Rate Transient Analysis

Theory/Software Course

RTA Theory / Software Course: Part 1

• Introduction
• Review of Traditional Decline Analysis Techniques
– Arps
– Fetkovich
• Modern Decline Analysis Theory
– Pseudo S.S. Equation for Oil
– Significance of Harmonic Equation
– Comparison of Constant Rate and Constant Pressure Conditions
– Concept of Material Balance Time
– Extending Concepts to Work for Gas Wells (Pseudo-time)
• Data Analysis Methods - Theory
– Blasingame
– Agarwal-Gardner
– NPI (Normalized Pressure Integral)
– Transient (tD format)
– Flowing Material Balance

1
 Traditional Decline Analysis

Exponential, Hyperbolic and

Harmonic Equations

exponential q = qie − Dit

hyperbolic
qi
q harmonic q=
(1 + bDit )1/ b
qi
q=
1 + Dit
t

2
 SPEE Definitions of Decline Rate

Nominal (true) Decline Rate:

dq
D = -(slope of line)/q

D = − dt
q q
q

D = − ln ( 1 − De)

t
Effective Decline Rate:
qi De = -(slope of line)/qi
q i − qf q
De =
qi
qf
−D
De = 1 − e
t
In RTA, “De” is refered to as “d”,
and is expressed as a percentage

Exponential Decline

-D (decline rate) is constant with time D = Kq0

Rate (q) has a

q = qie − Dit linear
relationship to
q = qi − DiQ Cumulative
Production (Q)

3
 Hyperbolic Decline

-D (decline rate) varies with time D= Kqb

Linear
qi relationship
q= cannot easily be
(1 + bDit )1/ b formed with
hyperbolic
parameters

Harmonic Decline

-D (decline rate) is proportional to rate D= Kq

Linear
qi
q= relationship
(1 + Dit ) between log rate
(q) and
cumulative (Q)

4
Notes About Recovery Mechanism and b Value
(from Arps)
-Single-phase liquid production, high-pressure gas, tubing-restricted
gas, poor waterflood performance: b = 0

-Solution gas drive: 0.1 < b < 0.4; depends on relative permeability
krg/kro curves

-Production data above bubble point should not be analyzed with data
below (Arps decline analysis is only valid when recovery mechanism
doesn’t vary with time)

-Typical gas wells: 0.4 < b < 0.5

-Conventional oil reservoirs under edge water drive (effective water

drive): b = 0.5

-Commingled, layered reservoirs: 0.5 < b < 1.0

-Field experience presented by Arps suggests 0.1 < b < 0.9

-Exponential decline appears to be a rare occurrence in nature, even

though it is the most commonly used decline technique

Fetkovich Theory

-Developed because traditional decline curve

analysis is only applicable when well is in
boundary dominated flow

- Fetkovich used analytical flow equations to

generate typecurves for transient flow, and
combined them with empirical decline curve
equations from Arps

-Resulting typecurves encompass entire

production life of well

5
 Fetkovich Theory – Depletion Stems

exponential
hyperbolic

harmonic
q
log(q)

t log(t)

Fetkovich Theory – Boundary Dominated and Transient

Transient flow- Analytical Stems

log(qDd)

Boundary Dominated Flow –

Empirical Stems

log(tDd)

6
 Modern Decline Analysis

Modern Decline Analysis

- Incorporates the effect of flowing pressure

- Uses pressure transient theory

- Relies on the equivalence between the

constant rate and constant pressure solutions

7
 Equivalence of Constant Pressure and Constant Rate

Pressure

pi
transient
transient
boundary
dominated
boundary (p.s.s)
dominated

time
increases
pwf

Constant Flowing Pressure Constant Rate

Production Welltesting

Modern Decline Analysis: How to Choose a Base Model

- Constant Pressure Model

- Emulates unrestricted flow to pipeline
- Most production data behaves this way
- Difficult to model analytically

- Constant Rate Model

- Emulates deliverability restricted
production
- Common assumption of welltesting
- Easier to model because many solutions
already exist in welltest literature

8
 Data Analysis Methods:
-Blasingame
-Agarwal Gardner
- Flowing Material Balance
-NPI
-Transient

Blasingame Typecurve Analysis

Blasingame typecurves have identical format to those of Fetkovich.

However, there are three important differences in presentation:

1. Models are based on constant RATE solution instead of

constant pressure

2. Exponential and Hyperbolic stems are absent, only

HARMONIC stem is plotted

3. Rate Integral and Rate Integral - Derivative typecurves

are used (simultaneous typecurve match)

Data plotted on Blasingame typecurves makes use of MODERN

DECLINE ANALYSIS methods:

- NORMALIZED RATE (q/∆p)

- MATERIAL BALANCE TIME / PSEUDO TIME

9
 Diagnostics using Typecurves

Radial Model
Blasingam e Typecurve Match

10-7
8
5
3
2

10-8
8
5
Transient
qDd 32 (concave up) Boundary Dominated
(concave down)
10-9
8
5

Base Model:
3
2

10-10
8
5
- Vertical Well in Center of Circle
3
2
- Homogeneous, Single Layer
10-11
8
5
3
2

4 56 8 -1 2 3 45 79 2 3 45 7 9 1 2 3 4 56 8 2 2 3 4 56 8 3 2 3 4 56 8 4 2 3 4 56 8 5 2 3 4 56 8 6 2 3 45 7 2 3 45 7
10 1.0 10 10
tDd
10 10 10 10 107

Diagnostics using Typecurves

Material Balance Diagnostics

Radial Model
Blasingam e Typecurve Match

10-7
8
5
3
2 Reservoir With
10-8
8
Pressure Support
5

qDd 32
Dual
10-9 Deple
Syste tion
8
5 m
3
2
Infin
10-10 Pre ite A
8 ssu ctin
re S g
Vo

5 upp
ort
lu

3
m

2
et
ric

10-11
8
5
Leaky Reservoir
3 (interference)
2

4 56 8 -1 2 3 45 79 2 3 45 7 9 1 2 3 4 56 8 2 2 3 4 56 8 3 2 3 4 56 8 4 2 3 4 56 8 5 2 3 4 56 8 6 2 3 45 7 2 3 45 7
10 1.0 10 10
tDd
10 10 10 10 107

10
 Diagnostics using Typecurves

Productivity Diagnostics
Radial Model
Blasingam e Typecurve Match

10-7
8
Increasing Damage (difficult to identify)
5
3
2

10-8
8
5
Productivity
qDd 3
2 Shifts (workover,
10-9
8 unreported
5
tubing change)
3 Well Cleaning Up
2

10-10
8 Liquid Loading
5
3
2

10-11
8
5
3
2

4 56 8 -1 2 3 45 79 2 3 45 7 9 1 2 3 4 56 8 2 2 3 4 56 8 3 2 3 4 56 8 4 2 3 4 56 8 5 2 3 4 56 8 6 2 3 45 7 2 3 45 7
10 1.0 10 10
tDd
10 10 10 10 107

Diagnostics using Typecurves

Transient Flow Diagnostics

Radial Model
Blasingam e Typecurve Match

10-7 Fracture Linear Flow

8
5 (Stimulated)
3
Transitionally
2 Dominated Flow (eg:
10-8
8 Channel or Naturally
5
Damaged Fractured)
qDd 3
2

10-9
8
Radial Flow
5
3
2

10-10
8
5
3
2

10-11
8
5
3
2

4 56 8 -1 2 3 45 79 2 3 45 7 9 1 2 3 4 56 8 2 2 3 4 56 8 3 2 3 4 56 8 4 2 3 4 56 8 5 2 3 4 56 8 6 2 3 45 7 2 3 45 7
10 1.0 10 10 10 10 10 10 107
tDd

11
 Diagnostics using Typecurves

“Bad Data” Diagnostics

Radial Model
Blasingam e Typecurve Match

10-7
8
5
∆p in reservoir is too low
3
-Tubing size too small ?
2 - Initial pressure too low ?
10-8 - Wellbore correlations
8
5
overestimate pressure loss ?
qDd 3
2

10-9
8
5
∆p in reservoir is too high
3
2 -Tubing size too large ?
10-10
- Initial pressure too high ?
8
- Wellbore correlations
5
3
underestimate pressure loss ?
2

10-11
8
5
3
2

4 56 8 -1 2 3 45 79 2 3 45 7 9 1 2 3 4 56 8 2 2 3 4 56 8 3 2 3 4 56 8 4 2 3 4 56 8 5 2 3 4 56 8 6 2 3 45 7 2 3 45 7
10 1.0 10 10
tDd
10 10 10 10 107

Flowing Material Balance

FMB analysis plots a normalized RATE versus normalized

CUMULATIVE PRODUCTION, on a LINEAR scale (x and y). No
typecurves are plotted.

The FMB methodology combines concepts from two individual

methods:

1. “Old” flowing material balance (after Mattar and

McNeil)

2. Agarwal-Gardner Rate vs. Cumulative production

typecurves

- The FMB plot provides an easy and effective way for estimating
fluid-in-place, using data that is BOUNDARY DOMINATED

- FMB methodology utilizes the concepts of material balance time

and pseudo-time.

12
Flowing Material Balance - “Old” Constant Rate Format

Initial pressure

Constant rate
(varying pressure)
p/z
flo
wi
ng
pr
es
su
re
s

OGIP

Cumulative Production

Flowing Material Balance- Exponential Decline

Constant pressure
(varying rate)
q

EUR OGIP

Cumulative Production

13
 Agarwal-Gardner Flowing Material Balance

variable rate and

variable pressure
q/∆p

OGIP

Cumulative/(C*∆p)

14