 |
 |
Hydraulic Conductivity of Compacted Lateritic Soil Treated with Blast Furnace Slag
Kolawole J. Osinubi Associate Professor, Dept. of Civil Engineering, Ahmadu Bello Univ., Zaria, Nigeria kosinubi@yahoo.com Adrian O Eberemu Research Assistant, Dept. of Civil Engineering A.B.U. Zaria, Nigeria |
ABSTRACT
Hydraulic conductivity tests were conducted on four compacted lateritic soil treated with up to 15% in stepped increment of slag content by weight of dry soil. The lateritic soil-slag mixtures were prepared at various moulding water contents, compacted and permeated in the laboratory. The hydraulic behaviour of the soil-slag mixtures were examined in which the effects of moulding water content, water content relative to optimum, initial degree of saturation at compaction, slag content and the Atterberg limits were investigated. The result of the tests showed that in a manner similar to natural soils, the hydraulic conductivity decreased with higher moulding water content. The hydraulic conductivity decreased on the wet side and increased on the dry side of the line of optimum for all compactive efforts used and slag content of the mixtures. Hydraulic conductivity decreased with higher slag content. The initial degree of saturation increased as the hydraulic conductivity decreased, but further increase in initial degree of saturation did not result in decrease in hydraulic conductivity values.
Keywords: Atterberg limits, Blast furnace Slag, Hydraulic Conductivity, Initial Degree of Saturation, Lateritic Soil.
INTRODUCTION
The improved geotechnical characteristics of compacted soils resulting from the beneficial reuse of some industrial by-products may be very important to some geotechnical applications such as engineered fill, pavement structure, landfill covers and lining systems. The development of alternatives for the beneficial reuse of industrial by products mostly brings economical and environmental benefits with the present increase in environmental awareness. Blast furnace slag (BFS), which is the by-product of the steel industry, is abundantly produced by the various foundries and steel industries in Nigeria. However, they have been scarcely used for engineering purposes with the bulk being placed in disposal pits or storage sites.
Compacted soil liners are earthen barriers used to minimize the movement of liquids into or from landfills, surface impoundments, and other facilities that contain materials that can contaminate groundwater. Because the primary purpose of compacted soil liners is to impede the
flow of fluids, the most significant factor affecting its performance is hydraulic conductivity. To be effective, a compacted soil liner should have low hydraulic conductivity, which in many cases is considered to be less than 1 x 10-9 m/s (Daniel 1993). Laboratory studies by Osinubi and Nwaiwu (2002, 2005, and 2006) indicated that fine-grained lateritic soils can be used as hydraulic barrier layers. However, soil property variations within lateritic profiles have been found to affect their use in construction. Therefore, the implication of these variations on the construction of treated lateritic soil liners and covers is necessary.
This paper presents the results of a laboratory study on the hydraulic conductivity of compacted lateritic soil treated with up to 15% ground BFS as a suitable hydraulic barrier material for the containment of environmental pollutants emanating from municipal solid waste (MSW) sites. In addition, work was carried out to investigate how the hydraulic conductivity is affected by the moulding water content, water content relative to optimum, initial degree of saturation, slag content and Atterberg limits.
MATERIALS AND METHOD
Soils
The soil sample used in the study is a natural material that is reddish brown in colour. A study of the geological and soil maps of Nigeria after Akintola (1982) and Areola (1982), respectively, shows that it belongs to the group of ferruginous tropical soils derived from acid igneous and metamorphic rocks. This is in agreement with D’Hoore (1954). Soil sample was obtained from a burrow pit by method of disturbed sampling in Zaria (latitude 11º 15’ and longitude 7º 45’ E), Nigeria. The soil is classified as lean clay with sand (CL) according to the Unified Soil Classification System (USCS) (ASTM D2487). The specific gravity of the soil is 2.76, with percentage passing BS No. 200 sieve is 77.5%. X-ray diffraction (XRD) and differential thermal analysis (DTA) of soil from the study borrow area reported by Osinubi (1998a,b) showed that the clay mineralogy is predominantly kaolinite. The pH of the soil is 6.7. The Atterberg limit on portion passing 425 (m aperture sieve are as follows: liquid limit 42.2%, and plastic limit 32.4%. The soil had natural moisture content of 5.8% at the time of sampling. The particle size distribution curve of the natural soil shown in Fig. 1
Figure 1. Particle size distribution curve for the natural lateritic soil sample.
Slag
Bulk samples of blast furnace slag were obtained from the foundry of Defence Industries Corporation of Nigeria in Kaduna Sate, Nigeria. The samples were ground to particle sizes passing through BS No. 200 sieve before use. The material is non plastic and non hazardous, with typical chemical composition showing 29-42% SiO2, 5-23% Al2O3, 0-21% MgO, 0.1-3.8% Fe2O3, 0.1-22% MnO, 0.4-2.5% S and 29-48% CaO.
Preparation and Testing of Specimens
Compaction
All the specimens tested were prepared by mixing the relevant quantities of dry soil and ground BFS which were percentages of dry weight of soil (i.e., 0%, 5%, 10% and 15%). The maximum slag content used was based on results from several preliminary tests, that showed changes in plasticity and a reduction in maximum dry density after an initial increase and the compaction difficulties found using higher slag content.
In general, specimens were moulded using different moulding water contents and different compactive efforts similar to the various compaction efforts that might be achieved in the field (i.e., the standard Proctor and modified Proctor compactive efforts in accordance with BS 1377: 1990, as well as the West African Standard (WAS) compaction in accordance with the Nigerian General Specification (1997)). Using the 1000 cm3 mould for standard Proctor, five batches of soil passing a 4.75 mm sieve aperture each weighing about 2.5 kg was placed in a tray and mixed with tap water. Compaction was in 3 layers applying 27 blows per layer using a 2.5 kg rammer falling from a height of 300 mm. For the modified Proctor compactive effort, the same mould was used but with a 4.5 kg rammer falling from a height of 450mm applying 27blows/layer and compacting in 5 layers. The WAS compaction level involved the use of the same mould but a 4.5 kg rammer falling from a 450 mm height on five different layers receiving 10 blows each in accordance with Osinubi (1998a).
Hydraulic Conductivity
The hydraulic conductivity test was carried out using the falling head hydraulic conductivity apparatus (Head, 1992). A relatively short sample was connected to a standard pipe, which provided both the head of water and the means of measuring the quality of water flowing through it. Compacted samples at 10%, 12.5%, 15%, 17.5%, and 20% moulding water contents at the three compactive efforts, respectively, were soaked in distilled water for a minimum period of 96hrs to allow for full saturation. During soaking, the samples were not allowed to swell vertically. After saturation, test specimens were connected to a permeant liquid (distilled water) for permeation and hydraulic gradients that ranged from 5 to 15. The test lasted from 2 to 7 days and was only discontinued when hydraulic conductivity readings were steady. Readings were taken intermittently and the change in water height under certain time intervals measured. This test was carried out for all four soil-slag mixes.
RESULTS AND DISCUSSION
The hydraulic behaviour of the soil-slag mixtures was examined in which the effects of moulding water content relative to optimum, initial degree of saturation at compaction and slag content as well as Atterberg limits were investigated.
Moulding Water Content
The relationship between hydraulic conductivity and compaction moulding water contents are shown in Figs. 2–5. In a manner similar to natural clays, the hydraulic conductivity in all cases decreased as the moulding water increased. This continued up to points where the hydraulic conductivity started again to increase. The reason for this initial decrease with the increase in water content can be attributed to increasing water that helped in deflocculating the particle structure thus reducing voids. Consequently, the arrangement of individual particles, which were influenced by the moulding water content, controlled the hydraulic conductivity (Lambe, 1958).
Olsen (1962) further suggested that most of the flow of water in compacted clay occurs in relatively large pore spaces located between pads or clods of clay rather than between the particles of clay within the clods. According to his clod theory, soft wet clods of soil that are easier to remould results in smaller inter clod voids and have lower hydraulic conductivity. At similar compaction water contents, lower hydraulic conductivities were generally obtained with higher compaction efforts. The maximum hydraulic conductivity of 1 x 10-9 m/s required for hydraulic barrier liners were obtained at 13.5% moulding water content for standard Proctor and as well as below 10% moulding water content for both WAS and modified Proctor efforts. In summary, the addition of slag, as well as increases in compaction energy and moulding water content decreased the hydraulic conductivity.
Figure 2. Variation of hydraulic conductivity with moulding water content for natural soil.
Figure 3. Variation of hydraulic conductivity with moulding water content for soil with 5% Slag content
Figure 4. Variation of hydraulic conductivity with moulding water content for soil with 10% Slag content
Figure 5. Variation of hydraulic conductivity with moulding water content for soil with 15% Slag Content
Water Content Relative to Optimum
Hydraulic conductivity varies greatly between the dry and wet side of the line of optimum. It generally decreased on the wet side and increased on the dry side of optimum for all compactive efforts and slag contents. Figs. 6-9 show variations of hydraulic conductivity with water content relative to optimum. The results are consistent with that of natural clay as reported by Mitchell et al. (1965) as well as Benson and Daniel (1990).
The natural soil compacted between 10 and 20% moulding water content at the energy of the standard Proctor gave hydraulic conductivity < 1 x 10-9 m/s from 0.25% dry of optimum to above 3.5% wet of optimum; for WAS compactive effort, from 2% dry to above 4% wet of optimum and for modified Proctor from 3% dry to 7% wet of the line of optimum. For soil treated with 5% slag content and compacted between 10 and 20% moulding water content, at the energy of standard Proctor hydraulic conductivity values of 1 x 10-9 m/s were obtained from 2.5% dry of optimum to greater than 4.3% wet of the optimum line. For WAS effort, from 3% dry to much greater than 7% wet of the line of optimum. For 10% slag treatment, compacted to the same moulding water content as above and at the energy of standard Proctor and WAS, it gave the required minimum hydraulic conductivity and less from –3.6% to 3.56% and 4.5% and above, respectively from the dry to wet side of the line of optimum. Modified Proctor compactive effort gave much less than 2.7% to much more than 7.3%. For 15% slag treatment of soil, compacted at the same moulding water content as those above, < 1 x 10-9 m/s hydraulic conductivity values were recorded for standard Proctor at –3% to above 3.46%; WAS, from below –5.4 to above +4.6 % and, modified Proctor from below –3% to above 7% of the line of optimum from dry to wet.
In all cases of stepped addition of slag to natural soil, the water content relative to optimum on the wet side of optimum increased as the hydraulic content decreased. Furthermore, as the compactive effort increased, the permissible required hydraulic conductivity values shifted to the dry side from the wet side of optimum. The main reason for this behaviour can be explained based on the particle orientation theory proposed by Lambe (1958) and the clod theory proposed by Olsen (1962) that were discussed earlier.
Figure 6. Variation of hydraulic conductivity with moulding water content relative to optimum for natural soil
Figure 7. Variation of hydraulic conductivity with moulding water content relative to optimum for soil with 5% Slag Content
Figure 8. Variation of hydraulic conductivity with moulding water content relative to optimum for soil with 10% Slag Content
Figure 9. Variation of hydraulic conductivity with moulding water content relative to optimum for soil with 15% slag content
Initial Degree of Saturation at Compaction
Initial degree of saturation can be used to examine hydraulic conductivity, because it generally decreases as the water content is increased beyond the line of optimum irrespective of the compactive effort. As the water content is increased the initial saturation increases. Further increase in initial saturation does not always result in decrease in hydraulic conductivity because if the water content is increased far above the optimum, the hydraulic conductivity will increase while initial saturation may remain constant. Furthermore, samples compacted at the same initial saturation but with different compactive efforts will have different fabrics and hence different hydraulic conductivities. When soils are compacted at higher initial degrees of saturation they are compacted wetter relative to the line of optimum and generally have lower hydraulic conductivity.
Variations of hydraulic conductivity with initial degrees of saturation for all samples are shown in Figs. 10 – 13. For the natural material, initial saturation of 64.7% and beyond satisfied the minimum required for barrier materials at 66.9% and above, below 51.2% and beyond 84.6% standard Proctor, WAS and for modified Proctor compactive efforts, respectively. On treatment of soil with 5% slag, initial saturation of 57.2% and beyond, for standard Proctor; 73.6% and beyond at WAS, below 46.9% and beyond 97.3% for modified Proctor compactive efforts all gave the minimum hydraulic conductivity values required. For 10% slag treatment of soil, it was obtained at an initial degree of saturation of 59.5% and beyond for both standard Proctor and WAS compaction efforts, while it was below 68.8% and beyond 99.3% for modified Proctor compactive energy. For 15% slag treated soil, the initial degree of saturation ranged from 41.8 to 96.0%, 47.4 to 99.8%, 59.5 to 100% for standard Proctor, WAS and modified Proctor compactive energies, respectively.
A hydraulic conductivity value of 1·10-9 m/s was obtained for all initial degree of saturation results with modified Proctor and WAS, while it was obtained from 68.8% and above at standard Proctor compactive effort. Generally for all cases of slag treatment, as the initial degree of saturation increased, the hydraulic conductivity decreased and this is in conformity with the works of Boutwell and Hedges (1989), Benson et al. (1994, 1999), and Abichou et al. (2000).
Figure 10. Variation of hydraulic conductivity with initial degree of saturation for natural soil
Figure 11. Variation of hydraulic conductivity with initial degree of saturation for soil with 5% Slag Content
Figure 12. Variation of hydraulic conductivity with initial degree of saturation for soil with 10% Slag Content
Figure 13. Variation of hydraulic conductivity with initial degree of saturation for soil with 15% Slag Content
Slag Content
The effects of slag content for the three compactive efforts at different moulding water contents on hydraulic conductivity are shown in Figs. 14–16. In all cases the hydraulic conductivity decreased with higher slag contents. From Fig. 14 it was observed that values < 1 x 10-9 m/s were obtained at 15%, 17.5%, and 20% moulding water contents, respectively for all the slag treatments. At WAS compactive effort, the hydraulic conductivity values < 1 x 10-9 m/s were obtained at 12.5, 15, 17.5 and 20% moulding water content for 5 to 15% slag treatment of soil. Compaction at the modified proctor compactive energy yielded minimum hydraulic conductivity required in all cases between 0 and 15% slag treatment of soil.
For all the three compactive effort, the hydraulic conductivity decreased with higher slag content, due to the increase in the pH value of the moulding water content as a result of the partial dissociation of calcium hydroxide. The calcium ions in turn combined with the reactive silica or alumina from the laterite or both when they are present, to form insoluble calcium silicate or aluminates or both when they are present, to form insoluble calcium silicates or aluminates or both, blocking the flow of water through soil voids. Furthermore, cementing is supposed to increase with increase in slag content, thus decreasing hydraulic conductivity. This decrease might also be due to precipitation of calcium carbonate in the voids as the ionized calcium reacts with the dissolved carbon dioxide in the water.
Figure 14. Variation of hydraulic conductivity with slag content for standard Proctor compaction.
Figure 15. Variation of hydraulic conductivity with slag content for West African Standard compaction.
Atterberg Limits
Samples compacted at the three energy levels increased in hydraulic conductivity as the liquid limit increased. This is supported by the positive statistical correlation of 0.36 obtained from the correlation matrix between logarithm of hydraulic conductivity and the liquid limit. On the other hand, with increase in plasticity index, the hydraulic conductivity decreased. These trends were expected because liquid limit and plasticity index are related to the mineralogy of the soil. An increase in the presence or more active minerals from higher slag content generally corresponded to a decrease in the size of micro-scale pores resulting from the formation of more insoluble calcium silicate or aluminates, thus leading to reduced hydraulic conductivity.
Figure 16. Variation of hydraulic conductivity with slag content for modified Proctor
CONCLUSION
The relationship between hydraulic conductivity and compaction water content and compaction energy for laterite-slag mix exhibits trends similar to those of natural clays. The hydraulic conductivity decreased as the moulding water content increased. It further decreased on the wet side of optimum and increased on the dry side of optimum irrespective of the compactive effort. As the initial degree of saturation increased during compaction so did the hydraulic conductivity decrease irrespective of the slag content or compaction energy adopted. The hydraulic conductivity decreased as the compactive effort increased further shifting to the dry side of the line of optimum at which the maximum permissible hydraulic conductivity of 1 x 10-9 m/s occurred thus allowing more materials to be compacted dry of the optimum. Increasing the compactive effort reduced the initial degree of saturation at which the maximum permissible hydraulic conductivity occurred. The addition of slag helped in reducing the hydraulic conductivity of all treatments. The hydraulic conductivity reduced as the plasticity index of the soil-slag mixture increased.
REFERENCES
 | |
| © 2006 ejge | |