Characterizing Chronologically Aged Basic Oxygen Furnace Slags as Aggregates and Their Use in Asphalt Concrete Mix as Filler

Abstract

Before using basic oxygen furnace slag (BOFS) in any engineering application, it is important to determine its properties. The chemical composition, mineralogy, and physical properties of BOF slag are subject to large fluctuations as a result of different raw additives, different compositions of the molten iron used for producing the steel, oxygen stirring of a molten pool, uneven temperature fields, and other complex physical conditions. Thus, in this research, the engineering properties of BOF slag aggregates with different ages were identified, and then the feasibility of BOF slag aggregates in mortar application was investigated. It was found that stockpiled BOFS was harder than fresh BOF slag, which had lower aggregate crushing values and lower LA abrasion values. Also, stockpiled BOFS showed less expansion than fresh BOF slag regardless of water and 1 M NaOH solution immersion. The chemical reaction between f-CaO, f-MgO, and water due to weathering in the field extremely reduced the expansion of BOFS submerged in water. BOFS may contain reactive silica, which causes an alkali–silica reaction (ASR). Stockpiled BOFS (100%, 75%, 50%, 25%) used as a mineral filler in asphalt concrete mix brought about low tensile strength at break up (crack), low compressive strength at +50 °C, poor cohesion, and residual porosity. However, these parameters were partially improved with the use of a thermostable adhesive additive for road bitumen based on polyphosphoric acid esters. The comprehensive assessment in this study indicates that while some mixtures meet the specified criteria for certain properties, there are challenges, particularly regarding crack resistance and cohesion, that need to be addressed to fully align with the standard. Adjustments to the mixture proportions, the exploration of alternative additives, and the use of different types of fillers may be necessary to achieve the desired properties, especially in terms of crack resistance and cohesion.

1. Introduction

In modern global development, environmental sustainability is crucial, especially in metallurgy, which generates significant waste, like slag, sludge, and dust. Millions of tons of industrial waste, including steel slags, are typically disposed of in landfills, depleting land resources and incurring high management costs. Researchers in countries like China, the United States, and South Korea have explored using basic oxygen furnace (BOF) slags in construction, but this has not been widely studied in Kazakhstan. BOF slags’ chemical composition and properties can vary due to factors like raw materials and production conditions.

Industrial steel by-products (slags) are successfully used as aggregate substituents in concrete production and roadway construction [1]. Steel slags have good shearing resistance and high density and are suitable aggregates for different civil engineering works, such as earth cover and embankment construction [2]. It is said that BOF and electric arc furnace (EAF) slags are beneficially used in asphalt mixtures due to having a “rough surface texture, high angular shape, high specific gravity” [3]. Moreover, due to the “rich content of Fe₂O₃, CaO, SiO₂ and MgO” in slags, these are quite good for cement, fertilizer production, and soil refining purposes [4]. The use of steel slags is energy efficient and can reduce carbon dioxide emissions by almost 40% compared to commonly used cement [5].

BOF slags are non-metallic solid wastes consisting of silicates, alumina silicates, calcium aluminum silicates, and iron oxides [6]. It was found that highly oxidized elemental iron is commonly detected in BOF slags, as some portion of molten iron cannot be recovered into the steel [7,8]. Despite the presence of harmful metals in BOF slags, a high phosphorus content must also be dealt with, as it minimizes steel quality [9]. Thus, special attention must be paid to these issues when using slags. However, slag components may vary from batch to batch, depending on the cooling, treatment, and storage conditions [10]. In addition, calcium silicates in BOF slags, such as “dicalcium silicate (Ca₂SiO₄), tricalcium silicate (Ca₃SiO₅), and wollastonite (CaSiO₃)” make these slags acquire cementitious characteristics [10]. BOF slags contain around 25–55 wt.% of lime, which is problematic [11]. Namely, BOF and EAF slags are often stockpiled outside, where they are subjected to the hydration of lime (CaO) and magnesium (MgO) oxides [7]. This results in “free lime volume instability”, which restricts the usage of BOF and EAF in conditions where moisture is present. There are examples of road surfaces with slags that are hydrated and then become fractured due to the aforementioned free lime expansion [9]. A 10% volume increase in BOF slag in such conditions and the generation of hydroxides (Ca(OH)₂ and Mg(OH)₂) at ambient temperatures were detected [12]. The iron content in BOF slags is equal to a 14–30 wt.%, which restricts BOF slag use in cement production [13]. However, in order to avoid volume instability of BOF, it is recommended that slags undergo aging or weathering processes before their application; these processes may take some time [14].

Moreover, BOF slags with limited applicability also impose a problem regarding storage and management, as more land spaces are required to accommodate yearly increasing amounts of BOF slags and high management expenses to deal with BOF’s environmental issues. In 2000, in Australia and New Zealand, 64% of BOF slags (510,000 tons) were utilized in “sealing aggregate, asphalt aggregate, base, subbase, construction fill, subsoil drains and grit blasting” [13]. In some other countries, BOF slags are used as fertilizer for agricultural purposes by increasing the phosphorus amount in the slags [15]. BOF slags are also used as soil conditioners [16] and in glass production [15]. In Kazakhstan, only a few researchers have dealt with BOF slag reuse. BOF slags from JSC “ArcelorMittal Temirtau” (Temirtau, Kazakhstan) with a pH of 12.07, conductivity of 1087 µs/cm, and a specific gravity of 2.09 are used to produce BOF-enhanced cement mortar [17].

In addition, an ordinary asphalt concrete mix must contain coarse aggregate, fine aggregate, filler, and bitumen (asphalt binder). Several studies have been considered where BOF slags are used as coarse aggregates, fine aggregates, and filler. However, it must be noted that the substitution of natural aggregates (conventional materials, such as basalt, limestone, granite, dolerite, and others) with BOF slags in asphalt mixes should be carried out for either the fine aggregate fraction only or the coarse aggregate fraction only [18]. The authors highlighted that “if slag is used to replace both the coarse and fine aggregate components in HMA, the mix is likely to have a lot of air voids, necessitating the use of high quantities of bitumen; this could lead to bulking and flushing problems in a road pavement”.

This refers to the development of mix designs with steel slags, such as BOF slags. However, the application of only BOF slags as both coarse and fine aggregates in asphalt mixtures is not a good decision. Asphalt with such an aggregate skeleton will exhibit poor workability, as BOF slags have high angularity and significantly rough texture. Therefore, it would be rational to use BOF slags in combination with other conventional aggregates [19,20]. For example, using a combination of BOF slag and limestone as coarse aggregates makes asphalt mixture more resistant to fatigue failure and plastic deformation [21].

In general, many Kazakhstani studies have claimed that the problem with the utilization of basic oxygen furnace slags has not been resolved yet in the whole country. Most of the studies have performed only a preliminary characterization analysis on basic oxygen furnace slags and no further analysis has been carried out. Thus, in this research, the engineering properties of BOF slag aggregates were identified, and the feasibility of BOF slag aggregates in mortar application was investigated. Several types of BOF slags were considered: stockpiled BOF, fresh BOF, CO₂-exposed BOF. BOF aggregates were analyzed in comparison with an ordinary river sand material. Moreover, in this research, mineral filler made with combinations of stockpiled BOF and MP1 (traditional filler) with the proportions of 100–0%, 75–25%, 50–50%, and 25–75% was used in asphalt concrete mixes and tested.

2. Research Scope

The aim of this research study was to characterize different types of BOF slags and conventional river sand material as aggregates and then evaluate them in mortar application. BOF slags have different ages. As shown in Figure 1, this aim was achieved by a two-phase, comprehensive laboratory evaluation: (i) basic aggregate performance characterization and (ii) mortar application evaluation.

The other aim of this research study was to analyze physical–mechanical properties of AC mixes with BOF slag filler and its combinations with MP1 (mineral powder MP-1 is obtained by grinding mineral rocks, such as limestone, marble, dolomite, and other sedimentary rocks). The tested parameters were the residual porosity of AC, %; water saturation, % by volume; ultimate compressive strength at a temperature of 20 °C, MPa; ultimate compressive strength at a temperature of 50 °C, MPa; ultimate compressive strength at a temperature of 0 °C, MPa; shear resistance according to the coefficient of internal friction at a temperature of 50 °C, MPa; shear resistance by the cohesion in shear at a temperature of 50 °C, MPa; Crack resistance according to the tensile strength at break up (crack) at a temperature of 0 °C, MPa; water resistance, % (30–60 min) as stated in ST RK 1225. The definitions of these parameters are explained below.

Residual Porosity (%): The residual porosity indicates the amount of void space within the compacted asphalt mixture. This parameter is important because excessive porosity can lead to reduced durability and increased vulnerability to moisture infiltration, which can cause premature cracking and degradation of the pavement.

Water Saturation (% by volume): Water saturation measurement helps assess the potential for moisture damage within the asphalt mixture. High water saturation levels can weaken the asphalt–aggregate bond and lead to stripping, rutting, and other forms of distress, compromising the longevity of the pavement.

Ultimate Compressive Strength: Compressive strength measures the ability of the asphalt mixture to withstand load-bearing pressures. Different temperature conditions (20 °C, 50 °C, 0 °C) simulate real-world temperature variations. This information is essential for designing pavements that can handle various traffic and environmental conditions without deformation or structural failure.

Shear Resistance: Shear resistance parameters (coefficient of internal friction and cohesion in shear) assess the asphalt mixture’s ability to withstand lateral forces and shear stresses. These properties are important for preventing pavement deformation, rutting, and shear-related failures under traffic loads.

Crack Resistance: Tensile Strength at Break (0 °C, MPa): This parameter evaluates the asphalt mixture’s ability to resist cracking, especially at lower temperatures. Cold-temperature cracking is a significant concern, and measuring the tensile strength at low temperatures helps predict the mixture’s behavior in cold climates.

Water Resistance (%): Water resistance testing assesses how well the asphalt mixture repels water and resists moisture-related distress. Proper water resistance is crucial for preventing moisture-induced damage, including cracking, stripping, and weakening of the asphalt–aggregate bond.

3. Experimental Program

3.1. Materials

The procedure for sampling basic oxygen furnace slag (BOFS) was derived from ASTM C1420. This method is established in advance of a sample collection. Prior to selecting samples, a visual examination of the BOFS was conducted to assess its condition and to determine the appropriate methodology and equipment required for the effective collection of the stockpiled BOFS. Chemical analyses of the fresh and stockpiled BOFS from the plant revealed variations in their elemental oxide weight percentages, depending on the production period. Consequently, the choice of sample locations must be representative of the stockpile while also taking into account the critical lifespan of the pile (as depicted in Figure 2). Factors considered in the sampling process include the location within the stockpile, the outward appearance, variations in color, and areas with similar levels of weathering.

ASTM C1420 outlines three sampling techniques: random sampling, location-specific sampling, and condition-specific sampling. In this context, location-specific sampling was chosen because it provides the most representative samples. Each sample container (batch) was assigned a sample label containing essential information, including the sample code, container number, date, facility location, and any relevant remarks.

An evaluation of the various sample extraction methods was conducted. Given the apparent hardness of the stockpiled BOFS, it was determined that the extraction procedures needed to be carefully considered. Sample collection was carried out using an excavator bucket (front-end loader) for simple digging. Plastic bags were then filled with BOFS collected from different locations within the pile, each representing different aging periods.

Each sample was extracted from a depth of approximately 1 m. The estimated age of the stockpiled BOFS ranged from 1 to 40 years. Fresh BOFS samples, on the other hand, were obtained directly from the plant, specifically from the beta interducts, and were estimated to be up to 1 month old.

Another type of BOF used in this research was a lab-harvested or processed BOF, which is a CO₂-exposed BOF slag with an age of 3 months. BOF aggregates were aged in full water submersion. The aging period was 2.5 months. For the “dry cycle” aging (weathering) conditions of BOF slags, the carbon dioxide content in the air was approximately 330~450 ppm (Figure 3).

For ordinary materials, river sand was chosen for comparison with fine BOF aggregates. Pre-determined reactive siliceous river sand was used in expansion tests for assessing its mortar application. The specific gravity and absorption capacity of the river sand were 2.68 and 2.87%, respectively.

For the preparation of asphalt concrete mixes, the traditional asphalt mix of type B (the mineral aggregate part makes up 40–50% of the AC mass) with a maximum gravel aggregate size of 20 mm, bitumen content of 5.2%, and mineral filler of 7% that was prepared according to ST RK 1225 and tested according to ST RK 1218 was mixed with stockpiled BOF filler with proportions of 100, 75, 50, and 25% with ordinary MP1 powder.

Activated mineral powder brand MP1 from limestone was produced by “Zhartas-SN” in accordance with GOST 16557. This powder is used in road construction as a component of asphalt concrete and other organo-mineral mixtures (filler in the construction of roads and airfield runways). Physical–mechanical properties of BOF filler and traditional MP1 filler are defined as well in accordance with ST RK 1221. The produced stockpiled BOF filler is shown in Figure 4, and the process of mixing two types of fillers (MP1 + BOF) in different proportions is shown in Figure 5.

Also, in this study, “DAD-KT” thermostable adhesive additive to road bitumen based on polyphosphoric acid esters was used in the asphalt concrete mix with BOF slag filler (Figure 6). DAD-KT is used in road construction to improve the adhesion between oil road bitumen and aggregate materials (both acidic and basic rocks), while also retaining the improved adhesive properties of bitumen from 7 to 10 days. The addition of this road adhesive additive into bitumen organic binders (from 0.1% to 1.2%) makes it possible to significantly improve the quality of pavement. It has a positive effect on the durability of asphalt concrete pavement by preventing peeling and spalling.

3.2. Material Testing and Analysis Methods

The experimental program encompassed a series of tests, including the determination of the oxide composition and examination of the particle mineralogy and morphology of BOF slags through XRD, XRF, and grain-size analysis. Additional tests involved the assessment of aggregate properties, such as absorption, specific gravity, moisture content, soundness (sulfate attack and freeze and thaw tests), aggregate crushing value, and abrasion resistance. Moreover, an evaluation of the expansion potential of mortar bars containing BOF slag and river sand (at 14 and 28 days) was conducted to ascertain both the fresh and hardened properties of these mortar bars.

As previously detailed, the research team had the opportunity to analyze the oxide composition of BOF slags using the Axios mAX X-ray fluorescence (XRF) spectrometer from PANalytical. This machine employs the SST-mAX X-ray tube, featuring close coupling and ZETA Technology. For both the X-ray fluorescence (XRF) and X-ray diffraction (XRD) analyses, 45-micron-sized samples of BOF slags were prepared using a Proctor compaction apparatus. X-ray diffraction (XRD) analysis with Cu Kα radiation (utilizing the Rigaku SmartLab System) was employed, with a scan range set at 5–70° and a continuous scan type with a sampling interval of 0.03°, allowing for the identification of the primary mineral constituents.

The grain-size distributions of BOF steel slags and river sand conformed to ASTM 2007c, ASTM E 105, and ASTM E 122.

The aggregate absorption, specific gravity, and moisture content of the BOF slag samples were determined within 3 days in accordance with ASTM C 127, ASTM C 128, and ASTM C556. The specific gravity of the BOF slags is expressed as the bulk specific gravity, bulk specific gravity (SSD) (saturated surface-dry), or the apparent specific gravity of the BOF samples. The specific gravity and absorption determination of the fine BOF aggregates were performed using the cone test.

The aggregate soundness test followed the guidelines in ASTM C 88. Determining the resistance to freezing and thawing of the BOF slag samples involved the preparation of a 3% NaCl solution in accordance with LS-614. Both coarse and fine BOF samples were prepared similarly to the soundness test procedure. The BOF samples were placed in containers, and a 3% NaCl solution was added before subjecting them to environmental testing within the chamber at Nazarbayev University’s Geotechnical laboratory.

The BOF aggregate crushing value was determined following the guidelines of the BS 812-110. Samples of BOF slag retained on a 9.5 mm sieve were selected for the experiment and manually compacted within a cylinder mold (as shown in Figure 7).

An expansion test was conducted in accordance with ASTM D 4792. BOF slag aggregates were compacted according to the D 1883 (Figure 8).

Then, the samples in the CBR molds were placed inside a hot water bath at a temperature of 70 ± 3 °C for a period of 14 days. During these 14 days, the volume expansion levels of the compacted specimens inside the water were recorded 4 times a day (Figure 8).

An accelerated mortar bar expansion test was performed in accordance with ASTM C 1260. BOF and river sand-added mortar bar specimens were placed in sealed containers filled with 1 M NaOH solution at 80 °C temperature, and also in sealed containers with water at an 80 °C temperature (Figure 9). The expansions of these mortar bars in two conditions were observed within 14 and 28 days.

AC mix samples with coarse and fine aggregates of gravel and mineral filler of a combination of BOF slag—MP1 traditional powder were prepared (Figure 10).

The tested parameters were the residual porosity of AC, %; water saturation, % by volume; ultimate compressive strength at a temperature of 20 °C, MPa; ultimate compressive strength at a temperature of 50 °C, MPa; ultimate compressive strength at a temperature of 0 °C, MPa; shear resistance according to the coefficient of internal friction at a temperature of 50 °C, MPa; shear resistance by the cohesion in shear at a temperature of 50 °C, MPa; crack resistance according to the tensile strength at break up (crack) at a temperature of 0 °C, MPa (Figure 11); water resistance, % (30–60 min) in accordance with ST RK 1225 and ST RK 1218.

4. Results

4.1. Aggregate Characterization

The bulk chemical analyses of the fresh and stockpiled BOF slags are reported in

Table 1. Bulk chemical composition of fresh and stockpiled BOF slags.

Composition	Fresh BOF	Stockpiled BOF
MgO	1.75	1.31
Al2O3	0.51	1.69
SiO2	5.88	7.72
P2O5	1.31	1.73
SO3	0.21	0.3
K2O	0.09	0.05
Na2O	-	-
CaO	47.40	47.76
TiO2	0.75	0.69
MnO	4.59	2.82
Fe2O3	35.46	34.44
CuO	0.05	0.04
ZnO	0.47	0.19
SrO	0.02	0.02
BaO	0.03	-
PbO	0.03	0.02
Au2O3	-	0.08
Bi2O3	-	-
ClO2	0.11	0.09
Cr2O3	0.33	0.31
Nb2O5	0.01	0.01
V2O5	0.97	0.65
PtO2	0.02	-
NiO		0.02
CoO	-	0.04
MoO3	-	-
ZrO2	-	0.01
Ga2O3	-	0.01
Sum	100.00	100.00

. Variation in the chemical composition of both the fresh and stockpiled BOFS over a 3-month period is evident. Lower SiO₂, Al₂O₃, Fe₂O₃, and SO₃ and higher CaO and MgO generally characterize the fresh BOFS compared to those in the stockpiled BOFS.

From the X-ray powder diffraction (XRD) analysis, the mineralogy of the fresh and stockpiled BOF slags was determined, and the results are presented in Figure 12. The main difference between the fresh and stockpiled BOFS was in the abundance of free lime (CaO) and calcite (CaCO₃) and a small amount of portlandite (CaOH)₂) in the fresh BOFS, whereas the stockpiled BOFS contained a high amount of portlandite instead of free lime. Both the fresh and stockpiled BOFS contained relatively high amounts of quartz (SiO₂), wustite (FeO), magnetite (Fe₃O₄), free lime, and portlandite, and minor amounts of calcium silicate (Ca₂SiO₄) and srebrodolskite (Ca₂Fe₂O₅).

Amorphous content existed in both the fresh and stockpiled BOFS. A higher amount of portlandite and calcite in both types of BOFS can be explained by the following chemical reaction.

CaO_(S) + CO_2(g) -> CaCO_3(S) ∆H = −178.3 kJ/mol CO₂

(1)

Ca(OH)_2(S) + CO_2(g) -> CaCO_3(S) + H₂O_(l) ∆H_r = −113.1 kJ/mol CO₂

(2)

Figure 13 displays various particle sizes (with mesh sizes) of stockpiled BOF slags that were sifted following the ASTM C 136 guidelines. Regardless of the specific layer within the pile or the age of the BOF slags, these materials exhibited an angular shape and possessed a rough texture. Consequently, when BOF slags are employed as aggregates in applications such as Portland cement concrete, hot mix asphalt concrete, or base courses, they have the potential to enhance their stability and strength through aggregate interlocking. However, it is important to note that their angular and rough-textured nature may result in reduced workability compared to rounded and smooth-textured aggregates, making it more challenging for them to move smoothly past each other.

Figure 14 illustrates BOF slags that have undergone an extended aging process, resulting in weathered and aged products with a distinctive light brown coloration. As previously mentioned, the principal crystalline phases found in pure BOF slags include wüstite, magnetite, free lime, larnite, and brownmillerite. While calcium silicate (Ca₂SiO₄) within BOF slags can impart some hydraulic properties, the presence of other inert phases and free lime, along with magnesium, makes BOF slags challenging to use as aggregates due to their tendency to undergo volume instability. However, the carbonation of BOF slags can lead to the transformation of calcium silicate and free lime into calcite, addressing the issue of volumetric instability.

The carbonation reaction that led to the aged products happened due to Equations (1)–(3):

Ca₂SiO₄ + CO₂ + H₂O -> CaCO₃ + SiO₂ + H₂O

(3)

Table 2. Moisture content, absorption capacity, and specific gravity of coarse BOF slags.

Material	Soundness Test (%)	Freeze and Thaw Test (%)	Moisture Content (%)	Specific Gravity Gsb-od	Specific Gravity Gsb-SSD	Specific Gravity Gsa	Absorption (%)
2nd Batch Fresh BOF	3.01	1.49	3.07	3.28	3.33	3.45	1.51
2nd Batch Stockp. BOF	0.63	1.46	4.02	3.50	3.54	3.65	1.14
1st Batch Fresh BOF	1.36	4.52	2.68	2.83	2.95	3.23	4.43
1st Batch Stockp. BOF	1.37	5.25	3.94	3.14	3.2	3.32	1.78

and

Table 3. Moisture content, absorption capacity, and specific gravity of fine BOF slags and sand.

Material	Soundness Test (%)	Freeze and Thaw Test (%)	Moisture Content (%)	Specific Gravity Gsb-od	Specific Gravity Gsb-SSD	Specific Gravity Gsa	Absorption (%)
2nd Batch Fresh BOF	3.03	2.97	3.07	3.26	3.42	3.87	4.82
2nd Batch Stockp. BOF	4.14	4.25	4.02	3.16	3.32	3.78	5.17
1st Batch Fresh BOF	1.46	3.40	2.68	3.36	3.47	3.77	3.24
1st Batch Stockp. BOF	1.65	11.62	3.94	3.02	3.11	3.34	3.17
River sand			0.23	2.58	2.66	2.8	3.02

present data on the moisture content, absorption capacity, and specific gravity of both coarse and fine BOFS. Regarding the coarse BOFS, it was observed that the stockpiled BOFS exhibited a higher overall moisture content and specific gravity when compared to fresh BOFS, but it showed a lower capacity to absorb water. It is essential to highlight that the stockpiled BOFS had undergone a weathering process, resulting in the presence of more aged products, including intense hydration and carbonation products, as well as a higher in situ moisture content. Additionally, the fresh coarse BOFS was more porous than stockpiled BOFS, leading to its higher water absorption capacity.

In contrast, when examining fine BOFS, it is evident that the stockpiled BOFS had a lower total moisture content and specific gravity, but it exhibited a higher water absorption capacity compared to the fresh BOFS. This difference can be attributed to the fact that the products generated through the weathering (aging) process had a higher specific surface area, were lightweight, and included substances like portlandite and calcite. Both the fine fresh and stockpiled BOFS, however, possessed a higher specific gravity than the river sand.

The tables above also provide data on the soundness properties, including sulfate and freeze–thaw resistance. According to ASTM C 88, an aggregate is considered to have good sulfate resistance if it exhibits less than an 18% total loss before and after exposure to sodium sulfate (Na₂SO₄). Additionally, Method of Test for Freezing and Thawing (F-T) of Aggregate suggests that an aggregate with excellent freeze–thaw (F-T) resistance typically experiences mass loss ranging from 4% to 18% before and after the test.

It is worth noting that the fine stockpiled BOF slag demonstrated excellent sulfate and freeze–thaw resistance. However, it is important to highlight that it also possessed a higher absorption capacity and moisture content when compared to river sand. In general, it can be seen that BOF slags may vary from batch to batch. Several factors can contribute to this batch-to-batch variability, including:

• The composition of the raw materials used in the BOF process can fluctuate. Variations in the composition of iron ore, scrap metal, and other inputs can lead to differences in the resulting slag.
• Factors like temperature, pressure, and reaction kinetics during the BOF process can impact the characteristics of the slag. Variations in these conditions from one production run to another can result in different slag properties.
• The cooling rate and post-production processing of the slag can influence its structure and properties. Different cooling rates or processing methods can lead to variations in the final product.
• Complex chemical reactions can occur during the formation of slag. These reactions may be influenced by impurities, additives, or other factors, leading to batch-to-batch differences.
• Environmental conditions, such as humidity and exposure to external elements, can affect the properties of slag materials, especially when they are stockpiled or stored for extended periods.
• Quality control measures and practices within the manufacturing or processing facility can impact the consistency of slag products. Variations in quality control procedures can result in batch-to-batch differences.

The aggregate crushing value (ACV) indicates satisfactory resistance to crushing under the roller during construction and adequate resistance to surface abrasion under traffic. It was found that the ACV in the fresh BOFS was equal to 32.7, whereas the ACV in the stockpiled BOFS was equal to 30.9. However, it is suggested that fresh coarse BOFS is more porous than other aggregates, which is supported by its high water absorption capacity. As previously stated, the weathering (aging) process converts hydrated calcium silicate, free lime, and calcium hydroxide (Ca(OH)₂) into calcium carbonate (CaCO₃) by reaction with CO₂. The formed calcite is stronger than Ca(OH)₂ and fills out the inherent voids in fresh BOFS. As a result, the stockpiled BOFS has higher strength and denser structure than the fresh BOFS.

The Los Angeles aggregate abrasion value indicates the toughness and durability of aggregate subjected to impact and abrasion (

Table 4. LA aggregate abrasion results of BOF slags.

BOF Type	Percent Loss (%)
Fresh BOFS	38.7
Stockpiled BOFS	26

). This value can provide insight into how aggregates will stand up to wear and tear over time when the aggregate is used for each construction application. Aggregates with low LA abrasion values are stronger and more resistant to abrasion.

As can be seen, fresh BOFS has the highest abrasion value (percent loss) than other stockpiled BOFS. Thus, stockpiled BOFS has better toughness and abrasive properties. As stated earlier, fresh BOFS is more porous than other aggregates, which is supported by its high water absorption capacity. Also, the weathering (aging) process and hydration in fresh BOFS produce CaCO₃, which fills out the inherent voids in fresh BOFS and produces calcium silicate hydrate (C-S-H). This result matches the aggregate crushing value result.

This finding aligns with the observations of reduced porosity, improved density, and strength resulting from the aging process and the formation of CaCO₃. Therefore, stockpiled BOF slags are likely to be a more suitable option for construction applications where resistance to abrasion is a critical factor.

4.2. Mortar Application of BOF Slags

Figure 15 illustrates the expansion characteristics of mortar bars made of fresh and stockpiled BOFS submerged in water under a controlled temperature of 80 °C. Pre-determined reactive siliceous river sand was used for a comparative study in the expansion test in both water and 1 M NaOH solution. The expansion of the mixture containing river sand at 80 °C water was 0.01% at both 14 and 28 days, respectively. This river sand is innocuous in normal conditions, and not in a high-alkali environment. However, the mortar bars containing fresh and stockpiled BOFS had higher expansions at 14 and 28 days. For example, the mortar bars containing fresh BOFS exceeded the threshold value of 0.1% based on ASTM C 1260/C 1567 (14-day expansion) and were broken even before 14 days. The expansion of the stockpiled BOFS mortar was below 0.1% at 14 days. However, when the expansion period was extended to 28 days, the expansion of all BOF specimens also exceeded the threshold value of 0.1%, except the processed BOF (lab-harvested fresh BOF, which was water-submerged). This volumetric expansion was due to the formation of expansive portlandite minerals.

During the hydration process, the reaction of free calcium oxide (f-CaO) with water induced the formation of Ca(OH)₂ and CaCO₃, leading to excessive expansion and cracks in the mortar bar specimen (Figure 16).

Figure 17 presents the expansion characteristics of the mortar bars containing fresh and stockpiled BOFS, which were submerged in 1 M NaOH solution under a controlled temperature of 80 °C. As expected, the expansion of all mortar mixtures was beyond a limit value of 0.1% expansion at 14 days, indicating reactive and volumetric expansion field performance behaviors. In this case, both portlandite and alkali–silica reaction gel formations occurred together and resulted in more severe expansion. Again, the expansion of stockpiled BOFS was less than that of fresh BOFS. Therefore, when BOFS materials are used as an aggregate, it seems necessary to conduct an aging process that eliminates or minimizes deleterious expansion.

As stated previously, one of the critical challenges facing the utilization of BOFS as an aggregate in construction applications is its volume instability due to the formation of Ca(OH)₂ or magnesium hydroxide (Mg(OH)₂) caused by the reaction between f-CaO or f-MgO and water. Mg(OH)₂ crystalline products along with Ca(OH)₂ arise.

Figure 15 and Figure 17 present the expansion characteristics of mortar bars containing stockpiled BOFS. Interestingly, the expansion of all mortar mixtures was below a limit value of 0.1% expansion at both the 14- and 28-day periods, indicating innocuous field performance behavior. This result indicates that the formation of portlandite had already taken place in the stockpile. There was no more deleterious portlandite formation, or less formation occurred, leading to less expansion of the mortar bar. However, when all mortar mixtures were submersed in 1N NaOH solution, all samples exceeded a limit value of 0.1% expansion at 14 days, indicating reactive and volumetric expansion field performance behavior. This indicates that they exhibited reactive and volumetric expansion in this aggressive alkaline environment. This expansion was attributed to alkali–silica reaction (ASR), which is a chemical reaction between alkalis in the solution and reactive siliceous aggregates in the mortar or concrete. To minimize the expansion caused by ASR in these conditions, the text suggests the use of supplementary cementitious materials, such as fly ash or ground granulated blast furnace slag. These materials can help mitigate the expansion by reacting with the alkalis in the solution and reducing the potential for ASR.

Moreover, the potential expansion of BOF aggregates from hydration reaction by using compacted cylindrical specimens is demonstrated in Figure 18. As expected, the expansion of stockpiled BOFS is lower than that of fresh BOFS. This result matches the mortar bar expansion result. This difference in expansion is expected and may be attributed to the aging or curing process of BOFS.

Figure 18 demonstrates that when using compacted cylindrical specimens, the stockpiled BOFS showed lower expansion compared to the fresh BOFS, which is consistent with the findings from the mortar bar expansion tests. This consistency strengthens the evidence that stockpiled BOFS is less prone to expansion, which can be an important factor to consider in construction applications where the potential for expansion can be problematic.

4.3. Characterization of Traditional MP1 and BOF Fillers

The physical and mechanical properties of traditional MP1 filler were identified according to ST RK 1276 (

Table 5. Characterization of MP1 filler.

	Grain Composition % by Mass, Smaller < 1.25 mm	Grain composition % by Mass, Smaller < 0.315 mm	Grain Composition % by Mass, Smaller < 0.071 mm	Porosity, %	Bitumen Capacity Index, g	Swelling of a Sample from a Mixture of Mineral Powder with Bitumen, %	Humidity, % by Mass
Requirements of ST RK 1276 for MP1 Filler	>100	>90	>80	<28	<50	<1.5	<0.5
Experimental results of MP1 Filler	99.82	99.02	85.39	26.5	40.2	0.45	0.12

). It is made by crushing sedimentary rocks and bituminous rocks (limestones, dolomites and sandstones containing natural bitumen, sands and other soils bound by bitumen).

BOF filler is considered an MP3 powder from non-carbonate rocks and industrial waste (inactivated), as stated in ST RK 1276. Stockpiled BOF filler is produced by grinding crushed BOF slag in a ball mill. It was found that the stockpiled filler barely corresponded to the requirements of ST RK 1276 in comparison to the BOF filler made from the fresh BOFS.

4.4. Mineral Filler Application of BOF Slags

The testing of AC samples made of stockpiled BOF filler and MP1 (limestone) filler in the proportions of 100–0%, 75–25%, 50–50%, 25–75%, and 25–75% with 0.4% DAD-KT additive produced the following results of the tensile strength at break up (crack), as shown in Figure 19.

According to ST RK 1225, the crack resistance of AC mixes must meet the criteria of 3.5–6.5 MPa, which indicates that none of the mixtures corresponded to this range. From these values, it seems that the mixture with a 50–50 proportion of BOF filler and MP1 filler had the highest tensile strength at break (3.18), indicating better crack resistance compared to the other mixtures. However, it did not meet the crack resistance criteria currently outlined in the standard. Thus, in order to enhance the crack resistance of the asphalt concrete mixtures, considerations may need to be given to adjusting the mixture proportions, exploring alternative additives, or utilizing different types of fillers. Potential achievement of the crack resistance values specified within the range mentioned in ST RK 1225 could result from modifications to the mixture design.

The compressive strength values of 5 AC mixes are given in Figure 20, which shows that the mixture with 25% BOF slag filler and the other mixture of 25% BOF plus 0.4% added DAD-KT correspond to the criteria of compressive strength of more than 1.3 MPa, as indicated in ST RK 1225 (Figure 20). The mixture’s strength is influenced by factors such as the particle size distribution, reactivity, adhesion properties, and overall compatibility.

Moreover, it was found that the residual porosity values of the five mixes were 11.63%, 10.35%, 8.59%, 7.11%, and 5.0%, respectively, for 100–0%, 75–25%, 50–50%, 25–75%, and 25–75% with 0.4% of DAD-KT additive mixtures. The residual porosity was established by the test method that determines the level of pores in a compacted asphalt mixture and is calculated based on the average density (ρ_avg) and the true density (ρ_true). Only the last mixture (25% BOF filler with 0.4% DAD-KT additive) provided a barely suitable result of 5.0%, as the limits for this criteria are 2.5 to 5.0%, as stated in ST RK 1225. Also, poor cohesion was demonstrated for the mixes of 100–0% and 75–25% (BOF-MP1 filler), as they had cohesion of less than 0.38, whereas the standard dictates this value to be more than 0.38. A cohesion value below the prescribed threshold implies that the particles within the mixtures lack effective bonding, potentially leading to compromised pavement integrity and performance. Poor cohesion in mixtures with BOF slag filler (100–0% and 75–25%) could be due to the irregular shapes and angular nature of BOF slag particles. This can hinder effective interlocking and bonding between particles, resulting in weak cohesion.

However, other standard physical and mechanical properties of five AC mixes fully correspond to the criteria set in ST RK 1225. For example, in Figure 21 the compressive strength at 20 °C is given and all five mixtures have values of CS more than 2.5 MPa.

Also, all five AC mixes have adequate values of water saturation (1.5–4.0 MPa), compressive strength at 0 °C (less than 13 MPa), internal friction coefficient (tgᵩ), which is not less than 0.83, water resistance (not less than 0.85%). This implies that:

• Water saturation within the specified range indicates that these mixtures are resilient to moisture-related issues, which is crucial for the long-term durability of pavement structures;
• Compressive strength values below 13 MPa at 0 °C suggest that these mixtures can perform well under cold temperature conditions. This is important for preventing cold temperature-induced cracking;
• An internal friction coefficient above 0.83 indicates that these mixtures have good shear resistance and internal stability, which is essential for withstanding traffic loads without excessive deformation or rutting;
• A water resistance value not less than 0.85% suggests that these mixtures are effective at resisting the detrimental effects of water infiltration, which is critical for maintaining pavement integrity.

The findings underscore the importance of meticulous mixture design and the inclusion of appropriate additives. The mixture with 25% BOF filler and 0.4% DAD-KT additive emerged as the most promising due to its alignment with the residual porosity and cohesion criteria set by ST RK 1225. This emphasizes the need to balance the components effectively to achieve the desired properties and performance in asphalt concrete pavement.

However, addressing the challenges associated with BOF filler’s particle shape, compaction, and adhesion is crucial to ensure that the resulting mixtures meet or exceed performance standards and requirements.

The comprehensive assessment in this study indicates that while some mixtures meet the specified criteria for certain properties, there are challenges, particularly regarding crack resistance and cohesion, that need to be addressed to fully align with the standard. Adjustments in the mixture proportions, the exploration of alternative additives, or the use of different types of fillers may be necessary to achieve the desired properties, especially in terms of crack resistance and cohesion.

5. Conclusions

Before using basic oxygen furnace slag (BOFS) in any engineering application, it is important to determine its properties. The chemical composition, mineralogy, and physical properties of BOF slag are subject to large fluctuations as a result of different raw additives, different compositions of the molten iron used for producing the steel, oxygen stirring of a molten pool, uneven temperature fields, and other complex physical conditions. Moreover, the expansion characteristics of BOFS due to these inherent high free CaO and MgO contents in the presence of water and the oxidation of the metallic content should be coped with because its volume instability is a considerably unsafe factor and limits the utilization of BOF slag in transportation engineering applications. Based on all the test results, the following results were obtained:

• Stockpiled BOFS was harder than fresh BOF slag, which had lower aggregate crushing values and lower LA abrasion values. The stockpiled BOFS showed less expansion than the fresh BOF slag regardless of water and 1 M NaOH solution immersion.
• The chemical reaction between f-CaO, f-MgO, and water due to weathering in the field extremely reduced the expansion of BOFS submerged in water. BOFS may contain reactive silica, which causes an alkali–silica reaction.
• Indeed, it is evident that basic oxygen furnace (BOF) slags can exhibit variability from one batch to another due to raw material compositions, process conditions, cooling and post-production processing, chemical reaction and environmental conditions, and quality control practices.
• The mixture with 25% BOF filler and 0.4% DAD-KT additive emerged as the most promising due to its alignment with the residual porosity and cohesion criteria set by ST RK 1225. Adjustments to the mixture proportions, the exploration of alternative additives, or the use of different types of fillers may be necessary to achieve the desired properties, especially in terms of crack resistance and cohesion.