PEGN 422 Well Test Analysis and Design
Fall 2016
Module 1: Introduction
Dr. Erdal Ozkan
Elio Dean
Colorado School of Mines
Golden, CO
Module 1 Objectives and Outcomes
Objective
To provide attendees a high level summary of critical topics associated with well testing
By the end of this Module, participants should be able to:
Explain the role of well testing in petroleum engineering
Define the objectives of well testing
Describe the general approaches for reservoir characterization
Explain well testing applications for field development and reservoir management
List different types of well tests and compare their advantages and disadvantages
Understand fundamental methodology of transient flow analysis
Explain the motivation and use of pressure derivatives in well testing
Describe the bases and methodologies of manual and computerized well tests analyses
Introduction
Petroleum Engineer
A Hybrid Discipline
1. Calls upon the contribution of professionals with different expertise and background
2. Requires a series of highly-technical decisions, evaluations and operations performed by the team of special
expertise and creativity
Major Branches
1. Drilling Engineering
2. Reservoir Engineering
Well testing is a sub-function of reservoir engineering
3. Production Engineering
Introduction
Oil and Gas Reservoir Engineering
Basic Job of a Reservoir Engineer:
Determine
1.
The hydrocarbons in place
2.
The rate they can be produced at the surface
3.
The ultimate recovery factor
4.
The consequences of stimulation
These questions are answered repeatedly throughout the life of the field
Introduction
The work of a Reservoir Engineer
Begins with collecting:
The geological and geophysical information
(geological description, faults, reservoir or drainage
boundaries)
The engineering data
(production rates, wellhead pressures, cumulative
production, well completion, etc.)
Then:
Estimation of initial hydrocarbon volume in place
(by volumetric procedures)
Introduction
The work of a Reservoir Engineer
Knowing the hydrocarbon reserves, determine the rate they can be produced
Productivity index (single-phase, oil flow):
qo = J p - pwf
J: productivity index
qo: production rate
pwf: bottomhole flowing pressure
p: average reservoir pressure
Introduction
The work of a Reservoir Engineer
Basic question of reservoir engineering:
qo = J p - pwf
is low productivity because of
low average reservoir pressure, p?
low productivity index, J?
Introduction
The work of a Reservoir Engineer
Productivity index, J, for single-phase, oil flow:
J=
-3
qo
(p- p )
wf
Productivity index depends on:
7.08 10 kh
=
1
4A
Bm ln g
+
s
2
2
e C Arw
Formation flow capacity, kh
Fluid viscosity,
Well condition (skin effect), s
Drainage area, A
Well location
Introduction
Well Testing
Well testing is a branch of reservoir engineering.
General objectives of well testing:
Characterizing formation
Evaluating and predicting well performance
Inflow and outflow performance curves for systems analysis
[http://petrowiki.org/Nodal_analysis]
Introduction
Objectives of Well Testing
rw
Conventional Objectives
Estimate:
Pe
Formation flow capacity, kh
Formation damage due to drilling and completion
Average reservoir pressure
Hydraulic connectivity of the reservoir
Distance to a physical boundary
Pw
Drainage area
Production capacity of the well
Pw
re
Introduction
10
Objectives of Well Testing
Core samples
characterized
small area
Contemporary Use of Well testing
As a tool to aid:
Well test
characterize
larger area
Reservoir characterization
(scale and averaging issues)
Reservoir (Flow) modeling
(non-uniqueness of responses)
High resolution Static Model vs. Scaled Dynamic Simulation Model
Introduction
[http://petrowiki.org/File%3AVol6_Page_094_Image_0001.png]
11
Use of Well Testing for Reservoir Description
Reservoir Characterization
Objective:
Obtain a better description of the reservoir
Typically:
i) variations in permeability, porosity
ii) barriers to flow (faults, boundaries, shale lenses, etc.)
Introduction
12
Reservoir Characterization
i.
Distribution of properties to the entire field
Measurements (observations) are only at limited points
ii.
Scaling problem
Are measurements only valid for the scale they have been measured?
iii.
Parameter variable
Intensive vs. Extensive
Deterministic vs. Stochastic
Introduction
State Input Response
Dependent vs. Independent
13
Reservoir Characterization
Data from well testing and other sources
i)
Other sources (cores, logs, seismic, etc.)
Static measurements (small scale, local)
ii) Well test
Dynamic measurements (large scale, volume of investigation)
Introduction
14
Reservoir Characterization
Well test data
Concerns:
i)
Input data from static measurements are used in the interpretations
ii)
Well test estimates of properties represent an average in the volume of investigation
iii) Many applications require small-scale (local) properties
Introduction
15
Reservoir Characterization
Well test data
Permeability Averaging:
i)
Upscaling: Assigning permeability values to unsampled locations in the reservoir
Concern: Characterization of reservoir by a finite number of point measurements (core-estimates) of permeability
ii)
Downscaling: Estimating local values of permeability from distributions (averages)
Concern: Meaning of well-test-estimated (average) permeability
Introduction
16
Reservoir Characterization
Well test data
Permeability Upscaling (simple techniques):
N
i)
ii)
Arithmetic average:
Harmonic average:
k=
k h
i i
i=1
k A=
ht
1
N
i=1
iii) Geometric average:
Introduction
(layers in parallel)
(layers in series)
1
Ai ki
1 N
log k = logki
N i=1
(random or log normal permeability distribution)
17
Reservoir Characterization
Well test data
Permeability Upscaling (simple techniques):
Concerns:
I.
Assumption of steady flow
II.
Alteration of original core properties
III.
Core plug is assumed to represent a volume that is more than 100 times greater than its own volume
IV.
Cores are usually taken from the best part of the formation
Introduction
18
Types of Well Tests
The type of the test depends on the information expected from it.
Some common tests:
Drawdown Tests
Buildup Tests
Injection Tests
Falloff Tests
Interference Tests
Drill Stem Tests (DST)
Introduction
19
Types of Well Tests
Drawdown Tests
Produce the well at a constant (or variable) rate and record pressure change as a function of time
rate
pressure
time
Introduction
20
Types of Well Tests
Drawdown Tests
Information Obtained:
The reservoir behavior
Permeability and formation damage
Properties of natural and induced fractures
Reservoir pressure
Location of the boundaries (drainage area)
Introduction
21
Types of Well Tests
Drawdown Tests
Limitations:
It is difficult to keep the rate constant
Unstabilized conditions before the test interfere with the pressure responses
Introduction
22
Types of Well Tests
Buildup Tests
After a period of production, shut in the well and record pressure buildup as a function of time
rate
pressure
time
Introduction
23
Types of Well Tests
Buildup Tests
Information obtained
Same as drawdown tests
Limitations
Loss of production during the test
Introduction
24
Types of Well Tests
Injection Tests
Record pressure response to the injection of a fluid into the reservoir
pressure
rate
Introduction
time
25
Types of Well Tests
Injection Tests
Information obtained
Same as drawdown and buildup tests
Mobility of fluid banks
Limitations
Multiphase flow effects complicate interpretation
Introduction
26
Types of Well Tests
Falloff Tests
Observe pressure drop when the injection stops
pressure
time
rate
Introduction
27
Types of Well Tests
Falloff Tests
Information obtained
Same as injection tests
Limitations
Multiphase flow effects complicate interpretation
Introduction
28
Types of Well Tests
Interference Tests
Observe the pressure change in shut-in well(s) due to production from a neighboring well
Introduction
29
Types of Well Tests
Interference Tests
Information obtained
Hydraulic connectivity
Formation properties
Directional changes in properties
Limitations
Takes considerably longer
Introduction
30
Types of Well Tests
Drill Stem Test (DST)
DST is a sequence of drawdown and buildup tests
Introduction
31
Types of Well Tests
Drill Stem Test (DST)
Information obtained
Similar to drawdown and buildup
1.
2.
3.
4.
5.
Limitations
the reservoir fluids,
initial or average reservoir pressure,
effective permeability to the flowing fluid,
skin factor (mechanical damage), and
an estimate of the potential flow rates.
Analysis requires special techniques
Momentum and friction effects interfere with the data
Unstabilized conditions resulting from recent drilling
Introduction
32
Methodology of Transient Flow Analysis
General Methodology
Create a change in production rate
(starting, stopping, or changing the rate of production )
Measure the response of the reservoir at a point (well) in the reservoir
(measured reservoir response may be pressure, rate, or both)
Analyze the reservoir response by comparing with the known (theoretical) responses
Introduction
33
Example:
Flowing Time
t, hr
Flowing Pressure
pwf, psi
3500
2917
2900
2888
2879
2869
7.5
2848
10
2830
15
2794
20
2762
1500
30
2703
20
40
2650
Wellbore radius, rw, ft
0.33
50
2597
Porosity, f, fraction
0.18
60
2545
15 x 10-6
70
2495
80
2443
1.2
90
2392
100
2341
Pressure-Transient Test Data
Measured pressure vs.
time data during a
flow (drawdown) test
Formation, fluid, and
well properties required
for analysis
Flow rate, q, stb/d
Formation thickness, h, ft
Total compressibility, ct, psi-1
Viscosity, m, cp
Formation volume factor, B, rb/stb
Introduction
34
Methodology of Transient Flow Analysis
Key Points of Well-Test Interpretation
For the success of the analysis, accurate data are essential
Measurements do not directly give the desired information; the data need to be analyzed and interpreted
Interpretation is based on the comparison of the measured responses with a theoretical model
The theoretical model is a solution of governing flow equation (diffusion equation)
Introduction
35
Methodology of Transient Flow Analysis
Pressure Derivative Analysis
Derivative idea has been introduced by many researchers: Chow (1952), Jones (1957, 1962),
Carter (1966), Prasad (1979), Tiab and Kumar (1980), Bourdet et. al. (1983)
Similar to pressure vs. time data, pressure derivative vs. time data display special characteristics
during different flow regimes
Pressure derivative has been an integral part of pressure transient analysis since the 1980s
Introduction
36
Methodology of Transient Flow Analysis
Motivation of Pressure Derivative Analysis:
Derivative magnifies small changes in pressure vs. time response
G(t) = g(t') dt'
f, g
F, G
g(t)
f(t)
F(t) = f(t') dt'
0
t
Introduction
t
37
Methodology of Transient Flow Analysis
Calculating Pressure Derivatives
Calculation of pressure derivative from field data magnifies the noise also
f
df
lim
t 0 t
dt
f
df /dt
f /t
t
Introduction
: noise in data
f df
lim
lim
t 0
t
dt t 0 t
If is relatively constant,
/ t increases as Dt 0
38
Methodology of Transient Flow Analysis
Calculating Pressure Derivatives
Modern application of pressure-derivative ideas are in terms of logarithmic derivative
dp
dp
p p
t
d ln t
dt
It is recommended to calculate the logarithmic derivatives using the differences in log time
dp
d ln t
Introduction
dp
d t
39
Methodology of Transient Flow Analysis
Calculating Pressure Derivatives
Pressure change, Dp
Bourdet et. al. algorithm (1984, SPE 12777 )
2
i
1
Dx1
Dp1Dx2
For smoother derivatives,
x1 , x2 L
Logarithm of time, log t
Introduction
Dp2
p
p
x2
x1
dp x 1
x 2
dx
x1 x2
Normally, L 0.2
40
Methodology of Transient Flow Analysis
Calculating Pressure Derivatives
Bourdet et. al. algorithm (1984, SPE 12777 )
The logarithmic derivatives of measured pressure vs. time data can be computed by the following formula:
Dpi - Dpi-1
Dpi+1 - Dpi
ti
dDp
=
ti+1 - ti ) +
ti - ti-1 )
(
(
d lnt ( ti - ti-1 ) + ( ti+1 - ti ) ti - ti-1
ti+1 - ti
To obtain smoother derivatives, the time intervals (ti - ti-1) and (ti+1- ti) should not be too small.
Bourdet et al. suggested:
t
t
ln i +1 and ln i L
ti
ti -1
Introduction
0 L 0.5
(usual value: L = 0.1)
41
Methodology of Transient Flow Analysis
Calculation of Pressure Derivatives
Horne algorithm (1995)
ln
t
t
p
ln
t
t
t
dp
i i k
i j
i j i k i pi
ln ti j ti lnti ti k
d ln t i ln ti j ti ln ti j ti k
lnti j ti pi k
lnti ti k lnti j ti k
ln ti j ln ti 0.2 and ln ti ln ti k 0.2
Introduction
42
Methodology of Transient Flow Analysis
Manual Analysis Techniques
Straight Line Analysis
Theoretical models indicate certain straight line relations when pressure is plotted as a function of time on specific
coordinates. The slopes of the straight lines are functions of the properties sought
Different flow regimes require different plots
Linear flow:
p vs.
Bilinear flow:
p vs. t 1 4
Boundary dominated flow:
Radial flow:
Introduction
p vs. t
p vs. log t
43
Methodology of Transient Flow Analysis
Manual Analysis Techniques
Straight Line Analysis
Example: Radial, single-phase flow toward a fully penetrating well of radius rw in an infinite-acting reservoir
Theoretical model:
pwf pi pwf
162.6qBm
k
log t log
3
.
23
0
.
87
s
2
kh
fm
c
r
t
w
pwf (or pwf ) vs. log t should give a straight line with slope
162.6qBm
m
kh
kh can be calculated from the slope and the skin factor, s, can be calculated from the theoretical model equation
Introduction
44
Methodology of Transient Flow Analysis
Manual Analysis Techniques
Straight Line Analysis
p, psi
Example: Radial, single-phase flow toward a fully penetrating well of radius rw in an infinite-acting reservoir
pwf
162.6qBm
kh
m log t log
3.23 0.87s
2
fmct rw
log t
This is known as semi-log analysis
Introduction
45
Methodology of Transient Flow Analysis
Manual Analysis Techniques
Type-Curve Matching
Uses the generalized solutions (theoretical models) in the form of type curves
The generalized solution (type curve) is in terms of dimensionless variables
Example:
kh
pD
p
141.2qBm
tD
Introduction
2.637 10 4 k
fct mrw2
t
46
Methodology of Transient Flow Analysis
Manual Analysis Techniques
The basis of type-curve matching:
kh
log pD = log
+ log pi - p
141.2qBm
log
tD
rD2
= log
2.637 10-4 k
fmcr
+ logt
A log-log plot of pi - p vs. t (the field curve) should look exactly like the log-log plot of pD vs. t D r
2
D
The shift of the field curve in the vertical and horizontal axes can be used to determine formation properties
Introduction
47
Methodology of Transient Flow Analysis
Manual Analysis Techniques
log p
log pD
Type-Curve Matching
Field plot
M
x
pDM, D pM
tDM, tM
Type curve
log t
log tD
pD
kh
=
Dp
141.2qBm
M
-4
t r2
2.637
10
k
D D
=
t
f ct mr 2
M
Reservoir properties may be estimated from the match point values
Introduction
48
Methodology of Transient Flow Analysis
Computerized Analysis
Basis: Inverse solution
field
input
assumed
input
reservoir
mechanism
reservoir
theoretical
model
computed
response
response
Generate a set of synthetic responses assuming the unknown properties and underlying flow model
Find the best combination of the assumed properties and flow model that yields the best regression fit
between the synthetic and measured responses
Introduction
49
Methodology of Transient Flow Analysis
Computerized Analysis (Regression Analysis)
Assume reservoir properties and generate pwf vs t data (synthetic)
log pwf
log t
1. Initialization (no match).
log pw f
log t
3. 3rd iteration (improving).
Introduction
log pwf
log t
2. 2nd iteration (no match).
log pwf
log t
4. (after many iterations) Final
match, may not be unique.
50
Methodology of Transient Flow Analysis
Computerized Analysis (Regression Analysis)
Introduction
51
End of Module 1
Back-up Slides and Extra Examples
Dr. Erdal Ozkan
Elio Dean
Colorado School of Mines
Golden, CO
