BP Exploration

Completion Design Manual

(Volume 1)

Contents

Chapter 1 How to Use the Manual

Chapter 2 Developing a Statement of Requirements

Chapter 3 Conceptual Completion Design

Chapter 4 Determining the Optimum Well Performance

Section 4a Reservoir Performance
Section 4b Near Wellbore Performance
Section 4c Multiphase Flow
Section 4d Artificial Lift
BP Exploration How to use the Manual – Chapter 1
Completion Design Manual CONTENTS

INTRODUCTION 1

HOW TO GET THE BEST FROM THE MANUAL 7

BP Exploration How to use the Manual – Chapter 1
Completion Design Manual

INTRODUCTION

The aim of this manual is to increase the awareness,

throughout all the technical disciplines within the company,
of the importance of the completion design process on overall
well productivity and hence field profit.

An interdisciplinary approach to completion design, involving

Reservoir Engineering, Petroleum Engineering, Production
Engineering, Drilling Engineering, Production Geology and
Field Development is necessary. To achieve optimum well
productivity, it is important to understand the whole
completion design process and how key decisions in different
disciplines can have a major impact on this productivity.

This manual has reviewed and systemized this process in

order to provide a design system which takes the user from
establishing the aims for the well through all the analyses
and key decisions up to the stage where the hardware is run
into the ground.

The proper consideration of all the parameters involved in

the␣ completion design process is fundamental in achieving an
optimum design. The selection of the correct elastomer in the
completion equipment can, for example, have as significant
an impact on the overall well productivity as adopting the
correct procedures to maximize the near wellbore
performance.

Many of the decisions taken in the completion design process

are interrelated. An initial decision on say the appropriate
architecture of the completion may change as a result of a
subsequent review of downhole production chemistry
problems and available well servicing and workover
techniques. The process is not therefore sequential and many
studies need to be carried out in parallel. However, three
distinct phases in the process can be identified:

• Conceptual completion design.

• Detailed completion design.
• Procurement and installation.

Page 1
How to use the Manual – Chapter 1 BP Exploration
Completion Design Manual

The activities associated with each phase are illustrated in

Figures 1.1, 1.2 and 1.3. As can be seen from Figure 1.4, the
manual covers the whole process and is divided into eleven
key topics and eight subsidiary topics.

The conceptual design process takes the user through the

analyses and key questions which need to be addressed prior
to the detailed design stage. In establishing the conceptual
designs, the user must employ the whole process, as detailed
in the manual, on a conceptual basis. This allows many of
the dilemmas associated with interrelated decisions to be
resolved at an early stage. The results of the conceptual
design are then used as the basis for the final and more
detailed design.

To fully exploit a field’s potential, the conceptual completion

design should commence at the field appraisal stage (see
Figure 2.1 in Chapter 2 – Developing a Statement of
Requirements). Well productivity and completion design can
have a fundamental effect on field design and development
and must therefore be addressed at an early stage.

As more wells are drilled and more information becomes

available, or as well conditions change, it is important to go
back and review the initial statement of requirements and
modify the conceptual design. In this way, the completion
design can be optimized around the new conditions or any
further information. The optimum completion design for any
field is not fixed in time and it should change to either meet
the new conditions or as a result of a better understanding of
the environment.

With simple mechanical workovers costing in the order of

£0.5 million from a North Sea platform or at the North Slope
and £3 million for subsea wells, the adoption of a more
systematic approach to completion design should provide the
opportunity for significant cost saving by reducing the
number of workovers. In addition, there may be many wells
with additional potential that could be exploited by utilizing a
more rigorous view of the near wellbore performance.

Page 2
BP Exploration How to use the Manual – Chapter 1
Completion Design Manual

Statement of requirements

Establish design basis

Determine the optimum well

performance

Establish the configuration

of the conceptual
completion design

Perform tubing movement and

stress calculations

Review material requirements

Investigate near wellbore

performance problems

Brief review of well

servicing/workover philosophy

Establish broad budget costs

Finalize conceptual design

NOTE: More complicated completion problems may require additional stages in

the conceptual design process eg workover problems, equipment and
material availability etc.

Figure 1.1 – Conceptual Completion Design Process

Page 3
How to use the Manual – Chapter 1 BP Exploration
Completion Design Manual

Conceptual completion design

Review any changes to the Statement of Requirements

Modify design basis

Finalize strategy to ensure optimum well

performance throughout field life

Finalize near wellbore design

Develop the structure of the final completion design

Perform tubing movement/stress calculations

Establish final material requirements

Specify completion fluids

Review any production chemistry problems

Select appropriate tubulars

Design of surface equipment

Select types of completion equipment

Review workover and well servicing philosophy

Review procurement – establish lead times

Prepare budget costs

Establish completion procedures

Prepare final completion design report

Figure 1.2 – Detailed Completion Design Process

Page 4
BP Exploration How to use the Manual – Chapter 1
Completion Design Manual

Detailed completion
design

Prepare technical
specifications

Issue enquiry

Technical evaluation of bid

Commercial evaluation
of bids

Presenting a recommendation
to a bid committee

Pre-production meeting
with vendors

Issue final order

QA and inspection

Prepare procedures for

onshore make-up

Witness onshore
make-up

Run completion

Figure 1.3 – Procurement and Installation Process

Page 5
How to use the Manual – Chapter 1 BP Exploration
Completion Design Manual

KEY TOPICS ASSOCIATED SUBSIDIARY TOPICS

1. How to use the manual

2. Developing a Statement
of Requirements

3. Conceptual
Completion Design a. Reservoir performance

b. Near wellbore performance

4. Determining the Optimum
Well Performance c. Multiphase flow

d. Artificial lift

5. Architecture of Completions
a. Optimizing the completion
in the near wellbore region

6. Tubing Movement and

b. Downhole production
Stress Calculations chemistry

7. Selection of Tubulars a. Materials

8. Selection of
Completion Equipment

9. Procurement of Equipment

10. Running the Completion a. Completion fluids

11. Well Servicing and

Workover Philosophies

Figure 1.4 – Format of the Manual

Page 6
BP Exploration How to use the Manual – Chapter 1
Completion Design Manual

HOW TO GET THE BEST FROM THE MANUAL

A distinction has been made in the manual between the

eleven key topics and the eight subsidiary topics. The
subsidiary topics, although no less important to the success
of the well, are areas where the completion engineer can
supplement his working knowledge with the specialist
assistance available in the company.

The key topics are those areas a completion designer needs

to␣ be fully conversant with and are therefore dealt with in
greater detail.

To cope with the large scope and overall length of the manual
and to make the manual manageable, each key topic has
been split into five colour coded subsections. Figure 1.5
illustrates the breakdown. Subdividing the key topics in this
manner allows the reader to quickly locate the subsection
appropriate to his knowledge and interest. It is felt that all
users, irrespective of discipline, should read the summary
and practical guidelines for all topics in order to achieve a
greater awareness of the whole process.

The first section in any chapter will deal with the theory or
background information associated with each topic. This
section will primarily be used by readers new to the topic or
by users with a particular problem which requires further
theoretical investigation. The techniques section details the
actual procedures which should be employed, and includes
details of how to use the appropriate software or company
procedures. The worked examples illustrate how the
techniques are employed in a real situation. The summary
and practical guidelines essentially provide a quick reference
containing rules of thumb associated with any topic. The
final subsection contains a comprehensive bibliography and
list of references together with a list of company specialists
who can be contacted for further assistance. The aim of this
section is to provide a means of obtaining further information
and specialist advice on problems not covered in sufficient
detail in the manual.

A complete chapter in the manual will contain the five

subsections of the key topic, supported, if appropriate, by the
associated subsidiary topic as illustrated in Figure 1.6.

Page 7
How to use the Manual – Chapter 1 BP Exploration
Completion Design Manual

The manual is not a ‘cook book’, in that it does not contain a

number of recipes for perfect completion design to suit any
situation. However, by employing the analysis techniques and
following the total process described in the manual, the
optimum design solution can be found for any set of well
conditions.

Page 8
BP Exploration How to use the Manual – Chapter 1
Completion Design Manual

Each Key Topic contains five colour coded subsections

CONTACTS,
REFERENCES,
ETC

Example: Determining the SUMMARY AND

Optimum Well PRACTICAL
Performance GUIDELINES

WORKED
EXAMPLE

TECHNIQUES

THEORY

CONTACTS,
REFERENCES,
ETC

SUMMARY AND
Example: Architecture of PRACTICAL
Completions GUIDELINES

WORKED
EXAMPLE

PROCEDURE

BACKGROUND

Figure 1.5 – Key Topics

Page 9
How to use the Manual – Chapter 1 BP Exploration
Completion Design Manual

Each chapter contains the subdivided Key Topic supported as necessary by the
associated Subsidiary Topic(s)

b SUBSIDIARY TOPIC

a SUBSIDIARY TOPIC

CONTACTS,
REFERENCES,
ETC

SUMMARY AND PRACTICAL

GUIDELINES

WORKED EXAMPLE

TECHNIQUES

THEORY

Figure 1.6 – Structure of Complete Chapter

Page 10
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual CONTENTS

BACKGROUND

INTRODUCTION 1

IMPORTANCE OF ESTABLISHING OBJECTIVES 4

DIFFICULTIES IN ESTABLISHING PRECISE AND 5
AGREED OBJECTIVES

DEVELOPING AN SOR – CONSIDERATIONS 7

INTRODUCTION TO THE SYSTEMS APPROACH 7

TECHNIQUES

INTRODUCTION 13

PROCEDURE FOR DEVELOPING AN SOR 14

PHASE 1: REVIEWING DATA BASE AND ESTABLISHING 14
INITIAL OBJECTIVES
PHASE 2: INTERDISCIPLINARY DISCUSSIONS 19
PHASE 3: ANALYSIS OF THE COLLECTED ERC AND 20
EXAMINATION OF ALL OBJECTIVES
PHASE 4: RECONCILIATION AND MODIFICATION OF 21
INITIAL SOR
PHASE 5: WRITE THE FINAL SOR 21
Developing a Statement of Requirements – Chapter 2 BP Exploration
CONTENTS Completion Design Manual

WORKED EXAMPLE

MILLER COMPLETION SOR 23

PHASE 1 23
PHASE 2 25
PHASE 3 26
PHASE 4 27
PHASE 5 27

SUMMARY AND PRACTICAL GUIDELINES

SUMMARY 31

CONTACTS, REFERENCES, ETC

CONTACTS 33
REFERENCES 33
BIBLIOGRAPHY 33
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

INTRODUCTION

The two initial stages in completion design, namely,

developing a statement of requirements (SOR) and conceptual
completion design are critical in achieving an optimum
design. Stated in simple terms, developing an SOR is the
process of defining the problems and establishing the
achievable aims, whereas conceptual design is the process of
translating these aims into the basis for detailed design. The
way these two initial stages should tie in with field or well
development planning is shown in Figure 2.1.

Prior to commencing a systematic completion design, it is

important to establish the realistic aims for the well. This key
step is often overlooked, and without any valid objectives or
with only vague requirements, it is impossible to optimize or
evaluate the completion design. It is important to recognize
that the aims for an initial development well are often very
different from the aims for a later infill well or subsequent
workover, and these aims should be reviewed as the
environment changes.

Developing an SOR is the process of taking the initial

objectives from all the disciplines and determining whether
these objectives can be achieved within the environment and
constraints. In carrying out this process, it is important to
determine whether the perceived constraints are in fact real
or valid.

An SOR is developed from a list of detailed objectives,

additional data requirements, resources, environment factors
and constraints for a given well completion. The SOR
considers all aspects of production operations and reservoir
depletion, both present and future, as determined by the
exploration, reservoir, drilling, completion, and production
(including field operations and process) groups that are
involved in the exploitation of the field. Designing the
completion from an SOR helps to ensure that the well can
play an effective role in the depletion of the reserves.

To assist in this, a procedure has been developed which

provides a logical and coherent method of listing the
parameters influencing completion design. This system is
(*Reference 2.1)
based on the classification developed by Patton and Abbot*.
In classifying parameters in this way, it is possible to identify

Page 1
Developing a Statement of Requirements – Chapter 2 BP Exploration
BACKGROUND Completion Design Manual

Figure 2.1 – Initial Stages of Completion Design

Page 2
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

the factors which need to be controlled or modified in order

to␣ meet the optimum objectives. A proforma, in which each
discipline can list the parameters in terms of objectives,
environment, constraints and resources has been
established. The completed proformas are then used by
the␣ completion group to establish an agreed SOR.

Page 3
Developing a Statement of Requirements – Chapter 2 BP Exploration
BACKGROUND Completion Design Manual

IMPORTANCE OF ESTABLISHING OBJECTIVES

The fundamental design goals for a completion are to:

• Achieve a desired (optimum) level of production

or␣ injection.

• Provide for adequate maintenance and surveillance

programmes.

• Be as simple, reliable and safe as possible.

• Be as flexible as possible for future operation or changes

in well duty.

• Effectively contribute its role in the reservoir development

plan in conjunction with other wells.

• Minimize initial capital costs and operating costs.

The overall objective for any completion is to safely maximize

long-term profitability. This goal usually means balancing
the requirement for maximum production against constraints
such as: cost, well location, special material requirements,
reservoir limitations, production problems, well treatment
and workovers.

While minimizing initial costs is clearly an objective, it cannot

overwhelm the goal of long-term profitability. It should be
understood that an expensive completion option may provide
the best long-term profitability. Figure 2.2 compares the cash
flow realized for two hypothetical completions. It can be seen
that the expensive completion achieves higher positive cash
flows earlier than the completion that minimizes initial costs.
Additionally, the cash flows remain higher for a longer time.
That is, the goal of long-term profitability is often realized by
the more costly completion.

Page 4
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

Figure 2.2 – Designing a Completion for Long-term Profit

DIFFICULTIES IN
ESTABLISHING
PRECISE AND
AGREED
OBJECTIVES
Optimizing completion design requires the best estimate
available of the individual well behaviour throughout its life.
Traditionally, individual well performance, production
problems and characteristics have not been addressed at an
early enough stage in the field development plan.

This situation is created by a combination of factors. The

reservoir engineering group’s objective concerning field
development is different from that of the completion groups.
The reservoir department has, among other responsibilities, a
goal of describing the performance of a field as a whole,
whereas the completion group’s objective is to design a
completion that is optimized for profit throughout the field
life. Therefore, the completion group needs to evaluate the
design on a well-by-well basis.

There is often an understandable reluctance on the part of

the reservoir engineering group to ‘guesstimate’ the likely
performance of individual wells. This reluctance is based on
the understanding that there is often insufficient data in the

Page 5
Developing a Statement of Requirements – Chapter 2 BP Exploration
BACKGROUND Completion Design Manual

early stages to determine the individual well performance

with any degree of accuracy. The limitations and inaccuracies
in full field reservoir simulation often make the detailed
analysis of individual well performance irrelevant in terms of
overall reservoir performance.

As a result, the reservoir and geoscience groups set the plan

and objectives for the field and for separate wells, but are
generally not concerned with production problems, well
maintenance or detailed operations. The two groups are
largely unconcerned with details of a singular well other than
a change in well duty and major changes in production rate.
However, the completion group has to address the individual
well performance and potential problems in order to meet
field objectives.

This section demonstrates that by discussing the objectives

with the different groups involved, a realistic SOR can be
developed before the start of the completion design process.

Page 6
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

DEVELOPING AN SOR – CONSIDERATIONS

Developing an SOR is a process of taking the initial well

completion objectives, established by each of the groups
involved in the field development planning, and determining:

• Whether these objectives are valid and will result in the

optimum individual well completion.

• Whether the goals can be practically, safely and

economically achieved.

Furthermore, the completion designer must analyse the

environment and resources of the physical well system along
with a given set of constraints, and write an SOR that can
satisfy the projected performance of the reservoir.

To ensure that all the parameters influencing the completion

design are considered, a ‘systems approach’ is advocated.
INTRODUCTION
TO THE
SYSTEMS
APPROACH
The systems approach to planning well completions is a
method of solving decision problems by categorizing and
evaluating available data. Alternatives available for the
completion are considered and compared on the basis of long
term profitability for the well. Part of the systems approach
considers the uncertainties involved in implementing the
available alternatives.

Definitions
of System
Terminology
Objective
An objective defines the purpose of the system. It is also a
standard with which to compare performance. The overriding
objective of a well completion is to safely maximize the
profitability of the well throughout its life.

Environment
The environment of a system consists of the natural setting,
physical properties and administrative factors that affect the

Page 7
Developing a Statement of Requirements – Chapter 2 BP Exploration
BACKGROUND Completion Design Manual

system and cannot be controlled. For practical purposes,

environments are limited by specific boundaries. The
boundaries of a well system are generally considered to be
between the drainage radius and the wellhead choke.

However, some environmental factors occurring outside the

well system boundaries still affect the system, eg government
regulations. Examples of environmental factors within the
system are the fluid and rock properties. In order to design a
completion, a detailed knowledge of the likely environmental
factors is required. Tables 2.1 to 2.4 list the environmental
factors in a well system and indicate the various sources for
gathering the data.

Constraints
System constraints are inherent parts of the system which
can limit its performance. Some control can be exerted over
system constraints. Some examples are scale, sand
production and corrosion. The factors that create these
constraints include conditions for carbonate precipitation,
acid gases, and poor or no cementation of sand grains, all of
which are part of the environment since they cannot be
controlled. However, the results of these factors, scale, sand
or corrosion problems, can be controlled depending on the
well system and completion design.

When defining system constraints it is important to

determine whether the perceived constraints are in fact real,
or have been arbitrarily established by preconceived ideas or
biases.

Page 8
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

REQUIRED ENGINEERING DATA:

ENVIRONMENT AND DATA RESOURCES

ENVIRONMENT DATA RESOURCE

Fluid Properties: Oil, water and gas analyses

PVT, chemical composition, etc
Oil, gas, water

Rock Properties: Core analyses:

θ, k, Sw Routine/log analysis
Relative permeability Special
Wettability Special
Intergranular cementation Special
Formation solubility Laboratory testing
Formation strength Mechanical properties log
Fracture gradient Rock mechanics:
breakdown ISIP
Fracture orientation Rock mechanics:
structural geology
Grain size distribution Sieve analysis

Initial reservoir pressure Well test/RFT

Reservoir drive mechanism, Geology and production

present and future characteristics

Drainage volume properties: Transient well tests:

Pressures: pwf, p*, pskin Drawdowns
Effective permeability Flow and build-up
Reservoir continuity Multirate (flow-after flow)
Reservoir heterogeneity Isochronal
Formation damage DST
Inflow performance pR, pwf and production test
Gas deliverability Back-pressure test

Fluid entry (into or out of wellbore): Production logs (cased hole):

Type Temperature
Amount Flowmeter
Location Gradiometer
Combination

Geothermal profile Wireline temperature survey

Rate-sensitive sand production Increasing multirate tests

History: Production characteristics Production records:

Rates: oil, water, GOR Rate vs time, semi-log plot
Cum’s: oil, water, gas Rate vs cum prod., log-log
Ratios: WOR, GOR Ratios vs cum’s, log-log

History: Mechanical Well files; review of offset

Completion histories (if necessary)
Workover(s)

History: Subsurface equipment Well files; subsurface frequency

failures failure charts

Table 2.1

Page 9
Developing a Statement of Requirements – Chapter 2 BP Exploration
BACKGROUND Completion Design Manual

REQUIRED GEOLOGICAL DATA: ENVIRONMENT AND

DATA RESOURCES

ENVIRONMENT DATA RESOURCE

Trap classification Geophysical/geological data

Strike and dip of formation Geological studies/dip meter log

Rock properties: Open hole logs, mud logs,

θ, Sw, gross and net pay cuttings
Gas/oil, water/oil contacts

Clay, silt and shale types: Petrophysical analysis

mineral content X-ray diffraction

Fractures, vugs, solution channels Core analysis

Stratification, bedding planes Cores and logs

Other formations of interest

for present or future operations:
Additional hydrocarbon
bearing zones
Aquifers
Potential gas storage
Thief zones
Salt or anhydrite zones

Table 2.2

REQUIRED FORMATION DAMAGE DATA:

ENVIRONMENT AND DATA RESOURCES

ENVIRONMENT DATA RESOURCE

Clay swelling Laboratory testing

Particle plugging and tilting Flow and backflush tests

Wettability changes Special core analysis

Emulsion or water blocks Bottom hole sample of liquid

Deposits:
Produced oil: Oil, water, and treating fluid
Asphaltenes (acids/frac fluids) compatibility
Paraffins analyses and bottom hole samples
Produced water:
Scale
Salt

Potential formation damage Compatibility tests:

Liquid to liquid
Liquid to formation

Table 2.3

Page 10
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual BACKGROUND

REQUIRED DRILLING DATA:

ENVIRONMENT AND DATA RESOURCES

ENVIRONMENT DATA RESOURCE

Washouts, cavities Caliper log

Lost circulation zones Daily drilling records
High pressure zones Daily drilling records
Analogous fields

Table 2.4

Resources
System resources include money, materials, manpower,
equipment and certain system characteristics that can be
utilized and controlled to meet the system objectives. Some
resources such as reservoir pressure and temperature and
reservoir drive may seem to be part of the environment.
Resources can often be controlled, changed or improved,
whereas the environment is a given and non-controllable
resource. The more controllable and flexible resources come
from outside the system, ie materials, manpower and money.

Each of the groups must supply the completion planner

with␣ their perceptions of the environment, resources and
constraints (ERC) of the well system. These are then used as
the basis for developing an agreed SOR.

A developed SOR must state the objectives that can be

achieved with the given environment, constraints and
resources, and establish any additional data requirements.

Failure to develop the SOR based on a systems approach

with input from all groups involved in the field development
will undoubtedly result in a completed well that cannot
effectively fulfill its individual role in the reservoir
depletion␣ plan.

Page 11
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual TECHNIQUES

INTRODUCTION

The following procedure is recommended for developing an

SOR. Obviously, the exact nature of the procedure will
depend on the size of project and number of people and
disciplines involved. However, the main thought processes
outlined here should be applied universally. The procedure
has been split into five phases, namely:

1) Reviewing data base and establishing initial objectives.

2) Interdisciplinary discussions.

3) Analysis of collective views and objectives.

4) Reconciliation and modification of initial SOR.

5) Write the final SOR.

Page 13
Developing a Statement of Requirements – Chapter 2 BP Exploration
TECHNIQUES Completion Design Manual

PROCEDURE FOR DEVELOPING AN SOR

PHASE 1:
REVIEWING
DATA BASE AND
ESTABLISHING
INITIAL
OBJECTIVES
The first step in developing the SOR is to understand the
initial objectives of the client (geoscience and reservoir
departments).

Table 2.5 illustrates the origin and process of developing a

well completion SOR once the initial objectives of the client
are identified.

After the client objectives are known, the next step is to have
each discipline involved make a thorough review of the
available data base and set its initial objectives independent
of the other disciplines. This phase includes:

• Review of existing data (see Tables 2.1 to 2.4 for potential

data resources):

– Determine what data is available.

– Determine its accuracy and reliability and whether

it␣ must be extrapolated or applied to other parts of
the␣ field.

– Ascertain the risks of conditions changing from well to

well; ie, what is the degree of reservoir heterogeneity?

• Determining additional data requirements.

• Determining the need for further studies or analyses, or

setting the objectives for designing appraisal well tests.

• Categorizing, per systems approach, all the parameters as

each discipline perceives them, assisting the completion
engineer in reconciling the objectives and validating the
constraints.

Page 14
 Completion Design Manual
BP Exploration
PHASES FOR DEVELOPING AN SOR
PHASE 1 2 3 4 5
COMPREHENSIVE DISCUSSIONS ANALYSIS FEEDBACK WRITE SOR
REVIEW OF PREVIOUS BETWEEN PLANNING
EXPERIENCE GROUPS

ACTIVITY • Exploration, Reservoir, • Each group presents: • Validate ERCs. • Meet with all groups. Write developed SOR
Drilling, Completion and Objectives, ERC. • Analyse objectives • Present initial SOR and Conceptual
Production Groups • Reconcile obvious for viability. with alternatives. Completion Plan.
establish their initial conflicts. • Identify and quantify • Discuss affects of risk
objectives. Review uncertainty. and uncertainty.
data base. • Determine if • Answer questions
• Identify good ideas, constraints are real. pertaining to SOR.
procedures. • Establish further data • Reconcile differences,
• Avoid previous requirements. make compromises.
mistakes. • Develop IPRs and TPCs
for overall well system
behaviour, considering
Table 2.5
Page 15

future conditions.
• Establish alternatives
and limitations.

Developing a Statement of Requirements – Chapter 2

RESULTS EACH GROUP: • Establish achievable • Initial SOR for • Agreed objectives,
• Develops own goals for SOR. review by well ERC: for SOR.
objectives. • Agree on role well planning groups.
• Establishes will play in field.
ERC.

GROUPS ALL ALL COMPLETION ALL COMPLETION

INVOLVED

GUIDELINES

1. The goal of this process is to arrive at an SOR, not for each group to get its way. Groups must work together.
2. Present everything that is known and germane at the Phase 2 meeting. Each group must do this. If anything is held back it will complicate things later.
3. Avoid wishful thinking: use logical reasoning.

TECHNIQUES
4. Keep uncertainty and limitations reasonable.
5. Bias and experience must be tempered to project at hand.
6. Each group should realize that SOR is a compromise of everybody's ‘must haves’, ‘wants’ and ‘would like to haves’.
Developing a Statement of Requirements – Chapter 2 BP Exploration
TECHNIQUES Completion Design Manual

• Evaluating how critical parameters and conditions will

change in the future; eg, reservoir pressure, WOR, GOR,
corrosion, sand, scale, emulsions, wax, need for artificial
lift, well duty and workover and recompletion
requirements.

• Establishing the group’s initial objectives for the

particular well completion based on their analyses of
the␣ ERC.

The above re-evaluation will be done with a view towards

identifying good ideas and procedures used before and
incorporating them into the new project, and avoiding
previous mistakes.

This review will include data from the discovery well and any
appraisal/development wells in order to establish objectives
for further drilling and recommendations for any future work.

At the conclusion of Phase 1 the various groups involved

should have the following documents for presentation in
Phase 2. The parameters specified under each classification
are examples to illustrate how each group may view the
project. The list cannot be comprehensive because what may
be a constraint in one situation could prove to be a resource
in another.

From
Geoscience␣ and
Reservoir Groups
Statement of Objectives
Producing rate
Bottom hole location
Depth to zone(s) of interest
Plateau rate
Future recompletion of other zones
Potential change of well duty in future

Statement of Environment
(Give known values or best estimates possible, along with
assumptions)
Depth to zone(s) of interest
Trap classification(s)
Strike and dip of formation(s)
Rock properties (porosity, permeability, thickness, wettability
and saturations)

Page 16
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual TECHNIQUES

Fluid properties (analyses: PVT, water, gas):

Corrosive potential (CO2, H2S)
Paraffin potential
Asphaltene potential
Scale potential
Hydrate potential
Lithology:
Formation solubility
Clay, silt and shale types
Mineral content
Fractures, vugs, solution channels
Stratification, bedding planes

Statement of Resources
(Give best estimates, along with assumptions)
Reservoir drive
Reservoir pressure and temperature
General reservoir description (as expected):
thickness/permeability/porosity/etc.
Reserves: OOIP, OGIP
IPR and/or deliverability of subject well
Alternative completions

Statement of Constraints
Maximum offtake rates
Minimum bottom hole flowing pressure
From
Drilling␣ Group
Statement of Objectives
Well cost and duration
Well type, eg land, platform, subsea
Surface location
Deviation, kick-off point, measured depth,
horizontal displacement
Casing programme

Statement of Environment
Temperature and pressure profiles
Fluid properties
Thief zones
Aquifers
Salt or anhydrite sections

Page 17
Developing a Statement of Requirements – Chapter 2 BP Exploration
TECHNIQUES Completion Design Manual

Statement of Resources
Drilling budget
Type of rig

Statement of Constraints
Cost
Time
Drilling problems, eg hole stability
Environmental factors
Governmental regulations
From
Completion
Group

Statement of Objectives
Well production rate
Field production profile
Completion life
Type of well, eg platform, subsea etc

Statement of Environment
Fluid properties (as above)
Geothermal gradient
Reservoir pressure and temperature
Other zones, aquifers etc

Statement of Resources
Budget
Alternative completions
Artificial lift
Workover and well servicing
Well inflow performance
Reservoir drive
Stimulation

Statement of Constraints
Fluid properties (as above)
Formation damage
Rate sensitive sand production
Governmental Regulations

Page 18
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual TECHNIQUES

From Field
Operations/
Development
Statement of Objectives
Field production profile
Number of wells
Type of wells, eg platform, land etc
Total project cost

Statement of Environment
Surface pressure and temperatures
Fluid properties

Statement of Resources
Project budget

Statement of Constraints
Minimum wellhead pressure
Production facilities, eg topside weight offshore
Simultaneous operations, eg drilling and production
Governmental regulations

PHASE 2:
INTER-
DISCIPLINARY
DISCUSSIONS
Following Phase 1, the necessary discussions between the
exploration, reservoir, drilling, completion, and production
groups can begin. At these meetings each group presents
their objectives and ERC as they have specified them.

From these presentations, the completion designers can

combine the ERC from each group and reconcile obvious
conflicts in objectives and establish a broad outline for the
initial SOR. After the reconciliation of any conflicting
objectives, what remains are the proposed workable
objectives the completion design group uses to develop an
SOR. At this time the combined group will agree on the role
this completion is intended to play in the reservoir
development plan.

Page 19
Developing a Statement of Requirements – Chapter 2 BP Exploration
TECHNIQUES Completion Design Manual

PHASE 3:
ANALYSIS OF
THE COLLECTED
ERC AND
EXAMINATION
OF ALL
OBJECTIVES
In Phase 3 the completion design group takes the information
resulting from Phase 2, and compares the data and
assumptions with what is known of the ERC from the most
recent well(s). The comparison is made to ensure that the
most relevant data is used and estimates provided from
previous wells are␣ applicable.

In the process of developing an initial SOR, the clients’

goals␣ will be studied to determine whether they can be
economically and effectively achieved within the ERC. This is
accomplished by validating or analysing the ERC submitted
by the geoscience and reservoir groups.

A similar validation and analysis effort is performed on the

ERC submitted by the drilling and operation groups.

After these reviews, the completion designer will be able to

determine which of the perceived ERCs, submitted by the
various groups, are applicable to the completion design.

In addition, the completion designer will establish further

data requirements to be included in the SOR, such as
additional coring, drill stem tests (DSTs), more open hole
logging suites, sample analyses, cased hole testing and
logging, to be included in the current project.

It is essential that the appraisal drilling and testing

programmes are driven by the requirements for sufficient,
accurate data which can be used to describe the reservoir
and well behaviour in order to arrive at an optimum
completion design.

Using information from DSTs and/or cores, inflow

performance (IPR) of the well can be estimated (see Chapter␣ 4
– Determining the Optimum Well Performance). The IPR is
used with tubing performance curves␣ to optimize tubing size
for production.

Page 20
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual TECHNIQUES

Having progressed through the above procedures in this

phase, the completion designer is ready to write the initial
SOR in which will be itemized:

• The achievable objectives that have survived the

re-evaluation process.

• The well system environment as it is now understood.

• The resources now available, both system and outside.

• The meaningful constraints.

• The recommended tubing and casing sizes.

Additionally, any risks and uncertainties need to be pointed

out so all the planners will have a comprehensive knowledge
of all the requirements and significant limitations of the
completion. At this point, all the groups must agree to these
requirements and limitations prior to starting the conceptual
design.

PHASE 4:
RECONCILIATION
AND
MODIFICATION
OF INITIAL SOR
WITH THE
OTHER GROUPS
During this phase the completion designer presents the
initial SOR along with alternatives, risks and uncertainties.
The presentation will cover what effects these risks and
uncertainties may have on the completion and what
requirements and limitations are acceptable while still
satisfying the SOR. These discussions will also consider
probable well life and well duty, safety procedures, workover
philosophy, production goals and limitations, well operation
and regulatory requirements.

PHASE 5: WRITE
THE FINAL SOR
The completion group agrees upon the content and writes the
final SOR. This final SOR forms the basis for the conceptual
completion design.

Further, the SOR must establish the criteria used to

determine the optimum completion.

Page 21
Developing a Statement of Requirements – Chapter 2 BP Exploration
TECHNIQUES Completion Design Manual

The final SOR contains:

• Statement of well objectives as agreed by all groups.

• Statement of ERC as agreed.

• Statement of additional data required.

• Statement of applicable safety, environmental and

regulatory requirements.

• Statement of budgetary and schedule requirements.

Page 22
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual WORKED EXAMPLE

MILLER COMPLETION SOR

The following example illustrates the SOR development

process for a production well in the Miller field in the
North␣ Sea.
PHASE 1
Tables 2.6, 2.7 and 2.8 present the objectives and ERC data
from each group from Phase 1. Note that the other items
discussed in this phase are also presented in the tables.

GEOSCIENCE GROUP OBJECTIVES AND ERC FROM PHASE 1

OBJECTIVES
Producing rate: 17 000 stb/d

Bottom hole location: Northwest section of reservoir and at least the same distance from
the platform as the discovery well, approximately 2.5 km.

Depth to zone of interest: 3970 m TVD subsea

Plateau rate: 14 000 stb/d

Future recompletion of
other zones: No specification

Potential well duty

and change: No specification

Determine formation characteristics in this area of reservoir

ENVIRONMENT
Trap classification: Stratigraphy (sandstone fan)

Strike and dip: Strike S 20 E, dip 4 deg (estimated on bottom)

Refer to geological report for complete geological discussion

Rock properties: Gross thickness 70 to 90 m, good porosity and permeability, both

of which are believed to improve to the northwest.

Lithography: Jurassic submarine fan sandstones of the Brae formation, poorly

sorted, well cemented, fine-grained to conglomerate with expected
inter-fingered siltstones and shales on the edges of the field –
predominantly montmorillinite clays and some illite and bentonite.
No significant secondary porosity or fractures are expected. Little
stratification in the centre of the fan, becoming significant
toward lateral boundaries.

RESOURCES
Refer to the BP Drilling Proposal

Table 2.6

Page 23
Developing a Statement of Requirements – Chapter 2 BP Exploration
WORKED EXAMPLE Completion Design Manual

RESERVOIR GROUP OBJECTIVES AND ERC FROM PHASE 1

OBJECTIVES
Producing rate: 13 000 stb/d

Bottom hole location: One (1) km NNE of the platform. This location should penetrate zone
up dip of OWC

Depth to zone of interest: 4020 m TVD subsea

Plateau rate: 13 000 stb/d

Future recompletion of
other zones: None

Potential well duty

and change: Future injector late in life

Require subsurface PVT sample and cores

ENVIRONMENT
Reservoir pressure
and temperature: 6100 psi and 250°F at 4085 m subsea

Rock properties: Logs and preliminary side wall core (SWC) analysis show porosity of
18 to 21% and permeability of 30 to 600 md, average of 200 md

Fluid properties: Oil, water and gas properties and analyses (recombined sample from
well 16/7b-24 as reported in the PVT report for this well). H2S
approximately 200 ppm, CO2 negligible There were problems
experienced in collecting the recombined sample, and the PVT report
is suspect.

Corrosion potential: Yet to be determined

Wax problems: No wax deposition report has been made, but an analogous field has
a wax appearance temperature of 170°F and a wax content of 13.7%.

Asphaltene potential: None

Scale potential: No valid water analysis available. First available water sample should
be analysed.

Wettability: Waiting on special core analysis from discovery well

RESOURCES
OOIP: 31 mmstb

EUR: 10.8 mmstb 23 Bcf (assuming pressure maintenance)

GOR: 2136 scf/stb from DST #3 in discovery well

PI: 19 stb/d/psi (estimated)

Natural water drive: Aquifer estimated to be 11 times size of oil field

CONSTRAINTS
Formation damage, poorer quality reservoir rock in certain areas, limited number of wells for gas injection,
insufficient data

Table 2.7

Page 24
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual WORKED EXAMPLE

DRILLING, COMPLETION AND PRODUCTION GROUPS (INCLUDING FIELD

OPERATIONS AND TOPSIDERS) OBJECTIVES AND ERC FROM PHASE 1
OBJECTIVES
Producing rate: 12 400 stb/d, 26.5 mmscf/d
Location:
Surface: Slot 3
Bottom hole: East (UTM) 406329
North (UTM) 6532264
Drilled depth: 4020 m TVD subsea
1062 m NNE of platform
4480 m MD (approximate)
KOP to be determined

Multirate tests and flowing gradient runs for IPR and TPC validation

ENVIRONMENT
Other formations of interest: 180 m salt section immediately above pay, aquifer below pay. No
thief zones. Potential disposal zone at approximately 1980 m.

Sand production: None expected

CONSTRAINTS
Formation damage: Damage experienced with fresh water muds, recommend using oil-
base mud.

Estimated flowline
pressure and temperature: 800 to 1100 psig and 200 to 240°F

Table 2.8

PHASE 2
During Phase 2 of the SOR development, the following
conflicts were resolved:

• Producing rate expected to be 12 400 stb/d, design for

15␣ 000 stb/d.

• Well location agreed to be the north-east location.

• Minor errors in categorizing systems approach parameters

were corrected, and a list of the ERC for the completion
was made and agreed upon.

• Compute analyses of DSTs of discovery well showed

permeability to average 170 md, rather than the␣ 200 md
previously reported from hand calculations on␣ location.

Page 25
Developing a Statement of Requirements – Chapter 2 BP Exploration
WORKED EXAMPLE Completion Design Manual

The use of oil-base muds remained unresolved. It was

decided that the following are the proposed achievable
objectives for the completion design group to use in the SOR:

• Producing rate is designed for 15 000 stb/d.

• Surface location: slot 3.

• Bottom hole location is: East (UTM) 406329

North (UTM) 6532264.

• Depths: TVD = 4020 m subsea

drilled = 4480 m
(estimated) (KOP to be determined)
step out = 1062 m, NNE of platform

• A subsurface PVT sample will be collected.

• Multirate tests will be included in the completion plan.

• Well duty will start as a producer and change to injector

late in life. Well design must reflect this.

• Well design to include possible recompletion for disposal

zone at 1980 m.

• Cores to be cut. Reservoir engineering to select intervals.

PHASE 3
The completion design group’s execution of Phase 3 showed
that data and assumptions used were relevant and most
estimates were reasonable. The objectives can be
economically and effectively achieved. There was no bias in
the ERC from the different groups.

Additional data required by the completion group is

as␣ follows:

• Sonic log for synthetic seismogram.

• Change in open hole log suite for oil-base mud.

• Should the well penetrate the OWC, an RFT log will

be␣ run.

• Cased hole pulsed neutron log will be run.

Page 26
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual WORKED EXAMPLE

• Accurate bottom hole temperature and geothermal profile

log to be run.

• Multirate transient well tests will be run upon completion.

• A reservoir limit test will be run.

• Measured flowing gradient tests will be run to validate one

of the vertical multiphase flow correlations available. Until
such validation is done, the Orkiszewski correlation which
works well in an analogous field will be used for sizing
tubing in this well.
PHASE 4
By the end of Phase 3 and the beginning of Phase 4,
the␣ completion group wrote the initial SOR to include
the␣ following:

• Objectives – as above with additional data requirements.

• Well system environment – as above.

• List of available resources.

• List of constraints to be discussed with other groups.

• Risks and uncertainties:

– corrosion – hydrate potential
– scaling potential – formation damage
– oil-water contact and ability to correct
– coning potential – reservoir quality
– gas toxicity and thickness

During Phase 4, discussions ensued between the disciplines

to come to agreement about the workable objectives,
additional data requirements and what could be done to
minimize the risks and uncertainties at that time.

PHASE 5
Phase 5 resulted in a written SOR that included
the␣ following:

Objectives:
• Producing rate is designed for 15 000 stb/d.

• Surface location: slot 3.

Page 27
Developing a Statement of Requirements – Chapter 2 BP Exploration
WORKED EXAMPLE Completion Design Manual

• Bottom hole location is: East (UTM) 406329

North (UTM) 6532264.

• Depths : TVD = 4020 m subsea

drilled = 4480 m
(estimated) (KOP to be determined)
step out = 1062 m, NNE of platform.

• A subsurface PVT sample will be collected.

• Multirate tests will be included in the completion plan.

• Well duty will start as a producer and change to injector

late in life. Well design must reflect this.

• Well design to include possible recompletion for disposal

zone at 1980 m.

• Cores to be cut. Reservoir engineering to select intervals.

• Sonic log for synthetic seismogram.

• Change in open hole log suite for oil-base mud.

• Should the well penetrate the OWC, an RFT log will

be␣ run.

• Cased hole pulsed neutron log will be run.

• Accurate bottom hole temperature and geothermal profile

log to be run.

• Multirate transient well tests will be run upon completion.

• A reservoir limit test will be run.

• Measured flowing gradient tests will be run to validate one

of the vertical multiphase flow correlations available. Until
such validation is done, the Orkiszewski correlation which
works well in an analogous field will be used for sizing
tubing in this well.

Page 28
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual WORKED EXAMPLE

Environment
• Trap classification: Stratigraphy (sandstone fan).

• Strike and dip: Strike S 20 E, dip 4 deg (estimated on

bottom).

• Refer to the Drilling Proposal for complete geological

discussion.

• Rock properties: Gross thickness 70 to 90 m, good

porosity and permeability, both of which are believed to
improve to the north-west.

• Lithology: Jurassic submarine fan sandstones of the Brae

formation, poorly sorted, well cemented, fine-grained to
conglomerate with expected inter-fingered siltstones and
shales on the edges of the field – predominantly
montmorillinite clays and some illite and bentonite. No
significant secondary porosity or fractures are expected.
Little stratification in the centre of the fan, but becoming
significant toward lateral boundaries.

• Reservoir pressure and temperature: 6100 psi and 250°F

at 4085 m subsea.

• Rock properties: Logs and preliminary side wall core

(SWC) analysis show porosity of 18 to 21% and
permeability of 40 to 500 md, average of 170␣ md.

• Fluid properties: Oil, water and gas properties and

analyses (recombined sample from well 16/7b-24 as
reported in the PVT report). H2S approximately 200 ppm,
CO2 negligible. There were problems experienced in
collecting the recombined sample, and the␣ PVT report
is␣ suspect.

• Corrosion potential: Yet to be determined.

• Wax problems: No wax deposition report has been made,

but an analogous field has a wax appearance temperature
of 170°F and a wax content of 13.7%.

• Asphaltene potential: None.

• Scale potential: No valid water analysis is available. First

available water sample should be analysed.

Page 29
Developing a Statement of Requirements – Chapter 2 BP Exploration
WORKED EXAMPLE Completion Design Manual

• Wettability: Waiting on special core analysis from

discovery well.

• Other formations of interest:

– 180 m salt section immediately above pay.

– Aquifer below pay.
– No thief zones.
– Potential disposal zone at approximately 1980 m.

• Sand production: None expected.

Resources
• Refer to the BP Drilling Proposal.

• OOIP: 31 mmstb.

• EUR: 10.8 mmstb, 23 Bcf (assuming pressure

maintenance).

• GOR: 2136 scf/stb from DST #3 in discovery well.

• Productivity index: 19 stb/d/psi (estimated).

• Natural water drive: Aquifer estimated to be 11 times size

of oil field.

Constraints
• Formation damage: Damage experienced with fresh water
muds, recommend using oil-base mud.

• Estimated flowline pressure and temperature: 800 to

1100 psig and 200 to 240°F.

Uncertainties
• Corrosion.
• Scaling potential.
• Oil-water contact.
• Coning potential.
• Gas toxicity.
• Hydrate potential.
• Formation damage and ability to correct.
• Reservoir quality and thickness.

Page 30
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual SUMMARY AND PRACTICAL GUIDELINES

SUMMARY
• An SOR is developed from a list of detailed objectives,
additional data requirements, resources, environment
factors and constraints for a given well completion.

• It is important to establish objectives in order to optimize

and evaluate the resulting well completion.

• Varying concerns of each group involved in providing

information for a completion SOR create difficulties in
establishing objectives.

• During the SOR development programme the designer

must determine:
– Whether these objectives are valid and will result in
the optimum individual well completion.
– Whether the goals can be practically, safely and
economically achieved.

• Systems logic is essentially a process of checking and

rechecking one’s reasoning. With this in mind, we can
outline five basic considerations that need to be
remembered when thinking about a system for SOR
development:
– The total system objective and the performance
measures of the whole system.
– The system’s environment – part of the system over
which no control can be exerted, but which impacts on
the␣ system.
– The system’s constraints – over which some control
may be exerted.
– The resources of the system that can be used to
accomplish the objectives.
– The outside resources that can be brought in to be
used for accomplishing the objectives (money,
materials, manpower).
There are many ways to think about systems but this list
is both informative and minimal.

Page 31
Developing a Statement of Requirements – Chapter 2 BP Exploration
SUMMARY AND PRACTICAL GUIDELINES Completion Design Manual

• The procedure for developing an SOR can be divided into

five major stages:

Phase 1: Reviewing data base and establishing initial

objectives

Phase 2: Interdisciplinary discussions

Phase 3: Analysis of the collected ERC and examination

of all objectives

Phase 4: Reconciliation and modification of initial SOR

with the other groups

Phase 5: Write the final SOR

Page 32
BP Exploration Developing a Statement of Requirements – Chapter 2
Completion Design Manual CONTACTS, REFERENCES, ETC

CONTACTS
Phil Murray, SPE Well Technology Studies, Aberdeen x2898.

REFERENCES
2.1 Patton, L D and Abbot, W A: Well Completions and
Workovers – The Systems Approach, 2nd Edition,
Energy Publications, Dallas, TX (1985)
BIBLIOGRAPHY
Murray, P: ‘Systematic Completion Design – the Key to
Increased Well Performance’, presented at the 1989 Joint
Engineering Technology Exchange Conference, Montgomery,
TX (April 2 to 7, 1989)

Golan, M and Whitson, C H: Well Performance, International

Human Resources Development Corporation, Boston (1986)

Nind, T E W: Principles of Oil Well Production, McGraw-Hill

Book Company, Tulsa, OK (1984)

Page 33
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual CONTENTS

BACKGROUND

BACKGROUND 1

TECHNIQUES

INTRODUCTION 3
DETERMINING THE RANGE OF WELL PERFORMANCE (CHAPTER 4) 3
NEAR WELLBORE PERFORMANCE (CHAPTER 4, SECTIONS 4b AND 5a) 5
ARCHITECTURE OF COMPLETIONS (CHAPTER 5) 6
TUBING MOVEMENT AND STRESS CALCULATIONS (CHAPTER 6) 7
SELECTION OF TUBULARS AND MATERIALS (CHAPTER 7 AND
SECTION 7a) 8
SELECTION OF COMPLETION EQUIPMENT (CHAPTER 8) 9
WELL SERVICING AND WORKOVER PHILOSOPHIES (CHAPTER 11) 9

WORKED EXAMPLE

MILLER CONCEPTUAL COMPLETION DESIGN 11

BACKGROUND 11
DETERMINING THE RANGE OF WELL PERFORMANCE 11
NEAR WELLBORE PERFORMANCE 13
ARCHITECTURE OF COMPLETIONS 14
TUBING MOVEMENT AND STRESS CALCULATIONS 17
SELECTION OF TUBULARS AND MATERIALS 18
SELECTION OF COMPLETION EQUIPMENT 19
WELL SERVICING AND WORKOVER PHILOSOPHIES 19
FURTHER WORK 20
Conceptual Completion Design – Chapter 3 BP Exploration
CONTENTS Completion Design Manual

SUMMARY AND PRACTICAL GUIDELINES

SUMMARY 23

CONTACTS, REFERENCES, ETC

CONTACTS 25
BIBLIOGRAPHY 25
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual BACKGROUND

BACKGROUND

Conceptual design involves utilizing the complete design

process outlined in the manual on a conceptual basis. At this
stage, key questions and decisions are identified and resolved
on the basis of available information. This confirms whether
the objectives in the Statement of Requirements (SOR) can be
achieved and develops conceptual solutions. Figure 3.1 shows
how the conceptual completion design should take place in
conjunction with the evolution of the field development plan.
The early analysis of the development well performance
during conceptual design is an important part of the field
development planning. The results of this analysis can have a
major impact on the production profile, number of wells,
process design and hence project economics.

The objective in the conceptual design differs from detailed

design in that the required accuracy is lower. The main aim
in the conceptual phase is to identify the risks and attempt to
quantify them to a level where sufficient confidence can be
placed in achieving the SOR. It should be remembered that in
the early stages of a project other estimates (eg capital costs)
have a relatively wide range of accuracy. The work carried out
in the conceptual design phase should be to the same level of
accuracy. The conceptual design can be regarded as the first
stage in converging on the optimum completion. The detailed
design further refines conceptual solutions based on
improved understanding and more detailed engineering. This
further refinement starting from appropriate conceptual
solutions should result in an economic and technically
optimum well completion.

The purpose of this chapter is to demonstrate how the whole

design process can be applied at a lower level to conceptual
design. Each section of the manual is taken in turn and the
design considerations, questions and data requirements that
need to be addressed in order to perform a conceptual
completion design are highlighted. This procedure is not only
appropriate for major developments, it is important to use a
similar approach, albeit on a smaller scale, for the smaller
developments. Time spent in getting the completion right will
pay dividends, especially in smaller fields where well
performance is critical to the economic viability.

Page 1
Conceptual Completion Design – Chapter 3 BP Exploration
BACKGROUND Completion Design Manual

FIELD COMPLETION SINGLE WELL

DEVELOPMENT STAGE DESIGN ACTIVITY DEVELOPMENT STAGE

Development
DEVELOPING STATEMENT Well
prospect OF REQUIREMENTS prospect

‘Defining the problems and

establishing achievable aims’

Use proforma to gather

well/field data

Discuss perspectives and

objectives with all disciplines.
Define the completion data
requirements for appraisal
drilling/further work.
Feasibility studies/ Economic
appraisal drilling feasibility studies

Use further information to

develop the SOR

CONCEPTUAL
COMPLETION DESIGN

‘Translating aims into

the basis for design’

Establishing basis for conceptual

completion design

Perform conceptual design as

per programme

Complete conceptual
completion design
Detailed well
Field plan including
development plan casing design
Review any changes to field
development plans and
assess impact on the basis
for detailed design

COMMENCE DETAILED DESIGN

Figure 3.1 – Initial Stages of Completion Design

Page 2
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual TECHNIQUES

INTRODUCTION

Although the stages in the conceptual design are presented

sequentially, many of the decisions that need to be taken are
interrelated and in practice many of the evaluations need to
be carried out in parallel. This in fact is the main justification
for performing a conceptual completion design, in that it
allows the dilemmas associated with conflicting requirements
to be understood and resolved at an early stage.
DETERMINING
THE RANGE OF
WELL
PERFORMANCE
(CHAPTER 4)

Inflow
Performance
(Chapter 4 and
Section 4a)
Conceptual Design Consideration
– Predict the range of initial ‘near ideal’ (minimum near
wellbore effects) inflow performance and determine how
this performance is likely to change with time.

Data Required
– All well test and formation evaluation data from appraisal
or analogous wells, including RFTs, cores and logs.

Questions To Be Addressed
– What is the appropriate inflow performance relationship?
– What is the likely range of inflow performance and the
impact on the project?
– How will the reservoir drive affect well performance?
– How will the initial reservoir pressure vary?
– What is the likely change in GLR and water cut?

Page 3
Conceptual Completion Design – Chapter 3 BP Exploration
TECHNIQUES Completion Design Manual

Outflow
Performance
(Chapter 4 and
Section 4c)
Conceptual Design Considerations
– In order to determine production rates, a method of
accurately modelling tubing performance is required. This
model should be used to determine the initial range of
tubing performance and predict how this performance will
change with time.

Data Required
– PVT samples and flash data.
– Flowing gradient surveys from appraisal wells or analogous
wells to validate pressure drop prediction method.
– Initial view on likely wellhead flowing pressure (from SOR).
– Desired completion life (from SOR).
– Answers to questions on inflow performance raised earlier.

Questions To Be Addressed
– Is there an appropriate method for predicting tubing
performance and what are the potential errors and their
impact on the predictions?
– What is the likely range of tubing performance over the
area of the field and with time?
Combining the
Inflow and
Outflow
Performance
(Chapter 4)
Conceptual Design Considerations
– Confirm to an acceptable range of accuracy that the rates
in the SOR can be met. Determine the likely range of well
performance over the field life.

Data Required
– Results from above inflow and outflow analyses.

Page 4
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual TECHNIQUES

Questions To Be Addressed
– Is the range of initial rates in the SOR economically and
technically feasible?
– Is the selected WHFP the economic optimum. If not, what
are the repercussions of changing it?
– What is the appropriate tubing size for the range of inflow
performance and desired production rates?
– Will this tubing size sustain flow at the required rate over
the completion life?
– Is re-completing with a smaller tubing size economically
attractive?
– Is artificial lift likely to be a requirement? If so, what is the
conceptual technique?

NEAR
WELLBORE
PERFORMANCE
(CHAPTER 4,
SECTIONS
4b␣ AND␣ 5a)
Conceptual Design Considerations
– The likely well performance determined in the previous
section was based on achieving ‘near ideal’ inflow
performance (ie near zero skin). In the conceptual phase,
the aim is to identify wells where the near wellbore
performance is likely to be less than ‘ideal’ (positive skin),
or where there is a potential for better than ‘ideal’, ie
stimulation (negative skin).

Data Required
– Appraisal well tests should be designed to acquire all the
data necessary to avoid near wellbore problems in the
development wells.

Questions To Be Addressed
– Identify the potential types of near wellbore damage,
possible stimulation methods and potential sand
production problems?
– Evaluate the sensitivity to perforating techniques?
– Identify if near wellbore performance is a critical factor in
the field development and attempt to quantify the risks?

Page 5
Conceptual Completion Design – Chapter 3 BP Exploration
TECHNIQUES Completion Design Manual

ARCHITECTURE
OF
COMPLETIONS
(CHAPTER 5)
Conceptual Design Considerations
– Based on the above work, alternative designs for the
reservoir/wellbore interface need to be developed.
– Determine the number of zones to be completed and likely
method of production, ie segregated or commingled.
– Based on the tubing stress analysis and an evaluation of
the potential well servicing and workover techniques (see
below) alternatives for the casing/tubing interface should
be evaluated.
– Provide input to casing design.

Data Required
– Formation evaluation data from previous or analogous
wells.
– Results from well performance, near wellbore performance
and tubing stress analyses.
– Required completion life.
– Brief evaluation of potential well servicing and workover
philosophies.

Questions To Be Addressed
– Do all the alternative structures satisfy all the
requirements?
– Are the completions simple, safe, reliable and flexible?
– Do the completions have the ability to cope with
unforeseen problems?
– Does the design life of the completion tie in with the
production profile forecasts? (See Table 3.1.)

Page 6
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual TECHNIQUES

WELL LIFE EXAMPLES

Completions for Long Life: > 15 Years

Subsea wells or unmanned satellite
Oil wells in large single zone fields equipped with
artificial␣ lift
Sweet gas wells with limited production problems
Injection wells with few injection problems
Multizone completions
Enhanced recovery projects

Completions for Normal Life: 5 to 15 Years

Flowing oil wells from large or over pressured zones
Gas wells and injection wells with probable production
problems (sand, water, scale, corrosion)
Oil wells equipped with gas lift or hydraulic pumps
early␣ in␣ life
Zones with limited reserves (multiple zones)

Completions for Short Life: < 5 Years

Wells with small reserves per interval (Texas Gulf Coast)
Wells with electric submersible pumps or beam pumps
Wells where economics favour routine workovers rather
than complex completions

Temporary Completions: < 1 Month

Prospect test well
Drill stem tests
Production tests (appraisal wells)

Table 3.1
TUBING
MOVEMENT AND
STRESS
CALCULATIONS
(CHAPTER 6)
Conceptual Design Considerations
– Identify whether standard stock tubulars are suitable for
the application.
– Can tubing movement be eliminated.

Data Required
– Pressures and temperatures associated with all likely
well␣ conditions.

Page 7
Conceptual Completion Design – Chapter 3 BP Exploration
TECHNIQUES Completion Design Manual

Questions To Be Addressed
– Will the selected strength of tubular satisfy all the
operating conditions with acceptable design factors?

SELECTION OF
TUBULARS AND
MATERIALS
(CHAPTER 7
AND␣
SECTION␣ 7a)
Conceptual Design Considerations
– Identify the alternative materials that can provide the
required completion life given the well conditions.
– Is the desired completion life and workover frequency
realistic?
– Evaluate the economic trade-off between corrosion
resistant materials and workover frequency.

Data Required
– Accurate samples and analyses of well fluids or reliable
data from analogous wells.

Questions To Be Addressed
– Is the data representative of conditions throughout
the␣ field?
– Is there any experience of using the same materials in
similar conditions that can be applied to the situation?
– Can material of suitable strength be made in the desired
corrosion resistant material?
– Will conditions change with time?
– Have all the fluids likely to be used been considered?
– Does the selected material and completion life provide the
lowest risk and best economic return?
– Is the lead time on the selected material compatible with
project timing?

Page 8
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual TECHNIQUES

SELECTION OF
COMPLETION
EQUIPMENT
(CHAPTER 8)
Conceptual Design Considerations
– This activity should be performed as part of the detailed
design process. However, if the completion requires special
equipment, care should be taken to ensure that the
equipment can be designed, tested and built within the
time frame of the project.

WELL SERVICING
AND WORKOVER
PHILOSOPHIES
(CHAPTER 11)
Conceptual Design Considerations
– Identify the techniques required to maintain the well
throughout its life.

Data Required
– Likely well duty.
– Likely production chemistry problems.

Questions To Be Addressed
– Are the conceptual completion designs compatible with the
well maintenance philosophy?

Having completed the analysis outlined in this Techniques

section, the final stage in the conceptual design is to
establish broad budget costs and approximate lead times.

Page 9
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

MILLER CONCEPTUAL COMPLETION DESIGN

This worked example is not intended to represent a complete

conceptual completion design as presented in a conceptual
design report, but rather to highlight the key issues and main
considerations behind the suggested design(s).
BACKGROUND
The Miller field lies mainly in block 16/7b and 16/8b of the
UKCS. The reservoir is an Upper Jurassic submarine fan
sandstone formation with an oil-water contact at 4090 m
subsea. There are plans for 31 wells, of which 11 would be
injectors.

The reservoir fluid has an overall three stage flash GOR of

about 1895 scf/stb and contains 11% by weight of CO2 and
up to 750 ppm of H2S in the gas phase. Corrosion resistant
alloy (CRA) materials will be required to avoid excessive CO2
corrosion, hydrogen embrittlement and sulphide stress
corrosion cracking.

The formation water contains high levels of bicarbonates

which will, with reduction in pressure at the wellbore and up
the tubing, result in calcium carbonate scale deposition. In
addition, the high levels of barium will result in severe
barium sulphate scaling. The potential for scaling in Miller is
supported by the experiences in the neighbouring South Brae
field.

DETERMINING
THE RANGE
OF WELL
PERFORMANCE

Inflow
Performance
The reservoir performance was based on a reservoir
simulation model, and the productivity indices, PIs, were
found to range from less than 10 stb/d/psi to almost
200 stb/d/psi.

Normally, sufficient data is not available at the conceptual

design stage to get ‘meaningful’ IPRs from a full field reservoir
simulator. In these cases, more reliance must be placed on
estimates using the radial flow equation and maps of

Page 11
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

permeability and net pay prepared by the geologists.

However, a large number of appraisal wells were drilled and
tested on Miller, enabling the reservoir simulator to be used
to estimate both the initial IPRs and the variation of inflow
performance with time as a function of number of wells,
reservoir pressure, water production, etc.

A ‘near ideal’ skin value, equal to the lowest experienced in

appraisal well DSTs, was assumed for all wells. However,
production problems like water cut and uncontrolled scaling
will have severe detrimental effects on the well productivities.

Initial reservoir pressure in Miller was 7300 psi, but due to

depletion in South Brae which has a common aquifer,
pressure is expected to be around 6100 psi at the time of
Miller start-up. By commencing water injection at high rate
simultaneously with first production, it is expected that
reservoir pressure will increase rapidly to approximately
7000␣ psi.

The water flood is expected to encroach from the periphery

rather than from the underlying aquifer, and the low oil
viscosity gives a favourable mobility ratio. These factors imply
a rapid rise in water cut following water breakthrough.
Outflow
Performance
Lift curves were generated for 4 1/2in, 5 1/2in and 7in
tubing, with a minimum wellhead flowing pressure of
800␣ psia and water cuts up to 80%.

The Orkiszewski multiphase flow correlation with oil

continuous phase at all water cuts, was used to create the lift
curves. Only limited and inaccurate flowing pressure data
was available to validate the flow correlations, and hence this
data was discarded. However, based on past experience in
similar conditions, the Orkiszewski correlation was thought
to have the best chance of predicting appropriate lift curves.
Generally, measured flowing gradient surveys should be run
at several rates in order to validate the flow correlations.

PVT properties were estimated using black oil correlations

tuned to available multistage flash data. This resulted in PVT
parameters very similar to those generated using
compositional methods.

Page 12
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

Combining
the␣ Inflow
and␣ Outflow
Performance
The decision on tubing size depends on the initial
productivity index, the timing of water breakthrough and the
decline of productivity with increasing water cut.

By superimposing the productivity index lines on the

generated lift curves, expected well production rates over the
completion life can be obtained. As a result, it was found that
7in tubing was the optimum tubing size for wells having a PI
of 30 stb/d/psi or more, whereas 5 1/2in was suitable where
the PIs were less than 30 stb/d/psi.

For wells having lower productivities, it was found that

reducing the tubing size from 5 1/2in to 4 1/2in would
prolong well life at high water cuts. However, it could act as a
restriction when the wells are dry, resulting in significant oil
production being deferred. Hence, only 5 1/2in and 7in
tubing were recommended as the optimum tubing sizes.

The 5 1/2in completions might be choked back to maintain

bottom hole flowing pressure above the bubble point pressure
(pb = 4900 psi) when the productivity declines. This should,
however, not be done until after the time that the reservoir
has been repressurised (pR = 7000 psi).

Water injection wells should be completed with 7in tubing

strings in order to achieve sufficient injection rates without
excessive pressure drops and tubing velocities.

No provision for artificial lift was required.

NEAR
WELLBORE
PERFORMANCE

Formation
Damage
Mechanism
The main near wellbore performance problem in Miller is
believed to be calcium carbonate and barium sulphate scale
deposition in and around the perforation tunnels, although
the exact mechanisms are not yet fully understood.

Page 13
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

Both salt water, fresh water and oil based mud have been
used to drill the reservoir section in the appraisal wells. The
experience so far shows a trend towards smaller skin factors
where oil based mud was used. This indicates that part of the
formation damage problem in Miller may be barium sulphate
scale deposition in the near wellbore area.
Perforating
Method
A number of perforating options were investigated. The need
for underbalanced perforating to minimize the perforating
skin, and hence the pressure drop across the perforations,
was evident to reduce the amount of calcium carbonate scale
being deposited in the perforations.

The use of tubing conveyed perforating (TCP) guns run with

the completion, requires the drilling of 100 to 150 m of sump
to accommodate the spent guns. This was rejected mainly
due to the drilling damage risk of weighting up mud to drill
the sump whilst the reservoir zone pressure was itself
depleted. The option of using underbalanced TCP on a
temporary string requires the killing of the well after
perforating in order to run the permanent completion. This
option is dependent on a non-damaging kill fluid being
identified.

The agreed optimum perforating method for Miller was,

therefore, to perforate underbalanced with TCP on a
temporary string, with through-tubing wireline perforating
retained as back-up.
ARCHITECTURE
OF
COMPLETIONS
The main criteria which influenced the overall design
philosophy behind the Miller production completions were:

• Minimum number of restrictions to reduce scaling

tendencies.

• Minimum wireline interventions.

• Ability to perform through-tubing workovers.

• Pressure integrity of the completion string to prevent

corrosion of the carbon steel 9 5/8in casing by reservoir
fluid.

Page 14
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

In addition, the completion designer should always try

to␣ keep the completion as simple, safe, reliable and flexible
as␣ possible.
Reservoir/
Wellbore
Interface
In order to maintain flexibility to shut off identified water
producing zones and hence reduce scaling, perforated
completions appeared to be the obvious choice for Miller.

No problems were expected with commingled production

since no potential pressure barriers, ie thick extensive shales,
had been identified within the reservoir.

A composite 10 3/4 – 9 5/8in casing string will be run,

terminating above the reservoir and with a liner through the
reservoir section. The 10 3/4in section of casing is required
to accommodate 7in tubing retrievable downhole safety
valves.

To limit the effects of scale, the completion must:

• Not encourage scale deposition.

• Allow water producing intervals to be shut off or

chemically inhibited.

• Facilitate scale removal as easily as possible.

Thus, internal diameters will be maximized to provide full

bore, and hence allow large scale removal tools and bridge
plugs to be run into the liner through the completion. With
this in mind, 6 5/8in and 5in liners were proposed for the 7in
and 5 1/2in producing wells respectively. In addition, the full
bore will reduce scaling tendencies in that localized pressure
drops will be minimized.

There is no requirement or advantage in having a liner which

is stepped down in size from the tubing size for the water
injection wells. Therefore, 7in liners will be suitable for the
injectors.

A tailpipe with gauge hanging nipple that extends into the

liner was found to be inadvisable since it would restrict the
completion diameter and could risk cementation of the

Page 15
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

tailpipe in the liner by scale. In addition, the nipple is likely

to be unusable once scaling commences.
Casing/Tubing
Interface
A moving seal completion is required to accommodate the
estimated tubing movement resulting from the expected
service conditions.

In order to maximize the internal diameter through the

completion and to ease wireline access, a PBR/seal assembly
will be preferred to an overshot/seal assembly. The seals and
PBR will be run pinned together to protect the seals during
installation in the well.

Completion string pressure integrity is particularly important

in order to protect the 9 5/8in casing from corrosive reservoir
fluids. The extra cost of running CRA material for the
complete 9 5/8in casing is unacceptably high and should be
avoided if possible. However, completing the lower 30 m of
the 9 5/8in in CRA material is recommended, although the
extra cost may again make this unattractive.

Tubing pack-off can be obtained using a permanent

hydraulically set packer with a long PBR assembly anchored
into the packer. The packer is run close to the top of the liner
by having a short set of static seals below the packer which
sting into a short PBR incorporated at the top of the liner. By
completing this way, the packer, anchor latch, long PBR and
dynamic seals can all be run pinned together with the seals
protected, and the packer can be set without having to
expose the upper moving seals. However, the use of a packer
leaves a dead space in the 9 5/8in annulus below the packer,
which cannot be monitored for leaks of corrosive reservoir
fluid to the 9 5/8in.

The packer could be omitted to avoid this dead space by

incorporating an anchor in the lowermost PBR and seals.
This would, however, involve separation of the long PBR and
dynamic seals during completion installation, which would
compromise the philosophy of maximizing protection of
these␣ seals.

Page 16
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

TUBING
MOVEMENT
AND STRESS
CALCULATIONS
The load cases expected to give maximum stresses or tubing
movements are:

• Plugged tubing pressure test to 5000 psi.

• Annulus pressure test to 5000 psi.
• Production well treatment or water injection at 7500 psi.
• Production at maximum rate.
• Tubing retrieval.

The maximum injection and well treatment pressure of

7500␣ psi will be backed up by an annulus pressure of 3500
psi so that the tubulars and completion components above
the packer can be 5000 psi rated. A tubing retrieval load case
was not considered pending more information. Calculations
were performed for:

• 5 1/2in, 23 lb/ft tubing with 6.000in seal bore.

• 7in, 35 lb/ft tubing with 7.375in seal bore.

The stress levels were calculated using the extremes of the

pressures and temperatures expected for the various load
cases, and the higher value of thermal expansion for fully
austenitic material was used.

Maximum tubing movements calculated were as follows:

• 5 1/2in : 37 ft (+13 ft and -24 ft)

• 7in : 31 ft (+11 ft and -20 ft)

It is impractical and inadvisable to allow this range of free

tubing movement. Although it cannot be eliminated
altogether in Miller, it can be reduced by changing the
installation procedures, reducing the seal bore diameter or,
more significantly, applying slack-off weight.

The tubing stress calculations performed to date indicate

that␣ the highest stress is found at the surface end of the
tubing during annulus pressure testing. However, application
of slack-off will increase any compression at the bottom of the
tubing, and this might therefore become the highest stressed
area.

Page 17
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

When analysing CRA material tubulars, effects like

temperature dependent yield stress and anisotropic
mechanical properties should also be considered. For Miller,
these effects must be fully evaluated before the tubing stress
analysis is finalized and a recommendation on suitable
tubulars is issued.
SELECTION OF
TUBULARS AND
MATERIALS
Corrosion of the downhole tubulars will be kept to a
minimum by the use of suitable materials, rather than by use
of chemical injection techniques. The selected tubulars must
therefore be capable of dealing with the corrosive reservoir
fluids, ie high levels of CO2 and H2S.

In production wells, the 6 5/8in or 5in liners and the

components in the liner hanger and pack-off will be of
corrosion and stress cracking resistant alloys.

For the tubing material, it is expected that the choice will be

between duplex alloys and fully austenitic alloys, depending
on corrosion studies at RCS. Availability of various strengths
for any suitable CRA material tubulars is not considered to
be a problem.

The tubing accessories will probably be manufactured from

Inconel 718 or Incoloy 925 due to their superior machining
properties compared to chrome steel.

By choosing these expensive CRA materials for both the

tubulars and tubing accessories, the completion life will
be␣ maximized.

Back-flow in water injectors is not considered to be a

problem, and hence the tubulars for the water injectors
may␣ be manufactured from standard service material.

Internally flush premium couplings (eg VAM) should be used

for all tubulars exposed to the reservoir fluid.

The decision on which elastomers are suitable for the

corrosive environment and resistant to the chemical
inhibitors is pending results from RCS. However, Marathon
have had good experience with Kalrez, Teflon and Peek in
similar conditions on South Brae.

Page 18
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

The merits of coated tubulars should also be considered.

SELECTION OF
COMPLETION
EQUIPMENT
This activity should be carried out in the detailed completion
design phase. However, with conditions like Miller, if special
equipment or equipment material is needed, it needs to be
addressed at an early stage to ensure enough time is
available to design, manufacture and test the special item(s).

Tubing retrievable downhole safety valves (TRSCSSV) are

recommended for Miller completions. Tubing retrievable
valves have no external packings to fail and have proven to be
more reliable in the past than wireline retrievable valves.
They also offer a minimal restriction to flow because of their
large through-bore, an important feature in view of the
scaling potential in Miller. In the event of a failure of the
tubing retrievable valve, a lock-out mechanism and nipple
profile would allow the landing of a wireline valve as back-up.

The valve can have an equalization facility, but it is

recommended this is not incorporated in order to increase
reliability.
WELL SERVICING
AND WORKOVER
PHILOSOPHIES
Wireline intervention will be kept to a minimum for the
following reasons:

• Extremely corrosive reservoir fluid.

• The wells are generally deep.

• Some wells are highly deviated.

• There are potential scaling problems.

• The size and weight of 7in toolstrings.

However, some wireline work will be needed for reservoir

monitoring purposes.

Annulus injection of scale inhibitor would complicate the

completion design significantly and would also increase
costs. This method of inhibition would not protect the near

Page 19
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

wellbore formation, the perforation tunnels nor the liner from

scale. For these reasons, no provision for continuous
downhole injection has been made in the design. The
currently preferred method of scale inhibition is by regular
squeeze treatments into the formation, using an inhibitor
which adsorbs onto the rock and slowly releases during
production. Chemicals suitable for the treatment of these
scales are being investigated by RCS.

Because of the scaling potential of the Miller formation water

and in view of the likely difficulty in finding effective scale
inhibitors, ready access to the liner to allow easy scale
removal and bridge plug setting is given a high priority.
FURTHER WORK
Three of the major areas where further work is needed are:

• Materials.
• Workovers and well servicing.
• Completion equipment.
Materials
Whichever metallurgy is selected for the tubulars, there will
be a trade off between material strength, weight and expense.

Workovers and
Well Servicing
Scale prevention measures, including zone isolation and
inhibitor squeezes, need a more detailed review. Methods to
be investigated include bridge plugs and retrievable straddle
packers, run on either wireline or coiled tubing. Limitations
due to depth, deviation, injectivity, pressure and corrosion
must be considered.

Scale removal methods also need to be reviewed. Areas to be

considered include:

• Hardness of scale to be removed.

• Evaluation of various tools such as mills and downhole

motors.

• Suitability of coiled tubing versus work strings.

• Arguments concerning well kill versus scale removal in

live wells.

Page 20
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual WORKED EXAMPLE

• Requirement and timing of full scale workovers.

• Rig time availability.

Completion
Equipment

Tubing movement reduction and any means to ensure the

integrity of the tubing completion to protect the 9 5/8in
casing require further work.

The merits of retaining or omitting a permanent packer need

further consideration.

The design for 7in tubing retrievable downhole safety valves

is still being developed and requires continual review along
with any operational experience in the near future.

Page 21
Conceptual Completion Design – Chapter 3 BP Exploration
WORKED EXAMPLE Completion Design Manual

POSSIBLE INTERNAL DIAMETERS

COMPLETION ITEM
7" COMPLETION 5 1/2" COMPLETION

TUBING HANGER 6.125 4.900

(WITH BPV PROFILE)

TUBING, 7" AND 5 1/2" 5.950 4.625

(POSSIBLY SPECIAL DRIFT)

NIPPLE 5.875 4.562

(FOR WIRELINE INSERT VALVE)

TUBING RUN DHSV (DEEP SET) 6.000 4.562

(9.5" AND 8.375" OD)

10 3/4" x 9 5/8" CASING CROSSOVER

TUBING, 7" AND 5 1/2" 5.879 4.545

NIPPLE 5.813 4.437

LONG SEAL MANDREL 6.004 4.750

POLISHED BORE RECEPTACLE 7.375 6.000

ANCHOR LATCH (OPTION) 6.031 4.875

PERMANENT PACKER (OPTION) 6.000 4.750

NIPPLE 5.750 4.312

ANCHOR LATCH (OPTION)

SEAL MANDREL 6.004 4.750

POLISHED BORE RECEPTACLE 7.375 6.000

LINER HANGER AND PACKER

9 5/8" SHOE

LINER SHOE, 6 5/8" AND 5" 5.666 5.666

THE REQUIRED NIPPLES ARE INDICATED IN THE COMPLETION, ALTHOUGH THE SELECTION
OF NIPPLES AND NIPPLE SIZES NORMALLY BELONGS TO THE DETAILED COMPLETION DESIGN PHASE.

Figure 3.2 – Miller Conceptual Completion Design (7in and 5 1/2in, 5000 psi)

Page 22
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual SUMMARY AND PRACTICAL GUIDELINES

SUMMARY
The early analysis of development well performance as part of
the conceptual completion design is an important part of the
field development planning. This analysis can have a major
impact on the production profile, number of wells, process
design and hence project economics.

The potential of appraisal well tests to provide data on near

wellbore and tubing performance should be fully exploited. A
completion design is only as good as the input data, and care
should be taken to ensure a high level of accuracy.

The work performed in the conceptual design should have the

same level of accuracy as the field development plan.

Within the constraints of the well duty, completion life and

well maintenance requirements, the completion should be
kept as simple as possible.

The conceptual design should consider using as much of the

company’s inventory of supplies and materials as possible,
but not to the extent of constraining the optimum completion.
For example, do not use extra heavy wall J-55 casing simply
because it is available from inventory and will take the same
tensile load as thinner walled N-80. The initial cost savings of
using inventory stock may haunt the production department
for the life of the well when they cannot run the proper size
pump through the heavy walled casing.

Consider the operating costs associated with the conceptual

completion design. Design in flexibility, minimum moving
parts and maximum safety. Design out any unnecessary
complexity and maintenance problems.

Page 23
BP Exploration Conceptual Completion Design – Chapter 3
Completion Design Manual CONTACTS, REFERENCES, ETC

CONTACTS
Phil Murray, SPE Well Technology Studies, Aberdeen x2898.
BIBLIOGRAPHY
Patton, L D and Abbot, W A: Well Completions and Workovers
– The Systems Approach, 2nd Edition, Energy Publications,
Dallas, TX (1985)

Golan, M and Whitson, C H: Well Performance, International

Human Resources Development Corporation, Boston (1986)

Nind, T E W: Principles of Oil Well Production, McGraw-Hill

Book Company, Tulsa, OK (1984)

Allen, T O and Roberts, A P: Production Operations – Well

Completions, Workover and Stimulation, Vol 1 and 2, Oil &
Gas Consulatants, Tulsa, OK (1989)

Murray, P: ‘Systematic Completion Design – the Key to

Increased Well Performance’, presented at the 1989 Joint
Engineering Technology Exchange Conference, Montgomery,
TX (April 2 to 7, 1989)

Page 25
 Chapter 4 - Well Performance

BP Exploration

Jonathan Bellarby

Independent Completions Engineering Ltd.

20 October 1998
 BP Completion Design Manual - Section 4

Contents
4.1 Introduction .......................................................................................................................3
4.1.1 The Definition of Well Performance ...........................................................................3
4.1.2 Inflow Performance Relationship (IPR) ......................................................................4
4.1.3 Vertical Lift Performance (VLP)..................................................................................5
4.1.4 Overall Pressure Drops ..............................................................................................5
4.1.5 Well Design for the Entire Life of the Well..................................................................5
4.1.6 Check List for Well Performance Modelling Using Prosper .......................................7
4.2 PVT ...................................................................................................................................9
4.2.1 Direct use of Laboratory Data...................................................................................11
4.2.2 Equation of State ......................................................................................................12
4.2.3 Empirical Correlations (Untuned) .............................................................................13
4.2.4 Empirical Correlations (Tuned).................................................................................14
4.3 Inflow Performance .........................................................................................................16
4.3.1 Radial Inflow Equation and Skin...............................................................................16
4.3.2 Vogel Inflow Performance ........................................................................................17
4.3.3 Fetkovich ..................................................................................................................17
4.3.4 Jones ........................................................................................................................17
4.3.5 Hydraulically Fractured Well.....................................................................................17
4.3.6 Horizontal Wells........................................................................................................18
4.3.7 Perforation and Deviation Skin .................................................................................19
4.3.8 Multiple Zone Completions .......................................................................................21
4.3.9 Gravel Pack Completions .........................................................................................21
4.4 Tubing Performance .......................................................................................................23
4.4.1 Which Correlation Should I Use? .............................................................................24
4.4.2 Using Correlations to Predict When a Well Dies......................................................27
4.4.3 Tuning Pressure Drop Correlations ..........................................................................27
4.4.4 Natural Flow .............................................................................................................28
4.4.5 Shut-in Conditions ....................................................................................................28
4.4.6 Artificial Lift ...............................................................................................................29
4.4.7 Heat Transfer ...........................................................................................................36
4.4.8 Erosion .....................................................................................................................37
4.4.9 Tubing Size Selection ...............................................................................................39
4.5 Well and Field Optimisation ............................................................................................44
4.5.1 Practical vs. Theoretical Optimisation ......................................................................46
4.6 Glossary ..........................................................................................................................48

20 October 1998 Page 2

 BP Completion Design Manual - Section 4

4.1 Introduction
The purpose of this section of the manual is to define well performance and how it is
assessed and modelled. It has been written from a users viewpoint rather than any
theoretical perspective. The more theoretical aspects are covered in sections 4a, 4b, and
4c.
This section of the manual should be used in conjunction with the well performance
software “Prosper” manual. This section does not attempt to cover any of the mechanics
of using Prosper, rather what issues need to be addressed when attempting accurate and
representative well performance modelling. There are some areas of overlap, between the
Prosper manual and this section, however these have been included to emphasise
particular points.

4.1.1 The Definition of Well Performance

Following the review of the reservoir and assessing what is known about it, the planning
process then focuses on well design. A major component of this is the selection and
design of the completion to be used. There are a large number of tools and techniques
available to the completion engineer to address the specific properties of the reservoir and
to maximise the productivity of the planned wells. However, central to the entire well
design process is the means of defining well performance.
The classic approach to expressing well performance is a graph of the well bottomhole
flowing pressure versus the produced fluid rate. The following example is for an oil well:
Figure 1 - Typical Inflow and Outflow Plot

Inflow v Outflow Curves

7000

6300

5600

4900

4200
Pressure (psig)

3500

2800

2100

1400

700

0
0 4000 8000 12000 16000 20000
Liquid Rate (STB/day)

20 October 1998 Page 3

 BP Completion Design Manual - Section 4

Superimposed on the graph are two well properties. The blue line represents what the
reservoir will deliver to the well and is termed the inflow performance relationship (IPR).
The concept was introduced in the 1920’s with the development of the bottomhole
pressure gauge. From the gradient of this line the productivity index, or PI, can be
calculated and is quoted as stock tank barrels per day per psi drawdown. This gives a
means of comparing productivity between wells. The red line defines what the bottomhole
flowing pressure has to be in order for the well to produce at the rate on the bottom axis.
This relationship accounts for the pressure required to get the produced fluids up through
the tubing used to construct the well and is termed the vertical lift performance (VLP) or
tubing performance. If the datum point for each system is the same, the point at which the
two lines intersect defines the rate at which the well will flow and the bottom hole pressure
at this point. This is an example of nodal analysis, so called because the whole system
performance is described by two independent subsystems. The two subsystems are
upstream and downstream of a common node. It is possible for this node to anywhere in
the system, for example when looking at facilities performance as well, it is often useful to
define the node as the wellhead or where the flowline meets the manifold. Such concepts
are explored more in the well optimisation section (4.5).
Having defined a means of describing well performance, the design of the well can be
considered in order to optimise this. In this context, optimisation isn’t necessarily the same
as maximise. Certainly high rate wells can be designed, however, this may have
implications on the performance of the reservoir throughout field life that may reduce
overall hydrocarbon recovery. The key issue is that a means of reviewing well
performance is available to contribute to the overall review of the best production strategy
for the reservoir.
As shown in the graph above, there are two parameters to consider when assessing well
performance:
The Inflow into the well - described by the IPR
The outflow through the well to surface - described by the VLP
An analogous relationship exists for injection wells.

4.1.2 Inflow Performance Relationship (IPR)

The reservoir will have specific properties defined by the geological processes occurring
as it was laid down. Similarly, the hydrocarbon fluid will have properties defined by its
source and the environment in which it was converted from Kerogen. What the
completions engineer has control over is the geometry of the well, what sections are
allowed to flow, the order in which this occurs, and the selection of the interface between
the reservoir and wellbore. For example, a horizontal well may be selected with flowing
sections towards its end (toe) and beginning (heel). The toe may be produced first to
ensure maximum clean-up from these sections through a wire wrapped completion to
prevent the production of sand from the reservoir.
As well testing became common a large effort took place within the industry to define
mathematical equations that describe the IPR based on reservoir and fluid properties.
From this work, reservoir and fluid parameters have been identified that control the inflow.
In turn these can be used in order to predict the potential IPRs of new wells as a basis of
effective well design programs. This chapter will discuss the various IPR relationships
available and criteria for their applicability to a given well. How the appropriate IPR can
then be used to evaluate the potential well performance for a specific well design will then
be discussed with a proposed methodology.

20 October 1998 Page 4

 BP Completion Design Manual - Section 4

4.1.3 Vertical Lift Performance (VLP)

The VLP or tubing performance is controlled by both the properties of the produced
hydrocarbon and the construction of the well. Whilst the completions engineer has no
control over the hydrocarbon characteristics, defined by the pressure - volume -
temperature relationship or PVT, they have options to use different tubing sizes and
artificial lift methods in order to optimise production. The two most common methods of
artificial lift are using gas lift, where produced gas is injected at a series of points down the
well, or running a pump of some sort. Both methods have the effect of reducing the
hydrostatic column pressure in the well. Without this to overcome, produced fluids can be
far more easily flowed to surface.
The dimensions of the tubing that the produced fluids flow through to surface, the material
used, the degree of corrosion, and other factors such as scale build up also effect the
VLP. Clearly the completion engineer has far more control over these along with the well
head flowing pressure. By considering all of these parameters, the most appropriate well
design can be identified.

4.1.4 Overall Pressure Drops

It is useful to take an overall view of the pressure drops in the field / well. This enables the
real influences on the field / well performance to be identified. For example in a gas well in
a tight reservoir, the majority of the pressure drops are frictional in the reservoir. In an oil
well, hydrostatic pressure drops in the tubing are more important. This allows both the
effort in prediction to be focused on the areas where it maters, but also the areas where
the completion designer can have an impact on improving performance.
Figure 2 - Generalised Pressure Drops in Hydrocarbon Production

Reservoir Completion Facilities

Reservoir Sandface Choke

Pressure

Oil
Pressure

Gas

Separator
Pressure

Distance from Separator

Hydrostatic pressure (for oil wells)
Friction dominates Friction dominates
dominates, but also significant for gas wells

4.1.5 Well Design for the Entire Life of the Well

A major issue to be considered during the well and completion design process is how the
well should perform for the entire period of its operational life. Factors that impact on this
are:
• The performance of the reservoir throughout the field life. For example, pressure
support (or lack of it) or water / gas coning.
• Reservoir constraints. For example on the Harding field, the reservoir is so conductive
that rates in excess of 40,000 bpd could be achieved. However, gas and / or water
coning would occur at these rates and therefore the wells do not need to be designed
to handle these rates. Likewise sometimes bottom hole pressures are deliberately
maintained above the bubble point in order to prevent relative permeability effects.

20 October 1998 Page 5

 BP Completion Design Manual - Section 4

• The fluids handling and how this may change e.g. there is no point in designing a large
well to cope with high GORs if the wells will have to be constantly choked back
because the gas can’t be managed at surface.
• Production problems. Scales, asphaltenes, wax etc. all impact the well design and
productivity. Asphaltene deposition occurs at a certain pressure and therefore the
bottom hole pressure is maintained above this point in order to prevent asphaltene
depositing in the reservoir or at the bottom of the well (e.g. perforations).
• The likelihood of workovers. This will allow for a tubing size change or implementation
of artificial lift.

20 October 1998 Page 6

 BP Completion Design Manual - Section 4

4.1.6 Check List for Well Performance Modelling Using Prosper

Before starting out on well performance modelling, it is useful to pull together as much
information on the reservoir, the fluids and the expected conditions. The check list below
may be a useful starting point:

PVT
R Bubble point or dew point
R Oil density at standard conditions
R Formation Volume Factor at Reservoir Pressure and bubble point
R GOR (if available for below the bubble point, this must be constant composition
data) or CGR (again constant composition data is a must), alternatively CGR for
gas wells (with CGR values below the dew point, these should again be for
constant composition expansion).
R Oil viscosity at reservoir, bubble point and separator conditions (constant
composition)
R Gas gravity and viscosity
R CO2,, N2 and H2S concentrations
R Water salinity or density
R Emulsion data (viscosity vs. water cut)
R Confidence level in the PVT data i.e. sample method and information on
consistency checks
Reservoir
R Geometry - no flow boundaries and well position
R Well inclination w.r.t. reservoir
R Net reservoir penetration and how this is arranged i.e. partial penetration
R Horizontal permeability, large (reservoir) and small (inches) scale kv/kh
R Relative permeability (oil / water / gas) if available
R Reservoir net thickness
R Porosity and UCS (for perforation performance determination)
R Pressure and datum depth and any expected variation over time
R Expected water cuts and GORs / CGRs - with any variations over time
R Any formation damage tendencies or responses to well kills
Well Test Data
R Completion schematic with dimensions, deviation data and depths
R Fluid rates - water, gas, oil - how they are measured and the confidence in them -
ideally also a variety of flowrates, covering likely rates over field life.
R Surface
and downhole pressures and temperatures - ideally not just spot
measurements, but covering both ranges in time and depth
Completion
R Liner size and weight, tubing size(s) and weights
R Casing sizes and depths (especially if accurate temperature data is required)
R Completion fluid (density and whether oil or water)
R Potential restrictions (e.g. wireline retrievable sub-surface safety valves)
R Deviation Survey
R Water depth (if offshore), RT elevation
R Tubing metallurgy and condition (scale and / or corrosion) or direct roughness data

20 October 1998 Page 7

 BP Completion Design Manual - Section 4

Artificial Lift
R Gas lift - compressor discharge and wellhead pressure(s) and gas gravity
R ESP - power availability, setting depth, casing size
R Pump performance data (if available)
Surface
R Separator pressure
R Flowline dimensions and lengths (including manifolds) from tree to separator
R Flowline /manifold geometry (90° bends and the like)
R Choke sizes

20 October 1998 Page 8

 BP Completion Design Manual - Section 4

4.2 PVT
The Pressure Volume Temperature (PVT) relationship describes how a fluid behaves
under changing conditions. With an accurate PVT relationship, the density, viscosity and
gas-oil ratio for the fluid under expected pressures and temperatures can be reliably
extracted. This is then used to determine the inflow performance and more importantly the
tubing performance.
As a hydrocarbon fluid is produced, the temperature and pressure changes. These
changes will initially only cause the oil or gas to change viscosity and density. At a certain
point however the fluid will change from a single phase to two phase. For a black-oil fluid,
the gas will start to come out of solution at the bubble point. For a condensate,
condensate will start to come out of solution. For a given fluid composition, these points
(the saturation points) define the phase envelope. Outside the phase envelope, the fluid is
single phase, inside the fluid is two phase. The critical point is the point on the phase
diagram where to the left the fluid that first comes out of solution is a gas. To the right, the
fluid that first comes out of solution is a liquid. Figure 3 shows an example of a phase
envelope.
Figure 3 - Example Phase Envelope

Pressure (psig)

4000 Critical Point

Black or
3600 volatile oil 649.6 (degrees F)
Critical Point
3200 Retrograde Dry gas 3351.7 (psig)
condensate
2800

2400 Cricondentherm
882.2 (degrees F)
2000
858.3 (psig)
1600

1200
Cricondenbar
800 431.5 (degrees F)

400
3913.1 (psig)
0
-200 0 200 400 600 800 1000 1200 1400 1600
Temperature (degrees F)

Bubble Point System

The definitions of the different fluids is:

20 October 1998 Page 9

 BP Completion Design Manual - Section 4

• A black-oil or volatile oil is one where with a drop in pressure, gas will come out of
solution. The difference between a black oil and a volatile oil is purely arbitrary and
relates to the higher GOR and formation volume factor of a volatile oil compared to a
black oil. The FVF for a volatile oil will be above approximately 1.5., below this it will be
a black-oil
• A retrograde condensate is one where with a drop in pressure, liquid will first come out
of solution. Note most retrograde condensates will exhibit the behavior where as the
pressure is reduced further, the liquid may vaporise again.
• A gas is where a drop in pressure will not result in the phase envelope being crossed.
These definitions apply to the conditions in the reservoir (isothermal). In the tubing,
cooling will occur and therefore the fluid will almost certainly cross the phase envelope at
some point. Therefore a gas will produce some liquid (condensate) before it reaches the
surface.
It is possible to have all of these fluids with the same composition, it is just the initial
pressure and temperature that may change. For example in the phase envelope (Figure
3), a reservoir temperature of 650°F or below would result in a black or volatile oil. A
reservoir temperature of between 650°F and 858°F would be a retrograde condensate
and a reservoir temperature of above 858°F would be a dry gas. Note in this example a
black-oil fluid is likely as most reservoir temperatures are below 650°F!
Below the saturation pressure, the proportions of the phases will change. This can be
examined on the phase envelope plot (as in Figure 4).
Figure 4 - Fluid Behavior From Reservoir to Surface

Black
4000 Oil
3600 Gas

3200

2800

2400
Pressure (psig)

2000
Gas
Gas
40%

1600
60%
Gas

1200
20%

800
Gas
80%

400

0
-200 0 200 400 600 800 1000 1200

Temperature (degrees F)

As the fluid proportions and properties change, this will clearly effect the well
performance. The reservoir or near wellbore performance will be effected by:
1. The viscosity
20 October 1998 Page 10
 BP Completion Design Manual - Section 4

2. The expansion or contraction of the fluid (included in the formation volume factor)
3. Any relative permeability effects due to liquid or gas break-out (note Prosper can not
handle gas relative permeabilities and these will have to be estimated separately)
The tubing performance on the other hand will be effected by:
1. The density of the fluid(s) - accounted for by the FVF and oil / gas gravity
2. The proportion of gas to liquid (the GOR or CGR)
3. The viscosity (to a lesser extent)
The PVT data is therefore critical to the well performance predictions. Historically most
errors in well performance prediction have been attributed to poor or inaccurate PVT data.
Regardless of which PVT system is used to describe the fluids within Prosper, it is vital
that the original laboratory PVT data is both representative of the reservoir fluids and
covers well performance issues as well as reservoir performance issues:
1. Ensure that the well has cleaned-up adequately (as evidenced by stable flow with a
steady GOR and water cut).
2. There are two options, either bottom hole single phase samples or separator samples,
recombined according to the GOR. Sampling at the separator introduces more errors
than downhole sampling as the proportions of oil and gas have to be accurately
measured. Which ever method is chosen, ensure that the fluid is single phase at the
sandface otherwise condensate or gas breakout may make the produced fluids
unrepresentative. Sampling of fluids where the reservoir pressure is close to the
bubble or dew point is always going to be error prone and this uncertainty must be
acknowledged.
3. The value of accurate samples is huge. For example on Pompano, approximately US
$20 million could have been saved if the paraffin content had been accurately known
and the expensive TFL completions avoided. The only oil sample was unfortunately
“lost” in the laboratory.
4. Multiple samples should be checked against each other. The bubble point is the most
useful consistency check.
5. The PVT analysis should include a constant composition expansion experiment for
conditions between the bubble point and separator pressure. Note, the reservoir
engineer is more likely to be interested in differential expansion where at each stage of
the expansion, the produced gas is removed and therefore the composition changes.
The two methods will produce different GORs, and Formation Volume Factors and
hence different predictions about well performance. Constant composition is more
valid in the tubing as the oil and gas will be in constant contact with each other.
6. Any PVT model may be ideal for the reservoir or the facilities, but not the tubing. For
example the correct fluid density may not be critical for either the reservoir or the
facilities model. However for the tubing the correct fluid density over a large range of
pressures and temperatures is absolutely vital. Likewise, the reservoir fluids do not
need to include a large sensitivity to temperature and facilities correlations do not
necessarily need to cover a large range in pressure.

4.2.1 Direct use of Laboratory Data

It is possible to use laboratory data if it is available over the expected range of pressures
and temperatures. Interpolation will be performed between pressure and temperature data
points, but so long as there are sufficient data points and that for each temperature, there
is a data point at the point bubble, the errors will not be substantial. The real problem
comes from the inflexibility this approach introduces. It does not allow the user to alter the
composition to any extent, i.e. gas lift, or changing GORs can not be incorporated.

20 October 1998 Page 11

 BP Completion Design Manual - Section 4

4.2.2 Equation of State

An Equation of State (EoS) is a mathematical method for modelling a fluid based on the
components in the fluid. As a reservoir hydrocarbon contains hundreds of different
components, the EoS is usually limited to the major components or groups of
components. The properties of these components and these “pseudo components” must
be specified. The equations of state were originally developed for pure substances. With
time their use was extended to mixtures. With mixtures (e.g. reservoir fluids) some
method of introducing a measure of the polar and other interactions between pairs of
dissimilar molecules is required. The binary interaction coefficients change the ideal
Equation of State to match the reality of many mixtures.
The Equation of State used by Prosper is the Peng-Robinson (P-R). It has been well
established that the capability of two-parameter equations of state, such as P-R, in
predicting the liquid density can be improved by introducing the “volume shift” parameter.
The method is particularly attractive because it does not change the predicted phase
equilibrium, but affects the phase densities by shifting the volume axis. The accurate
prediction of density is vital for tubing performance predictions. However this technique is
arbitrary and non rigorous / unscientific. The inclusion of the third parameter in EoS, may
deteriorate the predicted density at some conditions. This could occur for systems with
high concentrations of supercritical compound(s), particularly methane. Moreover, using a
constant shift parameter for light hydrocarbons, the phase densities can not be predicted
accurately.
With an accurate EoS, in theory, the properties of the mixture can therefore be calculated
for any pressure and temperature.
Some of the specific issues concerning equation of state models are:
1. They may give the impression of being highly accurate. Like any model, they are only
as good as data they are based on.
2. They are only valid for a fixed input composition. Although they are very good at
predicting the phase to phase transfers vital for separator performance, they require a
feed composition. Therefore each EoS is only valid for a single GOR. They are
therefore difficult to use for gas lift completions and any model where the reservoir
GOR may change (e.g. gas cap expansion).
3. The equations of state are often produced specifically for reservoir or facilities
engineers they may therefore be invalid for tubing conditions. This is particularly the
case if they have been tuned excessively using volume shift or binary interaction
coefficients. In the example below (Figure 5), the bubble point is ridiculously high at low
temperatures and may indicate other potential problems at likely conditions.

20 October 1998 Page 12

 BP Completion Design Manual - Section 4

Figure 5 - Example of Unreliable PVT Data Caused by Excessive Tuning

Phase Envelope
Pressure (psig)
50000 Critical Point
45000 656.7(degrees F)

40000 3704.9 (psig)

35000

30000 Cricondentherm
886.3 (degrees F)
25000
904.4 (psig)
20000

15000
Cricondenbar
10000 317.9 (degrees F)

5000 4858.2 (psig)

0
-100 0 100 200 300 400 500 600 700 800 900
Temperature (degrees F)

Bubble Point System

4. It is known that many EoS models poorly represent the liquid density of fluids and
errors of around 5% are commonplace. Errors in liquid densities would not significantly
effect either reservoir models or most facilities models, but will seriously effect the
tubing performance.
5. An EoS model is not particularly accurate at determining the viscosity of fluids. This
can be improved by the entering of “critical volumes” for each component.

4.2.3 Empirical Correlations (Untuned)

Various correlations are available that attempt to predict a fluids properties based on the
fluids oil and gas gravity, and the GOR. Each correlation has been designed for a
particular range of fluids and is purely empirical in nature.
The correlations available in Prosper are:
Black Oil FVF, GOR
• Glaso
• Standing
• Lasater
• Vazquez-Beggs
• Petrosky et al
Black oil Viscosity
• Beal et al
• Beggs et al
• Petrosky at al

20 October 1998 Page 13

 BP Completion Design Manual - Section 4

Gas Viscosity (note if a gas is chosen, all the condensate drop out is assumed to occur
in the separator - NOT in the tubing). Therefore any condensate hold-up problems can
NOT be analysed.
• Lee at al
• Carr et al
Any of these correlations can be used directly without tuning so long as the GOR and
densities are known. It would be possible to examine each of these correlations in turn to
see which correlation was developed to model a similar fluid to one in question. However
this implies a complete lack of understanding of the reservoir fluids. At the very least, the
bubble point should be matched and the correlation chosen accordingly.
Retrograde Condensate
Prosper has its own Retrograde condensate empirical model that is capable of predicting
fluid properties and condensate gas ratios (CGRs) below the dew point. Even more than a
black oil model, this model is only realistic if it is matched to real conditions.

4.2.4 Empirical Correlations (Tuned)

The value (and risk) of correlations is that they offer a means of extracting a lot of data
(fluid properties) from very little information. The Black-oil correlations for example only
require an oil and gas density and a GOR in order to produce fluid properties. Where
these correlations become very powerful is when they can be tuned to match
experimental results at specific conditions. The correlations then become applicable over
a particular range of conditions.
In order to tune the correlations, key information is required. In order of priority this is:
1. Bubble point pressure at reservoir temperature.
2. Reservoir condition oil FVF.
3. FVF below the saturation pressure (constant composition expansion data).
4. GOR / CGR below the saturation point.
5. Oil viscosity at reservoir, bubble point and wellhead conditions (constant composition
expansion).
6. The above properties at temperatures below the reservoir temperature.
The tuning process is simple and automated within Prosper. A multiplier and a constant
are applied to the property in question (e.g. GOR) such that the resulting error (standard
deviation) is minimised. With a limited amount of data, an exact match is possible at the
data points. With more data, a best fit will be achieved.
Note if the tuning required is large, the correlations are being used outside of their
intended “envelope”. This will make their use under different conditions questionable. For
example if only data is available at the reservoir conditions and large tuning factors are
required in order to get a match, it is possible that the correlation will be completely in
error under much lower conditions. In the example below (Figure 6), the untuned data is
unrealistic at reservoir conditions. However, by tuning it (excessively) only to the reservoir
conditions, the data is now a poor match at wellhead conditions. The same issue is very
likely when the tuning is performed at a specific GOR and then the same (tuned)
correlation is used at a different GOR.

20 October 1998 Page 14

 BP Completion Design Manual - Section 4

Figure 6 - Example of Tuning the Viscosity

0.8

0.7

Untuned
Viscosity (cP)

0.6
Data
0.5

0.4

0.3

0.2
Tuned to
reservoir Tuned to reservoir
0.1 viscosity and wellhead
viscosity
0
800 1200 1600 2000 2400 2800 3200 3600 4000 4400
Pressure (psi)

20 October 1998 Page 15

 BP Completion Design Manual - Section 4

4.3 Inflow Performance

4.3.1 Radial Inflow Equation and Skin

The fundamental equation of liquid flow from the reservoir into the completion is the radial
inflow equation or Darcy equation:
0.00708kh( PR − Pw )
Q=
r 
Bµ ( Log  e  + S )
 rw 
Where:
Q = flow rate (bpd)
k = permeability (md)
h = reservoir thickness (ft)
Pr = reservoir pressure (psi)
Pw = bottom hole flowing pressure (psi)
B = formation volume factor or FVF (dimensionless)
µ = viscosity (cP)
re = drainage radius
rw = wellbore open hole radius
S = skin (dimensionless). The skin is a measure of any additional pressure drops that are
present if the well
This defines the flowrate for a given pressure drop for a vertical well fully penetrating a
reservoir interval with a constant pressure outer boundary. The skin is a catch all term for
any additional pressure drop. The kh and skin can be determined from a well test.
This equation can be simplified even further, if all the constants are grouped together:
Q = J ( PR − Pwf )
Where J is the productivity index or P.I.
These equations are only applicable above the bubble point and for oil wells.
In Prosper, this equation uses the Dietz shape factor and drainage area rather than the
reservoir / no flow boundary radius. The Dietz shape factor accounts for the fact that the
drainage area may not be circular. Each shape (round, square, rectangular etc.) will have
its own number - for example a well in the centre of a circular drainage area has a Dietz
shape factor of 31.6. A table of shapes and factors is included in the Prosper help files.
The skin combines various different aspects of the well performance:
• Formation damage skin
• Perforation skin
• Partial completion skin (i.e. if not all of a reservoir interval has been completed)
• Deviation skin
• Stimulation effects
• Reservoir heterogeneity skin
• Multiphase flow effects
• Darcy effects of sand or gravel filled perforations
• Non Darcy (or turbulence) effects associated with any of the above.
20 October 1998 Page 16
 BP Completion Design Manual - Section 4

Note it is not possible to treat any of these effects in isolation. For example a well
perforated in a small proportion of the reservoir will make any poor perforation effects
more pronounced.
More details on the inflow performance is found in the BP Near Wellbore Performance
1
Manual

4.3.2 Vogel Inflow Performance

This empirical equation is used for producing wells below the bubble point:

  Pwf   Pwf  
2

Q = Qb + ( Qmax − Qb )1 − 0.2  −  

  Pb   Pb  

Where Qb and Pb are a flowrate and a corresponding bottom hole bubble point pressure.
In order to use this correlation therefore test data must be available.

4.3.3 Fetkovich
Q = J o ( Pr2 − Pwf2 )n
Jo is the P.I. above the bubble point adjusted for any relative permeability effects
This empirical relationship has similar application to the Vogel IPR and should be used
below the bubble point.

4.3.4 Jones
( Pr − Pwf ) = aQ 2 + bQ
This empirical equation is an expansion of the Darcy inflow equation to include rate
dependent or non Darcy effects and is therefore applicable to oil and gas wells. The b
term is the same for the Darcy equation. The a term accounts for the flow velocity:

0.0005359 Bo2 ρ
a=
k 1.201hp2rw
Where hp is the completed interval.
Note the Jones equation does not account for any flow convergence around perforations
or for any case that is not a simple open hole completion.

4.3.5 Hydraulically Fractured Well

A fractured well performance will depend on getting the fluids to the fracture face and
getting the fluids along the fracture. The first is dependent on the reservoir permeability
and the fracture length, the second more dependent on the fracture conductivity.
The (dimensionless) fracture conductivity (Fcd) is dependent on the fracture width and
proppant permeability:
K f Wf
Fcd =
KX f

1
Near Wellbore Performance Manual WEO-W08 September 1990

20 October 1998 Page 17

 BP Completion Design Manual - Section 4

Where Kf and K are the permeabilities of the proppant and reservoir respectively, W f is the
fracture width and Xf is the fracture half length (distance from the well to the tip of the
fracture). Fracture widths are usually in the order of 0.15 to 0.75 inch. A fully effective
fracture should have an Fcd > 10. Note the permeability of proppant is much less under
real conditions than measured in the laboratory due to compression, gel residues, fines
entrapment, long term crushing, non Darcy effects or contamination. A useful rule of
thumb is to reduce the lab derived permeability value by a factor of 10. The fracture width
will also be reduced by embedment of the proppant into the rock. In severe cases, this
can lead to a severe reduction in the effectiveness of the fracture and techniques for
extending the width of the fracture to compensate (i.e. tip-screen out techniques) should
be used.
The model used within Prosper is a transient model and accounts for the early time being
dominated by fracture conductivity and the later time being dominated more by overall
reservoir performance. An example of the effect this has is shown in Figure 7.
Figure 7 - Propped Fracture Transient Performance
4000
1:Time (days)
3600
0=0.18
1=0.32
2=0.58
3200 3=1.05
4=1.90
5=3.42
2800 6=6.16
7=11.10
8=20.00
8 9=40.00
2400
Pressure (psig)

9 0
7
2000 5

6
1600 1
4 2

1200 3

800

400

0
0 2000 4000 6000 8000 10000 12000 14000 16000 18000

Liquid Rate (STB/day)

In this example, only after 20 days, has the well settled down to close to steady state
performance.

4.3.6 Horizontal Wells

Prosper uses three models. The first is the Horizontal well model of Kuckuk and Goode. It
assumes a horizontal well in a closed rectangular drainage volume bounded by sealing
surfaces. The pressure drop in the wellbore itself is not accounted for. This is acceptable
for cases where the drawdown is an order of magnitude greater than this frictional
pressure drop.
The second model is similar except that a constant pressure upper boundary is assumed.
This would be the case for a well with an active gas cap. Such a case will give higher
productivities than an upper no flow boundary.

20 October 1998 Page 18

 BP Completion Design Manual - Section 4

The third method allows for pressure drops along the completion interval itself. A number
of horizontal well models are available including the preferred ones of Goode and
2
Wilkinson and Kuckuk and Goode. Note both of these formulations are valid even when
the horizontal wellbore length approaches the length or width of the reservoir. The method
of Goode and Wilkinson is less recommended when frictional pressure drops are
included, but is preferred when friction is ignored (infinite conductivity). Note in order to
Use Goode and Wilkinson with Infinite conductivity, set the roughness to 0 (this turns off
friction, rather than assuming friction for a smooth pipe. The other horizontal well models
are included for completeness and should not be used in practice.
Note the Wong and Clifford model for a deviated well will be available in Prosper v.6 by
late 1998.

4.3.7 Perforation and Deviation Skin

Prosper uses three options for models to calculate the skin in a deviated, perforated well.
Locke’s method is the original model and is included in Prosper for completeness.
Mcleod’s method includes an option of over or underbalanced simplification. The crushed
zone permeability is reduced to 10% for over balanced perforating, and 40% for
underbalanced perforating. This is rather arbitrary and a more detailed discussion is found
in section 5 of this manual. Karakas and Tariq is a more general technique and is
recommended for most applications. It can deal with high deviations - so long as radial
flow is reached away from the wellbore and therefore the well is not too long in
comparison to the reservoir.
The input data for Karakas and Tariq is rather involved. The possible provenances of the
various data inputs are included below:

2
Inflow Performance of Partially Open Horizontal Wells P.A. Goode and D.J. Wilkinson SPE 19341 August
1991

20 October 1998 Page 19

 BP Completion Design Manual - Section 4

Table 1 - Karakas and Tariq Data Input

Perforation Diameter This is the diameter in the rock (not the casing). It can be
extracted from the gun vendor with a correction for the rock
strength. Alternatively the equation
D = EH N 80 [327
. − 0.61 ln(UCS )] 3, for deep penetrating
charges where D is the perforation hole diameter, UCS is the
unconfined compressive rock strength and EHN80 is the
entrance hole diameter in N80 steel (available from perforation
catalogues).
Shots per Foot For underbalanced perforating this will be the gun’s shot per
foot. If overbalance perforating with dirty fluids, the open
perforation may only be 0.1 - 0.25 the nominal shot per foot.
Perforation Length This is a property of the gun and charge, with corrections for
confining stress and rock strength. These corrections are vital
and the equations will soon be incorporated into Prosper but
are available in THoR and from the gun companies. Typical
numbers are between 12 and 24 inches
Damaged Zone This will depend on the mud and rock properties and how the
Thickness wells are drilled. Typical values are 3 to 12 in. The actual
values. The value is not particularly critical - so long as the
perforations extend behind this. The actual thickness can be
determined from filter cake tests on core or estimated from
filtrate depth of invasion calculated from mud losses - difficult
to achieve in practice.
Damaged Zone This is the damage caused by the mud filtrate invasion into a
Permeability reservoir. The ratio between this and the formation permeability
can be anything from 0 to 99% damage with an average of
around 0.8. Core flood tests (return permeability) are vital for
this. Note that the damage profile is also important. For
example with a surfactant adsorption damage mechanism, the
damage is likely to be most severe near the wellbore - where it
has least effect in a perforated completion, by the time the
filtrate reaches the total invasion depth, the surfactant may be
completely adsorbed and therefore no further damage occurs.
Crushed Zone From CAT scans and thin sections of perforated core, this is
Thickness typically 0.5" thick.
Crushed Zone Assuming an optimum perforation underbalance, this should be
Permeability the same as the formation permeability, otherwise a figure of
up to half the formation permeability can be used.
Shot Phasing A choice of the engineer! Low values (60 - 90 degrees) will
normally give optimum perforation efficiency.
Deviation A choice of the engineers!
Penetration This is the fraction of the wellbore that is open to flow. Prosper
assumes that it the top interval only that is completed (i.e. from
the top boundary of the reservoir downwards).
Vertical Permeability There is some uncertainty around what figure should be used.
In THoR and KT Perf, there was a natural division between
short and large scale k v/kh. Short scale Kv/kh applied to the
scale of the perforations i.e. on the scale of a few inches. Large
scale kv/kh applied to the effects of anisotropy on the deviation
or partial penetration skin. Prosper assumes that the large and
small scale kv/kh are the same. In a vertical fully completed
well, the small scale kv/kh should be used, but in a high angle or
partially completed well, the large scale effects may be more

3
Underbalance criteria for Minimum Perforation Damage L.A. Behrmann SPE 30081, May 1995

20 October 1998 Page 20

 BP Completion Design Manual - Section 4

important and the lower large scale kv/kh should be used.

Therefore some judgment is required in order to pick the
correct vertical permeability. This anomaly should be corrected
when the method of Wong and Clifford is included in Prosper
v.6
Wellbore Radius This is the drilled open hole radius.

4.3.8 Multiple Zone Completions

Prosper has various options for looking at multizone completions. It can either include or
exclude the frictional pressure drop between each zone. Note that although crossflow is
included - it is only in the wellbore - no fluid flow is allowed between different layers in the
reservoir.

4.3.9 Gravel Pack Completions

Prosper can handle both cased and open hole gravel packs. It can also deal with open
hole completions with sand control but no gravel. The subject of gravel packing can not be
dealt with in detail here - only the basic drivers on productivity and how to calculate it for a
gravel pack completion. The BP sand control guidelines contain much more detail.

4.3.9.1 Cased Hole Gravel Packs

A cased hole gravel pack is where screens are installed in cased and perforated well.
Gravel is pumped to fill the volume between the casing and the screen and also - critically
- the perforations themselves.
The cased hole gravel pack pressure drops include those of a cased and perforated
completion (i.e. phasing, perforation length and crushed zone permeability all have a
contribution to the pressure drop). However in addition, there will be a substantial
pressure drop through the gravel in the perforations and in the gravel between the casing
and the screen.
The models for cased hole gravel pack pressure drops therefore incorporate the
perforating parameters discussed in 4.3.7. The following variables are required for the
pressure drops in the gravel:
1. gravel pack permeability (this can be extracted from the table below (Table 2),
although due allowance must be given to downhole conditions - any debris or
compaction for example)
2. perforation diameter - this is critical - and the reason why big hole charges (1”
diameter) are preferred
3. shots per foot - again critical 12 spf is preferred (dependent on liner size)
4. gravel pack length - in Prosper this is the distance from the sandface (not the casing)
to the screen O.D. This means that the liner size is NOT included in any of the
calculations and therefore the validity of these calculations.
5. perforation interval
6. perforation efficiency

20 October 1998 Page 21

 BP Completion Design Manual - Section 4

Table 2 - Gravel Pack Permeability (clean, uncompressed gravel)

US Mesh Range Sieve opening (mm) Permeability (Darcy)

6/10 2 - 3.360 2703
8/12 1.68 - 2.38 1969
10/20 0.84 - 2 652
12/20 0.84 - 1.68 630
16/30 0.589 - 1.19 250
20/40 0.42 - 0.84 171
40/60 0.25 - 0.42 69
50/70 0.21 - 0.297 45

Note, the criticality of getting a large number of large diameter perforations filled with high
permeability gravel. In the event that the perforations are not effectively packed, they will
fill with reservoir sand. The permeability of the loose reservoir sand is likely to be much
lower than properly sorted and clean gravel.

4.3.9.2 Open Hole Gravel Pack

An open gravel pack where a screen is installed open hole and the volume between the
screen and the open hole is packed with gravel. The same model can also be used to
model an open hole sand control completion where the formation collapses around the
sand screens.
The productivity of an hole gravel pack will be dependent largely on the permeability of the
gravel or formation sand. If there is no gravel and the formation is heterogeneous, the
collapsed sand may contain significant fines or clay and the permeability will be reduced
as well as these fines potentially plugging the screens.

20 October 1998 Page 22

 BP Completion Design Manual - Section 4

4.4 Tubing Performance

The pressure drop in any tubing or conduit for any fluid is a function of three components:
• Gravity head
• Friction loss
• Acceleration
∆ptotal = ∆p gravity + ∆p friction + ∆pacceleration
The acceleration loss term is normally small and is ignored unless there are large
amounts of expansion (i.e. a well producing to atmospheric pressure).
The friction is dependent on the fluids (i.e. the PVT), the completion (I.D. and roughness)
and of course the flowrate. Frictional pressure drops may be typically 10% of the overall
pressure drop for a properly sized oil well completion, although for a gas well may be
significantly higher. The roughness of the tubing will depend on the condition and
metallurgy.
Table 3 - Roughness of Tubing

Tubing material Roughness Information Source

CeramKote 54 0.000027 in National Engineering Laboratory
Report FRI001
Tuboscope TK 236 0.00008 in Tuboscope coatings manual
New 13% Cr L80 0.0006 in
New Carbon steel L80 0.0029 in Measured on BP stock
Lightly corroded steel 0.01 in
Heavily corroded / scaled 0.1 Estimate based on 2.5 mm pitting
steel or scale

The gravity term is the weight of the fluid. For a single phase system, it is simply the
density of the fluid. For multiphase systems, the density is a function of the relative
proportions of the different phases. If the phases are travelling at the same speed (no
slip), then the proportion of each phase will depend on the PVT properties and the
pressure. In reality the gas will travel faster than the liquid and this will increase the
proportion of liquid (termed the hold-up). Correlations are used to determine the amount
of slip and therefore the proportion of each phase. The correlations are dependent on
accurate PVT data in order to make valid predictions of the slippage.
The relative contribution of the friction and the density will vary with production rate (Figure
8) and also tubing size. Note that the average density reduces as the flowrate increases.
This is because the slippage will reduce until the hold-up approaches the no slip hold-up.
Below a certain rate the density increases markedly and will approach the liquid density at
zero flowrate. This is because the slippage is excessive and in fact the liquid will not be
produced at a fast enough rate to be produced out of the well. When this happens the
liquid builds up until it is produced out in a large slug. This is called heading and is outside
of the scope of the pressure drop correlations available for steady state flow - this is
because it is a transient or time dependent phenomena.

20 October 1998 Page 23

 BP Completion Design Manual - Section 4

Figure 8 - Friction and Density Contributions to the Pressure Drop

Total
Pressure Drop (psi / ft)

Density

Friction

Flowrate (bpd)

This transient phenomena has four implications:

1. All the Prosper pressure drop correlations will be invalid in this flow area.
2. Any well tests flowing at these rates can not be used for comparison or tuning of the
flow correlations.
3. The completion and well design should avoid flowrates in this area by appropriate
choice of tubing size (see section 4.4.9).
4. You can not predict the shut-in wellhead pressure by using the total pressure drop at a
zero rate. This will seriously under predict the shut-in pressure.
As the gravity term is typically 70-90% of the total pressure drop inside a correctly
designed oil completion, getting the density of the fluids correct is vital - the PVT data
must therefore be at least accurate when it comes to the oil density, formation volume
factor and GOR.
There are a number of techniques used to calculate the hold-up:
The simplest (e.g. Gilbert’s curves) are empirical correlations based on field data.
Of greater complexity are correlations that attempt to predict a flow regime and then use
correlations

4.4.1 Which Correlation Should I Use?

The vertical lift performance correlations available are extensive. However each
correlation has been developed to satisfy the prediction of certain systems. No single
correlation is available that is ideal for all wells. It is therefore very beneficial if the
correlations can be short-listed based on their intended use (i.e. flow regime or oil density
for example). From the short-list it is then possible to select a correlation that most
accurately matches real data e.g. a well test.

4.4.1.1 Fancher-Brown Correlation

This is a strange correlation in that its purpose is not to accurately predict the tubing
performance! It is a non slip (or homogeneous) flow equation. This means that it will
always over-predict the fraction of the pipe occupied by the gas. As a result it will tend to
under predict the pressure drop where slippage occurs. It will therefore only accurately
predict the pressure drop under high flow bubble regime conditions or single phase flow
(e.g. above the bubble point).

20 October 1998 Page 24

 BP Completion Design Manual - Section 4

The importance of this correlation is that it should under predict pressure drop. Therefore
when plotted against real data it should lie to the left of the real (measured) data. If it lies
to the right it indicates either errors in the measured data, errors in friction or more likely
errors in density predictions i.e. PVT problems.

4.4.1.2 Hagedorn and Brown

This relatively simple flow correlation does not use a flow map. Instead it is based on data
from a 1400 ft test well with air, water and dead crude over a range of viscosities from 10-
110 cp. It covers bubble and slug flow and therefore should not be used for gas wells or
high GOR wells. The original Hagedorn and Brown correlation should never be used,
however refinements in Prosper have avoided some of the original problems. They also
allow this correlation to be used for deviated wells. This correlation should not be used to
predict the minima of the lift curve. It tends to under predict the minima. There are also
errors in large bore wells with a 35-70° deviation where oil / water slippage may occur.
Despite these constraints it is frequently used and accurate under the conditions it was
designed for.

4.4.1.3 Duns and Ros

This correlation uses a flow map (i.e. determines the flow regime) to determine the hold-
up and friction. This correlation should therefore apply to all flow regimes. In reality in
works best in gas and gas condensate wells - especially in mist flow and performs poorly
in oil wells where slug flow occurs or at low rates. The loading rate prediction for gassy
wells (>6,000 scf/stb) is generally good.

4.4.1.4 Orkiszewski
4
This correlation is an amalgam of various correlations including Duns and Ros . It works
well over a wide range of conditions and often appears as the closest correlation to match
well test results. Unfortunately there is a discontinuity in the predictions at a mixture
velocity of 10 ft/s. This results in a jump in predicted pressures (as seen in Figure 9). This
means that the correlation is of little practical use.
Figure 9 - Orkiszewski VLP Predictions

4400

4200
Pressure (psia)

4000
4.5" Tubing

3800

3600

5.5" Tubing
3400

3200

3000
0 4000 8000 12000

Liquid Rate (STB/day)

4
Predicting Two Phase Pressure Drops in Vertical Pipe J. Orkiszewski JPT June 1967 SPE 1546

20 October 1998 Page 25

 BP Completion Design Manual - Section 4

4.4.1.5 Beggs and Brill

5
This was the first comprehensive study of multiphase flow for inclined pipes . Small
diameter pipe (up to 1.5in) was used in experiments where the flow regime was observed.
The correlation over predicts liquid hold-up in wells and pipelines however it is in
widespread use as it applies to vertical, horizontal and inclined pipes.
The main use of the Beggs and Brill correlation is for quality control. As it overpredicts the
hold-up, it over estimates pressure losses -typically by between 10 and 15%. This means
that actual data should lie to left of the Beggs and Brill correlation. If results lie to the right
of the correlation, possible causes are:
1. Inaccurate PVT data - particularly too low a liquid density.
2. Under recorded water cuts or over estimated gas rates.
3. Too low a friction factor.
Caution: arbitrarily altering any parameter in order to get a match must be avoided. There
may be a temptation to increase roughness in order to get consistency. However this
should only be done if there is physical or analytical evidence that corrosion, scale or other
deposits may be effecting the condition of the bore of the production conduit.

4.4.1.6 Ansari
6
This correlation is recently available in Prosper. It has been available in THoR and
Multiflo for some time. It is a “mechanistic” model in that it predicts pressure drops based
on the mechanism i.e. the flow regime. This is similar to the more sophisticated traditional
methods such as Duns and Ros. However newer mechanistic models take this further by
eliminating more empirical steps and consider the transitional regimes further.
The Ansari correlation should therefore cover all flow regimes, all sizes of tubing, all
deviations and everything from heavy oil to gas. In a comparison of this model across a
7
number of oil and gas fields , at a variety of conditions, this correlation was the most
widely applicable - although by no means perfect, with average absolute errors of around
7%. There is still concern that this correlation (and many others) is not fully applicable for
deviated wells.
This correlation finds greatest application with low GOR, heavy or medium oil fluids such
as Harding, Foinaven and Andrew. It is acceptable for use on higher GOR fluids and
gasses, but correlations such as Gray are often better for gas.

4.4.1.7 Petroleum Experts Correlations

Petroleum Experts have developed there own hybrid correlations. According to them
“Petroleum Experts” uses the Gould et al flow map Flow Map and for the various flow
regimes the following correlations are applied:
• Bubble flow: Wallis and Griffith
• Slug flow: Hagedorn and Brown
• Annular Mist flow: Duns and Ros
• Transition: combination of slug and mist
Petroleum Experts 2 includes the features of the PE correlation plus original work on
predicting low-rate vertical lift performances and well stability.

5
A Study of Two-Phase Flow in Inclined Pipes H. Beggs and J. Brill May 1973 JPT
6
A Comprehensive mechanistic Model for Upward Two-phase Flow in Wellbores. A. Ansari, et al SPE 20630
Sept 1990
7
An Evaluation of Recent “Mechanistic” Models of Multiphase Flow for Predicting Pressure Drops in Oil and
Gas Wells J. Pucknell at al SPE 26682 Sept 1993

20 October 1998 Page 26

 BP Completion Design Manual - Section 4

4.4.1.8 Gray
This correlation was developed for vertical gas condensate wells and was originally used
in the API 14B subsurface safety valve sizing program. The published limits of the
1
correlation are a flow velocity of 50 ft/sec, 3 /2" tubing or less, and a condensate ratio of
under 50 bbl/MMscf. In reality however, Gray has proven accuracy in low to moderate
CGR gas wells and is useful for predicting liquid loading. It is empirical and uses its own
PVT module for condensate and liquid prediction. It is therefore not compatible with
compositional PVT models.

4.4.2 Using Correlations to Predict When a Well Dies

The correlations also vary considerably in their ability to accurately predict the minimum of
the lift curve. Figure 10 shows a comparison of the correlations used in Prosper. Note the
difference in the shape of the curves at low flowrates. The minima predicted varies
between about 2,000 bpd and 8,000 bpd.
Figure 10 - Correlation Comparison
5000
Petroleum Experts 3
4500
Fancher Brown
4000 Orkiszewski
Hagedorn Brown
3500 Duns and Ros Modified
Beggs and Brill
3000
Pressure (psig)

Mukerjee Brill

2500 Ansari
Petroleum Experts
2000 Petroleum Experts 2

1500

1000

500

0
10000 20000
Liquid Rate (STB/day)
In general the following caveats apply:
1. Don’t use the correlation outside of its intended regime i.e. don’t use Duns and Ros for
oil.
2. Avoid using Orkiszewski, Fancher Brown, Beggs and Brill or Hagedorn Brown for
predicting tubing lift minima.
3. In the absence of a verified correlation for the field, Ansari has proven good at
predicting minima over a wide range of fluids.

4.4.3 Tuning Pressure Drop Correlations

Prosper has the capability to “tune” the correlations to match real data. This it does by
regression on two variables:
Parameter 1 - this is a multiplier on the slippage and therefore alters the hold-up and
therefore the gravity term. For a discrepancy more than 5%, the density term itself is
altered.
20 October 1998 Page 27
 BP Completion Design Manual - Section 4

Parameter 2 - this is a multiplier for the friction term.

Extreme caution must be exercised if tuning the VLP correlations. It is possible to get a
reasonable match with any data, simply by tuning the correlation. Tuning the VLP
correlation is only valid under the following conditions:
1. There is absolute confidence in the PVT data and that it is valid for the conditions the
well was flowing under when it was tested. For example if the PVT data is for a certain
GOR and the well test GOR has risen, the confidence in the extrapolated PVT data will
be reduced.
2. The flowrates (oil, water and gas), BHP, BHT, THT and THP are accurately known
during the tests and are steady. In order to confirm the accuracy of these
measurements, the type of meters, the flow period and whether the meters have been
recently calibrated should be critically analysed and the pressure data not confined to a
spot reading. The pressure should be watched for at least 10 minutes to ensure that it
is not fluctuating and effected by instability or heading.
3. The flow regime should not be severe slug or any other unstable state. This can be
checked by plotting the predicted BHP vs. rate. If the test point lies close to or to the
left of the minima of the curve, the correlations will not accurately match the test data
and therefore tuning the correlations will probably render invalid any predictions for
conditions outside of the test condition.

4.4.4 Natural Flow

Prosper can easily be used to predict the natural flow performance of a well over the well’s
lifetime using the system performance calculation.
Issues to watch are:
• Ensure that the datum depth for the completion is the same as that used to reference
the reservoir pressure. It is accurate to use the mid depth of the completed interval.
However, unless friction in the liner is large, then it is not critical where the datum is, so
long as it is consistent.
• It is not normally necessary to include restrictions in modelling the completion. Nipples,
packers, gas lift mandrels and tubing retrievable safety valves will cause a negligible
additional pressure drop. Wireline retrievable valves should be included and any
change in tubing size.

4.4.5 Shut-in Conditions

Determining the wellhead shut-in pressure is an important part of well performance
prediction. It is not easy! The wellhead shut-in pressure is required for:
• Determining the flow control requirements at surface i.e. the wellhead and tree ratings.
• Required treating pressures e.g. scale inhibitor squeezes, fracture treatments etc.
A number of techniques are available:
1. Assume a gas gradient to surface and the maximum reservoir pressure. This will give
you the worst possible case under all conditions (so long as the lowest gas gradient
and the highest possible reservoir pressure are used). Note, remember that the gas
gradient has a very strong (almost linear) relationship with pressure. The easiest
method of determining the gas gradient is to use Prosper and perform a gradient
prediction with a gas fluid (no condensate). A low rate (e.g. 0.1 mmscf/d) rate can be
used. The gravity pressure loss can then be extracted from the variables available and
plotted (Figure 11).

20 October 1998 Page 28

 BP Completion Design Manual - Section 4

Figure 11 - Gas Pressure Gradient Calculations Using Prosper

Gradient traverse - plot ( 19 Jun 98 13:08)

0 Case 1 50 psi Wellhead
Case 2 100 psi Wellhead
1000 Case 3 500 psi Wellhead
Case 4 1000 psi Wellhead
Case 5 7000 psi Wellhead
2000

3000
True Vertical Depth (feet)

4000

5000
Fluid G a s
Flow Tubing
6000 T y p e Producer

7000

TH Temp 80.0 (degrees F)

8000 BH Temp 180.0 (degrees F)
B H M D 1 2 0 0 0 . 0 (feet)
BH TVD 9 1 0 0 . 0 (feet)
9000

10000
0 100 200 300 400 500 600 700 800 900 1000 1100

Gravity Pressure Loss (psi)

Clearly this will frequently tend to be an over estimation of the wellhead pressure,
particularly if the fluids are not gassy or a lot of water is being produced. What
frequently happens is that the high gas oil ratios are associated with depletion and the
therefore high gas oil ratios (and therefore a simplification of gas only) do not combine
with high reservoir pressures to give high wellhead pressures. However as a first pass
it is quick, simple and conservative.
2. The next approach is a bit more subtle. If the BHP vs Rate could be extrapolated down
to a zero rate without any of the problems associated with varying hold-up at low rates,
we would be close to extracting a realistic pressure gradient. This is easily achieved
using the Fancher - Brown no slip correlation. This correlation will give the highest
proportion of in-situ gas and therefore will be conservative. This gradient will be
realistic for conditions immediately after a shut-in. Unfortunately phase segregation will
occur - this can result in higher pressures than immediately after . Fortunately the oil
will always drop in comparison to the gas. This means that there will be no further
phase transition from oil to gas. Therefore as the standard volume of gas is known at
the time of shut-in, it is conservative to assume that the amount of space occupied by
the gas remains the same. This then means that the standard gas volume = (1-total
hold-up) x completion volume. The actual gas volume is this standard gas volume
corrected to the pressure in the well.
3. Use a transient model - if you can find one?

4.4.6 Artificial Lift

The well performance aspects of artificial lift need to be integrated with all the other
aspects of artificial lift completion design. These are covered in section 5 - Architecture of
Completions.
Prosper can be used to assess the options for gas lift or Electrical Submersible Pumps or
to troubleshoot the performance of such systems when installed.

20 October 1998 Page 29

 BP Completion Design Manual - Section 4

4.4.6.1 Gas Lift

Gas lift works on the principal of lowering the hydrostatic pressure by lowering the
average density of the produced fluid by introducing gas.
This is easily modelled in most well performance predictions - so long as varying gas oil
ratios can be modelled accurately. This automatically excludes the effective use of EoS
models, where a different model is required for each composition.

4.4.6.1.1 Injection Depth

Because gas will occupy a greater volume at lower pressures, injecting a small amount of
gas will have a relatively large effect at shallow depths. Injecting gas deeper will always be
beneficial. Because the deeper gas will be at a higher pressure, more gas may however
be required or be beneficial. This is demonstrated in Figure 12.
Figure 12 - The Benefit of Increasing Depth of Gas Lift Injection
10000

9500
1

9000 1

0
8500
0
Liquid Rate (STB/day)

1
8000
0
Gaslift Gas Injection Rate
7500 Curve 0 = 4 (MMscf/day)
Curve 1 = 8
7000 0
1
6500

6000

5500

5000 01
0 2000 4000 6000 8000 10000

Injection Depth (feet)

When considering what is the optimum gas lift depth, then the other completion design
aspects must be considered. These include casing design and packer setting depth.
These are discussed in some detail in section 5.
Note, the benefit of increasing depth of injection will depend on the pressure drop
correlation in use. This is particularly the case for high angle wells. There is some concern
that the current models may not be entirely realistic when it comes to multiphase
modelling at high angles - the benefit of injecting gas very deep along highly inclined
wellbores may be over estimated by the models. This could result in gas lift mandrels
being placed where they can only be serviced by tractors or coiled tubing, without
sufficient justification. Figure 13 shows a plot of injection depth vs. liquid rate for a typical
deviated well profile. Note the big difference in the benefit of deep gas lift for correlations
such as Hagedorn and Brown compared to the Ansari correlation. In this instance it is
suspected that the Ansari correlation is closer to the truth, due to its ability to handle a
greater range of flow regimes.

20 October 1998 Page 30

 BP Completion Design Manual - Section 4

Figure 13 - Gas Lift Injection Depth vs. Rate for 5 Untuned Correlations
15000

13500

12000
Orkiszewski
10500 0
0 Hagedorn Brown
Liquid Rate (STB/day)

0
0 0 0
9000 0
0 0 0 0
0
0 0
0 00
7500 0
0 0 0 0
0
0
6000 Ansari 0 0
0
4500
Petroleum 0 0
3000 Experts 2
0
1500 Duns and Ros Modified

0
0 4000 8000 12000 16000
Injection Depth (feet)

4.4.6.1.2 Gas Injection Rate

Increasing the amount of gas injected will increase the liquid production levels - up to a
point. The hydrostatic pressure will lower with more gas, but the frictional pressure will
rise. There comes a point when increasing the gas results in the frictional pressure
increasing more than the hydrostatic pressure decrease. This point is very condition and
well specific. The maximum liquid production rate is not going to be the optimum. When
designing the field facilities, there will be a cost associated with building gas compression.
There will also be an operating cost associated with gas compression. This will include
servicing and power costs. This means that the optimum production rate will be somewhat
below the maximum. This can be determined by determining the cost vs. benefit of the
gas and plotting this line on the performance curve (Figure 14). When operating these
wells, the availability of the gas must also be considered. This is covered in section 4.4.9.4

20 October 1998 Page 31

 BP Completion Design Manual - Section 4

Figure 14 - Injection Rate vs. Production Rate

11400
1 1
10800 1 1
1
10200 Maximum liquid
1 Optimum gas injection production for deep gas
9600 rate for deep gas lift lift
1
Liquid Rate (STB/day)

9000
Injection Depth
1
Curve 0 = 3000 (feet)
8400 Curve 1 = 12000

7800
1

7200 0 0 0 0
0 0
0 0
6600 0
Optimum gas Maximum liquid
6000 injection rate for production level for
01 shallow gas lift shallow gas lift
5400
0 3 6 9 12 15 18 21 24 27 30

Gaslift Gas Injection Rate (MMscf/day)

4.4.6.1.3 Gas Lift Design

Having determined the optimum gas injection depth and rate, well performance
predictions can be used to determine the optimum gas lift design. Note at this point it is
very important to consider a life of well approach, rather than a specific condition. Prosper
can be used to determine the gas lift design with respect to well performance. However it
is strongly recommended that the gas lift company or gas lift expertise is included at an
early stage so that the whole system (wells, facilities and reservoir) can be integrated and
best practice incorporated.
The purposes of the gas lift design are:
1. Produce a gas lift design that achieves the optimum production rate
2. Produce a design that is reliable and does not introduce heading problems, instability
or start-up problems.
3. Produce a design that is simple to troubleshoot and optimise
4. Minimise costs by avoiding unnecessary gas lift mandrels.
Prosper will assist in the first of these issues:

20 October 1998 Page 32

 BP Completion Design Manual - Section 4

Figure 15 - Gas Lift Design Plot

0
1000 GASLIFT DESIGN DATA
Design Situation : New Well
2000 Valve Type : Casing Sensitive
Dome Pressure Correction
(above 1200psig) : Yes
3000
Vertical Lift Correlation : Petroleum Experts 3
True Vertical Depth (feet)

4000
Maximum Gas Available : (MMscf/day) 6.000
5000 Flowing Top Node Pressure : (psig) 500.0
Operating Injection Pressure : (psig) 2700.0
Valve Water Cut : (percent) 90.000
6000 ACTUAL Liquid Rate : (STB/day) 7209.4
ACTUAL Oil Rate : (STB/day) 720.9
ACTUAL Gas Injection Rate : (MMscf/day) 5.975
7000 ACTUAL Injection Pressure : (psig) 2200.0
8000
Valve
9000

10000
Orifice

11000

12000
0 400 800 1200 1600 2000 2400 2800 3200 3600

Pressure (psig)

The design capabilities of Prosper are quite extensive and will not be gone into in detail
here. Some points that may be useful are:
1. There is a choice of valve type. All gas lift valves are sensitive to both tubing and
casing pressure. However casing pressure operated valves are more sensitive to
casing pressure and vice versa. The choice should not be dictated by performance and
uncertainty. A casing sensitive valve requires a drop in casing pressure to close the
valve, therefore potentially more valves may be required or a higher unloading
pressure. Figure 15 shows a plot of a gas lift design using a casing pressure drop to
close the unloading valves. However it is less sensitive to changes in reservoir
properties as the casing pressure is much easier to determine than the tubing
pressure. Tubing pressure operated valve systems are conversely harder to design as
they are more dependent on changes in reservoir properties.
2. The number of valves required is dependent on the separator or wellhead pressure
during the unloading and the assumption about the static gradient of the load fluid. The
gradient may be the completion fluid gradient, however this can produce a design that
is overkill for all other conditions. Consider the possibility of designing the gas lift
system so that the unloading (i.e. removing the completion fluid from the tubing) can
be performed by a nitrogen lift and routine kick-off performed as normal. This means
that the static gradient would become the worst case static gradient of the reservoir
fluid. This can be assumed to be 100% formation water or whatever.
3. Consider the point at which gas lift is required. During the early field life, there may be
little or substantial benefit in gas lift depending on the conditions and especially the
GOR of the produced fluid.
4. Do not design the system just for high rates and high reservoir pressures. This will
push the use of large orifice sizes which may introduce heading problems later on in
the well life.
5. Depending on the access difficulties, it can be useful to consider changing the orifice
valve position with time. As the reservoir depletes, the use of a deeper injection
position may be possible and beneficial. This implies that either two orifice valves are
run from day 1 or the lower mandrel is run with a dummy valve.

20 October 1998 Page 33

 BP Completion Design Manual - Section 4

6. The pressure drops with gas between the casing and the tubing are usually not
significant. However Prosper has the capability to include these as required.

4.4.6.2 Electrical Submersible Pumps (ESPs)

A pump can be introduced into the completion in order to increase the energy of the
system. The pumps are centrifugal multistage pumps powered by a 3 phase motor. The
speed of the motor is dependent on the frequency of the power supply. The well
performance aspects of such pumps and motors are addressed in this section, however it
is important to integrate the facilities, completion design and reservoir issues. These are
considered further in section 5.
Prosper works in a similar manner to the hand calculations that can be performed for ESP
design:
1. First the total dynamic head requirement is calculated. This is head required to lift the
fluids from the suction to the discharge of the ESP. The head is required rather than
the pressure as the pump will deliver a constant head independent of the fluid,
whereas the pressure will be dependent on the fluid being pumped.
2. A pump stage is then selected in order to fit inside the casing and deliver the flowrate.
The number of stages is then chosen in order to deliver the total dynamic head. With
pump efficiency this determines the motor power requirements.
3. A motor and cable are then chosen to deliver the pump power requirements. With
motor efficiency and cable losses, this determines the surface power requirements.
4. Sensitivities are performed to see over what range of conditions the pump / motor can
operate. If necessary a variable speed drive can be used in order to extend the range
of conditions the pump / motor can cope with.
The following issues must be considered during this process:
1. The position of the pump must be considered and its implications - see section 5.
2. The power requirements will depend largely on the fluids being pumped and the
reservoir pressure. This must be assessed over time, so that facilities are not under
designed. In one case, the reservoir depletion was not accounted for in the facilities
1
design and as a result the power available was only /4 of what it should have been.
3. The design of the pump / motor should be based on expected changes and uncertainty
in conditions - but over a much shorter time frame than the facilities design. The
average run life of an ESP is 2-4 years. There is therefore no point in designing it for
the reservoir pressure expected in 10 years time!
4. Free gas at the pump suction may seriously limit the performance of the pump. This
will either limit the available drawdown or require gas handling pumps or a gas
separator. Prosper can output the free gas liquid ratio at the pump intake.
5. The fluids will have an effect on the pump efficiency. With emulsions, the pump
efficiency may only be 10%, compared to a typical figure of 40% for less viscous
8
fluids . This poor efficiency is however not all bad news. The other 90% will be
transferred to heat energy in the fluid and may increase free gas above the pump,
lower viscosities and aid in oil / water separation at surface. Prosper can not directly
perform these calculations, however the energy “lost” will be known and can easily be
converted into heat energy of the fluid by using the heat capacity of the fluid. This
temperature increase can then be manually entered into Prosper.
6. The range of pumps and motors in the Prosper database is not complete and may not
be up-to-date. All designs should be checked by an ESP vendor(s) independently to
see if the same recommendations are made.

8 th
Energy Management in ESP Wells A. Simpson 5 European ESP Roundtable January 1998

20 October 1998 Page 34

 BP Completion Design Manual - Section 4

4.4.6.3 Jet Pumps

Jet pumps can not be modelled in Prosper. However, they are an effective means of
artificial lift and should not be ignored. They work on the venturi principle (Figure 16). The
power fluid is pressurised and exits the nozzle at high velocity. The velocity of this fluid
creates a low pressure area in the throat. This low pressure draws in the reservoir fluid
and the fluid exits the throat into the diffuser, where expansion converts velocity into
pressure.
Figure 16 - Jet Pump Performance

Diffuser

At
Throat
An

Power Fluid

Reservoir
fluids
Nozzle

Power Fluid Pressure

9
The performance of the jet pump is dependent on :
1. The density of the power fluid
2. The nozzle area (An)
3. The throat area (At)
4. The power fluid rate and pressure
5. The head requirements i.e. reservoir performance, tubing performance, wellhead
pressure and power fluid rate.
Note as the power fluid is always mixed with the reservoir fluids it will have to be produced
and this will effect the tubing performance - particularly if a heavy power fluid such as
water is used.
BP have access to a jet pump simulator program called “Jet Pump Simulator”! It is a
simple model using a few vertical lift performance options, a few untuned PVT
correlations, a simple string and a straight line PI. It is however useful for artificial lift
screening and jet pump selection. Various sensitivities can be performed using this model,
and the two jet pump parameters (throat and nozzle areas) can be selected and input as
well as. The program will then calculate the operating point (flowrate) and pressures with
these details. In the example below (Table 4), the power fluid is water and the sensitivities
have been performed on the nozzle and throat combinations

9
Fundamentals of Oilwell Jet Pumping A.W. Grupping, J.L.R. Coppes, J.G. Root SPE 15670 1986

20 October 1998 Page 35

 BP Completion Design Manual - Section 4

Table 4 - Example Output from the Jet Pump Simulator

Nozzle Throat Formation Power Pump FBHP Pump Efficiency

Number identifier Rate Fluid suction (psia) power (fraction)
(stb/d) rate press. fluid inlet
(stb/d) (psia) press.
(psia)
3 0 1750 3939 3922 4297 10098 .222
3 1 2441 3989 3763 4138 10098 .308
3 2 2926 4024 3652 4027 10098 .375
3 3 2991 4029 3637 4012 10098 .385
3 4 2354 3983 3783 4158 10098 .297
3 0 2169 5126 3826 4201 10095 .240
3 1 2925 5196 3652 4027 10095 .331
3 2 3328 5233 3559 3934 10095 .384
3 3 3131 5215 3605 3980 10095 .357
3 4 1940 5104 3878 4254 10095 .215
The throat and nozzle identifiers are the manufacturers classification of the throat and
2
nozzle areas. For example a 3 Nozzle has an area of 0.004 in . Note there are differences
between the manufacturers in their throat / nozzle identifiers and the corresponding sizes.
Contact the well design team in SPR for further details of how to use and access this
program.

4.4.7 Heat Transfer

The accurate modelling of heat transfer is important for a number of reasons:
1. The temperature will effect the well performance as it will determine the viscosity of
fluids and the amount of gas held in solution.
2. Predictions of the maximum wellhead temperature so that this can be used for subsea
or facilities input (e.g. materials, pipeline performance or production chemistry impact).
3. Production chemistry impact in the completion. For example wax, hydrate and
sometimes scale formation will be dependent on the temperature of the fluids being
produced.
4. Materials selection. Certain materials such as elastomers will have temperature limits.
If these materials are used away from the reservoir section it may be unduly
pessimistic to assume that the temperature is still that of the reservoir.
5. For modelling of tubing and casing stresses.
For these reasons, it is worthwhile spending some effort in determining the heat transfer.
Heat transfer will comprise:
1. The transport of fluids by production or injection
2. Conduction between the fluids, the tubing, the casing, cement and the rock.
3. Convection of fluids in the annuli
4. Phase to phase transfers - e.g. oil to gas
5. Expansion or contraction of fluids (especially gases)
We have three main choices for heat transfer modelling:
1. Assume a fixed temperature profile e.g. as extracted from well test data. This will often
be sufficient for determining the pressure drops in the completion so long the flowrates
are similar to the well test data.

20 October 1998 Page 36

 BP Completion Design Manual - Section 4

2. Assume a constant heat transfer coefficient. Prosper has the capability to model heat
transfer with one overall heat transfer coefficient. Typical values are 4-6 BTU/h/ft2/F.
The heat capacity of the fluids is also required. The heat capacity of the oil will vary
considerably (factor of 2) between a light and a heavy oil. This is probably the best
method to use where there is some temperature data from well tests. The well test
configuration can be modelled and a trial and error approach used to match the heat
transfer coefficient with the measured temperature data. Care should be taken with
offshore wells however as an overall heat transfer coefficient will likely over simplify the
model.
3. Model the well in detail. Either Prosper or Welltemp (part of the Wellcat) package can
be used for detailed temperature modelling. Considerable detail is required including
all casing strings, annular contents, and rock types and properties. Unless all of this
information is rigorously included, the accuracy will not be any higher than option 2
above. This option is the only method that will allow an estimate of the transient heat
transfer effects rather than steady state.
Further details on heat transfer is contained in section 4c.

4.4.8 Erosion
Corrosion and erosion are fundamentally related to the flowrate and flow type. Therefore
Prosper is the ideal tool in which to assess the chances of erosion or erosion-corrosion.
There are no exact or theoretical relationships between the flowrate and the resulting
erosion. However, BP have further developed the empirical and conservative API erosion
limit with corrections for flow type, metallurgy, corrosion and sand.
For multiphase flow, a C factor is used, where:
C
VE =
ρm
Where ρm is the mixture density in lbs/ft , VE is the erosional velocity limit (ft/s) and C
3

depends on the metallurgy and flow type.

The erosional velocity (ft/s) should be compared against the mixture velocity (V m), where:
qL + qg
Vm = = vsl + vsg
Ap
3
Where: qL and qg are the volumetric flowrates of liquid and gas (in ft /s)
2
Ap is the cross sectional area (ft )
vsl and vsg are the superficial liquid and gas velocities
The superficial liquid and gas velocities and the mixture density can be extracted from
Prosper by generating the full output from the gradient calculations. Using Excel, this can
be converted into a C factor vs. depth. This can then be compared against the C factors
suggested in Table 5.
Table 5 - Erosional Velocity Limits

Gas Liquid Multiphase

Nominally Sand V=70 m/s (if dry and C=250 (c/s), 300 C=135 (c/s), 300
Free* completely free of solids) (13Cr), 350 (DSS) (13Cr), 350 (DSS)
Sand Containing** Contact SPR for more C=150 (c/s), C=200 C=100
details (13Cr and DSS)
10
Table 5 has been taken from the BP Erosion guidelines .
Notes to the table

10
Erosion Guidelines Revision 2.0 (1996) ESR.97.ER.002 John Martin BP Sunbury

20 October 1998 Page 37

 BP Completion Design Manual - Section 4

* Sand free is defined as less than 1 parts per thousand barrels (pptb)
** The present velocity limits for "sand containing" conditions should only be applied up
to 275 pptb. Above this level the appropriate specialist should be consulted.
C/S - carbon or low-alloy steels, 13Cr - 13%Cr steels, DSS - duplex stainless steels
An example of extracting the C factor from Prosper is shown in Figure 17.
Figure 17 - C factor vs. Depth Extracted from Prosper for a Variety of Conditions

900

800
10,000 bpd Gas Oil Ratio 2000.0 (scf/STB) Tubing/Pipe Diameter 3.80 (inches)
10,000 bpd Gas Oil Ratio 6000.0 (scf/STB) Tubing/Pipe Diameter 4.80 (inches)
10,000 bpd Gas Oil Ratio 6000.0 (scf/STB) Tubing/Pipe Diameter 3.80 (inches)
700 10,000 bpd Gas Oil Ratio 2000.0 (scf/STB) Tubing/Pipe Diameter 4.80 (inches)
5,000 bpd Gas Oil Ratio 2000.0 (scf/STB) Tubing/Pipe Diameter 4.80 (inches)
5,000 bpd Gas Oil Ratio 6000.0 (scf/STB) Tubing/Pipe Diameter 3.80 (inches)
C=236
600 C=100

500
C factor

400

300

200

100

0
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Depth (m)

20 October 1998 Page 38

 BP Completion Design Manual - Section 4

4.4.9 Tubing Size Selection

4.4.9.1 Natural Flow Production

There are various rules of thumb for selecting tubing sizes. As a first pass, Table 6 can be
used:
Table 6 - Flowrate Range for Common Tubing Sizes
Tubing Liquid Flowrate Gas Flowrate Range
Size Range
3
2 /8" Less than 1 mbd Less than 2 mmscf/d
7
2 /8" 1-3 mbd 2-8 mmscf/d
1
3 /2" 2-6 mbd 3-14 mmscf/d
1
4 /2" 3-10 mbd 6-30 mmscf/d
1
5 /2" 5-17 mbd 8-50 mmscf/d
7" 7-25 mbd 12-150 mmscf/d
5
9 /8" 15 + mbd 20+ mmscf/d
Note that these ranges are very generic and especially in gas wells, the choice of tubing
size is very dependent on the wellhead pressure.
A more rigorous method is to look at the minima in the liquid rate vs. pressure plot for the
minimum flowrate and the system performance plot for the maximum. As discussed in
section 4.4. this predicts a change a flow regime. As a rule of thumb, avoid conditions
where the flowrate is below this minima. However, this does not mean the well will not flow
below this point. As shown in Figure 18, a high PI well will likely not flow below the minima.
This is because the two curves do not intersect. A high pressure, low PI reservoir on the
other hand will be able to flow because it only intersects the tubing curve at one point.
However the flow is likely to be slug or severe slug. A moderate PI and pressure well
actually intersects the curve at two points. The left hand point is unstable in that any
perturbation will feedback and push the flowrate either higher or lower. Thus starting the
well off may be difficult. Once started, flow should be possible, but again slugging is
possible. As the two curves intersect at a low angle, this slugging could cause severe
instability.

20 October 1998 Page 39

 BP Completion Design Manual - Section 4

Figure 18 - Using Tubing Performance Curve Minima

Bottom Hole Pressure (psia)

Hig
h
Pr
es
su
re,

Mo rve
Lo

de
rat e Cu
w

e nc
PI ma
PI

an erfor
dp P
res ing
su Tub
re
Low Press
ure, High
PI

Liquid Rate (bpd)

Both the minimum and maximum rates of a range of tubing sizes can be plotted on one
graph (Figure 19).
Figure 19 - Example of Minimum and Maximum Flowrates for Various Tubing Sizes
8000

7500

7000 Initi
al R
6500 Pre eservo
ssu
re ir
6000 /2"
41
/8"

/2"

5500 /2"
27

31

51
Pressure
(psig)

5000
7"
4500

4000
Res
ervo
3500 afte ir Pres
r5y s
Res ears ure
erv
3000 afte oir Pre
r 10 ss
yea ure
2500 rs

2000
0 6000 12000 18000 24000 30000
Liquid Rate (STB/day)
In this example, the reservoir depletion is indicated by the changing inflow performance
curves. The minima of the tubing curves is also very clear, with 7" having a minima at
1
9,000 bpd, whereas 3 /2" tubing has a minima at around 2,000 bpd. The maximum rates
1
for initial reservoir pressures are 25,000 bpd for 7" and 11,000 bpd for 3 /2". However the
constraints of the erosional velocity must also be superimposed on this model.

20 October 1998 Page 40

 BP Completion Design Manual - Section 4

With initial reservoir pressure, all the tubing sizes are stable, likewise after 5 years.
However after between 5 and 10 years, the wells will cease to flow on their own. In this
instance there is not much difference in when any of the tubing sizes shown become
unstable. This is because the productivities are high. With lower productivities, or for
example gas break out in the reservoir, the benefit of a smaller tubing size may be more
apparent.
For gas wells, the same approach can be used, however great care must be taken with
selecting the PVT representation of the gas. The PVT section (4.2) discusses the pitfalls
in the various Black-oil correlations and their assumption that the gas density can be
increased to represent the condensate.
The most important point about using this method is that care must be taken with the flow
correlation. The correlations section (section 4.4.1) indicates which correlations can be
used to accurately predict the instability region.
The next improvement in this approach that could be made is to examine the flow
behavior to the left of the minima to see if it is operationally acceptable to facilities. This is
covered in section.
One aspect of well performance in producers that is often overlooked is the location of the
bubble point and the impact that this has on tubing size. There will be no minima in the lift
curve if the fluid is single phase (i.e. all liquid or all gas). In these circumstances, the
bigger the tubing the better (from a well performance stand point). For example, in one
field, the bubble point was only 90 psi, therefore the fluid was single phase all the way to
surface. The development plan proposed had a 7" liner, but 4.5" tubing as this was
appropriate to the production rates expected. In this case, 7" tubing would not only have
reduced pressure drops slightly, but would have led to a monobore completion, with the
only downside, the increased cost of the 7" tubing and tree.
The same approach can be used for liner sizing. If the fluids will be single phase in the
liner and a long liner is to be used, then from a well performance stand point, a large liner
(e.g. 7") can be accepted. However this is not the only consideration for liner sizing and a
full discussion in contained in section 5.

4.4.9.2 Tapered Strings

Tapered strings can be useful. With the lowest pressure at surface, any evolved gas will
increase and expand the higher up the tubing one goes. Therefore frictional pressure
drops will be highest close to surface. At this point however, velocities will also be higher
and therefore larger tubing close to surface can not only be tolerated better, but has the
greatest effect.
How does one determine the depth of a tubing size change? Figure 20 shows a plot of
tubing performance for a 1000 ft section of tubing at different pressures flowing with a
moderate GOR. The top curve is at 4000 psi, the bottom curve at 400 psi and the other
curves every 400 psi. The two curves at each pressure are for 5.5" and 7" tubing. Note
that as the pressure increases, the minima of the curve moves to the right. At lower
pressure, the pressure drops are also highest at the higher rate. These two phenomena
indicate that there is a benefit of using the larger tubing size close to surface. In this
example, if the tubing was 7" for pressures less than 800 psi and 5.5" for pressures
greater than this, the completion would give stable flow down to around 5,000 bpd. This is
the same as if the tubing was 5.5" throughout. However the pressure drops have been
minimised.

20 October 1998 Page 41

 BP Completion Design Manual - Section 4

Figure 20 - Tubing Performance Curves at Different Pressures

Pressure (psig)

2000 6000 10000 14000

Liquid Rate (STB/day)

4.4.9.3 Pumped Production

The effect of tubing size on pumped production wells is much less than in the case of
natural flow wells or gas lifted wells. Generally, so long as there is no instability or slugging
below the pump, and the pump can deal with a large head of liquid above the pump, there
should be minimal problems in using a larger tubing size than for the case of a naturally
11
flowing well. Texaco have reported problems in pumps placed above long high angle
wellbores where slugs of water have built up in an undulating well profile. The danger here
is that if there was free gas, a slug of gas would cause gas lock or severe pump wear.

11 th
Uses of Reliable Downhole Monitoring In the Captain Field D.J. Cohen, 5 European ESP Round Table 1998

20 October 1998 Page 42

 BP Completion Design Manual - Section 4

If severe slugging occurs above the pump, there is the potential for the wellbore to fill with
essentially pure liquid. Centrifugal pump stages are designed to be able to handle such a
condition, however this will increase the power requirement of the motor and must be
addressed in the ESP or hydraulic submersible pump design. The slugs are then a
problem for the facilities rather than a potential cause of the well dying. If the slugging
results in the loads on the pump varying, then in a severe case, this could lead to the
pump being periodically operated outside of the upthrust / downthrust envelope and
increased pump wear occurring. In the case of positive displacement pumps, the slugging
may place additional loads on the pump and rods and again the pump unit should be
designed to be able to handle a 100% liquid (water) head.
This phenomena means that much larger tubing can be effectively used. This is of
significant value in the case of electrical or hydraulic submersible pumps deployed inside
5 5
the tubing by means of coiled tubing or cable. The use of 7", 7 /8" or even 9 /8" tubing then
becomes possible.

4.4.9.4 Gas Lift Production

For gas lift conditions, the choice of tubing size is again a balance:
• Too small a tubing size and adding gas may increase friction more than it decreases
the hydrostatic pressure. The result could be a well that flows better without gas lift, or
where the production rate is severely constrained.
• Too large a tubing size and the well may suffer from heading problems or instability.
The first issue is easy to quantify using Prosper. The second issue can be addressed
using Dynalift.
Dynalift is a transient well performance model. When set up with details of the well, the
inflow performance, the gas lift valve etc. it can calculate likely performance in terms of
slugging and heading. Dynalift is available for BP from Rob Perry.

20 October 1998 Page 43

 BP Completion Design Manual - Section 4

4.5 Well and Field Optimisation

Well optimisation does not just apply to gas lifted or pumped wells. Wherever there is a
control that will effect the well rates or a resource (e.g. power), that control should be
optimised. For example a single naturally flowing well producing to a fixed pressure facility
is controlled and optimised by the choke, a gas lifted well is optimised by the gas lift
injection rate, an ESP by the electrical frequency and the choke and a jet pump by the
power fluid injection rate. Note in each of these examples, the control can be changed at
will and as from surface. There is also another form of optimisation that can be applied for
well interventions. For example, the tubing size can be changed with a tophole workover,
the pump changed out or the gas lift valve operating depth altered.
On a single well basis, optimisation is fairly straight forward. The optimum rate is the one
where the production cash flow is maximised. Note that this doesn’t necessarily mean the
highest production rate as the expenditure used to get the fluids from the reservoir to the
point of sale must be accounted for. For example, the rate can be increased by fully
opening the choke and lowering the separator pressure, however this may introduce
additional pumping costs or result in additional flaring. The same applies for gas lifted or
pumped production where the gas compression or power requirements must be
addressed. Figure 14 shows an example of optimising a single gas lifted well.
Where more than one well is involved, the process becomes a lot more complex. This
time the field cash flow must be maximised. This must account for:
1. The value of the oil production
2. The cost of water production
3. The effect of increased production on pipeline / flowline pressure drops
4. The cost of power or gas compression
5. Facilities limits (oil, water, gas, power)
6. Downhole production constraints (asphaltene, scale, multiphase in the reservoir etc.,
coning etc.)
7. Facilities performance - e.g. what happens if the separator pressures are changed
In Figure 21 and Figure 22, an example optimisation is shown for 4 gas lifted wells with
varying water cuts. The first plot shows the performance curves for 4 wells. Note that two
of the wells can not flow without the assistance of gas lift. In order to optimise the
allocation of a limited amount of gas to these wells, care must be taken. Figure 22 shows
the actual optimised system. Note that it is not simply a case of adding gas to the well
where there is the most benefit. Well D actually benefits the most (in terms of liquid rate)
from gas lift due to its high water cut, but it is best to leave this well shut in until 12
MMscf/d of lift gas is available. At this point, some of the other wells have to be slightly
robbed of gas in order to get this well going.

20 October 1998 Page 44

 BP Completion Design Manual - Section 4

Figure 21 - Gas Lift Performance Curves for 4 Wells

Well A Oil
14000
Well B Oil
12000 Well C Oil
Well D Oil
10000
Oil Production

8000

6000

4000

2000

0
0 1 2 3 4 5 6 7 8
Gas Lift Injection Rate

20 October 1998 Page 45

 BP Completion Design Manual - Section 4

Figure 22 - Example of a Gas Allocation System for these Same 4 Wells

20 70000
Well D Gas Injection Rate
Allocated Gas Injection Rate Well C Gas Injection Rate
Well B Gas Injection Rate
Well A Gas Injection Rate
15
Well D Production Rate 60000
Well C Production Rate
Well B Production Rate
Well A Production Rate
10
50000

40000

0 10

12

14

16

18

20
0

30000

-5

optimised Production Rate

20000
-10

10000
-15

-20 0
Gas Injection Rate

4.5.1 Practical vs. Theoretical Optimisation

There are two means of optimising a field performance:
Practical Methods
This relies on well tests in order to measure the effect of changing control values on the
well or facilities. For example in order to assess the effect a change in separator pressure
has on a well, a multi-rate well test can be performed and a plot of wellhead pressure vs.
rate produced. The wellhead pressure can then be mapped to the separator pressure.
This approach relies on accurate well tests over a range of conditions for each well and
then a method (by hand or computer) of optimising the total field performance. The well
tests would have to be repeated on a regular basis and whether conditions changed. Such
a situation is not always possible due to the lack of a test separator or it being required for
other purposes or the shear number of wells.
Theoretical Methods
Prosper and GAP have the ability to model fairly complex fields with a variety of different
wells and to optimise the allocation of resources in the optimum way.

20 October 1998 Page 46

 BP Completion Design Manual - Section 4

In order to do this accurately, the wells and facilities must be accurately modelled and the
performance predictions must be accurate. Particular care should be exercised in tuning
the lift curve correlations on individual wells. Personally I recommend that a generic well
model is constructed to best match all the wells and that if tuning is required it is then
based on a large dataset and individual measurement errors on each well should not
grossly effect the result. See section 4.4.3 for further details on tuning correlations.
Combined Method
A theoretical model such as GAP is flexible in that it can offer predictions and answers to
what-if scenarios and is easily changed. It is vital that the model is however compared
with reality. I particularly recommend the use of multi-rate well tests on gas lifted wells as
frequently there can be discrepancies between theory and reality. If there are
discrepancies then these can be investigated in terms of the model, facilities measuring
errors and the likes.

20 October 1998 Page 47

 BP Completion Design Manual - Section 4

4.6 Glossary
BHP .................................... Bottom hole pressure
BHT .................................... Bottom hole temperature
CGR ................................... Condensate gas ratio (the volume of dissolved liquid held in
the gas)
FVF..................................... Formation volume factor. This is the ratio between the volume
of a fluid at reservoir conditions and at standard conditions.
The is also called Bo
GOR ................................... Gas Oil Ratio (the volume of dissolved gas held in solution in
the liquid)
head ................................... The height a column of fluid can be supported. It is related to
the pressure by the density of the fluid
Liquid hold-up..................... The fraction of the pipe occupied by the liquid
IPR ..................................... Inflow Performance Relationship
Solution GOR ..................... The ratio of gas remaining in solution to the oil
Slippage ............................. The phenomena whereby two phases travel at different
velocities
Superficial Velocity ............. The velocity of a phase assuming that it takes up the entire
pipe area
THP .................................... Tubing head pressure
THT .................................... Tubing head temperature
UCS.................................... Unconfined compressive strength
VLP..................................... Vertical Lift Performance (or tubing performance)
w/c .............................. Water cut or base sediment and water (BS&W)

20 October 1998 Page 48

BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

CONTENTS

INTRODUCTION 1

IDEALIZED RADIAL FLOW EQUATION 2

RADIAL FLOW CHARACTERISTICS 2
PARAMETERS IN RADIAL FLOW EQUATION 8
OIL WELL PRODUCTIVITY EQUATION 16
GAS WELL PRODUCTIVITY EQUATION 18
EFFECTS OF WATER PRODUCTION 23
EFFECTS OF PRESSURE DEPLETION 29
EFFECT OF GAS PRODUCTION IN OIL WELLS 32
MULTIZONE IPRs 39

REFERENCES 41
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

INTRODUCTION

The completion engineer must have an understanding of the

basic equations controlling inflow performance in order to:

• Develop inflow performance relationships (IPRs).

• Validate IPRs produced by reservoir simulation, and to

ensure that rate dependent and near wellbore skin effects
have been incorporated in a realistic manner.

• Understand the probable variations in well IPRs that

can␣ be expected across the field and during the life of
the␣ project.

• Evaluate the impact of various completion, stimulation

and workover techniques on the skin effects and, hence,
on well deliverability.

• Appreciate the major areas of uncertainty and therefore

the probable range of IPRs that can be expected from any
given well.

• Define the magnitude of the skin effects involved in

unsatisfactory well performance and, hence, identify
opportunities for increasing production by reperforation,
stimulation or improved completion practices.

• Determine if adequate provision has been made for the

factors that cause actual IPRs to differ from the idealized
model used in many simple reservoir engineering
calculations, eg perforation configurations (refer to
Section␣ 4b – Near Wellbore Performance), two-phase and
non-Darcy flow effects and the effects of saturation and
pressure changes over time. These will predominantly be
seen as skin effects in well test results, but may not
indicate formation damage.

• Develop IPR estimates for infill wells and workovers based

on proper modification of the well test data from offset and
appraisal wells.

The basic theory of reservoir performance is presented in this

section, while the effects of the near wellbore performance on
the total skin are detailed in Section 4b.

Page 1
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

IDEALIZED RADIAL FLOW EQUATIONS

RADIAL FLOW
CHARACTER-
ISTICS
Flow from reservoirs is characterized as transient, pseudo-
steady state or steady state flow depending on whether the
pressure response initiated by opening the well has reached
the drainage area boundary and on the type of boundary.
Transient flow occurs when the well is initially opened (or has
a significant rate change), and is a result of the pressure
disturbance moving towards the outer boundary of the
drainage area. During this period, production conditions at
the wellbore change rapidly, and the bottom hole flowing
pressure (BHFP), pwf , decreases exponentially with time.
Most drill stem tests (DSTs) and many well tests are
conducted under transient flow conditions and consequently
the observed productivity will often be greater than that
which will be seen in long-term production. Therefore, to
compute meaningful IPR values from short duration well
tests, appropriate corrections must be applied to compensate
for the transient flow behaviour, as well as the possible
variation in the skin effects.
When the disturbance meets the outer boundary, flow
becomes steady state or pseudo-steady state. If the boundary
is a constant pressure boundary, the reservoir pressure pR
will not change with time, and the flow is termed steady
state. If the boundary is a no-flow boundary, the reservoir
pressure pR will decline as a result of depletion, and the
resulting flow is termed pseudo-steady state. Steady state
equations should be used for oil wells where a waterflood is
in operation, or there is a strong natural water drive. Pseudo-
steady equations are used in the other cases. For a more
detailed discussion on this topic see Sections 2.2, 2.6 and 2.7
(*Reference 4a.1) of Golan & Whitson*.
Stabilization
Time
When the BHFP, pwf , appears to be constant, or declining
only slowly proportionately with time, the reservoir is said to
have achieved stabilized flow conditions. Under these
circumstances, the pseudo-steady state flow equations
presented in this section may be used to predict the long
term deliverability of a well.

Page 2
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

TRANSIENT
TRANSIENT

PSEUDO-STEADY PSEUDO-STEADY
pwf STATE p
wf STATE

log t t

Figure 4a.1 – Variation of pwf with Time

The time at which the flow regime changes from transient to

pseudo-steady state depends on the size and shape of the
well’s drainage area. For a well in the centre of a circular
reservoir, the stabilization time, ts, can be estimated from:

θ µ ct r e2
(Equation 4a.1) t s = 948 (hours)
k
For wells producing from reservoirs of different shapes, the
stabilization time is given by:

θ µ ct A
(Equation 4a.2) t s = 3790 tDApss (hours)
k
where: θ = porosity, fraction
µ = viscosity, cp
ct = total fluid compressibility, psi-1
re = drainage radius, ft
A = drainage area, ft2
k = effective permeability, md
t DApss = the dimensionless time to stabilization for
different geometry and well placements
(*Reference 4a.1)
(see Table 2.4 of Golan and Whitson* or
(*Reference 4a.2) Table C.1 of Earlougher*)

Page 3
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Typical time to stabilization for different reservoir conditions

is indicated below:

• Prolific oil well producing from 5000 md reservoir in the

centre of a 300 acre circular drainage area – 4 hours

• Moderate permeability gas well producing at 10 000 ft

from a 50 md reservoir in the centre of a 640 acre
drainage square – 8 days

• Stripper oil well producing from a 10 md reservoir in the

centre of a 40 acre drainage square – 4 days

The Radial
Flow Equation
Consider an idealized well drilled in a fully oil bearing
reservoir rock of constant thickness and uniform
permeability. Assuming the producing interval is completely
penetrated by a circular wellbore, oil production will be
characterized by radial flow geometry. Under pseudo-steady
state conditions, the reservoir pressure declines very slowly
with cumulative production and Darcy’s equation for a
circular drainage area when expressed in practical oilfield
units, becomes:

p R – p wf
(Equation 4a.3) q o = 0.00708 k oh (stb/d)
µo B o
ln .472 rr e + S'
w

where qo = oil rate, stb/d

pR = volumetric average reservoir pressure, psi
pwf = bottom hole flowing pressure, psi
ko = effective oil permeability, md
h = net pay, ft
µo = live oil viscosity at reservoir conditions, cp
Bo = formation volume factor, bbl/stb
re = drainage radius, ft
rw = wellbore drainage radius, ft
S' = total skin effect, dimensionless (S' = S + Dq).
This accounts for departures from ideal flow
conditions, see Section 4b – Near Wellbore
Performance. (If there are no rate dependent
skin terms, S may be substituted for S'.)

For a more detailed derivation of Equation 4a.3, refer to Golan

(*Reference 4a.1) and Whitson*, Section 2.2.

Page 4
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

In some petroleum engineering texts, Equation 4a.3 may be

written as:

k oh p R – p wf
(Equation 4a.4) qo = (stb/d)
141.2 µ oBo ln rr e – 0.75 + S
w

Both expressions are equivalent mathematically. Equation

4a.4 is merely a rearrangement of the logarithm term of
Equation 4a.3 and assumes no rate dependent skin effects.
The steady state equation is obtained by simply replacing the
–0.75 in the above equation by –0.5 (or the 0.472 in the log
term by 0.606) and should be used where a waterflood or a
strong natural water drive is present.
Gas Well Radial
Flow Equation
Darcy’s law may also be used to describe the inflow
performance of a natural gas well under pseudo-steady state
conditions:

0.0397 k gh p R – p wf
(Equation 4a.5) qg = (scf/d)
µg Bg ln .472 rr e + S'
w

where: qg = gas flowrate, scf/d

kg = effective gas permeability, md
µg = gas viscosity, cp
Bg = gas formation volume factor, cf/scf

Since gas properties have a strong pressure dependence, the

gas formation volume factor, Bg , and the gas viscosity, µg, are
normally incorporated with the pressure in a single term
(*Reference 4a.3) called the real gas pseudo-pressure, m(p). Al Hussainy et al*
defined this as:

p
p
(Equation 4.18) m(p) = 2 dp
pb
µZ

where the integration limits are substituted with the pressure

range being considered, normally pR and pwf for inflow
calculations.

Page 5
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Hence:

k gh m p R – m p wf
(Equation 4a.6) qg = (mscf/d)
1422 T ln rr e – 0.75 + S'
w

where: qg = gas flowrate in mscf/d, ie 103 scf/d

Z = gas deviation factor, dimensionless
T = reservoir temperature, °R
S' = total skin including non-Darcy effects,
dimensionless

Although the pseudo-pressure integral appears daunting,

it␣ may be easily calculated by numerical integration as
(*Reference 4a.4) illustrated in the following table after Dake (Table 8.1)*.

CALCULATION OF PSEUDO-PRESSURE INTEGRAL

Pseudo-
PVT Data Numerical Integration Pressures
p
µ 2p 2p 2p 2p
p Z ∆p × ∆p m = Σo ∆p
µZ µZ µZ µZ
(psia) (cp) (psia)2/cp
400 .01286 .937 66391 33196 400 13.278 x 106 13.278 x 106
800 .01390 .882 130508 98449 400 39.380 x 106 52.658 x 106
1200 .01530 .832 188537 159522 400 63.809 x 106 116.467 x 106
1600 .01680 .794 239894 214216 400 85.686 x 106 202.153 x 106
2000 .01840 .770 282326 261110 400 104.444 x 106 306.597 x 106
2400 .02010 .763 312983 297655 400 119.062 x 106 425.659 x 106
2800 .02170 .775 332986 322985 400 129.194 x 106 554.853 x 106
3200 .02340 .797 343167 338079 400 135.231 x 106 690.084 x 106
3600 .02500 .827 348247 345707 400 138.283 x 106 828.367 x 106
4000 .02660 .860 349711 348979 400 139.592 x 106 967.958 x 106
4400 .02831 .896 346924 348318 400 139.327 x 106 1107.285 x 106

2p is the average of 2p for the current row and the previous row.
µZ µZ
Table 4a.1

The pseudo-pressure approach is valid for all pressure

ranges, and given that most test analyses and deliverability
estimates are today conducted using computer software,
there is no real reason not to use this correct approach.

Page 6
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

However, where the use of pseudo-pressures is not feasible or

convenient (eg in the field), the pressure squared approach is
still sometimes used. In this case, the gas formation volume
factor, Bg, is evaluated at the average of the reservoir and
flowing pressures:

(Equation 4a.7) Bg = ZT
pR + pwf
35.37
2

Substitution of Equation 4a.7 into Equation 4a.5 and

collection of the two sets of pressure terms,

pR + pwf p 2 – p wf 2
p R – p wf = R
2 2

yields:

k gh p R2 – p wf 2
(Equation 4a.8) qg = (mscf/d)
1422 µ g ZT ln .472 rr e + S'
w

where: qg = gas flowrate, mscf/d

S' = effective skin including turbulence effects
T = reservoir temperature, °R
µg = gas viscosity at the average of pR and pwf , cp
Z = gas deviation factor at the average of pR
and pwf

Amplification of the derivation of Equation 4a.8 can be found

(*Reference 4a.1) in Golan and Whitson*, Section 2.4.

Note that the gas well IPR will always be a curved function
of pwf .

This pressure squared form of the IPR equation is only

accurate for pR <1500 psi (depending on the gas composition)
although if only small pressure changes are involved (as in
high permeability wells), the results are usually satisfactory
even outside this range.

Page 7
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

PARAMETERS
IN RADIAL
FLOW
EQUATION

Permeability
The absolute permeability, k, of a reservoir rock is dependent
primarily on the rock structure and the effective confining
stress. It is a constant so long as one single-phase fluid (oil,
water, gas or air) saturates the entire porous media, flowrates
are low and the rock stress remains constant.

The existence of a second fluid blocks off a portion of the flow

channels and causes greater resistance to the flow of one
fluid in the presence of the second fluid. Since hydrocarbon
reservoirs contain more than one fluid, the ease with which,
for example, oil moves in the presence of connate water (and
perhaps free gas) is defined as the effective oil permeability
(ko) used in the radial flow equation. It is important to
recognize that the sum of effective permeabilities in a multi-
phase system is always less than the single-phase absolute
permeability measured in routine core analysis.

A D Where:
A = k o at S w = 0. Absolute permeability of the rock (k)

B = Connate water saturation S wc (usually 0.1 to 0.5)

EFFECTIVE PERMEABILITY

EFFECTIVE PERMEABILITY

X = k w at S w = S wc . Effective permeability of oil at

kw
connate water saturation. Frequently referred to
TO WATER, kw

X as the end point relative permeability to oil. Notice

TO OIL, k o

that the effective permeability of water at this

point is zero.
Y
ko
C = Residual oil saturation S or (usually 0.1 to 0.5)

Y = k w at S w =(1-S or) . Effective permeability of water

at residual oil saturation. Frequently referred to
as the end point relative permeability to water.
Notice that the effective permeability of oil at this
point is zero.
0 B C 1 D = k w at S w = 1. Absolute permeability of the rock (k)
WATER SATURATION, S w
At any value of S:
k o + k w = k e< k

Figure 4a.2 – Effective Permeability Curves

Page 8
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

The concept of relative permeability is useful for quantifying

the changes in effective permeability of an individual phase
under conditions of changing fluid saturations. Oil relative
permeability is commonly defined as follows:

effective oil permeability

kro = ko =
k absolute rock permeability

Similar expressions can be written for water and gas relative

permeabilities, krw and krg, respectively. Relative
permeabilities are, in fact, ratios whose values are always ≤1.

Permeability measurements routinely conducted on core

samples are measured under unstressed, single-phase
conditions. Therefore, the results from conventional core
analysis must be corrected to account for differences in
stress levels and fluid saturations between the laboratory and
formation conditions. The correction factors are obtained
from special core analysis and vary with rock type and depth.
Generally, the correction factors range between 0.2 and 0.8.
Core air permeabilities should also be reduced for the
Klinkenberg effect, where gas slippage at high flowrates
makes apparent core permeabilities too high.

In-situ kh estimates derived from well test and DST data are
therefore preferred when making theoretical IPR calculations.
Even then, well test kh values may have to be adjusted to
take into account:

• The downhole commingling of several zones that were not

in communication during the limited test period.

• The variations in thickness, h, saturation, Sw, and sand

quality,k, between the test well and its offsets. For this,
trends will be mapped in regions of good well control,
often using some geological model for guidance. Although
no true mathematical relationship exists between
porosity, θ, and permeability, k, for sandstone reservoirs,
a semi-log relationship is often assumed. On large pools,
this assumption may be made during reservoir modelling
and the results used to predict kh (or the well IPRs). The
variation of IPR with time is assumed to depend on the
effect of saturation changes on the effective permeability,
changes in reservoir pressure and the effect of workovers
which plug off lower zones.

Page 9
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Net Pay
Thickness
The net pay thickness, h, of a producing formation is the
average over the drainage area. It is usually based on a
combination of the analysis of open hole logs and the
geologist’s isopach maps (lines of equal pay thickness).
Portions of the formation which will not contribute
substantial quantities of hydrocarbons should be excluded.
These can usually be identified based on high water
saturations (from a resistivity log), poorly developed
porosities (from the density, neutron and sonic logs) and
poor sand development (from gamma ray (GR) or
spontaneous potential (SP) logs). Varous ‘net-to-gross’
criteria are used to define these ‘non-pay’ portions. The ratio
of net pay to total or ‘gross’ pay in a well is known as the
‘net-to-gross’ ratio. In a deviated well, or a well intersecting a
steeply dipping formation, the net pay value must be
adjusted to remove the apparent thickening caused by not
penetrating the pay zone perpendicular to the bedding plane.

Net pay, h, should not be confused with the perforated

interval, hp. Generally speaking, perforation of only a portion
of the pay section will be seen in the skin term and will not
affect the kh value unless vertical permeability is very poor
due to shale layers or tight streaks being continuous. The
net pay effectively contributing to a specific set of
perforations will depend, to great extent, on the geological
model, the shale and sand continuity, the presence of any
vertical fractures and the degree of effective zonal isolation
by the cement.

Oil Formation
Volume Factor
Oil formation volume factor, Bo, is defined as the ratio of the
volume of oil (plus gas in solution) at reservoir pressure and
temperature to the volume of oil at standard conditions in
the stock tank. Hence:

Bo ≥ 1.0
Shrinkage = 1/Bo

Bo data should be obtained from the reservoir group but can

also be read directly from PVT analyses or from correlations
(*References 4a.5 and 4a.6) such as those published by Standing (1952)* and Glaso*. A
good set of correlations is provided in MULTIFLO. Refer to
Chapter 4 Theory – Outflow Performance.

Page 10
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

It is important that the proper solution GOR is used in

determining Bo. This may be considerably below the
producing GOR measured on well tests if the formation is
producing free gas or if the well is on gas lift.

The gas formation volume factor, Bg, is a function of

pressure, temperature and Z, and its derivation is discussed
in the section on the gas well radial flow equation.
Fluid
Viscosity
The viscosity term, µ, appearing in Darcy’s law always
represents the behaviour of the fluid at reservoir conditions
and, hence, must be corrected for the effects of temperature,
pressure and dissolved gas or liquids. It is also obtained from
the reservoir engineers, PVT reports or correlations.

Dead oil viscosity (which contains no dissolved gas) is

primarily a function of oil gravity and reservoir temperature.
The heavier the oil, the greater the sensitivity of oil viscosity
to temperature. For example, heating a 15° API oil containing
little or no gas from 100°F to 130°F will decrease its dead oil
viscosity from 700 cp to 200 cp. Higher temperatures will
continue to decrease the dead oil viscosity (30 cp at 220°F;
3 cp at 500°F) which is the reason for steaming heavy oil
wells.

In light and medium oils, the viscosity of live oil (ie containing
dissolved gas) is directly dependent on the volume of
(*Reference 4a.7) dissolved gas (refer to Beal* and MULTIFLO for dead and live
oil viscosity correlations, respectively).

The viscosity of formation water is a function of temperature

and dissolved solids. µw is considerably lower under reservoir
(*Reference 4a.8) conditions than at surface (refer to Matthews and Russell*).

The viscosity of free gas is far below that of liquids. Since gas
is extremely compressible, its viscosity is strongly dependent
on both reservoir temperature and pressure as well as on the
gas composition and gravity. In the absence of laboratory
measured PVT data, µg can be estimated from gas
pseudo-reduced pressure and pseudo-reduced temperature
using the correlations developed by Carr, Kobayashi and
(*Reference 4a.9) Burrows*.

Page 11
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Reservoir
Temperature
The reservoir temperature, TR, may be extrapolated from the
data recorded on well logs or recorded during testing. The
temperature of a reservoir can also be estimated from the
expression:

TR = TGround + geothermal gradient × depth

Common geothermal gradients are in the range of

1° to 2°F/100 ft but may be as high as 7°/100 ft. For nearly
all reservoir and production calculations, the reservoir
temperature must be expressed in absolute terms,°R:

°R = °F + 460
Drainage and
Wellbore Radii
The drainage radius, re, is primarily a function of well
spacing. It will also depend, to some degree, on the relative
production rates of the well and its offsets.

Wellbore radius, rw, is the radius of the open hole section in

which the casing was set, since this is where radial flow must
terminate. For onshore operations it is common to assume
rw ≈ 0.3 ft, while offshore it is usually between 0.33
and 0.5 ft.

The ratio of re/rw always appears as a logarithmic term and,

as illustrated below, is a relatively insensitive parameter.

Area Drainage Radius rw (ft) ln

(acres) re (ft) (.472 re/rw)

40 660 0.33 6.9

80 933 0.33 7.2
160 1320 0.33 7.5
640 2640 0.33 8.2

For the drainage areas and wellbore sizes normally

encountered in an onshore oilfield (where re ≈ 800 to 1200ft),
the assumption ln (.472 re/rw) ≈ 7 is generally valid. Hence,
Equation 4a.3 can be rewritten as:

Page 12
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

0.00708 k oh p R – p wf
(Equation 4a.9) qo ≈ (stb/d)
µoBo 7 + S'

Since gas is highly mobile compared to oil, gas well spacings

and hence re values tend to be larger (re = 2000 to 2640 ft)
than those of oil wells.

Hence:

ln .472 rr e ≈ 8
w

So that Equation 4a.8 can be approximated by:

k gh p R2 – p wf 2
(Equation 4a.10) qg ≈ (mscf/d)
1422 µ g ZT 8 + S'

Skin Effects
A deviation from the ideal flow model, which predicts a
logarithmic pressure change with distance from the well,
frequently occurs in the near wellbore region and is referred
to as the total skin effect S'.

rs
rw

ACTUAL PRESSURE
PRESSURE WHEN NO
SKIN PRESENT
p

SKIN REGION

∆ pskin = ADDITIONAL PRESSURE

DROP DUE TO SKIN EFFECT
p wf

Figure 4a.3 – Skin Effect

Page 13
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

This effect may be either positive, as a result of formation

damage, non-Darcy flow effects or ineffective or incomplete
perforation of the well’s pay section; or negative as a result of
locally increased fluid conductivities due to stimulation,
natural fractures, very effective perforations or hole deviation.
The various causes of skin and their combined effect on well
productivity are discussed at some length in Section 4b –
Near Wellbore Performance, while the techniques used to
achieve low or negative values are discussed in Section 5a.

The additional pressure loss between the ideal and actual

BHFPs, ∆pskin, serves as the definition of the skin effect, ie:

∆ptotal = ∆pDarcy + ∆pskin (psia)

For convenience, the skin effect, S', is defined in terms of the

parameters of the radial flow equation:

.00708 kh ∆ p skin
(Equation 4a.11) S' =
q µB

Simply stated, the skin effect is a concept which petroleum

engineers have introduced to account for the difference
between the ideal and actual inflow performance. Most of the
influence of the skin effect occurs in the near wellbore region
and is therefore discussed in Section 4b – Near Wellbore
Performance.

Most classical reservoir engineering equations were developed

for low rate wells, hence many textbooks show the total
Darcy (or non-rate dependent) skin, S, in these equations.
However, in high rate and high GLR oil wells and in gas wells,
the effects of high flow velocities must be included as an
equivalent, rate dependent skin (Dq), so that:

(Equation 4a.12) S' = S + Dq

The influence of skin effects on an oil well productivity index,

J, can be quite dramatic. It should be emphasized that the
skin value is the only term in the IPR equation that can be
altered by applying different completion practices. Hence, it is
the skin term which is of major interest to the production
engineers.

Page 14
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

h o = 30ft
6000
k o = 100md
µ o = 0.5cp
5000 B o = 1.5

BOTTOM HOLE PRESSURE, psi

r w = 0.35ft
r e = 1500ft
4000
p R = 3000psi
SKIN = various
3000

s = -5
2000 s=
s 0
=
+5

s
s=
s = +10

=
+1
1000 +2

0
0
0

0
0 2500 5000 7500 10 000 12 500 15 000
OIL RATE , stb/d

Figure 4a.4 – Possible Well PIs for Various Skin Factors

Typically, skin factors can be expected to be as follows:

Situation Typical Skin Factor

Unstimulated, undamaged well 0

Very poorly completed well +20 to +500
Damaged well +2 to +20
Good initial, unstimulated
completion +1 to –1
Lightly acidized 0 to –2
Typical deviated well 0 to –3
Natural fractures, or
small propped frac –3.0 to –5
Large frac in low
permeability formation –5.0 to –6.5
Minimum possible skin –7.0

Page 15
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Reservoir
Pressure and
Bottom Hole
Flowing Pressure
The reservoir pressure, pR, is either found from an RFT log
(for uncased wells) or from pressure build-up test data, with
test data being the preferred method. For production
engineering calculations, it may be necessary to correct
reservoir pressure data from datum depth to the mid-point of
perforations (or top of perforations) using the fluid gradient
estimated for the reservoir.

Pressure correction = (Datum depth – midpoint perforation

depth) × vertical pressure gradient

where depths are vertical depths.

Vertical pressure gradient = SG × 0.433 (psi/ft)

= Measured pressure gradient
along hole/cos θ

where: θ = deviation angle

OIL WELL
PRODUCTIVITY
EQUATION

Productivity
Index

The empirical productivity index, PI, developed in the Theory

section can be derived from the radial flow theory.

The PI (J) is defined as:

q
(Equation 4.1) J =
p R – p wf

Rearranging Darcy’s law, Equation 4.2.

(Equation 4.2) Jo = 0.00708 k oh = k oh

µoBo ln .472 rr e + S' 141.2 µ oBo ln rr e – 3 + S'
w
w 4

(stb/d/psi)

Page 16
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

Using the simplifying assumptions for re/rw as derived in the

section on drainage and wellbore radii:

(Equation 4a.13) J o ≈ 0.00708 k oh (stb/d/psi)

µoBo 7 + S'

In certain applications, the more rigorous form of the

equation may need to be used.

The skin is obtained from well tests along with koh. However,
if the productivity index has changed with time and there is
no change in water cut and the well’s BHFP is above the
bubble point, the new skin can be calculated from the
following equation where Jo is the new oil productivity index.
This equation is derived from 4a.13.

(Equation 4a.14) S' ≈ 0.00708 k oh – 7

µ o Bo Jo

Where considerable variation in pay zone thickness can occur

across a field, it is often useful to tabulate the specific PIs of
the nearby wells:

qo
(Equation 4a.15) Specific PIs = (stb/d/psi ⋅ ft)
p R – p wf h

The average specific PI for an area can be used to predict the

potential of an infill well from the net pay zone thickness
(although no account is taken of permeability variation).

Examination of Equation 4a.13 reveals that many of its terms

are constants, or change only slowly with time.
Consequently, the effects of formation damage, stimulation
treatments and/or changes in completion efficiency can be
represented by the equation:

(Equation 4a.16) J 1 = 7 + S'2

J2 7 + S'1

This equation can be used to estimate the impact of an

improved completion technique or a workover operation on
well inflow, so that a well deliverability improvement can be
estimated for use in the supporting economics.

Page 17
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

As discussed previously, the total skin term, S', includes not

only formation damage and completion effects, S, but also
both two-phase flow and turbulence effects. Therefore, in
many wells, the total skin, S', is not a constant but varies
with drawdown and rate, leading to curvature of the well’s
IPR. If this is significant, other equations will be needed to
describe the shape of the IPR. These are discussed in the
Theory section of Chapter 4. Similarly, gas and water
production and pressure depletion will affect the IPR and are
discussed later.
GAS WELL
PRODUCTIVITY
EQUATION
As discussed earlier, although Darcy’s law may also be used
to describe the production performance of a natural gas well
under pseudo-steady state conditions, the total skin, S',
includes a significant rate dependent term, Dq. Therefore, the
gas well IPR is really a quadratic equation in terms of q.

(Equation 4.19) m p R – m p wf = Aq g + Bq g2 (psia/cp)

so that:

0.5
–A + A 2 + 4B m p R – m p wf
(Equation 4a.18) qg = (mscf/d)
2B

The variable A is referred to as the Darcy (or laminar) flow

coefficient, while B is denoted as the non-Darcy (or inertial)
flow coefficient. Inspection of the form of Equation 4.19
indicates that a plot of:

m p R – m p wf vs. qg on normal coordinate paper

qg
should produce a straight line with a slope equal to B and a
y-intercept given by A.

In the field, the quadratic form of Equation 4.19 may be used,

especially if pR < 1500 psi or pR and pwf > 3500 psi. In this
case:

(Equation 4.17) p R 2 – p wf 2 = Aq g + Bq g2 (psia2 )

and the IPR will be plotted as follows:

Page 18
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

WELLBORE FLOWING PRESSURE, Pwf, psia

pR

AOF

GAS RATE, q g , mscf/d

Figure 4a.5 – Typical Gas Well IPR

It should be noted that the gas well IPR is always curved,

and␣ the degree of the curvature will be strongly dependent
upon the magnitude of the non-Darcy term and the
production rate.

These deviations from Darcy’s law in gas wells were

(*Reference 4a.10) recognised years ago by Rawlins and Schellhardt*, engineers
from the US Bureau of Mines, who developed the classic
‘back-pressure’ equation relating gas flowrate to flowing
pressure:
n
(Equation 4.16) q g = C p R 2 – p wf 2 (mscf/d)

where: n = delivery exponent between 1.0 for laminar

conditions and 0.5 for high velocity flow
conditions

C = deliverability coefficient, (mscf/d)/psi2

From Equation 4.16, it is clear that the maximum theoretical

flowrate of any gas well can be defined as the absolute open
flow, AOF, given by:

(Equation 4a.19) AOF = C p R 2 n (mscf/d)

These expressions were developed empirically, and there is

therefore no well accepted theoretical basis for calculating n.
Similarly C can only be calculated theoretically when n
approximates to unity.

Page 19
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Comparison of Equations 4.16 and 4.17 also illustrates why

n␣ varies between 1.0 and 0.5. For very low flowrates,
Aqg >> Bq g2, so pR2 – pwf2 ≈ Aqg, and the n value ≈ 1.
Conversely, for very high flowrates Aqg << Bqg2, so
pR2 – pwf2 ≈␣ Bq g2 and the n␣ value ≈ 0.5. Hence, the value of n
is an indicator of the flow regime within the gas reservoir.

The back-pressure equation (4.16) and the Forcheimer

equation (4.17) are now being replaced by the more
theoretically sound pseudo-pressure equation (4.19).

In order to define the values of A and B in Equation 4.19

(or C and n in Equation 4.16), it is necessary to flow a well at
three or more flowrates. Unfortunately, many gas wells, as a
result of their low permeabilities, can take several days or
more for their flowrates to stabilize. Consequently, testing at
a number of rates can take an excessive amount of time.
Several test methods, collectively known as isochronal tests,
are designed to avoid this problem. These all involve flowing
the wells for a fixed period of time at each rate so that the
same drainage radius is investigated. After each flowrate the
well is shut-in for a fixed time interval. Various plots can be
constructed from the results, of which Figure 4a.6 is an
example based on the use of Equation 4.19.

LINE THROUGH
STABILIZED POINT
TO OBTAIN A
m (pR ) - m (p wf )
qg LINE THROUGH POINTS
TAKEN AT END OF EACH
FLOW PERIOD
USED TO OBTAIN B

A SLOPE = B

qg

Figure 4a.6 – Determining Gas Well IPR Parameters

Page 20
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

Conditions at the end of each isochronal flow period are

plotted on this graph. The pseudo-pressure drawdown
(m(pi ) – m(pwf )) is calculated using the shut-in pressure (pi ) at
the start of each flow period. The slope of the line drawn
through the unstabilized points is used to calculate B. To
establish A, the well is produced at a single flowrate until
pseudo-steady state is reached. This gives the stabilized point
in Figure 4a.6. A line through this stabilized point is drawn
parallel to the ‘unstabilized’ line. The intercept of this
stabilized line with the y–axis gives the value of A. This
approach is not exact and is only one example of an
isochronal test method. For further details on gas well testing
(*Reference 4a.11) refer to Slider* or the ERCB* gas well testing manual.
(*Reference 4a.12)
A and B are related to ‘true’ skin and the turbulence
coefficient by the following equations:

1422 T ln 0.472 rr e + S
w
(Equation 4a.20) A = psi 2/ c p / m s c f / d
k gh

(Equation 4a.21) B = 1422 T D

k gh

or

3.16 × 10 –12 β γg T
(Equation 4a.22) B = p s i 2/ c p / m s c f / d2
µwf h p rw
2

where: β = inertial coefficient at the perforated

interval, ft–1
hp = effective perforated interval, ft (this
depends on the number of perforations
effectively open)
µwf = gas viscosity at pwf , cp
γg = reservoir gas gravity (air = 1)

An inertial coefficient correlation (log β vs log k) should be

determined based on laboratory core flow tests, as the results
vary widely between wells and fields. In the absence of
(*Reference 4a.13) specific field data, the correlations published by Katz*
(*Reference 4a.4) (Equation 4a.23) or by Dake* (Equation 4a.24) can be used.

Page 21
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

4.11 × 10 1 0
(Equation 4a.23) β =
k p1.33

2.73 × 10 1 0
(Equation 4a.24) β =
k p1.1045

where: kp = effective permeability in the perforated

zone, md

The difference between these correlations illustrates the high

degree of uncertainty over the β factor even in undamaged
core material. However, the β factor required in these
calculations is that in the immediate wellbore area where
permeability may have been affected by completion
operations.

Note that the skin factor, S, appearing in Equation 4a.20

should not include the rate dependent flow effects (Dq), which
are directly accounted for in the B term.

As discussed previously, the total skin, S', obtained from gas

well tests is:

(Equation 4a.12) S' = S + Dq

Hence, the non-Darcy term, D, can be directly related to B or

can be calculated theoretically by:

B k gh
(Equation 4a.25) D = (mscf/d)–1
1422 T

or

2.22 × 10 –15 β γg k g h
(Equation 4a.26) D = (mscf/d)–1
h p rw µw f
2

Page 22
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

Comparison of the coefficients measured in well tests with

theoretical calculations can help identify the major causes of
poor productivity and the potential improvements in well
performance to be expected from stimulation or reperforation.
Damaged gas wells respond extremely well to stimulation
since the treatments will effect both the Darcy skin, S, and
the non-Darcy flow coefficient, B, which is dependent on the
effective perforated interval and the inertial effects caused by
permeability damage in the near wellbore area.

EFFECTS
OF WATER
PRODUCTION

Water
Production
in Oil Wells
As discussed in the Theory section of Chapter 4, the
production engineer is often more interested in the gross PI
than in the net oil productivity, when water encroachment
can be expected as a result of natural water drive or in
response to a waterflood scheme.

An increasing water cut may indicate gradually increasing

water saturation in the oil zone as oil is displaced by water –
this is known as the fractional flow of water. Fractional flow
is controlled by the relative permeability of oil and water as a
function of increasing water saturation.

The oil and water may also be flowing in a segregated

fashion, whereby some reservoir layers have completely
watered-out while others are still producing dry oil.
Segregated flow also occurs in most layered zones and in
wells that cone water. The water cut in segregated flow is
controlled by the relative thickness of reservoir pay that has
been flooded out.

Page 23
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

WATER CONING HIGH PERMEABILITY STRINGERS

OWC

SEGREGATED FLOW

WATER FLOOD ADVANCE TRANSITION ZONE

FLOOD FRONT

TRANSITION
OWC

FRACTIONAL FLOW

WATERED OUT/WET ZONE LOW WATER CUT ZONE

HIGH WATER CUT ZONE OIL BEARING ZONE

Figure 4a.7 – Segregated and Fractional Flow Regimes

Under these conditions, the gross PI radial flow equation

under pseudo-steady state flow can be written as:

(Equation 4a.27) Jgr = 0.00708 k o h o + 0.00708 k w h w

µo Bo ln .472 rr e + S' o µw Bw ln .472 rr e + S'w
w w

+ 0.00708 k w h o (stb/d/psi)
µw Bw ln .472 rr e + S' o
w

where: ko ho = effective oil permeability thickness in oil

zone, md ft
kw hw = effective water permeability thickness in
water zone, md ft
kw ho = effective water permeability thickness in
oil zone, md ft
S'o = oil zone skin
S'w = water zone skin

The first two terms describe the segregated flow of oil and
water through separate reservoir intervals, while the first and
third term characterize the fractional flow of formation water
due to saturation changes in an oil zone or a transition zone.

Page 24
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

The segregated flow of oil and water dominates in many

coning, cusping and water breakthrough situations. This can
be modelled using a separate IPR curve for each produced
fluid. The curves are summed at various pwf values to develop
a gross IPR curve, as depicted in Figure 4a.8.

1.0

WATER CUT
GROSS
p wf 0.8
0.6
WATER
0.4
OIL 0.2
0
q GROSS q

IPR CURVE WATER CUT PLOT

Figure 4a.8 – Segregated Flow IPR Curve and Water Cut Plot

In this figure, the pressure in the water-bearing interval is

shown as being greater than that of the oil zone. Accordingly,
at very low pressure drawdowns only water will be produced
provided there is no cross-flow, which there often is. As the
BHFP is reduced, the water cut will vary with the gross
production rate (refer to Figure 4a.8). Extreme caution must
therefore be exercised in modelling this situation in well
performance predictions.

Fractional flow describes the process whereby water is

created from a transition zone, which may have been created
by water encroachment into an oil bearing reservoir. If the
layers are homogeneous and isotropic, as assumed in most
reservoir models, this might also occur with water influx from
an active aquifer or artificially through the process of
waterflooding. As the water saturation in the section
gradually increases, the effective permeability to oil, ko,
decreases whereas the effective permeability to water, kw,
increases (refer to Figure 4a.2). Equation 4a.27 describes the
total fluid PI function under fractional flow conditions. Since
most of the terms in Equation 4a.27 are constants, then Jgr is
proportional to:

Page 25
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

ko + kw
µo B o µw B w

The water-oil ratio will be the ratio of these two mobility

terms.

Figure 4a.9 shows how to determine the gross PI, Jgr, under
fractional flow conditions. This figure demonstrates how the
PI initially decreases with increasing water cuts, but at
higher water saturations the gross PI may start to rise.

µ o = 0.6 cp B w = 1.3 rb/stb

k ro k rw
PI α + µ w = 0.3 cp B w = 1.0 rb/stb
µo Bo µ wB w
S wirr = 0.30 S or = 0.30

Sw k ro k rw k ro / µ o Bo k ro / µ w B w PI Sw / PI Sw = o Water
Cut
0.30 1.00 0.000 1.282 — 1.000 0.000
0.40 0.92 0.010 1.179 0.033 0.945 0.027
0.45 0.75 0.025 0.962 0.083 0.815 0.079
0.50 0.40 0.050 0.641 0.167 0.630 0.207
0.55 0.15 0.070 0.192 0.233 0.332 0.548
0.60 0.06 0.100 0.077 0.333 0.315 0.812
0.65 0.02 0.135 0.026 0.450 0.371 0.945
0.70 0.00 0.180 — 0.600 0.468 1.000

1.0 1.0
PI Sw / PI ( S w = 0)

0.8 0.8
WATER CUT

0.6 0.6
qg

0.4 0.4

0.2 0.2

0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
WATER CUT SW

Figure 4a.9 – Reservoir Performance under

Fractional Flow Conditions

In summary, for the more common segregated flow, simply

summing the oil and water PI will describe the gross fluid
production. In fractional flow, the gross PI is proportional to
kB total.
µ

Page 26
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

(d)
(a) DURING LATER PRODUCTION
BEFORE PRODUCTION WATER PERCOLATES PAST SHALES AGAIN AND
AND DURING VERY EARLY PRODUCTION FINGERS ALONG HIGH PERMEABILITY ZONES —
VERTICAL EQUILIBRIUM — IDEAL SITUATION VERTICAL EQUILIBRIUM LOST

PRODUCTION TOP RESERVOIR PRODUCTION TOP RESERVOIR

K K
J J

H H

F F

E E
ORIGINAL OWC 2217 m ORIGINAL OWC 2217 m

(b) (e)
DURING LATER PRODUCTION DURING LATER PRODUCTION
WATER PERCOLATES PAST SHALES BRIDGE PLUGS SET AND VERTICAL
AND VERTICAL EQUILIBRIUM LOST EQUILIBRIUM RE-ESTABLISHED

PRODUCTION TOP RESERVOIR PRODUCTION TOP RESERVOIR

K K

J J

H H

F F

ORIGINAL OWC 2217 m E E

ORIGINAL OWC 2217 m

(f)
(c) CURRENT POSITION
DURING LATER PRODUCTION WATER PERCOLATES PAST SHALES AND FINGERS
VERTICAL EQUILIBRIUM ALONG HIGH PERMEABILITY ZONES AGAIN.
RE-ESTABLISHED BY BRIDGE PLUGS FINAL LOSS OF VERTICAL EQUILIBRIUM

PRODUCTION TOP RESERVOIR PRODUCTION TOP RESERVOIR

K K

J J

H H

F F

ORIGINAL OWC 2217 m E ORIGINAL OWC 2217 m E

Figure 4a.10 – Conceptual Water Movement on the Forties Field

Page 27
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Modelling water production using reservoir simulators is

often a complex exercise, particularly with multiple layers.
Knowing how the well behaves and exactly modelling this
behaviour is difficult, and hence predictions from simulators
can be prone to errors. When designing wells for this
situation, close cooperation and discussion with the reservoir
engineer is required to ensure that the likely variation in the
prediction of water influx is addressed.

Figure 4a.10 illustrates the concept of water movement in a

typical north sea reservoir.
Water
Production
in Gas Wells
Natural gas under reservoir conditions is always saturated
with water vapour, which condenses during production.
The␣ water content of gas is dependent on reservoir
conditions. It increases with temperature and pressure. The
volume of condensed water is relatively constant for any
given wellhead conditions, but increases with compression
and/or declining temperature and is not usually critical until
velocities become low.

Production of formation water is much more significant to

productivity calculations and may vary with rate. It may
result from either fractional or segregated flow, or from poor
zonal isolation. Fortunately, formation water usually can be
readily identified on the basis of its dissolved solids content.

Fractional flow of connate water will occur when the water

saturation becomes sufficiently high to create a mobile water
phase with a non-zero effective water permeability. As long as
the water saturation does not change significantly, the kw/kg
ratio will also remain constant, as will the observed
producing water-gas ratio (WGR). This fractional flow of water
is analogous to that described previously for oil zones.

The segregated flow of gas and water can be described using

separate IPR curves for both the gas and water phases in
much the same manner as discussed above for the
segregated flow of oil and water. However, in this case it
makes little sense for the two curves to be combined. The
dominant fluid (wet gas or gassy water) is used to develop the
TPC curve intersection, and the corresponding bottom hole
pressure is used to determine the production rate of the
secondary fluid, and hence the formation WGR to which the

Page 28
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

condensed WGR must be added. Note that the TPC must be

based on the actual WGR, and therefore this process is
iterative.

TPC FOR WGR = 6.0 bbl / mmscf

pR

p wf
GAS

WATER

0
10 20 30 40 50
q g (mmscf/d)
0
100 200 300 400 500
q w (bbl/d)

85
FLOWING WGR = = 4.7 bbl/mmscf
18
CONDENSED WGR = 1.3 bbl/mmscf

TOTAL WGR = 6.0 bbl/mmscf

Figure 4a.11 – Wet Gas Well IPR Curve

EFFECT OF
PRESSURE
DEPLETION

Oil Wells
To develop meaningful production forecasts and appropriate
artificial lift designs, estimates of expected IPR curves
throughout the reservoir life are frequently required. These
IPR predictions must account for declining reservoir pressure
as well as the effects of changing fluid saturations.

Where a reservoir model has been developed, simulation

results can be used to predict the future IPRs as discussed in
the Theory section of Chapter 4.

However, on smaller fields the production engineers may

have to develop their own estimates, for which a number of
methods have been derived.

Page 29
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

• Standing’s Method
To make this generally applicable, and to avoid confusion
(*Reference 4a.14) with measured PI, J', Standing (1970)* defined a term J*,
as the PI at low drawdowns derived from the pseudo-
steady state radial flow equation, Equation 4a.3. This
means the theoretical PI with no two-phase flow effects,
which is the case at low drawdowns.

Under solution gas drive conditions, the only reservoir

parameters that are likely to change with reservoir
pressure are the solution GOR and gas saturation, Sg.
Therefore, the ratio of future J* to present J* is obtained
from Equation 4a.3 as:

(Equation 4.13) J *f u t u r e = k r o / µoBo f u t u r e

J *present k r o / µoBo present

where: kro = relative oil permeability

• Eickmeier/Fetkovich
(*Reference 4a.15) Fetkovich* further simplified this by suggesting that the
change in relative permeability was directly proportional
to the change in the static reservoir pressures. Since the
viscosity and Bo effects are often relatively small and tend
to cancel out, then:

(Equation 4a.28)
J *f u t u r e = p R future
J *initial p R initial

(*Reference 4a.16)
Eickmeier* reportedly modified this concept and
suggested that a more realistic expression was:

(Equation 4a.29) J *i n i t i a l = p R initial m

J *future pR future

It can be shown theoretically that the exponent m can

vary between 1 and 3. It would be 3 where Fetkovich’s IPR
(*Reference 4a.17) model applies, while Vogel’s work* suggested an m value
closer to 2.5. Eickmeier’s empirical example from the
House Mountain Field in Canada, however, shows a less
rapid deterioration in the IPR with m being 1.66.

Page 30
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

• Empirical Methods
Empirical correlations have been used successfully to
predict future IPR curves in solution gas drive reservoirs.
The most common one is now often attributed to Kermit
(*Reference 4a.18) Brown*.

A plot of log PI vs. cumulative oil recovery will often result

in a straight line. For best results, the PIs should be
determined at the same limited drawdown. The method is
strictly empirical, but is based on numerous field
observations.

(*Reference 4a.19) Uhri and Blount* have also proposed a graphical, pivot
point method by which two well tests at different reservoir
pressures can be used to develop a forecast of the ongoing
IPR decline trend.

In fields where reservoir pressure is maintained by

waterflooding and/or active water drive, the gross PI often
shows little or no decline until workover operations either
result in formation damage or reduce the contributing
interval, as a result of squeeze operations. This is not true
in heavy oil fields where large differences in oil and water
viscosity can cause large changes in the gross PI.

However, with the increasing availability of reservoir

simulation results, these methods of guesstimating future
IPRs are less frequently needed.
Gas Wells
As for an oil well, the shape of gas well IPRs changes over the
life of the reservoir. Strictly speaking, both coefficients in the
quadratic form of the gas deliverability equation (A and B) will
vary with the pressure decline since they are functions of
pressure-dependent gas properties, such as viscosity µ and
gas deviation factor Z.

(Equation 4a.30) A = A
µgZ future µgZ initial

(Equation 4a.31)
B = B
Zf u t u r e Zi n i t i a l

Page 31
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

0.07

0.06
µg Z
0.05
µ g or µ g Z, cp

0.04 1.6
1.5
0.03 1.4

Z FACTOR
µg 1.3
0.02 Z 1.2
1.1
0.01 1.0
0.9
0 0.8
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10 000

PRESSURE, psi

Figure 4a.12 – Example of Relationship between µg, Z and p

It is simpler to use the pseudo-pressure form of the IPR

(ie␣ based on Equations 4.19, 4a.20 and 4a.25), because if the
turbulence coefficient, D, is constant, there will be no change
in the values of A and B as the pressure declines.

In highly overpressured zones, expansion of the connate and

shale bound water may lead to further deterioration of the
gas relative permeability, while rock compaction may lead to
decreasing rock or fracture permeability. Condensate dropout
as the BHFP drops below the dewpoint will also cause a
deterioration in the IPR. Therefore, A and B should be
regularly re-evaluated, by testing, through the life of the
reservoir.

EFFECT OF GAS
PRODUCTION
IN OIL WELLS

Vogel’s Two-
phase Flow IPR
Once the BHFP of a producing oil well drops below the
bubble point pressure, pb, the liberated gas causes a
build-up␣ of gas saturation in the vicinity of the wellbore.
Moreover, the volume occupied by this gas will be a function
of the flowing pressure. The resulting reduction of oil relative

Page 32
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

permeability causes lower well productivity and,

consequently, a curved IPR.

The effects of IPR curvature are most significant at high

drawdowns (eg for oil wells which are pumped off). The
estimation of PI curvature is therefore critical for the design
of artificial lift equipment, especially in depletion and gas cap
drive reservoirs. The most commonly used method to predict
IPRs for solution gas drive reservoirs is that proposed by
(*Reference 4a.17) Vogel*, based on the results of a large number of simulation
runs. He found that the maximum rate, qmax, under these
conditions was likely to be only 55% of that predicted by the
straight line PI.

J *p R
(Equation 4a.32) q max = (stb/d)
1.8

1.0 0% DRAWDOWN = BUBBLE POINT PRESSURE

0.8

0.6
p wf / p R

LINEAR PI
0.4
VOGEL IPR
100%
0.2 DRAWDOWN

0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

q / q max 9 q max
q max 5

Figure 4a.13 – Relationship between Vogel’s Reference Curve

and a Linear PI

Vogel’s simulation studies showed that the curvature of the

IPR as a result of two-phase flow could be described by a
general reference curve:

q pwf pwf 2
(Equation 4.3)
q max = 1 – 0.2 pR – 0.8 pR

Page 33
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

VOGEL IPR EQUATION 2

qo p wf p wf
= 1 — 0.2 — 0.8
qo max pR pR

1.00
p wf
pR

0.80
=
BOTTOM HOLE FLOWING PRESSURE
AVERAGE RESERVOIR PRESSURE

0.60

0.40

0.20

0
0 0.20 0.40 0.60 0.80 1.00
PRODUCING RATE q o
=
MAXIMUM PRODUCING RATE q o max

Figure 4a.14 –␣ Vogel’s Reference Curve for Solution Gas Drive Reservoirs

Page 34
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

This relationship does not apply to wells producing above the

bubble point pressure or to wells with significant skin
(S > +3).

When a well is tested at high drawdown, curvature of the IPR

will be seen as an apparent skin S'. Two-phase skin effects
are dealt with more fully in Section 4b – Near Wellbore
Performance.

Combining the effects of two-phase flow with the problem of

(*Reference 4a.16) predicting future IPRs, Eickmeier* extended Vogel’s work to
develop a family of generalized IPR curves based on his
empirical field data.

1.0
p wf / p R , BOTTOM HOLE WELL PRESSURE, AS
A FRACTION OF RESERVOIR PRESSURE

0.8

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1.0

q/ q m , PRODUCING RATE AS A
FRACTION OF MAXIMUM RATE

Figure 4a.15 – Eickmeier's Generalized IPR Curves

(Equations 4.4 As discussed in the Theory section, Neely developed an

and 4.5) extension to Vogel’s equation, which has been published by
(*Reference 4a.18) Brown*, for cases where the IPR passes through the bubble
point (see Chapter 4, Theory, page 11).

An alternative and common approach to this problem of

predicting the IPR of a well that is initially above the bubble
point pb, and where there is no well test data below pb, is to:

1) Use PVT data or correlations to estimate the bubble

point pb.

Page 35
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

2) Use the straight line PI (measured or theoretical) to give

the production rate at pb (qb) and to approximate the
rate (qA) just below the bubble point (say at 90% pb).

3) Estimate the incremental rate, qi, obtained below the

bubble point:

qi = qA – qb

4) Enter the Vogel curve in the usual manner using pb as the

reservoir pressure and qi as the rate, and estimate qmax.
Note that in this case, qmax is the maximum incremental
rate, and the total rate is:

qtmax = qb + qmax

5) The incremental and total rates of any other bottom hole

pressure can be calculated in a similar fashion.

Vogel’s reference curve and equation only apply to wells with

a skin factor of zero, or, stated another way, with a flow
efficiency of 1.0. To account for damage, or predict the
(*Reference 4a.14) improvement expected from stimulation, Standing (1970)*
published a modification to Vogel’s generalised curve as
shown in Figure 4a.16.

Flow efficiency can be defined from Figure 4a.3 as:

ln rr e – 3
p – p' w f w 4
(Equation 4a.33) FE = R = ≈ 7
pR – pwf 7+S
ln rr e – 3 + S
w 4

where: p'wf = equivalent undamaged flowing pressure

pwf = actual (measured) flowing pressure

Although Standing originally published his curves for FE

from 0.5 to 1.5, the high FE curves are generally viewed as
(*Reference 4a.18) incorrect and a revision has been published by Brown*,
apparently based on unpublished work by Harrison, as
shown in Figure 4a.16.

Page 36
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

1. CALCULATE FLOW EFFICIENCY, FE

p wf
2. ENTER PLOT AT
pR
3. READ OFF FRACTION OF MAXIMUM VOGEL FLOWRATE

1.0
FRACTION OF RESERVOIR PRESSURE p wf / p R

FRACTION OF RESERVOIR PRESSURE p wf / pR

1.0 0.9

0.8 FLOW
BOTTOM FLOWING PRESSURE,

BOTTOM FLOWING PRESSURE,

EFFICIENCY
0.8 0.7

FLOW 0.6
0.6 EFFICIENCY
0.5

1.1

1.5
1.3

1.7

2.1

2.3

2.5
1.9
0.

0.4
9
0.

1.2
1.0

1.4

1.6

2.0

2.2
1.8

2.4
0.4
0.5

1.0
0.

0.3
0.6

0.2 0.2

0.1

0 0
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

FRACTION OF MAXIMUM VOGEL FLOWRATE FRACTION OF MAXIMUM VOGEL FLOWRATE

STANDINGS CORRELATION FOR WELLS WITH HARRISONS EXTENSION OF STANDING's WORK
FE VALUES NOT EQUAL TO 1 TO INCLUDE OTHER FE VALUES

Figure 4a.16 – Corrections to the Vogel Curve for Non–Zero Skins

Generalized
IPR Curves
Because of the problems of using the Vogel/Standing model
in practical situations where the flow efficiency is uncertain,
it may be better to use the generalized IPR curves presented
in the Theory section of Chapter 4 for oil wells producing
below the bubble point, if multirate test data is available.

(Equation 4.9) q = C p R 2 – p wf 2 n (stb/d)

or

(Equation 4.11) p R – p wf = aq + bq 2 (psia)

Page 37
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

Wells Producing
Free Gas
If the producing gas-oil ratio (GOR) is greater than that
expected on the basis of the solution GOR at that specific
reservoir pressure, then free gas is being produced from
the␣ reservoir.

Once the reservoir pressure drops below the bubble point,

high GOR can be expected because of the diversion of gas
liberated in the immediate wellbore area before it can reach
the secondary gas cap. Under these conditions, the GOR is
often a function of drawdown.

In some instances, a high producing GOR may also be

indicative of the production of free gas from an associated gas
cap. Such a situation is more likely to occur under reservoir
conditions favouring gas coning or cusping. Coning is a
vertical distortion of the GOC near the wellbore, which
develops when the drawdown exceeds the existing gravity
forces. Coning occurs most frequently under conditions
where vertical permeability is enhanced due to vertical
fracturing (either natural or hydraulic) or uniform isotropic
sands without shale barriers. Poor cement bonds lead to a

p wf

TPC FOR GOR = 3300 scf/stb

FREE GAS PRODUCTION = 9 000 000 scf/d

p wf
OIL PRODUCTION = 3000 stb/d

FREE GAS EQUIVALENT GOR = 3000 scf/stb

SOLUTION GOR = 300 scf/stb

OIL
PRODUCING GOR = 3300 scf/stb
0
1000 2000 3000 4000 5000 6000 7000
q (stb/d)
o
0
2 4 6 8 10 12 14
FREE GAS RATE, q (mmscf/d)
fg

Figure 4a.17 – IPR of an Oil Well Producing Free Gas

Page 38
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

similar effect due to the high vertical transmissibility close to

the wellbore. Cusping or tonguing is a lateral distortion of the
GOC down-dip in high permeability streaks.

If it can be ascertained that a well is coproducing from both

the oil zone and gas cap, separate IPR curves should be
generated for the liquid and gas production, so that the pwf for
the oil phase can be used to determine the producing GLR to
be used in the outflow simulation. The TPC should be based
on the actual GOR and, hence, this becomes an iterative
process.
MULTIZONE IPRs
Strange looking IPRs can be expected where a well has been
completed with downhole commingling or in a stratified
reservoir composed of a series of layers of markedly different
permeabilities.

p R1

p R2

p R3

p wf

COMPOSITE IPR

LOW k MEDIUM k HIGH k

ZONE ZONE ZONE

Figure 4a.18 – Multizone IPRs

If there is no vertical communication between the zones,

except through the wellbore, production will be drawn
primarily from the highest permeability zone with the result
that the static pressure in this interval will decline more
rapidly than the pressures in other zones. Conversely, the
lowest permeability section will maintain the highest reservoir
pressure due to its limited production. This phenomenon is

Page 39
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

referred to as differential depletion. In this situation, the

relative contribution from each zone will be a function of
drawdown and of the IPRs of the individual layers.

When the well is shut-in, cross flow will occur between the
higher pressured, lower permeability layers and the lower
pressured, higher permeability sections, in response to the
pressure differentials.

Production allocation and reservoir performance prediction

and management in this type of completion is therefore
extremely difficult.

Page 40
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4a – RESERVOIR PERFORMANCE

REFERENCES

4a.1 Golan, M and Whitson, C H: Well Performance,

International Human Resources Development
Corporation, Boston, NY (1986)

4a.2 Earlougher, R C Jr: Advances in Well Test Analysis,

Monograph Series, SPE, Dallas (1977)

4a.3 Al-Hussainy, R, Ramey, H J Jr and Crawford, P B:

‘The Flow of Real Gases Through Porous Media’,
J␣ Pet␣ Tech (May 1986) 624; Trans, AIME (1966) 237,
637–642

4a.4 Dake, L P: Fundamentals of Reservoir Engineering,

Elsevier Scientific Publishing Company, Oxford (1978)

4a.5 Standing, M B: Volumetric and Phase Behaviour of

Oilfield Hydrocarbon Fluids, Reinhold Publishing Corp,
New York (1952)

4a.6 Glaso, O: ‘Generalized Pressure-Volume-Temperature

Correlations’, J Pet Tech (May 1980), 785–795

4a.7 Beal, Carlton: ‘The Viscosity of Air, Water, Natural

Gas, Crude Oil and Its Associated Gases at Oil-Field
Temperatures and Pressures’, Trans, AIME (1954)
165, 94–115

4a.8 Matthews, C S and Russell, D G: Pressure Buildup

and Flow Tests in Wells, Monograph Series, Society of
Petroleum Engineers of AIME, Dallas (1967) 1,
Appendix G

4a.9 Carr, N L, Kobayashi, R, and Burrows, D␣ B: ‘Viscosity

of Hydrocarbon Gases Under Pressure’, Trans, AIME
(1954) 201, 264–272

4a.10 Rawlins, E L and Schellhardt, M A: ‘Back-Pressure

Data on Natural Gas Wells and Their Application to
Production Practices’, US Bureau of Mines
Monograph 7 (1936)

Page 41
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4a – RESERVOIR PERFORMANCE Completion Design Manual

4a.11 Slider, H C: Worldwide Practical Petroleum Reservoir

Engineering Methods, Penwell Publishing Company,
Tulsa, OK (1983)

4a.12 Energy Resources Conservation Board, Calgary,

Alberta, Canada: Theory and Practice of Testing of
Gas␣ Wells , Third Edition, 1975

4a.13 Katz, D L et al: Handbook of Natural Gas Engineering,

McGraw-Hill Book Company, Inc (1959), 47–50

4a.14 Standing, M B: ‘Concerning the Calculation of Inflow

Performance of Wells Producing from Solution Gas
Drive Reservoirs’, J Pet Tech (Sept 1971), 1141–1142

4a.15 Fetkovich, M J: ‘The Isochronal Testing of Oilwells’,

SPE 4529, SPE of AIME, 1973

4a.16 Eickmeier, J R: ‘How to Accurately Predict Future Well

Productivities’, World Oil (May 1968), 99–106

4a.17 Vogel, J V: ‘Inflow Performance Relationships for

Solution Gas Drive Wells’, J Pet Tech (Jan 1968),
83–93

4a.18 Brown, K E, The Technology of Artificial Lift Methods,

Vols 1 and 4, Pennwell Publishing Company, Tulsa,
OK (1977)

4a.19 Uhri, D C and Blount, E M: ‘Pivot Point Method

Quickly Predicts Well Performance’, World Oil
(May␣ 1982), 153–164

Page 42
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

CONTENTS

INTRODUCTION 1

COMPONENTS OF TOTAL SKIN 2

COMPONENTS OF DARCY SKIN 5

FORMATION DAMAGE SKIN 5
PERFORATION SKIN 7
PARTIAL COMPLETION SKIN 11
DEVIATION SKIN 14
STIMULATION SKIN 16
SKIN DUE TO RESERVOIR HETEROGENEITY 18
SKIN DUE TO MULTIPHASE FLOW 19

NON-DARCY SKIN 22
ONSET OF NON-DARCY FLOW 23
NON-DARCY SKIN IN GAS WELLS 23
NON-DARCY SKIN IN OIL WELLS 25
PRACTICAL CONSIDERATIONS WITH RESPECT TO NON-DARCY SKIN 25

EFFECTS OF GRAVEL PACKS 27

SKIN EFFECTS FROM AN INTERNAL GRAVEL PACK IN OIL WELLS 27
SKIN EFFECTS FROM AN INTERNAL GRAVEL PACK IN GAS WELLS 30
SKIN EFFECTS FROM EXTERNAL GRAVEL PACKS 31

IMPLICATIONS FOR COMPLETION DESIGN AND 33

WORKOVER PLANNING

REFERENCES 34
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

INTRODUCTION

As discussed in the Theory section of Chapter 4 and in

Section 4a – Reservoir Performance, the near wellbore
performance, caused by the well␣ completion operations,
modifies the ideal reservoir performance to create the well’s
actual IPR. This is represented by the total skin term, S', in
the theoretical inflow performance equation and in the results
of a well test analysis.

In most wells, the total skin, S', is the only IPR parameter
that can be changed by drilling, completion, stimulation,
workover and production practices, and is therefore a primary
concern of production engineers.

Initial well deliverability predictions and tubing selection

are␣ usually based on the assumption that good completion
practices will achieve a near optimum IPR with minimal, or
negative, skin. However, not all completions can meet this
objective, and the production engineer needs a methodology
for identifying the probable causes of a non-ideal inflow
performance. Recommendations can then be developed on
the␣ best remedial treatment for an existing well, and on
improved completion practices to avoid similar problems in
future wells.

The techniques used to minimize skin damage, or to attain a

negative skin, are discussed in Section 5a – Optimizing the
Completion in the Near Wellbore Region, while this section
focuses on estimating the magnitude of the various
components of skin.

In specific cases, reservoir management considerations,

production targets or technical problems dictate that a non-
ideal completion configuration should be used. The
production engineer here needs procedures to estimate the
detrimental consequences on the inflow performance.
Similarly, in some fields, limited production targets or
allowables, coupled with marginal economics for high cost
completion or stimulation operations make the attainment of
a near optimum IPR difficult to justify. In these cases, the
production engineer can utilize the techniques presented in
this section to estimate the skin to be expected with various
completion options, in order to develop alternative forecasts of
well deliverability and lifting cost in support of the
recommendations for the most cost-effective completion.

Page 1
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

COMPONENTS OF TOTAL SKIN

There are a number of factors which combine to cause the

skin effect seen in many poorly completed or damaged wells.

Lp DRILLING DAMAGE
CEMENT (Part of Formation Damage)

CASING

CONVERGING

FLOW FROM
h
h p

PARTIAL

COMPLETION

rw NOTE: EXAGGERATED SCALE

rd
DEEP FORMATION DAMAGE
rs

rw = WELLBORE RADIUS
rd = RADIUS OF DRILLING DAMAGE
rs = RADIUS OF FORMATION DAMAGE
Lp = EFFECTIVE LENGTH OF PERFORATION

Figure 4b.1 – Positive Skin Effects

The total skin, S', measured in a well test may include both
low rate, Darcy skin effects, S, and high rate, non-Darcy flow
effects, Dq , as discussed in Section 4a – Reservoir
Performance.

The total skin is given by:

(Equation 4a.12) S' = S + Dq

Page 2
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

Each of these components may be the result of a combination

of factors:
• Formation damage, Sd.
• Perforation skin, Sp.
• Partial completion skin, Sc.
• Deviation skin, S"c.
• Stimulation effects, Ss.
• Reservoir heterogeneity effects, Sh.
• Multiphase flow effects, Sm.
• Darcy effects of sand or gravel filled perforations, SG.
• Non-Darcy effects associated with native formation
permeability and any of the above, Dq.
The effects are generally additive, but cannot be summed
directly as the causes of one type of skin directly affect the
causes of other types of skin. Skins due to near wellbore
effects (such as perforation skin) can, however, be added to
more distant effects (such as partial penetration), eg:

(Equation 4b.1) S = h Sp + Sc
hp

where: h = net pay thickness (isopach), ft

hp = perforated interval length, ft
(*Reference 4b.1) The h/hp multiplier was proposed by Jones and Watts* to
reflect the increase in velocity that could be expected across
the damaged zone of a partially completed well.
Non-Darcy skins are also cumulative. The presence of, for
example, drilling damage, a low virgin permeability and low
perforation density will combine to create a non-Darcy skin
greater than either would produce on its own.
The effect of multiphase flow in the near wellbore area due to
producing an oil well below the bubble point (ie, the Vogel
effect), or a gas condensate well below the dewpoint (ie,
condensate blockage effect), may be included as either a
Darcy component, Sm, or a non-Darcy component, Dmq,
although neither accounts for the real flowrate dependence
of␣ multiphase flow skin.
These skin effects create restrictions to flow through the near
wellbore area and cause a decrease from the ideal IPR to the
actual IPR.
A fuller description of how to evaluate the combined effects of
the various components of skin is presented in Chapter 2 of
the Near Wellbore Manual.

Page 3
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

TYPICAL OIL WELL IPR

RADIAL FLOW EQUATION TURBULENCE

I (NON-DARCY) EFFECTS (Dq)
p R - p wf = J q
pR TWO PHASE
FLOW EFFECTS (S m )

pb DARCY
IDEAL IPR
SKIN EFFECT(S)

ACTUAL IPR 2
p R - p wf = aq + bq
p wf

q
PRODUCTION LOST
DUE TO EFFECTIVE SKIN (S')

TYPICAL GAS WELL IPR

IDEAL IPR
RADIAL FLOW EQUATION
PR m(p r ) — m(p wf ) = Aq

P wf TURBULENCE
EFFECTS (S TB )
DARCY
ACTUAL IPR SKIN EFFECT(S)
2
m(p R ) - m(p wf ) = A q+ B q

Q
PRODUCTION LOST
DUE TO EFFECTIVE SKIN (S')

Figure 4b.2 – Effect of Skin on Well IPRs

Page 4
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

COMPONENTS OF DARCY SKIN

In this subsection, we shall consider the theoretical

magnitude of the various components of Darcy skin. Each
skin factor is computed independently as though the subject
well was otherwise a non-damaged, fully penetrating open
hole in a uniform isotropic reservoir producing a single-phase
incompressible fluid under a low pressure differential in
accordance with Darcy’s law. Although the reliability of these
theoretical calculations is somewhat nebulous since many of
the input parameters are uncertain, they are useful in a
qualitative sense to identify the most probable causes of the
observed skin, or to compare the relative magnitude of the
skin to be expected from various completion procedures.
Moreover, they can be valuable in determining if a remedial
operation to improve the IPR is likely to be successful, and in
selecting the most appropriate course of action. For example,
a reperforation or a small volume clean-up treatment with
low strength acid and ball sealers may be more effective than
a full scale matrix acidization with mud acid if the skin is
caused by plugged or inadequate perforations (refer to
Section 5a – Optimizing the Completion in the Near Wellbore
Region), but not if it is due to deep damage; whereas none of
these treatments will be effective if the skin is caused by
multiphase flow or partial completion effects.
FORMATION
DAMAGE SKIN
Formation damage skin, Sd, occurs when the permeability in
the near wellbore area is reduced as a result of various fluid-
fluid and fluid-rock interactions. This is often due to the
invasion of dirty or incompatible fluids during drilling,
completion, workover or well kill operations. The damage can
result from both physical and chemical effects, as discussed
in Sections 5a – Optimizing the Completion in the Near
Wellbore Region and 5b – Downhole Production Chemistry.

Page 5
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

PARTICLES IN WELLBORE
FLUIDS BLOCK PORE THROATS
CEMENT

ku

CASING
ks SOLIDS IN
WELLBORE
FLUIDS
SHALE
SWELLING
SAND GRAINS

SAND GRAINS
ADJACENT TO
WELLBORE SHALE SWELLING
SOLIDS IN
WELLBORE TRAPPED
FLUIDS HIGH WATER
CAN BLOCK
PERFORATIONS SATURATION

PRODUCED FINES
TRAPPED
HIGH WATER
SATURATION
rw
rs

PRODUCED
FINES

Figure 4b.3 – Schematic of Formation Damage Effects, Sd

The formation damage skin, Sd, can be estimated by Hawkins’

(*Reference 4b.2) equation*:

k u – ks
(Equation 4b.2) Sd = ln rr s
ks w

where: Sd = formation damage skin

ku = undamaged formation permeability, md
ks = skin zone permeability, md
rs = radius of damaged zone, ft

ks can be obtained from coreflood tests. rs may, in some

cases, be estimated from resistivity logs where it is assumed
that the depth of invasion equals the depth of damage. It is
also possible to make an estimate of rs from special ‘fluid loss’
experiments. Limits on possible values of rs may be obtained
from the mud used to drill the relevant section of the well.

This equation is used for converting the results of core

damage and acidization tests for application in field
situations. It is also useful to show how very high skins can
be generated as ks approaches zero, irrespective of the depth
of the damage. The equation should be used as an indication,
as no account is taken of perforations, heterogeneity etc.

Page 6
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

By combining the Hawkins’ equation with the filter theory,

presented in Section 5a – Optmizing the Completion in the
Near Wellbore Region, it is possible to determine the
minimum acceptable quality for clear completion and
workover fluids that may be lost into the perforations in
significant volumes (eg, during gravel packing or workovers
using clear brine).

There is a great deal of uncertainty involved with trying to

separate the perforation skin, Sp, from the formation damage,
Sd. Nevertheless, an attempt should be made to determine
whether the damage was probably caused during the drilling
or completion operations, and to select the most appropriate
remedial treatment. True formation damage may be more
difficult to treat or to bypass than plugged or damaged
perforations, and involves higher costs and risks (refer to
Section 5a).

In the past, formation damage was attributed primarily to

drilling and cementing operations. However, as discussed in
Section 5a, the use of good mud engineering practices should
ensure that drilling damage is localized and easily bypassed
using high powered perforating guns.

However, it should be noted that practices that are good for

avoiding formation damage are not necessarily followed when
drilling a well, either because of cost considerations, the use
of additives to avoid unrelated problems or because the mud
has not been tested for compatibility with the formation or
downhole fluid loss characteristics.

Where the perforations extend beyond the damaged zone, the

effect of the formation damage on well productivity will be
greatly reduced. The effects of this type of shallow damage
(eg, from drilling with good muds) may be incorporated within
the perforating skin component, Sp, by adjusting the effective
length of the perforations.

PERFORATION
SKIN
Perforation skin, Sp, is caused by additional pressure drops
as the radial flow of reservoir fluids deviates to spherical/
cylindrical flow into the perforations and crosses the crushed
zone around the perforation tunnel to enter the wellbore. The
perforation skin factor, Sp, will increase significantly where the
shot density is low, or effectively low because of perforations

Page 7
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

being plugged. Perforation geometry and perforation quality

will also affect this skin component.

Until recently, it was common to combine perforation damage

together with formation damage. However, trying to separate
the two components allows a better understanding of the
importance of good perforating practices and explains why
simple remedial treatments (eg, reperforation or clean-out
acid jobs with low strength acid and ball sealers) often yield
substantial productivity improvements.

The effects of the perforation parameters on well deliverability

and skin have been extensively discussed in the literature.
Because of the complexity of modelling a complex three-
dimensional flow system, electrolytic analogs or finite element
mathematical models are generally used, although Karakas
(*Reference 4b.3) and Tariq* have recently published a set of semi-analytical
equations that are included in the BP Near Wellbore Manual.
Several perforating companies offer simulation services either
on a commercial or promotional basis, or have presented
nomograms based on the results of their model studies.
However, all of these methods can only be used in a
qualitative fashion and with caution, as many of the input
parameters are uncertain.

The most extensively used model is the nomogram published

(*Reference 4b.4) by Locke* in 1981. Tariq* subsequently confirmed Locke’s
(*Reference 4b.5) productivity predictions, but pointed out that the results
were slightly optimistic (5 to 10%) due to an insufficiently fine
grid size, and that in high rate wells, further impairment
could be expected due to inertial effects.

This nomogram is useful for illustrating the importance of

penetrating beyond the drilling damage and having at least
four shots per foot open for production.

To use this nomogram semi-quantitatively, one must

estimate the downhole length of the perforation tunnel, the
thickness and permeability of penetrated damage, kd, the
crushed zone permeability ratio and the effective shot
density. Most of these factors are unknown and only partially
controllable, and many engineers feel that this negates the
value of the process. However, the uncertainties can be
reduced by using sensitivity analysis and the results of
previous performance matching studies. In the absence of
any better data, the following assumptions, based on the

Page 8
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

NOMOGRAPH FOR PRODUCTIVITY RATIO

0.5, 0.25 PERF. DIAM. (in.)

18 0.4 11.8
PERFORATION LENGTH (h)

0.33
15 0.5 7.9

5.3
12 0.6

12" BOREHOLE 0.7

9

SKIN FACTOR
3.4
o
0 PHASING
o

PRODUCTIVITY RATIO
6 90 PHASING 2.0 DAMAGED ZONE
DAMAGED ZONE THICKNESS (in.)

0.8
DIAMETER
3 0 0.9 0.9
d
b' - c'
0 3 o
1.0 0.0 OPEN HOLE
120
DIAMETER
0 o 180 o 90 o
6 1.1 -0.7
Ø = PHASE ANGLE
b-c 6" BOREHOLE CRUSHED
9 ANGULAR
1.2 -1.3
ZONE DIAMETER
0.4 0.1 (k d/ ku ) PERFORATION
PHASING
SPACING
o
8 spf, 0 PHASING (DEPENDENT ON
(k c / k u )
SHOT DENSITY)
CRUSHED ZONE PERFORATION
(0.1)
RELATIVE DIAMETER
PERMEABILITY
(k c / ku ) (0.2)
PERFORATION
LENGTH
(0.4)

(1.0)

4
2 1

PERFORATIONS PER FOOT

Figure 4b.4 – Locke’s Nomogram for Perforation Skin

analysis of the skin factors measured in a number of

exploration well tests, can be used:

Drilling damage zone thickness

= 3in (onshore)
= 12in (in North Sea)
Drilling damage zone permeability ratio (kd/ku)
= 0.8 (average based on Sunbury coreflood tests)
Crushed zone permeability ratio (kc/ku)
= 1.0
Effective shot density
= Gun shot density (if perforated underbalance and
produced immediately)
= 0.1 to 0.25 gun shot density (if perforated overbalance
in dirty fluids or if perforations are subsequently
plugged)

Page 9
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

Downhole penetration depth

(*Reference 4b.6) = API RP 43* Section II total target penetration, from
which the actual casing, cement thickness and gun
clearance have been subtracted. If the actual
underbalance is significantly less than that suggested
(*Reference 4b.7) by King et al*, the sensitivity of reducing the
perforation length by half should be considered.

For major developments, reservoir cores should be shot at

reservoir conditions in the laboratory to establish realistic
penetrations. Drilling damage should be estimated from
coreflood tests, resistivity logs etc.

The usefulness of Locke’s nomogram is to compare the effects

of various perforation procedures on well productivity. A
range of typical perforation skin values can be estimated as
follows:

TYPICAL PERFORATION VALUES

Method Perforation Skin

High drawdown, 5in tubing conveyed perforation

(TCP) at 12 spf (90°) with 22 gm charges -1 to 0

Low drawdown, 2 1/8in through-tubing perforation

at 0° phasing with 13 gm charges at 4 spf +1 to +3

Low drawdown or overbalance perforating with

5in casing guns at 4 spf with 22 gm charges at
90° phasing in non-damaging fluids +1 to +5

Overbalance perforating with 5in casing guns at

4 spf (90°) with 22 gm charges in unfiltered or
damaging fluids +5 to +20

Table 4b.1

Aside from considerations of perforation length, shot density

and minimization of the crushed zone by inducing temporary
sand failure, perhaps the most significant parameter to
incorporate into the model is the percentage of perforations
actually open to flow. Once perforations become plugged, only
a small proportion (10 to 25%) may be opened up using
drawdown alone since the differential decreases once a few
perforations start to flow. If the plugged perforations are
randomly distributed, this leads to a reduction in the effective
shot density. However, where entire intervals remain plugged,

Page 10
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

the result will be even more dramatic and will be seen as a

partial completion effect, Sc.

The simulation results have significant implications for the

selection of perforating gun type (TCP, through-tubing or
casing gun), perforating conditions (under vs. overbalanced)
and perforating fluids, as discussed in Section 5a –
Optimizing the Completion in the Near Wellbore Region.

By attempting to match the observed skin value from a well

test with a perforation simulation or nomogram, it is possible
to postulate:

1) If feasible perforation damage or plugging would result in

an Sp value of the required magnitude.

2) The minimum value of Sp that could be expected with the

current perforation configuration, and therefore the
advisability of reperforation.

3) The sensitivity of the conclusions to the input

assumptions.

4) The most appropriate remedial operation.

As with any model study, reliability will improve with local

experience and history matching.

PARTIAL
COMPLETION
SKIN
In situations where less than about 85% of the total net pay
thickness is open, flow will converge to enter the perforated
interval and cause a significant additional pressure drop. The
resultant skin, Sc, is variously called the partial completion
skin, partial penetration skin, geometric skin or pseudo-skin.

If the well has significant formation damage or positive

perforation skin, partial completion will also compound these
skin effects, as shown in Equation 4b.1.

The influence of partial completion can be estimated using

(*Reference 4b.8) the correlations developed by Brons and Marting*.

Page 11
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

28

24

20

10 0
1 000 0
Sc

00
16

100
30
50
12
h' h
rw rw

8
20
10
4
5
2
0 1 hp
0 0.2 0.4 0.6 0.8 1.0 b=
h

(a) (b) (c)

30ft
15ft
0.25ft 6ft

150ft 150ft 150ft

15ft

75ft

Examples of partial well completion showing; (a) Well only partially penetrating the formation;
(b) Well producing from only the central portion of the formation; (c) Well with 5 intervals open to production.

Figure 4b.5 – Partial Completion Skin – Brons and

Marting Chart

Vertical Projection of the Total Interval Perforate

b =
Net Pay Thickness

h p cos θ
=
h

h Symmetry Element Thickness of Producing Zone

rw = Wellbore Radius

Page 12
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

where: θ =
well angle measured from the normal to the
bedding plane (usually near vertical). This
is illustrated by the three examples of
Figure 4b.5.
b = 30/150 = 0.2 (for all 3 cases)
h/rw = 150/0.25 = 600 and Sc = 18 for (a)
h/rw = 75/0.25 = 300 and Sc = 16 for (b)
h/rw = 15/0.25 = 60 and Sc = 9.5 for (c)

(*Reference 4b.9) More rigorous analyses have been presented by Odeh* (1976)
(*Reference 4b.10) and Streltsova-Adams* which confirmed that the partial
completion skin was strongly dependent on the open interval
length, the ratio of horizontal to vertical permeability, other
skin effects and more weakly on the location within the pay.

(*Reference 4b.11) Odeh* (1980) subsequently reformatted his charts into a

general equation which clearly indicates the importance of
the ratio of horizontal to vertical permeability.

0.825
S'c = 1.35 h –1 ln h k / k v + 7 – 1.95
h p cos θ

(Equation 4b.3) – ln rw c 0.49 + 0.1ln h k / k v

where: S'c = partial completion skin

h = pay thickness, ft
hp = length of the perforated interval, ft
θ = well deviation angle, deg
k = horizontal permeability, md
kv = vertical permeability, md
rwc = corrected wellbore radius, ft

If the perforations do not start at the top of the pay:

(Equation 4b.4) r wc = r w e 0.2126 Z m /h + 2.753

(ft)

where: Zm = vertical distance from the top pay to the

middle of the perforated interval, ft
rw = wellbore radius, ft

If perforations are at the top of the zone Z m = 0.5 h p :

rwc = rw (ft)
(Equation 4b.5)

Page 13
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

In most texts, including the original, hp is used in place of

hp cos θ, since the wells are assumed to be vertical.

The concept of limited entry depends on there being a

significant degree of vertical permeability. In highly stratified
reservoirs with no vertical permeability between the sand
layers, the net pay will depend on the interval perforated, hp.
Indeed in the above equations, h refers to the formation
between the nearest layers of zero permeability (eg shales)
and not the total formation thickness as is commonly
assumed. Hence, it is essential to get perforations into each
layer if the reserves are to be fully drained.

When a partial penetration skin is quoted, care should be

taken to ensure that its subtraction from the total skin does
not produce an unrealistic value (eg less than –7). If it does, a
mistake has been made!

Partial penetration skin can be very large and should

generally be avoided. However, a zone may be intentionally
partly perforated as part of a completion strategy, for
example, to avoid coning of water or gas. The effects of the
limited perforated interval on well productivity must then be
incorporated into the production forecast and tubing design.

A high total skin factor calculated from a test on a partially

completed well does not necessarily indicate formation
damage or poor perforations, or make the well an obvious
candidate for stimulation treatment. Each of the contributing
factors to the total skin must first be quantified.

The ability of the reservoir model to calculate partial

penetration effects depends on the number of layers used in
setting up the grid and the assumptions used in developing
the well skin or IPR parameters. This should be discussed
with the reservoir engineer. In general, single well model
results will include partial completion effects and predict the
resultant skin, while for full field models, the reservoir
engineer must input the effects of partial penetration within a
specific layer (although the model will handle the effect
between layers).
DEVIATION SKIN
The converse of partial completion is the apparent thickening
of the completion interval in deviated wells and in steeply
dipping formations. This will be seen as a deviation skin (S"c).

Page 14
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

For a deviated well, perforated throughout a formation, the

(*Reference 4b.12) following expression was derived by Cinco et al* to describe
the skin factor due to well deviation, S"c:

2.06 1.865
(Equation 4b.6) S"c = – θ ' – θ' log h k
41 56 100 rw kv

where: S"c = deviation skin

θ' = tan - 1 k v tan θ

k
θ = well angle measured from the normal to the
formation bedding plane (normally near
vertical)
h = net pay thickness, ft
k = horizontal permeability, md
kv = vertical permeability, md
rw = wellbore radius, ft

0
15°
-1
DEVIA 30°
TION
ANGL
E θ
-2

45°
-3
S c"
-4

60°
-5

-6
75°
-7 4
2 3
10 10 10 10
h
hD = rw

Figure 4b.6 – Cinco’s Nomogram for Deviation Skin

This expression is only applicable to angles of up to 75°. It

can also be applied for apparent thickening due to steeply
dipping formations.

Page 15
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

In the general case of a deviated well that is partially

(*Reference 4b.13) perforated, Cinco-Ley* has shown that the total partial
completion skin, Sc, is the sum of the effects of partial
perforation and deviation:

(Equation 4b.7) Sc = S'c + S"c

Since deviation effects are the converse of partial completion,

it is often sufficient to handle this by using the measured
along-hole perforation length in calculating S'c and ignoring
the S"c term, unless hp>h. This approach has been informally
recommended by several authors of the classic papers
on␣ skin.

Deviation effects may be ignored in reservoir modelling.

Deviated well performance could therefore exceed model
study predictions in this respect.

An obvious extrapolation from the benefits of high deviation

is the use of horizontal wells to increase well productivity, or
to reduce drawdown in order to inhibit coning. Since the flow
pattern is no longer radial and will be highly dependent upon
vertical permeability and reservoir continuity, the prediction
of horizontal well performance falls outside of this general
discussion. However, it is worth noting that flow effects in the
horizontal section will cause variations in drawdown along
the completion.
STIMULATION
SKIN
To produce at their most economic flow rates, many wells
require fracture stimulation treatments (refer to Section 5a –
Optimizing Completion in the Near Wellbore Region). This is a
standard completion practice for wells with permeabilities
less than about 25 md. Fracturing can also be an effective
means to bypass deep formation damage. Fracture
stimulation should always result in a negative skin.

To determine the economics of a fracture stimulation or plan

future field developments, the production engineer must
estimate the productivity improvement expected from the
(*Reference 4b.14) stimulation job. A simplified approach developed by Prats*,
assumes the fracture is equivalent to an increase in the
effective wellbore radius within the conventional radial flow
equations. The resultant skin factor is derived as:

Page 16
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

(Equation 4b.8) Ss = ln 2xrw

f

where: xf = propped length of one wing of the

fracture, ft

This approximation applies only to a moderately sized

fracture (xf ≤ 15% re) that has infinite flow capacity in
comparison to the formation (kf wf > 1000 kr), where wf
is␣ the␣ width of the fracture and kf wf is the fracture
conductivity. This only applies where the flow regime
remains␣ pseudo-radial.

Longer fractures result in linear flow patterns, which

theoretically should give a higher productivity increase.
However, because of the increasing flow resistance in the
fracture, Equation 4b.8 can often still be used as a reasonable
first approximation.

STANDARD FRACTURE
x f < 15% r e

x f ONE WING LENGTH

xf CAN USE DARCY's LAW

FRACTURE IS LIKE AN INCREASED WELL SIZE

S = In 2r w
xf

FLOW IS RADIAL x f = 2rwe -s

LARGE FRACTURE
xf
x f > 15% r e

FRACTURE CURVE MUST BE USED

TO ESTIMATE IMPROVEMENT

FLOW IS LINEAR

Figure 4b.7 – Effect of Fractures on Flow Pattern

in the Reservoir

For formations with higher permeabilities (above about 5 md)

and for longer fractures, the conductivity of the fracture itself
(kf wf) plays an important role in determining the ultimate
productivity increase. Consequently, for large scale jobs

Page 17
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

(xf ≥ 30% re), fracture performance simulators (such as HFS)

must be used to predict the productivity improvement and
resultant skin. However, preliminary estimates of the
stabilized productivity improvement can be obtained using the
(*Reference 4b.15) curves developed by McGuire and Sikora*, presented in the
(*Reference 4b.16) SPE Hydraulic Fracturing Monograph*. Using these curves,
the post fracturing skin can be approximated as:

(Equation 4b.9) Ss ≈ 7 –7
J f/ J o

where: Jf = stabilized productivity of the fractured

well, stb/d/psi
Jo = stabilized productivity of an undamaged
well (S = 0), stb/d/psi

An excellent discussion of post-fracture performance is given

(*Reference 4b.17) by Economides and Nolte* in their Reservoir Stimulation
Manual.

In stimulating very tight rock (k < 1 md), the transient post

fracture production is a significant factor in the treatment
economics and should be used in designing the completion.
Although this can be estimated using the type curves
presented by Economides and Nolte, it is better to use a
fracture and well performance simulator, such as the HFS
program available through the BP Research Centre at
Sunbury.

Although acidization is a stimulation process, it does not

usually result in a significant stimulation skin component. As
discussed in Section 5a – Optimizing the Completion in the
Near Wellbore Region, acid jobs are intended to remove skin
damage and restore the undamaged IPR. In some cases, this
may result in the well achieving its natural negative skin
tendencies (eg, in deviated or naturally fractured wells).

SKIN DUE TO
RESERVOIR
HETEROGENEITY
Many formations, especially carbonates, are naturally
fractured and will exhibit a negative skin once the drilling
damage is bypassed by acidization or by effective
perforations. A similar effect is seen when high permeability
streaks act as a feeder for the rest of the pay. An unexplained
negative skin is therefore generally attributed to reservoir

Page 18
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

heterogeneity, Sh. The magnitude of the resultant skin will

generally be in the range of –1 to –4.

The contrast between vertical and horizontal or radial

permeability, a type of anisotropy, will also affect the skin.
This is accounted for by the use of a correction factor
k / k v in the partial completion terms. Several
investigations have shown how this will also affect the
convergent flow term in the perforation skin. For simplicity,
(*Reference 4b.4) this is ignored in the Locke* nomogram, but is included in
(*Reference 4b.3) most simulators as discussed by Karakas and Tariq*. It is
self-evident that the shot density should be increased as the
anisotropy increases. As discussed in Section 5a, it is
recommended to use 12 spf if the vertical permeability is less
than 1% of the horizontal permeability.
SKIN DUE TO
MULTIPHASE
FLOW
As discussed in Section 4a – Reservoir Performance,
multiphase flow in the reservoir will affect the relative
permeability in the near wellbore area, and will be seen as a
skin effect in well test results. This is generally termed a two-
phase flow effect in oil wells and a condensate blockage effect
in gas wells. The multiphase flow skin is actually a drawdown
dependent term and will therefore be seen in the rate
dependent term, Dq, if multirate testing is undertaken.
However, because of the prevalence of single rate testing in
low rate wells, where these effects are most significant, it is
often modelled as a Darcy term, Sm.

In an oil well that is producing below the bubble point

without high rate flow effects, the multiphase flow skin, Sm,
can be back-calculated from the apparent linear PI, J',
corresponding to a point on the curved IPR.

(*Equation 4.3) By using Vogel’s equation* for the curved IPR of an

undamaged well (ie, all other skin terms are zero), it can
be␣ shown that:

(Equation 4b.10) Sm ≈ 12.6 –7

p
1 + 0.8 pw f
b

(The derivation of this equation is discussed in the BP Near

Wellbore Completion Manual.) Substituting pwf/pb values for

Page 19
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

1.0 0% DRAWDOWN S m = 0.0

0.8 A
PP 25% DRAWDOWN S m = + 0.875
A
R
EN VO
T G 50% DRAWDOWN S m = + 2.0
P / P avg

0.6 LI EL
N
EA IP
R R IDE
IP (S AL
R o= LIN 75% DRAWDOWN S m = + 3.5
0.4 ,J EA
'(S 0. RI
0) PR
m ,J
= * (S
5.
6) o= 100% DRAWDOWN S m = + 5.6
0.2 0.0
)

0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

q / q max

So = OTHER SKIN S m = TWO PHASE SKIN S = TOTAL SKIN

Figure 4b.8 – Multiphase Flow Skin in Oil Wells Below the Bubble Point

key points on the IPR curve specifies the expected range

of␣ Sm:

Pwf/Pb Multiphase Flow Skin (Sm)

1.0 0.0
0.75 +0.9
0.5 +2.0
0.25 +3.5
0.0 +5.6

This shows that if a well is tested with a bottom hole flowing

pressure well below the bubble point, a small positive skin
does not necessarily indicate formation damage or poor
perforation efficiency.

As discussed in Section 4a – Reservoir Performance, the

existence of skin damage reduces the two-phase flow effects,
causing the oil well IPR to become more linear and reducing
the multiphase flow component.

Condensate blockage results from the build-up of liquid

around the wellbore of a gas well produced below the dew
point. The effects are generally most significant in low
permeability zones (<50 md).

Page 20
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

The size and effect of the condensate bank depends on how

the relative permeability to gas and liquid varies with
(*Reference 4b.18) saturation. Although Golan* presents an equation for
estimating the radius of the condensate bank after a given
flow time, it is generally better to conduct reservoir studies on
the effects using a single well model.

The Hawkins’ equation (Equation 4b.2) can be used to predict

the multiphase flow skin from the radius of the condensate
bank and the relative permeability to gas in the condensate
invaded zone. As a first approximation, the gas permeability
at the critical liquid saturation is often used, if a saturation
profile is not available.

Page 21
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

NON-DARCY SKIN

As discussed in the Introduction to this chapter and in the

Theory section of Chapter 4, high rate flow effects will also
cause an additional rate dependent pressure drop in the near
wellbore area that will be seen in the total skin, S', from a
well test. If several build-ups have been recorded from
different rates (as in a modified isochronal test), the Darcy
and non-Darcy components can be separated using
Equation␣ 4a.12 by plotting the total skin versus test rate.

RATE DEPENDENT SKIN DUE TO NEAR WELLBORE TURBULENCE

+ 75 -3
D = 2.1 x 10 / (stb/d)

∆ SKIN
SLOPE = =D
∆ RATE
APPARENT SKIN, S'

+ 50

BUILD-UP AFTER MAXIMUM

RATE TEST AT 16 000 stb/d

BUILD-UP AT 9500 stb/d

+ 25

BUILD-UP AT 3300 stb/d

S
DARCY SKIN FACTOR S = 10
0

0 5000 10 000 15 000 20 000 25 000

FLOWRATE, stb/d

Figure 4b.9 – Rate Dependent Skin Due to

non-Darcy Flow Effects

It is useful to be able to estimate the magnitude of the non-

Darcy effects theoretically for situations where only a single
rate test is available, and in order to estimate the
improvements that can be achieved by good completion
practices or remedial treatments.

Page 22
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

ONSET OF NON-
DARCY FLOW
(*Reference 4b.19) As early as 1937, Muskat* showed that deviations from Darcy
flow could be expected if the Reynold’s number exceeded 1.0.
(*Reference 4b.18) Golan* has presented a more practical limit as the point
where the high rate flow effects exceed 5% of the total
pressure drop, given by the following equations:

19 × 10 3 µo rw hp2 ln rr e + S
w
(Equation 4b.11) q o, HVF > stb/d
ρ o Bo h k / k a

1.4 × 10 3 µg rw hp2 ln rr e + S
w
(Equation 4b.12) q g, HVF > mscf/d
γg h k / k a

where: qo, HVF = critical oil rate for high velocity flow
effects, stb/d
qg ,HVF = critical gas rate for high velocity flow
effects, mscf/d
rw = wellbore radius, ft
hp = perforated interval open to flow, ft
h = total net pay thickness, ft
µ = gas or oil viscosity, cp
γg = reservoir gas gravity, (air = 1)
ρo = live oil density at flowing pressure, lb/ft3
Bo = oil FVF at flowing pressure, bbl/stb
k = formation permeability, md
ka = near wellbore permeability, md
S = Darcy skin factor

From this, it can be seen that gas wells, and oil wells
producing large volumes of free gas at the formation face,
will␣ quickly reach non-Darcy flow conditions. However,
non-Darcy flow effects do not become significant in oil wells
with normal GLRs, unless rates exceeds some 100 stb/d/ft,
or an internal gravel pack is installed (as discussed later).

NON-DARCY SKIN
IN GAS WELLS
The gas well productivity discussion in Section 4a – Reservoir
Performance presents the derivation and general form of the
non-Darcy skin component, D, in terms of the non-Darcy flow
coefficient, B.

Page 23
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

B kg h
(Equation 4a.24) D = m s c f / d- 1
1422 T

2.22 × 10 –15 γg k g h β
(Equation 4a.25) D = ×r m s c f / d- 1
h p2 µ w f w

where: β = inertial coefficient at the perforated

interval,␣ ft–1
hp = effective perforated interval, ft (this
depends on the number of perforations
effectively open)
µwf = gas viscosity at pwf , cp
γg = gas gravity, (air = 1)

The problems of estimating the β factor are discussed in

Section 4a. In fact, separate β factors should be used for the
damaged zone, βd, and reservoir rock, βr, so that it can be
shown that:

2.22 × 10 –15 γg k g h β d r d – rw
(Equation 4b.13) Dd = × (mscf/d)–1
h p2 µ w f r d rw

2.22 × 10 –15 γg k g h r –r
(Equation 4b.14) Dr = × β r re r d (mscf/d)–1
h p µw f
2 e w

where: Dd = non-Darcy skin component due to damage,

mscf/d–1
rd = damage zone radius, ft
Dr = non-Darcy skin component in the reservoir
rock, (mscf/d)–1
re = drainage radius, ft

[NB: Usually (re – rd) ≈ (re – rw) ≈ re]

There may be additional non-Darcy components associated

with the complex flow patterns around perforations, which
may be increased by plugged perforations, gravel packs and
small scale heterogeneity.

Page 24
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

NON-DARCY SKIN
IN OIL WELLS
A similar set of equations can be developed for the non-Darcy
skin components (D in [stb/d]–1) in a high rate oil well:

1.635 × 10 –16 ρo k oh Bo β d r d – r w
(Equation 4b.15) Dd = × r d rw (stb/d)–1
µo h p2

1.635 × 10 –16 ro koh Bo β r r e – r d

(Equation 4b.16) Dr = × r e rd (stb/d)–1
µo h p2

[NB: Usually re – rd ≈ re]

The same core flow measurements or correlations (Equations

4a.22 and 4a.23) can be used for estimating β in oil wells, as
were discussed for gas wells in Section 4a – Reservoir
Performance.
PRACTICAL
CONSIDERATIONS
WITH RESPECT
TO NON-DARCY
SKIN
In practice, it is extremely difficult to estimate the non-Darcy
skin effects because of uncertainties over:

• The near wellbore permeability in the perforated interval.

• The extent and magnitude of any damage.

• The number of perforations effectively producing.

• The relationship between β and the measured rock

properties.

However, theoretical calculations of the non-Darcy skin from

the reservoir rock properties, assuming that all other damage
can be removed, can provide a lower boundary.

It is also important to recognize how sensitive the non-Darcy

skin is to the effective perforation interval, the near wellbore
permeability and the flow rate. Good perforating and clean-up

Page 25
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

procedures and stimulations designed to open up plugged

perforations therefore have a dramatic effect on high rate
wells.

Because of the uncertainties in estimating the non-Darcy

skin, it is strongly recommended that a modified isochronal
test, with build-ups from several test rates, be used in testing
high rate or high GLR oil wells and gas wells.

Page 26
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

EFFECTS OF GRAVEL PACKS

The following subsections describe the skin effects associated

with internal and external gravel packs in both oil and gas
wells. In higher rate wells, non-Darcy flow effects will
dominate. However, high Darcy skins can be expected from
poorly designed packs. For an extensive discussion on the
productivity of gravel packed wells, the reader is referred to
(*Reference 4b.18) Golan*, Section 3.6 and to the BP Near Wellbore Manual.
Gravel pack configurations and methods are discussed in
Section 5a – Optimizing the Completion in the Near Wellbore
Region.
SKIN EFFECTS
FROM AN
INTERNAL
GRAVEL PACK
IN OIL WELLS
In an internal gravel pack, high permeability gravel
(50 to 1300 D) is placed in the perforation tunnels and
annular space between the casing and a screen, and is sized
to inhibit formation sand from invading the pack and
reducing its permeability.

SCREEN CASING CEMENT

PERFORATIONAL TUNNEL

GRAVEL / SAND INTERFACE

FORMATION SAND

VOID CREATED BY WASHING

FILTER SIZE

LINEAR FLOW PATH

SAND
GRAVEL

Figure 4b.10 – Internal Gravel Pack Configuration

Page 27
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

The gravel filled perforations have a finite conductivity, kgAp,

causing a flow restriction and a resultant skin, S'g, since the
cross-sectional flow area is small. Using the flow equations
(*Reference 4b.20) for a linear flow system, Saucier* showed that the pressure
losses across a cased hole gravel pack are:

µo L Bo qo Bo q o 2
(Equation 4b.17) ∆p G = + 9.1 × 10 –13 βG L ρ o psi
1.127 × 10 –3 k G Ap Ap

where: Ap = total area of all open perforations, ft2

kG = effective downhole permeability of the
gravel, md
L = length of linear flow path, ft
ρo = downhole live oil density, lbs/cu ft
βG = inertial flow coefficient for gravel, ft–1

Substituting this expression into Equation 4a.11, the

generalized equation for skin effect, we obtain the gravel pack
skin equation:

S'G = 2 π kh L + 6.43 × 1 0 –15 kh β G L ρ o o 2o

(Equation 4b.18)
B q
k G Ap µo Ap
so that:

(Equation 4b.19) SG = 2 π kh L
k G Ap

(Equation 4b.20) DG = 6.43 × 10 –15 kh β G L ρ o Bo 2 s t b/ d –1

µo Ap
Four main parameters influence the magnitude of gravel pack
skin:

• Downhole Gravel Permeability (kG)

Estimated values for typical downhole gravel
permeabilities are as follows:

Gravel Size Downhole Permeability

US Mesh (md)
8 – 12 1 000 000 – 1 500 000
10 – 20 300 000 – 600 000
20 – 40 100 000 – 200 000
40 – 60 25 000 – 75 000

Page 28
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

• Length of the Linear Flow Path

To allow for convergence, the linear flow path is assumed
to be about 1in longer than the distance from the OD of
the cement sheath to the casing ID.

Bit Size – Casing ID + 2

(Equation 4b.21) L = (ft)
24

• Area of Linear Flow Path

The perforation area is:

π EH 2
(Equation 4b.22) Ap = h p Νp Ep ft 2
576

where: EH = diameter of the perforation or entrance

hole (EH), in
Ep = perforation efficiency (ie, fraction of
perforations open to flow)
Np = shot density, spf

Plugging of the perforations with sand, LCM material, dirt

or fluid residues is the main cause of high skins in gravel
packed wells. In designing a gravel pack, it is generally
assumed that 30 to 70% of the perforations remain
plugged, depending on the perforation clean-up and
gravel placement technique (refer to the example in
Chapter 4).

• The Inertial Flow Coefficient for Gravel (βG)

The effective downhole βG factor for gravel is generally
recognized as being higher than that estimated with the
(*Reference 4b.21) Katz correlation (Equation 4a.21). Cooke* proposed the
following correlation for the standard 20:40 gravel:

3.4 × 10 1 2
(Equation 4b.23) β G= (ft–1 )
k G1.54

This results in βG of 68 000 (ft–1) if kG = 100 000 md.

However, several experts privately recommend using even
higher values of βG (100 000 to 200 000 ft–1) to reflect the
intermixing that must occur at the sand gravel interface.

Page 29
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

SKIN EFFECTS
FROM AN
INTERNAL
GRAVEL PACK
IN GAS WELLS
The pressure losses and skin effect from an internal gravel
pack in a gas well are strongly dependent on flowrate, since
high velocity flow effects will occur in the reservoir rock as
well as in the gravel filled perforations. Moreover the
computations are complicated by the dependence of gas
(*Reference 4b.22) properties on pressure. McLeod* shows that the pressure
drop across the gravel pack is:

8936 µ g T Z L q g
p 2wfs – p 2 w f =
k G Ap

2
qg
(Equation 4b.24) + 1.25 × 10 –10 γg Z T L β G psi 2
Ap

substituting into Equation 4a.8 yields:

2 π kh 8.79 × 10 –14 γg L βG kh q g
(Equation 4b.25) S'G = +
k G Ap Ap2 µ g

so

(Equation 4b.26) SG = 2 π khL

k G Ap
and

γ g L βG kh
(Equation 4b.27) DG = 8.79 × 10 –14 m s c f / d –1
A µg
2
p

where: pwfs = sand face flowing pressure, psi

qg = gas rate, mscf/d
DG = non-Darcy skin coefficient, (mscf/d)–1
L = length of linear flow path, ft (refer to
Equation 4b.20)

Page 30
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

Ap = total area of all open perforations, ft2 (refer

to Equation 4b.21)
βG = inertial flow coefficient for gravel, ft–1 (refer
to Equation 4b.22)

The gas properties (µg, Z) should be evaluated at the average

of the bottom hole flowing pressure and the sand face
pressure.
SKIN EFFECTS
FROM EXTERNAL
GRAVEL PACKS
An external gravel pack is achieved by first underreaming the
pay section and then placing gravel between the sand face
and a screen suspended in the centre of the borehole.

OPEN HOLE GRAVEL PACKS

UNDERREAMED UNDERREAMED
WINDOW BELOW CASING

Figure 4b.11 – External Gravel Packs

Due to the high permeability radial flow areas associated with

an external gravel pack, very little pressure drop occurs
across the pack unless sand gravel intermixing occurs.
Reasonable estimates of well productivity can be made by
assuming the effective wellbore radius, r'w, is the
underreamed diameter. Alternatively, the Hawkins’ equation
(4b.2) can be used to calculate the gravel pack skin, using an
assumed downhole gravel permeability.

Page 31
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

The major source of flow restriction in an external pack is

associated with damage which may occur at the interface of
the pack and the underreamed hole. Such damage can arise
from the build-up of a filter cake during underreaming, the
use of LCM during tripping or by sand invasion into
improperly sized gravel. This damage is difficult to assess or
treat. Consequently, the high theoretical productivities of
external gravel packs can be much reduced. However, if such
damage can be avoided, the productivity of an external gravel
pack will exceed that of the equivalent internal gravel pack,
especially in high rate gas wells.

Page 32
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

IMPLICATIONS FOR COMPLETION DESIGN AND

WORKOVER PLANNING

In this section, we have considered a number of the factors

which contribute to a well’s near wellbore performance.
Specifically, we have examined a number of components of
the total skin. Clearly, in planning a workover or stimulation
treatment, a well’s productivity may be improved by either
reducing the Darcy skin factor, S, or the non-Darcy
coefficient, D.

The skin factor can be decreased by avoiding, removing

or␣ bypassing formation damage using good perforating
practices, and by avoiding partial completion skin. Hydraulic
fracturing can be used to create a negative skin while large
acid treatments may restore naturally occurring negative
skins in carbonate formations.

The non-Darcy component, D, can be decreased by

minimizing any reduction in permeability in the perforated
zone, hence reducing the inertial coefficient, β, and by
maximizing the perforated interval, hp, and the number of
perforations open. Since hp appears as a squared term in the
denominator of the expression for non-Darcy skin, increasing
hp can decrease rate dependent skin effects dramatically.
Good perforating and clean-up operations and stimulations
designed to open up plugged perforations can therefore
greatly improve gas well performance.

Page 33
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4b – NEAR WELLBORE PERFORMANCE Completion Design Manual

REFERENCES

4b.1 Jones, L G and Watts, J W: ‘Estimating Skin Effect

in␣ a Partially Completed Damaged Well’ J Pet Tech
(Feb 1971), 249–252

4b.2 Hawkins, M F, Jr: ‘A Note on the Skin Effect’, Trans,

AIME (1956) 207, 356–357

4b.3 Karakas, M and Tariq, S M: ‘Semianalytical

Productivity Models for Perforated Completions’,
SPE␣ 18247, 63 ATC , Oct 1988

4b.4 Locke, S: ‘An Advanced Method for Predicting the

Productivity Ratio of a Perforated Well’, J Pet Tech
(Dec␣ 1981) 2481–2488

4b.5 Tariq, S M: ‘Evaluation of Flow Characteristics of

Perforations Including Non-Linear Effects with the
Finite-Element Method’, SPE Prod Eng, (May 1987)
104–112

4b.6 API RP 43: ‘Recommended Practice for Standard

Procedure for the Evaluation of Well Perforators’, 4th
Edition (August 1985)

4b.7 King, G A, Anderson, A and Bingham, M: ‘A Field

Study of Underbalance Pressures Necessary to
Obtain␣ Clean Perforations using Tubing-Conveyed
Perforating’, SPE 14321, 60th ATC, Las Vegas,
(Sept␣ 1985)

4b.8 Brons, F and Marting, VE: ‘The Effect of Restricted

Fluid Entry on Well Productivity’, J Pet Tech
(Feb␣ 1961), 172–174; Trans, AIME 222

4b.9 Odeh, A S: ‘Pseudo-Steady-State Flow Capacity of

Oilwells with Limited Entry and with an Altered Zone
Around the Wellbore’, SPE 6132, 51 AFTC, New
Orleans, (1976)

4b.10 Streltsova-Adams, T D: ‘Pressure Drawdown in a Well

with Limited Flow Entry’, SPE 7486, 53 AFTC,
Houston (1978)

Page 34
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4b – NEAR WELLBORE PERFORMANCE

4b.11 Odeh, A S: ‘An Equation for Calculating Skin Factor

Due to Restricted Entry’, J Pet Tech (June 1980),
964–965

4b.12 Cinco, H, Miller, F G and Ramey, H J: ‘Unsteady State

Pressure Distribution Created by a Directionally
Drilled Well’, J Pet Tech (Nov 1975), 1392–1400

4b.13 Cinco-Ley, H: ‘Pseudo-Skin Factors for Partially-

Penetrating Directionally-Drilled Wells’, SPE 5589,
50␣ AFTC, Dallas, (1975)

4b.14 Prats, M: ‘Effects of Vertical Fractures on Reservoir

Behaviour – Incompressible Fluid Case’, Soc Pet Eng J,
(June 1961), 105–118

4b.15 McGuire, W J and Sikora, V J: ‘The Effect of Vertical

Fractures on Well Productivity’, Trans, AIME (1960)
219, 401–403

4b.16 Howard G C and Fast, C R: Hydraulic Fracturing, SPE

Monograph, Volume 2, SPE, Dallas 1970

4b.17 Economides, M J and Nolte, D G: Reservoir

Stimulation, Schlumberger Education Services,
Houston (1987)

4b.18 Golan, M and Whitson, C H: Well Performance,

International Human Resources Development
Corporation, Boston, NY (1986)

4b.19 Muskat, M: The Flow of Homogeneous Fluids

Through␣ Porous Media , McGraw-Hill Book Co Inc,
New␣ York (1937)

4b.20 Saucier, R J: ‘Gravel Pack Design Considerations’,

SPE␣ 4030, 47th AFTM (1972), J Pet Tech (Feb 1974),
205–212

4b.21 Cooke, C E: ‘Conductivity of Fracture Proppants in

Multiple Layers’, SPE 4117, 47th AFTM, 1972,
SPE␣ Reprint Series 5A , Vol II, Dallas (1978), 220–226

4b.22 McLeod, H O and Crawford, H R: ‘Gravel Packing High

Rate Completions’, SPE 11008, 57th AFTC, New
Orleans (1982)

Page 35
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

CONTENTS

INTRODUCTION 1

MULTIPHASE FLOW CONCEPTS 4

PHASE BEHAVIOUR 5
DEFINITION OF VARIABLES 9
FLOW PATTERNS 18
LIQUID HOLD-UP 24
DIMENSIONAL ANALYSIS 24

PRESSURE GRADIENT PREDICTION 26

CONSERVATION PRINCIPLES 26
PRESSURE GRADIENT PREDICTION METHODS 28

COMPUTING ALGORITHM 37

TEMPERATURE PREDICTION 40
WELLBORE HEAT TRANSFER 41
TEMPERATURE PREDICTION 42

PRESSURE DROP ACROSS CHOKES 45

PRESSURE GRADIENT CURVES 48

REFERENCES 50
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

INTRODUCTION

The accurate design of oil and gas well tubing strings requires
the ability to predict flow behaviour in the wells. All wells
produce a mixture of gas and liquids, regardless of whether
they are classified as oil wells or gas wells. This multiphase
flow is significantly more complex than single-phase flow.
However, the technology to predict multiphase flow behaviour
has improved dramatically in the past decade. It is now
possible to select tubing sizes, predict pressure drops and
calculate flowrates in wells with acceptable engineering
accuracy.

The common occurrence of multiphase flow in wells can be

discussed with the simplified production system shown in
Figure 4c.1. Fluids entering the wellbore from the reservoir
may be an undersaturated oil or a single-phase gas. Free
water may accompany the fluids as a result of water coning,
waterflooding or production of interstitial water. Alternately, a
free gas saturation in an oil reservoir may result in a gas-
liquid mixture entering the well. Retrograde condensation can
result in hydrocarbon liquids condensing in a gas condensate
reservoir so that a gas-liquid mixture again enters the
wellbore. Even when single-phase gas or liquid flow exists
near the bottom of a well, multiphase flow can occur
throughout most of the wellbore. This is due to evolution of
gas from oil or condensation of gas with reduction of pressure
and temperature as the fluids flow up the well.

Although many of the wells drilled on land tend to be nearly

vertical, wells drilled offshore and in other hostile
environments such as the Arctic are directional. Inclination
angles can vary from vertical near the surface to horizontal
near the production zone. Flowrates of gas, oil and water vary
widely. Pipe diameters can be as small as 1in and as large as
9in. Flow can also occur in a casing-tubing annulus. Depths
can range from a few hundred feet to over 20 000 ft (6096 m).
Pressures can be as small as a few atmospheres or as high as
20 000 psia. Temperatures can be over 400°F (204°C) or
approach the freezing point of water. Oil viscosities can be
less than 1 cp or greater than 500 000 cp.

Page 1
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

FLOWING SEPARATOR
WELLHEAD
PRESSURE TO SALES

STOCK
TANK

INCLINED FLOW
HORIZONTAL FLOW

IN UBIN
VERTICAL

CL
T

IN G
INTAKE ED FLOW THROUGH
POROUS MEDIA
pR , k, IPR

Figure 4c.1 – Multiphase Flow Production System

Most wells contain some type of well control device that

requires produced fluids to flow through a restriction. This
can vary from a bottom hole choke to a subsurface safety
valve or a remotely controlled variable sized surface choke.
Wells can be produced by artificial lift mechanisms involving
a submersible pump or gas injection.

The broad variations in flow variables encountered in

producing wells have made the development of prediction
methods much more difficult. Techniques that work for gas
condensate wells do not necessarily work for oil wells.
Assumptions that are valid for some wells are totally invalid
for others.

Numerous methods have been proposed for predicting

multiphase flow in wells. Some are very general, and others
are applicable over a very narrow range of flow variables.
Some are very empirical in nature, and others make an
attempt to model the basic phenomena encountered. It is
vitally important that persons engaged in designing wells are
aware of the limitations of all methods. Thus, the primary
objective of this subsidiary topic is to present the important
methods for predicting flow behaviour in wells and discuss
their limitations and ranges of applicability.

Page 2
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

Engineers engaged in designing wells should have a thorough

understanding of the calculation procedures necessary to
optimize their design calculations. This includes coupling the
wells to both surface facilities and reservoirs. It includes
understanding computer algorithms, and it includes an
understanding of fluid physical properties and mass transfer
for multicomponent mixtures. A significant part of this
section is devoted to these concepts.

Page 3
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

MULTIPHASE FLOW CONCEPTS

When two or more phases flow simultaneously in pipes, the

flow behaviour is much more complex than for single-phase
flow. The phases tend to separate due to differences in
density. Shear stresses at the pipe wall are different for each
phase as a result of their different densities and viscosities.
Expansion of the highly compressible gas phase with
decreasing pressure increases the in situ volumetric flowrate
of the gas. As a result, the gas and liquid phases normally do
not travel at the same velocity in the pipe. For upward flow,
the less dense, more compressible and less viscous gas phase
tends to flow at a higher velocity than the liquid phase, a
phenomenon known as slippage. However, for downward flow
the liquid often flows faster than the gas.

Perhaps the most distinguishing characteristic of multiphase

flow is the variation in the physical distribution of the
phases, a characteristic known as flow pattern or flow
regime. During multiphase flow through pipes, the flow
pattern that exists depends on the relative magnitudes of the
forces that act on the fluids. Hydrostatic, turbulence and
surface tension forces vary significantly with flow rates, pipe
diameter, inclination angle and fluid properties of the phases.
Several different flow patterns can exist in a given well as a
result of the large pressure and temperature changes that the
fluids encounter. Especially important is that the pressure
gradient varies significantly with flow pattern. Thus, the
ability to predict flow pattern as a function of the flow
parameters is of primary concern.

Analytical solutions are available for many single-phase flow

problems. Even when empirical correlations are necessary,
ie␣ for turbulent flow friction factors, the accuracy of
prediction is excellent. The increased complexity of
multiphase flow logically resulted in a higher degree of
empiricism for predicting flow behavior. Many empirical
correlations have been developed for predicting flow pattern,
slippage between phases, friction factors, etc for multiphase
flow in pipes. Virtually all of the existing standard design
methods rely on these empirical correlations. However, since
the mid 1970s dramatic improvements in understanding the
fundamental mechanisms that govern multiphase flow have
taken place. These have resulted in new predictive methods
that rely much less on global empirical correlations.

Page 4
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

In this section, basic definitions of flow parameters unique to

multiphase flow in pipes are introduced and discussed. Flow
patterns are described in detail, including methods available
for predicting their occurrence. The pros and cons of using
empirical correlations based on dimensional analysis and
dynamic similarity are presented.
PHASE
BEHAVIOUR
A complex mixture of hydrocarbon compounds or
components can exist as a single-phase liquid, single-phase
gas or as a two-phase mixture, depending on the pressure,
temperature and the composition of the mixture. Unlike a
single component or compound, like water or carbon dioxide,
a multicomponent mixture will exhibit an envelope rather
than a single line on a pressure-temperature diagram when
two phases exist simultaneously. A typical phase diagram
for␣ a multicomponent hydrocarbon system is given in
Figure␣ 4c.2 . The shapes and ranges of pressure and
temperature within actual envelopes vary widely with
composition.

Several terms in Figure 4c.2 require definition.

• Critical Point. Pressure and temperature at which gas and

liquid are indistinguishable, and bubble and dewpoint
curves merge together.

• Bubble Point Curve. Curve that separates single-phase oil

from a two-phase mixture.

• Dewpoint Curve. Curve that separates single-phase gas

from a two-phase mixture.

• Cricondentherm. Highest temperature possible for a

two-phase mixture to exist.

• Liquid Volume Fraction. A curve within the two-phase

envelope that represents a constant volume fraction of
liquid in the hydrocarbon mixture.

Figure 4c.2 permits a qualitative classification of the types of

reservoirs that are encountered in oil and gas systems.
Typical oil reservoirs have temperatures below the critical
temperature of the hydrocarbon mixture. Gas condensate
reservoirs normally have temperatures between the critical
temperature and the cricondentherm for the hydrocarbon

Page 5
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

4000
OIL GAS-CONDENSATE GAS
RESERVOIRS RESERVOIRS RESERVOIRS

CRICONDENTHERM = 250 °F
A

Tc = 127 °F
3500 B
CRITICAL
POINT

TION
RESERVOIR PRESSURE, psia

3000

DUC
B1
DE

PRO
W
NT PO
POI
2500 IN

H OF
B LE T
UB

PAT
B
% B2
80
2000
%
40
ID %
QU ME 20
1500 I
L LU
VO
%
10

5%
1000
B3 0%

A2
500
0 50 100 150 200 250 300 350
RESERVOIR TEMPERATURE, °F

Figure 4c.2 - Multicomponent Phase Diagram

mixture. Dry gas reservoirs have temperatures above the

cricondentherm.

Many condensate fluids exhibit retrograde condensation, a

phenomenon in which condensation occurs during pressure
reduction rather than with pressure increase as for most
gases. This abnormal or retrograde behaviour occurs in the
shaded region within the two-phase envelope of Figure 4c.2.
For example, a reservoir with initial pressure and
temperature at point B will remain as single-phase gas as
depletion occurs until the dewpoint pressure at point B1 is
reached. Further pressure reduction will result in
condensation of gas until point B2 is reached. This
corresponds to the maximum liquid volume fraction of 10%.
Subsequent pressure reduction results in vaporization of the
condensed liquid and a reduction in liquid volume fraction

Page 6
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

until eventually a second dewpoint pressure is reached when

all of the liquid has vaporized.

Mass transfer occurs continuously between the gas and the

liquid phases within the two-phase envelope of Figure 4c.2 as
pressures and temperatures change. All attempts to describe
mass transfer assume that equilibrium exists between the
phases. Two approaches have been used to simulate mass
transfer for hydrocarbon systems: the ‘black oil’ (or constant
composition model) and the (variable) compositional model.
Each is described in the following sections.
Black Oil Model
The term black oil is a misnomer and refers to any
hydrocarbon liquid phase that is relatively non-volatile. This
would include hydrocarbons produced from oil reservoirs in
Figure 4c.2. These oils are typically dark in colour, have API
gravities less than approximately 40°API and undergo
relatively small changes in composition within the two-phase
envelope. Thus, a better description might be a constant
composition model.

For black oils with associated gas, a simplified parameter

called the ‘solution gas-oil ratio’, Rs, has been defined to
account for gas that dissolves (condenses) or evolves (boils)
from solution in the oil. Rs is expressed as scf/stb and can be
measured in the laboratory or determined from empirical
correlations. A summary of available correlations is given in
Table 4.9 in the Techniques section of Chapter 4.

A second parameter called the ‘oil formation volume factor’,

Bo, has also been defined to describe the shrinkage or
expansion of the oil phase. Oil volume changes occur as a
result of changes in dissolved gas and also because of the
compressibility and thermal expansion of the oil. By far the
most important factor causing volume change is dissolved
gas. Bo is expressed as bbl/stb and can also be measured in
the laboratory or predicted with empirical correlations. A
summary of available correlations is given in Table 4.9 in the
Techniques section of Chapter 4.

Once the black oil model parameters are known, oil density
and other physical properties of the two phases can be
calculated. Methods for predicting these properties are given
in the BP MULTIFLO Manual.

Page 7
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

When water is also present, solution gas-water ratio, RSw,

and␣ water formation volume factor, Bw, can be defined.
Correlations for these parameters and physical properties of
the water are also given in the BP MULTIFLO Manual. The
amount of gas that can be dissolved in water and the
corresponding possible changes in water volume are much
smaller than for gas-oil systems.
Compositional
Model
For volatile oils and condensate fluids, vapour-liquid
equilibrium (VLE) or ‘flash’ calculations are the only method
with acceptable accuracy for describing mass transfer. Given
the composition of a fluid mixture or ‘feed’, a VLE calculation
will determine the amount of the feed that exists in the
vapour and liquid phases, and the composition of each
phase. From these results, it is possible to determine the
quality or mass fraction of gas in the mixture. Once the
composition of each phase is known, it is also possible to
calculate the densities, enthalpies and viscosities of each
phase and the interfacial tension. VLE calculations and
prediction of other pressure-volume temperature (PVT)
properties can be performed with the BP GENESIS
flowsheeting package.

VLE calculations are considered more rigorous than using

black oil model parameters to describe mass transfer.
However, they are also much more difficult to perform. If a
detailed composition is available for a gas-oil system, it is
possible to generate black oil parameters from VLE
calculations. However, the nearly constant composition that
results for the liquid phase and the increased computation
requirements make the black oil model more attractive for
non-volatile oils.

Volumetric
Flowrates
After mass transfer calculations are completed, it is possible
to calculate the in situ volumetric flowrates of each phase.
For the black oil model, volumetric flowrates are determined
from:

(Equation 4c.1) q o = q oSCBo

(Equation 4c.2) q w = q wSCBw

(Equation 4c.3) q g = q gSC– q oSCRs – q wSC R sw Bg

Page 8
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

where: qo = volumetric flowrate of oil, ft3/sec

qw = volumetric flowrate of water, ft3/sec
qg = volumetric flowrate of gas, ft3/sec
Bo = oil formation volume factor, bbl/stb
Bw = water formation volume factor, bbl/stb
Bg = gas formation volume factor, ft3/sec
Rs = solution gas/oil ratio, scf/stb
Rsw = solution gas/water ratio, scf/stb

and the subscript ‘sc’ means ‘standard conditions’,

ie␣ 14.7␣ psia and 60°F.

Bg is derived from the engineering equation of state to be:

p SC Z T
(Equation 4c.4) Bg =
p ZSC T SC

where: pSC = standard conditions pressure, psia

ZSC = standard conditions gas compressibility
factor
TSC = standard conditions temperature

For the compositional model, volumetric flowrates are

calculated from

w t 1–x
(Equation 4c.5) qL =
ρL

qg = wt x
(Equation 4c.6)
ρg

where: qL = volumentric flowrate of liquid, ft3/sec

qg = volumetric flowrate of gas, ft3/sec
wt = mass flowrate, lbm/d
x = physical property correction factor
ρL = density of liquid, lbm/ft3
ρg = density of gas, lbm/ft3

DEFINITION
OF VARIABLES
When performing multiphase flow calculations for wells,
single-phase flow equations are often modified to account for
the presence of a second phase. This involves defining
mixture expressions for velocities and fluid properties that
use mixture weighting factors based on either volume or

Page 9
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

mass fractions. The choice of variables and weighting factors

often depends upon the flow pattern predicted to exist. Each
of these parameters is discussed in the following sections.

Mixture
Weighting
Factors
When gas and liquid flow together up a well, the gas normally
travels faster than the liquid. Under steady state conditions
this results in a reduced area fraction for the gas phase and
an expanded area fraction for the liquid phase. Thus,
slippage of gas past liquid results in an accumulation of
liquid in the pipe, a phenomenon known as liquid hold-up.
Liquid hold-up can be defined as the fraction of a pipe cross
section or volume increment that is occupied by the liquid
phase. The difference in velocities of the gas and liquid
phases can be estimated from a variety of equations
developed from experimental data for a limited range of flow
conditions. A more common procedure is to develop empirical
correlations for predicting liquid hold-up for a broad range of
flow conditions.

For the case of equal phase velocities, or no-slip conditions,

the volume fraction of liquid in the pipe can be calculated
analytically from a knowledge of the in situ volumetric
flowrates given in the previous section. Thus,

q
(Equation 4c.7) f Lns = q +L q
L g

where: f Lns = no-slip liquid hold-up

and where qL is the sum of the oil and water flowrates for the
black oil model, or is given by Equation 4c.5 for the
compositional model. If free water exists when using the
compositional model, the water flowrate must be added to the
oil or condensate flowrate to account for all the liquid. Since
the no-slip liquid hold-up can be determined rigorously, it is
commonly used as a correlating parameter for other
multiphase flow parameters, such as fL.

When oil and water flow simultaneously in pipes, with or

without gas, it is possible for slippage to occur between the
oil and water phases. However, this type of slippage is
negligible compared to the slippage that can occur between

Page 10
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

gas and liquid. As a result, the oil fraction in a liquid phase is

normally calculated from:

q
(Equation 4c.8) f o = q +o q
o w

where: fo = volume fraction of oil in liquid

The water cut, fw, based on in situ rather than stock tank
flowrates, is simply 1–fo .

When performing calculations with the compositional model,

it is more common to deal with mass fractions than volume
fractions. Consequently, a no-slip quality can be obtained
from the results of a VLE calculation as follows:

VMg
(Equation 4c.9) x =
VMg+ LML

where: V = vapour mole fraction, moles gas/feed

Mg = molecular weight of gas, lbm/mole
L = liquid mole fraction, moles liquid/feed
ML = molecular weight of liquid, lbm/mole
Velocities
As described previously, for multiphase flow in wells the
individual phase velocities are normally quite different. Only
for the case of the highly turbulent dispersed bubble flow
pattern, in which the fluids exist as a homogeneous mixture,
are the phase velocities essentially equal. For all other cases,
significant slippage can occur between the gas and liquid.
Under steady state flow conditions, slippage will result in a
disproportionate amount of the slower phase being present at
any given location in the well. As mentioned previously, the
in situ volume fraction of liquid, fL, must be determined from
empirical correlations. Important correlating parameters are
superficial or (pseudo) liquid and gas velocities, vSL and vSG,
which assume a given phase occupies the entire pipe area.

Thus,

qL
(Equation 4c.10) vSL =
Ap

Page 11
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

qg
(Equation 4c.11) vSg =
Ap

A total or mixture velocity can then be defined as:

qL + qg
(Equation 4c.12) vm = = vSL + vSg
Ap

where: vSL = superficial liquid velocity, ft/sec

vSg = superficial gas velocity, ft/sec
Ap = cross-sectional area of pipe, ft2
vm = mixture velocity, ft/sec

Velocities averaged over time and pipe cross-section for each

phase can be calculated from a knowledge of the time and
space averaged liquid hold-up obtained from the empirical
correlations. Thus:

(Equation 4c.13) vL = vSL

fL

vSg
(Equation 4c.14) vg =
1 – fL

A slip velocity can be defined as the difference between the

actual phase velocities:

(Equation 4c.15) vs = vg – vL

where: vL = liquid velocity, ft/sec

fL = liquid hold-up
vg = gas velocity, ft/sec
vs = slip velocity, ft/sec

A variety of other velocities are encountered in multiphase

flow that pertain to flow mechanisms in specific flow
patterns. Examples are rise velocities of small bubbles and
larger bullet shaped bubbles that often occur in bubble and
slug flow patterns. These variables are used primarily in the
calculation of mixture density.

Page 12
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

Mixture Fluid
Properties
Regardless of whether the black oil or the compositional
model is used, it is possible to calculate the density and
viscosity of each phase (oil, water and gas) and the gas-oil
and gas-water surface tensions. A variety of methods have
then been used to define mixture fluid properties.

Oil-Water Mixture
For the black oil model case, and for the compositional model
case when free water exists, oil and water properties are
combined as follows:

(Equation 4c.16) ρ L = ρ o fo + ρ w fw

(Equation 4c.17) σL = σo fo + σw fw

(Equation 4c.18) µL = µo fo + µw fw

where: ρ = density, lbm/ft3

µ = viscosity, cp
and subscripts o = oil
w = water
L = liquid
σL = gas-liquid mixture surface tension,
dynes/cm
σo = gas-oil surface tension, dynes/cm
σw = gas-water surface tension, dynes/cm
fo = volume fraction of oil in liquid
fw = volume fraction of water in liquid

The primary use of viscosity is in calculation of Reynolds

numbers for predicting friction factors. Since the pressure
drop due to friction is a minor component of the total
pressure drop in most wells, viscosity is not normally
considered to be a significant factor. Nevertheless, studies
have shown that Equation 4c.18 is often not valid for the
viscosity of two immiscible liquids like oil and water.
Although oil-water mixtures often behave as non-Newtonian
fluids, an alternate approach is to treat the mixture as a
Newtonian fluid with an apparent viscosity which, when used
in the traditional Reynolds number expression, will yield an
accurate friction factor from the Moody diagram shown in
Figure 4c.3.

Page 13
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

Numerous studies have been published following this

(*Reference 4c.1) approach, beginning with the paper by Woelflin*, which
predicted an exponential increase in mixture viscosity with
increasing water fraction. An attempt to generalize this trend
resulted in the so called API emulsion viscosity equation

(Equation 4c.19) µL = µoe 3.6 f w

where: µL = liquid viscosity, cp

µo = oil viscosity, cp

0.1

0.09
LAMINAR CRITICAL TRANSITION
FLOW ZONE ZONE FULLY ROUGH ZONE
0.08
0.05
0.07
0.04
0.06
0.03
LAM

0.05 0.02

RELATIVE ROUGHNESS, D
e
INAR 64
f = Re

0.015
FRICTION FACTOR, f

FLO

0.04
0.01
W

0.08
0.006
0.03 Re
cr 0.004

0.025
0.002

0.02 0.0001
0.0008
0.0006
0.0004
0.015
0.0002
SMOOTH
PIPES 0.0001

0.000,05
0.01

0.009
0.008 0.000,01
3 2 3 4 5 6 8 4 2 3 4 5 6 8 5 2 3 4 5 6 8 6 2 3 4 5 6 8 7 2 3 4 5 6 8 8
10 10 10 10 10 10
e e
= 0.000,001 = 0.000,005
pVD D D
REYNOLDS NUMBER, Re µ

Figure 4c.3 – Moody Diagram

Page 14
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

Unfortunately, Equation 4c.19 does not properly describe the

phenomenon of wall shear stress. The viscosity of a
dispersion or emulsion has been found to depend mainly
upon determining which phase is continuous. The apparent
liquid viscosity will then be governed primarily by the
viscosity of the continuous phase since this is the phase that
predominates at the pipe wall where most of the friction
losses occur. Other factors such as the dispersed phase
viscosity and the droplet size distribution of the dispersed
phase have also been found to be important.

For some oil-water systems, it has also been observed that

the viscosity of the liquid mixture can be several times
greater than the oil viscosity when the continuous phase is
oil but the water fraction is approaching the point where an

4
10

LAMINAR FLOW
REGIME LINE

3
10
OIL VISCOSITY, cp

2
10

10

1
0 0.2 0.4 0.6
INPUT WATER FRACTION REQUIRED
TO INVERT THE MIXTURE

Figure 4c.4 – Water Fraction Required to Invert an

Oil-Water Mixture

Page 15
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

inversion of the dispersion or emulsion will occur so that

water becomes the continuous phase. This increase has been
attributed to two different causes: a deviation from
Newtonian behaviour near the inversion point and an energy
loss requirement to invert the mixture that is not accounted
for in the conservation equations used. Thus, the energy loss
appears as an additional pressure loss that is seen as an
increase in apparent viscosity when using a Newtonian fluid
flow model. The inversion point of an oil-water mixture has
been found to occur at water fractions ranging from 0.2 to
0.5 with inversion taking place at lower water fractions when
oil viscosities are high. A plot demonstrating this for
experimental data from a variety of sources is given in
Figure 4c.4. For example, if the oil viscosity at a given
location in a pipe is 10 cp, then oil will be the continuous
phase for water fractions up to about 0.4. For higher water
fractions, water will be the continuous phase. Furthermore, it
has been found that inversion can be accomplished at even
lower water fractions by the use of surfactants.

Although the most common way of treating the apparent

viscosity of an oil-water mixture is with Equation 4c.18, a
more accurate method would be to use the oil viscosity when
oil is the continuous phase and the water viscosity when
water is the continuous phase. Figure 4c.4 could then be
used to estimate the water fraction at which inversion takes
place. An even better alternative would be to conduct flow
tests on actual crudes and water to determine the rheological
characteristics of the fluids and the probable inversion point.
This approach is further complicated by the possibility of a
gas phase changing the water fraction at which inversion
takes place.

Gas-Liquid Mixture
Numerous equations have been proposed for describing PVT
properties of gas-liquid mixtures. In general, these equations
are referred to as ‘slip’ or ‘no-slip’ properties, depending upon
whether fL or fLns is used as the volumetric weighting factor.
Thus, for the case of two-phase viscosity:

(Equation 4c.20) µs = µL fL + µg 1 – f L

(Equation 4c.21) µs = µL fL × µ g 1 – fL

(Equation 4c.22) µn = µL fLns + µg 1 – f Lns

Page 16
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

where: µs = slip liquid mixture viscosity, cp

µg = gas viscosity, cp
µn = no-slip liquid mixture viscosity

Although Equation 4c.20 has never been used in multiphase

flow design correlations, Equation 4c.21 is used by the
Hagedorn and Brown correlation described later. All other
empirical correlations use Equation 4c.22.

The following expressions have been used by various

investigators to calculate multiphase flow mixture densities.

(Equation 4c.23) ρ s = ρ L fL + ρ g 1 – f L

(Equation 4c.24) ρ n = ρ L fLns + ρ g 1 – f Lns

(Equation 4c.25) f 2
1 – f Lns 2
ρk = ρL L ns
+ ρg
fL 1 – fL

where: ρs = slip mixture density, lbm/ft3

ρn = no-slip mixture density, lbm/ft3
ρk = ‘kinetic’ mixture density, lbm/ft3

Equation 4c.25 uses the subscript k because it appears in the

kinetic energy term for the specific case of a homogeneous
mixture linear momentum conservation equation.

When performing temperature change calculations for

multiphase flow in wells, it is necessary to predict the
enthalpy of the multiphase mixture. Most VLE calculation
methods, including the BP GENESIS flowsheeting package,
have a provision to predict the enthalpies of the gas and
liquid phases. If enthalpies are expressed per unit mass, then
the enthalpy of a multiphase mixture can be calculated from

(Equation 4c.26) h = h L 1 – x + h gx

where: h = mixture enthalpy, BTU/lbm feed

hL = liquid enthalpy, BTU/lbm liquid
hg = gas enthalpy, BTU/lbm gas

Page 17
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

FLOW
PATTERNS
As a multiphase mixture flows up a well, the gas and liquid
can be distributed in a variety of configurations called flow
patterns. Predicting the flow pattern that occurs at a given
location in a well is extremely important since the empirical
correlation or mechanistic model used to predict pressure
drop varies with flow pattern. Numerous investigators have
described flow patterns in wells and made attempts to predict
when they occur. Consequently, a concensus of how to
classify flow patterns does not exist.
Flow Pattern
Classification
For upward multiphase flow of gas and liquid, most
investigators now recognize the existence of four flow
patterns:

• Bubble flow.
• Slug flow.
• Annular-slug transition (churn) flow.
• Mist (annular) flow.

These flow patterns, shown schematically in Figure 4c.5,

are␣ described below. Slug and churn flow are sometimes
combined into a flow pattern called intermittent flow.

• Bubble Flow
Bubble flow is characterized by a uniformly distributed
gas phase as discrete bubbles in a continuous liquid
phase. Based on the presence or absence of slippage
between the two phases, bubble flow is further classified
into bubbly and dispersed bubble flows. In bubbly flow,
relatively fewer and larger bubbles move faster than the
liquid phase due to slippage. On the other hand, in
dispersed bubble flow numerous tiny bubbles are
transported by the liquid phase, causing no relative
motion between the two phases. Dispersed bubble flow is
often referred to as froth flow.

Page 18
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

BUBBLE SLUG CHURN ANNULAR

FLOW FLOW FLOW FLOW
(ANNULAR-SLUG (MIST FLOW)
FLOW)

Figure 4c.5 – Description of Vertical Flow Patterns

• Slug Flow
Slug flow is characterized by a series of slug units, each of
which is composed of a gas pocket called a Taylor bubble,
a plug of liquid called a slug and a film of liquid around
the Taylor bubble flowing downwards. For vertical flow,
the Taylor bubble is an axially symmetrical bullet-shaped
gas pocket that occupies almost the entire cross-sectional
area of the pipe. The liquid slug, carrying distributed gas
bubbles, bridges the pipe and thus separates two
consecutive Taylor bubbles.

• Annular-Slug or Churn Flow

Churn flow is a chaotic flow of gas and liquid in which
both the Taylor bubbles and the liquid slugs are distorted
in shape. Neither of the phases appears to be continuous.
The continuity of the liquid in the slug is repeatedly
destroyed by a high local gas concentration. Typical of
churn flow is the oscillatory or alternating direction of
motion of the liquid phase. Previous investigators often
considered churn flow as a transition region between slug
flow and mist flow.

Page 19
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

• Mist or Annular Flow

Annular flow is characterized by a continuous gas phase
in a central core with the liquid flowing upwards, both as
a thin film along the pipe wall and as dispersed droplets
in the core. At high gas flowrates, more liquid becomes
dispersed in the core, leaving a very thin liquid film
flowing along the wall. The interfacial shear stress acting
at the core-film interface and the amount of entrained
liquid in the core are important parameters in annular
flow. Some investigators have called this flow pattern
mist␣ flow.

Flow Pattern
Occurrence
The following is a description of a typical sequence of flow
patterns in an oil wellbore. Starting with a flowing bottom
hole pressure above the bubble point pressure, only a liquid
phase exists at the bottom. As the liquid moves upward, it
experiences pressure decline resulting in the liberation of
some of the gas dissolved in the liquid phase. The liberated
gas appears as small bubbles in the continuous liquid phase,
which characterizes the bubble flow pattern. As the flow
continues upward, further decrease in pressure and
temperature occurs, resulting in gas expansion and the
liberation of more solution gas from the oil phase. This
creates more and larger bubbles that start coalescing with
each other. The coalescence creates larger Taylor bubbles
separated by the continuous liquid phase. The slug flow
pattern thus occurs. Further continuation of the upward
motion of the flow into the region of lower pressure causes
the expansion of Taylor bubbles along with the liberation of
more gas from liquid slugs. This creates a chaotic two-phase
flow, defined earlier as churn flow. The churn flow continues
to exist until the gas flowrate is sufficiently high to push the
liquid against the pipe wall. This characterizes the existence
of annular flow.

Thus, due to the continuous changes in pressure,

temperature and mass transfer between the two phases, the
flow pattern having a continuous liquid phase at the bottom
hole can be completely transformed into a flow pattern with a
continuous gas phase at the wellhead. This progressive
change of flow patterns is shown in Figure 4c.6.

Page 20
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

ANNULAR FLOW
(MIST FLOW)

CHURN FLOW
(ANNULAR-SLUG FLOW)

SLUG FLOW

BUBBLE FLOW

LIQUID FLOW

Figure 4c.6 – Typical Flow Pattern Changes in a Well

Flow Pattern
Prediction

Methods to predict the occurrence of the various flow

patterns in wells have fallen into two categories, empirical
maps and mechanistic models.

Most early attempts were based on conducting experimental

tests in small diameter pipes at low pressures using air and
water. Flow patterns were observed, and the values of various
flow parameters at the transition between flow patterns were
determined. Empirical flow pattern maps were then drawn
that could be used to predict the transitions. Some of the
more frequently used maps developed with this approach are
shown in Figures 4c.7 and 4c.8. The coordinates in Figure
4c.7 are defined in a later section on Dimensional Analysis.

Page 21
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

2
10

10
FROTH FLOW
5

N Lv REGION I REGION II REGION III

N
TIO
Ls

NSI
1

TRA
BUBBLE
5 FLOW
Lm

2
PLUG HEADING SLUG
-1 FLOW FLOW MIST 1
10
-1 2 5 2 5 2 5 2 2 5 3
10 1 10 10 10
N Gv
ρL 0.25 ρL 0.25
N Lv = 1.938 VSL N Gv = 1.938 VSG
σL σL

Figure 4c.7 – Duns and Ros Flow Pattern Map

ρG
1 1
ρ σ /4 /3
L WA
Y =
ρ σ
X = ρA
W L

X, Y — PHYSICAL PROPERTIES CORRECTION FACTORS

100

10
ft / sec

BUBBLE
1
YVSL

SLUG ANNULAR
0.1 MIST
FROTH

0.01
0.1 1 10 100 1000
XVSG ft / sec

Figure 4c.8 – Aziz et al Flow Pattern Map

Page 22
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

The second method for predicting flow patterns considers the

basic mechanisms that are important in causing a flow
pattern change. A typical map for a specific set of geometrical
parameters and PVT properties is given in Figure 4c.9 and is
based on the Taitel et al model. The following mechanisms
have been found to govern flow pattern transitions:

• Bubble-Dispersed Bubble
Turbulence due to high liquid velocities causes a
homogeneous, no-slip behaviour.

• Bubble-Slug
Increased gas superficial velocity results in the liquid
hold-up falling to a critical value of 0.75, below which gas
bubbles coalesce to form Taylor bubbles.

• Dispersed Bubble-Slug
Superficial gas velocity reaches a sufficiently high value
for no-slip liquid hold-up to have minimum possible value
for dispersed bubble flow of 0.48. Below this, gas bubbles
must coalesce regardless of turbulence.

• Annular Mist
Gas velocity is just high enough to drag largest liquid
droplets and the liquid film on the pipe wall upward.

20
10
DISPERSED BUBBLE
SUPERFICIAL LIQUID VELOCITY, m/s

C
B

1 BARNEA
BUBBLY TRANSITION

ANNULAR

0.1 D
A
SLUG OR CHURN

0.01

0.002
0.02 0.1 1 10 100

SUPERFICIAL GAS VELOCITY, m/s

Figure 4c.9 – Taitel et al Vertical Flow Pattern Map

Page 23
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

Flow pattern transitions predicted with empirical maps often

do not accurately predict flow patterns beyond the range of
variables used to create the maps. This is probably the major
source of errors in predicting pressure gradients for flow
pattern dependent correlations. The mechanistic model
approach appears to be valid for all ranges of flow variables,
but must be validated more extensively prior to widespread
adoption in the industry.

LIQUID HOLD-UP
Liquid hold-up is the most critical variable in the calculation
of pressure drop in wells. This is because of its importance in
calculating mixture density in Equation 4c.23 for use in the
hydrostatic head component in the pressure gradient
equation.

Only for homogeneous flow when no slippage occurs and the

gas and liquid travel at the same velocity can liquid hold-up
be predicted rigorously. For all other cases, liquid hold-up
will depend on the flow pattern and the degree of slippage
that exists. This has led to numerous empirical correlations
that are usually a function of flow pattern and probably valid
for a limited range of flow conditions. As in the case of flow
pattern prediction, the most successful liquid hold-up or
mixture density predictions are based on mechanistic models
that attempt to describe the complex interactions between
the gas and liquid phases.

DIMENSIONAL
ANALYSIS
The use of dimensional analysis to solve engineering
problems that cannot be solved analytically is a powerful tool
that has been used successfully in many engineering fields,
including single-phase fluid mechanics and heat transfer.
The increased complexity of multiphase flow has logically
resulted in even broader use of the same principles. When
performed properly, and with the availability of appropriate
experimental data, use of dimensionless groups to scale flow
behavior beyond the range of data used to develop empirical
correlations can be very successful. Failure to include all
pertinent variables will always result in empirical correlations
with limited accuracy.

Many different dimensionless groups are encountered in the

multiphase flow literature. Commonly encountered are
groups that are also used in single-phase flow such as
Reynolds numbers, Froude numbers, Prandtl numbers,

Page 24
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

relative roughness, etc. The classic dimensional analysis for

(*Reference 4.14) multiphase flow in wells was performed by Duns and Ros*.
They identified 13 variables that were important, resulting in
10 independent dimensionless groups. Of these, they
determined that the following four were especially important
in determining flow pattern and/or liquid hold-up:

ρL 0.25
(Equation 4c.27) N Lv = 1.938 vSL
σL

ρL 0.25
(Equation 4c.28) N g v = 1.938 vSg
σL

ρL 0.5
(Equation 4c.29) N D = 120.872 D
σL

(Equation 4c.30) N L = 10.15726 µL 1

ρ L σL 3

where: NLv = dimensionless liquid velocity number

NGv = dimensionless gas velocity number
ND = dimensionless pipe diameter number
D = pipe inside diameter, ft
NL = dimensionless liquid viscosity number

Equations 4c.27 to 4c.30 have already been converted to oil

field units. Most subsequent investigators made use of the
Duns and Ros dimensionless groups to develop empirical
correlations for flow pattern and liquid hold-up.

Page 25
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

PRESSURE GRADIENT PREDICTION

CONSERVATION
PRINCIPLES
The basis for calculating pressure drop during fluid flow in
pipes is conservation of mass and linear momentum. This is
true for both single-phase and multiphase flow. Following is
the development of the pressure gradient equation for steady
state single-phase flow.
Conservation
of Mass
Conservation of mass simply means that for a given control
volume such as the segment of pipe shown in Figure 4c.10,
the mass in minus mass out must equal the mass
(*Reference 4c.2 Knudsen) accumulation*.

dL
dZ

dX

Figure 4c.10 – Control Volume

For a constant diameter pipe:

∂ρ ∂
(Equation 4c.31) + ρυ = 0
∂t ∂L

For steady state flow, rv = constant, and no mass

accumulation can occur. Equation 4c.31 then becomes
(Equation 4c.32)
d ρv = 0
dL

Page 26
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

Conservation
of Linear
Momentum

Application of Newton’s First Law to fluid flow in pipes

requires that the rate of momentum out minus the rate of
momentum in (efflux) plus the rate of momentum
(*Reference 4c.2 Knudsen) accumulation in a given pipe segment must equal the sum of
all forces on the fluids*. Figure 4c.10 defines the control
volume and pertinent variables.

Conservation of linear momentum can be expressed as

∂ ρv ∂ ρv 2 ∂p ρ g sin θ
(Equation 4c.33)
g + g =– – τ πD – gc
∂t c
∂L c
∂L A

where: gc = gravitational constant, lbf ft/lbm sec2

g = acceleration of gravity, ft/sec2

The first term on the left hand side of Equation 4c.33

represents the rate of momentum accumulation, and the
second term represents the rate of momentum efflux, which
can be expanded as follows
∂ ρv 2 v d ρ v + ρ v dv
(Equation 4c.34) =
∂L g c gc dL gc dL

Combining Equations 4c.32 to c.34 and assuming steady

state flow [∂(ρv)/∂t = 0] to eliminate the rate of accumulation
of linear momentum, gives:
ρ v dv dp πD ρ g sin θ
gc dL = – dL – τ A –
(Equation 4c.35)
gc

Solving for the pressure gradient, one obtains

(Equation 4c.36) dp ρ g sin θ f ρ v 2 ρ v dv

= – gc – –
dL 2 gcD gc dL

where: f = Moody friction factor

Page 27
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

where the wall shear stress was used as

ρ v2
(Equation 4c.37) τ = f
8 gc

Equation 4c.36 is also frequently called the mechanical

energy balance equation. Thus, the steady state pressure
gradient equation is a result of applying the principles of
conservation of mass and linear momentum.

Equation 4c.36 clearly shows that the steady state pressure

gradient is made up of three components. The first term in
Equation 4c.36 is the pressure gradient due to elevation
change (often called hydrostatic head or elevation
component). It is normally the predominant term in wells and
contributes from 80 to 95% of the pressure gradient. The
second component is due to friction or shear stress at the
pipe wall and normally represents 5 to 20 % of the total
pressure drop in wells. The final component is due to change
in velocity (often called acceleration or kinetic energy
component). It is normally negligible and can become
significant only if a compressible phase exists at relatively
low pressures.

For flow up a well, pressure always drops in the direction of

flow. It is common to adopt a sign convention that pressure
drop is positive in the direction of flow. Equation 4c.36 must
then be multiplied by –1 to yield a positive pressure gradient.
PRESSURE
GRADIENT
PREDICTION
METHODS

Two fundamentally different methods have been used to

predict multiphase flow behavior in pipes. The most rigorous
approach is to write equations for conservation of mass,
linear momentum and energy for each phase. This results in
a six-equation formulation that must be solved
simultaneously with appropriate auxiliary expressions for
variables such as interfacial friction factor, liquid
entrainment, film thickness, etc. Although this approach has
had significant success in predicting transient flow behavior
in multiphase pipelines, it has limited application in wells
where the gas and liquid phases are seldom continuous. The
second approach is to treat the gas-liquid mixture as a
pseudo single-phase that obeys the pressure gradient

Page 28
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

equation for single-phase flow given in Equation 4c.36. All

methods for predicting flow behaviour in wells rely to some
extent on this method.
Homogeneous
Methods
Homogeneous methods assume that the gas and liquid flow
at the same velocity in the well. For this case the liquid
hold-up in the pipe is the no-slip liquid holdup as predicted
by Equation 4c.7. All of the early multiphase flow correlations
were homogeneous methods and differed only by the
empirical correlations used to obtain a multiphase flow
friction factor. The resulting pressure gradient equation is

dp ρ n g sin θ f TP ρ n vm 2
(Equation 4c.38) = gc +
dL 2 gc D

where: fTP = two-phase Moody friction factor

Three of the correlations in this category, the Poettmann and

(*References 4c.3-4c.5) Carpenter*, Baxendell and Thomas* and Fancher and Brown*
correlations all attempted to correlate friction factor with the
(*Reference 4.19) numerator of a mixture Reynolds number. The Cornish*
correlation uses a friction factor from the Moody diagram for
a mixture Reynolds number. Of these, only Cornish is
available in MULTIFLo.

These methods cannot predict accurate pressure gradients

except for high velocities where dispersed bubble flow exists,
and should not be used as a generalized correlation.

Slip Flow
Methods
Slip flow methods recognize that the gas and liquid do not
travel at the same velocity, but are independent of flow
pattern. The most commonly used methods in this category
(*References 4.21, 4.20) were presented by Hagedorn and Brown* and Gray*. For this
method, the pressure gradient is obtained from the following
equation:

ρ s g sin θ f TP ρ TP vm 2
gc +
dp 2 gc D
(Equation 4c.39) =
dL 1 – EK

where EK is a dimensionless kinetic energy term defined by:

Page 29
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

ρ TP d vTP
g
(Equation 4c.40) EK = c dL
dp
dL

Thus, the only difference between the homogeneous and slip

correlations is in the selection of densities and friction factor
and the inclusion of the kinetic energy component.

Hagedorn and Brown Correlation

(*Reference 4.21) The Hagedorn and Brown* correlation was developed from
data obtained from a 1400 ft deep vertical test well. Fluids
used were air and either water or dead crudes with viscosities
of 10, 30 and 110 cp. A major limitation of the study was that
liquid hold-up was not measured but was calculated based
on Equation 4c.39 representing the measured pressure
gradient. The resulting pseudo liquid hold-up values were
then correlated using the four Duns and Ros dimensionless
groups given by Equations 4c.27 to 4c.30.

Several modifications have been made to the Hagedorn and

Brown correlation over the years that have greatly improved
its accuracy. The original correlation should never be used.
However, the modified correlation has proven to be one of the
best available. The following is a summary of the
modifications.

• The Hagedorn and Brown liquid hold-up correlation often

gives values less than the no-slip liquid hold-up. When
this occurs, the no-slip liquid hold-up must be used.

• The Hagedorn and Brown correlation has been found to

perform poorly for the bubbly flow pattern. If Griffith and
(*References 4c.6-4c.7) Wallis* predict bubble flow to exist, the Griffith* bubble
flow correlation should be used. Note that this is identical
(*Reference 4.25) to the bubble flow correlation used in the Orkiszewski*
correlation.

• The kinetic energy pressure gradient component

recommended by Hagedorn and Brown predicts too large
a value. The kinetic energy component should be
neglected in bubble and slug flow.

(*Reference 4.14) • If Duns and Ros* predict mist flow or the transition
region␣ to occur, then the Duns and Ros correlation should
be used.

Page 30
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

Gray Correlation
(*Reference 4.20) Gray* developed a slip model based on Equation 4c.39 for use
in the API 14B Subsurface Controlled Subsurface Safety
Valve sizing computer program. Empirical correlations were
presented for predicting liquid hold-up and pseudo wall
roughness when mist flow occurs. A revised kinetic energy
term may also have been used.
Flow Pattern
Dependent
Methods
All of the more modern correlations for predicting pressure
gradients in wells recognize the importance of both flow
pattern and slippage. Empirical maps or correlations are first
used to predict the flow pattern that exists at a given location
in the tubing. Equation 4c.39 is then used to calculate
pressure gradient, with different liquid hold-up and friction
factor correlations being used for each flow pattern.

Kinetic energy effects are normally considered only for mist

flow, and the Duns and Ros formulation is almost always
used. For this case, the dimensionless kinetic energy
parameter is defined as

vm vSG ρn
(Equation 4c.41) EK = gc p

Equation 4c.41 clearly shows that kinetic energy effects are

important at low pressures. Like the Hagedorn and Brown
formulation, this one also tends to overpredict. Equation
4c.41 can be related to the Mach number in compressible
flow. When it reaches a value of 1.0, the pressure gradient
becomes indeterminate and a shock would be established in
the pipe. Unfortunately, Equation 4c.41 approaches 1.0 too
rapidly, resulting in an overprediction of the pressure
gradient.

Duns and Ros Correlation

(*Reference 4.14) The Duns and Ros* correlation was the first of the
comprehensive flow pattern dependent studies. It was based
on measured pressure drop and liquid hold-up data obtained
in a 187 ft high, transparent pipe with diameters ranging
from 1.25 to 5.60in. Fluids used were air and either water or
a hydrocarbon liquid, and most tests were conducted at near
atmospheric conditions. The empirical flow pattern map in

Page 31
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

Figure 4c.7 was developed from the data. In addition,

empirical correlations for a dimensionless slip velocity
number were developed for bubble and slug flow, from which
slip velocity and thus liquid holdup can be calculated for use
in Equations 4c.23 and 4c.39. A unique treatment for
predicting friction losses in mist flow was presented that
involved increasing the roughness of the pipe and decreasing
the area available for flow to account for the liquid film on the
pipe wall. No modifications to the Duns and Ros correlation
have been published, although Shell Oil Co. has made at
least three improvements.

• In 1963 Ros modified the correlation using data from 17

high GOR vertical oil wells. The nature of the revisions is
unknown.

• From 1974 to 1976, a joint study conducted by Shell and

Mobil resulted in the Moreland Mobil-Shell Method
(MMSM). Data from 40 vertical and 21 deviated wells was
used to develop simplified liquid hold-up correlations for
the bubble and slug flow patterns. The remainder of the
method is based on the Duns and Ros correlation.
Improved statistical results are claimed and the resulting
computer program has been marketed to other
companies. This method is not currently available in
MULTIFLO.

(*Reference 4.41) • Reinicke et al* developed a method for predicting pressure

drops in high water cut gas wells that involved using the
Duns and Ros mist correlation and interpolating
procedure for the transition region, but using the Gray
criteria for the mist-nonmist flow pattern transition. This
method is not available in MULTIFLO.

Aziz, Govier and Fogarasi Correlation

(*Reference 4.22) In 1972 Aziz et al* presented a correlation based on the flow
pattern map given in Figure 4c.8 and liquid hold-up and
friction factor correlations developed from laboratory data.
The liquid hold-up correlations make use of mechanistic
models for bubble rise velocities in bubble and slug flow.
If␣ mist flow is predicted by Figure 4c.8, the Duns and Ros
mist flow correlation is used.

Beggs and Brill Correlation

The first comprehensive study of multiphase flow for inclined
(*Reference 4.23)
pipes was presented by Beggs and Brill in 1973*. Pressure

Page 32
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

drop and liquid hold-up measurements were obtained for flow

of air and water through 90 ft long transparent pipe with
diameters of 1.0 and 1.5in. Empirical liquid hold-up
correlations for horizontal flow and a correction factor for
inclination angle were developed as a function of the flow
pattern that existed for horizontal flow. A multiphase friction
factor correlation was also developed that is based on smooth
pipe and also includes a minor discontinuity that probably
does not affect accuracy. The Duns and Ros kinetic energy
component was recommended for all flow patterns, even
though it is normally small except for the mist (distributed)
flow pattern.

This correlation has been found to overpredict liquid hold-up

in wells and pipelines. However, the availability of a single
correlation for both wells and pipelines has resulted in
widespread use. In an attempt to improve liquid hold-up
predictions, Palmer conducted a study in which natural gas
and water was flowed through a hilly terrain flowline with
inclination angles of +10 to –10 deg from horizontal. Simple
correction factors to the Beggs and Brill liquid hold-up
predictions were developed which included multiplying all
uphill liquid hold-up values by 0.924. Other modifications
that are normally used with the Beggs and Brill correlation
are:

• Use of a rough pipe rather than a smooth pipe

normalizing friction factor. This will have little effect on
pressure loss in wells but gives improved answers for
pipelines.

• A revised horizontal flow pattern map is used which

eliminates a discontinuity in liquid hold-up when
changing from segregated to intermittent flow.

Orkiszewski Correlation
After concluding that none of the existing pressure gradient
correlations was sufficiently accurate, Orkiszewski combined
the Griffith bubble flow correlation, the Duns and Ros mist
flow correlation and a new development for slug flow into a
comprehensive pressure gradient prediction method. The slug
flow development combined the Griffith and Wallis Taylor
bubble rise velocity correlation with a new parameter called
the liquid distribution coefficient to obtain the slip mixture
density and the friction component. Empirical equations for
the liquid distribution coefficient, depending on the

Page 33
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

continuous liquid phase and the mixture velocity, were

developed using the data of Hagedorn and Brown. A
discontinuity in the liquid distribution coefficient can occur
when changing equations at a mixture velocity of 10 ft/s.
This discontinuity can cause a convergence problem in the
computing algorithm described later for calculating pressure
drop.

(*Reference 4.32) • Chierici et al* recorrelated the Griffith and Wallis Taylor
bubble rise velocity data in an attempt to improve the
Orkiszewski correlation. The statistical performance of the
(*Reference 4.24) result was worse than the original correlation*. (Not
available in MULTIFLO.)

(*Reference 4c.8) • Triggia* eliminated the discontinuity where the liquid

distribution coefficient equations change by translating
one equation while retaining the slope. This removed the
convergence problem, but the statistical performance of
the result was slightly worse than the original correlation.
(Not available in MULTIFLO.)

• The Orkiszewski correlation with oil as the continuous

liquid phase at all water cuts is also available in
MULTIFLO.

Mukherjee and Brill Correlation

(*Reference 4c.9) In 1985 Mukherjee and Brill* presented the results of an
experimental study on multiphase flow in inclined pipes. Air
with either kerosene or lube oil was flowed through
transparent 1.5in diameter pipe at angles from +90 to –90
degrees from horizontal. Liquid hold-up was measured with a
new capacitance type sensor. A series of empirical
correlations using the Duns and Ros dimensionless groups
was developed for predicting flow pattern at all inclination
angles. For each flow pattern, empirical correlations were
developed for liquid hold-up and friction factor. Like the
Beggs and Brill correlation, it has enjoyed some use since it
is applicable to both wells and pipelines. This method is not
currently available in MULTIFLO.
Mechanistic
Models
Many of the flow pattern dependent correlations described
above included mechanistic models for part of their
development. This included models (or lower orders of
empirical correlations) for such parameters as bubble rise

Page 34
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

velocity in bubble flow, Taylor bubble rise velocity in slug flow

and film thickness in mist (annular flow). Not included were
parameters such as length of slugs, collision rates of small
gas bubbles, liquid entrainment fraction in mist flow, film
velocity surrounding a Taylor bubble, etc. Of greatest
importance is that none of the correlations incorporated an
adequate mechanistic model for predicting flow pattern
transitions. It is of little use to have mechanistic models to
predict flow behaviour for different flow patterns if equivalent
accuracy is not available for predicting which flow pattern
exists.

The development of mechanistic models for predicting flow

patterns has permitted formulating comprehensive
mechanistic models for predicting pressure gradients at all
inclination angles encountered in wells. Especially important
(*Reference 4.15) has been the work by Taitel et al* and Barnea*.
(*Reference 4.18)

Mechanistic models for predicting flow behaviour in one flow

(*References 4c.10-4c.16) pattern have been presented by numerous investigators*.
However, only recently have attempts been made to combine
(*References 4c.17-4c.20) these into comprehensive models*. Before these models can
be adopted as reliable design tools, further validation is
required using field data from a broad spectrum of well types.

Caetano (TUFFP)
(*Reference 4c.17) Caetano* conducted an experimental investigation of
multiphase flow in a 52.5 ft long vertical annulus consisting
of 3.0in ID casing and 1.66in OD tubing. Tests were
conducted with both air-water and air-kerosene mixtures and
with both concentric and fully eccentric annuli. Modifications
to the Taitel et al flow pattern prediction model were made for
flow in an annulus. Mechanistic models were formulated
based on experimental observations and on existing models
for specific flow patterns in pipe flow. The resulting
mechanistic model should be valid for vertical wells in which
production is through the casing-tubing annulus. Future
modifications may be necessary for deviated wells. This
model will shortly become available in MULTIFLO.

Ozon et al
In a joint study in Boussens, France by ELF Aquitaine, IFP
and Total-CFP, a comprehensive mechanistic model was
developed for predicting pressure drop in wells at any
inclination angle. Tests were conducted in 98 ft long pipes
with diameters of 3.0 and 6.0in using condensate, crude oil

Page 35
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

and water as liquid phases and natural gas. Details of the

model were only partially presented since the resulting
computer program, WELLSIM, is proprietary and may be
marketed by the sponsors.

Hasan and Kabir

(*Reference 4c.19)
Hasan and Kabir* made minor changes in the flow pattern
prediction models of Taitel et al. For each flow pattern,
mechanistic models and correlations developed by numerous
investigators were combined to yield pressure gradient
equations. The resulting comprehensive model compared
favorably with laboratory scale liquid hold-up and pressure
gradient data from three independent studies. This model is
not currently available in MULTIFLO.

Ansari (TUFFP)
(*Reference 4c.20)
Following an approach similar to Hasan and Kabir, Ansari*
developed a comprehensive mechanistic model that was
based primarily on combining the studies of Barnea* for flow
(References 4.18, 4c.14-
4c.16) pattern prediction and Ansari and Sylvester*, Alves et al *
and Sylvester* for pressure gradient prediction. The TUFFP
well databank was expanded to include high capacity wells
from the Prudhoe Bay Field. The resulting databank of 1775
flowing well surveys was used to evaluate the comprehensive
model in comparison to various empirical correlations. The
Ansari model is available in MULTIFLO.

Page 36
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

COMPUTING ALGORITHM

Calculating the pressure profile in a well involves integrating

the pressure gradient equation from either the bottom of a
well to the top or vice versa. The most commonly used
algorithm is to divide the well into a given number of length
increments and assume a constant average pressure gradient
is valid over each length increment. The procedure is trial
and error since the average pressure gradient is a function of
average pressure, which depends on the pressure drop across
the increment. The total pressure drop becomes:
L

n
(Equation 4c.42) ∆p =
dp
dL
dL ≈ ∑ dp
dL i
∆ Li
i =1
0

where: L = length, ft

A flow chart of the procedure for one length increment is

shown in Figure 4c.11. Note that after guessing the pressure
at the end of the length increment (i+1), a decision must be
made on how to treat the temperature. If a known
temperature distribution is to be used, the temperature can
be calculated directly at location i+1. If the temperature
profile is not known, the temperature at location i+1 must be
determined from either a numerical or approximate analytical
integration of an enthalpy gradient equation developed from
conservation of energy principles. The enthalpy gradient
equation and the approximate analytical solution are
presented in a later section.

The algorithm outlined in Figure 4c.11 is valid for any piping

system and phase behaviour model. Once an average
pressure and temperature is assumed for the increment,
mass transfer calculations can be performed with either the
black oil model or the compositional model. PVT properties
can then be determined and volumetric flow rates can be
calculated. Any of the available pressure gradient correlations
or mechanistic models can then be employed to determine
the average pressure gradient in the increment. After
successfully converging on the pressure, one ‘marches’ on to
the next increment until the end of the pipe is reached.

Page 37
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

START

GUESS p i +1

YES TEMP
PROFILE
KNOWN

NO

CALCULATE
Ti + 1 NO
ENTHALPY
BALANCE

YES
CALCULATE T i + 1
GUESS T i +1 BY COULTER &
BARDON EQUATION

CALCULATE
dq / dL & dh / dL
h i + 1 = h i ± dh
dL
. ∆L

FIND T i + 1 THAT
MATCHES h + FROM
h i + 1 = hL (1 - x) + hg x

NO CONVERGE
ON T i + 1

YES

CALCULATE dp / dL
p i + 1 = p i ± dp
dL
. ∆L

NO CONVERGE
ON p i + 1

YES

NEXT INCREMENT

Figure 4c.11 – Calculation of Pressure Traverse when Temperature

Distribution is Unknown

Page 38
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

If pipe diameter, inclination angle or flowrate (ie for gas lift

installations) changes at any location in the pipe, a new
increment must begin and appropriate changes made in the
data.

The BP MULTIFLO computer program has been written to

utilize a similar algorithm with virtually every production

Page 39
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

TEMPERATURE PREDICTION

situation encountered in oil and gas wells.

The steady state energy balance is illustrated by the control
volume in Figure 4c.12. The energy of a fluid entering the
control volume, plus any shaft work done on or by the fluid,
plus any heat added to or removed from the fluid must equal
the energy of the fluid leaving the control volume.

HEAT EXCHANGER
U'1 dE Q U'2
dt
p1V1 p2V2

mv12 mv22
2 gc Z2 2 gc
PUMP OR TURBINE
mgZ 1 mgZ 2
Z1
gc gc
REFERENCE PLANE
W'

Figure 4c.12 – Flow System Control Volume

Energy conservation per unit mass can thus be expressed as:

p1 2 g Z1 p 2 g Z2
(Equation 4c.43) U1 + + v1 + + Q' + W = U 2 + 2 + v2 +
J ρ 1 2g c J gc J J J ρ 2 2g c J gc J

where: U = internal energy, BTU/lbm

J = mechanical equivalent of heat,
778 ft lbf/BTU
Z = vertical elevation, ft
Q' = heat flowrate, BTU/hr
W = work, ft lbf

Expressing Equation 4c.43 in differential form gives

p gdZ
(Equation 4c.44) dU + 1 d + vdv + + dQ' + dW = 0
J ρ gc J gc J J

Page 40
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

This form of the energy balance equation is difficult to apply

because of the specific internal energy term. However, from
thermodynamics:

p
dU = dh – 1 d
(Equation 4c.45)
J ρ

Combining Equations 4c.44 and 4c.45, neglecting any shaft

work and solving for the specific enthalpy gradient gives:

(Equation 4c.46) dh = – dQ' – vdv – g sin θ

dL dL dL

Due to the strong dependence of dh and dQ on temperature,

Equation 4c.46 is used to determine temperature change
when fluids flow through pipes. Normally the kinetic energy
term is negligible. Therefore, for a horizontal pipe, an
increase in enthalpy of a fluid equals the heat transferred to
the fluid from the surroundings. Also, if no heat transfer
occurs, an increase in elevation causes a decrease in
enthalpy and a corresponding decrease in temperature.

WELLBORE
HEAT TRANSFER
When hot reservoir fluids enter a wellbore, they immediately
begin losing heat to the cooler surrounding rock as they flow
to the surface. The surrounding rock gradually heats up,
reducing the temperature difference and thus the heat
transfer between the fluids and the rock. Eventually, for a
constant mass flowrate, the entire earth surrounding the well
reaches a steady state temperature distribution. Prediction
of the fluid temperature in the wellbore as a function of depth
(and time) is necessary to determine fluid physical properties
required to calculate pressure gradients.

Due to the high thermal conductivity and relatively small

radial distance between the flowing fluids and the borehole
wall, heat transfer in this region can be considered steady
state. Thus, all heat lost by the fluids instantaneously flows
through the wellbore and into the surrounding rock.

A simple expression for the total heat loss from the fluids in
(*Reference 4c.21)
the tubing can be estimated from Newton’s Law of cooling*:

(Equation 4c.47) Q = 2π r t o ∆ L U ∆T

Page 41
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

where U is an overall heat transfer coefficient. Heat transfer

within the tubing and in a fluid filled annulus is primarily
due to convection. Heat transfer through the tubing and
casing walls and through a cement filled annulus between
the casing and borehole wall is primarily due to conduction.
Heat transfer into the surrounding rock is by heat
conduction, and is a transient process.

In differential form, Equation 4c.47 can be expressed in terms

of specific heat lost per length of pipe for use in Equation
4c.46:

dQ' U π dt o T f – Te
(Equation 4c.48) = w
dL

where: dto = tubing outside diameter, in

Tf = temperature of fluids in tubing, °F
Te = geothermal earth temperature, °F
TEMPERATURE
PREDICTION
Prediction of temperatures in wells requires application of
conservation of mass, momentum and energy principles.
This can be accomplished by coupling the pressure gradient
and enthalpy gradient equations given by Equations 4c.39
and 4c.46. The complexity of these equations prevents
rigorous analytical solution, and a numerical solution
procedure is described in Figure 4c.11. However, an
(*Reference 4c.22) approximate analytical solution given by Ramey* can be
modified for use in flowing wells. In addition, Shiu and
(*Reference 4c.23)
Beggs* developed an empirical simplification to the Ramey
method that is relatively easy to use, but potentially
inaccurate. Although numerous other publications have
appeared, primarily pertaining to steam injection or
producing wells, all are additional modifications to the Ramey
method or to calculating heat transfer coefficients. Following
is a summary of the modified Ramey solution.

Ramey presented an approximate solution to the wellbore

heat transmission problem pertaining to the injection of
single-phase hot fluids. Neglecting kinetic energy effects,
friction and assuming an incompressible fluid, Ramey solved
a combined pressure gradient and enthalpy gradient equation
to obtain an expression for fluid temperature as a function of
depth and time. When converted to the case of a directional
producing well, the resulting equation is:

Page 42
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

T L , t = T R – gG L sin θ + gG A sin θ 1 – e –L / A +

(Equation 4c.49) T bh – T R e –L /A

where: Tr = reservoir temperature, °F

Tbh = bottom hole temperature, °F

and where L is measured from well bottom and A, a term

called the relaxation distance, is defined as:

wc p 1
(Equation 4c.50) A = .
2 π r t oU

The last term in Equation 4c.49 accounts for possible changes

in temperature at the bottom of the well due to flow through
perforations or a gravel pack.

Ramey also presented an equation to be used for an ideal gas

being injected in vertical wells. When modified for a
directional well producing a real gas, and assuming a
constant pressure gradient, the resulting equation is:

g
T L , t = T R – gG L sin θ + A sin θ – gG e –L /A – 1 +
gcJ cp

dp –L /A
(Equation 4c.51) T bh – T R e –L /A – A η e – 1
dL
where h is the Joule-Thompson coefficient that describes
cooling (or heating) by expansion.

Calculation of flowing temperatures as a function of depth

and time can be very tedious due to the complexity of the
overall heat transfer coefficient in Equation 4c.50. Shiu and
Beggs proposed the use of an empirical correlation for A that
was developed from a broad set of flowing temperature
surveys. The resulting equation is independent of time:

(Equation 4c.52) A = 0.0149 w t 0.5253

d t i –0.2904 API 0.2608 γg 4.4146 ρ L 2.9303

Equation 4c.51 is recommended when calculating

temperature changes in wells flowing multiphase mixtures
and it is now available in MULTIFLO. Mixture properties
required for use in Equation 4c.51 were discussed earlier.

Page 43
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

MULTIFLO uses a trial and error procedure to predict

temperature changes in wells that is a modification of the
(*Reference 4c.24) Fontanilla and Aziz* method for wet steam injection wells.
The Fontanilla and Aziz method for injection wells represents
an improved Ramey method in that frictional losses and
kinetic energy effects can be included by using multiphase
flow correlations for pressure gradient. In addition, fluid
physical properties and the overall heat transfer coefficient
were permitted to vary with depth. The resulting method
should give results similar to Equation 4c.51.

Page 44
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

PRESSURE DROP ACROSS CHOKES

Restrictions to flow are normally installed downstream from

wellheads to provide a means of controlling flow rates from
wells. These chokes can have a variety of shapes. When fluids
flow through the chokes, velocities increase significantly and
a permanent unrecoverable pressure loss occurs. If velocities
in the choke reach sonic conditions, critical flow is said to
occur and flow behaviour becomes independent of
downstream conditions. Equations for critical and subcritical
flow are completely different, so it becomes necessary to
predict the boundary between the two conditions. Ashford
(*Reference 4c.25) and Pierce* developed the following equation for the
downstream to upstream pressure ratio, p2/p1, at the critical
flow boundary.

–k
2R 1 R1 1 – X cb – X c + 1 X –e
c –1
–1/k b
k 1 + R1 X c
(Equation 4c.53) Xc =
R1

where: b = k–1
k

e = k+1
k

R1 = vS g 1
vS L 1
Figure 4c.13 shows a plot of the solution to Equation 4c.53 for
different values of specific heat ratio, k. It is clear that, as
single phase gas flow is approached at high values of vSg, the
accepted value of 0.50 to 0.55 for Xc is predicted. However, as
the in situ gas/liquid ratio declines, much lower pressure
ratios (higher pressure drops) are required to reach critical
flow. The region above the curves represents subcritical flow,
and region below the curve is critical flow.

Page 45
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

0.7

k si 04 k=1.04
0.6 k si 20 k=1.20
SUB-CRITICAL

x c , CRITICAL PRESSURE RATIO p2 /p i

k si 40 k=1.40
FLOW
0.5

0.4 ASHFORD

CRITICAL FLOW
0.3

0.2

0.1

0.0
0.001 0.01 0.1 1 10 100 1000 10 000 100 000

R, GAS - LIQUID RATIO, m / m 3

3

Figure 4c.13 – Critical Flow Boundaries for Multiphase Flow

Through Chokes

(*Reference 4c.26) For critical multiphase flow, the empirical correlation of

Gilbert* is still considered adequate. Thus,

10.0 R p0.546 q Lsc

(Equation 4c.54) p1 = 1.89
d ch

where: Rp = producing gas-liquid ratio, scf/stb

dch = choke diameter, 64ths of in

For critical single phase gas flow through a choke, the

following equation can be used:

C p1
(Equation 4c.55) q gsc =
γg T 1

where: γg = gas specific gravity

where the coefficient c depends on base pressure and choke

size. For non-circular chokes, it is necessary to calculate the
equivalent circular choke diameter that would give the same
choke area. Table 4c.1 gives values of C for a standard

Page 46
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

pressure of 14.4 psia. These values must be multiplied by

1/␣ Z to correct for a non-ideal gas, where Z is determined at
upstream conditions.

CHOKE COEFFICIENTS FOR SINGLE PHASE GAS

Orifice Size (in) C

1/8 6.25
3/16 14.44
1/4 26.51
5/16 43.64
3/8 61.21
7/16 85.13
1/2 112.72
5/8 179.74
3/4260.99

Table 4c.1

Critical flow is difficult to obtain for single phase liquid

flow␣ because of the very high velocities needed to reach
sonic␣ conditions.

Subcritical gas, liquid or multiphase flow is much more

difficult to predict accurately than critical flow. Application of
the Bernoulli and continuity equations with a discharge
coefficient to account for all non-idealities is all that can be
justified. It is then necessary to obtain experimental or field
data to determine appropriate discharge coefficient values.
Thus:

ρ n1 vmB1 2
(Equation 4c.56) ∆p =
2 gc CD2

where: CD = discharge coefficient

Page 47
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

PRESSURE GRADIENT CURVES

Before the advent of powerful computer programs and

personal computers for predicting multiphase flow behaviour
in wells, engineers frequently used pressure gradient curves
to size tubing and design gas lift installations. A typical set of
gradient curves is shown in Figure 4c.14. Books of similar
curves have been developed to cover a broad range of pipe
diameters, liquid production rates and water fractions. A
graphical overlay procedure and interpolation between curves
is used to predict flow behaviour for a given well.

Generation of gradient curves requires making several

serious and limiting assumptions. Included are:

• Selection of one multiphase flow correlation that may not

be accurate for the field in question. Most available
gradient curves were generated with a modified Hagedorn
and Brown correlation.

• Selection of one inclination angle. All available gradient

curves were generated for vertical wells and will not be
valid for directional wells.

• Selection of one API gravity, one gas specific gravity and

one water specific gravity.

• Selection of specific empirical correlations to predict mass

transfer between phases and fluid physical properties.

• Selection of a temperature distribution that is

independent of flowrate.

These assumptions are so restrictive that the use of gradient

curves is no longer recommended. Programs such as
MULTIFLO have removed all of the limiting assumptions
inherent in gradient curves.

Page 48
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

PRESSURE x 100 psi

3500 psi
160 PSI 8 16 24 32 40 48 56

1840 ft VERTICAL FLOWING PRESSURE GRADIENTS

(OIL PERCENT = 50)
2 TUBING SIZE = 2.441in ID
PRODUCTION RATE = 1000 STB/D
GAS SPECIFIC GRAVITY = 0.65
AVERAGE FLOWING TEMPERATURE = 150°F
OIL API GRAVITY = 35.0 API
WATER SPECIFIC GRAVITY = 1.07
4

6
DEPTH x 1000 ft
12 000 ft

10

12

14 50

100
GAS/LIQUID RATIO, SCF/STB
16
200

18 300

400

20
30
20 0
15
12

10

80

60

50
0
00
00
00

00

GIVEN DATA: SOLUTION:

2 7/8in OD TUBING (2.441in ID) 1. Find the equivalent depth corresponding to 160 psi wellhead pressure. To do
LIQUID FLOW RATE = 1.000 STB/D (50% WATER) this, proceed vertically downward from 160 psi at zero depth until intersecting the
DEPTH = 12 000 ft 400-scf/bbl GLR line. This is at a depth of 1840 ft. Note the pressure scale is
PRODUCING GOR = 800 SCF/STB is in 800 psi increments and the depth scale is in 2000 ft increments.
PRODUCING GLR = 800/2 = 400 SCF/STB
WELLHEAD PRESSURE = 160 PSI 2. Add the equivalent depth of 1840 ft to the well depth of 12 000 ft and obtain 13 840 ft.

3. From 13 840 ft on the vertical scale, proceed horizontally to the 400 scf/stb line
FIND THE FLOWING BOTTOM-HOLE PRESSURE. and read a flowing pressure of 3500 psi

Figure 4c.14 – Typical Pressure Gradient Curves

Page 49
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

REFERENCES

4c.1 Woelflin, W: ‘The Viscosity of Crude-Oil Emulsions’,

API Drill and Prod Pract (1942), 148

4c.2 Knudsen, J G and Katz, D L: Fluid Dynamics and Heat

Transfer, McGraw Hill Book Co, New York (1958)

4c.3 Poettmann, F H and Carpenter, P G: ‘The Multiphase

Flow of Gas, Oil and Water Through Vertical Flow
Strings with Application to the Design of Gas-Lift
Installations’, API Drill and Prod Pract (1952),
257–317

4c.4 Baxendell, P B and Thomas, R: ‘The Calculation of

Pressure Gradients in High-Rate Flowing Wells’,
SPE␣ J␣ Pet Tech (October 1961), 1023–1028

4c.5 Fancher, G H Jr and Brown, K E: ‘Prediction of

Pressure Gradients for Multiphase Flow in Tubing’,
Soc Pet Eng J (March 1963), 59–69

4c.6 Griffith, P and Wallis, G B: ‘Two-Phase Slug Flow’,

ASME J Heat Transfer (August 1961), 307–320

4c.7 Griffith, P: ‘Two-Phase Flow in Pipes’, Special Summer

Program, Mass Inst Tech, Cambridge, Mass (1962)

4c.8 Brill, J P: ‘Discontinuities in the Orkiszewski

Correlation for Predicting Two-Phase Pressure
Gradients in Wells’, accepted for ASME J Energy
Res␣ Tech (1989)

4c.9 Mukherjee, H and Brill, J P: ‘Pressure Drop

Correlations for Inclined Two-Phase Flow’, ASME J
Energy Res Tech (Dec 1985)

4c.10 Fernandes, R C, Semiat, T and Dukler, A E:

‘Hydrodynamic Model for Gas-Liquid Slug Flow in
Vertical Tubes’, AIChE J (Nov 1986), 29, 981–989

4c.11 Orell, A and Rembrand, R: ‘A Model for Gas-Liquid

Slug Flow in a Vertical Tube’, Ind Eng Chem Fund
(1986), 25, 196–206

Page 50
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4c – MULTIPHASE FLOW

4c.12 Oliemans, R V A, Pots, B F M and Trompe, N:

‘Modeling of Annular Dispersed Two-Phase Flow in
Vertical Pipes’, Int J Multiphase Flow (Oct 1986), 12,
711–732

4c.13 Yao, S C and Sylvester, N D: ‘A Mechanistic Model for

Two-Phase Annular-Mist Flow in Vertical Pipes’,
AIChE J (June 1987), 33, 1008–1012

4c.14 Sylvester, N D: ‘A Mechanistic Model for Two-Phase

Vertical Slug Flow in Pipes’, ASME J Energy Res Tech
(Dec 1987)

4c.15 Ansari, A, M and Sylvester, N D: ‘A Mechanistic Model

for Two-Phase Bubble Flow in Vertical Pipes’, AIChE J
(August 1988), 34, 1392–1394

4c.16 Alves, I N, Caetano, E F, Minami, K and Shoham, O:

‘Modelling Annular Flow Behavior for Gas Wells’,
Presented at ASME Annual meeting, Chicago, Ill
(Nov␣ 1988)

4c.17 Caetano, E F: ‘Upward Vertical Two-Phase Flow

Through an Annulus’, PhD Dissertation, The
University of Tulsa (1985)

4c.18 Ozon, P M, Ferschneider, G and Chwetzoff, A: ‘A New

Multiphase Flow Model Predicts Pressure and
Temperature Profiles in Wells’, SPE 16525/1,
Presented at Offshore Europe 87, Aberdeen (1987)

4c.19 Hasan, A R and Kabir, C S: ‘A Study of Multiphase

Flow Behavior in Vertical Wells’, SPE J Prod Eng
(May␣ 1988), 263–272 (a correction to this paper was
published in the following issue (June 1989))

4c.20 Ansari, A M: ‘A Comprehensive Mechanistic Model for

Two-Phase Flow in Vertical Pipes’, MS Thesis, The
University of Tulsa (1988)

4c.21 Bird, R B, Stewart, W E and Lightfoot, E N: Transport

Phenomena, John Wiley and Sons, New York (1960)

4c.22 Ramey, H J: ‘Wellbore Heat Transmission’, J Pet Tech

(April 1962)

Page 51
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4c – MULTIPHASE FLOW Completion Design Manual

4c.23 Shiu, K C and Beggs, H D: ‘Predicting Temperatures

in Flowing Wells’, ASME J Energy Res Tech
(March 1980)

4c.24 Fontanilla, J P and Aziz, K: ‘Prediction of Bottom-hole

Conditions for Wet Steam Injection Wells’, J Cdn Pet
Tech (March-April 1982)

4c.25 Ashford, F E and Pierce, P E: ‘The Determination of

Multiphase Pressure Drops and Flow Capacities in
Down-Hole Safety Valves (Storm Chokes)’, SPE 5161,
Presented at SPE Annual Fall Meeting, Houston, TX
(October 1974)

4c.26 Gilbert, W E: ‘Flowing and Gas-Lift Well Performance’,

API Dr and Prod Prac (1954), pg 126

Page 52
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

CONTENTS

INTRODUCTION 1

GAS LIFT 8
GAS LIFT PROCESS AND APPLICATIONS 8
IMPLICATIONS FOR COMPLETION DESIGN 12
COMMON PROBLEMS 13

ELECTRICAL SUBMERSIBLE PUMPS (ESP) 14

ESP EQUIPMENT AND APPLICATIONS 14
PUMP AND SYSTEM PERFORMANCE 18
IMPLICATIONS FOR COMPLETION DESIGN 22
COMMON PROBLEMS 23

HYDRAULIC PUMPING SYSTEMS 25

BASIC CONCEPTS AND APPLICATIONS 25
IMPLICATIONS FOR COMPLETION DESIGN 30

ROD PUMPING SYSTEMS 31

BASIC EQUIPMENT AND APPLICATIONS 31
IMPLICATIONS FOR COMPLETION DESIGN 37

SCREW PUMP SYSTEMS 38

PLUNGER LIFT 40

REFERENCES 42
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

INTRODUCTION

As discussed in the Theory section of Chapter 4, the

application of artificial lift will simply displace the tubing
performance curve (TPC) downwards, so that a lower bottom
hole flowing pressure (BHFP), pwf, can be achieved at any
given rate (see Figure 4d.1).

Figure 4d.1 – Effect of Artificial Lift on TPC

Any artificial lift system can therefore be viewed simply as a

point of energy injection into the system. Upstream and
downstream of the artificial lift system, well behaviour is
governed by the laws of fluid flow discussed in Chapter 4
and␣ Sections 4a, 4b and 4c.

Energy can also be added into the system by reservoir

pressure maintenance. Reservoir development optimization
studies are therefore used to determine the relative technical
and economic benefits of these options as well as the
optimum timing of the associated investments. In many
fields, both pressure maintenance and artificial lift are used,
often allowing artificial lift installation to be deferred (see
Figure 4d.2). In poorer zones, however, artificial lift may be
required immediately to achieve the production and economic

Page 1
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.2 – Production System Pressures on a Waterflood under Artificial Lift

targets by rate acceleration, whether pressure maintenance is

installed or not. Thus, just as tubing size is critical in the
high productivity index (PI) wells (>5 stb/d/psi), so
minimization of the BHFP is the design objective in low PI,
low pressure wells.

Artificial lift may therefore be installed for a number

of␣ reasons:

• To offset the effects of declining reservoir pressures.

• To offset the effects of increasing water production in both

oil and gas wells.

• To overcome the excessive friction pressures associated

with the production of viscous or waxy crude, or with the
attainment of high offtake rates.

Page 2
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.3 – Application of Artificial Lift Methods based on

Well Performance

• To kick-off high gas-liquid ratio (GLR) wells that tend to

die when shut-in or during heading cycles.

• To accelerate production by reducing the back pressure

against which the reservoir must produce.

• To continuously remove wax or scale build-up from the

tubing walls. Although we normally associate artificial
lift␣ primarily with oil wells, it may also be used for
dewatering gas wells, for the production of water supply
wells, and for powered dump-flood operations.

Selection of the appropriate artificial lift method is based on a

number of factors, but particularly well performance
characteristics (see Figure 4d.3). The main artificial lift
systems and their applications are listed in Table 4d.1.

Page 3
 SECTION 4d – ARTIFICIAL LIFT
Determining the Optimum Well Performance – Chapter 4
MAIN ARTIFICIAL LIFT SYSTEMS

Lift System Common Applications Normal Depth Rates Min BHFP Problems
Range (ft) (stb/d)

Continuous gas lift Offshore wells 12 000 10–50 000+ (0.05)–0.15 Gas supply
Deviated wells psi/ft Hydrates
Deep wells Single wells
High GLR wells Casing design
Moderately productive wells Depletion

Electrical submersible High rate, low GLR wells 12 000 500–20 000 100 psi Casing size
pumps (ESPs) Water supply wells (5000) (500–10 000) Handling
High WOR wells BHT >300°F
Alternative to high rate Sand
Gas lift High GLR
Table 4d.1

Hydraulic jet pump Mod-low GLR wells 12 000 100–8000+ 20% net lift Casing size
Page 4

Mod-good PI (5000) (25 000) High THIP

Deviated wells Fluid volume
Alternative to gas lift Triplex pumps

Hydraulic piston pump Deep wells 12 000 10–2000 Wellhead Cleanliness

Low GLR wells (18 000) (<1000) pressure Casing size
Deviated wells ((<8500)) ((2000–8000)) High THIP
Alternative to rod pumps Triplex pumps

Rod pumps Onshore wells 8000 2–1500 Wellhead Doglegs

Wells with low CIBHP (12 000) (<500) pressure Fatigue
Low PI wells (<0.1 stb/d/psi) High GLR

Completion Design Manual

Sand

Screw pumps Heavy oil wells (<18° API) 5000 1–500 Wellhead BHT >250°F
Shallow/light wells (1–1200) pressure High GLR

BP Exploration
Sandy wells (1–400)

Plunger lift Dewatering gas wells 12 000 1–500 100-500 psi Low GLR
High GLR/low rate oil wells Solids
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

(*Reference 4d.1) According to Brown*, the percentage use of these lift methods
in the USA on wells producing more than 10 stb/d is:

• Gas lift 53%

• Rod pumping 27%
• Electrical submersible pumps (ESP) 10%
• Hydraulics 10%

Selection of the optimum artificial lift method for a particular

well or field is not only dependent upon the capital costs and
production capabilities but also on the operating costs and
the well servicing frequency and costs (refer to Table 4d.2).
Estimation of the run life is the most difficult aspect of such
evaluations since it is usually a function of the operating
environment. This is particularly true for ESPs, where run
times vary from less than two months to five years,
decreasing with temperature, restart frequency, sand
production, GLR, and lack of experience.

Similarly, some systems have a better ability to cope with

production problems than others and this may also influence
the choice of lift system, or necessitate the selection of special
materials or ancillary equipment.

Consideration of artificial lift requirements during the

well␣ planning stage is required not only where it will be
immediately needed, but also to enhance efficiency of future
installations. Often relatively minor modifications to the well
or completion design can greatly enhance the long-term
production capacity and significantly reduce future costs.
Such considerations include:

• Casing size required to accommodate ESPs, hydraulic

power fluid supply conduits, downhole gas separation and
venting, high rate gas supply and the subsurface safety
valves (SSVs) for dual conduit systems.

• Casing coupling required to provide a gas-tight seal for

the use of high pressure gas lift (premium casing
connections should be employed).

• Size and positioning of liners and packers with respect to

the setting depth of the lift equipment.

• The provision of a sump to permit rod pumps to be set

below the pay zone.

Page 5
 SECTION 4d – ARTIFICIAL LIFT
Determining the Optimum Well Performance – Chapter 4
TYPICAL ARTIFICIAL LIFT PERFORMANCE SUMMARY

Service
Frequency Service
(yrs/job) Cost Rating of Ability to Handle Problems

High High
>250°F GLR Wax Sand Corrosion Scale Visc Doglegs

Rod Pumps 1 L–M 5 3 3 3 3 4 3 2

Table 4d.2

Gas Lift 1.5 L 4 5 2 4 4 3 2 5

Page 6

ESP
– <150°F 2
– 150-250°F 1 H 2 3 4 2 2 3 2 3
– >250°F 0.5

Hydraulic
– Jet 1.5 L 4 4 5 4 5 5 3 4
– Piston 0.5 M 4 2 4 2 4 2 5 4

Plunger 1 VL 4 5 4 1 2 1 1 5

Completion Design Manual

Screw Pump 1 M 1 2 3 5 3 3 5 3

5 – Excellent H – High L – Low

BP Exploration
1 – Poor M – Medium VL – Very Low
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

• The use of a tubing anchoring system and pump seating

nipple on wells that will be converted to rod pump without
a tubing pulling job.

• The pre-installation of properly spaced gas lift mandrels

and annulus SSVs, where required.

• The pre-installation of multiple conduits with downhole

sleeves or pump seating nipples in wells that will be
converted to hydraulic pump.

• The use of a uniform bore on wells that may have plunger

lift installed.

• The possibility of accelerating production by early

installation of the lift equipment.

In the remainder of the chapter, the basic principles and

applications for each of the lift systems will be discussed.

Page 7
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

GAS LIFT

GAS LIFT
PROCESS AND
APPLICATIONS
Continuous gas lift simply supplements the natural flow
process by adding additional gas into the produced fluid to
reduce the hydrostatic head component and hence, the back
pressure on the formation.

The injection gas is normally supplied in a closed loop

system, in which it is collected from the separators,
compressed, dried (if necessary), and supplied to the well (see
Figure 4d.4).

Figure 4d.4 – Gas Lift Systems

Page 8
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

The lift gas is normally injected down the annulus and into
the tubing through a gas lift valve (GLV), which contains a
non-return check. In some wells, the gas is supplied via the
tubing with production being taken either up the annulus, or
through a second tubing string (concentric or parallel to the
supply string).

The other less common concept, also shown in Figure 4d.4, is

the intermittent gas lift system. This is used to produce low
volumes of liquid (<350 stb/d) from wells with low BHFP
(<0.1 psi/ft). In this case, the liquid is produced in slugs by
periodically opening a motorized flowline valve and allowing
the gas accumulated in the annulus to suddenly expand
through the GLV to surface, bringing the accumulated liquids
with it. A standing valve is used to prevent the gas from
flowing into the formation. Since liquid fallback is an obvious
limitation, this technique is often combined with the plunger
lift method discussed at the end of the chapter.

For continuous or intermittent lift, it is advantageous to

locate the gas injection point as deep as possible. The depth
of the operating GLV (OGLV) is limited by:

• The gas supply pressure.

• The flowing pressure in the tubing at the intended

offtake␣ rate.

• The depth of the packer and deepest gas lift mandrel.

• The differential across the valves required to keep the

upper valves closed (20 psi/valve) and to ensure that
injection at the OGLV is stable (50 to 500 psi).

Figure 4d.5 illustrates the fundamentals of gas lift design and

operation. The figure shows pressure on the horizontal axis,
and true vertical depth on the vertical axis.

The available kick-off pressure gradient is first plotted on the

graph. This kick-off pressure is usually higher than the
normal operating pressure of the gas lift system. Next, a kill
fluid gradient is plotted starting at the flowing tubing head
pressure of the production system. The intersection of the kill
gradient with the kick-off pressure gradient determines the
location of the first unloading GLV.

Page 9
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.5 – Example Gas Lift Design

As shown in the figure, gas passes through the first GLV and
lightens the fluid column to surface, according to the total
gas liquid ratio (TGLR) curve shown. As the fluid gradient in
the tubing changes, the gas in the casing moves down to the
second GLV, unloading kill fluid from casing. As the gas
reaches the second GLV and begins to lighten the fluid
column, the first GLV closes, so that all gas is passing
through the second GLV.

The process continues until the kill fluids in the casing

annulus have been displaced, and the gas is passing through
the OGLV. Once the well has been unloaded, the operating
pressure on the casing will drop below the initial kick-off
pressure.

At some time during the unloading procedure, the reservoir

will have begun to produce fluids. The production rate from

Page 10
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

the well is a function of a number of parameters, including

reservoir pressure, PI, water cut and gas injection rate. After
the well is producing at a stabilized rate, production should
be optimized by varying the gas injection rate.

In performing a system performance analysis, the Production

Engineer must determine:

1) The optimum injection depth.

2) The available lift gas supply rate.

3) The resultant production rate and hence the effective

injection GLR (IGLR). This is obviously an iterative
process.

As discussed in the Theory section of Chapter 4, increasing

the GLR initially decreases the bottom hole pressure on the
(*Reference 4d.2) TPC. Gilbert* showed that for any constant rate of flow and
tubing size, there is an optimum GLR (OGLR) which provides
the minimum BHFP.

Since the GLR requirements are very much subject to the law
of diminishing returns, most gas lift systems are designed
either on the basis of a near optimum GLR (NOGLR) that
achieves a BHFP within 20 to 50 psi of the minimum, or on
the basis of available injection gas supply volumes, Qi, (see
Figure 4d.6). It should be apparent that:

1) Liquid rate (q) depends on the IPR and attainable BHFP.

2) Total GLR = Producing GLR + Injection GLR ≤

Optimum␣ GLR

3) IGLR = Qi/q

To evaluate the performance of a gas-lifted well:

1) Estimate the gas injection depth using the flowing

gradients and the available lift gas pressure, if defined.
(Note: Care should be taken in selecting the appropriate
pressure drop prediction method.)

2) Generate a series of TPC at varying GLRs, and water

cuts,␣ if appropriate, using the gas injection point defined
in Step 1.

Page 11
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

3) Overlay the IPR and TPC curves to define the production

rate and hence the required lift gas supply rates and
pressures.

Figure 4d.6 – Effect of Total GLR on Well Outflow Performance

IMPLICATIONS
FOR
COMPLETION
DESIGN
Higher lift gas supply pressures are now commonly used to
reduce or eliminate the requirement for unloading valves to
kick-off wells. However, this greatly increases the risk of gas
leakage through the tubing and casing couplings. Couplings
with improved gas sealing capacity are therefore generally
preferred. Because of the large spiral void area, buttress
threaded couplings (BTC) commonly give problems in high
pressure gas lift operations and should be avoided.

Page 12
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Wireline retrievable GLVs set in side pocket mandrels (SPM)

are now generally used to reduce well servicing costs and
risks (refer to Chapter 8 – Selection of Completion
Equipment, for discussion of SPMs). Since these have a large
OD, there are sometimes clearance problems in small casings
and liners.

The valve spacing is designed based on a true vertical

projection of the well. SPM setting depths must be properly
converted to measured depth. Note that the deeper valves
may be as close as six joints apart, so that spacing in this
region is particularly critical.

On wells equipped with an SSV, some annulus blowout

protection is provided by the check valve in the GLV.
However, in critical locations like offshore this barrier is not
sufficiently reliable. In these locations an injection valve or
tubing retrievable SSV (TRSSV) is run below the annulus
access on the tubing hanger. Alternatively a sub-mudline
tubing hanger/packer with either an integral annulus safety
valve, or with a separate gas supply string and conventional
SSV, may be used . All of these options significantly effect the
casing, tubing and wellhead design, so that the probability of
needing gas lift must be identified early in the well planning
process.

COMMON
PROBLEMS
The most common problem with gas lift design is spacing the
valves too far apart because of underestimating the pressure
losses in the gas supply system and the back-pressure
caused by slugging. In operation, the most common problem
is gas leakage through the upper valves, and/or through
tubing or casing couplings.

Page 13
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

ELECTRICAL SUBMERSIBLE PUMPS (ESP)

ESP EQUIPMENT
AND
APPLICATIONS
ESPs have their greatest application in moving large volumes
(>500 stb/d) of low GLR fluid (<100 scf/stb). They are
particularly attractive for water supply wells, high water cut
producers and high deliverability undersaturated oil wells.
However, units are available at pumping rates as low as
90␣ stb/d. Where pump intake pressure is below the bubble
point, modified design procedures and improved downhole
gas separators/compressors allow effective operation at up to
1000 scf/stb.

The units are best suited to maintaining a more or less

steady offtake rate within ±25% of the optimum rate. Variable
frequency drives (VFD) and/or wellhead chokes can be used
to increase the range of pumping rates (50 to 190% of
optimum), but increase capital and operating costs
considerably.

Figure 4d.7 – ESP Pump Stage

Page 14
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

A submersible pump is a multi-staged centrifugal pump

connected by a short shaft to a downhole electric motor. Each
stage consists of a rotating impeller and stationary diffuser
(see Figure 4d.7).

Energy transfer is accomplished by the rotating impeller,

imparting a tangential and radial motion to the fluid (kinetic
energy). The diffuser then converts the high kinetic energy
into lower velocity, pressure, or potential energy. Each stage
consisting of an impeller and diffuser is about 3 to 5in long.
There may be 50 to 500 stages in a pump.

The ESP is a dynamic displacement pump in which the

differential pressure or total dynamic head (TDH) developed
by the pump is a function of pump flowrate. TDH is
approximately the sum of the head developed by each stage,
which can be obtained from the manufacturers published
test data (see Figure 4d.8).

(Equation 4d.1) TDH = Ns Hs

where: Ns = number of stages

Hs = head per stage

The standard test is made using fresh water (SG = 1.0,

µ = 1.0 cp) and reported in terms of pump head per stage
(Hs). Corrections can be applied for the average produced
fluid gravity (SG) to determine the actual pump differential
pressure (∆pp) developed at a given pump flowrate.

(Equation 4d.2) ∆pp = 0.433 (SG) (TDH)

(Equation 4d.3)
∆pp = 0.433 (SG) Ns Hs

The pump characteristics are based on a constant rotation

speed, which depends on the frequency of the AC supply:

• 3500 RPM with 60 Hertz

• 2915 RPM with 50 Hertz

Because of these high rotation speeds, it is obviously very

important to evaluate and control sand production and the
formation of emulsions, if ESP installations are to be
successful.

Page 15
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.8 – Typical Design Curve for a Submersible Pump Stage

It can be seen from the above that the ESP delivery capacity
will vary with:

• The well IPR characteristics.

• The reservoir pressure.

• The surface back-pressure.

• The electrical supply frequency.

The typical electrical submersible pumping unit assembly

consists of a three-phase electrical motor, seal section,
possibly a rotary or reverse flow gas separator, multi-staged
centrifugal pump, motor cable, power cable, drain valve and

Page 16
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

check valve. Surface equipment includes a three-phase

transformer, a motor controller, and a wellhead pack-off for
the power cable.

Figure 4d.9 – Typical ESP Installations

Whenever possible, the installation is designed to facilitate

downhole separation of any free gas, which is vented up the
annulus. This is particularly important where the downhole
gas volume exceeds 10% of the total fluid volume, unless
oversized initial stages are used.

Most submersible pump installations are simply run on the

tubing and landed-off above the perforations or open hole
section. The motor is run at the bottom so that the produced
fluid will dissipate the heat generated. If the pump has to be
set below the perforations or in relatively large casing, a
shroud (or flow tube) is used to draw the produced fluid past

Page 17
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

the motor at sufficient velocity (>1 ft/sec) to provide cooling.

Bottom discharge pumps are also available for powered dump
floods.

In some locations a safety valve operated by the pump

discharge pressure can be mounted below the pump to
provide a positive shut-off when the pump is pulled and to
meet SSV regulations. A cable suspended pump was
developed to reduce pulling costs in offshore applications.
This pump is locked into the tubing shoe with dogs and a
sealing element supports the fluid head. The main problem
with the cable suspended pumps has been failures and earth
leakages in the cable, pot-head and wellhead splice. Gas
penetration and armour corrosion problems are the main
cause, since the cable is exposed to direct contact with the
produced fluids at high temperatures. This system is not
therefore commonly used.

Submersible pumps are available from a number of

(*References 4d.3-4d.5) manufacturers (TRW-Reda*, Centrilift-Hughes*, Westech-
Dresser*, etc) all of whom provide design services,
engineering manuals, and customer support. It can be
advantageous to select the supplier at an early stage and
then to work with that supplier on finalizing the design and
equipment selection. It is also highly desirable to include
electrical engineers in the design and operations groups, as
this can lead to considerable improvements in performance
and permits a total systems approach to power generation,
distribution and utilization.

PUMP AND
SYSTEM
PERFORMANCE
It is usual to select the largest pump that will fit into the
production casing or to size the production casing to
accommodate the pump. Small casing or liners will limit the
maximum feasible horsepower and pump capacity (see
Table␣ 4d.3 ).

Page 18
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

ESP CAPACITY RANGES

Casing Pump Motor Typical Max TDH

Size OD OD Rate Power Capacity
(in) (in) (in) (stb/d) (HP) (ft)

4.5 3.38 3.75 100–1900 50–125 5000–12 000

5.5 4.0 4.50 200–5000 100–300 5000–12 000
7.0 5.62 5.43 1000–16 000 200–650 5000–12 000
8.625 6.75 7.38 4000–26 000 400–850 3000–10 000
10.75 8.62 N/A 12 000–33 000 500–1020 2000–5000
13.375 11.25 N/A 24 000–100 000 500–1030 2000–3500

Table 4d.3

In performing a system analysis for an ESP installation, it is

important to consider the effects of GLR and viscosity. If
downhole gas separation is to be used, the TPC under
pumping may be considerably different from that for the
flowing well. Similarly, downhole viscosities above 30 cp may
not only effect the TPC, but also reduce the pump efficiency,
rate, and to a lesser extent, head capacity.

Two approaches are commonly used in evaluating an

ESP␣ system:

1) To preselect the production target and corresponding

BHFP and determine the TDH and pump size/depth
required to meet this rate. This is often carried out by
plotting the pressure traverses above and below the pump
(see Figure 4d.10).

2) To preselect the maximum pump horsepower, or number

of stages, and determine the attainable pump rate with:

• A fixed IPR and various tubing sizes.

• A fixed tubing size and various IPR options.

In this case, the pump performance curve is often plotted

below the system performance curves. An example of the
use of this approach to optimize the number of stages for
a maximum pump horsepower is shown in Figure 4d.11.

Page 19
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.10 – ESP Design for a Preselected Rate

Page 20
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.11 – ESP Design for a Preselected HP

Page 21
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

In the above example, it can be seen that using the 4.5in

tubing has the potential for increasing rate by some
300 stb/d (4%). The critical issue, however, is the attainment
of a near optimum IPR. A skin of only +5 would reduce the
potential by some 1400 stb/d (20%), despite the fact that the
well still cannot be pumped-off without the use of a much
larger tandem pump and motor.

The operating life of the pump is severely affected by up-

thrust on the bearings if the pump head drops below the
value corresponding to the maximum recommended pump
rate. Some installations are therefore equipped with a
variable choke at surface to apply back-pressure in the event
that the reservoir pressure or performance increases with
time, as in some waterflood operations. Alternatively, a
variable speed drive (VSD) can be used to adjust the pump
speed, without exceeding the thrust limits.

A minimum BHFP of 100 to 200 psi is normally

recommended to prevent cavitation or gas locking. Where an
oversized unit tends to pump-off the well, because of
declining pressure or IPR, a time clock can be used to pump
intermittently. However, a large number of restarts (50 to
200) significantly reduces run life.
IMPLICATIONS
FOR
COMPLETION
DESIGN
Effective heat removal and design of the insulation materials
for the actual operating temperatures and environment are
often cited as the key to efficient submersible pumping
operations. This is particularly critical where temperatures
exceed 250°F. The clearance past the pump should be small
enough that the fluids are flowing by at more than 1 ft/sec.
In larger casings, a shroud is used to achieve these velocities.
The pump and motor must also be properly centralized for
this purpose.

The power supply cable has low inherent strength and must
be properly supported by the tubing. Two cable clamps per
joint are recommended. The material used for the clamps
must be compatible with the operating environment and
armour material. Centralizers or protective cable clamps
should be used to provide crush protection through doglegs
and build-up sections.

Page 22
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

The pump should be set in a reasonably straight section

of␣ hole to minimize bending stresses on the shaft. A slightly
different position should be used on each run to minimize
the␣ casing corrosion caused by the eddy currents around
the␣ motor.

The casing or liner size and well deviation profile will

obviously limit the size of unit that can be considered.
Conversely, the future use of ESPs favour the selection of
large size production casing during the planning phase.
Similarly, the clearance required for the cable and protectors
may limit the tubing size that can be used.

Careful attention must be paid to the need for downhole gas

separation, especially in situations where a downhole SSV is
also required (refer to Figure 4d.9).

Many companies are now making the downhole cable splices

in a workshop, where they can be tested and radiographed.

Modified tubing hangers, packers and wellheads are

obviously required to provide a conduit for the cable and/or
a␣ pre-installed electrical connector.

COMMON
PROBLEMS
The main problem with ESP operations is short run times
and consequently high service costs and production losses.
These are commonly caused by:

• Poor handling procedures.

• Inadequate system analysis and design leading to the unit

operating outside the recommended range.

• Inadequate cable insulation for the operating

environment.

• Pumping significant quantities (>10%) of free gas without

enlarged intake stages.

• Sand production or corrosion.

• Frequent restarts without a soft start controller or VSD.

• Scale formation on the pump impellers.

Page 23
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

• Omission of the viscosity corrections in designing a

system to handle emulsions, waxy crude, or heavy oil.

• Unstable supply voltage.

• Poor quality control.

Page 24
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

HYDRAULIC PUMPING SYSTEMS

BASIC
CONCEPTS AND
APPLICATIONS
Since it is not particularly sensitive to temperature, depth,
deviation or severe operating environments, hydraulic
transmission has obvious attractions as an alternative to
electrical or mechanical means of transferring energy
downhole. The major problems are, however, the need for at
least two reasonably large conduits to minimize fluid friction
losses, the difficulty of maintaining a clean, solids-free power
fluid, and the high capital and maintenance costs for high
pressure pump units.

Hydraulic pumping has won the greatest acceptance as an

alternative to gas lift in areas with limited gas availability. It
is a viable alternative to rod pumping for producing deep
(>8000 ft) or deviated wells.

The two main systems, the jet pump and piston pump, are
interchangeable in most installations, providing a high degree
of flexibility to cope with changing well conditions. Moreover,
in many cases, the bottom hole pump assembly can be
circulated in or out of the well using a set of swab cups as
a␣ pump-down motor. Alternatively, the units can be run and
pulled with wireline. This permits low downhole service costs
and minimal downtime for pump servicing or replacement.

The most commonly used method to obtain two conduits is

to␣ take production up the annulus between the tubing and
casing. However, where this is not feasible because of
corrosion or safety concerns, a dual string completion is
used. For this, the two tubing strings may be installed either
concentrically or in parallel. Examples are shown in
Figure␣ 4d.12.

Page 25
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.12 – Downhole Schematics for Hydraulic Pumping

Page 26
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

The annulus may also be used for gas venting, especially for
the piston pumps, which are sensitive to gas locking.
Occasionally, in low rate installations, a third conduit is
installed to permit the use of a closed power fluid system.

If an SSV is required, it can either be installed conventionally

or placed below the pump and operated by the power fluid
supply pressure.
Piston Pump
The hydraulic piston pump is a closely coupled reciprocating
engine and pump, which performs rather like a steam engine.
Single and double acting pumps are available. Internal check
and shuttle valves control the reversals and exhaust the
spent power fluid.

Essentially, piston pumping is the competitor to rod

pumping, and is capable of pumping a well off at rates of up
to 8000 stb/d in large tubing, but is normally used to
produce <2000 stb/d. Piston pumps are commonly selected
for deviated or deep wells (8000 to 18 000 ft), although very
high surface pressures are needed below 12 000 ft.

The piston pump is a positive displacement pump for which

the net lift (H) or total dynamic head is constant for a given
pump/engine (P/E) diameter ratio. The net lift, maximum
rated speed (40 to 120 SPM depending on size) and
displacement are obtained from the manufacturer’s
catalogue. Pump speed can be adjusted by the rate at which
the power fluid is supplied. The stroke length is normally
relatively short (12 to 24in), although long stroke units (46 to
96in) are available.
Jet Pump
The jet pump employs no moving parts but imparts
momentum onto the produced fluid by a venturi system with
carefully sized jet, throat and diffuser. By varying the sizes of
these components, the system can handle volumes from 100
to 15 000 stb/d, although the free pump systems are limited
to less than 8000 stb/d with 4 1/2in tubing. To avoid
cavitation, the pump must be submerged by at least 20% of
the TDH. It is therefore best suited to reasonably productive
wells or wells with restricted offtake targets.

The jet pump can withstand a much poorer quality of both

power and produced fluids; it also has the ability to handle
fairly high gas-oil ratios (GOR) (3000 scf/stb). However, the

Page 27
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

pump has a fairly low efficiency (33 to 66%), so that large

volumes of fluid have to be handled in high rate installations.

Pump performance is a complex function of GOR, pump

intake pressure, supply pressure and supply rate. System
design and selection of the optimum pump for the application
is normally carried out using the manufacturer’s computer
program. However, the technical support group in the
northern North Sea also have a program for jet pump sizing.
At the maximum surface supply pressure (usually 5000 psi),
production rate decreases as the TDH or net lift increases,
and the power fluid requirements increase. The jet pump is
therefore a dynamic displacement pump, like the ESP.

Preliminary calculation of the pump intake or output curve

(*Reference 4d.1) can be made by hand. Brown* presents a curve of the head
ratio across the jet pump versus the flow ratio. The maximum
attainable performance is summarized in Table 4d.4.

MAXIMUM PERFORMANCE FROM A JET PUMP

Head Ratio Flow Ratio

Pump output pressure –

Pump intake pressure Reservoir production rate
Downhole power fluid pressure – Power fluid rate
Pump output pressure

0.45 0.5
0.25 1.0
0.17 1.5
0.10 2.0

Table 4d.4

The maximum desirable power fluid supply pressure and rate

can often be fixed from surface equipment considerations
(commonly p <5000 psi, qPF <4500 stb/d are used).

The corresponding bottom hole pressures will depend upon

the completion configuration and can be computed with
MULTIFLO. In performing these calculations, it is important
to add the power fluid rate to the production to obtain the
total discharge rate. The pump intake curve (PIC) can then be
generated using Table 4d.4 and plotted against the well IPR.

Page 28
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.13 – Jet Pump Design Curve

Turbine Pumps
A hydraulic turbine pump has been developed by Weir
(*Reference 4d.6)
Pumps Ltd* in the UK, as an alternative to the ESP for
producing very large volumes of fluid (2000 to 100 000
stb/d).

It is based on the same principle as the ESP, but uses a

hydraulic turbine to rotate the shaft at very high rates (5000
to 10 000 rpm). This allows much higher lift capacities (head
and volume) per stage, and hence units are much shorter
(10% of equivalent ESP). Since the rate can be controlled with
the hydraulic supply pressure and volume, the operating
window is very large (10 to 100% rate, and 20 to 150% TDH
at reduced rates).

This concept went into operation during the early 1980s.

Since then a number of problems have been experienced with
the units installed in the Forties field. The main problems
appear to result from both the high rotating speed and
inappropriate metallurgy. Based on this experience there are
still a number of problems that need to be resolved before the
pump becomes a viable artificial lift technique.

Page 29
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

IMPLICATIONS
FOR
COMPLETION
DESIGN If multiple strings are to be used in a hydraulic pump
installation, the casing size is obviously a critical limitation.
Concentric completions are generally favoured because they
are more space efficient and provide a large return conduit
size. If the annulus is used for the produced fluid, special
attention must be paid to corrosion protection, materials
selection, and de-oxygenization of the power fluid.

The two conduits must be properly sized for the intended

rates to minimize friction losses.

Gas venting requirements may make it desirable to have a

third conduit (usually the annulus).

Where SSVs are required, there is the option of installing

them below the pump and using the power fluid pressure for
SSV control. Moreover, if the well is incapable of flow, it may
be possible to obtain a waiver of the SSV requirements.

Page 30
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

ROD PUMPING SYSTEMS

BASIC
EQUIPMENT AND
APPLICATIONS
On land, the most common artificial lift system for low rate
wells is rod or beam pumping. Typically this is used for
stripper wells (<10 stb/d) and for shallow wells (<8000 ft)
producing less than 500 stb/d, although rod pumps are
sometimes used to produce up to 1500 stb/d.

There are three main elements to the rod pumping system

(see Figure 4d.14): the bottom hole pump, the sucker rod
string and the surface pumping unit.

The annulus is normally left open and used to vent any free
gas that can be separated downhole. Tubing is used as the
production conduit to protect the casing from wear and
corrosion. The tubing is generally anchored to the casing and
pulled into tension to minimize tubing movement, buckling
and rod wear.

There are two main types of bottom hole pumps (see

Figure␣ 4d.15 ).

The tubing pump has a higher capacity, but requires the

entire tubing string to be pulled to service the working barrel
or a stuck standing valve.

The rod or insert pump is the most commonly used, because

it can be fully serviced by pulling only the rods. It also
provides higher compression ratios and therefore has less
problems in handling gas.

The pump displacement (PD) is defined by the plunger stroke

(Sp), the pump speed (N), the plunger diameter (D), and the
amount of liquid fillage and/or slippage past the plunger
(Ep = 0.7 to 0.95)

(Equation 4d.4)
PD = Ct x Sp x N x D2 x Ep

where: Ep = efficiency of pump

Ct = correction for terms, pi, etc
Ct = 0.1166 for oilfield units (in, SPM, in2, stb/d)

Page 31
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

Figure 4d.14 – Rod Pumping System

Page 32
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.15 – Two Types of Bottom Hole Rod Pumps

Because of rod stretch and dynamics, Sp will rarely be the

same as the surface stroke, S. The load-displacement plot
forms the basis for pump design and analysis. The fluid load,
Fo, carried by the rods on the upstroke will depend on the net
lift, H, which is the vertical distance from the operating fluid
level (OFL) in the annulus to surface, plus the equivalent
head of any surface backpressure. It will also depend on the
fluid gravity (SG) or density. The API recommends ignoring
the area of the rods when calculating this load:

(Equation 4d.5) Fo = Ct x SG x D2 x H

where: Ct = 0.340 in oilfield units (SG, in2, ft, lbsf)

This load can be estimated from dynamometer surveys,

which measure the rod load versus displacement at the

Page 33
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

surface. These serve as the most effective means of

diagnosing pump problems.

The surface loads on the polished rod will include the fluid
loads, the rod weight, and the dynamic forces. Rod loads are
significant in deeper wells. A tapered rod string is therefore
often employed which involves using thicker rods on the
upper sections, where loads are higher. Buoyancy will vary
throughout the stroke, but is normally calculated on the
downstroke when the travelling valve is open.

Acceleration and friction will also be involved in determining

the dynamic loads. The peak polished rod load on the
upstroke will generally be significantly higher than the sum
of the rod and fluid loads. Similarly, on the downstroke, the
minimum polished rod load will be less than the buoyant
weight of the rods (see Figure 4d.16).

During the up-stroke, rod stretch occurs which can reduce

pump displacement, while on the downstroke, the dynamics
result in overtravel of the rods at higher pumping speeds.
The␣ stroke efficiency is therefore a function of pump speed as
well as the rod loading. The dynamics also cause the rods to
oscillate (harmonically) like a stiff spring.

Figure 4d.16 – Rod Loads and Dynamometer Cards

Page 34
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Since typical pumping speeds of 8 to 15 SPM correspond to

4.2 to 7.9 million cycles per year, the design of a rod
pumping installation focusses on the minimization of fatigue
failures. Corrosion exacerbates fatigue so the allowable
loading is also a function of the operating environment.

Typically, the surface pumping unit is of the beam type,

however a number of other concepts have been developed to
convert the rotary motion of the prime mover to linear motion
of the polished rod.

The reducer gearbox is the critical and most expensive item

in a beam pumping unit; its torque rating (in thousands of in.
lbs) limits the capacity of the system in terms of net lift and
pump rate. Moreover, proper balancing of the pumping unit
is required to maintain a positive torque on the gearbox and
to prevent fatigue of the gear teeth.

Thus, while the rod pumping system is, in principle, a

positive displacement pump, with net lift being independent
of pump rate, in practice the surface unit rating limitations
result in a threshold beyond which the TDH capacity
decreases with rate.

System design is therefore an iterative and complex process

which is normally carried out on the computer. The API have
used their program to generate a set of design curves,
(*Reference 4d.7) published in RP 11L* and have provided some generalized
(*References 4d.8, 4d.9) results in Bulletins 11L3* and 11L4*, which can be used as a
convenient starting point for design work.

However, it should be noted that the production rates in API

11L4 are based on the assumption of 100% volumetric
efficiency and higher pump rates (15 to 25 SPM) on the
smaller units than are normally accepted in the field (8 to
15␣ SPM). It is therefore advisable to enter the curve with a
range of rates between 100 to 200% of the intended target,
for scoping-out the required equipment capacity.

The API design method is not reliable for heavy oil wells
(<20°␣ API), unless proper corrections are applied for the fluid
friction load caused by the high viscosities and the lack of rod
weight on the downstroke. Rod fall problems generally
restrict pump rates to 1.5 to 2.5 SPM and favour the use of
long stroke, variable speed pumping units, such as the
hydraulic pump jacks. The high viscosities also lead to sand
production problems, poor valve performance, poor gas
separation and poor lubrication at standard clearances
(0.003in).

Page 35
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual
Figure 4d.17 – API Bulletin 11L4, Composite Design Curves for
Rod Pump Installations
Page 36
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

IMPLICATIONS
FOR
COMPLETION
DESIGN
If free gas will be present downhole at the intended BHFP, it
is highly desirable to have an open annulus through which
gas can be vented. Therefore, packers should not be installed
in wells that can be expected to be converted to rod pump
within a short period of time (<2 years), unless they are
essential for zonal isolation or safety reasons.

Where wells are to be pumped-off, ie pumped below the

bubble point pressure, the bottom hole pump should be set
below the perforations to aid gas separation, maximize the
drawdown, and minimize blockage of the perforations by fill.
However, immediately after a fracture stimulation treatment,
it is usual to set the pump above the perforations to avoid it
being sanded-in by returning proppant.

In high capacity wells that cannot be pumped-off, the pump

may be set at a relatively shallow depth to reduce the rod
loads, provided that gas separation can be effectively
achieved.

The downhole geometry should be designed to facilitate gas

separation by having a fluid downflow velocity which is much
less then the gas bubble rising velocity predicted by Stokes
Law (approximately 0.3 ft/sec). In wells with moderate to
high viscosities (20 to 5000 cp) and high GLR wells, downhole
gas separation may be enhanced by using a suction tube or
gas anchor.

Where significant tubing-to-casing clearances exist (>0.75in),

the tubing should be anchored and pulled into tension to
prevent buckling on the upstroke and the resultant rod and
tubing wear near the pump. By preventing tubing stretch,
this also improves the stroke efficiency. The tubing design
calculations presented in Chapter 6 – Tubing Movement and
Stress Calculations can be used to determine the required
overpull. However, many engineers use the charts developed
(* Reference 4d.10) by Lubinski and Blenkarn* to determine the required tension
or stretch. Proper tensioning of the tubing is critical in wells
deeper than 3000 ft, especially where production leads to a
significant increase in wellhead temperature (>10°F).

Page 37
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

SCREW PUMP SYSTEMS

Screw or progressive cavity pumps consist of a rubber stator

and stainless steel rotor. These are both spirally machined to
provide a chain of individual cavities between the matching
surfaces. As the rotor turns, the cavity progresses along the
spiral from the inlet to the outlet carrying the fluid with it
and thereby providing a positive displacement (see
Figure␣ 4d.18 ).

The rotary drive is transmitted through conventional sucker

rods (3/4in or 7/8in Class D rods) from surface. A
conventional 32 mm (1 1/4in) polished rod is used with a
rotary stuffing box mounted on the flow tee. Drive can be
provided by:

• Hydraulic drive on the wellhead.

• Belt drive via a torque limiting hub.
• Direct wellhead mounted electric motor.
Typically these units have a differential pressure capacity of
900 to 2600 psi and achieve rates of 5 to 500 stb/d.
However, capacities up to 1500 stb/d may be feasible. Their
greatest application has been in pumping heavy oil wells and
viscous emulsions where conventional pump capacities are
limited by rod fall problems. Moreover, these pumps can
handle large quantities of sand (up to 50%) at low rotational
speeds.

Because of their low capital and operating costs, screw

pumps are taking over from rod pumping in many low rate,
shallow oil wells producing light and medium crude. They are
also very convenient for running pump tests on exploration
wells.

The production rate is proportional to the rotational speed

and can be obtained from manufacturers charts. Rotational
speed is dependent on fluid viscosity and sand content, but
can vary from 50 to 100 rpm in sandy heavy oil to over
500␣ rpm in clean light oil.

The stator rubber must be selected for the operating

environment. Nitrite rubber is limited to temperatures less
than 176°F, while synthetics such as Nysar can handle
temperatures up to 356°F. However, it is not yet possible to
use these pumps on thermal projects.

Page 38
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.18 – Screw Pump Installation

Page 39
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

PLUNGER LIFT

Plunger lift is a lift system for high GLR wells that produce
liquids at relatively low rates (< 500 stb/d). It requires an
area where gas energy can be stored and then supplied to the
tubing at very high rates. Usually the tubing-casing annulus
is used for this purpose, but a natural or induced fracture
system may be adequate. The gas energy is used to drive a
piston, or plunger, carrying a small slug of liquid up to the
surface. After production of the tail gas, the well is closed-in
and the plunger falls back to bottom (see Figure 4d.19).

This lift method is particularly useful for dewatering low rate

gas wells. Requirements include:

• GLR >500 scf/stb.

• PI <1 stb/d/psi.
• Plunger velocity 700 to 1000 ft/min.

Plunger lift efficiency decreases with depth and PI, but

increases with tubing size. The tubing must be carefully
drifted and swaged prior to installation. The system has
been␣ successfully used below a SSV and in sour gas wells
(36% H2S).

Page 40
BP Exploration Determining the Optimum Well Performance – Chapter 4
Completion Design Manual SECTION 4d – ARTIFICIAL LIFT

Figure 4d.19 – Plunger Lift Installation

Page 41
Determining the Optimum Well Performance – Chapter 4 BP Exploration
SECTION 4d – ARTIFICIAL LIFT Completion Design Manual

REFERENCES

4d.1 Brown, K E et al, The Technology of Artificial Lift

Methods, Penwell Publishing Co, Tulsa, 1980

4d.2 Gilbert, W E, ‘Flowing and Gas-Lift Well Performance’,

API Drilling and Production Practices, 126, API, 1954

4d.3 TRW Reda, Submersible Pumps for the Petroleum

Industry, Bartlesville, Oklahoma, 1983

4d.4 Centrilift Hughes, Electrical Submersible Pumps and

Equipment, Tulsa, Oklahoma, 1983

4d.5 Dresser Westech, Electrical Submersible Pumps,

Shawnee, Oklahoma, 1984

4d.6 Weir Pumps Ltd, Weir Hydraulic Downhole Pumping

System, Glasgow, Undated

4d.7 API, ‘Recommended Practice for Design Calculations

for Sucker Rod Pumping Systems’, API RP 11L, API
Production Department, Dallas, 1977

4d.8 API, ‘Sucker Rod Pumping Design Book’, API Bulletin

11 L3, API Production Department, Dallas, 1970

4d.9 API, ‘Curves for Selecting Beam Pumping Units’,

API␣ Bulletin 11 L4 , API Production Deparment,
Dallas,␣ 1970

4d.10 Lubinski, A, Blenkarn, K A, ‘Buckling of Tubing in

Pumping Wells, Its Effects and Means of Controlling
It’, TP 4482, SPE Reprint Series 12 Artificial Lift, SPE,
Dallas, 1975

Page 42