VOLUME FIFTY SIX
DEVELOPMENTS IN PETROLEUM SCIENCE
WELL COMPLETION
DESIGN
By
Jonathan Bellarby
SPE (Society of Petroleum Engineers)
NACE International and
TRACS International Consultancy Ltd.
Aberdeen, UK
Amsterdam

Boston

Heidelberg

London

New York

Oxford
Paris

San Diego

San Francisco

Singapore

Sydney

0910111213 10987654321
ACKNOWLEDGEMENTS
It is one thing to think that you know a subject but quite another to confidently write it
down, secure in the knowledge that no one will challenge you later. I definitely fall into the
former category. I assert that there are no experts in completion design, but there are experts
in specialities within completion design. It is to many of these experts that I have turned for
guidance and verification. I thank Alan Holmes, Paul Adair, Andrew Patterson, Mauricio
Gargaglione Prado, Simon Bishop, John Blanksby, Howard Crumpton, John Farraro, Tim
Wynn, Mike Fielder, Alan Brodie and Paul Choate for their specialist support and reviews.
Their constructive criticism and ideas were essential. It should be apparent from the
references that a considerable number of people inadvertently provided data for this book.
In particular, the Society of Petroleum Engineers (SPE) is a tremendous depository of
technical knowledge, primarily through technical seminars and papers, but also with
technical interest groups and distinguished authors.
This book was written over a two-year time period; much of that time was spent holed
up in a log cabin in the mountains of Western Canada. This involved a not-inconsiderable
disruption to my family who joined me on our ‘sabbatical’. I cannot imagine a more
welcoming and inspirational place than the small town of Canmore, Alberta. There was no
better way of cur ing writer’s block than a run through the woods behind the house, even in
the snow or avoiding bears. It is perhaps telling that a photograph of the area even makes its
way into the book.
When teaching courses or writing books on subjects like completion design, it becomes
apparent that clear, colour drawings are essential. The process of generating these drawings is
worth explaining. I would usua lly dump my thoughts into a hand drawing, with text that
would scrawl, pipe that would wave over the page and perforations that looked like a
seismograph trace in an earthquake. These scribbles would then be neatly transposed into
the drawings you see today. My long-suffering wife Helen was almost solely responsible for
these professional transformations and I owe her an enormous debt.
Jonathan Bellarby
xiii

1.2. Safety and Environment
Safety is critical in completions; people have been killed by poorly designed or
poorly installed completions. The completion must be designed so as to be safely
installed and operated. Safe installation will need to reference hazards such as well
control, heavy lifts, chemicals and simultaneous operations. This is discussed further
1
in Section 11.4 (Chapter 11). Safe operation is primarily about maintaining well
integrity and sufficient barriers throughout the well life. This section focuses on
design safety.
It is common practice to perform risk assessments for all well operations. These
should be ingrained into the completion design. The risk assessments should not
just cover the installation procedures but also try to identify any risk to the
completion that has a safety, environmental or business impact. Once risks are
identified, they are categorised according to their impact and likelihood as shown in
Figure 1.1. Most companies have their own procedures for risk assessments,
defining the impact in terms of injuries, leak potential, cost, etc., and likelihood in
terms of a defined frequency. Mitigation methods need to be identified and put in
place for any risk in the red category and ideally for other risks. Mitigation of a risk
should have a single person assigned the responsibility and a timeline for
investigation. It is easy to approach risk assessments as a mechanical tick in the
box procedure required to satisfy a company’s policy; however, when done properly
and with the right people, they are a useful tool for thinking about risk. Sometimes,
risks need to be quantified further and numerically. Quantitative Risk Assessments
(QRAs) attempt to evaluate the risk in terms of cost versus benefit. QRAs are
particularly useful for decisions regarding adding or removing safety-related
equipment. Clearly, additional expertise with completion engineering is required
for these assessments. Such expertise can assist in quantifying the effect of leaks,
fires, explosions, etc., on people, nearby facilities and the environment.
Example – annular safety valves
Annular safety valves are used to reduce the consequence of a major incident on a

The barriers do not necessarily need to be mechanical barriers such as tubing; they
can include mud whilst drilling or the off switch of a pumped well. Examples of
barriers during various phases of well construction and operation are shown in
Table 1.1.
The primary barrier is defined here as the barrier that initially prevents
hydrocarbons from escaping; for example, the mud, the tubing or the Christmas
tree. The secondary barrier is defined as the backup to the primary barrier – it is not
Table 1.1 Examples of barrier systems through the life of the well
Example Primary Barrier Secondary Barrier
Drilling a well Overbalanced mud capable of
building a filter cake
Casing/wellhead and BOP
Running the upper
completion
Isolated and tested reservoir
completion, for example
inflow-tested cemented
liner or pressure-tested
isolation valve
Casing/wellhead and BOP
Pulling the BOP Packer and tubing Casing, wellhead and tubing
hanger
Isolated reservoir completion,
for example deep-set plug
Tubing hanger plug. Possible
additional barrier of
downhole safety valve
Operating a naturally
flowing well
Christmas tree Downhole safety valve

Ideally, pressure testing should be in the direction of a potential leak, for
example, pressure testing the tubing. Sometimes this is not practical. If there is
anything (valve openings, corrosion, erosion, turbulence, scale, etc.) that can affect a
barrier then the bar rier should be tested periodically. This applies to the primary
barriers and often to the secondary barriers as well (e.g. safety valve).
1.2.2. Environmental protection
Completions affect the environment. Sometimes this is for the worse, and
occasionally for the better. The environmental impact of completion installation is
covered in Section 11.4 (Chapter 11), including waste, well clean-ups and harmful
chemicals. The design of completions has a much greater environmental effect.
1. An efficient completion improves production but also reduces the energy
consumption (and associated emissions) required to get hydrocarbons out of the
g round.
2. Well-designed completions can reduce the production of waste materials by
being able to control water or gas production.
3. Completions can be designed to handle waste product reinjection, for example
drill cuttings, produced water, non-exported gas, sulphur or sour fluids.
Sometimes this disposal can be achieved without dedicated wells. These
combination wells are covered in Section 12.6 (Chapter 12).
4. Carbon capture and sequestration will likely become a big industry. Carbon
sequestration may not be associated with oil and gas developments, for example
injection of carbon dioxide from a coal power station into a nearby saline aquifer.
Carbon sequestration may also involve active or decommissioned oil and gas
reservoirs. Regardless, sequestration requires completions. Sequestration is
discussed in Section 12.9 (Chapter 12).
Safety and Environment4
P
P
P
Primary

drilling with mud. Possible
pressure test during
completion operations.
How tested
Inflow tested during
completion installation.
Subsequent periodic inflow
tests.
Pressure tested during
drilling (with mud).
Sometimes tested during
completion annulus pressure
test (brine). Not routinely
tested during operations.
Possible test as part of
leak off test of next hole
section. Monitoring of
annulus pressure (except
subsea wells).
Christmas tree
valves
Tree connection
to tubing hanger
Tubing and body
of completion
components
Packer
Casing under
packer
Figure 1.2 Example of a well barrier schematic.

rig owners,
rig crew, etc.)
Geologists,
petrophysicists
and geophysicists
Facilities, process
and plant operations
Service sector Management
Reservoir engineers
Completion
engineers
Specialists
(metallurgists,
chemists, etc.)
Commercial
analysts
Figure 1.3 Te a m i n t e g r a t i o n .
The Role of the Completion Engineer6
export routes. So all a completion engineer has to do is fit the completion into the
casing and produce the fluid to a given surface pressure. Many opportunities for
improvement are lost this way.
A vital role of completion engineers is to work with the service sector. The
service sector will normally supply the drilling rig, services (wireline, filtration,
etc.), equipment (tubing, completion equipment, etc.), consumables (brine,
proppant, chemicals, etc.) and rental equipment. Importantly, the service sector
will provide the major ity of people who do the actual work. Inevitably, there will be
multiple service companies involved, all hopefully fully conversant with their own
products. A critical role of the completion engineer is to identify and manage these
interfaces personally, and not to leave it to others.
For small projects, a single completion eng ineer supported by service companies

Introduction 7
For each piece of data, understand where it comes from, what the uncertainty
range is and how it might change in the future. A large range of uncertainty
promotes completions that can cope with that uncertainty. For example, if it is
not known whether an aquifer will naturally support oil production, the
possibility of water injection requires consideration. Water injection wells do not
necessarily need to be designed, but consideration is required for converting a
producer to an injector or for dealing with associated water injection issues
(souring, scaling, etc.).
Appraisal wells are frequently overlooked as opportunities for completion
eng ineers. Their primary purpose is to reduce uncertainty in volumetric
estimations. These wells are also an opportunity to try out the reservoir completion
technique that most closely matches the development plan. For example, if the
development plan calls for massive fracturing of development wells, some of the
appraisal wells should be stimulated. This adds value by reducing uncertainty in
production profiles emanating from tentative fracturing designs and provides data
on which to base improvement of the completion.
Reservoir parameters
(pressure, temperature,
production profiles,
water cuts, etc.)
Project and
commercial
(timeframes, profitability
drivers, license constraints)
Drilling
(trajectory, casing, muds,
formation damage)
Facilities
(throughput, pressures,

additional million dollars on a corrosion-resistant completion or to install a cheaper
completion that is expected to be replaced in 10 years’ time. If the completion fails,
a rig has to be sourced and a new completion installed; this costs money and a delay
in production. The time value of money reduces the impact of a cost in 10 years. In
the case of the onshore well producing at lower rates where a workover is cheaper,
this workover cost is less than the upfront incremental cost of the high-specification
Production
Time
(years)
W
a
t
e
r
O
i
l
o
r
g
a
s
High reliability to reduce
operating costs and
maintain plateau.
Declining reservoir pressure and
onset of water; possible artificial
lift requirement to maintain plateau.
Minimising production
decline through artificial

Simple, reliable equipment

Minimisation of well interventions, for example water shut-off, by improved
completion design
The problem is that these two requirements are conflicting. Remotely shutting
off water can be achieved by smart wells (Section 12.3, Chapter 12) for example, but
this clearly increases complexity and arguably reduces reliability. A balance is
required.
1.6. The Design Process
Many operators have their own internal processes for ensuring that designs are
fit for purpose. There is a danger that such processes attempt to replace competency,
that is, the completion must be fit for purpose so long as we have adhered to the
process. Nevertheless, some elements of process are beneficial:

Pulling together the data that will be incorporated into the design. This
document can be called the statement of requirements (SoR). The SoR should
incorporate reservoir and production data and an expectation of what the
completion needs to achieve over the life of the field.
Table 1.2 Economic examples of completion decisions
The Design Process10

Writing a basis of design. This document outlines the main decisions made in the
completion design and their justification. The table of contents of this book gives
an idea of the considerations required in the basis of design. This document can
form the basis of reviews by colleagues (peer review), internal or external
specialists and vendors. The basis of design should include the basic installation
steps and design risk assessments. It is often useful to write the basis of design in
two phases with a different audience in mind. The outline basis of design covers
major decisions such as the requirement for sand control, stimulation, tubing size
and artificial lift selection. These decisions affect production profiles, well

Well trajectory and inclination

Open hole versus cased hole

Sand control requirement and type of sand control

Stimulation (proppant or acid)

Single or multi-zone (commingled or selective)
Introduction 11
Barefoot
Pre-drilled or
slotted liner
Cemented and
perforated liner
or casing
Open hole
sand control
screens/gravel
pack
Cased hole
gravel pack or
frac-pack
Figure 1.6 Reservoir co mpletion methods.
Tubingless
completion
Tubing
completion
without packer
Tubing

coverage of open hole completions and the specifics of perforating and stimulation
(proppant and acid).
2.1. Inflow Performance
Inflow performance is the deter mination of the production-related pressure
drop from the reservoir to the rock face of the reservoir completion. This section
serves as an introduction to inflow performance for open hole wells. The details of
inflow performance related to cased and perforated wells are discussed in Section
2.3.4. It is useful to determine, in outline, the inflow performance for different well
geometries for the reservoir as part of selecting completion strategies such as open
hole versus cased hole. Inflow performance also allows a value comparison of
different reservoir completions such as a vertical, hydraulically fractured well
compared to a long, open hole horizontal well. Although inflow performance
might appear to be the remit of the reservoir engineer, an integrated approach is
required – many aspects of completion design affect inflow performance and must
be assessed.
Understanding fluids (shrinkage, viscosity, gas to oil ratios, etc.) is an integral
part of inflow performance. Section 5.1 (Chapter 5) includes a detailed discussion of
the behaviour of hydrocarbon fluids.
The starting point for inflow performance is to consider pressure drops in a
cylinder of rock as shown in Figure 2.1.
The pressure drop through the rock is dependent on the flow rate, viscosity,
cross-sectional area of the rock and the length of the section. Whilst investigating
the hydraulics of water flow through sand beds, Henry Darcy (French scientist
1803–1858) suggested that the pressure drop also depends on a property of the sand,
i.e. permeability (k). The unit of Darcy is named in his honour, although the
p
o
p
i
q

, the formation volume factor, that is the conversion from stock
tank conditions to reservoir conditions (res bbl/stb) (see Section 5.1.3, Chapter 5 for
more details on oil behaviour and shrinkage). m
o
, the viscosity of the oil (cp); l,the
length of the rock sample (ft); A, the cross-sectional area of the rock (ft
2
); k
o
,the
per meability of the rock to oil (md); and p
i
Àp
o
, the pressure drop between the inlet
and outlet.
This equation and the ones that follow can be converted to fluid flow involving
mixtures of oil and water by incorporating a flow rate term for water with an
appropriate water formation volume factor (close to 1), water viscosity and water
permeability.
This equation has its uses – for example the pressure drop through tubing full of
sand or perforations packed with gravel. However, for reservoir flow in a vertical
well with a horizontal reservoir, flow is radial as shown in Figure 2.2.
This radial flow accelerates the fluids as they move from the effective drainage
area and approach the wellbore. Correcting (integrating) for the geometry of the
flow in the idealised conditions shown in Figure 2.2, the inflow performance is
given by:
q
o
¼

w
, the wellbore flowing pressure.
Horizontal
reservoir
Open hole, vertical,
undamaged well
k
o
h
q
o
r
e
p
e
p
w
r
w
μ
o
Figure 2.2 Radial in£ow.
Inflow Performance16
The outer pressure (p
e
) has been replaced with the average reservoir pressure
(
p
r
). This correction introduces 0.472 into the logarithm. The difference between

temperature (R); z, the gas compressibility factor at the average pressure and
temperature; k
g
, the permeability to gas.
The square relationship to pressure derives from the gas law – low pressures
create high volumes and hence high velocities.
These equations also define the pressure profile through a reservoir. An example
is shown in Figure 2.3 for an oil well and in Figure 2.4 for a gas well.
Marked on the charts are the points where 50% of the pressure drop occurs –
around 26 ft for the oil example and only 5.3 ft for the gas example. The gas
example has been manipulated to give the same drawdown as the oil example, that
is 5000 psi. The low bottom hole pressure creates gas expansion and thus the
different shape and large pressure drop near the wellbore. In reality, in the gas case,
the situation would be even more severe due to turbulent flow.
A plot of drawdown and rate creates the inflow performance relationship (IPR).
For the two examples shown in Figures 2.3 and 2.4, the IPRs are shown in Figure
2.5 and 2.6.
Half the pressure
drop occurs
within 26 ft
Assumptions:
Oil well, semi-steady state
8500 bpd
8.5 in. open hole diameter
100 ft thick, 100 md formation
Viscosity 4 cp
Oil formation factor 1.2
Average reservoir pressure 5565 psia
0
0

o
p
r
À p
w

(2.4)
The PI is a function of the fluids, the rock and the geometry of the reservoir and
well. It can be measured by a multi-rate well test – assuming that each rate step
achieves near pseudo steady state. Oilfield units are bpd/psi.
For a gas well, there is no straight line and therefore no PI. In fact, the oil inflow
relationship is only valid above the bubble point and assumes a constant viscosity
0
0
1000
2000
3000
Pressure (psia)
4000
5000
6000
500 1000
Distance from centre of well (ft)
1500 2000
Half the pressure
drop occurs
within 5.3 ft
Assumptions:
Gas well, semi-steady state
25 Mmscf/D

shown in Eq. (2.5).
q
o
¼
0:00708k
o
h p
r
À p
w

m
o
B
o
lnð0:472r
e
=r
w
ÞþS

(2.5)
A negative skin factor represents superior inflow performance to a vertical
undamaged open hole well. Given that ln(0.472r
e
/r
w
) is typically between 7 and 8,
the skin factor can never go far below around À5. Conversely, a blocked well has an
infinitely positive skin. The skin factor incorporates all aspects of near-wellbore

6000
Reservoir pressure = 5565 psia
4000
2000
Bottom hole flowing pressure (psia)
0
0 10000 20000
Rate (Mscf/D)
30000
Absolute open
flow (AOF)
Drawdown 5000 psi
Figure 2.6 Example gas in£ow performance.
Reservoir Completion 19
For example a skin factor of À4 is equivalent to converting an 8.5 in. diameter
borehole to a 38.7 ft diameter borehole. This visualisation also works the other way
round – it is surprising how little difference altering the borehole size makes.
If the degree and depth of damage is known, the skin factor can be calculated:
S ¼
k
k
d
À 1

ln
r
d
r
w


in the inflow equation and is relevant to both oil and gas flow.
A selection of the shapes given by Odeh is shown in Figure 2.8. A more generalised
form for a variety of other shapes and mixed flow/no-flow boundaries is given by
Yaxley (1987).
Figure 2.7 E¡ective drainage areas and virtual £ow boundaries.
Inflow Performance20
For example, for the triangular drainage area drained by well x4inFigure 2.7,
the pseudo steady-state inflow performance for oil would be approximated by:
q
o
¼
0:00708k
o
h p
r
À p
w

m
o
B
o
ln 0:472 Â 0:604
ffiffiffiffi
A
p
=r
w

þ S

r
À 0:8
p
w
p
r

2
(2.10)
where q
o(max)
is calculated from well tests and is the same as the absolute open flow
(AOF) potential.
Example. Figure 2.9 shows an example for a saturated reservoir (reservoir pressure
equals bubble point pressure) with the following parameters:

Well test bottom hole pressure ¼ 3500 psia at 7800 stbpd.

Average reservoir pressure ¼ 4800 psia.
From Eq. (2.10), q
o(max)
is 18189 stbpd. From this figure, the rest of the inflow
performance can be calculated as shown in Figure 2.9.
Standing (1971) modified Vogel’s relationship for undersaturated fluids. A
straight-line inflow performance is used above the bubble point and a revised
relationship used below the bubble point [Eq. (2.11)].
q
o
À q
b

¼
1:8ðq
oðmaxÞ
À q
b
Þ
p
b
(2.12)
5000
4000
3000
2000
Bottom hole flowing pressure (psia)
1000
0
0 5000 10000
Rate (stbpd)
15000 20000
AOF = 18189 stbpd
Well test result
Figure 2.9 Vogel in£ow performance relationship example for a saturated £uid.
Inflow Performance22