- Email: [email protected]
Lubrication of water-based clay suspensions
Tribology Research: From Model Experiment to Industrial Problem G. Dalmaz et al. (Editors) 9 2001 Elsevier Science B.V. All rights reserved.
331
Lubrication of Water-Based Clay Suspensions B. J. Briscoe a, p. M. Cann b, A. Delfino a, G. Maitland c ~Department of Chemical Engineering, Imperial College, London, UK bDepartment of Mechanical Engineering, Imperial College, London, UK cSchlumberger Cambridge Research, Cambridge, UK This paper describes preliminary studies of the lubricating behaviour of a simple water-based rock drilling fluid (bentonite clay suspension) in metal contacts. Friction measurements have been carried out on a modified tribometer (Zircon model, Glacier Metals, UK) with a model contact between a rotating shaft and a loaded planar counterface. These experiments were designed to investigate the tribology of the contact between the drillstring and the metal wall of an oilwell. Measurements are reported for a range of loads and contact velocities and for different compositions of the fluid. The result are presented in the form of classical Stribeck-Hersey curves in order to identify the lubrication regime and to illustrate the combined effects of load and speed on the friction coefficient. Optical interferometry measurements have also been carried out, using a model counterface of a metal sphere upon an optical plate, in order to visualise the flow of the suspension through the contact and also to measure the film thickness and the film thickness profiles. Two basic lubrication regimes are identified: at high loads a regime characterised by the deposition of layers of solid clay onto the contacting surfaces and at low loads a regime in which the main lubricating action is provided by the base fluid. In the transition between the two regimes, an intermediate region is found to be characterised by changes in the fluid composition and rheology within the contact. The general trend of the Stribeck curve is obtained and a peculiar scattering of the data is evident in the region between the boundary lubrication regime and the mixed lubrication region. The intrinsic nature and the complex rheology of the fluid appear to be the parameters that may control this effect and in part define the lubrication regime.
1. T h e Frictional P r o b l e m in Oil Drilling In oil well technology special fluids are used during drilling, completion and t r e a t m e n t operations. These fluids are in general complex multicomponent dispersions or solutions and they are required to fulfill many different roles.[1-3]. For example, pressure control in the wellbore, lubrication and cooling of the drill bit, removal of the cuttings from downhole locations and prevention of blowouts. These fluids are formulated to have carefully controlled rheological and tribological properties as these have a profound effect upon the overall drilling and well operation [1,4,5]. In modern drilling technology it is desirable to improve the conventional lubricating action of the mud since increasingly severe conditions are encontered in extended-reach (directional) oil wells. It is generally accepted t h a t torque and drag forces in directional wells are primarily caused
by sliding friction between the drillstring and either the metallic casing or the drilled strata. The friction coefficient is defined as the ratio of the frictional force and the sidewall normal contact force. Friction coefficients can be estimated from field measurements and are generally in the range 0 . 1 - 0.4 [6]. In drilling application it would be desirable to reduce the friction in order to reduce the energy dissipation and the wear of the drillstring, as well as the prospect of drillstring seizure and breakage. This paper presents the preliminary results of an investigation of the lubricating behaviour of water-based drilling muds in metal contacts. It also a t t e m p t s to clarify the lubricating role of the mud in the contact between the metal drillstring and the wall sections of an extended-reach deviated wellbore. The study has been carried out experimentally using a typical tribometer (Zircon Model, Glacier Metals, UK),
332
modified to give a model lubricated contact between a rotating shaft and a housing. The friction has been measured over a range of rotational velocities and normal loads for different mud compositions which reflect, in part, well bore environments. An optical interferometry technique has been used to measure film thickness and to visualise the flow of the lubricant through the contact in order to provide a better understanding of the lubrication mechanism.
~~.td
IN
Exte:.~ed-Reach
0il Well
2. T h e Drilling P r o c e s s and Drilling Fluids Drilling fluids serve many different purposes [13]. The ones concerned with the present work are: 9 lubricating (and cooling) of the drillstring and the bit to minimise driving torque, drag and pipe sticking and rupture. 9 cleaning the bottom of the hole and carrying the rock debris to the surface. 9 in general maintaining the integrity of the hole, drillstring, casing and tubing. 9 providing an hydrostatic pressure that prevents blowout and the release of gases during well operations.
2.1. Friction in D i r e c t i o n a l Wells A directional well (figure 1) is one in which a significant deviation in the drilling path is deliberately initiated in order to reach particular zones in the formation. In certain situations it is the only practical and economic way of reaching the production area. For example, in multiple offshore drilling or when it is desired to reach inaccesible locations or to increase the exploitation of a reservoir. In deviated wells, as these wells are termed, drag and torque forces tend to be more troublesome and in deep and highly deviated wells the reduction of those forces can be critical in order to achieve satisfactory well completion. 2.2. T y p e s of Drilling Fluids Drilling muds can be defined as a suspension of solids in a liquid phase where the liquid can be water or oil. It is generally possible to distinguish
Figure 1. Schematic of an extended-reach deviated oil well between three main types of drilling fluids: waterbased muds, oil-based muds and emulsion muds. This study focuses mainly on water-based fluids as they are the most widely used at present. Water based drilling fluids contain at least four main components: the continuous phase (water), the dispersed colloidal phase (bentonite clay), the chemical components (additives, polymers) and the inert components (weighting materials). The continuous phase provides the initial viscosity and the medium into which the other components are dispersed. The colloidal phase enhances viscosity and develops a yield point through the absorption of water. The chemical components control the relevant physical properties of the mud and the inert phase provides high gravity solids to increase the mud weight. The polymers are usually selected from a wide range of products such as carboxymethylcellulose and hydroxyethylcellulose while the weight materials are generally barite, galena, iron oxides or limestone. The composition is variable and the clay content is generally 3 - 7% by weight. The rheological properties of the mud are very important as they influence the frictional pressure
333
Mud Cake
Dri11pipe
Mud
provide useful insights into the problem. In this paper two fundamental aspects will be considered: the measurement of friction (friction reduction is the most important aspect in the drilling application) and the mechanism of lubrication and the nature of the film formed.
Drillstring
3. Materials ments
Figure 2. Schematic of a section of the well hole drop, torque and pipe sticking, cutting transport and the conditions at the boundary of the mudcake. 2.3. A i m of the S t u d y The development of deviated wells presents severe technical problems and increasingly it is becoming important to optimise the lubrication properties of the drilling fluids. In particular it is highly desirable to be able to improve the lubricity of the drilling fluid without modifying other physical properties (viscosity, degree of flocculation etc.) that are required for other fundamental roles such as the transport of cuttings up to the surface. The aim of this work is to study and characterise the lubricating behaviour of water-based drilling fluids and to try to improve their lubricating effect. The system considered is illustrated in figure 2. The drillstring rotates inside the well hole which is filled with mud and contacts the walls of the well. Realistic simulations of the real contact problem are difficult due to the scale and the complexity and variability of the problem (types of drilled rocks, types of different drillstrings etc). The contact conditions are highly variable but the system can be basically considered as a pure sliding, highly loaded contact. This paper therefore has focused upon studying the basic tribological properties of the mud rather than the complete system. Nevertheless, it is reasonable to believe that even in this way the study will be able to
and E x p e r i m e n t a l Arrange-
The mud was prepared by dispersing a common type of bentonite clay (Whitegel, IK) in water (2% ,3% and 5% clay weight fraction). The clay was slowly added to the water while mixing with a high-shear rate mixer (Silverstone, UK). At least 45 minutes of mixing were necessary to obtain a well-dispersed and reproducible suspension. The mud was then allowed to swell for 24 hours before use.
3.1. Zircon T r i b o m e t e r The frictional behaviour of water-based drilling fluids in metal contacts was studied using the apparatus shown schematically in figure 3. It was a typical tribometer (Glaciers-Metals, Zircon model) that was modified in order to obtain a lubricated contact between the shaft and a planar metal counterface. Mild steel samples (4 x 4 cm) were used for the counterface and were loaded against a 25 mm diameter rotating mild steel shaft. The normal load was applied by means of dead weights ranging from 0.05 kg to 0.8 kg (maximum Hertzian pressure ~ 100 MPa) and the speed of the shaft was 0.2 to 0.9 m/s ( 8 0 - 360 rpm ). A piezo-electric force tranducer (Kistler, Germany) was attached to the sample holder in order to record the frictional force after suitable calibration. A plastic box (made of commercial PMMA) was fitted with suitable seals onto the shaft in order to contain the fluid. In this arrangement, the shaft could be submerged in the mud providing a nominally completely flooded contact. The fluid reservoir was equipped with a stirrer in order to control the shear history of the mud and investigate its effect upon the lubrication process; this part of the study is not described.
334
co
Zircon Tribometer
mpeccromecer
load
Arm
\
glamm dimk
Transducer
~ ~ ~
f Metal Sample
alur~um
/ miliaa
1ubz-lom~e film B(:ee2 J~t22
.Z4yez"
I
kPropeller ~o~~t~ng Shaft
Figure 3. Schematic of the Zircon tribometer 3.2. C a l i b r a t i o n The transducer was calibrated attaching a metal wire going through a pulley so that dead weights could be used. Different settings of the sensitivity of the transducer were adopeted in order to encompass the appropriate ranges for the forces to be measured. 3.3. O p t i c a l I n t e r f e r o m e t r y The film thickness measurements were carried out using a modified optical interferometry techinque [7]. A diagram is shown in figure 4. The contact was formed by a steel ball rolling against a glass disc. The underside of the disc was coated with a chromium layer overlaid with silica. Monochromatic or white light was shone into the contact through the glass disc. Some of the light reflects off the surface whilst some passes through the lubricant film in the contact to be reflected back off the other surface. The two reflected beams then optically interfere to an extent dependent on the difference in path length, and thus the lubricant film thickness, to produce an interference pattern. A microscope coupled to a CCD video camera was positioned directly above the contact. The camera was connected to a capture board so that images from the contact could be taken and stored for analysis. This technique produced a contour map of the contact film thickness and previous studies [8] have established a colour-film thickenss calibration that was used in this study.
\z
Figure 4. Optical interferometry techinque for film thickness measurement The preliminary optical studies consisted of testing 5% mud in both rolling and sliding conditions. The load was 20 N and the velocities ranged from 0.025 to 0.4m/s. Images were captured during the test in order to visualise the flow of the suspension through the contact and film build up. Also, from the image data, it was possible to estimate the central film thickness. 4. R e s u l t s and D i s c u s s i o n 4.1. Friction M e a s u r e m e n t s The frictional force between the shaft and the counterface has been measured for a range of loads and rotating speeds. The frictional force in the contact was measured as a function of time. Signal noise that arose from uncertain sources were removed before any further analysis was carried out; also significant spurious vibrations were introduced by the drive system. The data were then averaged to obtain the mean value of the frictional force and to calculate the friction coefficient. Data is presented for 3 different mud compositions and the friction coefficient is plotted against the normal load W for different speeds (see figures 5, 6 and 7). The range of values of the friction coefficient is very high (generally between 0.05 and 0.4), suggesting that the contact works in the so-called mixed lubrication regime and for the highest loads in the boundary lubrication regime. As is normally expected in lubricated metal contacts, the
335
friction coefficient increases with increasing load and with decreasing sliding velocity. There seems to be a "transition" in the trend of the data at about 1 . 5 - 2 N between the two regimes. It can be noted that, for loads higher that 1.5 N, the friction coefficient is more affected by the change in the sliding velocity. A similar trend was found for the other compositions. With the more dilute muds, the friction coefficient was higher, indicating a less efficient lubricating action. If pure water was used, a decrease in the friction coefficient was observed but the wear of the surfaces was found to be substantially higher than for mud-lubricated contacts. This was probably be due to the presence of a solid layer of clay deposited onto the surface (especially for higher loads) that protected the contacting surfaces. In this case the friction phenomena is found to be dependent upon the solid contact of clay particles and the friction coefficient tends to be higher probably due to an increase of the area of solid contact [9,10]. More work on this important aspect of the problem is currently been carried out.
Friction Coefficient for 3% Mud 0.45 0.4
, .....................................................................................................................
+i~
0.3
......................................................
~:
0.25
+
.~_,,
0.2
~ ......................
+ ......
0.1
o.os
' i
0 0
0.4
..................................
0.35
03
"~
0.25
~
" )
+
....
~ 0.2
'
~
i
a
1
2
" ~
DI ~
- .............................
- ............................... - .....................................................
)
................
3
. . . . . . . .
i
i
4
5
6
Friction Coefficient for 5% Mud '
0.4
'
'
................................................................................
035
i
...............................................................................................
0.3
8
..................................................
"
+!
El
....
i
....
-
.
.
.
.
.
.
.
x:~
x ..
i
i• ,, i ! ~ ........................ "........................................................
9 ...............................
~3
~+ .............................~ : ...........................Oi ~ ........................................................
0
...........;~i
................
x
v =m0/.S5v3= om. 7/ s
i ......................... ; : .......... ~ .................
.....
. . . . . . . . . .
0.25
+
v=0.37 m/s
! ..................................................
.....................................................................................................
'0
v = 0 . 2 1 ,,,'Is
i
. ............................... -:"......................................................................................
i
i
i
1
2
.
.
.
.
.
.
.
.
.
'
---
i
3
4
5
6
--
0
........................
Figure 7. Friction Coefficient for 5% Mud
.............................
i
0 []
v=O 3 7 m / s Xv=O 53 m/s v=0.7 m/s
........................................................
:; ..........................
W
:
.............................................................................................
8
:
........................................................................................................................... +
;E=
.......... x !
.................................."'+'" ................ i .................~:
X
5
~ X~:
+ x Y~
o-
................... ~ ..................................................
0.45
0
,,,',,,,Is
n
Figure 6. Friction Coefficient for 3 % Mud
o +v=0.21
. []
W
005
+"
0
~
...................................................................................
"
'
x
v=o~7 m/,
~ . .............................. i ................................................................
x.:.:::~.............................................................................
o.15 ......... 9 ......... ~--.--a
o. 1
'
X
~i~ . . . . . . . . . . . . .
)~
•
............ x .........
0.15
Friction Coefficient for 2% Mud
t +
v = O. 3-7- m / s v=0.53 m/s
i
"0
+
'
v = 0 . 2 1 ~n/s
.......................................................................................... -~:.............................. ~...................................................
0.35
0.2
0 45
~........
o15
'
i
'
'
.
............................
5. O p t i c a l M e a s u r e m e n t s
ol ........~...........: •
0.05
.............................................................................. i : i
0 0
, 1
i 2
3
4
J 5
W
Figure 5. Friction Coefficient for 2 % Mud
6
5.1. V i s u a l i s a t i o n Figure 8 show a selection of images taken from water-based mud lubricated point contacts, with the inlet on the left. The circle represents the Hertzian contact area with diameter of about 300 microns.
336
Images were captured during the same run, for different speeds and for a load of 20 N. A comparison between rolling and sliding conditions is considered and from these images the film thickness data were obtained. The image at the top shows the starting condition, with the Hertzian contact area and the typical blue colour indicating the zero-point for the film thickness measurements. Under pure rolling conditions, at low speeds, a fairly uniform film distribution throughout the contact is encountered. However, for the sliding condition, a fairly thick (> 150nm) deposition of solid film in the track is evident and it increases with running time. In these conditions, the film appears to be load-carrying. This is not seen in the rolling case. At high speeds, very different film formation patterns are seen. Moreover, in the rolling case, it can be noted that, starting from a speed higher than 120mm/s, a build up of the film occurs at the side of the contact (,~ 80nm). It appears to have a very similar film thickness pattern to that seen with with starved fluid lubricant systems [8]. Our case can be considered a very particular type of starvation that occurs as a consequence of the nature of the fluid while the entry region is still completely flooded. At speeds approaching 400mm/s, the film deposition at the edges of the contact is more and more significant while at the centre the film collapses. Regarding the sliding condition, the more critical and severe nature of the contact is evident from the images. Quite different film distribution patterns between rolling and sliding conditions were observed. This is likely to be due to different entrainment and deposition conditions. The process appears to be dominated by the formation of deposited layers of clay particles that are loadcarrying under some conditions. The conclusions drawn from these preliminary optical investigations illustrate and clarify the lubricating behaviour of the muds. It is evident that there is no traditional fluid film formation following either the hydrodynamic or the elastohydrodynamic theories. Since aqueous clay suspensions are nonNewtonian fluids and have a pressure-viscosity
coefficient that is nearly zero (6.68.10 -1~ m2/N), no classical EHD behaviour is expected. This is also confirmed by the traction data which show high friction coefficients and, in the range of contact parameters close to typical mixed regime, a quite highly scattered distribution of the values. Therefore the lubrication regime encountered can be considered semi-fluid (boundary components seems to be important) and classified as a mixed starved non-Newtonian regime. A particular type of starvation occurs where the entry conditions are affected not by the inlet supply of lubricant but by the fluid nature and rheology. It also appears that the fluid going through the contact undergoes some sort of degradation since "dry" clay particles are found to remain deposited on the surfaces. This effect is particularly evident as the contact conditions approaches the boundary lubrication regime. An important issue in the improvement of the entry conditions might be the control of the effect of the degree of flocculation and the thixiotropy of the mud. The modification of the shear history of the suspension may change, to a certain extent, the lubricating action of the mud. Also, the boundary lubrication part of the process may require some detailed investigation and may turn out to be determinantin the study of the problem. 5.2. F i l m T h i c k n e s s
Measurements
The film thickness profiles shown in figure 9 were obtained from tests in rolling conditions and correspond to the set of images on the left in figure 8. The film thickness profiles were taken along the centreline orthogonal to the direction of rolling. The data confirm the qualitative observation obtained from the images. The films are extremely thin (< 300nm) and their shape is somewhat irregular and, by no means, resemble the classic EHD shape. There seems to be a build-up of film thickness with increasing contact rolling speed. Almost the same situation is encountered in grease-lubricated starved contacts [11,12], possibly due to the similar rheological nature of the fluids. Both are mainly constituted by two phases and they both display complex theologies.
337
Figure 8. Silica disk images for rolling and sliding
338
Film thickness for 5% mud (roiling contact) 600 I I / 50o
L
/
'
'
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
' z e r o point v=25mm/s v= 100 m m / s
: .............................................................................
~
;
i
V=200 m m / s
v=300 mm/s
+ x ;~ [] 9
~- ,oo ..................~.....................:....................:i........................i..................i....................................................... ~" 300
- -
200
-§
,~-
-
-
-:~-~I~
Z
100
0
0
50
100
150
200
250
300
350
400
position (microns)
Figure 9. Measured film thickness for 5% mud (rolling contact) 5.3. D i s c u s s i o n o f R e s u l t s The friction data can be better analysed if presented in the form of classical Stribeck curves. The curves were obtained by plotting the measured friction coefficient versus the group ~TU/W, the product of viscosity and velocity divided by the load. This has been done for three different mud compositions, with clay contents of 2%, 3% and 5%. The Bingham equation T --
TB
contributions act together. A quite remarkable result is the fact that these curves (10) show an overall trend that is similar to the "traditional" Stribeck curves obtained for a conventional lubricant such as a mineral oil. For example, figure 11 shows a Stribeck curve obtained on the Zircon machine running mineral oil (Shell Tellus 46). From figure 10, the data seems to collapse fairly well onto a master curve for both high and low values of the ratio rlU/W (in figure 10 regions I and III), while there is a quite substantial scatter for the intermediate values (region II). This might be due to the complex nature and rheology of the fluid that, under those conditions of load and speed, affects the entry of the particles and as a direct consequence the film formation and the overall lubrication process. This seems to be a particular type of starvation effect, not caused by the typical reduction of fluid inlet but by the intrinsic rheological nature of the fluid. In certain contact conditions (zone II), the clay particles do not continuously enter the contact area but are either squeezed to the side of the contact or they accumulate in the inlet as aggregates until they are sheared and pass through. This may cause the scatter of the data in the central portion of the Stribeck curves. For low values of the ratio ~TU/W (high loads
n t- ?~plZY
was used to evaluate the effective viscosity TB
~e/S
-
-V
+ rlpl
3 % mud 5 % mud
The Bingham parameters 7x and ~pZ (yield stress and plastic viscosity) for different clay contents were obtained from [13,14] where similar muds were studied; TB
=
0.60 exp(33.46@~) [Pal
Tip I
---
4.43exp(10.55r
Stribeck Curves
(1)
(2)
Since the inlet shear rate is known to be very high (105 to l0 s s - l ) , it is possible to calculate the effective viscosity by simply neglecting the first term in equation 1. The portions of Stribeck curve obtained corresponds approximately to the mixed (or partial lubrication) regime where both fluid and boundary
.
"~
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
"t2 ...................................... . ..................................:.............................. i o i~?fx i ~ ! ! ~~i_x 2 IL i i ,I i
0.3
8 g
9 0
,......................................................
.
.
.
.
.
.
.
.
.
.
.
0.2 .r-
%io ~(
oo
.J'
wn
,-,X ~
"'- X
"
9 XO OI O X i
i ................................ X ............ ~ "I"
0.002
9
9
:
: :
9
X "0 0 I .I'x "0" "0" "0" ......... X ........ X I: . . . . . . . . . . . . . . . . . ,
I
0
i X
0.004
0
,0
I
i
0.006
0.008
0
nU/W
Figure 10. Stribeck curve for 2, 3 and 5% mud
0.01
339
and low velocities, zone I) the boundary lubrication mechanism seems to prevail. Particles enter the contact and contribute to the load carrying. For low loads and higher velocities (zone III), the lubrication regime tends to approach the fluid regime and this allows for a more regular and relatively thick film build up and the data are more stable. In this case the particles rarely enter the contact and the main lubrication mechanism is provided by the continuous phase alone. These features of the curve are encountered for the three different mud composition and the scatter seems to be more accentuated as the clay content decreases. An important factor to be considered is also the influence of the shear history of the mud, its degree of flocculation and the preparation. No control on the shear state of the mud was attempted in these preliminary studies and also no additives were used to control flocculation. These two factot may be of key importance and will be included in more detailed future studies. The shear history control might affect the result and also explain the scatter encountered in the Stribeck curves. The reproducibility of the data was also found to be strongly dependent on the preparation protocol adopted for the suspension and on its flocculation state. In particular it was relatively difficult to obtain highly reproducible results since many of measurements were affected by the nonuniform shear state of the suspension. In the future work more reproducible systems will be obtained by controlling the shear history of the m u d .
6. C o n c l u s i o n s The design of drilling fluids has become, in recent times, a crucial issue in drilling technology. In particular, great interest is focused upon their potential lubricating role. This paper has presented the preliminary results from a study aimed to characterise the lubricating behaviour of water-based drilling muds and to examine the potential improvements of their lubricity. The techniques adopted were friction measurement, film thickness measurement and visualisation of the flow through a model contact.
Stribeck
Curve
,
Jr
+
,,-, t._~
0.03
~:~
0.025
for Mineral
Oil Shell
,
-: . . . . . . . .
:.
. . . . . . . . . . . . . . . .
Tellus
46
!
.
.
'
.
--
i.
.
.
.
.
.
.
0 o 0.02 ._
"6
+
.c-
~"
0.015
0.01
i
........ - ~ + .....................................................
~++
, ..................................
+
i ...............................................................
;
i
................~:~..+..<;+..;.................................................. i.................................i............................ .
.
.
+++ ++ . .
.
.
.
++t' -t-
+•
+-r,
I
I
0.005
0.01
"t.
.
+
0 015
+
.
+
0.02
.
_
0.025
nv/W
Figure 11. Tellus 46
Stribeck curve for mineral oil Shell
The results of this work can be summarised as follows: 9 The measured friction coefficients are in the range 0.1 +0.4, corresponding to the boundary lubrication regime. 9 From the analysis of the Stribeck curves and the film thickness measurements, three basic regimes can be identified. A regime which is essentially controlled by the deposition of clay layers onto the surfaces (at high loads and low speeds) and a regime in which the lubricating action is provided mainly by the base fluid. A transitional regime is identified when passing from high loads to low loads, characterised by a less predictable behaviour, due to changes in the contact fluid composition. 9 The issue of "starved" lubrication appears to be important for these fluids and reflects the fact that the entry conditions, and thus the rheology of the fluid, strongly affect the formation of the lubricant film. 9 The optical studies indicate that the system does not actually "starve" even at high speeds because the inlet appears to be always flooded (no fluid-air meniscus is detected). Rather, a sort of phase separation takes place when the suspension flows
340
through the contact. In particular, the clay aggregates seem to flow through the contact in a quite disordered fashion and the starvation phenomenon is caused by the fact that, at high speeds, the particles are pushed to the side of the contact leaving a very thin and irregular film in the central region.
.
.
6. .
o
7. A c k n o w l e d g e m e n t s .
The authors would like to thank Schlumberger Cambridge Research for providing financial support to this project.
REFERENCES 1. 2.
G. Maitland. Brit. Soc. Rheol. Bull., 33:7898, 1991. H. Darley and G. Gray. Composition and
Properties of Drilling and Completion Fluids. 3.
Gulf Publishing Co., Houston, 1988. G. Chillingarian and P. Vorabatur. Drilling and Drilling Fluids. Elsevier, Amsterdam, 1981.
10. 11. 12. 13. 14.
F. Fordham et al. Drilling and Pumping Journal, 6, 1988. K.Adamson et al. Oilfield Review, 10, 1998. C. A. Johancsik et al. SPE Reprint Series, 30. Directional Drilling. R. C. Wayte G. J. Johnstone and H. A. Spikes. ASLE Trans, 34:187-194, 1991. J. H. Hutchinson P. M. E. Cann and H. A. Spikes. STLE Trib. Trans., 39:915-921, 1996. B. Scrouton B. J. Briscoe and R. F. Willis. Proc. Roy. Soc. Lond., A333:99, 1973. B. J. Briscoe. Solid-Solid Interactions, pages 67-80, 1996. Imperial College Press. P. M. E. Cann. Tribol. Trans., 39:698-704, 1996. P. M. E. Cann. Proc. 22nd Leeds-Lyon Symposium, 1996. B. J. Briscoe et al. Phil. Trans. R. Soc. Lond. A, 348:179-207, 1994. B. R. Ren. PhD thesis, Imperial College, London, 1992.
331
Lubrication of Water-Based Clay Suspensions B. J. Briscoe a, p. M. Cann b, A. Delfino a, G. Maitland c ~Department of Chemical Engineering, Imperial College, London, UK bDepartment of Mechanical Engineering, Imperial College, London, UK cSchlumberger Cambridge Research, Cambridge, UK This paper describes preliminary studies of the lubricating behaviour of a simple water-based rock drilling fluid (bentonite clay suspension) in metal contacts. Friction measurements have been carried out on a modified tribometer (Zircon model, Glacier Metals, UK) with a model contact between a rotating shaft and a loaded planar counterface. These experiments were designed to investigate the tribology of the contact between the drillstring and the metal wall of an oilwell. Measurements are reported for a range of loads and contact velocities and for different compositions of the fluid. The result are presented in the form of classical Stribeck-Hersey curves in order to identify the lubrication regime and to illustrate the combined effects of load and speed on the friction coefficient. Optical interferometry measurements have also been carried out, using a model counterface of a metal sphere upon an optical plate, in order to visualise the flow of the suspension through the contact and also to measure the film thickness and the film thickness profiles. Two basic lubrication regimes are identified: at high loads a regime characterised by the deposition of layers of solid clay onto the contacting surfaces and at low loads a regime in which the main lubricating action is provided by the base fluid. In the transition between the two regimes, an intermediate region is found to be characterised by changes in the fluid composition and rheology within the contact. The general trend of the Stribeck curve is obtained and a peculiar scattering of the data is evident in the region between the boundary lubrication regime and the mixed lubrication region. The intrinsic nature and the complex rheology of the fluid appear to be the parameters that may control this effect and in part define the lubrication regime.
1. T h e Frictional P r o b l e m in Oil Drilling In oil well technology special fluids are used during drilling, completion and t r e a t m e n t operations. These fluids are in general complex multicomponent dispersions or solutions and they are required to fulfill many different roles.[1-3]. For example, pressure control in the wellbore, lubrication and cooling of the drill bit, removal of the cuttings from downhole locations and prevention of blowouts. These fluids are formulated to have carefully controlled rheological and tribological properties as these have a profound effect upon the overall drilling and well operation [1,4,5]. In modern drilling technology it is desirable to improve the conventional lubricating action of the mud since increasingly severe conditions are encontered in extended-reach (directional) oil wells. It is generally accepted t h a t torque and drag forces in directional wells are primarily caused
by sliding friction between the drillstring and either the metallic casing or the drilled strata. The friction coefficient is defined as the ratio of the frictional force and the sidewall normal contact force. Friction coefficients can be estimated from field measurements and are generally in the range 0 . 1 - 0.4 [6]. In drilling application it would be desirable to reduce the friction in order to reduce the energy dissipation and the wear of the drillstring, as well as the prospect of drillstring seizure and breakage. This paper presents the preliminary results of an investigation of the lubricating behaviour of water-based drilling muds in metal contacts. It also a t t e m p t s to clarify the lubricating role of the mud in the contact between the metal drillstring and the wall sections of an extended-reach deviated wellbore. The study has been carried out experimentally using a typical tribometer (Zircon Model, Glacier Metals, UK),
332
modified to give a model lubricated contact between a rotating shaft and a housing. The friction has been measured over a range of rotational velocities and normal loads for different mud compositions which reflect, in part, well bore environments. An optical interferometry technique has been used to measure film thickness and to visualise the flow of the lubricant through the contact in order to provide a better understanding of the lubrication mechanism.
~~.td
IN
Exte:.~ed-Reach
0il Well
2. T h e Drilling P r o c e s s and Drilling Fluids Drilling fluids serve many different purposes [13]. The ones concerned with the present work are: 9 lubricating (and cooling) of the drillstring and the bit to minimise driving torque, drag and pipe sticking and rupture. 9 cleaning the bottom of the hole and carrying the rock debris to the surface. 9 in general maintaining the integrity of the hole, drillstring, casing and tubing. 9 providing an hydrostatic pressure that prevents blowout and the release of gases during well operations.
2.1. Friction in D i r e c t i o n a l Wells A directional well (figure 1) is one in which a significant deviation in the drilling path is deliberately initiated in order to reach particular zones in the formation. In certain situations it is the only practical and economic way of reaching the production area. For example, in multiple offshore drilling or when it is desired to reach inaccesible locations or to increase the exploitation of a reservoir. In deviated wells, as these wells are termed, drag and torque forces tend to be more troublesome and in deep and highly deviated wells the reduction of those forces can be critical in order to achieve satisfactory well completion. 2.2. T y p e s of Drilling Fluids Drilling muds can be defined as a suspension of solids in a liquid phase where the liquid can be water or oil. It is generally possible to distinguish
Figure 1. Schematic of an extended-reach deviated oil well between three main types of drilling fluids: waterbased muds, oil-based muds and emulsion muds. This study focuses mainly on water-based fluids as they are the most widely used at present. Water based drilling fluids contain at least four main components: the continuous phase (water), the dispersed colloidal phase (bentonite clay), the chemical components (additives, polymers) and the inert components (weighting materials). The continuous phase provides the initial viscosity and the medium into which the other components are dispersed. The colloidal phase enhances viscosity and develops a yield point through the absorption of water. The chemical components control the relevant physical properties of the mud and the inert phase provides high gravity solids to increase the mud weight. The polymers are usually selected from a wide range of products such as carboxymethylcellulose and hydroxyethylcellulose while the weight materials are generally barite, galena, iron oxides or limestone. The composition is variable and the clay content is generally 3 - 7% by weight. The rheological properties of the mud are very important as they influence the frictional pressure
333
Mud Cake
Dri11pipe
Mud
provide useful insights into the problem. In this paper two fundamental aspects will be considered: the measurement of friction (friction reduction is the most important aspect in the drilling application) and the mechanism of lubrication and the nature of the film formed.
Drillstring
3. Materials ments
Figure 2. Schematic of a section of the well hole drop, torque and pipe sticking, cutting transport and the conditions at the boundary of the mudcake. 2.3. A i m of the S t u d y The development of deviated wells presents severe technical problems and increasingly it is becoming important to optimise the lubrication properties of the drilling fluids. In particular it is highly desirable to be able to improve the lubricity of the drilling fluid without modifying other physical properties (viscosity, degree of flocculation etc.) that are required for other fundamental roles such as the transport of cuttings up to the surface. The aim of this work is to study and characterise the lubricating behaviour of water-based drilling fluids and to try to improve their lubricating effect. The system considered is illustrated in figure 2. The drillstring rotates inside the well hole which is filled with mud and contacts the walls of the well. Realistic simulations of the real contact problem are difficult due to the scale and the complexity and variability of the problem (types of drilled rocks, types of different drillstrings etc). The contact conditions are highly variable but the system can be basically considered as a pure sliding, highly loaded contact. This paper therefore has focused upon studying the basic tribological properties of the mud rather than the complete system. Nevertheless, it is reasonable to believe that even in this way the study will be able to
and E x p e r i m e n t a l Arrange-
The mud was prepared by dispersing a common type of bentonite clay (Whitegel, IK) in water (2% ,3% and 5% clay weight fraction). The clay was slowly added to the water while mixing with a high-shear rate mixer (Silverstone, UK). At least 45 minutes of mixing were necessary to obtain a well-dispersed and reproducible suspension. The mud was then allowed to swell for 24 hours before use.
3.1. Zircon T r i b o m e t e r The frictional behaviour of water-based drilling fluids in metal contacts was studied using the apparatus shown schematically in figure 3. It was a typical tribometer (Glaciers-Metals, Zircon model) that was modified in order to obtain a lubricated contact between the shaft and a planar metal counterface. Mild steel samples (4 x 4 cm) were used for the counterface and were loaded against a 25 mm diameter rotating mild steel shaft. The normal load was applied by means of dead weights ranging from 0.05 kg to 0.8 kg (maximum Hertzian pressure ~ 100 MPa) and the speed of the shaft was 0.2 to 0.9 m/s ( 8 0 - 360 rpm ). A piezo-electric force tranducer (Kistler, Germany) was attached to the sample holder in order to record the frictional force after suitable calibration. A plastic box (made of commercial PMMA) was fitted with suitable seals onto the shaft in order to contain the fluid. In this arrangement, the shaft could be submerged in the mud providing a nominally completely flooded contact. The fluid reservoir was equipped with a stirrer in order to control the shear history of the mud and investigate its effect upon the lubrication process; this part of the study is not described.
334
co
Zircon Tribometer
mpeccromecer
load
Arm
\
glamm dimk
Transducer
~ ~ ~
f Metal Sample
alur~um
/ miliaa
1ubz-lom~e film B(:ee2 J~t22
.Z4yez"
I
kPropeller ~o~~t~ng Shaft
Figure 3. Schematic of the Zircon tribometer 3.2. C a l i b r a t i o n The transducer was calibrated attaching a metal wire going through a pulley so that dead weights could be used. Different settings of the sensitivity of the transducer were adopeted in order to encompass the appropriate ranges for the forces to be measured. 3.3. O p t i c a l I n t e r f e r o m e t r y The film thickness measurements were carried out using a modified optical interferometry techinque [7]. A diagram is shown in figure 4. The contact was formed by a steel ball rolling against a glass disc. The underside of the disc was coated with a chromium layer overlaid with silica. Monochromatic or white light was shone into the contact through the glass disc. Some of the light reflects off the surface whilst some passes through the lubricant film in the contact to be reflected back off the other surface. The two reflected beams then optically interfere to an extent dependent on the difference in path length, and thus the lubricant film thickness, to produce an interference pattern. A microscope coupled to a CCD video camera was positioned directly above the contact. The camera was connected to a capture board so that images from the contact could be taken and stored for analysis. This technique produced a contour map of the contact film thickness and previous studies [8] have established a colour-film thickenss calibration that was used in this study.
\z
Figure 4. Optical interferometry techinque for film thickness measurement The preliminary optical studies consisted of testing 5% mud in both rolling and sliding conditions. The load was 20 N and the velocities ranged from 0.025 to 0.4m/s. Images were captured during the test in order to visualise the flow of the suspension through the contact and film build up. Also, from the image data, it was possible to estimate the central film thickness. 4. R e s u l t s and D i s c u s s i o n 4.1. Friction M e a s u r e m e n t s The frictional force between the shaft and the counterface has been measured for a range of loads and rotating speeds. The frictional force in the contact was measured as a function of time. Signal noise that arose from uncertain sources were removed before any further analysis was carried out; also significant spurious vibrations were introduced by the drive system. The data were then averaged to obtain the mean value of the frictional force and to calculate the friction coefficient. Data is presented for 3 different mud compositions and the friction coefficient is plotted against the normal load W for different speeds (see figures 5, 6 and 7). The range of values of the friction coefficient is very high (generally between 0.05 and 0.4), suggesting that the contact works in the so-called mixed lubrication regime and for the highest loads in the boundary lubrication regime. As is normally expected in lubricated metal contacts, the
335
friction coefficient increases with increasing load and with decreasing sliding velocity. There seems to be a "transition" in the trend of the data at about 1 . 5 - 2 N between the two regimes. It can be noted that, for loads higher that 1.5 N, the friction coefficient is more affected by the change in the sliding velocity. A similar trend was found for the other compositions. With the more dilute muds, the friction coefficient was higher, indicating a less efficient lubricating action. If pure water was used, a decrease in the friction coefficient was observed but the wear of the surfaces was found to be substantially higher than for mud-lubricated contacts. This was probably be due to the presence of a solid layer of clay deposited onto the surface (especially for higher loads) that protected the contacting surfaces. In this case the friction phenomena is found to be dependent upon the solid contact of clay particles and the friction coefficient tends to be higher probably due to an increase of the area of solid contact [9,10]. More work on this important aspect of the problem is currently been carried out.
Friction Coefficient for 3% Mud 0.45 0.4
, .....................................................................................................................
+i~
0.3
......................................................
~:
0.25
+
.~_,,
0.2
~ ......................
+ ......
0.1
o.os
' i
0 0
0.4
..................................
0.35
03
"~
0.25
~
" )
+
....
~ 0.2
'
~
i
a
1
2
" ~
DI ~
- .............................
- ............................... - .....................................................
)
................
3
. . . . . . . .
i
i
4
5
6
Friction Coefficient for 5% Mud '
0.4
'
'
................................................................................
035
i
...............................................................................................
0.3
8
..................................................
"
+!
El
....
i
....
-
.
.
.
.
.
.
.
x:~
x ..
i
i• ,, i ! ~ ........................ "........................................................
9 ...............................
~3
~+ .............................~ : ...........................Oi ~ ........................................................
0
...........;~i
................
x
v =m0/.S5v3= om. 7/ s
i ......................... ; : .......... ~ .................
.....
. . . . . . . . . .
0.25
+
v=0.37 m/s
! ..................................................
.....................................................................................................
'0
v = 0 . 2 1 ,,,'Is
i
. ............................... -:"......................................................................................
i
i
i
1
2
.
.
.
.
.
.
.
.
.
'
---
i
3
4
5
6
--
0
........................
Figure 7. Friction Coefficient for 5% Mud
.............................
i
0 []
v=O 3 7 m / s Xv=O 53 m/s v=0.7 m/s
........................................................
:; ..........................
W
:
.............................................................................................
8
:
........................................................................................................................... +
;E=
.......... x !
.................................."'+'" ................ i .................~:
X
5
~ X~:
+ x Y~
o-
................... ~ ..................................................
0.45
0
,,,',,,,Is
n
Figure 6. Friction Coefficient for 3 % Mud
o +v=0.21
. []
W
005
+"
0
~
...................................................................................
"
'
x
v=o~7 m/,
~ . .............................. i ................................................................
x.:.:::~.............................................................................
o.15 ......... 9 ......... ~--.--a
o. 1
'
X
~i~ . . . . . . . . . . . . .
)~
•
............ x .........
0.15
Friction Coefficient for 2% Mud
t +
v = O. 3-7- m / s v=0.53 m/s
i
"0
+
'
v = 0 . 2 1 ~n/s
.......................................................................................... -~:.............................. ~...................................................
0.35
0.2
0 45
~........
o15
'
i
'
'
.
............................
5. O p t i c a l M e a s u r e m e n t s
ol ........~...........: •
0.05
.............................................................................. i : i
0 0
, 1
i 2
3
4
J 5
W
Figure 5. Friction Coefficient for 2 % Mud
6
5.1. V i s u a l i s a t i o n Figure 8 show a selection of images taken from water-based mud lubricated point contacts, with the inlet on the left. The circle represents the Hertzian contact area with diameter of about 300 microns.
336
Images were captured during the same run, for different speeds and for a load of 20 N. A comparison between rolling and sliding conditions is considered and from these images the film thickness data were obtained. The image at the top shows the starting condition, with the Hertzian contact area and the typical blue colour indicating the zero-point for the film thickness measurements. Under pure rolling conditions, at low speeds, a fairly uniform film distribution throughout the contact is encountered. However, for the sliding condition, a fairly thick (> 150nm) deposition of solid film in the track is evident and it increases with running time. In these conditions, the film appears to be load-carrying. This is not seen in the rolling case. At high speeds, very different film formation patterns are seen. Moreover, in the rolling case, it can be noted that, starting from a speed higher than 120mm/s, a build up of the film occurs at the side of the contact (,~ 80nm). It appears to have a very similar film thickness pattern to that seen with with starved fluid lubricant systems [8]. Our case can be considered a very particular type of starvation that occurs as a consequence of the nature of the fluid while the entry region is still completely flooded. At speeds approaching 400mm/s, the film deposition at the edges of the contact is more and more significant while at the centre the film collapses. Regarding the sliding condition, the more critical and severe nature of the contact is evident from the images. Quite different film distribution patterns between rolling and sliding conditions were observed. This is likely to be due to different entrainment and deposition conditions. The process appears to be dominated by the formation of deposited layers of clay particles that are loadcarrying under some conditions. The conclusions drawn from these preliminary optical investigations illustrate and clarify the lubricating behaviour of the muds. It is evident that there is no traditional fluid film formation following either the hydrodynamic or the elastohydrodynamic theories. Since aqueous clay suspensions are nonNewtonian fluids and have a pressure-viscosity
coefficient that is nearly zero (6.68.10 -1~ m2/N), no classical EHD behaviour is expected. This is also confirmed by the traction data which show high friction coefficients and, in the range of contact parameters close to typical mixed regime, a quite highly scattered distribution of the values. Therefore the lubrication regime encountered can be considered semi-fluid (boundary components seems to be important) and classified as a mixed starved non-Newtonian regime. A particular type of starvation occurs where the entry conditions are affected not by the inlet supply of lubricant but by the fluid nature and rheology. It also appears that the fluid going through the contact undergoes some sort of degradation since "dry" clay particles are found to remain deposited on the surfaces. This effect is particularly evident as the contact conditions approaches the boundary lubrication regime. An important issue in the improvement of the entry conditions might be the control of the effect of the degree of flocculation and the thixiotropy of the mud. The modification of the shear history of the suspension may change, to a certain extent, the lubricating action of the mud. Also, the boundary lubrication part of the process may require some detailed investigation and may turn out to be determinantin the study of the problem. 5.2. F i l m T h i c k n e s s
Measurements
The film thickness profiles shown in figure 9 were obtained from tests in rolling conditions and correspond to the set of images on the left in figure 8. The film thickness profiles were taken along the centreline orthogonal to the direction of rolling. The data confirm the qualitative observation obtained from the images. The films are extremely thin (< 300nm) and their shape is somewhat irregular and, by no means, resemble the classic EHD shape. There seems to be a build-up of film thickness with increasing contact rolling speed. Almost the same situation is encountered in grease-lubricated starved contacts [11,12], possibly due to the similar rheological nature of the fluids. Both are mainly constituted by two phases and they both display complex theologies.
337
Figure 8. Silica disk images for rolling and sliding
338
Film thickness for 5% mud (roiling contact) 600 I I / 50o
L
/
'
'
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
' z e r o point v=25mm/s v= 100 m m / s
: .............................................................................
~
;
i
V=200 m m / s
v=300 mm/s
+ x ;~ [] 9
~- ,oo ..................~.....................:....................:i........................i..................i....................................................... ~" 300
- -
200
-§
,~-
-
-
-:~-~I~
Z
100
0
0
50
100
150
200
250
300
350
400
position (microns)
Figure 9. Measured film thickness for 5% mud (rolling contact) 5.3. D i s c u s s i o n o f R e s u l t s The friction data can be better analysed if presented in the form of classical Stribeck curves. The curves were obtained by plotting the measured friction coefficient versus the group ~TU/W, the product of viscosity and velocity divided by the load. This has been done for three different mud compositions, with clay contents of 2%, 3% and 5%. The Bingham equation T --
TB
contributions act together. A quite remarkable result is the fact that these curves (10) show an overall trend that is similar to the "traditional" Stribeck curves obtained for a conventional lubricant such as a mineral oil. For example, figure 11 shows a Stribeck curve obtained on the Zircon machine running mineral oil (Shell Tellus 46). From figure 10, the data seems to collapse fairly well onto a master curve for both high and low values of the ratio rlU/W (in figure 10 regions I and III), while there is a quite substantial scatter for the intermediate values (region II). This might be due to the complex nature and rheology of the fluid that, under those conditions of load and speed, affects the entry of the particles and as a direct consequence the film formation and the overall lubrication process. This seems to be a particular type of starvation effect, not caused by the typical reduction of fluid inlet but by the intrinsic rheological nature of the fluid. In certain contact conditions (zone II), the clay particles do not continuously enter the contact area but are either squeezed to the side of the contact or they accumulate in the inlet as aggregates until they are sheared and pass through. This may cause the scatter of the data in the central portion of the Stribeck curves. For low values of the ratio ~TU/W (high loads
n t- ?~plZY
was used to evaluate the effective viscosity TB
~e/S
-
-V
+ rlpl
3 % mud 5 % mud
The Bingham parameters 7x and ~pZ (yield stress and plastic viscosity) for different clay contents were obtained from [13,14] where similar muds were studied; TB
=
0.60 exp(33.46@~) [Pal
Tip I
---
4.43exp(10.55r
Stribeck Curves
(1)
(2)
Since the inlet shear rate is known to be very high (105 to l0 s s - l ) , it is possible to calculate the effective viscosity by simply neglecting the first term in equation 1. The portions of Stribeck curve obtained corresponds approximately to the mixed (or partial lubrication) regime where both fluid and boundary
.
"~
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
"t2 ...................................... . ..................................:.............................. i o i~?fx i ~ ! ! ~~i_x 2 IL i i ,I i
0.3
8 g
9 0
,......................................................
.
.
.
.
.
.
.
.
.
.
.
0.2 .r-
%io ~(
oo
.J'
wn
,-,X ~
"'- X
"
9 XO OI O X i
i ................................ X ............ ~ "I"
0.002
9
9
:
: :
9
X "0 0 I .I'x "0" "0" "0" ......... X ........ X I: . . . . . . . . . . . . . . . . . ,
I
0
i X
0.004
0
,0
I
i
0.006
0.008
0
nU/W
Figure 10. Stribeck curve for 2, 3 and 5% mud
0.01
339
and low velocities, zone I) the boundary lubrication mechanism seems to prevail. Particles enter the contact and contribute to the load carrying. For low loads and higher velocities (zone III), the lubrication regime tends to approach the fluid regime and this allows for a more regular and relatively thick film build up and the data are more stable. In this case the particles rarely enter the contact and the main lubrication mechanism is provided by the continuous phase alone. These features of the curve are encountered for the three different mud composition and the scatter seems to be more accentuated as the clay content decreases. An important factor to be considered is also the influence of the shear history of the mud, its degree of flocculation and the preparation. No control on the shear state of the mud was attempted in these preliminary studies and also no additives were used to control flocculation. These two factot may be of key importance and will be included in more detailed future studies. The shear history control might affect the result and also explain the scatter encountered in the Stribeck curves. The reproducibility of the data was also found to be strongly dependent on the preparation protocol adopted for the suspension and on its flocculation state. In particular it was relatively difficult to obtain highly reproducible results since many of measurements were affected by the nonuniform shear state of the suspension. In the future work more reproducible systems will be obtained by controlling the shear history of the m u d .
6. C o n c l u s i o n s The design of drilling fluids has become, in recent times, a crucial issue in drilling technology. In particular, great interest is focused upon their potential lubricating role. This paper has presented the preliminary results from a study aimed to characterise the lubricating behaviour of water-based drilling muds and to examine the potential improvements of their lubricity. The techniques adopted were friction measurement, film thickness measurement and visualisation of the flow through a model contact.
Stribeck
Curve
,
Jr
+
,,-, t._~
0.03
~:~
0.025
for Mineral
Oil Shell
,
-: . . . . . . . .
:.
. . . . . . . . . . . . . . . .
Tellus
46
!
.
.
'
.
--
i.
.
.
.
.
.
.
0 o 0.02 ._
"6
+
.c-
~"
0.015
0.01
i
........ - ~ + .....................................................
~++
, ..................................
+
i ...............................................................
;
i
................~:~..+..<;+..;.................................................. i.................................i............................ .
.
.
+++ ++ . .
.
.
.
++t' -t-
+•
+-r,
I
I
0.005
0.01
"t.
.
+
0 015
+
.
+
0.02
.
_
0.025
nv/W
Figure 11. Tellus 46
Stribeck curve for mineral oil Shell
The results of this work can be summarised as follows: 9 The measured friction coefficients are in the range 0.1 +0.4, corresponding to the boundary lubrication regime. 9 From the analysis of the Stribeck curves and the film thickness measurements, three basic regimes can be identified. A regime which is essentially controlled by the deposition of clay layers onto the surfaces (at high loads and low speeds) and a regime in which the lubricating action is provided mainly by the base fluid. A transitional regime is identified when passing from high loads to low loads, characterised by a less predictable behaviour, due to changes in the contact fluid composition. 9 The issue of "starved" lubrication appears to be important for these fluids and reflects the fact that the entry conditions, and thus the rheology of the fluid, strongly affect the formation of the lubricant film. 9 The optical studies indicate that the system does not actually "starve" even at high speeds because the inlet appears to be always flooded (no fluid-air meniscus is detected). Rather, a sort of phase separation takes place when the suspension flows
340
through the contact. In particular, the clay aggregates seem to flow through the contact in a quite disordered fashion and the starvation phenomenon is caused by the fact that, at high speeds, the particles are pushed to the side of the contact leaving a very thin and irregular film in the central region.
.
.
6. .
o
7. A c k n o w l e d g e m e n t s .
The authors would like to thank Schlumberger Cambridge Research for providing financial support to this project.
REFERENCES 1. 2.
G. Maitland. Brit. Soc. Rheol. Bull., 33:7898, 1991. H. Darley and G. Gray. Composition and
Properties of Drilling and Completion Fluids. 3.
Gulf Publishing Co., Houston, 1988. G. Chillingarian and P. Vorabatur. Drilling and Drilling Fluids. Elsevier, Amsterdam, 1981.
10. 11. 12. 13. 14.
F. Fordham et al. Drilling and Pumping Journal, 6, 1988. K.Adamson et al. Oilfield Review, 10, 1998. C. A. Johancsik et al. SPE Reprint Series, 30. Directional Drilling. R. C. Wayte G. J. Johnstone and H. A. Spikes. ASLE Trans, 34:187-194, 1991. J. H. Hutchinson P. M. E. Cann and H. A. Spikes. STLE Trib. Trans., 39:915-921, 1996. B. Scrouton B. J. Briscoe and R. F. Willis. Proc. Roy. Soc. Lond., A333:99, 1973. B. J. Briscoe. Solid-Solid Interactions, pages 67-80, 1996. Imperial College Press. P. M. E. Cann. Tribol. Trans., 39:698-704, 1996. P. M. E. Cann. Proc. 22nd Leeds-Lyon Symposium, 1996. B. J. Briscoe et al. Phil. Trans. R. Soc. Lond. A, 348:179-207, 1994. B. R. Ren. PhD thesis, Imperial College, London, 1992.