EF Nr.

194010

Bundesamt fr Energie
Office fdraI de Inergie
Ufficio federale deIIenergia
Ufflzi federal denerqia

Biomass Programme

Furnace Design and Combustion Control to

Reduce Emissions and Avoid Ash Siagging
International Energy Agency, Biomass Combustion Activity
1997
Final Report of the Triennium 1995

Thomas Nussbaumer
Swiss Federal Institute of Technology (ETH), CH
Verenum Research, Langmauerstrasse 109, CH
SWITZERLAND

8092 Zrich and

8006 Zrich

on behalf of the
Swiss Federal Office of Energy

March 1998

Final Report

Contents

Summary

.5

2 Introduction

2.1 Motivation

2.2 Objectives

2.3 Definitions

3 IEA Contributions
3.1 Experiments on a moving grate furnace in Switzerland

8
8

3.2 Experiments for the investigation of ash cieposition in USA

10

3.3 Experiments with a whole bale furnace in Denmark

11

4 Fuel Chararcteristics of Biomass

1 2

4.1 Composition of biomass

12

4.2 Triangular diagram combustiblewaterash

14

4.3 Nitrogen content in the 1 uel

15

4.4 Content of sulfur, chiorine and potassium in the fueE

15

5 Furnace Design and Process Control

1 7

5.1 Fundamentals of biomass combustion

17

5.2 Automatic wood furnaces

20

5.3 Furnaces for herbaceous fuels

22

5.4 Process control

23

6 Emissions and Depositions

26

6.1 Nitric oxide emissions

26

6.2 Particle emissions

27

6.3 Particle Separation in a rotating particle separator

29

6.4 Chlorine emissions

31

6.5 Deposit formation

32

6.6 Dioxins and furans

34

7 Reterences

35

Summary

The aim of the present project was to describe the main influences of the fuel characteristics, the
furnace design, and the Operation conditions on the formation of emissions during biomass com
bustion. Furthermore the parameters which can cause ash slagging and which influence the formation
of depositions should be identified. The relevant parameters were investigated in different projects
within the framework of the IEA Biomass Combustion Activity. Detailed results from different research
projects have been contributed 1 rom the Swiss Federal Institute of Technology and Verenum
Research from SWITZERLAND and from Sandia National Laboratories from USA. Furthermore several
informations have been contributed from all other IEA members during the IEA seminars and project
meetings from 1995 to 1997.
The present study gives an overview on typical fuel characteristics of wood and herbaceous fuels such
as straw, miscanthus and herbage-grass (hay). lt is shown that herbaceous biomass has significantly
higher contents of nitrogen, chlorine, potassium and ash. Furthermore the ash softening temperature
15 lower than tor native wood.
The main process steps during combustion are described in the report and typical biomass furnaces
are presented. lt is shown that the formation of unburnt pollutants (CC, HC, PAH) can be avoided by
complete combustion. An appropriate furnace design is needed with agood mixing quality between
combustion air and combustible gases. Advanced combustion control technologies can be applied to
guarantee optimum combustion conditions in practice.
NO emissions are formed from the fuel bound nitrogen. Therefore specitic measures in the com
bustion process such as air staging or fuel staging are necessary to reduce NO formation. In the
opposite to CO emissions, NO emissions cannot be reduced below a certain level by primary
measures. To meet bw emission standards, secondary measures can be applied.
Particles from biomass combustion are mainly in the aerosol range < 1 jim. To achieve 10w particle
concentrations in the flue gas, fabric or electric filters are needed. As an alternative to filters, a rotational
particle separator has been investigated which offers a better Separation efficiency than cyclones.
If herbaceous fuels are burnt, significant concentrations of chlorine can be found in the fuel and in the
flue gas. Most of the chlorine is emitted as salt (mainly as KCI) which can lead to depositions on the
furnace walls and in the boiler. Due to the bw ash softening temperature of herbacous biomass,
significant ash slagging can occur on the grate. Since there is a conflict between achieving a complete
burnout (high temperature) and avoiding ash slagging (10w temperature), the furnace design and
Operation tor such fuels is different from native wood. Special measures such as cooled walls in the
grate section of the furnace and flue gas recirculation are considered for these fuels.

Introduction

2.1

Motivation

The flue gas emissions and the combustion efficiency are influenced by the fuel characteristics, the
furnace design and the operation conditions. For woody biomass, appropriate combustion techno
logies are available which ensure high combustion quality and high efficiency and which are able to
guarantee Operation with 10w maintenance effort. The Operation fl practice can be optimized with an
appropriate combustion control system.
However, there are still furnaces available which do not consider the specific combustion characteristic
of biomass sufficiently. Furthermore fuel specific pollutants as nitrogen oxides can be emitted also in
furnaces with appropriate combustion design. Furthermore the combustion of biomass with 10w ash
softening temperature like straw, grass, and miscanthus, ash slagging on the grate and ash depositions
on the the furnace walls and in the boiler can cause severe Operation problems and damage the
furnace.
To reach a good combustion quality a high temperature in the combustion chamber is necessary, while
10w temperatures are needed to avoid slagging. Therefore a conflict between the requirements to
reach 10w emissions and to avoid Operation problems can be found.

2.2

Objectives

In the present project the main influences between furnace design, operation conditions, emissions,
efficiency and ash behaviour are studied. Results from experiments and data from the literature are
used to consolidate principles of furnace design and operation. For that purpose the influence of the
following parameters on emissions, ash slagging and ash deposition have been investigated:

overall excess air ratio and resulting combustion temperature

influence of over- and under-stoichiometric conditions in the gasification zone
(primary excess air ratio)
quality of the primary air distribution in the gasification zone and of the homogenousity
of the fuel bed (avoiding wholes in the fuel bed with continuous movement of the grate)
cooling of the furnace walls and cooling of the grate and
flow direction of fuel and gas in grate furnaces (co-, cross- and counter-current).

2.3

Definitions

During the thermal treatment of biomass in the furnace, the processes of drying, pyrolysis, gasification
and oxidation take place. To define the different process steps of thermochmical conversion, the ratio
of fuel to air is described by the excess air ratio ? as follows:

rii,
mVSh

amount of air supplied

stoichiometric amount of air

The characteristics of the different process steps can be described by the excess air ratio and the
temperature as shown in Table 2.1.

Process

Excess air

[]
Combustion

>

Gasification

0.2

Pyrolysis

Table 2.1

<

<

Temperature
[C]
800

1 300

Main
product
hot exhaust
gas

Most important applications

of the main product
Production of hot water, steam, and hot
gas. Steam and hot gas for electric power
prod uction.

0.5

700

900

producer gas
of high
thermal value

Fuel tor gas engine, gas turbine,

sythesis gas (e.g. as basis tor fuel or
methane), and fuel cells

(< 0.2)

400

700

liquid of high
thermal value
(pyrolysis oil)

Fuel for diesel engine or gas turbine.

Gas is used internally in the process,
charcoal accumulates as a by-product.

<

Characterization of thermochemical conversion processes combustion, gasification, and

pyrolysis. The excess air is the ratio of supplied to stoichiometric amount ot air needed
tor a complete combustion. *For pure pyrolysis, 2 = 0. In reality a part of the conversion
products gets burnt right away, therefore can also be slightly bigger than 0.

IEA Contributions

Different investigations on the described subject were carried out in Switzerland by ETH and Verenum
which have been funded by the Swiss Federal 0ff ice of Energy as part of the biomass research
programme. Furthermore the IEA members were asked to contribute with experimental data from
combustion experiments in furnaces and laboratory equipment under varying conditions on emissions,
ash siagging, deposition formation and deposition reduction.
An important contribution to the present project came from Sandia National Laboratories (USA), which
is involved in research on deposit formation in biomass fired boilers. Furthermore a research scientist
from ETH was at Sandia during approx. 6 months to carry out specific combustion experiments on a
plug flow laboratory reactor (Multi Fuel Combustor, MFC) which gave the opportunity to compare
deposition formation found in grate furnaces in Switzerland with depositions found in biomass fired
power stations in the USA and respective laboratory experiments in the MFC.
The main experiments which contributed to the present project are described in the present sections.

3.1

Experiments on a moving grate furnace in Switzerland

Combustion experiments have been carried out in Switzerland in cooperation with the furnace
manufacturer SCHMID AG, Eschlikon:

Combustion tests with pellets and briquetts of grass (hay from extensively used land) and
miscanthus in comparison to wood fuels.

Monitoring of the main emissions (continously: CC, C0

, N0; particle concentration, particle size
2
distribution and further emissions (HCI, NH
3 PCDD/F etc.) are measured discontinously)
Temperature measurements are carried out at several locations at the grate and at the side walls to
investigate the influence of the temperature an ash slagging

Figure 3.1

Moving grate furnace with 450 kWth (SCHMID AG, Eschlikon, Switzerland)

Mass flows of residues are determined and the composition of the residues are analysed and
compared to regulations tor the use as tertilizers (bottom ash and cyclone ash) and for landtilling
(filter ash).

The experiments showed essentially the tollowing results.

Grass and miscanthus can be burnt with high combustion quality and 10w C0-emissions (< 1 00
3 at 11 Vol.-% 02). However severe Operation problems can be found with grass and
mg/Nm
miscanthus due to ash siagging on the grate, ash deposition at the side walls of the primary
combustion chamber and depositions and fouling in the heat exchanger tubes. Ash siagging on the
grate can be avoided with improved grate design and ash moving equipment. Furthermore lt is
assumed that water-cooled walls in the primary combustion chamber are necessary to avoid
depositions. However air-cooling of the side walls was not sufticient to avoid ash depositions.

For the heat exchanger tubes the injection of pressurized air or vapor tor a periodical cleaning can
was succestully applied. However it the depositions exceed a certain temperature, agglomeration
and condensation can lead to very strong depositions which cannot be removed by pressure
shocks. Theretore the heat exchanger inlet temperature must be limitied. Investigations from USA
(Baxter et al.) show, that tor superheaters a temperature

>

850C is critical while tor temperatures

<

750C no severe problems are tound. lt is assumed that the gas temperature in hot water heat
exchangertubes may be higher due to lower wall temperatures than in superheaters. However no
quantitative data are known.

If the energy is used tor space heating only, lower heat exchanger inlet temperature are not of
relevance tor the etticiency. Theretore it is proposed that an efficient water cooling ot the primary
combustion chamber, which leads to an energy output of approx. 50% ot the overall heat, can be
applied. Furthermore flue gas recirculation can be applied as a measure to limit the temperature.
However experiments so far were not satisfying since a large amount ot recycled tlue gas is
necessary for an effective temperature reduction.

Furthermore a test equipment has been built to carry out precipitation tests with ditterent types ot cloth
filters under varying conditions (temperature, pressure drop, velocity etc.) and using ditterent additives
tor the adsorption of HCI. To investigate the deposition formation, experiments were carried out in the
grate turnace as weil as in a 100 kW under stoker.

Placement of the deposit

sampling probe
Secondary air inlet holes

Service doors

Hearth with primary air inlet

Figure 3.2

100 kWth underfeed stokerturnace (TIBA-MUELLER AG, Balsthai, Switzerland)

3.2

Experiments tor the investigation ot ash deposition in USA

As an opportunity within the IEA Biomass Combustion Activity, a PhD student from Switzerland (H.
Kaufmann) carried out a research visit at Sandia National Laboratories in Livermore (CA, USA) in the
group of L. Baxter. The aim of the research visit was to start an improved exchange of experience and
to carry out specific experiments with the biofuels used in tests in Switzerland at the mufti-fuel
combustor equipment at Sandia.
The multi-fuel combustor allows to carry out deposition experiments with different types of fuels, which

are injected as small particies (coal, biomass etc.) or as gaseous fuels. The combustion conditions
(temperature, excess air) can be varied and the exhaust gas is used at different temperatures to
simulate ash deposition conditions as they are found in boilers. The deposition rate on a water cooled
heat exchanger tube can be measured by an electronic balance. Furthermore the composition of the
deposition can be analysed on-line with spectrometric methods (FTIR, laser-spectroscopy) and off-line,
after a certain time of exposures, by analysing the composition of the depositions.

Heatedf
insulated
modules

Pcsv
Laser beam

to Exliaust

Figure 3.3

Laboratory plug flow reactor (MFC) for at Sandia National Laboratories.

Since many experiments of this type have been carried out at Sandia with direct exposure of the heat
exchanger tubes in the exhaust gas, mainly two specific questions were investigated within the
present project:
1. A comparison between grass combustion in grate furnaces and in the MFC
2. The difference between depositions with direct exposition of the heat exchanger tube in the flue
gas and with a preliminary extraction of the arge particle. This objective is important for a comparison
with depositions found in grate furnaces since an effective extraction of large particles on the grate
can be achieved.

10

During the research visit of H. Kaufmann at Sandia, respective tests with grass from Switzerland have
been carried out. The resuits are described in detail in [Kaufmann et al. 1998].

Figure 3.4

3.3

Arrangement of sampling cylinder A and protection cylinder B in the outlet of the MFC

Experiments with a whole bale furnace in Denmark

For the use of grass or miscanthus in grate furnaces the fuels must be transformed to pellets or
briquetts. In the overall economy, the pellet fabrication is a relevant cost factor (costs in Switzerland
app. CHF 110.lt

ECU 70.lt

USD 851t). Therefore the application of whole bales of grass as they

are used in agriculture is a favourable Option.

Straw boilers with whole bale furnaces are mainly in use in Denmark, where different types of furnaces
and feeding systems have been developped. For large units (>2
while for smaller units (>200 kW

3 MWth) cigar burners are used,

2 MWth) special feeding Systems are used to cut the straw bales into

slides and feed the furnace with one slide (app. 20 cm) of a bale.
Since there are only few experiences with grass, combustion tests were carried out on a whole bale
furnace with bale-cutter of app. 1.5 MWth using grass as it was used in the combustion tests in Switzer
land. However the combustion tests were not succesfull due to unsufficient ignition of the grass in the
primary combustion chamber. The manufacturer of the whole bale furnace assumes, that the primary
combustion chamber has to be adopted by adding a radiation surface to increase the temperature.
However an increase of the temperature in the primary combustion chamber can also lead to slagging
and deposition problems which are not found in straw firings. Unfortunately it is not possible to answer
the questions whether grass can be used in whole bale furnaces without slagging and deposition
problems by the experiments carried out as a part of this project. Therefore further experiments for a
comparison between straw and grass were carried out in the grate furnace in Switzerland.

11

4.1

Fuel Chararcteristics of Biomass

Composition of biomass

Table 2.1 shows the typicat composition of wood, grass, miscanthus and straw. The thermal value and
the combustion properties are mainly dependent on the percentage of organic substance in the fuel.
For biogenic fuel water and ash content are the most important parameters. Furthermore, the
combustion characteristic is dependent on the physical properties as size and specific surface. Not
only the percentage of ash is important, but also its melting properties. Wood ash has a much higher
melting point than ash of grass, miscanthus and straw.
The emissions of pollutants and the required measures to be taken are dependent on the content of
nitrogen, sulfur, chiorine and potassium, which is much higher for grass, miscanthus and straw than 1cr
wood. A correlation 01 the nitrogen content with the sum of potassium, chiorine and sulfur content can
seen in Figure 4.4.It can also been seen, that the content of potassium, chlorine and sulfur is de
creasing with increasing lignification and demineralisation. The content of potassium, chiorine and
sulfur is in general higher in grass than in straw and higher in straw than in miscanthus. Phosphor and
potassium content influence the content of nutrients in the residues. The content of lead, cadmium,
chrome, cobalt, copper, molybdenium, nickel, mercury and zinc influence the amount of pollutants in
the residues and the heavy metal emissions in the flue gas. For particle emissions of less than 1 50
3 significantly less than 2 mgINm
mgINm
3 of lead and zinc and significantly less than 0.2 mg/Nm
3 of
mercury can be reached.

12

U nt

Wood

Grass

Miscanthus

Straw

Influence*

Ash content medium

[weight-%]

0,2

9,8

2,43

6,0

Ash contenttypical

[weight-%]

0,11

612

14

4,57,5

Watercontent

[weight-%]

1060

520

520

520

ThermalvaIueH

[MJ/kg
,
1
j

18,5

16,8

17,9

17,5

Sinterbeginn

[C]

1180

870

840

830

Erweichungspunkt

[C]

1470

970

880

940

Haibkugelpunkt

[C]

1600

1040

1040

1100

Fliesspunkt

[CJ

1640

1120

1100

1170

**

**

[weight-%]

50

[weight-%]

[weight-%]

44

[weight-%]

0,08

1,90

0,58

0,54

[weight-%]

0,01

0,20

0,08

0,14

Ci

[weight-%]

<0,001

0,44

0,15

0,37

[weight-%]

0,11

2,45

0,47

1,20

1,2,3

Si

[weight-%j

2,14

1,60

0,91

Ca

[weight-%]

0,61

0,20

0,29

1, 3

[ppm]

720

Pb

[ppm]

1,1

0,7

2,5

0,5

Cd

[ppm]

<0,1

0,10

0,11

0,09

Cr

[ppm]

<

0,5

Co

[ppm]

<0,1

Cu

[ppm]

1,9

7,6

2,9

2,9

Mo

[ppm]

<0,2

0,9

0,4

0,5

Ni

[ppm}

<

Hg

[ppm]

Zn

[ppm]

Table 4.1

0,27

**

1,2

11

0,01

0,01

0,02

38

21

Typical composition of wood, grass, mscanthus and straw.

Data from [Hasler et al. 1996] and [Hofbauer 1994]
*

1: Combustion technique, 2: flue gas cleaning, 3: residue quality

The differences of C, H and 0 content in different tree species are small. According to
Browning [1963] five different leaf and coniferius trees only show the following range:
C of 48.9 to 50.2 weight-%, H of 5.8 to 6.1 weight -% and 0 of 43.4 to 44.5 weight -%.

**

As the contents of C, H and 0 is only little different for different plants, the values for wood
can be taken as an assumption for grass, miscanthus and straw.

13

4.2

Triangular diagram combustiblewaterash

For a characterization of the fuel regarding the combustion properties, the percentage of combustible
matter, the water content and the ash content are of importance. These fuel properties are shown in
the triangular combustiblewaterash diagram. Furthermore the diagram allows to show the operational
range of different combustion systems. Figure 4.1 shows the thangular diagram for native wood, bark
and grass, miscanthus and straw. The combustion properties of wood and grass are significantly
different. Grass has a high ash content but with the current fuel storrage and a 0w humidity (w < 20
weight-%). Native wood chips have a bw ash content but usually a higher humidity (w = 25% 55%).
Berk has a medium ash content and a high humidity (up to 60%). Waste wood can have high or bw
amount of ash or water, depending on its origin.
Under stoker furnaces can be used tor wood with a 10w ash content and a bower or higher humidity.
They usually reach up to 1 2 MW. Different types of grate furnaces are mostly used tor wood woth a
high ash content and a high water content, as they have a better ash removal and therefore a higher
operational range.
-

Herbaceous fuels

Grate furnace

Native wood, bark

Under stoker furnace

100

100
80

80

Ash
[weight-%]

60

Water

Ash

Water

[weight-%]

[weight-%j

[weight-%j

60

40

40

20

20

0
0

Figure 4.1

cP

Combustible

Combustible

[weight-%]

[weight-%]

Triangular diagram combustible water ash [Biollaz and Nussbaumer 1996].

left hand side: compostion of the fuel (wood and grass)
right hand side: operational range of under stoker and grate furnaces
-

14

4.3

Nitrogen content in the fuel

Figure 4.2 shows the frequency distribution of nitrogen content in grass, miscanthus, straw and wood.
The nitrogen content is increasing with an increasing protein content. The ranges of wood and
miscanthus are comparable (wood: 0.05 0.65 weight-%; miscanthus: 0.06 0.87 weight-%). Straw
has a significantly higher nitrogen content, which is in a fairly small range (0.4 1.0 weight-%). Grass,
including hay for feeding, have the highes nitrogen content in a large range (0.7 4.9 weight-%). The
treatment of the soil has a big influence on the nitrogen content of grass. As long as the soil is enriched
with nutrients, the grass has a high nitrogen content. A decrease of the nitrogen content in the grass
can only be expected after several years of extensive farming and after a decrease of the nutrients in
the soil.

60
.Grass
Miscanthus

50

0
S
traw

> 40
0

0
W
ood

30
20
cl)

10
0

ii

cJ-cDQcJj
d d d o

Figure 4.2

4.4

oc

CDO

cJ

cJ

CJ

N-content [weight-%]

C)

C)

0
C)

C)

C)

Frequency distribution of the nitrogen content of grass, miscanthus, straw and native
wood.

Content of sulfur, chiorine and potassium in the fuel

The sulfur content is dependent on a few proteins of the plant and therefore only covers a small
range.Chlorine and potassium are mainly assimilated as salts. Sulfur, chlorine and potassium from the
fuel form salt (KCI, 4
S0 at the combustion. The sum of potassium, chtorine and sulfur is taken to get
2
K
)
an indication on the amount of particies (safts) that are to expect. Figure 4.3 shows the frequency
distribution of the sum of potassium, chlorine and sulfur of grass, miscanthus, straw and native wood.
Compared to beech and coniferous trees, a higher content of K, Cl and S was found in fast growing
trees as eucalyptus or poplar (0.04
0.43 weight-%). The content is continuously increasing from
miscanthusto strawto grass (miscanthus: 0.11 1.1 weight-%, straw: 0.17 2.3 weight-%, grass 1.78
4.86 weight-%). Straw is a good example to show that these sdubstances are soluble in water. Straw
for atraw furnaces in Denmark is divided in yellow straw (that has not been rained on) and grey straw

15

(that has been rained on). The grey and dry straw has better combustion properties and is therfore
preffered for combustion.
Younger plants have a higher content of substances being needed for their growth, as they have a
higher number of cells still growing. Therefore a higher content of sulfur, chlorine and potassium can
be found at higher nitrogen content (see Figure 4.4).

70

Grass
i Miscanthus
Straw
c Wood

60
>

50

40
c3G)

30

4-

20

cl)

10
0
ID
CJ
0

ID

ID

LX)
Lf)
0
0
ID-QCJ
CJ

CJ

rocj

ID

IDCJ
CJ

ID

ID

OCLOC)
C)
c) c)

ID

0CJLf)

ID

r
.

0
0
1

Content of K+CIS [Gew.-%]

Figure 4.3

Frequency distribution of the sum of potassium, chlorine and sulfur for grass, miscanthus,
straw and natve wood.

5.0

.
.

4.0

.:.
.

3.0

:. : >

2.0

A %

Grass
Miscanthus
Straw

1.0

E
(1)

0.0
0.0

1.0

2.0

3.0

4.0

N-content [weight-%]
Figure 4.4

Correlation of nitrogen content and the sum of potassium, chlorine and sulfur tor grass,
miscanthus and straw.

16

Furnace Design and Process Control

5.1

Fundamentals of biomass combustion

The goal of combustion is to transform all of the chemical energy of wood into heat by oxidation. The
combustion of wood consists of several steps. At first the wood is dried by the heat in the combustion
chamber and decomposed to gases and charcoal by contact with the combustion gas. With the supply
of air the charcoal is then gasified to carbon monoxide. The gasses emitted by the wood and the
charcoal finally oxidize with the combustion gas in a gas flame.
For electric power and heat production the hot flue gasses pass their energy on to a working medium in
a heat exchanger, e.g. to water for producing steam. The heat can also be passed on to another gas in
a hot gas heat exchanger. The heated gas can then be used in a Stirling engine or in a closed cyclic
process to produce mechanical power.
With modern combustion technology it is nowadays possible to get an almost complete combustion of
the gases and thus an energy use that is poor in emissions.

Wood
(CHmOnNo)

Primary air (2 <1)

i.

Gasification

Combustible gases
CO, H
, 4
2
CH NH
,
, HCN, NO, etc.
1

Secondary air (2

<

Oxidation

1)

Flue gas
C0 HO, NOx
,
2

Figure 5.1

Two stage combustion of wood with primary air for gasification and secondary air for burn
out of the combustible gases [Nussbaumer 1989].

17

The combustion technology has little requirements to the fuel characteristics. The wood can be of
different sizes, even with a high content of fine material, and it can contain up to 60% of humidity. Even
impurities and contamination do not cause any problems, if the plant is professionally planned and
realized. The combustion technology tor plants of 20 kW to 100 MW is tully developed. Good
construction as weil as controlling and regulation of modern plants allows 10w emissions of unburnt
pollutants (carbon monoxide, organic carbons) even at changing bad. Very efficient technobogies exist
tor reducing nitrogen oxides, dust, and other pollutants resuiting from impurities in the wood, in the
exhaust gases. The limits for the measures for reducing emission are primarily given by the investment
and operating cost.
The amount of emitted pollutants is very 10w if a reasonable effort is made in emission control.
Especiaily dust and nitrogen oxide emissions are of importance. The additional cost of denitrification
and dedusting at larger plants is economically acceptable. The ash of the combustion of
uncontaminated wood can be used as a tertilizer. Residues from emission control have to be disposed
of. Waste water from plants with exhaust condensation has to be treated before being fed into the
sewage system.

Co

[-.1

Excess air ratio

Figure 5.2

Carbon monoxide emission as a function of the excess air ratio.

a)simple manuallycharged log wood boiler
b) downdraft boiler tor log wood
c)automatic furnace with combustion technology (as of 1990)
d) automatic furnace with enhanced combustion technology (as of 1995)
Automatic furnaces with appropriate combustion control technique can be operated at
Optimum conditions.

18

10

fl
9

%mNJ1

l2

8Ai

.I 1

\ 1

ttt

t
6

Figure 5.3

jI

:ti (I1I t

Sectional view of a modern log wood boller with downdraft principe

(TIBA, Bubendorf, SWITZERLAND).
1
3
5
8

Log wood storage, 2Drying zone

Pyrolysis and gasification zone, T = 150 600C, 4GIow bed, T = 800C
Primary air inlet, 6Secondary air inlet, 7 Oxidation zone, T = 600 1000C
Combustion chamber, 9Heat exchanger, 10 FIue gas exit, T = 180- 200C
-

19

5.2

Automatic wood furnaces

The most important designs for automatic wood furnaces can be distinguished as follows (Table 5.1):

Performance range

Type

Fuel

Water content
(wet basis)

Under stoker furnace

20 kW 2.5 MW

Native wood chips, waste wood chips

5%

50%

with a < 1%
Moving grate furnace

150 kW

100 MW

All kinds of wood, also

5%

60%

wirha1%
Preoven with grate

20 kW

1.5 MW

Dry waste wood chips

8%

35%

also witha> 1%
Bubbling fluidized bed furnace

Circulatingfluidized bed
furnace (CFB)
Dust burner

5 MW 15 MW

d < 10 mm

5%

15MW100MW

d< lOmm

5%60%

5 MW10 MW

d <5 mm

usually
<

Dust burner in coal power plant total 100 MW 1 GW

Wood: d <24 mm

Wood max. 10%

Table 5.1

60%

20%

usually
<20%

Overview of the most important designs of wood furnaces (a

ash content).

Several types of automatic and manually fed biomass furnaces exist. The most common furnaces used
tor wood chips are unders stoker and grate furnaces (Figure 5.4 and Figure 5.5). Under stoker furnaces
are built from 20 kWth to 2 MWth and they are suited tor fuels with 10w ash content and water contents on
wet basis up to approx. 50%. Moving grate furnaces are built from approx. 200 kWth to 20 MWth and
they can be used for fuels with higher ash content and water contents up to approx. 55%60%.
Furthermore dust burners, bubbling fluidized bed and circulating fluidized bed furnaces are used for
biomass combustion, especially for dry fuel which is grinded to small particle sizes. Due to high
investment cost, these furnaces are applied for larger plants (Figure 5.6, Figure 5.7, Figure 5.8 and
Figure 5.9).

20

Figure 5.4

Under stoker furnace:

1 fuel supply, 2 hearth, 3 primary air inlet,
4 secondary air inlet, 5 combustion
chamber, 6 heat exchanger, 7 insulation,
8 cyclone, 9 ash discharge.

Figure 5.5

Grate furnace:
1 fuel supply, 2 grate, 3 primary air inlet, 4
secondary air inlet, 5 combustion
chamber, 6 heat exchanger, 7 insulation,
8 cyclone, 9 ash discharge.

L
Brennstoff

Figure 5.6

Dust burner tor fine wood particies.

Figure 5.7 Bubbling fluidized bed combustor.

Gas.
1....

Gas

L45

FJ[
FUeI
A[

FIXED BED

BUBBLING

CIRCULATING

/t

LOI
(-p)

LOG (veloclty)

Figure 5.8

Principle of fixed and fluidized bed.

Figure 5.9 Circulating fluidized bed combustor.

21

5.3

Furnaces for herbaceous fuels

Hearbaceous fuel requires different fuel supply Systems than wood, especially straw. Straw Is mainly
used in Denmark, where lt is pressed into rectangular bales at the harvest. The straw bales can be burnt
as a whole, corresponding to a manually fed log wood boiler in its combustion properties. The alter
natives for a controlled combustion would be a cigar burner (see Figure 5.10) which is pushing the
whole bales the way that only their front face is burning or a turnace with bale cutter (see Figure 5.11),
that always cuts a part of the bale and continuously pushes lt on the grate. A good continuous
combustion, even tor smaller performances, can be achieved with this kind of furnace.

Figure 5.10 Cigar bumertor straw bales

Figure 5.11 Bale furnace with bale cutter (Linka)

The storage and the combustion of bales requires a specialized technology and logistic. The energy
density of bales is much less compared to wood chips and therefore the technology requires much
more space. An alternative to the combustion ot bales are pellets of herbaceous tuels. Pellets have a
higher energy density and can be used in moveing grate furnaces, as the handling is much easier. Due
to the high content in ash and to the high risk ot siagging, turnaces tor herbaceous tuels have a cooled
combustion chamber and a powerfull ash discharge system. Furthermore, measures to prevent
depostition or tor automatic cleaning have to be taken.

22

5.4

Process control

Elf iciency improvement for combustion processes demands bw fluegas temperature and bw oxygen
content. In biomass combustion, the amount of unburnt pollutants, i.e. CC, HC, PAH and soot, is
highly influenced by the excess air ratio. At excess air ratios below a certain Optimum value, the unburnt
pollutants rapidly increase by a factor of 10 or 100. With increasing excess air ratio above an optimal
value, the amount of unburnt pollutants slowly increases (see Figure 5.1 and Figure 5.3). Control
technique on biomass furnaces has progressed very much since the last ten years. While the first gen
eration of automatic wood furnaces were operating only at nominal heat output, the next generation
could also run at 50% 01 the nominal heat output. The standby hours were again remarkably reduced
by bad control technique being able to vary the heat output between 30% and 100% of the nominal
heat output. To operate at high efficiency and 10w emissions, an aciequate combination of bad and
combustion control technique was necessary [Nussbaumer 1989, Good 1992].
The most important combustion control techniques used in combination with bad controb are control of
the temperature in the combustion chamber (flame temperature control) and control of the excess air
ratio (lambda control). These control techniques dont use any information about the actual amount of
unburnt pollutants and therefore the settings of the combustion control techniques used today are
usually conservative, i.e. the setpoint tor the excess air ratio is rather high to avoid incomplete com
bustion. The Optimum excess air ratio is influenced by the furnace design and the operating
conditions, such as heat demand and fuel properties (humidity, specific weight). Experiments showed,
that the correlation between CO emissions and excess air ratio shifts to higher excess air with
decreasing heat output or with increasing fuel humidity (Figure 5.1). Therefore selftuning setpoint opti
mization algorithms were developed to find the optimum excess air ratio for different furnace desings
and changing operating conditions.
100000
]
3
[mg/m
lower heat output
and/or
higher tue! humidity

10000

co

x:_

1000

100
range 01
optimum setpoint

10
0

[-1

excess air ratio ?

Shift of the correbation between CO emissions and excess air ratio due to changing
operating conditions.
Because the operating conditions permanently change, the new generation of combustion control
technique should include information about the amount of unburnt polbutants to guarantee a high
efficiency and abow emission level without manual interaction. Information about CO emissions can be
used abone. However an increase of CO emissions is easier to interpret, if lambda or the flame
temperature is absoknown.
An advanced combustion control technique ACCT has been deveboped as an additionab control boop to
the existing combinations of bad and combustion control technique [Good and Nussbaumer 1998].
ACCT is based ona CO/Lambda control algorithm. With the use of additional information by measuring
Figure 5.1

23

CO emissions, the setpoint of the lambda control loop is permanently optimized. One possibility of
setpoint optimization is to reduce the lambda setpoint until an increase of CO emissions is detected.
Another is shown in Figure 5.3. Inside a certain range around the actual setpoint lambda control does
not change the secondary air. Data of lambda and CO emissions are permanently measured. With
exponential forgetting factors, new data are stronger weighted than older data. The shifting of the
correlation between CO emissions and excess air ratio due to changing operating conditions can be
detected and the setpoint will be adapted. Therefore ACCT maximizes the efficiency with respect to
the emissions of unburnt pollutants.
increase
secondaty air

Fuel Input
Primary
Air Supply

pont

ruvvuv

Load Control

PID [ry
Control
Air Supply

no action

decrease
secondary air

j
3
[mg/m
lflflO

Furnace

Co

co

$2

setpoInt

34

[-15

excess alr ratio

Figure 5.2 CO/Lambda control with setpoint opti

mization [Good & Nussbaumer 1998].

Figure 5.3 Principles of the controllers action.

ACCT was applied to a 1 MW understoker furnace equipped with bad control and flame temperature
control (Figure 5.4). Measurements of emissions and efficiency were done during a whole heating
period. CO/Lambda control and the setpoint optimization were first performed by reference
measurements of 02 and CO emissions with paramagnetic and infrared analyzers. In a second step
ACCT waspertormed by the use of 10w cost, in-situ sensors detecting CO emissions and excess air ra
tio. The gas sensor detecting CO emissions is a solid-state, semi-conductor sensor mainly composed
)which detects gases through an increase in electrical conductivity when
2
of sintered tin dioxide (SnO
reducing gases areabsorbed on the sensors surface, heated at 400 C The lambda sensor is based
on zirconium dioxide (ZrO
). At temperatures above 500 C the zirconium dioxide as a solid-state
2
electrolyte becomes aconductor of 02 ions.
.

Figure 5.4

Schematic of a 1 MW understoker furace equipped with ACCT [Good and Nussbaumer

1998] (TIBA-MUELLER AG).
24

Measurements on the 1 MW understoker furnace equipped with ACCT during a whole heating period
showed that the efficiency was above 90 % tor the whoe range ot the heat output. Especially at part
bad Operation the efficiency was improved by up to 5% compared to flame temperature control. CO
emissions averaged over one heating period were below 50 mg/Nm
3 which represents a reduction by a
factor of 5 compared to flame temperature control shows, that the selftuning setpoint optimization
leads to lower excess air ratios, thus resulting in higher efficiency and lower CO emissions.It has been
seen that there is no need tor accurate measurement of CO emissions and lambda. ACCT can there
tore be realized by application of 10w cost, in-situ sensors. They can be used tor control and monitoring
purposes over long-term periods without calibration and maintenance. Signal drifting of the sensors
has no negative etfect on the setpoint optimization algorithm. For typical fuel costs and an expected
etticiency improvement over the whole year of 2 to 4 % the investments tor ACCT are estimated to be
repaid in 2 to 5 years.
For biofuels with bw ash melting points such as grass, miscanthus or urban waste wood, ACCT can be
extended in order to avoid an exceeding ot the ash melting temperature.

100

1-

uJ

._

:fftLEEf60

t
a

Ten,p-Controt with COlLsnibda-Control

Ten,p.-Control

a 2-step bad controt, 1988

ZLE

0
0
1 flflfl
-

A Temp-Control with CO/Larnbda-Control

Temp.-Controb
!i 2-step bad controb, 1988

50___,______,____
40

30

50

60

70

80

100

90

30

Heat Output [%j

Figure 5.5

Figure 5.6

------
-

A Temp.-Control with CO/Lambda-Control

Temp.-Control
2-step bad control, 1988

-J

:-_J

11
30

40

50

60

70

80
Heat

Figure 5.7

90

50

60

70

80

90

100

Heat Output [%]

Etticiency versus heat output,

averaged over one heating period.

40

100

Output [%]

Excess air ratio Lambda versus heat

output, averaged over one heating
period.

25

CO emissions versus heat output,

averaged over one heating period.

Emissions and Depositions

Emissions are dependent on the fue, the design of the furnace, the contro of the furnace and on the
cleaning measures applied to the flue gas. A few emissions and possibilities tor their reductions are
presented.

6.1

Nitric oxide emissions

Nitric oxides from combustion processes are formed in three different reactions: Thermal NO is formed
at high temperature from the molecular nitrogen in the air, prompt NO can be formed during the
combustion of hydrocarbons and fuel NO,, are formed from the nitrogen caontained in the fuel. NO
emissions from biomass combustion originate mainly from fuel nitrogen, since combustion temperature
are usually below 1300 C and thermal NO,, is therefore of minor importance (see Figure 6.1).
Typical Biomass Corrdx,stion

Thermal NO and Prompt-NO

800
700

600
500
400
300
200

H.,b.c.oLi.

!f&.ao

100
0
T

Fuel NO,,, thermal NO,, and prompt NO,, versus combustion temperature
[Nussbaumer 1996].

Figure 6.1

To minimize NO by primary measures, the fuel nitrogen is reduced to molecular nitrogen in zones with
an excess air ratio of smaller than 1. The most promising primary measures for the reduction of NO
emissions are air staging, fuel staging and flue gas recirculation. NO reductions of 40% 50% for bw
and upto 75% tor high nitrogen content can be reached by air staging with a reduction chamber tor the
following conditions: primary excess aur ratio = 0.7 0.8, temperature 1,1000 1200C, residence
time 0.3 0.5 s. For air staging, flue gas recirculation is proposed to control the temperature and
advanced combustion control technmologies are used to ensure an Operation at optimum primary
excess air. However, air staging is not suited tor biofuels with bw ash melting point since the high
temperature can lead to siagging. Fuel staging shows a similar potential of NO,, reduction and it is mainly
feasable for barger plants [Nussbaumer 1996].
Possible secondary measures tor the reduction of NO emissions are the selective catalytic and non
catalytic reduction (SCR, SNCR) by injection of ammonia or urea. 60% 80% NO,, reduction can be
reached with SNCR. The NO,, reduction is Iimited by the ammonia slippage and an accurate process
-

26

control is necessary to ensure the temperature windiw of 840C 920C. Up to 95% NO reduction can
be ach ieved with SCR.
For large plants, combinations of 10w NO and denox-techniques are considered. Furthermore,
biofuels can be used as effective reburn fuels for the NO reduction in coa fired power stations.
The potential of different primary and secondary measures at different fuel nitrogen content is shown in
Figure 6.2.
-

2
NOx asNO

1000i

NO
3
[mg/Nm

2
atll%0

Conventional
ibustion
LOWNO with
Air Staging and
Reductlon Chamber

V/SCR

10
0.01

Grass
Straw
Urban Waste Wood
Miscanthus 4
UFChipBOardS
NativeWood

0.1

[wt.-%]

10

FueI-N
Figure 6.2

6.2

Comparison of NO emissions versus fuel nitrogen tor different measures.

Particle emissions

As the efficiency of particle separators and filters is dependent on the size of the particies, it is
important to know the particle size distribution of fly ash. The particle size distribution of fly ash was
determined with an Andersen cascade impactor for various fuels in different furnaces and under
different operating conditions in [Hasler and Nussbaumer 1997 & 1998].
Particle diameters are calculated as aerodynamic diameters (unit particle density of p = 1.0 kg/dm
). For
3
native wood, chipboard fuel, urban waste wood and hay, un[modal particle size distributions with a
mean diameter of less than 0.25 jm were found. More than 80% of the particle mass was found to be
aerosols with a diameter of smaller than 1im (see Figure 6.3). Bark is the only fuel found so far which
exhibits a bimodal particle size distribution. Flue gas recirculation Ieads to lower amounts of aerosol
particies in under stoker furnaces.
Fly ash from herbage-grass and wheat straw has been analysed closer with regard to composition and
shape and is tound to consist mainly of potassium chloride (KCI). The shape of the fly ash consists of
crystalline particies having the typical cubic shape of the main compound, KCI. This is in clear
Opposition to fly ash from wood, which appears in amorphous shape (see Figure 6.4).

27

Native wood

Irnbd 2.72
TotI 4451 0ontnt

75
=

Native wood / bark mixture

78 rng/Nr<3 (11% 02)

= 1.81
To511 d<st 005tent

75
=

72 rng/Nnt3 (11% 02)

50

z
E
25

25

1.)

(3

0
<0.22

0.22

0.35

0.61

1.28

2.03

3.01

4.48

0.21

7.22

8)

=
Q
0.21

150

Urban waste wood

ntbds 328
Totsl dont sontont

0.34

0.60

1.27

2.02

2.99

4.46

7.19

Aerodyn. particle diameter [sm1

Aerodyn. particle diameter [)Jm]

Hay pellets

206 rng/Nm3 )l 1% 02)

6mbd=a5
lolol dost oontont

125

100

300
=

250

200
150

50

100

25

<0.22

0.22

0.35

0.61

1.28

2.04

3.03

4.50

7.26

<0.26

0.26

Aerodyn. particle diameter [pmj

Figure 6.3

519 tng/Nnt3 111% 02)

0.41

0.71

1.48

2.35

3.47

5.16

8.30

Aerodyn. particle diameter [)im]

Particle size distribution during the combustion of native wood, a mixture of native wood
and bark, urban waste wood and hay peflets in a 450 kW moving grate furnace at full bad
[Hasler and Nussbaumer 1998].

(a)

(b)

Figure 6.4
Fly ash particies from herbage-grass (a), wheat
straw (b) and wood (c) collected on polycarbonate
filters [Kaufmann and Nussbaumer 1998].
(c)

28

6.3

Particle separation in a rotating particle separator

The Rotating Partiole Separator RPS [Brouwers 1995 & 1997] is a new design of a particle separator
which has been investigated for the use in biomass furnaces [Hasler et al. 1997 and 1998].

Top view ot the RPS

RPS with cyclone and rotating cylinder

Rotating cylinder with axial capillaries

Figure 6.5: Diagram of the RPS (tangential version) [Brouwers 1995]

The particle collection efficiency has been experimentally verified as a function of the dimensionsless
partiole diameter. The effective d
0 max was found to be higher by a factor of approximately 1.5 than the
calculated value d
1 The RPS will also separate smaller particles than the
it is designed tor, but
with a lower efficiency. In oase of the so-called tangential design appropriate tor flue gas filtering the
separation efficiency tor particles with d = 0.5
is approx. 70%, tor d = 0.25
the efliciency is
approx. 50%.
Cleaning of the particle baden cylinder channels is done by injecting compressed air from the top ot
the rotating filter element through a nozzle. Dedustirig can be made during filter Operation.
The nominal operating conditions for the RPS tested are as follows:

Filter element rotating speed 360 rad/s

Gas temperature 200C
Flow rate 1800 Am
/h
3

The calculated value tor d

0

max

is 1.8 jim.

29

In a 450 kW moving grate furnace, the particle collection efficiencies have been determined during the
combustion of saw dust (Figure 6.6). The RPS was operated at 350 radls (3300 rpm) and the flow rate
/h.
3
was 900 Am

100

0
>
0

80

*0

* Measurement na.

cl)
c

60

c)

40

oMeasurement no.

4J-_-_-______

0
0
0)
()

20

ccl

0
0.00

2.00

4.00

6.00

8.00

[-1
Figure 6.6: RPS particle collection efficiencies from test runs during saw dust combustion

A series of particle mass measurements has been made in the clean gas of the RPS with variing angular
velocitites of the filter element. Native wood chips were used as a fuel. The liow rate entering the RPS
/h. Raw gas particle concentrations were observed from 50 to 120 mg/Nm
3
was 900 Am
3 in previous
tests. Although the residual particle emissions exhibit some variations, there is a tendency of lower
emissions with increasing angular velocity. In total 11 measurements were made and two values
exceeded the target dust concentration of 50 mg/Nm
3 (at 11% 02).

70

target value

60
50
0

40

00
0

0)
0
0

cl)
(.)
(cl

30
0

20

10
0

200

250

300

350

400

450

Angular velocity [rad/s]

Figure 6.7: Residual particle emissions after RPS during the combustion of native wood chips in a 450
kW moving grate furnace

30

6.4

Chlorine emissions

As the combustion of herbage-grass, miscanthus, hay and straw causes higher emissions of particies,
more deposition is encountered in the boiler and in the heat exchanger. The high particie emissions
mailny resuit from the volatile minerals contained in grass and straw. Suifur is emitted as a salt 4
S0 or
2
(K
)
as SO
2 and SO
. if the combustion is incompiete sulfur is also emitted as H
3
S. Chiorine can occur in the
2
fiue gas in the form of sait (KC1, NaC1) or as HC1 as weil as in dioxines and organic chiorine compouds.
Own measurements and data from literature show that most of the chiorine is emitted as salt. The
concentration of HCI is usuaiiy between 20 and 120 mg/Nm
3 (see Figure 6.8). lt is assumed, that the
10w HCI concentrations result from the high concentrations of K and Na, that react with the chiorine.

120

100

(1)

Herbage-gr.
Timo1he
Feedgr. mix
W. straw
> Hemp
OMisc.
Sedge

80
ci,

60

0.1

v
ci,

40

0
0

20

0.01
0

0.005

0.01
0.015
Fuel chiorine content [kg/kg DS]

Figure 6.8

0.0025 0.005 0.0075

0.01

Fuel chiorine content [kg/kg DS]

HCI emissions as a function of the chiorine content in the fuel (left hand side) and the
transfer rate of fuel chiorine to HCI (right hand side) [Kaufmann 1997]

31

calculated tor 11 vol% 02

5000

Dust
[mg/Nm
]
3

Area sum of contents

4000

3000

Grass
Straw
Miscanthus
Wood

2000

Measured dust concentrations

in raw gas
Grass

1000

Slraw

111111111

Sum of K, Cl, S content in fuel [weight-%]

Figure 6.9

Dust content in the raw gas (flue gas after the cycione) tor combustion of grass, straw and
miscanthus. The maximum concentration of salts for the complete emission of potassium,
chiorine and sulfur in the tlue gas is given tor comparison.
*

Deposition, condensation, crystallization

Figure 6.10 Path of potassium, chlorine and sulfur ot the fuel in the furnace.

6.5

Deposit formation

During the combustion of herbage-grass (hay), miscanthus and straw in a grate furnace, intensive
deposit formation was observed on the combustion chamber walls (brick) and on the boiler walls (steel).
Although it is obvious, that the deposits are caused by the fly ash, the formation mechanisms of de
positions are not yet known in detail. The aim of the present investigation was to analyze the chemical
and morphological properties, to identifiy the mechanisms of deposition formation, and to propose
measures for deposition reduction. Depositions from bioniass are accumulations of fine particles in the
32

nanometer range, whose chemism is dominated by the properties of potassium chiorine, the major
compound found in the deposition. Among the known formation mechanisms, thermophoresis
revealed to be the crucial process for depositions from biomass combustion. The resuits of the
investigation are presented in [Kaufmann et al. 1998]. Table 6.1 gives an overview of the composition
of the different depositions found in the Multif Fuel Combustor at Sandia and in a grate furnace
operated in Switzerland.
Table 6.1

Composition of the sticky deposition on a single cylinder in the plug flow reactor compared
to the composition of deposits in the bauer of a 450 kWth grate furnace in [wt.-%]
MFC

MFC

Grate furnace

Fieldgrass

Miscanthus

Fieldgrass

Sticky deposit

Sticky deposit

Fuel
Deposition

End of pipe

Beginn of pipe

ER

121.5

22.3

121.5

22.3

22.3

22.3

Dim.

[wt.-%]

[wt.-%]

[wt.-%]

[wt.-%]

[wt.-%]

[wt.-%]
53.9

27.72

30.1

10.3

iZi

46.8

6.76

1.94

11.5

1.45

12.32

2.7

CI

12.4

13

8.37

24.7

33.4

Si

18.25

10.0

21.7

20.5

1.09

0.14

Ca

9.58

7.51

4.57

1.48

0.16

Na

0.82

1.16

11.5

1.13

0.660

0.71

4.62

2.59

1.39

1.49

1.05

0.07

Mg

2.82

2.15

1.13

1.16

0.316

0.02

Ti

0.086

0.07

0.04

0.04

0.069

<0.01

Fe

123

0.7

1.12

126

1.29

<0.01

Al

0.84

1.32

0.8

0.79

0.40

0.16

0.13

0.4

0.48

0.75

0.066

0.016

Sum:

85.6

71.7

67.7

55.6

90.4

91.4

Figure 6.11 FIy ash deposition on bauer surfaces of the grate furnace
[Kaufmann and Nussbaumer 1998].

33

6.6

Dioxins and furans

Polychlorinated dioxins and furans (PCDD and PCDF) have been determined during the combustion of
wood and herbaceous biofuels in different furnaces (Nussbaumer and Hasler 1996]. PCDD/F are
found to be a consequence of the de novo synthesis in the temperature window between 180C and
500C. Carbon, chiorine, cataiysts (copper) and oxygen are necessary for the formation of PCDD/F.
Emissions from non contaminated wooci were usually below 0.1 ng TE/Nm
3 at 11 vol.-% 02, while from
3 were found, ranging from 0.03
the combustion of urban waste wood emissions of app. 2 ng TEINm
. The combustion of charcoal in a wood stove resulted in the very bw PCDD/F emissions
3
18 ng TE/Nm
, while combustibbe household waste Iead to extremely high emissions. Although
3
of 0.028 ng TE/Nm
. This is
3
herbaceous biofuels have high chiorine contents, PCDD/F are usualiy below 0.1 ng TEINm
explained by the high alkali content, which Ieads to the formation of saits (KCI, NaCI) and to a lower level
of gaseous chiorine for the de novo synthesis.
Primary measures to avoid dioxin emissions are a complete burn out of the fly ash and an Operation of
the combustion at 10w air excess (02 < 11 vol.-%, prefereably <6%) and under stable conditions. For
automatic wood furnaces an efficient particle Separation iS needed and the filter must be operated weil
below the temperature of the de novo synthesis. Furthermore adsorptives such as activated lignite can
2 and PCDD/F in filters. However illegal incineration of
be used for a combined extraction of HCI, SO
wastes in wood furnaces must be strictly avoided.

dust burner
moving grate
understoker
preoven

log wood boiler

ooen chimney
wood stove

0.00

0.20
0.10

0.40
0.60
0.80
PCDD/F emission [ng TEINm
]
3

1.00

Figure 6.12 PCDD/F emissions in the clean gas from the combustion of uncontaminated wood at 11
vol.-% 02, 1-TEF acc. to NATO-CCMS [Nussbaumer and Hasler 1996].

34

References

Baxter, L.; Miles, Th. ; Miles, Th. (Jr.); Jenkins, B.; Dayton, D.; Milne, Th.; Bryers, R.; Oden, L.: Alkali
Deposits found in Biomass Bollers, Sandia National Laboratory, Livermore (CA, USA) 1996
Baxter, L.L., Jenkins, B.M., Miles, T.R. e.a. (1994). Alkalis in alternative biofuels. FACT, Vol. 18,
Combustion Modeling, Scaling and Air Toxins, ASME
Biollaz, S.: Messen des Verweilzeitspektrums in der Nachbrennkammer von Holzfeuerungen zur
Modellierung der Kohlenmonoxid-Oxidation, Diss ETH Nr. 12383, Zrich 1997
Biollaz, 5.; Nussbaumer, Th.: Einsatz von Rostfeuerungen fr Holz und Halmgter. In: Nussbaumer,
Th. (Hrsg.): Feuerungstechnik, Ascheverwertung und Wrmekraftkopplung, 4. Holzenergle-Sym
poslum, 18.10.1996, ETH Zrich, ENET, Bundesamt fr Energiewirtschaft, Bern 1996, 941
Brouwers, J.J.H. (1995). Rotational particle separator: A New Method for Separating Fine Particles and
Mists from Gases. In: Chem. Eng. & Technol. 19, 1-10, 1996
Brouwers, J.J.H. (1997). Particle Collection Efficiency of the Rotational Particle Separator. In: Powder
Technology, 92, 89-99
Browning, B.L.: The Chemistiy of Wood, Interscience Publisher, John Wiley & Sons, New York and
London 1963
Bruch, C.; Nussbaumer, Th.: CFD Modeling of Wood Furnaces. Biomass forEnergy and Industty. lOth
European Conference and Technology Exhibition, June 8 111998. Wrzburg, Germany.
Flarup, J.: Ballenfeuerungen fr Halmgter. In: Nussbaumer, Th. (Hrsg.): Feuerungstechnik, Asche
verwertung und Wrmekraftkopplung, 4. Holzenergie-Symposium, 18.10.1996, ETH Zrich,
ENET, Bundesamt fr Energiewirtschaft, Bern 1996
Good, J. (1992). Verbrennungsregelung bei automatischen Holzfeuerungen, Ph.D. Thesis ETH Nr.
9771, Zurich.
Good, J.; Nussbaumer, Th.: Efficiency Improvement and Emission Reduction by Advanced
Combustion Control Technique (ACCT) with CO/Lambda Control and Setpoint Optimization.
Biomass for Energy and lndustry. lOth European Conference and Technology Exhibition, June 8
111998. Wrzburg, Germany.
Gronli, M.: A Theoretical and Experimental Study of the Thermal Degradation of Biomass. The Norwe
gian University of Science and Technology, Trondheim 1996
Hasler, P.: Modern log wood boiler with fuzzy logic control. Biomass for Energy and lndustry. lOth
European Conference and Technology Exhibition, June 8 111998. Wrzburg, Germany.
Hasler, Ph.; Nussbaumer, Th.: Landwirtschaftliche Verwertung von Aschen aus der Verbrennung von
Gras, Chinaschllf, Hanf und Stroh. Bundesamt fr Energiewirtschaft, Bern 1996
Hasler, Ph.; Nussbaumer, Th.: Partikeigrssenverteilungen von Flugstaub bei der Verbrennung und
Vergasung von Biomasse, Bundesamt fr Energiewirtschaft, Bern 1997
Hasler, Ph.; Nussbaumer, Th.: Particle Size Distribution of the Fly Ash from Biomass Combustion.
Biomass tor Energy and lndustiy. lOth European Conference and Technology Exhibition, June 8
111998. Wrzburg, Germany.
Hasler, Ph.; Nussbaumer, Th.; Schaffner, H.P.; Brouwers, J.J.H.: Reduction of Aerosol Particles in
Flue Gases from Biomass Combustion with a Rotational Particle Separator RPS. Biomass tor Energy
and !ndustry. lOth European Conference and Technology Exhibition, June 8 111998. Wrzburg,
Germany.
-

35

Hofbauer, H.: Charakterisierung von biogenen Brennstoffen und Verwertung von Holzaschen. In Th.
Nussbaumer (Hrsg): Neue Erkenntnisse zur thermischen Nutzung von Holz, 3. Holzenergle
Symposium, ETH Zrich, Bundesamt fr Energiewirtschaft, ENET, Bern 1994
Kaufmann, H. (1997). Chlorine compounds in emissions and residues from the combustion of herba
ceous biomass. PhD 12429, Swiss Federal Institute of Technology, Zrich
Kaufmann, H., Nussbaumer, T., Baxter L.L., Young, N.. Deposition formation on a single cylinder
during combustion of fieldgrass, miscanthus and pine (in preparation)
Kaufmann, H.; Nussbaumer, Th.: Characteristics and Formation of FIy Ash Particles in Biomass
Furnaces. Biomass tor Energy and Industnj. lOth European Conference and Technology
Exhibition, June 8 111998. Wrzburg, Germany.
Kaufmann, H.; Nussbaumer, Th.; Baxter, L.; Young, N.: Mechanisms of Deposition Formation during
Biomass Combustion. Biomass tor Energy and lndustry. lOth European Conference and
Technology Exhibition, June 8 111998. Wrzburg, Germany.
Keller, R.: Primrmassnahmen zur NQ-Minderung bei der Holzverbrennung mit dem Schwerpunkt der
Luftstufung, Diss ETH Nr. 10514, Zrich 1994
Nussbaumer, Th., Hasler, P.: Dioxin emissions in biomass combustion. In A.V. Bridgwater and D.G.B.
Boocock (Eds.): Developments in Thermochemical Biomass Conversion, Blackie Academic,
Chapman & Hall, London 1997, ISBN 0 7514 0350 4, 14921506
Nussbaumer, Th., Hustad, J.: Overview on biomass combustion. In A.V. Bridgwater and D.G.B.
Boocock (Eds.): Developments in Thermochemical Biomass Conversion, Blackie Academic,
Chapman & Hall, London 1997, ISBN 0 7514 0350 4, 12291 243
Nussbaumer, Th.: Biomass Combustion: Basic Principles and Application of Grate-Furnaces, Under
stoker Furnaces and Dust Combustors, Biomass Summer School, Schloss Stein, Styria (Austria),
29.7. 2.8. 1996
Nussbaumer, Th.: Primary and secondary measures for NO reduction in biomass combustion. In A.V.
Bridgwater and D.G.B. Boocock (Eds.): Developments in Thermochemical Biomass Conversion,
Blackie Academic, Chapman & Hall, London 1997, ISBN 0 7514 0350 4, 14471461
Nussbaumer, Th.: Schadstoffblldung bei der Verbrennung von Holz, Diss ETH Nr. 8838, Zrich 1989
Nussbaumer, Th.: Verbrennung und Vergasung von Energiegras und Feldholz. In: J.L. Hersener
(Red.): Schlussbericht Projekt Energiegras/Feldholz, Bundesamt fr Energiewirtschaft, Bern 1 997,
42 72
Nussbaumer, Th.; Good, J.; Jenni, A.; Koch, P.; Rutschmann, Ch.: Projektieren automatischer Holzteuerungen, Pacer, BfK, Bern 1995, 191 S., ISBN 3-905232-63-4
Salzmann, R.; Meuli, M.; Good, J.; Nussbaumer, Th.: Einsatz von CO-Sensoren zur Verbrennungsre
gelung. Heizung Klima, Nr. 3 1998, AT Verlag Aarau
Salzmann, R.; Meuli, M.; Good, J.; Nussbaumer, Th.: Regelung von Feuerungsanlagen mit Messung
von CO und Lambda. Heizung Klima, Nr. 2 1998, AT Verlag Aarau
Schweizer, P.: Fuzzy Logic zur Regelung von Stckholzfeuerungen. In: Nussbaumer, Th. (Hrsg.):
Neue Erkenntnisse zur thermischen Nutzung von Holz, 3. Holzenergie-Symposium, ETH, Zrich,
Bundesamt fr Energiewirtschaft, Bern 1994
Spliethoff, H.; Siegle, V.; Hein, K.: Zufeuerung von Biomasse in Kohlekraftwerken: Auswirkungen auf
Betrieb, Emissionen und Rckstnde, Technik und Kosten. In: Nussbaumer, Th. (Hrsg.): Feue
rungstechnik, Asche verwe rtung und Wrme-Kraft-Kopplung, 4. Holzenergie-Symposium, ETH
Zrich, 1996
-

36