See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/234004270
Waste ammunition as secondary mineralizing
raw material in Portland cement production
Article in Cement and Concrete Composites · December 2005
DOI: 10.1016/j.cemconcomp.2005.10.001
CITATIONS
READS
15
88
1 author:
Konstantinos G. Kolovos
Hellenic Army Academy
36 PUBLICATIONS 322 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Green Binder Materials for Civil Engineering and Architecture Applications View project
All content following this page was uploaded by Konstantinos G. Kolovos on 21 December 2016.
The user has requested enhancement of the downloaded file.
Cement & Concrete Composites 28 (2006) 133–143
www.elsevier.com/locate/cemconcomp
Waste ammunition as secondary mineralizing raw material
in Portland cement production
K.G. Kolovos
*
National Technical University of Athens, School of Chemical Engineering, Labs of Inorganic and Analytical Chemistry,
9 Heroon Polytechniou St., 15773 Athens, Greece
Received 27 January 2005; accepted 26 September 2005
Available online 21 November 2005
Abstract
This paper presents a laboratory scale simulation that aims to investigate the possibility of partially substituting ordinary cement raw
mix with waste ammunition materials (WAM), originated from a shooting range in Athens, Greece, in Portland cement clinker production. One reference and twelve modified mixtures, containing 0.5%, 1.0%, 1.5% and 2.0% w/w of three blends of corresponding types of
waste ammunition materials, were examined. It was concluded that the three used WAM blends, improve remarkably the burnability of
the cement raw mixture, even though in a different extent, without affecting considerably the hydration rate and the cement properties. In
spite of the high volatile matter in the WAM, primarily due to high levels of lead present, incorporation degree of the heavy metals present in the WAM blends in the mineralogical clinker compounds was rather high during the sintering process. Leaching tests showed that
the heavy metal concentrations in the leachates were kept low.
2005 Elsevier Ltd. All rights reserved.
Keywords: Waste ammunition; Cement raw mix; Clinker; Mineralizers; Volatility
1. Introduction
The environmental impact of toxic heavy metals such as
Pb (lead), Cu (copper), Zn (zinc), Sb (antimony), Sn (tin),
Ni (nickel), Cr (chromium), Cd (cadmium), Hg (mercury),
As (arsenic) etc., has been extensively reviewed in the past.
One specific source of the above metals, especially Pb and
Sb, is different kinds of manufactured ammunition. Due to
its special properties such as softness, high specific gravity,
and low cost for production, lead is considered as a suitable
material for making bullet cartridges and airgun pellets and
therefore difficult to be replaced by other materials. Antimony is added in varying proportions in order to harden
the bullet, depending on the characteristics needed. On
the other hand, the case of the cartridge consists of different
*
Present address: 43 Liakateon St., 114 74 Athens, Greece. Tel.: +30
210 6426339.
E-mail address: kolovosk@central.ntua.gr
0958-9465/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.cemconcomp.2005.10.001
brass alloys and contains varying levels of Cu, Zn, Ni, Sn,
Fe, Sb, Cd etc.
Published studies have shown the random composition
of bullet alloys due to different types of produced ammunition [1–5]. Different concentrations for trace elements can be
detected even for cartridges of the same type and manufacturer [6]. Taking into account that in some cases the crushed
bullets or fragments settle down in the wider area of the
shooting range and remain without been collected manually, they may enter and contaminate, in the long run, the
soil layer. Even if to the authorsÕ knowledge very few quantitative data are known with regard to the amounts of accumulated spent ammunition in shooting ranges, according to
a recent published work [7], in Finland, a country that not
only produces ammunition for recreational shooting but
also has one of the biggest number of functional shooting
ranges in Europe, a mean value of more than 7000 kg of
shot material/year/shooting range and approximately the
same mean value for bullet cases is speculated. These magnitudes vary, of course, from country to country and the
134
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
above numbers might probably be found increased in countries with higher population of active shooters involved not
only with Olympic shooting events but also with other
sports that incorporate shooting (p.e. biathlon).
In most of the countries that produce ammunition, the
major portion goes to military purposes, and although very
little are known, the used material in some cases is collected
and recycled. On the other hand, ammunition that is manufactured to be used for other purposes, such as clay and
target shooting, generally is not recycled after its usage,
due to the extended heterogeneity (accumulation of different types of spent ammunition), variable chemical composition and the relatively high cost for manual collection (in
comparison to the cost for manufacture) that the waste
ammunition material (WAM) exhibits. Hence spent ammunition (both crushed bullets, fragments and cartridge cases)
is more likely to be left in the shooting range for a future
disposal in dumps or sanitary landfills. Therefore from
the above remarks it is conceivable that WAM has to be
considered as a potential environmental hazardous waste
material and precautions should be taken considering its
disposal in the environment. Despite of the extended literature on remediation technologies for industrial wastes
prior to disposal in landfills or directly on site [8], little published data are available concerning the fate of WAM. One
option, apart from recycling, could be the use of the waste
ammunition as a secondary raw material for the production of Portland cement clinker.
The use of alternative raw materials in cement industry
that can be used as mineralizers or fluxes, in order to
improve the burnability of the raw mix, and the effect of
minor components on the formation of the clinker has been
broadly discussed and reviewed adequately [9–18]. Previous
relative work from the author has led to the conclusion
that certain foreign elements, despite their low concentration, exert a remarkable effect on the sintering process by
improving the reactivity of the raw mix and affecting the
crystal structure and texture of clinker minerals, and the
hydration reactions [17–25]. It must be noted, however,
that when several elements are jointly present in the
mixture, the reactivity is not additively changed and any
secondary material that is expected to have a positive effect
on the reactivity must be individually investigated.
This work aims to investigate the capability to substitute
cement raw mix with WAM in cement clinker production,
as a mineralizing additive. The addition of WAM in the
cement raw meal will introduce varying concentrations of
minor foreign elements, such as Pb, Cu, Zn, Sn, Sb etc.,
hence the possible use of that waste material in industrial
scale will cause environmental concerns with regard to
heavy metal emissions, especially for Pb, Cu and Zn. In
previous published works it was well established that lead
compounds are quite volatile under sintering conditions
[18,26–30] but small amounts can still be retained in clinker. Pb oxide acts as a retarder during hydration [13]. On
the other hand Cu in the form of CuO acts as mineralizer
by decreasing the melt temperature by at least 50 C, even
for 0.5% w/w addition in the raw meal, and favors the combination of free lime even at 1100 C [31]. Similarly, the
addition of ZnO improves the burning behavior of the
raw mix and accelerates the clinker formation. ZnO is also
known to act as a mineralizer, leading to higher activity
and promoting the solid reactions as well as the formation
of alite by increasing the amount of liquid phase [32–34].
To avoid the aforementioned problems, a varying content
of three blends prepared by properly treated WAM types,
not exceeding 2.0% w/w, was introduced in a conventional
cement raw meal. The reactivity of the modified mixtures,
the sintering reactions and the structure of selected sintered
products were investigated. In addition, the effect of the
added waste material on the hydration rate and some of
the mechanical–physical properties of the produced
cements were examined. Additionally, leaching tests were
carried out in order to assess any potential environmental
risk from a possible application of a cement originated
from such modified clinkers.
2. Experimental
2.1. Materials
Three different types of WAM were used in this study:
Type 1 (T1) was manually collected from ten individual
butt stoppers of 10 m paper target devices, and corresponds to 4.5 mm (0.17700 ) caliber crushed airgun pellets
and fragments. Type 2 (T2) was manually collected from
ten individual butt stoppers of 25 and 50 m paper target
devices, and corresponds to 5.6 mm (0.2200 ) and 7.62–
9.65 mm (0.30–0.3800 ) caliber crushed bullets and fragments. Type 3 (T3) corresponds to 5.6 mm (0.2200 ) and
7.62–9.65 mm (0.30–0.3800 ) caliber cartridge cases that were
manually collected from the firing points of the 25 and
50 m outdoor ranges. For all types of WAM, the bulk
material (roughly 1 kg) was immersed in an ultrasonic bath
with HNO3 1.0% for 10 min, in order to remove the soluble
Table 1
Average chemical composition (% w/w) of different types of waste
ammunition material
Element (% w/w)
T1a
T2b
T3c
Pb
Sn
Cu
Zn
Sb
Fe
Al
Ni
Cr
As
Hg
Bi
99.06
0.01
–
–
0.64
–
–
0.01
–
–
–
–
96.38
2.52
0.12
0.03
0.83
0.006
0.004
0.001
0.001
0.01
0.001
–
0.06
0.05
70.32
27.91
0.02
0.03
0.02
0.20
0.03
0.001
0.004
0.001
a
b
c
Type 1 WAM.
Type 2 WAM.
Type 3 WAM.
135
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
Table 2
Characteristics of the industrial cement raw mix and the produced clinker
Raw mix
Clinker
Chemical
composition
(% w/w)
Potential mineral
composition
(Bogue) (% w/w)
Moduli
C3 S
C2 S
C3 A
C4AF
LSFb
SMc
ARd
HMe
SiO2
Al2O3
Fe2O3
CaO
MgO
K2O
LOIa
a
b
c
d
e
13.76
3.23
2.45
43.11
0.55
0.28
35.62
71.1
7.7
6.9
11.6
0.981
2.42
1.32
2.22
LOI = loss on ignition.
LSF = lime saturation factor.
SM = silica modulus.
AM = alumina modulus.
HM = hydraulic modulus.
WAM types in a laboratory swing mill for 1 h, in order to
be employed as secondary raw material for clinker production. The selected proportions of T1–T3 used for the production of the three blends, as well as their codification
are given in Table 3. Table 3 gives also the composition
and sample codification of the modified raw mixtures prepared for this study. One reference and twelve modified
mixtures were prepared by mixing the reference sample
with 0.5%, 1.0%, 1.5% and 2.0% w/w of B1–B3 WAM
blends in a laboratory swing mill for 1 h. Homogeneity
was ascertained by dosing the added elements, in some
indicative samples.
2.3. Burning procedure
impurities (residues from the primer mixture, paper target
pieces, dust, soil etc.), rinsed afterwards with deionized
water and then dried at 105 C for 24 h in an laboratory
electrical oven. The dried material was ground in a laboratory ultrafine attrition mill, to a particle size <90 lm. The
chemical composition of each type of ground WAM was
obtained using atomic absorption spectroscopy (AAS)
and is presented in Table 1. Ordinary Portland cement
raw meal of industrial origin was used (residue at 90 lm:
15%). The chemical composition of the raw mix as well
as the mineral composition (according to Bogue) and the
moduli of the obtained clinker are presented in Table 2.
2.2. Sampling strategy and preparation
Three different blends of WAM (B1, B2 and B3 respectively) were prepared, by intensively mixing T1, T2 and T3
All samples were pressed to form pellets, placed in a
Pt-dish and were thermally treated at 1100, 1200, 1300,
1350, 1400 and 1450 C for 20 min in an electrical furnace
and then were rapidly cooled. Sintering and cooling conditions were kept strictly constant.
2.4. Measuring and monitoring techniques
The sintering reactions in all modified raw mix samples
were recorded by means of differential thermal analysis
using a Mettler Toledo TGA/SDTA 851 instrument. The
temperature was raised at a constant rate (10 C/min) from
ambient to 1450 C. The experiments were conducted in a
static atmosphere. The effect on the burnability was evaluated on the basis of the unreacted lime content in samples
sintered at the above-mentioned temperatures. The sintered
pellets were ground and analyzed by the ethylene glycol
method [35,36] in order to estimate the free CaO (fCaO)
content in the final sintering products and using a Siemens
Table 3
Codification and mixture compositions (% w/w) of the produced WAM blends and modified raw mixes
WAM Blends
BD1a
BD2b
BD3c
35% T1, 35% T2, 30% T3
25% T1, 25% T2, 50% T3
15% T1, 15% T2, 70% T3
M1-05
M1-10
M1-15
M1-20
0.5% BD1,
99.5% OCRMd
1.0% BD1,
99.0% OCRM
1.5% BD1,
98.5% OCRM
2.0% BD1,
98.0% OCRM
M2-05
M2-10
M2-15
M2-20
0.5% BD2,
99.5% OCRM
1.0% BD2,
99.0% OCRM
1.5% BD2,
98.5% OCRM
2.0% BD2,
98.0% OCRM
M3-05
M3-10
M3-15
M3-20
0.5% BD3,
99.5% OCRM
1.0% BD3,
99.0% OCRM
1.5% BD3,
98.5% OCRM
2.0% BD3,
98.0% OCRM
Raw mixes
a
b
c
d
BD1 = blend 1.
BD2 = blend 2.
BD3 = blend 3.
OCRM = ordinary cement raw meal.
136
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
D-5000 X-Ray Diffractometer, with nickel-filtered Cu Ka1
radiation (k = 1.5405 Å), in order to identify the mineralogical phases formed during sintering.
The content of the added minor elements in the samples,
sintered at 1450 C, was determined, using atomic absorption spectroscopy (AAS), in order to evaluate the incorporation degree of the main foreign elements in clinker
phases. Prior to measurement, 1 g of each sample was dissolved according to Maczkowski method.
Scanning electron microscopy was used in selected samples in order to examine the texture of the obtained clinkers
at 1450 C and the distribution of the foreign elements in
their main phases. A JEOL JSM-5600 scanning electron
microscope, interfaced to an OXFORD LINK ISIS 300
energy dispersive X-ray spectrometer (EDXS) was used.
Experimental conditions involved 20 kV accelerating voltage and 0.5 nA beam current.
2.5. Cement hydration and properties
The clinkers were interground with 5.0% w/w gypsum in
a pro-pilot plant ball mill of 5 kg capacity. The gypsum was
of industrial origin (98% w/w Ca2SO4 Æ 2 H2O). The fineness of all the produced cements was found to be in the
range between 3500 and 3700 cm2/g (Blaine). The compressive strength of the samples (EN 196-1) as well as the
consistency of standard paste, the setting time and the
soundness (EN 196-3) was determined.
The cements were mixed with water in order to prepare
cement pastes. A water-to-cement ratio (W/C) of 0.4 was
retained for all pastes and deionized water was used. After
a period of 6 h the cement cores were put in polyethene
containers (vials), sealed hermetically and wet-cured at
20 C. Samples hydrated for periods of 1, 2, 7, and 28 days
were subjected to acetone and isopropyl ether treatment
and then dried for 24 h in vacuum. The hydrated samples
were ground to pass through a 54 lm sieve and were studied by means of XRD in order to identify the hydrated
products. In addition, thermogravimetry analysis (TGA)
was used for the evaluation of the hydration rate. The samples were heated from 20 to 600 C at a constant rate of
15 C/min in an atmosphere of carbon dioxide free nitrogen, flowing in 50 cm3/min. A limited number of samples
were treated up to 1000 C in order to find out if there
was any carbonation of Ca(OH)2.
2.6. Leaching tests
The procedure used for the leaching tests was the ANSI/
ANS-16.1-1986. The leachant was deionized water with a
conductivity <6 lX/cm in which the specimens, with a volume of leachant to external geometric surface area of the
specimen (195 cm2) ratio of 10 cm, were immersed in individual plastic vials for 1, 2, 7, 28 and 90 days. Plastic vials
were kept in watertight containers full of water, which were
stored in a curing chamber, thus assuring maximum possible humidity. When leaching time was over, leachates were
extracted from the plastic containers and put in refrigerator
to avoid evaporation. The concentration of the main heavy
metals Pb, Cu, Zn, Sn and Sb (presented in the modified
raw mixtures) was determined by induction-coupled
plasma emission spectrometry (ICP). pH was kept constant
at the range of 7.0 by feed-back control and addition of
0.1 N HNO3 or 0.1 N NaOH solutions.
3. Results and discussion
3.1. WAM characterization
Although the results derived from chemical analysis of
T1, T2 and T3 types concern a waste material that is found
in Greece, the basic conclusions are applicable and results
of the same magnitude are expected in relevant cases in
other countries of the European Union, since the same
ammunition types are widely used across Europe and
USA for the same purposes. Additional chemical analyses
carried out in WAM, collected from the same shooting
ranges in different posterior periods, showed a rather consistent composition of the WAM in the main constituent
elements.
As it can be seen from the results of Table 1, the addition of T1 and T2 WAM in the raw material will introduce
significant quantities of lead as the main foreign element,
while T3 will introduce significant quantities of Cu and
Zn. In the high oxidizing-high temperature kiln environment most of the heavy metals form oxides. Since Pb exhibits high volatility during sintering, in order to avoid high
lead emissions and to enhance the mineralizing effect of
Cu, Zn and other trace elements, it was decided to prepare
blends of T1–T3 WAM, enriched in T3, and to limit their
use in the cement raw meal to fractions that will not exceed
2.0% w/w.
The production of these three blends containing different proportions of Pb, Cu, Zn, Sn and Sb elements is an
easy way to fix problems encountered from the variability
of WAM chemical composition.
3.2. Thermal studies of modified clinkers
In the DTA curve of a cement raw mix the following
stages are distinguished:
• a broad endothermic effect attributed to the dehydroxylation of clays (200–600 C),
• a strong endothermic effect around 900 C caused by the
decomposition of the limestone,
• one or more exothermic reactions (1200–1350 C) connected with the belite (C2S) formation,
• an endothermic reaction just after the last exothermic
reaction (1300–1400 C) associated with the partial
melting of the sample.
The first two stages depend on the chemical and mineralogical characteristics of the clay and the limestone respec-
137
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
tively, and as it was expected, they are identical in all samples. The most important stages are the third and the
fourth which are directly associated with the clinkerization
process.
Fig. 1 presents the DTA curves of the pure sample and
the modified raw mixes containing 2.0% w/w of the three
waste ammunition blends.
The evaluation of the DTA curves of all modified raw
meals led to the following remarks:
• In all modified samples the reactions associated with the
decomposition of CaCO3 (in the temperature range of
800–900 C) and clinkerization (1200–1450 C) are
recorded, suggesting satisfactory burning and clinkerization of all samples.
• The added WAM blends do not affect the decomposition of CaCO3 and the belite formation. In all modified
mixtures the formation of the melt is shifted to lower
temperature and overlapped with the belite formation.
This fact indicates that the constituents of the added
WAM blends are dissolved in the liquid phase, affecting
mainly the formation and the properties of the melt and
therefore change the reactivity of the mixture at high
temperatures.
3.3. Burnability of the raw mix
The effect of the added WAM blends on the overall burnability of the raw mixtures was evaluated on the basis of
the unreacted lime (fCaO) content after sintering at varying
temperatures. Table 4 presents the fCaO content in relation
to the sintering temperature and the WAM blend content
in the modified raw mixes. Using the fCaO values of the
sintered samples, the burnability capacity (BC) of the raw
mix can be estimated from Eqs. (1) and (2) [37].
C ¼ 2fCaO1100 C þ 2fCaO1200 C þ 3fCaO1300 C
þ 4fCaO1350 C þ 4fCaO1400 C þ 2fCaO1450 C
ð1Þ
BC ¼ 600=C
ð2Þ
The resulted values of BC are also given in Table 4. The
ratio of the burnability capacity of the sintered samples
to the burnability capacity of the pure sample, in relation
to the doping WAM blend concentration in the raw mix
is presented in Fig. 2. BC ratio greater than 1 indicates that
the added mineral favors the sintering process.
Table 4
fCaO and BC values of sintered products
Raw mix
BCa
fCaO (%)
Temperature ( C)
1100
1200
1300
1350
1400
1450
Reference
M1-05
M1-10
M1-15
M1-20
39.71
39.67
39.41
38.72
37.70
31.77
29.84
29.07
28.52
27.11
14.47
13.12
12.51
11.71
10.33
8.62
8.10
7.61
7.52
6.81
4.88
4.11
4.00
3.76
3.43
1.99
1.77
1.52
1.34
1.10
2.46
2.60
2.68
2.76
2.94
M2-05
M2-10
M2-15
M2-20
39.51
39.11
38.41
37.10
29.61
28.07
26.52
24.11
12.91
12.09
9.86
8.11
7.67
7.02
5.85
5.12
3.86
3.61
2.86
2.27
1.53
1.39
0.91
0.63
2.65
2.78
3.18
3.52
M3-05
M3-10
M3-15
M3-20
39.07
38.21
36.70
35.10
21.43
20.11
17.38
15.45
11.51
10.41
6.33
5.71
5.55
4.31
2.15
1.58
2.51
1.84
1.52
0.62
1.01
0.77
0.31
0.17
3.16
3.45
4.31
4.72
a
BC = burnability capacity.
2.20
2.00
BC ratio
1.80
M1
M2
M3
1.60
1.40
1.20
1.00
0.80
0.0
0.5
1.0
1.5
2.0
WAM doping concentration (%w/w)
Fig. 1. DTA curves of (a) pure, (b) M1-20, (c) M2-20, and (d) M3-20
modified samples.
Fig. 2. BC ratio in relation to the waste ammunition material content in
the modified cement raw mixes.
138
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
As it is seen (Table 4), the addition of the WAM
blends in the conventional cement raw meal improves
the burnability of the raw mix. Surprisingly, a positive
effect is recorded even from 1100 C, attributed to the
presence of Pb, Cu, Zn, Sn and Sb in the raw mix. A significant decrease in the fCaO content is recorded at 1450,
1400 and 1350 C (by 11–45%, 16–30% and 6–21% in M1,
23–68%, 21–53% and 11–41% in M2 and 49–91%, 49–87%
and 36–82% in M3, respectively), implying a remarkable
impact of the added blends on the reactions that proceed
in the presence of the liquid phase. This is also reflected
on the BC values (Fig. 2), which are mainly based on
the fCaO content at 1400 and 1350 C. On the overall,
even 0.5% w/w addition of any WAM blend is considered
to improve the burnability of the raw mix, even in low
temperatures. The positive effect of the waste material
seems to be somehow proportional to the doping concentration. Optimum results were achieved in the case of M3
modified raw mixes, followed by M2 and M1, where the
Cu and Zn fractions in the raw meal were significantly
higher.
There is no previous literature on the effect of Pb, Sn
and Sb compounds on the reactivity of the cement raw
mix. The mineralizing action of CuO and SnO2 seems to
be related to their incorporation in solid solutions with calcium oxide and/or an acceleration of the reactions among
the four main oxides, rather than the decomposition of calcium carbonate. The dark colour of all sintered samples,
even at 1100 C, indicates a high degree of combination
of the main constituents and especially of Fe2O3. Previous
study by the author reported improvements in the reactivity of the CaO–SiO2–Al2O3–Fe2O3 system, already at
1100 C, by the addition of 1.0% w/w CuO in the synthetic
raw mix [18]. In fact, Odler and Abdul-Maula [38,39] as
well as Kakali et al. [31] have found CuO to be an efficient
mineralizer and flux at 1450 C, but in a lesser extend, compared with 1200 C. ZnO is acting both as mineralizer and
flux, by accelerating clinker formation, leading to higher
activity and promoting the solid reactions as well as the
formation of alite by increasing the amount of liquid phase
[32–34].
3.4. Incorporation degree of the minor elements in clinker
The incorporation degree of the added foreign elements
Pb, Cu, Zn and Sn was obtained by comparing the theoretical content with the data from the chemical analysis in
samples sintered at 1450 C. The theoretical content was
calculated using the data from the chemical analysis of
the T1–T3 WAM types and the ordinary cement raw meal,
taking into account the measured loss on ignition (LOI).
Results are given in Table 5.
The contents of lead measured in M1–M3 clinkers show
an average incorporation degree of 46%, with a maximum
of 60%, a value that should be considered high, in comparison to relative published data [26–30]. As lead content
decreases from M1 to M3 samples, the incorporation
Table 5
Incorporation degree (%) of Pb, Cu, Zn and Sn in clinker
Sintered sample
Incorporation degree (%)
Pb
Cu
Zn
Sn
M1-05
M1-10
M1-15
M1-20
41.8
37.1
36.3
28.6
100.0
100.0
98.9
99.4
100.0
98.4
97.5
94.2
100.0
100.0
100.0
99.4
M2-05
M2-10
M2-15
M2-20
53.1
48.3
45.9
43.7
97.6
98.2
95.3
90.3
95.3
91.1
90.0
86.2
100.0
99.5
99.3
99.7
M3-05
M3-10
M3-15
M3-20
60.2
56.2
57.3
55.6
93.3
90.1
85.8
83.9
89.1
82.8
77.6
73.3
99.9
99.6
99.1
98.9
degree for lead seems to be stabilized. On the other hand
the incorporation degree for copper and zinc fluctuates
between 84–100% and 73–100% respectively. Tin and antimony seem not to be volatilized at all.
It must be noted that the thermal treatment of the samples was performed as a batch operation in a laboratory
electrical furnace. In industrial practice the evaporation
of heavy metals depends on the composition of the raw
meal, burning conditions and atmosphere (oxidizing or
reducing) during the burning process. Even though in
many cases the presence of alkalies or chlorine in industrial
raw meals increases the volatility of certain elements that
form volatile chlorides or alkali salts which participate in
the so-called alkali cycle, their retention within clinker
phases depends upon the process type and the extent
to which precipitator/bag filter dust is returned to the
kiln. Thus, in a suspension preheater kiln with nearly
100% return of dust to the kiln and a dust collecting system
meeting current environmental requirements, an almost
complete retention of such elements could be accomplished.
3.5. XRD studies of modified clinkers
All samples sintered at 1450 C have the structure of a
typical clinker. The dominant phases (alite, belite, calcium
aluminate and ferrite) were well crystallized, giving peaks
at the common 2h values. Additionally, no detection of
undesired compounds (c-C2S) was recorded. Differentiations among samples were restricted in calcium aluminate
and ferrite, due to the selective incorporation of the constituents of the WAM blends in these phases and the changes
that occur in their chemical composition and structure.
Indications concerning the formation of PbSiO3 phase were
recorded in clinker matrices sintered at 1450 C that correspond to M1 and M2 raw mixtures, enriched in B1 and B2
WAM blends. Similar indications concerning the partial
formation of the compound CaCu2O3 were also recorded
in the XRD patterns of samples doped with more than
1.0% w/w of the B3 WAM blend.
139
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
3.6. SEM studies of modified clinkers
SEM studies were conducted in order to examine the
structure of selected clinkers obtained and the distribution
of the foreign elements in their main phases. The doping
concentration of 2.0% w/w was chosen in order to
strengthen the effect of the added foreign elements on the
structure of clinker and facilitate their detection by SEM
in the clinker compounds. The SEM observations are summarized in Table 6, while the SEM photos of the pure and
modified clinkers originated from raw mixes doped with
2.0% w/w B1, B2 and B3 WAM blends, burned at
1450 C, are given in Fig. 3. Codification for clinkers simply follows the codification for the corresponding modified
raw mixes.
The photos were selected to be representative as far as
the size and the shape of alite and belite crystals are concerned. With the spot analysis of EDXS, the composition
of the principal phases was analyzed at five to eight points.
Owing to instrumental limitations, accuracy in analyses of
the minute grains in the interstitial phases proves difficult
to achieve. In any case, the comparison of the spot analysis
in each clinker phase can lead to only qualitative indications concerning the distribution of the minor elements in
the individual clinker compounds.
From Fig. 3 it is concluded that the addition of the three
WAM blends results in variations in the appearance of
both alite and belite crystals that seem to be enhanced as
the raw mix is enriched in Cu and Zn. Alite was developed
in large compact crystals, which tend to appear more prismatic and angular in shape, especially as the doping concentration of Pb is lowered (M2-20 and M3-20 sintered
samples), in contrast to the slightly angular hexagonal outline of C3S grains in reference and M1-20 samples. Belite as
well was uniformly distributed, forming bigger in size and
rounded in shape crystals. The amount of interstitial
matrix in all tested samples was generally adequate, having
a fine-crystalline structure.
In contrast to Pb, which was preferentially found in the
silicate phases (especially in alite), EDXS analyses showed
Table 6
Size and shape of alite and belite crystals in modified clinkers
Sintered sample
Alite
Belite
Size (lm)
Shape
Size (lm)
Shape
Reference
20–30
Compact, prismatic, with slightly rounded outline
5–10
M1-20
M2-20
30–40
40–60
Larger in size, compact, rounded at the rims
Compact, prismatic, with slightly angular hexagonal outline
5–10
5–15
M3-20
40–60
Angular, rather elongated, prismatic
5–15
Small,
roundish
Round
Bigger in
size, round
Bigger in
size, round
Fig. 3. SEM photos of (a) pure, (b) M1-20, (c) M2-20, and (d) M3-20 modified clinkers.
140
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
that Cu, Zn, Sn and Sb are selectively incorporated in the
interstitial material, as their doping concentration in the
modified raw mixtures is increased. EDXS analyses performed on selected spots at the rims of alite crystals indicated a high Pb and Si content (PbO:SiO2 approximately
equals 1:1) in M1-20 sintered sample and a high Cu and
Ca content (CaO:CuO approximately equals 1:2) in M320 sintered sample, also seen in the XRD patterns of that
particular samples. However the low content of PbSiO3
and CaCu2O3 compounds and the overlapping of the clinker peaks do not permit a safe identification.
3.7. Hydration rate of cement pastes
All pastes were studied by means of a TG analyzer. The
weight loss up to 550 C, which corresponds to the total
water incorporated in the cement paste, was determined.
The Ca(OH)2 content, which is directly related to the
hydration of silicate compounds, was also measured. The
weight loss in the range 600–700 C, if any, corresponds
to the decomposition of CaCO3 and it has to be converted
to the equivalent Ca(OH)2 and then to equivalent H2O.
The carbonation of the pastes may take place during the
preparation of the paste or during the grinding of the paste
previous to the TG measurement. The water combined in
the hydration products (other than calcium hydroxide) corresponds to the weight loss up to 300 C. Any changes of
this amount indicate that the kind and/or the stoichiometry
of the hydration products are changed.
Table 7 presents the content of the total compound
water, calcium hydroxide and water in the hydration products in the case of C3 pastes (originated from M3 clinkers),
in relation to the age of hydration.
Table 7
Total compound water, calcium hydroxide and water in the hydration
products, in relation to the age of hydration for C3 cement pastes
Cement paste
Hydration age (days)
1
2
7
28
–
C3-05
C3-10
C3-15
C3-20
Compound water (% w/w)
15.47
17.97
16.90
17.10
16.69
18.75
15.44
15.55
18.01
14.36
14.92
17.33
14.02
15.63
16.75
23.55
24.88
23.88
22.72
22.35
–
C3-05
C3-10
C3-15
C3-20
Ca(OH)2 (% w/w)
20.39
21.49
20.56
17.02
19.61
18.75
18.64
20.51
16.03
19.45
26.68
24.98
24.17
23.76
23.27
–
C3-05
C3-10
C3-15
C3-20
Water in hydrated products (% w/w)
10.51
12.74
11.01
12.10
12.55
12.93
10.67
10.99
12.32
9.82
9.93
11.61
10.12
10.90
11.47
24.21
23.93
23.39
23.53
21.72
17.06
18.80
18.00
16.94
16.69
As it was extracted from the evaluation of the TG tests,
small additions of the added B1–B3 WAM blends (up to
1.0% w/w) do not significantly affect the total rate of
hydration. As the waste material content in the raw mix
increases, the hydration tends to slow down, as it is indicated by the progressive decrease of water and calcium
hydroxide content in all pastes. The stronger effect of the
added materials on the calcium hydroxide content indicates
that they affect especially the hydration of silicate compounds. When heavy elements, such as Pb, Cu and Zn,
are present in cement, the early hydration reactions are
inhibited, probably through the formation of amorphous
compounds that cover the unreacted cement grains. The
hydration reactions start again after the crystallization
and sedimentation of these compounds, a phenomenon
that takes place when the Ca2+ and OH concentration
becomes high enough [40–42]. The hydration process, however, of cements produced by clinkers containing the above
elements is different, since their incorporation in the crystal
lattice of clinker minerals delays their dissolution rate during hydration.
3.8. XRD studies of hydrated samples
In all samples, a gradual increase of Ca(OH)2 content,
accompanied by a corresponding decrease of the anhydrous calcium silicate compounds was observed. The formation of ettringite (Ca6Al2(SO4)3(OH)12 Æ 26H2O), which
is the first reaction product between calcium aluminates
and gypsum, is confirmed in all samples even after 1 day
of hydration. Differentiations are observed after the age
of 2 days. The transformation of ettringite to the monosulfate salt (Ca4Al2SO10 Æ 12H2O) seems to be delayed, proportionally to the percentage of each WAM blend
addition in the raw meal. The most likely reason for the
delay is the reduced rate of hydration of the aluminate
phase. This extends the period of time required to exhaust
the CaSO4 supply. The monosulfate was initially detected
in the pure and the 0.5% w/w doped samples after 2 days
of hydration, in the 1.0% and 1.5% w/w doped samples
after 7 days and in the 2.0% w/w doped sample after 28
days. The rest of the calcium silicate and calcium aluminate
hydration products do not present distinct peaks due to
their low degree of crystallinity and/or their small amounts.
3.9. Cement properties
Table 8 presents the compressive strength development
and setting time of the produced cements in relation to
the content of the added blends of waste material.
The cements prepared from the corresponding modified
clinkers present almost the same initial strengths (1 day),
while a positive effect is observed at 2, 7 and 28 days. From
all cement samples studied, those that came from B3 blend
presented a higher strength development than the reference
one. The 1.0% w/w addition seems to have the optimum
effect in all three cement types, while in higher percentages
141
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
Table 8
Compressive strength development (N/mm2) and setting time (min) of
cements in relation to the percentage of the added waste ammunition
material blends
Sample
Compressive strength (N/mm2)
Setting time (min)
Age (days)
1
2
7
28
Initial
Final
Reference
C1-05
C1-10
C1-15
C1-20
14.6
14.6
14.9
14.7
14.3
23.0
23.2
23.6
23.3
24.4
38.9
39.4
40.8
40.0
39.9
53.3
53.9
54.5
54.1
53.9
95
97
101
112
118
145
152
156
165
167
C2-05
C2-10
C2-15
C2-20
14.1
14.9
14.4
14.3
23.5
24.7
24.5
24.3
39.5
40.5
39.7
38.6
54.0
54.9
54.8
54.4
97
100
106
110
141
146
152
161
C3-05
C3-10
C3-15
C3-20
14.4
15.0
14.7
14.7
26.1
28.3
27.1
26.4
42.2
43.7
42.4
41.9
55.1
57.8
56.5
55.8
90
95
99
104
143
151
157
160
varies from 0.8 mm to 1.9 mm while the limit according to
EN 197-1 is 10 mm.
3.10. Leaching tests
Table 9 presents results obtained from the leaching tests.
It is confirmed that the clinker matrix is capable of stabilizing Pb to a great extent, since concentrations of such
element found in leachates are in the range of parts per
billion (ppb). Similar to that, very little Cu, Zn and Sn
are found in the leaching solutions, even in the age of 90
days. No antimony, mercury, cadmium, chromium, arsenic
or bismuth were detected in the leachates corresponding to
cements with the highest WAM blend doping addition
(2.0% w/w), in all ages. The above behavior is attributed
to the high degree of incorporation of the added heavy
metals in the crystal lattice of clinker minerals and the
delay of their dissolution rate during hydration.
4. Conclusions
of addition (1.5% and 2.0% w/w) the positive effect seems
to be inhibited.
The addition of the WAM blends leads to a slight
increase of the setting time proportionally to the WAM
blend content. This may be due to formation of insoluble
Pb, Cu and Zn hydroxides that probably act as retarders
during the hydration process by coating the silicate grains
and prevent further hydration.
The addition of the secondary waste materials has no
effect on the water demand of the cement paste. The standard consistency of all pastes varies from 25.9% to 26.5%
in all samples.
The soundness of the tested cements was satisfactory.
The expansion measured according to Le Chatelier method
The disposal of waste ammunition materials deriving
from recreational shooting and/or military activities should
be considered as a serious problem. The well-known and
generally used methods for waste disposal such as open
dumping and landfilling present various disadvantages.
With the present technique, almost all quantities of spent
ammunition can be incorporated into the raw materials
for the manufacture of Portland cement clinker with practically complete elimination of toxic impact in environmental matrices during disposal and without affecting the
quality of the final cement product. The waste ammunition
blends may be used successfully as a secondary raw material for partial substitution of the conventional raw materials, leading to resources conservation and ensuring the
quality of the produced cement.
Table 9
Concentration (mg/l) of Pb, Cu, Zn and Sn in leachates
Paste
Hydration age (days)
Hydration age (days)
1
2
7
28
90
1
2
7
28
90
C1-10
C1-20
C2-10
C2-20
C3-10
C3-20
Pb (mg/l)
0.039
0.063
0.022
0.046
0.012
0.022
0.043
0.099
0.033
0.068
0.017
0.033
0.076
0.165
0.049
0.076
0.028
0.051
0.150
0.223
0.099
0.124
0.039
0.091
0.197
0.245
0.140
0.176
0.087
0.111
Cu (mg/l)
n.d.a
0.011
n.d.
n.d
0.011
0.011
n.d.
0.019
n.d.
0.018
0.017
0.033
0.011
0.016
0.015
0.037
0.028
0.051
0.023
0.033
0.029
0.047
0.039
0.091
0.021
0.042
0.040
0.059
0.087
0.131
C1-10
C1-20
C2-10
C2-20
C3-10
C3-20
Zn (mg/l)
n.d.
n.d.
n.d
n.d.
n.d
0.010
n.d.
n.d.
n.d
n.d.
0.010
0.019
n.d.
n.d.
0.016
0.019
0.018
0.020
0.011
0.015
0.024
0.031
0.041
0.066
0.016
0.023
0.029
0.042
0.063
0.072
Sn (mg/l)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
0.014
0.015
n.d.
n.d.
n.d.
n.d.
0.017
0.029
0.015
0.019
n.d.
n.d.
a
n.d. = not detected.
142
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
From the present study the following conclusions can be
drawn:
• The addition of the waste ammunition material promotes the consumption of free lime and improves the
burnability of the raw mix, especially during the final
stage of the sintering. This positive effect is proportional
to the doping concentration.
• The contents of lead measured in the modified clinkers
show an average incorporation degree of 46%, a value
that should be considered high, in comparison to relative published data. Almost all amounts of Cu, Zn, Sn
and Sb in the modified raw mixes are retained in clinker.
• Except from Pb, which was preferentially found in the
silicate phases, Cu, Zn, Sn and Sb are mainly concentrated in the melt, affecting the growth environment of
alite crystals and modifying their shape and size.
• Despite the slight retarding effect on the hydration process, the added waste ammunition blends improve
strength development and do not affect the physical
properties of the cements. Optimum results were
obtained in the case of the Cu and Zn enriched blend,
in the doping concentration of 1.0% w/w.
• The fixation of the heavy metals, present in the waste
ammunition material, in the clinker minerals, led to considerably low concentrations detected in the leachates
for all ages tested.
References
[1] Brunelle RL, Hoffman CM, Snow KB. Comparison of pistol bullets
by atomic absorption: preliminary study. J Assoc Off Anal Chem
1970;53(3):470–4.
[2] Peters CA, Havekost DG, Koons RD. Multielemental analysis of
bullet lead by inductively coupled plasma-atomic emission spectrometry. Crime Lab Digest 1988;15(2):33–8.
[3] Peele ER, Havekost DG, Peters CA, Riley JP, Halberstan RC.
Comparison of bullets using the elemental comparison of lead
component. In: Proc int symposium on the forensic aspects of trace
evidence; 1991. p. 57–68.
[4] Blacklock EC, Sadler PA. Shot-gun pellet identification and discrimination. Forensic Sci Int 1978;12(2):109–17.
[5] Guinn VP. JFK assassination: bullet analyses. Anal Chem
1979;51(4):484A–93A.
[6] Dufosse T, Touron P. Comparison of bullet alloys by chemical
analysis: use of ICP-MS method. Forensic Sci Int 1998;91(3):197–
206.
[7] Levonmäki M, Kairesalo T. Do steel shots raise a chromium problem
on shooting range areas? ISSF News 2002;2:9–10.
[8] Tchobanoglous G, Theisen H, Vigil S. Integrated solid waste
management. New York: McGraw Hill; 1993.
[9] Blumentahl M. The use of scrap tyres in the US cement industry.
World Cem 1992;45(12):14–20.
[10] Dawson B. Emerging technologies for utilising wastes in cement
production. World Cem 1992;45(12):22–4.
[11] Sprung S. Relieving the environment through the utilization of
secondary raw materials. Zem-Kalk-Gips 1992;45(5):213–21.
[12] Liebl P, Gerger W. Benefits and limitations when using secondary
materials. Zem-Kalk-Gips 1993;46(10):632–8.
[13] Bhatty JI. Role of minor elements in cement manufacture and
use. Skokie, Illinois: Portland Cement Association; 1995.
[14] Timashev VV. The kinetics of clinker formation. The structure and
composition of clinker and its phases. In: Proc 7th international
congress on the chemistry of cement. Paris, France, vol. 1; 1980. p.
I-3/1–I-3/20.
[15] Bucchi R. Features on the role of minor compounds in cement
clinker. Part I. World Cem 1981;12(6):210–31.
[16] Moir GK, Glasser FP. Mineralisers, modifiers and activators in the
clinkering process. In: Proc 9th international congress on the
chemistry of cement. Delhi, India, vol. 1; 1992. p. 125–52.
[17] Kolovos K, Tsivilis S, Kakali G. The effect of foreign ions on the
reactivity of the CaO–SiO2–Al2O3–Fe2O3 system. Part I: anions. Cem
Concr Res 2001;31(3):425–9.
[18] Kolovos K, Tsivilis S, Kakali G. The effect of foreign ions on the
reactivity of the CaO–SiO2–Al2O3–Fe2O3 system. Part II: cations.
Cem Concr Res 2002;32(3):463–9.
[19] Tsivilis S, Kakali G, Stamatakis M, Kolovos K, Voglis N, Choupa K,
et al. Use of selected minerals as secondary raw materials in cement
industry. Part I: effect on the burnability of raw mix. Key Eng Mater
2001;206–213:1891–4.
[20] Tsivilis S, Kakali G, Stamatakis M, Kolovos K, Voglis N, Choupa K,
et al. Use of selected minerals as secondary raw materials in cement
industry. Part II: effect on clinker structure. Key Eng Mater
2001;206–213:1895–8.
[21] Kakali G, Tsivilis S, Kolovos K. Factors affecting the reactivity of the
CaO–SiO2–Al2O3–Fe2O3 mixture. Key Eng Mater 2001;206–
213:1899–902.
[22] Perraki M, Perraki Th, Kolovos K, Kakali G, Tsivilis S. Secondary
raw materials in cement industry. Evaluation of their effect on the
sintering and hydration processes by means of thermal analysis. J
Therm Anal Cal 2002;70(1):143–5.
[23] Kolovos K, Tsivilis S, Kakali G. SEM examination of clinkers
containing foreign elements. Cem Concr Comp 2005;27(2):163–70.
[24] Kakali G, Tsivilis S, Kolovos K, Voglis N, Aivaliotis J, Perraki T,
et al. Use of secondary mineralising raw materials in cement
production. The case study of a wolframite–stibnite ore. Cem Concr
Comp 2005;27(2):155–61.
[25] Kakali G, Tsivilis S, Kolovos K, Choupa K, Perraki T, Perraki M,
et al. Use of secondary mineralising raw materials in cement
production. The case study of a stibnite ore. Mater Lett
2003;70(20):3117–23.
[26] Krčmar W, Linner B, Weisweiler W. Investigations into the behaviour of trace elements during clinker burning in a rotary kiln system
with grate preheater. Zem-Kalk-Gips 1994;47(10):600–5.
[27] Weisweiler W, Krčmar W. Heavy metal balances for a cement kiln
plant with grate preheater. Zem-Kalk-Gips 1990;43(3):142–52.
[28] Sprung S, Rechenberg W. Levels of heavy metals in clinker and
cement. Zem-Kalk-Gips 1994;47(5):258–63.
[29] Sprung S, Kirchner G, Rechenberg W. Reaction of poorly volatile
trace elements in cement clinker burning. Zem-Kalk-Gips
1984;37(10):513–8.
[30] Shirasaka T, Hanehara S, Uchikawa H. Influence of six minor and
trace elements in raw material on the composition and structure of
clinker. World Cem 1996;49(3):102–15.
[31] Kakali G, Parissakis G, Bouras D. A study on the burnability and the
phase formation of PC clinker containing Cu oxide. Cem Concr Res
1996;26(10):1473–8.
[32] Tsuboi T, Ito T, Hokinoue Y, Matsuzaki Y. The effect of MgO, SO3
and ZnO on the sintering of Portland cement clinker. Zem-Kalk-Gips
1972;25(9):426–31.
[33] Kakali G, Parissakis G. Investigation of the effect of Zn oxide on the
formation of Portland cement clinker. Cem Concr Res
1996;25(1):79–85.
[34] Guangllang X, Wenxi H, Zhongyuan L, Guangren Q. The effect of
ZnO on burning of Portland cement clinker in rapid heating-up
burning. In: Proc 9th int congress on the chemistry of cement, Delhi,
India, vol. 1; 1992. p. 372–8.
[35] Taylor HFW. Cement chemistry. 2nd ed. London: Thomas Telford
Publishing; 1997.
K.G. Kolovos / Cement & Concrete Composites 28 (2006) 133–143
[36] Arjunan P, Kumar A. Rapid techniques for determination of free
lime and free magnesia in cement clinker and portlandite hydrates.
Cem Concr Res 1994;24(2):343–52.
[37] Blaise R, Musikas N, Tiedrez H. Nouvelle méthode de détermination
cinétique de lÕ aptitude à la cuisson dÕ un cru de cimenterie. Rev Mater
Constr 1971;674–675:287–95.
[38] Odler I, Abdul-Maula S. Effect of mineralizers on the burning process
of Portland cement clinker. Part I: kinetics of the process. Zem-KalkGips 1980;33(6):132–6.
[39] Odler I, Abdul-Maula S. Effect of mineralizers on the burning process
of Portland cement clinker. Part II: mode of action of the mineralisers. Zem-Kalk-Gips 1980;33(6):278–82.
View publication stats
143
[40] Arligue G, Grandet J. Influence de la composition dÕ un ciment
Portland sur son hydratation en presence de zinc. Cem Concr Res
1990;20(4):517–24.
[41] Tashiro C, Takahashi H, Kanaya M, Hirakada I, Yoshiba R.
Hardening properties of cement mortar adding heavy metal compound and solubility of heavy metal from hardened mortar. Cem
Concr Res 1977;7(3):283–90.
[42] Uchikawa H, Hanehara S, Himi H. Behaviour of heavy metal
elements in the hardening of cement paste. In: Proc of the 10th int
congress on the chemistry of cement, Gothenburg, Sweden, vol. 2;
1997. p. 7.