Emissions from Tyre Fires

(1)

Emissions from Tyre Fires

SP Fire Technology SP REPORT 2005:43

SP Swedish National T

(2)

(3)

(4)

Abstract

To assess the emissions to air and water from a fire in tyres, four medium-scale fire experiments were performed. Each test involved 32 car tyres. Two different storage set-ups were used. In two of the tests, water or water with foam was applied. Extensive gas sampling and gas analyses were performed during the tests. This included analyses for inorganic gases, volatile organic compounds (VOC), polycyclic aromatic hydrocarbons (PAH), polychlorinated dibenzodioxins and furans (PCDD/ PCDF), particles, and metals. After the tests with water application, the run-off water was analysed for the same types of species, and also some common water quality parameters. The report presents both time-resolved and integrated results from these analyses. The application of water affected the result in that way that the yield of organic species in the fire gases increased. The yield of particles, however, decreased. The yields of VOC, PAH, and PCDD/PCDF in the run-off water increased significantly with the addition of foam solution compared to the case with only water.

Key words: tyre fire, gas analyses, extinguishing water, PCDD/F, PAH, metals,

SP Sveriges Provnings- och SP Swedish National Testing and Forskningsinstitut Research Institute

SP Rapport 2005:43 SP Report 2005:43 ISBN 91-85303-75-5 ISSN 0284-5172 Borås 2005 Postal address: Box 857,

SE-501 15 BORÅS, Sweden

Telephone: +46 33 16 50 00 Telefax: +46 33 13 55 02

E-mail: info@sp.se

(5)

Preface

This work was sponsored by the Swedish Rescue Services Agency (Räddningsverket, HNR 55).

The technicians at SP Fire Technology are acknowledged for their efficient manner of performing the fire tests. Thanks to Lars Rosell, Kristofer Gustafsson, and Stefan Österberg who assisted with the gas sampling and measurements. The City of Borås, Svensk Däckåtervinning AB, and RagnSells assisted both with information and help to obtain the tyres used as fuel in the fire tests. For this assistance these organizations are gratefully acknowledged.

A reference group was connected to the project. The members of the reference group were:

Ann Lundström, Environmental authorities of the City of Göteborg (Miljöförvaltningen, Göteborg)

Ingela Höök, The County Administration of Västra Götaland (Länsstyrelsen i Västra Götalands län)

Måns Krook, The fire department of Malmö (Malmö Brandkår)

Niklas Johansson, Swedish Environmental Protection Agency (Naturvårdsverket) Peter Andersson, The Swedish Fire Protection Association (Svenska

brandförsvarsföreningen)

The work presented in this report was part of a research project concerning also the spread of emissions to air, soil, and water. Within this project also the following reports were produced:

Lönnermark, A., Andersson-Sköld, Y., Axelsson, J., Haeger-Eugensson, M., Palm Cousins, A., Rosén, B., and Stripple, H., "Emissioner från bränder - Metoder, modeller och mätningar", Räddningsverket, P20-470/07, Karlstad, 2007 (in Swedish).

Lönnermark, A., and Blomqvist, P., "Emissions from Fires in Electrical and Electronics Waste", SP Swedish National Testing and Research Institute, SP REPORT 2005:42, Borås, Sweden, 2005.

Lönnermark, A., Stripple, H., and Blomqvist, P., "Modellering av emissioner från bränder", SP Sveriges Provnings- och Forskningsinstitut, SP Rapport 2006:53, Borås, 2006 (in Swedish).

Haeger-Eugensson, M., Tang, L., Chen, D., Axelsson, J., and Lönnermark, A., "Spridning till luft från bränder", IVL Svenska Miljöinstitutet, IVL rapport B-1702, Göteborg, 2006 (in Swedish).

Rosén, B., Andersson-Sköld, Y., and Starzec, P., "Emissioner från bränder - Spridning till mark och vatten", Statens geotekniska institut, SGI Varia nr 568, Linköping, 2006 (in Swedish).

The Swedish Rescue Services Agency is acknowledged for their financial support and the reference group and the project members for the work during the project.

(7)

Sammanfattning

Allt fler material och produkter blir förbjudna att lägga på deponier. I stället skall de återanvändas eller återvinnas på något sätt. Detta innebär att det skapas många mellanlager där använda produkter samlas i väntan på att transporteras till nästa steg i återvinningskedjan. Sådana lager innebär en brandrisk och detta kan också innebära en hälso- och miljörisk beroende på vilka ämnen som produceras i branden. Ett exempel på en sådan produkt är bildäck.

I det aktuella arbetet har fyra brandförsök med bildäck genomförts för att analysera utsläpp i form av brandgaser och släckvatten. Varje försök innehöll 32 bildäck. Två olika uppställningar användes. Tre försök genomfördes med däck placerade i en stor hög (på samma vis i varje försök) och ett försök med däcken placerade i fyra staplar. I två av försöken med däcken placerade i hög analyserades brandgaserna under en period då uppställningen vattenbegöts. I ett av dessa försök blandades skumvätska i vattnet. De ämnen som analyser utfördes för var oorganiska gaser, flyktiga organiska ämnen (VOC), polycykliska aromatiska kolväten (PAH), polyklorerade dibensodioxiner och dibensofuraner (PCDD/PCDF), partiklar och metaller och några andra utvalda

grundämnen. Både brandgaserna och släckvattnet analyserades. Denna rapport innehåller resultaten från dessa analyser. Resultaten visar bland annat att vattenbegjutningen ökar utbytet av organiska ämnen medan utbytet av partiklar i brandgasen minskade.

Skuminblandningen ökadet kraftigt utbytet av VOC, PAH och PCDD/PCDF i släckvattnet jämfört med fallet med enbart vatten.

(8)

(9)

1 Introduction

Since 1994 there is a regulation in Sweden stating that anyone who professionally

produces, imports or sells tyres is responsible for retrieval and disposal of used tyres in an environmentally friendly way [1]. This is administrated by Svensk Däckåtervinning AB (SDAB) and performed by its entrepreneur Ragn-Sells. After the 1st of July 2002

landfilling of whole used tyres (excluding bicycle tyres and tyres larger than 1400 mm) is not allowed in Sweden [2]. According to an EU directive, after 1 July 2003, whole tyres are not allowed to be placed in landfills and after 1 July 2006, chunked or shredded tyres are not allowed to be placed in landfills [3]. Even if this means fewer tyres in landfills, there are many small and a number of large storage sites for used or shredded tyres, before the tyres are transported to the next stage in the recycling chain. This means a potential fire hazard that is of interest to study and there are examples of fires that have already occurred at such storage sites.

There have been fires in large storages of tyres, e.g. in the Rhinehart tire dump (USA) in 1983 [4], and the two large fires in Canada (Hagersville and Saint Amable) in 1990 [5]. Some of the conclusions from these fires were that they were difficult to extinguish and that emissions from open tire fires can have a serious impact on health and the

environment. The pollutions include air emissions via the fire smoke and contamination of soil and water due to the run-off water and pyrolytic oil released from the burning tyres. There are examples of measurements of emissions from real tyre fires. Reisman reports airborne emissions from three cases (Rhinehart (1983), Somerset, Wisconsin (1986), and Everest, Washington (1984)) [6]. There are, however, several difficulties in sampling and analysing fire gases from real tyre fires. Firstly, one has to have equipment and personnel available at the time of the fire. Secondly, it is difficult to sample in a representative way. The sampled concentration corresponds to a concentration at that position at that time, but the concentration can exhibit significant variation spatially. Thirdly, it is difficult to measure the plume flow and to estimate the total release of emissions. Finally, the momentary burning rate might be unknown and therefore it can be impossible to calculate yields that could be used to estimate total emissions for the actual case or for other cases.

For these reasons it is often advantageous to perform controlled fire experiments where all fire gases can be collected and analysed and where the mass loss rate can be measured. A number of test series have been performed with one or two tyres or with pieces of tyres [7-10]. These tests simulate small scale scrap tyre fires and do not necessarily correspond to a large scale tyre storage case. Furthermore, these tests did not include the effect of extinguishment on the emissions. Therefore it was considered interesting to perform medium scale tyre fire tests, within SRV project HNR 55, where both the effect of geometry of the set-up and the effect of extinguishment were studied. These tests are presented in this report.

Note that the tests presented in this report are numbered from Test T5 to Test T8. The reason for this is that the presented tests were a part of larger test series, also including tests with electrical and electronics waste. Those tests are, however, presented in a separate report [11].

(10)

2 Commodities and experimental set-up

SDAB allowed SP to use some collected tyres for research and Ragn-Sells delivered approximately 200 tyres. Most were car tyres represented among these tyres, but there were also some larger tyres and some motorcycle tyres. Whole car tyres, as similar to each other in size as possible, were selected from the delivered tyres.

Two different experimental configurations were used: heap and pile. Both set-ups represent common ways to store used tyres. There are also other types of storage, e.g. container, but since such storage was not expected to be as common as the other two, it was not used in the experimental series.

In both set-ups the tyres were placed on a steel pan, 2 m × 2m. The steel pan was placed on steel beams lying on load cells (see Figure 2.1 and Figure 2.3). The whole set-up was positioned above a concrete pan (3 m × 4 m) to collect the extinguishing water. In both set-ups 32 used rubber tyres were used. The tyres varied somewhat in size, but tyres that were as similar as possible were used. In the tests with a heap of tyres, the tyres were placed in different layers in the same way in each test (see Figure 2.2). The total weight of the tyres for each test is presented in Table 2.1. Each tyre contained two rings of steel wire. The look and weight of these varied, but as an average the weight of each ring was approximately 150 g. This means that 9 kg to 10 kg in each set-up was constituted by these non-combustible metal rings.

Load cell

(11)

Figure 2.2 The different layers of the heap set-up. Load cell Load cell Load cell Burner 1.25 1.40

Figure 2.3 Experimental set-up of the pile of tyres.

Table 2.1 Total weight of tyres in each test.

Test id Type of set-up Weight [kg]

T5 Heap 245.6

T6 Heap 247.0

T7 Heap 239.2

T8 Pile 245.9

A square gas (propane) burner (17 cm × 17 cm) was placed in the concrete pan and ignited the tyres through a hole (22 cm × 22 cm) in the bottom of the steel pan. A piece of Promatect® board ran in tracks on the outside of the bottom of the steel pan. This was used to cover the hole in the bottom after the gas burner was switched off. The heat release rate of the ignition burner was approximately 25 kW.

In test T5 the concrete pan was protected by rock wool insulation (thickness 20 mm), see Figure 2.4.

(12)

Figure 2.4 Experimental set-up in test T5.

The gas temperature was measured with type-K thermocouples at two positions along the vertical centre-line of the set-up, 0.75 m and 1 m above the bottom of the steel pan. In test T8 a third thermocouple 1.4 m above the bottom of the steel pan was added. In the tests T5 and T6 the thermocouple diameter was 0.25 mm and in the tests T7 and T8 0.5 mm. A pre-test (test T0) was performed to study the fire behaviour of a heap of tyres. In this test 24 tyres (173.4 kg) were used. Most of the analyses described in this report were not performed for test T0, but the test protocol is given in Appendix 1 and the time-resolved heat release rate in Appendix 2.

(13)

3 Experimental procedure

Four different tests were performed. In two tests water application was used, of which one was with foam solution (AR-AFFF, 3 %) added to the water. The water application is described in the result section. The test set-ups were described in the previous section. The test series was run in the following way:

Test T0: pre-test; 24 tyres in a heap; water application

Test T5: 32 tyres in a heap; no water application

Test T6: 32 tyres in a heap; water application

Test T7: 32 tyres in a heap; water/foam application

Test T8: 32 tyres in four piles (8 in each pile); no water application

Each test started with background measurements of the time resolved parameters for two minutes. After this period the gas burner was ignited (time zero) and was allowed to burn for two minutes. After the gas burner had been switched off the hole in the steel pan was covered by the Promatect® H board. At this time the accumulation (or non-time resolved) sampling was started, provided the test did not include water application. In experiments with water application, the accumulation sampling was started when the water application was started. The time at which the accumulation sampling was ended depended on the type of species to be analysed and on how much had been collected. The exact times are presented in Table 3.1.

Two of the tests were run with water application on the heap set-up. In one of the

extinguishing tests, foam solution was added (3 %) to the water. The time for starting the water/foam application was based on the free burning test. The aim was to start the application after the maximum heat release rate was reached, but before the HRR had decreased too much.

Table 3.1 Events during the fire tests.

T5 T6 T7 T8 Burner on 0:00 0:00 0:00 0:00 Burner off 2:00 2:00 2:00 2:00 Soot sampling on 2:00 12:00 12:00 2:00 TENAX on 2:00 12:20 12:00 2:00 Dioxins on 2:00 12:00 12:00 2:00 Hg on 2:00 - - -

Soot sampling off 36:00 40:00 40:00 40:00

TENAX off 22:00 36:20 36:00 22:00

Dioxins off 36:00 40:30 40:30 36:00

Hg off 22:00 - - -

Between each test a propane gas burner (2.5 to 3 MW) was used to heat the system during ten minutes to avoid memory effects from the previous test.

(14)

4 Measurements, sampling and analysis

methods

4.1 Gas analysis

4.1.1 Collection of smoke gases and HRR measurement

The fire gases were collected by a hood connected to SP’s industry calorimeter [12, 13], and guided through a duct where the flow rate and gas temperature were measured and the gases were analysed. The flow rate was measured using a bidirectional probe [14].

Gas sample Gas analysis Gas analysis Data logger 8 m 6 m Ø 1 m Measurement station a) 2.05 0.47 0.73 0.27 0.145 0.28 0.535 ELPI Dioxins, PAH Particles total FTIR, Tenax O₂, CO, CO₂ Hg p,T Laser Gas flow b)

Figure 4.1 Experimental set-up in the SP fire hall with a) the hood system and b) details on the measurement station. (Dimensions in m)

The industry calorimeter is used to measure the heat release rate (HRR) from fires. The method used to calculation the HRR is based on oxygen consumption [15, 16]. At the measurement station in the duct the concentrations of O2, CO2, and CO are measured for

(15)

calculation of HRR. Further, at this station the extra probes, for other analyses as described in Section 4.1.2, were added.

4.1.2 Analysis of the smoke gases

The smoke gases were analysed for inorganic gases (e.g. CO, HCl etc.), volatile organic compounds (VOC), polycyclic aromatic hydrocarbons (PAH) and polychlorinated dibenzodioxins/furans (PCDD/PCDF). The amount and the size distribution of particles in the smoke gases and the metals on the particles were also analysed.

Many smaller species can be analysed using FTIR spectroscopy, which gives time resolved concentration information. This method was used for CO2, CO, HCl, HBr, HF,

SO2, NH3, HCN and NOx (=NO+NO2). The instrument used was a BOMEM MB 100

with a 0.922 L gas cell with a path-length of 4.8 m. The spectral resolution of the instrument was 4 cm−1. Four averaged spectra were collected per minute. Further details of this instrument have been presented elsewhere [17]. The applicability of the FTIR technique for measurements of fire gases has previously been examined in a European research project [18] and proved to be valid.

Measurements of VOC species were conducted by adsorption of fire gases on Tenax (~200 mg) adsorbent tubes (Perkin Elmer). Two adsorbent lines in parallel were used, with different sampling flows. The sampling flows were both low with this method, generally below 100 mL/min. To minimize the risk for losses in sampling, each line had a backup sampling tube. The definition of VOC species using this method includes a range of non-polar or slightly polar small-medium sized hydrocarbon species with a molecular weight of approximately 75–200 amu. The adsorbents were subsequently analysed using thermal desorption and high resolution gas chromatography (HRGC). The GC column was split for both FID (flame ionisation detector) and MS (mass selective detector) detection. Individual species were identified from the MS data, and quantified from the FID data.

The higher molecular weight hydrocarbons (PAH, PCDD/ PCDF, etc.) were collected using a large sampling-volume system. The sampling flow normally collected with this system was 20 L/min. The sampling system consisted of a heated fibreglass filter, a water-cooled condenser with a condensate bottle, and a large adsorbent cartridge containing XAD-2 (~50 g). Further, all parts of the sampling system in contact with sample gases were made of quartz glass, and thoroughly cleaned at the chemical

laboratory prior to use. Organic species with a high boiling point tend to be adsorbed on particles in smoke gases. Hence, it is important that the sampling method used collects a representative sample concerning particle size distribution. This was considered in the tests presented here. The diameter of the sample-probe tip was, together with the

sampling flow, adjusted to obtain a sampling speed in the orifice of the tip that was equal to that of the gas speed in the fire gas duct. In this manner iso-kinetic sampling was achieved.

The analysis methodology used for the larger hydrocarbons was in all cases based on high-resolution gas chromatography (HRGC) and mass fragmentography. The

quantification procedure was carried out using internal standards, which implies that the results are compensated for losses due to sample preparation. PAH: The samples were prepared by using modified US EPA3580 “Waste dilution” method. The determination of PAH was performed using the modified US EPA 8270. Twenty-one different PAH species (128–300 amu) were determined using this method. PCDD/ PCDF: Six specific PCDD isomers and nine specific PCDF isomers were quantified from their MS

(16)

was determined from the specific mass number. Extraction and analysis was conducted in accordance with EN 1948.

The real-time particle production and particle size distribution were measured using an ELPI (electrical low-pressure impactor). In the ELPI, the particles are charged using a Corona charger before they reached the low-pressure impactor, which has a number of electrically isolated collection stages. The total amount of particles was determined by weighing the amount collected on a quartz fibre filter. The collected particulate matter was analysed for metals and halogens. The results were also compared with the measured extinction of a laser beam by the smoke in the exhaust tube.

4.2 Analysis of extinguishing water

In the tests with extinguishment, representative samples were taken from the collected water. Metals and some other elements were analysed using ICP-MS screening analysis. The water samples were analysed for volatile organic compounds, VOC, and less volatile organics, semi-VOC. The samples were extracted in two different ways: a) For analysis of semi-VOC, a specific amount of the sample was extracted with hexane. The extract was analysed with GC-MS. b) For analysis of VOC, part of the sample was heated in an airtight ampule. Samples were taken from the gas phase and analysed with GC-MS. PAHs were analysed using GC/MS. Dioxins and furans were analysed according to method SS-EN-1948-2/3:1996 (final analyses performed with gas chromatography and high resolution mass spectroscopy (HRGC/HRMS)).

4.3 Analysis of fire debris

Included in the main analyses was analysis of the fire debris of the tests T1 and T4 for the content of PCDD/PCDF. These were analysed according to method

SS-EN-1948-2/3:1996 (final analyses performed with gas chromatography and high resolution mass spectroscopy (HRGC/HRMS)). Later the fire debris for all tests were analysed and then (in addition to PCDD/PCDF) also PAH, and metals were included in the analyses. The results from these analyses are reported elsewhere [19].

(17)

5 Results

5.1 Heat release rate and temperature

The time resolved results for HRR and gas temperatures are given in Appendix 2. The maximum HRR and peak temperatures are summarized in Table 5.1.

Table 5.1 Maximum HRR and maximum gas temperatures at different heightsa). The heights were measured from the bottom of the steel pan.

Test HRRmax [kW] Tmax,75cm [ºC] Tmax,100cm [ºC] Tmax,140cm [ºC] T5 3722 1246 1292 - T6 3609 1318 1363 - T7 3686 1275 1141 - T8 3607 1072 1231 1057

a) Observe that the height of the set-up was different in T8 from the other tests.

5.2 Mass loss

The mass loss was registered by load cells during the tests and for the tests without extinguishment (T5 and T8) the evaluation is straight forward. The consumed mass during a certain time period could be taken directly from the difference in the load cell signal. The application of water/foam complicates the situation, and average values of the heat of combustion for the tyres were used to estimate the mass loss during the time periods of water application. These estimated mass losses are connected with a larger uncertainty than the cases without water application, but they are assumed to be sufficiently accurate to give interesting information for the calculation of yields (see below). The calculated and estimated mass losses for different time periods are summarized in Table 5.2.

Table 5.2 Mass losses in kg for different time periods during the tests. The time periods (in min) in the parenthesis are measured from ignition.

Tests T5 112 (2-41) 109 (2-36) 92.1 (2-22) 110 (2-38) T6 42.5 (2-12) 75.6 (12-40)a) 73.4 (12:20-36:20)a) 66.1 (13:20-36:20)a) T7 49.5 (2-12) 66.4 (12-40)a) T8 118 (2-43) 116 (2-40) 85.3 (2-22) 113 (2-36) a) Water application.

5.3 Gas analyses

In many cases it is the concentration of different species that are of interest. However, since the concentrations in a real fire depend on a number of parameters, e.g. position, size of fire, topology, wind, etc., it is not a very good parameter to use to compare the results from different tests and different commodities. Instead, a more useful parameter is yield, defined as:

tot x x

m

Y

Δ

=

(18)

where mx is the mass produced of species x and Δmtot is the mass of combustible material

consumed. The yield can then be used to compare the production of various species from different material or different set-ups. It can also be used in emission modelling for fires.

5.3.1 Inorganic gases

The inorganic gases measured during the tests were carbon monoxide (CO), carbon dioxide (CO2) and sulfer dioxide (SO2). The concentrations of CO and CO2 were

measured with two different types of instruments: FTIR and NDIR. In Figure 5.1 a comparison is made between the CO2 concentrations in test T5 measured with the two

different instruments. The results from the two instruments were in close agreement.

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 10 20 30 40 50 60 Time (min) CO 2 (% ) NDIR FTIR

Figure 5.1 Comparison between CO2-concentration measured during T5 with

NDIR-instrument and FTIR spectrometer, respectively.

Table 5.3 Yield of inorganic gases measured with FTIR.

Test Time perioda) (min) YCO2 (g/kg) YCO (g/kg) YSO2 (g/kg) T5 2 – 41 2009 48.8 23.6 T5 2 – 36 1991 47.9 23.6 T5 2 – 22 1904 47.5 23.0 T5 2 – 38 1996 48.1 23.6 T6 2 – 12 1772 57.4 20.1 T6 12 – 40 b) 2012 57.3 24.8 T6 12:20 – 36:20b) 1991 56.1 24.3 T6 13:20 – 36:20 b) 2012 56.5 24.3 T7 2 – 12 1751 58.1 20.1 T7 12 – 40 b) 1991 53.6 25.2 T8 2 – 43 2071 58.9 24.3 T8 2 – 40 2059 58.4 24.3 T8 2 – 22 1973 60.1 24.4 T8 2 - 36 2047 58.2 24.4 a) From ignition.

(19)

In Table 5.3 the yields of CO2, CO, and SO2 are presented for different time periods. The

time periods were chosen both to correlate with sampling periods for other species (see below) and to represent periods with or without water application. Some selected time periods are also representative for the entire test. The time-resolved data from the FTIR measurements is presented in Appendix 2.

HCl was not detected by the FTIR, but was trapped and detected (as Cl–) in the soot filter before the FTIR. This has been converted into a total production of HCl and is presented in Table 5.4.

Table 5.4 Yield of chloride (presented as HCl) trapped on the soot filter before the FTIR.

Test Time perioda) (min) YHCl (g/kg) T5 2 - 41 0.41 T6 2 - 40 0.43 T7 2 - 40 0.51 T8 2 - 43 0.42 a) From ignition

5.3.2 VOC

In Table 5.5 the yield of different VOC species are presented. Sampling was performed during a specific time period in each test and the results are averages for this time period.

Table 5.5 Yield of VOC [g/kg].

Analysis T5 T6 T7 T8 Time period 2-22 12:20 – 36:20 12 - 36 2 - 22 Benzene 1.8 1.6 0.3 1.9 Toluene 0.1 0.5 0.8 0.1 Phenyl ethyn 0.1 0.1 0.2 0.1 Styrene 0.1 0.3 0.4 0.1 Phenol 0.2 0.0 0.0 0.2 Benzonitrile 0.0 0.0 0.0 0.0 Indene 0.1 0.2 0.3 0.1 Biphenyl 0.0 0.1 0.1 0.0 Total VOCa) 3.1 6.7 12.6 3.3

a) Naphthalene is reported together with the PAHs and is not included in the VOC data. Note that not all VOC species included in total VOC were identified.

The production of VOC is increased during water application, by a factor of between two and four. The relative amount of the different VOC species is also changed.

(20)

5.3.3 PAH

Selected PAHs, both carcinogenic and others, were analysed for and the results are presented as yields in Table 5.6. The results are presented graphically in Figure 5.2.

Table 5.6 Yield of PAHs [mg/kg].

Analysis T5 T6 T7 T8

Time period [min:s] 2-36 12-40:30 12-40:30 2-36

Benz(a)anthracene 23 50 41 25 Benzo(a)pyrene 27 52 43 32 Benzo(b)fluoranthene 24 39 34 25 Benzo(k)fluoranthene 31 53 45 32 Chrysene/Triphenylene 30 62 52 32 Dibenz(a,h)anthracene 3.4 6.0 5.4 3.2 Indeno(1,2,3-cd)pyrene 29 40 36 26 PAH, total carcinogenic 160 300 252 180 Acenaphtene 1.0 3.8 5.2 1.1 Acenaphthylene 100 255 252 180 Anthracene 22 58 52 27 Benzo(ghi)perylene 23 31 25 20 Phenanthrene 165 297 264 170 Fluoranthene 93 170 144 91 Fluorene 28 63 59 30 Naphthalene 320 562 613 370 Pyrene 79 148 132 77

PAH, total others 860 1600 1600 980

PAH, totalt incl

naphthalene 1000 1900 1800 1200 PAH, totalt excl

naphthalene 700 1300 1200 790 PAH 0 50 100 150 200 250 300 350 B enz( a )a n thr a cen e B enzo( a) py rene Be nzo( b )f luor an th ene B e n zo( k) fluor ant he ne C h rysene/ Tr iphe ny le ne D ib enz( a ,h )a n th racene Inde no( 1, 2, 3-cd) p yr en e Acen apht en e A cenap ht hyl ene An th ra ce n e Be nz o (gh i)p e ry lene P h e nan tr ene Fl uo ra nt hene Fl uor ene P yr ene P A H y iel d [ m g/ k g ] T5 T6 T7 T8

(21)

Background sampling was performed after the test series with electronic waste. Gases were sampled from the duct with during a time period of 25 min. The amounts of PAHs found from the background sampling were negligible compared to the amounts found in the fire tests.

The water application increases the yield of PAH with a factor of approximately two.

5.3.4 Dioxins and furans

A selected number of PCDD/PCDF congeners were analysed for. In Table 5.7 the results are presented as yields both for the individual congeners and as total toxic equivalences. The results in Table 5.7 are affected by large uncertainties and should be used with care since the PCDD/PCDF level in the background samples was of the same order of

magnitude as for the sample taken during the tests T5 to T8 (see Table 5.8). In some cases the largest amount found was in the background sample. It seem as the PCDD/PCDF level decreases during the test series indicating a contamination of the system, whose effect decreases as it is burnt away. This would in that case indicate that the separate heating of the system performed between each test was not enough to totally avoid the PCDD/PCDF on these low levels. However, the decrease of PCDD/PCDF in the tests T6 and T7 can also be due to the water applications, since the amount of particles was lower in the tests with water application (see Section 5.3.5).

Table 5.7 Yield of PCDD/PCDF [μg/kg].

Analysis T5 T6 T7 T8

Time period (min) 2 - 36 12 - 40.5 12 - 40.5 2 - 36

2378 TCDD BD BD BD BD 12378 PeCDD BD BD BD BD 123478 HxCDD BD BD BD BD 123678 HxCDD BD BD BD BD 123789 HxCDD 0.0020 BD 0.0030 BD 1234678 HpCDD 0.0029 0.0041 0.0028 0.0022 OCDD 0.0059 0.0059 0.0047 0.0068 2378 TCDF 0.0122 0.0035 0.0028 0.0015 12378 PeCDF 0.0079 0.0034 BD BD 23478 PeCDF 0.0067 0.0038 BD 0.0018 123478 HxCDF 0.0070 BD BD BD 123678 HxCDF 0.0063 0.0031 BD BD 123789 HxCDF 0.0027 BD BD BD 234678 HxCDF 0.0047 0.0024 BD BD 1234678 HpCDF 0.0093 0.0047 0.0030 0.0041 1234789 HpCDF 0.0029 0.0047 BD BD OCDF 0.0107 0.0127 0.0064 0.0051 TCDD-ekv I-TEQ Lower Bound 0.0072 0.0031 0.0006 0.0011 TCDD-ekv I-TEQ Upper Bound 0.0100 0.0073 0.0071 0.0043 TCDD-ekv Nordic 0.0093 0.0072 0.0070 0.0043 TCDD-ekv Eadon 0.0122 0.0082 0.0078 0.0046

(22)

Table 5.8 Total amount of PCDD/PCDF ”produced” [μg]. Analysis Background T5 T6 T7 T8 Time period (min) 25 min 2 - 36 12 - 40.5 12 - 40.5 2 - 36 2378 TCDD 0.150 BD BD BD BD 12378 PeCDD BD BD BD BD BD 123478 HxCDD BD BD BD BD BD 123678 HxCDD BD BD BD BD BD 123789 HxCDD BD 0.218 BD 0.199 BD 1234678 HpCDD BD 0.311 0.313 0.184 0.245 OCDD 1.346 0.645 0.449 0.311 0.766 2378 TCDF 1.421 1.321 0.265 0.184 0.174 12378 PeCDF 0.561 0.855 0.257 BD BD 23478 PeCDF 1.122 0.730 0.289 BD 123478 HxCDF 0.396 0.761 BD BD BD 123678 HxCDF 0.509 0.684 0.232 BD BD 123789 HxCDF BD 0.295 BD BD BD 234678 HxCDF 0.613 0.513 0.184 BD BD 1234678 HpCDF 0.464 1.010 0.353 0.199 0.466 1234789 HpCDF 0.247 0.319 0.353 BD BD OCDF 0.329 1.165 0.962 0.423 0.577 TCDD-ekv I-TEQ Lower Bound 0.9 0.8 0.2 0.04 0.1 TCDD-ekv I-TEQ Upper Bound 1.2 1.1 0.6 0.5 0.5 TCDD-ekv Nordic 1 1 1 0 0 TCDD-ekv Eadon 1 1 1 1 1

BD = below detection limit

5.3.5 Particles

The smoke or particle production was measured with three different methods: a manual gravimetric method, ELPI, and a laser light extinction method. These methods are somewhat different in what they measure, but the parameter measured will be called “particles” for all three methods. The total amount of particles measured with the manual gravimetric method is presented in Table 5.9, where the yield is also given. The time resolved production of particles was measured with an ELPI. The particle concentration as function of time and the particle size distribution at different times is presented in Appendix 2. In Table 5.11 the concentration values from the ELPI have been integrated over the same time period as was used for the soot sampling with the manual gravimetric method. These values can be compared to the total amount measured with the gravimetric method and presented in Table 5.9. In both cases the amount of particles is lower in the test with water application. However, the magnitudes of the results differ between the methods. To extend the comparison of the different methods, results are compared to the laser light extinction method in Table 5.11. The laser method gives the smoke production rate (SPR) in m2/s. The time-resolve results are presented in Appendix 2.These results have been used to calculate the total amount of particles (smoke) in m2. This parameter

(23)

then has to be converted into kg. The conversion factor depends on the burning material, but Choi et al. summarized conversion factors for different materials showing that for several material a value of 8000 m2/kg is a representative value [20]. This value has been used to get the calculated results presented here. The calculations of the ELPI results include an assumed density of the particles of 2 g/cm3.

The comparison in Table 5.11 indicate that the laser and the ELPI show similar values while the manual gravimetric method gives very low values in the tests T5, T6, and T7. There is, however, no indication that there was anything special happening during the sampling of particles during these tests. Therefore, it has not been possible to relate the differences to differences in measuring techniques or problems during sampling or analysis. In general the manual gravimetric method is supposed to be more accurate than the other two methods. The main conclusion that can be drawn from all three methods is that the particle concentration is lower in the tests with water application.

Table 5.9 Total amount and yield of particles measured with manual gravimetric method.

Test id Start Stop Amount on filters [mg] Concentration [mg/m3n] Total amount [g] Yield [g/kg] T5 2 36 64.5 40.6 1739 16.0 T6 12 40 28.9 22.0 799.7 10.6 T7 12 40 22.4 17.1 623.7 9.39 T8 2 40 613.3 353.6 17263 149

Table 5.10 Total amount of particles measured with the ELPIa) (g).

Test id ELPI

T5 19000 T6 10320 T7 7516 T8 16500

a) Integrated over the same time periods as for the gravimetric method (see Table 5.9). The stages

1 to 10 of the ELPI have been included in the evaluation. The D50%-value (i.e. 50 % collection

efficiency of this aerodynamic diameter for this stage) of stage 10 was 2.41 μm. The higher stages

(larger diameters) were excluded since the uncertainty is large due to a smaller amount of particles collected on these stages.

Table 5.11 Comparison of yields of particles based on three different methods (g/kg).

Test id MGMa) ELPI Laser

T5 16.0 174 162

T6 10.6 136 113

T7 9.39 113 114

T8 149 142 161

a) Manual gravimetric method.

5.3.6 Metals

The filters used to collect particles in the manual gravimetric method, were after

evaluation of total particle yield analysed for the content of selected metals. In Table 5.12 the results are presented both as concentration of metals on the particles and as amount of metal on particle per kg consumed fuel (yield). The five metals with the highest yield for

(24)

each test are highlighted (written with bold text) in the table. For all four tests zink is the major component. Other metals with high yields are nickel, chromium, lead, barium, molybdenum, cobalt, and antimony. An interesting result is the high yields of barium and lead in the test with foam application (test T7).

Table 5.12 Metals on particles (the five highest yields presented in bold).

Metal T5 T6 T7 T8 Concen-tration on particle [mg/kg particle] Yield [μg/kg fuel] Concen-tration on particle [mg/kg particle] Yield [μg/kg fuel] Concen-tration on particle [mg/kg particle] Yield [μg/kg fuel] Concen-tration on particle [mg/kg particle] Yield [μg/kg fuel] As <0.3 <5 <0.7 <7 2.46 23.1 0.65 97 Ba 79.1 1270 125 1320 714 6710 15 2240 Cd 1.35 21.6 1.25 13.2 129 1220 4 600 Co 14.4 231 41.5 440 30.8 289 19 2800 Cr 91.5 1460 72.7 770 67.0 629 85 13000 Cu 34.1 546 22.8 242 22.8 214 14 2100 Mo 20.2 322 21.5 227 19.2 180 20 3000 Ni 89.9 1440 72.7 770 58.0 545 120 18000 Pb 24.8 397 16.6 176 893 8380 23 3400 Sb 69.8 1120 16.6 176 5.36 50.3 11 1600 Se 4.03 64.5 7.27 77.0 2.63 24.7 7 1000 Tl 0.031 0.496 0.104 1.10 0.0446 0.419 0.13 19 V <2 <25 <3 <37 33.9 319 1 150 Zn 1810 29000 2910 30800 2280 21400 3900 581000 Be <0.02 <0.2 <0.04 <0.4 <0.05 <0.4 <0.005 <0.7 Ga 0.687 11.0 1.34 14.2 9.35 87.8 0.38 57 Ge <0.5 <7 <1 <11 <1 <13 0.34 51 Rb 0.131 2.09 0.231 2.45 0.474 4.46 0.19 28 Y 0.605 9.68 0.306 3.25 0.227 2.13 0.0935 13.9 Zr 1.22 19.5 2.44 25.9 3.31 31.1 0.23 34 Ag <1 <10 15.5 164 2.25 21.1 <0.01 <2 Sn 3.68 58.9 2.34 24.8 0.920 8.64 1 150 Te <0.3 <5 <0.7 <7 <0.9 <8 <0.01 <2 Ce 0.0755 1.21 0.176 1.87 0.180 1.69 0.092 14 Nd 0.0804 1.29 0.187 1.99 0.358 3.36 0.12 18 Re <0.03 <0.5 <0.07 <0.7 <0.09 <0.8 <0.005 <0.7 Au <0.3 <5 <0.7 <7 <0.9 <8 <0.01 <2 Bi <0.3 <5 <0.7 <7 <0.9 <8 0.13 19

(25)

5.4 Water application

5.4.1 Extinguishment

A water applicator with nine nozzles was used for the water application. The nozzles were positioned in three rows with three nozzles in each row. The distance between the rows and between the nozzles in each row was 45 cm. The applicator was placed so that the openings of the nozzles were situated 20 cm above the highest point of the top tyre in the heap of tyres. A total water flow of 5 L/min was used. A calibration test showed that the water density on a plane 20 cm beneath the nozzles was approximately 2 L/m2/min. The water flow rate was chosen to affect the fire but not to extinguish it. The reason for this strategy was to be able to collect extinguishing water that has been affected by the combustion. The water density is also representative of that one would expect when fighting a fire from a distance.

The fire was very difficult to extinguish with water. The heat release rate became somewhat lower than the free burning tests, but the difference was not very large. In test T6, some of the nozzles were not working properly, i.e. they did not give the correct spray. The water ended up on the fuel; it was only the total spray pattern that differed from the one in the case of normal functioning. In test T6 approximately 140 L water were used during the water applicator phase and 98 L for the manual

extinguishment. The corresponding values for T7 were 140 L and 42.3 L, respectively. The value 140 L is based on the calibration values (5 L/min). The high temperature in the beginning of the water application phase made the water vaporize in the nozzle. This fact, together with the low flow rate, made the adjustment of the flow rate difficult. A check with a total flow meter indicated that the total flow was lower, but approximately the same for both tests.

After test T6, 49 L water was collected for analyses; after test T7 20 L were collected for analyses. A large amount of water was vaporized, but some of the water was also trapped in only partly consumed tyres and this can be one of the explanations for the difference in amount of collected extinguishing water after the tests.

After the fire there were three types of debris: partly unburned tyres, metal rings from the tyres, and some kind of powder unwilling to mix with water.

5.4.2 Water analyses

Water was applied during a selected time period in the tests T6 and T7. In test T7 foam (AR-AFFF) was mixed with the water (3 % foam). The results of the analysis of the run-off water from these tests are presented in this section. Some common water parameters are presented in Table 5.13. The results from the analysis of tensides in test T7 is also included in Table 5.13. In Table 5.14, Table 5.15, Table 5.16, and Table 5.17 the results from the analyses of VOC and semi-VOC, PAH, PCDD/PCDF, and metals, respectively, are given.

(26)

Table 5.13 Analyses of various common water parameters. Analysis T6 T7 Particle matters [mg/L] 140 250 pH, 25 ºC 7.5 6.7 Conductivity, 25 ºC [mS/m] 132 189 BOD7 [mg/L] 280 2000 COD(Cr) [mg/L] 700 4600 TOC [mg/L] 160 1500 Nitrogen, total N [mg/L] 28 55 Phosfor, total P [mg/L] 0.23 0.62 AOX [μg/L] <250 <2000 EOX [μg/L] <1 3.8 Tensides, anion [mg/L] NA 210 Tensides, cation [mg/L] NA 3.0 Tensides, nonion [mg/L] NA <0.3

Table 5.14 VOC and semi-VOC in the runoff water [μg/L].

Analysis T6 T7 Chlorobenzene <1 <1 Dichlorobenzenes <2 <2 Trichlorobenzenes <2 <2 Tetrachlorobenzenes <2 <2 Pentachlorobenzene <1 <1 Hexachlorobenzenes <1 <1 Benzene <10 <10 Toluene 13.6 110 Ethylbenzene <5, traces 26 Xylenes <15, traces 54

Alkylated benzenes larger than xylene

<20 140

Nonylphenol <5 <5

Di-n-butyl phthalate <1 2.7

Benzyl butyl phthalate <1 <1

Bis(2-ethylhexyl) adipate 2.1 <1

Diethyl hexyl phthalate 4.2 25

Alkylated naphtalenes 73 640

Unpolar aliphatic hydrocarbons

600 18000 Total conc. of extractable

organic material (μg organic carbon per L)

15000 270000

Other compounds searched for were 2-chlorotoluene, 4-chlorotoluen, bromobenzene, trichloroethylene, 1,3-dichloropropane, 1,1,1-trichloroethane, bromodichloromethane, hexachlorobutadiene, 1,2-dibromo-3-chloropropane, 1,1,2,2-tetrachloroethane, dimethyl phthalate, diethyl phthalate, di-n-octyl phthalate. All of these had concentrations below 1 μg/L.

(27)

Table 5.15 PAH in the run-off water [mg/L] Analysis T6 T7 Benz(a)anthracene 0.0006 0.043 Benzo(a)pyrene 0.0006 0.039 Benzo(b)fluoranthene 0.0006 0.041 Benzo(k)fluoranthene 0.0002 0.013 Chrysene/Triphenylene 0.0007 0.045 Dibenz(a,h)anthracene <0.0001 0.0051 Indeno(1,2,3-cd)pyrene 0.0003 0.019

PAH, total carcinogenic 0.0030 0.21

Acenaphtene 0.0018 0.014 Acenaphthylene 0.0017 0.020 Anthracene 0.0007 0.025 Benzo(ghi)perylene 0.0004 0.023 Phenanthrene 0.0025 0.098 Fluoranthene 0.0017 0.094 Fluorene 0.0009 0.017 Naphtalene 0.0056 0.045 Pyrene 0.0018 0.10

PAH, total others 0.017 0.44

Table 5.16 PCDD and PCDF in the run-off water [ng/L]

Congener T6 T7 2378 TCDD 0.0045 <0.0020 12378 PeCDD <0.002 0.0088 123478 HxCDD <0.002 0.0054 123678 HxCDD <0.002 0.0049 123789 HxCDD <0.002 0.014 1234678 HpCDD 0.0020 0.043 OCDD 0.0032 0.070 2378 TCDF 0.0093 0.049 12378 PeCDF <0.0002 0.0084 23478 PeCDF 0.0022 0.010 123478 HxCDF <0.002 0.0073 123678 HxCDF <0.002 0.0073 234678 HxCDF <0.002 0.0046 123789 HxCDF 0.0025 0.019 1234678 HpCDF 0.0033 0.026 1234789 HpCDF <0.002 0.018 OCDF <0.002 0.045

TCDD-ekv I-TEQ Lower Bound 0.0068 0.022

TCDD-ekv I-TEQ Upper Bound 0.0092 0.024

TCDD-ekv Nordic 0.0091 0.024

(28)

Table 5.17 Metals in the run-off water [μg/L]. The concentration for all other elements was less than 1 μg/L. The five metals with the highest concentration in each test are presented in bold. Element T6 T7 Element T6 T7 Aluminium, Al 240 160 Lithium, Li 110 81 Antimony, Sb 58 36 Magnesium, Mg 6500 3700 Arsenic, As <1 <1 Manganese, Mn 450 360 Barium, Ba 210 96 Molybdenum, Mo <1 <1 Lead, Pb 29 37 Sodium, Na 100000 190000 Boron, B <1 <1 Neodymium, Nd <1 2 Bromine, Br 167000 48000 Nickel, Ni <1 2 Cerium, Ce <1 <1 Rubidium, Rb 92 57 Europium, Eu <1 <1 Selenium, Se <1 <1 Phosphorus, P 4 3 Silver, Ag 4 5 Iodine, I 440 177 Scandium, Sc 1 <1 Iron, Fe 8600 8800 Strontium, Sr 320 90 Cadmium, Cd <1 <1 Sulfur, S 7300 5100 Calcium, Ca 120000 41300 Tin, Sn 4 4 Potassium, K 32000 19000 Titanium, Ti 200 93 Silicon, Si 260 150 Vanadium, V <1 <1 Cobalt, Co 1000 770 Bismuth, Bi <1 2 Carbon, C 460 9000 Tungsten, W <1 2 Copper, Cu 72 75 Yttrium, Y <1 <1 Chromium, Cr <1 <1 Zinc, Zn 20000 12000 Lanthanum, La <1 <1 Zirconium, Zr <1 <1

(29)

6 Conclusions

Four tests with used tyres were performed. Car tyres as equal as possible to each other wer chosen to give as similar conditions as possible in each test. The tyres were placed in two different set-ups, in a heap and in four piles, respectively. During two of the three tests with the heap set-up, water was applied. In one of these tests foam solution was added to the water.

The experimental set-up (heap or piles) did not seem to significantly affect the results. The manual gravimetric method gave indeed much higher particle yields in test T8 (piles), but the comparison between the different methods used to determine the

production of particles showed that the ELPI and the laser method correlated better with each other than with the manual gravimetric method and also more with what could be expected. Only in test T8 the result of the manual gravimetric method correlated with the other methods. One should, however, keep in mind the differences in what the different methods measure.

The maximum HRR was relatively similar for the different experimental set-ups. However, the time to reach the maximum HRR was longer in the test with piles of tyres. The application of water increases the yields of organic species. This is the case for VOC and PAH, For particles the case is the opposite, i.e. a lower yield of particles are

measured when water is applied compared to the case without water. The reason for this is probably that the smoke is “washed” by the water. The addition of foam did not significantly affect the results from the gas analyses, maybe with the VOCs as an exception where the yield increased. In the water analyses there were, however, large differences. The yields of VOC, PAH, and PCDD/PCDF in the run-off water increased significantly with the addition of foam solution compared to the case with only water. On the particles in the fire gases zinc, nickel and chromium dominate among the metals. Other metals with high yields are lead and barium. Note that this ordering is made only from the yields with no regards to their effect on the environment or health. In the run-off water the metals sodium, calcium, potassium, and zinc were found in highest

concentration. Another element found in high concentration was bromine.

It is always interesting to compare experimental results to other test series where the conditions or scales may vary. One test series suitable for comparison is the one described by Lemiieux and Ryan where shredded tyres (5.1 cm × 5.1 cm) and tyres cut in chunks (1/6 – 1/4 tyre) [8]. When comparing their results with the results presented in this report one general conclusion is that the smaller the pieces of tyres, the higher the yields of VOC and PAH. The reason for these results is probably the compactness of the fuel set-up and thereby the availability of the oxygen. This is important to keep in mind when comparing different scales and different types of storage.

(30)

7 References

1. SFS 1994:1236, "Förordning om producentansvar för däck", 1994 (in Swedish). 2. SFS 2001:512, "Förordning om deponering av avfall", 2001 (in Swedish). 3. "Council Directive 1999/31/EC of 26 April 1999 on the landfill of waste", In

Official Journal of the European Communities, 1999.

4. Yamaguchi, E., "Emissions from Open Tire Fires",

http://www.p2pays.org/ref/11/10504/html/intro/openfire.htm, 13 October 2000.

5. "Tyre fires - a pile of polution", Fire Prevention, 241, July/August, 22-26, 1991. 6. Reisman, J. I., "Air Emissions from Scrap Tire Combustion", United States

Environmental Protections Agency, EPA-600/R-97-115, 1997.

7. Lemieux, P. M. and DeMarini, D., "Mutagenicity of Emissions from the

Simulated Open Burning of Scrap Rubber Tires", U.S. Environmental Protection Agency, Control Technology Center, office of Research and Development, EPA-600/R-92-127, 1992.

8. Lemieux, P. M. and Ryan, J. V., "Characterization of Air Pollutants Emitted from a Simulated Scrap Tire Fire", Journal of the Air & Waste Management

Association, 43, 1106-1115, 1993.

9. Lemieux, P. M., Lutes, C. C., and Santoianni, D. A., "Emissions of organic air toxics from open burning: a comprehensive review", Progress in Energy and

Combustion Science, 30, 1-32, 2004.

10. Büthe, N., "Combustion effluents of rubber",

http://www.buethe.onlinehome.de/research.htm.

11. Lönnermark, A. and Blomqvist, P., "Emissions from Fires in Electrical and Electronics Waste", SP Swedish National Testing and Research Institute, SP REPORT 2005:42, Borås, Sweden, 2005.

12. Dahlberg, M., "The SP Industry Calorimeter: For rate of heat release measurements up to 10 MW", SP Swedish National Testing and Research Institute, SP REPORT 1992:43, Borås, Sweden, 1993.

13. Dahlberg, M., "Error Analysis for Heat Release Rate Measurement with the SP Industri Calorimeter", SP Swedish National Testing and Research Institute, SP REPORT 1994:29, Borås, 1994.

14. McCaffrey, B. J. and Heskestad, G., "Brief Communications: A Robust

Bidirectional Low-Velocity Probe for Flame and Fire Application", Combustion

and Flame, 26, 125-127, 1976.

15. Huggett, C., "Estimation of Rate of Heat Release by Means of Oxygen Consumption Measurements", Fire and Materials, 4, 2, 61-65, 1980.

16. Parker, W. J., "Calculations of the Heat Release Rate by Oxygen Consumption for Varios Applications", National Bureau of Standards, NBSIR 81-2427, Gaithersburg, USA, 1982.

17. Blomqvist, P., Lindberg, P., and Månsson, M., "TOXFIRE - Fire Characteristics and Smoke Gas Analyses in Under-ventilated Large-scale Combustion

Experiments: FTIR Measurements", SP Swedish National Testing and Research Institute, SP REPORT 1996:47, Borås, Sweden, 1998.

18. Hakkarainen, T., Mikkola, E., Laperre, J., Gensous, F., Fardell, P., Le Tallec, Y., Baiocchi, C., Paul, K., Simonson, M., Deleu, C., and Metcalfe, E., "Smoke Gas Analysis by Fourier Transform Infrared Spectroscopy - Summary of the SAFIR Project Results", Fire and Materials, 24, 101-112, 2000.

19. Lönnermark, A., "Analyses of Fire Debris after Tyre Fires and Fires in Electrical and Electronics Waste", SP Swedish National Testing and Research Institute, SP REPORT 2005:44, Borås, Sweden, 2005.

20. Choi, M. Y., Mulholland, G. W., Hamins, A., and Kashiwagi, T., "Comparisons of the soot Volume Fraction Using Gravimetric and Light Extinction

(31)

Appendix 1 Test protocols

Test id: T0 – Pre test

-2:00 Measurement start

00:00 Propane burner on 02:00 Propane burner off

06:35 Small “explosion”

12:25 Small “explosion”

17:37 Water applicator on

24:49 Almost completely extinguished

25:50 Manual extinguishment

Test id: T5 – Heap of tyres

00:00 Propane burner on

01:00 Flames approximately 1 m above the steel pan. 02:00 Propane burner off

Sampling of soot, dioxins, Hg and on TENAX started 04:00 Flames approximately 3 m above the steel pan. 06:25 Flames 5-6 m above the steel pan.

07:15 Flames occasionally 7 m above the steel pan. 10:00 Flames approximately 8 m above the steel pan.

12:00 A small amount of smoke outside the collecting hood, but it seems to be sucked into the hood again.

<14:00 The wire with the thermocouples fell down. 17:00 Flames approximately 5 m above the steel pan.

Some melted tyre drips on the insulation under the front left corner. A small pool fire was formed, but was self extinguished relatively quickly.

20:00 Small “explosions” could be heard now and then. 21:00 Most of the tyres had collapsed into a thick layer.

22:00 Sampling of Hg and on TENAX stopped

24:30 Flames approximately 2 m above the steel pan. 36:00 Sampling of soot and dioxins stopped.

41:00 Manually extinguished

52:30 Reignited

63:00 Extinguished with foam

Test id: T6 – Heap of tyres, water application

00:45 Flames approximately 1 m above the steel pan. 04:00 Flames more than 3 m above the steel pan. 12:00 Propane burner off

Sampling of soot and dioxins started

Water application started. In the beginning most of the water is vaporized, but after some minutes more water is reaching the tyres

(32)

13:00 Dripping, but no pool fire is formed

19:00 The wire holding the thermocouples has burnt off

27:00 From one of the nine nozzles the water is dripping rather than spraying. 31:00 The smoke is turning white.

34:30 The fire is now controlled, but for the corners.

36:20 Sampling on TENAX stopped

40:00 Water application stopped Soot sampling stopped 40:30 Sampling of dioxins stopped

Test id: T7 – Heap of tyres, water application with foam

00:33 Flames approximately 0.75 m above the steel pan 01:10 Flames approximately 1 m above the steel pan 03:00 Flames approximately 5 m above the steel pan 04:00 Flames 5-6 m above the steel pan

06:00 Flames approximately 6 m above the steel pan; some sparks into the hood 12:00 Water/foam application starts

Sampling of soot and dioxins and on TENAX started 13:00 One tyre is hanging down on one side of the set-up 16:45 Flames approximately 4 m above the steel pan

A small pool fire (10 cm × 30 cm) in the concrete pan. 20:00 The pool fire and the hanging tyre are no longer burning 26:30 The production of white smoke is increasing.

32:00 The water applicator is moved somewhat further in since some of the foam now is landing in the concrete pan and not on the tyres

34:00 It is burning in the left front corner and in the back of the set-up

40:00 Soot sampling stopped

40:05 Water/foam application off. 40:30 Sampling of dioxins stopped

48:00 Still some smoke is coming from the set-up ELPI and FTIR are switched off.

Test id: T8 – Piles of tyres

00:20 Flames approximately 0.75 m above the steel pan 00:30 Flames approximately 1 m above the steel pan

01:30 Flames approximately 1.4 m above the steel pan (top of the set-up). 02:00 Sampling of soot and dioxins and on TENAX started

03:00 Flames 0.5 m above the top of the set-up. 04:00 Flames 1 m above the top of the set-up. 05:20 Flames 2 m above the top of the set-up. 06:15 Flames 4 m above the top of the set-up.

08:45 The flames are reaching the outer border of the set-up at the top, through the point where the piles meet each other.

(33)

12:00 The tyres are collapsing somewhat.

20:45 Everything is burning as a heap with the exception of three tyres in the front to the right which are still standing as a pile. The outer side of this pile is not burning.

24:00 The flames approximately 3 m above the steel pan, sometimes between 4 and 5 m.

26:00 The flames approximately 3 m above the steel pan

28:20 The flames start to reach the outside of the small pile (still standing) in the front right.

29:30 The flames approximately 2 m above the steel pan A small pool fire in the concrete pan.

36:00 Sampling of dioxins stopped 40:00 Sampling of soot stopped

(34)

results from measurements of heat release rate, temperature within the experimental set-up, particle concentration as function of time, particle concentration as function of aerodynamic particle diameter (at different times), smoke production rate (SPR), and gas concentrations (CO2, CO, and SO2) are given. For the preparation tests T0 only heat

release rate and smoke production rate are presented.

T0 – Pre test 0 500 1000 1500 2000 2500 3000 3500 4000 0 10 20 30 40 50 HRRtot [kW] HRRconv [kW] H ea t re le a se ra te [ kW ] Time [min] 0 50 100 150 200 0 5 10 15 20 25 30 SPR [m2/s] SP R [ m 2 /s ] Time [min] T5 – Heap of tyres 0 500 1000 1500 2000 2500 3000 3500 4000 0 10 20 30 40 50 HRRtot [kW] HRRconv [kW] H ea t re le a se ra te [ kW ] Time [min] 0 500 1000 1500 0 10 20 30 40 50 0cm [o_C] -25cm [o_C] T em per at u re [ o C] Time [min] 1 10 100 1000 104 105 106 107 108 0 10 20 30 40 50 P ar tic le c o nc en tr at io n [1 /c m 3 ] Time [min] 0.01 1 100 104 106 108 0.01 0.1 1 10 Time = 5 min Time = 10 min Time = 15 min Time = 20 min P ar tic le c o nc en tr at io n [1 /c m 3 ]

Aerodynamic particle diameter [μm]

0 50 100 150 200 0 10 20 30 40 50 SPR [m2/s] SP R [ m 2 /s ] Time [min]

(35)

0 50 100 150 200 250 0 10 20 30 40 50 Time (min) CO 2 ( g /s ) 0 2 4 6 8 10 0 10 20 30 40 50 Time (min) C O ( g /s ) 0.0 1.0 2.0 3.0 4.0 5.0 0 10 20 30 40 50 Time (min) SO 2 (g/s)

(36)

T6 – Heap of tyres, water application 0 500 1000 1500 2000 2500 3000 3500 4000 0 10 20 30 40 50 HRRtot [kW] HRRconv [kW] H ea t re le a se ra te [ kW ] Time [min] 0 500 1000 1500 0 10 20 30 40 50 0cm [o_C] -25cm [o_C] T em per at u re [ o C] Time [min] 1 10 100 1000 104 105 106 107 108 0 10 20 30 40 50 P ar tic le c o nc en tr at io n [1 /c m 3 ] Time [min] 0.01 1 100 104 106 108 0.01 0.1 1 10 Time = 5 min Time = 10 min Time = 15 min Time = 20 min P ar tic le c o nc en tr at io n [1 /c m 3 ]

0 50 100 150 200 0 10 20 30 40 50 60 SPR [m2/s] SP R [ m 2 /s ] Time [min]

(37)

(38)

T7 – Heap of tyres, water application with foam 0 500 1000 1500 2000 2500 3000 3500 4000 0 10 20 30 40 50 HRRtot [kW] HRRconv [kW] He at r e le a se r a te [k W] Time [min] 0 500 1000 1500 0 10 20 30 40 50 0cm [o_C] -25cm [o C] T e m per at ur e [ o C] Time [min] 1 10 100 1000 104 105 106 107 108 0 10 20 30 40 50 P ar tic le c o nc en tr at io n [1 /c m 3 ] Time [min] 0.01 1 100 104 106 108 0.01 0.1 1 10 Time = 5 min Time = 10 min Time = 15 min Time = 20 min P ar tic le c o nc en tr at io n [1 /c m 3 ]

0 50 100 150 200 0 10 20 30 40 50 60 SPR [m2/s] SP R [ m 2 /s ] Time [min]

(39)

(40)

T8 – Piles of tyres 0 500 1000 1500 2000 2500 3000 3500 4000 0 10 20 30 40 50 HRRtot [kW] HRRconv [kW] He at r e le a se r a te [k W] Time [min] 0 500 1000 1500 0 10 20 30 40 50 0cm [o_C] -40cm [o C] -65cm [o_C] T e m per at ur e [ o C] Time [min] 1 10 100 1000 104 105 106 107 108 0 10 20 30 40 50 P ar tic le c o nc en tr at io n [1 /c m 3 ] Time [min] 0.01 1 100 104 106 108 0.01 0.1 1 10 Time = 5 min Time = 10 min Time = 15 min Time = 20 min P ar tic le c o nc en tr at io n [1 /c m 3 ]

0 50 100 150 200 0 10 20 30 40 50 SPR [m2/s] SP R [ m 2 /s ] Time [min]

(41)

(42)

Appendix 3 Photos from the tests

Figure A3.1 Set-up in test T0. Figure A3.2 Test T0.

Figure A3.3 Test T0. Figure A3.4 Water application in test T0.

Figure A3.5 The square propane burner used for ignition.

(43)

Figure A3.7 Test T5. Figure A3.8 Fire debris after test T5.

Figure A3.9 Test T6. Figure A3.10 Test T6 with water application.

(44)

Figure A3.13 Set-up for test T7. Figure A3.14 Test T7.

(45)

(46)

SP Fire Technology SP REPORT 2005:43 ISBN 91-85303-75-5 ISSN 0284-5172

technical investigation, measurement, testing and certfi cation, we perform

research and development in close liaison with universities, institutes of technology and international partners.

SP is a EU-notifi ed body and accredited test laboratory. Our headquarters are in Borås, in the west part of Sweden.

SP Swedish National Testing and Research Institute

Box 857

SE-501 15 BORÅS, SWEDEN

Telephone: + 46 33 16 50 00, Telefax: +46 33 13 55 02 E-mail: info@sp.se, Internet: www.sp.se

Emissions from Tyre Fires