Read PIPESIM User Guide text version

PIPESIM Suite

User Guide

Proprietary Notice

Copyright 1985 - 2005 Schlumberger. All rights reserved. No part of this document may be reproduced, stored in a retrieval system, or translated in any form or by any means, electronic or mechanical, including photocopying and recording, without the prior written permission of Schlumberger. Use of this product is governed by the License Agreement. Schlumberger makes no warranties, express, implied, or statutory, with respect to the product described herein and disclaims without limitation any warranties of merchantability or fitness for a particular purpose.

Patent information

Schlumberger ECLIPSE reservoir simulation software is protected by US Patents 6,018,497, 6,078,869 and 6,106,561, and UK Patents GB 2,326,747 B and GB 2,336,008 B. Patents pending.

Service mark information

The following are all service marks of Schlumberger: The Calculator, Charisma, ConPac, ECLIPSE 100, ECLIPSE 200, ECLIPSE 300, ECLIPSE 500, ECLIPSE Office, EDIT, Extract, Fill, Finder, FloGeo, FloGrid, FloViz, FrontSim, GeoFrame, GRAF, GRID, GridSim, NWM, Open-ECLIPSE, PetraGrid, PlanOpt, Pseudo, PVTi, RTView, SCAL, Schedule, SimOpt, VFPi, Weltest 200.

Trademark information

Silicon Graphics and IRIX are registered trademarks of Silicon Graphics, Inc. OpenGL® and the oval logo are trademarks or registered trademarks of Silicon Graphics, Inc. in the United States and/or other countries worldwide. OpenInventor and WebSpace are trademarks of Silicon Graphics, Inc. IBM, AIX and LoadLeveler are registered trademarks of International Business Machines Corporation. Sun, SPARC, Solaris, Ultra and UltraSPARC are trademarks or registered trademarks of Sun Microsystems, Inc. Macintosh is a registered trademark of Apple Computer, Inc. UNIX is a registered trademark of UNIX System Laboratories. Motif is a registered trademark of the Open Software Foundation, Inc. The X Window System and X11 are registered trademarks of the Massachusetts Institute of Technology. PostScript and Encapsulated PostScript are registered trademarks of Adobe Systems, Inc. OpenWorks and VIP are registered trademarks of Landmark Graphics Corporation. Lotus, 1-2-3 and Symphony are registered trademarks of Lotus Development Corporation. Microsoft, Windows, Windows NT, Windows 95, Windows 98, Windows 2000, Windows XP, Internet Explorer, Intellimouse, Excel, Word and PowerPoint are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries. Netscape is a registered trademark of Netscape Communications Corporation. AVS is a registered trademark of AVS Inc. ZEH is a registered trademark of ZEH Graphics Systems. Ghostscript and GSview are Copyright of Aladdin Enterprises, CA. GNU Ghostscript is Copyright of the Free Software Foundation, Inc. Linux is Copyright of the Free Software Foundation, Inc. IRAP is Copyright of Roxar Technologies. LSF is a registered trademark of Platform Computing Corporation, Canada. VISAGE is a registered trademark of VIPS Ltd. Cosmo is a trademark and PLATINUM technology is a registered trademark of PLATINUM technology, inc. PEBI is a trademark of Veritas DGC Inc./HOT Engineering GmbH. Stratamodel is a trademark of Landmark Graphics Corporation. GLOBEtrotter, FLEXlm and SAMreport are registered trademarks of GLOBEtrotter Software, Inc. CrystalEyes is a trademark of StereoGraphics Corporation. Tektronix is a registered trade mark of Tektronix, Inc. GOCAD and JACTA are trademarks of T-Surf. Myrinet is a trade name of Myricom, Inc. This product may include software developed by the Apache Software Foundation (http://www.apache.org). Copyright (c) 1999-2001 The Apache Software Foundation. All rights reserved. MPI/Pro is a registered trademark of MPI Software Technology, Inc. The TGS logo is a trademark of TGS, Inc. LAPACK is Copyright 1999 Society for Industrial and Applied Mathematics, Philadelphia, PA, http://www.netlib.org/lapack/.

Contact information

Web:

www.sis.slb.com

Support: Service Desk Note: Information in this document is subject to change without notice. Companies, names and data used in examples herein are fictitious unless otherwise noted.

Contents

3

Table of Contents

Proprietary Notice ............................................................................ 2 Patent information............................................................................ 2 Service mark information ................................................................ 2 Trademark information .................................................................... 2 Contact information ......................................................................... 2

TABLE OF CONTENTS ....................................................... 3 DOCUMENT CONVENTIONS ............................................ 10 PIPESIM HOT KEYS ........................................................ 10 1 INTRODUCTION .......................................................... 15

1.1 Setting up ............................................................................. 15 1.1.1 Before you run setup ...................................................... 15 1.1.2 Running setup ................................................................ 17 1.1.3 Changing Options after quitting setup ............................ 17 1.2 Documentation..................................................................... 17 1.2.1 PIPESIM additional documentation ................................ 17 1.2.2 Case Studies .................................................................. 18 1.2.3 Online Help..................................................................... 18 1.3 PIPESIM overview................................................................ 19 1.3.1 Modules .......................................................................... 20 1.3.2 Options ........................................................................... 23 1.4 File Management.................................................................. 25

1.5 Security ................................................................................ 26 1.5.1 Stand-alone security (dongle)......................................... 26 1.5.2 LAN Security................................................................... 27

PIPESIM

4

Contents

1.6 1.7 1.8

New features ........................................................................ 28 Schlumberger Support Services ........................................ 28 What to do next.................................................................... 28

2

2.1 2.2 2.3

MODEL OVERVIEW .................................................... 31

Steps in building a model ................................................... 31 Starting PIPESIM.................................................................. 31 Units System ........................................................................ 31

2.4 Fluid data.............................................................................. 32 2.4.1 Black Oil ......................................................................... 32 2.4.2 Compositional................................................................. 34 2.4.3 Steam ............................................................................. 35 2.5 Model components overview.............................................. 35 2.5.1 Model & Component limitations...................................... 39 2.6 2.7 2.8 Flow correlation ................................................................... 40 Run an operation ................................................................. 40 Saving & Closing PIPESIM.................................................. 41

2.9 How to build models............................................................ 41 2.9.1 Fluid calibration .............................................................. 41 2.9.2 Pipeline & facilities.......................................................... 42 2.9.3 Well Performance ........................................................... 45 2.9.4 Network Analysis ............................................................ 48 2.9.5 Production Optimization ................................................. 50 2.9.6 Field Planning................................................................. 50 2.9.7 Multi-lateral ..................................................................... 51

3

FLUID & MULTIPHASE FLOW MODELING ............... 52

3.1 Black Oil ............................................................................... 52 3.1.1 Lasater............................................................................ 52

PIPESIM

Contents

5

3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.1.7 3.1.8 3.1.9 3.1.10 3.1.11

Standing ......................................................................... 53 Vazques and Beggs ....................................................... 53 Glasø .............................................................................. 54 Coning ............................................................................ 55 Liquid Viscosity............................................................... 56 Dead Oil Viscosity .......................................................... 56 Live Oil Viscosity ............................................................ 57 Undersaturated Oil Viscosity .......................................... 58 Oil/Water Mixture Viscosity............................................. 59 Gas Viscosity.................................................................. 60

3.2 Compositional...................................................................... 60 3.2.1 EOS (Equations of State) ............................................... 60 3.2.2 Viscosity model............................................................... 61 3.2.3 BIP (Binary Interaction Parameter) Set .......................... 63 3.2.4 Hydrates ......................................................................... 63 3.3 Pressure Drop Calculation.................................................. 65 3.3.1 Flow regimes .................................................................. 66 3.3.2 Single Phase Flow Correlations ..................................... 69 3.3.3 Vertical Multiphase Flow Correlations ............................ 70 3.3.4 Horizontal Multiphase Flow Correlations ........................ 76 3.4 References ........................................................................... 80

4

RESERVOIR, WELL & COMPLETION MODELING ... 87

4.1 Vertical Completions ........................................................... 87 4.1.1 Liquid Reservoirs............................................................ 87 4.1.2 Gas and Gas Condensate Reservoirs............................ 89 4.2 Horizontal Completions ...................................................... 91 4.2.1 Effect of Pressure Drop on Productivity.......................... 91 4.2.2 Single Phase Pressure Drop .......................................... 94 4.2.3 Multiphase Pressure Drop .............................................. 95 4.2.4 Inflow Production Profiles ............................................... 95 4.2.5 Steady-State Productivity ............................................... 96 4.2.6 Pseudo-Steady State Productivity .................................. 99 4.2.7 Solution Gas-Drive IPR ................................................ 101

PIPESIM

6

Contents

4.2.8 4.3

Horizontal Gas Wells .................................................... 101

Multiple Layers / Completions.......................................... 103

4.4 Artificial Lift........................................................................ 104 4.4.1 Gas Lift ......................................................................... 104 4.4.2 ESP Lift......................................................................... 105 4.5 Tubing................................................................................. 105

4.6 Chokes................................................................................ 106 4.6.1 Ashford-Pierce.............................................................. 106 4.6.2 Omana.......................................................................... 107 4.6.3 Gilbert, Ros, Baxendall, Achong and Pilehvari............. 108 4.6.4 Poettmann-Beck ........................................................... 109 4.6.5 Mechanistic Correlation, ............................................... 110 4.6.6 API 14-B Formulation ................................................... 112 4.7 Heat transfer....................................................................... 113

4.8 Reservoir Depletion........................................................... 113 4.8.1 Volume Depletion Reservoirs ....................................... 113 4.8.2 Gas Condensate Reservoirs ........................................ 115 4.9 References ......................................................................... 115

5

5.1 5.2 5.3

FIELD EQUIPMENT ................................................... 119

Compressor........................................................................ 119 Expander ............................................................................ 120 Single Phase Pump ........................................................... 121

5.4 Multiphase Boosting ......................................................... 121 5.4.1 Multiphase Boosters ­ Positive Displacement Type..... 126 5.4.2 Twin Screw Type Multiphase Boosters ........................ 127 5.4.3 Progressing Cavity Type Multiphase Boosters............. 129 5.4.4 Multiphase Boosters ­ Dynamic Type .......................... 130 5.4.5 Helico-Axial Type Multiphase Boosters ........................ 131 5.4.6 Contra-Rotating Axial Type Multiphase Booster........... 133

PIPESIM

Contents

7

5.4.7 5.5 5.6 5.7 5.8

Alternative approach..................................................... 134

Separator ............................................................................ 135 Re-injection point .............................................................. 135 Heat Transfer...................................................................... 135 References ......................................................................... 135

6

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8

OPERATIONS ............................................................ 139

Check model ...................................................................... 139 No operation....................................................................... 139 Run model .......................................................................... 140 System Analysis ................................................................ 140 Pressure Temperature profile........................................... 140 Flow correlation matching ................................................ 140 Wax Prediction................................................................... 141 Nodal Analysis ................................................................... 141

6.9 Artificial Lift Performance................................................. 142 6.9.1 Well Performance Curves............................................. 143 6.9.2 Optimization module performance curves .................... 143 6.10 Gas Lift Design & Diagnostics ......................................... 145 6.10.1 Check for Gas Lift instability ......................................... 145 6.11 6.12 6.13 6.14 Horizontal well analysis .................................................... 148 Reservoir tables................................................................. 148 Network analysis ............................................................... 149 Production Optimization ................................................... 149

PIPESIM

8

Contents

6.15 Field Planning .................................................................... 150 6.15.1 Dynamic Eclipse link..................................................... 150 6.15.2 Look-up tables .............................................................. 152 6.15.3 Compositional tank models .......................................... 153 6.15.4 Event handling.............................................................. 154 6.16 Multi-lateral well analysis.................................................. 155

6.17 Post processor................................................................... 155 6.17.1 Graphical plots.............................................................. 155 6.17.2 Tabular data ................................................................. 156 6.17.3 Onscreen data .............................................................. 156 6.18 References ......................................................................... 156

7

CASE STUDIES ......................................................... 159

7.1 Pipeline & facilities Case Study ­ Condensate Pipeline 161 7.1.1 Task 1. Develop a Compositional Model of the Hydrocarbon Phases .................................................................. 161 7.1.2 Task 2. Identify the Hydrate Envelope.......................... 162 7.1.3 Task 3. Select a Pipeline Size ...................................... 163 7.1.4 Task 4. Determine the Pipeline Insulation Requirement 165 7.1.5 Task 5. Screen the Pipeline for Severe Riser Slugging 167 7.1.6 Task 6. Size a Slug Catcher ......................................... 170 7.1.7 Data Available .............................................................. 172 7.2 Well Performance Case Study ­ Oil Well Design............ 175 7.2.1 Task 1. Develop a Calibrated Blackoil Model ............... 175 7.2.2 Task 2. Develop a Well Inflow Performance Model...... 180 7.2.3 Task 3. Select a Tubing Size for the Production String 180 7.2.4 Data Available .............................................................. 182 7.3 Network Analysis Case Study ­ Looped Gas Gathering Network ...................................................................................... 7-184 7.3.1 Task 1. Build a Model of the Network........................ 7-184 7.3.2 Task 2. Specify the Network Boundary Conditions ... 7-189

PIPESIM

Contents

9

7.3.3 7.3.4 7.4 7.5 7.6

Task 3. Solve the Network and Establish the deliverability 7-190 Data Available ........................................................... 7-192

Optimization .................................................................... 7-194 Field Planning ................................................................. 7-194 Multi-lateral...................................................................... 7-194

8

INDEX ......................................................................8-194

PIPESIM

10 Document conventions

Conventions

<edit/copy> - used to denote commands enter into the computer from either Microsoft Windows operating systems or PIPESIM

PIPESIM

Conventions THIS PAGE LEFT BLANK INTENTIONALLY

11

PIPESIM

12

Conventions

PIPESIM Hot Keys

File Create New Well Model Create New Pipeline Model Create New Network model Open model Open engine file Save model Close PIPESIM Text Edit Export to Engine file Purge Engine Files Simulation Run model Restart Model Check model Windows New Model Window Close Active Window Go to Next Window Go to Previous Window Tools Print Access Help Editing/General Access Pull-down menus Cut Copy Paste Delete Select All Find Sticky key mode

PIPESIM

CTRL+W CTRL+ CTRL+N CTRL+O CTRL+T CTRL+S ALT+F4 CTRL+T CTRL+E CTRL+Y CTRL+G CTRL+R CTRL+E CTRL+W CTRL+F4 CTRL+F6 or CTRL+TAB CTRL+SHIFT+F6 or CTRL+SHIFT+ TAB CTRL+P F1 ALT or F10 CTRL+X CTRL+C CTRL+V Del CTRL+A CTRL+F SHIFT

Conventions Zoom in Zoom out Zoom Full View Restore View SHIFT+Z SHIFT+X SHIFT+F SHIFT+R

13

PIPESIM

Field Equipment

15

1 Introduction

Welcome to Schlumberger's PIPESIM - the integrated Petroleum Engineer and Facilities package for Design, Operation and Optimization. 1.1 Setting up You install PIPESIM on your computer by using the program SETUP.EXE. The setup up program installs ESI

M

16 · · · ·

Field Equipment A mouse 32Mb of RAM Microsoft Windows 98 or higher The PC system date is set to the current date. The security system uses the current PC date.

The recommended system requirements are: · Pentium III processor 600MHz · 3Gb hard disk · A 4x CD-ROM drive · A SVGA display running in 1024x768 and 256 colors · A 2 button mouse · 64Mb of RAM · Microsoft Windows 2000 1.1.1.2 Check the PIPESIM package The following items should be in the PIPESIM package: · PIPESIM User Guide · PIPESIM Additional Notes · PIPESIM Service Pack Notes (if applicable) · PIPESIM Installation Guide · PIPESIM CD · Registration form (also available on our web site) · Software license reference number. This should be quoted on all correspondence. If any of the above are missing then please contact your nearest Schlumberger office. 1.1.1.3 Make backup copies Before you run the install procedure please back up copies of any important data stored on your PC. You are also encouraged to make a back up copy of the install CD. 1.1.1.4 Read the additional notes document The additional notes' document (shipped with the package) lists any changes to the User Guide since its publication.

PIPESIM 2000

Field Equipment 1.1.2 Running setup When you run the setup program To start Setup

17

Once you have installed PIPESIM the following links will be created on the Programs menu; · Schlumberger · PIPESIM · GOAL · FPT · HoSim · Documentation · OpenLink · Utilities · B26 to P2K Converter · Security utilities · User defined DLL registry editor · Plotting utility 1.1.3 Changing Options after quitting setup You can run they setup program as many times as you like to install, re-install or remove components. However, only 1 copy of PIPESIM can be installed on a single PC. 1.2 Documentation 1.2.1 PIPESIM additional documentation In addition to this User Guide the following documentation is available to assist users in using PIPESIM or some of its modules. The latest versions of these documents are available from any Schlumberger support office or can be downloaded directly from the Schlumberger web site in Adobe Acrobat PDF format. 1.2.1.1 Artificial lift Performance curve The optimizer module utilizes artificial lift performance curves to model the wells. These can be created by a suitable Nodal analysis software package.

PIPESIM 2000

18

Field Equipment

1.2.1.2 User Defined Multiphase flow correlation The user can create their own multiphase flow correlations and link these into PIPESIM. 1.2.1.3 OpenLink A collection of COM object that allows PIPESIM to be accessed from 3rd party applications, e.g. Microsoft Excel, Visual basic, etc. An up to date list of features and functionality can be obtained from the Schlumberger web site, along with all the necessary documentation. 1.2.1.4 PVT file format The composition can be transferred from third party applications directly into PIPESIM, provide that it is supplied in the correct format. This document details that format. 1.2.1.5 Sentinel LM Security The LAN version of PIPESIM utilizes Sentinel LM License manger as its security system The Sentinel LM Administrators Guide can be of assistance to IT personnel. Note: This User Guide does not cover the menus or dialogs that are used within the software. These are covered, in detail, in the Help system, supplied with PIPESIM. 1.2.2 Case Studies The PIPESIM installation installs sample models on to your hard disk. 1.2.3 Online Help You can access Help through; · the Help Contents command, · by searching for specific topics with the Help Search tool · pressing F1 to get context-sensitive Help. 1.2.3.1 Help contents For information on Help topics, choose Contents from the Help menu or press F1 and click the Contents button. You can use the Contents

PIPESIM 2000

Field Equipment

19

screen to jump to topics that tell you how to useaBT5/TT4 1- Tc 0.000 Tc 0 Tw 13.98

t c

m

p

e

k

a

n

s

hh

.

h

m

(

t

ee

l

l

e

S

n

t

e

)

T

a

j

r1

3

.

c

9

8

h

0

m

0

b

1

3

u.

9

T

t

c

t

0

6

o 0

5

0

n

0

7

2

1u

7

2

s

7

.

e

5

c

a

a

n

-

g

e

t

p

T

20

Field Planning Multi-lateral well Multi-zone wells

Field Equipment

This release of PIPESIM does not have all modules fully integrated, i.e. Production Optimization (GOAL), Field Planning (FPT), Multilateral well (HoSim). 1.3.1 Modules PIPESIM consists of the following modules: · Pipeline & Facilities · Well Performance Analysis · Network Analysis · Production Optimization (GOAL) · Field Planning (FPT) · Multi-lateral (HoSim) 1.3.1.1 Pipeline & Facilities A comprehensive multiphase flow model with "System Analysis" capabilities. Typical applications of the module include: · multiphase flow in flowlines and pipelines · point by point generation of pressure and temperature profiles · calculation of heat transfer coefficients · flowline & equipment performance modeling (system analysis) 1.3.1.2 Well Performance analysis A comprehensive multiphase flow model with "Nodal & System Analysis" capabilities. Typical applications of the module includes: · Well design · Well optimization · Well inflow performance modeling · Gas Lift Design · ESP Design · Gas lift performance modeling · ESP performance modeling · Horizontal well modeling (including optimum horizontal completion length determination) · Injection well design · Annular and tubing flow

PIPESIM 2000

Field Equipment

21

1.3.1.3 Network analysis module Features of the network model include: · unique network solution algorithm to model wells in large networks · rigorous thermal modeling of all network components · multiple looped pipeline/flowline capability · well inflow performance modeling capabilities · rigorous modeling of gas lifted wells in complex networks · comprehensive pipeline equipment models · gathering and distribution networks 1.3.1.4 Production Optimization (GOAL) This module allows production optimization of an artificial lifted (gas lift or ESP) oil field to be performed given a number of practical constraints on the system. The module will predict the optimum artificial lift quantity (lift gas or ESP speed) so as to optimize oil production from the entire field. As an alternative to calculations based on produced oil the optimization can be performed on gross liquids, gross gas or revenue. The program models the full network on a point-by-point basis, and offers a choice of flow correlation options for multiphase flow. In addition to being able to optimize field production it includes a unique production prediction mode, which allows current field production rates and pressures to be predicted and the results compared directly against actual field data. The module has been primarily developed for use by operations staff in the day-to-day optimization and allocation of lift gas for complex multi-well networked configurations. GOAL has been designed with to allow answers to specific problems to be easily obtained. This could be, for example, when a well is shutin and the extra quantity of lift gas or horse power is made available. The module can then be used to determine the best re-allocation of the lift gas to the remaining wells, while taking into account any production constraints, to optimize the total production.

PIPESIM 2000

22

Field Equipment

To allow the day-to-day modeling of the system to be performed quickly, modeling of the wells and the optimization process have been separated. This allows answers to specific problems, by examining a number of scenarios, to be generated in a very short time. Input is taken from individual well performance models created from a multiphase flow simulator, in the form of well performance curves. These performance curves should be generated and checked before being included in the model. To obtain the correct solution the pressure drop must be correctly accounted for along the surface network. This is simulated by the use of (tuned) industrial standard multiphase flow correlation's to predict the pressure loss and liquid hold-up in the pipeline. In its production prediction mode of operation it can be used to validate the individual well gas lift or ESP lift performance curves by using them to predict current production rates. Results are displayed in tabular form, graphical plots or by utilizing the sophisticated graphical user interface to display a variety of rates and pressures. The solution provides a comprehensive report that includes the required gas injection rate for each well or required operating speed for each well, the flow rate and pressure at each manifold in the system and economic data. Full features of the model include: · interfaces with the well Analysis module · solves multi-well commingled scenarios · allows well production performance modeling · offers operator decision support functions · Black Oil only 1.3.1.5 Multi-lateral wells (HoSim) HoSim is designed to model horizontal and multilateral heterogeneous wells in detail. The software uses a rigorous network solution algorithm to solve horizontal and multilateral wells as gathering networks.

PIPESIM 2000

Field Equipment

23

The program enables detailed horizontal well models to be built quickly and easily through a graphical user interface. The user can define various IPR relationships, and specify a detailed well description. Certain equipment models, which are common to the pipeline and facilities module, are available such as chokes, gas lift, ESP's and also separators, compressors, pumps etc. Fluid description can be either black oil or compositional and different fluids can be specified which are mixed together using appropriate mixing rules. Specifying either an outlet pressure or an outlet flow rate (or a range of values for a batch run) to run the model. Results can be displayed either as text (point values) or graphically for any part of the model. 1.3.1.6 Field Planning (FPT) Allows the network module to be coupled to a "reservoir model" to model reservoir behavior over time. In addition conditional logic decision can be taken into account, i.e. bring well 56 on steam in year 5, etc. The reservoir may be described as either; · Black oil tank model · Compositional tank model · look-up tables · Commercial reservoir simulator · Commercial material balance program 1.3.2 Options In addition to the above basic modules a number of options are available. 1.3.2.1 Compositional option Allows a PVT package to be used to determine the fluid properties. Options are · SIS Flash (provided by Schlumberger) · Multiflash (provided by InfoChem)

PIPESIM 2000

24

Field Equipment · SPPTS (for Shell users only)

The compositional options have the following features; · Standard library of 50+ components · Petroleum Fraction · Phase envelope generation · Dew point line · Bubble point line · Critical point · Hydrate formation line (if present) · Ice formation line (if present) · Quality lines · EOS · Peng-Robinson (standard and advanced) · SRK (standard and advanced) · Corresponding EOS · SMIRK (limited access) · Stand alone flash (PT, PH, etc) details · Viscosity models · Pederson · LBC In addition the Multiflash option has the following features; · Multiple Bubble point matching · Multiple Dew point matching · Multiple Viscosity data matching · Multi-stage flashing · Setting of BIPs · Emulsion options · User defined BIPs 1.3.2.2 OLGAS 2000 Utilizes the steady-state version of the multiphase flow correlation from Scandpower as used in OLGA Transient. This option has 2 versions; (i) 2-phase and (ii) 3-phase.

PIPESIM 2000

Field Equipment

25

1.3.2.3 ECLIPSE 100 Allows the Field Planning module to use the ECLIPSE 100 (Black Oil) reservoir simulator to model the reservoir performance. The system has been designed so that ECLIPSE can reside on UNIX or PC. 1.3.2.4 ECLIPSE 300 Allows the Field Planning module to use the ECLIPSE 300 reservoir simulator (Compositional) to model the reservoir performance. The system has been designed so that ECLIPSE can reside on UNIX or PC. 1.3.2.5 MBAL Allows the Field Planning module to use the material balance program Mbal (from Petroleum Experts) to model the reservoir performance. 1.4 File Management PIPESIM uses the following to store data; · ASCII files · Binary files · Microsoft Access Database. Input data (*.BPS, *.BPN, *.PGW, *.FPT,*.HSM) Contains all the data that is necessary to run a model. This includes data for; units, fluid composition, well IPR, system data, etc. The support team requires these files when support queries are made. Output data (*.OUT, *.SUM) Contains program output data in different formats. Transfer files (*.PLT, *.PLC, *.PWH, *.PBT, *.TNT, *.PST) Files that transfer data from one PIPESIM module to another. PVT table (*.PVT) A file that contains a single stream composition and a table of fluid properties for a given set of pressure and temperature values. This file can (if required) be created by a commercial PVT package e.g. Multiflash, Hysys, PVTSim, EQUI90, etc. or via the compositional module in PIPESIM.

PIPESIM 2000

26

Field Equipment

Database files (*.MDB) Microsoft Access Database file that contains; · Black Oil fluid data, · ESP performance curves · User defined pump and compressor curves Units file (*.UMF) Units files. Used to store user defined unit sets. These files can be passed from user-to-user. 1.5 Security Stand-alone (single PC) versions of PIPESIM are protected from unauthorized use by means of either a license file or a hardware security module (generally referred to as a 'dongle' or 'bit lock'). Local Area Network (LAN) versions are normally protected via License Manager software. 1.5.1 Stand-alone security (dongle) When the program executes the dongle must be attached to the parallel port of the computer otherwise it will not run. The dongle remains the property of Schlumberger while in use by customers, and are not replaceable if lost. You can connect another device (or more Schlumberger dongles) to the parallel port while the dongle is still attached to it without affecting the operation of the device or the dongle. Do this simply by plugging the device into the back of the dongle. If you already have another program protected by a similar dongle, they can both be plugged into the port at the same time, and should not interfere with each other. The dongle is quite robust, so no particular care need be taken in handling it. Users are able to view the Schlumberger software modules licensed on their dongles by using the Dongle Utility. On start-up of the utility, the attached dongle license details for the various software modules are displayed. When renewing or purchasing additional software licenses you will need to update the licenses on your dongle(s) by receiving instructions from Schlumberger.

PIPESIM 2000

Field Equipment

27

The dongles have an internal timing mechanism to enforce the license periods. It is important NOT to set your PC's clock into the future and run PIPESIM, as the dongle will prevent you from using PIPESIM after you have set your clock back. If you do accidentally do this, contact Schlumberger for information on how to "reset" your dongle.

28

Field Equipment

1.6 New features You are advised to review the Release Notes document supplied with your version of the software for a complete list of new features. 1.7 Schlumberger Support Services Schlumberger offers full technical support for PIPESIM from our offices worldwide. Please see the web site for your nearest support center or contact the support centers in the United Kingdom or in Houston (USA).

Center United Kingdom mailto:[email protected] America [email protected] Tel +44 1293 55 68 97 +1 713 513 2037

To offer the best and fastest support our preferred method for support services is via email. 1.8 What to do next Depending upon your needs the following is recommended; New users · Familiarize yourself with the all PIPESIM modules, their function and application. · Work through the case studies for your particular area of interest Existing users · Read the Release Notes document to obtain an overview of new features.

PIPESIM 2000

Field Equipment THIS PAGE LEFT BLANK INTENTIONALLY

29

PIPESIM 2000

Model Overview 2 Model Overview

31

2.1 Steps in building a model The steps involved in building a PIPESIM model are slightly different for each module but follow the same basic steps. · Select units · Set fluid data · Calibrate data (optional) · Define components in the model · Well components (completion, tubing) · Pipeline component · Field equipment · Set heat transfer options · Select multiphase flow correlation · Perform an operation · Analyze the results · Graphical · Tabular · Via schematic 2.2 Starting PIPESIM The PIPESIM GUI can be run from the start menu <start/program files/Schlumberger/PIPESIM>. 2.3 Units System The built in units system allows you the flexibility to select any variable and define the unit of measurement to be used. Thus you can use this feature to modify the units system to match reports or data supplied by a service company or to simply customize the units system to suit your own personal preferences. Two non-customizable unit sets are provided; · Engineering (oil field) and · SI. In addition the customizable unit sets are available. Any number of customized unit sets can be created and saved (each one to a different external data file) under a new name. These customized files can be provided to other PIPESIM users.

PIPESIM

32

Model Overview

The units system used for any particular model is saved with the model data, thus allowing models to be moved easily. Any unit set can be set as the default for new models or new sessions of PIPESIM. 2.4 Fluid data One of the first things that you need to do before using PIPESIM is to decide what type of fluid system you are going to use. PIPESIM can model the following fluid types: · Gas · Gas condensate · Liquid · Liquid & Gas · Steam The fluid can be described by one of the following methods; · Fully Compositional · Black Oil correlations · Steam tables The fluid model that you use will depend upon: · Properties of the fluids in the system · Flow rates and conditions (pressure & temperature) at which the fluid(s) enter and leave the system. · Available data, etc. For a quick screening study where the accuracy of the physical properties is not essential, we advise the user to use a Black oil fluid model specification. 2.4.1 Black Oil Black oil fluid modeling utilizes correlation models to simulate the key PVT fluid properties of the oil/gas/water system. These empirical correlation's treat the oil/gas system as a simple two component system - unlike the more rigorous multi-component compositional model methods. The hydrocarbon is treated simply as a liquid

PIPESIM 2000

Model Overview component (if present) and a gas component related to stock tank conditions. All that is needed for most applications is a minimum of production data, oil gravity, gas gravity, solution gas/oil ratio and, if water is also present in the system, the watercut. Black oil fluid modeling is appropriate for use with a wide range of applications and hydrocarbon fluid systems. In general, the basic black oil correlations will provide reasonable accuracy in most PVT fluid property evaluations over the range of pressures and temperatures likely to be found in production or pipeline systems.

33

However, care should be taken when applying the black oil approach to a highly volatile crude or a condensate where accurate modeling of the gaseous light ends is required. In this case, the user should consider the use of compositional modeling technique that describes the fluid as a multi-component mixture. In order to increase the accuracy of the basic black oil correlations for modeling multiphase flow, PIPESIM provides the facility to adjust salient values of a number of the most important PVT fluid properties to match laboratory data. These PVT fluid properties are considered the single most important parameters affecting the accuracy of multi-phase flow calculations. Calibration of these properties can greatly increase the accuracy of the correlations over the range of pressures and temperatures for the system being modeled. This facility is optional, but the above calibrations will significantly improve the accuracy of the predicted gas/liquid ratio, the flowing oil density and the oil volume formation factor. If the calibration data is omitted, however, PIPESIM will calibrate on the basis of oil and gas gravity alone and thus, there will be a loss in accuracy. It should be noted that the black oil calibration feature is only applicable to oil fluid types, as it is not appropriate for a gas fluid type. The following blackoil correlations are available: · Solution gas and bubble point pressure: Lasater, Standing, Vasquez and Beggs, Kartamodjo, Khan, or Glasø.

PIPESIM

34

Model Overview · Oil formation volume factor of saturated systems: Standing, Vasquez and Beggs, or Glasø. · Oil formation volume factor of undersaturated systems: Vasquez and Beggs, or Glasø. · Dead oil viscosity: Beggs and Robinson, Glasø, or User's data. · Live oil viscosity of saturated systems: Chew and Connally or Beggs and Robinson. · Live oil viscosity of undersaturated systems: Vazquez and Beggs, Kousel, or None. · Viscosity of oil/water mixtures: Inversion, Volume Ratio, or Woelflin. · Gas viscosity: Lee et al. · Gas compressibility: Standing, or Hall and Yarborough.

2.4.2 Compositional For compositional fluid modeling of hydrocarbon fluids and associated gas and water components, PIPESIM uses a PVT modeling package. Compositional fluid modeling is generally regarded as more accurate, but also more expensive in terms of time and computer resources than black oil modeling. It is justified for problems involving volatile fluids needing rigorous heat transfer calculations. However, the black oil modeling approach can often give satisfactory results with volatile fluids. Oil systems contain in reality many thousands of pure components, consisting of a spectrum of molecules with different carbon numbers and large numbers of different isomers. It would be impossible to model the behavior of such systems by explicitly defining the amount of each of these molecules, both because of the excessive computing power needed and the fact that laboratory reports could not possibly supply all this information. Since the alkane hydrocarbons are non-polar and therefore mutually relatively ideal, lumping them together in the form of a number of 'pseudo-components' results in fairly accurate phase behavior and physical property predictions.

PIPESIM 2000

Model Overview

35

Petroleum fractions are normally defined by splitting off sections of a laboratory distillation of the C7+ mixture. Curves of boiling point, density and molecular weight are produced from which the properties of the individual pseudo-components may be derived. Petroleum fractions are characterized by either; · Measured Properties; · boiling point (BP), · specific gravity (SG) and · molecular weight (MW). T · Critical Property · critical temperature (TC), · critical pressure (PC), · acentric factor (Omega) and · specific gravity (SG). Further details of the equations used, etc can be found in the PIPESIM help system. 2.4.3 Steam For steam systems (production and injection) PIPESIM uses the GPSA stream tables. When modeling stream systems the pressure and quality are required. If the quality is superheated (quality =100%) or sub-cooled (quality=0%) then the temperature is also required. 2.5 Model components overview A PIPESIM model is built (via the GUI) by adding components (from the toolbox) to the model window. Components are divided into 2 groups; · Node type components · Boundary nodes - Must be on the edge of the system and can only have one connection either leaving (source) or entering (sink). · Internal nodes - Cannot be on the edge of the system and can have any number of connections. · Linking type components - Joins 2 node type components

PIPESIM

36

Model Overview

Node type components are connected by linking components and thus must be added to the model first. The components available depend upon the modules purchased. Details on the inputs for each component can be found in the help system. A full list of components and their type is listed below. Pipeline & facilities module Component Type Description Source Boundary The point where the fluid enters the Node system. Flowline Link A flowline to a point where it meets another flowline (with different characteristics) or another object. Maybe horizontal or inclined and surrounded by air, water or both; insulated or bare Riser Link A description of the riser (vertical or near-vertical - up or down) to a point where it meets another riser or another object. Pump Internal A single or multistage pump for the Node pumping of liquids. Multiphase Node A multiphase booster. Booster Separator Internal Allows fluid separation to take place in Node the model. It is a two-phase separator, (i.e. gross liquids, water or gas). The removed fluid can be re-injected back into the network model via the injection point component. A single or multistage centrifugal gas compressor An expander.

Compressor Expander

Internal Node Internal Node

PIPESIM 2000

Model Overview Heat exchanger Choke Generic Equipment Injection point Internal Node Internal Node Internal Node Internal Node Allows a change in temperature and pressure to be modeled A device to restrict the flow of fluids.

37

Multiplier/Adder Spot report

Internal Node Internal Node

Keyword tool

Internal Node Link

Connector

A general device that can alter the pressure or temperature. Allows a side stream (compositional only) to be injected into the main stream. The incoming pressure and flowrate (along with the composition) are required. Changes the flowrate by the amount specified. Allow key pieces of information to be retrieved at any point (between links) in the system. This component has no effect on the temperature or pressure in the system. Allows engine keywords to be inserted into a model. A full list of the keywords can be found in the Help system under keyword reference. Joins to nodes without having any effect on the calculations, i.e. a zero length piece of pipe.

Well Performance module Component Type Description Vertical Boundary Describes the well IPR and the completion Node reservoir static pressure for a vertical completion. These are then used to determine the bottom hole pressure. Horizontal Boundary Describes the horizontal completion, completion Node the IPR and the reservoir static pressure. These are then used to determine the bottom hole (heal) pressure Tubing Link Joins the reservoir top the surface. The fluid can flow either through the tubing

PIPESIM

38

Model Overview or outside the tubing (inside the casing) or both. The tubing may also have down hole equipment installed. The point in the system where the (nodal) analysis is to be conducted. The model is then broken into two parts; inflow to the NA point and outflow from the NA point.

Nodal analysis point

Node

Network module Component Type Description Production well Boundary Models the source as a production well. Node The well is (normally) defined from the sand face to the point where it joins another object, i.e. well head, manifold, etc. Generic source Boundary The point where a fluid enters the Node system. Can be used when a well is modeled from the well head. Injection well Boundary Models the sink as an injection well, Node including tubing and completion. Generic sink Boundary The point where the fluid leaves the Node systems. A model may have any number of sinks. Node Node A point in the system where 1 or more branches meets Branch Link Connects 2 or more nodes, sources or sinks. Any combination of flowline, riser or pieces of equipment can be used to describe a branch. When connected between a well and a node the resulting branch has no physical meaning Re-injection Node Connects 3 branches; node 1 - the incoming fluid stream 2 - the outlet stream 3 - the stream removed by the separator. All the fluid removed from the separator is re-injected. The re-

PIPESIM 2000

Model Overview injected stream can be upstream or downstream of the separator. 2.5.1 Model & Component limitations The following limitations; General: · Maximum number of components in a stream: Pipeline & facilities · Maximum number of sources: · Maximum number of sinks: · Maximum number pipe coatings: · Maximum number of nodes for a pipeline or riser: Well Performance · Maximum number of completions: · Maximum number of sinks · Maximum number tubing coatings: · Maximum number of nodes for a tubing: · Maximum number of geothermal survey points: · Maximum number of tubing strings: · Detailed model: · Simple model: Network · Maximum number of wells / branches: · Maximum number of nodes: · Maximum number of PVT files: · Maximum number of compositions: · Maximum number of Black Oil compositions: · Maximum number of PQ data points: Field Planning · Maximum number of stored timesteps: · Maximum number of auxiliary properties: · Maximum number of Eclipse models: · Maximum number of network models: 50 1 1 4 101 10 1 10 100 100 20 4 unlimited unlimited 500 1,000 1,024 30 256 1,500 1 5

39

PIPESIM

40 · · · · ·

Model Overview Maximum number of events: Maximum number of schedule 'bean' lists: Maximum number of look-up tables: Maximum number of data lines in all look-up tables: Maximum number of tank reservoirs: 2,500 99 500 1500 50 500 400 1 500

Production Optimization (GOAL) · Maximum number of wells/branches: · Maximum number of nodes: · Maximum number of sinks: Multi-lateral (HoSim) · Maximum number of multi-laterals:

2.6 Flow correlation Flow correlations are used to determine the pressure drop and holdup in the system Flow correlations are split in to the following section; · Single phase · Multiphase - vertical · Multiphase - horizontal A number of flow correlations have been proposed over the years. In addition to the standard supplied flow correlations user's can create and add their own multiphase flow correlation in to PIPESIM via the user DLL facility. The linkages are documented in the user defined flow correlations document, which can be obtained from Schlumberger or down loaded from our web site. 2.7 Run an operation Select the operation that is relevant to the model developed. The simulation will commence and the post-processor can then be used to analyze the results.

PIPESIM 2000

Model Overview

41

2.8 Saving & Closing PIPESIM When PIPESIM is closed all files (models) that have been modified during the session are checked and an option to save any that have changed is presented to the user. 2.9 How to build models This section provides a brief overview of the steps involved in building a model with each of the basic PIPESIM modules. See the PIPESIM Help system " How do I..." section for full details on setting up the basic models. PIPESIM can build the following basic models; · Pipeline and facilities · Production well · Single completion well · Multiple completion well · Horizontal completion well · Injection well · Sub-surface and surface Networks · Gathering systems · Looped systems · Distribution systems · Multi-lateral wells · Production · Injection 2.9.1 Fluid calibration 2.9.1.1 Black Oil The following basic steps are required to calibrate the black oil defined fluids; · Select the units set of your preference · Enter the basic fluid data · Enter the Bubble Point data · Enter the Advanced calibration data (optional) · Run the operation. · Save the model!

PIPESIM

42

Model Overview

In a network model the calibration data is "mixed" at junctions to provide average calibration data for the resulting stream. 2.9.1.2 Compositional The following basic steps are required to calibrate the compositionally defined fluids; · Select the units set of your preference · Enter the basic fluid data (library components, petroleum fractions) · Produce the phase envelop (for reference) · Select the quantity to match to; Bubble Point or Dew point · Enter the matching data · Select viscosity matching options if applicable · Enter the viscosity data · Run the matching operation · Update the composition · Produce the new phase envelop · Save the model! 2.9.2 Pipeline & facilities The following basic steps are required to build a pipeline & facilities model; · Select the units set of your preference · Add the necessary components to the model (source, flowline, equipment, etc) and defined the necessary data. · Define the fluid specification (black oil or compositional). · Define the flow correlation to use. · Save the model! One the basic model has been developed a number of operations can be performed or the model can be utilized in additional PIPESIM modules. 2.9.2.1 Correlation matching The following basic steps are required to determine the most suitable horizontal multiphase flow correlation; · Build the pipeline & facilities model. · Select the Correlation matching operation · Determine the boundary condition to compute

PIPESIM 2000

Model Overview · · · · Select suitable Horizontal correlations Enter any known measured pressure and temperature values Run the operation. Save the model!

43

Insure that the most suitable correlation is then selected from the horizontal flow correlation list for subsequent simulations. 2.9.2.2 Pressure/Temperature profile The following basic steps are required to determine the pressure or temperature profile along the system; · Build the well performance model. · Select the Pressure/Temperature profile operation · Determine the boundary condition to compute · Select any sensitivity parameters · Enter the sensitivity parameters · Run the operation · Save the model! 2.9.2.3 Equipment/Flowline sizing (1 parameter) The following basic steps are required to size a flowline/riser or a piece of equipment; · Build the pipeline and facilities model. · Include the flowline/equipment/riser to be sized. · Select the Pressure/Temperature profile operation · Select the sensitivity parameter · Enter the data for the sensitivity parameter · Run the operation. · Save the model! 2.9.2.4 Equipment/Flowline sizing (Multiple parameter) The following basic steps are required to size a flowline/riser or a piece of equipment; · Build the pipeline and facilities model. · Include the flowline/equipment/riser to be sized. · Select the System Analysis operation · Select the multiple sensitivity · Select the x-axis and sensitivity parameters

PIPESIM

44

Model Overview · Enter the data for the sensitivity parameters · Decide if the sensitivity parameters are permuted or change in step. · Run the operation. · Save the model!

2.9.2.5 Multiphase booster design The following basic steps are required to complete a multiphase booster design; · Build the pipeline and facilities (including the well if required) model. · Include the multiphase booster. · Perform the analysis (nodal, PT profile, etc) with the booster inactive. · Invoke the generic Multiphase booster option and set the booster parameters. Details on efficiency factors are supplied in the help system. · Re-run the analysis. · Verify that multiphase booster van enhance production. · Decide upon the Multiphase booster type required (Helico Axial or Twin Screw). · For twin screw boosters · Select the generic twin screw module · Enter the required data and re-run the analysis · PIPESIM will automatically select the most suitable size of the twin screw booster. · Select the Twin screw booster module · Select the nominal booster as recommend by the previous operation · Enter the data required data and re-run the analysis · Select the vendor Twin screw module · Enter the data required data and re-run the analysis · For Helico Axial boosters · Enter the required a data and re-run the analysis · Save the model!

PIPESIM 2000

Model Overview

45

2.9.3 Well Performance The following basic steps are required to build a well model (single or multiple completion); · Select the units set of your preference · Determine the completion of the well · Single · Multiple · Horizontal · Add the necessary components to the model (completion, tubing, etc) and defined the necessary data. · Define the fluid specification · Define the flow correlation to use. · Save the model! Once the basic model has been developed a number of operations can be performed or the well model can be utilized in additional PIPESIM modules.

2.9.3.1 Correlation matching The following basic steps are required to determine the most suitable vertical multiphase flow correlation; · Build the well o.aTm(well dules. )TjETEMC/P <</MMCID 7 >>BDCBT/C2_1 1 T

46

Model Overview · Determine the inflow and outflow parameters. · Run the operation. · Save the model!

2.9.3.3 Pressure/Temperature profile The following basic steps are required to determine the pressure or temperature profile along the system; · Build the well performance model. · Select the Pressure/Temperature profile operation · Determine the boundary condition to compute · Select any sensitivity parameters · Enter the sensitivity parameters · Run the operation · Save the model! 2.9.3.4 Equipment/Tubing sizing (1 parameter) The following basic steps are required to size tubing or a piece of equipment; · Build the well model. · Include the tubing/equipment to be sized. · Select the Pressure/Temperature profile operation · Select the sensitivity parameter · Enter the data for the sensitivity parameter · Run the operation. · Save the model! 2.9.3.5 Equipment/Tubing sizing (Multiple parameter) The following basic steps are required to size tubing or a piece of equipment; · Build the pipeline and facilities model. · Include the tubing/equipment to be sized. · Select the System Analysis operation · Select the multiple sensitivity · Select the x-axis and sensitivity parameters · Enter the data for the sensitivity parameters · Decide if the sensitivity parameters are permuted or change in step. · Run the operation.

PIPESIM 2000

Model Overview · Save the model! 2.9.3.6 Artificial Lift analysis The following basic steps are required to analysis the effects of artificial lift on a well; · Build the well performance model. · Insure that the gas lift or ESP lift depth has been set. · Select the Artificial Lift operation · Select the sensitivity parameters · Run the operation · Save the model! 2.9.3.7 Well performance curves for GOAL The following basic steps are required to create well performance curves for the Optimization module (GOAL); · Build the well performance model. · Insure that the gas lift or ESP lift depth has been set. · Select the Artificial Lift operation · Select the GOAL curve format · Enter the required data · Run the operation. · Save the model!

47

The resulting data transfer files (*.PLT & *.PWH) are required by the optimization model. These files must then be transferred (manually) to the required optimization (GOAL) directory. 2.9.3.8Well performance curves for Network Solver The following basic steps are required to create well performance curves for the Network module (GOAL); · Build the well performance model. · Select the Well Performance operation · Select the sensitivity parameters · Enter the required data · Run the operation · Save the model! The resulting data transfer files (*.WPI) are required by the network model if the well is to be represented by a performance curve. These

PIPESIM

48

Model Overview

files must then be transferred (manually) to the required network directory. 2.9.3.9 Reservoir Tables The following basic steps are required to create reservoir look-up tables; · Build the well performance model. · Select the reservoir tables operation · Select the reservoir simulator · Enter the required data · Run the operation. · Save the model! The resulting ASCII file can then be used directly by the reservoir simulator. 2.9.3.10 Horizontal completion length The following basic steps are required to determine the optimal horizontal completion length; · Build the well (horizontal) performance model. · Select the Horizontal completion length operation · Enter the required data · Run the operation. · Save the model! 2.9.3.11 Gas Lift Rate v's Casing head pressure The following basic steps are required to analysis the effects of gas lift rate on the casing head pressure for a well; · Build the well performance model. · Insure that the gas lift depth and quantity has been set. · Select the Gas Lift rate v's casing head pressure operation · Select the sensitivity parameters · Run the operation · Save the model! 2.9.4 Network Analysis 2.9.4.1 Fluid properties In a network model different fluid descriptions can not be used, i.e. the model must be either black oil, compositional or steam.

PIPESIM 2000

Model Overview

49

Each source can have it's own fluid description or use shared data. 2.9.4.2 Boundary Conditions In order to solve the network model the correct number of boundary conditions must be entered. Boundary nodes are those that have only one connecting branch, e.g. production well, injection well, source and sink. The number of boundary conditions that are required for a model is known as the models Degrees of Freedom. This is computed by the total number of boundary nodes, i.e. number of well (production and injection) + number of sources + number of sinks. For example a 3 production well system producing fluid to a single delivery point has 4 degrees of freedom (3+1) regardless of the network configuration between the well and the sink. Each boundary can be specified in terms of; · Pressure · Flowrate OR · Pressure/Flowrate (PQ) curve. To enable the system to be solved 1: the number of Pressure, flowrate or PQ specifications must equal the degrees of freedom of the model. 2: At least 1 pressure must be specified 3: All each source (production well & source) the fluid temperature must be set. For example the above 3 well / 1 sink model could be specified as; · Well 1: Reservoir pressure, reservoir temperature · Well 2: Reservoir pressure, reservoir temperature · Well 3: Reservoir pressure, reservoir temperature · Sink: Delivery pressure OR · Well 1: Reservoir pressure, Flowrate, reservoir temperature · Well 2: reservoir temperature

PIPESIM

50

Model Overview · Well 3: Reservoir pressure, reservoir temperature · Sink: Delivery pressure

OR · · · · Etc. 2.9.4.3 Network model The following basic steps are required to build a network model; · Select the units set of your preference · Develop the network model (wells and surface facilities). Prebuilt models of wells/flowline can be used. · Set the fluid properties · Set the boundary conditions · Save the model! 2.9.5 Production Optimization The following basic steps are required to build an optimization c (GOAL) model; · Select the units set of your preference · Develop the surface network model · Set the outlet pressure · Develop individual well models · Create well performance curves for each well · Save the model! See the GOAL Used Guide for details on; · building an optimization model · Calibrating the surface network · Calibrating the individual well models · Optimizing the field · Applying field constraints 2.9.6 Field Well 1: Flowrate, reservoir temperature Well 2: Flowrate, reservoir temperature Well 3: Flowrate, reservoir temperature Sink: Delivery pressure

Model Overview

51

· · · · · · · ·

· Tanks · Tables · Reservoir simulator · Set the name of the host UNIX workstation · Material balance program Develop the network model (well and surface network) or models. Link the wells to the reservoir description. Specify any flowrate constraints Define the time dependent events. Define the conditional based events. Select any auxiliary properties that are to be stored during the simulation and analyzed in the post-processor. Set the convergence tolerance Save the model!

See the FPT Used Guide for an example of building a Field Planning model. 2.9.7 Multi-lateral The following basic steps are required to build a multi-lateral well model; · Select the units set of your preference · Add the necessary components to the model (horizontal well section, branch, etc) and defined the necessary data. · Define the fluid specification (black oil or compositional). · Define the flow correlation to use. · Save the model! See the HoSim Used Guide for an example of building a multi-lateral well model.

PIPESIM

52

Model Overview

3 Fluid & Multiphase Flow Modeling This section defines the fluid models and flow correlation modeled available in PIPESIM. 3.1 Black Oil Fluid properties can be predicted by black-oil correlations that have been developed by correlating gas/oil ratios for live crude's with various properties, such as oil and gas gravities. The selected correlation is used to predict the quantity of gas dissolved in the oil at a particular pressure and temperature. The black oil correlations have been developed specifically for crude oil/gas/water systems and are therefore most useful in predicting the phase behavior of crude oil well streams. When used in conjunction with the calibration options, the black oil correlations can produce accurate phase behavior data from a minimum of input data. They are particularly convenient in gas lift studies where the effects of varying GLR and water cut are under investigation. However, if the accurate phase behavior prediction of light hydrocarbon systems is important, it is recommended that the more rigorous compositional models is employed. 3.1.1 Lasater A correlation developed in 1958 from 158 experimental data points. The data points spanned the following ranges: pb (bubble point pressure): 48 to 5,780 psia TR (reservoir temperature): 82 to 272 °F g API (API gravity): 17.9 to 51.1 °API g g (gas specific gravity): 0.574 to 1.223 Rsb (solution gas at bubble point pressure): 3 to 2,905 scf/STB 3.1.1.1 Bubble point pressure Step 1: Calculate Mo (molecular weight of the stock tank oil) For API <= 40: Mo = 630 - 10g API For API > 40: Mo = 73,110(g API)-1.562 Step 2: Calculate yg (mol fraction of gas) yg = (Rsb/379.3)/(Rsb/379.3 + 350g o/Mo) where g o = oil specific gravity Step 3: Calculate the bubble point pressure factor (pbg g/TR)

PIPESIM 2000

Fluid & Multiphase Modeling For yg <= 0.6: pbg g/TR = 0.679 exp(2.786yg) - 0.323 For yg > 0.6: pbg g/TR = 8.26yg3.56 + 1.95 Step 4: Calculate pb pb = (pbg g/TR )(T/g g) 3.1.1.2 Solution gas Rs = 132755 g o yg/(Mo(1 - yg))

53

3.1.2 Standing Standing presented an equation to estimate bubble point pressures greater than 1,000 psia. The correlation was based on 105 experimentally determined bubble point pressure of California oil systems. The data points spanned the following ranges: pb (bubble point pressure): 130 to 7,000 psia TR (reservoir temperature): 100 to 258 °F gAPI (API gravity): 16.5 to 63.8 °API g g (gas specific gravity): 0.59 to 0.95 Rsb (solution gas at bubble point pressure): 20 to 1,425 scf/STB 3.1.2.1 Bubble point pressure Step 1: Calculate yg (mol fraction of gas) yg = 0.00091TR - 0.0125g API Step 2: Calculate pb pb = 18(Rsb/g g)0.83 x 10yg 3.1.2.2 Solution gas Rs = g g (p/(18 x 10yg))1.204 3.1.2.3 Oil formation volume factor - saturated systems Step 1: Calculate F (correlating factor) F = Rs (g g /g o)0.5 + 1.25T Step 2: Calculate Bo (oil formation volume factor in bbl/STB) Bo = 0.972 + 0.000147F1.175 3.1.3 Vazques and Beggs Vasquez and Beggs used results from more than 600 oil systems to develop empirical correlations for several oil properties including bubble point pressure.

PIPESIM

54

Fluid & Multiphase Modeling

Approximately 6,000 measured data points were collected across the following ranges: pb (bubble point pressure): 50 to 5,250 psia TR (reservoir temperature): 70 to 295 °F g API (API gravity): 16 to 58 °API g g (gas specific gravity): 0.56 to 1.18 Rsb (solution gas at bubble point pressure): 20 to 2,070 scf/STB 3.1.3.1 Bubble point pressure pb = (Rsb/(C1g g exp(C3g API/( TR + 460))))1/C2 where for g API <= 30: C1 = 0.0362, C2 = 1.0937, C3 = 25.724 g API > 30: C1 = 0.0178, C2 = 1.187, C3 = 23.931 3.1.3.2 Solution gas Rs = C1 g g pC2 exp((C3 g API )/(T + 460)) where for g API <= 30: C1 = 0.0362, C2 = 1.0937, C3 = 25.724 g API > 30: C1 = 0.0178, C2 = 1.187, C3 = 23.931 3.1.3.3 Oil formation volume factor - saturated systems Bo = 1 + C1 Rs + C2 (T - 60)(g API/g gc) + C3 Rs (T - 60)(g API/g gc) where for g API <= 30: C1 = 4.677e-4, C2 = 1.751e-5, C3 = -1.811e-8 g API > 30: C1 = 4.67e-4, C2 = 1.1e-5, C3 = 1.337e-9 3.1.3.4 Oil formation volume factor - undersaturated systems Bo = Bob exp(co (pb - p)) 3.1.4 Glasø Glasø developed PVT correlations from analysis of crude oil from the following North Sea Fields:Ekofisk Stratfjord Forties Valhall COD 30/7-2A

PIPESIM

Fluid & Multiphase Modeling 3.1.4.1 Bubble point pressure and solution gas pb = f 1 [(Rs /g g )0.816 (T 0.172/g API 0.989)] 3.1.4.2 Oil formation volume factor - saturated systems Bob = f 2 [Rs (g g/g o)0.526 + 0.968T] 3.1.4.3 Oil formation volume factor - undersaturated systems Bt = f 3 [Rs (T 0.5 /g g0.3) g oA p-1.1089] Where A = 2.9 x 10-0.00027Rs

55

3.1.5 Coning In order to simulate gas and/or water breakthrough from the reservoir, flowrate-dependent values of GOR and watercut may be entered. In a homogeneous reservoir, analysis of the radial flow behavior of reservoir fluids moving towards a producing well shows that the rate dependent phenomenon of coning may be important. The effect of increasing fluid velocity and energy loss in the vicinity of a well leads to the local distortion of a gas-oil contact or a water-oil contact. The gas and water in the vicinity of the producing wellbore can therefore flow towards the perforation. The relative permeability to oil in the pore spaces around the wellbore decreases as gas and water saturation increase. The local saturations can be significantly different from the bulk average saturations (at distances such as a few hundred meters from the wellbore). The prediction of coning is important since it leads to decisions regarding: · Preferred initial completions · Estimation of cone arrival time at a producing well · Prediction of fluid production rates after cone arrival · Design of preferred well spacing

PIPESIM

56

Fluid & Multiphase Modeling

3.1.6 Liquid Viscosity There are four steps to calculating the liquid viscosity as follows: 1 Calculate the dead oil viscosity at atmospheric pressure and the flowing fluid temperature. The methods available for calculating dead oil viscosity are: Beggs and Robinson, Glasø method, or Users data. Calculate the saturated live oil viscosity at the flowing fluid pressure and temperature assuming that the oil is saturated with dissolved gas. The methods available for calculating live oil viscosity

2

Fluid & Multiphase Modeling

57

3.1.7.1 Beggs and Robinson method Beggs and Robinson used results from 600 oil systems to develop relationships for dead and live oil viscosity. 460 dead oil observations and 2,073 live oil observations were taken. The range of data analyzed was as follows: p (pressure): 50 to 5,250 psia T (temperature): 70 to 295 °F g API (API gravity): 16 to 58 °API Rsb (solution gas at bubble point pressure): 20 to 2,070 scf/STB Dead oil viscosity is calculated as follows: m OD = 10x - 1 where x = yT-1.163 y = 10z z = 3.0324 - 0.02023 gAPI 3.1.7.2 Glasø method Dead oil viscosity is calculated as follows: mOD = c(loggAPI)d where c = 3.141(1010 )T-3.444 d = 10.313(logT) - 36.447 3.1.7.3 User's data method A curve is fitted through the supplied data points of the following form: Log(mOD) µ (1/T) 3.1.8 Live Oil Viscosity The following live Oil Viscosity methods are available · Chew and Connally · Beggs and Robinson

PIPESIM

58

Fluid & Multiphase Modeling

3.1.8.1 Chew and Connally Chew and Connally used results from 457 oil systems to develop relationships for live oil viscosity. The range of data analyzed was as follows:p (pressure): 132 to 5,645 psia T (temperature): 72 to 292 °F Rsb (solution gas at bubble point pressure): 51 to 3,544 scf/STB Live oil viscosity is calculated as follows:mOb = AmODB where A and B are given by the following table: Rs (cu ft/bbl) A 0 1.000 50 0.898 100 0.820 200 0.703 300 0.621 400 0.550 600 0.447 800 0.373 1,000 0.312 1,200 0.273 1,400 0.251 1,600 0.234 3.1.8.2 Beggs and Robinson Live oil viscosity is calculated as follows: mOb = AmODB where A = 10.715(Rs + 100)- 0.515 B = 5.44(Rs + 150)- 0.338 3.1.9 Undersaturated Oil Viscosity 3.1.9.1 Vasquez and Beggs Undersaturated oil viscosity is calculated as follows:m = mOb(p/pb)m

PIPESIM

B 1.000 0.931 0.884 0.811 0.761 0.721 0.660 0.615 0.578 0.548 0.522 0.498

Fluid & Multiphase Modeling where m = 2.6p1.187 exp(-8.98x10-5 p - 11.513) For dead oils at high pressures the Vasquez and Beggs correlation overestimates the viscosity: Use Kousel. 3.1.9.2 Kousel method Undersaturated oil viscosity is derived from the equation Log(mp/ma) = p/1000(A + Bma0.278) Where A and B are parameters entered by the user.

59

Suggested values for A and B are 0.0239 and 0.01638 respectively. m a is the viscosity of the oil at the same temperature and atmospheric pressure. 3.1.9.3 No calculation The undersaturated oil viscosity is assumed to be the same as the saturated live oil viscosity at the same temperature and pressure. 3.1.10 Oil/Water Mixture Viscosity 3.1.10.1 Inversion method The inversion method assumes that the continuous phase changes from oil to water at a given watercut cutoff point. This means that, at a watercut below or equal to the cut-off value, water bubbles are carried by oil, and the mixture assumes the same viscosity as that of the oil. At a watercut above the cut-off value, oil bubbles are carried by water, and the mixture assumes the same viscosity as that of the water. 3.1.10.2 Volume ratio method Mixture viscosity is calculated as follows mm = mO Vo + mw Vw where mO = oil viscosity Vo = volume fraction of oil mw = water viscosity Vw= volume fraction of water

PIPESIM

60

Fluid & Multiphase Modeling

3.1.10.3 Woelflin method The Woelflin option assumes that the continuous phase changes from emulsion to water at a given watercut cutoff point. This means that, at a watercut below or equal to the cut-off value, an emulsion forms and the emulsion viscosity is given by the Woelflin equation for emulsions. At a watercut above the cut-off value, oil bubbles are carried by water, and the mixture assumes the same viscosity as that of the water. The Woelflin equation is as follows mm = mO (1 + 0.0023 Vw2.2 ) 3.1.11 Gas Viscosity 3.1.11.1 Lee et al. Method Gas viscosity is calculated as follows: mg = Kexp(Xr y) where K = (7.77 + 0.0063M)T1.5/(122.4 + 12.9M + T) X = 2.57 + 1914.5/T + 0.0095M Y = 1.11 + 0.04X M is the gas molecular weight r is the gas density 3.2 Compositional 3.2.1 EOS (Equations of State) Equations of state describe the pressure, volume and temperature behavior of pure components and mixtures. Most thermodynamic and transport properties are derived from the equation of state. The following equations of state are available:· SRK (advanced and standard) · PR (advanced and standard) · SMIRK 3.2.1.1 Soave-Redlich-Kwong The standard SRK equation is; P = (NRT/(V - b)) + (a/(V(V + b)))

PIPESIM

Fluid & Multiphase Modeling

61

The values of "a" and "b" in the above equations are derived from functions of the pure component critical temperatures, pressures, and acentric factors. The advanced implementation of SRK contains additional nonstandard features. These include the ability to match stored values for the liquid density (Peneloux correlation) and the saturated vapor pressure and a choice of mixing rule. 3.2.1.2 Peng-Robinson The standard PR equation is; P = (NRT/(V - b)) + (a/(V2 + 2bV - b2)) The values of "a" and "b" in the above equations are derived from functions of the pure component critical temperatures, pressures, and acentric factors. The advanced implementation of PR contains additional nonstandard features. These include the ability to match stored values for the liquid density (Peneloux correlation) and the saturated vapor pressure and a choice of mixing rule. 3.2.1.3 SMIRK The Shell SPPTS package uses the SMIRK equation of state. 3.2.2 Viscosity model The following methods are available to predict the liquid and gas viscosity; · Pederson · LBC (Lohrenz-Bray-Clark) These are not available when using SMIRK (SPPTS) Preliminary testing has shown the Pedersen method to be the most widely applicable and accurate for oil and gas viscosity predictions. Both methods are based on the corresponding state theory.

PIPESIM

62

Fluid & Multiphase Modeling

The choice of the equation of state has a large effect on the viscosities predicted by both methods. The LBC method is more sensitive to equation of state effects than the Pedersen method. 3.2.2.1 Lower Alkanes Predicted liquid viscosities using LBC and Pedersen methods have been compared to experimental data for Methane and Octane as a function of both temperature and pressure and for Pentane as a function of temperature. For both Methane and Pentane the Pedersen method predictions show close agreement with experimental data. For Octane, the Pedersen and LBC methods give comparable results. For the aromatic compound, Ethyl Benzene, the Pedersen method is not as good as the LBC method. 3.2.2.2 Higher Alkanes The results for higher alkanes Eicosane and Triacontane are mixed: the Pedersen method is adequate for Eicosane whereas the LBC method is slightly better than Pedersen for Triacontane. For Triacontane both LBC and the Pedersen methods are inadequate. However, in the majority of cases the higher hydrocarbons should be treated as petroleum fractions rather than as single named components. 3.2.2.3 Petroleum Fractions The LBC method describes viscosity as a function of the fluid critical parameters, acentric factor and density. The LBC model is therefore very sensitive to both density and the characterization of the petroleum fractions. 3.2.2.4 Water The Pedersen method suffers the same drawback as the LBC method in that it is unable to predict the temperature dependence of water, a polar molecule. To overcome this problem, the Pedersen method has been modified especially for water so that it now accurately models the viscosity of water in the liquid phase. This was achieved by the introduction of a temperature-dependent correction factor. However the prediction of the viscosity of the gas phase is also affected, though in only a minor way.

PIPESIM

Fluid & Multiphase Modeling

63

3.2.2.5 Methanol Neither the LBC nor the Pederson method can deal with polar components with the Pederson method slightly worse than the LBC method. This is not surprising, as both methods were developed for non-polar components and mixtures. The Pedersen method works best with light alkanes and petroleum mixtures in the liquid phase. It performs as well or better than the LBC method in nearly all situations. 3.2.2.6 Emulsion The following options are available for handing emulsions; · Inversion method · Volume ratio method · Woelflin method The methods are as described for Black Oil emulsions. 3.2.3 BIP (Binary Interaction Parameter) Set Binary Interaction parameters (BIPs) are adjustable factors which Are used to alter the predictions from a model until these reproduce as closely as possible the experimental data. BIPs apply between pairs of components. The SRK and PR EOS (being cubic equations of state) require only a single BIP, kij, in the model description. The closer the binary system to ideality the smaller the size of kij, which will be zero for ideal systems. It is unlikely that the value of kij will be greater than 1, although it is possible for it to be negative. 3.2.4 Hydrates Natural gas hydrates are solid ice-like compounds of water and light components of natural gas. They form at temperatures above the ice point and are therefore a serious concern in oil and gas processing operations. The phase behavior of the systems involving hydrates can be very complex because up to six phases must normally be considered. The behavior is particularly complex if there is significant mutual solubility between phases. The hydrate model uses a modification of the RKS equation of state for the fluid phases plus The van der Waals and Platteeuw model for the hydrate phases. The model can explicitly represent all the effects of the presence of inhibitors.

PIPESIM

64

Fluid & Multiphase Modeling

Note: you must explicitly include water in the mixture if you wish to do hydrate calculations. The amount of water may influence the results of the calculations, particularly when inhibitors or watersoluble gases are present. The main features of the model are: · The description of the hydrate phase behavior uses a thermodynamically consistent set of models for all phases. · The vapor pressures of pure water are reproduced. The following natural gas hydrate formers are included: METHANE, ETHANE, PROPANE, ISOBUTANE, BUTANE, NITROGEN, CO2 AND H2S. The thermal properties (enthalpies and entropies) of the hydrates are included, permitting flashes involving these phases. The properties of the hydrates have been fixed by investigating data for natural gas components in both simple and mixed hydrates to obtain reliable predictions of both structure I and structure II hydrates. The properties of the empty hydrate lattices have been investigated and the most reliable recent values have been adopted. Proper allowance has been made for the solubilities of the gases in water so that the model parameters are not distorted by this effect. This is particularly important for Carbon Dioxide and Hydrogen Sulphide which are relatively soluble in water. Correct thermodynamic calculations of the most stable hydrate structure have been made. The model has been tested on a wide selection of open literature and proprietary experimental data. In most cases the hydrate dissociation temperature is predicted to within 1 degree Kelvin. Hydrate inhibitors decrease the hydrate formation temperature or increase the hydrate formation pressure in a given gas mixture. The model includes parameters for the commonly used inhibitors such as Methanol, and the glycols MEG, DEG and TEG. A new mixing rule has been developed for the SRK equation of state to model the inhibitors' effects on the fluid phases.

PIPESIM

Fluid & Multiphase Modeling

65

The treatment of hydrate inhibition has the following features. The model can represent explicitly all the effects of inhibitors, including the depression of hydrate formation temperature, the depression of the freezing point of water, the reduction in the vapor pressure of water (i.e. the dehydrating effect) and the partitioning of water and inhibitor into the oil, gas and aqueous phases. The model has been developed using all available data for mixtures of water with Methanol, MEG, DEG and TEG. This involves simultaneously representing hydrate dissociation temperatures, depression of freezing point data and vapor-liquid equilibrium data. The solubilities of hydrocarbons and light gases in water/inhibitor mixtures have also been represented. There is no fundamental difference between calculations with and without inhibitors. To investigate the effect of an inhibitor it must be added to the list of components in the mixture and the amount must be specified just as for any other component. It is not possible to specify the amount of inhibitor in a particular phase, only the total amount in the mixture. This is because the inhibitor will be split among the different phases present at equilibrium with the amount in a particular phase depending on the ambient conditions and the amounts of other components present in that phase This is exactly what happens in reality. The amount of inhibitor typically needed would be approximately 35% by mass of inhibitor relative to water. 3.3 Pressure Drop Calculation The Pressure change in a flow device is determined from the general momentum equations; · Elevation: conversion of fluid potential energy into hydrostatic pressure. · Friction: shear stress between pipe wall and fluid(s) · Acceleration: changes in velocity of the fluid. This leads to the equation:(dp/dl) = elevational + frictional + accelerational = (dp/dl)elev+ (dp/dl)fric+ (dp/dl)acc. where = friction factor = fluid density = -( g sin )/gc + - 2/2gcd + - /(gc d /dl)

PIPESIM

66

Fluid & Multiphase Modeling = fluid velocity g = gravitational constant at current altitude gc = universal gravitational constant = flow angle d = pipe diameter

The contribution from the major terms; elevational and frictional can be summarized as; · In well · Elevation term (85-100%) · Frictional (0-15%) · In pipes · Elevation term (0-30%) · Frictional (70-100%) For single phase flow the accelerational term is negligible and is assumed to be zero. Thus the above equation reduces to an elevational and frictional term. In the simultaneous transportation of liquid (oil & water) and gas along a single pipe (or well bore) the basic pressure drop equation is the same as for single phase flow with mixture density and friction factor specific to the correlation in which they are used. 3.3.1 Flow regimes Flow Regimes Classification for Vertical Two Phase Flow The general problem of predicting the pressure drop for the simultaneous flow of gas and liquid is complex. The problem consists of being able to predict the variation of pressure with elevation along the length of the flow string for known conditions of flow. Multiphase vertical flow can be categorized into four different flow configurations or flow regimes, consisting of bubble flow, slug flow, slug-mist transition flow and mist flow. A typical example of bubble flow is the liberation of solution gas from an undersaturated oil at and above the point in the flow string where its bubble point pressure is reached.

PIPESIM

Fluid & Multiphase Modeling

67

In vslug, both the gas and liquid phases significantly contribute to the pressure gradient. the gas phase exists as large bubbles almost filling the pipe and separated by slugs of liquid. In transition flow, the liquid slugs between the gas bubbles essentially disappear, and at some point the liquid phases becomes discontinuous and the phase becomes continuous. The pressure losses in vtrans are partly a result of the liquid phase, but are more the result of the gas phase. vannular is characterized by a continuous gas phase with liquid occurring as entrained droplets in the gas stream and as a liquid film wetting the pipe walls. A typical example of mist flow is the flow of gas and condensate in a gas condensate well. Vertical bubble flow

Vertical Slug Flow

Vertical Transition Flow

PIPESIM

68

Fluid & Multiphase Modeling

Vertical Annular/Mist Flow

Flow Regimes Classification for Horizontal Two Phase Flow Prediction of liquid holdup is less critical for pressure loss calculations in horizontal flow than for inclined or vertical flow, although several correlations will require a holdup value for calculating the density terms used in the friction and acceleration pressure drop components. The acceleration pressure drop is usually minor and is often ignored in design calculations. As in the vertical flow, the two-phase horizontal flow can be divided into the following flow regimes: smooth (smooth, wavy), Intermittent Flow (plug and slug) and Distributive Flow (bubble and mist). Smooth

Wavy

Slug

Elongated bubble/Plug

Annular/Mist

Bubble

PIPESIM

Fluid & Multiphase Modeling

69

3.3.2 Single Phase Flow Correlations There are a number of different methods available for calculating the friction factor (). In all cases Re is the Reynolds number (Re) is given by Re = d/µ Where = fluid density = fluid velocity d = pipe diameter µ = fluid viscosity 3.3.2.1 Moody For liquid or gas For laminar flow (Re < 2000) = 64/Re For turbulent flow (Re > 2000) -0.5 = 1.74 - 2log((2 /d) + (18.7/Re 0.5)) where = pipe roughness 3.3.2.2 AGA For gas only. For laminar flow (Re < 1000): = 64/Re For turbulent flow: (0.25 )-0.5 = 4log10(3.7d/ ) For transition flow: (0.25 )-0.5 = 4log10(Re/(0.25 )-0.5) - 0.6 The boundary between transition and turbulent flow is a function of the Reynolds number and friction factor.

PIPESIM

70

Fluid & Multiphase Modeling

3.3.2.3 Panhandle 'A' For gas only (0.25 )-0.5 = 6.872Re0.07305 3.3.2.4 Panhandle 'B' For gas only (0.25 )-0.5 = 16.49Re0.01961 3.3.2.5 Hazen-Williams For liquid water only = (1/192)(150/ m)0.15d-0.17 (in Engineering units) 3.3.2.6 Weymouth For gas only (0.25 ) = 0.00272d-1/3 (in SI units) 3.3.3 Vertical Multiphase Flow Correlations The following vertical multiphase flow correlations are available: 3.3.3.1 Ansari The Ansari model was developed as part of the Tulsa University Fluid Flow Projects (TUFFP) research program. A comprehensive model was formulated to predict flow patterns and the flow characteristics of the predicted flow patterns for upward two-phase flow. The comprehensive mechanistic model is composed of a model for flow pattern prediction and a set of independent models for predicting holdup and pressure drop in bubble, slug, and annular flows. The model was evaluated by using the TUFFP well databank that is composed of 1775 well cases, with 371 of them from Prudhoe Bay data. 3.3.3.2 Baker Jardine Revised Baker Jardine & Associates (now is part of Schlumberger) have developed a correlation for two phase flow in gas-condensate pipelines with a no-slip liquid volume fraction of lower than 0.1. This model represents no major advance in theory, but rather a consolidation of various existing mechanistic models, combined with a modest amount of theoretical development and field data testing. The model uses the Taitel Dukler flow regime map and a modified set of the Taitel Dukler momentum balance to predict liquid holdup. The

PIPESIM

Fluid & Multiphase Modeling

71

pressure loss calculation procedure is similar in approach to that proposed by Oliemans, but accounts for the increased interfacial shear resulting from the liquid surface roughness. The BJA correlation is used for pressure loss and holdup with flow regime determined by the Taitel Dukler correlation. The BJA correlation has been developed by Baker Jardine & Associates specifically for applications involving low liquid/gas ratios, e.g. gas/condensate pipelines. The BJA correlation is not recommended for systems having a nonslip liquid volume fraction greater than 0.1 Users should note that while quite extensive testing of the correlation against operating data has been undertaken for horizontal and inclined flow, the test data for vertical flow is not so comprehensive. 3.3.3.3 Beggs & Brill Original The Original Beggs & Brill correlation is used for pressure loss and holdup. The flow regime is determined by either the Beggs & Brill or Taitel Dukler correlation. The Beggs & Brill correlation was developed following a study of two-phase flow in horizontal and inclined pipes. The correlation is based upon a flow regime map that is first determined as if the flow was horizontal. A horizontal holdup is then calculated by correlations, and this holdup is corrected for the angle of inclination. The test system included two 90 ft long acrylic pipes, winched to a variable elevation in the middle, so as to model incline flow both upwards and downwards at angles of up to 90°. 3.3.3.4 Beggs & Brill Original, Taitel Dukler map As Beggs & Brill original, but utilizing the Taitel Dukler flow map 3.3.3.5 Beggs & Brill Revised As above except that the revised version of the Beggs & Brill correlation is used, with rough pipe friction factors, holdup limiters and corrective constants as proposed by Palmer and Payne. The following enhancements to the original method are used; (1) an extra flow regime of froth flow is considered which assumes a no-slip holdup, (2) the friction factor is changed from the standard smooth pipe model, to utilize a single phase friction factor based on the average fluid velocity.

PIPESIM

72

Fluid & Multiphase Modeling

3.3.3.6 Beggs & Brill Revised, Taitel Dukler map As Beggs & Brill Revised, but utilizing the Taitel Dukler flow map 3.3.3.7 Brill & Minami The Brill & Minami Holdup correlations can be used with any flow map correlations except Mukerjee & Brill and No Slip. 3.3.3.8 Duns & Ros The Duns & Ros correlation is used for pressure loss and holdup with flow regime determination by either the Duns & Ros or the Taitel Dukler correlations. The Duns & Ros correlation was developed for vertical flow of gas and liquid mixtures in wells. Equations were developed for each of three flow regions, (I) bubble, plug and part of froth flow regimes, (II) remainder of froth flow and slug flow regimes, (III) mist flow regime. These regions have low, intermediate and high gas throughputs respectively. Each flow region has a different holdup correlation. The equations were based on extensive experimental work using oil and air mixtures. 3.3.3.9 Duns & Ros, Taitel Dukler map As Duns & Ros, but utilizing the Taitel Dukler flow map 3.3.3.10 Govier & Aziz The correlation of Aziz, Govier, and Forgasi is used for pressure loss, holdup, and flow regime. The Govier, Aziz & Fogarasi correlation was developed following a study of pressure drop in wells producing gas and condensate. Actual field pressure drop v. flowrate data from 102 wells with gas-liquid ratios ranging from 3,900 to 1,170,000 scf/bbl were analyzed in detail. The phase conditions in the well bore were determined by standard flash calculations. Pressure-gradient data for flow under single-phase conditions were compared with conventional predictions, and found generally to confirm them. For the test in which two-phase conditions were predicted throughout the well bore, the field data were compared with several wholly empirical prediction methods, with a previously proposed method, and with a new prediction method partly based on the mechanics of flow. The new prediction method incorporates an empirical estimate of the distribution of the liquid phase between that flowing as a film on the wall and that entrained in the gas core. It employs separate

PIPESIM

Fluid & Multiphase Modeling

73

momentum equations for the gas-liquid mixture in the core and for the total contents of the pipe. 3.3.3.11 Gray The Gray Vertical Flow correlation is used for pressure loss and holdup. This correlation was developed by H E Gray of Shell Oil Company for vertical flow in gas and condensate systems which are predominantly gas phase. Flow is treated as single phase, and dropped out water or condensate is assumed to adhere to the pipe wall. It is considered applicable for vertical flow cases where the velocity is below 50 ft/s, the tube size is below 3½-in, the condensate ratio is below 50 bbl/mmscf, and the water ratio is below 5 bbl/mmscf. 3.3.3.12 Hagedorn & Brown The correlation of Hagedorn & Brown is used for pressure loss and holdup. There is a choice of either Beggs & Brill, Duns & Ros or Taitel Dukler flow regime determination. The Hagedorn and Brown correlation was developed following an experimental study of pressure gradients occurring during continuous two-phase flow in small diameter vertical conduits. A 1,500 ft experimental well was used to study flow through 1-in, 1¼-in, and 1½-in nominal size tubing. Tests were conducted for widely varying liquid flowrates, gas-liquid ratios and liquid viscosities. All of the correlations involve only dimensionless groups, which is a condition usually sought for in similarity analysis but not always achieved. BJA consider the use of the original correlation unwise, as it can grossly underestimate liquid holdup. Users are advised to use the Hagedorn & Brown Revised correlation. 3.3.3.13 Hagedorn & Brown, Duns & Ros map As Hagedorn & Bown, but utilizing the Duns & Ros flow map 3.3.3.14 Lockhart & Martinelli

3.3.3.15 Lockhart & Martinelli, Taitel Dukler map As Lockhard & Martinelli, but utilizing the Taitel Dukler flow map 3.3.3.16 Mukherjee & Brill: The Mukerjee & Brill correlation is used for Pressure loss, Holdup and flow map. Note: selection of alternative flow maps and/or holdups will

PIPESIM

74

Fluid & Multiphase Modeling

cause unpredictable results. The Mukherjee & Brill correlation was developed following a study of pressure drop behavior in two-phase inclined flow. For bubble and slug flow a no-slip friction factor, calculated from the Moody diagram, was found adequate for friction head loss calculations. In downhill stratified flow, the friction pressure gradient is calculated based on a momentum balance equation for either phase assuming a smooth gas-liquid interface. For annularmist flow, a friction factor correlation was presented that is a function of holdup ratio and no-slip Moody friction factor. Results agreed well with the experimental data and correlations were further verified with Prudhoe Bay and North Sea data. 3.3.3.17 NOSLIP Correlation The NOSLIP correlation assumes homogeneous flow with no slip between the phases. Fluid properties are taken as the average of the gas and liquid phases and friction factors are calculated using the single phase MOODY correlation. Note: selection of alternative flow maps and/or holdups will cause unpredictable results. 3.3.3.18 OLGA-S 2000 Steady State OLGAS is based in larger part on data from the SINTEF two-phase flow laboratory near Trondheim, Norway. The test facilities were designed to operate at conditions that approximated field conditions. The test loop was 800 m long and 8 inches in diameter. Operating pressures between 20 and 90 barg were studied. Gas superficial velocities of up to 13 m/s, and liquid superficial velocities of up to 4 m/s were obtained. In order to simulate the range of viscosities and surface tensions experienced in field applications, different hydrocarbon liquids were used (naptha, diesel, and lube oil). Nitrogen was used as the gas. Pipeline inclination angles between 1° were studied in addition to flow up or down a hill section ahead of a 50m high vertical riser. Over 10,000 experiments were run on this test loop during an eight year period. The facility was run in both steady state and transient modes. OLGAS considers four flow regimes, stratified, annular, slug and dispersed bubble flow and uses a unique minimum slip criteria to predict flow regime transitions. This correlation is available to all members of the SINTEF syndicate, and to nonmembers on payment of the appropriate royalty fees.

PIPESIM

Fluid & Multiphase Modeling

75

A separate document is available that details OLGA-S 2000. This can be downloaded from our web site. 3.3.3.19 Orkiszewski The Orkiszewski correlation is used for pressure loss, holdup, and flow regime. The Orkiszewski correlation was developed for the prediction of two phase pressure drops in vertical pipe. Four flow regimes were considered, bubble, slug, annular-slug transition, and annular mist. The method can accurately predict, to within 10%, the two phase pressure drops in naturally flowing and gas lifted production wells over a wide range of well conditions. The precision of the method was verified when its predicted values were compared against 148 measured pressure drops. Unlike most other methods, liquid holdup is derived from observed physical phenomena, and is adjusted for angle of deviation. 3.3.3.20 Shell SIEP Correlations These correlations are provided by Shell International Exploration & Production (SIEP) and are for Shell or Shell approved clients only. Correlations available; · MMSM · GZM 3.3.3.21 Shell SRTCA Correlations These correlations are provided by Shell International Oil Products and are for Shell or Shell approved clients only. Correlations available; · SRTCA two-phase · STRCA two-phase slugging · STRCA two-phase slugging & slug DP · STRCA three-phase · STRCA three-phase & water-oil dispersion 3.3.3.22 GRE Mechanistic Model BP This correlation is provided by BP and is available for general use.

PIPESIM

76

Fluid & Multiphase Modeling

3.3.4 Horizontal Multiphase Flow Correlations The following horizontal multiphase flow correlations are available: 3.3.4.1 Baker Jardine Revised Baker Jardine (is now part of Schlumberger) have developed a correlation for two phase flow in gas-condensate pipelines with a no-slip liquid volume fraction of lower than 0.1. This model represents no major advance in theory, but rather a consolidation of various existing mechanistic models, combined with a modest amount of theoretical development and field data testing. The model uses the Taitel Dukler flow regime map and a modified set of the Taitel Dukler momentum balance to predict liquid holdup. The pressure loss calculation procedure is similar in approach to that proposed by Oliemans, but accounts for the increased interfacial shear resulting from the liquid surface roughness. The BJA correlation is used for pressure loss and holdup with flow regime determined by the Taitel Dukler correlation. The BJA correlation has been developed specifically for applications involving low liquid/gas ratios, e.g. gas/condensate pipelines. The BJA correlation is not recommended for systems having a nonslip liquid volume fraction greater than 0.1 3.3.4.2 Beggs & Brill Original The original Beggs & Brill correlation is used for pressure loss and either the BBO or the BJA correlation is used to calculate holdup. Flow regime is determined by either the Beggs & Brill or Taitel Dukler correlation. The Beggs & Brill correlation was developed following a study of two-phase flow in horizontal and inclined pipes. The correlation is based upon a flow regime map which is first determined as if the flow was horizontal. A horizontal holdup is then calculated by correlations, and this holdup is corrected for the angle of inclination. The test system included two 90 ft long acrylic pipes, winched to a variable elevation in the middle, so as to model incline flow both upwards and downwards at angles of up to 90°. 3.3.4.3 Beggs & Brill Original, Taitel Dukler map As Beggs & Brill Original, but utilizing the Taitel Dukler flow map

PIPESIM

Fluid & Multiphase Modeling

77

3.3.4.4 Beggs & Brill Revised As above except that the revised version of the Beggs & Brill correlation is used, with rough pipe friction factors, holdup limits and corrective constants as proposed by Palmer and Payne. The following enhancements to the original method are used; (1) an extra flow regime of froth flow is considered which assumes a no-slip holdup, (2) the friction factor is changed from the standard smooth pipe model, to utilize a single phase friction factor based on the average fluid velocity. 3.3.4.5 Beggs & Brill Revised, Taitel Dukler map As Beggs & Brill Revised, but utilizing the Taitel Dukler flow map 3.3.4.6 Brill & Minami: The Brill and Minami Holdup correlations can be used with any pressure loss and any flow map correlations except Mukherjee & Brill and No Slip. 3.3.4.7 Dukler, AGA + Flanigan The AGA & Flanigan correlation was developed for horizontal and inclined two phase flow of gas-condensate gathering systems. The Taitel Dukler flow regime map is used which considers five flow regimes, stratified smooth, stratified wavy, intermittent, annular dispersed liquid, and dispersed bubble. The Dukler equation is used to calculate the frictional pressure loss and holdup, and the Flanigan equation is used to calculate the elevational pressure differential. 3.3.4.8 Dukler , AGA + Flanigan (Eaton holdup) As Duker, AGA + Flanigan but with liquid holdup calculated according to the Eaton correlation. 3.3.4.9 Duns & Ros, Taitel Dukler map The Duns & Ros correlation is used for pressure loss, with a choice of either Duns & Ros or BJA holdup. Flow regime determination is either the Duns & Ros or the Taitel Dukler correlations. The Duns & Ros correlation was developed for vertical flow of gas and liquid mixtures in wells. Equations were developed for each of three flow regions, (I) bubble, plug and part of froth flow regimes, (II) remainder of froth flow and slug flow regimes, (III) mist flow regime. These regions have low, intermediate and high gas throughputs

PIPESIM

78

Fluid & Multiphase Modeling

respectively. Each flow region has a different holdup correlation. The equations were based on extensive experimental work using oil and air mixtures. 3.3.4.10 Lockhart & Martinelli

3.3.4.11 Lockhart & Martinelli, Taitel Dukler map As Lockhard & Martinelli, but utilizing the Taitel Dukler flow map 3.3.4.12 Mukherjee & Brill The Mukherjee & Brill correlation is used for Pressure loss, Holdup and Flow Map. Note: selection of alternative flow maps and/or holdups will cause unpredictable results. The Mukherjee & Brill correlation was developed following a study of pressure drop behavior in two-phase inclined flow. For bubble and slug flow, a noslip friction factor calculated from the Moody diagram was found adequate for friction head loss calculations. In downhill stratified flow, the friction pressure gradient is calculated based on a momentum balance equation for either phase assuming a smooth gas-liquid interface. For annular-mist flow, a friction factor correlation was presented that is a function of holdup ratio and no-slip Moody friction factor. Results agreed well with the experimental data and correlations were further verified with Prudhoe Bay and North Sea data. 3.3.4.13 NOSLIP Correlation The NOSLIP correlation assumes homogeneous flow with no slip between the phases. Fluid properties are taken as the average of the gas and liquid phases and friction factors are calculated using the single phase MOODY correlation. Note: selection of alternative flow maps and/or holdups will cause unpredictable results. 3.3.4.14 OLGA-S 2000 Steady-State: OLGAS is based in larger part on data from the SINTEF two-phase flow laboratory near Trondheim, Norway. The test facilities were designed to operate at conditions that approximated field conditions. The test loop was 800 m long and 8 inches in diameter. Operating pressures between 20 and 90 barg were studied. Gas superficial velocities of up to 13 m/s, and liquid superficial velocities of up to 4 m/s were obtained. In order to simulate the range of viscosities and

PIPESIM

Fluid & Multiphase Modeling

79

surface tensions experienced in field applications, different hydrocarbon liquids were used (naptha, diesel, and lube oil). Nitrogen was used as the gas. Pipeline inclination angles between 1° were studied in addition to flow up or down a hill section ahead of a 50m high vertical riser. Over 10,000 experiments were run on this test loop during an eight year period. The facility was run in both steady state and transient modes. OLGAS considers four flow regimes, stratified, annular, slug and dispersed bubble flow and uses a unique minimum slip criteria to predict flow regime transitions. This correlation is available to all members of the SINTEF syndicate, and to nonmembers on payment of the appropriate royalty fees. A separate document is available that details OLGA-S 2000. This can be downloaded from our web site. 3.3.4.15 Oliemans The Oliemans correlation was developed following the study of large diameter condensate pipelines. The flow regime is predicted using the Taitel Dukler flow regime map, and a simple model, which obeyed the correct single phase flow limits was introduced to predict the pressure drop. The model was based on a limited amount of data from a 30-in, 100-km pipeline operating at pressures of 100 barg or higher. The Oliemans pressure loss correlation can be used with the Eaton, BJA, BRIMIN1 or BRIMIN2 holdup correlations. 3.3.4.16 Xiao The Xiao comprehensive mechanistic model was developed as part of the TUFFP research program. It was developed for gas-liquid twophase flow in horizontal and near horizontal pipelines. The model is able first to detect the existing flow pattern, and then to predict the flow characteristics, primarily liquid holdup and pressure drop, for the stratified, intermittent, annular, or dispersed bubble flow patterns. The model was tested against a pipeline data bank. The data bank included large diameter field data culled from the AGA multiphase pipeline data bank, and laboratory data published in literature. Data included both black oil and compositional fluid systems. A new correlation was proposed which predicts the internal friction factor under stratified flow.

PIPESIM

80

Fluid & Multiphase Modeling

3.3.4.17 Shell SIEP Correlations These correlations are provided by Shell International Exploration & Production (SIEP) and are for Shell or Shell approved clients only. Correlations available; · GZM 3.3.4.18 Shell SRTCA Correlations These correlations are provided by Shell International Oil Products and are for Shell or Shell approved clients only. Correlations available; · SRTCA two-phase · STRCA two-phase slugging · STRCA two-phase slugging & slug DP · STRCA three-phase · STRCA three-phase & water-oil dispersion 3.3.4.19 GRE Mechanistic Model BP This correlation is provided by BP and is available for general use. 3.4 References Multiflash for Windows - User Guide. Infochem. Aziz, K., Govier, G. W. and Forgasi, M.: "Pressure Drop in Wells Producing Oil and Gas," J. Cdn. Pet. Tech. (July-Sept. 1972) 38-48. Baker, A., Nielsen, K., and Gabb, A.: "Pressure Loss, Liquid-Holdup Calculations Developed," Technology, Oil & Gas Journal (Mar. 14, 1988). Beal, C.: "The Viscosity of Air, Water, Natural Gas, Crude Oil and its Associated Gases at Oil Temperatures and Pressures," Trans. AIME (1946) 94.

PIPESIM

Fluid & Multiphase Modeling

81

Beggs, H. D., and Brill, J. P.: "A Study of Two Phase Flow in Inclined Pipes," J. Pet. Tech. (May 1973) 607-617. Beggs, H. D. and Robinson, J. R.: "Estimating the Viscosity of Crude Oil Systems," J. Pet. Tech. (Sept. 1975) 1140-1. Brill, J. P. et al.: "Analysis of Two-Phase Tests in Large Diameter Flow Lines in Prudhoe Bay Field," SPEJ (June 1981). Brill, J. P. and Beggs, D. H.: Two-Phase Flow in Pipes, 6th Edition, University of Tulsa, Tulsa, Oklahoma, December 1988. Brown, K.E.: The Technology of Artificial Methods, Penwell Publishing Company, Tulsa, Oklahoma, 1984. Chew, J. and Conally, C. A. Jr.: "A Viscosity Correlation for Gas Saturated Crude Oils," Trans., AIME (1974) 23. Dukler, E. A., et al.: "Gas-Liquid Flow in Pipelines, I. Research Results," AGA-API Project NX-28 (May 1969). Duns, H., and Ros, N. C. J.: "Vertical Flow of Gas and Liquid Mixtures in Wells," 6th. World Pet. Congress (1963) 452. Eaton, B. A.: "Prediction of Flow Patterns, Liquid Holdup and Pressure Losses Occurring During Continuous Two-Phase Flow in Horizontal Pipelines," Trans., AIME (1967) 815. Fetkovich, M.J. and Vienot, M.E.: "Shape Factors, CA, Expressed as a Skin, sca," JPT (February 1985) 321-322. Flanigan, O.: "Effect of Uphill Flow on Pressure Drop in Design of TwoPhase Gathering Systems," Oil and Gas J. (March 10, 1958) 56, 132.

PIPESIM

82

Fluid & Multiphase Modeling

Glaso, O., "Generalized Pressure Volume Temperature Correlation," J. Pet. Tech. (May 1980) 785. Golan, M. and Whitson, C.H.: Well Performance, International Human Resources Corporation, Boston, MA (1986). Hagedron, A. R. and Brown, K. E.: "Experimental Study of Pressure Gradients Occurring During Continuous Two-Phase Flow in SmallDiameter Vertical Conduits," J. Pet. Tech. (April 1965) 475-484. Katz, D. L. et al.: Handbook of Natural Gas Engineering, McGraw Hill Book Co., Inc., New York (1959). Lasater, J. A.: "Bubble Point Pressure Correlation," Trans., AIME (1958) 379. Lee, A. L. et al.: "The Viscosity of Natural Gases," Trans., AIME (1966) 997. Lockhart, R. W. and Martinelli, R. C.: "Proposed Correlation of Data for Isothermal Two-phase, Two-Component Flow in Pipes," Chem. Eng. Prog. (January 1949) 45, 39. McLeod, H. O.: "The Effect of Perforating Conditions on Well Performance," JPT (Jan. 1983). Manhane, J. M., Gregory, G. A. and Aziz, K.: "A Flow Pattern Map for Gas-Liquid Flow Pattern in Horizontal Pipes," Int. J. of Multiphase Flow. Minami, K. and Brill, J. P.: "Liquid Holdup in Wet Gas Pipelines," SPE J. Prod. Eng. (May 1987).

PIPESIM

Fluid & Multiphase Modeling

83

Mukherjee, H. and Brill, J. P.: "Liquid Holdup Correlations for Inclined Two-Phase Flow," JPT (May 1983) 1003-1008. Mukherjee, H. and Economides, M. J.: "A Parametric Comparison of Horizontal and Vertical Well Performance," SPE paper 18303 presented at the Annual Technical Conference and Exhibition in Houston, October 1988. Muskat, M.: The Flow of Homogeneous Fluids Through Porous Media, I.H.R.D.C., Boston (1937). Norris, L.: "Correlation of Prudhoe Bay Liquid Slug Lengths and Holdups Including 1981 Large Diameter Flowlines Tests," Internal Report Exxon (October 1982). Oliemans, R. V. A.: "Two-Phase Flow in Gas-Transmission Pipeline," ASME paper 76-Pet-25, presented at Pet. Div. ASME meeting Mexico City (Sept. 1976). Orkiszewski, J.: "Predicting Two-Phase Pressure Drops in Vertical Pipes," J. Pet. Tech. (June 1967) 829-838. Palmer, C. M.: "Evaluation of Inclined Pipe Two-Phase Liquid Holdup Correlations Using Experimental Data," M.S. Thesis, The University of Tulsa (1975). Payne, G. A.: "Experimantal Evaluation of Two-Phase Pressure Loss Correlations for Inclined Pipe," M.S. Thesis, The University of Tulsa (1975). Renard, G. I. and Dupuy, J. M.: "Influence of Formation Damage on the Flow Efficiency of Horizontal Wells," SPE paper 19414 presented

PIPESIM

84

Fluid & Multiphase Modeling

at the Formation Damage Control Symposium, Lafayette (February 1990). Scott, S. L., Shoham, O., and Brill, J. P.: "Prediction of Slug Length in Horizontal Large-Diameter Pipes," SPE paper 15103 (April 1986). Standing, M. B.: Volumetric and Phase Behavior of Oil Field Hydrocarbon Systems, Society of Petroleum Engineers, (1977) 121. Standing, M. B.: "A General Pressure Volume-Temperature Correlation for Mixtures Of California Oils and Greases," Drill. and Prod. Prac., API (1947) 275. Standing, M. B. and Katz, D. L.: "Volumetric and Density of Natural Gases," Trans., AIME (1942) 140. Taitel, Y. and Dukler, A. E.: "A Model for Predicting Flow Regime Transitions in Horizontal Gas-Liquid Flow," AICHE J. (vol. 22, no. 1) (Jan. 1976) 47-55. Vasquez, M., and Beggs, H. D.: "Correlations for Fluid Physical Property Prediction," SPE paper 6719, presented at the 52nd Annual Technical Conference and Exhibition of the Society of Petroleum Engineers, Denver, Colorado (1977). Woelflin, W.: "The Viscosity of Crude-Oil Emulsions," Drill. and Prod. Prac., API (1942) 148.

PIPESIM

Fluid & Multiphase Modeling THIS PAGE LEFT BLANK INTENTIONALLY

85

PIPESIM

Reservoir, Well & Completion Modeling

87

4 Reservoir, Well & Completion Modeling The well modeling components of PIPESIM are; · Completion · Vertical · Horizontal · Tubing · Deviation survey · Gas lift injection point or points · ESP lift point · chokes 4.1 Vertical Completions Inflow performance relationships (IPRs) have been developed to model the flow of fluids from the reservoir, through the formation, and into the well. They are expressed in terms of Pws (static reservoir pressure), Pwf (flowing bottom hole pressure), and Q (flowrate). 4.1.1 Liquid Reservoirs 4.1.1.1 Fetkovich / Normalized back pressure Is a development of the Vogel equation to take account of high velocity effects. The equation is follows: Q = Qmax(1 - (Pwf/Pws)2)n, where Qmax is the open flow potential, i.e. the liquid flowrate when the bottom hole pressure is zero, and n is the PI coefficient 4.1.1.2 Jones The Jones equation is Pws - Pwf = AQ2 + BQ. Where A is the turbulent coefficient and B is the laminar coefficient. The coefficients must satisfy A => 0 and B=> 0.

PIPESIM

88

Field Equipment

4.1.1.3 Pseudo-Steady state / Darcy The Pseudo Steady-state equation is given as Q = kh(Pws - Pwf)/(141.2µoBo(ln(Re/Rw) - 0.75 + s))) where s = skin k = formation permeability h = formation thickness µ = liquid viscosity B = formation volume factor Re = Drainage radius Rw = wellbore radius Alternatively, the skin (and related turbulence coefficient) values can be calculated by describing the completion. 4.1.1.4 (Straight line) Well productivity Index The productivity index relationship is Q = J(Pws - Pwf) where J = productivity index. 4.1.1.5 (Straight line) Well productivity Index (with Vogel correction below bubble point) Below the bubble point pressure, the relationship can be modified to take account of evolved gas. The correction is to apply the Vogel relationship below the bubble point. 4.1.1.6 Vogel Was developed to model saturated oil wells. The equation is as follows: Q = Qmax(1 - (1 - C)(Pwf/Pws) - C(Pwf/Pws)2), where Qmax is the absolute open flow potential, i.e. the liquid flowrate when the bottom hole pressure is zero, and C is the PI coefficient. The value of C is usually around 0.8. 4.1.1.7Hydraulic Fracture See Help system for details.

PIPESIM

Field Equipment

89

4.1.1.8Multi-rate tests In addition multi-rate test data can be utilized so that the modeled inflow matches the actual measured inflow in the well. Two types of multi-rate test are available; · multi-point - A 'flow-after-flow' test sequence. Static pressure is taken as a constant throughout the test period. · Isochronal - This type of test is normally performed in reservoirs with low permeability where the time taken to reach stabilized flow conditions is unacceptably long (e.g. low permeability sands). Isochronal testing is performed by periods of flowing followed by shutting-in of a well (normally with increasing rate). The wellbore flowing pressure is recorded during each flow period at a specific time (e.g. if the time is 4 hours, then the test is referred to as a 4hour isochronal test). Due to the long stabilization time normally associated with the isochronal test, reservoir conditions need not return to the original static pressure. Hence a different static reservoir pressure is recorded. Multi-rate test data can be applied to the following; · Multi-rate Fetkovich · Multi-rate Jones 4.1.2 Gas and Gas Condensate Reservoirs 4.1.2.1 Back pressure / C and n Developed by Rawlins and Schellhardt in 1935 after testing 582 wells. The equation is Q = C(Pws2 - Pwf2)n. 4.1.2.2 Forchheimer The Forchheimer equation is; Pws2 - Pwf2 = FQ2 + AQ. Where F is the turbulence coefficient and A is the laminar coefficient. The coefficients must satisfy F => 0 and A=> 0. 4.1.2.3 Jones The Jones equation is :

PIPESIM

90

Field Equipment

Pws2 - Pwf2 = AQ2 + BQ. Where A is the turbulent coefficient and B is the laminar coefficient. The coefficients must satisfy A > 0 and B=> 0. 4.1.2.4 Pseudo-Steady state / Darcy The Pseudo Steady-state equation is given as Q = kh(Pws2 - Pwf2)/(1422µTz(ln(Re/Rw) - 0.75 + s))) where s = skin k = formation permeability h = formation thickness µ = gas viscosity T = temperature Z = z factor Re = Drainage radius Rw = wellbore radius Alternatively, the skin (and related turbulence coefficient) values can be calculated. 4.1.2.5 (Straight line) Well productivity Index The productivity index relationship is Q = J(Pws2 - Pwf2) where J = productivity index. 4.1.2.6 Hydraulic Fracture 4.1.2.7Multi-rate tests Multi-rate test data (as descried above) can be applied to the following; · Multi-rate Back pressure / C and n · Multi-rate Forchheimer · Multi-rate Jones · Multi-rate (Straight line) Well Productivity Index

PIPESIM

Field Equipment

91

4.2 Horizontal Completions This section focuses on the reservoir engineering aspects of horizontal well technology. The pressure drop in horizontal wells and its effect on well performance will be discussed. The steady state and pseudo-steady state analytical solutions on the productivity of horizontal wells will also be reviewed for both oil and gas wells. The main purpose of drilling horizontal wells is to enhance production. There are also many circumstances that lead to drilling horizontal wells (Cooper, 1988): · Thin reservoirs - The increased area of contact of the horizontal well with the reservoir is reflected by the productivity index (PI). Typically, the PI for a horizontal well may be increased by a factor of 4 when compared to a vertical well penetrating the same reservoir. Heterogeneous reservoirs - When irregular reservoirs exist, the horizontal well can effectively intersect isolated productive zones which might otherwise be missed. A horizontal well can also intersect vertical natural fractures in a reservoir.

·

Reduce water/gas coning - A horizontal well provides minimum pressure drawdown which delays the onset of water/gas breakthrough. Even though the production per unit well length is small, the long well length provides high production rates.

·

Vertical permeability - If the ratio of vertical permeability to horizontal permeability is a high, a horizontal well may produce more economically than a vertical well.

·

4.2.1 Effect of Pressure Drop on Productivity In reservoir engineering calculations, the horizontal wellbore is treated as an infinite conductivity fracture, i.e. the pressure drop along the well length is negligible. However, in practice, there is a pressure drop from the toe (tip-end) of the horizontal wellbore to the heel (producing-end) so as to maintain fluid flow within the wellbore (see Figure 4.1). Dikken (1989), Folefac (1991) and Joshi (1991) have recently addressed the effect of wellbore pressure gradient on horizontal well production performance.

PIPESIM

92

Field Equipment

Figure 4.1 Along-hole pressure gradient of a horizontal well (Joshi, 1991) Dikken (1990) and Folefac (1991) contend that the assumption of constant pressure wellbore is reasonable for single phase laminar flow but is no longer valid when turbulent or multiphase flow occurs. Folefac (1991) showed that a typical well with the following properties: o = 800 kg/m3; µ = 1.0 cp; d = 0.1968 m; and Q = 5000 RB/d gives a NRe of 4000 which is well above the turbulence transition limit of 2000. In most practical situations, Dikken (1990) asserts that horizontal wells will exhibit non-laminar flow. In addition, the pressure drop will be even greater when multiphase flow exists. Joshi (1991), thus, asked the question: What is the magnitude of the wellbore pressure drop as compared to pressure drop from the reservoir to the wellbore? If the wellbore pressure drop is significant as compared to the reservoir drawdown, then the reservoir drawdown, and consequently, the production rate along the well length will change. Thus, there is a strong interaction between the wellbore and the reservoir. The reservoir flow and wellbore equations must be solved simultaneously as shown in Figure 4.2.

PIPESIM

Field Equipment

93

Figure 4.2 Schematic of reservoir and flow relationship (Joshi, 1991) The coupled equations were solved by Dikken (1990) analytically by simplified boundary conditions, notably, no inflow from the toe-end. Folefac (1991) used a Black Oil type model that involved a finite volume technique. Folefac (1991) concluded that the well length, wellbore diameter and perforated interval had the most profound effect on the level of pressure drop in the wellbore. Folefac (1991) pointed out that the wellbore pressure profile is non-linear with respect to the well length. This is because the mixture momentum equation has a non-linear term in velocity, the friction force. This in turn will result in an uneven drawdown in the reservoir that is otherwise considered homogenous. Furthermore, Folefac (1991) showed that as the wellbore radius increased from 64.5 mm (2.5") to 114.3 mm (4.5"), the rate at which pressure dropped along the wellbore became nearly constant. This is mainly due to the turbulent flow being converted to laminar flow by drilling a larger size hole. Joshi (1991) mentions other situations where wellbore pressure drop is considerable: · High flowrates of light oil (10,000 to 30,000 RB/d).

· ·

High viscous crude's (heavy oils and tar sands). Long well lengths.

The wellbore pressure drop effects well deliverability and in turn influences well completion and well profile design. The need to accurately calculate well flowrates and wellbore pressures is therefore, essential. Joshi (1991) lists a few remedies to minimize high wellbore pressure drops: · Drilling a larger diameter hole would dramatically reduce the pressure drop. The reason being that for single phase flow - P 1/d5. For example, Joshi (1991), states " for a given production rate, by increasing the well diameter twofold, the pressure drop can be reduced at least thirty-two fold".

·

Varying the shot density of a cemented hole or the slot size of a slotted liner would control production rates and minimise pressure drop along the wellbore

PIPESIM

94

·

Field Equipment Gravel packs are used in high permeability reservoirs. If the well is completed with a slotted liner, the slots should be placed as far apart as possible. Joshi (1991) states that "this will let the gravel pack act as a choke and facilitate maintaining minimum pressure drop across the well length".

Therefore, by selecting the appropriate well geometry, hole size and length, wellbore pressure drops can be minimized. 4.2.2 Single Phase Pressure Drop Assuming that the horizontal wellbore can be treated as a horizontal pipe, the single phase flow pressure drop calculation for oil flow can be written as follows:

p = (114644 x10 -5 ) fm q 2 L / d 5 .

(4.1)

where, p fm q L d

= pressure drop, psia = Moody's friction factor, dimensionless = fluid density, gm/cm3 = flowrate, RB/d = horizontal length, ft = internal diameter of pipe, inches

For gas flow, however, the pressure drop calculations are more complex. This is due to friction, which could change the temperature of the gas as it travels through the wellbore. Moreover, density and viscosity are strong functions of gas pressure and temperature. This would result in a changing pressure drop per foot length of a well along the entire well length. The Weymouth equation for dry gas is the simplest equation to estimate pressure drop in a horizontal pipe

qg = 15320 ( p12 - p2 2 ) d 16 / 3 g TZL

(4.2)

where qg p1 p2 L

PIPESIM

= gas flowrate, scfd = pipe inlet pressure, psia = pipe outlet pressure, psia = pipe length, miles

Field Equipment

95

T = average temperature, oR Z = average gas compressibility factor d = pipe diameter, in g = oil volume formation factor, RB/STB Also, several multiphase correlations (Brill, 1988) are applicable for a single-phase flow of either oil or gas. 4.2.3 Multiphase Pressure Drop There is very little discussion on multiphase pressure drop in horizontal wells. Folefac (1991) studied the effect of two phase flow (hydrocarbon liquid and water are treated as one phase with identical velocity but averaged properties). The pressure drop along the horizontal wellbore was similar to that for single phase flow. However, the pressure drop was higher than for single phase flow for the same volume of fluid intake. For a horizontal pipe, Brill (1988) has discussed numerous multiphase flow correlations. Slip velocities between phases make these equations more complex than single phase flow equations. In general, Joshi (1991) states that, "different multiphase correlations may give different values of the pressure drop". The various correlations should be compared with actual pressure drop data. However, measuring the pressure at both ends of a horizontal well and calibrating the data is very difficult. There is a definite need for further study on multiphase flow in horizontal wells. 4.2.4 Inflow Production Profiles Horizontal wellbore pressure drops also depend upon the type of fluid inflow profiles. Figure 4.3 shows some horizontal well fluid inflow profiles. On the basis of well boundary condition and reservoir heterogeneity, several profiles are possible. Joshi (1991) examined the effect of different fluid entry profiles on the wellbore pressure drop. Depending on the type of profile, Joshi concluded that the total pressure drop varied from 6 psi to 14.5 psi but it was not large enough to effect the wellhead pressure.

PIPESIM

96

Field Equipment

Figure 4.3 Horizontal Well Inflow Profiles (Joshi, 1991) 4.2.5 Steady-State Productivity The simplest forms of horizontal well productivity calculations are the steady-state analytical solutions, which assume that the pressure at any point in the reservoir is constant over time. According to Joshi (1991), even though very few reservoirs operate under steady-state conditions, steady state solutions are widely used because:

· ·

Analytical derivation is easy. The concepts of expanding drainage boundary over time, effective wellbore radius and shape factors allows the conversion to either transient or pseudo-steady state results to be quite straightforward. Steady-state mathematical experimentally. results can be verified

·

Giger (1984), Economides (1989), Mukherjee (1988) and numerous others have developed solutions to predict steady-state productivity. Most are similar in form to the equation given by Joshi (1988) who simplified the 3-D Laplace equation (2p=0) by coupling two 2-D problems. This was based on the assumption that a horizontal well drains an ellipsoidal volume around the wellbore of length L as shown in Figure 4.4.

PIPESIM

Field Equipment

97

Figure 4.4 Horizontal Well Drainage Pattern For isotropic reservoirs (kh=kv),

qh = 0.007078k h hp /( µ o Bo ) a + a 2 - ( L / 2 )2 h ln[ ] + ( h / L)ln[ ] L/2 2rw

(4.3)

and

a = (L / 2 )[0.5 + 0.25 + (2reh / L) 4 ]0.5

(4.4)

where qh p L h rw reh µo Bo kh

= flowrate, STB/day = pressure drop, psi = horizontal well length, ft = reservoir height, ft = wellbore radius, ft = drainage radius of horizontal well, ft = oil viscosity, cp = oil volume formation factor, RB/STB = horizontal permeability, md

If the length of the horizontal well is significantly longer than the reservoir height, i.e. L >> h, then the second term in the denominator of equation (4.3) is negligible and the solution simplifies to

qh = 0. 007078k h hp /( µ o Bo ) r ln[ eh ] (L / 4 )

(4.5)

Muskat (1937) suggested a simple transformation to account for permeability anisotropy. An effective permeability, keff, is defined as

k eff = k v k h

(4.6)

To account for vertical anisotropy, the reservoir thickness can be modified as follows

PIPESIM

98

h=h kh kv

Field Equipment (4.7)

In addition, the influence of well eccentricity (distance from the center of the reservoir in the vertical plane) was also implemented. Thus, equation (4.3) was transformed as follows 0.007078k h hp /( µ o Bo ) qh = (4.8) 2 2 2 2 2

ln[ a + a - ( L / 2) (h / 2 ) + ] ] + (h / L)ln[ 2rw L /2

where

=

kh kv

(4.9)

and is the horizontal well eccentricity (offset of the well from the center of the pay zone) in feet. Productivity comparisons of a horizontal well to that of a vertical well can easily be made by using equation (4.8). In converting the productivity of a horizontal well into that of an equivalent vertical well, an effective wellbore radius can be calculated, rw,eff rw,eff = rw exp(-s) (4.10) The effective wellbore radius is defined as the theoretical well radius, which will match the production rate. Joshi (1991) assumed equal drainage volumes, reh=rev, and equal productivity indices, Jh=Jv to give the following for an anisotropic reservoir

rw,eff = reh (L / 2 ) a[1+ 1- ( L / 2 a )2 ] + [(h / rw )](h / L)

(4.11)

In this way, controlling parameters like well length, permeability and formation thickness can be used to screen potential candidates for further simulation studies. Renard (1990) studied the effect of formation damage around the wellbore and modified the steady-state equation to include skin. Renard (1990) concluded that due to the lower productivity index per unit length in horizontal wells, the effect of skin damage is not as

PIPESIM

Field Equipment

99

pronounced as it is in vertical wells. Celier (1989) came to the same conclusion with respect to the effect of non-Darcy flow. 4.2.6 Pseudo-Steady State Productivity It is often desirable to calculate productivity from a reservoir with unique boundary conditions, such as a gas cap or bottom water drive, finite drainage area, well location, etc. In these instances pseudosteady state equations are employed. Pseudo-steady state or depletion state begins when the pressure disturbance created by the well is felt at the boundary of the well drainage area. Dake (1978) and Golan (1985) describe the pseudo-steady state flow of an ideal fluid (liquid) in a closed circular drainage area. Rearranging the equation gives the familiar vertical well productivity

qv = khp / 141.2 µ o Bo ln [2.2458 A / (CA Rw 2 )] + s + sm + Dqv

(4.11)

where sm

= mechanical skin factor due to drilling and completion related well damage. s = total skin due to perforations, partial penetration and stimulation. CA = shape factor Dqv = near wellbore turbulence factor

The above equation can be reduced to the following single-phase pseudo-steady state equation for oil flow (assuming s=0, sm=0 and Dqv=0),

qv = khp / 141.2 µ o Bo r ln[( e ) - 0. 75] rw

(4.13)

Equation (4.13) is for a vertical well which is located in the center of a circular drainage area. Fetkovich (1985) wrote the shape factor in terms of an equivalent skin. This skin was expressed by choosing a reference shape factor of a well at the center of circular drainage area

s CA = ln[ C A,ref / C A ]

(4.14)

The horizontal well shape factor depends on the following

PIPESIM

100

· · ·

Field Equipment drainage area shape. well penetration. dimensionless well length, LD = (L/h)(kv/kh)0.5.

Joshi (1991) explains that the well performance approaches a fully penetrating infinite-conductivity fracture when the horizontal well length is LD > 10. Babu (1989), Goode (1989) and Mutalik (1988) have developed methods to calculate pseudo-steady state productivity for single phase flow in horizontal wells. Shape factors were used to arbitrarily locate the well within a rectangular bounded drainage area and the reservoir was bounded in all directions. Mutalik's model assumed the horizontal well as an infinite conductivity well (i.e. the wellbore pressure drop is negligible). Babu's model assumed uniform-flux boundary condition. Goode's model used an approximate infinite conductivity solution where the constant wellbore pressure is estimated by averaging the pressure values of the uniform-flux solution along the well length. Goode (1989) also considered the effects of completion type on productivity. Their model allowed for cased completion, selectively perforated completion, external casing packers to selectively isolate the wellbore and slotted liner completion with selectively isolating zones. Babu (1989) looked upon the problem as a partially penetrating vertical well, which is turned sideways. The derived pseudo-steady productivity equation is

qh = 0.007078b k x k z p /( µ o Bo ) A1 ln[ ]+ lnC H - 0. 75 + s R rw

(4.15)

where b sR CH kx

= extension of the drainage volume in the direction along the well axis, ft = skin factor due to partial penetration. = geometric shape factor defined by Babu (1989) = permeability in the horizontal plane perpendicular to the well axis, md

PIPESIM

Field Equipment kz A1 rw = permeability in the vertical plane, md = drainage area in the vertical plane, ft2 = wellbore radius, ft

101

The equation is derived from a very complex general solution. It requires the calculation of the CH and SR. The geometric shape factor accounts effect of permeability anisotropy, well location and relative dimensions of the drainage volume. The skin accounts for the restricted entry associated with the well length. Babu (1989) reported an error of less than 3% when compared to the more rigorous solution. 4.2.7 Solution Gas-Drive IPR Cheng (1990), Joshi (1991) and Bendakhlia (1989) have studied the inflow performance relationship (IPR) for solution gas-drive reservoirs. Bendakhlia followed the same approach used by Vogel for vertical wells and developed the following equation

qo q o,max = [1- V( p wf p ) - (1- V)( wf ) 2 ]n pR pR

(4.16)

Equation (4.16) can be used under the assumptions of Vogel's original IPR correlation. The parameter V and n were correlated as a function of recovery factor. 4.2.8 Horizontal Gas Wells The preceding sections have dealt with oil flow. However, horizontal wells are also appropriate for gas reservoirs. For example, in highpermeability gas reservoirs wellbore turbulence limits the deliverability of a vertical well. The most effective way, according to Joshi (1991), to reduce gas velocity around the wellbore is to reduce the amount of gas production per unit well length which can be accomplished by horizontal wells. Joshi (1991) describes two methods for the relationship between pressure and flowrate.

· ·

The gas flowrate is proportional to the pressure square terms. Al-Hussainy (1966) defined a pseudo-pressure m(p). The gas flowrate is directly proportional to the pseudo-pressures which is defined as

PIPESIM

102

Field Equipment

m( p ) = 2

p

0

p dp µz

(4.17)

Joshi (1991) did a comparison of the two methods. Below reservoir pressures of 2500 psia, either method can be employed. However, above 2500 psia, the pseudo-pressure should be used. The steady-state equation for gas flow is

0.007027k h h(p e - p wf ) qh = r ln[ e ]µZT rw,eff

2 2

(4.18)

where qh = gas flowrate, mmscf/day pe = pressure at external radius, psia pwf = wellbore flowing pressure, psia kh = horizontal permeability, md h = reservoir height, ft re = drainage radius, ft rw,eff = effective wellbore radius, ft µ = average viscosity, cp Z = average compressibility factor T = reservoir temperature, oR The pseudo-steady state gas flow equation can be written as follows (Joshi, 1991)

qh = 0.007027kh(p r - p wf ) r [ln[ e ]- 0.75 + s + s m + s ca - c + Dq h ]µZT rw 2.222x10 -15 ( g k a h )

2 2

(4.19)

D=

µ pwf rw h

2 p

(4.20) (4.21) (4.22)

= 2.73x1010 k

-1.1045 a

or

= 2.33x1010 k

-1.201 a

PIPESIM

Field Equipment where qh pr pwf s sm sca c k h re rw µ Z T µpwf g hp ka

103

= gas flowrate, mmscf/day = average reservoir pressure, psia = wellbore flowing pressure, psia = negative skin due to horizontal well = mechanical skin damage = shape related skin factor = shape fact conversion constant = permeability, md = reservoir height, ft = drainage radius, ft = wellbore radius, ft = average viscosity, cp = average compressibility factor = reservoir temperature, oR = viscosity at well flowing conditions, cp = high velocity flow coefficient, 1/ft = gas gravity = perforated interval, ft = permeability in the near wellbore region, md

Equation (21) and (22) are from Golan (1986) and Brown (1984), respectively. The above equations are based upon circular drainage area. The turbulence term, Dq, accounts for the extra pressure drop in the near wellbore region due to the high gas velocity. This term was neglected when dealing with oil flow. In addition, the term makes the solution of equation (19) iterative. 4.3 Multiple Layers / Completions Multiple layers can be modeled with PIPESIM. Each layer can have the following, different, properties; · Static Pressure · Temperature · Depth · IPR specification · Fluid description

PIPESIM

104

Field Equipment

The IRR for each individual layer can be specified using any of the standard completion options (described above). Similarly, the fluid description for each individual layer can be specified using the standard black oil or compositional fluid descriptions. PIPESIM performs the fluid mixing in the wellbore and also calculates inter layer pressure drops. 4.4 Artificial Lift Artificial lift is the process of assisting the production of fluids from the reservoir by reducing the static head in the well bore. There are a number of methods available for doing this; · Gas Lift · Electrical Submersible pumps (ESP) · Rod Pump Given their wider operating range and wider established application in the oil and gas industry, the modeling of artificial lift in PIPESIM has been limited to gas lift and ESP. 4.4.1 Gas Lift Gas lift can be described as a simple single injection point or by defining the gas lift valves as equipment in the tubing description. With the single injection point description, the user explicitly specifies the injection gas flowrate (and no details of the gas lift valves or ports are required). In this mode of operation it is assumed that the casing pressure is sufficient to inject all the lift gas at the specified depth. Alternatively, if gas lift valves are described as part of the tubing description, then PIPESIM will calculate the injection gas throughput for each valve (dependent on the casing, tubing and dome pressures and valve temperature) PIPESIM contains a database of gas lift valve details for most of the commonly used gas lift valves from various manufactures.

PIPESIM

Field Equipment 4.4.2 ESP Lift ESP's are modeled via an ESP performance curve that shows the relationship between flowrate, head and efficiency. This data is supplied at a set pump speed and number of stages. The most common ESPs used in the oil & gas industry have been made available within PIPESIM via a database. The manufacturers covered are: · Reda · ODI · Centrilift · Ramco Alnas · Trico For each manufacturer a number of models are available. The user can vary the following for each ESP; · Speed · Number of stages · Head factor to match the exact ESP in-situ.

105

In addition the user can extend the database by adding new ESP's curve data in the form of flowrate, head and efficiency. 4.5 Tubing The production of the fluids from the reservoir to the surface is via a series of tubing strings. The tubing allows the modeling of; · Straight tubing · Deviated tubing · Changes in pipe diameter · Tuning, Annular or Tubing and Annular flow · Gas lift injection (single and multi-point) · ESP lift point · Down hole equipment (SSSV, choke, separator, etc)

PIPESIM

106

Field Equipment

4.6 Chokes The pressure drop through a restriction is based on the following; · Fluid properties computed from upstream pressure · Heat capacities of the two phases computed from the upstream conditions The sonic velocity if the fluid is then computed from the heat capacity ratio, Cp/Cv. If the actual throat velocity is greater than the sonic velocity then the flow is critical. If it is less then it is sub-critical. The correlations used in each regime can be selected. Note: The downstream pressure can not be determined in the case of critical flow. If critical flow is determined in the case where the outlet pressure has been specified then the choke downstream pressure is computed from the flowrate and the outlet pressure. 4.6.1 Ashford-Pierce The correlation of Ashford and Pierce [1975] is valid for critical and sub-critical flows.

qo = 351Cd e2 .

= ( Bo + Fwo )

-1

2

n -1 n T1 z1 ( R - Rs ) 1 - e n + 198.6 p1 (1 - e) × 0 + 0.000217 g Rs + Fwo w n - 1 = -1 T1 z1 ( R - Rs )e n 0 + 0.000217 g R + Fwo w 198.6 + p1

[

]

1

2

[

]

where qo C de

PIPESIM

- oil flow rate at standard conditions (bbl/d) - choke discharge coefficient - choke diameter (64th in.)

Field Equipment Fwo Bo n p1 p2 R Rs T1 z1 e g o w - Water to oil ratio (WOR) - oil formation factor volume factor (bbl/STB) - specific heat ratio - upstream choke pressure (lb/ft2) - downstream choke pressure (lb/ft2) - producing GOR (scf/STB) - solution GOR at p1 and T1 (scf/STB) - upstream choke temperature (oR) - gas compressibility factor at T1 and p1 - choke downstream to upstream pressure ratio, p2/p1 - gas specific gravity at T1 and p1 - oil specific gravity at T1 and p1 - water specific gravity at T1 and p1

107

Assumptions: · polytropic expansion of gas-liquid mixture · equal gas and liquid velocities at the throat · incompressible liquid phase · liquid dispersed in a continuous gas phase · negligible friction losses Recommended values for discharge coefficient (C) are: Choke size (64th in.) 32 24 20 12 8 C 0.95 0.95 0.976 1.2 1.2

4.6.2 Omana The correlation of Omana [1969] is valid for critical flow. The original equation is:

N qL = 0.263 N -3.49 N Pl 3.19 Qd 0.657 N D1.8

where

PIPESIM

108

N qL

Field Equipment

= 184q L . L

o L 1.25

N =

G

L

1

N pl = 174 × 10 -2 P1 .

L L

Qd =

1 1 + R1

N D = 120.872 Dc

L L

Final re-arranged equation:

o q L = 1953 × 10 -3 ( L ) .

-1.245

( L )1.545 (1 + R1 )

-0.657

( Dc )1.8 ( G ) -3.49 ( P1 ) 3.19

NqL ND Npl Qd R1 Dc P1

- Omana liquid volume rate number - Omana diameter number - upstream pressure number - Omana dimensionless production number - density at upstream conditions(lb/ft3) - surface tension at upstream conditions (dynes/cm) - In situ GOR (ft3/ft3) - choke diameter (ft) - upstream pressure (psia)

subscripts G - gas L - liquid 4.6.3 Gilbert, Ros, Baxendall, Achong and Pilehvari The correlation proposed by Gilbert, Ros, Baxendall, Archong and Pilehvari [Ghassan, Maha, 1991] are valid for critical flow.

PIPESIM

Field Equipment The equations proposed are all of the form

o qL = aP (GOR) -b d c (1) 1 where P1 - upstream pressure (psia) o q L - liquid flow rate at standard conditions (STB/D) GOR - producing GOR (scf/STB) d - choke diameter (64ths in.) a,b,c - empirical coefficient given below

109

Correlation Gilbert Ros Baxendall Achong Pilehvari

A 0.1 0.05747 0.10460 0.26178 0.021427

B 0.546 0.5 0.546 0.650 0.313

c 1.89 2.00 1.93 1.88 2.11

4.6.3.1 PDVSA modification Recently a modification, by PDVSA, was made to equation (1) to incorporate another parameter "e" to better match their field data. For the all above correlation's e=1.

o qL = (aP (GOR) -b d c ) - e 1

This modification has been implemented in PIPESIM via the engine keyword tool. The parameters are proprietary. 4.6.4 Poettmann-Beck The correlation of Poettmann & Beck [1963] is valid for critical flow.

o qo =

5.61 o + 0.0765 G (GOR ) L

88992 Ac

0

.

1 9273.6 P1 0.4513( R1 + 0.766) . V1 (1 + 0.5m1 ) R11 + 0.5663

where

R =

1 1

0.00504T1 z1 ( GOR ) - ( Rs )1

0

(

)

P1 Bo

PIPESIM

110

m1 = 1 1 + R11

Field Equipment

1G 1L

V1 =

L

m1

q - oil flow rate (STB/D) Ac - choke cross-sectional area (ft2) P - pressure (psia) - specific gravity at P1 & T1 GOR - gas to oil ratio (scf/STB) Rs - Solution gas (scf/STB) B - formation volume factor - density (lb/ft3) T - temperature (oR) z - compressibility factor subscripts L G 1 o - liquid - gas - at upstream conditions - oil

Superscripts o - at standard conditions

4.6.5 Mechanistic Correlation, The mechanistic correlation, Brill & Beggs, is valid for critical and subcritical flows.

pTP = p L L + pG G qL p L = 2 g c 144 C L Ac

L

2

qG p G = 2 g c 144 YCG Ac

L

2

PIPESIM

Field Equipment

2 p - p1 d2 Y = 10 - 0.41 + 0.35 (1 / K ) 2 . p1 d1

111

C=

Cd d 1- 1 d2

4

Total pressure drop for the two-phase system is given by:

pTP 1 + CdL = p L G YC dG

2 - 1

where

d 4 qm p L = L 1 - 1 2 d 2 8083d1 CdL

G - no-slip fraction of free gas in the stream approaching the choke - no-slip fraction of liquid in the stream approaching the L choke qL - liquid flow rate (ft3/sec) qG - gas flow rate (ft3/sec) Ac - choke cross-sectional area (ft2) p1 - pressure upstream of choke (psi) p2 - pressure downstream of choke (psi) - density (lbm/ft3) C - flow coefficient Cd - discharge coefficient Y - compressibility factor d1 - upstream tubing diameter (same units as d2) d2 - orifice diameter (same units as d1) K - ratio of specific heats (cp/cv) Subscripts L - liquid G - gas

PIPESIM

112 TP 1 2

Field Equipment - two-phase - at upstream conditions - at downstream conditions

4.6.6 API 14-B Formulation The API 14-B formulation, Brill & Beggs, is similar to the mechanistic formulation, with the addition of the following assumptions and is valid for critical flow. 1) Liquid flow through the choke is incompressible. The discharge coefficient is constant with a value of 0.85. 2) Sub-critical gas flow through the choke is adiabatic and compressible. The discharge coefficient is constant with a value of 0.9. 3) Sub-critical two-phase compressible flow is described by weighting the liquid and gas orifice flow equations with the no-slip fraction of free gas G in the stream approaching the choke. 4) The density and flow rates of each phase can be replaced by a noslip mixture density, NL , and a total mixture flowrate, qm.

pTP 1 + CdL = p L G YC dG

2 - 1

where

d 4 qm p L = N 1 - 1 2 d 2 8083d1 CdL

CG = 0.9 CL = 0.85 Using the above equations we get:

. 1121 p tp = p L 1 + G 2 - 1 Y

where

p L = N d qm 1- 1 2 d 2 6870.55d1

4

PIPESIM

Field Equipment

113

- no-slip fraction of free gas in the stream approaching the G choke qm - total mixture flow rate (ft3/sec) PL - liquid phase pressure change (psi) PG - gaseous phase pressure change (psi) N - no-slip mixture density (lbm/ft3) CdG - discharge coefficient for the gas phase CdL - discharge coefficient for the liquid phase Y - compressibility factor d1 - upstream tubing diameter (same units as d2) d2 - orifice diameter (same units as d1) Subscripts L - liquid G - gas TP - two-phase 1 - at upstream conditions 2 - at downstream conditions 4.7 Heat transfer The effects of heat transfer in the well bore can be modeled by the use of an overall heat transfer coefficient. The heat transfer coefficient is relative to the outside pipe diameter. The surrounding ambient temperature can also be entered. 4.8 Reservoir Depletion The field planning module of PIPESIM can take into account the depletion of the reservoir over time. 4.8.1 Volume Depletion Reservoirs There is assumed to be no change in the reservoir volume occupied by hydrocarbons during depletion of the reservoir. The material balance equation, expressed at standard conditions for a given volume of production Gp and consequent drop in the average reservoir pressure p = p i - p is given by [Dake - 1978]

PIPESIM

114 Production = (sc) or

Gp = G - G E Ei

Field Equipment Gas Initially in Place (sc) Un-produced Gas (sc)

where: Gp is the cumulative production expressed at standard conditions G is the gas initially in place at standard conditions E is the gas expansion factor after cumulative production Gp Ei is the gas expansion factor at initially un-depleted reservoir conditions For fields units at standard conditions of p=14.17psia, T=520ºR and Z=1

E = 35.37 p ZT

and by using the equation of state for a real gas pV = ZnRT we can re-write the material balance equation as

p pi G p = 1 - Z Zi G

The initial conditions pi, Zi and G are input from the user The cumulative production, Gp, can be computed from the flow rate that the network module calculates, and the flowing time (time-step) specified. In the case of multiple wells in the tank Gp is simply the sum of the flow rates from wells in that reservoir over flowing time. The p/Z term can now be evaluated and correlations at reservoir pressure for the specified fluid composition can now be used to

PIPESIM

Field Equipment evaluate pressure for the (constant) reservoir temperature and volume.

115

The model assumes that the well flows at a constant rate between each time-step. 4.8.2 Gas Condensate Reservoirs The dry gas material balance as described above may be used to model gas condensate reservoirs. When the pressure falls below dew point, liquid hydrocarbons are deposited in the reservoir. Since FPT is a fully compositional simulator the new 2-phase z-factor for the reservoir will be automatically calculated. 4.9 References Ghassan, H. A., and Maha, R. A., "Correlations developed to predict two-phase flow through wellhead chokes", The journal of Canadian Petroleum Technology, Volume 30, N0. 6, 1991 F. H. Poettman and R. L. Beck, "New Charts Developed to Predict Gas-Liquid Flow through Chokes", World Oil, March 1963, 95-101. Two-Phase Flow in Pipes (Dr. James P. Brill, Dr. H. Dale Beggs), course notes, pp 6-8 through 6-12 Two-Phase Flow in Pipes (Dr. James P. Brill, Dr. H. Dale Beggs), course notes, pp 6-36 through 6-39 Al-Hussainy, R., Ramey Jr., H. J. and Crawford, P. B.: "The Flow of Real Gases Through Porous Media," JPT (1966) 624-636. Ashford, F.E. and Pierce, P.E. : "Determining Multiphase Pressure Drops and Flow capacities in Down-Hole Safety Valves", Journal of Petroleum Technology, Paper No. SPE-5161, September, 1975 . Babu, D. K. and Odeh, A. S.: "Productivity of a Horizontal Well," SPE Reservoir Engineering (November 1989) 417-421.

PIPESIM

116

Field Equipment

Bendakhlia, H. and Aziz, K.: "Inflow Performance Relationships for Solution-Gas Drive Horizontal Wells," SPE paper 19823 presented at the Annual Technical Conference and Exhibition, San Antonio, October 1989. Celier, G. C. M. R., Jouault, P. and de Montigny, O. A. M. C.: "Zuidwal: A Gas Field Development With Horizontal Wells," SPE paper 19826 presented at the Annual Technical Conference and Exhibition in San Antonio, October 1989. Cheng, A.M.: "Inflow Performance Relationships for Solution-GasDrive Slanted/Horizontal Wells," SPE paper 20720 presented at the Annual Technical Conference and Exhibition, New Orleans, September 1990. Cooper, R.E., and Troncoso, J.C.: "An Overview of Horizontal Well Completion Technology," SPE paper 17582 presented at the International Meeting on Petroleum Engineering, Tianjin, China, November 1988. Dake, L.P.: Fundamentals of Reservoir Engineering, Elsevier Scientific Publishing Co., New York, 1978. Dikken, B.J.: "Pressure Drop in Horizontal Wells and its Effect on Production Performance," JPT (November 1990) 1426-1433. Economides, M.J., McLennan, J.D., Brown, E., and Roegiers, J.C.: "Performance and Stimulation of Horizontal Wells," World Oil, (July 1989) 69-76. Folefac, A. N., Archer, J. S. and Issa, R. I.: "Effect of Pressure Drop Along Horizontal Wellbores on Well Performance," SPE paper 23094

PIPESIM

Field Equipment presented at the Offshore Europe Conference held in Aberdeen (September 1991).

117

Ghassan, H. A., and Maha, R. A., "Correlations developed to predict two-phase flow through wellhead chokes ", The Journal of Canadian Petroleum technology, Volume 30, No 6, 1991 Giger, F. M., Reiss, L. H., and Jourdan, A. P.: "The Reservoir Engineering Aspects of Horizontal Drilling," SPE paper 13024 presented at the Annual Technical Conference and Exhibition in Houston, September 1984. Goode, P. A. and Wilkinson, D. J.: "Inflow Performance of Partially Open Horizontal Wells," SPE paper 19341 presented at the SPE Eastern Regional Meeting, Morgantown, WV, October, 1989. Gurley, D. G., Copeland, C. T. and Hendrick, J. L.: "Design Plan and Execution of Gravel-Pack Completion," J. Pet. Tech. (Oct. 1977). Jones, L. G. and Slusser, M. L.: "The Estimation of Productivity Loss Caused by Perforation - Including Partial Completion and Formation Damage," SPE paper 4798 (1974). Joshi, S. D.: Horizontal Well Technology, Penwell Publishing Company, Tulsa, Oklahoma (1991). Joshi, S. D.: "A Review of Horizontal and Drainhole Technology," SPE paper 16868 presented at the Rocky Mountain Regional Meeting in Casper, WY (May 1988). Lockhart, R. W. and Martinelli, R. C.: "Proposed Correlation of Data for Isothermal Two-phase, Two-Component Flow in Pipes," Chem. Eng. Prog. (January 1949) 45, 39.

PIPESIM

118

Field Equipment

McLeod, H. O.: "The Effect of Perforating Conditions on Well Performance," JPT (Jan. 1983). Muskat, M.: The Flow of Homogeneous Fluids Through Porous Media, I.H.R.D.C., Boston (1937). Mutalik, P. N., Godbole, S. P. and Joshi, S. D.: "Effect of Drainage Area Shapes on Horizontal Well Productivity," SPE paper 18301 presented at the Annual Technical Conference and Exhibition, Houston (October 1988). Omana, R. et al., "Multiphase Flow Through Chokes", SPE 2682, 1969 Poettman, F. H. and Beck, R. L. "New Charts Developed to Predict Gas-Liquid Flow Through Chokes", World Oil, March 1963, 95-101 Pots, B. F. M., Bromilow, I. G. and Konijn, M. J. W.: "Severe Slug Flow on Offshore Flowline/Riser Systems," SPE paper 13723, (March 1985). Renard, G. I. and Dupuy, J. M.: "Influence of Formation Damage on the Flow Efficiency of Horizontal Wells," SPE paper 19414 presented at the Formation Damage Control Symposium, Lafayette (February 1990).

PIPESIM

Field Equipment

119

5 Field Equipment

5.1 Compressor The basic compressor model uses centrifugal and reciprocating compressor equations to determine the relationship between inlet pressure and temperature, outlet pressure and temperature, flowrate, power, and efficiency. It is also possible to use built in, or user developed compressor curves to describe the relationship between differential pressure, flowrate, and efficiency for a range of compressor speeds. If compressor curves are used, therefore, the compressor speed and number of stages become a additional factors. At least one parameter must be supplied. This could be: · outlet pressure · differential pressure · pressure ratio (Pout/Pin) · power (shaft power) · speed and number of stages (if using curves) The remaining quantities will then be calculated using compressor equations. If more than one value is supplied, then the parameter which leads to the smallest compressor differential pressure will be used, and all other supplied parameters will be discarded. The main centrifugal compressor equations used are as follows: Adiabatic Route Head = (ZavgRTin/(M(k-1)/k))((Pout/Pin)((k - 1)/k) - 1) where k = Cp/Cv Polytropic Route Head = (ZavgRTin/(M(n-1)/n))((Pout/Pin)((n - 1)/n) - 1) where n = 1/(1 - ((Cp/Cv - 1)/(eCp/Cv))) Mollier Route (compositional cases only) Head (Hout - Hin)

PIPESIM

120

Field Equipment

where the values of Hout and Hin are obtained from isentropic compression from Pin to Pout 5.2 Expander The basic expander model uses centrifugal expander equations to determine the relationship between inlet pressure and temperature, outlet pressure and temperature, flowrate, shaft power, and efficiency. It is also possible to use built in, or user developed expander curves to describe the relationship between differential pressure, flowrate, and efficiency for a range of expander speeds. If expander curves are used, therefore, the expander speed and number of stages become a additional factors. At least one parameter must be supplied. This could be: · outlet pressure · differential pressure · pressure ratio (Pin/Pout) · power (shaft power) · speed and number of stages (if using curves) The remaining quantities will then be calculated using centrifugal expander equations. If more than one value is supplied, then the parameter which leads to the smallest expander differential pressure will be used, and all other supplied parameters will be discarded. The main expander equations used are as follows: Adiabatic Route Head = (ZavgRTin/(M(k-1)/k))((Pout/Pin)((k - 1)/k) - 1) where k = Cp/Cv Polytropic Route Head = (ZavgRTin/(M(n-1)/n))((Pout/Pin)((n - 1)/n) - 1) where n = 1/(1 - ((Cp/Cv - 1)/(eCp/Cv))) Mollier Route (compositional cases only) Head (Hout - Hin)

PIPESIM

Field Equipment where the values of Hout and Hin are obtained from isentropic compression from Pin to Pout

121

5.3 Single Phase Pump The basic pump model uses centrifugal pump equations to determine the relationship between inlet pressure and temperature, outlet pressure and temperature, flowrate, shaft power, hydraulic power and efficiency. It is also possible to use built in, or user developed pump curves to describe the relationship between differential pressure, flowrate, and efficiency for a range of pump speeds. If pump curves are used, therefore, the pump speed and number of stages become a additional factors. At least one parameter must be supplied. This could be: · outlet pressure · differential pressure · pressure ratio (Pout/Pin) · power (shaft power) · speed and number of stages (if using curves) The remaining quantities will then be calculated using centrifugal pump equations. If more than one value is supplied, then the parameter which leads to the smallest pump differential pressure will be used, and all other supplied parameters will be discarded. The main pump equations used are as follows: Hydraulic Power Flowrate x Differential Pressure

Hydraulic Power = Shaft Power x Efficiency 5.4 Multiphase Boosting Multiphase boosting technology (also referred to as multiphase pumping technology) for the oil and gas industry has been in development since the early 1980s, and is now rapidly gaining acceptance as a tool to optimize multiphase production systems [1]. Particularly for the development of satellite fields, multiphase

PIPESIM

122

Field Equipment

boosting has been recognized as a promising technology: rather than separation, gas compression, liquid pumping and use of dual flow lines back to the host facility, multiphase boosting enables the full (non-separated) well stream to be boosted in a single machine. Besides the thus realized simplification of the production system, the potential cost reductions could make development of marginal fields economic. Since 1990, well over one hundred multiphase boosters have been installed worldwide, with the vast majority of the installations based onshore or offshore topsides. Over the years, the development of multiphase boosting has led to three types of boosters being commercially available: - twin screw type multiphase boosters - progressing cavity type multiphase boosters - helico-axial type multiphase boosters The first two types mentioned belong to the category of positive displacement type pumps and the third type to the category of dynamic type pumps.

PIPESIM

Field Equipment

Traditional Approach The incoming fluid is separated in its constituent gas and liquid phases. The separated liquids are pumped up to the required pressure and exported via the liquid export line. Separated gas is compressed up to the required pressure and exported via the gas export line. Alternative Approach The incoming fluid is separated in its constituent gas and liquid phases. The separated liquids are pumped up to the required pressure and separated gas is compressed up to the required pressure, before the two phases are recombined and exported via a multiphase export line. Multiphase Boosting The incoming fluid is directly boosted up to the required pressure without separation of the gas and liquid phases, and exported via a multiphase export line.

123

Figure 5-1 Multiphase boosting vs. Traditional approaches Multiphase boosters are pumps/compressors that can accommodate fluids composed of 100% liquid to 100% gas, and anywhere in between. Although commonly referred to as multiphase pumps, the terminology used in this document is `multiphase booster' to recognize the fact that also 100% gas can be handled by this equipment (albeit with some restrictions, as outlined in later chapters of this document). Figure 3.1 depicts the difference between multiphase boosting technology and the more traditional technology of separation, pumping and compression. The rationale for employing multiphase boosters stems from two basic factors: (1) Production Enhancement ­ accelerated and/or incremental hydrocarbon production as a result of lowering the backpressure on the well(s);

PIPESIM

124

Field Equipment

(2) Pressure Boosting ­ increasing fluid pressure for transportation over long distances or to move fluid from low pressure systems to higher pressure systems. In many cases, there will be a combined effect of the two factors, e.g. lowering the backpressure on a well by use of a multiphase booster provides at same time a higher pressure available at the inlet to the flowline. To demonstrate the principle of multiphase boosting, take the example of a well which is connected via a flowline and riser to the inlet separator on the host facility. See Figure 5-2.

Figure 5-2 Simplified production system Based on estimates of the pressure drop across the tubing string, and given the production characteristics of the formation and the IPR of the well, the curve of tubing-head pressure pth against rate for an individual well can be obtained; this curve is known as the tubinghead pressure (THP) curve. Similarly, based on estimates of the pressure drop across flowline and riser, and given the pressure at the inlet separator of the host facility, the curve of required flowline inlet pressure against rate can be obtained; this curve is known as the outflow curve.

PIPESIM

Field Equipment

125

Figure 5-3 demonstrates the principle of tubing-head pressure curve and outflow curve; the point of intersection of the two curves is the system operating point, i.e. pressure and production rate at the wellhead.

P ro d u c tio n S y s te m A n a ly s is

7 0 .0

6 0 .0

O u tflo w c u r ve

Pressure at wellhead (bara)

5 0 .0

4 0 .0

3 0 .0

T H P c u r ve

2 0 .0

1 0 .0

0

0

5 .0

1 0 .0

1 5 .0

2 0 .0

2 5 .0

P ro d u c tio n ra te (kg /s)

Figure 5-3 Production system analysis: THP curve and outflow curve From Figure 3.3, it can be seen that the system operating point involves a tubing head pressure of 39 [bara] and production rate of 5 [kg/s]. We can however also see from the THP curve that the flowing potential of the well is far greater than the production rate of 5 [kg/s], should the back pressure on the well be lower than the 39 [bara]. Assuming we could install a booster that allows us to provide a `boost' of 20 [bar] to the well fluids directly downstream of the wellhead, the outflow curve shown in Figure 5-3 will change to that shown in Figure 5-4. The new system operating point involves a tubing head pressure of 24 [bara] and production rate of 10 [kg/s], i.e. through the boosting of the well stream production has increased by 100%.

PIPESIM

126

Field Equipment

Production System Analysis

70.0

60.0

Outflow curve - No boosting

Pressure at wellhead (bara)

50.0

40.0

30.0

Outflow curve - Boosting 20 bar

20.0

10.0

THP curve

0 5.0 10.0 15.0 Production rate (kg/s) 20.0 25.0

0

Figure 5-4 Production system analysis: the effect of multiphase boosting visualized Through the type of analysis outlined in Figure 5-3 and Figure 5-4, the effect of multiphase boosting on production system operating point (tubing head pressure, production rate) can readily be established, as can be the multiphase booster operating point and power requirement. Further details of this analysis, in particular with respect to the system analysis tool PIPESIM, are given in Chapter 3. 5.4.1 Multiphase Boosters ­ Positive Displacement Type Positive displacement type pumps work on the basis of pressure being added hydrostatically rather than dynamically, which results in these pumps being less sensitive to fluid density than dynamic type pumps. As a result of this, positive displacement type pumps appear to figure higher in surface applications than dynamic type pumps, because with surface applications fluids tend to show higher gas fractions and a greater tendency for density change than in subsea applications [2]. Although initially piston type pumps were also considered for use as multiphase boosters, the commercial development of positive displacement has concentrated on two types only: (1) twin screw type multiphase booster (2) progressing cavity type multiphase booster The majority of positive displacement type multiphase boosters on the market are of the twin screw type, with the remainder being of the progressing cavity type. Within the Shell EP Group of Operating

PIPESIM

Field Equipment

127

Companies, no progressing cavity type multiphase boosters have been installed thus far. This chapter will therefore predominantly address the working principle of twin screw type multiphase boosters, but mention will be made of the progressing cavity type also. 5.4.2 Twin Screw Type Multiphase Boosters The twin screw type booster, also referred to as two-spindle screw pump, works on the basis of liquid carried between the screw threads of two intermeshing feed screws and displaced axially as the screws rotate and mesh. In principle, the intermeshing screws form chambers [3], which are: - filled with fluid at the pump suction side; - closed to capture the amount of fluid that has entered the chamber at pump suction; - transported to the discharge side of the pump; - opened to the outlet system once the chamber has reached the pump discharge port. Figure 5-5 shows an example of a twin screw type pump.

Figure 5-5 Twin screw type pump It should be noted that, unlike screw type compressors, the volume of the chambers is not reduced on its way from pump suction side to pump discharge side, i.e. there is no in-built compression in the twin screw type multiphase boosters. Pressure build-up by the twin screw type multiphase booster is entirely caused by the fact that a definite amount of fluid is delivered into the outlet system with every revolution of the feed screws; the pressure developed at pump discharge is thus solely the result of resistance to flow in the outlet system. However, due to the pressure differential between pump discharge and pump suction, an internal leakage in the pumping elements

PIPESIM

128

Field Equipment

results and causes a pressure gradient across the moving chambers. This internal leakage causes the pump net flow to be less than its theoretical capacity, as demonstrated in pump performance curves (see Figure 5-6).

Twin Screw Multiphase Pump - Performance Curve (valid for GVF=0%, p1=1 bara)

Shaft power [kW] 600 Flow rate [m3/h] 400 200 0 0 10 20 30 40 50 60 70 Pump differential pressure [bar] 1500 1000 500 0

Flow rate Shaft power

Twin Screw Multiphase Pump - Performance Curve (valid for GVF=85%, p1=1 bara)

Shaft power [kW] 600 Flow rate [m3/h] 400 200 0 0 10 20 30 40 50 60 70 Pump differential pressure [bar] 1500 1000 500 0

Flow rate Shaft power

Figure 5-6 Pump performance curves (typical) As can be seen from Figure 5-6, pump flow rate is dependent on pump differential pressure: the higher the pump differential pressure, the higher the internal leakage, and thus the lower pump flow rate. The theoretical capacity of the pump, i.e. the flow rate if no internal leakage is present, is the flow rate found for zero pump differential pressure ­ for the pump represented in Figure 6, the theoretical flow rate is 500 m3/h; the difference between theoretical flow rate and actual flow rate is the internal leakage, also called `pump slip'. As an example, for the pump represented in Figure 6, GVF=0%, the actual flow rate for pump differential pressure 40 bar is 400 m3/h, i.e. pump slip is (500 ­ 400) = 100 m3/h. Given the relative insensitivity of flow rate to differential pressure, especially for higher GVF, the twin screw multiphase booster is sometimes referred to as a `constant flow rate' pump.

PIPESIM

Field Equipment

129

As can also be seen from Figure 5-6, pump flow rate is dependent on GVF also, whereas the effect of GVF on pump shaft power is less pronounced. Whereas Figure 3.6 may suggest that an unlimited variety of twin screw multiphase pumps is available to cover an unlimited amount of (differential pressure / flow rate)-combinations, in practice however a number of physical limitations applies: pump differential pressure is typically limited to 70 bar to avoid excessive deflection of feed screws and possible contact between rotating screws and stator housing; pump flow rate (total volumetric flow rate at pump suction) at present limited to approximately 2000 m3/h per single pump; gas volume fraction at pump suction typically limited to 95% maximum (for GVF>95%, some form of liquid re-circulation is typically required to maintain GVF-suction at 95% maximum); pump inlet pressure and outlet pressures restricted by casing design pressure and seal design pressure. 5.4.3 Progressing Cavity Type Multiphase Boosters The progressing cavity type pump (also known as single-rotor screw pump) operates on the basis of an externally threaded screw, also called rotor, turning inside an internally threaded stator (see Figure 7); the most simple configuration is the one whereby there is one lead on the rotor and two leads on the stator, commonly referred to as a 1:2 ratio element profile. Other configurations are also feasible, provided that the stator has one more lead than the rotor [4].

Figure 5-7 Moyno® progressing cavity pump

PIPESIM

130

Field Equipment

As with the screw type pump, as the rotor rotates within the stator, chambers are formed and filled with fluid and progress from the suction side of the pump to the discharge side of the pump conveying the process fluid. The continuous seal line between the rotor and the stator helix keeps the fluid moving steadily at a fixed flow rate proportional to the pump rotational speed. Application of the progressing cavity type pump for multiphase boosting has been less widespread than the twin screw type multiphase booster, and flow rates and differential pressures are typically lower than those achievable with the twin screw type. Claimed to be the largest progressing cavity type pump for multiphase applications is Moyno's R&M Tri-Phaze® System, capable of transferring multiphase flows up to 29,000 bbl/day (192 m3/h) at differential pressures up to 300 psi (20.7 bar). Through the installations of various pumps in series/parallel arrangement, higher flow rates and higher differential pressures are achievable, however at the expense of complexity [4]. Given their wider operating range and wider established application in the oil and gas industry, the modeling of positive displacement type multiphase boosters in PIPESIM has been limited to the twin screw type multiphase booster only. 5.4.4 Multiphase Boosters ­ Dynamic Type Dynamic type pumps work on the principle of pressure being raised by adding kinetic energy to the fluid, which is then converted to pressure. The actual increase in pressure is directly proportional to the density of the pumped fluid, i.e. the higher the fluid density, the higher the pressure increase. Because of this, dynamic type pumps are more sensitive to fluid density than positive displacement type pumps, and tend therefore to be used in applications with lower maximum gas volume fractions than positive displacement type pumps, e.g. in subsea applications. The commercial development of dynamic type multiphase boosters has concentrated on the helico-axial type, based on helico-axial hydraulics developed and licensed by Institute François du Petrole (IFP). For very high gas volume fractions (GVF>95%), there is also

PIPESIM

Field Equipment the contra-rotating axial (CRA) machine, originally developed by Framo Engineering AS and Shell.

131

The design of the helico-axial type pump has further concentrated on the driver mechanism for subsea use, and led to the availability of electric motor driven units as well as hydraulic turbine driven units. For onshore or offshore topsides applications, other driver types can also be used. 5.4.5 Helico-Axial Type Multiphase Boosters The helico-axial type multiphase booster features a number of individual booster stages, each consisting of an impeller mounted on a single rotating shaft, followed by a fixed diffuser. In essence, the impeller imparts kinetic energy to the fluid, which is converted to pressure in the diffuser. The impeller blades have a typical helical shape, and profile of the open type impeller and diffuser blade arrangement are specifically designed to prevent the separation of the multiphase mixture inside the pump [5]. Figure 5-8 shows an example of a helico-axial pump stage.

Figure 5-8 Helico-axial pump stage The boosting capabilities of the helico-axial type booster are a function of GVF-suction and suction pressure, as well as speed, number of impeller stages and impeller size. See Figure 5-9. The quoted flow rates and speed limitation represent present technology status.

PIPESIM

132

Field Equipment

Figure 5-9 Helico-axial type multiphase booster ­ Pressure boosting potential As can be seen from Figure 3.9, the pressure boosting capability drastically reduces for higher GVF. Also, for reduced speed or reduced number of stages, the pressure boosting capability will be less than the maximum shown in Figure 3.9. For a given pump with given number of stages, speed and impeller diameter, pump performance curves can be provided as shown in Figure 3.10. These curves are valid for given GVF-suction, p-suction and fluid density only; for differing GVF-suction, p-suction and fluid density, new performance curves will apply.

PIPESIM

Field Equipment

133

Booster differential pressure

Helico-axial type multiphase booster - Performance curve (valid for given GVF, p -suction and fluid density)

Best efficiency line

Maximum booster differential pressure

ax

.D

P

lin

e

90%

80%

Ma

M

xim

spee

spee

Figure 5-10 Pump performance curve (typical) Practical operating limits of the helico-axial type multiphase booster are [6]: pump differential pressure typically limited to 70 bar pump flow rate (total volumetric flow rate at pump suction) at present limited to approximately 1500 m3/h per single pump; gas volume fraction at pump suction typically limited to 95% maximum; pump inlet pressure 3.4 bara minimum; pump outlet pressure restricted by casing design pressure and seal design pressure. 5.4.6 Contra-Rotating Axial Type Multiphase Booster The CRA operates on the basis of axial compressor theory, but rather than having one rotor and a set of stator vanes, the CRA employs two contra-rotating rotors. The inner rotor consists of several stages mounted on the outside of an inner cylinder. The outer rotor consists of several stages on the inside of a concentric, larger diameter cylinder. See Figure 5-11.

im u M in ms d pee

um

d

Total volumetric flow rate at suction

d

s pee d

PIPESIM

134

Field Equipment

Figure 5-11 Contra-rotating axial (CRA) compressor The exact mechanism underlying pressure build-up inside the CRA compressor have not yet been fully understood, nor are sufficiently mature design rules available for the scale-up of CRA performance to larger flow rates. Flow rates that can be handled by the CRA are of same order of magnitude as for helico-axial type multiphase booster, however achievable differential pressures (maximum 20 bar) and realized efficiencies (approximately 25%) are significantly less than what's achievable with conventional boosting systems. Given their wider operating range and wider established application in the oil and gas industry, the modeling of dynamic type multiphase boosters in PIPESIM has been limited to the helico-axial type multiphase booster only. 5.4.7 Alternative approach The alternative approach described in Figure 5-1 has also been implemented in PIPESIM. This generic booster splits the fluid into liquid and gas and pumps the liquid and compresses the gas. Efficiency values for the compressor efficiency have been obtained from field data and are available in the help system.

PIPESIM

Field Equipment

135

5.5 Separator Placing a separator in the model removes up to 100% (by volume) of the gas, water or liquid (oil plus water) phase. The % efficiency (or efficiency fraction) refers to the amount of that material removed. For example, a 90% efficient water separator removes 90% of the water. From that point onward, flow of the remaining fluids will be modeled. 5.6 Re-injection point Works in conjunction with a separator in a network model only. All the fluid removed from the separated will be re-injected. The following must be defined; · The incoming, outgoing and separated branches. · Separated stream inlet temperature if different from the separator temperature · An estimate of the flowrate for the separated stream. 5.7 Heat Transfer 5.8 References [1] How multiphase pumping can make you money K.C.Oxley, J.M. Ward, W.G. Derks Paper presented at Facilities 2000 Conference, New Orleans 1999 [2] Success grows in pumping high-gas-fraction multiphase fluids B. Butler Petroleum Engineer International, July 1999 Pump Handbook, 2nd edition J. Karassik et al. McGraw-Hill Inc., 1986 Progressing cavity multiphase pumping systems: expanding the possibilities K.Z. Mirza Paper presented at BHR Conference Multiphase '99

[3]

[4]

PIPESIM

136 [5]

Field Equipment Innovations in multiphase hydrocarbon operations C. de Marolles, J. de Salis Article from www.pump-zone.com, 1999 Satellite multiphase boosting ­ Multiphase boosting study Siep-RTS, ABB Lummus Global Shell report SIEP 98-5463

[6]

PIPESIM

Operations THIS PAGE LEFT BLANK INTENTIONALLY

137

Operations 6 Operations The operations of PIPESIM available for each module are · Pipeline & facilities module · Check model · No operation · Run model · System analysis · Pressure Temperature profile · Flow correlation matching · Wax prediction · Well Performance module · Check model · No operation · Run model · System analysis · Pressure Temperature profile · Flow correlation matching · Nodal analysis · Reservoir tables · Artificial lift analysis · Well Performance Curves · Network module · Check model · Run model · Restart model · Abort run 6.1 Check model Allows the model to be check for missing input data input before a simulation is performed.

139

6.2 No operation Allows a model to be built and saved with no associated operation. This is mainly for use with Schlumberger's Production data management software ProdMan.

PIPESIM 2000

140 6.3 Run model Run the selected operation.

Operations

6.4 System Analysis The systems analysis operation enables the user to determine the performance of a given system for varying operating conditions on a case-by-case basis (4.f. Pressure/Temperature Profiles where performance is evaluated on a point-by-point basis). Results of the system analysis operation are provided in the form of plots of a dependent variable (e.g. outlet pressure) versus an independent variable (e.g. flow rate). Families of X-Y curves can be generated for the system by varying either a single sensitivity variable (e.g. watercut) or through permutations of a group of sensitivity values. The ability to perform analysis by combining sensitivity variables in different ways makes the system analysis operation a very flexible tool for plotting data on a case-by-case basis. A typical systems analysis type plot is shown below.

Outlet Pressure

Watercut=30% Watercut=60% Watercut=90% Flow Rate

Figure 6.1 Typical Systems Analysis Plot 6.5 Pressure Temperature profile Pressure and temperature profiles of the system can be generated as a function of distance and along the system. Both temperature and pressure profiles are generated on a node-by-node basis for the system 6.6 Flow correlation matching This option allows the user to match well test data against each correlation for a particular system, hence allowing the most suitable correlation to be determined for each system model.

PIPESIM 2000

Operations

141

6.7 Wax Prediction The wax prediction operation in PIPESIM was is at present, only available to Shell (and Shell approved companies) and to BP (and BP approved companies). 6.8 Nodal Analysis PIPESIM has been designed as a nodal analysis tool so, rather than just provide single point solutions to individual flow problems, the model allows the user to perform sensitivity studies and generate system performance curves. Such graphical system analysis techniques are essential in well performance modeling and in optimizing the design of complex pipeline systems. This comprehensive nodal analysis capability has been achieved without compromising the rigorous finite element solution techniques necessary in generating accurate pressure and temperature profiles throughout the system. In essence, the objective of nodal analysis is to combine the various components of a given oil or gas production or transportation system in order to optimize the various components in the system. This is done by splitting the system at the point of interest known as the nodal analysis point and performing a solution for pressure at the nodal analysis point on the upstream (Inflow) and downstream (Outflow) sub-systems. The point at which there is no pressure differential at the nodal analysis point for the sub-systems is known as the operating point for the given system. This can be represented graphically by the intersection point of the inflow and outflow performance curves as shown in Figure 3.1. Optimization of the system is conducted by investigating the effect on the operating point of varying key system parameters.

PIPESIM 2000

142

Operations Inflow

Pressure

Outflow

NA Point

Flowrate Figure 6.2 Nodal Analysis Inflow/r7 T5rvesPoint

CBT/TT4

1

Tf0.000nNA Pcan be at anymp(NA P(betTm(1 Tf6.0004 Tw 13.98 0 0 13.06j1 465m414.620(ween components) in thensystem Tm( )T cSEMC/P <</4CID 0 >>BDCC2_1/TT4 1 Tf0.0000 Tc 0 Tw 13.98 0 0 13.)Tj6 54.363184.759<08Tj0>m(CBT/TT4 1 Tf9.0007 Tc -1.0051 Tw 13.98 0 0 1108 Tf254.363184.7596 Tm(Insensi/ReityPcan flow )Tj0.0012 Tc 22.0051 Tw 13.98 0 0 1239.52 694.363184.7596be er anymiTm(Invariabl142 )Tj13.98 0 0 138 l17394.363184.7596e Tm( )T cSEMC/P <</5CID 0 >>BDCC2_1/TT4 1 Tf0.0000 Tc 0 Tw 13.98 0 0 13.)Tj6 54.982796.759<08Tj0>m(CBT/TT4 1 Tf1.0007 Tc -3.0051 Tw 13.98 0 0 1108 Tf254.982796.7599 Tm(Outsensi/R142 )Tj13.98 0 0 1184.032954.982796.7599v142 )Tj13.98 0 0 13.84052 54.982796.759(ityPcan benflow )Tj0.0007 Tc 0.0051 Tw 13.98 0 0 12696 5MCI4.982796.759(er anymoTm(Outvari142 )Tj13.98 0 0 1386.52 694.982796.759(able Tm( )T cSEMC/P <</6CID 0 >>BDCC2_1/TT4 1 Tf0.0000 Tc 0 Tw 13.98 0 0 13.)Tj6 54.322591.839<08Tj0>m(CBT/TT4 1 Tf8.000n)Tj3.0051 Tw

Thnalalso F

s al(Ous

thennecessary

o

weflow

f

)Tj9.0007

Tc

t

-5.0051

h

Tw

e

13.98

n 0

0

f

139

l

c6822.28985596.759(ll o

w performance

c5rvesPto Oper

bencreatednflow

)Tj4.000.oTOpmized. Tm( )T cSEMC/P <</14CID 0 >>BDCBT/TT4 1PIPESI142 (Ou(Oe from thenstandardnflow )Tj0.0007 TcTm( 0.0051 Tw 13.98 0 0 1256.56 3318gu9601.839(system model data withnflow7 Tc 00.0017 Tw 13.98 0 0 14.846833318gu9601.839(thenaddiOper2 Nwithnflow )Tj1.0007 Tc 08.0051 Tw 13.98 0 0 13.)Tj275170.8804.5002artifi142 )Tj13.98 0 0 1116.43535170.8804.5002c142 s parameters, such asnwellheadPflow Fssure, )T

cSEMC/P watercut,

tubing

<</and

fl142

<</3CID

0

>>BDCC2_1/TT4

Operations

143

6.9.1 Well Performance Curves These can be created for us in the Network solver to produce faster solution times. A curve is created that represents the performance of the well under certain conditions. The network solver will then utilize this curve instead of modeling the well directly. 6.9.2 Optimization module performance curves As part of the artificial lift operation performance curves for the optimization module, GOAL, can be created. The curves are of the general form x-axis : lift quantity y-axis: liquid flowrate sensitivity variable: system outlet pressure, normally the well head, but see below on well head chokes. The lift quantity should be set so that it spans the working range of values. For gas lift this should include the case of zero injection gas, i.e. can the well flow naturally? The liquid flowrate will be computed at all the lift quantity rates for a set system outlet pressure. In order to utilize the performance curves in GOAL the system performance needs to be ascertained at different system outlet pressures. These pressures should span the normal working system outlet pressure (normally well head or manifold pressure). Typically 4/5 values are required. 6.9.2.1 Well head chokes The choking back of gas lifted wells is rare in the oil industry, but in real-life operations, some gas lifted wells have to be choked back due to instabilities of the wells. Therefore, GOAL offers several ways to modeling gas lifted wells that are choked back.

PIPESIM 2000

144

Operations

Wellhead Choke Manifold

Flowline Wellhead

Well

As GOAL uses gas lift performance curves the individual well models can be developed to model a well to either: 1. the wellhead, upstream of a well head choke or 2. the manifold that the well is connected to (including a wellhead choke and associated flowline between the well and the manifold). It is normally recommended that the well performance curves are modeled to the manifold, i.e. the choke is included in the well model. However, if any of the following situations are to be studied in GOAL then the well must be modeled to the choke. · A maximum liquid constraints into individual wells · Choke optimization · Pressure calibration Method 1: GOAL model with wells modeled to the manifold

PIPESIM 2000

Operations

145

Model 2: GOAL model with wells modeled to the well head

6.10 Gas Lift Design & Diagnostics PIPESIM is capable of performing gas lift designs for both new mandrel spacing and also for existing mandrel spacing. The user has considerable flexibility over the design method and design parameters to use. For a new spacing the mandrel depths are computed and for a new design the port size and test rack pressures. PIPESIM contains a database of gas lift valve details for most of the commonly used gas lift valves from various manufactures. The gas lift diagnostic operation can be used to analyze the performance of an existing gas lifted installation (or a proposed new design). For any selected operational conditions (e.g. tubing and casing head pressures), the status and gas throughput for each valve will be computed. This operation will also take into account the throttling behavior of the valves. Linking to the production database via the ProdMan module can further enhance the functionality of this module. 6.10.1 Check for Gas Lift instability Unstable operational conditions may occur in a continuous gas lift well because the characteristics of the system are such that small perturbations can degenerate into huge oscillations in the flow parameters. Therefore, a clearly defined mechanism is required to

PIPESIM 2000

146

Operations

show the relative importance of the different factors involved, and help to assure stable flow conditions at the design phase or to decide what actions to take in order to stabilize an unstable gas lift well. Unified instability criteria were developed by Alhanati et al. (1993) for continuous gas lift wells to overcome the drawbacks in previous developments. The unified criteria can be used for all possible flow regimes for the gas-lift valve and surface gas injection choke. The unified criteria were developed using a number of simplifying assumptions, and therefore they should not be considered as highly accurate or that they can be applied to every type of instability experienced in a gas lift installation. However, the criteria cover a number of common cases encountered in the industry and certainly indicate what can be done to improve operating instability. Assumptions of the model: · constant pressure at the gas injection manifold which is upstream of the surface injection choke. · adiabatic flow through the choke In the unified criteria, two sets of criterion were defined, namely C1 & C2, and both must be greater than zero for stable gas lift operation.

rv 2 - rv C1 = F 1. - 1 + F 3. . Fc µv µv rv rv C 2 = F 1. - 1 + µ v Fc

where

F1 =

(C

(

B f . g. q 2 go . J

D

ApY

)

2 v

.Va . q fo

F3 =

(q (

fo

+ q go . At Pto . q fo f - g .g

) )

Fc =

CD ApY

(

Pto Pco

r ( 2 - rv ) . ch ch v µ ch 2 r ( 2 - rv ) C D A p Y . ch v µ ch

)

2

+ CD ApY

(

)

2

)

rv =

µv =

( zT ) t ( zT ) c

rch =

Pco Pm

µ ch =

( zT ) c ( zT ) m

Nomenclature

PIPESIM 2000

Operations At Bf CD J Va g Pco Pto qfo qgo Pm Y T r z t g µ Cross sectional area of tubing (in2) Volume factor for reservoir fluids at injection point Gas Valves Discharge coefficient. Default = 0.8 Productivity index (stb/d/psi) Volume of tubing-casing annulus (ft3) Acceleration of gravity (ft/s2) Steady state casing pressure (psia) Steady state tubing pressure (psia) Steady state reservoir fluids flow rate (stbd) Steady state injected gas flow rate (mmscf) Gas injection manifold pressure (psia) Gas expansion factor Temperature (F) ratio of pressures gas compressibility factor reservoir fluids density (lb/in3) injected gas density (lb/in3) ratio of the products zT

147

SUBSCRIPTS v gas lift vale ch gas injection choke t tubing c casing m manifold In order to utilize this feature from the well model must be developed with the following included; · Well IPR is modeled by the PI method · Casing inside diameter is set · Port diameter. The inside diameter of the Gas Lift injection valve that is currently being used. · Surface injection pressure From this additional data the well model will automatically calculate the steady state casing and tubing pressures. The (GOAL) Gas Lift performance curves should then be developed as normal and the Alhanati factors will be automatically be generated.

PIPESIM 2000

148

Operations

The factors can be viewed graphically for any well by select the Alhanati Criterion for the y axis from the series option within the plotting utility PSPLOT. Both factors can be displayed on then same plot, if required, by adding a second series.

1.

P IP E S IM P lo t A u g 0 9 1996

W e l l P A 1 3 - AL ilc h s a n: aJ At in h o u s e ( K e n e d to B i

0001)

1.

Al ha na ti Cr

0.

0.

0.

0.

0.

1.

2.

3.

4.

5.

T o ta l In je c tio n G a s

P I P E S I M fo r W in d o w s © B a k e r J a r d in e & A s s o c ia te s

Figure 6-3 Alhanati Criterion 6.11 Horizontal well analysis The Horizontal well operation is an integral part of PIPESIM's reservoir-to-surface analysis. This option allows the user to predict hydraulic well bore performance in the completion. The multiple source concept used leads to a pressure gradient from the blind-end (Toe) to the producing-end (Heel) which, if neglected, results in overpredicting deliverability. The reduced drawdown at the Toe results in the production leveling off as a function of well length and it can be shown that drilling beyond an optimum length would yield no significant additional production. Several inflow performance relationships are available. These are solved with the wellbore pressure drop equations to yield the changing production rate along the well length. To use this operation a horizontal well completion must be included in the system model. 6.12 Reservoir tables It is often necessary, for the purposes of reservoir simulation, to generate VFP curves for input to a reservoir simulation program. The

PIPESIM 2000

Operations

149

VFP curves supply the simulator with the necessary data to define bottom hole flowing pressures and tubing head pressures as a function of various parameters such as flow rate, GOR, watercut, surface pressure and the artificial lift quantity. The reservoir simulator interface allows you to write tabular performance data to a file for input into a reservoir simulation model. Currently, the following reservoir simulators are supported: · ECLIPSE · PORES · VIP · COMP4 · MoReS (Shell's in-house reservoir simulator) The effects of variations of up to five parameters can be investigated and reported and all combinations of the variables entered by the user are used to generate the tables. Tabular data is then created in a format specific to the reservoir simulator selected. Note: Users may wish to model flow networks in their reservoir simulator, by generating VFP curves items of well tubing, flowline or riser. This will not result in an accurate model of the surface network as temperatures at network connections will not be modeled correctly. Schlumberger also has a dynamic link to reservoir simulators via the Field Planning module (FPT). 6.13 Network analysis The basic stages involved in developing a model of a field are: · Build a model of the field, including all wells and flowlines. · Specify the boundary conditions · Run the model 6.14 Production Optimization The basic stages involved in developing an optimization model of a field are: · Build a model of the field, including all wells and flowlines.

PIPESIM 2000

150

Operations

· Develop individual artificial lift performance curves for the wells in the model. Even if the wells are not on artificial lift a performance curve is required. These performance curves can be created by any approved Nodal Analysis software package. The recommended program is the well performance module of PIPESIM. · Calibrate the models developed. This involves obtaining field data so that the individual performance curves can be calibrated and checked. The Perform Prediction mode should be used for this. · Optimise the system. Once the wells and surface network have been calibrated an optimisation can be performed. See the GOAL User Guide for full details. 6.15 Field Planning The reservoir can be modeled by either; 1: the GeoQuest EclipseTM reservoir simulation program (via the Open Eclipse link) or 2: a single, or series of, look-up tables or 3: compositional tank models. The network models are constructed using the network module and solved using its calculation engine. 6.15.1 Dynamic Eclipse link The network module models the surface Network from the bottom hole conditions to the supply/distribution point while Geoquest's Eclipse reservoir simulator is used to model the reservoir. FPT passes flowrate targets to Eclipse and the network in order to try to converge on bottom-hole conditions.

PIPESIM 2000

Operations

151

PROS: · An industry standard simulator simulates the reservoir. · Phase flowrates are dependent on current flowrates from all wells and reservoir history. · Full account can be taken of the reservoir geometry and aquifer behavior etc. CONS: · Simulation time is significantly longer. · Need to set-up the communication link from the Eclipse simulator based on a UNIX workstation to FPT based on a PC. · Need to purchase OpenEclipse from Geoquest and install it properly. · It is much harder to converge on a solution between the network and Eclipse. Capabilities: · Can model deliverability systems that have pressure specified sinks. · Can model blackoil Eclipse reservoir models in both Engineering and SI units. · Can flowrate constrain all source wells. Limitations: · Cannot model surface networks which have flowrate specified sinks. · Cannot model compositional Eclipse models.

PIPESIM 2000

152

Operations

Construction of the overall Eclipse linked model involves first providing the name of the Eclipse model and on which server/workstation it is located on the Network. This model contains the time stepping information that will be used to control the surface network and also decides when wells will be turned on or off. This field planning data can be overridden by events defined in the field events editor. It also contains the flowrate and pressure limits that are to be imposed upon the wells. These can be ignored in deliverability mode where the maximum capability of the surface network is used to calculate the flow from each well, or obeyed in the usual running mode. A number of network models can be linked to the Eclipse model, so injection and production networks can be modeled separately. The surface injection network can be ignored which significantly reduces simulation time. 6.15.2 Look-up tables Reservoir properties are taken from a table defined in an ASCII text file, which provide pressure (and optionally pressure and watercut) as a function of cumulative production of oil, liquid, or gas.

Sample decline curve

5000 4000 3000 2000 1000 0 70 60 50 40 30 20 10 0

0 5 10 15 PROS: Cumulative liquid production [mmstb] · Very fast reservoir modeling Pressure [psia] GOR [scf/stb] Watercut [%] as no iteration is required unless conditional logic in the field planning demands that a timestep be run again. · Tables can be generated in other packages such as Excel, by Eclipse, by MBAL etc. and then read into FPT. · This is the easiest form of reservoir modeling to set-up and use. · Everything is included in the FPT package, no third party software is required.

CONS: · Phasic flowrate behavior is NOT dependent on total flowrate.

PIPESIM 2000

Operations

153

User supplied composition, initial volumetric

Here the reservoirs are modelled by defining the geometry of a simple cylinder containing a user-supplied volume of fluid (either in terms of liquid or gas). Given a user supplied composition, this tank is then depleted via wells mapped to

154

Operations

decline and possible composition changes. Simple aquifer models and fluid injection options are also available. PROS: · Relatively straightforward to set-up with no third party software. · Full compositional modeling is performed upon the fluid in the reservoir to obtain the correct pressure. CONS: · The watercut in the tank model cannot be changed without injecting a fluid stream containing water. Capabilities: · Simple aquifer (influx rate or volume replacement) and fluid injection options are available. · Product streams can be gas, liquid, or the tank mixture. Limitations: · Aquifer influx does not cause a gradual watering out of the well but a sharp cut off when the aquifer is deemed to have raised the water level in the reservoir to the well perforation point. · Simple tank geometry is assumed. A tank is merely a cylinder that does not account for any pore volume reduction as fluid is taken from the reservoir. 6.15.4 Event handling · FPT allows events to be specified either at certain timesteps, or conditionally upon targets being reached, or exceeded etc, e.g. if the watercut in branch XXX goes above 95%, shut well Y off. · Flowrate constraints can be imposed on individual wells in the network models. These wells will be automatically choked back (if necessary) to meet production requirements. · Gas lift rates, well PI values, and compressor horsepower settings can be set and/or changed from the Events Editor. · The look-up table editor now enables the user to specify a case study mode for FPT enabling different scenarios to be run in batch mode and the results analyzed in the postprocessor. · Group flowrate constraints imposed in an Eclipse input file can be honored by the FPT.

PIPESIM 2000

Operations

155

See FPT User Guide for full details. 6.16 Multi-lateral well analysis See the HoSim User Guide for full details. 6.17 Post processor The post processing is conducted via one of the following methods; · Graphical plots · PIPESIM graphical utility · Microsoft Excel · Tabular data · Standard text editor · Microsoft Excel · Onscreen data · PIPESIM GUI 6.17.1 Graphical plots Graphical plots are the most common method used to view data (input and results) from PIPESIM. Input data may be viewed graphically to show; · Tubing profile · Flowline profile · Inflow performance relationship Calculated data may be viewed graphically to show; · Phase envelop · Calculated Inflow performance relationship · PVT data · Simulation results · System data - data that changes as a result of some input, i.e. system outlet pressure as a function of well PI, etc. · Profile data - data that changes along the system profile, i.e. pressure, temperature, etc.

PIPESIM 2000

156

Operations

6.17.2 Tabular data Tabular data is in the form of text (ASCII) output files. These can be viewed from with PIPESIM or via a standard text editor. They can also be printed. 6.17.3 Onscreen data The input and output data from any object can be obtained via the screen schematic. In addition results from the network module can be obtained via the output report tool. 6.18 References Alhanati et al. (1993) B Wilkens, M Apte, G Broze (1999) User's Guide for the wax Deposition Option in PIPESIM. Project R13-0511.000.

PIPESIM 2000

Operations THIS PAGE LEFT BLANK INTENTIONALLY

157

PIPESIM 2000

Case Studies 7 Case Studies

159

The PIPESIM software comes preloaded with a number of case studies that demonstrates some of its capabilities, some of which are fully documented here. The full list of case studies is; Pipeline & facilities Case Study ­ Condensate Pipeline · Compositional · Phase envelope creation · Hydrate envelope · Pipeline sizing · Pipeline insulation · Slugging · Slug catcher sizing Well Performance Case Study ­ Oil Well Design · Black Oil fluid calibration · Well IPR · Tubing sizing Network Analysis Case Study ­ Looped Gas Gathering Network · Compositional · Network model · Boundary conditions · Establish field deliverability Optimization Field Planning Multi-lateral

PIPESIM

Case Studies

161

7.1 Pipeline & facilities Case Study ­ Condensate Pipeline A subsea pipeline is to be designed to transport condensate from a satellite platform to a processing platform. Compositional analysis of the condensate has been obtained. The engineer is asked to perform the following tasks:- Develop a compositional model of the hydrocarbon phases. - Add the aqueous phase to the compositional model and identify the hydrate envelope. Hydrates are to be avoided by operating the pipeline above the hydrate formation temperature. - Select a pipeline size. - Determine the pipeline insulation requirement. - Screen the pipeline for severe riser slugging. Severe riser slugging is to be avoided. - Size a slug catcher. The engineering data available is given at the end of this case study. 7.1.1 Task 1. Develop a Compositional Model of the Hydrocarbon Phases A compositional fluid model allows the fluid physical properties to be estimated over the range of pressures and temperatures encountered by the fluid. The fluid model is made up of individual pure library components such as methane, and petroleum fractions. Petroleum fractions are used to estimate the behavior of groups of heavier pure components. The hydrocarbon phase envelope can be plotted on pressure and temperature axes. The following steps are to be carried out:- Add the pure hydrocarbon components. - Characterize and add a petroleum fraction. - Generate the hydrocarbon phase envelope. After starting PIPESIM use the <File/New/pipeline and facilities model> menu to open a new model and save this in the training directory (e.g. as file c:\training\ps02.bps). Use the <setup/compositional...> menu to enter the pure components given at the end of the case study. Select the pure hydrocarbon components from the component database. Multiple selection is possible by holding down the control key. When all pure hydrocarbon components have been selected, press the "Add>>" button. When the number of moles of the pure components have been added, select the "Petroleum Fractions" tab and characterize the petroleum fraction "C7+" by entering the BP, MW, and SG in row 1. Then press the "Add to composition>>"

PIPESIM

162

Case Studies

button and enter the number of moles for C7+ under the "Component Selection" tab. Generate the hydrocarbon phase envelope by pressing the "Phase Envelope" button. The following plot should be obtained:

7.1.2 Task 2. Identify the Hydrate Envelope Certain fluid compositions show a tendency to form hydrate compounds in the presence of water. These compounds can cause line blockages. The tendency to form hydrates is dependent also on pressure and temperature. In this study, hydrate formation is to be avoided by operating above the hydrate formation temperature at all times. The following steps are to be carried out: - Add the aqueous component. - Generate the hydrate envelope. First it is necessary to add the aqueous component, pure water. Use the <setup/compositional...> menu to select "water" and press the "Add>>" button. Enter the water concentration of 10% volume ratio (bbl/bbl). Generate the aqueous phase envelope and the hydrate formation line by pressing the "Phase Envelope" button. The following plot should be obtained:

PIPESIM

Case Studies

163

Note that hydrates tend to form in the region on or to the left of the hydrate line. In this study, hydrate formation will be avoided by operating the pipeline at temperatures above 75 °F at all times. 7.1.3 Task 3. Select a Pipeline Size Find the smallest pipeline I.D. that will allow the design flowrate of 10,000 STB/d of condensate to be transported from the satellite platform whilst maintaining an arrival pressure of not lower than 1,000 psia at the processing platform. The pipeline sizes available are 8", 10", or 12" I.D. as described in the data section at the end of the case study. This can be determined as follows: - Use the pressure temperature profiles operation to calculate the pressure drop for each of the three pipeline size options. First it is necessary to add a source to the model. This is done by pointing and clicking on the source button at the top of the screen and then pointing and clicking in the work area. A source appears as shown below. Alternatively the wizard feature can be used.

source button

source

To enter data relevant to the source double click on the object. Enter the inlet pressure of 1,500 psia and the inlet temperature of 176 °F.

PIPESIM

164

Case Studies

Now add a boundary node to represent the arrival point at the processing platform. boundary node button boundary node

Then add nodes to represent each end of the pipeline: node button node

Connect the model together by pointing, clicking and dragging using the riser and flowline buttons: riser button flowline button

Completed Model Note that the red outline indicates that essential data is missing for that component. Double click on "Riser_1" to enter the riser details i.e. horizontal distance and elevation difference (length is automatically computed), I.D., roughness, overall heat transfer coefficient and ambient temperature. Repeat this for "Flowl_1" and "Riser_2". Select the <operations/pressure-temperature profiles...> menu and set up the operation so that the calculated variable is outlet pressure. Set the Inlet pressure 1,500 psia and the Liquid Rate to 10,000 STB/d. The sensitivity variable is Pipeline ID with values of 8", 10", and 12", this select the component as "Flowline_1" , the variable as "ID" and enter the sizes. Press the Run Model button when all the data has been added. The following plot should be obtained (the axis may have to be changed to show Total Distance v's Pressure):

PIPESIM

Case Studies

165

It can be seen that a 10" is the smallest pipeline size that will satisfy the arrival pressure condition of at least 1,000 psia. Note: Don't forget to now set the flowline ID to 10" for all subsequent simulations. 7.1.4 Task 4. Determine the Pipeline Insulation Requirement Find the smallest thickness of thermal insulation that can be used to insulate the pipeline and maintain an arrival temperature of not less than 75 °F. This minimum arrival temperature is required to prevent the formation of hydrates. The insulation has a thermal conductivity of 0.15 Btu/hr/ft/°F and a thickness of 0.75" or 1". This can be determined as follows: - Use the pressure temperature profiles operation to calculate the temperature profile for the design and turndown flowrate cases with 0.75" thermal insulation thickness. - Re-run the model with 1.0" thermal insulation thickness and compare the temperature profiles. Double click on "Flowl_1". Select the "Heat Transfer" tab, and then select the "Calculate U" sub-tab. Enter the heat transfer data given at the end of the case study, and add a layer of insulation with a thermal conductivity of 0.15 Btu/hr/ft/°F and a thickness of 0.75". Press the "OK" button. Select the <operations/pressure-temperature profiles> menu and set up the operation so that the calculated variable is outlet pressure, and the

PIPESIM

166

Case Studies

sensitivity variable is System data/liquid rate with values of 5,000 and 10,000 STB/d. Run the model and configure the output to obtain the following plot:

Re-run the model using a thermal insulation thickness of 1". Configure the output to obtain the following plot:

It can be seen that 1" insulation is required to maintain an arrival temperature of 75 °F.

PIPESIM

Case Studies

167

Note: Don't forget to now set the insulation thickness to 1" for all subsequent simulations. 7.1.5 Task 5. Screen the Pipeline for Severe Riser Slugging Severe riser slugging is likely in a pipeline system followed by a riser under certain conditions. The elements leading to severe riser slugging are: 1. The presence a long slightly downward inclined pipeline prior to the riser. 2. Fluid flowing in the "stratified" or "segregated" flow regime (as opposed to the usual "slug" or "intermittent" flow regime). 3. A slug number (PI-SS) of lower than 1.0. The PI-SS number can also be used to estimate the severe riser slug length from the equation: slug length = riser height/PI-SS number. Severe riser slugging is to be avoided in this case. The necessary information can be extracted from the model as follows:- Configure the model output such that slug information, and flow regime maps are printed for the fluid at the riser base. Select the <setup/define output...> menu and check the "slug output pages" box. Set "number of cases to print" to 2. Add a report tool to the model in place of node "N2". This can be done by first selecting a report tool and placing it in the work area. report tool button report tool

Then reconnect "Flowl_1" to the report tool by first clicking on the middle of "Flowl_1". You will see that highlight boxes appear at either end of the flowline. Move the mouse over the right hand highlight box, and the mouse pointer changes to an "up arrow" shape (). The line can then be dragged from "N2" and dropped onto the report tool as shown below.

PIPESIM

168

Case Studies

Reconnecting the flowline to the report tool Similarly reconnect "Riser_2" to the report tool. Delete "N2", and reposition the report tool as shown below.

l Modified model Double click on the report tool and check the option "flow map". Select the <operations/pressure-temperature profiles> menu and re-run the model. Select the <reports/view output> menu and check the PI-SS number at the riser base for both flowrate cases. It can be seen that the PI-SS number is higher than 1.0 at the riser base in both cases. In the turndown flowrate case the PI-SS number is 1.18 as shown below:

PIPESIM

Case Studies

169

Check the riser base flow regime maps in the output file to see if the flow is in the "stratified" or segregated region. It can be seen that flow is in the intermittent (normal slugging) flow regime. The turndown case flow map is shown below:

PIPESIM

170

Case Studies

It can be seen that the segregated region has been avoided and the likelihood of severe riser slugging is reduced. Note: Don't forget to save the final model! 7.1.6 Task 6. Size a Slug Catcher Having established that normal slug flow is expected, it is now necessary to size a slug catcher. The size will be determined by the largest of three design criteria: 1. The requirement to handle the largest slugs envisaged (chosen to be statistically the 1/1000 population slug size). 2. The requirement to handle liquid swept in front of a pig. 3. Transient effects, i.e. the requirement to handle the liquid slug generated when the production flow is ramped up from 5,000 to 10,000 STB/d. This can be achieved as follows: - Review the simulation output to establish the slug catcher volume required for each of the three design criteria and select the largest volume. Review the output file and it can be seen that the turndown case generates larger slugs.

PIPESIM

Case Studies

171

. As shown above, the 1/1000 slug length is 1,781.2 ft, which gives a slug volume of 971.5 ft3. Now select the <reports/view summary> menu and check the liquid swept in front of a pig ("liquid by sphere").

PIPESIM

172

Case Studies

It can be seen that the turndown case gives the larger volume of 279.1 bbl or 1,567 ft3. Now calculate the liquid generated when the flow is ramped up from 5,000 STB/d to 10,000 STB/d. This is the difference in total holdup between the two cases, i.e. 692 - 623 = 69 bbl or 522 ft3. Therefore the pigging volume of 1,567 ft3 is the determining design case. 7.1.7 Data Available Layout: Condensate flows down a 400 ft x 10" ID riser from the satellite platform to the seabed, along a 5 mile pipeline, and up a 400 ft x 10" ID riser to the processing platform. Boundary Conditions: Fluid inlet pressure at satellite platform Fluid inlet temperature at satellite platform Design liquid flowrate Maximum turndown Minimum arrival pressure at processing platform Minimum arrival temperature at processing platform Pure Hydrocarbon Components: Component Methane Ethane Propane Isobutane Butane Isopentane Pentane Hexane Petroleum Fraction: Name Boiling Point (°F) C7+ 214 Molecular Weight 115 1,500 psia 176 °F 10,000 STB/d 5,000 STB/d 1,000 psia 75 °F.

Moles 75 6 3 1 1 1 0.5 0.5 Specific Gravity 0.683 Moles 12

PIPESIM

Case Studies Aqueous Component: Component Water

173

Volume ratio (%bbl/bbl) 10 Roughness (") 0.001 0.001 0.001 10/1000 5 miles 0 0.5" 0.001" 50 °F 0.2 Btu/hr/ft2/°F 50 Btu/hr/ft/°F 0.15 Btu/hr/ft/°F 0.75" or 1.0" water 1.64 ft/sec 0 (half buried) 1.5 Btu/hr/ft/°F 0 -400 ft +400 ft 10" 0.5" 0.001" 50 °F 0.2 Btu/hr/ft2/°F

Pipeline Sizes Available: I.D.(") Wall thickness (") 8 0.5 10 0.5 12 0.5 Pipeline Data: Height of undulations Horizontal distance Elevation difference Wall thickness Roughness Ambient temperature Overall heat transfer coefficient Pipeline Insulation Study Data: Pipe thermal conductivity Insulation thermal conductivity Insulation thickness available Ambient fluid Ambient fluid velocity Burial depth Ground conductivity Data for Risers 1 and 2:Horizontal distance Elevation difference (Riser_1) Elevation difference (Riser_2) Inner diameter Wall thickness Roughness Ambient temperature Overall heat transfer coefficient

PIPESIM

174

Case Studies

THIS PAGE LEFT BLANK INTENTIONALLY

PIPESIM

Case Studies

175

7.2 Well Performance Case Study ­ Oil Well Design An oil reservoir has been discovered in the North Sea. A vertical well has been drilled, a test string inserted and flow characteristics measured. Fluid properties at stock tank and laboratory conditions have been obtained. Reservoir simulations have been performed to predict the change in watercut over the field life. The reservoir pressure will be maintained by water injection and the preference is to avoid the use of artificial lift methods. The engineer is asked to perform the following tasks: - Develop a blackoil model to match the laboratory data. It is necessary to develop a method of predicting the fluid physical properties so that the pressure losses and heat transfer characteristics can be calculated. - Develop a well inflow performance model applicable throughout field life. This provides a relationship between the reservoir pressure, the flowing bottom hole pressure and flowrate through the formation. - Select a suitable tubing size for the production string. The engineering data available is given at the end of this case study. 7.2.1 Task 1. Develop a Calibrated Blackoil Model No analysis work can be carried out until a blackoil fluid model has been developed. This allows all of the fluid physical properties to be estimated over the range of pressures and temperatures encountered by the fluid. These physical properties are subsequently used to determine the phases present, the flow regime, the pressure losses in single and multiphase flow regions, and the heat transferred to or from the surroundings. The following steps are to be carried out:- Obtain a partially calibrated blackoil model using the stock tank and bubble point properties. - Plot the partially calibrated oil formation volume factor (OFVF) over a range of pressures and temperatures to identify any differences between the measured and the predicted properties. Any discrepancies will lead to fluid flow modeling errors. - Apply calibration to the OFVF above the bubble point pressure and observe how the property curves are corrected. - Apply calibration to the OFVF below the bubble point pressure and observe how the property curves are corrected. - Apply calibration to the oil viscosity using first the measured dead oil data and then further tuning with live oil data. - Apply calibration to the gas viscosity and the gas compressibility.

PIPESIM

176

Case Studies

After starting PIPESIM use the <File/new/well> menu to open a new well performance model and save this in your training directory (e.g. c:\training\...). Use the <setup/blackoil...> menu to enter the stock tank oil properties and the bubble point properties given at the end of the case study. Help on the definitions and valid ranges of these stock tank properties can be obtained by selecting the button from the dialog header bar and clicking on the relevant data entry field. Press the "OK" button and save the model. Use the <setup/blackoil/advanced calibration data> menu and press the "plot PVT data..." button (note: do not enter the advanced calibration data at this stage). Use the <series> menu to plot the oil formation volume factor on the y axis. The following plot should be obtained:

The partially calibrated curve for a temperature of 210 °F shows that the predicted OFVF is higher than the measured value both above and below the bubble point pressure. At 4,269 psia the predicted value is 1.52 compared to the measured value of 1.49 and at 2,000 psia the predicted value is 1.41 compared to the measured value of 1.38. Therefore further calibration is required. Apply OFVF calibration above the bubble point pressure. The measured value is 1.49 @ 4,269 psia and 210 °F. The following plot should be obtained:

PIPESIM

Case Studies

177

Apply OFVF calibration below the bubble point pressure. The measured value is 1.38 @ 2,000 psia and 210 °F. The following plot should be obtained:

Calibration of the oil viscosity requires two dead oil data points. The uncalibrated default approach is to use the Beggs and Robinson correlation which gives values of 1.561 cP @ 200 °F and 23.27 cP @ 70 °F. The Beggs and Robinson correlation uses the oil API gravity to

PIPESIM

178

Case Studies

predict two dead oil data points based upon data obtained from around 2,000 data points from 600 oil systems. Plot the un-calibrated oil viscosity. The following plot should be obtained:

In this case it can be seen that the predicted oil viscosity value at a temperature of 70 °F and 14.7 psia is 23.27 cP as specified by the Beggs & Robinson correlation. This is significantly different from the measured dead oil data and would lead to errors in the prediction of pressure loss. Open the <setup/blackoil/viscosity data> menu and select the correlation option "user data". Enter the two measured values of 0.31 cP @ 200 °F and 0.8 cP @ 70 °F. The following plot should be obtained:

PIPESIM

Case Studies

179

It can be seen that the predicted oil viscosity value at a temperature of 70 °F and 14.7 psia is 0.8 cP consistent with the laboratory dead oil data. Open the <setup/blackoil/advanced calibration data> menu and enter the live oil calibration data of 0.29 cP @ 2,000 psia and 210 °F. The following plot should be obtained:

It can be seen that the predicted oil viscosity value at a temperature of 210 °F and 2000 psia is 0.29 cP consistent with the laboratory live oil data.

PIPESIM

180

Case Studies

Proceed to calibrate the gas viscosity and the gas compressibility using the calibration data given earlier. 7.2.2 Task 2. Develop a Well Inflow Performance Model A straight line productivity index (PI) method is considered adequate in this case because the fluid flows into the completion at a pressure considerably above the bubble point and no gas comes out of solution at this stage. This applies throughout field life and the productivity index is not expected to change. The PI will not be affected by changes to the reservoir pressure because the reservoir pressure is to be maintained by water injection. The PI will not be affected by changes to the watercut through field life because the oil and water have similar mobilities in this reservoir structure. The following step is to be carried out: - Use the drill string test data to obtain a representative productivity index. First it is necessary to add a vertical completion to the model. This is done by pointing and clicking on the vertical completion button at the top of the screen and then pointing and clicking in the work area. A vertical completion appears as shown below.

vertical completion button

vertical completion

Double click on the vertical completion in the work area to enter data relevant to that item. Enter the static reservoir pressure of 4,269 psia and the reservoir temperature of 210 °F. Press the "calculate/graph..." button and enter the drill string test data given below. Press the "plot IPR" button and this will calculate a productivity index to be used throughout the analysis work. 7.2.3 Task 3. Select a Tubing Size for the Production String Find the smallest tubing size that will allow this production plan to be met on the basis that the production string will not be replaced during field life. The tubing sizes available are 3½", 4½" or 5½" for which the I.D.'s are 2.992", 3.958" and 4.892". This can be determined as follows: - Use the systems analysis operation to generate a plot of oil flowrate against watercut for each of the three tubing sizes.

PIPESIM

Case Studies

181

- Overlay the production plan data and identify the smallest size that allows this plan to be met. First it is necessary to extend the model to include a tubing string. Add a boundary node to the model by pointing and clicking on the boundary node button at the top of the screen and then pointing and clicking in the work area: boundary node button boundary node

Then use the tubing button to connect the well to the boundary node: tubing button

Completed Model Note that the red outline indicates that essential data is missing for that component. Double click on the tubing to enter the well depth and the tubing thickness, roughness, overall heat transfer coefficient and ambient thermal gradient. Select the <operations/systems analysis> menu and set up the operation so that the calculated variable is liquid rate. The x axis variable is watercut with values of 0, 12, 20, 35, 40, 47, 54 and 60%, representing the various stages of field life. The sensitivity variable is tubing I.D. with values of 2.992", 3.958" and 4.892". Configure the output to give the water cut against the stock-tank oil rate at the outlet (this is achieved via the series option of PSPLOT):

PIPESIM

182

Case Studies

It can be seen that 4½" tubing is the smallest size that will satisfy all of the production plan conditions. Note: Don't forget to now set the tubing ID to 3.958 to reflect the 4½" tubing for all subsequent simulations. 7.2.4 Data Available Reservoir Conditions: Reservoir pressure 4,269 psia, Reservoir temperature 210 °F Stock Tank Oil Properties: Watercut 0%, GOR 892 scf/STB, Gas SG 0.83, Water SG 1.02, API 36.83 Bubble Point Properties: Pressure 2,647 psia, Temperature 210 °F, Solution Gas 892 scf/STB Blackoil Calibration Data: OFVF (above bubble point pressure) OFVF (below bubble point pressure) Dead oil viscosities Live oil viscosity Gas viscosity Gas compressibility (Z) 1.49 @ 4,269 psia and 210 °F 1.38 @ 2,000 psia and 210 °F 0.31 cP @ 200 °F and 0.8 cP @ 70 °F 0.29 cP @ 2,000 psia and 210 °F 0.019 cP @ 2,000 psia and 210 °F [email protected] 2,000 psia and 210 °F

PIPESIM

Case Studies

183

Deviation Survey: The well is vertical from the well head on the sea bed. Mid perforations are at a depth of 9,500 ft from the well head. The ambient temperature varies linearly between 210 °F at mid perforations and 60 °F at the wellhead. The minimum casing inner diameter is 10". The generally accepted overall heat transfer coefficient of 2 BTU/hr/ft2/°F for wellbores can be used throughout. Minimum Pressure Allowed at the Wellhead: 300 psia Multiphase flow correlation Beggs & Brill revised Production Strings Available: I.D. (") Wall thickness (") 2.992 0.5 3.958 0.5 4.892 0.5 Drill String Test: Oil Flowrate (Q), sbbl/d 2,000 3,000 4,000 5,000 Roughness (") 0.001 0.001 0.001

Flowing Bottom Hole Pressure (Pwf), psia 4,186 4,152 4,106 4,072

Production plan obtained from reservoir simulation: Year Watercut (%) Oil Flowrate, sbbl/d 0 0 12,000 4 12 10,500 5 20 9,400 6 35 7,500 7 40 7,000 8 47 6,000 9 54 5,000 10 60 4,300

PIPESIM

7-184

Case Studies

7.3 Network Analysis Case Study ­ Looped Gas Gathering Network The deliverability of a production network is to be established. The network connects three producing gas wells in a looped gathering system and delivers commingled product to a single delivery point. The engineer is asked to perform the following tasks:- Build a model of the network. - Specify the network boundary conditions. - Solve the network and establish the deliverability. The engineering data available is given at the end of this case study. 7.3.1 Task 1. Build a Model of the Network The following steps are to be carried out:- Enter the engineering data for the first well. - Copy the data to wells 2 and 3. - Modify the data for well 3. - Specify the composition at each production well. - Connect the network together. - Define the engineering data for each branch. After starting PIPESIM use the <file/new/network> menu to open a new network model and save this in your training directory (e.g. as file c:\training\pn01.bpn). Use the production well button to place Well 1 in the work area as shown below.

production well button

production well

Double click on Well 1 to reveal the components as shown below:

PIPESIM

Index

7-185

Double click on the vertical completion to enter the inflow performance data. Enter a gas PI of 0.0004 mmscf/d/psi2. The reservoir temperature and pressure are defined below. Double click on the tubing, and define a vertical tubing with a wellhead TVD of 0 and mid perforations TVD and MD of 4500 ft. The ambient temperatures are 130 °F at mid perforations and 60 °F at the wellhead. The tubing has an I.D. of 2.4". Note that the essential data fields are shown in red outline (if the fields are not outlined, then data entry in these fields is optional). Close the view of Well 1 to return to the network view. Select "Well 1" and using the commands <edit/copy> <edit/paste> copy "Well 1" to "Well 2" and "Well 3". Position the new wells as shown below:

You will see that Wells 2 and 3 have adopted the data of Well 1. Double click on Well 3 and modify the completion and tubing data. Double click on the vertical completion to enter the inflow performance data. Enter a gas PI of 0.0005 mmscf/d/psi2. Double click on the tubing, and define a vertical tubing with a wellhead TVD

PIPESIM

7-186

Case Studies

of 0 and mid perforations TVD and MD of 4900 ft.. The ambient temperatures are 140 °F at mid perforations and 60 °F at the wellhead. The tubing has an I.D. of 2.4". Close the view of Well 3 to return to the network view. The next step is to define the compositions at the production wells. Wells 1 & 2 are producing from the same reservoir and have the same composition. Well 3 has a different composition as shown in the data section at the end of the case study. The most efficient way define the compositions is to set the more prevalent composition (i.e. that for Wells 1 and 2) as the global composition and then to specify the composition of Well 3 as a local variant. The composition of Wells 1 and 2 is the same as that for the pipeline and facilities case study 2 and can be imported. First save the current network model. Open the pipeline and facilities case study (e.g. c:\training\ps02.bps). Use the <setup/compositional...> menu and the export button to export the composition to a file called "comp1.pvt". Now close the pipeline and facilities case study. In the network model, use the <setup/compositional...> menu and the import button to import comp1.pvt as the global composition. Click the right mouse button over Well 3, select fluid model and modify the composition to be locally defined as given at the end of this case study. Now position the sink and some junction nodes. Note that holding down the "Shift" key whilst placing junction nodes allows multiple placement, you should release the "Shift" key before the final placement. The network should now look like this:

PIPESIM

Index

7-187

Using the branch button connect J1 to J2. To do this, click on the branch button, then hold down the left mouse button over J1 and drag the mouse pointer to J2 before releasing the left mouse button.

branch button

branch connected

Double click on the arrow in the center of "B1" to enter data for that branch. Now double click on the flowline to enter data. Close the "B1" window to return to the network view. As the looped gathering lines are all identical, the data for branch "B1" can be propagated to the other looped gathering lines. Select "B1" by clicking on the arrow in the middle of the branch and using the commands <edit/copy> and then <edit/paste> copy "B1" to "B2", "B3", and "B4". Position the new branches as shown below:

PIPESIM

7-188

Case Studies

In order to reconnect a pasted branch, first pick the arrow in the middle of the new branch. You will see that highlight boxes appear at either end of the branch. Move the mouse pointer over the right hand highlight box, and you will see that the mouse pointer changes to an "up arrow" shape (). This end of the branch can then be dragged and dropped onto a junction node. Now connect the wells to the adjacent junction node and connect "J4" to the sink as shown below:

Now enter the components and data for branch "B5". Branch "B5" comprises a liquid separator with an efficiency of 100%, a compressor with a pressure differential of +400 psi and an efficiency of 70%, an after-cooler with an outlet temperature of 120 °F and a delta P of 15 psi, and flowline sections. The equipment is located at "B5" as shown below:

PIPESIM

Index Note: You should use the connector together. to join the equipment

7-189

7.3.2 Task 2. Specify the Network Boundary Conditions First it is necessary to summarize the rules for specification of network boundary conditions. The network solver solves the fluid pressures, temperatures, and flowrates around a network for a userspecified set of boundary conditions. The following definitions are used: Lone Node: A lone node is a node with only one branch connected, i.e. a production well, an injection well, a source or a sink. Boundary conditions: The fluid pressure, temperature, and flowrate at each lone node in the network. The following rules apply: Rule for Temperatures: The fluid temperature at all sources and the static reservoir temperature at all production wells must be specified by the user. The fluid temperature at all sinks and injection wells are always calculated. Rules for Pressures and Flowrates: There are two rules for specification of pressure and flowrate boundary conditions: Rule 1 - Degrees of Freedom. The total number of flowrates, pressures and PQ curves specified must equal the total number of lone nodes. Rule 2 - At Least one Pressure. A least one pressure must be specified at one of the lone nodes. All unspecified pressures and flowrates are calculated by PIPESIMNet. In this case study, the above rules are satisfied by the following; - Specify all the fluid inlet temperatures

PIPESIM

7-190

Case Studies

- Specify all the fluid inlet pressures and the delivery pressure. Use the <Setup/boundary conditions> menu to specify the boundary conditions below: Node Well_1 Well_2 Well_3 Sink_1 Pressure 2,900 psia 2,900 psia 3,100 psia 800 psia Temperature 130 °F 130 °F 140 °F (calculated)

Note that all of the flowrates will be calculated. 7.3.3 Task 3. Solve the Network and Establish the deliverability First it is necessary to explain the network tolerance. A network has converged when the pressure balance and mass balance at each node is within the specified tolerance. The calculated pressure at each branch entering and leaving a node is averaged. The tolerance of each pressure is calculated from the equation: Ptol = I(P - Pave.)/Pave. x 100%I If all Ptol values are within the specified network tolerance then that node has passed the pressure convergence test. This is repeated for each node. The total mass flowrate into and the total mass flowrate out of a node are averaged. The tolerance is calculated from the equation: Ftol = I(Tot. mass flowrate in - Tot. mass flowrate ave.)/Tot. mass flowrate ave. x 100%I If the Ftol value is within the specified network tolerance then that node has passed the mass convergence test. This is repeated for each node. When all of the above conditions are satisfied, the network has converged.

PIPESIM

Index In this case study, the following steps are required: - Set the network tolerance. - Run the model. - View the tabular reports. - View the graphical reports.

7-191

Use the <setup/options/network iterations> menu to set the network tolerance to 1%. Save the model and then press the run button .

When the network has solved you should get the message "pn01 Finished OK". Press the "OK" button. Press the report tool button simulation. and you will see the results from the

More comprehensive tabular reporting is available using the summary file button . Select the branch from well "W3", branch "B3" and branch "B5". Hold the "Shift" key down in order to effect a multiple selection. Then press the system plot button . The following pressure profile for these three branches should be obtained. The effect of the compressor at "J4" on the system pressure can be seen:

PIPESIM

7-192

Case Studies

7.3.4 Data Available Layout: The network is laid out as shown below:

Completion and Tubing Data: Gas PI Wellhead TVD Mid Perforations TVD Mid Perforations MD Tubing I.D. Wellhead Ambient Temperature Mid Perforations Ambient Temperature Heat Transfer coefficient Wells 1 & 2 0.0004 mmscf/d/psi2 0 4500 ft 4500 ft 2.4" 60 °F 130 °F 0.2 Btu/hr/ft2/F Well 3 0.0005 mmscf/d/psi2 0 4900 ft 4900 ft 2.4" 60 °F 140 °F 0.2 Btu/hr/ft2/F

Pure Hydrocarbon Components (Wells 1 & 2): Component Moles Methane 75 Ethane 6 Propane 3

PIPESIM

Index Isobutane Butane Isopentane Pentane Hexane Petroleum Fraction (Wells 1 & 2): Name Boiling Molecular Point (°F) Weight C7+ 214 115 1 1 1 0.5 0.5 Specific Gravity 0.683 Moles 12

7-193

Aqueous Component (Wells 1 & 2): Component Volume ratio (%bbl/bbl) Water 10 Pure Hydrocarbon Components (Well 3): Component Methane Ethane Propane Isobutane Butane Isopentane Pentane Hexane Petroleum Fraction (Wells 3): Name Boiling Molecular Point (°F) Weight C7+ 214 115 Aqueous Component (Well 3): Component Water Moles 73 7 4 1.5 1.5 1.5 0.5 0.5 Moles 10.5

Specific Gravity 0.683

Volume ratio (%bbl/bbl) 5

Data for Looped Gathering Lines (B1, B2, B3, and B4): Rate of undulations 10/1000 Horizontal distance 30,000 ft Elevation difference 0 ft

PIPESIM

8-194

Case Studies 6" 0.5" 0.001" 60 °F 0.2 Btu/hr/ft2/°F Liquid 100% 400 psi 70% 120 °F 15 psi 10/1000 10,000 ft 0 ft 8" 0.5" 0.001" 60 °F 0.2 Btu/hr/ft2/°F

Inner diameter Wall thickness Roughness Ambient temperature Overall heat transfer coefficient Data for Deliver Line (B5): Separator type Separator efficiency Compressor differential pressure Compressor efficiency Aftercooler outlet temperature Aftercooler delta P Flowline Rate of undulations Flowline Horizontal distance Flowline Elevation difference Flowline Inner diameter Flowline Wall thickness Flowline Roughness Flowline Ambient temperature Flowline Overall heat transfer coefficient Boundary Conditions: Node Well_1 Well_2 Well_3 Sink_1 Pressure 2,900 psia 2,900 psia 3,100 psia 800 psia

Temperature 130 °F 130 °F 140 °F (calculated)

7.4 Optimization See the GOAL User Guide for optimization case studies. 7.5 Field Planning See the FPT User Guide for Field Planning case studies. 7.6 Multi-lateral See the HoSim User Guide for Multi-lateral case studies.

8 Index

PIPESIM

Index Alhanati instability criteria ... 146 Artificial Lift ESP Lift ........................... 105 Gas Lift ........................... 104 Performance ................... 142 Back pressure IPR................ 89 Bit lock...... See Security Device Black Oil correlations ....................... 52 fluid type ........................... 32 Building a model ................... 31 C and n IPR .......................... 89 Chokes ............................... 106 Compositional EOS .................................. 60 fluid type ........................... 34 Compressor ........................ 119 Coning .................................. 55 Darcy IPR ............................. 90 Dongle ...... See Security Device Expander ............................ 120 Fetkovich,liquid IPR............. 87 Flow correlation Multiphase - horizontal...... 76 Multiphase - vertical .......... 70 Single Phase..................... 69 Flow regimes ........................ 66 Fluid calibration .................... 41 Black Oil............................ 41 Compositional ................... 42 Fluid data.............................. 32 Forchheimer gas, IPR........... 89 Gas Lift Design............................. 145 Diagnostics ..................... 145 instability ......................... 145 Horizontal Completions ........ 91 How to ... Analyis a field over time .... 50

8-195 Analyse artificial lift requirements ................. 47 Analysis a production well. 45 Calibrate a fluid ................. 41 Create GOAL curves......... 47 Create reservoir tables...... 48 Design a Multiphase Booster ...................................... 44 Develop a pipeline & facilities model ............................ 42 Find the optimal completion length ............................ 48 Match data to a flow correlation ..................... 42 Model a multi-lateral well .. 51 Perform a field wide optimization ................... 50 Perform a Nodal Analysis . 45 produce a pressure / temperarture plot ........... 43 Set boundary conditions ... 49 Size equipment ................. 43 Inflow Performance .............. 87 Jones gas, IPR ..................... 89 Jones liquid, IPR .................. 87 Limitations of Model & Component ....................... 39 Model components overview 35 Multiphase Boosting ........... 121 Contra-Rotating Axial...... 133 Dynamic Type ................. 130 Helico-Axial ..................... 131 Positive Displacement Type .................................... 126 Progressing Cavity.......... 129 Twin Screw ..................... 127 Multiple Layers / Completions ........................................ 103 Multi-rate tests gas IPR ............................. 90

PIPESIM

8-196

Case Studies Steam, fluid type................... 35 Straight line PI liquid, IPR..... 88 Stream Re-injection ............ 135 Support Services .................. 28 Units System ........................ 31 Viscosity Gas ................................... 60 Liquid ................................ 56 Live Oil.............................. 57 Viscosity Dead Oil............................ 56 Vogel, IPR ............................ 88 Well PI, IPR .......................... 88

liquid IPR .......................... 89 Nodal Analysis.................... 141 Future IPR ...................... 142 Liquid Loading line .......... 142 Oil/Water Mixture Viscosity... 59 Optimization module performance curves ........ 143 Pressure Drop Calculation.... 65 Pseudo-Steady state IPR ..... 90 Pseudo-Steady state, IPR .... 88 Security Device .................... 26 Separator............................ 135 Single Phase Pump ............ 121

PIPESIM

Information

PIPESIM User Guide

196 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

109225


You might also be interested in

BETA
Oracle Argus Safety
Payroll Malaysia (PY-MY)