Microseismic Clouds: Modeling and Implications
Microseismic Clouds: Modeling and Implications Ian Palmer Higgs-Palmer Technologies 12 September 2012 SPE 154903 Higgs-Palmer Technologies
1
Acknowledge Co-Authors • Zissis Moschovidis, PCM Technology, Tulsa, Oklahoma • Aaron Schaefer, Aetman Engineering, Houston, Texas
Higgs-Palmer Technologies
2
The Revolution: Shale Gas and Shale Oil
Higgs-Palmer Technologies
3
Wow!!
Higgs-Palmer Technologies
4
Years of Shale Gas Resource: North America
Kevin Smith “Unlocking BC’s Shale Gas Potential”, 6th Annual Shale Gas Symp. (Canadian Institute), Calgary, January 2010. Technologies Higgs-Palmer
5
Cost Advantage of Natural Gas Kevin Smith “Unlocking BC’s Shale Gas Potential”, 6th Annual Shale Gas Symp. (Canadian Institute), Calgary, January 2010.
Higgs-Palmer Technologies
6
Gas Price 10-Year Low In January 2012* • NEW YORK — Natural gas prices are rebounding from 10-year lows as producers cut back and colder weather forces homeowners to turn up the heat. • The price of natural gas futures rose Tuesday to finish at $2.55 per 1,000 cubic feet. The futures contract dropped as low as $2.32 on Thursday, the lowest since Feb. 25, 2002. Natural gas rose this week after one of the largest producers, Chesapeake Energy Corp., announced it would slow down production this year. And weather forecasts showed a chilly mix of rain and snow from the Southwest to the Great Lakes. More than half of homeowners use natural gas for heat, and investors are betting they'll need to crank up furnaces as temperatures drop. • The price of natural gas has plunged as drillers expanded their reach in North America, tapping vast underground shale layers that are rich in gas and oil. Supplies in storage have grown well beyond the five-year average for this time of year. Prices are about 44 percent lower than at the same time last year, and experts say U.S. supplies will continue to test the country's storage capacity because of weak demand. Higgs-Palmer Technologies
*CHRIS KAHN, AP, 24 January 2012
7
Frac Equipment Limitations
Beckwith, R. “Hydraulic fracturingHiggs-Palmer – the fuss, the facts, the future”, JPT Dec Technologies 2010, p 34
8
What is Microseismic? • Shear failure on natural fracture or weak plane (next slide) • Part of energy release is in form of a tiny sound wave (microearthquake) • This burst of energy can be picked up by sensitive geophones if not too far away • And several geophones can locate the source of the microseismic burst by triangulation Higgs-Palmer Technologies
9
Shear stress movement large perm increase
Peak-on-peak stops crack from closing completely
Microseismic burst to geophones Higgs-Palmer Technologies
10
Ian Palmer: March 2012
11
Microseismic Qualitative Info • About where a fracture stimulation goes • And how far it extends • What kind of coverage?
Higgs-Palmer Technologies
12
Salehi, I., GTI (RPSEA project) Ian Palmer: March 2012
13
Questions All events in plan view with events scaled as bubble sized by magnitude. Questions: 1. Why are microseismic events large along Well C? 2. Why are there no microseismic events to the northeast? 3. Why does the cloud of microseismic events tend to go towards the southwest? 4. Why is the geometry of the microseismic events different near well A?
Maxwell, SPE 140449
Higgs-Palmer Technologies
14
Events located during an 8-well 143-stage hydraulic fracture operation. Events are colored by fracture stage. A total of 144,057 microseismic events were captured using multiple near-vertical and horizontal sensor arrays.
ESG performed real-time around-the-clock microseismic monitoring of 143 fracture stages over a period of 43 days.
Higgs-Palmer Technologies
15
Microseismic Quantitative Info
NEW !
• Largely ignored but….. • Can give permeability during injection into a fracture network (next slide) • Permeability is surprisingly large (100s of md) • Can give aperture width of fractures in the network • Can decide if proppant can fit into these fractures or not Higgs-Palmer Technologies
16
Fracture network in roadway
Higgs-Palmer Technologies
17
Fracture network in Barnett Shale Primary frac direction (red) roughly N45oE.
Fracture network spacing ≈ 70 ft
King et al, SPE 119896, 2008
Higgs-Palmer Technologies
18
Microseismic Quantitative Info
NEW !
• Largely ignored but….. • Can give permeability during injection into a fracture network (last slide) • Permeability is large (100s of md) but most of this is lost when we turn the well on (> production perm
Seek implications for proppant, etc
Higgs-Palmer Technologies
26
Pattern of microseismic bursts after one-stage stimulation of a horizontal well (left panel). The dotted line outlines the stimulation volume. Match microseismic pattern injection perm in SRV
We assume quasiuniform microseismic distribution
Geometry of flow model (rate transient analysis) to match gas rate versus time (right panel). Match gas rate production perm in SRV
Higgs-Palmer Technologies
27
What we are Finding: in a Nutshell 0.5” Virgin
Fluid flow
5”
flowrate ↑ 1000 x
During injection
flowrate ↑8x
1” During production Ian Palmer: March 2012
28
Case Study: Barnett Shale • Horizontal well in Parker County • Depth ≈ 5,000 ft
Higgs-Palmer Technologies
29
Multi-stage sequential fracs through two wells in Barnett shale. Each color is the microseismic spread from one frac stage in one well.
King et al, SPE 119896, 2008 Higgs-Palmer Technologies
30
Weak planes are random, eff = 81%, Cfr = 0.46, φi = 0.019%, Ki = 229 md.
Shear failure extends out 900 ft
Image from DomAnal Higgs-Palmer Technologies
31
Results • ϕi = 0.019% (this is fracture porosity). Φi has to be small because microseismic spread (associated with frac fluid) is large after pump time of only 3 hours (ie, cannot be matrix porosity). • Ki is 229 md for random orientation of weak planes (this is fracture permeability). Ki has to be large to raise the pore pressure enough to induce failure all the way out to the microseismic envelope. • A large Ki and a small Φi imply fracture-controlled flow (not matrix-controlled flow). Higgs-Palmer Technologies
32
Fracture Network: Aperture and Spacing NEW !
• We have been able to calculate aperture widths (and spacing) in the fracture network (next slide). • Can we use this info to optimize proppant?
Higgs-Palmer Technologies
33
Fracture spacing = ?? and open fracture width = ?? μ Fracture spacing
Fracture width
Fluid injection during frac treatment
Injection porosity = 0.019% and injection permeability = 229 md Higgs-Palmer Technologies
Fracture spacing and width are uniquely determined by injection porosity and permeability
34
Image from DomAnal
Fracture spacing in Parker County network Higgs-Palmer Technologies
35
Fracture opening width in Parker County network
Image from DomAnal
Higgs-Palmer Technologies
36
Implications for Proppant • Where is the proppant? • Is it doing any good? • Is it right-sized? Is it getting into the fractures in the fracture network?
Higgs-Palmer Technologies
37
Is it here??
Is it here??
Cipolla, C. “Fracture modeling issues in unconventional gas reservoirs (tight sands and38 Higgs-Palmer Technologies shales), SPE-ATW, Barossa Valley, Australia, 14 October 2008.
Dominant central fracture
Proppant Distribution Evenly distributed: eg, lightweight proppant
Is it here?? Is it here??
Pillar distribution due to proppant fallout Cipolla, C. “The relationship between fracture complexity, reservoir properties, and fracture Higgs-Palmer Technologies treatment design”, SPE 115769, ATCE, Denver, CO, 21-24 Sept 2008.
39
BUT FIRST: Can the proppant even get into the fractures in the fracture network?
ANSWER: It depends on the size of a proppant grain, and the opening width of the fracture(s) Higgs-Palmer Technologies
40
If we know the fracture opening width…
Means should use more 100mesh and less 40-70 mesh proppant
100–mesh is first stage, followed by 40-70 mesh second stage
Higgs-Palmer Technologies
41
Equation to Determine Proppant %
NEW !
% 100-mesh = 140 - 0.4 x (average fracture width) • where % means (lbs 100-mesh) / (lbs total for 100-mesh + 40-70) • average fracture width (microns) during injection comes from injection permeability and porosity when the geomechanics model is matched to the microseismic pattern.
Average aperture = 168 µ % 100-mesh = 72% Higgs-Palmer Technologies
42
Implications • Most shale-frac treatments use 100-mesh followed by 40-70 mesh proppant. • Our result suggests should use relatively more 100-mesh and less 40-70 proppant. • Can we take this analysis further and tailor the proppant to prop a greater number of fractures in the created fracture network?
Higgs-Palmer Technologies
43
What we have discussed today
ess to work
Proppant Choice Table: Preliminary
Travel distance within network **
Conduct ivity
Durability of resulting fracture
Suitable at elevated temperature?
Suitability in trunk
Cost
Suitability for multiphase flow
Acceptable to screenout or fully pack frac?
Suitability at high stress
Erosivit
M
M
better
Y
M
M
Larger diameter likely needed near wellbore unless continuous voids are sustained
y
HH
M
M
L
medium
N
L
L
Only if continuous voids between pillars are sustained
y
ML
H
MH
LL
medium
N
LL
L
Marginally acceptable even if pillars are sustained
y
M
H
H
L
poor
N
LLL
H
Only if uniform PML achieved and sustained
No
LL
L
HH
LL
poor
N
LLLL
H
No
No
LL
L
L
M
medium
N
M
L
Y if low stress and wide fractures
y
L
H
L
H
better
Y
HH
M
Y
y
H
M
* There are more than 50 mines supplying frac sand, and more than 70 ceramic plants. Quality varies tremendously and buyer should be aware of enormous quality variation with some samples providing less than 20% the conductivity of similarly named proppants. This table assumes top quality material in each category. ** Note that all proppants will fall into a settled bank unless frac geometry interferes with settling, or if “forced closure” procedures capture particles prior to setting. Travel distance is relative, and it should NOT be presumed that even 100 mesh ULWP is neutrally buoyant or will be suspended until fracture closure, unless densified frac fluids are used. [ULWP is not neutrally buoyant in typical frac fluid]. Flowback resistance: Most field data indicate larger proppant diameters are more flowback resistant, and some fields require stronger proppants to avoid proppant flowback. Curable resins and other additives can effectively control proppant flowback.
Conclusions • Matching a cloud of microseismic data with geomechanics model which assumes leakoff of frac fluid causes shear failure • Injection permeability during a frac stimulation is 100s of md and injection porosity is < 0.1% (this is fracturedominated flow) • Assuming a fracture network average fracture spacing ~6 ft and width ~168µ • This width can be used to select % of 100-mesh ahead of 40-70 proppant • Goal is to save more of injection perm: production perm in SRV is < 0.1 md in this case, and size of SRV is much less than microseismic dimensions Higgs-Palmer Technologies
45
Acknowledgements • We thank Xiaowei Weng for reviewing this work and making valuable comments. • Mike Vincent offered many helpful comments on this work. • Funding for this project is provided by RPSEA through the “Ultra-Deepwater and Unconventional Natural Gas and Other Petroleum Resources” program authorized by the U.S. Energy Policy Act of 2005. Higgs-Palmer Technologies
46
Shale gas/oil revolution
THE END
The singularity: 2045
47
1
Acknowledge Co-Authors • Zissis Moschovidis, PCM Technology, Tulsa, Oklahoma • Aaron Schaefer, Aetman Engineering, Houston, Texas
Higgs-Palmer Technologies
2
The Revolution: Shale Gas and Shale Oil
Higgs-Palmer Technologies
3
Wow!!
Higgs-Palmer Technologies
4
Years of Shale Gas Resource: North America
Kevin Smith “Unlocking BC’s Shale Gas Potential”, 6th Annual Shale Gas Symp. (Canadian Institute), Calgary, January 2010. Technologies Higgs-Palmer
5
Cost Advantage of Natural Gas Kevin Smith “Unlocking BC’s Shale Gas Potential”, 6th Annual Shale Gas Symp. (Canadian Institute), Calgary, January 2010.
Higgs-Palmer Technologies
6
Gas Price 10-Year Low In January 2012* • NEW YORK — Natural gas prices are rebounding from 10-year lows as producers cut back and colder weather forces homeowners to turn up the heat. • The price of natural gas futures rose Tuesday to finish at $2.55 per 1,000 cubic feet. The futures contract dropped as low as $2.32 on Thursday, the lowest since Feb. 25, 2002. Natural gas rose this week after one of the largest producers, Chesapeake Energy Corp., announced it would slow down production this year. And weather forecasts showed a chilly mix of rain and snow from the Southwest to the Great Lakes. More than half of homeowners use natural gas for heat, and investors are betting they'll need to crank up furnaces as temperatures drop. • The price of natural gas has plunged as drillers expanded their reach in North America, tapping vast underground shale layers that are rich in gas and oil. Supplies in storage have grown well beyond the five-year average for this time of year. Prices are about 44 percent lower than at the same time last year, and experts say U.S. supplies will continue to test the country's storage capacity because of weak demand. Higgs-Palmer Technologies
*CHRIS KAHN, AP, 24 January 2012
7
Frac Equipment Limitations
Beckwith, R. “Hydraulic fracturingHiggs-Palmer – the fuss, the facts, the future”, JPT Dec Technologies 2010, p 34
8
What is Microseismic? • Shear failure on natural fracture or weak plane (next slide) • Part of energy release is in form of a tiny sound wave (microearthquake) • This burst of energy can be picked up by sensitive geophones if not too far away • And several geophones can locate the source of the microseismic burst by triangulation Higgs-Palmer Technologies
9
Shear stress movement large perm increase
Peak-on-peak stops crack from closing completely
Microseismic burst to geophones Higgs-Palmer Technologies
10
Ian Palmer: March 2012
11
Microseismic Qualitative Info • About where a fracture stimulation goes • And how far it extends • What kind of coverage?
Higgs-Palmer Technologies
12
Salehi, I., GTI (RPSEA project) Ian Palmer: March 2012
13
Questions All events in plan view with events scaled as bubble sized by magnitude. Questions: 1. Why are microseismic events large along Well C? 2. Why are there no microseismic events to the northeast? 3. Why does the cloud of microseismic events tend to go towards the southwest? 4. Why is the geometry of the microseismic events different near well A?
Maxwell, SPE 140449
Higgs-Palmer Technologies
14
Events located during an 8-well 143-stage hydraulic fracture operation. Events are colored by fracture stage. A total of 144,057 microseismic events were captured using multiple near-vertical and horizontal sensor arrays.
ESG performed real-time around-the-clock microseismic monitoring of 143 fracture stages over a period of 43 days.
Higgs-Palmer Technologies
15
Microseismic Quantitative Info
NEW !
• Largely ignored but….. • Can give permeability during injection into a fracture network (next slide) • Permeability is surprisingly large (100s of md) • Can give aperture width of fractures in the network • Can decide if proppant can fit into these fractures or not Higgs-Palmer Technologies
16
Fracture network in roadway
Higgs-Palmer Technologies
17
Fracture network in Barnett Shale Primary frac direction (red) roughly N45oE.
Fracture network spacing ≈ 70 ft
King et al, SPE 119896, 2008
Higgs-Palmer Technologies
18
Microseismic Quantitative Info
NEW !
• Largely ignored but….. • Can give permeability during injection into a fracture network (last slide) • Permeability is large (100s of md) but most of this is lost when we turn the well on (> production perm
Seek implications for proppant, etc
Higgs-Palmer Technologies
26
Pattern of microseismic bursts after one-stage stimulation of a horizontal well (left panel). The dotted line outlines the stimulation volume. Match microseismic pattern injection perm in SRV
We assume quasiuniform microseismic distribution
Geometry of flow model (rate transient analysis) to match gas rate versus time (right panel). Match gas rate production perm in SRV
Higgs-Palmer Technologies
27
What we are Finding: in a Nutshell 0.5” Virgin
Fluid flow
5”
flowrate ↑ 1000 x
During injection
flowrate ↑8x
1” During production Ian Palmer: March 2012
28
Case Study: Barnett Shale • Horizontal well in Parker County • Depth ≈ 5,000 ft
Higgs-Palmer Technologies
29
Multi-stage sequential fracs through two wells in Barnett shale. Each color is the microseismic spread from one frac stage in one well.
King et al, SPE 119896, 2008 Higgs-Palmer Technologies
30
Weak planes are random, eff = 81%, Cfr = 0.46, φi = 0.019%, Ki = 229 md.
Shear failure extends out 900 ft
Image from DomAnal Higgs-Palmer Technologies
31
Results • ϕi = 0.019% (this is fracture porosity). Φi has to be small because microseismic spread (associated with frac fluid) is large after pump time of only 3 hours (ie, cannot be matrix porosity). • Ki is 229 md for random orientation of weak planes (this is fracture permeability). Ki has to be large to raise the pore pressure enough to induce failure all the way out to the microseismic envelope. • A large Ki and a small Φi imply fracture-controlled flow (not matrix-controlled flow). Higgs-Palmer Technologies
32
Fracture Network: Aperture and Spacing NEW !
• We have been able to calculate aperture widths (and spacing) in the fracture network (next slide). • Can we use this info to optimize proppant?
Higgs-Palmer Technologies
33
Fracture spacing = ?? and open fracture width = ?? μ Fracture spacing
Fracture width
Fluid injection during frac treatment
Injection porosity = 0.019% and injection permeability = 229 md Higgs-Palmer Technologies
Fracture spacing and width are uniquely determined by injection porosity and permeability
34
Image from DomAnal
Fracture spacing in Parker County network Higgs-Palmer Technologies
35
Fracture opening width in Parker County network
Image from DomAnal
Higgs-Palmer Technologies
36
Implications for Proppant • Where is the proppant? • Is it doing any good? • Is it right-sized? Is it getting into the fractures in the fracture network?
Higgs-Palmer Technologies
37
Is it here??
Is it here??
Cipolla, C. “Fracture modeling issues in unconventional gas reservoirs (tight sands and38 Higgs-Palmer Technologies shales), SPE-ATW, Barossa Valley, Australia, 14 October 2008.
Dominant central fracture
Proppant Distribution Evenly distributed: eg, lightweight proppant
Is it here?? Is it here??
Pillar distribution due to proppant fallout Cipolla, C. “The relationship between fracture complexity, reservoir properties, and fracture Higgs-Palmer Technologies treatment design”, SPE 115769, ATCE, Denver, CO, 21-24 Sept 2008.
39
BUT FIRST: Can the proppant even get into the fractures in the fracture network?
ANSWER: It depends on the size of a proppant grain, and the opening width of the fracture(s) Higgs-Palmer Technologies
40
If we know the fracture opening width…
Means should use more 100mesh and less 40-70 mesh proppant
100–mesh is first stage, followed by 40-70 mesh second stage
Higgs-Palmer Technologies
41
Equation to Determine Proppant %
NEW !
% 100-mesh = 140 - 0.4 x (average fracture width) • where % means (lbs 100-mesh) / (lbs total for 100-mesh + 40-70) • average fracture width (microns) during injection comes from injection permeability and porosity when the geomechanics model is matched to the microseismic pattern.
Average aperture = 168 µ % 100-mesh = 72% Higgs-Palmer Technologies
42
Implications • Most shale-frac treatments use 100-mesh followed by 40-70 mesh proppant. • Our result suggests should use relatively more 100-mesh and less 40-70 proppant. • Can we take this analysis further and tailor the proppant to prop a greater number of fractures in the created fracture network?
Higgs-Palmer Technologies
43
What we have discussed today
ess to work
Proppant Choice Table: Preliminary
Travel distance within network **
Conduct ivity
Durability of resulting fracture
Suitable at elevated temperature?
Suitability in trunk
Cost
Suitability for multiphase flow
Acceptable to screenout or fully pack frac?
Suitability at high stress
Erosivit
M
M
better
Y
M
M
Larger diameter likely needed near wellbore unless continuous voids are sustained
y
HH
M
M
L
medium
N
L
L
Only if continuous voids between pillars are sustained
y
ML
H
MH
LL
medium
N
LL
L
Marginally acceptable even if pillars are sustained
y
M
H
H
L
poor
N
LLL
H
Only if uniform PML achieved and sustained
No
LL
L
HH
LL
poor
N
LLLL
H
No
No
LL
L
L
M
medium
N
M
L
Y if low stress and wide fractures
y
L
H
L
H
better
Y
HH
M
Y
y
H
M
* There are more than 50 mines supplying frac sand, and more than 70 ceramic plants. Quality varies tremendously and buyer should be aware of enormous quality variation with some samples providing less than 20% the conductivity of similarly named proppants. This table assumes top quality material in each category. ** Note that all proppants will fall into a settled bank unless frac geometry interferes with settling, or if “forced closure” procedures capture particles prior to setting. Travel distance is relative, and it should NOT be presumed that even 100 mesh ULWP is neutrally buoyant or will be suspended until fracture closure, unless densified frac fluids are used. [ULWP is not neutrally buoyant in typical frac fluid]. Flowback resistance: Most field data indicate larger proppant diameters are more flowback resistant, and some fields require stronger proppants to avoid proppant flowback. Curable resins and other additives can effectively control proppant flowback.
Conclusions • Matching a cloud of microseismic data with geomechanics model which assumes leakoff of frac fluid causes shear failure • Injection permeability during a frac stimulation is 100s of md and injection porosity is < 0.1% (this is fracturedominated flow) • Assuming a fracture network average fracture spacing ~6 ft and width ~168µ • This width can be used to select % of 100-mesh ahead of 40-70 proppant • Goal is to save more of injection perm: production perm in SRV is < 0.1 md in this case, and size of SRV is much less than microseismic dimensions Higgs-Palmer Technologies
45
Acknowledgements • We thank Xiaowei Weng for reviewing this work and making valuable comments. • Mike Vincent offered many helpful comments on this work. • Funding for this project is provided by RPSEA through the “Ultra-Deepwater and Unconventional Natural Gas and Other Petroleum Resources” program authorized by the U.S. Energy Policy Act of 2005. Higgs-Palmer Technologies
46
Shale gas/oil revolution
THE END
The singularity: 2045
47
