Sustainable concrete production: Partial aggregate replacement with electric arc furnace slag

2.1 Effects of EAF steel slag on the properties of fresh concrete

Certain aggregate properties of the EAF steel slag should also be comparable to those of natural aggregates used in concrete. Additional research examined the water absorption of EAF steel slag aggregate where it was found to have less than a 5% greater water absorption [10]. In Santamaría et al. [11], the authors evaluated the hardened densities of each mixture as another parameter to consider and explained the importance of ensuring correlation to their fresh densities.

In a study conducted by Maharaj et al. [12]; the fresh densities of concrete with 30–100% replacement of coarse aggregate with EAF steel slag were 2,434–2,651 kg/m³, both surpassing the control sample’s fresh density of 2,355 kg/m³. Additionally, it was found that slump height decreased with the increasing content of slag. Similarly, studies [13,14] found that increasing EAF steel slag content when replacing fine aggregate decreased the workability, especially after 50% replacement. The writers of both [13,14] recommended the use of admixtures to decrease the negative effect. These fresh concrete properties are the first steps to any test involving the formation of a concrete mixture.

2.2 Methods of incorporating EAF steel slag into concrete

There are different techniques for incorporating EAF steel slag into concrete as shown by previous research which include the following: partial or total replacement of fine natural aggregates, coarse natural aggregates, specified sizes of coarse aggregates, or both fine and coarse aggregates. Qasrawi et al. [13] investigated replacing fine aggregate with EAF steel slag in concrete with design strengths of 25–45 MPa, where they found that the mechanical properties had the best performance at a 50% replacement. Olonade et al. [15] examined the partial and total replacement of fine aggregate with EAF steel slag for 20 MPa strength concrete. It was found that 50% replacement was optimum and had similar findings on the workability. However, the studies [12,14] conducted investigations where they found 20 and 30 MPa strength concrete, respectively. They both found that the optimum replacement ratio was 30% after replacing fines with EAF steel slag in 10% increments from 0 to 50%. Furthermore, Borole et al. [16] also replaced fines for 30 MPa strength concrete and closed by stating their optimum replacement ratio was 25%. Despite producing the same concrete grade, the optimum replacement ratios of slag varied. This was due to the inconsistent characteristics of EAF steel slag, justifying the need to perform these tests locally.

To continue the discussion of implementation strategies, there is literature that explored coarse aggregate replacement with EAF steel slag. Similar to the studies that explored fine aggregate substitution, a study by Maharaj et al. [12] aimed to identify an optimum replacement ratio for the fresh and hardened properties of concrete. The authors found that there was not a specific ratio that allowed all properties to produce optimal results. Diotti et al. [17] realized this and used a 30% replacement ratio to obtain satisfactory physical and mechanical performances of their concrete. Moreover, the engineering properties of concrete were found to benefit from total replacement by EAF steel slag. Literature [18] reported this, as did Maharaj et al. [12] later on. The former authors though, claimed that establishing a controlled concrete grade for analysis would permit more precise conclusions. Despite the accurate results obtained from designing a specific concrete grade, the researcher focused more on the change in strength than on aiming for a desired value. Instead, a standard mix design was chosen based on the available resources.

The final incorporation method examined by existing literature involved replacing portions of both fine and coarse natural aggregate with slag. In [19], it was deduced that the mechanical performance of concrete was commendable for the partial substitution of both types of natural aggregates with EAF slag. Other literature followed with a similar study, examining the properties of concrete when its fine aggregate, coarse aggregate, and total aggregate, are replaced by EAF slag. They determined that it would be worthwhile to replace either fine or coarse aggregate with slag, but replacing both causes detrimental effects on concrete strength [20]. Desirable conclusions may result from a study that incorporates partial replacement of both fine and coarse natural aggregate.

2.3 Effects of EAF steel slag on the mechanical properties of hardened concrete

As a building material, the mechanical properties of concrete must meet specified requirements since these properties are what influence the durability and structural integrity of any structure made from it. This study investigates the plausibility of altering the composition of concrete mixtures with a new material as aggregate. Thus, a thorough evaluation of the effects of this substitute material on the engineering characteristics of concrete is critical. The mechanical properties emphasized are compressive strength and tensile strength. A comprehensive analysis of literature discussing the results of EAF steel slag concrete properties follows.

Compressive and tensile strength tests are not optional when assessing the mechanical properties of concrete, therefore they were evaluated in all pertinent sources. In Qasrawi et al. [13], it was observed that depending on the replacement ratio and concrete grade, incorporating EAF slag as fine aggregates can cause a 1.1–1.3 times improvement in 28-day compressive strength, and a 1.4–2.4 times improvement in 28-day tensile strength in comparison with traditional concrete. Research papers [7,14–16] conducted subsequent fine aggregate replacement studies which yielded comparable findings to the increase in strengths with similar reasoning. In addition, marginal improvements in compressive strength were found more so than in tensile strength in studies by the authors of [12,17] with incremental coarse aggregate replacement. These outcomes imply that replacement ratios containing a blend of both fine and coarse aggregate can be an effective replacement method, and that there is an optimum coarse aggregate size rather than a relationship between aggregate size and mechanical properties.

2.4 Concrete mix design optimization to include EAF steel slag

The emphasis of this study is to produce an advanced form of concrete, so it is critical to establish the necessary parameters to obtain its desired characteristics. One such parameter is the gradation of aggregate sizes. The writers of Qasrawi et al. [13] outlined the necessity of grading aggregates for concrete production to obtain desirable characteristics of the concrete, such as workability and mechanical strength.

Authors from [21,22] found that larger aggregate sizes resulted in stiffer and more unworkable concrete mixtures. Further, studies have shown that an irregular distribution of coarse aggregates in concrete mixtures can significantly impact its compressive strength. The researchers in [23] tested and compared the compressive strengths of concrete mixtures made with 10 and 20 mm coarse aggregates. They found that the larger aggregate size had a 45.7% greater compressive strength. A similar comparative experiment was conducted in [24] where it was concluded that compressive strength increases with increasing coarse aggregate size from 6 to 20 mm but drops with 25 mm particles. This showed that while an excess of coarse aggregate can increase strength, a lack of the same can produce concrete that displays weaker strengths. The aim of this paper is to examine the effects of concrete strength with the incorporation of slag aggregates into the mixture. Thus, aggregate grading must be done to eliminate other factors affecting concrete mixture properties and hardened concrete strength.

In addition, the water to cement (w/c) ratio is another important parameter for mix design. Researchers investigated the compressive and tensile strengths of EAF steel slag aggregate concrete produced in Trinidad [12]. The water to cement ratios used were 0.4, 0.45, and 0.5 and discovered that the concrete with a w/c ratio of 0.4 had the best performance in the compressive strength test, with a value of 80.5 MPa. This research project did not involve assessing the effect of the w/c ratio on the mechanical properties of EAF steel slag concrete but used steel slag from similar sources. Thus, a w/c ratio of 0.4 was considered.

2.5 Identification of gaps and future research

The aforementioned discussion of EAF steel slag explained that its physical and chemical properties depended on the raw material used to produce the steel. The diverse locations of the studies, as well as the vast mix design options for concrete production, meant an exact optimum ratio could not be decided. It also meant that the specific steel-making material and concrete grade influences the relationship between replacement ratios and their effect on the mechanical properties. Furthermore, EAF steel slag is known to be susceptible to physical and chemical changes when exposed to different environmental conditions. Due to the storing conditions being unknown, it was also impossible to know the exact influence that the EAF slag would have had on the properties of the concrete. Proper evaluation of these effects can be done if the EAF slag produced from the same plant is used. Generalized conclusions for each grade can only be drawn if the EAF steel slag used complies with some international standards for manufacturing and storing. The insufficiency of the local steel plant, Arcelor Mittal Company, to provide such information posed limitations in this dissertation. It is also noted that this company has not been in operation for some years and any existing EAF steel slag stored was expected to be in unattended and unmaintained conditions.

Consequently, to incorporate EAF steel slag into concrete, the articles reviewed focused on the partial replacement of fine aggregates, or coarse aggregates. The few studies that replaced both aggregate types did it in individual mixtures, despite there being countless existing literature supporting that method. This research perceives the benefits of replacing both aggregate types to indicate that replacing a blend of both would also produce concrete with satisfactory mechanical performance. For example, EAF steel slag can be added in volumes of 15, 30, 50, and 100%, with each percentage consisting of a specified amount of fine and coarse aggregate.

3.1 Materials

The materials used in the concrete mixtures were blended cement, water, natural aggregates, and EAF steel slag used as both fine and coarse aggregates, which were all locally sourced. The blended cement used was from Trinidad Cement Limited. The natural coarse and fine aggregates used were quartzite obtained from the quarry in Valencia, Trinidad, and the EAF slag was acquired from stockpiles at Arcelor Mittal, in Point Lisas, Trinidad. The EAF steel slag was used as both fine and coarse aggregates.

3.2 Specimen preparation procedure

3.2.1 Sieve analysis of natural aggregate and EAF steel slag aggregate

The natural coarse and fine aggregates were graded according to the ASTM C33 standard to ensure a suitable size for concrete production. The findings from the sieve analysis conducted on the fine and coarse slag aggregates are illustrated in Tables 1 and 2 respectively. Figure 1 illustrates the gradation curves of the aggregates.

Table 1

Particle size distribution for fine EAF slag aggregate

Sieve size (mm)	Individual fine NA weight retained (g)	Percentage of fine NA weight retained (%)	Required mass of EAF slag from 13 kg (kg)
10	6	0.30	0.04
5	31	1.55	0.20
2.36	296	14.80	1.92
1.18	310	15.50	2.02
0.6	317	15.85	2.06
0.3	553	27.65	3.59
0.15	397	19.85	2.58
0.075	89	4.45	0.58
TOTAL	1,999	99.95	12.99

Table 2

Particle size distribution for coarse EAF slag aggregate

Sieve size (mm)	Individual coarse NA weight retained (g)	Percentage of coarse NA weight retained (%)	Required mass of EAF from 20 kg (kg)
25	0	0.00	0.00
20	266	13.30	2.66
10	1,190	59.50	11.90
5	163	8.15	1.63
2.36	85	4.25	0.85
PAN	296	14.80	2.96
TOTAL	2,000	100	20.00

Figure 1

Grading of natural and EAF slag aggregates.

Figure 1 represents the grading of the fine and coarse natural aggregates and fine, coarse, and stockpile steel slag aggregates with reference to the upper and lower limits as defined by the ASTM C33 standards. The fine natural aggregates met the grading requirements since it fell within the range of the upper and lower limits. However, the coarse natural aggregates fell slightly out of the boundary for the 10 and 5 mm sieves, but the grading curve was still deemed acceptable.

The grading curve of the slag aggregates directly from the stockpile fell between the upper coarse limit and the lower fine limit. It showed that 59.2% of the stockpile aggregate was 5 mm, so the stockpile slag had more fine than coarse aggregates. It was therefore separated into the relevant particle sizes and graded as coarse and fine aggregates accordingly.

Grading the steel slag aggregate was recommended by Pellegrino and Gaddo [19], where the writers explained the importance of aggregate grading to produce desirable workability and mechanical properties. The grading curves of the fine and coarse slag aggregates can also be seen in Figure 1 both falling very closely within their respective upper and lower limits.

3.3 Specific gravity, particle water absorption, and aggregate density of natural aggregate and EAF slag aggregate

Specific gravity, particle water absorption, and aggregate density tests were performed on the fine and coarse natural and slag aggregates. These tests were conducted in accordance with the ASTM C127-15 standards. The aggregate impact value (AIV) test was performed in accordance with ASTM D5874 testing procedure.

3.3.1 Specific gravity

Table 3 and Figure 2 illustrate the specific gravity values of the natural aggregates and the EAF slag aggregates. According to Table 3 and Figure 2, the specific gravity values of the slag aggregates are greater than that of the natural aggregates for both coarse and fine. The oven dried (OD), saturated surface dried (SSD), and apparent specific gravity (ARD) of the slag aggregates were overall greater than that of the natural aggregates. These results match the trends of the values obtained by testing from Maharaj et al. [12]. It can be inferred from these findings that the slag aggregates are stronger than the natural aggregates and should produce a higher-strength concrete.

Table 3

Specific gravity of natural and EAF slag aggregates

Aggregate property	Coarse natural aggregate	Coarse EAF slag	Fine natural aggregates	Fine EAF slag aggregates
Specific gravity (OD)	2.59	3.68	2.59	2.82
Specific gravity (SSD)	2.60	3.71	2.62	2.92
Specific gravity (ARD)	2.63	3.80	2.65	3.12

Figure 2

Specific gravity of natural and EAFSS aggregates.

3.3.2 Water absorption

Table 4 and Figure 3 display the water absorption of each aggregate type and material, where the fine slag aggregates are shown to have the highest water absorption of 3.39%, which is 4.2 times greater than the fine natural aggregates. The coarse slag aggregates had a 1.15 times higher water absorption than coarse natural aggregates. Research in San-José et al. [10] found that the water absorption of slag aggregates was greater than that of natural aggregates but by less than 5%.

Table 4

Water absorption of natural and EAF slag aggregates

Aggregate property	Coarse NA	Coarse EAFSS	Fine NA	Fine EAFSS
Water absorption (%)	0.55	0.84	0.81	3.39

Figure 3

Water absorption of natural and EAFSS aggregates.

The water absorption of aggregate depends on its chemical makeup. Yildirim and Prezzi proposed that the chemical composition of steel slag aggregates depends on the raw materials used to make the steel that it is made from, which can differ for every country and every steel plant [21]. Thus, the difference in results is likely due to the different chemical properties of the slag aggregate used by the authors and that which was sourced locally. This idea was supported by the findings of Maharaj et al. [12] since the steel slag utilized in the research was sourced locally and had a 4.5 times greater water absorption than the natural aggregates.

The steel slag aggregates have an overall higher water absorption than the natural aggregates; however, this implies that these aggregates are able to remove the water in the concrete mixture needed for proper hydration of the cement. The water-to-cement ratio of concrete does not account for any water that is absorbed by aggregates; it is only enough for proper hydration of the cement matrix. Therefore, it can be deduced that increasing steel slag aggregate content may eventually cause the strength of the concrete to decrease.

3.3.3 Aggregate density

Table 5 and Figure 4 show the aggregate densities of each type and classification of aggregate used in the experiment. The slag aggregates are shown to have an overall higher density than the natural aggregates. This suggests that it would require a smaller amount of the steel slag aggregate to fill a given volume than it would for natural aggregates.

Table 5

Aggregate density of natural and EAFSS aggregates

Aggregate property	Coarse NA	Coarse EAFSS	Fine NA	Fine EAFSS
Density (OD) (kg/m3)	2,583	3,670	2,588	2,816

Figure 4

Aggregate density of natural and EAFSS aggregates.

The higher density of the steel slag aggregates infers that it is stronger than the natural aggregates and could create stronger concrete. From this, it can be deduced that increasing the amount of steel slag aggregate could increase the strength of the concrete specimens.

3.3.4 AIV

Table 6 and Figure 5 illustrate that steel slag aggregates have a lower AIV value than natural aggregates, with values of 30.52 and 37.44%, respectively. A lower value means that a smaller proportion of the steel slag aggregates was fractured during the test, so it is stronger than the natural aggregate. From this, it can be deduced that the slag aggregates could produce stronger concrete specimens than natural aggregate.

Table 6

Aggregate impact test

3/8″ EAFSS aggregates			3/8″ NA
Mass	Test 1	Test 2	Mass	Test 1	Test 2
Initial mass (g)	332.0	383.0	Initial mass (g)	348.0	332.0
Retained mass (g)	230.5	262.0	Retained mass (g)	217.5	204.0
Mass passed (g)	99.5	119.0	Mass passed (g)	128.5	126.0
AIV (%)	29.97	31.07	AIV (%)	36.93	37.95
Average AIV (%)	30.52		Average AIV (%)	37.44

Figure 5

AIVs for each aggregate material.

3.4 Mix design of concrete

Table 7 shows the proportions of the fine and coarse natural and slag aggregates incorporated into each concrete mixture.

For Table 7, the following ratios were used:

• Fine to coarse aggregate: 0.4/0.6,

• Aggregate-to-cement ratio: 3.5, and

• Water-to-cement ratio: 0.4.

Each mixture was to produce six concrete specimens, three cubes (100 mm × 100 mm × 100 mm), and three cylinders (100 mm diameter × 203.2 mm height). This mix design procedure was used, rather than designing a specific grade of concrete since the aim of this project was to examine the variation of concrete strength with increasing the slag aggregate content.

3.5 Testing of fresh properties of concrete

3.5.1 Workability

The results of the slump test illustrated in Table 8 showed that the 100% natural aggregate mixture had a 0.00 mm slump height, which was further depicted in Figure 6. The slump height remained similar to the 25 and 50% slag aggregate mixtures.

Table 8

Slump height of each concrete mixture

Concrete mix (percentage of EAF slag) (%)	Slump height (cm)
0	4.2
10	0.5
15	0.5
25	0.0
50	0.0

Figure 6

Slump height of mixture containing 0% EAF slag.

The recorded data challenge the expected decrease in the workability with increasing slag content [14,20]. However, while working with the mixtures during the experiment, the mixtures were observed to become stiffer and more unworkable with increasing amounts of slag aggregate.

A slump measuring 0.00 mm in height indicates maximum stiffness, so any increase in the stiffness of the mixture caused by adding slag to the mixtures would not have been reflected by the slump height. Thus, based on this theory and the visual observations, it can be deduced that the slag aggregate did, in fact, decrease the workability of the mixtures. This deduction is further supported by the water absorption result since the higher water absorption of the steel slag aggregates than the natural aggregates also suggests that the workability should have decreased, once the water-to-cement ratio remained the same.

4.1 Compressive strength

Table 9 and Figure 8 display the effects of increasing slag aggregate content on the compressive strength of concrete. The graph shows an increase in strength from 0 to 25% slag content and then a drastic decrease from 25 to 50% slag content. The 25% slag aggregate mixture had a 1.15 times improvement in the 28-day compressive strength than the 100% natural aggregate mixture.

Figures 8–10 illustrate the variation of the density, compressive strength, and tensile strength respectively with increased content of the EAF slag aggregates.

Figure 8

Density variation with EAF slag content.

Figure 9

Compressive strength (MPa) variation with EAF slag content.

Figure 10

Tensile strength (MPa) variation with EAF slag content.

The compressive strength values increased as the percentage of the EAF slag aggregates increased, the highest value of the compressive strength was calculated to be 58.70 MPa using 50% EAF slag. This finding is in good agreement with the work done in Pellegrino et al. [25].

However, the density of the hardened concrete was highest using 25% EAF slag aggregates and lowest using 50% EAF slag aggregates. This finding is in contradiction with the work done by researchers in Qasrawi et al. [13] found that replacing fine aggregates in 25 MPa concrete produced concrete with compressive strengths 1.1–1.3 times greater than 100% EAF slag aggregates. Furthermore, papers [12,17] replaced coarse aggregate with slag and also found improved compressive strength results. Those three papers support the improved compressive strength of the 25% slag aggregate mixture in this study since it replaced both fine and coarse aggregates.

The anomaly in this finding can be justified by Table 6 in which the increases in the coarse EAF slag aggregate content would have initiated increases in the air spaces between the fine and the coarse particles. With 50% EAF slag aggregates, 30% of the EAF slag was coarse aggregates, and 20% fine aggregate, there would not have been adequate fine slag content to fill in the void spaces between the slag particles. This could have been attributed to a reduction in the overall density value of the concrete produced. This could also be justified from Table 10.

Table 10 displays the mathematical relations formulated between the tested parameters (density, compressive strength, and tensile strength) values and the percentage of the EAF slag aggregates.

It is evident that the R ² values are highest having a value of 0.663 for the linear prediction of compressive strength variation with the percentage of EAF slag. It can be deduced that the linear approximation is a good prediction for the compressive strength variation. This linear prediction can be attributed to an increase in the compressive strength values as the percentage of EAF slag content increases.

5.1 Recommendations

While this study followed standard testing procedures and methods that are directly applied in the field, there are some additions to this experiment that can improve its quality and the accuracy of its outcomes. They are as follows:

1. Aggregate properties such as porosity, fineness modulus, and resistance to abrasion should be studied to better understand the differences between EAF steel slag aggregates and natural aggregates.

2. Chemical testing of the EAF steel slag aggregate can be done to better understand its content. This would foster a prediction to be made with the cement and water mixture in concrete. The findings of such a test can potentially explain ways of increasing the tensile strength of concrete.

3. Investigations on the long-term performance of slag aggregates in concrete, and their performance when exposed to different environments can also be done. This would reveal areas for practical application.