Dioxin production during the thermal treatment of meat and bone meal residues

Article abstract of DOI:10.1016/j.chemosphere.2004.09.089

Safe animal by-product disposal is a priority target as a result of the Bovine Spongiform Encephalopathy crisis in the European beef industry. One such disposal option is the incineration of by-product material such as meat and bone meal residues (MBM) for the purpose of energy recovery. Although currently applied, the thermal decomposition of MBM wastes has not been scientifically studied until now. A series of experiments has been performed to study the thermal behavior of MBM both in inert (N₂) and reactive atmosphere (air), both by thermogravimetry and in a horizontal laboratory furnace. As a general trend, MBM gives low PCDD/F values, compared with incineration of other wastes. Maximum yield of pollutants is observed at a nominal temperature between 700 and 800°C.

Full text of DOI:10.1016/j.chemosphere.2004.09.089

Chemosphere 59 (2005) 85–90
www.elsevier.com/locate/chemosphere
Dioxin production during the thermal treatment of meat
and bone meal residues
*
Juan A. Conesa , Andres Fullana, Rafael Font
Department of Chemical Engineering, University of Alicante, Ap. 99, E-03080 Alicante, Spain
Received 20 October 2003; received in revised form 21 September 2004; accepted 23 September 2004
Abstract
Safe animal by-product disposal is a priority target as a result of the Bovine Spongiform Encephalopathy crisis in the
European beef industry. One such disposal option is the incineration of by-product material such as meat and bone
meal residues (MBM) for the purpose of energy recovery. Although currently applied, the thermal decomposition of
MBM wastes has not been scientifically studied until now. A series of experiments has been performed to study the
thermal behavior of MBM both in inert (N2) and reactive atmosphere (air), both by thermogravimetry and in a hor-
izontal laboratory furnace. As a general trend, MBM gives low PCDD/F values, compared with incineration of other
wastes. Maximum yield of pollutants is observed at a nominal temperature between 700 and 800ꢀC.
ꢁ 2004 Elsevier Ltd. All rights reserved.
Keywords: Pyrolysis; Combustion; Animal wastes; PCDD/F; Thermogravimetry
1. Introduction
to achieve total inactivation. Germany proposed a proc-
ess for the inactivation of the prions trough the use of
destructive conditions (steam pressure > 3bar; tempera-
ture > 133ꢀC; residence time > 20min) and this process
has been applied in many European countries, including
Spain and Switzerland. The process however has many
drawbacks, since these conditions destroy a major part
of the infection agent, but not in totality. In many coun-
tries, brain and spinal cord (where large amounts of
prions are concentrated) are removed and treated
separately.
Nowadays meat and bone meal (MBM) is being de-
stroyed due to the possibility of being responsible for
the transmission of the bovine spongiform encephalopa-
thy (BSE), due to the presence of a protein species called
ÔprionsÕ, that cause Creutzfeldt-Jacob disease (Ômad cows
diseaseÕ).
Import and export of MBM to/from/within the Euro-
pean Union has been banned since December 2000.
Feeding MBM to cattle, sheep or goats has been banned
within the European Union since July 1994.
Dedicated incineration plants for MBM, as currently
used in e.g., Belgium and England, only appear reason-
able when sufficient quantities of MBM can be guaran-
teed in the long term. The most common technique is
co-incineration, mainly in cement kilns. In the USA
there are about 30 separate sites where cement kilns
are burning hazardous waste derived fuel (US EPA,
1999).
The inactivation of the infection agent is very diffi-
cult. Only strong sodium hypochlorite treatment seems
*
Corresponding author. Tel.: +34 965 903 400x2324; fax:
+34 965 903 826.
E-mail address: ja.conesa@ua.es (J.A. Conesa).
0045-6535/$ - see front matter ꢁ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2004.09.089

86
J.A. Conesa et al. / Chemosphere 59 (2005) 85–90
From a thermal point of view, MBM is a good fuel
2. Methods and materials
(net calorific value approximately 17000kJkgꢀ1, similar
to wood). It contains approximately 5% water, 22% ash,
40% carbon, 8% nitrogen, 6% hydrogen, 0.6% sulfur and
0.5% chloride, mainly as common salt. This composition
however raises concerns about dioxin when incinerating
this waste. Furthermore, MBM contains some metals
(mainly copper, lead and chromine), known to act as
catalysts in the PCDD/F synthesis reactions (Everaert
and Baeyens, 2002).
The material employed (MBM) was obtained from a
cement plant situated close to our working center and
owned by the CEMEX group. The plant requires
15tonhꢀ1 of coke and approximately 10% is replaced
by MBM materials. Table 1 presents the analysis of
the MBM sample, performed in a Carlo Erba Instru-
ment (model CHNS-O EA110), and also chloride data
obtained by X-ray fluorescence. As commented before,
most of the chloride in MBM is present as common salt
(NaCl). The MBM was well mixed in a mortar. The
amount of fatty compounds (measured by extraction
in a mixture 1:1 of dichloromethane: toluene for 24h)
is approximately 12wt%. The net calorific value
(16900kJkgꢀ1) was measured in a calorimetric bomb
(Leco Instruments) (average of two measurements).
MBM has a brownish color, a bulk weight of
680kgmꢀ3 and an intense sweet odor. It should be noted
that, if stored improperly in damp conditions, MBM
provides an ideal medium for a variety of bacteria, fungi
or vermin. The average particle size used in this study is
0.125mm.
The feeding rates in cement kilns vary from country
to country. For example, in Spain the limit is 15% of
the energy needed in the kilns, but there is no limit in
Switzerland. Incineration of MBM and tallow is seen
as useful by the cement industry, where the kiln offers
good conditions for complete combustion in terms of
temperature and time spent in the kiln. Substituting pri-
mary fossil fuels (coal or lignite) has moreover environ-
mental and economic advantages (Scheuer, 2003).
Although industrially applied, the thermal decompo-
sition of MBM wastes has not yet been scientifically
studied in detail. Papers found in literature mainly deal
with the reprocessing of the combustion residue. Knud-
sen et al. (2003) affirm that the reprocessing of the resid-
ual phosphate is difficult due to its high chloride content.
Chaala and Roy (2003) suggest its use for agricultural
soil enrichment in minerals and as a soil moisturizer.
Deydier et al. (2003) evaluated the use of the combustion
residue as a low cost substitute for materials in the proc-
ess of lead sequestration from water effluents. Concern-
ing the MBM combustion process, a paper has been
found dealing with the behavior of MBM pellets in a flu-
idized bed combustor (McDonell et al., 2001). Beck et al.
(2004) showed the effect of MBM co-combustion in a
catalyzed reactor, showing an increase in catalyst activ-
ity. Other interesting effect in the co-combustion pro-
cesses is the minimization of the emission of nitrogen
compounds when bone meal is present (Goeran et al.,
2002).
The thermogravimetric experiments were carried out
in a Netzsch Thermobalance, model TG209 controlled
by a PC system. The atmosphere used for pyrolysis
was nitrogen and for combustion was synthetic air.
The mass of the samples used was 10mg approximately.
Experiments were carried out with heating rates of 10,
20 and 30Kminꢀ1 over a range of temperatures that in-
cluded the entire range of solid decomposition, i.e., 80–
800ꢀC.
The experiments of the second part of the study were
conducted in a horizontal furnace shown in Fig. 1. The
programmed temperature of the runs was varied be-
tween 600 and 1100ꢀC. Fig. 1 also shows a typical tem-
perature profile of a combustion experiment (at 850ꢀC).
In each experiment, 0.1g of MBM was placed in the
sample holder and combusted by introducing the sample
From a thermal point of view the solid phase under-
goes two important steps in a combustion process: (i) a
pyrolysis stage, which devolatilizes the solid feed to yield
volatiles (gases and tars) and a solid char fraction; (ii) a
combustion stage, involving heterogeneous reactions of
the char to yield gaseous products and an inert residue
(ash). The pyrolysis and combustion stages may be
sequential or simultaneous, depending on the feature
of the process considered (Conesa et al., 1998; Font
et al., 2001).
holder inside the furnace at a specific velocity
(0.46mmsꢀ1) and maintained inside the furnace for
20min. After passing through the furnace, the reactor
gas was collected in an adsorptive trap containing
XAD-2 resin. After each experiment, the adsorptive trap
was extracted using toluene and the extract was analyzed
using a High Resolution Gas Chromatography–High
Table 1
Analysis of the raw material
The present work has two different and important
parts in the study of the thermal decomposition of this
material: (i) the thermogravimetric behavior of the
MBM both in nitrogen and air atmospheres is pre-
sented; (ii) a series of experiments in a horizontal labora-
tory furnace has been performed including the analysis
of the pollutants produced (mainly dioxins and furans).
C
H
S
N
Oa
Clb
0.8
40.4 6.4
Moisture 3.5%
0.5
Ash 28.7%
7.8
11.9
a
By difference.
Measured by X-ray fluorescence.
b

J.A. Conesa et al. / Chemosphere 59 (2005) 85–90
87
0.25pggꢀ1 I-TEQgꢀ1 MBM, i.e., 0.25ppt. Bearing in
mind the fat content (12%) this represents 2.08pggꢀ1
fat, similar to the content found in another MBM sam-
ple by Eljarrat et al. (2000). Other authors, nevertheless,
found a much lower content of 0.09pggꢀ1 (Guruge
et al., 2003).
Temp. (ºC)
900
FURNACE
700
Gas input
Gas
outlet
500
300
100
Magnet
Sample holder
3. Results and discussion
Distance
3.1. Thermogravimetric study
Fig. 1. Layout of the flow reactor system with associated
temperature profile.
Fig. 2 illustrates the experimental curves for Thermo-
Gravimetry (TG) and Derivative ThermoGravimetry
(DTG) (dw/dT) for pyrolysis and combustion at 10, 20
and 30Kminꢀ1. The first fact that calls our attention
is that both processes coincide until a definite tempera-
ture. From this temperature onwards, the combustion
process goes beyond the pyrolysis process, giving a
higher weight within the temperature range of 375–
545ꢀC at 10Kminꢀ1. This fact can only be explained
by an increase of weight due to the incorporation of oxy-
gen molecules into the structure of the MBM being
decomposed (by chemical reaction or adsorption), form-
ing an intermediate compound that is later decomposed.
This behavior has been explained by using a model com-
prising three different processes taking place when the
material decomposes (both in pyrolysis and in combus-
tion); furthermore in the combustion, another process
appears overlapping the third process observed in pyro-
lysis (Conesa et al., 2003).
Resolution Mass Spectrometry AutoSpec NT apparatus
for dioxin and furans production, equipped with a
Hewlett-Packard GC. The spectrometer was operated
in the electron impact ionization mode at a resolving
power of 10000. 13C-labeled PCDD/F standards (Wel-
lington Labs, Canada) were used. A quantitative deter-
mination was performed by the isotope dilution
method based on the relative response factors previously
obtained from three standard solutions (EPA1613-CS1,
CS2 and CS3).
A blank experiment has been done using the same
experimental conditions (reactor, temperature, time,
standards. . .) but with no sample. Assuming that 0.1g
of sample were introduced, the result of this blank run
shows a total amount of 0.1pgI-TEQgꢀ1 of sample as
the lower detectable limit. This quantity is much lower
than the obtained when MBM is present, as will be
shown later.
Chaala and Roy (2003) suggested that the first proc-
ess appearing in the pyrolysis is due to the dehydratation
of the animal flour; the second process would be pro-
duced by the evaporation of low molecular weight
Prior to the decomposition runs, the MBM material
has been analyzed for dioxin and furan content by using
the EPA 1613 method. The total ÔbackgroundÕ concen-
tration of these compounds in the raw material is
100
80
Pyrolysis
Combustion
10 K/min
60
30 K/min
40
20
0
20 K/min
30
230
430
630
830
Temperature (ºC)
Fig. 2. Thermogravimetric behavior at the three heating rates used in the study.

88
J.A. Conesa et al. / Chemosphere 59 (2005) 85–90
compounds, whereas the third process is a consequence
of the degradation of the bone portion, and the complete
decomposition of the remaining intermediates.
an 80% of the I-TEQ value obtained. There is a very
low yield of octachlorinated species, i.e., lower chlorin-
ated homologues are mainly formed. Although biologi-
cal matrices like meat contain traces of copper
containing enzymes as essential elements, the major part
of the chloride present in the MBM sample is in the form
of NaCl, as commented above. Addink et al. (1998)
already showed the limited chlorinating ability of NaCl
for PCDD/F formation, i.e., the increase in the amount
of NaCl does not vary the yields of PCDD/F.
3.2. Pollutant production
Fig. 3 shows the dioxin and furan production at sev-
eral temperatures, expressed in terms of (sum of all con-
geners from tetra to octachlorinated species) produced
per gram of MBM decomposed (pggꢀ1). Table 2 pre-
sents the yields obtained for each chlorination degree,
from tetra to octachlorinated. From these data, I-TEQ
values could be calculated and are also presented in
Table 2. As it is known, only 2,3,7,8 congeners contrib-
ute in the calculation of the I-TEQ value. Although it is
not shown in the tables, in almost all the experiments
sum of PCDD, PeCDF and HxCDD represents almost
Yields of PCDD/F found in this work ranged be-
tween 3000pggꢀ1 and 9000pggꢀ1. In terms of total
I-TEQ the values ranged approximately between 5 and
45pgI-TEQgꢀ1 (compared to the 0.25pgI-TEQgꢀ1 of
the raw material). There is hence a significant formation
of PCDD/F both in pyrolysis and combustion. The val-
ues obtained are apparently high in comparison with
source concentrations (US EPA, 2001) in industrial
incinerators for sewage sludge (270pggꢀ1) and munici-
pal waste incineration (20–1000pggꢀ1). The comparison
of data among raw materials is difficult, since the results
could strongly be a function of the specific conditions of
the system used. Using the experimental method of this
paper, Fullana et al. (2002) found concentrations of
25000pggꢀ1 using sewage sludge as material, whereas
Samaras et al. (2000, 2001) found 10 times higher
yields of PCDD/F for refused derived fuel (RDF) using
a horizontal reactor similar to the one used in the
present work and working at very high temperatures
(ffi1000ꢀC). Apparently, MBM gives low PCDD/F val-
ues, approximately 10% compared with incineration of
other materials. A likely explanation can be found in
the work of Samaras et al. (2000) and the following pa-
pers of this group. Meat contains large amounts of pep-
tides, which undergo formation of urea and ammonium
salts under incineration, which have been shown to be
good inhibitors for PCDD/F.
10000
9000
Furans-Pyrolysis
Dioxins-Pyrolysis
Furans-Combustion
8000
Dioxins-Combustion
7000
6000
5000
4000
3000
2000
1000
0
600
700
850
950
975
1100
Temperature (ºC)
It is important to remark that the yields obtained in
pyrolysis runs are unexpectedly high. From data in
Fig. 3. Sum of all congeners of dioxins and furans from MBM
in the horizontal reactor at several temperatures.
Table 2
Yield of each species as a function of temperature and pyrolysis/combustion condition, expressed as pggꢀ1
T
TCDF PeCDF HxCDF HpCDF OCDF TCDD PeCDD HxCDD HpCDD OCDD Total
I-TEQ
(ꢀC)
Pyrolysis
Pyrolysis
Pyrolysis
Pyrolysis
Pyrolysis
700
850
87.5
1.9
71.6
60.9
3.7
65.7
189.9
45.5
11.3
2863
17.2
0.294
1.65
0.052
44.8
69.1
6.9
72.1
60.2
7.9
45.1
234.8
28.8
1.95
17.1
0.319
0.840
0.098
0.079
0.199
7.7
20.2
950
975
8.2
6.38
18.1
70.5
11.6
6.62
0.796
1.74
102.9
31.9
40.4
22.6
3.48 0.029
6.8
1613
132.5
6.9
107.6
3076
1100
10.6
0.058
16.1
Combustion
Combustion
Combustion
Combustion
Combustion
600
700 1434
28.4
286.9
5840
2995
410.2
52.1
225.4
644.0
308.4
76.2
105.4
311.2
101.4
16.2
0.489
6.8
131.2
476.4 6025
311.9
1101.5
2107
172.5
2.6
92.2
282.3
52.3
20.5
33.7
82.2
0.299
3.1
1.7
5.62
25.7
43.0
1.31
850
950
975
294.4
29.9
2.4
0.591
1.2
2213
403.5
59.9
1126
948.5
271.1
62.9
2.6
0.129
1.4
1855
625.5
65.9
333.1
0.611
3.3
297.2
325.3
1.40
5.01
Combustion 1100
41.7
534.1
102.5

J.A. Conesa et al. / Chemosphere 59 (2005) 85–90
89
Table 2, the I-TEQ value obtained in the present work in
the absence of oxygen (pyrolysis) represents 50% of the
value obtained in combustion runs. This is surely related
to the amount of oxygen present in the MBM itself. As
presented in Table 1, tested MBM contains a 11.9% of
oxygen, i.e., a total of 1.19 · 1011 pg of oxygen/gram of
sample, that is an amount obviously very much higher
than the amount needed to form thousands of pg of
dioxins/furans: part of these oxygenated organic com-
pounds could form precursors like phenols and esters
that could produce dioxins and furans in the absence
of oxygen. The amount of oxygen needed to dioxin for-
mation is very low.
(German Federal Environmental Agency, http://
www.umweltbundesamt.de/).
Acknowledgement
Support for this work was provided by CYCIT-
Spain, Research project PPQ2002-00567.
References
Addink, R., Espourteille, F., Altwicker, E.R., 1998. The role of
inorganic chloride in the formation of PCCD/F from
residual carbon on incinerator fly ash. Environ. Sci.
Technol. 32, 3356–3359.
In relation to the temperature behavior, we found
the maximum production of PCDD/F between 700
and 800ꢀC. Bearing in mind that the reactor is non-
isothermal, there is no proof that the PCDD/F forma-
tion occurs at this temperature. Note that the
maximum total yield is obtained at 700ꢀC, but the
maximum in terms of I-TEQ is produced in the runs
performed at 850ꢀC, due to the dominant presence
of HxCDF, HxCDD and TCDD. It is know that the
production of products of incomplete combustion is
higher at 700–900ꢀC (Fullana et al., 2000). These com-
pounds react in the post-combustion cooling zone to
form PCDD/F. At temperatures higher than 950ꢀC
the I-TEQ decreases drastically to less than 10pg
I-TEQgꢀ1. It has been reported that under less-than-
optimum conditions (such as insufficient oxygen),
PCDD/Fs may be formed at temperatures that nor-
mally would exceed the threshold temperature of
800ꢀC for PCDD/F formation (Gullett et al., 1990;
Walker and Huntley, 1997). Insufficient oxygen within
the furnace frequently leads to the formation of vari-
ous PAH-type compounds, especially during the
phases of devolatilization.
Altwicker, E.R., 1996. Formation of PCDD/F in municipal
solid waste incinerators: laboratory and modelling studies.
J. Hazad. Mater. 47, 137–161.
Beck, J., Brandenstein, J., Unterberger, S., Hein, K.R.G., 2004.
Effects of sewage sludge and meat and bone meal co-
combustion on SCR catalysts. Appl. Catal. B: Environ. 49,
15–25.
Chaala, A., Roy, C., 2003. Recycling of meat and bone meal
animal feed by vacuum pyrolysis. Environ. Sci. Technol. 37,
4517–4522.
Conesa, J.A., Font, R., Fullana, A., Caballero, J.A., 1998.
Kinetic model for the combustion of tire wastes. Fuel 77,
1469–1475.
Conesa, J.A., Fullana, A., Font, R., 2003. Thermal decompo-
sition of meat and bone meal. J. Anal. Appl. Pyrol. 70, 619–
630.
Deydier, E., Guilet, R., Sharrock, P., 2003. Beneficial use of
meat and bone meal combustion residue: an efficient low
cost material to remove lead from aqueous effluent. J.
Hazard. Mater. B 101, 55–64.
Eljarrat, E., Caixach, J., Rivera, J., 2000. Determination of
PCDDs and PCDFs in different animal feed ingredients.
Organohalogen Compd. 47, 330–334.
Fig. 3 also illustrates that the ratio PCDD/PCDDF is
close to one, whereas the typical concentration of PCDF
in combustion of wastes is several times higher that the
concentration of PCDD. Data show that there is a slight
variation of the ratio PCDD/PCDF with temperature,
but it is too small to discuss it from a mechanistic point
of view. The ratio D/F is found to be higher than one,
only in the case precursor-synthesis (chlorobenzenes or
chlorophenols). A high value of the D/F ratio is an indi-
cator that a precursor of PCDD is present in the original
sample (Altwicker, 1996; Everaert and Baeyens, 2002).
Some chlorobenzenes and chlorophenols are used as
bactericide in waste disposal to prevent decomposition
of organic material. In this sense, Ghorishi and Alt-
wicker (1996) found a similar PCDD/F distribution dur-
ing combustion of dichlorobenzene, which is sound with
presence of a precursor in the MBM. The analysis of
these compounds was unfortunately not performed in
the raw MBM, but a reference to a similar material con-
taining 0.55mgkgꢀ1 of organic chlorine has been found
Everaert, K., Baeyens, J., 2002. The formation and emission of
dioxins in large scale thermal processes. Chemosphere 46,
439–448.
Font, R., Fullana, A., Conesa, J.A., 2001. Analysis of the
pyrolysis and combustion of different sewage sludges. J.
Anal. Appl. Pyrol. 58–59, 931–945.
Fullana, A., Font, R., Conesa, J.A., Blasco, P., 2000. Evolution
of products in the combustion of scrap tires in a horizontal
laboratory scale reactor. Environ. Sci. Technol. 34, 2092–
2099.
Fullana, A., Font, R., Conesa, J.A., Sidhu, S., 2002. Investi-
gation of catalytic activity of sewage sludge combustion ash
for PCDD/Fs formation. Organohalogen Compd. 56, 131–
134.
Ghorishi, S.B., Altwicker, E.R., 1996. Rapid formation of
polychlorinated dioxins/furans during the heterogeneous
combustion of 1,2-dichlorobenzene and 2,4-dichlorophenol.
Chemosphere 32 (1), 133–134.
Goeran, O., Wang, W., Ye, Z., Bjerle, I., Andersson, A., 2002.
Repressing NOx and N2O emissions in a fluidized bed
biomass combustor. Energy Fuels 16 (4), 915–919.

90
J.A. Conesa et al. / Chemosphere 59 (2005) 85–90
Gullett, B.K., Bruce, K.R., Beach, L.O., 1990. The effect of
metal catalysts on the formation of polychlorinated
dibenzo-p-dioxin and polychlorinated dibenzofuran precur-
sors. Chemosphere 20, 1945–1952.
Samaras, P., Sdodras, G., Sakellaropoulos, G.P., Blumen-
stock, M., Schramm, K.M., Kettrup, A., 2001. Toxic
emissions during co-combustion of biomass-waste wood-
lignite blends in an industrial boiler. Chemosphere 43, 751–
755.
Guruge, K.S., Seike, N., Yamanaka, N., Miyazaki, S., 2003.
Contamination status of animal feeds by PCDDs, PCDFs
and dioxin-like PCBs. Organohalogen Compd., 60–65.
Knudsen, N.O., Henriksen, N., Hundebol, I., Wieck-Hansen,
K., 2003. Co-combustion of meat and bone meal with
natural gas. VGB PowerTech. 83 (11), 81–82.
Scheuer, A., 2003. Utilization of alternative fuels and raw
materials (AFRs) in the cement industry. Cement Int. 1 (1),
48–50.
US EPA, 1999. Technical Support Document for HWC MACT
Standards, vol. I. Description of Source Categories, US
EPA Office of Solid Waste and Emergency Response.
US EPA, 2001. Database of sources of Environmental Releases
of Dioxin-like compounds in the United states, EPA/600/C-
01/012, EPAs Office of Research and Development.
Walker, W.J., Huntley, S.L., 1997. A literature review of
formation and release of PCDD/Fs from gas manufactur-
ing: a previously unidentified source? Chemosphere 35,
1409–1422.
McDonell, K., Desmond, J., Leahy, J.J., Howard-Hildige, R.,
Ward, S., 2001. Behaviour of meat and bonemeal/peat
pellets in a bench scale fluidised bed combustor. Energy 26,
81–90.
Samaras, P., Blumenstock, M., Lenoir, D., Schramm, K.W.,
Kettrup, A., 2000. PCDD/F prevention by novel inhibitors:
addition of inorganic S- and N-compounds in the fuel
before combustion. Environ. Sci. Technol. 43, 751–755.