Evaluating Shale Gas Wells Using Rate and Flowing Surface ... Shale Gas... · Rate and Flowing...

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

[email protected]

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

Flow Rate Permeability x Area / Distance

Shale Gas ProductionWhat Made It Possible?

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

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 OGIP

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

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

• 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!)

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


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

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


Early rate profile very “flat” due to decline in FTP from opening choke


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


P/z Ramagost




OGIP = 18 BSCF

Straight line (BDF) reached at Gp < 0.5 BSCF (< 3 mths)

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


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


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


P/z Ramagost




OGIP = 19 BSCF

Workover resulted in contribution from previously “plugged” perforations

Good straight line trend despite multiple choke changes

FMB Method – Multiple ExamplesFMB Method – Multiple Examples

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


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?

PI MethodPI Method

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?

The SearchThe Search

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

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

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


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)


PI versus Cumulative Production

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)




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


Good straight line trend reached after < 0.5 BCF (< 3 mths)


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


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)



Profile impacted by large choke changes

Good straight line trend despite multiple choke changes

Rate increasing!

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)


MULTIPLE WELLS


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

RESERVOIR SIMULATIONK Layout and BHP distributionRESERVOIR SIMULATION

K Layout and BHP distribution

DEHAN15 DEHAN15

DEHAN15

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)

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

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.

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?

Happiness is a Straight

Line!

Evaluating Shale Gas Wells Using Rate and Flowing Surface ... Shale Gas... · Rate and Flowing...

Documents

Transcript of Evaluating Shale Gas Wells Using Rate and Flowing Surface ... Shale Gas... · Rate and Flowing...