Tutorial_Complete.pdf

Petroleum Experts
IPM Tutorials
IPM 5.0
August, 2005
Tutorial Examples

2 - 205 Tutorial Guide
PETROLEUM EXPERTS LTD
The information in this document is subject to change as major improvements
and/or amendments to the program are done. When necessary, Petroleum
Experts will issue the proper documentation.
The software described in this manual is furnished under a licence
agreement. The software may be used or copied only in accordance with the
terms of the agreement. It is against the law to copy the software on any
medium except as specifically allowed in the license agreement. No part of
this documentation may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, recording, or
information storage and retrieval systems for any purpose other than the
purchaser's personal use, unless express written consent has been given by
Petroleum Experts Limited.
All names of companies, wells, persons or products contained in this
documentation are part of a fictitious scenario or scenarios and are used
solely to document the use of a Petroleum Experts product.
Address: Registered Office:
Petroleum Experts Limited Petroleum Experts Limited
Spectrum House Spectrum House
2 Powderhall Road 2 Powderhall Road
Edinburgh, Scotland Edinburgh, Scotland
EH7 4GB EH7 4GB
Tel: (44 131) 474 7030
Fax: (44 131) 474 7031
Email: edinburgh@petex.com
Web: www.petroleumexperts.com

Tutorial Guide 3 - 205
TUTORIAL GUIDE
1 Introduction..............................................................................................................................4
1.1 About This Guide ...............................................................................................................4
1.2 How to use this guide.........................................................................................................4
1.2.1 Symbols and conventions............................................................................................5
2 Dexterity Examples..................................................................................................................6
2.1 GAP Gas Network Example...............................................................................................6
2.1.1 STEP 1 : Initialise GAP................................................................................................6
2.1.2 STEP 2 : Initialise GAP Method Options......................................................................8
2.1.3 STEP 3 : Define GAP Model Schematically...............................................................10
2.1.4 STEP 4 : Define the Well ...........................................................................................11
2.1.5 STEP 5 : Calculate the Well IPR and VLP.................................................................13
2.1.6 STEP 6 : Solve the GAP Network..............................................................................16
2.1.7 STEP 7 : Material Balance Prediction........................................................................18
2.1.8 STEP 8 : Constraints .................................................................................................22
2.2 PROSPER Gas Well Example...........................................................................................29
2.2.1 STEP 1 : Initialise PROSPER......................................................................................29
2.2.2 STEP 2 : Initialise PROSPER Method Options ...........................................................30
2.2.3 STEP 3 : Initialise PVT Data ......................................................................................31
2.2.4 STEP 4 : Initialise Well Inflow and Equipment...........................................................32
2.3 MBAL Gas Reservoir Example ........................................................................................41
2.3.1 STEP 1 : Initialise MBAL ...........................................................................................41
2.3.2 STEP 2 : Initialise MBAL Method Options.................................................................43
2.3.3 STEP 3 : Initialise PVT Data ......................................................................................44
2.3.4 STEP 4 : Initialise Tank Parameters..........................................................................45
2.4 GAP Gas-Lifted System ...................................................................................................47
2.4.1 Introduction ................................................................................................................47
2.4.2 Step 1: Setting up the System ...................................................................................48
2.4.3 Step 2: Drawing the system.......................................................................................52
2.4.4 Step 3: Describing the wells.......................................................................................58
2.4.5 Step 4: Describing the Network .................................................................................62
2.4.6 Step 6: Allocating the Amount of Gas Available ........................................................66
2.4.7 Step 7: Analysing the results .....................................................................................68
2.4.8 Step 8 – Associated Water / Gas Injection Model......................................................71
3 Physics Examples .................................................................................................................86
3.1 Gas Lift Example..............................................................................................................86
3.1.1 Setting Up the Basic Model........................................................................................87
3.1.2 Matching Test Data and Data Quality Check Methods..............................................99
3.1.3 Designing a New Gas Lift Well ................................................................................110
3.1.4 Using QuickLook as a Diagnostic Option to Check the Gas Lift Design..................116
3.2 PROSPER ESP Example ................................................................................................120
3.2.1 Setting up the Basic Model ......................................................................................121
3.2.2 Matching Test Data and Data Quality Check Methods............................................131
3.2.3 Designing an ESP for this Problem..........................................................................136
3.2.4 Using ESP QuickLook as a Diagnostic Option to check an Existing ESP Design...143
3.3 MBAL Gas History Matching Example...........................................................................152
3.3.1 STEP 1 : Initialise MBAL .........................................................................................153
3.3.2 STEP 2 : Add Production History Data ....................................................................153
3.3.3 STEP 3 : Material Balance Introduction...................................................................153
3.3.4 STEP 4 : Material Balance Matching .......................................................................154
3.3.5 STEP 5 : Relative Permeability Matching ................................................................162
3.4 MBAL Oil History Matching Example.............................................................................168
3.4.1 PVT Data..................................................................................................................169
3.4.2 Setting up the Basic Model ......................................................................................169
3.4.3 Matching to Production History data in MBAL.........................................................179
3.4.4 Running Sensitivity Analysis on the Tank Model .....................................................186
3.4.5 Using Simulation Option to Quality check the History Matched Model....................188
3.5 Performing a Production Prediction starting from a history matched model ..................189

1Introduction
This document contains tutorials for the Petroleum Experts’ software:
PROSPER, MBAL and GAP. The tutorials will lead you through a number of
program examples. If you are relatively new to the software, then these will
allow you to use the software immediately and will provide a good overview of
the programs’ functionality.
The tutorials are split into two sections. The first set of tutorials is for
beginners and focuses on the dexterity skills needed to use the programs.
The second set focuses on the physics and engineering issues relating to the
programs.
1.1 About This Guide
The guide assumes you are familiar with basic Windows operations and
terminology.
The screen displays used in this guide are taken from the examples provided
with the software. On occasion, the data files may vary from the examples
shown as updates to the program are issued. Where major amendments or
changes to the program require further explanation, the corresponding
documentation will be provided.
What is in this guide:
• Chapter 2, ‘Dexterity Examples’, lists tutorials that concentrate on
basic use of the programs, but do not emphasise the physics of the
engineering problems concerned.
• Chapter 3, ‘Physics Examples’, lists tutorials that assume the user
understands the basics of the programs, but concentrates on the
physics of the engineering problems.
1.2 How to use this guide
If you have not used Petroleum Experts’ software before you should work through
all the examples in Chapter 2 before moving on to Chapter 3.
If you are comfortable with using the Petroleum Experts software you may still find
the tutorials in Chapter 3 useful.

TUTORIAL GUIDE
1.2.1 Symbols and conventions
Throughout the user guide, special fonts and/or icons are used to
demonstrate specific steps, instructions and procedures in the program.
PETEX program The term PETEX program is used when the comment is
applicable to MBAL or GAP.
ALL CAPS Represent DOS directories, file names, and commands.
Italics Used to highlight certain points of information.
Keycap Bold fonts are used to indicate a specific action to be taken.
For example: ‘Click Done to exit the window.’
Menu ⏐ Command To avoid repeating the phrase ‘Click the File menu and
choose the Open command’, we use the File - Open
convention instead.
∫ Emphasises specific information to be entered or be aware
of.
This keyboard icon marks step-by-step instructions.

This symbol is a reminder to click the RIGHT mouse button.
Clicking the right mouse button performs specific functions
in MBAL, depending on the active dialogue box or plot. If
you do not have a right mouse button, holding down the
SHIFT key while you click the mouse button performs the
required function.

2 Dexterity Examples
This section contains the following tutorials:
GAP Gas Network Example:
This example builds a simple gas network system and runs a production
prediction. It uses PROSPER to model the wells in the system and MBAL to
model the tanks.
PROSPER Gas Well Example:
This example is used within the GAP Gas Network example to show how to
set up the well models required in the gas network system. It can also be run
in isolation.
MBAL Gas Reservoir Example:
This example is used within the GAP Gas Network example to show how to
set up the tanks in the gas network system. It can also be run in isolation.
GAP Gas Lift Example:
This self-contained example builds and optimises an oil production system
using gas lift.
All the example files can be found under:
C:Program FilesPetroleum ExpertsIPM 5.0Worked examplesDexterity
examples
2.1 GAP Gas Network Example
This tutorial example is designed to provide a step-by-step introduction to the
GAP program. The emphasis is on the data entry required to model and
analyse the production potential of a dry gas producing reservoir (no
condensates). The actual data is of little importance: for clarity it has been
chosen to be minimal. However, the systematic method used to generate the
GAP model using PROSPER and MBAL is an important element of the tutorial.
The PROSPER and MBAL phases are separate modules referenced from within
this GAP tutorial.
2.1.1 STEP 1 : Initialise GAP
This section describes how GAP is started and how the location of the
required associated files is initialised.
Start the GAP program by running GAP.EXE, which can be found in the
Petroleum Experts directory (default C:Program FilesPetroleum ExpertsIPM 5.0).

TUTORIAL GUIDE
See the GAP manual for more details on how to start GAP. The version of
GAP being used may be checked by selecting, Help | About GAP.
Figure 1: About GAP – version and build information
The command options (File, Options etc.) at the top of the GAP window are
laid out in a logical order (left to right) that reflects the order in which
operations will usually be performed. Check that File | Directories and check
that they point to the current versions of PROSPER and MBAL respectively. The
PROSPER and MBAL applications can also be found in the Petroleum Experts
directory.
Figure 2: Directories settings
Note that files saved with these versions of GAP, PROSPER and MBAL will not
be readable by previous versions. It is recommended, therefore, that the File |
Directory | GAP and MBAL Data Directory options are set to point to
directories that are exclusively used to store data files created with the current
software versions.

When GAP is started a new file is initialised (unless otherwise specified in the
preferences). If you wish you can ensure initialisation by selecting File | New
or the toolbar accelerator to start a new file.
It is important to ensure that consistent units are used throughout, particularly
when data generated by PROSPER and MBAL are incorporated into a GAP
model. Oilfield units will be used for this example. Select Options | Units to
view the units used by GAP for input and output, the data validation ranges,
and output precision. Near the top of the screen within the table heading
select Oilfield for both input and output units (as shown in Figure 3), and
then select OK.
Figure 3: Oilfield Units for Input and Output
2.1.2 STEP 2 : Initialise GAP Method Options
In this section the scope of calculations that GAP will be asked to perform will
be defined.
This example has a dry gas reservoir feeding a delivery pipeline 10000 ft
away. No production history is available, but the extent and composition of the
reservoir has been estimated, allowing a material balance prediction to be
performed.

TUTORIAL GUIDE
Select Options | Method to set the GAP calculation method.
Figure 4: Setting the options
Set Prediction type to On. This tells GAP how to update reservoir pressures
during a Prediction calculation. A constant reservoir pressure can be specified
by selecting None, in which case no prediction calculations are performed: in
this case, the model represents the system at an instant in time and no tank
models are required. Optimisation may be performed for both predictive
(depleting reservoir) and non-predictive (constant reservoir) calculations.
Set System type to Production, Prediction method to Pressure and
temperature, Optimisation method to Production, and Track Compositions
to No. The completed method screen is shown above. This completes the
GAP calculation method set-up. Select OK to return to the main GAP window.

2.1.3 STEP 3 : Define GAP Model Schematically
In this section the components required to model a simple gas production
network are specified. The properties of the components and reservoir fluids
are entered at a later stage using PROSPER and MBAL.
The model will consist of a reservoir tank, a gas production well and a pipeline
connecting the well manifold to the delivery pipeline. It is recommended that
the GAP model be specified from the separator (delivery pipeline) end towards
the reservoir, allowing the complexity to develop naturally as the model is
entered. Since this example is very simple it makes little difference in what
order the components are created.
The toolbar ( )
is used to create and modify components on the network schematic. Note that
when they are selected they remain active until they are unselected. The
exception to this is the ‘Delete’ icon, which must be selected for each deletion.
To identify an icon, hold the mouse cursor over it until a yellow box appears
with a short description of the icon function.
Select the Separator icon and click the left-hand mouse button in the
main GAP display area towards the top right. Give the resulting node the
name ‘Separator’ when requested. The name is just a label and is not
required, but it is generally a good idea to identify the nodes in this way. A
separator is considered by GAP to be the end of the production chain and will
be allocated a pressure at a later stage. It does not have to actually be a
separator, rather a convenient delivery point where a known pressure exists.
Next, select the Joint icon and add a joint. Name this ‘Manifold 1’. Place
the manifold to the left of the separator. Place a second “Manifold 2” to the left
of this. A joint is any manifold or intersection where pipes converge. Every
pipe must have both ends connected to joints.
Select the Well icon and add a well below the Manifold 2. Name this
‘Well’. Select the Tank icon and add a Tank, named ‘Tank’ below the
well.
Finally link the components together by selecting the Link icon and
dragging the left-hand mouse button between two components. Connect the
Tank to its Well, and the Well to its Manifold 2, then manifold 2 to 1 for the
pipe, finally connect the Manifold 1 to the Separator. Note that a pipe
component has been inserted between the two Manifolds. No pipe
components have been entered between the Tank, Well, and Manifold 2 since
any piping between these components is assumed to be implicitly defined by
the Well.

TUTORIAL GUIDE
Deselect the Link icon to prevent adding more links.
The basic model layout has been specified: additional components can easily
be added or deleted as the model is refined later. The GAP screen display
should look something like the following image.
Figure 5: Schematic Network Diagram
Use File | Save As to save the work done this far to a GAP file (Gasres.GAP)
in a suitable directory.
2.1.4 STEP 4 : Define the Well
In this step we specify the physical characteristics of the well and perforation
interval that will define the flow from the reservoir to the wellhead (‘Manifold’ in
this example). Although data may be entered directly into GAP, PROSPER will
be used to enter the well properties. The advantage of using PROSPER is
principally that the VLPs and IPRs can be generated later by automatic batch
calls to PROSPER from GAP. VLPs and IPRs are elements that describe the
performance of a well. We recommend the use of ‘VLP/IPR Intersection’ as
the well model.
Please refer to the GAP manual for more details on well models and available
options.
Additionally, matching to production history and sensitivity analysis of the well
can be performed using PROSPER.
Double click the left-hand mouse button within the Well component on the
GAP display screen. An Equipment Data Entry screen is then displayed. All of
the model components can be seen in the right window and data entry for any

component can be made by selecting the required component with a left
mouse click. The red crosses show where insufficient data has been entered.
Within the Well data entry screen change the Well Type to Gas Producer
(which changes the well colour from green to red). Select the Model as VLP /
IPR Intersection.
Figure 6: Equipment Data Entry Screen immediately before starting PROSPER
Select the “Run PROSPER” button and wait for PROSPER to load. Check that
the correct version of PROSPER has loaded, otherwise check File | Directory
from within GAP. Go to the PROSPER Gas Well Example (See Section 2.2
below) to set up the PROSPER model.
Go to the PROSPER Gas Well Example now – Section 2.2.
Once the PROSPER exercise is complete return to GAP from PROSPER by
selecting GAP. It is recommended that any changes made to a PROSPER file
are saved before returning to GAP.
On returning to GAP after creating a PROSPER file enter its file name in the
PROSPER Well File field of the Equipment Data Entry screen. Use the Browse
button to locate the file. The output (e.g. Gasres.OUT) PROSPER file should be
used in preference to the input (Gasres.SIN) and analysis (Gasres.ANL) files.
If the full drive and path to the PROSPER file are not entered in the Well File
field, then GAP will look for the PROSPER file in the directory defined by File |
Directories.

TUTORIAL GUIDE
Notice that the status of the Well File field has changed from Invalid to Valid.
To further check that the PROSPER file is properly located select Run PROSPER
and then return to GAP by selecting GAP once the PROSPER file has loaded.
Select OK to return to the GAP main window.
Now save the GAP file by clicking on and selecting Yes to the overwrite
confirmation.
2.1.5 STEP 5 : Calculate the Well IPR and VLP
The Inflow Performance Relation (IPR) and Vertical Lift Performance (VLP)
data can now be generated automatically by batch calls to PROSPER.
Select Generate | Generate Well VLPs | All | Data to specify the ranges of
data for which data should be generated. PROSPER is called to load the
values it already has stored. Select the Edit button next to the ‘Well’ label.
Enter the following data ranges:
Figure 7: VLP Data Ranges
Note that the Populate buttons can be used to create the tables rather
than the data being entered manually.

This table covers the range of possible gas production rates, manifold (well
head) pressures and water to gas ratios (WGR) that may occur. Since the
gas is a dry gas, the condensate to gas ratio (CGR) will always be 0.
Select OK and then Generate to perform the calculations using PROSPER: this
may take some time. Select OK | OK when the calculations have completed.
Go back to the main screen. Double-click on the Well icon to bring up the well
summary screen. Notice that the colour of the box next to the word ‘VLP’ has
turned green. This indicated that the VLP generation has been completed.
Figure 8: VLP in well screen showing OK
The inflow performance relationship (IPR) of the well has been described in
the PROSPER well model. We need to transfer the IPR from PROSPER to
GAP. For gas wells, while importing the IPR from PROSPER to GAP, GAP will
take three points from the PROSPER IPR, and fit the three points with either
Forcheimer or C n IPR method (Defined by user in the IPR screen of the
well in GAP).
In GAP, when selecting Generate | Generate well IPRs From PROSPER
means open the PROSPER file, read three points from PROSPER IPR, and fit
the points with the selected IPR method in GAP.
To gererate the IPR for this well, click on | Generate | Generate well IPRs
from PROSPER in the main GAP window.

TUTORIAL GUIDE
You will see a screen as below:
Figure 9: Generate error message
This is because we have not select any well yet. Click on | All to select all the
valid wells in the model (in this case only one).
Figure 10: Select wells message
The screen above shows you the selected wells. Click on | Generate to
proceed.

The IPR generation will begin and when it finishes, you will see the message:
Figure 12: IPR generation finish message
Click | OK to go back to the main screen. Double-click on the Well icon to
bring up the well summary screen.
Now save the GAP file by clicking on and selecting Yes to the
overwrite confirmation.
2.1.6 STEP 6 : Solve the GAP Network
This section describes the solving of the Network by the allocation of a
separator pressure from which an unconstrained flow can be calculated. This
is a precursor to the material balance calculation in which the reservoir
pressure reduces as the reservoir fluid is produced.
Double-click the left-hand mouse button within the Well on the main GAP
display area. Notice that all of the Data Summary flags are green and the Well
has a green tick next to it in the list of components on the right side of the
Equipment Data Entry screen. If this is not the case, then the VLP and IPR
have not been calculated correctly and STEP 5 should be repeated carefully.
Before the Network solver can be performed, the pipe must be defined.
Double click the left-hand mouse button over the pipe on the main window
and select Input at the bottom of the screen, followed by the Description tab,
leaving the default Environment properties. Enter the following pipe data:
Length = 10000 ft
TVD downstream = 0 ft
TVD upstream = 0 ft
Inside diameter = 6 INS
Roughness = 0.0006 INS
Select OK to return to the main screen.

TUTORIAL GUIDE
Figure 14: Pipe Data Input
Perform the Solve Network from the main menu and put Pressure 1
Separator Pressue at 1300 psig. Select Next | Calculate, when the
calculation is finished select Main. Provided that the pipe line is not a bottle
necked the pipe icon will stay blue. If there is a bottle neck the icon would
turn red.
The results can be seen by hovering the mouse over each node. The
following information will be seen: Pressure; Temp.; Qo; Qg; Qwat; Qginj; and
dP for the exit point of that item.

Figure 15: Check solution
If a pressure is allocated to the Separator, then the flow within the network
can be calculated by GAP tracing back through the component PCs from the
separator towards the well.
Now the results can be viewed. Select Results | Detailed | All Items and
note the gas production rate of 70 MMscf/day. Select OK to return to the main
GAP window.
confirmation.
2.1.7 STEP 7 : Material Balance Prediction
In this section a tank model is defined using MBAL, and a material balance
prediction of flow and pressure decline is undertaken.
Go to the MBAL Gas Reservoir Example now – see Section 2.3. The
standalone version of MBAL must be used to generate the tank model.
Double Click on the tank and then select Run MBAL in the Summary Screen.
If MBAL has been accessed from GAP, upon returning to GAP from MBAL, the
path of the MBAL file should be displayed in the Tank Summary Screen of

TUTORIAL GUIDE
GAP. However, if you prepare the MBAL file by running MBAL independently,
then you have to specify the file path of the MBAL model for GAP manually. To
do this, double click the left-hand mouse button over the Tank within the main
display, and select Browse to locate the MBAL file (GASRES.MBI). Notice
that the Tank component now has green tick beside its name in the right side
of the Equipment Data Entry screen.
Figure 16: Tank summary screen
The material balance tank model is now in place and a prediction can be
performed. A straightforward prediction will be run first, with no constraints or
events occurring during the production. This is essentially the same as the
Solve Network calculation performed previously, except that a material
balance calculation is performed after each time step to update the reservoir
pressure and PVT properties.
Select Prediction | Run Prediction and set the following time control data.
• Start Date 01/01/2005
• End Date 01/01/2020
• Step Size 1 Year(s)
Select Next | Next and allocate a Separator pressure of 1300 psig. Select
Next | Calculate and allow the Solve Network cycle to be performed for each
of the 15 time steps requested.

Figure 17: Calculation screen
When the run is complete select Back | Back.
The results from the material balance prediction can be viewed by selecting
Plot Nodes, and highlighting the Separator, Manifold 1 and 2 and Well nodes
from the resulting list. Since the components are all in series, the flow
parameters should be identical for each node and have overlying curves.

TUTORIAL GUIDE
Figure 18: Select nodes to plot
Click on Plot and a plot window will appear. Select Variables and look at the
plot for Gas rate by highlighting it and selecting OK. Plot the Water rate,
Reservoir pressure and Cum Gas Production graphs.

Figure 19: Results – Gas rates
The initial peak gas rate should be 70 MMscf/day, and the peak water rate
should be 43 STB/day. This water is the vaporised and connate water
released as the reservoir depressurises and the water and formation rocks
contract. Select Main to return the main GAP window.
No constraints have been entered for this system, and it is recommended that
none are entered until the potential of the system has been established. At
this point the user should consider design options, potential problems and
possibly sensitivity analysis.
confirmation.
2.1.8 STEP 8 : Constraints
In this section a constraint will be applied to the maximum flow rate that can
be passed through the separator. One of the wells will initially have to be
choked back to satisfy the separator constraint.
Select the well icon and add a new well next to the current well.

TUTORIAL GUIDE
The already defined well properties (VLPs, IPR, PVT etc) can be copied to the
new well by holding the Ctrl key down while selecting the first well with the left
mouse button, and dragging the mouse over the new well.
Add a link between the new well and the Manifold 2 using the link icon, and
then deselect the link icon.
Figure 20: Adding a second well
Enter the Summary Data Entry screen for the second well by double-clicking
on the icon with the left-hand mouse button. Change its label to Well 2 in the
top left of the screen, and then click on the green area next to dP Control in
the lower part of the screen. These buttons are accelerators to different
screens of equipment input data. Set the “dP Control” box to Yes and the
Delta P Choice to Calculated. This will simulate the presence of a well head
choke that allows GAP to reduce the flow from the well and meet any
constraints imposed on the system.

Figure 21: Setting the well to controllable choke
Select OK. The potentially choked (controllable) well will have a ring around it.
Figure 22: Controllable well has a thin red circle around it
Enter the Separator data entry screen by double-clicking the left-hand mouse
button over the separator icon. Navigate to the Constraints data entry by
clicking on the Constraints accelerator in the lower half of the Equipment Data

TUTORIAL GUIDE
Entry screen. Enter a Max gas production of 100 MMscf/day, and then select
OK.
Figure 23: Setting constraint at separator
The separator constraint is shown on the display as two inward pointing
arrows, as shown in the figure below.

Figure 24: Schematic Diagram with Controllable Well and Constrained Separator
Select Solve Network | Next click on the Optimise and Honour Constraints
option:
Figure 25: Calculation screen showing Optimise checkbox

TUTORIAL GUIDE
and Calculate to solve the system with constraint, then Main when the
calculation has finished. As you have two wells and one is selected with dP
control, using optimise the solver will control Well 2 to achieve the constraint
set at the Separator. Go to Results | Detailed | All Wells. The Gas
production for the two wells can then be noted. Use Next to move to Well 2
and note that the production has been choked back to 33 MMscf/day to
achieve the constraint at separator.
Notice that the pipe icon has changed colour from blue to red. This indicates
that this pipeline is bottlenecking the system.
Figure 26: Bottle-neck pipeline
Double click on the pipe and select Results | Network Solver tab screen and
note that the Bottle Neck flag shows Choked. From this we see that the
combined flow from the two wells can be greater, but the pipe between the
Manifold 2 and Manifold 1 is bottlenecking the system.

Figure 27: Bottle neck flag
If a prediction is now done (selecting to honour the constraints), then Well 2
will be choked back as long as the potential of the system is greater than the
constraint set:
Figure 28: Well results
Save the GAP file using and select Yes to overwrite the current file.

TUTORIAL GUIDE
2.2 PROSPER Gas Well Example
PROSPER program. The emphasis is on the data entry required to model a dry
gas producing well for inclusion into a GAP model. See the GAP Gas Network
Example for further details. Since it is hoped that this example will be used as
a phase in the GAP Gas Network Example, it is anticipated that PROSPER will
have been loaded from within GAP. However if that is not the case, this
example can be run using the standalone version of PROSPER.
PROSPER is a single well characterisation program. Its output is principally
Inflow Performance Relations (IPRs) and Vertical Lift Performances (VLPs).
These relations respectively describe the inflow to the well sandface from the
reservoir and the outflow from the well sandface to a manifold (or well head)
at the top of the well. These pressure and flow correlations are heavily reliant
on the PVT (Pressure, Volume, and Temperature) characteristics of the
produced fluid. Using Inflow and Outflow, we know the behaviour of the well in
terms of the flow rates vs. bottom hole pressures for a given mean reservoir
pressure.
In addition PROSPER has tools to match known correlations to observed
production history and perform detailed sensitivity analyses.
2.2.1 STEP 1 : Initialise PROSPER
If PROSPER has not been started from with GAP, then start the PROSPER
program by running PROSPER.EXE, which can be found in the Petroleum Experts
directory (default C:Program FilesPetroleum ExpertsIPM 5.0). See the
PROSPER manual for more details on how to start PROSPER.
Check that the current version of PROSPER has been loaded. The version of
PROSPER being used can be seen in the title bar.
The command options (File, Options etc.) at the top of the PROSPER window
are laid out in a logical order (left to right) that reflects the order in which
operations will usually be performed.
Select File | New to start a new file if required. If this option is not available,
then PROSPER has already started a new file.
Note that files saved with this version of PROSPER will not be readable by
previous versions. Select File | Preferences followed by the File tab. It is
recommended that the Default Data Directory field is set (using the Browse
button) to point to a directory that is exclusively used to store data files
created with the current software version.
Now select the Units tab. It is important to ensure that consistent units are
used throughout, particularly when data generated by PROSPER may be
incorporated into an MBAL or GAP model. Oilfield units will be used for this

example. Ensure that Input Units and Output Units box have Oilfield
selected. Select Done to return to the main PROSPER window.
Figure 29: Preferences screen
2.2.2 STEP 2 : Initialise PROSPER Method Options
In this section the type of well and reservoir fluid that PROSPER will use are
defined. Their detailed specification will be entered later.
This example has a dry gas producing well. Select Options | Options to
display the System Summary screen. This screen is primarily used so that
PROSPER can provide only the relevant screens as the model is constructed.
Set the options shown below and click Done.

TUTORIAL GUIDE
Figure 30: Completed System Summary Screen
2.2.3 STEP 3 : Initialise PVT Data
This simple example will use an unmatched Black Oil PVT to characterise the
reservoir fluid. Select PVT | Input Data to enter the PVT data. Notice the
options to match correlations to data, or to use lookup tables of PVT data.
Enter the following data and select Done.
Gas gravity 0.59
Separator pressure 100 psig
Condensate to Gas Ratio 0 STB/MMscf
Condensate gravity 50 API
Water to Gas ratio 0 STB/MMscf
Water salinity 10000 ppm
Mole Percent H2S 0 %
Mole Percent CO2 0 %
Mole Percent N2 0 %
Reservoir Pressure 11500 psig
Reservoir Temperature 230 deg F

Figure 31: PVT Data Screen
The condensate gravity of 50 API will not be used for a dry gas, however a
value greater than 5 is required by default. See the PROSPER manual on
details of how to change unit range defaults.
2.2.4 STEP 4 : Initialise Well Inflow and Equipment
This step defines the properties of the reservoir and well that will determine
the flow rate of the produced fluid for a given reservoir pressure and well head
pressure.
Select System | Equipment (Tubing etc) to input the well properties. Select
All, and then Edit. Enter the following deviation survey data describing a
vertical well profile down to a depth of 17350 ft. Click Done when the
deviation survey data has been entered.
Measured Depth (ft) True Vertical Depth (ft)
0 0
17350 17350

TUTORIAL GUIDE
Figure 32: Deviation Survey
No surface equipment will be entered for this simple model. Note that all
equipment between the well head and manifold defined in GAP would in
general have to be entered here. Select Cancel.
Enter the following tubing and casing data in the downhole equipment screen,
and then click Done.
Type Measured depth (ft) Inside diameter (in) Roughness (in)
X’mass tree 0 - -
Tubing 17250 2.992 0.0006
Casing 17350 6 0.0006

Figure 33: Downhole Equipment
This model is performing a pressure and temperature calculation, therefore
the temperature of the surrounding formations and a mean heat transfer
coefficient are required. Enter the following linear geothermal gradient and
then select Done.
Measured Depth (ft) Formation temperature (o
F)
0 60
17350 230
Overall heat transfer coefficient 3 BTU/ft2/F/hr

TUTORIAL GUIDE
Figure 34: Geothermal Gradient
For Average Heat Capacity, we use the default values. Click on Done to
accept the values.
We are now back to the main Equipment Entry screen. Select Summary |
Draw Downhole to view a schematic of the downhole equipment that has
been entered. Select Main to save the input data and return to the main
PROSPER window.

Figure 35: Downhole equipment sketch
The data required to calculate VLPs has been defined, although the VLP
correlation function has not yet been entered. Now the reservoir inflow must
be characterised by defining an IPR.
In PROSPER main screen, select System | Inflow Performance to open the
IPR Input screen. Highlight the Petroleum Experts Reservoir Model and
Enter Skin By Hand for the Mechanical/Geometrical Skin. Set the following
data in the lower right of the screen.
• Reservoir Pressure 11500 psig
• Reservoir Temperature 230 degrees F
• Water Gas Ratio 0 STB/MMscf
• Condensate Gas Ratio 0 STB/MMscf

TUTORIAL GUIDE
Figure 36: Inflow performance model selection
Select the Input Data button at the top right of the screen and enter the
following data within the Reservoir Model tab.
• Reservoir Permeability 20 md
• Reservoir Thickness 100 feet
• Drainage Area 2500 acres
• Dietz Shape Factor 31.6
• Wellbore Radius 0.354 feet
• Perforation Interval 30 feet
• Time 100 days
• Reservoir Porosity 0.2
• Swc 0.2
Select the Mech/Geom Skin tab and enter a value of 2. Select Calculate. An
IPR plot showing the inflow to the well as a function of the well’s sandface
pressure will be shown. An AOF of 143 MMscf/day is shown as the cutoff
(maximum) flow.

Figure 37: Inflow performance relation plot
When building field models and the answer for AOF is 150 MMscf/day, then
note the y-axis minimum pressure is likely not be at 0 (zero). This is not
maximum flow. Check the limits for maximum AOF. Select Main and then
select File | Preferences and tab Limits. Note the Maximum AOF for GAS
and change it to say 2000. Select Done | System | Inflow Performance |
Calculate, the displayed AOF will now be OK. The y-axis minimum pressure
will be 0 (zero), i.e. maximum flow. Select Main to save the data and return to
the main PROSPER window.
There is no production history to match data with so we will move on to the
calculation phase to check that the IPR and VLP that had been defined
appear reasonable. Select Calculation | System (IPR+VLP) | 3 variables
and enter the following data.
• Top Node Pressure 1500 psig
• Water Gas Ratio 0 STB/MMscf
• Condensate Gas Ratio 0 STB/MMscf
• Vertical Lift Correlation Petroleum Experts 2
• Solution Node Bottom Node
• Rate Method Automatic - Linear

TUTORIAL GUIDE
Figure 38: System calculation entry screen
Please note that although a WGR of 0 was entered, the program will use the
Minimum WGR calculated in the PVT screen (vaporized water).
If matching had been performed, the correlations would have been chosen
and fitted using non-linear regression. As it is, unmatched correlations will be
used.
Select Continue | Continue | Calculate and allow PROSPER to perform the
calculation. Confirm the end of the calculation by clicking OK. Scroll right
(towards the bottom of the screen) within the Results display until the dP
Friction and dP Gravity columns are shown. Notice that for moderate and
large gas flow rates, the frictional pressure drop within the well dominates the
gravitational pressure drop to such an extent that these flow rates are unlikely
to ever be achieved, suggesting that perhaps a larger diameter well should be
considered. Select Plot to display the results.

Figure 39: IPR and VLP Curves
The X-axis shows the produced gas flow rate and the Y-axis shows the well
sandface pressure. The reservoir pressure has been set to 11500 psig and
the well head pressure to 1500 psig. For these pressures, the IPR (green
curve) and VLP (red curve) intersect at a well sandface pressure of 8564 psig
and flow rate of 73 MMscf/day, these being the flow conditions that the well
would actually achieve (i.e. the unique flow pressure solution that lies on both
the IPR and VLP curves). If the mouse cursor is moved within the plot, the X
and Y coordinate values are displayed at the top right of the screen.
The relatively steep gradient of the VLP curve compared with the IPR curve
indicates that most of the available pressure drop from the reservoir to the
well head will be within the well as a consequence of its large frictional
resistance.
Select Finish and note that the solution values are displayed on the right of
the Calculation Output screen.
Select Main and File |Save As to save the PROSPER data. Enter the file name
(Gasres.OUT) in a suitable directory, remembering not to overwrite a
PROSPER file generated with an earlier version of PROSPER, and then select
Done.
If PROSPER was being run from GAP, select GAP and return to the GAP Gas
Network Example documentation, otherwise select File | Exit.
If you are following the example for building a GAP, PROSPER, MBAL model,
you have been directed to jump to this Section 2.2 from Section 2.1.4 (pg.
14). You have now completed Section 2.2. Go back to pg. 14 now.

TUTORIAL GUIDE
2.3 MBAL Gas Reservoir Example
MBAL program. The emphasis is on the data entry required to model a dry
gas reservoir for inclusion into a GAP model. See the GAP Gas Example 2.1
for further details. This example should be run using the standalone version of
MBAL.
MBAL is a reservoir analysis tool that uses the production history of a
reservoir and the PVT characteristics of the production fluid to perform mass
balance calculations to estimate the Stock Tank Original Oil In Place
(STOOIP) and identify the driving mechanisms within the reservoir (fluid
expansion, formation expansion and aquifer inflow). Good PVT
characterisation and production history are usually an essential input to an
MBAL calculation, but for this tutorial example a minimum of input data is
required.
With respect to the GAP Gas Example, the purpose of the MBAL model is to
define the reservoir characteristics so that material balance prediction
calculations can be performed by GAP.
2.3.1 STEP 1 : Initialise MBAL
Start the MBAL program by running MBAL.EXE, which can be found in the
Petroleum Experts directory (default C:Program FilesPetroleum ExpertsIPM 5.0). See
the MBAL manual for more details on how to start MBAL.
If you are starting from GAP, double click on the tank and select Run MBAL
button at the top of the screen.

Figure 40: MBAL main screen
Check that the current version of MBAL has been loaded. Select Help | About
MBAL Package to check the version number.
Select File | New to start a new file, and then select Tool | Material Balance
to start an MBAL material balance session.
The command options (File, Tool, Options etc.) at the top of the MBAL
window are laid out in a logical order (left to right) that reflects the order in
which operations will usually be performed. Note that files saved with this
version of MBAL will not be readable by previous versions. It is therefore
recommended that the File | Data Directory option is set to point to a
directory that is exclusively used to store data files created with the current
software version.
It is important to ensure that consistent units are used throughout, particularly
when data generated by MBAL may be incorporated into a GAP model. Oilfield
units will be used for this example. Select Units to view the units used by
MBAL for both input and output, as well as the expected data ranges. Select
Oilfield for both input and output units, and then select Done.

TUTORIAL GUIDE
Figure 41: Oilfield Units for Input and Output
2.3.2 STEP 2 : Initialise MBAL Method Options
In this section the type of reservoir fluid and tank model that MBAL will use will
be defined. Their detailed specification will be entered later.
This example has a dry gas reservoir. Select Options to display the System
Options screen. This screen allows MBAL to guide you through the model set-
up by only presenting the relevant screens as it is constructed. Set the options
shown in the figure below and then select Done.

Figure 42: Completed System Options Screen
2.3.3 STEP 3 : Initialise PVT Data
This simple example will use an unmatched Black Oil PVT to characterise the
reservoir fluid. Select PVT | Fluid Properties to enter the PVT data. Note the
options to match correlations to data, or to use lookup tables of PVT data.
The PVT data used by MBAL must be the same as that used by PROSPER if an
integrated GAP model involving MBAL and PROSPER is to be used. To aid this
process, MBAL can import the PVT data used by PROSPER by using the Import
button to import a PVT file generated by PROSPER (e.g. GASRES.PVT). If this
is done, then the same matching to correlations or tabulated values must be
initialised within MBAL.
If data is not imported from a PROSPER generated PVT file, then enter the data
as shown in the figure below, and select Done. The condensate gravity of 50
API will not be used for a dry gas, but a value greater than 5 is required by
default. See the MBAL manual for details on how to change unit range
defaults.
Gas gravity 0.59
Separator pressure 100 psig
Condensate to Gas ratio 0 STB/MMscf

TUTORIAL GUIDE
Condensate gravity 50 API
Mole percent H2S 0 %
Mole percent CO2 0 %
Mole percent N2 0 %
Figure 43: PVT Data Screen
2.3.4 STEP 4 : Initialise Tank Parameters
This step defines the physical properties of the reservoir required for material
balance calculations.
From MBAL main screen, select Input | Tank Data to input the tank
properties. Add the following parameters to each of the available tabs within
the Tank Input Data screen. Use the Validate button at the bottom of the
screens to validate the data input.
2.3.4.1 Tank Parameters
• Tank Type Gas
• Temperature 230 degrees F
• Initial Pressure 11500 psig
• Porosity 0.2 fraction
• Connate Water Saturation 0.2 fraction
• Water Compressibility Use Corr 1/psi
• Original Gas In Place 600 Bscf
• Start of Production 01/01/2005

2.3.4.2 Water Influx
• Model None
2.3.4.3 Rock Properties
• Check the From Correlation button.
2.3.4.4 Rock Compaction
• Click on “reversible”
2.3.4.5 Relative Permeabilities
• Rel Perm. From Corey Functions
• Water Sweep Eff. 100 percent
Residual Saturation End Point Exponent
Krw 0.2 0.8 2
Krg 0.01 0.9 1.5
Note that the residual saturation for the water relative permeability
corresponds to the connate water saturation.
There is no Pore Volume vs Depth or Production History data to be entered.
Select Done when the data has been entered.
With no production history, no history matching is possible. Select File | Save
As to save the MBAL data. Enter the file name GasRes.MBI in a suitable
directory, remembering not to overwrite a file generated with an earlier version
of MBAL.
If MBAL was being run from GAP, select GAP. Otherwise, select File | Exit.
If the GAP Gas Example is being followed then return to the GAP
documentation, otherwise the MBAL Oil History Matching Example in the
Physics section may be used to demonstrate the history matching features
available in MBAL.
If you are following the example for building a GAP, PROSPER, MBAL model,
you have been directed to jump to this Section 2.3 from Section 2.1.7 (pg.
22). You have now completed Section 2.3. Go back to pg. 22 now.

TUTORIAL GUIDE
2.4 GAP Gas-Lifted System
The main objective of this example is to show how GAP can be used to
optimize the gas lift allocation to gas lifted wells in a simple production
system, and hence optimize the total oil production from the field.
2.4.1 Introduction
This tutorial not only offers a step-by-step guide to the setting up of the
problem, but also gives an overview of other GAP functionality that can be
used alongside, in addition to, or instead of the given approach. These points
will be made in the body of the text.
It is advised that the step-by-step guide is followed on the computer, entering
data as requested. We would encourage browsing around the system as you
proceed through the guide: this can be a useful way of learning about other
features of GAP that are not described here. For more detail on a particular
feature, please refer to the main GAP manual.
Menu commands are described in this tutorial using (for example) the
following scheme: File | Exit means select the Exit option from the File drop-
down menu item.
2.4.1.1 Definition of the Problem
The system that we are to set up is as follows:
• An oilfield has two gas lifted wells: well GL#1 and well GL#2.
• Each well is tied back to the riser base via a 1500 ft flowline.
• Each flowline has an ID of 5 inches.
• The riser is 500 ft long and has an ID of 10 inches.
• The platform is at 500 ft above the seabed. The seabed is assumed to be
flat.
2.4.1.2 Step-by-Step Approach
We summarise here the standard steps taken in building a network model
from scratch. Although the actual steps taken in building a model will vary
depending on the model, the following list gives an indication of the amount of
work that needs to be done to set up an accurate field reproduction.
The steps to be followed are:
• Setting up the system.
• Drawing the system.
• Setting up the well models.

• Describing the surface network.
• Generating the inflow performances from existing well models.
• Generating lift curves for the wells.
• Allocating optimally the amount of gas available.
• Analysing the results.
2.4.2 Step 1: Setting up the System
At the end of this step we want to have initialised GAP ready for construction
of the network. The steps are:
• Start a new file.
• Set up the optimisation method.
• Set up the units.
• Set up the gas injection source.
2.4.2.1 Starting a New File
To start a new file, choose File ⏐ New. This option clears the current screen
display and resets the program workspace to initial values.
2.4.2.2 Setting up the Optimisation Method
To set-up the optimisation method, choose Options | Method
Select as Input Parameters:
• Prediction: None
One can run predictive models in GAP, either using a simple decline curve
model or by linking to Petroleum Experts’ MBAL program to perform Material
Balance calculations. Connectivity to Petroleum Expert’s REVEAL
numerical simulator can also be done.
• System type: Production
Water and gas injection systems can also be modelled. When performing a
prediction run, these injection systems can be associated with a production
system to provide voidage replacement (for example) into the producing
reservoirs.
• Prediction Method: Pressure and temperature
This allows GAP to perform pressure and temperature drop calculations in
pipeline models.
• Optimisation Method: Production
You can also optimise with respect to revenue (in which case you must enter
value parameters here).

TUTORIAL GUIDE
• Track composition: No
GAP allows fluid compositions to be tracked from the well (or reservoir)
level to the top node. In this example, only black-oil properties will be
reported.
Figure 44: System options
The Ok button can now be clicked to finish this step.

2.4.2.3 Setting up the units
To set-up the input and output units, click on Options | Units and select the
unit system you want to use.
Clicking on the cell below the Input or Output column header (defaulted to
Oilfield) will yield a selectable list of available units systems. This example will
be worked in oilfield units throughout.
For more information, please refer to the online help or the GAP manual.
Figure 45: Setting up unit system
Click on Ok to complete this step.

TUTORIAL GUIDE
2.4.2.4 Setting Gas Injection Sources
If you have gas injection in your system, either for gas lifted wells or for gas
injection into a reservoir during a prediction run, then it is convenient to set the
gas injection parameters at this stage.
GAP maintains a list of gas injection sources with different gas gravities and
impurity levels (and compositions if compositional tracking is enabled). These
can be edited by selecting Options |Injection fluids. When a new file is
created, a default entry is supplied with a specific gravity of 0.7 and no
impurities. You may edit this entry, or add your own in the table.
Figure 46: Gas lift gas properties setting
When you set up your gas lifted well models, you will be able to apply any
source in the list to the well in question.

2.4.3 Step 2: Drawing the system
By the end of this step, we want to have a basic schematic set up on the main
screen. The equipment data can be entered once the network is in place.
2.4.3.1 Adding Wells
To create the wells icon, click on the ‘Add Well’ icon, from the toolbar.
One may now click on anywhere on the screen and a well icon will be created
at that point. Whenever an equipment icon is created, a label can be entered.
Click OK once the well name is entered.
The first well will be labelled GL#1 and the second well GL#2. Users are
encouraged to use real well names as labels for their wells.
If you want to move a well icon from one place to another on the screen, then
hold down the |Shift key, point the cursor to the well icon and then move it to
the desired place. Alternatively, select the Move tool from the toolbar and drag
the item to the new location.
A well icon can be deleted by clicking on the Delete button on the toolbar and
then clicking on the well icon that needs to be deleted. If a piece of equipment
needs to be removed from the system temporarily, then it is preferable to
Mask the item – select the Mask tool from the toolbar to achieve this.
Figure 47: Two wells have been added to the system
See the GAP manual for more details on user interface functionality.

TUTORIAL GUIDE
Â TIP: Buttons (such as ‘Add Well’) can be selected by clicking the
right-hand mouse button in the window area to create a drop-down
menu. Alternatively, the tools are also selectable from the toolbar
buttons.
2.4.3.2 Tie-backs
Joints are needed to hook up the wells to the tiebacks. Joints (or manifolds)
are used as connection tools in GAP.
To create a joint icon, select the ‘Add Joint’ option from the toolbar. Click on
the screen at the position where you would like the icon to be placed (above
each well icon, for instance).
The joint to be connected to the well GL#1 will be labelled WH1-GL#1, and
the second joint will be labelled WH2-GL#2. Again, users are encouraged to
use real joint names as labels for their manifolds.
We will also need to create a third joint that will gather the production from the
tiebacks. This will be labelled ‘Collector’.
We will also need to create a fourth joint that represents the riser top. This will
be labelled ‘Riser top’.
These joints will be connected together with pipes at a later stage.
Figure 48: Joints have been added to the system

2.4.3.3 The Platform
The platform is represented as a separator.
To create a separator icon, select the ‘Add Separator’ option from the tool bar.
Click on the required location on the screen and an icon will be created, as
above. This separator will be labelled ‘Platform’.
2.4.3.4 Pipes/Links
Â TIP: Pipes are created using the ‘Add Link’ tool from the toolbar. The
reason for this name is that this tool can also be used to create logical
connections (for example, well to reservoir, or compressor to
manifold): whether a pipe or a logical connection is made depends on
the equipment being connected
In order to connect the different equipment, we will now select the ‘Add Link’
option from the toolbar.
To hook up the well GL1 to the joint WH1-GL1, situate and click the cursor in
the centre of the well icon GL1, and drag a connected to the WH1-GL1 icon.
Repeat the process with GL2 and WH2-GL2.
Repeat the process between WH1-GL1 and Collector and WH2-GL2 and
Collector.
Link the Collector manifold to the Riser Top: this will become the Riser.
Finally, link the Riser top to the Platform.

TUTORIAL GUIDE
Figure 49: The whole system
Â Note: At this point, you will see that the pipes between the wells and
the collector manifold, and the collector and the riser top, contain an
icon to represent the flow-line data, whereas those between the wells
and the wellhead manifolds, and that between Riser top and Platform
do not. This is because GAP expects the well model to include all
equipment up to the well head, and so does not attempt to model any
pressure drops here. Pressure drops are modelled for all other pipes
and depend on a pipeline description, as described below.
Labelling
The tieback between WH1-GL1 and the collector will be labelled ‘TieOne’.
The tieback between WH2-GL2 and the collector will be labelled ‘TieTwo’.
To label the tieback between WH1-GL1 and the collector, double-click on click
on the pipeline icon between the joint WH1-GL1 and the collector and enter in
the label field ‘TieOne’. Click on Ok to complete. Repeat for the pipe between
WH2-GL2 and the collector. Repeat for the pipe between the collector and the
riser, and label this ‘Riser’.

Figure 50: Labelling the pipeline

TUTORIAL GUIDE
2.4.3.5 Other Drawing Options
The user interface is configurable in several ways.
• You may want to give a name to the model. For this, click on Options |
System Summary and enter the title ‘Tutorial GAP Example’ (for
example). This text will now appear as a heading for the system
network on the application screen.
• Clicking the right hand mouse button in the application screen and
selecting the Fonts option can change the screen fonts. Clicking the
right hand mouse button in the title can change the title font.
• Selecting Icon Sizes from the same drop-down menu can change the
sizes of the icons on the GAP screen. This may be useful if building a
large model.
Consult the user manual or online help for more options.
Figure 51: Labelling the system
The basic schematic is now set up, as shown above. The next step is to
describe the equipment comprising the network.

2.4.4 Step 3: Describing the wells
By the end of this step, we want to have each network well modelled
accurately.
It is recommended that the system is described from the wells to the top node.
There are various quality checking functions that can be performed at the well
level prior to building the whole system. These will be demonstrated in the
following chapters.
Â TIP (note on entering equipment data): The basic means of entering
data is from the equipment data entry screen. This can be accessed
by double clicking on any equipment icon. The data entry screen
consists of a data entry area and a list of network equipment.
Descriptions of several pieces of equipment can be entered in one
edit session by clicking on the entries in the equipment list to bring
up different entry screens.
2.4.4.1 Entering Well Data
Well GL1
To enter the equipment data entry screen, double click on the Well GL1 icon.
This will lead you to the ‘Well Data Entry - Summary screen’ for this well.
Enter the following data:
• Label: GL1
This has already been set when the icon was created.
• Mask: Include In System
• Well Type: Oil Producer (Gas Lifted)
• Well file: C:Program FilesPetroleum ExpertsIPM 5.0Worked
examplesDexterity examples GL#1.out
• Wells can be modelled using Petroleum Experts’ PROSPER package, as in
this example. Enter the above PROSPER well file in this field, either
typing it directly or using the ‘Browse’ button to invoke a file browser.
• Model: VLP/IPR intersection
Click on Ok to complete this step, or navigate directly to the next well.
Well GL2
The above comments apply also to Well GL2. Enter the following data:-
• Label: GL2
• Mask: Include In System
• Well Type: Oil Producer(Gas Lifted)
• Well file: C:Program FilesPetroleum ExpertsIPM 5.0Worked
examplesDexterity examples GL#2.out
• Model: VLP/IPR intersection

TUTORIAL GUIDE
Â Note: All the data here was entered on the summary screen. Note
that the data entry screen is divided into three parts as indicated
from the toggle buttons at the bottom right of the screen: Summary,
Input, and Results. Click on the input and results buttons and have a
look at the various categories of data that are available for entry or
viewing: for example, the first tab on the input screen allows you to
set up the gas lift injection source for the well.
2.4.4.2 Generating IPRs From Existing PROSPER Well
Models
By entering a well model file on the summary screen, we have associated this
well with a PROSPER well model stored on disk. We can now import IPR data
for the well directly from PROSPER.
When IPRs are transferred, GAP receives three points that lie on the PROSPER
IPR along with PVT parameters and reservoir pressure. GAP then performs a
match to this data to obtain the PI.
To transfer the well IPRs from the existing PROSPER well models, click now on
Generate | Transfer Well IPRs from PROSPER and then follow the on-screen
instructions. The following screen will be displayed:
Figure 52: Generate error message
This is because we have not selected any well yet. Click on All to select all the wells.

Figure 53: Select wells message
The screen above shows you the selected wells. Click on | Generate to
proceed.
The IPR generation will begin and once started, the IPRs are transferred as a
batch job and no user intervention should be required. When it finishes, you
will have to minimise PROSPER in order to see the message:
Figure 55: IPR generation finish message
Click | OK to go back to the main screen. Double-click on the Well icon to
bring up the well summary screen. Notice that the colour of the box next to the
word ‘IPR’ has turned green. This indicated that the IPR generation has
completed.
Now save the GAP file by clicking on , save the file as ‘Tutorial Gas Lift
Example.GAP’.

TUTORIAL GUIDE
Â Note: IPR parameters can be entered by hand and matched from the
IPR input screen. From the well data entry screen, select the input
button and navigate to the IPR tab. As you will note, this has been
filled in automatically by the above process.
2.4.4.3 Importing Existing Lift tables to the Well
Models
A well is basically defined by an inflow and an outflow; the inflows (IPRs) have
been already transferred to the wells in the above procedure.
In order to import/assign the VLP to the well GL#1, double-click on the GL#1
well icon, click on the VLP item (should be red if not valid) and, using the
Browse button select the file Program Files Petroleum Experts Samples
Worked Examples Dexterity examples GL#1.VLP. Note clicking on the VLP
item on the summary screen is equivalent to selecting the Input button
followed by the VLP tab.
Figure 56: Assign the VLP files to the well
Repeat for the second well with file: Program Files Petroleum Experts
Samples Worked Examples Dexterity examples GL#2.VLP
Lift curves can be plotted or inspected by clicking on the Plot buttons of the
VLP screen.

Â Note: In this case we are simply assigning pre-calculated VLP files to
the lift curve entries of the wells. In general use, you would have to
create these files. Once a PROSPER file has been assigned to the
well, lift curves can be calculated by PROSPER by selecting Generate
| Generate Well VLPs. Alternatively, GAP can import .TPD files
(generated by PROSPER) to make .VLP files. To do this, click on
Import on the VLP screen and select the required import file.
2.4.5 Step 4: Describing the Network
At the end of this step, all the remaining network equipment will be modelled.
In the following discussion we enter true vertical depths (TVDs) with respect
to the platform. Thus we define the platform to be at zero ft TVD such that the
collector and tiebacks are at 500 ft TVD.
2.4.5.1 Riser Description
To describe the riser, click on the pipeline icon between the collector and the
platform, labelled ‘Riser’. This will lead to the ‘Pipe Data Entry - Summary
Screen’.
Enter the following data:
• Correlation: Petroleum Experts 4
• Correlation Coefficients: 1 and 1 (default)
Now go to the input screen (by clicking on the ‘Input’ button) and enter the
following:
• Environment: default
This can be used to set up special pipe environmental quantities such as
ambient temperature or heat capacities. The default entries are suitable
for our requirements.
Finally, we enter the physical description of the pipe. Go to the ‘Description’
tab and enter the following:
• Enter 0 ft for the downstream TVD (Platform)
• Point the cursor to the first cell in the second row in the ‘Segment
Type’ column and select ‘Line pipe’:
• Length: 500 ft
• TVD: 500 ft
• ID: 10
• Roughness: 00006 (default)

TUTORIAL GUIDE
Figure 57: Riser description
Â TIP: If you have real data for your pipeline flows, then it is a good
idea to match the correlation that you are using to this data. To do
this, click on the Match button of the description entry screen and
follow the instructions detailed in the on-line help or the user manual.
Â TIP: If you do not enter pipe data, then the pipe will be treated as a
simple logical connection between two nodes, and zero pressure
drop will be modelled across it. GAP does not insist that you enter
pipe data.
Click on Ok to complete this, or navigate to the next pipe.

2.4.5.2 Description of the tie-back ‘TieOne’
The above process detailed for the riser is repeated for the other system
pipes.
The pipeline description is:
• Enter 500 ft for the downstream end (Collector)
• Select ‘Line pipe’ in the first cell in the second row in the ‘Segment
Type’ column.
• Length: 1500 ft
• TVD: 500 ft
• ID: 5
Figure 58: Tie One Description
Navigate to the final pipe.

TUTORIAL GUIDE
2.4.5.3 Description of the tie-back ‘TieTwo’
The pipeline description is:
• Enter 500 ft for the downstream end (Collector)
• Select ‘Line Pipe’ in the first cell in the second row in the
‘Segment Type’ column.
• Length: 1500 ft
• TVD: 500 ft
• ID: 5
Figure 59: TieTwo Description
Now click on Ok to complete this.

2.4.6 Step 6: Allocating the Amount of Gas Available
We are now in a position to allocate gas lift for optimum production. In this
step, we want to determine (given a total quantity of available gas) the
optimum amount of gas to be injected in each well.
In order to perform an optimisation, click on | Solve Network and then enter
different amount of gas lift gas available.
Figure 60: Specifying cases with different gas lift gas available
Gas available (MMscf/d)
0
3
6
10
20
Click on | Next. For the platform pressure, enter 250 psig.

TUTORIAL GUIDE
Figure 61: Specifying the separator pressure
Click on | Next | Calculate. Make sure that the ‘Optimise and Honour
Constraints’ check box is ticked before the calculation is started. GAP is going
to allocate the available gas to the wells to maximise the oil production. When
the calculation is finished, click on | Main to go back to the main screen.

2.4.7 Step 7: Analysing the results
To see the effect of the optimised injection of increasing amount of lift gas,
click on Results | Detailed | All Separators and Injection Manifolds and a
screen similar to this is displayed:
Figure 62: Allocation results
The natural flow production of this production network system is about 4000
BOPD. With 6 MMscf/day of gas injection, an optimal allocation would
increase the production to around 4990 BOPD.
We also see from these results that increasing the total gas injection beyond
10 MMscf/day does not increase the amount of production. The maximum
production available from this system is nearly 5000 BOPD.
A plot of oil production against lift gas injection can be displayed by clicking on
Plot.

TUTORIAL GUIDE
Figure 63: Allocation results
The optimal contribution/distribution between the wells can be viewed by
clicking on Results | Summary | All Wells. Select ‘Injected Gaslift’ from the
spin box at the top of the screen to display how the amount of gas injection to
each well varies with total amount available. You may click on Plot for a
graphical view. Select as variables:
• Gas available: MMscft/d
• Y axis variable: Oil produced

Figure 64: Select variable to plot
Figure 65: Oil produced from each well
Â TIP: You can view and plot allocation results for any node in the
system by entering its data entry screen in the usual manner and
then clicking on the Results button. The first tab displays the
Allocation results. Press Plot to obtain a plot of these results.

TUTORIAL GUIDE
2.4.8 Step 8 – Associated Water / Gas Injection Model
In this section a water injection model will be linked to the previously build
production model. A materal balance prediction will be run on this model.
2.4.8.1 Set the Production Model
The production model used is the gas lifted production model previously
created, to which a reservoir has been added in order to be able to run a
material balance prediction.
The procedure to add a reservoir has been described in the first GAP example
developed in the tutorial.
The tank needs to be described in the production model. Select Browse to
locate the MBAL file (TUTORIAL GAP EXAMPLE_TANK.MBI).
Figure 66: Schematic Diagram of the Production Model
In order to be able to run a material balance prediction, we need to set the
model to be a predictive model. To do so, go to Options | Method and select
On with the scrollbar related to Prediction.

Figure 67: System Option Setting Screen
The relative permeabilities corresponding to the reservoir have been
described in the tank model. Double click on the well, go to Input | IPR | More
and set Prediction Fractional Flow Rel Perm to From Tank Model.

TUTORIAL GUIDE
Figure 68: Schematic Diagram of the Production Model
The production model is then set in order to run a material balance prediction.
Use File | Save As to save the work done this far to a GAP file (TUTORIAL
GAP EXAMPLE.GAP) in a suitable directory.
2.4.8.2 Create the Water Injection Model
The first step is to create an independent GAP model to model the water
injection system.
Go to File | New to create a new GAP file.
In order to set the model to water injection go to Options | Method and select
Water Injection in the system type scroll bar. As this model is going to be
linked with a production model and as a material balance prediction is going
to be run, select On in the prediction scrollbar. Click OK to validate the data.

Figure 69: Water Injection Model Settings
The next step is to implement the elements constituting the model : reservoir,
water injection wells, injection lines and injection manifold.
In the main GAP screen, click on the icon and add a tank.
Using the icon, add a well.
Using the icon, add a water injection manifold (a injection temperature
must be entered).
Using the icon, add two joints in between the water injection manifold
and the well.
Using the icon, link all the elements together.

TUTORIAL GUIDE
The network described on the following figure is obtained.
Figure 70: Schematic Diagram of the Water Injection Network
Use File | Save As to save the work done this far to a GAP file (WaterInj.GAP)
We need now to specify the physical properties of the different elements
constituting the system. The procedure has been explained in detail for each
element when the production network has been created.
The tank needs to be described as the same tank used in the production
model. Select Browse to locate the MBAL file (TUTORIAL GAP
EXAMPLE_TANK.MBI).
The water injection well needs to be created using the same procedure
described for the production wells. IPR and VLPs must be generated as for
any other type of well.
The injection flow line can be characterised by implementing some pipe data :
pipe length, pipe inside diameter, pipe inlet and outlet TVD as normal.
The following network can then be set up :

Figure 71: Schematic Diagram of the Water Injection Network
The red circles around the tank and the well are not present anymore,
confirming the validity of the data input on each element of the system.
Use File | Save As to save the work done this far to a GAP file (WaterInj.GAP)
2.4.8.3 Link the Production and the Injection System
The next step is to link the production model and the water injection model.
To do so, open the production model file in GAP. Go to Options | Method and
tick the box corresponding to Associated Injection Models | Water Injection.
The browsing box will then be available. Browse the water injection model
previously built. The path corresponding to this file will appear.

TUTORIAL GUIDE
Figure 72: Linking Production and Water Injection Models
Click OK. Both the production and injection models are going to appear in the
GAP main window.
One way of visualising both systems in the GAP main window, go to Window |
Tile Vertically.

Figure 73: Schematic Diagram of both Production and Water Injection Network
It is now possible to make modifications on each model using the same GAP
session.
Use File | Save As to save the work done this far to a GAP file. Each model
will be saved separately as shown by the following screen. Click Continue if
you wish to save the production and water injection models in the same
directories chosen previously. If this is not the case, simply alter the file path
name in the saving screen.
Figure 74: Saving Both Production and Water Injection Systems

TUTORIAL GUIDE
2.4.8.4 Running the Material Balance Prediction
A material balance prediction can now be run.
Using the icon, start the material balance prediction process.
The first screen enables the selection of the prediction starting date, ending
date and step size.
Figure 75: Prediction Screen Settings
Several options are available in order to control the water injection (i.e. or gas
injection).
- target pressure input : this will control the water injection so that the
reservoir pressure never goes under the target pressure entered.
- Voidage replacement input : this will control the water injection
taking in account a voidage replacement scheme, defined by a
percentage of voidage replacement entered by the user.
- Water recycling : this option enables to inject a defined percentage
of the produced water
- Fixed Rate : this option enables to inject a defined rate of water.
In order to respect these constraints, the injection well needs to be set as
controllable.
To do so, select Main what will enable you to come back to GAP main window.
Go in the water injection model window, right click on the well and select
Controllable. A red circle will appear around the well which confirms that the
well can be choked back by the software.
Come back into the prediction run by using again the icon : then set a
fixed water injection rate of 3000 STB / d.

Figure 76: Prediction Screen Settings
Click Next to go to the next prediction screen : It summarises the input data
for the tank chosen.
Figure 77: Prediction Screen
Select Next and allocate the amount of gas available for gas lift purposes.

TUTORIAL GUIDE
Figure 78: Gas Lift Gas Allocation
Select Next and allocate a separator pressure of 250 psig.
Figure 79: Separator Pressure Allocation
Select Next and allocate a injection manifold pressure of 2000 psig.

Figure 80: Water Injection Manifold Pressure Allocation
Select Optimise and Honour Constraints and Calculate. This allows the Solve
Network cycle to be performed for each of the 15 time steps requested, while
respecting the constraints implemented.
Figure 81: Material Balance Prediction Calculation Screen

TUTORIAL GUIDE
As soon as the calculation is finished, select Main and return to the main GAP
window.
To inspect the results, double click on the tank and select MBAL Results. This
enables accessing the global prediction results for the tank. To check that the
constraint on the water injection rate as been respected, select Plot | Variables
and choose the variables you want to display on the plot, here Average Water
Injection Rate Vs. Time.
Figure 82: Selection of the Variables displayed on the Plot
Select Done and the plot is displayed. It is then noticeable that the constraints
on the water injection rate set previously as been fulfilled.

Figure 83: Average Water Injection Rate Vs. Time
2.4.8.5 Associated Gas Injection Model
A similar procedure can be followed to set up GAP surface network model
associated with a Gas Injection System, as shown on the following
screenshots.
Figure 84: Associated Gas Injection Model Settings

TUTORIAL GUIDE
Figure 85: Associated Gas Injection Model Example

3Physics Examples
This section contains the following tutorials:-
• PROSPER Gas Lift Example:
This example builds a PROSPER well model including a gas lift system.
It also shows how to design the gas lift system.
• PROSPER ESP Example:
This example builds a PROSPER well model including an ESP and
shows how to design the ESP.
• MBAL Gas History Matching Example:
This example shows how to run the history matching section. It also
includes Fw matching and verification of the water cut using the
production prediction. It is a continuation of the MBAL gas example in
the dexterity section.
• MBAL Oil History Matching Example:
This example builds an MBAL tank model and shows how to perform
the history matching.
C:Program FilesPetroleum ExpertsIPM 5.0Worked examplesPhysics
examples
3.1 Gas Lift Example
This example assumes that the user is already familiar with setting well
models in PROSPER.
examplesGas lift
Objective:
In this model the objective is:
1. Quality check the test / production data that is available. This
quality check is based on what is possible physically.
2. Based on the checked data, we select and build our PVT and flow
models.
3. Design a new gas lift system for this well.
4. Use QuickLook option of PROSPER for performance diagnosis.
Methodology:

TUTORIAL GUIDE
The single well model will be built step by step and at each step any available
test / production data available will be used to validate the model. Also as we
progress through the example, new test data will be added and checked
against data previously entered. In case of conflicts, reasoning on what is
possible physically will be used to RESOLVE this conflict.
Data Available:
PVT Data:
• Temperature = 250.0 deg F
• Bubble Point Pb = 2200.0 psig
• GOR at Pb = 500 scf/stb
• Oil FVF at Pb = 1.32 rb/stb
• Oil viscosity at Pb= 0.4 cp
• Oil gravity = 39.0 API
• Gas gravity = 0.798
• Water Salinity = 100,000 ppm
Gradient Data:
Data Set 1
• Well head pressure = 264.0 psig
• Water cut = 20.3 %
• Liquid rate = 6161.0 stb/day
• GOR = 432 scf/stb
• Gas Lift = 0 MMscf/day
• Injection depth = 13000 ft
• Pressure @ 14800 ft = 3382.0 psig
Data Set 2
• Well head pressure = 264.0 psig
• Water cut = 20.3 %
• Liquid rate = 1100.0 stb/day
• Gas Lift = 1.0 MMscf/day
• Injection depth = 8000.0 ft
• Pressure @ 1500 ft = 500.0 psig
3.1.1 Setting Up the Basic Model
We are going to set up a model with the following options:
Fluid Oil and water
Method Black oil
Separator Single-stage
Emulsions No
Hydrates Disable warning
Water viscosity Use default correlation
Flow type Tubing flow
Well type Producer

Artificial lift method Gas lift
Type No friction loss in annulus
Predict Pressure and temperature (offshore)
Model Rough approximation
Range Full system
Output Show calculating data
Well completion type Cased hole
Gravel pack No
Inflow type Single branch
Gas coning No
Figure 66: Setting up the options
Select the Option menu in PROSPER and select the following options:
Then select | PVT | Input Data and enter the following data:
Solution GOR 500 scf/stb
Oil gravity 39 API
Gas gravity 0.798
(no impurities no gas)

TUTORIAL GUIDE
Figure 67: Entering PVT parameter
Click the Match Data button on the above dialog and enter the PVT match
data that we have:
Temperature 250 degree F
Bubble point 2200 psig
GOR @ bubble point 500 scf/stb
Oil FVF @ bubble point 1.32 rb/stb
Oil viscosity @ bubble point 0.4 cp

Figure 68: Entering PVT lab data
Click Done on the above dialog to go back to the PVT input dialog. Then
perform the match calculation by clicking the Regression button and then the
Match All button.
Figure 69: PVT matching
After finishing the PVT match, click the Parameters button to view the
statistics and select the best correlation for PVT modelling. Based on the
theses regression parameters (parameter 1, which is multiplier and parameter
2 which is a shift factor) and standard deviation, select the best model. Ideally
the std deviation should be very small, parameter 1 should equal 1.0 and
parameter 2 should equal zero.

TUTORIAL GUIDE
Figure 70: Matching parameters
Based on the results, we might want to use the Beggs et al correlation for
viscosity modelling and Glaso for all other properties. Click on | Done | Done
to go back to the main PVT screen. Select the correlations to use in the main
PVT screen.

Figure 71: Select the correlation used
Once this is done, click the Done button to return to the main window.
Now click on the | System | Equipment (Tubing etc.) menu option and input
the equipment data:
Deviation Survey
It is given the deviation survey as follow:
Measured depth (ft) True vertical depth (ft)
0 0
1000 1000
2500 2405
6500 5322
15200 11500

TUTORIAL GUIDE
Figure 72: Deviation survey
Down hole Equipment
It is given the down hole equipment as follow:
Type Measured depth (ft) Internal diameter (in) Roughness (in)
X’mass tree 0
Tubing 14500 3.96 0.0006
Casing 15200 6.00 0.0006

Figure 73: Downhole equipment
Geothermal Gradient
It is given the Geothermal gradient as follow:
Measured depth (ft) Formation temperature
(degree F)
0 50
15200 250
Overall heat transfer coefficient 8 BTU/hr/ft2/F

TUTORIAL GUIDE
Figure 74: Geothermal gradient
Â Note: There is no surface equipment. Hence we can leave the surface
equipment section alone. Also, leave the heat capacities to the default
values. Click on | Done to exit to the main screen.
Next click on System | Inflow Performance and select the IPR model and
enter the basic parameters:
Reservoir model Darcy
Mechanical / Geometrical skin Enter by hand
Reservoir pressure 3844 psig
Reservoir temperature 250 degree F
Water cut 20.3 %
Total GOR 500
Relative permeability No

Figure 75: Select reservoir model
Then enter the IPR data as follow:
Reservoir Permeability 100 md
Reservoir thickness 100 ft
Drainage area 100 acres
Dietz shape factor 31.6
Well bore radius 0.354 ft

TUTORIAL GUIDE
Figure 76: Entering parameters for the reservoir model
Click on the tab labelled ‘Mech/Geom Skin’ and a screen prompting for a skin value
will occur. Enter a skin of 0.

Figure 77: Entering skin
Click on the Calculate button to get the following IPR plot:

TUTORIAL GUIDE
Figure 78: IPR plot
Click the Main menu item on the IPR plot in order to get back to the main
PROSPER window. Next click on the System | Gaslift Data menu item and
enter the gas lift data as follows.
Gaslift gas gravity 0.7
Mole percent H2S 0%
Mole percent H2S 0%
Mole percent H2S 0%
GLR injected 0 scf/stb
Gas lift method Optimum Depth of injection
Maximum Depth of injection 13000 ft
Casing pressure 1900 psig
DP across valve 100 psi
Figure 79: Gaslift specification
We have selected an optimum depth of injection, but want to limit the injection
depth to 13000 feet, which is our packer depth. Also we know that we will
have gas lift gas available at 1900 psig at casing head. Click on Done to
complete this and to go back to the main PROSPER screen.
We should now save the file. For this we click on | File | Save As, and name
the file as GLIFTG.OUT for instance in your working directory.
3.1.2 Matching Test Data and Data Quality Check
Methods
The first thing that we will do is to quality check our data. Let us try to use
data set one as defined at the start of the tutorial. The first step would be to
check on Data Set 1. For this, Select the Matching | Correlation

Comparison | Tubing menu option and enter the following data, selecting
correlations as highlighted:
Well head pressure 264.0 psig
Water cut 20.3 %
Liquid rate 6161.0 stb/day
GOR 432 scf/stb
GOR free 0 scf/stb
Gas Lift gas rate 0 MMscf/day
Injection depth 13000. ft
Pressure @ 14800 ft 3382.0 psig
Correlations Duns and Ros Modified
Hagedorn Brown
Fancher Brown
Petroleum Experts 2
Petroleum Experts 3
Figure 80: Correlation comparison
Then click the Calculate button and the Calculate button again on the next
dialog. Once we perform the calculations and plot the results, we get the
following plot:

TUTORIAL GUIDE
Figure 81: Correlation comparison plot
If we notice the bottom right hand corner of the plot, the test data point lies to
the left of the pressure traverse generated by the Fancher Brown correlation.
But the Fancher Brown correlation is a non-slip correlation, so it predicts least
pressure drops. However the plot indicates that our test point requires lesser
pressure drops than Fancher Brown so there is obviously something wrong.
This means that the PVT model we have and the test data are in conflict. If we
look at the test data itself, we can see that we are reporting a GOR of 432
scf/stb at a reservoir pressure of 3844 psig, whereas the solution GOR is 500
scf/stb and the bubble point is 2200 psig. One of the items of data is incorrect.
However in this case we know the PVT data are correct so the reported GOR
must be wrong. Hence we change the GOR and redo the calculation.

Figure 82: Change GOR
We will see the following results

TUTORIAL GUIDE
Figure 83: Results of Changing GOR
Once this change is made, the test data point does fall on the right of the
Fancher Brown correlation, and we can proceed with the use of this test data.
The next step in building the model will be matching a correlation to the test
data that we have and then use the matched correlation in the analysis. We
will try to use data point two for this purpose – as defined at the start of this
example.
If we use data point two in a similar way to data point one in the Correlation
Comparison dialog, and perform the correlation comparison calculation, we
can check how this test data point compares to the standard correlations. The
plots are:

Figure 84: Results of Data point 2
The test data point lies to the right of the Duns and Ross Modified (DRM)
correlation. Like Fancher Brown (FB), the DRM correlation represents the
other extreme of the pressure drop i.e. maximum pressure loses. Thus if a
point lies to the right of the DRM, we are expecting pressure drops greater
than DRM. The other point to note is that for the same well head pressure and
IPR, with gas lift we are getting lower flow rates than without gas lift as
indicated by data point one. It could be that the data point is wrong or the PVT
data are incorrect. However we already know that our PVT data are correct, so
the data point must be incorrect.
Since we already have another test data point (Data set 1) we will match the
correlations to that data point. The matching process consists in reproducing
the test data point by matching the two components of pressure drop i.e.
gravity and friction by using multipliers (parameter 1 and parameter 2) for
each correlation. The correlation that matches best will selected to model flow
in the tubing. Select the Matching | Matching | IPR/VLP (Quality Check)
and enter test data point 1 in the screen as shown below:
Well head pressure 264.0 psig
Tubing head temperature 132.8 degree F
Water cut 20.3 %
Liquid rate 6161.0 stb/day
GOR 500 scf/stb
GOR free 0 scf/stb
Gas Lift gas rate 0 MMscf/day
Injection depth 13000. ft
Pressure @ 14800 ft 3382.0 psig

TUTORIAL GUIDE
Figure 85: Entering the match data
Then click the Match VLP button and select the following correlations on the
next dialog:
• Hagerdorn Brown
• Petroleum Experts 2
Figure 86: Selecting correlations to match
Then click on Match button to perform the matching calculation. Once we
have performed the match calculations, the new match parameters (seen by
pressing the button Statistics) are:

Figure 87: Matched parameters
We will use Petroleum Experts 2 as the vertical lift correlation.
We have now matched VLP to the test data. We should next look at the IPR.
We must make sure that the IPR can supply the rate that we are getting. We
can first use the tuned VLP correlation to calculate the bottomhole flowing
pressure for the same conditions as the test data (same rate, water cut, GOR,
well head pressure, etc.). Since IPR is a plot of bottom hole flowing pressure
vs. liquid rate, we have a test point on the IPR now, which is the test liquid
rate vs. the calculated bottomhole flowing pressure using the tuned VLP
correlation.
All these can be done in the VLP/IPR matching section. From the correlation
matched parameters screen, click on | Done | Done, you will go back to the
VLP/IPR matching main screen. From there click on the button ‘VLP/IPR’.

TUTORIAL GUIDE
Figure 88: Calculate the BHFP
The purpose of this screen is to calculate the bottomhole flowing pressure for
the test conditions. First we make sure that the right tubing correlation is
selected. In our case, we are going to use Petroleum Experts 2. Then hit on
Calculate to start the calculation. The results of the calculation will be shown
and the calculated bottomhole flowing pressure will be shown.

Figure 89: Calculated BHFP for the test rate
We can now hit on IPR to go to the IPR section.

TUTORIAL GUIDE
Figure 90: IPR section
Hit on Calculate to plot both the tuned VLP and IPR on the same plot and
compare them to the test data.

Figure 91: Comparison of the current IPR model with the test data
The square box is the test point. The VLP and IPR should be intersecting at
that point. The errors are displayed on the right of the screen. We can now
adjust the IPR model to reduce the errors.
There is no fixed method to adjust the IPR. It depends on the conditions. For
instance, if we are uncertain about the reservoir pressure, we can adjust the
reservoir pressure. If we think that the value of skin has changed, we can
adjust the skin value. In this exercise, we are going to change the reservoir
pressure.
Click on Finish to close the plot window. Change the reservoir pressure in the
IPR main screen to 3876 psig and hit on | Calculate again. We will see that
the error has been reduced to a very small value and we have matched the
IPR.
This finishes our matching of test data and data quality section. Go back to
the main screen, and save the file as GliftG1.out.
3.1.3 Designing a New Gas Lift Well
Note: we will design for water cut of 50%. The gas available is 6 MMscf/day
@ 1900 psig injection pressure at the top node.

TUTORIAL GUIDE
Select the Design | Gas Lift | New Well menu item. Supply the following
input data. We are asking for the gas lift valves to be casing sensitive. We
also have selected the valves to be designed in such a way that they open at
casing pressure:
Design rate method Calculate from max production
Design rate 20000 stb/day
Maximum gas available 6 MMscf/day
Maximum gas during unloading 6 MMscf/day
Flowing top node pressure 250 psig
Unloading top node pressure 250 psig
Operating injection pressure 1900 psig
Kick off injection pressure 1900 psig
Desired dP across valve 200 psi
Maximum depth of injection 13500 ft
Water cut 50%
Minimum spacing 500 ft
Static gradient of load fluid 0.45 psi/ft
Minimum transfer dP 25%
Maximum port size 32/64 ths inch
Safety for closure of last unloading valve 0 psi
Valve type Casing sensitive
Min CHP decrease per valve 20 psi
Valve settings All valves Pvo = gas
pressure
Dome pressure correlation above 1200 psig Yes
Check rate conformance with IPR Yes
Vertical lift correlation Petroleum Experts 2
Surface pipe correlation Dukler Flannigan
Use IPR for unloading No
Orifice sizing on Calculated dP at orifice

Figure 92: Gas lift design
Once the valve type has been selected, press Continue. Then generate the
gas lift performance curve by clicking the Get Rate button and then the Plot
button at the top of the screen. This generated performance curve is as
shown:

TUTORIAL GUIDE
Figure 93: Gas lift performance curve
The performance curve of a gas lift design plots the oil rate produced with
increased gas injection rates. To understand the shape of this curve, we will
appeal to the notion that we do gas lift to decrease the pressure loss in the
tubing string by decreasing the gravity component of pressure drop. The
greater is the amount of gas injected; the lighter will be the fluid column.
However as the amount of gas injected increases, we also will be increasing
the other component of pressure drop, i.e. the friction. A stage reaches in
injection when any further increase in gas injection will increase friction
component more than it will decrease gravity component. After this stage any
increase in gas injection will decrease production rates. Thus the PC curve
will go up and then come down as shown above and will have a maximum oil
production rate and the gas injection required corresponding to this rate will
be optimum. If we look at the performance curve we see that at a gas lift rate
of 6 MMscf/day the oil production is around 4440 stb/day. The maximum oil
production of 4600 stb/day occurs for gas lift rate of approximately 9.3
MMscf/day. From this plot PROSPER determines the gas lift required for
maximum oil production. This is the optimum gas lift rate for this well. In case
the available gas is higher than the optimum gas required, the program will
only inject the optimum gas into the well, which is 9.3 MMscf/day in this case.
In case the available gas is less than optimum gas, the actual available gas
value will be used. If we proceed with design at this stage by pressing the

Design on the following screen, PROSPER will design this case using 6.0
MMscf/day of gas.
Figure 94: Gas lift design
To see a plot of the Gas lift design click the Plot button at the bottom of the
above screen and the following plot will appear.

TUTORIAL GUIDE
Figure 95: Gas lift design plot
If we press Finish we will exit the plot and return to design screen. On this
screen if we press the Results button, the following screen with the results of
gas lift design will appear. On the screen given below if you press Calculate,
PROSPER will calculate the dome pressure settings for you.
Figure 96: Gas lift design results
This finishes the gas lift design. Go back to the main screen, and save the file
as GliftG2.out.

3.1.4 Using QuickLook as a Diagnostic Option to
Check the Gas Lift Design
Â Note: In this section we are going to use the QuickLook option as a
diagnostic tool and see if the existing gas lift set up is performing as
per design. We will use the design of the previous section; selected on
the basis of a 50% future water cut and try to do the study for 20.3 %
current water cut.
We will start from the Matching – QuickLook menu option. To start with let
us say we have current flowing conditions as per data set one defined at the
start of the tutorial.
To start with let us say we have the following oil rate, water cut, pressure, lift
gas rate and temperature data for the well:
Tubing head pressure 264 psig
Tubing head temperature 160.7 degree F
Liquid rate 6161 stb/day
Water cut 20.3 %
Total gas rate 6.555 MMscf/day
Gas injection rate 4.1 MMscf/day
Casing head pressure 1750 psig
Figure 97: QuickLook entry screen

TUTORIAL GUIDE
To enter the valve data, press the Valves button on the above screen. The
following screen appears:
Figure 98: Valve depth specification
We can transfer the valve data from the design we have just done by pressing
Transfer on the above screen and pressing Gas Lift Design on the screen
below. Note that this is the gas lift design got for the case when we fixed oil
production to 4000 stb/day in the previous section.
Figure 99: Transfer valve data
Next click on Done | Calculate | Calculate | OK | Plot to get the following
diagnostic plot:

Figure 100: QuickLook calculation plot
To analyze this plot, let us examine the QuickLook principle. In this method
we calculate well pressure traverses for both tubing and casing pressure in
two directions: one beginning from the wellhead and going to the sand face,
and the other going from the sand-face up to the wellhead.
To change the assumptions, we must understand the factors that affect these
traverses.
The downward gradients are based on measured data (THP, CHP, gas and
liquid flow rates, WC, GOR), while the upward gradients depend on the inflow
(in the case of the tubing pressure) and on the pressure drop across the
orifice (as regards the casing pressure).
If our assumptions about the gas lift rates, oil flows etc., are correct, the two
pressure traverses should be identical. If not we have to change these
assumptions until we get identical traverses.
In Figure 100 we see that the tubing traverse calculated starting from the
flowing bottomhole pressure is higher than the measured tubing traverse. This
suggests that the inflow potential is too high, so the reservoir pressure should
be lower than considered. Let’s decrease the reservoir pressure down to 3050
psi.
The QuickLook calculated is now:

TUTORIAL GUIDE
Figure 101: QuickLook calculation plot 2
The tubing curves now overlap.
Next let us see how pressure traverse curves compare in the casing above
the orifice.
The calculated upward casing traverse is now smaller than the measured one.
This suggests that the pressure drop across the orifice for some reason (like
scaling) has increased. So, in order to match the two gradients, a smaller
orifice diameter can be chosen. Let us decrease it to 22/64” and re-perform
QuickLook calculations, the plot then looks like below:

Figure 102: QuickLook calculation plot 3
These results are good and based on this we can predict the oil flow rates and
gas injection rates fairly. This method also can be used to trouble shoot and
check the performance of operating wells, if we have down-hole and reliable
flow measurements available.
Save the file as GliftG3.out.
3.2 PROSPER ESP Example
This example presumes that the user is already familiar with setting up well
models in PROSPER.
examplesESP
Objective:
In this model the objective is:
• Quality check the test / production data that is available. This quality
check is based on what it possible physically.
• Based on the checked data select build our PVT and flow models.
• Design an ESP system for this well.
Methodology:

TUTORIAL GUIDE
The single well model will be built step by step and at each step any test /
production data available will be used to validate the model. Also as we
progress through the example, new test data will be added and checked
against data previously entered. In case of conflicts, reasoning on what is
possible physically will be used to RESOLVE this conflict.
Data Available:
PVT Data:
• Solution GOR = 392.0 scf/stb
• Oil Gravity = 37.66 API
• Gas Gravity = 1.045
• Water Salinity = 94333.9 ppm
• Temperature = 205 deg F
• Bubble Point Pb = 1361.0 psig
Pressure
Psig
GOR
scf/stb
Oil FVF
rb/stb
Oil Viscosity
cp
1361.0 392 1.289
3215 392 1.25 0.66
Gradient Data:
Data Set 1
• Well head pressure = 334 psig
• Water Cut = 6 %
• Liq. Rate = 5200 stb/day
• GOR free = 0 scf/stb
• Pressure @ 7677.2 ft = 2329.0 psig
3.2.1 Setting up the Basic Model
Run PROSPER and go to the Option menu in PROSPER and select the following
options:

Figure 103: Setting the model option
In this screen we have specified:
• The handled fluid is an oil
• PVT behaviour will be modelled as a Black oil, with a single stage
separation scheme.
• This is an offshore well
• We want to do temperature predictions using rough approximation
method.
• The fluid flows through the tubing
• No emulsion forms
• For the moment we have no ESP in the system; the reason will be
given further on.
• It is a cased hole with no gravel pack.
• We also do not have gas coning and the inflow is simple and not
multilateral.
Next in the main screen go to the PVT | Input Data and fill in the PVT data as
indicated in the available data section.

TUTORIAL GUIDE
Figure 104: PVT input data
Since we do have lab data, we should match them to the existing Black oil
Correlation. Enter the PVT match data that we have by clicking the Match
Data button on the above screen. Enter the match data as follows:
Figure 105: PVT match data
Once you have entered measured data, go back to the previous dialog by
clicking the Done button and perform the matching calculation by clicking on

the Regression button and then the Match All button. The program does a
regression analysis on all the entered data with all standard black oil
correlations that are available in PROSPER.
To display the regression parameters and standard deviations for all the
correlations, click on Parameters.
Figure 107: Match parameters

TUTORIAL GUIDE
Based on the these regression parameters (parameter 1, which is multiplier
and parameter 2 which is a shift factor) and standard deviation select the best
model. Ideally the std deviation should be very small, parameter 1 should
equal 1.0 and parameter 2 should equal zero. In this case we select Beggs et
al correlation for viscosity modelling and Standing for all other proprties. Once
this is done, click the Main button to go back to the main window. Please set
the correlations in the main PVT screen (see below)
Figure 108: Choose the correlation
After finishing the PVT match i.e. providing the model with an adequate fluid
description, we have to specify our well bore. To do so in the main screen,
select the System | Equipment (Tubing etc.) menu item and input the
equipment data as follows:
Deviation Survey:
It is given the deviation survey as follow:
Measured depth (ft) True vertical depth (ft)
0 0
463.3 463.3
2399.9 2368.4
3450.1 3256.6
4649.9 4100.1
5200.1 4467.5
6899.9 5673.9
7450.1 6079.7
8687.7 7280.2

Figure 109: Deviation survey
This survey is taken as the basis for calculating true depths, in the model. It is
recommended to use zero of this survey as the reference depth for all further
entries. If we use the zero of deviation survey as the reference depth, then it
is easier to enter well bore data as all the depths indicated in the well bore
data refer to deviation survey. Once we have supplied the deviation survey,
the well bore details are entered as follows:
Down-hole equipment:
It is given the down hole equipment as follow:
Type MD (ft) Tubing ID (in) Tubing OD (in) Casing ID (in)
X’mas tree 59.4 - - -
Tubing 689.0 3.96 4.5 8.68
SSSV - 2.13 - -
Tubing 7660.8 3.96 4.5 8.68
Restriction - 2.31 - -
Tubing 7677.2 3.96 4.5 8.68
Casing 7860.9 - - 8.68

TUTORIAL GUIDE
Casing 8169.3 - - 6.18
Casing 8687.7 - - 3.96
All roughness of tubing / casing = 0.0006 in
Figure 110: Downhole equipment
The data regarding tubing outside diameter will be used further on, when the
ESP option will be selected.
Next we want to specify the geothermal gradient and the overall heat transfer
coefficient, as we are doing temperature predictions as well.
Geothermal Gradient:
It is given the Geothermal gradient as follow:
Measured depth (ft) Formation temperature
(degree F)
59.4 60
8687.7 205
Overall heat transfer coefficient 3 BTU/hr/ft2/F

Figure 111: Geothermal gradient
Note: There is no surface equipment. Leave the heat capacities to default
values.
After specifying the well bore we will want to build an inflow model into the
well bore. So, next select the System | Inflow Performance menu item and
select the IPR model and enter the IPR data as shown in the following
dialogs:
Reservoir model PI Entry
Reservoir pressure 2468 psig
Reservoir temperature 205 degree F
Water cut 6 %
Total GOR 392
Relative permeability No

TUTORIAL GUIDE
Figure 112: Selecting the IPR model
We have selected the simple PI model for the Inflow Performance. For this
model to supply input data if we click on the Input Data button of this screen
we get the following screen and we can enter a PI of 7.19 STB/day/psi.

Figure 113: Entering PI value
Click the Calculate button – this should plot the following IPR:

TUTORIAL GUIDE
Figure 114: IPR plot
3.2.2 Matching Test Data and Data Quality Check
Methods
The first thing that we will do is to try to quality check our data. Let us try to
use data set 1 defined at the start of the tutorial.
At this point let us recall that for ESP models for standard traverse
calculations, PROSPER takes the sand face as the first node for calculations.
But in our test data we have a wellhead pressure as reference and do not
know our bottom hole pressure. The other fact is that our test point is above
the ESP, and for flow correlations, in the tubing above ESP the flow is like
natural flow but with a higher bottom hole pressure. This is why, when building
the basic model, we did not select the ESP pump option.
After matching we can revert to ESP option.
Always considering no artificial lift (by selecting the None option for the
Artificial Lift Method in the Options dialog), the first step would be to check
how this test point compares to the gradient plot. Select the Matching |
Correlation Comparison | Tubing menu item and supply the following data,
selecting correlations as highlighted.

Figure 115: Correlation comparison
Once we perform calculations (by clicking the Calculate button and then the
Calculate button again on the next dialog) and plot the results (by clicking the
Plot button), we get the following plot:

TUTORIAL GUIDE
Figure 116: Gradient plot
Figure 116 shows that in the region at the bottom right of the plot, the test
data point lies on the right of the Pressure traverse generated by the Duns
and Ros Modified (DRM) correlation.
Now DRM correlation gives maximum pressure drop as already discussed in
the previous gas lift example. But since the plot indicates that our test point
requires higher pressure drop than DRM, there is something wrong. This
means that the PVT model we have and the test data are in conflict. One of
the two sets of information is incorrect. However, in this case let us say we
know that our PVT data is correct so there are inconsistencies in the test point.
The first thing that we can do is go back to the source of the test data and
check again for the numbers. In this case we are reporting a water cut of 6%
which is quite low for normal cases where we are thinking of going to artificial
lift. As a matter of fact the water cut in this example is 34%. If we make this
change to the input data of the Tubing Correlation Comparison and repeat the
correlation comparison we get the following plot:

Figure 117: Gradient plot 2
Once this change is made the test data point does fall on the left of DRM
correlation, and we can proceed with the use of this test data point.
The next step in building the model will be matching a correlation to the test
data that we have. The matching process consists in reproducing the test data
point by matching the two components of pressure drop i.e. gravity and
friction by using multipliers (parameter 1 and parameter 2) for each
correlation. The correlation that matches best will be selected to model flow in
the tubing. Select the menu item Matching | Matching | IPR/VLP (Quality
Check) and enter test data 1 in the screen as shown below (note that we will
now use 34% for the water cut rather than the 6% in the original data):

TUTORIAL GUIDE
Figure 118: VLP/IPR matching data
Then perform matching by clicking the Match button. Select the following
correlations from the list:
• Hagedorn Brown
Then click the Match button again to calculate the match parameters. Once
we have performed the match calculations, the new match parameters (seen
by pressing the Statistics button) are:

Figure 119: VLP matched parameters
We are using Petroleum Experts 2 as the vertical lift correlation – note that we will
make the selection later in the tutorial so make a note of which one to use.
This finishes our matching test data and data quality check section. At the end
of this save the file as espg.out. At this stage remember to change the lift
method to ESP in the option screen.
3.2.3 Designing an ESP for this Problem
Â Note: We will design for a water cut of 60% and a delivering a design
rate of 9000 STB/day against a wellhead pressure of 100 psig. We
want the pump placed at a depth of 7660 feet. A cable roughly around
7710 feet will be needed to go up to the pump.
Select on the main screen menu Design | ESP | Design.
After entering the tubing outside diameters in the Downhole Equipment
screen, supply the following input data. We will start with the assumption that
no gas separation is needed at the pump inlet.
Pump depth 7660 ft
Operating frequency 60 Hz
Maximum OD 6 in

TUTORIAL GUIDE
Length of cable 7710 ft
Gas separator efficiency 0 %
Design rate 9000 stb/day
Water cut 60 %
Top node pressure 100 psig
Motor power safety margin 0 %
Pump wear factor 0
Pipe correlation Beggs and Brill
Tubing correlation Petroleum Experts 2
Figure 120: ESP design parameter
Press the Calculate button on the above dialog and the Calculate button
again on the next dialog and PROSPER will calculate the pump head, power etc
as shown below:

Figure 121: ESP design calculation
Once pump calculations are finished, check the validity of the assumption of
no gas separation at pump inlet by using the Dunbar plot. The plot is
activated by pressing the Sensitivity button and the following plot appears:
Figure 122: Sensitivity plot
The different lines on the Dunbar plot are for different levels of gas separation
efficiency at pump intake. The separation efficiency assumed is okay for this
criterion if it the point of operation falls above the Dunbar line. In this case, the
pump operating point with zero separation at inlet (shown by the square

TUTORIAL GUIDE
symbol) falls above the Dunbar line, which implies that we do not need a gas
separator at the pump inlet. In case it was otherwise, we should use
separation efficiency in the ESP Design input dialog above and repeat the
pump calculations until we get the pump intake point above the Dunbar line.
After checking for gas separation requirements, we proceed with the design
by pressing the Finish button on the plot, the Done button followed by the
Design button and we should be in the following dialog:
The design at this stage consists in determining an adequate pump, motor
and cable that can handle this load.
Figure 123: Selecting the suitable pump, motor, and cable
Note that you should have built your pump, motor and cable database
by now. If you have not, go to Design | ESP | Pump database and you will
get the following screen:

Figure 124: Editing the pump database
On this screen use Import | Append to import a pump database. There are
some databases provided with the program in the samplesPROSPER
directory. You can load the motor and cable databases in a similar fashion.
The design screen will select from the database, the equipment that can do
the job. We will select one combination out of these. In the first selection we
select the pump, then the motor and finally the cable.
Figure 125: Selecting pump

TUTORIAL GUIDE
From the available pumps, let us select the REDA SN8500 model. Next we
will select a motor for this pump as shown below:
Figure 126: Selecting motor
From the motor selection let us select the Reda 540_90-0_Int 400HP 2116V
113A motor. Next we will select a cable.
Figure 127: Selecting cable

Based on the selection available we select #1 Copper as our cable. This
stage completes the ESP design and the results are displayed in the same
screen in terms of current required etc as shown below:
Figure 128: ESP design details
We have selected a REDA SN8500 pump with 137 stages to do the job. If we
click on the Plot button it display the pump performance curve also indicating
the limits of operation of this pump.

TUTORIAL GUIDE
Figure 129: ESP pump plot
The point on this plot shows the design operating point on the pump
performance plot. This finishes a new ESP design.
3.2.4 Using ESP QuickLook as a Diagnostic Option to
check an Existing ESP Design
Â Note: In this section we are going to use the QuickLook option as a
diagnostic tool. We will see how we can monitor the performance of
an ESP for an intermediate water cut that has been designed on the
basis of 60% future water. What we have is reliable down-hole
measured data for pump intake and discharge pressure in this case. A
liquid rate of 6523 STB/day with a well head pressure of 345 psig has
been observed for this well.
We will start by selecting the menu option Matching | QuickLook. The
measurements indicate a water cut of 60%. The pump is same as designed
in the previous section. We will assume that the pump runs at 60 Hz still has
no wear factor. On the basis of these measurements we can supply the data
as shown below:
Tubing head pressure 345 psig
Liquid rate 6523 stb/day
Water cut 60 %

Produced GOR 392 scf/stb
Static bottom hole pressure 2468 psig
Pump depth 7660 ft
Operating frequency 60 Hz
Length of cable 7710 ft
Gas separation efficiency 0 %
Number of stages 137
Pump wear factor 0 (fraction)
Downhole data:
Pump discharge pressure (MD = 7660 ft) = 2725 psig
Pump suction pressure (MD = 7660 ft) = 1025 psig
Figure 130: ESP QuickLook
Once we have given the measured pressure data and the flow rates, through
the pump, to start QuickLook press the Calculate button, the Calculate
button again and then the Plot button. You should get the following plot:

TUTORIAL GUIDE
Figure 131: ESP QuickLook plot
To analyse this plot, let us examine the QuickLook principle. In this method
we calculate well pressure traverses in two directions, one beginning from the
well head and going to the sand-face, and the other going from the sand face
up all the way to the wellhead. In case our assumptions about the pump depth
and oil flows etc. are correct, the two pressure traverses should be identical
and overlap. If not we have to change these assumptions until we get identical
traverses.
To start the process of diagnosis, we can think of the pump as a tie point for
the system, where the inflow up to the pump and the lift above the pump are
tied with each other. For a given wellhead pressure, the pump discharge
pressure depends only on the weight and frictional loss of the fluid above the
pump. In our case we see that, for the measured well head pressure, the
pump discharge pressure we got is slightly lower than that measured as
indicated by the circle area on the above plot.
The section of the well can be considered as a naturally flowing well with
bottomhole pressure equal to the pump discharge pressure.
So, in order to match the downward discharge pressure point with the
measured one, we can go back to the simple well configuration (choosing No
Artificial Lift from the Option screen) and tune the flow correlations using the
measured data at the pump discharge as test data:
Tubing Head Pressure: 345 psig
Tubing Head Temperature: 174 deg F
Water Cut: 60 %

Liquid Rate: 6523 STB/day
Gauge Depth: 7660 ft
Gauge Pressure: 2725 psig
GOR: 392 scf/STB
GOR free: 0 scf/STB
Enter these data in the VLP/IPR Matching screen:
Figure 132: VLP/IPR Matching screen
Then perform the Correlation Comparison to quality check the test data. The
test data point is within the limits given by the Duns and Ros Modified and
Fancher and Brown correlations, as shown in the following plot:

TUTORIAL GUIDE
Figure 133: Correlation Comparison
The choice of the correlation is done following the guidelines given by the
PROSPER manual.
Now, back in the VLP/IPR screen, we click on Match, choose the Petroleum
Experts 2 correlation and click on Match again.
Figure 134: VLP/IPR Matching
The match parameters are calculated. Clicking on Statistics, the match
parameters are displayed:

Figure 135: Match parameters
At this point, it is possible to re-select the ESP Artificial Lift Method from the
Option menu and go to the QuickLook section.
Click on Calculate⎮Calculate, then Plot. The following QuickLook plot is
displayed:

TUTORIAL GUIDE
As it is shown, now the discharge pressure point on the measured downward
gradient is matched.
Now the pump inlet pressure must be matched. Figure 136 shows that the dP
across the pump calculated in the downward gradient is bigger than the
measure one (given by the distance between the two blue squares). The
reason of this could be, for example, the pump wear, which decreases the
pump performance.
So, let’s choose a pump wear factor of 0.18 to be entered in the QuickLook
main screen:

Figure 137: ESP QuickLook main screen
Then Calculate⎮Calculate and Plot. The following QuickLook plot is
displayed:

TUTORIAL GUIDE
The downward gradient is now matched.
All what we have to do now is to match the upward gradient to this. The
upward gradient is calculated starting from the bottomhole pressure given by
the IPR. So, it depends on the inflow.
In order to match the upward calculated gradient, let us go to the IPR section
and decrease the PI to 6.3 STB/day/psi. Then, back in the QuickLook section,
calculate and display again the QuickLook plot:

This match is acceptable. This method can thus be used to trouble shoot and
check the performance of operating wells, if we have down-hole and reliable
flow measurements available.
Save this file as espg1.out.
3.3 MBAL Gas History Matching Example
This tutorial example is designed to provide a continuation of the step by step
introduction to the MBAL program, following on from the MBAL Gas Reservoir
Example. The emphasis here is on the steps required to match a production
history using a material balance model to estimate the original gas in place
within a dry gas reservoir.
The driving mechanisms within the reservoir will be identified, increasing
understanding of the reservoir’s potential production. Relative permeability’s
for gas and water will be estimated by matching fractional water production to
simulated water saturation and tested by performing a prediction calculation.
examplesGas history matching

TUTORIAL GUIDE
3.3.1 STEP 1 : Initialise MBAL
Start the MBAL program by running MBAL.EXE, which can be found in the
PETEX directory (default C:Program FilesPetroleum ExpertsIPM 5.0). See the
MBAL manual for more details on how to start MBAL.
Check that the current version of MBAL has been loaded. Select Help |
About MBAL Package to check the version number.
Select File | Open to open the file created from the MBAL Gas Reservoir
Example in the Dexterity section - GASRES.MBI. Immediately save this as a
new file (GASRES2.MBI) using File | Save As.
3.3.2 STEP 2 : Add Production History Data
Production history data is entered and an aquifer model is initialised in this
section. Enter the production history shown below in Table 1 by selecting
Input | Tank Data and selecting the Production History tab. This data is
also contained in an EXCEL spreadsheet named GASRES2.XLS and the data
(select cells A5:F25) may be copied and pasted into MBAL using a right
mouse click to select Copy within EXCEL and then Paste in MBAL.
Alternatively the standard shortcuts Ctrl C and Ctrl V may be used to copy
and paste data.
After evaluating the possibility of the existence of an aquifer, the following
data can be used as a starting point:
Model Hurst-van Everdingen Modified
System Radial Aquifer
Reservoir thickness 100 ft
Reservoir radius 5000 ft
Outer / Inner radius ratio 5
Encroachment angle 360 degree
Aquifer permeability 20 md
3.3.3 STEP 3 : Material Balance Introduction
A very brief introduction to the material balance method is included here.
The governing principal is mass conservation as the reservoir is produced.
This may be restated as:
The volume of material removed by production at reservoir conditions is
replaced by fluid/formation expansion and possibly aquifer inflow
The equation below represents this volume (material) balance at reservoir
conditions (pressure and temperature):
F = N.Et + We

F is the produced fluid volume at reservoir conditions. Good production
history and PVT is required to estimate this quantity.
N is the original oil/gas volume in place, which can be estimated by
geological investigations.
Et is the expansion of the reservoir fluid and water, and formation rock
compaction following the depressurisation of the reservoir as it is
produced. Good reservoir pressure history and PVT is required to
estimate Et. Note that for oils, good PVT (Bo) above the bubble point
is especially important, since the compressibility of undersaturated
liquid oils is relatively small.
We is the volume of aquifer water entering the initial reservoir volume.
In general, if good PVT and production history is available, F and Et are
reasonably well known. Also, an initial estimate of N can usually be made.
The question then is to refine correlations for We and the value of N to match
the production data. The material balance equation above can be rearranged
to perform non-linear regression on N and aquifer model parameters.
The quality of the PVT and production history data is vital to the material
balance calculations. However for simplicity, this example uses an
unmatched Black Oil PVT and a fictitious production history.
3.3.4 STEP 4 : Material Balance Matching
The production history data will be matched to a material balance model using
judgment and non-linear regression. The drive mechanisms within the
reservoir and the Original Gas In Place (OGIP) will be estimated.
The History Matching starts considering no aquifer is present. So, the Tank
Input data screen, we choose No Aquifer.
It is suggested that the MBAL window is made full screen to aid viewing.
Select History Matching | All. Three screens appear graphically illustrating
the fit of the material balance model to the production data and reservoir
pressure.

TUTORIAL GUIDE
Figure 140: History matching plots
The material balance model is defined by the correlations and parameters
entered in the Input | Tank Data screens. All of these may be altered at any
time to improve the fit, but only the OGIP and the aquifer model parameters
may be modified by non-linear regression. This reflects the observation that
these are generally the least well known variables.
It is very important that a systematic methodology is followed, based on an
understanding of the material balance model, rather than a series of
regressions. It should be understood that the regression solutions are not
necessarily unique and work better if their values prior to regression are not
too far from a solution. Therefore the interpretation of the graphical
representations of the material balance model must be used to refine the
model before regression should be used.
An aquifer model is present, but is providing no water (We = 0). Highlight the
Graphical Method window by clicking the left mouse button within its title bar,
and select Method | Cole ((F-We)/Et). Recall the material balance equation
(F-We)/Et = N (F/Et = N, when We = 0). If the material balance model was
well fitted to the production data then the Cole plot should be a horizontal line
with a value equal to N (OGIP).

Figure 141: Graphical Method showing Cole Plot
The initial rise indicates an increasing apparent value for N. The expansion
(Et) of the reservoir fluids/formation alone is not sufficient to maintain the
reservoir pressure. There is more energy in the reservoir than the current
material balance calculation is predicting. These equivalent statements imply
the reservoir pressure is being maintained by another mechanism, the likely
candidate is an aquifer.
An aquifer model can be added to the model, based on the initial estimates
given in the introduction:

TUTORIAL GUIDE
Figure 142: Initializing an aquifer model
Also note that the rise of the Cole curve does not continue, but levels out and
then starts to fall at later times. This can be interpreted as the aquifer inflow
slowing at around data point 5 and stopping near data point 10. This means
that around this time the outer boundary of the aquifer has been ‘felt’.
At early times, the well will not ‘see’ the aquifer. Therefore the early values on
the Cole curve indicate minimum values for N. Select Display | Scales and
set the Y-axis bottom value to 600 Bscf and select Done. An extrapolation of
the Cole curve towards the Y axis is difficult, emphasising the importance of
early data recording. However, an extrapolation to the Y-axis would suggest a
value for N larger than 600 Bscf, perhaps nearer 750 Bscf.
Next, the value for N shall be updated to 750 Bscf and the aquifer model will
be started. Highlight the Analytical Method window and select Regression
from the menu toolbar.

Figure 143: Modified OGIP and Outer/Inner Radius
Set the Gas in Place to 750 Bscf and the Outer/Inner Radius to 5 and select
Done to display the recalculated material balance.

TUTORIAL GUIDE
From the Graphical Method screen (top right of Figure 144) it is clear that the
aquifer that has been added is too strong, it is providing too much energy to
the system. The extrapolation of the Cole curve to the Y-axis suggests a
possibly reduced OGIP (N).
Look at the WD function Plot screen (top left of Figure 144). This shows a
dimensionless time (tD) and dimensionless aquifer inflow volume (Q). The
‘elbow’ of this curve occurs at the point where the aquifer cannot supply
additional water; the boundary of the aquifer has been ‘felt’. Recall that the
initial Cole curve suggested this occurred between data points 5 and 10.
Move the cursor within the WD function Plot screen and double click left. This
alters the Outer/Inner Radius parameter of the aquifer model, altering the
displays in the other Method screens. In particular, notice that the Analytical
Method gas production/pressure curve moves. By double left clicking in the
WD function plot screen, try to select an aquifer Outer/Inner Radius parameter
that shows a reasonable fit to the production displayed in the Analytical
Method screen. An Outer/Inner Radius of about 2.1 works quite well, but the
‘elbow’ on the WD function Plot is not between data points 5 and 10.
Highlight the Analytical Method window and select Regression from the
toolbar. Set the Outer/Inner Radius to 2.1 in the start (left) column and select
Done to view the results.
Figure 145: Modified Outer/Inner Radius

The data points at very early times may only be reflecting responses from
regions in the vicinity of the well and don’t necessarily show responses of the
entire reservoir, therefore the material balance would not be expected to show
the complete OGIP until the pressure signal from the producing well has had
time to permeate the entire reservoir. It is possible that the Graphical Method
screen is showing this effect at early times. The signal time to permeate the
reservoir can be estimated from the diffusivity and reservoir dimensions. The
diffusivity, D=k/ϕµc (ft2s-1) relates the radial pressure response at a distance
r and time t from the well source by the equation P∝exp(-r2/4Dt).
For this example the first data point shown (point 2) is one year after the start
of production and can probably be expected to reflect the whole reservoir’s
response, suggesting that the aquifer model still requires some fine tuning.
Note also that the ‘elbow’ of the WD Function Plot is not reflecting correctly
the time at which the aquifer energy is exhausted.
However, recognising the points noted above, the material balance model is
now not too far from being consistent with the production data and non-linear
regression may be used to refine the model parameters.
Highlight the Analytical Method window and click on Regression. Check the
Gas in Place, Outer/Inner Radius, Encroachment Angle and Aquifer
Permeability boxes to regress on.

TUTORIAL GUIDE
Figure 147: Select variables to regress on
Select Calc to start the regression. When it finishes, copy the ‘Best Fit’ values
to the ‘Start’ values by clicking the left pointing arrows (Figure 147). You may
select the left pointing arrow between the ‘Start’ and ‘Best Fit’ headings to
copy all of the regressed values. Select Done to view the changes.
Do not regress on combinations of parameters that are simply multiplied by
each other in the aquifer model. For example, the Hurst-van Everdingen
aquifer constant contains the product of porosity, reservoir thickness,
encroachment angle and the square of the original reservoir radius. If a
regression is performed on pairs of these parameters, then the regression will
not converge easily, particularly if the initial values are not close to a solution.

The match now is very good (Figure 148), the aquifer model and OGIP are
consistent with the production history. Note however, that although the actual
values calculated for the aquifer model taken together describe the aquifer
well, the individual parameter values do not in themselves necessarily
correspond to reality. These parameters are not a unique set that characterise
the aquifer.
Select Finish | File | Save.
3.3.5 STEP 5 : Relative Permeability Matching
In this section, the effective relative permeability of water will be obtained by
matching the fractional water flow obtained from the production history to the
fluid saturations calculated by a material balance simulation.
Relative permeability’s were not used during the material balance matching
and are not used during the simulation calculation, since the produced water
and gas are input as part of the production history. The simulation is merely
providing the water saturation within the tank model, resulting from a material
balance simulation.
Select History Matching | Run Simulation | Calc to run a material balance
simulation of the production history. Select OK when the calculation has
completed.

TUTORIAL GUIDE
Figure 149: Simulation results
Return to the main MBAL display by clicking Done. Next, select History
Matching | Fw Matching to display the matching screen.

Figure 150: Fw matching screen
Within this screen, the fractional water flow is plotted as a function of water
saturation. The water breakthrough point can be set by a double left click at
an appropriate saturation (a dashed green line is shown at the new
breakthrough saturation). Note that a breakthrough point below the connate
water saturation (indicated by a grey line) is not possible. Additionally
parameter values can be entered by selecting Parameters. Leave the water
breakthrough saturation at the connate water saturation (0.2).
A region of the display can be enlarged by holding down the left mouse button
and dragging it to select the desired region. The original display can be
redrawn by selecting Redraw. Production history data points may be selected
by holding down the right mouse button and dragging it to select the desired
points. The weighting of the selected points may be altered, or excluded from
use in the regression.
Select Regress, then Parameters to display the matched parameters (Figure
151). Select Finish | Yes to save the matched Corey coefficients.
Figure 151: Matched Fw Parameters
It is now desirable to perform a material balance predition to check that the
fraction flow of water is sufficiently well characterised by the matched relative
permeability model. A prediction calculation in its simplest form requires a
history of oil or gas production rates (copied from the simulation calculation),
from which everything else is calculated. Of particular interest will be the
predicted WGR (Water Gas Ratio).
Select History Matching | Run Simulation | Report and check report to
Clipboard with the ‘Tab delimited’ format.

TUTORIAL GUIDE
Figure 152: Reporting data
Select Layout | Hide All and then highlight Time and Average Gas Rate.
Figure 153: Selecting items to report
Select Done | Report to save the results to the clipboard. Select Done to
return to the main MBAL window and open EXCEL and paste the the contents
of the clipboard into it. The simulation average gas rates are recorded at the
end of each time step.
From within MBAL select Production Prediction | Prediction Setup and set
the prediction method as shown in Figure 154 and select Done. Ensure that

Use Relative Permeabilities is checked and the Prediction End is set to End
of Production History.
Figure 154: Prediction Setup Screen
Select Production Prediction | Production and Constraints and copy the
production gas history (by selecting Copy) into the production constraint
screen (Figure 155) and select Done.

TUTORIAL GUIDE
Figure 155: Production Prediction Screen
Select Production Prediction | Reporting Schedule. Here, we need to
specify how frequent do we need MBAL to report the results. We can set to
automatic. Hence, click on Done to accept automatic reporting. Then, hit on
Prediction | Run Prediction | Calc and OK | Plot when the calculation has
completed. Select Variables and highlight streams Simulation and
Prediction, and plot Pressure, then select Done to view the plot.

Figure 156: Comparing simulation and prediction results
For this example the prediction is very good and some confidence in future
predictions can be expected. If the prediction does not model the fractional
productions well, then the fractional flow can be rematched using different
data point weightings or Corey parameters altered by hand.
Select Finish | Done | File | Save to complete this tutorial example.
3.4 MBAL Oil History Matching Example
This example presumes that the user is familiar with setting up single tank
models in MBAL.
In this exercise the objectives are:
• Quality-check the production data that is available. This quality check
is based on what is possible physically.
• Based on the checked data build MBAL model and identify various
drive mechanisms and fine tune Oil in place (OIP) estimates using
History matching techniques.
• Quality-check the fine tuned MBAL model selected after history match
using techniques available in MBAL software.

TUTORIAL GUIDE
Methodology:
The oil reservoir model will be built step by step. At each step if any laboratory
or field data is available, its quality will be checked and then it will be used in
the model.
The available data is described in the following sections.
examplesOil history matching
3.4.1 PVT Data
(@ 250 deg F)
• Bubble point (Pb) = 2200 psig
• Solution GOR = 500 SCF/STB
• FVF@ Pb = 1.32 RB/STB
• Oil Visc.@ Pb = 0.4 cP
• Oil gravity = 39 API
• Gas grav. = 0.798
• Water Salinity = 100,000 PPM
3.4.1.1 Production data
This data is contained in an Excel file OILRES1.XLS. Later in this chapter a
description on how to transfer the data from Excel into MBAL will be provided.
3.4.2 Setting up the Basic Model
Note: If you comfortable with setting up basic MBAL single tank models, you
can skip this section and go to the next section. The basic tank model that this
section is used to set up is under the file name “res1.mbi”.
• Start MBAL and select the menu option File | New.
• On the menu bar go to Tools and click on Material Balance.
• On the menu bar go to Options and following screen appears. Fill the
screen with the following details:

Figure 157: Setting the option
In this screen, you have defined oil as the main fluid, selected a simple tank
model and will enter the production history by tank. You also do not want to
do compositional tracking for this reservoir.
Then again on the menu bar go to PVT | Fluid Properties and supply the
following data:

TUTORIAL GUIDE
Figure 158: PVT input
In this section we have specified the Black oil properties of the oil as given in
the PVT data available section. We also have specified water salinity and
indicated that the produced gas has no CO2, H2S or N2 in it.
In the previous screen we also have indicated that we want the PVT behaviour
to be predicted by Glaso and Beggs et al correlations. Since we do have
laboratory measured data for this fluid at bubble point conditions as given in
available PVT data section, we will match the lab data to the correlations that
we are selecting. In the PVT Input dialog, press the Match button and the
following screen appears and we can enter measured data at bubble point as
indicated in the following screen:

Figure 159: Entering PVT match data
After we have entered the data, if we click on Match | Calc this will start the
matching process. Before you hit Calc, ensure that the Match all option as
shown in the screen below is ticked. This will make the program match the lab
data to all the correlations available in MBAL.

TUTORIAL GUIDE
Once this is done, click the Match Param button to look at parameters of
each of the correlations and the standard deviation, to see which of them
does the best job. In this case we select Glaso for bubble point, GOR and
FVF calculations; and Beggs for viscosity calculation.

Figure 161: Matched parameters
At this stage you have finished specifying the PVT properties of the fluid of
your tank. Now the next step is building your tank model. In the main menu
bar go to Input | Tank Data, and supply the following information. This is the
basic information about the reservoir that you must have.
Tank type Oil
Tank name Tank01
Temperature 250 degree F
Initial pressure 4000 psig
Porosity 0.23
Connate water saturation 0.25
Water compressibility Use Corr
Initial gas cap 0
Original oil in place 206 MMSTB
Start of production 01/01/1998
Figure 162: Tank input screen – tank parameters
In this screen we have said that we have a reservoir which has oil as its
primary fluid, it is at a temperature of 250 deg F, the initial pressure of this
reservoir was 4000 psig. The average porosity within the reservoir is 23%.
The connate water saturation is 25%. Note that the initial gas cap field is not
available to be edited. This is because, in PVT section we specified that at 250

TUTORIAL GUIDE
F, the bubble point was 2200 psig and at 4000 psig the reservoir will be
undersaturated and will have no gas cap. The program on the basis of the
tank temperature, pressure and the PVT section will determine whether the
reservoir is undersaturated or not. In case it is not you will also require an
initial estimate of the gas cap.
You also will be required to enter an initial estimate of Oil in Place, obtained
from geological surveys for example. The screen also requires entering a
start production date for this reservoir.
Next information to be supplied is the aquifer support to the field. As there is yet no
evidence to suggest the presence of an aquifer, this will be left to “None” for the time
being.
The next information about the tank that we will have to enter is the rock
properties. We can enter rock compressibility by hand, we can use the
correlations to evaluate rock compressibility for us or we can enter
compressibility as a function of pressure in table form, if we have the data. In
this example we select to use correlations.
Figure 163: Tank input screen – Rock compressibility
The next data you have to enter is the relative permeability data. The relative
permeability data is used in prediction calculations only. It is used to find WC
and producing GOR, which are basically water fractional flow and gas
fractional flow and depend on the water and gas saturation in the tank. If we

have an initial gas cap and we are producing from it, we should use the total
reservoir volume including that of gas cap to find saturation. If this is the case
it should be selected in the screen of Tank Parameters. The relative
permeability data can be entered as a table or as a Corey function. If you click
on the Rel Perm from combo box, both these options are revealed. In this
case we enter the Corey functions.
Phase Residual Saturation
(fraction)
End Point
(fraction)
Exponent
Water 0.25 0.7 1.5
Oil 0.15 0.8 1.3
Gas 0.02 0.9 1
Figure 164: Tank input screen – Relative permeability
You also enter water and gas sweep efficiency values in this screen. These
values are used to find the velocity at which the OWC and GOC contact
move, when the monitor contacts option is selected.
The last data that we have to supply is the production history of the reservoir
as shown in the following screen. Note that this can be copied from the Excel
file OILRES1.XLS.

TUTORIAL GUIDE
Figure 165: Tank input screen – History matching
At this stage we have specified all the input data to the reservoir and we
should check if everything is in order. To do so, if we go to the main screen
menu and select Input | System summary the following screen appears and
tells us if we have any missing or invalid data entry. In this case everything is
okay.

Figure 166: Tank input summary
This finishes our setting up of basic tank model. It is advisable to save the file
at this point. Next step would be to fine-tune the model, in terms of identifying
and quantifying its various drive mechanisms.

TUTORIAL GUIDE
3.4.3 Matching to Production History data in MBAL
The first thing to do is to see whether our production history data is consistent
with our PVT data. In the PVT section we indicated that the bubble point was
2200 psig and the solution GOR was 500 Scf/STB. If we go to the production
history screen in the tank input data, we can click on the option Work with
GOR at the bottom of the dialog and the gas rates are converted into
producing GOR values.
Figure 167: Checking consistency of data
In the production history, if you scroll down on this screen, you will see that
the reservoir pressure is always above 2200 psig. Thus there is no free gas in
the tank and hence the producing GOR should be the one coming from
solution. Indeed in this case all the gas rates covert into GOR values which
are nearly 500 SCF/STB. Thus the data is consistent with the PVT. In case it
is not so, we need to go back to PVT or source of production data and try find
the reason for the anomaly.
Once we are sure that the production history data is consistent with the PVT,
the next step is to see how the model that we have set up compares to the
history data. At this stage it is important to note that we are after a model
which performs well on each method and not only on some. Thus we should
start with History Matching | All. This produces the following plots.

Figure 168: History matching
Note that in the graphical methods the plot shown in the screen above is the
Campbell plot. You may not get this initially. You should click on the graphical
method screen and in the menu bar of the above screen as shown appears.
Select Method | Campbell Plot.
The first plot is called the energy plot. It indicates contribution of various drive
mechanisms towards production with time.
The second plot is the Campbell plot, which is a graphical technique used to
find oil in place given a production history and known drive mechanisms
(diagnostics).
The last plot is the analytical plot. On the y-axis of the analytical plot is the
tank pressure and on the x-axis is the primary phase production (in this case
oil). The data points are the actual pressures with oil production that we have
entered in the history. The blue line indicates the response of the model
according to the data entered in the Tank Data screen.
Based on the response of the Campbell plot, the presence of an aquifer is
very likely (source of energy). Therefore an aquifer model can be selected in
the tank data section:

TUTORIAL GUIDE
Figure 169: Initialising an aquifer model
Going back to History Matching/All:

If we look at the analytical plot, it indicates that with the current aquifer model,
we are predicting production rates higher than those actually observed. Thus
we may have a weaker aquifer. We can decrease the strength of the aquifer
either by accessing on the tool bar of the previous screen Input | Tank data
and decreasing the aquifer inner to outer radius ratio (rD). We can accomplish
the same, in the WD plot of the above screen, if we double click at a the
smaller rD value of 4.0, as indicated in the plot below, so that on the analytical
plot the actual history points and the solid line come closer as shown below.

TUTORIAL GUIDE
Figure 171: History Matching Plot
At this stage if you look at the analytical plot, we can see that the match we
are getting is quite good. On this plot the model selected and the history
entered seem to be in good agreement with each other. But on the other
hand, the model we have selected does not fit well on the Campbell plot,
where ideally we should get a horizontal straight line. This is the reason that
we recommended at the start that while doing history match, all the screens
should be used simultaneously and a model that fares well according to all
different methods be used.
At the end of this step we are very close to representing the reservoir
behaviour and we can fine-tune it by doing a regression analysis. To activate
regression analysis button, we have to click on the analytical plot and in the
menu bar of the above screen select Regression, which will activate the
following screen

Figure 172: Regression
In all of the variables we can change on the above screen, we select the
parameters that influence aquifer behaviour and the OIP itself. This is a good
choice for this case because from the energy plot we see that these are the
two major components of the energy of the system. We do not change the
compressiblity because we believe the correlations do a good job for this
case.
At the end of regression we get the following best match.

TUTORIAL GUIDE
Figure 173: Transferring the regressed data
Once the regression is finished, the best fit data should transferred to start
column by clicking on the transfer button which is the arrow button between
the Start and Best Fit text.
After transferring the data if we click on Done we get the following plots:

Figure 174: History matching done
The model got at this stage in terms of OIP and various drive mechanisms
seems to satisfy all the methods and is thus acceptable
Save the file as Oilres.mbi.
3.4.4 Running Sensitivity Analysis on the Tank Model
Once at the end of history matching we have selected a model in terms of OIP
and various drive mechanisms, it is important that we do some sort of analysis
on the figures we have arrived at check our confidence in these figures. This
is allowed by doing a sensitivity analysis on the model. In sensitivity analysis
what we do is that we try to see how sensitive our model is to the change in
parameters that we have fixed by history match.
If we go to the main menu and select History Match | Sensitivity the
following screen appears. In this case we want to study the effect of changing
the OIP place only. We are trying to see the effect of changing it between the
values of 180 and 250.

TUTORIAL GUIDE
Figure 175: Sensitivity
On this screen if the “Plot” button is selected, the following plot is obtained.
Figure 176: Sensitivity plot

On the x-axis is the OIP and on the y-axis is the standard deviation in terms of
predicted production rates over the history. The presence of a minimum
shows the uniqness of the solution.
We can similarly do sensitivity analysis on other drive mechanisms like the
aquifer parameters for this case and find about our confidence in those figures
as well.
3.4.5 Using Simulation Option to Quality check the
History Matched Model
At the end of running sensitivity analysis the next step we have identified our
confidence in the drive mechanisms. At this stage it must be noted that in the
regression analysis that we did in analytical plot, we take the tank pressure
and non primary phase production and with the model calculate the error in
production rate of primary phase, oil in this case.
In simulation what we do is exactly opposite. With the given model we have
fixed we take all the phase rates from the history and try to predict the
pressure, phase saturation in the tank and other tank parameters. If our model
is acceptable the last test it should satisfy is that it should be able to
reproduce pressure as well.
If in the main menu we select History Matching | Run simulation |
Calculate , the program does calculations. At the end of calculation if you hit
Plot, the following plot appears:-
Figure 177: Comparison between simulation and history

TUTORIAL GUIDE
This plot has the pressure with time plotted both from simulation and
production history data. In this case both are identical and thus the match
attained is good.
Â Note: The model is not ready at this stage to go ahead with
predictions and study various development alternatives. Fractional
flow matching in order to create pseudo relative permeability curves
should be done and also the verification of these as demonstrates in
the previous example.
3.5 Performing a Production Prediction starting from
a history matched model
The following example shows how to perform prediction runs with MBAL.
The file GasTank1.mbi has already been history matched.
examplesProduction Prediction
Statement of the problem
An MBAL model has been created and history match performed. It will be used to
generate production predictions for two cases:
Case 1: Follow the current production pattern – WHFP = 800 psig
Case 2: Reduce the well head pressure to 100 psig from 01/01/2004
Input Data
Well Inflow Data
In this example, the well Inflow is represented by the C n model. Any
sophisticated IPR model can be collapsed to the C n method. For more
details on Inflow models, please refer to PROSPER, the well bore modelling
package.
Well Name C- factor (Mscf/d/psi2
) N-Exponent
Producer#1 0.027 0.85
Producer#2 0.002 0.95
Producer#3 0.005 0.9
Generic Well Outflow tables:
In this example, it will be assumed that all wells have the same lift tables.
Lift tables can be generated with PROSPER and then imported in MBAL. They
have already been prepared and can be found in:
C:ProgramFilesProgramFilesPetroleumexpertsIPM5.0SamplesWorked
ExampleGasTank Well Lift Tables.TPD

Cases to study
Case 1: WHFP = 800 psig
Case 2: WHFP is reduced to 100 psig from 01/01/2004 by installing a
compressor in the field.
Step-by-Step procedure for a prediction run with MBAL
Step 1: Prediction set-up
In order to perform a prediction, select |Production Prediction |Prediction
Setup and make the following changes:
Figure 178: Prediction Calculation Setup
Select |Done to complete this.
Step 2: Boundary conditions at surface
Now select |Production Prediction|Production and Constraints and enter date and
Manifold Pressure as shown in the screenshot below:

TUTORIAL GUIDE
Figure 179: Prediction Production and Constraints
The manifold pressure is the pressure in the node furthest from the sandface
in the well model used to generate the lift tables. In this example, the manifold
pressure is the well head pressure.
Now select |Done to complete this.
Step 3: Well Type Definition
Each well is defined by a VLP and an IPR. VLP stands for vertical lift
performance, whereas IPR stands for Inflow Performance Relationship.
Select |Production Prediction|Well Type Definition, the following screen is
accessed:
Figure 180: Well Input Data

Clicking on the “ ” button, a well is created with the default name “Well1”
that can be overwritten.
Here we will overwrite the default well name with “Producer#1P”
Make sure that the well type is correct. Here the well type is “Dry Gas
Producer”.
This is how the screen looks like:
Figure 181: Well Input Data
Select the Next button to proceed to the Inflow section.
Step 4: Inflow Performance Tab
Once entered the Inflow tab screen, make the following changes:

TUTORIAL GUIDE
Figure 182: Well Inflow Data
To assign the relative permeabilities derived during the fractional flow
matching, select |Edit|Copy. In this way, a list of all permeability tables
available in the MBAL model is displayed:
Figure 183: Relative Permeability Data Transfer
Here select the appropriate one and then select |Copy and then the Corey
parameters are assigned to the well.

Now click on |Done. MBAL will ask if the water breakthrough saturation is to
be copied:
Figure 184: Water Breakthrough copy
Now select |Yes to continue with |Next|Next (skipping the More Inflow).
Note: How to derive relative permeability curves by fractional flow matching:
After running a simulation, select History Matching⎪Fw Matching from the
main menu toolbar. The follow plot is displayed:
Figure 185: Fw Matching plot
From the menu toolbar of the plot screen, select Well, and then the well to
match, for example Producer#1:

TUTORIAL GUIDE
Figure 186: Fw Matching plot
In the plot area, double click on a point corresponding to the desired
breakthrough water saturation (for example the minimum water saturation,
which is 0.29 in this case). After that, click on Regress, so that the program
can perform a regression to fit the points calculated by the simulation, and a
set of Corey function parameters are calculated, which give the same
fractional flow.

Figure 187: Fw Matching plot while performing regression

TUTORIAL GUIDE
Step 5: Lift tables import
Figure 188: Well Outflow Data
In this screen click on the |Edit button and then select |Import. Using the
browser, import the TPD file provided.
TPD files are lift tables files for Petroleum Experts applications (GAP, REVEAL,
MBAL).
Figure 189: TPD file import

Clicking on |Open, a statistics about the flow tables is shown:
Figure 190: TPD file import
Select |Done and this completes the setup of this well.
Repeat the same process for all the 3 wells.
This is how the main screen will look like:
Figure 191: MBAL final model

TUTORIAL GUIDE
Step 6: Well scheduling.
In order to schedule the wells, select |Production Prediction|Well Schedule
and make the following changes:
Figure 192: Well Schedule
Click on |Done to validate the screen.
Step 7: Reporting schedule
For this example, select |Production Prediction|Reporting Schedule and make
the following changes:

Figure 193: Reporting Schedule
The “Keep History” button allows to have the full history stream along with the
prediction stream for comparison purposes. Keep Automatic Reporting
Frequency and click on Done to exit the screen.
Step 8: Running the prediction
Choose |Production Prediction|Run Prediction|Calculate, then Ok. The results
of the calculation are displayed:

TUTORIAL GUIDE
Figure 194: Run Production Prediction
Click on |Plot|Variables and make the following choices:
Figure 195: Plot Variables

Afterwards, click on |Done. The following plot is displayed:
Figure 196: Average Gas Rate Plot
Now starting from this model it is possible to make forecasts on multiple field
scenarios.
Step 9: Saving a case a stream within MBAL
From the plot toolbar choose |Finish. It is possible to save the results of each
prediction, so that they can be reviewed later and compared to other
scenarios.
In the Run Prediction screen, click on Save:
Figure 197: Save Prediction
Then click on|Add and overwrite the stream name as suggested below:

Figure 200: Save Results screen
After clicking on |Done and then |Plot|Variables, make the following
selections, for instance:
Figure 201: Plot Variables
Then |Done and the following plot is displayed:

TUTORIAL GUIDE
Figure 202: Average Gas Rate Plot for Case 1 and 2
This plot shows a comparison between Case 1 and Case 2.
The example is now terminated.
Please save the file as:
C:ProgramFilesPetroleumExpertsIPM5.0WorkedexamplesPhysicsexample
sProduction predictionGas Tank1 solved.mbi

Tutorial_Complete.pdf

More Related Content

What's hot (20)

Similar to Tutorial_Complete.pdf (20)

Recently uploaded (20)

Tutorial_Complete.pdf