MPD Course Manual

Advanced MPD
Well Design
Chapter 1 – Introduction
INTRODUCTION 1-1 Rev. 1.0

Table of Contents
1 Introduction to the Managed Pressure Drilling Well Design and Operations Course1-
3
1.1 MPD Defined ................................................................................................... 1-4
1.1.1 Pressure Gradients ................................................................................... 1-4
1.1.2 Illustrative Example of MPD ................................................................... 1-7
1.2 IADC UBO and MPD Sub Committee ............................................................ 1-9
1.2.1 Risk Classification Level ....................................................................... 1-10
1.2.2 Application Type ................................................................................... 1-11
1.2.2.1 Level A Systems ................................................................................ 1-11
1.2.2.2 Level B Systems ................................................................................ 1-11
1.2.2.3 Level C Systems ................................................................................ 1-12
1.2.3 Fluid Type .............................................................................................. 1-12
1.3 Drivers for Considering MPD........................................................................ 1-12
1.4 MPD Options ................................................................................................. 1-14

1 Introduction to the Managed Pressure Drilling Well
Design and Operations Course
Controlling the pressure in a wellbore while drilling for oil
or gas is an operation that draws its roots from Colonel
Drake’s Spindletop well drilled in the late 1800’s.
Arguably thought of as the moment which gave rise to the
modern Blow Out Preventer (BOP) and drilling fluid
industries, the geyser of oil at Spindletop graphically
demonstrated that geological pressure must be respected in
order to access what lies beneath.
Historically, the most apparent way to control subterranean

pressure was to take advantage of hydrostatics. Altering the
density of the drilling fluid, or mud, allowed the driller to
keep those hydrocarbons at bay while drilling and reach the
desired target ‘safely’. Drilling in this manner continued
for decades, until three key characteristics began to emerge: Figure 1: The blowout at
Spindletop.
1. Reservoir pressures do not remain at ‘virgin’ conditions
and eventually decline below the hydrostatic pressure of even the lightest drilling
fluids,
2. Exerting excessive hydrostatic pressure on a formation could (in some cases) cause
terminal loss of drilling mud, and
3. Although hydrostatically over-pressuring the reservoir kept hydrocarbons in place,
the exerted pressure on the porous rock had (in some cases) limited the reservoir’s
productive potential through damaging the near wellbore.
These characteristics, amongst a host of others, prompted the drilling industry to explore
new avenues of annular pressure control. Technological advances in surface equipment
resulted in a host of new terms and acronyms thrust upon drillers and engineers to
describe these new techniques:
Mud Cap Drilling
Dual-gradient Drilling
Air Drilling
Managed Pressure Drilling (MPD)
Underbalanced Drilling (UBD)
The common theme linking these different drilling techniques is the attempt to actively or
proactively control the annular wellbore pressure profile. In contrast, conventional
drilling practices react to changing wellbore conditions by altering the mud weight
according to observations of differences in mud volumes (kicks or losses). All of the
terms mentioned in the bulleted list above have come to signify specific annular pressure

control techniques, and thus none can be used as an umbrella term to adequately describe
them all.
1.1 MPD Defined

In conventional drilling operations, the mud weight is selected such that its static gradient
is higher than the exposed formation pressure. The system is open, returning the fluid to
atmospheric tanks. When circulating, the pressure imposed on the formation increases
due to frictional pressure of fluid moving in the wellbore. The Bottom Hole Pressure
(BHP) is controlled by the following equation:
BHP = ∆PGravity + ∆PFriction (1)
Where:
∆PGravity = hydrostatic pressure due to mud weight
∆PFriction = friction pressure due to circulation
While conventional drilling uses only fluid density to manage pressure, MPD uses a
combination of surface pressure, fluid density, friction, and energy terms to balance the
exposed formation pressure. The addition of specialized MPD equipment like the
Rotating Control Device (RCD) and MPD choke enable the application of surface
pressure to achieve the desired annular pressure profile. Other variables are now
introduced into the pressure equation:
BHP = SP + ∆PGravity + ∆PFriction + ∆PEnergy * (2)
Where: BHP = bottom hole pressure

SP = surface pressure
∆PGravity = hydrostatic pressure due to mud weight
∆PFriction = friction pressure due to circulation
∆PEnergy = pressure changes as a result of the energy of another device (ie; sea
floor pump)
The point is that conventional drilling only uses gravity and friction, while MPD uses the
other components of the equation to manage the pressure.
1.1.1 Pressure Gradients

Pressure gradient curves are commonly used to map out the subterranean pressure
profiles. Pore pressure may be thought of as the pressure limit which traps subterranean
fluids, while fracture pressure delimits an upper bound in pressure, above which the rock
would fracture as a result of the injection of fluids. Pore and frac pressures can vary with
depth and are typically non-linear. The pressure exerted by a single phase fluid in a
*
Pressure change due to acceleration, ∆PAccelaration , is also considered in high energy applications. The
pressure effects due to acceleration are negligible, and therefore not discussed in this manual.

wellbore follows a linear gradient, or slope. A column of static, unmoving fluid follows a
pressure gradient dependent upon its density. Pumping the same fluid will alter the slope
of the fluid gradient due to the additional friction in the system. At any given depth, the
pressure exerted by pumping fluid will be higher than that of a static fluid. These fluid
gradients are commonly referred to as static and dynamic fluid gradients. The term
Equivalent Circulating Density (ECD) is used to describe the equivalent static density
of a fluid if it were to follow the dynamic fluid gradient curve. ECD is comprised of the
static mud weight and friction pressure term, and is usually expressed in pound per gallon
(ppg).
Figure 2 and Figure 3 illustrate the differences between conventional and managed
pressure drilling pictorially. Figure 2 depicts a traditional pore-fracture gradient window
for a hypothetical well. Static and dynamic fluid gradient curves are superimposed.
Simplistically, the next casing point is highlighted at the bottom of the figure where the
dynamic circulating density line approaches the fracture gradient.
Pressure
Static Fluid Gradient

(Pumps Off)
Dynamic Fluid Gradient

(Pumps On)
Pore
Pressure
Frac
Depth
Pressure
Casing Point
Drilling window
Maximum Depth for
Next Casing Point
Figure 2: Conventional drilling fluid gradients & casing setting points.

The margin between the static and dynamic gradients defines the drilling window, shown
shaded.
Figure 3 shows the same hypothetical well, demonstrating an MPD option with a static
mud weight gradient (blue) less than pore pressure. Note that although the dynamic

gradient (black) results in an overbalanced state while, the system is statically
underbalanced at some points in the wellbore. Application of surface back pressure
however, results in a shift of the static gradient (orange) above pore pressure. In this
example, an anchor point pressure at the bit is chosen, whereby the pressure is matched
with pumps on and pumps off at this point. Higher in the wellbore, this curve crosses
frac pressure, however the result is inconsequential as this point is behind casing.
The one dramatic result demonstrated in this figure is the ability to drill deeper with the
same mud system, thus extending the casing depth or reaching the planned target.
Surface pressure
Pressure

(Pumps Off)

(Pumps On)
Pore
Pressure
Frac
Depth
Pressure
Casing Point
Plus Surface Pressure
MPD window
Statically Maximum Depth for

UB Next Casing Point
Figure 3: MPD fluid gradients and drilling window.
Applied back pressure MPD wells may be distinguished by the chosen fluid density,
whereby the resultant annular pressure system may be statically underbalanced, or
statically overbalanced. As alluded to previously, the ‘static’ term in this case refers to
the state of the rig’s mud pumps, whereby static means ‘pumps off’ and dynamic means
‘with rig pumps on’, both with no applied back pressure. The chosen weight of the fluid,
therefore will dictate the complexity in further planning of an MPD well relative to
inherent risk. A statically overbalanced well is considered less risky as the well is
controlled with the rig’s pumps off simply by the weight of the drilling fluid. In a
statically underbalanced system, the well may ONLY be controlled with the application

of further surface back pressure when the rig’s pumps are off. The risk of an unplanned
influx from the wellbore must then be accounted for appropriately
1.1.2 Illustrative Example of MPD

Pore and Fracture Pressure Profile (EMW)
EMW (ppg)
7 8 9 10 11 12 13 14
0
Pore Pressure
Fracutre Pressure
2000
MW
4000
6000
Depth (ft)
8000
10000
12000
14000
Figure 4 - Pore and Fracture Pressure Profile in terms of Equivalent Mud Weight, for Illustrative Example

Pore and Fracture Pressure
Pressure (psi)
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
0
Pore Pressure
Fracture Pressure
2000
4000
6000
Depth (ft)
8000
10000
12000
14000
Figure 5 - Pore and Fracture Pressure Profile in terms of Pressure, for Illustrative Example
The Figures above show the EMW PP and FG, and the same data in terms of pressure,
for an example well. The pressures and depths do not sound impressive (10.5 ppg
reservoir at 12.500+ ft), but you do not need high depths, pressures and temperatures to
have challenging wells!
In Figure , the casing seat selection based on the max MW for each interval is illustrated-
showing proposed casing shoe depths. We clearly see two problem areas in conventional
drilling:
In the section between 2000-4000+ ft, there is some risk of losses as the MW is too close
to the fracture pressure at 2600 ft.
In the section below 10,000 ft, there appears to be no mud weight to allow us to drill all
the way to TD, because of that pesky little pressure ramp we see below 10.000 ft! There
is a very high probability that we will need to set liner very soon (about 10300 ft), before
we can continue drilling. If we could just deepen the previous casing shoe slightly (i.e.,
bring it to say 10,400 ft), we can drill the rest of the well without setting pipe, thus
eliminating the liner, as well as risks associated with drilling.
MPD is certainly applicable in the section below 4,100 ft. A combination of fluid density
and surface back pressure can navigate the tight pore-frac margin, and shoe can be set at
10,400 ft. (this can be illustrated using Figure ). This allows for the next section to be
drilled conventionally to TD. The result: reach TD in 3 strings rather than 4.

MPD may also be advisable in the section above (2000 ft – 4100 ft). Although we can
probably squeak by, there is risk of lost circulation, especially considering ECD effects.
With MPD, although the shoe may not be deepened, the risk of lost circulation is greatly
reduced.
1.2 IADC UBO and MPD Sub Committee

The IADC Underbalanced Operations & Managed Pressure Drilling Committee strives to
promote the safe and efficient application of underbalanced and managed pressure
drilling operations worldwide.
In early 2008, the sub-committee ratified the definition of Managed Pressure Drilling,
distinguishing it from underbalanced operations. As taken from the IADC website:
Managed Pressure Drilling (MPD) means an adaptive drilling process used to control
precisely the annular pressure profile throughout the wellbore. The objectives are to
ascertain the downhole pressure environment limits and to manage the annular hydraulic
pressure profile accordingly. MPD is intended to avoid continuous influx of formation
fluids to the surface. Any flow incidental to the operation will be safely contained using
an appropriate process.
1. An MPD process employs a collection of tools and techniques which may mitigate
the risks and costs associated with drilling wells that have narrow downhole
environmental limits, by proactively managing the annular hydraulic pressure profile.
2. MPD may include control of back pressure, fluid density, fluid rheology, annular
fluid level, circulating friction, and hole geometry, or combinations thereof.
3. MPD may allow faster corrective action to deal with observed pressure variations.
The ability to control annular pressures dynamically facilitates drilling of what might
otherwise be economically unattainable prospects.
The key phrase that has been emphasized in the development of this definition is “MPD
is intended to avoid continuous influx of formation fluids to the surface”. This distinction
draws a line between techniques based upon pressure bias, with the mid point being
reservoir pressure. Planned operations whereby the resultant pressure bias is below
reservoir pressure resulting in formation fluid influx are referred to as UBD. Others may
be referred to as MPD.
One of the main achievements of the committee is the adoption of a standard well
classification system for MPD & UBO. The classification is three part alpha numeric,
and resembles the bit grading system. The ultimate goal of this classification is to specify
the equipment required to drill each type of well safely. At the time of writing, the IADC
sub-committee has reviewed a draft copy of the recommended practices for
underbalanced operations, which will be ultimately submitted to the American Petroleum
Institute (API) for review and ratification. A similar document for MPD operations is
currently being drafted.

The classification uses an alpha numeric system to describe the following:
Table 1: Elements of the MPD & UBO well classification system
Risk Level Application Type Fluid Type

0 A 1
1 B 2
2 C 3
3 4
4 5
5
Table 2, Table 3, and Table 4 below define a matrix which classifies the majority of
known MPD & UBO applications. As stated above, the IADC system combines the risk
management categories previously defined (Levels 0 to 5) with a sub-classifier for
application type to indicate if wells are drilled MPD, underbalanced, or using mud cap
techniques (A-C). In order to provide a complete method of classifying the type of
technology used for one or more sections of a well, or multiple wells in a particular
project, a third component of the classification system addresses the fluid type used (1-5).
1.2.1 Risk Classification Level

The first tier of the system is used to characterize risk. The following table describes the
6 Classification Levels identified by the IADC. The levels are described based on the
ability of the well to flow and the comparative pressure rating of the primary barrier to
the maximum anticipated surface pressure.
Table 2. IADC UBO Committee Risk Classification Levels

Level Description
0 Performance enhancement only; no hydrocarbon containing
zones
1 Well incapable of natural flow to surface, inherently stable
and a low-level risk from a well control point of view.
2 Well capable of natural flow to surface but enabling
conventional well control methods and has limited
consequences in the case of catastrophic equipment failure.
3 Geothermal and non-hydrocarbon production. Maximum
shut-in pressures are less than MPD/UBO equipment
operating pressure rating. Catastrophic failure has immediate
serious consequences.
4 Hydrocarbon production. Maximum shut-in pressures are
less than MPD/UBO equipment operating pressure rating.
Catastrophic failure has immediate serious consequences.
5 Maximum projected surface pressures exceed MPD/UBO
operating pressure rating but are below BOP stack rating.
Catastrophic failure has immediate serious consequences.
1.2.2 Application Type

The Application Types are described in three levels. These are tabulated below.
Table 3. IADC Application Types

Level Description
A Managed Pressure Drilling
B Underbalanced Operations
C Mud Cap drilling
1.2.2.1 Level A Systems

Most MPD methods use similar equipment and techniques that underbalanced operations
use however, the circulating system is normally designed to be only slightly overbalanced
to the reservoir at all times, even when pumps are off. These methods generally remove
the safety margins of conventional drilling operations and, although the well is still
drilled overbalanced, the system is designed to handle produced fluids in the event of a
kick and often drilling can continue in this situation.
The primary distinction is that applied pressure in MPD operations results in annular
pressure above the pore pressure, where UBD the applied pressure is below the pore
pressure, in at least one point of the well
1.2.2.2 Level B Systems

An underbalanced state occurs when the annular bottom-hole circulating pressure of the
drilling fluid is less than the pore pressure of the formation being drilled. If underbalanced
conditions are maintained while drilling, formation fluids enter the well when a permeable
horizon is penetrated, thus leading to production during drilling.

In drilling underbalanced, surface equipment such as a rotating control head, high pressure
flow lines, UBD choke manifold, separator system and flare lines allow for the produced
hydrocarbons and associated surface pressures to be processed safely.
1.2.2.3 Level C Systems

Mud Cap Drilling is a technique used to drill through producing zones without returns to
surface i.e. total losses. Fluid is simultaneously pumped down the drill string at a
sufficient rate to cool the bit and transport cuttings up to the loss zone as well as down the
annulus at a rate sufficient to keep produced fluids from migrating to surface.
1.2.3 Fluid Type

Fluid systems have been categorized by the IADC by the following system:
Table 4. IADC Fluid Types

Level Term SG Description
1 Gas drilling 0.00-0.02 Drilling process using only gas
as the drilling medium; no
intentional fluid added.
2 Mist drilling 0.02-0.07 Drilling with liquid entrained
in a continuous gaseous phase;
typical mist systems have
<2.5% liquid content.
3 Foam drilling 0.07-0.6 Drilling with a two-phase fluid
and a continuous liquid phase
generated from the addition of
liquid, surfactant, and gas;
typical foams range from 55%
to 97.5% gas.
4 Gasified 0.55-0.9 Drilling with a gas entrained in
liquid drilling a liquid phase.
5 Liquid 0.8 and Drilling with a single liquid
drilling above phase.
For example, a 2-A-4 well describes a well which is able to flow to surface under its own
energy (2) and is drilled MPD (A) with two phase fluid (4).
1.3 Drivers for Considering MPD

The primary drivers for considering MPD and UBO techniques can be summarized as:
Drilling enabling
Maximize production through minimizing reservoir impairment
Reservoir characterization
Minimizing NPT and ILT
Reduced drilling cost

Where the predominant value derived from underbalanced operations has been
maximizing production and increasing reservoir characterization, the value in MPD is
primarily in drilling enabling. A comparison of the three broad variants of MPD and
UBO is shown in Table 5.
Table 5: A relative comparison of operational facets relevant to MPD and UBO. The colors depict relative
applicability, not absolute. For example, it is possible to detect kicks in a UB well, however, they are
harder to detect than an MPD well.
OBD MPD UBO Mud Cap
Regulatory Policy
Drilling Problems
ROP Improvements
Well Control
Reduce Formation Damage
Reservoir Characterization
Surface Equipment Complexity
Kick Detection and Control
Tripping and Completions
Regulatory Policy – The simple fact that the underlying principle of MPD
operations is to maintain an annular pressure above pore pressure eases the minds
of most Regulators. In mud cap operations, gas migration from the reservoir is
continually fought where UBO allows hydrocarbons to flow surface, complicating
regulations somewhat. In some parts of the world, studies have proven that UB
operations are inherently safer than conventional (overbalanced) drilling
operations, and the local regulators have agreed with them. As the techniques are
not yet as universally accepted as conventional drilling, work is required when
entertaining each technique. Regulations governing MPD and UB operations
remain regional, not global, and therefore may differ from region to region and
country to country.
Drilling Problems – Each of the drilling techniques tackles drilling problems in
their own way, and all positively impact the drilling process. The reduction in
required mud weight in MPD and UBO can minimize or eliminate stuck pipe and
losses as well as navigate narrow pore-frac gradients. Where terminal losses are
inevitable, mud cap drilling allows troublesome zones to be navigated.
ROP Improvements – Studies and practice alike have demonstrated that
lightening the drilling fluid reduces the chip hold down effect and increases on
bottom ROP. The lower the Bottom Hole Circulating Pressure (BHCP), the
greater the ROP.
Well Control – MPD demonstrates a relatively higher degree of well control, as
the goal in MPD design is to remain above pore pressure at all times. The
comments made in the Regulatory Policy item hold true for Well Control.

Reduce Formation Damage – On the reservoir damage scale, UBO is the clear
leader. The negative pressure bias clearly lends to a situation where no drilling
induced formation damage occurs, as the fluid systems used are very benign and
no filter cake is formed on the borehole wall. Drilling blind with terminal losses
in mud cap operations represents the opposite end of the scale as drilling fluid,
cutting, produced fluid, and heavy annular mud are forced into the formation.
MPD fits somewhere between the two as the hydrostatic head is reduced, but
maintained above pore pressure. Damage is therefore lessened, but not
eliminated.
Reservoir Characterization – The influx of hydrocarbons offers a unique
opportunity to fingerprint the reservoir while drilling UB. As MPD operations are
dynamically overbalanced, this opportunity does not exist. In certain cases, where
a statically UB fluid is used, the annular pressure may be reduced to ‘sample’ the
pore pressure profile discretely thus hinting at a form of characterization. Mud
cap operations offer extremely limited forms of RC, only identifying large
cavernous fractures that take everything pumped and produced.
Surface Equipment Complexity – UBO is considered the ultimate pressure
control environment, and in addition to the pressure control equipment used in
MPD or Mud Cap operations, a complete fluid processing plant is required to deal
with reservoir influx.
Kick Detection and Control - It is possible to detect kicks in a UB well,
however, they are harder to detect than an MPD well. Kick detection in
Pressurized Mud Cap operations is possible, but not so in Floating Mud Cap.
Tripping and Completions – Tripping and running completions present two of
the most difficult operations in MPD and UBO. The complexity of the operations
increase if some amount of surface back pressure must be applied to trip in and
out of the hole. As typically lower pressures are required in MPD wells, the
operations are comparatively easier than UB and mud cap.
1.4 MPD Options

Managed Pressure Drilling is all about pressure balance. Altering the annular pressure
profile in MPD operations can be achieved many ways, a shown in Figure , whereby
individual techniques may be classified by the mechanism in which pressure control is
realized.
Annular Friction methods primarily address systems that focus on directly controlling
annular friction pressures, or the ECD, to control BHCP. ECD can be maintained during
times when the rig pumps are off through continually circulating fluid. Conversely, the
ECD can be ‘enhanced’ by either increasing or decreasing the ECD as desired
Surface Pressure methods are rather self-explanatory and affect the overall pressure
profile by applying back pressure at surface, usually by means of a choke.

Figure 6: MPD pressure control options according to defining pressure control mechanism.
Energy Change methods alter the annular pressure profile by means of a mechanical
device, either by separate annular mud lift pumps or drillstring deployed jet pumps.
Density design methods focus on the weight of the drilling fluid as the primary focus of
pressure control. Fluid density methods in which circulation is maintained include low
head and dual gradient, whereas non-circulating methods are best demonstrated by mud
cap techniques.
Which MPD method suits a particular project is the result of detailed engineering and
planning, balancing well objectives with design constraints and economics. The
following chapters shall present an approach to answer these questions.

Chapter 2 – An Overview of Geologic
Pressure
GEOLOGIC PRESSURE 2-1 Rev. 1.0

Table of Contents
2 Understanding Geologic Pressure ............................................................................ 2-3
2.1 Normal Pressure ............................................................................................... 2-3
2.2 Abnormal Pressure ........................................................................................... 2-3
2.2.1 Compaction .............................................................................................. 2-3
2.2.1.1 Terzaghi’s Experiments ....................................................................... 2-3
2.2.2 Other Causes of Abnormal Pressure ........................................................ 2-4
2.2.2.1 Diagenesis ............................................................................................ 2-4
2.2.2.2 Density Effects ..................................................................................... 2-4
2.2.2.3 Fluid Migration .................................................................................... 2-5
2.2.3 Overburden Calculations ......................................................................... 2-6
2.3 Subnormal Pressure ......................................................................................... 2-6
2.3.1 The Geology of Karsts ............................................................................. 2-7
2.4 Pore Pressure Prediction Methods ................................................................... 2-8
2.4.1 Pre-Well Sources ..................................................................................... 2-8
2.4.1.1 Compaction Trend Analysis ................................................................ 2-8
2.4.1.2 Eaton Exponent Method ...................................................................... 2-8
2.4.1.3 Equivalent Depth Method .................................................................... 2-9
2.4.1.4 Cross Plot Pressure Methods ............................................................... 2-9
2.4.1.5 Seismic Processing Overview ............................................................ 2-10
2.4.1.6 Dix Equations..................................................................................... 2-11
2.4.2 Prediction Methods While Drilling ........................................................ 2-11
2.4.2.1 ROP Indicators ................................................................................... 2-11
2.4.2.2 Lagging Indicators ............................................................................. 2-12
2.5 Fracture Pressure Prediction Methods ........................................................... 2-12
2.5.1 Reservoir Depletion Effects ................................................................... 2-13
2.6 Wellbore Stability .......................................................................................... 2-13
2.6.1 Principal Stresses and Well Direction.................................................... 2-14
2.7 Building Pore and Fracture Pressure Curves ................................................. 2-16
2.8 References: ..................................................................................................... 2-18

2 Understanding Geologic Pressure
This chapter is a brief overview of geologic pressure with an emphasis on abnormal
pressure. Abnormal pressure refers to formation pressure where the pressure of the fluid
in the rock pore space exceeds the pressure of a column of water to surface. References
on abnormal pressure related to drilling include Mouchet and Mitchell (1989) and Dutta
(1987). A formation pressure in excess of 9 ppg is considered abnormal for wells on the
Gulf Coast of the USA.
2.1 Normal Pressure

Normal pressure refers to formation pressure where the pressure of the fluid in the pore
space is equal to the hydrostatic column of formation water. The formation water may
contain some level of total dissolved solids (TDS) that will affect the density of the water.
Therefore abnormal pressure is somewhat of a relative term depending on the chlorides
or TDS of the formation fluid.
2.2 Abnormal Pressure

There are several causes of abnormal pressure identified in the literature, (Fertl, 1976),
including mechanical compaction disequilibrium, diagenetic, phase changes of minerals,
tectonism, aquathermal pressuring and organic metamorphism. Some of the main causes
of abnormal pressure are compaction, diagenesis, density effects and fluid migration.
These are discussed in more detail below.
2.2.1 Compaction
Compaction or mechanical compaction disequilibrium is considered the most important
cause of abnormal pressure and has received the most attention. The process is dynamic
and is caused by a rapid burial of thick sections of normally pressured shale in a marine
environment. As the sections are buried, the overburden pressure is increased, causing a
reduction in porosity and permeability. In low permeability material such as clay and
shale, the water can not be expelled from the formation at a rate fast enough to relieve the
increase in pressure due to burial. The fluid trapped in the shale ends up supporting part
of the overburden load and becomes abnormally pressured.
2.2.1.1 Terzaghi’s Experiments

Laboratory experiments by Terzaghi and Peck (1948) proposed a concept to explain the
concept of abnormal pressure based on the role of drainage in clay compaction. The
concept in equation form is:
S=σ+P
Where: S = Total overburden (or total stress)

σ = Matrix support (or effective stress)
P = Fluid pressure
The concept is shown schematically in Figure 1 where the matrix support and fluid are
represented by springs. Part of the overburden load is supported by the rock matrix. The

rest of the overburden is supported by the fluid trapped in the formation since it can not
escape. The dark horizontal sections are metal plates and the middle plate has
perforations to allow fluid to move between the upper and lower sections. The system is
sealed and when S, the overburden load is increased simulating burial, the fluid pressure
increases since it is not allowed to escape.
Figure 1 – Terzaghi’s Experiment
2.2.2 Other Causes of Abnormal Pressure

There are other possible causes of abnormal pressure including diagenesis, density effects
and fluid migration.
2.2.2.1 Diagenesis
Diagenesis consists of changes in rock chemistry with time, temperature and pressure that
can create a pressure seal or cap and possibly an increase in pressure.
2.2.2.2 Density Effects

Abnormal pressure due to density effects is caused by a column of lower density fluid
(oil or gas) that requires a higher mud weight than normal pressure at the top of the
reservoir to balance the pressure resulting from the loss in hydrostatic pressure of the gas
column. The increase in equivalent mud weight (EMW) depends on the height of the
hydrocarbon column and the density. This is illustrated in Figure 2 where the water
below the gas zone is at normal pressure of 8.3 ppg. The BHP at 9,000 ft is 4,230 psi due
to the 1,000 ft of gas column. Therefore the EMW at 9,000 ft is 9 ppg to balance the BHP
at the top of the gas zone resulting in abnormal pressure.

0 ft Surface – Ground Level
9,000 ft
Gas from 10,000 to 9,000 ft with 0.1 psi/ft gradient
BHP = 4330 – 100 = 4230 psi at 9,000 ft
EMW at 9,000 ft = 4230 psi / 9000 ft / 0.052 = 9.0 ppg

10,000 ft
Water gradient 0.433 psi/ft = 8.33 ppg
BHP = 4330 psi at 10,000 ft
Figure 2 – Density Effects Example
2.2.2.3 Fluid Migration

Another cause of abnormal pressure is fluid migration from deep to shallow formations
due to faults or a wellbore that creates a flow path. The fluid may be normal pressure at
the deeper depth but if the deep zone is in communication with a shallower zone and the
fluid is a gas with a low gradient, the upper zone can become charged with the higher
pressure. This results in abnormal pressure in the upper zone. This is illustrated in Figure
3.
Variation in heights of the surface location relative to the formation depth also has an
effect on the equivalent mud weight required to balance the BHP. This could be due to
hills or mountains for example.

7,000 ft
4500 psi
EMW = 12.4 ppg
10,000 ft
4500 psi
EMW = 8.7 ppg
Figure 3 – Fluid Migration Example
2.2.3 Overburden Calculations

The overburden pressure S can be calculated from density log data by integrating the data
down to the depth of interest. If a log is not available, overburden can be estimated using
the Eaton relationship for overburden.
S = d * (1.02 − 0.17 * E −0.000085*d )

Where:
d = true vertical depth
2.3 Subnormal Pressure

The BHP in a formation can be subnormal or less than normal pressure for several
reasons including depletion of the hydrocarbon reservoir, downthrust faults or surface
elevation effects.
An example of surface elevation effects is shown in Figure 4. The pressure is normal

since the formation outcrops at surface. The formation is deeper relative to the well
location resulting in a subnormal EMW at the well site of 6.7 ppg.
Downthrust faults can push a normally pressured formation downward while maintaining
the same BHP. The EMW based on a constant elevation surface location is now reduced
due to the increased in vertical depth.

10,000 ft
8,000 ft
Aquifer
Normal pressure due to 8,000 ft of fluid gradient;

8000 ft * 8.33 ppg * 0.052 = 3,465 psi
Subnormal pressure at the wellsite;

3,465 psi / 10,000 ft / 0.052 = 6.7 ppg EMW
Figure 4 – Subnormal Pressure Example
2.3.1 The Geology of Karsts

A special case of subnormal pressure which poses a particular problem for drilling is
karsts. Karsts are subterranean caves, or sink holes, usually found in limestone
structures. Karsts are primarily formed as a result of carbonation, whereby the limestone
is ‘weathered’ through a reaction with carbonic acid.
Figure 5 - Example of the formation of karsts.

Further karstification can occur from the bacterial degradation of oil, which produces
CO2 and H2S as byproducts. Depending on the size of the karts, encountering one while
drilling usually results in total losses as there is typically no pressure support in these
structures. Karst networks may also interrupt hydrocarbon producing sequences,
requiring that they be navigated safely to reach target depth. The combination of total
losses, presence of H2S and potential for kicks lends to the use of mud cap techniques for
wells to be drilled in karstic areas.
2.4 Pore Pressure Prediction Methods

A number of methods are available to access abnormal pressure. Porosity data is used in
several methods to estimate the depth of abnormal pressure and the magnitude of
abnormal pressure. Examples of logs that indicate porosity include sonic, resistivity (and
conductivity) and density. Seismic data is also an indicator of porosity.
2.4.1 Pre-Well Sources

Seismic and offset well information including electric log data. Various methods to
estimate pressure from log have been developed for certain areas around the world.
2.4.1.1 Compaction Trend Analysis

The general method to evaluate data for abnormal pressure is to establish a normal
pressure trend line. Deviations from the normal trend are an indication of abnormal
pressure. This is illustrated in Figure 6. The normal pressure shale is indicated by a
constant slope down to the top of the abnormal pressure section. Log data is plotted on a
semi-log scale with depth. The magnitude of the deviation from the normal trend line can
be calibrated based on actual data to get an estimate of the magnitude of pressure change.
Resistivity Conductivitiy Sonic Density Dc Exponent
Abnormal Pressure Indicator
Log Rsh Log Csh Log Dt φN dc

Ω m2/m (µ
µ mho/m) (µ
µ sec/ft) %
Figure 6 – Trend Analysis
2.4.1.2 Eaton Exponent Method

The Eaton Exponent method (Eaton 1975) uses a power law relationship for the ratio of
the observed measurement compared to the normal trend.

n
V 
PP = Pobs − (Pobs − Phyd )*  i  (1)
 Vn 
Where:
PP = Predicted pore pressure
Pobs = Overburden pressure
Phyd = Hydrostatic pressure (normal)
Vi = Interval velocity observed
Vn = Normally compacted shale velocity
n = Empirically derived exponent
2.4.1.3 Equivalent Depth Method

The Equivalent Depth Method is based on the Terzaghi equation:
S = σ + PP (2)
Where:
S = Overburden pressure
σ = Matrix pressure
PP = Predicted pressure
Assuming σ1 = σ2, an increase in S will result in an increase in pressure.
Measurement
D equivalent
Depth
Figure 7 – Equivalent Depth Method
2.4.1.4 Cross Plot Pressure Methods

An example of a resistivity cross plot is shown in Figure 8. To determine the pressure in
abnormal pressured zones, the ratio of normal resistivity to observed resistivity is

calculated. This value is entered on the X axis and the pressure is determined from the
pressure model for a given area on the Y axis.
Data from various basins has been evaluated and calibrated to indicate the magnitude of
abnormal pressure.
Resistivity Cross Plot Correlations
0.400
GOM - Hottman
GOM - Miocene
0.500 Texas - Vicksburg
Texas - Wilcox
Texas - Frio
0.600
South China Sea
0.700
0.800
0.900
1.000
0.00 1.00 2.00 3.00 4.00 5.00 6.00
Ratio of Rn/Ro
Figure 8 – Cross Plot Example
2.4.1.5 Seismic Processing Overview

The additional water trapped in the pore spaces affects the velocity of the seismic waves
in each interval.
Additional water in the interval reduces the velocity and increases the transit time in an
abnormally pressured interval.
Correlations can be used to estimate the magnitude of the pressure in the shale section.

2.4.1.6 Dix Equations
The Dix equations are used to convert two way travel time and stacking velocity to
interval velocity and interval transit time. The Dix equations are shown below:
V22t2 − V12t1 1000000 Vint (t2 − t1 )

V int = T int = Dz =
t2 − t1 Vint 2000
Where:
V1, V2 = stacking velocities (ft/sec) t1, t2 = travel time (msec)

Vint = interval velocity (ft/sec)
Tint = interval transit time (µsec/ft) Dz = zone thickness (ft)
Knowing the pulse transit time and interval velocity the thickness of each interval can be
estimated.
The depth of each interval starting at the surface can be estimated by adding the thickness
of each interval.
The interval transit time in units of time per unit depth (microseconds/ft) can be plotted
vs depth like sonic log data.
The deviation from the normal pressure trend line indicates abnormal pressure. Overlay
curves for a given basin are used to estimate pore pressure.
2.4.2 Prediction Methods While Drilling

A number of methods are used to predict pressure while drilling. The list of real time
methods includes rate of penetration (ROP), ‘d’ exponent, corrected ‘d’ exponent and
other exponents based on ROP and drilling parameters.
Some surface measurements that indicate changes in pressure are torque, drag, overpull,
hole fill and indications of flow from the formation.
Some indicators that lag the current drilling depth are mud gas, background, connection
and trip gas, mud chlorides, flow line temperature, shape of cuttings and shale density.
These indicators all depend on the drilling fluid to bring the data to surface and depend
on the depth and circulating rate.
2.4.2.1 ROP Indicators

ROP indicators can be measured in real time and interpreted to estimate a change in pore
pressure. However there are a number of factors that affect ROP including changes in
drilling parameters like bit weight and RPM, changing formation strength, bit dulling,
etc. ROP normally decreases with depth due to increased compressive strength and bit
dulling in a normal pressure shale section. The indication of increased pore pressure is a
change in the normal dulling trend to no dulling trend to an increase in ROP with the
same drilling parameters. The reason for the change in ROP is the mud overbalance is
reduced as the pore pressure increases with the same mud density.
2.4.2.2 Lagging Indicators

The lagging indicators listed above are measured and evaluated along with the ROP
indicators. Increase in connection gas from one connection to the next is an indicator of
increased pore pressure. Background gas, observation of cuttings and measurements of
changes in the mud returns are indicators that confirm the real time analysis from ROP
indicators.
2.5 Fracture Pressure Prediction Methods

A number of fracture pressure prediction models are available. The general form of the
relationship is:
F = kiσ + PP (3)
Where:
F = Fracture pressure
ki = Matrix stress coefficient
σ = Matrix pressure
PP = Pore pressure
Figure 9 shows ki values vs. depth for a number of fracture strength models.
Figure 9 – Effective Stress Coefficients

2.5.1 Reservoir Depletion Effects
The reduction in pore pressure is associated with a reduction in horizontal stress.
Consequently, the fracture closure stress (lower bound of fracture pressure) reduces,
sometimes significantly enough to affect future drilling and well design. This is a well-
recognized phenomenon, and is often discovered empirically during hydraulic fracturing
of depleted reservoirs. It is common to define a linear stress-path coefficient, the ratio of
the reduction in fracture pressure to that in pore pressure. Numerous published data
(Salz, 1977; Teufel, et. al., 1991; Kristiansen, 1998) indicate that reduction in fracture
pressure can be 50% to 80% of the magnitude by which the pore pressure reduces. For
example, Salz (1977) presents the following relationship between pore pressure and
fracture gradient as a function of depletion, in the Vicksburg formation in South Texas.
Figure 10 shows this relationship. The initial pore pressure gradient was in the range of
0.8-0.9 psi/ft, with the corresponding fracture pressure being around 1.0-1.05 psi/ft. With
around 40% depletion, when the pore pressure drops to 0.5 psi/ft, the fracture gradient
falls to 0.8 psi/ft, in this case.
Figure 10. Pore-Pressure and fracture gradient relationship in the Vicksburg Formation,
South Texas (after Salz, 1977)
2.6 Wellbore Stability

The make up of the Earth’s crust is complex and not homogeneous, as is proven by
tectonic shifting of plates and the presence of fault lines around the globe. Unlike the
equivalent stresses due to hydrostatic pressures at a given subsea depth, subterranean
geological stresses are not equal, or are anisotropic.*
*
Imagine a small cube submerged under water at a depth of 100 meters. The pressure exerted on each side
of the cube can be thought of as equal. Now imagine the same cube inserted into a vise, and pressure
applied on two of its sides. Those sides would see induced compressive stresses, while the unconstrained
faces would see a different stress state (in this case tensile). This approximates subterranean anisotropic
stresses.
2.6.1 Principal Stresses and Well Direction
If subterranean stress were isotropic, wellbore stability would be much less of an issue.
Stress anisotropy, however, greatly impacts the stability of wellbores and can ultimately
affect well placement, well direction, drilling envelopes and even mud programs.
Principal stresses are usually described in vertical and horizontal planes, as shown in
Figure 11: Principal stress states. There are three primary stress states, denoted as:
σ vert = Overburden weight acting down

σ hmax = Maximum horizontal stress applied in an azimuth direction
σ hmin = Minimum horizontal stress acting in a 90 degree offset to σ hmax.
Figure 11: Principal stress states.

In shallow or highly stressed areas, σ hmax may not be vertical. When a well is drilled
directionally, stress magnitudes must be translated to borehole alignment. This means the
maximum stress magnitude and orientation will be different than a vertical well, and
stability analyses should be conducted to determine appropriate well placement and
related mud program. Figure 12 shows the results of such an analysis, comparing the
required mud weights to maintain adequate borehole stability for three different rock
strengths. The circular graphs are read in the following manner:
The concentric circles show the equivalent deviations of the wellbore from the
vertical. The centre point is equivalent to a vertical well. The next circle is for a
30° well, the next 60°, and outer circle 90°.

The points around the circumference of the circle are relative azimuthal directions
of a wellbore. The points at which the arrows refer to in the figure show a
horizontal well placed at an azimuth of 330°.
The color gradient on each graph relates to the mud weight scale beneath each
UCS=7000 UCS=8000 UCS=9000
Figure 12 - Results of a wellbore stability analysis, showing the variation in required mud
weights for three different Ultimate Compressive Strengths (UCS) of the rock drilled.
From this example, if the rock had an Ultimate Compressive Strength (UCS) of 7,000 psi,
a mud weight of approximately 6.5 ppg would be required to drill the horizontal well.
Comparatively, if the rock had a UCS of 9,000 psi, a lower mud weight of 5.5 ppg could
be tolerated and still maintain wellbore stability.
Inclusion of wellbore stability analyses introduces another constraint into the MPD
design process. Pore pressure and frac gradient curves can be defined by many pre-well
and post well sources. If stability is an issue, the lower bound of the drilling window may
not be the pore pressure curve but the required borehole stability mud weight curve, as
shown in Figure 13.

Pressure
Pore
Pressure
Frac
Depth
Pressure
Casing Point
Wellbore stability Reduced Drilling window

curve
Figure 13 - Overlay of wellbore stability data on pore pressure / fracture gradient curves.
2.7 Building Pore and Fracture Pressure Curves

One of the first tasks for designing MPD wells is to build pore pressure and frac gradient
curves form available data. In some cases, the data may be plentiful, while in others it is
scarce. In either case, each well is different, and determination of pore pressures and frac
gradients are predictions rather than an exact science. Degrees of uncertainty will exist
when analyzing the pore and frac data, and must be taken into account during the design
stages. Figure 14 shows an example of pore pressure data from numerous sources
displayed on the same graph. The highlighted yellow region shows the maximum
uncertainty in the readings, in this case equivalent to 1.5 ppg. Discussions with the
drilling and subsurface teams help to determine how sensitive the design needs to be with
respect to pressure uncertainty.

0
500
1000 Method 1 PP
Method 1 FG
1500
Offset Mwt
2000
Design PP
2500 Design FG
Method 2 PP
3000
Method 2 FG
3500 Offset Shoe Tests
4000 Seismic PP
Realtime PP
4500
Real time FG
5000 Act Mwt
5500 Actual Shoe Test
MDT Results
6000
6500
7000
7500
Depth
8000
8500
9000
9500
10000
10500
11000
11500
12000 Maximum
12500
Uncertainty
13000
13500
14000
14500
15000
15500
16000
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 18.0
Equivalent Mwt (ppge)
Figure 14 - Building pore pressure curves from various sources results in varying degrees
of uncertainty, which must be taken into account during MPD design. In this case, the
maximum uncertainty is 1.5 ppg.

2.8 References:
Dutta, N.C., ed. 1987. Geopressure. Geophysics reprint Series. Society of
Exploration Geophysicists, Tulsa Oklahoma.
Eaton, B.A., The Equation for Geopressure Prediction from Well Logs, SPE
5544.
Fertl, W.H., 1976, Abnormal Formation Pressures. Elsevier Science Publ. Co.,
Inc., New York.
Foster, J.B. and Whalen, H.E., Estimation of Formation Pressures from Electrical
Surveys – Offshore Louisiana, Journal of Petroleum Technology, Feb. 1966.
Mouchett, J.P. and Mitchell, A., ed. 1989. Abnormal Pressures While Drilling. Elf
Aquitaine Edition Boussens.
Terzaghi, K. and Peck, R.B., 1948, Soil Mechanics in Engineering Practice, John
Wiley & Sons, Inc., New York.

Chapter 3 – MPD Variants
MPD Variants 1 Rev. 1.1

Table of Contents
3 MPD Variants ............................................................................................................. 3
3.1 Drivers for MPD .............................................................................................. 3
3.2 Technique Description ..................................................................................... 4
3.2.1 Annular Friction .......................................................................................... 5
3.2.1.1 Long Chokes (a.k.a Choking Drill Collars) ............................................ 5
3.2.1.2 Continuous Circulation System (CCS) and Continuous Circulation
Valves (CCV).......................................................................................................... 6
3.2.1.3 Concentric Casing Injection ................................................................... 7
3.2.2 Surface Pressure Methods ........................................................................... 8
3.2.3 Fluid Density Methods .............................................................................. 12
3.2.3.1 Low Head Drilling ................................................................................ 12
3.2.3.2 Dual Gradient Drilling .......................................................................... 12
3.2.3.3 CAP M .................................................................................................. 14
3.2.3.4 Mud Cap Drilling.................................................................................. 15
3.2.3.5 Floating Mud Cap Drilling ................................................................... 15
3.2.3.6 Pressurized Mud Cap Drilling .............................................................. 16
3.2.4 Delta Energy .................................................................................................... 19
3.2.3.7 Subsea Mud Lift Drilling (SMD) ......................................................... 20
3.2.3.8 Drillstring Deployed Systems ............................................................... 21

Chapter 3
3 MPD Variants
In order to determine what MPD application is suitable for a subject prospect, an
understanding of MPD drivers is essential. Keeping the primary drivers in mind will aid
an MPD well designer in evaluating and comparing the various tools and techniques
presently available. This chapter introduces these drivers and introduces the basic
variants of MPD, from applied back pressure MPD through to dual gradient and mud cap
drilling.
3.1 Drivers for MPD

The primary driver for the implementation of managed pressure drilling (MPD)
techniques is the desire to combat and mitigate non-productive time associated with
drilling through narrow pore pressure-fracture gradient margins, or severe variations in
pore or fracture pressures. Even when casing points are selected such that circulating
pressure in each hole section is within these margins, uncertainties in pore and/or fracture
pressure can still result in either a kick, or a loss, or both. Consequential problems like
stuck pipe can also arise from such situations. According to one survey1, lost circulation,
stuck pipe and kicks caused over 82% of all down time in deep wells. Clearly, the benefit
of reducing these problems, especially in high rig-rate deepwater drilling, is significant.
Since narrow pore pressure-fracture gradient margins necessitate the use of several
casing/liner points, an important benefit of MPD is its ability to extend the casing points,
thereby possibly eliminating some of these strings. Since fewer casing points are required
to reach TD, clearances between strings improves, with the resulting benefits of reduced
equivalent circulating density (ECD) and reduced losses. Furthermore, final hole size can
increase.
MPD also confers the benefits that arise from minimizing the degree of overbalance-
such as reduced invasion (and hence reduced damage) and improved rate of penetration.
These benefits can sometimes drive, or further buttress, the selection of MPD, even in
situations where the pore pressure-fracture gradient margin is not narrow and adequate
kick and trip margins are available for conventional drilling. Low overbalance is also of
benefit in highly depleted formations that are not underbalanced drilling candidates, but
still present lost circulation and stuck pipe problems when drilled conventionally.
Because MPD operations can maintain constant bottom hole pressure that is close to (but
greater than) the pore pressure, it is possible to more quickly identify and control kicks
than is the case in highly overbalanced conventional drilling.
In summary, the key drivers for MPD are the following:
Reduce non-productive time arising from kicks, lost circulation and stuck-pipe
when the well is drilled conventionally.
Minimize overbalance to
o Increase ROP
o Avoid differential sticking
1
James K. Dodson Study, www.infogulf.com, 11/2005.

Chapter 3
o Prevent lost returns

o Reduce invasive formation damage
Maintain constant BHP to avoid wellbore ballooning & wellbore instability
Extend the depth between casing setting points
o Narrow kick tolerances
o Deplete tight gas zones containing nuisance gas
Enable faster kick detection because of better flow measurements (especially
beneficial for deepwater and HPHT exploration)
Enable dynamic well control methods
It should be noted, however, that MPD is not feasible in all situations. Careful planning
and pre-engineering is required to properly assess the feasibility and risks of MPD, and to
determine whether another drilling technique is more appropriate for the given situation.
3.2 Technique Description

Balancing well objectives with design constraints will begin to direct a well designer to
methods in which to achieve the desired annular pressure profile.
BHP = SP + ∆PGravity + ∆PFriction + ∆PEnergy
The pressure balance equation can be used to categorize the tools available for each
method, as shown in Figure 1. The tools and details of equipment used are discussed in
each corresponding Chapter 5

Chapter 3
Figure 1 - Using the pressure balance equation to categorize MPD control options.
3.2.1 Annular Friction
3.2.1.1 Long Chokes (a.k.a Choking Drill Collars)

The inclusion of choking drill collars into the drill string effectively adds a “bottleneck”
in the wellbore. In doing so, the increased friction through the bottleneck imposes
addition backpressure between them and the BHA (see Figure 2). This backpressure is
only imposed while there is circulation in the well, therefore they do not provide a
constant BHP. Their original purpose was to oppose the increase in WHP brought out by
flush production from underbalanced drilling.

Chapter 3
Dynamic fluid gradient with no

collars
Dynamic gradient with collars
Depth
Increased BHP
Figure 2 - Dynamic fluid gradient affected by the inclusion of Choking Drill Collars
(CDC) in the string, acting as a long choke.
3.2.1.2 Continuous Circulation System (CCS) and Continuous

Circulation Valves (CCV)
Continuous circulation of fluid during connections may be accomplished through
specialized equipment, such as the rig floor mounted Continuous Circulation System
(CCS) and drill string installed Continuous Circulation Valves (CCV). Both systems
ensure that flow down the drill string remains uninterrupted during normal periods of
“pumps off”, such as connections (Figure 3). In doing so, a steady state flow regime is
attained in the well bore with a constant BHP being maintained.

Chapter 3
BHP
Connection
Connection
Connection
Figure 3 - BHP due to continued circulation through connections
By not exposing the well to the normal pressure surges of pumps off / pumps on, the well
bore remains more stable and issues such as connection gas, and kaleoning are avoided.
The drilling fluid can be statically underbalanced with the presence of continuous ECD
keeping the well overbalanced.
3.2.1.3 Concentric Casing Injection

An increase in friction can also be accomplished by the addition of more fluid into the
wellbore through a concentric casing annulus, as shown in Figure 4. An increase in
annular friction may be desired in cases where reservoir pressures are high, and
combinations of mud weight and ECD used to control annular pressure. Concentric
casing injection of fluid adds increased flexibility in pressure control as the rate at which
fluid is pumped can be controlled from surface. Concentric injection has the further
benefit of circulating fluid both during drilling when pumps are on and during
connections when pumps are off.

Chapter 3
Static drilling fluid gradient

Dynamic drilling fluid gradient
Static concentric fluid gradient
Dynamic concentric fluid gradient
Static drilling / dynamic concentric
fluid gradient
Pressure
Depth
Static concentric gradient

for illustration purposes
Concentric fluid continually

circulated. Reduction in
pressure due to absence of
friction from drilling fluid
Static / dynamic gradients

shifted due to addition of
heavier concentric fluid
Static / dynamic gradients
for conventional drilling
Figure 4 - Concentric casing injection of fluid. In this example, the concentric fluid is
heavier than the drilling fluid.
3.2.2 Surface Pressure Methods

Referred to in this manual as Applied Back Pressure, utilizes surface pressure controlled
through choke position to actively manage BHP. By adjusting the choke position, the
effective area for returning mud and cuttings is altered. The pressure drop occurs sharply
over this discrete area.
Since the choke position is generally variable (unless on a fixed bean choke), pressure on
the well bore can be instantly altered. This has the same effect as instantly altering the
effective mud weight. A key attribute of this method is the ability either trap pressure on
a connection or restrict the flow from a second flow source and imposed the required
pressure back on the well.
Figure 5 illustrates how ABP holds the BHP at the bit constant as ECD is lost from the
wellbore on a connection.

Chapter 3
Example –
For a well of 10,000 ft TVD, a 16 ppg mud is used for drilling. The ECD for this
mud and wellbore geometry is 0.7 ppge. In order to maintain the same BHP while
drilling when the pumps are shut down, this ECD is replaced with a surface
pressure of approximately 365 psi.
Figure 6 demonstrates a further application of APB whereby the use of a statically

overbalanced mud would see the fracture gradient exceeded half way down the proposed
hole section. By using a statically underbalanced mud, the ECD is less, and its circulating
gradient steeper. By applying back pressure at the choke either while circulating, and
most certainly, with the pumps off, the well bore pressure can be managed without
allowing an influx or fracturing the formation.
For the applications described above, the choke manifold is being used to replace either a
lack of ECD or all ECD when the pumps are off. There are several ways in which this
can be achieved. The choke either closes in to fully shut as the mud pumps are
consequently shut down, or the choke begins to throttle against a reduced flow rate
passing through it. The reduced flow rate comes from an alternate source other than the
well; normally from an auxiliary pump effectively pumping across the top of the well.
Both methods impose the additional pressure on surface to keep the well bore pressure in
the desired condition until the circulation down the well resumes.
What should be understood though, is once the chokes are fully open, they are unable to
relieve any further pressure from the well. Therefore in situations where partial losses are
experienced, and the choke is fully open, the ABP method is unable to reduce the well
bore pressure. This can only be achieved by reducing the mud weight and then
considering adjustment of the ABP as losses abate and drilling continues.

Chapter 3
Static fluid gradient

Dynamic fluid gradient
RCD
13 5/8" Pressure
Depth
Static gradient shifts from

the application of choke
pressure when pumps are
shut off
Increase in BHP is equal

to the pressure applied at
surface
Figure 5 - Example of Applied Back Pressure MPD.

Chapter 3
Surface pressure
Pressure

(Pumps Off)

Depth
(Pumps Off)
Pore
Pressure
Frac
Pressure
Casing Point
MPD window

Figure 6 - Example of a statically underbalanced fluid used in applied back pressure

MPD.
Subtle variants of applied back pressure MPD have appeared recently, along with new
descriptive names and commercial acronyms. DAPC, or Dynamic Annular Pressure
Control, is a technique employing computer controlled annular back pressure, while
CBHP (Constant Bottom Hole Pressure) and PCP (Point of Constant Pressure) are semi-
automatic techniques. Each variant has inherent benefits and must be evaluated
depending on the well characteristics.
In general, surface back pressure may be applied in the following manner:
Manually – Simply a system whereby an operator manually controls the annular

pressure by opening or closing the drilling choke valve.
Semi-Automatically – back pressure calculated by an engineer using hydraulics
software, and choke pressure altered appropriately.
Automatically – A PLC programmed with hydraulics software, connected to the
choke and back pressure pump controls the desired annular pressure
automatically.

Chapter 3
3.2.3 Fluid Density Methods
3.2.3.1 Low Head Drilling

Low head drilling is a term which refers to near balance drilling through subnormally
pressured formations. As the formation pressure is subnormal, most low head projects
require pumping of a gas phase to lower the hydrostatic pressure.
3.2.3.2 Dual Gradient Drilling

Dual gradient drilling is a variant of managed pressure drilling used in ultra-deep water,
where the pore pressure/fracture gradient margin is narrow. In conventional drilling, a
single mud column extends from the rig floor to the bottom of the hole, resulting in a
bottom hole pressure commensurate with the depth and density of the drilling fluid.
Pressure gradients are referenced to the rig floor.
In a dual gradient system, two fluids of differing densities are used to obtain the desired
bottom hole pressure. A slightly heavier mud fills the wellbore annulus from the mudline
to the well’s total depth. The second fluid, usually seawater, fills the riser annulus from
the mudline to the rig floor. The rate of the primary mud is controlled by hole cleaning
and drilling requirements. The lighter fluid rate is controlled to manage the density of
mixture in the riser annulus, and hence the bottom hole pressure.
There are two primary methods for creating a dual gradient. A subsea pumping system
may be installed on the seafloor to provide the energy to lift the mud from the wellbore
annulus back to the surface in the riser return lines, sometimes known as Mud Lift or
Riserless Mud Recovery (RMR). The riser return lines are independent of the riser
annulus itself. A comparison of Mud Lift and riser drilling against riserless drilling is
shown in Figure 7. In riserless drilling, lower annular pressures are achieved by allowing
all drilled returns to enter the sea at the sea floor. This has obvious environmental impacts
that would have to be taken in to account when choosing a drilling fluid. The alternative
method is pumping a lighter fluid down a secondary annulus that mixes with the primary
fluid at the sea floor. All gradients are referenced to the seafloor, and the margins
between fracture gradient and pore pressure are much greater while drilling the well.
Either open or closed fluid systems may be used in dual gradient drilling.

Chapter 3
Courtesy AGR
Figure 7 - Comparison between riserless and dual gradient drilling.

Chapter 3
Courtesy AGR
Figure 8 - Annular pressure and required casing scheme comparison between dual
gradient and riser drilling.
Figure 8 graphically depicts the benefits associated with dual
gradient, Mud Lift, drilling. The segregation of the drilling
fluid returns at the sea floor increases the options for heavier
fluids in the sub-sea floor annulus. This ultimately leads to
fewer casing strings and increased hole size in the pay zone.
3.2.3.3 CAP M
Continuous Annular Pressure Management (CAP M) is a
form of dual gradient drilling that employs a different fluid
pumped in the concentric annulus to impact the annular
pressure profile. CAP M is a process patented by
TransOcean. It is a dilution based technique that uses a light
weight drilling fluid to dilute the mud returns on the riser.
The mud weight in the riser is lighter than what is being
circulated down through the drill pipe. The two fluid
densities are then separated at the surface with the mud
processing equipment, which is a continuous process.
Figure 9 - Description of CAP M technique.

Chapter 3
3.2.3.4 Mud Cap Drilling

annulus at a rate sufficient to keep produced fluids from migrating to surface. This
annular fluid is referred to as the “mud cap” since it prevents migration of the drilling and
inflow fluids up the annulus to the surface. The cuttings, pumped fluids and any produced
fluids are pumped back into the formation at the loss zone. It is important to note that
this system does not try to prevent or cure loss of circulation.
Mud Cap Drilling is used in areas where severe lost circulation problems and kicks are
prevalent and it is unsafe to allow hydrocarbons to flow to surface i.e. in the event of the
presence of sour gas. The system ensures the produced hydrocarbons, injected drilling
fluid, cuttings and mud cap mud all enter the loss zone.
Mud Cap Drilling can be classified under two types listed below:
Floating (Dynamic) Mud Cap Drilling

Pressurized Mud Cap Drilling
3.2.3.5 Floating Mud Cap Drilling

Floating (or dynamic) mud cap drilling uses an annular fluid whose density is greater
than the fracture gradient of the formation, so that there is fluid loss into the formation.
The annulus fluid level is below surface and there is no annular surface pressure. The
fluid system is therefore ‘open’. This method prevents migration of produced gas to
regions above the loss zone but requires a large volume of mud and mud materials as
mud is continuously pumped down the annulus. Maintenance of a proper mud cap is by
trial and error. Migration of gas to the surface is difficult to detect and requires constant
vigilance. The method can also be expensive because of the need for weighted mud cap.

Chapter 3
Sacrificial drilling fluid (“Sac”) is

pumped down drill string
Annulus mud
weight is greater
than pore pressure,
resulting in the top
of fluid
‘somewhere’ in the
wellbore, thus
producing a floating
cap. Mud is
continually pumped
at a rate to arrest
gas migration from
surface.
Sac and all cuttings are forced

into fractures / karsts / high perm
zones
Figure 10 - Floating Mud Cap drilling.
3.2.3.6 Pressurized Mud Cap Drilling

Pressurized mud cap uses a lower density annular fluid, often referred to as Light
Annular Mud (LAM) and allows the annular injection rate to be optimized. As a result,
less fluid is lost to the formation. The annulus fluid level is at surface and annular surface
pressure is used the force the fluid into the loss zone. As there is annular surface pressure,
a rotating control head is required, and thus a closed fluid system is required. Unlike
dynamic mud cap drilling, the system allows continuous monitoring of annulus
conditions so drilling is NOT blind.
The advantage of the Pressurized Mud Cap system (in comparison with dynamic mud cap
systems) is the control of surface pressure to monitor the migration of fluids into the loss
zone. The rate that fluids are pumped down the annulus can be optimized. In addition,

Chapter 3
the fluid lost to the formation will be lighter, cheaper and is potentially less damaging to
the formation.
Sacrificial drilling fluid (“Sac”) is

pumped down drill string
Annulus is closed, positive

pressure shown on casing
RCD
13 5/8"
Light Annular Mud (LAM) fills

annular space
Sac and all cuttings are forced

into fractures / karsts / high perm
zones
Figure 11 - Initial set up for PMCD once total losses have been observed.

Chapter 3
RCD
13 5/8"
The annulus
lightens as the
wellbore deepens,
causing gas influx
if hydrocarbon
zones are present.
The gas bubbles
enter the annulus
and cause a
pressure increase at
surface. Casing
pressure is allowed
to increase to a
predetermined
value.
Sac and all cuttings continue to

enter fractures.
Figure 12 - Onset of gas migration during PMCD.

Chapter 3
RCD
13 5/8"
Once maximum
casing pressure is
seen, the annulus is
opened and LAM
pumped down the
annulus at a rate
faster than gas
bubble migration.
Once casing
pressure reduces to
designed drilling
pressures, LAM
pumps are shut off
and casing shut in
again.
Flush procedures continue when

casing pressure increases to
maximum value.
Figure 13 - Gas bubble flushing procedure.
3.2.4 Delta Energy

By adding an additional energy source into the standard equation for BHP, the BHP can
be altered. The standard equation for surface pressure is:
Surface Pressure (SP) = Bottom Hole Pressure (BHP) – ∆ Pgrav - ∆ Pfric - ∆ Paccel.
The previous section showed how manipulating the surface pressure provided the desired
BHP. MPD using delta energy adds and addition component to the above equation. This
energy comes from pumps of one form or another. Thus the equation becomes:
Surface Pressure (SP) = Bottom Hole Pressure (BHP) – ∆ Pgrav - ∆ Pfric - ∆ Paccel + ∆ Ppump

Chapter 3
For current applications of this technology, the surface pressure on the annulus side of the
well could be zero. The following are examples of MPD methods that use pumps as the
‘Delta Energy’ component.
3.2.3.7 Subsea Mud Lift Drilling (SMD)

This technology was developed to access reserves in deeper waters. The challenge has
always been the narrow pore pressure fracture gradients in the upper unconsolidated
formations directly below the sea floor. The pressure exerted by a full column of drilling
fluid in the riser limits how deep each hole can be drilled before having to case off.
SMD removes the column of mud in the riser and replaces it with a equivalent column of
sea water of considerably less weight. All returns from the well are diverted on the sea
floor using an RCD to sea water driven mud lift pump. The pump then lifts the returns to
surface for conditioning.
Figure 14 illustrates a typical pressure profile expected using the SMD equipment. The
red line shows drilling fluid in the well bore annulus bellow the sea bed at the same
gradient as the dashed black line. However, the lines is shifted to the left since the
addition pressure imposed by a full column of circulating drilling fluid has been
eliminated and replaced by the blue hydrostatic column of sea water in the riser. The red
line then diverts to the right since the pump boosts the mud from the sea floor to surface.
The assumption here is the pressure loss for drilling fluid in the riser is the same as the
return hose between the subsea lift pump and surface. This gradient can alter depending
on geometry and roughness of both the riser and return hose. Pressure at surface need
only be atmospheric to allow returns to discharge across the shakers.

Chapter 3
Figure 14 - Pressure gradient of the SMD system.
3.2.3.8 Drillstring Deployed Systems

Drill string deployed systems have been developed to reduce the hydrostatic head. Such
systems, which reduce ECD in the same manner as SMD systems described in the
previous section, use jet pump concepts to move fluid up the wellbore. Some tools
incorporate a moving wellbore (casing) seal to separate annular gradients, while others do
so by simply pumping fluid up.

Chapter 4
Chapter 4 – Applied Surface Pressure MPD
Applied Surface Pressure MPD 4-1 Rev. 3.0

Chapter 4
Table of Contents
4.1 Introduction ...................................................................................................... 4
4.2 Engineering and Well Design .......................................................................... 5
4.2.1 The Hydrostatic Effect ................................................................................ 6
4.2.2 Frictional Effect ........................................................................................... 6
4.2.3 Acceleration Effect ...................................................................................... 8
4.2.4 Pump Energy ............................................................................................... 9
4.2.5 Applied Back Pressure ................................................................................ 9
4.2.6 Controlling BHP .......................................................................................... 9
4.3 Calculating MPD Pressure Drop .................................................................... 10
4.4 Approaches to MPD Well Design .................................................................. 12
4.4.1 Define the Problem .................................................................................... 13
4.4.2 Define the Objectives ................................................................................ 13
4.4.3 Define the Design Constraints................................................................... 13
Hole Cleaning ....................................................................................................... 14
4.4.4 Identify Design Considerations ................................................................. 14
4.4.5 Introducing the Anchor Point .................................................................... 14
4.4.6 Data Gathering and Analysis ..................................................................... 16
4.4.7 Perform MPD Engineering........................................................................ 17
4.4.7.1 Procedure for determining the required Mud Weight for a single phase
MPD Operation ..................................................................................................... 20
4.4.7.2 Optimizing Mud Weights and Surface Pressures ................................. 24
4.4.7.3 Temperature Effects on MPD Design .................................................. 25
4.4.8 Specify Equipment .................................................................................... 27
4.4.8.1 Factors Of Safety In Design ................................................................. 27
4.4.9 Define Strategies ....................................................................................... 30
4.5 Design Process Layout ................................................................................... 30
4.5.1 Write the MPD Program ........................................................................... 30
4.6 Equipment ...................................................................................................... 32
4.6.1 · Rotating Control Devices ........................................................................ 32
Passive and Active Systems ...................................................................................... 33
RCD Selection Considerations ................................................................................. 34
4.6.2 Operational Considerations ....................................................................... 36
API SPEC 16RCD .................................................................................................... 37
Vendors ..................................................................................................................... 37
4.6.3 · Chokes and Choke Manifolds ................................................................. 37
Choke Lines .......................................................................................................... 39
4.6.4 Control Systems ........................................................................................ 39
Dynamic Annular Pressure Control DAPC® ....................................................... 41
Secure Drilling® .................................................................................................... 41
Halliburton GeoBalance® ..................................................................................... 42
4.6.5 Design Considerations............................................................................... 43
4.6.6 Drillstring Tools ........................................................................................ 43
Drillstring valves ................................................................................................... 43

Chapter 4
4.6.7 Tripping Equipment .................................................................................. 45

Rig Assist and Snubbing equipment ..................................................................... 45
Downhole Annular Valves .................................................................................... 46
4.6.8 Fundamentals of MPD Well Control ........................................................ 48
Pressure Indicators ................................................................................................ 49
Kick Detection ...................................................................................................... 49
Determining the Kick Intensity............................................................................. 50
Well Control Matrix .............................................................................................. 50
Conventional Well Control Techniques................................................................ 52
Dynamic Well Control .......................................................................................... 53
Specific Well Control Techniques ........................................................................ 54
Applied Surface Pressure MPD ............................................................................................................... 54

Chapter 4
4.1 Introduction
MPD Pressure
Control Options
Annular Surface Density ∆ Energy

Friction Pressure
Mechanical lift
device
Circulating Non
Long choke Circulating SMD (Mud Lift)
Continuous Manual (choke) Turbolift
circulation of fluid Semi-automatic
CCS (CBHP)
CCV Automatic (DAPC)
Concentric Mud Cap
injection
Floating
Low head (single / Pressurized
multi-phase)
CAPM
Variable density (dual
gradient)
Surface pressure methods (sometimes referred to as Applied Back Pressure - ABP) is the
keystone to all other Managed Pressure Drilling methods, as almost all variants
incorporate it in one form or another. As such, an understanding of surface pressure
methods is critical in order to fully understand the others. In this chapter, detail is
provided on engineering principles, the steps in well design, and the equipment required.

Chapter 4
4.2 Engineering and Well Design
As MPD is a departure from conventional drilling practice, a review of engineering

principles is important to highlight differences between the two types.
As stated previously, MPD is an adaptive drilling process, wherein greater flexibility in

controlling the annular pressure profile is designed into the operation by controlling one
or more of the parameters that affect BHP. The basic approach of pressure management
is best understood by looking at the pressure gradient in the annulus. From energy
balance over a control volume between the bottom hole and surface location, the BHP in
the circulating annulus of any drilling operation (see Figure 1) is given simply by
BHP = ∆Pgrav + ∆Pfric + ∆Pacc ± ∆Ppump + ABP (1)
Where:
ABP is the applied back pressure,

∆Pgrav is the gravitational or hydrostatic loss of pressure,
∆Pfric is the frictional pressure loss,
∆Pacc is the acceleration loss, and
∆Ppump is the gain or loss of pressure energy from pumps or other energy devices
in the flow path (negative sign when energy is added to the flowing fluid).
In essence then, MPD is the control or management of one or more of the above
parameters that affect the BHP.
ABP
BHP
Figure 1: Single-Phase Flow in an Annulus

Chapter 4
4.2.1 The Hydrostatic Effect

∆Pgrav is the gravitational or hydrostatic loss of pressure. In single phase incompressible
flow, ∆Pgrav is related to the density of the fluid by the simple expression
∆Pgrav = ρ gh (2)
where ρ is the density of the fluid, g is the gravitational constant, and h is the elevation
difference between the bottom hole location and the surface location. If density is
constant, this reduces to the familiar equation
∆Pgrav = 0.052 × MW × TVD (3)
where MW is the mud weight in pounds per gallon (ppg), and TVD is the true-vertical
depth (in feet) to the bottom hole location.
In reality, however, density is not constant with depth. It changes in response to both
temperature (decreasing as temperature increases) and pressure (increasing as pressure
increases). For this reason, especially in deep, HPHT wells, it is important to account for
the variation in density while calculating the gravitational effect.
In practice, gravitational effect can only be controlled by modifying the density.
4.2.2 Frictional Effect

The frictional pressure loss ∆Pfric is simply the pressure loss due to hydraulic friction. It
is sometimes referred to as AFP (Annular Frictional Pressure) in some MPD literature.
The frictional pressure loss is related to the fluid Reynolds number Re and the absolute
roughness ε (in dimensions of length). For a Newtonian fluid, the Reynolds number is
simply given by
ρvD
Re = (4)
µ
where, in consistent units.
ρ is the density of the fluid,
v is the velocity,
D is a characteristic length (the diameter in a circular conduit, The difference between
the OD and ID for an annulus, and an equivalent hydraulic diameter† in a non-
circular conduit), and
†
Equivalent hydraulic diameter is defined as
4 × Cross Sectional Area
Dh = .
Wetted Perimeter
It is commonly used to model flow in non-circular ducts. For an annulus, it reduces to the difference
between the ID of the outer pipe and the ID of the inner pipe.

Chapter 4
µ is the absolute viscosity (which is the a fluid property, and is the constant of
proportionality between the shear stress and shear rate of the fluid, by the definition
of Newtonian fluids),
The non-dimensional Reynolds number is essentially the ratio of the inertial forces to the
viscous forces in a fluid.
In Newtonian fluids, the relationship between the frictional pressure loss, Reynolds
number Re, and the roughness ε is cast in terms of a friction factor f,
16
f = , Re < 2000 (5)
Re
1 ε /d 2.51 
= −2.0 log + , Re > 4000 (6)
f  3.7 Re f 
 
In single phase flow, two flow regimes are recognized- a laminar flow regime, for Re <
2000, and a turbulent flow regime for Re > 4000 (with the region between Re = 2000 and
Re = 4000 treated as a “transition” regime). Equation (5a) above is used for Laminar
flow and Eq. (5b) for turbulent flow.
The non-dimensional friction factor f is related to the pressure drop ∆Pfric itself by
ρv 2
∆Pfric = f (7)
2 gD
Thus, the strategy is to find the Reynolds number (and a roughness) for the given flow
and geometric conditions, and depending upon the flow regime, finding the friction factor
f from Eqs. (5), and finally finding ∆Pfric from Eq. (6).
Since Eq. (5b) is implicit in f, it needs to be evaluated iteratively. Before the advent of
computers, this was clearly a cumbersome affair, leading to the development of
nomographic representations of the relationship between f, Re and ε. The most famous of
these is the Moody diagram (White, 1986), reproduced below as Figure 2. In this
diagram, for a known Re, f can be read off for a given roughness. Then using Eq. (6), the
frictional pressure loss is calculated.
The frictional pressure loss can be controlled by controlling the rate, density or geometry.
For given density and geometry, the only means of controlling frictional loss is through
rate. As Eq. (6) shows, frictional loss is proportional to the square of velocity, which is a
good thing!

Chapter 4
Figure 2: The Moody Diagram for a Single Phase, Newtonian fluid (after White, 1986)
Non-Newtonian Fluids: Many MPD fluids are non-Newtonian, that is, they exhibit a
non-linear relationship between the shear stress and shear rate (i.e., a single absolute
viscosity does not adequately represent the behavior of the fluid under a shearing stress).
The strategy to find the pressure gradient is the same as in Newtonian flows. The problem
then becomes one of finding the Reynolds number. Several models have been proposed
for this definition, from plastic viscosity-yield point fluids, to power-law fluids. In all of
these, a suitable rule for the relationship between the shear stress and shear rate is first
proposed, and the Reynolds number then based on this relationship, rather than a single
absolute viscosity. A complete description of non-Newtonian flow modeling is beyond
the scope of this document, and does not serve to illustrate the physics any better.
4.2.3 Acceleration Effect

Pressure loss due to acceleration is negligible in all single-phase systems as acceleration
is not significant in such fluids except at points of major change in cross section area. In
multi-phase flows, however, gas expansion can lead to a significant acceleration effect1.
However, it is still a minor contributor to the pressure gradient, and usually cannot be
used as a control parameter.
1
Multi-phase flow engineering is introduced and briefly discussed in the Low Head MPD chapter.

Chapter 4
4.2.4 Pump Energy

In some situations, a pump (or other energy device, such as a jet pump) can be used to
add energy to the flowing fluid, or the fluid is made to drive an energy device on its way
up the annulus. In “dual gradient” drilling, a pump or similar device is used to add
pressure energy to the fluid flowing up. In Eq. (1), the term ∆Ppump is negative when
energy is added to the fluid. In other words, a pump adds pressure energy to lift the fluid,
thus reducing the BHP. This can be used as an effective means of controlling (usually
reducing) the BHP, by lifting heavy drilling fluids without having to increase the BHP.
4.2.5 Applied Back Pressure

Finally, the applied back pressure, as the name suggests, is simply the pressure applied at
the surface through a choke. Increasing this pressure increases the BHP proportionately,
and it is therefore an excellent control parameter.
4.2.6 Controlling BHP

An important goal of all drilling is to manage the bottom hole pressure. Different drilling
techniques (and different variations of MPD) control the BHP by manipulating one or
more of the parameters affecting BHP, as shown in Eq. (1). Table 1 gives a brief list of
the different techniques and the parameters available for BHP control in that technique.
Table 1. Pressure Management In Different Drilling Techniques
Technique ∆Pgrav ∆Pfric ∆Ppump ∆Pacc ABP
Conventional Drilling X X - - ZERO

Classical MPD X X - - X
Mud Cap Drilling X X X
Dual Gradient Drilling X X X - X
In conventional overbalanced drilling, the ABP is by definition zero, and only density
and friction loss are available as control parameters. Changing density means changing
the mud weight, which takes time. Moreover, for the full impact of density change to be
felt in terms of BHP, the new density fluid has to circulate all the way to surface. This
means that in practice, BHP control through change in density is slow. Frictional pressure
loss can be changed more easily, by changing the flow rate. The main limitation of this
approach is the minor impact of frictional pressure loss on BHP in large-clearance annuli.
In tighter clearances (or if the clearance is deliberately reduced by using, for example,
large OD drill collars), frictional pressure loss can have a significant impact on BHP
control. This is particularly evident in Casing While Drilling, where very narrow
clearances between casing OD and hole ID allow for precise control of BHP solely
through change in flow rate. It should be remembered that there is a lower limit of rate
governed by hole cleaning requirements, and an upper limit dictated by the downhole
motors and equipment used.

Chapter 4
In classical MPD, where a rotating control device is used to allow applied back pressure,
greater control is now available. Density and friction are still available as control
parameters, just as in conventional drilling. Moreover, since ABP is now a control
parameter, far greater flexibility is available in the design of the operation. This is a
common MPD variation.
In Mud Cap drilling techniques, the goal is to maintain a column of fluid to surface even
though the return fluid is lost in a lost circulation zone. This is done by pumping fluid
down the annulus. A combination of fluid density, rate and surface pressure are used to
achieve a full column to surface. However, in this case, BHP is disconnected from the
surface pressure by the loss zone, and both BHP and ABP are controlled instead by the
fracture pressure in the loss zone.
In Dual-Gradient Drilling systems, a pump or similar device is used along the return
flow path to add pressure energy to the fluid. The main advantage here is that we need
not suffer the high BHP consequences of increasing mud weight. Using the pump also
reduces the stand-pipe pressure, as some of the energy that is normally provided by the
drilling pump is now being provided by the return pump. Dual gradient systems are most
attractive in deep water wells.
Several other variations of MPD have been suggested and used, and no doubt, several
more will be invented in time. Regardless of the technique however, Eq. (1) always
holds, and is the most intuitive way to see how a given MPD technique works.
4.3 Calculating MPD Pressure Drop

Calculating the pressure drop in a well designed for MPD operations is relatively straight
forward. As a review, consider the following conventional well:
1. The well is vertical, drilled to 10,000 feet with a mud weight of 9.5 pounds per
gallon (ppg). The well geometry is such that the calculated friction in the
annulus is 500 psi, while the friction in the drill pipe is 1,800 psi.
2. Given this information, it is possible to calculate the following:
• Return surface pressure (SP)
• The static bottom hole pressure (BHP), or pressure when the mud pumps
are off.
• The dynamic bottom hole pressure, or pressure when the pumps are on.
• The stand pipe pressure (SPP).

Chapter 4
Calculating SP – Remember that this is a conventionally drilled well, and by

definition, the well must have an open return line.
The surface pressure therefore is 14.7 psia, or 0 psig. SPP

Calculating Static BHP – A common method to calculate BHP using SP
common oilfield units as presented here is to use the following equation:
lb
BHP ( psi ) = 0.052 * MW ( ) * TVD ( ft )
gallon
For this example, the calculation becomes:
BHPstatic = 0.052 * 9.5 ppg * 10,000 ft = 4,940 psi
Calculating Dynamic BHP – The characteristic contributing to BHP

when the mud pumps are on is the pressure due to friction of the moving
fluid in the wellbore annulus: BHP
Annulus friction = 500 psi. Figure 3
BHPdynamic = BHPstatic + Friction = 4,940 psi + 500 psi = 5,440 psi
Calculating Stand Pipe Pressure – In order to calculate SPP, only the friction
pressures present in the annulus and the drillpipe need to be known. The pressure
due to the hydrostatic column of the drilling fluid balances on the drillipipe and
annulus side.
SPP = FPdrillpipe + FPannulus = 1,800 psi + 500 psi = 2,300 psi
Now consider the same well, drilled using MPD techniques. The well geometry,
mud weight, and friction pressures are the same, however, a rotating control device is
now used. The surface pressure is therefore not zero, but is prescribed by design. In
this example, the surface pressures are:
SP pumps on = 400 psi

SP pumps off = 900 psi
Recall that the RCD seals on the drillpipe, and a variable choke placed in the return
flowline facilitates the increased surface pressures. During drilling, the chokes may
be partially closed to attain the desired pressure. When the pumps are off (during a
connection), the chokes are fully closed, thereby trapping pressure in the wellbore.
Think of the trapped pressure as a wellbore pressure test, using the rig’s mud pumps
as the pressure test pump, pumping against the closed choke valve.

Chapter 4
Calculating Dynamic BHP – An extra term is now introduced to calculate BHP with the
pumps on:
lb
BHP ( psi ) = 0.052 * MW ( ) * TVD ( ft ) + FPannulus ( psi ) + SPpumpson ( psi ) RCD SP
gallon DPP 13 5/8"
BHP = 0.052 * 9.5 ppg * 10,000 ft + 500 psi + 400 psi = 5,840 psi
Calculating Static BHP – When the pumps are off, the friction pressure term is lost,
however, the surface pressure term is added.
BHP = 0.052 * 9.5 ppg * 10,000 ft + 0 psi + 900 psi = 5,840 psi
Note that the resultant bottom hole pressures in both cases are equal, due to the fact
that the surface pressure is increased when the friction pressure is lost.
As can be seen, calculating MPD pressures is relatively simple. Complexity is

introduced when a designer must balance all imposed constraints to find the optimum
BHP
balance of mud weights and pressures to address the drilling objectives.
Figure 4
4.4 Approaches to MPD Well Design

MPD well design is a methodical process not unlike conventional well design. It should be
noted that in MPD well design, the casing geometry becomes a constraint that the MPD design
must be bound by. There is little change to the casing design in an MPD well, unless a drill
though isolation valve is used2.
The steps followed in the MPD design process may be summarized as follows:
1. Define the Problem

2. Define the Objectives
3. Define the Design Constraints
4. Identify Design Considerations
5. Perform MPD Engineering
6. Specify Equipment
7. Define strategies
8. Design process layout
9. Write the MPD Program
The approach may be viewed as a simple, inverted triangle. The first steps are very broad, with
the focus narrowing on each subsequent step, culminating in a detailed MPD program. The first
step is to understand the problem.
2
Down Hole Isolation Valves (DHIV) are discussed later in this chapter and changes to the

Chapter 4
4.4.1 Define the Problem

The initial step for an MPD designer is to firmly understand the problems faced during
conventional drilling. These problems are what drive an Operator to consider using MPD
techniques in the first place. A misunderstanding of the challenges faced can cause a
deviation in the path of the MPD design and the resultant drilling solution.
4.4.2 Define the Objectives

Once the problems have been identified, the designer moves on to what is wished to be
accomplished by using MPD. Drivers for considering MPD were introduced in Chapter
3, and are re-iterated here. The objectives could be one or any combination of the below:
Reduce non-productive time arising from kicks, lost circulation and stuck-pipe
when the well is drilled conventionally.
Minimize overbalance to
o Increase ROP
o Avoid differential sticking
o Prevent lost returns
o Reduce invasive formation damage
Maintain constant BHP to avoid wellbore ballooning
Extend the depth between casing setting points
o Narrow kick tolerances
o Deplete tight gas zones containing nuisance gas
Enable faster kick detection because of better flow measurements (especially
beneficial for deepwater and HPHT exploration)
Enable dynamic well control methods
4.4.3 Define the Design Constraints

The key design criteria for MPD are summarized below:
The annular pressure in the open section should be greater than the pore pressure
and less than the fracture pressure profiles. Further, the annular pressure should
be adequate to satisfy borehole stability requirements.
The minimum drilling window (minimum gap between the annular pressure and
fracture pressure) should be acceptable.
The surface pressures (injection pressure and surface backpressure) should be
within the limitations of available rig equipment (pumps and rotating control
devices) while drilling, making connections and tripping.
Adequate hole cleaning should be established for each hole section. Hole cleaning
can be achieved either by rate or viscosity management.
The fluid selected should be such that it is compatible with the formation; exhibits
minimal “sag”; and can handle periodic gas influxes without breaking down.
The designer attempts to satisfy all of the above constraints while designing the
operation. When it becomes impossible to satisfy all of the constraints, the designer must

Chapter 4
investigate alternatives, or accept the compromise while procedurally minimizing the

risks arising from such compromise.
Hole Cleaning
In single-phase fluids, particularly those with additives to improve their rheological
properties, the common approach to assessment of hole cleaning is the Cuttings Transport
Ratio (CTR) Criterion. For a known (or anticipated) cuttings size, the CTR is defined as
Vsc
CTR = 1 −
Vmean
where Vsc is the cuttings slip velocity and Vmean is the mean velocity of the fluid. Slip
velocity is calculated from the known cuttings size and fluid properties.
4.4.4 Identify Design Considerations

Where design constraints bound the design by what is physically possible to achieve,
design considerations are ‘softer’ issues that further bound the design either by policy or
good engineering practice. Examples of design considerations include corporate policy
on tripping with pressure at surface to the recommended surface pressure for optimum
RCD sealing element performance.
An additional design consideration is whether the fluid selected exerts hydrostatic

pressure in excess of pore pressure without any surface backpressure applied. In some
situations, “statically overbalanced” fluids are stipulated, rather than indicated. It must
be recognized, however, that as long as continuous overbalance is procedurally assured, it
should not matter whether such overbalance comes from fluid density or applied back
pressure, or a combination thereof.
4.4.5 Introducing the Anchor Point

The Anchor Point is an important design consideration in the MPD well design process.
It is a physical depth in the wellbore at which the static and dynamic pressures are the
same. The location of the anchor point is defined by the designer.
In depleted fields where both pore pressure and formation strength has declined, there is a
need to navigate down through a narrow pressure window. By doing so, additional casing
points can be avoided. The pressure window profile determines where the anchor point
must be assigned. If the pore pressure / fracture gradients diverge with depth, the anchor
point must be set at the top of the MPD section to protect the casing shoe. If the
gradients narrow, or converge, then the anchor point should be placed at the bit. In this
case, a wider pressure margin exists at the casing shoe.

Chapter 4
Figure 5: Anchor Point selection

In problem sections where a pressure ramp exists somewhere (neither at the top or
bottom), the anchor point is set to that point.
Many factors alter the location and pressure of the anchor point including:
Mud weight
Pump rate
Applied static surface pressure
Applied dynamic surface pressure
Altering any combination of these can alter the anchor point’s location and value. The
following table summarizes how the anchor point is affected by changing these
parameters.

Chapter 4
Table 2 Impact of changes to Anchor point

Increase in… Anchor Point Depth Anchor Point Pressure
Choke Pressure
Decreases Decreases
Pumps ON
Choke Pressure
Increases Increases
Pumps OFF
Decreases Increases
Mud Weight
Pump Rate Decreases Decreases
4.4.6 Data Gathering and Analysis

Data gathering for an MPD well or campaign is a critical step in well design. Data to be
collected includes:
Any and all formation pressure information, including uncertainties

Complete current well construction information
Drilling fluid data
Rig specifications
Offset well records
Rock stability pressure
Data gathering should hopefully be able to answer the following questions:
Are there drilling related issues that would be improved or eliminated by the use
of MPD in this well or field?
Does conventional drilling cause formation damage that MPD can reduce or
eliminate?
Are any formation or reservoir-related problems likely by using MPD?
o Wellbore stability problems
The information gathered at this stage will help to answer the following questions:
Can MPD conditions be practically achieved?

o Able to achieve the desired bottom hole pressure at the desired depth
within a suitable tolerance

Chapter 4
o Able to achieve satisfactory hole cleaning while managing the BHP.

o Will it affect or impact motor and / or BHA performance
o Availability of MPD equipment to accomplish the objectives.
Can an appropriate fluid system be designed and maintained?
o Fluid / formation compatibility
o Fluid / surface system compatibility
o Foamer / defoamer compatibility if considering foam drilling.
Any special considerations that pose as obstacles’?
o High temperature
o Safety and regulatory considerations
o Logistics and equipment availability
Is the well or campaign economically viable?
Understanding as much about the geology for the prospective well is paramount.
In many instances, the efforts of MPD are forfeited since the geology did not
evolve as expected.
Appreciation of pressure sensitive formations, gas bearing shales, and large
fractures is critical to allow the elected MPD technology to perform as required. If
a formation is pressure sensitive, there may be value is holding constant BHP
across these formations. This depth then becomes the anchor point if the casing
shoe and bit depth are less pressure sensitive.
If partial losses are anticipated then mud weight will need to be selected such that,
when encountered, back pressure on an ABP choke manifold can be reduced to
reduce losses. Obviously this reduction is carried up the well, and so
understanding pore pressures above this point will be necessary. A loss zone may
also provide stuck pipe concerns, and for the similar reasons, the ability to reduce
surface back pressure and release the BHA from the well bore proves invaluable.
Motor and bit performance on offset wells should be collated to determine what
flow rate windows will likely be in effect for the well’s MPD sections.
Appreciation of bit and motor performance in conventional offset wells can then
be carried forward as benchmarks for correlation against MPD performance.
4.4.7 Perform MPD Engineering

Now that the problem has been constrained, the designer can progress with performing
initial MPD calculations. A recommended approach for performing these calculations is
as follows:

Chapter 4
Setup flow modeling software – As MPD well design is an iterative process, it is

recommended to use hydraulic flow modeling software. Although the hydrostatic
calculations are relatively easy to perform, accurate friction calculations are more
involved. Inputs to the software are:
• Well trajectory,
• Casing design,
• Fluid properties,
• Temperature profile,
• Drill string design,
• BHA design,
• Pump rate

Chapter 4
Surface pressure
Pressure

(Pumps Off)

(Pumps Off)
Pore
Pressure
Frac
Depth
Pressure
Casing Point
MPD window

Figure 6: MPD Pressure Window
Hydraulic modeling forms the core of ‘Front End’ engineering work, and is integral to
the chosen design and the technical feasibility study. During the data gathering and
analysis stage, data for pore pressure, formation strength and formation breakout strength
will have been determined. With MPD, the objective is to use the proposed well bore
geometry and BHA to conduct a series of flow modeling sensitivities to ensure wellbore
pressure remains inside the pressure window, as shown in Figure 6.
UBD projects generally use multiphase flow models to determine operating parameters
and size the surface equipment. MPD, in many cases, uses a single phase fluid, and so a
single phase model will prove adequate. The approach to single phase modeling will be
different to what the drilling engineer is used to. The software needs to have the
functionality of being able to run simulations with a fixed surface pressure and in some
cases, be able to fix a pressure in the well bore (i.e. casing shoe). Consideration also
needs to be given to pressure loss in the surface lines coming out of the well. Pressure is
normally being monitored at a choke manifold, with hose or hard pipe linking it to the
BOP’s. Considerable pressure loss can occur in these lines depending on the pipe’s
internal diameter and deviated flow path.
The objective of the design is to optimize the combination of mud density and pump rate.
A lower fluid density results in a steeper pressure curve and smaller degree of

Chapter 4
overbalance at bottom. A lower fluid density also results in a higher pump pressure that
in turn results in lower pump rates, and lower frictional pressure loss in the wellbore. In
addition, the lower density mud results in higher annular surface pressure to maintain
constant bottom holes pressures, particularly during connections and trips.
4.4.7.1 Procedure for determining the required Mud Weight for a single
phase MPD Operation
1. Setup flow modeling software – (from Chapter 4, section 4.4.6 Perform MPD
Engineering)…
2. Select the Anchor Point – Identify the bottom hole set point for the design.
Set anchor point at top of the section if PP-FP window diverges with depth
Set anchor point at the bit if PP-FP window converges with depth
If there is a regression on PP or FP, creating a narrower window at any point
along the planned open hole section, set anchor point at the depth of the smaller
difference between PP and FP
3. Select the target Annular Pressure at the Anchor Point depth (Pann(AP)) – It
usually will be just slightly above the PP and FP. However, the decision might be
affected by other considerations, generally identified in accordance with the general
strategy of the well. Some of the considerations are:
Factors that move the target Pann(AP) towards the FP:

• Possibility of toxic gases associated with a potential influx;
• tight regulatory/environmental policies regarding flow from the reservoir;
and/or
Factors that move the target Pann(AP) towards the PP:
• High risk/cost of losses (total losses and/or too expensive mud);
• desire to minimize overbalance to reduce formation damage;
• plan to perform some type of pore pressure determination during the operation
and/or;
• surface set up prepared to handle inflow from the reservoir.
• Propensity for differential stickers
• Hard rock is low drilling
4. Select the target dynamic well head pressure WHPdyn – This point can be
calculated by setting a point just above pore pressure at the previous shoe and just
below fracture gradient at section TD draw line though the X axis.
5. Verify surface pressure is within design criteria - There are different criteria
regarding the design of the dynamic well head pressure in a MPD operation. Factors
such as the general strategy of the whole operation, pressure uncertainty, or
equipment availability, may the design. Following is some discussion regarding this:
General criteria should be to design for WHPdyn to be as low as possible, while
keeping a desirable working margin which allows reduction of the BHP –by

Chapter 4
reducing WHP– below the expected pore pressure at any point along the open
hole section (for a statically underbalanced design), or to a pressure enough to
stop or reduce the potential fluid losses (for a statically overbalanced design).
A second criteria would be to use the common practice among some MPD
designers, which is to design for a maximum WHPdyn of psi, as the wear in the
sealing elements of the RCD is said to increase abruptly above this limit.
A third approach, when there is any preliminarily selected equipment (based on
availability, previous agreements, etc), will depend on maximum equipment
rating and recommended working ranges from the vendor. If no recommended
value is available from the vendor, a good practice is to design the WHPdyn to a
maximum of 50% of the dynamic pressure rating, at the planned RPM3.
In general, whichever is the approach used, the design WHPdyn should not exceed
70% of the RCD dynamic rating.
6. Determine the Mud Weight required – Choosing the mud weight for MPD
operations is an iterative process, and has traditionally involved guess work for the
first choice. An engineered, methodical approach is presented to remove as much of
the guess work as possible. The approach is predicated on an initial over estimation of
the desired mud weight and iterating towards the final solution (see Figure 7):
As a starting point, calculate an equivalent mud weight (EMW), so that its

gradient, plus the WHPdyn, equals the target Pann(AP).
EMW =
(PAnn ( AP ) − WHPdyn )
(8)
0.052 × TVD AP
This initial mud weight is overestimated because it includes the friction pressure
losses; but this will be adjusted during the iteration process.
Run the simulation model with the calculated EMW, and find the annular friction
pressure from surface to the anchor point depth (∆Pf(1)). Add this ∆Pf(1) to the
WHPdyn, to obtain a static well head pressure (WHPst(1)).
WHPst (1) = WHPdyn + ∆Pf (1) (9)
This initial wellhead pressure is overestimated, as the friction is calculated with

an overestimated mud weight. This will be adjusted during the iteration process.
Now, obtain a new mud weigh (MW(1)), so that its gradient, plus the WHPst(1),
equals the target Pann(AP).
3
Please see Pressure ratings and factors affecting the life of the sealing elements on the RCD, in
SectionXX

Chapter 4
MW(1) =
(P
Ann ( AP ) − WHPst (1) )
(10)
0.052 × TVD AP
Perform a second run of the simulation model with the calculated MW(1), to find a
second friction pressure from surface to the anchor point depth (∆Pf(2)). Add this
∆Pf(2) to the WHPdyn, to obtain an adjusted static well head pressure (WHPst(2)).
WHPst ( 2 ) = WHPdyn + ∆Pf ( 2 ) (11)
Calculate the new adjusted mud weight MW(2), as per the previous calculation
using the latest WHPst calculated. To check if it is necessary to iterate again,
compare the last mud weigh to the previous one. If the difference between the two
of them is less than what is practically manageable in field conditions for mud
density, the latest value is good enough for the design. This would mean, in field
units, a difference of less than 0.05 ppg.
MW( 2 ) − MW(1) ≤ 0.05 ppg (12)

Chapter 4
(P Ann ( AP ) − WHP dyn )

EMW =
0 . 052 × TVD AP
WHPst (1) = WHPdyn + ∆Pf (1)
MW(1) =
(P Ann ( AP ) − WHPst (1) )
0.052 × TVDAP
WHPst ( 2 ) = WHPdyn + ∆Pf ( 2 )
MW( 2 ) =
(P
Ann ( AP ) − WHPst ( 2 ) )
0.052 × TVD
MW( 2 ) − MW(1) ≤ 0.05 ppg
Figure 7: Process to determine required Mud Weight for a MPD operation.
7. Verify all the parameters – Make sure that the calculated WHPst does not exceed the
80% of the dynamic rating or the 70% of the static rating of the RCD. Also, plot the
static and dynamic pressure gradients and compare with the PP-FP windows, to
confirm that they are within the operational envelope along the open hole section.
Depending on the degree of uncertainty in the pore and fracture pressure profiles,
sensitivity analyses may have to be performed. Performing the same steps as outlined

Chapter 4
above for P10, P50, and P90 cases for pressures will result in the ‘worst case’ scenario
and a robust well design.
4.4.7.2 Optimizing Mud Weights and Surface Pressures

By modelling lighter mud and differing flow rates, a set of surface pressures and fluid
gradients will be collated. Figure 8 shows the range of bottom hole pressures covered by
using 11.3 ppg, 12.8 ppg, 14.3 ppg and 15.5 ppg mud weights. The boundaries in each of
the cases above assume a natural back pressure of 100 psi in the MPD system and a
maximum of 1,000 psi surface pressure while the mud pumps are on. The wellhead
conditions for each of the cases are the same.
Figure 8: Four mud weight operating ranges within a PP-FG window

The selected mud properties and flow rates must then be used in hole cleaning analysis to
ensure the desired mud weights and flow rates are capable of cleaning the hole clean all
the way to TD. If not, then either the mud weight, mud properties or flow rates will need
to be adjusted.
If hole cleaning is adequate, the key pressure points on the system must be examined.
With a lighter fluid, stand pipe pressure will be higher and may exceed the capabilities of
either the pumps or standpipe. It is not advisable to continuously pump at higher
pressures or rates due to the excessive wear on equipment. Lighter fluids in the annulus
mean higher surface pressures beneath the RCD and in surface lines up to the choke.
Higher pressures affect the longevity of the RCD stripper rubbers, and so a high
frequency of elements changes and additional cost. It is prudent to try and limit, where
possible, surface pressures below the rig floor purely from a safety standpoint.

Chapter 4
The designed flow rates must also suit to the operating window of any down hole motor
or turbine being used. Motor selection is based on drilling performance, but the scenario
that must be avoided is requiring a high flow rate through motor that subsequently
imposes a greater friction in the annulus. The result being the choke is 100% open and
unable to maintain the required pressure at the pivot point.
With MWD tools, simulation or testing should be conducted to ensure pumps can be
brought up and down at suitable rates that both obtain a survey and also maintain the
desired constant pressure at the pivot point. If a tight control of pressure the pivot point
is required, it may be necessary to deploy a wired telemetry system in the BHA or
consider running a casing gauge in the previous casing.
A casing gauge would stream a continuous pressure to the MPD control system or
Operator. The telemetry system provides a continuous signal to the MPD control system
or Operator all the time the pumps are running.
4.4.7.3 Temperature Effects on MPD Design

Temperature plays an important role in flow behavior. The properties and PVT behavior
are dependent upon the temperature (and pressure), and thus, the final pressure
calculations are a function of temperature. The temperature and pressure are coupled, as
the complete energy equation includes pressure terms as well as heat transfer terms. In
general, thermal problems in an MPD operation can be physically understood from
Figure 9. Cold fluid enters the drill pipe, and returns through the annulus, along with any
formation fluid. The formation fluid enters the wellbore at the bottom hole temperature at
the point of entry, and mixes with the injected fluid returning in the annulus. Heat is
transferred between the fluid in the annulus, the injection fluid in the drill pipe, and the
formation. In general, the formation tends to heat the injected fluids- i.e., the formation is
the source of heat, and the fluids absorb the heat. The intervening casings, cement, and
drill pipe, as well as the fluids themselves, act as resistances in this heat transfer.

Chapter 4
Figure 9: Wellbore heat transfer during MPD

If the wells thermal profiles are unknown, a thermal analysis should be performed using
software such as WELLCAT’s cyclical pressure and heat transfer computation with
complex heat transfer. In this program, thermal analysis is performed using a finite
difference formulation of the energy equation, and imposes a 2-dimensional grid on the
problem space. The energy equation in finite difference form is satisfied at each grid
point, for each time step. The overall heat transfer coefficient is defined at some radial
location in the wellbore (usually drill pipe OD), and accounts for the resistance offered to
heat flow because of all the components between the centerline of the wellbore and the
outer boundary of the wellbore (boundary between wellbore and formation). Then, the
heat transfer can be modeled as:
q U A T OD DP = Σ∆ , (13)
In this equation, q is the heat transfer rate, U is the overall heat transfer coefficient, A is
the area available for heat transfer (surface area of drill pipe, since U is defined at OD of
drill pipe), and ∆T is the temperature difference. Since fluid properties depend upon
temperature and pressure, and since heat transfer is affected by the fluid properties, the
problem is solved iteratively until both pressure gradient and temperature are consistent.
Multiphase flow and PVT correlations are used to solve the pressure problem. The
solution steps through time (and space), and finds the temperature and pressure at the end
of specified time. If the temperature and pressure are inconsistent, the process is repeated
with an updated guess of temperature, until convergence.

Chapter 4
4.4.8 Specify Equipment

Once a design has been completed, the specification of the required equipment can be
developed. The equipment includes, but is not limited to:
a. RCD specs
b. MPD choke specs (rating, auto, semi-auto, manual)
c. Line sizing
d. Separator sizing / mud gas sep. calcs (if required)
e. Snubbing equipment
f. BHA (NRV’s)
g. Additional equipment
4.4.8.1 Factors Of Safety In Design

In practical design, it is necessary to specify equipment with adequate safety margin.
Margins should be applied to the pressure, volume, and rates specified for each piece of
equipment. The margin also depends upon the safety criticality and the likely uncertainty
in the design ranges that result from the modeling. Summarized below are the safety
margins and levels of redundancy we recommend for different parameters and equipment
components in an MPD operation.
Error! Reference source not found. lists the recommended margins for the injection
equipment. In addition to the safety margin in the pressure rating of the equipment, the
volume requirements must also be specified with a margin of safety. In addition,
redundant equipment should at least be identified and made easily accessible (if not
physically on location, at least within timely reach) for equipment susceptible to
breakdown such as pumps and compressors.
Table 3 Recommended Safety Margins, Injection Equipment
Parameter/ Safety Margin / Comments

Equipment Redundancy
Pressure 1.25 times Max Injection Max Injection Pressure based on
Pressure minimum fluid density
Pumps Unit redundancy On site, or immediate availability
Compressors
Pump rate 1.25 times max pump rate
Table 4 lists the safety margins recommended for the downhole equipment. However, in
specifying the motor, a margin of safety should be imposed on specified temperature and
volume rating. It is prudent to use the static bottom hole temperature (BHT) in the
specification of motor or MWD equipment temperature ratings, since the BHT could be
reached in MPD during prolonged periods of very low or no flow (dynamic flow checks).

Chapter 4
Table 4 - Recommended Safety Margins, Downhole Equipment

MWD – 1.1 times Static BHT Likely during connections,
temperature Unit redundancy dynamic flow checks and trips
rating
Motor – 1.1 times Static BHT Likely during connections,
temperature Unit redundancy dynamic flow checks and trips
rating
Error! Reference source not found. lists the recommended safety margins for the
safety-critical pressure control equipment. Notice that for both chokes and Rotating
Control Head margins, the design basis and margins of safety may not always be
achievable.
Table 5-Recommended Safety Margins, Surface Pressure Control Equipment

Rotating Static: 1.1 times MASP MASP gas to surface may be
Control head - Dynamic: 1.5 times Max difficult to obtain
Ratings Back pressure during
operations
Temperature: 1.25 times
maximum anticipated
surface temperature
Rotating Project specific Redundancy is generally required
Control head – of wear-sensitive parts such as
redundancy elastomers. Unit redundancy is
not common, but may be
desirable for long MPD
operations in remote locations.
This could include a spare
bearing assembly.
Choke Based on MASP from May be difficult. Alternative
Manifold - reservoir at its highest designs for safety (high pressure
Pressure expected pressure. separation system).
Rating
Choke – Alternative leg redundancy Almost all MPD chokes give a
Redundancy alternative leg redundancy
BOP Based on MASP from Usually Rig BOP already
Equipment reservoir designed to this at higher
Reservoir Pressure

Chapter 4
Table 6 - Recommended Safety Margins, Surface Returns Handling Equipment

Piping 1.25 times MASP upstream Based on piping friction loss
of separator analysis
1.25 times working pressure Sizing based on minimizing
downstream of separator working pressure (or system back
pressure)
Pumps 1.25 times max working Max working pressure based on
pressure if re-injecting pressure loss and receiving system
Unit redundancy pressure
1.5 times anticipated
maximum flow rates
Solids handling 1.25 times solids rate at
highest ROP
Table 6 lists the recommended margins of safety for surface returns handling equipment

Chapter 4
Figure 10 – Typical MPD P&ID
HOLE FILL
(to be disconnected
during MPD Ops)
4.4.9 Define Strategies

Three MPD operations which differ significantly from conventional drilling are the act of
drilling itself, tripping, and well control. As MPD is a new technology for most
Operators, policies will probably not exist covering them. As such, strategy documents
should be developed to bridge the gap between what has been already accepted and MPD
operations. The strategy documents are high level, and set the guidelines for how the
operations will be carried out. These documents will form the reference documents for
detailed procedures and on site written work instructions.
4.5 Design Process Layout

The Process and Instrumentation Diagram (P&ID) is a critical document in an MPD
operation. The P&ID serves as the blueprint for the MPD rig up and details every line,
valve and sensor that will be added to the rig. This diagram will be used by the MPD
service provider during rig up, and should be audited during the commissioning phase.
4.5.1 Write the MPD Program

The final step is to collate all of the information from the well design into the drilling
program for the MPD section. Based upon the strategy documents and the P&ID,
detailed procedures are included as part of the MPD program. It is important to note that
the MPD procedures will include sufficient detail, identify each valve by unique name
and its position (open or closed) during each operation. The following Error! Reference
source not found. shows an example of just one MPD procedure.

Chapter 4
Table 7 – Example – MPD procedure

Step Action Responsible party
1 Space out a tool joint above the rig floor Driller
2 Turn of the mud pumps Driller
Shut the well in at the MPD choke and isolate the
choke.
a. WFD
3 Ensure the choke does NOT open
Close FLM6 or FLM9 to ensure a positive seal
b. Establish pressures ALL
a. Divert flow to rig choke manifold
b. Open Valve BOP7 (HCR) Driller
c. Close FL1 WFD
4 Open FLC1 or FLC2 to bleed off any residual
pressure to zero in primary flow line.
d. WFD
Monitor any pressure build up on the data header
to ensure FL1 is holding.
a. Read and record the stabilized SIDPP**, SICP. ALL
Continue to rotate the string and reciprocate
b. Driller
5 slowly (do not pull a tool joint through the RCH).
Complete required mud density calculations as
c. ALL
per conventional well control kill sheets
Circulate kick out of the hole through rig choke
6 manifold, maintaining the new constant SPP. Driller
Once well is dead with new calculated back

a. WFD
pressure, divert flow back to MPD choke
7 b. Open FL1 WFD
c. Close BOP7 Driller
8 Drill ahead Driller

Chapter 4
4.6 Equipment
4.6.1 · Rotating Control Devices
Rotating Control Device (RCD) is considered as the main enabling piece of equipment
for MPD operations, and is normally placed above the conventional BOP stack for most
MPD operations. They are also known as Rotating Flow Diverter
(RFD), Rotating Control Head (RCH), Rotating Flow Head (RFH),
Rotating BOP (RBOP). The function of the RCD is to provide a seal
between the wellbore and the atmosphere, while allowing the pipe
to move (up/down and rotate) and diverting the returns flow from
the well to a contained system. For many MPD operations, the
RCD in conjunction with the flow control choke are part of the
primary barrier for well control. This is the case when the
operation is planned to be statically underbalanced. This is
discussed further in the Well Control Section of this manual.
The rotating control head is typically mounted on top of the rig’s

existing annular preventer. Therefore, it is imperative that
allowance is made on the substructure below the rig floor to
accommodate this additional equipment. The amount of spacing
required is dependent upon the selected model of rotating
control head.
There are several vendors that offer RCDs, including major

service companies, as well as smaller companies. However, almost
all rotating control devices work off of the same basic design. They
all have a packing element that forms the seal around the drill pipe.
These packing elements are expendable elements that must be
replaced as they wear. For the majority of heads in use today, the
packing element rides on a bearing system that allows the
packer to rotate with the pipe. There are still several brands of Figure 11: Exploded view of a
rotating diverters for which the packer element does not rotate Weatherford Rotating Control Device
with the pipe. An example of this type of equipment is the
Weatherford Dual Annular Preventer that utilizes two non-rotating preventers. Another
example is the Washington Series 1500, specifically designed for geothermal drilling
applications requiring the stripping of multiple diameter drill strings. The upper rubber
rotates with the drill string and seals the Kelly and drill pipe, while the lower rubber is
stationary and is used to seal on larger diameter drill string components (drill collars).

Chapter 4
Passive and Active Systems
There are two basic types of rotating control heads,

passive and active. The passive devices rely on a force
fit of the packing element around the drill pipe. Passive
units are designed to be pressure energized by the well
pressure, which means the higher the well pressure, the
tighter the seal. Figure 13 depicts the Weatherford
Passive RCD.
As shown, the stripper rubber is the sealing element

around the drill pipe. The bearing assembly allows the
rubber to rotate with the drill pipe, minimizing wear.
The use of a Kelly and Kelly driver creates positive
movement of the stripper rubber assembly.
Some passive systems have a hydraulic unit with the

package. For these systems the power pack does not
supply operating pressure for the seal system. It is for
supplying lubrication for the bearing assembly.
Additionally, the hydraulic unit can supply opening and
closing pressure for the bearing assembly
locking mechanism and refrigerating
fluid for units fitted with such
systems, respectively. Some
passive systems are much
simpler in design, and don’t use
any additional equipment. These
Figure 12: An example of a passive are typically the lower pressure
Weatherford RCD unit. RCDs.
Active rotating heads rely on hydraulic pressure to energize the

packing element around the drill pipe. As the element wears or the
annular pressure rises, the hydraulic pressure can be increased in
order to increase the sealing pressure.
Active RCD assemblies have a hydraulic control unit and control

panel, which controls the closing force on the stripper Figure 13: An example of an active
rubber. The accumulator system used to control the RCD RCD unit.
should be independent of the rig’s unit. Figure 14, depicts
a cutout of the Northland RBOP, while Figure 14 depicts the cutout of the Shaffer
PCWD®. Only the PCWD® is certified as an API Annular Preventer.

Chapter 4
Figure 14: The Shaffer active RCD, which is the only RCD to receive an API monogram as a recognized
blowout preventer.
RCD Selection Considerations

Because the rotating control head acts as one of the primary barriers to well pressure, the
design has critical safety implications. Generally rotating heads have three important
specifications:
1. All RCDs have pressure ratings for static and rotating conditions. The static
pressure rating is the maximum pressure the equipment is designed to control with
no pipe movement. The dynamic pressure rating includes rotation of the drill
string and packing element, while drilling. The dynamic pressure rating in some
cases reported by the manufacturer at various rotating velocities, decreasing the
rating as the rotation increases. Figure 15 shows an example of dynamic pressure
ratings versus rotational speed, in this case for the W-7800. The dynamic rating
should be based on the maximum anticipated surface pressure during drilling
operations. Static pressure rating should be based on the maximum anticipated
surface pressure.

Chapter 4
Figure 15: An Operating envelope for one model of an RCD. Operating envelopes will vary from model
to model.
2. Temperature rating, or the maximum allowable temperature of the returning fluid,
which is often a function of the elastomeric components used in the manufacture
of the rotating control head. The temperature rating should be based on the
maximum expected temperature of the surface returns, which is obtained by
performing a thermal analysis. In the absence of thermal analysis, the temperature
rating should consider the production return temperatures, since they represent the
highest returns temperature possible, in the event of well inflow.
3. Sealing elements material should be selected accordingly to the anticipated fluids

composition present in the operation. Although main fluids will be the injected
ones (drilling fluid or sacrificial fluid in case of mud cap operations), there is
possibility in MPD operations to have formation inflow during limited periods of
time. This inflow will be different in composition and temperature than the
drilling fluid, and should be taken in account for material selection.
The maximum opening of the rotating control device should also be a consideration in
selecting the equipment. The maximum OD of all tools that will be run should be
considered, including wear bushings, packers and tubing hangers that will be run during
the completion. Having to rig down the RCD prior to running them will increase the rig
time and compromise the ability to remain hydrostatically underbalanced, if this is the
case of the operation.
Many RCDs are not designed to have a column of fluid above the packing element. Fluid
above the packing element can cause damage to the top seal therefore most RCDs do not
have a bolt pattern above them. This will be an issue if plans call for the use of a
snubbing unit.

Chapter 4
Many RCD systems have an assembly that replaces the packing element during various
operations. These include tripping sleeves or snubbing assemblies that will protect the top
seal and allow tripping and logging operations.
4.6.2 Operational Considerations
Increasing the life cycle of rotating devices and the elastomeric seals are major
consideration when designing an efficient MPD operation. Well planners should make
every attempt to eliminate potential problems areas for rotating devices. MPD experience
thus far has revealed an enormous cost to operators when preparations were not made to
protect the rotating device and the sealing systems. The tangible cost for downtime,
mobilization, and the risk of safety problems warrant that special attention is given to the
following:
Ensure alignment of rig over well bore during rig move operations. Misalignment of
the rig to the wellbore creates an eccentric wear and side loading that drastically
shortens the life cycle of elastomers and bearing components.
Once a rig is over the well, ensure the alignment of the stack is perpendicular to the
floor. In other words, ensure correct stack alignment.
Ensure fluid compatibility with elastomeric products.
Ensure operating temperatures do not exceed manufacturer’s specifications.
Ensure all tubular goods that are to be run into the well are smooth and free of heavy
surface corrosion. Pitted pipe, nicks and tong marks will severely affect the service
life of rubber goods, especially when used under high pressure.
Check hard or soft banding on proposed string for burrs and poor installation. If
possible, send to a machine shop and have them turned down to smooth (do not
remove) banding material.
Use manufacturers’ recommended lubricating procedure for elastomers in use. Many

common rig recipes of stripping lubricant such as pipe dope mixed with gear oil can
cause more harm than good. In many cases water is the best material for lubrication.
Ensure that operating limits are observed with regard to pressure, rotating speed and
temperature.
When tripping or stripping through a rotating device with nitrogen flowing across the
wellhead in order to maintain pressure, consider using a mist pump in the nitrogen
flow stream to provide lubrication.

Chapter 4
Consider use of pipe with no API grade identification grooves. Fill in the grooves
with soft banding material and grind smooth.
Sealing element life will be affected by wellbore pressure, especially during tripping.
API SPEC 16RCD

In February 2005, The American Petroleum Institute API issued the API SPEC 16RCD
Specification for Drill Through Equipment Rotating Control Devices, in an effort to
regulate and standardize the design, construction and factory testing of the RCD
equipment across the industry. It does not provide standards or specifications for field use
or field testing.
Most equipment currently in the market has been designed and built before the API
SPEC 16RCD issue date, and does not conform to the standards. At the moment of issue
of this manual only the RFH 5000® from Optimal Pressure Drilling Services is designed,
built and factory tested as per API specs for RCD.
Vendors
The following is a list of the rotating control heads that are currently on the market:
Table 8
Rotating Control Head Description
Weatherford
• 7000/7100/7800/8000/9000 Passive
• RPM 3000 Active
Northland RBOP (Weatherford) Active

Shaffer PCWD Active
Strata Energy Passive
Optimal Pressure Drilling Passive
Smith Passive
Hydril Passive
Pruitt Passive
Washington Passive
Diamond Passive
4.6.3 · Chokes and Choke Manifolds

The primary purpose for the flow control choke system is to provide a controllable flow
path through an adjustable choke to maintain the bottom hole pressure required for
drilling or well control operations. As such, it is vital for pressure control and is
considered part of the primary barrier for well control in many MPD operations.
It is important to emphasize that use of rig well control choke is not acceptable for MPD
operations because the choke is subject to erosion. Also, a rig well control choke does not

Chapter 4
have a sufficient internal diameter to be used as a flow control choke. The smaller ID of a
well control choke will lead to excessive backpressure, plugging and erosion. Emergency
well control equipment should never be used for routine operations.
All chokes should be equipped with tungsten carbide orifice assemblies. Wear sleeves or
flow couplings can be installed downstream of each choke to minimize erosion due to
turbulence.
The main difference between UBD/MPD flow control chokes and well control chokes is
the maximum opening through the choke. Most well control operations are done at slow
pump rates. MPD flow control chokes must be designed for higher flow rates and
continuous operations in a harsh environment. To minimize turbulence downstream of
the choke (which leads to high rates of erosion and high pressure drop across the choke,
causing back pressure upstream of the choke) larger diameter flow control chokes are
desired. Some flow control chokes are actually hydraulic rubber sleeves. This type of
system is cheaper, easier to replace and more resistant to erosion. Other systems include
globe valves or large diameter adjustable chokes. Positive bean chokes should not be
permitted for drilling or MPD operations.
Two types of adjustable chokes are used for choke manifolds - hand operated and remote
controlled hydraulic chokes. Many UBD/MPD choke manifolds include one manual and
one remote controlled choke. In the group of remote controlled chokes, some have been
designed to be automatically controlled through Programmable Logic Control (PLC). The
operator can pre-set an upstream pressure, and the choke will adjust its opening if flow
variation occurs, to maintain the set pressure (see Figure 17).
These automated chokes can be a great resource in MPD operations, allowing

maintenance of constant backpressure, and thus, a reasonable constant Bottom Hole
Pressure (BHP). However, this feature has to be carefully monitored if a fluid
composition change occurs in the wellbore. (for example a gas kick). In this case, the
appropriate reaction would be to reduce the choke opening to increase the backpressure,
and thus increase the BHP (to regain the overbalanced condition) and to compensate for
the loss of hydrostatic in the wellbore. But the reaction of the automated choke will be to
increase the opening, to compensate the increase in pressure caused by the additional
flow. This reaction will cause a drop in BHP, then allowing more inflow into the
wellbore.
Some vendors offer automated MPD choke systems, built in special application
packages, Weatherfords such as AtBalance’s Dynamic Annular Pressure Control
DAPC®; Halliburton’s Geobalance®; or Secure Drilling®. These choke systems control
the bottom hole pressure, pressure upstream of the choke, or drill pipe pressure by
opening or closing the choke. They are discussed in depth later in this chapter.

Chapter 4
Choke Lines
The sizing of the lines, both upstream and downstream of the choke, should be based as
per API RP 14E. However, these recommended practices are known to be conservative,
as they cover permanent installation of offshore
production platform piping systems. Being MPD systems
are temporary installations, design could be more
tolerant if a periodical inspection program is in
place, and if wall thickness monitoring on site is
possible during operation.
4.6.4 Control Systems
Figure 16: Schematic of MI-Swaco Super Autochoke
Figure 17: P&ID of an automated MPD control system, courtesy of @balance.

The control system is at the core of any MPD equipment setup, whether it is as simple as
a RCD and a choke manifold, or as complex as these items in addition to a back pressure
pump, flow meter and software control algorithm (shown here). The type of equipment
layout depends on many factors, with the simpler systems used in a more benign and
forgiving drilling environment and the more complex ones in higher risk wells. The
systems were introduced in Chapter 3, and may be categorized as follows:
Manually – a system whereby an operator manually controls the annular pressure

by opening or closing the drilling choke valve.

Chapter 4
Semi-Automatically – back pressure calculated by an engineer using hydraulics

software, and the choke automatically adjusts to a pre-set surface pressure.
Automatically – A PLC programmed with hydraulics software, connected to the
choke and back pressure pump controls the desired annular pressure
automatically.

Chapter 4
Dynamic Annular Pressure Control DAPC®

The heart of this system is a flow control choke tied to a recirculation pump, both
controlled by an automated computerized system. During drill pipe connections the pump
takes fluid from the suction tanks, and discharges upstream the choke. Both the pump and
the choke are controlled by the
automated system, which uses
a combination of flow rates and
choke openings to create the
required additional
backpressure required to
maintain a constant bottom
hole pressure. The system
compensates for the loss of
annular friction, replacing it for
friction through the choke
restriction.
The DAPC system (Figure 19)

is tied to the real time annular
bottom hole pressure from a
MWD tool. When pressure
Figure 18: Halliburton DAPC automated back pressure pump
reading from BHA is not
available, the model ties its
response to the results of flow
simulation software, which
runs in the background all the time, and calibrated with the PWD information when
available. The DAPC system is
designed to work only with
single phase or mulitphase
systems.
Secure Drilling®
The Secure system is available in two
options, Standard and Special. The
standard version is a kick detection and
control system, designed for
hydrostatically overbalanced
operations. When a kick is detected, the
system can be set to automatically
circulate the influx out, based on
standpipe pressure (Figure 20).
The special version is designed for

hydrostatically underbalanced MPD
Figure 19: Secure drilling choke control manifold. The mass flow
meter (silver U shaped object) is included in this skid

Chapter 4
operations, where the choke is automatically controlled to maintain a preset backpressure.

If an influx is detected, the system alerts the operator, who can decide to engage the
automatic set up for circulate the influx out, and then resume operation with a different
preset backpressure, or set directly the
backpressure if no preset has been programmed previously.
Both options rely on their patented kick detection system Micro-flux, based on a Coriolis
mass meter tied into software that compares flow in and out and alerts the operator if
there is a kick. The system also includes a transient single phase flow simulator, used to
calculate the friction losses. The signal from a PWD tool can be used to compare and
calibrate the flow simulator. Secure Drilling, in both of its options is designed to deal
with single phase liquid systems.
Halliburton GeoBalance®
Figure 20: Halliburton GeoBalance Choke manifold skid
Recently launched, GeoBalance offers four different levels of service for MPD.
GeoBalance Self-Managed (Level 1) provides the operator with a RCD and a dual
hydraulic adjustable choke manifold. GeoBalance Automated (Level 2) provides remote
software-controlled choke.
GeoBalance Optimized (Level 3) adds optimization services, transient flow simulators

and wellbore integrity modeling and monitoring. It integrates data captured by the
equipment and by third party services companies. The system will then control back
pressure to maintain BHP within the pore and fracture pressure, as well as wellbore
stability limits. A back pressure pump can be added to the system, and tied to be
controlled by the software to maintain BHP constant during drill pipe connections.
GeoBalance Sigma® (Level 4) incorporates surface separation and injection equipment to

allow drilling in low pressure zones.

Chapter 4
4.6.5 Design Considerations

The main design criteria for the flow control choke system for a MPD operation should
include the following:
Pressure rating, based on the maximum anticipated surface pressure for the
operation.
Maximum opening of the chokes, suitable for the anticipated flow rate and
minimum backpressure desired.
Material compatibility with anticipated flow, such as sour service.
Serviceability: the set up should be designed to allow working on a plugged or
washed choke while flowing through another one without having to shut down the
operation.
If automated choke is considered for the operation, make sure their operation
response is taken in account in the procedures for kick events.
The choke need to be sized properly, considering pressures, flow velocity, fluid
composition including solids, and pressure drop.
Potential for formation of gas hydrates in case of wet gas flows through the
chokes.
Fluid freezing problems when drilling in sub-freezing areas. This problem may
require the addition of a steam line to maintain temperatures during low flow
conditions.
Material in the return fluids like sand, shale, pipe scale, and rubber that can cause
plugging of the chokes and erosion at bends in the lines and downstream of the
chokes. A minimum of 7 times the diameter of piping is required downstream of
choke outlets.
Choke and choke manifold construction specs are governed by API SPEC 16C.
4.6.6 Drillstring Tools
Drillstring valves
Non-return valves (NRVs) are essential additions to the drill string for MPD operations.
(Figure 22). In most hydrostatically underbalanced operations, they provide the only
barrier between the reservoir and surface from inside the drill string. Scenarios including
the injection of gas to reduce the hydrostatic head require that a minimum of two (2),
preferably three (3) drill string NRVs be included in the downhole assembly to prevent
flow back up the drill string during both connections and
stripping/tripping. A seating nipple above these two NRVs
is additionally recommended in the event the NRVs fail. A
plug (blanking or check valve type) can then be run, if
required, to isolate the drill string from well pressure
Years of underbalanced drilling experience in Canada have

proven that a proper NRV insert must be used. This
experience has shown that the best type of NRV insert is
the non-ported, flapper style. Other types of NRVs used

Figure 21: Drillstring NRV: Flapper and Dart
Chapter 4
have either “washed out” in relatively short periods of time or

became plugged with drill solids and/or barite.
Flapper style NRVs also allow for passage of wire line tools and
coring balls, if required. Thus, it is suggested that the NRVs to be
used in the drill string be of the flapper type, although one dart
(plunger) style should be run at the bottom of the drill string (just
above the bit or core barrel). All drill string NRV valves should be
spring loaded, especially in horizontal wells. The dimension of the
wire line tools and the coring balls that may be used should be sized
the minimum ID of the NRV.
Figure 22 NRV bleed-off tool
They also serve as an important safety device by limiting the

available inventory of fluid in the drill string that can flow back. It is
recommended that the NRV be placed in the drill string just prior to
commencing drilling. Additional NRVs should be placed about
every 500 feet as required to minimize bleed down time during
connections. These will be taken out of the string during trips and
moved to the top of the drilling section when back on bottom.
A special tool is required for relieving trapped pressure below a NRV during tripping
operations. This tool is installed into the float sub and a pin is screwed down to open the
NRV. The pressure-relieving tool has a side outlet to allow for the safe bleeding of
pressure. Rig floor manifolds must include a method to bleed the string to a tank or a
designated safe area.
Conventional drill string float subs are not suitable for MPD operations. These types of
float subs do not mechanically lock the NRV insert in place inside the float sub. Rather,
the friction between the NRV seal and the polished bore sealing area secures the NRV
insert. While drilling, the NRV insert is also held in place by the drill string connection
(the pin end of the drill string above the float sub). During tripping operations however,
trapped pressure below the NRV can exert a force greater than the holding friction
thereby propelling the insert from the sub when it is disconnected from the drill string.
This situation obviously can create an unsafe release of pressure and fluids from inside
the drill string.
Significant drill string float sub advancements have been made to mechanically lock the
NRV insert in place. Float subs are now designed incorporating a groove on top of the
NRV insert so that a snap ring can be installed to mechanically lock it in place. For
additional safety, a short drill string sub may be installed on top of the float sub.
Drill string NRV temperature and pressure ratings must also be considered. Standard
rubber seals are rated between 175 and 200ºF. Special high temperature seals made of the
elastomer Viton are rated to 450ºF. Pressure ratings for NRV seals are as high as
10,000 psi. Sour (H2S) service NRVs are also available.

Chapter 4
Figure 23 A cutaway diagram of a wireline retrievable Non-Return Valve (NRV). The sealing mechanism
can be retrieved in the event that emergency drillpipe intervention operations are desired.
Wire line retrievable NRVs are available from some vendors. These allow the NRV insert
to be retrieved in the event that the drill string becomes stuck, allowing for wire line
recovery operations. The NRV may also be retrieved and replaced in the event of leaks
Figure 24. If suitable wire line retrievable NRVs are available, they are recommended to
be used.
Prior to running any float sub into the well, a pressure test must be performed to ensure
its integrity and functionality.
4.6.7 Tripping Equipment
Rig Assist and Snubbing equipment

Some MPD operations may require trippins pipe while maintaining wellhead pressure. In
this case, it will be necessary to install a rig assist snubbing package or a snubbing
package. This will allow pipe to be tripped in and out with pressure at surface. It is
recommended that the rig assist-snubbing unit only be used in the pipe light mode and
that the conventional rig pipe handling system is used at all other times. This ensures that
interface issues are minimized and conventional pipe handling systems are utilized as
much as possible. The depth of pipe light can be calculated by multiplying the cross
sectional area of the drill pipe by the Maximum Anticipated Surface Pressure and diving
this by the buoyed weight per foot of the drill string (Equation 14). Figure 24 shows the
rig assist snubbing equipment.
Pipe Light Calculation
ADP × MASP
PLD = (14)
BW ft
Snubbing units must be sized for the maximum force that will be required to hold pipe
coming out of the hole and the maximum force required to snub pipe back in the hole.
This force can be calculated by calculating the maximum anticipated surface pressure
during snubbing operations by the cross sectional area of the BHA (Equation 15).

Chapter 4
Snubbing Force Calculation
Fs = ABHA × Ps (15)
Figure 24.Tesco Rig Assist Snubbing Unit
Downhole Annular Valves

Hydrostatically underbalanced MPD operations require constant backpressure in order to
maintain the control of the well.
In an effort to facilitate tripping and other operations under pressure, equipment providers
have developed downhole annular valves. These are designed to close the hole once the
bit has been pulled above the valve, isolating the open hole from the casing annulus. This
allows the annulus to be de-pressurized above the valve, allowing tripping without
pressure. There are two types of valves currently in the market; mechanical and
hydraulic.
Halliburton’s QTV (Figure 25) mechanical valve uses a shifting sleeve that travels with
the bit. When running in the hole, the sleeve locks in a dedicated profile inside the valve,
and pushes the valve spring loaded flapper open. The sleeve stays in the valve during the
rest of the pipe trip. When pulling out of the hole, the bit picks the sleeve, allowing the
valve flapper to close.

Chapter 4
Weatherford’s DDV (Figure 26) hydraulic valve uses control lines all the way to surface.
Applying control fluid pressure to the closing port forces an internal sliding sleeve
upwards, allowing the spring loaded flapper to close. To open, control fluid pressure is
applied to the opening chamber, moving the sleeve down, forcing the flapper open.
Both types of valves can be run in permanent or temporary installation. In permanent

installation, the valve is cemented in the well and stays as part of the casing string. In
temporary installation, the valve is run as part of a retrievable tie-back string. After
drilling operations are finished, the temporary string is pulled back out of the hole, and
the valve serviced to be used in a next operation.
Deployment valves have met with varied degrees of success. Designers face reliability
problems, application barriers such as dimensions, and application problems in terms of
safety.
Some of the problems and limitations include:
Hydraulic valves need to run control line all the way to surface, making the casing
running process more complicated. It also requires modifications to the wellhead
to allow passage of the control lines to the exterior.
Mechanical valves require them to be run on drill pipe or other solid tubulars for
opening, limiting some operations, such as wireline logging.
Valves need additional clearance in previous casing to be run. This could limit the
maximum number of casing strings.
Reliability concerns about flapper seal.

Chapter 4
Figure 25. Halliburton Quick-Trip Valve QTV
Figure 26. Weatherford Downhole Deployment Valve DDVOperations & Well Control
4.6.8 Fundamentals of MPD Well Control

By definition MPD is intended to avoid continuous influx of formation fluids to the
surface. Any influx incidental to the operation will be safely contained using an
appropriate process. Therefore, the definition of a kick and the requirements for well

Chapter 4
control are the same for MPD as they are for UBD. In both cases, the definition of a kick
is the unwanted influx of formation fluid into the wellbore. The intent of well control is
to increase the BHP such that further influx will not occur and to safely remove the influx
from the wellbore.
As many MPD operations have a narrower margin between the formation pressure and
the applied BHP, the probability of having a well control event increases. This is offset
by the improved kick detection methods available and the speed at which an event can be
handled.
What is different in many MPD applications is the reaction and actions that must be
taken, if an influx occurs. For conventional drilling systems, increasing the BHP requires
increasing the density of the fluid. For many of the MPD systems available today,
increasing the BHP can be accomplished by simply increasing applied back pressure to
the system, increasing pump rate or changing the mixing ratio of dual gradient system. In
each of these cases, the change in applied BHP can be done almost instantly, significantly
reducing the impact of the well control event.
Pressure Indicators
With most MPD systems, the same indicators of an increase in formation pressure are
normally available. These include:
Increase in penetration rate,

Change in the shape, size and amount of cuttings,
Increase in rotary torque,
Increase in drag,
Increase in gas content,
Variation from normal “d” exponent4,
Increase in flowline temperature,
Decrease in shale density,
Increase in Chloride content.
The major exception to this rule is MPD systems that do not circulate returns back to
surface (Mud Cap and PMCD).
Kick Detection
In conventional drilling operations, indications that a kick has occurred are:
Flow Rate increase from the wellbore,
Volume Gain in the pits,
4
The ‘d’ exponent is a factor for evaluating drilling rate and predicting abnormal pore pressure zones.
Normally the d exponent should increase with depth, but deviation from this trend may indicate a zone of
abnormal pressure.

Chapter 4
Continued Flow with the pumps off,

Speed of pumps and or pump pressure change,
Gas or oil Show.
The normal first reaction is to shut down the pumps and verify whether the well is
flowing with the pumps off. If the well is flowing, this is a positive indication that a kick
has occurred, followed by the well being shut in to initiate well control.
In most MPD operations, the same indications of an influx occurring are available, in
some form or another. The biggest change is that nearly every method of well control
utilizes a Rotating Control Device (RCD). This precludes the common procedure of
looking down the wellbore and observing the flow to get a positive indication of flow.
For MPD applications, positive indication of flow with the pumps off must be done at a
point downstream of the RCD.
As many of the MPD systems are statically underbalanced by design, require continuous
pumping or a positive pressure at surface to maintain a constant BHP, or have a heavier
density fluid in the drill pipe than the annulus, the concept of watching for flow with the
pumps off must be modified. In these cases, a change in differential flow rate from the
well must be used as a positive indication that a kick has occurred. For this reason, many
of the currently available MPD systems incorporate highly accurate meters, such as the
coriolis meter, and intelligent software, which makes detecting a kick easier. Detecting
kicks earlier, allows the size of the kick to be kept to a smaller size, making the handling
of the kick much easier.
Determining the Kick Intensity

For conventional operations, the first reaction once a kick has been detected is to shut the
well and determine the required MW to balance the BHP. Although some MPD
operations will utilize a trial and error method to quickly balance the BHP, obtaining a
direct or indirect reading of the BHP is the only way to positively determine the required
changes to the system to balance the BHP. The actual steps required to determine the
BHP changes depending on the system being used.
Well Control Matrix

For all MPD operations, a well control matrix should be developed. The purpose of the
matrix is to set criteria as to when it is acceptable to continue drilling, under what
conditions it is acceptable to conduct dynamic well control, and at what point it is
required to switch to conventional well control. In some geographical locations, a well
control matrix is required under governing regulations.
The Minerals Management Service, which governs drilling in federal waters off of the
United States, requires a Managed Drilling Operations Matrix for all MPD operationsi.
The Managed Pressure Drilling Notice to Lessees and Operators (NTL) gives the
following example:

Chapter 4
Table 9: MPD Well Control Matrixi

Surface Pressure Indicator
(See Chart 2 Below)
> Planned Back
MPD Drilling At Planned Pressure &
At Planned Drilling Connection ≥ Back pressure
Matrix < Back Pressure
Back Pressure Back Pressure Limit
Limit
Increase pump rate,

mud weight, or both
No Influx Continue Drilling Continue Drilling Pick up, shut in,
AND reduce surface
pressure to planned or evaluate next action
contingency levels
(See Chart 1 Below)
Increase pump rate,

Influx Indicator
Increase back Increase back

pressure, pump rate, pressure, pump rate, mud weight, or both
Operating AND reduce surface Pick up, shut in,
mud weight, or a mud weight, or a
Limit pressure to planned or evaluate next action
combination of all combination of all
contingency levels
Cease Drilling. Cease Drilling.
< Increase back Increase back
Planned pressure, pump rate, pressure, pump rate, Pick up, shut in, Pick up, shut in,
Limit mud weight or a mud weight or a evaluate next action evaluate next action
combination of all combination of all
≥
Planned Pick up, shut in, Pick up, shut in, Pick up, shut in, Pick up, shut in,
Limit evaluate next action evaluate next action evaluate next action evaluate next action
Table 10: Chart 1 of MMS Well Control Matrixi

Chart 1
Defined Limits for Interval ft to ft TVD
Influx State No Influx None
Operating Limit Low
< Planned Limit Medium
≥ Planned Limit High
Influx Rate No Influx None
Influx Indicator
Operating Limit Light

< Planned Limit Moderate
Influx Duration No Influx None
Operating Limit Low
Volume Gain No Influx None
Operating Limit Low
> Planned Limit High
• Influx indicator can be any or a combination of the factors shown in Table 10.
1. Pit gain is an absolute indicator.
2. Equipment must be used which can measure the influx rates to an acceptable
tolerance.

Chapter 4
Table 11: Chart 2 of MMS Well Control Matrixi

Chart 2
Surface Planned Drilling Back Pressure
Pressure Planned Connection Back Pressure
Indicator Back Pressure Limit
Equipment must be used which can measure the surface pressures to an acceptable
tolerance.
Table 12: Other Indicators from MMS Well Control Matrixi
Other Indicators
Well Control
Triggers
“Other Indicators” list signal that should be considered planned limits and require
immediate remedial actions.
The criteria for developing the well control matrix should be in concert with the basis of
design for the well and the actual system and equipment being used. Although it is
appropriate in some cases to revert to conventional well control for all kicks, in most
cases this would be highly inefficient and in some case it could lead to significant
problems.
Conventional Well Control Techniques

Once an influx is detected, the primary response for MPD wells should always be to shut
in the well and follow procedures to directly read the formation pressure. The only
exception to this would be if the influx is significantly small enough to allow Dynamic
Well Control in accordance with the MPD Well Control Matrix. The major complication
in reading the formation pressure is the presence of Non-Return Valves, floats or flow
stop valves that are present in most MPD operations. For certain methods of MPD, such
as dual gradient drilling, there have additional complications, which will be addressed
below.
Once the formation pressure is ascertained, the method for bringing the well back in
balance and circulating out the influx will need to be determined. The decision must also
be made whether to bring the influx out under the control of the MPD system, or whether
it will be handled through the rig’s well control system. If there is any doubt as to the
capacity or competency of the MPD system, the safest decision is always to utilize the
rig’s well control equipment and follow standard well control protocol. It must always be
remembered that the MPD equipment is not designed or rated as blow out control
equipment and most MPD personnel are not certified for well control.

Chapter 4
For conventional drilling, the only method for controlling the BHP is increasing the fluid
density. Conventional well control theory has several methods for achieving pressure
balance, including:
The Driller’s Method

Wait and Weight Method
The Concurrent Method
The difference between these are when to circulate out the influx from the well, when to
pump the higher density fluid to balance the pressure. All three methods are constant
bottom hole pressure methods and rely on surface pressure to balance the pressure until
fluid with the proper density is circulated around the well.
The advantage that MPD has is there are normally several other parameters that can be
changed to balance the pressure, depending on the type of MPD being employed. These
include:
Surface Pressure (Applied Surface Pressure)
Pump Rate (for SMD, Turbolift, Mud Lift, etc.)
Mixing ratio (for dilution based systems)
Annular Friction (CCS, CCV, etc.)
The method used will be dependent on many factors, including the size and intensity of
the influx and the capacity of the MPD system to operate under a higher load.
Dynamic Well Control

Dynamic well control is the term used to describe actively bringing a well back in
balance, without shutting down and getting a direct reading of the reservoir pressure.
This is only practical in MPD applications, as the changes in the applied BHP can be
done very quickly. The advantage of dynamic well control is the significant reduction on
Non-Productive Time (NPT) before normal operations can be restored. In some cases,
operators will go back to drilling while the influx is being brought out of the well.
As the reservoir pressure is not measured directly, dynamic well control has an inherent
level of risk, in either under estimating or over estimating the pressure required to
balance the formation pressure. Although there is a level of science in determining the
change in BHP required, there is also a level of trial and error in the process.
Due to the inherent risk in dynamic well control, it should only be employed when small
influxes have occurred and corrective actions can be made before significant expansion
of the influx has occurred. It is also important that all participants be prepared to switch
to conventional well control at any point that there is a question on how the well is
responding.

Chapter 4
One of the easiest examples where dynamic is with a statically underbalanced MPD
operation that is using surface pressure to balance the well. If an influx occurs, the well
can be brought back in balance quickly by increasing the surface pressure, providing
there is sufficient pressure capacity in the system to safely handle the higher pressure. In
this case, the primary indication of a well control incident would be an increase in the
flow out of the well and a slight gain in the pit volume. If the gains are significantly
small (as noted in the MPD Well Control Matrix), the choke pressure would be staged up
until the flow in matches the flow out and there is no further gain in the pits. Once this
balance point is reached, the influx would be circulated out, keeping the standpipe
pressure constant.
Specific Well Control Techniques

As stated previously, the method for balancing the well will depend on many factors,
including the type of MPD being employed. Below the well control option normally
used when Applied Surface Pressure MPD method is used.
Applied Surface Pressure MPD

Applied Surface Pressure5 MPD systems employ a combination of surface pressure and
fluid density to achieve the desired BHP. If an influx occurs, either the fluid density was
too low, the surface pressure is too low or both. The advantage of most Applied Surface
Pressure systems is that they have a highly accurate flow meter and have system control
software that give positive indication of an influx quicker than would be detected with a
conventional system. The other major advantage is that once an influx has been detected,
the well can very quickly be brought back into balance by increasing the choke pressure.
The disadvantages of a statically underbalance MPD system from a well control stand
point is all system will have an NRV or float in the drill string, making direct reading of
the formation pressure more difficult. In addition, if the influx is not properly
indentified, many choke based systems will react by opening the choke, inducing an ever
increasing influx.
5
Applied Surface Pressure and Applied Back Pressure refer to the same technique. The term ABP has
been used previously in this manual.

Chapter 4
Applied Surface Pressure - Kick
Pre-Kick Annular Pressure Post Kick SIDPP Post Kick SICP
Depth (ft)
0 Delta SIDPP 2000 4000 6000 8000 10000 12000 14000
0
2000
4000
6000
Pressure (psi)
8000
10000
12000
14000
Kick Height
16000
Figure 27: Example of Kick with an Applied Surface Pressure MPD System
Once an influx has occurred (Figure 27), the first step will be to decide whether a
dynamic kill or conventional well control will be used. Assuming the kick is sufficiently
larger to require conventional well control, the next step will be to shut the well in and
allow the well to come into pressure balance. The decision must be made whether to shut
in on the RCD or close one of the BOPs and whether to shut the well in on the rig’s
choke or stay on the MPD choke (note in most cases, the decisions go hand in hand, as
closing a BOP also isolates the MPD choke). Once the well is in pressure balance (time
required will depend on the permeability of the formation), the formation pressure must
be determined. As the drill string will have an NRV or float, a process of opening the
float must be followed to get a reading of the standpipe pressure. The process is as
follows:
Pressure the pipe up in small increments, kicking the pump on and off.
Monitor both the standpipe pressure and casing pressure as the drill string is
pressured up.
The opening of the valve will be noted by a small dip or break back in the
standpipe pressure. It may also be noted by an increase in the casing pressure, as
the pipe and the annulus come in communication.
Once the valve is open, the pumps are shut down. The standpipe pressure that the
pressure bleeds back to is the initial SIDPP.

Chapter 4
Assuming the drill pipe is full of a clean single density fluid, the bottom hole pressure
can then be determined by:
BHP = SIDPP + MW * 0.052 * Dtvd
Where:
BHP = bottom hole pressure (psi)
SIDPP = Initial Shut In Drill Pipe Pressure
MW = Mud Weight (ppg)
After the BHP is known, the decision must be made as to what parameter will be used to
balance the well; either increased density, increased choke pressure or a combination of
both.
If there is sufficient pressure capacity in the system, the easiest and quickest method is to
keep the MW constant and balance the well with increased choke pressure. In this case,
the final choke pressure will be equal to the SIDPP plus what ever margin is desired. The
first circulation of the Driller’s Method is used to circulate out the influx. Once the
influx is out of the hole, operations can continue as planned.
If there is not sufficient pressure capacity in the MPD System, the density will need to be
increased to make up the difference in pressure. The new fluid density can be calculated
by:
MWnew = (SIDPP-Ptarget)/(0.052 * Dtvd) + MWold
Where:
MWnew = New Mud Weight (ppg)
MWold = Mud Weight at the time of the kick (ppg)
Ptarget = Target surface pressure, pumps off (psi)
The well can be brought in balance either using the two circulation Driller’s Method, the
Concurrent Method or the Wait and Weight Method
i
United States Department of the Interior, Minerals Management Service, “Notice to Lessees and operators
of Federal Oil, Gas and Sulfur Leases in the Outer Continental Shelf, Gulf of Mexico OCS Region,
Managed Pressure Drilling”, NTL No. 2008-G07, May 15, 2008.

Chapter 5 – Offshore MPD
1 Rev. 1.0
Table of Contents
5.1 Introduction ...................................................................................................... 3
5.2 Dual Gradient ................................................................................................... 3
5.1.1 Subsea Mud Lift Drilling (SMD) ................................................................ 4
5.1.1.1 Engineering and Well Design ................................................................. 4
5.1.1.1.1 Example ………………………………………………………….6
5.1.1.2 Operations & Well Control..................................................................... 7
5.1.1.2.1 Kill Weight Fluid Density ……………………………………….8
5.1.2 CAP M......................................................................................................... 8
5.1.2.1 Engineering and Well Design ................................................................. 8
5.1.2.1.1 Example ………………………………………………………….9
5.1.2.2 Operations & Well Control................................................................... 11
5.1.2.2.1 Connection and trip induced kicks ……………………………..14
5.2.1 Continuous Circulating Strategies ............................................................. 14
5.2.1.1 Operations & Well Control................................................................... 16
5.3 Equipment ...................................................................................................... 17
5.3.1 Sea Floor Pumps........................................................................................ 17
5.3.2 Marine Risers ............................................................................................ 18
5.1.3 Virtual Riser® ........................................................................................... 19
5.1.4 RCD Installation from a Jack-Up .............................................................. 19
5.1.5 Riser Loads ................................................................................................ 19
5.1.5.1 Pressure Limit ....................................................................................... 20
5.1.5.2 5.4.6 Effective Tension ........................................................................ 20
5.1.5.3 Impact of internal pressure on riser design ........................................... 21
5.1.5.4 Emergency Disconnect ......................................................................... 21
5.1.5.5 Recoil .................................................................................................... 22
5.1.5.6 Fatigue .................................................................................................. 22
5.1.5.7 Slip Joint ............................................................................................... 22
5.1.5.8 Isolation ................................................................................................ 23
5.2 Offshore Operations ....................................................................................... 23
5.2.1 Space Limitations ...................................................................................... 23
5.2.1.1 Equipment ............................................................................................. 23
5.2.1.2 Personnel .............................................................................................. 24
5.3 References ...................................................................................................... 25
2 Rev. 1.0
Chapter 5
MPD Pressure
Control Options

Friction Pressure
Mechanical lift
device
Circulating Non
Long choke Circulating
SMD (Mud Lift)
circulation of fluid
Semi-automatic
CCS (CBHP)
Concentric Mud Cap
injection
Floating
multi-phase)
CAPM
gradient)
5.1 Introduction
Applying MPD techniques to offshore prospects can have significant impact on well
costs where operating day rates are high in comparison to land based operations. Any
Invisible Lost Time (ILT) saved while drilling has an amplified effect on the drilling
AFE. Combating the effects wellbore breathing by employing constant bottomhole
pressure techniques and reducing the number of required casing strings are two of the
ways that MPD can impact offshore operations.
5.2 Dual Gradient

As with Low Head MPD discussed in a previous chapter, there are many forms of dual
gradient drilling. As the name infers, two fluid gradients are designed and managed to
manipulate the BHP.
Conventional drilling requires the static mud column to hold back formation pressure. In
offshore wells the static mud column imposes significantly more pressure against the
formation since it occupies the mud riser (sea bed to surface). However, in the upper
sections of the borehole, there generally exists a reduced formation overburden which
makes for a narrow pressure window between pore pressure and fracture gradient. The
premise of dual gradient is to reduce or remove this additional pressure imposed by the
riser mud column. In doing so, the number of casing strings, and thus cost, can be
Offshore MPD 3 Rev. 3.0

Chapter 5
reduced. The simplest form of dual gradient would be to allow all returns to spill onto
the sea bed. However cost and environmental legislation limit this approach.
Figure 1- Reduced casing strings due to dual gradient. Scientific Drilling, No. 6, July 2008
Dual gradient is not necessarily accomplished through the injection of an additional fluid
or gas, it can also be achieved through the injection of glass or thermoplastic beads that
get recycled and re injected. For this section, we discuss two principle methods of dual
gradient, namely Mud Lift and Continuous Annular Pressure Management (CAP M).
5.1.1 Subsea Mud Lift Drilling (SMD)
5.1.1.1 Engineering and Well Design

SMD uses a subsea pump to lift the returns and cuttings from the sea bed. In doing so, it
removes the additional well bore pressure that would have been exerted from mud in the
riser. This is now replaced by a column of sea water of reduced weight and gradient.
SMD was investigated by a Joint Industry Project (JIP)1 for water depth applications
from 4,000 to 10,000 feet. The JIP performed one test deployment of the system
successfully, however, for various techno-economic reasons widespread use has been
negligible.
1
SPE 71357, SubSea Mudlift Drilling Joint Industry Project: Delivering Dual Gradient Drilling
Technology to Industry, K.L. Smith et al.

Chapter 5
Design considerations for these well types become a complex mass balance equation. The
mud weight and injection flow rate need to provide the required back pressure on the
formation. It is possible to then add an applied back pressure in the sea floor by installing
a subsea choke that is operated by an ROV. The flow rate into the well must not exceed
the capacity of the sea bed pump to lift the returns to the rig. Systems also exist that
exclude a riser and subsea RCD and allow the subsea pump to lift all returns to surface
through a hose.
Thermal calculations are also important for deep offshore wells where the density of the
fluid is constantly changing with the abrupt changes in surrounding temperatures.
Trip tank design must be addressed with subsea mud lift since two separate mud
gradients are being used. For the riser, that is full of sea water, it may be prudent to have
two trip tanks. One is used to continuously pump across the riser. The second tank should
be kept half full and tied in to monitor for any gains / losses in the riser that could come
about from RCD leaks. Another trip tank is then required for the mud system. This
monitors gains / losses on the well, and is also tied into a high horsepower fill up pump.
U-tubing is another aspect that needs design consideration. There has been development
into string valves that will hold a column of mud in the string from surface to the sea
floor. If these were not included, the mud in the drill string from surface to the sea floor
would continue to U tube on connections, and overload the sea floor mud pump.
The unique aspect of the design of SMD wells rests with the specification of the seawater
powered fluid pumps, Figure 2. The governing pressure balance equation for the pump
is as follows:
PSWD = PDMW + PFML + PFSWL
Where:
PSWD = seawater power fluid discharge pressure
PDMW = 0.052 xWDx ( PPG MUD − PPGSW )
= Pressure differential between column of mud and seawater at the sea bed
PPGMUD = mud weight in pounds per gallon
PPGSW = seawater weight in pounds per gallon
WD = water depth
PFML = friction pressure in the mud return line
PFSWL = friction pressure in the seawater powerline

Chapter 5
Figure 2 - Details of SMD pumping system.
Note that this is the pressure equation of the feed pump, not of the seafloor pump itself.
SMD systems operate under the principle of pressure, not volume, so as to maintain
annular pressure appropriately. The seafloor pump is instrumented with pressure gauges
on the inlet and outlet side, and is controlled / monitored by a PLC. If a well control
event is noted, the system must revert to volumetric control to ensure that kicks can be
controlled properly.
5.1.1.1.1 Example
Example - Find the required mud weight and sea floor pump specification for an offshore
MPD project in 10,000 feet of water. The target bottom hole pressure is 15,000 psi, and
TVD is 15,000 feet below the mud line (25,000 ft). The frictional loss in mud return line
is 400 psi.
Step 1 - Calculate sea floor pump inlet pressure
Pinlet = Phyd @ sea floor
=10,000ft*8.4 ppg*0.052
=4,368 psi
Step 2 - Calculate fluid density
(Target BHP – Phydsw) / ((TVD-WD)*0.052)
(15,000 psi – 4,368 psi) / (15,000*0.052)
=13.6 ppg
Step 3 – Calculate pump output pressure
= Pmw + Pf
= 13.6 ppg*0.052*10,000 ft + 400 psi
= 7,488 psi

Chapter 5
Step 4 - Pump Specification

• = Pout-Pin
• =7,488 psi - 4,368 psi = 3,120 psi
5.1.1.2 Operations & Well Control

Similar to the dilution based systems, pump based system utilize a high density fluid in
the drill string. This makes the use of SIDPP for determining reservoir pressure
impossible. During normal drilling activities, pump based MPD systems control the BHP
by controlling the inlet pressure to the pump. The energy imparted on the fluid decreases
the annular pressure. The pump speed becomes the primary indicator of an influxi, ii. If
an influx occurs, the pump must increase to keep the inlet pressure at the pump constant.
Other methods of kick detection are also available for mud lift systems, including;
Drilling break,
Pit gain,
Decrease in circulating pressure,
Increase in rotary torque, drag and fill,
Increase in string weight.
One method not available is the size and shape of the cuttings, as mud lift systems utilize
a cuttings grinding system in front of the pump. Flow from the well with the mud pumps
off may or may not be an indicator, depending if a flow stop valve is utilized.
Similar to most forms of well control, the first decision that must be made once a kick is
detected is whether to shut in the well or utilize a dynamic kill procedure to control the
flow. If the kick is sufficiently small, as defined by the Well Control Matrix, control of
the BHP can be achieved instantly by changing the inlet pressure at the pump. The inlet
pressure can be increased until the subsea pump rate matches the mud pump rate.
In most instances, the well will be shut in for more conventional well control activities.
Shut in procedures must be well thought out, as a hard shut in can induce fracturing the
well if a flow stop valve is not in place or is leaking. This is because the hydrostatic head
of a full column of mud in the drill string will normally exceed the fracture gradient of
the formation. Therefore, a modified driller’s method would be used which would
include:
Switch control of the Subsea pump from inlet pressure control to pump speed
control,
Slow pump speed to pre-kick rate and record the inlet pressure, and drill pipe
pressure,
Switch control of the subsea pump to manual and hold the drill pipe pressure
constant while circulating the influx out of the wellbore,
Once the influx is circulated out of the well, the new inlet pressure could be set at
the pressure at the end of the circulating process. Alternatively, the mud weight
could be increased to offset the increase in reservoir pressure.

Chapter 5
5.1.1.2.1 Kill Weight Fluid Density

Unlike conventional drilling well control, it is not possible to directly determine the BHP
from measuring the SIDPP with the well shut in. The ability to estimate the reservoir
pressure with the pumps off depends on whether a flow stop valve is employed.
If a flow stop valve is used, the theoretical SIDPP can be estimated by measuring the
change in the opening pressure, pre and post kick. If no flow stop valve is employed, the
theoretical SIDPP can be estimated by the change in the subsea pump inlet pressure pre
and post kick.
The new mud weight required to balance the formation must be modified to take into
account that the equivalent density must be calculated with respect to the sea floor.
Therefore the kill weight mud is calculated by:
KWM = SIDPPt
+ MWold
0.052 * ( Dtvd − Dw )
Where:
KWM = Kill Weight Mud (ppg)
SIDPPt = Theoretical Shut In Drill Pipe Pressure (psi)
Dtvd = True Vertical Depth (ft)
Dw = Water Depth (ft)
5.1.2 CAP M
5.1.2.1 Engineering and Well Design

CAP M is a dual gradient MPD concept developed by Transocean. It is also referred to
as a dilution based system. In place of injecting a gas into the riser, or using a subsea
mud pump to lift the returns, a second lighter mud (LM) is injected into the riser. By
mixing the drilling mud, or Heavy Mud (HM) and lighter mud, a third “mud weight” is
developed. The summation of the drilling fluid in the open hole plus the mixed mud in
the riser gives an equivalent mud weight at the bit lower than what a single gradient
would exert.
The advantages of this system are:
No major rig modifications
Retains operability in that it is simple enough to switch from single gradient
drilling to dual gradient drilling
Lower capital investment required.
Minimal subsea equipment to operate and maintain.
Designing a well using CAP M requires navigating the mud weight curve through the
pore pressure / frac gradient as shown in Figure 6. To achieve this either the mud weights
or the respective pump rates can be adjusted.

Chapter 5
Heavy Fluid = 14.0 ppg Dual Gradient Configuration
Mix Fluid = 10.7 ppg
Light Fluid = 9.0 ppg 5,950 ft of 10.7 ppg mud in the riser.
Priser = Mud Wt x 0.052 x TVD

= 10.7 x 0.052 x 5,950
= 3,311 psi
ML
6,000 ft
There is 5,050 ft of 14.0 ppg mud in the open

hole, so the bottom hole pressure at TD is:
BHP = Priser + Popen hole
= Priser + (14.0 x 0.052 x 5,050)
= 3,311 + 3,676
11,000 ft BHP with SG 14.0 ppg mud was 8,008 psi
Figure 3- Dual gradient calculations for Light and Heavy Mud weights.
Dual gradient wells can be planned with or without applied surface back pressure,
depending on the project drivers. The choice of LM and HM fluids is primarily
dependent upon the desire to remain statically overbalanced or underbalanced. Steps in
choosing mud weights are best demonstrated by example.
5.1.2.1.1 Example
Example - Figure 3 breaks down the fluid gradients above and below the mud
line for a well in 5,950 feet of water drilled to 11,000 feet. Above the mudline
injecting a 9 ppg LM into the riser mixes with a 14 ppg heavy drilling mud. The
result is a 10.7 ppg mix fluid, exerting 3,311 psi at the mud line. The 14ppg HM
is effectively dispersed through the open hole, exerting a pressure due its
hydrostatic column of 3,676 psi at the bottom of the hole. The total bottom hole
pressure is the sum of the HM in the open hole and the LM above the mud line, or
3,311 + 3,676 = 6,987 psi. When compared to the pressure exerted by a single
gradient of 14 ppg mud at the same depth (8,008 psi), the DG system is
considerably lighter.
Conceptually, Figure 4 depicts how the pressure gradients vary with depth for dual
gradient drilling. The yellow curve shows the Light Mud gradient while the green curves
show the heavy Mud. The blue line shows the resultant mixture fluid density returning to

Chapter 5
surface. The impact that lightening the fluid in the annulus is graphically shown as a
reduction in bottom hole circulating pressure. When analyzed as equivalent mud weights
(eMW) instead of pressure gradients, curves as shown in Figure 5 and Figure 6 are
developed. What is immediately noted is that the eMW curves do indeed ‘curve’ with
depth, as the calculation point for the mud weight is effectively ‘reset’ at the depth in
which it encounters the lighter fluid. As two fluids of different densities are stacked one
upon another, the resultant eMW does not vary linearly, but parabollicaly.
Figure 4 - Pressure gradient variances for dual gradient drilling.

Chapter 5
Figure 5 - Various mud weight curves below the mud line
Figure 6 - Navigating the mud weight through the narrow pressure window

The major issue with dilution based MPD systems like CAP M is the ability to determine
the kick intensity, if a kick has occurred. This problem occurs because the drill pipe will

Chapter 5
almost always be overbalanced, even when a kick occurs. As knowledge of the kick
intensity is required to determine the actual BHP and the density required to balance the
reservoir pressure and the fluid density in the annulus is contaminated any time a kick has
occurred, it is impossible to directly measure the reservoir pressure.
There are two methods for estimating the BHP. The first method is to observe the
pressure at the BOP upon initial shut in. If the kick is sufficiently small, the BHP can be
calculated from the BOP pressure plus the hydrostatic of the heavy mud from the BOP to
TD. The estimated BHP can be confirmed by slowly pump down the drill string until
there is a positive indication that the Flow Stop Valve has opened and comparing the
difference in the opening pressure before the kick to the BHP after the kick (Figure 7).
The difference can be added to the BHP before the kick to estimate the BHP of the
reservoir.
Delta-Crack Open Press

SICP
Pressure
Drill pipe shut in

Annulus shut in
Fi
ng
Depth - TVD
er
Cr
Pr
ac
in
k
tD
op
at
en
a
af
te
rk
ic
k
Influx pressure
Figure 7 - Cartoon of Kick in Dilution Based MPD

The important point is that this is only an estimate, compared to the direct measurement
measured in conventional operations. For this reason, it is recommended that the driller’s
method be used to circulate out the kick. Once the kick is out of the annulus and up in
the choke line, the well can be shut in and a direct measurement taken. A more accurate
measurement can be made when the first circulation of the driller’s method is complete,
and the kick is out of the well.
Extreme care must be taken prior to shutting in the well on a suspected kick that occurs
when the pumps are brought down. It is critical that the correct determination be made
that the flow is a result of a kick, not a failure of the Flow Stop Valve. Shutting in the
well with a failed Flow Stop Valve will result in extreme annular pressure that will
probably result in fracturing the formation. The major difference in a kick and a failed
FSV will be noted in the SPP. For a kick, the FSV will trap and hold pressure as the

Chapter 5
pumps are brought down. For a failed FSV, the SPP will quickly drop and eventually go
on a vacuum as the fluid U-Tubes to the annulus.
Another major complication of a dual gradient kick is the fact that proper dilution must
be maintained throughout the kill. It must be recognized that not maintaining proper
dilution may result in breaking down the shoe. As closing the BOP will normally isolate
the dilution line, the dilution fluid must be pumped down the kill line. The dilution ratio
must be kept constant during the first circulation to avoid complicating the kill process.
The general steps for shutting in and completing the first circulation are:
1. Stop Rotating,
2. Pick up off bottom,
3. Stop Heavy mud Pumps,
4. Stop light Fluid Pumps,
5. Verify that FSV has not failed,
6. Close upper annular,
7. Open the BOP choke line valves,
8. Record trapped DPP, SICP, Pit Gain and BOP Pressure,
9. Confirm space out and close hang-off rams,
10. Slack off and land drill pipe on rams,
11. Monitor riser for flow at the drilling manifold (indicating leaking BOP),
12. Open Kill line and line up light fluid on the kill line,
13. Bring the kill line pump up to speed, circulating down the kill line and up the choke
line, while maintaining the BOP pressure constant,
14. Bring the heavy pumps up to kill rate while maintaining constant pressure at the BOP,
15. Once the pumps are up to speed, continue to circulate the influx out of the hole
keeping the DPP constant.
After the first circulation is complete and the BHP known, there are several methods for
killing the well. These are:
Increasing the heavy mud density,

Increasing the light mud density,
Changing the mixing ratio.
If suitable flexibility exists, changing the mixing ratio can be done quickly, but the full
benefits of the dilution are lost. In most cases, the mud weight will be increased to
compensate for the increased reservoir pressure. In this case, the kill weight fluid density
can be calculated by:

Chapter 5
KWM = SIDPPt
+ MWold
0.052 * ( Dtvd − Dw )
Where:
SIDPPt = Theoretical Shut In Drill Pipe Pressure (psi)
Dw = Water Depth (ft)
Note that unlike conventional calculations, the KWM is calculated from the dilution point
to TD. Using this method, the dilution ratio must also be modified to keep the density of
the riser fluid constant.
5.1.2.2.1 Connection and trip induced kicks

One of the biggest risks in drilling with dual gradients is failure to control proper density
in the riser during connections and trips. For single phase systems, the problem will
occur if the light fluid pump is not turned off when the heavy fluid pump is turned off.
Continued pumping of the light fluid will result in a lightening of the fluid column, which
could result in a kick.
The proper fluid density must also be maintained in the well bore to maintain safe BHP
during trips. This is complicated by the fact that two fluid densities are present in the
wellbore. Normal steps for tripping are:
Pump out of the hole to the shoe, maintaining the proper pressure plus trip margin
at the BOP.
Once at the shoe, the FSV is normally bypassed, allowing the fluid in the drill
string to U-tube to the annulus. Once the fluid stabilizes with the proper trip
margin at the BOP, the light fluid pump is turned off.
Continue tripping out of the hole, filling periodically with heavy mud through the
drill pipe, maintaining the proper trip margin at the BOP.
Once the bit is in the riser, the RCD element is pulled and the trip tank is used to
continue filling the riser with mixed fluid density.
5.2.1 Continuous Circulating Strategies

Drill pipe connections are a source of disruption during MPD operations. The main effect
is the loss of friction along the annulus when mud pumps are stopped for connections.
This is combined and aggravated in the case of compressible fluids as part of the
circulating system, because of the gas/liquid segregation, as well as the time required to
stabilize flow and pressure oscillations after resuming circulation. Multiple strategies
have been developed to minimize these problems, and compensate for the loss of friction
in the annulus. These strategies include automated choke systems and back pressure
pumps. However, avoiding the flow interruption would be the best scenario. Coiled
tubing is a good example of continuous circulation, and an excellent match for MPD

Chapter 5
operations. Unfortunately, it is not always an option, due to its limitations in hole size and
reach. Due to its specific nature, CT equipment is not discussed in this manual.
For jointed pipe drilling, different strategies have been designed to maintain circulation
while making connections. One of these strategies involve
continuous circulation valves, CCV (Figure 8), that allow to
circulate through a side port on the valve, down the drill string,
while an additional single/stand of drill pipe is added on top of the
valve to continue drilling. A valve must be installed on top of
each drill pipe single/stand before the continuous circulation
operation starts. When a connection is to be performed, a hose is
connected to the side inlet of the valve, the flow from the mud
pumps will then be switched from the top inlet to the side inlet
and top drive can then be disconnected and a new single/stand
installed. To continue drilling, the operation is reversed.
A similar principle, but involving more complex equipment is

provided by National Oilwell Varco’s Continuous Circulation
System CCS™. In this case, a device including a series of two
pipe rams and one blind ram is installed on the rig floor, with
lateral mud inlet and bleed off port. When connection is ready to
be made, the lower and upper rams close around the pipe. The Figure 8: An example of a
saver sub is then disconnected from the drill pipe, without flapper type Continuous
stopping the mud pumps. Once disconnected, mud is started to be Circulation Valve
pumped through the lateral inlet, between the lower and the blind
ram. Flow from the standpipe is stopped, and the blind ram is closed. After relieving the
pressure, the upper ram can be opened and the pipe lifted to pick up the new stand/single.
A rig assist snub unit is built-in on top of the CCS, to prevent the pressure inside the
chamber to push the pipe out, once the connection is broken.

Chapter 5
Figure 9: The rig floor mounted Continuous Circulation System. Drilling fluid flow is
diverted from the drillpipe to an enclosed chamber in the unit, allowing it to flow through
a the drillpipe while making a connection.
To make up the new pipe on the drill string, the tool joint is placed in the chamber, and
the upper ram closed against the pipe. After pressure equalization, the blind ram can be
opened and flow through the stand pipe is resumed, while stopping flow through the
lateral inlet. Connection is made up, pressure is released and both upper and lower rams
are open. Then drilling can be resumed.
Both continuous circulation strategies (CCV and CCS) have shown good success in
maintaining the bottom hole pressure reasonable constant during connections. The
downsides are that connection time is increased significantly, impacting total operation
time. Also, the preparation time is considerable. The CCV requires that a valve be made
up on top of every stand (or single) that is going to be used. The CCS requires the
installation / operation of a considerably sized unit on the rig floor, with additional
personnel.

For continuous circulation based systems, a kick is handled similarly to a conventional
well control event. The well is shut in and the BHP is determined from the SIDPP. The
required mud weight to balance the kick is calculated by:

Chapter 5
KWM = SIDPP
+ MWold
0.052 * Dtvd
Where:
SIDPP = Shut In Drill Pipe Pressure (psi)
The influx is then circulated out using either the driller’s method or the wait and weight
method.
5.3 Equipment
Equipment commonly used in offshore MPD applications are the surface RCD and an
MPD choke manifold previously discussed in this manual. Specialty equipment such as
sea floor pumps and continuous circulating devices are discussed in this section.
5.3.1 Sea Floor Pumps

Pumps are placed on the sea floor, attached to the wellhead. They take the fluid returning
on the annulus, and pump through separate lines to the floater rig on surface. Sea floor
pumps are usually high pressure electric driven pumps, with enough capacity to pump all
the returns, including cuttings, all the way to surface. The use of sea floor pumps
eliminates the requirement of the riser. Figure 10 shows the pumps of the Riserless Mud
Recovery RMR®, by AGR. These are capable of pumping up to 1500 gpm back to
surface. Depending on the water depth of the operation, successive pumping units will be
deployed to cope with the hydrostatic head. The pumps require a surface control unit.
Figure 10 - AGR Sea floor pumps for RMR® system

Sea floor pumps have been used in two different arrangements. In the first one, the pump
is connected to the wellhead, which remains open to the sea water column on the top. The
pumping speed is controlled to lift the same amount of mud that is being pumped down
the hole by the mud pumps. Dedicated sensors in the well head detect if the mud level
stays constant above the suction point. If it decreases without changing the surface
pumping speed, it could indicate a loss circulation scenario. If, on the other hand, the

Chapter 5
level increases, it possibly means that a kick is occurring and

evaluation needs to be done, possibly leading to a well control
situation. This scenario is basically applicable for top holes.
Error! Reference source not found. depicts the funnel placed
above the wellhead for the open RMR system. This funnel hosts
the pressure and level sensors, as well as a ROV-friendly side
outlet, and light and video camera for monitoring.
The other scenario where sea floor pumps are used includes the
use of a rotating head above the subsea BOP stack. The
pumping speed is controlled automatically to maintain a set
wellhead pressure in this case. This pre-set wellhead pressure
can be changed at any moment, and the system will react with
variation of the pumping speed to maintain the pressure. This Figure 11 The funnel used in
system provides the well planner with an ample range of riserless mud recovery
possible well head pressure, from just equal to the sea water systems.
gradient, up to the RCD working ratings.
5.3.2 Marine Risers

Marine risers provide MPD planners with means to adjust the fluid gradient in off shore
wells. Possibilities include annular injection of fluids at the mud line level, and placement
of a RCD on top of the riser, to facilitate annulus back pressure. Sea floor injection in the
liner offers options ranging from lower density fluids (including gases) to higher density
fluids. These options allow the manipulation of the hydrostatic column, as well as the
friction losses, along the riser. The operational benefits derived of the use of these options
are discussed in depth in the Dual Gradient section of this manual.
Some equipment limitations in the use of marine risers for MPD operations include:
Pressure rating limitations: riser pressure ratings need to be revised, because

light fluid injection could lead to a riser collapse situation; or the injection of
heavier fluid, possibly combined with the use of RCD on top of the riser could be
limited by riser burst rating, or by the flexible joint pressure rating.
U-tube effect: As well as the system using sea floor pumps, the marine riser
systems used to inject lighter fluid in the returns path face the issue of U-tube
effect due to an unbalanced fluid column when mud pumps are stopped. To cope
with this phenomenon, a drill string valve has been designed. The spring in the
valve allows flow only when the differential pressure across the valve exceeds
certain pre-set value. This value will be exceeded only when the mud pumps are
started.

Chapter 5
5.1.3 Virtual Riser®

Weatherford’s Virtual Riser system includes the use of a
RCD on the sea floor, specifically for surface holes. It
diverts returns flow from the well head to the sea, hence only
compatible fluids can be used with this system. The use of a
RCD on the sea floor allows to handle a pressurized
wellbore for MPD operations, and its major advantage is that
allows dealing with shallow abnormally pressured aquifers.
Subsea installations of RCDs are still novel at the time of
writing this manual. The RCD shown in this figure has been
bench tested to a simulated water depth of 7,500 feet for 168
hours at a pressure differential of 2,000 psi.
5.1.4 RCD Installation from a Jack-Up

The height of the base of the rig floor above the annular BOP must be
sufficient to accommodate the height of the RCH. For most jack-up rigs,
this is not an issue, as the rig can be jacked up to accommodate the
height required. Figure 13: Weatherford Subsea
Virtual Riser shown in a test rig.
Figure 14 shows a typical rig up of the BOP stack and RCH below a
jack-up rig. Note the two ESD valves immediately after the RCH.
Figure 14 View of Stack and RCH Under a Jack-Up Rig
5.1.5 Riser Loads

The loads on the riser are impacted by many of the MPD techniques. In particular,
techniques that impose a back pressure on the riser will impact riser loading. The main
difference between the loads imposed by MPD and those imposed during Conventional

Chapter 5
Drilling is the change in the riser internal pressure conditions. In general, the internal
pressure may increase (as in the Applied Back Pressure technique), although in some
techniques and situations (such as CAPM, dual gradient or MPD in sub-hydrostatic
reservoirs), the internal pressure may actually decrease in comparison to conventional
drilling.
MPD can impact the following riser design considerations:
Effective tension and operating window,

VME Stress
Emergency Disconnect
Riser Recoil
Additional piping
Fatigue
Slip joint.
5.1.5.1 Pressure Limit

As noted, the pressure load on the riser can change during MPD operations. Although the
pressure limit for the body can be calculated, the controlling limit will usually be the
connector. Many of the connector manufactures are not willing to give a value for the
pressure rating of the connectors. For 21” OD, 1” Wall riser, the limit most used is 1500
psi. In MPD design, it is important to estimate the maximum allowable pressure load
(Burst and Collapse) on the riser, and ensure that the loads remain below these limits
during all of the MPD operations (including tripping).
5.1.5.2 5.4.6 Effective Tension

The concept of effective tension is familiar in the offshore structures industry as it is
crucial in riser design. A detailed discussion of effective tension is beyond the scope of
this chapter. However, readers are referred to the excellent papers by Lubinski, et. al.2,
Sparks3 and Mitchell4 for an appreciation of the importance and use of the concept.
Briefly, effective tension is the axial force acting on the pipe cross section when the
capped-end pressure forces (sometimes referred to as the “pressure-area” forces) are
excluded. Mathematically, it is defined simply as
Feff = Freal − p i Ai + p o Ao (Eq - 1)

Where:
• Feff = Effective force
• Freal = Real force
• pi = Internal pressure
2
Lubinski, A., Althouse, W. S., and Logan, J. L., “Helical Buckling of Tubing Sealed in Packers”, SPE
Petroleum Transactions (Journal of Petroleum Technology), June 1962, pp. 655-670. Also archived as
SPE 178.
3
Sparks, C. P., “The Influence of Tension, Pressure and Weight on Pipe and Riser Deformations and
Stresses”, ASME Journal of Energy Resources Technology, Vol 106, March 1984, pp. 46-53.
4
Mitchell, R. F., “Fluid Momentum Balance Defines the Effective Force”, SPE/IADC 119954

Chapter 5
• Ai = Area that internal pressure is acting on, πri2

• po = External Pressure
• Ao = Area that external pressure is acting on, πro2
The purpose of this definition is to differentiate between fluid-conveyed axial forces, and
axial forces conveyed directly to the pipe cross section. Where as real force acting on the
tubular is the force that would be measured by an imaginary load cell placed across the
axial cross section of a tubular immersed in a fluid, effective force is a “fictitious”
quantity defined as above, to aid in determination of buckling tendency for bodies subject
to both fluid pressures and axial forces.
Buckling is influenced by forces caused by pressure as well as real axial forces (such as
weight, overpull or slack off, thermal forces), and their combined effect is best described
by using the notion of effective force. Note, from the Feff equation, that the effective
force can be negative (i.e., buckling is possible) even when the real force is positive (i.e.,
the stress state is tensile), depending upon the internal and external pressure, and the ID
and OD areas of the pipe. Conversely, even when real force is compressive, buckling
may not occur. This is because of the pressure effect on the stability of unsupported pipe-
external pressure tends to stabilize pipe, while internal force tends to destabilize pipe.
This is the reason why although the compressive stress increases with depth in a purely
hydrostatic environment (such as when a pipe is lowered into an infinitely deep sea), the
pipe cannot buckle as long as the density of the pipe material is greater than that of the
fluid (the reader can verify that in this case, the effective compressive force at the free
end of the pipe is zero regardless of the depth).
5.1.5.3 Impact of internal pressure on riser design

Chief among many of the advantages of effective tension is the fact that it controls onset
of buckling. Therefore in riser resign, effective tension controls the operating window.
As the applied back pressure acts on the inside of the riser, the effective tension on the
riser will be reduced. The impact can be substantial, depending upon the riser geometry
and pressure magnitude. A decrease in effective tension will increase the propensity for
buckling. If possible, this must be compensated for by increasing riser tensioning. If the
load is not compensated for, rotation of the flex joint will increase, reducing the operating
window.
Internal pressure will also impose additional hoop stress, and hence the triaxial (von
Mises Equivalent or VME) stress on the riser. However, in a VME sense, the impact of
the additional hoop stress is minimal as the stress is balanced by the axial and bending
stresses. The actual VME stress state must be checked against API RP 16Q
recommendations, which limit the maximum VME stress to 40% of minimum yield
strength (σy)during normal operations, and 67% of σy in emergency conditions.
5.1.5.4 Emergency Disconnect

Risers typically have a limit on maximum pressure load allowable to enable emergency
disconnect. This is normally set in terms of “maximum mud weight” allowable at a given

Chapter 5
depth rating of riser. For example, a rating of 16 ppg at 10,000 ft equates to a differential
pressure rating of ~3,980 psi at the disconnection point.
In the above example, adding 1500 psi of pressure at surface would lower the allowable
mud weight to ~13.1 ppg.
Disconnect may still be possible at the higher burst pressure, but it needs to be checked in
the design stages.
5.1.5.5 Recoil
Recoil calculations will have to be checked, as the recoil with a rotating control head
(RCD) and different pressure loading will be different. Additional piping is required due
to the application of MPD, which needs to be designed with adequate allowance for
movement.
5.1.5.6 Fatigue
Fatigue design is unlikely to be affected by MPD, as long as the standard API 16Q limits
on VME are satisfied (40% of σy normal operating, 67% of σy emergency).
5.1.5.7 Slip Joint

For MPD operations, the slip joint must either be locked while drilling, the inner barrel
removed or the RCD must be placed below the slip joint (see Figure 15 for an illustration
of the RCD rig-up). This is because the RCD must be fixed relative to the earth or the
pipe and tool joints will move uncontrollably through the RCD element with the motion
of the rig. This will lead to early failure of the element.
The slip joint is also locked due to the low pressure rating of the slip joint seals when
moving. Without out locking the slip joint, pressure capacity will be severely
compromised.

Chapter 5
Figure 15: Typical RU of rotating control head from a floating drilling unit.
5.1.5.8 Isolation
Many operators have chosen to isolate the marine riser with a higher pressure inner riser.
This inner riser (usually 13 3/8” or 9 5/8” casing) is landed in a specially built housing
that is run at the top of the riser. The lower part of the inner riser is sealed upper annular
preventer.
Utilizing an inner riser eliminates the pressure load on the marine riser, but adds
additional compression loading on the system that must be compensated for with
additional tensioning.
5.2 Offshore Operations

Offshore MPD operations create special problems that require extensive engineering and
planning. The majority of the problems stem from the confined space in which the
project must be done. This creates issues with heat, escape routes, zoning and exhaust.
5.2.1 Space Limitations
5.2.1.1 Equipment
A major issue in designing an MPD project offshore is the limited space available on a
rig. Although most of the equipment is containerized for offshore applications, the
equipment must be placed on the rig such that it will not interfere with normal drilling
operations. This usually requires critical timing for the placement of the equipment. The

Chapter 5
equipment must be place so that piping is minimized, in particular the need for turns. As
there is potential for hydrocarbons, it must be zone 2 rated and isolated from any non-
zone rated equipment.
Weight of equipment is also a critical factor in underbalanced equipment for offshore
equipment. Skid weights can exceed 20 tons. It is critical that the crane capacity vs.
boom length be checked against the wait and location for each skid.
5.2.1.2 Personnel
Space limitations for personnel on offshore installations is also critical. By law, every
person on the rig must have a space on a lifeboat in case of an emergency. There are
seldom-excess living accommodations available on a jack-up rig during drilling and
completion activities.
MPD drilling operations requires 2 to 5 specialists to operate and manage the equipment.
Planning is critical to assure that the people that are required for each activity are
available on the rig. In some cases this requires elimination of non-essential personnel
during MPD operations.

Chapter 5
5.3 References
i
Smith, K., Gault, A., Witt, D., Weddle, C., “SubSea MudLift Drilling Joint Industry Project, Delivering
Dual Gradient Drilling Technology to Industry”, SPE 71357, SPE Annual Technology Conference, 30
September 2001.
ii
Shubert, J., Juvkam-Wold, H., Weddle, C., Alexander, C., “HAZOP of Well Control Procedures Provides
Assurance of safety of the SubSea MudLift Drilling System”, IADC/SPE 74482, IADC/SPE Drilling
Conference, 26-28 February 2002.

Chapter 6 – Low Head MPD
Low Head MPD 6-1 Rev. 1.0

Table of Contents
6.1 Introduction ..................................................................................................... 3
6.2 Engineering and Well Design ......................................................................... 3
6.2.1 Multiphase Flow Fundamentals ................................................................. 3
6.2.1.1 Two Phase Fluids................................................................................... 4
6.2.2 Flow Regimes ............................................................................................. 4
6.2.3 Calculating Pressure Drop in Multiphase Flow ......................................... 6
6.2.4 Multiphase Flow Simulators ...................................................................... 7
6.2.5 Well Design .............................................................................................. 10
6.2.5.1 General Modeling Strategy .................................................................. 10
6.2.5.2 Injection Modeling .............................................................................. 11
6.2.6 The Low Head “Operating” Window....................................................... 12
6.2.7 Gas Injection Methods ............................................................................. 14
6.2.7.1 Standpipe Injection .............................................................................. 15
6.2.7.2 Parasite String Injection....................................................................... 16
6.2.7.2.1 Pressure Gradient ....................................................................... 17
6.2.7.3 Concentric Casing Injection ................................................................ 17
6.2.7.4 Concentric Drill Pipe Injection ............................................................ 18
6.2.8 Foam Drilling ........................................................................................... 19
6.2.8.1 Pressure Gradient ................................................................................. 22
6.3 Equipment ..................................................................................................... 22
6.3.1 Gas Injection Options and Equipment .......................................................... 22
6.3.1.1 Natural Gas .......................................................................................... 22
6.3.1.2 Nitrogen ............................................................................................... 23
6.3.1.2.1 Membrane Nitrogen ................................................................... 23
6.3.1.2.2 Cryogenic Nitrogen .................................................................... 24
6.3.1.3 Exhaust Gas ......................................................................................... 25
6.3.2 Degassers and Separators .............................................................................. 27
6.3.2.1 Types of Separators ............................................................................. 27
6.3.2.2 Separation Equipment Selection and Design....................................... 28
6.3.2.2.1 Separator Sizing ......................................................................... 29
6.3.2.2.2 Vessel size considerations .......................................................... 29
6.3.2.2.3 Separator lines sizing ................................................................. 30
6.3.2.2.4 Mud leg calculation .................................................................... 31
6.3.3 MWD Tools .................................................................................................. 32
6.3.3.1 Data Transmission ............................................................................... 33
6.3.3.2 Downhole Tool Limitations................................................................. 34
6.4 Operations & Well Control ........................................................................... 34
6.4.1 Impact of Drill String Connections .......................................................... 34
6.4.2 Well Control in Low Head MPD .................................................................. 35

Chapter 6
MPD Pressure
Control Options

Friction Pressure
Mechanical lift
device
Circulating Non
SMD (Mud Lift)
Semi-automatic
CCS (CBHP)
Concentric Mud Cap
injection
Floating
multi-phase)
CAPM
gradient)
6.1 Introduction
Low pressure reservoirs often represent challenges to drill. Usually, along with the pore
pressure reduction comes a reduction of the fracture pressure, likely to create a narrow
window scenario. This situation is increasingly common in mature fields, where
operators are finding more problems to drill nearly every well.
Some of those MPD candidates require the injection of a gas to lighten the fluid column
in order to sufficiently decrease the hydrostatic pressure. In these cases, all of the rigour
of underbalanced well design must be applied to a low head MPD well, except for the
fact that hydrocarbons are not flowing to surface. A fundamental understanding of multi-
phase flow fundamentals is therefore essential.
6.2.1 Multiphase Flow Fundamentals

Having reviewed briefly single-phase incompressible flow modeling, we are now in a
position to appreciate the rationale and development of multiphase flow models.
When multiphase flows began receiving interest (primarily in Nuclear and Chemical
Engineering applications), the strategy was to use the single phase flow modeling logic
and approach, reducing the problem to one of finding the appropriate Reynolds number
for the flow. Since density, phase and viscosity change with pressure and temperature in
multiphase flows, this is not as straightforward as it sounds. Also, flow regimes in

Chapter 6
multiphase flow are more complex, and do not lend themselves to simple
characterizations such as laminar and turbulent flow. Experimental and theoretical work
have shown that flow regimes are far more complex in two- and multiphase flow, and are
dependent upon the angle of inclination as well as the phase behavior of the fluids.
6.2.1.1 Two Phase Fluids

“Gasified” liquids are a combination of liquid and gas where the liquid is the continuous
phase. Gasified liquids do not have surfactants that bind the gas within the liquid.
The main advantages of gasified liquids are the lower cost, the ability to recycle the
liquids, better environmental properties and simplicity. Foam systems also create a lower
frictional loss compared to a comparable foam system.
The main disadvantage of the foam system comes from the fact that the gas is not bound
within the liquid. This allows the gas to segregate from the liquid and create slugs. The
gasified liquids have a much lower solids carrying capacity.
Either fresh water or oil may be used as the base liquid of a two phase fluid. The various
gases used to lighten the fluid column: Nitrogen, CO2, Natural Gas, or Exhaust Gas. The
main criteria used in the selection of gas are cost, corrosion, safety, availability, and
logistics. The most commonly used gas is nitrogen, either cryogenic or membrane.
6.2.2 Flow Regimes

A simplistic illustration of the flow regimes in a two-phase vertical upflow is given in
Figure 1. Assume that a two-phase system begins flow from the bottom of the conduit
and flows towards the top. It is obvious that the pressure reduces as flow progresses
upwards. As long as the pressure in the conduit is higher than the bubble point of the
two-phase fluid, the fluid stays in a single phase and acts as a single-phase fluid (although
the fluid is now compressible). This flow regime is represented as SP in Figure 1. When
the pressure falls below the bubble point, the gas phase escapes, and the flow regime is
characterized by gas bubbles dispersed in a liquid phase. This is referred to as the
“bubble flow” regime, represented in the figure as BUB. As pressure falls further, the gas
bubbles expand, coalesce and alternate slugs of gas and liquid characterize the flow
regime. This is referred to as “Slug Flow” and is represented by SLUG in the figure.
Finally, on further reduction of pressure, the gas phase occupies most of the cross section
of the conduit with the liquid phase forming a thin film along the wall and dispersing
small droplets in the continuous gas phase. This is referred to as an Annular Mist flow
regime (see AM in Figure 1).

Chapter 6
AM
SLUG
BUB
SP
Figure 1 - Flow Regimes in Two-Phase Vertical Upflow

Clearly, the flow behavior in each of these regimes is entirely different in nature, and has
to be modeled as such. To do so, of course, one has to recognize the flow regimes
themselves, as a function of some intrinsic physical parameters of the flow and geometry
(analogous to the way in which the single phase flow regimes are recognized as functions
of Reynolds number).
In horizontal flow, the flow regimes are, as would be expected, different from those in
vertical flow. Flow regimes such as wavy, slug, stratified, and annular mist have been
recognized in horizontal flows. Figure 2 depicts the flow regimes normally associated
with multiphase flows in horizontal conduits. Flow regimes are ever refined and
improved by workers in the area, through experimental and theoretical work.
Stratified
Figure 2 - Flow Regimes in Horizontal Multiphase Flows

Chapter 6
6.2.3 Calculating Pressure Drop in Multiphase Flow

Starting from the energy equation, it is clear that the terms contributing to the pressure
loss, namely gravitational, frictional and acceleration, are all more complex to model in a
multiphase flow. The simplistic pressure relationship is illustrated in Figure 3
SP = BHP –∆Pgrav – ∆Pfric - ∆Pacc
BHP
Figure 3 - Relationship between Bottom Hole and Surface Pressure in the Annulus (Flow Direction
Upwards)
For instance, in multiphase flows with gas and liquid components, the gravitational or
“hydrostatic” term, which takes into account the density effect on pressure, depends upon
the pressure, temperature and phase behavior (or PVT behavior) of the components in
each phase. This means that multiphase flow models must incorporate a PVT model.
The simplest PVT model is of course, the ideal gas law for gas phase, and an
incompressible assumption for the liquid phase. The PVT behavior in complex non-
reactive hydrocarbon mixtures is complicated further by the need to account for the
differing phase behavior of each component in the non-reactive mixture. Several more
sophisticated models have been developed for multiphase flow, primarily for
hydrocarbon liquid and gas components.
Gas
Liquid
Figure 4 - Liquid Holdup (Area Occupied by Liquid  Pipe Area)
Liquid hold-up is the proportion of conduit cross section occupied by the liquid. It is
defined as the ratio of the area of the conduit occupied by the liquid phase to the total
flow area of the conduit, as illustrated in Figure 4. In addition, in multiphase flows, gas
may travel at a greater relative velocity than the liquid, causing a slip between gas and
liquid. This means that a smaller area of the pipe is required to maintain mass balance of
gas (given mass rate of gas requires smaller area at higher velocities). This has the effect
of increasing the liquid held up at that location, and therefore the average density at that
location. This in turn has the effect of increasing the bottom hole pressure. Multiphase

Chapter 6
flow models must have a means to model liquid holdup accurately, and liquid holdup
predictions are a major goal of multiphase flow models.
The liquid and gas fractions are related by
hL = (1 − hg ) (1)
where hL is the liquid fraction (or liquid holdup), and hg is the gas fraction. And the
density of the mixture of gas and liquid phase is given by
ρ m = ρ L hL + ρ g (1 − hL ) (2)
where ρ L , ρ g , and ρ m are the liquid density, gas density and mixture density,
respectively. Clearly, from the earlier discussion of flow regimes, liquid holdup depends
upon the flow regime.
The frictional term can be calculated using familiar single-phase arguments, but this too
has to be fine-tuned to the flow regime, as well as temperature and pressure.
The acceleration effect, usually ignored in single-phase flows, could play a role in
multiphase flows, due to the presence of gas.
In general, hydrostatic (or gravitational) pressure dominates real wellbore flow pressure
losses. This means accurate calculation of liquid holdup is critical to a reasonable
estimate of the pressure gradient, and hence the bottom hole (or surface) pressure.
6.2.4 Multiphase Flow Simulators

Due to the complexity of this process, underbalanced flow modeling is usually performed
using multiphase flow computer programs. The typical goal of these programs is to
predict the pressure (and temperature) profile in a multiphase fluid flowing in a conduit.
From the foregoing discussion, it is apparent that a complete multiphase flow program
should include the following models (at a minimum):
Model for flow regime (or flow map) prediction, which predicts the two-phase flow
regime at any given location in the conduit
Model for liquid holdup, which calculates the amount of liquid held up at any given
location in circulating two-phase flow
Model for two-phase flow frictional pressure loss. Although hydrostatic pressure loss
dominates the total pressure loss, frictional pressure drop is an important contribution to
the bottom hole pressure, and should be accurately modeled.
Model that considers behavior of foams

Chapter 6
Thermal model, a PVT model, and a set of property correlations to allow the program to
account for the effect of temperature and pressure on the phase and properties of the
fluids
Model to predict hole cleaning performance and requirements
Several such programs are available commercially. Most of these, however, have been
developed to study upflow during production, or the flow of multiphase fluids in surface
production facilities. Very few of these programs can be readily adapted to study
underbalanced/managed pressure drilling. Among commercially available programs, the
Wellflo7 program developed by Neotechnology Consultants, Dynaflo Drill by Rogaland
Research, Hubs by Signa Engineering, and Mudlite by Maurer Engineering, are the most
common. Their sophistication and performance vary widely, and none of them has all the
components listed above for a complete underbalanced drilling model. Some of them
have very few multiphase flow correlations available, while some have proprietary
correlations included. Care must be exercised in the selection and use of drilling
simulators.
Blade Energy Partners uses the Wellflo7 program from Neotechnology Consultants.
Wellflo7 is one of the most sophisticated multiphase flow programs currently available
for underbalanced drilling modeling on the market. It originated in the pipeline
multiphase flow models developed by Neotechnology Consultants, and has been
enhanced over the years to be particularly user-friendly for underbalanced drilling
applications. Wellflo7 has several multiphase flow correlations available, including Olga
(it is the only commercially available underbalanced drilling simulator, to our knowledge,
that has licensed Olga). It also has sophisticated PVT and property models, as well as a
thermal model. It includes different pre-defined elements to model chokes, BHAs, and
point sources. It explicitly allows inclusion of inflow, which is calculated based on either
user-provided inflow performance relationship, or one of several in-built IPR equations.
The density of the fluids, the gas-to-liquid ratio and the flow rate of the circulating fluids
(both injected and produced) influence annular bottom hole pressure in a multiphase
circulating system. One purpose of multiphase flow modeling is to investigate the effect
of these parameters on the bottom hole pressure and determine if it is possible to create
underbalanced conditions. This is best illustrated by an example. Figure 5, below, shows
an example of a curve plotting annular bottom hole pressure versus varying gas
injection/inflow rates, for a fixed annulus surface pressure and a constant liquid injection
rate. Injecting varying amounts of gas varies the density of the injected fluid.

Chapter 6
Impact of Gas Injection Rate on Bottomhole Pressure
2500
Bottomhole Pressure (psia)
2000
1500
1000
Hydrostatically Friction
500 dominated Dominated
0
0 250 500 750 1000 1250 1500
Gas Injection Rate (SCF/min)
Figure 5 - Impact of Gas Injection on Bottom Hole Circulating Pressure
This figure shows that for a constant liquid rate, the annular bottom hole pressure at the
bit initially decreases as gas is introduced into the circulating system. The introduction of
gas clearly decreases the density and hence the hydrostatic pressure of the injected fluid.
However, both increasing fluid flow friction and decreasing liquid fraction reduction in
the fluid gradually offset this drop in hydrostatic pressure. Thus, the rates of change of
bottom hole pressure decreases as the gas rate increases. This portion of the curve is
referred to as being “hydrostatically dominated”.
As gas rate is increased, the rate of change of bottom hole pressure decreases on the
hydrostatically dominated portion of the curve. The lower the rates of change of bottom
hole pressure, the more stable the circulating system. It is important to note that an
unstable circulating system will frequently slug and surge, and experience periods of no
annular flow as the well loads up and subsequently unloads. This instability can create
large oscillations in annular bottom hole pressure.
Ultimately, the rate of change of bottom hole pressure becomes zero, and the annular
bottom hole pressure reaches its minimum. Increasing the gas rate can no longer reduce
bottom hole pressure. The increase in frictional pressure caused by higher velocities of
the fluid becomes greater than the decrease in the density of the fluid mixture.
Consequently, the system becomes “friction dominated”. From this point, an increase in
the gas injection rate will actually increase the annular bottom hole pressure.
By varying the surface pressure, the density and the injection rate, it can be determined if
it is feasible to achieve the desired circulating pressure window for the given pore
pressure / frac gradient curve. This is usually a primary goal of multiphase flow
modeling.

Chapter 6
6.2.5 Well Design

Design of multi-phase low head systems are almost identical to UB systems, except for
the fact that hydrocarbons are not flowing into the wellbore. Multi-phase flow
complexities do require that the same rigor be followed in a Low Head design as for a
UB design.
Complete details of multi-phase flow modeling can be found in literature1; however, the
culmination of the design is the operating envelope shown later in this chapter. .
6.2.5.1 General Modeling Strategy

Figure 6 gives a simple flowchart for the general strategy in modeling. The basic goal in
low head modeling is to seek a range of operating parameters (gas and liquid injection
rates) such that the following conditions are satisfied at a minimum:
Bottom hole circulating pressure is above pore, within an acceptable range

Adequate hole cleaning is achieved
Injection parameters are within the specifications of injection equipment
Equivalent flow through BHA is within their limits
1
Advanced Underbalanced Well Design Manual, Blade Energy Partners.

Chapter 6
Determine Reason
For MPD
Input Surf
Pres. Gas
& Liq.
Rates Modify
Annular Parameters
Specify
Modeling (Rates, Fluid
OR Equipment
Type, etc) and
Input re-run
BHP, Gas
& Liq.
Rates
BHP Surface Gas rates water
Pressure. rate hc rates
Minimum Liquid BHP
Velocity
Are
Is overbalance &
overbalance, hole
Minimum Velocity
cleaning, & motor
Acceptable?
Acceptable?
Injection
Pressures and
Model Injection Equivalent Flow
Pressures thru motors
Figure 6: Basic flowchart for Low Head modeling

Although flow is down drill pipe, around the bit, and up the annulus to the choke, it is
customary to conduct the analysis in two parts- the annular side, and then the drill pipe
side, with a pressure continuity condition imposed at the bit. Analysis generally begins
with the consideration of upflow in the annulus. The questions asked in this analysis are:
For a given backpressure at the choke, what would be the bottom hole circulating
pressure (at different combinations of gas and liquid injection rates)?
For a given desired bottom hole circulating pressure, what is the backpressure (for
different combinations of gas and liquid injection rates)?
Hole-cleaning performance may be assessed using any of the methods described earlier. .
6.2.5.2 Injection Modeling

Once a reasonable operating range is identified such that low head is achievable with
sufficient hole cleaning, the injection side is modeled (for the operating range just
identified). In injection modeling, the goals are to ascertain the injection pressures (and
thus the injection equipment rating) needed for the preliminary range of operating
parameters, as well as to ensure that the equivalent flow through motors and other BHA
components is not beyond (or below) the recommended range. Since the achievable BHP
range has already been determined, injection analysis can proceed. In practice, the

Chapter 6
highest likely circulating BHP and the highest likely injection rates are considered to
determine injection requirements (as they are likely to place the highest injection pressure
requirements). Injection modeling is also a subset of multiphase flow analysis, except that
the pressure regimes in the drill pipe are typically higher than in the annulus (thus
reducing the in-situ gas to liquid ratios). Moreover, liquid hold-up is more easily
determined, as the flow is downward (in the direction of gravity). Therefore, injection
analysis is typically more forgiving of multiphase flow correlation choices than annular
“upflow” analysis.
Typically, motor limits are specified in a given MPD operation. If they are not, injection
analysis allows the specification of the desired equivalent flow for the motor. Figure 7
illustrates the limits on surface injection rates, for a given motor limit (in this case,
300 gpm maximum). The maximum allowable gas injection rate is clearly a function of
the liquid rate, as well as the pressure at the motor inlet (which is a function of the
injection pressure). The relationship is linear in the figure, as the ideal gas law has been
used to model the gas behavior with temperature and pressure.
Max Gas Injection Rate
3500
3000
300 GPM Motor Equivalent Flow Limit
BHP = 1200 psi, Motor and BHA Losses = 500 psi
Gas Injection Limit (SCFM)
2500
2000
1500
1000
500
0
0 100 200 300 400 500 600
Liquid Injection Rate (gpm)
Figure 7: Illustration of Surface Injection Rate Limitations for a 300 GPM Motor
6.2.6 The Low Head “Operating” Window

As the flow chart indicates, the process described above is iterative, since a range of
liquid and gas rates may have to be considered to satisfy the requirements for low head.
Once a reasonable range of operating parameters has been identified, they can be
presented in a single chart, as shown in Figure 8. This has become the canonical way of

Chapter 6
presenting the results of flow modeling. The figure is specific to the example problem
considered later in this manual. It shows the variation in the bottom hole pressure, as a
function of injection gas and liquid rates, with the reservoir pressure and target maximum
bottom hole circulating pressure while MPD also shown. Also in the figure are the
minimum gas/liquid rates allowable to obtain sufficient annular liquid velocity to achieve
adequate hole cleaning (based on the rule of thumb discussed above), and the maximum
allowable gas/liquid rates to be within the motor limits. It is also possible to indicate on
the same graph the minimum allowable liquid/gas rates to satisfy the minimum motor
requirements, and a minimum allowable circulating BHP (which can be set, for instance,
by the maximum draw down limitations for the formation being drilled). With all this
information on the same graph, it becomes possible to identify a “low head operating
window”, and this is the first step in the modeling exercise for any low head operation.
For instance, in Figure 8, we can say that to be within the motor limits and to achieve
adequate hole cleaning, while still being within the motor limits and staying in the
“frictionally dominated” region of the curves, the low head operating window indicates a
gas rate between 500 scfm and 800 scfm, with a liquid rate between 100 gpm and
125 gpm. Complete details of multi-phase flow modeling can be found in literature2;
however, the culmination of the design is the flow envelope, as shown.
2
Advanced Underbalanced Well Design Manual, Blade Energy Partners.

Chapter 6
1600
50 gpm 75 gpm
1500
Mud pump curves
1400 100 gpm 125 gpm
1300 150 gpm

Bottom Hole Circulating Pressure (psi)
1200
1100
1000 Operating window

Upper circulating
900 pressure limit
800
700
600
Reservoir Pressure
500
400
300
Min. motor limit
200
Max. motor limit
100 Hole cleaning limit
0
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
Gas Injection Rate (scfm)
Figure 8: Low Head operating window, depicted in the grey shaded area bounded by hole cleaning limit
and maximum motor throughput, as well as upper and lower pressure boundaries.
6.2.7 Gas Injection Methods

At the design stage, excluding foam drilling, there are principally four methods to
consider in order maintain the BHP low enough to be just above a low formation
pressure. These methods are the same as those adopted for underbalanced drilling:
Standpipe injection,
Parasite string injection,
Concentric string injection
Concentric drill pipe injection.
Each method has advantages and disadvantages.
The overall benefits, cost and risk, along with the impact on bottom hole pressure and
annular velocities, must be considered in selecting the appropriate injection method.

Chapter 6
6.2.7.1 Standpipe Injection

A conventional method for managed pressure drilling is to mix the gas and liquid at
surface and inject the mixture down the drill string. This method requires the fewest
modifications to the well design, but can impact MWD data transmission up the drill
pipe.
For multiphase systems, the presence of gas in the drill pipe will hamper or eliminate the
signal from a conventional MWD. Since gas is a compressible fluid, it tends to dampen
out the pulsed signal. For gas ratios above 20% to 28%, the signal will be lost. This
problem can be overcome by using Electromagnetic MWD or wired telemetry. EMWD
sends the signal through the earth to surface instead of through the fluid column. It is not
impacted by the gas in the drill string, and makes the system better suited to very low
head wells where higher volumes of aerated injection fluid are required. It is however
impacted by salt formations, and their presence must be verified during the design stage.
Wired telemetry uses couplings in the tool joints linked by data cables inside the pipe.
Boosters can be inserted into the string as required to boost the transmitted data signal. A
transmitter on the MWD then takes signals from the MWD tool and transmits them
through the couplings and wire to a received in the top drive. This provides for
continuous data transmission at high rates even while the mud pumps are off.
Standpipe injection method means gas will pass through the BHA. Therefore, motor
selection needs to consider the equivalent flow rate through the motor since motor
performance will be reduced when compared to single phase liquid injection. Further
consideration must also be given to the type and size of stator rubber and the impact of
explosive decompression.
Injecting gas down the drill pipe tends to loosen and slough mud and scale inside the drill
pipe. It can also become entrained into a protective coating the pipe may have. When
pressure is relieved on the drill string (as for a connection), the entrained gas will rapidly
expand and separate the coating from the drill pipe wall. The fact that gas will be pumped
down the drill pipe also requires a gas tight connection.
Connections take longer when using standpipe injection. It is necessary to bleed off all
charged pressure within the drill string prior to breaking the connection. Modifications
may need to be made to the rigs standpipe manifold, or a purpose build injection / bleed
off manifold fabricated. The bleed down line must vent the gas to a safe area. This is
normally to the separator or the flare line. Drill string floats positioned at intervals in the
string reduce the volumes of gas bled off, as well as the time to bleed off the stand pipe
manifold.

Chapter 6
Figure 9: A photo showing a parasite gas injection string clamped to the outside of casing. The gas
injection port can be seen at the left of the photo.
6.2.7.2 Parasite String Injection

Parasite string injection utilizes a sacrificial small diameter string as the conduit for gas
injection. The parasite string is run on the outside of the casing string to the gas injection
point. The parasite string is normally not run into the open hole, to prevent damage to the
string.
When designing a parasite string injection well the following issues should be
considered:
The depth of the injection port is optimal to ensure the well can be drilled near balanced.
It has to be deep enough such that excessive rates of gas injection are not required to
reduce the BHP to the required value. Simulations must be conducted on the I.D. of the
injection string and what injection pressures will be required. As the gas is contained
within the string, it is possible to run the injection point below the top of cement. If the
required BHP cannot be achieved, then simulations must be conducted on parasite string
I.D., surface pressure injection rates, setting depths. The string I.D. may start to affect the
choice of casing I.D. and setting depths.
It is normally deemed prudent to install a side entry sub with a back pressure valve. This
prevents fluid from entering the gas injection line. To implement this form of MPD, it is

Chapter 6
necessary to plan ahead with regards the previous casing string and well head
modifications. As the parasite string is run as a permanent fixture, concern exists over the
creation of a leak path if the backpressure valve or a leak in the string occurs in the
future.
The main advantage of the parasite injection method is that only liquid is pumped down
the drill pipe. This allows the use of conventional MWD tools and BHA, and eliminates
concerns over material selection and power generation for the motor. It also allows for a
more direct measurement of BHP and drill string problems from surface.
Connections are also quicker with parasite string injection systems. As the drill string
only contains a single-phase non-compressible fluid, the time to bleed off the string is
reduced considerably. The use of a parasite string also eliminates the effect of gas on
scale and dried mud on the drill pipe. This eliminates the requirement to condition drill
pipe prior to drilling
.
If the fluid used is kill weight fluid, a well can be brought under control by stopping the
injection of gas. If desirable, it is possible to continue gas circulation during trips and
connections.
A major concern of the parasite injection method is the possibility of damage to the
parasite string. The parasite string can be crushed or crimped during the running of the
casing. A leak in the string at one of the termination points or in the string is also a
concern. There is also the possibility that the string can be plugged. In either case, the
damage may cause this method to become ineffective.
6.2.7.2.1 Pressure Gradient

A major factor in parasite string well design is that the gas enters the wellbore at a point
off bottom. This limits the degree of pressure reduction that can be obtained by the use
of this method. Due to the compressibility of gas, the impact of injecting the gas off
bottom is limited as long as the injection point is below 5000 feet. Having the injection
point below this point will impact the bottom hole pressure, but only marginally.
A major impact of injecting the gas off bottom will be the impact on annular velocity and
hole cleaning. As the gas is injected off bottom, the total fluid out of the bit is lower
from a given target bottom hole pressure. This will lower the annular velocity in the
open hole and may create a hole-cleaning problem. This issue may be dealt with by
increasing both the gas and liquid volumes pumped, which may increase the cost of the
project (gas requirements, chemical additions, etc.).
6.2.7.3 Concentric Casing Injection

Concentric casing injection is the technique of pumping the gas down an un-cemented
casing-casing annulus. The gas enters the wellbore at the base of the un-cemented string.
The un-cemented string may be an inner string of casing run for the purpose of creating
the annulus or may be an un-cemented tieback casing string. Gas may enter the wellbore

Chapter 6
at the base of the casing, through a gap left between the tieback and the liner, through a
perforated joint near the base of the tieback or through an entry sub containing a
backpressure sub. In most cases, the un-cemented casing string is removed from the well
following the MPD activities.
The advantages and disadvantages of utilizing concentric casing injection are similar to
parasite injection include:
Conventional MWD can be used
Conventional motors can be use with no impact on power
Eliminates sloughing scale and dried mud
Faster connections
Limited depth of injection
Cost
Impact on BHP and Annular velocities
Possibly higher gas volume requirements
Concentric casing injection has the additional advantage that concerns over damage or
loss of the injection method are minimized and the elimination of the need to modify the
wellhead. Injection pressures are also generally lower utilizing concentric string
injection. Due to the smaller hole size created by the concentric strings, hole clean is
improved.
The main disadvantage not seen with parasite string injection is the cost of pulling the
casing string upon completion. Concentric casing designs have the additional problem of
severe pressure oscillations during and after connections if a backpressure valve is not
used. The volume between the casing strings acts as an accumulator that can cause
pressure oscillations for up to an hour after the connection.

The issues of pressure gradient and annular velocity are the same as parasite string
injection.
6.2.7.4 Concentric Drill Pipe Injection

Concentric drill pipe strings are designed with two separate flow paths that allow gas and
liquid to be pumped down the drill pipe in separate pipes. Gas enters the wellbore at a pre
determined point above the BHA through a cross over with an injection port. The drill
string is then crossed over to conventional drill pipe and BHA below the injection port.

Chapter 6
Figure 10 - Concentric drill pipe injection

The advantages of concentric drill pipe design are similar to parasite and concentric
string design. In addition, no permanent changes to the well are required and no long
term impact on the life of the well.
System cost is one of the main disadvantages of concentric drill pipe design. Special drill
pipe is required. In addition, modifications to the surface injection system, including a
special Kelly or top drive are necessary. For this reason only a few concentric drill pipe
projects have been undertaken. There is very little concentric drill pipe available in the
world.

The impact on the pressure gradient and annular velocity is dependent on the depth
selected for gas entry into the annulus. If the point selected is near the bit, the pressure
gradient will be similar to drill string injection. If the point selected is further up the
hole, the pressure gradient will be similar to parasite string or concentric string injection.
6.2.8 Foam Drilling

Foam drilling requires the generation of foam on surface that is then injected down the
drill string. Invariably the foam is created by using a mist pump to inject the base fluid,
membrane or cryogenic units to inject the nitrogen gas, and a chemical pump to inject the
foaming agent (surfactant).
In modeling foam, advantage is taken of the empirically observed fact that the apparent
viscosity of foam is strongly correlated to the foam quality, which is defined as the ratio

Chapter 6
of the gas phase volume to the total volume. The typical relationship of the apparent foam
viscosity to the quality is shown qualitatively in Figure 11. The figure shows the increase
in the apparent viscosity in terms of an apparent viscosity ration, the ratio of the foam
apparent viscosity to the basic 2-phase viscosity of the underlying fluids. As can be seen,
for qualities less than 50%, no advantage is gained from the foam, as its apparent
viscosity is the same as that of the underlying two-phase system. As quality increases
beyond 50%, the impact of foaming the fluid is clear. For qualities between about 80%
and 90%, the apparent viscosity can be an order of magnitude higher than the two-phase
viscosity, giving foam its remarkable ability to carry cuttings (but also to add to the
pressure loss). As the quality of the foam increases, the carrying capacity of foam will
increase, to a limit of +/- 97.5%. At this point, the foam will break down into a mist and
the carrying capacity will drop significantly.
It must be noted that the figure below is only qualitative, and the exact relationship
between the apparent viscosity and quality depends upon the individual foam, the
pressure and temperature regime, and the impact of pH and contaminants.
Apparent viscosity ratio
Max viscosity close to 90%

(µfoam/µ2-ph)
2-ph viscosity Gas-phase

viscosity
1
50 75 100
Quality % (gas volume / total volume)
Figure 11 - Relationship between apparent viscosity and quality of a typical foam
The use of foam as a drilling fluid presents a further set of problems with regards
generating foam on the injection side and breaking the foam on the returns side.
Flow modeling for MPD, as with conventional drilling, must also consider the injection
side of the well. For most single phase MPD operations, the standpipe injection pressures
will be similar to conventional drilling. Greater attention must be given to modeling when
injecting a second fluid phase down a concentric casing string or parasite string.

Chapter 6
Figure 12 – Foam generation using mist pumps, surfactant totes and cryogenic N2.
The surfactant binds the gas within the structure of liquid. As the gas is bound within the
liquid, the gas and liquid move together. This increases the velocity of the liquid. Due to
the structure of foam, it has a tremendous solids carrying capacity. Foam has better hole
cleaning characteristics than a conventional mud. Due to these characteristics, the
minimum velocity for hole cleaning is set at 100 feet/min. Hole cleaning with velocities
as low as 30 feet/min has been successful over short distances.
Foam has some disadvantages. In addition to the cost of the surfactant and the additional
equipment required, foam adds a degree of complexity to the system. Foams can break
down with contamination from acid gasses, hydrocarbons and brines. Since there is
always the chance of one of these entering the wellbore, selection and testing of a foam
system is critical. Foams also lose their stability with temperature. A further disadvantage
in foam systems is the higher relative viscosity. This increases the frictional pressure loss
in the wellbore.
Breaking the foam structure must be accomplished when it reaches the surface to
effectively separate the fluid constituents (gas, water, liquid hydrocarbons and solids).
Failure to effectively break the foam will cause separation problems and will overload the
separator. The foams can either be broken chemically or mechanically. In the past, foam
systems could not be re-generated after they are broken. This caused high disposal costs
and environmental concerns. The introduction of recyclable foams has eliminated many
of these concerns.

Chapter 6
Most foam systems are anionic. This precludes the uses of cationic corrosion inhibitors.
Foam systems can be further defined as stable foams, which contain only water gas and a
surfactant, and stiff foams, which include gels and polymers to increase the viscosity of
the system. Foam boosters and foam stabilizers can also be added to the system to
increase the half-life of the foam.
6.2.8.1 Pressure Gradient

Unlike the previous four examples of low head MPD drilling, foam drilling provides a
linear pressure gradient in the well bore. This pressure gradient can be adjusted by
applying a different surface pressure at the MPD choke manifold.
6.3 Equipment
6.3.1 Gas Injection Options and Equipment
When required annular pressure profile for a MPD operation is lower than what is
achievable with pure liquid columns, one of the options to reach a lower hydrostatic head
is to mix gas with the drilling fluids at some point in the wellbore, either injected through
the drill pipe via the standpipe, or in the annulus via parasite string or concentric casing.
This section covers the gas injection equipment related to the low head MPD
applications.
6.3.1.1 Natural Gas

Natural gas in a pure form is an ideal fluid for low head systems. Since the system is free
of oxygen, it is also relatively safe to use as hydrocarbon gases are not flammable without
oxygen. If natural gas is available at the site it should be considered for use as it is
generally the lowest cost system available, especially if a gas recovery system is
employed. If the pipeline pressure is not high enough to inject straight into the standpipe,
a high pressure gas compressor will be required. These compressors require the gas to be
dry and thus often additional scrubbers are required to ensure all liquid is knocked out
prior to compression.
Besides the cost advantage of natural gas it also has the advantage of extending the
operability of mud pulse MWD equipment. If crude or diesel are proposed drilling fluids,
the natural gas will saturate the crude (gas goes into solution) thus minimizing the free
gas in the drill pipe. This will lower the gas volume fraction and increase the operability
of the MWD equipment. For these two specific reasons it is suggested that the option to
utilize natural gas for drilling first be pursued. If natural gas is ruled out, due availability
or safety concerns (i.e. coiled tubing applications), then more costly options must be
evaluated.

Chapter 6
6.3.1.2 Nitrogen
There are two common sources of Nitrogen used for low head operations: Cryogenic and
molecular filtration. The selection of either one will be based on availability, cost per
volume delivered, expected volume of gas to be required.
6.3.1.2.1 Membrane Nitrogen

The most commonly used gas source for low head drilling is membrane generated
nitrogen. Membrane systems simply obtain the required nitrogen on location from air. A
complete membrane generation system consists of feed compressors, a cooling system
(ambient temperature dependant), the membrane units, booster compressor and
depending on the injection pressure (typically above 2000 psi), a high pressure booster
compressor may be needed. A recent design, trailer-mounted Nitrogen membrane system
is shown in Figure 13. This particular, compact design combines the feed compressor, the
Nitrogen unit and the booster compressor.
Figure 13: Trailer-mounted Nitrogen integrated unit, courtesy of Optimal Pressure Drilling.
Membrane technology utilizes the differential rates of molecular diffusion of gases

through semi-permeable fibers to separate enough of the oxygen from the air to result in a
non-combustible product stream (Figure 14). The advantages of this system are the lower
per-unit cost for operations that will require large volumes of nitrogen, and logistics.
Once the initial investment has been made, a membrane system can generate large
volumes of nitrogen at a continuous rate for a low per-unit volume cost. The system is
portable and can be moved from location to location with the rig. This eliminates
concerns with movement of large volumes of materials during the drilling program.

Chapter 6
Figure 14: Contents of a nitrogen membrane filter cartridge. Nitrogen molecules are allowed to pass
axially through the filter, while trace gases (like oxygen) are expelled radially.
The system is approximately 50% efficient. This means that feed compressors must
produce twice the volume of air that is required.
The disadvantages of this system are the large mobilization cost, the size of the
equipment, maintenance and the presence of oxygen. The size of the system and the
number of components are a concern with a membrane system. Oxygen is another
concern with membrane separating systems. Although the oxygen can always be set
below the explosive limit for a sweet well, there is still some oxygen in the system
(usually 3% to 7%). If a membrane system is employed, a corrosion inhibitor system
must be available in the case of water production. Even low concentrations of oxygen can
be extremely corrosive.
To meet the continuous demand of a drilling program, especially in remote well

locations, it is recommended that 100% redundancy be provided for routine maintenance
of compressors and equipment failure.
6.3.1.2.2 Cryogenic Nitrogen

Liquid nitrogen is an ideal candidate for the gas phase in mixed fluid system. It is inert,
safe and easy to use. Due to the advantages, it has been utilized in many mixed fluid
systems for UBD and MPD. Liquid nitrogen is usually transported and stored in bulk
tanks. A liquid converter system is utilized to convert the liquid to a gas at the pressure
required for the fluid system, Figure 15. The liquid converters are equipped with flow and
pressure meters to accurately control the flow of nitrogen.

Chapter 6
Cryogenic nitrogen is
transported to the well site as
a liquid. Cryogenic tanks are
necessary for transporting the
liquid nitrogen to location,
because the boiling point of
nitrogen (at atmospheric
conditions) is -321°F. The
pumping unit consists of a
diesel driven, positive
displacement pump and a
heat exchanger. The liquid
nitrogen is pumped from the
cryogenic tank through a heat
exchanger that evaporates the
liquid, and discharged as an
80°F to 120°F Gas. Small
Figure 15: Cryogenic nitrogen operator with unit. Note the ice build
up on the liquid nitrogen transfer lines and pumps due to evaporative
units typically are able to
cooling. deliver 1,100 scfm at
pressures up to 3,000 psi, but
as the delivery pressure increases towards the unit’s pressure rating of 4,500 psi, the
delivery rate will fall. Larger units are capable of delivery rates to 6,000 scfm at pressures
up to 8,000 psi.
The major disadvantages to the system are the availability, logistics, transportation and
cost. Care must be taken in storage as evaporation may result in a 20% reduction in
stored volume per day. Unless cryogenic nitrogen is readily available in the field, or a
small volume of nitrogen is required for start-up, it is not likely an economically viable
alternative to membrane generated nitrogen.
Cryogenic nitrogen spills are very dangerous, because of its extremely low temperature.
It may cause severe lesions if becomes in contact with skin or limbs. Also, in contact with
equipment it can cause steel embrittlement, thus seriously damage decks, containers, etc.
6.3.1.3 Exhaust Gas

Northland Energy developed a gas supply package that takes the exhaust from the gas
compressors and uses this as the feed gas. For this, the system compresses the exhaust
gas from a propane fueled engine. Since the engines are normally driven by propane
(opposed to diesel) the exhaust is relatively clean. The exhaust gas is treated to remove
the residual oxygen and hydrocarbons leaving behind an inert gas comprised of 87%
nitrogen and 13% carbon dioxide. Exhaust gas technology eliminates oxygen (due to
combustion), a potential corrosion mechanism, and requires less equipment on location,
typically only a single skid required. Figure 16 depicts exhaust gas system schematics.

Chapter 6
Figure 16. A schematic depicting the flow path through an exhaust gas system. Exhaust
gas systems were used commercially in the Canada in the mid 1990s, however few units
remain in operation due to problems maintain the correct gas mixture ratio for
combustion.
6.3.1.4 Factors of Safety Gas Delivery

Table 1 adjusts the recommended margins for injection equipment. In addition to the
safety margin for pressure the volumune requirements must be set with a factor of safety.
Table 1Recommended Safety Margins, Injection Equipment

Gas 1.5 times Max Injection Used for Membrane generator,
Availability - Rates compressed air, exhaust gas
Rate Based
Gas 2 times Total
Volume Cryogenic Nitrogen. 100%
Availability – needed redundancy is for project
Volume Based availability. Based on lowest
expected ROP. Logistics and
supply will dictate field
availability.
Pressure 1.25 times Max Injection Max Injection Pressure based on
Pressure
Pumps and Unit redundancy On site, or immediate availability
Compressors

Chapter 6
6.3.2 Degassers and Separators

Separation systems are commonly placed downstream
of the flow control choke manifold. They
can be as simple as the widely used
mud/gas separator, present in most
drilling rigs nowadays, or as complex as
the Weatherford (formerly Expro) high
pressure separation system, which
includes high, medium and low pressure
gas and/or solids separation, with
complex instrumentation and
pressure/level automated controls.
6.3.2.1 Types of Separators

There are two basic designs for
separators, vertical and horizontal.
Vertical separators are best for separating gas
from liquid. Horizontal separators are the
optimum design for the separation of liquids of
various densities. Sizing of separators is based
on the rate of fluid pumped, the rate of fluid that
will be produced, type of fluid pumped and
produced, hole size, and footage drilled.
While a vast majority of vertical separators are Figure 17: A cutaway diagram of an
instrumented vertical 2-phase separator.
atmospheric, pressurized vertical separation
systems are not uncommon. However, most pressurized separators are horizontal. They
can include pressure control valves (PCV) and level control valves (LCV). Depending on
the design, they may be able to separate gas from liquids (2 phase separators), or gas
from liquids, and then oil from water (3 phase separators). Some separation systems are
designed to remove also solids from the stream (4 phase separators). These are more
commonly used in UBD operations. Figure 18 depicts the schematics of a horizontal 4-
phase separator.
Separators rely basically in gravity and difference in densities between the different fluids
they are intended to separate. Then, diffusers, baffles, vortex, or other strategies can be
added to the system to help improve the separation capacity.

Chapter 6
Figure 18: A cutaway of a horizontal 4-phase separator. The term ‘4’ phase refers to solids, gases, drilling
liquids, and produced liquids.
All separators are pressure vessels and should be designed and maintained under the
ASME pressure vessel code. The ASME code controls the design, inspection and
certification of pressure vessels. They must be protected from excess pressure by use of
certified pressure safety valves (PSV) and PSV discharge lines must be vented to a safe
area. The separator system must be designed to protect from liquid discharge into the gas
effluent or gas discharge into the liquid effluent.
Regardless of the type of separator used, it must be protected from flow back from the
vent line or flow line. During connections and shut downs, reverse flow back to the
separator can be caused as the system cools, drawing air into the system. This can create
an explosive mixture if hydrocarbons are present. Continuous purging of the line,
backpressure valves, or flame arrestors can be used to protect the vessels.
6.3.2.2 Separation Equipment Selection and Design

Selection of a vertical versus horizontal separator will rely on the maximum potential for
gas flow from the well at any given time, being this gas either injected or produced. If an
eventual underbalanced situation creates the potential for high gas flow, due to high PI,
then a pressurized horizontal separator should be considered. Most single phase MPD
operations are more likely to require a vertical, atmospheric separator.
Some MPD operations flow permanently through the separator, while some others use
the separator only as contingency equipment. These are more likely to be the statically
overbalanced operations. Provisions must be taken to avoid solids build up in the bottom
of the separator, or plugging the lines exiting the separator. Strategies may include
periodical flush, sparge lines and pumps, line size design to maintain a minimum flow
velocity, among others. Special attention to solids build up should be given on operations
flowing permanently through the separator.

Chapter 6
6.3.2.2.1 Separator Sizing

The rules governing the separator sizing are included in API RP 521. Although they are
not specifically designed for drilling separation equipment, the provided guidelines allow
to calculate separator and lines sizes, as well as maximum flow rates, to avoid carryover
(liquid going into the gas discharge), and to avoid blow by (gas going through the liquid
discharge).
6.3.2.2.2 Separator Factor of Safety

Table 2 list the recommended margins of safety for surface handling equipment. Rate
capacity safety margins for the flare are slightly higher than for the separator since
separators could be bypassed with flow directed to the flare in an emergency.
Rate capacity safety margins for the flare are slightly higher than for the separator since
the separators could be bypassed with flow directed to the flare in the event of an
emergency.
Table 2 – Recommended Safety Margins, Surface Returns Handling Equipment

Separator - 1.25 times Maximum Max rates at highest likely draw-
Rate Anticipated Rates down
Separator - 1.25 times System Back High pressure separator in high
Pressure Pressure rate, high pressure reservoirs
Flare - rate 1.5 times Maximum Self-igniting. May have noise and
Anticipated Rates temperature restrictions.
Piping 1.25 times MASP upstream Based on piping friction loss
of separator analysis
1.25 times working pressure Sizing based on minimizing
downstream of separator working pressure (or system back
pressure)
Pumps 1.25 times max working Max working pressure based on
pressure if re-injecting pressure loss and receiving system
Unit redundancy pressure
1.5 times anticipated
maximum flow rates
Solids handling 1.25 times solids rate at
highest ROP
6.3.2.2.3 Vessel size considerations

All separators must guarantee that the gas/liquid mixture stays in the vessel enough time
to allow the gas to break free from the liquid. This time is known as “retention time”, and
is affected by the total flow in the separator, the gas/liquid ratio, the properties (density
and viscosity) of both the gas and liquid, and the size and shape of the vessel.

Chapter 6
6.3.2.2.4 Separator lines sizing

The principles for separator line sizing starts with the Fundamental Flow Equation, which
can be used to predict the pressure drop for a gas, in steady-state, adiabatic, flow along
pipe.
2 2
To ( P1 − P2 )
Q = 77.54 * ( ) * d 2.5 (3)
Po GT f LZ f
Where:
Q = gas flow rate, (SCFD)
L = pipe length, mi
D = inside diameter of pipe, in.
P1 = upstream pressure, psia.
P2 = downstream pressure, psia.
Pb = base pressure, psia (usually 14.7 psia)
Tb = base temperature, °R (usually 60+460 = 520 °R)
Tf = average flowing temperature of gas, °R
G = gas specific gravity (Air = 1.00)
Z = gas compressibility factor at the flowing temperature and pressure, dimensionless
f = friction factor, dimensionless
Weymouth’s equation, Adapted from Bernoulli’s flow equation for pipeline gas flow,
incorporating pipeline efficiency factor.
2 2
T d 16 / 3 * ( P1 − P2 )
Q = 433.5 * E ( b ) (4)
Pb GT f LZ f
Where:
Q = gas flow rate (scf/D)
E = pipe efficiency factor
d = inside diameter of the pipe (inches)
Tb = standard temperature (520 °R)
Pb = standard pressure (14.7 psia)
G = gas gravity (air = 1)
Tf = Flowing temperature
L = Length of pipe (miles)
Zf = average compressibility factor (Weymouth used Za = 1)
P1,P2 = the inlet and outlet pressures (psia)

Chapter 6
6.3.2.2.5 Mud leg calculation

Adapted from Eq. 4, Eq. 5 can be used to calculate the height of liquid U-tube seal
required at the bottom of a mud gas separator.
.58 * Q 2 * T f * L
5.333
+ P22 − P2
d
h= (5)
.433 * g
P1 = .433 * h * g (6)
Where:
h = height of mud leg (feet)
L = length of vent line (feet)
Q = flow rate in MMscf/d
Tf = Flowing temperature in degrees Rankin
d = Internal diameter of the vent line (inches)
P2 = Vent line outlet pressure (psi)
g = Specific gravity in the U-tube

Chapter 6
Figure 19. Example of mud gas separator with long liquid leg for higher rate MPD operations.
6.3.3 MWD Tools

Horizontal drilling and directional control while drilling are widely used technologies,
frequently included in MPD operations. Additionally, since the main objective of MPD
operations is to maintain control of the pressure profile along the hole, it is especially
helpful to have real time readings of the annular pressure from the bottom of the hole.
Any type of MPD application, friction, density, surface pressure or ∆energy will be
greatly supported by having real time PWD readings.
Another important use of PWD readings is to calibrate the flow simulators. This way the
estimation of the friction pressure along the hole can be more accurate.
Also, the possibility of having the bottom hole pressure readings available in the memory
of the downhole tool will help to calibrate a posteriori the simulation model or the
procedures, for cases when the signal is interrupted or has poor quality. This happens
normally when using mud pulse tools and circulation is suspended for connections and
other operations.

Chapter 6
6.3.3.1 Data Transmission

As commonly known, most MWD/LWD tools transmit their data by means of pressure
pulses, either negative or positive, through the mud system inside the drill pipe. A
pressure sensor is attached to the standpipe manifold, allowing the surface unit to decode
the data and provide the required information. This is a proven and widely used
technology, provided by many vendors worldwide.
Operations requiring gas injection through the drill pipe will face the problem of
dampening the pulses emitted by the MWD tool, due to the gas compressibility. There is
a maximum gas fraction that will allow pulse data transmission to surface, and is
commonly set as 12%; however some experiences set this limit as high as 26%. The use
of flow restrictors or signal boosters has proven effective to this purpose.
When gas volume flow through the drill string prevents pulse data transmission, some
other alternatives are available, although they have significant limitations.
Electromagnetic MWD tool sends radio signals through the earth, instead of
sending pulse signals. The radio signals are received on surface by a
properly placed antenna. When applicable, this technique has proven a
great help for data transmission.
Unfortunately, the EM signal has a limited reach through formations,

limited to approximately 7000-8000 ft (approx 2100-2400 m) If there is
any resistive formation, the reach is even shorter.
In this case, one of the options is to run an

extended range antenna, which connects a wireline
to a higher point in the well, where the EM
transmitter is placed. Then, radio signal should be able
to reach surface, and receive the real time data. Figure
20 illustrates the basic EM MWD setup and the use of
the Extended Reach Antenna.
Another option is to use wireline with wet connectors.

However, this option has significant practical limitations,
including the time to retrieve the wireline and running
it back for every connection, not being uncommon to
have problems to hook the connectors back together.
The extended connection time is negative for
Figure 20: Electromagnetic MWD gasified systems, as fluids in the annulus segregate
systems transmit data wirelessly back to further, aggravating slugging and pressure transient
surface. problems.
Wired drillpipe is a newer solution to the data transmission problem. It has inductive and
receiving coils on each tool joint, wired together via a conductor built in the pipe body.
Each joint receives the electrical signal from the joint below, being able to conduct the

Chapter 6
electrical signal to the receiving

sensor at surface. This technology
opens a whole new possibility of
bidirectional high speed data transfer
between the BHA and surface. Signal
attenuation has been addressed by
adding battery powered signal
boosters when needed (typically
every 15 stands). Interesting
possibility about these signal
boosters is that they can include
temperature and pressure sensors,
thus providing an entire distribution
Figure 21: A diagram of hard wired drillpipe. The data profile along the hole. The main
cable is integral to the pipe itself, with contact coils in the downside of this technology at the
tooljoint to facilitate signal transmission from joint to joint. moment is the renting and
Courtesy of Intelliserv.
mobilization cost.
6.3.3.2 Downhole Tool Limitations

Downhole tools used for MPD operations generally have the same pressure and
temperature ratings than the conventional tools. However, MPD operations often involve
critical conditions, such as HPHT or high H2S contents. Also, gasified fluid systems have
less thermal capacity than pure liquid systems, which means that the tools in gasified
systems will reach circulating temperatures that are closer to formation temperature,
compared as to conventional environments. There is limitation on quantities of specific
tools (such as PWD) being built for different sizes.
Because of these reasons, significant limitations could exist for the use of downhole
tools; hence this often requires more extensive planning and research than usual.
6.4 Operations & Well Control

6.4.1 Impact of Drill String Connections
In drilling with jointed pipe, it is important to note that drill string connections interrupt
the steady state of a circulating system, and as such, must be considered when designing
a circulating system for an underbalanced well. Figure 22 illustrates actual annular
bottom hole pressures recorded on an oil well drilled underbalanced. The period
illustrated includes the making of a drill string connection.

Chapter 6
3,500
Annular Bottomhole
Pressure (kPa) 3,300 ANNULUS SHUT-IN
3,100
2,900
DRILL STRING
2,700 DRILLING CONNECTION
2,500
0 5 10 15 20 25 30 35 40 45 50 55
Time (min)
Figure 22 - Annular Bottom Hole Presure During Drilling of an Oil Well
It can be seen that during a drill string connection, the pressure initially drops due to loss
of the friction gradient when circulation is stopped. As fluids separate, however, both in
the annulus and in the drill string, liquid slugs form. Upon restarting circulation, the
annular bottom hole pressure increases due to fluid acceleration overcoming inertia and
the subsequent lifting of the liquid slugs. This pressure cycling is repeated during every
drill string connection or during periods of no circulation.
6.4.2 Well Control in Low Head MPD

Typically, wells drilled using Low Head MPD are classified as level 0, 1 (and less
likely 2) from a risk perspective in the IADC UBD/MPD Well Classification System.
This means that they are low risk operations, with minimum or no chance of hydrocarbon
flow to surface.
In the event of a kick or inflow during a Low Head MPD operation, the fastest and most
effective response will be an increase on the backpressure, followed by a reduction in the
gas injection.
The logics of the well control process are the same as discussed in the Applied Back
Pressure section in this Manual, in Chapter 4. Please refer to that chapter for additional
details.

Chapter 7 – Mud Cap Drilling
1 Rev. 4.0
Table of Contents
7.1 Introduction ...................................................................................................... 3
7.2 Engineering and Well Design .......................................................................... 4
7.2.1 Evaluating Losses ........................................................................................ 4
7.2.2 Wellbore Flipping ....................................................................................... 5
7.2.3 Downhole design ......................................................................................... 6
7.2.4 Drill string Pump Rate and Pressure Requirements .................................... 6
7.2.5 Light Annular Mud Selection ...................................................................... 6
7.2.6 Wellbore Pressure Evaluation ..................................................................... 7
7.2.7 Gas Influx into Wellbore ............................................................................. 7
7.2.8 Controlling Gas Migration .......................................................................... 9
7.2.9 Estimated Mud Usage ................................................................................. 9
7.2.10 Affect of Annular Pressure on Drill string Weight ........................................ 10
7.3 Gas Migration Discussion .............................................................................. 10
7.3.1 Gas Migration Studies in the Petroleum Industry ..................................... 10
7.3.2 Study of Counter-Current Gas-Liquid Flow During Bullheading
Operations ................................................................................................................. 11
7.3.3 Rise Velocity Studies for Taylor Bubbles in Fluid Mechanics ................. 14
7.3.4 Results of the Gas Migration Rates for an Example Well ........................ 18
7.3.5 Minimum Bullheading Rate to Arrest Gas Migration ............................... 19
7.4 Equipment ...................................................................................................... 20
7.5 Operations & Well Control ............................................................................ 20
7.5.1 PMCD........................................................................................................ 20
7.5.2 Floating Mud Cap ...................................................................................... 21
7.5.3 Trips .......................................................................................................... 21
7.5.3.1 Gunk Plugs ........................................................................................... 21
7.5.3.2 Disappearing Plugs ............................................................................... 21
7.5.3.3 Down Hole Isolation Valve .................................................................. 22
7.5.3.4 Snubbing out of the hole ....................................................................... 22
7.5.3.5 Floating Mud Cap ................................................................................. 22
7.6 References ...................................................................................................... 23
2 Rev. 4.0
Chapter 7
MPD Pressure
Control Options

Friction Pressure
Mechanical lift
device
Circulating Non
SMD (Mud Lift)
Semi-automatic
CCS (CBHP)
Concentric Mud Cap
injection
Floating
multi-phase)
CAPM
gradient)
Figure 1
7.1 Introduction
annulus at a rate sufficient to keep produced fluids from migrating to surface. This
annular fluid is referred to as the “mud cap” since it prevents migration of the drilling and
inflow fluids up the annulus to the surface. The cuttings, pumped fluids and any produced
fluids are pumped back into the formation at the loss zone. It is important to note that
this system does not try to prevent or cure loss of circulation.
Mud Cap Drilling is used in areas where severe lost circulation problems and kicks are
prevalent and it is unsafe to allow hydrocarbons to flow to surface i.e. in the event of the
presence of sour gas. The system ensures the produced hydrocarbons, injected drilling
fluid, cuttings and mud cap mud all enter the loss zone.
Mud Cap Drilling can be classified under two types listed below:
Floating (Dynamic) Mud Cap Drilling

Pressurized Mud Cap Drilling
Mud Cap Drilling 3 Rev. 1.0

Chapter 7

The design of a Floating Mud Cap (FMCD) or Pressurized Mud Cap (PMCD) well
centers on fluid design, fluid storage, casing strings and getting out of the well. Unlike
other forms of MPD, it does not look to manage a series of formation problems. In many
cases, both forms of mud cap drilling can be designed for but never used on the well. The
formation (normally a Limestone Karst) needs to be such that total losses are taken. In
many cases the losses are not severe enough to switch to a mud cap regime.
As a rule mud cap operations should only be used if open fractures with a diameter of at
least 3 times cutting diameters are present. Attempting mud cap operation with
insufficient fractures will lead to packing off and stuck pipe.
7.2.1 Evaluating Losses

Once lost returns are encountered, the severity of the losses and the cause of the losses
will determine the appropriate steps to be taken and the eventual plugging procedure.
This section provides several decision points and the recommended course of action for
each decision point.
While drilling ahead in the conventional mode the primary decision points when lost
returns are encountered will be;
1. Are losses due to a large Karst structure? Is this a highly permeable layer or is
this due to hydrostatic/circulating pressures alone? In thick gas laden carbonate
structures friction pressure for horizontal wellbores and pure hydrostatic plus
friction pressure for vertical wellbores, create an increasing overpressure with
depth and an increase in the potential for losses. Reducing the ECD of the drilling
fluid will typically minimize these losses. Losses due to large Karst features will
have less of dependency on fluid dynamics and will generally have static losses
only slightly less than dynamic losses. If static losses are not present or
significantly less than dynamic losses then the losses are not due to Karst
Features. PMCD is not applicable in situations where the formation does not have
the ability to receive continuous injection of drill cuttings and fluids.
2. If the losses are from a Karst feature, then will the suspension of drilling and the
pumping of a Light Annular Mud (LAM) into the structure result in an acceptable
loss rate of annular fluid once PMCD operations begin? The term loss rate in this
case is directed more towards the volume of fluid required to prevent gas
migration. Safety, economics and logistics will be the key decision making
criteria for this decision. Injectivity rates and pressures need to be investigated
prior to initiating PMCD operations.
3. Can the losses be repaired with standard LCM treatments or other techniques and
the well returned to conventional operations? This will probably not be the
course of action since once losses are experienced in the carbonates the use of
LCM treatments has minimal effect.

Chapter 7
4. What fluid density can be used to safely drill ahead in the PMCD mode? This
will be the critical decision for the onsite staff. This decision will impact both
economics and logistics and safety.
Determining the probable cause of the losses is key to understanding the steps forward. It
is however, difficult to determine in many of these wells since identification of a Karst
feature can be masked. Parameters such as ROP vs. Weight On Bit or String Weight
should be evaluated for Karst identification. A rapid rise in Total String Weight
accompanied with an abrupt decrease in Weight on Bit, plus a gain in ROP could signal
entry into a Karst feature.
The most ambiguous decision will be to discontinue drilling while seawater is injected
into the annulus in an attempt to fill the formation and reduce the loss rate of the more
expensive annular cap fluid. This decision could also include the stripping of the bit back
to the shoe since surface pressure required to inject seawater may be above the RCD
stripping pressure and the annular may be closed as a safety measure. It is estimated that
no less than 8 hours and up to 24 hours will be required to determine if this activity will
aid in the reduction of the losses. Measurements from the annular pressure sub will be
critical in determining the long term effect of the injection.
7.2.2 Wellbore Flipping

A major concern when drilling highly permeable carbonate reservoirs is the possibility of
bore hole fluid exchange or the rig description of “flipping”. Flipping, occurs when
losses to a Karst reduce the hydrostatic pressure, allowing inflow of reservoir fluid.
In reservoirs where Karsts exist the flow potential of the reservoir is extremely high and
compounds the problem. Karst features accept fluid at a very high rate from the wellbore
and in deviated wells it is normally thought that this occurs on the low side of the
wellbore due to gravity. At the same time that fluid is entering the formation from the
well, gas feeds into the wellbore, in an attempt to equalize the pressure. Since the
reduction in pressure due to friction is significant for a long section of wellbore the
higher pressured gas is able to displace the wellbore fluid within the confines of the
wellbore and migrate up the wellbore at a velocity equal to the velocity of the falling
fluid. In some cases this inversion has occurred so rapidly that surface detection was
almost impossible, even with the latest in kick detection equipment.
Once the large gas bubble has entered the wellbore and the fluid velocity slows due to the
pressure equalization, the energized gas bubble continues a rapid rise to the surface which
normally results in a severe fluid and gas flow at the surface. When using a surface stack
this poses a serious problem but the event is even more serious when working with a sub
sea stack since the gas is normally above the stack by the time the event is recognized. It
is for this reason that these procedures are based on a conservative approach. When
indications of massive fluid losses are indicated the first approach will be to maintain
hole fill at a predetermined rate and be prepared to secure the well using the annular BOP
or rams to reduce the surface impact in the event of gas migration.

Chapter 7
7.2.3 Downhole design

As will be discussed in some of the following sections, the objective of MCD is to not
cure the losses but more to manage the losses so as to prevent gas from coming to
surface. Gas entering the well bore will eventually need to driven back into the formation.
This is achieved by pumping a fluid down the annulus. The gas is driven into the
formation based on velocity
Annular pump rate requirements are dictated by the velocity of the gas migration and the
annular fluid velocity required to overcome this migration. Gas migration is a function of
densities, gravity, pressure, viscosity, surface tension and wellbore geometry. Gas
migration is discussed in detail in Section 7.3.
7.2.4 Drill string Pump Rate and Pressure Requirements

The sacrificial fluid (SAC) which will be injected down the drill string will be required to
lift the cuttings from around the BHA back to the point of losses. Since the SAC fluid
will be low rheology seawater, velocity will be the primary hole cleaning component. As
in UBD operations when using clear fluids, it is possible to provide very good hole
cleaning if the fluid velocity is maintained above 250 fpm in horizontal wells and 180
fpm in vertical wells. The second criterion for drill string pump rates is the BHA
requirements. Ensure sufficient pump rate for downhole motors.
Standpipe pressure will respond in the PMCD role the same as it does in the conventional
operations. Standpipe pressure is the sum of all frictional pressure losses in from the top
drive through the sting motor etc and then up through the annulus to the surface. During
PMCD operations the fluid path will only be part way up the annulus, before it enters the
formation. Although the frictional annular losses above the loss losspoint will be
removed, they will be replaced by the annular downhole injection pressure observed at
the rock face. Since the bottom hole injection pressure should be at formation pressure in
this application, the standpipe pressure will only be affected by a slight amount from
conventional drilling operations. The major factor will be the hydrostatic pressure inside
the drill string will be less than formation pressure so there will normally be trapped
pressure when bleeding off for a connection.
7.2.5 Light Annular Mud Selection

Light Annular Mud (LAM) selection is primarily a cost and logistics based decision.
Although fluid invasion and reduction of formation damage are a part of the selection
process, cost and ability to mix at sufficient rates become the major selection criteria. The
density of the LAM is selected to achieve a surface pressure that can be maintained
below the working pressure of the RCD. The LAM must be designed to minimize casing
wear during the drilling operations.
Saltwater is generally an acceptable base fluid for a LAM (if offshore) if the expected
pore pressures are encountered. If the injection pressure looks like it will exceed the
working pressure of the RCD, the LAM density should be increased.
Consideration should also be given to adding a bentonite gel so as to provide viscosity to

the base saltwater. This will aid in displacing the gas from the wellbore and more

Chapter 7
importantly assist in minimizing casing wear. Other products are available to add
lubricity to further minimize casing wear. The use of oil based mud for the LAM has
proven beneficial in retarding the migration of gas, as the gas is miscible in the fluid
system, thereby reducing the number of pump/flush cycles.
If a weighted LAM is required due to higher pore pressures, a barite may need to be
added to the fluid to achieve the required density.
7.2.6 Wellbore Pressure Evaluation

There is value in being able to understand the gas migration rates vs. annular fluid
injection rates in the well bore. To do this, inclusion of an annular pressure sub should be
included in the design. Error! Reference source not found. describes the indicators and
probable causes when using an annular pressure sub.
Table 1- Annular pressure sub indicators for PMCD
PWD Annular Drill Probable Cause Action
Reading Surface pipe
Pressure Surface
Pressure
Increases Increases Increases
Formation is Plugging – Reduce annular
Could be observed when a Injection and
higher formation pressure is Evaluate
encountered and formation
plugging is suspected in the
previous loss zone
Stays the Increases Stays the Gas Bubble migrating in Increase Annular
Same Same Annulus Injection Rate
Decreases Decreases Decreases High loss rate into formation Continue drilling
or lower bottom hole ahead if other
pressure formation associated factors are not a
with losses to other formation problem.
– Impossible to determine if
cross flow is occurring.
7.2.7 Gas Influx into Wellbore

Determining the Volume of Gas Influx and Target Surface Pressure
Displacing reservoir fluids back into the reservoir is the key operation in PMCD
operations. One must first determine what is the allowable volume of gas and the
corresponding pressure change when that volume enters the well? Since it is difficult to
determine the volumes in the closed system, we will use the indicated pressures to
determine when high rate annular injection is required. These pressures are based on
basic well control fundamentals of gas migration and hydrostatic pressure changes due to
influx.
For this operation several pressure limitations are given for the equipment that will be in
operation. The most critical of these components is the RCD which, for example, has a
dynamic pressure rating of 1000 psi. This will be the maximum allowable annular

Chapter 7
pressure when the pipe is rotating or moving up and down. For safety, this pressure rating
will be reduced by 100 psi. With a maximum dynamic pressure of 900 psi set as the
annular pressure limit, the amount of gas influx for various LAM weights can be
calculated.
Example –. A rig drilling in a limestone formation known to contain karst like

features is outfitted with equipment for mud cap drilling. The rig is drilling 8 ½”
hole with 5 7/8” drillpipe.
Total losses have been reported at a TVD of 4,245 feet. The observed bottom
hole pressure when the losses started was 2,485 psi. A Light Annular Mud
(LAM) has been chosen with a gradient of 0.46 psi/ft for PMCD operations. The
density of the influx gas is assumed to be 0.065 psi/ft. It has been determined that
the maximum Shut In Casing Pressure (SICP) while drilling is 850 psi.
Determine the volume of LAM required to flush the influx back into formation.
The LAM volume can be derived from the size of the influx using the following
formula:
2 2
SICPincr IDhole − OD pipe
InfluxVolume(bbls ) = *
LAM density − Influxdesnity 1029.4
Where:
SICPincr = Amount of annular surface pressure increase observed
LAMdensity = Light Annular Mud density (psi/ft)
Influxdensity = Estimated gradient of gas influx (psi/ft)
ID hole = Internal Diameter for section of hole containing the influx
OD pipe = Outside Diameter for drill string
Step 1 - Calculate the anticipated surface pressure (or the shut in casing
pressure)for the chosen LAM.
 P  
SICP =  res  − LAM  * TVDres
 TVDres  
 2,485 psi  
  − 0.46 psi / ft  * 4,245 ft = 532 psi

 4,245 ft  
The calculated or target annular surface pressure should be 532 psi with a full
column of 0.46 psi/ft LAM in the annulus.
Step 2- Determine the increase in SICP.

Chapter 7
It has been noted that the maximum allowable surface pressure will be 850 psi.
The increase in SICP is therefore:
SICPincr = SICPcurrent-SICPtarget
SICPincr = 850 psi – 532 psi = 318 psi
Returning to the first equation;

2 2
SICPincr IDhole − OD pipe
LAM density − Influxdesnity 1029.4
318 psi 8.5 2 − 5.875 2

0.46 psi / ft − 0.065 psi / ft 1029.4
= 29.5 bbls
Since barrels of influx is an indication of the minimum volume of mud required to

displace the gas back to the formation, this is the minimum volume that will be pumped
at a high rate during the displacement operation.
7.2.8 Controlling Gas Migration

The primary tool to control gas migration in the wellbore is liquid velocity; and pump
speed. Obviously, the best technique to eliminate the potential for gas migration would
be the continuous injection of Light Annular Mud, but this is seldom a feasible solution.
Instead, when well conditions allow, the gas is displaced from the annulus with a high
pump rate and then injection is reduced or stopped and the annulus pressure is observed.
An increase in annular pressure will be observed over time, this is due to gas migration in
the annulus. When the annular pressure rises to a predetermined value, fluid is injected
into the annulus, at the prescribed rate, and the gas is forced back into the reservoir.
Pumping is then shut down and the annulus pressure is allowed to rise again. This process
is repeated continuously until the drilling depth is achieved.
It is impossible to determine the amount of time this cycle will take in the planning stage,
but will be easily observed during the well operations. The important point is to look for
this cycle, and optimize it, to reduce the overall cost of material required to perform the
operation. After continued injection of seawater and LAM the high perm features may
not react as fast as they did when first drilled. Again, optimizing this time period, and the
fluid used to maintain the well, should be critical.
7.2.9 Estimated Mud Usage

The volume of mud required during an operation is unknown. It is only possible to
understand the rigs ability to mix and provide a continuous supply and to make decisions
with regards acceptable mud usage. This time can be extended with continuous mixing
on the rig. If mud usage is too high to allow continued injection, it may be necessary to

Chapter 7
pull the bit to the shoe and inject seawater in an attempt to fill the pore space and reduce
fluid losses. If seawater injection is not effective, a second line of pills, LCM or fiber
cement should be considered as an option to reduce fluid losses.
7.2.10 Affect of Annular Pressure on Drill string Weight

Based on the estimated bottom hole pressure and the use of LAM, annular surface
pressure will be positive for most of the operation. The presence of annular surface
pressure will create a piston force across the end area of the drill string and will result in
an additional upward force.
Changes in annular pressure will have a direct effect on the actual weight on bit and this
effect should be considered when LAM densities and surface pressure are altered. This
change should be noted on the weight indicator and must be allowed for, while drilling,
to maintain desired weight on bit.
7.3 Gas Migration Discussion

Gas migration poses a serious well control issue, as well as an operational issue for
PMCD operations. An understanding of gas bubble migration and the optimal methods
to arrest that migration is essential in MPD operations.
The “Taylor” bubble is defined as a large bubble i.e. for which λ > 0.6 and its terminal
velocity is not a function of λ but a function of the pipe diameter D. λ is the ratio of the
sphere-volume equivalent bubble diameter d to the pipe diameter D [1].In a two-phase
flow pattern called the ‘slug-flow’, gas slugs or Taylor bubbles occupy almost the entire
pipe cross section as the gas flow rate is fairly high [2]. The Taylor bubbles are axially
separated by a liquid slug in which small bubbles are dispersed. The liquid confined by
the Taylor bubble and the tube wall flows around the bubble as a falling film [2]. The gas
migration rate through mud/water in the event of a gas-kick can be identified as the
bubble rise velocity of the Taylor bubble.
7.3.1 Gas Migration Studies in the Petroleum Industry

Various authors [3, 4, 5, 6 and 7] have published literature on gas-migration rates during
gas kicks. Johnson and White [3] reported a surprising finding based on their experiments
that bubbles rise faster in drilling mud than in water despite the increased viscosity.
Johnson and Cooper [4] discuss the effect of geometry and deviation on gas-migration
rates. Tarvin et al. [5] and Johnson et al. [6] measured migration rates in laboratory large-
scale tests and found them to be much greater than commonly reported in the industry.
This was accompanied by simulation of field kick data. Their results showed that gas
migrates as fast as 6000 ft/hr through drilling mud and not 1000 ft/hr as earlier believed
in the industry, for gas void fraction of more than 10 %. Grace et al. [7] have presented
field examples of gas migration where surface pressure is used to predict influx behavior
and gas migration rates.
The fundamental theory and equations these studies have incorporated include the work
of Davies and Taylor [3] and Zuber and Findlay [3, 4, and 8]. Davies and Taylor derived
the equation for “Taylor” bubble velocity for large bubbles that fill the pipe by

Chapter 7
considering inviscid flow around the bubble nose. This velocity uses the pipe diameter as
the scaling parameter. The equation is,
ν t = 0.35 ( g ( ρ l − ρ G ) D / ρ l …….. (1)
where νt= “Taylor” bubble velocity, ρl =liquid density and ρG = gas density
Zuber and Findlay derived a model that stated,
vG = C 0 v H + v s ……… (2)
where vG = mean gas velocity and ν s =gas-bubble slip velocity relative to a stationary
fluid. The homogeneous velocity, νH is defined as
v H = (qG + ql ) / A …… (3)
where qG and ql are volumetric flow rates of gas and liquid phases respectively. They
showed that the distribution parameter Co would range from 1.0 to 1.5 and this was later
verified by experiments.
7.3.2 Study of Counter-Current Gas-Liquid Flow During Bullheading

Operations
Louisiana State University (LSU) conducted an experimental study to investigate the gas
removal efficiency of the bullhead method for well-control based on the work by Zuber
and Findlay [8]. The primary focus of the investigation was the downward displacement
of gas by liquid, which requires consideration of counter-current flow behavior.
Experiments were conducted in a full-scale well, using a computer controlled down-hole
fracture simulation system; water and low-viscosity drilling mud were used for bullhead
fluids. Experiments were also conducted in an inclined flow loop so that the effect of
vertical deviation angle could be investigated. Figure 2 illustrates the results for the
removal efficiency of gas as a function of the average annular velocity pumped. A
removal efficiency of 100 % implies that the downward annular velocity was greater than
the upward slip velocity of the largest bubbles formed. At a given injection rate, the
removal efficiency of mud is much higher than that of water.

Chapter 7
Figure 2- Removal Efficiencies for LSU Experiments [9]

Co-current gas-liquid flow in long vertical or inclined annuli can be represented on a plot
of the average gas velocity as a function of the average mixture velocity (Zuber-Findlay
plot). The work considers that for a gas kick in an annulus of constant cross sectional
area, the average mixture velocity in the gas contaminated kick region can be taken as
equal to the velocity of the uncontaminated liquid above the gas kick region. The use of
Zuber-Findlay plots was extended to counter-current gas migration by considering a
downward pumping velocity to be negative and a conventional upward circulation
velocity to be positive. Similarly, an upward gas velocity is considered positive while a
downward gas velocity is considered to be negative.
Figure 3 is a Zuber-Findlay plot that combines data taken in the LSU inclined flow model
(6-in ID) and data taken in the research well (6-in. ID). In addition, published data
(Johnson and White [1991]) taken at the Schlumberger Cambridge Research facility in a
larger (8-in. ID) model of a concentric annulus is also plotted. The lower data point
shown (solid square) corresponds to the minimum downward pumping velocities for
100% gas removal efficiencies in the research well experiments. The solid line represents
the results that would be predicted by the computer model developed in their work.

Chapter 7
Figure 3 - Gas Velocity for Water in a Vertical Annulus [9]
Figure 4 is the Zuber-Findlay plot obtained for a mud in a vertical annulus. The
counter-current gas slip velocity in a mud during bullheading operations was found to be
much less than for water.
Figure 4 - Gas Velocity for Mud in a Vertical Annulus [9]

Chapter 7
The work conducted at LSU will help to evaluate when bullheading kill operation will
have a high chance of success for controlling underground blow-outs as one can now
predict the countercurrent gas slip velocities through both Newtonian and non-Newtonian
fluids.
7.3.3 Rise Velocity Studies for Taylor Bubbles in Fluid Mechanics

Pioneering work has been done in the field of fluid mechanics to study the rise velocity of
Taylor bubbles since the original work by Davies and Taylor (1950). Davies and Taylor
conducted experiments illustrating that small bubbles and drops are nearly spherical, but
large ones are nearly spherical caps with rather spherical upper surfaces and concave
lower surfaces [10]. When a bubble rises through a fluid, it is deformed from the
spherical shape by the flow of the fluid around it and by the hydrostatic pressure due to
gravity [10]. Figure 5 contains photographs of Taylor bubbles rising through different
viscosity fluids [11]. As seen in the figure, the Taylor bubbles exhibit a rounded nose and
in water give rise to an unsteady and irregular tail.
Figure 5 - Photographs of Taylor bubbles rising through 76.2 mm inside diameter pipe
filled with different viscosity fluids: a) water ;b) Purolub 150 oil ( 480 mPa s);

Chapter 7
Dumitrescu (1943) and Davies and Taylor (1950) ignored frictional and capillary effects
and considered only the kinetic energy of the liquid falling around the bubble [11]. The
equation published by Dumitrescu relates the bubble velocity U through a liquid in a
vertical circular tube to the tube diameter and the acceleration due to gravity by the
following equation:
U = 0.351 gD …….. (4)
This equation generally agreed with the equation later presented by Davies and Taylor.
The equation developed by Davies and Taylor contained the constant 0.328 in place of
the constant 0.351 in equation 4. Davies and Taylor (1950) analyzed the related problem
of the rise of a spherical cap bubble in an unbounded liquid without the restraining effect
of the tube walls [11]. They found the value 0.47 instead of 0.328 for long gas bubbles
associated with the effects of liquid drainage on the tube walls [11]. Griffith and Wallis
(1961) investigated experimentally the two-phase flow slug flow through various sizes of
round pipes and found that their results in agreement with that of Dumitrescu [11].
Nicklin, Wilkes and Davidson (1962) conducted experiments to observe the effect of
bubble length on rise velocity and concluded that the rise velocity of the Taylor bubble is
independent of the slug length and is modified by a net flow of liquid across a section
above the slug [11]. Brown (1965) made an experimental and theoretical study of the
effect of liquid viscosity on the terminal rise velocity of Taylor bubbles [11]. The original
form of Brown’s equation is
 1 + ND − 1 
U = 0.35 gD 1 − 2 
 (5)
 ND 
where
 ρ2g 
N = 3 14.5 l 2  ………(6)
 µl 
N is dimensionless and ND is dimensionless. U= “Taylor” bubble rise velocity, ρl =liquid
density, µl =liquid viscosity, D = pipe diameter
The applicability limits of equation 5 were established empirically as:

2
ρ l gD 2   
Surface Tension: 1 − 2 1 + ND − 1   > 5.0
4σ   ND 
  
Viscosity: ND > 60
Zukoski (1966) performed an extensive experimental study to observe the effects of

surface tension, viscosity, density, pipe diameter and angle of inclination [11]. Wallis
(1969) proposed a general correlation for Taylor bubble rise velocity in terms of all the
relevant variables:

Chapter 7
0.5
 Dg ( ρ l − ρ g 
U = k  ……….. (7)
 ρl 
0.345  m
(3.37 − Eο )
and k = 0.3451 − e
− 0.01 R
1 − e  ……… (8)
  
where R is the buoyancy Reynolds number:
R=
[D 3
g (ρ l − ρG )ρ l ]
0.5
…………… (9)
µ
m is a function of R and takes on the following values:
R>250 m=10
18<R<250 m=69R-0.35
R> 18 m=25
Joseph D.D. [12] applied the theory of viscous potential flow to the problem of finding
the rise velocity of U of a spherical cap bubble. He proposed the following equation for
rise velocity,
0.5
U 8 ν (1 + 8s ) 2 16 sσ 32ν 2 
=− + 1 − 2 s − 2
+ 3
(1 + 8s )2 
gD 3 gD 3 3  ρgD gD 
……………… (10)
where R= D/2 is the radius of the cap, ρ and υ are the density and kinematic viscosity of
the liquid, σ is the surface tension and s is the deviation of the free surface from perfect
sphericity. Funada T. et al. [13] modeled the rise velocity of long gas bubbles in round
tubes by an ovary ellipsoidal cap bubble rising in an irrotational flow of a viscous liquid.
Viana et al.(2003) [11] collected data from 255 experiments available in the published
literature and 7 new experiments based on rise velocity of long gas bubbles in round
pipes and processed them in log-log plots of normalized rise velocity called the Froude
number, Fr versus buoyancy Reynold’s number R (eqn.9) for fixed ranges of the Eotvos
number, Eo.
Froude number is proportional to the ratio of inertial forces to the gravitational forces and
is given as,
Fr = U /(gD ) 0.5 ……….. (11)
Eotvos number is propotional to the ratio of the gravitational forces to the surface tension
forces and is given as

Chapter 7
Eo = gρ l D 2 / σ ………… (12)
where D is the pipe diameter, σ is the interfacial surface tension.
The log-log plot of Fr versus R is illustrated in Figure 6 [11]. The data sorts into a
flat region for large R (>200), a slope region for small R (<10) and a transition region in
between.
Figure 6 - Log Fr versus Log R quoted from Viena et al. [11] for Eo >6 [11]
This plot was used to construct power laws of the type Fr = α ( Eo) R β ( Eo ) to collapse the
data on the rise velocity of long bubbles in round tubes. The composition of these
separate power laws emerge as bi-power laws for two separate flow regions for large and
small buoyancy Reynold’s number. For large R (>200),
0.58
Fr = 0.34
(1 + 3805 / Eo )
3.06
………………. (13)

Chapter 7
For small R (<10),
9.494 × 10 −3
Fr = R 1.026
(1 + 6197 Eo )
2.561
0.5793
………………. (14)
The composite correlations for large and small R were joined in an overall universal
correlation which is given as,
……………….. (15)
The performance of the universal correlation was evaluated by comparing the values
predicted by it to experiments. Almost all the values fall within the 20 % error line and
most of the data is within 10% of predicted values. Funada T. et al. [13] regarded that the
universal correlation accurately predicts the rise velocity and further improvement cannot
be expected from modeling. Therefore the universal correlation proposed by Viana et al.
[11] was developed as a spreadsheet application for the calculation of the gas migration
rates for the gas-water/mud system applicable for bullheading in the event of a gas-kick.
7.3.4 Results of the Gas Migration Rates for an Example Well

The universal correlation based on the work by Viana et al. (2003) forms the basis of the
gas migration rate calculations. The scenario being modeled is the gas migrating upwards
in the annulus after a kick has occurred. An accurate prediction is extremely important as
the rates will determine how fast the gas reaches the surface which needs to be arrested
by bullheading water through the annulus. The liquid properties are that of water. The
hydraulic diameter (hole diameter/ ID of casing-OD of drillpipe) is used for gas
migration calculations in the annular geometry. The gas migration rates for the geometry
of the sample well are as follows:

Chapter 7
Table 2 - Gas Migration Results for the Sample Well

Geometry Gas migration rate
(ft/hr)
1. 14.75” hole 7700
2. 14.75” hole x 5.875” 5973

drillpipe
3. 12.615” ID casing 7121
4. 12.615” ID casing x 5205

5.875”drillpipe
5. 12.25” hole 7018
6. 12.25” hole x 5062

5.875” drillpipe
7. 8.681” ID casing 5907
8. 8.681” ID casing x 3359

5.875” drillpipe
9. 8.5” hole 5846
10. 8.5” hole x 3248

5.875” drillpipe
The results indicate that gas migration rate varies significantly with the pipe diameter.
Gas moves faster when closer to the surface, much faster than 6000 ft/hr which is the rate
reported in the literature [5 and 6]. If the requirement is that of no-gas on the surface,
then one needs to determine the minimum bullheading rate that will stop the gas
migration.
7.3.5 Minimum Bullheading Rate to Arrest Gas Migration

Watson et al. [14] proposed an equation for the minimum kill rate for bullheading so
that the velocity of the pumped fluids exceeds any upward gas migration rate. The
equation is,
(q kr )min = U . A cs ……….(16)
where U= the gas migration velocity through the largest flow conduit of cross-sectional
area Acs.

Chapter 7
As an example, if the kick is taken in the 12.5” hole, then the largest flow conduit would
be the inner diameter of the 13.375” casing which is 12.615”. From Table 1, the gas
migration velocity is 7121 ft/hr and Acs is 0.87 ft2. Using equation 16,
(q kr )min = 18 bbl/min
= 770 gal/min
This is the minimum bullheading rate to ensure complete gas displacement.
Similarly, if the kick is taken in the 8.5” hole, then the largest flow conduit would be the
inner diameter of the 9.625” casing which is 8.681”. From Table 1, the gas migration
velocity is 5907 ft/hr and Acs is 0.41 ft2. Using equation 16,
(q kr )min = 7.2 bbl/min
= 302 gal/min
Any kill rate higher than 302 gal/min will be a good selection to stop gas migration.
7.4 Equipment
The primary equipment used for PMCD operations is the Rotating Control Device, which
has been discussed extensively in Chapter 4.
7.5 Operations & Well Control

7.5.1 PMCD
PMCD uses a hydrostatically underbalanced fluid and surface pressure to maintain BHP
at the fractured zone. As all returns are pumped back into the formation, kicks are not
normally an issue, unless the ability to pump down the annulus is lost. From a design
standpoint, a kick can also be described as requiring higher annular pump pressure to
inject the reservoir fluids back into the formation. This can become a major issue if the
equipment (annular pump and rotating control head) does not have sufficient pressure
rating for the higher pressure.
If the reservoir pressure is too high, or more correctly the fluid injection pressure at the
fracture is too high (should be basically the same in a fractured formation), the annular
fluid density must be increased to bring the surface casing pressure down. The new
required fluid density can be calculated by:
( Pann − PTraget )
MWnew = MWold + (17)
0.052 * D fracture
Where:
MWnew = Required mud weight to achieve target pressure (ppg)
MWold = Original fluid density (ppg)
Pann = Annular Pressure with original mud weight (psi)
Ptarget = Target Annular Pressure with New Mud Weight (psi)
Dfracture = Depth of fracture that annulus fluid is pumped into (ft)

Chapter 7
The more common problem in PMCD operations occur when attempts to bridge the
annulus to allow a trip fail. It is not uncommon to take several attempts to create a bridge
to allow a trip.
7.5.2 Floating Mud Cap

A kick in a floating mud cap operation occurs when the fluid used to create the cap is not
sufficient to hold back the rise of gas in the annulus. Although this can occur because a
zone of higher pressure is intersected, it is more common that the rate of annular pumping
is not sufficient and the gas reaches the bubble point, unloading the cap or the fractures
have become plugged.
Common well control technique is to close in the well on the BOP and bull head the kick
back into the formation. The mud weight can then be readjusted to create a safer cap.
7.5.3 Trips
Like many non-standard operations, drilling is not the biggest design hurdle, tripping is.
Particular attention must be made how to make bit trips and how to run completions in
the hole. This is of particular concern as many operators and rig contractors have policies
requirement a stable fluid column and no surface pressure on trips. Several options have
been used in the past, with advantages and disadvantages for each.
7.5.3.1 Gunk Plugs

One of the first applications of PMCD occurred on the Mobil NSO platform, offshore
Indonesiai. This same procedures is being run successfully on many PMCD projects to
day. Mobil utilized GUNK (Diesel Oil Bentonite) plugs set just inside the last casing
shoe. The GUNK plugs effectively isolated the open hole, allowing the wellbore above
the plug to be circulated clean and the string tripped out of the hole with no pressure.
When going back in the hole the string is run back in the hole to just above the plug. The
RCD element is put back in place and the GUNK plug is washed through.
The same procedure can be used to run a slotted liner across the zone (if desired), but a
system must be in place to allow the plug to be washed through. This normally includes
an inner string (to allow circulation out the shoe), a drilling shoe and a method for
rotating the pipe through the plug.
Although this process has been used successfully on many projects, it is not uncommon
to have to set several GUNK plugs before on set successfully. It must be understood that
the plugs are being set in dynamic situations (total losses and gas migration). For many
cases, the first plug slows down the losses, allowing a better chance on subsequent plugs
to get good isolation from the open hole.
7.5.3.2 Disappearing Plugs

There are many “disappearing plugs” that are available on the market. Most of these are
created using cross link polymers. The plugs are set across the last casing shoe, similar to
the GUNK plug. The main difference between the GUNK plugs and Disappearing Plugs
is that the cross link polymers will break over time. This means that if the timing is
correct, there is no reason to have to wash through the plugs. The downside to these

Chapter 7
plugs is getting the timing correct. The risk is that the plug breaks down at the wrong
time.
7.5.3.3 Down Hole Isolation Valve

Down hole isolation valves have been used successfully to isolate the open hole from the
cased hole section.
7.5.3.4 Snubbing out of the hole

The only method that ensures that managed pressure can be maintained throughout the
drilling of the well is stripping out of the hole to the pipe light depth and snubbing out the
remainder of the string. The biggest draw-back of utilizing snubbing for trips is the
requirement for a snubbing unit once the string becomes pipe light.
The use of snubbing is not usually the preferred tripping method for PMCD due to the
wear and number of elements required to strip under high pressure, the additional cost
and complexity of utilizing a rig assist snubbing unit, and since the snubbing force
exceeds the capacity of a push pull machine.
7.5.3.5 Floating Mud Cap

Some operators have decided to strip out under pressure, managing surface pressure by
pumping LAM as required. Once the pipe gets to the pipe light depth, the density of the
LAM is increased to create a floating mud cap. This allows the pipe to be tripped the rest
of the way out of the hole. It should be noted, that this method will not satisfy the
requirement for a stable fluid column and no pressure for trips.

Chapter 7
7.6 References
1. Tomiyama A., Nakahara Y., Adachi Y. & Hosokawa S., “ Shapes and Rising
Velocities of Single Bubbles rising through an Inner Subchannel”, Journal of
Nuclear Science and Technology, Vol. 40, No.3,March 2003
2. Hasan A.R. and Kabir C.S., “A Study of Multiphase Flow Behavior in Vertical
Wells”, SPE 15138, May 1988
3. Johnson A.B. and White D.B., “ Gas Rise Velocities During Kicks”, SPE 20431,
Dec 1991
4. Johnson A.B. and Cooper S., “ Gas Migration Velocities During Gas Kicks in
Deviated Wells”, SPE 26331, Oct 1993
5. Tarvin J.A., Hamilton A.P., Gaynord P.J. and Lindsay G.D., “Gas Rises Rapidly
Through Drilling Mud”, IADC/SPE 27499, Feb 1994
6. Johnson A., Cooper I., Bailey T and McCann D., “Gas Migration : Fast, Slow or
Stopped”, SPE/IADC 29342, March 1995
7. Grace R.D.,Grace, Shurshen J.L, “Field Examples of Gas Migration Rates”,
IADC/SPE 35119, March 1996
8. Zuber N. and Findlay J.A., “Average Volumetric Concentration in Two-Phase
Flow Systems”, “Average Volumetric Concentration in Two-Phase Flow
Systems”, Journal of Heat Transfer, Nov 1965.
9. Bourgoyne T.M., Koederitz W.L. and Bacca H., “An Experimental Study of
Bullheading Operations for Control of Underground Blowouts”, Jan 2001
10. Miksis M.J., Vanden –Broeck J. and Keller J., “Rising Bubbles”, Journal of Fluid
Mechanics, Vol 123, April 1982
11. Viana F., Pardo R., Yanez R., Trallero J. and Joseph D., “Universal Correlation
for the Rise Velocity of Long Gas Bubbles in Round Pipes”, J Fluid Mechanics,
494,2003
12. Joseph D.D., “Rise Velocity of a spherical cap bubble”, J. Fluid. Mech. , vol 488,
Feb 2003
13. Funada T., Joseph D.D.,Maehara T. and Yamashita S., " Ellipsoidal Model of the
rise of a Taylor bubble in a round tube", International Journal of Multiphase Flow,
2005
14. Watson D., Brittenham T. and Moore P.L.," Advanced Well control", SPE
Textbook Series Vol 10,2003
15. Smith K.L., Gault A.D., Witt D.E., Weddle C.E. “Subsea Mud Lift Drilling Joint
Industry Project: Delivering dual gradient drilling technology to Industry, SPE
71357, 2001
16. Torsvoll A., Horsrud P., Reimers N., “Continuous Circulation During Drilling
Utilizing a Drill String Intergrated Valve – A Continuous Circulation Valve”,
IADC/SPE 98947, 2006
17. Meyers G., “Ultra-Deepwater Riserless Mud Circulation with Dual Gradient
Drilling”, Scientific Drilling No 6. Pg 48- 51, July 2008
i
Quitzau, R., Brand, P., Tarr, B., Frink, P, Laeuchtenberg, C., “System for Drilling an Offshore Sour Gas
Reservoir”, SPE/IADC 52808, 1999 SPE/IADC Drilling Conference, 9-11 March 1999, Amsterdam, The
Netherlands.

Chapter 8 – Procedures
MPD Proceudures 8-1 Rev. 2.0

Table of Contents
8 MPD Procedures ...................................................................................................... 8-3
8.1 Procedure Development ................................................................................... 8-3
8.1.1 Planned Procedures .................................................................................. 8-4
Pressure testing and commissioning procedure ................................................... 8-4
Connection procedure .......................................................................................... 8-5
RCD Element Change Out ................................................................................... 8-5
Dynamic Flow Check .......................................................................................... 8-6
Tripping in / POOH ............................................................................................. 8-7
POOH for PMCD................................................................................................. 8-8
Other Operating Procedures ................................................................................. 8-9
8.1.2 Contingency Procedures .......................................................................... 8-9
Catastrophic RCD element failure ....................................................................... 8-9
MPD Choke Plugging ........................................................................................ 8-10
NRV Failure ....................................................................................................... 8-10
Other Contingency Procedures .......................................................................... 8-11
8.2 Procedure Format ........................................................................................... 8-11
8.3 Managing Procedures at the Rig Site ............................................................. 8-13
8.4 The Procedure Lifecycle ................................................................................ 8-14

8 MPD Procedures
This section discusses the approach needing to be taken in creating and executing the
numerous MPD procedures required at the well site. Failure to give due diligence to the
planning and philosophy of MPD procedures will have a detrimental impact once
operations start. Development, management and improvement of procedures should be
considered as a segment of ‘Project Management’ that needs continuous attention. The
procedures will not be developed, printed and then considered as final. They are living
documents that continue to be improved and controlled through the Management of
Change process.
The fundamental purpose of a procedure is the same as a recipe for making Lasagna; to
ensure the end objective is achieved, it is necessary to follow the steps methodically. The
challenge at a well site is to ensure all personnel involved in the activity understand the
procedure, and are aware of their responsibilities. While sounding obvious, conducting a
procedure can easily go wrong resulting in frustration and possible damage / harm to
personnel, equipment or the formation.
This section discusses the approach to developing procedures as part of planning for an
MPD well(s). It discusses the types of procedures needed, how they should be developed
and checked, and issues to consider at the rig site when they are being undertaken.
Sample procedures are presented at the end. These have been simplified with the
specifics removed to promote clarity.
8.1 Procedure Development

Development of procedures occurs during the planning stage of the well / campaign. A
prudent approach is to split procedures into two broad categories; planned and
contingent. The planned procedures are those activities you know will be conducted at
some point in the MPD operation. These would include tripping in hole, POOH, RCD
element change out etc. Contingent procedures, as the name infers, are procedures you
have prepared in the event of something going awry. Such events would include NRV
failure, catastrophic failure of the RCD element etc. It is important to have procedures in
place for such events, because should the event occur, time is generally of the essence,
and an operator’s first reaction may be the incorrect one.
An initial list of procedures can normally be compiled and an outline developed for each
one. After conducting the HazId and HazOP, additional procedures will normally be
recognized as required, and these will be added to the list. The reason for this is that upon
identifying a potential hazard in the process, only two methods generally exist to mitigate
their occurrence; either engineer / design the hazard out, or develop a procedure to ensure
the hazard does not occur.
Prior to developing the procedures, a thorough rig visit should be undertaken. During this
visit, it is necessary to speak with the Tool Pushers, Drillers, and Derrick man etc to
understand how they run their rig. The procedures need to mold in around rig operations
as best as possible. To do this, it is important to understand how the standpipe manifold is

normally aligned, what are the preferred rig pumps to be used, how the Derrick man runs
his tanks and pumps. It is also worthwhile understanding the PVT system and what, if
any, tanks are not monitored etc. The rig visit will allow the process flow diagram to be
created and all interfaces between the rig and MPD identified.
Merging the draft process flow diagram with how the rig personnel run the rig will allow
the person / people developing MPD procedures arrive at a workable set of draft
procedures quicker. Failure to do this invariably results in certain procedures either being
unworkable or requiring rig crew / service personnel to change their standard set up /
method of operation. There will be an element of this, but if it can be limited, then the
simpler the transition into drilling MPD.
8.1.1 Planned Procedures

As stated earlier, planned procedures are those procedures expected to occur at some
point during the well site operations. The following are a sample of the likely MPD
procedures to be used on an ABP MPD project. They are simplified to give an example
of what is required from each procedure.
Pressure testing and commissioning procedure

1. Develop a pressure testing sequence.
2. Line up the mud pumps to flush all MPD related lines to confirm all valves are open
and no obstructions.
3. Walk all the lines with the PFD and ensure the valves are all correctly labeled.
4. Rig in the pressure test pump at the designated tie in point.
5. Line up the valves for the initial body test.
6. Conduct a low and high pressure body test on all the required lines and connections.
Chokes valves etc should have been pressure tested ahead of time.
7. Starting from the furthest point of the pressure test pump, start closing valves and
opening up valves downstream of them. Conduct low and high pressure tests on all
the valves. Make sure valves are being tested from the direction they are expected to
hold pressure from.
8. Continue to work forward following the pressure testing sequence. Ensure leak paths
behind the valves exist.
9. For pressure testing the RCD, install either new element on DP or use a test cap
(dependent of the model of RCD). Need to coordinate this test in with the BOP
testing and will possibly switch from a pressure test pump to a cement unit.
Commissioning
10. Flow through both chokes to ensure they are operating correctly and that the position
indicators are calibrated.
11. Function test any remote starts for pumps.

12. Use each rig pump to flow through the MPD flow meter and into the trip tank. Use
this opportunity to calibrate the pumps and flow meters against the know volume of
the trip tank.
13. If required calibrate the control system against the well bore.
Certain MPD systems work through a PID controller and their system needs to be
calibrated against the well bore and the fluid to be drilled with. If WBM is in the well,
and the intention is to drill MPD with OBM / SBM, then the well will need to be
circulated over ahead of calibrating the control system against the well. The objective is
to allow the controller to better respond the well bore geometry and the dampening effect
of the drilling mud. By calibrating the system, the chokes should not overshoot /
undershoot when trying to trap the required pressure. This is an important step and time
should be taken to get the chokes and system tuned.
Connection procedure
1. Work pipe as required for hole cleaning. Obtain torque and drag values, as well as,
pick up and slack off values.
2. Start to bring the mud pumps down while the MPD choke starts to trap the required
back pressure to compensate for the loss of ECD.
3. Once the pumps are off, set slips and bleed off the standpipe pressure
NOTE: If there is continued flow, the NRVs may have failed. Close the IBOP and
inform the Tour Pusher. Refer to NRV failure procedure.
4. Make the connection.
5. Start rotating at 10-15 rpm and slowly bring mud pumps on.
NOTE: There will be a delay between the mud pumps coming on the stand pipe
increasing as the stand is filled.
6. As ECD is reintroduced into the well, the MPD choke will start to open and allow
back pressure being applied on the well to reduce.
7. Once full drilling rate is reached, confirm MPD system is ready, and resume drilling.
8. Mud Logger – Advise the bottoms up strokes / time.
RCD Element Change Out

1. Conduct TBT prior to start. This must be conducted by the RCD hand
2. If holding backpressure, switch to the secondary flow line.
3. Confirm WHP is remaining constant
4. Isolate the primary flow line
5. Apply pressure to the well via the secondary flow line.
6. Isolate the RCD from the well and MPD choke.
7. Bleed all pressure from beneath the RCD element

8. Unlatch the RCD element / bearing
9. Pull the RCD element /bearing to above the drill floor
10. Remove the element from the drill pipe
11. Install the new element assembly onto the stand.
12. Lower the RCD seal assembly into bearing / spool.
13. Latch the assembly
14. Inflow test the RCD seal.
15. Open the Annular and isolate the secondary flow line.
Dynamic Flow Check

If drilling with a statically underbalanced fluid it is not possible to bleed off all surface
pressure to conduct a flow check. An influx will be taken. This procedure allows for a
dynamic flow check to be conducted whereby pressure is applied to the well, but mud is
being drawn from and discharged to the rigs trip tank. Therefore any gain in the trip tank
when circulating within a discrete flow loop must be coming from the well.
This method can also be used to step pressure down off the formation to better
understand formation pressures.
1. Confirm the trip tank is approximately half full
2. Define the starting back pressure on the well
3. Switch the suction and discharge valves to start drawing from the trip tank and
discharging into the trip tank, isolating the suction pit.
4. Establish a starting level in the trip tank.
5. Monitor this level and confirm if the well is gaining or losing while continuing to
apply the same back pressure on the well.
If there is a gain in the trip tank:
6. An influx has been taken, and steps must be taken to circulate it out.
There is a drop in the trip tank:

7. Inform the Driller that the well is taking fluid.
8. Start to reduce the back pressure applied to the well in increments. Remain lined up
on the trip tank and monitor for a reduction in losses.
If losses do note reduce, continue to reduce the back pressure until at zero surface
pressure. At this stage, inform the DSV that you are attempting to reduce losses,
and that he may need to prepare for pumping LCM.

Tripping in / POOH
Tripping in hole
At this stage the RCD seal element is not installed and the well is being circulated across
the trip tank using the trip tank pump and flow line. The well is being monitored by the
Mud logger and Driller. Once the BHA has been made up, it can be RIH taking returns
through the rigs flow line.
1. Line up the downstream side to take returns.
2. Run in 10 stands, and check if the NRVs are holding.
3. If the NRVs have failed, then POOH and change out the NRVs. If the NRVs hold
pressure, then resume RIH.
4. RIH taking pipe displacement back to the trip tank.
5. Install the RCD seal assembly with the bit 300’ above est. TOC else two stands
short of the casing shoe.
6. Line up to take returns through the MPD equipment
7. Start stripping in hole through the RCD element now taking returns through the
MPD choke until two stands off bottom. Wash down the last two stands and then
circulate bottoms up.
8. Upon completion of a bottoms up circulation divert returns direct to the header box.
Pulling out of Hole
For this example it was assumed the well was to be circulated over to a kill weight fluid
prior to POOH. There obviously occasions where this is neither desirable nor possible.
The well will need to be circulated over to a statically overbalanced mud prior to POOH.
This includes the need to verify swab calculations, and hold a Tool Box Talk ahead of
POOH.
Upon the decision to POOH, the following steps should be taken:

1. Circulate the well clean while reciprocating the pipe over a single.
2. Weigh up the mud in preparation for circulating the well over to a kill weight fluid.
Provide the MPD Service Company with the new mud thus allowing them to
complete their step down chart for a controlled circulation.
3. Circulate the well over to kill weight mud. The MPD system should be able to
control the bottom hole pressure at a constant pressure.
NOTE: Driller will need to circulate the kill and choke lines over to the new kill weight
mud.
4. Once the circulation is complete, the WHP should have reduced to zero. It will now
be possible to flow check the well. This can be done via a flow rate or pressure.

5. POOH following the same precautions for swabbing as per a normal drilling
operation. Slug the drill pipe if needed.
6. Once inside the casing, flow check and then remove the RCD seal assembly.
7. Line up to the rigs flow line and onto the trip tank and POOH conventionally. The
RCD element is normally removed once inside the casing to limit unnecessary wear
on the element and all increase the speed of POOH.
POOH for PMCD

The above procedure described a common means of POOH for an ABP MPD operation
when the mud weight is simply increased to reduce the WHP to zero. For PMCD where
total losses have been incurred and WHP is being reduced by pumping a Light Annular
Mud down the annulus. To POOH in this instance it is necessary to isolate the well with a
plug of some form and then displace the well over to a kill weight fluid above. The
following is a procedure used to POOH on a PMCD well.
1. Position drill bit 50’ above Loss Circulation Zone.

2. Ensure mixing pit, hopper, lines, etc. are completely clean and dry. If water mixes
with the Bentonite-Diesel mixture hydration will begin and the plug may set
prematurely.
3. Prepare Bentonite – Diesel mix in the batch tank
5,000 lbs Bentonite
32 bbls of Diesel
Add barite to achieve desired fluid density
4. Continue pumping down annulus at 280 gpm w/ LAM.

5. Mix cement with bentonite-diesel mixture at cement unit.
a. Mix 9,400 lbs of cement into Bentonite-Diesel Mixture
6. Pump 10 bbl diesel spacer down drill string.
7. Pump 50 bbl DOB mixture into drill string
8. Pump 10 bbl diesel spacer down drill string
9. Displace DOB to drill bit at a rate of 2 bpm with kill weight mud.
10. Slow annulus injection to 2 bpm as DOB slurry exits the bit to achieve a 1:1 ratio of
DOB and mud.
11. Monitor both drill string and annulus pressures. If annulus pressure starts to
increase, shutdown annulus injection and start to POOH.
12. Continue to pump down drill string to slowly force pill out. Displace from drill
string the displacement volume of the stand(s) pulled out of the pill.
13. Pull bit clear of pill. DP injection pressure should decrease when pill exits bit.
14. Allow pill to hydrate for 15 minutes. Maintain annulus pressure.
15. Attempt to circulate KWM through drill string to ensure hole is full of fluid.
Monitor well for losses. If losses still present, prepare to set another DOB plug.
16. Once well is static with a full column of kill weight mud, prepare to POOH.

Other Operating Procedures
Section 8.1.1 elaborated on some of the main procedure likely to be conducted during an
ABP MPD project. The list below names some of the others that could be expected:
Start up Procedure – This specifies the starting valve alignment, what the initial
flow rates need to be, who is stationed where etc. It may also cover issues around
radio usage, and other checks to be closed out before starting to drill MPD.
Drillers Instructions – A dedicated procedure for the Driller that highlights his
responsibilities during MPD.
Kick Detection and Management – Defines what constitutes a kick and then
provided guidance to all as to how it is going to be circulated out of the well.
Valve Manipulation – describes the sequencing for switching flow from one path
to another. It highlights who actually turns what valves and which ones are
opened first.
MPD Fingerprinting – With many HPHT wells, it is necessary to conduct a series
of ‘fingerprinting’ exercises prior to drilling out the shoe as well as when a
change in bottom hole conditions is seen. This procedure would provide the
necessary steps required and define what final outputs are to be produced.
8.1.2 Contingency Procedures

Contingency procedures are similar to an insurance policy, in that you need to pay into it
but hope you don’t have to use it. When developing the procedures, focus should be
placed on securing the situation and only then executing the events to get back to
operating. The procedures need to be simple in their steps.
The following are samples of simplified contingent procedures one could expect on an
ABP MPD project.
Catastrophic RCD element failure

Sudden failure of the RCD sealing element is unlikely but nevertheless still possible. The
only option with a sudden RCD failure is to isolate pressure from the RCD and change it
out. The challenge is to react swiftly enough to stop pumping and isolate the well with
the BOPs without taking an influx. The basic procedure is to:
1. Shut down the mud pumps and close the annular. Open the MPD choke.
2. Isolate the RCD and align to the secondary flow line.
3. Change out the RCD element as per the RCD Element change out procedure.
It is preferred to line up to the secondary flow line to as to regain direct control of
pressure on the well bore. The element change out may take some time, and it is
important to be in a position to both detect if an influx was taken, and then also be able to
circulate it out. Depending on rheology and ROP, it may even be necessary to circulate
bottoms up just to prevent cuttings settling out.

MPD Choke Plugging
Occasionally a piece of debris or cuttings will start to pack off across the choke manifold.
Depending on the design of the MPD control system, the choke will automatically open
to compensate for the increased backpressure brought about from the plugging material.
However, once the choke is fully open, the pressure will resume increasing. Quite often
choke opening is sufficient for the debris to clear on its own. However, if it does not self
clear, then steps must be taken to switch to the secondary choke leg. It is prudent for the
driller to pick up and circulate when conducting the following steps.
1. Open the second leg on the MPD choke.
2. Manually attempt to clear the plugging choke.
3. Isolate the plugged choke using the upstream gate valve.
4. Bleed off trapped pressure and pull the choke out. At this stage the secondary choke
is now managing pressure on the well.
5. Clean out the debris and make a note of what caused the blockage.
6. Reinstate the primary choke and inflow test. This can be done by opening the
upstream gate valve.
This procedure, as with most stated here may vary with the type of MPD system being
employed. Generally though, it is worth seeing of the plugging material can be clear by
“rocking” the choke open and closed. Depending on the downhole pressure tolerance
though, this may not be an option.
NRV Failure
The impact of complete failure of all non return valves in the BHA is varied depending
on the type of MPD operation being performed. If the well is being drilled with a
statically overbalanced mud, where pressure is being kept constant during periods of
“pumps off”, the impact will be minimal. The impact will be that on connections pressure
on the annulus can no longer be held, and so “constant BHP” is no longer possible.
If drilling using a statically underbalanced fluid or doing PMCD, the impact is greater.
With ABP MPD and a statically underbalanced mud, it is necessary to hold back pressure
during connections. This applied back pressure is transmitted down the annulus and up
the drill string. Normally, when bleeding off the standpipe pressure prior to breaking off
the Kelly / TDS, the NRVs would close and seal pressure from below. If they have failed,
then string pressure will not bleed off and it is not possible to proceed with the
connection.
When the failure occurs also dictates the actions necessary. If the Kelly / TDS is still
made up, the first option should be to pump through the NRVs to see if the blockage can
be cleared. If not, then the well will need to circulated over to a KWM and then POOH.
Depending on the pressures being managed, it may be possible to slug the pipe with a
heavy slug and pull out into the casing first. Either way, eventually the NRVs will need
replacing.

If failure should occur when the Kelly / TDS has been removed i.e. the decision then
needs to be whether the Kelly / TDS can be made back up quickly and safely or should
the stab in valve be installed. Essentially, once either has been installed, the well can then
be brought back under control.
Other Contingency Procedures

Section 8.1.2 elaborated on some of the contingent procedures that have been known to
occur more than once. The list below names some of the others that could be expected:
Bit nozzle plugging – this affects the standpipe pressure and flow rate. Any
change in flow rate affects the ECD and thus BHP. The choke will start to react to
this.
Mud Pump Washout – Flow in is generally measured off pumps strokes.
However, If the actual flow rate in is now diminished, the ECD will be less as will
the BHP. The flow meter will detect this drop in flow out, and assume losses.
MPD Pump Failure – If the MPD system incorporates an auxiliary pump, and it
fails during a connection, the choke will look to suddenly close in an effort to trap
pressure. If the pump does not start as the rig pumps start to shut down, the rig
pumps can be allowed to continue pumping until the problem is resolved.
Drill String Leak – A drill string leak is a problem with conventional drilling.
With MPD, the added issue can be a reduction in ECD and subsequent reduction
in BHP.
Rig power failure – If rig power suddenly fails the emergency power tends to only
provide lighting. Therefore unless the MPD equipment is operating of a dedicated
generator, the equipment will not work. Residual energy stored in accumulator
bottles and a UPS may be enough to allow the choke to be closed in if the rig
pumps shut down. The procedure needs to consider the sequence of bringing the
power back up so as to avoid the pumps coming on with the MPD choke manifold
shut in and perhaps bypassed.
Annulus pack off – This would effectively act as a down hole choke. There would
be a reduction in surface pressure causing the choke to close, while at the same
time the BHP would have increased for the same flow rate due to the restriction.
Stuck pipe – Differentially stuck pipe can actually be released using MPD
depending on the pressure boundaries. However, mechanically stuck pipe will
require jarring, and this will be done through the RCD element. This is not ideal
and will need to be reviewed on a case by case basis.
8.2 Procedure Format

Easy to overlook or underestimate, but if consideration is given to the procedure format,
their execution at the rig site can be made simpler. A few experienced based, pointers on
this subject are:
Use Arial font size 12 with 1.5 line spacing. A Driller behind his console wearing
safety classes / reading classes, surrounded by noise and trying to review a set of
procedures encased in grease smeared, laminated sleeves does not need to be
challenged with small italic text.

State the purpose of the procedure on the first page. This is just a few lines
explaining to the reader what the procedure does. It sets the scene and allows the
rig guys to better understand what is ahead of them. It also confirms to them that
they are about to conduct the correct procedure.
Break the procedure down in to clear steps with each step having a unique bullet
point number. Do not use bullet dots as used here for this, since it is difficult to
reference “we are at the 4th dot down”. It is simpler to say, “we are at step 4”
There is merit is stating what the step is going to be, and then providing the
individual activities required by whom to undertake the step. See Figure 1. Error!
Reference source not found.
Figure 1- Suggested format for MPD procedures
Use basic English and keep the sentences short and concise.
Track the revision for the procedure on the front of the document. As the project
develops, changes will be made, and it is imperative that all personnel are
working from the same procedure.
Ensure the procedures refer to the Process Flow Diagram. The valves on the PFD
should have unique labels and the procedures need to refer to these labels. Most
people find it simpler to follow a procedure if they can visually refer to a diagram
showing the flow paths.
Where possible, use the valve number and endorse it with a title that the common
user of the valve knows it as. i.e. “Driller – Open B8, the upper choke line HCR
valve”.
The procedure format should be tailored to the people are going to use them. If the
majority of people are French or Arabic speaking etc, then procedures should be available
in that language. If a procedure becomes to long and complicated, then consider splitting

into two smaller procedures. The procedures provide a tangible place to record some of
the potential hazards associated with conducting the operation. If these are stated at the
beginning of the procedure, they can be referred to in tool box talk ahead of time.
8.3 Managing Procedures at the Rig Site

While developing procedures, continued thought must be given as to how the procedure
is physically going to be enacted at the rig site. It is important to envisage who will be
stood where, who will have radios, what everybody’s first language is, how noisy it will
be around the individuals work area etc. All this should factor into the thought process for
procedure development. A few issues to consider are:
The Driller has pumps and a brake to physically hold, as well as a series of dial
and gauges to watch, plus a floor crew to manage. He cannot be expected to stand
reading out procedures to the MPD Operators, the Mud Loggers, and the
Cementer etc.
A central person needs to orchestrate the procedures. This person should be able
to read the procedures out, be able to read ahead and anticipate what it going to be
happening, be sat in an environment that is both quiet and free from interruption
etc. In most cases a control cabin is ideal. The same person should be able to see
the process flow diagram, and if necessary, be able to mark the PFD valves as
open or closed.
Who is going to need a radio and who will need head sets for the radios? The
cement unit and drill floor can often be noisy places; it is counter productive if
personnel need to keep asking for clarification because they cannot hear the
instructions.
Where are the drilling parameters being displayed. Does everyone who needs to
see a pressure or flow rate, have access to that value. If not, cannot someone else
be given the task of confirming a pressure has bled off etc.
Is someone who may be assigned to a procedure always going to be available?
The Tool Pusher and Company Man are normally occupied in managing the rig.
Tasking them to a specific procedure is not advisable.

8.4 The Procedure Lifecycle
Figura 2 -Procedure Lifecycleis an example of how the procedures should be managed
through a project.
Figura 2 -Procedure Lifecycle

Chapter 9 – Special Problems
1 Rev. 2.0
Table of Contents
9 Special Design Considerations ................................................................................ 9-3
9.1 Pressure Margin for Drilling ......................................................................... 9-3
9.2 Swab and Surge ............................................................................................. 9-4
9.3 Density and Viscosity Effects ....................................................................... 9-6
9.4 Fluid Selection for HPHT Conditions......................................................... 9-10
9.5 Wellbore Breathing ..................................................................................... 9-11
2 Rev. 2.0
Chapter 9
9 Special Design Considerations

In this section, we look at some of the special considerations in the design of MPD wells.
These considerations arise from not only the nature of managed pressure drilling
techniques, but also the type of wells where MPD is usually called for, such as
deepwater, HPHT, and ERD wells. MPD is also indicated in situations where formation
(pore or fracture) pressures exhibit multiple variations with depth- such as in the case
when a low-pressure zone is above a high-pressure zone, or several such formations are
stacked.
9.1 Pressure Margin for Drilling

By definition in overbalanced drilling, the mud weight is selected such that the
circulating pressure is between the pore pressure and the fracture pressure. The minimum
pressure margin between the circulating pressure and the fracture pressure is referred to
as the “drilling window”, and is indicative of the risk of lost circulation during drilling1.
While margins in excess of 1 ppge are common in conventional wells, the luxury of large
margins is not available in HPHT, Deep Water, or Extended Reach wells. In such wells,
the friction loss from circulating up deep annuli increases the ECD and thus reduces the
window. In deep water wells, as a result of the overburden being replaced by sea water,
the fracture pressure is less developed, and the margin between pore pressure and fracture
gradient is narrow in deep water wells.
Figure 1 shows the logic in specifying a drilling window. The purpose of the drilling
window is to provide safety margins for various effects that cause possible decrease or
increase in annular pressure. These include desired overbalance margin over pore
pressure, mud weight uncertainty, swab margin, surge margin, and additional safety
margin for other effects. The typical acceptable margin for deep water, HPHT and ERD
wells, based on industry experience, is about 0.3-0.4 ppge above pore pressure, and 0.7-
0.8 ppge (0.08 – 0.1 SG) below fracture gradient. It is becoming increasingly tolerable to
drill with such low margins in many of the deeper offshore wells.
Typical Drilling Window in HPHT

OB margin, MW uncertainty, Addl Safety margin,
0.1-0.2 ppg 0.1 – 0.2 ppg 0.15 – 0.2 ppg
ECD and Surge, 0.5 ppg min

(surge is typically 0.2 ppg)
Swab margin, MUD WEIGHT

~ 0.05 ppg
Pore Pressure (usually Frac pressure (from

P50 expectation is used) FIT, LOT or FG)
Figure 1. Typical Drilling Windows in Deep, HPHT Wells

1
In a similar manner, the gap between the mud weight and pore pressure is commonly referred to as the
“Kick Margin”.
Special Problems 9-3 Rev. 2.0

Chapter 9
Figure 2 depicts the drilling windows in some wells in the North Sea. As the figure
shows, many fields are being drilled with drilling windows of less than 0.15 SG. In fact,
the Shearwater wells are drilled with a very narrow margin, and in one of the wells, the
drilling window was as low as 0.57 ppge (0.07 SG)2. At a depth of 17,000 ft, this is only
500 psi.
HPHT Drilling Window vs Depth
3000
Tyr Proposed Minimum

3500
4000
Depth TVDRKB m
Shell Norge HPHT 2001

Egret
4500
2/4-14
Kingfisher
Shearwater Erskine
Heron
5000
5500
Commander
Elgin / Franklin
6000
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Drilling Window SG
Figure 2. Typical Drilling Windows in Some North Sea Wells

The problem of narrow margins is exacerbated in multi-layered or stacked reservoirs
where some formations are preferentially depleted, creating large pressure variations
during the progress of drilling. As formations deplete, their fracture pressure also drops
(as discussed in a later section). Frequently, several liners are needed to combat the
situation. MPD is a potential solution in such situations. MPD techniques can allow
navigation through the pressure variations and eliminate some of the liners, while
reducing the potential for losses or kicks that are frequent in such situations.
9.2 Swab and Surge

Surge and swab are among the most critical design considerations in MPD well design.
Both these effects are well known even in conventional drilling, but with the narrow pore
pressure-fracture gradient margins typical in MPD, they become especially important. As
noted in Figure 1, swab/surge account for a large part of the desired drilling window.
2
After SPE 59175, Drilling Windows in Certain UKCS Wells

Chapter 9
When a tubular such as a drill string or casing is tripped into or out of a wellbore filled
with a fluid, transient pressure fluctuation occur, and the fluid pressure at any given depth
can oscillate above and below the static pressure. Fluctuations above the static pressure
are referred to as a pressure surge, while fluctuations below the static pressure are
referred to as a swab. These are analogous to the familiar “water-hammer” problem that
arises out of sudden closure of a valve on a flowing stream of fluid3.
Swab and surge are well studied in the literature. Lubinski et. al.4 were one of the first
people to publish a detailed treatment of the problem for the petroleum industry.
Mitchell5 describes the development a transient model to predict swab and surge
pressures. A combination of fluid inertia, drill pipe elastic response and wellbore
(formation) elastic response contribute to these transient effects. Mitchell’s approach has
since been incorporated into commercial swab-surge simulator tools.
While it is natural to associate a surge with downward pipe movement (trip into hole) and
a swab with upward (out-of-hole) pipe movement, it is important to recognize that swab
and surge are transient phenomena. Even though the initial response during a trip-in is a
surge (caused by, among other effects, fluid compression) , it is immediately followed by
a secondary swab as the pressure falls below the static pressure on its way down (much
like a spring whose end is extended beyond its equilibrium position and then let go).
This phenomenon was described by Rudolf and Suryanarayana6, and later validated with
field measurements by them7. The effect is illustrated in Figure 3, which his for a deep,
HPHT well while tripping one stand of 5” drill pipe (bit at 13,341 ft) into the 8 ½” hole
section at typical speeds, with well TD at 22,000 ft. The secondary swab that
immediately follows a pressure surge is clearly seen. More alarming is the magnitude of
these fluctuations. The pressure surge is nearly 1 ppge, and the swab is nearly 0.3 ppge.
Alarming indeed, when viewed in the context of the narrow margins in deep, HPHT wells
shown in Figure 1, and the even narrower margins that are typical in MPD wells.
3
Wylie and Streeter, Fluid Mechanics
4
Lubinski, A., Hsu, F. H., and Nolte, K. G., “Transient Pressure Surges Due to Pipe Movement in an Oil
Well”, Reprinted in “Developments in Petroleum Engineering- The Collected Works of Arthur Lubinski”,
Gulf Publishing, 1987, pp. 277-332.
5
Mitchell, R. F., “Dynamic Surge/Swab Pressure Predictions”, SPEDE, September 1988, pp 325-333.
6
Rudolf, R. L., and Suryanarayana, P. V., “Kicks Caused by Tripping In Hole In Deep, High-Temperature
Wells”, SPE 38055.
7
Rudolf R. L., and Suryanarayana, P. V., “Field Validation of Swab Effects while Tripping in the Hole in
Deep, High Temperature Wells” SPE 39395.

Chapter 9
Figure 3. Transient Pressure Signature While Tripping in Hole, Illustrating the

Secondary Swab (After SPE 38055)
In HPHT wells, temperature itself causes significant reduction in effective fluid density
(and hence hydrostatic pressure). This, when combined with the swab/surge effects, can
lead to undesirable (and avoidable) kicks or lost circulation events.
In planning and designing MPD wells, therefore, great care must be exercised during
trips, and in pipe handling. Tripping pipe is a wearisome operation, especially in deep
wells, and the tendency to “speed-up” the trip, especially with the bit deep in the well, is
understandable. In MPD operations, however, this tendency should be avoided.
Maximum trip speeds should be estimated using appropriate software tools or programs,
and the effects of temperature should be included in these estimates.
9.3 Density and Viscosity Effects

Drilling fluids are complex fluids consisting of different chemical additives that must
remain stable at high temperatures and pressures. Water-based drilling fluids often
contain bentonite clay and thermal instability is often associated with bentonite-based
fluids. Oil-based fluids often contain organophilic clays that do not experience the large
changes in physical properties with increasing temperature observed in bentonite-based
fluids [Fisk & Jamison (1989)]. Many drilling fluid products are susceptible to thermal
degradation at elevated temperatures. High-temperature gelation occurs in both OBM and
WBM in addition to high-temperature fluid loss [Oakley et al. (2000)]. In WBM, clay
(bentonite) flocculation causes gelation and the situation is further compounded by
thermal degradation of thinners, a drop in pH and an increase in filtrate loss. In OBM, the
interaction of colloidal particles (clays and fluid-loss additives) and breakdown of
emulsifiers may cause gelation. Rheology of the drilling mud needs to be controlled at
HPHT conditions as it may lead to poor hole cleaning, barite sag and a non-uniform
density profile in the annulus. Cesium Formate is being recognized as a potential drilling
fluid for HPHT wells due to its no sag potential and low solids content.

Chapter 9
In HPHT wells, high temperature conditions cause the fluid in the wellbore to expand
while high pressure conditions cause fluid compression. These two conditions have
opposing effects on the equivalent circulation density (ECD) of the drilling fluid and
therefore on the estimation of the bottomhole pressure [Harris & Osisanya (2005)].
Hence, the temperature and pressure effects on ECD must be considered in HPHT wells.
Figure 4 illustrates the difference in ECD estimation with and without consideration of
temperature and pressure. As depth increases, the ECD dependent on pressure and
temperature continues to decrease relative to the ECD assuming constant fluid properties.
This trend is a result of the more dominant effect of thermal expansion, as opposed to
compression from increased pressure [Harris & Osisanya (2005)].
Figure 4. Equivalent Circulating Density in a 17200-ft well [Harris & Osisanya (2005)]
Accurate prediction of ECD in HPHT wells requires the knowledge of the pressure and
temperature dependence of the rheological properties of the drilling mud and the accurate
temperature profile in the well [Rommetveit & Bjorkevoll (1997)]. In the absence of
laboratory data, models are available that predict pressure and temperature dependence of
rheology. Figure 5-Figure 8 illustrates results from a simulator for a 5000 m deep well
and a sea depth of approximately 100 m. In Figure 5 and Figure 6, the density profiles for
OBM and WBM vary with depth, with the density of WBM being more temperature
dependent compared to OBM. In Figure 7 and Figure 8, the equivalent viscosity profiles
for OBM and WBM vary with depth. In the lower part of the well, the temperature is
higher and so are the shear rates. The high shear rate viscosity of OBM decreases with
increasing temperature however the viscosity of WBM is nearly independent of
temperature. Clearly, calculation of density, viscosity and pressure profiles is more
challenging in HPHT wells. (Note that similarly, the temperature will impact the thermal
properties of packer fluids during production).

Chapter 9
Figure 5. Calculated Mud Density Profile for WBM Along the Well with Geothermal
Temperature Profile and Circulation Temperature Profile [Rommetveit & Bjorkevoll
(1997)]
Figure 6. Calculated Mud Density Profile for OBM Along the Well with Geothermal
Temperature Profile and Circulation Temperature Profile [Rommetveit & Bjorkevoll
(1997)]

Chapter 9
Figure 7. Calculated Equivalent Viscosity Profile for WBM Along the Well with
Geothermal Temperature Profile and Circulation Temperature Profile [Rommetveit &
Bjorkevoll (1997)]
Figure 8. Calculated Equivalent Viscosity Profile for OBM Along the Well with
Geothermal Temperature Profile and Circulation Temperature Profile [Rommetveit &
Bjorkevoll (1997)]

Chapter 9
PVT models for wellbore fluids have been developed, and in particular the Zamora et al.
(2000) approach of characterizing the dependence of fluid density on temperature and
pressure is commonly applied in thermal analyses of wellbores. PVT models for wellbore
fluids have been developed, and in particular the Zamora et al. (2000) approach of
characterizing the dependence of fluid density on temperature and pressure is commonly
applied in thermal analyses of wellbores. They propose a quadratic, six-parameter curve
fit for density in terms of temperature and pressure. The parameters of the equation are
determined experimentally. Care should be exercised when extrapolating outside the
experimental data; second order terms may blow up to unrealistic values. It should be
checked that the Zamora model fits the experimental data reasonably well. If not, a
standard interpolation/extrapolation method can be considered, but this should be used
with caution.
A number of correlation based models have been published and can be used when good
PVT data at actual pressures and temperatures is not available for the fluid(s) under
consideration. Such models should be used with care, especially when pressure and/or
temperatures outside the recommended range of the correlation used. There are several
PVT models for crudes (Glassø (1980), Standing (1977)) and brines (Kemp and Thomas
(1987)) but for muds, it is general practice to use the Zamora et al. (2000) approach, or
conduct tests to obtain the density and viscosity curves as a function of temperature and
pressure. Zamora et. al. propose a curve fit for density ρ, in terms of temperature T and
pressure P,
ρ (P, T ) = ρ w [(aoT + bo ) + (a2T + b1 )P + (a2T + b2 )P 2 ] (1)
The six coefficients of the equation, commonly known as the “Zamora” coefficients, are
determined experimentally.
9.4 Fluid Selection for HPHT Conditions

Composition of HPHT Fluids can be considerably different from that of conventional
reservoirs. High pressure and temperature conditions increase the volatility of heavy
compounds that result in gas condensates being richer in heavy compounds. Furthermore,
volatility of water also increases with temperature and hence water can become a major
constituent of HPHT fluids and therefore should be taken into account while studying the
phase behavior of HPHT fluids [Gozalpour et al. (2005)]. Figure 9 compares measured
viscosity of a model volatile oil at a temperature of 392 oF and at pressures above the
bubble point, without water and with 5.4 mole % of water [Danesh (2002)]. A 20%
increase in viscosity is observed in the presence of water. Therefore, higher water content
in HPHT fluids is an important consideration in their phase behavior unlike conventional
reservoir fluids. Many HPHT fluids also have considerable solids content.

Chapter 9
Figure 9. Measured Viscosities of Volatile Oil at Different Pressures With and Without
Water at 200 oC [Danesh (2002)]
9.5 Wellbore Breathing

Another important phenomenon that is particularly evident in MPD (and HPHT) wells is
the so-called “Wellbore Breathing”. The term refers to the flow of fluid from the
formation into the wellbore during non-circulating periods (such as during the making of
a connection). During circulation in deep wells with reduced annular clearances, the
friction loss (the equivalent circulating density, or ECD, effect) can be significant. In
MPD, the effect of minor fluid loss is exacerbated by the narrow margins. During this
period, some invasion of drilling fluid and near-wellbore pressure charging of the
reservoir is likely. Since the bottom-hole pressure drops during non-circulating periods
(flowing friction being absent), there could be periods of underbalance with respect to
near-wellbore formation conditions, leading to a flow back. The elastic contraction of the
wellbore on reduction of bottom hole pressure, and the increase in fluid temperature as it
returns to geothermal conditions, also contribute to this effect. This flow back can be
interpreted as a kick, although this is not a true kick. As soon as circulation resumes, or
pressure balance is achieved, the effect ceases. It can also lead to large volume flow back
during long-lasting non-circulation events such as trips. Typically, most of the flow back
is of the invaded filtrate.
The primary consequence of wellbore breathing is the potential downtime it can lead to
when influx from wellbore breathing cannot be differentiated from a kick. All three
effects- flowback of fluid due to temporary underbalance with respect to near-wellbore
pressure, expansion of fluid due to increase in temperature, and squeezing of the fluid as
the wellbore elastically contracts in response to reduced BHP- combine to cause fluid
gains of the order of barrels in a typical well. Since kick margins are so tight in MPD
wells, this can understandably lead to concern, and time-consuming corrective action that
may not be necessary.

Chapter 9
The flowback volume can be substantial, even with limited degree of underbalance. This
is illustrated below. This is the result of simulations where the fluid exchange during
drilling overbalance and non-circulating events is tracked. The simulator has ultra-fine
grids near-wellbore, and is used for investigation of invasive damage, but has been
adapted to study the wellbore breathing problem in this case. The situation modeled here
is for a the 8 ½” hole section of a deep well with pore pressure of 14000 psi. During
drilling, the degree of overbalance is 1500 psi, and the fracture pressure is close to 15500
psi. During the simulated non-circulating event, BHP momentarily drops to 13500 psi
and the near wellbore pressure charge during drilling is depleted as the pressures
eventually reach equilibrium. The figure shows that, as expected, the flow rate reduces
with time. However, the cumulative influx can be as high as 5 barrels for a 10 hour non-
circulating interval. Of course, the cumulative influx and the flow rate depend upon the
specific conditions of pressure and reservoir properties, but this example illustrates the
order of magnitude of the flow back. Other workers have shown that surface fluid gain
from thermal and wellbore elasticity effects is also of the same order of magnitude.
Thus, it is not unreasonable to expect 5-10 bbls of gain solely caused by wellbore
breathing.
5
Accumulative Flow rate (STB)
Flow back of invaded mud

3 filtrate
0
0 2 4 6 8 10 12 14
Time of flow back (hours)
Figure 10: Wellbore breathing during a non-circulation event, caused by influx.

Chapter 10
Chapter 10 – Economics of an MPD Project
Economics of an MPD Project 10-1 Rev. 1.0

Chapter 10
Table of Contents
10 Economics of an MPD Project ........................................................................... 10-3
10.1 Benefits of MPD ............................................................................................ 10-5
10.1.1 Elimination of a Casing String.................................................................... 10-5
Casing Point Example ........................................................................................ 10-6
10.1.2 Differential Sticking.................................................................................... 10-7
Differential Sticking Example ........................................................................... 10-8
10.1.3 Recovery from Stuck Pipe .......................................................................... 10-8
10.1.4 Lost Returns ................................................................................................ 10-9
10.1.5 Improved ROP .......................................................................................... 10-10
ROP Example................................................................................................... 10-11
10.1.6 Reduced Formation Damage..................................................................... 10-12
10.1.7 Formation Instability................................................................................. 10-12
10.1.8 Ballooning ................................................................................................. 10-12
10.1.9 Improved Hole Cleaning ........................................................................... 10-13
10.1.10 Ability to Drill Further ............................................................................ 10-13
10.2 Costs of MPD vs. Conventional Drilling ..................................................... 10-13
10.2.1 Project Economic Example ....................................................................... 10-13
10.3 References .................................................................................................... 10-16

Chapter 10
10 Economics of an MPD Project

Like all technologies employed in drilling a well, the intent is to make sure the benefit
gained from the technology, outweighs the cost of implementing it. In some instances,
MPD may be the enabling technology, meaning the well could not be drilled utilizing
conventional drilling techniques. An example of this would be a well that has an
extremely tight pore pressure to fracture gradient window. The window could be so
narrow, that the number of casing strings required to reach total depth would be
impractical. In this case, the cost of the technology is justified as long as the project
economics meet the minimum threshold set by the operator.
In many cases, particularly deep water applications, the major savings is the elimination
of a casing string or strings and the associated flat time associated with each casing point.
This includes the time associated with logging, tripping and running pipe. Elimination of
one casing string allowing the well to be down sized can reduce the total cost of the well
by 10% to 25%, depending on the ratio of the spread rate to tangible cost.
In other cases, MPD is utilized as a methodology to either drill more efficiently or to

reduce Non-Productive Time (NPT). A survey of wells drilled offshore between 1993
and 2002 showed that the average NPT had risen to 22% for wells drilled in the Gulf of
Mexico1. Allowing for the fact that current worldwide drilling budgets are estimated at
$234 billion per year, NPT could account for $52 billion in wasted expenditures yearly, if
one were to extrapolate these results. A breakdown of the causes of the 22% NPT is
shown in Figure 1.
Although MPD will never eliminate all NPT, it can significantly impact the largest root
causes of NPT worldwide:
Stuck pipe,
Lost returns,
Wellbore stability and
Ballooning shales.
In addition, MPD can help further reduce cost by minimizing Invisible Lost Time (ILT)
and improve drilling efficiency. Where NPT characterizes time associated to equipment
failures and non-drilling events, ILT events address drilling inefficiencies not related to
NPT. In fact, ILT can mask the true cost associated with NPT. A break down of the
major causes of NPT that could be minimized or eliminated by the effective
implementation of MPD showed that over 40% of that NPT could be affected, accounting
for $20.8 billion a year (Figure 2).

Chapter 10
Gas Flow
Shallow Water 0.3% Other Stuck Pipe
1.1% 11.1%
1.7%
Wellbore
Instability Twist Off
0.6% 4.2%
Rig Failure
20.9% Kick
9.7%
Casing /
Wellhead Failure
5.3% Directional
Completion
Wait On 5.6%
Weather
12.3% Cement
Sloughing Shale Squeeze
1.9% Chemical 10.0%
Lost Circulation Problems
12.8% 2.5%
Figure 1: Break down of major causes of NPT for offshore GOM wells greater than
15,000ft TVD and in less than 600ft water depth.
MPD
Shallow Water Wellbore

/ Gas Flow, Instability,
2.0% 0.6%
Twist Off, 4.2%

Lost
Circulation,
12.8%
Kick, 9.7%
Stuck Pipe,
11.1%
Figure 2: NPT that could be directly addressed by adopting MPD techniques. This
accounts for over 40% of total drilling NPT.

Chapter 10
MPD has also been shown to improve ROP. Examples have been cited savings in excess
of 23% based on the ROP benefits of MPD2
10.1 Benefits of MPD

The typical benefits of MPD are:
Elimination of a casing string
Decreased probability of differential sticking
o Reduced Differential Pressure
o Improved Mud Properties (improved filter cake)
The ability to recover easier from stuck Pipe
o Instantaneous reduction in bottom hole pressure
Reduced probability of Lost returns
o Reduced Differential Pressure
Improved ROP
o Lower Differential Pressure
o Lower solids content of the mud
Decreased formation damage
o Less Differential Pressure
o Less solids content of the mud
o Reduce the use of LCM
Lower probability of formation stability problems
o Lower Pressure Fluctuation
Mitigate the problems associated with Ballooning
o Lower maximum pressure
o Less pressure fluctuation
Improve Hole Cleaning
o Eliminate the need to reduce pump rates to control dynamic losses.
o MPD can effectively disconnect the problem of equivalent circulating
density (ECD) and circulating rate.
Increase the length of hole that can be drilled
o Reduce the BHP during drilling operations
o Improved hole cleaning
Some of the benefits of MPD are discussed in more detail below.
10.1.1 Elimination of a Casing String

The design and cost of a well are directly impacted by the number of casing strings
required to be able to safely drill the well to Total Depth (TD). It is estimated that
eliminating a single casing point will reduce the cost of a well by 10% to 25%, depending
on the ratio of daily cost to tangible cost (the higher the daily cost the lower the savings).
The savings include the cost of the string eliminated, time for evaluating the section and
running the casing and the smaller volume of fluids like mud and cement. Although
some casing points are required by regulation, mainly in the case of protecting fresh
water, most casing points are dictated by the strength of the formation and the formation
pressures that will be encountered.

Chapter 10
In selecting optimum casing points, the well should be designed from TD back towards
surface. As a general rule, a casing point is required at any point where the required mud
weight to control wellbore pressures will exceed the minimum fracture gradient exposed
in the open section, taking into account uncertainty in the pressure and gradient.
Therefore, the tighter the difference in the fracture gradient and the wellbore pressure and
greater the increase in wellbore pressure along a section, the more strings of casing that
will be required to reach TD. In narrow fracture gradient – pore pressure environments,
the total number of casing strings can be significantly impacted by utilizing MPD.
Casing Point Example

Figure 3 graphically show an example of casing point selection for a deep water well. For
conventional design a total of minimum six casing/liner strings will be required to reach
TD. Conversely, using dual gradient MPD, the number of casing/liner string can be
reduced to four. 26” casing and 11-3/4” liner can be eliminated from the conventional
design using MPD.
Figure 3: Example of Eliminating Number of Casing Strings through MPD

Eliminating a string of casing has significant impact on cost directly and can also have
indirect impact on the cost. At a minimum, a casing point will require 2 days to POOH,
run casing, cement, NU on new wellhead and RIH to drill out. In most cases, a minimum
of 5 days is associated with a casing point where logging is required. In deep water and
HPHT applications, it is not unusual to have 10 to 15 days associated with a casing point.
In addition to time related cost, the tangible cost of the casing, wellhead and cementing
hardware must be included. If narrow clearance liners are required, the requirement to
open up a hole size may also add significantly to the cost.
Finally, the indirect cost must be calculated by the impact of hole size on the project.
Once the number of strings is determined and the final casing side at TD is known
(usually dictated by completion size), each of section hole sizes are determined. If it is
possible to eliminate a casing point, the starting hole size for the well can be reduced.

Chapter 10
This impacts project cost by minimizing fluid volumes (mud and cement), results in
increased ROP and decreases cost per foot for casing, as the weight is reduced.
10.1.2 Differential Sticking

Historically, differential sticking is one of the largest causes of NPT. The cost is not only
in the lost time spent trying to free the stuck pipe, but must include the risk of fishing and
sidetracking of the hole if free the stuck pipe is not successful. The potential loss of the
BHA must also be included. The bottom hole assemblies used in many operations today
can run in the millions of dollars.
During conventional drilling operations, the pressure exerted by the fluid in the annulus is
always greater than the formation pressure. When the drill string comes in contact with
the wall cake opposite a permeable formation zone of lesser pore pressure, the drill string
can get stuck to the wall cake against the wall of the hole. The hydraulic force now acts
across the isolated portion of the drill string, holding it in place. The forces holding the
pipe against the formation are proportional to the differential pressure and the area of
contact of the pipe against the wall cake.
MPD will reduce the probability of stuck pipe by lowering the differential pressure
between the annular pressure and the formation pressure. In addition, as there is less
solids in most MPD fluids as compared to conventional fluids, a thinner and tighter wall
cake is created.

Chapter 10
Differential Sticking Example

In the example shown in Figure 4, the differential pressure between annular pressure and
formation pressure for conventional drilling is 817 psi, whereas for MPD it is 216 psi.
Pressure (psi)
8000 9000 10000 11000 12000 13000
12000
Pore Pressure
12500
Conventional Drilling
13000
MPD
13500 Conventional Static Mud
TVD (ft)
Pressure
14000
14500
15000
216 601
15500 Overbalance
16000 817
16500
17000
Figure 4: Overbalance Comparison between MPD and Conventional Drilling
Assuming a 12 ¼” hole and the pressure working across a 30 foot 8” drill collar, with a
wall cake of ¼”, the sticking force for conventional drilling can be calculated as
1,418,566 lbs. Reducing the differential pressure down to 216 psi for MPD, reduces the
sticking force to 375,043 lbs, or 26% of the conventional case. Finally, if using a lower
density fluid allows that wall cake to be reduced to 1/8”, the sticking force will drop to
264,541 lbs (19% of conventional).
10.1.3 Recovery from Stuck Pipe

Once pipe has become stuck, the immediate actions taken to free the pipe will have a
great impact on the probability of freeing the stuck pipe. Studies have shown that if the
pipe is not freed within 12 to 24 hours, the probability of freeing the stuck pipe reduces
dramatically.
There are three major methods used to free stuck pipe:

Mechanically jarring the pipe,
Chemically reducing the wall cake or,
Reducing the differential pressure.

Chapter 10
Several operators have shown that if a slight underbalance condition can be created
across the stuck pipe, it is easier to free the pipe. Some operators pump a low density
fluid around the open hole to reduce the bottom hole pressure. After the pipe is freed, the
influx must be circulated out or bullheaded back into the formation and the fluid system
conditioned prior to continuing normal operations. Although the method has been found
effective, the full process can take several days.
MPD allows change from overbalance to underbalance in minutes by opening the choke,
thereby lowering the surface pressure. Once the pipe is freed and the influx bullheaded
back into the formation, the system can be returned to operational mode in hours, instead
of days. Figure 5 shows the impact of surface pressure on bottomhole pressure when
freeing stuck pipe.
Impact of Surface Pressure on Bottomhole

Pressure
Pressure (psi)
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000 13000
0
2000 Pore Pressure

100 psi SP
200 psi SP
300 psi SP
4000 400 psi SP
500 psi SP
TVD (ft)
600 psi SP
6000 700 psi SP
800 psi SP
900 psi SP
1000 psi SP
8000 1100 psi SP
1200 psi SP
1300 psi SP
10000 1400 psi SP
1500 psi SP
12000
14000
16000
Figure 5: Impact of Surface Pressure on Bottomhole Pressure when Freeing Stuck Pipe
10.1.4 Lost Returns

Along with stuck pipe, lost returns are historically the greatest cause of NPT. Lost
returns occur when either a highly permeable formation is intersected, such as natural
fractures or vugs, or the pressure imposed on the formation by the drilling fluid exceeds
the fracture gradient of the formation. MPD lowers the probability of lost returns by
reducing differential pressure across the formation. Depending on the degree of losses
encountered with conventional techniques, MPD may eliminate losses, reduce the extent
of the losses or extend the length of the section drilled before losses occur. The ability to
control losses by varying pressure will also mitigate problems associated with trying to
fight lost returns using conventional methods. Currently losses are fought with lost
Chapter 10
circulation materials and by reducing the circulating rate. In the event of significant
losses, gunk squeezes are utilized. In some cases, the fluid system is converted from an
oil based system to a water based system when fighting severe losses. The use of lost
circulation material, gunk squeezes and water based fluids damages the properties of the
fluid system and can damage the permeability of the formation.
10.1.5 Improved ROP

It is common knowledge in the drilling industry that lower the differential pressure from
the annular fluid to the formation, the higher the rate of penetration. Bourgoyne and
Young demonstrated the relationship between ROP and differential pressure in 19743,4.
They theorized that the hydrostatic pressure of the drilling fluid exerts a force against the
rock that is being penetrated, thus requiring more energy to remove the rock. At the same
time, a filter cake is deposited due to spurt loss of drilling fluid. The bit cutters must
remove this deposit along the formation being penetrated.
Bourgoyne and Young created a correlation between mud weight and ROP. Using their
correlations, graphs can be generated to estimate the improvements that can be expected
with reduced mud weights (Figure 6).
Rate of Penetration vs. Mud Gradient (ECD)
Mud Gradient ECD (ppg) Pore pressure (ppge)
120
Rate of Penetration (ft/hr)
100
ROP 45 ft/hr with MPD
(ECD: 14.7 ppg)
80
ROP 20 ft/hr with
60 conventional Drilling
(ECD: 15.34 ppg)
40
20
0
13.5 14 14.5 15 15.5 16
Mud Gradient ECD (ppg)
Figure 6: Bourgoyne and Young Correlation for ROP improvement
Moore also studied the relationship between overbalance and ROP. He theorized that the
solids content of the fluid system was the main cause of reduced ROP5.
As the drilling of rock is always done with the pumps on, the ROP is negatively impacted
by ECD. MPD both lowers the differential pressure across the formation and reduces the
solids content of the fluid system. Saponja, et al, used the theories of Bourgoyne and

Chapter 10
Moore to justify the implementation of managed pressure drilling in Canada6. The use of
MPD increased ROP by a factor of 2.5, reducing the overall drilling cost by 20%.
ROP Example
Figure 7 shows the schematic of a 16,100 ft vertical well. The well is cased off by 9-5/8”
casing at 8900 ft and the surface equipment set-up allows drilling of the 8-1/2” hole
section through reservoir in MPD mode using a back-pressure choke. The top of reservoir
is at 15,100 ft with a pore pressure of 11228 psi (14.3 ppg EMW).
Figure 7: Well Schematic – Comparison of Overbalance between Conventional

Drilling and MPD

Chapter 10
Assuming a minimum kick margin of 0.4 ppg from well control perspective, it is required
to maintain a circulating pressure equivalent to 14.7 ppg EMW at the reservoir top during
all operations. If the well is drilled conventionally, the drilling fluid density has to be 14.7
ppg to ensure that the minimum overbalance is maintained with static mud in the well
during non-circulating periods such as pipe connections or tripping. During drilling, the
overbalance will be higher due to annular friction pressure, which will depend on the
circulation rate, annular geometry and drilling fluid density. The hydraulics of this
example is modeled using Neotec’s Welflo7 program7. For a flow rate of 450 gpm and
14.7 ppg mud used in conventional drilling, the circulating pressure at the reservoir top is
12045 psi (ECD: 15.34 ppg) with an overbalance of 817 psi. Conversely, for same flow
rate of 450 gpm and a lighter mud of 12.74 ppg, using a back pressure of 1000 psi
through MPD choke, the circulating pressure equivalent to 14.7 ppg ECD can be
maintained at the reservoir top resulting in a much lower overbalance of ~ 216 psi
(Figure 4).
Using the Bourgoyne and young Correlation shown in Figure 6 it can be seen that ROP
can be increased to 45 ft/hr from 20 ft/hr, if overbalance to pore pressure (14.3 ppge) is
reduced by MPD with a lower ECD of 14.7 ppg than 15.34 ppg as in the case of
conventional drilling.
10.1.6 Reduced Formation Damage

MPD can reduce formation damage by lowering the differential pressure and by having
fewer solids in the fluid system. Lower differential pressure will help control the depth
of invasion into the formation. Lower solids content will allow for a tighter filter cake
and will be less damaging to the formation. In addition, MPD can reduce the use of
damaging LCM.
The reduction of solids and fluid invasion to the reservoir can result in improved
productivity.
10.1.7 Formation Instability

In conventional drilling operations, the starting and stopping of circulation, for activities
like connections, creates transient pressure fluctuations. These fluctuations exacerbate
formation stability problems by creating stress cycles in the formation.
MPD avoids pressure fluctuations by manipulating the surface pressure to maintain a

constant bottom hole pressure. The elimination of the pressure fluctuation will mitigate
formation instability events.
10.1.8 Ballooning
Ballooning, or wellbore breathing, is the term given formations that take fluid as they are
subjected to high annular pressure and will give the fluid back when the pressure is
reduced. In critical operations, ballooning can lead to excessive NPT or ILT as the wells
is circulated to determine if the gains are a kick or ballooning. Incorrect interpretation of
ballooning normally leads to raising the mud weight. This has the knock on effect of
lower ROP and the risk of lost returns and stuck pipe.

Chapter 10
MPD will mitigate the problems associated with ballooning by eliminating the high and
low pressure events. If the BHP is kept constant, ballooning will not occur.
10.1.9 Improved Hole Cleaning

In many instances, the on set of lost circulation is mitigated by the addition of LCM and
by reducing the circulating rate. The circulation rate is reduced to lower the ECD,
thereby reducing the differential pressure across the formation. As reducing the
circulating rate is detrimental to optimum hole cleaning, additional time is spent
circulating the hole clean.
MPD eliminates the correlation between circulation rate and induced pressure.
Therefore, MPD allows circulating at optimum rates for hole cleaning.
10.1.10 Ability to Drill Further

By eliminating the impact of ECD, MPD will allow longer hole sections to be drilled
before the fracture gradient of the formation is reached. The lower annular pressure and
improved hole cleaning may allow higher angle wells to be drilled, extending the
reservoir section.
10.2 Costs of MPD vs. Conventional Drilling

The cost of MPD systems vary greatly. Nearly all MPD systems utilize a Rotating
Control Device (RCD). The cost of a RCD varies from $1200 per day for low pressure
static devices on land to $5000 per day for high pressure devices offshore.
The applied back pressure systems range from $3,000 to $10,000 per day depending on
the level of automation, the inclusion of the back pressure pump. The normal crew
compliment for these systems range from 2 people to 6 people per day. This equates to
an additional $2,400 to $9,000 per day. In total, the cost of an applied back pressure
system can range from $7,500 for a manual system to $20,000 for a fully automated
system.
For pressurized mud cap operations, the only cost is the rotating control head. If the rig
does not have sufficient pumping capacity, a rental pump may also be required.
The Varco CCS system is currently priced at $20,000, including 4 operators. The
Riserless Mud Recovery System is billed out at approximately $900,000 per well, plus
mobilization and personnel. Many of the other MPD products are not fully commercial,
so pricing structures are not fixed.
There is also a cost associated with designing the MPD project. For simple projects, this
cost may be insignificant or included in the equipment rental. For more complicated or
higher risk projects, the engineering and project management cost can exceed $500,000.
What is critical is to verify that the gains from implementing the technology will be
greater than the cost.
10.2.1 Project Economic Example

One of the benefits of MPD is the reduction in operating days of the well. As discussed in
Section 6.1.5, the operational time savings can come from reduced drilling time due to
higher ROP as well as reduced tripping time because of less number of bit-trips to

Chapter 10
complete a hole section. In MPD operations, because of operational time savings, the
total number of days can be significantly reduced in comparison to conventional drilling.
As mentioned previously the savings must be high enough to offset the total cost for
implementing the technology. The total cost of a MPD project and its economic viability
will significantly depend on this additional MPD related equipment and service cost. It is
therefore important to perform a detailed cost analysis of a MPD project and compare
with the cost of well drilled conventionally.
In what follows is a simplified cost analysis for drilling a 7200 ft 8-1/2” hole section of
an example onshore well using MPD versus conventional drilling. Table 1 shows the
comparison of operating days between MPD and conventional drilling for this hole
section. In this example, drilling time for MPD is reduced to 6.67 days from 15 days for
conventional drilling because of higher ROP of 45 ft/hr against an estimated 20 ft/hr for
conventional drilling (Figure 6). The average number of bits decreases from 9 to 4, due
to the higher ROP, assuming a constant number of hours for each bit. Tripping time for
MPD is also reduced to 4 days as compared to 6.8 days for conventional drilling because
of less number of bit trips. This is slightly offset by the slower tripping time (1800 ft/hr
compared to 1400 ft/hr., caused by working under pressure. The overall savings is 11.1
operating days.
Table 1: Comparison of Operating days between MPD and Conventional Drilling for an
example onshore well

Chapter 10
.Table 2 shows a cost comparison between MPD and conventional drilling based on the
savings on operating days as above. According to this cost estimate, for a spread rate cost
of $ 50,000 per day, a savings of $ 0.6 million can be realized on standard rig operating
cost. Considering the additional expense on specialized MPD equipment and services, a
net savings of $ 0.35 million is expected for MPD operation. For this example, the MPD
project is economically viable with cost savings potential.
Table 2: Economic Comparison of MPD vs. Conventional Project

Conventional Drilling MPD Operation
Operating Days Operating Days
21.8 10.7
Daily Daily Total
Cost Parameters Total Operating Lump sum Lump sum
Operating Total Cost Operating Operating Total Cost
Cost Cost Cost
Rate Rate Cost
US $ / day US $ US $ US $ US $ / day US $ US $ US $
MPD Equipment and Services (Dynamic
$13,000 $139,100 $139,100
Annular Pressure Control)
Rotating Control Head $2,000 $21,400 $21,400
MPD well Design $50,000 $50,000
Spread Rate Cost (Rig Day Rate, Fuel,
$50,000 $1,090,000 $1,090,000 $50,000 $535,000 $535,000
Mud Logging, Supervisor, Trailers etc.)
Total Cost $50,000 $1,090,000 $1,090,000 $65,000 $695,500 $50,000 $745,500
Cost Savings for MPD Operation $344,500

Chapter 10
10.3 References
1
Martin, M. D., “Managed Pressure Drilling Techniques and Tools”, Thesis, Office of
Graduate Studies of Texas A&M University, May 2006.
2
Hareland, G., Olea, I., Shirkavand, F., Teichrob, R., Kutsamsi, A., “Advanced Drilling
Simulation Proves managed-Pressure Drilling (MPD) Economical in Gasfield
Developments in Western Canada”, CIPC/SPE Gas Technology Symposium, 16
to 19 June 2008.
3
Bourgoyne, A.T., Chenevert, M.E., Millheim, K.K., Young, F.S. “Applied Drilling
Engineering, SPE Text Book Series, Vol. 2”, 2005.
4
Bourgoyne, A., Young, F., “A Multiple Regression Approach to Optimal Drilling and
Abnormal Pressure Detection”, SPE 4238, August 1974.
5
Moore, P., Drilling Practices Manual, Penwell Publishing Company, 1986, 247-190,
343-362.
6
Saponja, J., Adeleye, A., Hucik, B., “Managed Pressure Drilling (MPD) Field Trials
Demonstrate Technology Value”, IADC/SPE Managed Drilling Conference,
April 2005.
7
Wellflo7 documentation, Neotechnology Consultants, Calgary, Canada.

Chapter 11
Chapter 11 – HSE and Project Management
HSE & Project Management 11-1 Rev. 2.0

Chapter 11
Table of Contents
11 HSE and Project Management ........................................................................... 11-3
11.1 Overview ........................................................................................................ 11-3
11.2 HSE Elements ................................................................................................ 11-3
11.2.1 Operating Procedures .................................................................................. 11-3
11.2.2 Safety Elements .......................................................................................... 11-4
11.2.3 Hazard Identification (HazId) ..................................................................... 11-4
Bow Tie Diagrams ............................................................................................. 11-6
11.2.4 HazOp ......................................................................................................... 11-7
11.2.5 Critical Review of Drilling Procedures....................................................... 11-9
11.2.6 Drill the Well on Paper (DWOP) ................................................................ 11-9
11.2.7 Crew Training ............................................................................................. 11-9
11.3 Regulatory Authorities ................................................................................... 11-9
11.3.1 Review of NORSOK Regulations for MPD ............................................... 11-9
11.3.2 MMS and MPD ......................................................................................... 11-11
11.3.3 UK Health & Safety Executive (HSE) and Canadian Energy Resource
Conservation Board (ERCB) ............................................................................... 11-12
11.4 Project Management Elements .................................................................... 11-12
11.4.1Project Scoping .......................................................................................... 11-13
Technical Feasibility and Basis of Design (BOD) ........................................... 11-13
Components of the Basis of Design ................................................................. 11-14
11.4.2 Design and Planning ................................................................................. 11-16
Drafting Operational Procedures ..................................................................... 11-17
11.4.3 Implementation ......................................................................................... 11-18
11.5 Appendix A – IADC HSE Guidelines ......................................................... 11-19

Chapter 11
11 HSE and Project Management

11.1 Overview
Both managed Pressure Drilling and Underbalanced Drilling are drilling operations with
an HSE focus. As such, HSE management structure impacts contract and project
management. This chapter introduces the HSE elements of MPD and how it incorporates
in the project management framework.
11.2 HSE Elements

Once equipment is designed and operating procedures developed, an integrated hazard
and operability (HazOp) assessment should be conducted to ensure potential hazards are
effectively managed and that appropriate safety controls have been put in place. The
HazOp should occur with sufficient lead-time to allow implementation of changes to
equipment and procedures prior to operational startup. The meeting should be attended
by members of the MPD project team, service companies, operations staff from the rig,
and a third party HazOp chairman who facilitates formal review of the project. Issues
and follow-up actions identified during a HazOp are formally documented and “closed
out” before operations begin. Equipment design is only one element of a comprehensive
MPD safety management system. Other key HSE elements include:
Procedures for special rig operations and emergency contingencies

Integrated safety control systems for effective site management
Formalized hazard review called HazId
Formalized safety review process called HazOP
Formalized procedure and program review session called DWOP
Specialized training for rig crews and service personnel
What follows is a brief overview of the HSE framework process. A comprehensive

review of HSE guidelines and specifics on HazIds and HazOps as published by the IADC
is located in Appendix A – IADC HSE at the end of this chapter.
11.2.1 Operating Procedures

The development of procedures, which include all services, is essential to ensuring a safe
and cost-effective operation. The cornerstone to an operation’s systematic procedures is
an operational strategy document which links the company’s core operational policies
with critical managed pressure operations, namely:
Tripping strategy
Well control strategy
Based on the strategy document, normal and contingency operating procedures should
integrate into one document with a mechanism for change control. Training and drills on
procedures are essential to bring all rig crews up to standards on these procedures.
Further detail on project management structure is included in Section 11.4 and examples
of the procedures that must be developed are listed in Chapter 12.

Chapter 11
11.2.2 Safety Elements

A number of key safety elements should be designed into the layout of MPD surface
equipment, which is discussed below. Not all are required for each operation or
installation, but all must be considered at the design stage:
Gas Detection – Comprehensive coverage of flammable and toxic gas detectors
should be provided at both the surface equipment and rig floor to warn of gas
releases.
Fire Detection – The surface equipment should have a fusible loop system of fire
detection, which automatically initiates emergency shutdown of the surface
equipment and initiates deluge.
Emergency Shutdown System – The surface equipment should be protected by an
emergency shutdown system, which isolates the surface equipment from the well
in the event of an emergency. ESD are also automatically activated by abnormal
process parameter (level and pressure), fire detection or manually by push buttons
located around the surface equipment and drilling rig.
Water Deluge System – Fixed water deluge should be provided for the surface
equipment and drill floor, which can be remotely activated to extinguish fires and
maintain equipment below failure temperatures in the event of a fire.
Liquid Segregation – All surface equipment should be located within a bundled
area to prevent the development of running fires in the event of a hydrocarbon
liquid spillage.
Breathing Apparatus – Breathing apparatus should be provided at the surface
equipment and drill floor to allow escape from a toxic environment.
Zoning – Zoning of electrical equipment is specified in API RP 500B. This
classifies areas around rigs and hydrocarbon containing structures for the safe
installation of electrical equipment.
11.2.3 Hazard Identification (HazId)

An assessment of the potential major hazards should be performed early on in the project.
The objective is to give consideration to the major hazards that could occur during an
MPD operation and qualitatively reviewing the measures provided to prevent their
occurrence and to recover from their consequences. This includes a critical review of the
provision of fire and gas detection measures, active fire protection and breathing
apparatus as well as the evacuation routes and muster points. In this way a consensus can
be arrived at as to the acceptability of the risk posed by the hazard and any additional
preventive or mitigating measures required rendering the risk tolerable.
One method is to follow systematic ISO approach (Figure 1) to identify what new
hazards are to be introduced to the operation and collect them in a hazard register, .
Hazards are then ranked according to risk severity using a risk matrix, (Figure 2.) A
sample identification of a hazard is shown below:
Hazard - Reservoir fluid
Threat – Intentional loss of primary well control barriers (I.e. mud column)
Exposure – All personnel on rig floor, about ten times a day
Possible Consequences – Loss of well control
Escalation Factor – Blowout
Control of Hazard – Barrier Policy; Training
Chapter 11
H-25.04 Cold stress

H-25.05 High humidity
H-25.06 Vibration
No. HAZARD DESCRIPTION H-25.07 Work stations
H-01 Hydrocarbons H-25.08 Lighting
H-01.01 Crude oil under pressure H-25.09 Incompatible hand controls
H-01.02 Hydrocarbons in formation H-25.10 Awkward location of workplaces and
H-01.03 LPGs
machinery
H-01.04 LNGs
H-01.05 Condensate, NGL H-25.11 Mismatch of work to physical abilities
H-01.06 Hydrocarbon gas H-25.12 Mismatch of work to cognitive
H-01.07 Crude oil at low pressure abilities
H-01.08 Wax H-25.13 Long and irregular working
No. HAZARD DESCRIPTION
H-01.09 Coal hours/shifts
H-14 Open Flame
H-02 Refined Hydrocarbons H-25.14 Poor organisation and job design
H-14.01 Heaters with fire tube
H-02.01 Lube and seal oil
H-14.02 Direct fired furnaces H-25.15 Work planning issues
H-02.02 Hydraulic oil
H-02.03 Diesel fuel
H-14.03 Flares H-25.16 Indoor climate
H-02.04 Aviation fuel, petrol H-15 Electricity H-26 Psychological Hazards
H-15.01 Voltage > 50 - 440V in cables
H-03 Other Flammable Materials H-26.01 Living on the job/away from family
H-15.02 Voltage > 50-440V in
H-03.01 Cellulosic materials H-26.02 Working and living on a live plant
equipment
H-03.02 Pyrophoric materials H-26.03 Post traumatic stress
H-15.03 Voltage > 440V
H-03.04 Carbon fibre reinforced
material
H-15.04 Lightning discharge H-27 Security Related Hazards
H-15.05 Electrostatic energy H-27.01 Piracy
H-03.04 Dry vegetation
H-16 Electromagnetic Radiation H-27.02 Assault
H-04 Explosives
H-16.01 Ultraviolet radiation
H-04.01 Detonators H-27.03 Sabotage
H-16.02 Infra red radiation
H-04.02 Conventional explosives H-27.04 Crisis
H-16.03 Microwaves
H-04.03 Perforating gun charges H-27.05 Theft, pilferage
H-16.04 Lasers
H-04.04 Explosive gases
H-16.05 E/M radiation: high voltage ac H-28 Use of Natural Resources
H-05 Pressure Hazards
cables H-28.01 Land take
H-05.01 Bottled gases under pressure
H-05.02 Water under pressure H-28.02 Water
H-05.03 Non hydrocarbon gas under H-28.03 Air
pressure in pipeworks H-28.04 Trees, vegetation
H-28.05 Gravel
H-29 Medical
H-29.01 Medical unfitness
H-29.02 Motion sickness
H-30 Hazardous Goods
H-30.01 Dangerous goods in transport
activities
Figure 1: Sample ISO hazard listing.
Probability of Occurrence
A B C D
Unlikely - Has never Occasional - Has Probable - Could Frequent - Could
occurred in industry. happened in the happen on this happen on this
company at least project at least once project several times.
once. or in the company
several times.
4
Risk
3
2
1
Intolerable - Highest priority, immediate attention required.

Excessive - Higher priority, further evaluation required.
Manageable - Monitor under continuous improvement program.
Negligible - No further action required, monitor for improvement.
Figure 2: Sample Risk matrix for hazards.

Chapter 11
The process which can be followed is shown diagrammatically in Figure 3:

1. Hazards are reviewed and collected in a hazard register through a HazId
workshop / meeting.
2. Identified hazards can be classified according to the risk they impose on the
project by using a risk matrix.
3. High risk items must be further investigated to reduce imposed risk. These
identified risks can be addressed in the company’s safety case, addressed in
appropriate procedures, and in training.
4. Medium risk items must be discussed further to ascertain that the risk they impose
is manageable and As Low As Reasonably Practicable (ALARP).
5. Low risk hazards are not to be ignored, and should be vigilantly monitored during
the process to ensure they do not escalate to a medium or high risk.
Potential Risk Risk Management Level of Control

Level Objective
HAZID High Risk Reduce to medium or low risk

HSE Case
Systems
Hardware
W orkshop Hazard and Effects Demonstrate risk reduction s to Procedures
Register (THESIS) Medium ALARP
Training
Risk
HSE - MS
Continue to manage for
Reviews Low Risk improvement
Procedures
Training
Figure 3: HazId process.

A tool which is used to analyze High and Medium risk hazards is referred to as the Bow
Tie diagram.
Bow Tie Diagrams

Bow tie diagrams are referred to as such as they resemble a bow tie once complete
(Figure 4: Sample Bow Tie diagram.) It visually demonstrates and communicates the link
between controls and the management system.
The presence of a hazard could result in a top line event (presence of hydrocarbon fluid,
with loss of mud column, could result in a blowout). Threats that may cause the top line
event to occur are collected on the left side of the diagram. Barriers to each threat are
then systematically reviewed and placed in front of the threat. One can very quickly see
that where less than two barriers exist, a potential problem could arise. A further barrier
should be put in place.

Chapter 11
The right side of the diagram lists the possible consequences of the top line event
occurring. Should such an unfortunate incident occur, a plan to mitigate the severity of
those consequences must be put in place. Each consequence is then analyzed and
mitigating recovery measures also listed for each one.
Prevention Mitigation
Figure 4: Sample Bow Tie diagram.
11.2.4 HazOp
A HazOp is a formalized, systematic method for identifying possible hazardous
operational problems or hazardous situations that may occur during normal plant
operation or transient periods of operation, such as start-up, shutdown, commissioning,
and maintenance. The process addresses hazard and operability issues associated with
surface facilities.
There are many methods for performing a HazOp. The main goal is to perform a
systematic review of the process chaired by an independent facilitator who is familiar
with the methodology. The process identifies deviations from the design parameters,
identifies the consequence of an occurrence, identifies safeguards that are in place,
quantifies the probability of occurrence, and quantifies the consequence if the deviation
occurs (risk to property, the environment and/or life). If the combination of probability
of occurrence and consequence of occurrence are outside of guidelines set by the
company, a plan must be identified to bring these with in the company’s guidelines. The
methodology includes a formal ACTION CLOSEOUT PROCESS to identify and resolve
potential safety issues raised during the review before operational start-up.
The central document in the HazOp process is the Process and Instrumentation Diagram,
often referred to as the P&ID. The P&ID allows the analysis* of nodes within the surface
facility to be analyzed, usually encompassing specification breaks (ie; the high pressure
piping node will include all pipework to the choke, but not include low pressure piping
after the choke). Figure 5 highlights a high pressure node, while Figure 6 shows a low
pressure piping node.
*
Complete process description of the HazOp process is included in the IADC HSE Guidelines, Appendix
A – IADC HSE .
Chapter 11
It should be noted that a properly executed HazOp is a review of the planned design. It is
not a meeting or forum to re-design the facility. As the P&ID is the central document, its
accuracy and intent is critical. The analyzed and completed P&ID becomes a frozen
document, with any alterations planned to it done so under a supervised Management of
Change process. The P&ID becomes the process map for rig up and a key
troubleshooting guide should contingency procedures during drilling be enacted.
Note that the process does not include sub-surface piping (ie; casing, tubing, downhole
jewelry) and does not include a procedure review.
Figure 5: Sample P&ID highlighting the primary flow path node (in blue) to the choke
manifold.
Figure 6: Low pressure piping node, in blue.

Chapter 11
11.2.5 Critical Review of Drilling Procedures

As neither the HazId nor HazOp deal specifically with the procedures that have been
developed, a specific workshop is held to do just that. The next task in the HSE
framework is the systematic review of the developed normal operating, contingency and
SIMultaneous OPerationS (SIMOPS) procedures (Section 11.3). On finishing this review
cycle, the procedure documents, like the P&ID are frozen, and should only be changed
through the MOC process.
11.2.6 Drill the Well on Paper (DWOP)

Following the review of the MPD specific operating procedures, the next task is to roll up
everything that has been prepared to date, and step through each phase of the drilling
process. The DWOP combines together, drilling, managed pressure operations, and
completions procedures in one exercise. Key operational staff from the Operator, drilling
contractor and MPD contractor makes up the team for the DWOP. Any final tweaks to
the developed procedures are brought to the surface in this meeting.
11.2.7 Crew Training

The final task prior to MPD operations is crew training. Successful completion of this
task usually results in signing off on the HSE certificate†, allowing operations to
commence. In the project management framework, crew training can occur at various
stages:
The MPD planning team (Project managers, drilling engineers, operations
managers, and subsurface team) undertakes an MPD well design course to
understand the various design aspects and equipment specifications of an MPD
project.
The drilling operations group undertakes practical, specialized, scenario training
specific to the project. Scenarios are built in a drilling simulator and the well
drilled via computer.
After rig up of the MPD equipment, the rig crew undergoes procedural and
familiarization training on the rig site. Dry runs of various operations, such as
RCD element change out, are planned for both the day and night crews.
11.3 Regulatory Authorities

11.3.1 Review of NORSOK Regulations for MPD
In conventional drilling, annular barriers are totally independent of each other. In MPD,
since both barriers require pressure containment at surface, they share common wellbore
elements (WBE) below the BOP rams. This dependency also happens in all live well
operations, such as snubbing or production well logging. As such, specific review of
common WBEs and their risk exposure is required by Norwegian regulatory authority
NORSOK. The process adopted by NORSOK is a rigorous HazId analysis of the WBEs
with an assurance that any exposed risk is as low as practicably tolerable.
†
The actual documentation for cataloguing the steps of the HSE framework are usually dictated by the
Operating Company, and as such, the nomenclature of the various certificates may vary.
Chapter 11
Figure 7 - In conventional drilling, the drilling fluid and BOPs are considered independent
barriers (blue); while in MPD operations (shown on the left) common Well Bore Elements are
shared as pressure containing devices.
In this sense, Norway regulations allow MPD, as long as the analysis and review is
carried out. A sample HazId is shown in Table 1.

Chapter 11
Table 1 - Example HazId analysis of common Wellbore Elements for MPD operations
11.3.2 MMS and MPD

The US Department of the Interior Minerals Management Service Gulf Of Mexico OCS
Region (MMS) has issued a Notice to Lessees document regarding the use of MPD in the
Gulf of Mexico, and is included in the appendix of this document. The document
outlines the criteria required for a Lessee to conduct MPD operations in the Gulf of
Mexico with either a statically UB or statically OB drilling fluid. It uses a planning
operational matrix as a guideline (Figure 8).

Chapter 11
Figure 8 - Guideline operational matrix for MPD in the Gulf of Mexico as sanctioned by
the US regulatory authority, the MMS.
11.3.3 UK Health & Safety Executive (HSE) and Canadian Energy

Resource Conservation Board (ERCB)
UBD operations have matured in both the UK and Canada, along with their regulatory
policies. Both countries have policy documents drafted to address UB operations, but
have not (to date) developed specific MPD policies. As UBD may arguably be
considered the ultimate managed pressure environment, any issues arising with MP
operations are covered under existing regulation.
11.4 Project Management Elements

Experience has proven that planning, implementing and sustaining a high-quality MPD
project requires effective management and provision of a skillful team through each
phase of the project. A project management template provides an excellent starting point
for the MPD project manager to plan and implement the MPD project and optimally
utilize and manage project resources. In addition to delivering an on-time and on-budget
project, an effective project manager should also leave behind a well-documented legacy
of the project, which includes everything from the Concept Selection (Basis of Design),
the Detailed Design (P&IDs etc.), the Implementation (well reports etc.) and culminating
in the evaluation of results.
As mentioned in the previous section, management of HSE issues and overall project
management are closely linked. The project framework may be broken down into three
broad phases, as shown in
Figure 9:

Chapter 11
Figure 9: Generic MPD project timeline.
11.4.1Project Scoping
The initial phase serves as a foundation to the rest of the project, in which a business case
(if required) is established for drilling MPD. The Candidate Selection phase high grades
the best potential MPD candidates and eliminates the poor ones. The Technical
Feasibility section then outlines the means to drill the selected candidates which results
in the blueprint for drilling the well called the Basis of Design (BOD) document. As
technical success alone is not a defining characteristic of a successful project, an
Economic Feasibility is also conducted. In cases where MPD is already fully accepted,
work may begin directly with the Technical Feasibility section. A fundamental goal of
this first phase is to provide enough information to allow the management team to answer
the question “Should we or should we not proceed with this project?” If the integrated
business case points to a “YES” answer, the Project Approval decision gate is passed.
Technical Feasibility and Basis of Design (BOD)

As the Basis of Design document forms a cornerstone of the overall project, a closer look
at it and the Technical Feasibility phase is warranted.
If a sound business case for MPD already exists for field in question and asset buy-in at
all levels is already obtained, this is the likely first phase the Operator requires. During
the initial data gathering trip meetings will be held with the drilling and asset teams to
understand the drivers for considering pressure control techniques for drilling the target
fields.
The technical feasibility forms the cost element of the Project Scoping phase as it details
the required equipment for MPD. The objective of the technical feasibility study is to
determine if managed pressure drilling is technically feasible and economically viable.
The rigorous study should be conducted using a state-of-the-art multiphase flow analysis
program (ie; Wellflo7 or other), as well as a drill string mechanics program. It will result
in a preliminary design, detailing:

Chapter 11
suitability of candidate,
drilling fluid selection,
conveyance/drilling technique selection and technical feasibility,
equipment requirements, and
sensitivity analyses.
The sensitivity analyses will focus on the pressure control available, and whether
contingency UBD or Mud-Cap operations will be required, depending upon the pore and
fracture pressure profiles in the drilled section. The final outcome of this task is a
complete Basis of Design for MPD Operations‡, which is a critical document in any
MPD project§, and drives the further steps in the project, such as project management,
HSE, detailed engineering, procedure development and implementation.
This phase may not be skipped in any MPD project as the Basis of Design forms the
reference document for the rest of the campaign. If is absent, or created in error, the rest
of the project is held hostage to this weakness.
Components of the Basis of Design

The following describes a methodology used for developing a basis of design document
which ultimately provides insight into the complexity of an MPD candidate, circulating
system design, equipment suitability and well controllability. This methodology is based
on theoretical basis for hydraulic / multiphase flow analyses as well as practical
understanding and experience gained in MPD operations worldwide. The modeling
sequence must be applied in every technical feasibility study of a MPD candidate.
However, the extent of reservoir uncertainty for a candidate will dictate the degree of
sensitivity analyses required to develop the sought “robust” design.
The following are the key tasks and deliverables:
1. Gather data required for the analysis and to become familiar with the problem.
This will require a trip to the intended rig (if selected) to ascertain usability and
identify modifications.
2. Drill string/coiled tubing mechanics analysis should consider the following:
Set-down weight‡ , overpull available, and surface equipment requirements,
Fatigue damage to the string,
Recommendation on string design and material selection,
Alternative conveyance methods should also be considered at the technical
feasibility stage. Hole cleaning considerations are handled separately (see below).
‡
Mykytiw, C. G. et. al., “Practical Use of a Multi-phase Flow Simulator for Underbalanced Drilling
Applications Design”, SPE 91958, presented at the IADC/SPE Drilling Conference, Houston, Texas,
October 2004.
§
Suryanarayana, P. V. et. al., “Basis of Design for Coiled Tubing Underbalanced Thru-Tubing Drilling in
the Sajaa Field”, SPE 87146, to be presented at the IADC/SPE Drilling Conference, Dallas, Texas, March
2004
‡
Maximum set-down weight just prior to lock-up is indicative of the margins available during drilling.
Maximum overpull prior to yield is indicative of margin available in pick-up.
Chapter 11
3. The first step in MPD flow modeling is to model the annular pressure profiles and
velocities. If low pressure formations are present, this may require multiphase
drilling fluids. In this case, the modeling could be performed using a multiphase
flow simulation program; otherwise a competent hydraulic simulator could also
be used. This is done to determine the MPD design anchor point for static
gradient and define the window of static / dynamic gradients against the pore /
frac gradient. The range of pressure control available and its tolerance to pore
and fracture pressure uncertainty should be evaluated.
4. Determine the MPD Operating Window. Once the flow simulations are
completed, taking into account all relevant well information, an operating window
may be developed. Only those combinations of liquid, gas and surface choke
pressures that meet the pressure requirement, minimum liquid velocity
requirement and are within motor limits, are acceptable.
5. Contingency Operations Design: It is customary in MPD to seek an operating
window that ensures overbalanced condition with minimum degree of
overbalance across the entire open section. However, this is not always feasible,
especially in mature fields. In such a case, uncertainties in either pore or fracture
pressure profiles can lead to either a lost circulation event or an underbalance
event. These should be incorporated into the design envelope. If underbalance is
not tolerable, for example, then the contingency situation will be to pump fluid
down the annulus to manage the drilling operation (“Mud Cap Contingency”). If,
on the other hand, underbalance conditions are acceptable and desirable, a UBD
operating envelope will have to be developed as a contingency, and the equipment
selection will reflect this contingency.
6. Summarize the parameters of the MPD equipment including safety factors. These
specifications may be used to tender to assure the Operator receives a “fit for
purpose” package to drill the target field. The specifications should be written in
tender language as an appendix to the Basis of Design.
7. Recommendations for specialized procedures such as connection, tripping,
running casing, and completion: It is not enough in MPD design to evaluate only
drilling conditions. Often, these special operations are more problematic than
drilling itself. Therefore, in this task, evaluation of different strategies for these
special operations, and recommend viable strategies are developed. These would
be advanced during the FEED phase.
8. Identify possible roadblocks and concerns for the implementation of the project.
9. Final Report - Basis of Design for MPD Operations. This document will be
“frozen” for document control purposes and amended by the project manager as
required and the project advances. The BOD is a key document for future
reference and the creation of the drilling program.

Chapter 11
11.4.2 Design and Planning

Otherwise known as Front End Engineering and Design (FEED), the Design and
Planning phase starts with the information contained in the BOD and advances the
project after approval. It is in this phase where the HSE framework becomes more
clearly visible as key HSE milestones appear (HazId, HazOp, and site based crew
training). A detailed FEED project plan is shown in Figure 10. Once all key deliverables
are complete, final Safety Approval is granted prior to conducting MPD operations.
In this phase, the project manager is responsible for:
Developing implementation plans for the project continuously managing the

timeline. The project manager may be considered the focal point for the MPD
project and principal technical advisor to the Operator’s engineers and
management for all technical and economic aspects of the project.
Developing detailed flow simulations to assure that a stable flow system is
possible under all likely situations within the uncertainty of the data available.
This should be performed using an industry-standard hydraulics/multiphase flow
simulation.
Eliminating roadblocks and concerns identified in the Basis of Design.
Developing tripping and well control strategy documents.
Developing concise equipment specific set of operational and emergency
procedures that will detail steps that must be taken under a given situation. These
will detail who is responsible for each step and a valve operation strategy. This is
critical for the pre-job training as well.
Preparing and facilitating the HazId/HazOp including development of action
tracking list and final report.
Conducting weekly update meetings with key project personnel to identify key
tasks, deliverables and timing to maintain a manageable critical path for project
delivery.
Optimizing tripping and completion methodology.
Developing detailed engineering of DS/CT, including string optimization,
logistics, string monitoring requirements, and safety considerations.
Evaluating well controllability.
Developing key technical and commercial performance indicators for post
evaluation review.
Coordinating with service contractor to development well specific project
drawings.
Specify data gathering and documentation requirements.
Preparing the MPD section of the proposed drilling program.
Conducting on-site training as required.
Conducting post-mortem and after action reviews to address operational issues
and apply improvements suggested from initial operations.

Chapter 11
Figure 10: Sample project plan for the FEED phase.
Drafting Operational Procedures

One key component of the FEED phase is the development of operational procedures.
Although there are generally accepted operational practices for MPD, procedures must be
specifically tailored for each project. The development of operational procedures is
foundational, working from broad strategy documents through to truncated work
instructions.
The first step is the establishment of strategy documents which highlight, at a very high
level, what approach the Operator would like to take with regards to specific critical
MPD tasks such as well control and tripping. These strategy documents define the basis
for all other developed procedures and tie together the company’s standard operating
procedures, corporate philosophy and in some cases may be governed by insurance
requirements. Some examples are:
An Operator may only be allowed to consider a lighter mud; however it must be
statically overbalanced. Such a governing philosophy affects both the overall
design as well as operating procedures.
Changing out the RCD rubber element can not be performed with pressure below
the annular due to corporate policy restrictions. Both equipment specification
(inclusion of a DDV) as well as operational procedures (continue circulating
without shutting in or pull above DDV) would be affected.
Once the strategy documents are ratified, detailed operating procedures can be developed.
Both normal operating and contingency procedures are developed. The structure of
theses procedures is very comprehensive, and usually includes reference to the P&ID,
valve numbers and valve lineups (prior to and post operations). The comprehensive
nature of these procedures requires that appropriate time be allocated to the process. It
should be noted that although the custodian of the procedures is the MPD project
manager, they are not developed in a vacuum. Input from the senior drilling engineer, rig
foreman, and MPD contractor is critical. The procedures are published in controlled
manuals, and copies kept on the rig site and the team’s central office. An example of
these operating procedures is included in Chapter 8 - Operations.

Chapter 11
Despite the most rigorous planning, operational reality rarely mimics design exactly.
This, in combination with the efficient nature of drilling, requires that procedures be
succinct and readily available. Written Work Instructions or WWIs are usually drafted
by the Company Man or MPD Supervisor for operations following the current one.
These instructions are based upon the comprehensive procedure manuals, but are edited,
truncated, and adapted to operational reality. Note that the authority of the procedure
manuals is not violated. Any major change to the procedures must follow the appropriate
Management of Change process.
11.4.3 Implementation
The Implementation phase is self explanatory and is the culmination of all planning. The
importance of a properly executed MPD program should not be downplayed, as it has
been found that projects which failed to meet expectations did so largely due to poor
attention to the drilling plan.

Chapter 11
11.5 Appendix A – IADC HSE Guidelines

MPD Course Manual

Uploaded by

Copyright:

Available Formats

MPD Course Manual

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MPD Course Manual

Uploaded by

Copyright:

Available Formats

Advanced MPD

INTRODUCTION 1-1 Rev. 1.0

INTRODUCTION 1-2 Rev. 1.0

Historically, the most apparent way to control subterranean

INTRODUCTION 1-3 Rev. 1.0

1.1 MPD Defined

BHP = ∆PGravity + ∆PFriction (1)

BHP = SP + ∆PGravity + ∆PFriction + ∆PEnergy * (2)

Where: BHP = bottom hole pressure

1.1.1 Pressure Gradients

INTRODUCTION 1-4 Rev. 1.0

Static Fluid Gradient

Dynamic Fluid Gradient

Figure 2: Conventional drilling fluid gradients & casing setting points.

INTRODUCTION 1-5 Rev. 1.0

Static Fluid Gradient

Dynamic Fluid Gradient

Statically Maximum Depth for

Figure 3: MPD fluid gradients and drilling window.

INTRODUCTION 1-6 Rev. 1.0

1.1.2 Illustrative Example of MPD

INTRODUCTION 1-7 Rev. 1.0

INTRODUCTION 1-8 Rev. 1.0

1.2 IADC UBO and MPD Sub Committee

INTRODUCTION 1-9 Rev. 1.0

Table 1: Elements of the MPD & UBO well classification system

Risk Level Application Type Fluid Type

1.2.1 Risk Classification Level

Table 2. IADC UBO Committee Risk Classification Levels

INTRODUCTION 1-10 Rev. 1.0

1.2.2 Application Type

Table 3. IADC Application Types

1.2.2.1 Level A Systems

1.2.2.2 Level B Systems

INTRODUCTION 1-11 Rev. 1.0

1.2.2.3 Level C Systems

1.2.3 Fluid Type

Table 4. IADC Fluid Types

1.3 Drivers for Considering MPD

INTRODUCTION 1-12 Rev. 1.0

INTRODUCTION 1-13 Rev. 1.0

1.4 MPD Options

INTRODUCTION 1-14 Rev. 1.0

INTRODUCTION 1-15 Rev. 1.0

GEOLOGIC PRESSURE 2-1 Rev. 1.0

GEOLOGIC PRESSURE 2-2 Rev. 1.0

2.1 Normal Pressure

2.2 Abnormal Pressure

2.2.1.1 Terzaghi’s Experiments

Where: S = Total overburden (or total stress)

GEOLOGIC PRESSURE 2-3 Rev. 1.0

Figure 1 – Terzaghi’s Experiment

2.2.2 Other Causes of Abnormal Pressure

2.2.2.2 Density Effects

GEOLOGIC PRESSURE 2-4 Rev. 1.0

BHP = 4330 – 100 = 4230 psi at 9,000 ft

EMW at 9,000 ft = 4230 psi / 9000 ft / 0.052 = 9.0 ppg

BHP = 4330 psi at 10,000 ft

Figure 2 – Density Effects Example

2.2.2.3 Fluid Migration