Organic content

More than 250 laboratory organic content tests were undertaken and the results for peat, organic soil, and clay are tabulated in Table 25.6. The classification of the soft soils in terms of peat, organic soil, and clay has been based on field observations and confirmed by the laboratory testing.

While no specific boundaries were set for each type of material, the peat generally showed an organic content of < 20% with a recorded maximum of 87%, and the organic soils generally had organic contents of between 15% and 35%. Most of the clays had organic contents of > 15%. A distinctive feature of the CKE peat was its fibrous texture.

The organic contents for the peat, organic soil, and clay are presented later in Figs. 25.A.1–25.A.3 in the appendix.

Dry unit weight

The dry unit weights of the soft soils derived from the laboratory testing are shown in Table 25.7. Out of a total of 80 peat samples tested, the dry unit weight of 80% of the samples were below 0.8 t/m³ and only 2 samples had dry unit weights of greater than 1.5 t/m³. The mean dry unit weight for 80% of the samples below 0.8 t/m³ is 0.41 t/m³.

It appears that the dry unit weights obtained for the peat samples along the length of the CKE alignment are slightly higher than the values of 0.35 and 0.33 t/m³ for Peliyagoda and Muthurajawela, respectively, as reported by Ray et al. (1986). The dry unit weight distribution for peat, organic soil, and clay are presented later in Figs. 25.A.4–25.A.6 in the appendix.

Moisture content and Atterberg limits

The moisture contents (MC) for the CKE soft soils are presented in Table 25.8. MC for all undisturbed samples were extracted from samples saturated in the consolidation ring. As the groundwater level is very close to the surface and most of the soft soil samples were collected below the groundwater table, it is assumed that the MC obtained from the saturated samples would be representative of the in situ condition.

As significant moisture losses were observed during collection of some peat samples, it can be expected that the MC obtained from the disturbed samples would be lower than that obtained from the undisturbed samples. However, such a trend was not evident. The distribution of MC with depth obtained from testing of disturbed and undisturbed samples of peat, organic soil, and soft clay are presented later in Figs. 25.A.7–25.A.9 in the appendix. Atterberg limits for the soft soils are tabulated in Table 25.9.

The liquid limit for the majority of the peat and organic soil samples were above 50%, indicating that these materials are highly plastic. Out of 241 clay samples tested for the liquid limit, less than 10% of the values were below 35%, and 88% of the values were between 35% and 100%. Two peat samples, three organic soil samples, and six clay samples had liquid limits in excess of 100%, indicating the extent of compressibility of the soft soils.

The liquid limits, plastic limits, and natural moisture contents for the soft soils along the CKE alignment are presented in Fig. 25.3. Three organic soil samples had MC greater than 400% and one clay sample had an MC in excess of 200%. These results have been excluded in the figure for clarity of the presentation. It is also noted that MC and Atterberg limits tested from samples collected away from the main CKE alignment are not included in Fig. 25.3.

Figure 25.3 shows the variation of the natural moisture content along the proposed CKE alignment. High MC encountered in the areas of Peliyagoda (Ch 0 + 500–3 + 000), Welikadamulla (Ch 5 + 300–6 + 900), and Muthurajawela (particularly Ch 11 + 000–15 + 000) indicate the extent of CKE soft soil affinity to water. It has been reported that most of the peaty soils tested in Sri Lanka have MC between 50% and 500% (Tennekoon et al., 1993). The latest work for the CKE project includes about 300 samples tested for MC and shows that the natural moisture contents of the Colombo peats could be higher than previously encountered.

The undrained shear strength, S_u, of the soft soils encountered along the CKE alignment was assessed by reviewing the in situ vane shear tests, pocket penetrometer tests, and the unconsolidated undrained (UU) triaxial tests. The vane shear and UU test results for peat, organic soil, and clay are tabulated in Table 25.10 and presented in a graphical form later in Figs. 25.A.10–25.A.12 in the appendix.

One of the objectives of the 2003 investigation was to assess the strength gain of the material below the constructed embankments. Therefore, some of the undrained shear strength testing undertaken in 2003 was on material that had been subjected to embankment loading. It is evident that the undrained strength increases for peat and organic soils, as shown in Table 25.10. Figures 25.A.10 and 25.A.11 show that when the isolated upper values for the pre-2003 results are discounted, the strength gains due to consolidation are evident.

The undrained strengths from the 2003 investigation for soft clay in Table 25.10 appear to be less than what was obtained prior to construction. However, a more detailed review of the results for 2003 indicates that only five tests were undertaken in clay and most of them were undertaken in the Peliyagoda area where the soft soil materials extend to depths of 10–12 m and are very soft. The limited undrained clay strength results available for this area indicate that they are comparable with the results obtained in 2003. This is in line with the fact that the embankments in this area are not constructed or still at ground level.

The results in Table 25.10 also indicate that the in situ vane shear strengths are generally higher than the laboratory UU triaxial test results.

Drained strength

The drained strength parameters (i.e., c′ and ϕ′) interpreted from the consolidated undrained (CU) triaxial tests are tabulated in Table 25.11. For the 2003 investigation, difficulties in sample preparation for the soft soils were encountered and as such only one CU triaxial test was undertaken. This result also appeared to be erroneous and has therefore been excluded from the analysis.

Only a few CU triaxial tests were undertaken in the investigations prior to 2003. The results indicate that the peats exhibit a small effective cohesion generally in the range of 5–8 kPa and the organic soils and clay show an effective cohesion in the range of 8–12 kPa. While the effective friction angle for four of the peat samples are between 22° and 25°, the effective friction angle for three of the organic soil samples lie between 25° and 27°. The effective friction angles for the clay samples are generally between 23° and 25°, with two values greater than or equal to 29°.

It is noted that there were difficulties in recovering undisturbed samples of very soft peat and organic soil. Therefore, the results shown in Table 25.11 may not be representative of the soft soils.

The void ratios extracted from the consolidation testing for the CKE soft soils are presented in Table 25.12. The void ratio of peat was generally between 1 and 8. This shows that for most of the peat samples the volume of voids within the soil mass was between 50% and 90%, indicating extremely high compressibility of the peats.

Ray et al. (1986) states void ratios of 5.08 and 4.16 for peat in Peliyagoda and Muthurajawela, respectively. The results from this study show that the average void ratios for the Peliyagoda and Muthurajawela areas are 4.80 and 4.36, respectively, and compare well with the previous findings. The results also show that the void ratio, and thus the compressibility, is less for the organic soils than for the peats, while the soft clays are the least compressible.

The testing in 2003 focused more on the areas with deep soft soils, particularly at Peliyagoda, Welikadamulla, and Muthurajawela. The pre-2003 investigations covered the entire alignment, which also includes less compressible areas. As a result, the void ratios appear to be higher from the 2003 investigation.

Compression and recompression indices

The compression and recompression indices extracted from the consolidation testing for the soft soils are presented in Table 25.13. The compression indices as shown in the table indicate that the compressibility of the soft soils decreases from peat to organic soil to soft clay. The mean recompression index is between 7% and 12% of the compression index.

Overconsolidation ratio

The overconsolidation ratios (OCRs) for the soft soils as interpreted from the oedometer tests are presented in Table 25.14. The preconsolidation pressure was determined from the oedometer tests.

The OCR for peat was generally between 0.4 and 7. The OCRs > 7 were from peat samples collected from less than 0.5 m depth. These high pre-2003 values are most likely due to effects of desiccation or variation of the water table. When these high OCR values are removed, the mean OCR for the peat is 1.5, indicating that the peat is generally normally consolidated or slightly overconsolidated.

The OCR for organic soil was generally between 0.5 and 7. The OCRs > 7 were from samples at shallow depths of < 0.3 m. If these values are ignored, the mean OCR for organic soil is 2.1. The OCR for clay was generally between 0.5 and 10. The OCRs > 10 were from samples at shallow depths of < 0.6 m. If these values are removed, the mean OCR for the soft clays is 3.5.

Coefficient of volume compressibility

The coefficient of volume compressibility, m_v, extracted from the consolidation testing and the back-analysis findings are presented in Table 25.15. In determining which laboratory m_v value to use in the statistical analysis, the proposed embankment heights were used to estimate the final loading at the test sample location and the m_v value corresponding to the expected final stress was chosen for the analysis. The back-analysis of the monitoring data took into consideration the reduction of m_v as the height of the embankment increased.

The laboratory test results and the back-analyzed m_v values were plotted against vertical stress, as shown in Fig. 25.4 for peat, organic soil, and soft clay. The back-analyzed m_v values for peat were in the range of 0.5–7.6 m²/MN and were generally higher than that for organic soil (0.6–5 m²/MN) and clay (0.2–2.5 m²/MN), indicating greater compressibility for peat than for organic soil and clay.

The lower and the upper bounds of the m_v range assumed in the back-analyses for peat, organic soil, and clay are shown in Fig. 25.5.

Coefficient of secondary consolidation

The modified secondary compression indices, C_ɛα = C_α/(1 + e₀), for the soft soils have been calculated using the C_α values derived from the laboratory testing and are tabulated in Table 25.16.

The data from the pre-2003 investigations appear to be questionable and are excluded from the analysis. Figure 25.6 shows the distribution of the back-analyzed C_α/(1 + e₀) for peat, organic soil, and clay along the alignment. The back-analyzed C_α/(1 + e₀) values were generally in the range of 2–8% for peat, 1–4% for organic soil, and 1–2% for soft clay.

Figure 25.7 shows the relationship between the modified secondary compression index (from back-analysis) and the natural water content for the peat, organic soil, and clay of the CKE project together with other published data. When assessing the range of values for the organic soils, one MC result of 10% was excluded from the results due to its isolated extremity. The C_α/(1 + e₀) for the CKE soft clays appear to be higher than that for other reported clays.

Laboratory oedometer tests with extended loading stages were undertaken in 2003 to assess creep settlements in normally and overconsolidated soft soils. The test procedure generally consisted of stepped load increments and calculation of the corresponding void ratios. At a loading stage approximately equal to the proposed embankment load, the loading was held for an extended period (generally ~ 7 days) for the sample to consolidate well beyond the completion of the primary consolidation stage, and thus well into the secondary compression stage.

Thereafter, the loading was increased again as per the normal testing procedure. During the unloading stage, the load was again held for an extended period at the approximate proposed embankment load. The creep settlement for the overconsolidated soils were found to be negligible compared with the normally consolidated soils. The summary of the test results for normally consolidated soils under equivalent embankment loads is presented in Table 25.17.

The coefficients of consolidation, c_v, for the soft soils from laboratory testing and back-analysis are shown in Table 25.18. In determining which laboratory c_v value to use in the statistical analysis, the proposed embankment heights were used to estimate the final loading at the test sample location and the c_v value corresponding to the expected final stress was chosen for the analysis. These laboratory test results and the back-analyzed c_v values were plotted against vertical stress as shown in Fig. 25.8.

The back-analysis revealed that the c_v values reduced as construction of the embankment progressed. This is considered reasonable, and is consistent with research findings (Mesri and Rokhsar, 1974). In general, the c_v values for peat were found to be between 2 and 20 m²/year and 1–10 m²/year for organic soil and 0.5–5 m²/year for clay. This suggests that the peat is more permeable than the organic soil and clay.

Figure 25.8 indicates that the back-analyzed c_v values for the peat, organic soil, and soft clay are notably higher than that obtained from the laboratory testing. The lower and the upper bounds of the c_v range assumed in the back-analyses for peat, organic soil, and clay are shown in Fig. 25.9.

Permeability

Table 25.19 presents the results of the rising head field permeability tests undertaken in the 2003 investigation. The field permeability test results indicate that the coefficient of permeability of peat is in the order of 10^− 6–10^− 7 m/s. Tennekoon et al. (1993) quote coefficient of permeability values ranging from 1 × 10^− 5 m/s for fibrous peats to 1 × 10^− 7 m/s for amorphous peats.