This study adopted a laboratory-based experimental research design aimed at evaluating the effectiveness of Cement Kiln Dust (CKD) and Lime Kiln Dust (LKD) as stabilizing agents for expansive clay soils used as subgrade material in road construction. The design involved systematic characterization of raw materials, controlled stabilization experiments at varying dosages, and comprehensive geotechnical testing to assess improvements in engineering properties. This experimental approach enabled direct comparison between untreated (neat) and treated (stabilized) soils under standardized laboratory conditions, following relevant ASTM and British Standard (BS) procedures. The methodology was structured to ensure reproducibility, reliability, and scientific rigor.

The primary materials used in this study were expansive clay soils, Cement Kiln Dust (CKD), and Lime Kiln Dust (LKD). Expansive clay soils, representing problematic subgrade materials common in Kenya’s highland regions, were collected from selected road alignment sites in Nyandarua County. CKD and LKD were sourced from major cement and lime manufacturing plants due to their pozzolanic and calcium-rich properties, making them suitable alternatives. All materials were collected, handled, and stored in accordance with ASTM and BS standards to preserve their chemical and physical integrity before testing.

Thirty samples of the expansive soils were collected from three different road projects along their alignments (10 samples per project spaced at 500 M apart - 5 KM section) in Nyandarua County: RWC 567 (Ndaragwa Constituency), RWC 589 (Ol Kalou Constituency), and RWC 596 (Kinangop Constituency). A stratified random sampling technique, guided by NTCP 103.1 standards, was employed. Stratification was based on subgrade depth, moisture exposure zones, and observable in-situ soil conditions. Sampling points were randomly selected within each stratum to minimize bias and ensure representative coverage. A minimum separation distance of 50 km was maintained between project sites to capture spatial variability in clay mineralogy, swelling potential, and moisture regimes. Samples were sealed in airtight polythene bags and labeled with location coordinates, depth, moisture condition, and project ID.

CKD samples were obtained from three major cement manufacturing plants in Athi River, Machakos County: Bamburi Cement, Simba Cement, and East African Portland Cement (EAPC). LKD samples were collected from lime production facilities in the same region. Three 50 kg samples of each stabilizer were collected from each plant. The plants were purposively selected because they utilize locally produced clinker, ensuring comparable chemical reactivity and pozzolanic potential. Samples were packed in sealed cement bags and transported under controlled conditions to prevent moisture contamination and maintain homogeneity.

Comprehensive characterization was performed on both the expansive clay soils and the stabilizers (CKD and LKD) to establish baseline engineering properties.

For Expansive Clay Soils:

Natural Moisture Content (BS 1377: Part 2).Atterberg Limits (Liquid Limit, Plastic Limit, and Plasticity Index) using cone penetrometer and rolling methods (BS 1377).Linear Shrinkage (BS 1377).Particle size distribution (BS 1377).Standard and Modified Proctor Compact Tests (AASHTO T99 and T180).California Bearing Ratio (CBR)- unsoaked and soaked (AASHTO T193).

Chemical composition (ASTM C114).Fineness, specific gravity, and moisture content (ASTM C311).All tests followed standardized procedures outlined in BS 1377 and AASHTO specifications. Samples were prepared in accordance with BS 1377 Part 1 (drying, pulverizing, quartering, and sieving) before testing.

Stabilization trials were conducted by combining predetermined percentages of CKD, LKD, and CKD-LKD blends with the expansive clay soils. Mixture preparation followed BS 1924 guidelines for cementitious and lime-based soil stabilization. The stabilizers were first blended with dry soil to ensure uniform distribution. Tested and approved water was then added to achieve the Optimum Moisture Content determined from compaction tests. The mixtures were thoroughly mixed, sealed in airtight containers, and allowed to cure before testing. Performance was assessed by comparing untreated and stabilized samples across varying stabilizer dosages.

Stabilized samples underwent the same geotechnical tests as natural soils, including Atterberg Limits, Linear shrinkage, Compaction characteristics, and California Bearing Ratio (CBR)—unsoaked and soaked. These tests quantified improvements in plasticity, volumetric stability, compaction behaviour, and load-bearing capacity resulting from kiln dust stabilization.

Laboratory data were analyzed using descriptive statistics to evaluate the effectiveness of kiln dusts and their blends in stabilizing expansive clay soils. Descriptive statistics were computed to quantify improvements in key geotechnical properties, including Atterberg limits (Liquid Limit, Plastic Limit, and Plasticity Index), Linear Shrinkage, compaction characteristics (Optimum Moisture Content and Maximum Dry Density), and strength properties (California Bearing Ratio—unsoaked and soaked). Comparative analysis was performed to evaluate the relative performance of CKD, LKD, and CKD-LKD blends at different dosages (i.e., 5%, 10%, 15%, and 20%). Trends were visualized using line graphs and bar charts to illustrate the relationship between stabilizer dosage and key engineering properties. This systematic approach ensured objective evaluation of stabilization effectiveness, identification of optimum stabilizer type and dosage, and reliable interpretation of results for practical application in road subgrade improvement.

This study evaluated the relationship between the increase of Cement Kiln Dust (CKD), Lime Kiln Dust (LKD), and 50 percent (50:50) CKD and LKD on the Atterberg limits of the expansive clay soil. Liquid limit (LL), Plastic limit (PL), Plasticity Index (PI), and Linear Shrinkage (LS) were evaluated because they are the primary indicators for determining plasticity of a soil and swelling potential.

Figure 1. Effect of stabilizers on Atterberg limits on expansive clay soils.

Results in Figure 1 show a consistent decrease in LL for all three materials (CKD, LKD, CKD-LKD blend) with increasing stabilizer content. The LL decreased for CKD-treated soil from 72.0 percent at 0 percent to 41.3% at 20%. The LL of soil treated with LKD was similarly decreased to 48.9% at 20%, while the blended CKD-LKD was decreased to 41.1% at the same stabilizer content. The reduction of the LL indicates that the ability of the soil to retain water has decreased. An engineer would consider using soils with lower liquid limits. The Plastic Limit (PL) shows consistent upward movement with increased stabilizer quantity. For example, PL increases from 28%, no stabilizer (untreated soil), to 34.5% (20% CKD), 42.6% (20% LKD), and 35.6% (50:50 CKD: LKD blended) stabilizers. Moreover, as the amount of stabilizer added increases, the Plasticity Index (PI) decreases significantly from a PI of 44.0% for untreated to 6.8%, 6.3%, and 5.5% for each of the three stabilizer additions (CKD, LKD, 50%: 50% mixture, respectively) for a combined reduction of PI of approximately 38.5% to 39.5%. The substantial reduction in PI indicates to civil engineers that the amount of plasticity and associated degree of swelling of expansive clay has been reduced significantly. This is important because high plasticity soil can undergo volume changes over wetting and drying; this can cause pavement failure due to cracking and/or upheaving.

For Linear Shrinkage (LS), a similar decreasing trend occurred with increased amounts of stabilizer added to the soil. The LS decreased from around 21% in untreated soil to approximately 3% at 20% stabilizer content across various types of stabilizers. This decrease in LS illustrates that there is a considerable reduction in shrink-swell potential, a major concern when working with expansive clays when constructing roads. In terms of engineering, the reduction in plasticity and shrinkage characteristics of the stabilized soil suggests improved construction workability due to reduced swelling potential and increased long-term stability as a subgrade material. Of the stabilizers tested, CKD-LKD showed slightly higher reductions in plasticity than either material individually, indicating an increased combined stabilizing effect. Overall, the testing showed that greater amounts of kiln dust help to improve the consistency limits and volumetric stability of expansive clays, thus improving their ability to be used as subgrade materials in roadways.

The study evaluated the compaction characteristics of the expansive clay soil by determining the Optimum Moisture Content (OMC) and Maximum Dry Density (MDD) at different percentages of kiln dust. These metrics are essential in pavement engineering since they determine the moisture conditions needed to achieve maximum soil density and stability during field compaction.

The compaction results in Figure 2 indicate that increasing the content of

Figure 2. Effect of stabilizers on OMC and MDD.

Cement Kiln Dust (CKD), Lime Kiln Dust (LKD), and the 50:50 CKD-LKD progressively improved the compaction characteristics of the expansive slay soil. For all stabilizers, the Optimum Moisture Content (OMC) decreased while the average maximum dry density (MDD) increased with increasing stabilizer content. The untreated soil recorded and OMC of 24.5% and an MDD of 1.63 g/cm³. At 20% stabilizer content, CKD reduced the OMC to 16.5% and the MDD increased to 1.77 g/cm³. Similarly, the CKD-LKD blend reduced the OMC to 17.0% and increased the MDD to 1.76 g/cm³ whereas LKD achieved an OMC of 19.9% and an MDD of 1.72 g/cm³.

The LKD-only treated soil also showed similar improvements, but the overall decreases in the OMC and increases in the MDD were not as significant as with CKD and/or blended stabilizers. The additional drying density (MDD) shows that adding kiln dust makes the treated soil denser than it was before. This change occurs because larger soil aggregates have formed, and free spaces in the soil have been filled with the small kiln ash particles. Among the stabilizers tested, CKD showed superior compaction characteristics and produced the highest MDD and the lowest OMC at high stabilizer concentrations. The CKD-LKD blend performed comparably, indicating the potential for using these 2-material combinations as a stabilizing agent. LKD improved compaction; however, the effects of LKD on compaction were not nearly as pronounced as with CKD. The improvements in compaction characteristics that occurred suggest a benefit of kiln dust stabilization to the structural integrity and stability of the expansive clay soil based on engineering principles. The result is a subgrade layer that provides a stronger and more durable foundation to support pavement loadings due to a higher dry density and lower moisture requirements.

The California Bearing Ratio (CBR) test was conducted to examine the improvement in the load-bearing ability of the expansive clay soil when stabilized with kiln dust. CBR is among the most essential metrics used in pavement design since it indicates the subgrade soil’s ability to support traffic loads.

The California Bearing Ratio (CBR) results in Figure 3 show that kiln dust stabilization notably improved the load-bearing capacity of the expansive clay soil especially under soaked conditions. The untreated soil exhibited unsoaked and soaked California Bearing Ratio (CBR) values of 4.8% and 2.7%, respectively, confirming its poor suitability as a pavement subgrade. Increasing stabilizer content improved CBR values even though the extent of improvement differed among the stabilizers. Lime Kiln Dust (LKD) produced the least improvement while unsoaked CBR remained almost the same (4.8% - 4.9%) and CBR increased progressively to 13.5% at 20% of stabilizer. On the contrary, CKD increased the unsoaked CBR to 7.5% and the soaked CBR to 19.8%, while the 50:50 CKD-LKD blend achieved the highest soaked CBR of 20.8% and an unsoaked CBR of 6.8% at the same stabilizer content. Among the stabilizers evaluated, CKD showed the

Figure 3. Effect of kiln dust on California bearing ratio (CBR).

greatest improvement in unsoaked CBR, whereas the CKD-LKD blend produced the highest soaked CBR. This finding indicates that CKD-LKD blend provides complementary stabilization benefits under saturated conditions. The increase in soaked CBR from 2.7% for the untreated soil to 20.8% for the blended stabilizer represents a very high improvement in load-bearing capacity, demonstrating the effectiveness of kiln dust stabilization in improving subgrade performance.

The notable increase in CBR may be due to cation exchange, flocculation-agglomeration, and pozzolanic reactions initiated by the calcium-rich kiln dusts. These cementitious products bind the soil and give it greater stiffness and resistance to penetration when loaded. Clay particle flocculation and agglomeration also reduce plasticity of the soil and enhance its internal structure. The increase in soaked CBR value is of major importance for engineers, as subgrade clay may be exposed to moisture throughout its lifetime. Thus, the increased soaked CBR value means that the stabilized clay will have sufficient strength to resist deformation and failure due to moisture. Ultimately, using kiln dust to stabilize expansive clay will allow engineers to convert weak expansive clay into acceptable subgrade materials for building roads, resulting in greater durability and performance of the road.

Stabilization of expansive soils with CKD, LKD, or CKD-LKD blends notably enhanced the soil's engineering properties for subgrade use. LKD was the most effective stabilizer, achieving the greatest reductions in plasticity index (from 44% to 6.3%) and swell potential (up to 97%), and the highest gains in unsoaked and soaked CBR (65% and 13.5%, respectively) and compaction density (MDD = 1.72 g/cm³ at 20% dosage). The CKD-LKD blend showed a synergistic effect, combining the rapid early reaction of LKD with long-term pozzolanic contribution of CKD, resulting in balanced improvements in strength, compaction, and dimensional stability. CKD alone produced moderate improvements in plasticity, CBR, and density, but it is still suitable for low-to-medium traffic subgrade layers. Kiln dust stabilization effectively changed weak, expansive soils into a denser, stronger, and dimensionally stable material suitable for road subgrade material, showing technical effectiveness and practical application in road construction.

The road authorities, such as KeNHA, KeRRA, and KURA, should encourage partnerships with cement manufacturers to secure locally produced CKD and LKD, ensuring consistent supply and quality of stabilizers. Lime Kiln Dust (LKD) should be the preferred stabilizer for expansive clay soils requiring rapid strength gain and high resistance to swelling, particularly in medium- to high-traffic road subgrades. CKD-LKD blends can be used where both cost efficiency and long-term pozzolanic stabilization are desired, providing a balance between early and later strength development. Cement Kiln Dust (CKD) is suitable for low-traffic road subgrades or for projects where industrial by-product utilization is a priority, though higher dosages may be required.

Contractors should apply strict moisture conditioning around the new OMC range and compaction at the achieved MDD benchmarks, supported by controlled CKD storage similar to cement handling, to avoid strength inconsistency due to premature hydration or moisture contamination before mixing. Kenya’s road materials manuals and highway standards should be updated to formally classify CKD as an approved soil stabilization additive, with testing routine validation standardized in all pavement construction projects.