε h h = 0 Total vertical displacement change: ε z = ε Thermoelastic displacement change: ε z = Poroelastic displacement change: ε z zt a L 1+ ν = α 1 ν + ε 1+ ν 1 ν zp T ( 1 2ν ) E p Total stress change: σ h = σ ht + σ Thermoelastic stress change: σ h = a L E 1 ν T Poroelastic stress change: 1 2ν σ h = α p 1 ν hp
Poroelastic/thermal effects on pfoc Origin: P and T gradients cause change in the confining stress acting on the fracture also called back stress. As a result, fracture propagation pressure changes during pumping and closure Analytical estimates in 2-D 2 D slices (Hagoort( Hagoort,, Smith, ) based on linear elasticity and elliptic flow pattern Energy methods (numerical evaluation of integrals) is more general Fully coupled solution of fracture opening, leak-off and poro/thermoelastic stresses around frac is the most rigorous method Important for high leak-off situations
Poroelastic/thermal effects Consequences for history matching/design: In conventional fracturing pressure and thermal effects are opposing tend to cancel out In high leak-off fracturing larger effects result; either P or T can dominate In thermal operations P and T compound each other - very large increases in stresses can result Poroelastic component tends to reduce the containment (stress contrast) for height growth In many design programs, only the poroelastic component is considered In layered formations, differential back stresses exist; the analytical methods are difficult to apply In waterflood fracturing it is often termed thermal fracturing
Below Fracturing Pressure Stress dependent properties Fault stability Caprock Integrity
Stress dependent properties Permeability and porosity can be strongly affected by changes in effective stress Benefits: Increased injectivity Increased storage capacity
Stress dependent properties with injection Example Brine disposal into Oriskany sandstone (k=0.01-0.03 md, φ =1.5%, but likely microfractured) ) (SPE REE, April 1999) At fracture face: 1000x k, 3 x f! Porosity ratio φ/φ init vseff stress 3 Permeability vs eff stress 1000 Normalized Pore Volume, dpv/pvi, (fraction) 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 Sample 1 @ 5245.8 Oriskany Sample 2 @ 5266.0 Oriskany Sample 3 @ 5383.7 Heldelberg Sample 4 @ 5391.7 Heldelberg Sample 5 @ 5634.4 Anhydrite Sample 6 @ 5636.4 Anhydrite approx fit (y=13.515 x ^ -0.3297) Permeability (md) 100 10 1 0.1 results from PTA Best fit exponential 1 0 500 1000 1500 2000 2500 3000 Net Stress (decreasing only), psi 0.01 2000 2500 3000 3500 4000 4500 5000 5500 6000 pressure (psia)
Fractured Reservoirs Fracture deformations stongly stress dependent Permeability of fracture system can induce permeability anisotropy Increased permeability of fractures can cause premature breakthrough
Fracture (single joint) deformation Normal deformation Shear deformation Joint behavior under normal load is non-linear: k n d n = σ d v Deformation of the matrix+fracture σ n = normal effective stress ν = Deformation normal to joint plane k n = normal stiffness of the joint Deformation of the fracture only
Example: Fractured Reservoir Kxx (ARMA/USRMS 06-1101)
Fault Stability Fault mechanics dependent on effective normal stress and shear stress Initial and dynamic stress state Increased pressure due to injection can cause critical loading of faults Limiting injection pressures to avoid failing faults can be prohibitive to maintaining voidage
Example: Fault Reactivation StrLev = τ τ max τ max = c + S Stress Level > 1 = FAILURE ' n tan φ
Above Fracturing Pressure Induced planar fractures Shear Fracturing/dilation Micro-cracking cracking Material behavior controls the type of failure and fracturing Effects: Increased injectivity Altered stress state Increased risk of failure out-of of-zone
Some possible models of injection into geomaterials a) primary SPF, no formation failure (classical fracture mechanics) b) combination of SPF and microfracturing or microchanelling (hard rock, granular media) c) Homogenized model (fracture + enhanced zone + unfailed zone) - successfully used in oil sands, chalk, coal d) Induced fracturing or microchanelling model - in-situ fracture networks develop based on the local stress state induced by injection
Possible models of injection into geomaterials e) Jointed rock model -joint deformation and permeability increases as a function of effective stress state. Used to explain observed directionality in waterfloods(primarily hard rock).
Evidence for secondary fractures or dilation Injection/fall-off off testing in oil sands: PTA methods (Leshyschyn( Leshyschyn) 3 regions clearly visible Reservoir,.91 md Dilation, 5 md Fracture, 420 md
Evidence for the dilation of microfractured rock around SPF Example Brine disposal into Oriskany sandstone (k=0.01-0.03 md, φ =1.5%, but likely microfractured) ) (SPE REE, April 1999) At fracture face: 1000x k, 3 x φ! permeability multiplier 10000000 1000000 100000 10000 1000 100 10 Fracture flow Reservoir flow 1000 Permeability (md) Normalized Pore Volume, dpv/pvi, (fraction) 100 10 1 3 2.8 2.6 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0 500 1000 1500 2000 2500 3000 Net Stress (decreasing only), psi results from PTA Best fit exponential Sample 1 @ 5245.8 Oriskany Sample 2 @ 5266.0 Oriskany Sample 3 @ 5383.7 Heldelberg Sample 4 @ 5391.7 Heldelberg Sample 5 @ 5634.4 Anhydrite Sample 6 @ 5636.4 Anhydrite approx fit (y=13.515 x ^ -0.3297) 1 0.00 500.00 1000.00 1500.00 2000.00 2500.00 min. eff. stress (psia) 0.1 0.01 2000 2500 3000 3500 4000 4500 5000 5500 6000 pressure (psia)
PWRI/CRI: Propagation of induced fractures Field evidence: Injectivity of wells declines with time (sea water or PW injection) Consequence: Dynamic fracture develops even in very high kh formation with time Competition between mechanisms: Permeability of reservoir can increase by 1-21 2 orders of magnitude due to re-opening of fractures - can cause creation of dual porosity media Damage can cause decrease of permeability by orders of magnitude Need to handle these aspects in a coupled manner Large economic impact on EOR projects (max inj pressure, number of wells, facilities design )
Coupled nature of the PWRI/CRI Damage reduces permeability and increases pressure gradients Stresses change as a result of cooling (thermal stress) and pressure gradients (poroelastic stress) - competition Fracture propagation is controlled by leak- off perm and frac propagation pressure injection process Damage k=f(v) Frac growth Cooling P gradients P foc changes Leak-off dp Stress changes Injection pressure
Conceptual model of PWRI/CRI injection Dilation zone Compaction zone Internal cake (k damage in the formation) Fracture tip plugging L f External plugging on fracture face (filter cake) Mechanisms of PWRI or CRI injection at frac pressure in soft formations
Modeling of permeability damage Types of damage: External (filter cake on the surface) Internal (particle deposition in the formation) Formulation of internal damage: Deep bed filtration theory (DBF) Empirical velocity (throughput) based model (VBM) There is an equivalence between the models (SPE 79695); the empirical model is easier to use
Damage Model in the Simulator Generalized 3-parameter 3 equation for k : k / k 0 1 R R min min = (1 +α( V / A) n = shape coefficient α = damage coefficient R min = asymptotic damage ratio (max damage) In the simulator, cumulate volume of water V at block faces as V = Σ Q k t k 1 n ) (k) damaged/(k)initial (k) damaged/(k)initial 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Damage model with constant a =0.5, R min =0 0.1 1 10 100 Volume througput/area Damage model with constant alpha=0.5, n=1 0.1 1 10 100 Volume througput/area n=1.0 n=0.6 n=1.4 Rmin=0.0 Rmin=0.1 Rmin=0.2
Coupled geomechanical model for PWRI with damage PWRI Injection Stress-strain model Multiphase thermal reservoir model Cooling reduces stress + Pfoc Dynamic fracture develops Progressive damage on frac wall
Example: Water disposal in the Masila Block, Yemen Current water production 1,000,000 BPD, will increase in the future Disposal cost is a significant part of economics PWRI at fracture pressure is being considered: Can we achieve 150,000 BPD/ well? What will be the surface pressures?
Experience with PWRI in the Masila Block Injectivity of wells declines with time Vertical wells Horizontal wells Vertical wells with propped fractures Damage mechanisms Permeability damage (external and internal) common to many PWRI projects Check valve effect unique, not well understood (attributed to fines mobilization) Typical of many PW injection projects
Injectivity history Camaal 30 160 140 Matrix acid jobs Propped frac Injectivity index (bbls/psia/day) 120 100 80 60 40 20 0 0 100 200 300 400 500 600 700 time from 02/01/99 (days)
Matching Haru 4 PW injector (SPE 77600) 2700 100000 BHP, data and simulated (psia) 2500 2300 2100 1900 1700 1500 0 100 200 300 400 500 600 700 800 time (days) Observed BHP case 1 Case 2 Case 3 inj rate 90000 80000 70000 60000 50000 40000 Case α (1/ft) n n R min 1 - median a L 0.0005 1 0.1 2 - high a L, low conductivity 0.002 1 0.035 3 - high a L, high conductivity 0.2 1 0.003 Injection rate (BWPD) Coupled problems: More variables, therefore must be constrained by more data to be unique
Nature of non-uniqueness permeability permeability Damage effect a L effect Pressure s y Damage effect a L effect Pressure s y Low damage small fracture High damage large fracture Laboratory measurements of a L make the solution unique
Geomechanical effects caused by reservoir development Stress re-orientation Long term changes of stresses and frac pressure Fracture containment
Due to Stress reorientation Reservoir depletion or injection global changes Local poroeastic/thermoelastic thermoelastic well effects local changes Fracture loading (stress field from fracture opening) Effect on Fracture initiation (breakdown) pressure and initiation plane Fracture propagation pressure Fracture direction Examples: Areal stress reorientation due to existing fracture (Wong, 1988) Initiation of a steam fracture from horizontal well (Settari( et al., 2001)
Effect on fracture containment Depletion typically increases containment Pre-injection and heating/cooling can be used to manipulate stress but only close to the wellbore small fracs Initial horiz. stress Altered horiz. stress Zone 1 - pressured shales Zone 2 - depleted altered initial
Example: Stress changes with water injection
Stress Changes and Fracture Containment -8100 6000 6500 7000 7500 8000 8500-8200 -8300 t= 0 t= 100 t= 600 t= 1200-8400 Depth -8500-8600 -8700-8800 -8900 Smin (psi)
Summary of effects important for Reservoir Management Accounting and designing for stress dependent permeability, porosity and/or fracturing can increase injectivity and reduce development costs Increased injectivity may reduce # of injectors or allow pressure maintenance enhancing production Understanding fault mechanics and failure may allow optimization of maximum injection pressures Environmentally friendly disposal of drilling waste can be controlled Understanding altered stress state can be used to reduce risks associated with fracturing and infill drilling Accounting for induced anisotropy of permeability can be used to plan injector placement and optimize sweep Quantifying geomechanical effects requires additional data