Thixotropy Of Bentonite Dispersions

This article was downloaded by: [HEAL-Link Consortium] On: 6 August 2008 Access details: Access Details: [subscription number 786636650] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Energy Sources, Part A: Recovery, Utilization, and Environmental Effects Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713770930

Investigations on the Thixotropy of Bentonite Suspensions V. C. Kelessidis a a Department of Mineral Resources Engineering, Technical University of Crete, Chania, Greece Online Publication Date: 01 January 2008

To cite this Article Kelessidis, V. C.(2008)'Investigations on the Thixotropy of Bentonite Suspensions',Energy Sources, Part A:

Recovery, Utilization, and Environmental Effects,30:18,1729 — 1746 To link to this Article: DOI: 10.1080/15567030701456261 URL: http://dx.doi.org/10.1080/15567030701456261

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Energy Sources, Part A, 30:1729–1746, 2008 Copyright © Taylor & Francis Group, LLC ISSN: 1556-7036 print/1556-7230 online DOI: 10.1080/15567030701456261

Investigations on the Thixotropy of Bentonite Suspensions V. C. KELESSIDIS1

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1

Department of Mineral Resources Engineering, Technical University of Crete, Chania, Greece Abstract The thixotropic behavior of water bentonite suspensions has been investigated using Wyoming and Zenith bentonites at 5.0% and 6.42% concentrations with 0.0 M, 0.01 M, and 0.1 M NaCl following API preparation protocols. Rheological data was collected with a Couette viscometer, deriving first the D0 curve, from 600 to 3 rpm, followed by a typical thixotropic loop, getting the U curve from 3 to 600 rpm and then the D curve from 600 to 3 rpm. The Herschel-Bulkley model describes very well all experimental data. Yield stresses did not vary, but variations of flow consistency and of flow behavior indices were detected for the three curves. All suspensions with no salt and 0.01 M salt exhibited anti-thixotropy, estimated from the D0–U curves, while the 0.1-M suspensions showed no thixotropy. Thixotropy indices vary and depend on bentonite type and concentration and on the presence and amount of salt. The mechanisms of thixotropy and implications on drilling fluid performance are discussed. Keywords thixotropy

bentonite, drilling fluids, Herschel-Bulkley, rheology, suspensions,

Introduction Thixotropy has been defined as the variation of shear stress (shear rate) with time when a fluid is subjected to a constant shear rate (shear stress) (McMillen, 1932a, 1932b, 1932c; Mewis, 1979; Barnes, 1997). It is a characteristic behavior of two-phase systems (Nguyen and Boger, 1985; Lemke et al., 1999), like clay suspensions, which contain non-spherical particles (Barnes, 1997; Nakaishi and Yasutomi, 1994; Luckham and Rossi, 1999). The influence of time of mixing, time of measurements, and time of shearing on rheological properties of these two-phase systems normally is not taken into account in fluid flow computations, but it has been the subject of continuous experimental and theoretical research from early years, among many by Ambrose and Loomis (1933), Moore (1959), and Singhal and Malik (1964), to recent times by Cheng (2003), Li et al. (2003), Labanda et al. (2004), Roussel et al. (2004), and Bekkour et al. (2005), among many others. Constitutive equations relating shear stress to shear rate with a structural parameter using structural kinetics approach to thixotropy have been discussed and analyzed by Moore (1959), Cheng and Evans (1965), Sestak et al. (1982), Toorman (1997), Mujumdar et al. (2002), Dullaert and Mewis (2006), and Galindo-Rosales and Rubio-Hernandez (2006).

Address correspondence to Vassilios C. Kelessidis, Department of Mineral Resources Engineering, Technical University of Crete, 73100 Chania, Greece. E-mail: [email protected]

1729

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1730

V. C. Kelessidis

Thixotropy is a reversible process and occurs because of the microstructure of the suspensions. For many materials, when they are sheared at a constant rate, the shear stress decreases with time because the structure is broken down. When shearing is removed, the material slowly rebuilds its structure, but the time necessary for complete structure restoration is much higher (McMillen, 1932a; Chavan et al., 1975). These systems are termed thixotropic or with positive thixotropy (Barnes, 1997). There are materials though where the opposite is true, and the stress increases with time upon application of constant shear rate. This phenomenon is termed anti-thixotropy or negative thixotropy (Barnes, 1997), and it has been reported for clay suspensions (Heckroodt and Ryan, 1978) as well as for certain simulated waste slurries (Chang and Smith, 1996). Negative thixotropy may result from temporary aggregation of suspension particles because of increased collisions due to shearing (Barnes, 1997), or when certain flocs become looser and more open under the action of shear, as reported for ferric oxide suspensions in mineral oil (Kanai and Amari, 1994). Combination of positive and negative thixotropy, sometimes termed complex thixotropy (Hou et al., 1998; Li et al., 2003), has been reported for clay water suspensions, but some investigators have attributed this to the effect of measuring time, with negative thixotropy observed for short measuring times and positive thixotropy observed for long measurement times (Nakaishi and Yasutomi, 1994). Water bentonite suspensions are encountered in a variety of industries and are the main ingredients in oil well drilling fluids. They exhibit a yield stress, and it is well accepted and well documented that yield stress fluids are also thixotropic and shear thinning (Barnes, 1997; Møller et al., 2006). There is little information, however, on their thixotropic behavior viewed from the drilling fluid industry point of view, although studies indicate an increase in the static gel strength of water bentonite-based drilling fluids with time (Speers et al., 1987), while Kok and Alikaya (2005) have shown that the addition of polymers also increases gelation time of drilling fluids. In general, thixotropy is not addressed when dealing with drilling fluids (Bourgogne et al., 1991), not only because of the complexity of the issue and the unavailability of models taken into account (Barnes, 1997), but also because it is assumed that, after the intensive shearing experienced by the fluids in the drilling process, thixotropy is kept at least to a minimum (Bourgogne et al., 1991). It is for this reason that the drilling fluid industry specifies comprehensive testing protocols (API, 1993; API, 2000) for preparation and laboratory as well as field testing of clay suspensions and, in particular, of water bentonite suspensions. According to these specifics, the mixture is subjected to very high shearing prior to rheological measurements, thus probably destroying the structure of the suspension. Deformation prehistory of the suspensions is very significant for studying and understanding thixotropy (Barnes, 1997), and consistent initial but also testing conditions are essential not only for testing for thixotropy but also for comparing results of other research work and even different tests in the same laboratory. The API procedures thus ensure, to the best of experimental accuracy at least, consistency in measurements, so that cross-comparisons can be made with more certainty. Further evidence has been recently provided by Møller et al. (2006), where it is stated that only by controlling the aging history of colloidal samples by large preshearing, reproducible results can be obtained. In oil well drilling operations, the fluids experience high shear rates in the drill pipe, where they flow downward, and extremely high shear rates when they go through the nozzles of the drill bits. There, the shear rates are of the order of thousands of reciprocal seconds before they enter the annulus, where they experience low shear rates on their way to the surface. This article attempts to provide some evidence that, despite high shearing in

Thixotropy of Bentonite Suspensions

1731

the laboratory, thixotropy is still observed in water bentonite suspensions, and it depends on the type of bentonite used and the concentration as well as on environmental variables like the amount of salt present in the suspensions. Data is therefore presented for the thixotropy of water bentonite suspensions after intensive shearing, with full rheograms derived by going from high to low shear rates followed by the standard thixotropic loop for two bentonites at two concentrations and three salt concentrations.

Experimental

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Materials Two commercial bentonites used by the drilling fluid industry have been utilized in this study: a Wyoming bentonite that is a naturally occurring sodium montmorillonite, kindly provided by Halliburton-Cebo Holland, and a Zenith bentonite, a calcium montmorillonite that is converted to sodium montmorillonite after treatment, kindly provided by S&B Industrial Minerals, Greece. Two concentrations have been tested, a 5.0% and a 6.42% w/w bentonite in water, of similar order to concentrations used by the drilling industry. Tests were performed in freshwater suspensions and in saline water, with electrolyte (NaCl) concentrations of 0.01 M and 0.1 M. The particle size of both bentonites was less than 70 m, with the majority of particles with diameters around 10 m.

Instruments and Methodology The American Petroleum Institute standard preparation and testing protocols (API, 1993; API, 2000) were followed. Bentonite was mixed with deionized water in a high-speed mixer (11,000 rpm) for a total mixing time of 20 min, and it was then stored in a container for 16 h at room temperature for complete hydration of bentonite particles. Prior to testing, the suspension was sheared for 5 min at high shear (11,000 rpm), and it was then poured in the viscometer cup. Rheological data was obtained with a continuously varying rotational speed Grace M3500 rotating viscometer (Houston, TX) at temperature of 25ıC and speeds of 600, 300, 200, 100, 6, and 3 rpm. Inner cylinder diameter was 1.7245 cm and outer rotating cylinder diameter was 1.8415 cm, thus giving a diameter ratio of ı D 1:06785. The maximum Newtonian shear rate in the viscometer was 1,021 sec 1 , while the minimum was 5.1 sec 1 . For the high speed mixer, the highest Newtonian shear rate was estimated around 2,300 sec 1 . The measurements began immediately after pouring the sample into the viscometer cup, starting from the high shear rate of 600 rpm and going down to 3 rpm, thus giving the D0 curve. The time of measurement at each rotational speed was 60 sec, for a total of six readings, which were taken at 10-sec intervals. Each of the six readings was then averaged and associated with the particular rotational speed. A 60-sec interval followed at zero shear rate, and then a full hysteresis thixotropy loop began by first deriving the up curve (curve U), going from 3 rpm up to the maximum 600 rpm and then continuing with the down curve (curve D) from 600 rpm down to the 3 rpm. The full time evolution of the rotational speed together with shear stress measurements for a particular sample is shown in Figure 1. Total test duration was 19 min. The Herschel-Bulkley rheological model, given by  D y C K. P /n ;

(1)

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1732

V. C. Kelessidis

Figure 1. Shear rate and shear stress time evolution for a particular sample.

where P is the shear rate, y is the yield stress, K is the flow consistency index, and n is the flow behavior index, has been used to describe each one of the three rheograms. For every suspension and particular salt concentration, three sets of rheological parameters are obtained by standard non-linear regression techniques, although other approaches have also been reported to derive the three rheological parameters for the Herschel-Bulkley model (Kok and Alikaya, 2003, 2004, 2005; Kelessidis et al., 2006). The HerschelBulkley model describes very well the experimental data derived not only from the work of this laboratory (Kelessidis et al., 2005), but also from other work for water bentonite suspensions (Coussot and Piau, 1994; Bekkour et al., 2005). The goodness-of-fit has been determined using two statistical indicators, the correlation coefficient, Rc2 , and the sum of square errors normalized by the yield stress of the suspensions, NSSE, defined as NSSE D

P .i Oi2 / ; y2

(2)

where i , Oi are the measured and predicted shear stress values. The thixotropy index is defined as the area enclosed between the up curve (U) and the down curve (D). The area of the hysteresis loop (A) essentially represents power per unit volume (Perret et al., 1996) because A D 
P ;

(3)

        1 N 1 N
m 1 ŒJ  power ŒAŒDŒP a ŒD ŒD ŒD 3 D : 2 3 s m s m s Œm Œs volume

(4)

with the units of (A) in the SI system,

Thixotropy of Bentonite Suspensions

1733

So the area represents the power required to bring the given volume of the material to a more fluid state. This power is expended in breaking up the gel of the fluid. Large hysteresis loop areas represent highly thixotropic material, while a null hysteresis loop area represents either a non-thixotropic material or material for which the structure has been completely broken up, and there was not enough time to rebuild, at least within the timeframe of the experiment. The thixotropy index has been estimated with two different methods. The first one is by integration of the Herschel-Bulkley equations, as

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

A D AU

AD D

Z

Pf

U d P 0

Z

Pf

D d ; P

(5)

0

with the subscripts U and D referring to the up and down curves, respectively. The second method is by computing the area of the trapezoids between two successive points and summing them all up. The computations using either method have shown that the results differ by less than 5%. The values given in this work are derived using the integration method.

Results The full rheogram and the thixotropic loop for the Zenith bentonite suspensions of 6.42% and 5.0% concentration are shown in Figure 2 for 0.0 M salt concentration, in Figure 3 for the 0.01 M salt, and in Figure 4 for the 0.1 M salt concentration. The data are shown with the error bands computed as one standard deviation from the six measurements taken at each shear rate. Many interesting features can be distinguished in all figures, with small and large variations observed for the suspensions at different salt concentrations. For the 6.42% bentonite concentration with no salt, curve D0 is below curve U, exhibiting an inverse thixotropy, or anti-thixotropy, if one considers the D0–U as a thixotropic loop. Analyzing through the standard thixotropic loop U–D, normal thixotropic behavior is observed, and the loop is similar to the one described by Perret et al. (1996), with the time spent at the highest speed kept to a minimum. It is worth noticing the coincidence of the D curve with the D0 curve for shear rates less than 511 sec 1 . The 5.0% bentonite suspension exhibits similar behavior to the 6.42% for the D0–U curves, with D0 being lower than the U curve, although to a smaller extent. Differences, however, are observed with the standard thixotropic loop, the area between U–D curves, being smaller for the 5% suspensions because the points at the two highest shear rates coincide with the points for the 6.42% bentonite suspension. The results from the suspensions with the 0.01 M salt concentration (Figure 3) show similar characteristics to the 0.0 M concentration for both bentonite concentrations. However, a big change is observed when one considers the results for the 0.1 M salt concentration (Figure 4) for both bentonite concentrations. The D0 curves are now higher or slightly higher than both the U and the D curves for all conditions. Shear stress values at the starting shear rates for the D0 curve almost coincide with shear stresses measured at the ending shear rates for the U curve, while ending values of the D0 curves coincide with ending values of the D curves. This behavior is observed for both concentrations. The U–D curves show anti-thixotropy for shear rates less than 400 sec 1 and no thixotropy for higher shear rates for both concentrations.

V. C. Kelessidis

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1734

Figure 2. Shear stress-shear rate thixotropic curves for 6.42% and 5% Zenith suspension with no salt. pH of the suspensions was 9.64 and 9.55, respectively.

The results for the Wyoming bentonite (Figures 5, 6, and 7) are similar to the Zenith bentonite results. The D0 curves are again lower than the U and D curves for both concentrations, for the 0.0 M and for the 0.01 M salt concentrations. They almost coincide or are slightly higher than both U and D curves at the 0.1 M salt concentration for the 6.42% and the 5.0% bentonite concentrations, respectively.

1735

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Thixotropy of Bentonite Suspensions

Figure 3. Shear stress-shear rate thixotropic curves for 6.42% and 5% Zenith suspension with 0.01 M salt. pH of the suspensions was 9.41 and 9.39, respectively.

For each of the three curves, D0, U, and D, the Herschel-Bulkley rheological parameters have been computed, and the goodness-of-fit, as determined from the values of the correlation coefficient and the normalized sum of square errors, has been very good for all samples. In Figure 8, the two statistical indices are shown for the 6.42% bentonite concentration for both bentonites and at all salt concentrations tested, with similar results

V. C. Kelessidis

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1736

Figure 4. Shear stress-shear rate thixotropic curves for 6.42% and 5% Zenith suspension with 0.1 M salt. pH of the suspensions was 8.83 and 8.74, respectively.

obtained for the 5% bentonite concentration. For the 6.42% concentration, the correlation coefficients are greater than 0.99 for all samples except the Zenith bentonite at 0.01 M salt concentration for which it is close to 0.98. Similarly, all normalized sum-of-square errors are very small and range from 0.05% to a maximum of 0.65% of the yield stress of each suspension, indicating the very good description of rheological data with the Herschel-Bulkley model.

1737

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Thixotropy of Bentonite Suspensions

Figure 5. Shear stress-shear rate thixotropic curves for 6.42% and 5% Wyoming suspension with no salt. pH of the suspensions was 8.95 and 8.79, respectively.

The yield stresses computed from the different rheograms for all samples are shown in Figure 9. No significant variations are observed among the D0, the U, and the D curves for each condition. Similar results have been obtained for the 5% concentrations. Relevant comparison for the flow consistency index should be made together with the flow behavior index because of the strong correlation among the two, particularly with

V. C. Kelessidis

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1738

Figure 6. Shear stress-shear rate thixotropic curves for 6.42% and 5% Wyoming suspension with 0.01 M salt. pH of the suspensions was 8.69 and 8.60, respectively.

respect to the shape of the rheogram. In fact, the slope of the curve is derived from Eq. (1) and is given by d D .K/.n/. P /n 1 : d P

(6)

1739

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Thixotropy of Bentonite Suspensions

Figure 7. Shear stress-shear rate thixotropic curves for 6.42% and 5% Wyoming suspension with 0.1 M salt. pH of the suspensions was 8.14 and 8.19, respectively.

The slopes of all curves have been computed for all conditions and are shown versus the shear rate in Figure 10 for the 6.42% concentration for both bentonites. All slopes for Wyoming bentonite are larger than the slopes for the Zenith bentonite for all conditions, meaning more shear thinning behavior of the former with respect to the latter. The degree of thixotropy is estimated from the variation in the slopes among the three curves (D0, U,

V. C. Kelessidis

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1740

Figure 8. Goodness-of-fit of Herschel-Bulkley equation to rheological data of all bentonite suspensions for 6.42% bentonite concentration. (a) Correlation coefficients, (b) normalized sum of square errors.

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Thixotropy of Bentonite Suspensions

1741

Figure 9. Yield stresses computed from the different rheograms for the first down curve (D0), the up curve (U), and the second down curve (D) for all bentonite suspensions.

and D) that is observed, and there is less variation at high shear rates than at lower shear rates for both bentonites. Similar results have been obtained for the 5.0% concentration. The thixotropy indices, computed as described above from the area between the U and the D curves, are shown in Figure 11 as a function of salt concentration for all samples tested. Most samples show positive thixotropy with the exception of the 5.0% Zenith suspension at 0.1 M salt concentration, which exhibits negative thixotropy. The 5.0% suspensions with 0.0 M salt of both bentonites and the 6.42% Zenith suspensions with 0.1 M salt show negligible thixotropy. The Zenith samples of both concentrations and the 5% Wyoming samples show a maximum thixotropy index at the 0.01 M salt concentration, while the 6.42% Wyoming suspension shows a local minimum at the same salt concentration. When no salt is present, there is no difference in the thixotropic indices among the two bentonites tested, with the indices having high values for the 6.42% bentonite concentration and very low values for the 5.0% bentonite concentrations. For the suspensions with salt, differences are observed among the two bentonites, with the most significant at 0.01 M and for the 6.42% bentonite concentration.

Discussion The results presented above show specific trends, with all of these systems exhibiting complex thixotropic behavior, meaning combination of positive and negative thixotropy (Hou et al., 1998; Li et al., 2003). For both bentonites and both concentrations with no salt in the suspension (0.0 M), the initial rheograms obtained (curves D0) after intensive preshearing give rheograms with the lowest shear stresses at all shear rates, compared to the subsequent U and D curves. The D0 curve is always lower than the U curve, indicating rebuilding of the three-dimensional structure of the suspensions, thus giving anti-thixotropic behavior. The structure rebuilding primarily results from clay platelet interactions (Perret et al., 1996), probably from face-to-face or edge-to-edge associations because the pH of the suspensions is greater than the pH of the isoelectric point of these montmorillonite suspensions (van Olphen, 1977; Tombácz and Szekeres, 2004). Anti-thixotropy of clay suspensions, however, has also been attributed to the build-up of

V. C. Kelessidis

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1742

Figure 10. Slope of the rheograms for the Wyoming and Zenith 6.42% suspensions.

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

Thixotropy of Bentonite Suspensions

1743

Figure 11. Thixotropy index, the area of the hysteresis loop between the U and the D curves as a function of salt concentration for the two bentonites and the two concentrations.

edge-to-face structures at high shear rates, which hold larger void volume in the structure (Heckroodt and Ryan, 1978), but this is not the case in the particular work because it is observed at low shear rates. For the low salt concentration of 0.01 M, similar behavior to the behavior exhibited by the 0.0 M suspensions is observed as far as anti-thixotropy is concerned, with the departure of the U curve from the D0 curve being more significant at the highest bentonite concentration, and more intensive for Zenith compared to Wyoming bentonite. Further shearing to get the thixotropic loop, destroys the structure again for both 0.0 M and 0.01 M salt concentrations, as it has been also reported by Perret et al. (1996), and the material exhibits normal thixotropic behavior after it has been subjected to high shear rates and then to low shear rates. Thus, the significant step to observe restructuring, particularly for clay suspensions, is that the material should be sheared at low shear rates to allow partial rebuilding, as demonstrated also by Li et al. (2003). The intensive preshearing before measurements, where the material is subjected to an estimated shear rate of 2,300 sec 1 for an extended period of time (5 min), most probably has broken down the structure completely. Restructuring, however, is observed as the shear rates are decreased and only after 6.5 min have passed, which was the time necessary for the completion of the D0 curve. Following then with the U curve, destruction of the structure is observed again at shear rates of similar magnitude as during preshearing (1,021 sec 1 ). The degree of destruction diminishes with decreasing bentonite concentration, similar to the results reported by Singal and Malik (1964) and Nakaishi and Yasutomi (1994). In addition, the degree of destruction diminishes with increasing salt concentration (up to 0.01 M) and is more intense for Wyoming bentonite compared to Zenith bentonite. It has been suggested that, if the flow of bentonite suspensions is strong enough, it will partially destroy the structure, leading to a viscosity decrease and thus enabling faster structure destruction (avalanche effect), while the opposite holds at low shear rates,

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1744

V. C. Kelessidis

with viscosity increasing as the shear rates become less (Møller et al., 2006). Hence, equilibrium between rebuilding at low shear rates and destruction at high shear rates would be expected with this cycling process. The time of shearing, however, for this to occur depends on the materials used and can be significant, as even after 1 h of shearing, equilibrium was not reported by Møller et al. (2006). The duration of the full test in this work, including the time spent to preshear the material, was 24 min; hence, equilibrium between destruction and restructuring could not be expected during testing in this work. The suspensions with 0.1 M salt concentrations reversed the above trends for both concentrations and both bentonites tested. At this high salt concentration, flocculation of the bentonite particles is expected (Luckham and Rossi, 1999), and the suspensions exhibit either very small thixotropy when going from the D0 to the U curve, especially the Zenith bentonite samples, or no thixotropy at all for all other cases. In conditions of high salinity, the hysteresis loop also disappeared in the study of Perret et al. (1996), and the authors indicated that the salt-induced flocculation structure was built up very rapidly and reversibly when shearing, in a similar fashion observed in this study, either by preshearing or when shearing during measurements. The presence of electrolyte at high concentration leads to the compression of the electric double layer around the clay platelets (van Olphen, 1977; Luckham and Rossi, 1999; Tombácz and Szekeres, 2004), and since no thixotropy is observed at this concentration for both bentonites and both concentrations, it can be concluded that it is the electric double layer, at least partially, responsible for the thixotropic behavior of these bentonite suspensions. Furthermore, the more bentonite particles in the suspensions, the stronger the thixotropy, while the variation observed in thixotropy for the different bentonites may be attributed to the fact that one is naturally occurring sodium montmorillonite while the other one is treated. The antithixotropy observed is not due to shearing, as it has been reported for various flocculated systems (Barnes, 1997) but due to the cycling of shearing and the passing through the low shear rates. This conclusion is significant for the drilling fluids because after they are sheared through the drill bits at very high shear rates, they experience very low shear rates in the annulus for prolonged periods, even hours, and during this time, structure will be rebuilt, significantly increasing the viscosity. Thus, thixotropic models will have to be implemented into the modeling of flow behavior of drilling fluids.

Conclusion The thixotropic behavior of water-bentonite suspensions has been investigated at different bentonite concentrations in the presence of NaCl. These suspensions have exhibited thixotropic behavior, despite intensive preshearing, with the degree of thixotropy depending on the bentonite used, the bentonite concentration, and the presence and amount of salt. Three rheograms have been derived for each suspension after intensive preshearing, from high to low, from low to high, and from high to low shear rates. The HerschelBulkley rheological model described very well the experimental data, and for each of the three curves, the three model parameters have been estimated. No significant variations have been seen for the yield stresses of the suspensions among the three obtained curves. The rheogram slopes, however, indicative of the combined variation of the flow consistency and flow behavior indices, differ, with fewer variations at high than at low shear rates. Negative thixotropy is observed for all suspensions at 0.0 M and 0.01 M but not at 0.1 M salt concentrations when creating the hysteresis loop from high to low to

Thixotropy of Bentonite Suspensions

1745

high shear rates, indicating partial rebuilding of the structures because of the exposition of the suspensions to low shear rates. Thixotropic indices computed from the normal thixotropic loop, low to high to low shear rates, have been computed and vary from positive to negative. The higher the bentonite concentration, the higher the index for the suspensions with no salt, while in the presence of salt, variations of the thixotropy index are obtained among the two bentonites, with the degree of variation dependent on bentonite concentration.

Acknowledgment Experimental data has been collected by Mrs. C. Tsamantaki.

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

References Ambrose, H. A., and Loomis, A. G. 1933. Fluidities of thixotropic gels: Bentonite suspensions. Physics 8:265–273. American Petroleum Institute Specifications 13A. 1993. Specification for drilling fluid materials, Washington, DC. American Petroleum Institute Specifications 13I. 2000. Recommended practice standard procedure for laboratory testing of drilling fluids, Washington, DC. Barnes, H. A. 1997. Thixotropy—A review. J. Non-Newtonian Fluid Mech. 70:1–33. Bekkour, K., Leyama, M., Benchabane, A., and Scrivener, O. 2005. Time-dependent rheological behavior of bentonite suspensions: An experimental study. J. Rheol. 49:1329–1345. Bourgogne, Jr., A. T., Chenevert, M. E., Millheim, K. K., and Young, Jr., F. S. 1991. Applied Drilling Engineering. SPE Textbook Series, Vol. 2, Richardson, TX. Chang, C., and Smith, P. A. 1996. Flow-induced structure in a system of nuclear waste stimulant slurries. Rheol. Acta 35:382–389. Chavan, V. V., Deysarkar, A. K., and Ulbrecht, J. 1975. On phenomenological characterization of thixotropic behavior. Chem. Eng. J. 10:205–214. Cheng, D. C.-H. 2003. Characterization of thixotropy revisited. Rheol. Acta 42:372–382. Cheng, D. C.-H., and Evans, F. 1965. Phenomenological characterization of the rheological behavior of inelastic reversible thixotropic and antithixotropic fluids. Br. J. Appl. Phys. 16:1599–1617. Coussot, P., and Piau, J.-M. 1994. On the behavior of fine mud suspensions. Rheol. Acta 33:175– 184. Dullaert, K., and Mewis, J. 2006. A structural kinetics model for thixotropy. J. Non-Newtonian Fluid Mech. 139:21–30. Galindo-Rosales, F. J., and Rubio-Hernandez, F. J. 2006. Structural breakdown and build-up in bentonite dispersions. Appl. Clay Sci. 33:109–115. Heckroodt, R. O., and Ryan, W. 1978. Clay suspensions with negative thixotropy. Trans. J. Brit. Ceram. Soc. 77:176–183. Hou, W. G., Sun, D. J., and Han, S. H. 1998. Study on the thixotropy of aluminum magnesium hydroxide-Na-montmorillonite suspension. Colloid Polym. Sci. 276:274–277. Kanai, H., and Amari, T. 1994. Negative thixotropy of ferric oxide suspensions. Rheol. Acta 3:303– 310. Kelessidis, V. C., Maglione, R., Tsamantaki, C., and Aspirtakis, Y. 2006. Optimal determination of rheological parameters for Herschel-Bulkley fluids and impact on pressure drop, velocity profiles, and penetration rates during drilling. J. Pet. Sci. Eng. 53:203–224. Kelessidis, V. C., Mihalakis, A., and Tsamantaki, C. 2005. Rheology and rheological parameter determination of bentonite-water and bentonite-lignite-water mixtures at low and high temperatures. Proceedings of the 7th World Congress of Chemical Engineering, Glasgow, July 10–14.

Downloaded By: [HEAL-Link Consortium] At: 14:47 6 August 2008

1746

V. C. Kelessidis

Kok, M. V., and Alikaya, T. 2003. Rheological evaluation of polymers as drilling fluids. Pet. Sci. Technol. 21:113–123. Kok, M. V., and Alikaya, T. 2004. Determination of rheological models for drilling fluids (a statistical approach). Energy Sources 2:153–165. Kok, M. V., and Alikaya, T. 2005. Effect of polymers on the rheological properties of KCL/polymer type drilling fluids. Energy Sources 27:405–415. Labanda, J., Marco, P., and Llorens, J. 2004. Rheological model to predict the thixotropic behavior of colloidal dispersions. Colloids Surf. A: Physicochem. Eng. Aspects 249:123–126. Lemke, T., Bagusat, F., Kohnke, K., Husemann, K., and Mo, H.-J. 1999. Time dependent viscosity of concentrated alumina suspensions. Colloids Surf. A: Physicochem. Eng. Aspects 150:283– 287. Li, S. P., Hou, W. G., Xiao, J. C., Hu, J. F., and Li, D. Q. 2003. Influence of measuring conditions on the thixotropy of hydrotalcite-like/montmorillonite suspension. Colloids Surf. A: Physicochem. Eng. Aspects 224:149–156. Luckham, V., and Rossi, S. 1999. The colloidal and rheological properties of bentonite suspensions. Adv. Colloid Interface Sci. 82:43–92. McMillen, E. 1932a. Thixotropy and plasticity. I. The measurement of thixotropy. J. Rheol. 3:75– 94. McMillen, E. 1932b. Thixotropy and plasticity. II. An empirical equation expressing thixotropic fluidity changes. J. Rheol. 3:163–178. McMillen, E. 1932c. Thixotropy and plasticity. III. The effect of thixotropy upon plasticity measurements. J. Rheol. 3:179–195. Mewis, J. 1979. Thixotropy—A general review. J. Non-Newtonian Fluid Mech. 6:1–20. Møller, P. C. F., Mewis, J., and Bonn, D. 2006. Yield stress and thixotropy: On the difficulty of measuring yield stress in practice. Soft Matter 2:274–283. Moore, F. 1959. The rheology of ceramic slips and bodies. Trans. Br. Ceram. Soc. 58:470–494. Mujumdar, A., Beris, A. N., and Metzner, A. B. 2002. Transient phenomena in thixotropic systems. J. Non-Newtonian Fluid Mech. 102:157–178. Nakaishi, K., and Yasutomi, R. 1994. Rheological study on thixotropic flow: Effect of time scale on flow type measurement. Appl. Clay Sci. 9:71–79. Nguyen, Q. D., and Boger, D. V. 1985. Thixotropic behavior of concentrated bauxite residue suspensions. Rheol. Acta 24:427–437. Perret, D., Locat, J., and Martignoni, P. 1996. Thixotropic behavior during shear of a fine-grained mud from Eastern Canada. Eng. Geol. 43:31–44. Roussel, N., Le Roy, R., and Coussot, P. 2004. Thixotropy modeling at local and macroscopic scales. J. Non-Newtonian Fluid Mech. 117:85–95. Sestak, J., Houska, M., and Zitny, R. 1964. Mixing of thixotropic fluids. J. Rheol. 26:459–476. Singhal, J. P., and Malik, W. U. 1964. Studies on the thixotropic behavior of “Akli” bentonite suspensions. Rheol. Acta, Band 3, Heft 8, 127–131. Speers, R. A., Holme, K. R., Tung, M. A., and Williamson, W. T. 1987. Drilling fluid shear stress overshoot behavior. Rheol Acta 26:447–452. Tombácz, E., and Szekeres, M. 2004. Colloidal behavior of aqueous montmorillonite suspensions: The specific role of pH in the presence of indifferent electrolytes. Appl. Clay Sci. 27:75–94. Toorman, E. A. 1997. Modeling the thixotropic behavior of dense cohesive sediment suspensions. Rheol. Acta 36:56–65. van Olphen, H. 1977. An Introduction to Clay Colloid Chemistry. New York: Wiley.