Professional Documents
Culture Documents
Drilling Problems N Drilling Optimization
Drilling Problems N Drilling Optimization
Drilling Objectives
Drilling Objectives
The goals of any drilling venture are safety, minimized cost and a usable completion.
Companies may have different ideas ofhow best to attain these goals, and drilling
practices may vary according to location, rig type, hole conditions or other
factors.But the goals themselves are always the same.
Safety is the primary concern in drilling an oil or gas well. Protection of personnel
supersedes all other well objectives, even when it means altering the well plan, incurring
unexpected costs or delaying operations. Failure to make safety the top priorityon a rig
can and does result in accidents, disabling injuries and deaths.
The second priority in drilling is to protect the well and the surrounding
environment. We must anticipate potential problems and include provisions in
the well plan for minimizing blowout risks or other dangers, and we must
continually monitor operations once the rig moves on location.
Environmental issues are having an ever-growing impact on all oil and gas
operations. In California and other parts of the United States, for example,
public opposition--usually based on environmental concerns--has curtailed or
even halted drillingand production activities. Those projects that are allowed
to proceed are governed by a maze of complex regulations, many ofwhich
carry severe penalties for non-compliance.
Brown (1992) provides an excellent summary of environmental issues in the
petroleum industry. Although his article focuses onU.S. regulations, his basic
premise--that environmental management must be an integral part of daily oil
and gas operations--applies anywhere. As energy and the environment
become matters of public policy throughout the world, the bestthings that
engineers or rig supervisors can do are to work closely with their companies'
environmental specialists, becomefamiliar with applicable regulations and
make sure that their operations are in compliance.
Minimized cost is the basis for most drilling optimization efforts. When looking at this
objective, we must consider the total wellcost--it does no good to save money on one
procedure if the result is a more expensive well.
Effective cost control involves establishing a balance among well objectives.
While we should never jeopardize safety in an effort to keep costs down,
neither should we spend money unnecessarily. By the same token, while we
must provide a usable well, we have to do so within realistic budget
constraints.
Whether or not an expense is justified is something the engineer must decide
for each well. Intermediate casing strings thatmight be necessary in one field,
for instance, may be wasted expenditures in another field. Concentrated well-
A usable completion should be the outcome of any drilling operation, whether the well is a
producer, an injector or simply a source of information. Even a safely drilled, low-cost well
is not entirely successful if it does not meet the needs that led to it being drilled in the first
place. As a minimum standard, a well should have:
o no irreparable damage to the hole or producing formation ;
o a sufficiently large hole diameter for running completion equipment or carrying
out other post-drilling activities.
We may look at our basic drilling goals from two perspectives: dealing with drilling
problems and optimizing normal operations.
A drilling problem is any occurrence or condition that stands in the way of well
objectives. It could involve anything from weather to transportation delays to
blowouts. In this discussion, we concentrate on problems that occur as part of the
drilling process itself.
A comprehensive, thoroughly researched well plan is our best defense against hole
problems. Every aspect of the plan, from pre-spud activities to completion
procedures, should be designed to anticipate and control these problems. An
effective wellplan requires thorough research and data collection, and a careful
analysis of these data. Data sources may include:
bit records
mud records
mud logs
IADC drilling reports ("tour sheets")
scout tickets
log headers
production histories
seismic studies
well surveys
geologic contours
databases or service company files
The engineer should use all available sources to identify and plan for potential hole
problems. Each source may containimportant information not available elsewhere.
The preparation and effort that goes into well planning should carry over into daily
operations. Rig personnel must continuallymonitor drilling performance for early
indications of problems and make sure that the appropriate problem-solving
tools,materials and technical expertise are either at the rig or available on short
notice.
Drilling optimization involves using available resources to minimize overall cost,
subject to safety and well completionrequirements. Part of this, of course, entails
preventing or successfully solving hole problems. But optimization efforts
alsoencompass "normal" operations. The key measure of performance in this area is
the cost to drill a given interval.
In field units,
(2)
where P = pressure, psi
MW = mud weight or density, lbm/gal (U.S.)
TVD = true vertical depth, ft
When expressed in SI units (with pressure given in kPa), 0.052 becomes 0.00980.
This hydrostatic pressure, which we control by adjusting the mud weight, is our
primary means of counteracting formation, or pore pressure.
Equation 2 is adequate for determining the pressure of a static, homogeneous mud
column. In many instances, however, the mud is not static; circulation and pipe
movement can significantly affect wellbore pressure. Nor is the mud column
necessarily homogeneous; during cementing operations or mud changeovers, for
example, the well contains fluids of different densities. To determine the actual
wellbore pressure, we have to account for these variables. We can simplify our task
by defining equivalent mud weight (EMW) or equivalent circulating density(ECD):
(3)
(4)
where EMW = equivalent mud weight, lbm/gal (U.S.)
ECD = equivalent circulating density, lbm/gal (U.S.)
Pwell = sum of hydrostatic and applied pressures, psi
Pann = the frictional pressure loss in the annulus. This corresponds to the
pump pressure minus the pressure losses through the surface equipment, drill
string and drill bit.
TVD = True vertical depth, ft
When expressed in SI metric units (with pressure given in kPa), 0.052 becomes 0.00980.
Example 1: Equivalent Circulating Density
Determine the Equivalent Circulating Density and the effective wellbore pressure for
the following conditions:
Mud weight = 12.5 lbm/gal
TVD =11,300 ft
Calculated annular pressure loss = 150 psi
Solution:
Pore pressure is normally equal to the formation depth times the pressure gradient of
the native formation water:
(5)
where = formation fluid density, lbm/gal (U.S.)
= fluid pressure gradient, psi/ft
TVD = True Vertical Depth, ft
When expressed in SI units (with pressure given in kPa), 0.052 becomes 0.00980.
The fluid pressure gradient g typically ranges from 0.433 psi/ft [9.80 kPa/m] for
fresh water to 0.465 psi/ft [10.54 kPa/m] or higher for brine.
Example 2: Determining pore pressure
Determine the pore pressure of the following formation.
Normal pressure gradient = 0.46 psi/ft
Measured depth = 12,500 ft
TVD=10,000 ft
Solution:
In many instances, we may encounter abnormal pore pressures, which deviate from
established area trends.
Differential pressure (P) is simply the difference between wellbore an dformation
pressure:
(6)
When the wellbore and formation pressures are approximately equal, P is zero. We
refer to this condition as balanced ( Figure 1 ).
Figure 1
This pressure of 4,600 psi is pushing from the formation to the well. To keep
formation fluid from entering the wellbore, the mud column must exert at least 4,600
psi in the other direction -- from the well to the formation. Our required mud weight
is therefore:
Balanced drilling with an unweighted mud system is a viable option in areas or depth
intervals having well-established pressure trends and little blowout risk. In lowpermeability, high-strength formations, operators may even drill underbalanced
(P<0) using air, gas, foam or low-density liquids ( Figure 2 , underbalanced
conditions; wellbore pressure is less than pore pressure).
Figure 2
Because of favorable rock mechanics effects and the fact that there is no chip holddown caused by high differential pressure, balanced or underbalanced drilling can
result in penetration rates much higher than those we could attain with a weighted
mud system. Figure 3 illustrates how penetration rates tend to decrease with
increasing differential pressure.
Figure 3
Figure 4
That is, we have to drill overbalanced (P>0) to protect against sudden fluid level
drops, unintentional mud weight reductions or other circumstances that might
decrease wellbore pressure.
In addition, we need a safeguard against abnormal formation pressures. In many
areas (e.g., south Louisiana, south Texas, the North Sea, parts of the Middle East), it
is not uncommon to see P values on the order of several thousand psi.
Example 4: Differential pressure
Determine the differential wellbore pressure at 15,400 ft TVD on a U.S. Gulf Coast
well, where the mud weight is 14.0 lbm/gal and the formation pressure is 7,600 psi.
Solution:
Disadvantages of overbalanced drilling are lower penetration rates and higher mud
costs, along with increased potential for formation damage and stuck pipe. Up to a
point, this is a fair trade-off, since weighted muds allow us to better control
formation pressure and maintain hole integrity. We need to keep in mind, however,
that our drilling objectives also include optimizing penetration rates and reducing
expenses. We should not compromise these objectives unnecessarily by using
heavier muds than safety requires.
Formation fracture gradients define our upper wellbore pressure limit. Exceeding this
limit can cause lost circulation, resulting in formation damage and induced fractures.
Severe fracturing can cause the wellbore fluid level to drop, thereby creating a
blowout risk.
Fracture pressure is related to overburden stress, which is equal to the rock matrix
pressure plus the pore pressure:
(7)
where S = overburden stress
= matrix stress at depth of interest
Ppore = pore pressure at depth of interest
Over the years, a number of researchers have developed theoretical methods for
determining fracture gradients. Matthews and Kelly (1967), building on the landmark
work of Hubbert and Willis (1957), developed the following relationship for
sedimentary rocks:
(8)
where frac = fracture gradient, psi/ft [kPa/m]
D = depth, ft [m]
Ki = matrix stress coefficient for the depth at which the value of s would be
the normal matrix stress (dimensionless)
Matthews and Kelly based this equation on the assumption that the cohesiveness of a rock matrix
is generally related to matrix stress, and thus varies only with the degree of compaction. Their
procedure for determining the fracture gradient is as follows:
1. Determine the pore pressure (based on seismic measurements, well log anlaysis or
drilling data correlations).
2. Assume a gradient of 1.0 psi/ft [22.62 kPa/m] for the overburden, and
calculate the matrix stress s.
3. Determine the depth (Di) at which s would have a normal value, assuming
that S=1.0 psi/ft [22.62 kPa/m] and the normal stress gradient equals 0.535
psi/ft [12.10 kPa/m]. Therefore,
(9)
When depth is expressed in meters and s is expressed in kPa, 0.535 becomes 12.10.
Figure 5
Gulf Coast Sands), determine Ki and calculate frac from Equation 8 (note that
Figure 5 is empirically generated from field data).
Example 5: Determining fracture gradient
Using the Matthews and Kelly procedure, determine the fracture gradient just below
the casing seat for the following Louisiana Gulf Coast well.
Casing set at 6,650 ft TVD
Formation pressure = 3,300 psi
Solution:
In the field, we can estimate the minimum fracture gradient at each new casing point
by performing a leak-off test as follows:
1. Close the blowout preventer and apply pressure down the drill pipe in small
increments, using a low-volume pump.
2. Continue pumping small volumes of mud until the formation begins to take
fluid, or until the pressure reaches a pre-set test limit.
3. Plot pressure versus the pumped volume to determine the initial fracture
pressure.
Example 6: Leak-off test-- calculating fracture gradient (after Adams, 1985):
Determine the fracture gradient at a well's casing point (6,750 ft. TVD), given the
information in Table 1., below (mud weight = 12.5 lbm/gal).
Solution:
Figure 6 (Results of leak-off test from Example 6) indicates that the formation begins
Figure 6
The fracture pressure is equal to this applied pressure plus the hydrostatic wellbore
pressure:
Pressure, psi
0
4
100
190
280
370
460
550
640
730
820
850
880
(2)
where MW is the mud weight and steel is the density of steel, which equals 65.5
lbm/gal [7860 kg/m3]. Table 2.2 lists buoyancy factors for mud weights ranging from
8.33 to 16.0 pps.
Example 1: Buoyancy effects
Determine the maximum weight-on-bit provided by 450 ft of 7 3/4-inch o.d., 144
lb/ft drill collars for both of the following mud weights. (Assume that all of the drill
string compression is in the drill collars.)
9.5 lbm/gal
16.0 lbm/gal
Solution:
MW,
[kg/m3]
Buoyancy
lbm/gal
Factor
8.33
[1000]
0.873
9.0
[1078]
0.862
9.5
[1138]
0.855
10.0
[1198]
0.847
10.5
[1258]
0.840
11.0
[1318]
0.832
11.5
[1378]
0.824
12.0
[1438]
0.817
12.5
[1498]
0.809
13.0
[1558]
0.801
13.5
[1618]
0.794
14.0
[1678]
0.786
14.5
[1738]
0.778
15.0
[1798]
0.771
15.5
[1858]
0.763
16.0
[1918]
0.756
Table 1. Buoyancy Factors for Various Mud Weights (Based on Steel Density of 65.5
lbm/gal [7860 kg/m3])
Rheological Models
Rheology, a term often used in drilling engineering, is the study of the flow or
deformation of matter. We generally describe flow or deformation in terms of shear
stress and shear rate.
We may best understand these concepts by considering a fluid located between two
parallel plates separated by a distance d ( Figure 1 ).
Figure 1
If we apply a force to the upper plate while holding the lower plate stationary, the
upper plate attains a constant velocity (v), which depends on the applied force (F),
the distance between the plates, the surface area (A) and the fluid viscosity ():
(1)
Shear stress () is equal to the force per unit area,
(2)
while shear rate () is equal to the fluid velocity divided by the distance between the
plates:
(3)
These parameters define the flow relationships used to characterize drilling mud
behavior: the Newtonian, Bingham Plastic and Power Law models. (A fourth
relationship, the Three Parameter model, combines elements of the Bingham Plastic
and Power Law models.)
Newtonian fluids exhibit direct proportionality between shear stress and shear rate
( Figure 2 ):
Figure 2
(4)
The proportionality constant () is the fluid viscosity. Although this model has often
been used to describe drilling fluid behavior, it is an oversimplification; most drilling
fluids do not exhibit this straight-line relationship.
A common method for describing the shear stress/shear rate relationship is the
Bingham-Plastic model ( Figure 3 ):
Figure 3
(5)
where
p = plastic viscosity (PV), cp
y = yield stress or yield point (YP), lbf /100 ft 2 (1.0 cp = 0.001 Pa-s; 1 lbf /100 ft 2
= 0.4788 Pa)
Note that a certain initial pressure is required to initiate fluid movement, and that the
mud's circulating pressure depends on the initial yield point and the plastic viscosity.
We can determine these quantities using a rotational viscometer and the following
equations:
(6)
(7)
where 600and 300 = the viscometer readings at rates of 600 and 300 RPM,
respectively, and PV is given in cp.
The Bingham Plastic model thus depends on a linear relationship (the difference
between the two viscometer readings) to describe drilling fluid flow. Although widely
used, its application to turbulent flow conditions, where fluid behavior deviates from
linear relationships, is limited.
The Power Law model is a non-linear relationship, which is more descriptive of real
drilling fluid behavior ( Figure 4 ):
Figure 4
(8)
where
(9)
n = flow behavior index = 3.32 x log( 600/300)
(10)
While it is more realistic than the Bingham Plastic model, the Power Law model is
also more complex.
Flow Regimes
The flow regimes most commonly encountered in drilling are laminar, turbulent and
transitional.
In laminar flow, fluid behaves as a series of parallel layers moving at uniform or nearuniform velocities. There is no large-scale movement of fluid particles between layers.
The fluid layers nearest the center of the pipe or annulus generally move faster than the
layers adjacent to the hole wall ( Figure 1 , Laminar flow profile in drill pipe/hole annulus).
Figure 1
Figure 2
The most common method for determining a fluid's flow regime is to calculate its
Reynolds number. In its simplest form (i.e., for the case of a Newtonian, non-elastic
liquid inside pipe) we may define the Reynolds number as
(1)
where = fluid density, kg/m3
(2)
= inside diameter of pipe, m
= fluid viscosity, Pa-s
For annular flow, the Reynolds number becomes
(3)
where d2 = hole or casing diameter, m
d1 = outside diameter of pipe, m
(4)
The term 0.816(d2-d1) in Equation 3 is the equivalent circular diameter of a slot
representation of the annulus (Bourgoyne et al, 1986).
In field units (where is given in lbm/gal, v in ft/s, q in gal/min, d, d1 and d2 in
inches and in cp), Equations 1 through 4 become:
(Pipe flow):
(5)
(6)
(Annular flow):
(7)
(8)
As a general guideline, Reynolds numbers of less than 2,100 indicate laminar flow,
while Reynolds numbers greater than 4,000 indicate turbulent flow. Between these
values, flow is considered transitional.
In field applications, we identify flow regimes by determining the critical Reynolds
number or critical velocity (vc) at which flow changes from laminar to turbulent. For
example, we may calculate vc and compare it to the actual fluid velocity (v). If vc <
v, flow is laminar, while if vc >v, it is turbulent. Because the transition between
laminar and turbulent flow may not be clear-cut, it often becomes necessary to
calculate a range of critical velocities to determine the flow regime.
Example 1: Flow regimes
A 10.5 lbm/gal drilling mud with a viscosity of 30 cp is circulating at 250 gallons per
minute in an 8 3/4 inch diameter wellbore. Determine the flow regime inside 4.5 inch
o.d., 16.60 lb/ft drill pipe (3.826 inch inside diameter), and in the drill pipe/hole
annulus. For this example, assume that the mud behaves as a Newtonian fluid.
Solution:
Drill pipe:
(Pipe flow)
(9)
(Annular flow)
(10)
(Pipe flow)
(11)
(Annular flow)
(12)
Using these equations, the criterion for turbulence is the same as for Newtonian
fluids, with laminar flow occurring below a Reynolds number of 2100.
Power Law Fluids
For power law fluids, we may use correlations developed by Dodge and Metzner
(1959):
(Pipe flow):
(13)
(Annular flow):
(14)
(13 SI)
(14 SI)
The turbulence criterion for power law fluids is based on a critical Reynolds number
( Nrec ), which depends on the value of the flow behavior index. A simple equation
for estimating the critical Reynolds number at the upper limit of laminar flow is
(15)
For the region between transitional and turbulent flow, the critical Reynolds number
is
(16)
Note: In the field, determine both the critical and the actual flow velocities. If v c<v the flow is
turbulent, while if vc v, it is laminar. If the critical and actual velocities are approximately equal,
then perform pressure loss calculations for both flow regimes, and use the results that give the
larger pressure loss.
We use this information to maximize the pressure loss through the bit and thereby
maximize the hydraulic energy at the bottom of the hole. Table 1 and Table 2
summarize the equations for determining pipe and annular pressure losses based on
Bingham Plastic flow, while Table 3 and Table 4 list the pressure loss equations for
Power Law fluids.
Fluid velocity (v)
Oilfield Units:
SI Units:
Critical velocity (vc)
Oilfield Units:
SI Units
Pressure Loss for
laminar flow (Pds )
(Use for v < vc)
Oilfield Units:
SI Units:
Pressure Loss for turbulent flow
(Use for v vc)
Oilfield Units:
SI Units:
Nomenclature (SI units in brackets):
Oilfield Units:
SI Units:
Oilfield Units:
SI Units
Pressure Loss for laminar flow (Pds )
(Use for v < vc)
Oilfield Units:
SI Units:
Pressure Loss for turbulent flow
(Use for v vc)
Oilfield Units:
SI Units:
Nomenclature:
Table 2 Annular pressure loss equations for a Bingham Plastic fluid [After Bourgoyne
et al (1986) and Adams (1985)]
Fluid velocity (v)
Oilfield Units:
SI Units:
Oilfield Units:
SI Units:
Pressure Loss for laminar flow (Pds )
(Use for v < vc)
Oilfield Units:
SI Units:
Pressure Loss for turbulent flow
(Use for v vc)
Oilfield Units:
SI Units:
Nomenclature:
K = consistency index
(dimensionless)
Oilfield Units:
SI Units:
Critical velocity (vc)
Oilfield Units:
SI Units:
Pressure Loss for laminar flow (Pds )
(Use for v < vc)
Oilfield Units:
SI Units:
Pressure Loss for turbulent flow
(Use for v vc)
Oilfield Units:
SI Units:
Nomenclature:
K = consistency index
(dimensionless)
Table 4 Annular pressure loss equations for a Power Law fluid [After Bourgoyne et al
(1986) and Adams (1985)]
We can find the pressure loss through surface equipment (DPsurf) by treating it as
an equivalent length of drill pipe. To determine this equivalent length, we simply
match our surface equipment specifications to one of the four groups shown in Table
2.7. For example, if a rig has Group 4 surface equipment and uses 5-inch, 19.5 lb/ft
drill pipe, we would, for calculation purposes, add 579 feet to the actual drill pipe
length.
Component
Typical Combinations
Case 1
Case 2
Case 3
Case 4
i.d.,in
L,ft
i.d.,in
L,ft
i.d.,in
L,ft
i.d.,in
L,ft
[cm]
[m]
[cm]
[m]
[cm]
[m]
[cm]
[m]
40
3.5
40
45
45
[7.62]
[12.192]
[8.89]
[12.192]
[10.16]
[13.716]
[10.16]
[13.716]
Drilling
45
2.5
55
55
55
hose
[5.08]
[13.716]
[6.35]
[16.764]
[7.62]
[16.764]
[7.62]
[16.764]
Standpipe
Swivel,
2.5
2.5
washpipe,
[5.08]
[1.219]
[6.35]
[1.524]
[6.35]
[1.524]
[7.62]
[1.829]
2.25
40
3.25
40
3.25
40
40
[5.715]
[12.192]
[8.255]
[12.192]
[8.255]
[12.192]
[10.16]
40
gooseneck
Kelly
Drill pipe: Length of surface connections, expressed as equivalent ft [m] of drill pipe
437 [133.2]
161 [49.1]
761 [232.0]
479 [146.0]
340 [103.6]
816 [248.7]
579 [176.5]
Drill bit pressure losses do not result primarily from friction forces, rather, they are
due to the acceleration of the drilling fluid through the bit nozzles. We may express
the bit pressure drop in psi as
(2)
where
q = circulation rate, gal/min (U.S.)
MW = mud weight, lbm/gal
Cd = nozzle discharge coefficient (dimensionless)
AT = total nozzle area, in2
In the SI system, where Pbit is expressed in kPa, 12,031 becomes 2000.
Although the nozzle discharge coefficient Cd varies according to nozzle type and size,
a value of 0.95 is acceptable for most field calculations.
The following example illustrates how we can apply pressure loss relationships.
Example 1
Determine the bit pressure drop for the following well:
Total Depth=9,950 ft
Casing: 9 5/8 inch, 43.50 lb/ft (8.755 in. i.d) cemented at 6,500 ft.
Open hole: 8 1/2 inch from 6,500 ft to 9,950 ft (T.D.)
Drill string:
Drill pipe: 9500 ft. of 4 1/2 inch, 16.60 lb/ft (3.826 in. i.d.)
Bottomhole assembly: 450 ft. of 6 3/4 inch o.d. x 2 1/4 inch i.d. drill collars and tools
Mud properties:
Mud weight (MW)=10.5 lbm/gal
Plastic viscosity (PV)=35 cp
Yield point (YP)=6 lbf/100 ft2
Assume Bingham Plastic fluid
Circulating rate=300 gal/min
Pump pressure=2,200 psi
Surface Equipment:
Standpipe: 45 ft x 4 in. i.d.
Hose: 55 ft x 3 in. i.d.
Swivel, washpipe, gooseneck: 5 ft x 2.5 in. i.d.
Kelly: 40 ft x 3.25 in. i.d.
Solution:
1. Note first that the surface components correspond to a Case 3 equipment combination
(Table 2.7). The pressure loss through these components is equivalent to 479 ft. of 4 1/2
in., 16.6 lb/ft drill pipe. We can therefore combine Psurf with our calculation of Pds.
2. Determine the pressure losses inside the drill string (Pds). This involves
separate calculations for the drill pipe and the bottomhole assembly:
a) Drill pipe:
= 605 psi
(Note that we accounted for surface pressure losses by adding in the drill pipe
equivalent length of 479 ft)
b) Bottomhole assembly:
= 340 psi
Pds=Pdp+Pbha=605+340=945 psi
3. Determine the pressure losses in the annulus. (DPann). This involves separate
calculations for the cased-hole section (surface to 6,500 ft), the drill pipe/hole
annulus (6,500 - 9,500 ft) and the BHA/hole annulus (9,500 - 9,050 ft).
a) Cased-hole annulus (surface-6500 ft):
= 73 psi
b) Drill pipe/hole annulus (6,500 ft - 9,500 ft):
Assume laminar flow (critical velocity will be close to that calculated for the casedhole annulus)
= 38 psi
c) BHA/open hole annulus:
= 31 psi
Total annular pressure losses: DPann=73 + 38 + 31 = 142 psi
4. Determine pressure loss at bit:
Pbit=Ppump-(Psurf+Pds)-Pann=2200 - 945 - 142= 1,113 psi
Figure 1
The resulting piston effect generates a surge pressure that is added to the
hydrostatic pressure. Excessive surge pressures can increase the wellbore pressure
to such a degree as to induce lost circulation.
Conversely, when we pull pipe out of a well, mud flows down the annulus to fill the
resulting void. This causes a suction effect, generating a swab pressure that can
lower the differential pressure and possibly bring formation fluid into the wellbore.
Figure 2 (Typical pressure surge pattern for a joint of casing lowered into a wellbore)
shows the pressure fluctuations that resulted from lowering casing into a mud-filled
wellbore, and illustrates the significance of surge and swab effects.
Figure 2
Figure 3
To apply the clinging constant, we need to know the mud velocity in the annulus. For
a closed drill string, this is equal to
(1)
where vpipe = pipe velocity
d2 = hole diameter
d1 = pipe outside diameter
For open-ended pipe,
(2)
= 2.07 ft/s
d1/d2= 10.75/14.75=0.73; from Figure 3 , K=0.44 (assume laminar flow)
Psurge = 67 psi
Hole Cleaning
From the time that rotary drilling came into its own with the dramatic Spindletop
discovery of 1901, it was destined to ultimately replace the once dominant cable tool
methods. Rotary drillers found out early on that they had a major advantage over
cable tool operators: the ability to continuously remove drilled cuttings (as opposed
to periodically removing them with a bailer).
Cuttings removal and transport are still primary functions of the drilling mud.
Additionally, the mud must be able to hold these cuttings in suspension when
circulation stops. The mud's effectiveness at removing cuttings significantly affects
drilling efficiency.
Drilled cuttings vary in size and density according to formation lithology, differential
pressure, the cutting action of the bit and other factors. They are usually heavier
than the drilling mud, and therefore tend to slip down through the annulus, back
toward the bottom of the hole. A mud's ability to transport cuttings -- that is, its
carrying capacity -- is related to the difference between the annular velocity and the
slip velocity with which the cuttings fall ( Figure 1 , Carrying capacity depends on a
mud's annular velocity and the slip velocity with which cuttings fall back to bottom).
Figure 1
Moore (1974), Chien (1971) and Walker and Mayes (1975) are among those who
have developed correlations for estimating slip velocity (this section presents Moore's
correlation by way of example). To use these methods, we need to determine the
average densities and diameters of drilled solids by visually inspecting representative
cuttings, or by doing a sieve or screen analysis.
We can use the following relationship to estimate the slip velocity of a particle
suspended in a Newtonian fluid:
(1)
where vs = slip velocity, ft/s
dp = particle diameter, inches
p = particle density, lbm/gal
(2)
where = fluid viscosity, cp
In SI units (kg/m3, m/s,m, Pa-s), 928 reduces to 1.
We can then obtain Cdrag from Figure 2 (Particle drag coefficient as a function of the
particle Reynolds number).
Figure 2
(3)
In SI units, 82.87 becomes 0.3267.
For (NRep) >2,000 (turbulent flow), Cdrag has a constant value of about 1.5, and
Equation 1 reduces to
(4)
In SI units, 1.54 becomes 2.945.
For intermediate particle Reynolds number values (i.e., 10< (NRep)
<100) Cdrag =
, and
(5)
In SI units, 2.90 becomes 0.7053.
The term a in these equations is the apparent viscosity, given in units of centipoise
and defined as follows:
(6)
(7)
(8)
(6 SI)
Moore recommends Equation 5 for use in routine field applications. Note that this
method for determining slip velocity may require several iterations.
Example 1: Slip velocity (Moore's method)
Calculate the slip velocity of drilled cuttings having an average diameter of 0.25
inches and a density of 2.65 g/cc (22.1 lbm/gal), given the following mud and hole
data:
Hole size=12.25 inches
Drill pipe=5 inches
MW=10.2 lbm/gal
Annular Velocity=60 ft/min (1.0 ft/s)
n (from viscometer readings)=0.8
K=150 cp equivalent
Solution:
1. Determine apparent viscosity (a):
= 88.1 cp
2. Assume an (NRep) range and then check the results obtained:
Let (NRe p) <1:
=0.7 ft/s
Verification:
vs = 0.39 ft/s
Verification:
Exercises
1. What mud weight is required to balance formation pressure and provide a
400 psi trip margin at a depth of 12,300 ft (TVD) and a fluid pressure
gradient of 0.45 psi/ft?
solution
2. Calculate the maximum bit weight that would be available using 750 ft of
Soln
3. Calculate the plastic viscosity (PV), yield point (YP), flow behavior index (n) and
the consistency index (K), given viscometer readings of 34 at 300 RPM and 62 at 600
RPM.
Soln
4. Using the Bingham-Plastic flow model, determine the critical velocity and flow regime for a
12.25 x 7.75 inch open hole/drill collar annulus:
Mud properties:
MW=10.2 lbm/gal
PV=32 cp
YP=8 lbf/100ft2
Circulation rate: Q=300 gallons/minute
Soln
Using the Bingham-Plastic flow model, determine the critical velocity and flow regime
for a 12.25 x 7.75 inch open hole/drill collar annulus:
5. Using the Bingham Plastic model, calculate the frictional pressure loss in a 10,000 foot
section of 5-inch, 19.50 lb/ft (i.d.=4.276 inches) for the following circulating conditions:
Q=350 gallons/minute
MW=13.5 lbm/gal
PV=26 cp
YP=9 lbf /100ft2
Soln.
Soln..
Equivalent velocity
ve = vmud -Kvpipe
Surge pressures:
Drill pipe:
Drill Collars:
chemical diagenesis
tectonic activity
presence of salt structures
differences in fluid density
fluid migration between zones
presence of artesian systems
Sedimentary layers form through rock particle deposition. As each new layer forms,
it puts more weight on underlying layers and forces them downward. The underlying
layers are thus subjected to successively increasing overburden pressure.
If an underlying layer is normally pressured, increased overburden causes the rock
matrix to compact, thereby decreasing the pore space and forcing fluid into adjacent
permeable formations. This compaction increases the rock density and matrix
strength, enabling the rock to support the added pressure. As long as the fluid
escapes fast enough to compensate for the increased pressure, there is a balance
between the overburden and the rock matrix stress.
Compaction can only take place, however, if the pore fluid has somewhere to go. If a
low permeability barrier (e.g., a shale layer) slows or prevents fluid from escaping,
then the pore volume cannot decrease quickly enough to provide adequate rock
matrix strength. Since the rock matrix cannot support the higher overburden
pressure, the pressure increase is transmitted to the trapped formation fluid,
resulting in an overpressured condition ( Figure 1 , Abnormal pore pressure due to
undercompaction of sediments. The lower sandstone formation is surrounded by lowpermeability shales, which prevent fluid from escaping).
Figure 1
Figure 2
Figure 3
Because free water occupies more space than bound water, it becomes
overpressured if held in by a low-permeability seal.
Figure 4
Figure 5
In this case, faulting and uplift have moved a formerly buried formation from an area
of high overburden stress to one of lower overburden stress. The fault has placed
this permeable formation against an impermeable shale, effectively forming a trap.
The result is an abnormal pressure condition.
Salt beds are a leading cause of abnormal pressures. They not only form
impermeable barriers, but they are plastic and have very little matrix strength.
Instead of supporting overburden stress, a salt bed transmits the entire load to
underlying formations.
Salt domes, because they are plastic and usually less dense than surrounding rocks,
can literally flow upward into shale sections, overcompacting shallow formations and
creating unusually high pressures.
Abnormal pressures can result from fluid density differences in zones having
connected permeability. Figure 6 (Abnormal pressure due to fluid density difference)
illustrates a situation where the pressure in a gas-producing interval having a
significant dip is higher than the native salt water gradient.
Figure 6
Under certain circumstances, fluid migrating from a relatively deep formation can
charge a shallower, lower-pressured formation. When this occurs, the fluid that may
have been normally pressured at its original depth will be overpressured at the
shallower depth. These circumstances could include leakage across a fault, casing
leaks, channeling through cement or underground blowouts.
An Artesian water system may result in abnormal pressures by transmitting
hydrostatic pressure from a high elevation to a lower elevation through a continuous,
water-bearing zone ( Figure 7 , Abnormal pressure in an Artesian water system). This
type of overpressuring cannot be predicted using conventional techniques.
Figure 7
Figure 1
It follows that in an overpressured formation, where porosities are abnormally high, we should
see interval transit times that are above normal ( Figure 2 , Interval transit time plot indicating
abnormal pressure trend, based on seismic data from south Texas Frio trend).
Figure 2
While this is a qualitative indication of abnormal pressure, we can arrive at quantitiative presure
estimates by analyzing seismic data in more detail.
We can also obtain interval transit times from sonic logs. A major benefit of seismic
techniques, however, is that they provide "pre-spud" data, which can be used during
the early stages of well planning.
As with seismic interpretation, the objective of log analysis is to detect the high
porosities characteristic of abnormally pressured formations. We can do this by
obtaining electrical (resistivity/conductivity), sonic, density and other measurements
from open-hole or cased-hole tools. With the exception of measurement-whiledrilling (MWD) data, well logs are used for "post-drilling" evaluations.
Because water has lower electrical resistivity (i.e., higher conductivity) than rock
matrix particles, an increase in the fluid-filled pore volume due to overpressuring
should result in a resistivity decrease. Figure 3 (Generalized shale resistivity plot)
illustrates this deviation from the normal, linear shale resistivity trend.
Figure 3
(Note that since the abnormal pressures originally developed in shale sections, only
resistivities of clean shales are used in constructing resistivity plots. )
To estimate formation pressure from resistivity measurements, Hottman and Johnson
(1965) recommend determining the ratio between observed and normal rock
resistivities. using an empirical correlation ( Figure 4 , Empirical correlation of
reservoir fluid pressure gradients versus ratio of noremal to observed shale
resistivities) and the following three-step procedure:
1.
Figure 4
Plot the shale resistivity versus depth on semi-log paper to establish a normal resistivity
trend.
2. Note the depth at which the plotted points deviate from the established
trend. This indicates the top of the overpressured zone.
3. Determine the pressure gradient at the depth of interest.
a. Extrapolate the trend line to determine the "normal" resistivity at that depth.
b. Calculate the ratio of the observed resistivity to the extrapolated
resistivity.
c. Use Figure 4 to determine the formation pressure.
Example 1: Abnormal pressure prediction using resistivity data (from Adams, 1985):
Given the resistivity data in Table 1, below, determine:
(a) the top of the abnormal pressure zone
(b) the formation pressure at a depth of 11,600 ft
Resistivity, -m
Depth, ft
Resistivity, -m
Depth, ft
0.54
4,000
0.80
10,400
0.64
4,600
0.76
10,700
0.60
5,600
0.58
10,900
0.70
6,000
0.45
11,000
0.76
6400
0.36
11,100
0.60
7,000
0.30
11,300
0.70
7,500
0.28
11,600
0.74
8,000
0.29
11,900
0.76
8,500
0.27
12,300
0.82
9,000
0.28
12,500
0.90
9,700
0.29
12,700
0.84
10,100
0.30
12,900
Solution:
These data are plotted in Figure 5 .
Figure 5
The deviation from the normal pressure trend occurs at about 9,700 ft. Extrapolated
"normal" resistivity at 11,600 ft. is 1.23 -m, while the observed resistivity is 0.28
-m. The ratio of these resistivities is (1.23/0.28) = 4.39. Hottman and Johnson's
correlation ( Figure 4 ) gives a formation pressure corresponding to about 17.1 ppg
EMW, or 10,314 psi.
Overlay plots, consisting of parallel lines that represent formation pressure in terms
of mud weight, are available for resistivity or conductivity curves, as well as for sonic
logs. ( Figure 6 , Shale resistivity overlay).
Figure 6
-- trip gas the increased gas circulated out of the hole after tripping
pipe.
Increased gas concentrations have commonly been interpreted as signs of abnormal
pressure. There are, however, a number of other factors contributing to these increases,
which compromise their reliability as pressure indicators. (Adams, 1980).
Mud resistivity and chloride content Changes in these parameters may serve
as secondary indicators of increasing porosity and, therefore, abnormal
pressure. For mud resistivity to serve as an indicator, the mud salinity must
be measurably different from that of the formation fluid.
Shale density Normally, shale density should increase with increasing depth.
By measuring the bulk density of the drilled cuttings, the mud logger can
detect deviations from this trend.
These tools, once at the frontiers of technology, have become integral to many drilling operations.
They offer the advantage of real time (i.e., instantaneously transmitted) data, including resistivity
and formation conductivity, lithology (from gamma ray measurements), annular temperature, and
other abnormal pressure indicators.
Defining the relationship between penetration rate and various rig parameters (e.g.,
weight-on-bit, rotary speed, mud properties) is a basic goal of drilling research. The
models and correlations developed for this purpose are mainly empirical, and range
from "rule-of-thumb" field calculations to sophisticated computer simulations.
Perhaps the most well-known drilling model used in the industry is the dc exponent,
developed from Bingham's drilling rate equation (Bingham, 1965):
(1)
where
Jordan and Shirley (1966) simplified and modified this relationship, assigning a value
of "1" to the formation drillability constant and replacing Bingham's bit weight
constant with a d-exponent:
(2)
Assuming constant mud weights and normally-pressured formations, a plot of d
versus depth typically shows increasing d-exponent values with increasing depth . In
abnormally pressured formations, d should increase at a slower rate, or even begin
to decrease, with depth.
Rehm and McLendon (1971) took into account the effect of increasing mud weight by
modifying the d-exponent:
(3)
where EMWn = Equivalent mud weight of normally pressured formation.
We may plot dc exponents as a function of depth on semi-log paper, and then use
overlays to determine formation pressures ( Figure 7 , Typical dc-exponent plot).
Figure 7