Rate Transient Analysis

1- 4: TRADITIONAL DECLINE ANALYSIS 23-32: RADIAL TYPE CURVES

1. Traditional (Arps) Decline Curves Exponential Decline: 23-24: Radial Flow Model: Type Curve Analysis
3. Exponential Decline
§ Decline rate is constant.
§ Log flow rate vs. time is a straight line. Dt 23. Calculations for Oil All radial flow type curves are based on the same reservoir model: 24. Calculations for Gas
log q = log qi -
Harmonic (b=1)
§ Flow rate vs. cumulative production is a straight line. 2.303 (Agarwal-Gardner Type Curves) § Well in centre of cylindrical homogeneous reservoir. (Agarwal-Gardner Type Curves)

Rate
qi § Provides minimum EUR (Expected Ultimate Recovery). § No flow outer boundary.
q=
1 + Di t 141 .2 Bo mo q / Dp § Skin factor represented by rwa. 1 . 417 ´ 10 6 T q / Dp p
Hyperbolic (0<b<1) Hyperbolic Decline: k= ( ) match k= ( ) match
q qi h qD h qD
q= 1 § Decline rate is not constant (D=Kq ).
b Time re
(1 + b Di t) b
§ Straight line plots are NOT practical and b is determined by q = qi - DQ 0 .00633 k t c 0.00633 k t ca
( ) match
nonlinear curve fit. ( ) p ( fmg ct ) i t DA rw
p fmo ct t DA match rw rwa
rwa = , s = ln ( )
rwa = , s = ln ( ) rwa

Rate
Exponential (b=0) (reD) match
b value Reservoir Drive Mechanism (reD ) match rwa
q = qi e -Dt
0
Single phase liquid (oil above bubble point)
§ Information content of all type curves (Figures 25-32) is the same. 0 .00633k t ca p r 2 f hS gi
t Single phase gas at high pressure 0. 00633k t c re = ) match , G = e
(
re = ( ) match , p r f hSoi
2
§ The shapes are different because of different plotting formats. p ( fmg ct ) i t DA Bgi
Cumulative Production
N= e
0.1-0.4 Solution gas drive pfmo ct t DA Boi § Each format represents a different “look” at the data and
0.4-0.5 Single phase gas Note: Gas calculations are ITERATIVE because of pseudo-time.
emphasizes different aspects.
0.5 Effective edge water drive
2. Decline Rate Definitions 4. Harmonic Decline
0.5-1.0 Commingled layered reservoirs 25-26: Blasingame
No Straight Line
Harmonic Decline:
Nominal Decline Rate Effective Decline Rate
25. Blasingame: Rate (Normalized) § qDd and tDd definitions are similar to Fetkovich. 26. Blasingame: Integral-Derivative
q1 - q 2 § Decline rate is directly proportional to flow rate (b=1).

Rate
D=-
1 dq
De = § Normalized rate (q/Dp or q/Dpp) is plotted.
§ Log flow rate vs. cumulative production is a straight line. 1
q1 § Three sets of type curves: t Dd
q dt q1
1. qDd vs. tDd (Figure 25).
q Ddi =
t Dd ò
0
q Dd dt Dd
Summary: Time Increasing rwa
q q reD 2. Rate integral (qDdi) vs. tDd (has the same shape as qDd). d ( q Ddi )
§ Boundary-dominated flow only. Di 3. Rate integral-derivative (qDdid) vs. tDd (Figure 26). qDdid = -
q2 § Constant operating conditions. log q = - Q + log qi d (ln t Dd )
2. 303qi
§ Developed using empirical relationships. 2p

qDd

qDdid
Rate
1 Year
Ha § In general: qDd = qD bDpss , tDd = t
§ Quick and simple to determine EUR. r rm b Dpss DA
t t q Dd = q D ln( reD - 0 . 5 ) , reD = e on
§ EUR depends on operating conditions. rwa ic
(U § bDpss is a constant for a particular well/reservoir configuration.
-D
De = 1 - e § Does NOT use pressure data. nit
Sl
Cumulative Production 2tD op
§ b depends on drive mechanism. tDd = e)
(ln reD - 0.5)( reD2 - 1)

t Dd t Dd
5-10: FETKOVICH ANALYSIS
27-28: Agarwal-Gardner

5. Analytical: Constant Flowing Pressure § qD and tD definitions are similar to well test. 6. Analytical: Constant Flowing Pressure § qD and tDA definitions are similar to well testing. 28. Agarwal-Gardner: Integral-Derivative
27. Agarwal-Gardner: Rate (Normalized)
§ Convenient for transient flow. Increasing rwa § Normalized rate (q/Dp or q/Dpp) is plotted.
Transient § Results in single transient stem but multiple boundary- Boundary-Dominated § Three sets of type curves: Boundary-
• Single Curve Increasing rwa
dominated stems. • Single Curve
• Exponential
1. qD vs. tDA (Figure 27). Transient Dominated
reD
Boundary-Dominated 2. Inverse of pressure derivative (1/pDd) vs. tDA (not shown).
• Multiple Curves
fm 3. Inverse of pressure integral-derivative (1/pDid) vs. tDA (Figure 28).
• Exponential
qDd
qD

Notes:

qD

1/pDid
d ( p D)
Transient
1. Pressure derivative is defined as pDd =
§ qDd and tDd definitions are convenient for production data • Multiple Curves
qD =
1
=
141.2qBμ Ha
rm d (ln t DA ) 1 t DA
p Di = ò
on
analysis. pD kh( p i - pwf ) ic p D dt DA Un
(U
nit 2. Inverse of pressure derivative is usually too noisy and inverse of t DA 0 it
§ Convenient for boundary-dominated flow. Slo
pe pressure integral-derivative is used instead. Slo
pe
0.00633 kt d ( p Di )
§ Results in single boundary-dominated stem but multiple tDA = )
pDid =
transient stems. f μct A d (ln t DA )
tD 2t D t Dd
q Dd = qD ln( reD - 0 .5) t Dd = t DA t DA
(ln reD - 0.5)( reD2 - 1)
29-30: Normalized Pressure Integral (NPI)
7. Empirical: Arps Depletion Stems 8. Empirical: Arps-Fetkovich Depletion Stems
29. NPI: Pressure (Normalized) § pD and tDA definitions are similar to well testing. 30. NPI: Integral-Derivative
§ Normalized Pressure (Dp/q or Dpp/q) is plotted rather than
1 kh( pi - pwf ) 1 t DA
pD =
qD
=
141.2qB μ op
e normalized rate (q/Dp or q/Dpp). p Di =
t DA ò0
p D dt DA Boundary-
Dominated
Harmonic
it
Sl § Three sets of type curves: Transient
e
Un 1. pD vs. tDA (Figure 29). d ( p Di ) Sl
op
pDid = nit
Harmonic
2. Pressure integral (pDi) vs. tDA (has the same shape as pD).
Rate

Replot on Log-Log Scale d (ln t DA )

Rate

U
Hyperbolic
3. Pressure integral-derivative (pDid) vs. tDA (Figure 30).

pD

pDid
Hyperbolic

Exponential
Exponential reD

Increasing rwa
Time Time
t DA t DA

Summary:
10. Fetkovich/Cumulative Type Curves 9. Fetkovich Type Curves 31-32: Transient-Dominated Data
§ Combines transient with boundary-dominated flow.
§ Transient: Analytical, constant pressure solution. Boundary-Dominated
Rate ulativ
e
§ Boundary-Dominated: Empirical, identical to traditional (Arps). • Concave down 31. Rate (Normalized) § Similar to Figures 27 & 28 but uses tD instead of tDA. This format is 32. Integral-Derivative
Cum
§ Constant operating conditions. • Empirical useful when most of the data are in TRANSIENT flow.
§ Used to estimate EUR, skin and permeability. § qD and tD definitions are similar to well testing. 1 tD d ( p Di )
§ EUR depends on operating conditions. § Normalized rate (q/Dp or q/Dpp) is plotted.
p Di = ò p D dt D pDid =
QDd

fm tD 0 d (ln t D )
§ Does NOT use pressure data. § Three sets of type curves:
qDd
and

Zero Slope
§ Cumulative curves are smoother than rate curves. 1. qD vs. tD (Figure 31).
q Dd

§ Combined cumulative and rate type curves give more unique Transient 2. Inverse of pressure integral (1/pDi) vs. tD (not shown).

1/pDid
• Concave up 3. Inverse of pressure integral-derivative (1/pDid) vs. tD (Figure 32).
qD

match (Figure 10). • Theoretical

tDd
Increasing re
QDd = ò 0
q Dd d t Dd
t Dd t Dd

tD tD

11-14: MODERN DECLINE ANALYSIS: BASIC CONCEPTS

11-12: Material Balance Time
33-40: FRACTURE TYPE CURVES
11. Comparison of qD and 1/pD § Material Balance Time (tc) effectively converts constant pressure 12. Equivalence of qD and 1/pD
solution to the corresponding constant rate solution. 33-37: Finite Conductivity Fracture
§ Exponential curve plotted using Material Balance Time becomes
Constant rate solution harmonic. 33. Rate § Fracture with finite conductivity results in bilinear flow (quarter 34. Integral-Derivative
(1/pD) is harmonic. § Material Balance Time is rigorous during boundary-dominated flow. slope).
Ha Ha
lf lf S
Slo
qD and 1/pD

lop
qD and 1/pD

pe
xe e Increasing FCD xe
Actual Rate Decline Constant Rate Qua
rter Increasing FCD xf xf
Q 1 t Slop
tc = = ò qdt e
kw
Un

Un

q q 0 xe § Dimensionless Fracture Conductivity is defined as: FCD = f xe

it S

it S

1/pDid
Constant pressure solution
Q § Constant pressure solution (qD) is converted to kxf
qD
lop

lop

(qD) is exponential. constant rate (1/pD) using Material Balance Time (tc).
Q
e

§ Constant rate and constant pressure solutions using tc

Increasing xe
§ Fracture with infinite conductivity results in linear flow (half slope). Increasing xe
are harmonic.
Actual Time (t) Material Balance Time

Un
it
(t c) = Q /q
Un

Slo
tD t D (Material Balance Time) 0. 00633kt 0. 00633kt
it S

t Dxf = t Dxf =

pe
fm c t xf2 fm c t xf2
lop
e

13-14: Type Curve Interpretation Aids

t Dxf § For FCD>50, the fracture is assumed to have infinite conductivity. t Dxf

13. Concept of Rate Integral Rate (Normalized) 14. Derivative and Integral-Derivative
q
§ Combines rate with flowing pressure. 35. Elliptical Flow: Integral-Derivative 36. Elliptical Flow: Integral-Derivative 37. Elliptical Flow: Integral-Derivative
Dp
Rate Integral = Integral (Normalized Rate) Integral-Derivative
Increasing a/b
§ Smoothes noisy data but attenuates the 1 tc q
ò
b
Rate Integral

Actual Rate = dt Increasing a/b

reservoir signal. tc 0 Dp
Rate

a xf
Ha
Increasing a/b lf S
lop
Derivative (Normalized Rate) e
q

qDid
qDid
qDid

§ Amplifies reservoir signal but amplifies noise d( )

Dp
Actual Time (t) Actual Time (t) - Derivative Quarter
as well. d (ln t c )
Slope

§ Rate Integral is cumulative average rate. Integral-Derivative (Normalized Rate)

d 1 tc q
§ Rate Integral is an effective way to remove noise. § Smoothes the scatter of the derivative. - (
d (ln tc ) t c ò 0 Dp
dt )
FCD = 5.0 FCD = 50.0
Time FCD = 0.5

t DA t DA t DA

38-40: Infinite Conductivity Fracture

15-18: GAS FLOW CONSIDERATIONS
38. Blasingame: Rate and Integral-Derivative 39. NPI: Pressure and Integral-Derivative 40. Wattenbarger: Rate
15-16: Pseudo-Pressure
Ha Hal 0.00633kt
lf S
lop re
f Sl
ope t Dye =
15. Darcy’s Law Gas properties vary with pressure: 16. Pseudo-Pressure (pp) e re fmct ye2
§ Z-factor (Pseudo-Pressure, Figures 15 & 16) e xe=2xf
xf Increasing xf op Increasing yw / ye
§ Liquid (Constant Viscosity): § Viscosity (Pseudo-Pressure & Pseudo-Time, Figures 15, 16 & 18) Increasing xf xf
it
Sl
qDdid

qD ( ye / xe)
pDid

Un
§ Compressibility (Pseudo-Time, Figures 17 & 18)
§ Gas: Viscosity and Z-factor are not constant.
and
and

Define Pseudo-Pressure (pp):

Pseudo-Pressure (pp)

pD
qDd

ye
p pdp § Pseudo-pressure corrects for changing viscosity and Z-factor with
pp = 2ò
Un
it S
Boundary- lop
p0 μgZ pe
Un

pressure. Slo Dominated yw e

it

alf
§ In all equations for liquid, replace pressure (p) with pseudo- Transient
Sl

H
op

Pseudo-Pressure (pp) corrects for changing viscosity pressure (pp).

e

( µ g ) and Z-factor with pressure. t Dd t DA t Dye

1.417 ´ 10 qT
6
1
Note: For gas, qD = p =
kh( ppi - ppwf )
Darcy’s Law for Gas: D pp µ q D

Pressure ( p)

41-43: HORIZONTAL WELL TYPE CURVES

17-18: Pseudo-Time

§ Compressibility represents energy in reservoir. 41. Blasingame: Integral-Derivative 42. Blasingame: Integral-Derivative 43. Blasingame: Integral-Derivative
17. Gas Compressibility Variation 18. Pseudo-Time (ta)
§ Gas compressibility is strong function of pressure (especially at
1 1 ¶Z 1 LOW PRESSURES). § In all equations for liquid, replace Time (t) with Pseudo-
cg = - » § Ignoring compressibility variation can result in significant error in Time (ta). xe xe
p Z ¶p p xe
L Increasing L L L
original gas-in-place (G) calculation. t dt Increasing L
§ Pseudo-time (ta) corrects for changing viscosity and compressibility t a = ( μg c t ) i ò μg c t
Increasing L
Gas Compressibility

0
xe xe xe
with pressure.
qDdid

qDdid
qDdid

§ Pseudo-time calculation is ITERATIVE because it depends on µg and § Convert material balance time (tc) to Material Balance
ct at average reservoir pressure, and average reservoir pressure Pseudo-Time (tca).
depends on G (usually unknown). 1 ta ( μ g ct )i t qdt
q ò0 ò
t ca = qdt a = L
q 0 μ g ct L L
Note: Pseudo-time in build-up testing is evaluated at well flowing = 0.1 = 0.3 = 1.0
xe xe xe
pressure NOT at average reservoir pressure.
Note: µg and ct are evaluated at average reservoir pressure
t Dd t Dd t Dd
Pressure (unlike build-up testing).

19-22: FLOWING MATERIAL BALANCE 44-45: WATER-DRIVE TYPE CURVES

20. Gas: Determination of bpss 44. Blasingame: Rate 45. Agarwal-Gardner: Rate
19. Oil: Flowing Material Balance
Oil Np 20
pi - pwf = + qbpss Infinite Aquifer
Transient ct N M=10 (Constant Pressure)
Boundary-Dominated
q Np 1 Reservoir
=- +
pi - pwf c t Nbpss ( pi - p wf ) bpss
Normalized Rate

Boundary-Dominated Increasing rwa

)

qD
qDd

bpss M=0 (Volumetric Depletion)

Note: bpss is the inverse of productivity index and is constant during § Mobility ratio (M) represents the strength of the aquifer.
boundary-dominated flow. Increasing M
N k aq μres
(

Increasing M
M =
kres μaq
aG pa 2 pi
a=
ta
Gas
ppi - ppwf = + qbpss Transient ( μg ct Z ) i
Gpa = ò 0
qdt a r eD = 48 § M = 0 is equivalent to Radial Type Curves (Figures 25-32).
M=0 (Volumetric Depletion)
Gi
Np /c t ( p i - pwf ) t DA
a Gpa /q t Dd
Normalized Cumulative Production
p pi - ppwf aGpa
= + bpss
q qG
21. Gas: Flowing Material Balance p p = p pwf + qbpss 22. Procedure to Calculate Gas-In-Place NOMENCLATURE
Calculating G is ITERATIVE:
1. Estimate G; plot p/Z vs. Gp from pi/Zi to G. a semi-major axis of ellipse Gp gas cumulative production pDi dimensionless pressure integral re exterior radius of reservoir xe reservoir length
Summary: A area Gpa pseudo-cumulative production pDid dimensionless pressure integral- reD dimensionless exterior radius of xf fracture half length
§ Uses flowing data. No shut-in required. 2. At any time, Gp is known. Determine p at Gp from p/Z b hyperbolic decline exponent or semi - h net pay derivative reservoir ye reservoir width
§ Applicable to oil and gas. plot. minor axis of ellipse k permeability pi initial reservoir pressure rw wellbore radius yw well location in y-direction
bDpss dimensionless parameter kaq aquifer permeability pp pseudo-pressure rwa apparent wellbore radius Z gas deviation factor
§ Determines hydrocarbon-in-place, N or G. 3. Obtain µg and ct at p.
bpss inverse of productivity index kf fracture permeability pp pseudo-pressure at average reservoir s skin Z gas deviation factor at average
§ Oil (N): Direct calculation. 4. Convert t to ta and pwf to ppwf (Figures 18 & 16). B formation volume factor kh horizontal permeability pressure Sgi initial gas saturation reservoir pressure
§ Gas (G): Iterative calculation because of pseudo-time. Bgi initial gas formation volume factor kres reservoir permeability ppi initial pseudo-pressure Soi initial oil saturation Zi initial gas deviation factor
G
§ Simple yet powerful. 5. Determine bpss from Figure 20.
Bo oil formation volume factor kv vertical permeability ppwf pseudo-pressure at well flowing t flow time constant
§ Data readily available (wellhead pressure can be converted to 6. Determine p from pp = ppwf + qbpss. Boi initial oil formation volume factor K constant pressure ta pseudo-time porosity
bottomhole pressure). 7. Plot p/Z vs. Gp and determine new G. cg gas compressibility L horizontal well length pwf well flowing pressure tc material balance time viscosity
§ Supplements static material balance. ct total compressibility M mobility ratio q flow rate tca material balance pseudo-time
§ Ideal for low permeability reservoirs. 8. Repeat steps 2-7 until G converges. ct total compressibility at average N original oil-in-place qD dimensionless rate tD dimensionless time aq aquifer fluid viscosity
reservoir pressure Np oil cumulative production qDd dimensionless rate tDA dimensionless time g gas viscosity
D nominal decline rate p pressure qDdi dimensionless rate integral tDd dimensionless time g gas viscosity at average reservoir
De effective decline rate p average reservoir pressure qDdid dimensionless rate integral-derivative tDxf dimensionless time pressure
Di initial nominal decline rate p0 reference pressure qi initial flow rate tDye dimensionless time o oil viscosity
FCD dimensionless fracture conductivity pD dimensionless pressure Q cumulative production T reservoir temperature res reservoir fluid viscosity
G original gas-in-place pDd dimensionless pressure derivative QDd dimensionless cumulative production w fracture width

Oil field units; qg (MMSCFD); t (days)

Copyright © 2008 Fekete Associates Inc. Printed in Canada