Emission factors for PM2.5, CO, CO2, NOx, SO2 and

STOTEN-21849; No of Pages 10 Science of the Total Environment xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Emission factors for PM2.5, CO, CO2, NOx, SO2 and particle size distributions from the combustion of wood species using a new controlled combustion chamber 3CE Francisco Cereceda-Balic a,b,⁎, Mario Toledo a,c,d, Victor Vidal a,b, Fabian Guerrero a, Luis A. Diaz-Robles c,d, Ximena Petit-Breuilh e, Magin Lapuerta f a

Environmental Chemistry Laboratory (LQA), Center for Environmental Technologies (CETAM), Universidad Técnica Federico Santa María, Av. España 1680, Valparaiso, Chile Department of Chemistry, Universidad Técnica Federico Santa María, Av. España 1680, Valparaiso, Chile c Department of Mechanical Engineering, Universidad Técnica Federico Santa María, Av. España 1680, Valparaiso, Chile d Department of Chemical Engineering, Universidad de Santiago de Chile, Chile e Núcleo de Energías Renovables, Universidad Católica de Temuco, Chile f E.T.S. Ingenieros Industriales, Universidad de Castilla-La Mancha, Ciudad Real, 13001, Spain b

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Reproducibility and repeatability were studied for a new combustion chamber. • Optimization of combustion parameters was made on the new 3CE. • Pollutant emission factors for burning dry softwood and hardwood were determined. • Emissions of PM2.5 were determined for pre-ignition, flame and smoldering stages.

a r t i c l e

i n f o

Article history: Received 26 September 2016 Received in revised form 14 December 2016 Accepted 19 January 2017 Available online xxxx Editor: D. Barcelo

a b s t r a c t The objective of this research was to determine emission factors (EF) for particulate matter (PM2.5), combustion gases and particle size distribution generated by the combustion of Eucalyptus globulus (EG), Nothofagus obliqua (NO), both hardwoods, and Pinus radiata (PR), softwood, using a controlled combustion chamber (3CE). Additionally, the contribution of the different emissions stages associated with the combustion of these wood samples was also determined. Combustion experiments were performed using shaving size dried wood (0% humidity). The emission samples were collected with a tedlar bag and sampling cartridges containing quartz fiber filters. High reproducibility was achieved between experiment repetitions (CV b 10%, n = 3). The EF for PM2.5 was 1.06 g kg−1 for EG, 1.33 g kg−1 for NO, and 0.84 g kg−1 for PR. Using a laser aerosol spectrometer

Abbreviations: EF, emission factors; NO, Nothofagus obliqua; PR, Pinus radiata; EG, Eucalyptus globulus; SEP, stage emission process; OPC, optical particulate counter; Ec, combustion efficiency; TEC, total emissions contribution; 3CE, controlled combustion chamber for emissions; CR, concentration ratios; MM, molecular markers; AOA, atmospheric organic aerosol; EI, emissions inventories. ⁎ Corresponding author at: Environmental Chemistry Laboratory (LQA), Center for Environmental Technologies (CETAM), Universidad Técnica Federico Santa María, Av. España 1680, Valparaiso, Chile. E-mail address: [email protected] (F. Cereceda-Balic).

http://dx.doi.org/10.1016/j.scitotenv.2017.01.136 0048-9697/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Cereceda-Balic, F., et al., Emission factors for PM2.5, CO, CO2, NOx, SO2 and particle size distributions from the combustion of wood species using a new cont..., Sci Total Environ (2017), http://dx.doi.org/10.1016/j.scitotenv.2017.01.136

2

F. Cereceda-Balic et al. / Science of the Total Environment xxx (2017) xxx–xxx

Keywords: Particulate matter Emissions factors Residential wood combustion Combustion chamber

(0.25–34 μm), the contribution of particle emissions (PM2.5) in each stage of emission process (SEP) was sampled in real time. Particle size of 0.265 μm were predominant during all stages, and the percentages emitted were PR (33%), EG (29%), and NO (21%). The distributions of EF for PM2.5 in pre-ignition, flame and smoldering stage varied from predominance of the flame stage for PR (77%) to predominance of the smoldering stage for NO (60%). These results prove that flame phase is not the only stage contributing to emissions and on the contrary, pre-ignition and in especial post-combustion smoldering have also very significant contributions. This demonstrates that particle concentrations measured only in stationary state during flame stage may cause underestimation of emissions. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Wood has been commonly used for heating and food preparation because it is an accessible and low-cost renewable energy source. Nevertheless, wood burning may be an important source of air pollution. Examples of these types of pollution problems are shown in countries where winters are cold and different types of wood are used for heating purposes (Glasius et al., 2006; Hellén et al., 2006; Hedberg et al., 2002) and where this type of fuel is much cheaper in comparison with fossil fuels. In Zurich, Switzerland, wood has been estimated to contribute 10% of sub- micrometric particles (bPM1.0) and 20% of hydrocarbons (Lanz et al., 2007). In Chile, wood is the main fuel for the residential sector (20% of the total primary energy) (Ministry of Environment, Chile, 2012). This percentage is even higher in the south of the country. In Araucanía region and Los Lagos region in Chile, wood fuel is estimated to contribute 62 and 95% of PM2.5 respectively (Ministry of Environment, Chile, 2012). In the city of Temuco, PM2.5 wood contributes 80% to 90% of the total PM10, which is very high compared with 30% to 60% in Santiago, presenting air quality problems with serious consequences on public health (Cereceda-Balic et al., 2012; Díaz– Robles et al., 2008; Kavouras et al., 2001; Peters, 2005; Sanhueza et al., 2006; Sanhueza et al., 2009). Wood fuel generates a series of pollutants, such as volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), inorganic compounds (Johansson et al., 2004), and high emissions of particulate matter of different aerodynamic sizes (Alves, 2008). According to the US Environmental Protection Agency (EPA), biomass fuel generates a third of the fine particles (b 2.5 μm; PM2.5) in the atmosphere at a global level (Pope, 2000). Additionally, epidemiological studies across the world clearly show a relation between the increase in concentration of PM2.5 in the atmosphere and the increase in mortality rates (Pope, 2000), where sub micrometrical particles (b 1 μm; PM1.0) stand out. These smallest particles are of greatest interest because of their special implications for human health (Díaz-Robles et al., 2014; Lipsky and Robinson, 2006). As a product of wood burning, a variety of hazardous chemical compounds (polycyclic aromatic hydrocarbons (PAHs), anhydrosaccharides, carboxylic acids, among others) are emitted in gaseous phase and are also associated with particulate matter (PM10, PM2.5 and PM1.0) (Iinuma et al., 2007; Jenkins et al., 1996; Samburova et al., 2016). These pollutants, once released to the atmosphere, are mixed with those generated by other anthropogenic and natural sources, making it difficult to identify and quantify the original emission sources for each pollutant. Because of this, researchers have dedicated to identify and determine emission factors (EF) and concentration ratios (CR) for selected compound families, such as PAHs, as well as to obtain specific molecular markers (MM) for each type of atmospheric organic aerosol (AOA) generated by each type of emission source, like, for example, those coming from the biomass burning (Alves et al., 2011; Fine et al., 2002; Fine et al., 2004a; Fine et al., 2004b; Gonçalves et al., 2010; Gonçalves et al., 2011; McDonald et al., 2000; Nel, 2005; Simoneit and Elias, 2000). Through the use of specific EF for each type of source it is possible to create more faithful emissions inventories (EI), which allow for an estimation of the real impact of a given source on urban air quality, on the population health and on the climate change. In particular, EF, CR and MM from burning biomass, especially

firewood, allow, for example, for the identification and knowledge of the amount of standard pollutants (NOx, SO2, CO, PM10, PM2.5) generated in cities from the use of firewood fuel as heating and/or food preparation. Studies carried out by different authors (Alves et al., 2011; Fine et al., 2002; Fine et al., 2004a; Fine et al., 2004b; Gonçalves et al., 2010; Gonçalves et al., 2011; McDonald et al., 2000; Nel, 2005; Simoneit and Elias, 2000; Hays et al., 2002; Hellén et al., 2008) have put emphasis on the high variability of emission factors for PM2.5, which can be summarized in Table 1. The values for EFPM2.5 determined by different researchers vary widely between each other, probably due to important differences in their fuel characteristics, different burning devices (wood-burning stove and chimney), shaving sizes, moistures and wood types (origin and species), obtaining incomparable results. The goal of this work was to analyze the burning of wood representative from central-southern region of Chile, such as Eucalyptus globulus (EG) (introduced hardwood), Nothofagus obliqua (NO) (native hardwood) and Pinus radiata (PR) (introduced softwood), shaving size, with 0% humidity, in order to analyze the emissions profile at different stages of the combustion process and to determine emission factors for particulate matter PM2.5 and combustion gases (CO2, CO, NOx y SO2). In this work a Controlled Combustion Chamber for Emissions (3CE) was designed to control the most important variables affecting each stage emission process (SEP) in order to obtain reproducible, repeatable and comparable experiments and results. A complete picture of the emission process (pre-ignition, flame and smoldering stages) can be taken, without using diluted samples, and thus avoiding problems and artifacts associated with use of dilution tunnels (Kinsey et al., 2009; Lipsky and Robinson, 2006). 2. Material and methods 2.1. Controlled combustion chamber for emissions (3CE) Combustion tests were run in 3CE (Fig. 1), designed and manufactured by CETAM – UTFSM (Patent application No 843-2008, 2010 in Chile and International Patent Application No PCT/CL 00058, 2010, pending). 3CE is a thermally insulated combustion chamber designed to analyze combustion physicochemical parameters allowing to determine combustion efficiency and energy potential, together with the characterization of combustion emissions (gases and particles). 3CE consists of four functional units constructed in 316 stainless steel of high mechanical and thermal resistance: pre-chamber, combustion chamber, cooling zone, and emission collection system (Fig. 2). The parameters controlled by 3CE are flow and quality of oxidant, combustion chamber temperature (insulated oven), wood ignition procedure (by using an electrical resistance), sampling collection system and fuel mass by batch. The airflow received by 3CE is divided into primary air (directly injected into the combustion chamber) and secondary air (right after the combustion zone but previously to the sampling bag). Mass flow controllers were used to maintain constant airflows during the entire combustion test. The emission collection system of 3CE consists of a Tedlar bag coupled to the combustion chamber, which acts as a lung with an adaptation of four sampling ports designed for installation of 4 sampling cartridges equipped with quartz membrane filter


F. Cereceda-Balic et al. / Science of the Total Environment xxx (2017) xxx–xxx Table 1 Emission factors of PM2.5 from the hardwoods and softwoods combustion using residential appliances. Researchers

Residential appliances

Biomass

Amaral et al., 2014

Fireplace

Alves et al., 2011

Fireplace

Hardwoods Softwoods Hardwoods Softwoods Hardwoods Softwoods Hardwoods

Woodstove Fine et al., 2004a

Fireplace

Fine et al., 2002

Woodstove

Gonçalves et al., 2011

Fireplace Woodstove

Hays et al., 2002

Open burn

Hedberg et al., 2002 McDonald et al., 2000

Woodstove Woodstove

Emission factors (PM2.5 g kg−1)

3.18 ± 1.35 5.66 ± 1.03 9.9–13.4 14.2 4.2–15.1 16.3 0.88 ± 0.16–3.4 ± 0.5 Softwoods 1.1 ± 0.2–2.0 ± 0.3 Hardwoods 3.3–6.8 Softwoods 1.6–3.4 Hardwoods 7.8 ± 6.2–21 ± 10 Softwoods 6.9 ± 3.6 Hardwoods 5.8 ± 3.9–13 ± 8.3 Softwoods 5.2 ± 4.3 Foliars 10 ± 3.9 fuels Hardwoods 0.1–2.6 Hardwoods 2.3–8.3 Softwoods 2.9–9.0

and polyurethane foam filters (PUFs) for sample collection. Samples collected on cartridges allow for the identification and quantification of atmospheric organic and inorganic species present in aerosols, such as PAHs, VOCs, PCBs, elements, ions, among others (Cereceda-Balic et al., 2012; Kavouras et al., 2001; Sanhueza et al., 2006). The emission collection system of 3CE is equipped with a portable gas analyzer Testo 350-XL, Germany, with sensors for measurement of CO2, (IR sensor), CO, NOx and SO2 (electrochemical sensors). For particle size distribution measurement, a laser aerosol spectrometer Grimm Environmental Dust Monitor Model 107 (OPC, Germany) was used. This instrument classifies particles in 31 channels, from 0.265 μm to 34.00 μm, based on the intensity of the light scattered by the sampled particles at an angle of 90° with respect to the emitting laser beam (with wavelength of 675 nm), and then counts particles with an optical counter. To measure PM emissions, the gas stream is previously diluted using a Dilutor Emission Sampling System Grimm Aerosol Model 7.917 connected to an OPC from Grimm.

3

and similar wood shavings a semi-industrial machine was used (GENCO Energy Machinery, Type TFS-158, China). The obtained shavings were dried (0% humidity) in a temperature-regulating stove (80 °C, 24 h) until constant mass was achieved. 2.3. Determination of biomass composition Biomass samples were analyzed chemically (Table 2) to obtain lignin, cellulose and hemicellulose contents. Lignin was isolated using the Klason method (Raiskila et al., 2007), cellulose was quantified using the Kurshner-Hoffer method, and hemicellulose content was calculated by subtracting the content of cellulose to holocellulose (JaymeWise method). Using a Fourier transform infrared spectroscopy (FTIR model BX-II Perkin Elmer), functional lignin groups were identified and analyzed (Boeriu et al., 2004; García et al., 2009; Popescu et al., 2009; Sjoström, 1993). Elemental analysis (C, H, N and S) was made on a LECO Truspec analyzer, and Chlorine was determined with a ion chromatograph Methrom IC Pro, after digesting samples in a microwave digester Milestone Ethos 1. Proximate analysis was made in a thermogravimetric analyzer TGA Q500 from TA Instruments, yielding moisture, volatile matter, fixed carbon and ash. Higher heating values were measured with a Parr 1351 bomb calorimeter. Elemental and proximate compositions, and heating values of all tested samples are listed in Table 2. 2.4. Combustion parameters Based on the elemental analysis of the biomass, the stoichiometric flow required for combustion was determined for each experiment (Nussbaumer, 2003). Subsequently, optimal-point parameters for the operation were established: primary airflow, secondary airflow, flame temperature, wood mass, excess air coefficient (λ), oven temperature, and expected CO2 concentration. Table 2 shows that the variations in elemental analysis between the selected species are minimal. However, from stoichiometry, the mass of oxygen needed for combustion can be determined (Nussbaumer, 2003): C 1 H 1:56 O0:61 þ 1:09O2 → CO2 þ 0:78H 2 O ðEucalyptus globulusÞ

ð1Þ

C 1 H 1:56 O0:61 þ 1:09O2 → CO2 þ 0:78H 2 O ðNothofagus obliquaÞ

ð2Þ

C 1 H 1:37 O0:57 þ 1:06O2 → CO2 þ 0:68H 2 O ðPinus radiataÞ

ð3Þ

2.2. Generation of firewood and dry samples Wood samples were obtained from local certified distributors. Then, each wood species was separately transported to the laboratory in sealed plastic containers appropriately labelled. Once in the laboratory, wood samples were stored at room temperature separated by wood species in labelled and pre-cleaned coolers. To generate a reproducible

2.5. Filters for the collection of particulate material The type of filters required for sampling during the combustion tests was previously evaluated, taking into account effective PM collection. To achieve this, combustion tests were carried out in 3CE and samples

Fig. 1. Design of controlled combustion chamber "3CE".


4


Fig. 2. Experimental scheme for the wood combustion in controlled combustion chamber 3CE and sampling system.

were taken using teflon filters (TF), glass-fiber filters (GF) and quartzfiber filters (QF) with and without heat treatment. Before combustion tests, quartz fiber filters were prefired (Barnstead International muffle, model 1300) at 600 °C for 6 h to eliminate residual carbon (Gonçalves et al., 2010; Gonçalves et al., 2011) and then used to collect samples in a combustion test at 3CE. Five combustion tests were run by taking four samples in each one using 10 g of EG (shavings size, 0% humidity) and in similar optimized operational conditions. Then, sampled filters were weighed gravimetrically in a room specially conditioned to 25 °C and 50% HR, using a double-range analytical scale (Sartorius BP211D, Germany): 0 to 89 (±0.01 mg), 89 to 200 (±0.1 mg), incorporating a system of static elimination (Metter Toledo by Haug. Type: 01.7782.100. No: V41370/08, Germany). Selection criteria for filter media to be used were mechanical behavior and high reproducibility in PM collection (CV b 5%). Results indicate that quartz fiber filters with heat treatment present the best performance, achieving 2.27% variation coefficient between tests (n = 4).

2.6. Validation of sampling reproducibility In order to evaluate the reproducibility in the sampling ports built in the 3CE samples collection system, combustion tests were carried out using 4 sampling cartridges, all equipped with quartz and polyurethane foam filters. Wood burning tests were carried out using wood shavings (10 g) at 0% humidity and in optimized operation conditions defined beforehand. The filters previously described where then analyzed gravimetrically (same procedure than Section 2.5) in order to measure the particulate matter (PM2.5) mass collected, thus effectively verifying some variability for the 4 sampling ports (4 cartridges). Variability coefficient (CV, n = 3) was calculated and used as reproducibility criteria. Values of CV b 5% between sampling ports and CV b 10% between combustion tests were obtained in this validation process and are

considered acceptable as a quality criterion for the rest of this research (Papp et al., 1998). 2.7. Generation and analysis of blanks In order to avoid memory effects between combustion tests, a deep cleaning procedure (physical and chemical) for 3CE was designed and applied. The cleaning stages consisted in cleaning all of 3CE's 316 stainless steel components: pre–chamber, combustion chamber, cooling zone and emissions collection zone with a solution of Extran® MA 03 (alkaline detergent, phosphate free, Merck, 2% volume) using ultrapure water as dissolvent. Then, 3CE combustion chamber was assembled and the oven turned to 80 °C in order to dry it. In addition, a blank test was carried out which corresponds to a simulation of the combustion process, with the same conditions as in the combustion experiments, but without firewood. Blank test emissions were collected in sampling cartridges equipped with quarts membrane and polyurethane foam filters (PUFs). Additionally, real time measurement of PM2.5 concentration and particle size distribution was carried out using Grimm OPC, and combustion gases was measured using TESTO. Once blank test was finished, sampled filters were weighted (same conditions as in Section 2.5) and then extracted for hydrocarbons analysis on CETAM-UTFSM following the procedure described by Cereceda-Balic et al., 2012 and Kavouras et al., 2001. After the efficiency of each of the cleaning stages (physical and chemical) was verified, a blank combustion process was performed using 3CE at the same conditions as in the combustion tests. During the blank combustion test, no emissions of combustion gases and fine particulate matter were observed. Therefore, the cleaning procedures implemented for the combustion chamber and accessories were considered appropriate, thus guaranteeing that the results of combustion gases and PM2.5 emission factors presented no memory effects that might lead to overestimations. Also, chromatograms obtained from

Table 2 Elemental analysis (dry basis), proximate analysis (dry basis), chemical analysis (dry basis), heating values and moisture content of woods. Biomass

Pinus radiata Eucalyptus globulus Nothofagus obliqua

Moisture content (dry basis, % weight)

Elemental analysis (dry basis, % weight)

Proximate analysis (dry basis, % weight)

Chemical analysis (dry basis, % weight)

C

Volatile matter

Cellulose Hemicellulose Lignin HHV (MJ kg−1)

H

N

S

O

Fixed carbon

Ash

Heating values

7.12 5.95

50.83 5.81 0.05 0.017 38.77 81.60 49.94 6.50 0.08 0.022 40.53 87.93

13.88 9.14

4.52 39.9 2.93 41.0

24.2 19.4

27.4 28.1

20.96 19.26

7.83

49.68 6.46 0.11 0.020 40.21 87.51

8.97

3.52 35.8

28.7

27.3

18.01



5

Table 3 Combustion tests of Eucalyptus globulus for selection of primary air. Tests biomass (n = 5)

Actual air flow (L min−1)

Process duration (min)

Maximum temperature (°C)

Maximum CO2 (% volume)

λ

1 2 3 4 5

10.66 15.99 21.33 26.66 31.99

40 40 40 40 40

284.15 422.69 511.34 457.75 484.38

9.35 9.85 9.15 6.92 6.82

1 1.5 2 2.5 3

blank process showed no presence of hydrocarbons, thus discarding any memory effect for organic compounds. 2.8. Determination of emissions factors Emission factors were determined as the total number of particles emitted (np) or the total mass of a specific gaseous species i emitted (mi), divided by the waste biomass consumed (mbw) during the experiment. The total number of particles emitted was calculated (Eq. (4)) by integrating the product of the total air volume flowrate, V_ a (primary plus secondary), and the instantaneous particle number concentration emissions (cp in particle number per unit volume) over the measurement period, which was extended to 1500 s, since it was checked that all emissions were negligible beyond this limit. Finally, the total mass of gases emitted was determined (Eq. (5)) by integrating the product of the total air mass flowrate and the instantaneous mass fractions of gases (yi). Since the volume or mass flowrates remain unchanged during the experiment, they can be removed from the integrals in all equations. In the case of gaseous species, Eq. (5) needs mass fractions to be transformed into mole fractions because it is mole fractions (xi) what gas analysers provide, and mass flowrate can be written as a function of the pressure and temperature of the entering air flow (p and T):

np −1 ¼ EF p #kg ¼ mbw

V_ a

t¼1500 Z

cp dt t¼0

ð4Þ

mbw t¼1500 Z

EF i gkg

−1

V_ a W i

_ a dt yi m

m ¼ i ¼ mbw

t¼0

mbw

¼

t¼1500 Z

xi dt t¼0

ð5Þ

mbw RT=p

with V_ a being the total air volume flowrate, Wi the molecular weight of the measured gas, R the universal gas constant, and i being CO2, CO, NOx and SO2.

Fig. 3. Emissions during experiments with Eucalyptus globulus, Nothofagus obliqua and Pinus radiata. A) Carbon dioxide. B) Carbon monoxide.

3. Results and discussion 3.1. Optimization of the combustion parameters Average flame times for each combustion test were 420 s, 370 s and 390 s for PR, EG, and NO, respectively. Considering the relationship between the air volume and flame time (Vair/tflame) the stoichiometric

air flows (λ = 1) were identified (Nussbaumer, 2003) as 10.40 dm3 min− 1; 10.84 dm3 min− 1; 10.52 dm3 min− 1 for EG, NO, and PR, respectively. Since these stoichiometric flows were very similar, only tests using 10 g of EG (0% humidity, shaving size) as fuel were performed to determine the optimum excess air. When stoichiometric air flow (10.40 dm3 min− 1, λ = 1) was used, no flame was observed

Table 4 Combustion parameters for selected species at 0% moisture. (CI = (tn − 1, 95% ∗ S/√n). Tests (n = 3)

Grain size

Time of ignition (s)

Time of flame (s)

Time of smoldering(s)

Temperature of combustion process Maximum temperature (°C) (°C)

Combustion efficiency (%)


Shaving Shaving

12 ± 4 29 ± 3

420 ± 2.3 370 ± 1.4

1080 ± 2.4 1130 ± 1.3

267 ± 30 283 ± 74

527 ± 9 534 ± 18

97 ± 7 93 ± 6

Shaving

58 ± 6

390 ± 4.2

1110 ± 4.1

239 ± 23

453 ± 4

91 ± 4


6


Table 5 Emission factors (g kg−1) of combustion gases (CI = (tn − 1, 95% ∗ S/√n). Tests (n = 3)

CO

CO2

NO

NOX

SO2


49.35 ± 7.27 38.98 ± 6.07 41.99 ± 4.09

1947.52 ± 161.84 1701.62 ± 118.31 1801.67 ± 423.58

0.38 ± 0.08 0.54 ± 0.09 0.56 ± 0.14

0.41 ± 0.10 0.58 ± 0.10 0.61 ± 0.16

0.48 ± 0.20 0.37 ± 0.21 0.41 ± 0.13

because the amount of oxygen was insufficient for a complete biomass combustion (Nussbaumer, 2003). From tests with different excess air ratios (λ) the optimum amount of injected primary air to the combustion chamber 3CE (which remains approximately constant for 40 min, since the flue gas remains exactly constant regardless the biomass composition) was identified as that leading to maximum peaks of temperature and CO2 concentration; and minimum CO concentration. Table 3

shows that the test with 21.33 dm3 min−1 (λ = 2) produces highest combustion temperature and therefore fastest degradation of the cellulose, hemicellulose and lignin structures, these being the structures that produce highest pollutant emissions (mainly VOCs) (Papp et al., 1998; Hosoya et al., 2008; Kuo et al., 2008; Mamleev et al., 2007; Sellivan and Ball, 2012). Concentrations of CO2 (Table 3) were similar for the following injections of primary air: 10.66 dm3 min−1 (λ = 1), 15.99 dm3 min−1 (λ = 1.5) and 21.33 dm3 min−1 (λ = 2), with values of 1.12 × 105 mg m−3, 1.18 × 105 mg m−3, 1.09 × 105 mg m−3, respectively. However, the primary air flow of 10.66 dm3 min− 1 and 15.99 dm3 min−1, did not reach temperatures above 500 °C, because the oxidant flow in the primary air injection was not sufficient for the complete thermal degradation of cellulose, hemicellulose and lignin (Hosoya et al., 2008; Kuo et al., 2008; Mamleev et al., 2007; Sellivan and Ball, 2012; Shen and Bridgwater, 2010). Maximum sampling flow that can be achieved by 3CE is 40 dm3 min−1, which is determined by the sampling flow of each cartridge (10 dm3 min−1) installed in the four sampling ports in the Tedlar bag. Therefore, for EG, NO and PR, additionally to the 21.33 dm3 min−1 primary air, 18.67 dm3 min−1 of secondary air was also injected for the tests to reach equilibrium and to avoid any overpressure or vacuum in the chamber. Table 4 shows the chemical combustion efficiency (EC). This value is obtained as the ratio between experimental CO2 emissions (EF CO2 g kg−1 × g wood to be burned) and the theoretical CO2 emissions (obtained assuming complete combustion, and Wbw being the molecular weight of the biomass with unitary carbon atom, consistently with Eqs. (1–3)).

EC ¼

mCO2 experimental EF CO2 mbw W ¼ ¼ EF CO2 bw W CO2 mCO2 theoretical W CO2 mbw W bw

ð6Þ

Results observed in Table 4 indicate high combustion efficiency, with values above 90%. Therefore under this condition, most of the carbon in the wood structures are oxidized to CO2, lowering emissions of organic atmospheric aerosols and PM (Hosoya et al., 2008; Kuo et al., 2008; Mamleev et al., 2007; Sellivan and Ball, 2012; Shen and Bridgwater, 2010).

Fig. 4. Emissions during experiments with Eucalyptus globulus, Nothofagus obliqua and Pinus radiata. A) NOX. B) SO2.

Fig. 5. Emission factors of PM2.5 originated by hardwoods and softwoods combustion.



7

Fig. 7. Duration of SEP for experiments with Eucalyptus globulus (EG), Nothofagus obliqua (NO) and Pinus radiata (PR). Fig. 6. FTIR result. Ratios of O-H groups and C-O groups.

2.01 g kg−1 ± 0.30 g kg− 1and NO: 2.37 g kg−1 ± 0.54 g kg−1. EFPM2.5 obtained with the OPC were (average, n = 3): PR: 0.84 g kg− 1 ± 0.10 g kg−1; EG: 1.06 g kg− 1 ± 0.13 g kg− 1 and NO: 1.30 g kg−1 ± 0.25 g kg−1. EFPM2.5 obtained with the laser OPC and gravimetrically were submitted to statistical tests of medians comparison and Kolmogorov Smirnov (Guerrero et al., 2013) indicating that there are no statistically significant differences between them with 95% probability. The gravimetric EFPM2.5 for NO and EG were 1.58 and 1.34 times higher than EFPM2.5 for PR. This can be explained because of their different compositions in cellulose, hemicellulose and lignin (Hosoya et al., 2008; Kuo et al., 2008; Mamleev et al., 2007; Sellivan and Ball, 2012; Shen and Bridgwater, 2010), which can be seen in Table 2, where the concentration of cellulose and lignin present in EG is 1.5 and 1.2 times higher than in PR, respectively and the hemicellulose concentration in NO is 1.2 times higher than in PR, this can explain the higher EFPM2.5 of these hardwoods. Pine (softwood) is a conifer with lignin much more complex than that of hardwoods. It is demonstrated by FTIR analysis (Fig. 6), were ratio between the O-H groups of guaiacil and siringil structures and C-O groups in PR is 5 times higher than for NO and 3.5 times higher than EG. It is believed that there is a correlation between the types of functional groups with the compounds generated by incomplete combustion, meaning that softwood (PR) is producing less PM2.5 than hardwoods (NO and EG), in part, because of their differences in chemical composition. EFs of PM2.5 obtained in this study are not comparable with emission factors for PM2.5 found in the literature (Alves et al., 2011; Fine et al., 2002; Fine et al., 2004a; Fine et al., 2004b; Gonçalves et al., 2010; Gonçalves et al., 2011; McDonald et al., 2000), where values range from 2.3 g kg−1 to 20.2 g kg−1 for combusted hard and soft wood were obtained. These differences are mainly due to combustion characteristics, wood shavings, wood moistures, types of dilution ducts, sampling procedures, use of different combustion appliances (fireplaces and wood stoves), and different operating conditions of the combustion appliances, among others. This reinforces the need to

3.2. Emission factors for combustion gases (CO2, CO, SO2, and NOX) Maximum average concentrations in the flue gas after the injection of secondary air were observed (n = 3) for CO2 (EG: 1.15 × 105 mg m−3; NO: 1.22 × 105 mg m− 3; PR: 1.36 × 105 mg m−3) and minimum concentrations were observed for CO (EG: 1.31 × 103 mg m− 3; NO: 2.06 × 103 mg m−3; PR: 1.90 × 103 mg m−3) during the flame phase (Fig. 3a–b). Emission factors obtained for the combustion gases can be seen in Table 5. The maximum EFCO2 was obtained for PR, mainly because with a high combustion efficiency at maximum temperatures above 500 °C a complete oxidation is expected for the structures of cellulose, hemicellulose and lignin, leading to lower emissions of fine particulate matter. Furthermore, it should be considered that during thermal fragmentation of cellulose, hemicellulose and lignin, a series of heterolytic and homolytic breakage is generated, which creates VOCs, SVOCS and anhydrosaccharides. These, through elimination reactions such as decarbonylation and de-carboxylation gives rise to CO and CO2 and thus to their emission factors (Hosoya et al., 2008; Kuo et al., 2008; Mamleev et al., 2007; Sellivan and Ball, 2012; Shen and Bridgwater, 2010). Concentrations of NOx, and SO2 (Fig. 4a–b) depend on nitrogen and sulphur content in the biomass. Biomass contains low concentrations of combustible sulphur and nitrogen which results in lower emission factors for these gases compared to CO (Hrdlicka et al., 2016; Bignal et al., 2008; Forbes et al., 2014).

3.3. Emission factors for fine particulate matter (PM2.5) Fig. 5 shows emission factors for particulate matter from different species of wood shavings size obtained in real time (Grimm OPC) compared with gravimetric EFs. The results for EFPM2.5 obtained gravimetrically (average, n = 3) were: PR: 1.54 g kg− 1 ± 0.32 g kg− 1; EG:

Table 6 PM2.5 emission factors and contribution of the stages, obtained with OPC. Pre-ignition EFPM2.5 (g kg Pinus radiata Eucalyptus globulus Nothofagus obliqua

0.004 0.009 0.004

Flaming −1

)

%

EFPM2.5 (g kg

0.48 0.85 0.31

0.65 0.56 0.52

Smoldering −1

)

%

EFPM2.5 (g kg

77.61 52.88 39.59

0.18 0.49 0.78

Total −1

)

%

EFPM2.5 (g kg−1)

21.91 46.27 60.10

0.84 1.06 1.30


8


have a standardized combustion equipment such as that presented in this work (3CE) and the possibility of conducting studies of different wood type and testing different combustion conditions in the same equipment so as to obtain reproducible and comparable results. 3.4. Contribution of EFPM2.5 in each SEP: ignition, flame, and smoldering phase In Table 6 and Fig. 7, PM emitted in each SEP (pre-ignition, flame and smoldering phase) are shown. This innovative analysis helps to determine the PM2.5 contribution of each emission stage, demonstrating that particulate matter concentration measured only in stationary state may cause underestimation of emissions. As observed in Table 6, PR is the wood species that generates highest emissions in the flame phase with an EFPM2.5 of 0.65 g kg− 1 ± 0.08 g kg−1 (77.61% of Total Emissions Contribution (TEC)), followed by EG and NO with values of 0.56 g kg−1 ± 0.05 g kg−1 (52.88% of TEC) and 0.52 g kg−1 ± 0.03 g kg−1 (39.59% of TEC), respectively. In the smoldering phase, NO generates the highest emission with a EFPM2.5 of 0.78 g kg−1 ± 0.10 g kg− 1 (60.10% of TEC), followed by EG and PR with values of 0.49 g kg−1 ± 0.09 g kg−1 (46.27% of TEC) and 0.18 g kg−1 ± 0.02 g kg−1 (21.91% of TEC), respectively. These results demonstrate that emissions contribution from flame phase and smoldering phase are very significant. Smoldering phase contribution is between 22% to 60% while the flame phase contribution is between 40% to 78%. Finally, the ignition stage, generates an emission of PM2.5 always below 1%. Thus, all of the emission the stages are important in order to properly evaluate the emission of particulate matter and must be considered to obtain the EFPM2.5. Considering these results, in order to obtain reliable and accurate EF, test samples should be taken in a transient manner, considering the entire combustion process (pre-ignition, flame and smoldering phase). This way, emission factors for PM2.5 will not be underestimated, and leading to more accurate emission inventories, especially in cities where biomass burning is a major source of pollution with a high contribution of PM2.5 (Kavouras et al., 2001; Lipsky and Robinson, 2006). 3.5. Number and particle size (μm) emitted during the combustion process Fig. 8a–c shows the behavior (number and size distribution in time) of particles in the combustion process. It can be seen that the finest particles (b 0.5 μm) predominate in all SEP, which have a bimodal behavior with a maximum number of particles emission in the flame stage (the average time of EG was 29 s to 256 s; NO: 58 s to 206 s and PR: 12 s to 285 s). Then, the number of emitted particles decreases, and in the smoldering stage (average time period for EG was 370 s to 1500 s; NO: 390 s to 1500 s and PR: 420 s to 1500 s) a new peak of particles smaller than 0.5 μm can be observed. Parametric statistical tests (tests of multiple ranges) and nonparametric tests (Kruskal - Wallis) were applied to all particle size distribution data below 0.5 μm obtained from any stage. It was found that the particle number distribution in all three emission stages (pre-igniton, flame, and smoldering) showed statistically significant differences between them with 95% probability. Table 7 Percentage (%) of particles numbers for different sizes (b0.5 μm) during an integrated combustion process, including pre-ignition, flame and smoldering phase.

Fig. 8. Particle concentration (# / cm3) by size during combustion process. A) Pinus radiata. B) Eucalyptus globulus. C) Nothofagus obliqua.

Particle size (μm)

Biomass Pinus radiata (%)

Eucalyptus globulus (%)

Nothofagus obliqua (%)

0.265 0.290 0.325 0.375 0.435 0.475

33 25 24 13 4 1

29 23 24 14 6 3

21 19 21 17 10 6



Flame stage showed the higher amount of particles emitted below 0.5 μm followed by smoldering stage, and finally pre-ignition stage. Table 7 shows the different percentages of fine particles that are emitted in the 3CE. All studied wood species have highest percentage of particles within a size of 0.265 μm (between 21% to 33%). It can be considered that the smaller the particle size, the deeper they can travel into the respiratory system and therefore can cause greater public health demage. PR, with 33% particles emission in the range of 0.265 μm, emits the largest number of particles of that size, followed by EG (29%) and NO (21%). This information is very important to assess the effects on the health of the population because according to the literature (Peters, 2005; Sanhueza et al., 2006; Sanhueza et al., 2009; Pope, 2000; Díaz-Robles et al., 2014; Lipsky and Robinson, 2006; Lei et al., 2005, O'Neill et al., 2005; Kennedy et al., 2007), deaths due to exposure to fine particles are significant and an important public health problem arises in many countries around the world due to this type of atmospheric pollutant. 4. Conclusions The home-made 3CE controlled combustion chamber (PCT patent pending) has demonstrated to be a reliable equipment to obtain reproducible and comparable combustion tests, allowing to control and modify critical variables that affect the combustion process. It has been used for emission measurements from combustion of Eucalyptus globulus, Nothofagus obliqua, and Pinus radiata at 0% humidity with shaving size. Combustion conditions were optimized, using as criteria the highest CO2 emission and highest combustion temperature among the air flows tested. Combustion efficiencies were always higher than 90%. The results obtained with optimized operating conditions resulted in the lowest EFPM2.5 for Pinus radiata (1.54 g kg−1 ± 0.32 g kg−1, softwoods), followed by Eucalyptus globulus (2.01 g kg−1 ± 0.30 g kg−1, hardwoods) and Nothofagus obliqua (2.37 g kg−1 ± 0.54 g kg−1, hardwoods). Differences in emissions can be explained by differences in their concentration of hemicellulose, lignin and cellulose. Significant contributions of PM2.5 were generated during every stage (pre-ignition, flame and smoldering). This demonstrates that particle concentrations measured only in stationary state during flame stage may cause underestimation of emissions. The methodology proposed allows for high accuracy in the determination of the EF values of PM2.5 produced by the combustion of biomass. Finally, the percentages of fine particles emitted from combustion of studied wood species were measured, and all wood species analyzed showed highest percentage of particles in the lowest range size of 0.265 μm (between 21% to 33%). PR, with 33% of the emitted particles in the range of 0.265 μm, emits the largest number of particles of that size, followed by EG (29%) and NO (21%). Considering that the smaller the particle size, the deeper they can penetrate into the respiratory system and can cause greater public health effects, these results are useful to evaluate not only the environmental impacts but also the health effects of wood consumption in the central -southern region of Chile. Acknowledgments The authors would like to thank FONDEF D08-I-1147, Fondecyt No. 1131028, Fondecyt No. 1120791 and No. 1161793, Projects USM-DGIP (PIIC), Conicyt – PAI/grant for the national competition for PhD thesis 2014-78141103, and Conicyt – “Programa Atracción de Capital Humano Avanzado Extranjero MEC” Project No. 80140096 for their financial support. References Alves, C.A., 2008. Characterisation of solvent extractable organic constituents in atmospheric particulate matter: an overview. An. Acad. Bras. Cienc. 80 (1), 21–82. Alves, C., Gonçalves, C., Fernandes, A.P., Tarelho, L., Pio, C., 2011. Fireplace and woodstove fine particle emissions from combustion of western Mediterranean wood types. Atmos. Res. 101 (3), 692–700.

9

Amaral, S.S., Junior, J.A., Costa, M.C., Neto, T.G., Dellani, R., Leite, L.H., 2014. Comparative study for hardwood and softwood forest biomass: chemical characterization, combustion phases and gas and particulate matter emission. Bioresour. Technol. 164 (1), 55–63. Bignal, K.L., Langridge, S., Zhou, J.L., 2008. Release of polycyclic aromatic hydrocarbons, carbon monoxide and particulate matter from biomass combustion in a wood-fired boiler under varying boiler conditions. Atmos. Environ. 42 (39), 8863–8871. Boeriu, C., Bravo, D., Gosselink, R., Van Dam, J., 2004. Characterisation of structure-dependent functional properties of lignin with infrared spectroscopy. Ind. Crop. Prod. 20 (2), 205–218. Cereceda-Balic, F., Fadic, X., Llanos, A.L., Domínguez, A.M., Guevara, J.L., Vidal, V., DíazRobles, L.A., et al., 2012. Obtaining polycyclic aromatic hydrocarbon concentration ratios and molecular markers for residential wood combustion: Temuco, a case study. J. Air Waste Manage. Assoc. 62 (1), 45–51. Díaz–Robles, L.A., Ortega, J.C., Fu, J.S., Reed, G.D., Chow, J.C., Watson, J.C.A., et al., 2008. Hydrid ARIMA and artificial neutral netwoorks model forecast particulate matter in urban areas: the case Temuco Chile. Atmos. Environ. 42 (32), 8331–8340. Díaz-Robles, L.A., Fu, J.S., Vergara Fernández, A., Etcharren, P., Schiappacasse, L.N., Redd, G.D., Silva, M.P., 2014. Health risks caused by short term exposure to ultrafine particles generated by residencial wood combustión: a case study of Temuco, Chile. Envirom. Int. 66, 174–181. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2002. Chemical characterization of fine particle emissions from the fireplace combustion of woods grown in the southern United States. Environ. Sci. Technol. 36 (7), 1442–1451. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004a. Chemical characterization of fine particle emissions from fireplace combustion of woods grown in the Midwestern and western United States. Environ. Eng. Sci. 21 (3), 387–409. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004b. Chemical characterization of fine particle emissions from the woodstove combustion of prevalent United States tree species. Environ. Eng. Sci. 21 (6), 705–721. Forbes, E.G.A., Easson, D.L., Lyons, G.A., McRoberts, W.C., 2014. Physico-chemical characteristics of eight different biomass fuels and comparison of combustion and emission results in a small scale multi-fuel boiler. Energy Convers. Manag. 87 (1), 1162–1169. García, A., Toledanoa, A., Serranoa, L., Egüesa, I., González, M., Marín, F., et al., 2009. Characterization of lignins obtained by selective precipitation. Ser. Purif. Technol. 68 (2), 193–198. Glasius, M., Ketzel, M., Wåhlin, P., Jensen, B., Monster, J., Berkowicz, R., et al., 2006. Impact of wood combustion on particle levels in residential area in Denmark. Atmos. Environ. 40 (37), 7115–7124. Gonçalves, C., Alves, C., Evtyugina, M., Mirante, F., Pio, C., Caseiro, A., et al., 2010. Characterisation of PM10 emissions from woodstove combustion of common woods grown in Portugal. Atmos. Environ. 30 (35), 4474–4480. Gonçalves, C., Alves, C., Fernandes, A.P., Monteiro, C., Tarelho, L., Evtyugina, M., et al., 2011. Organic compounds in PM2.5 emitted from fireplace and woodstove combustion of typical Portuguese wood species. Atmos. Environment 45 (27), 4533–4545. Guerrero, F., Vidal, V., Cubillos, F., Espinosa, D., Cereceda–Balic, F., 2013. Emission factors for PM2.5 generated from the combustión Eucalyptus globulus and Nothofagus obliqua on ideal conditions using a cotrolled combustión chamber 3CE. AWMAs 106 th Annual Conference and Exhibition, Chicago, United States. Hays, M.D., Geron, C.D., Linna, K.J., Smith, N.D., Schauer, J.J., 2002. Speciation of gas-phase and fine particle emissions from burning of foliar fuels. Environ. Sci. Technol. 36 (11), 2281–2295. Hedberg, E., Kristensson, A., Ohlsson, M., Johansson, C., Johansson, P.Å., Swietlicki, E., et al., 2002. Chemical and physical characterization of emissions from birch wood combustion in a wood stove. Atmos. Environ. 36 (30), 4823–4837. Hellén, H., Hakola, H., Pirjola, L., Laurila, T., Pystynen, K.H., 2006. Ambient air concentrations, source profiles and source apportionment of 71 different C2–C10 volatile organic compounds in urban and residential areas of Finland. Environ. Sci. Technol. 40 (1), 103–108. Hellén, H., Hakola, H., Haaparanta, S., Pietarila, H., Kauhaniemi, M., 2008. Influence of residential wood combustion on local air quality. Sci. Total Environ. 393 (2), 283–290. Hosoya, T., Kawamuto, H., Saka, S., 2008. Diferent pyrolytic pathways of levoglucosan in vapor – and liquid/solid – phases. J. Anal. Appl. 83 (1), 64–70. Hrdlicka, J., Skopec, P., Dlouhy, T., Hrdlicka, F., 2016. Emission factors of gaseous pollutants from small scale combustion of biofuels. Fuel 165 (1), 68–74. Iinuma, Y., Bruggemann, E., Gnauk, T., Muller, K., Andreae, M.O., Helas, G., Parmar, R., Herrmann, H., 2007. Source characterization of biomass burning particles: the combustion of selected European conifers, African hardwood, savanna grass, and German and Indonesian peat. J. Geophys. Res. Atmos. 112. Jenkins, B.M., Jones, A.D., Turn, S.Q., Williams, R.B., 1996. Emission factors for polycyclic aromatic hydrocarbons from biomass burning. Environ. Sci. Technol. 30, 2462–2469. Johansson, L.S., Leckner, B., Gustavsson, L., Cooper, D., Tullinh, C., Potter, A., 2004. Emission characteristics of modern and old-type residential boilers fired with wood logs and wood pellets. Atmos. Environ. 38 (25), 4183–4195. Kavouras, G.I., Koutrakis, P., Cereceda–Balic, F., Oyola, P., 2001. Source apportionment of PM10 and PM2.5 in five Chilean cities using factor analysis. J. Air Waste Manage. Assoc. 51 (3), 451–464. Kennedy, R., Elton-Marshall, T., Sendzik, T., 2007. Tobacco smoke pollution in outdoor hospitality environments–a study of PM2.5 levels in restaurants and bars. Proceedings of the 13th Annual Society for Nicotine and Tobacco. Research Conference, Austin Texas, United States. Kinsey, J.S., Karilier, P.H., Dong, Y., 2009. Evaluation of methods for the physical characterization of the fine particle emissions from two residential wood combustión appliances. Atmos. Environ. 43 (32), 4959–4967.


10


Kuo, L., Helbert, B.E., Louchouran, P., 2008. Can levoglucosan be used characterize and quantify char/charcoal black carbon. Org. Geochem. 39 (10), 1466–1478. Lanz, V.A., Alfarra, M.R., Baltensperger, U., Buchmann, B., Hueglin, C., Prévot, A.S.H., 2007. Source apportionment of submicron organic aerosol at an urban site by factor analytical modelling of aerosol mass spectra. Atmos. Chem. Phys. 7 (6), 1503–1522. Lei, Y.C., Hwang, J.S., Chan, C.C., Lee, C.T., Cheng, T.J., 2005. Enhanced oxidative stress and endothelial dysfunction in streptozotocin-diabetic rats exposed to fine particles. Environ. Res. 99 (3), 335–343. Lipsky, A.M., Robinson, A.L., 2006. Effects of dilution on fine particle mass and partitioning of semivolatile organics in diesel exhaut and wood smoke. Environ. Sci. Technol. 40 (1), 155–162. Mamleev, V., Bourbigot, S., Yuen, J., 2007. Kinetic analysis of the thermal decomposition of cellulose: the main step of mass loss. J. Anal. Appl. Pyrolysis 80 (1), 151–165. McDonald, J.D., Zielinska, B., Fujita, E.M., Sagebiel, J.C., Chow, J.C., Watson, J.G., 2000. Fine particle and gaseous emission rates from residential wood combustion. Environ. Sci. Technol. 34 (11), 2080–2091. Ministry of Environment, 2012. Official Environment Status Report. Santiago de Chile. http://www.mma.gob.cl/1304/articles52016_InformeEsradoMedioAmbiente_ Completo_2011.pdf. Nel, A., 2005. Air pollution-related illness: effects of particles. Science 308 (5723), 804–806. Nussbaumer, T., 2003. Combustion and co-combustion of biomass: fudamentals, technologies and primary measures for emission reduction. Energy Fuel 17 (6), 1510–1521. O' Neill, M.S., Veves, A., Zanobetti, A., Sarnat, J.A., Gold, D.R., 2005. Economides PA. Diabetes enhances vulnerability to particulate air pollution-associated impairment in vascular reactivity and endothelial function. Circ. Res. 111 (22), 2913–2920. Papp, M.L., Elkins, J.B., Musick, D.R., Messner, M.J., 1998. Data quality objectives for the PM2.5 monitoring data. U.S. Environmental Agency Portection. Research Triangle Park, NC.

Peters, A., 2005. Particulate matter and heart disease: evidence from epidemiological studies. Toxicol. Appl. Pharmacol. 207 (2), 477–482. Pope, C.A., 2000. Review: epidemiological basis for particulate air pollution health standards. Aerosol Sci. Technol. 32 (1), 4–14. Popescu, C., Popescu, M., Singurel, G., Vasile, C., Argyropoulus, D., Wilfor, S., 2009. Vibrational spectroscopy and X-ray diffraction methods to establish the differences between hardwood and softwood. Carbohydr. Polym. 77 (4), 851–857. Raiskila, S., Pulkkinen, M., Laakso, T., Fagerstedt, K., Löija, M., Mahlberg, R., et al., 2007. FTIR spectroscopic prediction of klason and acid soluble lignin variation in Norway spruce cutting clones. Silva. Fenn. 41 (2), 351–371. Samburova, V., Connolly, J., Gyawali, M., Yatavelli, R.L.N., Watts, A.C., Chakrabarty, R.K., Zielinska, B., Moosmuller, H., Khlystov, A., 2016. Polycyclic aromatic hydrocarbons in biomass-burning emissions and their contribution to light absorption and aerosol toxicity. Sci. Total Environ. 568, 391–401. Sanhueza, P., Vargas, C., Mellado, P., 2006. Impact of air pollution by particulate matter (PM10) on daily mortality in Temuco, Chile. Rev. Med. Chil. 134 (6), 754–761. Sanhueza, P.A., Torrealba, M.A., Díaz-Robles, L.A., Schiappacasse, L.N., Silva, M.P., Astelle, T.D., 2009. Particulate air pollution and health effects for cardiovascular and respiratory causes in Temuco, Chile: a wood – smoke - polluted urbana real. J. Air Waste Manage. Assoc. 59 (12), 1481–1488. Sellivan, A.L., Ball, R., 2012. Thermal decomposition and combustion chemistry of cellulosic biomass. Atmos. Environ. 47 (1), 133–141. Shen, D.K., Bridgwater, A.V., 2010. Study on the pyrolitic behavior of xilan – based hemicellulosa using TG-FTIR and PY – GC – FTIR. J. Anal. Appl. 87 (2), 199–206. Simoneit, B.R.T., Elias, V.O., 2000. Organic tracers from biomass burning in atmospheric particulate matter over the ocean. Mar. Chem. 69 (3–4), 301–312. Sjoström, E., 1993. Wood chemistry. Fundamentals and Applications, second ed. Academic Press, Inc., San Diego, p. 133.
