Barite Sag: Measurement, Modeling,

and Management

P.A. Bern, SPE, BP-Amoco; Eric van Oort, SPE, and Beatrice Neustadt, SPE, Shell; Hege Ebeltoft, SPE, Statoil;

Christian Zurdo, SPE, Elf; and Mario Zamora, SPE, and K.S. Slater, SPE, M-I Drilling Fluids

Summary

A joint industry project was established to study barite sag mecha-
nisms and to develop field guidelines to manage the conse-
quences. A simple empirical model was developed to compare sag
potential for a wide range of fluid types. In the study, physical
properties of the mud, wellbore conditions, and characteristics of
the weighting material were shown to have a large influence on
sag behavior. The study also included direct measurements of the
properties of settled weight-material beds. These results provide
new insight into the mechanisms of barite sag and how best to
manage problems in the field.

Data from the tests clearly demonstrate that the parameters af-

fecting sag are interrelated and seldom act in isolation. For all
muds tested, the highest sag occurred at low annular velocities
over angles from 60 to 75°. Drillpipe rotation was particularly
beneficial in minimizing barite settlement. Rotation also assisted
in re-distributing barite deposits formed on the low side of the
hole.

The improved understanding of the mechanisms of barite sag

enabled development of practical field guidelines. Case history
studies presented in the paper demonstrate how the results of the
work together with better field monitoring have been successfully
applied to manage the effects of barite sag in high-pressure/high-
temperature and extended-reach drilling operations.

Introduction

Barite sag is the undesirable fluctuations in mud weight that occur
due to downhole settling of the weighting agent. The problem has
been exacerbated by the increased frequency of high-angle wells
with the associated increase in particle settling rate which occurs
in inclined fluid columns

共Boycott effect

1

兲. Barite settlement, re-

ferred to as sag, occurs both statically and dynamically.

Operational consequences of barite sag can be severe. Potential

problems include mud-weight variations

共in and out兲, well-control

difficulties, downhole mud losses, induced wellbore instability,
and stuck pipe. The situation is particularly acute in high-angle
wells where the allowable mud weight window can be restricted
by mechanical wellbore stability considerations.

2

A special flow loop was used to quantify barite sag under simu-

lated field conditions of annular flow and drillpipe rotation. Tests
were conducted on 20 different fluids provided by project partici-
pants. Data were used to develop a simple empirical relationship
to predict sag behavior with time. Also, an existing sag model was
found that is helpful for correlating laboratory and field results.

Most importantly, results from the study were combined with

field observations to develop guidelines on minimizing barite sag.
Case histories are given where the operational guidelines have
been successfully employed to mitigate barite sag problems.

Laboratory Measurement of Sag Potential

Sag potential was measured in over 130 tests using the purpose-
built flow loop shown schematically in Fig. 1. It was shown pre-
viously that sag potential is directly related to the weight loss of
the circulating fluid in this flow loop.

3

Mud-weight measurements

were adjusted for temperature effects to accurately determine sag-

induced density changes. Additional investigations included the
effects of chemical- and clay-based rheology modifiers, fluid den-
sity, weighting agent density, particle size, and low-shear rheol-
ogy.

Test Protocols. Test parameters included angle

共0 to 90°兲, annu-

lar velocity

共25 to 200 ft/min兲, pipe rotation 共0 to 250 rpm兲, ec-

centricity

共0 to 0.8兲, and time. The standard test protocol shown

graphically in Fig. 2 was established to evaluate the combined
effects of these key factors under a common set of conditions.
Designed to observe multiple interactions among key parameters,
the protocol was based on seven time steps between which flow
rate and pipe rotation were changed systematically. Annular ve-
locities were recorded as nominal values. Other test protocols
were used on a few fluids to examine certain characteristics in
greater detail.

Flow rate and rotation were maximized during the first and last

time segments to stabilize the test fluid, clean up the flow loop,
and qualify each test. Segments 2 to 3 lowered flow rate in two
steps without rotation in order to induce bed formation. Segments
4 to 6 measured the potential for removing sag beds, first by
increasing pipe rotation, and then flow rate. All standard tests
were run at 0.8 eccentricity since preliminary tests showed less
response when the pipe was centralized.

Test Fluids. The 20 test fluids represented a variety of mud types,
formulations, weights, suppliers, and geographical sources. One
of the project goals was to select muds from active directional
wells

共⬎30°兲 using weighted muds 共⬎12 lbm/gal兲 in order to

provide immediate feedback to operations. No other stipulations
were made on field muds. Some laboratory-modified and
laboratory-prepared fluids were also tested. Physical properties of
all test fluids are listed in Tables 1 and 2.

Mechanical-Parameter Results. Fig. 3 shows results for Mud 6
using the standard protocol at four inclinations

共45, 60, 75, and

90°

兲. Circulating fluid density change 共corrected to 120°F兲 is plot-

ted vs. time. All fluids tested responded similarly although the
magnitudes varied.

Most, if not all, of the sag-bed formation occurred at low flow

rates and no rotation. Initial pipe rotation to 75 rpm consistently
had the greatest effect on removing the beds. Doubling the rotary
speed and then the annular velocity helped as expected, but to a
much lesser extent.

Time-segment 3 in Fig. 3, during which the sag was greatest,

most clearly demonstrates the effects of angle. For Mud 6, the
order of decreasing sag severity was 60, 45, 75, and 90°. While
each individual mud tested exhibited slightly different behavior,
the angle at which the maximum sag occurred was consistently in
the range 60 to 75°. This is consistent with previously reported
data.

3

Fluid-Parameter Results. Rheology, density, weight material,
and chemical treatments were the key fluid parameters investi-
gated. These parameters emerged from an earlier study as the
most significant variables affecting sag that can easily be con-
trolled in the field.

3

Fig. 4 illustrates the effect of low-shear rhe-

ology of Muds 7a to 7d, which were different fluids with similar
formulations. Higher low-shear-rate rheological properties sys-
tematically diminished sag tendency. The low-shear-rate rheologi-
cal parameter used for this study is based on the Fann data read-
ings at 6 and 3 rpm. The low-shear-rate yield point (

␶

LS

) is

defined by

Copyright © 2000 Society of Petroleum Engineers

This paper (SPE 62051) was revised for publication from paper SPE 47784, first presented
at the 1998 IADC/SPE Asia Pacific Drilling Technology Conference held in Jakarta, 7

–

9

September. Original manuscript received for review 7 September 1998. Revised manu-
script received 18 October 1999. Paper peer approved 21 December 1999.

SPE Drill. & Completion 15

共1兲, March 2000

1064-6671/2000/15

共1兲/25/6/$5.00⫹0.50

25

␶

LS

⫽共2⫻

␪

3

兲⫺

␪

6

.

共1兲

Other rheological parameters such as viscoelastic properties are
also believed to influence sag.

4,5

Steady-state rheological param-

eters were used in this laboratory study since these are readily
available at the rig site. Low-shear-rate yield points (

␶

LS

) for the

muds at 120°F were 4, 6, 8, and 11 lbf/100 ft

2

, respectively. Mud

7a sagged so severely that the barite bed avalanched at low-flow
rates, causing bed recirculation near the end of the test section.

The effects of fluid density can be seen in Fig. 5. The densities

of Muds 8a and 8e were 13.4 and 15.5 lbm/gal; values of

␶

LS

at

120°F were 9 and 8 lbf/100 ft

2

, respectively. Only results at 75°

inclinations are shown to simplify the graph. The results indicate
very similar trends with both muds although the higher-density
mud shows slightly higher sag.

Fig. 6 compares sag response of three different weight materi-

als in the same 18 lbm/gal oil-based mud

共Muds 9a to 9c兲. The

testing protocol of annular velocity and drillpipe rpm is shown in
Fig. 6. Specific gravity of the barite, hematite, and manganese
tetroxide materials were 4.2, 5.0, and 4.8, respectively. The data
show that particle mass is more important than specific gravity.
Hematite particles were the largest, while the manganese tetroxide
particles were the smallest. The unusual hematite response be-
tween 60 and 90 minutes was caused by severe slumping during
the long static period.

The comparison between chemical- and clay-based rheology

modifiers can be seen in Fig. 7. Mud 8b with the chemical modi-
fier had a

␶

LS

of 11 lbf/100 ft

2

; Mud 8c, which was viscosified

with a clay-based product, had a

␶

LS

of 10 lbf/100 ft

2

. These

limited data infer that the clay-based modifier was more effective
at reducing sag.

Sag-Bed Study. Physical properties and characteristics of depos-
ited sag beds were examined from samples taken from the sag

flow loop. Tests were conducted on a 17.6 lbm/gal oil-based mud
weighted with a hematite/barite blend. Preserved beds were ob-
tained from three window sections cut from the lower side of the
sag flow loop.

The in-situ sag beds had a pudding-like consistency. Attraction

among the inert particles was clearly weak, and beds were easily
fluidized by minor disturbances like those generated by logging
and tripping operations. This behavior may explain why:

䊉

Sag beds are more responsive to removal by velocity and pipe

rotation than most cuttings beds.

䊉

Sag beds tend to flow rather than slide when placed at angle.

䊉

Sag beds can still ‘‘slump’’ downwards at angles up to 75°,

about 10 to 15° greater than cuttings beds.

6

The flow loop sag beds contained more fine solids

共⬍6

␮兲 and

coarse solids

共⬎32

␮兲 than medium solids. The particle-size dis-

tribution of the mud was narrower. Average bed density was ap-
proximately 20 lbm/gal.

Modeling of Sag Results

It has been shown that the Sag Register is a suitable model to
monitor sag in the field.

7

The same approach, taken to normalize

and correlate test results, produced this slightly modified version:

Fig. 1–Schematic diagram of the dynamic sag flow loop.

Fig. 2–Annular velocity and pipe rotary speed schedule for
standard test protocol.

TABLE 1– DRILLING FLUID TYPES

Mud ID

Drilling Fluid

1

Synthetic base, 14.8 lbm/gal, North Sea

2

Synthetic base, 15.9 lbm/gal, North Sea

3

Diesel base, 16.6 lbm/gal, Louisiana

4

Lab-prepared calcium bromide drill-in fluid, 12.9
lbm/gal

5

Diesel base, 16.6 lbm/gas, Louisiana

6

Low-toxicity oil base, 14.2 lbm/gal, North Sea

7a

Synthetic base,15.3 lbm/gas, Gulf of Mexico

7b

7a treated to improve

␶

LS

7c

Synthetic base, laboratory reformulation of 7a to
improve

␶

LS

7d

Synthetic base, laboratory reformulation of 7a to
further improve

␶

LS

8a

Low-toxicity oil base, 13.4 lbm/gal, North Sea

8b

8b treated with base-oil, barite, and chemical-base
rheology modifier, 13.5 lbm/gal

8c

8a treated with base-oil, barite, and clay-base rheology
modifier, 13.4 lbm/gal

8d

8a reformulated to 16.3 lbm/gal

8e

8d reformulated to 15.5 lbm/gal

8f

8a treated with base-oil, barite, and clay-base
rheology modifier, 13.4 lbm/gal

8g

8a treated with base-oil, barite, wetting-agent and clay-
base rheology modifier, 13.4 lbm/gal

9a

Low-toxicity oil base, barite weighted, 18.0 lbm/gal,
laboratory prepared

9b

Low-toxicity oil base, hematite weighted, 18.1
lbm/gal, laboratory prepared

9c

Low-toxicity oil base, magnese tetroxide weighted,
18.1 lbm/gal, laboratory prepared

Mud A

Low-toxicity oil base, 11.6 lbm/gal

26

Bern et al.: Barite Sag

SPE Drill. & Completion, Vol. 15, No. 1, March 2000

1/R

st

⫽1/exp共50

*

⌬m

t

/m

兲.

共2兲

Data from a previous study

3

were analyzed to characterize Eq.

1, since there were no single-parameter tests included in the cur-
rent project. Fig. 8 includes plots of three sample 1/R

st

curves for

Mud A described in the reference. The properties of the mud are
listed in Table 2. The following model was found to be an excel-
lent match for the wide range of data:

1/R

s

⫽Y

0

⫹共1⫺Y

0

兲

*

共1⫺t兲

a

.

共3兲

‘‘Y

0

’’ and ‘‘a’’ have physical significance. The value of 1/R

s

after 1 hour of testing is ‘‘Y

0

’’

共assumed to be the equilibrium

value

兲, t is time in hours, and the exponent ‘‘a’’ defines the rate of

sag.

Model results superimposed on the data in Fig. 8 demonstrate

the excellent fit. Values of ‘‘Y

0

’’ increased with increasing pipe

rotation and velocity, indicating decreasing sag for increased flow
and pipe rotation. Plots of ‘‘Y

0

’’ vs. angle for Mud A demon-

strate maximum sag

共minimum ‘‘Y

0

’’

兲 occurs for angles in the

range 60 to 75°.

Operational Guidelines

One of the primary project goals was the development of field
guidelines for sag management and prevention. Guidelines were

established from results of this study, data in the public literature,
and field observations and experiences. The impact of barite sag
can be minimized by attention to detail at the planning and execu-
tion stages of a well. In particular, four key areas should be ad-
dressed: well planning, mud properties and testing, operational
practices, and wellsite monitoring procedures.

Well Planning. For critical wells, it is important to assess all
options available at the planning stage. In practice, barite sag is
just one of the many key issues that needs to be considered. The
overall well plan often requires some form of compromise to op-
timize the overall well objectives.

Well Type. The most critical wells have extended-reach geom-

etry where the margin between pore pressure and fracture gradient
can be small. In these wells, the maximum viscosity is limited by
the need to control equivalent circulating density

共ECD兲.

Well Environment. Temperature and pressure are critical for

drilling fluid design. High temperatures cause the drilling fluid to
thin, which can increase sag tendency. It is important to ensure
that viscosity measurements are taken under high-pressure/high-
temperature

共HP/HT兲 conditions. For critical wells, consider mak-

ing HP/HT viscosity measurements at the wellsite.

Angle and Well Profile. Sag can occur in wells drilled with

inclinations

⬎30°. The most critical region for sag is 60 to 75°.

Casing Design. If possible, avoid casing designs which give

rise to low annular velocities in deviated sections. In cases where
a 6 in.

共152.4 mm兲 hole is drilled through a 7 in. 共177.8 mm兲

drilling liner, recognize the increased potential for barite sag to
occur in the larger, previous casing/drillpipe annulus due to low
annular velocities.

Hole Diameter. Barite sag has been shown to occur over a

wide range of annular velocities and well angles. Most problems

have occurred in 12

1
4

in.

共311.2 mm兲, 8

1
2

in.

共215.9 mm兲, and 6 in.

共152.4 mm兲 sections.

Fig. 3–Mechanical parameter sag results for Mud 6.

Fig. 4–Effect of low-shear-rate yield point on sag.

Fig. 5–Effect of fluid density on sag.

TABLE 2– DRILLING FLUID RHEOLOGICAL PROPERTIES

Mud

ID

MW

(lbm/gal)

Oil/Water

PV

(cp)

YP

(lbf/100 ft

2

)

␶

LS

at

150°F

(lbf/100 ft

2

)

1

14.8

67/33

29

22

6

2

15.9

75/25

53

26

6

3

16.6

86/14

48

11

2

4

12.9

¯

10

5

0

5

17.1

80/20

35

20

12

6

14.2

65/35

42

24

8

7a

15.3

80/20

22

8

3

7b

15.34

81/19

27

11

5

7c

14.84

78/22

28

14

7

7d

15.27

78/22

28

23

10

8a

13.35

84/16

28

20

10

8b

13.49

93/7

19

19

11

8c

13.38

93/7

21

16

8

8d

16.34

86/14

35

26

11

8e

15.52

87/13

28

15

7

8f

13.35

86/14

8g

13.35

86/14

24

23

10

9a

18.02

79/21

43

12

5

9b

18.14

77/23

28

7

1

9c

18.08

78/22

28

27

10

Mud A

11.6

78/22

12

8

3

Bern et al.: Barite Sag

SPE Drill. & Completion, Vol. 15, No. 1, March 2000

27

Mud Properties and Testing. Correct choice of fluid properties
is vital for optimizing drilling performance. It is also important
that appropriate screening tests, such as viscometer sag tests,

8

are

carried out at the pre-drilling phase to ensure drilling fluids remain
stable under all conditions encountered within the well.

Drilling Fluid Type. Experience from field operations is that

sag can occur with all fluids: oil-based, synthetic-based, and
water-based.

Drilling Fluid Density. Field experience has shown that barite

sag can occur over a relatively wide density range

共12 to 20 lbm/

gal, 1.44 to 2.40 sg

兲.

Rheology. Increasing low-shear viscosity

共Fann 6 and 3 rpm

readings, and gels

兲 normally helps reduce sag. There is evidence

which suggest that ‘‘clay’’-type products are more effective than
fatty acids as low-shear enhancers. Sufficient free water must be
available for rheology modifiers to function.

Yield Stress. A minimum drilling fluid yield stress is required

to provide adequate barite suspension. Low-shear-rate yield point
(

␶

LS

) is generally a good indicator of yield stress for most drilling

fluids.

9

Based on available data, the minimum

␶

LS

should be in the

range of 7 to 15 lbf/100 ft

2

共3.4 to 7.2 Pa兲.

3

Increasing oil or

synthetic content tends to thin the drilling fluid and can increase
sag tendency. It is important that the clay content or rheology
modifier concentration be increased to compensate for any loss in
viscosity.

Surfactant Concentration. Levels of wetting agent in non-

aqueous drilling fluids must be sufficient to prevent barite from
agglomerating into large clusters. Overtreatment must be avoided
to prevent undesirable viscosity reduction. Sufficient free water
must be available for wetting agents to function.

Fluid-Loss Additives. Field problems have been experienced

with adverse reactions between certain fluid-loss additives and
drilling fluid viscosity. This reinforces the need for pilot testing to
assess specific drilling fluid formulations and interactions.

Weight-Material Selection. Tests demonstrate that sag depends

upon the mass of the weight material. It is possible to control sag
by using smaller particles, or particles of lower density. In water-
based drilling fluids, there is a greater tendency for smaller par-
ticles to agglomerate. Different types of weight material may have
different suspension characteristics.

Particle-Size Control. A broad particle-size distribution can

help minimize sag. Very large particles should be eliminated, as
these can settle naturally under static conditions.

10

Avoid exces-

sive processing with solids-control equipment that can lead to an
undesirable

共narrow兲 particle-size distribution.

Operational Practices. The success or failure of a drilling opera-
tion is governed by wellsite actions. There are many parallels
between managing barite sag and eliminating stuck pipe. Rapid
and appropriate action by all rig crew members is vital.

11

Rotary vs. Sliding. Data clearly demonstrate that sag is worst

when the drillpipe is stationary and eccentric, particularly for
wells in the range 60 to 75°, where even high annular velocities
can have difficulty preventing sag. While sliding, periodically use
rotary wiper trips to stir up any deposited barite beds.

Time Between Trips. Recognize that sag increases with time.

Beds formed at lower angles

共⬍60°兲 tend to slump more and

cause greater problems.

Staging in Hole. Staging into the hole can reduce the associ-

ated problems if sag occurs. Drilling fluid should be circulated at
each staging depth until the density stabilizes. If the drilling fluid
density is stable, circulate a minimum time of 1.5 bottoms-up.

Conditioning Drilling Fluid. If excessive density swings are

observed at surface, stop and condition the fluid. Allow a mini-
mum of two total circulations to provide effective chemical treat-
ments and ensure all density fluctuations are eliminated. Avoid
overtreating the drilling fluid prior to running casing and cement-
ing. Excessive dilution can dramatically increase sag tendency.

Operations with Low Shear. Recognize that barite sag is pre-

dominantly a dynamic settling phenomenon which occurs under
conditions of low shear rate

共e.g., low circulation rates, running

casing, or running logs

兲. Any operations that induce low shear can

accelerate sag. Ensure proper drilling fluid condition, particularly
for extended low-shear operations.

Sag-Bed Properties. Recognize that sag beds are very different

from cuttings beds. Sag beds are more fluid than cutting beds, and
tend to flow rather than slide. The fluidized nature of sag beds
allows them to be readily dispersed with a combination of high-
flow rates and drillpipe rotation. In laboratory testing, static sag
beds exhibited fragile gel strengths and were easily fluidized.

Wellsite Monitoring Procedures. For critical wells, it is vital
that all rig crew members are vigilant. Careful observation and
rapid communication of any deviations from expected behavior

Fig. 6–Influence of weight materials on sag.

Fig. 7–Comparison of sag tendency with clay and chemical vis-
cosifiers.

Fig. 8–Sample plots of experimental data and modeling results.

28

Bern et al.: Barite Sag

SPE Drill. & Completion, Vol. 15, No. 1, March 2000

can increase chances of success. It is also important that appro-
priate wellsite measurement methods are established and reported.

Drilling Fluid Density. At critical stages, ensure drilling fluid

density

共in and out兲 are measured at 15 minute intervals while

circulating bottoms-up after a trip. Always record mud tempera-
ture and correct the density for temperature. Use a pressurized
drilling fluid balance to obtain accurate data. Ensure the balance is
operating properly and recently calibrated.

Standpipe Pressure. Carefully monitor trends in standpipe

pressure. Fluctuations may occur as slugs of light and heavy drill-
ing fluid pass through the bit nozzles. Changes can occur from
both hydrostatic

共U-tube effect兲 and differences in frictional pres-

sure loss.

Torque and Drag. High torque and overpulls may indicate that

a barite bed is forming a restriction on the low side of the hole.

Drilling Fluid Losses and Gains. Ensure that active pit vol-

umes are monitored accurately. Unexpected losses may occur as
heavy spots of drilling fluid in the annulus reach near-vertical
sections of the well and rapidly increase hydrostatic pressure. The
opposite effect can occur with light drilling fluid, which may
cause the well to flow.

Pressure Measurements. Consider using real-time pressure

measurements to determine true downhole density. The tool
should be configured to give pumps-off data on re-initiating flow.
Analysis of stored data is very useful for post-well analysis.

General Guidelines. The following are valuable aids to control
and monitor sag while executing the well plan.

Sag Register. Sag Register, an exponential ratio of density

change, is a proven measure of sag tendency in the field and
laboratory.

Sag Time Dependency. Sag behavior can be characterized by

two parameters. One describes the rate of sag, while the other
describes the ultimate extent of the sag. No simple correlation
exists between drilling fluid properties and these characteristic
parameters at this time. Sag beds can be laid down very quickly.

Sag-Bed Removal. Deposited sag beds can be removed by

higher annular velocities

共⬎100 ft/min兲 combined with drillpipe

rotation

共⬎50 rpm兲 to re-disperse weight material in the main flow

stream of the drilling fluid. The flow rate and rotation should be
maintained until drilling fluid weight stabilizes.

Drilling Fluid Thixotropy. Rapid development of gel strength

can reduce sag. Since the time to make and break gels is different,
shear history can be important. The order in which annular veloc-
ity and rotation vary can significantly influence sag behavior.

Case Histories of Sag Management

Well A. This was a HP/HT well with a 17.9 lbm/gal oil-based
mud. The well was S shaped with a tangent angle of 40°. Sag was
experienced which resulted in lost circulation problems when
drilling on bottom. Samples taken at the shakers showed that the
mud had sagged with a difference in density ranging from a mini-
mum of 16.6 lbm/gal to maximum of 20.7 lbm/gal.

Fig. 9 shows the lagged mud density profile determined from

samples taken at 500 m intervals. The minimum density coincides
with the start of the high-angle section

共approximately 40°兲. The

maximum density occurs above the bottom of the well and is
approximately 500 m above the drop-off point.

During the drilling of the remainder of the well, careful moni-

toring of the lagged mud weight was used to determine the extent
of the sag. Clay-based gelling agents were used to control the
level of sag. The improvement in fluid quality with time is dem-
onstrated in Fig. 10. These data show the degree of sag

共normal-

ized for static periods without circulating

兲 as a function of time.

The data demonstrate that sag was not totally eliminated, but was
managed down to a lower level. This enabled the well to be
drilled successfully.

Well B. This well was drilled with a 13.2 lbm/gal oil-based fluid
through the 8

1
2

in. interval at a maximum angle of 60°. Tripping

into the hole after the well had been static for 5 days showed that
the mud had sagged. Fluid circulated at the previous casing shoe
showed variations between 12.7 and 13.7 lbm/gal. Running into
open hole was tight and mud losses were encountered.

The drilling fluid was treated with low-shear rheology modifier.

This increased

␶

LS

from 8 to 11 lbf/100 ft

2

. The decision to use

the low-shear enhancer rather than clay was taken to avoid large
increases in viscosity and, hence, ECD. However, the low-shear
enhancer was ineffective at reducing a sag and lost circulation
continued throughout the remaining drilling phase. Further losses
were encountered when running the 7 in. liner. Mud-weight varia-
tions in the partial returns showed the same fluctuations between
12.7 and 13.7 lbm/gal.

The interpretation of events was that the initial concentration of

oganophillic clay was too low. The addition of low-shear en-
hancer was not effective at stabilizing the mud properties and
sagging continued. The degree of sagging was probably exacer-
bated by the low-shear-rate regime experienced while running
casing. Subsequent experience has shown that increasing the clay
concentration is more likely to be successful in reducing sag. This
has now been adopted as the preferred option to combat sag.
Increases in ECD have been successfully managed by slight re-
ductions in circulation rate.

Well C. This was an extended-reach well drilled using a 15.0
lbm/gal synthetic-based fluid in the 8

1
2

in. section. The well was S

shaped with a maximum angle of approximately 60°. Sag was
initially observed on a trip back into the hole with return mud
weights varying from 14.6 to 15.7 lbm/gal when circulating at the
9

5
8

in. casing shoe.

The normal procedure adopted on this operation for re-

distribution settled barite was to ream down at normal circulating
rates of 400 gpm. However, pressure while drilling data taken
earlier in the well showed that the normal reaming rate of 5 min
added an additional 250 psi surge pressure in the annulus. Similar
increases in surge pressure have been observed in the field using
pressure while drilling measurements.

12

Fig. 9–Sag profile for Well A.

Fig. 10–Normalized sag behavior for Well A.

Bern et al.: Barite Sag

SPE Drill. & Completion, Vol. 15, No. 1, March 2000

29

Based on the assumption that the mud weight in the annulus

may have increased to above 15.8 lbm/gal, it was decided that
there was a high risk of inducing losses if the normal reaming
down procedure was adopted.

Two options were considered to mitigate the risk of losses. The

first option was to reduce the circulating from 400 to 300 gpm in
an attempt to reduce ECD. The second option was to maintain the
initial circulation rate, but to control the reaming down rate to
minimize the associated pressure surges.

The option of reducing flow rate to 300 gpm would have led to

an annular velocity in the 10

3
4

in. casing of below 110 ft/min.

This low velocity was considered as likely to contribute to the sag
problem. The second option of staging in the hole at a controlled
reaming rate of 1 stand per hour was adopted. Reaming was con-
tinued for a period of 1 to 1.5 bottoms-up to ensure that the mud
weight in the annulus had fully stabilized. Each period of reaming
was followed by running-in 10 stands at a controlled rate of 2
min/stand. This process was repeated until the bit was on bottom
and drilling ahead recommenced.

Throughout the staging-in process, mud weights were mea-

sured every 10 minutes and pits were carefully monitored for
losses. Pipe rotation and circulation during reaming was effective
at re-distributing settled barite beds and stabilizing the mud
weight. The procedure allowed the remainder of the section to be
drilled without any further problems.

Conclusions

䊉

In practice, barite sag can be minimized by attention to detail

in the following key areas: well planning, mud properties, opera-
tional practices, and wellsite monitoring.

䊉

Barite sag and hole cleaning are related in principle, but are

distinguished by their bed characteristics.

䊉

Barite beds are more responsive to removal by mud velocity

and pipe rotation than most cuttings beds.

䊉

Significant barite sag was measured at angles as high as 75°.

The most critical range was 60 to 75°.

䊉

Barite sag can be exacerbated by low annular velocities, and

eccentric and stationary drillpipe.

䊉

In some cases, sag reduction can be more sensitive to drill-

string rotation than annular velocity.

䊉

Sag behavior with time can be predicted with a simple em-

pirical equation.

䊉

With time, an equilibrium is established between dynamic

barite deposition and bed erosion.

Nomenclature

R

st

⫽ modified Sag Register for experimental data

R

s

⫽ modeled modified Sag Register

m

⫽ initial mud weight, M/L

3

, lbm/gal

⌬m

t

⫽ m-mud weight at time t, M/L

3

, lbm/gal

Y

0

⫽ value of R

st

after 1 hour of testing

t

⫽ time, T, hours

a

⫽ model parameter describing sag rate

␶

LS

⫽ low-shear-rate yield point, M/LT

2

, lbf/100 ft

2

␪

3

⫽ Fann 3 rpm dial reading, M/LT

2

, lbf/100 ft

2

␪

6

⫽ Fann 6 rpm dial reading, M/LT

2

, lbf/100 ft

2

Acknowledgments

The authors thank their respective companies for allowing publi-
cation of the work. We are also grateful to Sanjit Roy of M-I
L.L.C. for his valuable assistance in conducting the laboratory
studies and analyzing the raw data.

References

1. Boycott, A.E.: ‘‘Sedimentation of Blood Corpuscles,’’ Nature (Lon-

don)

共1920兲 104, 532.

2. McLean, M.R. and Addis, M.A.: ‘‘Wellbore Stability: The Effect of

Strength Criteria on Mud Weight Recommendations,’’ paper SPE
20405 presented at the 1996 SPE Annual Technical Conference and
Exhibition, New Orleans, 23–26 September.

3. Bern, P.A. et al.: ‘‘The Influence of Drilling Variables on Barite

Sag,’’ paper SPE 36670 presented at the 1996 SPE Annual Technical
Conference and Exhibition, Denver, Colorado, 6–9 October.

4. Saasen, A. et al.: ‘‘Monitoring of Barite Sag Important in Deviated

Drilling,’’ Oil & Gas J.

共1991兲 89, No. 34, 43.

5. Sassen, A. et al.: ‘‘Prediction of Barite Sag Potential of Drilling Flu-

ids from Rheological Measurements,’’ paper SPE/IADC 29410 pre-
sented at the 1995 SPE/IADC Drilling Conference, Amsterdam, 28
February–2 March.

6. Seeberger, M.H. et al.: ‘‘Oil Muds in Large-Diameter, Highly Devi-

ated Wells: Solving the Cuttings Removal Problem,’’ paper SPE/
IADC 18635 presented at the 1989 SPE/IADC Drilling Conference,
New Orleans, 28 February–3 March.

7. Zamora, M. and Jefferson, D.T.: ‘‘Controlling Barite Sag Can Reduce

Drilling Problems,’’ Oil & Gas J.

共1994兲 92, No. 7, 47.

8. Jefferson, D.T.: ‘‘New Procedure Helps Monitor Sag in the Field,’’

paper ASME 91-PET-3 presented at the 1991 Energy Sources Tech-
nology Conference and Exhibition, New Orleans, 20–24 January.

9. Hemphill, T.: ‘‘New Viscometer Allows for Improved Understanding

of Drilling Fluid Behavior at Low-Shear Rates,’’ presented at the
1999 AADE Annual Technical Forum, Houston, 30–31 March.

10. Alfenore, J. et al.: ‘‘Selecting the Optimum Fluid for HPHT Condi-

tions,’’ presented at the 1996 HP/HT Going into Production Confer-
ence, London, 18–19 June.

11. Bradley, W.B. et al.: ‘‘A Task Force Approach to Reducing Stuck

Pipe Costs,’’ paper SPE/IADC 21999 presented at the 1991 SPE/
IADC Drilling Conference, Amsterdam, 11–14 March.

12. Swanson, B.W. et al.: ‘‘Measurement of Hydrostatic and Hydraulic

Pressure Changes During HPHT Drilling on Erskine Field,’’ paper
SPE 38480 presented at the 1997 Offshore Europe Conference, Ab-
erdeen, 9–12 September.

SI Metric Conversion Factors

ft

⫻ 3.048

*

E

⫺01 ⫽ m

ft/min

⫻ 5.08

*

E

⫺03 ⫽ m/s

gal/min

⫻ 6.309 020

E

⫺05 ⫽ m

3

/s

in.

⫻ 2.54

*

E

⫺02 ⫽ m

lbf/100 ft

2

⫻ 4.788 026

E

⫺01 ⫽ Pa

lbf/in.

2

共psi兲 ⫻ 6.894 757

E

⫹03 ⫽ Pa

cp

⫻ 1.0

*

E

⫺03 ⫽ Pa

•s

lbm/gal

⫻ 1.198 264

E

⫹02 ⫽ kg/m

3

*

Conversion factors are exact.

SPEDC

Peter A. Bern is a senior engineer with BP-Amoco’s Upstream
Technology Group in Sunbury-on-Thames, UK. He has worked
in a number of areas related to fluids technology. Bern holds a
degree in physics from the U. of Bristol, UK. Eric van Oort is a
staff research scientist with Shell E&P Technology Applications
and Research. He acts as the coordinator for Shell’s Global
Hole Stability Team. van Oort holds a PhD degree in physical
chemistry from the U. of Amsterdam, The Netherlands. Beat-
rice Neustadt is a production technologist with Shell E&P Tech-
nology Application and Research Center in Rijswijk, The Neth-
erlands. She works in the area of drilling and completion fluids.
Neustadt holds a PhD degree in chemistry from Halle U., Ger-
many. Hege Ebeltoft is a senior staff engineer in Statoil’s Well
Construction Group in Stavanger. She works in areas related
to fluids and wellbore stability. Ebeltoft holds a PhD degree in
physical chemistry. Christian Zurdo is Drilling and Completion
Manager with Elf-Petroland in The Hague. He served on the
1991–92 Forum Series in North America Steering Committee.
Mario Zamora is Manager Applied Engineering for M-I Drilling
Fluids in Houston. He has broad experience in drilling, drilling
fluids, and computer technology. Zamora holds a BS degree
from the U. of Texas. Currently a member of the Drilling Con-
ference Program Committee, he has served on several SPE
committees including the Publications Coordinating Commit-
tee and the Editorial Review Board. Kenneth Slater is Team
Leader for M-I Drilling’s Applied Engineering Laboratory in
Houston. He works on projects related to fluids and fluids test-
ing. Slater holds an MS degree in mechanical engineering
from the U. of Houston.

30

Bern et al.: Barite Sag

SPE Drill. & Completion, Vol. 15, No. 1, March 2000