REDESIGNING COALBED METHANE RESERVOIR · PDF fileREDESIGNING COALBED METHANE RESERVOIR ENGINEERING: “a reflective analysis”! Turgay Ertekin Penn State University Fall

REDESIGNING COALBED METHANE RESERVOIR ENGINEERING: “a reflective analysis”

Turgay Ertekin Penn State University

Fall 2014 Energy Exchange Seminar Series January 22, 2014

§ 

A MULTIPLE CHOICE QUESTION: “REVISITING THE RESERVOIR ENGINEERING SCHOOL FOR COALDBED METHANE PRODUCTION” IS NECESSARY, BECAUSE:

AS AN ACADEMICIAN, I NEED TO CONDUCT ACADEMIC EXERCISES

EXISTING PUBLISH OR PERISH SYNDROME IN ACADEMIA DEMANDS IT

OTHERWISE, I WOULD NOT BE ABLE TO MAKE THIS PRESENTATION

MY FATHER OWNS COAL PROPERTIES AND WOULD LIKE TO MAXIMIZE HIS PROFIT NONE OF THE ABOVE

§ 

§ 

§ 

§ 

SOME ALTERNATIVE ANSWERS . . .

THE PHYSICS OF FLOW IN A COALBED METHANE RESERVOIR DIFFERS FROM ITS CONVENTIONAL PARTS PRODUCING COALBED METHANE AHEAD OF MINING DECREASES METHANE EMISSIONS TO ATMOSPHERE PRODUCING COALBED METHANE CONTRIBUTES TO NATION’S ENERGY BUDGET SIGNIFICANTLY PRODUCING COALBED METHANE AHEAD OF MINING OF COAL INCREASES MINE SAFETY AND MINING EFFICIENCY UNMINEABLE COALSEAMS CAN PROVIDE POTENTIAL CO2 SEQUESTRATION SITES

* P* P* P* P* P

ACKNOWLEDGEMENTS §  TRW §  U.S.STEEL §  AMOCO §  GRI §  DOE/NETL §  MERIDIAN OIL

DIVERSITY IN COALBED METHANE TECHNOLOGY

§  EXPLORATION GEOLOGISTS §  DRILLING AND COMPLETION ENGINEERS §  LOG ANALYSTS §  STIMULATION ENGINEERS §  PRODUCTION ENGINEERS §  ECONOMISTS §  LEGAL AND JUDICIAL BODIES

ALONG WITH RESERVOIR ENGINEERING, COALBED METHANE PRODUCTION PRESENTS A SPECIAL CHALLENGE TO:

REVISITING THE COALBED METHANE RESERVOIRS

§  HOW METHANE IS STORED IN COAL?

§  HOW CAN THIS STORED METHANE BE RELEASED EFFICIENTLY?

§  AND ONCE IT IS RELEASED, HOW DOES THE METHANE FLOW TO THE WELL?

IT IS NECESSARY TO UNDERSTAND:

PENN STATE STRATEGY

? ? ?

? ?

? ? ? ?

? ? ?

? ?

? ? ? ?

HOW DID WE FEEL AT THE ONSET?

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS

FIRST-GENERATION NUMERICAL

MODELS

SECOND- GENERATION NUMERICAL

MODELS

TYPE CURVES FOR

PRODUCTION ANALYSIS

PRESSURE TRANSIENT ANALYSIS

PROCEDURES THIRD-

GENERATION MODELS

EVOLUTION OF COALBED METHANE RESEARCH AT PENN STATE

PURE BASIC RESEARCH

1975-85

USE INSPIRED BASIC RESEARCH

1985-2000

PURE APPLIED RESEARCH

2000-

RANDOM WALK RESEARCH

YES

YES

NO

NO

CONSIDERATION OF USE

QU

EST

FOR

UN

DER

STA

ND

ING

PENN STATE STRATEGY


MODELS

SECOND- GENERATION NUMERICAL MODELS

TYPE CURVES FOR

PRODUCTION ANALYSIS


PROCEDURES

1979-1985

ESTABLISHING FLOW

MECHANISMS

ESTABLISHING FLOW MECHANISMS

DESORPTION DIFFUSION CONVECTION

VE

P

HENRY’S ISOTHERM gHE pVV =

LANGMUIR’S ISOTHERM gL

gLE

pppVV+

=

FREUNDLICHS ISOTHERM ngfE pVV =

DESORPTION KINETICS

DIFFUSION

§  GAS ADSORBED IS ONLY A FUNCTION OF PRESSURE IN THE CLEAT NETWORK

§  GAS DESORPTION IS INSTANTANEOUS

§  GENERALLY PREDICTS OPTIMISTIC RESULTS

EQUILIBRIUM SORPTION MODELS:

§  FREE AND ADSORBED GAS COMPOSITIONS ARE IDENTICAL, SELECTIVE ADSORPTION AND DESORPTION DO NOT OCCUR

§  GAS TRANSPORT IN THE MICROPORES IS A DIFFUSION PROCESS

§  SURFACE DESORPTION IS INSTANTANEOUS

NON-EQUILIBRIUM SORPTION MODELS:

NON-EQUILIBRIUM SORPTION MODELS

PSEUDO-STEADY STATE " Similar to Warren and Root Approach

" Relatively small computational work

" Good for long term predictions ] i ( ) [ p V V

1

dt dV

E i - - = λ

UNSTEADY STATE " Most rigorous models (de Swaan’s approach)

" Require additional computational work

" Good for all situations including short term

MATRIX

CLEAT NETWORK

DIFFUSION

SLAB ELEMENTS

MATRIX

CLEAT NETWORK

DIFFUSION

CYLINDRICAL ELEMENTS

MATRIX

CLEAT NETWORK

DIFFUSION

SPHERICAL ELEMENTS

FACE CLEAT

BUTT CLEAT

CONVECTION

ESTABLISHING THE FLOW MECHANISM

§  POTENTIAL FIELD §  MACROSCOPIC - DARCIAN FLOW

§  CONCENTRATION FIELD §  RANDOM MOLECULAR - FICKIAN FLOW

IN TIGHT FORMATIONS TWO FLOW FIELDS CONTROL THE GAS FLOW DYNAMICS :

p1 < p2 r1 = r2

p1

r1

p2

r2

p1 < p2 r1 < r2

p2

r2

p1

r1

MULTIMECHANISTIC FLOW

§  IN DOUBLE-‐POROSITY, SINGLE-‐PERMEABILITY SYSTEMS,

MULTIMECHANISTIC FLOW DOMINATES WHEN 10-5 md < k < 10-1 md

IS MULTIMECHANISTIC FLOW IMPORTANT?

DIFFUSION FLOW

DOMINATES

DARCIAN FLOW

DOMINATES

MULTIMECHANISTIC FLOW

§  DRIVE MECHANISMS EXERTED BY THE PRESSURE AND CONCENTRATION FIELDS ARE ACTING IN PARALLEL

§  FLOW THROUGH PRESSURE FIELD OBEYS DARCY’S LAW

§  FLOW THROUGH CONCENTRATION FIELD OBEYS FICK’S LAW

§  NO CHROMATOGRAPHIC SEPARATION OF GAS TAKES PLACE DUE TO INDIVIDUAL DIFFUSIVITIES OF THE GAS CONSTITUENTS

CHARACTERISTICS:

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS


MODELS


TYPE CURVES FOR

PRODUCTION ANALYSIS


PROCEDURES

1979-1985

1980-1987

FIRST GENERATION NUMERICAL MODELS

§  SINGLE-COMPONENT GAS

§  SINGLE-PHASE OR TWO-PHASE

§  DOUBLE-POROSITY, SINGLE-PERMEABILITY

BASIC CHARACTERISTICS:

SINGLE-WELL, ONE-PHASE SINGLE-WELL, TWO-PHASE

MULTI-WELL, TWO-PHASE MULTI-WELL, TWO-PHASE WITH MINING ACTIVITY

SINGLE-WELL MODELS

CYLINDRICAL COORDINATES

ELLIPTICAL COORDINATES

STIMULATED VERTICAL WELL

HORIZONTAL WELLS DRILLED FROM A SHAFT BOTTOM

STIMULATED VERTICAL WELL

MULTI-WELL MODELS

HORIZONTAL BOREHOLE

IMPERMEABLE BARRIER

FRACTURED WELL

UNSTIMULATED WELL

MODEL PERFORMANCE

0

5

10

15

20

25

30

35

0 100 300 500 700 900 1100 1300TIME, DAYS

GA

S D

ES

OR

PT

ION

RA

TE

, M

SC

F/D

AY

GA

S P

RO

DU

CT

ION

RA

TE

, M

SC

F/D

AY

0510152025303540

WA

TE

R P

RO

DU

CT

ION

R

AT

E, B

BL

/DA

Y

0

500

1000

1500

2000

0 200 400 600 800

TIME, DAYS

PRO

DU

CED

GA

S, M

SCF ACTUAL WELL HISTORY

UNCONVENTIONAL GAS RESERVOIR MODELING

CONVENTIONAL GAS RESERVOIR MODELING

HISTORY MATCHING

PREDICTION

MARY LEE COAL SEAM

MODEL PERFORMANCE

MODEL PERFORMANCE

APPROACH

0 %

2.0 md

3.1 %

30 %

UNCONVENTIONAL GAS RESERVOIR

CONVENTIONAL GAS RESERVOIR

APPROACH

60 %

9.0 md

82 %

0.1 %

INITIAL GAS SATURATION

MACROPORE PERMEABILITY

POROSITY

CRITICAL WATER SATURATION

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS


TYPE CURVES FOR

PRODUCTION ANALYSIS

1979-1985

1987-1991


MODELS

1980-1987


PROCEDURES


I O S

DIRECT PROBLEM

INVERSE PROBLEM

I x S O O I S

q

t

PI

Pwf

t

Coal Seam Conventional Reservoir

COAL SEAMS VS CONVENTIONAL RESERVOIRS



§  DESORPTION §  DUAL POROSITY NATURE

§  DEPENDENCE OF PERMEABILITY ON PRESSURE

§  DIFFUSIONAL & LAMINAR FLOW

SHORTFALLS OF CLASSICAL WELL TEST MODEL FOR COAL SEAMS:


§  DIMENSIONLESS GROUPS §  LAPLACE TRANSFORMATION

§  INVERSION §  NUMERICAL §  APPROXIMATE ANALYTICAL

SOLUTION PROCEDURE :

SUMMARY OF SOLUTIONS DEVELOPED


PSEUDO STEADY-STATE SORPTION/DIFFUSION

UNSTEADY-STATE SORPTION/DIFFUSION

NUMERICAL INVERSION

APPROX. ANALYTICAL INVERSION

NUMERICAL INVERSION

FOR 6 POSSIBLE COMBINATIONS OF BOUNDARY CONDITIONS


Pre

ssur

e, p

si2

Log (time), hr ACTUAL PREDICTED

k 20.00 mD 20.03 mD

φ 0.020 0.027

τ 100 hr 90 hr

V L 14.0 scf/cf 15.5 scf/cf

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS


1979-1985

1988-1992


MODELS

1980-1987

TYPE CURVES FOR

PRODUCTION ANALYSIS


PROCEDURES 1987-1991

DECLINE CURVE ANALYSIS

§  THE DEVELOPED DECLINE CURVES PROVIDE A PRACTICAL TOOL WHICH CAN BE USED IN PREDICTING THE PERFORMANCE OF COAL SEAMS AND THEIR CHARACTERISTICS.

PATTERN RECOGNITION:

DEVELOPMENT OF TYPE CURVES

§ EXPRESS TWO-‐PHASE EQUATIONS IN THE FORM OF A SINGLE RELATIONSHIP

§ PUT THE EXPRESSION INTO A DIMENSIONLESS FORM

§ GENERATE TYPE CURVES USING A NUMERICAL MODEL

0.01

0.1

1

1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10

Dim

ensi

onle

ss G

as F

low

Rat

e

Dimensionless Time

wrr

0.2pp

70%S

i

wf

gi

=

=

TYPE CURVE FOR GAS FLOWRATE

0.01

0.1

1 1.0E+04 1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09 1.0E+10

Dimensionless Time

Dim

ensi

onle

ss G

as F

low

Rat

e

wrr

0.2pp

50%S

i

wf

gi

=

=

1.0E+05

1.0E+06

1.0E+07

0.01 0.1 1 10 100 1000

Gas

Flo

w R

ate

(SC

F/D

)

Time (Days)

TYPE CURVE MATCHING

DECLINE CURVE ANALYSIS

EXAMPLE TYPE CURVE % CHANGE

p (psi) 435.0 500.0 13.0 pwf (psi) 97.0 100.0 3.0 h (ft) 15.5 10.0 55.0 Sgi (%) 30.0 30.0 0.0 VL (cc/gm) 8.35 13.70 39.1 PL (atm) 10.24 11.40 10.2 t (sec) 5.63 x 105 2.0 x 107 97.2

TYPE CURVE MATCHING:

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS

1979-1985

1988-1992


MODELS

1980-1987

TYPE CURVES FOR

PRODUCTION ANALYSIS




MODELS

1988-1999

SECOND GENERATION NUMERICAL MODELS

§  MULTI-COMPONENT GAS

§  DOUBLE-POROSITY, DOUBLE-PERMEABILITY

BASIC CHARACTERISTICS:

MULTI-‐COMPONENT GAS MULTI-‐MECHANISTIC FLOW

MULTI-WELL

COMPOSITIONAL MODELING

§  ADSORPTION/DESORPTION ISOTHERMS OF MULTICOMPONENT GAS MIXTURES ARE STRONGLY DEPENDENT ON GAS COMPOSITION

§  COMPOSITION OF THE ADSORBED GAS IS SIGNIFICANTLY DIFFERENT FROM THAT OF THE FREE GAS

IS THERE A WIDE SPECTRUM OF COMPONENTS ? ARE COMPOSITIONAL PHENOMENA DOMINANT ?

COMPOSITIONAL MODELING THERMODYNAMICS OF MIXED-GAS ADSORPTION IS

ANALOGOUS TO VAPOR-LIQUID EQUILIBRIA:

VAPOR

LIQUID

FREE GAS

ADSORBED GAS

COMPOSITIONAL MODELING

CALCULATION OF MIXED-‐GAS ADSORPTION FROM SINGLE-‐GAS ISOTHERMS:

CO2

C2H6

CH4 N2

PRESSURE PRESSURE

ADSORBED

VOLUME

ADSORBED

VOLUME

PURE COMPONENT ISOTHERM

MULTI-COMPONENT ISOTHERM

TOTH ALGORITHMS

UNILAN

PENN STATE STRATEGY

ESTABLISHING FLOW

MECHANISMS

1979-1985

1988-1992


MODELS

1980-1987

TYPE CURVES FOR

PRODUCTION ANALYSIS




MODELS

1988-1999

?

PENN STATE STRATEGY

1988-1992 1987-1991

ESTABLISHING FLOW

MECHANISMS

1979-1985


MODELS

1980-1987

TYPE CURVES FOR

PRODUCTION ANALYSIS


PROCEDURES


MODELS

1988-1999

INVERSE SOLUTION

METHODOLOGIES UTILIZING SOFT

COMPUTING PROTOCOLS

2001- COMPUTATIONALLY

ENHANCED NUMERICAL

MODELS

2008-

§ GENERATE A WIDE SPECTRUM OF FORWARD SOLUTIONS USING ANALYTICAL AND NUMERICAL TOOLS DEVELOPED

§ CONSTRUCT ARTIFICIAL NEURAL NETWORK TOPOLOGIES THAT CAN BE USED FOR RESERVOIR CHARACTERIZATION AND PRODUCTION FORECASTING PURPOSES

§ TRAIN THE ARTIFICIAL NEURAL NETWORK

§ USE THE ARTIFICIAL NEURAL NETWORK FOR PREDICTION PURPOSES

DEVELOPMENT OF AN EXPERT SYSTEM

§ RESERVOIR CHARACTERIZATION §  PERMEABILITY

§  POROSITY

§  SORPTION CONSTANTS

§  RELATIVE PERMEABILITY

§ FIELD SCALE APPLICATIONS §  OPTIMIZED FIELD DEVELOPMENT

§  VIRTUAL WELL TESTING

§  PRODUCTION FORECASTING

POTENTIAL USES OF ARTIFICIAL NEURAL NETWORKS

Coal seam characteristics and desired field performance

optimized field development strategies

coal seam characteristics

Field performance and field development parameters

§  USE OF THE DEVELOPED TOOLS IN ESTIMATING THE EXPECTED OUTCOME OF A CO2 SEQUESTRATION PROJECT IN SHALE GAS AND COAL SEAM RESERVOIRS

Coal seam or shale gas reservoir and CO2 properties

Prediction of CO2 sequestration performance indicators

POTENTIAL USES OF ARTIFICIAL NEURAL NETWORKS – a project screening tool

APPLICATIONS OF COMPUTATIONALLY ENHANCED

NUMERICAL MODELS – local grid refinement protocols


NUMERICAL MODELS – CO2 sequestration in coal seams

CO2 Sequestration profiles

0.00E+00

1.00E+06

2.00E+06

3.00E+06

4.00E+06

5.00E+06

6.00E+06

7.00E+06

8.00E+06

0 200 400 600 800 1000 1200

Time, Days

Rat

e, S

CF/

D

0.00E+00

2.00E+08

4.00E+08

6.00E+08

8.00E+08

1.00E+09

1.20E+09

1.40E+09

1.60E+09

Am

ount

, SC

F

CH4 Production Rate CH4 Production Rate-Prim CH4 Cum ProdCO2 in place CH4 Cum Prod-Prim


NUMERICAL MODELS – CO2 sequestration in shale reservoirs

SUMMARY

§  MULTIMECHANISTIC FLOW CONCEPT IN TIGHT AND ULTRA-TIGHT SYSTEMS

§  NUMERICAL MODELS WHICH TAKE INTO ACCOUNT FULL PHYSICAL DESCRIPTION OF THE FLOW DYNAMICS

§  PTA PROCEDURES THAT CAN BE USED IN IN-SITU CHARACTERIZATION OF COAL SEAMS INCLUDING ITS SORPTION CHARACTERISTICS

§  TYPE CURVES FOR TWO-PHASE RATE TRANSIENT PRESSURE TRANSIENT APPLICATIONS IN COAL SEAMS AND SHALE GAS RESERVOIRS

PENN STATE’S CONTRIBUTIONS:

§  USE OF THE DEVELOPED TOOLS IN CO2 SEQUESTRATION APPLICATIONS

§  USE OF THE DEVELOPED TOOLS IN OPTIMIZED PRODUCTION OF COALBED METHANE AND SHALE GAS RESERVOIRS

SUMMARY

PENN STATE’S FUTURE PLANS

EPILOGUE

§  IMPROVED OUR UNDERSTANDING OF THE BEHAVIOR OF ADSORBED GASES ON COAL AT A MACROSCOPIC LEVEL

§  PROVIDED A QUALITATIVE AND QUANTITATIVE UNDERSTANDING OF THE PHYSICS OF COAL/METHANE INTERACTIONS AND FLUID FLOW MECHANISMS IN COALBED RESERVOIRS

§  ESTABLISHED A PLATFORM FOR TRAINING A NEW SCHOOL OF COALBED RESERVOIR SCIENTISTS AND ENGINEERS

PENN STATE’S REVISITING OF RESERVOIR ENGINEERING OF COLABED METHANE PRODUCTION HAS:

§  CBM RESERVE ESTIMATES VARY; HOWEVER A RECENT ESTIMATE FROM THE U.S. GEOLOGICAL SURVEY PREDICTS ALMOST 1,000 TRILLION CUBIC FEET OF METHANE WITHIN THE US.

§  AT A NATURAL GAS PRICE OF US$3.00 PER THOUSAND SCF, THAT VOLUME IS

WORTH US$3.0 TRILLION. AT LEAST 100 TRILLION CUBIC FEET OF IT IS ECONOMICALLY VIABLE TO PRODUCE.

§  CBM CURRENTLY ACCOUNTS FOR NEARLY 10% OF U.S. ANNUAL GAS

PRODUCTION AND APPROXIMATELY 12% OF ESTIMATED TOTAL U.S. NATURAL GAS RESERVES.

EPILOGUE

ACKNOWLEDGEMENTS

GREGORY R. KING Ph.D. 1985 DAVID J. REMNER M.S. 1984 JAMES E. KOLESAR M.S. 1985 WONMO SUNG Ph.D 1987 KEMAL ANBARCI Ph.D 1991 SHAHAB MOHAGHEGH Ph.D 1992 ADWAIT CHAWATHE Ph.D 1995 JULIO MANIK Ph.D 1999 TIMOTHY E. KOHLER Ph.D 1999 XIULI DONG M.S. 2003 OLUFEMI ODUSOTE M.S. 2003 BURCU GORUCU M.S. 2005 KARTHIK SRINIVASAN M.S. 2008 PROB THARAROOP Ph.D. 2010 VAIBHAV RAJPUT M.S. 2012 DENNIS ALEXIS Ph.D. 2013 ERHAN ASLAN Ph.D. 2013

REDESIGNING COALBED METHANE RESERVOIR · PDF fileREDESIGNING COALBED METHANE RESERVOIR ENGINEERING: “a reflective analysis”! Turgay Ertekin Penn State University Fall

Documents