Journal of Hazardous Materials 209–210 (2012) 84–91

Contents lists available at SciVerse ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Reuse of steel slag in bituminous paving mixtures

Sabrina Sorlini ∗ , Alex Sanzeni, Luca Rondi
Department of Civil, Architecture, Territory and Environmental Engineering, University of Brescia, Via Branze 43, Brescia, 25123, Italy

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents a comprehensive study to evaluate the mechanical properties and environmental
Received 29 April 2011 suitability of electric arc furnace (EAF) steel slag in bituminous paving mixtures. A variety of tests were
Received in revised form executed on samples of EAF slag to characterize the physical, geometrical, mechanical and chemical prop-
22 December 2011
erties as required by UNI EN specifications, focusing additionally on the volumetric expansion associated
Accepted 28 December 2011
with hydration of free CaO and MgO. Five bituminous mixtures of aggregates for flexible road pavement
Available online 9 January 2012
were designed containing up to 40% of EAF slag and were tested to determine Marshall stability and
indirect tensile strength. The leaching behaviour of slag samples and bituminous mixtures was evalu-
Keywords:
Steel slag
ated according to the UNI EN leaching test. The tested slag showed satisfactory physical and mechanical
Bituminous mixtures properties and a release of pollutants generally below the limits set by the Italian code. Tests on volume
Aggregates stability of fresh materials confirmed that a period of 2–3 months is necessary to reduce effects of oxides
Volumetric expansion hydration. The results of tests performed on bituminous mixtures with EAF slag were comparable with
Leaching behaviour the performance of mixtures containing natural aggregates and the leaching tests provided satisfactory
results.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction a study conducted by Sorlini et al. [18], based on results of a

previous research [20], was carried out to evaluate the steel slag
A recent study of the European Commission [1] showed that the recovery in road foundations. The slag was subjected to a chemical
percentage of primary slag per ton of steel produced in the UE is characterization in order to investigate the release of pollutants
in the range of 12–15%, whilst the percentage of secondary slag is and the phenomenon of volume expansion. In 2010, De Windt
around 3%; these values are also confirmed by a research conducted et al. studied the kinetics of steel slag leaching through batch tests
in Germany in 2010 [2]. In Italy the production of primary slag in at liquid to solid ratios of 10 and 100 over a 30-day period and
2009 was estimated about 3 million tons [3]. More information proposed a geochemical model validated with experimental data
about steel slag production is reported in other studies [4–6]. [21]. The evaluation of the mechanical properties on steel slag was
If not recovered, the steel slag is disposed of in landfills, thus done in several research projects [6,8,9,13,15,16,19], where results
increasing the amount of waste to be dumped. Reuse of steel slag showed similar characteristics between the mixtures containing
as recycled material can be seen as a way to limit the amount of natural aggregates and steel slags.
dumped wastes and allows to reduce the consumption of natural In this study, an extensive experimental campaign was executed
materials (i.e. natural aggregates). For these reasons, in the last two to investigate the possibility of steel slag recovery as a recy-
decades, a large number of experimental studies have been con- cled aggregate for the production of bituminous paving mixtures.
ducted to investigate the opportunity to partially substitute natural The experimental investigation included a geometrical, physical,
aggregates with steel slag as construction material [7–12]. mechanical and chemical characterization of electric arc furnace
The first studies on steel slag recovery in asphalt mixtures (EAF) slag and bituminous mixtures containing a fraction of slag.
reported in literature dates back to the beginning of the 1970s [7]. Special attention was given to the leaching properties and the vol-
Most of the researches in this field covered chemical, physical and ume stability due to the hydration of free lime and magnesium
geometrical properties [8,9,13–19]. In 2001, Motz and Geiseler [4] oxide.
studied the possibility of reusing steel slag in road paving, focusing
on the leaching tests to evaluate the environmental behaviour 2. Materials and methods
and on expansion tests to determine the volume stability. In 2003,
2.1. Steel slag

∗ Corresponding author. Tel.: +39 0303711299; fax: +39 0303711312. The experimental study was carried out on the slag of first fusion
E-mail address: [email protected] (S. Sorlini). (i.e. primary slag), a by-product of the steel production process

0304-3894/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.jhazmat.2011.12.066
 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91 85

Table 1
Tests performed to investigate chemical properties of steel slag.

Test requirement (reference standard) Steel slag sample

A B C D E F G H

0/20 14/32 0/32 0/32 0/100 0/32 0/32 14/32 0/22

M. D. ‘98 and Chemical composition x x – – – x – – –

M. D. 06/186 Oxides content (%) x x – – x x – – –

Lightweight organic contaminators (UNI EN 1744-1) x x – x – – – – –

Volume stability Dicalcium silicate disintegration (UNI EN 1744-1) – – – x – – – – –
Iron disintegration (UNI EN 1744-1) x – – x – x – – –
Expansion of steel slag (UNI EN 1744-1) – – – – – – – – x
UNI EN 13043 Magnesium oxide (UNI 196-2) – – – – x x – – –
Free lime Complexometry method (UNI EN 1744-1) x – – – – – – – –
Acidimetry method (UNI EN 1744-1) – – – – x – – – –
Extraction with ethylene glycol (UNE 80243/86) x – – x – – – – –
Leaching test Single stage test (UNI EN 12457-2) x x x x x – x x –

at the “Feralpi S.p.A.” factory, located in Lonato (Brescia, northern course 20%”). Mixtures of aggregates were composed using first
Italy). The steel is produced by EAF through direct fusion of ferrous fusion steel slag with a percentage in the range of 20–40% by weight
scrap, reclaimed body parts of vehicles and the addition of lime, and natural aggregates of various size and shape (see Table 2 for
iron alloys, coal and oxygen. After collecting the slag from the fur- mixture composition). The asphalt binder used was traditional
nace, a period of 3–4 months is allowed for aging in an unprotected bitumen of class 50/70 (UNI EN 1426:2002 or CNR BU n. 24/71);
open area and subsequently the material is sieved and transported aggregates were hot-mixed (approximately 160 ◦ C) with 3.8–5.5%
to the company that operates the recovery. Aging and exposition of bitumen depending on the type of pavement layer (Table 2).
to weather conditions are necessary to minimize subsequent volu-
metric changes due to oxidation. One month after transportation, 2.3. Experimental program
the slag is crushed into two grain size fractions named “slag 0/20”
(0–20 mm diameter) and “slag 14/32” (14–32 mm diameter). In the Italian legislation, the reference for the recovery of spe-
Experiments were carried out on a number of samples obtained cial wastes (such as steel slag) is the Legislative Decree 2006/152
through the previously described process between July 2006 and [22], amended and supplemented by Legislative Decree 2008/04.
July 2008. In the following discussion, slag samples will be labelled The specific law that rules over the recovery and reuse of some
with letters from A to H (Table 1). Samples A, B, C, E, F and G (July special wastes, including steel slag, is the Ministerial Decree dated
2006–December 2007) had grain size fractions “0/20” and “14/32” February 5th 1998 [23], amended and supplemented by Minis-
(in Table 1, label “0/32” indicates a mix, 50–50% by weight, of terial Decree 2006/186 [24]. Since 2004, the aggregates obtained
the two main fractions); sample D (January 2007) had grain size from natural or recycled materials have been certified with the CE
fraction 0–100 mm, obtained by mixing the above slag fractions mark, in accordance with UNI EN 13043:2004 [25] if used for the
with a coarser fraction provided by steel factory; sample H (July realization of roads, airfields and other trafficked areas.
2008), with grain size of 0–22 mm (required by UNI EN 1744-1), In compliance with UNI EN 13043:2004, a large number of tests
was selected for studying the effect of aging on the slag volumetric were carried out to evaluate geometrical, physical (mechanical) and
stability. chemical properties of the steel slag.

2.2. Bituminous mixtures 2.3.1. Geometrical and physical properties of steel slag
Geometrical properties comprise size, shape and angularity of
Five experimental bituminous mixtures for flexible road the aggregates. Particle size is essential for the subsequent compo-
pavements were composed using quantities of aged steel slag sition of the mixtures of aggregate, whereas shape and angularity
as aggregate, available in the grain size ranges 0–20 mm and are essential parameters for obtaining mixtures of high stability
14–32 mm. Two mixtures were prepared for the road base (respec- and roughness; the presence of elements with sharp edges gives
tively referred to as “Road base 30%” and “Road base 40%”), two mix- the material a high friction angle, it affects the roughness of the
tures for the base course (respectively “Base course 30%” and “Base surface layers and reduces pavement rutting under the traffic load
course 40%”) and one mixture for the wearing course (“Wearing [26].

Table 2
Experimental bituminous mixtures: grain size composition and quantity of bitumen.

Aggregates Particle size (mm) Mixtures – percentage composition (% by weight)

Road base 30% Road base 40% Base course 30% Base course 40% Wearing course 20%

Slag
Slag 0/20 0–20 15 20 30 40 20
Slag 14/32 14–32 15 20 – – –
Natural aggregate
Filler 0–0.075 5 5 5 5 7
Crushed 0/4 0–4 24 22 30 30 32
Crushed 2/8 2–8 15 10 15 10 20
Crushed 6/14 6–14 – – 10 8 21
Crushed 10/20 10–20 11 12 10 7 –
Round 14/32 14–32 15 11 – – –

Bitumen class 50/70 (% by weight) 3.8 4.2 4.5 4.8 5.5

86 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91

Regarding the physical properties, the value of the Los Angeles graded aggregate material for bases or subbases for highways or
test (UNI EN 1097-2 Tests for mechanical and physical properties airports” [30].
of aggregates—Part 2: Methods for the determination of resistance
to fragmentation) is required for the assessment of the resistance 2.3.3. Physical–mechanical properties and leaching of bituminous
to fragmentation. The micro-Deval test (UNI EN 1097-1 Tests for mixtures
mechanical and physical properties of aggregates—Part 1: Determi- The following documents were adopted as a reference: Bul-
nation of the resistance to wear (micro-Deval)) provides a measure letin n. 178/95 National Research Council (CNR BU 178/95) [31],
of the resistance to wear by friction, and the accelerated polishing a popular national standard that identifies road pavement types;
coefficient is used to assess the ability of the material to be con- the standard Technical Specifications document of the Province of
sumed on the surface and its ability to become slick. In addition, it Brescia [32]; the standard Technical Specifications of the “Brescia-
is necessary to determine the particle surface absorption, the affin- Verona-Vicenza-Padova Motorway” Company [33].
ity between aggregate and bitumen, the sensitivity to frost action The tests carried out for the characterization of the experimental
(depending on material nature and structure) and the behaviour bituminous mixtures comprise both the Marshall test for studying
in case of high temperature changes (during mixing with bitumen the characteristics of stability, and the indirect tensile strength test
and road construction). on specimens of wearing course bituminous mixtures. The percent-
age of residual voids was measured, after compaction, in order to
2.3.2. Chemical properties of steel slag assess the leaching properties of conglomerates. In addition, the
The chemical characterization of the steel slag was carried out release of pollutants was verified through the compliance test of
with tests listed in Table 1, in agreement with the requirements of extraction in a single stage provided by the method UNI 12457-2.
the Ministerial Decree February 5th 1998, for the recovery of non-
hazardous waste. The release of pollutants was evaluated through 3. Results and discussion
the compliance test provided by the Ministerial Decree 2006/186
and according to the method UNI 12457-2 [27] of extraction in 3.1. Tests performed on steel slag
a single stage in demineralized water (liquid/solid ratio, L/S = 10)
for a period of 24 h. Chemical requirements were also investigated 3.1.1. Results of the geometrical and physical characterization
according to UNI EN 13043. tests
The following three tests were performed to evaluate the The results of the characterization tests, performed on sam-
presence of chemical components that affect volume stability: ples A and B of slag 0/20 and 14/32, are summarized in Table 3;
determination of dicalcium silicate and iron disintegration of air- a comparison with typical values referring to natural aggregates
cooled blast-furnace slag and direct measurements of steel slag is proposed. Fig. 1 shows the grain size distribution curves of the
expansion. The latter test is the most significant for this type of fractions studied. Regarding the geometrical requirements of the
waste and allows to detect its predisposition to expansion due to aggregates, the test results for the assessment of the flakiness and
a late hydration of lime and magnesium oxide. A testing time of 7 shape indexes indicated that, due to the crushing process during
days (168 h), for EAF slag, must be allowed and the values of expan- maturation stages, the percentage of flat or non-polyhedral ele-
sion should be compared with the categories specified in UNI EN ments is fairly limited (1–6%), contributing to ensure high stability
13043. The analysis of free CaO, which, together with MgO, may characteristics of the mixtures.
be responsible for the volume expansion of the slag, was carried In terms of physical requirements, the Los Angeles test gave val-
out with the complexometry method, the acidimetry method and, ues of about 22–23%, whilst the micro-Deval test resulted in the
according to the Spanish standard UNE 80243/86 [28], by extraction range 6.5–9.5%, showing that the steel slag has a fair resistance to
with ethylene glycol. fragmentation and wear and can be used in any layer of a flexible
Five expansion tests were performed – 15, 30, 60, 120 and pavement; the results were below the 25% limit set by the Techni-
190 days after production – on slag sample H (July 2008), cal Specifications of the Province of Brescia, although it appeared
specifically selected for studying the effects of aging on vol- higher than typical values obtained with traditional aggregates
ume stability. The expansion tests were carried out following such as basalt and porphyry. Since the result of the Los Angeles
the method UNI EN 1744-1 [29] and first proposed by Motz and test is one of the most important parameters for road use, tests of
Geiseler in 2001 [4]. The results were compared with the limit resistance to fragmentation were performed weekly on the mate-
value proposed by ASTM D 2940-98 “Standard specification for rial collected in the period from January 2007 to the end of April

Table 3
Tests results to determine geometrical and physical properties of slag.

Geometrical and physical requirements Unit Steel slag sample Natural aggregates

A B

0/20 14/32 0/20 14/32

Flakiness index (EN 933-3) – 4 6 1 3 –

Shape index (EN 933-4) – 4 3 2 4 –
Percentage of crushed and broken surfaces (EN 933-5) % 100 100 100 100 –
Resistance to fragmentation, Los Angeles (EN 1097-2) % 22 23 23 23 10–20
Resistance to wear, micro-Deval (EN 1097-1) % 6.5 8.6 9.3 9.5 5.0–8.0
Resistance to polishing (EN 1097-8) – 0.44 – 0.47 – 0.40–0.50
Aggregate unit weight (EN 1097-6) kN/m3 20.97 16.70 20.92 16.32 –
Specific gravity (EN 1097-6) kN/m3 39.1 39.0 39.6 39.2 26.0–28.0
Particles water absorption (EN 1097-6) % 1.7 1.5 2.2 2.0 1.0–2.0
Freeze–thaw resistance (EN 1367-1) % 0.3 0.5 1.1 2.1 0.2–0.8
Thermal shock resistance (EN 1367-5) % 1 – 2 – 1–3
Aggregate–bitumen affinity (prEN 12697-11) % 15 – 10 – 5–10

Note: – = test not carried out/data not available.

 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91 87

100
Slag A 0/20 of the slag (Table 4). For most of the elements (Cu, Crtot , Zn, Ba,
90 Slag B 0/20 Co, V, Fe, Mn) a greater concentration was identified in sample
Slag A 14/32
80 Slag B 14/32
A and E with grain size range 14–32 mm, whilst concentration of
other elements, such as Cu, Crtot , B, did not vary with grain size.
Passing percentage (%)

70
The different results in different samples are due to the hetero-
60 geneous characteristics of the materials fed into the electric arc
50 furnace.
40
The total content of oxides (SiO2 , CaO, Al2 O3 , MgO and FeO, see
Table 4) was in the range 59.8–89.2%; only in three cases this per-
30
centage was higher than the minimum value of 80% required by the
20 Ministerial Decree of 1998 for the recovery of such wastes with a
10 simplified procedure [23,24]. The greater percentage was due to the
0
oxides of calcium, silicon and iron and, generally, metal and oxide
0.01 0.1 1 10 100 concentrations were higher in the 14/32 fraction than in the 0/20
Sieve size (mm) one. The concentration of lime (expressed as total CaO) and mag-
nesium oxide (MgO), which are the most significant parameters for
Fig. 1. Grain size distribution curves of steel slag samples A and B. volume stability, were in the range 16–22% and 2–3%.
The analysis of free lime was carried out according to differ-
2007 and the results confirmed the fair mechanical properties of the ent methods. The Spanish method [28] provided a value of 0.07%
material. Results were also comparable with the values obtained in on sample C and a value equal to zero on sample A of slag 0/20.
other recovery research projects [4,14,15,18,20]. The result of the The acidimetry method [29] on slag D showed a concentration of
Los Angeles test may be affected by the structure of the slag and free lime of 0.33% and the only sample analysed with the com-
the presence of inner voids in the grains due to the cooling process. plexometry method [29] did not provide any results. Data obtained
The values of the accelerated polishing coefficient (APC) were in were lower than the limit value of 4% indicated in the literature
the range 0.44–0.47 and comparable to those of natural materials [4] and reported by German law as the limit for the recovery of
such as porphyry (APC = 0.45–0.48) or basalt (APC = 0.42–0.45); the slags; furthermore, these values were lower than those reported in
particle water absorption was between 1.5 and 2.2%, in the range a previous study by Sorlini et al. on the same residues [18].
of traditional materials for the construction of road pavements. The dicalcium silicate disintegration was evaluated by the anal-
Other features, including the freeze–thaw resistance and the ysis of the slag surface that, exposed to UV radiation, always showed
aggregate–bitumen affinity, were comparable with the values usu- a uniform colour, typical of the stable samples and not subjected to
ally obtained from natural aggregates. The value of the aggregate the phenomena of disintegration.
unit weight was approximately 20.9 kN/m3 ; the specific gravity The evaluation of the iron disintegration, due to the hydrolysis
of the grains, instead, was in the range from 39.1 to 39.6 kN/m3 , of iron and manganese sulphides in the sample placed in water for 2
approximately 1.5 times the specific gravity of natural soil (Table 3). days, did not show significant phenomena of cracking or crushing.
Coarse lightweight organic contaminants, larger than 2 mm,
3.1.2. Results of the chemical characterization were analysed in two samples and they always resulted lower than
High concentrations of total chromium, barium, iron, man- the maximum value of 0.1% allowed for the category mLP 0.1 of the
ganese and aluminium were found in the chemical composition standard UNI 13043.

Table 4
Steel slag chemical composition and percentage of oxides content.

Parameter Unit Steel slag sample

A D E

0/20 14/32 0/100 0/20 14/32

Lead (Pb) mg/kg 26.6 30.0 – 29.3 25.9

Antimony (Sb) mg/kg <0.005 <0.005 – 53.200 59.300
Copper (Cu) mg/kg 1.1 250.1 – 272.8 323.6
Cadmium (Cd) mg/kg 3.5 4.5 – 21.7 21.3
Total chromium (Crtot ) mg/kg 2,428.3 4,850.4 – 871.3 933.6
Arsenic (As) mg/kg <0.001 <0.001 – <0.001 0.400
Selenium (Se) mg/kg <0.01 <0.01 – <0.01 <0.01
Mercury (Hg) mg/kg <0.001 <0.001 – <0.001 <0.001
Zinc (Zn) mg/kg 199.5 215.8 – 259.7 301.2
Barium (Ba) mg/kg 1,276.7 1,412.7 – 1,252.1 1,233.7
Beryllium (Be) mg/kg <0.001 <0.001 – <0.001 <0.001
Cobalt (Co) mg/kg 1.1 4.8 – 3.7 5.1
Nickel (Ni) mg/kg 16.6 187.5 – 156.1 105.4
Vanadium (V) mg/kg 222.6 268.8 – 307.2 368.9
Aluminium (Al) mg/kg 33,875.7 35,761.4 – 30,222.5 29,959.5
Boron (B) mg/kg 217.60 225.70 – <0.01 <0.01
Iron (Fe) mg/kg 256,713 359,353 – 301,466 34,229
Manganese (Mn) mg/kg 27,845 34,875 – 32,776 35,361

Total silicon (SiO2 ) % 12.90 17.20 10.25 18.59 19.10

Calcium oxide (CaO) % 18.65 22.10 17.93 18.28 16.46
Aluminium oxide (Al2 O3 ) % 5.32 6.35 7.20 5.80 5.84
Magnesium oxide (MgO) % 2.00 2.35 3.10 2.53 2.02
Iron oxide (FeO) % 20.92 29.84 50.71 35.71 36.71
88 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91

Table 5
Concentrations of the pollutants in the leachate, obtained in compliance with UNI EN 12457-2.

Parameter Unit Steel slag sample Limits of M. D. 06/186

A B C D F G

0/20 14/32 0/32 14/32 0/100 0/32 14/32

pH – 7.1 6.8 9.9 10.9 11.7 11.2 11.3 5.5–12.0

Mercury ␮g/L <1.0 <1.0 0.9 <1.0 <1.0 <1.0 <1.0 1.0
Vanadium ␮g/L 118 120 91 180 77 110 120 250
Selenium ␮g/L <10.0 <10.0 1.0 <1.0 <5.0 <5.0 1.5 10.0
Zinc mg/L <0.01 0.07 <0.01 <0.01 <0.01 0.07 <0.01 3.00
Total chromium ␮g/L 6 15 8 <5 <10 15 30 50
Copper mg/L 0.007 0.010 <0.005 <0.010 <0.010 <0.010 <0.010 0.050
Barium mg/L 0.27 0.43 0.23 0.26 0.30 0.05 0.30 1.00
Arsenic ␮g/L <1 <1 <5 <10 <10 <10 <10 50
Cadmium ␮g/L <1.0 <1.0 <0.5 <3.0 <3.0 <3.0 <3.0 5.0
Fluorides mg/L 0.15 0.16 0.40 0.20 0.10 1.00 0.10 1.50
Chlorides mg/L 2.0 2.9 2.7 1.0 <1.0 3.0 1.0 100.0
Sulphates mg/L 6.8 12.0 5.2 <1.0 3.0 3.0 3.0 250.0
Nitrates mg/L 1.4 5.1 1.5 3.0 <3.0 0.5 <1.0 50.0
DOC mg/L 21.3 29.7 9.0 <1.0 11.0 <5.0 <1.0 30.0

Note: for all samples the following concentrations were also found: no asbestos, Pb < 5 ␮g/L, Ni < 10 ␮g/L, Be < 1 ␮g/L, Co < 5 ␮g/L, and CN < 50 ␮g/L.

3.1.3. Results of the leaching test 1.0 Test 1

Test 2
The leaching tests (UNI EN 12457-2) performed on all available

Volumetric expansion (%)

Test 3
samples provided values below the limits indicated in the Min- 0.8
Test 4
Test 5
isterial Decree 2006/186 (Table 5). For most of the parameters, ASTM limit value
higher pollutant concentrations were measured in the finest grain
0.6
size fraction slag, due to its higher specific surface. Vanadium, total εv = 0.5%
chromium and nitrate showed a variable concentration amongst
0.4
the different samples. The release of organic matter (dissolved
organic carbon, DOC) of slag sample A reached concentrations close
to the limit, whilst it remained well below in other cases. 0.2

3.1.4. Volumetric expansion test results 0.0

The results of the volumetric expansion, obtained from five tests 0 40 80 120 160 200
carried out on slag H (specifically selected for studying the aging Time after tapping (d)
effects on the volumetric stability), are shown in Fig. 2 as a function
Fig. 3. Volumetric expansion as a function of aging time.
of time in a logarithmic scale, to illustrate that significant expan-
sion occurs within 12 h of the test. Each curve was obtained from
tests performed at different time intervals after slag production (15, material for bases or subbases for highways or airports”. To ensure
30, 60, 120 and 190 days) and represents the average of two tests acceptable volume stability, an aging period of at least 2–3 months
carried out on specimens obtained from the same sample. Fig. 3 is therefore advisable [34].
shows the volumetric expansion measured at the end of each per-
formed test, as a function of the slag aging period, and it indicates
that the volumetric expansion εv (at the end of the test) is in inverse 3.2. Results of tests on bituminous mixtures
proportion to the slag aging time: Test 1, carried out on the “fresh
slag” (approximately 15 days of aging), gave εv ≈ 0.6% at 168 h; Test Tables 6–8 show the results of the tests on conglomerate sam-
5, performed on the aged one (approximately 6 months after tap- ples for road base, base course and wearing course, prepared by
ping), gave a εv ≈ 0.1%. Only Tests 3, 4 and 5 met the requirement using steel slag and natural stone aggregates in percentages indi-
of ASTM D 2940-98 “Standard specification for graded aggregate cated in Table 2. Each table refers to a type of conglomerate with
a given quantity of slag. The slag sample identification letter is
reported and a comparison with acceptance limits given by the
CNR Bulletin 178/95 and other reference documents is suggested.
In addition, for each type of mixture, the comparison with results
of the same tests conducted on mixtures with natural aggregates is
proposed.
Tests performed on bituminous mixture samples for the road
base (slag percentage of 30 and 40%, Table 6) provided values above
minimum requirements set by the CNR Bulletin n. 178/95 and
by the Technical Specifications documents. In particular, for both
mixtures, the stability and stiffness values, obtained in the Mar-
shall test, were significantly higher than the limits of acceptance.
The comparison with the data obtained from mixtures with natu-
ral aggregates showed that the performance of the experimental
bituminous mixtures was similar or even better than the tradi-
tional mixtures. Compared with the natural aggregate mixtures,
Fig. 2. Volumetric expansion as a function of test time. the experimental ones gave a higher percentage of residual voids
 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91 89

Table 6
Road base test results and comparison with natural aggregate mixtures.

Parameter Unit Mixture Road Mixture Road Mixture Road Natural aggregates Reference values
base 30% base 30% base 40% mixture

Slag content % by weight 30 30 40 0 CNR BU 178/95a Motorway spec.

Slag sample A B A –

Marshall stability daN 1,564 1,721 1,694 1,558 >800 >900

Marshall flow mm 2.7 2.9 2.9 3.2 1.5–3.0 –
Marshall stiffness daN/mm 595 591 595 511 >250 >250
Residual voids % 5.6 5.1 4.4 2.9 4.0–7.0 4.0–7.0
Voids filled with bitumen % 62.3 66.4 71.3 77.3 <80.0 –
Apparent density g/cm3 2.633 2.685 2.725 2.496 – –
a
Similar values are adopted in Brescia Province Technical Specifications.

Table 7
Base course test results and comparison with natural aggregate mixtures.

Parameter Unit Mixture base Mixture base Mixture base Natural aggregates Reference values
course 30% course 30% course 40% mixture

Slag content % by weight 30 30 40 0 CNR BU 178/95a Motorway spec.

Slag sample A B A –

Marshall stability daN 1,681 1,660 1,709 1,487 >1,000 >1,100

Marshall flow mm 2.8 2.8 2.7 2.3 1.5–3.0 –
Marshall stiffness daN/mm 602 607 639 657 >300 >300
Residual voids % 6.9 6.9 6.0 2.4 3.0–6.0 3.0–7.0
Voids filled with bitumen % 61.3 61.5 66.2 82.3 <80.0 <80.0
Apparent density g/cm3 2.571 2.608 2.648 2.469 – –
a
Similar values are adopted in Brescia Province Technical Specifications.

(4.4–5.6% rather than 2.9% of the natural aggregate mixture), and from this experimental mixture. In particular, values of Marshall
a lower percentage of voids filled with bitumen (62–71% against stability (even after submersion in a water bath at 60 ◦ C for 24 h)
the 77% of the natural aggregate mixture). Due to the high specific and indirect tensile strength were higher than the limits of accep-
gravity of the steel slag, the density of the experimental mixtures tance, showing suitable characteristics of mechanical resistance,
was 7–9% higher than the traditional mixtures. comparable to the performance of traditional mixtures. The value
Tests performed on bituminous mixture samples for the base of the apparent density was 7% higher than the natural aggregate
course (steel slag percentage of 30 and 40%, Table 7), provided mixture.
acceptable results as well. The stability and stiffness values, As expected, the release of pollutants from the experimental
obtained by the Marshall test, were significantly higher than the bituminous mixtures, complied with UNI EN 12457-2, was below
limits of acceptance, indicated by the CNR Bulletin n. 178/95 and the limits of the Ministerial Decree 2006/186 for all samples. The
by the Technical Specifications, and even better than the values concentrations of all pollutant elements were widely below the
obtained from the tests done on the natural aggregate mixture. The limits, as reported in Table 9. The presence of the steel slag deter-
residual voids resulted high and the apparent density of the Mar- mined an increase in pH, concentration of vanadium, zinc, total
shall test specimens were 6–8% higher than mixtures with natural chromium, fluorides, chlorides and nitrates for all mixtures. The
aggregates. most critical parameter, even if widely below the limit, was again
A summary of test results performed on samples of wearing vanadium, for which no enclosing or binding effect due to the bitu-
course is reported in Table 8. Satisfactory results were obtained men was appreciated.

Table 8
Wearing course test results and comparison with natural aggregate mixtures.

Parameter Unit Mixture Wearing Mixture Wearing Natural aggregates Reference values
course 20% course 20% mixture

Slag content % by weight 20 20 0 CNR BU 178/95a Motorway spec.

Slag sample A B –

Marshall stability daN 1,806 1,734 1,506 >1,200 >1,200

Marshall flow mm 2.9 2.8 2.8 1.5–3.0 –

Marshall stiffness daN/mm 624 615 528 >350 >350

Residual voids % 4.8 2.4 2.0 3.0–6.0 3.0–6.0

Voids filled with bitumen % 73.2 84.5 86.0 <80.0 –
Apparent density g/cm3 2.550 2.601 2.456 – –
Stability after submersion %; daN 95.5; 1,725 94.0; 1,636 – – –

Indirect tensile strength N/mm2 1.05 0.97 – >0.70 0.60–1.00

a
Similar values are adopted in Brescia Province Technical Specifications.
90 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91

Table 9
Concentrations of the pollutants in the leachate, obtained in compliance with UNI EN 12457-2.

Parameter Unit Natural aggregates mixtures Mixtures with steel slag Limits of M. D. 2006/186

Road base Base course Wearing course Road base Base course Wearing course

Bitumen % 4.2 4.8 5.5 4.2 4.8 5.5 –

pH – 8.4 8.8 8.4 10.5 9.5 10.6 5.5–12.0
Vanadium ␮g/L 2.4 9.6 2.8 147.4 99.0 146.0 250.0
Zinc mg/L 0.003 0.002 0.001 0.002 0.018 0.023 3.000
Total chromium ␮g/L 2.2 7.5 2.6 16.1 12.2 6.0 50.0
Copper mg/L 0.001 0.002 0.037 0.002 <0.001 <0.001 0.050
Barium mg/L 0.079 0.028 0.242 0.120 0.035 0.041 1.000
Beryllium ␮g/L <1 <1 <1 <1 <1 3 10
Arsenic ␮g/L <1.0 4.0 4.4 2.3 <1.0 5.0 50.0
Fluorides mg/L 0.043 0.040 0.043 0.152 0.260 0.260 1.500
Chlorides mg/L 0.500 0.446 0.421 0.734 2.300 2.300 100.000
Sulphates mg/L 3.250 2.859 3.133 6.081 1.700 2.700 250.000
Nitrates mg/L 0.378 0.474 0.267 15.665 0.900 3.500 50.000

Note: for all samples the following concentrations were also found: no asbestos, Pb < 5 ␮g/L, Ni < 10 ␮g/L, Se < 10 ␮g/L, Co < 5 ␮g/L, CN < 50 ␮g/L, Hg < 1 ␮g/L, Cd < 1 ␮g/L, and
DOC < 10 mg/L.

4. Conclusions References

Several tests were executed to characterize physical, geomet- [1] European Commission, Draft Reference Document on Best Available Tech-
niques for the Production of Iron and Steel, Integrated Pollution Prevention
rical, mechanical and chemical properties of EAF slag; five bitumi- and Control, 2009.
nous mixtures of aggregates (containing up to 40% of EAF slag) were [2] Badische Stahlwerke GMBH, Environmental Declaration 2010, BSW, 2010.
designed and tested. From a mechanical point of view, the fragmen- [3] Federacciai, L’industria siderurgica italiana nel 2009 – Relazione annuale,
Pubblicazioni Federacciai. http://www.federacciai.it/Pdf/RelazioneAnnuale/
tation and wear resistance provided acceptable results (22–23% Relazione 2009 1.pdf, 2009.
and 6.5–9.5% respectively); the values of the accelerated polish- [4] H. Motz, J. Geiseler, Products of steel slags. An opportunity to save natural
ing coefficient (0.44 and 0.47) and water absorption (1.5–2.2%) resources, Waste Management 21 (2001) 285–293.
[5] Euroslag, Legal Status of slags Position Paper, European Slag Association.
were comparable with values obtained with natural materials, http://www.euroslag.org/, 2006.
such as porphyry and basalt. The bituminous mixtures contain- [6] P. Ahmedzade, B. Sengoz, Evaluation of steel slag coarse aggregate in hot mix
ing steel slag showed mechanical characteristics similar or even asphalt concrete, Journal of Hazardous Materials 165 (2009) 300–305.
[7] J. Emery, Steel slag utilization in asphalt mixes, in: National Slag Association,
more satisfactory than the mixtures obtained using natural aggre-
MF 186-1, 1984.
gates (measured values of Marshall Stability generally higher than [8] U. Bagampadde, H.I.A. Wahhab, S.A. Aiban, Optimization of steel slag aggregates
1600 daN, average Marshall flow of 2.8 mm). From a chemical point for bituminous mixes in Saudi Arabia, Journal of Materials in Civil Engineering
of view, slag samples had a relevant content of total chromium 11 (1999) 30–35.
[9] Y. Xue, S. Wu, H. Hou, J. Zha, Experimental investigation of basic oxygen furnace
(up to 4850 mg/kg), barium (up to 1413 mg/kg), aluminium (in slag used as aggregate in asphalt mixture, Journal of Hazardous Materials B138
the range 30,000–35,000 mg/kg), iron (up to 359,353 mg/kg) and (2006) 261–268.
manganese (in the range 28,000–33,000 mg/kg). The release of pol- [10] P. Chaurand, J. Rose, V. Briois, L. Olivi, J.L. Hazemann, O. Proux, J. Domas, J.Y.
Bottero, Environmental impacts of steel slag reused in road construction: a
lutants from the leaching test cannot be regarded as negligible crystallographic and molecular (XANES) approach, Journal of Hazardous Mate-
although it resulted lower than the limits indicated in the Min- rials B139 (2007) 537–542.
isterial Decree 2006/186: for instance the release of mercury in [11] S. Wu, Y. Xue, Q. Ye, Y. Chen, Utilization of steel slag as aggregates for stone mas-
tic asphalt (SMA) mixtures, Building and Environmental 42 (2007) 2580–2585.
sample B was 0.9 ␮g/L (with a limit of 1.0 ␮g/L), vanadium in sam- [12] M. Etxeberria, C. Pacheco, J.M. Meneses, I. Berridi, Properties of concrete using
ple C was 180 ␮g/L (with a limit of 250 ␮g/L), total chromium in metallurgical industrial by-products as aggregates, Construction and Building
sample G was 30 ␮g/L (with a limit of 50 ␮g/L) and DOC in sample Materials 24 (2010) 1594–1600.
[13] I.M. Asi, H.Y. Qasrawi, F.I. Shalabi, Use of steel slag aggregate in asphalt concrete
A was 29.7 mg/L (with a limit of 30 mg/L). Free lime and magne- mixes, Canadian Journal of Civil Engineering 34 (2007) 902–911.
sium oxide concentrations, responsible for expansive phenomena, [14] E. Marchi, Studio e raccolta dati sul processo relativo alle scorie di acciaieria uti-
were 16–22% and 2–3% respectively. Measurements of volumetric lizzate per la produzione di aggregati per conglomerati bituminosi speciali per
le pavimentazioni stradali, Master Degree Thesis, University of Trieste, 2006.
expansion (εv = 0.62%, obtained from fresh material 15 days after
[15] L. Morfini, L. Pagani, Loppe di produzione dell’acciaio: sviluppo industriale per
tapping) resulted in inverse proportion to the aging time and sug- produzione di aggregato da riciclo, L’Edilizia 142 (2005) 36–37.
gest that a period of 2–3 months must be allowed before reuse. The [16] M. Pasetto, N. Baldo, Experimental evaluation of high performance base course
release of pollutants measured in leaching tests performed on the and road base asphalt concrete with electric arc furnace steel slags, Journal of
Hazardous Materials 181 (2010) 938–948.
experimental bituminous mixtures resulted acceptable, except for [17] S. Porisiensi, M. Maschietto, P. Grandesso, Scorie di acciaieria: un nuovo aggre-
total chromium (16.1 ␮g/L) and vanadium (147.4 ␮g/L), and gener- gato ad elevate caratteristiche per conglomerati bituminosi speciali, Rassegna
ally lower than that of the slag alone. del bitume 50 (2005) 53–57.
[18] S. Sorlini, C. Collivignarelli, M. Vezzola, Recupero di scorie nei rilevati stradali,
Recycling 1 (2003) 101–106.
Acknowledgements [19] H. Ziari, M.M. Khabiri, Preventive maintenance of flexible pavement and
mechanical properties of steel slag asphalt, Journal of Environmental Engineer-
ing and Landscape Management 15 (2007) 188–192.
The study was conducted under the research agreement “Recov- [20] M. Del Fabbro, M. Stefanutti, C. Cerchia, Nuovi prodotti dalla lavorazione delle
ery of steel slags in bituminous mixtures” between the University scorie di forno elettrico ad arco: caratteristiche fisico-chimiche e possibili
of Brescia and the Company “Vezzola S.p.A.” (Lonato, Brescia, Italy). impieghi in campo geotecnico e stradale, Quarry & Construction 9 (2001).
[21] L. De Windt, P. Chaurand, J. Rose, Kinetics of steel slag leaching: batch tests and
The authors wish to thank Eng. Aceti Enea (Vezzola S.p.A.) for his modeling, Waste Management 31 (2010) 225–235.
collaboration in the experimental activity and the company “Feralpi [22] Legislative Decree 04/03/2006 n. 152, Norme in materia ambientale, Published
Siderurgica S.p.A.” (Lonato, Brescia) for providing the test materi- on Gazzetta Ufficiale n. 88 dated 04/14/2006 – Supplemento Ordinario n. 96.
[23] Ministerial Decree 02/05/1998, Individuazione dei rifiuti non pericolosi sotto-
als. The authors wish to thank Eng. Bertoglio Vera Stefania, Eng.
posti alle procedure semplificate di recupero ai sensi degli articoli 31 e 33 del
Colosio Enrico, Eng. Gervasoni Giulio and Eng. Ragnoli Carla for Decreto Legislativo 5 Febbraio 1997 n. 22, Published on Gazzetta Ufficiale n. 88
collaborating in the research project. dated 04/16/1998 – Supplemento Ordinario n. 72.
 S. Sorlini et al. / Journal of Hazardous Materials 209–210 (2012) 84–91 91

[24] Ministerial Decree 04/05/2006 n. 186, Regolamento recante modifiche al [30] American Society for Testing and Materials, ASTM D 2940-98: Standard spec-
decreto ministeriale 5 Febbraio 1998 “Individuazione dei rifiuti non pericolosi ification for graded aggregate material for bases or subbases for highways or
sottoposti alle procedure semplificate di recupero, ai sensi degli articoli 31 e 33 airports, 1998.
del Decreto Legislativo 5 Febbraio 1997 n. 22”, Published on Gazzetta Ufficiale [31] Researches National Council, Catalogo Italiano delle pavimentazioni, CNR 178,
n. 115 dated 05/19/2006. Roma, 1995.
[25] UNI EN 13043, Aggregates for bituminous mixtures and surface treatments for [32] Technical Specifications document of the Province of Brescia (ai sensi art. 45
roads, airfields and other trafficked areas, January 2004. D.P.R. 12/21/1999 n. 554), Ex S.S. 572 Desenzano Salò – messa in sicurezza
[26] P. Ferrari, F. Giannini, Ingegneria stradale. Corpo Stradale e Pavimentazioni, e razionalizzazione dei flussi transitanti, Provincia di Brescia, Area Tecnica,
ISEDI, 1996. 2005.
[27] UNI EN 12457-2, Leaching: compliance test for leaching of granular waste [33] Technical Specifications of the “Brescia-Verona-Vicenza-Padova Motorway”
materials and sludges—Part 2: One stage batch test at a liquid to solid ratio Company, Lavori di manutenzione alle pavimentazioni, Autostrada Brescia-
of 10 L/kg for materials with particle size below 4 mm (without or with size Verona-Vicenza-Padova S.p.A., 2005.
reduction), October 2004. [34] V.S. Bertoglio, Studio sperimentale sulla stabilità di volume delle scorie di
[28] UNE-80243/86, Contenido en cal libre, 1986. acciaieria, Master Degree Thesis, University of Brescia, 2008.
[29] UNI EN 1744-1, Tests for chemical properties of aggregates—Part 1: Chemical
analysis, January 2010.