- Email: [email protected]
The presence of zinc in Swedish waste fuels
Waste Management 33 (2013) 2675–2679
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
The presence of zinc in Swedish waste fuels Frida Jones a,b,⇑, Mattias Bisaillon c, Daniel Lindberg a, Mikko Hupa a a
Åbo Akademi University, Biskopsgatan 8, FI-20500 Åbo, Finland SP Technical Research Institute of Sweden, Box 857, SE-501 15 Borås, Sweden c Profu AB, Götaforsliden 13, 431 34 Mölndal, Sweden b
a r t i c l e
i n f o
Article history: Received 20 December 2012 Accepted 19 July 2013 Available online 4 September 2013 Keywords: Zinc Solid waste Waste-to-Energy plants Fuel characterization
a b s t r a c t Zinc (Zn) is a chemical element that has gained more attention lately owing to its possibility to form corrosive deposits in large boilers, such as Waste-to-Energy plants. Zn enters the boilers in many different forms and particularly in waste, the amount of Zn is hard to determine due to both the heterogeneity of waste in general but also due to the fact that little is yet published specifically about the Zn levels in waste. This study aimed to determine the Zn in Swedish waste fuels by taking regular samples from seven different and geographically separate waste combustion plants over a 12-month period. The analysis shows that there is a relation between the municipal solid waste (MSW) content and the Zn-content; high MSW-content gives lower Zn-content. This means that waste combustion plants with a higher share of industrial and commercial waste and/or building and demolition waste would have a higher share of Zn in the fuel. The study also shows that in Sweden, the geographic location of the plant does not have any effect on the Zn-content. Furthermore, it is concluded that different seasons appear not to affect the Zn concentrations significantly. In some plants there was a clear correlation between the Zn-content and the content of other trace metals. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Zinc (Zn) enters Waste-to-Energy (WtE) plants with the different waste fuels. It is hard to determine a general form of Zn since waste mixtures are very heterogeneous. The Zn-content depends on numerous factors, such as household consumption patterns, industrial/commercial activity, season, economic climate, and possibly geographical location. In the database AvfallsAtlas,1 information from different waste fraction characterization studies have been collected and summarized, giving the possibility to show the waste fractions with the highest contribution to the share of Zn in waste. In Table 1 different waste fractions are listed with the minimum and maximum reported Zn-content in the database. The merging of different types of materials within each fraction, such as a summary of all plastic materials, also means that the origins of the waste fractions are merged. Therefore, all of the fractions presented in Table 1 include waste collected as Municipal Solid Waste (MSW), Industrial and Commercial Waste (ICW), and Building and Demolition Waste (BDW). From here on the three-letter abbreviations will be used. ⇑ Corresponding author. Address: SP Technical Research Institute of Sweden, Energy Technology, Box 857, SE-501 15 Borås, Sweden. Tel.: +46 (0) 10 516 57 67. E-mail address: [email protected] (F. Jones). 1 AvfallsAtlas (WasteAtlas) is a closed database developed by Profu AB, www.profu.se/indexaa.htm. Contact person Mattias Bisaillon ([email protected]) or Johan Sundberg ([email protected]). 0956-053X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2013.07.023
The internal variation of Zn within each fraction also indicates how difficult a general determination of Zn-content in a waste mixture would be, for example, in the Textile/leather/rubber-fraction where the Zn-content varies from around 30 to 18,000 mg/kg waste. Furthermore, although the difference between the materials in the textile/leather/rubber-fraction is obvious there is also a major difference within the glass fraction with reported measurements varying between close to 0 and 10,000 mg/kg. A study of the details behind Table 1 reveals that the materials contributing the most to the Zn-content in waste are rubber tires, glass, plastics treated with flame retardants, and PVC plastics. Zinc oxide (ZnO) is a frequently used chemical as an additive in numerous products (Klingshirn et al., 2010; Moezzi et al., 2012). The most dominant application of ZnO is within the rubber industry where it is used in many types of rubber, and has been for over a century (Depew, 1933a, 1933b, 1933c). Predominantly ZnO is used in the car industry, activating the sulfur crosslinking and also for improving the absorption of frictional heat in tires, as well as being used in belts, windshield wipers, hoses, oil seals, trim and mountings International Zinc Association-Zinc Oxide Information Centre (2012). In addition, ZnO is also used in heat resistant glass, cooking wear glass, and specialty glass, such as photochromic lenses. ZnO is also a main ingredient for the production of a flame retardant zinc borate (XZnOYB2O3) which exists in various forms, differing by the Zn/B-ratio and number of attached hydrates. Zinc borate is added to a number of materials and products, such as plastics, textiles, and rubber (Shen et al., 2008).
2676
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Table 1 Waste fractions and their Zn-content, summarized information from the database AvfallsAtlas.
a b
Waste fraction
Minimum (mg/ kg)
Cardboard/paper/corrugated cardboard Plastics Kitchen and garden waste Textile/leather/rubber Glass Wood/RWWa/CCAb/sleepers Scrap/electronic waste Sanitary products/diapers Other combustibles Other non-combustibles
13 15 14 32 <1 27 101 14 150 12
Maximum (mg/ kg) 160 10,000 125 17,640 10,000 539 210 29 322 150
RWW = Recovered waste wood. CCA = pressure-treated wood with copper, chromium and arsenic.
Traditionally, ZnO has been used in the production of photocopy paper and linoleum; however, nowadays these materials are produced differently, and paint which was once a dominating area of use for ZnO, only represents a small part of the present usage International Zinc Association-Zinc Oxide Information Centre (2012). However, even if the production of many products might have changed, and the usage of ZnO has shifted to other areas, the ZnO once added in different materials will most likely still appear in the waste streams today when the materials are discarded. Another Zn-compound (zinc stearate, Zn(C18H35O2)2) is used as a heat stabilizer in common materials, such as PVC. Often it is used in combination with other chemical compounds like those of Ca, Ba or K (MacKenzie and Willis, 1983; Minagawa, 1989). Owing to the flexibility, mechanical properties, and electrical properties in materials stabilized with the Ca/Zn-combination it is used in a variety of PVC based plastics, such as potable water pipes, healthcare products, water bottles, cable covering, and toys PVC Europe (2012). All these components are possible fractions in all three waste fractions (MSW, ICW, and BDW). In combination with Ba the most common materials include flexible foils, flooring, wall covering, fabric coating, and footwear. Moreover, the K/Zn-combination is used as a stabilizer for foam layers in cushion flooring, foamed wallpaper and foamed fabric coating PVC Europe (2012). All these are common products that can be found in MSW, ICW, and BDW. Material recovery is conducted, some at the source on household level and some at treatment facilities, but with methods vary due to differences between municipalities in Sweden. However, after pretreatment Zn-content are still present in various concentrations in all of the three waste streams. During combustion Zn from the fuel is released and is available to form gaseous products in the flue gas. Previous studies indicate that Zn readily forms ZnCl2, which melts at low temperatures creating sticky melts and increased deposit formation on boiler heat exchanger surfaces (Sarofirm and Helble, 1993; Verhulst et al., 1996; Ljungdahl and Zintl, 2001; Backman et al., 2005; Niemi et al., 2006; Åmand et al., 2006; Enestam et al., 2011; Enestam, 2011; Sjöblom, 2011; Bankiewicz et al., 2012). Elled et al., 2008, suggests that the formation of ZnCl2 is thermodynamically favoured between 450 and 850 °C at reducing conditions while at oxidizing conditions, the formation is initiated at 400 °C and gradually increases with temperature. However, the calculations also show that reducing conditions increase the release of gaseous Zn-content to the flue gas but in the case of oxidizing conditions, the retention of Zn in the solid ash is strong. The most common form of Zn, ZnO, is thermally stable with a melting temperature of 1975 °C. ZnO is insoluble in both water
and alcohol (Verhulst et al., 1996). This hinders volatilization at combustion temperatures under the melting temperature. However, at elevated temperatures it readily reacts with hydrochloric acid (HCl) to form gaseous ZnCl2:
ZnOðsÞ þ 2HClðgÞ ! ZnCl2 ðgÞ þ H2 OðgÞ Verhulst et al. (1996) have shown with thermodynamic equilibrium calculations that the amount of chlorine highly affects the volatility of Zn. In addition, they suggest that the presence of water decreases the volatilization of Zn by shifting the equilibrium to the left in reaction. This means that in the presence of HCl in the flue gas also could contribute to the formation of ZnCl2. In a study of Zn in fly ash aerosols and ash deposits in a full scale boiler, Enestam et al. (2011) doped wood chips with ZnO to increase the Zn-content. They showed that Zn can be released from its oxide form during the combustion process.
2. Methods To obtain a good overview of the Zn-content in Swedish waste fuel, seven waste combusting plants spread over Sweden with 1000 km between the southernmost and the northernmost were investigated in this project. There were two Bubbling Fluidised Bed (BFB) boilers and five grate fired boilers. This ratio is in accordance with what the distribution of Sweden’s waste combustion plants were at the time for the study approximately 25% were FB-boilers and plants with grate fired boilers. At each of the seven plants studied, six solid waste samples were collected over a twelve month period. The samples are mixtures of MSW, ICW, and BDW, either shredded and mixed (plant A and B) or in its original form directly from the bunker (plant C–G). In Table 2, the plants are described with the type of combustion technique and the mean average of MSW during the sampling campaign, the table also illustrates which months the samples were taken. The time for sampling was based on plant availability, which leads to irregularity in time for some of the plants.
2.1. Plants A and B Fuel that is combusted at BFB-boilers needs to undergo pretreatment before combustion, to ensure a particle size suitable for the combustion technique. The incoming fuel is often a mixture of MSW, ICW, and to some extend BDW, but all fractions are shredded and mixed at a preparation site, before being placed in the bunker and subsequently fed into the boiler via an overhead crane. The sampling procedure was highly facilitated by the pre-treatment of the waste which makes the fuel mixture more homogeneous even if this fuel is much more heterogeneous than, for example, pure biomass fuels. In addition to this, in Sweden some sorting has normally already been done at source by the households to exclude material like glass, metals, paper, cardboard and food waste. The extent to which sorting is done is different in the various regions in Sweden when it comes to which and how much of the material that is separated before the waste is collected to go to combustion. When the fuel entered the boiler system, it is possible to collect a fuel sample via a hatch that is placed close to where the fuel drops down towards the combustion bed. The samples at plant A and B were collected by repeatedly inserting a shovel directly into the falling waste stream in a specific pattern until a sample of 30 kg was gathered. Samples were taken according to the information in Table 2.
2677
F. Jones et al. / Waste Management 33 (2013) 2675–2679 Table 2 The seven waste combusting plants included in the project and the time for sampling. Plant ID
Combustion technique
Mean average MSW (%)
Sampling marked with x 2010 S
A B C D E F G
BFB BFB Grate Grate Grate Grate Grate
25 35 100 100 55 40 40
x
2.2. Plants C–G Extracting a representative sample from a grate-fired boiler is more complicated due to the even higher heterogeneity of the waste when it is not pre-treated and mixed, as in the BFBs. In the grate boilers, the fuel is partly treated, some pieces of the ICW and the BDW might be too large and will therefore be shredded. As well as in the BFB:s the waste is sorted at the source, in various percentages, depending on the region the waste is collected. The heterogeneity of the fuel requires a systematic method that will give a good overview of the bunker content. In this study, a dividing method based on the standard DD CEN/TS 15442:2006 (2006) was employed for the sampling. The sampling procedure started with a mixing of the incoming material in the bunker (by using the overhead crane) over a given time-period, varying between the plants but normally between one to 3 days. After the mixing the waste material was placed strategically close to an opening of the bunker so that a sample of five to seven tonnes could be extracted. The sample was then shredded and mixed twice, resembling the treatment of BFB waste fuel. After this, the sample was spread on clean ground in a square shape of approximately 10 10 m. The square was divided into two halves, one half was disposed of and the remaining half was once again spread out on the same area. As before, one half was removed and this dividing procedure was repeated until each of the four possible halves of the square were removed and the height of the remaining square was about 20 cm. Subsequently, the square was divided into smaller squares of approximately one square meter. From each one of these smaller squares one sample was carefully taken with a shovel, with the aim to secure samples all the way from the bottom to the top of the waste pile. The total weight of the final samples collected at the five grate-fired boilers was approximately 30 kg. The halving process during the sampling procedure is described schematically in Fig. 1. All 42 solid waste fuel samples were chemically analyzed at an accredited laboratory with respect to fuel composition. The samples were prepared by grinding and mixing before a sub-sample was taken from the 30 kg-sample by using a coning and quartering method according to DD CEN/TS 15443:2006 (2006). From this subsample, the mass content of C, H and N was determined with a gas analyzer, while the water soluble anions and cations were analyzed with Ion Chromatography (IC) in combination with a Conductivity Detector (TCD). Metals and trace metals were analyzed with
Fig. 1. The sampling of solid waste from a grate-fired boiler. The waste is extracted from the bunker and placed on the clean ground in a square shape. One halv is removed and the waste was squattered once again. This is repeated until all possible four halves have been removed.
2011 O
N
D
x x x
x x x
x x x
x x x
J
F x x
x x x
x x x
M x x x
xxx
A
M
x
x x x x
x x x x
J
J
A x
x x x x x
Inductively Coupled Plasma-Optical Emission Spectroscopy (ICPOES). All analytical standard methods are listed in Table 3. 3. Results and discussion The chemical analyses provided detailed data about the fuel composition of the incoming waste at the seven plants and the mean averages for the ultimate analyses of all 42 samples are presented in Table 4. As can be seen from this table the fuel mixtures do not differ greatly and do not seem to be dependent on either combustion technique (plant A and B compared to the rest) or share of MSW in the mixture. 3.1. Combustion technique, share of MSW and geographical relevance In Table 5 the analysis results for Zn are presented. The results show significant differences between the samples, both between samples from the seven plants and also within samples from one plant. There is no clear difference between the waste mixtures fed to the two different combustion techniques. However, for the two plants with 100% MSW (C and D) there are a lower mean average and median in the Zn-content over the whole sampling campaign. This indicates that MSW contains less Zn than ICW and/or BDW. When excluding the samples showing the highest and lowest Zn-concentration (min/max-samples) for each plant, the two boilers with 100% MSW are still the ones with the lowest mean average, while plant E still has the highest average. Plant F, G, A, and B have mean averages closer to each other and change places in the middle of the list when the min/max-samples are excluded. Furthermore, plant E that has the lowest share of MSW and also has the highest mean average and median but also the highest standard deviation. This is related to that plant E has one extreme sample, namely sample E6 containing 15,000 mg/kg Zn, increasing both the mean and the median. However, even when excluding sample 6, plant E has the highest mean average over the sampling period. Considering all samples, the mean average is 1100 mg/kg ds owing to the outlier of 15,000 mg/kg ds from Plant E. When excluding the min/max-samples it goes down to 800 mg/kg ds, which is much closer to the median of all samples, 830 mg/kg ds. The geographical spread of the seven plants does not contribute to any differences; neither does the size of the population in the city where the plants are situated. The Zn analysis results suggest that the concentration of Zn in the mixed waste fuel, considering longer time periods, generally origins from the ICW and BDW than the MSW. Nevertheless, MSW can include materials with high levels of Zn, making single samples peak and giving high analysis results for the MSW. 3.2. Correlations to other elements A correlation study within each of the combustion plants was performed where the Zn-content were correlated to each of the
2678
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Table 3 Analysis methods used for the chemical analysis. Parameter
Analysis method
Moisture content Ash S, Cl C, H, N Calorific value Ash-forming elements (Al, Si, Fe, Mn, Ti, Ca, Mg, Ba, Na, K, P) Trace elements (As, Pb, Cd, Cr, Cu, Co, Ni, Zn, V, Mo, Sb)
SS-EN 14774-2:2009 SS-EN 14775:2009 SS-EN 15289:2011 SS-EN 15104:2009 SS-EN 14918:2010 Mod. ASTM D 3682:2006 Mod. ASTM D 3683:2011
Table 6 The only elements with a clear correlation to Zn, the numbers refers to the correlation factor between Zn and the element. Plant Id. Element
C
Cu Mn Pb As Fe
0.90
D
F
0.98 0.98 0.98 0.95
3.3. Variation over time other elements analyzed in the waste samples (see Table 3 for all analyzed elements). No clear pattern was visible for correlations with Zn when comparing all waste mixtures. It is therefore not possible to point out a special cause or origin, which would have facilitated exclusion of this waste stream to avoid Zn in the fuel. However, in three of the plants; C, D, and F, there were good correlations with one or more elements, suggesting the waste consisted of material which combine the elements, see Table 6. The elements that correlate well with Zn in three of the plants are all metals, except the metalloid Arsenic. In both Plant C and D the fuel mixture was 100% MSW, meaning that neither ICW nor BDW has contributed to the share of metals. In addition, Plant C and D are also the plants with the lowest mean average content of Zn. The elements occurring together with Zn in these two plants; Cu, Mn, Pb, and Fe are most likely originating from small Waste Electrical and Electronic Equipment (sWEEE) that should have been excluded from the MSW by sorting at source. With more than half of the waste mixture consisting of Industrial and commercial/ BDW in Plant F, it is hard to determine the dominating source of the Zn–Fe combination, but the alloy is widely used in, for example, the automotive industry (Marder et al., 1991) which indicates it originates elsewhere than from MSW.
Owing to the significant difference in the mean average for all 42 samples when including or excluding the clear outlier E6 (1124 compared to 799 mg/kg ds) the following data analysis is done by using the Zn-content normalized against the mean average that is closest to the median, excluding the min/max of all samples. Fig. 2 presents the variation in Zn-content over the year the sampling campaign was performed. The mean average for all samples is represented by 1 on the logarithmic scale and all variation around it is the scattering of the different Zn-content in the samples. The first samplings took place in the (Swedish) autumn and the very last sample was taken late summer, in the end of August. For the two fluidised bed plants, and the plants operating on 100% MSW, the Zn-content stays relatively stable over the sampling campaign. Two of the plants with the highest share of ICW/ BDW, Plant E and F both have slightly higher Zn-content during the winter, possibly indicating that the waste mixtures contain more Zn during this time period. However, no general trends of Zn either increasing or decreasing depending on the month of the year or season can be concluded with only a few samples from each plant and time period.
Table 4 Results from the ultimate fuel analysis. ID. (%) Analyzed parameter
Plant MSW
A 25%
B 35%
C 100%
D 100%
E 55%
F 40%
G 40%
Moisture Ash C H N S Cl Higher heating value (constant volume) Lower heating value (constant pressure)
wt.% (as) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) MJ/kg (ds) MJ/kg (ds)
35 20 47 6.2 1.2 0.41 0.80 24.37 20.37
37 21 46 6.0 1.1 0.61 0.80 24.47 20.43
45 20 46 6.3 1.2 0.15 0.75 25.50 20.70
43 19 46 6.2 1.3 0.20 0.68 26.07 20.47
35 24 44 5.7 1.1 0.40 0.73 23.20 19.00
35 24 45 5.9 1.0 0.57 0.88 23.12 19.25
37 17 50 6.6 0.9 0.29 0.84 24.48 22.18
as = as received, ds = dry sample
Table 5 Results from the Zn-analysis of each solid waste fuel sample. Minimum and maximum-analyses for each plant are in italics. The mean and median are given for each plant and for all samples, as well as for all samples where the minimum and maximum analyses of all samples have been subtracted. Plant Id.
Sample
1
2
3
4
5
6
Mean
Median
A B C D E F G All samples All samples – min/max
mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg
960 830 950 250 850 630 1100
1000 550 840 250 640 450 820
750 860 480 380 1090 1600 1800
910 1020 420 300 1400 1500 640
930 660 260 1250 950 840 450
630 1000 450 280 15,000 580 660
860 820 560 450 3300 930 910 1100 800
920 840 460 290 1000 730 740 830 830
(ds) (ds) (ds) (ds) (ds) (ds) (ds)
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Fig. 2. The variation of zinc in the waste fuel mixes during the full sampling campaign on a logarithmic scale. The figure is based on the normalization of all samples by the mean average excluding the minimum and maximum samples. In the figure the four time periods are related to the Swedish autumn (September– November), winter (December–February), spring (March–May) and summer (June– August).
4. Conclusions A summary of the waste database AvfallsAtlas shows that Zn originates from both MSW as well as ICW, and from BDW. A study of the different waste fractions in the database suggests that the highest contribution of Zn comes from rubber tires, glass, plastics treated with flame retardants, and PVC plastics. A large national waste sampling campaign, conducted at seven separate waste combustion plants with two different combustion techniques, and spread over 1000 km in Sweden was performed to study the presence of Zn in Swedish waste fuels. Six samples were taken at each plant and analyzed at an accredited laboratory according to standard methods. The trace analysis shows that there is a relation between the MSW-content and the Zn-content; high MSW-content gives lower Zn-content. The study also shows that the geographic location of the plant in Sweden does not have any effect on the Zn-content. Furthermore, it is concluded that different seasons appear not to affect the Zn-concentrations significantly. When studying the correlations between Zn and other chemical elements no general pattern is observed when considering all waste mixtures. However, for three of the plants there was a clear correlation between Zn and other elements; in one plant there was a strong correlation with CU, in another a correlation between Zn and Fe, and in the third one a correlation between Zn and the following three elements: Mn, Pb, and As. This suggests that Zn together with these other elements often originate from the same waste fraction in these particular WtE plants. Acknowledgements Funding for this project has been received from the seven participating plants, all which are greatly acknowledged. In addition, this project has been financially supported by the Swedish Waste Management (Avfall Sverige) and The Swedish Energy Agency. References Åmand, L.-E., Leckner, B., Eskilsson, D., Tullin, C., 2006. Ash deposition on heat transfer tubes during combustion of demolition wood. Energy Fuels 20 (3), 1001–1007.
2679
ASTM International Standard ASTM D3682-01, 2006. Standard Test Methods for Major and Minor Elements in Combustion Residues from Coal Utilization Processes, doi:http://dx.doi.org/10.1520/D3682-01R06. ASTM International Standard ASTM D3683-11:2011 Standard Test Method for Trace Elements in Coal and Coke Ash by Atomic Absoption, doi:http://dx.doi.org/ 10.1520/D3683-11. Backman, R., Hupa, M., Hiltunen, M., Peltola, K., 2005. Interaction of behaviour of lead and zinc with alkalis in fluidized bed combustion or gasification of waste derived fuels. In: Proceeding of FBC 2005, 18th International Conference on Fluidized Bed Combustion, May 22–25, Toronto, Ontario, Canada. Bankiewicz, D., Enestam, S., Yrjas, P., Hupa, M., 2012. Experimental studies of Zn and Pb induced high temperature corrosion of two commercial boiler steels. Fuel Process. Technol. available online since 10 January 2012 on:. CEN Solid recovered fuels. Methods for sampling. DD CEN/TS 15442:2006 ISBN: 978 0 580 50257 6, European Committee for Standardization CEN, Brussels. CEN Solid recovered fuels. Methods for laboratory sample preparation. DD CEN/TS 15443:2006. ISBN: 978 0 580 50256 9, European Committee for Standardization CEN, Brussels. Depew, H.A., 1933a. Zinc oxide in rubber. Ind. Eng. Chem. Res. 25 (4), 370–374. Depew, H.A., 1933b. Zinc oxide in rubber II choice and testing of zinc oxide for rubber compounding. Ind. Eng. Chem. Res. 25 (5), 532–534. Depew, H.A., 1933c. Zinc oxide in rubber III manufacture of zinc oxide suitable for rubber compounding. Ind. Eng. Chem. Res. 25 (5), 534–538. Elled, A.L., Åmand, L.-E., Eskilsson, D., 2008. The fate of zinc during combustion of demolition wood in a fluidized bed boiler. Energy Fuels 22 (3), 1519–1526. Enestam, S., 2011. Corrosivity of hot flue gases in the fluidized bed combustion of recovered waste wood. Academic dissertation. Åbo Akademi University, Turku, Finland. Enestam, S., Boman, C., Niemi, J., Boström, D., Backman, R., Mäkelä, K., Hupa, M., 2011. Occurrence of Zinc an lead in aerosols and deposits in the fluidized-bed combustion of recovered waste wood. Part 1: samples from boilers. Energy Fuels 25 (4), 1970–1977. International Zinc Association-Zinc Oxide Information Centre, . Visited 25th April 2012. Klingshirn, C.F., Meyer, B.K., Waag, A., Hoffmann, A., Geurts, J., 2010. Zinc Oxide – From Fundamental Properties Towards Novel Applications. Springer Series in Material Science, Springer, Heidelberg Dordrecht London New York. . Ljungdahl, B., Zintl, F., 2001. Demolition wood and sewage sludge. Problem identification related to bed sintering and emissions during combustion in a fluidized bed. Värmeforsk Service AB, Report 753, Stockholm November 2001 (In Swedish). MacKenzie, M.W., Willis, H.A., 1983. An infra-red spectroscopic study of the stabilization of poly(vinyl chloride) by zinc and calcium stearates. Eur. Polym. J. 19 (6), 511–517. Marder, A.R., Mingyan, G.U., Notis, M.R., 1991. The effect of continuous heating on the phase transformations in zinc-iron electrodeposited coatings. Metall. Trans. A 22 (8), 1737–1743. Minagawa, M., 1989. New developments in polymer stabilization. Polym. Degrad. Stab. 25 (2), 121–141. Moezzi, A., McDonagh, A.M., Cortie, M.B., 2012. Zinc oxide particles: synthesis, properties and applications. Chem. Eng. J. 185–186, 1–22. Niemi, J., Enestam, S., Mäkelä, K., 2006. The influence of lead and zinc on superheater deposit formation in fluidized bed combustion of demolition wood. In: Proceedings of FBC 2006, 19th International Conference on Fluidized Bed Combustion, May 21–24, Vienna, Austria. PVC Europe, http://www.pvc.org/en. Visited 25th April 2012. Sarofirm, A.F., Helble, J.J., 1993. The impact of ash deposition on coal fired plants. In: Williamson, J., Wigley, F. (Eds.). Taylor & Francis, Washington, DC, pp. 67–582. Shen, K.K., Kochesfahani, S., Jouffret, F., 2008. Zinc borate as multifunctional polymer additives. Polym. Adv. Technol. 19, 469–474. Sjöblom, R., 2011. Hypotheses and mechanisms for development of deposits containing zinc and lead in conjunction with combustion of wood waste. Värmeforsk Service AB, Report 734, Stockholm March 2011 (In Swedish). Swedish Standard Institute, 2009. Solid biofuels – Determination of moisture content – Oven dry method – Part 2: Total moisture – simplified method SS-EN 14774:2:2009. Swedish Standard Institute, 2009. Solid biofuels – Determination of ash content SSEN 14775:2009. Swedish Standard Institute, 2009. Solid biofuels – Determination of total content of carbon, hydrogen and nitrogen – Instrumental methods SS-EN 15104:2009. Swedish Standard Institute, 2010. Solid biofuels – Determination of calorific value SS-EN 14918:2010. Swedish Standard Institute, 2011. Solid biofuels – Determination of total content of sulphur and chlorine SS-EN15289:2011. Verhulst, D., Buekens, A., Spencer, P.J., Eriksson, G., 1996. Thermodynamic behavior of metal chlorides and sulfates under the conditions of incineration furnaces. Environ. Sci. Technol. 30 (1), 50–56.
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
The presence of zinc in Swedish waste fuels Frida Jones a,b,⇑, Mattias Bisaillon c, Daniel Lindberg a, Mikko Hupa a a
Åbo Akademi University, Biskopsgatan 8, FI-20500 Åbo, Finland SP Technical Research Institute of Sweden, Box 857, SE-501 15 Borås, Sweden c Profu AB, Götaforsliden 13, 431 34 Mölndal, Sweden b
a r t i c l e
i n f o
Article history: Received 20 December 2012 Accepted 19 July 2013 Available online 4 September 2013 Keywords: Zinc Solid waste Waste-to-Energy plants Fuel characterization
a b s t r a c t Zinc (Zn) is a chemical element that has gained more attention lately owing to its possibility to form corrosive deposits in large boilers, such as Waste-to-Energy plants. Zn enters the boilers in many different forms and particularly in waste, the amount of Zn is hard to determine due to both the heterogeneity of waste in general but also due to the fact that little is yet published specifically about the Zn levels in waste. This study aimed to determine the Zn in Swedish waste fuels by taking regular samples from seven different and geographically separate waste combustion plants over a 12-month period. The analysis shows that there is a relation between the municipal solid waste (MSW) content and the Zn-content; high MSW-content gives lower Zn-content. This means that waste combustion plants with a higher share of industrial and commercial waste and/or building and demolition waste would have a higher share of Zn in the fuel. The study also shows that in Sweden, the geographic location of the plant does not have any effect on the Zn-content. Furthermore, it is concluded that different seasons appear not to affect the Zn concentrations significantly. In some plants there was a clear correlation between the Zn-content and the content of other trace metals. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction Zinc (Zn) enters Waste-to-Energy (WtE) plants with the different waste fuels. It is hard to determine a general form of Zn since waste mixtures are very heterogeneous. The Zn-content depends on numerous factors, such as household consumption patterns, industrial/commercial activity, season, economic climate, and possibly geographical location. In the database AvfallsAtlas,1 information from different waste fraction characterization studies have been collected and summarized, giving the possibility to show the waste fractions with the highest contribution to the share of Zn in waste. In Table 1 different waste fractions are listed with the minimum and maximum reported Zn-content in the database. The merging of different types of materials within each fraction, such as a summary of all plastic materials, also means that the origins of the waste fractions are merged. Therefore, all of the fractions presented in Table 1 include waste collected as Municipal Solid Waste (MSW), Industrial and Commercial Waste (ICW), and Building and Demolition Waste (BDW). From here on the three-letter abbreviations will be used. ⇑ Corresponding author. Address: SP Technical Research Institute of Sweden, Energy Technology, Box 857, SE-501 15 Borås, Sweden. Tel.: +46 (0) 10 516 57 67. E-mail address: [email protected] (F. Jones). 1 AvfallsAtlas (WasteAtlas) is a closed database developed by Profu AB, www.profu.se/indexaa.htm. Contact person Mattias Bisaillon ([email protected]) or Johan Sundberg ([email protected]). 0956-053X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2013.07.023
The internal variation of Zn within each fraction also indicates how difficult a general determination of Zn-content in a waste mixture would be, for example, in the Textile/leather/rubber-fraction where the Zn-content varies from around 30 to 18,000 mg/kg waste. Furthermore, although the difference between the materials in the textile/leather/rubber-fraction is obvious there is also a major difference within the glass fraction with reported measurements varying between close to 0 and 10,000 mg/kg. A study of the details behind Table 1 reveals that the materials contributing the most to the Zn-content in waste are rubber tires, glass, plastics treated with flame retardants, and PVC plastics. Zinc oxide (ZnO) is a frequently used chemical as an additive in numerous products (Klingshirn et al., 2010; Moezzi et al., 2012). The most dominant application of ZnO is within the rubber industry where it is used in many types of rubber, and has been for over a century (Depew, 1933a, 1933b, 1933c). Predominantly ZnO is used in the car industry, activating the sulfur crosslinking and also for improving the absorption of frictional heat in tires, as well as being used in belts, windshield wipers, hoses, oil seals, trim and mountings International Zinc Association-Zinc Oxide Information Centre (2012). In addition, ZnO is also used in heat resistant glass, cooking wear glass, and specialty glass, such as photochromic lenses. ZnO is also a main ingredient for the production of a flame retardant zinc borate (XZnOYB2O3) which exists in various forms, differing by the Zn/B-ratio and number of attached hydrates. Zinc borate is added to a number of materials and products, such as plastics, textiles, and rubber (Shen et al., 2008).
2676
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Table 1 Waste fractions and their Zn-content, summarized information from the database AvfallsAtlas.
a b
Waste fraction
Minimum (mg/ kg)
Cardboard/paper/corrugated cardboard Plastics Kitchen and garden waste Textile/leather/rubber Glass Wood/RWWa/CCAb/sleepers Scrap/electronic waste Sanitary products/diapers Other combustibles Other non-combustibles
13 15 14 32 <1 27 101 14 150 12
Maximum (mg/ kg) 160 10,000 125 17,640 10,000 539 210 29 322 150
RWW = Recovered waste wood. CCA = pressure-treated wood with copper, chromium and arsenic.
Traditionally, ZnO has been used in the production of photocopy paper and linoleum; however, nowadays these materials are produced differently, and paint which was once a dominating area of use for ZnO, only represents a small part of the present usage International Zinc Association-Zinc Oxide Information Centre (2012). However, even if the production of many products might have changed, and the usage of ZnO has shifted to other areas, the ZnO once added in different materials will most likely still appear in the waste streams today when the materials are discarded. Another Zn-compound (zinc stearate, Zn(C18H35O2)2) is used as a heat stabilizer in common materials, such as PVC. Often it is used in combination with other chemical compounds like those of Ca, Ba or K (MacKenzie and Willis, 1983; Minagawa, 1989). Owing to the flexibility, mechanical properties, and electrical properties in materials stabilized with the Ca/Zn-combination it is used in a variety of PVC based plastics, such as potable water pipes, healthcare products, water bottles, cable covering, and toys PVC Europe (2012). All these components are possible fractions in all three waste fractions (MSW, ICW, and BDW). In combination with Ba the most common materials include flexible foils, flooring, wall covering, fabric coating, and footwear. Moreover, the K/Zn-combination is used as a stabilizer for foam layers in cushion flooring, foamed wallpaper and foamed fabric coating PVC Europe (2012). All these are common products that can be found in MSW, ICW, and BDW. Material recovery is conducted, some at the source on household level and some at treatment facilities, but with methods vary due to differences between municipalities in Sweden. However, after pretreatment Zn-content are still present in various concentrations in all of the three waste streams. During combustion Zn from the fuel is released and is available to form gaseous products in the flue gas. Previous studies indicate that Zn readily forms ZnCl2, which melts at low temperatures creating sticky melts and increased deposit formation on boiler heat exchanger surfaces (Sarofirm and Helble, 1993; Verhulst et al., 1996; Ljungdahl and Zintl, 2001; Backman et al., 2005; Niemi et al., 2006; Åmand et al., 2006; Enestam et al., 2011; Enestam, 2011; Sjöblom, 2011; Bankiewicz et al., 2012). Elled et al., 2008, suggests that the formation of ZnCl2 is thermodynamically favoured between 450 and 850 °C at reducing conditions while at oxidizing conditions, the formation is initiated at 400 °C and gradually increases with temperature. However, the calculations also show that reducing conditions increase the release of gaseous Zn-content to the flue gas but in the case of oxidizing conditions, the retention of Zn in the solid ash is strong. The most common form of Zn, ZnO, is thermally stable with a melting temperature of 1975 °C. ZnO is insoluble in both water
and alcohol (Verhulst et al., 1996). This hinders volatilization at combustion temperatures under the melting temperature. However, at elevated temperatures it readily reacts with hydrochloric acid (HCl) to form gaseous ZnCl2:
ZnOðsÞ þ 2HClðgÞ ! ZnCl2 ðgÞ þ H2 OðgÞ Verhulst et al. (1996) have shown with thermodynamic equilibrium calculations that the amount of chlorine highly affects the volatility of Zn. In addition, they suggest that the presence of water decreases the volatilization of Zn by shifting the equilibrium to the left in reaction. This means that in the presence of HCl in the flue gas also could contribute to the formation of ZnCl2. In a study of Zn in fly ash aerosols and ash deposits in a full scale boiler, Enestam et al. (2011) doped wood chips with ZnO to increase the Zn-content. They showed that Zn can be released from its oxide form during the combustion process.
2. Methods To obtain a good overview of the Zn-content in Swedish waste fuel, seven waste combusting plants spread over Sweden with 1000 km between the southernmost and the northernmost were investigated in this project. There were two Bubbling Fluidised Bed (BFB) boilers and five grate fired boilers. This ratio is in accordance with what the distribution of Sweden’s waste combustion plants were at the time for the study approximately 25% were FB-boilers and plants with grate fired boilers. At each of the seven plants studied, six solid waste samples were collected over a twelve month period. The samples are mixtures of MSW, ICW, and BDW, either shredded and mixed (plant A and B) or in its original form directly from the bunker (plant C–G). In Table 2, the plants are described with the type of combustion technique and the mean average of MSW during the sampling campaign, the table also illustrates which months the samples were taken. The time for sampling was based on plant availability, which leads to irregularity in time for some of the plants.
2.1. Plants A and B Fuel that is combusted at BFB-boilers needs to undergo pretreatment before combustion, to ensure a particle size suitable for the combustion technique. The incoming fuel is often a mixture of MSW, ICW, and to some extend BDW, but all fractions are shredded and mixed at a preparation site, before being placed in the bunker and subsequently fed into the boiler via an overhead crane. The sampling procedure was highly facilitated by the pre-treatment of the waste which makes the fuel mixture more homogeneous even if this fuel is much more heterogeneous than, for example, pure biomass fuels. In addition to this, in Sweden some sorting has normally already been done at source by the households to exclude material like glass, metals, paper, cardboard and food waste. The extent to which sorting is done is different in the various regions in Sweden when it comes to which and how much of the material that is separated before the waste is collected to go to combustion. When the fuel entered the boiler system, it is possible to collect a fuel sample via a hatch that is placed close to where the fuel drops down towards the combustion bed. The samples at plant A and B were collected by repeatedly inserting a shovel directly into the falling waste stream in a specific pattern until a sample of 30 kg was gathered. Samples were taken according to the information in Table 2.
2677
F. Jones et al. / Waste Management 33 (2013) 2675–2679 Table 2 The seven waste combusting plants included in the project and the time for sampling. Plant ID
Combustion technique
Mean average MSW (%)
Sampling marked with x 2010 S
A B C D E F G
BFB BFB Grate Grate Grate Grate Grate
25 35 100 100 55 40 40
x
2.2. Plants C–G Extracting a representative sample from a grate-fired boiler is more complicated due to the even higher heterogeneity of the waste when it is not pre-treated and mixed, as in the BFBs. In the grate boilers, the fuel is partly treated, some pieces of the ICW and the BDW might be too large and will therefore be shredded. As well as in the BFB:s the waste is sorted at the source, in various percentages, depending on the region the waste is collected. The heterogeneity of the fuel requires a systematic method that will give a good overview of the bunker content. In this study, a dividing method based on the standard DD CEN/TS 15442:2006 (2006) was employed for the sampling. The sampling procedure started with a mixing of the incoming material in the bunker (by using the overhead crane) over a given time-period, varying between the plants but normally between one to 3 days. After the mixing the waste material was placed strategically close to an opening of the bunker so that a sample of five to seven tonnes could be extracted. The sample was then shredded and mixed twice, resembling the treatment of BFB waste fuel. After this, the sample was spread on clean ground in a square shape of approximately 10 10 m. The square was divided into two halves, one half was disposed of and the remaining half was once again spread out on the same area. As before, one half was removed and this dividing procedure was repeated until each of the four possible halves of the square were removed and the height of the remaining square was about 20 cm. Subsequently, the square was divided into smaller squares of approximately one square meter. From each one of these smaller squares one sample was carefully taken with a shovel, with the aim to secure samples all the way from the bottom to the top of the waste pile. The total weight of the final samples collected at the five grate-fired boilers was approximately 30 kg. The halving process during the sampling procedure is described schematically in Fig. 1. All 42 solid waste fuel samples were chemically analyzed at an accredited laboratory with respect to fuel composition. The samples were prepared by grinding and mixing before a sub-sample was taken from the 30 kg-sample by using a coning and quartering method according to DD CEN/TS 15443:2006 (2006). From this subsample, the mass content of C, H and N was determined with a gas analyzer, while the water soluble anions and cations were analyzed with Ion Chromatography (IC) in combination with a Conductivity Detector (TCD). Metals and trace metals were analyzed with
Fig. 1. The sampling of solid waste from a grate-fired boiler. The waste is extracted from the bunker and placed on the clean ground in a square shape. One halv is removed and the waste was squattered once again. This is repeated until all possible four halves have been removed.
2011 O
N
D
x x x
x x x
x x x
x x x
J
F x x
x x x
x x x
M x x x
xxx
A
M
x
x x x x
x x x x
J
J
A x
x x x x x
Inductively Coupled Plasma-Optical Emission Spectroscopy (ICPOES). All analytical standard methods are listed in Table 3. 3. Results and discussion The chemical analyses provided detailed data about the fuel composition of the incoming waste at the seven plants and the mean averages for the ultimate analyses of all 42 samples are presented in Table 4. As can be seen from this table the fuel mixtures do not differ greatly and do not seem to be dependent on either combustion technique (plant A and B compared to the rest) or share of MSW in the mixture. 3.1. Combustion technique, share of MSW and geographical relevance In Table 5 the analysis results for Zn are presented. The results show significant differences between the samples, both between samples from the seven plants and also within samples from one plant. There is no clear difference between the waste mixtures fed to the two different combustion techniques. However, for the two plants with 100% MSW (C and D) there are a lower mean average and median in the Zn-content over the whole sampling campaign. This indicates that MSW contains less Zn than ICW and/or BDW. When excluding the samples showing the highest and lowest Zn-concentration (min/max-samples) for each plant, the two boilers with 100% MSW are still the ones with the lowest mean average, while plant E still has the highest average. Plant F, G, A, and B have mean averages closer to each other and change places in the middle of the list when the min/max-samples are excluded. Furthermore, plant E that has the lowest share of MSW and also has the highest mean average and median but also the highest standard deviation. This is related to that plant E has one extreme sample, namely sample E6 containing 15,000 mg/kg Zn, increasing both the mean and the median. However, even when excluding sample 6, plant E has the highest mean average over the sampling period. Considering all samples, the mean average is 1100 mg/kg ds owing to the outlier of 15,000 mg/kg ds from Plant E. When excluding the min/max-samples it goes down to 800 mg/kg ds, which is much closer to the median of all samples, 830 mg/kg ds. The geographical spread of the seven plants does not contribute to any differences; neither does the size of the population in the city where the plants are situated. The Zn analysis results suggest that the concentration of Zn in the mixed waste fuel, considering longer time periods, generally origins from the ICW and BDW than the MSW. Nevertheless, MSW can include materials with high levels of Zn, making single samples peak and giving high analysis results for the MSW. 3.2. Correlations to other elements A correlation study within each of the combustion plants was performed where the Zn-content were correlated to each of the
2678
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Table 3 Analysis methods used for the chemical analysis. Parameter
Analysis method
Moisture content Ash S, Cl C, H, N Calorific value Ash-forming elements (Al, Si, Fe, Mn, Ti, Ca, Mg, Ba, Na, K, P) Trace elements (As, Pb, Cd, Cr, Cu, Co, Ni, Zn, V, Mo, Sb)
SS-EN 14774-2:2009 SS-EN 14775:2009 SS-EN 15289:2011 SS-EN 15104:2009 SS-EN 14918:2010 Mod. ASTM D 3682:2006 Mod. ASTM D 3683:2011
Table 6 The only elements with a clear correlation to Zn, the numbers refers to the correlation factor between Zn and the element. Plant Id. Element
C
Cu Mn Pb As Fe
0.90
D
F
0.98 0.98 0.98 0.95
3.3. Variation over time other elements analyzed in the waste samples (see Table 3 for all analyzed elements). No clear pattern was visible for correlations with Zn when comparing all waste mixtures. It is therefore not possible to point out a special cause or origin, which would have facilitated exclusion of this waste stream to avoid Zn in the fuel. However, in three of the plants; C, D, and F, there were good correlations with one or more elements, suggesting the waste consisted of material which combine the elements, see Table 6. The elements that correlate well with Zn in three of the plants are all metals, except the metalloid Arsenic. In both Plant C and D the fuel mixture was 100% MSW, meaning that neither ICW nor BDW has contributed to the share of metals. In addition, Plant C and D are also the plants with the lowest mean average content of Zn. The elements occurring together with Zn in these two plants; Cu, Mn, Pb, and Fe are most likely originating from small Waste Electrical and Electronic Equipment (sWEEE) that should have been excluded from the MSW by sorting at source. With more than half of the waste mixture consisting of Industrial and commercial/ BDW in Plant F, it is hard to determine the dominating source of the Zn–Fe combination, but the alloy is widely used in, for example, the automotive industry (Marder et al., 1991) which indicates it originates elsewhere than from MSW.
Owing to the significant difference in the mean average for all 42 samples when including or excluding the clear outlier E6 (1124 compared to 799 mg/kg ds) the following data analysis is done by using the Zn-content normalized against the mean average that is closest to the median, excluding the min/max of all samples. Fig. 2 presents the variation in Zn-content over the year the sampling campaign was performed. The mean average for all samples is represented by 1 on the logarithmic scale and all variation around it is the scattering of the different Zn-content in the samples. The first samplings took place in the (Swedish) autumn and the very last sample was taken late summer, in the end of August. For the two fluidised bed plants, and the plants operating on 100% MSW, the Zn-content stays relatively stable over the sampling campaign. Two of the plants with the highest share of ICW/ BDW, Plant E and F both have slightly higher Zn-content during the winter, possibly indicating that the waste mixtures contain more Zn during this time period. However, no general trends of Zn either increasing or decreasing depending on the month of the year or season can be concluded with only a few samples from each plant and time period.
Table 4 Results from the ultimate fuel analysis. ID. (%) Analyzed parameter
Plant MSW
A 25%
B 35%
C 100%
D 100%
E 55%
F 40%
G 40%
Moisture Ash C H N S Cl Higher heating value (constant volume) Lower heating value (constant pressure)
wt.% (as) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) wt.% (ds) MJ/kg (ds) MJ/kg (ds)
35 20 47 6.2 1.2 0.41 0.80 24.37 20.37
37 21 46 6.0 1.1 0.61 0.80 24.47 20.43
45 20 46 6.3 1.2 0.15 0.75 25.50 20.70
43 19 46 6.2 1.3 0.20 0.68 26.07 20.47
35 24 44 5.7 1.1 0.40 0.73 23.20 19.00
35 24 45 5.9 1.0 0.57 0.88 23.12 19.25
37 17 50 6.6 0.9 0.29 0.84 24.48 22.18
as = as received, ds = dry sample
Table 5 Results from the Zn-analysis of each solid waste fuel sample. Minimum and maximum-analyses for each plant are in italics. The mean and median are given for each plant and for all samples, as well as for all samples where the minimum and maximum analyses of all samples have been subtracted. Plant Id.
Sample
1
2
3
4
5
6
Mean
Median
A B C D E F G All samples All samples – min/max
mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg mg/kg
960 830 950 250 850 630 1100
1000 550 840 250 640 450 820
750 860 480 380 1090 1600 1800
910 1020 420 300 1400 1500 640
930 660 260 1250 950 840 450
630 1000 450 280 15,000 580 660
860 820 560 450 3300 930 910 1100 800
920 840 460 290 1000 730 740 830 830
(ds) (ds) (ds) (ds) (ds) (ds) (ds)
F. Jones et al. / Waste Management 33 (2013) 2675–2679
Fig. 2. The variation of zinc in the waste fuel mixes during the full sampling campaign on a logarithmic scale. The figure is based on the normalization of all samples by the mean average excluding the minimum and maximum samples. In the figure the four time periods are related to the Swedish autumn (September– November), winter (December–February), spring (March–May) and summer (June– August).
4. Conclusions A summary of the waste database AvfallsAtlas shows that Zn originates from both MSW as well as ICW, and from BDW. A study of the different waste fractions in the database suggests that the highest contribution of Zn comes from rubber tires, glass, plastics treated with flame retardants, and PVC plastics. A large national waste sampling campaign, conducted at seven separate waste combustion plants with two different combustion techniques, and spread over 1000 km in Sweden was performed to study the presence of Zn in Swedish waste fuels. Six samples were taken at each plant and analyzed at an accredited laboratory according to standard methods. The trace analysis shows that there is a relation between the MSW-content and the Zn-content; high MSW-content gives lower Zn-content. The study also shows that the geographic location of the plant in Sweden does not have any effect on the Zn-content. Furthermore, it is concluded that different seasons appear not to affect the Zn-concentrations significantly. When studying the correlations between Zn and other chemical elements no general pattern is observed when considering all waste mixtures. However, for three of the plants there was a clear correlation between Zn and other elements; in one plant there was a strong correlation with CU, in another a correlation between Zn and Fe, and in the third one a correlation between Zn and the following three elements: Mn, Pb, and As. This suggests that Zn together with these other elements often originate from the same waste fraction in these particular WtE plants. Acknowledgements Funding for this project has been received from the seven participating plants, all which are greatly acknowledged. In addition, this project has been financially supported by the Swedish Waste Management (Avfall Sverige) and The Swedish Energy Agency. References Åmand, L.-E., Leckner, B., Eskilsson, D., Tullin, C., 2006. Ash deposition on heat transfer tubes during combustion of demolition wood. Energy Fuels 20 (3), 1001–1007.
2679
ASTM International Standard ASTM D3682-01, 2006. Standard Test Methods for Major and Minor Elements in Combustion Residues from Coal Utilization Processes, doi:http://dx.doi.org/10.1520/D3682-01R06. ASTM International Standard ASTM D3683-11:2011 Standard Test Method for Trace Elements in Coal and Coke Ash by Atomic Absoption, doi:http://dx.doi.org/ 10.1520/D3683-11. Backman, R., Hupa, M., Hiltunen, M., Peltola, K., 2005. Interaction of behaviour of lead and zinc with alkalis in fluidized bed combustion or gasification of waste derived fuels. In: Proceeding of FBC 2005, 18th International Conference on Fluidized Bed Combustion, May 22–25, Toronto, Ontario, Canada. Bankiewicz, D., Enestam, S., Yrjas, P., Hupa, M., 2012. Experimental studies of Zn and Pb induced high temperature corrosion of two commercial boiler steels. Fuel Process. Technol. available online since 10 January 2012 on: