Professional Documents
Culture Documents
Review Articles
Jürgen Vehlow*, Britta Bergfeldt, Hans Hunsinger, Helmut Seifert and Frank E. Mark1
Forschungszentrum Karlsruhe GmbH, Institute for Technical Chemistry, Karlsruhe, Germany
1 Dow Europe, Horgen, Switzerland
Introduction
DOI: http://dx.doi.org/10.1065/espr2003.02.147
Abstract
Bromine is the third element in the group of halogens and
its chemical properties are similar to those of the by far more
Intention, Goal, Scope, Background. The halogen bromine is abundant chlorine. It is used for its chemical and biological
far less abundant than chlorine, but it can be found at high activity in special applications like photographic chemicals
concentrations in special materials like flame retarded plastics.
The fate and effects of Br in waste incineration are not well
and pharmaceuticals. Major Br products are flame retard-
understood. It may have similar implications like Cl for the ants and, hence, it is found in high concentrations, in part,
volatilisation of heavy metals and the formation of low volatile in some textiles and primarily in engineering plastics in the
organic compounds. Due to its lower oxidation potential, there electrical and electronic, as well as in the building industry.
is a risk of formation of elementary Br2 in the offgas. End-of-life products from these sectors are main Br sources
Objective. Co-combustion tests of different types of Br contain- in municipal solid waste.
ing plastic waste materials (up to 22%) and MSW in the
TAMARA pilot plant for waste incineration were conducted to There is not much valid information found in literature on
investigate the Br partitioning and the influence of Br on metal the effects of Br in waste incineration. The main reason for
volatilisation. this lack of knowledge is its low concentration in municipal
Methods. The Br inventory of the fuel mix was elevated to approx. solid waste (MSW). However, since the disposal and espe-
1 wt-%. All input and output mass flows of the furnace have been cially the incineration of chlorine containing plastic waste is
sampled and the partitioning of Cl, Br, S, and a number of heavy of high public concern due to suspected dioxin formation, it
metals, has been calculated on the basis of closed mass balances. is no surprise to see this discussion being extended to waste
Results and Discussion. Organically-bound Br was typically re- fractions with a higher Br inventory. Other potential effects
leased to more than 90% into the raw gas. Elementary Br2 was are the formation of elementary Br2 in the offgas and the
detected at high Br levels. Its presence was always analysed when volatilisation of some toxic heavy metals, both in view of
all SO2 in the raw gas was oxidised to SO3. Br enhances the
emission implications.
volatilisation of metals like K, Zn, Cd, Sn, Sb, and Pb out of the
fuel bed principally in the same way as Cl. The tests gave strong Consequences could be expected for the disposal of plastic
indication that the promoting influence of the halogens on metal waste from electrical and electronic equipment (WEEE), since
volatilisation is more pronounced than that of the fuel bed tem-
about 11% of the plastic material is flame retarded and some
perature. The volatilised metals are condensated on the fly ashes
and are discharged along with the filter ashes. applications make use of brominated organic compounds.
Other waste streams with a significant Br level due to their
Conclusions. As long as a surplus of SO2 is present in the raw
gas no Br2 is formed. Although the halogen induced transfer out flame retardancy are rigid polyurethane (PUR) and extruded
of the fuel bed causes high concentrations of volatile metals in polystyrene foams (XPS) from the building sector.
the filter ashes, a recovery is not economically feasible for the
The Forschungszentrum Karlsruhe and industrial project
time being. The volatilisation gives no rise to metal emission
problems as long as efficient dedusting is achieved. groups under the umbrella of APME, the Association of
Plastics Manufacturers in Europe, have conducted semi-
Recommendation and Outlook. If there is a risk of Br2 forma-
tion, in wet scrubbing a reducing agent has to be added to the technical tests on the behaviour of various fractions of Br-
neutral scrubber for efficient abatement. Filter ashes should be containing end-of-life plastic materials in co-combustion
disposed of in a way that enables access for recovery in the fu- with municipal solid waste (MSW) (Mark 2000, Vehlow et
ture. The exact volatilisation characteristics of the various metals al. 1995, 2000, 2002). The tests were performed in the
have to be studied in future using specifically tailored experiments. TAMARA pilot plant for waste incineration located at the
Institute for Technical Chemistry in Karlsruhe, Germany.
Keywords: Bromine; chlorine; co-combustion; heavy metals; par-
The most important findings concerning the Br partition-
titioning; plastic waste; volatilisation; waste from electrical and
electronic equipment; waste incineration ing and the influence of Br on metal volatilisation are de-
scribed in this paper.
ESPR – Environ Sci & Pollut Res 10 (5) 329 – 334 (2003)
329
© ecomed publishers, D-86899 Landsberg, Germany and Ft. Worth/TX • Tokyo • Mumbai • Seoul • Melbourne • Paris
Bromine in Waste Incineration Review Articles
1 Experimental Methods and Materials (Vehlow et al. 2002). The materials used were mixed WEEE
All experimental investigations have been performed in the shredder residue (2 types), shredded television back and front
Karlsruhe TAMARA pilot plant for waste incineration. plates (2 types), and shredded printed wiring boards.
TAMARA is a scaled-down model of a grate combustion
furnace equipped with a boiler, a fabric filter and a two- 2 Parameters Controlling Element Volatilisation
stage wet scrubbing system. The geometry of the combus-
tion chamber can be varied between a contra, centre and All waste fractions contain metals. Their concentration varies
parallel flow configuration by installation of variable roof and so does their speciation. All different species from pure
elements. All configurations have been used during the tests. metals to even rare compounds may be analysed. In plastic,
metals serve various purposes: they are used as catalysts, fill-
The thermal capacity of TAMARA is 0.5 MW at a through- ers, flame retardants, or pigments, depending on their field of
put of 200–300 kg/h. The basic fuel comprises typically 75 application. Some plastic waste fractions, like Automotive
wt-% of shredded and homogenised residential waste and Shredder Residues (ASR), may contain pure metals as well as
25 wt-% of pelletised refuse derived fuel (RDF). This fuel inorganic and organic metal compounds (Mark 1998).
mix well represents today's Central European municipal solid
waste without separating or pre-sorting activities. Metals of concern are Cu, Zn, Cd, Sn, Sb, Hg, and Pb. The
emission of these metals is subject to most legislative regula-
The investigation of Br partitioning and the effects of Br on
tions of air emission from waste incineration. That is why
metal volatilisation were major topics of the co-combustion
trials with Br containing plastic waste fractions. The test the volatilisation of these metals into the raw gas was an
programmes in TAMARA hence comprised an extensive important topic to be investigated in the co-combustion tests.
sampling and analysis programme in order to establish bal- The extremely volatile Hg was omitted, since it was only
ances not only for halogens, but also for metals. Samples present in traces in the co-combusted plastics.
were taken from all input and output mass streams follow- The partitioning of elements in waste incineration is con-
ing the recommendations of the International Ash Working trolled by a number of parameters:
Group (International Ash Working Group 1997). Fly ashes • their speciation in the fuel,
were not taken from the fabric filter discharge flow, but • the total chemical composition of the fuel, that is the concentra-
isokinetically sampled in a gas duct downstream of the boiler. tion and the speciation of major potential reactants with the ele-
ment in question,
From industrial waste incineration, it is known that, at high • the temperature they are exposed to in the fuel bed on the grate
Br concentrations in the fuel, elementary Br2 may be found in and
the raw gas. Br2 is insufficiently absorbed in wet scrubbing • the time of this exposure.
systems if no reducing agent is added to the neutral scrubber.
For the species-specific analysis of the total Br concentration The fate of an element in a well defined, closed and homo-
in the raw gas, hence, a special sampling train was used. The geneous chemical system can in principle be predicted by
train comprised a condenser, a first acidic scrubber for HBr calculation of the thermodynamic equilibrium (Alvin 1977,
absorption, and a second neutral scrubber with the addition Frandsen 1994). The preferentially formed compounds are
of Na2S2O3 for the absorption of elementary Br2. characterised by negative free formation enthalpies. Ther-
modynamic data for many systems are available meanwhile
In two research programmes, plastic foams from the building and commercial programmes can also be found for such
sector (PUR and XPS), flame retarded by brominated organic calculations (Barin 1996).
compounds like HBCD and brominated polyols, were co-
combusted (Vehlow and Mark 1995). The addition of foams In the case of waste incineration, however, the chemical sys-
was limited to 3 wt-% which accounted for approx. 20–30 tem cannot be sufficiently defined, it is not homogeneous
vol-% of the input due to the low density of the material. nor closed. The chemical speciation of elements in the waste,
Various types of WEEE were subject of two further test pro- the temperature in the fuel bed, and the respective residence
grammes (Vehlow and Mark 2000). In the first programme, time are widely unknown. Thermodynamic calculations are
up to 12 wt% of 3 plastic materials from the consumer (small therefore not meant to deliver precise transfer numbers for
appliances, kitchen machines, PC, TV), commercial (medium a specific element in waste combustion. Such calculations,
size parts, PC, fax machines), and industrial sector (mix of nevertheless, supply valuable information concerning which
electrical housings, printed circuit boards, main frame com- compounds are most likely formed, and thermodynamic
puters) were added to the basic fuel. A fourth synthetic mate- modeling is a useful tool to establish overall strategies for
rial simulated the composition of the average European poly- the control of element fluxes.
mer mixture for this sector today. The maximum Br inventory
in the fuel mix during this campaign was approx. 0.25 wt-%. Experimental investigations in waste incineration plants also
need to consider a number of factors which may limit their
A second WEEE test campaign was conducted with an ad- accuracy:
dition of other end-of-life flame retarded plastic materials.
• the waste fuel is rather inhomogeneous, even if its input is well
This time up to 22 wt% of the plastic waste were added to controlled by the crane operator,
the reference waste fuel in order to increase the Br inventory • the mass flow of the solid residues like the grate ashes in full-
of the fuel mix up to 1 wt% and to check for the option of scale plants is mainly taken from long-term averages and its
Br recovery from the effluents of wet scrubbing systems error can be estimated to exceed 10%,
• the sampling and sample pretreatment – again first of all for the boiling points of the bromides of the above mentioned met-
inhomogeneous solids – may cause errors of more than 20%, als are slightly lower then those of the respective chlorides,
• the fuel bed temperature is unknown in full-scale plants, and as is documented in the compilation in Table 1. In MSW
even in TAMARA it is only measured at 8 central positions along
the grate length.
incineration, no specific influence of Br will be detected, since
the Br level in MSW is typically lower than that of Cl by a
The analytical error, in most cases, can be kept <1% and can factor of 50–100 and the formation of bromides can be ne-
– compared to these major sources of error – be neglected. glected. However, it must be expected that, at an elevated
Br inventory of the fuel, those compounds are formed and
Taking these impediments into account, the total error of the
are to a great extent volatilised out of the fuel bed.
inventory of homogeneously distributed major constituents
of the waste fuel like Cl is estimated to be at least ± 10%. The A comparison of chemical effects needs to be based on the
determination of metals like Cu or Pb, which may partly be molar rather than on the mass concentration of the halo-
present as metallic species in the input and in the grate ashes, gens. The atomic weight of Br is approx. 80 amu compared
will suffer from higher errors and be on the order of ± 20% to approx. 35.5 amu of that of Cl. This fact, furthermore, is
and more. Hence, the following discussion of correlations be- impeding the evaluation of the influence of Br against that
tween metal partitioning, halogens, and temperatures can only of Cl on the volatilisation of metals.
provide indications and further theoretical as well as experi-
mental work is needed to establish a more exact description
3 Results and Discussion
of the interdependencies of the various parameters control-
ling the fate of elements in waste incineration. 3.1 Partitioning of Br
A number of experimental results of element partitioning The prevailing combustion products of halogens in MSW
derived from sampling experiments in full-scale plants have incineration are their hydrides (e.g. HCl, HBr), which are
been published (Brunner 1986, Schneider 1986, Reimann released into the gas phase. The transfer of Cl into the raw
1989, Angenend 1990, Belevi 2000). The data obtained re- gas is approx. 80–90% of its inventory in the fuel (Interna-
veal a significant correlation between the Cl inventory of tional Ash Working Group 1997). Br seems to stay in the
the fuel and the volatilisation of a number of metals. The bottom ashes to a greater extent. Its balance is depicted in
same correlation was found in the TAMARA pilot plant Fig. 1 for the co-combustion of XPS foams at 950°C. The
which can be operated under much more defined conditions partitioning found here is typical for all co-combustion tests
than a full-scale waste incinerator (Vehlow 1996). The pro- with Br containing plastics. It indicates that Br bound in or-
moting effect of Cl on metal volatilisation is easily explained ganic matrices is almost totally transferred out of the fuel bed.
by the fact that the chlorides typically have lower boiling This can be explained by the fact that halogen containing or-
points and higher vapour pressures than most other metal ganic compounds already decompose at lower temperature in
compounds. Furthermore, the formation of chlorides is very the pyrolysis and gasification zone on the grate, and release
likely in the fuel bed in a waste incinerator, due to the rather halogen hydrides or short chain halogen containing organic
high Cl concentration of 0.5–1 wt-% in MSW. compounds into the gas phase (Lenz 1979). Inorganic halo-
According to its close chemical relationship, Br should have gen compounds are typically decomposed or evaporated at
the same effect as Cl for most volatile metals. In fact, the much higher temperature and hence have a greater tendency
to stay in the bottom ashes (Hunsinger 1994).
Table 1: Boiling points of selected halogenides
The total Br concentration measured in the raw gas during
Compound b.p. in °C
the first WEEE campaign reached approx. 250 mg/m3. Up
KCl 1407 to this level, HBr was the only detected Br species. In the
KBr 1380
FeCl2 1012
FeBr2 967
Cu2Cl2 1367
Cu2Br2 1345
ZnCl2 732
ZnBr2 650
CdCl2 960
CdBr2 863
SnCl2 652
SnBr2 620
SbCl3 283
SbBr3 280
PbCl2 950
PbBr2 916 Fig. 1: Balance of Br during XPS co-combustion at 950°C
Fig. 5: Volatilisation of Sn as a function of the accumulated Cl+Br concen- Fig. 7: Volatilisation of Sn as a function of the Cl and Br inventory of dry fuel
tration and the maximum fuel bed temperature