DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2019
Emissions from a fast-pyrolysis bio-oil fired boiler: Comparison of
health-related characteristics of
emissions from bio-oil, fossil oil and wood
Sippula, O
Elsevier BV
Tieteelliset aikakauslehtiartikkelit
© Elsevier Ltd.
CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/
http://dx.doi.org/10.1016/j.envpol.2019.02.086
https://erepo.uef.fi/handle/123456789/7561
Downloaded from University of Eastern Finland's eRepository
Accepted Manuscript
Emissions from a fast-pyrolysis bio-oil fired boiler: Comparison of health-related characteristics of emissions from bio-oil, fossil oil and wood
Olli Sippula, Kati Huttunen, Jouni Hokkinen, Sara Kärki, Heikki Suhonen, Tuula Kajolinna, Miika Kortelainen, Tommi Karhunen, Pasi Jalava, Oskari Uski, Pasi Yli- Pirilä, Maija-Riitta Hirvonen, Jorma Jokiniemi
PII: S0269-7491(18)35037-1
DOI: https://doi.org/10.1016/j.envpol.2019.02.086 Reference: ENPO 12255
To appear in: Environmental Pollution Received Date: 13 November 2018 Revised Date: 18 February 2019 Accepted Date: 24 February 2019
Please cite this article as: Sippula, O., Huttunen, K., Hokkinen, J., Kärki, S., Suhonen, H., Kajolinna, T., Kortelainen, M., Karhunen, T., Jalava, P., Uski, O., Yli-Pirilä, P., Hirvonen, M.-R., Jokiniemi, J., Emissions from a fast-pyrolysis bio-oil fired boiler: Comparison of health-related characteristics of emissions from bio-oil, fossil oil and wood, Environmental Pollution (2019), doi: https://doi.org/10.1016/
j.envpol.2019.02.086.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M AN US CR IP T
AC CE PT ED
M AN US CR IP T
AC CE PT ED
Emissions from a fast-pyrolysis bio-oil fired boiler:
1
Comparison of health-related characteristics of emissions
2
from bio-oil, fossil oil and wood
3
Olli Sippula1,2*, Kati Huttunen3, Jouni Hokkinen4, Sara Kärki5, Heikki Suhonen1, Tuula Kajolinna4, 4
Miika Kortelainen1, Tommi Karhunen1, Pasi Jalava3, Oskari Uski3, Pasi Yli-Pirilä1, Maija-Riitta 5
Hirvonen3, Jorma Jokiniemi1 6
7
1Fine Particle and Aerosol Technology Laboratory, Department of Environmental and Biological 8
Sciences, University of Eastern Finland, P.O. Box 1627, FI-70211 Kuopio, Finland 9
2Department of Chemistry, University of Eastern Finland, Yliopistokatu 7, P. O. Box 111, FI-80101 10
Joensuu, Finland 11
3Inhalation Toxicology Laboratory, Department of Environmental and Biological Sciences, 12
University of Eastern Finland, P.O. Box 1627, 70211 Kuopio, Finland 13
4VTT Technical Research Centre of Finland, P.O. Box 1000, FI-02044 VTT, Espoo, Finland 14
5Fortum Power and Heat, Keilaniementie 1, 02150 Espoo, Finland 15
*Corresponding author: Tel.: +358403553397; E-mail address: olli.sippula@uef.fi 16
17
KEYWORDS 18
fine particles, aerosols, NOx, ash chemistry, heavy metals, PAH, aerosol toxicology, electrostatic 19
precipitator, Fast pyrolysis bio-oil, renewable energy, boiler, particle emissions 20
ABSTRACT 21
M AN US CR IP T
AC CE PT ED
There is currently great interest in replacing fossil-oil with renewable fuels in energy production.
22
Fast pyrolysis bio-oil (FPBO) made of lignocellulosic biomass is one such alternative to replace 23
fossil oil, such as heavy fuel oil (HFO), in energy boilers. However, it is not known how this fuel 24
change will alter the quantity and quality of emissions affecting human health. In this work, 25
particulate emissions from a real-scale commercially operated FPBO boiler plant are characterized, 26
including extensive physico-chemical and toxicological analyses. These are then compared to 27
emission characteristics of heavy fuel-oil and wood fired boilers. Finally, the effects of the fuel 28
choice on the emissions, their potential health effects and the requirements for flue gas cleaning in 29
small- to medium-sized boiler units are discussed.
30
The total suspended particulate matter and fine particulate matter (PM1) concentrations in FPBO 31
boiler flue gases before filtration were higher than in HFO boilers and lower or on a level similar to 32
wood-fired grate boilers. FPBO particles consisted mainly of ash species and contained less 33
polycyclic aromatic hydrocarbons (PAH) and heavy metals than had previously been measured 34
from HFO combustion. This feature was clearly reflected in the toxicological properties of FPBO 35
particle emissions, which showed less acute toxicity effects on the cell line than HFO combustion 36
particles. The electrostatic precipitator used in the boiler plant efficiently removed flue gas particles 37
of all sizes. Only minor differences in the toxicological properties of particles upstream and 38
downstream of the electrostatic precipitator were observed, when the same particulate mass from 39
both situations was given to the cells.
40
CAPSULE 41
Fast-pyrolysis bio-oil combustion generated high particulate mass concentrations but the particles 42
contained less PAH, heavy metals and induced less acute toxicity effects on a cell line, when 43
compared to particle emissions from heavy-fuel oil combustion.
44
1 INTRODUCTION 45
M AN US CR IP T
AC CE PT ED
Combustion processes are the main anthropogenic sources of fine particles, causing adverse 46
health effects (WHO, 2009; Pope et al., 2006) and climate effects (Gustafsson et al., 2009).
47
Exposure to fine particles has been associated especially with respiratory and cardiovascular 48
diseases (Hoek et al., 2013). The quantities, properties and the potential harmful effects of 49
emissions from combustion processes are known to be largely dependent on the fuel characteristics, 50
combustion conditions and flue gas after-treatment technologies (Lighty et al., 2000;Liu et al., 51
2017;Kaivosoja et al., 2014;Uski et al., 2014). At the moment, there is a globally increasing interest 52
in replacing fossil-oil fired energy boiler units with ones operated with renewable fuels. One of the 53
new renewable fuels developed to replace heavy fuel oil in industrial-scale energy production 54
applications is the fast pyrolysis bio-oil (FPBO) made from lignocellulosic biomass (Bridgwater et 55
al., 1999;Meier et al., 2013). The general benefits of FPBO are that it is considered renewable and 56
contains only small amounts of sulphur originating from the biomass feedstock (Bridgwater et al., 57
1999;Oasmaa et al., 2015). Therefore, replacement of heavy fuel oil with FPBO in energy 58
applications strongly reduces fossil CO2 and SOx emissions, but it is not known how this 59
replacement affects the overall characteristics of emissions and their potential public health effects.
60
In Finland, FPBO is currently used to replace heavy fuel oil in heat production. In new boiler 61
projects, FPBO can also be seen as an alternative for solid biofuels, which are mainly wood chips, 62
forest and saw mill residues and wood pellets. When new boiler investments are planned, it is 63
important to know the requirements for flue gas cleaning to comply with emission regulations, and 64
how the emission characteristics and quantities are affected by the fuel change. Regulatory 65
compliance with respect to emissions is an important question, particularly in small boilers for 66
which the relative cost of the flue gas cleaning technique is often substantial (Ohlström et al., 2006).
67
In addition, many of the small fossil fuel fired boilers are not equipped with particle filtration 68
systems, which needs to be considered when converting these boilers to operate on bio-oil. National 69
emissions limits for small- to medium-scale FPBO boilers exist in Finland and The Netherlands, 70
M AN US CR IP T
AC CE PT ED
which set limits on particulate matter, NOx and SO2 (Meier et al., 2013). In addition, the MCP 71
directive of the European Union (Directive (EU) 2015/2193) contains European level emission 72
limits for so-called medium combustion plants (1-50 MWth), which tightens the emission limits in 73
many countries in the coming years. Furthermore, for authorities and industry, it is important to 74
acknowledge the quality of the emissions to estimate the effects of emissions on air quality and 75
possible health effects. However, the recent detailed studies on the toxicological effects of various 76
combustion emissions clearly indicate that the traditional metrics used in flue gas emission 77
measurements provide only rough information on the potential of the emissions to induce adverse 78
health effects (Kaivosoja et al., 2013;Kocbach Bølling et al., 2009;Kasurinen et al., 2015;Oeder et 79
al., 2015). The particles from combustion processes often have a very complex chemical 80
composition, which makes it difficult to draw conclusions regarding the health hazards simply 81
based on emission concentrations or their chemical composition. Recent toxicological studies on 82
combustion emissions indicate that the key properties influencing toxicological responses of 83
combustion PM in the respiratory system include particle size, chemical composition and chemical 84
speciation (Kaivosoja et al., 2013;Kocbach Bølling et al., 2009;Uski et al., 2017). Heavy fuel-oil 85
fired boiler emissions are known to contain substantial amounts of heavy metals and polycyclic 86
aromatic hydrocarbons (PAH) (Sippula et al., 2009;Hays et al., 2009;Happonen et al., 2013), which 87
have been associated with strong cytotoxic and genotoxic effects in lung cells (Kaivosoja et al., 88
2013;Kasurinen et al., 2015). On the other hand, wood-fired boilers usually generate higher mass 89
concentrations of fine particles than fossil oil boilers (Sippula et al., 2009) and contain specific 90
transition metals, such as zinc, which has been associated with toxicological effects in lung cells 91
(Uski et al., 2014; Uski et al., 2015). Such detailed emission characteristics with respect to potential 92
health hazards have not been studied earlier for FPBO combustion.
93
To date, only a few demonstrations of utilization of FPBO in heat and power production have 94
been executed on a real commercial scale (Lehto et al., 2014). Only very limited published data on 95
M AN US CR IP T
AC CE PT ED
the emission factors and characteristics of flue gas emissions from FPBO combustion are available, 96
and most of the data originates from research laboratory settings, and short large-scale 97
demonstrations, and the available data represent mainly burner technologies based on air-assisted 98
atomization. Most of the emission data originates from studies carried out in laboratories with 99
miniaturized air-assisted atomization burners (Tzanetakis et al., 2011a;Tzanetakis et al., 100
2011b;Khodier et al., 2009;Zadmajid et al., 2017) or with other laboratory apparatuses such as drop- 101
tube furnaces (Feng et al., 2016;Feng et al., 2017). The current literature presents a large variation 102
in particle mass emissions, ranging between 50 and 550 mg/m3 (Oasmaa et al., 2015;Tzanetakis et 103
al., 2011b;Feng et al., 2017), depending mainly on the fuel quality and burner technology but likely 104
also due to different sampling and measurement protocols. Many of the studies conclude that ash- 105
forming matter, present in the FPBO, forms the majority of the particulate emissions, while 106
unburned fuel (soot, organics and char residues) may also form a substantial fraction of the PM 107
emissions (Tzanetakis et al., 2011a;Tzanetakis et al., 2011b). In addition, the presence of so-called 108
“solids”, i.e., small char particles in the fuel, can lead to enhanced particle emissions, and filtration 109
of solids from the fuel has been suggested to decrease PM emissions (Feng et al., 2016). NOx
110
emissions of FPBO have been demonstrated to originate mainly from fuel-bound nitrogen (Lehto et 111
al., 2014;Khodier et al., 2009). According to Baxter et al. (Baxter et al., 1995), the fuel nitrogen to 112
NOx conversion efficiency increases with increasing flue gas oxygen concentration and decreases 113
with increasing fuel nitrogen. The SOx emissions from FPBO combustion have been found to be 114
negligible (Lehto et al., 2014;Khodier et al., 2009). Overall, the results clearly indicate PM 115
emissions as the most challenging area in the emission abatement from FPBO combustion.
116
However, very limited information exists on particulate formation processes and the characteristics 117
of the particles such as particle size distributions and chemical composition. In addition, no data 118
have been published on the toxicological properties of emissions from FPBO combustion.
119
M AN US CR IP T
AC CE PT ED
In this work, a thorough analysis of particulate emissions from a real commercial-scale FPBO 120
boiler was carried out, including extensive emissions measurements and an analysis of particulate 121
physicochemical and toxicological characteristics. The measurements were carried out both 122
upstream and downstream of the electrostatic precipitator (ESP) to evaluate its functioning and 123
influence on emissions characteristics. Finally, the results were compared to emission data from 124
heavy fuel oil- and wood-fired boilers to enable discussing the effects of fuel choice on emissions, 125
their potential health effects and requirements for flue gas cleaning in small- to medium-sized boiler 126
units.
127
128
2 MATERIALS AND METHODS 129
2.1 Fast pyrolysis oil 130
The oil utilized in the study is the Fortum Otso® fast pyrolysis bio-oil (FPBO) produced from 131
lignocellulosic biomass in Fortum’s FBPO production plant in Joensuu, Finland (Description of the 132
process can be found in the supplementary material). FPBO is completely different from fossil oils, 133
as it contains a significant amount of oxygen-containing organic components and water. FPBO is 134
acidic, not fully distillable and sensitive to high temperatures (Oasmaa et al., 2015;Lehto et al., 135
2014). The FPBO utilized during the emission measurements was analysed for higher heating value 136
(ASTM D5291), ash content (EN ISO 6245), water content (ASTM E203) and elemental 137
composition, including C, H, N (EN 15104) and S (SFS-EN ISO 11885:09). The higher heating 138
value was on average 16.6 MJ/kg wet fuel. The water content was on average 29.1%, which is 139
typical for FPBO (Oasmaa et al., 2015). The ash, nitrogen and sulphur content was 0.15%, 0.3%
140
and 62 mg/kg wet fuel, respectively. All the measured parameters fulfil the limits set for FPBO in 141
the ASTM Burner Fuel Standard D7544 (Oasmaa et al., 2015) and in the new EN16900 standard 142
(SFS-EN 16900:2017). More detailed fuel analysis results are presented in the supplementary 143
material Table S1.
144
M AN US CR IP T
AC CE PT ED
2.2 Fast pyrolysis bio-oil boiler 145
The measurements were carried out at the Fortum Vermo district heating plant located in Espoo, 146
Finland, in which a 49 MW water tube boiler unit was modified to use FPBO. Technical details of 147
the FPBO combustion system are presented in the supplementary material Section 2. The plant is 148
equipped with an Ahlstom (ELPAC 2.1.) 2-field electrostatic precipitator to remove particles from 149
the flue gases. During the measurements, the boiler was run at approximately 83-95% of nominal 150
capacity, i.e., with 41-47 MW fuel load and 39.2-44.1 MW district heat production capacity.
151
Operational parameters were not varied during the test runs, and in each measurement, the capacity 152
was kept stable.
153
2.3 Measurements of particle and gas emissions 154
The gas emission measurements were carried out downstream from the convection pass and before 155
the ESP by using a Fourier transform infrared (FTIR, Gasmet Technologies Inc.) analyser. The 156
measured gas compounds were H2O, CO2, CO, CH4, NO, NO2, SO2, N2O, C2H6, NH3 and HCl.
157
Particulate mass concentration measurements were done both between the convection pass 158
and the ESP as well as downstream of the ESP. Altogether four different methods and sampling 159
setups were used to obtain a comprehensive analyses of emissions and to increase reliability of the 160
emission factors (Details in supplementary material Section 3 & Figure S2): (1) standard-based total 161
dust emissions according to the EN13284-1, (2) total particle mass emission, together with the 162
particle mass size distribution, using a Berner Low Pressure Impactor (BLPI) method (Sippula et 163
al., 2009a;Sippula et al., 2009b), (3) a gravimetric impactor (DGI, Dekati) for high-volume 164
particulate matter collection of PM1 and PM1-10 size fractions for toxicological, inorganic and PAH 165
analyses, according to the method described by Ruusunen et al. (2011), and (4) collection of PM1
166
filter samples for the analyses of gravimetric mass, and organic (OC) and elemental (EC) carbon 167
fractions in PM1, similar to the methodology described by Turpin et al. (2000). The particle 168
concentrations and size distributions were measured online using an electrical low-pressure 169
M AN US CR IP T
AC CE PT ED
impactor (ELPI, Dekati, 10 Lpm) which provided the number concentration for 12 aerodynamic 170
size bins ranging from 0.007 to 6.83 µm. Sintered impactor stages were used in the ELPI. The flue 171
gas was diluted for the ELPI using a porous tube diluter at a dilution ratio of 24 and 17 upstream 172
and downstream of the ESP, respectively.
173
All emission concentrations were normalized to 3% oxygen (18% CO2) at 0 °C temperature 174
and dry flue gas. All particle size distribution data were processed in a manner similar to earlier 175
studies (Sippula et al., 2009a).
176
177
2.4 Sample treatment and chemical analyses 178
The filters and impactor substrates were weighed before and after sampling to determine the 179
collected sample mass. The particulate matter from the DGI substrates was extracted by dispersing 180
the particles into methanol in an ultrasonic bath (Ruusunen et al., 2011). The samples were pooled 181
together to form two different particle size fractions: “PM1” for particles smaller than 1 µm in 182
aerodynamic diameter and “PM1-10” for particles larger than 1 µm in aerodynamic diameter. The 183
extracts were split into a number of glass tubes dedicated to different analyses. The methanol from 184
the tubes was evaporated using a stream of nitrogen gas.
185
Both the BLPI and DGI impactor samples were analysed with inductively coupled plasma 186
mass spectrometry (ICP-MS) for a variety of elements and with ion chromatography for water- 187
soluble anions. For elemental analysis, the samples were digested in HF-HNO3 solution and 188
analysed with ICP-MS, including the elements As, Ba, Ca, Cd, Cu, Fe, K, Mg, Na, Rb, Sb, Sr, Zn, 189
Ag, B, Be, Bi, Co, Cr, Li, Mn, Mo, Ni, Pb, Se, Th, Ti, Tl, U and V. For anion analysis, the samples 190
were extracted in deionized water and analysed with ion chromatography for Cl-, SO42-
, PO43-
, Br-, 191
F- and NO3-
. 192
X-ray diffraction analysis was carried out for the “PM1” sample. The sample was first 193
dispersed into methanol. The dispersion was then added to a zero-background X-ray diffraction 194
M AN US CR IP T
AC CE PT ED
sample holder a few drops at a time and dried in an oven at 40 °C. The dried sample was then 195
analysed with a Bruker D8 Discover XRD system using Cu Kα radiation (40 mA, 40 kV). The 196
diffractograms were measured for 2θ angles from 10 to 90° at a resolution of 0.05°. The crystalline 197
compounds were identified by comparing the diffractograms to those in the ICDD PDF-2 database.
198
Analyses of polycyclic aromatic hydrocarbon (PAH) compounds were conducted from PM1
199
and PM1-10 size fractions using a gas chromatograph mass spectrometer (6890N GC, equipped with 200
a 5973 inert Mass Selective Detector, Agilent Technologies) and an HP-17-MS column for the 201
separation of the compounds. The analysis was operated in the selected ion monitoring (SIM) mode, 202
and a total of 30 PAH compounds were analysed from the samples (Supplementary material Table 203
S3). The sample extraction and analysis procedure is described in detail in Lamberg et al. (Lamberg 204
et al., 2011).
205
The analyses of OC and EC fractions from the quartz fibre filter samples were performed with 206
a thermal-optical method using a Sunset carbon analyser (Sunset Laboratories, Inc.). The analysis 207
was based on the NIOSH 5040 procedure (NIOSH, 1999).
208
2.5 Toxicological analyses 209
Cell culture. A mouse macrophage cell line RAW264.7 obtained from American Type Culture 210
Collection (ATCC, Rockville, MD, USA) was cultured at 37 °C and 5% CO2 atmosphere in RPMI 211
1640 supplemented with 10% heat-inactivated foetal bovine serum (FBS), L-glutamine (2 mM) and 212
antibiotic (penicillin-streptomycin, 100 U/ml) (Gibco BRL, Paisley, UK). The cells were cultured in 213
flasks and refreshed every 2-3 days when confluent. Prior to the exposure experiment, the cells 214
were transferred to 12-well plates (Costar, Corning, NY, USA), 4 x 105 cells/well in 1 ml of culture 215
medium. After 24 hours, the detached cells were removed by refreshing the cell culture medium.
216
Experiment. Before the experiments, the particles were suspended in DMSO (20 µl/mg) and sterile 217
water (end particle concertation in suspension was 5 mg/ml). Samples were sonicated for 30 218
minutes before preparing a dilution series (15, 50, 150 and 300 µ g/ml) in the complete cell culture 219
M AN US CR IP T
AC CE PT ED
medium. Mouse macrophages were exposed to four doses selected based on earlier experiments 220
(Jalava et al., 2010) and appropriate controls (blank filter, negative and positive controls). After the 221
experiment, the metabolic activity of the cells was assessed with the MTT-test (Uski et al., 2014) 222
and acute cell death with propidium iodide (PI) staining (Uski et al., 2015). In addition, the well- 223
being of the cells was followed by measuring the amount of reduced free thiols (Kaddour et al., 224
2013). Indicators of disrupted cell function, such as cell cycle arrest and programmed cell death, 225
were analysed with a flow cytometric method (BD FACSCanto™ II (BD Biosciences, San Jose, CA 226
USA) from cells fixed with 70% ethanol (Uski et al., 2015). The genotoxicity of the samples was 227
assessed with the alkaline single cell gel electrophoresis (comet) assay and oxidative stress by 228
measuring the amount of reactive oxygen species (ROS) with the DCF method (Uski et al., 2015).
229
Inflammatory response was assessed by measuring the concentration of inflammatory mediators for 230
macrophage inflammatory protein MIP2 and tumour necrosis factor alpha (TNFα) with an ELISA- 231
based method according to manufacturers’ instructions (R&D Systems, Minneapolis, MN, USA).
232
The experiment was repeated four times. The results were analysed statistically with the Kruskal- 233
Wallis test (p<0.05) using IBM SPSS Statistics 20.0 (SPSS Inc., Chicago, IL, USA).
234
235
3 RESULTS AND DISCUSSION 236
3.1 Gaseous emissions and combustion conditions 237
The FPBO boiler was operated on average at 3.3-3.4% dry flue gas excess O2. The O2
238
concentrations varied only very slightly, between 3.2 and 3.6%. The flue gas moisture content was 239
17%. The CO emissions were very low (Table 1), with the average concentrations of measurements 240
ranging between 3 and 10 mg/m3 at 3% O2. The gaseous organic compounds and HCl measured 241
with the FTIR were also very low, most of the time below the detection limit of the analyser. SO2
242
emissions were on average 20 mg/m3, clearly below any limit values but higher than those usually 243
observed for wood-fired combustion plants (Sippula et al., 2009a), likely because the fuel sulphur 244
M AN US CR IP T
AC CE PT ED
was not reacting as efficiently with alkali and alkali earth metals as in wood combustion. NOx 245
emissions ranged between 300 and 353 mg/m3, corresponding to approximately 14% NOx
246
conversion efficiency of the fuel nitrogen. The NOx emissions were on a level similar to the level 247
measured earlier for 10 MW FPBO boilers (Lehto et al., 2014).
248
3.2 Particle mass emissions and size distributions 249
Table 1 summarizes the measured particle emission values with different methods both upstream 250
and downstream from the ESP. The total dust emissions (Total Suspended Particles, TSP) upstream 251
from the ESP were, on average, 250 and 160 mg/m3 with the EN-13284-1 standard-based filter 252
collection and with the BLPI method, respectively. The difference between the methods is likely 253
caused by losses of very large coarse particles in the BLPI sampling system. The measured TSP 254
emissions are mainly lower than those reported by Oasmaa et al. (2015) for 10 MW pyrolysis oil 255
boilers but higher than the emissions presented by Tzanetakis et al. (2011b) . However, overall, the 256
literature presents a very high range of total dust emissions for pyrolysis oil combustion, ranging 257
between 50 and 550 mg/m3 (Oasmaa et al., 2015;Tzanetakis et al., 2011b;Zadmajid et al., 258
2017;Feng et al., 2017). The fine particle mass emissions (PM1) measured with the BLPI and DGI 259
methods were 47 mg/m3 and 38 mg/m3, respectively. Thus, only 15-29% of the particulate matter 260
upstream of the ESP consists of fine particles. The TSP and PM1 emission concentrations 261
downstream from the ESP were 2.5-3.1 mg/m3 and 1.4-1.5 mg/m3, respectively, depending on the 262
measurement method. The ESP filtration efficiency for TSP and PM1 was 98-99% and 96-97%, 263
respectively. The slightly lower removal efficiency for PM1 is because at the particle sizes 0.2-1 264
µm, there is typically a so-called penetration window in ESPs, a particle size that is not efficiently 265
affected by diffusion charging or field charging (Sippula et al., 2009a). However, for adverse health 266
effects, this size range is particularly interesting, since those particles are deposited mainly in the 267
lower respiratory tract (Heyder, 2004).
268
M AN US CR IP T
AC CE PT ED
Figure 1 shows the particle mass size distributions in the flue gases before and after the ESP.
269
In unfiltered flue gas, two clearly distinct particle mass modes were observed. All the measurements 270
indicated a fine particle mass mode of approximately 60 nm, which is well in agreement with the 271
particle number size distribution measured with the ELPI and indicates that the fine particles are 272
mainly in the ultrafine (<100 nm) size range. However, the coarse particle mode, above 1 µm 273
particle diameter, clearly dominates the total particle mass concentration (Figure 1, a). The high 274
fraction of particles above 5 µm in aerodynamic diameter also indicates a high potential for 275
decreasing of the PM emissions simply with a cyclone or multicyclone (Ohlström et al., 2006).
276
A comparison of measurements before and after the ESP can be used to estimate the filtration 277
efficiency of the ESP as a function of particle size. The comparison of mass size distributions 278
indicates that downstream from the ESP, particle mass is more evenly distributed as a function of 279
particle size than before the ESP (Figure 1). The calculated particle size-dependent ESP filtration 280
efficiency curves (Supplementary material Figure S3) indicate the lowest filtration performance at 281
particle sizes between 0.1 and 3 µm, where the mass reduction efficiency drops below 98%. The 282
particle number-based measurements with ELPI show a clear increase in number-based geometric 283
mean diameter along the flue gas path, from 47 to 89 nm, and a strong decrease in particle number, 284
declining from 1.7x1014/m3 to only 3.1 x108/m3. The ELPI measurement-based ESP filtration 285
efficiency curve (Figure S3) has a size dependence similar to the mass-based curve but gives a 286
higher overall filtration efficiency. The high reduction efficiency of total particle number emission 287
can be partly explained by agglomeration of the particles along the flue gas channel but is also due 288
to the size-dependent particle collection efficiency in the ESP, where the particle penetration 289
window is above the GMD of the particle number size distribution. However, the ELPI 290
measurements include uncertainties because the presence of charged particles downstream from the 291
ESP may affect the results. In addition, the ELPI measurements could not be made at the same time 292
before and after the ESP.
293
M AN US CR IP T
AC CE PT ED
3.3 Particle chemical composition 294
According to the chemical analyses, both fine (PM1) and coarse particles (PM above particle 295
diameter of 1 µm) consisted predominantly of ash species (Figures 2 and S4). The organic carbon 296
(OC) content in the fine particles was, on average, only 0.8% of the weighed particulate mass, while 297
elemental carbon (EC) content in the fine particles was below the detection limit. Moreover, the ash 298
content of the total suspended particles (TSP), as measured with a heating method (DIN 51719), 299
was approximately 90% (Supplementary material, Table S4), indicating a low combustible carbon 300
content also for the coarse particles. If all the fuel ash-forming matter ended up as particulate matter 301
in the flue gas, the TSP mass concentration would be approximately 307 mg/m3 at 3% O2, while the 302
standard-based PM measurement showed a TSP concentration of 250 mg/m3 at 3% O2. Thus, the 303
fuel ash content could potentially be a good proxy for total suspended particle formation for a 304
FPBO boiler operating under efficient combustion conditions.
305
The major chemical species in PM1 were K, Ca, SO42-
and Cl-. In addition, Fe, Mg, 306
Mn, Na and Al were analysed in clearly noticeable amounts (Figure 2). XRD analysis of PM1 307
indicated the presence of crystalline KCl, K2SO4 and mixed sulphates such as K3Na(SO4)2 and 308
CaK2(SO4)2 (Supplementary material, Figure S5). A clear signal of crystalline MgO and weak 309
signal of CaSO4 crystals were observed, while other identified elements may also have been present 310
in an amorphous form or in concentrations too low to be detected by XRD.
311
Coarse particles (dp> 1 µm) were dominated by Ca, Mg, K, SO42- and chemically unidentified 312
matter (Supplementary material Figure S4). The unidentified fraction most likely contains various 313
silicate compounds that could not be analysed with the analysis methods used, as well as carbon 314
residues, which were estimated to be approximately 10% of the total particulate matter. PAH 315
concentrations in the PM1 and PM1-10 fractions were low, with a total analysed PAH concentration 316
of only 147 ng/m3 and 222 ng/m3 in PM1 and PM1-10 particle size fractions, respectively 317
(Supplementary material Table S6). The major PAH compounds were fluoranthene, pyrene and 318
M AN US CR IP T
AC CE PT ED
phenanthrene, which are also typical compounds for emissions from wood-fired boilers 319
(Supplementary material Figure S6).
320
3.5 Toxicological properties of emissions 321
Emissions from FPBO decreased the viability of the exposed cell cultures, which could be seen as a 322
decrease in the amount of both undamaged and metabolically active cells and an increase in the 323
number of damaged cells after exposure to the highest tested doses. The number of apoptotic or 324
mitotic cells was not affected, indicating that the exposure did not induce the apoptotic cell death or 325
the cell cycle arrest typically associated with DNA damage. However, the exposure did increase the 326
fragmentation of DNA at the highest dose level. The most prominent effect of the exposure was the 327
dose-dependent production of reactive oxygen species (ROS) and inflammatory markers, 328
significantly higher compared to control samples already at the lowest tested dose level. ROS and 329
inflammatory markers are both considered mediators of acute and subacute local irritation of 330
respiratory tissue, although sustained oxidative stress and inflammation can also lead to long-term 331
consequences associated with particulate exposure (Reuter et al., 2010).
332
Particle filtration in the ESP did not significantly alter the toxicological properties of the 333
emissions, as both samples collected before and after the ESP increased the production of ROS and 334
inflammatory mediators. A statistically significant difference was seen for increased production of 335
MIP-2 and lower metabolic activity of the cells in the samples collected downstream from the ESP, 336
which can be a result of size-dependent particle collection efficiency in the ESP, leading to a slight 337
increase in particulate mean diameter and consequent increased relative proportion of calcium in the 338
fine particle mass (Figure S4). However, considering the significant decrease in the particulate 339
matter concentration in the ESP, these differences probably have a limited role in the adverse health 340
effects induced by the emissions (Supplementary Material, Table S5).
341
3.6 Comparison of emission properties between boilers operating on pyrolysis oil, wood and 342
heavy fuel oil 343
M AN US CR IP T
AC CE PT ED
This chapter compares the emission characteristics from the FPBO boiler to our previous studies of 344
heavy fuel oil and wood-fired small boilers (Kaivosoja et al., 2013;Sippula et al., 2009a;Sippula et 345
al., 2009b) and with the literature on these emission sources. In addition, toxicological properties of 346
FPBO combustion-originated particles are compared with analyses from HFO and wood boiler 347
samples.
348
NOx emissions from the FPBO boiler (between 300 and 353 mg/m3) were clearly lower than 349
the NOx emissions from small heavy fuel oil fired boilers (Sippula et al., 2009b) and similar to or 350
lower than the NOx emissions from boilers operating on forest and saw mill residues (Sippula et al., 351
2009a). The European Medium Combustion Plant (MCP) Directive, which comes into force in 2018 352
for new plants and 2025 or 2030 for existing plants (Directive (EU) 2015/2193), limits the NOx
353
emissions at 650 mg/m3 for existing combustion plants and at 300 mg/m3 for new plants, indicating 354
that in new plants, there will be a slight need for further decreasing the NOx emissions, which could 355
be achieved either via further burner development or decreasing of the fuel nitrogen content.
356
Total particle mass concentrations in the FPBO boiler upstream of the ESP were lower or in 357
a range similar to the typical emissions in wood-fired grate boilers (Sippula et al., 2009a;Brunner, 358
2006) but 2-3 times higher than for heavy fuel oil-fired boilers (Sippula et al., 2009b). Ash content 359
in the FPBO is clearly lower than in wood residues. However, no bottom ash is formed, and thus a 360
higher fraction of the fuel ash contributes to suspending particulate matter. As a result, FPBO 361
combustion exhibits a relatively high fraction of coarse ashes in the total suspended particles. The 362
upcoming European MCP directive sets particulate emission limits for liquid fuels (other than 363
diesel) at 30 mg/m3 and 20 mg/m3 for existing and new 5-50 MW boilers, respectively. The ESP- 364
equipped boiler of this study clearly fulfils these requirements with FPBO. However, many small 365
oil-fired boilers (< 10 MW) are currently not equipped with particle filtration systems, and the 366
upcoming limits would indicate that these boiler units will require investments in particle filtration 367
systems when fired with both FPBO and HFO.
368
M AN US CR IP T
AC CE PT ED
PM1 concentrations from FPBO combustion were mainly lower than from wood-fired boilers 369
(Sippula et al., 2009a;Kaivosoja et al., 2013;Pagels et al., 2003). When compared to HFO boilers, 370
the PM1 concentrations were 1.7-3.4 times higher than in the study of Sippula et al. (Sippula et al., 371
2009b) and Happonen et al. (Happonen et al., 2013) but only 70% of the PM1 measured by 372
Kaivosoja et al. (Kaivosoja et al., 2013). In addition, PM chemical composition differs from wood 373
and HFO boilers. While HFO boiler PM emissions are dominated by sulphates, various transition 374
metals, elemental carbon and a variety of organic species (Sippula et al., 2009b;Happonen et al., 375
2013;Hays et al., 2009), the FPBO PM contained mainly ash species similar to wood boiler fly 376
ashes, with primary components of Ca, Mg, SO42- and K. However, the composition of the PM1 size 377
fraction in FPBO combustion also differs from that in wood combustion (Figure 2). In wood-fired 378
boilers, PM1 fine fly ash is dominated mainly by alkali metal sulphates, chlorides and carbonates 379
and to a smaller extent by zinc (Sippula et al., 2009a;Leskinen et al., 2014), while in the FPBO 380
boiler studied, PM1 contained several additional refractory elements, particularly Ca, Mg, Mn, Fe 381
and Al. This is likely explained by the higher combustion temperature in FPBO combustion when 382
compared to wood combustion. According to the calorific value, moisture and elemental 383
composition of the analysed fuel, as well as flue gas excess oxygen, the adiabatic flame temperature 384
was estimated to be approximately 1800 °C which is enough to volatilize these refractory elements 385
(see supplementary material section 5). In addition, the fuel particle pyrolysis and oxidation, and 386
consequent ash formation/transformation, does not occur in a fixed fuel bed as in grate boilers, 387
which likely enhances vaporization of ash-forming species from the fuel particles. A similar 388
presence of refractory species in PM1 has also been observed for pulverized wood combustion 389
(Sippula et al., 2008).
390
Particulate PAH concentrations in the FPBO flue gas were clearly lower than the particulate 391
PAH concentrations measured earlier for boilers operated with wood or heavy fuel oil. In particular, 392
the PAH emissions from the HFO boiler studied were substantially higher, with a total PM1 PAH 393
M AN US CR IP T
AC CE PT ED
emission factor of 0.2 mg/MJ (Kaivosoja et al., 2013), which is approximately 4600 times higher 394
compared to the emissions from the FPBO boiler. However, it should be noted that the earlier 395
measurements were made at smaller plants with 5-15 MWth size scales.
396
Fine particulate heavy metal emissions from FPBO combustion were clearly lower than 397
what had been measured from HFO-fired boilers and on a similar level with grate-fired wood 398
boilers (Kaivosoja et al., 2013;Sippula et al., 2009b). For trace metals, there are no emission 399
limitations for FPBO, but considering the metals limited by the European Waste Incineration 400
Directive (Directive 2000/76/EC) (including Cd, Tl, Sb, As, Pb, Cr, Co, Cu, Mn, Ni and V), only 401
the Mn emission was higher for FPBO than for HFO. Thus, replacing HFO with FPBO would 402
considerably decrease fine particulate heavy metal emissions, even in cases where FPBO and HFO 403
boilers would have the same degree of flue gas particulate filtration.
404
Particle size distributions showed a relatively high mass fraction of coarse particles for 405
FPBO flue gas. In contrast, the fine particle mode mean particle size was smaller than for wood 406
boilers (Sippula et al., 2009a) and similar to HFO boilers (Sippula et al., 2009b) (Supplementary 407
material Figure S7 and Table S7). The particle number concentrations upstream from the ESP were 408
roughly three times higher than the particle number concentrations measured for wood and HFO 409
boilers, possibly because the sampling point was immediately after the boiler convection pass, most 410
likely leading to less coagulation/agglomeration of particles than in the reference HFO and wood 411
boilers, in which the measurements were made from the stack or after the flue gas cyclone. The 412
FPBO particle number emissions decreased to a very low level with the use of the ESP.
413
Toxicological properties of the emissions from FPBO, wood and heavy fuel oils boilers were 414
tested in the same experimental setting, allowing for ranking of the overall toxicity of the emissions.
415
The PM1 samples from the HFO and wood boiler used in the toxicological analyses were the same 416
as those used in the studies of Kaivosoja et al. (2013) and Kasurinen et al. (2015). All studied 417
particulate emissions decreased the metabolic activity of the cells, with the effect of HFO clearly 418
M AN US CR IP T
AC CE PT ED
the strongest. Similar effects could also be seen in decreased numbers of viable cells and increased 419
proportion of damaged cells after exposure to FPBO and HFO. Neither of the exposures caused cell 420
cycle arrest, but HFO increased the number of apoptotic cells and fragmentation of DNA already at 421
the 50 µg/ml dose level, most likely due to the high PAH and transition metal content of the HFO 422
emissions. The production of ROS was increased after exposure to emissions from both pyrolysis 423
and heavy fuel oil, possibly reflecting the higher amount of metals in the emissions. The most 424
prominent effect of exposure to emissions from the FPBO boiler was the increased production of 425
inflammatory mediators, which was statistically significantly higher compared to both HFO and 426
wood boiler emissions.
427
Overall, the cytotoxicity of the emissions from FPBO were on the same level or lower than 428
the emissions from the wood boiler, and both FPBO and wood boiler emissions were clearly less 429
cytotoxic than emissions from the HFO. Importantly, the emissions from HFO were most potent in 430
increasing apoptosis and DNA damage in the exposed cells, suggesting that the emissions from 431
wood and FPBO boilers are less genotoxic. The increased production of ROS and inflammatory 432
mediators after exposure to particulate matter from FPBO indicate that the emissions from FPBO 433
are particularly capable of initiating inflammatory reaction in exposed tissues. The small emissions 434
after the ESP had larger inflammatory potential when compared to the same mass dose collected 435
before ESP. This phenomenon is also seen previously with ESP filtration in a wood-fired boiler 436
(Kasurinen et al., 2015), possibly of some importance since the particulate size range penetrating 437
through the ESP is also able to enter the deep lung (Heyder, 2004). Inflammation plays a significant 438
role in the development of both adverse respiratory and cardiac effects (Kelly and Fussell, 2015).
439
HFO emissions as well as wood emissions seemed to suppress the inflammatory reaction, a 440
phenomenon also seen previously with exposure to samples containing PAH compounds both in 441
vivo and in vitro (Happo et al., 2008;Jalava et al., 2009), which may lead to impaired 442
immunological function (Saravia et al., 2014). Furthermore, comparing the inflammatory potential 443
M AN US CR IP T
AC CE PT ED
of FPBO to that of HFO is likely affected by the acute toxicity of HFO, preventing the cells from 444
producing these mediators (Figure 3, Table 2).
445
446
4 CONCLUSIONS 447
Total PM emissions from the FPBO boiler without particle filtration were lower or on a similar 448
level to the total PM emissions that have been measured from wood-fired grate boilers but 2-3 times 449
higher than from heavy fuel oil-fired boilers. In contrast, NOx emissions were clearly lower than 450
from HFO-fired boilers and on a similar level to wood-fired boilers. Additionally, the PM1
451
concentrations in the flue gas were mainly higher in FPBO combustion than in HFO combustion, 452
and thus particle filtration systems (such as the ESP in this study) are needed to reach similar PM1
453
emission levels as in fossil oil-fired boilers. When comparing the measured emission levels of 454
FPBO combustion to the forthcoming emission limits in the European Medium Combustion Plant 455
(MCP) Directive, it can be concluded that there is especially a need for the controlling of particulate 456
emissions in FPBO-fired boilers. In the MCP-directive, the particulate emission limits for boiler 457
sizes 5-50 MW are lower than the PM1 measured upstream from the ESP in this study, which 458
indicates that particle filtration systems more efficient than cyclones are needed, at least for boiler 459
sizes starting from 5 MW. Since this study measured emissions from only one boiler plant operating 460
with one FPBO quality, more studies are needed in the future to form a complete picture of the 461
FPBO boiler emissions.
462
The 2-field electrostatic precipitator utilized had a high particle filtration efficiency, leading to 463
a relatively low PM emissions from the studied FPBO-fired boiler, clearly fulfilling the 464
forthcoming emission limits in the MCP-directive. ESP only slightly affected the particle size and 465
the overall toxicological properties of the emitted particles.
466
Both PM1 and total PM were formed almost entirely by ash species with a very low fraction 467
of unburned carbonaceous components, such as elemental carbon, organic carbon and PAHs, 468
M AN US CR IP T
AC CE PT ED
suggesting that the fuel ash content was a good proxy for PM emission formation in FPBO boilers 469
operated under efficient combustion conditions. The PM1 composition in FPBO combustion was 470
similar to the PM1 composition in wood-fired grate boilers, with the exception of higher 471
concentrations of refractory ash components (especially Ca, Mg, Fe and Mn). This finding is 472
supported by the theoretical equilibrium calculations and the higher adiabatic flame temperature 473
with FPBO in comparison to the grate firing of wood, leading to enhanced vaporization of 474
refractory ash species during combustion.
475
Overall, the cytotoxicity of fine particles from FPBO was lower than from a HFO boiler. The 476
HFO sample induced high genotoxic responses already at a low dose, most likely due to its high 477
PAH content, which was not observed with FPBO samples. Both HFO and FPBO particles 478
increased production of reactive oxygen radicals, reflecting the metal content in the particles. The 479
overall toxicity of FPBO emissions was on a level similar to the samples collected from a wood 480
boiler except for higher inflammatory and oxidative stress responses, suggesting a potential for 481
inducing acute respiratory irritation effects. Considering the consistently high toxicity of HFO 482
emissions shown in the majority of the measured endpoints and specifically the high potential for 483
genotoxic effects, FPBO emissions can be considered less harmful compared to emissions from the 484
HFO boiler. However, when assessing the adverse health effects of emissions, both the quantity and 485
quality of human exposure needs to be defined.
486
Currently in Finland, FPBO is used primarily for replacing the use of HFO in heat- and power- 487
producing boiler plants. This study shows that although the FPBO combustion emissions are 488
characterized by higher total suspended particulate matter and fine particulate matter (PM1) 489
concentrations than HFO combustion, FPBO combustion emits less of several health-hazardous 490
emission components such as PAHs and heavy metals. This feature was clearly reflected in the 491
toxicological properties of FPBO particle emissions, showing generally smaller toxicity effects on 492
the mouse macrophage cell line than HFO combustion particles. However, due to the relatively high 493
M AN US CR IP T
AC CE PT ED
PM formation in FPBO combustion, the conversion of small fossil oil boiler plants to operate on 494
FPBO in Finland will in most cases require additional particle filtration systems to fulfil the 495
national emission regulations and the upcoming EU regulations concerning small- to medium-sized 496
combustion boilers.
497
498
ACKNOWLEDGMENTS 499
Funding by Fortum Power and Heat, the European Regional Development Fund (Pyreus project, 500
Grant A70994), the Finnish Funding Agency for Innovation and the Academy of Finland is 501
gratefully acknowledged.
502
503
REFERENCES 504
505
Directive (EU) 2015/2193 of the European Parlament and of the council of 25 November 2015 on 506
the limitation of emissions of certain pollutants into the air from medium combustion plants. 2015.
507
Directive 2000/76/EC of the European parliament and of the council of 4 December 2000 on the 508
incineration of waste. 2000. Official Journal of the European Communities. L 332/91.
509
NIOSH Elemental Carbon (Diesel Particulate) Method 5040.
510
1999. The National Institute for Occupational Safety and Health, Atlanta, GA.
511
Baxter, L, Jenkins, B, Winter, F, 1995. Baseline NOx emissions during combustion of wood- 512
derived pyrolysis oils. Sandia report, SAND94-8780, UC-1409 . 513
Bridgwater, A.V., Meier, D., Radlein, D., 1999. An overview of fast pyrolysis of biomass. Organic 514
Geochemistry 30, 1479-1493 doi: 10.1016/S0146-6380(99)00120-5.
515
Brunner, T.T., 2006. Aerosols and coarse fly ashes in fixed-bed biomass combustion: formation, 516
characterization and emissions. Series Thermal Biomass Utilization Volume 7.
517
Feng, C., Gao, X., Wu, H., 2017. Emission of particulate matter during the combustion of bio-oil 518
and its fractions under air and oxyfuel conditions. Proceedings of the Combustion Institute 36, 519
4061-4068 doi: 10.1016/j.proci.2016.08.053.
520
Feng, C., Gao, X., Wu, H., 2016. Particulate matter emission from bio-oil incomplete combustion 521
under conditions relevant to stationary applications. Fuel 171, 143-150 doi:
522
10.1016/j.fuel.2015.12.043.
M AN US CR IP T
AC CE PT ED
Global Health Risks, Mortality and burden of disease attributable to selected major risks, World 524
Health Organization WHO, 2009.
525
Gustafsson, Ö., Kruså, M., Zencak, Z., Sheesley, R.J., Granat, L., Engström, E., Praveen, P.S., Rao, 526
P.S.P., Leck, C. & Rodhe, H., 2009. Brown clouds over South Asia: Biomass or fossil fuel 527
combustion? Science 323, 495-498 doi: 10.1126/science.1164857 528
Happo, M.S., Hirvonen, M., Hälinen, A.I., Jalava, P.I., Pennanen, A.S., Sillanpää, M., Hillamo, R., 529
Salonen, R.O., 2008. Chemical Compositions Responsible for Inflammation and Tissue Damage in 530
the Mouse Lung by Coarse and Fine Particulate Samples from Contrasting Air Pollution in Europe.
531
Inhalation Toxicology 20, 1215-1231 doi: 10.1080/08958370802147282.
532
Happonen, M., Mylläri, F., Karjalainen, P., Frey, A., Saarikoski, S., Carbone, S., Hillamo, R., 533
Pirjola, L., Häyrinen, A., Kytömäki, J., Niemi, J.V., Keskinen, J., Rönkkö, T., 2013. Size 534
distribution, chemical composition, and hygroscopicity of fine particles emitted from an oil-fired 535
heating plant. Environmental science & technology 47, 14468-14475 doi: 10.1021/es4028056.
536
Hays, M.D., Beck, L., Barfield, P., Willis, R.D., Landis, M.S., Stevens, R.K., Preston, W., Dong, 537
Y., 2009. Physical and Chemical Characterization of Residual Oil-Fired Power Plant Emissions.
538
Energy Fuels 23, 2544-2551 doi: 10.1021/ef8011118.
539
Heyder, J., 2004. Deposition of Inhaled Particles in the Human Respiratory Tract and Consequences 540
for Regional Targeting in Respiratory Drug Delivery. Proceedings of the American Thoracic 541
Society 1, 315-320 doi: 10.1513/pats.200409-046TA.
542
Hoek, G., Krishnan, R.M., Beelen, R., Peters, A., Ostro, B., Brunekreef, B., Kaufman, J.D., 2013.
543
Long-term air pollution exposure and cardio-respiratory mortality: A review. Environmental 544
Health: A Global Access Science Source 12 doi: 10.1186/1476-069X-12-43.
545
Jalava, P.I., Tapanainen, M., Kuuspalo, K., Markkanen, A., Hakulinen, P., Happo, M.S., Pennanen, 546
A.S., Ihalainen, M., Yli-Pirilä, P., Makkonen, U., Teinilä, K., Mäki-Paakkanen, J., Salonen, R.O., 547
Jokiniemi, J., Hirvonen, M., 2010. Toxicological effects of emission particles from fossil- and 548
biodiesel-fueled diesel engine with and without DOC/POC catalytic converter. Inhalation 549
Toxicology 22, 48-58 doi: 10.3109/08958378.2010.519009.
550
Jalava, P.I., Hirvonen, M., Sillanpää, M., Pennanen, A.S., Happo, M.S., Hillamo, R., Cassee, F.R., 551
Gerlofs-Nijland, M., Borm, P.J.A., Schins, R.P.F., Janssen, N.A.H., Salonen, R.O., 2009.
552
Associations of urban air particulate composition with inflammatory and cytotoxic responses in 553
RAW 246.7 cell line. Inhalation Toxicology 21, 994-1006 doi: 10.1080/08958370802695710.
554
Kaddour, H., Hamdi, Y., Vaudry, D., Basille, M., Desrues, L., Leprince, J., Castel, H., Vaudry, H., 555
Tonon, M., Amri, M., Masmoudi‐Kouki, O., 2013. The octadecaneuropeptide ODN prevents 6‐ 556
hydroxydopamine‐induced apoptosis of cerebellar granule neurons through a PKC‐MAPK‐ 557
dependent pathway. Journal of Neurochemistry 125, 620-633 doi: 10.1111/jnc.12140.
558
Kaivosoja, T., Jalava, P.I., Lamberg, H., Virén, A., Tapanainen, M., Torvela, T., Tapper, U., 559
Sippula, O., Tissari, J., Hillamo, R., Hirvonen, M.-., Jokiniemi, J., 2013. Comparison of emissions 560
and toxicological properties of fine particles from wood and oil boilers in small (20–25 kW) and 561
medium (5–10 MW) scale. Atmospheric Environment 77, 193-201 doi:
562
10.1016/j.atmosenv.2013.05.014.
563
M AN US CR IP T
AC CE PT ED
Kasurinen, S., Jalava, P.I., Tapanainen, M., Uski, O., Happo, M.S., Mäki-Paakkanen, J., Lamberg, 564
H., Koponen, H., Nuutinen, I., Kortelainen, M., Jokiniemi, J., Hirvonen, M., 2015. Toxicological 565
effects of particulate emissions – A comparison of oil and wood fuels in small- and medium-scale 566
heating systems. Atmospheric Environment 103, 321-330 doi: 10.1016/j.atmosenv.2014.12.055.
567
Kelly, F., Fussell, J., 2015. Air pollution and public health: emerging hazards and improved 568
understanding of risk. Environ Geochem Health 37, 631-649 doi: 10.1007/s10653-015-9720-1.
569
Khodier, A., Kilgallon, P., Legrave, N., Simms, N., Oakey, J., Bridgwater, T., 2009. Pilot-scale 570
combustion of fast-pyrolysis bio-oil: Ash deposition and gaseous emissions. Environmental 571
Progress & Sustainable Energy 28, 397-403 doi: 10.1002/ep.10379.
572
Kocbach Bølling, A., Pagels, J., Yttri, K.E., Barregard, L., Sallsten, G., Schwarze, P.E., Boman, C., 573
2009. Health effects of residential wood smoke particles: the importance of combustion conditions 574
and physicochemical particle properties. Particle and fibre toxicology 6, 29 doi: 10.1186/1743- 575
8977-6-29.
576
Lamberg, H., Nuutinen, K., Tissari, J., Ruusunen, J., Yli-Pirilä, P., Sippula, O., Tapanainen, M., 577
Jalava, P., Makkonen, U., Teinilä, K., Saarnio, K., Hillamo, R., Hirvonen, M., Jokiniemi, J., 2011.
578
Physicochemical characterization of fine particles from small-scale wood combustion. Atmospheric 579
Environment 45, 7635-7643 doi: 10.1016/j.atmosenv.2011.02.072.
580
Lehto, J., Oasmaa, A., Solantausta, Y., Kytö, M., Chiaramonti, D., 2014. Review of fuel oil quality 581
and combustion of fast pyrolysis bio-oils from lignocellulosic biomass. Applied Energy 116, 178- 582
190 doi: 10.1016/j.apenergy.2013.11.040.
583
Leskinen, J., Tissari, J., Uski, O., Virén, A., Torvela, T., Kaivosoja, T., Lamberg, H., Nuutinen, I., 584
Kettunen, T., Joutsensaari, J., Jalava, P.I., Sippula, O., Hirvonen, M.-., Jokiniemi, J., 2014. Fine 585
particle emissions in three different combustion conditions of a wood chip-fired appliance – 586
Particulate physico-chemical properties and induced cell death. Atmospheric Environment 86, 129- 587
139 doi: 10.1016/j.atmosenv.2013.12.012.
588
Lighty, J.S., Veranth, J.M. & Sarofim, A.F., 2000. Combustion aerosols: Factors governing their 589
size and composition and implications to human health. Journal of the Air and Waste Management 590
Association 50, 1565-1618 doi: 10.1080/10473289.2000.10464197.
591
Liu, L., Kong, S., Zhang, Y., Wang, Y., Xu, L., Yan, Q., Lingaswamy, A.P., Shi, Z., Lv, S., Niu, 592
H., Shao, L., Hu, M., Zhang, D., Chen, J., Zhang, X., Li, W., 2017. Morphology, composition, and 593
mixing state of primary particles from combustion sources - Crop residue, wood, and solid waste.
594
Scientific Reports 7 doi: 10.1038/s41598-017-05357-2.
595
Meier, D., van de Beld, B., Bridgwater, A.V., Elliott, D.C., Oasmaa, A., Preto, F., 2013. State-of- 596
the-art of fast pyrolysis in IEA bioenergy member countries. Renewable and Sustainable Energy 597
Reviews 20, 619-641 doi: 10.1016/j.rser.2012.11.061.
598
Oasmaa, A., van de Beld, B., Saari, P., Elliott, D.C., Solantausta, Y., 2015. Norms, Standards, and 599
Legislation for Fast Pyrolysis Bio-oils from Lignocellulosic Biomass. Energy & Fuels 29, 2471- 600
2484 doi: 10.1021/acs.energyfuels.5b00026.
601
M AN US CR IP T
AC CE PT ED
Oeder, S., Kanashova, T., Sippula, O., Sapcariu, S.C., Streibel, T., Arteaga-Salas, J.M., Passig, J., 602
Dilger, M., Paur, H.R., Schlager, C., Mülhopt, S., Diabate, S., Weiss, C., Stengel, B., Rabe, R., 603
Harndorf, H., Torvela, T., Jokiniemi, J.K., Hirvonen, M.R., Schmidt-Weber, C., Traidl-Hoffmann, 604
C., BeruBe, K.A., Wlodarczyk, J., Prytherch, Z., Michalke, B., Krebs, T., Prevot, A.S.H., Kelbg, 605
M., Tiggesbäumker, J., Karg, E., Jakobi, G., Scholtes, S., Schnelle-Kreis, J., Lintelmann, J., 606
Matuschek, G., Sklorz, M., Klingbeil, S., Orasche, J., Richthammer, P., Müller, L., Elsasser, M., 607
Reda, A., Gröger, T., Weggler, B., Czech, H., Rüger, C.P., Abbaszade, G., Radischat, C., Hiller, K., 608
Buters, J.T.M., Dittmar, G., Zimmermann, R., Schwemer, T., 2015. Particulate matter from both 609
heavy fuel oil and diesel fuel shipping emissions show strong biological effects on human lung cells 610
at realistic and comparable in vitro exposure conditions. PLOS One doi:
611
10.1371/journal.pone.0126536.
612
Ohlström, M., Jokiniemi, J., Hokkinen, J., Makkonen P., Tissari, J., 2006. Combating Particulate 613
Emissions in Energy Generation and Industry, Views and Conclusions from the FINE Particles - 614
Technology, Environment and Health Technology Programme.
615
Pagels, J., Strand, M., Rissler, J., Szpila, A., Gudmundsson, A., Bohgard, M., Lillieblad, L., Sanati, 616
M., Swietlicki, E., 2003. Characteristics of aerosol particles formed during grate combustion of 617
moist forest residue. Journal of Aerosol Science 34, 1043-1059 doi: 10.1016/S0021- 618
8502(03)00077-6.
619
Pope, C.A., Dockery, D.W., 2006. Health effects of fine particulate air pollution: lines that connect.
620
Journal of the Air and Waste Management Association 56, 709–742 doi:
621
10.1080/10473289.2006.10464485 622
Reuter, S., Gupta, S.C., Chaturvedi, M.M., Aggarwal, B.B., 2010. Oxidative stress, inflammation, 623
and cancer: How are they linked? Free Radical Biology and Medicine 49, 1603-1616 doi:
624
10.1016/j.freeradbiomed.2010.09.006.
625
Ruusunen, J., Tapanainen, M., Sippula, O., Jalava, P., Lamberg, H., Nuutinen, K., Tissari, J., 626
Ihalainen, M., Kuuspalo, K., Mäki-Paakkanen, J., Hakulinen, P., Pennanen, A., Teinilä, K., 627
Makkonen, U., Salonen, R., Hillamo, R., Hirvonen, M., Jokiniemi, J., 2011. A novel particle 628
sampling system for physico-chemical and toxicological characterization of emissions. Anal 629
Bioanal Chem 401, 3183-3195 doi: 10.1007/s00216-011-5424-2.
630
Saravia, J., You, D., Thevenot, P., Lee, G.I., Shrestha, B., Lomnicki, S., Cormier, S.A., 2014. Early- 631
life exposure to combustion-derived particulate matter causes pulmonary immunosuppression.
632
Mucosal immunology 7, 694-704 doi: 10.1038/mi.2013.88.
633
Sippula, O., Hokkinen, J., Puustinen, H., Yli-Pirilä, P., Jokiniemi, J., 2009a. Particle emissions from 634
small wood-fired district heating units. Energy and Fuels 23, 2974-2982 doi: 10.1021/ef900098v.
635
Sippula, O., Hokkinen, J., Puustinen, H., Yli-Pirilä, P., Jokiniemi, J., 2009b. Comparison of particle 636
emissions from small heavy fuel oil and wood-fired boilers. Atmospheric Environment 43, 4855- 637
4864 doi: 10.1016/j.atmosenv.2009.07.022.
638
Sippula, O., Lind, T., Jokiniemi, J., 2008. Effects of chlorine and sulphur on particle formation in 639
wood combustion performed in a laboratory scale reactor. Fuel 87, 2425-2436 doi:
640
10.1016/j.fuel.2008.02.004.
641