Emissions of Residential Wood Combustion in - Aalto

Aalto University
School of Science
Degree Programme in Engineering Physics and Mathematics
Ville-Veikko Paunu
Emissions of Residential Wood Combustion in Urban and Rural Areas of
Finland
Master’s Thesis
Espoo, May 23 2012
Supervisor:
Instructors:
Professor Peter Lund
Niko Karvosenoja D.Sc. (Tech.)
Mikko Savolahti M.Sc. (Tech)
Aalto University
School of Science
ABSTRACT OF
Degree Programme in Engineering Physics and Mathematics MASTER’S THESIS
Author:
Ville-Veikko Paunu
Title:
Emissions of Residential Wood Combustion in Urban and Rural Areas of Finland
Date:
Professorship:
Supervisor:
Instructors:
May 23 2012
Pages: vii + 63
Energy Sciences
Code: Tfy-56
Professor Peter Lund
Niko Karvosenoja D.Sc. (Tech.)
Mikko Savolahti M.Sc. (Tech)
Particulate matter (PM) is a significant threat of air pollution to human health
in Finland and Europe. Residential wood combustion is a major cause of PM
emissions. Therefore, the control of PM emissions is one of the most important
challenges related to air quality.
The goal of this thesis was to identify the characteristics of Finnish residential
wood combustion, study the PM2.5 emissions and the population exposure they
cause from different residential area types, and assess the emission reduction
options for the future.
The total PM2.5 emissions from residential wood heating was 8230 Mg a-1 , which
amounted to 26% of the total emissions in Finland in 2005. Supplementary wood
heating, i.e. stoves and masonry heaters, caused 70% of these. Non-urban areas
were responsible for 57% of the total emissions.
Supplementary heating caused 89% of the total PM2.5 exposure from RWC, with
80% of the total exposure coming from urban areas. In total, RWC was estimated
to have caused around 200 premature deaths in 2005. From population exposure
point of view, supplementary wood heating with stoves and masonry heaters was
much more significant than primary heating with boilers.
Since the RWC is increasing in the future, the reduction of the emissions is
important for the public health. The reduction measures should be targeted at
supplementary wood heating in urban areas, as it comprises the majority of the
population exposure. A viable way to reduce the emissions in the short term
could be to affect the operational practices through an information campaign or
legislation measures. A slower method is the renewal of the appliances with newer
models with lower emissions.
Keywords:
Residential wood combustion, masonry heater, sauna stove,
emissions, particulate matter, population weighted concentration, population exposure
Language:
English
ii
Aalto-yliopisto
Perustieteiden korkeakoulu
Tietotekniikan tutkinto-ohjelma
Tekijä:
Työn nimi:
Puun pienpolton
Päiväys:
Professuuri:
Valvoja:
Ohjaajat:
DIPLOMITYÖN
TIIVISTELMÄ
Ville-Veikko Paunu
päästöt kaupunki- ja maaseutualueilla Suomessa
23. toukokuuta 2012
Sivumäärä: vii + 63
Energiatieteet
Koodi:
Tfy-56
Professor Peter Lund
Niko Karvosenoja D.Sc. (Tech.)
Mikko Savolahti M.Sc. (Tech)
Hiukkaset on merkittävä terveysuhka Suomessa ja Euroopassa. Puun pienpoltto
on suuri hiukkaspäästölähde. Hiukkaspäästöjen hallinta onkin yksi tärkeimmistä
ilmanlaadun haasteista.
Tämän diplomityön tarkoitus oli määrittää Suomalaisen puun pienpolton
ominaispiirteet, tutkia pienpolton pienhiukkaspäästöjä ja niiden aiheuttamaa
väestöaltistusta eri asuinaluetyypeillä ja arvioida päästövähennyskeinoja tulevaisuudessa.
Kokonaispienhiukkaspäästöt puun pienpoltosta olivat 8230 Mg a-1 , joka oli 26%
Suomen kokonaispäästöistä vuonna 2005. Lisälämmitys, eli varaavat takat ja uunit, aiheuttivat 70% päästöistä. Maaseutuasutus vastasi 57% kokonaispäästöistä.
Lisälämmitys aiheutti 89% puun pienpoltosta johtuneesta kokonaispienhiukkasväestöaltistuksesta, ja 80% altistuksesta johtui kaupunkialueiden päästöistä.
Puun pienpolton arvioitiin aiheuttaneen 200 ennenaikaista kuolemaa vuonna
2005. Väestöaltistuksen näkökulmasta lisälämmitys varaavilla takoilla ja uuneilla
oli huomattavasti merkittävämpää kuin päälämmitys boilereilla.
Koska puun pienpoltto on lisääntymässä tulevaisuudessa, päästöjen
vähentäminen on tärkeää kansanterveyden kannalta. Vähennystoimet tulisi
kohdistaa kaupunkialueiden lisälämmitykseen, sillä se aiheuttaa suurimman
osan väestöaltistuksesta. Toteuttamiskelpoinen päästövähennyskeino lyhyellä
aikavälillä voisi olla laitteiden käyttötapoihin vaikuttaminen valistuskampanjoilla tai lakisäädäntötoimilla. Hitaampi keino on laitteiden korvaaminen uusilla,
vähäpäästöisillä malleilla.
Asiasanat:
Puun pienpoltto, päästöt, varaava takka, puukiuas, hiukkaset,
altistuminen
Kieli:
Englanti
iii
Acknowledgements
I wish to thank my instructors Niko Karvosenoja and Mikko Savolahti for
their superb guidance and support. They helped tremendously in the tight
schedule. Furthermore, I wish to thank the Director of the Consumption
and Production Centre Jyri Seppälä and the Head of the Environmental
Performance Unit Kimmo Silvo for the change to do my master’s thesis in
the Finnish Environment Institute, and all my colleagues for their support.
A special thanks to the supervisor of this thesis, Professor Peter Lund.
I would also wish to thank my family for their wonderful support and care
through my studies and life. A final thanks to my dear friend Lassi Ahlvik
for his comments on the thesis and life in general.
Espoo, May 23 2012
Ville-Veikko Paunu
iv
Abbreviations and Acronyms
CAFE
CMH
EC
FRES
MMH
NC
NMVOC
OC
OGC
PAH
PM
PMx
POM
PWC
RWC
SC
SVOC
VOC
VVOC
YKR
European Commission’s Clean Air for Europe program
Conventional masonry heater
Elemental carbon
Finnish Regional Emission Scenario model
Modern masonry heater
Normal combustion
Non-methane volatile organic compounds
Organic carbon
Organic gaseous compounds
Polyaromatic hydrocarbons
Particulate matter
Particulate matter, with particles which aerodynamical diameter is smaller than x µm
Particulate organic matter
Population Weighted Concentration
Residential wood combustion
Smouldering combustion
Semivolatile organic compounds
Volatile organic compounds
Very volatile organic compounds
Urban structure monitoring system
v
Contents
Abbreviations and Acronyms
v
1 Introduction
1
2 Emissions of Finnish Residential Wood Combustion
ances
2.1 Health Effects of Particulate Matter . . . . . . . . . . .
2.2 Finnish Residential Wood Heating Appliances . . . . .
2.2.1 Masonry Heaters . . . . . . . . . . . . . . . . .
2.2.2 Sauna Stoves . . . . . . . . . . . . . . . . . . .
2.2.3 Residential Wood Boilers . . . . . . . . . . . . .
2.2.4 Other Appliances in Finland . . . . . . . . . . .
2.3 Emissions of Air Pollutants . . . . . . . . . . . . . . .
2.3.1 Wood as Fuel . . . . . . . . . . . . . . . . . . .
2.3.2 Combustion Process . . . . . . . . . . . . . . .
2.3.3 Batch and Continuous Combustion . . . . . . .
2.3.4 Wood Combustion Emissions . . . . . . . . . .
2.3.5 Dispersion . . . . . . . . . . . . . . . . . . . . .
2.3.6 Factors Affecting Emissions . . . . . . . . . . .
2.3.7 Emissions from Finnish Appliances . . . . . . .
2.3.8 International Comparison . . . . . . . . . . . .
2.4 Emission Reduction Possibilities . . . . . . . . . . . . .
2.4.1 Advanced Combustion Technologies . . . . . . .
2.4.2 Combustion Practices and Legislation . . . . . .
2.4.3 Flue gas cleaning . . . . . . . . . . . . . . . . .
Appli.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3
3
5
6
8
9
11
13
13
15
16
17
19
20
21
23
24
25
26
27
3 Modelling of Population Exposure to Fine Particles from residential wood combustion
30
3.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.1.1 Emission Modelling . . . . . . . . . . . . . . . . . . . . 30
vi
3.1.2
3.1.3
Dispersion Modelling and Population Exposure Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Residential Areas . . . . . . . . . . . . . . . . . . . . . 39
4 Results
43
5 Discussion
5.1 Emissions and Spatial Distribution . . . . . . . . . . . . . . .
5.2 Population Exposure . . . . . . . . . . . . . . . . . . . . . . .
5.3 Emission Reduction . . . . . . . . . . . . . . . . . . . . . . . .
52
52
53
55
6 Conclusions
57
vii
Chapter 1
Introduction
Residential wood combustion (RWC) is a major cause of particulate matter
(PM) emissions. PM is a major threat of air pollution to human health
in Finland and Europe (Pekkanen, 2010). Therefore, the control of PM
emissions is one of the most important challenges related to air quality.
There are numerous different types of RWC appliances, all of which have
their own combustion and emissions characteristics. Devices that are the
main heating means for a house are used with continuous fuel feeding. Other
devices, which are used for supplementary heating or aesthetic effect, are
loaded batch per batch. This affects the controllability of the combustion
and, therefore, the emissions. Another important factor for the emissions
is operational practices. If optimal combustion conditions are not met, the
emissions can be multifold.
The particulate emissions from wood combustion have adverse health
effects. They cause premature deaths and restricted activity days, all of
which have serious social and economic impacts. Particle emissions from
RWC have typically strong local effects, so it is essential to know which kind
of surroundings the particles are emitted in. For example, high emissions in a
scarcely populated area might mean lower population exposure than smaller
emission in a dense residential area.
Residential wood combustion comprises 25% of anthropogenic PM2.5 emissions in Finland (Karvosenoja et al., 2008). Wood combustion usually acts as
a supplementary heat source in single-family houses. In Helsinki Metropolitan Area most of the wood is used in masonry heaters for supplementary
heating and in sauna stoves (Gröndahl et al., 2011). Primary wood heating,
i.e. houses that are primarily heated with wood, is less common in urban
than rural areas. Supplementary heating is often intermittent, and batch
combustion is the most common type. Primary heating, on the other hand,
is usually continuous, and is often done by continuous combustion.
1
This thesis addresses the following questions: (1) what are the characteristics of Finnish residential wood combustion and how do they affect the
particulate emissions? (2) How do RWC particulate emissions and population exposure differ in different types of areas in Finland? The study was
outlined to concentrate on fine particles, as they are the most important
emissions from RWC. In order to answer these questions, a literature review on RWC emissions measurements was conducted. The PM emissions
and population exposure were modelled using the Finnish Regional Emissions Scenario (FRES) model. The emissions and exposures from different
residential area types were compared. Possible reduction options and their
allocation was discussed.
2
Chapter 2
Emissions of Finnish Residential
Wood Combustion Appliances
The most important emissions from residential wood combustion (RWC) are
particulates. Particulate matter (PM), especially fine particles, have adverse
health effects on humans. They also have effects on the climate. Most
particulates have a cooling effect, as they reflect sunlight. However, black
carbon, which is an important component of particles from RWC, absorbs
sunlight and acts as a warming agent. The research of the health effects has
longer history than of the climate effects. This thesis concentrates on the
health effects on humans.
2.1
Health Effects of Particulate Matter
The health effects of particulate matter have been studied for decades. The
conclusion in the last decade has been, that the particulate matter emissions
are the most significant cause of negative health effects that come from the
environment (Salonen and Pennanen, 2006). The most harmful are particles with diameter that is smaller than 2.5 µm (PM2.5 ). These so called
fine particles penetrate deep into the lungs and the smallest can even enter
bloodstream.
Exposure to particulate matter cause numerous health effects. These
range from almost unnoticeable symptoms to serious diseases and death.
Milder symptoms are sore nose and throat. More serious and long-lasting
diseases include coronary disease, chronic obstructive pulmonary disease,
asthma and lung cancer, which develop slowly. Premature deaths are usually
caused by long-term exposure.
The health effects of particle matter depend on how deep into the system
3
Table 2.1: Some harmful effects of different particle sizes. (Salonen and Pennanen, 2006)
Particle size
Short-term exposure
Long-term exposure
Asthma and chronic obstruc- Chronic obstructive
Coarse particles
tive pulmonary disease worsen pulmonary disease?
(2.5-10 µm)
Respiratory infections
Asthma and chronic obstruc- Chronic obstructive
tive pulmonary disease worsen pulmonary disease
Fine particles
Coronary disease and cere- Atherosclerosis in(<2.5 µm)
brovascular disease worsen
tensifies
Respiratory infections
Lifetime shorten
Asthma?
Allergy?
Ultrafine particles Asthma worsen
No research results
(<0.1 µm)
Coronary disease worsen
they are able to penetrate. Harmful effects associated with different particle
sizes are presented in Table 2.1. Most of the biggest (>10 µm) particles are
filtered in the upper respiratory track. Particles smaller than that find their
way easier to the lungs and alveoli. From there, the ultrafine particles (<0.1
µm, PM0.1 ) can penetrate the bloodstream and further into the heart and
other organs. Particles cause damage to the cells, and the type of damage
depends on the structure of the cell. Possible effects include increased cell
death rate and increased inflammation risk, as well as physical damages, such
as ruptures and blockage in veins. (Salonen and Pennanen, 2006)
European Commission’s Clean Air for Europe (CAFE) program estimated, that fine particles caused 350 000 premature deaths within the population of EU (450 million) in 2000, shortening the lifespans on average by 8.1
months (CAFE, 2005). However, for certain groups with health problems
the shortening can be even 10 years. In addition, fine particles increased
medical and hospital treatments, and caused restricted activity days for tens
of millions. The annual economical costs were estimated to be 270-280 billion euros. In Finland, which was the cleanest country in the report, it was
estimated, that fine particles caused 1300 premature deaths and 600 chronic
bronchitises in the year 2000. On top of that, tens of thousands of people
suffering from cardiovascular diseases and also children were estimated to
suffer from restricted activity and increased medication use. Economic losses
were assessed to have been 1-2.9 million euros annually. It is notable that the
4
CAFE mainly estimated the effect of long-range transportation pollution. On
the PILTTI project (Ahtiniemi et al., 2010) it was estimated that RWC and
traffic emissions from Finland causes around a thousand premature deaths.
No threshold exposure for health effects have been found. Even exposure
to small concentration for a long time can have significant consequences, since
the amount of several particle types and, therefore, the effect in the system
is cumulative. This means that they worsen chronic cardiovascular diseases
all the time. Furthermore, high daily doses are believed to affect even after
one to two months after the exposure.
2.2
Finnish Residential Wood Heating Appliances
Typical Finnish RWC applications are masonry heaters and different stove
types. Finland has approximately 2.2 million small-scale wood-burning devices and, in addition, 1.5 million wood-burning sauna stoves (Tissari et al.,
2008a). In 2009, Finnish residential buildings used 51.1 PJ of wood fuel (Statistics Finland, 2010). On average, the annual RWC activity increase during
the 2000s has been 4% (figure 2.1) (Statistics Finland, 2010).
Figure 2.1: Fuel wood use in residential buildings in Finland.
Finland, 2010)
5
(Statistics
Figure 2.2 shows the residential wood fuel use by combustion appliance
type in 2008 in Finland. The biggest portion of wood was used in masonry
heaters and ovens, with log boilers second and sauna stoves third. While
the number of log boilers is smaller than masonry heaters and sauna stoves,
they are used for primary heating. Therefore, the wood use per appliance is
higher.
Figure 2.2: Fuel wood per appliance type use in residential buildings in
Finland in 2008 (based on (Torvelainen, 2009)).
The small-scale wood combustion appliances and burning habits in Finland differ significantly compared to other countries. In many places in Central Europe, for example, the aim is to produce heat slowly over a long period
of time, which means that the combustion is slow and the emissions high, if
heat accumulator is not used. In Finland, the combustion rate is high and
the operating time is short. (Tissari et al., 2008a)
2.2.1
Masonry Heaters
Tissari et al. (2008b) defines masonry heaters as “enclosed combustion appliances made of masonry products, a combination of masonry products or
soapstone”. They have a high mass (typically 800-3000 kg, but even up to
6000 kg), and usually an upright firebox, from which the exhaust gas flows
to an upper combustion chamber, then down through the ducts and from
the bottom or top of the heater in to the chimney. Figure 2.3 shows how
the air is taken into the combustion chamber. The large mass of the heater
efficiently stores the heat and releases it slowly into the surroundings (at the
average rate of 1-3 kW) in a period ranging from 10 hours to two days. Most
new detached houses in Finland are equipped with a masonry heater.
6
Figure 2.3: A modern masonry heater. The arrows indicate the intake
air. (Alakangas et al., 2008b)
The combustion process of masonry heaters differs from stoves and conventional fireplaces. They have qualities that reduce the emissions. The
surfaces of the firebox are hot and closed. The surfaces reflect heat back
into the flame, creating the gas turbulence that complete combustion needs.
Furthermore, secondary combustion is good in masonry heaters due to secondary combustion chamber, and efficiency is high because of the large mass.
Compared to open fireplaces masonry heaters’ air intake size is restricted and
the operating temperature is higher. (Tissari, 2008; Tissari et al., 2008a)
Masonry heaters can be divided into conventional masonry heaters (CMH)
and more developed modern masonry heaters (MMH). Modern masonry
heaters have some advantages compared to conventional ones (Tissari, 2008).
The primary airflow is controlled and there are small air inlet holes in the
grate. These holes have several advantages: the secondary air is spread to
surround the fuel batch, so mixing of air and combustion gases is increased;
decreased air flow through the grate decreases coarse particle ejection into
7
the flue gas; the overall excess air is reduced and thus the combustion temperature is higher, which reduces incomplete combustion during pyrolysis, and
the emissions are lower. Also, when the secondary air is preheated it may
decrease alkali metal compound release and improve secondary combustion.
In masonry heaters, the most problematic issues from the emission point
of view are too fast pyrolysis and too high combustion rate compared to the
air intake. Since air intake size is restricted, there may be an overall lack
of available oxygen. Gasification rate can be controlled by the primary air
supply, fuel moisture content and log and batch size. When combustion conditions are good, fine particle emissions are mostly composed of the vaporised
ash forming elements from the wood. (Tissari, 2008; Tissari et al., 2008b)
2.2.2
Sauna Stoves
Sauna stoves are very common in Finland. They are problematic from the
emission point of view. The structure is simplistic and combustion process
undeveloped. A picture of a wood burning sauna stove is presented in Figure 2.4.
Figure 2.4: A Harvia sauna stove. (Harvia, 2012a)
8
Because of a small firebox and no secondary combustion, the efficiency
of sauna stoves is low. Only half of the heat released is stored in the stones
in the stove. Heating need in a sauna room is temporarily high, so sauna
stoves operate with a high combustion rate. Furthermore, the air supply is
insufficient in relation to the high gasification rate, and this results in incomplete combustion. To reduce the emissions, the combustion technique should
be developed (for example, secondary air supply would lower emissions) or
secondary removal techniques used. (Tissari, 2008)
2.2.3
Residential Wood Boilers
Wood log boilers are used mainly in rural areas as primary heating devices. In
Finland, these are typically updraught (over-fire) boilers, which have higher
emissions than other log boiler types, and are sometimes used without a heat
storage tank. When a boiler is used without a heat storage tank, they are
often used with low thermal output using smouldering combustion, and this
causes high emissions. Other common boiler type is multi-fuel boilers, which
can burn oil, wood or pellets. They are sometimes used primarily for wood
log combustion with an updraught technique, and without a heat storage
tank. Finnish manufacturers have also some crossdraught boilers, which are
more developed and have lower emissions. However, these often lack control
techniques that Central European similar boilers have, and the combustion
process is more dependent on the user. Figure 2.5 shows schematics of updraught and downdraught boilers. (Tissari et al., 2005)
Wood chip boilers are mainly used to heat agricultural buildings and big
single buildings, such as schools in rural areas. They operate similarly to side
feed pellet burners, but with larger burner screws in order to allow the use of
larger fuel particles. Most typical new appliances are 100-200 kW devices for
agricultural buildings. The smaller scale boilers are almost exclusively stoker
burners. In these, the burner can be mounted partially inside the firebox of
the boiler, and partially outside, and only the hot flue gases are led to the
boiler. (Tissari et al., 2005)
Pellet boilers (figure 2.6) are still rare in Finland. Most common pellet
appliances follow the so called Swedish model. An oil burner of a former
oil boiler is replaced with a pellet burner, and a pellet storage is built next
to the boiler, from which the fuel is fed with a screw. Also the wood chip
burners can be used to burn pellets. (Tissari et al., 2005)
9
Figure 2.5: A schematic of (a) an updraught and (b) a downdraught
boiler. (Tissari et al., 2005)
Figure 2.6: A pellet boiler. (Pellettipojat, 2012)
10
2.2.4
Other Appliances in Finland
In addition to masonry heaters, there is a number of masonry cookstoves
(figure 2.7) in Finland, which often have a wood burning oven and a masonry
heater in the same structure. These appliances can be used for both cooking
and heating. Furthermore, there are some wood burning cookstoves in older
buildings. There are also some open fireplaces in Finland. These appliances
don’t have air intake control, and the emissions are often high.
Figure 2.7: A masonry cookstove. (Alakangas et al., 2008b)
Iron stoves (other than sauna stoves) are not as common in Finland than
in, for example, Central Europe. Iron stoves (figure 2.8) don’t store the heat
like masonry heaters. Therefore, if constant heating is needed, smouldering
combustion might be used. This causes increase in the emissions.
Pellet burners (fig 2.9), just like pellet boilers, are still rare in Finland. In
these appliances, the pellets can be fed manually or automatically. Especially
with automatic fuel feeding, the combustion process can be controlled well.
This allows optimal combustion conditions to be used over a long time, and
emissions to be low.
One quite common device in Finnish summer cottages is a water boiler.
In these boilers, the water is heated for domestic use, and used directly from
the boiler (instead of supplying it into a water pipe). The devices are rather
undeveloped, and combustion might be poor, causing high emissions.
11
Figure 2.8: An iron stove. (Lämpömaa, 2012)
Figure 2.9: A schematic of a pellet burner. (Alakangas et al., 2008b)
12
2.3
Emissions of Air Pollutants
The PM2.5 emissions from RWC are highly variable. Three key factors affecting the emissions in RWC can be identified: appliance type, fuel quality, and
burning practices and conditions. A more complete list of factors affecting
the emissions of RWC is presented in Table 2.2.
Although the appliance type gives an idea of the emissions of combustion,
there are differences between the same type of appliances, and especially appliance sub-types. For example, conventional masonry heaters have typically
higher emissions than modern ones. The main fuel characteristics affecting
emissions are moisture content, ash content, and wood species. For burning practices, emissions are influenced by several factors, such as log size,
batch size, arrangement of logs, and lighting. Most important factors in the
burning conditions are high combustion temperature, sufficient amount of
combustion air supply, and adequate mixing of combustion air and flue gas.
These conditions allow complete combustion, which reduces emissions. (Tissari, 2008)
Even though appliance type has important effect on the emissions, combustion practices can have even higher effect (Paunu et al., 2012). Poor
combustion practices, for example smouldering combustion or wet wood, can
significantly increase the emissions of any device. Therefore, all appliances
should be used with close to optimal parameters to achieve low emissions.
2.3.1
Wood as Fuel
Most wood species growing in Finland are suitable for fuel use. In residential
combustion used wood is usually in the form of logs. Also wood chips, pellets,
briquettes and hog fuel are used.
Fresh wood has typically 40-60% moisture content. In residential use,
wood is usually dried before use. The desired moisture content for wood
logs is around 15-20%. Dry pellets might have moisture content of 6%. For
wood chips, the moisture content might be even up to 60%. Wood has
a high content of volatiles, 80-90%. Therefore, it is a long-flame fuel and
needs a large combustion chamber. About 99% of the dry matter of wood
is carbon, hydrogen and oxygen. Nitrogen content is typically under 0.5%,
and sulphur low, less than 0.05%. The ash content of stemwood excluding
the bark is typically under 0.5%, or under 2% for coniferous trees. Bark has
higher ash content, from 1.6-3%. The sulphur ash content in general is lower
than in other solid fuels. The mineral content is usually less than 0.5%, main
compounds being calcium (Ca), potassium (K), magnesium (Mg), manganese
13
Table 2.2: Factors affecting emissions of residential wood combustion. (Tissari et al., 2005)
Factor
Characteristic
Effect
affecting emissions
Fuel
Moisture
Lowers combustion temperature
Ash content
Increases particulate emissions
Amount of gasifying Pyrolysis control more difficult,
substances
flame needs a lot of space
Log size
In continuous combustion affects
how constant the combustion is
In batch combustion affects firing
and gasification rate
Appliance
Firebox size, shape Affects draft conditions and comand materials
bustion temperature
Flue gas outlet dimen- Affects draft conditions
sions
Air supply
Affects the amount and mixing of
combustion air
Smokestack
Stack height, size and Affects draft conditions
shape
Combustion condi- Natural draft Affects
tions
combustion control
Flue gas residence Affects emission burnout
time
Combustion tempera- Affects emission burnout
ture
Air supply and mixing Affects flue gas and air mixing
Flue gas after- After-treatment appli- Affects emission quantity, and aptreatment
ances
pliance operation and use
Operating condi- Combustion rate
Affects gasification rate and
tions
amount of combustion gases
Fuel loading rate
Affects combustion rate and momentary power
Control devices
Affects fuel, air and appliance
power control
Appliance user
Operation
habits, Affects several combustion factors
garbage burning
14
(Mn), sulphur (S), chlorine (Cl), phosphorus (P), iron (Fe), aluminium (Al),
and zinc (Zn). (Alakangas, 2005; Tissari, 2008)
Typical heating values of Finnish wood logs are between 14 and 15 MJ
−1
kg for moisture content of 20% (Alakangas et al., 2008a). For wood pellets
with moisture content of 10% heating value is around 17 MJ kg−1 (Alakangas,
2005). The water content has clear impact on the heating value; the drier
the wood the higher the heating value. Compared to other solid fuels, the
heating value of wood is low, and wood also needs more storage space.
2.3.2
Combustion Process
Combustion is defined as a reaction where fuel reacts with oxygen, and heat
is produces by this chemical process. From fuel particle point of view the
combustion process consist of several phases: drying and heating of fuel, pyrolysis, firing and combustion. The first three phases consume heat. The
flaming combustion and the combustion of the residual char generate heat.
With wood fuels, combustion reactions happen mostly between gaseous products. In the residual char combustion the reactions are between gases and
carbon, which is on the surface of the solid char. (Tissari, 2008)
The first phase, the drying and pyrolysis, consists of the warming of the
fuel particle to drying temperature and, thereafter, vaporization of majority
of the water. When the moisture content of the fuel has dropped sufficiently,
the temperature of the fuel increases and the vaporization of the volatile hydrocarbons starts. Pyrolysis contains many complex chemical reactions, that
are parallel and sequential. The fuel constituents begin to hydrolyse, oxidise
and dehydrate, while the large structures, such as cellulose, hemicellulose
and lignin, begin to degrade. Several gaseous and liquid products are formed
during pyrolysis, e.g. volatile organic compounds (VOCs), H2 O, CO2 , H2
and CO. (Tissari, 2008)
Devolatilization of wood starts at 200◦ C, and the devolatilization rate
increases fast above that. Most volatiles have vaporized at 400◦ C, and devolatilization rate drops quickly. Decompositions of hemicellulose, cellulose
and lignin occur at 200-350◦ C, 250-450◦ C and 200-500◦ C, respectively. (Tissari, 2008)
The kindling of the combustion gases happen when the amount of heat
produced is higher than the heat loss to the environment. Fuel particle size
and moisture have a significant effect on the time it takes the kindling to begin. Moisture slows down the kindling, as the vaporization of water consumes
energy and the water vapour cools the surface of the fuel particle. Normally
the pyrolysis products burn around the fuel particle as a diffusion flame. This
generates heat for other pyrolysis reactions. As the heat generation increases
15
the fuel temperature rises. The combustion is accelerated until pyrolysis gas
production is slowed down. During the pyrolysis, the proportional share of
carbon compared to hydrogen increases, and the residual char combustion
starts. (Tissari, 2008; Tissari et al., 2005)
The last stage of the combustion is the flameless combustion of residual char. In this phase, the combustion happens on the surface of the fuel
particle. Biomass combustion has normally around 10-30% of residual char
content by dry weight. However, 25-50% of the total energy from the combustion comes from this phase. This phase also lasts for the longest time, as
the diffusion of oxygen to the char surface is slow. The reactions between the
gases and the char surface can also happen inside the fuel particle. Therefore,
the porosity of the fuel particle affects the duration of the combustion. (Tissari, 2008; Tissari et al., 2005)
2.3.3
Batch and Continuous Combustion
Combustion in an RWC appliance is either continuous or batch type, depending on the appliance. In batch combustion, the fuel is loaded in separate
batches, and the combustion starts from the first phase for each batch. The
emissions differ between batches and combustion phases. For the first batch,
fuel quality and combustion and firing practices have the strongest effect on
emissions. Usually the temperature of the firebox is low at the time of firing,
i.e. the high temperature that is needed for combustion is absent. To lower
emissions, firing should be done on top of the first batch. This forces released
gases to go through the flames and they are at least partly combusted. The
first batch burns slower than subsequent batches, and especially the firing
phase is longer than in the subsequent batches, so the combustion gases have
more time to combust. Therefore the emissions from the first batch are not
necessarily higher than from the following ones. As the firebox heats up as
the combustion progresses, the rate of pyrolysis increases, and this can cause
an increase in the emissions if pyrolysis gases are not combusted. (Tissari
et al., 2005)
In continuous combustion, the fuel is fed continuously, and possibly automatically, to the burning chamber. All combustion phases are in effect all
the time, and happen in the fuel layer. The combustion process is stable,
and controllability is better compared to batch combustion devices. Unstable
combustion can occur in situations where the process if interfered, appliance
use is intermittent, during cleaning, or when the appliance is used with partial load.
16
2.3.4
Wood Combustion Emissions
When the carbon in the wood is completely combusted, only carbon dioxide
(CO2 ) is formed. However, the combustion is often incomplete. Therefore,
part of the carbon is released as carbon monoxide (CO) and hydrocarbons
(Cx Hy ). Hydrocarbons can be grouped according to their boiling point: very
volatile, volatile, and semivolatile organic compounds, and particle phase
compounds (VVOC, VOC, SVOC, and POM, particle organic matter). Often
VOC means non-methane volatile organic compounds (NMVOC), from which
methane has been excluded. In measurements, the term organic gaseous
compounds (OGC) is also used. Hydrocarbons can also be grouped according
to the functional group. From these, polyaromatic hydrocarbons (PAH) are
the most important ones, as many of them are carcinogens and mutagens.
Nitrogen oxide (NOx ) emissions from residential wood combustion are mainly
from the nitrogen of the wood. The temperature of RWC seldom rises high
enough for NOx to form from nitrogen in the air. Sulphur emissions from
RWC are low, because the sulphur content of wood is typically low (below
0.05%). Water vapour (H2 O) forms as the moisture of the wood is vaporized
and also from hydrogen of the wood. Furthermore, combustion air has water
vapour, and this transfers to the flue gas. (Tissari et al., 2005; Tissari, 2008)
From good combustion, particulate emissions are formed mostly from
minerals in the wood. Poor combustion produces more incomplete combustion products, namely soot, tar, gaseous hydrocarbons and carbon monoxide.
Fine particles (PM2.5 ) that are formed in wood combustion are ash particles, such as alkali sulphates and chlorides, calcium oxides and metal oxides,
organic particles or soot particles with organics condensed on the surface.
Bigger particles consist of ash, charred remainders of the fuel particles, or
agglomerates formed from fine particles. (Tissari et al., 2005; Tissari, 2008)
The formation process of particles is illustrated in figure 2.10. Incomplete combustion may produce liquid or tar-like parts. These are formed by
gas-to-particle conversion of organic vapours in cooled flue gas. The organic
compounds can be in liquid or gaseous form, depending on the environmental
conditions. New particles may be formed by nucleation of heavy hydrocarbons. These hydrocarbons may also condense onto existing particles. The
latter process is more common. After the flue gases enter the chimney and,
further, the atmosphere, the condensation of the particles continues as the
combustion aerosols cool and are diluted. (Tissari et al., 2005; Tissari, 2008)
Fine ash particles are formed by homogeneous nucleation. This happens
after the flame, where the temperature and the vapour pressure of the ash
species decrease. The mineral compounds are easily released during the pyrolysis of the fuel. Temperature determines how much as particles is released:
17
Figure 2.10: Formation of (1) soot, (2) fine ash, (3) coarse particles, and
(4) particle organic matter (POM) in residential wood combustion. (Tissari,
2008)
higher temperature causes more ash particles to be released. Coarse particles (∼1-10 µm) are formed from ash compounds with low volatility, and
partially are unburnt char. At low temperatures the main formation process
18
for large ash agglomerates is agglomeration. When the temperature is sufficient, the ash compounds can melt and form regular ash droplets. Residual
fly ash particles lifted from the fuel bed form the super coarse particles (>10
µm). (Tissari, 2008)
Soot particles are mostly formed in the flame. Inside the diffusion flame
there are fuel-rich areas, where the soot particles are formed from hydrocarbons as shown in the branch (1) of figure 2.10. These fuel-rich zones always
exist in RWC, since the mixing of combustion gases and air is not sufficient.
Therefore the hydrocarbons are unable to oxidise. The soot particles originate from PAH compounds. The PAH compounds polymerize, grow, and
bond to the surface of core particles, finally forming primary soot particles.
After the soot particles are formed, most of them are burned in the oxygenrich area of the flame. However, small part of them are released unburned.
The size and amount of soot particles is determined by the extent of soot
oxidation. (Tissari, 2008)
2.3.5
Dispersion
The particle concentration and chemistry in the atmosphere caused by emissions depends on multiple factors. The topography of the environment and
the weather conditions have a strong effect on local concentration levels.
Other affecting factors are the source strength, distance from the source, the
atmospheric processes encountered during the transport, and mixing and
interaction with gases and particles from other sources during the transport (Pleijel, 2007). Lifetime of the particles in the atmosphere is quite
short, from few hours up to a week (Pleijel, 2009).
The chemical composition of the particles transforms still after they are
emitted to the atmosphere. When the hot exhaust gases are emitted, most
likely condensable gases in the exhaust gas form new particles. Particle
formation speed drops quickly after the emission due to rapid dispersion.
Some of the smallest particles may evaporate. After this, the emitted organic
gases can form condensable compounds through chemical reactions in the
atmosphere, and these can then condense on the particles around. This is a
much slower process. (Pleijel, 2009)
The temperature of the exhaust gas quickly drops after it leaves the chimney into the atmosphere. This enhances the condensation of different compounds on the surface of particles. The condensation depends heavily on temperature. It starts after the temperature drops below the boiling temperature
for the compound, and accelerates as the temperature falls, until the exhaust
gas is the in same temperature as the surrounding air. After this, condensation continues only with new condensable compounds produced by precursor
19
gases by chemical reactions. This and other particle growing processes are
made much slower by the strong dilution of the exhaust gases. (Pleijel, 2007)
New nanometre sized particles can also be formed by nucleation. When
the temperature of the exhaust gases drop fast and by much due to cool ambient air, a certain compound can become highly supersaturated. This can
also happen due to lack of surfaces for the gas to condense on. The concentration of the compound can be about 10 times at which the condensation
starts. These supersaturated gaseous compounds can then from new small
particles. (Pleijel, 2007)
As the emitted particles are transported further from the source, they
are exposed to the particles and gases in the natural atmosphere, and the
coagulation and condensation continues. The emissions are usually highly
diluted now, so the processes are much slower and depend strongly on the
concentration of participating components. Photochemical reactions induced
by sunlight produce strong oxidizing agents, such as ozone. Chemical reactions turn non-condensable gases to condensable. Cloud droplets offer a
reactive environment for the particle interactions. This can strongly affect
the chemical composition of particles. (Pleijel, 2007)
2.3.6
Factors Affecting Emissions
From the emission point of view complete combustion is favourable. Three
main parameters for complete combustion can be identified as follows: temperature of the combustion has to be high, air supply sufficient, and mixing
of combustion air and fuel gas adequate. (Tissari et al., 2008b)
When air intake is restricted in the combustion chamber, combustion
becomes smouldering. Also, if the burn rate is too high, the air supply
becomes insufficient and this leads to smouldering-like combustion. Smouldering combustion is typical, for example, in old wood boilers, which don’t
have heat-storage tanks, and other appliances without heat storing, such as
light metal stoves, since they are often used with a slow combustion rate with
restricted air to keep up the heating for a long time. An insufficient control
of primary air supply can lead to high particle organic matter (POM) and
elemental carbon (EC) emissions.
The combustion temperature affects the emissions in a complex way. If
the temperature is too low, the oxidation reactions are slower and combustion
compounds don’t burn out as completely as in higher temperatures. RWC
appliances demand an overall excess of oxygen for local oxygen concentrations
to be adequate for combustion reactions. However, the excess air lowers the
temperature due to inert nitrogen being heated in the air. A low temperature
and low local oxygen concentration both increase gaseous CO and volatile
20
hydrocarbon emissions. The temperature can also be too high. In Finnish
heaters it is typical that the air intakes are restricted and the combustion
temperature is high. This leads to a too high gasification rate relative to the
air intake size when large fuel batches are used. The supply of air is therefore
insufficient and combustion becomes incomplete and emissions rise. (Tissari
et al., 2008b)
The combustion temperature also affects the ash vaporization, so that in
high temperatures the amount of released ash particles is higher than in lower
temperatures. The chemical composition of ash is usually quite constant
between different pure wood fuels, so the temperature mainly determines the
fine ash emissions in RWC. (Tissari, 2008; Tissari et al., 2008b)
In batch combustion, the size of the fuel load affects the emissions. If the
fuel load is too big or logs are too small compared to the heater air intake,
combustion is incomplete. A large batch size increases gasification rate and
results in an insufficient air supply. Tissari (2008) measured that doubling
the batch size makes OGC emissions 4.0-, CO 2.2- and PM1 1.9-fold compared
to the initial batch size. Log size affected the emissions even more: small
logs caused 8.7-fold OGC, 2.3-fold CO and 4.8-fold PM1 emissions compared
to larger logs with the same batch size.
The results of log size are not applicable universally, however. The observations from open fireplace emissions (Dasch, 1982; Stern et al., 1992)
suggest that in these appliances bigger logs increase PM and CO factors.
Furthermore, Tissari et al. (2008a) stated that there didn’t seem to be a
direct proportionality between the size of the logs and their gasification rate,
and thus more research on the subject should be conducted.
Measuring emissions from RWC can be tricky, as the measurement methods may significantly affect the results. Furthermore, total PM is not a
suitable factor when comparing different RWC appliances, as the emissions
of total PM depend on the appliance and these particles occur randomly
in the flue gas, and every measurement method causes remarkable losses of
coarse particles. (Tissari, 2008)
2.3.7
Emissions from Finnish Appliances
Particulate emissions from masonry heaters and sauna stoves are mainly
PM1 , i.e. the aerodynamic diameter of the particles is below 1 µm (Tissari
et al., 2008a). Combustion conditions affect significantly PM1 mass emissions
as well as particle number and mass size distributions. In laboratory measurements made by Tissari et al. (2008a), the main difference between the
PM1 emissions of MMH and CMH comes from the firing phase. Otherwise
the PM1 emissions in the laboratory conditions were alike. PM1 emission
21
factors for both appliances, were 0.7 g kg-1 . The particle number emissions
were higher in MMH, but the particle size was smaller, so the overall PM1
emissions were similar. The PM1 emissions from MMH were 50% ash compounds and less than 20% organics. Tissari claimed that this indicated a
possibly different chemical composition of particles, and thus variable health
effects, compared to other appliances. In field measurements made by Tissari
et al. (2007), the PM1 emission factors for MMH and CMH were 0.7 g kg-1
and 0.6-1.6 g kg-1 , respectively. All in all, Tissari (2008) states that the CMH
emissions vary between 0.6 and 3.3 g kg-1 . Tissari found NOx emissions to
be relatively low with all wood fuels due to low combustion temperature.
Emissions differ greatly between different burning phases. The differences
between emission factors were 10- to 100-fold between the phases. Tissari
et al. (2008a) found that the biggest proportion of PM1 emissions come from
the firing phase, and the proportion grew higher from the first to the last
batch. Most OGC emissions happened also in the firing phase (18 gC kg-1
on average), and CO emissions were high as well (59 g kg-1 ). This was due
to a high gasification rate and resulting insufficiency of air supply and air
and fuel mixing in the firing phase. After the firing phase the PM1 emissions
decreased fast. OGC emissions were still significant in the combustion phase
(2 gC kg-1 ), but low in the burnt out phase (0.39 gC kg-1 ) and nearly nonexistent from the glowing embers (0.08 gC kg-1 ). CO emissions were lowest
in the combustion phase (13 g kg-1 ), but stayed high during the burnt out
phase (21 g kg-1 ) and from glowing embers (22 g kg-1 ). High CO emissions
were caused by the low diffusion rate of oxygen to the char and combustion
chamber cooling caused by the high excess air volume. The temperature
often dropped below the complete oxidation threshold for CO, i.e. 800◦ C.
Emissions increase from the first batch because the temperature of the
firebox rises and it accelerates the gasification rate of the wood. These reasons also decrease residual oxygen concentration. Thus, the supply of air
might not be sufficient, resulting in incomplete combustion and high soot
and organic carbon compound numbers in the emissions. OGC, CO and average PM1 emissions grow from batch to batch. The increasing combustion
temperature causes increase in the amount of all ash species. In Tissari’s
measurements, the last batch had 4.5 times higher PM1 emissions than the
first batch in a CMH.
Smouldering combustion (SC), i.e. combustion where air intake is restricted on purpose to slow down the combustion rate, significantly increases
the emissions from masonry heaters. Frey et al. (2009) measured six to seven
times higher POM emissions factors from SC than from normal combustion
(NC) in a common type of a small Finnish masonry heater. In SC, POM
made up 70% of the emissions, whereas in NC POM comprised 30% of the
22
emissions. EC comprised 32% ± 5% of total emissions in NC and 22-27%
in SC. The emission factors for EC were 0.9 ± 0.2 g kg-1 and 2.0-2.3 g kg-1 ,
respectively. The emission factors for POM were 0.9 ± 0.3 g kg-1 and 5.8-6.2
g kg-1 , respectively.
The emissions of CO and organic gaseous compounds (OGC) show the
completeness of secondary combustion. Tissari (Tissari, 2008; Tissari et al.,
2008b) found that the combustion temperature was lower in SC compared
to NC, which represented the best combustion practice. Smouldering combustion was achieved with small logs, big batches and closed air intakes in a
CMH. Most emissions were higher in SC, like OGC, POM, CO and PM1 , but
ash and particle number emissions were smaller, less than 50% compared to
NC emissions. Furthermore, SC had twice the particle size of NC. In addition, the particle composition was different between NC and SC. In NC, POM
composed 33% and EC 32% of the emissions, whereas in NC they made up
67-69% and 22-27% of the emissions, respectively. Fine ash emissions were
2-4 times higher in NC than SC.
There are few scientific measurements made on emissions of sauna stoves.
Tissari measured PM1 emissions from sauna stoves in laboratory (Tissari
et al., 2008a) and field (Tissari et al., 2007) conditions. The emission factors
were 5.0 and 2.7 g kg-1 , respectively. In laboratory, CO emissions were 55 g
kg-1 and OGC 10 gC kg-1 . In field measurements, CO emissions were clearly
higher, 120 g kg-1 , and OGC 13 gC kg-1 .
The combustion conditions have a clear effect on the composition of PM
emissions. Figure 2.11 shows how the composition differs between different
devices and combustion practices. Modern masonry heater represents the
best combustion, sauna stove a poor combustion and conventional masonry
heater something from between. Smouldering combustion is the poorest combustion quality, with the highest emissions. The most interesting point is the
ratio between elemental (EC) and organic carbon (OC). Organic carbon is
directly proportional to POM. When the combustion quality comes poorer,
the OC/EC ratio grows.
2.3.8
International Comparison
Compared to other parts of the world, masonry heaters are more common
in Finland. Pellet burners and other stoves than sauna stoves, on the other
hand, are rare in Finland in comparison. Since the appliance stock is different, the emission characteristics might be as well.
Emission-wise the Finnish supplementary heating devices have low to intermediate emissions. Masonry heaters, especially modern ones, have low
emissions. Sauna stoves, being much more simplistic devices, have quite
23
Figure 2.11: Chemical composition of PM1 samples from masonry heaters
and sauna stoves. (Tissari, 2008)
high emissions. In general, modern appliances, such as pellet burners, have
the lowest emissions. Old devices, such as traditional iron stoves, old log
boilers without a heat accumulator tank and open fireplaces, have high emissions. (Paunu et al., 2012)
Compared to Central Europe and Sweden, Finnish boilers are less developed and have higher emissions. Furthermore, legislation in Finland is less
strict than in some other countries. For example, Swedish legislation doesn’t
allow updraught boilers to be used without a heat accumulation tank. (Tissari et al., 2005)
2.4
Emission Reduction Possibilities
There are numerous ways to reduce the emissions from residential wood combustion. These include the development of combustion technologies, cleaning
of the flue gases, promoting better combustion practices and the use of legislation.
24
2.4.1
Advanced Combustion Technologies
The most straightforward way to reduce emissions from wood combustion is
to change the device into a more modern one with lower emissions. Especially the replacement of old open fireplaces and log boilers with newer models
may reduce the emissions significantly. The problem with this method is the
price of the new devices, and the long lifetime of RWC appliances. Thus, on
a larger scale, it doesn’t matter how good the new appliances are, as the old
ones will be replaced too slowly, if results are wanted in a few decades. By
assessment made by Savolahti et al. (2012), replacing conventional masonry
heaters with modern ones could reduce PM2.5 emissions by 1% from Finland’s total 2020 emission levels, with unit cost for the reductions being 253
ke/ton. Compared to other emission reduction options studied, replacement
of current devices with new ones is expensive, be it log or chip boilers with
boilers or masonry heaters with newer models.
Log boilers can be used with or without a heat accumulator tank. Equipping the tank allows the boiler to be used with optimal combustion rate
without compromising continuous heating. Boilers without heat tank have
to be operated with low combustion rate in order to produce heating at a
constant rate, and this leads to higher emissions. Savolahti et al. (2012) estimated, that installing accumulator tanks to all boilers would reduce Finnish
PM2.5 emissions by 5% of the total emissions of 2020. The unit cost for the
reductions was 2 ke/ton.
Harvia Oy has developed a sauna stove (Harvia GreenFlame, figure 2.12)
that should have lower emissions and lower fuel use than traditional models (Harvia, 2012b). They claim that especially CO and particulate emissions
are significantly lower. A patented automatic control mechanism controls the
air intake. In the firing phase, the combustion air is fed from under the fire.
When the combustion progresses and the stove heats up, the air is fed over
the fire. In the burn out phase, as the stove cools down, the air inlets are
closed. The mechanism is based on thermal expansion of metal. This device
shows, that sauna stoves can be developed further without compromising the
functionality of the stove.
There are some appliance types that are not so well known, but may
be good choices. One of them is a rocket stove. They have an L-shaped
internal design. Combustion chamber is at the end of a vertical or horizontal
fuel magazine, and chimney is above the combustion chamber. The aim is
a high combustion efficiency due to high combustion temperature and good
air draft. These type of appliances are used for cooking in many third-world
countries. A version of the device is a rocket mass heater, which combines
the design with the masonry heater. A schematic is presented in figure 2.13.
25
Figure 2.12: Harvia GreenFlame sauna stove. The blue arrows show the air
intake in the combustion phase. (Harvia, 2012b)
The emissions of rocket-type appliances should be low and efficiency high.
2.4.2
Combustion Practices and Legislation
Combustion practices have a high impact on emissions from RWC appliances.
This is especially true for batch combustion, in which the user is usually solely
responsible of the combustion parameters. However, it is hard to assess how
much of wood is combusted with poor practices, as it is unknown how much
users use their appliances in a non-optimal way.
Compared to technical measures, affecting the combustion practices could
be a fast way to reduce emissions of RWC. In practice, this could be done via
information campaign aimed at households that use wood as a fuel. However,
the effectiveness of this kind of campaign is hard to estimate. Savolahti et al.
(2012) assessed that even if the effect of such a campaign would be small,
the unit cost of PM2.5 emission reduce would be on the same level with the
cheapest technical measures. This implies that even if results are uncertain
prior to a campaign, it might still be worth a try.
Emissions can also be affected through legislative measures. For example, Germany and Austria have emission standards for new stoves. Limits for
Finland have been under discussion, but no measures have been taken yet.
However, European Union is preparing to add small-scale wood combustion
26
Figure 2.13: A schematic of a rocket mass heater. (ErnieAndErica, 2012)
devices into the Ecodesign Directive. As the appliances are replaced slowly,
such measures would take a long time to have a significant effect. Emission
limits could also be put on existing appliances, although supervising of such
a law could be hard. Other measures include the banning of the use of the
appliances for example during the time of general air quality problems. Legislation can also offer incentives to invest in appliances with lower emissions.
2.4.3
Flue gas cleaning
Electrostatic precipitators (ESP) are used in many power plants. In practice,
small-scale ESPs are options for the future, as they are not yet commercial
technology. There are other flue gas cleaning technologies in development,
but ESP seems like the most effective. According to calculation of Savolahti
et al. (2012), installing ESPs to all log boilers (with heat accumulator tanks)
in Finland could reduce the PM2.5 emissions by 2%, from the total emissions
of Finland in 2020, with a unit costs 40 ke/ton, respectively. The unit costs
for applying ESPs to chip or pellet boilers is significantly higher, making
them infeasible options. It should be kept in mind that the prices of the
27
commercial ESP units are uncertain, and it has significant impact on the
unit costs of the emissions reduction. A picture of an ESP fitted to a stove
is presented in figure 2.14(a).
(a)
(b)
Figure 2.14: Pictures of two main small scale wood combustion flue gas
cleaning: (a) an ”Airbox” ESP system installed to a stove (Hartmann et al.,
2011), and (b) a schematic of a catalytic combustor (Hukkanen et al., 2012).
One technology to reduce RWC emissions is already in use: catalytic
combustors have been widely used in United States. They have been found
to effectively oxidise carbon monoxide and light non-methane hydrocarbons.
One problem with using catalyst in RWC is the low temperature of the flue
gases in some appliances. This is especially problematic in the start-up phase,
when the emissions can be really high but temperature still low, preventing
the catalyst to work optimally in the most critical moment. This can be
solved by using heating systems, such as hot air injected through the catalyst, or placing the catalyst directly after the fire box instead of in the stack.
Furthermore, the catalyst can be fouled easily, and its activity can decrease
by the affected emissions. The use needs motivation from the user, as the catalyst needs to be cleaned regularly and might need surveillance. (Hukkanen
et al., 2012)
Hukkanen et al. (2012) studied the reduction of sauna stove emissions
28
with a commercial catalytic combustor (figure 2.14(b)). They measured emission drop for PM1 from initial 391 to 252 mg MJ-1 , and for CO and OGC
from 7900 to 6200 and 1500-1300 mg MJ-1 , respectively. The reduction occurred mainly during the gasification stage, except for CO. The same held
for OC and EC emissions, with OC emissions reduced more than EC. EC
requires a high temperature for oxidation, and was therefore affected less.
However, the catalyst increased the flue gas temperature enough so that the
oxidation was able to occur at all.
29
Chapter 3
Modelling of Population Exposure to Fine Particles from residential wood combustion
The emissions from residential wood combustion have an obvious effect on
population exposure to PM2.5 . in addition to the emission strength, another
important factor on the exposure is the population density; the same emission causes a bigger population exposure in a more densely populated area.
Therefore, it is crucial to know where the pollutants are emitted. The aim of
this chapter is to compare the PM2.5 emissions and exposures from residential
wood combustion (RWC) in different residential area types.
The spatial distribution and dispersion of PM2.5 emissions of RWC in
Finland was modelled. From the resulting concentration it was possible to
estimate the population exposures, and compare these in the different residential area types with different population densities.
3.1
3.1.1
Methodology
Emission Modelling
The emission and exposure assessments were done with the Finnish Regional
Emission Scenario (FRES) model (Karvosenoja, 2008). The FRES model
is an integrated assessment model of air pollution emissions and impacts in
Finland. The temporal and spatial resolutions for the model are one year
and 1 km x 1 km over Finland, respectively. The main objective of the FRES
model is to estimate the future emission scenarios, and evaluate the different
pathways of emission reduction possibilities.
30
Figure 3.1: Flowchart of the emission calculation in the FRES model (Karvosenoja, 2008).
31
A flowchart of the FRES model is presented in figure 3.1. The FRES
model uses both top-down and bottom-up approach for emission sources:
area sources and point sources, respectively. The point sources include large
plants with significant emissions on their own. Most emission sources are so
small in emission quantities and sometimes also hard to pinpoint that it is
not convenient to model them as point sources. Instead, these sources can be
handled as area sources. For an area, there is some known or estimated number of sources, which are represented by allocation surrogates. Residential
wood combustion are modelled as area sources.
In the FRES model, the emissions are calculated with activity levels,
emission factors, and emission control technology removal efficiencies and
utilization rates. For RWC, the activity is the annual wood use. For each
source, the rate of emission per activity unit is specified with emission factors.
The emission factors are determined so that the emissions represent the real
life emissions of the device as accurately as possible. The most recent year
in the model is 2005. This year was used in this study. The emission factors
are supposed to be constant, and the changes in the factors are modelled by
changing the emission control technology parameters or change in activity
division.
The PM2.5 emission factors for different appliances are presented in Table 3.1. The emissions of different RWC devices are studied in more detail
in section 2. Less developed boilers, i.e. log boilers without a heat accumulator tank, have high emissions. Modern devices, especially pellet boilers,
have low emission factors compared to other devices. Same holds for stoves
and masonry heaters: modern masonry heaters have a low emissions factor,
and fireplaces emit high emissions. Sauna stoves have quite high emissions,
due to poor combustion technology. The emission factors used in the FRES
model are based on literature, and are chosen to represent the real life usage.
The wood uses in the FRES model in the year 2005 for primary (boilers)
and supplementary (masonry heaters and stoves) wood heating are 14 and
30 PJ, respectively.
The calculation of the emissions is divided into several steps. The initial
step is the fuel wood use estimation for the whole of Finland. The wood use
is divided for different appliances types according to their estimated shares
in the year 2005 (Torvelainen, 2009). Every appliance type has it’s own
emission factor (Table 3.1), which gives the PM2.5 emissions per combusted
energy unit. Therefore, the country level emissions for a device are calculated
as:
EMa = Aa × EFa
32
(3.1)
Table 3.1: PM2.5 emission factors and activity levels in the FRES model for
different RWC appliances (Torvelainen, 2009).
Appliance
PM2.5
emission Activity
factor (mg MJ-1 ) (PJ)
Wood chip boiler
50
2.1
Pellet boiler
30
1.0
Log boiler with heat accumulator
80
7.3
Log boiler without heat accumulator 700
2.3
Farmhouse pellet boiler
30
0.14
Farmhouse log boiler with heat ac- 80
1.3
cumulator
Fireplace
800
2.5
Kitchen range stove
120
4.1
Conventional masonry heater
120
8.9
Masonry oven
120
7.3
Modern masonry heater
80
0.77
Sauna stove
200
6.7
where A is the activity (wood use), EF is the emission factor and a refers
to the appliance type. This way the emissions of the whole of Finland are
calculated for all appliance types, and then summarized to the source classes:
X
EMa
(3.2)
EMs =
a
where s refers to the source class. There are three RWC source classes:
primary and supplementary wood heating, and primary wood heating in
farmhouses. The classes are based on the appliance types, so the primarily
wood heated houses have also supplementary wood heating. Primary wood
heating class contains residential boilers, farmhouse primary wood heating
farmhouse boilers, and supplementary heating all the other residential wood
combustion devices (stoves, masonry heaters etc.). The farmhouse primary
wood heating is included because the wood use in farmhouse boilers is bigger
than in other residential house boilers (Torvelainen, 2009). The emissions
of the farmhouse boilers represent this difference, not the whole emissions
from the boilers, and are calculated separately. They are later included in
the total primary wood heating emissions.
To achieve the 1 km × 1 km resolution the emissions of the whole of
Finland are first distributed to municipalities according to source class activity, and then to the municipality’s area. Every municipality and 1 km ×
1 km cell has it’s own weighting factor, which is comes from the allocation
33
surrogates. The municipality emission is calculated by:
Sm,s
EMm,s = P
× EMs
m Sm,s
(3.3)
where m refers to the municipality, EMm,s is the emission of the municipality,
S
Pm the allocation surrogate for the emission source class in the municipality,
m Sm the total allocation surrogate in the country and EMs the total
emission of the source class. The emission of a cell is calculated similarly:
EMc,s =
Sc,s
× EMm,s
Sm,s
(3.4)
where c refers to the cell, EMc,s is the emission of the cell, Sc,s the allocation
surrogate for the emission source class in the cell, Sm the total allocation
surrogate in the municipality and EMm the total emission of the municipality.
The distribution base for RWC emissions depends on the source class.
The bases are listed in Table 3.2. In practice, the emission is first distributed
from the country level to municipalities, and in the next step within each
municipality to 1 km × 1 km cells. All the weighting factors have first been
calculated in the 250 m × 250 m resolution (the resolution for the residential
area type data), and then aggregated to the 1 km × 1 km resolution.
Table 3.2: Distribution bases for different source classes.
Source class
Allocation surrogate Source
Number of detached National
building
and
houses
dwelling register
Primary wood
Residential area type
Urban structure monitoring
heating
system
Heating degree day
Finnish Meteorological Institute
Number of detached National
building
and
houses
dwelling register
Residential area type
Urban structure monitoring
Supplementary
system
wood heating
Heating degree day
Finnish Meteorological Institute
Primary heating (only National
building
and
cell weighting factors)
dwelling register
Farmhouse primary
National
building
and
wood heating
dwelling register
34
For primary wood heating, the weighting factors of the municipalities and
cells are determined by three factors: number of detached houses, residential
area type and heating degree day. The number of the detached houses is
from the Population Register Centre’s National building and dwelling register (Mikkola et al., 1999) 2011 update. Residential area type is based on
urban structure monitoring system (YKR) (Ristimäki, 1999), and has three
options: large city urban (population of the city >20000), small city urban
(population of the city <20000), and rural. The weighting factors for different
area types are based on an annual wood use survey conducted by the Finnish
Forest Research Institute (METLA) (Torvelainen, 2009). The annual wood
uses of houses in different area types are presented in Table 3.3. The heating
degree day represents the different heating need in different parts of the country. Each municipality has it’s own heating degree day value in the FRES
model. The values come from the Finnish Meteorological Institute (Tainio
et al., 2009).
Table 3.3: The annual wood uses of houses in different residential area types.
The numbers are based on an annual wood use survey conducted by the
Finnish Forest Research Institute (METLA) (Torvelainen, 2009)
Residential area type
Annual wood use
(m3 /house)
Primary wood heating
Large city urban
9.3
Small city urban
16.1
Rural
19.0
Supplementary wood heating
Large city urban
2.5
Small city urban
3.0
Rural
5.0
The emission distribution for the supplementary wood heating follows
the same lines with primary wood heating. The basis is the detached house
number from the National building and dwelling register. In the municipality
level the distribution also takes into account the differences in wood use in
different residential area types (Table 3.3). Municipality level also uses the
heating degree day information. In the 1 km × 1 km level, the calculation also
takes into account the differences of wood use in houses with different primary
heating (Torvelainen, 2009). The wood uses for different primary heating
types are presented in Table 3.4. The emission distribution for supplementary
wood heating is structured so that the residential area type (and heating
35
degree day) is only taken into account in the municipality level, and the
primary heating type within each municipality in the 1 km × 1 km level.
Table 3.4: The annual wood uses of houses different primary heating. The
numbers are based on an annual wood use survey conducted by the Finnish
Forest Research Institute (METLA) (Torvelainen, 2009)
Primary heating
Annual wood use
(m3 /house)
District heating
1.4
Oil
2.9
Electricity
3.3
Other
6.9
The emission distribution for primary wood heating in the farmhouses is
altogether different. As with the two categories, the distribution weighting
factors are based on the National building and dwelling register. A 250 m ×
250 m cell is determined to contain a farmhouse if it contains both a detached
house and a agricultural building. Cells satisfying this condition get a weight
value of one. The cells are then aggregated into 1 km × 1 km resolution.
These are the cell weights, and the municipality weights are the sum of the
weights in the municipalities area weighted by the heating degree day factor.
When the emissions have been calculated in the 1 km × 1 km resolution,
the primary wood heating and primary wood heating in the farmhouses are
added together to form the total primary wood heating emissions, and they
are handled together from then on.
3.1.2
Dispersion Modelling and Population Exposure
Assessment
As well as emission modelling, the FRES model includes PM2.5 dispersion and
population exposure modelling. The PM2.5 dispersion modelling is based on
source-receptor matrices calculated with urban dispersion modelling system
(UDM-FMI) developed in Finnish Meteorological Institute (Karppinen et al.,
1998). The system includes Gaussian dispersion model. The emission source
size of the source-receptor matrices is 1 × 1 km2 , temporal resolution 1 hour,
and the assumed emission release altitude for residential wood combustion
7.5 m, which include the initial plume rise. The emission is dispersed into
a 41 × 41 km2 grid square domain, in the centre of which the emission lies.
The monthly and hourly variation is taken into account by temporal patterns
36
shown in figure 3.2. Daily variation is assumed to be constant for residential
wood combustion. (Karvosenoja et al., 2011)
Figure 3.2: Temporal patterns of residential wood combustion emissions for
monthly and hourly variation. (Karvosenoja et al., 2011)
The urban dispersion modelling system also includes MPP-FMI meteorological preprocessing model for meteorological input parameter estimation.
The meteorological preprocessing model takes data from ten synoptic weather
observation stations and two sounding stations as an input. The weather observation stations were chosen to represent different areas in Finland. Based
on these, ten source-receptor matrices, named by the weather observation
station locations, have been calculated. Figure 3.3 shows the allocation of
the different matrices in Finland. The calculated concentrations represent the
increase in the annual average concentration caused by an emission source.
(Karvosenoja et al., 2011)
The dispersion of the PM2.5 emissions is calculated by multiplying the
emission with the source-receptor matrices. Every cell has a source-receptor
matrix appointed to it, defined by the municipality the cell is in. The appropriate source-receptor matrix is placed on top of the emission cell, so that
the emission cell is in the centre. The source-receptor matrix is multiplied by
the emission, and the results are added to the underlying cells’ concentration.
In other words, the total concentration in a cell caused by a certain emission source class is the sum of the emissions multiplied by the appropriate
source-receptor matrix within 41 km × 41 km square grid from the cell:
X
Cc,s =
EMd,s × SRMa,b
(3.5)
d
37
where c to the cell, s refers to the source class, d to the cells within 41 km
× 41 km square grid from the cell c, a to the source-receptor matrix for the
municipality in question, b to the source-receptor matrix element, C is the
concentration in the cell, EM is the emission of a source cell, and SRM is
the appropriate source-receptor matrix.
Figure 3.3: Allocation of dispersion matrices in Finland.
Population exposure to PM2.5 is assessed with population weighted con38
centration (PWC). PWC is calculated by weighting the concentrations by
population densities:
PWC =
1 X
P opc Cc
P op
(3.6)
where P opc is the population of the cell, Cc is the concentration in the cell
and P op is the total population of the area studied (whole Finland in this
case). Population data was based on the locations of permanent residents
from the national building and dwelling register (Mikkola et al., 1999). The
population map for Finland is presented in figure 3.4.
In order to assess how important emissions from different residential areas were, an exposure/emission ratio was calculated. Exposure/emission
ratio represents how big of an exposure a unit emission causes. A big ratio
indicates, that the emissions cause big population exposure. The ratio is
calculated by:
PWC
(3.7)
EER =
EM
where P W C is the total population exposure from the residential area class
and EM is the total emission from the class.
3.1.3
Residential Areas
Finland was divided into seven residential area classes: block building, detached house, scattered detached house, other urban residential, villages, small
villages and rural residential area. Areas that were outside these classes
were defined as non-residential areas. A map of the classes is presented in
figure 3.5.
The residential area classes were based on urban structure monitoring
system (YKR) (Ristimäki, 1999) developed in the Finnish Environment Institute. The criteria for the classes are presented in Table 3.5. The residential
land use in a cell is represented by efficiency. The efficiency is calculated by
dividing the total residential floor space with the land area. The criteria are
applied for each cell, the size of which is 250 m x 250 m. The data was then
aggregated into 1 km x 1 km -resolution. The highest (most densely populated class) determined the class of the aggregated cell. It is notable that
the buildings in the FRES model can be situated in non-residential area.
The PM2.5 emissions were calculated for each residential area class for
primary and supplementary wood heating. The dispersion of the emissions
was modelled as well. From the resulting concentrations, the PWCs caused
by emissions from each class was calculated. In other words, the PWC of the
39
Figure 3.4: Population density (population/km2 ) in Finland.
40
Figure 3.5: Residential area classes in Finland. Zooms are, from top to
bottom, Rovaniemi, Kuopio, Nurmijärvi and Helsinki Metropolitan Area.
class represents the total exposure the emissions from that class cause in the
whole Finland, not only within the class.
41
Table 3.5: The criteria of the residential area classes per cell based on YKR.
Criteria
Block buildings
Residential floor space at least 40% of total floor
space and at least 400 km2 OR floor space at least
20% and at least 1000 km2 ; Efficiency at least 0.02;
block building floor space at least 60% of total residential floor space.
Detached
Same as above, except detached house flood space
at least 40% of total residential floor space.
Scattered detached At least one residential building and efficiency
smaller than 0.02.
Other urban
Not any of the classes above; area with at least 200
residents.
Village
At least six residential buildings in total in the cell
and adjacent cells; non-urban; at least 40 residents.
Small village
Same as above, except at least 20 residents.
Rural
Non-urban, non-village, non-water, at least one residential building within one kilometre.
Non-residential
Non of the above.
42
Chapter 4
Results
The PM2.5 emissions and population exposures were calculated for the eight
residential area classes. The total PM2.5 emissions from RWC was 8230
Mg a-1 , which corresponds 26% of the total PM2.5 emissions of Finland in
2005 (Hildén et al., 2008). The total PWC was 815 ng m-3 . The emissions
and PWCs for primary and supplementary wood heating were 2440 and 5790
Mg a-1 , and 93 and 722 ng m-3 , respectively.
The emissions from supplementary heating were 70% of the total emissions from RWC, but they comprised 89% of the total PWC. The share in
PWC was bigger, because the emissions of supplementary heating occurred
closer to high population density areas in contrast to primary heating.
Primary wood heating caused 30% of the total RWC emissions, but only
about 9% of the detached houses in Finland are primarily heated with wood.
These houses have, therefore, bigger share of the emissions than the building
stock. This reflects the fact that primary wood heating devices, i.e. boilers,
use more wood than supplementary heating devices, i.e. masonry heaters
and stoves, and the average emission factors are relatively high.
The PM2.5 emissions and concentration maps are presented in figures 4.3
and 4.4. The maps show that the spatial distribution of the emissions is
quite different for the primary and supplementary wood heating. The PM2.5
primary heating emissions are distributed somewhat evenly. The emissions
occurred mainly in rural areas, with highest emissions in Western Finland.
For supplementary wood heating, the cities clearly stand out from the
maps. The Eastern and Southern Finland have higher shares than in primary
heating case. As the highest emissions are in cities, they are close to areas
with high population density. Therefore, the emissions have higher impact
on the population exposure. In general the emissions from supplementary
heating are higher than form primary heating in urban areas.
The PM2.5 emissions, PWC caused by them and exposure/emission ratios
43
are presented for different residential area classes in Table 4.1. The shares of
emissions and PWCs from different residential area classes for primary and
supplementary wood heating are presented in figures 4.1(a) and 4.1(b).
(a)
(b)
Figure 4.1: Share of PM2.5 (a) emissions and (b) PWC from residential area
classes for primary and supplementary wood heating.
44
45
Table 4.1: PM2.5 emissions, PWC and exposure/emission ratios from primary and supplementary wood heating from
different residential area classes.
Source
Emission % of total PWC
% of total Exposure/emission ratio
-1
-3
(Mg a )
emissions (ng m )
PWC
(ng m-3 (Mg a-1 )-1 × 1000)
Primary
wood heating
Non-residential
1
<0.1
<0.1
<0.1
8.3
Rural
1150
47
17
19
15
Small villages
130
5
2.3
3
17
Villages
540
22
13
14
24
Other urban
20
1
1.6
2
86
Scattered detached 273
11
16
17
57
Detached
310
13
41
44
130
Block buildings
9
0.4
2.1
2
220
Supplementary
wood heating
Non-residential
3
<0.1
<0.1
<0.1
16
Rural
1730
30
35
5
20
Small villages
200
4
4.8
0.7
24
Villages
920
16
30
4
33
Other urban
70
1
13
2
190
Scattered detached 620
11
59
8
96
Detached
2150
37
540
75
250
Block buildings
100
2
40
6
400
The PM2.5 emissions from primary wood heating were highest in rural and
villages residential area classes, in general more scarcely populated areas,
and slightly lower in more densely populated areas. The emissions from
rural areas were 1150 Mg a-1 which comprises 47% of the total emissions
from primary wood heating. For villages and small villages the emissions
and shares were 540 and 130 Mg a-1 , and 22 and 5%, respectively. Scattered
detached and detached area classes had emissions of 273 and 310 Mg a-1 , with
shares of 11 and 13%, respectively. Each of the other urban, block building
and non-residential areas had less than 1% of the total emissions.
The distribution of the PM2.5 PWCs caused by primary wood heating
differ from the emissions. The areas with higher population densities increase
their significance. The PWCs for rural areas, which had the highest emissions
caused by primary wood heating, was 17 ng m-3 , which was 19% of the total
PWC. Villages and small villages had smaller shares of the total PWC than
the emissions, with PWCs of 13 and 2.3 ng m-3 and shares of 14 and 3%,
respectively. Scattered detached and detached areas increased their shares to
17 and 44% with PWCs of 16 and 41 ng m-3 , respectively. Other urban and
block building areas had PWCs and shares of 1.6 and 2.1 ng m-3 , and 2 and
2 %, respectively. The PWC of non-residential areas was negligible.
For supplementary wood heating, rural and detached residential classes
dominate the PM2.5 emissions. Detached areas had emissions of 2150 Mg
a-1 , which was 37% of the total emissions from supplementary wood heating.
Rural areas were the second biggest contributor, with emissions of 1730 Mg
a-1 and a share of 30%. Villages and small villages had emissions and shares
of 920 and 200 Mg a-1 , and 16 and 4%, respectively. For scattered detached
areas, the emissions were 620 Mg a-1 and the cut from the total 11%. Other
urban and block building areas had emissions that were less than 100 Mg
a-1 and they comprised less than 2% of the total emissions. Non-residential
areas had negligible emissions.
The exposures were concentrated into areas with dense population. Namely
detached residential areas, for which the PWC was 540 ng m-3 . That was
75% of the total PWC from supplementary heating. All the other classes
caused less than ten percent each of the total exposure. Block buildings and
scattered detached had PWCs and shares of 40 and 59 ng m-3 , and 6 and 8%,
respectively. Other urban areas had 2% of the total PWC with 13 ng m-3 .
Villages and small villages comprised 4 and 0.7% of the total PWC, with
30 and 4.8 ng m-3 . Rural areas caused a PWC of 35 ng m-3 , which meant a
share of 5%. Same as with emissions, the PWC of the non-residential areas
was negligible.
The rural, detached and villages area classes were the biggest emission
sources for both primary and supplementary wood heating. In both cases
46
they made up over 80% of the emissions. Only rural areas for primary heating had emissions higher than the four biggest contributor classes in supplementary heating. Each area class had higher emissions from supplementary
heating, the values being from 1.5- to 11-fold.
For PWC, detached areas were the biggest contributors for both heating types. It was also the second most densely populated residential area
class, which partly explained the highest PWC shares. Rural and villages
(i.e. scarcely populated) classes had higher shares in primary heating than
supplementary. Even though detached areas were the biggest contributor
to the PWC from primary heating, the actual PWC was smaller than the
PWC from supplementary heating scattered detached areas, emphasising the
higher exposures caused by supplementary heating. As with emissions, each
area class had higher PWC from supplementary than primary wood heating. Detached areas seemed to be the key source for PM2.5 from population
exposure point of view. Furthermore, supplementary wood heating in these
areas had the biggest impact on population exposure. These factors made
supplementary wood heating in detached areas important target for emissions abatement measures when the reduction of the health impacts of fine
particles are considered.
The exposure/emission ratio tells how big of an exposure a unit emission
causes. The exposure/emission ratios for different residential area classes
for primary and supplementary wood heating are presented in figure 4.2. In
general, the ratios were bigger in areas with higher population densities. The
ratios for non-urban areas (non-residential - villages) were close to each other,
and were much smaller than ratios for urban (other urban - block buildings)
areas. The emissions in non-urban areas exposed only few people due to
scattered population. The highest ratios were (in order from the smallest to
the biggest) for other urban, detached and block buildings. The emissions from
block buildings areas were relatively low, but the exposure/emission ratio
was high due to high population density in those areas. The detached area
had both high emissions and high exposure/emission ratio, highlighting it’s
importance from population exposure point of view. This further indicated
that emission control would be more important in higher population density
areas.
For primary wood heating, the exposure/emission ratios were lower in all
of the area classes compared to supplementary heating. In urban areas, the
ratios for supplementary heating were almost double compared to primary
heating. Therefore, a unit emission caused almost double the population exposure from supplementary heating. This indicated that most emission from
primary heating came from low population density areas, and even the emissions of primary wood heating that occurred in urban environment seemed
47
to occur in areas with lower population density compared to supplementary
heating. This supports the view that the emissions from supplementary wood
heating seemed to be more important from exposure point of view.
48
49
Figure 4.2: PM2.5 exposure/emission ratios for primary and supplementary wood heating in the different residential
classes.
50
Figure 4.3: PM2.5 (a) emissions and (b) concentration caused by primary wood heating.
51
Figure 4.4: PM2.5 (a) emissions and (b) concentration caused by supplementary wood heating.
Chapter 5
Discussion
5.1
Emissions and Spatial Distribution
The FRES model emission factors for residential wood combustion are chosen
to represent real life emissions. According to Tissari (2008), the normal combustion PM1 emission factors for modern and conventional masonry heaters
are 42 and 38-198 mg MJ-1 , respectively. The FRES model emission factors
are 80 and 120, respectively. The FRES emission factors are supposed to take
into account poor combustion, which causes higher emissions. Therefore, the
factors are higher than the smallest measurement values. It is hard to assess how much of the combustion is poor. Furthermore, only one emission
factors is used per appliance. Therefore, the factor has to represent the average emissions as accurately as possible. The emission factors of the FRES
model seem to fall between the extremes of the measurement results. In order
to further developed the values used in the model, for example combustion
practice surveys should be implemented. Since PM1 comprises 90% of the
PM emissions from residential wood heating, the PM1 and PM2.5 emission
factors can be compared to each other. More inaccuracy comes from the
assessment of combustion practices and their effect on the factors than from
the emissions of particles that fall in between 1 and 2.5 µm.
The total PM2.5 emissions of Finland in 2005 were 32 000 Mg a-1 (Hildén
et al., 2008). Thus, the emissions from residential wood combustion, 8230
Mg a-1 , were 26% of the total emissions. It is notable, that not only does
RWC have a big share of the total emissions, but the emissions also occur
within residential areas close to population, which makes them important
to population exposure. Furthermore, the pollutants are emitted from low
altitude, so the particles are less diluted than from high stacks, i.e. from
power plants.
52
The share of PM2.5 emissions from primary wood heating is higher than
the share of houses that are primarily heated with wood. There are approximately 90 000 houses heated primary with wood combustion, and little under
a million detached houses altogether. Where primary wood heating is used
in about 9% of all the detached houses, the emissions were about 30% of the
total emissions. This reflects the fact that the primary heating devices use
annually more wood per device. The emission distribution between primary
and supplementary heating followed the distribution of wood use, as 32% of
the wood fuel is used in primary heating.
The spatial distributions of the emissions from the two cases had a clear
difference. Primary wood heating was rare in big cities (especially in Helsinki
Metropolitan Area), and more common outside urban areas. West Finland
stood out with the highest emissions. Supplementary heating, in contrast,
had the biggest concentrations in the most populated cities. Therefore, the
lower urban wood use per appliance is not enough to compensate the higher
device numbers compared to non-urban areas. In other words, the prevalence
of supplementary wood heating appliances in urban areas cause higher PM2.5
concentrations compared to non-urban areas, even though their wood use per
device is smaller.
A high proportion, 74%, of the primary wood heating emissions occurred
in non-urban areas (non-residential - villages). In contrast, 51% of the emissions from supplementary heating were emitted in urban areas (other urban block buildings). Non-urban areas use more wood per residence, but the sheer
number of the houses is much bigger in urban areas. Primary wood heating
is much more common in non-urban areas than in urban (Torvelainen, 2009).
From population exposure point of view, this can be thought to be good. In
these areas, less people are exposed to the emissions, which, especially in the
cold months, are continuous.
5.2
Population Exposure
The PM2.5 population exposures differed significantly between primary and
supplementary wood heating, in contrast to the emissions. The PWC for
supplementary heating was 7-fold compared to primary heating. The main
reason for this was that more of the emissions of supplementary heating
occurred in more densely populated urban areas. Thus, more people were
exposed to the pollution. This highlights the importance of spatial modelling
of the emissions.
The effect of PM2.5 exposure to mortality can be estimated with the
PWC, exposure-response functions (ERF) and background incidences. Kar53
vosenoja et al. (2012) assumed the mean ERF for mortality to be 0.62%
change in non-accidental mortality per 1 µg m-3 change in PM2.5 concentration. Background non-accidental mortality for the year 2005 was obtained
from Statistics Finland (Official Statistics of Finland (OSF), 2010). Calculating with these assumptions, the total PWC from RWC (815 ng m-3 ) caused
200 premature deaths in 2005.
The CAFE estimate for premature deaths caused by PM2.5 in Finland
in the year 2000 is 1300. Therefore, the RWC would comprise the order
of magnitude 15% of this estimate. The CAFE estimation is mainly from
exposure to long-range transported pollution, as it takes into account the
local sources poorly. Long-range transport pollution cannot be controlled by
local measures. Considering other significant particle emission sources, the
PILTTI project estimated, that traffic caused about 800 premature deaths in
Finland in 2000. Traffic emissions are already controlled by legislation, and
fine particle emissions from traffic exhaust are decreasing. Large plants in
Finland have small impact on the population exposure (Karvosenoja et al.,
2010), because the particles are emitted in high altitude (because of high
stacks), and the emissions are relatively low due to flue gas cleaning.
In total, it can be estimated that RWC causes a significant part of the
total premature deaths from fine particles, with a share of at least 10%. RWC
emissions are not yet controlled by legislation in Finland, and it’s activity is
increasing in the future. Therefore, it’s share and significance is increasing.
In contrast to traffic and long-range transported particles, RWC emissions
have clear potential for abatement. There are undeveloped devices, and
incomplete combustion causes significant emissions.
Urban areas (other urban - block buildings) withheld 25% of the PM2.5
emissions of the primary wood heating, but 65% of the exposure. For supplementary wood heating the figures were 51% and 91%, respectively. From
the total emissions and exposure from RWC, the supplementary heating in
urban areas comprised 36% and 80%, respectively. All these indicate the
fact that supplementary stove heating in densely populated areas should be
in the focus when emission impacts and reductions are considered.
The exposure/emission ratios reflect the fact that the emissions in areas
with higher population density cause bigger population exposures. Furthermore, they indicate that supplementary wood heating is more significant
than primary heating for the population exposure. An emission of the same
amount causes higher population exposure from supplementary heating than
primary. This highlights the importance to concentrate the emission abatement measures to supplementary heating, i.e. masonry heaters and stoves.
54
5.3
Emission Reduction
Since the population exposure to PM2.5 from supplementary wood heating
is 7-fold compared to primary heating, it would be more efficient to target
the reduction measures to supplementary heating. Since the renewing rate of
masonry heaters is so slow, most feasible option for the reduction would be
the enhancement of the operation practices and flue gas cleaning. As flue gas
cleaning technologies are not yet widely commercially available, information
campaigns and legislation measures are the the most promising ways to affect
the emissions in the short term.
Since urban areas cause most of the population exposure to PM2.5 , the
information campaigns and legislation measures would be most efficient when
targeted there. The measures could take into account the population densities of the residential areas. Areas with higher density could have stricter
limits or have more emphasis in education.
Masonry heaters are the most common supplementary wood heating devices alongside with sauna stoves. Masonry heaters are typically made out of
masonry products, so once they are build, they tend to stay put. Therefore,
new appliances with lower emissions would affect the country-wide emissions
slowly. Improvement of operational practices would be the most viable reduction method. Especially fuel quality should receive attention. Burning
wet wood or waste in the device produces high emissions. Sauna stoves are
possibly renewed quicker, as they are easier to replace. For them, new appliances with set emission limits would be efficient way to cut down emissions.
Operational practices are important for sauna stoves as well.
Boilers are typically used with continuous combustion. This means that
the user affects the combustion conditions less than in batch combustion.
However, especially fuel quality (for example moisture content) has significant effect on the emissions, and poor quality can cause multifold emissions.
Therefore, education for boiler users could be important emission reduction
method. Technology wise, an important issue is the use of heat accumulator
tanks. Boilers without heat accumulator might be used with smouldering
combustion, which causes high emissions. Equipping all boilers with heat
accumulators would be a cost-effective emission reduction method. Flue gas
cleaning technologies, which are an option in the future, work better with
continuous combustion devices, and offer notable emission reduction potential.
Boilers or stoves in Finland don’t yet have binding emissions limits. However, the European Union’s Ecodesign Directive would affect Finland also.
The Directive is going to set emission limits to small scale wood combustion
55
devices. The effect of the Directive to the emissions depends on the renewal
rate of the devices.
Residential wood combustion has much emission reduction potential compared to other important particle emission sources. Traffic emissions are already controlled by legislation measures, and the traffic exhaust emissions
are decreasing. The emissions of big plants in Finland are controlled by legislation, and they have small impact on the population exposure to particles.
Long-range transported particles cause significant population exposure to
particles, but it cannot be controlled by local or national measures. Therefore, residential wood combustion should receive the attention of decisionmakers in order to decrease the adverse health impacts of fine particles.
56
Chapter 6
Conclusions
The goal of this thesis was to identify the characteristics of Finnish residential
wood combustion (RWC), study the PM2.5 emissions and the population
exposure they cause from different residential area types, and assess the
emission reduction options for the future. RWC was found to be an important
source for PM2.5 emissions, and it caused significant population exposure to
PM2.5 .
The total PM2.5 emissions from residential wood heating was 8230 Mg
-1
a , which amounted to 26% of the total emissions in Finland in 2005. Supplementary wood heating, i.e. stoves and masonry heaters, caused 70% of
these.
Emissions were studied in different residential area classes in order to
identify the most important source areas. Non-urban areas, with a share
of 74%, were more important emission source than urban for primary wood
heating, i.e. boilers. For supplementary wood heating, the urban areas made
up 51% of the emissions.
The picture of population exposure to PM2.5 was quite different from the
emissions. Supplementary heating caused 89% of the total PM2.5 exposure
from RWC, with 80% of the total exposure coming from urban areas. In total,
RWC was estimated to have caused around 200 premature deaths in 2005.
From population exposure point of view, supplementary wood heating with
stoves and masonry heaters is much more significant than primary heating
with boilers.
Since the RWC is increasing in the future, the reduction of the emissions is important for the public health. The reduction measures should be
targeted at supplementary wood heating, as it comprises the majority of the
population exposure. Supplementary wood heating has two main combustion
appliance types: masonry heaters and sauna stoves. For sauna stoves, the
renewal of the devices may offer emission reductions in the near future.With
57
sauna stoves, consumers should be directed to invest in new, low emission
technology. New stoves should also have legally binding emission limits. The
renewal of masonry heaters offers emission reduction potential as well. However, their renewal rate is slow, so the short term reductions might be small.
A viable way to reduce the emissions in the sort term could be to affect the operational practices. The emissions reduction potential seems to
be remarkable, as smouldering combustion causes significant increase in the
emissions. A potential way to affect the practices is through an information
campaign. Although the influence of such a campaign is hard to assess, it
is a cost-effective option even with small effects compared to other emission
reduction measures.
In the future, more studies are needed about how users operate their
RWC appliances. It would be crucial for the emission and health assessment,
as the emission factors and emission reduction potential could be estimated
with higher accuracy. Furthermore, residential wood combustion should be
studied more with combined health and climate effects assessment. The
increase of wood combustion is promoted as a measure to battle the climate
change. However, this is likely to increase the adverse health effects. A costbenefit analysis between different scenario options would shed light to the
real benefits of wood combustion promotion.
58
Bibliography
Ahtiniemi P., Tainio M., Tuomisto J. T., Karvosenoja N., Kupiainen K.,
Porvari P., Karppinen A., Kangas L., and Kukkonen J. Health risks from
nearby sources of fine particulate matter: Domestic wood combustion and
road traffic (piltti), 2010. Helsinki University print, Helsinki, Finland.
Alakangas E. Properties of wood fuels used in Finland - BIOSOUTH -project,
2005. Technical Research Centre of Finland, VTT.
Alakangas E., Erkkilä A., and Oravainen H.
Tehokas ja
ympäristöä säästävä tulisijalämmitys (in finnish), 2008a.
VTT,
http://www.biohousing.eu.com.
Alakangas E., Erkkilä A., and Oravainen H.
Tehokas ja
ympäristöä säästävä tulisijalämmitys: Polttopuun tuotanto ja käyttö,
2008b. VTT.
CAFE. The communication on thematic strategy on air pollution, 2005.
Commission of the European Communities.
Dasch J. M. Particulate and gaseous emissions from wood-burning fireplaces.
Environ Sci Technol, 16:16:636–45, 1982.
ErnieAndErica.
Ernie
and
erica
wisner:
rocketstoves
http://www.ernieanderica.info/rocketstoves. visited: 20.04.2012., 2012.
Frey A. K., Tissari J., Saarnio K., Timonen H. J., Tolonen-Kivimäki O.,
Aurela M. A., Saarikoski S. K., Makkonen U., Hytönen K., Jokiniemi
J., Salonen R. O., and Hillamo R. Chemical composition and mass size
distribution of fine particulate matter emitted by a small masonry heater.
Boreal Environ. Res., 14:255–271, 2009.
Gröndahl T., Makkonen J., Myllynen M., Niemi J., and Tuomi S. Tulisijojen käyttö ja päästöt pääkaupunkiseudun pientaloista (in finnish). HSY
Julkaisuja, page 39, 2011.
59
Hartmann H., Turowski P., and Kiener S.
Electrostatic precipitators for small-scale wood combustion systems - results from
lab- and field tests.
In IEA Task32-Workshop: ”Fine particulate emissions from small-scale biomass furnaces”. Central European Biomass Conference (CEBC), 26.-28.January 2011 in Graz
(www.biomasseverband.at/uploads/tx osfopage/WSF 5 Hartmann.pdf ),
2011.
Harvia. Harvia sauna - harvia m3. www.harvia.fi/content/fi/40/103/harvia
m3.html. visited: 19.04.2012., 2012a.
Harvia.
Harvia
sauna
harvia
greenfalme
http://www.harvia.fi/content/fi/40/4161/harvia%20greenflame.html.
visited: 20.04.2012., 2012b.
Hildén M., Karvosenoja N., Koskela S., Kupiainen K., Laine A., Rinne J.,
Seppälä J., Savolahti M., and Sokka L. Pitkän aikavälin ilmasto- ja energiastrategian ympäristöarviointi (in finnish). Suomen Ympäristö, 50:90,
2008.
Hukkanen A., Kaivosoja T., Sippula O., Nuutinen K., Jokiniemi J., and
Tissari J. Reduction of gaseous and particulate emissions from small-scale
wood combustion with a catalytic combustor. Atmospheric Environment,
50:16–23, 2012.
Karppinen A., Kukkonen J., Nordlund G., Rantakrans E., and Valkama I. A
dispersion modelling system for urban air pollution. Finnish Meteorological
Institute, Publications on Air Quality 28, Helsinki, 1998.
Karvosenoja N., Tainio M., Kupiainen K., Tuomisto J., and Kukkonen J.
Evaluation of the emissions and uncertainties of pm2.5 originated from
vehicular traffic and domestic wood combustion in finland. Boreal Environment Research, 13:465–474, 2008.
Karvosenoja N., Kupiainen K., Paunu V.-V., Savolahti M., Tohka A., Kangas
L., Kukkonen J., and Tuomisto J. T. Cost-efficient reduction of population
exposure caused by primary pm2.5 emissions in finland. Abstract book,
15th International Union of Air Pollution Prevention and Environmental
Protection Associations’ (IUAPPA) World Clean Air Congress ”Achieving
Environmental Sustainability in a Resource Hungry World”, September 1216, 2010 Vancouver, British Columbia, Canada, page 36, 2010.
60
Karvosenoja N., Holmberg M., Vanhala P., Paunu V.-V., Tainio M., and
Kukkonen J. Reduction of co2 by residential wood heating and tradeoff caused by increased population exposure to primary om2.5 in finland.
In 8th International Conference on Air Quality Science and Application,
short paper, USB-memory stick, 2012.
Karvosenoja N. Emission scenario model for regional air pollution. Monographs of the Boreal Environment Research, 32, 2008.
Karvosenoja N., Kangas L., Kupiainen K., Kukkonen J., Karppinen A.,
Sofiev M., Tainio M., Paunu V.-V., Ahtoniemi P., Tuomisto J. T., and
Porvari P. Integrated modeling assessments of the population exposure in
finland to primary PM2.5 from traffic and domestic wood combustion on
the resolutions of 1 and 10 km. Air Qual Atmos Health, 4:179–188, 2011.
Lämpömaa.
Lampömaa:
Kamiina,
kevyttakka
bari.
www.lampomaa.fi/kevyttakat+kamiinat/hase. visited:
20.04.2012.,
2012.
Mikkola A., Jaakkola O., and Sucksdorff Y. Slices-projekti eli valtakunnallisen maanmaankäyttö-, peitteisyys- ja maaperäineistojen muodostuminen
(the slices project: National classification of land use, land cover and soil,
and the production of databases (in finnish), 1999.
Official Statistics of Finland (OSF). Causes of death causes of death [epublication]. issn=1799-5078, 2010.
Paunu V.-V., Karvosenoja N., and Kupiainen K. Residential wood combustion in finland and emission factor database. In 8th International Conference on Air Quality Science and Application, short paper, USB-memory
stick, 2012.
Pekkanen J. Elin- ja ympäristön riskit suomessa. Ympäristö ja terveys, 3,
2010.
Pellettipojat.
Pellettipojat: Uusiutuvaa energiaa - pellettikattilat.
http://www.pellettipojat.fi/page6.php. visited: 20.04.2012., 2012.
Pleijel H. Transboundary air pollution: Scientific understanding and environmental policy in europe, 2007. Pozkal, Poland. ISBN 978-91-44-00471-6.
Pleijel H. Air pollution and climate change: Two sides of the same coin?
2009. Solna, Sweden. ISBN 978-91-620-1278-6.
61
Ristimäki M. Yhdyskuntarakenteen seurantajärjestelmä. ehdotus yhdyskuntarakenteen seurannan järjestämiseksi ja kehittämiseksi (in finnish), 1999.
Salonen R. O. and Pennanen A. Pienhiukkasten vaikutus terveyteen. tuloksia ja päätelmiä teknologiaohjelmasta fine pienhiukkaset - teknologia,
ympäristö ja terveys (in finnish), 2006. Work in progress. Expires June,
2004.
Savolahti M., Kupiainen K., and Karvosenoja N. Reduction potentials of
black and organic carbon emissions by technical and nontechnical measures
in finnish residential wood combustion. In 8th International Conference
on Air Quality Science and Application, short paper, USB-memory stick,
2012.
Statistics Finland. Energy statistics yearbook 2010, 2010.
Stern C. H., Jaasma D. R., Shelton J. W., and Satterfield G. Parametric
study of fireplace particulate matter and carbon monoxide emissions. Air
Waste Manage. Assoc., 42:777–783, 1992.
Tainio M., Karvosenoja N., Porvari P., Raateland A., Tuomisto J. T., Johansson M., Kukkonen J., and Kupiainen K. A simple concept for gis-based
estimation of population exposure to primary fine particles from vehicular
traffic and domestic wood combustion. Boreal Environment Research, 14:
850–860, 2009.
Tissari J. Fine Particle Emissions from Residential Wood Combustion. PhD
thesis, Kuopio University Publications C. Natural and Environmental Sciences 237, Kuopio University, Kuopio, Finland, 2008.
Tissari J., Raunemaa T., Jokiniemi J., Sippula O., Hytönen K., Linna V.,
Oravainen H., Pyykönen J., Tuomi S., Vesterinen R., Taipale R., Kolsi A.,
Nuutinen I., Kouki J., and K. Vuorio K. Puun polton pienhiukkaspäästöt.
loppuraportti, 2005.
Tissari J., Hytönen K., Lyyränen J., and Jokiniemi J. A novel field measurement method for determining fine particle and gas emissions from residential wood combustion. Atmospheric Environment, 41:8330–8344, 2007.
Tissari J., Hytönen K., Sippula O., and Jokiniemi J. The effects of operating
conditions on emissions from masonry heaters and sauna stoves. Biomass
and Bioenergy, 33:513–520, 2008a.
62
Tissari J., Lyyränen J., Hytönen K., Sippula O., Tappe U., Frey A., Saarnio
K., Pennanen A. S., Hillamo R., Salonen R. O., Hirvonen M.-R., and
Jokiniemi J. Fine particle and gaseous emissions from normal and smouldering wood combustion in a conventional masonry heater. Atmospheric
Environment, 42:7862–7873, 2008b.
Torvelainen J. Pientalojen polttopuun käyttö 2007/2008 (in finnish). Metsatilastotiedote (SVT Maa-, metsä ja kalatalous), 26:3, 2009.
63
Download PDF
Similar pages