This study focused on the feasibility of Tunisian sediment reuse as new road material. Experiments were conducted to investigate sediment reuse as road material. Three mixtures (raw sediments, stabilised sediments with 3% lime and stabilised sediments with 3% cement) were studied with regard to their mechanical performances. The first part of this paper introduces the characterisation of Zarzis sediments. The second part describes the laboratory experiments and the followed methods. The third part is dedicated to the analysis of results. The originality of this work is the use of a minimum percentage of lime, which was determined experimentally. Based on environmental tests, Zarzis sediments were uncontaminated. Thus, they can be reused directly without a decontamination treatment process. Results show that humic compounds complicated sediment treatment with cement. Therefore, cement did not present an appropriate binder for stabilising and solidifying the investigated sediments. Based on mechanical results, a mixture with 3% lime can be used as new road material.
Notation
Introduction
Sediments are defined as the union of organic or mineral particles transported, precipitated and accumulated at the bottom of lakes, rivers (river sediments) or harbours (marine sediments) (Adams et al., 1992). Sediments in estuaries and harbour areas are essential components of the marine ecosystem that ensure the development of various aquatic organisms. These sediments are in constant contact with pollutants coming primarily from industrial waste and maritime traffic. The pollutants trapped in sediments can migrate to the fauna, flora and individuals and damage their health. These can pose a major ecological and environmental hazard (Adams et al., 1992; Hu et al., 2006; Rosenberg, 1977).
Marine sediments must be periodically dredged in harbours all over the world to ensure sufficient depth of water, to maintain harbour waterways and to reduce the risk of pollution of the marine ecosystem. According to statistics (Ices, 2011; Le Guyader and Colin, 2012), the Netherlands presents the greatest quantity of dredged sediments with a volume of 122 million m3 over the next 10 years.
Dredging is the first step in harbour management. Before dredging, an environmental analysis is made. The results are used to decide on the management method: dumping at sea or depositing on land. Land management poses problems because of the large volume and level of contamination. Dredging operation managers are required to adopt the least damaging solution for the environment. Thus, the usual solutions are used – namely, dumping at sea, land disposal with or without confinement, hopper overflow dredging, dredging by resuspension of sediments and formation of artificial islands from dredged sediments. Dumping at sea is permitted only when the concentrations of heavy metals and organic pollutants in sediments are below acceptability thresholds fixed by regulations. However, hydrodynamic agents in the discharge site may contribute to sediment remobilisation towards the harbour. Therefore, hydrodynamic effects are considered in the choice of disposal sites at sea to avoid sediment remobilisation. Land disposal requires monitoring and permanent maintenance. Dumping at sea and land disposal are constrained by national and international regulations and conventions such as Directive 2008/56/EC of the European Commission (EC, 2008), the Oslo/Paris Convention (Ospar Convention, 1992), the London Convention (IMO, 1972, 2013) and the Barcelona Convention (2011). Hence, sediment reuse becomes a key step in the methodology of harbour management in order to balance the costs of dredging and eliminate the migration risk of contamination to the environment. To achieve this goal, the economical point of view must be considered in the recycling of sediments, but this is not considered in this paper.
Researchers have started to study ways to reuse marine sediments in various building construction materials such as clay bricks (Hamer and Karius, 2002; Lafhaj et al., 2008; Weng et al., 2003), cementitious materials (Agostini et al., 2007; Limeira et al., 2011; Wang, 2009) and paving blocks (Said et al., 2015).
In civil engineering, road engineering consumes a significant amount of aggregates with different mechanical properties. Indeed, the recommended minimum values of the immediate California bearing ratio (CBR) (ICBR) change from one layer to another. Mechanical properties, such as unconfined compressive strength (UCS), CBR values and indirect tensile strength (ITS), depend on the class of the road and the position of the material in the pavement structure (Scordia et al., 2008). Thus, a wide range of materials is used in road engineering. Road engineering is a solution for dredged sediment reuse not only for its cost-effectiveness but also for the large volume of sediments needed. The sources of aggregates are theoretically almost limited, and many of them remain unusable for various reasons such as inaccessibility, integration of deposits in urban areas or in classified sites, excessive operating costs or environmental sensitivity problems (Michel, 2006). However, the use of recycled materials, whenever possible, can be an advantageous alternative. Through the literature, many researchers (Boutouil, 1998; Boutouil and Saussaye, 2011; Colin, 2003; Duan, 2008; Kamali et al., 2008; Liang, 2012; Maher et al., 2006; Saussaye, 2013; Scordia et al., 2008; Silitonga, 2010; Zentar et al., 2008) studied the possibility of sediment reuse in road construction. Based on the mineralogical composition and physical characteristics of Dunkirk sediments, Wang et al. (2012) assessed the effects of cement and lime through modified Proctor compaction and UCS tests. The potential of sediments solidified with cement or lime for road construction is evaluated through a proposed methodology from the ICBR value. In their reported research work, Wang et al. (2012) characterised Dunkirk sediment as sandy soil. They affirmed that from the point of view of mechanics and applicability in road construction, 6% cement is an economic and reasonable amount to improve the mechanical performance of sediments. This statement had been justified before by Colin (2003). Another type of research work was conducted by Tran (2009) on the reuse of sediments on road construction that concerns silty clay materials. The study required the use of corrective sands in formulations in order to obtain better results in terms of compaction and compression.
In the present study, the research work aims to evaluate the feasibility of the reuse of raw dredged sediments as an alternative material for road construction. To achieve this goal, sediments were extracted from the Zarzis harbour in Tunisia. Then, an accurate characterisation of dredged sediments was carried out in order to assess and classify the investigated sediments. The mechanical behaviours of the three mixtures were evaluated and compared to French thresholds recommended in Treatment of Soils with Lime and/or Hydraulic Binders: Application to the Construction of Pavement Base Layers (GTS) (Sétra, 2000). The experimental results on tested formulations demonstrated the feasibility of the beneficial reuse of investigated sediments in road applications. For this reason, modified Proctor, ICBR, UCS and ITS tests were carried out. The originality of this work is the use of a minimum percentage of lime, which was determined experimentally, and the application to Tunisian harbour sediments. Zarzis sediments were uncontaminated, which presents an advantage for sediment recycling. Thus, the investigated sediments can be reused directly without performing a pretreatment process.
Materials and methods
Sampling site
The Zarzis harbour is located in the south-east of Tunisia as shown in Figure 1. It is approximately 30 km south of Djerba Island, 50 km from the Gulf of Gabes and 80 km from the border with Libya. The harbour provides commercial exchanges for the region dealing mainly in the export of marine salt and crude petroleum and import of white oil products (CJB Environnement Inc and EAM, 2006). The Zarzis harbour has an oil wharf located in the extension of the commercial quay on the west side. The harbour has been declared, since 1996, as one of the two free zones in Tunisia. This nomination was intended for enhancing the economic development of the region. However, silting is a major problem in the Zarzis harbour that limits ships from accessing the port. The harbour is composed of three sections. The navigation channel, the turning circle and the commercial basin were initially designed with 13, 12 and 12 m water depths, respectively.
Zarzis harbour corresponding to the studied site emplacement (© Google Earth, 2017)
Zarzis harbour corresponding to the studied site emplacement (© Google Earth, 2017)
The marine sediments tested in present study were sampled from the Zarzis harbour in Tunisia (Figure 1). Samples are designated, as shown in Figure 1, from P1 to P13. The sediments were taken by a diver by using manual sampling. After dredging, all sampled sediments were transported to the laboratory and homogenised to obtain a raw sediment. Raw sediment was stored in hermetic containers.
Materials
Cement and lime
Cement treatments are very usual. Since the mechanical requirements for road materials are relatively low (Czerewko and Cross, 2015; Scordia et al., 2008) and based on what is usually used for road construction in Tunisia, ordinary Portland cement CEM I 42.5R was chosen in present study. The lime used in this work was a quicklime because it is the most widely used form on road construction sites. Lime changes the physical and hydric characteristics (Leroueil and Le Bihan, 1996), structure and properties of a fine soil (Cai et al., 2006). It also acts on the organic matter (OM), which inhibits the hydration reaction of cement (Dubois et al., 2009, Mohd Yunus et al., 2017; Tremblay et al., 2002).
Raw sediments
Characterisation of the physical properties of raw sediments such as particle size distribution, water content, specific unit weight, methylene blue value and OM content was performed. The sediments were tested at least in triplicate, and the average values were determined.
Determination of the particle size distribution of Zarzis sediments was done by using wet sieving and hydrometer methods according to French standards NF P94-056 (Afnor, 1996) and NF P94-057 (Afnor 1992a). Then, water content determination was performed by desiccation at 50°C for 24–72 h based on the NF P94-050 standard (Afnor 1995), and the specific unit weight (γs) for each tested sample was determined using a helium pycnometer. In order to investigate the clay activity of the studied sediments and their water sensitivity, the methylene blue test was carried out using NF P94-068 (Afnor, 1998a). Afterwards, the OM content of the sample was evaluated by loss on ignition at 450°C referring to the XP P94-047 standard (Afnor, 1998b).
Table 1 gives results of the characterisation of Zarzis sediments. The tested sediments are mainly silty sands with an average of 75% sand, 18% silt and 7% clay fraction. Results show the presence of 12·52% OM, which indicates that Zarzis sediments are moderately organic. The methylene blue value, representing the adsorption capacity of methylene blue on the particle surfaces, is low and correlated with the grain size distribution. However, the plasticity of studied dredged sediments is high.
Physical properties of Zarzis raw sediments
| Water content: % | 210 |
| Sand percentage: % | 75 |
| Silt percentage: % | 18 |
| Clay percentage: % | 7 |
| Methylene blue | 1·20 |
| Liquid limit LL: % | 83·50 |
| Plastic limit PL: % | 66·13 |
| Plasticity index PI | 17·37 |
| OM: % | 12·52 |
| Particle density: g/cm3 | 1·70 |
X-ray fluorescence (XRF) and X-ray diffraction (XRD) analyses of raw sediments were carried out. They allow quantification (by percentage weight) of the main basic chemical components of the sediments.
XRF analysis showed that the raw sediments mainly consist of calcium (Ca), carbon (C) and silicon (Si) and smaller percentages of chlorine (Cl), magnesium (Mg) and aluminium (Al). Figure 2 illustrates the XRD data obtained. Raw sediments were mainly composed of quartz (SiO2) and calcium carbonate (CaCO3).
In order to reuse the sediments as road materials, it is necessary to perform environmental characterisation of the heavy metal concentration in the sediment leachates. Indeed, the solid fraction test assesses the total concentration of various trace elements considered as contaminants. The pollution degree of Zarzis sediments was evaluated using inductively coupled plasma (ICP) optical emission spectroscopy according to the EN 12457-3 standard (Afnor, 2002). Tunisian authorities apply the Netherlands standards to study the possibility of sea dumping of dredged sediment according to the results of heavy metal concentrations. Threshold values are respected as strict limit values without exception according to Zoute Bagger Toets (ZBT) values. The latter values were introduced for assessing whether dredged material can be dispersed in the marine environment (DGE, 2007). In the present study, heavy metals in raw sediments are assessed according to the ZBT values.
The concentrations of heavy metals in the leachate of Zarzis sediments (Table 2) were under ICP apparatus detection limit values. Thus, Zarzis sediments may be incorporated into the formulation of an alternative material without any pretreatment process.
Mechanical performance tests
The minimum recommended tests to determine the feasibility of a specific material to be used in road engineering are the modified Proctor test, I CBR test, unconfined compression test and indirect tensile test.
The modified Proctor test was performed using the NF P94-093 standard (Afnor, 1993a) in order to know the water content corresponding to the optimal compaction.
The I CBR index defines the capacity of a material to support vehicle or engine traffic during road construction. It also measures also the ratio of the force required for a circular piston to penetrate into a granular medium in a CBR mould at a speed of 1·27 ± 0·1 mm/min. The prescribed I CBR values for different road layers are specified in French standard NF P98-115 (Afnor, 1992b). For a subgrade layer, the minimum recommended value of I CBR is 20 (Sétra, 2000).
The U CS test was performed according to the NF P98-230-2 standard (Afnor, 1993b). The objective of the U CS test is the determination of the traffic ability of the subroad layer. According to the NF P98-115 standard, the traffic ability criterion is considered satisfactory since the U CS is greater than 1 MPa. Samples for U CS test were specifically manufactured using a poly(vinyl chloride) (PVC) sample holder. The specimens were prepared at the optimum water content. The samples were cylindrical with a diameter of 50 mm, a height of 100 mm and a slenderness equal to 2.
For the three mixtures, see also the section headed ‘Mechanical performance tests’ and Table 3; the U CS values were determined after 2, 7, 28, 60 and 90 d of maturation. The samples were stored in normal curing conditions in a room maintained at 20°C.
Designation of mixtures
| Mixture | Designation |
|---|---|
| Raw sediments | Raw sediments |
| Sediment stabilised with 3% lime | SS-3L |
| Sediment stabilised with 3% cement | SS-3C |
In order to study the mechanical tensile performance of the treated layer, ITS R t values were determined at 28 and 90 d of maturation. An ITS test was performed on monolithic samples. The specimens were prepared at the optimum water content and using a PVC tube as a mould. The dimensions of the samples for ITS tests are different from those of the U CS samples. The ITS samples were cylindrical with a diameter of 100 mm, a height of 100 mm and a slenderness equal to 1.
The tensile strength, R t, is calculated according to Equations 1 and 2
where R t is the tensile strength in megapascals; R tb is the compressive resistance along the sample length – that is, the sample height – in megapascals; F r is the failure strength in newtons; φ is the sample diameter in centimetres; and h is the sample height in centimetres.
The elastic modulus, E, is deduced from U CS testing. The mechanical road class is determined with the tensile strength and Young’s modulus of a treated material. The mechanical road class corresponds to the zone number, from 1 to 5, where the material is located in the classification chart defined in GTS (Sétra, 2000). Zone 5 corresponds to the lowest mechanical performance. This classification is important since the cost of layer design depends directly on it. Zone 5 is the minimum required, but contractors usually consider that it is important to be located at least in zone 3 in order to be of technical and economic interest (Scordia et al., 2008).
Mix design
An experimental study was performed on raw mixed sediments with different percentages of cement or lime to evaluate their reuse potential in road construction. The percentages of lime or cement mixed with fine sediments were fixed to 3% of the dry mass of raw sediments. The lime content was determined by the limit fixation test. This test consists of measuring the potential of hydrogen (pH) of a demineralised water and sediment solution. A percentage of quicklime was added until the pH became constant. Figure 3 shows that the pH stabilised for a lime dosage of 3%.
Achour (2013) used the lime saturation curve to determine the minimum percentage of lime. In his study, Achour (2013) investigated the feasibility of reusing sediments in road construction. Marine sediments were dredged from the Grand Port Autonome of Dunkerque and taken from the eastern basin Darse 6. In Achour’s (2013) research, the pH stabilised for a 1% lime dosage.
Three mixtures were tested in the present study as seen in Table 3. GTS (Sétra, 2000) recommends a maximum of 6% of cement addition for economic issues. For this reason, 3% of cement was fixed based on the present study.
The U CS test, indirect tensile test and elastic modulus determination were performed on each formulation. Raw sediment was mixed with binders (lime or cement) and the optimum water content by using a mixer for 5 min. The samples were cured in a specified room at a constant temperature of 20°C.
Results and discussion
The modified Proctor, ICBR, UCS, tensile strength and elastic modulus results are discussed in this section.
Figure 4 presents the modified Proctor compaction results for raw sediments, stabilised sediments with 3% lime (SS-3L) and stabilised sediments with 3% cement (SS-3C). It shows the dry density as function of water content. The maximum dry densities of raw sediments, SS-3C and SS-3L are 0·91, 0·93 and 0·95 g/cm3, respectively. The increase in the maximum dry density by adding cement is related to the pore-filling effect of cementitious materials from a series of chemical reactions in binder, water and sediments. Compared to raw sediments, the addition of lime increases the optimum water content from 22 to 27%. Indeed, lime hydrates rapidly, which results in increasing demand of the initial water mixture.
Dry density plotted against water content for raw sediments (black), SS-3L (red) and SS-3C (green)
Dry density plotted against water content for raw sediments (black), SS-3L (red) and SS-3C (green)
Figure 5 shows the immediate bearing capacity ratio (ICBR) as function of water content of the three mixtures. At the optimum water content, the ICBR values of raw sediments, SS-3C and SS-3L were 27, 24·3 and 59·4%, respectively. The ICBR value decreased from 27 to 24·3% with addition of cement. The variation in ICBR values depends on many factors such as mineralogical composition, pollutant content, water content, binder type and binder amount. Thus, the decrease in the ICBR value can be related to the organic content of raw sediments (12·52%). Indeed, the presence of humic compounds in a soil complicates its treatment with cement. Humic acid has an inhibitory action on the hydraulic setting of the cement (Kujala and Makikyro, 1996; Liang and Levacher, 2012; Rey et al., 2000; Tremblay et al., 2002). It is also known that the presence of silt or clay affects the reaction between the soil and cement. Miura et al. (2001) demonstrated that the water content, the cement content and the curing period influence the properties of cement-treated clays. The presence of silt, a material that neither swells nor compresses, should improve the engineering properties of cement-treated soft clay by reducing the potential for recompression and settlement of the clay (Chairat and Panich, 2014). This is can be also a justification for the decrease in the ICBR value for the SS-3C mixture.
ICBR plotted against water content for raw sediments (black), SS-3L (red) and SS-3C (green)
ICBR plotted against water content for raw sediments (black), SS-3L (red) and SS-3C (green)
The normal proceeding of cement stabilisation can be achieved by eliminating OM before treatment or by neutralising humic acid with a lime pretreatment. The reaction between lime and humic acid leads to the formation of calcium humates, which are not harmful with cement addition. Lime reduces the acidity of OM by neutralising pathogens that are responsible for presence of humic compound in sediments (Croise, 1964). It is recommended, for Zarzis sediment, to investigate combined lime-and-cement formulation in further work.
SS-3L presents the highest ICBR value (ICBR = 59·4%), thus corresponding to the best result. According to ICBR results for raw sediments, SS-3C and SS-3L can be used as subgrade layers according to the NF P 98-115 standard.
Figure 6 shows the evolution of UCS as a function of curing time. Depending on time, the assessment of the achieved mechanical performance differs from one formulation to another. It is clear that the mixture with 3% lime shows the highest values of UCS. Raw sediments and SS-3C show the same behaviour. The UCS values for raw sediments and SS-3C are almost the same. It is clear that cement does not influence sediment treatment for the UCS test.
Unconfined compression (UCS) plotted against time for raw sediments (black), SS-3L (red) and SS-3C (green)
Unconfined compression (UCS) plotted against time for raw sediments (black), SS-3L (red) and SS-3C (green)
A road layer is useful when its UCS is greater than 1 MPa (Sétra, 2000). From Figure 6, this value is not reached for raw sediments and SS-3C, and it is reached on the third day for SS-3L. This mixture presents a UCS equal to 2·78 MPa at 90 d, while raw sediments and SS-3C show, respectively, UCS values equal to 0·76 and 0·59 MPa for the same curing time (90 d).
Table 4 gives the tensile strength Rt and the elastic modulus E results of mixtures at 28 and 90 d of maturation. Figures 7 and 8 show the mechanical class of each mixture at 28 and 90 d of cure, respectively. According to Scordia et al. (2008), the reuse of sediments represented a real technical and economic interest if the envisaged treatments were at least in zone 3. However, it is possible to use a treatment in zone 4 or 5, but the thickness of the layer and the cost of its design become higher (Scordia et al., 2008). It can be observed that at 28 d, none of the binders make it possible to achieve the fixed level of performance (zone 3). At 90 d, SS-3L was located in zone 3. It is important to note that the classification zone was similar from 28 to 90 d of maturation for raw sediments and SS-3C. This similarity confirms the observations made during the compression tests where the maximum strengths of raw sediments and SS-3C were reached rapidly. In their research, Scordia et al. (2008) observed that at 28 d, none of the binders make it possible to achieve zone 3. On the other hand, at 90 d, treatments with the hydraulic binder Roc Sol and lime are located in zone 3.
Tensile strength, elastic modulus and classification at 28 and 90 d of maturation of mixtures: raw sediments, SS-3L and SS-3C
| Length of maturation: d | Mixture | Rt: MPa | E: MPa |
|---|---|---|---|
| 28 | Raw sediments | 0·12 | 1480 |
| SS-3L | 0·26 | 2854 | |
| SS-3C | 0·15 | 2206 | |
| 90 | Raw sediments | 0·14 | 1860 |
| SS-3L | 0·33 | 3713 | |
| SS-3C | 0·19 | 2295 |
Conclusion
This work aims to assess the feasibility of reusing dredged sediments in road construction. The tested sediments were dredged from a Tunisian harbour suffering from a silting problem. This issue generates a large amount of untreated sediments, so its reuse on road construction becomes an economic benefit. A subgrade layer consumes a large amount of raw materials, and the reserves of aggregates are theoretically almost limited. For these reasons, the choice of sediments to be reused in the subgrade layer was studied in this research.
Three mixtures were tested: raw sediments, stabilised sediments with 3% lime (SS-3L) and stabilised sediments with 3% cement (SS-3C).
The results show that the presence of humic compounds in sediments complicated their treatment with cement. Therefore, cement did not present an appropriate binder for stabilising and solidifying the tested sediments. SS-3L can be used as subgrade layer according to the ICBR results. The UCS demonstrated that a layer with the SS-3L mixture can be useful from the third day. Indeed, UCS reached a value equal to 2·78 MPa at 90 d of maturation, while raw sediments and the SS-3C mixture showed UCS values equal to 0·76 and 0·59 MPa, respectively, at the same curing time (90 d).
Based on tensile strength and elastic modulus at 90 d of maturation, SS-3L was located in zone 3 according to the mechanical performance chart of the GTS (Sétra, 2000). Thus, the reuse SS-3L formulation is of real technical and economic interest.
This study highlighted the possibility of the reuse of stabilised sediments as a subgrade layer. The SS-3L samples showed satisfactory mechanical performance results. Indeed, the ICBR value for SS-3L mixture was higher than the raw sediments and SS-3L mix results. From 2 to 90 d of maturation, UCS ranged from 0·45 to 2·78 MPa for SS-3L, from 0·2 to 0·76 MPa for raw sediments and from 0·15 to 0·59 MPa for SS-3C.
According to the recommendations of the technical guide GTS (Sétra, 2000) and based on ITS and elastic modulus, SS-3L was located in zone 3, at 90 d of curing. For Zarzis sediments, cement was not an appropriate binder for improving sediment performance, contrary to Wang et al.’s (2012) research. Indeed, in Wang et al.’s (2012) study, the unconfined compressive strengths and tensile strengths increased with cement content and curing time and the mechanical performance of sediments solidified with cement was much superior to that of sediments solidified with lime.
In the literature, many research works recommended the investigation of combined cement and lime mixture. Zentar et al. (2008) studied the feasibility of reusing sediments in combination with cement and lime as alternative road materials. The optimisation procedure used to design the mix has shown its ability to improve the main characteristics of the raw fine sediments. In terms of engineering properties, such as UCS, ICBR and elastic modulus, the added components have allowed the engineering characteristics of dredged fine sediments to increase substantially. Laboratory study has revealed the ability of the designed mix to be used as a sub-base material for a low intensity of road traffic. It is recommended to study the combined cement and lime mixture for Zarzis sediment in further works.
Acknowledgements
This study was funded by the partnership Hubert Cruien ‘Utique’ of the French Ministry of Foreign Affairs and the Tunisian Ministry of Higher Education and Scientific Research in Tunisia (Project Number 14 G 1116). The authors express their acknowledgements to the Tunisian Merchant Marine and Ports Office and the Tunis International Center for Environmental Technologies.
