• Ei tuloksia

Reduction of fine particle emissions from small-scale wood combustion using a novel heat exchanger system

N/A
N/A
Info
Lataa
Protected

Academic year: 2022

Jaa "Reduction of fine particle emissions from small-scale wood combustion using a novel heat exchanger system"

Copied!
66
0
0

Kokoteksti

(1)

REPORT SERIES IN AEROSOL SCIENCE

No. (2020)

REDUCTION OF FINE PARTICLE EMISSIONS FROM SMALL-SCALE WOOD COMBUSTION USING A NOVEL

HEAT EXCHANGER SYSTEM

JULIJA GRIGONYT -LOPEZ RODRIGUEZ

Department of Environmental and Biological Sciences Faculty of Science and Forestry

University of Eastern Finland Kuopio, Finland

Academic dissertation

To be presented, with the permission of the Faculty of Science and Forestry of the University of Eastern Finland, for public criticism in auditorium SN200, Snellmania Building at Yliopistonranta 1E, on December 4th, 2020, at 12 o'clock noon.

Kuopio 2020

(2)

Author’s Address: Fine Particle and Aerosol Technology Laboratory University of Eastern Finland

Department of Environmental and Biological Sciences P.O. Box 1627

FI-70211 Kuopio, Finland e-mail: julija.grigonyte@uef.fi

Supervisors: Professor Jorma K. Jokiniemi, Ph.D.

University of Eastern Finland

Department of Environmental and Biological Sciences

Associate Professor Olli Sippula, Ph.D.

University of Eastern Finland

Department of Environmental and Biological Sciences and Department of Chemistry

Reviewers: Associate Professor Mika Järvinen, Ph.D.

Aalto University

Department of Mechanical Engineering

Docent Patrik Yrjas, D.Sc.

Abo Akademi University Process Chemistry Center

Opponent: Doctor Terttaliisa Lind, Ph.D.

Paul Scherrer Institute

Nuclear Energy and Safety Research Department Villigen, Switzerland

20 Unigrafia Oy

ISBN 978-952-7276-44-0 (pdf version) Helsinki 2020

http://www.atm.helsinki.fi/FAAR

(3)

Acknowledgements

To my grandparents Nina and Anatolij and to my aunt Marina

This thesis work was carried out in the Department of Environmental and Biological Sci- ences at the University of Eastern Finland in Kuopio. The financial support for the study was provided by the Fortum Foundation personal grants 2012-2015.

I want to thank my supervisor Professor Jorma Jokiniemi, whom first accepted me to his group during my master’s studies, done partially at University of Eastern Finland within the Erasmus student exchange program. I also thank you for letting me to continue my PhD studies and for allowing me to get involved in many interesting projects at Fine Particle and Aerosol Technology Laboratory. I am also grateful to my co-supervisor Associate Professor Olli Sippula for his important contribution to improve my scientific writing skills and for explaining me the fundamental principles of combustion physics.

I would like to convey my deepest gratitude to Associate Professor Mika Järvinen and D.

Sc. Docent Patrik Yrjas for the revision of my thesis book. Your comments and feedback gave me new perspectives and broadened my ways of thinking about my scientific work. I am also grateful to Dr. Terttaliisa Lind for kindly accepting the invitation to act as the op- ponent in the public examination of my thesis.

I thank all co-authors and former colleagues at Fine Particle and Aerosol Technology La- boratory for your significant contribution to this work. I am grateful for your enthusiasm and fun moments during our measurement campaigns and coffee breaks. Kiitokset!

I am also grateful to all my friends whom actively participated in my life all these years. We shared many nice experiences and tough moments together. Our meetings, dinners, never ending laughter and simply talks pained my life in different hues. A i ! Kiitos! Gracias!

C o!

Thank you to my Cuban family Mercedes, Erasmo, Yumelkys, Daniela and Carly. I am grateful for your support and love. Gracias!

Huge thank you to my grandparents Michail and Neli for always waiting me coming from Finland and sharing nice talks with the cup of herbal tea and sweets. Thank you for your

support! o!

I am very thankful to my family, especially to my mom Svetlana, my dad Anatolij and my brother Denis. Geographically you are far, but no distance will weaken our bonds. You are always on my mind. You are one of the most important people that have shaped and molded me into the person I am today. Thank you for that endless support and your countless en- couragements to do and achieve my best. Also, I want to say whohohowho ('Thank you!' in dog language) to Uran - our family member and not less. You go, boy!

My endless gratitude goes to my grandmother Nina, my grandfather Anatolij and aunt Ma- rina. This work might have never been finished if were not for you. You have been my support team since the early 2006, when I moved in with you for my studies at Kaunas University of Technology. Thank you for always insisting that education is the most

(4)

powerful tool to change our life for good. You are the light of my soul. Now that the work is done, I want to say thank you for always pushing me to finalize my thesis book, even when I have been professionally engaged in a different work for more than 4 years.

!

Finally, to my support, my accomplice, my love, my husband Maykel. Thank you for be- lieving in me, even when there’ve been times in which I did not believe in myself. Thanks for this wonderful story we are living in; we started it together and both finished this path of our life together. Hand to hand and soul to soul we are now getting into new live experi- ences. Thank you for being the generator of my happiness and for taking me out of the bad moments as fast as they are noticed. Thanks for the endless motivation and encouragements and for showing me that things are not so serious as they might seem to be, even when they are. Muchas gracias mi amor. Te quiero!

11th of October, 2020 Julija Grigonyt

(5)

Reduction of fine particle emissions from small-scale wood combustion using a novel heat exchanger system

Julija Grigonyt -Lopez Rodriguez University of Eastern Finland, 2020 Abstract

The constrained use of fossil fuels for energy production and the growing demand for re- newable energy are the main targets for addressing climate change. Climate change policy remains steadily focused on reducing greenhouse gas emissions via several approaches, such as pricing carbon, increasing renewable energy needs and boosting energy efficiency. An alternative approach for constraining emissions that are produced by fossil fuels is increas- ing biomass utilization for energy supply. However, biomass combustion produces substan- tial fine particulate matter (PM1) emissions, especially in small-scale units that are operated with poor-quality fuel under poor combustion conditions and with no flue gas control sys- tem. Fine particles that are emitted into the ambient air influence radiative forcing and neg- atively impact health. Therefore, the effective mitigation of particle emissions from residen- tial combustion units is possible by establishing emission limit values for small-scale com- bustion and/or utilizing secondary flue gas cleaning technologies.

The primary objective of this thesis was to evaluate the potential application of a novel heat exchanger system to fine particle reduction in small-scale wood combustion appliances. The heat exchanger system included two types of emission abatement equipment, namely, a con- densing heat exchanger and a sonic shield charger, where the fine particles were retained mainly by thermophoretic and electrostatic forces. The physicochemical fine particle emis- sions from a 40-kW boiler that was equipped with a heat exchanger system were studied and compared to the fine particle emissions from a commercial firetube boiler. The second main objective was to compare the physicochemical properties of gaseous and fine particle emissions (dp< 1 m) from continuous and batch combustion appliances.

In continuous combustion, the fine particle emissions were in the range of 15-23 mg/MJ when combusting pellets and wood chips. The chemical composition of the fine particles was dominated by macroelements K, Na, and S. In addition, the PM1 that was obtained from the biomass combustion was enriched with microelements, such as Cl and Zn. The PM emissions were composed mostly of alkali metals, which is one of the features that enabled the realization of efficient combustion.

The fine particle emissions from the batch-wise combustion of softwood (spruce) and hard- wood (beech and birch) were in the range of 49-90 mg/MJ. The spruce combustion gener- ated the lowest PM1 emissions in comparison to hardwood combustion. The dominant frac- tions of fine particle matter from spruce combustion were elemental carbon (~50% of the total PM1 mass), organic carbon and inorganics (Na, Zn, Fe and K).

The fine particle emissions from modern boilers were moderate in comparison to batch-wise combustion in a masonry heater; however, to further minimize the PM1 emissions, second- ary measures were investigated.

The use of a condensing heat exchanger decreased PM1 by 40% in comparison to a conven- tional boiler. The integrated condensing heat exchanger and shielded corona charger system realized an 82% PM1emission reduction efficiency in comparison to a conventional boiler.

(6)

The highest PM1 reduction efficiency was 88%, which was realized by a shielded corona charger vs. a conventional boiler.

Fine particle emission reduction in small-scale wood combustion appliances is important and is becoming even more relevant due to the establishment of more stringent emission limits by the ecodesign directive, which will come into force soon. Therefore, emission abatement technologies and their applications in various small-scale combustion units should be further investigated, as they might provide a cost-feasible and efficient alternative for the adjustment of available boilers or stoves.

Keywords: Small-scale, Biomass, Combustion, Heat exchanger, Fine particles, Pellets, Electrical charging, Corona discharge, Thermophoresis, Masonry heater, Wood logs

(7)

Contents

1 Introduction ... 11

2 Objectives of the study ... 13

3 Scientific background ... 14

3.1 Fuels ... 14

3.2 Biomass combustion ... 15

3.3 Automatic fixed-bed combustion in small-scale biomass appliances ... 16

3.4 Batch combustion appliances ... 18

3.5 Combustion emissions ... 19

3.5.1 Particle formation during biomass combustion ... 19

3.5.2 Gaseous emissions ... 22

3.6 Flue gas cleaning devices in small-scale wood-fired appliances ... 23

3.6.1 Introduction to heat exchangers ... 24

3.6.2 Electrostatic precipitator and modifications ... 26

4 Methods ... 29

4.1 Combustion appliances ... 30

4.2 Condensing heat exchanger. Design and operating principle ... 32

4.3 Shielded corona charger. Design and operating principle ... 33

4.4 Particle measurements, characterizations and chemical analyses ... 34

4.4.1 Observation of the particle morphology ... 36

4.5 Measurement of gaseous emissions ... 36

4.6 Condensed water analyses ... 36

5 Results and discussion ... 37

5.1 Emission characteristics of combustion appliances ... 37

5.1.1 Experimental conditions in continuous and batch combustion appliances .. 37

5.1.2 Fine particle emissions and chemical composition in continuous and batch combustion appliances ... 39

5.1.3 Particle mass and number emissions and particle sizes in continuous and batch combustion appliances ... 41

5.1.4 Fine particle morphology and composition in continuous and batch combustion ... 41

(8)

5.2 Condensing heat exchanger and shielded corona charger systems as emission

reduction technologies ... 44

5.2.1 Fine particle reduction efficiencies in a novel condensing heat exchanger- shielded corona charger system ... 44

5.2.2 Effects of the particulate matter reduction systems on the PM1 emission chemical composition ... 46

5.2.3 Effects of particulate matter reduction systems on the number and mass size distributions ... 47

5.2.4 Particle morphology modification with the condensing heat exchanger- shielded corona charger system ... 50

5.2.5 Water balance, condensed water chemical composition and overall system thermal efficiency ... 51

6 Review of papers and the authors’ contributions ... 53

7 Conclusions and motivation ... 54

References ... 56

(9)

List of publications

This thesis consists of an introductory review, followed by four research articles. In the introductory part, the papers are cited according to their roman numerals. The publications are reproduced with the kind permission of respective journals.

I Grigonyte, J., Nuutinen, I., Koponen, T., Lamberg, H., Tissari, J., Jokiniemi, J., and Sippula, O. Evaluation of a Heat Exchanger Designed for Fine Particle Precipitation in Small-Scale Wood Combustion, Energy Fuels, Vol 28, Pages 6058-6065, 2014.

II Grigonyt -Lopez Rodriguez, J., Suhonen, H., Laitinen, A., Tissari, J., Kortelainen, M., Tiitta, P., Lähde, A., Keskinen, J., Jokiniemi, J., Sippula, O. A Novel Electrical Charging Condensing Heat Exchanger for Efficient Particle Emission Reduction in Small Wood Boilers, Renewable Energy, Vol 145, Pages 521-529, 2020.

III Leskinen, J., Ihalainen, M., Torvela, T., Kortelainen, M., Lamberg, H., Tiitta, P., Jakobi, G., Grigonyte, J., Joutsensaari, J., Sippula, O., Tissari, J., Virtanen, A., Zim- mermann, R., and Jokiniemi, J. Effective Density and Morphology of Particles Emit- ted from Small-Scale Combustion of Various Wood Fuels, Environ. Sci. Technol, Vol 48, Pages 13298-13306, 2014.

IV Kortelainen, M., Jokiniemi, J., Tiitta, P., Tissari, J., Lamberg, H., Leskinen, J., Grigonyte-Lopez Rodriguez, J., Koponen, H., Antikainen, S., Nuutinen, I., Zim- mermann, R. and Sippula, O.: Time-resolved chemical composition of small scale batch combustion from various wood species, Fuel, Vol 233, Pages 224-236, 2018.

(10)
(11)

11

1 Introduction

The limitation and regulation of greenhouse gases (GHGs) constitute one of the most sub- stantial international climate challenges (UNFCCC, 1992; Kyoto Protocol, 1997) with the prospects of enhancing the use of the renewable energy sources (RES), replacing fossil fuels and pursuing a more ambitious limit of 1.5 °C to avoid global warming, as established by the Paris Agreement (UNFCCC, 2015).

The RES Directive (Directive 2009/28/EC) for the EU specified a target of 20% of the total energy needs being supplied by renewables by 2020 (Finland 38%). The revised RES Pro- posal 2016 (Proposal for RES, 2016) sets the target as a final energy consumption of at least 27% renewables in the EU-27 by 2030.

Finland is among the leading countries in the use of biomass in energy production (EEA Report 2019). Finland’s target for the share of renewable energy is 38% of the final energy consumption in 2020, and this target was already reached in 2014. The use of biomass for energy can contribute to reduction of carbon dioxide emissions (one of the most important GHGs) and to energy security; however, it might have adverse effects on the environment and on human health.

Small-scale wood-fired heating systems for room heating and warm water supply are rec- ognized as one of the most significant sources of fine particulate matter emissions, namely, emissions of particles with an aerodynamic diameter (the aerodynamic diameter of irregular particle is defined as diameter of the spherical particle with a density of 1000 kg/m3 that has same settling velocity as the irregular particle) of <1 m (PM1) or <2.5 m (PM2.5), in many European regions (Johansson et al., 2004, Hellen et al., 2008, Plejdrup et al., 2016, Czech et al., 2018). The main reasons for high PM emissions are associated with incomplete com- bustion conditions as a consequence of the use of old and/or un-optimized wood-fired ap- pliances (Johansson et al 2004), wood quality and size (moisture content, compact or non- compact fuel, and chemical composition), insufficient air supply and/or uneven fuel mass load. For instance, modern wood-fired appliances typically make lower contributions to PM emissions (Johansson et al., 2004, Bäfver et al., 2011); however, they may also require sup- plemental improvements. In addition, based on environmental regulations and political de- cisions, the increase in biomass usage in small-scale appliances might significantly increase fine particle emissions in the near future. Therefore, this action is counterproductive regard- ing the further increase of biomass utilization in the small-scale heating sector since resi- dential biomass combustion for the public is presented as a sustainable, environmentally friendly and CO2-neutral heating technology.

PM (PM2.5 and PM1) that originates from biomass combustion consists of three main types of particles: carbonaceous solid material (soot), condensable organic compounds (tar) and inorganic particles from ash constituents (salts). According to epidemiological studies (Bol- ling et al., 2009, Arif et al., 2017), fine particles pose a risk to human health, especially to the health of the respiratory system, as they cause lung cancer and chronic lung and cardio- vascular diseases and can have additional carcinogenic effects (IARC, 2010). However, no

(12)

12

direct marker of wood burning is available for associating emissions and effects directly (WHO, 2013).

Particulate matter is found to be a substantial problem for both air quality and climate change, and it represents an uncertain factor in current climate scenarios (Fuzzi et al., 2015).

PM from combustion sources affects the climate through direct and indirect effects. The influence of fine particles on the climate depends on their physicochemical properties (Fuzzi et al., 2015). The direct, namely, warming, impact depends on the ability of the fine particles (e.g., EC or soot) to absorb sunlight and on whether they are in the air or in the surface of snow or ice. The cooling effect of the particles (e.g., sulphate particles) is linked with the cloud’s ability to reflect sunlight and occurs due to sunlight scattering. The indirect, or cool- ing, effect depends on the cloud condensation nuclei; the more suitable the morphology (particle size 0,1 m and spherical particle shape) of the condensation nuclei in the atmos- phere, the more cloud droplets there will be, and the better the cloud that reflects the light is formed (Chang et al., 2017; Hasekamp et al., 2019).

Nevertheless, no constrained emission limits are specified in the legislations or directives for small-scale appliances, in contrast to large-scale and industrial boilers (Nordic-Ecolabel- ling, 2014). However, regarding ecodesign requirements for solid-fuel boilers Directive 2009/125/EC sets more stringent PM emission limits for small-scale wood-fired appliances.

For instance, a new German ecolabel, namely, Blue Angel (Blue Angel, Stoves for wood 2020), is an example of tightened emission regulations for small-scale combustion appli- ances on the national level. It is a voluntary label; however, it is attractive to manufacturers for the production of efficient and low-emission heating appliances. Consequently, the stricter and more defined emission limit values will raise interest in flue gas filtering systems and their installation in small-scale wood-fired appliances. Secondary PM emission cleaning technologies that are used in large-scale combustion units include scrubbers, bag filters (fab- ric filters), electrostatic precipitators (ESPs), heat exchangers and catalytic converters. How- ever, only some are feasible for implementation in residential boilers due to their size, maintenance and cost-efficiency.

(13)

13

2 Objectives of the study

The objectives of this thesis were to evaluate the potential application of the condensing heat exchanger and shielded corona charger for fine particle reduction in small-scale wood combustion and to compare the emissions from various residential combustion appliances.

The individual objectives can be summarized as follows:

Paper I:

to evaluate the potential application of the novel heat exchanger to fine particle precipitation and latent heat recovery in small-scale wood com- bustion;

to determine the effects of the novel heat exchanger system on the PM physicochemical properties of a wood-fired boiler; and

to measure the particle reduction efficiency and ability to keep the con- densing heat exchanger walls clean by utilizing the scrubbing unit.

Paper II:

to investigate the efficiency and feasibility of the condensing heat ex- changer and shielded corona charger combined system in fine particle emission reduction;

to compare both the condensing heat exchanger and the sonic jet charger system with a reference firetube boiler;

to evaluate the performance of the particle charging process in decreasing emissions; and

to measure the effects of flue gas cleaning systems on emission character- istics.

Paper III:

to compare the particle physicochemical properties from pellet and batch combustion.

Paper IV:

to determine the time-resolved emissions of batch combustion.

(14)

14

3 Scientific background

3.1 Fuels

Biomass is a carbon-based renewable fuel that consists mainly of C (48-50%), H (6-6.5%), and O (38 -42%) (Parmar et al., 2017). In addition, biomass contains small amounts (< 1%) of N, which is the source of NOx; S, which is the source of SOx; and ash components such as Si, Al, Fe, Ti, Ca, Mg, Mn, K, P, S, Cl, and Na (Vassilev et al., 2017), (Figure 1).

The biomass composition and the boiler design and operation influence the composition of the ashes and, therefore, their ultimate behaviour. Fuels with large content of sulphur may cause low-temperature corrosion (70-400 °C) on the heat-exchanger surface via the conden- sation of acids and hygroscopic salts (Retschitzegger et al., 2015). Moreover, high-HCl and alkali salts (NaCl and KCl) may cause both high-temperature (450-900 °C) (Wiinikka et al., 2007) and low-temperature superheater corrosions (Nielsen et al., 2000, Enestam et al., 2013). Fuels with high contents of silica (SiO2) and alkali metals (Na and K) can form K/Na silicates, which induce slagging on the boiler surface (Jenkins et al., 1998). Slagging occurs in the boiler section that is directly exposed to the flame. Moreover, alkali metals (Na and K) can together with S, Cl and sometimes P, form eutectic salt mixtures, which may lower the melting point of the already formed ashes and cause slagging and fouling of the super- heater surfaces (Nutalapati et al., 2007, Melissari et al., 2014), which is mainly due to the condensation of volatile species that are vaporized in the previous boiler sections (Tortosa- Masia et al., 2005).

Wood fuel is suitable for combustion due to its low ash and nitrogen contents, in contrast to herbaceous biomasses, such as straw, miscanthus, and switch grass, which contain large amounts of N, S, K, and Cl and lead to higher emissions of NOx and particulates and in- creased ash, deposits and corrosion. Moreover, woody biomass is more suitable for smaller plants than herbaceous fuel (Nussbaumer et al., 2003). The typical wood forms for contin- uously fired small-scale appliances are pellets, wood chips, manufactured mineral fuel bri- quettes and wood logs. Pellets are a free-flowing material (Swietochowski et al., 2018) with a low moisture content (8 - 10%) and consistent size and geometry. Wood chips are a me- dium-sized solid material and are produced by cutting or chipping larger wood parts; the moisture content is typically 30-50% (Alakangas et al., 2016). Commonly used types of wood logs in small-scale appliances are beech, birch and spruce, which have a typical mois- ture content of 10 –20% (W179 Wood products information; Phyllis2 database for biomass).

(15)

15

Figure 1: Average composition of wood (Alakangas et al., 2016).

3.2 Biomass combustion

The biomass combustion process is a complex phenomenon that includes a chain of chem- ical and physical reactions. The combustion process of a biomass fuel consists of four main stages (Figure 2): drying, pyrolysis, char gasification and char combustion/oxidation. These processes might overlap with one another, depending on the combustion conditions and the combustion appliance. The small-scale solid fuel combustion processes can be divided into continuous combustion and batch combustion with natural or forced air draught supply (as discussed in detail in Sections 2.3.1 & 2.3.2). As an example, in combustion appliances with continuous fuel feeding, all these stages will occur simultaneously within the fuel batch.

However, in batch combustion, more pronounced separation between the volatilization and char combustion stages can be observed (Boman, 2005).

Nearly complete combustion can be realized when suitable amounts of air and fuel (which are expressed in terms of the air-fuel equivalence ratio or the excess air ratio, ) are mixed for a sufficient duration under suitable conditions of turbulence and temperature (TTT). Ide- ally, only a stoichiometric mixture is necessary for the full oxidation of the fuel; however, in practice, the excess air ratio in solid-fuel fired appliances must always exceed 1 for the realization of sufficient combustion efficiency. Typically, the excess air for small-scale combustion appliances is 1.5 - 2 to ensure efficient inlet air and flue gas mixing (Van Loo

& Koppejan, 2008, Obernberger et al., 2006).

In Figure 2 we can observe a two-stage combustion, where the primary air injection is ap- plied in the fuel bed and the secondary air injection is applicable in the combustion chamber.

An operation in low excess air l1 is possible only then, when the good mixing is determined.

(16)

16

Figure 2: Main reactions in two-stage biomass combustion with primary and secondary air.

(Adapted from Nussbaumer et al., 2003).

3.3 Automatic fixed-bed combustion in small-scale biomass appliances Biomass can be combusted using a stoker, grate combustion or fluidized-bed combustion (bubbling fluidized-bed combustion, circulating fluidized-bed combustion) or pulverized fuel combustion. The latter is for large-scale plants; therefore, it will not be discussed in this thesis. The use of a stoker is suitable mostly for small-scale combustion, such as burning wood chips or pellets, whilst the grate combustion is adaptable for small and medium scales.

Stoker burners are suitable for small-scale, single-house boilers with an output capacity of up to 100 kW. The maximum output capacity for stoker burners can be 3 MW. The fuel in a stoker burner (e.g., wood chips, pellets, grains or briquettes) is fed automatically into the burner and not directly into the boiler. There are several methods for distributing the fuel to the burner, and regarding these methods, stoker burners are classified as horizontally fed (side-fed) burners, underfed (bottom) burners and overfed (top) burners (Hartmann et al., 2003; Table 1). The fuel is provided to the burner by a screw auger feeder conveyor, and air is blown from under the fuel to provide primary air and from above the fuel to provide secondary air. Then, the large flame entered the boiler for the water heating application. The ash is removed at the end of the burning process, when it falls and collects in the ash box.

Depending on the output capacity of the burner, the fuel can be placed in silos manually or using a tractor.

(17)

17

Figure 3: Underfeed stocker with a screw feeder (Power and heat plants. Study prepared for the FAO).

Initially, the grate-firing was designed for coal combustion. Grate-firing is currently one of the most frequently used technologies for the combustion of biomass. This type of solid fuel combustion is suitable for heat plants with outputs of less than 30 MW. It can be used in boilers for single-family houses (15-40 kW) and larger buildings (40-400 kW) and in district heating plants (400 kW-20 MW). Grates are combustion surfaces, and they can differ in type according to the operating method. Stationary sloping grates, travelling grates, recip- rocating grates and vibrating grates are available. Stationary grates are suitable for applica- tions in which simple biomasses with low ash content and medium humidity are used. The fuel is fed onto a grate, where the primary air is supplied through the grate bars from below.

The grate has two main functions, namely, lengthwise transportation of the fuel and distri- bution of the primary air that is fed from beneath the grate. The primary air portion of the oxygen supply for combustion helps to dry the fuel, and later, when the combustion heats the grate, it cools it. For instance, the water pipes that are installed in a grate can also help cool the grate. Therefore, within the primary air feed through the fuel bed, drying, devolat- ilization and char combustion occur. Then, the secondary/tertiary air is supplied above the grate, and it plays important roles in mixing, burnout, and emissions. Secondary air staging is frequently applied in modern biomass-fired grate boilers to reduce the NOx emissions and lower the flame temperature peaks (Yin et al., 2008, Houshfar et al., 2011; Liu et al., 2013, Carroll et al., 2015). Moreover, according to Jalalabadi et al. (2017) and Carroll et al.

(2015), air staging demonstrates significant potential in reducing particulate (PM) and gas- eous (NO and CO) emissions. In addition, fuel staging, when it is fed in two levels, is a primary method for in situ reduction of NOx in biomass combustion (Nussbaumer et al., 2003).

Table 1 presents an accurate classification of automated biomass combustion technologies, as introduced and described by Hartmann et al. (2003).

(18)

18

Table 1: Explicit classification of automated biomass combustion technologies. (Adapted from Hartmann et al., 2003).

3.4 Batch combustion appliances

Batch combustion appliances are typically small, have an output of <50 kW and are fuelled by logs, lump wood or briquettes. There are three types of these appliances: stoves, inserts and hot water boilers. Enclosed stoves (mass <800 kg) are non–storing or semi–storing

(19)

19

wood heaters, and they are made with steel and sometimes covered with ceramic materials, stone or soapstone to increase heat storage (Tissari, 2008). A stove or insert directly warms a room via radiant heating by capturing the heat from the periodic combustion (re–loading) of wood logs and emitting it at a constant temperature for a long period. Stoves are fed manually mainly with firewood (logs). Modern stoves realize efficiencies of approximately 70 - 80% and are available with outputs from 3.5 to 20 kW. A heat-storing stove is a special design that can accumulate heat and radiate it over a long period after the fire has stopped burning.

3.5 Combustion emissions

Biomass combustion emissions can be divided into two main groups: particulate matter (PM) and gaseous emissions. Primary PM particles consist of a complex mixture of soot (elemental carbon, EC), organic material (OM) or condensable organic compounds (COCs) and inorganic material in the forms of salts and metal oxides. In addition, incomplete wood combustion generates substantial amounts of volatile organic compounds (VOCs). VOCs are precursors for secondary organic aerosol (SOA) (Heringa et al., 2011; Wu et al., 2017).

In complete combustion conditions, the particles are formed mainly from salts. Neverthe- less, during the start–up phase and improper operation, the amounts of COCs and soot can be elevated substantially in automatic wood combustion plants. Typically, incomplete com- bustion conditions occur in manually operated wood appliances, which result in the for- mation of COCs and/or soot, VOCs and CO (Nussbaumer et al., 2003).

Continuously fired appliances have lower CO and CxHy emissions than conventional wood stoves (Lamberg et al., 2011; Ozgen et al., 2014) due to the fully controlled combustion process. The PM emissions and the particle size distributions from small-scale combustion appliances strongly depend on the combustion conditions. The optimized and automated (continuously controlled) wood fuel combustion process induces a decrease in the total sus- pended particle (TSP) emissions and changes the PM size distribution (Kubica et al., 2004).

Based on studies of Boman et al. (2004), Hays et al. (2003) and Ehrlich et al. (2007), modern and optimized small-scale boilers emit mostly submicron particles (< 1 m), and the mass concentration of particles that are larger than 10 m is typically < 10% for small combustion appliances. Particulate emissions in pellet boilers are very low and mainly consist of inor- ganic matter, in contrast to wood stoves, for which the particle emissions are dominated by soot and organics (Van Loo & Koppejan, 2008).

3.5.1 Particle formation during biomass combustion

Aerosols are defined as two-phase systems that consist of particles and the gases in which they are suspended (Hinds, 1999). Aerosols ranges in particle size from 0.001 to over 100 m. There are many types of aerosols with varied chemical and physical properties and formation principles. Combustion aerosols can be classified according to PM size into

(20)

20

coarse particles (> 1 m), fine particles (< 1 m) and ultrafine particles (< 0.1 m), and according to origin, formation principle and chemical composition.

Figure 4: Images of coarse fly ash (left) and aerosol (right) particles from wood combustion in a large furnace (Van Loo & Koppejan, 2008).

3.5.1.1 Coarse fly ash particle formation

The major types of ashes that are formed in biomass combustion are bottom and fly ashes.

The fly ash fraction includes a fine mode and a coarse mode. The coarse mode constitutes a small portion of the ashes that are formed during the particle entrainment from the fuel bed;

thus, it might depend strongly on the primary air flow in a grate combustion system (Pagels et al., 2003). The coarse fly ashes vary in size from a few micrometres to 250 m (Obern- berger & Thek, 2010). For various boiler types and combustion conditions, one portion of the coarse ashes precipitate in plant sections, whereas the other portion forms the coarse fly- ash emissions. Typically, these particles are formed from low-volatility metals, such as Ca, Si, and P, along with smaller amounts of K, Na and Mn, in the form of oxides, sulphates or phosphates (Kelz et al., 2010). In small-scale combustion, these particles represent only 10 wt% of total particulate emissions (Kelz et al., 2010). In addition, the coarse particle mode strongly depends on the fuel type or fuel load (Wierzbicka et al., 2005). According to Jo- hansson et al. (2003) and Carlsson (2008), the coarse mode is absent from the flue gas or is present in very small amounts. Coarse PM with an aerodynamic diameter in the range of 2,5-10 m can be deposited mainly in the upper respiratory system (Cormier et al., 2006).

3.5.1.2 Fly ash particle formation

In comparison to coarse fly ash particle formation, the mechanism of fly ash particle for- mation is much more complex. Under suitable combustion conditions, inorganic elements such as K, Na, P, S, Cl, Zn and Pb, are released into the gas phase, where they may form small fly ash particles (< 1 m). Sulphates, chlorides, carbonates and oxides form in the gas phase. Once one of these compounds becomes supersaturated due to flue gas cooling or immoderate formation of the corresponding compound, gas-to-particle conversion via

(21)

21

nucleation and condensation begins. The growth of aerosol particles predisposes the system to vapour condensation, coagulation, agglomeration, surface reactions and adsorption (Sip- pula, 2010). Furthermore, the growth of particles via condensation depends on the chemical composition and initial diameter of the particles, along with the saturation ratio. Aerosol particle growth via coagulation and agglomeration occurs due to collisions of aerosol parti- cles with each other. Collisions occur due to Brownian motion, turbulence or external forces (Hinds, 1999). All these aerosol particle growth mechanisms might occur simultaneously.

The final particle size varies in the range of 50 nm-300 nm (Pagels et al., 2003). Moreover, in incomplete combustion, carbonaceous aerosol formation occurs. Carbonaceous aerosol particles are divided into soot particles (elemental carbon, EC) and particles that are com- posed of organic matter (OM/COCs). The OM fraction of particles is formed via the con- densation of organic vapours. Soot formation depends on the flame type, temperature, air- to-fuel ratio and characteristics of the fuel. According to Flagan et al. (1988), the soot par- ticles are agglomerates of small spherical particles. Moreover, the soot and inorganic parti- cles have a structural similarity, which is due to the common origin of these particles. How- ever, in contrast to inorganic ash formation, the mechanism of soot particle formation is not yet well understood, in which the hydrocarbon chemistry in the flame is highly complex.

The challenges also include the ability of soot to burn if exposed to oxygen at high temper- atures. When the combustion flue gas cools, the soot absorbs highly toxic compounds, such as polycyclic aromatic hydrocarbons (PAHs). PAHs are mutagenic and carcinogenic to hu- mans (Rengarajan et al., 2015). In summary, the inorganic aerosol is strongly influenced by the fuel type, while the carbonaceous aerosol can be reduced by optimizing the combustion process.

Figure 5: Ash formation mechanism in biomass combustion (Obernberger et al., 2010).

(22)

22 3.5.2 Gaseous emissions

The gaseous emissions that are generated by the combustion process are CO, CO2, hydro- carbons (HC), oxides of nitrogen (NOx) and sulphur oxides (SOx). In addition, water vapour is formed due to the fuel drying phase or hydrogen oxidation. CO2 is regarded as a major product of complete combustion. In addition, the carbon dioxide from wood combustion is not classified as a greenhouse gas. Hence, the biomass combustion process is regarded as CO2-neutral if a sustainable utilization is assumed (Obernberger et al., 2006). However, bi- omass combustion is not carbon-neutral since direct and indirect emissions of greenhouse gases (GHGs) other than CO2and indirect CO2can influence the carbon balance (Meyer et al., 2012). Incomplete combustion may lead to carbon monoxide emissions, and it is re- garded as a benchmark for the combustion efficiency. CO reduction can be realized with high combustion temperatures (> 850 °C), stoichiometric (theoretical) or excedentary total

, satisfactory mixing and sufficient retention time (> 1.5 s).

NOx emissions include NO, NO2 and N2O. The amount of N2O in modern fuel appliances is very low (Van Loo & Koppejan, 2008). NOx emissions play an essential role in atmos- pheric reactions and contribute to ground-level ozone (smog), SOA and acid rains (Camre- don et al., 2007). There are three types of NOx: fuel-derived NOx, thermal NOxand prompt NOx. For the formation of thermal NOx and prompt NOx in wood combustion, high temper- atures of >1300 °C are required. Therefore, the amounts of nitrogen oxides that form in modern appliances are minor due to the lack of high temperatures and nitrogen concentra- tions in the fuel. However, NOx might increase with increasing fuel N content (Ramirez- Diaz et al., 2014).

SOx contributes substantially to the acidification of forests. In biomass combustion, fuel- derived sulphur forms SO2 (along with SO3) and alkali and earth alkali sulphates. Conse- quently, sulphur (>75%) is released into the vapour phase, and due to rapid flue gas cooling in the boiler, sulphates condense on the fly ash particles and/or on the boiler walls. The efficiency of S fixation in the ash depends on the amounts of alkali (K and Na) and earth alkali metals (especially Ca) in the fuel and on the combustion technology and flue gas cleaning techniques. In addition, in power plants, where fuels with high S content are uti- lized, lime (CaCO3) and quicklime (CaO) are spread over the burning fuel to minimize the SOx emissions. In the presence of high SO2 emissions, the Cl is released due to sulphation of alkali and earth alkali metals. During the biomass combustion the alkali metals as K (and Na) are released into the gas phase as KOH and/or KCl. If potassium hydroxide is present in the flue gas, it may react further to KCl or/and K2SO4 with the presence of HCl and SO2

in the flue gas. As a result, the formation of KCl may strongly accelerate the corrosion, while K2SO4 has a little influence on the corrosion rate. Therefore, the additional sulphur is sometimes added to combustion process in order to force formation of the less corrosive K2SO4compound (Antunes et al., .2013).

(23)

23

3.6 Flue gas cleaning devices in small-scale wood-fired appliances The combustion of biomass fuels in small-scale heating appliances is an important source of fine particle emissions (Obaidullah et al., 2012). According to an IEA Bioenergy Task32 member countries survey on particle precipitation devices for residential combustion (Obernberger & Mandl, 2011), in some of the 32 studied European countries, the input to the total PM emissions by residential biomass combustion to the residential heating sector in some European countries exceeds 80 %. The dust emission limit values differ substan- tially among these countries. Therefore, due to residential boiler technology (Directive 2009/125/EC) and emission limit tightening, the development and introduction into the mar- ket of particle precipitation devices are essential.

The operating strategy of particulate filter devices is based on the passage of a gas stream with particles through the region, where external forces act on the particles and separate them from the gas stream. To select a suitable filtering device, four main factors must be considered: 1) particulate concentration in the gas stream; 2) the size distribution of the particles; 3) the gas flow rate and 4) particle emission limits (Flagan, 1988). Last, the total cost of the construction and operation of the cleaning device must be evaluated, as this cost (depending on the size of collector) is directly proportional to the volumetric flow of the gas that must be cleaned. Moreover, operational factors that might influence the cost of the de- vice are the pressure drop through the unit, the required power and the volume of the liq- uid/scrubber water if the wet filtering technique is used.

Filtering devices can use one or several physical mechanisms for the particulate matter pre- cipitation process. These include sedimentation, in which particles from the gas stream settle under gravity on chamber floor; migration of charged particles in an electric field, in which particles are charged and the electrostatic force induces them to migrate on the surface (plates) of the electrostatic precipitator (ESP); inertial deposition, in which a gas stream changes direction due to the flow around the object and particles continue to move along their original trajectories due to their inertia (cyclones, scrubbers and filters); and Brownian diffusion, as particles that are suspended in a gas are always in the Brownian motion and, thus, undergo natural collisions with other particles, where they adhere and are collected.

This separation mechanism is highly effective for small particles. In addition, two relatively weak deposition mechanisms, namely, thermophoresis (Chang et al., 1999) and diffusiopho- resis, can affect particle collection.

The principal parameter that plays an important role in the selection of the filtering device is the particle diameter (dp). This measure defines the efficiency of the filtering device (Van Loo and Koppejan, 2008).

A wide range of flue gas cleaning technologies are available for large- and medium-scale biomass boilers. These technologies include settling chambers (or gravity settlers) (Flagan, 1988), cyclones (Lanzerstorfer et al., 2015), electrostatic precipitators (Brunner et al., 2017, Ibarra et al., 2018; Aragon et al., 2015; Dastoori et al., 2013), bag filters (Ibarra et al., 2018, Aragon et al., 2015), spray chambers (scrubbers) (Van Loo & Koppejan, 2008; Bianchini et

(24)

24

al., 2016), ceramic filters (Risnes, 2002) and catalytic converters (Kaivosaja et al., 2012;

Ozil et al., 2009; Hukkanen et al., 2012). However, only some can be implemented and effectively used in small-scale wood combustion. Table 2 presents the PM removal efficien- cies of various cleaning devices for the specified particle sizes.

In Sections 4.2 and 4.3, the condensing heat exchanger (CHX) and the shielded corona charger will be discussed.

Table 2: Removal efficiencies for particles of various sizes in particle control devices (Van Loo & Koppejan, 2008).

Particle control technology Particle size ( m) Efficiency (%)

Settling chambers >50 <50

0 8

<

5

>

s e n o l c y C

0 9

<

5

>

s e n o l c y c it l u M

Electrostatic filters >1 <99

9 9

<

1

>

s r e tl if g a B

Spray chambers >10 <80

Impingement scrubbers >3 <80

Cyclone spray chambers >3 <80

Venturi scrubbers >0.5 <99

3.6.1 Introduction to heat exchangers

A heat exchanger (HX) is a device that is used to transfer heat between two or more fluids.

The fluids can consist of a single or two phases and, depending on the heat exchanger type, may be separated or in direct contact. Heat exchangers can be categorized in terms of the flow configuration and the construction type. There are four principal fluid flow classes:

counter flow, co-current flow, crossflow and hybrids, such as cross counterflow and multi- pass flow. Classification in terms of the construction type subdivides HXs into recuperative and regenerative heat exchangers (Figure 6). A recuperative HX has a separate flow path for each fluid, and the flow paths flow simultaneously through the HX with heat exchange across the wall. In contrast, a regenerative HX has a single flow path, through which cold and hot fluids pass alternatingly.

(25)

25

Figure 6: Heat exchanger classifications (Adapted from Brogan et al., 2011).

Various recuperative HXs are available, which are classified based on the type of fluid in- teraction: indirect contact, direct contact and special. In indirect-contact heat exchangers, fluids exchange heat through tubes or plates, whilst in direct-contact heat exchangers, fluids are not separated, and they exchange heat by being in direct contact.

The most common are tubular-shape heat exchangers due to the wide ranges of temperatures and pressures that can be implemented. The most prevalent are the shell and tube heat ex- changers. Each consists of a shell with traditional plain tubes. The fluids can be liquid or gases, where one is flowing inside the tubes and the other outside, within the shell. The fluids can flow in parallel or in a cross/counter flow. Heat exchangers of this type are typi- cally made of metal, but for various applications (such as pharmaceutical acids), they can be made of graphite, plastic or glass (Schou et al., 1997).

Heat exchangers are widely used in industrial processes, in electricity generation, and in heating, ventilation and air conditioning (HVAC) systems (Siegel & Nazaroff, 2003). Based on Zukeran et al. (2013), the combined condensing heat exchanger and electrostatic precip- itator system in diesel marine engines can reduce SO2and PM emissions by 76% and 78%, respectively. Moreover, according to Yrjöla et al. (2004), a heat exchanger can be used as a biomass drying system prior to the combustion process. In addition, heat exchangers can be used as air preheaters and flue gas cleaners in municipal solid waste (MSW) and hazardous waste incineration and biomass processing (Stehlik et al., 2007). However, few studies have been conducted on the use of heat exchangers in small-scale wood combustion for particu- late matter filtration and latent heat recovery. In the study of Gröhn et al. (2009), extensive modelling calculations for the condensing heat exchanger were conducted. It was found that the particle deposition varied based on thermophoresis and diffusiophoresis from 25 to 95%

depending on the temperature gradient and the condensation flux. According to de Best et al. (2008), aerosol emissions can be reduced by 45 - 70% in comparison to conventional boilers. Messerer et al. (2004 and 2007) concluded that a particle deposition efficiency of over 95% was realized and that deposition was dominated by thermophoresis and diffusion.

(26)

26

Thermophoresis is a dominant particle deposition mechanism, whereas Brownian diffusion is relevant for particles that are smaller than 0,1 m (Zhang & Ahmadi et al., 2000).

The crucial parameters that affect the deposition efficiency are the flue gas inlet temperature, the gas and water temperature differences, the particle retention time in the heat exchanger and the heat exchanger tube diameter (Messerer et al., 2004) and length. Furthermore, dep- osition increases with the flue gas moisture, without significant thermal losses, due to the recovery of latent heat in the condensing heat exchanger system. Therefore, in combusting dry fuel, the use of scrubber water to moisturize the flue gas is an alternative that facilitates the minimization of the PM deposition and flushes out the precipitates from heat exchanger inlets and walls.

In the combustion of pure wood, the corrosion phenomenon is not considered due to the low chlorine and sulphur contents in woody biomass fuel (Sippula et al., 2009); therefore, small amounts of alkali chlorides, HCl and SO2 are released into the flue gas. Fouling and slagging reduce the heat transfer with heat exchanger surfaces (Riedl et al., 1999) and accelerate corrosion and erosion of them, thereby inducing a reduced heat transfer efficiency and in- creasing the maintenance costs of the facilities (Pääkkönen, 2015). Fouling is associated with parameters such as the fuel and various combustion parameters, such as the air flow (Febrero et al., 2015).

3.6.2 Electrostatic precipitator and modifications

An electrostatic precipitator (ESP) is a flue gas cleaning device that uses electrostatic forces and consists of discharge wires and collecting plates. A high voltage is applied to the dis- charge wires to induce an electric field between the wires and the collecting plates, which also ionizes the gas around the discharge wires to supply ions. An ion collides with a particle and sticks to it, and the particle obtains its charge. Particles can be charged via two mecha- nisms, depending on their sizes: diffusion charging and field charging. The charging of par- ticles in unipolar gaseous ion density is referred to as diffusion charging (which is based on the effect of thermal agitation for small particles (<0.2 m) and weak applied electric fields).

Field charging is the mechanism for larger particles (> 0.5 m) and stronger electric fields.

Whenever a flue gas with fine particles passes between the collecting plates and discharge wires, particles in the gas are charged with ions. The Coulomb force that is generated by the electric field causes the charged particles to adhere to the collecting plates. Furthermore, the particles on the collecting plates can be removed via three methods: 1- rapping the collecting plates, 2-scraping off with a brush, or 3-washing off with water (wet type) and removing from the hopper for disposal.

The collection efficiency of the ESP depends strongly on the electrical properties of the particulate matter (electrical resistivity) that is being collected (Nussbaumer et al., 2010), the design parameters of the ESP (long electrodes and a large collection surface) and the combustion process (Nussbaumer et al., 2016). The collection efficiency of the PM in small- scale combustion and low-emission fuel might be at least 70% during normal operation (Carrol et al., 2017), and for large-scale appliances, it might exceed 90% (REF). In addition,

(27)

27

efficiency losses might occur due to re-entrainment, sneakage and back-corona. Re-entrain- ment is caused by rapping, and 10-15% of particles move back into the gas stream. Sneakage refers to the movement of a small portion of the gas flow (5-10%) around the charging zones while remaining uncharged due to design restrictions. Back-corona occurs if the electric field becomes sufficiently large to cause an electrical breakdown, which reduces the charge on the particles.

Figure 7: Operating principle of the parallel-plate electrostatic precipitator (Jaworek et al.

2007). HV – high voltage.

Various ESP types are widely used in large-scale industrial processes. All ESPs are divided into four main categories: plate-wire, flat-plate, tubular and wet precipitator. In large-scale appliances, in which large volumes of gas must be cleaned, the most suitable is the parallel- plate-type ESP design. However, for smaller gas-flow residential furnaces, the most suitable is the tubular-type electrostatic precipitator due to easier installation in or on top of the chimney (Report FutureBioTec, 2012).

Small-scale electrostatic precipitators can be divided according to the installation principle (Bologa et al., 2012); in a residential furnace, the precipitator might be placed between the furnace and the chimney, whereas for a chimney stove, it can be located next to the stove, in the flue gas duct or on the top of the chimney. ESPs can be used in automated wood chip/pellet boilers or in logwood boilers or stoves (Brunner et al., 2018).

Laitinen et al. (2016) introduced a new ESP technology, namely, a sonic jet charger, in which particles were removed based on diffusion charging and reached removal efficiencies of 80% for sub-micron particles. Moreover, in comparison to wire-type discharge elec- trodes, sonic jet-type chargers have the advantage of being shielded from the flue gas, which renders them suitable for systems in which the use of unshielded electrodes is not possible due to the flue gas properties. In addition, this type of ESP can be used if periodic cleaning

(28)

28

of the electrode would cause a problem, as the purged air protects it from contamination and corrosion. In comparison to wire-type electrodes, the sonic jet charger is more expensive.

These two types of ESPs (chimney top with manual cleaning and in-line with automated cleaning) were studied and evaluated by Carroll et al. (2017). It was found that removal efficiencies that exceed 70% are possible with an inline system with automated cleaning.

Moreover, for emissions from wood, which is regarded as a low-emission fuel, the precipi- tation efficiency could be maintained over a long exploitation time. A research study with a 50-kW wood pellet boiler and a 100-kW gasifier that was equipped with an ESP was con- ducted by Poškas et al. (2018). It was concluded that two these technologies produce differ- ent particle concentrations, and therefore, a high collection efficiency of 98-99% was real- ized with a system with flue gas (a pellet boiler). Due to the challenging operating conditions in the tests with a gasifier, a lower removal efficiency of ~75% was obtained.

According to a study on seven heating plants with tube-type and plate-type ESPs by Nuss- baumer et al. (2016), a removal efficiency that exceeds 95% is realizable for new plants with optimal ESP operation and low PM concentrations. Moreover, the control of the fuel input and the primary and secondary air is important for the realization of high PM precip- itation efficiencies. Bäfer et al. (2012) investigated the efficiency of an ESP (installed in the chimney) in both efficient and poor combustion conditions. The removal efficiencies with respect to the mass of the particles were 87% and 93% for efficient and poor combustion, respectively. The higher efficiency in the poor combustion was attributed to the lower flue gas temperature, which induced a longer particle residence time in the ESP.

In the study of Dastoori et al. (2013), a CFD model was developed for analysing the trajec- tories of the particles in a small-scale biomass appliance with an ESP. It was concluded that the PM collection efficiency depends on the flue gas velocity, the applied voltage and the particulate size. Moreover, the removal efficiency can be increased by changing the geom- etry of the system; for example, the maximum efficiency can be realized by increasing the length of the chimney.

Several negative issues are encountered in the use of the ESP technology: the back-corona effect (solution: decrease the dust resistivity); in submicron particle removal, re-entrainment from the collector electrode back into the flowing gas; and contamination of the discharge electrode, among others (Jaworek et al., 2007).

In summary, these flue gas cleaning devices realize high PM collection efficiencies that range from 50 to 98% (Bologa et al., 2012, Brunner et al., 2018, Laitinen et al., 2016 Migliavacca et al., 2014; Meiller et al., 2015). However, regardless of these examples, small-scale ESP versions remain under development (Obernberger & Mandl et al., 2011).

(29)

29

4 Methods

The work that is described here was conducted at the Fine Particle and Aerosol Technology laboratory at the University of Eastern Finland. In Papers I and II, the novel fine particle emission precipitation technology was utilized. Paper III focused on particle physicochem- ical properties and their differences in continuous versus batch combustions. In Paper IV, the time-resolved emissions of batch combustion were analysed.

Combining all these studies, the gas and particle emissions and subsequent physicochemical characterizations in various combustion conditions were examined. The objective of Papers I and II was to investigate the efficiency of emission cleaning technology in small-scale combustion. Schematic diagrams of the measurement systems are presented in Figure 8 and Figure 9.

Figure 8: Experimental setup with (A) a condensing heat exchanger and (B) a reference boiler (Paper I).

(30)

30

Figure 9: Experimental setup with a combined condensing heat exchanger and shielded co- rona charger system (Paper II).

4.1 Combustion appliances

All studies were conducted in small-scale combustion units of two types: solid biofuel boil- ers (pellets/wood chips) (Papers I-IV) and a modern masonry heater (Papers III and IV).

These two types of small-scale appliances are used for room and space heating. The masonry heater that was used in Papers III and IV heats a room/space via natural convection and radiation, whereas the heat that is produced by solid biofuel boilers is transformed into a fluid by heat exchangers and transported to other locations (radiators and underfloor heat- ing).

Figure 10: Schematic diagrams of combustion systems: (A) a moving grate multifuel boiler (Paper I and II), (B) a modern pellet boiler (Paper III) and (C) a modern masonry heater (Paper IV).

(31)

31 Moving Grate Solid Biofuel Boiler (A)

A novel modular small-scale combustion boiler was used in Papers I and II. The combus- tion unit included logic-controlled fuel feeding and air supply systems, a moving 3-step- grate burner with 8 primary air inlets and 16 secondary air inlets (Ariterm Multijet, 40 kW), a ceramic-insulated chamber with a quadratic cross-section of 0.15 m2and height of 0.8 m, a heat exchanger, a stack with a flue gas blower and a cooling circuit (Figure 8). The burner is designed for multifarious solid fuels (pellets and chips). The fuel feeding system, the burner and the heat exchanger were commercial products, and the remaining unit modules were self-made (Leskinen et al., 2015). The combustion reactor can simulate various com- bustion conditions with changes in the fuel and air flow rates and primary and secondary air settings. In addition, all the air flows were monitored with sensors (Schmidt). The flue gas from the combustion chamber was led to a horizontal passage, to a thermally insulated pipe and, finally, to the heat exchanger. Two heat exchangers were utilized in the study, namely, a conventional firetube boiler-type heat exchanger (Arimax 340 Bio) and a condensing tube heat exchanger (see Section 4.2 for a detailed description).

Pellet Boiler Biotech (B)

A pellet boiler with a 25-kW maximum load (Biotech Energietechnik GmbH, model PZ25RL) was used in Paper III as an automated fixed-grate boiler with an integrated burner and boiler. The boiler uses the top feeding principle, in which the pellets from the top are fed on the grate. The fuel feed is automatically controlled and ensures that there is no back burn from the furnace to the storage reservoir. Moreover, the boiler can be operated contin- uously for loads of between 6 and 25 kW. The boiler is equipped with logic-controlled fans for the combustion air supply. The boiler had a cylindrical furnace, and the primary air was supplied through holes in the grate at the bottom of the furnace. The width of the grate was approximately 110 mm with a hole diameter of 7 mm. The secondary air was supplied through 12 holes, each with a diameter of 15 mm, that were situated above the grate.

Masonry Heater (C)

Papers III and IV investigate batch combustion in a masonry heater of the modern type (Tulikivi Hiisi 4, 1.6 kW) with combustion air staging and a heat-retaining structure. Mod- ern masonry heaters (MMHs) differ from conventional heaters due to improvements in the combustion chamber construction and the separation of the combustion air into primary and secondary air flows. The air staging in MMHs, in comparison to conventional masonry heat- ers, improves the air mixing with the pyrolysis and gasification gases and results in more complete combustion in the flaming phase and, therefore, lower CO, OGC and particulate organic emissions (Tissari et al., 2008, Nuutinen et al., 2014). The considered MMH is com- posed of a 1300-kg soapstone, where heat is stored and released into the surroundings over a period of several hours (up to 48 h) after combustion has been stopped (Tissari, 2008).

The heater had a vertical firebox with a ceramic lining and a double-window glass door.

Additionally, the heater had a rack, which was placed in the firebox, where the wood logs

(32)

32

could be placed according to the instructions of the stove manufacturer. The flue gas in the fireplace flows according to a contraflow principle, i.e., from the upper combustion chamber downwards and out into the side channels and, subsequently, into the stack. The stack was located under a hood that was connected to a flue gas fan to induce an unforced draft, which was regulated by two dampers. Furthermore, the combustion air flow was allocated between the primary (through the grate), secondary (above the wood batch) and window flushing air portions with average proportions of 20%, 45% and 35%, respectively.

4.2 Condensing heat exchanger. Design and operating principle

A prototype of the condensing heat exchanger (CHX) was developed at the University of Eastern Finland (UEF), which was discussed in Papers I and II. The CHX has a water volume of 118 dm3, a fixed length of 745 mm and a baffle spacing of 150 mm. The CHX contains 121 tubes, each with an inner diameter of 8 mm (Figure 11). The flue gas was divided into all these tubes, and the gas flow was designed to be laminar. The CHX operates according to the counterflow principle; i.e., the flue gas flows downstream through the pipes from the combustion chamber, and water, which serves as a coolant, and a heat transfer agent (70% water and 30% glycol) flows upstream in the CHX. In addition, the CHX was optimized for high thermophoretic and diffusiophoretic deposition. The operating funda- mentals and the design of the CHX are extensively discussed in Gröhn et al. (2009), in which it was utilized as a secondary heat exchanger with a small-scale pellet unit. In Papers I and II, the CHX was installed directly after the combustion chamber to obtain the same range of flue gas inlet temperatures as are used in tube heat exchangers in small-scale boilers.

Furthermore, to avoid deposit formation in the CHX inlets and along the CHX, a scrubber unit was installed above the heat exchanger, and it was used during the cleaning periods (Figure 9). Droplets were generated via water injection at high pressure through a specially designed nozzle. After atomization, droplets vaporized to steam, condensed in the CHX and enhanced the water film flow on the heat exchanger walls. The condensed and scrubbing water flowed down the CHX walls and was collected in a container. From the container, all the water was pumped into a larger reservoir, which enabled the quantification of the water flow.

To determine the particle deposition and thermal efficiencies in the CHX, for comparison, reference measurements were conducted with a commercial firetube boiler (Arimax 340 Bio), which had a 40-kW heat capacity and a water volume of 210 dm3. In the reference setup, the CHX was replaced by a thermally insulated tube, thereby resulting a shift of the flue gas to the reference heat exchanger (Figure 8).

(33)

33 Figure 11: Schematic diagram of a heat exchanger.

4.3 Shielded corona charger. Design and operating principle

The shielded corona charger (SCC) that was used in Paper II was provided by Tassu ESP Ltd. (NASU electric diffusion charger). The SCC had an electrically insulated outer shell that was composed of a 200-mm long tube with a diameter of 26 mm (Figure 12). The outer shell was composed of a ceramic material (pure aluminium oxide) for proper insulation. The chamber contained 12 parallel-distributed sharp-needle corona discharge electrodes that produced ions. The chamber had connectors for both filtered compressed air (with a shield flow of 40 lpm) and a high-voltage supply. The optimized voltage was -20 kV, and the generated corona current was (0.2)-(0.6) mA. The corona discharge formed between the needles and the combustion chamber, which acted as a ground electrode. Surrounded by the flue gas, ions and particles collided, and charges were transmitted to the particles. Most of the charged particles precipitated on the CHX walls. In addition, several particles precipi- tated in the horizontal passage. Particles also deposited on the surface of the charger due to both the presence of naturally charged particles of opposite polarity and the temperature difference between the flue gas and the charger.

(34)

34

Figure 12: Schematic diagram and a photograph of the shielded corona charger.

4.4 Particle measurements, characterizations and chemical analyses Aerosol flow from the small-scale combustion units was diluted to the suitable concentra- tions and temperatures for the measurement instruments. The dilution was implemented us- ing a porous tube diluter (PRD) and either one or two ejector diluters (EDs) in Papers I-IV.

In addition, particle-free pressurized dried air was used as a dilution gas at ambient temper- ature. Moreover, based on the CO2 concentrations in the raw flue gas (which were measured via FTIR) and in the diluted sample, the dilution ratio (DR) after PRD and ED was adjusted to the desired value.

The particle emission measurements included the fine particulate matter emissions (PM1), which were determined using filter collections; direct mass measurements of the total sus- pended particles (TSP), which were obtained using a tapered element oscillating microbal- ance (TEOM ThermoFisher Scientific 1405) (only in Papers II and IV); and the number size distributions, which were measured via several techniques.

The PM1 filter sampling systems consisted of two (Papers II and IV) or three (Paper I) parallel filter collection lines. Particles were collected on quartz fibre filters (Pallflex, Tis- suequartz) for the organic and elemental (OC/EC) analyses and on Teflon membrane filters ((polytetrafluoroethylene (PTFE)) (Gelman Scientific, Teflo) for the estimation of the or- ganic gas phase vapour amount. The sampling collection time was 60-120 min.

Before and after collection, the filters were kept in a room with stable conditions (RH 40%

and 20 °C) for at least 24 h prior to weighting. Gravimetric analyses were conducted with a

Viittaukset

LIITTYVÄT TIEDOSTOT

Thirdly, the impact of sieve mesh size on the particle size distribution of wood chips from different age groups was compared by using an 80 mm × 80 mm sieve for a

The mean microfibril angle, the crystallinity of wood and the size of cellulose crystallites were determined as a function of the number of the year ring in Norway spruce (Papers

Shielded corona charger (SCC)-induced reduction efficiencies of particulate matter emissions measured upstream of the heat exchanger in different operating conditions (Paper II)... 32

Comparison of the whole combustion average particle mass emission factors (filter collection), ELPI particle number emission factors and ELPI particle GMD between different studies

hengitettävät hiukkaset ovat halkaisijaltaan alle 10 µm:n kokoisia (PM10), mutta vielä näitäkin haitallisemmiksi on todettu alle 2,5 µm:n pienhiukka- set (PM2.5).. 2.1 HIUKKASKOKO

Figure 9 summarizes the particle composition results from Papers II–IV and from some recent studies. Other studies also report quite good correlation between particle

Particle size distributions were measured by the means of an engine exhaust particle sizer (EEPS), but soot emissions, gaseous emissions and the basic engine performance were

Reduction efficiencies of fine particle (PM1) mass of different setups, at shielded corona charger (SCC) operating temperatures of 400–500 °C, in comparison to the reference