Biomass Pre-treatment for the Production of Sustainable Energy

– Emissions and Self-ignition

Acta Wexionensia No 88/2006 Bioenergy Technology

Biomass Pre-treatment for theProduction of Sustainable Energy

- Emissions and Self-ignition

Katarina Rupar-Gadd

Växjö University Press

Biomass Pre-treatment for the Production of Sustainable Energy – Emissions and Self-ignition. Thesis for the degree of Doctor of Technology, Växjö University, Sweden 2006

Series editors: Tommy Book and Kerstin Brodén ISSN: 1404-4307 ISBN: 91-7636-501-8 Printed by: Intellecta Docusys, Göteborg 2006

Abstract Rupar-Gadd, Katarina, 2006. Biomass Pre-treatment for the Production of Sustain-able Energy – Emissions and Self-ignition. Acta Wexionensia No 88/2006. ISSN: 1404-4307, ISBN: 91-7636-501-8. Written in English. Organic emissions with focus on terpenes, from biomass drying and storage were investigated by Solid Phase Microextraction (SPME) and GC-FID and GC-MS. The remaining terpenes in the biomass (Spruce and pine wood chips) after dry-ing were dependant on the drying temperature and drying medium used. The dry-ing medium used was steam or hot air; the drying temperatures used were 140°C, 170°C and 200°C. Steam drying at 170°C left more of the terpenes remaining in the wood chips, not emitting them into the drying medium. The terpenes emitted from storage of forest residues and bark and wood chips increased up to three-four or four-five months of storage, and then dropped down to approximately the same low level as the first month. The leachate taken from the forest residue pile contained 27µg PAH per liter. The SPME response for a monoterpene (α-terpene) at different temperatures, amounts and humidities was quantified. The highest concentration calibrated was 250 ppm and the lowest 9.4 ppm. There is a better linear agreement at higher temperatures (70°C and 100°C) than lower temperatures (below 40°C). Organic emissions from biofuel combustion were measured at three medium sized (~ 1MW) biomass fired moving grate boilers fired with different fuels: dry wood fuel, forest residues and pellets. The PAH emissions varied by almost three orders of magnitude between the three boilers tested, 2.8-2500 μg/m3. The varia-tion in PAH emission is most probably a result of boiler design and tuning of the combustion conditions. When comparing the contribution to self-heating from different wood materials by means of isothermal calorimetry with different metals added and stored at dif-ferent temperatures, the differences were quite large. Some of the samples re-leased as much as 600mW/kg, whereas others did not contribute at all to the self-heating. The storage temperature, at which the samples released the most heat, was 50°C. There was a peak in heat release for most of the samples after 10-30 days. Stepwise increase in temperature did not favour the heat release in the sample Dry Mix; the heat released was even lower than when it was directly put in the different storage temperatures. When metal is added, there is an increase in heat release, the reference sample without metal released 200mW/kg compared to 600mW when copper was added.

Key words: terpenes, SPME, biomass, pellets, self-ignition, storage, drying, calorimetry

qç=ãó=é~êÉåíë=få=ãÉãçêá~ã=

^åíçå=C=eÉêã~=oìé~ê=

5

List of publications This thesis is based on the following publications, appended in the thesis and re-ferred to in the text by their Roman numerals: I. Emission of volatile organic compounds (VOC) from Drying and

Storage of Biomass Katarina Rupar and Mehri Sanati Conference Proceedings of the 12th European Conference and Technol-ogy Exhibition on Biomass for Energy, Industry and Climate Protec-tion, Amsterdam, The Netherlands, 17-21 June, 2002.

II. The Release of Organic Compounds during Biomass Drying de-pends upon the Feedstock and /or Altering Drying Heating Medium Katarina Rupar and Mehri Sanati Biomass and Bioenergy, Vol 25, Issue 6, December 2003, p 615-622

III. The Release of Terpenes during Storage of Biofuels Katarina Rupar and Mehri Sanati Biomass and Bioenergy, Vol 28, Issue 1,January 2005, p29-34

IV. Boiler Operation Influence on the Emissions of Submicrometer-

Sized Particles and Polycyclic Aromatic Hydrocarbons from Bio-mass-Fired Grate Boilers Lena Lillieblad, Aneta Szpila, Michael Strand, Joakim Pagels, Katarina Rupar-Gadd, Anders Gudmundsson, Erik Swietlicki, Mats Bohgard and Mehri Sanati Energy & Fuels;2004;18(2); 410-417.

V. Solid Phase Micro Extraction Fibers, Calibration for use in biofilter

applications Katarina Rupar-Gadd, Mohammad Bagher Bagherpour, Göran Holm-stedt, Ulrika Welander and Mehri Sanati.

Biochemical Engineering Journal, accepted

6

VI. Spontaneous Combustion of Biofuels caused by Microbial Activity Katarina Rupar-Gadd, Lars Wadsö, Göran Holmstedt, Bror Persson, Per Blomqvist and Mehri Sanati

Conference Proceedings of the 2nd World Conference and Technologi-cal Exhibition on Biomass for Energy, Industry and Climate Protection, Rome, Italy,10-14 May, 2004

VII. Parameter Study Dependency on Heat released from Microbial Ac-tivity in the Self-ignition of Stored Biofuels Katarina Rupar-Gadd, Lars Wadsö, Jonas Widuch, Patrick Van Hees, Göran Holmstedt, and Mehri Sanati Fire and Materials Journal, submitted

Results related to this thesis are also presented in: Self-heating in biofuels; Chemical and biological processes Katarina Rupar-Gadd, Lars Wadsö, Xi Li, and Mehri Sanati manuscript A Parameter Study of Drying and Storage of Biomass. Katarina Rupar-Gadd Licentiate thesis, 3 June 2003 Hydrocarbon emission from storage of biofuel (In Swedish: Kolväteutsläpp vid lagring av biobränsle) Katarina Rupar and Mehri Sanati Värmeforskrapport 796, Dec 2002 Parameters Study in Laboratory Scale of Drying Biofuels in Particle Level (In Swedish: Parameterstudie av Biobränsle Torkning i Laboratorieskala på Partikelnivå) Mehri Sanati, Katarina Rupar, Ulrika Karlsson, Mostafa Faghihi Värmeforskrapport 748, Aug 2001

7

List of Contents

List of publications................................................................................................ 5 List of Contents..................................................................................................... 7 Background........................................................................................................... 9 Bioenergy.............................................................................................................. 9 Biofuels ............................................................................................................... 10 Wood................................................................................................................... 11

Combustion of wood .............................................................................. 11 Wood structure ....................................................................................... 11 Wood rots ............................................................................................... 12

Microorganisms .................................................................................................. 13 Fungi....................................................................................................... 13 Spores ..................................................................................................... 14 Bacteria................................................................................................... 14

Organic compounds ............................................................................................ 15 Terpenes ................................................................................................. 15 Environmental effects............................................................................. 16 Health Effects ......................................................................................... 19 Exposure limits ....................................................................................... 19 Odor limits.............................................................................................. 20

Projects................................................................................................................ 21 Emissions from combustion................................................................................ 22 Drying ................................................................................................................. 23

Dryers ..................................................................................................... 24 Storage and Self-Ignition .................................................................................... 25

Microbial activity ................................................................................... 25 Chemical processes................................................................................. 26 Self-heating............................................................................................. 26 Factors during storage............................................................................. 27 Energy/substance loss............................................................................. 27 Particle size............................................................................................. 27 Temperature............................................................................................ 27 Moisture.................................................................................................. 28 Oxygen ................................................................................................... 28 Nitrogen.................................................................................................. 28 Carbon dioxide ....................................................................................... 28 Spores ..................................................................................................... 28 Metals ..................................................................................................... 29

Experimental ....................................................................................................... 30 TGA .................................................................................................................... 30 GC-FID/GC-MS ................................................................................................. 30 PAH sampling..................................................................................................... 31 SPME.................................................................................................................. 32

8

Biomass Samples ................................................................................................ 34 Drying..................................................................................................... 34 Storage.................................................................................................... 34 Self-ignition............................................................................................ 35 Preparation of samples............................................................................ 35 Storage.................................................................................................... 36

Calorimetry ......................................................................................................... 37 Isothermal Calorimeter TAM Air ........................................................... 37

Results and Discussion........................................................................................ 39 Drying ................................................................................................................. 39

TGA........................................................................................................ 39 Comparison between different biofuels ................................................. 40 Comparison between different drying temperatures............................... 41 Comparison between different drying media.......................................... 44

Storage ................................................................................................................ 45 Temperature............................................................................................ 45 Precipitation............................................................................................ 46 Emissions to air ...................................................................................... 46 Emissions to water.................................................................................. 48 Conclusions ............................................................................................ 48

SPME.................................................................................................................. 49 Calibration curve for concentrations....................................................... 49 Temperature influence ............................................................................ 50 Conclusions ............................................................................................ 50

Emissions from Combustion ............................................................................... 51 PAH measurements ................................................................................ 51 Conclusions ............................................................................................ 52

Self-ignition ........................................................................................................ 53 Different materials .................................................................................. 53 Temperature parameter ........................................................................... 53 Dry Mix .................................................................................................. 53 Pellets 8mm ............................................................................................ 53 Aged Sawdust ......................................................................................... 53 Stepwise increase.................................................................................... 54 Metal parameter...................................................................................... 54 Dry Mix .................................................................................................. 54 Pellets 8mm ............................................................................................ 55 Aged Sawdust ......................................................................................... 55 Conclusions ............................................................................................ 55 Author’s contribution to presented articles............................................. 59

Future work......................................................................................................... 60 Acknowledgements............................................................................................. 61 References........................................................................................................... 62

9

Background The attempt to convert Sweden’s energy system has led to an increased interest in biofuels and, in Sweden, the supply of biofuels is good so it is likely to remain one of the most important energy sources in the future (Hillring 1999). The purpose of this work was to obtain knowledge on different parameters dur-ing biomass pre-treatment for the production of sustainable energy – with focus on organic emissions from drying and storage, and self-heating during storage. Hopefully, this knowledge can be applied on an industrial level to minimize the emissions of organic compounds and reduce the risk of self-ignition during stor-age.

Bioenergy Today, biofuels are responsible for approximately a fifth of Sweden’s entire net energy supply (Swedish Energy Agency 2005). This means that biofuels are the second largest energy source, after oil and nuclear power. Bioenergy resources provide approximately 14% of the worlds energy supply (IEA Bioenergy) Wood fuels are traditionally used for heating single-family houses, but there is also an established commercial market in the district-heating sector. Environ-mental taxes on carbon dioxide have made wood based biofuels a more cost-effective alternative to fossil fuels. The main reasons for this are both economi-cal and ecological (Bain et al. 1998, Bohlin et al. 1998, Hektor 1998, Hillring 1998, Rösch and Kaltschmitt 1999, Hillring 2000, Schneider and Kaltschmitt 2000). An advantage of bioenergy is that it does not contribute carbon dioxide to the atmosphere and thereby does not contribute to global warming. Organic materi-als contain carbon; plants take up carbon from the carbon dioxide in the air to be able to grow. When the plants are combusted, this carbon is released and returns to the atmosphere. Biofuels have stored their carbon recently and thus there is no increase in atmospheric carbon dioxide when combusted. A disadvantage of biofuels is that they normally have to be upgraded for use in small-scale combustion devices, like the ones found in single-family houses. For larger boilers in district heating power plants, the fuel does not have to be up-graded since those boilers are designed for fuel with large variations in moisture content. Here it is more important to have a fuel buffer, i.e. storage.

10

Biofuels According to Swedish Standard SS 187106, biofuel is fuel derived from biomass and biomass is defined as material with biological origin, which may or may not have been processed. Bioenergy is all energy where biomass is starting material. The attempt to convert Sweden’s energy system has led to an increased interest in biofuels. Different biomasses have been investigated, but the largest group is wood fuel, which can be divided into energy forest fuel, recycled wood fuel and forest fuel. Wood fuels can be divided into forest fuels and byproducts from the wood proc-essing industry. The different fuels have great variations in fuel quality and moisture content. There are many ways to categorize the fuel types; the follow-ing is a selection of the different fuels investigated during the project. Forest residue chips are made from branches and treetops. The moisture content of between 35-60 % varies with the time of year, composition and if the material has been stored before chipping. The major fraction size is smaller than 50 mm. Bark is a byproduct from sawmills when the bark is peeled off from the logs. The amount of bark varies with type of tree and with the age of the tree, but ap-proximately 10-15 % of a tree log is bark. There is also another type of wood chips that come from when the stem closest to the roots is reduced to the same dimension as the other parts of the stem. The bark and the root-reducing wood chips usually have a fraction size smaller than 50 mm and a moisture content of 40-65 %. Sawdust is produced when logs are divided into parts. It can be dried, or up to a moisture content of 50 %. Dry wood chips come from adjusting dried wood into desired dimensions. The wood is chipped into a fraction size less than 60 mm, and moisture content of 15-30 %. Pellets are made from different materials. Some are made entirely from the inner parts of the tree, some are made from bark, some are made from peat, and some are made from different mixings of materials. An example of a “mixed” pellet is 45% aged sawdust, 20% shavings, 15% new sawdust and 20% cellulose chips (aged).

11

Wood The biofuels investigated are wood fuels, made mainly from softwood, Norway spruce (Picea Abies) and Scots pine (Pinus Sylvestris).

Combustion of wood Heat causes wood to thermally decompose in a process called pyrolysis. The py-rolysis produces charcoal and volatiles. The volatiles are what actually burn dur-ing combustion. The slowest combustion is smoldering combustion, which emits heat but no light. Glowing combustion produce more heat and some light and flaming combustion is what we normally would think of when we think of fire.

Wood structure Wood is divided into two main groups; softwood and hardwood. The softwood trees have needles and the wood from is softer then the wood from broad leafed hardwood. The biofuels investigated in this thesis is mainly from softwood since this is the wood available in this region of Sweden. The softwoods investigated are Norway spruce (Picea Abies) and Scots pine (Pinus Sylvestris).

The stem of the tree store water, carbohydrates and minerals, conduct water and minerals and transports foods and hormones (Kozlowski and Pallardy 1997). The wood column inside the tree is called xylem. The young, outer part of the xylem, sapwood, conducts sap and water. In sapwood there are both dead tracheids, ducts which transport water and give strength to the tree and living parenchyma cells which store food. The parenchyma cells absorb oxygen which reacts with the food (starch and sugars). Due to this oxidation, the air in the sapwood only contains 10% oxygen compared to 21% in fresh air (Kubler 1987). The carbon dioxide level in the sapwood is raised from 0.03 % in fresh air to 10%. On aver-age, 10% of the sapwood cells are alive, and as the xylem gets older, the cells die and become heartwood, located at the center of the stem. The heartwood has a lower moisture content since it is cut off from water transport. Between the wood and the inner and outer bark, there is a vascular layer called cambium, which is the most respirating layer because all the cells are alive.

Wood is composed of cellulose, hemicellulose and lignin. Cellulose (40-50% by weight) gives wood its mechanical strength and is made up of chains of glucose residues. Hemicellulose (25-40%) is composed of branched polysaccharides. Lignin (20-35%) is a highly hydrophobic complex polymer. Lignin functions as a barrier against microbial attack (Tuomela et al. 2000, Andersson 2001). Lignin and hemicellulose binds the wood structure together.

12

Wood rots Degradation of lignin, cellulose and hemicellulose is performed by a variety of fungi, often through different kinds of rot. There are wet or dry rot, white, brown or black, they can grow fast or slowly and grow for as long as the substrate and environment allow. (Kitani and Hall 1989, Madigan et al. 2000, Andersson 2001). When white rot occurs, in the beginning the wood is red-brown and then be-comes yellow- white to white, appearing moist and fibrous. White rot fungi have the ability to degrade cellulose, hemicellulose and lignin, and wood degradation by white-rot fungi has been studied extensively, but is still not fully understood (Andersson 2001) Brown rot does not affect the lignin, only the cellulose and the hemicellulose; the wood turns dark, brown, shrinks, cracks into dices and eventually ends up as a brown powder. The wood degradation by brown rot fungi appears to involve both enzymatic and non-enzymatic reactions. Brown rot fungi can form hydro-gen peroxide, and one theory suggests that the peroxide diffuse and react with ferrous iron in the wood, and thereby initializing the decay. (Andersson 2001) Soft rot degrades cellulose, hemicellulose and may partially degrade lignin. The wood appears wet, spongy and hollow parts are formed in the cell walls of the wood fibres. The mechanical property of the wood is not as good as before, without obvious signs of attack. Plays an important role in decomposing wood components of buildings. Attacks the surface layer of wood chips, which be-comes softened. Appears typically in wood of high water content and high nitro-gen content. The damage can be extensive since they can grow at both high and low moisture content, even if they grow quite slowly. Lignin is mainly degraded by Basidiomycetes and Actinomycetes. (Kitani and Hall 1989) Basidiomycetes are rotfungi that can use cellulose or lignin as carbon source and energysource. Actinomycetes are bacteria that form multi cellular filaments and resemble fungi. (Tuomela et al. 2000) Cellulose is mainly de-graded by Trichoderma veride (T. reesei has higher cellulose activity) and As-pergillus niger. (Kitani and Hall 1989) Hemicellulose is mainly degraded by brown-rot fungi.

13

Microorganisms Both fungi and bacteria can colonize piles of comminuted biomass, mostly molds and actinomycetes (bacteria with growth pattern similar to fungi). Micro-organisms can be roughly divided into four groups by their temperature toler-ances: Psycrophile, mesophile, thermophile and hyperthermophile. Psycrophile have a maximum growth temperature below 20°C and a minimum growth temperature of 0°C or lower, mesophile have a temperature optimum be-tween 20 and 40°C. Thermophile have a growth temperature optimum above 45°C and hyperthermophile have a growth temperature optimum above 80° (Figure 1.)

Figure 1. Relation of temperatures to growth rates for Psycrophile, mesophile, thermo-phile and hyperthermophile microorganisms.

Fungi Fungi are chemoorganotrophs and use organic carbon for both energy and carbon purposes. They are morphologically larger than bacteria and are not as quickly reproduced. If competition in the form of bacteria is killed off by bacteriocides, this could lead to a “fungal bloom” which is next to impossible to control. Fungi are extremely adaptable and can survive in a water free environment for several years without growth. They respirate, but can also use anaerobic fermentation. Respiration generates more energy for the organism than anaerobic fermentation, resulting in a greater microbial growth. Molds are filamentous fungi that consist of many filaments (hyphae) that form a so-called mycelium. Hyphal branches grow up from the surface and into the air and form spores or conidia. This makes the surface of the mycelia look a bit

14

dusty, and the white color of the mycelia changes into the color of the conidia (could be black, blue-green, red, yellow or brown). Thermophilic fungi are known to be involved in composting and wood degrada-tion, but the mechanisms involved in decomposing biomass are not well under-stood. The knowledge about their physiology is fragmentary and some thermo-philic fungi have almost optimal growth at room temperature, and the belief that thermophilic fungi grow poorly at low temperatures should be reexamined (Maheshwari et al. 2000).The oxygen uptake of mesophilic fungi Aspergillus ni-ger was not affected by temperature changes between 15 and 40°C, fungi which can maintain an optimal metabolic rate over a temperature range will have an advantage (Maheshwari et al. 2000). Thermophilic fungi succed mesophilic fungi in composts or stored biofuels, and most of the initially available carbon source would have been depleted. Mould species isolated from air in sawmills or in stored wood chips are: Aspergillus, frequently encountered in buildings used for agricultural activities. Cladosporium, frequently found in a conifer wood environment Penicillium, a diversity of species in a conifer environment. (Simeray et al. 1997,Thörnqvist and Jirjis 1990, Jirjis 2005)

Spores If rot fungi decompose the wood, the molds release health hazardous spores in-stead; this is mainly an occupational hazard. Spores are airborne with a size range from 2 to 10μm (Monn 2001). During large-scale storage of wood chips, more than 1010 spores kg-1 dry mass is not uncommon (Hakkila 1989, Thörnqvist and Jirjis 1990, Richardson et al. 2002, Jirjis 2005). During the first months of storage, there will be a rapid increase in the number of colony forming units (vi-able spores, CFU), which also can reach as high as 1010 kg-1. The CFU in saw-mill air could be as high as 3300 m-3 (Simeray et al. 1997).

Bacteria Bacteria are single-celled with a size between 0,5 to 3,0μm, much smaller than fungi. Due to their small size, bacteria have a very large surface/volume ratio, which allows a quick transfer of soluble substrates into the cell. This results in bacteria often being more dominant than larger microorganisms like fungi. Ac-tinomycetes are bacteria that produce multicellular filaments and resemble fungi. They are present during composting and can sometimes become visible on the surface. Actinomycetes can degrade some cellulose and loosen lignin and can tol-erate higher temperatures than fungi. Actinomycetes can survive as spores (Tu-lomela et al. 2000). The dominant bacteria found in stored wood chips are ther-mophilic actinomycetes Thermoactinomyces vulgaris (Richardson et al. 2002).

15

Organic compounds Softwood consists of cellulose, hemicellulose, lignin and extractives. The extrac-tives are organic compounds, mostly terpenes. Volatile Organic Compounds (VOC) includes all organic compounds with a boiling point between 50-240°C, or 100-260°C for polar compounds. Polycyclic Aromatic Hydrocarbons (PAH) are a group of chemicals with two or more benzene rings; carcinogenic and with varying boiling points. PAHs are formed during incomplete combustion or from naturally occurring resin acids either during pulping processes or during biologi-cal treatments. They are semi volatile and can be found both in gas phase and on particles.

Terpenes

A terpene is a compound that can be divided into two or more isoprene units (Figure 2). An oxygen-containing terpene is called a terpenoid.

Figure 2 . Isoprene unit C5H8.

Terpenes are classified according to the number of isoprene units, see Table 1.

Table1. Classification of terpenes. Number of

isoprene units Formula

Monoterpenes 2 C10H16 Sesquiterpenes 3 C15H24 Diterpenes 4 C20H32 Triterpenes 6 C30H48 Tetraterpenes 8 C40H64 Polyterpene n (C5H8)n

Terpenes are components of essential oils obtained by distilling or extracting various parts of plants. These compounds have various uses for example in medicine, and the terpenes obtained from flowers are used in the manufacturing of perfume. Essential oils are straight-chain or cyclic compounds and may be mono- sesqui- or diterpenes. Essential oils are the source of most of the odors found in flowers, fruit and wood of many plants. Many are volatile and evaporate into the air, especially on warm days.

16

Environmental effects Emission of VOC is a potential environmental problem; it may contribute to air pollution, eutrofication or disturbing odors (Brown and Foote 1998, Hinds 1999, Petersson 1995, Corchnoy and Atkinson 1990). In many countries there are le-gal restrictions on the amounts of terpenes that may be released. Ground-level ozone and other photo-oxidants are formed by a chemical reaction between VOCs and nitrogen oxides (NOx) in the presence of sunlight. Anthropogenous emissions of VOC to air are in Sweden 500 00 ton /year (Jensen and Wolkoff 1996). In Sweden 1000-thousands of tons of photochemical very reactive ter-penes are yearly emitted from the pulp and paper industry to air (Strömvall and Peterson 1993a, Strömvall and Peterson 1993b, Jensen and Wolkoff 1996). Photo oxidation of terpenes produces blue haze (Kozolowski and Pallardy 1997). Blue haze is a phenomenon that can be seen over wooded hills during summer time or close to wood drying industries. This is due to the isoprene units emitted from wood drying and from the trees. The blue haze is caused by light scattering from an aerosol produced by photo oxidation of isoprene and other hydrocarbons (Strömvall and Petersson 1993a). Single particles are smaller than the light wavelength, which means that they cannot be registered optically, but it is possi-ble to see the collective optical effect. Organic compounds emitted in gas phase and condensating through homogenous and heterogeneous condensation form the small particles. Lately, the health effects of small airborne particles have been given attention; the particles are small enough to be transported into the lungs (Strömvall and Pe-tersson 1993b). There are several health effects associated with PM 10 and PM 2.5 and there are exposure regulations. In the presence of sunlight, hydrocarbons and nitrogen oxides (NOx) can give rise to ground level ozone. As alkenes, with comparatively high molecular weight, the bicyclic monoterpenes are during the summer transformed in less than a day, and monocyclic terpenes in less than an hour, and in some cases in some minutes. The quick formation of photo-oxidants can contribute to forest damage within a few miles distance from the emission source and also cause damage to crops. Terpenes have an important ecological function in trees; they work as signal sub-stances for insects, and either act as a defense against insects or are released in the sap when trees are injured. The terpene content varies between different spe-cies, parts of the tree, and growth localities (Borg-Karlson et al. 1993, Sjödin et al. 2000, Fäldt 2000, Fäldt 2001 et al., I,II,III). A selection of identified chemical compounds in spruce and pine wood chips is shown in Table 2. The difference between spruce wood chips with different growth localities is shown in Figure 3. During felling of trees, conifers emit unnaturally high levels of ter-penes, and the simultaneous emission of nitrogen oxides from the machines en-hances the local formation of photo-oxidants (Jensen and Wolkoff 1996).

17

Figure 3. Terpene content in spruce wood chips a) south of Sweden b) north of Sweden. Sampled and analyzed by SPME and GC-FID (II). (1) α-pinene (2) 3-carene (8) longi-folene When considering bioenergy in future energy systems it is important that infor-mation on the environmental effects is available as there are various administra-tive permissions needed (e.g. further use of by-products/wastes and the emis-sions of different substances released into the atmosphere). It may be necessary to clean the exhaust gas or condensate before releasing it into the environment.

M inutes 10 20 30

mVolts 30

20

10

10 20 30 Minutes

a) 30

20

10

1 3

8

b)

8 1 3

mVolts

18

Table 2. Chemical structure of identified compounds from spruce and pine wood chips. Compound Structure Compound Structure α-pinene

C10H16 CAS 80-56-8

D-longifolene C15H24 CAS 475-20-7

β-pinene

C10H16 CAS 18172-67-3

β-Cubebene C15H24

CAS 13744-15-5

3-carene C10H16

CAS 13466-78-9

γ-Muurolene C15H24

CAS 30021-74-0

D-limonene C10H16 CAS5989-27-5

α-Muurolene C15H24

CAS 31983-22-9

α-Longipinene C15H24 CAS 5989-08-2

γ-Cadinene C15H24

CAS 39029-41-9

Copaene C15H24 CAS 3856-25-5

β-Cadinene C15H24

CAS 523-47-7

α-Gurjuene C15H24 CAS489-40-7

Calamenene C15H22

CAS 483-77-2

Hexanal C6H12O CAS 66-25-1

O

19

Health Effects When terpenes are metabolized, epoxides are formed. These epoxides are in many cases reactive enough to cause damage. Limonene causes allergy (Birger-son et al. 1985, Karlberg and Lindell 1993), and turpentine and 3-carene are skin sensibilsating (Kasanen et al. 1999). It is likely that the compound has to be me-tabolized in the skin to be able to induce the allergic reaction. There have been a number of investigations on terpene levels and exposure in occupational environments such as sawmills, carpentry and joinery shops. The exposure measurements show large variations, the limits are sometimes ex-ceeded, and sometimes not (Dahlqvist et al. 1996a, Dahlqvist et al. 1996b, Eriks-son et al. 1997, Rosenberg et al. 1999, Teschke et al. 1999, Demers et al. 2000, Svedberg and Galle 2000). Many independent studies have shown an increased intensity of asthma, non-asthmatic airflow obstruction, upper and respiratory problems and sino-nasal cancer with exposure of softwood dust or monoterpenes (Falk-Filipsson 1995, Demers et al. 1997, Demers et al. 2000). The effects are more chronic than acute, (Eriksson et al. 1997) or there is a slight inflammatory reaction in the up-per airways (Dahlqvist et al. 1996a, Dahlqvist et al. 1996b). The formation of ground level ozone and photochemical oxidants also results in health effects. The two main effects are irritation of eye and mucus, caused by secondary products such as aldehydes, and diminished respiration. During peri-ods of high levels of oxidants in the air, there is a significant rise of the number of people seeking medical advice for respiratory problems (Warfinge 1997).

Exposure limits The hygienic limit is the highest acceptable average concentration of an air pol-lutant in the breathing air. The Swedish 8-hour occupational permissible expo-sure limit for terpenes is 150 mg/m3 (25 ppm), Table 3. The short time exposure value over a fifteen-minute period is 300 mg/m3 (50ppm) (AFS 2000). Table 3. Exposure limits for terpenes (AFS 2000:3) Compound CAS Exposure limit

Ppm mg/m3

Short time Ppm mg/m3

Terpenes 25 150 50 300 α-pinene 80-56-8 25 150 50 300 β-pinene 127-91-3 25 150 50 300 3-carene 13466-78-9 25 150 50 300 Limonene 138-86-3 25 150 50 300 Terpentine 8006-64-2 25 150 50 300

20

Odor limits The odor threshold for different mono terpenes are very different, Table 4. An odor threshold is defined as the concentration at which fifty percent of a group can smell it. Compared to the exposure limit, the terpene concentration can in-crease up to 60 times from when you first smell it before it reaches the exposure limitvalue. Table 4. Odor thresholds for terpenes. (Jensen and Wolkoff 1996) Compound CAS Odor threshold (μg/m3) α-pinene 80-56-8 3890 β-pinene 127-91-3 36000 3-carene 13466-78-9 2450 Limonene 138-86-3 2450

21

Projects This thesis is mainly based on three major projects:

• A drying project where the organic emissions from dried sawdust was investigated

• A storage project where organic emissions from stored wood chips was investigated

• A self-ignition project where the contribution to self-heating from mi-croorganisms was investigated

All of these projects focus on the biofuel before combustion, but field measure-ments of PAH emissions from biofuel combustion was also performed. In the drying and storage project, a sampling technique called Solid Phase Microextrac-tion (SPME) was used, and an investigation of the calibration of this instrument was also performed.

22

Emissions from combustion In Sweden there are a number of small biomass fired grate boilers used for dis-trict heating. These small boilers (<2MW) have particle emission requirements in the range of 100-150 mg/Nm3 and can meet these requirements using only a multi-cyclone for dust removal. The larger boilers with particle emission re-quirements of 50 mg/Nm3 need an additional removal device, commonly an elec-trostatic precipitator (ESP). Emissions of organic components such as PAH are a result of incomplete com-bustion. There is still limited knowledge about the effect of operating conditions on the emissions from these small boilers. The operating conditions influence on the emissions was investigated, measure-ments were carried out at 3 medium sized (~ 1MW) biomass fired moving grate boilers, used in commercial district heating units in the south of Sweden. The boilers where fired with different fuels: dry wood fuel, forest residues and pel-lets. To investigate the PAH levels in flue gas, sampling was performed with a sam-pling train consisting of a probe, a particle filter and a condenser followed by a solid adsorbent filter, a drying tower, a pump and a volume flow meter (IV).

23

Drying The reduction of water content in biomass, i.e. drying, results in the simultane-ous increase in thermal value, preservation potential, ease in storage and trans-portation, less negative impact on the environment and a more uniform combus-tion (Gislerud 1990, Obernberger 1998, McDonald et al. 2000). Large boilers have often been scaled for the use of biomass of varying moisture content, whereas small-scale combustors, gasification units and production of pellet and other processed biofuels, demand drier input materials (Wimmerstedt 1999). VOCs are emitted to the air already at a low temperature during biomass drying. When wood is dried, these compounds consist mainly of mono-terpenes. Ter-penes constitute about 90 % of the easily extractive material in wood. Other or-ganic matter released from wood biofuels, drying in liquid- and gas-phase, are a mixture of water vapor and compounds such as lightweight alcohol, carboxylic acids, methanol, and aldehydes (Cronn et al. 1983, Gevao and Jones 1998, McDonald et al. 1999, Otwell et al. 2000, Brammer and Bridgwater 2002). In certain boilers dimensioned for dry biofuels, and in small boilers, biofuel dry-ing is necessary for the process to function optimally, but drying is also an essen-tial part in biofuel processing plants. For existing and developed techniques for drying to work on the market, both the operational availability and the environ-mental performance must be improved. Biofuels can be taken from different sources and can be used in many types of plants to produce electricity and heat. Forest fuels and agricultural fuels are built up from cells that take up water and expand so they fill up the space in the cell structure. When the biomass is dried, the water is removed and the cells shrink, leading to packing of the cells; the biomass shrinks. Removal of the water also leads to removal of mass, making transport easier. Storage is also made easier due to both a microbial growth and decomposition being prevented, which would lead to substance loss and deterioration in fuel quality. A moist fuel normally contains 40-60 % water, but this can vary depending on the type of fuel, previous use and storage. There are also great variations between heartwood, sapwood and bark. The sapwood normally contains more water than the heartwood. Emissions from a biofuel dryer depend on a number of factors:

• The capacity of the dryer • The raw material composition (bark, type of wood) • Dryer parameters (drying temperature, residence time in the dryer, oxy-

gen level in the gas) • Cleaning device

24

In most cases, the drying is performed under such conditions that only volatile compounds are released (Fagernäs 1993). Besides water, the emitted compounds are organic compounds such as light-weight alcohols, carboxylic acids and ter-penes. The amount of light-weight alcohols and carboxylic acids emitted are de-pendent on the storage time, but in most cases small compared to the amount of terpenes. Volatile compounds can be formed during drying by decomposition of heavier compounds, but the probability of this taking place is small since the temperature during drying is usually not high enough for decomposition to occur. There are a number of reasons for drying biofuels:

• easier to store for longer periods • easier to transport • less emissions • smaller dimension of the combustion unit, more stable combustion con-

ditions • higher effective thermal value

Three parameters are significant in the drying process; the temperature, the resi-dence time and the mixing (tubulence) (Fagernäs 1993, Barry and Corneau 1999, Wu and Milota 1999, Otwell et al. 2000, Björk 2001) High temperature increases the NOx emission, but lowers the hydrocarbon emis-sion. These three parameters should be optimized during all drying processes in order to obtain a good and complete process with minimal emissions. The tem-perature influence on released terpenes was investigated in I and II.

Dryers There are many different types of dryers; they are usually classified by the way heat is supplied to the material, indirect or direct. The drying media used are hot air, flue gas or steam. Flue gas can come from a number of different sources like industrial processes or boilers. The influence of drying media on the release of terpenes was investigated in I and II.

25

Storage and Self-Ignition The majority of biofuels is directly transported to the thermal power plants and combusted within a few days. However, there is a need for a buffer for when no wood is being chipped for example during holidays or during the breaking up of the frost in the ground when heavy transports are not allowed on the small roads in the forests. Approximately 10 % of the yearly production in Sweden is stored 5-10 months, but almost twice that amount passes through a storage terminal in a shorter period. Transports longer than 50-100 km are not economically or environmentally vi-able, which means that the terminals must be placed near the power plants. This results in the terminal storage being placed close to where people live, often in small communities where there is an existing paved area, for instance a closed down factory of some kind. This results in people living close by complaining about odors and being worried if the emitted compounds are hazardous. There has also been a concern regarding fungal spores released from the storage. There is very limited knowledge today about the release of hydrocarbons during the storage of wood biofuels and thereby also limited knowledge about the health effects or environmental impact. There have been a number of investigations, but these have mainly focused on the substance loss or moisture- and temperature profiles within the piles during the storage (Nylinder and Thörnqvist 1980, Thörnqvist 1982, Thörnqvist 1983, Thörnqvist 1985, Thörnqvist 1986, Thörn-qvist 1987, Jirjis 1995, Nurmi 1995, Sanderson et al. 1997, Nurmi 1999, Le-htinkangas 2000). Small piles in the investigations are approximately 50-100 m3s, and large piles are approximately 600-2.000 m3s.

Microbial activity During storage of wood biofuels there is a microbial activity in the piles, but to what extent and how fast the growth occurs is determined by the access to nutri-tion in the material, the moisture content and the size of the material. Microbial growth on freshly harvested biomass is minimal, but colonization occurs as soon as the pile is built. The growth of micro fungi is larger in the first four months of storage (Thörnqvist 1983).The microorganisms produce heat and moisture and thereby enhance the factors that lead to further microbial growth. This leads to deterioration in fuel quality, substance losses, health effects due to releasing of spores and VOC, and adds to the self-heating of the material. The heat produced from different biomass samples was investigated in VI and VII. The self-heating can in some cases lead to self-ignition, but to reach a temperature where self-ignition can take place, heat from other sources than microbial is required. At higher temperatures different chemical processes take over from the biological.

26

Chemical processes Chemical oxidation processes require higher temperatures than the biological processes. Some chemical oxidation processes such as iron rusting or fat becom-ing rancid can occur at ambient temperatures. The temperature interval where wood oxidates with a noticeable heat generation is between 80° and 90°C, lower if the wood contains moisture. Between 100° and 300°, a number of investiga-tions have shown that the weight loss of woody samples is greater in an atmos-phere containing oxygen, an indirect symptom of oxidation (Kubler 1987). At temperatures above 100°C, wood starts to pyrolyse and becomes charred, the black residue of pyrolysed organic material is called charcoal. If a pile of stored material is dug open during ongoing pyrolysis, i.e. excess oxygen is added, the pyrolysis will turn into an overt fire. Pyrolysis and chemical oxidation progress exponentially as temperature rises. Chemical oxidation is what raises the tem-perature pyrolysis, but is there a difference if the chemical oxidation was pre-ceded by biological oxidation or not? Is the biological process necessary to in some way induce the chemical oxidation leading to self-ignition? The investiga-tions VI, VII and future work will hopefully give some answers to these ques-tions.

Self-heating Self-ignition in stored material is due to different heat releasing processes like microbial activity, change in moisture content and oxidation of volatile organic compounds (Figure 4). During storage of organic material, heat is produced (Hogland and Marques 2003, Riggle 1996). Self-heating in stored wood fuels is due to the living parenchyma cells continuing to respirate after the material is comminuted, causing initial heat development. The microbial activity in the inte-rior of the pile contributes to the temperature rise in the interval between 0-70°C, and heat is also formed when the differences in moisture content is evened out. Chemical oxidation processes are responsible for the temperature rise from ap-proximately 50°C. The temperature rise will in some cases lead to self-ignition. Why self-ignition occurs has not been established.

Figure 4. Microbial/chemical contribution to temperature rise in stored materials

27

Factors during storage Factors influencing self-heating and microbial growth are temperature, moisture content, particle size and size and form of the pile and duration of storage (Gray 2002, Kubler 1987, Thörnqvist and Jirjis 1990, Richardson et al. 2002). The processes taking place in the piles during storage of organic material consume oxygen and emit carbon dioxide and water.

Energy/substance loss Heat-generation processes reduce the energy value of the fuel; the energy loss is 10-24 % during 6-12 months storage, and the matter loss 2-17 %, depending on where in the pile the sample is taken. The matter loss is dependent on the initial moisture content and the decomposition is the greatest during the first months. If stored under cover (roof), the energy loss is only 5%. An energy gain has been noted in a 120m3 pile. (Thörnqvist 1985)

Particle size The comminution provides a much larger area for microbial colonization and storing of smaller particle size wood chips give a higher compaction and lower permeability and reduces the dissipatation of the released heat. Smaller particles mean higher temperature and more microbial growth. Larger particles mean lower temperatures and not as much microbial growth, until it starts, then the growth is really fast. (Jirjis 2005) Larger particle size of the wood chips inhibits the microbial growth due to faster drying (Hakkila 1989).

Temperature The activity of fungi and bacteria continues beyond temperatures of 60°C where the parenchyma living cells stops respirating, some fungi can tolerate tempera-tures of 60°C and bacteria as high temperatures as 80°C (Richardson 2002). The temperature directly affects the microbial activity, and the temperature in a pile is dependant on the initial moisture content. (Thörnqvist and Jirjis 1990) The temperature is the limiting factor; at freezing temperatures there is hardly any microbial growth. (Richardson et al. 2002)

In smaller piles with comminuted biofuels, there is a rapid temperature increase up to 60-70°C and the temperature will remain at that level for the storage pe-riod. If the proportion of needles and bark in the pile is increased, there will be an increase in the pile temperature. In one pile, compacted and stored under roof, the temperature in the lower central parts reached more than 300°C (instrument out of range) after four months then decreased after two weeks to 100°C at the end of the storage period.

28

Moisture Small sized shavings form close layers and trap moisture, compared to larger sized particles which will dry out (Hakkila 1987). When moisture content drops below 20%, water is found only in the cell walls and cannot be used by microor-ganisms. Drying will lower the microbial activity, but rewetting can revive dor-mant spores (Richardson et al. 2002). The material in larger piles will dry up in the center of the pile, but the total moisture content is increased or unchanged. Smaller piles will dry more than larger and the longer the storage period, the lar-ger the increase in moisture content. If stored under cover, the moisture content will decrease. If the pile is compacted and stored without cover, it will increase in moisture content. The moisture content will decrease more if the pile is un-covered and uncompacted, than compacted and covered. If covered with tarpau-lins, the moisture content will increase.

Oxygen The chimney effect in wood chip piles provides oxygen required for metabolic activity and different self-heating processes (Kubler 1987, Richardson et al. 2002). Small sized shavings, form close layers and restrict air movement (Kubler 1987, Hakkila 1989). Microorganisms consume oxygen down to a level of 1% (compared to 21% in fresh air), and some bacteria can survive under anaerobic conditions, and continue to produce heat, although not as much as under aerobic conditions. Covering the piles will restrict the access to oxygen and inhibit mi-crobial growth.

Nitrogen The microbial activity in a wood chip pile is to some extent dependant on the amount of nitrogen. Since there is more easily accessible nitrogen in needles and bark then wood, there should be a larger activity in needles and bark (Thörnqvist and Jirjis 1990).

Carbon dioxide The processes taking place in the piles during storage of organic material con-sume oxygen and emit carbon dioxide and water. The change of carbon dioxide and oxygen can give an indication as to the biological and chemical activity in the pile. The concentration in the pile quickly increases to a level of 30% during the first 24 hrs. The concentration then decreases to below 4%, but in the center of the pile, close to the ground, the level stabilizes at 15% (Thörnqvist and Jirjis 1990).

Spores During large-scale storage of wood chips, more than 1010 spores kg-1 dry mass is not uncommon (Hakkila 1989, Richardson et al. 2002, Jirjis 2005, Thörnqvist

29

and Jirjis 1990). If the material is too dry, the microorganisms can survive as spores and rewetting can revive dormant spores (Richardson et al. 2002).

Metals Metals can catalyze heat-generating reactions in organic material. Ferrous mate-rials accelerate the self-heating in piles of wood chips and are to be avoided in the piles (Kubler 1987, Thörnqvist 1987).

30

Experimental The main sampling techniques and methods for analyzing are described, see arti-cles for further details. The biomass samples, field measurements and lab scale, and the different parameters which were varied during the investigations are also described.

TGA A Thermo Gravimetric Analyzer (TGA) investigates the weight change of a ma-terial to determine the composition, thermal stability and other related phenom-ena. A high resolution TGA 2950 (TA Instrument) with high sensitivity and a temperature range up to 1000 °C was used. Thermo Gravimetric analysis was used to study how water and organic com-pounds are emitted from biofuels during drying (II). It is assumed that there is a simultaneous emission of water and organic compounds, but this was hard to de-termine despite the high-resolution instrument used.

GC-FID/GC-MS The emissions from the biomasses in both the drying and storage project (I, II, III) and the quantification (V) were analyzed on a Varian Saturn 2000 Gas Chromatograph (GC) equipped with a Flame Ionization Detector (FID) or a Sat-urn II ion trap Mass Spectrometer detector (MS). Reference compounds or the National Institute of Standards and Technology (NIST) mass spectral database was used for identification. For programming and specifications see I,II,III and V. The programming used the resulted in the division of peaks in the chroma-togram into three groups, the first around ten minutes consist mainly of monoter-penes, the second around twenty minutes consists mainly of sesquiterpenes and the third around thirty minutes consists mainly of diterpenes.

31

PAH sampling The sampling technique used for sampling PAH in the flue gas from 3 different medium sized (~ 1MW) biomass fired moving grate boilers is described below (IV). The sampling train consists of a probe, a particle filter and a condenser followed by a solid adsorbent filter, a drying tower, a pump and a volume flow meter (Figure 5). Before sampling, the sampling equipment including filters is cleaned and prepared with and internal standard. The probe is inserted into the stack and by choosing the proper probe diameter isokinetic sampling is achieved. The glass fibre particle filter is placed directly after the probe, outside the stack. The sam-pling gas is cooled below 20ºC and the condensate collected into a flask. A glass cartridge packed with XAD-2 and supported by a polyurethane foam (PUF) plug is used to collect the gas phase PAH downstream from the condenser. After each sampling cycle the sampling train is rinsed with acetone, which is added to the total sample. The sample treatment includes soxhlet and liquid-liquid extraction, filtration through silica gel column and dimethyl formamide (DMF) purification before GC-MS analysis.

Figure 5. PAH sampling train.

Probe Filter

Condenser

Filter Dryer

Pump

Volume Flowmeter

Condensate

32

SPME There are a number of methods for measuring concentrations of organic com-pounds in air, e.g. Tenax absorption tubes or Tedlar bags. One of the newer techniques is called Solid Phase Microextraction, SPME. The sampling in both the drying project and the storage project was performed with Solid Phase Microextraction (I, II, III), SPME, which is a solvent free technique invented in the early nineties (Pawliszyn and Liu 1987, Pawliszyn 1997, Pawliszyn 1999, Arthur and Pawlisyn 1990). SPME is easy to use and re-duces time and cost for sample preparation.

Figure 6. The SPME instrument. A one-cm silica fiber with an adsorbing/absorbing polymeric coating is attached to a piston in a syringe like holder that protects the fiber during storage or septa penetration (Figure 6). The extraction principle can be described as an equilib-rium process where the analytes are distributed between the fiber and the sample (Figure 7).

Figure 7. SPME adsorbtion/absorbtion in a vial followed by thermal desorption in an in-jector.

33

Liquid, solid or gaseous samples can be sampled by SPME. The fiber can be used for field sampling without any pre collecting of samples. The coating of the fiber is of the same sort used as stationary phase in columns in gas chromatogra-phy. Polar coatings such as poly acrylate (PA) or carbowax are suitable for sam-pling polar compounds and non-polar materials such as polydimethyl siloxane (PDMS) are suitable for non-polar samples. The PDMS has been used in envi-ronmental analytical chemistry to sample VOCs in surface water, wastewater and air samples (I,III, Namiesnik et al. 2000, Chai and Pawliszyn 1995, Augusto et al. 2001, Koziel and Novak 2002) Methods to quantify VOCs in air have been developed (Bartelt 1997, Augusto et al. 2001, Zhang and Pawliszyn 1993, Martos and Pawliszyn 1997, Martos et al. 1997, Gorlo et al. 1997), the (red) SPME fiber with 100μm PDMS was cali-brated by varying sampling time, temperature, relative humidity and analyte concentration (V.)

The treatment of samples was similar to earlier reports using SPME (Czerwinski et al. 1996, Eisert and Pawliszyn 1997, Koziel et al. 2000, Steffen and Pawliszyn 1996, Potter and Pawliszyn 1994, Namiesnik et al. 2000, Boyle et al. 2002). The dried biomass samples were weighed in 4-ml vials; the vials were filled to approximately two-thirds of the total volume (I,II). The samples were heated for 30 minutes at 50°C, after this the fiber was exposed and headspace vapor was sampled for ten minutes. Samples from the storage were collected by exposing the fiber directly above the piles for ten minutes (I,III).

34

Biomass Samples The biomass samples investigated are mainly softwood, Norway spruce (Picea Abies) and Scots pine (Pinus sylvestris). Samples were in the form of small or large wood chips, forest residues or wood pellets.

Drying The different samples in the drying project (I, II) were investigated to see how the drying medium and/or drying temperature would influence the amount and type of organic compounds that will remain in the samples after drying. From and environmental and economical point of view, it is desirable to have as much organic compounds as possible left in the biomass after drying. Any excess emission will lead to lesser calorific heat value in the biofuel and more emissions to clean in the flue gas or condensate from the dryer. The samples used were green samples of Norway spruce (Picea Abies) and Scots pine (Pinus sylvestris) wood chips. A spouted bed dryer (located at Karlstad University), with a capacity of 100-kg at atmospheric pressure, dried the samples (Granström 2002). In general, spruce samples were dried with steam, whereas pine samples were dried with both hot air and steam. Three drying temperatures were used. The drying temperature was kept around 140 ºC, 170 ºC and 200 ºC during this procedure and the drying media was recirculated. The samples were dried to approximately the same moisture content; all the samples were dried to moisture contents below 10%. A different sample of spruce wood chips and a sample of pellets were also analyzed for comparison to the projects samples. The difference in spectra between for example green and dried samples is believed to be the emissions during drying. The samples were investigated with Solid Phase Microextraction and analyzed with a Gas Chromatograph equipped with a Flame ionization Detector (GC-FID) and a Gas Chromatograph equipped with a Mass Spectrometer (GC-MS).

The dried samples were size fractionated into eight different fractions, prior to the TGA measurement. The moisture content of all the samples was determined by standard procedure SS187170.

Storage The fuels investigated in the storage project (I, III) were bark/wood chips and forest residues, stored in piles. Sampling was performed with Solid Phase Micro-extraction and analyzed with a Gas Chromatograph equipped with a Flame ioni-zation Detector (GC-FID) and a Gas Chromatograph equipped with a Mass Spectrometer (GC-MS).

35

The sampling period was 200 days, from June 2001 to January 2002. Samples were taken once a month from a terminal storage located outside of Växjö, in the south of Sweden. The piles investigated were approximately seven meters high and 50 meters long, but fuel was collected from the piles during the season. A few samples were also taken from the leachate from the pile containing bark/wood chips.

Self-ignition The biofuel samples used were a selection of the biofuels available from differ-ent manufacturers/importers in Sweden. The material is mainly softwood (Picea Abies and Pinus sylvestris). At a pellet manufacturer, two different sizes of pellets (six and eight mm) were produced from a mixture of shavings (sawmill dried), aged sawdust (stored out-doors for more than three months) and fresh sawdust. The mixture of the differ-ent biomasses is dried before pressed into pellets. The wood samples used were:

• Aged sawdust • Pellets 8mm • Dried mixture of sawdust before pressed to pellets (Dry Mix)

Preparation of samples The samples were prepared and stored in the 20 ml glass ampoules that are used for analysis in the calorimeter. Double samples were made, except for some of the reference samples (without metal) where only single samples were made. Sample size was approximately 2 g. Since the pellets are pressed under high pressure and temperature and the final moisture content is below 10%, there is no or very low activity in the samples. Since the samples investigated were very small (approximately 2 g), the samples were autoclaved to minimize random microbial activity, and inoculated with 7% w/w partially degraded forest residues to create a more uniform sample series. If the samples showed activity before inoculation, they were autoclaved one more time. The moisture content in the pellets is approximately 8-10%, which is too low for microbial activity. Initial attempts were made to increase the moisture content by storing the samples at higher relative humidity, 85.5%RH and 100%RH, but the moisture content did not increase enough. The samples moisture content were increased by adding 22 %w/w water, and stored at 100% RH.

36

Metal was added to see if any catalytic effects could be observed. The metals used were Manganese, Iron and Copper. Manganese was added as an oxide (MnO2), Copper was added as an oxide or a nitrate (Cu2O and Cu(NO3)*3H2O), and Iron was added as oxide, nitrate or citrate Fe2O, Fe(NO3)3*9H2O and C6H5FeO7*5H2O. Initially, the concentration was 10% w/w metal powder of the wood sample, but the concentration was lowered to 0.09% w/w metal ion of the total sample.

Storage The glass ampoules were placed in glass cans with lids together with an amount of water to control the relative humidity (100%RH). The glass cans were then placed inside an oven to control the storage temperature, Figure 8. The storage period stretches from –20 days to +74 days, day zero representing the day when the sample was inoculated with degraded forest residues, the nega-tive days in the storage period is when the sample was prepared and autoclaved before inoculation. The samples were stored for up to 74 days after inoculation. The storage temperatures were 20°C, 40°C, 50°C, 55°C and 60°C, with focus on 50°C, 55°C and 60°C. Before the samples were loaded into the calorimeter, air was added to the inside of the ampoules with a pipette to ensure aerobic condi-tions in the samples.

Figure 8 . Storage arrangement for biomass samples.

37

Calorimetry

Calorimetry, derived from the Latin calor meaning heat, is the science of meas-uring the amount of heat. All calorimetric techniques are based on the measure-ment of heat, and any process or sample that may generate (exothermic process) or consume (endothermic process) heat can be studied by calorimetry (Bjurman and Wadsö 2000).

Isothermal Calorimeter TAM Air In isothermal calorimetry the reaction environment is maintained at a constant temperature and the heat energy exchanged from a heat sink is measured. Heat flow between the reaction vessel and the heat sink is, ideally, identical to the thermal power produced in the reaction vessel.

P = dQ /dt

Q = ∫ P dt

The integral of thermal power P [W] over time (t) is the heat Q [J] which is pro-portional to the amount of sample m[g] (Figure 9)

Figure 9. The Thermal Power released from a sample vs. time. The integral of the curve is equal to the heat produced per gram sample.

Q

ther

mal

pow

er P

time t

38

The calorimeter used is an isothermal TAM Air, (Thermometric AB, Järfälla Sweden)(Figure 10). It has eight channels, temperature interval 5-60oC, a limit of detectability of 2 mW and a baseline stability of <±20μW in 24 hours. The samples are contained in 20 ml glass ampoules sealed with aluminum caps with rubber seals. The calorimetric measurements were made at 20°C, 40°C, 50°C, 55°C and 60°C. The measuring range was 60mW. Electrical calibrations accord-ing to the manufacturer's manual were executed approximately every third month for each temperature.

Figure 10. TAM Air isothermal calorimeter (from Thermometrics AB).

39

Results and Discussion

Drying

TGA Isothermal drying of green and hot air dried (170°C) spruce wood chips at three different temperatures (Figure 11). The temperatures were used: 50ºC, 70ºC and 90ºC. The shape of the curve in 11(a) where there is a slow drying procedure from 100 % weight down to 60 % could be due to the water loss through capil-lary forces and a simultaneous release of VOC (Bengtsson and Sanati 2004). The drying curves in 11(b) where the wood chips have already been dried (170°C) are probably due to reabsorbed water from the ambient air and possibly and con-tinued release of VOC.

Figure 11. Isothermal drying for 1 hr of spruce wood chips from north of Sweden a) green spruce wood chips b) hot air dried at 170°C spruce wood chips by means of TGA.

40

Comparison between different biofuels Pine and spruce contain different amounts of terpenes to begin with; the terpene content is higher in pine than spruce. The change of amount α-pinene present in the wood chips after drying at 170 degrees is shown in Figure 12.

Figure12. Amount α-pinene present in spruce and pine wood chips before and after dry-ing at 170°C.

0

500

1000

0 20 40 60 80MC wood chips (w/w %)

a-p

inen

/ m

g w

ood

chip

s (a

e /m

g od

w)

spruce

pine air

pine steam

41

Comparison between different drying temperatures The same material was dried at different temperatures and with different drying medium. The resulting chromatograms from investigating terpene remaining content in the dried wood chips together with the chromatogram for the wood chips before drying are shown in Figure 13,14 and 15. The pine wood chips dried in steam (Figure 13) at 170°C has the highest amount of remaining ter-penes in the wood chips, the drying temperatures 140°C and 200°C has removed more of the terpenes. The optimal drying temperature in this example is 170°C, since it leaves more of the organic compounds in the wood chips, not emitting them into the drying medium. Figure 13. Analysis of hydrocarbon emissions from fresh and steam dried pine wood chips by means of GC-FID; (a) fresh (b) steam dried in 140°C (c) steam dried in 170°C (d) steam dried in 200°C

(a)

(b)

(c)

(d)

5 15 25 minutes

mVolt 200 100

0 200 100

0 200 100 0

200 100 0

5 15 25 minutes

5 15 25 minutes

5 15 25 minutes

42

The pine wood chips dried in hot air (Figure 14) at 140°C has a slightly higher amount of remaining sesquiterpenes (around 20 minutes) in the wood chips, but the wood chips dried at 170°C has more of the monoterpenes left. The new peak at 7 minutes is hexanal.

Figure 14. Analysis of hydrocarbon emissions from fresh and hot air dried pine wood chips by means of GC-FID; (a) fresh (b) air dried in 140°C (c) air dried in 170°C (d) air dried in 200°C

( a )

( b )

( d )

( c )

mVolt 200 100 0

5 15 25 minutes 200 100 0

5 15 25 minutes 200 100 0 5 15 25 minutes 200 100 0

5 15 25 minutes

43

The spruce wood chips dried in hot air (Figure 15) all display a new peak of hex-anal at 7 minutes. All of the used temperatures effectively remove the terpenes from the wood chips during drying, emitting them into the drying medium.

Figure 15. Analysis of hydrocarbon emissions from fresh and hot air dried spruce wood chips by means of GC-FID; (a) fresh (b) air dried in 140°C (c) air dried in 170°C (d) air dried in 200°C

mVolt 30 20 10 0

( b )

( c )

( d )

( a )

40 20 0

5 15 25 minutes

5 15 25 minutes

5 15 25 minutes

5 15 25 minutes

40 20 0

40 20

44

Comparison between different drying media The difference between green pine wood chips, hot air dried wood chips and steam dried wood chips is shown in Figure 16. The new peak at seven minutes in the air dried sample was identified as hexanal. The steam drying leave more of the terpenes in the wood chips, compared to the air drying where the terpenes are emitted into the drying medium, not remaining in the wood chips. Drying by us-ing steam would be the best alternative in this case. Figure16 . Analysis of hydrocarbon emissions from fresh and in various media dried pine wood chips by means of GC-FID; (a) fresh (b) air dried in 170°C (c) steam dried in 170°C

5 15 25 minutes

mVolt 200 100 0

5 15 25 minutes

200 100 0

5 15 25 minutes

200 100 0

(a)

(b)

(c)

45

Storage

Temperature The ambient temperature and the temperature directly above the pile were meas-ured during sampling. Figure 17 shows the ambient temperature and the tem-perature directly above the bark/wood chips pile and the forest residue pile. The average temperature during 2000 was 7°C, 2 degrees higher than normal (SMHI 2000). The fall of 2000 was the second warmest over the past one hundred years (SMHI 2000). There is a peak in the temperature for both piles at the end of Oc-tober. This is due to heat released from biological or chemical processes in the piles. This could not be attributed to the respiration heat from the parenchyma cells in the comminuted wood, since exposure to 50°C for half an hour kills most of the cells and at 60°C they all die. The temperature inside a pile of wood chips quickly rises above 50°C and remains there for several weeks (Thörnqvist and Jirjis 1990). Cells in wood chips have been found to survive for 1 to 3 weeks (Kubler 1987).

Figure 17. Temperature ambient, above the bark/wood chips and Δ forest residue.

0

10

20

30

40

j j a s o n d j

month

degr

ee C

46

Precipitation The attempt to measure the precipitation on the site failed when the temperature dropped below the freezing point, so the precipitation data were taken from IVL Swedish Environmental Institute in Aneboda, not far from the site. The success-ful precipitation measurements were compared to the ones from IVL and found to correspond. Figure 18 shows the precipitation during the sampling period.

Figure 18. Precipitation from June 2001 to mid January 2002. SMHI, the Swedish Meteorological and Hydrological Institute rated the fall of 2000 as the fifteenth wettest of the past hundred years. The total precipitation during 2000 and 2001was 20-40% higher than normal (SMHI 2000, SMHI 2001).

Emissions to air The emissions from the bark/wood chips pile are high in the middle of the stor-age period and low in the beginning and end of the period. The emissions in Sep-tember do not follow this trend; the emissions were lower. The emissions from the forest residue are also high in the middle of the storage period and low in the beginning and end of the period. The emission level drops to a really low value already in November.

0

5

10

15

20

25

j j a s o n d j f

mm

47

May July September November January

May July September November January

Figure 19 shows how the emissions from the Bark/Wood chips pile vary from June to January. The emissions increase from June to August up to September where there is a sudden drop in the emissions. The emissions then continue to increase in October and November. The measurements in January show a de-crease in emissions, down to approximately the same level as in June.

Figure 19. Mono terpenes, Sesqui terpenes, Di terpenes

and Total emissions.

Figure 20 shows how the emissions from the Forest residue pile vary from June to January. The emissions increase from June to October. The measurements in November show a sudden drop in emissions; the level is the lowest measured during the storage period. The measurements in January show a small increase again, but only up to the levels in June-July (the beginning of the storage period).

Figure 20. Mono terpenes, Sesqui terpenes, Di terpenes

and Total emissions.

The highest emission is in October-November for the Bark/wood chips and in September/October for the forest residue.

0

500

1000

1500

2000

2500

MC

ount

s

0

500

1000

1500

2000

2500

MC

ount

s

48

Emissions to water Samples of leachate have been taken at two occasions from two piles, one with bark/wood chips and one with forest residues. The leachate taken from two dif-ferent piles are more similar than leachate taken from the same pile a month later. A sample from the bark/wood chips pile was analyzed for PAH (for more details see III). The total result was 27,27µg PAH per liter. There were some dif-ficulties in collecting leachate, even after high amounts of precipitation. For the extreme case where all of the precipitation that theoretically can leach through a pile give this concentration, the emission of PAH to water is approximately 3 mg/m3 wood chips. The PAH limit in drinking water is 0.10μg/l, but PAH emis-sions from biomass-fired boilers (~ 1MW) have been measured to 2.8-2500 μg/m3 (IV).

Conclusions The terpenes emitted from storage of forest residues and bark and wood chips in-creased up to 3-4 months or 4-5 months of storage, and then dropped down to approximately the same low level as the first month. The leachate taken from the forest residue pile contained 27µg PAH per liter.

49

SPME The SPME sampling method was used in both the drying project (I, II) and the storage project (I, III). The purpose of this study (V) was to investigate if the SPME fibres also could be used for quantification, not just identification. The previous studies quantified the terpenes by comparing the samples to each other, not by comparing the terpene content to a standard. This gives a change in con-centration between the samples, but not a definite concentration. The analyte used in the experiments was alpha-pinene, a hydrophobic organic compound commonly found in wood, and therefore found in wood storage facilities, wood processing industries and wood based biofilters (Misra et al. 1996, Mohseni and Allen 2000, Cohen 2001, Dhamwichukom et al. 2001). The calibration result was used to quantify the emissions from a lab-scale biofilter used to remove α-pinen in an air stream (Bagerpour et al. 2005).The SPME fibres were calibrated for different concentrations of alpha-pinene at different temperatures and relative humidities.

Calibration curve for concentrations The relation between the response of detection device (Mcounts) and the concen-tration in gas phase (ppm, mg/lit) is an important part of this investigation. This functionality is presented in Figure 21. The results imply a very good linear rela-tion between these two parameters. To obtain different concentrations of ana-lytes, a 5.5 liter air reservoir was charged by defined amounts of alpha-pinene in micrograms. Choosing a high volume reservoir reduces possible errors in the weighting of alpha-pinene, since the total amount will be equilibrated in the vol-ume of bottle. The experiments were performed at ambient temperature and rela-tive humidity.

Figure 21. The MCount response of the GC-MS vs. the concentration of α-pinene (Mg/l), dotted lines show 90% confidence interval.

0

20

40

60

10

30

50

70

0 0,1 0,2 0,3mg/lit

Mcount

50

Temperature influence Since the technique of sampling by the SPME fibres, is mainly depending on the mass diffusion processes, variation in the affecting parameters as like as tem-perature, pressure and some times presence of other competitive compounds, will change the out coming results of this method. Since in this study, the poly-meric stationary phase has a liquid nature, presence of other compounds should not cause any competitive effects. But at constant pressure, any change in tem-perature will have considerable effects on the results. The calibration curves for temperatures 24°C, 40°C, 70°C and 100°C are pre-sented in Figure 22.

Figure 22. Calibration curve for 24 °C 40°C 70 °C and *100°C There is a better linear agreement at 70°C and 100°C than 24°C and 40°C de-grees. The difference in linear agreement is not believed to be caused by over-loading of the system since higher response values have been reached in other experiments, and still shows a good linear agreement (V). The highest concentra-tion that has been calibrated is 250 ppm and the lowest is 9,4 ppm. Higher con-centrations than 250 ppm were not within the linear range and lower concentra-tions could not be examined, the limiting factor being the scales in combination with the volume of the calibration vessel.

Conclusions The calibration for SPME is very time consuming work since the calibration needs to be performed for different temperatures, different humidities and differ-ent concentrations. The highest concentration calibrated was 250 ppm and the lowest 9,4 ppm. There is a better linear agreement at higher temperatures (70°C and 100°C) than lower temperatures (below 40°C).

05

1015202530354045

0 20 40 60 80 100

Conc. (ppm)

Mco

unt

51

Emissions from Combustion

PAH measurements Measurements were carried out at 3 medium sized (~ 1MW) biomass fired mov-ing grate boilers fired with different fuels: dry wood fuel, forest residues and pel-lets (IV). The boilers are O2 regulated with a preset value of 4-8 %. The tempera-ture of the flue gas is around 1000°C, after the heat exchanges the temperature has decreased to about 200ºC. The boilers were equipped with multi-cyclones for dust removal. Electrical low-pressure impactor (ELPI) was used to determine the particle num-ber concentration, low-pressure cascade impactor (DLPI) was used for the mass size distribution, and elemental analysis of the fly ash was made by Particle in-duced X-ray emission (PIXE). Emissions of polycyclic aromatic hydrocarbons (PAHs) were measured using the PAH sampling train (Table 5). Table 5 . PAH levels in sampled gas at recorded O2 levels and boiler loads.

Fuel gas (m3)

PAH (μg/m3 )

O2 halt (%)

Load(%)

Boiler 1 Dry Wood 2.182 120 7,2 49,4Boiler 1 Dry Wood 1.788 155 6,5 58,4Boiler 1 Dry Wood 1.542 69 6,2 50,5Average 115 Boiler 3 Moist Forest Residues 1.345 3,7 4,6 86,1Boiler 3 Moist Forest Residues 1.434 2,5 4,8 83,6Boiler 3 Moist Forest Residues 1.573 2,2 4,3 78,9Average 2,8 Boiler 4 Pellets 1.293 4100 6,3 66,3Boiler 4 Pellets 1.429 860 6,3 62,4Boiler 4 Pellets 1.182 2600 7,8 50,8Average 2520 Before PAH sampling, the sampling equipment including filters was cleaned and prepared with and internal standard. The sampling probe was inserted into the, downstream from the cyclone and isokinetic sampling was achieved by choosing the proper probe diameter. After each sampling cycle the sampling train was rinsed with acetone. The volume sampled was approximately 1 Nm3. All the ex-

52

periments were repeated three times. The samples were extracted by soxhlet and liquid-liquid extraction, concentrated and eluted through a silica gel solid-phase extraction and extracted by dimethyl formamide (DMF). The samples were ana-lyzed by GC-MS. The total PAH concentration was achieved by adding the 29 selected PAH. In addition to PAH, concentration of O2, temperature at the sampling point and boiler load (fuel feed rate) was recorded (Table 5.).

Conclusions The submicron particle emission was in the range of 50-75 mg/m3, and was not greatly influenced by load. The particle number concentration increased with in-creasing load. The PAH emissions varied by almost three orders of magnitude between the three boilers tested, 2.8-2500 μg/m3. It was difficult to identify any general parameters correlating to the PAH emissions. The variation in PAH emission is most probably a result of boiler design and tuning of the combustion conditions.

53

Self-ignition The results from the calorimetric measurements are presented in mW / (kg wood). All the samples have been autoclaved, inoculated with degraded forest residues, water was added and samples stored at 100%RH at different tempera-tures. Metal was added to some samples to observe any catalytic effects.

Different materials When comparing Aged Sawdust to Dry Mix and Pellets 8 mm, stored at 60 oC, there is no difference in the thermal power released between the different materi-als (VI). Most of the samples give a thermal power release between 10-30mW/kg. At 50 oC, the Aged Sawdust give a slightly lower value than the Dry mix and the Pellets 8mm. There is a peak in the values, 200-240mW/kg, of Dry mix and Pel-lets 8mm approximately 10 day after the samples have been inoculated with de-graded forest residues. After that, the values slowly decrease to a level of ap-proximately 100mW/kg after 45 days, and then decrease even more to a level of 35 mW/kg after more than 70 days. The initial level before inoculation was be-tween 10-50mW/kg.

Temperature parameter

Dry Mix There is a peak in the energy production after 10-30 days for the 50°C and 55°C series. The 20°C and 60°C series have a constant low level of energy produc-tion. The peak values in the 50°C and 55°C series are 240 and 233 mW/kg.

Pellets 8mm There is a peak in the energy production after 10 days for the samples stored in 50°C, the samples stored in 55°C and 60°C do not increase their energy produc-tion at all during the storage period.

Aged Sawdust Both the 50°C and the 60°C series have constant low values during the storage period, the 50°C series 30-40mW/kg and the 60°C series 10-30mW/kg .

54

Stepwise increase In an attempt to simulate a slow temperature increase which would be the case in a pile of wood chips, the Dry mix material (reference samples and with added metal) was first stored at 20°C, then at 40°C, 50 °C and 60°C (Figure 23.). Dou-ble samples were made, Figure 23 show only one of the two samples for each metal/reference. The aim was to create an environment where different microor-ganisms could succeed each other when the temperature changed. The time for each temperature interval was between 20-50days, which was maybe to long; the change in temperature was not made until the heat released already began to de-crease. There was an increase in heat released when changed from 20 °C to 40 °C, and when changed from 40°C to 50°C, but then there was a decrease when changed from 50°C to 60°C. The highest value of about 100mW/kg was reached at 50°C (except for a peak value of 130mw/kg at 40°C). These values are even lower than the values reached by storing the wood samples at these temperatures initially, i.e. having a slow stepwise increase in temperature did not favour the heat release in the samples.

Figure 23 . The heat released from Dry Mix during slow stepwise temperature increase. ο Mn, ∆ Cu, □ Fe, ◊ Ref.

Metal parameter

Dry Mix Comparison between the series stored in 20, 50, 55 and 60 °C (Figure 24). There is an increase in the thermal power released when the samples without metal are stored in 50 and 55 °C. When metal is added, there is a small increase in the samples with copper and iron stored in 60 °C s after 30-40 days and after 70 days for the sample with copper. The peak values after 10-40 days in the

0

50

100

150

6 26 41 48 56 65 79 93 98 106 113 120 126Time (days)

mW

/kg

20 20 20 40 40 40 40 50 50 50 60 60 60

Temp (deg C)

55

50°C and 55°C series with copper and iron added are up to 140 percent higher than the peak values in the reference series, with copper giving the highest val-ues.

Pellets 8mm Comparison between the series stored in 50°C, 55°C and 60°C (Figure 25). The series stored in 50°C clearly give a higher output of thermal power. The series in 55°C and 60°C have a low and constant output of thermal power. The absolutely highest output is when copper is added to the 50°C series. Iron and manganese give approximately the same values as the reference series. The peak values after 10-40 days in the 50°C series with metal added is up to 220 percent higher than the peak value in the reference series. Copper gives the highest values.

Aged Sawdust Comparison between the series stored in 50°C and 60°C (Figure 26). In the aged sawdust, there is no increase in the series stored in 50°C; both the 50°C and the 60°C series lie on a constant low level. There is a slight increase after 70 days in the series where copper is added. The series with copper added has a higher out-put for both temperatures, but mainly for the 60°C series.

Conclusions When comparing the different materials, Aged Sawdust does not give a contribu-tion to the self-heating, whereas the Dry Mix and the Pellets 8 mm respond very similar and give high peak values of 400-600mW/kg after 10-30 days. The sam-ples stored at 50-55°C released more heat than the samples stored at 20°C and 60°C. This would mean that the microorganisms added to the samples by inocu-lating with partly degraded bark will not grow and release heat at temperatures of 20°C or 60°C. Stepwise increase in temperature did not favor the heat release in the sample Dry Mix; the heat released was even lower than when it was directly put in the differ-ent storage temperatures. The highest peak value was reached during the 40°C period; Dry Mix with added Cu released 130mW/kg. Dry Mix with Cu, without the stepwise increase in temperature, released almost 600mW/kg at 50°C and 400mW/kg at 55°C. When metal is added, there is a high increase in the peak values after 10-30 days for the series stored in 50°C and 55°C. The Dry Mix with Cu reached as high as almost 600mW/kg at 50°C, and Pellets 8mm with Cu released 500mW/kg at 50°C. There is a slight increase in the value for the 60°C series where copper is added after 70 days.

56

Figure 24 Dry Mix (a)Mn (b) Cu (c) Fe (d) ref stored at 20°C, 50°C, 55°C and 60°C.

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0100

200

300400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

(a)

(b)

(c)

(d)

57

Figure 25 Pellets 8mm (a)Mn (b) Cu (c) Fe (d) ref stored at

50°C, 55°C and 60°C.

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

400

500

600

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

(a)

(b)

(c)

(d)

58

Figure 26 Aged sawdust (a)Mn (b) Cu (c) Fe (d) ref stored at 50°C and 60°C.

-100

0

100

200

300

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

-100

0

100

200

300

-40 -20 0 20 40 60 80 100Time (days)

mW

/kg

(c)

(d)

(a)

(b)

59

Author’s contribution to presented articles I. Emission of volatile organic compounds (VOC) from Drying and

Storage of Biomass Author planned and carried out the measurements, collected the data, evalu-ated and analyzed the data and wrote the contribution to the conference. Au-thor presented the contribution as a poster at the conference.

II. The Release of Organic Compounds during Biomass Drying depends upon the Feedstock and /or Altering Drying Heating Medium

Author planned and carried out most of the measurements, collected the data, evaluated and analyzed the data and wrote the article.

III. The Release of Terpenes during Storage of Biofuels Author planned and carried out the measurements, collected the data, evalu-ated and analyzed the data and wrote the article.

IV. Boiler Operation Influence on the Emissions of Submicrometer-Sized Particles and Polycyclic Aromatic Hydrocarbons from Biomass-Fired Grate Boilers

Author carried out the PAH measurements, collected the data and wrote a minor part in the article.

V. Solid Phase Micro Extraction Fibers, Calibration for use in biofilter applications

Author developed the sampling and GC method used, contributed to the planning of the measurements, carried out a minor part of the measurements, contributed to the evaluation and wrote part of the article.

VI. Spontaneous Combustion of Biofuels caused by Microbial Activity Author planned and carried out the measurements, collected the data, evaluated and analyzed the data and wrote the contribution to the confer-ence. Author presented the contribution as an oral presentation at the con-ference.

VII. Parameter Study Dependency on Heat released from Microbial Activ-ity in the Self-ignition of Stored Biofuels

Author planned and carried out the measurements, collected the data, evalu-ated and analyzed the data and wrote the article.

60

Future work The organic emissions from biofuels at storage temperatures (I,III), approxi-mately up to 80°C, and the organic emissions from biofuel drying (I,II), tem-peratures between 100 and 200°C, are the same type of emissions, VOC were measured. The emissions from biofuel combustion (IV) (temperatures around 1000°C in the gas during combustion) are different, the PAH emissions were measured. A project (starting April 2006) where organic emissions from fires in a pellet silo will be measured might be able to detect the transition from low temperature emissions to high temperature emissions. Simultaneous measure-ments with both SPME and Tenax adsorbtion tubes will add information to the SPME calibration curves in V.

The microbial activity in the biofuels (VI), temperatures from 20°C to 60°C is compared to the chemical activity above 60°C (VII). The next step is to see how much of the activity below 60°C actually comes from the microbes and what kinds of microbes are present. And is the temperature interval connecting the biological with the chemical processes dependant on microbial growth? That is; is it necessary to have had a massive microbial growth in the samples where an extreme temperature rise due to chemical processes takes place? Or are these two different and unrelated phenomena? To seek the answer to these questions is the next step in the project, to try and link the low temperature (20°C to 60°C) calo-rimetric experiments (VI, VII and IV ) to high temperature calorimetric experi-ments performed by Wadsö (Wadsö 2006 and unpublished). The investigations VI, VII and future work will hopefully give some answers to these questions. Identification of microorganisms and sampling of organic emis-sions during self-heating (SPME and other methods) will provide useful infor-mation. The information will be used as input data in modeling and simulating self-ignition.

61

Acknowledgements Värmeforsk (Research and Development of the Thermo Technology), the KK foundation, The Bioenergy group and CECOST Research School are gratefully acknowledged for the financial support. I would like to express my gratitude to my supervisor, Professor Mehri Sanati, for her support, enthusiasm and guidance. You are an inspiration! Michael S., it used to be just the two of us, thank you for the support and com-pany all those years, now we are not alone anymore! Ulrika W. for all the support, I hope I’ve guided you into “our” world; you have certainly made my life easier! Björn Z., my “other” mentor, I almost want to take up smoking just to be able to see you more! Special thanks to my chemistry teachers Stig S. (high-school), Susanne W., Åke N. and Håkan A. All the colleagues at TD, in both D and M house! You are all so nice, I just have to ask and you do everything for me! Micael C. and Stefan J., my computer care-takers, you are the best! All the people that have made my life in the lab more fun, especially Peter B. and Bagher, and Jan-Olof F. - don’t you dare to leave! Jan O. for looking after Matilda when I gave my lectures! Gunnar S., it’s entirely your fault I’m here! You can keep your job (for now!). Lars W. for introducing me into the world of calorimetry, and for being such a pleasure to work with! Ulrika K. for not abandoning me when you left me! Are you coming back? Linda H. for housing and feeding me in Lund, all those interesting discussions and keeping me company! Carina N. for discussions about chemistry and life! Ann N. for listening to me complaining and supporting me (hours and hours on the phone)! My sister Nena and her husband Sam, for putting up with me and supporting me! My parents- in- law, Jörgen and Inga-Maj for being there! My husband Pelle and my daughters Amanda and Matilda, you mean the world to me! Mom and Dad, I wish you were here!

62

References AFS 2000:3, (2000) Hygieniska gränsvärden och åtgärder mot luftföroreningar. Arbetar-

skyddsstyrelsen. ISBN 91-7930-357-9 Andersson, E. (2001) Colonisation and PAH degradation by wood-rotting fungi in con-

taminated soil. Ph. D. Thesis, Department of Biotechnology, Lund University 2001. ISBN: 91-89627-00-8

Arthur C.L., Pawliszyn J. (1990) Solid phase microextraction with thermal desorption us-ing fused silica optical fibers, Anal Chem 62 (1990) 2145-2148.

Augusto F., Koziel J., Pawlizsyn J. (2001) Design and Validation of Portable SPME De-vices for Rapid Field Air Sampling and Diffusion-Based Calibration, Anal. Chem. 73 (2001) 481-486.

Bagherpour M.B., Nikazar M., Welander U., Bonakdarpour B., Sanati M. (2005) Effects of irrigation and water content of packings on alpha-pinene vapours biofiltration per-formance, Biochem. Eng. J. 24 (2005) 185-193.

Bain R L, Overend R P, Craig K R. (1998) Biomass-fired power generation. Fuel Process-ing Technology 1998; 54: 1-16.

Barry AO, Corneau D. (1999) Volatile Organic Chemicals Emissions from OCB as a Function of Processing Parameters. Holzforschung 1999, 53, 441-446.

Bartelt R.J. (1997) Calibration of a Commercial Solid-Phase Microextraction Device for Measuring Headspace Concentrations of Organic Volatiles, Anal Chem 69 (1997) 364-372.

Bengtsson, P., Sanati, M. (2004) Evaluation of hydrocarbon emissions from heart- and sapwood of Scots pine using a laboraory-scale wood dryer. Holzforschung Vol. 58, pp. 660-665, 2004.

Birgerson B, Sterner O, Zimerson E. (1985) Kemiska hälsorisker –Toxikologi i kemiskt perspektiv, 1985. ISBN 91-23-01731-7

Bjurman J, Wadsö L. (2000) Microcalorimetric measurements of metabolic activity of six decay fungi on spruce wood as a function of temperature. Mycologia 2000; 92(1): 23-28.

Björk H. (2001) From laboratory to system-studies of superheated steam drying of bio-fuel. Dissertation, Chalmers University of Technology, Sweden, 2001, ISBN 91-7291-078-x.

Bohlin, F., Vinterbäck, J., Wisniewski, J. Wisniewski, J. (1998) Solid biofuels for carbon dioxide mitigation. Biomass and Bioenergy 1998, 15(4-5), 277-281.

Borg-Karlson A-K, Lindström M, Norin T, Persson M. (1993) Enantiomeric composition of Monoterpene Hydrocarbons in different tissues of Norway Spruce, Picea abies (L.) Karst. A Multi-dimensional Gas Chromatography Study. Acta Chemica Sacaninavia 1993; 47: 138-144.

Boyle R. R., McLean S., Brandon S., Pass G. J., Davies N. W. (2002) Application of solid-phase microextraction to the quantitative analysis of 1,8-cineole in blood and ex-pired air in a Eucalyptus herbivore, the brushtail possum (Trichosurus vulpecula), J. of Chromatograph. B 780 (2002), 397-406.

Brammer, J.G, Bridgwater A.V. (2002) The influence of feedstock drying on the perform-ance and economics of a biomass gasifier-engine CHP system. Biomass and Bioenergy 22 (2002) 271-281.

Brown W H, Foote C S. (1998) Organic Chemistry 2nd ,1998. ISBN: 0-03-020458-5

63

Chai M., Pawliszyn J. (1995) Analysis of Environmental Air Samples by Solid-Phase Mi-croextraction and Gas Chromatography/Ion Trap Mass Spectrometry, Environ. Sci. Technol. 29 (1995) 693-701.

Cohen Y. (2001) Biofiltration - the treatment of fluids by microorganisms immobilized into the filter bedding material: a review, Bio resource technology, 77 (2001) 257-274.

Corchnoy S.B.,.Atkinson R. (1990) Kinetics of the gas-phase reactions of hydroxyl and nitrogen oxide (NO3) radicals with 2-carene, 1,8-cineole, p-cymene, and terpinolene Environ. Sci. Techno, 24 (1990) 1497-1502.

Cronn D.R., Truitt, S. G., Campbell, M.J. (1983) Chemical characterization of plywood veneer dryer emissions. Atmospheric Environment 1983, 17(2), 201-221.

Czerwinski, J., Zygmunt, B., Namiesnik, J. (1996) Head-space solid phase microextrac-tion for the GC-MS analysis of terpenoids in herb based formulations Fresenius J. Anal. Chem. 1996, 356, 80-83.

Dahlqvist M, Palmberg L, Bergström B, Ekholm U, Eriksson K, Figler B, Larsson B-M, Malmberg P, Müller-Suur C, Siljerrud S, Sundblad B-M, Ulfvarson U, Zhiping W. (1996a) Akut exponering av luftföroreningar i sågverk hos friska försökspersoner. Ar-betslivsinstitutet, arbete och hälsa 1996. ISBN: 91-7045-376-4.

Dahlqvist M, Palmberg L, Palmberg P, Sundblad B-M, Ulfvarson U, Zhiping W. (1996b) Acute effects of exposure to air contaminants in a sawmill on healthy volunteers. Oc-cupational and Environmental Medicine 1996;53: 586-590.

Dhamwichukom S., Kleinheinz G.T., Bagley SW.T. (2001) Thermophilic biofiltration of methanol and a-pinene, J. of Industrial Micro. & Biotech. 26 (2001) 127-133.

Demers P A, Teschke K, Davies H W, Kennedy S M, Leung V. Exposure to Dust, Resin acids, and Monoterpenes in softwood Lumber Mills. AIHAJ 2000, 61: 521-528.

Demers P A, Teschke K, Kennedy SM. (1997) What to do About softwood? A Review of Respiratory Effects and Recommendations Regarding Exposure Limits. Am J Ind Med 1997; 31: 385-398

Eisert, R., Pawliszyn, J. (1997) Design of automated solid-phase microextraction for trace analysis of organic compounds in aqueous samples. J. Chromatogr. A 1997, 293-303

Eriksson KA, Levin JO, Sandström T, Lindström-Espeling K, Lindén G, Stjernberg NL.(1997) Terpene exposure and respiratory effects among workers in Swedish join-ery shops. Scand J Work Environ Health 1997; 23(2): 114-120.

Fagernäs L. (1993) Formation and behavior of organic compounds in biomass dryers. Bioresource Technology 1993, 46, 71-76

Falk-Filipsson A. (1995) Toxokinetict and acute effects of inhalation exposure to monoterpenes in man. Arbetsmiljöinstitutet, Arbete och Hälsa 1995:3. ISBN: 91-7045-294-6.

Fäldt J. (2000) Volatile constituents in conifers and conifer-related wood-decaying fungi. Dissertation 2000. ISBN 91-7170-608-9

Fäldt J, Sjödin K, Persson M, Valterova I, Borg-Karlson A-K. (2001) Correlations be-tween selected Monoterpene Hydrocarbons in the xylem of six Pinus (Pinaceae) spe-cies. Chemoecology 2001; 11: 97-106.

Gevao B, Jones K C. (1998) Kinetics and potential significance of polycyclic aromatic hydrocarbon desorption from creosote. Environ. Sci. Technol. 1998; 32: 640-646.

Gislerud O. (1990) Drying and Storing of Comminuted Wood Fuels. Biomass 1990; 22: 229-244.

Gorlo D., Wolska L., Zygmunt B., Namiesnik J. (1997) Calibration procedure for solid phase microextraction--gas chromatographic analysis of organic vapours in air, Ta-lanta, 44 (1997) 1543-1550.

Granström K. (2002) Emissions of volatile organic compounds during drying of wood. Licentiate dissertation 2002, Karlstad University, Karlstad, Sweden, ISSN 1403-8099.

64

Gray, B. (2002) Spontaneous Combustion and self-heating, in SFPE Handbook of Fire Protection Engineering, 3rd ed. ((P.J. DiNenno et.al. , eds), 2-211-2-228, 2002.

Hakkila, P. (1989) Utilization of Residual Forest Biomass. Springer-Verlag 1989, ISBN: 0-387-50299-8

Hektor B.(1998) Cost effectiveness of measures for the reduction of net accumulation of carbon dioxide in the athmosphere. Biomass Bioenergy 1998: 15(4-5); 299-309.

Hillring B.(1998) National strategies for stimulating the use of bioenergy: policy instruments in Sweden. Biomass and Bioenergy 1998; 14(5),425-437.

Hillring B. (1999) The Swedish wood fuel market. Renewable Energy, 1999; 16: 1031-1036.

Hillring B. (2000) The impact of legislation and policy instruments on the utilisation of wood fuels. Ecological Engineering, 2000; 16: S17-S23

Hinds W C. (1999) Aerosole Technology 2nd, 1999. ISBN: 0-471-19410-7 Hogland W., Marques M. (2003) Physical, biological and chemical processes during stor-

age and spontaneous combustion of waste fuel, Resources Conservation and Recycling 2003; 40: 53-69

IEA Bioenergy (2005) Annual report 2005, www.ieabioenrgy.com Jensen B, Wolkoff P. (1996) VOC odour threshold, 1996. National Institute of occupa-

tional health, Denmark. Jirjis R. (1995) Storage and drying of wood fuel. Biomass and Bioenergy, 1995; 9: 181-

190. Jirjis R. (2005) Effects of particle size and pile height on storage and fuel quality of com-

minuted Salix viminalis. Biomass and Bioenergy, 2005; 28: 193-201. Karlberg A-T, Lindell B. (1993) Nordiska Expertgruppen för gränsvärdesdokumentation,

107 Limonen. Arbetsmiljöinstitutet, Arbete och Hälsa 1993:14. Kasanen J-K, Pasanen A-L, Pasanen P. (1999) Evaluation of sensory irritation of Δ3-

Carene and Turpentine, and acceptable levels of monoterpenes in occupational and in-door environment. Journal of Toxicology and Environmental Health Part A 1999; 56: 89-114.

Kitani, O., Hall, C.W. (1989) Biomass handbook. Gordon and Breach Science 1989. ISBN: 2-88124-269-3

Koziel, J. Jia, M., Pawliszyn, J. (2000) Air Sampling with Porous Solid-Phase Microex-traction Fibers. Anal. Chem. 2000,72, 5178-5186.

Koziel J.A, Novak I. (2002) Sampling and sample-preparation strategies based on solid-phase microextraction for analysis of indoor air, Trends in analytical chemistry 21(12) (2002) 840-850.

Kozolowski, T T, Pallardy, S G. (1997) Physiology of woody plants, 2nd. 1997 ISBN:0-12-424162-X

Kubler, H. (1987) Heat Generating Processes as Cause of Spontaneous Ignition in Forest Products. Forest Products Abstract Vol.10 No. 11, 1987, 299-327

Lehtinkangas P. (2000) Storage effects on pelletised sawdust, logging residues and bark. Biomass and Bioenergy, 2000; 19: 287-293

Madigan, M.T, Martinko, J.M., Parker, J. (2000) Brock, Biology of Microorganisms, 2000. ISBN: 0-13-081922-0

Maheshwari, R., Bharadwaj, G., Iwara, K.B.M. (2000) Thermophilic fungi: Their Physi-ology and Enzymes. Microbiology and Molecular Biology reviews, 2000, Sept, 461-488.

Martos P.A., Pawliszyn J. (1997) Calibration of Solid Phase Microextraction for Air Analyses Based on Physical Chemical Properties of the Coating , Anal. Chem. 69 (1997) 206-215.

65

Martos P.A., Saraullo A., Pawliszyn J. (1997) Estimation of Air/Coating Distribution Co-efficients for Solid Phase Microextraction Using Retention Indexes from Linear Tem-perature-Programmed Capillary Gas Chromatography. Application to the Sampling and Analysis of Total Petroleum Hydrocarbons in Air, Anal. Chem. 69 (1997) 402-408.

McDonald G, Gifford J S, Dare P H, Steward D. (1999) Characterisation of the conden-sate generated from vacuum-drying of radiata pine wood. Holz als Roh- und Werkstoff 1999; 57: 251-258.

McDonald J D, Zielinska B, Fujita E M, Sagebiel J C, Chow J C, Watson J G. (2000) Fine Particle and Gaseous Emission Rates from Residential Wood Combus-tion.Environ. Sci. Technol. 2000; 34: 2080-2091.

Misra G., Pavlostathis S.G., Perdue E.M., Araujo R. (1996) Aerobic biodegradation of selected monoterpenes Appl Microbiol Biotechnol 45 (1996) 831-838.

Monn, C. (2001) Exposure assessment of air pollutants: a review on spatial heterogeneity and indoor/outdoor/personal exposure to suspended particulate matter, nitrogen diox-ide and ozone. Atmospheric environment 35, s 1-32

Mohseni M., Allen D. G. (2000) Biofiltration of mixtures of hydrophilic and hydrophobic volatile organic compounds - The fugacity approach, Chem. Eng. Sci., 55 (2000) 1545-1558.

Namiesnik J., Zygmunt B., Jastrzebska A. (2000) Application of solid-phase microextrac-tion for determination of organic vapours in gaseous matrices, J. of Chromatography A 885 (2000) 405-418.

Nurmi J. (1995) The effect of whole tree storage on the fuelwood properties of short-rotation Salix crops. Biomass and Bioenergy, 1995,8: 245-249.

Nurmi J. (1999) The storage of logging residue for fuel. Biomass and Bioenergy, 1999; 17: 41-47.

Nylinder N, Thörnqvist T. (1980) Storing of branches and tops of different fractions. A study carried out at Åtvidaberg Trävaru AB. Swedish University of Agricultural Sci-ences, Department of Forest Products, Report No 113, 1980. ISBN: 91-576-0437-1 (In Swedish with English summary.)

Otwell, L. P., Hittmeier, M. E., Hooda, U., Yan, H., Su, W., Banerjee, S. (2000) HAPs re-lease from wood drying. Environmental Science and Technology 2000, 34, 2280-2283

Obernberger I. (1998) Decentralized biomass combustion: state of the art and future de-velopment. Biomass Bioenergy, 1998; 14 (1): 33-56.

Pawliszyn J., Liu S. (1987) Sample introduction for capillary gas chromatography with laser desorption and optical fibers, Anal.Chem 59 (1987) 1475-1478.

Pawliszyn, J. Solid Phase Microextraction : theory and practice, 1997. ISBN: 0-471-19034-9

Pawliszyn, J. (1999) Applications of solid phase microextraction, 1999. ISBN: 0-85404-525-2

Petersson G. (1995) Kemisk miljövetenskap, 1995 2nd. ISBN 91-7197-071-1 Potter, D.W., Pawliszyn, J. (1994) Rapid determination of polyaromatic hydrocarbons and

polychlorinated biphenyls in water using solid-phase microextraction and GC/MS. En-viron. Sci. Technol. 1994,28, 298-305

Richardson, J, Björheden, R., Hakkila, P., Lowe, A.T., Smith, C.T. (2002) Bioenergy from Sustainable forestry, guiding principles and practice. Kluwer 2002, ISBN: 1-4020-0676-4

Riggle D. (1996) Controlling and preventing fires at compost facilities. BioCycle 1996; 37(5): 58-64

Rösch C, Kaltschmitt M. (1999) Energy from biomass- do non-technical barriers prevent an increased use? Biomass bioenergy 1999; 6: 347-356.

66

Rosenberg C, Ruonakangas A, Liukkonen T, Welling I, Jäppinen P. (1999) Exposure to Monoterpenes in Finnish Sawmills. American Journal of Industrial Medicine Supple-ment 1999, 1: 149-151.

Sanderson M A,. Egg RP, Wiselogel AE.(1997) Biomass losses during harvest and stor-age of switchgrass. Biomass and Bioenergy, 1997; 12: 107-114.

Schneider B, Kaltschmitt M. (2000) Heat supply from woody biomass- an economic analysis. Ecological engineerging 2000; 16: 123-135.

Simeray, J., Mandin, D., Chaumont, J.P. (1997) An Aeromycological Study of sawmills: Effects of type of Installation and Timber on Mycoflora and inhalation Hazards for workers. International Biodeterioration & Biodegradation, 1997, 40(1)11-17.

Sjödin K, Persson M, Fäldt J, Ekberg I, Borg-Karlson A-K. (2000) Occurrence and Corre-lations of Monoterpene Hydrocarbon Enantiomeres in Pinus sylvestris and Picea abies. Journal of Chemical Ecology, 2000; 26(7): 1701-1720.

SMHI (2000) Årets väder 2000, Väder och Vatten 13, ISSN 0281-9619 SMHI (2001) Årets väder 2001, Väder och Vatten 13, ISSN 0281-9619 Steffen, A., Pawliszyn, J. (1996) Analysis of Flavor Volatiles Using Headspace Solid-

Phase Microextraction J. Agric. Food Chem. 1996, 44, 2187-2193. Strömvall A-M, Petersson G. (1993a) Monoterpenes emitted to air from industrial barking

of Scandinavian Conifers. Environmental Pollution 1993; 79: 215-218. Strömvall A-M, Petersson G. (1993b) Photooxidant-forming monoterpenes in air plumes

from kraft pulp industries. Environmental Pollution 1993; 79: 219-223. Svedberg U, Galle B. (2000) Assessment of Terpene Levels and Workers’ exposure in

sawmills with Long Path FTIR. Applied Occupational and Environmental Hygiene 2000; 15(9): 686-694.

Swedish Energy Administration (2005) www.stem.se, Energiläget 2005. Swedish Standard SS 18 71 06 www.sis.se Swedish Standard SS 18 71 70 www.sis.se Teschke K, Demers P A, Davies H W, Kennedy S M, Marion S A, Leung V. (1999) De-

terminants of Exposure to Inhalable Particulate, wood Dust, Resin Acids, and Monoterpenes in a Lumber Mill Environment. Ann. Occup. Hyg. 1999; 43(4): 247-255.

Thörnqvist T. (1982) The importance of cover and air under-base with storage of fuel chips. The Swedish University of Agricultural Sciences, Department of Forest Prod-ucts, report No. 127, 1982. ISBN 91-576-1119-X (In Swedish with English summary.)

Thörnqvist T. (1983) Fuel chips change during one year of storage. Swedish University of Agricultural Sciences, Department of Forest Products, report No. 148, 1983. (In Swed-ish with English summary.)

Thörnqvist T. (1985) Drying and storage of Forest Residues for Energy Production. Bio-mass, 1985; 7: 125-134.

Thörnqvist T. (1986) Large-scale seasonal storage of tree fuel project- summary of phase 1. The Swedish University of Agricultural Sciences, Department of Forest Products, report No. 188, 1986. ISBN 91-576-2913-7 (In Swedish with English summary.)

Thörnqvist T. (1987) Spontaneous combustion in piles with comminuted fuel. Swedish University of Agricultural Sciences, Department of Forest Products, report No. 163, 1987. (In Swedish with English summary.)

Thörnqvist T., Jirjis R. (1990) Changes in fuel chips during storage in large piles. Swedish University of Agricultural Sciences, Department of Forest Products, report No. 219, 1990. (In Swedish with English summary.)

Tuomela M., Vikman M., Hatakka A., Itävaara M. (2000) Biodegradation of lignin in a compost environment: a review. Bioresource Technology 72, s 169-183.

67

Wadsö, L. (2006) Measuring chemical heat production rates of biofuels by isothermal calorimetry for hazardous evaluation modelling. 2006. Submitted to Fire and Materials

Wadsö, L. (2006) Unpublished material, 2006. Wu J, Milota M R. (1999) Effect of temperature and humidity on total hydrocarbon emis-

sions from Douglas-fir lumber. Forest Prod. J. 1999, 49(6), 52-60. Warfinge P. (1997) Miljökemi, Miljövetenskap i biogeokemiskt perspektiv. 1997(1.5)

ISBN 91-88558-18-5 Wimmerstedt R. (1999) Recent advances in biofuel drying. Chem. Eng. Process. 1999;

38: 441-447. Zhang Z, Pawliszyn J. (1993) Headspace solid-phase microextraction Anal.Chem 65

(1993) 1843-1852.