Society of Petroleum Engineers

Distinguished Lecturer 2005-06 Lecture Season

SPE DISTINGUISHED LECTURER SERIES

is funded principally
through a grant of the

SPE FOUNDATION
The Society gratefully acknowledges
those companies that support the program
by allowing their professionals
to participate as Lecturers.
And special thanks to The American Institute of Mining, Metallurgical,
and Petroleum Engineers (AIME) for their contribution to the program.
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 1/54

Society of Petroleum Engineers

Distinguished Lecturer 2005-06 Lecture Season

Performance-Based Reservoir
Characterization State-of-the-Technology
Thomas A. Blasingame, Texas A&M U.
Department of Petroleum Engineering
Texas A&M University
College Station, TX 77843-3116
+1.979.845.2292 t-blasingame@tamu.edu
Lecture materials located at:
link http://www.pe.tamu.edu/blasingame/data/
folder z_SPE_DL
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 2/54

Prelude: Performance-Based Reservoir Characterization

Q. What is "Performance-Based Reservoir Characterization?"
A. In the context of this lecture, "Performance-Based Reservoir Characterization" is the diagnosis, analysis, and interpretation of production
data taken from an individual well and the integration of the results
of this analysis with other forms of data (e.g., petrophysical data, results of well test analysis, geological data, etc.). Diagnosis/Interpretation:

z Discussion: Prelude

Generally pretty good,

mostly an issue of production data quality.
Analysis:
Modern tools for production data analysis are
excellent analytical,
semi-analytical, and
numerical modelling are
essentially without flaw.
Integration:
Evolving major issues
are data quality and the
scale(s) of the results.

Production analysis (PA) versus pressure transient analysis (PTA)? (same)

Weakness of PA?
(data quality/quantity)
Strength of PA?
(continuous monitoring, PTA becomes a subset of PA)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 3/54

Prelude: Guidelines for Production Analysis

z Guidelines: Performance-Based Reservoir Characterizations

REVIEW production data for consistency (allocations, accuracy, etc.).

REVIEW well history, particularly recompletions/stimulations.
GATHER/CORRELATE petrophysical data (core, logs, etc.).
PERFORM simplified analysis of production data (Arps, EUR, etc.).

REVIEW measured rate/pressures (quality check).

PERFORM model-based analysis of production (and well test) data.

INTEGRATE results at different scales (without reservoir simulation).

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 4/54

Primer: Reservoir Engineering

Q. What do I need to know for reservoir engineering ... ?
A. All you need is ... Darcy's law and material balance.
Darcy's Law:
1 kh p
q = ( )
r

141.2 B r
Oil Material Balance Eq.: (p>pb)
p = pi

1 Bo
Np
Nct Boi

Oil Diffusivity Eq.: (p>pb)

2 p

ct p
1 p
1
=
r 2 r r 0.0002637 k t
Oil Pseudosteady-State Flow Eq.: (p>pb)
o Bo 1 4 1 A
ln
+ s qo
p = p wf + 141.2
kh 2 e CA rw2

z Discussion: Primer Reservoir Engineering

Diffusivity equation?
(mass continuity (balance) + Darcy's Law)
Oil pseudosteady-state equation? (material balance + diffusivity equation)
Implications?
(origin of virtually all reservoir engineering relations)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 5/54

Orientation: PA Theory
Q. What is the theory to represent boundary-dominated flow behavior?
A. Combine material balance and pseudosteady-state flow equations.
The quantity Np/qo (or
Gp/qg) is known as "material balance time" is
rigorous for boundarydominated flow, and is a
very good approximation for transient flow.

Oil Material Balance Eq. (MBE): (p>pb)

1 Bo
p = pi mo , pss Np where mo , pss =
Nct Boi

Oil Pseudosteady-State Flow Eq. (PFE): (p>pb)

o Bo 1 4 1 A
ln
+ s
p = p wf + bo , pss qo where bo , pss = 141.2

2
kh 2 e CA rw

Production Analysis Relation:

(Gas approximate form)
(Oil analytical form)

( pi p wf )
Np
= bo , pss + mo , pss
qo
qo

m( pi) m( p wf )

z Discussion: Orientation PA Theory

qg

g , pss
= bg , pss + m

m( p) = 1

p
p
dp
pbase g z

Gp
qg

"Production Analysis Relation" implies? (plot: log[(pi-pwf)/qo] vs. log[Np/qo])

Use of "log-log" plot (i.e., log[(pi-pwf)/qo] vs. log[Np/qo])? (diagnosis/analysis)
Does material balance time work?
(yes very robust formulation)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 6/54

Orientation: Production Analysis (PA)

Production Analysis (PA) is
analogous to raising children
it is problematic every
case is unique (and potentially very frustrating). BUT,
the PA experience is also
potentially very rewarding ...
you always learn something
in the process.

z Discussion: Orientation Production Analysis (PA)

What are the major issues?

(pressure and rate data quality/quantity)
What are the benefits?
(est. reservoir properties/volume, rate prediction)
Where are the tools?
(commercial vendor products)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 7/54

Orientation: Production Analysis (PA) Data

Q. What is Production Analysis (PA)?
A. Combined analysis of rate and flowing bottomhole pressure data.
Rate Data:
Flowrates are measured
on a per-well basis for
most gas wells oil
flowrates are often allocated (this is a major
issue).
Pressure Data:
Measured bottomhole
pressure data are essentially non-existent
surface pressure data
are often available for
gas wells, flowing (surface) pressure data for
oil wells are rare at
best.

z Discussion: Orientation Production Analysis (PA) Data

PA is a passive technology can it yield high resolution results?

(yes)
What is/are the key data issue(s)? (q and pwf data accurate and correlated)
Erratic q and pwf data how to analyze? (boundary-dominated flow theory)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 8/54

Orientation: PA Data Analysis Example (Gas)

Q. Does material balance time work? (prove with an example)
A. Typical low productivity gas well example (mid-Continent US).
Transient Behavior:
Appears to reflect behavior of a vertical well
with a finite-conductivity vertical fracture.
m( p)

q g

Gp

g , pss
m

q g
trans

(1/4)

BDF/PSS Behavior:
Clear indication of
(approximate) material
balance behavior.
m( p)

q g

Gp
g , pss
m

bdf
q g

(1)

Material balance behavior is independent of

reservoir shape!

z Discussion: Orientation PA Data Analysis Example (Gas)

What are the limitations of this approach? (poor rate data, poor sampling)
What is the "best" data frequency? (minimum preferred frequency = daily)
What is the effect of pressure? (pressure data critical but in practice ...)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 9/54

Orientation: PA Objectives
Q. What are the objectives of Production Analysis (PA)?
A. Estimate reservoir properties and volume, and predict performance.
Transient Radial Behavior: (PTA/PA)

r(t)

(infinite-acting
radial flow behavior)

Common Characteristics:
zConstant rate (q).
zConstant reservoir properties.
zConstant fluid properties.
Differences:
zVolume ONLY from pseudosteady-state data.
zReservoir properties ONLY
from transient flow data.

p = ( pi p wf ) = btrns ,cr + mtrns ,cr ln(t )

d
d
[p] = mtrns ,cr
p' = t [p ] =
dln(t )
dt
mtrns ,cr = f (k , ...)

Pseudosteady-State Behavior: (PA)

btrns ,cr = f (k , s, ...)

re
dp
dr = 0
r
e

(closed outer
boundary)

p = ( pi p wf ) = bpss ,cr + m pss ,crt

d
d
[p] = m pss ,crt
p' = t [p ] =
dt
dln(t )
m pss ,cr = f ( N , ...) bpss ,cr = f (k , s, ...)

z Discussion: Orientation PA Objectives

In THEORY any differences in PA and PTA? (none, could use same tools)
In PRACTICE differences in PA and PTA? (PTA data is higher resolution)
In FUTURE what will happen with PA and PTA? (applications will merge)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 10/54

Orientation: PA Data Requirements

Q. What are the data requirements for Production Analysis (PA)?
A. Rate and pressure data, reservoir and fluid properties, well history.
Required Data:
zGood pressure data!
zConsistent rate!
zWell history.
zFluid properties.
Issues:
zEffect of well completion (see example).
zConversion of ptf
pwf (will remain a
major issue, consider
pwf measurement).
zAlways review well
history prior to performing PA.

z Discussion: Orientation PA Data Requirements

Issues with pressure?

(measured infrequently at surface)
Issues with rate?
(gas usually good, oil problematic, water poor)
Issues with well completion?
(review history, keep completion simple)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 11/54

History: Production Analysis (PA)

Past:
zEstimate reserves.
zSimple rate forecasts.
zSimple estimates of
reservoir properties.
Present:
zFull variable-rate/variable pressure capability analysis and
modelling (rigorous).
zRobust analysis
methods and very
good software.
Future:
zMeasured pwf data.
zLarger data volumes.
zWill still need tools for
poor (i.e., field) quality data.

z Discussion: History Production Analysis (PA)

PA up to 1970?
(estimate reserves, maximum rate, and future rates)
PA 1970-1990? (estimate reservoir properties simple models [pwf=con])
PA 1990-2000? (estimate reservoir properties general q and pwf profiles)
PA 2000-?
(interactive diagnosis, analysis-by-modelling, and forecasting)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 12/54

History: Production Analysis (PA) q(pwf)

From: Rawlins, E. L. and M. A. Schellhardt: Backpressure

Data on Natural Gas Wells and Their Application To

Production Practices, Monograph 7, U.S. Bureau of Mines,
Washington, DC, (1936).

Q. What is "deliverability" analysis, and why did it evolve?

A. Analysis of rate-pressure data, evolved to estimate maximum rate.
Well Deliverability:
z The original purpose
of well performance
analysis was to quantify well deliverability.

z The empirically de-

rived "deliverability"
equation is given by:

b. "Deliverability" Plot: (pav2-pwf2)

versus qg (log-log format)
used to estimate qg(pwf=0) (an
indicator of well performance
and recovery).
a. Wellbore Diagram: Ancient,
but still accurate.

2 n
q g = C ( p 2 p wf
)
This relationship is rigorous for low pressure gas
reservoirs (it can be derived from the p2-form of the
gas diffusivity equation
(n=1) for laminar flow).

z Discussion: History Production Analysis (PA) q(pwf)

Purpose of a q(pwf) relation?

(maximum rate, inflow performance)
Typical applications?
("deliverability" tests, darcy/non-darcy flow)
Limitations?
(rate a function of pressure (not time), laminar flow)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 13/54

History: Production Analysis (PA) q(t)

(Case 1)

Reserves by Oil-Well Production Curves, Department of the

Interior, U.S. Bureau Of Mines Bulletin 228 (1924).

From: Cutler, W.W., Jr.: Estimation of Underground Oil

Q. What is "rate-time" analysis, and what was its original purpose?

A. Analysis of rate-time data (graphically), originally used for taxation.
Concept:
Create an extrapolation
of the rate-time profile to
yield "estimated ultimate
recovery" (EUR) (i.e., to
provide a reserves estimate).
Limitation(s):
Early rate-time models
had no direct basis in
theory but ironically,
these models have been
proven in modern times,
and are regularly used
as standards for reserves estimation (i.e.,
the exponential and
hyperbolic rate-time
relations).

z Discussion: History Production Analysis (PA) q(t) (Case 1)

Theory for q(t)=qiexp(-Dit)? (... log(qo) vs. t straight-line data confirms model)
Theory for q(t)=qi/[(1+bDit)(1/b)]? (... other plots would be required to confirm)
Theory for "Averaged and Extrapolated" trend given by Cutler? (unknown)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 14/54

History: Production Analysis (PA) q(t)

(Case 2)

From: Cutler, W.W., Jr.: Estimation of Underground Oil

Reserves by Oil-Well Production Curves, Department of the

Interior, U.S. Bureau Of Mines Bulletin 228 (1924).

Q. "Non-exponential" q(t) behavior is it real?

A. See calibrated analysis using exponential and hyperbolic models.

a. "Rate-Time" Plot: qo versus t (or qg versus

t) hyperbolic behavior dominant, note
that "early" exponential is required.

b. "Rate-Cumulative" Plot: qo versus Np (or qg

versus Gp) apparent hyperbolic behavior,
analysis calibrated with q(t) data.

z Discussion: History Production Analysis (PA) q(t) (Case 2)

q(t) behavior?
("higher" resolution than qo(Np))
qo(Np) behavior?
("confirms" q(t) behavior also extrapolate Np(qo0))
Issue 1?(exponential and hyperbolic relations show early linear qo vs. log(t))
Issue 2?
(in concept apply to single well; in practice field/pool data)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 15/54

History: Production Analysis (PA) q(t)=qiexp(-Dit)

Q. Can the "exponential" rate-time relation (q(t)=qiexp(-Dit)) be derived?
A. Yes, see steps below slightly compressible liquid, pwf=constant.
The q(t)=qiexp(-Dit) form
Oil Material Balance Eq. (MBE): (p>pb)
is correct for boundary1 Bo
dominated flow behavior
p = pi mo , pss Np where mo , pss =
slightly compressible
Nct Boi
liquid, pwf=constant.
Oil Pseudosteady-State Flow Eq. (PFE): (p>pb)
o Bo 1 4 1 A
ln
+ s
p = p wf + bo , pss qo where bo , pss = 141.2
kh 2 e CA rw2

Steps:

1. Differentiate oil MBE and oil PFE with respect to time.

2. Assume: pwf = constant [i.e., d(pwf)/dt = 0].
3. Equate results 1st order o.d.e.
4. Separate/integrate.

mo , pss
5. Exponentiate result final form: q = qi exp( Di t ) Di =

bo , pss

z Discussion: History Production Analysis (PA) q(t)=qiexp(-Dit)

Is the q(t)=qiexp(-Dit) relation rigorous?

(oil (p>pb) yes, gas ?)
Other formulations?
(rate-cumulative: q(t)=qi DiNp)
Applications?
(estimate reserves (Np,max = Np(q0)), q(t) predicton)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 16/54

History: Production Analysis (PA) q(t)

Theory: (exp model)

Assume slightly compressible liquid, pwf=
constant use oil MBE
and oil PSS FE to yield:
(p<pb)
Curves," JPT (June 1980) 1065-1077.

From: Fetkovich, M.J.: "Decline Curve Analysis Using Type

Q. "Rate-Time" Plot: qo versus t (or qg versus t)?

(original purpose?)
A. Base "performance" plot for PA.
(originally used for taxation...)

q = qi exp( Di t )
Theory: (hyp model)
APPROXIMATE derivation using oil MBE
(p<pb), and oil PSS FE
(also p<pb):

q=

qi

(1 + bDi t ) (1 / b )

z Discussion: History Production Analysis (PA) q(t)

Is the q(t)=qiexp(-Dit) model supported by theory?

(yes liquid; pwf=con)
Is the q(t)=qi/[(1+bDit)^(1/b)] model supported by theory?
("not exactly")
Issues?
(exponential = conservative; hyperbolic = liberal use caution)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 17/54

History: Production Analysis (PA) (EUR)exp

Theory: (exp model)
The q(t)=qiexp(-Dit)
model (i.e., slightly compressible liquid, pwf=
constant) is integrated
to yield:
Curves," JPT (June 1980) 1065-1077.

From: Fetkovich, M.J.: "Decline Curve Analysis Using Type

Q. What is an Estimated Ultimate Recovery (or EUR) plot?

A. Plot q(t) vs. Np, extrapolate to zero rate using a straight line ((EUR)exp).

q = qi Di Np
Application: (EUR)exp
Plot q(t) versus Np, extrapolate trend using a
straight-line model to
Np(q=0) this gives
estimated ultimate recovery (EUR)exp.

z Discussion: History Production Analysis (PA) (EUR)exp

Is the estimated ultimate recovery (EUR)exp supported by theory?

(yes)
Is the (EUR)exp conservative or liberal?
((EUR)exp is always conservative)
Other issues? (use of other EUR models (e.g., hyperbolic) requires caution)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 18/54

History: Production Analysis (PA) EUR vs. qo,1yr

"Ancient" Technique:
The proposed correlation of EUR vs. qo,1yr was
used to estimate oil reserves from initial production performance
data.
Modern Application:
Approach is based in
theory EUR = f[k, s, xf,
... and contacted fluids
in-place (i.e., N or G)].

Revenue Act of 1918, Treasury Department United States

Internal Revenue Service (1919).

From: Manual for The Oil and Gas Industry Under The

Q. Origin and purpose of the EUR versus qo,1yr correlation?

A. Potential value as a correlation, but must quantify theory (N, k, s, etc).

Could be used as a "reservoir characterization"

tool to classify well performance.

z Discussion: History Production Analysis (PA) EUR vs. qo,1yr

Origin?
Theory?
Rationale?

T.A. Blasingame

(1919 production data correlation (>85 years old)!)

(Constant pwf (liquid) boundary-dominated flow conditions)
(Correlate reserves versus production (or reservoir properties))

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 19/54

History: Production Analysis (PA) Arps

Case

Cumulative Relation
q
Exponential: (b=0) q = qi exp( Di t )
Np = i [1 exp( Di t )]
Di
qi
qi
Np =
1 (1 + bDi t )1 (1 / b )
Hyperbolic: (0<b<1) q =
(1 b ) Di
(1 + bDi t ) (1 / b )
qi
qi
q=
Np =
ln(1 + Di t )
Harmonic: (b=1)
(1 + Di t )
Di

Rate Relation

Arps' observations:
(1945) 160, 228-247.

From: Arps, J.J: "Analysis of Decline Curves," Trans., AIME

Q. Theory for Arps' relations?

A. Arps derived the exponential and hyperbolic relations from loss ratio.

b=0

b=0

b=0.5
b=0.667
b=0.333

Reservoir is highly undersaturated (p>pb).

Gravity drainage and no free surface.
Gravity drainage with free surface.
Soln. gas-drive reservoir ( p vs. Np linear).
2
Soln. gas-drive reservoir ( p vs. Np linear).

Theory???

Loss Ratio:
a

1
q

D
dq/dt

Theory???

Loss Ratio Derivative:

b

d
[a ] d 1 d
dt
dt D
dt

/
dq
dt

z Discussion: History Production Analysis (PA) Arps

"Theory" for the Arps' relations?

(loss ratio (exp) and its derivative (hyp))
Validity of the Arps' observations? (only qualitative (except for p>pb case))
Graphical analysis using the hyperbolic relation? (only using Fetkovich TC)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 20/54

History: Production Analysis (PA) Fetkovich

Curves," JPT (June 1980) 1065-1077.

From: Fetkovich, M.J.: "Decline Curve Analysis Using Type

Q. What is the "Fetkovich" Decline Type Curve, and how is it used?

A. A composite of analytical (pwf=con) and empirical (Arps) solutions
used as a "type curve" (data overlay) to estimate reservoir properties.

Variables for the Fetkovich Decline Type Curve

t Dd =

q Dd =

0.00633 kt

ct rw 2 1 r 2 r 1
e 1 ln e
rwa 2
2 rw

q(t )
kh ( pi pwf )
r 1
141.2 B ln e
rwa 2

rwa = rwe s

Transient Stems: (left)

zInfinite-acting radial
flow model (pwf = con).
zq(t) is concave up.
Depletion Stems: (right)
zBounded circular reservoir (pwf = con).
zq(t) is concave down.
zb=0: pwf = con.
zb=1: qo = con. (qo/p).
zb>1: transient flow or
external drive energy.
Reservoir Properties:
zk y-axis match.
zN x&y-axis matches.
zs reD match.

z Discussion: History Production Analysis (PA) Fetkovich

An original purpose of the Fetkovich TC? (graphical solution of Arps Eqs.)

Use of "transient" stems?
(estimate reservoir properties k and s)
Use of "depletion" stems?
(estimate reservoir volume, predict rate)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 21/54

History: Production Analysis (PA) Carter

linear Gas Flow Systems: Constant Terminal Pressure

Case," SPEJ (October 1985) 719-728.

From: Carter, R.D.: "Type Curves for Finite Radial and

Q. What is the "Carter" Decline Type Curve, and how is it used?

A. A numerically-generated gas rate solution (pwf=con) used as a
"type curve" (data overlay) to estimate reservoir properties.

Variables for the Carter Decline Type Curve

t Dd =

q Dd =

0.00633 kt

gi cti rw 2 1 r 2 r 1
e 1 ln e
rwa 2
2 rw

q(t )
kh ( pi p wf )

rwa = rwe s

r 1
141.2 gi B gi ln e
rwa 2

Transient Stems: (left)

zNumerical flow model
(pwf = con).
zq(t) is concave up.
Depletion Stems: (right)
zq(t) is concave down.
zb=0: pwf = con.
zb=1: qo = con. (qo/p).
zb>1: transient flow or
external drive energy.
z: numerical gas flow
cases ( =f(pwf/(pi)).
Reservoir Properties:
zk y-axis match.
zG x&y-axis matches.
zs reD match.

z Discussion: History Production Analysis (PA) Carter

Genisis of the Carter TC?
Use of "transient" stems?
Use of "depletion" stems?

T.A. Blasingame

("correction" of Fetkovich gas flow solutions)

(estimate reservoir properties k and s)
(estimate reservoir volume, predict rate)

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 22/54

Modern PA: Orientation

"Modern production analysis (PA) is a mature and functional

process but we have to recognize that these methods may
make us overconfident in our abilities to diagnose, interpret,
and analyze a particular field data case."
Blasingame

z Discussion: Modern PA Orientation

Variable Rate/Pressure:
zMaterial balance time (equivalent constant rate time
function).
zDerived by applying the
superposition formulation
to the pseudosteady-state
solution.
Auxiliary Plotting Functions:
zTime-averaged rate function (rate integral function).
zDerivative, integral, and integral-derivative functions.
Semi-Analytical Solutions:
zAnsah, et al gas solutions/
Buba method.
zAdvanced methods for
hyperbolic rate analysis.
Adv. Decline Type Curves:
z"Palacio" type curve.
z"Doublet" type curve.
zCumulative production TC.

Major advances in PA?

(variable q/variable pwf analysis methods)

(auxiliary plotting functions to improve analysis)
(semi-analytical solutions advances from the Arps' relations)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 23/54

Modern PA: Material Balance Time (Theory)

Q. Concept of material balance time?
A. Constant-rate equivalent function (variable-rate constant rate case).

z Pressure Drop Form:

z Flowrate Form:

qo
1
p
=
= bo , pss + mo , pss t
qo
p bo , pss + mo , pss t
1 Bo
(slope)
mo , pss =
Nct Boi
o Bo 1 4 1 A
ln
+ s
bo , pss = 141.2
(intercept)
2

kh 2 e C A r
w

Np
(material balance time)
t=
qo
z Discussion: Modern PA Material Balance Time (Theory)

Value of material balance time? ("deconvolution" constant rate response)

p/qo versus Np/qo (pressure drop form)? (PTA analog (constant rate case))
qo/p versus Np/qo (rate decline form)?
(decline curve analysis format)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 24/54

Modern PA: Material Balance Time (Application)

Q. Application of material balance time?
A. Requires pressure and rate histories (accurate rates essential).
Theory:
Material balance time
and the general boundary-dominated flow
relation:
t =

[ q o /p ]

Np
qo

p
= bo , pss + mo , pss t
qo
qo
1
=
p bo , pss + mo , pss t

[ p /q o ]

[t = Np /qo ]

Application:
Plot p/qo versus Np/qo
and qo/p versus Np/qo
to correct for rate effects and to establish
boundary-dominated
flow behavior (BDFB).

z Discussion: Modern PA Material Balance Time (Application)

Effectiveness on raw data?

(note example excellent correlation)
p/qo versus Np/qo data?
(transient flow ?, BDFB yes!)
qo/p versus Np/qo (rate decline form)?
(transient flow ?, BDFB yes!)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 25/54

Modern PA: Auxiliary Plotting Functions (Theory)

Q. What are the "auxiliary" plotting functions for PA?
A. Time (or material balance time)-averaged pressure drop and flowrate
functions "integral" and "integral-derivative" forms are used.
Theory:
z Normalized Pressure Drop: z Normalized Flowrate:
"Pressure Derivative"

p
d p

=t
q
d
t
qo
o d
"Pressure Integral"

t
p
p
1
=

dt
qo i t 0 qo

"Rate Derivative"

qo
d qo
=t

d
t
p
d
p
"Rate Integral"

t
qo
qo
1
=
dt
p i t 0 p

"Pressure Integral-Derivative" "Rate Integral-Derivative"

The auxiliary functions (specifically the

integral and integral
derivative functions)
are designed to provide "smooth" data
functions which
should improve interpretation and analysis.
Application:
z"Pressure drop"
functions "Normalized PI" plot .
zFlowrate" functions
"Blasingame"
plot .

p
d p
qo
d qo

=t

=t

dt qo i
dt p i
qo id
p id

z Discussion: Modern PA Auxiliary Plotting Functions (Theory)

Theory for auxiliary plotting functions?
(just calculus ...)
Purpose of auxiliary plotting functions?
(more resolution for PA data)
Advice? (use BOTH pressure drop and flowrate functions (separate plots))
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 26/54

Modern PA: Auxiliary Plotting Fcns (Application 1)

Q. Application 1 What is the "Normalized PI" plot?
A. Pressure drop auxiliary functions versus material balance time.
Theory:
(p/qo) fcns versus Np/qo:
t =

Np

qo
p
d p
t
=


q
d
t
o d
qo
t
p
1
p
=
dt

q
t
o i
0 qo
p
d p

=t
q
d
t
qo i
o id

[ p /q o ]

[p/qo ] i
[p/qo ] d

Application:
Analysis approach same
as PTA look for characteristic behavior.

[p/qo ] id
[t = Np /qo ]

z Discussion: Modern PA Aux. Plotting Fcns ("Normalized PI" plot)

p/qo functions versus Np/qo? (PTA form: transient flow ?, BDFB yes!)
Validity of (p/qo)d function? (exceptional case downhole pressure data)
(p/qo)i and (p/qo)id functions?
(smooth and representative)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 27/54

Modern PA: Auxiliary Plotting Fcns (Application 2)

Q. Application 2 What is the "Blasingame" plot?
A. Rate auxiliary functions versus material balance time.
Theory:
(qo/p) fcns versus Np/qo:
t =

[ q o /p ] i

Np
qo

qo
d qo
=
t


dt p
p d
t
qo
qo
1
=

dt

p
t
i
0 p
qo
d qo

=t
p
d
t

p i
id

[ q o /p ]

[ q o /p ] d
[ qo /p ] id

[t = Np /qo ]

Application:
Transient flow behavior is
complex, but transition
and BDFB unique.

z Discussion: Modern PA Auxiliary Plotting Fcns ("Blasingame" plot)

qo/p versus Np/qo?

(rate decline form: transient flow ?, BDFB yes!)
Validity of (qo/p)d function?
((again) good downhole pressure data)
(qo/p)i and (qo/p)id functions?
(well-behaved and representative)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 28/54

Modern PA: Ansah, et al Gas Solution/Buba Method

Q. What is the Ansah, et al gas solution/Buba method?
A. A gas-specific solution for qg(Gp).
Theory:

2 p
wf
p

qg = C
z
z wf

p pi Gp
=
1

z
zi
G

Rate-Cumulative Result:
qg = qgi Di Gp +

1 Di 2
Gp
2 G

where
Di =

2 qgi
p /z
1 wf wf
pi / z i

z Discussion: Modern PA Ansah, et al Gas Solution/Buba Method

Validity of the Ansah, et al gas solution?

(general criteria: pi < 6000 psia)
Application of the Buba method?
(assortment of plotting functions)
Advantages?
(rigorous solution for gas flow cases (pwf = constant))

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 29/54

Modern PA: Hyperbolic EUR Methods (Type Curves)

Q. Straight-line or characteristic behavior for Arps hyperbolic relation?
A. Rate-cumulative ratios used to define "type curve" behavior.

z Discussion: Modern PA Hyperbolic EUR Methods (Type Curves)

Example gas case from Fetkovich (SPE 13169)?

("typical" gas case)
log[(qg/qgi)] versus log[1-(Gp/G)]?
(straight-line characteristic behavior)
Cartesian (qg/qgi) versus (Gp/G)?
(note that 0.4<b<0.5)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 30/54

Modern PA: EUR Methods (Schematic Behavior)

Q. Collection of modern EUR methods?
A. Comparison of methods illustrates behavior.

Liquid: (pwf = constant)

q g = q gi Di Gp

q gi

Di

Gas: (pwf = constant)

qg = qgi Di Gp +
Di =

1 Di 2
Gp
2 G

2 qgi
p /z
1 wf wf
pi /zi

Hyperbolic:

G p (1b)
q g = q gi 1

q gi

(1 b) Di

Harmonic:

z Discussion: Modern PA EUR Methods

Liquid (oil) flow behavior?
Gas flow behavior? (rigorous)
General hyperbolic relation?

T.A. Blasingame

i
q g = q gi exp
Gp
q gi

(straight-line rate-cumulative behavior)

(rate-quadratic cumulative behavior)
(straight-line log[qg] versus log[1-(Gp/G)])

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 31/54

Modern PA: Palacio Type Curve (Constant pwf case)

Q. Enhancement of the functions on the original Fetkovich type curve?
A. "Palacio" type curve uses auxiliary rate functions (better resolution).
z Auxiliary functions:
Rate integral:

t
1
qo ,i =
qo dt
t 0

Rate integralderivative:

qo ,id = t

d
qo ,i
dt

z Discussion: Modern PA Palacio Type Curve (Constant pwf case)

Auxiliary functions: Rate integral, integral derivative.

(smoothness/detail)
(time-averaged rate function)
Rate integral: qo,i = (1/t) Int(qo, 0, t).
(derivative-like rate function)
Rate integral derivative: qo,id = t |d/dt[qo,i]|.

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 32/54

Modern PA: Doublet Type Curve (Constant Rate Eq.)

Q. Type curve solution for variable-rate/variable pressure case?
A. "Doublet" type curve uses material balance time function.
z Auxiliary functions:
Material Balance
Time:

t
1
qo dt
t=
qo 0
Rate integral:
t q
qo
1
o
dt
=
p i t 0 p
Rate integralderivative:
qo
d qo

=t
p
d
t

p i
id

z Discussion: Modern PA Doublet Type Curve (constant rate eq.)

Material Balance Time: tmb = (1/q) Int(qo, 0, t). (material bal. "deconvolution")
Auxiliary Functions: Defined using qo and p. (accounts for qo and p = f(t))
Validity: Variable-rate/variable pressure drop cases. (rigorous for pss flow)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 33/54

Modern PA: Rate-Cumulative Type Curves

Q. Rate-cumulative type curves ... how can these help?
A. "Gross" comparison of rate-cumulative behavior.

z Discussion: Modern PA Rate-Cumulative Type Curves

Value?
(probably most value as a screening tool)
Pitfalls?
(VERY difficult to distinguish individual trends hyperbolic)
Implementation? (recommend use as a validation tool with Arps' analysis)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 34/54

Modern PA: Normalized PI Plot

Q. What is the "Normalized PI" plot (and how is it used)?
A. This is pressure transient analysis (PTA) analog analysis plot.
Plotting Functions:
(p/qo)id versus Np/qo:
(Pressure Integral-Derivative)

t
p
p
1
=
t =

dt
q o qo i t 0 qo
p
d p

=t
d
t
q
qo i
o id
Np

Orientation:
z(0) IARF
z(1/4) Fin. Con.
z(1/2) Inf. Con.
z(1/2) Horizontal
z(1) BDF

k
FcD
xf
Lh
N

z Discussion: Modern PA Normalized productivity index (PI) plot

How is the Normalized PI plot used? (diagnose characteristic flow regimes)

Transient flow behavior?
(radial flow, fractured wells, horizontal wells ... )
Boundary-dominated flow behavior?
(unique "unit-slope" behavior)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 35/54

Modern PA: "Blasingame" Plot

Q. What is the "Blasingame" plot (and how is it used)?
A. This is the decline curve analysis plot (i.e., modern Fetkovich plot).
Plotting Functions:
(qo/p)id versus Np/qo:
(Rate Integral-Derivative)
t
qo
qo
1
t =
=
dt
qo p i t 0 p
qo
d qo

=t
p
d
t

p i
id
Np

Orientation:
z(?) IARF
z(-1/4) Fin. Con.
z(-1/2) Inf. Con.
z(?) Horizontal
z(-1) BDF

k+s
FcD
xf
?
N

z Discussion: Modern PA "Blasingame" plot

How is the "Blasingame" plot used? (diagnose characteristic flow regimes)

TF behavior? (fractured wells, horizontal wells ... IARF not unique (k,s tied))
BDF behavior? (unique transition from TF, unique "unit-slope" at late times)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 36/54

Modern PA: Analysis-by-Modelling

Q. What is "Analysis-by-Modelling?"
A. Interactive (dynamic) analysis using a specified reservoir model.
Issues:
z Variable-rate case: Constant Rate Solution

p = pi 141 .2

zThese superposition

B n

( q j q j 1 ) p sD ,cr (t D t D , j 1 )

kh j =1
z Variable pressure drop case:

1 kh n
q=
( pi p wf , j ) q D ,cp (t D t D , j 1 )
141 .2 B j =1

z Approach:

Constant Pressure Solution

relations are rigorous

i.e., for EXACT inputs, we will obtain
EXACT results.
zThe superposition relations must be modified
for the gas case OR,
the gas (or multiphase)
case(s) can be generated NUMERICALLY.

Assemble time-pressure-rate (TPR) data.

Perform quality control (particularly on pressure data).
Establish initial reservoir model using normalized PI/Blasingame plots.
Generate: rate data + model pwf(t); pressure data + model q(t)
Generate: FORECAST of production and/or pressure performance.

z Discussion: Modern PA "Analysis-by-Modelling

Issues?
(superposition is rigorous q and pwf data must be accurate)
Typical case? (pressure data quality is less-than-optimal interpretation)
Unexpected benefit(s)? (modelling validates/disputes q and pwf data quality)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 37/54

Modern PA: Analysis-by-Modelling (Oil Case)

(1/3)

Points to Consider:
zDAILY data (pressure is
measured continuously, with a single value
reported per day).
zRates appear to be accurate and well-correlated with pressure.

z Discussion: Modern PA Oil Case

Origin of data? (DAILY data: rate surface, pressure downhole gauge)

Likelihood of a successful analysis?
(VERY HIGH)
Rate decline at late times?
(damage, gas-blocking, ?)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 38/54

Modern PA: Analysis-by-Modelling (Oil Case)

(2/3)

Q. Behavior of "normalized PI" and "Blasingame" plots for this case?

A. Both plots perform EXTREMELY well driven by data quality.

a. "Normalized PI" Plot: (p/qo) functions

versus Np/qo extraordinary agreement
between data and reservoir model.

b. "Blasingame" Plot: (qo/p) functions versus

Np/qo excellent data-model performance
(note that even derivative function is valid).

z Discussion: Modern PA "Normalized PI" and "Blasingame" plots

Transient flow data?

(Transition regime is apparent, NOT transient flow)
Boundary-dominated flow?
(very strong agreement all data functions)
Influence of late-time effect? (damage?) (not significant compared to BDF)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 39/54

Modern PA: Analysis-by-Modelling (Oil Case)

(3/3)

Q. Performance of final model match? (i.e., rate and pressure functions)

A. Extraordinary agreement of model and data again, data quality ...
Comment:
zExcellent data quality
yields results similar to
pressure transient
analysis (PTA).
zThis case strong advocates continuous
rate and pressure
measurement the
data are worth the cost.
zDownhole pressure
measurement must be
considered if the economics support the
implementation.
zThere is no substitute
for vigilance in data
acquisition.

z Discussion: Modern PA Final model/data match

Rate match?
(EXCELLENT near perfect match until late times)
Pressure match?
(virtually perfect)
Accounting for "damage?" (use PTA, or "cheat" use variable skin effect)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 40/54

Modern PA: Analysis-by-Modelling (Gas Case) (1/4)

Q. Analysis-by-modelling for gas cases?
A. Same general procedure, must use pseudopressure/pseudotime, or
numerical (or semi-analytical) gas solutions.

Points to Consider:
zUsing both PA and PTA
can yield complimentary analyses.
zAccurate rate and pressure data are required.

z Discussion: Modern PA Gas Case

Origin of data?
(DAILY data: rate surface, pressure surface)
Likelihood of a successful analysis? (high to very high, data well-correlated)
Integration of PTA data? (designed for production, modelling for shut-ins)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 41/54

Modern PA: Analysis-by-Modelling (Gas Case) (2/4)

Q. Behavior of "normalized PI" and "Blasingame" plots for this case?
A. Both plots perform EXTREMELY well driven by data quality.

a. "Normalized PI" Plot: (pp/qg) functions

versus Gp/qg excellent agreement in
data and model functions.

b. "Blasingame" Plot: (qg/pp) functions versus

Gp/qg excellent agreement note that the
qDd functions converge (confirms BDF).

z Discussion: Modern PA "Normalized PI" and "Blasingame" plots

Transient flow data?

(confirms fractured well behavior, moderate FcD)
Boundary-dominated flow?("convergence" (Blasingame plot) confirms BDF)
What is required to achieve similar results?
(accurate q and pwf data)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 42/54

Modern PA: Analysis-by-Modelling (Gas Case) (3/4)

Q. How well can a gas case be modeled?
A. Depends on the data can be excellent...
Comment:
zModelling captures
individual shut-ins,
for both rate and
pressure data.
zSuch performance
REQUIRES very
accurate rate and
pressure data.
zThis case is an exception (low permeability reservoir
permits use of daily
data). High permeability reservoirs
require continuous
rate and pressure
measurements.

z Discussion: Modern PA PA model/data match

Rate match?
(EXCELLENT near perfect match until late times)
Pressure match?
(virtually perfect including shut-ins)
Accounting for "damage?" (use PTA, or "cheat" use variable skin effect)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 43/54

Modern PA: Analysis-by-Modelling (Gas Case) (4/4)

Q. Analysis of individual pressure transient tests?
A. Should be straightforward vigilance in data acquisition is required.
Comment:
zExcellent match on log-log
plot (pressure drop functions) and very good match
of the entire production
pressure history.
zDaily rate and pressure
data are sufficient for this
low permeability reservoir
case.

z Discussion: Modern PA PTA model/data match

PTA match?
(VERY GOOD note that production pwf data also matched)
Comparison of PA and PTA results? (minor differences, due to data quality)
etc.?
(note that pressure history is matched for PTA (entire history))

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 44/54

Summary: Production Analysis (PA)

"The future looks familiar ... We should be able to perform

continuous reservoir monitoring using production analysis
(PA). The theory and tools are well-established the trick is
vigilant data acquisition and quality control"
Blasingame

PA Present:
zStrong roots in theory.
zVery good software tools in
the market.
zFair to good data acquisition practices.
PA Future:
zImproved data acquisition
systems, particularly
permanent bottomhole
pressure measurement.
zIntegration of analytical/
numerical tools numerical models may govern the
future of production data
analysis.
z"Event" analysis (PTA) will
fade and "continuous" data
analysis will dominate.

z Discussion: Summary Production Analysis (PA)

PA Theory?
(variable q/variable pwf analysis methods proven)
PA Tools?
(software products are robust and easy to use)
PA Future? (better data acquisition, numerical models, continuous analysis)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 45/54

Summary: Production Analysis (PA) Issues

Q. Major issues for PA?
A. Rate data, pressure data, and well completion history.
Pressure:
zMeasure consistently.
zCheck for relevance
(i.e., is the pressure
measured in the correct location?).
zPressure conversions
to bottomhole conditions are always approximate (at best).
Flowrates:
zDaily rates are best (as
a practical minimum).
zBe sure to check rates
for accuracy.

z Discussion: Summary Production Analysis (PA) Issues

Pressure data are critical!

(ptf pwf conversion hard to model/correlate)
Consistent rate measurements!
(rates frequency/accuracy is an issue)
Review the well history! (note that several completion changes can occur)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 46/54

Summary: Production Analysis (PA) Analysis

Q. Analysis approach for PA?
A. Not like PTA in PA data quality determines "analyzability" of data.
Interpretation:
zBoundary-dominated
flow data should "selfinterpret."
zTransient flow data
should also "self-interpret" however, a
major issue is the
quantity (i.e., frequency) of transient flow
data.
Modelling:
zSimple is always best.
zDo not use approximations in modelling
(if at all possible).

z Discussion: Summary Production Analysis (PA) Analysis

Interpretation of production data should be obvious. (if complex, beware)

(do not impose an analysis)
DATA should provide analysis.
Keep the reservoir/well model simple as possible.
(simple = appropriate)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 47/54

Summary: Production Analysis (PA) Pressures

Q. Pitfalls of pressure data as these relate to PA?
A. Put simply pressure data (quality/quantity) remain weakest link.
Due Diligence:
zCorrelate pwf versus q
this should give
some indication of
"correlation."
zNo correlation no
analysis/interpretation.
Advice:
zPressure is the weak
link but ... try to
perform an analysis
regardless, at least
validate that pressure
data are relevant or
not.
zReconstruct a pwf(t)
profile by any reasonable means.

z Discussion: Summary Production Analysis (PA) Pressures

Reality: All pressure data are suspect.
(scrutinize ALL pressure data)
(ptf pwf conversion issue)
Fantasy: Pressure at bottomhole conditions.
(rate becomes issue)
Future: pwf (bottomhole measurements) common.
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 48/54

Summary: Production Analysis (PA) Flowrates

Q. Pitfalls of flowrate data as these relate to PA?
A. Flowrate data beware of allocations/frequency of measurements.
Due Diligence:
zAllocated flowrates
always have issues.
Advice:
zFlowrate data are
generally better than
pressure data you
should be able to
obtain an analysis.

z Discussion: Summary Production Analysis (PA) Flowrates

Reality: Rates are generally better than pressures. (allocations are an issue)
(rate-pressure mismatch)
Fantasy: Rates measured at high frequency.
Future: Downhole flowrate measurement. (continuous monitoring/analysis)
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 49/54

Current Assessment: Production Analysis (PA)

"Technology ... is a queer thing. It brings

you great gifts in one hand, and it stabs
you in the back with the other."
C.P. Snow (1905-1980)

z Discussion: Current Assessment Production Analysis (PA)

Analysis?
(Very good tools for PA and PTA (incremental improvements))
Modelling? (Analytical modelling is sufficient, numerical modelling evolving)
Data Issues? (Improvement in data handling (good/bad data, large volumes))
Data Acquisition? (Lack of innovation in testing/monitoring methodologies)
Integration?
(Breakthrough to integrate PA/PTA results with other data)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 50/54

Current Assessment: PA Reality Check

From: Simulator Parameter Assignment and the Problem

of Scaling in Reservoir Engineering Halderson (1986).

Q. What is the "reality check" with regard to PA?

A. It is just time-pressure-rate data, do not expect a miracle.

Reservoir Engineering Model

Works 95+ percent of the time...
Why? Pressure and volume
averaging of reservoir properties.
When does it not work? High
contrast in reservoir properties.
Actual Reservoir Model
Complex geology.
Complex fluid behavior.
Poor lateral (and vertical) continuity.
Challenges:
z Technical data acquisition/management.
z Operational "reservoir management" (use of the data).

z Discussion: Current Assessment PA Reality Check

Reservoir scale issues?
(petrophysical data, PTA, PA, seismic data, etc.)
Results from PTA/PA?
("reservoir-scale" flow character)
Key to success?
(high precision/frequency reservoir performance data)
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 51/54

Current Assessment: Future Work in PTA/PA

Q. What are the prospects for PTA/PA?
A. Most important issue is data quality/frequency characterization.

From: Simulator Parameter Assignment and

the Problem of Scaling in Reservoir
Engineering Halderson (1986).

From: How Heterogeneity Affects

Oil Recovery Weber (1986).

Crossplot: ko (PA and PTA) vs. klog mean

Santa Barbara Field (Venezuela)..

z Discussion: Current Assessment Future Work in PTA/PA

Additional reservoir/well models.

(elliptical flow, moving boundary, ...)
Full incorporation of PVT character.
(volatile oil, gas condensate, ...)
Reservoir scaling for PTA/PA.
(scaling of petrophysical data?)
Handling poor quality rate/pressure data.
(major issue at present)
Continuously measured pwf data. (this is coming ... high frequency rates?)
Multiple well analysis (integration). (analytical (material balance) or models)
Coupling of analysis/interpretation (3D/3P models). (over-determined case)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 52/54

Society of Petroleum Engineers

Distinguished Lecturer 2005-06 Lecture Season

Performance-Based Reservoir Characterization

State-of-the-Technology

End of Presentation
Thomas A. Blasingame, Texas A&M U.
Department of Petroleum Engineering
Texas A&M University
College Station, TX 77843-3116
+1.979.845.2292 t-blasingame@tamu.edu
Lecture materials located at:
link http://www.pe.tamu.edu/blasingame/data/
folder z_SPE_DL
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 53/54

Society of Petroleum Engineers

Distinguished Lecturer 2005-06 Lecture Season

SPE DISTINGUISHED LECTURER SERIES

is funded principally
through a grant of the

SPE FOUNDATION
The Society gratefully acknowledges
those companies that support the program
by allowing their professionals
to participate as Lecturers.
And special thanks to The American Institute of Mining, Metallurgical,
and Petroleum Engineers (AIME) for their contribution to the program.
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide 54/54

Selected References: Production Data Analysis

References Production Data Analysis:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.

Lewis, J.O., and Beal, C.H.: "Some New Methods for Estimating the Future Production of Oil Wells," Trans. AIME (1918) 59, 492-525.
Manual for The Oil and Gas Industry Under The Revenue Act of 1918, Treasury Department United States Internal Revenue Service (1919).
Cutler, W.W., Jr.: Estimation of Underground Oil Reserves by Oil-Well Production Curves, Department of the Interior, U.S. Bureau Of Mines Bulletin 228 (1924).
Roeser, H.M.: "Determining the Constants of Oil-Production Decline Curves," Trans AIME (1925) 71, 1315-1321.
Larkey, C.S.: "Mathematical Determination of Production Decline Curves," Trans AIME (1925) 71, 1322-23.
Johnson, R.H. and Bollens, A.L.: "The Loss Ratio Method of Extrapolating Oil Well Decline Curves," Trans. AIME (1927) 77, 771.
Lindsly, B.E., and Berwald, W.B.: Effect of Vacuum on Oil Wells, U.S. Department of Commerce, Bureau of Mines, Bulletin 322 (1930).
Allen, R.E.: "Control of California Oil Curtailment," Trans AIME (1931) 92, 47-66.
Rawlins, E. L. and M. A. Schellhardt: Backpressure Data on Natural Gas Wells and Their Application To Production Practices, Monograph 7, U.S. Bureau of Mines, Washington, DC, (1936).
Arps J.J.: "Analysis of Decline Curves," Trans. AIME (1945) 160, 228-247.
Gurley, J.: "A Productivity and Economic Projection Method Ohio Clinton Sand Gas Wells," JPT (Nov. 1963) 1183-1188.
Slider, H.C.: "A Simplified Method of Hyperbolic Decline Curve Analysis," JPT (Mar. 1968) 235-236.
Stewart, P.R.: "Low-Permeability Gas Well Performance at Constant Pressure," JPT (Sept. 1970) 1149-1156.
Gentry, R.W.: "Decline-Curve Analysis," JPT (Jan. 1972) 38-41.
Gringarten, A.C.: "Reservoir Limits Testing for Fractured Wells," paper SPE 7452 presented at the 1978 SPE Annual Technical Conference and Exhibition, Houston, TX., 1-3 October 1978.
Fetkovich, M.J.: "Decline Curve Analysis Using Type Curves," JPT (March 1980) 1065-1077.
Nind, T.E.W.: Principles of Oil Well Production, McGraw-Hill Book Company, 2nd Ed., 1981, 31-44.
Carter, R.D.: "Type Curves for Finite Radial and linear Gas Flow Systems: Constant Terminal Pressure Case," SPEJ (October 1985) 719-728.
Fraim, M.L., Lee, W.J., and Gatens, J.M., III: "Advanced Decline Curve Analysis Using Normalized-Time and Type Curves for Vertically Fractured Wells," paper SPE 15524 presented at the 1986
SPE Annual Technical Conference and Exhibition, New Orleans, LA, 05-08 October 1986.
Fetkovich, M.J., Vienot, M.E., Bradley and M.D., Kiesow, U.G.: "Decline Curve Analysis Using Type Curves Case Histories," SPEFE (Dec. 1987) 637-656.
McCray, T.L.: Reservoir Analysis Using Production Decline Data and Adjusted Time, M.S. Thesis, Texas A&M U., College Station, TX (1990).
Palacio, J.C. and Blasingame, T.A.: "Decline Curve Analysis Using Type Curves Analysis of Gas Well Production Data," paper SPE 25909 presented at the 1993 Joint Rocky Mountain
Regional/Low Permeability Reservoirs Symposium, Denver, CO, 26-28 April 1993.
Doublet, L.E., Pande, P.K., McCollum, T.J., and Blasingame, T.A.: "Decline Curve Analysis Using Type Curves Analysis of Oil Well Production Data Using Material Balance Time: Application
to Field Cases," paper SPE 28688 presented at the 1994 Petroleum Conference and Exhibition of Mexico held in Veracruz, Mexico, 10-13 October 1994.
Shih, M.-Y. and Blasingame, T.A.: "Decline Curve Analysis Using Type Curves: Horizontal Wells," paper SPE 29572 presented at the 1995 Joint Rocky Mountain Regional/Low Permeability
Reservoirs Symposium, Denver, CO, 20-22 March, 1995.
Doublet, L.E. and Blasingame, T.A.: "Evaluation of Injection Well Performance Using Decline Type Curves," paper SPE 35205 presented at the 1996 SPE Permian Basin Oil and Gas Recovery
Conference, Midland, TX, 27-29 March 1996.
Crafton, J. W.: "Oil and Gas Well Evaluation Using the Reciprocal Productivity Index Method," paper SPE 37409 presented at the 1997 SPE Production Operations Symposium, held in Oklahoma
City, Oklahoma, 09-11 March 1997.
Agarwal, R.G., Gardner, D.C., Kleinsteiber, S.W., and Fussell, D.D.: "Analyzing Well Production Data Using Combined Type Curve and Decline Curve Analysis Concepts," paper SPE 49222
prepared for presentation at the 1998 SPE ATCE, New Orleans, LA, 27-30 September 1998.
Marhaendrajana, T. and Blasingame, T.A.: "Decline Curve Analysis Using Type Curves Evaluation of Well Performance Behavior in a Multiwell Reservoir System," paper SPE 71514 presented
at the 2001 Annual SPE Technical Conference and Exhibition, New Orleans, 30 September-03 October 2001.
Araya, A. and Ozkan, E.: "An Account of Decline-Type-Curve Analysis of Vertical, Fractured, and Horizontal Well Production Data," paper SPE 77690 presented at the SPE Annual Technical
Conference and Exhibition held in San Antonio, Texas, 29 September-02 October 2002.
Pratikno, H., Rushing, J.A., and Blasingame, T.A.: "Decline Curve Analysis Using Type Curves: Fractured Wells," paper SPE 84287 presented at the 2003 Annual SPE Technical Conference and
Exhibition, Denver, CO., 05-08 October 2003.
Mattar, L. and Anderson, D.: "A Systematic and Comprehensive Methodology for Advanced Analysis of Production Data," paper SPE 84472 presented at the SPE Annual Technical Conference
and Exhibition held in Denver, Colorado, U.S.A., 05-08 October 2003.
Anderson, D. and Mattar, L.: "Practical Diagnostics Using Production Data and Flowing Pressures," paper SPE 89939 presented at the SPE Annual Technical Conference and Exhibition held in
Houston, Texas, U.S.A., 26-29 September 2004.
Fuentes-Cruz, G., Camacho-Velzquez, R., and Vsquez-Cruz, M. : "Pressure Transient and Decline Curve Behaviors for Partially Penetrating Wells Completed in Naturally Fractured-Vuggy
Reservoirs, paper SPE 92116 presented at the 2004 SPE International Petroleum Conference in Mexico, Puebla, Mexico, 08-09 November 2004.

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide
55/54
(Appendix)

Selected Vendors: PA Software

z Vendor A: Fekete

Product: RTA
Contact: Ed Ferguson (ed@fekete.com)
web:
www.fekete.com
z Vendor B: Kappa Engineering
Product: Topaze
Contact: Kevin Siggery (siggery@kappaeng.com)
web:
www.kappaeng.com
z Vendor C: Weatherford/eProduction Solutions
Product: PanSystem
Contact: Carol Marini (Carol.Marini@e-petroleumservices.com)
web:
www.ep-solutions.com

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide
56/54
(Appendix)

Rules for Life: Blasingame

Simple Rules for Life:

Important Rules for Life:

zRighty-tighty lefty-loosey zAlways work harder than

solves most problems in life. those you work for.
zNever own anything that
zYou can learn a lot from a
Chihuahua (no fear, all love). eats while you sleep.
zNever own anything that
zDon't mow the grass until
needs repainting.
the city tells you to.
zLeadership is not defined by
vision, strength, integrity, or
courage JUST PASSION.
zIf you have to herd cats, then
be a rat.
zIf you have to hold yourself
out as an example, then
make sure to be a bad one
"An empty stomach is not a good political
(Mark Twain).
advisor moral know your convictions,
but take care of yourself as well"
Mahatma Gandhi (+ Blasingame)

T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide
57/54
(Appendix)

Rules for Mathematics: Blasingame

Mathematics:
(for reservoir engineering)
1. Cheat (approximate, truncate, etc.). (except for
material balance)
2. Take the Derivative.
3. Take the Integral.
4. Take the Laplace Transform.
5. Go back to Rule 1 (Cheat).

Favorite Numbers:
z0, 1, .
(What else do we need?)
T.A. Blasingame

Performance-Based Reservoir Characterization State-of-the-Technology

Slide
58/54
(Appendix)