Shale Experience from a global gas perspective
Shale Gas Experience from a Global Gas Company Perspective 25 October 2011
Alex Gabb
Agenda • BG Group business and Shale Gas.
• Overview of Shale. • How is shale gas appraised and developed? • Technological challenges that BG considers need to be addressed in order to put more science into what has been a largely empirical understanding to date.
2
Global LNG: growing a global business
UK Italy
USA
Japan
Egypt
Trinidad & Tobago
China
Singapore* Nigeria
EG
Tanzania Brazil
Australia
Chile *exclusive right to supply
Equity position Existing long term supply source
Liquefaction under construction
Potential liquefaction
Existing import capacity
Potential import capacity
Long term customer
Global assets, supply and markets
3
Shale Gas Basins of the World
4
Shale – An Outcrop
5
What makes a good shale? Matrix Porosity/Permeability
Reservoir Pressure Typically Overpressured; 0.6 psi/ft upwards.
Porosity >4% & Micro/Nanodarcies
Gas In Place
Containment
Free and Adsorbed ‘Resource Density’
Frac’ Containment
Thermal Maturity Degree of ‘Cooking’
‘Fracability’ Able to Initiate and Propagate and Complex Fracture Network
Organic Richness High TOC >2% and Adsorbed Gas Content
Unlikely that you can get your shale to work if you don’t have all of these!
6
Shale Formations – Finding the right sort!!
Matrix Permeability & Porosity 100000
10000
1000
Conventional Oil ~100 mD - 10D
100
Permeability (mD)
10
Conventional Gas ~100 – 0.1 mD 1 0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
0.1
Typical Limit of ‘Standard Core’ Measurements.
0.01
0.001
0.0001
0.00001
Shale Gas ~0.001 – 0.0001 mD = Microdarcies – Nanodarcies!
Porosity (Frac.) Conventional Oil
Shale
8
Shale in Microscopic Detail Quartz + Other Minerals
Pore
Phyllosilicates
Connected Pores (Blue); Kerogen (Green); Isolated Pores (Red)
Kerogen
500 nm
• Pore structure has similar dimensions to the gas molecules themselves. • Darcy Flow versus Diffusive Flow.
Gas within complex pore system with even more complex flow physics.
Adsorption/Desorption Mechanisms
Adsorption
– Adhesion of a single layer of gas molecules to the internal surface of the coal or shale matrix. – Physical Process versus Chemical Desorption
– The process whereby adsorbed gas molecules become detached from the pore surfaces and take on the kinetic properties of free gas.
10
How do we measure Gas Content? • Determined from Canister Tests.
200 Gas Content (scf/ton)
• At Reservoir Temperature? Possibly not! • Lost Gas + Measured Gas + Crushed Gas. • In shales this is a combination of free and adsorbed gas. • Varies according to TOC%; so need to take enough samples to characterise the reservoir interval. This is more than one sample!
80 40
2
4
6
8
10
12
14
16
18
20
TOC (Wt. %)
300
Methane
48.5 scf/ton
Gas Storage Capacity, scf/ton
Methane Storage Capacity, scf/ton
120
0
Methane Isotherm Results
18.9 scf/ton
50
New Albany Shale
0
70 60
Antrim Shale
160
TOC = 4.97 wt. %
40 30 20 TOC = 1.98 wt. %
10
Carbon Dioxide
Ethane
Mixture
200
250
200
150
100
50
0
0 0
500
1,000
1,500
2,000
2,500
Pressure, psia
Data should be considered qualitative
3,000
0
1,000
2,000
3,000
Pressure, psia
4,000
5,000
Gas Transport in Shale : An Analogy
12
Shale Gas Log Responses GR 0-200
NPHI-RHOB
DT
RT-RXO
•
Borehole conditions are typically good; low clay content = hole stability.
•
GR is very high; Marcellus shales > 800 API
•
Density – Neutron
– •
•
USA Haynesville
Less Shale separation; Cross-over due to gas kerogen and lack of water.
Resisitivity; Usually quite high > 20 ohmm
–
Low Rt caused by water-wet shales or graphite (in some over-mature areas)
PEF; Often > 30 due to heavy minerals
Kerogen
Rock (Density Log)
Total Porosity Total Porosity (GRI Method)
Matrix V Clay
Bulk Minerals
Clay Layers Structural Water (OH) -
Kerogen
Caliper/PEF
Clay Surfaces & Interfaces Hydration or Bound Water
Small Pores
Large Pores
Capillary Water Hydrocarbon Pore Volume
Irreducible or Immobile Water
Total Saturation (GRI Method) Modified from Hill et al, 1969
13
Shale Gas Reservoir Core Analysis Core Description Wellsite Canister Homogeneous Wholecore sections (0.3 to 0.5m, canisters at reservoir temp.)
Wireline or fast retrieval of core recommended to minimised lost gas
Measure Gas Composition during Desorption (90 days approx.)
Select Fresh State Wholecore Samples for Rock Mechanics Tests (Triaxial Static Tests, Vp & Vs)
Langmuir Isotherm Analysis
Half canister sample
Halve and quarter sample using diamond saw Quarter canister sample
Grain Density, Porosity, Total Organic Content, Rock Eval Pyrolysis, Vitrinite Reflectance, XRD & XRF Mineralogy
Select Fresh State Core Samples (approx 500g)
Quarter canister sample
Residual/Crushed Gas Analysis
Fresh State Bulk Density Matrix Permeability
Crush Sample
Dean-Stark Analysis for Sw, So & Sg
High Pressure Mercury Injection SEM (Argon/iron beam milling of surface) Fluid Sensitivity Tests (Clay swelling & fracture flow tests)
Grain Volume and Grain Density
Gas Shale Core Analysis • Standard or conventional methods of core analysis for porosity, saturations, and permeability are unsuitable for Gas Shales
• Porosity requires sample cleaning of a plug and a Boyle’s Law porosity using Helium
– Difficult to take plugs in many shales due to bedding plane partings – Measurement requires equilibrium to be obtained which requires a long time in nano-darcy permeability and diffusion rates • Permeability measurements on plugs have the same problem and must use pressure decay techniques for the low permeability ranges
Principals of Hydraulic Fracturing Objective: Create a high conductivity crack within the reservoir • Rock is split using liquid that is pumped under high pressure
• Tiny split or fracture held open using proppant
• Gas flows from the fracture
It is also imperative that the fracture system stays open.
16
What System Do We End Up with? Complex Fracture System Complex Well Geometry in a Tight Reservoir
Complex Pore System 17
Key Differences : Conventional and Shale Gas Characteristic
Conventional Gas
Shale
Gas Generation
Gas is generated in the source rock and then migrates into the reservoir.
Gas is generated and trapped within the source rock.
Gas Storage Mechanism
Compression.
Compression and adsorption.
Gas Produced
Free gas only.
Free and adsorbed gas.
• Minimal transient period followed by a long boundarydominated flow period. • Production rates are mainly relatable to permeability and declining reservoir pressure..
• Very long transient (linear) flow period that can extend many years. In some cases, it is debatable if boundary-dominated flow will ever be fully realized. • Production rates are mainly relatable to the success of creating a large fracture network around a long horizontal wellbore and to the matrix permeability.
• Recovery factor = 50% – 90%
• Recovery factor = 15% – 40%
Production Performance
Recovery Factors
Conventional Gas
Shale Gas
18
What did Arps intend? And what do we do? • Most decline curve analysis is based on the Arps Equation (or set of equations!) which was presented in 1945. b = 0; Exponential Decline
0 < b < 1; Hyperbolic Decline
b = 1; Harmonic Decline
• Supposed to be a constant pressure steady-state solution; in a shale gas well typically we would not have this condition. • b = 1 intended as a special case since implies infinite recovery at infinite time; this implies an unbounded system; the use of b>1 is common place in the production analysis of shale gas wells. • Shale Gas well is almost always in TRANSIENT FLOW .. Arps is intended for a STABILISED FLOW scenario.
19
Typical gas shale production profiles 25000
Typical Duration of Production Dataset
Field
Haynesville
Marcellus
IP (mscf/d)
20000
5000
Di (%)
80
68
B
1.1
1.3
Dt (%)
6
6
EUR30 (bscf)
7.380
4.301
IP30 (mscf/d)
17676
4537
Gas Rate (mscf/d)
20000
15000
10000
Typical Duration of Production Forecast (with minimal understanding of reservoir physics!!) Risks • Water and/or Condensate Hold-Up. • Well Integrity (or Lack of) • Reservoir Compaction
5000
0 0
5
10
15
20
25
30
Year Haynesville
Marcellus
Gas Desorption and Diffusive Flow leads to long production ‘tail’
20
Rate Transient Analysis 10 days
1
100s days
2
5000
4500
4000
3500
exposed fracture surface area.
1
3000 Gas Rate (mscf/d)
Characterized by infinite-acting linear flow into
Characterized by quasi-steady depletion of SRV
2500
2000
1500
2
1000
3
500
4
0
1
2
3
4
5
6
7
8
9
Year
10000s days Characterized by transient infinite-acting linear flow into external faces of SRV.
4
0
10
3
1000s days Boundary-dominated flow characterized by quasisteady flow from the depletion volume into external faces of SRV
Analysis Options • Stretched Exponental Production Decline (SEPD) – Avoids the requirement for key parameters to vary with time but may require a large population of wells in order to constrain parameters effectively.
– Ref. John Lee, Valko, Ilk et al.
• Root Time Methodologies – OK if the well is in transient linear flow but the deviation from linear trend indicates transition to boundary dominated flow.
• Simulation – A large number of input parameters; which are poorly defined and constrained, not least the characteristics of the Stimulated Rock Volume (SRV).
– BUT also in-situ matrix permeability and desorption-diffusion coefficients.
22
Shale Gas Development • Gas
shale developments utilise horizontal wells and multi-stage fracs. • This is repeated many times over a large area. • These developments are CAPEX and human resource intensive. • For success, the process needs to be efficient. • Well planned appraisal and pilot production stages are essential. • Success is NOT guaranteed; so ‘offramps’ need to be clearly defined before moving to a development.
23
Technical Challenges – Five Key Ways to Improve • Prediction of SRV permeability prior to treatment. – Combination of Geomechanics and flowing well behaviour.
• Better understanding of transport physics. – Prediction of Diffusivity and Darcy flow in shales.
• Understanding of Transient Behaviour of a producing well. – Can the geometric attributes of the SRV be defined from flowing rate and pressure data?
• Impact of Liquids on Shale Gas Well Deliverability. – Liquids rich plays are becoming more attractive (as a hedge against low gas prices); how does the physics differ from a pure gas play?
• Water Management. – Shale Gas is a BIG water consumer; and water is becoming a scarce commodity.
24
Questions?
25
Alex Gabb
Agenda • BG Group business and Shale Gas.
• Overview of Shale. • How is shale gas appraised and developed? • Technological challenges that BG considers need to be addressed in order to put more science into what has been a largely empirical understanding to date.
2
Global LNG: growing a global business
UK Italy
USA
Japan
Egypt
Trinidad & Tobago
China
Singapore* Nigeria
EG
Tanzania Brazil
Australia
Chile *exclusive right to supply
Equity position Existing long term supply source
Liquefaction under construction
Potential liquefaction
Existing import capacity
Potential import capacity
Long term customer
Global assets, supply and markets
3
Shale Gas Basins of the World
4
Shale – An Outcrop
5
What makes a good shale? Matrix Porosity/Permeability
Reservoir Pressure Typically Overpressured; 0.6 psi/ft upwards.
Porosity >4% & Micro/Nanodarcies
Gas In Place
Containment
Free and Adsorbed ‘Resource Density’
Frac’ Containment
Thermal Maturity Degree of ‘Cooking’
‘Fracability’ Able to Initiate and Propagate and Complex Fracture Network
Organic Richness High TOC >2% and Adsorbed Gas Content
Unlikely that you can get your shale to work if you don’t have all of these!
6
Shale Formations – Finding the right sort!!
Matrix Permeability & Porosity 100000
10000
1000
Conventional Oil ~100 mD - 10D
100
Permeability (mD)
10
Conventional Gas ~100 – 0.1 mD 1 0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
0.1
Typical Limit of ‘Standard Core’ Measurements.
0.01
0.001
0.0001
0.00001
Shale Gas ~0.001 – 0.0001 mD = Microdarcies – Nanodarcies!
Porosity (Frac.) Conventional Oil
Shale
8
Shale in Microscopic Detail Quartz + Other Minerals
Pore
Phyllosilicates
Connected Pores (Blue); Kerogen (Green); Isolated Pores (Red)
Kerogen
500 nm
• Pore structure has similar dimensions to the gas molecules themselves. • Darcy Flow versus Diffusive Flow.
Gas within complex pore system with even more complex flow physics.
Adsorption/Desorption Mechanisms
Adsorption
– Adhesion of a single layer of gas molecules to the internal surface of the coal or shale matrix. – Physical Process versus Chemical Desorption
– The process whereby adsorbed gas molecules become detached from the pore surfaces and take on the kinetic properties of free gas.
10
How do we measure Gas Content? • Determined from Canister Tests.
200 Gas Content (scf/ton)
• At Reservoir Temperature? Possibly not! • Lost Gas + Measured Gas + Crushed Gas. • In shales this is a combination of free and adsorbed gas. • Varies according to TOC%; so need to take enough samples to characterise the reservoir interval. This is more than one sample!
80 40
2
4
6
8
10
12
14
16
18
20
TOC (Wt. %)
300
Methane
48.5 scf/ton
Gas Storage Capacity, scf/ton
Methane Storage Capacity, scf/ton
120
0
Methane Isotherm Results
18.9 scf/ton
50
New Albany Shale
0
70 60
Antrim Shale
160
TOC = 4.97 wt. %
40 30 20 TOC = 1.98 wt. %
10
Carbon Dioxide
Ethane
Mixture
200
250
200
150
100
50
0
0 0
500
1,000
1,500
2,000
2,500
Pressure, psia
Data should be considered qualitative
3,000
0
1,000
2,000
3,000
Pressure, psia
4,000
5,000
Gas Transport in Shale : An Analogy
12
Shale Gas Log Responses GR 0-200
NPHI-RHOB
DT
RT-RXO
•
Borehole conditions are typically good; low clay content = hole stability.
•
GR is very high; Marcellus shales > 800 API
•
Density – Neutron
– •
•
USA Haynesville
Less Shale separation; Cross-over due to gas kerogen and lack of water.
Resisitivity; Usually quite high > 20 ohmm
–
Low Rt caused by water-wet shales or graphite (in some over-mature areas)
PEF; Often > 30 due to heavy minerals
Kerogen
Rock (Density Log)
Total Porosity Total Porosity (GRI Method)
Matrix V Clay
Bulk Minerals
Clay Layers Structural Water (OH) -
Kerogen
Caliper/PEF
Clay Surfaces & Interfaces Hydration or Bound Water
Small Pores
Large Pores
Capillary Water Hydrocarbon Pore Volume
Irreducible or Immobile Water
Total Saturation (GRI Method) Modified from Hill et al, 1969
13
Shale Gas Reservoir Core Analysis Core Description Wellsite Canister Homogeneous Wholecore sections (0.3 to 0.5m, canisters at reservoir temp.)
Wireline or fast retrieval of core recommended to minimised lost gas
Measure Gas Composition during Desorption (90 days approx.)
Select Fresh State Wholecore Samples for Rock Mechanics Tests (Triaxial Static Tests, Vp & Vs)
Langmuir Isotherm Analysis
Half canister sample
Halve and quarter sample using diamond saw Quarter canister sample
Grain Density, Porosity, Total Organic Content, Rock Eval Pyrolysis, Vitrinite Reflectance, XRD & XRF Mineralogy
Select Fresh State Core Samples (approx 500g)
Quarter canister sample
Residual/Crushed Gas Analysis
Fresh State Bulk Density Matrix Permeability
Crush Sample
Dean-Stark Analysis for Sw, So & Sg
High Pressure Mercury Injection SEM (Argon/iron beam milling of surface) Fluid Sensitivity Tests (Clay swelling & fracture flow tests)
Grain Volume and Grain Density
Gas Shale Core Analysis • Standard or conventional methods of core analysis for porosity, saturations, and permeability are unsuitable for Gas Shales
• Porosity requires sample cleaning of a plug and a Boyle’s Law porosity using Helium
– Difficult to take plugs in many shales due to bedding plane partings – Measurement requires equilibrium to be obtained which requires a long time in nano-darcy permeability and diffusion rates • Permeability measurements on plugs have the same problem and must use pressure decay techniques for the low permeability ranges
Principals of Hydraulic Fracturing Objective: Create a high conductivity crack within the reservoir • Rock is split using liquid that is pumped under high pressure
• Tiny split or fracture held open using proppant
• Gas flows from the fracture
It is also imperative that the fracture system stays open.
16
What System Do We End Up with? Complex Fracture System Complex Well Geometry in a Tight Reservoir
Complex Pore System 17
Key Differences : Conventional and Shale Gas Characteristic
Conventional Gas
Shale
Gas Generation
Gas is generated in the source rock and then migrates into the reservoir.
Gas is generated and trapped within the source rock.
Gas Storage Mechanism
Compression.
Compression and adsorption.
Gas Produced
Free gas only.
Free and adsorbed gas.
• Minimal transient period followed by a long boundarydominated flow period. • Production rates are mainly relatable to permeability and declining reservoir pressure..
• Very long transient (linear) flow period that can extend many years. In some cases, it is debatable if boundary-dominated flow will ever be fully realized. • Production rates are mainly relatable to the success of creating a large fracture network around a long horizontal wellbore and to the matrix permeability.
• Recovery factor = 50% – 90%
• Recovery factor = 15% – 40%
Production Performance
Recovery Factors
Conventional Gas
Shale Gas
18
What did Arps intend? And what do we do? • Most decline curve analysis is based on the Arps Equation (or set of equations!) which was presented in 1945. b = 0; Exponential Decline
0 < b < 1; Hyperbolic Decline
b = 1; Harmonic Decline
• Supposed to be a constant pressure steady-state solution; in a shale gas well typically we would not have this condition. • b = 1 intended as a special case since implies infinite recovery at infinite time; this implies an unbounded system; the use of b>1 is common place in the production analysis of shale gas wells. • Shale Gas well is almost always in TRANSIENT FLOW .. Arps is intended for a STABILISED FLOW scenario.
19
Typical gas shale production profiles 25000
Typical Duration of Production Dataset
Field
Haynesville
Marcellus
IP (mscf/d)
20000
5000
Di (%)
80
68
B
1.1
1.3
Dt (%)
6
6
EUR30 (bscf)
7.380
4.301
IP30 (mscf/d)
17676
4537
Gas Rate (mscf/d)
20000
15000
10000
Typical Duration of Production Forecast (with minimal understanding of reservoir physics!!) Risks • Water and/or Condensate Hold-Up. • Well Integrity (or Lack of) • Reservoir Compaction
5000
0 0
5
10
15
20
25
30
Year Haynesville
Marcellus
Gas Desorption and Diffusive Flow leads to long production ‘tail’
20
Rate Transient Analysis 10 days
1
100s days
2
5000
4500
4000
3500
exposed fracture surface area.
1
3000 Gas Rate (mscf/d)
Characterized by infinite-acting linear flow into
Characterized by quasi-steady depletion of SRV
2500
2000
1500
2
1000
3
500
4
0
1
2
3
4
5
6
7
8
9
Year
10000s days Characterized by transient infinite-acting linear flow into external faces of SRV.
4
0
10
3
1000s days Boundary-dominated flow characterized by quasisteady flow from the depletion volume into external faces of SRV
Analysis Options • Stretched Exponental Production Decline (SEPD) – Avoids the requirement for key parameters to vary with time but may require a large population of wells in order to constrain parameters effectively.
– Ref. John Lee, Valko, Ilk et al.
• Root Time Methodologies – OK if the well is in transient linear flow but the deviation from linear trend indicates transition to boundary dominated flow.
• Simulation – A large number of input parameters; which are poorly defined and constrained, not least the characteristics of the Stimulated Rock Volume (SRV).
– BUT also in-situ matrix permeability and desorption-diffusion coefficients.
22
Shale Gas Development • Gas
shale developments utilise horizontal wells and multi-stage fracs. • This is repeated many times over a large area. • These developments are CAPEX and human resource intensive. • For success, the process needs to be efficient. • Well planned appraisal and pilot production stages are essential. • Success is NOT guaranteed; so ‘offramps’ need to be clearly defined before moving to a development.
23
Technical Challenges – Five Key Ways to Improve • Prediction of SRV permeability prior to treatment. – Combination of Geomechanics and flowing well behaviour.
• Better understanding of transport physics. – Prediction of Diffusivity and Darcy flow in shales.
• Understanding of Transient Behaviour of a producing well. – Can the geometric attributes of the SRV be defined from flowing rate and pressure data?
• Impact of Liquids on Shale Gas Well Deliverability. – Liquids rich plays are becoming more attractive (as a hedge against low gas prices); how does the physics differ from a pure gas play?
• Water Management. – Shale Gas is a BIG water consumer; and water is becoming a scarce commodity.
24
Questions?
25
