Department of Civil Engineering, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia

Abstract

The behavior and properties of sediments depend on their compositional characteristics and formation processes, as well as the environmental conditions during their geological history, i.e. post-formation processes. A vertical cut made in a hill in Dhahran, Saudi Arabia, reveals a vivid picture of the inherent heterogeneity of sediments that have been deposited at different geological ages. A review of the geology of the area, as well as laboratory tests, help to determine the possible causes of the variability of soil types and properties in the area. Laboratory tests include basic geotechnical tests, chemical tests, X-ray diffraction analysis, scanning electron microscopy, and thermal analysis. These tests are used to identify different rock types and soils from the face of the cut. The results of this study indicate that the material from this cut varies from clayey shale and limestone rock (Tertiary, lower Eocene) formed some 52 M.Y. to calcite-cemented sand and pure calcite rock formed in the Quaternary age.

Author Keywords: Construction materials; Heterogeneous materials; Mineral composition; Naturally cemented sand; Sedimentary rocks; Site characterization

Article Outline

1. Introduction

2. Geology of the area

3. Experimental

4. Results and discussion

4.1. Clayey shale

4.2. Cemented sand

4.3. Rocks

5. Summary and conclusions

Acknowledgements

References

1. Introduction

The behavior of geomaterials (soil and rock) depends on the properties of their constituents. Soil behavior includes different aspects, such as volume change, stability, erodability, and stress–strain characteristics. Various properties of soil depend on various compositional and environmental factors that determine their mineralogy, fabric, and structure (Mitchell, 1993). These factors depend on soil origin, formation or deposition processes, post-depositional changes, geological history, environmental conditions, and weathering processes.

Local soils are being used in Dhahran, Saudi Arabia in the construction of various types of pavement systems, as well as to support the foundations of different structures. These soils need to be characterized, and their engineering properties need to be explored in detail. The soil properties depend on their mineralogical constituents as well as their geological mode of formation. Various types of local soils exhibit widely different characteristics, even within the same site, in both vertical and horizontal directions. Soil exploration is very important in order to assess the soil properties, as well as to suggest ways and means to improve and stabilize poorly performing soils, for different construction purposes.

This case study focuses on a vertical cut made in one of the hills in the Dhahran area, in the Eastern Province of Saudi Arabia (Fig. 1). This cut was made in the summer of 1994, for the Al-Rabwah real estate development, Dhahran. It is located at 26° 21′ lat. N, 50° 08′ long. E, at an altitude of about 80 m above the present sea level and about 10 km from the shore of the Arabian Gulf. The cut lies in the N45E–S45W direction. The paper provides a brief description of the geology of the area and gives the results of an extensive experimental program used to identify and characterize different types of soils and rocks taken from the face of this cut.

2. Geology of the area

The Dhahran area is located in the Eastern Province of Saudi Arabia, which is part of the Arabian Shelf. The overall climate of the area is described as ‘arid’ (Al-Sayari and Zötl, 1978). The area is about 10 km from the western shore of the Arabian Gulf and rises progressively from sea level to a maximum altitude of about 150 m. The main geomorphologic feature is the Dammam dome, which rises to over 150 m from the edge of the Gulf; it is an anticlinal structure with a core excavated by erosion. Various formations outcropping in the area form concentric ellipses with their longer axis approximately along the N30W–S30E direction, Fig. 1.

The Dhahran area is formed geologically from a relatively recent sediment. It is part of the Arabian shelf, which was submerged beneath sea water during the transgression and regression cycles during the geological history of the Arabian Gulf, including the late Pleistocene–Quaternary age (about 0.4 M.Y. =400 000 years before the present) when the Arabian Gulf water stood at a level 150 m higher than its present mean (Al and Holm). The Arabian Gulf water has more salt content, with the major chemical constituents in (mg/ml) as follows: Na⁺ (20.7), Mg²⁺ (2.30), K⁺ (0.73), Ca²⁺ (0.76), Sr²⁺ (0.013), Cl⁻ (3.69), Br⁻ (0.121), SO²⁻₄ (5.12), and HCO⁻₃ (0.128) (Al-Amoudi et al., 1992).

The rocks exposed within the Dhahran area are very gently dipping Tertiary rocks, with a covering of Quaternary deposits in some locations. The oldest formation, the Paleocene–lower Eocene Umm er Rhaduma Formation, outcrops in the anticlinal Dammam Dome. The lithologies and stratigraphies for the various geological formations of the Dhahran area, from the base upward (Powers and Roger) are as follows:

Umm er Radhuma Formation (Tru): This is partially dolomitized aphanitic limestone, commonly fossiliferous and chalky, with thin calcarenitic intercalations and dolomite in its upper part. It outcrops at the core of a small anticline sitting on the Dammam dome (Tleel, 1973). This formation dates back to 67–54 M.Y.

Rus Formation (Tru): This outcrops widely in the core of the Dammam dome and is divided into three lithologic units:

1. gray to cream, compact, partially dolomitized limestone typically with quartz geodes at the top, about 21 m thick;

2. marl and limestone, with irregular masses of gypsum and quartz geodes at several levels, approximately 32 m thick;

3. white, soft, chalky limestone, about 4 m thick.

The top of the formation contains three intercalations of gypsiferous shale with chert nodules at the top. The Rus formation is of a lower Eocene age, which dates back to 54–52 M.Y.

Dammam Formation (Tdm): This is a lithologic and stratigraphic column of the Dammam formation that is depicted in Fig. 2, and is divided into five members:

1. the Midra Shale Member(sh), 3–8 m thick, composed of brown clayey shale, gray marl, and soft, impure limestone;

2. the Saila Shale Member(sh), 4.2–12 m thick, with gray, impure fossiliferous, calcarenitic limestone (0.6 m), overlain by brownish yellow clayey shale (3.6 m);

3. the Alveolina Limestone Member(ls), 0.9–12 m thick, composed of pale-orange to yellowish gray, rarely silicified, dolomitized limestone which is monocrystalline and partially recrystallized;

4. the Khobar Member(lsdl) with a lower marly unit (1.5 m thick) and an upper limestone unit (7.8 m thick); and

5. the Alat Member with an upper unit (9 m thick) comprising pale, chalky, porous dolomitic limestone with numerous mollusc impressions and silicified at the top (dlls); and a lower unit (6–18 m thick) of pale, fine-grained dolomitic marl (dl). The Dammam formation is of lower-middle Eocene age, which dates back to 52–36 M.Y.

Hadrukh Formation (Th): This formation is 90 m thick, composed mainly of marly sandstone, sandy marl, sandy clay, and sandy limestone; chert nodules occur in several layers, and small quantities of gypsum are also present; lenses of gravel occur locally. This formation is a Miocene–Oligocene formation that dates back to 36–20 M.Y.

Dam Formation (Td): This formation is 90 m thick and composed of pink, white, and gray marl, of red, green, and olive clay with intercalations of sandstone, of chalky limestone, and of shelly limestone. This formation is a Miocene formation dating back to about 20–2 M.Y.

Due to various erosion agents, most of the surfacial layers have been removed. The left-over part forms hills of different sizes and shapes. From the face of the cut made in one of these hills, a vivid picture of the variability of soil/rock materials in both horizontal and vertical directions can clearly be seen, i.e. the heterogeneity and the anisotropy (Fig. 3). These materials consist of clay, sand, and rock layers or pockets (Fig. 4). This figure shows both the sh-Tdm member (left, east) and the un-Tdm member (right, west), separated by clastic sediments. The lithologic symbol ‘sh’ represents ‘shale’, and ‘un’ represents ‘undifferentiated’ dlls-lsdl; where ‘dlls’ represents dolomite/dolomatic limestone, and ‘lsdl’ represents low magnesian limestone/dolomite. The strategraphic age ‘Tdm’ represents the Dammam Formation. The cut shows the opposite dip directions of the two members.

This study focuses on the left (east) side of the cut, the sh-Tdm member of the Dammam formation. This part contains layers of shale and limestone, as well as large lenses of white crystallized ‘calcite’, in addition to cemented sand. The structural geology of the shale and limestone layers gives a strike of about N 20°W and a dip of about 30°N. Various clay layers are segregated, and the clay material within each layer is highly fissured [Fig. 5(a)]. The sand is naturally cemented and can withstand a vertical cut of about 10 m. The rocks are of different types and colors. The brown rocks are cracked, while the white rocks contain solution cavities of various shapes and sizes [Fig. 5(b)]. This cut has been observed for the last 5 years, and no instability problem of any kind has occurred.

3. Experimental

In addition to site visits paid to this cut for visual inspections of various soils and rocks, many samples were collected from different materials to be used in the experimental investigation program. This program consisted of various laboratory tests to identify the composition of the different materials.

Tests were made on clayey shale samples to determine consistency limits, especially the liquid and plastic limits. The specific gravity was also found for some samples. Tests were conducted according to ASTM standard procedures. Also, the X-ray diffraction (XRD) technique was used to characterize minerals present in this clay. Furthermore, a scanning electron microscope (SEM) and Element Density Statistics (EDS) were used to determine the chemical elements present.

Various tests were made on samples from the naturally cemented sand. The particle size distribution of the crushed sand samples was obtained using sieve and hydrometer analyses. Because this sand is highly cemented, both dry and wet sieving were performed. A hydrometer test was performed on the material passing the #200 sieve (0.075 mm) from the wet sieving. Particle shape and surface texture for sand retained on the #200 sieve were investigated under a regular microscope. Consistency limits and specific gravity were also obtained for this material. Additionally, direct shear tests and triaxial tests were conducted on dry crushed sand samples. All tests were conducted according to the respective ASTM standard procedures. Further, XRD, SEM, and EDS studies were made on a sample of naturally cemented sand, as well as on the fine material passing the #200 sieve, in order to determine the mineralogy of the sand and the type of cementing material.

The rock type was initially identified by chemical methods. More sophisticated methods were used to identify the semi-transparent, crystallized ‘white’ rock using the XRD technique, SEM and EDS analyses, and simultaneous thermal analysis techniques of differential thermal analysis (DTA) and thermogravimentry (TG).

The XRD analysis was aimed at identifying the chemical/mineralogical phases (compounds) present in the sample and their relative amounts. The experimental procedure begins by grinding samples into a fine powder using an agate mortar and pestle. In this powdery form, many particles come into orientation and greatly improve the quality of the diffraction pattern. The diffraction pattern was generated by a θ–2θ scanning diffractometer. The operating conditions of the XRD analysis were a Cu fine focus tube at 40 kV and 30 mA, a divergence slit of 1°, scatter slit=1°, receiving slit=0.2 mm, a nickel filter, scanning speed and interval of data collection of 0.01° 2θ/s, and angle scanned (2θ) of 4–80°.

SEM and EDS analyses were aimed at obtaining a semi-quantitative analysis of the elements present in the different soil samples. The samples were mounted and sputter-coated with a thin gold film to improve their electrical conductivity. The JEOL JSM 840 scanning electron microscope was used for the microscopy and EDS analyses that were conducted under a test chamber pressure of about 0.5 mPa, an electron beam current of 3 nA, and a voltage of 19 kV. The Standardless Semi-Quantitative (SSQ) program was utilized to process the spectra of the samples and to obtain a weight percentage of the elements detected. Only elements from atomic number 11 (sodium) and higher can be detected by this instrument. Both area and spot analyses were carried out on the samples.

4. Results and discussion

4.1. Clayey shale

Results for the consistency limits of the clayey shale samples obtained from various layers are given in Table 1. The liquid limit (LL) varies from 202 to 46, the plastic limit (PL) varies from 121 to 22, and the plasticity index (PI) varies from 81 to 18. Soaked samples give higher values for LL, PL, and PI than ground samples. The specific gravity (G_s) of the clayey shale was found to be 2.824. The clay layers contain no sand, as indicated by wet sieving, which shows that 100% passes the #200 sieve.

Results from the XRD analysis are provided in Fig. 6 and Table 2. The approximate percentage weight fractions of the phases of different minerals are given in Table 2. The phase identification process involves calculating the ‘most likely match score’ for a given phase, based on peak intensity and peak position, when compared to a database of standard phases; while the weight fraction is calculated by comparing the intensity of the most intense peaks of that phase with standards. The intensities of the diffraction peaks are mainly governed by the relative amount of the material. Thus, it should be noted that the XRD is a semi-quantitative analysis technique primarily used on crystalline materials for determining the weight fractions of major crystalline phases, normally accurate down to 5%. The analysis of minor crystalline phases will be less accurate. The powder diffraction file (phase) number (PDF#) is also given in Table 2. The diffraction pattern is compared with the standard diffraction pattern for different phases established by the international committee on diffraction data (ICDD).

Results from the SEM and EDS analyses are shown in Fig. 7 and Table 3. Fig. 7(a) presents a SEM micrograph for area # 1 of a clayey shale sample with a magnification of 500×. The EDS spectrum from an areal analysis of this micrograph is shown in Fig. 7(b), and the weight percentage of elements determined from the SSQ analysis is given in column A1 in Table 3, which indicates that area # 1 has a high concentration of silicon (≈42%) and iron (≈31%). The SEM micrograph for area # 2 of the same shale sample is shown in Fig. 7(c), with its EDS areal analysis in Fig. 7(d). The elements present in area # 2 are given in column B1 in Table 3, which indicates that this area has a high concentration of chlorine (≈67%). Fig. 7(e) presents a SEM micrograph for a fresh fractured surface perpendicular to the bedding planes. The EDS spectrum from an areal analysis of this micrograph is shown in Fig. 7(f), and the elements found are listed in column C1 in Table 3, which indicates a high concentration of silicon (≈51%). The EDS for points # 1 and 2 from Fig. 7(e) are given in Fig. 7(g) and (h), respectively. The elements found in spots # 1 and 2 are provided in Table 3 column D1 and E1, respectively. Point # 1 has a high concentration of chlorine (≈72%), while point # 2 has a high concentration of silicon (≈60%).

4.2. Cemented sand

Typical curves for the particle size distribution of the sand, obtained from dry and wet sieving and hydrometer analyses, are shown in Fig. 8. Dry sieving shows that only 0.5% is finer than the #200 sieve, while wet sieving gives 14.0% passing the #200 sieve. A hydrometer analysis was conducted on the fine materials from wet sieving, and the results were used to complement the particle size distribution curve obtained from sieve analysis, Fig. 8. Consistency limits of the material passing the #200 sieve are provided in Table 1, which indicates that the values of LL, PL, and PI are very high compared to those obtained for most clayey shale samples. For wet sieving, the coefficient of uniformity, C_u, is 12.22, and the coefficient of curvature, C_c, is 4.94. Therefore, this sand is SC according to the Unified Soil Classification System (USCS), and A-2-7 according to the American Association of State Highway and Transportation Officials (AASHTO) classification system. This sand is naturally cemented by certain minerals as precipitated salts, which do not easily dissolve in water. Therefore, such a dry sand can be said to be a ‘collapsible’ soil, which requires certain precautions when used as a foundation or a construction material.

The particle size distribution shows that the material retained on the #200 sieve is almost uniform (poorly graded), which indicates that very considerable sorting has been done during the transportation and deposition processes. Also, the particle shape of these materials is seen under a regular microscope to have a high degree of rounding. Both characteristics of this sedimentary material reflect that the tranportational agent was air, i.e. aeolian sand (Lambe and Whitman, 1969). However, this sand has undergone some changes after deposition, and its surface texture reflects the long period of interaction with water.

The specific gravity of the sand (retained on the #200 sieve, from the wet sieving) was 2.728, while that of the fine materials was 3.004. These are higher than the usual specific gravity (G_s) for quartz sand, which is about 2.66. This is attributed to the presence of other minerals, especially in the fine materials, which are known to have a higher value of G_s than quartz, e.g. calcite, dolomite, and iron carbonate hydroxide. Lee et al. (1983) reported G_s values of 2.72–2.92 for calcite, 2.85 for dolomite, and up to 3.0 for iron-rich ferruginous soils.

Results of the direct shear tests on dry samples are shown in Fig. 9 in terms of the peak shear stress versus the normal stress, and the angle of internal friction versus shear deformation. The cohesion (c) is almost zero [Fig. 9(a)], because the cementation was destroyed upon crushing, and the samples were tested in a dry condition. The angle of internal friction (φ) was mobilized with shear deformation to a peak value of φ_p=43.7°, dropping to a residual value of φ_r=38°, Fig. 9(b). The full spectrum of φ was obtained, for cohesionless soils (c=0), in terms of normal and shear stresses (τ and σ) according to:

(1)

Results from the drained triaxial shear tests on dry samples are shown in Fig. 10, which reveal a peak angle of internal friction φ_trx=42°. It is well established in the literature that the direct shear test gives an angle of internal friction for plane strain (φ_ps) that is larger than φ_trx (Kullhawy and Mayne, 1990). Ladd et al. (1977) indicated that φ_ps is larger than φ_trx by 4–9° for dense sands, and by 2–4° for loose sands. Also, Lade and Lee (1976) presented the following relationship for φ_trx > 34°:

XRD results are shown in Fig. 11 and Table 4 for the gross sand sample, which show that quartz and albite are the main constituents, 45 and 30%, respectively, while calcite is only 20%. The XRD results for the fines (passing the #200 sieve) are shown in Fig. 12 and Table 5, which indicate that calcite is the main constituent (73%), while the quartz component is only 10%. This indicates that calcite is the coating material that cements the sand particles together.

SEM and EDS results are shown in Fig. 13 and Table 6 for a gross sample. Fig. 13(a) presents a typical SEM micrograph for a cemented sand specimen with a magnification of 50×. The EDS areal analysis of this micrograph is shown in Fig. 13(b), and the elements found are tabulated in column A2 in Table 6, which shows a high concentration of calcium (≈71%). This high percentage represents the surfacial material, which coats and cements the sand particles together. Fig. 13(c)–(e) present more magnified pictures that show the calcite precipitated on the surface of the sand particles, cementing them together. Fig. 13(f) presents an EDS point analysis for spot # 1 on a calcite crystal [ Fig. 13(e)], with a concentration of calcium of about 86%, as listed in column B2 in Table 6. Fig. 13(g) presents an EDS point analysis of spot # 2 on a sand particle [ Fig. 13(e)], with a concentration of silicon of about 68%, as given in column C2, Table 6. It is noticed that some sand particles have a porous surface [ Fig. 13(h)]. After examining their EDS [ Fig. 13(i)], it turns out that they are calcareous sand, with a calcium concentration of 76%, column D2, Table 6. The existence of some calcareous sand particles could be the result of the transgression–regression cycles of the Arabian Gulf during its geological history.

SEM and EDS results for fines found in the cemented sand are given in Fig. 14 and Table 7. Fig. 14(a) shows a typical SEM micrograph for fines found in the cemented sand, with a magnification of ×300. The EDS areal analysis of this micrograph is given in Fig. 14(b), and the elements found are listed in column A3 in Table 7, which indicates that the concentration of calcium is about 88%. Fig. 14(c) presents a more magnified picture of the fines, whereas Fig. 14(d) and (e) show the EDS at points # 1 and 2. The elements at these two points are given in Table 7, columns B3 and C3, respectively; with a major concentration of silicon (59%) at point # 1 and calcium (92%) at point # 2. This indicates that some of the quartz fraction has passed the #200 sieve.

4.3. Rocks

Since the rock samples are expected to be calcareous/carbonate rocks, based on the geology of the area, chemical tests were initially used. Different rock samples were identified by putting drops of hydrochloric acid (HCl) over them. Calcium carbonate (CaCO₃) rocks are identified by the effervescence of the carbon dioxide (CO₂) gas that comes out from them.

The results of the SEM study on the ‘white’ rock are presented in terms of photo micrographs, with magnifications of 50 and 200× in Fig. 15(a) and (b). These pictures show that this rock consists mainly of a semi-transparent pure calcite, as indicated by the EDS areal analysis of these micrographs [Fig. 15(c)]. In addition, the presence of white spots, whose EDS point analysis is given in Fig. 15(d), shows that they contain some sodium (Na), sulphur (S), chlorine (CL), and potassium (K), in addition to calcium (Ca).

Results of the XRD technique for the ‘white’ rock sample are presented in Fig. 16, which shows a sharp peak at 2θ=29.5°. The use of Bragg's law with n=1 and an X-ray wavelength λ=1.54 Å gives the following atomic basal spacing (d), which corresponds to CaCO₃ (pure calcium carbonate), i.e. calcite:

(3)

Results of thermal analysis techniques for the ‘white’ rock sample are provided in Fig. 17, in terms of the DTA curve (right axis) and the TG curve (left axis). From the DTA curve, a sharp endothermic peak exists between 765 and 950°C with a peak at 920°C. This peak is attributed to the decomposition reaction of calcium carbonate (solid) to calcium oxide (solid) and carbon dioxide (gas), as follows (Mackenzie, 1970):

The DTA is a convenient technique to differentiate between calcite and aragonite. Aragonite gives an irreversible endothermic peak in the range of 387–488°C, representing the phase transformation from aragonite to calcite (Mackenzie and Smykatz). Since the DTA ( Fig. 17) does not show another peak in that range, it follows that this calcium carbonate rock is not aragonite, but calcite.

However, the TG curve indicates that this reaction is associated with about a 44% reduction in the mass of the sample. This reduction in mass is due to the fact that CO₂ (gas) is lost. Calculating for compounds in the chemical reaction [Eq. (4)], CO₂ represents 43.96% of CaCO₃. It is well documented in the literature that decomposition of calcium carbonates is associated with a 44% reduction in mass (Haines, 1995).

5. Summary and conclusions

The variability in the type of geomaterials present in the investigated cut from clayey shale and limestone to cemented sand and calcite is attributed to their compositional characteristics, formation processes, and the various environmental conditions experienced during their geological history. The variability of soil properties observed in the Dhahran area can be explained in terms of the inherent heterogeneity and anisotropy caused by the existence of different types of geomaterials. Since such variation occurs in both the horizontal and vertical directions, even within a small area, extensive soil exploration is necessary to assess soil properties for various construction purposes. Studying the geology of the Dhahran area and a visual inspection of the face of the cut reveal that the exposed part of the investigated hill belongs to the Dammam formation (Tertiary, lower Eocene age), with some cemented sand deposits and crystallized calcite (Quaternary age). XRD and EDS results show that the yellow–brown shale from the Dammam formation is mostly palygorskite clay (80%). This shale is coated with halite (10%), and contains dolomite, ferroan (5%), and orthoclase (5%).

The sand pocket is cemented by calcite. This sand is poorly graded and consists of quartz (45%), albite (30%), and some fragments of calcium carbonates. Therefore, this sand could be an aeoleion sand drift that might have been saturated with water containing high concentrations of calcium and carbonate ions. However, the results of the sieve analysis indicate the importance of wet sieving for such naturally cemented sand. Wet sieving classifies this sand as SC instead of SP. Furthermore, shear strength tests give peak values of φ equal to 43.7 and 42° as obtained from direct shear and triaxial tests, respectively.

The results of this investigation suggest that the formation of the crystallized calcite at the interface between sand and clay or calcareous deposits (as shown in Fig. 4), may be relatively recent. Perhaps this calcite has crystallized as a result of the drying–wetting cycles of the solution (water with a high concentration of calcium and carbonate ions) trapped between the sand pocket and the other, less permeable deposits. Therefore, the calcium that had been leached from the calcium carbonate rocks precipitated as calcite at the interface between the high and low permeable soil formations. The precipitated crystals of calcite grow with drying–wetting cycles. The various techniques used in this study to determine soil mineralogy (i.e. XRD, SEM, and DTA) are successfully employed to complement each other in order to yield a more accurate assessment. This case study indicates that although geological maps are very useful for preliminary soil explorations, they need to be supplemented with field and laboratory tests for more reliable characterization of geomaterials.

Acknowledgements

The support provided by King Fahd University of Petroleum and Minerals in terms of laboratory, library, and computer facilities is greatly appreciated.

Fax: +966-3-860-2879; email: nshayea@kfupm.edu.sa