High-Pressure, High-Temperature Well Construction

Drilling and completing wells in HPHT environments is difficult and dangerous. But as

activity in these areas expands, transfering experience, expertise and knowledge

about the best techniques from person to person and well to well is essential to

reduce risks, increase safety and efficiency, and ultimately, improve financial returns.

Keelan Adamson
George Birch
Erhu Gao
Steve Hand
Colin Macdonald
David Mack
Anver Quadri
Aberdeen, Scotland, UK

For help in preparation of this article, thanks to John

Oliver, Dowell, Sugar Land, Texas, USA; and Demos
Pafitis, Anadrill, Sugar Land, Texas.
CemCADE, DensCRETE, FMP (Drilling fluids monitoring
package), MudCADE, MUDPUSH, PRISM (precision
recording for job supervision and monitoring) and
ULTIDRILL are marks of Schlumberger.

36 Oilfield Review

The principles of well construction in high-
pressure, high-temperature (HPHT) wells are
not significantly different from those used in
less demanding wells, but challenges remain
because of conditions that limit the range of
suitable materials and affect equipment
performance. The margins for error are
small and the potential consequences of fail-
ure are great.
Despite the challenges, interest in these
wells has remained high and the number of
HPHT wells has grown steadily. Reservoir
pressures in excess of 10,000 psi [68.9 MPa]
have been exploited in many parts of the
world, particularly in the search for gas.
High-temperature wells have been success-
fully drilled into reservoirs where tempera-
tures exceed 300°F [149°C] in Qatar, Ras al
Khaimah, Sudan and elsewhere. In China
exploitation of 500°F [260°C] reservoirs is
being planned for 1998.
Even more challenging conditions exist
where high pressures and temperatures are
present together as in Angola, the USA,
Yemen and the North Sea. In these regions it
is not uncommon for well temperatures
exceeding 350°F [177°C] to coexist with
pressure gradients requiring mud weights in
excess of 16 ppg [1.9 g/cm3]. The most
extreme wells in the North Sea (Ranger well
29/5B-4), Yemen (Shell Abbass 1 well)
and the USA (Sohio M.E. Coward well)
have temperatures above 400°F [204°C]
drilled with muds of up to 18.5 ppg
[2.22 g/cm3] (above).
Even with growing experience, many
aspects of drilling and completing HPHT
wells continue to demand special attention.
For example, secondary well control relies on
Summer 1998
surface equipment being able to function reli-
ably under extreme conditions. Blowout pre-
venter (BOP) elastomers and flexible hoses
must be rated to withstand the temperatures
and pressures for long enough to evacuate a
rig during the worst-case scenario (often con-
sidered to be complete expulsion of drilling
fluid from the well after loss of well control).
Evaluating HPHT wells requires special
logging and testing tools, with downhole
mechanical and electrical equipment
capable of withstanding harsh conditions of
elevated temperature and pressure, high-
temperature explosives for perforating, and
procedures for their successful operation.
These aspects of exploiting HPHT wells are
addressed elsewhere in this issue (see
“High-Pressure, High-Temperature Well
Logging, Perforating and Testing,” page 50).
Oilfield Review first evaluated HPHT
drilling in 1993.1 Much progress has been
made in the last five years. This article exam-
ines recent advances in HPHT well construc-
tion, drilling fluids and cementing techniques,
with examples of procedures and guidelines
for routine and contingency actions. We eval-
uate current well-control practices that
emphasize the application of computer simu-
lations to model downhole temperatures and
drilling fluid properties. When combined with
sophisticated hydraulic models, these simula-
tions help well engineers reliably predict
downhole pressures to within 1% of the
actual conditions. Techniques for measure-
ment of drilling fluid parameters at surface to
the same accuracy complement the models,
permitting control of fluid properties and ulti-
mately downhole pressures. Finally, we
review new cementing practices developed
specifically for HPHT wells.
■HPHT well locations
worldwide. HPHT wells
are concentrated in the
US Gulf Coast and the
North Sea, where reser-
voir pressures of 16,000
psi and bottomhole
temperatures of 350°F
are common. In Thai-
land, more than 45
high-temperature wells
have been drilled.
Planning for Success
Most of the hazards of drilling HPHT wells
are related to overpressured formations.
Ideally, such wells would be drilled with
high enough mud weight to give a comfort-
able safety margin over pore pressure. The
mud engineer’s job in formulating the mud
would then be relatively straightforward:
minimize formation damage and maximize
rate of penetration.
An overpressured formation becomes a
major problem when the formation fracture
pressure is close to that in the overpressured
zone. This results in drilling conditions
where kicks are easily taken, and where
fractures can be inadvertently initiated,
resulting in drilling fluid losses that are diffi-
cult to control.
These problems are clearly illustrated in the
Elgin and Franklin fields, operated by Elf
Exploration UK Plc, in the North Sea approx-
imately 240 km [150 miles] east of Aberdeen,
Scotland.2 Reservoirs in the Upper Jurassic
Franklin sands and the Middle Jurassic
Pentland sands are hot (up to 392°F [200°C])
and deep (18,370 ft [5600 m]). The pressure
gradient exhibits a marked increase below the
1. MacAndrew R, Parry N, Prieur J-M, Wiggelman JH,
Diggins E, Guichency P, Cameron D and Stewart A:
“Drilling and Testing Hot, High-Pressure Wells,”
Oilfield Review 5, no. 2/3 (April/July 1993): 15-32.
2. Gibson MT, Bergerot J-L and Humphreys A:
“Developing a Greater Understanding of the Well
Hydraulics in HP/HT Wells: An Updated Study of
the Wells in the Elgin/Franklin Fields,” presented at
Drilling ‘96: The 4th Annual Industry Forum,
Aberdeen, UK, March 27-28, 1996.
Houwen OH and Geehan T: “Rheology of Oil-Base
Muds,” paper SPE 15416, presented at the 61st SPE
Annual Technical Conference and Exhibition, New
Orleans, Louisiana, USA, October 5-8, 1986.
37
Kimmeridge claystone with reservoir pressure ards and costs, experience gained in previ- procedures for all critical eventualities and
of 16,330 psi [112.5 MPa] (below). These con- ous operations must be channeled into max- produced a comprehensive, practical HPHT
siderable drilling challenges are compounded imizing safety, eliminating unproductive manual.4 Open communication throughout
by the small margin—approximately 1400 psi time and increasing efficiency. With HPHT the discussions helped to achieve a consensus
[9650 kPa]—between fracture and pore pres- well costs of $20 million becoming relatively on best practices. One- and two-day training
sure in the lowermost intervals. A 1400-psi common, and with HPHT rig rates among courses familiarized personnel with the pro-
hydrostatic pressure window at 15,000 ft the highest in the industry, time spent cor- cedures, along with refresher sessions on the
[4750 m] corresponds to a small increment in recting well-control problems is expensive. rig. Feedback during the training sessions and
mud weight of 1.8 ppg [0.22 g/cm3]. Similar The cornerstone of this preventive planning as actual operations progress results in refine-
conditions—with even smaller margins— is the transfer of HPHT experience, captured ments and improvements to procedures.
exist in other North Sea wells, such as those in in guidelines, procedures and checklists. Much of the Mallard HPHT manual com-
the Heron project operated by Shell Expro, Some countries have legislation or govern- prises procedures for drilling, well control,
and in other HPHT fields around the world. ment recommendations to guide operators; suspension or abandonment, and emergency
Above all other considerations, risk reduc- other procedures are based on industry best planning—along with descriptions of person-
tion drives the planning needed to overcome practices. An example is the UK Institute of nel responsibilities. For example, the manual
these challenges. Given the potential haz- Petroleum “Model Code of Safe Practice” for spells out the need for skilled personnel to
HPHT wells.3 In addition to formal guide- ensure continuous and diligent monitoring of
lines, North Sea operators usually seek input key parameters when drilling the reservoir
Formation Formation pressure, psi from contractors and others experienced in section. Another requirement to minimize
1450 4350 7250 10,150 HPHT well construction. the number of unnecessary personnel on
Vertical depth, m

Shell Expro, and its lead contractor Sedco
Forex, adopted the policy of mobilizing
experienced crews in advance of particularly
challenging projects, such as their North Sea
NORDLAND
board is common to most HPHT operations.
A balance is needed. Equally important is
the description of communications. “Non-
routine” operations meetings, for example,

Mallard and Heron field developments. This
1000
policy ensures that HPHT crews have had
previous experience working as a team on
less challenging wells, which is seen as a
sound investment in achieving a satisfactory
CLAYSTONE
are required to discuss any nonstandard
operations and ensure that all parties clearly
understand the program and their roles in it.
During routine drilling operations the
hand-over between crew shifts must be

outcome on subsequent HPHT wells. controlled, because any small change

EOCENE

In planning the detailed program for the
Mallard wells, Sedco Forex, Shell Expro and
2000
other subcontractors met to work through
HORDLAND
in observed drilling parameters might indi-
cate a large potential problem. In addition
to specifying a routine pretour meeting,
the HPHT manual provides checklists to

ensure that essential information and trends

Transition Zone Pressures
Pressure, psi
0 1450 2900 4350 5800 7250 8700 10,150 11,600 13,050

4000 Formation
Hydrostatic
3000
Vertical depth, m

4500 Fracture
EKOFISK BALDER
PAL
M

5000 Unconformity
TOR

5500

4000
CHALK GROUP (LMST)
UPPER CRETACEOUS

0 10 20 30 40 50 60 70 80 90
Pressure, MPa
HOD

■Elgin pressure gradient. Overpressured intervals are char-

acteristic of HPHT wells. The formation pressure gradient
(left) is from the Elgin field in the Central Graben area of
the UK Sector of the North Sea. The top of the overpressured
gas/condensate reservoirs of the Upper Jurassic Franklin
HERR
HID

5000 sands and Middle Jurassic Pentland sands reaches 380°F

LC

with absolute static pressures greater than 16,000 psi. An

KIMM

CLST

unconformity at the base of the Lower Cretaceous (LC) for-

U. JURASSIC

mation coincides with the shale that is responsible for the

FRANKLIN

overpressure. The inset shows the predicted hydrostatic

SDST

pressure near TD as well depth increases. A large increase

in mud weight is needed at 5100 m to control formation
PENTLAND

1.0 1.5 2.0 2.5 Formation pressure gradient, psi/ft pressure. However, hydrostatic pressure then approaches
MJ

fracture pressure. Achieving adequate hydrostatic pressure

0 10 20 30 40 50 60 70 80 90 while avoiding fracturing the formation is a challenge for
Formation pressure gradient, MPa the HPHT mud engineer.

38 Oilfield Review
 ■Establishing
kick detection
Hole taking Pump Increase Increase guidelines. Sedco
Drill Gas cut
less mud pressure in return in pit Forex, acting as
break mud
than calculated decrease flow rate volume lead contractor for
Shell on the
Mallard develop-
ment, collaborated
with the operator
and contractors to
Flow check pool experience
in HPHT well
construction. In
order to focus the
No Yes attention of the rig
Shut in well using
Well flowing? personnel on
fast shut-in method
specific HPHT
problems, a system
No of Kick Alert Levels
Pit gain?
is implemented.
Monitor pressure
Personnel are
Yes advised of the risk
Investigate cause
level by notices
of indication
posted in key loca-
No Pressure tions around the
buildup? rig, by loudspeaker
announcements
In doubt Yes Yes and in the morning
about influx? report. This kick
Continue with recognition
well-control procedures flowchart from the
No Sedco 704 HPHT
Manual describes
how kicks should
Hole taken be handled after
No being recognized
correct volume RIH to bottom
(if tripping)? by observing one
or more of a num-
ber of indicators
(top row).
Yes

HP well
Yes
oil-base mud
drill break?

No
Circulating bottoms up
Yes taking returns through
Gas cut
choke prior to bubblepoint
mud?

No Continue drilling or Flow check well after drilling

tripping operations cautiously 15 ft or operating 5min

are communicated. An example is the
driller’s “Tour Checklist,” which requires the
driller to carry out a physical check of mud
density, viscosity and temperature every
15 minutes up to the tour change, and com-
municate this information to the mud engi-
neer, mud logger and drillers, as well as to
the new crew. Frequent checks of these
critical parameters may reveal any early
signs of potential trouble.
Kick control, the most critical of all proce-
dures, starts with an awareness of the risks
during each phase of drilling. Sedco Forex
Summer 1998
uses a three-level “Kick Alert Status” to focus 3. The Institute of Petroleum: “Well Control During the
the attention of rig personnel on specific Drilling and Testing of High Pressure Offshore Wells,”
(Part 17 of the Institute of Petroleum Model Code of
procedures (above). The highest alert Safe Practice in the Petroleum Industry), London,
level requires adherence to all procedures, England: John Wiley and Sons, May 1992.
precautions and checklists set out in the 4. SEDCO 704 HP/HT Manual, Mallard Development.
Sedco Forex, November 1997.
HPHT manual.
Prevention is, of course, better than a cure
for kicks, and procedures capture the mud
engineers’ calculations regarding maximum
pump rates, tripping speeds and tolerances
on parameters. The rig superintendent
checks these hydraulic calculations daily to
minimize the risk of a kick or of fracturing
the formation.
39
 Similar procedures were in place during Clearly the ability to predict these effects is Fluid Simulator
the recent drilling of three Heron wells for critical to the successful drilling of HPHT Simulation starts with a well model that
Shell Expro. Not a single kick was taken wells. Small but serious errors in computing includes a profile of temperatures in different
when drilling the reservoir sections, and this the drilling fluid pressure at the reservoir sections of the well and the heat transfer
success was largely attributed to the thor- may result from ignoring uncertainties due to between them under relevant flow condi-
oughness of the procedures established to either temperature or fluid properties. tions. The thermodynamics of the heat trans-
control drilling parameters, such as tripping Simulation of downhole temperature profiles fer can be modeled as a “countercurrent heat
speeds. On the one occasion that a kick was at all phases of the drilling operation is there- exchanger,” familiar to engineering students.
taken when drilling above the reservoir in a fore the key to understanding the behavior of
Mallard well, it was detected after only 8 bbl HPHT drilling fluids.
[1.3 m3] flowed, allowing prompt action
to control it. Formation Cement Drillpipe Drilling fluid Casing ■Heat transfer to
the wellbore. Cir-
In some cases, the conflicting demands culating mud
of pore pressure and fracture pressure are behaves as a
difficult to reconcile, and a risk analysis may countercurrent
be needed to determine which one is more heat exchanger.
The rate of heat
important to address. An informed decision
exchange between
might be made to risk fracturing the forma- the mud and the
tion, rather than take a kick, based on the rel- material at any
ative ease with which fractures can be particular depth
plugged or cemented. depends on the
temperature, ther-
mal conductivity
Drilling Fluid Properties and specific heat
Hole cleaning and control of formation pres- capacity of the
sures are critical functions of the drilling fluid. materials and on
the velocity of the
These are achieved when the mud pressure is
mud. In the pres-
sufficient to hold back pore pressure and the ence of casing,
viscosity is sufficient to transport cuttings from vertical conduc-
the bit to surface. Viscosity must also be suffi- tion of heat further
cient to hold mud solids in suspension.5 complicates the
temperature
Conventional calculations of downhole distribution.
pressure, which assume constant drilling
fluid properties, are both practical for day-to-
day use and accurate enough for routine
Temperature
wells. Downhole static pressures are easy
to calculate from mud weight measured at
surface, while additional pressures due to
2
circulation can be calculated using estab-
lished relationships between pump rate and
drilling fluid rheological properties.
Errors that result from ignoring variations in
Steady-stat
mud properties are small in relatively shal- e
circulation
low wells. In these settings, mud engineers 3
file
Depth

can concentrate on formulating drilling fluid tic p ro

properties for maximum rates of penetration
and optimal hole condition. Formations can
commonly withstand moderate overpressure
before being fractured, which permits mud
engineers to add a comfortable safety margin
when weighting the mud.
However, mud properties do vary with
downhole pressure and temperature, affect-
ing the accuracy of both surface measure-
ments and downhole estimations of mud
weight and viscosity. In HPHT wells these
variations can be significant because of the
limited safety margins available.
5. Fraser L, Stanger B, Griffin T, Jabri M, Sones G,
Steelman M and Valkó P: “Seamless Fluids Programs:
A Key to Better Well Construction,” Oilfield Review
8, no. 2 (Summer 1996): 42-56.
Sta
1
Time
■Transient temperature behavior. When left without circulation for more than 24 hours,
annulus mud temperature approaches the geothermal gradient. After tripping, circulation
rapidly reduces the temperature of the mud at TD (contour 1), while the temperature of
the mud returning to surface from near TD continues to increase (contour 2). The depth
at which the mud starts to cool off moves uphole with time to a point about one-third of
the well depth above TD (3). After approximately 6 hours, the circulating mud will have
reached a dynamic equilibrium, during which the temperature profile stays constant.
The transient temperature behavior must be simulated to predict total downhole pressure
after drilling begins.

40 Oilfield Review
 Circulating fluid moving along the well
gains or loses heat from or to its surroundings
(previous page, top). The rate of heat
Vertical Temperature Profile
exchange depends on the temperature and in Drillpipe and Annulus
velocity of the fluid, the thermal conductivity
Temperature
of the formation, the geothermal gradient in
the undisturbed reservoir, the specific heat
capacity of the mud and other factors. In the
presence of casing strings, significant vertical
conduction of heat further complicates the
temperature profile.

Depth
There is a net transfer of heat from the for-
mation to the mud as it goes down the well.
On reaching the bit, the mud is still cooler
than the surrounding formation. The mud
continues to heat up as it returns to surface
until it reaches a depth where the formation
temperature equals the mud temperature. Annulus at
Above this depth, the mud cools on its way 100 gal/min
to surface.
Steady-state temperature profiles can be
computed once a satisfactory model has
been developed. With time, thermal equilib-
rium can be achieved in two ways—after cir-
culation has ceased or under constant
circulating conditions. The static steady-state
profile approaches the geothermal gradient,
while the circulating temperature profile will
vary with pump rate.
The MudCADE program developed by
Dowell computes vertical temperature profiles
under various conditions. This program has
been validated with field data from numerous
other HPHT operations. The main inputs are
specific heat capacity and thermal conductiv-
ity of each component, and the main outputs
are the temperatures of the mud inside the
drillstring, and the mud inside the annulus,
between the drillstring and the casing (right).
Between steady-state circulation and stabi-
lized static conditions, the temperature pro- Drillpipe at Drillpipe at
250 gal/min 100 gal/min
files change with time (previous page,
bottom). In theory, after circulation ceases it
takes approximately 16 hours for the mud
temperature to approach within 10% of Annulus at
the geothermal gradient, while circulating 250 gal/min
temperatures can take over 6 hours to
equilibrate. The temperature simulator must
predict temperature transients to enable
computation of well pressure during and
after changes in pump rate. Where safety
margins above pore pressure are small, the
reduction in static pressure after circulation
may be considerable. Once the details of
well temperature are known, the effective
mud weight (EMW) can be computed from
the relationship between local density, pres-
sure and temperature. ■Mud temperature profiles. Mud density and viscosity vary with temperature, and must
be known accurately to compute static and dynamic pressures at each depth. Heat
transferred to the mud raises its temperature as mud descends the borehole. When the
mud arrives at the bit, it is below formation temperature and continues to heat up as it
returns up the annulus. At approximately one-third of the way up the well, the annulus
mud reaches the same temperature as that of the formation. Above this depth, heat is
conducted from the mud to the formation, cooling the mud as it returns to the surface.

Summer 1998 41
 Annular
pressure
loss to
Circulating surface
pressures
Depth
Fracture
Static pressure
pressure

Increasing pump rate Equivalent

circulating
pressure

Annular pressure losses Pressure

Static Circulating
pressure at TD pressure at TD

■Equivalent circulating density (ECD). During circulation, the pressure increment needed to overcome friction in the annulus and circu-
late the drilling fluid from a particular depth to surface represents an annular pressure loss (APL). APL increases with pump rate (left), and
drilling fluid viscosity, and adds to hydrostatic pressure, increasing the total pressure at total depth during circulation. Pump rates must
not allow mud pressure to exceed formation fracture pressure (right). At each pump rate an equivalent circulating density can be com-
puted that would result in the same total pressure at a particular depth. Since annular pressure losses are a function of viscosity and well
geometry, the concept of ECD is most useful when viscosity is well-defined and APL can be modeled as a function of pump rate.

Equivalent circulating density (ECD) is often 0.9

much higher than EMW in HPHT wells due T = 17.2˚C
to the small annular clearances between the
Specific gravity

drillpipe and hole wall (above). ECD is com-
puted from the dimensions of the annulus
and, for a given fluid viscosity, increases with
pump rate. The calculation becomes increas-
ingly complicated when changes of viscosity
with temperature are considered.
Computing Downhole Fluid Pressure
Instead of using EMW and ECD when calcu-
lating pressure in HPHT wells, it is more
accurate to consider static, dynamic and cut-
tings pressures as components of the total
downhole fluid pressure.
Static pressure—Static pressure is com-
puted by integration of hydrostatic pressures
at each depth. To achieve this, pressure-
volume-temperature (PVT) analysis is usually
performed on the mud or base fluid (above
right). Many base fluids used for oil-base
muds have high compressibility compared
with water-base muds.
By starting at the surface where the pres-
sure and temperature are known, the local
density of the fluid can be computed. The
predicted hydrostatic pressure and tempera-
ture permit the density at the next deeper
level in the well to be computed. At the well-
site, the measured mud weight is used as the
starting point, increasing the accuracy of the
initial conditions. With PVT data, static pres-
sure at each depth can be computed with
Dowell MudCADE and DSHyd software.
42
0.8 T = 100˚C
T = 140˚C
0.7
T = 170˚C
0.6
T = 205˚C
0.5
0 20 40 60 80 100 120
Pressure, MPa
■ULTIDRILL base fluid data. Prediction of hydrostatic pressure requires pressure-volume-
temperature (PVT) data for the mud in addition to an accurate simulation of downhole
temperature profile. The compressibility of a drilling fluid depends on its base fluid; the
solids are incompressible. This example is the PVT diagram for the low-toxicity, biodegrad-
able organic base fluid of the ULTIDRILL system, which has been used at 395°F and at
weights up to 19 ppg. The specific gravity of the base fluid under these conditions at a
depth of 16,000 feet, is 0.68. The same fluid (gold arrow) returned to surface temperature
and pressure has a specific gravity of 0.79, a 14% decrease in base fluid density at TD,
which is important in computing static pressure.
Dynamic pressure—The dynamic pressure viscometer tests (next page, middle).
term is more comprehensive than permitted Alternatively, the mud properties may con-
by the concept of ECD. It can account for form to established relationships, such as
annular pressure losses (APL) due to moving Bingham plastic model or an empirical
fluids, pipe velocity (swab and surge) and power law model with parameters chosen to
inertial pressure from string acceleration represent the specific mud behavior (next
when tripping and excess pressure required page, bottom).
to break thixotropic gels (next page, top). Modeling software such as the Dowell
Predicting the dynamic pressure contribu- DSHyd and MudCADE programs incorpo-
tion to the total pressure requires accurate rates the algorithms for computing dynamic
modeling of mud rheology. Several different pressure from either Bingham plastic or
relationships are available to relate shear power law models. The advantage of these
stress to shear rate, and define dynamic vis- (over more complex models) is that the rhe-
cosity at a given shear rate and temperature. ology parameters derived from them can be
Depending on the fluid, the mud engineer easily compared to wellsite measurements
selects an appropriate rheological model on made using viscometer readings.
the basis of fitting a curve to data from HPHT
Oilfield Review
 ∆Phydrostatic ∆Pdynamic + ∆PAPL +
∆Phydrostatic ∆Phydrostatic

(a) (b) (c)

■Swab and surge. Movement of a pipe through a viscous fluid causes shearing in the boundary layer adjacent to the pipe, accompanied
by a shear strain in the fluid. The shear strain is seen as a pressure difference, ∆P, in the fluid, which adds to the hydrostatic pressure (a).
Bottomhole pressure is reduced when pulling pipe out of hole (swabbing) (b) and increased while running pipe into the hole (surging) (c).
The changes in pressure are a function of fluid viscosity, well geometry and pipe velocity. Pulling pipe out of the hole at excessive speeds
can result in well pressure falling below formation pressure, inducing a kick. Conversely, running pipe in hole at excessive speeds may
result in the formation fracture pressure being exceeded. Modeling the dynamic pressures created during swabbing or surging enables
the mud engineer to determine safe tripping speeds.
■HPHT rheometers. A model of drilling fluid viscosity at downhole
Fann 70 Huxley- conditions is essential to account for dynamic forces correctly,
Bertram and begins with measurement of drilling fluid viscosity at down-
Rating Temperature 250˚C 300˚C
hole temperature and pressure. These measurements can be
made with either a Fann 70 or a Huxley-Bertram HPHT rheometer.
Pressure 20,000 psi 15,000 psi The Fann 70 rheometer is rugged enough for wellsite measure-
ments. The Huxley-Bertram is a coaxial cylinder laboratory
Additional Density/compressibility No Yes rheometer in which an outer rotating cylinder applies shear stress
measurements Long-term gel strength Yes Yes to a fluid that, in turn, directly transmits measurable shear to an
inner stationary probe. Temperature and pressure may be varied
independently, permitting modeling of most wellbore conditions.
Drawbacks Contamination possible Yes No
The Huxley-Bertram rheometer has five times the resolution at the
lower end of the shear stress range and can measure density,
compressibility and long-term gel strength under HPHT conditions.
Cuttings pressure—An additional compo-
nent of the total pressure is from cuttings
Rheology Analysis
accumulation, known as the cuttings pres- 200
sure. Although high-density muds used in
HPHT wells tend to reduce cuttings accu-
mulation, their contribution to total mud
pressure cannot be ignored. As cuttings are 150
2
Shear stress, lbf/100 ft

more dense than mud, any accumulation of

cuttings in the wellbore will increase the
mud pressure acting on the borehole. The 100
additional pressure from cuttings is affected
by rate of penetration (ROP), pump rate, cut-
tings size and distribution.
50 Fann data
High ROP increases accumulation of cut-
Bingham model
tings, and creates large cuttings, which have
Power law model
a higher settling velocity. Although both
these effects can be reduced by increasing 0
pump rate, they increase dynamic circulat- 0 100 200 300 400 500 600
ing pressure at the bit. Therefore, during Fann rotary speed, rpm
drilling there is a limit to how much the ■ULTIDRILL fluid behavior matched to power law model. A model of the relationship
dynamic pressure can be reduced before between shear stress and shear rate can be chosen after establishing the relationship at
benefits are overcome by increased cuttings a given set of downhole conditions (using an HPHT rheometer). In this example, ULTIDRILL
fluid behaves according to the power law model, so its viscosity can be predicted across a
pressure. When cuttings pressure becomes range of shear rates for the given conditions. Occasionally, drilling fluids may demonstrate
significant at maximum pump rates, it can be behavior similar to more than one model under different conditions. By making a series of
controlled by reducing ROP. laboratory tests, it is possible to predict viscosity across a range of downhole conditions.

Summer 1998 43
 Measured standpipe pressure, psi retain the desired downhole fluid properties.
10
0 290 580 870 1160 1450
1450
Reducing pump rate or mud viscosity while
maintaining the pump rate high enough to
clean the hole and prevent excessive cut-
Predicted standpipe pressure, MPa

Predicted standpipe pressure, psi

8 1160 tings pressure can help. The objective is to
find a pump rate that minimizes the contri-
butions from dynamic and cuttings pressure.
6 870 While dynamic pressure is controlled by
reducing the viscosity, care must be taken to
ensure that weighting material stays in sus-
4 580
pension. Mud solids falling out of suspension
cause density segregation of the drilling fluid
2 290 called sagging (below). The greatest problem
caused by sag is the poor control of bottom-
hole pressures. Unexpectedly high mud
0 0 weight can cause induced fractures and lost
0 2 4 6 8 10
circulation while low mud weight permits
Measured standpipe pressure, MPa
formation fluid influx and wellbore instabil-
■Measured versus predicted standpipe pressure. Accurate mathematical expressions ity. Sagging can occur in both dynamic and
are used to describe dynamic pressure losses in fluids described by the Bingham plastic static conditions but is most likely to occur
model and other models. After making wellsite viscosity measurements to define the
model, mud engineers compute pressure losses using software like the Dowell DSHyd under low shear-rate conditions before static
program. A comparison of standpipe pressure predicted in this way with measured viscosity is achieved.
pressure is shown. The average error in the predicted measurement is less than 2%. Reduction of total pressure by manipulat-
ing static pressure can be achieved only at
Total pressure—The total pressure is the ior to both models at the wellsite, and to the expense of the safety margin above for-
sum of static, dynamic and cuttings pressure. select the appropriate one, is of great benefit mation pore pressure. When the well is at a
Expressing the downhole pressure in this in accurately predicting APL. In practice, the critical stage, this margin may need to be fur-
general form covers all phases of the opera- DSHyd program routinely achieves average ther reduced while drilling, with additional
tion. The total pressure can be balanced differences of less than 2% between predicted circulating pressure preventing influx. The
between the lowest safe static pressure and and measured standpipe pressures (above). well would then have to be circulated to
the highest acceptable circulating pressure heavier mud prior to tripping. In such cases,
by achieving a compatible balance of the Pressure Control extreme care must be taken when making
different terms. The lowest pressures are Drillers try to avoid fracturing the formation, drillpipe connections, as the absence of
achieved while accelerating the pipe during but at critical depths the difference between dynamic pressure may permit influx of well
pulling out of hole (swabbing) after circulat- the pore pressure and the fracture pressure— fluid (connection gas).
ing bottoms up. The highest pressures occur mud pressure margin—is very small in some Manipulation of the drillstring must be
while drilling with high pump and penetra- wells, on the order of 500 psi [3.4 MPa]. If done carefully, lifting slowly to minimize
tion rates, while breaking circulation or total pressure approaches fracture pressure, swabbing, and steadily, because low accel-
while the pipe is accelerating during running the first option is to reduce the dynamic pres- eration minimizes inertial pressures. Pump
in through gelled fluid. sure. Control of downhole pressure within rate must also be carefully controlled when
During drilling, mud properties may vary the mud pressure window can be achieved in displacing lighter mud to heavier mud before
with time to such an extent that the rheologi- many ways. Viscosity, mud weight, mud tripping. During tripping the cuttings pres-
cal model might need to be changed. solids content, pump rate and ROP can be sure will be zero if the hole has been
Temporary variations in mud properties can manipulated to change the various terms in cleaned properly. The effects of tripping
cause the same fluid to behave as a power law the total pressure expression. speed and acceleration on the total pressure
fluid at one time and as a Bingham plastic Updated pressure predictions, based on can be predicted by a simulator such as the
fluid at another time, even in the same hole wellsite measurements, can aid in deciding DSHyd or MudCADE programs. Optimal
section. The ability to compare actual behav- which parameters need to be changed to speeds can be determined and used in plan-
ning the well construction.

■Sagging. In the absence of circulation,

gravity may cause weighting material
(barite) to fall, resulting in density segrega-
tion or sagging (a). In deviated wells, sag-
Low-density fluid ging may result in a barite bed on the low
side of the hole (b). Depending on the angle
of the wellbore and the strength of the bed,
the barite beds can slump down the low
Suspended barite side of wellbores like an avalanche (c). The
movement of the solids in the drilling fluid
Sagged (sedimented)
during sagging may result in a lowering of
barite
the viscosity by shear thinning, accelerat-
Slumped barite ing the process. Ultimately, slumping may
result in barite accumulation and a pro-
(a) (b) (c) (d) nounced density change within
the drilling fluid (d).

44 Oilfield Review
Choosing Oil- or Water-Base Mud ■Key drilling fluid
Drilling Fluid Property Required Performance in HPHT Wells
The decision to drill with oil- or water-base properties. Mud
mud is likely to be made before detailed Plastic viscosity As low as reasonably possible to minimize ECD engineers must
consider many
analysis of the mud properties is attempted. aspects of mud
Water-base muds (WBM) include conven- Yield stress and gels Sufficient to prevent sag, but not so high as to performance
tional bentonite muds and polymer systems, cause gelation, or high surge and swab pressures when designing
while oil-base mud can be an invert emul- an HPHT drilling
HPHT fluid loss As low as reasonably possible to prevent formation fluid system. The
sion, all-oil system, or synthetic oil-base. damage and risk of differential sticking table shows the
Each has advantages in cost, environmental effect of several
HPHT rheology Predictable in order to control sag, gelation and ECD
impact and drilling performance. key drilling fluid
The stability—as defined by rheology properties on mud
and fluid-loss control—of oil-base muds at Compressibility Must be known to estimate downhole pressures performance.
high temperatures is a clear advantage and ECD
in HPHT wells. Most are stable to at least Stability to Stable in the presence of gas, brine and cement
450°F [230°C] in 16-hour laboratory tests. contaminants
However, oil-base muds do have some sig- Gas solubility Needed for accurate kick detection and modeling
nificant disadvantages under HPHT condi-
tions. Possibly the greatest of these is the
solubility of gas in the base fluid, which Stability to aging Properties do not change over time under either
static or dynamic conditions
makes kick detection more difficult. A gas
influx while drilling will go into solution and Solids tolerance Properties insensitive to drilling solids
remain there without causing a significant
increase in mud volume until it is relatively Weighting Must be able to be weighted up rapidly if a kick
near the surface. On coming out of solution, is taken
the gas rapidly increases in volume, requir-
ing quick reactions to control the well. Also,
the thermal expansion of oil-base muds is

High viscosity
higher than water-base muds, which can

and gel
lead to pressurization of the annulus.
The conventional advantages and disad-
vantages of oil-base mud also apply to HPHT
wells: good protection against differential 6
sticking, reservoir protection and wellbore
stability in shales, claystones and salt.

et
H

ra
ig

p
h

m
pu
m

ULTIDRILL synthetic oil-base mud is

ud

gh
Hi
w
ei
gh

biodegradable and has low toxicity, reducing

t

the potential for environmental damage dur- 7

ing its use. It has been successfully used at
temperatures up to 395°F [202°C].
8
Optimizing HPHT Drilling
Careful engineering and monitoring are 9
4
needed to reconcile the demands made of 1
e
mud systems (above right). Procedures for rat Low
mp mu
using the mud must be consistent with pu dw
w 3 eig
practical tripping speeds while ensuring Lo ht
Low viscosity and gel

preservation of safety margins under all

conditions.
Mud engineering—Consideration must be
given to maintaining the drilling fluid in
good condition through efficient solids con-
trol, minimization of dilution requirements
and the reduction of fluid retention by cut-
tings (right). Although many criteria must be
considered when designing HPHT drilling
2
Fluid behavior
1 Low conditioning requirements
2 Low ECD and risk of fracture
3 High rate of penetration
4 Differential sticking minimized
5 Sagging minimized
■Optimizing drilling fluid properties. Three main
variables can be manipulated to achieve satisfac- 6 Good hole cleaning
tory drilling fluid performance: viscosity, mud 7 Maximum kick prevention
weight and pump rate. The illustration shows how 8 Good hole stability;
different optimizations of these variables lead to washouts minimized
different properties of the drilling fluid system. 9 Increased hydraulic
For example, a low risk of fracture (low ECD) is horsepower at bit
achieved with low mud weight, low viscosity and
gel concentration, and low pump rate.

Summer 1998 45

Input parameters
Hole Pump
cleaning pressure
yes yes
Achievable? Achievable?
no no
fluid, often one desirable property will be
achieved only at the expense of another. To
solve this problem, Dowell has developed a
method to optimize formulation properties
until all criteria can be met to the satisfac-
tion of the mud engineer (above).
Weighting up—Once the base fluid has
been selected, PVT data provide the basis for
local density modeling. The static pressure
will already have been computed so as to
exceed formation pore pressure by a mini-
mum safety margin at all depths. Although
this seems straightforward, variations in mud
density due to temperature and pressure
must be considered in determining the static
pressure. Further complications may be
encountered offshore where the minimum
static pressure is required to overcome pore
Barite sample A B C D
Specific gravity 4.25 4.23 4.27 4.22 4.30
Surface area, m2/g 1.3 1.1 4.1 1.7 0.4
Results in standard bentonite WBM after hot
rolling 16 hours at 400˚F
PV, cp 31 34 34 38 34
YP, lbf/100 ft2 16 4 20 0 0
API fluid loss, mL 11.5 6.0 14.0 11.5 2.5
Suspension
properties VG P VG G G
■Effects of barite quality. Susceptibility
to sagging in a drilling fluid is greatly
affected by the particle size of the weight-
ing material. The smaller the particle size,
the slower the speed with which it falls
through a static fluid, and static sag
is less likely to occur. The effects of barite
quality are summarized in the table,
which shows the plastic viscosity (PV),
yield point (YP), API fluid loss and quali-
tative particle suspension of a water-base
bentonite mud recipe made up with five
different barite samples. Suspension
properties have been rated as follows:
VG = very good, G = good, P = poor. The
barite particle size is reflected by the sur-
face area (m2/g). The smaller the particles,
the greater their surface area per gram,
so the finer barite has a larger surface
area (as in sample C).
46
Barite ECD and bit Other
sagging hydraulics criteria
yes yes
Tolerable? Achievable? All met?
no no no
Modify parameters
pressure when the marine riser is discon-
nected. The “riser margin” is a safety margin
needed when moving off a well in an emer-
gency, and increases the minimum dynamic
pressures used while drilling—especially in
deep water where the additional pressure
with the riser connected is high.
Weighting material is chosen to achieve the
required mud weight based on offset well
pressures and factors such as sag and gelling
behavior of the mud. Barite, milled to
achieve desired particle size (typically less
than 75 ␮m), is a common weighting agent.6
High-quality barite is essential for HPHT mud
because impurities or poor particle size dis-
tribution may result in problems that are mag-
nified in the HPHT environment (below left).
The presence of clay impurities in barite can
result in gelation at 275°F [135°C] as the
E clays “burn out.” Carbonate and iron miner-
als also cause changes in the mud as their
solubility with changing mud temperature
causes flocculation. Other chemical residues
lead to severe foaming problems at high tem-
perature. Sagging problems can be reduced
by specifying ultrafine particle size (less than
15 ␮m).
In high-density mud systems, barite can
account for up to 78% by mass and up to
45% by volume. When high mud weights
are required, the solids ratio may reach a
point where the mud becomes susceptible to
gelling in the event of fluid loss. The solids
ratio can be reduced by using high-density
weighting agents such as hematite. In any
case, the mud engineer routinely measures
solids content and fluid loss during mud
checks to ensure they do not exceed design
parameters during drilling.
Problematic gels—Gels and high-viscosity
fluids have considerable effects on the mud
pressure acting on the formation when trip-
ping in or out of the hole. They are controlled
by varying the appropriate mud additives to
prevent excessive gelation while ensuring
that weighting agents stay in suspension. If
the gel strength and viscosity can be deter-
■Dowell mud
design flow chart.
Dowell has devel-
oped a standard
procedure to opti-
mize mud rheology
and hydraulics.
Starting with hole
yes Optimum cleaning perfor-
design mance, a range of
parameters properties of the
proposed mud are
checked by model-
ing. Parameter
changes are made
as required until the
properties and per-
formance of the pro-
posed mud system
meet specifications.
mined at downhole conditions, their effects
on dynamic pressure can be simulated.
When influx cannot be prevented by restrict-
ing tripping speed and acceleration, special
procedures are needed to permit operations
to continue. The simplest procedure is to
increase the mud weight before pulling out,
but other techniques are possible.
If the predicted swab pressure is greater
than the trip margin, the pumps can be oper-
ated while tripping out. The additional
dynamic pressure resulting from maintaining
circulation can be made sufficient to achieve
the desired margin above pore pressure.
Alternatively, a heavy-weight slug can be
spotted to minimize the risk of a kick. With
careful design of both the volume and
weight of the slug, a hydrostatic pressure
profile is achieved that avoids fracturing the
formation while maintaining the desired
overbalance for tripping.
Excessive swabbing is not the only problem
associated with high gel strength. The pres-
sure increment needed to “break” (initiate)
circulation in the presence of gels can be
much higher than circulating pressures. In
the worst case, the formation fracture pres-
sure can be reached before the gel is broken.
Gel strength needs to be sufficient to support
the heavy particles in the mud but no more,
otherwise excessive pressures may be
needed to break the gel. The pressures
needed to break circulation can be modeled
at the design stage, and the implications
of these additional dynamic pressures
accounted for in the design.
Excessive gel strength can result in a further
hazard—trapped pressure (next page, top).
An influx of formation fluid below a gel will
not be observed as a flow at surface until the
gel is broken, by which time a large influx
may have occurred, resulting in a rapidly
deteriorating well-control situation. The prob-
lem is magnified by the ability of gas influx to
Oilfield Review
 for pump pressures below rig capacity to
enable dynamic kill, the intentional increase
of dynamic pressure by increasing annular
flow rates, to be used when attempting to
control the well.
Controlling sag—If hole cleaning is not a
concern, the mud engineer can focus on
assessing the possibility and effects of sag-
ging. Modeling sagging behavior is difficult,

Depth
Gel and is usually empirically assessed in lab
experiments and minimized as far as possi-
ble. Schlumberger Cambridge Research,
Cambridge, England has conducted ambient
Pressure
Gel Gel temperature and pressure experiments with a
Fluid dynamic sag tester from which guidelines on
influx
the effects of the type of weighting agents,
sag mechanisms and low-sag additives have
Pressure been derived. Procedures written into well
programs for mud conditioning, monitoring
■Gel support of pressure differentials. Absence of movement within a drilling fluid can of mud weight and tripping can all help min-
result in the formation of gels, which act as weak elastic solids. When breaking circula- imize the effects of sagging.
tion, some of the pressure beneath the gel will be taken by the pipe or borehole walls At the wellsite, the amount of sag occurring
(left). This pressure could be high depending on the gel strength and can result in forma-
tion fracture. In the case of an influx of formation fluid beneath a gel, the influx cannot be in a mud can be quantified from the heaviest
detected unless the differential pressure is high enough to break the gel. A significant and the lightest components of the circulat-
well-control hazard exists when the gel is eventually broken down during circulation, ing mud. After assessment of the amount of
resulting in a kick being taken (right). sag of the circulating mud, appropriate pro-
cedures are used to minimize it. In particu-
cause gelling in water-base muds under cer- planned overpressure. By modifying the pro- lar, if changes in mud weight indicate
tain conditions. If the gas contains carbon cedure for taking the mud sample tempera- density segregation is occurring, laminar
dioxide [CO2], the pH is reduced, causing ture during measurement of the mud weight, shear at low pump rates (which promotes
dispersants to become less effective, and the such errors have been largely eliminated. sag) should be avoided (below). Similarly,
carbonate and bicarbonate ions in the mud to Pump rates—Minimum pump rates to circulation at low pump rates prior to
continue promoting gelation. Freshwater gel clean the well are usually low due to the cementing should be minimized.
muds with high solids contents are particu- buoyancy of cuttings in high-density drilling Once sag behavior is defined, the hydraulic
larly susceptible to this effect. To minimize fluids. Therefore, in vertical boreholes, hole properties of the mud formulation can be
trapped pressure, the gel content must be cleaning is not usually a big concern in considered in the design process. The goal of
kept as low as practically possible. HPHT wells, and pump rates in HPHT wells the mud engineer is to design a mud that will
Controlling mud density—Frequent checks are more likely to be decided by other fac- continue to function between pore pressure
of mud density are essential to maintaining tors. Although low pump rates help maintain and fracture pressure at all times. These pres-
downhole pressures within the mud pressure low ECD, the well program may call for sure limits define the critical mud pressure
operating window. Recently a technique has higher rates to reduce bottom-up time and operating window and must include margins
been developed to correct surface measure- permit timely analysis of drilled cuttings to account for dynamic pressure resulting
ments of mud density for temperature effects lithology, background or connection gas, from swab and surge when tripping.
and has been used by Shell UK and Dowell and mud solids. It is good practice to design
in drilling Heron and Shearwater HPHT wells
in the UK sector of the North Sea.7 The ■Density segregation due to
drilling program for these wells specified that sagging. The graph shows the
the overbalance should be limited to 200 psi local mud weight recorded at
1000
the end of recording a suite of
[1.4 MPa]. Based on the conventional proce-
logs 67 hours after circulation
dure for measuring the mud weight, the error in a North Sea HPHT well.
2000
in mud-pressure gradient resulting from a After extensive sagging, the
temperature that was 5°F to 15°F [3°C to 8°C] weighting medium slumped to
Depth, m

different from the assumed temperature was 3000 4000 m and deeper. The mud
specific gravity was nominally
estimated at 1.5 pptf (psi per thousand ft) to 2.15. However, after the den-
4.5 pptf [0.0035 SG to 0.0104 SG]. With true 4000 sity segregation occurred, the
vertical depths (TVD) in excess of 15,000 minimum recorded was 1.89,
feet, the errors in downhole pressure were and the maximum 2.45. The
5000
Losses increase in local density near
estimated at up to 67 psi [460 kPa], far greater
TD resulted in wellbore insta-
than what was acceptable with such a small bility followed by losses,
6000
although the situation was
6. API specification for barite is that residue greater than controlled. [Adapted from Gib-
75 ␮m is a maximum of 3.0% by weight. 1.90 2.00 2.10 2.20 2.30 2.40 2.50 2.60
son, Bergerot and Humphreys,
Mud specific gravity reference 2.]

Summer 1998 47
■FMP equipment. The FMP
drilling fluids monitoring pack-
age is an online unit developed
by Dowell for measurement of
drilling fluid temperature, den- FMP Drilling Fluids Monitoring Package
sity, rheology and pH of the
water-base mud (left). All mea- Air
supply
surements are sampled for anal-
ysis using a PC running PRISM
software (right). The nonradioac-
tive densitometer measures the
resonant frequency of a U-tube
containing mud. Rheology is
measured by recording the pres-
sure drop across three rheology
flow loops of different diameters
at a constant flow rate. Observa-
tion of trends and the interdepen-
dence of parameters help distin- Densitometer
guish potentially hazardous
conditions from normal varia- Three rheology flow loops
tions, which might not be possi-
ble with infrequent sampling by
hand. In the example shown,
small variations in yield point
(YP), plastic viscosity (PV) and
pH have been caused by a
change in mud temperature.
Conventional sampling could
have resulted in unnecessary
adjustment of mud parameters if
the FMP data logger had not
been available.
Mud Stability
In HPHT wells, the temperature stability of
the mud is a key consideration in completing
the mud design. The degradation of mud
products is temperature and time dependent
and can affect all mud properties. Both oil-
base and water-base muds can suffer from
high-temperature gelation, though the mech-
anisms are different. Fluid loss increases with
temperature and is affected by both product
degradation and gelation. Finally, small
changes in solids content that result from
fluid loss can have a significant impact on
viscosity in the high-solids muds typical of
HPHT applications.
Temperature stability is tested by hot-
rolling mud samples in the lab. Samples are
placed in pressure vessels known as bombs
and heated to test temperature in an oven
while being continuously agitated by rolling.
A cycle of temperatures and pressures with
periodic assessment of properties tests the
ability of a mud formulation to withstand
drilling conditions. In some cases, samples
may be intentionally contaminated to test for
potential adverse interactions.
Quality control of raw materials is essential
if the drilling fluid is to behave as expected.
Particle size is important in defining sagging
and rheology, so it is critical that laboratory
tests be carried out on representative sam-
ples. Batch numbers and mixing procedures
ensure that the mud that goes in the well is
as close as possible to the proposed formula-
tion and to that tested in the lab.
48
(5) (3) (1) PRISM Output of Computed FMP Measurements (2) (4) (6)
100 690 12.0 50 50
Mud weight, pptf
90 140 (1)
Temperature, ˚F, reducing due to pumps off, 11.8
80 680 then cold mud in riser 40 40
(3)
70 120
11.6
Inlet strainer pressure, psi
670 (5) 30 30
60
(4)
50 100 11.4
Yield point, lbf/100 ft2
40 660 20 20
Plastic viscosity, cp 11.2
30 80 (6)
20 650 10 10
11.0
60
(2)
10 pH as read by FMP package,
varying inversely with temperature
0 640 10.8 0 0
0 am 2 am 4 am 6 am 8 am 10 am 0 am
Time of day
After establishing test results and optimiz- cement are formulated with a hierarchy of
ing the drilling fluid, the formulation of the both density and viscosity: the spacer should
drilling fluid can be recorded for mixing in be more viscous and more dense than the
bulk. In particularly sensitive conditions, the mud it displaces, and the cement more vis-
formulation includes batch numbers for cous and dense than the spacer.
products to permit traceability. Procedures The normally preferred turbulent flow tech-
for mixing and remixing the mud after aging nique for mud removal is impractical for
accompany the formulation to ensure con- HPHT applications due to the high viscosity
sistent performance in the field. and density of both spacer and cement. Flow
Data logger for mud control—To monitor rates in excess of 20 bbl/min [3.2 m3/min]
mud continuously during HPHT operations, would be needed to achieve turbulent
Dowell has developed a data logger called flow. In practice only 3 to 5 bbl/min [0.48 to
the FMP drilling fluids monitoring package 0.79 m3/min] is achievable, before the
(above). The FMP data logger records mud dynamic pressure generated would lead to
density, temperature, rheology (yield point the bottomhole pressures exceeding the
and plastic viscosity at a given temperature) formation fracture pressure.
and, in water-base mud, pH typically at 10- Modeling hydraulic flow during HPHT
minute intervals. cementing is essential to achieve the highest
The mud data are analyzed by PRISM pre- practical annular flow rates. Software, such
cision recording for job supervising and as the Dowell CemCADE program, simulates
monitoring software, and presented as a log the job based on well geometry, reservoir
of fluid parameters against time. The ability pore and fracture pressures, fluid densities
to see trends in these data is of great value in and rheology and casing strength. Several
cross-checking other measurements and combinations of pump rate, fluid densities
alerting the crew to changing conditions. For and rheology are modeled with manual iter-
example, the mud balances used on HPHT ation to find the widest possible pump rate
wells comprise matched components, range, the shortest time to complete the job
which, if accidentally mixed between sets, with the best chance of mud removal.
could result in uncalibrated readings. An Temperatures after circulating and during
inconsistent reading from such a balance cementing are also computed by CemCADE
could be seen quickly, and corrective action software for use in judging the slurry retarda-
taken immediately. tion needed.
Spacer design becomes more challenging
Cementing with oil-base mud. The spacer must be com-
After casing is set, the final step in constructing patible with both the mud and cement while
a new well is sealing the annulus between remaining stable at high temperature, retain-
the casing and borehole with cement. One of ing high viscosity and preventing excessive
the objectives of the cementing design is to fluid loss. Spacer and slurry must also be
remove the mud from the annulus with a compatible with downhole elastomers. Oil-
spacer and then to displace the spacer water emulsions, such as the Dowell MUD-
entirely with cement, leaving no channels or PUSH-XEO spacer are being successfully
other flaws. To achieve this, the spacer and used to tackle these demands in HPHT wells.
Oilfield Review
 Cement slurry design—Slurry design gener-
ally consists of formulating cement additives
to give the desired density, fluid-loss control,
rheology while pumping, the appropriate
setting time and adequate strength when set.
Slurry properties must ensure solids stay in
suspension, just as barite must be suspended
in mud. Failure to do so can result in loss of
well control and channeling on the high side
of deviated holes. As cement particles settle
to the low side, a continuous water channel
may form on the upper side of the hole, cre-
ating a path for gas migration.8
Cement slurry design involves special con-
siderations in HPHT wells. Neat cement
(without additives) is susceptible to loss of
strength and increasing permeability due to
shrinkage at temperatures above 230°F
[110°C]. To prevent this, part of the cement,
typically 25%, is routinely replaced by silica
flour (grain size 40 to 50 ␮m).9 Slurry density
of up to 17.5 ppg [2.1 g/cm3] can be achieved
by adjusting solids content. Slurry densities
greater than this are frequently needed in
HPHT wells, and can be achieved by adding
hematite to the slurry design (above right).
Avoiding unpredictable slurry performance
is the highest priority in cementing HPHT
wells. Long displacement times can be
expected because of the depth and low pump
rates needed to minimize dynamic pressure.
When these factors are combined with the
elevated temperatures encountered, there is
significant risk of premature cement thicken-
ing if the formulation is not appropriate. An
additional consideration with HPHT wells is
that in long liner intervals the temperature at
the top of the liner can be up to 50°F [28°C]
lower than at the bottom. However, retarding
the cement slurry adequately in the bottom of
the hole may result in excessive thickening
times in the top section.
Dispersant, retarder and other additives are
typically supplied as powder to be mixed into
the slurry at the wellsite. Many of these pow-
ders start life as liquids before being freeze-
7. Gao E, Estensen O, Macdonald C and Castle S:
”Critical Requirements for Successful Fluid
Engineering in HPHT Wells: Modeling Tools, Design
Procedures and Bottom Hole Pressure Management
in the Field,” paper SPE 50581, to be presented at the
SPE European Petroleum Conference, The Hague, The
Netherlands, October 20-22, 1998.
8. Bonnet A and Pafitis D: “Getting to the Root of Gas
Migration,” Oilfield Review 8, no. 1 (Spring 1996):
36-49.
9. Michaux M, Nelson E and Vidick B: “Cement
Chemistry and Additives,” Oilfield Review 1, no. 1
(April 1989):18-25.
Nelson EB: Well Cementing. Sugar Land, Texas, USA:
Schlumberger Educational Services, 1990.
Summer 1998
Cement
Silica Cement
Hematite
Weighting agent
(silica and hematite)
Stabilizer
■High-density cement slurry systems. To achieve densities up to 19.4 ppg [2.3 g/cm3],
hematite may be added to replace some of the silica and cement, thereby raising the den-
sity while retaining the same slurry porosity (left). When even higher densities are needed,
the approach is to reduce the slurry porosity by using varying particle sizes (right). By care-
ful grading of particles in DensCRETE slurry, a new solids system can be achieved in which
coarse hematite and silica are bound together by a relatively small amount of cement and
a stabilizing agent, in the same way that aggregate is bound by cement in construction
concrete. Slurry densities up to 22 ppg [2.6 g/cm3] have been achieved with this system.
dried into powders for ease of handling. In Quality assurance—Laboratory testing of
the last year, Dowell has reverted to using cement recipes is critical to ensuring the
the liquid forms directly for preparing sensi- success of the job, but is valuable only if the
tive HPHT slurries. These can be prepared mixing process at the wellsite can ensure the
and tested in bulk before being shipped to same formulation. For demanding HPHT
the wellsite. Homogenous liquids help jobs, the use of precisely the same ingredi-
improve control of the recipe, and volumes of ents as those tested is ensured by specifying
liquid can be measured more easily than lot numbers. While small quantities of addi-
weights of powder. Slurry samples taken at tives present no problem, large quantities
the wellsite are verified with laboratory tests. often do. Control of inventory to ensure
Liquid retarder has additional benefits. The availability of the same batch in sufficiently
Dowell retarder D161, for example, results large quantities can be an added complica-
in slurries that have long thickening times tion for the cementing contractor.
when moving, but which start to set when Best practice requires that all material be
pumping stops. This behavior permits over- available in double the required quantity at
retarding of the slurry, providing a significant the wellsite, in case slurry needs to be
safety factor against premature setting at the dumped for any reason after mixing (for
bottom of a long interval, while ensuring sat- example, due to contaminated mix water).
isfactory setting in 10 to 12 hours after
pumping stops at the top of an interval. The Balancing Act
Appropriate in long liners where a large Well construction strives to balance many
safety margin is needed, this approach has conflicting requirements. Drilling perfor-
been used by Shell in the North Sea in their mance must not compromise well control,
HPHT Shearwater field and elsewhere by and cementing pressure must not jeopardize
other operators. Liquid dispersants, used to formation integrity. In HPHT wells, the mar-
thin the slurry, are available to complement gins between these conflicting constraints
liquid retarders. are narrow and careful analysis is required to
Latex, for example, is used as an additive to plan for success. Implementation of the
control fluid loss and rheology in the slurry. HPHT well construction program requires
In cement it improves tensile strength and skilled personnel, guided by established pro-
prevents gas migration. Latex, supplied as a cedures, with accurate information on sur-
liquid of finely dispersed particles in water, face parameters. Software models and
may be needed as an additive at around downhole measurements help complete the
13% of the slurry volume. picture by predicting downhole fluid behav-
ior with enough accuracy to permit safe
drilling. With all these factors in place,
increasingly challenging HPHT wells will
continue to be successfully and safely con-
structed in the future. —DH,RR,RH
49