Analysis of the Influence of Excavated Soil Sand Characteristics on the Rheological and Mechanical Properties of Hydraulic Mortars

Abstract

This work investigates the effects of substituting natural sand with excavated soil sand in the formulation of hydraulic mortar developed from a self-compacting concrete (SCC). Four excavated soil sand deposits were studied to assess their physicochemical properties. Subsequently, a reference mortar (RM) was designed using the concrete equivalent mortar method. Furthermore, the effect of incorporating 30% of excavation soil sand under different moisture conditions (natural storage conditions, dry and saturated surface dry state) on the properties of mortar is studied. Spreading tests were carried out to observe how the rheological properties evolve over time. The study includes compressive and flexural strength tests at 2, 7, 14 and 28 days. The results showed that some sands had densities similar to those of natural alluvial sand, while others had lower densities. Water absorption values varied considerably from one sand to another, with some showing values ranging from 1% to 6%, while other sands had values of up to 10%. The results of spreading tests indicate that mortar made with sand in a saturated dry-surface state is more fluid than mortar made with sand in a dry state. Under all conditions, all mortars lose their fluidity over time. The variation in compressive strength among all excavated soil sand mortars compared to the reference mortar remained below 10% at 2 and 28 days, except for one sand with a high clay content. The incorporation of excavated soil sand at this percentage as a substitute for river sand led to an enhancement in the flexural strength of the mortar, with improvements of 40% and 50% observed for certain types of excavated sand. The statistical study revealed a strong relationship between the properties of the sand (in particular, the fines content and their nature, as well as the sand skeleton) and its saturation state, the flowability and the compressive strength of the mortar.

1. Introduction

Every year, France generates around 326 million t of waste [1], with a significant portion (224 million t) originating exclusively from the building and public works sector [2]. As the primary contributor to waste production in France, the construction sector positions the country as the leading producer of construction waste in Europe. Inert materials represent a substantial share of this waste, amounting to 211.3 million t. These inert materials are derived from various sources, including soil excavation, which generates an estimated 150 million t annually, and the demolition of concrete structures, contributing around 19 million t per year.

Meanwhile, France’s annual demand for aggregates is estimated at 456 million t. For concrete production alone, this demand reaches 125 million t per year, of which only 6.8 million t are met using recycled aggregates [3]. This leaves a significant shortfall of 118 million t that must be supplied by virgin materials.

Fine aggregates, commonly represented by sand, are the portion of aggregates ranging in size from 0.63 mm to 4 mm. They constitute 30–50% of the total volume of concrete [4]. To produce a high-quality concrete mix, it is essential for the fine aggregates to be clean, exhibit consistent properties, and to be free from chemicals or contaminants that could negatively impact the concrete’s rheological, mechanical and durability characteristics [5].

To date, a wide range of research studies have been carried out on recycled concrete aggregates (RCA) [6,7,8,9,10,11,12,13,14] to investigate their effects when incorporated into concrete. Nevertheless, the use of recycled sand derived from demolished building waste as a substitute for river sand in concrete is very limited, due to its inherent porosity, tendency for microcrack formation, and elevated fine particle content [15,16,17]. These properties vary according to the recycling methods used and the quality of the original concrete. The latest edition of the NF EN 206+A2/CN standard published on 5 November 2022 [18] allows the use of RCA from deconstruction for concrete structures, specifying the limits of use (article NA.5.1.3 Aggregates) for each type of recycled aggregate. For instance, a maximum replacement percentage of 10% of recycled concrete sand is authorized for concretes in exposure classes XC1 and XC2.

Although numerous studies have been conducted on RCA, aggregates derived from excavated soils are relatively underexplored in the literature and have not been studied in the context of self-compacting concrete (SCC). A detailed summary of the relevant literature, including the references, materials, problems in raw materials, treatment methods, and the behavior in terms of fresh properties, strength properties, and durability properties, are represented in Table 1. Excavated soils have a real density ranging from 2300 kg/m³ to 2700 kg/m³ [19,20,21,22,23]. Regarding water absorption, excavated soils have an absorption ranging from 10% to 35% [19,23,24], much higher than that of natural aggregates. These properties may vary depending on the composition of excavated soils with different mineralogical natures. Wu et al. [25] studied the effect of partial or total replacement (30%, 50%, 70% and 100%) of river sand by three batches of excavated soil sand (soil 1, 2 and 3) from different construction sites. As the percentage of excavated soil sand replacement increases, the compressive strength of cement mortar decreases for two soil batches, with a remarkable 26% reduction at 28 days for 70% replacement. In contrast, the cement mortar formulated with the third excavated soil showed a compressive strength comparable to that of a reference mortar. For flexural strength results, in the case of excavated soils 1 and 2, two trends were observed: for a replacement percentage of less than 30%, an increase in the replacement rate resulted in a 10% increase in flexural strength. On the other hand, for replacement rates above 30% (e.g., for a replacement percentage of 70%), flexural strength decreases as the replacement rate increases (e.g., an 11% reduction). In the case of excavated soil 3, the cement mortar based on excavated soil received a flexural strength comparable to that of a cement mortar based on natural fine aggregate. Wu et Zhang [26] studied the effect of incorporation of excavated soil sand (ES) as fine aggregates in concrete. At 7 and 28 days, within a range of replacement by ES of 0–70%, a maximum decrease of 11.4% and 8.9% of compressive strength was observed, respectively. At day 90, the compressive strength was not less than that of the concrete with NS.

To improve the properties of excavated soils, several methods have been explored in the literature: wet sieving [22], drying, grinding, and sieving [27], sediment separation equipment and soil sieving [28]. According to Wu et al. [27], the clay content decreased from 36% to 5.5% through soil processing via dry grinding and sieving. Priyadharshini et al. [22] studied the impact of wet sieving treatment on enhancing fine aggregates and cement mortar properties, using river sand (Sand > 63 µm: 99.2%, Clay < 2 µm: 0%, Fines > 2 µm < 63 µm: 0.8%); raw soil-1 (Sand: 71%, Clay: 15%, Fines: 14%) and raw soil-2 (Sand: 52%, Clay: 25%, Fines: 23%); and wet-sieved sand 1 (Sand: 71%, Clay: 15%, Fines: 14%) and 2, where 71% and 52% of the sand was recovered from the raw excavation soil-1 and 2, respectively, after wet sieving. The fine aggregate to cement (A/C) ratio of the mortar mix was varied from 3 to 5 by weight. The density of the mortar with wet-sieved excavated soil increases to approximately 1925 kg/m³ to 2150 kg/m³ for wet sieved sand 1 and 2, respectively, approaching that of the control mortar (2180 kg/m³). The compressive strength of mortar reduces with increasing clay content. The improvement in compressive strength with wet sieving was 107% and 161% for wet-sieved sand-1 and wet-sieved sand-2 compared to raw excavation soil-1 and soil-2, respectively. However, there is a reduction of 43% in the strength of mortar with wet-sieved sand-2 compared to the control mix. Moreover, given that excavated soil sand exhibits high water absorption similar to that of demolition concrete aggregates, the moisture content of the sand may also impact the fresh state of mortar and concrete. Many studies showed that the moisture state of recycled aggregates has a significant impact on concrete workability. Cyr et al. [29] investigated the evolution of conventional concrete rheology within the first 90 min. They examined how this evolution correlated with the water content of recycled aggregates, specifically when 30% of natural sand was replaced with recycled concrete sand. Three hydrous states of the recycled sand were characterized, defined based on the water absorption (WA) measured at 24 h (NF EN 1097-6 [30]): water contents of 120%, 87% and 33% of WA. They observed that the loss of slump over time is less when the recycled aggregates are in a hydrous state far below saturation. A moisture state close to the absorption coefficient gives the concrete the best flowability.

This study aims to enhance the utilization of construction waste by substituting alluvial sand with excavated soil sand in the formulation of mortars derived from self-compacting concrete (SCC) intended for precast industry, a domain that remains underexplored in existing literature. The primary objective is to investigate the feasibility of incorporating excavated soil sand into mortar formulations. The novelty of this research lies in two key aspects: (1) a comprehensive characterization of excavated soils to understand their variability and evaluate their potential as a substitute for alluvial sand in SCC-derived mortars; and (2) an assessment of the effects of incorporating 30% excavated soil sand on the properties of SCC mortar using the concrete equivalent mortar method. Additionally, the study investigates the influence of three distinct moisture states of the sand on the rheological and mechanical properties of the mortar. These findings aim to demonstrate the potential of excavated soil sand for concrete applications by first evaluating its performance in equivalent mortars, providing a foundation for future research on its integration into self-compacting concrete. Furthermore, this study addresses environmental challenges by promoting the reuse of construction waste.

Table 1. The literature summary of the use of excavation soil sand in mortar and concrete.

References	Material	Problems in Raw Materials	Treatment Methods	Behavior in	Fresh Properties	Strength Properties	Durability Properties
Huang et al. [28]	Excavated soil recycled fine aggregates (ESRFA)	Fine content, less roundness and less uniform particle size distribution than river sand	Sediment separation and sieving	Mortar	River sand mortar exhibited superior flowability to ESRFA mortar	The flexural and compressive strength of mortar increased at 30% of replacement and then decreased upon increasing the ESRFA content	Porosity, water absorption, and dry shrinkage increased at 30% of replacement then decreased for higher replacement
Wu et Zhang [26]	Sand from weathered residual soil of granite (RS) and recycled concrete coarse aggregates (RCA)	Lower fineness modulus, more porous and higher water absorption than natural sand for RS; lower packing density and higher crushing index and water absorption for RCA	Wet sieving or grinding and dry sieving to produce recycled sand (RS) from soil	Concrete	-	Faster increase in compressive strength of concrete containing RS than the concrete with NS, the compressive and flexural strength at 90 days were not lower than the concrete with NS; the 7 d, 28 d, and 90 d compressive strength generally decreased with the utilization of 50% RCAs	-
Wu et Xiong [20]	Manufactured sandstone (MSS) of excavated waste from underground construction	The sandstone specimen contained a significantly smaller proportion of quartz and a higher proportion of illite (non-expansive clay mineral)	Raw material crushing; dry sieving or washing for removing stone powder	Mortar	The workability of fresh mortar is dominated by the stone powder content of MSS, while the effect of surface roughness of MSS on the workability of mortar is negligible	The compressive strength of mortar group with 30% replacement of NS by MSS was similar to the control group without MSS, a loss of 20% is obtained with 50% MSS replacement	-
Priyadharshini et al. [22]	Excavation soil	Presence of clay and fines	Granulometry to improve granular skeleton. Wet sieving to remove clay and fines	Concrete	Up to 40% clay could be used for laterized concrete though lower percentages are preferable. Less W/C ratio results in poor compaction	Compressive strength reduced with increasing percentage of excavation soil.	Raw earth-based materials present high shrinkage and tendency to crack during drying.
Guan et al. [23]	Abandoned excavated soil that taken from a foundation pit	High moisture content, inorganic soil of low plasticity. About 95% of the soil mass has a particle size of less than 1 mm, and particles smaller than 75 m account for 64% of the total mass	Natural air-drying, then crushing, and passing through a 2 mm sieve	Foamed concrete	When the soil content was adjusted from 25% to 50%, the workability was reduced of 15.79%	A reduction of 80.73% (from 1.09 MPa to 0.21 MPa) when the soil content enlarged from 25% to 50%	-

Table 1. The literature summary of the use of excavation soil sand in mortar and concrete.

References	Material	Problems in Raw Materials	Treatment Methods	Behavior in	Fresh Properties	Strength Properties	Durability Properties
Huang et al. [28]	Excavated soil recycled fine aggregates (ESRFA)	Fine content, less roundness and less uniform particle size distribution than river sand	Sediment separation and sieving	Mortar	River sand mortar exhibited superior flowability to ESRFA mortar	The flexural and compressive strength of mortar increased at 30% of replacement and then decreased upon increasing the ESRFA content	Porosity, water absorption, and dry shrinkage increased at 30% of replacement then decreased for higher replacement
Wu et Zhang [26]	Sand from weathered residual soil of granite (RS) and recycled concrete coarse aggregates (RCA)	Lower fineness modulus, more porous and higher water absorption than natural sand for RS; lower packing density and higher crushing index and water absorption for RCA	Wet sieving or grinding and dry sieving to produce recycled sand (RS) from soil	Concrete	-	Faster increase in compressive strength of concrete containing RS than the concrete with NS, the compressive and flexural strength at 90 days were not lower than the concrete with NS; the 7 d, 28 d, and 90 d compressive strength generally decreased with the utilization of 50% RCAs	-
Wu et Xiong [20]	Manufactured sandstone (MSS) of excavated waste from underground construction	The sandstone specimen contained a significantly smaller proportion of quartz and a higher proportion of illite (non-expansive clay mineral)	Raw material crushing; dry sieving or washing for removing stone powder	Mortar	The workability of fresh mortar is dominated by the stone powder content of MSS, while the effect of surface roughness of MSS on the workability of mortar is negligible	The compressive strength of mortar group with 30% replacement of NS by MSS was similar to the control group without MSS, a loss of 20% is obtained with 50% MSS replacement	-
Priyadharshini et al. [22]	Excavation soil	Presence of clay and fines	Granulometry to improve granular skeleton. Wet sieving to remove clay and fines	Concrete	Up to 40% clay could be used for laterized concrete though lower percentages are preferable. Less W/C ratio results in poor compaction	Compressive strength reduced with increasing percentage of excavation soil.	Raw earth-based materials present high shrinkage and tendency to crack during drying.
Guan et al. [23]	Abandoned excavated soil that taken from a foundation pit	High moisture content, inorganic soil of low plasticity. About 95% of the soil mass has a particle size of less than 1 mm, and particles smaller than 75 m account for 64% of the total mass	Natural air-drying, then crushing, and passing through a 2 mm sieve	Foamed concrete	When the soil content was adjusted from 25% to 50%, the workability was reduced of 15.79%	A reduction of 80.73% (from 1.09 MPa to 0.21 MPa) when the soil content enlarged from 25% to 50%	-

2. Materials and Methods

2.1. Materials

2.1.1. Cement, Filler and Admixtures

Portland cement, designated CEM I 52.5 R CE CPE NF, with a density of 3.13 g/cm³ and a Blaine surface area of 4815 cm²/g, sourced from Lafarge, Usine de Saint-Pierre-la-Cour, France, was used. The powder fraction was supplemented with limestone fillers with a specific surface area of 4050 cm²/g and a density of 2.7 g/cm³, supplied by FACO—Carrière de Pareds, France. To meet the requirements of self-compacting concrete, a superplasticizer, MasterGlenium ACE 550, suitable for precast concrete, was employed, along with a hardening accelerator, MasterX-SEED 100, sourced from Master Builders Solutions, BASF France, Lisses, France.

2.1.2. Sand

Two types of sand were utilized in this study: natural alluvial sand and excavated soil sand obtained from various stripping sites (Figure 1). The references for each type of sand are provided in

Table 2. List of aggregate deposits.

Abbreviation	Particle Class (mm)	Type	Origin
NS	0/4	Alluvial	Pays de La Loire
ESS-SP	0/4	Excavation	Saint-Prouant
ESS-LB	0/4	Excavation	Le Boupère
ESS-SG	0/4	Excavation	Saint-Germain
ESS-LT	0/4	Excavation	La Trance

.

The natural alluvial sand used in this study, sourced from Loire-Atlantique, is a Pliocene siliceous alluvial sand that is washed and rounded, with a grain size distribution ranging from 0 mm to 4 mm and a low content of particles smaller than 0.063 mm. The excavated soils were sourced from construction sites located within a 30 km radius of AlainTP company’s separation platform, as shown in Figure 2. The separation of sand, lime, and clay was achieved using centrifugal and sedimentation processes [31]. Initially, the excavated soils underwent treatment involving immersion in water to separate aggregates from floating debris and to collect the suspended water. This water is then processed using centrifugal forces to separate lime, followed by sedimentation to settle clay at the basin’s bottom. This process facilitated the recovery of sand with reduced clay and lime content.

The “Swelling-Soil Shrinkage Hazards” map from the “Bureau de Recherches Géologiques et Minières” (BRGM) [32] (Figure 2) indicates varying levels of hazard for the excavation sites. The excavated site in Saint-Germain is located in an area with medium hazard. In contrast, the site at La Tranche-sur-Mer is located in an area where the hazard is presumed to be zero. Meanwhile, the sites in Saint-Prouant and Le Boupère are classified as being in areas with a low hazard level.

2.2. Physico-Chemical Characterization of Sand

The grain size distribution of the sands was determined in accordance with the NF EN 933-1 (2012) standard [33]. The fineness modulus and the fines content were calculated based on the percentages of material retained on a predetermined series of sieves following the NF EN 13139 (2003) standard [34]. The relative quantities of clayey or plastic fines and dust in the sand were determined using the sand equivalent test, which was conducted in accordance with the NF EN 933-8 (2015) standard [35]. In order to reveal the presence of clayey fines and determine their concentration, the methylene blue test was performed following the NF EN 933-9 standard [36]. To calculate the real density and the water absorption of sands, the NF EN 1097-6 (2014) standard [30] was employed. As regards the chemical aspects of these sands, the following standards are followed: NF EN 1744-1 [37] for water-soluble sulfate content and for water-soluble chloride content, and NF P 18-454 [38] for alkali content.

2.3. Composition of Mortar Mixtures

In this study, the concrete equivalent mortar method (CEM) developed by Schwartzentruber et al. [39] was employed to assess the impact of replacing natural sand with excavated soil sand by 30% on the characteristics of mortar in fresh and hardened states. The reference concrete used for this method is a self-compacting concrete (SCC) of class C40/45. The formulation approach of the CEM is summarized by determining a surface developed by aggregates modeled as perfectly spherical particles to simplify the calculation of the surface and volume of each granular class. The calculation of the CEM involves replacing coarse aggregates (diameters greater than 4 mm) with sand whose developed surface is equal to that of the gravel.

Additionally, the effect of different moisture states of sand on mortar properties was also investigated, including the dry state (DS), saturated surface dry state (SSD), and natural storage state (NSS). To achieve the dry state, the sand was dried in a ventilated oven at a temperature of 105 °C until a constant mass was obtained, following the method recommended in the NF 1097-6 standard [30]. For the saturated surface dry state, the sand was dried similarly and then mixed with water equivalent to its absorption capacity in a sealed pycnometer for 24 h at 20 °C. The pycnometer is thoroughly shaken during the first hours to ensure homogeneous distribution of water in the sand. The natural storage state of the sand is subject to variation in temperature and relative humidity, as it is stored outdoors.

Fifteen CEM mixtures were evaluated, including three reference mortars made with natural sand (M-NS), considering the three moisture states, as well as twelve mortars with a 30% volume replacement rate of natural sand (M-ESS-SP-30, M-ESS-LT-30, M-ESS-SG-30, M-ESS-B-30) for each of the four different sands in different moisture states. The formulations for the different sands are presented in

Table 3. Equivalent concrete mortar formulation.

Constituents	M-NS	M-ESS-SP-30	M-ESS-LT-30	M-ESS-SG-30	M-ESS-B-30
${\mathrm{W}}_{\mathrm{eff}}/\mathrm{C}$	0.56	0.56	0.56	0.56	0.56
${\mathrm{W}}_{\mathrm{eff}}/\mathrm{F}$	0.335	0.335	0.335	0.335	0.335
S/C	2.74	2.55	2.71	2.65	2.51

. The mortar was formulated while maintaining a constant ratio of 0.56 for effective water to cement (${\mathrm{W}}_{\mathrm{eff}}/\mathrm{C}$) and 0.335 for effective water to cement and filler (${\mathrm{W}}_{\mathrm{eff}}/\mathrm{F}$). The sand to cement ratio (S/C) was adjusted based on the mass of recycled sand required to maintain a constant sand volume to that of the reference mortar. The total water added was adjusted for the water absorption of the different sands to maintain a fixed effective water-to-cement ratio.

2.3.1. Fresh State Mortar Properties

The properties of fresh mortar were evaluated through spread tests (Figure 3), conducted in accordance with the NF EN 12350-8 (2019) standard [40]. These tests were performed at specific time intervals: immediately after mixing (t = 0 min), after 15 min (t = 15 min), and after 30 min (t = 30 min), without any further mixing during the resting period in the container. This approach aimed to simulate real conditions in the precast industry, where the concrete may remain at rest in a bucket before casting. The spread measurements were carried out using a specialized mini-cone designed for CEM characterization [39]. The dimensions of this cone were derived from the Abrams cone, scaled down with a homothetic ratio of two. The mini-cone had an upper diameter ${\mathrm{D}}_{\mathrm{upper}}=50\mathrm{mm}$, a lower diameter ${\mathrm{D}}_{\mathrm{lower}}=100\mathrm{mm}$, and a height h = 150 mm.

2.3.2. Properties of Hardened State Mortar

The studied properties of hardened mortars include flexural strength and compressive strength. Flexural strength tests were conducted according to the NF EN 1015-11/A1 (2019) standard [41] under three-point loading conditions on prismatic specimens until failure (Figure 4a). Prisms with dimensions of 160 mm × 40 mm × 40 mm were prepared in metal molds, demolded after 24 h, and then placed in airtight bags for a period of up to seven days. After this period, they were stored under relative humidity conditions of 65 ± 5% and at a temperature of 20 °C for the remaining curing period until testing. Compression strength tests were carried out according to the same standard [41] on one of the two halves of each prism resulting from the flexural strength tests (Figure 4b). Samples underwent flexural and compression tests at various intervals: at t = 2 days, t = 7 days, t = 14 days, and t = 28 days.

3. Results and Discussion

3.1. Properties of Sand

The properties of the different sands are summarized in Table 4. For each test, three samples were analyzed to ensure reliability and consistency of the results.

The graph in Figure 5 shows that the excavated soil sand “ESS-SP” has a similar particle size distribution to that of natural sand. However, the particle size distribution curve highlights variations among the sands derived from excavated soils.

The values of the fineness modulus for the various types of sand indicated that ESS-LB and ESS-SG sands suggest coarser sand, with values larger than 2.8, while the other sands exhibit values ranging between 1.5 and 2.8, indicating medium-fineness sand. The percentage of fines can have a significant impact on sand properties, such as water demand [22]. The excavation sand ESS-LT displays a fines content similar to that of natural sand, with a value of 1.31%, which is three times greater than that of natural sand. The excavation sand ESS-LB shows the highest fines content among various soil sources, with a value of 6.84%, which is approximately 484.62% higher than NS. According to the sand equivalent test (NF EN 933-8 standard [35]), all sands were deemed suitable for use in concrete, meeting the requirement of being greater than the lower limit of 60% as specified in the NF P 18-545 standard [42]. The methylene blue test (NF EN 933-9 standard [36]) is conducted to determine the activity and quantity of the clay fraction in the sand. Reactivity suggests that these fines are indeed clay. The results indicate that all sands, except for ESS-LB, comply with the NF P 18-545 standard [42] for use in concrete, which sets a limit for the methylene blue test of 1.5. ESS-LB exceeds this limit with a value of 1.77. The sand ESS-LT has the highest real density, which is relatively similar to NS, and the lowest water absorption compared to other excavated soil sands, approximately three times that of natural sand (2535 kg/m³ ± 3 kg/m³ and 1.31% ± 0.2%, respectively). On the other hand, the sand ESS-LB has the lowest real density and the highest water absorption, 2093 kg/m³ ± 98 kg/m³ and 10.9% ± 0.47%, respectively, reaching up to 17 times that of NS. The water-soluble sulfate content of the different sands is below the restrictive threshold of 0.2% (NF P 18-545 [42], October 2021). In France, the limit is set at 0.01% for the percentage of chlorides soluble in water in aggregates for cement-based concrete (NF P18-545 [42], October 2021). The chloride content values provided, expressed as a percentage, are significantly below the regulatory limit. Regarding the active alkali content, all sands exhibit a percentage below the 0.02% limit defined by Technical Specification FD P18-456 [43].

Table 4. Aggregates properties.

Property	NS	ESS-SP	ESS-LT	ESS-SG	ESS-LB
Real density in SSD condition (kg/m3)—NF EN 1097-6 [30]	2600	2106 ± 21	2535 ± 3	2348 ± 56	2093 ± 98
Packing density (kg/m3)—NF EN 1097-3 [44]	1650	1398 ± 11.13	1513 ± 9.6	1468 ± 14.7	1382 ± 18.5
Water absorption (%)—NF EN 1097-6 [30]	0.4	5.49 ± 0.53	1.31 ± 0.2	6.22 ± 0.58	10.9 ± 0.47
Fineness modulus—NF EN 933-1 [33]	2.73	2.74	2.27	3	3.32
Percentage of fines (%) ≤ 63 µm—NF EN 933-1 [33]	1.17	3.6	1.31	1.66	6.84
Sand equivalent value (%)—NF EN 933-8 [35]	80	72.64 ± 2.94	88.16 ± 2.03	82.12 ± 1.71	63.17 ± 1.54
Methylene blue test value—NF EN 933-9 [36]	0.4	1	0.52	1.12	1.77
Sulfates (%)—NF EN 1744-1 [37]	<0.036	0.048	0.08	0.052	0.096
Chlorides (%)—NF EN 1744-1 [37]	0.0011	0.0012	0.002	0.0007	0.0007
Active alkalis (% Na2O + 0.658x% K2O)—NF P 18-454 [38]	0.003	0.0067	0.0029	0.0047	0.0032

Table 4. Aggregates properties.

Property	NS	ESS-SP	ESS-LT	ESS-SG	ESS-LB
Real density in SSD condition (kg/m3)—NF EN 1097-6 [30]	2600	2106 ± 21	2535 ± 3	2348 ± 56	2093 ± 98
Packing density (kg/m3)—NF EN 1097-3 [44]	1650	1398 ± 11.13	1513 ± 9.6	1468 ± 14.7	1382 ± 18.5
Water absorption (%)—NF EN 1097-6 [30]	0.4	5.49 ± 0.53	1.31 ± 0.2	6.22 ± 0.58	10.9 ± 0.47
Fineness modulus—NF EN 933-1 [33]	2.73	2.74	2.27	3	3.32
Percentage of fines (%) ≤ 63 µm—NF EN 933-1 [33]	1.17	3.6	1.31	1.66	6.84
Sand equivalent value (%)—NF EN 933-8 [35]	80	72.64 ± 2.94	88.16 ± 2.03	82.12 ± 1.71	63.17 ± 1.54
Methylene blue test value—NF EN 933-9 [36]	0.4	1	0.52	1.12	1.77
Sulfates (%)—NF EN 1744-1 [37]	<0.036	0.048	0.08	0.052	0.096
Chlorides (%)—NF EN 1744-1 [37]	0.0011	0.0012	0.002	0.0007	0.0007
Active alkalis (% Na2O + 0.658x% K2O)—NF P 18-454 [38]	0.003	0.0067	0.0029	0.0047	0.0032

The percentage of fines, as well as their clay nature, measured by the methylene blue test, play a crucial role in water absorption (WA%). Sands with high clay fine content tend to absorb more water. To clearly demonstrate the impact of MB on WA%, we conducted a statistical analysis using the R language in R Studio software (version 4.3.2, 2023), employing the liner regression model function:

$$\mathrm{WA}=-2.57+7.72\times \mathrm{MB}$$

(1)

The model’s ability to predict the value is 99% (coefficient of determination ${\mathrm{R}}^2$, Equation (2)) and the correlation is shown in Figure 6. The standard deviation of the residuals, measured by the root mean squared error (RMSE), is 0.203 (Equation (3)). This aligns with the experimental standard deviation of water absorption measured on the same sand, which ranged from 0.2 to 0.5, confirming the validity of the model.

$${\mathrm{R}}^2={\displaystyle \frac{\sum {({\mathrm{y}}_{\mathrm{i}}-{\widehat{\mathrm{y}}}_{\mathrm{i}})}^2}{\sum {({\mathrm{y}}_{\mathrm{i}}-{\overline{\mathrm{y}}}_{\mathrm{i}})}^2}}$$

(2)

$$\mathrm{RMSE}=\sqrt{{\displaystyle \frac{1}{\mathrm{N}}}{\sum }_{\mathrm{i}=1}^{\mathrm{N}}{\left({\mathrm{y}}_{\mathrm{i}}-{\widehat{y}}_i\right)}^2}$$

(3)

where $\overline{y}$ and $\widehat{y}$ are, respectively, the predicted value and the mean value of the measured value y at the ith instant, and N represents the number of predictions.

3.2. Results of Fresh Mortar

The results of the spread of reference mortar manufactured with natural sand in different moisture states are represented in Figure 7. These results revealed that the mortar spread decreases when the sand is used in a dry state compared to saturated surface dry and natural moisture states. According to Meftah et al. [45], during mixing, dry sands (0% WA) absorb mixing water and the superplasticizer diluted in mixing water, which can influence workability. An increase in spread at T = 0 min is observed with increasing sand moisture content (saturated at 100% WA, oversaturated at 135% WA, and 335% WA, respectively).

The results of the fresh mortar spread tests, manufactured with 30% of excavated soil sands in different moisture states, are presented in Figure 8. Figure 8c shows the variation in the spread of mortar with sand in its natural storage state. It is relevant to note that mortars formulated with sands in the natural storage state significantly depend on the saturation state of the sand, whether it is under-saturated or over-saturated. However, a clear correlation is observed between the decrease in spread over time and the increase in the methylene blue value, indicating that higher clay content negatively affects the workability of the mortar. Sands with low values (0.4 and 0.51), as NS and ESS-LT, show a spread loss of about 10%. Sands with higher values (1 and 1.12), such as ESS-SP and ESS-SG present a loss of about 15%, while the sand ESS-LB with the highest value (1.77) shows a loss of about 20%. A study by Choudhary et al. [46] confirms this trend: higher clay fines content tends to decrease the workability of concrete.

The spread observations indicate that the SSD state of the sand is characterized by more significant spread, compared to the other states, particularly for sands with a high-water absorption rate (Figure 8b). Hung et al. [47] found that the use of recycled sand in the SSD state can be considered an effective solution to improve mortar consistency. Regarding the dry state (Figure 8a), it is observed that the spread tends to maintain its initial value at T = 0 min over time, with a percentage loss at T = 30 min ranging between 5 and 10%, whereas for other states, the loss reaches 20% at T = 30 min. Cyr et al. [29] found that the loss of settlement over time is less significant as the recycled aggregates are in a state of moisture far below saturation.

Statistical Study of Fresh State Results

Sand properties certainly influence the characteristics of fresh mortar, such as its spread (SF). Thus, according to the results, saturation state and total amount of water added during mixing have a significant impact. To deepen our understanding of the factors influencing mortar spread, we conducted a statistical analysis using the R programming language within R Studio. The linear regression was utilized to model mortar spread, considering several critical variables: time “T” (at 0, 15, and 30 min), specific surface area “SSA” (${\mathrm{m}}^2/\mathrm{g}$), intergranular porosity “IP” (%), methylene blue value “MBV”, sand saturation state “SSS” (0 for dry, 1 for saturated, <1 for undersaturated, >1 for oversaturated), and the quantity of water added during mixing. According to Benabed et al. [48], the addition of dune sand increases the flow time of self-compacting concrete (SCC) compared to alluvial sand, indicating higher viscosity. This effect is due to the increased specific surface area of the fine aggregates. Siyi Ju et al. [49] found in their study that aeolian sands, characterized by high intergranular porosity and a low fineness modulus, affect aggregate packing and require more water to fill the voids. This results in higher yield stress and plastic viscosity, and lower flowability of ultra-high-performance concrete. Several researchers have investigated the influence of methylene blue value on mortar and concrete performance. They mainly concentrated on workability and did find some correlations between MBV and workability [50,51,52].

Three linear regression models were developed for each moisture state (Figure 9). The first model, with an adjusted ${\mathrm{R}}^2$ of 89.49% and a residual standard error (RMSE) of 4.767 mm, demonstrates a strong fit and emphasizes the influence of various factors in the dry state. The second model exhibits a slightly lower Adjusted ${\mathrm{R}}^2$ of 84.57% and a higher RMSE of 8.62 mm, highlighting its relevance to the saturated surface dry state. The third model, which includes the saturation variable, achieves the highest adjusted ${\mathrm{R}}^2$ of 91.83% and an RMSE of 7.966 mm, indicating its effectiveness in capturing the spread behavior under natural storage conditions. Overall, these models collectively highlight the significant factors influencing mortar spread and their respective impacts across different storage states (

Table 5. Summary of residuals, minimum, and maximum values.

Model	Residuals (mm)	Min (mm)	Max (mm)
Dry state	−7.973 to 6.106	−7.973	6.106
Saturated surface dry state	−11.9786 to 13.355	−11.9786	13.355
Natural state of storage	−7.639 to 14.005	−7.639	14.005

).

A general model derived from this analysis (Figure 10), illustrating a correlation between these variables for the different moisture states combined (Equation (4)), was developed, achieving an ${\mathrm{R}}^2$ value of 84.5%. The analysis results showed highly significant p-values, with the overall model’s p-value (

Table 6. Overall model fit statistics of spread model.

Metric	Value
Multiple R2	0.8449
Residual standard error	9.799 mm
p-value	1.692 × 10−13

) well below the 0.05% threshold [53]. This indicates strong evidence against the null hypothesis, confirming that the model is statistically significant and effectively explains the variance in the dependent variable. The RMSE (root mean square error) value, indicating the mean standard deviation between the predicted and actual model values, is 9.799 mm, close to the maximum standard deviation of the experimental value of 7.2 mm, which validates the effectiveness of our model. When the blue methylene parameter is replaced by the water absorption, the analysis showed a coefficient of determination similar to the first model. These results validate the strong relationship between these two parameters as expressed in the linear regression model in Equation (1).

$$\mathrm{SF}\left(\mathrm{mm}\right)=124.4-1.3\times \mathrm{T}-7.9\times \mathrm{SSA}-4.9\times \mathrm{MBV}-8.2\times \mathrm{Added}\mathrm{water}+7.2\times \mathrm{IP}+12.8\times \mathrm{SSS}$$

(4)

3.3. Hardened State Mortar Results

3.3.1. Compressive Strength

The variation in the compressive strength of mortar with 30% excavated soil sand at different moisture states at 2, 7, 14 and 28 days is shown in Figure 11. The compressive strength gain is most significant between 2 and 14 days, with stabilization observed from 14 to 28 days, likely due to the effect of the hardening accelerator enhancing early hydration. The effect of the moisture state of sand on the compressive strength of mortar is most remarkable on the reference mortar with natural sand, with a strength of 37.57 ± 0.51, 41.33 ± 0.56, and 41.94 ± 0.61 MPa at two days, at dry state, natural state of storage, and saturated surface dry state, respectively. The lowest compressive strength is noted when natural sand is used in its dry state. This can be attributed to the absorption of the mixing water, combined with superplasticizer and setting accelerator, by the dry sand, notably, the correction for added water remains negligible due to the low water absorption of natural sand [45]. Consequently, this absorption reduces the available water for cement hydration and slows down the setting kinetics [54]. M. Nedeljkovic et al. [14] found in their study that the compressive strength of mortar made with natural sand in its dry state is slightly lower than that made with naturally stored sand. However, According to EN NF 206+A2 standard [18], a variation of 4 MPa (N/mm²) is accepted for a concrete referring to the same formulation.

The moisture state of the four excavated soil sands has a negligible effect on the compressive strength of the mortar at all tested intervals (2, 7, 14, and 28 days). Given the standard deviation of the compressive strength measurements, no significant differences were observed among the various moisture states. Le et al. [55] noticed that the initial saturation state of recycled sand has little influence on the mechanical properties if we work with the same ${\mathrm{W}}_{\mathrm{eff}}/\mathrm{C}$ ratio. As a variation of 4 MPa is accepted, consequently, a percentage variation of 9% is accepted at 2 days and 28 days for mortar with dry, SSD and NS sate sand.

All the excavated soil sand mortars showed a variation of less than 9% compared to the reference mortar, except for the sand from the “Le Boupère” site, ESS-LB, which showed the lowest strength among the samples studied, at early ages (e.g., 2 days), with a loss of 15.15% and 17.83%, compared with the reference mortar with sand at its natural and SSD state, respectively. The trend continues across all curing periods. The same trend is noticed at 28 days, with a loss of 15.87% and 19.95%. This could be attributed to the intrinsic characteristics of this sand, such as its clayey content (methylene blue value, 1.77), water absorption (10.98%), and low density (2093 kg/m³) compared to other sands.

3.3.2. Flexural Strength

The flexural strength of mortar is shown in Figure 12. For the reference mortar, the reductions in flexural strength are generally minor in the dry state compared to the natural and SSD state, with decreases up to 8% at 2 days. However, these differences tend to diminish over time, showing relative stability over the 28-day testing period. Mortars with excavated soil sands exhibit more substantial reductions in flexural strength in the dry state compared to both the natural and saturated states. For instance, ESS-LT experienced reductions ranging from 9.14% to 23.64% compared to its natural state and from −4.24% to −20.16% compared to its saturated state across the testing period. Similarly, ESS-SP, ESS-SG and ESS-LB mortars displayed reductions in flexural strength, when subjected to dry conditions of sands.

The increase in flexural strength with time is lower for the reference mortar with natural sand, regardless of the moisture state of the sand. The flexural strength of mortar increased with 30% of replacement of river sand with excavated soil sand, reaching 50% enhancement when 30% of ESS-SP was used at its natural and saturated state. According to Wu et al. [25], the flexural strength of excavated soil-based cement mortar showed that two different trends can be found in terms of the relationship between flexural strength and the percentage of replacement R; for an R of no more than 30%, this leads to an increase in the flexural strength in comparison with the reference mortar, for an R larger than 30%, the flexural strength decreases with an increase in R. This tendency is explained by the microstructures recorded by scanning electron microscope (SEM), the cement mortar made of river sand has a lot of C-S-H gel (i.e., the hydration products) and exhibits dense microstructures; a large amount of C-S-H gel can still be observed in the cement mortar with R = 30%, and the filling effect of excavated soil makes the microstructure of mortar denser, which leads to the decrease in porosity.

3.3.3. Statistical Study of Hardened State of Mortar

In this section, statistical analyses are also established to study the effect of sand properties and its moisture state on the compressive strength at different ages. The examination of the linear regression outcomes depicted in Figure 13 indicates that the model effectively accounts for the variations in compressive strength at 2, 7, 14, and 28 days, achieving a coefficient of determination (${\mathrm{R}}^2$) of 85%. The model’s root mean squared error (RMSE) was determined to be 3.914 MPa (

Table 7. Model summary statistics of compressive strength.

Metric	Value
Multiple R2	0.8523
Residual standard error	3.914
p-value	<2.2 × 10−16

), indicating the typical deviation between observed and predicted values.

The study takes into consideration the influence of different parameters on the model (Equation (5)). The fineness modulus “FM”, the sand equivalent value “ES”, the real density “${\rho }_{real}$” (kg/m³), the intergranular porosity, the quantity of water added during mixing, and the moisture state of sand showed an impact on the compressive strength of mortar mixes. This reveals the importance of fine content and its nature and property on the compressive strength. Choudhary et al. [46] studied the outcomes of effect of clay fines on strength of concrete with different sand with a percentage of clay fines of 0.25%, 0.5%, 1%, 1.5% and 2.0%. It was found that there is steady decrease in the compressive strength value, with a reduction of about 32% compared to reference concrete for mix with 2% clay fines [46]. The packing density also revealed an impact on the compressive strength, Xiao et al. [56] assumed that there exists approximately a linear relationship between the compressive strength and the mass density. Many studies showed that the moisture state of the aggregates can affect the strength of its concrete [47,57,58,59]. However, in this model, the coefficient of moisture state showed a high probability value (p-value) of 0.23. This means that the moisture state of the aggregates affects the hardened state of the mortar, but with low significance.

$${\mathrm{f}}^{\prime }\mathrm{c}\left(\mathrm{MPa}\right)=-304.7+0.8\times \mathrm{T}+0.1\times {\mathsf{\rho }}_{\mathrm{real}}+9.1\times \mathrm{FM}+3.6\times \mathrm{IP}-0.8\times \mathrm{SE}+73.6\times \mathrm{Added}\mathrm{water}+\mathrm{SSS}$$

(5)

4. Conclusions

In this study, excavated soil sand was evaluated as a partial replacement for natural river sand in mortars derived from self-compacting concrete, focusing on its physicochemical properties and their effects on the mortar’s rheological and mechanical performance. Based on the experimental results, the following conclusions were drawn:

• The physicochemical properties of excavated soil sand, including its particle size distribution, fineness modulus (2.27–3.32), fines content (up to 6.84%), and water absorption (up to 10.9%), exhibited significant variability. Higher fines and clay content correlate with increased water absorption, with a methylene blue value of 1.77 corresponding to the highest absorption rate.
• The rheological properties of mortar were significantly influenced by the moisture state of the sand. Mortars prepared with sand in the saturated surface dry (SSD) state exhibited better flowability compared to those with sand in the dry or natural storage state. The spreading behavior was strongly affected by the sand’s absorption capacity and fines content.
• Compressive strength tests revealed that mortars with 30% excavated soil sand replacement maintained stability across different moisture states. However, for ESS_LB sand, a 20% reduction in compressive strength was observed.
• Flexural strength results showed improvements of 20% to 50% when using excavated soil sand at a 30% replacement rate in the SSD state, demonstrating the potential for enhanced performance in structural applications.
• Statistical modeling revealed a strong relationship (R² = 85%) between the sand’s properties, particularly fineness modulus and methylene blue value, and the rheological and mechanical performance of the mortar. This predictive model provides valuable insights for optimizing mortar formulations.

These experimental results highlight the potential of excavated soil sand to replace natural river sand in mortar as a preliminary step towards its application in concrete, offering an environmentally friendly alternative that helps to conserve natural resources and minimize construction waste.

Another promising area for future work is the validation of these findings on a concrete scale. This will involve conducting SCC-specific tests, such as the L-box for passing ability and segregation resistance, to confirm the feasibility of using excavated soil sand in SCC formulations for industrial applications.