Performance-Based Evaluation of Supplementary Cementitious Material Synthesized with Basic Oxygen Furnace Slag and Ground Granulated Blast Furnace Slag

Abstract

Basic oxygen furnace slag (BOFS) is one of the major by-products of the steelmaking industry. Its limited utilization as a construction material is primarily attributed to its chemical properties, which hinder its stability and hydraulic activity due to its high free lime (f-CaO) content. This paper explores the performance of supplementary cementitious material (SCM) synthesized with ground granulated blast furnace slag (GGBFS), freshly produced BOFS (f-BOFS), and stockpiled BOFS (s-BOFS). A total of 10 mixtures with ordinary Portland cement (OPC) replacement percentages were assessed, maintaining a total replacement of 50% OPC, incorporating 15%, 25%, and 35% of each material by weight. The laboratory experimental program encompassed material characterization, fresh and hardened properties, pozzolanic activity, and durability assessment, with comparative studies conducted for each evaluation item. Test results indicate that f- or s-BOFS, when used with GGBFS, can be a viable alternative SCM with the potential for hydraulic activities and pozzolanic reaction. The newly synthesized SCMs demonstrated improved strength development in mortar mixtures. The mixture containing [15% f-BOFS + 35% GGBFS] achieved a 28-day compressive strength of 20.6 MPa, while the [25% BOFS + 25% GGBFS] blend reached a compressive strength of 19.7 MPa. These mixtures meet Grade 80 criteria as per ASTM C989/C989M Standard Specification for Slag Cement for Use in Concrete and Mortars. A performance-based ranking system was developed by integrating results from flowability, air content, strength activity index, drying shrinkage, alkali–silica reaction, and sulfate attack. The novelty of this work lies in assessing BOFS–GGBFS blends as SCMs using this multi-criteria approach to identify the most sustainable and technically viable mixtures. Moreover, the study highlights the influence of storage-induced weathering by directly comparing the reactivity and performance of f- and s-BOFSs in ternary blends, providing new insights into optimizing the utilization of slag. Notably, regardless of f- and s-BOFSs, proportions of [15% BOFS + 35% GGBFS] demonstrated superior strength development and achieved an excellent overall ranking. These findings confirm the potential of such slag blends as suitable SCMs for mortar and concrete applications, thereby advancing the sustainability and efficiency of cementitious materials.

1. Introduction

Ordinary Portland cement (OPC) production, a cornerstone of modern construction, is entangled with severe environmental concerns, prominently characterized by significant greenhouse gas emissions and substantial energy consumption. Cement manufacturing alone consumes approximately 5% of the world’s energy resources and accounts for 10% of anthropogenic CO₂ emissions [1,2]. Moreover, the annual production of over 30 billion tons of cement necessitates an extensive use of natural resources [3]. Additionally, OPC is the primary constituent of concrete, which is the most widely used construction material and the second most commonly used substance in the world, after water. In fact, for every person living on the planet, around 3–4 tons of concrete are produced yearly [4]. Exploring alternatives and ways to reduce the environmental impact of concrete, particularly OPC, is imperative to address this sustainability challenge.

Simultaneously, while integral to global development, the steel manufacturing industry poses ecological challenges. Accounting for 1.9% of global electricity consumption and contributing 8% of greenhouse emissions, the steel industry generates substantial waste, primarily in the form of steel slags [5]. The amount of slag accounts for about 15–20% of the total steel produced, depending on the steelmaking route and slag generation rate. This results in a global annual output of approximately 400 million tons of steel slags, which often end up in landfills, posing a threat to environmental sustainability [6,7]. Improper handling of these by-products without proper isolation can lead to detrimental ecological and public health impacts, including soil alkalization (leading to chlorosis, stunted growth, or even plant death), leaching of heavy metals, and contamination of surface and groundwater sources. Hence, there is a critical need for effective management and reuse strategies to mitigate the adverse effects of steel industry waste. Utilizing industrial by-products reduces waste and energy consumption, contributing to cost savings in production processes [8]. Of particular interest is the potential of steel and iron industry by-products to partially replace OPC in concrete, thereby decreasing cement consumption and increasing the utilization of industrial by-products. Following extraction and metal melting, slag emerges as a by-product with diverse compositional and structural properties, making it suitable for various construction and civil engineering applications.

In this context, integrating recovered materials into concrete production as supplementary cementitious materials (SCMs) emerges as a sustainable approach. SCMs have lower hydration reactivity than OPC, but their overall mass in concrete or mortar mixture might account for 20% to 70%, depending on their application [8,9]. Their utilization improves the overall performance of concrete and mortar. For example, GGBFS, one of the commonly used SCMs, possesses exceptional cementing properties and contributes to the setting and hardening of concrete when mixed with calcium-based materials, such as cement or lime, in the presence of water. Its benefits in concrete are widely established, lowering hydration heat, promoting long-term strength development, and increasing durability [10].

However, unlike GGBFS, the utilization of the other by-products from steelmaking plants, such as basic oxygen furnace slag (BOFS) and ladle slag (LS), is limited as SCMs because they have little or no binding properties despite containing reactive silica and alumina (coming from fluxes, such as bauxite or alumina-rich clays, added during the steelmaking and/or from the raw feed, including iron ore and coke ash). Moreover, BOFS, in particular, is readily available in regions with active steel production, such as Kazakhstan, where it remains an underutilized resource due to its high free calcium oxide (f-CaO) content. The properties of BOFS as SCMs depend on the basic oxygen furnace (BOF) process and the cooling method used [6]. Essentially, the BOF process is crucial in the steelmaking process, as it reduces the carbon content of the molten alloy and converts it into low-carbon steel. This is important because defects such as inclusions and blowholes may occur in steel due to higher carbon concentrations in the alloy [11]. Indeed, the converter oxidizes the impurities in the molten slag, thereby generating internal heat through an exothermic reaction that forms iron oxide and carbon monoxide. Furthermore, slag, which contains impurities from the molten metal, is generated by adding lime and fluorspar to the furnace. Finally, after the molten metal has been refined and its carbon concentration is acceptable, the steel is continuously poured from the furnace into a ladle and then into casting machines [12].

Moreover, the physical characteristics of slag are determined by the slag’s chemistry and the cooling process used. For instance, quenching produces vitreous/glassy slag, while slow cooling produces crystalline slag. The slag structure and its crystal sizes are affected by the type and speed of the cooling mechanism. For instance, it is not uncommon for the slag produced by slow cooling to have particles smaller than 100 mm, while extremely slow cooling yields particles smaller than 300 mm [6]. It is possible to utilize water-cooled slag in various applications, yielding porous and low-density products that can be used in specific applications based on their physical properties. Whereas, if cooling is processed under atmospheric circumstances, the slag tends to be hard and thick, which makes it ideal for use in concrete or other construction-related applications [13].

Because of steelmaking processes (BOF process turning pig iron to steel) and the cooling process, BOFS has poor cementing properties owing to low C₃S and β-C₂S content and unfavorable expansion properties due to free lime (f-CaO) and free magnesia (MgO) [14]. The poor hydraulic properties of BOFS also contribute to its high iron oxide content, which in turn forms one of the significant phases of slag, wustite (FeO), that does not react with water [15]. Several studies have been conducted to evaluate the physical, chemical, and mechanical properties of BOFS. Indeed, the composition and characteristics of BOFS vary significantly at different sites and the same site over various periods. The characteristics of the ongoing chemical reactions and mechanisms also differ based on the nature of the ore sources and the duration of the stockpiling period [16].

The mechanical characteristics of slag play an important role when it is used as a partial replacement for binders in cement. The chemical composition of BOFS is a good indicator of its mechanical properties; for example, the greater the basicity or alkalinity concentrations, the better the hydraulic characteristics. To qualify as a cementitious material, the ratio CaO/(SiO₂ + P₂O₅) must be more than 1.8 [12]. When Liu and Li [17] studied the mechanical characteristics of concrete containing BOFS, they discovered that replacing cement with 20% or more BOFS significantly decreased the concrete’s early-age strengths. In addition, Wang et al. [18] and Liu et al. [19] investigated the effects of steel slag on the mechanical characteristics of concrete. They stated that higher concentrations of steel slag led to detrimental impacts at later ages. However, Altun and Yilmaz [20] showed that the mixture with 30% BOFS had compressive and bending strengths comparable to those of OPC (100%) after 2, 7, and 28 days. They observed that strength can be attributed to several activation-related factors, including the sufficient fineness of the BOFS (specific surface area of 4000–4700 cm²/g), an alkaline and calcium-rich composition (high free CaO and MgO), and partial weathering of the slag. Cement mixtures may benefit from slag with a low MgO content, which was proven to slow the cement’s hydration and extend the mixture’s setting time. Moreover, Belhadj et al. [21] stated that the compressive strengths of mortar with BOFS cured at room temperature increased gradually, reaching 37 MPa and 50 MPa at 90 and 180 days, respectively. Their findings align with the XRD examination of hydrated phases, which showed that the C₂S phase remained consumed even after 90 days, increasing strength.

Research efforts have been directed toward understanding the physical and chemical properties of BOFS, acknowledging its variability across sites and over time, which is influenced by practices such as weathering and stockpiling that facilitate the hydration of CaO and MgO in steel slag. For instance, Mahieux et al. [22] state that weathering can stabilize the volumetric instability problems of free lime and magnesia in BOFS. Moreover, after slag is hydrated, calcium oxide (CaO) is transformed into hydrated lime (Ca(OH)₂), which readily reacts with the ambient CO₂ and produces calcite (CaCO₃). Thus, Belhadj et al. [21] claim these reactions decrease the CaO content over time. However, by comparing freshly produced BOFS (f-BOFS) and stockpiled BOFS (s-BOFS) samples through a series of laboratory tests, Yildirim and Prezzi [23] declared that the weathering of steel slag does not necessarily guarantee its volumetric stability. In fact, weathered slag may include only partially reacted f-CaO due to its finer granulometric distribution when compared with the BOFS particles. Thus, some lime particles may not have been accessible to water and remained unreacted [21]. Shi [24] found that specific materials for consuming and removing f-CaO must be used in conjunction with BOFS.

In comparison, Tsai et al. [25] investigated ternary mixtures of BOFS and GGBFS, analyzing the blended effect on their mechanical and cementitious properties. They stated that GGBFS stabilizes concrete volume because the Ca(OH)₂ precipitated from BOFS reacts with GGBFS in pozzolanic reactions. Moreover, Lu et al. [26] suggested the potential substitution of cementitious material in mortar mixtures by BOFS, further underscoring the need for innovative approaches to maximize the utility of this by-product in sustainable construction practices. Kabay et al. [27] also investigated the mechanical and durability properties of cementitious materials incorporating GGBFS and BOFS. The study highlighted the individual material’s pozzolanic behavior and activation potential when used as SCM. Meanwhile, the current research uniquely investigates their combined application in ternary systems based on systematically evaluating their physical, chemical, and durability properties. This approach provides a novel understanding of optimizing SCM combinations to achieve tailored performance characteristics in cementitious systems. Furthermore, unlike prior works, this study integrates a detailed assessment of both f-BOFS and s-BOFS, highlighting their potential viability despite the weathering effects.

2. Research Objective and Scope

This paper explores and evaluates the potential of integrating BOFS with GGBFS as a novel and sustainable approach to developing alternatives to conventional SCMs, with a focus on transforming by-products into valuable materials for concrete applications. Through this research, it is aimed to offer insights into the synergy between BOFS and GGBFS, enhancing their compatibility and performance in cementitious systems. As illustrated in Figure 1, the assessment of newly synthesized SCMs consists of 4 steps: (i) basic material characterization, (ii) investigation of fresh properties in mortar mixtures, (iii) evaluation of the pozzolanic activity of SCMs, and (iv) durability assessment.

Subsequently, after obtaining the results from all experimental programs, comprehensive comparative analyses of data were conducted. The relationships between testing parameters, such as flowability, air content, pozzolanic activity, expansion due to alkali-silica reaction (ASR), sulfate attack, and drying shrinkage, were examined by ranking them according to the best performance. All the required relationships that may account for the pozzolanic activity of BOFS and the difference between fresh and stockpiled BOFS, as well as its synergistic effect with GGBFS, were integrated into this performance-based ranking system by analyzing all mortar mixtures (OPC mixture, binary mixtures containing OPC, and the three materials individually, as well as ternary mixtures containing OPC and GGBFS with incorporated fresh or stockpiled BOFS).

3. Experimental Program

3.1. Material Preparation

This study employed four different materials as binders: OPC (PC500 D0—EASC 310.4-81), GGBFS, ground f-BOFS, and ground s-BOFS (aged for up to 1 year). As illustrated in Figure 2, material processing was conducted to obtain BOFS fines: initially, large-sized BOFS particles (50–100 mm) were crushed using a jaw crusher (BB 250XL, Retsch GmbH, Haan, Germany), reducing the size to <10 mm. The material was then ground using a ball mill (TM 300XL, Retsch GmbH, Haan, Germany) operated at a speed of 350 rpm for 30–45 min intervals. Finally, after each grinding period, the particles were sieved into different sizes, and the powdered forms of BOFS, which passed through the #200 sieve (75 μm), were used as SCMs. Furthermore, the specific gravity (SG) of cementitious materials was determined according to ASTM C188 [36], which uses a volumetric (Le Chatelier flask, Global Gilson Inc., USA) procedure with kerosene. The determined SG of OPC, GGBFS, fresh BOFS, and stockpiled BOFS is 3.14, 2.80, 2.90, and 3.08, respectively.

Additionally, graded sand was used as the fine aggregate to prepare all mortar mixture specimens. Before designing the mixture proportions, the aggregates were analyzed for fundamental physical parameters, including absorption capacity (AC) and specific gravity based on saturated surface dry conditions (SG_SSD), according to ASTM C128 [37], which employs a gravimetric (pycnometer) procedure. Consequently, the SG_SSD and AC of the aggregates were determined to be 2.59 and 0.56%, respectively. Then, based on the aggregate and SCM parameters, a sand-to-cement ratio of approximately 2.63:1 by weight was obtained. Moreover, the gradation chart of graded sand met the sand requirements stated in ASTM C778 [38].

3.2. Mixture Design

The main variables in the mixture design of mortars were the changes in the proportion of SCM used as a 50% replacement for cement. This choice was guided by the dual objectives of maximizing the incorporation of SCMs while maintaining acceptable mechanical and durability properties in the resulting mortar. The 50% replacement for cement was selected based on ASTM C989/C989M Standard Specification for Slag Cement for Use in Concrete and Mortars [39]. Then, as presented in

Table 1. Mixture proportions (by weight percent) of OPM and 50% cement replacement mixtures.

Mix No.	Mixtures	Mix No.	Mixtures
1	100% OPC	6	50% OPC + [35% f-BOFS + 15% GGBFS]
2	50% OPC + [50% GGBFS]	7	50% OPC + [50% s-BOFS]
3	50% OPC + [50% f-BOFS]	8	50% OPC + [15% s-BOFS + 35% GGBFS]
4	50% OPC + [15% f-BOFS + 35% GGBFS]	9	50% OPC + [25% s-BOFS + 25% GGBFS]
5	50% OPC + [25% f-BOFS + 25% GGBFS]	10	50% OPC + [35% s-BOFS + 15% GGBFS]

, 10 mortar mixtures with a water-to-cementitious material ratio (w/cm) of 0.47 were created to conduct all tests. They include an OPC mixture as a reference, three binary mixtures consisting of 50% cement replacements by GGBFS, f-BOFS, and s-BOFS, and six ternary mixtures developed with blended BOFS and GGBFS with 15%, 25%, and 35% by weight percentages of each material. Previous studies [2,3] regarding the reactivity and compatibility of these SCMs guided the selection of these specific blending ratios.

3.3. Test Methods

As shown in Figure 1, material characterization is crucial for evaluating the performance of individual materials, encompassing their chemical, physical, and mineralogical properties. The chemical composition of the cementitious materials, in terms of oxides, was determined using the XRF test with the xrFuse 6 electric fusion machine. Their crystalline phases were identified through XRD analysis, employing Cu Kα radiation on the Rigaku SmartLab System. The XRD test was conducted with a scanning range (2θ) of 5–70°, in continuous scan mode, and a sampling interval of 0.03°, thereby highlighting the primary mineral components of the samples. Additionally, the PSD of three cementitious materials was assessed using a laser scattering analyzer (Mastersizer 3000, Malvern Panalytical Ltd., Worcestershire, UK).

Moreover, the performance of the designed mixtures was evaluated through a series of tests that align with the C989, ASTM C618 Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete [40], and ASTM C311 Standard Test Methods for Sampling and Testing Coal Ash or Natural Pozzolans for Use in Concrete [41]. These specifications provide the guidelines for evaluating any SCMs.

The fresh properties of the mortars, including flowability (ASTM C1437 [28]), air content (ASTM C185 [29]), setting time (ASTM C191 [30]), and water requirement (ASTM C1403 [31]), were measured to assess their workability and behavior in the fresh state. The pozzolanic activity was evaluated through the strength activity index (SAI) based on compressive strength (ASTM C109 [32]), as shown in Figure 3a, as well as Chapelle test and thermo-gravimetric analysis (TG-DSC), while durability properties tests such as drying shrinkage (ASTM C596 [33]), sulfate attack (ASTM C1012 [34]), and ASR expansion (ASTM C1260 [35]) were evaluate by the length change in mortar bars, illustrated in Figure 3b.

4. Test Results and Discussion

4.1. Material Characterization

4.1.1. Chemical Composition Analysis

The chemical compositions of the SCMs used in this study are presented in

Table 2. Chemical compositions of BOFS, GGBFS, and OPC.

Oxide Content (%)	CaO	Fe2O3	SiO2	MgO	MnO	P2O5	Al2O3	TiO2	SO3	Others
1 f-BOFS	41.64 1	23.03	14.05	8.99	3.89	3.61	2.21	0.82	0.21	1.54
2 s-BOFS	43.93 2	24.67	13.38	6.16	2.80	3.91	2.93	0.85	0.24	1.12
GGBFS	40.51	0.34	32.25	10.46	0.31	-	11.40	1.42	2.00	1.31
OPC	69.19	2.96	17.42	2.05	0.12	0.09	4.36	0.20	2.61	0.7

Note: free lime content: ¹ f-BOFS (5.67%) and ² s-BOFS (0.21%).

. f- and s-BOFS mainly consist of 41.64% and 43.93% CaO, 23.03% and 24.67% Fe₂O₃, 14.05% and 13.38% SiO₂, and 8.99% and 6.16% MgO, respectively. Both BOFS have a higher percentage of Fe₂O₃ than GGBFS and OPC. It is clear because BOFS is a by-product of steelmaking, created by melting iron ore with lime or dolomite under gaseous oxygen-blowing cycles [42]. Furthermore, the main constituents of OPC are CaO (69.19%) and SiO₂ (17.42%). Even though GGBFS contains approximately the same proportion of CaO (40.51%) as both BOFS, its alumina (Al₂O₃) concentration is 4–5 times, and silica (SiO₂) concentration is two times greater than in both s- and f-BOFS, respectively. Manganese, phosphorus, titanium, and other cast iron impurities are also found in BOFS and GGBFS.

The significant levels of CaO in GGBFS and BOFS interact instantly with water to generate Ca(OH)₂, a cementing agent. It increases the pH of the pore solution, ensuring that the pozzolanic reaction occurs in an alkaline environment. Hydration products, such as C–S–H and C–A–S–H, are generated when Ca(OH)₂ is reacted with sufficient quantities of SiO₂ and Al₂O₃ from BOFS and GGBFS [25]. This suggests that these components may enhance the alkaline activation and pozzolanic activity in mortar or concrete mixtures. As stated earlier, the f-CaO in BOFS limits its use in construction applications, particularly when used as an aggregate. Thus, ASTM C114–24 Standard Test Methods for Chemical Analysis of Hydraulic Cement [43] were implemented to determine excessive f-CaO, which may result in delayed hydration, causing volumetric instability and potential cracking in the mixtures. Similarly, magnesium oxide (MgO) during hydration produces magnesium hydroxide (Mg(OH)₂), which aids in early strength gain. However, unhydrated MgO can later lead to expansion, posing durability risks. These effects underscore the importance of controlling hydration kinetics through proper curing conditions, pre-treatment of BOFS, or admixture optimization to minimize delayed expansion while harnessing the benefits of the material.

4.1.2. Mineralogical Analysis

Figure 4 shows the highly complicated structure of overlapping peaks in the XRD patterns of fresh and stockpiled BOFS samples. These results follow the chemical compositions of the slags. Although both BOFS include similar mineralogical elements, there were slight changes in a few peaks. In fact, the main difference between these materials is in the abundance of f-CaO and a small amount of portlandite (Ca(OH)₂) in fresh BOFS. In contrast, the stockpiled BOFS contains a high amount of Ca(OH)₂ and calcite (CaCO₃) instead of f-CaO. The following chemical reactions, Equations (1)–(3), can explain these mineralogical changes in both BOFS.

f-CaO + H₂O → Ca(OH)₂

(1)

f-CaO + CO₂ → CaCO₃

(2)

Ca(OH)₂ + CO₂ → CaCO₃ + H₂O

(3)

To put it another way, the presence of unstable phases (free lime or free periclase) in the f-BOFS sample might lead to volumetric instability of BOFS by interacting with water and producing portlandite and brucite (Mg(OH)₂). Expansion is caused by an increase in volume due to hydration products, which are larger in volume (lower density) than their precursors [44].

Due to higher iron oxide content, both f- and s-BOFS contain relatively high amounts of wustite (FeO) and magnetite (Fe₃O₄). Moreover, there are several peaks indicating quartz (SiO₂) and minor amounts of β-C₂S(Ca₂SiO₄) and srebrodolskite (Ca₂Fe₂O₅) due to higher silica and lime contents. Similar XRD patterns have been presented for BOFS by Chen and Lin [45]. Amorphous contents also exist in fresh and stockpiled BOFS, indicating the peaks have a lower intensity with a wider bandwidth. The presence of these phases contributes to the latent hydraulic and pozzolanic reactivity of the material. During blending with GGBFS, these disordered calcium–silicate and aluminosilicate phases can dissolve more readily under alkaline conditions, thereby enhancing the strength and durability of the mortar or concrete.

When GGBFS was analyzed with XRD, the most intense peaks were found at 2-theta values of 25.5° and 31.5°. These peaks are related to crystalline anhydrite (CaSO₄). Another CaSO₄ polymorph that is found in GGBFS is gypsum (CaSO₄·2H₂O). Furthermore, small amounts of CaCO₃, β-C₂S (2CaO·SiO₂), and yoshiokaite (Ca(Al, Si)₂O₄) were also found. Interestingly, in the OPC sample, several narrow peaks correspond to different types of crystals, including β-C₂S (Ca₂SiO₄), hemihydrate (CaSO₄·½H₂O), and monoclinic C₃S (M1 type) (Ca₅₄MgAl₂Si₁₆O₉₀). Other minerals that have been found less often are brownmillerite (Ca₂(Al, Fe)₂O₅), C₃A (3CaO·Al₂O₃), and arcanite (K₂SO₄).

Furthermore, the SG values of the cementitious materials can be correlated with their mineralogical composition and chemical content to understand the relationship between material density and chemical composition. The presence of heavy elements like calcium and iron, as well as the mineralogical phases present in the materials, can influence their SG values. For instance, the higher SG of OPC (3.14) can be attributed to its composition, which includes a high percentage of CaO and dense crystalline phases like β-C₂S, C₃S, and C₃A. The s-BOFS with an SG of 3.08 also contains significant amounts of calcium, iron, calcite, and wustite minerals, which are relatively dense and contribute to their relatively high density. A lower SG of f-BOFS than s-BOFS could be due to unreacted lime, porosity of particles, and less dense amorphous phases. While higher SG in s-BOFS may result from hydration and carbonation, it forms denser phases, such as calcite, that fill some pores within BOFS particles. Finally, GGBFS has the lowest SG among the materials due to its different mineralogical composition compared to the other materials. It is known to have a high silica content and lower iron content, which could contribute to its lower SG and lighter mineral phases, such as gypsum.

4.1.3. Particle Size Distribution

Figure 5 illustrates the particle size distributions of the cementitious materials used in this study. The features of cementitious materials, such as PSD and specific surface area, are critical because they influence their reactivity [46]. GGBFS has the smallest PSD and is smaller than the other cementitious materials, while OPC is the second-finest substance. Despite a slightly finer size distribution of f-BOFS compared to s-BOFS, both materials have a cumulative percentage passing of almost 82% through a 45 µm (#325 sieve), which confirms their pozzolanic material criterion, as required by ASTM C989 [39], which stipulates 80% passing through a #325 sieve.

Furthermore, the results of the mean diameters (D[3, 2]) of the particle size based on area-weighted mean and specific surface areas (SSA) were obtained for more detailed analyses of SDCM sizes. For instance, median values for fresh and stockpiled BOFS resulted in 21.1 µm and 24.9 µm, respectively, whereas those for OPC and GGBFS were 17.7 µm and 12.0 µm, respectively. Then, OPC had the largest D[3, 2] value of 9.56 µm, followed by stockpiled and fresh BOFS, which had 6.69 µm and 5.97 µm, respectively. As it was expected, GGBFS particles had the lowest D[3, 2] value of 5.34 µm. Moreover, the SSA indicated values of 627.8 m²/kg for OPC, 896.4 m²/kg for s-BOFS, 1005 m²/kg for f-BOFS, and 1123 m²/kg for GGBFS.

4.1.4. Microstructure Analysis

Figure 6 presents SEM images of GGBFS and BOFS materials. Both GGBFS and BOFS consist of angular to sub-angular particles. It should be noted that both GGBFS and BOFS materials were ground to obtain an average particle size finer than 45 µm. In particular, irregular and angular particles, as well as globular masses of particles in BOFS, primarily represent the amorphous constituents identified as aluminosilicate, with small amounts of Na and Ca associated. This indicates the unique microstructural characteristics of BOFS, which differentiate it from conventional SCMs, such as GGBFS. However, unlike conventional SCMs such as GGBFS and fly ash, this amorphous phase is not predominantly glassy in nature [47]. SEM observation of the present BOFS supports this view. It suggests that the surface texture of this amorphous material differs from the smooth glassy phase found in fly ash, which aligns with the chemical analysis, indicating the microstructural complexity of BOFS and its distinct surface texture compared to traditional SCMs.

The shape observed in BOFS particles significantly enhances the material’s reactivity by increasing the surface area available for pozzolanic reactions. This contributes to the formation of additional C–S–H and C–A–S–H gels, which are responsible for the development of strength. The microstructural characteristics of BOFS, including its higher angularity, have been linked to accelerated hydration rates. However, its rougher texture may lead to increased water demand, which should be accounted for in the mix design to ensure workability is maintained. In the case of GGBFS, the smoother and more uniform surface texture allows for efficient dispersion in the cement matrix, facilitating the pozzolanic reaction under alkaline conditions. The glassy phase in GGBFS promotes the long-term formation of dense C–S–H, contributing to enhanced compressive strength.

4.2. Fresh Properties

4.2.1. Flowability

Figure 7 represents all mortar mixtures’ relative flowability (Γ_m) values. They range from 1.30 to 3.82. In fact, the flowability of fresh mortar is affected by the PSD and surface characteristics of SCMs. The finer particles and smoother surfaces of SCMs can enhance flowability by reducing internal friction within mortar mixtures. As expected, the mixture containing 50% GGBFS has the highest Γ_m value, while the plain mixture has the lowest Γ_m value at a w/cm ratio of 0.47. It is well known that the workability of a concrete or mortar mixture containing GGBFS increases with the increasing replacement level of GGBFS [10,48]. In fact, fineness, smooth particle morphology, and narrow PSD of GGBFS enhance the packing density and reduce internal friction between particles, as well as its glassy surface, which consumes less water during early hydration compared to clinker phases, resulting in a more fluid and cohesive mix. This indicates that a concrete or mortar mixture containing GGBFS is easier to place, compact, and finish for a given workability.

Adding both f- and s-BOFS materials to the plain mixture also increases the flowability; however, their effect on flowability is less pronounced than that of GGBFS. Interestingly, the blended BOFS and GGBFS mixtures also showed higher flowability than the plain and individual mixtures containing f- or s-BOFS. Moreover, the mixtures containing f-BOFS have relatively higher Γ_m values than those with s-BOFS. However, different combinations of BOFS and GGBFS did not show much difference in the flowability of each mixture. For instance, the average Γ_m for [35% s-BOFS + 15% GGBFS] mixture is 2.08, while the average Γ_m for only the s-BOFS mixture is 1.43. It should be noted that both BOFS and GGBFS have angular and sub-angular particles.

4.2.2. Air-Content

Figure 8 shows the air content of all mortar mixtures. The plain mixture (100% OPC) had the highest air content, whereas the mixture containing 50% GGBFS presented the lowest. All the mixtures containing SCMs had a lower air content than the plain mixture. In fact, the overall mix design plays a crucial role in determining the final air content in the mortar or concrete mixture. Parameters such as w/cm, aggregate size, fineness, and use of SCM influence the mixture’s air content. Excess water in the mixture can create air bubbles during the mixing process. It is generally accepted that using SCM in the mortar/concrete mixture can reduce water demand, leading to lower air content [42,49,50].

Moreover, the fineness and specific surface area of SCMs also influence the air content in the fresh mortar mixture. Finer SCMs can entrain less air in the mortar mixture due to the pore-filling effect. For instance, 50% of GGBFS with a particle size of less than 10 µm may have filled large pores in the mortar system, resulting in the lowest air content.

Furthermore, the mixtures with 50% f- and s-BOFS also exhibited less air content than the plain mixture, although the overall particle size (D50) of both BOFS is larger than that of the cement. These results could be due to the difference in the surface mean (D[3, 2]) value between OPC and BOFS. For instance, the f- and s-BOFS have a smaller value of D[3, 2] than OPC, leading to a pore refinement effect, which implies a shift in large capillary pores by numerous fine pores. Moreover, at a later age, the pore structure becomes more refined, likely due to the filling and pozzolanic effects of BOFS resulting from the consumption of Ca(OH)₂, which creates more C-S-H. Indeed, this process produces additional C-S-H over time, further refining the pore structure by filling voids. This secondary reaction complements the physical pore-filling effect observed in fresh mixtures, leading to a denser matrix and reduced permeability at later ages.

For mixtures with a combined BOFS and GGBFS, the air content increased when the BOFS content increased, irrespective of the f- and s-BOFS content. This increase in air content may be attributed to the increase in BOFS particle size. However, among these combined BOFS and GGBFS mixtures, there was little difference in air content. Like binary mixtures, ternary mortar mixtures with s-BOFS exhibited a higher air content than those with f-BOFS. It should also be noted that higher air content in fresh mortar can impact flowability and setting time, as excessive air entrainment may reduce the mortar’s strength and increase its permeability.

4.2.3. Setting Time

Depending on the chemical and mineralogical properties of SCMs, their incorporation into mortar or concrete mixtures can affect the hydration kinetics of cementitious materials, leading to changes in the setting time and overall concrete properties. For instance, the use of ASTM Class F fly ash is known to prolong setting times due to its slower reaction rate, which can impact workability and construction schedules. Adjusting the proportions of SCMs is therefore critical to controlling setting time and ensuring optimal performance in practical applications.

Figure 9 presents the initial and final setting times of the mortar mixtures. Replacing OPC with 50% GGBFS, 50% f-BOFS, and 50% s-BOFS increased the initial and final settings of the mortar mixtures. Their setting times were retarded regarding the control mixture. For example, the initial setting time of these mixtures changed from 84 min to 185, 222, and 246 min, while the final setting time of these mixtures was retarded from 330 min to 420, 430, and 540 min, respectively. This retarding pattern coincides with the findings reported by several researchers. Shi [24] reported that the GGBFS mixtures typically have longer setting times. Similarly, Jexembayeva et al. [3] reported that an increased amount of BOFS in the mixture results in longer initial and final setting times due to the delayed hardening process caused by the reduced amount of cement and the low hydration rate of the additives.

Ternary mortar mixtures with combined GGBFS and BOFS yielded lower initial and final setting times compared to binary mixtures containing a single SCM. Interestingly, in all cases, the initial setting time in ternary mixtures is higher than that of the plain mixture. However, the final setting time in the ternary mixture is identical to or shorter than that of the plain mixture, except for the [35% s-BOFS + 15% GGBFS] mixture. Moreover, the setting time of the mixture containing GGBFS and f-BOFS decreased with the increase in fresh BOFS content up to a 35% replacement. In contrast, ternary mixtures containing s-BOFS presented the opposite trend. This result may be attributed to high f-CaO content in fresh BOFS. It should be noted that f-CaO content in f-BOFS is 5.67%, while s-BOFS has 0.21% f-CaO content. The replacement of cement with a higher amount of f-BOFS increases the f-CaO content in the mortar mixture. Therefore, the reduced final setting time is likely due to the instantaneous formation of ettringite resulting from the chemical reaction between sulfate and calcium [51].

4.2.4. Water Absorption

The surface area and reactivity of SCMs mainly influence water absorption in a fresh mortar mixture. SCMs with high reactivity require more water for proper dispersion and hydration, affecting the workability of the mortar mix. Moreover, higher water requirements due to SCMs can also impact flowability and air content in fresh mortar. Therefore, balancing water content with SCM properties is crucial for optimal workability and performance.

The pace at which water penetrates through the sample’s pores is referred to as the water absorption rate, which depends on the matrix’s pore distribution. Thus, after the weights of different mortars were recorded at 0, 15, 60, 240, and 1440 min, the water absorption values of each mixture were calculated. Additionally, to analyze the durability of the mortar samples for penetration, the water requirement was tested over a longer period, 20,160 min (approximately 14 days).

Figure 10 presents the water absorption of mortar mixtures with various replacement levels. Regardless of f- and s-BOFS, the mortar mixture containing 50% BOFS had higher water absorption than the other mixtures. These results may be due to the PSD and SSA of both BOFS materials, which have coarser PSD and higher SSA than the OPC mixture. The lowest water absorption was obtained in the [50% GGBFS] mixture. For instance, at 24 h (1440 min) measurement, [50% s-BOFS] and [50% f-BOFS] mixtures resulted in 18.6% and 20.8% higher absorption rates than the control mixture, while [50% GGBFS] exhibited a 17.9% lower value than the 100% OPC mixture, irrespective of all ages.

Moreover, a slight difference in water absorption rate was observed at the initial stage (up to 240 min) between ternary mixtures, regardless of whether they were f- or s-BOFS materials. However, as the time of water immersion increased, at the later stages, it became clear that the water absorption rate increased with an increase in BOFS percentage. For instance, the water absorption rate of the [15% s-BOFS + 35% GGBFS] mixture at 14 days (20,160 min) is 80.2%, while that of the [35% s-BOFS + 15% GGBFS] mixture is 97.65% at the same age. These results may be attributed to the fact that when the cement was replaced with a high amount of BOFS, the mixture, having a coarse particle size of BOFS, was less reactive, and its microstructure was not dense at an early age. Therefore, water can move into the mixture matrix more freely, increasing the water absorption rate [52,53]. It should also be noted that water absorption in hardened mortars is influenced by the total open porosity, pore-size distribution (particularly the volume of connected capillary pores), and the chemical nature of the hydration products. As presented in a later section, the dropping of compressive strength of the mixture with a high BOFS content can support this result.

4.3. Pozzolanic Reactivity

4.3.1. Strength Activity Index and Compressive Strength

The compressive strength of hardened mortar is significantly influenced by the type and proportion of SCMs used in the mixture. Adding SCMs like GGBFS can enhance long-term strength development by contributing to the formation of additional hydration products and denser microstructures. In fact, the material properties of SCMs influence the fresh properties of mortar mixtures, which in turn impact the compressive strength of hardened mortar. Understanding the interplay between fresh mortar properties and compressive strength is essential for optimizing mix designs, ensuring proper workability, and achieving desired strength levels in construction applications. By considering the effects of SCMs on both fresh and hardened properties, engineers can tailor mortar formulations to enhance the development of strength and overall performance.

Figure 11 presents the compressive strength at various curing ages for the binary mixtures with 50% replacement by GGBFS, f-BOFS, or s-BOFS, as well as for the ternary mixtures that combine these materials. The compressive strength of these mortar mixtures was compared with that of the control mixture (100% OPC mortar). The compressive strength of all mixtures increased over time, except the control mixture. At an early age (7 days), all mixtures containing GGBFS, f-BOFS, s-BOFS, or a combination of GGBFS and BOFS exhibited lower compressive strength than the control mixture. However, at 28 days, the mixture containing 50% GGBFS had a higher compressive strength than the control mixture. This result is due to the GGBFS property used in this study. Most of the CaO in GGBFS is bound as calcium silicate, calcium aluminate, and calcium aluminosilicate, similar to those found in cement. When these components are activated by a hydration product of cement, specifically Ca(OH)₂, they hydrate and contribute to the development of strength. As expected, the compressive strength of the GGBFS mixture continued to increase up to 91 days, reaching its highest value, which was 61% higher than that of the plain mixture.

On the other hand, replacing 50% of OPC with both BOFS materials does not contribute to strength development at either early or later ages, unlike GGBFS. For example, the 7-day compressive strengths of the f- and s-BOFS mixtures were 7.88 MPa and 7.55 MPa, respectively, which are 58% and 60% lower than the control mixture. The compressive strength gap between the control and BOFS mixtures decreased as hydration progressed. However, even at 91 days, the f-BOFS mixture had 45.1% lower compressive strength than the plain mixture, while the stockpiled one had 44.5% lower compressive strength than the plain mixture. The lower compressive strength in mixtures containing both BOFS materials can be attributed to the presence of CaCO₃, Ca(OH)₂, and Fe₃O₄ in BOFS, as determined by XRD analysis. Despite the CaO content being similar to that of GGBFS, the hydrated forms of CaO in BOFS do not fully contribute to the development of strength. Some work as inert materials that only physically fill in pores in the microstructure of the OPC-BOFS mixture system. Also, the relatively high amount of Fe₃O₄ in BOFS may negatively influence the hardening of the BOFS mixture. This result aligns with the findings of previous studies by other researchers [3,54].

The ternary mixtures containing BOFS and GGBFS had lower strength than the single GGBFS mixture but higher than those containing single BOFS at all ages. For instance, at 7 days, the strength of the [15% f-BOFS + 35% GGBFS] mixture was 11.88 MPa, which is slightly higher than the [50% f-BOFS] mixture (7.88 MPa), whereas the strength of the [25% s-BOFS + 25% GGBFS] mixture resulted in 13.23 MPa, which is lower than the [50% GGBFS] mixture (18.35 MPa). As the mortar mixture’s hydration continued, the strength values of the ternary mixtures increased drastically. At the final testing date, they exceeded the maximum compressive strength of the control mixture (21.96 MPa). For instance, the [15% f-BOFS + 35% GGBFS] mixture reached 26.91 MPa at 91 days, indicating a 223% increase compared with the [50% f-BOFS] mixture (12.06 MPa). Among the ternary mixtures with s-BOFS, the highest compressive strength was obtained from the [25% s-BOFS + 25% GGBFS] mixture, which achieved the strength of 23.83 MPa at the age of 91 days. It is essential to note that the combined use of BOFS and GGBFS can synergistically enhance compressive strength and provide guidance on how to utilize BOFS as a pozzolanic material effectively.

Figure 11 also presents the effect of different replacement ratios of BOFS on the compressive strength. The compressive strength of mortar mixtures decreases as the percentage of BOFS increases, except for the [25% s-BOFS + 25% GGBFS] mixture. This result may be due to the low contribution of the cementing property of BOFS, which contains inert phases such as dicalcium ferrite (C₂F) and Fe₃O₄. The higher the content of BOFS, the lower the hydraulic reactivity. Therefore, it is necessary to find methods to improve the strength development over the incremental content of BOFS in the combined BOFS-GGBFS mixtures. As suggested by many researchers [55,56,57,58], various methods can be applied. These include (i) making BOFS finer by grinding BOFS, (ii) activating BOFS chemically and thermally using calcium chloride dihydrate (CaCl₂…2H₂O) and sodium chloride (NaCl), and three different temperatures (20 °C, 40 °C, and 60 °C), and (iii) applying higher temperature carbonation curing.

4.3.2. Strength Activity Index

ASTM C311 [39] and ASTM C989 [37] define the strength activity index (SAI) as the ratio of the mortar mixture containing SCMs to the reference mortar mixture in terms of compressive strength. This means that the reference mixture, which is 100% OPC, is used as the denominator, and the mortar mixture containing SCMs is used as the numerator. The SAI is usually utilized to evaluate the pozzolanic packing effect of SCMs [59]. When the SAI is used to assess the slag activity index based on slag performance, the C989 classifies three different grades: Grade 80, Grade 100, and Grade 120. Whereas Grade 80 represents the lowest slag activity, Grade 120 represents the highest. Each grade requires the minimum percentage of the SAI at 7 and 28 days, which is as follows: Grade 80 (no value at 7 days and 70% at 28 days), Grade 100 (70% at 7 days and 90% at 28 days), and Grade 120 (90% at 7 days and 110% at 28 days). Thus, blended mixtures of this study were tested for pozzolanic properties using the SAI.

As presented in Figure 12, the compressive strength result of the control mixture at each age has a 100% strength activity index, while the strength activity indices of the other mixtures were calculated. For instance, based on the compressive strengths of the [50% GGBFS] mixture at 7 days (18.35 MPa) and 28 days (28.20 MPa), their SAI values were calculated as 97.2% at 7 days and 137.4% at 28 days. Therefore, the GGBFS used in this study can be classified as Grade 120 slag (SAI requirements), indicating its relatively higher hydraulic activity level and pozzolanic reactivity. As expected, the [50% f-BOFS] and [50% s-BOFS] mixtures resulted in the lowest compressive strengths with 41.8% and 40.0% SAIs at 7 days and 60.2% and 48.0% SAIs at 28 days, respectively. These f- and s-BOFS mixtures did not satisfy any slag grade presented in SAI requirements.

All ternary mixtures combined with f-BOFS and GGBFS satisfied the Grade 80 slag activity index standards because they had better compressive strength. The combinations with a larger quantity of GGBFS resulted in higher compressive strength and SAI. For example, the blended mixture [15% f-BOFS + 35% GGBFS] demonstrated a 7-day strength activity index of 63.0%, then 100.5% at 28 days, and 122.5% at 91 days, respectively. In addition, the mixtures [25% f-BOFS + 25% GGBFS] and [35% f-BOFS + 15% GGBFS] samples achieved 63.0% and 51.6% at 7 days, 95.4% and 81.0% at 28 days, and 120.1% and 100.4% at 91 days, respectively. The SAI of these ternary mixtures continuously increased to 91 days.

The SAI in the ternary mixtures containing GGBFS and s-BOFS also presented the same trend as those having GGBFS and f-BOFS. The ternary mixture with a higher amount of GGBFS has developed a higher SAI. Regardless of the BOFS content, the SAI in all ternary mixtures increased continuously up to 91 days, except for the [15% s-BOFS + 35% GGBFS] mixture. Although the [15% s-BOFS + 35% GGBFS] mixture showed substantial strength development up to 56 days (115.9%), the 91-day SAI of this mixture dropped to 101.45%, which is slightly smaller than the SAI of the [25% s-BOFS + 25% GGBFS] mixture (108.6% at 91 days). This behavior may be attributed to differences in late-age hydration kinetics between BOFS and GGBFS, as well as potential microstructural changes such as pore refinement and internal stress development that can influence strength gain at extended curing ages. Despite this slight reduction, the 91-day strength still surpasses the control strength, confirming that the mixture maintains adequate long-term performance. In fact, the mixture [15% s-BOFS + 35% GGBFS] belongs to the Grade 80 index, whereas the mixture [25% f-BOFS + 25% GGBFS] satisfies the Grade 100 criterion. However, the ternary mixture [35% s-BOFS + 15% GGBFS] does not belong to any grade.

In summary, the higher strength development in ternary mixtures compared to a binary mixture containing 50% BOFS and a control mixture at a later age may be attributed to the pozzolanic reaction, regardless of whether the BOFS materials are fresh or stockpiled. C-S-H and/or C-A-S-H are formed when the hydration product Ca(OH)₂ reacts with siliceous and aluminous minerals found in GGBFS and BOFS [25]. Finally, it is possible to employ BOFS as a pozzolanic material in the concrete and mortar mixtures by combining OPC, GGBFS, and BOFS.

4.3.3. Chapelle Test

Chapelle’s test [59] was used to determine the pozzolanic reactivity of GGBFS, BOFS, and the combined BOFS and GGBFS materials. It determines the lime consumption through the reaction of lime-SCMs (pozzolanic materials) made with a 1:1 mass proportion of lime to SCMs. Lime consumption is related to the amorphous or vitreous phase of the SCMs. Calcium consumption is also high if the SCM has high pozzolanic reactivity.

Figure 13 presents the calcium consumption results of different SCMs. As expected, [50% GGBFS] had the highest calcium consumption (286.0), which is 19.2% greater than that of [50% f-BOFS] and 15.8% greater than that of [50% s-BOFS]. All other f-BOFS mixtures have similar pozzolanic activity, ranging from 235.7 to 245.0. Interestingly, s-BOFS had a higher CaO consumption than f-BOFS, regardless of the combined content of GGBFS and BOFS. It should be noted that the s-BOFS contains higher Ca(OH)₂ content than f-BOFS because it was exposed to air and moisture. If BOFS reacts with water, Ca(OH)₂ is formed. Therefore, the Ca(OH)₂ remaining in s-BOFS can be easily consumed, eventually leading to higher CaO content.

4.3.4. Thermogravimetric Analysis

The thermogravimetric analysis provides information on the hydration products and phases of the mortar mixture over time. The evolution of hydration products, such as calcium silicate hydrates (C-S-H) and calcium hydroxide (Ca(OH)₂), influences the development of compressive strength. A higher content of C-S-H is typically associated with increased strength, while the presence of unreacted Ca(OH)₂ indicates incomplete hydration and lower strength. This test helps to assess the pozzolanic activity of SCMs by monitoring weight loss due to the decomposition of hydrated phases. Thus, additional hydration peaks or shifts in peak temperatures may correlate with higher SAI values and improved strength development.

The thermal properties of different mortar mixtures were determined using a simultaneous thermal analyzer (STA), which combines a thermogravimetric analyzer (TGA) for weight loss data and a differential scanning calorimeter (DSC) for heat flow data, after curing in water for 91 days at room temperature. The relative mass change (%) of mixtures with fresh and stockpiled BOFS is illustrated in Figure 14, and the results of heat flow (mW/mg) are shown in Figure 15, respectively.

Table 3. Mass loss at three dominant peaks in thermogravimetry.

Mixture	Mass Loss at Each Temperature (%)			Total Mass Loss (%)
	25–350 °C	350–450 °C	550–750 °C
100% OPC	7.27	1.13	1.1	9.50
50% GGBFS	4.03	1.16	1.70	6.89
50% f-BOFS	5.03	1.13	1.29	7.45
50% s-BOFS	4.34	1.24	1.30	6.88
15% f-BOFS + 35% GGBFS	8.58	1.05	1.04	10.67
25% f-BOFS + 25% GGBFS	7.77	1.11	0.83	9.71
35% f-BOFS + 15% GGBFS	4.22	1.12	1.42	6.76
15% s-BOFS + 35% GGBFS	8.68	0.95	1.04	10.67
25% s-BOFS + 25% GGBFS	8.02	1.12	0.86	10.00
35% s-BOFS + 15% GGBFS	5.53	1.14	1.48	8.15

summarizes the weight loss percentage at three dominant peaks.

Generally, thermogravimetry (TG) graphs exhibit three distinct steps. First, the highest weight loss is obtained in the 25–350 °C temperature range. The initial mass loss observed from ambient temperature to 110 °C is due to the evaporation of physically absorbed water from the pores. At the same time, the release of chemically combined water occurs from 110 °C to 350 °C, which includes the dehydration of the C-S-H gel. In this temperature range, the minimum mass loss (4.05%) was detected at the [50% GGBFS] mixture, while the maximum mass loss (8.58%) was found at the [15% s-BOFS + 35% GGBFS] mixture. Interestingly, the mixtures with higher BOFS content exhibited greater weight changes, regardless of whether the BOFS was fresh or stockpiled.

The next feature appeared in the temperature interval from 350 °C to 550 °C, which is associated with the dehydroxylation of Ca(OH)₂, as illustrated in Equation (2). This step is also accompanied by a mass loss, which is visible in the TG graphs in Figure 14 and Table 3. The mass loss due to Ca(OH)₂ decomposition is related to the pozzolanic reaction of each mixture and the development of strength. For instance, the amounts of Ca(OH)₂ decomposition were determined as 1.05%, 1.11%, and 1.12% for mixtures of [15% f-BOFS + 35% GGBFS], [25% f-BOFS + 25% GGBFS], and [35% f-BOFS + 15% GGBFS], respectively. These results indicate that more Ca(OH)₂ exists in the mortar mixture. In other words, increasing Ca(OH)₂ with increased BOFS replacement level implies less consumption in C-S-H formation due to pozzolanic reaction in the mortar mixture [60,61]. This finding supports the decrease in compressive strength at 91 days, regardless of whether the BOFS materials are fresh or stockpiled, as the BOFS content increases.

Ca(OH)₂ → CaO + H₂O

(4)

Finally, the third thermal event occurred in the 550 °C to 750 °C temperature interval. It is well known that the endothermic peak at these temperature ranges is attributed to the decarbonization of calcium carbonate (CaCO₃), as expressed in Equation (3) [62]. Mixtures containing more BOFS show relatively higher mass loss in CaCO₃ decomposition. For example, the [15% s-BOFS + 35% GGBFS] mixture has a 1.04% mass loss in CaO₃ content, while the [35% s-BOFS + 15% GGBFS] mixture has a 1.48% mass loss in CaO₃ content. Moreover, the mass loss of CaCO₃ decomposition in the [50% s-BOFS] mixture is higher than in the [50% f-BOFS] mixture. It should be noted that s-BOFS was exposed to air for a longer period than f-BOFS.

CaCO₃ → CaO + CO₂

(5)

4.4. Durability Properties

4.4.1. Alkali-Silica Reaction (ASR) Resistance

ASTM C1260/1567 test methods [35,63] provide assessment criteria of the ASR potential of aggregate by measuring the length change in mortar bars. If the expansion is less than 0.1% at 14 days, the aggregate is innocuous (low risk of ASR). If the expansion exceeds 0.2%, the aggregate is reactive (high risk of ASR). Finally, the aggregate is potentially reactive if the expansion is between 0.1% and 0.2%. The C1260/1567 test methods are also used to evaluate the ASR mitigation performance of SCMs through the same criteria [64]. Therefore, all mortar mixtures used the pre-determined reactive siliceous river sand in the ASR test.

Figure 16 presents mortar bar expansion characteristics due to ASR. The expansion of the control mixture displayed 0.23% and 0.32% at 14 days, which was considerably higher than that of the other mixtures. When the test period was extended to 28 and 56 days, the mortar bar’s expansion continued to increase, reaching 0.32% and 0.44%, respectively. However, the expansion of all binary and ternary blended mortar mixtures, including GGBFS and f- and s-BOFS, was less than 0.1% at 14 days, suggesting innocuous field performance behavior, except for the mixture [50% s-BOFS]. For example, the [50% GGBFS], [15% f-BOFS + 35% GGBFS], [15% s-BOFS + 35% GGBFS], and [35% s-BOFS + 15% GGBFS] mixtures expanded at the lowest rates of 0.019%, 0.014%, 0.021%, and 0.034% at 14 days, respectively. Although the testing period increased to 56 days, these four mixtures still exhibited less than 0.1% expansion. However, all other mixtures exhibit greater than 0.1% expansion at 56 days, which is a criterion for an indeterminate zone.

Interestingly, [25% s-BOFS + 25% GGBFS] and [25% f-BOFS + 25% GGBFS] mixtures have less than 0.1% expansion up to 28 days. They started to exceed the limit value at 42 days, resulting in the highest expansion among blended mortars at 56 days. These results indicate that an increase in BOFS content in the mortar mixture does not always result in reduced volume stability. Therefore, there may be an optimum combined content between GGBFS and BOFS materials to inhibit ASR expansion.

Based on the findings of the ASR tests, it can be inferred that the inclusion of BOFS fines may help avoid the detrimental expansion behavior of mortar mixes at an early stage, owing to the mixture’s reduced total alkali content. However, it was unable to properly stabilize expansion in later ages, most likely due to both ASR and residues of f-CaO and f-MgO, which create Ca(OH)₂ and Mg(OH)₂, respectively [65]. When BOFS is combined with adequate amounts of GGBFS, these materials produce pore refinement due to the reduced particle size and grain refinement through the pozzolanic process, which consumes precipitated Ca(OH)₂. Finally, the ternary combination of GGBFS and BOFS inhibits detrimental growth caused by ASR, f-CaO, and f-MgO. Additionally, it reduces the permeability of the mortar, reducing the possibility of exogenous alkali migrating into it [66].

4.4.2. Sulfate Attack Resistance

Figure 17 illustrates the expansion analysis of the sulfate attack test for approximately 6 months (189 days). Regardless of mixture types, the expansion of all mixtures was less than the ASTM C1012 test’s threshold value of 0.1% [34,67]. All other mortar combinations, including GGBFS and BOFS, expanded more slowly than the control mixture. The length of mortar bars containing BOFS was initially reduced but steadily increased over time. For instance, in cases of 50% GGBFS and f-BOFS, regardless of some fluctuations in the expansion, [50% GGBFS], [50% f-BOFS], [15% f-BOFS + 35% GGBFS], [25% f-BOFS + 25% GGBFS], and [35% f-BOFS + 15% GGBFS] mixtures reached 0.013%, 0.028%, 0.013%, 0.008%, and 0.023% at 6 months, respectively. Because the tricalcium aluminate (C₃A) amounts in BOFS and GGBFS are less than those of conventional OPC, the total quantity of C₃A in the blended mixtures is lowered. Siddique and Bennacer [66] stated that the decrease in the content of mono-sulfoaluminate and calcium aluminate hydrates in OPC mortars may reduce sulfate attack. Furthermore, GGBFS and BOFS contain amorphous silica and alumina that may react with Ca(OH)₂ and a sodium sulfate (Na₂SO₄) solution, resulting in a blended mixture with significantly less ettringite [56]. Lastly, the permeability of the blended mortar is significantly reduced due to the inclusion of pozzolanic components, resulting in decreased penetration of the Na₂SO₄ solution. As a result, binary and ternary mixtures with GGBFS and BOFS minimize the sulfate attack.

In contrast, the control and [50% s-BOFS] mixtures exhibited expansions of 0.035% and 0.038%, respectively, at the same age. It should be noted that the OPC mixture continuously produces Ca(OH)₂, and stockpiled BOFS contains more Ca(OH)₂ due to the weather effect (Equation (2)), which is more vulnerable to sulfate attack.

4.4.3. Drying Shrinkage

Figure 18 shows the drying shrinkage results of all mortar mixtures for approximately 7 months (203 days). The control mixture and the [50% GGBFS] mixture had higher drying shrinkage values than the other mixtures, which were 0.00121 μm and 0.00115 μm, respectively, at 203 days. In contrast, the smallest drying shrinkage occurred in the [50% f-BOFS] and [50% s-BOFS] mixtures, which are 0.00078 μm and 0.00084 μm, respectively. In other words, both mixtures with 50% BOFS replacement yielded approximately 35% lower drying shrinkage than the control mixture and 30% lower than mixtures made of 50% GGBFS.

Interestingly, the drying shrinkage of mortar mixtures combined with both BOFS and GGBFS increased as the percentage of GGBFS increased, except for the [25% s-BOFS + 25% GGBFS] mixture. This result may be attributed to the higher cementing properties of GGBFS compared to BOFS, which contains some inert phases, such as dicalcium ferrite (C₂F) and Fe₃O₄. Increasing GGBFS content in the mortar mixture enhances cement hydration and increases its strength. Therefore, the higher the GGBFS content, the higher the drying shrinkage is in the ternary mixtures.

In fact, drying shrinkage is a kind of independent stress deformation caused by the evaporation of internal free water in a hardened mortar or concrete mixture under the surrounding environment [68,69]. It is linked to the macro- and mesoscale pore structure of hydrated cement paste. The free water initially and quickly evaporates from the macropores, and the capillary stress (pressure) is minimal. This phenomenon corresponds to high initial drying shrinkage up to 10 days, as shown in Figure 18. Continuous evaporation reduces the amount of free water in the macropores. As a result, the water in the mesoscale pores gradually evaporates, leading to continuously increasing shrinkage due to a large capillary stress. This phenomenon is associated with drying shrinkage after 10 days, as shown in Figure 18.

5. Evaluation of Binary and Ternary Mixtures Using a Performance-Based Ranking System

Table 4. Summary of fresh and hardened properties of mortar mixtures and their ranking.

Mixtures	1 SAI (%)	Rank	2 ASR (%)	Rank	3 Sulfate Attack (%)	Rank	Drying Shrinkage (μm)	Rank	Flow-Ability (Γm)	Rank	Air Content (%)	Rank	Counting			Final Ranking (4 pts.)
													No of pts 3	No of pts 2	No of pts 1
100% OPC	100.0	-	0.325	10	0.0345	9	1.211	10	1.30	10	15.06	9	0	0	0	10 (0)
50% GGBFS	137.4	1	0.018	1	0.0127	3	1.151	9	3.82	1	10.10	1	4	0	1	1 (13)
50% f-BOFS	60.2	7	0.091	7	0.0278	8	0.784	1	1.87	7	12.96	5	1	0	0	6 (3)
50% s-BOFS	48.0	9	0.103	8	0.0385	10	0.837	3	1.43	9	15.22	10	0	0	1	8 (1)
15% f-BOFS + 35% GGBFS	100.5	2	0.037	3	0.0133	4	1.119	8	2.07	5	12.28	2	0	2	1	2 (5)
25% f-BOFS + 25% GGBFS	95.4	4	0.072	5	0.0085	1	0.986	5	1.89	6	12.88	4	1	0	0	5 (3)
35% f-BOFS + 15% GGBFS	81.0	6	0.105	9	0.0230	6	0.986	6	2.27	2	12.83	3	0	1	1	7 (3)
15% s-BOFS + 35% GGBFS	84.7	5	0.019	2	0.0135	5	0.810	2	1.78	8	13.02	6	0	2	0	3 (4)
25% s-BOFS + 25% GGBFS	97.3	3	0.076	6	0.0125	2	0.992	7	2.10	3	14.43	7	0	1	2	4 (4)
35% s-BOFS + 15% GGBFS	60.0	8	0.047	4	0.0275	7	0.917	4	2.08	4	14.57	8	0	0	0	9 (0)

Note ¹. Strength activity index (SAI) criteria (ASTM C989): minimum 70% at 28 days. ². Alkali-silica reaction criteria (ASTM C1260/C1567): maximum 0.1% at 14-day. The test period was extended up to 28 days. ³. Sulfate attack criteria (ASTM C1012 [34]): maximum 0.1% at 6 months. ⁴. The final ranking of each mortar mixture was calculated by comparing the points (pts) of each mixture obtained at different tests: top-ranking (3 pts), second-ranking (2 pts), and third-ranking (1 pt).

summarizes the fresh and hardened properties of a total of 10 mortar mixtures, comprising individual GGBFS, f-BOFS, s-BOFS, and blended BOFS and GGBFS, based on tests conducted according to the C989 and C311 specifications. Additionally, Table 4 presents the ranking of each mixture obtained from the threshold value of each test. The final ranking of mortar mixtures was determined by comparing the points obtained for each mixture in various tests. For instance, the best-performing mixture per test ranks 1, while the lowest-performing mixture ranks 10. Then, the top 3 mixtures per test receive points: the top-ranking mixture (the best-performing mixture) earns 3 points, while the second- and third-ranking mixtures earn 2 and 1 points, respectively. The mixture with the highest points is ranked as the number one, indicating it is the best mixture. The control mixture received no points for each test, resulting in last place on the list because it failed to meet the ASR criterion, which is an expansion of less than 0.1%.

As expected, the mixture containing 50% GGBFS performed best as an SCM, achieving four 1st-rankings in the SAI, ASR, flowability, and air content tests, and one 3rd-ranking in the sulfate attack test. However, its shrinkage property was ranked as one of the lowest, between [15% f-BOFS + 35% GGBFS] and the control mixture. Meanwhile, the mixtures with 50% f-BOFS and 50% s-BOFS showed the top performance in terms of drying shrinkage, ranking 1st and 3rd based on measurements up to 203 days, but did not meet the strength activity index threshold, and ultimately took sixth and eighth places, respectively.

Interestingly, the blended mixtures containing 15% BOFS and 35% GGBFS did not gain any top-ranking criterion, but the [15% f-BOFS + 35% GGBFS] mixture obtained two 2nd- and one 3rd-ranking items, whereas the mixture with s-BOFS at the same proportions gained two 2nd-rankings. Consequently, these two mixtures ranked second and third in the total SCM evaluation system. Moreover, it is worth noting that the mixtures combined with 25% BOFS and 25% GGBFS achieved the top two rankings in sulfate attack resistance, with the f-BOFS mixture ranking first and the s-BOFS mixture ranking second. Then, the [25% s-BOFS + 25% GGBFS] mixture obtained 4 points, with one 2nd ranking and two 3rd rankings, resulting in a fourth place overall. In contrast, the [25% f-BOFS + 25% GGBFS] mixture was ranked fifth with 3 points. Although the mixture containing [35% f-BOFS + 15% GGBFS] exhibited relatively good flowability and air content performance, it exceeded the ASR criterion when the ASR test period was extended to 28 days. Meanwhile, the [35% s-BOFS + 15% GGBFS] mixture did not achieve any rankings in any of the tests, resulting in one of the last places. These mixtures ended up with seventh and ninth rankings, respectively.

This stems from the inherent properties of slags and their synergistic interactions. The high alkalinity and f-CaO content in BOFS, particularly in the fresh state, contribute to secondary hydration when combined with GGBFS. This reaction enhances the formation of additional C-S-H or C-A-S-H, improving strength development and sulfate resistance. For s-BOFS, aging reduces the f-CaO content, thereby enhancing stability and improving drying shrinkage and sulfate resistance, albeit at the slight expense of reactivity and strength-related rankings. For instance, the [15% BOFS + 35% GGBFS] and [25% BOFS + 25% GGBFS] mixtures demonstrated optimal performance due to the complementary roles of the latent hydraulic activity of GGBFS and the alkaline activation of BOFS. These combinations promote pozzolanic reactions, reduce the risk of deleterious ASR expansions, and maintain workability and durability. In contrast, higher proportions of BOFS (e.g., 35%) in the mixtures can lead to challenges, such as increased f-CaO content, which can cause instability or higher ASR susceptibility, particularly for f-BOFS. Thus, it can be concluded that when the material newly synthesized with BOFS and GGBFS is used as the SCM at a 50% replacement rate with cement, either a combination of 15% BOFS and 35% GGBFS or 25% BOFS and 25% GGBFS can be accommodated as an SCM in the mortar and concrete application.

6. Conclusions

This paper explored the feasibility of newly synthesized materials, including f-BOFS, s-BOFS, and GGBFS, as a sustainable alternative to conventional SCMs. The key contribution lies in the detailed analysis of material properties and performance of binary and ternary blends, which have not been comprehensively explored in similar studies. Moreover, the findings revealed that weathering significantly influences the reactivity and overall performance of BOFS, underscoring the importance of comparing its fresh and stockpiled states of slag. This work also provided valuable insights into optimizing SCM combinations for enhanced performance, demonstrating that BOFS, despite its limited hydraulic reactivity, can synergistically improve early-age hydration rates and compressive strength when blended with GGBFS. Further, the main findings from the study include the following:

• Chemical composition analysis shows that f-BOFS has a higher f-CaO content than the s-BOFS, which are 1.34% and 0.21%, respectively. While f-BOFS has an abundance of CaO and a small amount of Ca(OH)₂, the s-BOFS contains a high amount of Ca(OH)₂ and CaCO₃ instead of CaO due to weathering in the field.
• Incorporating fresh and stockpiled BOFS materials into GGBFS reduced relative flowability but increased air content. Blending both f- and s-BOFS materials with GGBFS reduced drying shrinkage.
• The sole use of f- or s-BOFS as a 50% replacement for cement did not yield adequate strength development. However, mixtures incorporating f-BOFS exhibited slightly higher strength than stockpiled counterparts. The combined use of BOFS and GGBFS demonstrated enhanced strength development, meeting Grade 80 criteria according to the C989 Slag standards, except for the [35% s-BOFS + 15% GGBFS] mixture.
• A binary mixture incorporating s-BOFS alone exceeded the ASR criterion at 28 days, while the binary mixture with 50% GGBFS or f-BOFS mitigated ASR expansion. All ternary mixtures met the ASR and sulfate attack criteria at both 28 days and 6 months.
• The top-performing blends, consisting of fresh/stockpiled BOFS combined with GGBFS, either a combination of [15%BOFS + 35% GGBFS] or [25%BOFS + 25% GGBFS], can be accommodated as an SCM in the mortar and concrete application.

These findings highlight the potential of BOFS and GGBFS blends, especially at a 50% cement replacement rate based on the C 989, as a viable SCM for sustainable construction applications, reducing dependence on traditional cement and addressing industrial by-product utilization challenges. However, the study needs to extend various cement replacement ratios using the top four blended mixtures of [15% f-BOFS + 35% GGBFS], [15% s-BOFS + 35% GGBFS], [25% f-BOFS + 25% GGBFS], and [15% s-BOFS + 35% GGBFS] under different test conditions. For example, future research can extend this work by investigating the chemical activation of slag and the long-term durability of slag-based mortars/concrete under varying exposure conditions, including carbonation curing, freeze–thaw cycles, and standardized field exposure tests. Scaling these findings to pilot or field applications, including reinforced concrete structures, will validate their practical feasibility and contribute to sustainable construction practices globally.