Advances in Cement Research Advances in Cement Research, 2016, 28(7), 458–468

Volume 28 Issue 7 http://dx.doi.org/10.1680/jadcr.15.00149

Paper 1500149
From waste materials to products for use Received 30/12/2015; revised 21/03/2016; accepted 01/04/2016
in the cement industry Published online ahead of print 04/05/2016
Keywords: blended cements/recycled material/thermal methods/
Barthel, Rübner, Kühne, Rogge and Dehn analysis

ICE Publishing: All rights reserved

From waste materials

to products for use in
the cement industry
Maria Barthel Andreas Rogge
Graduate Scientist, Bundesanstalt für Materialforschung Head of Department, Bundesanstalt für Materialforschung
und -prüfung (BAM), Berlin, Germany und -prüfung (BAM), Berlin, Germany
(corresponding author: [email protected]) Frank Dehn
Katrin Rübner Professor and Head of Institute, Universität Leipzig, Institut für Mineralogie,
Senior Scientist, Bundesanstalt für Materialforschung und -prüfung (BAM), Kristallografie und Materialwissenschaft, Leipzig, Germany
Berlin, Germany
Hans-Carsten Kühne
Head of Division, Bundesanstalt für Materialforschung
und -prüfung (BAM), Berlin, Germany

Industrial wastes (bottom ash, crushed concrete fines, filter residue, paper ash and lignite fly ash) have potential for
use in building materials, for instance as raw materials for clinker production, as supplementary cementitious
materials (SCMs) or mineral additions in concrete. The properties of the products are dependent on the reactivity of
the waste materials used, which can be classified as inert, latent hydraulic or pozzolanic. In this study, waste materials
were first characterised. This was followed by theoretical considerations of the mentioned application options.
Experiments were limited to evaluation of potential as SCMs and, for this purpose, activity index measurements and
calorimetric and thermogravimetric analyses were performed. Finally, the synergetic effects of various waste
materials were considered. Paper ash (calcium oxide source) and filter residue (amorphous silicon dioxide source)
showed the best prospects for use as cementitious material components.

Introduction materials – bottom ash, crushed concrete fines, filter residue,

In industrial processes, many waste materials are landfilled paper ash and lignite fly ash. To classify the materials, they
even though they have the potential to be used in building were compared with frequently used by-products and cementi-
materials. The recycling of waste materials in the cement tious materials. Activity index measurements and calorimetric
industry is beneficial in terms of reducing costs and carbon and thermogravimetric analyses were then performed to evalu-
dioxide emissions, and has thus generated significant research ate the waste materials.
interest. In the cement industry, waste materials can be used
as raw materials for clinker burning, cement components or Waste materials with potential
supplementary cementitious materials (SCMs): the aim is to
Bottom ash from hard coal combustion (BA)
reduce the amount of ordinary Portland cement (OPC) used
In the hard coal combustion process in boilers heated by dry
through partial substitution with waste materials without a
firing at temperatures of 1100–1300°C, two fractions of ash
significant reduction in properties or even with improved
occur: 75% fly ash from the flue gas filter and 25% ash from
binder properties. However, despite the reuse potential of waste
the boiler bottom. In Germany, the amount of bottom ash
materials, their use can be challenging because of higher water
produced from hard coal combustion is around 410 000 t/year
demand or the presence of harmful components. Blends of
(VGB and Powertech, 2011). While fly ash is an established
waste materials could balance the drawbacks of each material
SCM, BA is mainly used in road construction or as lightweight
and, in addition, combined treatments of waste materials
aggregate. An overview of the application of coal BA in con-
could bring advantages due to synergy effects. The questions
crete was presented by Singh and Siddique (2013). Cheriaf
that need to be resolved are as follows.
et al. (1999) found that ground BA also exhibits pozzolanic
& Which materials could be combined? properties.
& Is it possible to create a mixture that is based only on
waste materials to develop a hydraulic binder? Crushed concrete fines (CF)
The demolition of concrete buildings and other constructions
The first step of this long-term study was to evaluate each results in large amounts of waste concrete. While waste con-
waste material separately. This paper considers five waste crete of large particle sizes can be reused as building materials,

458
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

the fine fraction (particle sizes smaller than 2 mm) is usually Gluth et al. (2014) studied the application of PA as a raw
landfilled due to its increased water absorption and negative material for geopolymer synthesis.
influence on fresh concrete properties (Splittgerber and Müller,
2002). Reformat et al. (2013) estimate that fines make up to Fly ash from lignite combustion (LFA)
20–40% of total concrete rubble; this corresponds to an annual Annually, Germany produces 8·47 million tons of fly ash from
amount of approximately 20 Mt in Germany (Basten, 2015). lignite coal combustion (VGB and Powertech, 2011). Fly ashes
Crushed concrete fines (CF) consist mainly of hydrated cement can be divided into calcium-rich (calcium oxide > 10%) and re-
and quartz sand, and can contain some limestone. This frac- duced calcium (calcium oxide < 10%) ashes. Their high sulfate
tion can be used as a raw material for clinker production content limits their use in cementitious systems. Mallmann
(Sanchez and Tudanca, 1998). In the burning process, the (2002) studied the influence of the great variation in com-
hydrated cement and the limestone decompose, and hydraulic position of lignite fly ashes (LFAs) from Germany. Due to the
clinker phases and free lime are formed. Furthermore, glassy state of the silica, silica-rich or alumosilica-rich fly ashes
untreated CF can be ground into small particle sizes (powder have the potential to be used as a pozzolanic SCM (Stark and
fraction < 0·15 mm) and then used as a filler (Angulo et al., Wicht, 2000). The free lime in calcium-rich fly ashes contrib-
2009). utes to its use in soil stabilisation or fertilisation. Mallmann
(2002) studied its use as a raw material for clinkerisation, while
Filter residue from chlorosilane production (FR) Pimraksa et al. (2009) reported that belite cements could be
In the production of pure silicon, organic silicone compounds produced by the hydrothermal treatment of the fly ash.
and highly dispersed silica, silanes and chlorosilanes are inter-
mediate products. From the thermal treatment of their distilla- Options of application
tion residues, the flue gas is washed with water. The acidic In the cement industry, hot and cold methods of use can be
wastewater is neutralised with lime milk. In this process, silica distinguished (Figure 1). In the hot method, the waste
residue precipitates. This is separated as filter sludge with a materials are calcined to produce clinker. In the cold method,
water content up to 85%. Annually, 5–10 kt of filter sludge per they can be used directly without thermal treatment but
producer are landfilled. Landfill costs for filter sludge (includ- possibly with an additional mechanical treatment as a SCM or
ing transport) are up to 25–40 E/t (Herbst and Rübner, 2012). as a mineral addition (MA) in concrete. The difference
For research purposes, dried filter sludge (filter residue (FR)) between SCMs and MAs is only the point of addition: SCMs
is used. The main component of this FR is amorphous silica, are added to the cement as a component of the binder,
which exhibits pozzolanic properties as well as a large specific whereas MAs are added directly to the concrete mixture.
surface area. Thus, FR is highly reactive but has a high water
demand. Due to its production process, the chloride content of The aim of this long-term study was to examine waste material
FR exceeds the limit valid in the DIBt approval principles for mixtures for the development of hydraulic binder systems. For
SCMs (DIBt, 2004). Many researchers have already examined this purpose, each waste material needed to be classified as a
the behaviour of silica residue in alternative binder systems (Fu first step; due to the variations in composition of most waste
et al., 2003; Gluth et al., 2013; Hartmann et al., 2014; Herbst materials, it was necessary to analyse every waste material.
et al., 2013; Park et al., 2007; Wang et al., 2001). Thereafter, the waste materials were evaluated in accordance
with different application possibilities. The potential use of
Waste paper sludge ash (PA) each material according to the three application options
To recycle paper, a de-inking process is necessary, and this (Figure 1) was considered theoretically by comparison of their
process produces waste paper sludge. Jung et al. (2011) esti- characteristic properties (chemical and mineralogical
mate that 25% of total paper recycling residues is de-inking
sludge, amounting to 1·15 Mt/year in Germany. Several
researchers have studied the activation of waste paper sludge at Cold method
Hot method
temperatures around 700°C. Waste paper sludge ashes contain (with or without
(thermal treatment)
kaolinite, which can be converted to metastable metakaolin mechanical treatment)
with pozzolanic properties. A comprehensive literature over-
view is presented by Frias et al. (2015). However, the larger
part of waste paper sludge is incinerated at temperatures above Use as
Use as Use as
reactive
800°C for energy recovery. The combustion residues (paper ash secondary raw material mineral
supplementary
for cement clinker addition (MA)
(PA)) have to be landfilled, at a cost of about 44 E/t (Jung production
cementitious
in concrete
material (SCM)
et al., 2011). Due to its mineralogical components (calcite,
belite and free lime), PA has the potential for application
in the cement industry. Some researchers have characterised Figure 1. Options of treatment and following application
PA and examined hydration behaviour after different con- possibilities of waste materials in cementitious systems
ditioning techniques (Bai et al., 2003; Segui et al., 2012) and

459
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

composition, particle size) with literature data. Additionally, Similarity to the waste materials described in the literature
their use as SCMs was experimentally tested. suggests the use of CF and LFA as secondary raw materials in
clinker production. In order to check the potential for appli-
cation of all the waste materials in this study (BA, CF, FR,
Use as secondary raw material for cement production PA, LFA), their chemical composition in comparison with
For cement clinker production, limestone, quartz sand and typical primary raw materials had to be determined first; the
clay serve as primary raw materials. The main chemical results are shown in Figure 2.
components are calcium oxide (CaO), silicon dioxide (SiO2),
aluminium oxide (Al2O3) and iron oxide (Fe2O3). Hydraulic From a chemical point of view, lime is almost 100% calcium
calcium silicates are formed during the burning process. oxide, quartz sand contains almost 100% silicon dioxide and
Aluminium oxide and iron oxide reduce the formation temp- clay is a mixture of silicon dioxide and aluminium oxide. As
erature of clinkers. In Germany, 5% of raw materials for shown in Figure 2, all the waste materials considered contain a
clinker production are waste materials (Basten, 2002). mixture of calcium oxide, silicon dioxide and aluminium
oxide + iron oxide. It thus appears that direct substitution with
For research purposes, Vogel et al. (2013) used up to 20% by one of the raw materials is not possible. However, an excess of
mass of waste materials (LFA, powdered aerated concrete and surplus chemical components could be compensated by a
so on) for cement clinker production. The requirements for reduction in other primary raw materials. PA is calcium-rich
adequate substitution were a minimal content of lime, a lime and has a considerable amount of aluminium oxide. LFA is
standard (KSt II) between 94·5 and 100 and a silica ratio (SR) also classified as calcium-rich because of its calcium oxide
between 2·7 and 3·2 (Table 4). The burning process was the content higher than 15%. However, the examined LFA con-
same as for normal clinker production, with temperatures up tained more silicon dioxide and aluminium oxide than the PA.
to 1480°C. The mixtures including limestone (79·5%), sand The BA, CF and FR consist mostly of silicon dioxide, and
(3·3%), LFA (12·1%) and powdered aerated concrete (5·0%) they could thus substitute a part of the quartz sand. In the
showed the same performance as ordinary clinker. Aouad burning process, hydraulically active phases such as calcium
et al. (2012) produced a clinker from 55% dredged sediments silicates develop. The rapid cooling process after burning also
(rich in silicon dioxide) in combination with calcium carbonate induces the formation of amorphous active components.
(CaCO3), iron oxide and aluminium oxide. After burning at However, some waste materials already possess reactive com-
1450°C, the same major crystalline phases as in OPC clinker ponents and it would be more efficient to use them without
developed. thermal treatment as SCMs.

Bottom ash (BA)

Crushed concrete fines (CF)
Filter residue (FR)
Paper ash (PA)
Lignite fly ash (LFA)
0 Clay (Taylor, 1997)
100 Quartz sand (Taylor, 1997)
Lime (Lothenbach et al., 2007)
OPC (Scholer et al., 2015)
25 Alumina cement (Taylor, 1997)
%
:M

75 Silica fume (Lothenbach et al., 2007)

Sili
de

Fly ash (Scholer et al., 2015)

co
oxi

Granulated blast-furnance slag (Scholer et al., 2015)

nd
ium

50
iox
ide
lc

50
Ca

:M
%

75
25

100
0
0 25 50 75 100
Aluminium oxide + iron oxide: M%

Figure 2. Calcium oxide/silicon dioxide/(aluminium oxide + iron

oxide) ternary phase diagram with compositions of different waste
materials in comparison with clinker raw materials, common
binder materials and SCMs

460
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

Use as supplementary cementitious material or smaller calcium oxide content than slag, both materials are
mineral addition located in the same region of the ternary phase diagram
Supplementary cementitious materials are added to burned (Figure 2) and it is assumed that their hydraulic properties are
clinker to produce composite cements. The clinker-to-cement similar. The C/S ratio of PA of 2·6 shows evidence of behav-
factor has been reduced from 0·86 to 0·75 since 2000 iour similar to that of cement (Figure 3). The hydraulic behav-
(Cembureau, 2012). SCMs can enhance the performance of iour depends strongly on whether the material is amorphous or
blended cement or concrete, for example by producing lower crystalline and x-ray diffraction (XRD) analysis was thus used
hydration heat, higher resistance against sulfate and chloride to determine the crystalline phases. The main components of
attack or lower permeability (Lothenbach et al., 2011). If a the considered waste materials are
proportion of OPC is substituted by waste materials without
strength loss, less clinker is needed and thus carbon dioxide & BA: quartz, anhydrite, mullite and amorphous components
emissions are reduced (Raab, 2010). Established SCMs are & CF: quartz
divided into inactive (fine limestone or quartz powder), pozzo- & FR: amorphous components
lanic (silica fume, natural pozzolans) or latent hydraulic & PA: calcite, quartz, lime, gehlenite, (belite)
(ground granulated blast-furnace slag, hard coal fly ash, & LFA: lime, anhydrite, quartz, gehlenite
natural pozzolans) materials. Their hydraulic potential
increases in this order. In general, the higher the calcium With the addition of SCMs, the quality and quantity of
oxide/silicon dioxide (C/S) ratio, the higher the hydraulic reaction products of cement change, and additional hydrate
potential. phases are formed after water addition. Cement hydration pro-
duces calcium silicate hydrates (C-S-H), portlandite, ettringite
Figure 3 shows the hydraulic behaviour of the examined waste and calcium aluminium carbonate hydrates (AFm phases:
materials. Materials rich in silicon dioxide (BA, CF, FR) have C3A.CaSO4.12H2O, C3A.CaCO3.11H2O). Based on thermo-
a C/S ratio close to zero. FR, CF and BA contain more than dynamic considerations, Lothenbach et al. (2011) calculated
80% silicon dioxide whereby the silicon dioxide can occur in the hydrate phases for cements blended with reactive waste
an amorphous state (FR), a crystalline state (CF) or a mixture materials. In blended cements, silica-rich materials with pozzo-
of both (BA). While amorphous silicon dioxide is highly lanic properties react with portlandite and with C-S-H phases.
pozzolanic (e.g. dispersed silica, silica fume), crystalline silicon Thus, the portlandite content and the C/S ratio of C-S-H
dioxide is almost inactive (quartz sand). Hence, it is assumed decrease. Calcium aluminium silicate hydrates (C-A-S-H) will
that FR is a highly reactive pozzolana, CF is almost inactive be formed by C-S-H and additional aluminium oxide.
and BA could have both properties. With a C/S ratio of 1·3, A higher content of aluminium components increases the for-
LFA can be classified as latent hydraulic, like the established mation of monocarboaluminate phases. Generally, the total
SCM ground blast-furnace slag. Although LFA has a slightly volume of hydrates increases.

Pozzolanic Latent hydraulic Cements

or materials
inactive
materials
Hydraulic activity

Hydration
Fast

Slow

0 0·5 1·0 1·5 2·0 2·5 3·0 3·5 4·0

Bottom ash (BA) Lignite fly ash Waste paper sludge ash
Concrete fines (CF) (LFA) (PA)
Filter residue (FR)
C/S ratio

Figure 3. Hydraulic activity of considered waste materials in

relation to calcium oxide/silicon dioxide ratio (after Herrmann et al.
(2015))

461
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

Based on a literature of the correlation between chemical com- Activity index

ponents and hydration products, the expected changes in phase Waste material mortars were mixed using 405 g cement (CEM I
composition by blending cement with the studied waste 42·5 R), 45 g waste material, 1350 g CEN standard sand and
materials are summarised in Table 1. While BA and CF 225 g water in accordance with EN 196-1 (DIN, 2005). The
consist mostly of quartz, they are nearly inert and do not influ- water/binder (w/b) ratio was 0·5. To ensure equal slump flows, a
ence the hydration products markedly. However, BA contains a superplasticiser based on polycarboxylate ether (PCE) was
small amount of amorphous components, which could contain added to the mixtures containing FR and PA. The mortar
amorphous silicon dioxide and aluminium oxide that form prisms were stored under water until testing. Their compressive
C-(A)-S-H as a hydration product of the blended cement. FR strengths were measured after 28 d and 91 d and the strength
has a high content of amorphous silicon dioxide (silica), which data were used to derive the activity index of the mortar samples.
increases the formation of C-S-H by decreasing the amount of Here, the activity index is the compressive strength of a mortar
portlandite. PA and LFA contain many components (free lime, in relation to that of plain mortar without waste material.
calcium carbonate, magnesium oxide (MgO)) that are assumed
to influence the hydration products. Free lime is highly reactive Calorimetry
and forms portlandite with water. Although calcium carbonate Pastes were made from 40 g cement, 10 g waste material
is non-hydraulic it reacts with cement to form additional and 25 g water. They were mixed for 2 min at 2000 rpm in a
mono- and hemicarbonates. Due to blending with PA or LFA, centrifugal mixer. In addition to the plain cement paste, pastes
magnesium oxide reacts with cement to form hydrotalcite-like with 10% by mass of quartz powder (QP) and 90% by mass of
phases. The amount of ettringite is increased by using LFA cement were mixed as comparative samples. Plastic ampoules
because of the calcium sulfate (CaSO4) content. PA could were filled with the pastes and these were then placed on heat
additionally increase the amount of C-S-H due to the small flux sensors in an isothermal calorimeter (TAM Air TA
content of hydraulic belite. Instruments). The released heat was measured in comparison
with an inert sand sample. Measurements were conducted at
Materials with mainly inert components (quartz, mullite or 20°C for 48 h.
gehlenite) do not influence the hydration products directly, but
they could have a filler effect and/or serve as additional nuclea- Thermogravimetric analysis
tion sites and so increase the strength of concrete (Lagerblad, Hardened samples used for the calorimetric measurements
2009; Lothenbach et al., 2011). were subjected to thermogravimetric analysis (Netzsch STA
449 C). In an alumina crucible, 20 mg of ground sample was
Experimental studies of the different waste heated to 1300°C in a nitrogen atmosphere. The heat rate was
materials 10 K/min. The change in sample mass was determined
Experiments for testing the SCM potential of waste materials automatically.
were carried out. The activity index was determined and
calorimetric and thermogravimetric analyses were used to
Results
assess mortar and cement paste samples containing the waste
materials. For all the tests, 10% by mass of cement was substi- Activity index of blended mortars
tuted with each waste material. The waste materials were used Compressive strength is a very common property used to assess
without thermal treatment (the cold method). building materials. The influence of waste materials on

Chemical component Waste material Expected phase changes

Silicon dioxide (amorphous) FR & Decrease of portlandite

& Formation of C-S-H with lower C/S ratio
CaO (free lime) PA, LFA & Increase of portlandite
CaCO3 PA, LFA & Formation of mono-/hemicarbonate
Aluminium oxide PA, BA & Formation of C-A-S-H from C-S-H with low C/S-ratio
& Decrease of ettringite, if low content of sulfur trioxide (SO3)
& Increase of AFm
Calcium sulfate LFA & Increase of ettringite
Magnesium oxide PA, LFA & Formation of hydrotalcite-like phase
Belite (2CaO.SiO2) PA & Increase of C-S-H

Table 1. Expected phase changes depending on the additional

chemical component of waste materials

462
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

compressive strength can be assessed through determination of minimum, with the main reaction starting after this point. In
activity index. An activity index of 90% indicates that the sub- this acceleration period, the heat flow increased to a maximum.
stitute material (10% by mass of cement in this study) does not The heat flow curves for the different pastes overlap, making
contribute to the compressive strength of the mortar. However, analysis difficult. For a better comparison of the different
achievement of maximum compressive strength is not essential setting behaviours of the pastes, the times taken to reach the
at all times, but depends on the building element. An activity minimum of the dormant phase and the maximum of the
index higher than 90% might indicate that the waste material acceleration phase are listed in Table 2. A shorter time to
contains reactive components and/or its particle fineness is achieve the maximum of the acceleration phase indicates an
higher than that of the cement. acceleration effect from the SCM, while a longer time indicates
a retarding effect on the setting behaviour.
As shown in Figure 4, the PA mortar achieved an activity
index of almost 90%, meaning that the amount of reactive The main reaction of cement with the highest heat flow
phases seems low. The activity index of the CF mortar was occurred between 2 and 30 h. Over this time period, the paste
lower than 90%, which is not surprising due to the inert main with QP showed a lower broad heat flow peak. This behaviour
component of CF (quartz). The activity indexes of BA and is typical for an inert SCM since only heat from cement
LFA both exceeded 90% after 91 d, indicating that BA and hydration is released. Due to the lower amount of cement in
LFA seem to be reactive. The mortar with FR showed the the blended pastes, the heat flow was lower than that of the
highest activity index after 28 d due to the high reactivity of plain paste. Inert materials do not react with water; the conse-
amorphous silica particles, but the activity index decreased at quence is a higher w/b ratio, which leads to delayed heat
91 d. A change in the crystalline structure hydrated phases release. The paste with QP showed the lowest cumulative heat
could provide an explanation for this. of all tested samples (Table 2). The behaviour of pastes con-
taining CF and BA was between the cement and the QP paste.
Calorimetry of blended pastes The BA paste showed a slightly higher heat flow than the CF
During the exothermic hydration reaction of cement, heat is paste and the maximum of the acceleration period was thus
released from the paste mixture. For reactive SCMs, the heat achieved earlier (Table 2). It seems that BA contains a few
release is higher than for a mixture with an inert SCM. reactive components; these are suggested to be amorphous
The results of the calorimetric measurements are shown in components, after XRD analysis. The cumulative heat of the
Figure 5. BA paste was significantly higher than that of the CF paste
after 168 h: it is assumed that the amorphous component of
During the initial period of cement hydration, heat flow the BA reacted as a pozzolan with the calcium hydroxide
increased quickly at first. This period was followed by a (Ca(OH)2) generated by cement hydration.
dormant period during which heat flow reduced to a
The LFA paste showed similar behaviour but achievement of
the maximum of the main reaction was delayed. The cumulat-
120 ive heat release until 48 h was lower in comparison with the
After 28 d
heat release of the pure cement paste, but was almost the same
After 91 d
100 for both pastes after 168 h. The reaction of the paste with
LFA was thus only delayed, although lime from the LFA
Activity index: %

80
reacted immediately with water. The additionally formed por-
tlandite might have possibly reacted with the cement com-
ponents, thus extending the main heat flow peak.
60

Free lime reacts in the PA paste, indicated by an additional

40
heat flow peak that occurred earlier than the main broad peak
of the plain cement paste. Furthermore, the PA paste may
20
show this behaviour due to the high alumina content because
calcium aluminate phases are known for their fast reaction.
0
The paste with PA and the paste with FR both achieved the
r

r
rta

rta

rta

rta

rta

maximum of the acceleration period much earlier in compari-

mo

mo

mo

mo

mo

son with the plain cement paste (Table 2). These pastes set
BA

CF

FR

PA

LFA

earlier. Nevertheless, after nearly 9 h, the FR paste had

Figure 4. Activity index of mortars with 10% by mass released more heat than the PA paste. Due to the hydration
substitution of cement with waste materials after 28 d and 91 d; reaction of cement, calcium hydroxide is produced, which
w/b = 0·5 immediately reacts with the amorphous silica of FR in a poz-
zolanic reaction. Consideration of the cumulative heat shows

463
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

CEM I 250

Cumulative heat of total sample mass: J/g

Heat flow of total sample mass: mW/g

CEM I–QP
CEM I–BA
2 200
CEM I–CF
CEM I–FR
CEM I–PA 150
CEM I–LFA CEM I
CEM I–QP
1 100 CEM I–BA
CEM I–CF
CEM I–FR
50
CEM I–PA
CEM I–LFA
0 0
6 12 18 24 30 36 42 48 20 40 60 80 100 120 140 160
Time after water addition: h Time after water addition: h
(a) (b)

Figure 5. Heat flow (a) and cumulative heat release

(b) of blended pastes with 10% by mass cement substitution with
waste materials and comparative pastes (100% cement and
10% cement substituted with quartz (QP)); w/b = 0·5

Mixture notation Time to minimum of Time to maximum of Cumulative heat of total sample mass: J/g
dormant phase: h acceleration phase: h
After 24 h After 48 h After 168 h

CEM I 1·9 10·8 120 163 225

CEM I–QP 2·4 12·4 94 133 190
CEM I–BA 2·3 11·4 109 145 202
CEM I–CF 2·1 12·2 98 137 194
CEM I–FR 1·7 7·7 119 156 214
CEM I–PA 3·7 6·9 100 140 211
CEM I–LFA 2·8 14·7 98 149 211

Table 2. Calorimetric characteristic data

that the heat release of the LFA paste increased slowly until it compounds (Scholer et al., 2015). It seems that aluminium
reached the values of the PA and FR pastes. These results compounds shift the degradation to higher temperatures
show that the pastes with PA, FR and LFA are all reactive, but in comparison with pure C-S-H. In contrast, the decompo-
they react at different hydration times. sition peak of calcium hydroxide is well known, which is
formed during the cement reaction. If the paste contains
Thermogravimetric analysis of hardened blended pastes pozzolanic components, which can react with calcium hydrox-
To obtain more information about the hydration products, ide, the DTG peak should decrease. This phenomenon is
hardened cement pastes with and without waste materials were shown in the FR paste, which contains mostly amorphous
examined by thermogravimetric analyses. Figure 6 shows the silica. Materials containing free lime (PA, LFA) should show
mass loss and the rate of mass loss (DTG) of the hardened higher amounts of calcium hydroxide due to the reaction of
paste samples. Significant mass losses were found in four lime with water. But the amount of the additionally formed
different temperature ranges. Every peak can be related to the calcium hydroxide seems to be low in comparison with that
decomposition of a particular paste component: Table 3 gives from cement hydration. All the other materials (QP, BA, CF)
an overview. appear not to show any calcium hydroxide consumption. Their
lower calcium hydroxide peak in comparison with the plain
Mass loss in the temperature range up to 200°C is linked to cement paste can be related to the 10% substitution of cement
the dehydration of C-S-H and partly includes aluminium with each waste material.

464
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

CEM I CEM I–QP CEM I–BA CEM I–CF

CEM I–FR CEM I–PA CEM I–LFA

100

95
Mass loss: %

90

85

80
0·2
0
–0·2
DTG: %/min

–0·4
–0·6 Calcium carbonate
–0·8
–1·0
C2(A)SH8
–1·2
C-S-H Calcium hydroxide
–1·4
100 200 300 400 500 600 700 800 900 1000 1100 1200
Temperature: °C

Figure 6. Results of thermogravimetric analyses (mass loss and

rate of mass loss (DTG)) of the sample mass of hardened pastes
with 10% by mass of cement substitution with waste materials
after 56 d of hydration; w/b = 0·5

Discussion and conclusion

Temperature zone: °C Reaction
Considerations about synergetic mixtures of waste materials
are based on a systematic evaluation of each material. Besides
20–130 Dehydration of CSH
the properties of the waste materials, their production quan-
130–180 Dehydration of C2ASH8
tities and disposal costs should also be considered for evalu-
400–490 Dehydration of calcium hydroxide
ation. While crushed concrete fines (CF) and lignite fly ash
650–725 Decarbonation of calcium carbonate
(LFA) are in the focus of utilisation due to the large quantities
Table 3. Temperature zones of DTG peaks and classification (in produced, the literature review shows that the chemical–
accordance with Bai et al. (2003) and Frias and Cabrera (2001)) mineralogical compositions of these materials fluctuate over a
wide range. As a basis for waste material mixtures, a better
choice would be paper ash (PA): the quantities of PA produced
The last peak, between 650°C and 750°C, is related to the dec- are not only larger than those of filter residue (FR) and
arbonation of calcium carbonate. This peak is slightly higher bottom ash (BA) but it also produces a constant annual
for the hardened FR paste. It is assumed that silicon dioxide is supply.
fully reacted and the excess remaining calcium hydroxide
reacts with carbon dioxide from the atmosphere to form The literature review and comparisons with other well-known
calcium carbonate. Additionally, the calcium carbonate peak is materials show that all the waste materials examined in this
decreased in the PA paste, which could be related to the high study (BA, CF, FR, PA and LFA) could be used as substitute
calcium carbonate content of the initial PA. Some calcium raw materials for clinker production. The oxide compositions
carbonate reacts with aluminium oxide to form monocarboalu- (calcium oxide, silicon dioxide, aluminium oxide and iron
minate (see Table 1), which is confirmed by the increased peak oxide) mainly indicate use of the waste materials via hot
between 130°C and 180°C for the PA paste. methods (i.e. as secondary raw materials). To balance the

465
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

deviations in composition from that of primary raw materials, they could be used as pozzolanic SCMs. On the other hand,
using mixtures of different waste materials seems to be a materials with free lime and/or calcium silicates (PA and LFA)
reasonable approach. Clinker composition is expressed in are also reactive but they showed hydraulic behaviour. The
terms of the lime saturation factor, silica ratio and alumina results of activity index and calorimetric and thermogravi-
ratio (Table 4). Among the waste materials studied, PA could metric measurements confirm almost all the theoretical con-
be used as the main component, due to it having the highest siderations. FR is a fine amorphous powder that is highly
calcium oxide content and appropriate silicon dioxide content. reactive. The activity index of FR was higher than 100% and
An addition of 14–16% of silica-rich waste material (BA, CF, the mass loss results show a high consumption of calcium
FR) is capable of achieving the desired silica ratio between 1·8 hydroxide, which gives evidence of a pozzolanic reaction. The
and 3·0. However, in comparison with OPC clinker, the results of activity index measurement and calorimetry show
alumina ratio is higher and the lime saturation factor is lower. that BA reacts with a delay. The amorphous component is
More calcium oxide is needed for a complete reaction with assumed to be aluminium oxide and/or silicon dioxide, which
silicon dioxide, aluminium oxide and iron oxide to form a could react as a pozzolan. The reaction seems to be insignifi-
technologically reasonable quantity of the four main clinker cant, which explains the missing calcium hydroxide consump-
phases (C3S, C2S, C3A and C2(A,F)). On the other hand, tion in the mass loss results and the relative cumulative heat of
clinker containing a larger amount of belite (C2S) or calcium only 90% (Figure 7). One reason for this reduced reaction
aluminate (e.g. mayenite) with a lower calcium/aluminium could be the coarse particle size. CF is finer but contains
ratio coupled with hydraulic properties could modify the raw quartz. Its inert character was confirmed by an activity index
mixture containing PA and silica-rich material. Alumina-rich less than 90% and hydration behaviour similar to the paste
cements are niche products used for special applications in the with inert quartz powder. The activity index of LFA was found
refractory industry. However, a low calcium oxide content to be higher than 90% after 28 d of hydration. Free lime reacts
could be corrected by using further waste materials based on immediately with water, but a higher calcium hydroxide
calcium oxide, for instance sludge from limestone washing content than the other pastes was not detected by mass loss
processes. analysis. It appears that some of the calcium hydroxide reacted
with silicon dioxide, consequently resulting in strength devel-
The burning process has the benefit that waste materials with opment. PA also contains free lime, which reacts very quickly.
high amounts of non-hydraulic mineral phases can be acti- In addition, a second peak in the heat flow curve appeared
vated. One example is PA, which consists mainly of calcite during the calorimetric measurements. This behaviour could
(calcium carbonate). At a temperature of 750°C, calcite be explained by hydration of reactive calcium aluminate due to
decomposes to carbon dioxide and free lime (calcium oxide) the high aluminium oxide content of PA. The calcium alumi-
while the latter reacts with silica to form C-S-H. nate reaction releases heat but does not contribute significantly
to the strength development of the mortars (activity index
Waste materials with amorphous components (FR and BA) 90%), as shown in Figure 7. The assumption of hydraulic be-
are presumed to be reactive without thermal treatment and haviour of PA as cement, which is based on the C/S ratio, was

Raw mixture Ideal value for

OPC production
PA–BA PA–CF PA–FR

BA content: % 14% — —
CF content: % — 16% —
FR content: % — — 14%
PA content: % 86% 84% 86%
Lime saturation factor
KSt = 100  CaO/(2·8  SiO2 + 1·2  Al2O3 + 0·65  Fe2O3) 50·8 49·8 49·8 90–102
Silica ratio
SM = SiO2/(Al2O3 + Fe2O3) 2·3 2·7 2·6 1·8–3·0
Alumina ratio
TM = Al2O3/Fe2O3 7·1 10·4 12·4 1·3–4·0

Table 4. Lime saturation factor, silica ratio and alumina ratio of

raw mixtures for clinker burning process based on waste materials
in comparison with the values of OPC clinker

466
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

CEM I–BA CEM I–CF CEM I–FR material blends in order to learn more about the influence of
CEM I–PA CEM I–LFA treatment (hydrothermal/thermal/mechanical) and synergistic
effects.
110
Activity index of mortar: %

REFERENCES
105
Angulo SC, Ulsen C, John VM, Kahn H and Cincotto MA (2009)
100 Chemical–mineralogical characterization of C&D waste
recycled aggregates from São Paulo, Brazil. Waste
95 Management 29(2): 10.
Aouad G, Laboudigue A, Gineys N and Abriak NE (2012)
90 Dredged sediments used as novel supply of raw material to
produce Portland cement clinker. Cement & Concrete
85 Composites 34(6): 788–793.
Bai J, Chaipanich A, Kinuthia JM et al. (2003) Compressive
80 strength and hydration of wastepaper sludge ash–ground
85 90 95 100 granulated blastfurnace slag blended pastes. Cement and
Relative cumulative heat of paste: % Concrete Research 33(8): 1189–1202.
Basten M (2002) Zementrohstoffe in Deutschland: Geologie,
Figure 7. Correlation between activity index of mortars and
Massenbilanz, Fallbeispiele. Bau+Technik, Düsseldorf,
cumulative heat of pastes relative to plain paste (CEM I)
Germany (in German).
Basten M (2015) Mineralische Bauabfälle Monitoring 2012.
In Bericht zum Aufkommen und zum Verbleib mineralischer
not confirmed. The main calcium oxide content is bound as Bauabfälle im Jahr 2012 (Pahl G and Schäfer B (eds)).
calcium carbonate, which is non-hydraulic but forms mono- Kreislaufwirtschaft Bau- Bundesverband Baustoffe- Steine
carboaluminate with aluminium oxide. In summary, highly und Erden e.V., Berlin, Germany, pp. 1–14 (in German).
reactive pozzolanic waste materials (FR) could be combined Cembureau (2012) Cements for a Low-Carbon Europe.
with waste materials that contain free lime. Cembureau, Brussels, Belgium.
Cheriaf M, Rocha JC and Pera J (1999) Pozzolanic properties of
In addition to the effect of waste materials on strength devel- pulverized coal combustion bottom ash. Cement and
opment, changes in setting behaviour should also be con- Concrete Research 29(9): 1387–1391.
sidered. While LFA has a retarding effect, PA produces an DIN (2005) EN 196-1: Prüfverfahren für Zement – Teil 1:
accelerating behaviour, similar to FR. Therefore, for adequate Bestimmung der Festigkeit.
setting behaviour, LFA, as well as CF and BA, require an DIBt (Deutsches Institut für Bautechnik) (2004) Grundsätze für
accelerating admixture. The interaction between SCMs and die Erteilung von Zulassungen für anorganische
admixtures should be examined in further studies. Betonzusatzstoffe (Zulassungsgrundsätze). Deutsches
Institut für Bautechnik, Berlin, Germany (in German).
Small particle sizes could influence hydration behaviour posi- Frias M and Cabrera J (2001) Influence of MK on the reaction
tively because of the nucleation function of fine particles. This kinetics in MK/lime and MK-blended cement systems at
fineness can also enhance performance due to a filler effect. 20°C. Cement and Concrete Research 31(4): 519–527.
Although inert materials with a small particle size contribute Frias M, Rodriguez O and De Rojas MIS (2015) Paper sludge,
physically to strength development, materials with a larger par- an environmentally sound alternative source of MK-based
ticle size and higher amounts of amorphous components could cementitious materials. A review. Construction and Building
provide a chemical contribution to the system. A mixture Materials 74: 37–48.
of waste materials could thus combine reactive components Fu X, Wang Y, Huang S, Hou X and Hou W (2003)
(e.g. FR) with other fine materials acting as nucleation sites or The influences of siliceous waste on blended cement
fillers (e.g. CF). In this case, an additional mechanical treat- properties. Cement and Concrete Research 33(6): 851–856.
ment could be beneficial for homogenisation of the waste Gluth GJG, Lehmann C, Rübner K and Kühne HC (2013)
material mixture. Geopolymerization of a silica residue from waste treatment
of chlorosilane production. Materials and Structures
This systematic examination of waste materials and theoretical 46(8): 1291–1298.
considerations concerning waste material mixtures indicate Gluth GJG, Lehmann C, Rübner K and Kühne HC (2014)
that PA is the most promising material for the production of a Reaction products and strength development of wastepaper
new reactive binder based on waste material mixtures. The sludge ash and the influence of alkalis. Cement & Concrete
next step of this long-term study is the examination of waste Composites 45: 82–88.

467
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.
 Advances in Cement Research From waste materials to products for use
Volume 28 Issue 7 in the cement industry
Barthel, Rübner, Kühne, Rogge and Dehn

Hartmann A, Petrov V, Buhl J et al. (2014) Zeolite synthesis Altbetonbrechsand als alternatives Kompositmaterial
under insertion of silica rich filtration residues from im Zement. Fachtagung Recycling R’13, Bauhaus-
industrial wastewater reconditioning. Advances in Chemical Universität Weimar, Weimar, Germany, pp. 1–10
Engineering and Science 4(2): 14. (in German).
Herbst T and Rübner K (2012) Nanoskaliger Sanchez JA and Tudanca S (1998) Reuse of building rubble in
Kieselsäureschlamm – Technologieentwicklungen zur cement manufacturing. In Sustainable Concrete: Reuse of
Verwertung in mineralischen Baustoffen mit dem Ziel der Recycled Concrete Aggregates (Dhir RK, Henderson NA
Verbesserung der Werkstoffeigenschaften, Teilprojekt: and Limbachiya MC (eds)). Thomas Telford, London,
Stoffcharakterisierung, hydraulische Bindemittel und UK, pp. 481–491.
Nachhaltigkeits- und Wirtschaftlichkeitsbetrachtungen: Scholer A, Lothenbach B, Winnefeld F and Zajac M (2015)
Schlussbericht zum Verbundvorhaben ‘NanoKiesel’. Hydration of quaternary Portland cement blends
Technische Informationsbibliothek u. containing blast-furnace slag, siliceous fly ash and
Universitätsbibliothek, Bundesanstalt für limestone powder. Cement & Concrete Composites 55:
Materialforschung und -prüfung, Berlin, Germany 374–382.
(in German). Segui P, Aubert JE, Husson B and Measson M (2012)
Herbst T, Rübner K and Seidemann M (2013) Utilization of Characterization of wastepaper sludge ash for its
siliceous filter residues in mineral building materials. valorization as a component of hydraulic binders. Applied
ZKG International 66(7–8): 58–65. Clay Science 57: 79–85.
Herrmann A, König A and Dehn F (2015) Vorschlag zur Singh M and Siddique R (2013) Effect of coal bottom ash as
Klassifizierung von alkalisch-aktivierten Bindemitteln partial replacement of sand on properties of concrete.
und Geopolymer-Bindemitteln. Beton (7–8): 390–395 Resources Conservation and Recycling 72: 20–32.
(in German). Splittgerber F and Müller A (2002) Identifying the type of
Jung H, Kappen J, Hesse A and Götz B (2011) Aufkommen und cement in hydrated concretes and mortars. ZKG
Verbleib der Rückstände aus der deutschen Papier- und International 7(7): 10.
Zellstoffindustrie- Ergebnisse der Rückstandsumfrage Stark J and Wicht B (2000) Zement und Kalk: Baustoff als
2010. Wochenblatt für Papierfabrikation 11: 3 (in German). Werkstoff. Birkhäuser, Basel, Switzerland (in German).
Lagerblad B (2009) Fillers and ultrafine fillers for low cement Taylor HFW (1997) Cement Chemistry. Thomas Telford,
concrete to reduce the carbon footprint. GDCh- Tagung London, UK.
Bauchemie. GDCh, Freiberg, Germany, pp. 1–9 VGB and Powertech (2011) Produktion und Verwendung von
(in German). Kraftwerksnebenprodukten aus Kohlekraftwerken in
Lothenbach B, Scrivener K and Hooton RD (2011) Deutschland im Jahr 2011. VGB and Powertech, Essen,
Supplementary cementitious materials. Cement and Germany (in German).
Concrete Research 41(12): 1244–1256. Vogel S, Kolditz K, Bellmann F, Ott D and Kralisch D (2013)
Lothenbach B, Winnefeld F, Alder C, Wieland E and Lunk P Ein Beitrag zur Reduzierung der CO2-Emissionen
(2007) Effect of temperature on the pore solution, beim Zementklinkerbrand- Teil1+2. Cement International
microstructure and hydration products of Portland cement 10(1): 10 (in German).
pastes. Cement and Concrete Research 37(4): 483–491. Wang L, Seals RK and Roy A (2001) Investigation of utilization
Mallmann R (2002) Entwicklung hydraulischer Bindemittel mit of amorphous silica residues as supplementary cementing
rheinischen Braunkohlenfilteraschen. Doctor rerum materials. Advances in Cement Research 13(2): 85–89,
naturalium, Universität Siegen, Siegen, Germany http://dx.doi.org/10.1680/adcr.2001.13.2.85.
(in German).
Park JS, Park YJ and Heo J (2007) Solidification and recycling
of incinerator bottom ash through the addition of
colloidal silica (SiO2) solution. Waste Management 27(9):
1207–1212.
Pimraksa K, Hanjitsuwan S and Chindaprasirt P (2009) Synthesis
of belite cement from lignite fly ash. Ceramics
International 35(6): 2415–2425.
Raab B (2010) Synthese und Charakterisierung nanoskaliger
hydraulisch hochreaktiver Phasen des Portland- und WHAT DO YOU THINK?
Tonerdezements. Doctor rerum naturalium, Martin- To discuss this paper, please submit up to 500 words to
Luther-Universität Halle-Wittenberg, Halle, Germany the editor at [email protected]. Your contribution will
(in German). be forwarded to the author(s) for a reply and, if
Reformat M, Linß E and Ludwig HM (2013) Potentiale considered appropriate by the editorial panel, will be
der Verwertung von mechanisch aktiviertem published as a discussion in a future issue of the journal.

468
Downloaded by [ Purdue Univ Lib TSS] on [20/09/16]. Copyright © ICE Publishing, all rights reserved.