PT 4

PRESSURE TRANSIENT ANALYSIS SOFTWARE

USER’S GUIDE

Copyright 1991-2007 The Comport Computing Company

All Rights Reserved

The Comport Computing Company

12230 Palmfree St., Houston, Texas 77034, U. S. A.
(713) 947-3363
www.comportco.com

Page 1
 PT 4 Software

LICENSE AGREEMENT

This software is protected by United States copyright law and international treaty
provisions. The ComPort Computing Company authorizes you to make archival
copies of the software for the sole purpose of backing-up your software and
protecting against software loss. You may move this software from computer to
computer freely, as long as there is no possibility of the software being used by
two different people at two locations at the same time. This software may not be
copied for the purpose of sales or distribution without the express written consent
of the ComPort Computing Company.

For technical support issues, please contact [email protected].

LIMITED WARRANTY

The ComPort Computing Company warrants the diskette or other distribution

medium to be free from defects in materials or workmanship for a period of 60
days from the date of purchase. In the event of such defect, the ComPort
Computing Company will replace the defective medium. Contact the ComPort
Computing Company for instructions and a return authorization.
The ComPort Computing Company shall not be liable for any damages arising
from the use of this product. All other warranties, express or implied, are
specifically disclaimed.

ACKNOWLEDGMENTS
Windows, Windows NT, Windows 95, Windows ME, Windows 2000 and Windows XP are
all registered trademarks of the Microsoft Corporation. Echometer is a registered trademark
of the Echometer Company.

Page 2
 TABLE OF CONTENTS

INTRODUCTION...................................................................................................4

INSTALLATION......................................................................................................5

USING PT..............................................................................................................6
Overview.........................................................................................................6
Menu Commands...........................................................................................6
File Menu.................................................................................................7
Analysis Menu.........................................................................................7
Window....................................................................................................7
Language.................................................................................................8
Help.........................................................................................................8
Test Data Window...........................................................................................8
General Data Tab.....................................................................................8
PVT Data Tab..........................................................................................9
Rates/Pressures Data Tab.......................................................................9
The PT Toolbars............................................................................................10
Cartesian Analysis.........................................................................................11
Square Root of Time Analysis.......................................................................12
Semilog Analysis...........................................................................................12
Log-log Analysis............................................................................................13
Adjust and Compare Analysis.......................................................................13
Fluid Systems and Pseudopressure.............................................................16
Multiple Rates...............................................................................................18
User Defined Units........................................................................................19
Importing Pressure Data...............................................................................20
Echometer® Data Analysis ..........................................................................21
Gage Data Editing .......................................................................................23

PRESSURE TRANSIENT ANALYSIS TUTORIAL...............................................24

Selecting and Loading Data..........................................................................24
Preliminaries - Log-log Analysis....................................................................25
Semilog Analysis...........................................................................................27
Wellbore Storage - Cartesian Analysis.........................................................29
Improving Things - Adjust and Compare.......................................................30
APPENDICES
I - PT NOMENCLATURE..............................................................................34
II - AN OVERVIEW OF PRESSURE TRANSIENT METHODS.....................36
III - COMPUTATION OF PRESSURE DERIVATIVES...................................38

Page 3
INTRODUCTION

Thank you for purchasing PT 4 Pressure Transient Analysis software. We

believe that you will find PT to be a full featured software package which will
greatly assist you in evaluating pressure transient well tests.

Features

Standard Win32 user interface

Choice of English or Spanish for menus and data labels
Data files in standard ASCII format
Backwards compatible with PT 2.80 and PT 3.x
Unlimited data points and 10 rate changes per test
Unlimited number of tests and analyses can be viewed simultaneously
Graphical data editing and preparation
Classical analysis methods
Cartesian (Storage, Reservoir limits)
Semilog (Horner, Miller-Dyes-Hutchinson)
Square root of time (Hydraulic fractures)
Log-log type curve matching
Adjust and Compare
Custom designed solutions
Radial and fractured wells
Radial flow, dual porosity, dual permeability
Wellbore storage, phase redistribution, momentum
Automated parameter searching
Graphical output on standard WindowsTM compatible printers
Network and corporate licenses available

The Personal Version of PT 4 is licensed for use on a single-user computer

system. Although it is not copy protected, software support, updates, and
documentation will not be provided to non-registered users. In addition, by using
a ‘pirated’ copy of this software, you are violating federal law and international
treaties.

Since an understanding of pressure transient analysis methods is integral to the

use of this software, this manual is divided into three major sections. The first
section provides a brief description of the program from a user’s viewpoint. It is
meant to be a quick introduction to the use of PT. The majority of this information
is also available on-line in the help files.

The second section shows how PT can be used to evaluate pressure transient
data. This section consists of an example test which is included in the
distribution package. A step-by-step procedure is given to provide an analysis of
the data from start to end.

Following the main text are several appendices. These contain more detailed
information on the equations and nomenclature used in PT and in well testing in

Page 4
general. This section is provided as backup documentation and as a quick
reference to the underlying mathematics.

In the following text it will be necessary to refer to key strokes and mouse
selections, as well as information displayed by the program. In most cases the
context will make it obvious what is being referred to, however, to avoid
confusion this manual uses the following type-face conventions:

Key strokes
Key strokes are bold and enclosed in brackets, for example [Y] means press the
key labeled ‘Y’, [Enter] means press the ‘Enter’ key, etc. When more than one
key must be pressed, this is indicated by ‘+’. Thus, [Alt]+[X] means press the
‘Alt’ and ‘X’ keys at the same time, while [Y] [Enter] means first press the ‘Y’ and
then the ‘Enter’ keys.

Menu Selection
Menu selections will be in bold, with sub-menu selections separated by a bar
character, |. For example, File | Open means select the File menu item, then
select the Open menu item from the drop-down menu.

Typed information
Information to be typed into the program will be indicated by bold type. For
example, typing the filename ‘test1.dat’ would be indicated by test1.dat.

Displayed information
Information displayed by the program will be indicated normal text. In either
case, the context will make clear what is being referred to. In some cases,
screen images will be used to show what is displayed.

INSTALLATION

PT 4 includes an automated installation utility which makes it easy to install.

Simply copy the distribution files to a subdirectory on your system and run the
SETUP program by double-clicking the Setup.exe file. Alternatively click the
Start button, select Run, browse to the location of the files and select Setup.exe,
and click OK.

The setup program will prompt you for some information and copy all needed
files to your hard disk. In nearly every case, the default directories should
suffice.

After the installation is complete, you will be able to run PT from the Start |
Programs Menu from Windows.

If you encounter any problems installing PT, please contact The ComPort
Computing Company for assistance.

Page 5
USING PT
Overview
The PT 4 system consists of a main window containing multiple tests and
analyses. To analyze a well test, a data set must be loaded, either from an
existing file or by entering data in the test data window. Once the test data has
been defined, any number of Classical, Log-log Type Curve, and/or Adjust and
Compare Analyses may be performed on the data.
It is important to understand how a test is modeled in PT. In the overall software
model, a well test consists of a single data set along with any analysis methods
that the user may choose. When an analysis method is selected, a new window
is opened and that particular analysis window will use the data for the test that is
active when the window is created. For this reason, an analysis method may
only be selected when a test data window is visible and active. In addition,
closing the test data window will also close any active analyses that apply to that
particular test.

**IMPORTANT**
When test data is edited, it is important to understand that the data being edited
is a local copy in memory, not the original test data. For changes in the data to
be dynamically recognized by PT, the test data must be updated using the
Update button. After updating, the analysis windows will reflect the changes the
next time that their windows are redrawn. The Update button only updates the
data in memory. To save the modified data set to disk, the File | Save menu
item must be used.

By opening several tests and analysis methods, it is easy to compare results and
data from different wells, different tests on the same well, or view the results of
various analyses on the same test. Each analysis window also contains special
controls and information which applies to the particular analysis technique.
There is no software limitation on the number of tests and analysis methods that
can be viewed at once. The limitation depends solely on the memory, display
area and disk capacity of the computer running PT.
In addition to the information which is displayed on each data or analysis window,
additional information and hints can be viewed at any time by pressing the [F1]
key to activate the context sensitive help system. Much of the information in this
section of the manual is available in the online help.

Menu Commands
Keyboard Shortcuts: To access the menu bars via the keyboard, press [Alt] +
the underlined letter of the menu title. To access menu items once the dropdown
menu appears, simply press the key which corresponds to the underlined letter
of the item.

Page 6
Remember, if an item is dimmed, or grayed, it is not applicable or available for
use at that time.
File Menu
New Create a new test data set. A new, blank test data window is
shown.
Open Open an existing data set. The data window for the test is shown.
Save Save the current data set to the same file.
Save As Save the current data set to a new file. You will be prompted for the
filename and location.
Close Close the current test data set.
Print Setup Sets the printer and configuration to be used for printing
Exit End PT

Analysis Menu
Classical | Cartesian
Selects the Cartesian analysis method and opens the analy-sis
window
Classical | Semilog
Selects the Semilog analysis method and opens the analysis
window
Classical | Sqrt Time
Selects the square root of time analysis method and opens the
analysis window
Log-Log Selects the log-log type curve analysis method and opens the
analysis window
Adjust & Compare
Selects the adjust and compare analysis method and opens the
analysis window
Echometer Selects the EchometerTM analysis method and opens the analysis
window (only applies to Echometer data sets)

Window
Tile Arrange the open windows by tiling them within the PT screen.

Page 7
Cascade Arrange the open windows by cascading them within the PT
screen.

Language
English Select English for all menus and data labels
Español Select Spanish for all menus and data labels.

Help
Contents Displays the Help system.
About Displays information about PT, including the version number.

Test Data Window

Title Enter a title for the test. The title can consist of any displayable
characters and can be up to 80 characters long. This is used for
display purposes only.
Update This button causes the test data to be updated. If changes are
made, the results are not saved for use by the various analysis
methods until the data is updated.

General Data Tab

Well Type Select Oil Well or Gas Well.
This determines which units are used for producing rates.
Echometer Indicate that the test data includes surface pressure and fluid level.
Gradient Enter the fluid gradient to be used for fluids in the anullus
(Echometer only).
Well Radius Enter the value to be used for the wellbore radius.
Porosity Enter the value to be used for the reservoir porosity as a fraction.
Thickness Enter the value to be used for the total formation thickness.
Units System
Select US Oil Field or SI Standard unit system.
User Defined Units
Enter names and conversion constants for desired units. See
Section titled “User Defined Units” for more details

Page 8
PVT Data Tab
Fluid System
This determines which definition of pseudo-pressure will be used in
any test analyses. Select from Normal Oil, Compressible Oil, Ideal
Gas, or Real Gas. If Normal or Compressible Oil is selected,
pressure will be used in the analyses. If Ideal Gas is selected,
pressure squared will be used, while real gas pseudo-pressure will
be used if Real Gas is selected. Note that the data requirements
vary depending upon which fluid system is selected.
Viscosity Enter the fluid viscosity to be used in the analyses. Note that for
Real Gas, viscosity data is entered in the gas PVT table instead.
Total Compressibility
Enter the total compressibility (fluid plus rock) to be used in the
analyses.
Fluid Compressibility
Enter the fluid compressibility to be used for analyses of tests. This
applies only to Compressible Oil fluid systems.
Bo Enter the formation volume factor for the fluid. This applies to
Normal and Compressible Oil systems only.
Std Temp Enter the standard temperature to be used in gas volume
calculations. This applies only to Ideal and Real Gas fluid systems.
Std Press Enter the standard pressure to be used in gas volume calculations.
This applies only to Ideal and Real Gas fluid systems.
Res Temp Enter the reservoir temperature to be used in gas volume
calculations. This applies only to Ideal and Real Gas fluid systems.
Avg Z Enter the real gas z-factor to be used in gas volume calculations.
This applies only to the Ideal and Real Gas fluid systems.
Gas PVT Table
For Real Gas fluid system, the calculation of the real gas pseudo-
pressure requires gas z-factor and viscosity data as a function of
pressure. The data to be used is entered in the table. The first
column should contain the pressure, while the second and third
columns should contain the corresponding viscosity and z-factor.
PT will use this data to compute the pseudo-pressure values used
in the analyses.

Rates/Pressures Data Tab

This tab contains 2 tables where the rate and pressure data to be analyzed are
entered.
Rate Table Rate data is entered up to the starting time of the test. Up to 10
rate variations can be represented. The first column contains the
time (actual, not time difference) that the rate period ends, while the

Page 9
 second column contains the rate in effect up to that time. The last
rate period always corresponds to the time period containing the
measured pressure data; the ending time of the last rate period
should be entered as END, indicating that the rate lasts until the
end of the test.

For a buildup or other test where the well is shut-in, the last rate
should be 0. For a simple drawdown test, there would be only a
single entry, while for a simple buildup test there should be 2
entries – the second being 0.
Pressure Table
The actual pressure data to be analyzed is entered into the
pressure table. The first column should contain the time in hours
since the last rate change, while the second column should contain
the corresponding pressure at that time.

The PT Toolbars
On the each of the analysis windows, a toolbar is visible to provide a quick way
to manipulate the data and perform the various other functions.

Curve Fit

Fits a curve or line to the selected data points. For the Classical
Analysis methods, a linear regression is performed to determine the
appropriate line. For the Adjust and Compare Analysis, a nonlinear
regression is performed.

Undo (Adjust and Compare Analysis only)

Revert to last model.

Show Derivatives

Toggles the display of derivative points and curves on the plot. If

derivatives are displayed, they will be hidden and vice versa.

Print

Prints the current plot and results to the default printer.

Curves (Log-log Analysis only)
Allows the selection of a type curve set for matching to the data.
When [None] is selected, the data can be only viewed.

Page 10
Select Method (Semilog and Cartesian Analysis only)
Allow the selection of the appropriate analysis method for the
current data set. For semilog methods, either Horner or MDH
methods may be selected, while for cartesian analysis, either
wellbore storage or reservoir limits analysis may be selected.

Arrow Buttons (Log-log Analysis only)

The arrow buttons allow the data to be moved relative to the type
curves to perform the actual type curve matching. Each time an
arrow button is pressed, the data is shifted in the appropriate
direction. The amount that the data is moved is controlled by the
adjustment factor (see below)
Adjustment Factor (Log-log Analysis only)
The adjustment factor determines how far the data moves when the
arrow buttons are pressed. A larger factor moves the data points
further, while a smaller factor moves the data by a smaller amount.

Cartesian Analysis
In cartesian analysis, data is plotted as pressure vs. time on a cartesian graph.
The data should form a straight line in two situations:
1) During wellbore storage domination at early times. The slope of the line is
proportional to the wellbore storage constant and the intercept is the flowing
pressure at the start of the test.
2) During reservoir voidage at long times. The slope of the line is proportional to
the reservoir volume being depleted by the well and the intercept depends on the
shape of the reservoir and the well position within it.

Method Choose the analysis method, either Wellbore Storage or Reservoir

Limit
Slope The slope of the displayed line. Pressing the [Fit Line] button will
determine the value by linear regression. Alternatively, a value may
be entered directly.
P 0 hr The intercept of the displayed line. Pressing the [Fit Line] button
will determine the value by linear regression. Alternatively, a value
may be entered directly.
Storage (Wellbore Storage Method only)
The estimated wellbore storage parameter value calculated from
the slope.
CD (Wellbore Storage Method only)
The estimated dimensionless wellbore storage parameter value

Page 11
 calculated from the slope.
Pore Vol (Reservoir Limit Method only)
The estimated reservoir volume being drained by the well. This is
calculated from the slope.
Area (Reservoir Limit Method only)
The estimated reservoir area being drained by the well. This is
calculated from the estimated reservoir volume and thickness.

Square Root of Time Analysis

In square root of time analysis, data is plotted as pressure vs. the square root of
time. The data should form a straight line when flow is dominated by linear
transient flow in the reservoir. This is usually encountered in hydraulically
fractured wells at early times.
Method Choose the analysis method, either Sqrt Time (for drawdown tests)
or Superposition Sqrt Time (for buildup tests).
Slope The slope of the displayed line. Pressing the [Fit Line] button will
determine the value by linear regression. Alternatively, a value may
be entered directly.
P 0 hr The intercept of the displayed line. Pressing the [Fit Line] button
will determine the value by linear regression. Alternatively, a value
may be entered directly.

Frac Length The fracture length estimated from the slope of the line.

Semilog Analysis
In semilog analysis, data is plotted as pressure vs. the logarithm of time. The
data should form a straight line when flow is dominated by radial transient flow in
the reservoir. There are several methods of analysis which differ in how the time
is adjusted to account for varying flow rates before the time of the test.
Method Choose the analysis method, either MDH or Horner analysis.
Slope The slope of the displayed line. Pressing the [Fit Line] button will
determine the value by linear regression. Alternatively, a value may
be entered directly.
P 1 hr The intercept of the displayed line. Pressing the [Fit Line] button
will determine the value by linear regression. Alternatively, a value
may be entered directly.
Prd Time (Horner method)
The value of producing time used in calculating the Horner time
values.
kh The estimated permeability-thickness calculated from the slope.

Page 12
k The estimated permeability calculated from the kh estimate and the
reservoir thickness.
Skin The wellbore damage estimated from the calculated permeability
and the intercept of the displayed line.
P* (Horner method)
The extrapolated static pressure estimated from the displayed line.

Log-log Analysis
In log-log analysis, data is plotted as the logarithm of the pressure change vs. the
logarithm of time. The plotted data is compared to theoretical curves and shifted
to obtain the best visual match. From the position of the data on the theoretical
curves, the reservoir and well parameters can then be inferred.
Curves Allows the selection of a type curve set for matching to the data.
When [None] is selected, the data can be only viewed.
Adjustment Factor
The adjustment factor determines how far the data moves when the
arrow buttons are pressed. A larger factor moves the data points
farther, while a smaller factor moves the data by a smaller amount.
tD/CD 1 hr The time matching parameter, which is the dimensionless
time/storage corresponding to 1 hour of real time.
pD 1 The pressure matching parameter, which is the dimensionless
pressure corresponding to 1 real pressure unit.
kh The permeability-thickness estimated from pD 1.
k The permeability estimated from the calculated permeability-
thickness and the reservoir thickness value.

Adjust and Compare Analysis

In adjust and compare analysis, data and theoretical curves are plotted on the
same graph as the logarithm of the pressure change vs. the logarithm of time.
The user then adjusts the parameters of the theoretical curves until the best
match to the data is obtained. From the values of the parameters, reservoir and
well properties can be inferred.

Well Tab
Model Select the type of well: Radial, Uniform Flux Fracture, Infinite
Conductivity Fracture or Finite Conductivity Fracture.
wfD kfD Dimensionless fracture conductivty (Finite Conductivty Fracture
only)
xf/rw Dimensionless fracture length (Fracture models only)

Page 13
Reservoir Tab
Reservoir
Select Radial, Dual Porosity, or Dual Permeability. This will
determine which reservoir equation is used in calculating the
theoretical curves.
tD/CD(1) This is the dimensionless time/storage corresponding to 1 hour real
time.
pD(1) This is the dimensionless pressure corresponding to 1 real
pressure unit.
Omega (Dual Porosity, Dual Permeability)
This is the porosity-thickness ratio of the major porous media in the
dual porosity model. A value of 0 indicates that all of the reservoir
volume is in the major interval, while a value of 1 indicates that all
of the reservoir volume is in the secondary interval.
Lambda (Dual Porosity, Dual Permeability)
This is the interporosity flow coefficient for the dual porosity or dual
permeability models. A large value indicates that fluids flow easily
between the two porous media, while a small value indicates that
flow between the media is restricted.
Kappa (Dual Permeability)
This is the permeability-thickness ratio of the major porous media in
the dual permeability model. A value of 0 indicates that all of the
reservoir flow capacity is in the major interval, while a value of 1
indicates that all of the reservoir flow capacity is in the secondary
interval.

Storage Tab
CaDe2S This is the dimensionless wellbore storage and skin damage
parameter. Large values indicate a large amount of wellbore
storage or damage, while small values indicate a small amount of
storage or an improved well condition.
Momentum Check to include momentum effects
CmD This is the dimensionless wellbore momentum parameter. When
the value is on the order of the value of CDe2S or larger, the effects
of momentum are significant. For smaller values, momentum
effects are masked by wellbore storage.
dPw0 This is the adjustment needed for the initial pressure.
dt0 This is the adjustment needed in the initial time.

Page 14
Phase Tab
Model Choose the type of phase redistribution: None, Fair or Hegeman.
CD/CaD This is the ratio of actual wellbore storage to apparent wellbore
storage. When there is no phase redistribution, the ratio is 1. For
typical phase redistribution situations, the ratio is greater than 1,
however, for thermal effects, the ratio may be either greater than or
less than 1.
CpD This is the maximum dimensionless phase redistribution pressure
rise. Values between 1 and 20 are typical, with larger values
causing the so-called "gas hump" in the pressure response.

Bndry Tab
Boundary Type
Choose either Inifinte (no boundary) or Single Fault.
Boundary Condition
Choose either Closed or Constant Pressure boundary
reD Dimensionless distance to boundary

Results
kh The permeability-thickness estimated from pD 1.
k The permeability estimated from the calculated permeability-
thickness and the reservoir thickness value.
Skin The wellbore damage estimated from the calculated permeability
and the time and wellbore parameters.
CD This is the estimated actual dimensionless wellbore storage. The
value is useful when comparing the results from cartesian wellbore
storage analysis.

Page 15
FLUID SYSTEMS AND PSEUDOPRESSURE

PT is internally formulated in terms of a generalized pseudopressure, so that

most types of reservoir fluids can be properly handled. In this version of PT, the
fluid systems supported are ‘Normal Oil’, ‘Compressible Oil’, ‘Ideal Gas’, and
‘Real Gas’. The only difference between the models is the definition of
pseudopressure which is used internally.
As shown by P. S. Fair, “Investigation of the Diffusivity Equation with General
Pressure-Dependent Rock and Fluid Properties” (submitted to SPE Dec. 1991),
the diffusivity equation can be written in terms of a general pseudopressure
function which applies not only to gases, but to any type of fluid and rock system
where the physical properties vary with respect to pressure. The use of this
general pseudopressure function greatly simplifies the formulation of pressure
transient analysis methods.
In PT, the form of the general pseudopressure function is
p

m p=∫ dp
p0 
2  kh
m D=c m−mi 
q
where c is an arbitrary constant.
Using this general definition, the typical types of fluids normally encountered in
practice can be handled by simply defining the appropriate relationships for
density and viscosity for each fluid type and then evaluating the resulting integral.
Normal Oil
In the normal oil fluid model, the standard assumptions of constant fluid viscosity
and small and constant fluid compressibility are used. In this case the integral
simplifies to k 0  p− p0 / . Note that the density ratio is the definition of
Bo and defining the arbitrary constant c = 1 gives the standard definition of
dimensionless pressure. In this case, the pseudopressure and pressure are
identical.

m p= p− pi
2  kh
m D=  p− pi 
q B
Compressible Oil
In the compressible oil fluid model, we remove the assumption of small
compressibility, but still assume that both viscosity and compressibility are
c f  p− p0
constant. The fluid density is given by the relation =0 e , where

Page 16
cf is the fluid compressibility. In this case the integral can be evaluated
analytically and the pseudopressure becomes

1
m p= [ 1−e c  p− p ]
f 0

cf
2 kh
mD = m−mi 
q B

Real Gas
In the real gas model, fluid density is expressed in terms of the real gas equation
of state and the viscosity is assumed to vary arbitrarily with pressure. Both the
real gas z-factor and viscosity must be specified versus pressure. To maintain
consistency with the standard literature definition of real-gas pseudopressure, a
factor of 2 is introduced, so the arbitrary constant c = 0.5. This yields the
following expression for pseudopressure:
p
p
m p=2∫ dp
p0 z
1 2 khT 0
m D= m−mi 
2 q p0 T

Ideal Gas

In the ideal gas fluid model, the z-factor and viscosity in the real gas model are
assumed to be constant. In this special case the integral can be evaluated
directly, yielding a pseudopressure function proportional to p2. This gives the
standard p2 formulation which is commonly used for gas wells at low pressure. In
this case the pseudopressure definition becomes
2 2
m p = p − p i 
1 2  khT 0
m D= m−mi 
2 q  z p0 T

Page 17
MULTIPLE RATES

Since many tests have varying rates, PT is setup to automatically handle up to

10 rate variations before the start of a well test. This does not mean that data
from all of the time periods can be analyzed on one plot. PT allows the analysis
of data from only a single constant rate flow period at a time.
However, in order to properly evaluate pressure data it is important that the
significant rate history be used in the analysis, since rate variations prior to the
start of a test can cause the test data to deviate significantly from the ideal case
where the rate is maintained constant.
For type curve analysis, the use of equivalent time is the only choice, since the
type curves themselves cannot be easily modified. For classical analysis
methods, the method itself defines how the rates are accounted for, for example,
the Horner method assumes a constant flow rate prior to a shut-in period, where
the rate is known to be zero. In the Adjust and Compare analysis, the rate
variations are automatically handled by the use of superposition, so the resulting
theoretical curves should be exact.
It should be noted that in PT, all tests are considered to have multiple rates.
Drawdown tests are handled internally as multiple rate test with only a single flow
period, buildup tests are handled as multiple rate tests with 2 flow periods (the
second rate is zero), etc.
In general, the data needed to properly specify the rate variation before the start
of the test is a sequence of times and rates entered into the Rate Table on Rate
and Pressure Tab in the Test Data window. Note that the time entered here
should be the total elapsed time, not the duration of the rate period. Note that
the final rate period always has an ending time specified as ‘END’, indicating that
the rate lasts until the end of the test.
For example a typical buildup test where the well is produced at a rate of 100
BOPD for 10000 hours prior to shut-in would be entered as 2 rate periods:
10000 100
END 0
In general, multiple rate periods would be specified by a sequence of entries as
follows:
<ending time of rate1> <rate1>
<ending time of rate2> <rate2>
...
END <rateN>
In this version of PT, a total of 10 rate variations can be represented for any test.

Page 18
USER DEFINED UNITS

In order to accomdate the various measurement units systems in use throughtout

the pertroleum industry, PT supports both US and SI standard units systems, as
well as user defined units. The default units system is set on the Data window,
General Tab by selecting the desired radio button. Note that US standard oilfield
units is the default when PT is started.
In the US unit system, volumes are expressed in barrels, pressure in psi, length
in feet, viscosity in centipoise, and permeability in millidarcies. In the SI unit
system, volume is in cubic meters, pressure in kilopascals, length in meters,
viscosity in pascal-seconds, and permeability in square micrometers. In all
systems, compressibility is expressed in reciprocal pressure units and time is in
hours.
Unfortunately, not all well test data use either unit system in its pure form and
units conversion is a chore that is best relegated to computer software. For this
reason, PT incorporates a flexible user defined units system which allows the use
of any particular units for any test. Internally, PT is written in the so-called ‘Darcy’
units, where volume is in cubic centimeters, pressure is in atmospheres, length in
centimeters, viscosity in centipoise, and permeability in Darcies.
To use the user defined units features in PT, the user must enter the appropriate
units information on the General Data Tab of the Test Data window. The
required input consists of the name of the unit and a conversion factor to be used
in the units conversion calculations. The name is used for display purposes only.
The units conversion factor for each unit is the number of units per equivalent
Darcy unit. PT will automatically apply these factors in all calculations. The
default standard units systems are as folows:
US Units
Pressure psi 14.696 psi/atm
Length ft 0.032808 ft/cm
Volume bbl 6.28981e-6 bbl/cm3
Permeability md 1000 md/D
Viscosity cp 1 cp/cp
SI Units
Pressure kPa 1013.25 kPa/atm
Length m 0.01 m/cm
Volume m3 1.0e-6 m3/cm3
Permeability µm 0.986923 µm/D
Viscosity Pa.s 1000.0 Pa.s/cp

Page 19
IMPORTING PRESSURE DATA
In order to more directly use data obtained from service companies in computer
files, PT 3.0 and later versions implement a reasonably flexible data import
mechanism. You can directly import pressure or casing pressure/fluid level data
from Echometer® *.exp files, as well as pressure data from text files in a general
columnar format. Version 3.1 and higher also support enhanced data editing
capabilities which are explained later in this document.
The data import feature of PT is accessed from the Test Data dialog,
Rates/Pressures Tab. When the mouse cursor is over the Pressures table,
pressing the right mouse button will display a pop-up menu with 3 choices:
Import Echometer BHP Data, Import Echometer Data, and Import BHP Data.
The first option will allow importing only the calculated BHP data from an
Echometer® *.exp file, while the second option will import the surface pressure
and fluid level data from an Echometer® *.exp file. The third option allows
importing BHP data from a wide range of service companies, as long as the data
is in a columnar format.
After selecting the type of data to import, a file directory dialog will be displayed.
Select or type the name of the file to import and press the Open button. For the
Echometer® data files, the data will be automatically imported and will appear in
the Pressure table.
For the Import BHP Data option, an Import BHP Data dialog box will appear
showing some information about the file and allowing choices to control the
amount and format of the data to be imported. The lower part f the dialog shows
a File Preview with line numbers for reference. In the top portion, you can
specify how many lines at the top of the file to skip (Skip 1st __ lines), which
columns contain the time and pressure data (Time Column and Pressure
Column), as well as the Time Format used in the file. If you have a large file, it
is also possible to specify Data Thinning parameters to avoid importing
enormous amounts of data. The default values of Max DT = 1 hr and Min DP = 1
psi mean that you want to import only pressure points that change by 1 psi or
more, but at least 1 point every hour.
When you have set the parameters, press the Import button to import the data.
(You can also press Cancel to abort.) If too much data was found in the file, a
message will be displayed suggesting that the data thinning parameters be used.
Simply select the import option again from the pop-up menu and try again. Each
time that you import data, the imported data replaces all of the data currently
loaded for the test.

Page 20
ECHOMETER® DATA ANALYSIS
In many beam pumped wells, it is impractical to directly measure bottom hole
pressures, since the rod string precludes running a wireline pressure gage. For
this reason, bottom hole pressures are often measured indirectly using surface
casing pressure and acoustically measured fluid level data. Additional information
concerning the acquisition of this data can be obtained from:
The Echometer Co.
5001 Ditto Lane
Witchita Falls, Texas 76302
USA.
PT 4.0 now supports an analysis method especially formulated for Echometer
data. Basically, a new well model has been introduced into the standard well
testing equations which allows the casing pressure and fluid level to be predicted
directly from the pressure transient equations, so that the raw data can be
analyzed directly. This reduces error in the analysis and also leads to a more
reliable analysis, since more data is being used. For more technical details of this
analysis method, contact The ComPort Computing Co.
To implement the new method, several changes to PT were made for additional
data requirements and a new item was added to the Analysis menu. On the test
data screen, an Echometer check box has been added, along with a fluid
gradient parameter. To specify a Echometer-type well test, simply check the box
and enter the estimated fluid gradient in psi/ft.
In addition, for an Echometer well test, the data entry table has been changed to
allow the input of both casing pressure and fluid level data versus time as shown
below. If a data file is available from Echometer (a file with an extension of
*.exp), the data can be automatically imported from the file by positioning the
mouse over the Pressure table, pressing the right mouse button and choosing
Import Echometer Data from the pop-up menu. Alternatively, choosing Import
Echometer BHP Data will import only the estimated bottom hole pressure data
from the file.

Once the data has been entered, standard analyses (Classical, Log-log Type
Curves, Adjust and Compare) can be performed as detailed in the users manual.
For these analyses, the bottom hole pressure is estimated from the fluid gradient,
casing pressure and fluid level data in a straight forward manner.
To perform an Echometer analysis, select Echometer from the Analysis menu.
An analysis window will be opened as shown below. Note that the upper curve
and data points in red represent the fluid level in the well, while the lower data
points and curve in blue represent the casing pressure. Two additional
parameters are shown on the Well Tab and a weighting factor appears at the
bottom right side of the screen.
Just as with the Adjust and Compare method, the analysis consists of adjusting
the reservoir and well parameters to obtain the best match to the data. This is
done in the same manner as in the Adjust and Compare Analysis. Please refer to

Page 21
that section of the user’s manual for details.
A brief explanation of the additional parameters is also in order. First, in order to
compute the fluid level and casing pressure from the well testing equations,
some additional parameters on the Well Tab are needed. The first, labeled
CD/CLD is the ratio of the wellbore storage to the liquid wellbore storage and is
usually about 0.9 to 1.0 if there is gas present in the annulus. A second
parameter, labeled Fgi, represents the gas fraction in the initil liquid column in
the annulus. Normally this parameter is small, but in wells producing large
quantities of annulus gas, may be close to 1.0. Finally there is a parameter
labeled DXi that represents the initial fluid column height. In a “pumped off” well,
this would correspond to the distance between the pump and the producing
interval.
Since this analysis method requires fitting 2 simultaneous data sets, casing
pressure in psi and fluid level in feet, a relative weighting factor, labeled Fld Lvl
Weight, is also needed. The value of this factor should reflect the relative
accuracy of the measurements. Theoretically, the value should be computed by
dividing the precision of the casing pressure measurements by the sum of the
precisions of the casing pressure and fluid level measurements, converted to psi
with the fluid gradient.
In other words, if the fluid level is measured with an accuracy of +/-5 ft and the
fluid gradient is 0.35 psi/ft, the fluid level precision is (5 ft)(0.4 psi/ft) = 2.0 psi. If
the casing pressure is measured with a precision of +/-0.1 psi, then the weighting
factor should be (0.1)/(0.1 + 2.0) = 0.0476. Tests indicate that values in the range
of 0.01 to 0.1 are typical and the results are not extremely sensitive to the exact
value in most cases.

Page 22
GAGE DATA EDITING
PT 3.1 and later versions contain an improved graphical means of editing raw
gage data to delete extraneous pressure measurements. It is now possible to
import pressure gage data (as described earlier), view the data on the screen,
select any extraneous data points or thin the data, and automatically delete the
selected data points. These editing features are available from the Analysis |
Classical | Cartesian window by pressing the right mouse button.
When the right mouse button is pressed with the mouse pointer over the plot
region of the Cartesian Analysis window, a pop-up menu should appear which
has 2 choices: Thin Data and Delete Inactive Data.
Selecting Thin Data presents a dialog on which data thinning parameters can be
entered. The Min DP parameter determines the minimum acceptable pressure
difference between points. For example, if Min DP = 10 psi, then adjacent data
points with a pressure difference less than 10 psi will be purged. The Max DT
parameter specifies the maximum time increment between data points. For
example, if Max DT = 2 hours, then at least 1 data point every 2 hours will be
retained, even if the pressure difference is less than that specified by Min DP.
Selecting Ok will mark as inactive the data points to be thinned. Note that Thin
Data does not delete the points, but only marks them as inactive.
Delete Inactive Data removes all inactive data points (colored gray on the
screen). After deleting the points, it is necessary to go to the Data window and
press the Update button for the changes to be registered in memory.

Page 23
PRESSURE TRANSIENT ANALYSIS TUTORIAL

This section of the User’s Guide describes step-by-step how actual pressure
transient data can be analyzed using PT 3.0 or later versions. Included on the
distribution diskette are several pressure transient data files. These should have
been automatically installed along with PT. The test data in these files was
extracted from the literature, as referenced in the test title.
Note that to a large extent pressure transient analysis is subjective. Without
knowing in advance the exact reservoir model and wellbore effects, it may be
possible to obtain several different analyses results for the same data, none of
which can be proven to be correct. Therefore, The ComPort Computing
Company cannot vouch for the results obtained from the analyses. The analyses
shown here are for illustrative purposes only, not to dispute results obtained
elsewhere.
In the following sections you will be guided through the evaluation of a published
set of pressure transient data. The user is encouraged to perform the steps in
these analyses while reading this manual. This will familiarize you with the
mechanics of using the PT software and may also provide some valuable insight
into a workable approach to the evaluation of actual data.
Selecting and Loading Data
To begin the analysis of a typical data set, start PT by selecting it from the Start
Menu (ie. with the mouse, press the Start button, select Programs, select the
PT 4.0 group, then select PT40. This should start the program with a blank
window, since no data has yet been loaded.
To load the data set that we will use in this example, select File | Open from the
menu, then choose TEST1.DAT from the list of files displayed. Either double-
click the file name with the mouse, or select the file and then select OK. You
should see the Test Data window for this well test as shown on the following
page.
On the General Data Tab, note that this is an oil well, the US units system is
used, and assorted data has been entered. The units conversion factors should
all be set to their default values.
The rest of the test data can also be viewed by clicking on the various tabs for
each category. The fluid PVT data is on the tab labeled PVT and the rate and
pressure data is found on the tab labeled Rates/Pressures. Note that a Normal
Oil fluid system has been selected and the necessary data has been entered for
viscosity, total compressibility and formation volume factor. The remaining items
are not used for this type of fluid. In addition, you can see that there is one rate
defined, since this is a drawdown test. The pressure data consists of 25 points
between 0 and 12 hours.
If we needed to make changes to the data, any needed editing could be done on
this window. In this case, however, the data should already be correct, so we
may proceed with our analysis.

Page 24
Preliminaries - Log-log Analysis
Now that the data for the test has been loaded, let’s take a look at it. While we
could jump directly to the Classical or Adjust and Compare analysis techniques, it
is often enlightening to view the data to get an idea of the characteristics of the
data we are dealing with. For this purpose, a log-log data plot is usually
extremely useful.
To view the log-log data plot, select Analysis | Log-log from the menu. The
Log-log analysis window should open displaying the data. Note that you can
resize or reposition the window however you like using the mouse. The display
is shown on the next page for reference.
Note that the blue circles on the data plot represent the pressure data, while the
pressure derivatives are represented by the red squares. If you do not want to
view the derivatives, select the Derivative button at the top of the window. Each
time this button is pressed, the derivative display is toggled.
In this particular test, we can see that the data exhibits a unit slope line at early
times, then flattens out at later times. The derivative data also rises at early
times, then curves over and approaches a nearly constant value. This is typical
of a test dominated by wellbore storage, so we don’t expect anything odd. From
theory we know that the flat part of the data can be evaluated using classical

Page 25
semilog methods.

Before we perform a semilog analysis, however, let’s attempt a type curve match.
This is easily done with PT by simply selecting the type curve set which we want
to use. From the drop-down box labeled Curves at the top of the window (which
currently indicates ‘None’), select the type curves which you want to match
against. In this case, we probably want to attempt a match with the Radial Flow
w/Derivatives type curves, so choose that. The type curves and data will be
drawn on the screen as shown below.
Note that the data appears to have the right shape, but we will need to move the
data to better match the curves. This is accomplished by pressing the arrow
buttons located at the top left of the window. Each time one of these buttons is
pressed, the data is moved in the indicated direction. To change the distance
that the data is moved, change the Adjustment Factor at the top right of the
window. A larger value moves the data a greater distance and vice-versa.
In this case, pressing the left arrow button moves the data to the left and into
fairly close alignment with the type curves. You can move the data however you

Page 26
like to obtain the best match. Note that each time the data is moved, the analysis
results are redisplayed on the screen.
In this case, with tD/CD = 27.7778 and pD(1) = 0.014715, the permeability of the
reservoir is estimated to be 9 md and CDe2S is about 1E+10

Semilog Analysis
Since we now have an idea that the test data appears reasonable and also have
an estimate of the permeability, we can go to the classical analysis methods to
check our work and perhaps refine our estimates.
To try a semilog analysis, we first must go back to the data set, since PT needs
to know what test we want to evaluate. (Remember, you can open any number
of tests at the same time, even though we only have one open at the moment.).
If the Test Data window is partially visible, simply click the mouse anywhere
within the window. Alternatively, select Window | Test Data from the menu.
With the Test Data window active, select Analysis | Classical | Semilog from
the menu.
The Semilog Analysis window should be opened, as shown and we are ready to
start our analysis by fitting a line to the late time pressure data.
Note that the MDH method is selected, since this is a drawdown test. If this were

Page 27
a buildup, Horner would have been selected as the default methods, but we
could chage that to MDH if desired. In this case, Horner does not apply, so we
cannot force the change.
To start out, all of the data points are colored blue, indicating they are active. To
select which points which we want to fit, we will need to deactivate the points we
do not want to use. In this case, the easiest way is to place the mouse cursur
over a point at the top left of the data plot and drag the mouse to the lower right
while holding down the left mouse button. As the mouse is moved, a rectangle
will be drawn. When the button is released, all of the data points inside the
rectangle will be changed to gray, indicating that they are no longer active.
Alternatively, clicking on any point with the mouse will toggle the color to gray,
deactivating the point. Clicking on the point again will toggle its color back to
blue, activating it again.
In this case, deactivate all of the points less than about 1.5 or 2 hours. Once
only the desired points are active, we can fit a line to them by pressing the
‘Curve Fit’ button at the top left of the window. This will fit a line to the selected
data and update the display. Using the points shown, the display should appear
as follows.

Page 28
Note that the slope of the line and the value of pressure at 1 hour are shown and
the permeability and skin have also been automatically calculated based on the
position of the line. If you want to manually adjust the line, the slope and
intercept can be entered. You can also change which points are active and fit
another line to the data.
In this case, however, it appears that the line is pretty reasonable, since the
standard deviation (at the bottom of the window) is less than 0.4 psi using 10
data points, so we will leave it alone.
Note that the permeability is estimated to be 9.05 md and the skin is estimated to
be about 5.7. This is encouraging, since the permeability agrees fairly well with
the estimate from the log-log analysis earlier.
Note also the buttons that appear on the bottom part of the right panel in the
window. These can be used to change the scale and display to view the data.
Pressing the Position buttons slides the scale right and left or up and down,
while pressing the Divisions buttons increases or decreases the number of
divisions shown. By using these, the graph can be adjusted to suit your tastes,
zooming in or out on parts of the data as desired. The position of the plot on the
display has no effect on the analysis results.
Wellbore Storage - Cartesian Analysis
In this well test, we noted from the log-log plot that the early time was dominated
by wellbore storage and that the early time data appears to form a unit slope line.
This means that a cartesian graph of the early time data should be linear and that
we can use Cartesian Analysis to determine the wellbore storage constant.
To perform a Cartesian Analysis, activate the Test Data window and select
Analysis | Classical | Cartesian from the menu. The Cartesian Analysis
window will be displayed.
Note that the Method selected at the top of the window is Wellbore Storage. If
we were looking at late time data, the Reservoir Limit method could be selected
instead. In most cases, however, we will want wellbore storage, as in this case.
Since we want to look at the early time data, it is necessary to zoom the plot to
show all of the early points. First, we will deactivate the late time points (greater
than about 1 hour) by dragging the mouse across them. Then select the
appropriate buttons to decrease the number of X-Axis Divisions to expand the
graph horizontally. As the graph expands, it appears that more of the points are
not linear, so we deselect those, also, until the only remaining points are at times
less than about 0.1 hours. Pressing the ‘Curve Fit’ button at the upper left fits a
line to the remaining points as shown.
Note that the standard deviation is about 0.64 psi, which appears acceptable.
The wellbore storage is also estimated as 0.00887 bbl/psi (which could be
checked against well tubular data, if we had any) and the dimensionless storage
is estimated to be about 8306.
Since we estimated the skin to be about 5.8 from the semilog analysis, we can
also check the CDe2S parameter with our match on the log-log plot. Using 8306

Page 29
for CD and 5.7 for S, the parameter is calculated to be 7.4x109. Since the log-log
match shows the data just below the curve labeled 1x1010, the log-log analysis
still appears to be valid and consistent.

Improving Things - Adjust and Compare

Now that we have performed a log-log type curve match and confirmed our
estimates using both semilog and cartesian analysis methods, we have a pretty
good idea that our analysis is valid and the parameters that we have estimated
are fairly close. At this point we might be satisfied, but if we really want to check
our analysis, we can use the Adjust and Compare Analysis method to check all
the data at once and fine tune our estimates. In addition, we can also perform
analyses on data which do not have a cartesian or semilog line and data for
which no type curve is available.
To start the Adjust and Compare Analysis, activate the Test Data window, then
select Analysis | Adjust/Compare from the menu. The Adjust and Compare
Analysis window will open displaying the data along with a curve showing a
theoreical solution to the well testing equations. In this analysis method, we will

Page 30
adjust the dimensionless parameters (with the help of PT) to match the entire
range of data as closely as possible.
To start out, we already know that the test exhibits wellbore storage and there is
no evidence of fractures, so on the Well Tab we will leve the default model as
Radial. We do have estimates for the parameters from our type curve match, so
we can use those as initial estimates in our evaluation by selecting the Res Tab
and typing in the values for pD = 27.778 and tD/CD = 0.01472 in the appropriate
boxes. Be sure to press [Enter] after typing each of the values and you should
see the curve redrawn with the updated parameters. Since we have also
estimated CDe2S, we can type its estimated value of 7.4E+9 after selecting the
Storage Tab on the upper right of the analysis window. The theoretical curve will
be redrawn to reflect the changes each time.
The display should look similar to that shown below.

Although the match doesn’t look too bad, the standard deviation is about 59 psi
(at the bottom of the screen), which seems much too large. We will let PT
improve the match for us by following these steps.
First, adjust the CDe2S parameter, simply by checking the box next to that
parameter. Then switch back to the Res Tab and check the boxes next to pD and
tD/CD. Each time, PT will automatically adjust the parameter to try to find the
best match to the data. Note that the standard deviation is reduced to about 8.7

Page 31
psi, which is much better. In addition, the permeability is estimated at about 9 md
and skin at 5.7, which is pretty close to our previous estimates.
But we can do even better. So far we have adjusted the parameters
independently. By letting PT perform a nonlinear search, all of the parameters
can be adjusted simultaneously, which is a much better numerical technique.
Unfortunately, the nature of the data and equations requires that we have a pretty
good estimate of the parameters before we turn PT loose. By adjusting the
parameters independently, we’ve come pretty close to a match, so PT should be
able to do its work from here.
To let PT find a better solution, simply press the ‘Curve Fit’ button at the upper
left of the window. PT will search through about 15 iterations of a nonlinear
search, updating the curve as it proceeds. When it’s finished, an improved
match is obtained, with a standard deviation of about 7.5 psi. Note that the
estimates of permeability and skin have changed to about 9.2 md and 6.1
respectively. We can ask PT to continue searching for an improved match by
pressing the 'Curve Fit' button repeatedly, or by allowing more iterations by
setting Max Iterations to a larger number, like 100. Go ahead and try it.
But, since PT handles much more complex analysis models, we can do even
better than this. Note that the derivative data appears to drop too rapidly and
flatten out too soon in comparison to the theoretical curve. This might be due to
wellbore effects, which we can model.
To model the effect noticed here, we will use a small amount of phase
redistribution. Note that since this is a drawdown test, we don’t expect phase
redistribution in its normal sense, but instead the efffect is probably due to
temperature changes in the wellbore. Discussion of this is beyond the scope of
this manual, but the interested reader is referred to the paper referenced in the
Appendix. (We could also add momentum effects, but that isn’t needed here.)
To include phase redistribution effects in the theoretical curve, simply go to the
Phase Tab and change the Model to Fair. Note that some additional parameters
appear to describe the phase redistribution effect and that the curve is redrawn to
effect the changes.
Since the amount of phase redistribution is small in this case, first lower the value
of CD/CaD from 10 to 1.2, then change the value of CpD from 10 to about 4. Note
that the curve shape adjusts each time and now has a much better shape.
However adjusting these parameters changed the match, so we go back to the
Reservoir Tab and readjust the tD/CD value by unchecking and rechecking the
box next to it. At this point the standard deviation has been reduced to about 4.0
psi, which is significant, and the curve has the right overall shape to match the
data.
To complete our analysis, we can let PT find a better match for us. First activate
the new parameters by going to the Phase Tab and checking the boxes next to
CD/CaD and CpD. Then press the ‘Curve Fit’ button again. PT will again adjust all
of the parameters for the best match it can find.
Note that the standard deviation has been reduced to about 2.6 psi and that the

Page 32
estimated permeability and skin are now 9.87 md and 6.89 respectively. These
estimates are fairly close to our prior estimates, so we can be pretty sure that our
analysis is valid. In addition, we have explained the slight deviation in the data
by introducing some wellbore effects and also shown that these effects do not
substantially change the evaluation results.
As stated before, the purpose of this analysis has been to illustrate the use of PT
for evaluating well test data. You can further adjust the parameters or try
changing the reservoir model and let PT continue improving the match until you
are satisfied that the best match has been obtained.
With about 1000 iterations, you should be able to attain a standard deviatin of
about 2.57 psi for this test and the permeability and skin would be around 9.45
md and 6.29 respectively.

Page 33
APPENDIX I - PT 3.1 NOMENCLATURE

The basic relations for pressure transient analysis consist of the diffusivity
equation, to describe flow within the reservoir, along with suitable boundary
conditions, to describe flow at the outer edge of the reservoir as well as into the
wellbore. In most cases, the reservoir is assumed to be infinite, so the outer
boundary condition is especially simple.
To simplify and standardize the diffusivity equation and boundary conditions,
dimensionless parameters are defined. The definition of these parameters,
expressed in Darcy units are as follows:

r kt
Radius r D= Time t D= 2
rw  c t r w
kh  p qw
Pressure pD= Rate q wD =
qB  qo B
C
Storage C D= 2
h ct r w

As documented in the paper “Generalized Wellbore Effects in Pressure Transient

Testing” (Fair, SPE Formation Evaluation, June 1996), the general equations for
describing flow into and pressure transient behavior within the wellbore during
single phase well testing are as follows:

Skin p wD = p D r D =1S qwD

Reservoir Inflow q wD =
 
 pD
rD r D=1

[  ]
2
dpwD dp D d q wD
Wellbore q wD =1−C D − −C mD
dt D dt D 2
dt D

The solution of these wellbore equations, along with the appropriate form of the
diffusivity equation, yields the general relationships used in pressure transient
analysis.
In addition, the common forms for the ‘phase redistribution pressure’ function
used in PT are defined as:

p  D =C  D  1−e 
−t D /  D
Fair

Page 34
Hegeman p  D =C  D erf t D / D 

Using either of these definitions, there is a linear early time behavior (in the
absence of momentum effects) which has the appearance of wellbore storage.
This apparent storage is defined as:

[ ]
−1
1 C D
Apparent Storage C aD= 
C D D

Summary of variables used:

B Formation volume factor
C Wellbore storage parameter
Cm Wellbore momentum parameter
Cφ Wellbore phase redistribution parameter
ct Total fluid compressibility
h Reservoir thickness
k Permeability
p Pressure
q Flow rate
r Radius
S Skin factor
t Time
x Length
α Phase redistribution time parameter
φ Porosity
µ Viscosity
Subscripts
a apparent
D dimensionless
f fracture
i initial
o surface
t total
w well
φ phase redistribution

Page 35
APPENDIX II - AN OVERVIEW OF PRESSURE TRANSIENT METHODS
The mathematical basis of pressure transient analysis lies in the partial
differential equations of single phase fluid flow. Since we are concerned mainly
with the pressure response at a producing well during time periods when
wellbore effects may be important, the radial form of the diffusivity equation is
usually used to model the reservoir flow behavior. Since wellbore effects are
important, mass and momentum balance relations for fluids in the wellbore
describe the wellbore boundary conditions. Since the reservoir is normally
extremely large in comparison with the size of the wellbore, an infinite reservoir is
often assumed.
In order to keep from solving the equations for every combination of parameters
that might be encountered, the equations are generally expressed in terms of
dimensionless quantities. Except for a few unusual circumstances, the same set
of dimensionless parameters are used throughout well test analysis and are
described in Appendix I.
Once the diffusivity equation and the associated boundary conditions are
specified in terms of dimensionless variables and a solution is obtained, the
process of pressure transient analysis becomes one of determining what values
of the dimensionless variables and parameters are needed to achieve a match
between the theoretical solutions and the measured data. This can be done in
several ways, leading to the three major methods used in pressure transient
analysis and implemented in the PT software.
Classical analysis methods rely on the observation that portions of the theoretical
pressure solutions form straight lines when plotted versus some type of time
related variable. The Cartesian plot is useful at early times, when wellbore
storage dominates and inflow from the reservoir is essentially filling the wellbore.
It is also useful at very late times, when production is draining the reservoir and
the average reservoir pressure is dropping with time. The square root of time
plot is useful when transient linear flow dominates, usually due to flow into a
vertical fracture intersecting the wellbore. The semilog methods are useful
during transient radial flow after wellbore effects have died out. In any of the
methods, fitting a straight line to the data is one way to determine the
parameters required to make the solution match the data. The slope and
intercept of the fitted line are used to compute the physical parameters.
Type curve analysis relies on the observation that the dimensionless time and
pressure parameters consist of a constant term multiplied by real time or
pressure and the fact that multiplying numbers is equivalent to adding logarithms.
When both the theoretical solutions and data are plotted on a log-log graph,
sliding the data vertically and horizontally is equivalent to adding or subtracting
logarithms and thereby adjusting the constant term of the dimensionless
parameters. When a suitable match is obtained visually, the parameters can be
determined from the ratio of the real and theoretical parameters at any point on
the graph. This is equivalent to fitting a line, except that curves are used and no
assumption of linearity in the solution is required.
The adjust and compare method, as implemented in PT, is more direct in

Page 36
determining the parameters of the solution. Instead of changing the problem into
one of finding a straight line or matching some set of theoretical curves, the data
and theoretical solutions are plotted on the same graph and the parameters of
the theoretical solution are adjusted until the best match is obtained. Using this
method, we are able to analyze data which exhibit no straight line by any
method, as well as in situations where no set of type curves has been prepared.
The values of the parameters of the theoretical solution directly determine the
desired analysis results.

Page 37
APPENDIX III - COMPUTATION OF PRESSURE DERIVATIVES
Although the technique is usually referred to as "pressure derivatives", the actual
quantity used in "pressure derivative" analysis is the derivative of pressure with
respect to the natural logarithm of time. This corresponds to the slope on a semi-
log plot, which always has more visual curve character than that of a log-log plot,
since the process of taking the logarithm of pressure differences also removes
most character from the data. Before development of the pressure derivative
techniques, semi-log methods were deemed more reliable for transient analyses
due to this loss of characteristic curvature with log-log methods.
Unfortunately, taking the derivative of measured pressure data is not trivial. The
simplest approach would be to simply divide each pair of pressure readings by
the difference in log-time, however, except for ideal data, the resulting values
nearly always have a high degree of scatter. The scatter is due to round-off and
measurement errors and makes use of the derivative information impractical.

p i1− p i
A two point method for estimating a derivative is f ' t =
t i1−t i
where p is the value measured at time t. If we expand with a Taylor series and
add a small error, we obtain

1 2 1 3
p i1= pit p ' t   t p ' ' t  t p ' ' ' t⋯
2 6
pi 1− p i 1 1 2 
= p ' t   t p ' ' t  t p ' ' ' t ⋯
t 2 6 t
p i1− p i 1 1 2 
p ' t= −  t p ' ' t−  t p ' ' ' t⋯
t 2 6 x

When the second derivative term is small, the error is dominated by and
t
since t should be small, the error can be very large.

To counteract this problem, a more complex algorithm is always used in

calculating pressure derivatives from actual data. The scatter is reduced by
using a higher order numerical method and by forcing time increments to be of a
minimum magnitude, thereby adding a smoothing effect to the derivative data.
The algorithm used is a second order Taylor's series approximation to the
derivative with non-evenly spaced points. To derive the algorithm, note that a
Taylor's series approximation to subsequent data points can be written as:
2 2
dp  t d p
p i1= pit i1  ⋯ and
dt 2 dt 2

Page 38
 2 2
dp  t d p
p i−1= pi− t i1  −⋯
dt 2 dt 2
If we hypothesize a relation of the form

dp
= Apt t i1  BptCp t− t i −1 
dt
then we can substitute the relations from the Taylor series expansion for p(t+∆ti+1)
and p(t-∆ti-1) to yield an expression for the derivative in terms of the adjacent
function values and the constants A, B, and C. By comparing the expression, we
require that the coefficients for all terms be zero, except the coefficient of the first
derivative term which should be 1, yielding 3 equations in A, B and C. By solving
the three equations for the coefficients simultaneously, a complete expression
can be obtained.
2 2
 t i1  p i− pi−1  t i −1  pi 1 − p i
p '  x=
 t i−1  t i1  t i−1 t i1
In this formulation, the second derivative error term vanishes, so we have a
formula that is more accurate, but it is still susceptible to small errors in the data
when ∆ti-1 and ∆ti+1 are small. We get around that problem by placing a lower
limit on the size of ∆ti-1 and ∆ti+1.

In effect, we use a more accurate difference formula to minimize estimation

errors, but make the differences larger to ensure stability.
In pressure transient analysis, instead of pressure derivatives, we are actually
interested in the semi-log slope, which is the derivative with respect to the
logarithm of time. The practical implication of that is the difference in t is
replaced by the difference in log-time. Even though time differences may be
large, the logarithmic difference may be very small. In practice, we usually
specify that the ratio of two times used in calculating the derivative are at least
10-20% different. This has the effect of requiring that the ∆ti in the differentiation
formula is on the order of 0.1 at the smallest and is usually adequate for most
test data.

dp dp
We also can take advantage of the fact that =t
d [lnt ] dt
Applying this relation, the formula normally used in calculating pressure
derivative data is
2 2
dp t i1−t i   p i− pi−1 t i−t i−1   p i1− p i 
=t i
d [ln t] t i1−t i t i −t i−1t i1−t i−1
where ti-1 < 0.9 t & ti+1 > 1.1 t.

Page 39