Introduction

Water plays a significant role in many human activities and industries, and to get the required quality, raw water must be treated in water treatment plants. Significant amounts of sludge are produced during the drinking water treatment processes, over 10,000 tons of sludge are produced every day worldwide (Liu et al. 2020a, b, c), which purify water via coagulation, sedimentation, and filtration processes (El-Didamony et al. 2014). Filter backwashes and sedimentation basins are the main sources of sludge. The quality and quantity of the generated sludge depend on the natural water quality and the dose of the coagulants used in the treatment process (Mahanna et al. 2015, 2018; Ahmad et al. 2016). The disposal of sludge produced from water treatment has become a major global environmental issue due to problems concerning land and the toxicity of such waste components (Nguyen et al. 2022). In many countries, the dewatered sludge is dumped into landfills, however, it poses a significant risk to the surrounding environment. Leachate can become contaminated with various chemicals and pollutants from the deposited sludge. This contaminated leachate can then seep into groundwater sources, potentially leading to long-term contamination and public health hazards. Moreover, the high cost associated with operating and maintaining landfills makes this method of waste disposal economically unsustainable in the long run. In addition, the disposal of WTS in waterways causes an increase in aluminum concentration, which has a negative impact on the environment and natural resources (Ahmad et al. 2017). So, there is a growing need for recycling WTS in numerous industrial and commercial manufacturing processes.

In addition to use the WTS in land applications (Ahmad et al. 2016), it has been investigated in construction materials manufacturing such as cement composites (Gomes et al. 2020), concrete blocks (Liu et al. 2020a, b), geopolymers (Luukkonen et al. 2019; Geraldo et al. 2017), sintering bricks (Heniegal et al. 2020; Benlalla et al. 2015; Amin et al. 2022), paving blocks (Liu et al. 2020a), tiles and eco-concrete blocks (Liu et al. 2020a, b, c), high alumina ceramics (Ferreira and Olhero 2002), glass ceramics (Toya et al. 2007), production of clinker (Husillos Rodríguez et al. 2011), a replacement for aggregate in concrete (Vembu et al. 2022; Feitosa et al. 2023; Kaish et al. 2021), and lightweight aggregates (Lee et al. 2021).

Using artificial or natural lightweight aggregates (LWA) to produce lightweight concrete (LWC) has many advantages. LWC has a higher strength/weight ratio and a lower density as compared to normal weight concrete, better fire resistance, lower coefficient of thermal conductivity, and enhanced durability properties (Helmy et al. 2023; Hussein et al. 2022; Bilodeau et al. 2004; Hwang and Hung 2005). Specific consideration has been given to the use of artificial LWAs manufactured either by the cold bond or sintering technique due to the decrease in natural resources. Various studies have proposed manufacturing LWA from WTS using the sintering method. Lee et al. reported that by adding 30% of the WTS and using sintering temperatures of 1150–1200 °C, LWAs were successfully manufactured with a density of 0.8–1.2 g/cm3. In addition, a concrete specimen with strength values of 20–45 MPa was achieved by using these aggregates (Lee et al. 2021). Huang et al. tried different samples of water treatment sludge to produce structural and nonstructural LWAs under a sintering temperature of 1150–1275 °C. For the resulting structural LWA and nonstructural LWA, the aggregates had particle densities of 1.35 and 0.98 g/cm3, and bulk densities of 726 and 518 kg/m3, respectively (Huang and Wang 2013). Sales et al. used lightweight coarse aggregate made of WTS and sawdust in manufacturing LWC. The produced LWC has a reduction in the thermal conductivity by 23% lower than that of conventional concrete (Sales et al. 2010). Since the sintering method consumes a large amount of energy and produces an enormous amount of pollutants, it is crucial from an economic and environmental standpoint to use a cold-bond manufacturing process to produce low-density lightweight aggregates (Tajra et al. 2018).

Cold-bonding technique is a cost-effective, and environmental process used to produce core–shell aggregates by encapsulating particles within a shell structure from different material sources (Ren et al. 2021). Additionally, the process of cold-bonded pelletizing is frequently used to recycle waste powders or solid waste into cold-bond lightweight aggregates (CBLAs) (Ferraro et al. 2021) or core–shell lightweight aggregates (CSLWAs) as green construction materials, which may reduce environmental loads and preserve natural resources (Tang and Brouwers 2018). CBLAs have a substantially higher density than sintered lightweight aggregates. Moreover, a variety of solid wastes are now included in the CBLAs raw materials such as rice husk ash, ground granulated blast furnace slag, coal fly ash (Tajra et al. 2019b), chromium-contaminated soil (Li et al. 2020), red mud, solid waste-based cementitious material (Zhang et al. 2023, 2021), and municipal woody biomass waste ash (Lin et al. 2023; Serpell and Zwicky 2022). Furthermore, the CSLWAs were manufactured by using multi-source solid wastes, including sewage sludge, waste glass powder, lithium slag, and contaminated soil (Gao et al. 2023). Furthermore, cement-fly ash with sodium alginate (Shang and Li 2020) and ultrafine fly ash with chromium-contaminated soil (Yang et al. 2021) were investigated to prepare CSLWAs. Moreover, many researchers have emphasized monitoring the properties of LWA through the implementation of core–shell structure design. Tajra et al. (2018) used the expanded perlite particles as a core to produce a low-density (510 kg/m3) structured LWA. Due to the low density of expanded perlite particles and the efficient cohesion with the cement matrix, it has been used as a core and as a powder in the shell to produce CSLWA applied in the manufacturing of LWC. The produced LWC have compressive strength and dry density ranging from 17.9 to 25.8 MPa and from 1115 to 1540 kg/m3, respectively (Tajra et al. 2019a, b). According to Peng et al. (2017), the water absorption of the CSLWA was reduced to 5.9% by coating with the white glue emulsion. Excavated soil and expanded perlite were used to develop high water absorption (26–55%) core–shell ceramsite, which can be applied in landscaping water-retaining materials (Dong et al. 2021).

This study aims to recycle solid waste sludge from water treatment plants into eco-friendly construction materials. This approach was presented by producing core–shell lightweight aggregates with low density and good mechanical properties from water treatment sludge and cement composite as a shell and expanded perlite as a core by the cold-bond method. Moreover, the effects of cement content on the aggregate properties were investigated. The process of producing LWA without sintering has decreased the consumption of energy and prevented the emission of harmful gas so it has a positive effect from an economic and environmental assessment. Furthermore, the production of lightweight concrete from investigated LWA is studied. The workability and mechanical properties of the produced LWC were accordingly examined.

Materials and methods

This study contains two phases of experimental work. The first phase was producing lightweight aggregates (LWA) by the core–shell method using expanded perlite, water treatment sludge, and cement. The second phase was producing lightweight concrete (LWC) using the produced LWA from the first phase.

Raw materials

The water treatment sludge (WTS) used in the production of the lightweight aggregates (LWA) was generated from a drinking water treatment plant in Meet Khames, Dakahlia, Egypt. The plant is a conventional system using alum as a coagulant. The WTS was taken from the sludge tank which receives sludge from settling tanks and filters backwash. Expanded perlite (EP) was obtained from The Egyptian Co. for Manufacturing Perlite. For LWC preparation, Ordinary Portland Cement (OPC) (CEM I 42.5 N), and superplasticizer (SP), were obtained from CMB company. The SP conforms to EN 934-2, ASTM C 494 type F, and ES 1899. The OPC conforms to EN 197-1 and ES 4756. The natural fine aggregate used for LWC preparation as filler material was quartz sand confirming to IS 383 with a specific gravity of 2.65. The collected sand was sieved through 2.0 mm and retained on a 0.075 mm sieve. The chemical composition of WTS powder was determined using energy-dispersive X-ray Spectroscopy (Oxford X-Max N 20, USA). Scanning electron microscopy (SEM) (JEOL; JSM-6510LV, Japan) using a microscope of field-emission electron with 20 kV accelerating voltage and magnification of 5× to 300,000× (printed as a 128 mm × 96 mm micrograph) was used to detect the microstructure of WTS powder and EP particles. To specify the acidity or alkalinity of sludge, pH was determined using a pH meter (AD8000 Professional Multi-Parameter).

Lightweight aggregates (LWA) production

In this study, the cold boning process was used to produce the LWA. The process involves the agglomeration of fine particles into pellets by the combined action of falling forces and capillary forces in a rotating pelletizer disk, where the proper amount of moisture is introduced to the raw materials. The cold-bonding method was used and preferred over the sintering method because it preserves the environment the opposite of the sintering method which consumes large amounts of energy and pollutes the environment. In addition, this method is characterized by high productivity in the shortest time depending on the capacity of the pelletizing disk device as well, resulting in round aggregates with high concrete workability. The schematic diagram of the LWA production is summarized in Fig. 1. The LWA production is conducted according to the method presented by (Tajra et al. 2018) with some modifications. Briefly, WTS was first collected from a sludge tank with high water content. The collected sludge was dewatered, and then sun dehydrated for four days. The dried sludge was grinded to about 7–10 mm particles to facilitate the drying process. The grinded particles were oven-dried at 110 °C for 6 h. After that, the dried samples were milled via a control jar ball mill and then sieved to get WTS powder smaller than 850 µm to enhance the pelletization process. The fabrication of LWAs was conducted by encapsulating EP particles used as a core within a composite of WTS and cement as a shell. To obtain a homogenous composite, the WTS and the cement were dry mixed for 3 min. The encapsulating process was done in a pelletizer disk with 40 rpm rotational speed and 40° angle. As shown in Table 1, five distinct cement to WTS ratios were utilized in the composite of shell; 10:90, 20:80, 30:70, 40:60, and 50:50. The production process for each ratio was conducted into two pelletization steps. The first step: the pelletizer disk was initially filled with 50 g of EP particles. Afterward, they were simultaneously sprayed with water and supplied with 500 g of composite. During the encapsulating process, the water was sprayed carefully with a content of 100 g in parallel with the supply of composite. The first pelletization step takes 15 min. The resulting pellets were left to dry for 24-h, then vacated using 4.75 mm sieve to get rid of the decimated particles. The second step: the resulting pellets were put into the pelletizer disk, then 150 g of water was sprayed with supplying 500 g of composite. This step also takes 15 min. The final pellets were carefully discharged from the pelletizer and transferred to a salver for 24-h. The LWAs were vacated using 4.75 mm sieve and selected for water curing. It is noted that as the cement percentage increases and the WTS percentage decreases, the quantity of granules produced increases, and the losses decrease. This happened as a result of the adhesive properties of cement, which are less in the sludge. Therefore, with an increase in the percentage of cement, adhesion is faster, and the losses decrease.

Fig. 1
figure 1

Schematic diagram of core–shell LWA production processes

Table 1 The proportions of the LWA mixture samples

2.3. Lightweight aggregates tests

The effect of the cement-sludge ratio in the mixture on the properties and manufacturing productivity of the aggregates produced was examined. Encapsulation efficiency, which is determined as the percentage between the pellets produced mass with a size ranging from 4.75 mm to 10 mm as shown in particle size distribution of the produced LWA (Fig. 2) and the total mass of the materials utilized in the manufacturing process, was used to evaluate productivity in this study. Additionally, the aggregate’s capability in terms of, loose bulk density, water absorption in 24-h, and maximum water absorption (after 28 days), were assessed. The oven-dry particle density was calculated by weighing a sample of 25 g of oven-dried aggregate (Wa), placing it in a graduated cylinder containing 70 ml of water (Vw), and reading the height in the volume of water immediately after adding the aggregate (Va+w). The loose bulk density was calculated according to ASTM C127 using Eq. (1).

Fig. 2
figure 2

Particle size distribution of the produced LWA

$$\mathrm{Loose\ bulk\ density }\left(\frac{{\text{gm}}}{{{\text{cm}}}^{3}}\right)=\frac{W{\text{a}}}{{V}_{{\text{a}}+{\text{w}}} - {V}_{{\text{w}}}}$$
(1)

The water absorption was calculated by reading the graduated cylinder after 24-h (Vw.24h) and 28 days (Vw.max) by adding the aggregates. The water absorption was calculated according to ASTM C127 using Eq. (2).

$$\mathrm{Water\ absorption }(\mathrm{\%})=\frac{V\mathrm{w }\gamma \mathrm{w }}{{\text{Wa}}}$$
(2)

The 28-d compressive strength of the LWAs was investigated by measuring the bulk crushing resistance, following EN 13055 Annex C. Before performing the strength test, the aggregate sample was first dried in an oven at 105 °C until a constant weight was achieved and then cooled down to room temperature. Next, the aggregate was compacted into a steel cylinder and loaded. Three samples of the same material were tested, and the mean value was taken into consideration. The crushing resistance was calculated using the following equation:

$${\text{Ca}} = \left( {L + F} \right)/A \, \left( {{\text{N}}/{\text{mm}}^{{2}} } \right)$$
(3)

where Ca is the bulk crushing resistance (MPa), L is the weight of the piston (N), F is the force recorded at 20 mm of compression for 100s (N), and A is the area of the piston (mm2).

To evaluate the extent of interaction between cement-sludge composite and perlite, scanning electron microscopy (SEM) was performed on a section of the aggregate. Moreover, the pH of LWAs was measured to evaluate the extent of the acidity or alkalinity effect of the aggregate when added to the concrete.

Lightweight concrete (LWC) production

In this study, two types of concrete mixtures were prepared: expanded perlite concrete (EPC) and lightweight aggregate concrete (LWAC). Based on the workability and consistency of the concrete, the ratio of materials was selected. The mass ratio of the EPC mixture is 1:2:3:1.1 of cement: sand: perlite: water. While, the LWAC mixture contains 450 kg cement, 600 kg sand, 600 kg LWA, and 200 kg water. In addition, the superplasticizer was used as 1.1% of cement weight. Each type of concrete was mixed using an Auto-Mortar mixer as shown in Fig. 3. The weight of the concrete produced must be less than 2000 kg/m3 to be considered lightweight concrete, according to BS EN 206/1-2013. The produced concrete was cured in water for 28 days for further testing.

Fig. 3
figure 3

Lightweight concrete production

Lightweight concrete (LWC) tests

During pouring, the slump, slump flow, and setting time of the concrete were determined according to GB/T50080-2002. Specimens of LWC with dimensions of 100 × 100 × 100 mm3 were obtained to conduct the 28-d compressive strength and dry density tests in accordance with EN 12390-3 and EN 12390-7, respectively. Furthermore, the splitting tensile strength of the concretes were measured using a 2000 kN concrete compression testing machine. The loading speed and size conversion factor were set at 0.5 MPa/s and 0.95, respectively, according to GB/T50081-2002. The average strength values of the three samples were calculated.

In addition, the thermal conductivity of 100 mm oven-dried cube samples was estimated via KD2 Pro using 3cm dual-needle sensors (SH-1) according to the standard specification ASTM D-5334. The KD2-PRO analyzer’s principle is based on gradually heating the needle while keeping track of its temperature. For KD2-PRO to produce more accurate results, the impact of ambient temperature on samples should be minimized. To reduce the impact of ambient temperature, the specimens’ surfaces were wrapped in plastic bags. The test sample was oven-dried at 45 °C for 48 h, the environmental conditions in the lab during the test were (temp. = 20 °C and R.H. = 55%). The apparatus accuracy, repeatability, and reproducibility are 1, 0.2, and 0.5%, respectively. The thermal resistance was calculated using Eq. (4).

$$Rt=\frac{L}{K}$$
(4)

where Rt (k/w) is the thermal resistance, L (m) is the plane thickness, and K (w/m K) is the thermal conductivity. Moreover, to assess the sound insulation of the produced concrete, a simplified procedure of calculation model was used to predict the weighted apparent sound reduction index. It concerns the weighting in accordance with EN 717-1. The weight sound reduction index Rw for concrete can be calculated by Eqs. (5) and (6) as a function of the mass per unit area.

$${R}_{{\text{w}}}=37.5{\text{lg}}\left({\text{m}}\right)-42$$
(5)
$$m=\mathrm{concrete\ densit }y\left(\frac{{\text{kg}}}{{{\text{m}}}^{3}}\right)\times \mathrm{plane\ thickness} ({\text{m}})$$
(6)

where Rw (dB) is the weighted sound reduction index, and m (kg/m2) is the surface mass. Furthermore, SEM was used to assess the extent of overlap between the aggregate particles 28 days of cured concrete. To assess the extent of aggregate particle distribution in the concrete, which influences the effectiveness of sound and thermal insulation, a section of the concrete was created and examined according to EN-12354.

Results and discussion

Raw materials characterization

Table 2 presents the chemical compositions of WTS. The chemical composition reveals that the main components of the WTS are SiO2, Al2O3, and Fe2O3. A higher concentration of SiO2 and Al2O3 in raw materials leads to an improvement in the strength of sintered products (Wang et al. 2009). SiO2 and Al2O3 are the primary components of WTS, as can be seen from the chemical composition. Therefore, adding cement to WTS can increase the strength of the pellets. The pH value of the sludge sample is 7.49, indicating that there is no acidic or alkaline activity of the sludge that would negatively affect the mixture over time.

Table 2 The chemical compositions of WTS

The SEM images of the surface morphology of WTS are shown in Fig. 4a, b. The images show the rough surface of the sludge granules, which promotes cohesion during the hydration process in the cement-sludge mixture used in the aggregate shell. The rough surfaces provide more contact points and interlocking opportunities, leading to improved mechanical interlocking and bond strength. The microstructure of perlite particles is shown in Fig. 4c, d. The SEM images show the honeycomb structure of the perlite particles which contain numerous voids. Furthermore, the fragility of the granule structure facilitates fragmentation during mixing with concrete components, reducing the size of existing voids.

Fig. 4
figure 4

SEM images of a, b WTS and c, d Perlite particles

Properties of produced LWA

To assess the properties of the produced LWA, the production efficiency, loose bulk density, pH, 24-h water absorption, 28-d water absorption, and 28-d bulk crushing strength as a function of cement ratio are presented in Fig. 5. Figure 5a shows the efficiency as the percentage of the mass of pellets produced with a size of more than 4.75 mm to the total mass of materials used in the manufacturing process, demonstrating that higher cement percentage result in fewer material losses. The production efficiency increased from 47.56% for 10% cement ratio to 97.3% for 50% cement ratio. This enhancement can be directly attributed to the role of cement as a binder, as increasing the cement ratio helps effectively hold the WTS particles together due to its stickiness (Tajra et al. 2018). Figure 5b shows the loose bulk density of LWAs. The aggregate density markedly dropped from 1.25 to 1.04 g/cm3 by decreasing the cement ratio from 50 to 10%. This drop in particle density is generally attributed to the low density of perlite, which occupies a large volume of the aggregate. While an increase in cement content increased bulk density. This is because the hydration products of cement fill the voids in the aggregate, resulting in fewer voids in aggregates with a higher cement content and a higher bulk density (Lin et al. 2023).

Fig. 5
figure 5

Lightweight aggregate properties a efficiency, b loose bulk density, c 24-h water absorption, d 28-d water absorption, and f bulk crushing strength

Figure 5c, d shows the 24-h and 28-d water absorption of LWA, respectively. The results indicate that the WTS–cement ratio affects water absorption. With an increase in the WTS ratio, water absorption increases. The 24-h water absorption of LWA50 (50% cement and 50% WTS) was equal to 12.4%, while the 24-h water absorption of LWA10 (10% cement and 90% WTS) was equal to 22%, as shown in Fig. 5c. This is attributed to the high porosity of WTS particles, which affects the water absorption of manufactured aggregate (Lee et al. 2021). In addition, the honeycomb-porous perlite significantly increases the water absorption capacity of LWAs, but with an increase in the cement content, the pores in the casing close, preventing water absorption through the perlite. Furthermore, comparing Fig. 5c, d, it can be observed that the 24-h water absorption is about 81.4% and 62.0% of the 28-d water absorption for LWA10 and LWA50, respectively. The 24-h water absorption serves as an indicator of the maximum water absorption (28-d). This can be used to specify the total amount of water required during concrete pouring by adding the amount of water absorbed by the aggregate to the amount of water required for the hydration process. Moreover, the pH value of the manufactured aggregate decreased from 11.73 to 10.16 with a decrease in cement content from 50 to 10% (Fig. 5e) due to the increase in WTS content, which has a neutral pH. Therefore, the application of WTS has a beneficial effect on the alkalinity of the aggregates.

Figure 5f represents the 28-d bulk crushing strength of all aggregate samples. As noted, the crushing resistance is primarily affected by the cement content. Increasing the cement ratio from 10 to 50% strongly improved the crushing strength, from 0.45 to 3.1 MPa. This improvement can be attributed to the actual cement hydration process, which results in the creation of calcium silicate hydrate (C–S–H), which increases with an increase in cement content. Besides, the hydration of cement produces calcium hydroxide (CH). Therefore, increasing the cement content also increases the amount of CH released, contributing to the improvement of the shell layer strength and enhancing the bulk crushing resistance of the LWAs (Tajra et al. 2018). Overall, compared to the previous studies with similar values of bulk density as shown in Table 3, it can be said that the produced LWA’s properties are acceptable

Table 3 Comparison of aggregate properties with previous studies

Microstructure of produced LWA

The scanning electron microscope (SEM) can be used to investigate the morphology and microstructure of the LWAs, assessing the porosity, size, and shape of the particles, as shown in Fig. 6. The LWA can be divided into three distinct parts: a rigid shell structure, a transition zone between the core and the shell structure, and the honeycomb-shaped expanded perlite core. The microstructure of LWA50, as shown in Fig. 6a, b, reveals the interference incident in the LWA50 particles between the cement–WTS composite and perlite, while Fig. 6c, d represents the interference incident in the LWA10 particles between the cement–WTS composite and perlite.

Fig. 6
figure 6

SEM images of ab LWA50, cd LWA10 and external shape and cross-section of ef LWA50, gh LWA10

Figure 6e, f, as well as g, h, shows the external shape and cross-section of LWA50 and LWA10, respectively. The strength of the interference manifests more clearly in LWA50 than in LWA10, as a result of increasing the cement ratio, which enhances the cohesion between the perlite core and the WTS–cement composite shell, thereby increasing the strength of the shell composite. Figure 6f, h shows the more regular shape of LWA50, as well as the regularity of the shell, compared to LWA10. Furthermore, larger pores and cracks can be noticed in the shell of LWA10 compared to LWA50. This is due to the hydration of the reacted cement, but some soil particles of WTS seemed not to be taken apart in the hydration by increasing the WTS particles content (Dong et al. 2021).

Properties of produced LWC

The properties of LWC made with produced LWA compared with LWC made with expanded perlite were evaluated. The workability of the produced concrete is listed in Table 4. EPC exhibited better slump and slump flow compared to LWAC. Furthermore, the setting time of the concrete produced using perlite was shorter than that produced using LWA. This can be attributed to the fact that the spherical LWAs flow better than the irregular perlite.

Table 4 The slump, slump flow, and setting time of the produced concrete

The dry density, 28-day compressive strength, strength/density ratio, and splitting tensile strength were estimated for each type of concrete. The dry density of LWAs concrete (LWAC) and expanded perlite concrete (EPC) was equal to 1840 kg/m3 and 1600 kg/m3, respectively, while the compressive strength after 28 days was 24 MPa for LWAC and 11 MPa for EPC. The main reason for the strength reduction of EPC is the lower strength of perlite compared to LWA, which can be attributed to variations in the pore structures of these aggregates. This results in EPC containing more voids, as confirmed by its lower density. The strength/density ratio of LWAC and EPC was 13 and 6.87%, respectively. The significant improvement in the strength-to-density ratio of LWAC meets the requirements of the standard for structural LWC according to EN 206-1. LWAC can also be categorized as a structural lightweight aggregate concrete based on its compressive strength, which meets the ACI 213. The high strength-to-density ratio allows LWAC to offer lightweight construction, improved load-bearing capacity, energy efficiency, thermal insulation, and ease of handling. These properties make LWAC a suitable choice for a wide range of applications. Furthermore, the splitting tensile strength of LWAC and EPC was 2.98 and 1.37 MPa, respectively, which can be attributed to the cohesion between the LWA and the cement mortar.

In this study, as a result of replacing of aggregates with LWAs, LWAs make up the majority of the concrete’s volume, which directly and significantly affects the material’s thermal characteristics. The thermal conductivity of LWAC and EPC is 0.72 and 0.64 W/m K, respectively. Compared to EPC, the thermal conductivity of LWAC increased by 12.5%. This increase can be attributed to the low density and highly porous structure of EPC compared to LWAC. The concrete thermal conductivity has been reported to range from 0.2 to 3.3 W/m K (Real et al. 2016; Yun et al. 2013). Up to 80% of the thermal conductivity of concrete can be decreased by replacing fly ash (Wu et al. 2015). According to Huang et al. (2013), concrete with fly ash replacement exhibits a 21% lower thermal conductivity than normal concrete. Additionally, the results demonstrate that the thermal conductivity of concrete created with LWAs was 78.2% lower than the highest thermal conductivity of concrete reported in previous studies, which was 3.3 W/m.k. The low thermal conductivity can be attributed to the implementation of WTS in the shell structure, which improved its thermal properties, and the low thermal conductivity of EP, which is used as a core structure. Furthermore, the findings of LWAC and EPC confirm that there is a relation between concrete’s thermal conductivity and its compressive strength. High-compressive strength concrete has high thermal conductivity due to a lack of pores.

In this study, the influence of using the LWA in concrete on the sound reduction index of a wall made of produced LWAC was investigated according to EN 12354. Assuming the thickness of the wall is 0.1 m, the sound reduction index of LWAC is 43 dB while EPC is equal to 40.65 dB. The sound reduction index increases by 5.7% in LWAC compared to EPC. Normally, to construct a sound insulation wall, the wall object is made of two cast in situ lightweight concrete layers that are separated by a thin layer of nonwoven fabric. This prevents the two types of concrete from mixing during the casting process, with the first layer providing structural properties and the second layer providing thermal insulation properties. To keep the nonwoven fabric in place when the concrete is in its liquid and thixotropic rheological phase and under high hydrostatic pressure, a steel grid is placed into the inner layer of the concrete. So, it is possible to replace all components of this wall by the wall with a single layer of LWAC, achieving both the structural purpose and thermal insulation at the same time. Overall, compared to the previous studies as shown in Table 5, it can be concluded that the produced LWAC properties are acceptable.

Table 5 Comparison of LWAC properties with previous studies

Microstructure of produced LWC

The crushed specimens of LWAC and EPC after exposure to the compressive strength test are shown in Fig. 7. The shape retention of LWAs in LWAC plays a critical role in enhancing the compressive strength of the concrete. It is clear that the LWA retained its original shape better than the EP, without any obvious damage. This indicates that the aggregates had sufficient strength-to-resist potential crushing caused by the concrete mixing. In addition, Fig. 7a, b shows that both EP and LWA were fractured along their diameter, with the clear appearance of the perlite in EPC and perlite cores in LWAC. As the LWAs are the weakest part of the LWAC, the concrete failure occurred there (Ke et al. 2009) rather than in the interfacial transition zone (ITZ) between the cement matrix and the LWAs. This demonstrates that the LWAC strength with a high cement content is primarily influenced by the LWA strength and failure of LWAC with a high cement content above 350 kg/m3 typically occurs in the LWA (Joseph and Ramamurthy 2009).

Fig. 7
figure 7

The distribution of aggregates in concrete (a) EPC, (b) LWAC

SEM was also conducted on the concrete specimens to show the microstructure of ITZ between the LWAs and EP, as well as the cement matrix, as shown in Fig. 8a, b and c, d, respectively. It is noted that the pores are concentrated only in the perlite core of LWA, while the spread of these pores increases throughout the EPC, which explains the lower compressive strength and higher intensity of EPC compared to LWAC.

Fig. 8
figure 8

SEM images of a and b LWAC, c and d EPC

Figure 8a, b shows a strong and continuous bond between the cement matrix and the LWA. The boundaries of both LWA and EP can be easily distinguished in Fig. 8a, b and c, d, respectively, due to their distinct microstructures, which contrast with the cement matrix. Furthermore, it is clear that the ITZ around LWA is denser than the ITZ encircling EP. This could be a result of the LWA shell internal curing from water storage (Tajra et al. 2019a, b).

Conclusions

In this study, a novel eco-friendly core–shell lightweight aggregate was performed by cold-bond technique from expanded perlite as a core, and water treatment sludge (WTS) and cement as the coating material. Based on the experimental results, the following conclusions can be obtained:

  • WTS content from 50 to 90% was beneficial to produce lightweight aggregates (LWA) with high efficiency.

  • The produced LWA had good single-particle crushing strength (3.1 MPa), bulk density (1.05 to 1.25 g/cm3), 24-h water absorption (12.4–22%), 28-day water absorption (22–27%), and pH values (11.73–10.16).

  • The produced LWA and expanded perlite (EP) were used to prepare lightweight aggregate concrete (LWAC) and expanded perlite concrete (EPC). The LWAC and EPC had densities of 1840 and 1600 kg/m3, compressive strengths of 24 and 11 MPa, and splitting tensile strengths of 2.98 and 1.37 MPa, respectively.

  • The results revealed a reduction in thermal conductivity of LWAC and EPC to be 0.72 and 0.64 W/m K, respectively.

  • The produced LWA concrete can be categorized as a structural and sound insulation concrete, with a high thermal conductivity.

  • These results can provide an environmentally friendly approach for converting WTS into high value-added LWA, with potential application in green lightweight concrete.