Slag/mud mixtures improve cementing operations in China

Dec. 23, 1996
Dahua Wu, Peiyan, Bozong Huang Chinese National Petroleum Corp. Tanggu, China The use of blast furnace slag, which is inexpensive and widely available throughout China, can with proper activators and retarders, solidify mud into an excellent cementing material. In many cases, the slag/mud slurry outperforms conventional Portland cements in oil and gas wells. Mud conversion to cement technology has been studied for half a century and has attracted the attention of the drilling and cementing
Dahua Wu, Peiyan, Bozong Huang
Chinese National Petroleum Corp.
Tanggu, China
The use of blast furnace slag, which is inexpensive and widely available throughout China, can with proper activators and retarders, solidify mud into an excellent cementing material.

In many cases, the slag/mud slurry outperforms conventional Portland cements in oil and gas wells.

Mud conversion to cement technology has been studied for half a century and has attracted the attention of the drilling and cementing industry worldwide, because of its economic, technical, and environmental benefits.1-3 Particularly since slag-mix technology was developed by Shell Oil Co. in 1991, it has been successfully used on more than 163 cementing jobs including primary, temporary abandonment, and sidetrack plug cements.4 5

The use of slag-mix has been somewhat controversial. Some experts claim slag-mix is the most important progress to date in mud-to-cement conversion and has become another choice for cementing practices.6 They also believed there were no fundamental limitations to its application downhole, and conceivably the material could be used for any well cemented. Other experts have different points of view and thought it might have limits for oil field use.7 In their studies, the basic mud had to be diluted by 60% or more with water before the blast furnace slag (BFS) was added. Their slag slurries showed a high incidence of cracking and apparent brittle nature, bad settling stability, and volume shrinkage.

To date, the Chinese National Petroleum Corp. (CNPC) has used mud solidification by slag successfully on 22 cementing jobs in the Sichun, Changqing, Jidong, and Shengli oil fields. The major purpose of these investigations was to determine the application of slag-mix technology to various cementing operations.

Blast furnace slag

The chemical composition, character, and other information on BFS have been reported in the literature.4 8-9 The BFS used in this study was the byproduct of China Capital Steel & Iron Co. (Table 1 [31159 bytes]).

Before the lab testing and field application, the slag had to be dried and ground to the desired fineness. A special activator system, BA-1 and BA-2, was adopted instead of soda ash and caustic. As a result, the BFS's fineness of 230 sq m/kg, instead of 400-650 sq m/kg, was sufficient for the slag and mud mixture to set and form cement with enough compressive strength.

Table 1 [31159 bytes] also lists the chemical composition of BFS discharged from 41 steel and iron plants in China in 1990. The total BFS production exceeds 30 million metric tons per year in China. BFS comes from more than 21 provinces, covering a greater area than that by oil well cement producers. The transportation costs for BFS is almost half that for conventional oil well cement.

The average quality coefficient Ko is 1.67. Ko is the oxides content ratio of BFS and represents the hydraulic ability of BFS:

Ko = (CaO + Al2O3 + MgO)/(SiO2 + MnO + TiO2)

The standard deviation is low, 0.149. BFS with Ko more than 1.4 makes up about 95% of the total produced, which is suitable for oil well cementing. Therefore, BFS is an available and cheap cementing material in comparison with oil well cement.

Blast furnace slag is a latent hydraulic material, which can be hardened by chemical activators or thermal energy. The chemical activators include alkali metal oxides and hydroxides, or Portland cement, combined with various salts and can control the setting time of slag/mud mixture. In general, a caustic, soda ash, and lime system is adopted, but they lead to severe flocculation of the mud system (Table 2 [45357 bytes]). This flocculation is the reason why the mud must often be diluted by water to accommodate the BFS. This heavy dilution results in little environmental impact from a BFS/mud mixture.7

The BA-1 and BA-2 system, a complex metal hydroxide, provides the minimum readings of Fann-35 viscometer of modified mud and BFS slurries, and the maximum 24-hr compressive strength development (keeping the same activator concentration with Na2CO3 and NaOH). So the dilution of mud has not been needed usually in CNPC's mud solidification experiences, unless the drilling fluid contained a high concentration of active clay. Obviously, the selection of the activator system is critical to the environmental benefits of slag-mix technology.

Mud solidification

From the Dagang and Bohai oil fields, three typical field muds were taken from various points in the drilling process and were converted into cement slurries by adding BFS, activators, and other chemical additives. The muds were a 1,160 kg/cu m freshwater/bentonite/hydrolyzed polyacrylonitrile (HPAN) polymer spud mud used for the surface hole, a 1,135 kg/cu m fresh water/

bentonite/polyacrylamide (PAM) and carboxymethyl cellulose (CMC) polymer mud used at the intermediate casing point, and a 1,920 kg/cu m sea water/KCl polymer mud used at the final stages of a well (Table 3 [90056 bytes]).

The BFS/mud mixtures were prepared according to the following procedure:

    1. Combine drilling mud and required dilution water, and mix. This process is needed only for high-density mud systems (Formulation 3) because of its high concentration of active clay and drilling solids.

    2. Mix activator BA-1 and BA-2, dispersant, and retarder for 2 hr or more to make the additives uniformly dissolve in the modified mud.

    3. Slowly add BFS during a 1-min period at low mixing speed, then continue mixing for another 35 sec at high speed.

    4. Pour the BFS slurries into rheometer cups, molds, and other test devices to determine the BFS slurries' properties according to American Petroleum Institute (API) recommended procedures.10 Because no cracking occurs in the tested BFS slurries under atmospheric or pressurized conditions, the strength development was measured with API standard crush tests.

BFS slurry thickening time can be tailored to satisfy the various downhole temperature demands by controlling the concentration of activator, dispersant, and thinner (also a retarder). The thickening time curve improves when the consistency can be developed from 30 BC (Bearden unit of slurry consistency) to 100 BC in less than 30 min. (The slurries often hardened in the cup before the chamber was cooled for disassembly.) The quick thickening time is beneficial for wells in which gas migration is a concern, because gas invasion often occurs at the transition period of liquid to solid.

Strength tests were performed on samples cured at 20° C. higher or lower than bottom hole circulating temperature. Table 3 [90056 bytes] also shows that BFS can be set to form enough compressive strength over a wide temperature range.

One advantage of this system, compared with conventional Portland cement, is that it can eliminate two-stage cementing operations and save material costs and rig time.9

The fluid-loss results indicate that BFS/mud mixtures also meet the demand of various cementing jobs on fluid loss control. It depends upon the fluid loss of the solidified mud and compatibility of mud fluid loss controller with BFS.

Settling stability

Settling stability has become an essential property for cementing fluids, especially for slurries designed and used in deviated and horizontal well bores, because unstable cement slurries or slag mixtures cannot provide effective zonal isolation.

The procedure to evaluate the settling stability is to measure API free water content.10 Free fluid is neglected because it cannot set to prevent or inhibit communication of fluid between the various formations penetrated by the well, and downhole temperature and pressure are excluded. This procedure is not a convincing measure of stability.

In 1990, BP Exploration developed a new method to test slurry stability by measuring the density segregation of setting cement. This method combined aspects of temperature, pressure, and deviation in a unique chamber to simulate downhole conditions.11 Settling particles often occur in low-density cement slurries, slurries with dispersants or retarders, and high-density slurries with weighting agents.

The finished BFS slurries of Formulations 1 and 3 in Table 3 [90056 bytes] were poured into a modified BP apparatus (20-cm cement column), placed and cured in a pressurized chamber for 24 hr at 20° C. higher than BHCT, and then disassembled and cut into seven sections to determine every section density.12 In this test, the volume shrinkage is also able to be measured if it exists.

Fig. 1 [54931 bytes] shows that volume shrinkage and particle settling barely occur in BFS slurries. Columns V and VI had little or no shrinkage, and there was only slight density variance between the top and bottom sections. In contrast, the Portland cement-based slurries (I-IV) had shrinkage and density variations. Note that the density variance between the top and bottom in sample VI is only 7 kg/cu m, even though its density is as high as 2,160 kg/cu m. It is difficult to formulate a conventional Portland cement and weighting agent to ensure the same settling stability as that of the slag mixture.

Fracture toughness

All set cements, including Portland cement, fiber cement, slag cement, and slag/mud mixtures, are intrinsically brittle materials. Their difference lies in fracture toughness, which is the resistance to crack propagation.

Excessive brittle cementing material will affect perforation performance. Fracture toughness of set cement can be evaluated by stress-strain curves and critical energy strength factor (Eic). CNPC has performed a great number of stress-strain tests on cement, latex-cement, and fiber cement. Eic is defined as the mechanical energy that specific volume samples (50.8 3 50.8 3 50.8 mm, except the shrinkage samples) can bear divided by compressive strength, while the samples are destroyed. The greater the Eic, the more acceptable the toughness of material.

Similarly, the flatter the stress-strain curve, the more acceptable the cement. These tests were conducted at a 100 kN/min loading rate. Table 4 [44516 bytes] shows that Eic of the slag mixture is between Class G cement and fiber cement. Its stress-strain curve is also sharper than that for fiber cement but flatter than that of Class G cement.

Slag or cement mixed with mud can provide more Eic than that of cement mixed with water. This is likely because the plastic, hydrated clay links slag or cement particles to improve the resistance to crack propagation. This also helps explain the higher compressive strength and improved volume stability for slag or cement/mud mixtures.

Microstructure

A scanning electron microscope (SEM) was used to examine the neat slag or cement slurry, slag mixture, and cement/mud mixture. These samples were prepared in the same manner and formulation as those in Table 4 [44516 bytes], but they were cured for 8 days.

The neat slag slurry had a network of interconnected particles, and fiber-like hydration products were discovered on the surface of slag particles. The slag mixture had an integrated structure composed of closely accumulated plates without holes between the plates. This resulted because the clay participated in structure formation by filling up the space between slag particles.

Class G cement had images similar to those of neat slag slurry, but it was composed of irregular particles. The hydration products were like needles formed on the surface, and more holes were observed than in the slag. No visible differences were discovered between the SEM images of the cement/mud mixture and neat cement slurry. The cement flocculated the hydrated clay.

These microstructure examinations helped explain the results from some of the tests above. Moreover, the SEM photos helped explain the lower permeability of slag mixture than cement.13

Strength retrogression

Both slag and cement can hydrate to form a gelatinous calcium silicate hydrate called C-S-H gel.14 At temperatures above 110° C., C-S-H gel is subject to metamorphosis and often converts to a phase called alpha dicalcium silicate hydrate (a-C2SH), which usually results in decreased compressible strength and increased permeability of the set cement. This phenomenon is known as strength retrogression.

The strength retrogression problem can be prevented by reducing the bulk time to silica ratio (C/S) in the cement slurry. Silica flour or fine silica sand is typically added to Portland cement slurry by 30-45% (by weight of cement), to reduce the C/S ratio to 1.0 or more preferably to 0.6 for higher curing temperatures. The C-S-H gel gradually forms minimal deterioration minerals known as tobermorite (C5S7H), xonotlite (C6S6H), gyrolite (C6S7H2), and truscotlite (C7S12H3), instead of converting to a-C2SH while the curing temperature is increased from 110 to 250° C.15

The C/S ratio of slag is less than that of Portland cement (determined from the chemical composition), so the BFS slurries will provide more temperature stability than cement. To make the BFS slurries suitable for deep oil and gas wells, geothermal wells, and thermal recovery well cementing jobs, the following two questions must be answered:

  • How much silica flour must be introduced into the slag mixture to prevent strength retrogression at elevated temperatures?

  • Can the strength retrogression be determined from the rule of keeping the C/S ratio between 0.6 and 1.0?

    Serial tests were performed on neat Class G cement and slag slurries with and without silica flour (Table 5 [39789 bytes]). The samples were hardened at 52° C. and atmospheric pressure for 24 hr and then cured for 3 days or further cured at 250° C. and 21 MPa for 3 days. The comparison of the compressive strength measured from various cured programs indicates that neat Class G cement and slag slurries exhibit complete strength retrogression, but it is more obvious in the former than the latter.

    Formulations 2-4 with silica flour above 10% (by weight of slag) retained high temperature strength, in accordance with theoretical data. Although long-term performance of BFS mixtures at elevated temperatures were not examined, the drilling literature has shown that there are no visible or tactile fractures and physical deterioration of BFS samples with silica during the 1-year-long retrogression study at 232° C.1

    Operating safety

    The sensitivity of BFS mixtures to additives and slag sources is one of the drilling industry's concerns. Although different additives and slag sources have severe influences on slurry properties, especially on thickening time, it should be a function of the material itself, not slag-mix technology.

    Moreover, there is no problem with using the same source of BFS and additives in field operations as those used in the lab tests. If different material sources were used in field operations and in lab tests, cementing jobs could have inconsistent results, even if conventional cement slurry were used.

    BFS should be mixed into a uniform material, similar to conventional cement, before use. Also, unstable additives should be avoided for both slag mixtures and conventional slurries.

    The different BFS sources have large influences on the compressive strength and flowability of mud/slag mixtures because of their different glass content and chemical components. The different sources of BFS had little affect on thickening time, which depends primarily on activators and retarders.

    Another concern was that the activator is an alkaline material, which requires caution from the field personnel. Whenever alkaline materials are used in the drilling process, extreme care must be exercised.

    The basic mud is a complex system with variable density, clay content, drilling solids, and additives. The content can change from well to well, making the formulation design for BFS slurries a time-consuming job. Consequently, solidified mud must be isolated. Considerable research is needed on the influence of these factors on slag slurry performance. When easily converted mud systems and additives are used, and the BFS slurry designers participate in the design, maintenance, and operation of drilling mud, this problem should be solved or minimized.

    The drilling mud is not as complex when wells are drilled in the same area because the drilling company often tends to use similar additives and mud composition. Thus, the BFS slurry formulation can refer to previous work on offset wells to reduce design time.

    Table 6 [28188 bytes] shows pertinent data from the slag slurries used on two adjacent wells, Well JZ-16-2-1 and Well JZ-16-4-2. Both were drilled with seawater/KCl/polymer mud systems by the Bohai Drilling Co. Although the densities were different, the formulations of their finished BFS slurries were amazingly similar. The slurries were designed for cementing at 95° C. BHCT.

    To examine the operating safety of the BFS slurry designed for JZ-16-2-1 in field use, the effects of fluctuating of activator, retarder, and BFS concentration on the finished slurry properties were studied (Table 7 [28192 bytes]). In field operations, it is possible that the concentration of BFS and additives do not completely match design, resulting in an uneven mix. Table 7 shows that the thickening time shortens by 31 min if the concentration of activator BA-1 is decreased by 1.5 kg/cu m. If the concentration of BA-1 is increased by 1.5 kg/cu m, the compressive strength drops, but to an acceptable level. The same conclusion can be drawn with increasing or decreasing the concentration of retarder MT-1 and BFS.

    Results

  • In China, blast furnace slag is an inexpensive and available cementing material.

  • Slag-mix technology should be considered not only as a viable mud solidification process, but also as a replacement for conventional Portland cement in some operations. But a significant effort should still be put forth to optimize slag-mix technology.

  • The amount of dilution needed depends on the adopted activator and active clay content in the basic mud system. If the BA-1 and BA-2 activator system is used, the mud generally does not need to be diluted, unless the drilled solids and active clay content are high.

  • If an appropriate activator is used, no volume shrinkage, particle settling, or cracking occurs in slag slurries.

  • Fracture toughness of slag mixture is between that of neat Portland cement and fiber cement, so the same attitude must be adopted toward use of slag with cement.

  • The slag mixture is more temperature stable than Portland cement. The addition of silica flour (10% by weight of slag) is enough for the BFS slurry to prevent strength retrogression at 250° C. The rule of keeping the C/S ratio of slag slurries between 0.6 and 1.0 can also be followed in designing a slag slurry for elevated temperatures.

  • Scanning electron microscope images indicate the hydrated clay participates in the interconnection of slag particles. Consequently, the slag/mud system has higher compressive strength, more acceptable toughness, and lower permeability than the neat slag system. Because cement flocculates the hydrated clay, there is no visible microstructural difference between the SEM images of cement/mud and neat cement. The effect of hydrated clay on cement slurry performance is less obvious than that of slag.

    Acknowledgment

    The authors wish to thank Li Baogui, Li Xizeng, Gao Yunhui, Gaoyu, and Zhangying at the Research Institute of Chinese National Petroleum Corp. for their help with this work. The authors also thank the directors of the Research Institute of CNPC for permission to publish this article.

    References

    1. Schlenmer, R.P., Branam, N.E., Edwards, T.M., and Valenziano, R.C., "Drilling Fluid Conversion: Mud Selection and Conversion Techniques," Paper 26324, presented at Society of Petroleum Engineers 68th Annual Technical Conference and Exhibition, Houston, Oct. 3-6, 1993.

    2. Wilson, W.N., Miles, L.H., Boyd, B.H., and Carpenter, R.B., "Cementing Oil and Gas Wells Using Converted Drilling Fluid," USP 4, 883, 125 (1989).

    3. Nahm, J.J., Javanmardi, K., Cowan, K.M., and Hale, A.M., "Slag Mix Mud Conversion Cementing Technology: Reduction of Mud Disposal Volumes and Management of Rig-site Drilling Wasters," Journal of Petroleum Science and Engineering, Vol. 11, No. 1, 1994.

    4. Hale, A.H., and Cowan, K.M., "Solidification of Water-Based Muds," USP 5, 048, 679 (1991).

    5. Leimkuhler, J.M., Rembow, F.H.K., Warren, P.B., Javanmardi, K., Ladner, B.S., and Smith, T.R., "Downhole Performance Evaluation of Blast Furnace Slag-Based Cements: Onshore and Offshore Field Applications," Paper 28474, presented at SPE 69th Annual Technical Conference and Exhibition, New Orleans, Sept. 25-28, 1994.

    6. Bell, S., "Mud-to-Cement Technology Converts Industry Practices," Petroleum Engineer International, September 1993.

    7. Benge, O.G., and Webster, W.W., "Blast furnace slag slurries may have limits for oil field use," OGJ, July 18, 1994.

    8. Cowan, K.M., Hale, A.H., and Nahm, J.J., "Conversion of Drilling Fluid to Cements with Blast Furnace: Performance Properties and Applications for Well Cementing," Paper 24575, presented at the SPE 67th Annual Technical Conference and Exhibition, Washington, D.C., Oct. 4-7, 1992.

    9. Cowan, K.M., "Solidify Mud to Save Cementing time and Reduce Waste," World Oil, October 1993.

    10. American Petroleum Institute Specification for Materials and Testing for Well Cements, Specification 10, 5th Edition, 1990.

    11. Greaves, C., and Hibbert, A., "Test improves measurement of cement slurry," OGJ, Feb. 12, 1990.

    12. Hu Zehua, Wu Zhongfu, Huang Bozong, "Test of Stability of Oil Well Cement Under HTHP Conditions," Drilling & Completion Fluids, Vol. 9, No. 5, pp. 72-74, 1992.

    13. Pessier, R.C., Javanmardi, K., Nahm, J.J., Leimkuhler, J.M., Dudley II, J.W., and Mody, F.K., "Evaluating Slag Solidified Mud Drillability and Bit Performance," World Oil, October 1994.

    14. Mueller, D.T., and Dickerson, J.P., "Blast Furnace Slag Technology: Features, Limitations, and Practical Application," Paper 28475, presented at SPE 69th Annual Technical Conference and Exhibition, New Orleans, Sept. 25-28, 1994.

    15. Nelson, E.B., "Thermal cements," from the book Well Cementing, edited by E.B. Nelson, Elsevier Science Publishing Co. Inc., 1990.

    The Authors

    Dahua Wu is an engineer with the Research Institute of Engineering & Technology at the Chinese National Petroleum Corp. in Tanggu, Tianjin, China. He has worked in cementing research for more than 5 years and holds an MS in polymer science from Tianjin University.

    Peiyan has worked with the Research Institute of Engineering & Technology at the Chinese National Petroleum Corp. in Tanggu, Tianjin, China, for the past 12 years. She is currently a cementing engineer. Peiyan has a BS in oil field chemistry from the Xinan Petroleum Institute.

    Bozong Huang is a professor and director of the cementing department at the Research Institute of Engineering & Technology at the Chinese National Petroleum Corp. in Tanggu, Tianjin, China. He has worked at the research center for past 20 years. He previously worked as a cement and concrete researcher for the Harbing Institute of the Academy of Science of China. He holds a BS in physicochemistry from Jiling University.

    Copyright 1996 Oil & Gas Journal. All Rights Reserved.