Shear rate has greater influence on cement slurry properties than total mixing energy

Paul Padgett
Halliburton Energy Services
Duncan, Okla.

Because field mixing equipment operates at significantly lower shear rates than typical laboratory equipment, slurry properties in the field frequently do not match the laboratory values, which can lead to a variety of cement-job problems.

Slurries designed under high-shear laboratory mixing conditions can result in the wrong combinations of additives or the wrong additive proportions.

Several years ago, researchers investigated the problem of variations in cement slurry properties between the laboratory and the field. These researchers theorized that when the mixing energy imparted to a slurry is equal for two mixing systems, the resulting slurry properties will also match.¹ Shortly thereafter, this theory was further advanced, and the relationship between rotational speed and residence time on slurry properties was also explored.² Soon after the mixing energy concept was introduced, an extensive series of laboratory and full-scale mixing tests on various cement slurries was conducted.

Laboratory testing was performed with a commercial laboratory blender and dispersator, and full-scale testing was done with a recirculating cement mixing system and downstream chokes.^{3 4} As a result of the testing, the following conclusions were reached:

• Some slurry properties, including rheology, free water, and settling, were responsive to mixing energy. This response did not occur with every slurry tested, however.

• Some slurry properties, such as fluid loss and thickening time, were only slightly affected by mixing energy. In some cases, the fluid loss declined slightly with increased mixing energy; in other cases, it increased. The thickening time, when it was affected, decreased slightly.

• At equal mixing energies, direct comparisons between the different devices (laboratory blender, dispersator, and field equipment) never led to matching results. The free-water values were highest with the field equipment and lowest with the laboratory blender. In all cases, the differences among mixing devices were much more pronounced than the differences among various levels of mixing energy.

• The 24-hr compressive strength was never shown to be a function of mixing energy, and it did not vary between mixing devices.

• Using chokes downstream of a recirculating cement mixer did not change the slurry properties whatsoever.

After this work, the basic conclusion was that mixing energy did help determine slurry properties. Because the different mixing devices influenced slurry properties to a much greater degree than total mixing energy, the conventional mixing-energy concept was found to be only marginally useful.

Coiled tubing

The mixing energy issue surfaced again when cementing through coiled tubing became a widespread practice. Continued failures of squeeze jobs prompted researchers to investigate how energy addition through the coiled tubing affected slurry performance.⁵ This study showed that pumping the slurry through coiled tubing and inattention to detail during batch mixing could lead to shortened thickening times.

Studies on the relationship between mixing energy and batch mixing followed.⁶ Researchers concluded that batch size affected thickening time even though the total mixing energy was held constant.

A series of full-scale tests conducted with coiled tubing did not produce any significant changes in thickening time or other slurry properties.⁷ These tests evaluated three different mixing systems (a standard recirculating cement mixing system or RCM, an RCM with an axial-flow mixing head, and a third-party paddle mixer), three different batch sizes (5, 7, and 21 bbl), and three different slurries.

In only one case were the data outside the expected range. This case was a latex slurry mixed with the paddle mixer. The thickening time was about 17% less than that for the blend and about 25% less than that for the RCM-mixed samples.

The conclusion reached was contrary to the work cited in References 5 and 6. These new tests indicated that slurry performance is not appreciably affected by batch size, mixing pumps, or pumping through coiled tubing and that a properly designed slurry, good mixing equipment, and a common-sense approach to job design and implementation are the essentials for successful cementing.

Recent tests

A series of full-scale tests with field equipment was completed for the sole purpose of investigating the mixing energy issue. The test setup is shown in Fig. 1 [86490 bytes].

Two slurry samples were pulled from each of four different sample points in the system while the system was running in a continuous-mixing mode at 3 bbl/min. Bulk-blend samples were pulled from the conveying line during mixing and were later mixed in the laboratory with a sample of the actual mixing water based on the American Petroleum Institute (API) Specification 10 procedure.⁸

Instruments were used throughout the test setup so that the mixing energy at each sample point and the thermal effects of the energy addition could be determined.

Two different slurries were run: neat Class H cement at 16.3 ppg and Class H cement with a 1% fluid-loss additive at 15.9 ppg. Three runs were made with the neat slurry, and two runs were made with the low fluid-loss slurry.

Operators varied the mixing energy level between the runs by controlling the recirculation pump and agitator speeds. After the holding tank was filled, the slurry was held and agitated for 15 min. The density was maintained at ±0.1 ppg with an automatic density control (ADC) system. Analysis of the data showed no correlation between property variations and these slight density fluctuations.

Slurry property data obtained from the tests are shown in Figs. 2 [11663 bytes], 3 [11772 bytes], 4 [13195 bytes], 5 [16355 bytes], 6 [13768 bytes]. The data points corresponding to the lowest mixing energy values were pulled from the premix compartment of the mixing system, while those with the highest mixing energies were pulled from the holding tank. The solid data points are from the blend mixed in the lab. The data reveal two key points:

• Little correlation exists between total mixing energy and any of the slurry properties measured.

• The free-water data show a strong dependence on the mixing device.

Mixing systems

One of the main differences between the field equipment (RCM) and laboratory equipment (blender) is the shear rate of the mixing environment. The maximum shear rate in an RCM system occurs in the centrifugal pump and is typically less than 2,000 sec-1, while the average is about 500 sec-1. Shear rates in the recirculation lines are approximately 1,500 sec-1, and in the impeller zone of the mixing tank the shear rates are considerably less.

Conversely, the laboratory blender, running at the 12,000 rpm recommended by API Spec 10, produces extremely high shear rates. The maximum is more than 30,000 sec-1, while the average is almost 4,000 sec-1.

Compared to the laboratory blender, the centrifugal pump is not a high-shear device, even though it was considered one in the past.

The shear history of the slurry, which accounts for the number of shear cycles the fluid is exposed to, is also a consideration. It has been suggested that circulating a batch tank with a centrifugal pump represents a cyclical shear, while the laboratory blender operates with continuous shear.⁶ In fact, both situations are cyclical. In the laboratory blender or any agitated vessel, the fluid is exposed to the maximum shear rate in the impeller zone and to considerably lower shear rates at other locations.

The pumping action of the impeller moves the fluid through the higher-shear impeller zone into the lower-shear regions of the vessel. Generally, the average shear rate in the vessel is an order of magnitude less than the average shear rate in the impeller zone. Whether a centrifugal pump or an agitator provides the circulation, the fluid is exposed to a wide variety of shear rates as it moves through the different zones of the mixing system.

Shear rate

With neat cements, the reason that shear rate is more important than total mixing energy is mainly the result of the forces required to break up the agglomerates that form during the initial wetting of the powder. Cement slurries with additives can also form a protective gel membrane around the cement particle that can inhibit hydration.⁹ Shear stresses imparted to the fluid cause the agglomerates to break down and damage or break the gel membrane. The shear stress required may decrease as the number of stress cycles increases; however, for each particle (or group of particles), a certain minimum stress will result in damage.

While examining test results shown in Figs. 2-6, research personnel saw that significant changes in the free water resulted from changes in the shear rate. For the neat slurry, the higher-shear mixing environment produced less free water because of the surface area increase resulting from aggregate and agglomerate breakdown. Conversely, in cement with fluid-loss additive, the high-shear environment produced more free water because of damage to the structure of the fluid-loss additive. In either case, total mixing energy was inconsequential.

Shear rate effects on thickening time have been investigated.10 On certain slurry types, the thickening time could be reduced by as much as 30-40% when the shear rate is increased. Although these tests were run to simulate different placement conditions and not the mixing process, the data further illustrate the importance of shear rate on slurry properties.

Wetting efficiency

The wetting efficiency of the mixing system plays an important role in the formation of the agglomerates. A system with high wetting efficiency will form fewer and smaller agglomerates. Wetting efficiency is difficult to determine and is not necessarily related to shear rate and mixing power.

Systems that provide high initial shear rates or high initial mixing power are more efficient than those that do not. Because the development of slurry properties is irreversible, applying shear late in the mixing process is often ineffective in improving slurry properties.

Laboratory investigation

To explore the shear rate relationship further, researchers ran a series of simple tests with neat Class H cement using a laboratory blender to vary the shear rate and mixing energy. The slurry was mixed at 16.3 ppg, and the rheologies and free water were measured. The results, shown in Figs. 7 [20005 bytes], 8 [17445 bytes], 9 [19639 bytes], 10 [20685 bytes], indicate the following:

• Free water is only slightly affected by mixing energy and it tends to increase with increasing energy.

• Free water is influenced to a much greater degree by the shear rate, and it tends to decrease with increasing shear.

• Yield point does not follow any particular trend with increasing mixing energy.

• As the shear rate increases, the yield point declines dramatically at the lower shear rates, then tends to increase.

From this small test, it was again concluded that shear rate is more important than mixing energy in controlling the slurry properties.

The mixing-energy theory neglects the importance of shear rate to the cement slurry mixing process:

E/m = tg

In this equation, E = energy, m = mass, = viscosity, t = time, and g = shear rate.

The equation demonstrates that if the residence time is increased, a low-shear device (jet mixer, batch mixer, etc.) can exert the same amount of mixing energy into a slurry as a high-shear device (laboratory blender). However, because it is the shear rate that is important to the slurry, the resulting properties will not necessarily be the same.

Results

The cement mixing process is still not completely understood. The various cement types and large number of additives contribute significantly to the complexity. Continued research will result in a better understanding of the mixing process; this research should provide better techniques for the design and testing of cement slurries and will lead to more accurate correlations between laboratory and field slurry performance.

References

1. Orban, J.A., Parcevaux, P.A., and Guillot, D.J., "Specific Mixing Energy: A Key Factor for Cement Slurry Quality," Paper 15578 presented at the 61st Annual Society of Petroleum Engineers Technical Conference and Exhibition, New Orleans, Oct. 5-8, 1986.

2. Vidick, B., "Critical Mixing Parameters for Good Control of Cement Slurry Quality," Paper 18895 presented at the SPE Production Operations Symposium, Oklahoma City, Mar. 13-14, 1989.

3. Wilson, M.A., "Mixing Energy-Effect on Cement Slurries," Halliburton Internal Report, C46-C001-89, Dec. 31, 1986.

4. Wilson, M.A., "Mixing Energy-Effect on Cement Slurries," Halliburton Internal Report, C46-C001-86, June 6, 1989.

5. Vidick, B., Nash, F.D., and Hartley, I., "Cementing Through Coiled Tubing and Its Influence on Slurry Properties," Paper 20959 presented at Europec '90, Oct. 23-24, 1990.

6. Vidick, B., Hibbert, A.P., and Kellingray, D.S., "Cement Slurry Batch Mixing: A Critical Step for Slurry Quality Control," Paper 7067 presented at the 24th Annual Offshore Technology Conference, Houston, May 4-7, 1992.

7. Heathman, J.F., Sands, F.L., Sas-Jaworsky, A., and Badalamenti, A.M., "A Study of the Effects of Mixing Energy Imparted on Cement Slurries by Field Equipment and Coiled Tubing," Paper 26573 presented at the 68th Annual SPE Conference and Exhibition, Houston, Oct. 3-6, 1993.

8. American Petroleum Institute, "Specifications for Materials and Testing for Well Cements," API Specification 10, fifth edition, July 1, 1990.

9. Rahman, A.A., and Double, D.D., "Dilation of Portland Cement Grains During Early Hydration and the Effect of Applied Hydrostatic Pressure on Hydration," Cement and Concrete Research, Vol. 12, 1982.

10. Surjaatmadja, J.B., Lee, L., Ripley, H.E., and Fulton, R.G., "Cement Analysis Under Simulated Pumping Conditions," Journal of Canadian Petroleum Technology, Vol. 32, No. 7, September 1993.

The Author