Waste Management 24 (2004) 603–612

www.elsevier.com/locate/wasman

Leaching from solid waste incineration ashes used

in cement-treated base layers for pavements
Z. Cai a, Dirch H. Bager b, T.H. Christensen a,*

a
Environment & Resources DTU, Technical University of Denmark, Building 115, DK-2800 Lyngby, Denmark
b
Research and Development Centre, Aalborg Portland A/S, P.O. Box 165, DK-9100 Aalborg, Denmark
Accepted 26 January 2004

Abstract

Waste incineration bottom ash and treated flue gas cleaning products mixed with 2.5% of cement (50 kg/m3 ) were tested in the
laboratory in terms of compressive strength and tank leaching tests over a 64-day period. Although the material displayed lower
mechanical strength than a reference concrete, the strength still was sufficient for use as a base layer for roads. The metal content in
the incineration-residue-based specimens was up to 100 times higher than in the reference concrete, suggesting that the mixed waste
incineration residue should be used only for dedicated purposes. The leaching of Cl and Na was increased by a factor of 20–100 from
the incineration-residue-based specimens as compared to the reference, while the leaching of K, Ca and SO4 was increased by a
factor of 2–10. The leaching of heavy metals was also higher from the incineration-residue-based specimens than from the reference
with respect to Cu (50 times), Cd, Pb and Zn (5 times), but not with respect to Cr and Ni.
The leaching curves did only allow for a closer evaluation of the leaching process in a few cases. The physical retention of the
constituents seemed to be the same in the reference as in the incineration-residue-based specimens. Heavy metal leaching was limited
by enhanced chemical retention in the incineration-residue-specimens as compared to the reference. Since no quality criteria in terms
of leaching from a monolithic material are currently available, the leaching issue must be evaluated case by case.
Ó 2004 Elsevier Ltd. All rights reserved.

1. Introduction
Incineration is a commonly used solid waste treat-
ment in many countries, e.g., Denmark, France, Japan,
Sweden and Switzerland. Most of the waste mass leaves
the incinerator as flue gas through the stack, but typi-
cally 20% of the mass remains as a bottom ash (BA) and
2–5% as flue gas ashes (FGA) collected as fly ash or air-
pollution control residues. Although these residues only
represent a small volume compared to the volume of
waste incinerated, the management of residues will be-
come a main issue as criteria for use of the residues or
landfilling most likely will become more specific and also
more restrictive in the future.
The main environmental issue is the risk of ground-
water and surface water contamination due to leaching
*
Corresponding author. Tel.: +45-4525-1603; fax: +45-4593-2850.
E-mail address: thc@er.dtu.dk (T.H. Christensen).
of salts and heavy metals from the residues. The BA has
in many cases been used as unconsolidated base for
roads and pavements (Chandler et al., 1997), while the
FGA is considered hazardous waste that needs special
landfills. According to the European Landfill Directive
(EU, 1999) hazardous waste, in this case FGA, cannot
be landfilled in the future without pre-treatment.
In a recent paper (Cai et al., 2003), we showed that
pretreated FGA could substitute fly ash from coal-
burning power plants in cement-treated base layers for
roads and pavements without significantly compromis-
ing compressive strength or leaching of salts and heavy
metals as investigated in tank leaching experiments. For
each m3 of concrete, about 110–120 kg Ferrox-treated
FGA (substituting for coal-burning ash) was mixed with
50 kg of cement and about 2100 kg of an ordinary sand
and gravel mixture. In this paper we want to investigate
further the possibility of substituting the sand–gravel
mixture with aged BA from waste incineration. If pos-
sible this would lead to the creation of a single waste

0956-053X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.wasman.2004.01.010
604 Z. Cai et al. / Waste Management 24 (2004) 603–612

stream, combining the BA and pretreated FGA, to be
utilised as a secondary construction material in terms of
a cement-treated base (CTB) for roads and pavements.
The expected high concentrations of heavy metals in the
CTB would probably require that it be used only for
dedicated purposes. Highways would potentially be
suited, since they are publicly owned, regularly main-
tained, the consolidated road base has little contact with
water and they often have their own recycling schemes.
In this paper the specimens produced included a
reference concrete containing coal-burning fly ash and
sand–gravel mixture and four combinations of pre-
treated FGAs and BAs from waste incineration. The
specimens were in duplicate subject to compressive
strength testing and to the tank leaching test according
to Dutch standards.
2. Materials and methods
2.1. Materials used in concrete preparation
The CTB was in all cases prepared from a recipe in-
volving per m3 of produced concrete: 50 kg of cement
(CEM I 52.5 R, Aalborg Portland), 80–100 l of water,
2150 kg of fluvio-glacial gravel or about 1800 kg of aged
bottom ash from waste incineration (BA) and about
110 kg of either fly ash from coal-burning power plants
(CBA) or Ferrox-treated flue gas ash from waste incin-
eration (FGA). The actual recipe is shown in Table 1.
The workability (subjectively evaluated) of the fresh
mixes was identical for all five recipes.
The BAs were obtained at the incinerator plant
Vestforbrænding I/S in Copenhagen, Denmark. BA1
was from an incineration line equipped with a mass
burning grate system followed by a rotary kiln for post-
treatment of the bottom ash. BA2 was taken from a new
incineration line that has a grate system only. The BAs
were aged for 3 months in order to obtain carbonation
of the samples and afterwards sieved and subject to
magnetic separation of iron prior to use in the CTBs.
Use of non-carbonated BA in CTB would weaken the
strength of the CTB because of mass expansion by
carbon dioxide up-take.
The two FGAs used were a semi-dry air-pollution-
control residue from the incineration plant Ama-
gerforbrænding I/S in Copenhagen, Denmark (FGA1)
and a fly ash from the incinerator plant Vestforbrænding
I/S in Copenhagen, Denmark (FGA2). The FGAs were
treated by the Ferrox-process prior to use in the CTBs
(Lundtorp et al., 2002a). The ashes were washed in a
ferrous sulphate solution and afterwards aerated to
transform the precipitated ferrous hydroxide into fer-
rihydrite that has a high capacity for binding the heavy
metals. The solids and solute were finally separated in a
combined filtration and washing step. The process is
attractive because it uses low-cost materials (the ferrous
sulphate is a waste product itself) and little water (about
3 m3 per ton), the technology is simple, about 85–95% of
soluble chloride, sodium and potassium are removed
and the leaching of heavy metals is reduced by 1–3 or-
ders of magnitude by a principle that is well documented
in geochemistry (Lundtorp et al., 2002a,b). The leaching
behaviour of Ferrox-treated FGAs has been docu-
mented over a range of pH-values and over time by
various leaching tests (Lundtorp et al., 2002b).
2.2. Production of CTB specimens
The CTB material was mixed in a laboratory paddle
pan mixer and cast into 150-mm cubes. The cubes were
filled in two layers, each compacted with an electronic

Table 1
Recipe and physical characterization of the five CTB specimens
Reference, Experiment 1, Experiment 2, Experiment 3, Experiment 4,
CBA + gravel CBA + BA1 CBA + BA2 FGA1 + BA1 FGA2 + BA2
CBT recipe per m3
Cement (kg) 50 50 50 50 50
Ash (kg) CBA:110 CBA:110 CBA:110 FGA1:110 FGA2:110
Bottom ash (kg) 0 BA1:1800 BA2:1830 BA1:1750 BA2:1810
Gravel (kg) 2150 0 0 0 0
Water (l) 92 100 80 80 80
Specimens used in leaching test
Weight (g) 2527 1848 1983 2112 2030
Surface area (cm2 ) 628 590 622 632 637
Volume (cm3 ) 1130 903 951 1033 1013
Internal surface (m2 /g) 4.4 10.7 11.1 14.3 10.4
Particle density (kg/l) 2.64 2.72 2.60 2.70 2.68
Bulk density (kg/l) 2.24 2.05 2.09 2.04 2.00
CBA, coal burning ash: fly ash from coal burning power plants; BA1, bottom ash 1, grate + rotary kiln, Vestforbrænding I/S, Copenhagen; BA2,
bottom ash 2, grate, Vestforbrænding I/S, Copenhagen; FGA1, ferrox-treated semi-dry air-pollution-residue from Amagerforbrænding I/S, Co-
penhagen; FGA2, ferrox-treated fly ash from Vestforbrænding I/S, Copenhagen.
 Z. Cai et al. / Waste Management 24 (2004) 603–612
Fig. 1. Photograph of monolithic specimen (Experiment 4) after 64
days of tank leaching test.
Kango hammer according to the Swedish ‘‘Kango-
Cube’’ method (Andersson, 1987). The cubes were
sealed and cured at 20 °C.
Compressive strengths were measured after 7 days
according to a Danish concrete testing procedures (DS,
423.23). At the age of 14 days, two prismatic subsamples
with an approximate volume of 1 l (about 8 cm
8
cm
15 cm) were cut out of each cube for use in the
tank leaching test. Weight, surface area and volume are
shown in Table 1. Fig. 1 shows one of the specimens
after completion of the tank leaching test.
2.3. Tank leaching test
Ten tank leaching tests (duplicate test of the five types
of CTB) were performed according to the Dutch Tank
Leach Test (NEN, 7345).
The specimen was submerged at room temperature
into a closed polyethylene tank containing a leachant
(deionised water weakly acidified at pH 4) at an ap-
proximate volumetric liquid-to-solid ratio of 4.2 l/l. The
leachant was renewed at set time intervals (0.25, 1, 2.25,
4, 9, 16, 36 and 64 days). Immediately after sampling,
pH (always above 11 in the leachant) and specific
conductivity were measured, the sample was filtered
(0.45 lm) and acidified with HNO3 to pH <2 for
conservation.
605
Each sample was analysed for Cl, SO4 , Na, K, Ca,
Mg, Pb, Cd and Cr. The amount leached was calculated
from measured concentrations and volume of water in
the tank and expressed as accumulated mass per m2 of
specimen surface as a function of time since start of
leaching.
For the analysis of other contaminants one sample
for each tank leaching experiment was obtained by
pooling equal volumes from the individual samplings.
The result was expressed as accumulated mass leached
per m2 of specimen surface after 64 days of leaching.
2.4. Physical–chemical characterization
2.4.1. Internal surface area
On crushed (<2 mm) samples, the internal surface
was measured in duplicate by N2 sorption according to
the BET-method by means of Micromeritics apparatus
controlled by Gemini 2375 (V4.01) software.
2.4.2. Particle density
The particle density of the crushed samples (<2 mm)
was calculated from sample weight and volume. The
sample volume was determined by helium gas expansion
in an EPS Helium Gas Expansion Porosimeter.
2.4.3. Availability test
In order to determine the content of contaminants in
the concrete potentially available for leaching, tests were
performed according to the Dutch Availability Leach
Test (NEN, 7341).
After termination of the tank leaching test, the cylin-
ders were crushed in a compression testing machine and
the crushed materials were dried at 105 °C for 2 days. A
jaw crusher crushed the material to a maximum particle
size of 4 mm, and from the homogenised material a
subsample was milled in a ball mill for 5 h to a particle
size <250 lm. The fine material was mixed with deion-
ised water at a liquid-to-solid (L/S) ratio of 100 l/kg and
pH was kept for 3 h at pH 7 by addition of HNO3 . The
eluate was separated from the mixture by filtration
(<0.45 lm) and stored. The filter containing the material
was mixed with fresh deionised water at a L/S ratio of
100 l/kg, and the mixture was kept for 3 h at pH 4 by
addition of HNO3 . The eluate was filtered and the two
eluates were combined into one sample for analysis.
The measured availabilities are important data in
the mathematical interpretation of the tank leaching
test data, and thus it is advisable to measure the
availability on the actual specimen. This can be done
only after the tank leaching test, but this cause no
significant error since the 64-day long tank leaching
test removes less than 1% the material in the
specimen.
606 Z. Cai et al. / Waste Management 24 (2004) 603–612
2.5. Analytical procedures
The following analytical procedures were used:
pH was measured by a Radiometer standard pH-
meter (PHM210) with a Radiometer combination elec-
trode (pH 106007-3,5-NC).
Specific conductivity was measured by a Radiometer
Conductivity Meter CDM21 with electrode.
Cl and SO4 were determined by HPLC ion chroma-
tography.
Ca, Mg, K and Na were measured by flame atomi-
sation on a Perkin–Elmer 2280 atomic absorption
spectrophotometer.
Cd, Pb and Zn in the leachant samples were measured
by graphite furnace atomic absorption spectrophotom-
etry (Perkin–Elmer 5000, deuterium background cor-
rection, HGA 400 graphite furnace, AS-1 automatic
sample injection) after automatic solvent extraction by
NaDDC (sodium diethyldithiocarbamate trihydrate) in
MIBK (methyl-isobutyl-ketone). The extraction proce-
dure was chosen in order to avoid matrix effects due
to high concentrations of salts and to allow for
preconcentration.
Cr was measured by graphite furnace atomic ab-
sorption spectrophotometry (Perkin–Elmer 5000, deu-
terium background correction, HGA 400 graphite
furnace, AS-1 automatic sample injection). Mg(NO3 )2
was used as a matrix modifier.
As, Cr, Ni, Pb, Se and V were quantified by ICP-MS
(VG PlasmaTrace 2 High Resolution ICP-MS) and Al,
Ba, Ca, Cd, Cu, Fe, K, Mg, Mn, Na and Zn by ICP-
EAS (Varian Vista SIM ICP-AES, AX CCD).
3. Modelling of leaching from monolithic materials
It is common to interpret tank leaching results ac-
cording to a diffusion model because this allows pro-
jection of large data sets into a few key parameters
useful in generalization of the experimental results.
3.1. Theory
The cumulative release of a diffusing constituent x
from a solid matrix as the CTB specimens can be de-
scribed, according to Crank (1989), by Eq. (1) based on
the assumptions that the concentration of the diffusing
constituent remains dilute at the surface boundary and
that the concentration of the diffusing constituent at
some distance from the surface boundary within the
material does not change, i.e, depletion does not occur
(Kosson et al., 1996):

1=2
Dex  t
Mxt ¼ 2qC0x ; ð1Þ
p diffusion-controlled. If the measured and the calculated
where Mxt is the mass of the constituent per surface area
of the specimen (mg/m2 ) released during a time interval
t (seconds), q is the bulk density of the specimen (kg dry
matter/m3 ), C0x is the concentration of the constituent x
available for leaching according to the availability leach
test (mg/kg dry matter) and Dex is the effective diffusion
coefficient of the constituent in the monolithic material
(m2 /s). The equation shows that the accumulated release
increases with the square root of the time.
After logarithmic transformation of Eq. (1), the log-
arithm of cumulative release as a function of the loga-
rithm of time can be described by a line with a slope of
0.5 according to the following equation:
(
1=2 )
1 Dex
logðMxt Þ ¼ logðtÞ þ log 2qC0x ð2Þ
2 p
The effective diffusion coefficient Dex can be described by
Dex ¼ D0x =ðRx sÞ; ð3Þ
where D0x is the diffusion coefficient of constituent x in
water, Rx is the chemical retention factor of constituent x
in the specimen and s is the physical retention factor of
the specimen. s is specific to the actual specimen and is
also called tortuosity.
For practical purposes, the effective diffusion coeffi-
cient is often expressed in terms of its negative logarithm
ðp ¼  logÞ by
pDex ¼ pD0x  pRx  ps: ð4Þ
3.2. Assessing data consistency with the diffusion model
The amount of a constituent x released from the
monolithic specimen during the tank leaching experi-
ment is calculated by accumulating the releases
measured at each time interval, Mxi (mg/m2 ). Mxi is
calculated (Mxi ¼ Cxi Vi =A) from the concentration mea-
sured in the leachant sample, Cxi (mg/l), the volume of
leachant, Vi (l), and the external surface area of the
monolithic specimen, A (m2 ).
For the evaluation of the characteristic leaching
behaviour of the different constituents (diffusion, de-
pletion, surface rinsing, etc.), the measured accumu-
lative release cannot be used since any deviations in a
given period accumulate in the subsequent period.
Alternatively, an arithmetical cumulative release of

constituent x at the time ti ; Mxti corresponding to the
end of the ith leaching interval can be calculated only
from the release in the ith interval, Mxi , assuming
diffusion control in both the ith and the previous
intervals:
pffiffiffi
 ti
Mxti ¼ Mxi pffiffiffi pffiffiffiffiffiffiffi : ð5Þ
ti  ti1
These values can be used to check whether the release is
 Z. Cai et al. / Waste Management 24 (2004) 603–612 607

cumulative releases are equal, the release is controlled by
diffusion (Chandler et al., 1997).
After plotting the logarithm of the calculated cumu-

lative release, Mxti against the logarithm of the time, ti
for all the eight leaching intervals, the slope of the re-
lation over the entire test duration and three time in-
tervals representing different ranges of the leaching
process can be determined. If the slope is in the range
0.35–0.65, bracketing the theoretical slope of 0.5, it is
concluded that diffusion controls the release of the
constituent from the monolith (Chandler et al., 1997;
NEN, 7345).
If the leaching-controlling mechanism is diffusion, the
effective diffusion coefficient (m2 /s) for a constituent x
can be calculated for leaching step i by
2 This will typically require a mean strength of 7 MPa.
p  ðMxi Þ
Dex;i ¼ 2 pffiffiffi pffiffiffiffiffiffiffi 2 :
4  ðC0x  qÞ  ð ti  ti1 Þ
Then, the mean effective diffusion coefficient of the
constituent can be calculated as
Pn
pDex;i
pDex ¼ i¼1 : ð7Þ
n
After having determined the effective diffusion coeffi-
cient for each constituent, the physical (sÞ as well as the
chemical (Rx ) retention factor can be determined ac-
cording to Eqs. (3) and (4). D0x is available in standard
physico-chemical tables and Rx is equal to 1 for inert
constituents. Considering, for example, chloride and
sodium as approximately inert constituents, it is possible
to identify s. Since s is a specimen characteristic and
hence is identical for all constituents, Rx can subse-
quently be estimated for the chemically retarded con-
stituents. For comparison, Kosson et al. (1996) provides
an overview of reported physical and chemical retention
factors.
4. Results and discussion
The compression tests showed that the CTBs con-
taining only waste incineration residues and cement
(2.5%) had lower strengths than the reference CTB, but
still above the requirement of a characteristic compres-
sive strength of 5 MPa after 7 days as normally required
by the Danish road authorities for use in road bases.
ð6Þ Table 1 shows that the particle densities were iden-
tical for all five specimens (2.60–2.72 kg/l), but the
specimens containing incineration residues had lower
bulk densities (2.00–2.08 kg/l as apposed to 2.28 kg/l)
and higher internal surfaces (10.7–14.3 m2 /g as opposed
to 4.4 m2 /g), which implies a higher porosity and hence a
lower compressive strength.
The availability tests (Table 2) showed that the CTBs
containing incineration residues had availabilities of
most constituents that were 5–100 higher if compared to
the reference CTB consisting of sand/gravel and cement
only. Thus, the leaching potential is substantially higher
from the incineration-residue-based CTBs and must be
evaluated in more details.
The actual leaching measured over time in the tank
leaching tests is presented as cumulative plots in

Table 2
Availability test (mg/kg) of the CTB specimens
Availability test (mg/kg) Reference, Experiment 1, Experiment 2, Experiment 3, Experiment 4,
CBA + gravel CBA + BA1 CBA + BA2 FGA1 + BA1 FGA2 + BA2
Aluminum, Al 110 3600 3500 3700 3650
Arsenic, As <0.2 <1 <1 <1 <1
Barium, Ba 8 165 140 135 160
Calcium, Ca 8300 45,100 40,200 55,500 46,450
Cadmium, Cd 0.4 2.3 1.8 7.7 2.0
Chloride, Cl <90 810 800 500 600
Chromium, Cr 0.2 3.6 3.3 3.4 3.4
Copper, Cu 8 470 370 380 610
Iron, Fe 21 910 980 490 750
Potassium, K 80 590 755 710 640
Magnesium, Mg 125 1355 1225 1285 1350
Manganese, Mn 6 120 1225 160 155
Sodium, Na 40 1210 1570 1850 1255
Nickel, Ni 0.3 6.6 5.5 5.9 8.1
Lead, Pb 0.2 120 110 240 155
Selenium, Se <0.2 <1 <1 <1 <1
Sulphate, SO4 345 13175 11,000 26,000 11,600
Zinc, Zn 7 890 610 1150 915
Vanadium, V 0.5 0.6 0.7 0.3 0.4
608 Z. Cai et al. / Waste Management 24 (2004) 603–612

Figs. 2 and 3. Table 3 presents the analysis performed to
identify which constituents are controlled by diffusive
leaching. Table 4 presents the cumulative flux after 64
days for a range of constituents and offers effective dif-
fusion coefficients for those constituents positively
identified to be released from the material according to
diffusion-controlled leaching.
4.1. Leaching of salts (Cl, SO4 , Ca, Mg, K and Na)
The measured cumulative release over time for cations
(K, Na and Ca) and anions (Cl and SO4 ) in the tank
leaching tests are presented in Fig. 2. The duplicate tank
leaching test showed consistent results revealing that the
release of Cl and Na were, compared to the reference,

Fig. 2. Measured cumulative release of Na, K, Ca, Cl and SO2 2

4 (mg/m ) from the tank leaching tests on monoliths (open and closed symbols
represent duplicates).

Fig. 3. Measured cumulative release of Cr and Pb (mg/m2 ) from the tank leaching tests on monoliths (open and closed symbols represent duplicates).
 Z. Cai et al. / Waste Management 24 (2004) 603–612 609

Table 3

Slopes (rc) and standard deviation of the slope (SD) for the linear plot of logðMxti ¼ f ðlogðti ÞÞ for assessing leaching mechanism from CTB specimens
based on the whole leaching range (all eight leaching steps)
Reference Experiment 1 Experiment 2 Experiment 3 Experiment 4
CBA + gravel CBA + BA1 CBA + BA2 FGA1 + BA1 FGA2 + BA2
Slope, rc SD Slope, rc SD Slope, rc SD Slope, rc SD Slope, rc SD
Ca 0.67 0.065 0.43 0.076 0.43 0.070 0.42 0.159 0.44 0.086
0.59 0.054 0.61 0.056 0.52 0.115 0.58 0.129 0.36 0.069
a a
Cl 0.21 0.032 0.25 0.032 0.34 0.105 0.20 0.045
0.13b 0.108 0.23 0.041 0.38 0.064 0.40 0.074 0.20 0.024
Cr 0.49 0.097 0.30 0.032 0.26 0.050 0.61 0.107 0.38 0.067
0.54 0.050 0.17 0.072 0.41 0.146 0.49 0.040 0.38 0.060
K 0.58 0.077 0.25 0.022 0.25 0.044 0.38 0.085 0.30 0.020
0.49 0.017 0.23 0.026 0.37 0.059 0.40 0.063 0.24 0.046
Na 0.42 0.028 0.30 0.110 0.31 0.106 0.50 0.068 0.34 0.107
0.47 0.047 0.33 0.141 0.43 0.076 0.52 0.089 0.29 0.111
a a
Pb 0.51 0.181 0.71 0.674 0.45 0.133 0.68 0.086
a a
0.61 0.011 0.86 0.218 – – 0.95 0.143
SO4 0.45b 0.112 0.18 0.053 0.10 0.027 0.34 0.123 0.26 0.054
0.31b 0.075 )0.04 0.038 0.30 0.085 0.37 0.073 0.30 0.059
Initial leaching range (leaching steps 1–3).
The upper value is for Duplicate A and the lower value for Duplicate B.
a
Leachant concentration below detection limit.
b
Late leaching range (leaching steps 6–8).

systematically about 20–100 times higher for those CTBs
containing incineration residues. K, Ca and SO4 were
also leached in greater amounts from the CTBs con-
taining incineration residues only, typically by a factor of
2–10. This increased leaching from the incineration-res-
idue-based CTBs is also shown in Table 4 in terms of
cumulative release in mg/m2 after 64-day leaching.
Regression analysis of the arithmetical cumulative
releases, as described in Section 3, are presented in Table
3 in terms of regression slopes and their standard devi-
ations. According to the Dutch Tank Leaching Test
procedure (NEN, 7345), slopes within the range 0.35–
0.65 can be considered to represent diffusion-controlled
leaching. Only Ca showed slopes consistent within this
interval. For the other constituents, the slopes were of-
ten low, suggesting that the leaching flux decreased
faster over time than if controlled by diffusion only. The
leaching of Cl was so low for the reference CTB that a
slope could not be determined precisely. The CTB in
Experiment 3 was the only incineration-residue-based
CTB that showed for all salts slopes within the interval
characteristic for diffusion-controlled leaching. This
may suggest that not only the materials used (the bot-
tom ash making up about 85% of the specimens was the
same in Experiments 1 and 3) but also the casting of the
CTBs may influence the leaching characteristics.
Effective diffusion coefficients (pDex ) calculated for
those cases where leaching was shown to be diffusion-
controlled are presented in Table 4. Na and Cl are
assumed to be retained in the material by physical
mechanisms only. All the pDex -values determined for Na
and Cl are in the range 10.0–10.5 with an average of
10.3. Subtracting the diffusion coefficient in water for
Na (pD0 ¼ 8:9) or Cl (pD0 ¼ 8:7) yields a physical re-
tention factor s of 101:4–1:6 (or ps equal to between )1.4
and )1.6). This value applies to all the CTBs in spite of
the material composition and also fits values reported in
another experiment where only the fine material was a
waste product (Cai et al., 2003). This suggests that the
larger internal surface of the CTBs containing inciner-
ation residues does not affect leaching, either because
internal fractures in the specimens have significant tor-
tuosity or the surfaces only represent closed cavities not
in contact with the external surfaces. For Ca, which is
the only other salt ion that shows consistency with
diffusion-controlled leaching, the effective diffusion co-
efficient pDex is slightly higher for the incineration-resi-
due-based CTBs (pDex;Ca ¼ 14:5–15:1) than for the
reference CTB (pDex;Ca ¼ 13:9–14:2), suggesting that the
chemical retention of Ca is somewhat stronger than or
at least the same as in the reference. This again implies,
in general, that the increased leaching of the salts from
the incineration-residue-based CTBs is not due to a
more open structure of the material (affecting physical
retention) nor to less efficient chemical binding in the
material (affecting chemical retention), but only to the
fact that the availability of the constituents in the ma-
terial as such, as determined in the availability test on
finely crushed material at low pH values, is significantly
higher in the incineration residues than in the material
usually used in concrete.
4.2. Leaching of heavy metals
The measured cumulative releases over time of Cr
and Pb in the tank leaching tests are presented in Fig. 3.
610 Z. Cai et al. / Waste Management 24 (2004) 603–612

Table 4
Measured cumulative releases during 64-day leaching of CTBs (Mx; 64 days in mg/m2 ) and estimated effective diffusion coefficients (pDex ;  log (m2 /s))
of constituents (x)
Reference Experiment 1 Experiment 2 Experiment 3 Experiment 4
CBA + gravel CBA + BA1 CBA + BA2 FGA1 + BA1 FGA2 + BA2
x Mx pDex Mx pDex Mx pDex Mx pDex Mx pDex

Al 165 13.1 4200 

4930 
2830 
3500 
   
115 13.6 5500 4610 2940 3770
As <3 b
<3 
<3 
<3 
<3 
    
<3 <3 <3 <3 <3
Ba 6.0 13.7 18.7 
14.8 
12.1 
16.4 
   
4.0 14.2 16.0 13.0 9.2 17.3
Ca 5370 13.9 13,400 14.6 10,000 14.6 12,700 14.6 13,740 14.5
4240 14.4 8280 15.0 9520 14.8 9760 15.1 15,750 14.5
Cd <0.12 >15.0 0.57 
0.56 
0.36 
0.40 
   
<0.12 >15.0 0.46 0.42 0.36 0.50
 a a a a
Cl 420 29,600 37,700 63,500 32,150
 a a
1100 31,100 32,150 10.4 64,850 10.4 33,750
a a
Cr 6.70 10.4 3.29 3.48 2.16 13.8 2.74 13.7
a
6.95 11.3 4.11 2.97 13.4 2.42 13.8 2.94 13.5
Cu <0.6 >16.3 27.5 
28.5 
38.1 
32.3 
   
0.6 14.2 30.1 25.7 45.8 35.0
Fe 24 13.0 11.0 
10.2 
17.7 
5.2 
   
30 13.5 8.3 8.3 12.2 9.0
a a a
K 3220 10.3 20,750 23,150 26,750 10.2 20,450
a a
3320 10.6 21,250 19,600 10.7 27,400 11.4 20,700
Mg 50 
11.9 
11.0 
9.7 
8.1 
    
51 13.9 9.2 7.7 9.6
Mn <6.0 >13.5 <6.4 
<6.2 
<6.0 
<5.8 
   
<6.0 >13.6 <6.5 <6.0 <6.0 <6.2
a a a
Na 1440 10.3 52,700 63,550 87,650 10.0 55,600
a a
1430 10.5 58,550 55,600 10.4 90,600 10.2 54,150
Ni <0.6 >13.1 3.1 
<0.62 
0.60 
<0.58 
   
<0.6 >12.8 <0.6 <0.60 0.60 <0.62
a a
Pb <0.6 >11.7 2.25 16.9 1.12 3.19 16.8 2.42
a a a
0.75 11.7 0.85 17.4 1.39 1.50 2.41
Se <3 
<3 
<3 
<3 
<3 
    
<3 <3 <3 <3 <3
a a a a
SO4 2080 11.6 5880 9040 22,380 5490
a a a
2130 12.1 9390 7030 22,980 13.6 6470
Zn <6.1 >14.0 <6.4 
6.4 
30.5 
8.27 
   
<6.1 >13.0 6.8 5.8 11.0 6.47
V 6 11.2 0.7 
8.1 
3.5 
4.1 
   
7 11.5 8.2 6.7 4.3 4.5
The upper value is for Duplicate A and the lower value for Duplicate B. The pDex values are provided only for constituents subject to diffusion-
controlled leaching.
a
pDe cannot be estimated because the leaching is not controlled solely by diffusion.
b
The concentrations in the availability test samples were below the detection limit.
*
Only a single pooled sample was analysed and thus the leaching mechanism could be evaluated and pDe was not calculated.

The duplicate tank leaching test showed consistent re-
sults revealing that the release of Cr and Pb was of the
same order of magnitude for the reference CTB and the
CTBs based on incineration residues. The other metals
were only determined in single pooled samples and the
results presented in Table 4 in terms of cumulative re-
lease in mg/m2 after 64-day leaching. The general picture
is that leaching of Cr and Ni from the incineration-
residue-based CTBs is not higher than in the reference,
while leaching of Cd, Pb and Zn is increased by a factor
of 5 and Cu by a factor of 50.
Regression analysis of the arithmetical cumulative
releases, as described in Section 3, is presented in Table 3
in terms of regression slopes and their SDs for Cr and
Pb. The slopes could not be determined in all cases for
Pb because several samples displayed concentrations
below the detection limit (<5 lg/l). The slopes for Cr
were in most cases within the range of 0.35–0.65 sug-
gested by the Dutch Tank Leaching Test (NEN, 7345)
to represent diffusion-controlled leaching.
Effective diffusion coefficients (pDex , Table 4) were
calculated in a few cases for Cr and Pb for those
 Z. Cai et al. / Waste Management 24 (2004) 603–612
leaching curves that had slopes within the interval
characteristic for diffusion-controlled leaching (0.35–
0.65). Because of lack of consistency regarding diffusion
as the leaching-controlling process for heavy metals, we
did not estimate effective diffusion coefficients for those
metals only measured in a single pooled sample. The
estimated effective diffusion coefficients showed much
increased chemical retention of Cr and Pb in the CTBs
based on incineration residues as compared to the ref-
erence CTB: pDex -values for Cr were 13.4–13.7 com-
pared to 10.4–11.3 for the reference and pDex -values for
Pb were 16.8–17.4 compared to 11.7 for the reference.
This suggests that the relatively modest leaching of
heavy metals, here in terms of Cr, Pb, Cd, Ni and Zn, is
due to enhanced chemical retention in the CTB speci-
mens containing incineration residues, eventually caused
by the iron oxides supplied in the Ferrox-treatment of
the flue gas cleaning residues. Cu was the heavy metal
leached the most and a comparison between accumu-
lated leaching and availability of Cu in the specimens
suggests that the retention of Cu in the incineration-
residue-based CTBs is of the same magnitude as in the
reference CTB. Leaching and availability of Cu are both
increased by a factor of approximately 50 in the incin-
eration-based CTBs as compared to the reference CTB.
5. Conclusions
Laboratory experiments showed that waste incinera-
tion residues including bottom ash (1800 kg/m3 ) and
treated flue gas cleaning products (110 kg/m3 ) mixed
with cement (50 kg/m3 ) would provide a base layer
suitable for roads since the required compression
strength can be obtained. The metal content, as ex-
pressed in the availability test, was up to 100 times
higher in the incineration-residue-based specimens than
in the reference concrete, suggesting that the mixed
waste incineration residue should be used only for
dedicated purposes and not as a low-quality substitute
in concrete in general.
Leaching from the specimens was measured over 64
days in tank leaching test. The leaching of Cl and Na
and salts was increased by a factor of 2–100 from the
incineration-residue-based specimens as compared to
the reference, while the leaching of K, Ca and SO4 was
increased by a factor of 2–10. Since NaCl in many
temperate climates is used on roads as a de-icing agent,
the environmental aspects of leaching of salts suppos-
edly should be focussed on K and SO4 . The leaching of
heavy metals was also higher from the incineration-
residue-based specimens than from the reference with
respect to Cu (50 times), Cd, Pb and Zn (5 times), but
not with respect to Cr and Ni.
The leaching curves did allow for a closer evaluation
of the leaching process only in a few cases. However, the
611
data on Na and Cl revealed that the physical retention
of the constituents was the same in the reference as in
the incineration-residue-based specimens, suggesting
that the lower strength and larger internal surface of the
incineration-residue-based specimens did not affect the
leaching. The higher leaching of the salt ions thus was
due to a higher content and/or availability of the salts in
the incineration-residues. In contrast, the leaching of
heavy metals seemed to be limited by enhanced chemical
retention in the incineration-residue-based specimens as
compared to the reference. For Cr, the chemical reten-
tion factor (R) was three orders of magnitude and for Pb
five orders of magnitude higher in the incineration-res-
idue-based specimens as compared to the reference.
Since no quality criteria in terms of leaching from a
monolithic material is currently available, leaching data,
as provided in this paper, must be subject to case-specific
evaluation to judge the environmental soundness of
using this material, for example, as a cement-treated
road base.
Acknowledgements
Financial support for this research project has been
granted by EU-LIFE 99-programme (LIFE 99 ENV/
DK000615), Babcock & Wilcox Vølund Aps., I/S
Amagerforbrænding, I/S Vestforbrænding, Aalborg
Portland A/S and the Technical University of Denmark.
The technical assistance of Bent Skov, Heidi Dwinge,
Lena Troelsen, Lene K. Jensen and Singh Ngyuen is
gratefully acknowledged.
References
Andersson, R., 1987. Bel€aggninger av v€altbetong – Provningsmetoder
och kontrol. CBI Report Ra 2.87, Cement och Betonginstitutet,
Stockholm.
Cai, Z., Jensen, D.L., Christensen, T.H., Bager, D.H., 2003. Reuse of
stabilized flue gas ashes from solid waste incineration in cement-
treated base layers for pavements. Waste Management and
Research 21, 42–53.
Chandler, A.J., Eighmy, T.T., Hartl
en, J., Hjelmar, O., Kosson, D.S.,
Sawell, S.E., van der Sloot, H., Vehlow, J., 1997. Municipal solid
waste incinerator residues. In: The International Ash Working
Group. Studies in Environmental Science, vol. 67. Elsevier,
Amsterdam.
Crank, J., 1989. The Mathematics of Diffusion, second ed. Oxford
University Press, New York.
DS 423.23: Hærdnet beton. Trykstyrke. Danish Standards Associa-
tion, Copenhagen.
EU, 1999. EU Council Directive 1999/31/EC of 26 April 1999 on the
Landfill of Waste. Official Journal L182 16, 1–19.
Kosson, D.S., Sloot, H.A., van der, Eighmy, T.T., 1996. An approach
for estimation of contaminant release during utilisation and
disposal of municipal waste combustion residues. Journal of
Hazardous Materials 47, 43–75.
Lundtorp, K., Jensen, D.L., Christensen, T.H., 2002a. Stabilization of
APC residues from waste incineration with ferrous sulfate on a
612 Z. Cai et al. / Waste Management 24 (2004) 603–612

simi-industrial scale. Journal of Air and Waste Management
Association 52, 722–731.
Lundtorp, K., Jensen, D.L., Sørensen, M.A., Mogensen, E.P.B.,
Christensen, T.H., 2002b. Treatment of waste incinerator air-
pollution-control residues with FeSO4 : Concept and product
characterization. Waste Management & Research 20, 69–79.
NEN 7341, 1995. Leaching characteristics of solid (earthy and
stony) building and waste materials-leaching test. In:
Determination of the Availability of Inorganic Components
for Leaching, first ed. Netherlands Normalisation Institute,
Delft.
NEN 7345, 1995. Leaching characteristics of soil and stony building
and waste materials-leaching test. In: Determination of the
Leaching of Inorganic Components from Building and Monolithic
Waste Materials with the Diffusion Test. Netherlands Normalisa-
tion Institute, Delft.