evaluating shale gas wells using rate and flowing surface ... shale gas... · rate and flowing...
Post on 08-Mar-2018
216 Views
Preview:
TRANSCRIPT
Evaluating Shale Gas Wells UsingRate and Flowing Surface Pressure Data
Evaluating Shale Gas Wells UsingRate and Flowing Surface Pressure Data
February 6, 2013February 6, 2013
ctucker@nsai-petro.com214-9695401
OUTLINEOUTLINE
1. Shale Gas Production - What made it possible?
2. Flowing Material Balance - Estimating Contacted OGIP
3. PI Method - Estimating Recoverable Gas
1. Shale Gas Production - What made it possible?
2. Flowing Material Balance - Estimating Contacted OGIP
3. PI Method - Estimating Recoverable Gas
Page 2
Flow Rate Permeability x Area / Distance
Shale Gas ProductionWhat Made It Possible?
Page 3
Conventional ReservoirGood K
Small Area = rw hLarge Distance
Radial Flow
Tight Gas ReservoirLow K
Increased Area = 2LhLarge Distance
Linear Transient Flow
Shale Gas Horiz Multi-Stage FracEnhanced K in SRV
Significantly Increased AreaSignificantly Decreased Distance
Complex Flow with Early BDF
SPE 136696 and 145080
SPE 7490
So which of these three parameters did the industry work on? How about ALL THREE!
Flowing Material BalanceFlowing Material Balance
Page 4
Flowing Material Balance (FMB)Flowing Material Balance (FMB)
WHAT IS IT?• Dynamic (performance-based) estimate of contacted OGIPHOW DO WE DO IT?• Estimate FBHP from rate and flowing tubing pressure• Estimate SIBHP from FBHP, rate and productivity index (PI)• Perform gas material balance
DOES IT WORK?• Have seen very “linear” FMB trends across numerous wells with
as little as 3 months of data for the Haynesville (little longer for the shallower plays)
• Estimated connected OGIP volumes appear reasonable• Estimated RF’s are typically 50-60% of contacted OGIP
WHAT IS IT?• Dynamic (performance-based) estimate of contacted OGIPHOW DO WE DO IT?• Estimate FBHP from rate and flowing tubing pressure• Estimate SIBHP from FBHP, rate and productivity index (PI)• Perform gas material balance
DOES IT WORK?• Have seen very “linear” FMB trends across numerous wells with
as little as 3 months of data for the Haynesville (little longer for the shallower plays)
• Estimated connected OGIP volumes appear reasonable• Estimated RF’s are typically 50-60% of contacted OGIPPage 5
FMB – What is it?FMB – What is it?
0
2
4
6
8
10
12
14
16
18
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 1 2 3 4
Gas Rate (M
MSCFD
)
P/Z and FBHP
(PSIA)
Cumulative Gas (BSCF)
FMB Illustration
P/Z
FBHP
Rate
Page 6
SIBHP estimated from
• Rate
• FTP, FBHP
• PI
Stress Dependent KStress Dependent K
0.0010
0.0100
0.1000
1.0000
0 2,000 4,000 6,000 8,000 10,000 12,000
K/Ki
BHP (PSIA)
Stress Dependent Permeability Relationship
Yilmaz and Nur
Initial BHP =11,000 psia
K/Ki = 1.0
Page 7
• FMB analysis requires a PI model• A stress-dependent decline in K is used for shale wells• Stress dependency must be determined independent of
FMB analysis – need SIBHP from subject well or analog
• FMB analysis requires a PI model• A stress-dependent decline in K is used for shale wells• Stress dependency must be determined independent of
FMB analysis – need SIBHP from subject well or analog
Rate and FTP vs Time – Example 1Rate and FTP vs Time – Example 1
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FTP
(PS
IA)
Gas
Rat
e (M
MSC
FD)
Months
Gas Rate FTP
Historical Gas Rate and FTP
Rate is constrained by high FTP Good pressure build-up indicating
SRV effective K is 0.1 md (not nanodarcies!)
Page 8
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 2 4 6 8 10 12 14 16 18 20
Nor
mal
ized
PI
P/Z
(psi
a) w
ith R
amag
ost C
orre
ctio
n
Cumulative Gas (BSCF)
P/z Ramagost
P/z Ramagost - Calculated from PI
Stress Normalized PI
Flowing Material Balance Analysis
FMB Method – Example 1FMB Method – Example 1
OGIP = 15 BSCF
Plotted points are NOT dependent on OGIP. They point to OGIP!
Solution algorithm is iterative.
Normalized PI data plot on straight line
Page 9
SIBHP survey confirmed OGIP trend line from FMB
Rate and FTP vs Time – Example 2Rate and FTP vs Time – Example 2
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20
FTP
(PSI
A)
Gas
Rat
e (M
MS
CFD
)
Months
Gas Rate FTP
Historical Gas Rate and FTP
Early rate profile very “flat” due to decline in FTP from opening choke
Page 10
FMB Method – Example 2FMB Method – Example 2
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 2 4 6 8 10 12 14 16 18 20
Nor
mal
ized
PI
P/Z
(psi
a) w
ith R
amag
ost C
orre
ctio
n
Cumulative Gas (BSCF)
P/z Ramagost
P/z Ramagost - Calculated from PI
Stress Normalized PI
Flowing Material Balance Analysis
OGIP = 18 BSCF
Straight line (BDF) reached at Gp < 0.5 BSCF (< 3 mths)
Page 11
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FTP
(PSI
A)
Gas
Rat
e (M
MS
CFD
)
Months
Gas Rate FTP
Historical Gas Rate and FTP
Rate and Pressure – Example 3Rate and Pressure – Example 3
Profile impacted by multiple large choke changes.
Would the method still work?
Rate is increasing or flat
Page 12
FMB Method – Example 3FMB Method – Example 3
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 5 10 15 20 25
Nor
mal
ized
PI
P/Z
(psi
a) w
ith R
amag
ost C
orre
ctio
n
Cumulative Gas (BSCF)
P/z Ramagost
P/z Ramagost - Calculated from PI
Stress Normalized PI
Flowing Material Balance Analysis
OGIP = 19 BSCF
Workover resulted in contribution from previously “plugged” perforations
Good straight line trend despite multiple choke changes
Page 13
FMB Method – Multiple ExamplesFMB Method – Multiple Examples
Page 14
2000
3000
4000
5000
6000
7000
8000
9000
0 1 2 3 4 5 6 7 8
P/Z
(psi
a) w
ith R
amag
ost C
orre
ctio
n
Cumulative Gas (BSCF)
MULTIPLE WELLSFMB Analysis
Observations/Questions?Observations/Questions?
• FMB method seems to work across multiple shale plays including Haynesville, Eagleford, Niobrara and Marcellus
• Is it a surprise the method still works with shallow, normally pressured shale plays?
• FMB trends have thus far remained straight suggesting GIIP is not increasing Negligible contribution from outside of SRV (so far) Free GIIP appears to be key, especially in early years which dominate
recovery and NPV
• Will the lines stay straight?• Will there be material contribution from outside SRV?
• FMB method seems to work across multiple shale plays including Haynesville, Eagleford, Niobrara and Marcellus
• Is it a surprise the method still works with shallow, normally pressured shale plays?
• FMB trends have thus far remained straight suggesting GIIP is not increasing Negligible contribution from outside of SRV (so far) Free GIIP appears to be key, especially in early years which dominate
recovery and NPV
• Will the lines stay straight?• Will there be material contribution from outside SRV?
Page 15
PI MethodPI Method
Page 16
The MotivationThe Motivation
• Shale gas wells are often rate-constrained during the early years of production
• Production rates can sometimes be near constant or even increasing as a result of choke changes
• Decline curve analysis methods that rely only on rate data are not reliable in such circumstances
• Thousands of US shale wells need to be analyzed each year. We wanted a simple forecast method that gave reliable estimates.
Was there a straight line out there somewhere?
• Shale gas wells are often rate-constrained during the early years of production
• Production rates can sometimes be near constant or even increasing as a result of choke changes
• Decline curve analysis methods that rely only on rate data are not reliable in such circumstances
• Thousands of US shale wells need to be analyzed each year. We wanted a simple forecast method that gave reliable estimates.
Was there a straight line out there somewhere?
Page 17
The SearchThe Search
Page 18
STRAIGHT LINESSheep, Engineers and Pseudo-Scientist’s
STRAIGHT LINESSheep, Engineers and Pseudo-Scientist’s
“Lets apply a little science to understand why the line is straight so we can more confidently predict the future and avoid any cliffs” Scott Rees – NSAI CEO, March 2011
Page 19
The PI MethodThe PI Method
Why?• Restricted rate and variable choke settings prohibit DCA• Productivity Index (PI) trends are predictable• Can address future operations (e.g. line pressure impact)• Diagnostic of problems/issues (e.g. offset interference, well damage etc)• EUR can be observed directly from the plot
How?• Convert FTP to FBHP using tubing flow equation• Calculate PVT properties including Pseudo Pressure m(p)• Calculate PI as follows
PI = Qgas / [m(pi) – m(pwf)]• Plot Log(PI) versus Gp
Why?• Restricted rate and variable choke settings prohibit DCA• Productivity Index (PI) trends are predictable• Can address future operations (e.g. line pressure impact)• Diagnostic of problems/issues (e.g. offset interference, well damage etc)• EUR can be observed directly from the plot
How?• Convert FTP to FBHP using tubing flow equation• Calculate PVT properties including Pseudo Pressure m(p)• Calculate PI as follows
PI = Qgas / [m(pi) – m(pwf)]• Plot Log(PI) versus Gp
Page 20
PI Method – Example 1PI Method – Example 1
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FTP
(PSI
A)
Gas
Rat
e (M
MSC
FD)
Months
Gas Rate FTP
Historical Gas Rate and FTP
0.00001
0.00010
0.00100
0.01000
0.10000
0 1 2 3 4 5 6 7 8
PI (
MM
SC
FD /
(10^
6 ps
ia^2
/cp)
Cumulative Gas (BSCF)
PI versus Cumulative Production
Page 21
EUR can be “seen” from plot
Transient increase in PI following shut-in
Rate restricted by high FTP
0.00001
0.00010
0.00100
0.01000
0.10000
0 1 2 3 4 5 6
PI (
MM
SC
FD /
(10^
6 ps
ia^2
/cp)
Cumulative Gas (BSCF)
PI versus Cumulative Production
PI Method – Example 2PI Method – Example 2
Page 22
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20
FTP
(PSI
A)
Gas
Rat
e (M
MSC
FD)
Months
Gas Rate FTP
Historical Gas Rate and FTP
Good straight line trend reached after < 0.5 BCF (< 3 mths)
PI Method – Example 3PI Method – Example 3
Page 23
0
2000
4000
6000
8000
10000
12000
1
10
100
1000
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FTP
(PSI
A)
Gas
Rat
e (M
MSC
FD)
Months
Gas Rate FTP
Historical Gas Rate and FTP
0.00001
0.00010
0.00100
0.01000
0.10000
0 2 4 6 8 10 12
PI (
MM
SC
FD /
(10^
6 ps
ia^2
/cp)
Cumulative Gas (BSCF)
PI versus Cumulative Production
Profile impacted by large choke changes
Good straight line trend despite multiple choke changes
Rate increasing!
Page 24
PI Method – Multiple Examples
0.00001
0.00010
0.00100
0.01000
0.10000
0 1 2 3 4 5 6 7 8
PI (M
MS
CFD
/ (1
0^6
psia
^2/c
p)
Cumulative Gas (BSCF)
MULTIPLE WELLS
PI versus Cumulative Production
Page 25
Rate vs Time Forecast – Example 3
0.00001
0.00010
0.00100
0.01000
0.10000
1.00000
0.01
0.10
1.00
10.00
100.00
1000.00
0 10 20 30 40 50 60 70 80 90 100
PI (
MM
SC
FD /1
0^6
psi
a^2
/cp)
Gas
Rat
e (M
MS
CFD
)
Qgas (history) PI (history) PI(forecast)
PI forecast
Gas forecast
Historical and Forecast Gas Rate and PI
FBHP forecast
Min. FBHP (line conditions)
Reservoir SimulationReservoir Simulation
Page 26
RESERVOIR SIMULATIONK Layout and BHP distributionRESERVOIR SIMULATION
K Layout and BHP distribution
DEHAN15 DEHAN15
DEHAN15
Page 27
RESERVOIR SIMULATIONBHP Match – Haynesville WellRESERVOIR SIMULATION
BHP Match – Haynesville WellHAYNESVILLE WELL
ECLIPSE Simulation History Match
5,000
6,000
7,000
8,000
9,000
10,000
11,000
12,000
Feb-10 Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10
BHP
(psi
a) a
nd G
as
Rat
e (M
cfd)
Gas Rate
BHP - Actual
BHP - Simulated (NSAI)
Page 28
RESERVOIR SIMULATIONPressure Build-up Match – Haynesville Well
RESERVOIR SIMULATIONPressure Build-up Match – Haynesville Well
BRIARWOOD 35-1ECLIPSE MATCH OF SIBHP SURVEY
9,300
9,400
9,500
9,600
9,700
9,800
9,900
10,000
10,100
10,200
1.0 10.0 100.0
Shut-in Time (days)
SIBH
P (p
sia)
ActualSimulated
Page 29
0
1
2
3
4
5
6
7
8
9
10
0.01
0.1
1
10
J-08 J-10 J-12 J-14 J-16 J-18 J-20 J-22 J-24 J-26 J-28 J-30 J-32 J-34 J-36 J-38 J-40 J-42 J-44 J-46 J-48 J-50 J-52 J-54 J-56 J-58 J-60
CU
M G
AS (B
CF)
GA
S R
ATE
(MM
CFD
)Gas Rate and Cumulative Gas
ECLIPSE versus PI Method
Actual
ECLIPSE
PI Model
RESERVOIR SIMULATIONRate Forecast – Haynesville Well
RESERVOIR SIMULATIONRate Forecast – Haynesville Well
ECLIPSE and PI model are in good agreement.
Page 30
ObservationsObservations
• PI method seems to work across multiple shale plays including Haynesville, Eagleford, Niobrara and Marcellus
• Method advantages Incorporates both rate and pressure data Simple to implement - only need Rate, FTP and Initial BHP data Provides EUR estimates with limited production history Easy to convert to rate versus time forecast
Will the lines stay straight?
• PI method seems to work across multiple shale plays including Haynesville, Eagleford, Niobrara and Marcellus
• Method advantages Incorporates both rate and pressure data Simple to implement - only need Rate, FTP and Initial BHP data Provides EUR estimates with limited production history Easy to convert to rate versus time forecast
Will the lines stay straight?
Page 31
Happiness is a Straight
Line!
Page 32
top related