Petroleum Experts

User Manual

IPM
Flow Assurance User Manual
Version 2.5
November 2020
 Flow Assurance
IPM - Flow Assurance User Manual
by Petroleum Experts Limited
 3

Copyright Notice
The copyright in this manual and the associated computer program are the property of Petroleum Experts
Ltd. All rights reserved. Both, this manual and the computer program have been provided pursuant to a
Licence Agreement containing restriction of use.

No part of this manual may be reproduced, transmitted, transcribed, stored in a retrieval system, or
translated into any language, in any form or by any means, electronic, mechanical, magnetic, optical or
otherwise, or disclose to third parties without prior written consent from Petroleum Experts Ltd., Petex House,
10 Logie Mill, Edinburgh, EH7 4HG, Scotland, UK.

© Petroleum Experts Ltd. All rights reserved.

IPM Suite, GAP, GAP Transient, PROSPER, MBAL, PVTp, REVEAL, RESOLVE, IFM, IVM, Model
Catalogue, OpenServer and MOVE are trademarks of Petroleum Experts Ltd.

We also recognise the registered trademarks of the following corporations that we may make reference to in
this manual: Microsoft, Schlumberger, Honeywell Process, Rock Flow Dynamics, Kongsberg Digital, AVEVA
SimSci, Halliburton, Stone Ridge Technology, CMG, AspenTech, Beicip-Franlab, ConocoPhillips, Emerson,
Shell, ExxonMobil, Saudi Aramco, BP, Chevron, WellDrill, SSI.

The software described in this manual is furnished under a licence agreement. The software may be used
or copied only in accordance with the terms of the agreement. It is against the law to copy the software on
any medium except as specifically allowed in the license agreement. No part of this documentation may be
reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying,
recording, or information storage and retrieval systems for any purpose other than the purchaser's personal
use, unless express written consent has been given by Petroleum Experts Limited.

Address:

Petroleum Experts Limited

Petex House
10 Logie Mill
Edinburgh, Scotland
EH7 4HG

Tel : (44 131) 474 7030

Fax : (44 131) 474 7031

email: [email protected]
Internet: www.petex.com

© 1990-2020 Petroleum Experts Limited

 I Flow Assurance

Table of Contents
0

Chapter 1 Introduction 2
1 Pre Requisites
...................................................................................................................................
of using this manual 2
2 Flow Assurance
...................................................................................................................................
Studies 2
3 Purpose
...................................................................................................................................
of the manual 2

Chapter 2 Steady state Flow Assurance Examples 5

1 The Production
...................................................................................................................................
System 5
2 Summary
...................................................................................................................................
of the issues faced by the field 6
3 Flow assurance
...................................................................................................................................
issues in the logie mill fields 7
4 Objectives
...................................................................................................................................
of the Model 7
5 Approach
...................................................................................................................................
to Case studies 8
6 All Cases
...................................................................................................................................
Integrated: the final solution 9
7 Application
...................................................................................................................................
Functionality 10
8 Erosion
...................................................................................................................................
during design and Operation of the field (Operating Envelopes) 10
Introduction
.......................................................................................................................................................... 10
Multiphase
.........................................................................................................................................................
pipeline sizing & Erosional Velocities 10
API Design
.........................................................................................................................................................
Standard (PROSPER) 10
An Alternative
.........................................................................................................................................................
to API RP 14E (PROSPER) 11
DNV.........................................................................................................................................................
Erosion Model (PROSPER) 11
Recommended
..........................................................................................................................................................
Design Approach 11
Example.......................................................................................................................................................... 12
Objective
......................................................................................................................................................... 12
Step.........................................................................................................................................................
1 Objective: Prepare the GAP model 12
Step.........................................................................................................................................................
2: Bottlenecking & pipe sizing 14
Step.........................................................................................................................................................
3: Quality check design in the Time domain 16
Step.........................................................................................................................................................
4: Manually detecting Erosion 17
Step.........................................................................................................................................................
5: Automated Erosion Detection using Workflows 20
Declare Variables
......................................................................................................................................... 20
Start Building
.........................................................................................................................................
Workflow 21
Flow Assurance
.........................................................................................................................................
Results from GAP 21
Data for Erosion
.........................................................................................................................................
Object 22
Erosion Calculator
......................................................................................................................................... 24
Run the
.........................................................................................................................................................
Model 25
Analysis
.........................................................................................................................................................
of Results 29
Mitigation Considerations
......................................................................................................................................... 31
9 Corrosion
...................................................................................................................................
Material Design (CO2) 31
The General
..........................................................................................................................................................
Approach 32
The Proposed
..........................................................................................................................................................
Approach 33
Worked Example
.......................................................................................................................................................... 33
Objective
......................................................................................................................................................... 33
Overall
.........................................................................................................................................................
Workflow 34
Step.........................................................................................................................................................
1: Preparing the files 34
Step.........................................................................................................................................................
2: Single Solve Analysis 37
 Contents II

Step.........................................................................................................................................................
3: Forecast Analysis 39
Step.........................................................................................................................................................
4: Building the Workflow 41
Step.........................................................................................................................................................
5: Publishing the Variables 47
Analysis ..........................................................................................................................................................
of Results 50
Corrosion
.........................................................................................................................................................
Rate Vs Time 50
Realtime
.........................................................................................................................................................
Corrosion Allowance Vs Corrosion Allowance 51
10 Separator
...................................................................................................................................
Stability 52
Introduction
.......................................................................................................................................................... 52
Objective.......................................................................................................................................................... 52
Example
.........................................................................................................................................................
Steps 53
Step 1 Objective:
.........................................................................................................................................
Prepare the GAP model 53
Step 2 Objective:
.........................................................................................................................................
Run a forecast 55
Step 3 Objective:
.........................................................................................................................................
Investigate the main trunk line to delivery point 56
Step 4 Objective:
.........................................................................................................................................
Evaluate different separator sizes at early field life 58
Step 5 Objective:
.........................................................................................................................................
Evaluate selected separator size at late field life 60
Conclusion
.......................................................................................................................................................... 65
11 Pipeline
...................................................................................................................................
Stability 66
Introduction
.......................................................................................................................................................... 66
POTS
.........................................................................................................................................................
Criteria 66
PE5 .........................................................................................................................................................
Stability Flag 67
Objective.......................................................................................................................................................... 67
Example..........................................................................................................................................................
Steps 67
Step.........................................................................................................................................................
1 Objective: Prepare the GAP model & Run a forecast 67
Step.........................................................................................................................................................
2 Objective: Manually detect Pipeline Instability 69
Step.........................................................................................................................................................
3 Objective: Automate detection of Pipeline Instability 73
Reset Flag ......................................................................................................................................... 75
Flow Assurance
.........................................................................................................................................
Calculation 75
FA Results Loop
......................................................................................................................................... 77
Get PE5 Stability
.........................................................................................................................................
Status 78
PE5 Stability
.........................................................................................................................................
= 1? 79
Action ......................................................................................................................................... 80
Terminator ......................................................................................................................................... 81
Step.........................................................................................................................................................
4 Objective: Run the RESOLVE forecast 82
Step.........................................................................................................................................................
5 Objective: Add mitigation into the visual workflow 83
Conclusion
.......................................................................................................................................................... 84
12 Well Stability
................................................................................................................................... 85
Introduction
.......................................................................................................................................................... 86
Objective.......................................................................................................................................................... 87
Approach.......................................................................................................................................................... 87
Example..........................................................................................................................................................
Steps 87
Prepare
.........................................................................................................................................................
the GAP model & Run a forecast 88
Manual
.........................................................................................................................................................
instability Detection 89
Mitigation
.........................................................................................................................................................
strategies 91
Surface
.........................................................................................................................................................
Jet Pump Mitigation Approach 92
Step 1 Objective:
.........................................................................................................................................
Prepare the GAP model 92
Step 2 Objective:
.........................................................................................................................................
Identify instability in a well. 93
Tubing string.........................................................................................................................................
Approach 95
SJP Approach
......................................................................................................................................... 95
Surface Jet ...................................................................................................................................
Pump 95
Artificial Lift................................................................................................................................... 98
Reservoir Injection
...................................................................................................................................
in GAP 98
Additional ...................................................................................................................................
Implications to reservoir injection 100
Automated.........................................................................................................................................
Detection and Mitigation using RESOLVE 101

November, 2020 Flow Assurance

II
 III Flow Assurance

Step 1 Objective:
...................................................................................................................................
Open the RESOLVE file 101
Step 2 Objective:
...................................................................................................................................
Import the variables 102
Step 3 Objective:
...................................................................................................................................
Setting up scenarios 106
Step 4 Objective:
...................................................................................................................................
Run forecast and view results 110
13 Asphaltene
...................................................................................................................................
Prediction and Mitigation 111
Introduction
.......................................................................................................................................................... 111
Modelling
..........................................................................................................................................................
Objective 112
Detection
.......................................................................................................................................................... 113
Step
.........................................................................................................................................................
1: Setup the GAP File 113
Step
.........................................................................................................................................................
2: Setup the RESOLVE file 114
Step
.........................................................................................................................................................
3: EOS-PVT (Data Object) 114
Step
.........................................................................................................................................................
4: Workflow Creation 117
Define the Operating
.........................................................................................................................................
Conditions 118
Calculate the
.........................................................................................................................................
Saturation Pressure for the Inlet and Outlet Conditions 118
Define the Conditions
.........................................................................................................................................
for Asphaltene Appearance 118
Run.........................................................................................................................................................
the Prediction 120
Analysis
.........................................................................................................................................................
of Results 122
Mitigation
.......................................................................................................................................................... 123
14 Hydrates
...................................................................................................................................
Prediction and Mitigation 123
Introduction
.......................................................................................................................................................... 123
General..........................................................................................................................................................
Approach to Hydrate Analysis 124
Recommended
..........................................................................................................................................................
approach to Hydrate Analysis 125
Example
..........................................................................................................................................................
Steps 126
Manual
.........................................................................................................................................................
Detection and Mitigation 126
Automated
.........................................................................................................................................................
Detection and Mitigation 130
Step
.........................................................................................................................................................
by step 130
Run.........................................................................................................................................................
the Model and View Results 144
15 Wax ...................................................................................................................................
Prediction and Mitigation 145
Introduction
.......................................................................................................................................................... 146
Recommended
..........................................................................................................................................................
Approach for analysis: 146
Modelling
..........................................................................................................................................................
Objective 147
Example
..........................................................................................................................................................
Step by Step 148
Step
.........................................................................................................................................................
1: Setup the GAP File 148
Step
.........................................................................................................................................................
2: Detection of Wax 149
Step
.........................................................................................................................................................
3: Mitigation Approaches 150
Step
.........................................................................................................................................................
4: REOLVE setup 151
Overall
.........................................................................................................................................................
Workflow 152
Building
.........................................................................................................................................................
the workflow 153
Step 1: Well
.........................................................................................................................................
Count 154
Step 2: Set.........................................................................................................................................
Loop 154
Step 3: Operation
......................................................................................................................................... 155
Step 4: If Statement
......................................................................................................................................... 158
Step 5: Field
.........................................................................................................................................
Model response 158
Step 6: line.........................................................................................................................................
cleaning required? 158
Step 7: increase
.........................................................................................................................................
the dP of the well 159
Step 8 ......................................................................................................................................... 159
Step 9 ......................................................................................................................................... 160
Analysis
.........................................................................................................................................................
of Results 161

Chapter 3 Logie Mill Production Field (Full Model) 165

1 Combining
...................................................................................................................................
the Previous Examples 165
Objectives
.......................................................................................................................................................... 165
 Contents IV

The Full..........................................................................................................................................................
Model using a Steady State Approach Only 165
Create
.........................................................................................................................................................
the Full Model Template 166
Data
.........................................................................................................................................................
Store Variables 169
Importing
.........................................................................................................................................................
Workflows 170
How to Save
.........................................................................................................................................
a Workflow 170
How to Import
.........................................................................................................................................
a Workflow 171
Defining
.........................................................................................................................................................
the System Options 173
Importing
.........................................................................................................................................................
Variables 173
Define
.........................................................................................................................................................
the Schedule 181
Run.........................................................................................................................................................
the Model 181
Analysis..........................................................................................................................................................
of Results using the Steady State Approach 181
Erosion
.........................................................................................................................................................
Results 182
Corrosion
.........................................................................................................................................................
Results 183
Pipeline
.........................................................................................................................................................
Stability 184
Well.........................................................................................................................................................
Stability 185
Wax
.........................................................................................................................................................
Formation 185
Hydrate
.........................................................................................................................................................
Formation 187
Asphaltene
.........................................................................................................................................................
Formation 188
2 Conclusions/Summary
................................................................................................................................... 188

Chapter 4 Water Chemistry: Reservoir and Wellbore

Scale 193
1 Introduction
................................................................................................................................... 193
2 Incompatabile
...................................................................................................................................
Water Injection 195
Objective
.......................................................................................................................................................... 195
Integrated
..........................................................................................................................................................
Model 196
Setting ..........................................................................................................................................................
up the Files 196
Water Chemistry
..........................................................................................................................................................
Data Objects 201
Water
.........................................................................................................................................................
Chemistry Data Object 201
Water
.........................................................................................................................................................
Chemistry Mixer 201
Water
.........................................................................................................................................................
Chemistry PVT Mixer 201
Adding
.........................................................................................................................................................
Water Chemistry and PVT Data Objects 202
Import or
..........................................................................................................................................................
Manually Complete Water Chemistry Data Objects 202
Import
.........................................................................................................................................................
Water Chemistry Object *.RDO File 202
Manual
.........................................................................................................................................................
Completion of the Water Chemistry “Mixer” Data Objects 203
Manual
.........................................................................................................................................................
Completion of the Water Chemistry PVT Mixer 203
Import
.........................................................................................................................................................
PVT Data Object’s Compositional PVT Data 204
Add Visual
..........................................................................................................................................................
Workflow Elements to RESOLVE Canvass 205
Visual
.........................................................................................................................................................
Workflow Functions 205
Import
.........................................................................................................................................................
Visual Workflow 205
Create
.........................................................................................................................................................
Visual Workflows 206
Linking
.........................................................................................................................................................
Wells, DataObjects and Visual Workflows 213
Importing
..........................................................................................................................................................
Application Variables into RESOLVE 214
Entering..........................................................................................................................................................
the RESOLVE Forecast Schedule 216
Analysis..........................................................................................................................................................
of Results 217
Direct
.........................................................................................................................................................
Access of REVEAL Results 217
RESOLVE
.........................................................................................................................................................
Published Variable Results 221
Published
.........................................................................................................................................................
Variable and Results Plotting in RESOLVE 221
View Forecast
.........................................................................................................................................
Results (table) 222
View Forecast
.........................................................................................................................................
Plots 223
Calculating
.........................................................................................................................................................
Barite Mass Deposition from Part-Per-Million Concentrations 226
Summary
.......................................................................................................................................................... 227

November, 2020 Flow Assurance

IV
 V Flow Assurance

3 Calcite
...................................................................................................................................
in Well 227
Objective
.......................................................................................................................................................... 227
Formation
..........................................................................................................................................................
and Physical Phenomenon 227
Design and
..........................................................................................................................................................
diagnosis 228
Integrated
..........................................................................................................................................................
Model 228
Setting ..........................................................................................................................................................
up the Files 230
Water Chemistry
..........................................................................................................................................................
Data Objects 231
Water
.........................................................................................................................................................
Chemistry Data Object 231
Water
.........................................................................................................................................................
Chemistry PVT Mixer 231
Adding
.........................................................................................................................................................
Water Chemistry and PVT Data Objects 232
Import or
..........................................................................................................................................................
Manually Complete Water Chemistry Data Objects 232
Import
.........................................................................................................................................................
Water Chemistry Object *.RDO File 232
Manual
.........................................................................................................................................................
Completion of the Water Chemistry PVT Mixer_2 233
Import
.........................................................................................................................................................
PVT Data Object’s Compositional PVT Data_2 233
Import
.........................................................................................................................................................
Application Variables 234
Setup
.........................................................................................................................................................
DataSets 235
Add Visual
..........................................................................................................................................................
Workflow Elements to RESOLVE Canvass 236
Import
.........................................................................................................................................................
Application Variables 237
Creation
.........................................................................................................................................................
of the Calcite Envelope 237
Checking
.........................................................................................................................................................
operating conditions against the Operating Envelope 238
Mitigation
......................................................................................................................................................... 241
Analysis..........................................................................................................................................................
of Results 242
Summary
.........................................................................................................................................................
and Analysis of Results 242

Chapter 5 Steady state and Transient Flow

Assurance Examples 247
1 Stability:
...................................................................................................................................
comparing Steady State Vs Transient simulations 247
Introduction
.......................................................................................................................................................... 247
Manual ..........................................................................................................................................................
Approach 247
Run Simulation
..........................................................................................................................................................
at Early Time 248
Early
.........................................................................................................................................................
Time (Steady State) 251
Early
.........................................................................................................................................................
Time (Transient) 253
Early
.........................................................................................................................................................
Time Analysis & Results 259
Run Simulation
..........................................................................................................................................................
at Late Time 259
Late.........................................................................................................................................................
Time results 261
Analysis..........................................................................................................................................................
& Results 261
2 Stability:
...................................................................................................................................
Automated initiation of transient simulations 263
Introduction
.......................................................................................................................................................... 263
Workflows
.......................................................................................................................................................... 264
PreSolve
......................................................................................................................................................... 264
Detect
......................................................................................................................................................... 265
Mitigation
......................................................................................................................................................... 266
Chapter

1
 2 Flow Assurance

1 Introduction

1.1 Pre Requisites of using this manual

This manual is centered on formulating a modelling strategy for flow assurance analysis using
a holistic approach. As such, this manual will assume that familiarity exists with all the
underlying applications used (MBAL, PROSPER, GAP, REVEAL, RESOLVE, LEDAFlow and
Excel). The physics behind the native functionality of these applications will only be briefly
discussed in the context of the case, as much more depth and general information is already
available in the respective underlying application user guide.

1.2 Flow Assurance Studies

Flow assurance is the proactive detection and mitigation of flow related physical phenomena
which unchecked will likely worsen, and eventually disrupt production. The severity of
disruption can vary from temporary shutdowns, to catastrophic containment failures.
Classically all such analyses were performed in transient point source models: common issues
to this isolated modelling approach are outlined below:
Approach: Often this type of analysis was driven by operational needs (i.e. phenomena
not considered a priority during the design phase, having to be considered subsequently.
Typical phenomena are surging, slugging, corrosion, erosion, hydrates and waxes
formations)
Data Used: The data used to feed the transient models rarely had anything to do with
reality. Using data from the last known calibrated conditions (last known measurements
of compositions, rates and pressures) does not paint an accurate picture.
Transparency of analysis: The flow assurance engineers performing the analysis have
limited to no interaction with any disciplines upstream of the wellhead. This is also true of
the tools and types of analysis performed. The impact of these studies is diluted as this
knowledge is not propagated to all involved in the production system, and can be siloed
into an operation analysis that is performed adhoc rather than continuously.
Types of Analysis: studies have historically focused on hydraulic related phenomena
like pipeline or riser slugging. Only in the last 10 years have PVT related issues (Waxes,
Hydrates, Asphaltenes) been considered in a more holistic way by thermodynamic and
transient models/operators. This has been driven in the past by the shift in our industry
towards exploration in deeper and harsher environments, containing more complex
fluids.

1.3 Purpose of the manual

To address the implicit assumptions and limitations of the above mentioned issues this manual
attempts to look at the subject of flow assurance in a holistic way. For example, on any one
day a production system may have workovers, maintenance and regular production activities

Flow Assurance November, 2020

 Introduction 3

occurring. There are some parts of this system that are best described using steady state
techniques, and others that are better captured using transient techniques. It is rare that any
one production system is in fully formed steady state or transient conditions, this is only true
for parts of the system.
Modelling strives to capture reality in physics based software and thus, any modelling
approach must capture both steady state and transient effects: without this, only a portion of
the system is understood. Practically this is difficult to do, and the approach to date by
engineers and software developers is to use one or the other.

Assuming steady state at all times has limitations, while assuming transient conditions at all
times is not practical in terms of computational time or effort. As such a platform that can
model steady state for most of the production system, and then initiate transient simulations
for any given part of the system is required. This can be achieved through the new GAP
transient, which enables some pipelines to modelled as transient and some as steady state.
This is also possible within RESOLVE: the vendor neutral integration platform used for solution
formulation in the upstream oil and gas production context. In RESOLVE, the initiation of the
transition between approaches must be controllable based upon (i) the user initiating
(manually) and (ii) based upon events in the model (dynamically). The triggered transient
calculation can then be performed using a third party transient simulator such as LedaFlow.

The manual is structured with the above in mind; an integrated steady state case study is first
presented which details erosion, corrosion, stability, asphaltenes, hydrates and wax. The
detection and mitigation techniques are then brought together into one model. The manual then
discusses predicting scale build up occuring with incompatible water injection using a reservoir
model and also detecting and mitigating calcite formation in a well.

The final section focusses on investigating Transient flow assurance problems, both on a well
level and network level.

© 1990-2020 Petroleum Experts Limited

Chapter

2
 Steady state Flow Assurance Examples 5

2 Steady state Flow Assurance Examples

2.1 The Production System

The production system is made up of two retrograde condensate reservoirs Logie Mill South-
West (LMSW) and East (LME), the former having been produced via two wells between 1998
and 2004, and the latter by a single well between 2000 and 2004. A schematic of the GAP
model is shown below:

The production fluid is delivered via the wells from both reservoirs, through a 39 km pipeline
that passes over undulating terrain as shown below before finally arriving at the delivery point:

© 1990-2020 Petroleum Experts Limited

 6 Flow Assurance

Key Questions regarding the operation of this field are outlined below. These will be
addressed individually in subsequent parts of the manual in the form of separate case
studies. The final case study, however, will bring together all the cases into a single
integrated model.

2.2 Summary of the issues faced by the field

The reservoirs are unconsolidated and thus sand production mitigation methods such as
gravel packs have been installed as standard in each well. Despite this operation
engineers have noted sand production. What is the impact of this?
Each Reservoir is currently producing a small amount of CO2 and operational engineers
are questioning the amount of corrosive thickness specified in the pipeline designs, and
its adequacy.
Process engineers are expecting stability issues with the long pipeline (slugging and
surging), and are also considering pigging operations. They have asked whether the
specified separator capacity can accommodate swept volumes from the pigging
operations.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 7

2.3 Flow assurance issues in the logie mill fields

The sand laden fluid would indicate that erosion may be an issue, particularly in early life
when the producing fluid velocities are high.
CO2 corrosion should be considered, in particular the detection and amount of protection
required.
Liquid accumulation in low points of the 39 km pipeline may result in stability issues in the
pipeline and the usual operational mitigation would be pigging the line. This in turn raises
the question of whether the separator capacity is sufficient to contain the surge volume.
The instability of the surface pipelines (increased backpressure) combined with reducing
fluid driving gradient over time (Reservoir pressure), may result in well instability (liquid
loading) issues that unchecked lead to well shut-in.
Hydrate, Wax and Asphaltene formation detection and mitigation mechanisms need to be
in place, as these phenomena are more common when retrograde condensates have a
low molecular weight for the c7+ plus fraction as is the case for the Logie Mill Fields.

2.4 Objectives of the Model

Detect flow assurance phenomena that may impinge the production systems ability to
produce, in day 1 of production and over the next 10-15 years of field life.
Propose and implement (where practical) Mitigation actions for the detected
phenomena in day 1 of production and over the next 10-15 years of field life. The
mitigation may be in the form of field management rules (to stay within operational
envelopes), or fundamentally changing the operation strategy (chemical inhibition)
Based upon events in the model, the model should be able to decide which modelling
approach (transient or steady state) is appropriate for a given element in the production
system.
Stay within the defined operational window over time to avoid known issues (see below
diagram)

© 1990-2020 Petroleum Experts Limited

 8 Flow Assurance

The challenges in early field life are different to those in later field life as the operational
window shrinks with time. For instance as the reservoir depletes there is less driving force and
well instability is more likely to be an issue. Considering the Logie Mill condensate production
described abover, if the decision in early field life was taken to use large tubing to maximize
production, the consequence in late field life would be an accelerated onset of well instability:
thus reducing the operational window.

2.5 Approach to Case studies

Each of the cases will have a similar construct:
Introduction to the flow assurance issue and its impact
How – if possible – this could be avoided at the design stage
How to DETECT and MITIGATE if these phenomena manifest (i.e. considering an
inherited production system that the current engineers had little to no design input on).
Any design carried out must consider the service life of the equipment and the conditions to
which it will be subjected to: this is summarized in the diagram below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 9

All of the above will be captured in the following worked examples.

2.6 All Cases Integrated: the final solution

All the sub cases are combined together in the final model and can be seen below. The
different flow assurance analyses are performed using data objects. The automated switch
between steady state and transient analysis is also specified on the same workspace. All of
this is brought together in RESOLVE: the vendor neutral platform for advanced solution
formulation.

© 1990-2020 Petroleum Experts Limited

 10 Flow Assurance

2.7 Application Functionality

This entire manual will not be bound by any one software application: the analysis and
phenomena detection and mitigation (regardless of where it is modelled) will draw upon many
different applications, and not be limited to steady state or transient theory alone. This will be
mostly achieved through the use of data objects.
The concept of data objects and workflows is simply to expose the native functionality within
the IPM tools in an easy and digestible manner, so that advanced analysis can be performed
in RESOLVE. As such the below examples will draw upon all the native flow assurance
functionality within the IPM tools, and these will be deployed across all platforms. For instance
the PROSPER engine may be used for detection of a phenomena in the surface network, and
this would be applied to all elements in GAP and controlled from RESOLVE. The RESOLVE
workflow would then be using this detection as a criteria to consider (i) mitigation actions in the
model as these event are dynamically detected, or (ii) performing transient analysis to see the
severity of the phenomenon and its plausible mitigation.

2.8 Erosion during design and Operation of the field

(Operating Envelopes)

2.8.1 Introduction
The general approach to performing design of any one piece of equipment in the oil and gas
industry is to establish a design criteria and expose this criteria to the overriding equations that
capture the physical response of the equipment. Historically this was the limit of the design
process, as many iterations of hand calculations were not practical. More recently, however,
software has provided a conduit to parameterize both the equations used and the input values,
allowing multiple scenarios (experimental designs) in the design stage to be considered. The
pipeline design standards give some guidance on the design and operation process and this is
outlined below:

2.8.1.1 Multiphase pipeline sizing & Erosional Velocities

The minimum fluid velocities in multiphase systems must be kept fairly high to ensure a
consistent mixture and prevent/minimise surging/slugging behaviour. As a rule of thumb the
minimum velocity should be kept to at least 10-15 ft/s (as described in the B3.14 design
standards) whereas the maximum velocity should be limited to 60 ft/s top inhibit noise (this
comes from piping design standards).
There are three methods available to predict erosional velocities in IPM:

2.8.1.2 API Design Standard (PROSPER)

Erosion can be caused by the repeated impact of solid particles on tubing and pipelines. To
avoid this we attempt to estimate the velocity at which erosion will occur. Normal practice is to
use equations outlined in API 14 E with a User input C value. This approach can be unreliable,

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 11

especially for clean production where the limiting value of C (125) can be too restrictive. In
practice, values of 1000 for C have been recorded in pipes where no erosion has been
detected.
An interesting paper that deals with the subject is "Criteria for Sizing Multiphase Flowlines for
Erosive/Corrosive Service" by S.J. Svedeman - SPE Paper 26569. In cases where sand is
present, even at very low velocities, erosion will be a problem and there are no guidelines on
the amount of reduction to the calculated velocity in these cases. Limiting the velocity in
flowlines to an unnecessarily low values will result in costly oversized flowlines.

2.8.1.3 An Alternative to API RP 14E (PROSPER)

A ConocoPhillips paper (An Alternative to API14E Erosional Velocity Limits for Sand Laden
Fluids – ASME Transaction - Energy Resources Technology Journal) challenges API14E on
the basis that it can be very conservative for clean service and is not applicable for conditions
where corrosion or sand are present. It proposes a simple alternative approach that has been
verified by a comparison with several multi-phase flow loop tests that cover a broad range of
liquid-gas ratios and sand concentrations. Values of S for different components are proposed
in the paper

2.8.1.4 DNV Erosion Model (PROSPER)

This is a new edition in IPM 10 to PROSPER (and the data objects in RESOLVE) and stems
from the empirical models created by DNV studying erosion due to sand laden fluid and its
equation and geometries are outline in the PROSPER user guide.

2.8.2 Recommended Design Approach

The approach advocated by Petroleum Experts is to perform the classical design technique of
satisfying the physical equations and design criteria by changing the design parameters (in the
pipeline context length, diameter). However once this process is complete, the established
design should be exposed to varying operating conditions that the design will have to withstand
over its operational life, and is best performed in Physics based tools that capture the
response of integrated systems.
The design approach as such usually starts with the smallest possible size and then
incrementally increases. The example outlined below goes through;
1. Preparing the GAP model for flow assurance calculation.
2. Checking bottlenecking and accordingly adjusting pipe sizing.
3. Quality checking the design using a prediction within GAP.
4. Manually detecting the erosion using the RESOLVE DNV erosion Data Object (step used as
a foundation for automating the process)
5. Using the DNV calculator data object and a Visual Workflow to automatically calculate the
erosion rate and hence the expected total erosion over the entirety of the prediction.

© 1990-2020 Petroleum Experts Limited

 12 Flow Assurance

2.8.3 Example

2.8.3.1 Objective
The objective of this exercise is to show how the erosion calculations native within PROSPER
can be used to quality check the design of pipelines (in GAP). This exercise first demonstrates
how to do this manually using a RESOLVE data object and entering the conditions in the
interface. It then shows the steps required to automate the process to understand the
cumulative erosion that can occur. Additionally, the advantage of the automated approach is
that RESOLVE can be used to perform multiple scenarios where the same logic could be
applied.

2.8.3.2 Step 1 Objective: Prepare the GAP model

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.1 Erosion\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Start GAP and go to File | Archive | Extract. Select the archive (*.gar) file.
If the "Open Master File?" question is prompted, select "No". This step ensures the
underlying models are extracted into the folder, in this case, the GAP model.
To extend the flow assurance studies in this GAP model, we will now continue to use the
Petroleum Experts 5 (PE5) mechanistic flow model for all trunk lines in the GAP model. PE5 is
an advanced mechanistic flow model which is capable of modelling any fluid type over any well
or pipeline trajectory.
We will use the execute OpenServer function to change all pipeline flow models to PE5. Select
the pipeline ‘LM2ToLM1’ and right-click on the correlation dropdown box to select ‘Open
Server’. Alternatively, hold down the ‘Ctrl’ key and right-click on the correlation dropdown box.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 13

In the new Open Server window, copy the OpenServer access string and value.

Click ‘OK’ to close the Open Server window and navigate to Edit | Execute Open Server
statement … from the main dropdown menu.

© 1990-2020 Petroleum Experts Limited

 14 Flow Assurance

In the new window, enter the pipeline correlation OpenServer string and value and click
‘Evaluate’. Use the $ operator in the PIPE syntax to automatically change all pipelines to PE5
– as shown below.

Click ‘Finished’ to return to the main GAP screen.

2.8.3.3 Step 2: Bottlenecking & pipe sizing

For the early life of the pipeline (‘Main Line’) a network solve can be performed to see if the
current 4” pipeline will result in bottlenecking.
When considering a fluid passing through a pipeline, there is an upper limit to how much mass
can be passed through, and this is defined as a “bottleneck”. In GAP a bottleneck is
essentially when the pressure increases upstream, and downstream no more rate can be
passed.
In reality all the multiphase flow correlations do not consider any of this, and this is usually the
domain of choke calculators. Fundamentally the bottleneck concept is based upon the sonic
velocity: i.e. critical velocity when no more rate can pass.
In GAP, flow assurance issues are indicated by a pink warning colour: Solving the network
with optimisation and using the tooltip shows that the pipelines in this network have various
warnings indicated, including bottlenecking (as shown below).

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 15

To address this, the next step would be to incrementally increase the pipeline size until this
bottle neck warning disappeared:
Increase pipeline size to 6”, 8” and 10” to see whether the warning message
disappears.
Another consideration worth having at this stage are the standard sizes of pipeline
available (from manufacturing mills), as this will play a role too.
This approach can be a false starter, as one could make the pipeline too large: the
pipeline is 39km in length and every reduction in diameter results in savings in the
material costs, installation costs, and the size of the support struts that hold the pipeline
in place when over ground (mostly the pipeline is buried but in some cases – roads,
rivers, other lines – this is not always true).
Considering all of the above, the trunk line diameter is increased to 10” and when checked in
early life this does not result in bottlenecking. The next step would be to see how the model
performs over a 16-year prediction period.

© 1990-2020 Petroleum Experts Limited

 16 Flow Assurance

2.8.3.4 Step 3: Quality check design in the Time domain

With the pipeline debottlenecked, we can now run a prediction until 2020 and check the
system response over time. The prediction will run from the end of history – 30/09/2004 – until
the start of 2020, using 2-month timesteps as shown below.

For long forecasts, the Rule based solver should be selected in the final prediction setup
window after specifying the dates.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 17

Inspecting the trunkline results, it can be seen no erosional velocity flags have been raised with
the new larger pipeline. We can also verify this using the RESOLVE DNV erosion object and
use this to automatically detect the erosion and consequently the total expected erosion.

2.8.3.5 Step 4: Manually detecting Erosion

1. Save and close the GAP file (as it will now be opened via RESOLVE)
2. Create the below RESOLVE model

© 1990-2020 Petroleum Experts Limited

 18 Flow Assurance

3. Double click on the GAP module and associate it with the GAP model (*.gap file
extension) from step 1. Set the model to use the rule based solver and to always save
prediction snapshots.

4. If we use the RESOLVE DNV Erosion Data Object, we can see any predicted erosion
rate for the flow conditions (if we assume a sand production of 5 lb per MMscf/d) is

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 19

extremely small and unlikely to cause any issues. The data can be entered into the
Erosion Data object manually, or automated by a workflow.
5. The Solve Network results for the MAIN LINE pipe can be used, to populate the erosion
calculator. These results can be accessed from GAP, as per the screenshot below:

6. When the inputs are entered from the results obtained above, the erosion rate can be
calculated, as shown below:

© 1990-2020 Petroleum Experts Limited

 20 Flow Assurance

2.8.3.6 Step 5: Automated Erosion Detection using Workflows

Using the data object we can also track the erosion rate for any one timestep, or over time
with a visual workflow. Continuing with the assumption that 5 lb per MMscf/d of sand
production exists, the objectives is to calculate the total erosion over time.
Using the Erosion data object, we can construct the visual workflow to report this using the
procedure below.

2.8.3.6.1 Declare Variables

First define the variables required in the workflow to populate the data object and calculate the
erosion rate and total amount of erosion. Note that the ’totalerosion’ variable will only be reset
when a new prediction run is started:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 21

2.8.3.6.2 Start Building Workflow

The workflow that will be built is shown below, and the steps to make this follow the diagram:

It should be noted that all inputs are case dependent, and so care should be taken in the
following steps.

2.8.3.6.3 Flow Assurance Results from GAP

Then begin constructing the workflow. First, add an operation to calculate the flow assurance
results from the GAP pipeline:

© 1990-2020 Petroleum Experts Limited

 22 Flow Assurance

The above is the equivalent of pressing the Flow Assurance button for each time step:

2.8.3.6.4 Data for Erosion Object

An operation to retrieve the superficial velocities and fluid densities from the end of the
pipeline, where (due to fluid expansion) the velocities will be highest i.e. the maximum possible
erosion rates. In addition the gas rate is retrieved, as we have specified a mass rate per unit
production of gas:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 23

These values are then assigned to appropriate inputs in the erosion calculator data object.
Note here we have assumed 5lb/MMscf/d of sand production by multiplying the gas rate by 5.
This can be changed or made a variable if desired:

© 1990-2020 Petroleum Experts Limited

 24 Flow Assurance

2.8.3.6.5 Erosion Calculator

Now that the erosion calculator is populated, the calculation is performed with an operation
element:

The erosion rate is then retrieved and multiplied by the two month timestep (note the rate is
given in inches/yr units, hence the 2/12 multiplier) to give the total erosion over the
timestep. This is added to the rolling total from the previous steps, before ending and moving
on to the next timestep:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 25

Connect the elements together.

2.8.3.7 Run the Model

Before running the model, the variables desired for plotting need to be declared:
1. Go to Variables | Import application variables… and add the Delivery pressure as per
the screenshots below:

© 1990-2020 Petroleum Experts Limited

 26 Flow Assurance

2. Now add the workflow variables to be plotted in a similar manner:

Flow Assurance November, 2020

Steady state Flow Assurance Examples 27

© 1990-2020 Petroleum Experts Limited

 28 Flow Assurance

3. Check the schedule has been defined appropriately:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 29

4. When the run is complete, save the file and create an archive from File | Archive |
Create.
2.8.3.8 Analysis of Results
When the model is now run, we can see the erosion rate can be tracked over time in addition
to the total amount of erosion (see below plot).

© 1990-2020 Petroleum Experts Limited

 30 Flow Assurance

The results can be considered in two phases:

Phase 1: Producing at production Plateau
Phase 2: Producing below production Plateau
Phase 1:
Initially, the rate of erosion increases as per the red line on the above graph. This can be
explained by considering a constant gas rate with increasing WGR results in gas occupying a
smaller space in the pipeline. Same rate, and lower area would equate to faster velocity, and
in the same way increasing the amount of erosion.
Phase 2:
After the production falls off plateau the field decline means the reduced rates also reduce the
amount of erosion (red line) occurring in the pipeline as less gas is being produced.
Ultimately, the total cumulative thickness (green line) eroded over the 16-year prediction is
approximately 0.000006” (0.0001524mm). This indicates that there is basically no erosion for
this assumed rate of sand production. We can also inspect the well results in GAP and note
that the wells LM5 and LM2 exceed their erosional velocities:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 31

2.8.3.8.1 Mitigation Considerations

Erosion is only an issue if the well is producing sand: sand production occurs as a result of
high shearing forces in the reservoir (i.e. high rates/drawdown) that is sufficient to overcome
the cementing forces that hold the grains in place.
One common operational mitigation is to limit the rate that a well can produce (usually using a
wellhead choke), which in turn limits the drawdown. To determine the maximum drawdown
there are three options that could be considered:
1. The sand failure calculator in PROSPER could be used – with some knowledge or
assumptions about rock properties this calculates a maximum drawdown that can be
made before sand production starts (using a simple empirical relationship).
2. Alternatively, the solids transport abilities of REVEAL can determine whether this is an
issue and whether mitigating actions are necessary.
3. Using some other constraint (e.g. in some parts of the world, government legislation is in
place to limit production for environmental reasons).
Should any of these models indicate (i) there will be solids produced whilst (ii) exceeding this
rate, possible mitigating actions may be:
Investigate maximum allowable drawdown and constrain the wells to respect this.
Consider other sand control methods alongside any already installed.

2.9 Corrosion Material Design (CO2)

© 1990-2020 Petroleum Experts Limited

 32 Flow Assurance

2.9.1 The General Approach

When designing a pipeline, an engineer will customarily have to determine the diameter and
wall thickness of the pipeline. Pipeline engineers frequently perform mechanical/hydraulic
design of pipelines referring to engineering design standards. Depending upon the scenario
being considered the engineer will refer to different standards for wall thickness (e.g. B31.3,
B31.4 or B31.8). One such example of B31.3 wall thickness calculation is outlined below:
Wall thickness calculations – using B31.3 Code
ANSI/ASME Standard B31.3 is a very stringent code with a high safety margin. The B31.3
wall thickness calculation formula is stated as:

Where:
= Minimum design wall thickness (inches)

= Corrosion allowance (inches)

= Thread or groove depth (inches)

= Allowable internal pressure in pipe (psi)

= Outside diameter of pipe (inches)

= Allowable stress for pipe (psi)

= Longitudinal weld-joint factor

= Derating factor

= Manufacturer allowable tolerance (%)

The main point of the equations in these standards is simply to ensure that the pipeline does
not buckle under its own weight considering the geometry and pressure of the fluid being
transported. Additional to this are a number of terms that consider manufacturing tolerances
and an allowable corrosion thickness (i.e. corrosion allowance), all of which serve to increase
wall thickness which has an economic implication.
The downstream engineers that perform this design are generally only given a basis of design
(BoD) which outline some basic data to perform the design, and it is rare to have access to
the field development plan, or even the upstream model upon which the field development is
based.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 33

2.9.2 The Proposed Approach

The design process of any element in the production system will is performed using specific
set of design conditions (otherwise there are too many degrees of freedom, and it becomes
difficult to pin down a specific design). However, once this design is complete it should be
exposed to the operating conditions that the element will be subjected to over its operational
life, to assess whether the design is adequate.
In the pipeline context this means to hydraulically/mechanically design a pipeline using the
design standards, and then subject the design to a 16 year forecast to assess whether the
design can stand the test of time. This model based approach is rarely done, and this lack of
rigor can have far reaching operational implications if said pipeline is constructed and
installed.
This section of the flow assurance manual will show how the CO2 corrosion allowance aspect
of pipeline design can be calculated using a model based approach. By knowing the corrosion
rate and the anticipated service life of the pipeline, a designer can calculate the extra
thickness required for corrosion resistance (i.e. a thickness that we accept will be corroded
away over the service life). The CO2 corrosion model in PROSPER should be used as a worst
case scenario, as outlined in the PROSPER user guide (Section 2.11.8.4).

2.9.3 Worked Example

2.9.3.1 Objective
The objective of this exercise is to show how CO2 corrosion calculations that are native within
PROSPER can be used to design long pipelines on the surface (in GAP), from RESOLVE. In
particular, this exercise demonstrates how to determine the corrosion allowance (i.e. extra
flowline thickness) required to ensure the integrity of the pipeline.
The advantage of this approach is that RESOLVE can be used to perform multiple scenarios
where the same logic could be applied.

© 1990-2020 Petroleum Experts Limited

 34 Flow Assurance

2.9.3.2 Overall Workflow

The overall workflow, which will be created in the following steps, is shown below:

2.9.3.3 Step 1: Preparing the files

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.2 Corrosion\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

1. Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder.
2. Setup up a new instance of RESOLVE
3. Add a GAP module from Edit System | Add Client Program | GAP, or alternatively by
following the screenshot below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 35

4. Double click on the GAP module and associate it with the GAP model (*.gap file
extension) extracted in step 1. Set the model to use the rule based solver and to always
save prediction snapshots.

© 1990-2020 Petroleum Experts Limited

 36 Flow Assurance

5. From Options | System Options set the forecast mode to “single solve/optimization
only”.

6. Add the PROSPER Corrosion Calculator Data Object from Edit System | Add Data |
Corrosion-Calculator, or alternatively by following the screenshot below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 37

2.9.3.4 Step 2: Single Solve Analysis

A single solve of the network will now be performed to determine the response of the main
trunk line. The results of this network solve will be used to populate the corrosion calculator
and determine the required corrosion allowance (i.e. extra pipeline thickness) for 16 years of
production.
1. Perform a single solve of the network from Run | Start, or pressing the short cut shown
below. (Please note, this may take a few seconds to complete.)

2. In the graphical view, double click on the corrosion calculator and populate it manually
from the solver results in GAP. The calculation options of the corrosion calculator should
be set to, calculate required thickness and calculate pH.

© 1990-2020 Petroleum Experts Limited

 38 Flow Assurance

The fluid properties (superficial velocities, density and viscosity) can be obtained from
the flow assurance calculator in the solver results. From the MAIN LINE pipeline, enter
the solver results and select the flow assurance button, as shown below.

3. With all the inputs defined, select Calculate to determine the corrosion allowance.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 39

The above approach of determining the corrosion allowance assumes a constant rate of
corrosion over a 16 year period. In reality this will not be the case, as over the lifetime of the
system, both the fluid properties and the CO2 content entering the flowline will change through
time.
To better determine the corrosion allowance required, a visual workflow can be used in
conjunction with a GAP forecast to track the ever changing corrosion rate and determine a
dynamic Corrosion Allowance.

2.9.3.5 Step 3: Forecast Analysis

The objective is now to setup a workflow that uses the Corrosion Calculator Data Object at
each timestep. At each timestep the rate of corrosion will be determined and used to compute
the Realtime Corrosion Allowance for the 16 year period.
1. Insert the OpenServer Data Object from Edit System | Add Data | OpenServer, or
alternatively by following the screenshot below:

© 1990-2020 Petroleum Experts Limited

 40 Flow Assurance

2. Input a DataStore Data Object from Edit System | Add Data | DataStore, or
alternatively by following the screenshot below:

Define two columns in the data store; Dynamic Corrosion Allowance and Constant Rate
Corrosion Allowance.

3. Input a Visual Workflow Module from Edit System | Add Client Program | Workflow
(or by following the screenshot below) and link from the GAP module to the Workflow
module.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 41

2.9.3.6 Step 4: Building the Workflow

Logic can now be defined within the Workflow to perform the following tasks:
Populate input data, calculate, and extract results from the Corrosion Calculator at each
timestep.
Record the Corrosion Allowance determined at the first timestep of the prediction.
Use the Corrosion Rate of each timestep to determine the Realtime Corrosion
Allowance.
The workflow that will be created in the following steps, is shown below:

© 1990-2020 Petroleum Experts Limited

 42 Flow Assurance

1. First define the variables required in the workflow. Go to the workflow, and define the
variables as shown below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 43

2. The first Operation, “Get Calc Inputs,” performs the flow assurance calculation before
extracting all required inputs for use in the Corrosion calculator. The flow assurance
calculation is performed using the generic OpenServer command shown below.

Each of the input variables (superficial velocities, viscosity, CO2 content, etc...) can then
be extracted using the relevant generic OpenServer functions, and given an appropriate
return assignment.

It should be noted that the OpenServer stings are case sensitive so ensure that the
correct equipment names are used.

© 1990-2020 Petroleum Experts Limited

 44 Flow Assurance

3. Set the input data within the Corrosion Calculator using the return assignments defined
above.

4. Setting up the Corrosion Calculation to run.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 45

5. Extract the corrosion rate from the calculator.

6. Define an IF function which recognizes if it is the first timestep of the prediction. If it is

the first timestep, the corrosion allowance for 15 years (i.e. single solve analysis) will be
extracted to the Data Store. If it is not the first timestep, the workflow will continue as
normal.

© 1990-2020 Petroleum Experts Limited

 46 Flow Assurance

7. After each timestep, the corrosion rate will be summed. At the end of the forecast, an
average corrosion rate will be determined by dividing this summation by the number of
timesteps.

8. Following the assignment to sum the corrosion rate at the end of each timestep, an IF
function which recognizes if the forecast has finished can be defined.

9. If it is not the end of the forecast, the run will continue to the next timestep. If it is the end
of the run, the Dynamic Corrosion Allowance will be determined.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 47

2.9.3.7 Step 5: Publishing the Variables

1. After setting up the workflow as per the above, specific variables, including the corrosion
rate, can be published for post forecast analysis from, Variables | Import Application
Variables.

© 1990-2020 Petroleum Experts Limited

 48 Flow Assurance

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 49

2. From, Options | System Options set the forecast mode to Full Forecast

3. Define the prediction schedule from, Schedule | Forecast Data.

© 1990-2020 Petroleum Experts Limited

 50 Flow Assurance

4. Run the prediction from Run | Start.

5. When the run is complete, save the file and create an archive from File | Archive |
Create.
2.9.4 Analysis of Results

2.9.4.1 Corrosion Rate Vs Time

The Norsk Hydro corrosion model implemented within the corrosion calculator considers not
only the effects of CO2, but also wall shear stress. As such, it can be seen that as the system
falls off plateau, the corrosion rate decreases (reduction in wall shear stress).

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 51

The fluctuations in corrosion rate during the first five years are inherent of the rule based
solver prioritizing the wells based on production (maximizing gas rate and penalizing based on
WGR). As wells LM1 and LM2 have a higher CO2 content, when they are choked back to
satisfy the rule based solver, a sudden decrease in corrosion rate is observed.

2.9.4.2 Realtime Corrosion Allowance Vs Corrosion Allowance

From the results in the Data Store, it can be seen that by tracking the rate of corrosion
throughout the forecast a more representative corrosion allowance is determined – in this case
we require less allowance.

The need to use this real-time approach is more profound when the asset in question is
changing throughout the forecast. For instance, drilling wells during the forecast (to meet
market demand) may significantly change the fluid properties/composition entering the pipeline
and must therefore be captured by tracking the corrosion rate.

© 1990-2020 Petroleum Experts Limited

 52 Flow Assurance

2.10 Separator Stability

2.10.1 Introduction
At the end of field life most reservoirs have undergone significant depletion, water production
is usually quite high; and pipelines with varying topography may result in a surging type
behavior (known as hydro dynamic slugging and induced from slippage). Providing that the
slug catcher/production separator is of adequate capacity this is not really an issue to the
downstream processing facilities. This will however, still impact the operation and as such
there are mitigation efforts often put in place in the field control.
The following exercises looks into the detection and mitigation procedures that are common
for separators, pipelines and wells.

2.10.2 Objective
The following vendor data has been provided, and the objective is to select a Separator
capacity that is suitable for the field in question. The field development plan has indicated that
the gas handling capacity of the downstream process is limited to 20 MMscf/d and High
pressure separator is to be operated at 655psig. With this in mind, the objective is to select
the size of separator:
Size
WP APPX. Liquid
DIA x Gas Capacity MMSCFD
Capacity
S/S PSIG WT. 400 600 800 1000 1200 1440
B/D
PSIG PSIG PSIG PSIG PSIG PSIG
720 1692 780 9.2 11.5
24" OD x 1000 2485 750 8.8 10.9 12.5 14.2
8’ 1200 2884 740 8.7 10.8 12.4 14 15.4
1440 3215 720 8.5 10.6 12.1 13.7 15.1 16.4
720 4031 1770 21.2 26.3
36" OD x 1000 5555 1750 20.4 25.3 29.2 33
9’ 1200 6464 1690 20.3 25.2 29 32.8 36.1
1440 7695 1630 18.9 23.5 27.1 30.6 33.7 34.2
720 5570 2420 28.1 34.9
42" OD x 1000 7441 2330 27 33.6 38.7 43.8
10' 1200 9184 2270 26.3 32.7 37.7 42.6 46.8
1440 10650 2230 25.8 32.1 37 41.7 45.9 49.8
720 7730 3170 35 43.5
48" OD x
1000 9773 3060 34 42.1 48.7 54.8
12’
1200 11964 2990 33.2 41.2 47.5 53.6 58.9

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 53

1440 14375 2910 32.2 40 46.2 52.2 57.2 62

To perform this investigation, the Slug Catcher Calculator will be used in conjunction with a
GAP prediction run. This approach ensures that the separator is correctly sized in the early
field lift and late field life.

2.10.2.1Example Steps

2.10.2.1.1 Step 1 Objective: Prepare the GAP model

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.3 Separator Stability\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Start GAP and go to File | Archive | Extract. Select the completed archive file (*.rsa) file
from the previous section in the samples folder.
When the "Open Master File?" question is prompted, select "No". This step ensures the
underlying models are extracted into the folder, in this case, the GAP model.
To extend the flow assurance studies in this GAP model, we will now continue to use the
Petroleum Experts 5 (PE5) mechanistic flow model for all trunk lines in the GAP model. PE5 is
an advanced mechanistic flow model which is capable of modelling any fluid type over any well
or pipeline trajectory.
We will use the execute OpenServer function to change all pipeline flow models to PE5. Select
the pipeline ‘LM2ToLM1’ and right-click on the correlation dropdown box to select ‘Open
Server’. Alternatively, hold down the ‘Ctrl’ key and right-click on the correlation dropdown box.

© 1990-2020 Petroleum Experts Limited

 54 Flow Assurance

In the new Open Server window, copy the OpenServer access string and value.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 55

Click ‘OK’ to close the Open Server window and navigate to Edit | Execute Open Server
statement … from the main dropdown menu.
In the new window, enter the pipeline correlation OpenServer string and value and click
‘Evaluate’. Use the $ operator in the PIPE syntax to automatically change all pipelines to PE5
– as shown below.

Click ‘Finished’ to return to the main GAP screen: all pipelines now have PE5 selected as the
multiphase flow correlation.

2.10.2.1.2 Step 2 Objective: Run a forecast

Navigate to Prediction | Run Prediction… from the main dropdown menu to open the prediction
run setup windows.
Set the prediction end date to 01/01/2020 with a 2 month step size.

© 1990-2020 Petroleum Experts Limited

 56 Flow Assurance

Click ‘Next’ and ‘Next’ again and ensure that the ‘Delivery’ pressure is set to 655 psig.

As this is a long term forecast, run the GAP forecast with the ‘Rule Based’ network solver.

Click ‘Main’ once the prediction has completed to return the GAP main screen.

2.10.2.1.3 Step 3 Objective: Investigate the main trunk line to delivery point
We are interested in the fluid rates and velocities at the delivery point in the GAP model. To
investigate and help select an appropriate separator size, we will compare the flow assurance
results at early and late life of the prediction.
To achieve this, we will use the slug catcher calculator in the flow assurance results.
Select the pipeline ‘MAIN LINE’ and enter the prediction results screen, as shown below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 57

We will first look at the flow assurance results on 30/09/2004 (early life) and select an
appropriately sized separator. Select the ‘View’ button on this date.
This will bring up the gradient traverse screen for this pipeline element. It will display all the
gradient results typically found in a PROSPER gradient calculation within this window.
At the node before the manifold, right-click on any result field and select ‘Slug Catcher
Calculator’ – as shown below.

© 1990-2020 Petroleum Experts Limited

 58 Flow Assurance

2.10.2.1.4 Step 4 Objective: Evaluate different separator sizes at early field life
The slug catcher calculator will automatically populate with input data from the selected node
in the gradient result.
We are interested in knowing of the separator is adequately sized to contain the liquid and gas
volumes. We will work from the largest to smallest separator sizes until we can find a
separator which is a reasonable size (i.e. sufficiently large without being oversized).
Switch the input method to ‘No Slugs’, this will assume that the separator inlet rate is at
steady-state conditions.
Under the geometry section, switch to ‘Dimensional’ and ‘Horizontal’ and enter the largest
separator size (i.e. Diameter 48 inches and (L/D) ratio: 0.625) and click on the
‘Calculate’ button.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 59

For the entered dimensions, the separator volume is 31.4 ft3 while the peak liquid volume is
9.1 ft3 (see above screen shot). As such, the separator is large enough to contain the liquid
volumes produced.
The liquid and gas velocity are such that there are also no carry over or carry under problems
(this phenomena and its calculation is described in more detail in the PROSPER user guide).
As the size of the separator is directly related to the cost, it is favorable to not over size the
separator. At late field life the production rates will decline and the current separator may be
too large – let’s test the next available separator size.
Change the diameter to 42” inches (as this is the next size down on our vendor data
table) and (L/D) ratio to 0.595 and click on the ‘Calculate’ button. As shown below.

© 1990-2020 Petroleum Experts Limited

 60 Flow Assurance

In this case, the separator volume is sufficient to contain the liquid production volumes.
However, there is potential for vapour to carry under to the separated liquid stream due to the
low estimated liquid velocity (see above screenshot).

2.10.2.1.5 Step 5 Objective: Evaluate selected separator size at late field life
Although the separator selected is adequate for early life, it is important to evaluate the impact
of such a separator on operations at late field life. During late field life, there is a greater
chance of well instabilities to occur and a potential for pipeline slugging to affect the operation
of the separator. We will use the slug catcher calculator to evaluate if the current sized
separator is able to hold the surge volumes that may occur due to pipeline slugging.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 61

Return the MAIN LINE prediction results and select ‘View’ from the flow assurance
column for the date 30/09/2012 (which is approximately half way through the prediction).
Open the slug catcher calculator for the node before the manifold. (If required, Step 2
can be reviewed for further guidance on this process.)
In a slugging pipeline, the size of the slug will vary with pipeline trajectory and diameter. As
such there are multiple slug sizes that could be evaluated. We will use the ‘Mean Slug’ method
to describe the slug flow characteristics.
Select the ‘Mean Slug’ input method, enter the separator geometry as per the selected
design and click ‘Calculate’ – as shown below.

© 1990-2020 Petroleum Experts Limited

 62 Flow Assurance

As can be seen from the results above, the current separator has an insufficient volume to
contain the slugs that form in this mid-life time step. This behavior (which is dependent on
multiple contributing factors, including fluid composition and properties, pressure, rates etc.)
extends for several time steps in the mid-life of the MAIN LINE pipeline. To address this, the
size separator must be increased further.
Change the diameter to 52” inches (as this is the next size down on our vendor data
table) and (L/D) ratio to 0.75 and click on the ‘Calculate’ button. As shown below.

As can be seen above, this separator size is now sufficient. There are other factors that have
to be considered before selecting a separator size. These include, though are not limited to
the physical footprint of the separator and whether this can be accommodated on the plant/

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 63

platform and the cost. As such the smaller separator size (48” diameter and 0.625 L/D ratio)
will be selected, with the assumption that a suitable slug catching method is in place to protect
the separator and downstream process.
To evaluate this size selection in late life:
Return the MAIN LINE prediction results and select ‘View’ from the flow assurance
column for the date 01/01/2020.
As can be seen in the screenshot below, slugging has been identified in the mid-region of the
MAIN LINE pipe line in later life. As can be seen from the flow regimes, these slugs dissipate
as they travel through the pipeline. This does not, however, indicate that the implementation of
typical slug catchers (i.e. finger type, vessel type, or parking loop) would not still be
necessary. As a result, further investigation would be required to determine the best slug
catcher design to protect the separator, and thus the downstream process, as mentioned
above. This further investigation is not considered in this example.

Open the slug catcher calculator for the node before the manifold. (If required, Step 2
can be reviewed for further guidance on this process.)
In a slugging pipeline, the size of the slug will vary with pipeline trajectory and diameter. As
such there are multiple slug sizes that could be evaluated. We will use the ‘Mean Slug’ method
to describe the slug flow characteristics.

© 1990-2020 Petroleum Experts Limited

 64 Flow Assurance

Select the ‘Mean Slug’ input method, enter the separator geometry as per the selected
design and click ‘Calculate’ – as shown below.
At the peak surge time, we can see that there is sufficient volume in the separator to contain
the slug and no carry over of liquid or carry under of vapour.

Using the ‘PlayBack’ slider scale, we can look at different points in time as the slug unit travels
through the separator. It can be seen that as the surge volume reduces, the estimated gas
velocity decreases while the estimated liquid velocity increases.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 65

It should be noted that given the low holdup volume there is potential for vapor carry under due
to liquid height or short length of the separator.

2.10.3 Conclusion
The selected separator is adequate for holding the liquid production and any surge volumes
due to pipeline instability when is it is assumed that an appropriate slug catching method is
implemented.
As a final check, we can evaluate the separator performance for the 1/1000 slug. This is a
slug flow characteristic from a normal distribution to give a worst case scenario.
In a worst case scenario, at peak surge, there is insufficient volume to contain the liquid

© 1990-2020 Petroleum Experts Limited

 66 Flow Assurance

production volume. This would have disastrous implications on the downstream operations,
and operational envelopes have to be established to avoid this scenario from occurring.

2.11 Pipeline Stability

2.11.1 Introduction
This case study will look at pipeline stability analysis: this has been classically the domain of
the transient simulation. In this case study the hydraulic stability of the producing fluid over
production life will be considered. There are other facets to this analysis, but they will be built
on gradually over the remaining course of this manual (to include transient simulation too)
Terrain induced riser slugging is well documented in literature, as such the mechanism is not
going to be discussed here. The impact of these slugs are directed at pipework, pipework
support struts, facilities incapable of accommodating surge volumes resulting in shutdowns and
liquid carry over resulting in cross contamination of separator stream.
In terms of modelling these phenomena both steady state and transient modelling should be
applied in series (not in parallel, so that one can initiate the other): there are two methods for
determining the onset of severe slugging in IPM:

2.11.1.1POTS Criteria
The POTS number calculated is a dimensionless number which is the ratio of the rate of
hydrostatic pressure buildup in the riser resulting from the growth of the slug against the rate
of gas pressure buildup in the flowline.
The POTS number has the form shown below:

Where:
LF is the length of flowline
ygF is the gas hold up in the flowline
wg is the gas mass flowrate
wL is the liquid mass flowrate

If the POTS number is greater than 1 then severe slugging is unlikely. If the POTS number is
less than 1 then severe slugging is likely.
The POTS number is sensitive to the ratio of gas and liquid mass rates, and in a condensate
system, the POTS number will likely always be greater than 1 and hence slugging is unlikely to
be detected according to this approach. This is not to say that it would not manifest, but rather
it is not detected.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 67

2.11.1.2PE5 Stability Flag

The PE5 mechanistic flow correlation discretizes the flowline and performs a momentum
balance over each phase travelling through the segment in question. This is physically and
mathematically more rigorous than the empirical correlation approach. One particular aspect
that falls out of this momentum balance approach to multiphase flow is the stability criteria.
The PE5 stability criteria works on the basis of the friction gradient calculated by the PE5
mechanistic flow model. The momentum balance performed by the PE5 mechanistic model
can capture liquid falling back in the opposite direction of flow (resulting in a change from a
positive to a negative frictional gradient). A positive friction gradient indicates stable flow while
the onset of unstable flow can be determined by a negative friction gradient.
The PE5 stability flag is designed to be conservative, with a 1e-4 psi/ft friction gradient
as the cut-off point.
If the flag = 0 then there is confidence that the flow is stable.
If the flag = 1 then it is possible that production is unstable (this can be investigated
further with a transient simulator).

2.11.2 Objective
To detect the onset of pipeline instability and propose mitigation actions to the operations
team. Instability is defined as inconsistent geometric phase behavior, subjecting the pipework
to intermittent stress loads that can be catastrophic if not properly accounted for during the
design phase. The most common example of this is “Severe slugging” which can occur at the
bottom of a riser base or at highly deviated sections of tubing/pipeline.
In this example, we will use the PE5 stability flag criteria to detect the onset of pipeline
instability.

2.11.3 Example Steps

2.11.3.1Step 1 Objective: Prepare the GAP model & Run a forecast

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.4 Pipeline Stability\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

© 1990-2020 Petroleum Experts Limited

 68 Flow Assurance

Start GAP and go to File | Archive | Extract. Select the completed archive (*.gar) file
from samples folder.
When the "Open Master File?" question is prompted, select "No". This step ensures the
underlying models are extracted into the folder.
Open the GAP model, navigate to Prediction | Run Prediction… from the main
dropdown menu to open the prediction run setup windows.
Set the prediction end date to 01/01/2020 with a 2 month step size.

1. Click ‘Next’ and ‘Next’ again and ensure that the ‘Delivery’ pressure is set to 655 psig.

2. As this is a long term forecast, run the GAP forecast with the ‘Rule Based’ network
solver.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 69

3. Click ‘Main’ once the prediction has completed to return the GAP main screen.

2.11.3.2Step 2 Objective: Manually detect Pipeline Instability

We will manually investigate the pipeline stability for ‘FromEast’ using the PE5 stability flag.
1. Select the pipeline ‘FromEast’ and enter the prediction results screen, as shown below.

© 1990-2020 Petroleum Experts Limited

 70 Flow Assurance

2. We will first look at the flow assurance results at 30/09/2008. Right-Click on the gradient
results and select Stability Criteria | PE5 Stability – as shown below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 71

3. The green dots in the PE5 Stability plot indicates that the fluid flow is stable.

© 1990-2020 Petroleum Experts Limited

 72 Flow Assurance

4. Click ‘Done’ and ‘Done’ to return to main pipeline prediction results. Select ‘Flow
assurance’ for the date 30/01/2015 and return to the PE5 Stability plot.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 73

The red dots indicate instability. This is likely caused by loss in fluid momentum due to a
reduction in gas rate production.
It is possible to automate the detection of the PE5 stability criteria through a visual workflow in
RESOLVE. This will be achieved in the next objective step.

2.11.3.3Step 3 Objective: Automate detection of Pipeline Instability

The objective of the visual workflow is retrieve the PE5 stability flag status. This will be
executed at the end of a GAP solve and will be achieved as follows in the visual workflow:
Perform the pipeline flow assurance calculation
Loop through the rows of the gradient results
Identify if the PE5 Stability flag is 1
Report PE5 Stability flag

1. To begin, open RESOLVE and create the workflow, as shown below, linking the
appropriate GAP file:

© 1990-2020 Petroleum Experts Limited

 74 Flow Assurance

2. Enter the visual workflow module and select the add variables button ‘abc’. In the new
window, define the three variables as shown below.

3. Click ‘OK’ to return to the main visual workflow screen. Using the workflow item palette,
recreate the workflow outlined in the figure below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 75

The item label and content will be systematically explained below.

2.11.3.3.1 Reset Flag

The ‘FLAG’ variable will hold the PE5 stability criteria status. Initially, this value will be reset to
zero.

2.11.3.3.2 Flow Assurance Calculation

Two generic OpenServer operations will be executed (1) perform the flow assurance
calculation (2) count the number of rows in the gradient results.
The flow assurance calculation for the pipeline in interest can be performed using the following
syntax:
"GAP.MOD[{PROD}].PIPE[{FromEast}].SolverResults[0].CALCFLOWASSURANCE()"
Within the operation element, select ‘Add global function’ and setup the OpenServer command
function as shown below.

© 1990-2020 Petroleum Experts Limited

 76 Flow Assurance

The number of rows in the gradient calculation can be found by performing a DoGet
OpenServer function with the following syntax:
"GAP.MOD[{PROD}].PIPE[{FromEast}].OUT.GRD.Results[0][0][0].MSD[0].COUNT"
Within the operation element, select ‘Add global function’ and setup the OpenServer command
function as shown below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 77

2.11.3.3.3 FA Results Loop

This element will setup a loop which will get the PE5 stability flag in the gradient results. The
setup of this element is shown below.

© 1990-2020 Petroleum Experts Limited

 78 Flow Assurance

2.11.3.3.4 Get PE5 Stability Status

With the operation element, the PE5 stability status can be retrieved by performing a DoGet
OpenServer function with the following syntax:
"GAP.MOD[{PROD}].PIPE[{FromEast}].OUT.GRD.Results[0][0][0].PE5STB["+I+"]"
The syntax has been broken down into string elements so that the row index in the gradient
results can be represented by the variable I value.
Within the operation element, select ‘Add global function’ and setup the OpenServer command
function as shown below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 79

2.11.3.3.5 PE5 Stability = 1?

The IF element checks if the FLAG variable value is equal to 1. If this is true this will be logged
during the RESOLVE forecast. If this is true, the following Action will be performed.

© 1990-2020 Petroleum Experts Limited

 80 Flow Assurance

2.11.3.3.6 Action
In the ACTION assignment element we will log a message to identify when the pipeline is
unstable during the RESOLVE calculation.
Within the argument box concatenate the message with the variable FLAG, as shown below.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 81

2.11.3.3.7 Terminator
These are exit points for the visual workflow.

© 1990-2020 Petroleum Experts Limited

 82 Flow Assurance

2.11.3.4Step 4 Objective: Run the RESOLVE forecast

1. Run the RESOLVE model, using the Schedule shown below:

2. Note the date when pipeline instability is detected by PE5.

We can see that the pipeline instability has been detected automatically by the visual workflow
on the 30/03/2014. This workflow could be extended to analyze the stability of all pipelines in
the GAP model.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 83

2.11.3.5Step 5 Objective: Add mitigation into the visual workflow

Pipeline instability is a result in loss of momentum of the fluid phases. In an attempt to mitigate
this one approach would be to increase the speed at which the fluids are travelling. This can
be achieved by lowering separator pressure or increasing production by lowering well head
chokes. In both of these cases an assessment would have to be made regarding condensate
banking and its impact: increasing the production rate may result in more liquid droplets no
longer being entrained in the main phase, thus causing a gradual loss in production.
In the remaining part of this exercise, the separator pressure will be lowered. Enter the
PipeStability visual workflow and enter the ‘Action’ variable assignment element.
Using the generic OpenServer object the separator pressure will be changed to 200 psig.

After the GAP separator pressure has been changed, we will want to re-solve the GAP
model. Double-click on Terminator-2 and change the Label to GAP.

© 1990-2020 Petroleum Experts Limited

 84 Flow Assurance

To avoid the workflow from resolving the GAP model in a later timestep once the separator
pressure has been dropped. The IF element will be modified as follows:

Now when the RESOLVE forecast is rerun, the change in separator pressure and resolve of
the GAP model can be seen in the calculation results, as shown below.

2.11.4 Conclusion
The above pipeline logic which detects and mitigates pipeline instability has also assisted the
well stability i.e. increase in production rates.
The plot below shows the change in field production volumes once the separator pressure has
been decreased.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 85

This concludes the pipeline stability example.

2.12 Well Stability

© 1990-2020 Petroleum Experts Limited

 86 Flow Assurance

2.12.1 Introduction
Well stability is described diagrammatically below:

During steady state production, the phases distribute themselves along the velocity profile
such that the fastest flowing phase (gas) is in the middle, and any heavier droplet forming
phases tend to arrange themselves closer to the sub-laminar layer (where friction is higher,
and velocity is lower). This is not really an issue so long as all the phases are moving up the
well towards the surface.
Overtime, as the energy of the reservoir depletes and the WGR increases, all phases travel
slower and the droplets start to collect along the pipe wall (this occurs as the VLP/IPR
intersection approaches the VLP minima). Again, this is not a serious issue as all the phases
are moving in one uniform direction.
At a certain point, the energy is not sufficient to maintain this and some of the liquid droplets
start to travel downwards (this is the onset of liquid loading) and can manifest at low
production rates (using PE5 this is detected by negative frictional gradient)
These droplets may coalesce and aggregate and turn into a thin film (liquid holdups in the
relevant well section less than 0.05 as a rule of thumb i.e these are the order of magnitudes
with which these phenomenon are to be considered). The thin film reduces the friction, which
the fast flowing phase experiences, resulting in a change in the pressure drop profile (in PE2
and PE5)

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 87

2.12.2 Objective
The objective of this exercise is to show how multiple methods (i.e. VLP/IPR intersection,
Turner velocity and PE5 mechanistic model) can be used to detect and in turn mitigate well
instability.

2.12.3 Approach
Predicting the onset of unstable production using PROSPER
When considering the stability of a well three things should be considered chronologically:
1. Where the VLP/IPR intersection occurs on the VLP curve
2. At low flow rates does the empirical Turner criteria predict liquid loading
3. At low flow rates do the mechanistic correlations predict negative friction gradient

2.12.4 Example Steps

© 1990-2020 Petroleum Experts Limited

 88 Flow Assurance

2.12.4.1Prepare the GAP model & Run a forecast

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.5 Well Stability\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Start GAP and go to File | Archive | Extract. Select the completed archive (*.gar) file
from samples folder.
When the "Open Master File?" question is prompted, select "No". This step ensures the
underlying models are extracted into the folder.
Open the GAP model, navigate to Prediction | Run Prediction… from the main
dropdown menu to open the prediction run setup windows.
Set the prediction end date to 01/01/2020 with a 2 month step size.

Click ‘Next’ and ‘Next’ again and ensure that the ‘Delivery’ pressure is set to 655 psig.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 89

As this is a long term forecast, run the GAP forecast with the ‘Rule Based’ network solver.

Click ‘Main’ once the prediction has completed to return the GAP main screen.

2.12.4.2Manual instability Detection

There are a few different ways to detect and view when the well is unstable within GAP; the
mouse flyover, the results table and the VLP/IPR intersection.

Mouse Flyover - Tooltip

The well status is displayed via the Tooltip when the mouse is moved over the respective well
(LM5) (shown below). This shows the status from the last network solve, if a prediction is run,
then this approach can be used by viewing the state of the network with 'Load Prediction
Snapshots'.

© 1990-2020 Petroleum Experts Limited

 90 Flow Assurance

Results Screen
The status result will show if any of the instability indicators have been triggered for both a
solve network and a prediction.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 91

VLP/IPR Intersection Playback

Through the View VLP/IPR option within the Results Screen, the VLP and IPR intersection for
a well can be plotted through time. This means that the well can be visually checked to see
whether it operates in the minimum point of stability region of the VLP curve. A technique often
used as one of the instability indicators.

2.12.4.3Mitigation strategies
There is more than one mitigation strategy possible, and the one selected will depend upon
the system being considered. Examples of possible approaches and how they could be
modelled are outlined below:
Considering a Surface Jet Pump (SJP)
o Providing that a forecast is being done, then consideration of an SJP is valid. In reality
the SJP would need to be designed and the point at which it comes on line determined
through forecasting. These activities require foresight and planning, and as such
implementing an SJP as an afterthought is not advisable.
o The SJP could be used to boost production (by using high pressure wells to assist low
producing wells)
Setup: Introduce an SJP into GAP, and mask it. Setup control logic in RESOLVE to
unmask at either a date (PRESOLVE) or based on an action (POSTSOLVE)
Reservoir Injection/Artificial Lift
o Whenever well stability is determined to occur, there are two possible approaches: (i)
reservoir injection – modifying the IPR or (ii) Artificial lift – modifying the VLP.
o In the retrograde gas condensate context, reservoir injection is usually limited to gas

© 1990-2020 Petroleum Experts Limited

 92 Flow Assurance

recycling to maximize condensate recovery. At this late stage of field life when the
reservoir pressure is well into decline, this is less fruitful.
Setup: Could recycle Gas from the production separator (see standard associated
injection example)
o Artificial lift should only be considered when there is sufficient reservoir deliverability.
Otherwise a scenario where more production is possible from the well, but the
reservoir cannot deliver it may occur.
Setup: Setup artificial lift wells in place of the current naturally flowing wells and
unmask them during the prediction either based on a date or on specific criteria
(instability).
Change tubing string
o Velocity strings are a common well workover performed in the field, as a smaller
tubing string results in faster phase velocities, thus circumventing the well instability.
Setup: Swap VLPs based on either a date or some predefined production criteria.
The impact of each of these strategies can be investigated using GAP or RESOLVE field
management logic and scenario manager. The steps below outline the methodology to first
detect the instability in both GAP and RESOLVE before detailing a step by step to investigate
the SJP strategy.

2.12.4.4Surface Jet Pump Mitigation Approach

2.12.4.4.1 Step 1 Objective: Prepare the GAP model

If this example is performed directly after completing the previous exercise regarding pipeline
stability, the GAP file will already be open. If not, the appropriate file can be accessed by
extracting the saved .gar file. To do this start GAP and go to File | Archive | Extract. Select
the completed .gar file the previous example, outlined in Pipeline Stability 66 .
If it has not already been done, perform a forecast from 20/09/2004 to 01/01/2020 in 2 month
timesteps using the Rule Based network solver (as it is a long prediction), making sure to save
the intermediate step result snapshots at each prediction step. This can be done through
Prediction | Run Prediction… | Next | Next | Next | Settings and enabled through the
option as per the screenshot below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 93

Select OK, then Calculate.

2.12.4.4.2 Step 2 Objective: Identify instability in a well.

As mentioned earlier, there are different methods which can be used to identify an unstable
well in GAP. The first is to view the Status of the well in the prediction results. The
screenshot below shows the prediction results of well LM5 and that it is unstable for the later
part of the prediction.

© 1990-2020 Petroleum Experts Limited

 94 Flow Assurance

Alternatively if a prediction snapshot is reloaded for a later date (30/03/2011 for example) the
status of the well can be ascertained by hovering over the well. To do this go to Prediction |
Reload Prediction Snapshot. The status of the well at that time step can be observed under
Warnings, as below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 95

The VLP/IPR intersection can also be viewed for this time step by returning to the well (LM5)
Summary screen and selecting Calculate | Initialise from solver results | Calculate | Plot. The
plot, shown below, shows the VLP and IPR are only just intersecting.

As outlined earlier there are multiple ways mitigation strategies.

2.12.4.4.3 Tubing string Approach

In this example, the tubing string size for well LM5 is already relatively small (2.06” and
2.441”) and as such reducing the tubing size will provide little benefit given the large time
period where the well is unstable. With this in mind, the SJP approach is investigated and the
steps to set this up are detailed below.

2.12.4.4.4 SJP Approach

Implementing an SJP is a short term solution and the pipelines are too long to make the
implementation of an SJP feasible. Nonetheless a SJP was implemented to illustrate the
short-term nature of this solution. The following steps were performed to do this:

2.12.4.4.4.1 Surface Jet Pump

1. Add an SJP and connect J1 to the high pressure inlet, connect J3 to the low pressure
inlet, and connect the pump outlet to J4, as shown below.

© 1990-2020 Petroleum Experts Limited

 96 Flow Assurance

2. Copy across the pipeline information for SW-Line, MAIN LINE and FromEast into the
appropriate pipelines connected to the SJP. This can be done by holding down Ctrl and
clicking on the pipeline then dragging and dropping it onto the new pipeline.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 97

3. The following input parameters were used for the SJP. For a more detailed discussion
on surface jet pumps, please refer to Section 2.5.11 of the GAP User Manual.
Additionally, a worked example can be found in Example 12 of the GAP User Manual.

4. As the SJP needs to come online at the beginning of the prediction, the Manifold joint
should be masked from the beginning:

© 1990-2020 Petroleum Experts Limited

 98 Flow Assurance

5. At this point the prediction should be run again and the status of well LM5 examined.

The results shown above show that by implementing the SJP from 30/09/2004, LM5 can be
stably produced for a further 6 months. Whilst improving the response of the well, this strategy
is clearly only a short term solution. By modelling the scenario first, the additional benefit can
be weighed against the cost of installing the surface jet pump before making a business
decision.

2.12.4.4.4.2 Artificial Lift

At this stage artificial lift methods may also be considered to periodically unload the well.
Although there are many different artificial lift options in reality, only continuous gas lift is
available to model in a retrograde condensate model in PROSPER. For this particular model,
however, continuous gas lift does not improve the production as the frictional pressure drop in
the tubing dominates. For this reason, this mitigation option will not be considered.

2.12.4.4.4.3 Reservoir Injection in GAP

Applying pressure support to the reservoir is another strategy that can be considered. For
gas condensate reservoirs, gas injection can be performed. The applicability of using gas
injection is dependent on several different factors, including the availability and cost of injection
gas.
To provide an idea of whether gas injection would improve the stability of the well, the Target
Pressure function will be used. This option can be enabled when running a prediction.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 99

1. Go to Prediction | Run Prediction… and enter the target pressure as 1200 psig and
the gas injection fraction as 100%, as shown below. This pressure was selected as it is
approximately the reservoir pressure of the last time step which the SJP was effective in
removing the well instability.

2. Run the prediction with the Rule Based network solver then access the prediction results
of well LM5, as shown below:

© 1990-2020 Petroleum Experts Limited

 100 Flow Assurance

It can be seen that be maintaining the reservoir pressure at 1200 psig the well becomes stable
for the full duration of the prediction. It is clear that increasing the target pressure (in
conjunction with the SJP already in place) has been an effective strategy. Doing this, however,
increases the amount of gas required and has further implications, which are discussed below.
The target pressure strategy can, of course, be investigated without the SJP in place too (as
will be looked at in the next exercise with RESOLVE scenario manager).

2.12.4.4.4.4 Additional Implications to reservoir injection

The produced gas would be used to provide pressure support to the reservoir. This type of
target pressure approach assumes the separator was sufficiently close to the reservoir, or
that an existing injection infrastructure was in place.
Additionally, the cost of all the equipment required for injection would have to be taken into
account. Further to these points, it should be determined if sufficient gas is produced at the
separator to supply enough pressure support to stabilize production in the well. If sufficient
gas is not produced then additional gas would need to be obtained, and the cost of this would
also be considered.
These are the implicit assumptions of any such simplistic voidage replacement, target
pressure analysis.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 101

Save the GAP model in the Completed folder (or a separate working folder) as this GAP
model will be used for the next section.

2.12.4.4.5 Automated Detection and Mitigation using RESOLVE

The status of wells can be checked dynamically during a run by using a Visual Workflow in
RESOLVE. When instability is detected (based on user specified criteria) then controlling logic
can be used to change the response of the system, should the user wish to do this.
The previous section outlined the types of mitigation analysis that can be performed through
GAP by investigating the results of each prediction. The methodology outlined below will show
how the detection process can be automated and the different scenarios (as were
investigated earlier) can be setup to run in a batch. The instability of the well can be plotted
through time.

2.12.4.4.5.1 Step 1 Objective: Open the RESOLVE file

For the following study, open the RESOLVE initial file found in the Well Instability| Initial Files
folder. This is the same RESOLVE file as finished in the previous example, Pipeline Stability 66 .
If opening an archived .rsa file go to File | Archive | Extract.

Going forward, the GAP model that was just setup in the well stability section will be used.
Hence ensure that the GAP instance in RESOLVE is loaded with the correct file through the
'Edit the GAP Case' interface as shown below.

© 1990-2020 Petroleum Experts Limited

 102 Flow Assurance

To perform the scenario study within RESOLVE, the first thing to setup is the automatic
detection of the well instability. The first thing to do is define the variables of interest. After this
has been setup, the scenario manager is used to setup the different strategy cases of
interest, these can then be run sequentially or simultaneously (using clustering).

2.12.4.4.5.2 Step 2 Objective: Import the variables

1. Go to Variables | Import applications variables… | Edit variables. In the OpenServer

variables tab import the gas rate and reservoir pressure for well LM5 as shown below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 103

2. Enter the OpenServer variable strings for the well status which indicate well instability, as
shown below:

© 1990-2020 Petroleum Experts Limited

 104 Flow Assurance

These OpenServer variable stings can be accessed from GAP. In GAP go to the network
solver results for well LM5 and right clicking in the Status results, as shown below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 105

3. In RESOLVE, once all the variables are entered click OK to return to the variable setup
screen. In this screen (shown below) check the box “Add to plot” for each of the
variables.

© 1990-2020 Petroleum Experts Limited

 106 Flow Assurance

2.12.4.4.5.3 Step 3 Objective: Setting up scenarios

The RESOLVE Scenario Manager is designed to be used to investigate different design
strategies. The scenarios can be setup using Event Driven Scheduling or Visual Workflows and
then run as a batch sequentially, as in this example, or simultaneously using PxCluster (for
more details please consult the RESOLVE User manual). The following example sets up the
original scenario (no SJP) for comparison against the second scenario with target pressure
and no surface jet pump.

1. Scenarios| Browse and Edit will bring up the scenario manager. To add a new scenario,
select the Scenarios header again and select 'Add empty scenario'.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 107

2. This scenario will be called 'Original' as it uses the original setup at the start of this
example, i.e. no SJP. The start of run workflow can be used to ensure the setup of the GAP
model is correct. This is setup with two operation elements as shown below.

The mask flags are set as shown below; this will ensure that the Surface Jet Pump is masked.

© 1990-2020 Petroleum Experts Limited

 108 Flow Assurance

The target pressure and injection fraction is also reset to FNA (this will delete the field entry)
and hence no target pressure will be maintained.

3. Now that the first scenario is setup, a second scenario can be added by right clicking the
'Original' scenario and selecting add copy of this scenario.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 109

The new scenario can be titled 'Target Pressure (No SJP)'. The only change that is needed is
in the start of run workflow is that the 'Setting target P and gas fraction' element will now
specify a target pressure of 1200psig and gas fraction of 100 (shown below).

The two scenarios are now setup and can be run as a batch through Run| Run Scenario(s).

© 1990-2020 Petroleum Experts Limited

 110 Flow Assurance

2.12.4.4.5.4 Step 4 Objective: Run forecast and view results

Run the scenarios through Run| Run Scenario(s)

In this example the forecast will be run until 01/01/2020 in 2 month time steps, this is as
default as this has been previously setup for the last example.
The results can then be view in Results | View Scenario Plots.
As an example, the PE5 instability flags for well LM5 are shown for three scenarios are shown
below; it can be seen that without the SJP, the target pressure will have to be increased, i.e.
gas will need to be injected from an earlier period.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 111

When the target pressure is set to 1200psig, it can be seen that the well is indicated as
unstable from 30/01/2009. When the target pressure is set to 1400psig, the well is unstable
for the period that it is being choked back to honour constraints. At later times, once the well
is fully open, the well is not indicated as unstable. When the target pressure is set as
1600psig, the well is not indicated as unstable for the entire duration of the prediction.

Deciding between a strategy to maintain a target pressure 1400psig and 1600psig is an ideal
candidate for a transient simulation study, where once the indicator is shown a transient
simulation can be triggered to determine whether the well will be steady state or transient.
This topic is covered later in the Steady state and Transient Flow Assurance Examples.

2.13 Asphaltene Prediction and Mitigation

2.13.1 Introduction
An asphaltene is a highly viscous and sticky material that precipitates from the oil phase.
There is more than one school of thought regarding the definition of an asphaltene, with no one
theory being clearly better than another. The most common of these theories are listed below:

An asphaltene is a chemical species that at room temperature is insoluble in n-Pentane

and n-heptane, but soluble in Benzene and Toluene – Laboratory definition

© 1990-2020 Petroleum Experts Limited

 112 Flow Assurance

An asphaltene is a solubility class and as such may cover a range of chemical species
(this explains why there is still debate over the definition). – Theoretical definition

An asphaltene is a solid – Theoretical definition

An asphaltene is a highly viscous liquid – Theoretical definition

An asphaltene is a complex mixture of aggregated species coated in a resin that keeps

them in the liquid phase. As soon as the resin is depleted, the asphaltenes aggregate
and settle from the bulk fluid phase. – Theoretical definition based on colloidal theory

The species itself, or the mechanism by which percipitation from the liquid phase occurs is
poorly understood. With the above in mind, it is exceptionally difficult to predict or mitigate
asphaltene occurrence, and yet they pose a very real issue to operators.

The following rules of thumb can be applied to asphaltene’s based upon our limited
understanding of them:

Lab tests of fixed compositions show that asphaltene occurrence tend to manifest close
to the bubble point (this is linked to colloidal theory, and suggests that as the lighter
components such as methane energize to the point where they can escape the liquid
phase, the outer resin is somehow destabilized).

It has also been shown that when an oil has a low C7+ fraction this is indicative of more
paraffinic (alkane) composition – which using the above lab definition – would result in an
increased likelihood of percipitation of asphaltenes. Due to the high paraffinic (alkane)
content, asphaltene and wax occurrences often go hand in hand. The De Boer Plot
(1998) [ref – Phase Equilibria by Pedersen] confirms this.

2.13.2 Modelling Objective

Despite the limited understanding of asphaltenes, a practical approach is required to detect
their formation. One approach could be to define an asphaltene envelope, and should

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 113

production conditions enter this then the user will be informed (or perhaps a mitigation action
from the field can be defined). The following example will look at how this process can be
automated within RESOLVE, this approach is scalable in terms of both investigating different
scenarios and also designing automatic mitigation.

2.13.3 Detection
With the above in mind, the detection method would be to create an operating envelope
(centered on the phase envelope) and detect if any element in the production system would
enter this area.

The idea would be to do this in a steady state forecast to understand when this would pose a
problem, and if this were an issue for the entire forecast, then there would be suitable
justification to consider mitigation techniques.

2.13.3.1Step 1: Setup the GAP File

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.8 Asphaltene Prediction and Mitigation\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder.

Check Options | Methods in GAP, and ensure that the “wax and hydrate warnings” are
enabled.

Solve the network.

© 1990-2020 Petroleum Experts Limited

 114 Flow Assurance

As such, Save and Close GAP as this will now be opened from RESOLVE.

2.13.3.2Step 2: Setup the RESOLVE file

Opening a new RESOLVE file, create the following Objects (Workflow Element, Data
Set object, Generic OpenServer object, EOS-PVT data object and the GAP instance)

Associate the GAP file closed in step 1 to this instance.

N.B. Additional nodes can be added by double clicking on the GAP instance and adding an
extra node in the Source/sinks tab. This process is outlined in more detail in Step 4: Workflow
Creation 117 . Step 3:

2.13.3.3Step 3: EOS-PVT (Data Object)

The GAP file being used is fully compositional. The analysis will be performed for the MAIN
LINE in the GAP model.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 115

1. To expose the data at the manifold in RESOLVE, double click the GAP instance, and go
the sources and sink Tab. Select manifold (see below)

© 1990-2020 Petroleum Experts Limited

 116 Flow Assurance

2. This will expose an additional output from GAP. It is this manifold element that should be
connected to the EOS_PVT object, as below:

Every time GAP is solved, it will automatically pass the composition from the manifold to the
data object.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 117

2.13.3.4Step 4: Workflow Creation

1. Create the workflow below:

2. Define the following variables:

© 1990-2020 Petroleum Experts Limited

 118 Flow Assurance

2.13.3.4.1 Define the Operating Conditions

1. Enter the following:

2.13.3.4.2 Calculate the Saturation Pressure for the Inlet and Outlet Conditions
The next step is to define a loop: this loop will cycle through the temperatures specified in the
first assignment box, and at each step calculate the phase envelope pressure.

2.13.3.4.3 Define the Conditions for Asphaltene Appearance

The asphaltene’s appearance window limits are somewhat arbitrarily decided: if guidance from
lab data is available then this should be used. In this example 100 psig above and below the
phase envelope are used.

1. In the IF function, define the following range for asphaltene appearance:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 119

If the operating conditions are within the asphaltene window at either the inlet or the outlet of
the pipeline, then this should be reported to the user, otherwise the run should continue:

2. In “Report to User” input the following:

If the manifold/joint pressure and temperature fall within this defined envelope, the following

© 1990-2020 Petroleum Experts Limited

 120 Flow Assurance

message can be seen in the message log window:

2.13.3.5Run the Prediction

The workflow is now ready for the prediction to be run. Before running the prediction:

1. Go to System | System options… and ensure the system options are as follows:

2. Go to Variables | Import application variables… and import the following workflow

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 121

variables and ensure they are also selected for plotting:

3. Go to Schedule and set the following:

© 1990-2020 Petroleum Experts Limited

 122 Flow Assurance

4. Run the prediction.

2.13.3.6Analysis of Results
It can be seen from the Calculation log that no timesteps show the warning for asphaltene
appearance. This is because the inlet and outlet conditions are significantly far away from the
calculated saturation pressures for the composition of that time step. As an example, the inlet
pressure and saturation pressure calculated for the inlet temperature is shown below:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 123

2.13.4 Mitigation
Mitigation techniques usually center on surface active agents: anything that can stop the
asphaltene grouping together and settling are used. This can involve drag reducers and anti-
agglomerants. Any field specific mitigation technique could be incorporated in a similar was to
those shown in previous examples.

2.14 Hydrates Prediction and Mitigation

2.14.1 Introduction
Hydrates are solid crystalline structures that at their center trap/host a gas molecule: the
crystalline structures are made from water molecules and thus exhibit hydrogen bonding. The
gas molecule is attracted by weak Van-Der-Waals forces to the water molecules. Hydrate
types are often classified according to the gas molecule/cavity size (i.e. type I, II or H). The

© 1990-2020 Petroleum Experts Limited

 124 Flow Assurance

general theory for the formation of a hydrate is considered in a two-step process when
calculating their occurrence. Step 1 is the formation of the empty lattice, while step 2 is the
gas molecule entering and stabilizing the lattice (i.e. adsorption of the gas molecule). There
has been subsequent research by Munck et al. to consider the changes in step 2 with
temperature; as this effects the rate that the gas molecule enters the lattice. This is discussed
in the PVTp manual in greater depth in Section 2.2.4.1. The Hydrafact models use the above
fundamental theory in combination with laboratory and field data to create proprietary models
that can predict the formation of hydrates with greater accuracy. Below are the hydrate
models available to predict the onset of hydrate formation.

2.14.2 General Approach to Hydrate Analysis

To date, most hydrate based simulation work has centered solely in the transient domain. For
example, when a pipeline is shut in, the temperature drops as the pipeline equilibrates with the
external temperatures and thus starts to enter the hydrate region. The hydrate region is
dominated by its temperature dependence (i.e. above specific temperatures step 1 and 2
happen in negligible amounts) and so the time for temperature loss usually indicates how long
the pipeline can remain shut in without issue. The below plot shows the hydrate formation
according to different models:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 125

2.14.3 Recommended approach to Hydrate Analysis

The approach adopted to date is reactionary and thus tries to solve a problem after it has
manifested. The recommended approach, however, is two phased, as outlined below:

Phase 1: During Operations

o Proactively ensure that the hydrate envelope is not entered

o If the hydrate curve is entered, then two approaches are usually considered (i)

chemical or (ii) Thermal mitigation

Phase 2: During Shut in

o Avoid Formation of hydrates by performing temperature modelling of the pipelines

o If Hydrates form – the removal of the hydrate obstruction uses a combination thermal,
chemical, and operational strategies to minimize the hydrate formation.

All of Phase 1 in the first instance should be modelled in steady state tools, and most of phase
2 should be modelled with transient tools. The below example outlines how to model Phase 1

© 1990-2020 Petroleum Experts Limited

 126 Flow Assurance

in terms of phenomena detection/mitigation on day 1 of production and over time as the field is
produced (as pressures drop, temperatures drop and water cut increases).

The overall steps are:

Detection

o Hydrate Check (Operating Envelope)

Mitigation

o Methanol Injection – RESOLVE

o Heating of line – Temperature profile of pipeline

2.14.4 Example Steps

2.14.4.1Manual Detection and Mitigation

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.7. Hydrate Prediction and Mitigation\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder

Move the mouse over any of the pink pipelines and a Hydrate Risk warning message can
be seen.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 127

Double click on the MAIN LINE pipeline and in the solve network results, scroll along to
view composition.

Select Phase Envelope | Calculate Hydrate… Wax | Plot

The green line, shown in the phase envelope below, indicate where the risk of hydrate
formation occurs.

© 1990-2020 Petroleum Experts Limited

 128 Flow Assurance

Run the forecast (Prediction| Run Prediction) ensuring the separator pressure is at
655 psig and the Rule Based network solver is used, as per the screenshot below:

Once complete, select the pipeline “LM2ToLM1” and navigate to the status results (as
shown below):

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 129

This step has shown how to manually detect the risk of hydrate formation, and that it is
relayed to the user from the upper level of the GUI interface (i.e. Pink pipelines, and mouse
overs).

Mitigation strategies often used include:

o More flow (less well head chokes) – using resolve to implement logic.

o Heated lines (very expensive) – changing the geothermal gradient.

o Chemicals – such as wax crystal modifiers, which keep the wax crystals from
agglomerating together and /or surface active agents like detergents and dispersants,
which are usually modelled as a P2 modifier.

© 1990-2020 Petroleum Experts Limited

 130 Flow Assurance

o Accept its going to occur, but pig the lines. This is usually the case, as hydrate
formation is often not considered in the correct phase of the design, i.e. we have to
live with it. Using GAP and RESOLVE, this is what the next section will show.

2.14.4.2Automated Detection and Mitigation

The focus of this exercise is to estimate the time and percentage of MEG required to
dose a flowline.

The objective of this example is to implement a workflow in RESOLVE so that we can

automate when MEG is required.

The criteria will be whenever GAP detects hydrates according to the hydrate curve
(previously shown), MEG will be added incrementally.

2.14.4.3Step by step
1. Open RESOLVE and add the following data objects:

2. In GAP add an injection element (fluid type water and temperature 100°F) and joint and

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 131

copy the existing pipe (MAIN LINE) to the new pipe (MAIN LINE_MEG) as follows. Ensure to
mask/disable the original MAIN LINE so that it will not be used.

N.B. a fixed injection rate can be entered in order to make the injection element valid, but this
will be a calculated variable and subsequently implemented by RESOLVE.

3. In the GAP object in RESOLVE add an extra node at the manifold

© 1990-2020 Petroleum Experts Limited

 132 Flow Assurance

We will then connect the source node to the base composition. This will auto pass the
composition at every timestep.

5. Go to the workflow and using the pallet create the following:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 133

After creating the above, define the following variables, which will be used:

© 1990-2020 Petroleum Experts Limited

 134 Flow Assurance

6. In the assignment reset the hydrate flag (required to move onto next timestep).

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 135

7. In the operation retrieve the hydrate flag in GAP pipe.

8. This flag can be found in the following GAP pipe.

© 1990-2020 Petroleum Experts Limited

 136 Flow Assurance

9. The if statement will check if the hydrate flag is 1 or 0. 1 means hydrates present and go
to sub flowsheet, 0 means no hydrates and continue to next step.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 137

Sub flowsheet 1

This flowsheet calculates the required MEG to add to the composition in RESOLVE.

10. In the operation we will get the pressure and temperature downstream of the main trunk-

© 1990-2020 Petroleum Experts Limited

 138 Flow Assurance

line (above).

11. In the assignment we will setup the hydrate calculator

The operation will calculate the hydrate curve by blending the water composition and base
composition together.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 139

12. The assignment will check if the hydrate flag is still 1 and retrieve the new MEG% added
in the water composition

13. The if statement will check if the hydrate flag is 1.

© 1990-2020 Petroleum Experts Limited

 140 Flow Assurance

14. The operation will add the inhibitor, MEG in this case, and incrementally increase the
added MEG by 0.25% until the hydrate flag disappears.

Sub flowsheet 2

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 141

This flowsheet will calculate based on a calculation what volume of MEG is required to be
added in GAP assuming the density of the MEG is 69.49lb/ft3. MEG will then be added to the
GAP model to calculate the dP in the pipe as a water since MEG in reality is primarily a water
based solution with alcohol.

15. The operation will retrieve the water rate from the injection element in GAP

© 1990-2020 Petroleum Experts Limited

 142 Flow Assurance

16. The mass of water and therefore mass of MEG required will be calculated as follows

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 143

17. The volume of MEG in STB/d is therefore Mass of MEG/Density of MEG, as below:

18. Finally the operation will set the calculated volume of MEG back into GAP.

19. The MEG% and volume of water injected can be added for plotting under Variables |
Import application variables... | Workflow:

© 1990-2020 Petroleum Experts Limited

 144 Flow Assurance

2.14.4.4Run the Model and View Results

To run the model, first a Schedule must be set. This can be accessed from the Schedule |

Forecast Data tab.

1. Enter the schedule as below and press OK:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 145

2. Run the model, and view the results for Vmeg and MEG%.

We can see that at 3 distinct points in time when MEG is required to prevent the formation of
hydrates in the pipe, each period requiring a different amount of MEG. The drop in volume of
Meg required is due to the drop in water being produced from the network.

As the incremental change in MEG mole % is 0.25, the three doses required are: 12, 12.25,
and 12.5 respectively. If a smaller incremental change is used, the frequency of dose
changes may increase, but over all it may prevent overestimation of MEG requirements. This
would be particularly important if an economical study was performed.

2.15 Wax Prediction and Mitigation

© 1990-2020 Petroleum Experts Limited

 146 Flow Assurance

2.15.1 Introduction
Waxes are multicomponent solid mixtures that precipitate/form based upon a complex
chemical and thermodynamic equilibrium that is not very well understood. By nature they are
usually the straight chain alkanes that precipitate out, and in multi-component oil samples the
wax formers usually range between C20-C50 (an indicative marker is a low C7+ fraction
density, indicating more alkanes).

In the absence of an in depth understanding of the forming mechanisms, work in this domain
since the 1980s has concentrated on making simplifying assumptions (using Ideal EOS
assumptions) combined with empirical correlations to predict which species and the
proportions of those species required to form waxes.

In PVTp, the solids calculation engine allows more than one method to describe the formation
of waxes based upon a composition. The first models used were those proposed by Won in
1985/1986 while the most recent (and recommended model) is the Pedersen wax model, as
this is generated and calibrated against live field samples. These models are outlined in the
Section 2.2.3.1 of the PVTp User Manual, and a list of further references can be found in
Section 2.2.3.2. As such in GAP the default hard coded wax model used is the Pedersen Wax
model.

The above summary of waxing models is true regardless of whether steady state or dynamic
modelling is performed (full reference list available in PVTp manual).

2.15.2 Recommended Approach for analysis:

As in most flow assurance studies, the first step is to investigate and detect whether the issue
could potentially occur through the lifetime of the field. Following this, various mitigation design

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 147

strategies can be investigated and modeled (in both steady state and transient studies) before
a strategy is finalised and implemented in the field.

Detection
Given that we understand that wax may drop out when the oil goes below the cloud point.
Possible wax deposition can be detected on this basis i.e. fluid temperature going below the
Wax Appearance Temperature. This detection can be performed within PROSPER, GAP and
the Wax Data Object within RESOLVE. To automate the detection of wax and then act upon
its detection, RESOLVE can be used (as will be shown in the design study below).

Mitigation
Once detected, actions can be performed to reduce/eliminate the deposition of wax. These
strategies fall under three main categories; thermal, chemical and mechanical. The example
below focuses on designing on a thermal basis, note that to perform the wax elimination
strategy on a chemical basis, a similar methodology as outlined in the Hydrates section could
be used.

2.15.3 Modelling Objective

The focus of this exercise is to design a system in order to reduce the risk of wax deposition.
As a first stage, a GAP prediction is performed to assess whether there is a risk of wax
during the lifetime of producing from the field.

After analysing the results, an assessment can be made on whether mitigation must be put in
place. In order to test the system with differing mitigation strategies, RESOLVE will be used
to automatically detect whether Wax has occurred and use Visual Workflows to control the

© 1990-2020 Petroleum Experts Limited

 148 Flow Assurance

model accordingly.

2.15.4 Example Step by Step

2.15.4.1Step 1: Setup the GAP File

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\2.6 Wax Prediction and Mitigation\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder

Check Options | Methods in GAP, and ensure that the “wax and hydrate warnings” are
enabled.

Solve the network

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 149

2.15.4.2Step 2: Detection of Wax

Mouse over any of the pink pipelines and see wax message.

Double click on the MAIN LINE pipeline and in the solve network results, scroll along to
view composition.

Select Phase Envelope | Calculate Hydrate… Wax | Plot

The red line, shown in the phase envelope below, indicate where the risk of wax
formation occurs.

© 1990-2020 Petroleum Experts Limited

 150 Flow Assurance

This step has shown how to manually detect waxing, and that it is relayed to the user from the
upper level of the GUI interface (i.e. Pink pipelines, and mouse overs).

2.15.4.3Step 3: Mitigation Approaches

The faster a fluid flows, the less time it has to lose heat and thus enter the wax
envelope. As such one mitigation strategy is to maximize production and minimize
unnecessary choking of the flow.

o This would be implemented using a visual workflow that would guide the GAP
algorithm (Rule based or optimizer) to apply viable solutions – i.e. choking only certain
parts of the system in specific increments.

If waxing is predicted from evaluation of a characterized EOS, then mitigation strategies

can be considered such as: pipeline insulation, pipeline heaters, etc. In severe cases
heat traced lines can also be investigated, and these are usually employed when flow is
stopped to keep the fluid temperature above the Wax appearance temperature.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 151

Chemical inhibition (wax crystal modifiers – to keep the wax crystals from agglomerating
together) with surface active agents like detergents and dispersants.

o This is usually modelled using a low friction multiplier value alongside a reduced
viscosity table for fluid properties.

One mitigation approach is to handle the wax purely from an operational standpoint: i.e.
accept that wax is going to occur and pig the lines that can be done so long as the initial
field development design basis included pipeline pigging capabilities.

o This can be captured using GAP and RESOLVE, and as such, this is what the next
section will show.

2.15.4.4Step 4: REOLVE setup

Open RESOLVE and setup the following workflow (i.e. put all the application on the
RESOLVE work canvas):

© 1990-2020 Petroleum Experts Limited

 152 Flow Assurance

The Excel file will populate the date when a pig is required. It will also show the number of
days from the last pig to determine the frequency and which well has resulted in a pig being
required.

2.15.4.5Overall Workflow
1. Go to the workflow and using the pallet create the following:

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 153

2. Define the variables to be used in this workflow as below:

2.15.4.6Building the workflow

© 1990-2020 Petroleum Experts Limited

 154 Flow Assurance

2.15.4.6.1 Step 1: Well Count

Set well count as an integer. This will count the number of well in GAP.

2.15.4.6.2 Step 2: Set Loop

The loop will loop through the wells one by one. Set ‘well’ as an integer.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 155

2.15.4.6.3 Step 3: Operation

1. Each operation will use the generic OpenServer function as shown below.

In case the access location for the OpenServer variables used in the above workflow are not
known, the following screenshots show where they can be found:

To access and enable the IPR dP shift for one of the wells, go to the well in GAP, select Input
| IPR | Ipr Layer and tick the box for IPR dP shift and enter zero, as shown below. Ensure to
do this for each every well. The OpenServer variable can be accessed by holding down Ctrl
and right clicking on the IPR dP shift value box.

© 1990-2020 Petroleum Experts Limited

 156 Flow Assurance

Do the same for the gas rate from the solve network results.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 157

The well label can be retrieved from the well Summary screen as follows:

© 1990-2020 Petroleum Experts Limited

 158 Flow Assurance

2.15.4.6.4 Step 4: If Statement

In the IF statement set if Qgas is less than 5 MMscf/d.

2.15.4.6.5 Step 5: Field Model response

In the No assignment for the relevant well we will incrementally increase the dP to simulate the
wax buildup

2.15.4.6.6 Step 6: line cleaning required?

In the yes assignment we will flag a pig is required for the relevant well. This will reset the
well dP and write the date and well name to excel.

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 159

2.15.4.6.7 Step 7: increase the dP of the well

We will set the previously calculated dP in the indexed well

2.15.4.6.8 Step 8
Once all 3 wells have been looped, we will go to a terminator which will move to the next
timestep and repeat the procedure.

© 1990-2020 Petroleum Experts Limited

 160 Flow Assurance

2.15.4.6.9 Step 9
The model can now be run.

1. Ensure the following schedule has been set:

2. Import the following application variables form Variables | Import application

variables…

Flow Assurance November, 2020

 Steady state Flow Assurance Examples 161

3. Run prediction.

2.15.4.7Analysis of Results
Reviewing the results we can see that the first pipe pigging is required on 30/07/2006 some
668 days after the start of the prediction. This is due to LM5 dropping below the 5 MMscf/d
gas rate, as such pipes connected to LM5 should be pigged. The dP is then reset to improve
the productivity but subsequent timesteps result in pigging being required every few months

© 1990-2020 Petroleum Experts Limited

 162 Flow Assurance

due to low productivity in LM5.

If we review the gas rates for each well (shown below) all three wells drop below the 5
MMscf/d minimum rate. LM5 we can see that after the pig the rate increases again, due to
resetting the dP shift value to zero. However, the rate drops again to the point where it is
always below the correlated 5 MMscf/d.

Well LM1 and LM2 also drop below the minimum rate, but at much later times. When this
happens the lines connected to LM1 and LM2 should also be pigged.

Flow Assurance November, 2020

Steady state Flow Assurance Examples 163

© 1990-2020 Petroleum Experts Limited

Chapter

3
 Logie Mill Production Field (Full Model) 165

3 Logie Mill Production Field (Full Model)

3.1 Combining the Previous Examples

3.1.1 Objectives
Until now, each flow assurance examples have been completed in isolation. Each detection
approach has been applied to the base model, and where appropriate, mitigation actions have
been performed. There are multiple flow assurance issues associated with this field that need
to be addressed in the design stages. It is unrealistic to consider each issue separately, as
the mitigation action for one issue may have a direct effect on the detection, and therefore
mitigation action, of another issue.

As such this section will look at how to integrate all of the individual RESOLVE workflows into
one master RESOLVE file. This file will perform each of the previously created workflows in
series at every timestep of the prediction, thus giving an accurate view of:

1. The design tolerances required.

2. The mitigation actions required and the scheduling of these actions.

3. The overall performance of the field.

From this point on only steady state modelling will be considered. Although transient modelling
is very important and should be considered, the application of transient tools is dependent on
the relevant licenses being available.

3.1.2 The Full Model using a Steady State Approach Only

The overall workflow that uses the steady state tools only will be created in a new RESOLVE
file and will take the following form:

© 1990-2020 Petroleum Experts Limited

 166 Flow Assurance

The following steps will outline how to create this model.

3.1.2.1 Create the Full Model Template

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\4.0 Full Model\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder. Once extraction is complete, close GAP.

1. Open a new instance of RESOLVE and populate the Graphical canvas with the object
shown above. This can be done by using the icons shown below:

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 167

2. Link the Excel object to the Wax.xlsx file, which can be found in '~\Petroleum Experts
\IPM 12.5\Samples\FlowAssurance\4.0 Full Model\Initial Files\'

3. Link the most recent GAP file for the system. This can be found in '~\Petroleum Experts
\IPM 10\Samples\FlowAssurance\4.0 Full Model\Initial Files\'

4. Within the GAP Model, made the changes highlighted below and save. These will ensure
that the previously made workflows will work properly without modifications being
required.

5. Ensure that the Rule Based solver option is selected:

© 1990-2020 Petroleum Experts Limited

 168 Flow Assurance

6. Ensure that the Manifold joint is exposed in RESOLVE by double clicking on the GAP
instance and following the screenshot below. Then create the link between the manifold
and the BaseComposition object:

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 169

7. Link all the objects up as per the above Full Model

3.1.2.2 Data Store Variables

The Corrosion data store (“CorrosionDataStore”) requires the following variables to be
defined. These can be accessed by double clicking on the data store object.

© 1990-2020 Petroleum Experts Limited

 170 Flow Assurance

3.1.2.3 Importing Workflows

Within each workflow object it is essential to import the appropriate workflow file (extension
*.vwk). If the previous examples have been completed it is possible the save an existing
workflow to be imported. Alternatively all the necessary workflows can be found in '~
\Petroleum Experts\IPM 12.5\Samples\FlowAssurance\4.0 Full Model\Initial Files\Workflows'.

The following steps will outline how to save and import a workflow.

3.1.2.3.1 How to Save a Workflow

If a workflow has been created in a previous example, it can be saved within the visual
workflow window by selecting the red save symbol:

This process should be repeated for all workflows.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 171

3.1.2.3.2 How to Import a Workflow

The following example shows the import of the Corrosion.vwk file into the “Corrosion”
workflow object.

1. Double click on the Corrosion workflow object to open the visual workflow window then
follow the screenshot below and link the appropriate *.vwk file:

2. When the Variable Mapping screen appears, select “OK”.

If there are unmapped variables, it may be necessary to “Purge unused variables”.

© 1990-2020 Petroleum Experts Limited

 172 Flow Assurance

3. Repeat this process until all workflows have been successfully imported.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 173

3.1.2.4 Defining the System Options

Within Options | System options… the following options should be selected:

3.1.2.5 Importing Variables

So that the results can be plotted and analysed, it is important to import the appropriate
variables. Any desired variables can be imported, but those which will be import in this
example are outlines below.

1. Go to Variables | Import application variables… | GAP | Edit variables and import

the following GAP variables:

© 1990-2020 Petroleum Experts Limited

 174 Flow Assurance

2. Select “OK” and select all the imported variables to be plotted:

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 175

3. Go to the Erosion tab and import and plot the following:

© 1990-2020 Petroleum Experts Limited

 176 Flow Assurance

4. In the Corrosion tab, import all of the variables and add them all to plot:

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 177

5. In the Wax tab, import and add the following variable to plot:

© 1990-2020 Petroleum Experts Limited

 178 Flow Assurance

6. In the Asphaltene tab, import and add the following variables to plot:

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 179

7. In the Asphaltene tab, import and add the following variables to plot:

© 1990-2020 Petroleum Experts Limited

 180 Flow Assurance

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 181

3.1.2.6 Define the Schedule

Select Schedule and enter the following forecast schedule:

3.1.2.7 Run the Model

The model is now ready to run. As with all the other examples, the model will be run using the
Rule Based Network solver. This is the recommended solver to use when performing long
term predictions, such as this.

As there are multiple workflows within this RESOLVE model it is important to keep in mind that
it may take some time to complete.

3.1.3 Analysis of Results using the Steady State Approach

The objective of combining all of the individual workflows is to see the effect of each mitigation
action on another. As such the result shown in this section will vary from those of the previous

© 1990-2020 Petroleum Experts Limited

 182 Flow Assurance

examples.

This section will not discuss the meaning of results in detail, as this is covered in the respective
examples, but will instead highlight the effects of different mitigation actions on the detection
and implementation of mitigation actions for other phenomena. For example, a change in
separator pressure to reduce pipeline instability will affect the flowrate through the pipelines,
and as such, affect the detection and mitigation of wax formation.

3.1.3.1 Erosion Results

As was previously seen in the specific erosion example, the amount of pipeline erosion caused
by sand production is minimal. In the Full Model, however, a new separator pressure is set on
30/03/2014. This changes the separator pressure form 655 psig to 200 psig. This can be
seen in the erosion results in the large spike in erosion rate and subsequent increase in the
total erosion amount, due to increased production caused by the reduced separator pressure.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 183

These results, although showing an increased total erosion over the period of the prediction,
are still insignificant. As such, no mitigating action was implemented.

3.1.3.2 Corrosion Results

The corrosion results have also been affected by the change in separator pressure (on
30/03/2014), as can be seen below:

The corrosion rate starts high, but falls as production drops off plateau. The gas rate at the
separator (Delivery) increases on 30/03/2014 (when the separator pressure decreases to 200
psig) as expected, but the corrosion rate decreases significantly on this date. This is because
more gas is now produced from LM5 which has a lower CO2 content than the other wells.

© 1990-2020 Petroleum Experts Limited

 184 Flow Assurance

3.1.3.3 Pipeline Stability

As previously mentioned, the separator pressure is changed in the 30/03/2014. This is due to
the detection of instability in the FromEast pipe line. As this pipe line was arbitrarily selected
for monitoring, it is possible that PE5 instability may have been detected earlier or later if
another pipeline was selected. A more thorough approach would be to monitor every pipeline,
and perform the mitigation action accordingly. However, in this example, only one pipeline was
monitored simply to highlight the functionality of the PE5 instability flag mechanism.

The screenshot below shows the message identifying the timestep when the PE5 flag was
detected and when the separator pressure was changed as a result.

In the following plot, it can be seen that the gas rate at the separator and from all of the wells
increases after the separator pressure has been decreased.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 185

3.1.3.4 Well Stability

Despite the drop in separator pressure which increases the gas rate produced by well LM5,
the increase in rate is not sufficient for the well to produce above the Turner Velocity.

Additionally, well LM5 has PE5 instability flags for the entire duration of the prediction. This
further indicates that some form of mitigation action is required. The most likely candidate for
this field (as discussed in the Mitigation strategies 91 section) is to use a form of pressure
support. However, it would be required to support the reservoir at its original pressure for
there to be any impact on well stability. As an economic study would be essential to
determine the feasibility of this; no mitigation actions have been performed in this example.

3.1.3.5 Wax Formation

Due to the composition of the produced fluid, length of the pipelines, and surrounding
temperatures used in this example, the formation of wax is inevitable. As discussed
previously, wax mitigation approaches 150 can include:

© 1990-2020 Petroleum Experts Limited

 186 Flow Assurance

Field management techniques to increase fluid thereby reducing heat loss.

Pipe line heating/lagging.

Chemical inhibition, to either suppress the cloud point or prevent agglomeration.

Implement a pigging schedule.

In this example it was chosen to implement a pigging schedule. This is ultimately saying that it
is excepted that wax formation may be a problem in this system, and that it will be dealt with
when it significantly affects production (using a field correlation). As can be seen below, the
minimum gas rate for each well has been set to 5 MMscf/day. When the rate drops below
this minimum, the pipelines downstream of that well will be pigged, thereby allowing my flow.
Due to the implementation of the other workflows, the schedule has changed, whereby LM2
does not drop below the minimum rate. This is caused by the separator pressure change
which has been highlighted on the plot below.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 187

Of course this is just one possible method. Each of the above options has the potential to be
very expensive. As a detailed economic study has not been performed, it is unclear which of
the above options would be the most economically viable. As such, no other method will
investigated in this example.

3.1.3.6 Hydrate Formation

As can be seen in the results plot below, until the separator pressure change takes affect on
30/03/2014, three different doses of MEG are required to prevent hydrate formation within the
MAIN LINE pipeline. However, after the separator pressure change the amount of MEG
required drops to zero. This is because the pressures and temperatures within the pipeline
(for the current composition) are such that hydrates do not form.

This is a prime example of why it is essential to run the full model with all detection and
mitigation methods active. Without properly investigating the implications of the changing the
separator pressure, for example, then it would have resulted in an over estimation of the
amount of MEG required for the duration of the prediction. Having evaluated the detection and
mitigation methods together, we are able to reduce the overall cost of MEG as less is
required.

© 1990-2020 Petroleum Experts Limited

 188 Flow Assurance

3.1.3.7 Asphaltene Formation

During the 16 year prediction, the flag for possible asphaltene formation in the MAIN LINE is
not observed. This means that the composition at the Manifold (where the differing
compositions from the different reservoirs met) generates a phase envelope which is
significantly far away from the operating conditions of the MAIN LINE pipeline.

The workflow used to detect the possible formation of asphaltenes has only been implemented
for the MAIN LINE pipeline. For a more thorough check this workflow could be extended to
monitor all the pipelines for the possible formation of asphaltenes.

3.2 Conclusions/Summary
In summary, this chapter has looked at:

1. How to create the full model with multiple workflow to detect and mitigate against
different flow assurance phenomena. This was done by importing previously created
workflows.

2. The effect of running each detection and mitigation technique simultaneously.

Each of the workflows deals with a separate method of detect and mitigation.

Erosion

Erosion can, in cases where significant amounts of sand are being produced, be a very
serious issue, and should be accounted for within the design phase. In the particular example,
however, sand production is not a significant issue and as such the erosional rates expected
are negligible. As a result of this no mitigation actions were taken.

Corrosion

In this example the CO2 corrosion allowance was evaluated for a pipeline. In this particular

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 189

case the amount of CO2 present in the produced fluid is relatively low, and as such the

corrosion rate is also low. Over time, as the field is in decline, the corrosion rate also
decreases.

Separator Stability

In this example, the flow assurance button within the prediction results was utilised to access
the Slug Catcher Calculator. This allowed an evaluation of the size of the separator to be
performed. As a result of this analysis a 48" separator with an L/D ratio of 0.625 was
selected. This size of separator was sufficiently large to cope with the surge volumes
experienced during operation.

Pipeline Stability

In this example, the FromEast pipeline was monitored for the PE5 stability flag. When this flag
was detected at any point in the pipeline then the mitigation action was to change the
separator pressure. This change in pressure caused the rate through that particular pipe (and
others) to increase, thereby removing the PE5 stability flag as stable flow was achieved.

Well Stability

The well stability example purely looked at manual detection and mitigation methods for well
LM5. Of the mitigation options that are available, the implementation of a surface jet pump
was investigated and ultimately disregarded as a viable option as the SJP was only effective
for the first 2 years of a 16 year prediction. When the SJP was used in conjunction with
reservoir injection the duration of stable flow was increased to approximately 9 years. As
further studies into the feasibility of reservoir injection were not performed, no mitigation
method was applied.

Wax

In this example the MAIN LINE a pigging schedule was decided as the course of action that

© 1990-2020 Petroleum Experts Limited

 190 Flow Assurance

would be taken. A field correlation was used to determine the frequency of pigs required, and
where in the system these pigs should be applied.

Hydrate

To mitigate against hydrate formation in the MAIN LINE pipeline, the injection of MEG was
considered. The workflow created utilised the Hydrate and Water Composition objects in
resolve to determine the mole % of MEG required to remove the risk of hydrate formation.
Over the course of the 16 year prediction 3 different dosages of MEG were required due to
the changing composition of the produced fluid.

Asphaltene

Asphaltene formation occurs when conditions fall within ± 100 psi of the saturation pressure
for a particular temperature. In this particular example, the MAIN LINE was monitored for the
presence of asphaltenes, but during the 16 year prediction none were detected.

Combining the separate workflows into a master RESOLVE file caused certain mitigation
actions to be delayed in time or become unnecessary. This is because mitigation actions
designed to account for a particular flow assurance phenomenon are affecting the detection of
another.

The main contributing factor to the observed results was the change in separator pressure
from 655 psig to 200 psig. This change was implemented in order to mitigate against pipeline
instability (within the FromEast pipeline). Almost all of the other detection and mitigation
techniques were affected by this change. For example:

Although the total erosion within the MAIN LINE is negligibly small, the reduction in
separator pressure causes a sharp increase in the erosion rate due to the increase fluid
velocities.

Flow Assurance November, 2020

 Logie Mill Production Field (Full Model) 191

Changing the separator pressure causes the rate of corrosion within the MAIN LINE to
reduce significantly, as more gas is allowed to flow from wells with a lower CO2 content.

In terms of wax formation, although the separator pressure change does not directly
affect the formation of the wax, it does have an effect on the pigging schedule.

When the separator pressure changed, the risk of forming hydrates was eliminated
within the MAIN LINE, and MEG injection was no longer required.

In this particular example, many of the techniques have focused on one particular element or
part of the system. For example, the MAIN LINE pipeline was targeted for hydrate formation,
erosion issues, and corrosion issues. In reality, however, these techniques would be applied
to the whole system, i.e. every pipeline and well. Since this example is purely to demonstrate
the flow assurance functionality within the tools, this was not necessary.

To conclude, it is essential that all flow assurance detection and mitigation techniques be
studied together in an integrated model. This will help to identify possible interactions between
methods and identify unforeseen responses. It would also be essential the apply these flow
assurance detection and mitigation techniques to all levels/aspects of the field.

© 1990-2020 Petroleum Experts Limited

Chapter

4
 Water Chemistry: Reservoir and Wellbore Scale 193

4 Water Chemistry: Reservoir and Wellbore Scale

4.1 Introduction
In many production environments such as the North Sea, production and operational issues
can arise from scale deposition in the reservoir itself through to the process facilities. Oilfield
scales typically consist of one or more types of inorganic deposit along with other matter.

Scale precipitation can occur due to any of the following reasons:

Auto-scaling: Changes in temperature can cause minerals to drop out of the formation
water. This is due to the fact that the solubilities of the minerals are functions of
temperature.

Incompatible waters: When waters with different chemical compositions mix, reactions can
occur which cause the formation of scale minerals. For example, injecting sea water
(containing sulphate ions) into reservoir water with a different composition (e.g. containing
Barium ions) can cause scale formation (Barite, i.e. Barium Sulphate) due to the mixing of
waters. Mixing of incompatible waters can also occur during production from multilayer
reservoirs with different water compositions, or in the surface network when mixing
production waters from different reservoirs.

Drying: Injecting a very dry gas can cause the reservoir water to be vaporised into the gas
phase. This increases the concentrations of the ions in the aqueous phase, which drives
scale formation. A typical example of this is scale formation at gas lift valves, where some of
the formation water vapourises into the gas phase causing scale dropout.

Wells producing water are likely to develop deposits of inorganic scales that will cause a
variety of flow assurance issues such as:
formation damage reducing permeability and porosity
blockages in perforations and sand control mechanisms
restricting and blocking of production tubing and flow lines
safety and well control equipment failure
gas lift mandrel blocking

artificial lift equipment wear

© 1990-2020 Petroleum Experts Limited

 194 Flow Assurance

induced corrosion underneath deposits in the production path

process equipment blocking

Scale is a crystalline deposit resulting from the precipitation of mineral compounds present in
the reservoir and production system. The precipitation of scales occur due to changes in the
ionic composition, pH, pressure and temperature of the fluids in the given system. When a
brine pressure reduces or brine temperature increases, a reduction in solubility of the salt
occurs and it will then precipitate and deposit as a solid, and is a self-scaling process. Calcite
deposition for instance generally falls under the self-scaling process definition, where the main
driver for its precipitation is the solubility of carbon-dioxide (CO2) from the water to the
hydrocarbon phases as pressure falls. The removal of carbonic acid from the water phase no
longer keeps the basic calcite dissolved and hence precipitates.

Carbonate scales such as calcium carbonate: Ca+2(aq) + H2CO3(aq) = CaCO3(s) +

CO2(aq) + H2O(l) are strongly pressure dependent, and will also precipitate at production
wells due to their dependency in pressure. Carbonate minerals are highly soluble in
hydrochloric acid, and can easily be dissolved and removed in the wellbore through bullheading
acid wash treatments.

Halite scaling is also a self-scaling process, and is caused by the reducing flowing fluid
temperature through the production path. Halite precipitation is further accelerated by
evaporation of water at gas lift valves caused from injection of the gas-lift-gas into the well
stream. Halite solubility in water decreases with decreasing temperature, favoring halite
dropout during the production of brines with high total dissolved solids (TDS). Halite
deposition further impacts gas lifted well performance from precipitating at the injection valve
(or orifice) reducing gas lift valve port areas, causing sub-optimum production and undesirable
multi-pointing injection situations resulting in well instability and production losses.

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 195

Reservoir brines are typically rich in cations such as barium, calcium and strontium, and when
mixed with injected seawater that is sulfate rich, leads to the precipitation of sulfate scales,
such as barium sulfate: Ba+2(aq) + SO4-2(aq) = BaSO4(s) and will deposit at, or near,
producing wells. Barium sulfate is one of the least soluble oilfield scales, and very difficult to
remove once formed. Strontium sulfate (SrSO4) will also form with barium sulfate in
sandstone reservoirs, in contrast, calcium sulfate (CaSO4) forms in limestone reservoirs.

Studies have shown that injecting untreated sea water into the producing formation mix with
the reservoir brine deep within the reservoir, resulting in barium ion stripping that would reduce
scale deposition at the producing wells. However, barium stripping does not necessarily occur
at far reaches in the reservoir, and the results of this study illustrate the necessary water
treatment requirements to avert future production consequences, costly squeeze treatments
for producing wells and equipment change-out. Treating the injected sea water to reduce the
sulfate ion concentration also reduces the nutrients that are available for growth of sulfate
reducing bacteria (SRB), averting the potential of future reservoir souring.

Further information on the background to scaling and water chemistry objects in RESOLVE can
be found in the RESOLVE manual.

4.2 Incompatabile Water Injection

4.2.1 Objective
The objective is to set up a fully integrated model to identify scale formation sites through
reservoir, wells and surface network that may arise from the injection of sea water that is
being used for reservoir pressure maintenance. The complete chemistry water cycle needs to
be captured, from injection of sea water + produced water to produced water breakthrough.

The results indicate the injection water filtering requirements at the field development design
stage that need to be implemented to prevent scale deposition, preventing future production
and field management issues.

The model will use RESOLVE to link the REVEAL reservoir simulator to the GAP well and
surface network model. RESOLVE Visual workflows will be used to dynamically pass the
producing well’s BS&W to the RESOLVE water chemistry data objects that will equilibrate the
compositions, identifying new species and concentrations forming from the mixing of reservoir
and sea water, and then dynamically update the water injection well’s water composition
during the simulation. Mineral formation will be captured as waters mix at the producing wells,
then deposit in the system such as the SubSea node in the GAP model.

© 1990-2020 Petroleum Experts Limited

 196 Flow Assurance

The GAP file uses inflow elements for the well inflow description, and the tubing to surface is
modelled using GAP pipe elements. Setting up the network model using inflow elements and
pipe to surface enables well nodes to be exposed to dynamically track solids and deposition
sites in the production system.

The gas lifted riser results will also show halite precipitation in early-time due to desaturation
of low produced BS&W in the riser due to gas-lift-gas injection, highlighting additional flow
assurance problems.

4.2.2 Integrated Model

The following screen shot illustrates what the RESOLVE file will look like when completed:

The following steps outline how to create and complete the file.

4.2.3 Setting up the Files

The files to complete this example can be found in: ‘~\Petroleum Experts\IPM 10\Samples
\FlowAssurance\5.1 Scale - Injection of Incompatible Waters\Initial Files’

1. Open GAP and select: File | Archive | Extract, then browse to the above file location to

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 197

open the WC_Example1.gar archive file. Once the GAP file has been opened, save and
close the file so it can be linked to a GAP instance in RESOLVE.
2. Open RESOLVE and select File | New, then add a GAP and REVEAL element to the
RESOLVE canvass using the Add Application tool bar icon shown in the following screen
shot:

a. Double-Click on the GAP instance and on the File Name path browse to the Initial
Files directory and link the WC_Example1.gap file.

b. Double-Click on the REVEAL instance and on the File Name path browse to the Initial
Files directory and link the WC_Example1_Chg_Inj_pH.rvl file, and select Start to
initialize the REVEAL instance as per the following screen shot:

© 1990-2020 Petroleum Experts Limited

 198 Flow Assurance

c. Once REVEAL has initialized the wells will now be visible on the RESOLVE canvass.
Double-Click again on the REVEAL instance as the Drainage region (advanced) IPR
model needs to be set up, as well as the Water Chemistry Tag Data mapping that now

becomes visible since RESOLVE knows that the REVEAL file has the water chemistry
option activated as per the following screen shots:

Flow Assurance November, 2020

Water Chemistry: Reservoir and Wellbore Scale 199

© 1990-2020 Petroleum Experts Limited

 200 Flow Assurance

3. Add the generic OpenServer object to the canvass via the Add DataObject tool pallet icon

a. The generic OpenServer icon enables certain calls and functions to be completed
between data objects that do not have in-build OpenServer functions available

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 201

4.2.4 Water Chemistry Data Objects

The Water Chemistry Data Objects have been designed to provide the water chemistry
capabilities of REVEAL to be accessible as calculations within RESOLVE. The water
chemistry data objects enable dynamic thermodynamic calculations to be performed at
specific GAP nodes for instance to understand precipitation of mineral species throughout the
production system during the simulation.

A comprehensive database of water based reactions are present with includes ionic species,
solid minerals and gases. The interaction of gases with the aqueous phase includes original
work to best represent the literature data on the solubility’s of gases such as CO2 and H2S in
water for various pressures and temperatures; the vaporization of water within the gas phase
is also captured through multi-phase calculations.

More information on the objects can be found in the RESOLVE manual.

4.2.4.1 Water Chemistry Data Object

The Water Chemistry Data Object stores a water composition and performs equilibration
calculations for the chemical species present in the water. Based on the entered water
composition and input pH, the equilibration operation calculates the equilibrium concentration of
the ions in the aqueous phase, amounts of solid precipitate, partial pressure of gases and final
pH.

4.2.4.2 Water Chemistry Mixer

The Water Chemistry Mixer data object performs a mixing calculation using the connected
Water Chemistry Data Object water compositions and flow rates. The mixing operation
calculates the concentrations of all species in the mixture based on the entered flow rates, and
equilibrates the resultant water composition and reports the equilibration results.

4.2.4.3 Water Chemistry PVT Mixer

The Water Chemistry PVT Mixer data object performs a mixing calculation using the
connected Water Chemistry Data Object’s water composition and Equation of State
description (EOS-PVT data object). The gas is assumed to be in equilibrium with the water,
and a two-way exchange of species can occur in this situation.

In particular, CO2, H2S and H2O can partition between the water and the hydrocarbon
phases, and can cause significant physical affects: for example injected CO2 from the gas-lift-
gas can dissolve in the water and change the pH, thereby preventing calcite scaling. If the
injected gas is dry, then liquid water will change to vapour, thereby causing drying and
increasing scaling possibilities. The amount of H2S is also important for souring issues.

The output of the mixing calculation is a final gas composition, and water composition, such
that the equilibria are satisfied.

© 1990-2020 Petroleum Experts Limited

 202 Flow Assurance

For further information on the water chemistry data objects, please see the relevant section of
the RESOLVE User Guide.

4.2.4.4 Adding Water Chemistry and PVT Data Objects

Add the water chemistry and PVT data objects and label them as per the following
screenshot:

4.2.5 Import or Manually Complete Water Chemistry Data Objects

4.2.5.1 Import Water Chemistry Object *.RDO File

Importing the *.RDO file automatically completes the Water Chemistry Data Object's input, the
following steps apply to import the *.RDO file:

1. Right-Click on each element and select Import from File, then browse to the Initial Files
folder and select the corresponding *.RDO file that have been named as per the PVT data
objects themselves.

NOTE: All Water Chemistry Objects can be automatically completed if desired from simply
importing the corresponding *.RDO file.

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 203

4.2.5.2 Manual Completion of the Water Chemistry “Mixer” Data Objects

1. Double-click on the WaterChemistryMixer element to access the input data

2. The Connection Name field showing the well names (Well1 and Well2 in left column) in the
below screen shot will automatically be filled in once the Well1 and Well2 have been linked
to the data object later on when completing the file
3. Set the Rate Type to Mass Solvent Rate
4. The Rate value (red column) will be updated on model initialization, and during the simulation

5. Complete the Mixer water chemistry mixer element in exactly the same way as outlined
above

As the model runs, the objects will be setup to automatically read the mass rates from GAP.

4.2.5.3 Manual Completion of the Water Chemistry PVT Mixer

1. Double-Click on the labelled WaterChemistryPVTMixer element, and set the Rate Type
field to Mass Rate Solvent and Mass Rate as per the following screen shot:

© 1990-2020 Petroleum Experts Limited

 204 Flow Assurance

a. Please note that once the connections in the model are made to the
WaterChemPVTMixer, the data object will automatically show the Connection Names
(Filtered and GasLift)

4.2.5.4 Import PVT Data Object’s Compositional PVT Data

1. Double-Click on the GasLift PVT data object to access the PVT input
2. Once opened, select Import PRP option shown below to import the EOS PVT data file
from the Initial Files folder

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 205

4.2.6 Add Visual Workflow Elements to RESOLVE Canvass

1. From the Add Application pallet, add the two Visual Workflow elements, and position
and name them as per the following screen shot:

4.2.6.1 Visual Workflow Functions

The function of the Setup visual workflow is to set mineral saturation indices (SI) for
magnesite and dolomite. The Dolomite SI is set to 10, essentially inhibiting dolomite
precipitation that is very slow to form, and the mineral will form, however it is not the focus of
the study. The Setup visual workflow also dynamically sets the water chemistry mixer and
PVT mixer’s pressures and temperatures using GAP pressures and temperatures from the
sub-sea gathering system.

REVEAL production and injection water rates are also dynamically captured and used as the
solvent rates that need to be defined in the water chemistry and PVT mixers.

The Filter visual workflow uses OpenServer operations that copy data from the Subsea water
chemistry object to the Filtered water chemistry object and trigger equilibration calculations,
enabling tracking of evolving mineral species during the simulation. The workflow also
emulates water treatment via the Dilute Mineral fractions concentrations operation.

4.2.6.2 Import Visual Workflow

If desired, the visual workflows can be imported into the visual workflow elements as per the
following steps:

© 1990-2020 Petroleum Experts Limited

 206 Flow Assurance

a. Open each workflow element, then select the Import Workflow from File option, then
browse to the Initial Files folder and select the appropriate *.VWK file:

4.2.6.3 Create Visual Workflows

The visual workflows can be created using the following steps:
1. Double-Click on the Setup visual workflow icon in the main RESOLVE canvass and add
an Assignment element then link the visual workflow elements together

2. Right-click on the newly added assignment element and select change label from the pop-
up menu and type Setup for the name
3. Double-click on the newly labeled Setup assignment element to access the assignment
variables, values and arguments. The completed Setup workflow will look as per the
following screen shot:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 207

4. Assess the first row in the Variable column, and type the first entry. Note that each time a
‘.’ is added, the Intellisense string writer automatically pops up enabling immediate selection
of the desired options as shown in the following figure:

© 1990-2020 Petroleum Experts Limited

 208 Flow Assurance

5. Select the DataModel Intellisense option, enter another ‘.’ and continue to complete the first
variable assignment entry of:
Well1.DataModel.Minerals.Data[“Magnesite”].InputSaturationIndex followed by the
Set equal to value of zero (0)
6. Complete the other assignment entries as per the above screen shot and return to the main
RESOLVE canvass
7. Double-click on the Filter visual workflow and add an Operation element and link the
workflow together as was done for the Filter visual workflow

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 209

8. Right-click on the newly added operation element and select the Change Label pop-up
menu option and enter Filter
9. Double-click on the operation element and enter the following Operations, the completed
Filter operation will look as per the following screen shot

© 1990-2020 Petroleum Experts Limited

 210 Flow Assurance

10. The following screen shots outline how to complete the Filter operation element data
entry:

Flow Assurance November, 2020

Water Chemistry: Reservoir and Wellbore Scale 211

© 1990-2020 Petroleum Experts Limited

 212 Flow Assurance

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 213

4.2.6.4 Linking Wells, DataObjects and Visual Workflows

Now that the reservoir and surface network models and all other elements have been added
and completed as per the above steps, they can now be linked together as per the system
schematic:

© 1990-2020 Petroleum Experts Limited

 214 Flow Assurance

4.2.7 Importing Application Variables into RESOLVE

RESOLVE offers several options to pass data between modules, as well as monitor variables
that can be used to trigger advanced field management logic for instance. To enable
RESOLVE to track or manage variables, they must first be published from each program as
outlined in the following steps:

1. From the RESOLVE main menu, select: Variables | Import Application Variables, then

set the following REVEAL variables:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 215

2. Select the GAP tab and import the following GAP variables:

© 1990-2020 Petroleum Experts Limited

 216 Flow Assurance

4.2.8 Entering the RESOLVE Forecast Schedule

1. From the RESOLVE main menu, select: Schedule | Forecast Data, and input the
following start and end-date, and forecast options:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 217

2. The model is now ready to run.

4.2.9 Analysis of Results

4.2.9.1 Direct Access of REVEAL Results

The REVEAL 3D results can be accessed by firstly selecting the opened instance of
REVEAL, then from the main menu select: Results | Reservoir 3D Graphics as shows in the
following screen shot:

© 1990-2020 Petroleum Experts Limited

 218 Flow Assurance

The REVEAL 3D reservoir grid plot variables, grid properties, etc., can be selected and
adjusted as shown in the following screen shot:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 219

Using the above reservoir 3D plotting property and results Play-Back options, the barium ion
rich brine water concentration as well as the sulfate rich water injection fluid at the start,
during, and at the end of the simulation can be viewed as shown in the following screen shots.
One can see that Well1 sees the barium ion and sulfate mixing first due to the breakthrough
of untreated seawater, resulting in barite deposition in the near wellbore and subsea network:

Mass Concentration Ba+2 in Reservoir Water at Time = 0-days

NOTE: To see Ba+2 variation at T=0-day, Right-Click on the scale and select: Scale |
Timestep Min/Max option.
Mass Concentration Sulfate at Time = 0-days (zero concentration since injection has
not started)

© 1990-2020 Petroleum Experts Limited

 220 Flow Assurance

As mentioned previously, barium stripping does not necessarily occur at far reaches in the
reservoir. Looking at a Z-direction slice across the reservoir, one can see that the sulfate rich
reservoir brine at Well1 mixes with the sulfate rich injection water in late-time due to injection
water breakthrough. The mixing of the incompatible waters at time of breakthrough causes
barite to precipitate in the near wellbore region as can be seen in the following screen shot:

Late-Time Well1 Near Wellbore Barite Formation from Incompatible Waters Mixing:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 221

4.2.9.2 RESOLVE Published Variable Results

The variables that were published in RESOLVE during the model construction enable the
dynamic data to be passed to the visual workflows that can be used to trigger events based
on given values, etc. The published variables where the plot check option is selected also
means that the data associated with the given published element can also be plotted, enabling
the GAP model nodes of interest to be plotted.

4.2.9.3 Published Variable and Results Plotting in RESOLVE

Both the published variable tabular results and direct plotting options are also available in
RESOLVE, where the results options can be accessed directly from the main menu RESULTS
menu, as shown in the following screen shot:

© 1990-2020 Petroleum Experts Limited

 222 Flow Assurance

4.2.9.3.1 View Forecast Results (table)

Selecting the first pop-up menu results option enables the tabular data to be reviewed, plotted,
saved and or exported as shown in the following screen shot:

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 223

4.2.9.3.2 View Forecast Plots

From the RESOLVE main menu, select: Results | View Forecast Plots, the following blank plot
and results menu options appear:

From either of the above RESOLVE plotting options, the SubSea Minerals variable can be
selected to plot the Barite-ppm to understand deposition in-time. One can see in the following
screen shot that deposition only starts after the injection water breakthrough at Well1, then
subsequently mixing again with the produced water from Well2 in the SubSea node and
gathering system in GAP:

Subsea Barite Deposition Compounded from Well2 Barium Ion Rich Water Mixing with Sulfate
Rich Well1 Produced Water at the Subsea Gathering System Node:

© 1990-2020 Petroleum Experts Limited

 224 Flow Assurance

The late-time barite deposition issues need to be dealt with at the design phase to aviod
catastrophic production issues, where interventions become further compounded from
reducing economic asset value.

The following screen shot illustrates the elimination of barite deposition at the Subsea node
due to the implementation of water treatment facilities removing the sulfate ion (SO -2):
4

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 225

The following screen shot shows the Gaslift Mandrel node Halite deposition that occurs in
early-time only from the desaturation of water vapor caused by the gas-lift-gas. Note that as
the water production in the riser increases the Halite precipitation stops:

Halite Precipitation due to Desaturation from Gas-Lift-Gas Injection

The Halite precipitation in the riser shown in the above screen shot is short-lived (~10-month),
and stops as the produced water increases, i.e., the dissolved salt remains dissolved in the
reservoir brine.

The following screen shot shows the varying late-time well Barite production. Note that is is
considerably lower than the SubSea node due to previous deposition and further mixing in the
reservoir reducing the Barium ion concentration:

Well Barite Production

© 1990-2020 Petroleum Experts Limited

 226 Flow Assurance

REVEAL results can be accessed directly to understand the changing species ion
concentrations by plotting both reservoir and production well’s data. Additional plots can be
opened to further understand deposition sites and the positive impact water treatment has on
mitigating mineral deposition, fluid pH, etc.

REVEAL wellbore results can also be accessed by selecting from the main REVEAL menu:
Results | Wellbore Results

4.2.9.4 Calculating Barite Mass Deposition from Part-Per-Million Concentrations

The maximum Barite concentration at the Subsea node is approximately 45ppm in the late-

time period, and can be translated to mass deoposition value using the following calculation
approach:

Barite Deposition SubSea (lbm/d) = (? solvent rate * Barite ppm)/1e6 = [lbm/d]

SubSea Barite (end of simulation)

Mass Rate Solvent (end of simulation) = 2573168lbm/d

End-of-Run Barite Deposition SubSea (lbm/d)

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 227

Barite deposition.

4.2.10 Summary
To conclude, the value the IPM suite of tools offer in being able to capture the in-depth physics
of complex systems enable detailed studies to be performed at the design phase, ensuring
that given production and field development strategies can be sustained over the long-term.
The study illustrated the importance of including the water chemistry physics of REVEAL to
understand what, and if any, water treatment requirements were required. The conformation
of implementation of necessary water treatment facilities enabled conformation of the
necessary field design requirements to avoid late-time catastrophic operational issues.

Further analysis invoking REVEAL rock mechanics also needs consideration to understand
changing well breakthrough times due to thermal fracture phenomena. For further information
on rock mechanics and thermal fracturing, please see the relevant section of the REVEAL
User Guide.

4.3 Calcite in Well

4.3.1 Objective
Calcium Carbonate (Calcite) scale can have a significant impact on producing/injecting
wells. This can cause blockages in perforations, reduce the diameter of the producing
wellbore, cause safety valve and choke failure, cause pump wear or reduce the effectiveness/
block gas lift valves. If the scale is not deposited, the suspended particles can cause problems
at surface by plugging filtration equipment and reducing oil-water separator efficiency.

Scale can be treated through various means, for example, mechanical (pigging or abrasive
jets) or chemical inhibitor. Carbonate scales such as calcite can be dissolved using
hydrochloric acid. In either case, there is a cost of deferred production and for deep subsea
wells, the above may be extremely costly to implement.

4.3.2 Formation and Physical Phenomenon

Dependent on the composition, pressure, temperature and total dissolved solids, finely divided

© 1990-2020 Petroleum Experts Limited

 228 Flow Assurance

calcium bicarbonate crystals can precipitate.

One of the principle driving forces in calcite formation is the CO2 partitioning between water
and hydrocarbon phases. When carbon dioxide dissolves in water, carbonic acid is produced.
This lowers the pH and ‘ties up’ the carbonate so that it is unable to form calcium carbonate.
Once the CO2 partitions to the hydrocarbon phase, the opposite process occurs in the water
where the pH then rises, calcium carbonate solubility falls and hence calcite precipitates.

It is therefore important to be able to capture this physical phenomena when modelling.

4.3.3 Design and diagnosis

Careful planning, design and mitigation through modelling can highlight whether calcite
formation is a risk and take steps to prevent costly unplanned remediation works in the field.
The first step is to detect the onset of calcite formation. This can be done with the Water
Chemistry Data Object within RESOLVE, allowing a dynamic calcite formation envelope to be
automatically created during a forecast based on the conditions.

Once detected, different mitigation strategies can be investigated. These can be in terms of
scheduled maintenance of some wells or could be an operational control on the wells to
ensure that they stay out of the formation envelope. Whilst the later approach may reduce the
flowrate by choking back the well, this option should be evaluated against the economics of
potential production workovers/deferment.

It is only through evaluating different options in an integrated approach, that the correct
strategic decision can be made going forward.

4.3.4 Integrated Model

The following example uses an integrated model within RESOLVE to calculate a calcite
formation envelope from the produced water composition. Logic is then put in place to prevent

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 229

operating within this formation envelope hence achieving the principle objective of Flow
Assurance.

Initial Set-up

The starting file is titled ‘Calcite Detection and Mitigation: Initial.rsa’, the GAP file is loaded
with a single well with inline choke. Whilst modelling should always be in context of the full
system, in this example, the GAP model is simplified in order to focus on the building of the
workflow. The focus of this example is to demonstrate how calcite can be detected and
mitigated against rather than building a workflow to cycle through wells and perform
mathematical optimisation.

With the Water Chemistry Data Object within RESOLVE, the scale envelope of the well can
be calculated. A workflow is then setup to assess whether the current operating conditions in
a well in GAP are within this envelope, if they are, automatic mitigation in terms of choke
control is employed.

The following screen shot illustrates what the RESOLVE file will look like once completed:

© 1990-2020 Petroleum Experts Limited

 230 Flow Assurance

The following steps outline how to create and complete the file.

4.3.5 Setting up the Files

The files to complete this example can be found in: ‘~\Petroleum Experts\IPM 10\Samples
\FlowAssurance\5.2 Scale - Calcite Detection and Mitigation\Initial Files’

1. Open GAP and select: File | Archive | Extract, then browse to the above file location to

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 231

open the Calcite.gar archive file. Once the GAP file has been opened, save and close the
file so it can be linked to a GAP instance in RESOLVE.
2. Open RESOLVE and select File | New

4.3.6 Water Chemistry Data Objects

The Water Chemistry Data Objects have been designed to provide the water chemistry
capabilities of REVEAL to be accessible as calculations within RESOLVE. The water
chemistry data objects enable dynamic thermodynamic calculations to be performed at
specific GAP nodes for instance to understand precipitation of mineral species throughout the
production system during the simulation.

A comprehensive database of water based reactions are present with includes ionic species,
solid minerals and gases. The interaction of gases with the aqueous phase includes original
work to best represent the literature data on the solubility’s of gases such as CO2 and H2S in
water for various pressures and temperatures; the vaporization of water within the gas phase
is also captured through multi-phase calculations.

More information on the objects can be found in the RESOLVE manual.

4.3.6.1 Water Chemistry Data Object

The Water Chemistry Data Object stores a water composition and performs equilibration
calculations for the chemical species present in the water. Based on the entered water
composition and input pH, the equilibration operation calculates the equilibrium concentration of
the ions in the aqueous phase, amounts of solid precipitate, partial pressure of gases and final
pH.

4.3.6.2 Water Chemistry PVT Mixer

The Water Chemistry PVT Mixer data object performs a mixing calculation using the
connected Water Chemistry Data Object’s water composition and Equation of State
description (EOS-PVT data object). The gas is assumed to be in equilibrium with the water,
and a two-way exchange of species can occur in this situation.

In particular, CO2, H2S and H2O can partition between the water and the hydrocarbon
phases, and can cause significant physical affects: for example injected CO2 from the gas-lift-
gas can dissolve in the water and change the pH, thereby preventing calcite scaling. If the
injected gas is dry, then liquid water will change to vapour, thereby causing drying and
increasing scaling possibilities. The amount of H2S is also important for souring issues.

The output of the mixing calculation is a final gas composition, and water composition, such
that the equilibria are satisfied.

© 1990-2020 Petroleum Experts Limited

 232 Flow Assurance

For further information on the water chemistry data objects, please see the relevant section of
the RESOLVE User Guide.

4.3.6.3 Adding Water Chemistry and PVT Data Objects

Add the water chemistry, PVT data objects, DataSets, OpenServer instance and GAP
instance and label them as per the following screenshot:

4.3.7 Import or Manually Complete Water Chemistry Data Objects

4.3.7.1 Import Water Chemistry Object *.RDO File

Importing the *.xml file automatically completes the Water Chemistry Data Object's input, the
following steps apply to import the *.xml file:

1. Open the water chemistry element and select Import, then browse to the Initial Files folder
and select the corresponding *.xml file that have been named as per the PVT data objects
themselves.

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 233

4.3.7.2 Manual Completion of the Water Chemistry PVT Mixer_2

1. Double-Click on the labelled WaterChemistryPVTMixer element, and set the Rate Type field
to Mass Rate Solvent and Mass Rate as per the following screen shot:

a. Please note that once the connections in the model are made to the
WaterChemPVTMixer, the data object will automatically show the Connection Names
(WaterIn and GasLift)

4.3.7.3 Import PVT Data Object’s Compositional PVT Data_2

1. Double-Click on the GasLift PVT data object to access the PVT input
2. Once opened, select Import PRP option shown below to import the EOS PVT data file
from the Initial Files folder

© 1990-2020 Petroleum Experts Limited

 234 Flow Assurance

4.3.7.4 Import Application Variables

The workflow will use the pressure and temperature results from the GAP model later and
hence the variables should be first imported into RESOLVE. This can be done through
Variables|Import Application Variables as shown in the screenshot below.

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 235

4.3.7.5 Setup DataSets

Before the DataSets are populated they must be first setup with the appropriate columns.

The CalciteFormationEnvelope must first have the columns defined as Temperature and
Pressure (where the calcite formation pressure will be calculated). After the columns are
defined, a temperature range must be entered within the Temperature column as shown
below. For this example, the Solubility index is set as 0.

The Results DataSet will also require to be setup with the appropriate columns (shown below)
but does not require any initial inputs.

© 1990-2020 Petroleum Experts Limited

 236 Flow Assurance

4.3.8 Add Visual Workflow Elements to RESOLVE Canvass

The Visual Workflow is used to execute three different tasks;
1. Creating the calcite formation envelope using the Water Chemistry Data Objects
2. Checking whether the well is currently operating within the Calcite formation envelope
3. Mitigates against the calcite formation through dynamically changing the wellhead choke.

The setup of this workflow is outlined as below.

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 237

4.3.8.1 Import Application Variables

Before starting the creation of the workflow, the following variables should be defined;

Note that the GRD results are defined as arrays.

4.3.8.2 Creation of the Calcite Envelope

The calcite envelope is formed by performing an water chemistry equilibration calculation for a
set of temperatures and across different pressures. For one single temperature, a sensitivity
on pressure is performed, once the calcite precipitation pressure is found, this pressure is
noted and the calculation continues. By performing this calculation for different temperatures, a
calcite precipitation operating envelope is found. This calculation is available within a single
operation and will be used for the following workflow.

The first three elements to use are shown below; the first solves the network so that the
current operating values can be retrieved, the second element sets up the calcite formation
envelope question and the third performs the envelope calculation as described above.

© 1990-2020 Petroleum Experts Limited

 238 Flow Assurance

4.3.8.3 Checking operating conditions against the Operating Envelope

Once the Calcite Formation envelope is known, the next step is to determine whether the well
is currently within the envelope, if so, calcite may form. The workflow employs the following
methodology;
Firstly retrieving the gradient pressure and temperature from GAP into their respective
workflow variables
Then using interpolating for each temperature to gain the relevant minimum pressure from
the Calcite formation envelope.
The logic to export the results is also included in this section and mitigation will be covered
later.

1/ Assignment Elements NumofPoints and GetGRDResults

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 239

2/ Minimum Pressure calculation (CalcMinP SubFlowSheet)

© 1990-2020 Petroleum Experts Limited

 240 Flow Assurance

3/ Report Results (SubFlowSheet) and raise flag if operating within Calcite Formation
envelope

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 241

4.3.8.4 Mitigation
With knowledge that the well is currently operating in the Calcite formation envelope, steps
can be taken to automatically mitigate against this. In this circumstance, the pressure in the
well is raised by a wellhead choke. However, in a network model, this will have an impact
across the entire system and hence may not be the most appropriate mitigation strategy. The
mitigation employed should be appropriate for the individual system and controls.

In this workflow, a 'mitigation' counter is used to inform the workflow whether mitigation is to
be conducted. Following this, the workflow then searches for the point in the wellbore that will
require the highest dP to go above the calcite formation envelope and sets that dP as the
wellhead choke value. The logic is designed to mitigate rather than optimise so a tolerance is
used to allow the well choke dP to be higher than strictly necessary.

1/ Mitigation elements. The CalcdPChoke SubFlowSheet is outlined below.

© 1990-2020 Petroleum Experts Limited

 242 Flow Assurance

2/ CalcdPChoke SubFlowSheet: Determining the choke setting required

4.3.9 Analysis of Results

4.3.9.1 Summary and Analysis of Results
Once complete, the Visual Workflow can then be run using the 'Run' arrow icon within the
workflow editor. As the Mitigate variable has been set initially to zero. The workflow will not

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 243

adapt the well head pressure to stay out above the minimum pressure. From the results plots
it can be seen that for the majority of the well, the actual pressure is lower than the minimum
and hence there is a risk of calcite deposition.

Once the initial investigation has been complete the mitigation variable flag can be set within the Get count(2)
assignment block to 1. This will tell the workflow that it is to perform mitigation via the increase of well head
pressure.

© 1990-2020 Petroleum Experts Limited

 244 Flow Assurance

Running the model after making this change yields the below results;

Flow Assurance November, 2020

 Water Chemistry: Reservoir and Wellbore Scale 245

It can be seen that the actual pressure is higher than the calcite formation pressure throughout the depth of the
well.

Within the Finished Example files for this particular example, a more rigourous workflow is provided that uses a
Form to indicate when mitigation may not be possible for given conditions but also will minimise the well head
pressure with accordance to the envelope too. This later steps ensures that the production is maximised within the
desired operating limits.

© 1990-2020 Petroleum Experts Limited

Chapter

5
 Steady state and Transient Flow Assurance Examples 247

5 Steady state and Transient Flow Assurance

Examples

5.1 Stability: comparing Steady State Vs Transient

simulations

5.1.1 Introduction
There are two fundamental components of any study, (i) the analysis required and (ii) the tools
used:

The analysis: The type of analysis should dictate the tools used, rather than the analysis
being dictated by what tools are available. The impact of any analysis must be understood in
the context of the upstream and downstream part of the system under study.

The Tools: The functional silo approach that exists between disciplines reduces the quality
and relevance of any analysis performed in isolation: the fluid does not know about our
departments, organisations or software and until that day occurs, our modelling approach
must replicate the process that the fluid undergoes as it travels through the production system
.

This section of the manual will attempt to address the classical limitations of the way in which
studies are approached.

5.1.2 Manual Approach

The approach that will be taken in this case is to consider the merits of transient versus steady
state analysis, specifically in the context of the Logie Mill Pipelines. To achieve this both types
of analysis will be run using the same data sets (the steady state data feeding the transient
simulation) at early and late field life conditions.

From the above analysis, the objective is to derive some rules of thumb regarding when to use

© 1990-2020 Petroleum Experts Limited

 248 Flow Assurance

which analysis.

5.1.3 Run Simulation at Early Time

This example starts from the files found in '~\Petroleum Experts\IPM 12.5\Samples
\FlowAssurance\3.1 Stability - Steady state Vs Transient\Initial Files'.
The sample files folder can also be accessed from the windows menu as shown below.

Open GAP and go to File | Archive | Extract and select the *.gar file from the samples
folder.

Edit Projects Paths (pressing Edit | Edit Project Paths) – this simply makes sure that
the archive file has been unpacked, and that GAP is picking up the correct underlying
application.

Run the full simulation (with Snapshots turned on): the below screen shots outline this
can be done from Prediction | Run Prediction… | Next | Next | Next | Settings and
enabling the option:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 249

Save the GAP file, and Close it down (it will now be opened from RESOLVE)

Open up the *rsl file provided (the GAP instance must point to the GAP file just closed
down). Please note that whilst this example starts from a pre-built file, a step-by-step
example of how to build a model linked with transient simulator can be found in Example
5.3 of the RESOLVE user manual.

© 1990-2020 Petroleum Experts Limited

 250 Flow Assurance

Double click on the LedaFlow instance and ensure that the following options have been
selected:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 251

5.1.3.1 Early Time (Steady State)

Load the first timestep (30/09/2004), see the separator rates and take note of them,
Press save

© 1990-2020 Petroleum Experts Limited

 252 Flow Assurance

Gap will the prompt a question: pressing yes saves this timestep as the home screen

Navigate to the MAIN LINE flow assurance results:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 253

In terms of the results from this steady state solve:

The pressure drop across the pipeline is 754 psig (pressure Manifold) minus 655 psig
(pressure at J4) is approximately 100 psig.

The GOR is 9801 scf/stb and the liquid rate is 2062 stb/day with a 1.14% water cut.

5.1.3.2 Early Time (Transient)

The RESOLVE file is setup to solve in single solve mode. This can be implemented
from Options | System options…:

© 1990-2020 Petroleum Experts Limited

 254 Flow Assurance

Pressing solve in RESOLVE will solve the GAP file, and pass the above values to feed
the Transient simulation.

The transient simulation is setup to receive the pressure and mass coming in, and it
calculates the pressure coming out over a much finer time scale (in this case the second
timescale for an entire day)

The below screen shot from LEDA can only be seen when LedaFlow is started as a separate
application, otherwise all the results are accessed through the RESOLVE interface.

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 255

The inlet and outlet pressures for GAP and LedaFlow are the same (as expected) and this can
be verified from the tabulated results in RESOLVE or from comparing the GAP screen and the
LEDAflow screens.

© 1990-2020 Petroleum Experts Limited

 256 Flow Assurance

The only real difference between the two simulations appears about 40,000ft along the
pipeline, after about an hour (3854 seconds shown in second screen shot below):

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 257

Pressure fluctuations appear to occur at the two high points of the pipeline profile:

© 1990-2020 Petroleum Experts Limited

 258 Flow Assurance

Changing the pressure axis to gas volume fraction shows slugging in the pipeline:

The results for the entire 86400seconds show the terrain induced parts:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 259

5.1.3.3 Early Time Analysis & Results

In the above context there seems to be good alignment between the pressure results
between the transient and steady state simulations.

The transient simulation does indeed indicate that there are some terrain induced effects
due to the ups and downs of the pipeline trajectory: however these are small in terms of
pressure fluctuations.

5.1.4 Run Simulation at Late Time

Performing the same steps as above.

Run the GAP model with snapshots saved.

Reload the LAST timestep in GAP (note rates and pressures going into the mainline) and
press save

© 1990-2020 Petroleum Experts Limited

 260 Flow Assurance

perform the LEDA flow simulation again by solving RESOLVE:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 261

5.1.4.1 Late Time results

The results are very similar to those found in early time:

It can be seen that there are slugs forming around the peaks in the pipeline, but again these
dissipate in terms of pressure/energy over the remaining length of the pipeline.

5.1.5 Analysis & Results

Comparing the transient response over steady state time lengths (1 day) shows the response
are similar in magnitude. Even when comparing early and late field life.

© 1990-2020 Petroleum Experts Limited

 262 Flow Assurance

The transient response shows greater variation in a smaller time scale as would be expected,
and it can be seen that changes in the gas volume fraction appear to occur over the pipeline
and the response to this can be seen in both GAP and LedaFlow.

The only real difference is that in early time the slugs dissipate smoothly, whereas in later time
this is not as true. This can be seen by comparing the two profiles directly:

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 263

5.2 Stability: Automated initiation of transient simulations

5.2.1 Introduction
Within each sub case of the flow assurance manual presented so far, there has been a
consistent approach of identifying the physical phenomena (detection) first and then identifying
operation strategies (mitigation) to avoid the scenarios presenting in the first place.

The current chapter is focused upon the integration of both steady state and transient
approaches over a prediction period. The below section thus focuses on how to:

Use the steady state model for the prediction, and providing that specific criteria are
detected in the model,

Initiate a transient simulation and,

Use a workflow to interpret the transient results, and if necessary modify the operating
strategy in the steady state model before continuing to the next time step.

The steps outlined below will vary depending upon the type of model and analysis being
performed, as such the below generic steps can be used in almost all contexts:

The RESOLVE model shown below illustrates how such a model could be connected:

© 1990-2020 Petroleum Experts Limited

 264 Flow Assurance

5.2.2 Workflows

5.2.2.1 PreSolve
This is simply setting the transient solver to not run. This will be changed as a post solve
action, should certain criteria be fulfilled.

Flow Assurance November, 2020

 Steady state and Transient Flow Assurance Examples 265

The below workflows show how to detect, and initiate the transient simulator (in this case
LedaFlow)

5.2.2.2 Detect
In this section we use the PE5 stability criteria to see whether the pipeline is showing stability.
If yes, carry on the run, if not then switch on LedaFlow:

The first thing is to establish whether stability is a concern or not:

If stability is not a concern (i.e. a value of 0) then continue the run. If stability is a concern,
then switch on LedaFlow:

© 1990-2020 Petroleum Experts Limited

 266 Flow Assurance

5.2.2.3 Mitigation
The action would be dictated by the type of analysis being performed, and may be to imitate a
routing change, a re-solve or some other such action.

Flow Assurance November, 2020