Torrefied and Carbonized Wood, Fuel Properties and Turn of Exothermic Reaction

Torrefied and carbonized wood, fuel properties and turn of exothermic reaction

Master’s Thesis Rautiainen Mari

Department of Forest Sciences Faculty of Agriculture and Forestry University of Helsinki

2014

Tiedekunta/Osasto  Fakultet/Sektion  Faculty Laitos  Institution  Department Faculty of Agriculture and Forestry Department of Forest Sciences

Tekijä  Författare  Author Rautiainen Mari Työn nimi  Arbetets titel  Title Torrefied and carbonized wood, fuel properties and turn of exothermic reaction Oppiaine Läroämne  Subject Forest Resource Science and Technology Työn laji  Arbetets art  Level Aika  Datum  Month and Sivumäärä  Sidoantal  Master’s thesis year Number of pages 11/2014 77p. Tiivistelmä  Referat  Abstract

Carbonization is thermochemical conversion, where biomass is thermally degraded in the absence of oxygen. Solid char, pyrolysis oil and non-condensable gases are produced from the biomass. Torrefaction is early phase of the carbonization in temperatures of 220–300 °C. Torrefied wood is promising as a renewable fuel for industrial use in coal co-combustion and gasification-combustion.

Torrefaction and carbonization increase the higher heating value and fuel properties of wood compared to untreated wood. There’s a lack of knowledge in torrefaction and carbonization effects to higher heating value, carbon content and turn from endothermic to exothermic reaction of conifer zone wood species .

Raw material was stemwood of birch (Betula pubescens) and pine (Pinus sylvestris) including bark. Trees were harvested from the Helsinki district and chipped , particle size 16 ≤ 8 mm . Samples were torrefied and carbonized at 250, 300, 350, 400 and 450 ˚C without nitrogen flow. Carbon content (%), higher heating value (MJ/kg), mass yield (%) and turn of endothermic to exothermic reaction were inspected.

Carbon content of untreated birch and pine increased from 47 % to 82 % (at 450 ˚C). Higher heating value exceeded 26 MJ/kg at 300 ˚C and 28 MJ/kg at 400 ˚C, reaching bituminous coal’s values. Mass yield declined to 45–54 % of the initial mass at 300 °C. In low temperature, gradual exothermic peak was observable. In higher temperatures peak was evident. Carbonization and torrefaction improved the higher heating value and carbon content of wood but decreased the solid char yield. Avainsanat  Nyckelord  Keywords birch, carbon content, carbonization, exothermic reaction, higher heating value, mass yield, pine, pyrolysis, torrefaction Säilytyspaikka  Förvaringsställe  Where deposited Department of Forest Sciences, Viikki Science Library Muita tietoja  Övriga uppgifter  Further information

Tiedekunta/Osasto  Fakultet/Sektion  Faculty Laitos  Institution  Department Maatalous ja metsätieteellinen tiedekunta Metsätieteiden laitos

Tekijä  Författare  Author Rautiainen Mari Työn nimi  Arbetets titel  Title Torrefioidun ja hiilletyn puun polttoaineominaisuudet ja eksoterminen reaktio Oppiaine Läroämne  Subject Metsävaratiede ja -teknologia Työn laji  Arbetets art  Level Aika  Datum  Month and Sivumäärä  Sidoantal  Pro gradu -tutkielma year Number of pages 11/2014 77s. Tiivistelmä  Referat  Abstract Hiiltäminen on termokemiallinen ilmiö, missä biomassa hajotetaan lämmön avulla hapettomassa tilassa. Hiiltä, pyrolyysiöljyä ja kaasuja muodostuu tällöin biomassasta. Torrefiointi on hiiltämisen miedompi vaihe, jossa käsittelylämpötila on 220–300 °C. Torrefioitu puu on lupaava teollisuuden uusiutuvan energian raaka-aine kivihiilen yhteispolttoon tai kaasuttamiseen.

Torrefiointi ja hiiltäminen lisäävät puun lämpöarvoa ja polttoaineominaisuuksia verrattaessa käsittelemättömään puuhun. Skandinaavisten puulajien torrefioinnin ja hiiltämisen vaikutuksia lämpöarvoihin ja hiilipitoisuuteen sekä eksotermisen reaktion hyödyntämistä lämmitysenergiassa ei ole tutkittu laajasti.

Raaka-aineena on käytetty koivun (Betula pubescens) ja männyn (Pinus sylvestris) kuorellista runkopuuta. Puut kaadettiin Helsingin alueelta ja haketettiin. Tutkittavaksi otettiin siivilöidyt hakepalat 16 ≤ 8 mm . Näytteet torrefioitiin ja hiillettiin 250, 300, 350, 400 ja 450 ˚C:ssa ilman typpivirtaa. Hiilipitoisuus (%), ylempi lämpöarvo (MJ/kg), massan saanto (%) ja eksotermisen reaktion ajoitus tutkittiin.

Käsittelemättömän koivun ja männyn hiilipitoisuus nousi 47 %:sta 82 %:n käsittelylämpötilan ollessa 450 ˚C. Ylempi lämpöarvo oli yli 28 MJ/kg 400 ˚C:n käsittelyn jälkeen, mikä vastaa kivihiilen lämpöarvoa. Jo 300 ˚C:n käsittely nosti lämpöarvon 26 MJ/kg. Massan saanto väheni 45-54 %:a alkuperäisestä massasta hiillettäessä 300 ˚C. Alemmissa käsittelylämpötiloissa (250–300 ˚C) eksoterminen reaktio oli havaittavissa. Korkeammissa lämpötiloissa lämpötilapiikki oli selvempi. Hiiltäminen ja torrefiointi paransivat puun lämpöarvoja ja hiilipitoisuutta, mutta laskivat kiinteän hiilen saantoa. Avainsanat  Nyckelord  Keywords eksoterminen reaktio, hiilipitoisuus, hiiltäminen, koivu, massan saanto, mänty, pyrolyysi, torrefiointi, ylempi lämpöarvo Säilytyspaikka  Förvaringsställe  Where deposited Metsätieteiden laitos, Viikin kirjasto

Muita tietoja  Övriga uppgifter  Further information

Preface

The thesis was part of the BalBic project, which aimed to develop industrial biocoal production (including both torrefied wood and charcoal) and markets as well as the sustainable use of bioenergy. Torrefied wood is thermally degraded wood used in energy production. The joint project between University of Helsinki, Forestry Development Centre Tapio and Latvian State Forest Research Institute “Silava” was financed partly by European Union trough Central Baltic Interreg IV A Programme 2007–2013. Work package of which thesis included researched manufacturing technologies of biocoal and behaviour of wood in different process conditions.

I would like to thank for enabling this thesis Mikko Havimo, Jukka Hyytiäinen and Marketta Sipi from Helsinki University’s Faculty of Agriculture and Forestry and also Arja Santanen, Marjut Wallner and Tony Kiuru helping me with the laboratory tests. I’m grateful to my family for all the support they have given me.

TABLE OF CONTENTS

List of abbreviations and explanations ...... 7

1. Introduction ...... 8

1.1 Renewable energy consumption ...... 8

1.2 Background and purpose of the study ...... 10

2. Literary study ...... 11

2.1 Chemical composition of wood ...... 11

2.2 Pyrolysis and process conditions ...... 14 2.2.1 Carbonization ...... 15 2.2.2 Torrefaction ...... 17 2.2.3 Temperature, holding time, heating rate and pressure ...... 17

2.3 Properties of charcoal and torrefied wood as fuel ...... 19 2.3.1 Heating value ...... 20 2.3.2 Mass yield ...... 22 2.3.3 Carbon and ash content ...... 23 2.3.4 Endo- and exothermic reaction ...... 26

2.4 Carbonization technologies ...... 28 Batch and continuous process and kiln types ...... 28 Internal, external heating and circulation gas ...... 30

3. Material and methods ...... 31

3.1 Pyrolysis tests ...... 31 3.2 Raw material ...... 32 3.2.1 Chip size and distribution ...... 33 3.2.2 Determination of moisture content of pre-dried wood chips ...... 35 3.2.3 Pyrolysis temperature ...... 36 3.3 Higher heating value...... 37 3.4 Mass yield ...... 38 3.5 Carbon content...... 38 3.6 Detecting exothermic reaction ...... 38 5

4. Results ...... 40

4.1 Higher heating value...... 40 4.2 Mass yield ...... 42 4.3 Carbon content...... 44 4.4 Exothermic reaction...... 46

5. Discussion ...... 49

5.1. Higher heating value...... 49 5.2. Mass yield ...... 52 5.3. Carbon content...... 53 5.4. Exothermic reaction...... 54 5.5. Analysis of sampling and profitability of pyrolysis ...... 55

6. Conclusions ...... 56

References ...... 58

Appendix 1, Chip size of untreated pine and birch including bark ...... 71

Appendix 2.1, Calculation of HHV of birch samples ...... 72

Appendix 2.2 Calculation of HHV of pine samples ...... 73

Appendix 3.1 Birch with bark torrefied at 300 ˚C and 350 ˚C ...... 74

Appendix 3.1 Birch with bark carbonized at 400 ˚C and 450 ˚C ...... 75

Appendix 3.2 Pine torrefied at 250 ˚C and 300 ˚C ...... 76

Appendix 3.2 Pine with bark carbonized at 350 ˚C and 400 ˚C ...... 77

List of abbreviations and explanations

charcoal a solid matter produced from wood in a carbonization process endothermic reaction absorbs energy from the system in the form of heat exothermic reaction release energy from the system in the form of heat HHV higher heating value (MJ/kg), amount of heat released during the complete combustion of a specified amount of fuel and the water vapour produced is condensed into liquid form peak temperature the highest reached temperature during torrefaction or carbonization process pyrolysis degradation of biomass in absence of oxygen. Collective term for torrefaction and carbonization in this study temperature shortened term for (slow) pyrolysis temperatures (torrefaction or carbonization temperatures, °C) in this study torrefaction is a thermochemical treatment of biomass at 220 – 300 °C

1. Introduction

1.1 Renewable energy consumption

The share of renewable energy of the total energy consumption is increasing from 2005 levels due to the EU energy targets in 2020 and 2050. Targets are to reduce greenhouse gas emissions, improve energy efficiency and increase the share of renewable energy resources (The EU climate and energy package 2011, Roadmap 2050 2011.) Fulfilling these targets renewable energy sources has to exploit and develop. Wood energy consumption is expected to increase from 346 million m³ by 66 % to 2020 depending on other renewable energy sources development and implementation of settled energy efficiency measures in EU27 (Mantau 2010). EU27 consist of 27 EU member states 1 January 2007 - 30 June 2013 (Glossary: EU enlargements. 2014). In 2030 wood energy consumption is estimated to be more than 700 M m³ (Figure 1).

Figure 1. Current and future amounts of wood energy consumption in EU27 (Mantau 2010).

Charcoal and torrefied wood could be renewable energy sources substituting coal and other fossil fuels by CO 2 neutral fuel, products which are non-fossil, biodegradable, organic and originate e.g. from plants. Released carbon dioxide emissions of CO 2 neutral fuels like wood can be partly offset by regrowth of the biomass (Lehtonen 2005, Schlamadinger & Marland 1996.) Due to low heating value of wood compared to fossil fuels it needs to be processed to be more competitive in a long transportation distances. Heating value is the amount of heat released during the combustion of a specified amount of combusted material. Forest residues and harvesting of young stands could be capitalized on energy processing. By carbonizing and torrefaction transportation costs could be declined (Shah et al. 2012). Increasing the forest residue collection accelerated procurement is demanded. Efforts in equipment development have been made in felling machines e.g. in small tree operations which have been challenging. (Hakkila 2006.)

In Finland wood-derived energy is an important source of energy. The growth of Finnish forest is approximately 35 million m 3/year more than natural drain and fellings. Even the use of wood fuels accounted approximately 24 % in 2012 of the annual primary energy supply in Finland, the total use of forest chips is low accounting 4 % of the total energy consumption (Ylitalo 2013, Ylitalo & Ihalainen 2013.) Logging residues from regeneration fellings were the major source of forest fuel accounting 8,7 mill. m³ of the total solid wood fuel use (18,7 mill. m³) in 2013 (Torvelainen et al. 2014). The rest of the forest fuel is from precommercial thinnings or unproductive hardwood stands, which could be utilized in the future if the procurement chain will be cost-effective (Ylitalo 2001, Torvelainen et al. 2014). For cost-effective procurement chain major investments are needed, because the competitiveness of wood-based fuels is not a sufficient level (Kärhä et al. 2010). There is potential for logging residue collection in regeneration felling areas especially in Central and Eastern Finland (Ranta 2005).

1.2 Background and purpose of the study

The thesis is a part of the Balbic project, which aim was to develop industrial biocoal markets and production in Southern Finland, parts of Estonia, Latvia and Sweden, understanding the existing raw material and its potential for CO 2 neutral fuel use through market and industrial aspects. Biocoal is manufactured through pyrolysis process from wood or other biomass. Pyrolysis is a form of heat treatment where biomass is thermally degraded in the absence of oxygen. The project partners were Forestry Development Centre Tapio and Latvian State Forest Research Institute “Silava” and the head of the project was University of Helsinki Department of Forest Sciences. The project was financed partly by European Union through Central Baltic Interreg IV A Programme 2007 – 2013 and Regional Council of Southwest Finland. The project aimed to reach the target groups such as forest industry, forest owners, research institutes bringing more information about economical transportation, international commercial good, seeking extensive and exact information about market values of biocoal industrial production (BalBic 2011.)

The aim of the study was to determine how pyrolysis temperature (from now on temperature in this study) affects to the wood fuel characteristics such as heating value (in this study higher heating value if not stated otherwise), mass yield and carbon content. Higher heating value (HHV) is one of the values which describe the amount of heat released during the pyrolysis, where mass yield describes the yield of torrefied and carbonized products. During carbonizing and torrefaction occur chemical endothermic and exothermic reactions. Endothermic reaction absorbs energy from the system, whereas exothermic reaction releases energy to the system in a form of heat. Energy released from the exothermic reaction could be utilized for initiation energy in torrefaction. The aim of this study was also to discover, where the reaction changes from the endothermic to exothermic reaction. These have not been discovered earlier with pine ( Pinus sylvestris) and birch (Betula pubescens) in pyrolysis temperatures 250–450 °C. Differences between these two species were also under examination. These two Northern coniferous forest belt tree species are significant in wood refinement industry in Scandinavia and Baltic Countries.

2. Literary study

2.1 Chemical composition of wood

Wood consists of chemical compounds cellulose, hemicelluloses, lignin and extractives, which are mainly carbon (C), oxygen (O), hydrogen (H) and sulphur (S).

(Panshin & De Zeeuw 1980.) The main chemical formula of cellulose is (C 6H10 O5)n (Li et al. 2009). Hemicelluloses are polysaccharides and include xyloglucans, xylans, mannans and glucomannans (Scheller & Ulvskov 2010). The structure of lignin is complex and heterogeneous and varies by species of plant (Vázquez et al. 1997). Lignin from aspen consists of 63,4 % carbon, 5,9 % hydrogen and 30 % oxygen whereas composition of cellulose is 42,11 C/%, 6,48 H/% and 51,42 O/%. (King et al. 1983, Properties viewer, cellulose 2014).

The chemical composition of wood varies by part of the tree and type of the wood and effects to the combustion reaction. Chemical composition is different in bark, roots and branches than in a stemwood also chemical composition varies in tension and compression wood, compared to normal wood. (Pettersen 1984.) On average wood consist 51–53 % of carbon, 40 % of oxygen, 5 % of hydrogen, 1 % of sulphur and nitrogen and approximately 2 % of ash. Ash contains non-organic compounds like Si, Ca, Al, Mg, Fe, Na, K, S, P. (Nikitin 1966.)

Figure 2. On the left side is cut-away drawing of wood cell wall structure (Wiedenhoeft 2010) and on the right side is three major components of wood cell wall (Murphy & McCarthy 2005).

Figure 3. The share of wood components in a three main region of wood cell structure (KnowPulp 2003)

Wood is organic matter which consists of cells. The cell wall structure of wood consist of middle lamella (ML), primary wall (P) and secondary wall (S1, S2 and S3) seen in Figure 2. Each layer (Figure 3) has three major components; cellulose, hemicellulose and lignin, but the share of each component varies in each section. (Panshin & De Zeeuw 1980.)

The relative amount of cellulose is highest in secondary wall (S2) and the amount of lignin is highest in middle lamella (Figure 3). The cell wall structures are different in each part of the wood e.g. heterogeneous bark and the share of wood components differ in each section of wood (Howard 1971.)

Needles, leaves and bark have higher extractive proportion and roots, stemwood and branches have relative high share of lignin (Table 1). Because of the share of lignin and extractives, heating value in those parts of wood is higher. (Nikitin 1966, Kärkkäinen 2007, Telmo & Lousada 2011). Higher heating value of pine stump is 22,4 MJ/kg, which is 2,8 MJ/kg greater than in stemwood, which might be reflecting that extractive proportion increases the heating value due to higher share of extractives in stump than stemwood (Nurmi 1993, Nurmi 1997). In Table 1 are presented averages of the proportions of the main chemical compounds (%) and their median absolute deviation is represented in parenthesis if more than two values were obtained and founded from the literature (Räisänen and Athanassiadis 2013).

Table 1. The main chemical composition (%) of Scots pine and Silver birch. The figures are collected from different sources. (Räisänen and Athanassiadis 2013).

Scots pine Cellulose Hemicelluloses Lignin Extractives Stemwood 40,7 (0,7) 26,9 (0,6) 27,0 (0,0) 5,0 (1,0) Bark 22,2 (3,2) 8,1 (0,4) 13,1 (5,4) 25,2 (5,2) Branches 32 32,00 21,5 (5,9) 16,6 (7,1) Needles 29,1 24,90 6,9 (0,8) 39,6 (1,3) Stump 36,4 28,20 19,50 18,7 Roots 28,60 18,90 29,80 13,30 Silver/Downy birch Cellulose Hemicelluloses Lignin Extractives Stemwood 43,9 (2,7) 28,9 (3,7) 20,2 (0,8) 3,8 (1,3) Bark 10,7 (0,3) 11,2 (0,5) 14,7 (3,9) 25,6 (1,1) Branches 33,3 23,40 20,8 (3,9) 13,5 (3,0) Leaves N/A N/A 11,1 (0,0) 33,0 (0,0) Stump 29,50 19,4 13,4 4,7 Roots 26 17,1 27,1 13,5

Chemical composition 100% 3,5 % 3,0 % Total extractives 90% 22,0 % 80% 27,7 % Lignin 70% 2,6 % 3,6 % Hemicelluloses, Other 60% 8,9 % polysaccharides 27,5 % Hemicelluloses, Glucuronoxylan 50% 16,0 % 40% 2,3 % Hemicelluloses, Glucomannan 30% Cellulose 20% 40,0 % 41,0 % 10% 0% Scots pine(Pinus sylvestris) Silver birch (Betula verrucosa)

Figure 4. Chemical composition of Scots pine and Silver birch dry basis (Sjöström 1993).

Observing Scots pine (Pinus sylvetris) and silver birch (Betula verrucosa) the most evident difference is in hemicelluloses composition (Figure 4). Hemicellulose consists of glucomannan, glucuronoxylan and other polysaccharides. The glucomannan composition of Scots pine is higher (16,0 %) than Silver birch (2,3 %), whereas birch contains more glucuronoxylan (27,5 %) than pine (8,9 %). Total share of silver birch hemicelluloses composition is 32,4 % where pine’s share is 28,5 %. Difference between lignin composition is 5,7 percentage units. (Nikitin 1966, Sjöström 1993.)

2.2 Pyrolysis and process conditions

Pyrolysis is thermal decomposition of biomass in the absence of oxygen. Depending on the process conditions the main products of pyrolysis are in a form of liquid (e.g. tar), solid (charcoal and torrefied wood) and gas. During the pyrolysis heat transfers from heat source to biomass resulting increase in temperature of biomass and release of volatiles. Some of the volatiles condensate producing liquids e.g. tar. Pyrolysis can divide into conventional or slow pyrolysis, fast and flash pyrolysis. Slow pyrolysis is carbonization or torrefaction of biomass whereas fast and flash produce liquids and gaseous products. (Babu 2008.) In this study pyrolysis is slow pyrolysis if not stated otherwise.

2.2.1 Carbonization

Carbonization and torrefaction differs basically only by the temperature. Torrefaction is thermochemical conversion in lower temperatures, from 220 to 300 ˚ C and carbonization starting from 400 ˚ C in the absence of oxygen. Products of the carbonization and torrefaction are in a form of solid, liquid and gas. Produced solid matter is charcoal or torrefied wood, liquid which contains pyrolysis oil and tar. Gas includes e.g. carbon dioxide and carbon monoxide. (Prins et al. 2006, Neves et al. 2011.)

During the carbonization process wood compositions of three main polymeric structure (hemicelluloses, cellulose and lignin) degrades differently according to the pyrolysis temperature (Prins et al. 2006, Babu 2008). Carbonization process can be divided into four phases according to temperature scheme. (Antal et al. 1996, Hakkila 2006, Babu 2008).

The first phase occurs in the temperature of 100–170 °C, where main process is the evaporation of free and bound water from the cells. Evaporation occurs above the boiling point of water (100 °C), at atmospheric pressure (Brito et al. 2008.) Among the water steam other vapours like terpenes evaporates. After completed water evaporation temperature increases more quickly to 170 °C. (Talvitie 1944.)

During the second phase, which takes place in temperature of 170–270 °C, most of the dry matter of wood distils, which can be seen as a mass loss. Hemicelluloses decompose in the temperature range of 225–325 °C (Antal & Gronli 2003, Prins et al. 2006, Almeida et al. 2010.) Wood begins to decompose in the temperature of 170 °C. Decomposition temperature of lignin is 250–550 °C, but minor decomposition can be observed already in lower temperatures (Talvitie 1944, Bergman et al. 2005.) Decomposition of lignin results mainly formation of carbon oxidises in this phase, carbon dioxide (CO 2) and carbon monoxide (CO) (Nikitin 1966). In this phase also chemically degraded water, which is developed from the chemical decomposition of wood, is released in a form of steam. Wood acid is composition of methanol, acetone and acetic acid. Part of the wood acid evaporates as a steam during the second phase. Evaporation of terpenes also continues. Compound of terpenes is called turpentine,

15 which is recovered from coniferous wood. Methanol, acetone, acetic acid and turpentine can be salvaged by cooling the vapour gases. (Talvitie 1944, Emrich 1985.)

These first two phases are mainly endothermic reactions, where the external heating is needed to maintaining the process and the energy is bind in the process. Feedstock is heated using external heat source e.g. oil burner. Exothermic reaction begins at the temperature of 270–280 °C. External heating is no longer needed, because exothermic reaction generates heat. Only activation energy is needed to start-up the process. Energy is released in exothermic reaction after the process is activated. (Antal and Gronli 2003, Alakangas 2011.)

During the third phase (350–400 °C) decomposition of wood is exothermic (Emrich 1985). Cellulose decomposes in the temperature of 305–375 °C (Prins et al. 2006, Almeida et al. 2010). The share of hydrocarbons in volatiles increases, especially the proportion of methane and the share of carbon monoxide decreases. Also hydrogen starts to develop. Methanol, turpentine and acetic acid distil, but the quality of turpentine degreases among the increment of temperature. Tar is produced substantial amounts depending on the lignin and cellulose decomposition in this phase. Tar distils to phase of liquid and gas. The heating value of the gases increases enormously. The form of the coniferous wood will be gained resin, which is the most valuable ingredient of tar. (Talvitie 1944.)

In the fourth phase distil only tar and gas during the process, because wood contains only lignin and cellulose in this phase. Volatiles contain mostly hydrocarbons and hydrogen. Temperature is from 450 °C to 500 °C or 600 °C depending on the used technology. The heating value of the gas is highest in the fourth phase when comparing all above-mentioned phases. (Talvitie 1944.)

2.2.2 Torrefaction

Torrefaction is early phase of the carbonization or thermal treatment in temperatures of 220–300 °C in holding time one hour or less at atmospheric pressure. Method is the same than in carbonization, but the pyrolysis temperature is lower. Synonyms for torrefaction are roasting, slow- and mild-pyrolysis. (Bergman et al. 2005, Ciolkosz and Wallace 2011.) Torrefaction can be divided into three classes according to the temperature light (220 °C), mild (250 °C) and severe torrefaction (280 °C) (Chen et al. 2011). In the torrefaction process hemicelluloses are the most reactive compounds. Lignin and cellulose decomposes in higher temperatures, but the reactions of polymers are overlapping. (Prins et al. 2006.) Torrefied wood is promising as a renewable fuel for industrial use in coal co-combustion, gasification-combustion and Fisher-Tropsch fuel production or in the small-scale use as barbeque fuel or firelighter. The advantages of torrefaction are improved higher heating value, grindability and possible hydrophobicity. Technologies used in torrefaction are the same as carbonization. (Bergman et al. 2005, Bridgeman et al. 2010, Ciolkosz and Wallace 2011.)

2.2.3 Temperature, holding time, heating rate and pressure

The yield of solid char, gases and liquids depend strongly on the process conditions; pyrolysis temperature, holding time, heating rate and raw material (Prins et al. 2006, Babu 2008). Peak temperature is the highest temperature reached in the carbonization or torrefaction process (Antal et al. 1996). High pyrolysis temperatures decline solid yield of char, but increase higher heating value, carbon and ash content. (Pach et al. 2002, Zanzi et al. 2002, Ferro et al. 2004, Almeida et al. 2010, Cuña et al. 2010)

Figure 5. Yields of charcoal (Y ch ,F), pyrolytic gas (Y G,F), pyrolytic

liquids (Yt ar ,F+Y H2O ,), and pyrolytic water (Y H2O ) as a function of pyrolysis temperature, dry and ash free basis (daf). Open dot is results of slow heating rate and solid dot of fast heating rates (Neves et al. 2011).

The yield of charcoal declines and the yield of gases increase in a function of the pyrolysis temperature (Figure 5). The yield of tar and pyrolysis water increases until 500 °C in both fast and slow pyrolysis. The dispersion of the collected data varies due to different pyrolysis conditions (e.g. holding time, heating rate, temperature, feedstock and reactor) and raw material (particle size and type of raw material). The conclusion between slow and fast pyrolysis can be drawn, in slow pyrolysis solid material is produced more than fast pyrolysis and vice versa with gases. The heating rates classified in the Figure 5 to slow and fast heating rates are questionable, due to literature’s partial heterogeneous and ambiguous information. (Neves et al. 2011.) The definition of slow and fast pyrolysis is also arbitrary (Mohan et al. 2006). In general, the fast pyrolysis is a process where organic substance is rapidly (10–200 °C/s) heated to 580–980 °C and in slow pyrolysis heating rate is 0,1–1 °C/s (pyrolysis temperature 280–680 °C). The holding time in slow pyrolysis is longer than fast pyrolysis (Babu

2008.) The product distribution is collected from wood e.g. birch, pine and beech, but also other organic material like bamboo, rice husk, rape seeds and sawdust. (Neves et al. 2011.)

Holding time has an impact on solid char yield, higher heating value and carbon content (Pach et al. 2002, Almeida et al. 2010, Antal & Gronli 2003). Increase in holding time decrease the yield of solid char. According to Pach et al. (2002) yield of solid char observed to be approximately 13 % higher in holding time of one hour than three hours in temperature of 280 °C. Holding time has a slight positive impact on higher heating value and carbon content. (Pach et al. 2002, Antal & Gronli 2003.)

Heating rate affects to the yield and composition of charcoal in addition to wood species. Rapid decomposition gives a lower yield of charcoal but greater yield of gases. (Alakangas 2011.) Effects of prevailing pressure in the process to decomposition of wood have been studied. The studies have shown that pressures above atmospheric pressure effects to exothermic reaction and reaction becomes very violent producing a slight increase in charcoal and gas volume. (Alakangas 2011, Antal et al. 1990.)

2.3 Properties of charcoal and torrefied wood as fuel

Solid pyrolysis products are combusted for energy generation (Babu 2008). Combustion is a chemical reaction, between fuel and an antioxidant. Basic combustion reaction is reaction between carbon and oxygen. During combustion energy is released in a form of heat and it can be salvaged and used in energy production. Toxic carbon monoxide (CO) is formed in incomplete burning where complete burning forms carbon dioxide (CO 2) which is one of the greenhouse gases. (Curkeet 2011.)

Analysing the properties of wood as renewable energy price, volume, availability and technical properties are under examination. Most important technical issues are the higher heating value, volume, carbon content and usability. Co-combustion of torrefied wood with coal has environmental and technical advantages. Torrefaction weakens the fibre structure of the biomass, improves the grindability and enables the biocoal to be processed with the coal at the power plant without extra efforts made in

19 the traditional coal-fired power plant (Beekes and Cremers 2012, Ohliger et al. 2013).

Environmental benefits of the biocoal are also known, as a CO 2 neutral fuel (Hakkila 2006). Profitably of wood as a fuel source can be promote via torrefaction or carbonization. The moisture content of green wood is relative high depending on the ambient humidity and the higher heating value relative low. (Bergman 2010, Boundy et al. 2011). Torrefaction reduce moisture content and transportation costs (Shah et al. 2012). Torrefied wood is claimed to be hydrophobic (Bourgeois & Doat 1984, Felfli et. al. 2005). Hence the storing among coal in outdoors could decline the expenses.

2.3.1 Heating value

There are multiple ways to define the heating value of wood. Generally heating value is the amount of energy stored in a system per mass or volume unit (kJ/kg, kJ/mol, kcal/kg, Btu/lb) (Demirbas 2005). In this study the amount of energy is observed per mass unit (MJ/kg) and defined as higher heating value (HHV).

The higher heating value or gross calorific value (GCV) or higher calorific value (HCV) is the quantity of energy released when one unit of fuel is completely combusted and the water vapour produced is condensed into liquid form in the temperature of 25 °C. The heat of vaporisation is recovered. Lower heating value (LHV) or net calorific value (NCV) or lower calorific value (LCV) is the quantity of heat liberated by the complete combustion of one unit of fuel when water vapour produced remains as a vapour and the heat of vaporisation is not recovered. (Kärkkäinen 2007.) Moisture and hydrogen content affects the difference between higher and lower heating values. The higher is the moisture and hydrogen content, the more obvious is the difference between those heating values. The heating value can be calculated in three different forms fuel as received, dry (db) or dry and ash free basis (daf). (Demirbas 2005.)

Heating value depends on the chemical composition of wood and the temperature. The fewer elements contain oxygen, the higher calorific value is. Carbon and hydrogen content classifies the quality of fuel. Heating value differs in different parts of wood due to different chemical composition. Lignin has higher heating value (25,5 MJ/kg) than cellulose (17,4–18,2 MJ/kg). Bark contains more lignin and fats, because of that

20 the heating value of bark is higher than in sap- and heartwood. Heating value is also higher in needles and stumps, because of the high extractive composition. (Kärkkäinen 2007.) Spruce bark contains more carbon and hydrogen and less oxygen, due to that higher heating value is higher in bark than in stemwood (Demirbas 2005).

Higher heating value increases with the increase in temperature. Higher heating value of untreated stemwood varies from 18,3 MJ/kg to 22,4 MJ/kg with pine, spruce, birch and aspen (Kärkkäinen 2007). HHV of pine chips increases from 19,5 MJ/kg to 25,4 MJ/kg and logging residue chips from 19,8 MJ/kg to 26,4 MJ/kg within the torrefaction temperature of 225–300 °C. (Manunya and Sudhagar 2011).

30 Logistic regression equation Measured heating value

Higher heating MJ/kg value, Higher 29

28 200 400 600 800 1000

o Temperature, C

Figure 6. Higher heating value of spruce in a function of temperature. Drawn after data from Demirbas (2001b).

Temperature does not affect to the higher heating values significantly in temperatures above 600 °C, and it remains in the level of 31,4–32 MJ/kg with spruce. Increase in higher heating value is most dramatic in temperatures of 400–600 °C. When increasing the temperature from 300 °C to 900 °C, the higher heating value increases by 10 % from 28,8 MJ/kg. (Figure 6.)

2.3.2 Mass yield

Mass yield or solid char yield degreases with the increase in pyrolysis temperature (Pach et al. 2002). Mass yield is one of the main parameters when evaluating carbonization process. Mass loss occurs mainly during the endothermic reaction due to hemicellulose decomposition. In exothermic reaction mass loss is negligible. (Park et al. 2010.) Also the holding time has an impact to the mass yield. According to Pach et al. (2002) elongated holding time in lower torrefaction temperatures already decrease to the yield of solid char.

Solid yield of torrefied pine wood chips decreased from 89 % to 52 % in temperature range of 225–300 °C. The most dramatic decomposition of wood occur temperatures above 275°C. (Manunya and Sudhagar 2011). The solid yield of soft- and hardwoods decrease from 35,6 % to 22,7 % in temperatures of 280–880 °C. (Demirbas 2001a).

1,0

0,8

0,6

0,4 Relative Relative mass, %

0,2

0,0 0 200 400 600 800

o Temperature, C

Figure 7. Mass loss of spruce in function of temperature (Havimo 2010).

Thermogravimetric analyser (TGA) is used to measure weight changes in a material as a function of temperature or time under a controlled athmosphere (PerkinElmer, Inc. 2010). Thermogravimetric analysis of spruce shows that the mass loss is most dramatic in the temperature of 200–300 °C. At 300 °C mass yield is approximately 30 % and at 400 °C mass yield is lover than 10 %. (Figure 7.)

2.3.3 Carbon and ash content

High carbon content of fuels is valuable in combustion reaction and energy generation. Carbon content increases with increase in pyrolysis temperature. Carbon content of the wood varies according to the wood species. Untreated wood contains approximately 45–50 % carbon in dry basis (Lamlom and Savidge 2003.) Torrefied wood contains 50–60 % carbon depending on the wood species and torrefaction temperature (Table 2, Phanphanich and Mani 2011). When comparing untreated birch and pine, pine has slightly greater carbon content than birch. Carbon content of charcoal is around 80 % in temperature of 400 °C (Demirbas 2001a). The carbon content increases reaching almost 95 % carbon content in temperature of 1 000 °C (Table 2.) In Figure 10 most dramatic increase in carbon content and decrease in oxygen content can be seen in temperature of 300–500 °C. Lower oxygen content indicates more valuable charcoal (Antal and Gronli 2003). Torrefaction increases the carbon content of wood by 2 to 24 % in temperatures 230–280 °C. Carbonization increases the carbon content in higher temperatures (> 300 °C) by 42–89 %. In temperature of 600 °C and 650 °C the carbon content is 88–89 %. (Table 2.)

Table 2. Properties of torrefied wood and charcoal. Values are collected from sources:

1 Pach, et al. (2002), 2 Zanzi et al. (2002), 3 Ferro et al. (2004), 4 Cuña Suárez et al. (2010)

Figure 8. Effect of temperature to beech char mass fraction %. Solid line represent heating rate of 2 °C/min and dashed line 10 °C/min (Antal and Gronli 2003).

From 200 to 300 °C carbon content of beech char is approximately 50 %. The most dramatic increase in carbon content is between 300–400 °C. At 400 °C carbon content reached 70 % mass fraction and increased to 90 % at 700 °C. Temperature decrease oxygen and hydrogen content. Carbon content increases when heating rate decreases from 10 °C/min to 2 °C/min (Figure 8.)

Ash content is non-wanted feature, because it reduces the heating value of charcoal and torrefied wood due to the lower amount of combustion material. Ash content increases slightly with increase in pyrolysis temperature. The share of ash content in the dry mass of wood differs according to the wood species and part of the wood. Ash content is greater in leafs, bark and branches than in stemwood, because of the greater composition of inorganic material. (Kärkkäinen 2007.) Ash content of untreated softwoods is 0,5 % and hardwoods 0,4 % (Demirbas 2001a). Ash content of torrefied pine is 0,24 – 0,43 % and torrefied birch lower than 0,75 % (Pach et al. 2002, Ferro et al. 2004, Manunya and Sudhagar 2011, Table 2). Torrefied logging residues have slightly greater ash content, but the content is lower than 2,3 % (Manunya and Sudhagar 2011). In temperature of 800 °C ash content remains at 4 % . (Table 2.) In comparison to ash content of bituminous coal (6,5 %) the share of ash in wood is

25 relatively small (Manunya and Sudhagar 2011). Due to minor share of ash in torrefied wood and charcoal ash content was not observed more closely in this study.

2.3.4 Endo- and exothermic reaction

Endo- and exothermic reactions are formed in thermodynamic systems. Four laws of thermodynamics characterize thermodynamic systems and define fundamental physical quantities like temperature and energy. First law of thermodynamics states that the total amount of energy remains constant, but the form of the energy can be changed from one to another. That’s the principal in endothermic and exothermic reactions. The enthalpy change ( H) is used for calculation of total energy of a thermodynamic system. Enthalpy change is energy used in bond breaking reactions minus energy released in bond making products. (Cengel and Boles 2007.) Enthalpy changes of endothermic reactions are positive because energy is used in chemical bond breaking reactions. Endothermic reaction absorbs energy from surroundings and heat flows into the system from surroundings. If the enthalpy change is negative reaction is exothermic and chemical bonds are made when heat generated in a reaction flows from the system out to the surroundings. Exothermic reaction releases energy. (Chang 2006.) Released energy from exothermic reaction could be used for initial heating in carbonization or torrefaction.

Wood pyrolysis is complex continuum of endothermic and exothermic reactions (Park et al. 2010). Endo- and exothermic reactions of wood major constituents (cellulose, lignin and hemicellulose) occur in different temperature. Peaks of exothermic reaction can occur in different temperatures depending on the chemical composition of wood. (Strezov et al. 2003.) Endothermic decomposition of wood is followed exothermic reaction. Ferro et al. (2004) found that the exothermic reaction on goes at temperatures between 180 ˚C and 270 ˚C (raw material pine, lucern, sugar cane, bagasse, wood and straw pellets) where as Ciuta et al. (2014) reported exothermic behaviour above 305 ˚C. Strezov et al. (2007) found that thermal decomposition of biomass was exothermic at temperatures above 230 °C. Single exothermic peaks are detected at 320, 350, 360 and 413 ˚C (Ciuta et al. 2014, Strezov et al. (2003).

Explanations of exothermic behaviour differs significantly. Exothermic reaction has been explained to be result of secondary tar cracking reactions, lignin or dehydrocellulose decomposition (Bilbao et al. 1996, Di Blasi et al. 2001, Milosavljevic & Suuberg 1995, Milosavljevic et al. 1996, Strezov et al. 2003). According to Park et al. 2010 experimental and numerical tests secondary tar cracking and lignin decomposition have only small impact on the centre temperature peak and exothermic intermediate solid decomposition would be responsible.

2.4 Carbonization technologies

Carbonization technologies can be divided upon feeding system into batch and continuous process or upon heating method to internal, external or circulation gas.

Batch and continuous process and kiln types

Charcoal production technologies can be divided to continuous and so called batch- method techniques according to the operational base. In batch method feedstock is charged in the beginning of the reaction. After carbonization retort is discharged and charged again with raw wood. During the carbonization retort is not charged or discharged. Continuous kiln is charged and discharged continuously. (Ciolkosz and Wallace 2011.)

Figure 9. Traditional batch charcoal kilns (British hardwood charcoal 2009, Hall 2014, Warne 2009).

Kiln types can be separated into traditional and more advanced kiln types. Traditional kiln types are simple, easy to build and maintain. Typical for traditional kiln types is that kilns are grouped together to improve labour productivity (Figure 9). In traditional kiln types by-products are not recovered and exploited e.g. as energy for the process through indirect or direct heating of the wood. In batch types kilns controlling the emissions are difficult due to unstable emission conditions. (Trossero et al. 2008a, Trossero et al. 2008b, Brown 2009.)

Figure 10. More advanced continuous process charcoal kilns (Industrial charcoal making 1985, Summers 2010).

In more advanced kiln types, like Carbo Twin Retort, Waggon and Lambiotte Retort, emissions are controlled (Figure 10). There’s difference also in the oxygen contact between traditional and advanced kiln types. In more advanced kiln types or retorts oxygen contact with the biomass is prevented to ensure high yield of charcoal. In traditional kilns air intake is controlled, but not prevented. In more advanced kiln types at least combustion gases are reclaimed but also liquids, and the process is semi- continuous or continuous. Retorts are used in more advanced kiln types and among industrial charcoal manufacturing. (Trossero et al. 2008b.)

Internal, external heating and circulation gas

Carbonization or combustion technologies can divide also according to heating method. Heating can be internal, external or re-circulated. In the internal heating systems raw material is combusted under controlled air inlet. This is the most common and traditional method of charcoal production. Kilns operating by internal heating are traditional kilns e.g. earth mound, concrete and pit kilns but also portable, movable and fixed metal kilns (Figure 9). In external heating system retort is heated outside and there’s no air inlet into the retort. Biomass is carbonized in absence of oxygen. Examples of this method are metal retorts like Van Marion Retort. Heating with re- circulated heating gas, hot gas passes through the raw material charge. Charcoal and by-product yield is high, but the method is more expensive than above-mentioned. (Emrich 1985.)

3. Material and methods

3.1 Pyrolysis tests

Pyrolysis was accomplished by especially for this purpose manufactured carbonization pot using external heating method. The technology used was batch process (Figure 11.) Indicators, here thermocouples, were placed inside the pot in outer and inner surface of the pot for tracing and saving the variation of temperature. Also one of the indicators was placed in the exhaust pipe.

Figure 11. Carbonization pot, on the left is the drawing of the pot, and on the right is Figure of pot in the first trial run

Electricity was used for heat production and initiation energy of the process. Resistors were placed in outer zone of the pot. Heat was conducted to the inner layers of the pot. Liquids were collected, but not measured. Gases produced in the system were left out of the study. Gases could be used for power production and initiation energy of the process.

The pot was charged once in the beginning of the carbonization and discharged after carbonization which is characteristic for batch processes. Feedstock (M 1) of birch was 1,41–1,69 kg and pine 1,55–1,64 kg in each batch. External heat was kept on until the peak in inner temperature and decrease in inner temperature was detected therefore holding time was uncontrolled. After turning off the electric heating the pot cooled down unassisted. Torrefaction is exothermic reaction in the absence of oxygen. Nitrogen flow is used for keeping the reaction absence of oxygen (Pach et al. 2002, Ferro et. al. 2004). Pre-tests, where chips were carbonized with and without nitrogen flow in the same test device, showed that the nitrogen in this context does not affect to the solid char yield. Due to results of pre-tests the nitrogen flow was left out of the study. Heating rate was 20 ˚C/min. Slow heating rate was optimal for gaining high yield of char. In this study only solid char products were studied. Liquids and gaseous products were left out of the study.

3.2 Raw material

The Finnish forest land area is dominated by pine (67 %), spruce 22 % and birch 10 % (State of Finland's Forests 2012). The largest potential for increasing the forest chips for energy is in small trees in Finland. According to the different calculation methods the greatest potential for forest chips is in whole trees, delimbed stems and integrated harvesting with pulp wood of pine and birch. Potential of spruce for forest chips is in stumps and crown. (Anttila et. al. 2014, Asikainen et al. 2014.) In North Europe, Russia and Belarus birch is commercially most valuable broadleaf tree species and it can be grown in pure or mixed coniferous stands (Hynynen et al. 2010). Birch is light demanded tree species, for production of high quality timber pre-commercial thinnings are required. Whole trees from pre-commercial thinnings could be used for energy production (Cameron et al. 1995, Niemistö 1995). From these three dominant tree species birch and pine were taken into the study, because of the potential for forest chips, economic importance, and lack of the knowledge in earlier studies.

Raw material used is stemwood of birch ( Betula pubescens) and pine ( Pinus sylvestris) including bark. Stumps, leaves and branches were not included in to the study. Raw material was harvested from the Helsinki district (60°12'59.8"N and 25°01'25.8"E),

32 pruned and chipped with the tractor chipper at the site. The height of the harvested birches were approximately 15 m. Samples were pre-dried three to four days in temperature of 50–70 ˚C to prevent degradation of wood chips. After pre-drying the samples were packed in paper backs and stored in room temperature. Moisture content of wood chips was measured before pyrolysis.

The density of wood, which differs in wood species and part of the wood, affects to the heating value. Heating value increases with increase in wood density. Density has an impact on the thermal conductivity. Thermal conductivity increases with direct relation to the increment of density. Heating values of wood could be presented in volume units, which take into account the different densities of wood. (Kärkkäinen 2007.) Density of pine (410 kg/m³) differs from birch density (475 kg/m³) but the difference between these two tree species is not substantial (Repola 2006). For comparing the impact of density to the higher heating value other tree species e.g. spruce (385 kg/m3) or aspen (376 kg/m 3) would have been more applicable (Herajärvi & Junkkonen 2006, Repola 2006). The density calculations were left out of this study.

3.2.1 Chip size and distribution

The quality requirements and classes of solid biofuels are defined in European standards (EN 14961-1). Typical length of wood chips for solid biofuel is 5–50 mm, cut in a shape of sub-rectangular (Alakangas 2011).

The particle sizes are classified more closely to P-classes (P16, P45, P63 and P100) according to standard prEN 15149-1. The sizes of wood chips in main fraction, which is min 75 w% (weight percent), are in P16 3,15 ≥ 16 mm, P45 8 ≥ 45 mm, P63 8 ≥ 63 mm and P100 16 ≥ 100 mm. P-classes are classified to non-industry (A) and industry (B) use of wood chips. (Table 3.) Also the share of fines and coarse fraction, sulphur, nitrogen, chlorine, ash and moisture content are defined in the standard. Bulk density was recommended to be stated if chips are traded by volume basis. (Solid biofuels 2010, 2010, Alakangas 2011.)

Table 3. Specification of size for wood chips as fuel (Solid biofuels 2010)

Dimensions (mm) Main fraction Fines fraction, Coarse fraction, (w -%), max. length of (minimum 75 w- w-% (< 3,15 mm) particle, mm %), mm a

P16A c 3,15 16 mm and all < 31,5 mm P16B c 3,15 45 mm and all < 120 mm The cross sectional area of the oversized particle <1 cm² P31,5 8 45 mm, and all < 90 mm The cross sectional area of the oversized particles < 2 cm² < 6 % > 63 mm and maximum 3,5 % > P45A c 8 63 mm and maximum 3,5 % > P45B c 8 100 mm, all < 350 mm The cross sectional area of the oversized particles <10 cm² P100 c 16 200 mm, all < 350 mm The cross sectional area of the oversized particles <18 cm² a The numerical values (P-class) for dimension refer to the particle sizes (at least 75 w-%) passing through the mentioned round hole sieve size (ISO XXXXX). b For logging residue chips, which include thin particles like needles, leaves and branches, the main fraction for P45B is 3,15 < P < 45 mm, for P63 is 3,15 < P < 63 mm and for P100 is 3,15 < P < 100 mm and amount of fines (< 3,15 mm) may be maximum 25 w-%. c Property classes P16A, P16B, P31.5 and P45A are for non-industrial and property class P45B, P63 and P100 for industrial appliances. In industrial classes P45B, P63 and P100 the amount of fines may be stated from the following F04, F06, F08.

The chip size and fraction was determined from one kilogram sample of each raw material separately. The one kilogram sample was taken from 30 l paper sack where the chips were stored before torrefaction or carbonization. The sample was chosen by random because of representativeness and heterogeneity. The fraction of the chips was defined by batches. Each batch was shaken 20 seconds per sieve. Sieve distribution was >31,5 mm, 31,5 mm, 16 mm, 8 mm, 4 mm, 2mm, 1mm, 0,5 mm and < 0,5mm.

The mass of the samples were measured before and after the sieving and the particle size distribution was calculated.

Dominant sieve in every raw material was 16 mm, accounting 36–49 % of the sample and 89–92 % of the raw material particle size was greater than 8 mm (Appendix 1). Fraction of pine was P16A and birch was P45A. (Alakangas 2009, Alakangas 2010.) Particle size 16 ≤ 8 mm was used in this study for torrefaction and carbonization.

3.2.2 Determination of moisture content of pre-dried wood chips

Determination of moisture content is based on ISO 589 –standard method. Determination of moisture content were accomplished by weighting two pre-dried samples of each raw material before and after drying. Accuracy of weighting was 0,01 g and the weight of samples were 30 g before drying. Samples were dried 24 hours in Memmert oven in temperature of 105 ± 2˚C. The samples were weighted directly after drying, when the samples were hot. Moisture content (M ar ) is calculated using formula (1.) (Alakangas 2000).

(1.) = × 100

where Mar is moisture content (%)

m1 is mass before drying (g)

m2 is mass after drying (g)

Moisture content of pre-dried wood chips varied from 5,1–5,4 % (Table 4). Pine samples PB1 and PB2 exceeded the set time in the oven, due to that samples PB3 and PB4 were measured. Average moisture content (Average, %) of pine and birch samples are calculated from both 1-2 and 3-4 samples. In the Table 4 are presented also the weight of each aluminium foil dish (Dish, g), where the chips were placed during drying, mass before drying (m 1) and mass after drying with dish (m2 +dish). The weight of each dish was subtracted from the mass after drying with dish to gain only wood chip mass after drying (m 2) and moisture content (M ar ).

Table 4. Moisture content of pre-dried wood chips

3.2.3 Pyrolysis temperature

Pyrolysis temperatures were determined along the previous studies to be 250, 300, 350, 400 and 450 ˚C. One run was accomplished in each temperature with pine and birch comprising ten runs in total. The target was to examine especially torrefaction temperatures (250 and 300 ˚C), because of the non-existent studies and due to most dramatic mass loss in temperatures between 200 and 400 ˚C. Temperature scale 200– 400 ˚C wanted to explore more closely and one temperature above the scale wanted to take in study to confirm there’s no significant change after 400 ˚C. Due to the feature of carbonization device, the holding time could not be regulated. Heating rate was 20 °C/min.

3.3 Higher heating value

Higher heating value was determined among SFS-EN 14918 standard (Solid Biofuels, Determination of calorific value) unless otherwise stated. Test samples, sieve less than 1 mm, were closed in combustion bags (IKA Werke C12, higher heating value 46344 J/g, RSD 0,02 %). The bomb process was carried out with Parr Adiabatic Calorimeter GWB, Model nro 1241EA.

The test sample was placed with fuze in to the closed vessel in to a can filled with RO- water (reverse osmosis water), which is purified water using semipermeable membrane-technology. The temperature of RO-water was 24–26 °C. The can was placed in the Adiabatic Calorimeter and the bomb was activated. After launch results, variation of carbon (C) and heating value (MJ/kg) were entered. According to the SFS- EN 14918 standard after launched bomb crucibles need to place in the oven for ensuring complete burning of the samples and accurate test result of heating value. All samples were burned completely in the Adiabatic Calorimeter, on this account crucibles were not needed to place in oven in 105 °C and 560 °C for correcting the heating value.

Average of higher heating value was calculated from three samples in each temperature except from pine treated at 400 °C. Four samples (treated at 400 °C) was taken into the testing by accident. Heating value of combustion bags were calculated and deducted from HHV of samples. Benzoic acid was used for calibration of the calorimeter. (Appendix 2.1 and Appendix 2.2.)

3.4 Mass yield

Mass of each batch was measured before and after pyrolysis (torrefaction or carbonization). Dry and ash free (daf) mass yield (n M) or solid char yield is calculated as follows:

(2.) = where Mo is oven dry mass before pyrolysis (g)

Mt is mass after pyrolysis (g)

3.5 Carbon content

Carbon content, which is valuable in energy terms among higher heating value, were determined using Elementar vario MAX CN-device. From each temperature three samples were measured. In total 15 samples (sieve < 1 mm) were weighed before measuring in clean and tare crucibles. The range of weight was 0,127–0,199 g. Nitrogen content was automatically measured, but it was not meaningful to observe it in this project. Three samples of each temperature was measured and average values were calculated. Also reference samples of pine and birch, exceptionally without bark was measured.

3.6 Detecting exothermic reaction

Reactions that produce heat are exothermic processes in chemistry and physics. Heat flows from the system into the surroundings and an increase in temperature of surroundings are detected when exothermic reaction occurs (Cengel & Boles 2007.) Also after endothermic reaction the center temperature of biomass increases rapidly exceeding the surface temperature (Park et al. 2010). For detecting the change of temperature and exothermic reaction three indicators, in this context thermocouples were placed into the system. Thermocouples are used for measuring temperature differences between two points. Measuring requires that one of the junctions is so called cold junction, which is maintained at a controlled reference temperature and other ones senses the temperatures to be measured. (Pico Technology 2014.)

Two of the thermocouples were placed inside the test device among wood chips for measuring the temperatures inside the system and one of them was placed for measuring the surroundings of the system in an exhaustion pipe. Thermocouples which were placed inside the test device were in the middle of the pot called from now on inner and the other one in the outer surface (outer). The outer junction was placed approximately 3 centimeters from the side of the pot. Also cold junction was placed for compensation. Tracing the time and temperature were started when the new batch was sealed and the external heating was turned on. The temperature was recorded in one minute timescale from the junctions. Temperatures with timescale were automatically saved. External heating was turned off after detecting peak in inner temperature and observing the decline in inner temperature after peak. External heat turned off approximately 2 hours after detection of peak temperature in inner thermocouple.

4. Results

4.1 Higher heating value

Higher heating value of untreated birch was 17,3 MJ/kg (Figure 12). Torrefaction in 250 °C increased the HHV to 20,3 MJ/kg. In temperatures above 300 °C heating value were above 26,1 MJ/kg exceeding 28,2 MJ/kg at 450 °C. In temperature of 350 °C average heating value increased 0,4 MJ/kg from 300 °C (26,1 MJ/kg). At 400 °C average heating value was 27,3 MJ/kg. (Appendix 2.1, Figure 12.) Regression line describes phenomena significantly with R 2 = 0,83. (Figure 12.)

In Figure 13 are represented higher heating values of pine samples. The higher heating value of untreated pine was 18,2 MJ/kg (Figure 13). Torrefaction of pine increased the HHV to 20,2 MJ/kg and 26,2 MJ/kg in 250 °C and 300 °C respectively. In temperatures 350 °C and 400 °C the higher heating value was approximately 27 MJ/kg. In the extreme carbonization temperature (450 °C) the average of higher heating values was 27,6 MJ/kg. Variation was substantial at the temperature of 450 °C, where the measured HHV changed from 26,5 MJ/kg to 29,8 MJ/kg. R-squared of regression line was 0,94. (Appendix 2.2, Figure 13.)

Figure 12. Higher heating value of birch

Figure 13. Higher heating value of pine

4.2 Mass yield

Mass yield of torrefied and carbonized wood declines according to the increase in the temperature (Figure 14 & 15). Most dramatic mass loss occurs during the endothermic reaction, in temperatures 250–300 °C. Pyrolysis in 250 °C dropped the mass yield of birch from 100 % to 78 %. In 300 °C mass yield is 45 %. In temperatures 350 °C and 400 °C mass yield is 38 % and 36 % respectively. Mass yield of 31 % was measured in 450 °C. (Figure 14.)

Systematic decline of mass yield is evident from temperature 250 °C to 400 °C. Mass yields of pine in fixed temperatures (250, 300, 350 and 400 °C) were 86 %, 54 %, 38 % and 26 % respectively. Rising peak (mass yield 33 %) in temperature 450 °C is observable . (Figure 15.)

In lower temperatures (250 and 300 °C) mass yield of pine is slightly greater than birch (8 to 9 percentage units). When difference in higher temperatures is negligible, excluding exceptionally low mass yield of pine in 400 °C, which is 10 percentage units lower than mass yield of birch. Regression lines of mass yield of birch (R² = 0,86) and pine (R² = 0,91) approximates well real data points meaning relatively high dependency. (Figure 14 and 15.)

1,0

0,8

0,6

0,4 Mass yield ratio yield Mass

0,2 Mass yield Regression line (R² =0,86)

0,0 200 250 300 350 400 450 500

Temperature (° C)

Figure 14. Mass yield of torrefied and carbonized birch

1,0

0,8

0,6

0,4 Mass yield ratio yield Mass

0,2 Mass yield Regression line (R² =0,91)

0,0 200 250 300 350 400 450 500

Temperature (° C)

Figure 15. Mass yield of torrefied and carbonized pine

4.3 Carbon content

Untreated birch contains 47 % of carbon (Figure 16). Carbon content followed the pattern of higher heating value and increased according to temperature gaining 82 % share of carbon (average of three samples) at 450 °C (Figure 16). Most remarkable increase of carbon content is from 250 °C to 300 °C, when carbon content increased to 72 % at 300 °C. At 350 °C carbon content was 77 % and 79 % in 400 °C. (Figure 16.)

Untreated pine contained 48 % of carbon (Figure 17). The average carbon content of three samples at 250 °C was 54 %. At 300 °C treated pine gained 71 % share of carbon and in higher temperatures (350, 400 and 450 °C) increase was 5, 7 and 10 percentage units from 300 °C value (71 %) accordingly. Regression lines of carbon content (R² = 0,99) approximates well real data points meaning relatively high dependency. (Figure 16, Figure 17.)

Carboncontent (%) Regression line (R²=0,9924) 55 Sample 1 Sample 2 Sample 3 50

45 0 200 300 400 500

Treatment temperature (º C)

Figure 16. Carbon content (%) of birch

Carboncontent (%) Regression line (R²=0,9968) 55 Sample 1 Sample 2 50 Sample 3

45 0 200 300 400 500

Treatment temperature (º C)

Figure 17. Carbon content (%) of pine

4.4 Exothermic reaction

Exothermic reaction is detectable from the line charts when the inner temperature exceeds the outer temperature. External heating was kept on 7–9 hours in torrefaction temperatures. In higher temperatures external heating was on approximately 3–4 hours.

The inner temperature exceeded the outer temperature in time of 5 h 47 min (240 °C) with birch torrefied at 250 °C (Figure 18, Table 5). Maximum temperature (258 °C) detected from the inner thermocouple was reached after 6 h 56 min, when the difference between inner and outer temperature was 11 °C. After 9 hours external heating, which was set to be 250 °C, turned off and the pot cooled down unassisted. (Figure 18.)

In higher temperatures (350–450 ˚C) the peak of temperature it’s more obvious than in lower temperatures (250 ˚C and 300 ˚C) with both birch and pine (Figure 18, Figure 19, Appendix 3.1, Appendix 3.2). Inner temperature remained around 100 ˚C for one hour and 20 minutes after turning on the external heating for one hour and 20 minutes with pine. After two hours inner temperature increased above the outer temperature (456 ˚C). The maximum temperature 492 ˚C detected from the middle of the chips reached after two minutes after exceeding the outer temperature. (Figure 19.) All the other line charts of the birch and pine exothermic reactions are represented in Appendix 3.1 (birch) and Appendix 3.2 (pine).

300 Outer Inner 250 Exhaust pipe Cold junktion

200

150

100 Temperature (° C) Temperature(°

0 0:00 4:00 8:00 12:00

Time (hours)

Figure 18. Birch with bark torrefied at 250 ˚C

600 Outer Inner Exhaust pipe 500 Cold junktion

400

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00

Time (hours)

Figure 19. Pine with bark carbonized at 450 ˚C

In torrefaction temperatures (250 and 300 ˚C) the inception of exothermic reaction took approximately 4 h 20 min to 6 hours both pine and birch (Table 5). In 350–450 ˚C the starting point of the exothermic reaction took 2 to 2 hours 40 minutes. (Table 5.) The exothermic reaction could not detect at 450 ˚C with birch, because the junction placed in the middle of the pot bent and variation of inner and outer temperatures could not trace properly (Appendix 3.1). Inner temperatures exceeded the outer temperature near the appointed temperatures (Table 5).

The maximum inner temperature of birch was equivalent in temperatures 350 ˚C and 400 ˚C. The maximum of inner temperature increased from the appointed temperature by 112 ˚C with birch treated at 350 ˚C, which was the highest temperature difference observed between appointed and maximum temperatures. Highest temperature difference (67 ˚C) between inner and outer temperature was also with birch treated at 350 ˚C. In other temperatures inner temperature exceeded the outer temperature by 6– 42 ˚C. Temperature differences were higher with birch. With pine the increase of maximum temperature from the appointed was 63–65 ˚C in temperatures 300–400 ˚C. In 250 ˚C temperature the maximum inner temperature was 8 ˚C above appointed temperature with birch and with pine inner temperature remain 2 ˚C under the appointed temperature. In the extreme temperature of pine (450 ˚C) maximum inner temperature was 42 ˚C higher than appointed temperature. (Table 5.)

Table 5. Detecting the exothermic reaction, time (h:min) and temperature when inner temperature (T i) exceeded the outer temperature (T o), peak temperature of inner temperature (Peak T i) and maximum difference between inner and outer temperature (Max

Ti-To), where Temp. is appointed temperature of pyrolysis

BIRCH PINE Max Temp. Ti exceeded T o Peak Ti Max T i- Ti exceeded T o Peak Ti Ti-To (°C) (°C) To (°C) (°C) (h:min) Ti (°C) (h:min) Ti (°C) (°C) 250 5:47 240 258 16 6:05 241 248 6 300 4:24 278 379 43 4:17 299 363 20 350 2:41 348 462 67 2:34 327 413 39 400 2:05 396 460 42 2:32 432 465 23 450 no result 2:02 456 492 34

5. Discussion

5.1. Higher heating value

Chemical composition has an effect to the higher heating value. Resins, fats and lignin increase the HHV, because the higher heating values of those components are higher than celluloses. The water content affects only to the lower heating value, but not to the higher heating value, because HHV is calculated from the dry mass. (Kärkkäinen 2007.)

Comparing the higher heating values of pine and birch at appointed temperatures to values found from literature is intractable due to the lack of knowledge in those temperatures and tree species. From the literature was found measured higher heating values of pine and birch in low torrefaction temperatures, but in carbonization temperatures parallel values could not found. Higher heating values of pine and birch at carbonization temperatures are compared to Eucalyptus dunnii, soft- and hardwood values.

According to Pach et al. (2002) the higher heating value of birch at 250 °C is 18,8–20 MJ/kg, depending on the holding time. In the holding time of 3 hours the higher heating value was 20 MJ/kg, which is comparable to the values of this study (19,6– 20,9 MJ/kg) (Appendix 2.1). At 280 °C the HHV was 21,1 MJ/kg which is closest value found from the literature to birch treated at 300 °C (25,3–26,7 MJ/kg) (Pach et al. 2002, Appendix 2.1).

The higher heating value of pine treated at 250 °C was 18,4–19,4 MJ/kg according to Pach et al. (2002) and Ferro et. al. (2004) whereas results in this study were 19,3–20,9 MJ/kg. The HHV of pine treated at 280 °C was 19–22,7 MJ/kg, when at 300 °C HHV was 25,5–26,9 MJ/kg. The difference in higher heating values is considerable at 280 °C and 300 °C even those temperatures are not totally comparable. The difference in results depends strongly on the holding time which is 1–3 hours in Pach et al. (2002) and Ferro et. al. (2004) studies. The holding time in this study was undefined, due to

49 test device feature. Because of the undefined holding time samples were longer under pyrolysis conditions.

Higher heating values of hardwood and softwood treated in carbonization temperatures 300 °C and 400 °C were 2–2,5 MJ/kg higher than HHV of pine and birch (Demirbas 2001a). Higher heating value of Eucalyptus dunnii was measured at 300, 350 and 450 °C gaining the HHV of 29,2 MJ/kg, 26,7 MJ/kg and 30,7 MJ/kg respectively (Table 4). The values are approximate and cannot compare with values of birch or pine. The discrepancies in higher heating values of soft- and hardwood might be result of non-accurate information about the raw material used. The concept soft- and hardwood is extensive due to differences inside in both tree scales. The variation in the wood properties already inside the softwoods is wide, but because of the lack of knowledge the results of soft- and hardwood are suggestive. The higher heating values of E. dunnii than birch are explicable with higher cellulose 76,7 % and lignin content 27,9 % in sapwood (Cetinkol et al. 2012). The initial HHV of E. dunnii was 18,4 MJ/kg which was nearly the initial pine’s HHV. The differences also might depend on different process conditions.

Higher heating value (26 MJ/kg) increased most dramatically in temperature of 300 °C both pine and birch reaching higher heating value of bituminous coal. Higher heating value of bituminous coal is 24-33 MJ/kg (Coal Basics 2014). In temperatures 350 °C the average values increased only 0,4–0,6 MJ/kg from the values of 300 °C and 1,3–1,4 MJ/kg at 400 °C with pine and birch. Higher heating value of pine sawdust was 24,4 MJ/kg and 28,6 MJ/kg in fast pyrolysis in holding time of 15 min at 400 °C and 475 °C respectively (Azargohar et al. 2013). Measured HHV was lower than average of pine HHV in this study at 400 °C (27,4 MJ/kg) but higher at 475 °C (27,6 MJ/kg at 450 °C). From the initial 17,3 MJ/kg (birch) and 18,2 MJ/kg (pine) higher heating value increased 62,9 % and 52,1 % at 450 °C respectively. Maximum higher heating value of birch was 28,4 MJ/kg and pine 29,8 MJ/kg (Appendix 2.1, Appendix 2.2). Dispersion is noticeable with pine samples in 450 °C, which might be the result of non-automated sample preparation (Figure 13). The difference between pine samples at 450 °C was 3,2 MJ/kg. (Appendix 2.2).

There are slight differences between the average higher heating values of pine and birch. In relation to the untreated samples birch heating value in each temperature increased more than pine. The difference of the average higher heating value between pine and birch is 0,1–0,3 MJ/kg at 250–400 °C and 0,6 MJ/kg at 450 °C. The average higher heating value of pine is greater in temperatures 300–400 °C and the average HHV of birch at 250 °C and 450°C. Systematic difference in those tree species cannot be detected in this study. Differences between pine and birch samples might depend on the chemical composition of the wood or density.

The density of wood affects to the energy amount in volume unit (kWh/i-m3, kWh/p- m3). Because the density of birch on average (475–478 kg/m³) is greater than pine (410–435 kg/m³), higher heating value of birch is greater per volume unit (Repola 2006). The density of the samples were not measured. Density of wood influence directly to the thermal conductivity of wood. When the density increases also the thermal conductivity increases (Steinhagen 1977.) The share of lignin and extractives is greater with pine than birch, which would predict higher HHV of pine (Table 1). Pine has relatively higher lignin content than birch but also higher extractive share, that increase the HHV of wood (White 1987).

In conclusion, the results of higher heating value of this study were in torrefaction temperatures higher than earlier studies have shown. More reliable result would have gained if e.g. five samples of each temperature would have been examined. Also the density of these two species would have been valuable to measure for comparing these two tree species. Elemental analysis or chemical composition of these two tree species would have been valuable to measure for comparing result to other tree species. Comparison of tree species to other tree species would have been easier because of the lack of knowledge with pine and birch.

5.2. Mass yield

During thermochemical conversion biodegradable element loses mass according to the temperature. Wood consists mostly cellulose (40 %), hemicelluloses (25–35 %) and lignin (20–30 %) (Stenius 2000). Most dramatic mass loss occurs during the endothermic reaction, temperatures below 350 °C, correspond to cellulose and hemicelluloses decomposition. Ciuta et al. (2014) found the highest mass loss in temperature range of 350–400 °C. Hemicelluloses start to degrade in temperature of 230 °C, cellulose following at 260 °C and lignin above 300 ºC (Chen and Kuo 2010, Industrial charcoal making 1985). Small amount of lignin decomposition occurs already at 240 ºC (Shoulaifar et al. 2014). Higher mass yield of pine in lower temperatures (250 and 300 °C) is the result of pine’s lower hemicelluloses share. Mass yield begin to stabilize after 350 °C. Mass yield declined at 300 °C both pine and birch like earlier studies have indicated (Figure 14 and 15.)

Solid mass yield of birch was 78 % in 250 ºC, whereas Pach et al. (2002) estimated that solid mass yield was approximately 82–85 % at 250 ºC. According to Shoulaifar et al. (2014) mass yield of birch was 81 % at 240 ºC and 76 % at 255 ºC. Remarkable was that at 300 °C mass gain was 45 % and in 20 °C lower temperature (280 °C) the solid char yield was 65–69 % (Pach et al. 2002, Shoulaifar et al. 2014). Differences might depend on holding times. Holding times in Pach et al. (2002) study was 1, 2 and 3 hours and the holding time in Shoulaifar et al. (2004) was 35–45 min, whereas in this study the holding time was uncontrolled. External heating was kept on 9 hours in 250 °C, before the heating was turned off. Also in other temperatures external heating source was kept on a long period of time, decreasing the solid mass yield.

Solid mass yield of pine was 83–88 % at 250 ºC and 67–78 % at 280 ºC (Pach et al. 2002). Solid mass yield of torrefied pine in this study was 86 % in 250 ºC and 54 % in 300 ºC. Comparing the solid mass yield of pine treated at 300 ºC is mass yield remarkable lower than in Patch etc. 2002 results. The remarkable lower mass yield at 300 ºC than 280 ºC might be result of longer holding time. In earlier studies have reported sharp decrease in mass yield (32–40 %) with birch in temperature range 278– 430 ºC and with pine in 220–310 ºC. (Ciuta et al. 2014, Braadbaart & Poole 2008). The reportedly results are comparable with the results of birch and pine except the

52 result at 400 ºC there is a drop in mass yield (26 %). Braadbaart & Poole (2008) reported 32 % mass yield of pine at 400 ºC.

In lower temperatures the mass yield is greater with pine than birch (Figure 14 and 15). The share of hemiselluloses of Scots pine is lower (20 %) than birch (34,2 %), that might explain that in temperatures 250 °C and 300 °C mass yield of pine is greater than birch (Nikitin 1966).

5.3. Carbon content

Carbon content followed the pattern of higher heating value reaching 70 % share of carbon treated at 300 °C both pine and birch (Figure 16 and 17). Differences between these two species are negligible.

In 250 °C carbon content of birch was 54–55 % whereas Pach et al. (2002) estimated that carbon content of birch without bark is approximately 52–53 % and pine 51–52 %, where in this study carbon content of pine was few percentage units higher. In carbonization temperatures (300, 350, 400 and 450 °C) carbon content of beech was approximately 55 %, 68 %, 75 % and 78 % and carbon content of oak 70–78 % respectively (Antal & Gronli 2003, Braadbaart & Poole 2008). Carbon content of beech and oak remains under measured carbon content of birch and pine.

Slight difference in carbon content might depend on the difference in bark content in raw material. Bark of the wood contains both birch and pine more carbon than stemwood. The carbon content of stemwood of untreated pine is 51,2–51,8 % and bark is 53,2–53,5 %, whereas stemwood of birch is 49,7 % and bark of birch is 55,7 % (Laiho and Laine 1997, Tolunay 2009). From the above-ground biomass of Scots pine 70,2 % of the carbon is in stemwood and 7,8 % in bark, the rest of the carbon is in needles, dead and living branches (Tolunay 2009). Even the carbon content in bark is higher than in stemwood the share of bark in samples is relatively low. The relative share of bark in birch logs is 11,5 % and pine logs 12,2 % in South-Finland (Heiskanen & Riikonen 1976, Kellomäki & Salmi 1979). Because of the share of bark in samples

53 and carbon content of untreated samples 47–48 %, impact of bark to carbon content is minor.

Carbon content of untreated samples were 47–50 % with pine, birch, beech and oak (Pach et al. 2002, Ferro et al. 2004). Different process conditions affects to the carbon content. Increase in holding time increase carbon content (Wannapeera et al. 2011). Higher carbon content of the samples in this study is result of process conditions. Elongated holding time increased the carbon content of the samples. Holding time in this study was uncontrolled and compared to previous studies total time under torrefaction or carbonization conditions was longer. Longer time under process conditions explain slight higher carbon content.

5.4. Exothermic reaction

Reported temperature range of exothermic reaction and the difference between inner and surface temperature of solid matter varies widely due to different chemical composition of wood. According to earlier studies, starting point of exothermic reaction occur in temperature of 180–305 °C (Alakangas 2011, Ciuta et al. 2014, Ferro et al. 2004, Nikitin 1966, Strezov et al. 2007).

Exothermic behaviour existed in all torrefaction and carbonization cases, but the centre temperature peak in higher temperatures (300–450 °C) is more distinct than 250 °C. Even the cut of inner and outer temperature of birch torrefied at 450 ˚C could not detected, the exothermic reaction has occurred before the external heating was on two hours. There is an evident peak in inner temperature and exhaust pipe, which indicates the exothermic reaction. Endothermic reaction continues longer when temperature is lower, and reaction is more rapid with higher temperatures. At 250 °C increase of inner temperature is dilatory. Plateau during the endothermic reaction at 100 °C was not evident in all cases due to evaporation of free and bound water already in pre-drying of samples (Appendix 3.1 & 3.2, Figure 18 & 19).

With hardwoods like oak, beech and alder the increase of temperature in exothermic reaction was under 60 °C, but with chestnut and pine it was 83 °C and 136 °C

54 respectively. (Nikitin 1966). In temperatures 365–558 °C the surface temperature of maple exceeded the inner temperature by 10–70 °C (Park et al. 2010). Difference between center peak and surface temperature with birch was 25 ˚C or 50 ˚C depending on the particle size. Pyrolysis temperature was 290–410 ˚C. (Ciuta et al. 2014.) In this study difference between inner and outer temperatures was higher with birch (67 °C) and varied from 6 to 42 °C in other cases.

In lower pyrolysis temperatures exceeding the outer temperature took relatively long 4–6 hours. Comparing pine and birch in lower temperatures (250 °C and 300 °C) birch have slightly more explicit peak than pine. Inner temperature exceeded the outer surfaces temperature at 250 °C, but the increase of temperature was not violent, which is peculiar to exothermic reaction. (Figure 18, Appendix 3.1 & 3.2). The heating rate was 20 °C/min but due to particle size and process conditions heating of biomass was controlled by heat conduction.

The increase of inner temperature was extremely high with birch carbonized at 350 ˚C. Inner temperature was 112 ˚C higher than adjusted temperature. Elemental analysis or chemical composition of raw material would have been important to measure for explicating the phenomenon.

5.5. Analysis of sampling and profitability of pyrolysis

Torrefaction demands initiation energy for exothermic reaction. It is questionable, is it cost efficient to torrefied or carbonize chips before using it power production. If the transportation distances are long, then it might be efficient, for reducing the transportation costs. Untreated wood contains water more than torrefied or carbonized wood and increases the weight of the chips. Reducing the transportation costs requires torrefaction or carbonization near the cutting area. Transportation costs and torrefaction or carbonization costs should calculate. Also storage and conversion or utilization will benefit from torrefaction, but net benefit of torrefaction must be carefully accounted in each stage. Endothermic reaction took approximately 6 hours, before exothermic reaction occurred at 250 ˚C. External heating is needed for maintaining the endothermic reaction.

Three samples constricted the extrapolating the carbon content and heating value. Five samples of each temperature and tree species would have given more reliable knowledge of carbon content and heating value. But three samples of each temperature was sufficient for this study. The results of carbon content, heating value and can extrapolate but the exothermic reaction should study more. Inner temperature exceeded the outer temperature close to appointed temperature. Timing of exothermic behaviour was dependent on initial heating temperature. If elemental analysis should have been made, it would have been easier to analyse and compare raw material and results to other organic biomass. One batch in each pyrolysis temperature restricted the comparison of mass yield and exothermic reaction at 450 ˚C. The results gave more basic information properties of torrefied and carbonized birch and pine.

6. Conclusions

Carbonization and torrefaction improve the higher heating value and carbon content of wood. Carbon content of untreated birch and pine samples were 17 MJ/kg and 18 MJ/kg. Torrefaction at 250 °C already improved wood fuel properties (HHV 20 MJ/kg, C approx. 54 %) and the mass yield was relatively high (78 % and 86 %). At 300 °C considerable increase in carbon content (71-72 %) and heating value (26 MJ/kg) is detected and the mass yield was 45 % and 54 %. Torrefied wood (at 300 °C) reached HHV of coal. In higher than 300 °C temperatures, increase in carbon and higher heating value stabilize and mass loss is substantial.

Mass loss is the most dramatic during the endothermic reaction temperatures below 350 °C mainly due to completion of hemicellulose and cellulose decomposition. Mass yield declined to 30 % from the initial mass at 450 °C. Carbon content of birch and pine increased from the untreated 47 % carbon content to 82 % (carbonized at 450 ˚C). Uncontrolled holding time resulted loss in solid char yield.

Exothermic peak was detected in all temperatures (250–450 ˚C). In higher temperatures peak was evident. The initiation of exothermic reaction in torrefaction temperatures took several hours. Maintaining the endothermic reaction consume

56 energy, before exothermic reaction occurs. Exothermic reaction releases energy, but will via torrefaction or carbonization gain more energy than consume it. It’s questionable is torrefaction and carbonization cost- and energy-efficient. This is the reason why gases produced in the reaction should examine and exploit in torrefaction or carbonization process. Also the energy used in torrefaction, economically profitable transportation distances and hydrophobic would require more examination. Holding time affected to the mass yield, due to that holding time would be important to be regulated.

Variation between these two tree species in higher heating value, mass yield, carbon content and behaviour of exothermic reaction was negligible. Differences are slight but not systematic.

References

Alakangas, E. 2000. Suomessa käytettävien polttoaineiden ominaisuuksia. VTT Research notes 2045: 25–33. Available from: http://www.vtt.fi/inf/pdf/tiedotteet/2000/T2045.pdf.

– 2011. European standards for fuel specification and classes of solid biofuels. Solid Biofuels for Energy. Springer. p. 21–41.

Almeida, G., Brito, J.O. & Perre, P. 2010. Alterations in energy properties of eucalyptus wood and bark subjected to torrefaction: The potential of mass loss as a synthetic indicator. Bioresource technology 101: 9778–9784.

Antal, M.J., Croiset, E., Dai, X., DeAlmeida, C., Shu-LaiMok, W. & Norberg, N. 1996. High-yield biomass charcoal. Energy & Fuels 10: 652–658.

– & Gronli, M. 2003. The art, science, and technology of charcoal production. Industrial & Engineering Chemistry Research 42: 1619–1640.

– , Mok, W.S.L., Varhegyi, G. & Szekely, T.A.F. 1990. Review of methods for improving the yield of charcoal from biomass. Energy & Fuels 4(3).

Anttila, P., Nivala, M., Laitila, J., & Korhonen, K.T. 2014. Metsähakevarat ja metsähakkeen käyttö. In publication: Bioenergiaa metsistä: Tutkimus- ja kehittämisohjelman keskeiset tulokset. Asikainen, Ilvesniemi & Muhonen. Working Papers of Finnish Forest Research Institute 289: 13-20.

Asikainen, A. & Anttila, P. 2013. Riittääkö metsähake? Finnish Forest Research Institute. Metsätieteen aikakausikirja 4/2013: 685–688.

Azargohar, R., Jacobson, K., Powell, E. & Dalai, A. 2013. Evaluation of properties of fast pyrolysis products obtained, from Canadian waste biomass. Journal of Analytical and Applied Pyrolysis 104: 330-340.

Babu, B.V. 2008. Biomass pyrolysis: a state-of-the-art review. Biofuels Bioproducts & Biorefining-Biofpr 2: 393–414.

BalBic. 2011. [web-page] About BalBic. Available from: http://www.balbic.eu/. [cited 19.12.2011].

Braadbaart, F. & Poole, I. 2008. Morphological, chemical and physical changes during charcoalification of wood and its relevance to archaeological contexts. Journal of Archaeological Science 35(9): 2434-2445.

Beekes, M., & Cremers, M. 2012. [Web-page]. Torrefaction Cracks the Biomass Challenge. Available from: http://www.renewableenergyworld.com/rea/news/article/2012/06/torrefaction- cracks-the-biomass-challenge. [cited 5.8.2014].

Bergman, P., Boersma, A., Zwart, R. & Kiel, J. 2005. Torrefaction for biomass co- firing in existing coal-fired power stations. Energy Research Centre of the Netherlands. Report No. ECN-C-05-013.

Bergman, R. 2010. Drying and control of moisture content and dimensional changes. Wood handbook—wood as an engineering material. General Technical Report FPL– GTR–190: 13-1-13-20.

Bilbao, R., Mastral, J.F., Ceamanos, J. & Aldea M.E. 1996. Modelling of the pyrolysis of wet wood. Journal of analytical and applied pyrolysis 36: 81–97.

Boundy, R.G., Diegel, S.W., Wright, L.L. & Davis, S.C. 2011. Biomass energy data book. Edition 4. U.S. Department of Energy. 254 p.

Bourgeois, J. & Doat, J. 1984. Torrefied wood from temperate and tropical species. Advantages and prospects. Bioenergy 84. Proceedings of conference 15-21 June 1984, Goteborg, Sweden. Volume III. Biomass conversion 1984. p. 153-159.

Bridgeman, T.G., Jones, J.M., Williams, A. & Waldron, D.J. 2010. An investigation of the grindability of two torrefied energy crops. Fuel 89: 3911-3918.

British hardwood charcoal. 2009. [web-page]. Available from: http://www.countryside-jobs.com/Focus/Articles/1109_5.html. [cited 5.8.2014].

Brito, J.O., Silva, F.G., Leão, M.M. & Almeida, G. 2008. Chemical composition changes in eucalyptus and pinus woods submitted to heat treatment. Bioresource technology 99: 8545-8548.

Brown, R. 2009. Biochar production technology. In publication: Lehmann, J. & Joseph, S. Biochar for environmental management: science and technology. Earthscan. London, Sterling, Va. p. 127–139.

Cameron, A., Dunham, R. & Petty J. 1995. The effects of heavy thinning on stem quality and timber properties of silver birch (Betula pendula Roth). Forestry 68:275- 286.

Cengel, Y.A. & Boles, M.A.B. 2007. Thermodynamics: An Engineering Approach Sixth Edition (SI Units). McGraw Hill Higher Education. ISBN 10: 0071257713 / 0- 07-125771-3.

Cetinkol, O.P., Smith-Moritz, A.M., Cheng, G., Lao, J., George, A., Hong, K., Henry, R., Simmons, B.A., Heazlewood, J.L. & Holmes, B.M. 2012. Structural and Chemical Characterization of Hardwood from Tree Species with Applications as Bioenergy Feedstocks. Plos One.

Chang, R. 2006. General chemistry: the essential concepts. McGraw-Hill. Dubuque, Iowa. ISBN 0-07-111567-6.

Chen, W.H., Hsu, H.C., Lu, K.M., Lee, W.J. & Lin, T.C. 2011. Thermal pretreatment of wood (Lauan) block by torrefaction and its influence on the properties of the biomass. Energy 36: 3012–3021.

– & Kuo, P.C. 2010. A study on torrefaction of various biomass materials and its impact on lignocellulosic structure simulated by a thermogravimetry. Energy 35: 2580-2586.

Ciolkosz, D. & Wallace, R. 2011. A review of torrefaction for bioenergy feedstock production. Biofuels, Bioproducts and Biorefining 5: 317-329.

Ciuta, S., Patuzzi, F., Baratieri, M. & Castaldi, M.J. 2014. Biomass energy behavior study during pyrolysis process by intraparticle gas sampling. Journal of Analytical and Applied Pyrolysis 108: 316-322.

Coal Basics. 2014. [web-page]. Coal Marketing International Ltd. Available from: http://www.coalmarketinginfo.com/coal-basics/. [cited 5.8.2014].

Cuña Suárez, A., Tancredi, N., Pinheiro, P.C.C. & Yoshida, M.I. 2010. Thermal analysis of the combustion of charcoals from Eucalyptus dunnii obtained at different pyrolysis temperatures. Journal of Thermal Analysis and Calorimetry 100: 1051- 1054.

Curkeet, R. 2011. [Conference proceedings]. Wood Combustion Basics. Intertek. Available from: http://www.epa.gov/burnwise/workshop2011/WoodCombustion- Curkeet.pdf. [cited 7.2.2014].

Demirbas, A. 2001a. Biomass to charcoal, liquid, and gaseous products via carbonization process. Energy Sources 23: 579-587.

– 2001b. Carbonization ranking of selected biomass for charcoal, liquid and gaseous products. Energy Conversion and Management 42: 1229-1238.

– 2005. Yields and heating values of liquids and chars from spruce trunkbark pyrolysis. Energy Sources 27: 1367-1373.

Di Blasi, C., Branca, C., Santoro, A. & Hernandez, E.G. 2001. Pyrolytic behavior and products of some wood varieties. Combustion and Flame 124(1-2): 165–177.

Emrich, W. 1985. Handbook of charcoal making: the traditional and industrial methods. Dordrecht D. Reidel Publishing Company. Solar energy R&D in the European Community. Series E, Energy from biomass 7.

State of Finland's Forests. 2012. [Web-page] Finnish Forest Research Institute. Available from: http://www.metla.fi/metinfo/sustainability/c4-tree-species.htm. [cited 5.8.2014].

Felfli, F.F., Luengo, C.A., Suárez, J.A. & Beatón, P.A. 2005. Wood briquette torrefaction. Energy for Sustainable Development 9: 19-22.

Ferro, D.T., Vigouroux, V., Grimm, A. & Zanzi, R. 2004. Torrefaction of agricultural and forest residues. Conference publication 4 II-0185-FA Cubasolar, Guantánamo, Cuba.

Glossary: EU enlargements. 2014. European Commission. Eurostat. Available from: http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Glossary:EU_enlarge ments. [cited 23.10.2014].

Hakkila, P. 2006. Factors driving the development of forest energy in Finland. Biomass and Bioenergy 30(4): 281-288.

Hall, S. 2014. Charcoal kilns. [web-page]. Available from: http://www.panamintcity.com/panamint/charcoalkilns.html. [cited 5.8.2014].

Havimo, M. 2010. [unpublished communiqué]. Thermogravimetric analysis of spruce.

Heiskanen, V. & Riikonen, P. 1976. Havusahatukkien kuoren määrä ja siihen vaikuttavat tekijät. Summary: Bark amount in coniferous sawlogs and factors affecting it. Folia Forestalia. 250:1–67.

Heräjärvi, H., & Junkkonen, R. 2006. Wood density and growth rate of European and hybrid aspen in southern Finland. Baltic Forestry 12(1): 2-8.

Howard, E.T. 1971. Bark structure of the southern pines. Wood Science, Vol. 3(3): 136-148. Available from: http://www.srs.fs.usda.gov/pubs/ja/ja_howard008.pdf.

Hynynen, J., Niemistö, P., Viherä-Aarnio, A., Brunner, A., Hein, S. & Velling P. 2010. Silviculture of birch (Betula pendula Roth and Betula pubescens Ehrh.) in northern Europe. Forestry 83:103-119.

Industrial charcoal making. 1985. FAO. Rome. FAO forestry paper. ISSN 0258- 6150; 63. ISBN 92-5-102307-7.

Kellomäki, S. & Salmi, J. 1979. Koivutukkien kuoren määrä. Summary: Bark quantity of birch logs. Folia Forestalia 413: 1–16.

King, H-H., Solomon, P.R., Avni, E. & Coughlin, R.W. 1983. Modeling Tar Composition in Lignin Pyrolysis. Symposium on Mathematical Modeling of Biomass Pyrolysis Phenomena Washington, D.C. p. 1

KnowPulp. 2003. [e-Learning environment]. Principal of chemical pulping. Chemical composition of wood fiber. Available from : http://www.knowpulp.com/english/demo/english/pulping/cooking/1_process/1_princ iple/frame.htm. [cited 24.7.2014].

Kärhä, K., Elo, J., Lahtinen, P., Räsänen, T., Keskinen, S., Saijonmaa, P., Heiskanen, H., Strandström, M. & Pajuoja, H. 2010. Availability and Use of Wood-Based Fuels in Finland in 2020. Publications of the Ministry of Employment and the Economy. Energy and Climate 66: 2010.

Kärkkäinen, M. 2007. Puun rakenne ja ominaisuudet. Metsäkustannus. Helsinki. p. 245-247.

Laiho, R. & Laine, J. 1997. Tree stand biomass and carbon content in an age sequence of drained pine mires in southern Finland. Forest Ecology Management. 93: 161-169.

Lamlom, S.H. & Savidge, R.A. 2003. A reassessment of carbon content in wood: variation within and between 41 North American species. Biomass and Bioenergy 25: 381-388.

Li, R., Fei, J., Cai, Y., Li, Y., Feng, J. & Yao, J. 2009. Cellulose whiskers extracted from mulberry: a novel biomass production. Carbohydrate polymers 76: 94-99.

Lehtonen, A. 2005. Carbon stocks and flows in forest ecosystems based on forest inventory data. Dissertationes Forestales 11. University of Helsinki, Department of Forest Ecology.

Mantau, U., Saal, U., Prins, K., Steierer, F., Lindner, M., Verkerk, H., Eggers, J., Leek, N., Oldenburger, J., Asikainen, A. & Anttila, P. 2010. EUwood - Real potential for changes in growth and use of EU forests. Final report. Hamburg/Germany, June 2010. 160 p.

Manunya, P. & Sudhagar, M. 2011. Impact of torrefaction on the grindability and fuel characteristics of forest biomass. Bioresource technology 102: 1246-1253.

Milosavljevic, I. & Suuberg, E.M. 1995. Cellulose thermal decomposition kinetics: global mass loss kinetics. Industrial and Engineering Chemistry Research 34 (4): 1081–1091.

–, Oja, V., Suuberg, E.M. 1996. Thermal effects in cellulose pyrolysis: relationship to char formation processes. Industrial and Engineering Chemistry Research 35 (3): 653.

Mohan, D., Pittman, C.U. & Steele, P.H. 2006. Pyrolysis of wood/biomass for bio- oil: A critical review. Energy & Fuels 20: 848-889.

Murphy, J.D. & McCarthy, K. 2005. Ethanol production from energy crops and wastes for use as a transport fuel in Ireland. Applied Energy 82(2): 148-166.

Neves, D., Thunman, H., Matos, A., Tarelho, L.& Gomez-Barea, A. 2011. Characterization and prediction of biomass pyrolysis products. Progress in Energy and Combustion Science 37: 611-630.

Niemistö, P. 1995. Influence of initial spacing and row ‐to ‐row distance on the growth and yield of silver birch (betula pendula). Scandinavian Journal of Forest Research 10: 245-255.

Nikitin, N.I. 1966. The chemistry of cellulose and wood. Israel Program for Scientific Translations. Jerusalem. p. 496-585

Nurmi, J. 1993. Heating values of the above ground biomass of small-sized trees. Acta Forestalia Fennica 236. 30 p.

– 1997. Heating values of mature trees. Acta Forestalia Fennica 256. 28 p.

Ohliger, A., Förster, M., & Kneer, R. 2013. Torrefaction of beechwood: A parametric study including heat of reaction and grindability. Fuel 104: 607-613.

Pach, M., Zanzi, R. & Björnbom, E. 2002. Torrefied biomass a substitute for wood and charcoal. 6th Asia-Pacific International Symposium on Combustion and Energy Utilization.

Panshin, A.J. & De Zeeuw, C. 1980. Textbook of wood technology : structure, identification, properties, and uses of the commercial woods of the United States and Canada. McGraw-Hill series in forest resources. New York. 722 p.

Park, W.C., Atreya, A. & Baum, H.R. 2010. Experimental and theoretical investigation of heat and mass transfer processes during wood pyrolysis. Combustion and Flame 157: 481-494.

Pettersen, R.C. 1984. The chemical composition of wood. The chemistry of solid wood 207: 57.

Phanphanich, M. & Mani, S. 2011. Impact of torrefaction on the grindability and fuel characteristics of forest biomass. Bioresource technology 102: 1246-1253.

Pico Technology. 2014. [web-page]. Thermocouple Application Note. Available from: http://www.picotech.com/applications/thermocouple.html. [cited 5.8.2014].

Prins, M.J, Ptasinski, K.J. & Janssen, F.J.J.G. 2006. Torrefaction of wood: Part 1. Weight loss kinetics. Journal of Analytical and Applied Pyrolysis 77: 28-34.

Properties viewer, cellulose. 2014. [web resource]. chemicalize.org. ChemAxon. Available from: http://www.chemicalize.org/structure/#!mol=cellulose&source=calculate. [cited 12.11.2014]

Ranta, T. 2005. Logging residues from regeneration fellings for biofuel production–a GIS-based availability analysis in Finland. Biomass and Bioenergy 28: 171-182.

Repola, J. 2006. Models for vertical wood density of Scots pine, Norway spruce and birch stems, and their application to determine average wood density. Silva Fennica 40: 673-685.

Roadmap 2050. 2011. [web-page]. European Comission. Available from: http://ec.europa.eu/clima/policies/roadmap/perspective/index_en.htm.[cited 9.11.2011].

Räisänen, T. & Athanassiadis, D. 2013. [Journal]. Basic chemical composition of the biomass components of pine, spruce and birch. Available from: http://www.biofuelregion.se/UserFiles/file/Forest%20Refine/1_2_IS_2013-01- 31_Basic_chemical_composition.pdf. [cited: 5.8.2014].

Scheller, H.V. & Ulvskov, P. 2010. Hemicelluloses. Annual Review of Plant Biology 61: 263-289.

Schlamadinger, B. & Marland, G. 1996. Full fuel cycle carbon balances of bioenergy and forestry options. Energy Conversion and Management 37(6–8): 813-818.

Shah, A., Darr, M.J., Medic, D., Anex, R.P., Khanal, S. & Maski, D. 2012. Techno- economic analysis of a production-scale torrefaction system for cellulosic biomass upgrading. Biofuels Bioproducts & Biorefining-Biofpr 6: 45-57.

Shoulaifar, T., DeMartini, N., Willfor, S., Pranovich, A., Smeds, A., Virtanen, T., Maunu, S-L., Verhoeff, F., Kiel, J. & Hupa, M. 2014. Impact of Torrefaction on the Chemical Structure of Birch Wood. Energy & Fuels 28(6): 3863-3872.

Sjöström, E. 1993. Wood chemistry: fundamentals and applications. Orlando, FL. Academic Press. 223p.

Solid biofuels. 2010. [Worksheet for standard]. Fuel specifications and classes, Part 1: General requirements. ISO. Available from: http://www.google.fi/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CB8Q FjAA&url=http%3A%2F%2Fstandardsproposals.bsigroup.com%2FHome%2FgetP DF%2F757&ei=qxnZU7fLKKT8ywOk7IDoAQ&usg=AFQjCNHKE4zWOwZUp97 Jk3tSeZoX5f5v0w&bvm=bv.72185853,d.bGQ. [cited 30.7.2014].

Steinhagen, H.P. 1977. Thermal conductiveproperties of wood, green or dry, from - 40° to 100°. USDA Forest Service General Technical Report FPL 9. U.S. Department of agriculture forest service. Madison. Available from: http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr09.pdf. [cited 2.5.2013].

Stenius, P. 2000. Papermaking science and technology. Book 3, Forest products chemistry. Fapet. Helsinki. Publ. in cooperation with the Finnish Paper Engineers’ Association and TAPPI. p.350. ISBN:952-5216-00-4.

Strezov, V., Moghtaderi, B. & Luca, J.A. 2003. Thermal study of decomposition of selected biomass samples. Journal of Thermal Analysis and Calorimetry 72: 1041- 1048.

– , Patterson, M., Zymla, V., Fisher, K., Evans, T.J. & Nelson, P.F. 2007. Fundamental aspects of biomass carbonisation. Journal of Analytical and Applied Pyrolysis 79(1–2): 91-100.

Summers, D. 2010. [web-page]. Charcoal, Oil Shale and the Ecoshale process in Utah. Available from: http://bittooth.blogspot.fi/2010/04/charcoal-oil-shale-and- ecoshale-process.html. [cited 5.8.2014].

Talvitie, A. 1944. Kemiallinen teknologia 1.osa. Werner Söderström Osakeyhtiö. Porvoo. p. 212-213.

The EU climate and energy package. 2011. [web-page]. European Comission. Available from: http://ec.europa.eu/clima/policies/package/index_en.htm. [cited 9.11.2011].

Telmo, C. & Lousada, J. 2011. The explained variation by lignin and extractive contents on higher heating value of wood. Biomass and Bioenergy 35:1663-1667.

Tolunay, D. 2009. Carbon concentrations of tree components, forest floor and understorey in young Pinus sylvestris stands in north-western Turkey. Scandinavian Journal of Forest Research 24: 394-402.

Torvelainen, J. Ylitalo, E. & Nouro, P. 2014. [Web-publication]. Puun energiakäyttö 2013. Finnish Forest Research Institute. Forest Statistical Bulletins 31/2014. Available from: http://www.metla.fi/metinfo/tilasto/julkaisut/mtt/2014/puupolttoaine2013.pdf. [cited 1.10.2014].

Trossero, M., Domac, J., Kulisic, B., Karan, M. & Medic, S. 2008a. Environmental aspects of charcoal production in Croatia. TCP/CRO/3101 (A) Development of a

68 sustainable charcoal industry. FAO. Available from: http://s3.amazonaws.com/zanran_storage/www.drveniugljen.hr/ContentPages/46443 012.pdf. [cited 10.11.2011].

–, Domac, J. & Siemons, R. 2008b. Industrial charcoal production. TCP/CRO/3101 (A) Development of a sustainable charcoal industry. FAO. Available from: http://www.google.fi/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CCgQ FjAB&url=http%3A%2F%2Fpvbiochar.org%2Fforum%2Findex.php%3Faction%3 Ddlattach%3Btopic%3D1335.0%3Battach%3D234&ei=fkBjVPyMMIvlaqbogLAL &usg=AFQjCNHn1gIDYY7vV0xTnWiXdCR2Q0Ogsw&bvm=bv.79189006,d.d2s. [cited 10.11.2011].

Vázquez, G., Antorrena, G. González, J. & Freire S. 1997. The influencing of pulping conditions on the structure of acetosolv eucalyptus lignins. Journal of Wood Chemistry and Technology 17(1–2): 147–162.

Wannapeera, J., Fungtammasan, B. & Worasuwannarak, N. 2011. Effects of temperature and holding time during torrefaction on the pyrolysis behaviors of woody biomass. Journal of Analytical and Applied Pyrolysis 92: 99–105.

Warne, K. 2009. [web-page]. Green charcoal in Chame. Available from: http://laststands.kennedywarne.com/green-charcoal-in-chame. [cited 5.8.2014].

White, R.H. 1987. Effect of Lignin Content and Extractives on the Higher Heating Value of Wood. Wood and Fiber Science 19: 446-452.

Wiedenhoeft, A. 2010. Chapter 03: Structure and Function of Wood. In publication: Wood Handbook, Wood as an Engineering Material. General Technical Report FPL- GTR-190. Madison, WI. Department of Agriculture and Forest Service. Forest Products Laboratory: 508 p. Available from: http://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr190/chapter_03.pdf. [cited 5.8.2014].

Ylitalo, E. 2001. Puupolttoaineen käyttö energiantuotannossa vuonna 2000. Finnish Forest Research Institute. Forest Statistical Bulletin 574.

– 2013. Energy. In Publication: Finnish Statistical Yearbook of Forestry, 2013. Finnish Forest Research Institute. Available from: http://www.metla.fi/julkaisut/metsatilastollinenvsk/index-en.htm.

– & Ihalainen, A. 2013. Forest resources. In Publication: Finnish Statistical Yearbook of Forestry, 2013. Finnish Forest Research Institute. Available from: http://www.metla.fi/julkaisut/metsatilastollinenvsk/index-en.htm.

Zanzi, R., Sjöström, K. & Björnbom E. 2002. Rapid pyrolysis of agricultural residues at high temperature. Biomass and Bioenergy 23: 357-366.

Appendix 1, Chip size of untreated pine and birch including bark

Pine with bark Mass (g) 1. 2. 3. Sieve (mm) batch batch batch 4. batch TOTAL (%) >31,5 0 2,97 6,28 1,55 10,8 1 % 31,5 27,63 38,12 34,42 13,5 113,7 11 % 16 155,29 135,12 150,02 52,36 492,8 49 % 8 63,31 68,16 93,49 51,27 276,2 28 % 4 19,25 14,31 23,78 12,33 69,7 7 % 2 11,07 4,68 8,21 4,55 28,5 3 % 1 3 0,86 1,69 0,71 6,3 1 % 0,5 0,78 0,45 0,48 0,46 2,2 0 % 1000,1 100 %

Birch with bark Mass (g) 1. 2. 3. Sieve (mm) batch batch batch TOTAL (%) >31,5 22,48 85,64 13,33 121,5 12 % 31,5 74,62 67,46 83,65 225,7 23 % 16 102,69 135,49 126,13 364,3 36 % 8 45,67 77,28 82,14 205,1 21 % 4 8,78 18,28 20,13 47,2 5 % 2 3,34 8,33 14,59 26,3 3 % 1 0,86 1,21 4,18 6,3 1 % 0,5 0,39 0,69 1,7 2,8 0 % 999,1 100 %

Appendix 2.1, Calculation of HHV of birch samples

Explanation of sample name: BBtempX, where BB = Birch with bark, temp = pyrolysis temp. (°C), X = sample number HHV, bag 46344,00 J/g Sample Pyrolysis Bag weight HHV 1 , Sample + C change HHV 2 Sample HHV Average temp. (°C) (g) bag bag weight measured weight (g) (MJ/kg) HHV (MJ/kg) (g) (MJ/kg) (MJ/kg)

BB2501 250 0,14 6,61 0,68 1,85 27,52 0,54 20,91 BB2502 250 0,14 6,52 0,75 1,94 26,13 0,61 19,61 BB2503 250 0,14 6,60 0,64 1,71 26,87 0,50 20,27 20,26 BB3001 300 0,14 6,61 0,59 1,93 32,78 0,45 26,17 BB3002 300 0,14 6,51 0,60 1,88 31,83 0,46 25,33 BB3003 300 0,14 6,29 0,81 2,64 32,96 0,67 26,67 26,05 BB3501 350 0,14 6,59 0,72 2,38 33,35 0,58 26,77 BB3502 350 0,14 6,68 0,59 2,00 34,03 0,45 27,35 BB3503 350 0,14 6,67 0,86 2,75 32,07 0,72 25,40 26,51 BB4001 400 0,14 6,59 0,67 2,20 33,13 0,53 26,54 BB4002 400 0,14 6,65 0,75 2,61 35,02 0,61 28,36 BB4003 400 0,14 6,65 0,65 2,18 33,77 0,51 27,12 27,34 BB4501 450 0,14 6,54 0,70 2,41 34,94 0,56 28,40 BB4502 450 0,14 6,55 0,73 2,49 34,30 0,59 27,75 BB4503 450 0,14 6,37 0,63 2,16 34,74 0,49 28,36 28,17 BB untreated 0,15 6,74 0,69 1,65 24,04 0,55 17,30 Benzoic acid 0,15 7,09 1,15 3,33 29,25 1,00 22,16

HHV= HHV2- HHV 1

Appendix 2.2 Calculation of HHV of pine samples

Explanation of sample name: PBtempX, where PB = Pine with bark, temp = pyrolysis temp. (°C), X = sample number HHV, capsule 18813,00 J/g HHV, bag 46344,00 J/g Sample Pyrolysis Bag HHV 1 , Sample + C change HHV 2 Sample HHV Averag temp. (°C) weight bag bag measured weight (g) (MJ/kg) e HHV (g) (MJ/kg) weight (g) (MJ/kg) (MJ/kg)

PB2501 250 0,14 6,55 0,67 1,79 27,01 0,53 20,46 PB2502 250 0,14 6,68 0,60 1,63 27,61 0,45 20,92 PB2503 250 0,15 6,77 0,70 1,80 26,04 0,55 19,27 20,22 PB3001 300 0,14 6,34 0,56 1,81 32,60 0,42 26,26 PB3002 300 0,13 6,17 0,50 1,65 33,14 0,37 26,96 PB3003 300 0,14 6,31 0,63 1,97 31,78 0,49 25,47 26,23 PB3501 350 0,14 6,47 0,67 2,22 33,69 0,53 27,22 PB3502 350 0,14 6,45 0,70 2,30 33,27 0,56 26,82 PB3503 350 0,14 6,45 0,67 2,16 32,82 0,53 26,37 26,80 PB4001 400 0,14 6,41 0,73 2,47 34,08 0,59 27,67 PB4002 400 0,14 6,39 0,62 2,09 33,99 0,48 27,60 pb4002 400 0,14 6,53 0,65 2,20 33,93 0,51 27,40 PB4003 400 0,14 6,39 0,61 2,02 33,50 0,47 27,11 27,44 PB4501 450 0,14 6,34 0,63 2,26 36,10 0,50 29,75 PB4502 450 0,14 6,42 0,71 2,33 32,97 0,58 26,55 PB4503 450 0,14 6,61 0,67 2,19 33,14 0,53 26,53 27,61 PB untreated 0,14 6,62 0,65 1,60 24,77 0,51 18,15 Benzoic acid 0,14 6,40 1,02 2,96 29,22 0,88 22,83

HHV= HHV 2- HHV 1 73

Appendix 3.1 Birch with bark torrefied at 300 ˚C and 350 ˚C

400 Outer BIRCH, Inner 300 °C Exhaust pipe Cold junktion

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00 8:00

Time (hours)

500 BIRCH, Outer Inner 350 °C Exhaust pipe 400 Cold junktion

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00

Time (hours)

Appendix 3.1 Birch with bark carbonized at 400 ˚C and 450 ˚C

500 BIRCH, 400 °C Outer Inner Exhaust pipe 400 Cold junktion

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00

Time (hours)

500 BIRCH, 450 °C Outer Inner Exhaust pipe Cold junktion 400

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00

Time (hours)

Appendix 3.2 Pine torrefied at 250 ˚C and 300 ˚C

300 PINE, 250 °C

250

200

Outer Inner 150 Exhaust pipe Cold Junktion

100 Temperature (° C) Temperature(°

0 0:00 2:00 4:00 6:00 8:00 Time (hours)

400 Outer Inner PINE, 30 0 °C Exhaust pipe Cold junktion 300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00 8:00

Time (hours)

Appendix 3.2 Pine with bark carbonized at 350 ˚C and 400 ˚C

500 PINE, 35 0 °C Outer Inner Exhaust pipe 400 Cold junktion

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00 8:00 Time (hours)

500 Outer PINE, 40 0 °C Inner Exhaust pipe Cold junktion 400

300

200 Temperature (° C) Temperature(°

100

0 0:00 2:00 4:00 6:00

Time (hours)