Deformation Characteristics of Coastal Plain Sands at a Water Treatment Plant Site, Calabar, South-Eastern Nigeria.
E. O. Esu
Department of Geology, University of Calabar, Nigeria
Civil Engineering Department, Federal Airports
The 25.4 mm permissible settlement value for sandy soils was in dispute at a water treatment site works in Calabar, South-eastern Nigeria. The terrain is underlain by Tertiary-Recent sediments of the Benin Formation of the Niger Delta. Preliminary soil tests had shown a predominantly sandy soil with up to 40% fines (materials passing sieve no 200). Consolidation tests carried out on undisturbed soil samples obtained from the site for the water filter unit shows a preloaded soil with consolidation indices (mv, Cc, k) having values comparable to reported values for some clays; for example the Upper London Blue clay and the Cincinnati clay. The implication of this to the assumed permissible value of 25.4mm for the proposed structure is the subject of this article.
Keywords: Consolidation, sands, fines, Compresibility, Permeability, Settlement, Calabar, Nigeria.
Consolidation tests are often carried out as part of site investigation to predict the settlement behavior of intending structure on the soil on which it would be founded. It is common practice in Foundation Engineering design to limit permissible settlement for isolated footing to 25.4mm and 40-65 mm for raft foundation on sandy soil (Skempton and MacDonald, 1956). This value is closely related to the standard penetration ‘N’ value of the soil. (Peck et al, 1974 Lambe and Whiteman 1978). Ambiguity exists as to the nature (normally or over consolidated) and type of sandy soil such principle is applicable to; especially as natural sand deposits usually always contain silt or clay. (Consolidation characteristics of soil are influence apart from the magnitude of the applied stress by particle sizes, shapes, and mineral skeleton arrangements). The influence on consolidation characteristics of the presence of a reasonable percentage of fines up to 40% within a sandy matrix is not sufficiently dealt with in available works. Further more, clayey soil understandably receives more attention in literature than its sandy counterpart. This paper using consolidation tests carried out at a water treatment plant site seeks to examine such a situation.
Background and Site Description
By an International Bank loan to the Cross River State Government in South-eastern Nigeria, an Urban Water Scheme involving construction of New water treatment units, storage tanks and new pipe networks is presently being carried out at Calabar, the capital of Cross River State by a consortium of three international contractors, handling the different aspects of the job. The water treatment plant site is located in Ediba, Calabar. Calabar has geographical coordinates given by Latitude 04?55' and 05?05' N and Longitude 08-15' and 08-25'E.
Geologically, Calabar is underlain by the Benin Formation; one of the Formations of the Tertiary - Recent sediments of the Niger Delta (Short and Stauble, 1967) and is locally referred to as Coastal Plain Sands.
Table 1. Sieve Analysis and Atterberg Limits of soil samples from filter locations
The type section of the Coastal Plain Sands according to Allen (1967), is made up of fine grained sands, pebbles moderately sorted with local lenses of fine grained poorly cemented sands and gravels with clay and shale intercalations. The sands are sub angular to well rounded.
The site for the water treatment units (consisting of aerators, sedimentation tank, filter and clarifiers) is located between 67.00 m-69.0 m above mean sea level. Preliminary soil investigation at the site proposed for siting the water treatment units indicate a soil with standard penetration test ‘N’ values in the loose to medium dense consistency up to 7.0 m, with ‘N’=20 typical from about 4.0 m to 7.20 m; and dense to very dense (N>30) from around 8.0 m-13.0 m (Figure 1).
Figure 1. Typical standard penetration test signature for the water treatment plant site
Drill record showed soil from ground level to 7.20 m to be brown clayey sand and between 7.20 m to 13.60 m to be clayey sand interspersed with gravel. This latter layer has wide variations in horizon and thickness around the site from the drill records available. In some cases the first layer, which is clayey sand, goes up to 12.0 m deep. Sieve analyses showed the presence of significant amount of fines (soil passing sieve no 200) up to 45% (Table 1) in some cases within a predominantly sandy soil. The presence of this appreciable amount of fines led to objection by the contractor as to the applicability of the 25.4mm permissible settlement assumed in designing the water treatment unit structures, which are to be of reinforced concrete. This value was assumed not only to satisfy the limit states requirements (serviceability and ultimate) but also a settlement value that the connecting pipes between the treatment units can tolerate without buckling or being damaged. Consequently, undisturbed soil samples were collected from the proposed locations for both the filter and clarifier plants for consolidation tests. The floating ring consolidometer was employed for the tests. This paper deals with the analysis and interpretation of data collected for the filter plant site.
Table 2. Consolidation Parameters for the 4.0 m Depth Saturated Sample
The filter has an overall dimension of 71.90 m by 17.90 m, and is 4.0 m in height, with a base and vertical wall thickness of 0.40 m. This structure consists of eight cells imposing a surcharge load of approximately 55kPa. An `N' value of about `10' is required for this surcharge load from a chart (Figure 2) by Terzaghi and Peck (1967), that will limit the permissible settlement to 25.4mm, whereas ‘N’=20 is recorded for this site at the bottom of the filter which is designed to be 4.0 m below the existing ground level. The existing ground level is at an average elevation of 68.50 m above mean sea level (amsl). The consultant for the project places their confidence on this preliminary information until the contractor raises the objection due to the presence of significant amount of fines in the soil on the site.
Figure 2. Settlements of footing from standard penetration resistance, N (Terzaghi and peck 1948)
This study has the objectives of:
Comparing values of consolidation parameters, coefficient of consolidation Cv., permeability k, and coefficient of volume compressibility mv, and compression index Cv, obtained from consolidation test with those obtained in previous works especially for clay soils.
Determining whether the 25.4mm permissible settlement value for a sandy soil is valid in the case of the soil under investigation; and
Predicting the likely total settlement for the proposed filter structure from consolidation test carried out on soil samples obtained from the proposed site for the filter.
Field Procedure and Data Acquisition
Undisturbed samples were obtained at the filter location at 4.0 m and 6.0 m depths. The base of the filter is to be at about 4.0 m below the existing ground level. This guided the depth at which soil samples were obtained. Due to non-availability of mechanical soil boring equipment at the site for collecting undisturbed soil sample, the consultant ordered the soil sample to be taken from excavated pits.
Table 3. Consolidation parameters for the 4.0 m depth unsaturated sample
The samples considered undisturbed were taken from 4.0 m and 6.0 m depths using different pits for each depth. Care was taken in locating the pit. The pit for the 6.0 m depths was located on the alignment of pipe coming into the filter from the clarifier units at some distance from the filter. The other was located at some distance from the short edge of the proposed filter position. A bucket excavator was employed to remove the overburden until 3.0 m and 5.0 m depths. The exposed surfaces at these depths were hand shoveled down to 4.0 m and 6.0 m. The four sides were carefully chiseled down to 5.0 m and 7.0 m. The base of the exposed sample blocks was carefully chiseled to produce a cube of 0.40 m sides. The resulting sample blocks were immediately gauze-waxed right in the pit to conserve their natural moisture. Laboratory samples were obtained from each of the sample blocks. Consolidation tests were carried out on both saturated and unsaturated samples based on procedure described in ASTM D 2345 - 70 using the loading sequence 10, 30, 50, 100, 200, 400, 800, 1600, and 3200kPa, and the floating ring type consolidometer. Other tests such as natural moisture content, consistency limits, specific gravity, and bulk density were determined in accordance with relevant ASTM standards.
Table 4. Consolidation parameters for the 6.0 m depth saturated sample
Sieve analysis and Atterberg limits laboratory tests were carried out on some samples. Sieve analysis was based on ASTM D422-63 standards while Atterberg limits tests (liquid and plastic) were based on ASTM D423-66 and D424-57 standards respectively. Table 1 locates the soil in the sandy region, but with appreciable amount of fines between 38.1% and 45.5%. Atterberg Limits on minus 40 (US sieve size) for the soils are also presented in Table 1 and a plot on the plasticity chart places the soil dominantly in the Inorganic clays and silts with medium plasticity. The soil is thus classified as SC and SM.
Table 5. Consolidation parameters for the 6.0 m depth unsaturated sample
Results and Analysis
For each load in a loading cycle, relevant graphs were plotted and computations of consolidation parameters made using a software package agreed upon by the contractor, the consultant and the author. For the sake of completeness, the formulas used for computation on "cohesive" soils (Lambe and Whitman,1979) are given in the following.
Coefficient of consolidation,
Coefficient of compressibility,
Coefficient of volume compressibility,
H = sample height
t90 = time for which 90% of the primary consolidation takes place
e2, e1, = final and initial void ratios respectively.
P2, P1, = final and initial pressure on soil sample
gw, = unit weight of water.
The software uses Taylor’s method to estimate coefficient of consolidation Cv, since it is known to give a conservative value (Lambe and Whiteman, 1979). Tables 2, 3, 4, 5 display values of computed consolidation parameters. Casagrande’s method was used to estimate preconsolidation pressure. Figures 3 and 4 show typical graphs and construction procedures. The values obtained for the overburden and preconsolidation pressures for the samples at 4.0 m and 6.0 m depths are presented in Table 6. It is necessary to determine whether the soil is preconsolidated or not because computations of settlement based solely on the overburden pressure might be misleading if the consolidation status of the soil is overlooked (Peck et al, 1974).
Figure 3. Estimation of preconsolidation pressure for the 4.0 m depth saturated soil sample using Casagradre method
Figure 4. Estimation of preconsolidation pressure for the 6.0 m depth saturated soil sample using Casagradre method
The coefficient of consolidation (cv), coefficient of volume compressibility (mv), permeability (k), initial void ratio (e), surcharge load, are parameters required for estimating likely settlement of structure placed on a soil. Since a linear relationship exist between cv, mv and k, a linear correlation of these data was carried out (Table 6). The correlation process allows:
Table 6. Overburden, preconsolidation pressure and correlation between some consolidation indices
Table 7. Representative dial gauge readings and settlements for the unsaturated soil samples
Table 8. Classification of soils based on coefficient of permeability1
The quality of data obtained from consolidation test for a sample to be examined, and the degree of dependence of one of the parameters on another to be evaluated.
Part of the dial gauge reading for some of the consolidation test load sequence are presented in Table 7 to show that much settlement takes place within the first 15 seconds after change of load, (a complete load cycle is 24 hours). This initial settlement can be regarded as immediate settlement and is estimated to be about 80% of the total settlement. The remaining settlement is due to primary consolidation, which is time dependent and is also primarily influenced by the compressibility characteristics of the soil. Since from Tables 2 - 5 the compressibility mv, of the unsaturated samples is high than for the saturated samples, estimation of immediate settlement at the corner of the filter structure is made for 6.0 m depth for the unsaturated sample assuming a rigid base slab, using the elastic equation (Tomlinson, 1980);
ri = immediate settlement
B = width of foundation
Ed = 'deformation' modulus
µ = poisson’s ratio
qn = foundation pressure
Ip = influence factor (from chart by Terzaghi, 1943, Figure 5).
Using µ = 0.28, Ed = 2702.7kPa, qn = 55kPa, Ip = 0.98.
From which ri = 32.89 cm
Figure 5. Influence coefficients for settlement of rectangular loaded area (From Terzaghi 1943)
Settlements due to primary consolidation were calculated for the 4.0 m and 6.0 m depths for the unsaturated samples assuming a compressible layer of 6.0 m and 8.0 m (values of settlement and time period in brackets in Table 7). The time period for 90% consolidation was also calculated assuming a double drainage path and using the one-dimensional consolidation theory equation.
S = Settlement
Cc = Compression index
e0 = Initial void ratio
DP = Difference between existing load on soil and imposed surcharge
P1 = total load on soil (including preconsolidation load)
H = thickness of compressible stratum (m)
t90 = Time period for 90% consolidation
T = Time factor (0.85 here).
The soil under study is not too expansive as indicated by low plasticity in index (PI) between 20% and 27%, which is less than 35% classified as moderately expansive, by Holtz and Kovacs (1981).
For both saturated and unsaturated samples at both depths compressibility mv ranges from 0.11 m2 MN-1 to 0.77m2 MN-1 for the load range 800kPa to 30kPa. Although coefficient of compressibility is higher for the unsaturated samples, from Table 6 coefficient of correlation ‘r’, between Cv and mv is -0.1145 and 0.8528 for the saturated sample, which are less than 0.6595 and 0.9070 for the unsaturated sample. It follows that the settlement is more reliably estimated in the unsaturated state than the saturated state for these types of soils. Comparing the mv values obtained for the load range 30-800kPa for all the soil samples with values obtained for various clays reported by Tomlinson (1980), Mckinlay (1982), the values fall within what is classified as clays of medium compressibility (that is values of mv between 0.10 m2 MN-1 to 0.30 m2 MN-1). Examples cited as clays with medium compressibility include Upper Blue London clay and weathered oxford clay (Tomlinson, 1980).
From Tables 2-5 the Oedometer permeability k obtained for both types of soil samples at the two different depth can be classified as being very low (10-5-10-7cm/sec) using Terzaghi and Peck, (1967) soil classification based on permeability (Table 8). As indicated by Table 6 there is good degree of correlation between k, and Cv with ‘r’ having a least value of 0.3601 and a maximum value of 0.9239, thus affirming that Cv is quite uniquely related to the permeability of a loaded soil. An independent evaluation of k through experimental method other than using the relation k = Cv mv ?w will therefore help in estimating reliable value for Cv.
Compression index Cc, with values from Tables 2-5 fall between 0.100 and 0.308 for both saturated and unsaturated samples. They also become significant at surcharge load above 200 kPa and are within the range of values reported by Lambe and Whitman, (1979) for clay soils having comparable plastic and liquid limit values to the soil under study; for example Montana Clay has a Cc of 0.21, plastic limit of 28% and the liquid limit of 58%. Cincinnati clay with Cc of 0.17 has a plastic limit of 12% and liquid limit of 30%; while Boston Blue clay (remolded) with Cc of 0.21, has a plastic limit of 20% and liquid limit of 41%.
The drainage pattern within the substratum of the foundation is very crucial to estimating settlement for a given structure on soil as it determines the value of ‘H’ (drainage path) in consolidation theory. It is necessary therefore to determine the drainage pattern with reasonable degree of certainty.
Implications of the Results
From preliminary computations of immediate settlement and the settlements due to primary consolidation (Table 9), the total settlement of the proposed filter structure is clearly not within acceptable limits required to meet ultimate limit state requirement and pipe work connection. Proposed options include;
Table 9. Settlement computations for 4.0 m and 6.0 m unsaturated samples, and time for 90% degree of consolidation
The preloading of the soil with earthworks
The installation of large diameter piles to the clayey sand layer containing gravel.
Placement of grouted small diameter bored piles designed as end bearing, to the gravelly layer with settlements from primary consolidation estimated at a maximum of 6.0cm; and the structures allowed to stand for about six months before pipes are connected between various units.
Time constrains and delays do not encourage the first option. However, this controversy could have been easily resolved using electrical resistivity geophysical method. There was controversy over the different depths, thickness and continuity of the gravelly layer as indicated by three drill records available. Due to this controversy option II was not adopted, as there were fears that large load from big diameter piles could lead to punching shear in this layer. The third option was eventually adopted.
It is the practice within the study area to design the foundation of structures sited within the Coastal Plain Sands assuming a sandy soil. The study illustrates a case where the permissible settlement design criteria for foundations on such sandy soil do not apply even though the soil being dealt with is predominantly sandy. Detail soil consolidation characteristics need to be determined for soil within the study area.
This is also a case study of a situation where the presence of significant amount of soil fines like clay or silt up to 40% within a predominantly sandy matrix affects significantly the consolidation characteristics of soil as indicated by the values of mv, k and Cv obtained for the soil investigated. These values are well within the range of values reported in previous works for clays.
We are grateful to the contractor Messrs IMPRESIT BAKOLORI and the Consultants SGI Consulting Nig. Limited for the opportunity given us to participate in this aspect of the project as independent witness to the consolidation tests carried out to resolve the dispute.
The following symbols are used in this paper.
av = Coefficient of compressibility;
B = Width of foundation;
Cc = Compression index;
cv = Coefficient of Consolidation;
Ed = Oedometric Modulus;
e1 = Initial void ratio of soil sample;
e2 = Final void ratio of soil sample;
H = Sample height or thickness of compressible stratum;
DH = Height change in soil sample;
Ip = Influence factor;
k = Coefficient of Permeability;
mv = Coefficient of volume compressibility;
Pc = Preconsolidation Pressure;
DP = Difference between existing load and imposed surcharge load;
P1 = Total load on soil;
P1 = Initial pressure on soil sample;
P2 = Final pressure on soil sample;
qn = Foundation pressure;
S = Settlement of soil due to consolidation;
T = Degree of consolidation time factor;
T90 = Time Factor for 90% consolidation;
t90 = Time period for 90% consolidation;
e = Soil sample strain;
ri = Immediate settlement (Elastic);
gw = Unit weight of water
µ = poisson ratio
|© 2003 ejge|