Environmental Technology

in the context of solid waste incineration

October-November 2021
DISCLAIMER

The booklet you have in hand, the extracts from study reports and the reproductions
of documents it contains are intended exclusively for the teaching of the course
ChimH409 at the Polytechnic School of Brussels (ULB) and may not be distributed
outside the University.

Le présent fascicule, les extraits d'études et les reproductions de documents qu'il

contient sont exclusivement destinés à l'enseignement du cours ChimH409 à l'Ecole
Polytechnique de Bruxelles et ne peuvent être diffusés hors de l'Université.

Brussels, November 2021

 Part I. Waste incineration
Introduction
Economic growth generally implies growing demand for raw materials, energy and other natural
resources, and growing amounts of materials that end up as waste if not properly managed. Main
concerns relate to:
 the pressures exerted on natural assets (finite resources);
 the negative environmental impacts from the extraction, processing and use of materials;
and, at end-of-life of the products:
 the negative impacts that inappropriate waste management has on human health and the
environment. We documented the latter aspects in the ChimH302 class, namely: air, soil and water
pollution; climate change; degradation of natural habitats, biodiversity and ecosystems.

Addressing these concerns at the right geographical scale(s) can materialize after enacting and
implementing the right regulatory texts and legislation. Nowadays the underlying important concept in
this context is the so-called Waste Hierarchy (initially suggested by Dutch politician Lansink).

Fig. F01 The Waste Hierarchy

The strategy is thus a hierarchy of five treatment steps, categorized according to their environmental &
sustainability impact, and is now at the core of the waste management policy in the EU. The reader is
referred in this respect to the European Waste Management Directive 2008/98/EC.

In practice this strategy will rely on the consistent deployment of various environmental technologies. We
shall cover a number of these in this class, with a special focus on Waste-to-Energy (WtE) plants. From
an environmental engineering point of view, WtE installations are highly informative because, firstly,
they significantly contribute to the "circular economy" way of doing things. Secondly they have
themselves to be equipped with an impressive array of depollution equipment that we are going to
document in this booklet. These techniques and pieces of equipment are usually shown to the students
through an in-depth technical visit of the Brussels-Energy WtE plant, quai Monnoyer in Neder-over-
Hembeek (Fig. F02). It is the reason why a number of illustrations directly coming from the Brussels-
Energy WtE plant have found their place in the booklet.

1
ChimH409 Nov2021
Fig. F02. The Brussels-Energy plant gets rid of our domestic refuse, produces electricity and provides
district heating to the nearby area. It has to do all this while complying with stringent environmental
standards (www.bru-energie.be).

1.1 Recent evolution of the industry

1.1.1 Multi-objectives

In old times, incineration of waste was only perceived as a way to eliminate the waste (reducing both
volumes and masses), avoid bad smells and prevent diseases. In the last fourty years the combustors have
considerably progressed in the field of environmental protection and this is reflected in Table F01. But in
the last twenty years, sustainability preoccupations (and care for recovery of valuable resources as we just
mentioned) have become paramount. What we have to describe now are genuine Waste-to-Energy
conversion plants (the modern variant to the incinerators of the past).

Table F01. Time-history of environmental technologies put in place to clean flue gases generated by
incinerators.

#0 < 1970 no real preoccupation for air pollution (no relevant regulatory framework)
#1 1970-1985 Flue gas cleaning is very basic: a dry electrostatic precipitator removes particles
together with the pollutants directly associated to them (no treatment of acid gases)
#2 1975-1990 Semi-dry flue gas cleaning with injection of lime (no activated carbon at this stage)
and subsequently removal of particles in a baghouse filter.
#3 1990-2000 Wet flue gas cleaning following particle removal in an electrostatic precipitator. The
system provides wet scrubbing in an acidic scrubber system (pH ≈ 1.5) and
subsequent wastewater cleaning.
#4 2000- Wet flue gas cleaning (as described in #3) with an additional 'alkaline' scrubber
(pH ≈ 6.5) followed by a bag house filter with activated carbon injection for
removal of dioxin and mercury and selective noncatalytic reduction (SNCR) of
nitrogen oxide (de˗nitrification / deNOx)
#5 2005- Semi-dry flue gas cleaning (as #2 but with significant increased efficiency) with
addition of activated carbon and SNCR for deNOx
#6 since 2010 Semi-dry flue gas cleaning and dioxin filter (as described in #5) followed by flue
gas condensation and selective catalytic reduction (SCR) instead of SNCR for

2
ChimH409 Nov2021
 nitrogen oxide removal (deNOx)

It has to be noted that in an EU context, neither #1, #2 and #3 air pollution control technologies would
suffice to fulfil current legal requirements of today (refer to IED Industrial Emissions Directive
2010/75/EU, Table F04 at the end of the booklet).

Given the thorough cleaning generally practiced, it seems unlikely that the limit values imposed by the
EU directives would significantly evolve even further. Yet, it cannot be excluded that still new parameters
would be brought forward, such as nanoparticles and nitrous oxide (N2O).

1.1.2. Types of burners most often encountered in practice

As indicated by Buekens (2013) thermal treatment of waste has now evolved into convincing mature
technology worldwide, with:
 fluidized bed burning units for sewage sludge (namely, the biosolids coming from the treatment of
wastewater), as well as for co-firing wastes with extremely dissimilar properties;
 mechanical grate incinerators as standard in municipal solid waste (MSW) incineration;
 rotary kiln plants (or even cement rotary kilns) for firing various industrial and hazardous waste.

Actually each of these units still has some technical limitations. Fluid bed units require steady, size-
reduced feed and may experience loss of fluidization in the presence of low-melting ash and at too high
temperatures. Mechanical grate stokers must support the waste during combustion, yet have difficulty in
coping with both very wet and high-calorific waste. For rotary kiln units, gas phase mixing and wear are
major problem areas. There are thus further developments and engineering advances which are still
needed in the domain.

1.2. Waste in large volumes that we have to address

We shall speak in particular of three types of wastes.

1.2.1. Sewage sludge

The biological treatment of wastewater from domestic and many industrial sources involves a series of
process steps aimed at either converting undesirable pollutants from a solubilized state to a solid form or
removing the solids from the bulk flow. Although the processes used may involve some reduction in the
pollutants (e.g., through biological oxidation), the majority of the pollutant mass that enters the plant exits
the wastewater treatment plant as a constituent of biological sludge or biosolids.
The categories of biosolids include
 Primary treatment sludge—the sludge generated through settling early in the treatment sequence;
 Secondary treatment sludge—the excess biomass generated in the activated sludge tanks under
aerobic conditions through biological reproduction and growth whereby soluble pollutants are
converted into cell mass;
 Tertiary treatment sludge—biomass generated in advanced wastewater treatment such as
denitrification.

Other solid streams are often generated in the course of wastewater treatment though in lesser quantities
than the sludges:
 scum: the floatable material accumulating on clarifiers or recovered in aerated grit chambers at
plant inlet;
 screenings: the rags, twigs, and other large solids screened from the entering wastewater, and
 grit: the coarse sandy/silty solids removed following screening.

3
ChimH409 Nov2021
Having captured the solids, some plants move directly to some kind of ultimate disposal system such as
landfill or direct land application.

Fig. F03. The modern routes followed nowadays to valorize wastewater sludges

Most plants interpose one of several processes ahead of disposal in order to accomplish several ends:
pasteurization or other step to kill off pathogenic organisms and stabilization of the sludge such that
continued vigorous biological activity stops or is slowed. Common processes to accomplish these ends
include lime addition and digestion. Lime addition involves mixing lime (calcium oxide) with the sludge
to raise both the temperature (from the heat of hydration/slaking) and the pH to kill pathogenic organisms
and stop biological activity. This often involves addition of 20% to 30% lime (by weight on dry solids),
so a considerable increase in both mass and ash content occurs. In case of land application of sludge for
soil amendments, the lime addition is especially welcomed by farmers (who can then limit the lime
quantities that they have to purchase externally).

Digestion remains one of the important sludge to resources routes (Fig F03). It involves continuing
biological degradation of organic compounds bound in the sludge. The process can be conducted under
aerobic conditions (with excess oxygen) or anaerobic conditions (oxygen-deficient). The latter involves
thermophilic, methane-forming organisms and, consequently, generates fuel values as digester gas that
have potential application in incineration or other energy-intensive processes as well as for sludge
preheating. Clearly, the conversion of a portion of the biomass to methane fuel (the digester gas is about
50% methane and 50% CO2) results in some depletion of the fuel value (higher ash-to-combustible ratio)
of the stabilized sludge solids. Further, digestion adversely impacts the dewatering characteristics of the
sludge.

The means for ultimate disposal of wastewater treatment solids has evolved over the years. For
wastewater treatment plants that are in or near to rural areas with easy access to farmlands and that
generate sludge having low concentrations of the heavy metals mercury, cadmium, chromium, lead, and
nickel, application of liquid sludge (2% to 5% solids), dried sludge products or composted sludge to
croplands has historically been encouraged by many regulatory authorities and found to be cost-effective
and generally acceptable to the sludge users. In more urbanized areas where the distance to and scale of
land application would be unreasonable and/or where intensive industrialization has led to the presence of

4
ChimH409 Nov2021
significant concentrations of heavy metals in the sludge, incineration in dedicated burners (such as the
ones detailed below) followed by ash landfilling is often practiced.

As one might suspect from the above, the characteristics of wastewater treatment sludge are strongly
related to (1) the mix of domestic, commercial, and industrial wastewater types involved and (2) the
process flowsheet of the treatment plant. Thus, a bedroom-community wastewater plant usually generates
a highly organic, clean sludge, whereas a heavily industrialized city generates a sludge that can be heavily
contaminated with undesirable heavy metals and, even, toxic organic pollutants. Most historic urban
agglomerations, especially in Europe, rely on combined sewers where runoff from storms is carried (and
treated) through the same sewer system as the sanitary sewage. Thus, contaminants and the content of
inert, soil-derived materials in the sludge (scoured from the sewer lines during high storm flows) are
dependent on rainfall patterns. Changes and upsets in the treatment plant can significantly alter the
characteristics of treatment and the performance of dewatering equipment. The conclusion one must draw
is that the characteristics of wastewater sludges are exceedingly variable. Therefore, flexibility in the
ability to respond to changes is an important process feature of a satisfactory sludge incineration system.

In addition to this, the elemental chemical composition of biosolids differs significantly from that of most
other wastes considered for incineration. This derives from its nature as a biomass comprised massively
of POU (proteins of unicellular organisms, typically bacteria). This tends to increase the phosphorous,
nitrogen, and, to a degree, sulfur content when compared to typical refuse materials.

1.2.2 Municipal solid waste and assimilated

The waste being led to an urban incineration facility often consists of several types of waste, such as
household waste, some bulky waste, commercial and institutional waste and some industrial waste. This
is because a lot of non-domestic waste actually bear a close resemblance to household refuse and can thus
be handled and mass-burned alongside them without problem.

Table F02. Typical fractions of municipal solid waste (MSW).

Hawf : ash and water free calorific value

5
ChimH409 Nov2021
In some cases sewage sludge or even clinical waste are considered to be incinerated within the same
facilities, after some technical adaptations being brought primarily to the feeding system. Street-sweeping
residues and construction and demolition waste containing a large fraction of incombustible matter are
obviously less suited for incineration.

The different waste types received at the incinerator have significantly different characteristics. The
different fractions are listed in Table F02 showing the marked differences in terms of heating value,
moisture content and ash content. These are the three key variables in characterizing waste as a fuel (see
further discussion in Section 1.3). Design and operation of an WtE also need to take into consideration
that the waste amounts and characteristics can evolve significantly over time, on week-season-year
timespan. The increase of the selectivity of the refuse collection in a city like Brussels (orange-blue-
yellow-white bags) obviously has consequences in terms of the heating value of what is sent year after
year to the WtE furnaces.

1.2.3 Refuse-derived fuel

Now, rather than firing waste as it comes, one can convert it into storable fuel, following a suitable
sequence of operations (Fig. F04), composed of primary and secondary shredding, grading, wind sifting
and screening, magnetic and eddy-current separation, etc. Suitable combinations of such operations may
convert municipal solid waste, packaging, wood, paper and plastics, etc., into better manageable and
storable refuse-derived fuel (RDF) with more predictable characteristics and specifications, such as
HHV, and proximate and ultimate analysis. RDF assumes different forms, such as fluff, powdered (after
adding embrittling agents), or densified, i.e., in bales, pellets (Fig. F05).

6
ChimH409 Nov2021
Fig F04. Separation table (Umweltbundesamt 2018) used as one of the steps to pretreat garbage just after
shredding. A combination of aerodynamic classification with vibrations allows to selectively recover the
light fraction which is the one which contains most of the organic fraction and heating value that we want
to concentrate in the RDF. The heavy fraction recovered on the left has a marked metallic component and
will be exposed to ferromagnetic and eddy-current separation; after that it will be disposed of (no passage
through a furnace as it would be useless).

Fig F05. Refuse-derived-fuel (RDF) pellets in their final molded form, ready to be incinerated in a
fluidised bed burner

There are efforts, spearheaded especially in North America, to try to standardize RDF, to improve its
acceptance and access to the energy markets. In Europe operators are not convinced by RDF, weary of the
high costs of investment & repair needs associated to the process of preparing the pellets. Fire and
explosion hazards have also been associated with the shredding and drying steps.

1.3. The characteristics of solid waste which could position it as a valid fuel

Waste varies erratically in composition and properties and these greatly influence the selection of WtE
furnaces, heat recovery, and flue gas cleaning.

A. Waste Component Analysis

For many waste streams, an acceptable and useful analysis is obtained by weighing the waste after
separation into visually definable components. For municipal solid waste, the component categories often
include newsprint, corrugated cardboard, ‘‘other paper,’’ food waste, yard waste, aluminum, metal
(excluding aluminum), glass, leather, rubber, textiles, plastics, and miscellaneous. The use of these waste
categories has obvious advantages in the separation step. Further, the weights in each category are often
useful in assessing recycling processes and in monitoring trends in waste sources and composition.

B. Proximate Analysis
The balance between moisture, combustible, and ash content, and the volatilization characteristics of the
combustible fraction at high temperatures are important properties affecting combustor design and
operation. Quantitative knowledge of these properties gives considerable insight into the nature of the
pyrolysis and combustion processes for the waste considered to be burned. A simple and relatively low-
cost laboratory test that reports these properties is called the proximate analysis. The procedure includes
the following steps:

7
ChimH409 Nov2021
It should be recognized that the value reported as ‘‘moisture’’ includes not only free water but also,
inadvertently, any organic compounds (e.g., solvent) with significant vapor pressure at 110°C. The value
for ‘‘volatile matter’’ includes organic compounds driven off or pyrolyzed but may also include water
driven off from hydroxides or hydrates. ‘‘Fixed carbon’’ includes the weight of carbon left behind as a
char, but the value may be reduced by weight gain in the crucible from oxidation of metals.

C. Ultimate Analysis
For fuels the term ultimate analysis refers to an analysis routine that reports moisture (loss in weight, for
solid fuels at 105°C), combined water (equivalent to the oxygen), and the content of several elements.
Carbon, available or net hydrogen (hydrogen other than in moisture and combined water), oxygen, total
sulfur, nitrogen, and ‘‘ash’’ are always reported. Oxygen is generally determined by difference, and thus
both the value for ‘‘percent oxygen’’ and the ‘‘combined water’’ value may be in error.
In calculations relating to waste disposal systems, it may be appropriate to request analyses for chlorine
and for environmentally significant elements (such as arsenic, beryllium, cadmium, chromium, nickel,
and lead) and for other elements that could influence the combustion process or would be important to air
pollution or water pollution assessments. In general, several different samples are used to generate the
total analysis report.

Fig F06. Tanner's ternary diagram. Waste within the shaded area can be combusted without auxiliary fuel

8
ChimH409 Nov2021
 D. Thermochemical Analysis and Thermogravimetry
The heat of combustion of wastes is, clearly, important information in incineration system analysis.
However, one must be cautious in accepting even the mean of a series of laboratory values (noting that a
bomb calorimeter uses only about 1 g of sample) when there are problems in obtaining a representative
sample. For both municipal and industrial incineration systems, the analyst must also recognize that
within the hardware lifetime waste heat content will almost inevitably change: year to year, season to
season, and, even day to day. This strongly suggests the importance of evaluating the impact of waste
variation on system temperatures, energy recovery, etc.

On the basis of aforementioned balances, the amount and identity of the incineration products can now be
derived.
The proximate analysis splits the waste to be fired up into three parts: (1) moisture W, (2) ash A, and (3)
combustibles C. The first reports to the flue gas, the second forms the residues, whereas the third is
converted into combustion products also reporting to the flue gas. Starting from W% + A% + C% = 100,
Tanner represented waste composition in a triangular diagram, in which a zone of auto-combustible MSW
is identified (Fig. F06).

Moist refuse is difficult to ignite. Ash content confines the reduction in weight achievable and determines
the burden of residue extraction; important is the composition of ash and its softening and melting
behavior at high temperature.

Volatile matter will lead to flaming combustion and fixed carbon to glowing combustion, each of these
two modes showing their specific demands.

Data from these analyses will allow us to establish the necessary material balances (Fig F07), as well as
estimating a sufficiently representative heating value.

Table F03. Typical analysis results found for a municipal solid waste

Based on a MSW ultimate analysis such as the one given in Table F03, experimental equations are
available to predict the lower heating value. The equation of Schwanecke (1976) is widely used:

The percentages must have the same reference as the LHV. For instance in order to obtain a LHV
representing the total waste amount, the percentages given must also refer to the total waste amount,
including moisture and ash. Another possible reference is the ash and water free heating value (as we
have used in Table F02): it just needs to be consistent on both sides of the equation.
9
ChimH409 Nov2021
1.4 WtE in practice

1.4.1. A rapid look at the energy recovery from waste

It is good to have an overall view of the WtE plant thermodynamically. Typical operational data are those
reported in the Best Available Technology document on waste incineration by the EU IPPC directive. The
value of the LHV (lower heating value) of the waste is taken as 10.5 MJ/kg (BREF 2019). As a result of
the incineration of one Mg (ton) of waste, the sensible heat contained in the flue gas can largely (thermal
efficiency ηtherm typically 75–85%) be recovered in waste heat boilers (under various technical
configurations that we do not detail here, see BREF 2019). The steam, once having reached 40
bar / 400°C, is fed into a steam turbine linked to a generator (Fig.F02) producing electrical power. Per ton
of waste, a well-designed professionally-operated WtE plant produces 650 kWh of electricity, which
corresponds to a gross energy efficiency of 22%. With an in-plant consumption of 150 kWh per ton of
waste, this typically results in an electricity production of 500 kWh (net efficiency of 17%) which can be
exported -at market price- to the general electricity grid.
As illustrated in Fig.F02, the overall thermodynamic performance (and use of resources) by the WtE plant
can be increased further by delivering district heating under the form of hot water. The gain of
performance is marginally altered by the loss in electricity production (Murer et al 2009). The feed
temperature sent to nearby buildings and infrastructures is around 90°C whereas the return temperature
(after warm water use) is around 45°C. Potentially, per ton of waste the heat which can be delivered in
this way amounts to 800 kWh. We say potentially because the market for that is often found to be
precarious. One of the problems is that the user (i.e. the client) should be geographically close enough to
the WtE to avoid all costs linked to overstretched conduits and excessive heat losses along the way.

Most of the time, when sizing WtE installations, higher capacity is achieved by providing parallel lines of
identical capacity and make. WtE work continuously all year round 24/7. Spreading capacity over several
parallel independent combustor lines allows more flexibility, in case of shutdown of one train or of
variable supply of waste.

Mass balances directly relate input streams to output streams. One Mg (ton) of municipal solid waste
(MSW) requires 6.5–7.8 Mg (5,000–6,000 Nm3) of combustion air. Typically, a mass-fired WtE of the
reciprocal-grate type (as described in detail in Section 1.4.3.3) generates:
• 250–350 kg of bottom ash,
• 5–15 kg of boiler slag,
• 20–40 kg of fly ash,
• 5–15 kg of neutralization salts,
• 7–8.6 Mg of flue gas.

Fig. F07. Mass balance is always the prime methodological step in the consideration of any environmental
technology problem.

10
ChimH409 Nov2021
1.4.2. Basics of engineered/controlled combustion

Before going into the pollution aspects, some combustion concepts should be clear in our mind.
Incinerators are thermal units: liberating more combustion heat also requires supplemental combustion
air. A simple rule of thumb states that this amount is directly proportional to the higher heating value,
whatever the waste fired. Combustion air is supplied to the furnace with several purposes: primary air
activates the fire bringing oxygen to the reaction surroundings, whereas secondary air (also termed over-
fire air) is injected at high speed (typically 100 m/s) to induce mixing and turbulence, as far as its
momentum reaches. Increasing primary airflow accelerates combustion until a point at which higher
cooling supersedes this stimulating effect. Air is often used for cooling the mechanical grates but water-
cooled grates are also in use.

In addition to the delivery of sufficient air, incineration requires suitable levels of the three T’s, i.e.,
Temperature, residence Time, and Turbulence. Turbulence is required to sustain the required macro- and
micro-scale mixing to bring together combustibles and air oxygen. Conditions during incineration vary
according to the technology employed and the characteristics of the waste fired. Some combustors feature
active heat and mass transfer so that combustion takes place much faster, e.g. in fluidized bed burning.
These require, however, size-reduced waste, i.e., preliminary shredding and grading, so that the residence
time provided allows either complete burnout even of the largest particles or their recycling after
separation.
Temperature ranges from as low as about 750°C (bed temperature of fluidized bed combustion) to more
than 1,200°C (destruction of hazardous waste, such as PCBs).

Residence time at high temperature is only few seconds (generally 2–3 s) for flue gas. Solid waste and its
combustion residue have a much longer residence time, from about a minute in fluidized bed combustion
(time required to dry, heat and burnout the ash) to typically 50-60 minutes on a mechanical grate; yet,
much depends on the time required for drying and heating (Fig. F08).

Fig. F08. Thermal treatment of solid waste: by-default description of what should happen. Solid fuels first
dry, and then thermally decompose while heating, with evolving volatiles sustaining flaming combustion
and the charring residue much slower glowing combustion (see photograph in Section 1.4.3.3).
Converting fuel or waste into volatiles and fixed carbon is an essential step (pyrolysis) in their
combustion. In practice, these steps proceed partly in parallel, rather than in the strict sequence suggested
by the sketch.

11
ChimH409 Nov2021
Some codes prescribe minimum values for temperature and time (e.g., 850°C for 2 seconds), or they limit
the amount of products of incomplete combustion in flue gas (carbon monoxide, CO; total organic carbon,
TOC) and carbon in residues. Combustion is never entirely complete, even though – thermodynamically
– equilibrium approach could come close to unity. In practice, when most combustibles are burned, the
rate of heat generation drops, temperature falls and combustion slows down and eventually stops. The
reason for further completing combustion is strictly environmental: Products of incomplete combustion
(PICs) are major atmospheric pollutants and responsible for reduced visibility, photochemical smog, as
well as soot or black carbon formation.

Fig. F09. Auxiliary burners

The combustion chamber is provided with auxiliary burners (Fig. F09) designed to: raise the temperature
in the combustion chamber prior to starting up from cold and during shutting down of the plant; ignite the
refuse at the start-up of the plant following shut down periods; and comply with the IED requirements for
reducing flue gas emissions, i.e. whenever the temperature falls below 850°C a standby burner will
operate automatically to maintain the temperature whilst there is waste on the grate.

Flue Gas Composition

Flue gas composition is determined by several factors of influence. The most important one is waste
composition: all entering elements will also leave the plant (Fig. F07), whether as flue gas or as solid
residue. Mass balances, together with waste composition data, allow estimating the flue gas and the
residue composition, even though some assumptions are needed regarding the distribution of the relevant
elements over the various output streams. A second factor is the technology used: mass burning of MSW
yields much more bottom-ash (typically 25–30 wt.% of MSW) than fly ash (2–3 wt.% of MSW).
Fluidised-bed incineration of the same MSW will turn this relation in favour of fly ash, which may reach,
e.g., 10–12 wt.% of MSW.

One of the methods presented in part 2 of the booklet, with Fig.S01 giving the separation efficiencies
attainable in practice, is selected to separate fly ash particles from the gas admitted to the stack. Fly ash is
considered hazardous, because it accumulates volatilized heavy metals (e.g., Hg, Cd, Pb, and Zn), as well
as PICs, some of which are semi-volatile, such as polycyclic aromatic hydrocarbons (PAHs) and
dioxins i.e., polychlorinated dibenzo-p-dioxins (PCDD) and dibenzofurans (PCDF), listed and targeted
for removal and destruction by the Stockholm Convention on principal organic pollutants (POPs).
12
ChimH409 Nov2021
Thermal processes have been applied to detoxify such residues (Sakai & Hiraoka 2000), yet this treatment
is expensive.

1.4.3. Types of burners and mode of operation

1.4.3.1. Multiple Hearth Furnaces (MHF)

Waste which is too fluid or pasty cannot be burned on a mechanical grate because most of it would pass
through the grate openings. One of the option to incinerate sewage sludge (under a slurry form) is to use a
MHF, i.e. a vertical incinerator through which the fluid waste is progressively descending under a well-
defined and controllable zonation of events (Fig. F10).

Fig. F10. The MHF is a vertical cylindrical construction, composed of a number of circular hearths
mounted one above the other. Each hearth contains an air-cooled rabble arm, driven from a common
central shaft. Blades, fitted to the slowly rotating rabble arm move the material forward, either toward the
center or toward the periphery, until it passes over a discharge aperture and falls onto the lower hearth.
The retention time of the charge is varied by changing the speed of rotation of the rabble arms.

13
ChimH409 Nov2021
Multiple hearth furnaces consist of a series of superimposed hearths, solids being fed on top and
descending stepwise by gravity, after describing a spiraling movement on each hearth, starting at the
discharge point of the higher hearth and ending at that to the lower hearth (Fig. F10). Gases generally
mount, in countercurrent to the movement of solids, aided by buoyancy.

The feed material is charged onto the upper hearth and slowly makes its way down, falling from one
hearth to the next, while it is progressively dried, heated, ignited, combusted, and finally cooled by the
combustion air as depicted in Fig. F11. The combustive air is introduced in part or all at the bottom of the
furnace, preheated by the ash on the lower hearth(s), and partly consumed on successive combustion
hearths. There is much contact surface with air and this surface is periodically renewed by the passing
rabble arms with attached plates, plowing through the material.

The resulting flue gas is cooled by the incoming feed and leaves the unit toward possible post-combustion
and cleaning.

Fig. F11. Vertical distribution of temperature inside a MHF

Auxiliary burners are used for preheating and adapting and controlling the temperature profile at what is
exactly needed as a function of the waste composition. In fact the hearths may be individually heated and
the number, temperature, rotation rate and size of each hearth determines the residence time and
combustion conditions allowing to achieve complete calcination. The atmosphere is controlled by
balanced introduction of air, or recirculation.

MHF are also very suitable to destroy/incinerate industrial waste slurries, for instance the spent liquors to
be discarded in the pharmaceutical industry.

14
ChimH409 Nov2021
In brief, MHFs are a proven technology that allows a flexible adaptation of operating conditions on each
hearth. This versatility is less available in rotary kilns. However, because of their configuration, they are
relatively expensive comparatively to the other incineration configurations. They also take long times to
preheat and shutdown.

Fig. F12. Multiple hearth furnace in its industrial setting. The individual hearths are lined with refractory
brick, and the rabble arms are typically a force cooled metal alloy. The entire structure is enclosed in a
cylindrical brick-lined steel shell.

15
ChimH409 Nov2021
1.4.3.2. Fluidized-bed burners

Fig.F13 Stationary bubbling fluidized-bed burner

Smooth fluidization is obtained only at a gas velocity of three to four times umf. With rising gas
velocity, entrainment of fines becomes more important, depending on their terminal falling
velocity. The entrainment of bed particles specifies an upper limit of gas velocity.
Depending on particle size (often 0.3–0.8 mm), gas velocities typically range between 0.3
and 5 m/s.

In practice
A fluidized bed incinerator requires steady, size-reduced feed so the waste needs to be
pretreated first. RDF pellets are naturally ideal. Many installations successfully burn
wastewater sludges having reached a sufficient level of dehydration (min. 30% wt). As shown
in Fig. F13, the stationary bubbling fluidised-bed burner consists (starting below and moving
upward) of an empty plenum chamber; a distributor supporting the granular bed; the bed of
granular material to be fluidized; a freeboard zone ensuring disengagement of entrained
particles and post-combustion, and adapted feeding (admission of material to be burned); start-
up and auxiliary heating burners. The bed material consists of a graded fraction of clean heat-
resistant material, such as sand, more seldom alumina, limestone, dolomite, or ash. Pollutants,
such as SO2, can be removed in situ by simply feeding limestone or dolomite into the bed.
Secondary air is injected into the freeboard zone to complete the combustion. It also provides
the swirl required for mixing.

The great merit of fluidized bed systems is that they intrinsically provide for intensive mass
and heat transfer between the solid particles to be incinerated and the flow of gas, thus
enabling highest efficiency and complete chemical reactions. Also, the three key parameters
turbulence, temperature, and time can be technically controlled in fluidized bed systems within
narrow ranges. Because of precise control of mass flow (air flows as well as waste supply) the
maximum temperature can be kept within a very narrow range of around 900°C, just above the
minimum temperature necessary for complete gasification and combustion. Therefore,
combustion in the fluidized bed systems allows for very low formation of NOx (nitrogen oxide)
due to the absence of thermal formation by oxidation of nitrogen gas supplied with air for
combustion. Such undesirable reactions do not occur at temperatures below 1000°C.

The combustion process in a fluidized bed and the “freeboard” above it are complete. Thus,
there are very low concentrations of unburned organic substances in the solid residues and in
the flue gas. Due to the precise process control, the excess air ratio (i.e., the excess oxygen
content in the flue gas) after complete combustion can be kept very low including the option
for controlled recirculation of cleaned flue gas extracted after the filter system (necessary
removal of pollutants) back into the fluidized bed system. Reduced quantity of excess air
minimizes the overall heat losses and atmospheric emissions of the system.

So they have many advantages compared to other burners. Fluidized (bubbling) bed
combustors are highly adaptable, allowing to burn (or gasify, if air supply is sub-
stoichiometric) solid, pasty, melting, liquid, sludge, slurried, and gaseous waste,
simultaneously and at unusually low temperatures. The bed material provides a thermal
flywheel that effectively copes with short-term fluctuations in feed rates and quality.
The principal limitations are the relatively important pressure drop, as well as bed
agglomeration, in the presence of tacky ash or salts. MSW need to be pretreated (reduction in
size). In case of very large WtE installations fluidized-bed burners are not necessarily the best
option to select: they are considered by practitioners too expensive in this case compared to
mass-fired grate combustors.
Fig. F14. Fluidised-bed burner with internal gas recirculation (© Integral 2014)
1.4.3.3. Grate combustion of “as-received” MSW

In a mass burn WtE, the MSW collection trucks discharge their load into the waste bunker where it is
mixed by an overhead crane and is temporarily stored for several days. The primary combustive air is
extracted from this waste bunker area. Any methane which would form from a beginning of
fermentation of the wastes while in the pit is thus extracted as well (and incinerated). The strategy has
also the advantage of maintaining a slight negative pressure in the bunker hall. This prevents the risks
of odour and dust emissions from the bunker area (air is only coming in and not going out).

The filling hopper is used as a continuous waste supplier. It is filled in batches by the overhead crane,
keeping the combustor active 24 hours a day, all year round (with the exception of planned stoppages
for maintenance). As indicated in Fig. F15, the pile of packed waste temporarily accumulated in the
feed chute plays the role of air seal. Then, through the action of the ram feeder, the direction of
movement of the waste is changed from vertical in the feed chute to horizontal. The waste, compacted
in the feed chute, is loosened in this process and pushed onto the grate in a fully controlled way (see
below).
Different types of grate systems exist on the market and differ by the way the waste is conveyed
through the different zones in the combustion chamber. Each grate system has to fulfil requirements
regarding primary air feeding, conveying velocity and raking, as well as mixing of the waste.

The air required for combustion is fed as "under-fire" or primary air from below the grate, passing
through the grate bars into the fuel bed (Fig. F08). The air can be preheated to pre-dry the waste if
necessary. The grate type illustrated and explained in Fig. F20 & F21 is a reverse-acting stoker grate,
designed and refined over the years by leading company Martin GmbH. It consists of alternate fixed
and reciprocating grate bar steps. The reciprocating steps move slowly upwards against the downward
flow of refuse to achieve reverse overall agitation. The grate is subdivided into several zones which are
individually supplied with primary combustive air. The exact air flow to each zone is thus controlled to
suit combustion conditions. Reverse overall agitation assures uniformity of the refuse layer and stable
combustion condition, for stability of combustion temperature and of steam generation in the boiler.

It is important to note that, comparatively to conventional fuels such as coal, municipal waste has a
high volatile content. This means that the volatile gases are driven off and burn above the grate with
only a small part of the actual incineration taking place on or near the grate. Secondary air (also termed
over-fire air) is injected from the sides at high speeds (#8 in Fig. 15), guaranteeing turbulent mixing
and an excellent burnout of the flue gases, thereby eliminating volatile organic compounds. TOC
sensors and CO sensors are used downstream to monitor the flue gases and check continuously that the
full combustion is complete at all time.

The average residence time of solids on the grate is 50-60 minutes. Upon completion of the
combustion, the remaining inert heavy ashes pass over the clinker roller and fall into the bottom ash
extractor (Fig. F22). The discharger is filled with water that is both quenching and cooling the bottom
ash to approximately 85°C, but is also providing a seal against air intake. The water used to fill the
discharger is recycled water (coming notably from the flue gas treatments). From there the coarse
combustion residues are taken to a bunker or storage area for metals recovery (Fig. 23-26) and for
further use.

In recent WtE installations an infrared camera mounted at the boiler ceiling delivers instantaneous
(two-dimensional) thermographic images of the surface temperature of the fuel bed. It allows to
accurately control the under-fire air distribution as a function of the position of the main combustion
zone, the ram-feeder speed and grate movement. An adequate spatial positioning of the combustion
zone is key to have maximum transfer of heat to the boiler. Nowadays it is also preferred to use as few
external air as possible, in order to limit the flow rates of exhaust gases which need to be depolluted
(refer, for instance, to Equation 7 on page Suppl.10). There is thus increasing use of recirculation gas
and of disconnect between turbulent mixing and oxygen supply (Martin et al 2015).
Fig. F15
 Fig. F16. Inspection of the interior of the oven (Brussels-Energy)

Fig. F17. This is what we see through the observation window into the combustion chamber. The
visualisation allows to grasp the difference between flaming combustion occuring in the image background
and glowing combustion of the much smaller objects (on the left-side) in the foreground. Glowing
combustion can be sustained at a much lower concentration of oxygen than the 10% required for flaming
combustion.
Fig. F18. Garbage trucks tip waste directly into the waste bunker, which is made of impervious concrete.
Storage bridges the gaps between the schedules of collection rounds and continuous firing (Brussels-
Energy).
Fig. F19 Incinerator feed should always be made as homogeneous and constant as possible, e.g., by
mixing, blending, and for municipal solid waste ageing, to provoke moisture transfer and to account for a
wide difference of flammability between easily igniting plastics on the one hand and moist biogenic
(vegetal) waste on the other. A traveling crane allows mixing waste of different origins, stacking waste
against the bunker wall, and then feeding it into the hopper on top of the load shaft. Loading the hopper
can be semi-automatic. Some cranes are fitted with a weight sensing device.
It is also the responsability of the crane operator to detect (and put aside) oversized concrete or metal
parts or other objects which could endanger the furnace operation, such as gas canisters, ammunition etc.
Sometimes at the Brussels-Energy plant, tall tree trunks find their way to the waste bunker as well.
Fig. F20. Inspection of a Martin reverse-acting grate burner before startup. The angle of the grate is 26°
relating to the horizontal.

Fig. F21 The logic of the reverse-acting grate is that solid waste will ignite more easily when an already
existing glowing mass is pushed back underneath it (©Martin GmbH 2013). This is the role of the
moving grate. Movement of the solid waste is from left to right (through the action of gravity over the
declivity angle 26°).
Fig. F22. When having reached the clinker roller, the hot, burned-out bottom ash drops from the grate
end to a water bath in the ash discharger below the grate, where complete quenching occurs.

Fig. F23. Temporary storage of the bottom ash. In this area, bottom ash, ferrous metals and oversized
objects are stored separately (Brussels-Energy).
Fig. F24. A magnetic separator drum is first used to remove the ferromagnetic materials out of the coarse
combustion residues. Screens are also used to remove bulky objects (those containing unburned organics can
be fed back into the furnace).

As we have seen in the ChimH302 class, it is especially important to recycle aluminum metal because the
production of Al from bauxite (the principal ore of aluminum) requires a lot of energy.
After ferromagnetic separation, the non-ferrous metals (Al, Cu, Zn...) are recovered by eddy current
separation. The separation mechanism works by exposing conductive, nonferrous metallic particles to a
time-varying magnetic field, which in turn gives rise to electrical currents throughout their volumes. The
relative motion between the current and the magnetic field gives rise to a force (the Lorentz force), which
subsequently deflects the Al / Cu / Zn particles away from the non-conductive matters. The separation effect
thus results from the combination of electric and magnetic force on a particle due to electromagnetic fields.

Fig. F25a & F25b. Principle and operation of the eddy current separator to recover non-ferrous metals.

In practice, the time-varying magnetic field is generally achieved by rotating a cylindrical array of
permanent magnets with alternating polarity (Fig. F25a). If the magnetic array is both strong enough and
fast enough, metallic particles will actually lift off the feed belt (Fig. F25b) and hurl over the mechanical
splitter nearby. The process works well at extracting particles of aluminum, copper, brass, and zinc out of the
coarse combustion residues (ISWA 2015).

Afterwards the bottom ash (without any remaining metal parts) will be weathered and graded, and then
ultimately recycled as an aggregate material as construction material; it is sufficiently stable and inert to be
used in sub-road construction and embankments, for instance. It is also used to establish massive noise-
barriers along highways (ISWA 2015).
26
Schmelzer et al (1995, 1996) showed that it was possible to further increase the
efficiency of metal recovery from incineration bottom ash. The dry treatment they
suggested is shown in Fig. F26.

Fig. F26. Enhanced process for metal recovery of incineration bottom ash residue.

A wet process was developed subsequently:

Schmelzer, G (1995). Separation of metals from waste incineration residue by application

of mineral processing. In: Proceedings of the XIX International Mineral Processing
Congress, vol. 4. pp. 137-140.

Schmelzer, G., Wolf, S., Hoberg, H. (1996). New wet treatment for components of
incineration slag. Aufbereitungs-Technik/Mineral Processing 37 (4), 149–157.
27
 Part II. Air pollution control engineering

2. Air pollution control engineering: Applied flue-gas cleaning and control systems

As has been seen in the Environmental Pollution class Chim-H302, typical air contaminants
generated by human activities are volatile organic compounds and other ozone precursors
(NOx and CO), acid rain precursors (SOx and NOx), numerous hazardous materials
(carcinogenic organic and heavy metals), and total suspended particulates (TSP). Particles of
particular concern are, firstly, those less than 10 μm in diameter (PM10) which reduce
visibility and can be associated with numerous chemical contaminants, and, secondly, those
less than 2.5 μm in diameter (PM2.5) which are able to affect human health the most adversely
because of their very small size. Nowadays PM10 and PM2.5 are the particles which fall under
more and more stringent regulations in the European Union and abroad.
The need to provide effective removal of both particles and gaseous pollutants introduces
numerous constraints in the determination of the best technology for air pollution control
following industrial combustion processes. Let's start with particles.
For such fine particulate matter, conventional separation techniques such as gravity settlers or
aerocyclones are not sufficient: other, more elaborate technologies have to be foreseen and
implemented.
Figure S01 gives the range of separation efficiencies (Cin-Cout)/Cin which can reasonably be
attained using these enhanced techniques. The values of efficiencies are primarily determined
by the median particle size (and particle size distribution) of the dust contained in the flue
gases. We are now going to present these techniques with the necessary detail, starting with
the separation principle on which they are based.

Fig. S01. Separation efficiencies of the techniques available to eliminate dust particles from
flue gases

Suppl.1
Electrostatic precipitators as key contributors to air pollution control
and removal of fine particulates and aerosols from dirty gas flow
An electrostatic precipitator (ESP) is a separation device like a gravity settler or an
aerocyclone, but it uses electrostatic forces to move particles entrained within a waste gas stream
to the walls of the collector plates. As will be detailed below, the precipitation process involves
1) charging particles by means of ions produced in a corona discharge, 2) separating the charged
particles from the gas stream in an imposed electric field, 3) collecting the particles on a grounded
surface, and 4) removing the collected particles by either knocking them off or washing then off
with water or another absorbing liquid medium. The technique is very useful for the removal of
fine dusts (between 1 and 10 µm, but mostly less than 5 µm) from different kind of waste gases
with very high efficiency (approaching 99.9 %).
Electrostatic precipitators are unique among gas cleaning equipment in that the forces
separating the particulates from the gas stream are applied directly to the particles themselves
rather than to the overall gas flux to be handled. Hence the energy required to effect the
separation is considerably less than for other types of gas cleaning apparatus. Gas pressure drops
through an electrostatic precipitator may be of the order of 3 mbar or less, as compared with
pressure drops of up to 25–250 mbar for scrubbers and baghouses (fabric filters). This
fundamental advantage of electrostatic precipitation results in its widespread use in applications
where large gas volumes are to be handled and where high efficiencies (refer to Fig. S01) are
required for collection of small particles and aerosols. It will be shown later, however, that the
particle capture efficiency of an ESP is highly sensitive to changes in the values of the incoming
gas flow rate and this has to be taken into account both in terms of initial design and operation of
the separation device. Another downside of dry electrofilters is that they occupy vast spaces.
A wet precipitator is smaller but has the issue of the disposal of the waste slurry.

Fig. S02

Suppl.2
 Typical configuration

The wire/plate configuration consists of an array of high-voltage discharge wires located

midway between grounded, parallel-plate collecting electrodes. A schematic of the wire/plate dry
ESP is shown in Figure S03. A large ESP for a solid waste incinerator or a power plant would
consist of many of these flow channels arranged in parallel. Most large-scale dry ESPs used in the
industry have negatively-wired emitting electrodes (though positive wiring exists). This is the
reason why we focus on them here.

Figure S03. Single-stage high-voltage wire-plate ESP: typical configuration and nomenclature.

Wires are either hung from the top of the ESP (in the case illustrated they are weighted) or
supported in a frame. Wires are made from noncorrosive materials like tungsten or alloys of steel
and copper.

How the separation works

Fig. S04 can help to understand how a wire/plate dry electrofilter is working. Electrostatic
precipitation requires maintenance of a high potential difference, typically 40–60 kV DC. The
collecting electrode is grounded and the discharge electrode is brought to, say, - 40kV. Since the
radius of curvature of the wire is small compared to the distance from the wires to the plates, the
electric field strength E (expressed in Volt/meter) is very high near the wire.

Suppl.3
The applied voltage is increased until it produces a discharge (corona) which can be seen as a
luminous blue glow in the immediate proximity of the discharge electrode. This is where the so-
called avalanche multiplication (of the attack of the electrons on the ambient gas molecules)
occurs.

Figure S04. Electrical charging and migration of particles toward the vertical collection surface
(simplified representation, see text). Ideal separation conditions are met when the electrical
resistivity ρ of the dust particles remains in the intermediate range 107-1010 Ω.cm (i.e. neither too
high, nor too low ρ values).

Within the negative corona discharge, electrons are accelerated by the very strong electrical field
and strike and ionize gas molecules. Each collision of a fast-moving electron and a gas molecule
generates an additional electron and a positively charged gas ion. These additional free electrons
are the ones responsible for the avalanching effect which is essential for the local atmosphere
energization.
It has to be noted first that the positive gas ions generated in the ionization process move back
toward the discharge electrode. Some of these positive gas ions will deposit on particles inside
the corona and charge them positively. These positively charged particles deposit on the negative
discharge electrodes, requiring them to be cleaned periodically.

Suppl.4
A bit farther away from the discharge electrode, where the electrical field strength (E) is lower,
electrons released in the corona discharges are captured by gas molecules. These negatively-
charged gas ions move rapidly toward the grounded collection plates. Some of these gas ions are
captured by particles, charging them negatively as shown in Fig. S04. Particles thus obtain a
negative charge (q). Note that the capacity of individual dust particles to accumulate these
negative electric charges is directly proportional to their surface area, meaning that large particles
will be more strongly affected by the applied electrical field. As they carry charge q and are
exposed to the electrical field E, the path of the particles is now modified by the electric force
which is equal to qE. The net effect of qE is to force the particles towards the grounded vertical
collecting plate (Fig. S04). When the particles reach the collection plates, most of their negative
charge is neutralized. Actually a portion of the charge is kept (see discussion below) which
contributes to the intermolecular cohesive and adhesive forces that will hold the particles stuck to
the collection surfaces. A cake-like layer of ash is then able to accumulate on the vertical plates,
building up to a thickness that would preferably be restrained to (about) 2 cm maximum. If the
dust layer becomes too thick, there is a risk that the accumulated layer would act as an insulator,
reducing the flow of the electric field lines.

Fig. S05. Side-mounting rapping hammers. The rapping is carried out at regular programmed
intervals (the frequency of which can be modulated as a function of actual operating conditions).

The next step is to dislodge the trapped particles from the collector surface. Periodically, the large
plates are mechanically vibrated or rapped with automated hammers (either top- or side-mounted
systems as illustrated in Fig. S05), causing the particles to drop into the hoppers below the plates.
They have then to be conveyed at regular intervals out of the hoppers (Fig. S02). They will be
stored in the ash silos for a certain amount of weeks, before being trucked and ultimately
disposed.
In some industries the recovered ash has a composition which is sufficiently close to the initial
raw material so that it can be considered, after some processing, to be reinjected in the process.
This is, naturally, not the case for the fly ashes generated by MSWI considering the high
compositional heterogeneity of both flue gases and solid wastes processed.

Suppl.5
Fig. S06. Rapping hammers used to get rid of the particles which would deposit on
the discharge electrodes. Due to the strongly negative voltage, an insulator in this
case is necessary.

The particle electro-migration velocity w is the speed at which a particle, once charged, migrates
toward the grounded collection electrode. Knowledge of its value is obviously important in the
design of an ESP (see Design Parameters section).

Suppl.6
In some instances the electrical resistivity of the dust particles needs to be manipulated

When they are exposed to an electric field, particles having differing composition will display
differing resistance to transferring charge (both accepting and giving up charges). This will also
change according to the flue gas operating conditions (temperature, but also moisture).
The relevant property of particles in this context is their resistivity ρ, expressed in units of Ω.cm
at a certain reference temperature. In an ESP, where particle charging and discharging are key
functions, resistivity is an important factor that significantly affects collection efficiency. While
resistivity is an important phenomenon in the inter-electrode region where most particle charging
takes place, it has a particularly important effect on the dust layer at the collection electrode
where discharging occurs. On one hand, particles that exhibit high resistivity are difficult to
charge. But once charged, they do not readily give up their acquired charge on arrival at the
collection electrode. On the other hand, particles with low resistivity easily become charged and
readily release their charge to the grounded collection plate.
These are the reasons why, for efficient work of an ESP, the resistivity of the particles should be in
the so-called intermediate range, which is between 107 and 1010 Ω.cm. Operation at a lower or higher
resistivity leads to a decrease in removal efficiency.

To improve efficiency it is possible to perform chemical conditioning of the flue gases (with
moisture, by addition of SO3 gas, sodium or ammonium salts) or to work at a lower operating
temperature (with caution because of the possible condensation and corrosion).

The risk of de novo synthesis of toxic organo-halogenated compounds

With the principle “3T+E”, which means Temperature higher than 850 °C, residence Time longer
than 2 s, Turbulence, and Excessive air in the furnace, the organic pollutants initially present in
the solid waste are mostly destroyed. However, it is important to be aware that there is a
downside to the accumulation of fly ash on the collecting plates of the electrostatic precipitator,
in direct connection with the physico-chemical conditions characterizing that specific
environment. The preoccupation is a risk of de novo formation of polychlorinated dibenzo-p-
dioxins (PCDD) and dibenzofurans (PCDF). Firstly, there is always some residual carbon char
coming from the burner (a product of incomplete combustion): if they are present these organic
fragments can act as organic precursors to the dioxin/dibenzofuran molecule. Secondly, the MSW
waste components provide a source of chlorine, and of metals which ultimately find their way
into the fly ash and thus into the fly ash cake allowed to accumulate on the ESP collection plates.
The unfortunate consequence, as documented in field and lab studies, is a de novo formation of
PCDD/F when the organic precursors are adsorbed on the surface of the fly ash cake and undergo
heterogeneous reactions catalysed by metals (especially copper). It was shown that the highest
production rate of PCDD/F was occurring at a temperature of around 300°C.

A first strategy to deal with this is to a) reduce releases of unburnt carbon by better compliance
with the "3T+E" rule ; b) reduce residence time in the ESP; and c) maintain the temperature at the
ESP collecting plates below 250°C. The second strategy is to capture the dioxins in a subsequent
pollution control unit: fine-grained activated carbon, in particular, is an adsorption material which
displays high affinity for dioxin-like molecules.

Suppl.7
Fig. S08 The generation of the corona discharge can be enhanced by giving special geometry
(and curvature radius) to the emitting electrode. Rigid discharge electrodes designed as these
ones (© Babcock-Wilcox Inc.) contribute to the efficient charging and separation of fly ashes
having resistivity properties outside of the recommended intermediate range.

Another development in the industry is an energy management system in which an opacity

meter in the stack is used to monitor the particulate output. The particulate output is held
constant by raising or lowering electric power as needed.

In the case of a waste-to-energy plant, the selection of an electrofilter (to eliminate fine
particulates) is often fostered by the fact that electric energy is produced locally anyway.

Suppl.8
ƒ Gas and particles move in the x-direction at constant velocity u.
ƒ The particles are uniformly distributed in the y- and z-directions at every x direction.
ƒ The charging and collection electric fields are constant and uniform, i.e. the particles attain
an electric drift velocity w in the y-direction very quickly. The determination of w will be
discussed in a later section.
ƒ Once particles reach the collection plate, they stay there.
L

D/2
u

wire ∆x

Figure S09. Top view of a one-half of flow channel of a wire-plate precipitator. Particle is shown
moving at velocity u in the flow stream direction x and with electric drift velocity w in the
transverse direction y. The height of the precipitator in the z-direction is H.

With these assumptions, we can write a mass balance on an element ∆x.

D D
(u H)Cx − (u H) ⋅ C x+ ∆x = w ⋅ Cx +∆x / 2H ⋅ ∆x . (3)
2 2

This equation simply states that the difference between the mass of particles flowing into and out
of the slice must equal the flux of particles in the y direction times the area normal to the flux.
The rhs of the equation is the mass removed. Dividing by ∆x and taking the limit as ∆x goes to
zero, we obtain

− uHD dC
= wHC (4)
2 dx

Separating variables and integrating, yields

C x =L dC L w2H
∫ = −∫ dx , (5)
C x =0 C 0 uHD

CL  − wA 
= exp  , (6)
C0  Q 

Suppl.9
where A is the plate area in one channel (2HL) and Q is the volumetric gas flow (uHD) evaluated
at the temperature and pressure inside the precipitator. Thus the collection efficiency is
expressed as

CL
= 1 − e(
−wA Q )
η = 1− (7)
C0

This equation shows the relationship between collecting plate area and volumetric flow rate. The
specific collection area is defined as A/Q. Our next step will be to estimate w.

Electric Migration Velocity w

Particles that acquire an electrical charge from the ions created in the corona discharge,
experience an electrical force oriented along the electric field lines. This force is given by
& &
Fe = qE (8)

where q the charge on the particle and the electric field is that in the collection section of the
precipitator. The particle also experiences a drag force due to movement of the particle relative
to the gas (Fig.S07). The gas exerts a drag force on the particle. The drag force is written as

& 1
FD = C D A p ρ f w 2p (9)
2

Note that the velocity w in this expression is the relative

velocity of the particle to the gas or the electric drift velocity.

The drag coefficient is a function of the particle shape and of the flow regime as defined by the
Reynolds number. If we assume that the particle is in the Stokes flow regime which is valid for
spherical particles with Reynolds numbers less than one, then

24 24
CD = = (10)
Re ρ f pw
D
µ

The Stokes flow drag coefficient is sometimes corrected for very small particles less than 5
microns in diameter to take into account noncontinuum effects (Cunningham, ProcRoySocA1910).

If Fig.S07 is considered as equilibrium conditions, then the drag force must equal the electric force:

Suppl.10
Note that equation (13) is a vector equation. The electromigration velocity has both a
magnitude and a direction.
It suggests that, knowing the charge q and the value of the electric field, we should,
theoretically at least, be able to predict the electromigration velocity. We could adapt the
necessary electric power in order to achieve our particle separation aims (while trying
naturally to maintain the energy expenses at a minimum). In other words, in our effort of
designing the electrostatic precipitator, the value of coming from equation (13) could now
be introduced in equation (7).

In practice, however, rather than a theoretical prediction, it is preferred to infer the value of
the electric drift velocity based on tests conducted experimentally on the dust to be handled.
That notably allows to include the potential beneficial effects of the chemical conditioning
aimed at manipulating the dust resistivity ρ (as discussed above). There are also (gross) design
calculations which just consider that experimental values of recuperated / imported from a
very similar installation (e.g. the exact same furnace working also on MSW) are applicable as
such. This reflects the fact that practitioners have much more confidence in equation (7),
called the Deutsch-Anderson equation, than in equation (13) which is considered
insufficiently robust in its present stage of development.

Designing a dry ESP in practice

This being said we can now come back to what equation (7) is telling us (note the double
minus sign):

• the electromigration velocity has a role but is not so easy to manipulate in day-to-day
operation: we have to consider it as an imposed input rather than a variable.

• the separation efficiency (also called the particle capture efficiency) of an dry
electrofilter will considerably increase if the collection plate has a larger area A. This
increase is exponential. For the capture of fine particles, it is therefore not uncommon
to have individual plates as high as 6 m (and 2 m width) with an inter-distance in the
range 0.15 ≤ D ≤ 0.35 m. The consequence of that is dry ESPs take an enormous
space, as illustrated in Fig.S10.

• If, during operation of the plant, there are significant changes in the volumetric flow
rate of the flue gases, this will immediately impair the particle capture efficiency. The
larger the flow rate Q increase, the less effective will be the actual particulate capture
in the ESP. Note in this respect that Q is in the denominator of the exponentiated term.
Concurrent devices, such as baghouse filters, should be preferred if there are
significant changes of Q over time. Note that aerocyclones will also perform badly if

Suppl.11
 the incoming airflow velocities are not maintained sufficiently close to their design
values (number of vortex gyres internally would not be as expected).

Fig. S10. In practice, dry ESPs are very large "empty" boxes (through which flue gases will
pass with minimal pressure losses).

Wet electrostatic precipitators

Wet electrostatic precipitators are based upon the same separation working principle as
electrostatic precipitators. They allow to obtain the gain in efficiency depicted in Fig. S01.
With this design, the precipitated dust on the collector plates is washed off using a liquid,
usually water. As a variant, another absorbing liquid may be used as a function of the actual
content of the flue gases. The liquid aspersion may be done continuously or periodically. This
technique operates satisfactorily in cases where moist or cooler flue-gas enters the
electrostatic precipitator. Comparatively to its dry counterpart, the wet separator has an
overall size significantly smaller. It is also recommended when particles to be separated are
sticky or have a very high resistivity ρ. In addition to that, the risk of reentrainment of
particles departing from the collecting surfaces (which is often a preoccupation in dry ESP,
especially for particles having a very low resistivity) does not exist thanks to the regular liquid
rinsing. The downturn of the wet electrofilter is that the rinsing liquids have to be disposed of.
[ 1, UBA 2001 ]. As a general rule in Environmental Technology we hate dealing with
residues which are heterogeneous (either in composition or physical state) because it increases
handling costs and it usually discards the recycling-recovery routes we could have considered
otherwise.

Condensation electrostatic precipitators

The condensation electrostatic precipitator is used to deposit very fine, solid, liquid, or sticky
particles and aerosols, for example in the flue-gas from hazardous waste incineration plants.
Unlike conventional wet electrostatic precipitators, the collecting surfaces of condensation
electrostatic precipitators consist of vertical plastic tubes arranged in bundles, which are
externally watercooled to force condensation of the aerosols and other very fine particles.

Suppl.12
.

The dust-containing flue-gas is first cooled down to dew-point temperature in a quench by

direct injection of water and then saturated with vapour. By cooling the gases in the collecting
pipes further down, a thin, smooth liquid layer forms on the inner surface of the tubes as a result
of condensation of the vapour. This is electrically earthed and thus serves as the passive
electrode.

Particles are deposited by the influence of the electric field between the discharge electrodes
suspended in the tube axes and the condensation layer in continuous flow. At the same time the
condensation layer also causes continuous removal of deposited particles from the deposition
area. Even water-insoluble dust and poorly wettable soot are washed off. The constantly
renewed wetting prevents dry spots and sticking, which can cause sparking (electrical
discharges between the electrodes). Avoiding sparking allows for a higher deposition voltage,
which in turn leads to an improved and consistently high deposition performance (see Figure
2.34).

Source: [ 1, UBA 2001 ]

Figure 2.34: Condensation electrostatic precipitator

2.5.3.4 Ionisation wet scrubbers

The purpose of the ionisation wet scrubber (IWS) is to remove various pollutants from the flue-
gas flow [ 1, UBA 2001 ]. The IWS combines the principles of:

 electrostatic charging of particles, electrostatic attraction and deposition of aerosols

(smaller than 5 µm);
 vertical deposition of coarse, liquid and solid particles (larger than 5 µm); and
 absorption of hazardous, corrosive and odorous gases.
Fig. S11. Wet Electro-Venturi separators at the Brussels-Energy plant
 Chapter 2

The IWS system is a combination of an electrostatic filter and a packed scrubber. It is reported
to require little energy and has a high deposition efficiency for particles in the submicron as well
as the micron range.

A high-voltage zone is installed before each packed tower stage. The function of the high-
voltage zone is to ionise the particles (dust, aerosols, submicron particles) contained in the flue-
gas. The negatively charged particles induce opposing charges on the neutral surface of the
wetted packing material and the falling water drops. As a result, they are attracted onto the
liquid surfaces and are then washed out in the packed section. This is referred to as Image/Force
attraction (IF attraction), i.e. attraction through electron shift. Hazardous, corrosive and odorous
gases are also absorbed in the same scrubber fluid and chemically combined to be discharged
with the scrubber effluent.

Another type of ionisation wet scrubber includes a venturi. The pressure changes that occur
through the venturi allow the fine particles to grow and the electrode charges them. They are
then collected by the dense layer of water droplets projected by a nozzle, serving as a collecting
electrode. [ 74, TWG 2004 ]

2.5.3.5 Bag filters

Bag filters, also called baghouse filers or fabric filters, are very widely used in waste
incineration plants. Filtration efficiencies are very high across a wide range of particle sizes. At
particle sizes below 0.1 microns, efficiencies are reduced, but the fraction of these that exist in
the flue-gas flow from waste incineration plants is relatively low. Low dust emissions are
achieved with this technology. It can also be used following an ESP and wet scrubbers.
[ 74, TWG 2004 ]

The compatibility of the filter medium with the characteristics of the flue-gas and the dust, and
the process temperature of the filter are important for effective performance. The filter medium
should have suitable properties for thermal, physical and chemical resistance (e.g. hydrolysis,
acid, alkali, oxidation). The gas flow rate determines the appropriate filtering surface, i.e.
filtering velocity.

Mechanical and thermal stress on the filter material determines service life, energy and
maintenance requirements.

In continuous operation, there is gradual loss of pressure across the filter media due to the
deposit of particles. When dry sorption systems are used, the formation of a cake on the media
helps to provide the acid removal. In general, the differential pressure across the filter is used to
monitor the need for cleaning. Periodic replacement is required when the residual lifetime is
achieved or in the case of irreversible damage (e.g. an increasing loss of pressure may be caused
by an irreversible deposit of fine dust on the filter material). Several parameters help to control
the lifetime of the bags: pressure drop drift, visual or microscopic analysis, etc. Potential leaks
in the bag filter will also be detected by the increased emissions or by some process disturbance.
[ 64, TWG 2003 ]

Bag filter systems are usually composed of several chambers where sections can be isolated one
by one for maintenance. It is common practice to design the system to operate at full capacity
when one section is out of service to facilitate on-load maintenance and reduce downtime.
However, there may still be constraints on how much repair work can be completed with the
incinerator on-line, for example the outside of the off-line section may be too hot to allow safe
personnel access due to the adjacent sections being in service.

Waste Incineration 97
Chapter 2

Source: [ 1, UBA 2001 ]

Figure 2.35: An example of a bag filter

Selection of filter bag material

The filter material is selected to be suited to the physical and chemical conditions under which it
will operate.

The key characteristics of fabrics for use in gas filtration include maximum operational
temperature and resistance to acids, alkalis and flexing (due to bag cleaning). Gas humidity can
also affect the strength and dimensional stability of the fabrics, due to hydrolysis. Several basic
fibre properties are summarised below; some may be coated or impregnated with special
chemicals (e.g. sulphur). [ 74, TWG 2004 ]

98 Waste Incineration
 Chapter 2

Table 2.15: Operational information for different bag filter materials

Maximum Resistance
Fabric temperature Physical
Acid Alkali
(C) flexibility
Cotton 80 Poor Good Very good
Polypropylene 95 Excellent Excellent Very good
Wool 100 Fair Poor Very good
Polyester 135 Good Good Very good
Nylon 205 Poor to fair Excellent Excellent
PTFE 235 Excellent Excellent Fair
Polyimide 260 Good Good Very good
Fibreglass 260 Fair to good Fair to good Fair
NB:
Not all of these materials are commonly used in incineration – see operational data below.
Source: [ 2, InfoMil 2002 ] [67, Inspec, 2004]

Increasing the temperature may lead to the melting of any plastic components in the fabric
material, and the potential for fires. High humidity in the flue-gas may cause the filter materials
to stick together and lead to shutdowns. [ 74, TWG 2004 ] PTFE covering of sheets/foils can be
used to improve the removal of such sticky salts and solid particles from the bags. Operational
improvements in semi-wet systems (see also Section 2.5.4) are reported to have been achieved
by using PTFE in a MSWI facility in Prague (Czech Republic) and in Schwandorf (Germany).

Several filtration media are reported to not be commonly used in MSWI, e.g. cotton, wool,
propylene. In MSWI, the main media are polyimide, PPS (rarely), PTFE, and fibreglass (with or
without PTFE coating). Some fibres may be combined (e.g. polyimide and PTFE for higher
resistance at high temperatures).

Chemical reactions in the absorbent media may affect the operational temperature. The quality
of the scrim is also of importance, as well as the fibre quality. [ 2, InfoMil 2002 ], [ 64, TWG
2003 ]

Special filter bags may include catalytic elements for the reduction of NOX and/or for the
destruction of PCDD/F (see Sections 4.5.4.5 and 4.5.5.4).

2.5.3.6 Cyclones and multi-cyclones

[ 64, TWG 2003 ]

Cyclones and multi-cyclones use centrifugal forces to separate dust from the gas stream. Multi-
cyclones differ from single cyclones in that they consist of many small cyclone units. The gas
flow enters the separator tangentially and leaves from a central port. Solids are forced to the
outside of the cyclone and collected at the sides for removal.
Cyclones on their own cannot achieve the required dust emission levels. They can however play
an important role, where applied as a pre-deduster before other flue-gas treatment stages, in
reducing the dust load to be finally treated. Energy requirements are generally low as there is
almost no pressure drop across the cyclone.

A major advantage of cyclones is their wide operational temperature range and robust
construction. Erosion of cyclones, particularly at the point of impingement of dirty flue-gases,
can be an issue where the flue-gas is more heavily loaded with particulate, and particularly
where bed material escapes from fluidised bed plants. Circulating fluidised beds usually
incorporate a cyclone for the removal and recirculation of the bed material to the furnace.

Waste Incineration 99
 Fig. 2.35b. Principle of the Venturi scrubber (with the absorbing liquid forced into the Venturi throat)

99b
 Chapter 2

2.5.4.1 Wet processes

Wet flue-gas cleaning processes use different types of scrubber design, for example:

 jet scrubbers;
 rotation scrubbers;
 venturi scrubbers;
 dry tower scrubbers;
 spray scrubbers;
 packed tower scrubbers.

The scrubber solution is (in the case of water-only injection) strongly acidic (typically pH 0–1)
due to acids forming in the process of deposition. HCl and HF are mainly removed in the first
stage of the wet scrubber. The effluent from the first stage is recycled many times, with little
fresh water addition and a bleed from the scrubber to maintain acid gas removal efficiency. In
this acidic medium, deposition of SO2 is low, so a second scrubber stage is required for its
removal.

Removal of sulphur dioxide is achieved in a washing stage controlled at a pH close to neutral or

alkaline (generally pH 6–7) in which caustic soda solution or milk of lime is added. For
technical reasons this removal takes place in a separate washing stage, in which, additionally,
there occurs further removal of HCl and HF.

If the treated waste contains bromine and iodine, these elements can be deposited from the flue-
gas flow if waste containing sulphur is combusted simultaneously. In addition to sulphur
compounds, water-soluble salts of bromine and iodine will form, which can be deposited
through the wet SO2 flue-gas cleaning processes. Additionally, the deposition of elementary
bromine and iodine may be improved by specific employment of reductive washing stages
(sulphite solution, bisulphite solution). In any case, it is important to be aware of which wastes
contain iodine or bromine.

If milk of lime or limestone is used as a neutralising agent in the wet flue-gas cleaning stages,
sulphate (as gypsum), carbonates and fluorides will accumulate as water-insoluble residues.
These substances may be removed to reduce the salt load in the waste water and hence reduce
the risk of encrustation within the scrubbing system. Residues of the cleaning process (e.g.
gypsum) can be recovered. When using a caustic soda solution there is no such risk because the
reaction products are water-soluble. If NaOH is used, CaCO3 may form (depending upon water
hardness), which will again lead to deposits within the scrubber. These deposits need to be
removed periodically by acidification.

The diagram below shows a typical two-stage wet scrubbing system. The number of scrubbing
stages usually varies between one and four with multiple stages being incorporated in each
vessel.

Waste Incineration 101

Chapter 2

Figure 2.36: Diagram of a two-stage wet scrubber with upstream dedusting

Wet scrubbing increases the moisture in the flue-gas and therefore the plume visibility,
especially with lower ambient temperature and higher humidity levels. Increasing the
temperature of the flue-gases provides one way of reducing plume visibility, as well as
improving the dispersion characteristics of the release; reheating the flue-gases requires
however some energy. Depending on the flue-gas moisture content and atmospheric conditions,
plume visibility is greatly reduced above stack release temperatures of 140 °C. Reducing the
moisture content of the flue-gases by the use of condensing scrubbers (see Section 2.4.4.5) also
reduces the plume visibility.

Waste water from wet scrubbers

To maintain scrubbing efficiency and prevent clogging in the wet scrubbing system, a portion of
the scrubber liquor must be removed from the circuit as waste water. This waste water must be
subjected to special treatment (neutralisation, precipitation of heavy metals), before discharge or
use internally. Mercury removal is given special attention. Volatile mercury compounds, such as
HgCl2, will condense when the flue-gas is cooled, and dissolve in the scrubber effluent. The
addition of reagents for the specific removal of mercury provides a means of removing it from
the process.

In many plants fitted with wet abatement systems, the waste water produced is evaporated in the
incineration plant by spraying it back into the flue-gas as a quench in combination with a dry
dust abatement system. This avoids waste water emissions from the flue-gas treatment system
and avoids waste water treatment costs. Other techniques to avoid water releases are described
in Section 4.6.1.

102 Waste Incineration

 Chapter 2

2.5.4.2 Semi-wet processes

In the version of this process that is often called semi-dry, water and hydrated lime are injected
separately, typically into a conditioning tower and/or duct. In the spray absorption version, the
absorption agent is injected either as a suspension or solution into the hot flue-gas flow in a
spray reactor (see Figure 2.37).

This type of process utilises the heat of the flue-gas for the evaporation of the solvent (water).
The reaction products generated are solid and need to be deposited from the flue-gas as dust in a
subsequent stage, e.g. bag filter. Compared to dry processes, these processes are typically more
efficient and require less overdosing of the sorbent.

Here, the bag filter is also an important part of the process. Plumes are also rarely visible with
this technique.

Source: [ 1, UBA 2001 ]

Figure 2.37: Operating principle of a spray absorber

A system which falls between the normal dry and semi-wet systems is also applied. This is
sometimes known as semi-dry with CFB (circulating fluidised bed) reactor. These systems re-
inject into the inlet flue-gas a proportion of the solids collected in the bag filter. Water is added
at a controlled rate either directly to the collected fly ash and reagent or into the reactor
upstream of the filter in such a way as to ensure that solids remain free-flowing and not prone to
stickiness or scaling. The reactor design might be as a circulating fluidised bed (CFB) reactor or
a simple plug-flow reactor (duct). No slurry handling is required (in contrast to semi-wet
systems) and no effluents are produced (in contrast to wet systems).

The recycling of reagent reduces the demand for reagent and the amount of solid residue
produced. It can also be applied to dry and semi-wet systems.

Waste Incineration 103

Chapter 2

Another system that is not completely dry or semi-wet or semi-dry uses the condensation effect
on solid particles recirculated into the flue-gas upstream of the fabric filter. This effect can be
achieved by steam injection or by cooling the solids prior to re-feeding.
[78, ESWET, 2015]

2.5.4.3 Dry processes

In dry sorption processes, the absorption agent is fed into the reactor as a dry powder. The dose
rate of reagent may depend on the temperature, the SO2/(HCl+HF) ratio, the process conditions,
and on the reagent type. The reaction products generated are solid and need to be removed from
the flue-gas as dust in a subsequent stage, normally a bag filter.

The overdosing of reagent leads to a corresponding increase in the amount of residues, unless
reagent recirculation is carried out. Significant improvements have been made over the course
of the last 15 years to the techniques for the recirculation of the unreacted fraction, allowing for
the reduction of the stoichiometric excess.

If there is no pre-dedusting stage (e.g. electrostatic precipitator), particles are removed with the
used reagent and reaction products. The cake of reagent that forms on fabric filters provides for
effective contact between flue-gas and absorbent, thus acting as a second reactor stage.

Plumes are rarely visible with this technique.

Figure 2.38: Schematic diagram of a dry FGC system with reagent injection into the FG duct and
downstream bag filtration

104 Waste Incineration

 Chapter 2

2.5.4.4 Direct desulphurisation

[ 1, UBA 2001 ] Desulphurisation in fluidised bed processes can be carried out by adding
absorbents (e.g. calcium or calcium/magnesium compounds) directly into the incineration
chamber. Additives such as limestone dust, hydrated lime and dolomitic dust are used. The
system can be used in combination with downstream flue-gas desulphurisation.

The arrangement of the jets and the injection speed influence the distribution of the absorbents
and thus the degree of sulphur dioxide deposition. Part of the resulting reaction products is
removed in filter installations downstream; however, a significant proportion remains with the
bottom ashes. Therefore, direct desulphurisation may impact on bottom ash quality.
[ 64, TWG 2003 ]

Ideal conditions for direct desulphurisation exist in a cycloid furnace due to the constant
temperature.

On its own, this technique does not lead to compliance with the required SO2 emission levels.
However, it is useful as a pre-treatment and contributes to meeting the lowest emission levels
when applied in combination with other techniques.

The amount of residue from the flue-gas cleaning system itself can be reduced, resulting in
lower disposal costs.

Absorption (and adsorption) of pollutants can also be performed in a (circulating) fluidised bed
reactor into which residues and reagents are recirculated in the combustor at a high rate.
Recirculation of flue-gas keeps the gas flow above a minimum level in order to maintain
fluidisation of the bed. The bed material is separated in a bag filter. Injection of water reduces
the consumption of absorbents (and hence the production of residues) significantly.
[ 74, TWG 2004 ]

2.5.4.5 Selection of alkaline reagent

Various alkaline reagents (and combinations) are used in FGC systems of waste incineration
plants.

Calcium hydroxide is used in all types of FGC systems: wet, semi-wet (or semi-dry) and dry.
For (semi-)wet systems, milk of lime can be prepared on site in a quicklime slaking unit or
starting from hydrated lime. In (semi-)dry processes, hydrated lime or high specific surface
(HSS) hydrated lime are used directly. [7, TWG 2017]. Sodium bicarbonate is applied to a
range of mainly dry systems. Sodium hydroxide and limestone are generally only applied to wet
FGC systems.

The advantages and disadvantages of the different reagents, usually strongly influenced by the
overall technology selection, are summarised in the following table.

Waste Incineration 105

Chapter 2

Table 2.16: Comparison of features of various alkaline reagents

Reagent Advantages Disadvantages Comments/other data
 Higher cost/kg
reagent
 Variable cost
(quarterly)  Only used in wet
 The waste water systems
 Highly reactive with acid gases
Sodium produced by wet  Well suited to
 Low consumption rates
hydroxide sodium hydroxide variable inlet
 Low solid waste production
processes contains concentrations, e.g.
soluble salts to be HWI
handled and treated
 Highly corrosive
material
 High reactivity can be reached
with HSS lime
 Handling can be
Quicklime  Possibility to operate at higher  Residue recirculation
problematic in non-
and temperature with HSS lime is possible and
optimised systems
hydrated  Lower cost/kg reagent allows reducing
lime  Limited options to
 Low solubility residues specific consumption
recycle the residues
 Can allow gypsum recovery
from wet scrubbers
 Not widely applied
 Medium reactivity  Wet limestone in MSWI
 Lower cost/kg reagent processes generate a  Mainly used in wet
Limestone  Low solubility residues purge (suspension) systems
 Can allow gypsum recovery that needs to be  Sometimes used in
from wet scrubbers handled and treated fluidised bed
systems
 Highly reactive with SO2 and  Higher soluble part in
HCl residue
 Low consumption rates  Soluble solid residues
(stoichiometric ratio 1.05–1.20 ) formed can be
problematic for
 Low residue production  Typical reagent
disposal (but use in
depending on the stoichiometric consumption ranges
chemical industry is
ratio are 6–12 kg/t waste
possible)
 Purification and reuse of residue incinerated for MSW
Sodium  Higher cost than and 15–40 kg/t for
possible and applied
lime/kg reagent
bicarbonate  Effective over wide FGC HW
 Depending on the
operating temperature range  Residues without fly
particle granularity of
(120–300 °C) and boiler ash are on
the reagent as
 High operating temperature average 70–75% of
supplied, a grinding
range and high efficiency on SO2 the reagent weight
device may be
may increase compatibility with
required and affect
SCR
availability due to
 No water injection/humidity possible fouling
control required problems
Source: [ 64, TWG 2003 ] [ 74, TWG 2004 ]

The temperature range which is adequate for the dry sodium bicarbonate process is linked to the
transformation of sodium bicarbonate to sodium carbonate; this phenomenon increases the
surface area and porosity of the reagent and hence its reactivity. The phenomenon is observable
from temperatures around 80-100 ºC depending on the particle size, but higher temperatures are
required to ensure reaction kinetics are acceptable. From 140 ºC the kinetics are generally fast
enough, with experience of further increases in reactivity at temperatures of 160–180 ºC.

106 Waste Incineration

 Chapter 2

In some cases, mixed FGC systems are implemented: they can operate either with HSS lime or
sodium bicarbonate, allowing for better control of reagent costs despite a possible need to
compromise on process optimisation. [ 74, TWG 2004 ]

The co-injection of hydrated lime with sodium bicarbonate is also possible, and has recently
been applied in plants in Germany and in the Netherlands, to make up for the lower reactivity of
sodium bicarbonate with HF and thus improve the continuous control of HF emissions. The co-
injection of hydrated lime has been reported to also potentially reduce the overall use of
reagents and associated operating costs [ 99, EuLA 2015 ].

The overall abatement cost is determined both by the reagent cost (unit cost per kilogram of
reagent and amount required ) and by the residue treatment/disposal cost. The flue-gas
composition (which affects the stoichiometric ratio of the different possible reagents/processes),
the unit price per kilogram of reagent, as well as the availability and cost of residue
treatment/disposal options are important factors influencing the total abatement cost.

Limestone, quicklime, hydrated lime, enhanced (high-surface) hydrated lime, sodium hydroxide
and sodium bicarbonate are all used in a wide variety of incineration plants throughout Europe
and elsewhere.

2.5.5 Techniques for the reduction of emissions of oxides of nitrogen

[ 3, Austria 2002 ]
Nitrogen oxides (NOX) may be formed in three ways:

 Thermal NOX: During combustion a part of the air nitrogen is oxidised to nitrogen
oxides. This reaction only takes place significantly at temperatures above 1 300 °C. The
reaction rate depends exponentially on the temperature and is directly proportional to
the oxygen content.
 Fuel NOX: During combustion a part of the nitrogen contained in the fuel (including in
the waste) is oxidised to nitrogen oxides.
 Formation of NOX via radical reaction (prompt NOX): Atmospheric nitrogen can also be
oxidised by reaction with CH radicals and intermediate formation of HCN. This
mechanism of formation is of relatively low importance in waste incineration.

Waste Incineration 107

Chapter 2

Source: [ 3, Austria 2002 ]

Figure 2.39: Temperature dependence of various NOX formation mechanisms in waste incineration

2.5.5.1 Primary techniques for NOX reduction

[ 1, UBA 2001 ]
NOX production can be reduced using furnace control measures that:

 prevent oversupply of air (i.e. prevent the supply of additional nitrogen);

 prevent the use of unnecessarily high furnace temperatures (including local hot spots);
 optimise combustion control for the homogenisation of the combustion conditions and
to avoid sharp temperature gradients.

2.5.5.1.1 Air supply, gas mixing and temperature control

The use of a well-distributed primary and secondary air supply to avoid the uneven temperature
gradients that result in high-temperature zones and, hence, increased NOX production is a
widely adopted and important primary technique for the reduction of NOX production.

Although sufficient oxygen is required to ensure that organic materials are oxidised (giving low
CO and VOC emissions), the oversupply of air can result in additional oxidation of atmospheric
nitrogen, and the production of additional NOX.

Achieving effective gas mixing and temperature control are important elements.

108 Waste Incineration

 Chapter 2

2.5.5.1.2 Flue-gas recirculation

This technique involves replacement of around 10–20 % of the secondary combustion air with
recirculated flue-gases. NOX reduction is achieved because the supplied recirculated flue-gases
have a lower oxygen concentration and therefore lower flue-gas temperature which leads to a
decrease of the nitrogen oxide levels. [ 74, TWG 2004 ]

2.5.5.1.3 Oxygen injection

The injection of either pure oxygen or oxygen-enriched air provides a means of supplying the
oxygen required for combustion, while reducing the supply of additional nitrogen that may
contribute to additional NOX production.

2.5.5.1.4 Staged combustion

Staged combustion has been used in some cases. This involves reducing the oxygen supply in
the primary reaction zones and then increasing the air (and hence oxygen) supply in later
combustion zones to oxidise the gases formed. Such techniques require effective air/gas mixing
in the secondary zone to ensure CO (and other products of incomplete combustion) are
maintained at low levels.

2.5.5.1.5 Natural gas injection (reburn)

[70, USEPA, 1994]

Natural gas injection into the over-grate region of the furnace can be used to control NOX
emissions from the combustor. For MSWIs, two different natural-gas-based processes have
been developed:

 reburning – a three-stage process designed to convert NOX to N2 by injecting natural gas

into a distinct reburn zone located above the primary combustion zone;
 methane de-NOX – this technique injects natural gas directly into the primary
combustion unit to inhibit NOX formation.

2.5.5.1.6 Injection of water into furnace/flame

A properly designed and operated injection of water either into the furnace or directly into the
flame can be used to decrease the hot spot temperatures in the primary combustion zone. This
drop in peak temperature can reduce the formation of thermal NOX.
[ 74, TWG 2004 ]

2.5.5.2 Secondary techniques for NOX reduction

[ 1, UBA 2001 ] In order to achieve compliance with NOX emission limits, it is common for
secondary measures to be applied. For most processes, the application of ammonia or
derivatives of ammonia (e.g. urea) as a reduction agent has proved successful. The nitrogen
oxides in the flue-gas basically consist of NO and NO2 and are reduced to N2 and water vapour
by the reduction agent.

Reaction equations:

4 NO + 4 NH3 + O2  4 N2 + 6 H2O
2 NO2 + 4 NH3 + O2  3 N2 + 6 H2O

Waste Incineration 109

 Chapter 2
 
Two processes are important for the removal of nitrogen from flue-gases - selective non-
catalytic reduction (SNCR) and selective catalytic reduction (SCR).
 
Both NH3 and urea are applied as aqueous solutions. For safety reasons, NH3 is normally
supplied as a 25 % solution. Urea can also be applied in the pure solid form.
 
 
 
2.5.5.2.1 Selective non-catalytic reduction (SNCR) process
 
In the selective non-catalytic reduction (SNCR) process, nitrogen oxides (NO + NO2) are removed by
injecting the reducing agent (typically ammonia or urea) into the furnace. The reactions occur at
temperatures between 850 °C and 1 000 °C, with zones of higher and lower reaction rates within this
range. The operating principle of the technique is illustrated in Figure 2.40.
 

 
Source: Martin J.J., Koralewska R., Wohlleben A. (2015). Advanced solutions in combustion-based WtE technologies. Waste
management, 37, 147-156
 
Figure 2.40: SNCR operating principle
 
 
 
Reducing NOX levels by more than 60–80 % using SNCR requires a higher addition of the reducing
agent. This can lead to emissions of ammonia, also known as ammonia slip. The relationship between NOX
reduction, ammonia slip and the reaction temperature is given in Figure 2.41 below.
 
 
 
 
 
 
110 Waste Incineration
 Chapter 2

Source: [3, Austria, 2002] [ 64, TWG 2003 ]

Figure 2.41: Relationship between NOX reduction, production, ammonia slip and reaction
temperature for the SNCR process

Figure 2.41 shows that, at a reaction temperature of 1 000 °C, the reduction of NOX would be
about 85 %, with an ammonia slip of about 15 %. In addition, at this temperature there would be
a production of NOX, from the incineration of the injected NH3, of about 25 %.

Figure 2.41 also shows that, at higher temperatures (with ammonia), the percentage of NOX
reduction is higher, and, while the ammonia slip is lower, the NOX produced from the ammonia
rises. At high temperatures (> 1 200 °C), NH3 itself oxidises and forms NOX. At lower operating
temperatures, the NOX reduction is less efficient and the ammonia slip is higher.

Application of urea instead of ammonia in SNCR leads to relatively higher N2O emissions in
comparison with ammonia reduction. [ 64, TWG 2003 ]

In order to ensure an optimum utilisation of ammonia at varying loads, which cause varying
temperatures in the combustion chamber, NH3 can be injected into several layers.

When used with wet scrubbing systems, the excess ammonia may be removed in the wet
scrubber. The ammonia can then be recovered from the scrubber effluent using an ammonia
stripper and fed back to the SNCR feed system.

Waste Incineration 111

Chapter 2

The effective mixing of flue-gases and NOX reduction reagent is important for optimisation of
the SNCR process and there must be sufficient gas residence time to allow the NOX reduction
reactions to occur.

In the case of pyrolysis and gasification processes, optimisation of SNCR is achieved by

injecting the reagent into the syngas combustion zones with a well-controlled temperature and
effective gas mixing.

Selection of reagents for SNCR

The reagents used for SNCR are ammonia and urea.

Obtaining a good understanding of temperature profiles in the combustion chamber is

fundamental to the selection of the reagent.

New plants can be specifically designed to achieve stable and predictable combustion
conditions, allowing for optimal injection locations for the reagent, thus making it possible to
use ammonia with the maximum environmental benefit (i.e. highest peak NOX reduction at
lowest N2O emissions). This also applies to existing plants that have stable and well-controlled
combustion and temperature profiles in the furnace.

Existing plants that experience difficulties in stabilising combustion conditions (e.g. for design,
control or waste type reasons) are less likely to be in a position to optimise the reagent injection
(location, temperature, mixing) and may therefore benefit from the use of urea. However, if
temperatures above 1 000 °C are anticipated, the N2O production rate with urea becomes more
significant.

In cases where the advantages and disadvantages are finely balanced, storage and handling
hazards may have a greater impact on the final reagent selection.

The relative advantages and disadvantages of ammonia and urea are outlined in the table below.
The reagent selection needs to take account of a variety of process operational, cost and
performance factors to ensure that the optimal one is selected for the installation concerned.

Table 2.17: Advantages and disadvantages of urea and ammonia use for SNCR
Reagent Advantages
 Higher peak NOX reduction
potential (if well optimised)
Ammonia
 Lower N2O emissions (10–
15 mg/Nm3)
 Wider effective temperature range
(750–1 000 °C) makes temperature
Urea control less critical
 Less storage and handling hazards
 Lower cost per tonne of waste
Source: [62, Tyseley 2001] [ 64, TWG 2003 ]
Disadvantages
 Narrower effective temperature
range (850–950 °C) therefore
greater optimisation is required
 Handling and storage hazards
higher
 Higher cost per tonne of waste
 Odour of residues if in contact with
humidity
 Lower peak NOX reduction
potential (compared to ammonia
when optimised)
 Higher N2O emissions (25–
35 mg/Nm3) and hence GWP

The urea/N2O reaction is very dependent on temperature, with as much as 18 % of the NOX
removed appearing as N2O at 1 000 °C, while it is negligible at 780 °C.

Ammonia is reported to be marginally more expensive to use than urea. Handling and storage
requirements for ammonia liquid, gas and solutions are generally more stringent and, hence,

112 Waste Incineration

 Chapter 2

more expensive than for urea, which can be stored as a solid - this contributes to the cost
differential between the two reagents.

Storage of gaseous or liquefied ammonia is subject to stringent safety requirements, which may
cause extra costs. In most cases ammonia is used as a solution, as in this case the safety
requirements are less onerous.

The use of urea is only cheaper for relatively small plants. For larger plants, the higher storage
cost of ammonia may be fully compensated by the lower chemical cost.

2.5.5.2.2 Selective catalytic reduction (SCR) process

In the selective catalytic reduction (SCR) process, an ammonia- air mixture (the reduction
agent) is added to the flue-gas and passed over a catalyst, usually a mesh (e.g. platinum,
rhodium, TiO2, zeolites). [ 74, TWG 2004 ] When passing through the catalyst, ammonia reacts
with NOX to give nitrogen and water vapour.

To be effective, the catalyst usually requires a temperature of between 150 °C and 450 °C. The
majority of systems used in waste incinerators currently operate in the range of 180–250 °C so
as to minimise the need to reheat the flue-gas, even if more catalyst surface is necessary and
there is a greater risk of fouling and catalyst poisoning at these lower temperatures. In some
cases, catalyst-temperature-regulated bypasses are used to avoid damage to the SCR unit.
[ 74, TWG 2004 ], [7, TWG 2017]

SCR gives high NOX reduction rates (typically over 90 %) at close to stoichiometric additions of
the reduction agent. For waste incineration, SCR is mainly applied in the clean gas area, i.e.
after dedusting and acid gas removal. For this reason, the flue-gases generally require reheating
to the effective reaction temperature of the SCR system. This adds to the energy requirements of
the flue-gas treatment system. However, when SOX levels in the flue-gas have already been
reduced to a very low value at the inlet of the SCR section, reheating may be reduced
substantially or even omitted. Heat exchangers are used to reduce additional energy demand.

After a wet FGC system, droplets may be removed to prevent salt deposits inside the catalyst.
Due to the risk of ignition resulting from the accumulation of CO in the catalyst, safety
measures are of importance, e.g. bypasses, CO control. [ 74, TWG 2004 ]

Low-temperature SCR requires low levels of SOX (generally below 20 mg/Nm3) to minimise the
risk of ammonium sulphate formation. In the event of ammonium sulphate deposition on the
catalyst layers, the deposits are removed by thermal regeneration. The thermal regeneration
should be carefully controlled because the salt sublimation may lead to peaks in the releases to
air for some pollutants, e.g. SO2, NOX, NH3, which can be avoided by recirculation of the
exhaust gas from regeneration or by controlling the heating rate. [ 74, TWG 2004 ]

SCR is sometimes positioned directly after the ESP, to reduce or eliminate the need for
reheating the flue-gas. When this option is used, the additional risk of PCDD/F formation in the
ESP (typically when the ESP is operated at temperatures above 220–250 ºC) must be
considered. Such operation can result in increased PCDD/F emissions in the gas stream leaving
the ESP and in the ESP residues. However, since SCR can also be used to provide for partial
PCDD/F destruction, PCDD/F emissions downstream of the SCR are generally low.

Waste Incineration 113

Chapter 2

Source: [ 3, Austria 2002 ]

Figure 2.42: SCR operating principle

The flue-gases discharged from the reactor may be directed through a gas-gas heat exchanger to
preheat the entering gases in order to maintain the operating temperature of the catalyst and to
save some of the imported energy.

2.5.6 Techniques for the reduction of mercury emissions

2.5.6.1 Primary techniques

Mercury is highly volatile and therefore almost exclusively passes into the flue-gas stream.

The only relevant primary techniques for preventing emissions of mercury to air are those which
prevent or control, if possible, the inclusion of mercury in the waste:

 efficient separate collection of waste that may contain heavy metals, e.g. cells, batteries,
dental amalgams;
 notification of waste producers of the need to segregate mercury;
 identification and/or restriction of receipt of potential mercury-contaminated wastes:
o by sampling and analysis of wastes where this is possible;
o by targeted sampling/testing campaigns;
 where such wastes are known to be received - controlled addition to avoid overload of
the abatement system capacity.

2.5.6.2 Secondary techniques

Mercury vaporises completely at a temperature of 357 °C and remains gaseous in the flue-gas
after passing through the furnace and boiler. Inorganic mercury (mainly Hg2+ as a chloride and,
in specific cases, other halogenic compounds) and elemental mercury are affected differently by
FGC systems and detailed consideration of the fate of both is required. [ 1, UBA 2001 ]

The selection of a process for mercury abatement depends on the mercury content and the
chlorine content of the feed material. At higher chlorine contents, mercury in the crude flue-gas
will increasingly be in the ionic form which can be deposited in wet scrubbers. This is a
particular consideration at sewage sludge incineration plants where raw gas chlorine levels may
be low. If, however, the chlorine content in the (dry) sewage sludge is 0.3 mass-% or higher,
only 10 % of the mercury in the flue-gas [ 74, TWG 2004 ] is elemental. The presence of an
SCR system in the high-dust configuration can also support the oxidation of elemental mercury
and its subsequent capture in the downstream elements of the FGC system.

114 Waste Incineration

 Chapter 2

Elemental mercury can be removed from the flue-gas stream by the following:

 Transformation into ionic mercury by adding oxidants and then deposition in a wet
scrubber - the effluent can then be fed to waste water treatment plants with heavy metal
deposition, where the mercury can be converted to a more stable form (e.g. HgS), which
is more suitable for final disposal [ 74, TWG 2004 ].
 The alternative option is direct deposition on brominated or sulphur-doped activated
carbon, hearth furnace coke, or zeolites. Adsorption on activated carbon can take place
in a fixed-bed absorber or on activated carbon injected into the gas flow. In the latter
case the carbon accumulates on the bag filters used to collect it, where it forms a filter
cake. The contact time and mixing with the sorbent increase when the flue-gas passes
through the filter cake, enhancing mercury capture.

Tests have shown that sulphur dioxide neutralisation in the furnace by adding limestone can
reduce the proportion of metallic mercury, making overall mercury removal from the gas stream
more efficient.

In incineration plants for municipal and hazardous wastes, the chlorine content in the waste is
usually high enough to ensure that mercury is present mainly in the ionic form. However,
specific inputs of certain waste may change the situation and metallic mercury may need to be
removed with the use of specific techniques, as mentioned above.

High-mercury wastes
For the incineration of waste with a high mercury content in hazardous waste incineration
plants, a mercury deposition level of 99.9 % can only be ensured when highly chlorinated waste
is also incinerated in an appropriate proportion to the mercury load. Multistage wet scrubbing
processes are typical of this type of plant and these can achieve almost total mercury removal
from the flue-gas.

High total chlorine loads (approximately 4 wt-% input) and a therefore high interim Cl2 supply
lead to high mercury chlorination levels and a mercury deposition level of close to 100 %. With
lower chlorine loads, the mercury deposition reduces rapidly.

Source: [ 1, UBA 2001 ]

Figure 2.43: Relationship between fraction of mercury present in elemental form and the raw gas'
chloride content at a hazardous waste incineration plant

Waste Incineration 115

 Fig.S12 A modern Waste-to-Energy plant
(annual capacity: 150,000 Tons municipal solid waste incinerated per year)
Table F04.

Source: IED Industrial Emissions Directive 2010/75/EU

Note: To avoid willful dilution with ambient air (to make concentration figures look lower), analytical data
for the gases emitted at the stack always have to be expressed at some standard concentration of either
oxygen (e.g., 11 vol.-% O2) or carbon dioxide (e.g., 6 or 8 vol.-% CO2).
Further reading

BREF (2019) by Neuwahl F., Cusano G., Gomez Benavides J., Holbrook S., Roudier S. Best Available
Techniques (BAT) Reference Document for Waste Incineration. JRC EUR 29971 EN;
doi:10.2760/761437.

Buekens, A. (2013). Incineration technologies. Springer Science & Business Media.

Damgaard, A., Riber, C., Fruergaard, T., Hulgaard, T., & Christensen, T. H. (2010). Life-cycle-assessment
of the historical development of air pollution control and energy recovery in waste incineration. Waste
Management, 30(7), 1244-1250.

IWSA (2015) by Kahle K., Kamus B. Bottom ash from WtE plants: Metal recovery and utilization.
International Solid Waste Association Report, 30p.

Martin, J.J., Koralewska, R., Wohlleben, A. (2015). Advanced solutions in combustion-based WtE
technologies. Waste management, 37, 147-156.

Murer, M.J., Spliethoff, H., Van Berlo, M.A.J., De Waal, C.M.W. & Gohlke, O. (2009). Comparison of
energy efficiency indicators for energy-from-waste plants. In Sardinia 2009 Symposium.

Sakai S, Hiraoka M (2000) Municipal solid waste incinerator residue recycling by thermal processes.
Waste Manag 20:249–258

Schmelzer, G (1995). Separation of metals from waste incineration residue by application of mineral
processing. In: Proceedings of the XIX International Mineral Processing Congress, vol. 4. pp. 137-140.

Schmelzer, G., Wolf, S., Hoberg, H. (1996). New wet treatment for components of incineration slag.
Aufbereitungs-Technik/Mineral Processing 37 (4), 149–157.

Schwanecke, R. (1976). Formulas and nomograms for use of the technical guide for air pollution control.
Wasser, Luft und Betrieb, 20(11).

Tanner R. (1965) Die Entwicklung der Von Roll-Muellverbrennungsanlagen. Schweizer Bauzeitung

83(16).

TWG, Comments to Draft 1 of the WI BREF, 2017.

UBA, Draft of a German Report for the creation of a BREF document "waste incineration", 2001.

Umweltbundesamt (2018). Best Practice Municipal Waste Management. Texte 40/2018. Dessau-Rosslau,
310p.

Van Caneghem, J., Brems, A., Lievens, P., Block, C., Billen, P., Vermeulen, I., Baeyens J. &
Vandecasteele, C. (2012). Fluidized bed waste incinerators: Design, operational and environmental issues.
Progress in Energy and Combustion Science, 38(4), 551-582.