INHIBITION OF PCDD/F AND PCB FORMATION IN CO- COMBUSTION

Transcription

1 INHIBITION OF PCDD/F ND PCB FORMTION IN CO- COMBUSTION M. Pandelova a, D. Lenoir a, K.-W. Schramm a, b a GSF-Forschungszentrum für Umwelt und Gesundheit, Institut für Ökologische Chemie, Ingolstädter Landstr. 1, D-8764 Neuherberg, Germany b TUM, Wissenschaftszentrum Weihenstephan für Ernährung und Landnutung, Department für Biowissenschaften, Weihenstephaner Steig 23, D-830 Freising, Germany schramm@gsf.de; pandelova@gsf.de BSTRCT This study addresses co-combustion of coal-solid waste mixtures in pilot and laboratory-scale combustors, with emphasis on monitoring of toxic chlorinated hydrocarbon emissions such as polychlorinated dibenzo-p-dioxins and -furans (PCDD/F) and polychlorinated biphenyls (PCB). The objectives of the work are to investigate the so-called primary measures technique. Thermally resistant inorganic compounds were added directly to the fuel as inhibitors of PCDD/F formation. The fuel-types used in this study are lignite coal, pre-treated municipal solid waste and PVC. (NH 4 ) 2 SO 4 and (NH 4 ) 2 S 2 O 3 were employed as the most effective inhibitors for PCDD/F and PCB formation. Low cost and high efficiency favour them as useful for a full scale combustion unit. 1

2 1. INTRODUCTION Co-firing of solid waste with coal in power plants is a promising technique for countries lacking waste incinerators because of special environmental and economical reasons. The energy efficiency of coal for electric power, coupling of heat and local heating in power plant is substantial higher compared to the waste used as fuel in municipal waste incinerators. lthough incineration of small percentage solid waste used as substituted fuel has become an important method in coal power plants. The potential of such a substituted fuel type is high enough for proposed co-firing of waste with coal and could help to reduce the waste disposal, preserve fossil fuel coal resources, and reduce uncontrolled burning at dumping sites or backyard burning in all countries. Co-combustion of solid waste and coal also reduce the emissions of greenhouse gasses such as CO 2 and CO. However, the application of this process poses some environmental risks due to the chemical nature of the compounds formed from solid waste during co-combustion. For example, high concentrations of polychlorinated dibenzo-p-dioxins (PCDD), polychlorinated dibenzofurans (PCDF) and polychlorinated biphenyls (PCB) are produced during waste incineration [1]. ccording to a report released in 1994 by the United States Environmental Protection gency (EP), dioxin is a potential humans carcinogen and exposure to dioxin can cause immune system damage and interfere with regulatory hormones [2, 3]. Over the last few years, control over the release of dioxins into the atmosphere has become more widespread, which has led to an increasing consensus in Europe for an emission limit in flue gas of 0.1 ng/m 3 based on the TEQ value [4]. In this study, we used a laboratory-scale furnace to perform a series of experiments designed to prevent the formation of PCDD/Fs and PCBs formation by the use of inhibitors. The aim of the work was to identify the most effective inhibitors of PCDD/F formation that were with low cost effective and had low toxicity. 2. MTERILS ND METHODS 2.1 Reagents The fuel-types of fuel used in this study are lignite coal from Puertollano (Spain), pre-treated municipal solid waste (Rethmann Plano GmbH), and used polyvinyl chloride (PVC) (waste from ground carpet). The lignite coal was 80% by weight percentage content of the total fuel. Such a high amount of lignite coal was chosen because of the interest in coal combustion. lthough addition of solid waste supposes to be sufficient for dioxin formation, in the laboratory scale furnace the emissions were still not very high. Thus, the addition of PVC to the lignite coal / solid waste mixture was necessary to generate a PCDD/F-level that can potentially reduced by inhibitors at least 3 orders of magnitudes. The chlorinated compounds were 20% by weight of the fuel mixture with a 13.3 : 6.7 ratio of the solid waste to PVC. In the inhibitor experiments, the inhibitor is added as 10 % of the fuel. The remaining 90% fuel used is homogenised very well with additives in the sampling boat. lthough addition of solid waste to the coal has been reported to be optimal for dioxin formation, the dioxin emissions concentration in the laboratory - scale furnace was not very high. Thus, the addition of PVC to the lignite coal / solid waste mixture was mandatory. The 20 different compounds are investigated as inhibitors (Tab. 1). TBLE 1: lphabetic indicator of inhibitors used for laboratory scale experiments without inhibitor H H 2 NSO 3 H O P 2 S B TiO 2 I S P Na 2 S H 2 O C ZrO 2 J H 2 NSO 2 NH 2 Q Cr 2 O 3 D Na 2 MoO 4 2H 2 O K pomace R (NH 4 ) 2 SO 4

3 E l 2 O 3 L (CH 3 ) 2 N 3 P(O) S (NH 4 ) 2 HPO 4 F H 2 NSO 4 H M N(CH 2 CH 2 OH) 3 T (NH 2 ) 2 CO+S G NaVO 3 N Na 2 WO 4 2H 2 O U (NH 4 ) 2 S 2 O 3 The two most effective inhibitors for PCDD/Fs and PCBs formation were also tested at an amount of %, 3 % and 1 % of the fuel pparatus laboratory - scale horizontal split-tube furnace (Split Tube furnace (3-zone) type HZS & TVS, Carbolite, England) (Fig.1) was, is used for the experiments. The laboratory reactor consisted of a quartz tube with ID = 14 cm and a length of 10 cm, ¾ of the length of which was placed into the heating zone of the furnace. The sampling boat was pushed from the cooled part of the quartz tube to the middle of the heated zone when the desired temperature was achieved FIGURE 1: Scheme of the laboratory scale furnace (1-inlet with air flow, 2-sampling boat, 3-toluene impingers, 4-glass wool filter, -vacuum pump). The trace organics in the flue gases occurring during the experiment were trapped into 3 impingers in series each filled with 70 ml toluene. The first impinger was previously spiked with 100 l CEN-sampling PCDD/Fs standard. The combustion experiments are performed according to the same procedure each time in order to achieve reproducible results. The furnace temperature was 400 C, the air flow was 2 L / min, the weight of the sample was 10 g and the duration of the experiment was 30 min Dioxin Measurement Once the combustion experiment is completed the sample is processed for further specific clean-up steps [, 6]. The tetra- to octachloro- isomers of PCDD/F and the tetra- to hepta- PCB were identified and quantified in pg / g combusted material Statistical nalysis Kovach Computing Service, Statistical Package 3.11g, [7] was used for the Principle Component nalysis (PC). PC reflected the effect of the 20 different inhibitors on the PCDD/F congener patterns emitted. No transformation of the data is performed during the statistical analysis. 3

4 3. RESULTS ND DISCUSSION PC was used in order to give a better view of the similarity between the samples using the concentrations of PCDD and PCDF congeners. The matrix for PCDDs consisted of 37 variables (PCDDs congeners) and 2 cases (samples) and the matrix of PCDFs is comprised of 63 variables (PCDFs congeners) and 2 cases (samples). The statistical score plot for the laboratory-scale an experiment with 20 inhibitors is shown in Fig. 2. PC case scores, PCDD O 1.4 xis 2 J H F L UT R I DP S G 0.9 N 0. C MK E B Q xis 1 PC case scores, PCDF xis 2 UR T FH D I G JL S P M O C N K E B Q xis 1 FIGURE 2. Score plot of PCDD and PCDF of samples treated without any inhibitor and 20 samples with 10% inhibitor 4

5 Two main groups of samples were observed in both score plots. The further left a sample is in the statistical plot, the lower the concentration of PCDD/Fs is. The samples without inhibitor were found on the right side of the statistical plot in Fig. 2. Neighbouring these toxic samples were other points that could be attributed to the metal oxide inhibitors Cr 2 O 3, TiO 2, l 2 O 3 and pomace. However, they also belong to the samples with higher PCDD/Fs concentrations. Near to the ordinate axis and between both groups, N-, S- and metal oxide inhibitors as such P 2 S, N(CH 2 CH 2 OH) 3, Na 2 WO 4 2H 2 O and ZrO 2 were plotted. These four compounds were additives that manifested low inhibitory effects or samples with high variation in the PCDD/Fs amount. group that was highly agglomerated, especially in the case of PCDF, is on the left hand side. It included samples with 10 % inhibitor concentration, where all inhibitors were N-and S-containing compounds, and others such as NaVO 3, Na 2 MoO 4 2H 2 O, (NH 4 ) 2 HPO 4, (CH 3 ) 2 N 3 P(O), S and Na 2 S H 2 O. The points were very close to each other, which was explained by the similarity in their pattern and the amount of PCDD/Fs formed. Here, the N-and S- containing compounds were placed much more to the left in the score plot. This effect correlated to Fig. 3, where the same compounds showed a strong inhibitory effect for I-TEQ PCDD/Fs concentrations in the flue gas. The total amounts of PCDDs, PCDFs and PCBs generated during experiments with a mixture of lignite coal, solid waste and PVC were substantial enough to investigate the effect of inhibition. The average I-TEQ of the sum of PCDD/Fs and PCB was about 16 pg/g fuel or about 2. ng/nm 3 (Fig.3). Such experiments were performed five times and the results showed a relatively low standard deviation. The experimental results with 10% inhibitor showed very different I-TEQ values; some of the compounds led to similar and / or greater amounts of PCDD/Fs and PCBs than the samples without inhibitor (Fig. 3). Other additives showed a very strong inhibitory effect in the flue gases. Two compounds could reduce the dioxins emission up to 98-99%. 2 24,43 20 I-TEQ *, pg/g fuel ,94 16,24 12,46 12,4 8,94 7,93 4,99 4,68 3,89 0 2,96 2,66 2,6 1,68 1,27 1,11 1,10 0,98 0,3 0,34 0,22 Q B E K C N M O G L D P I J H F S T R U * sum PCDD/F and PCB FIGURE 3. I-TEQ* (pg/g) values of the flue gas during thermal treatment of lignite coal, solid waste and PVC for the samples without inhibitor and 20 different compounds used at 10% inhibitor of the fuel. (NH 4 ) 2 SO 4 and (NH 4 ) 2 S 2 O 3 showed stronger inhibitory effects of PCDD/F and PCB formation (Fig. 3). Therefore, experiments wherein these inhibitors were studied at %, 3 % and 1 % of the total fuel weight were performed in triplicate (Fig. 4). In addition, the experiment without any inhibitor was repeated in order to obtain an actual value during that time of the

6 investigation. The average calculated concentration for the sum PCDD/F and PCB I-TEQ was 17. pg/g with a maximum value of 20. pg/g and minimum value of 12.7 pg/g. (NH 4 ) 2 SO 4 and (NH 4 ) 2 S 2 O 3 were effective inhibitors at 10% and % of the fuel for PCDD/Fs and PCBs (Fig.4). If the percentage of these substances was decreased further, the inhibition of chlorinated micropllutants formation was also decreased. ccording to Fig. 4, (NH 4 ) 2 SO 4 might also have reduced the PCDD/Fs and PCBs in flue gas emission at 3 % but not as effectively as at higher percentages. 2 I-TEQ*, pg/g fuel ,46 6,23,3 2,88 1,86 0,4 0,34 0,48 0,78 10% R 10% U % R % U 3% R 3% U 1% R 1% U * sum PCDD/F and PCB Figure 4. I-TEQ* pg/g fuel in the flue gas for the samples without inhibitor () and with inhibitor R and U at 10 %, %, 3 % and 1 % of the fuel. Inhibitor-letter correlations can be found in Tab CONCLUSIONS Twenty different compounds subsumed into four different types of groups according their chemical nature are studied as inhibitors of PCDD/F and PCB in flue gases. High concentrations of PCDD/F and PCB are detected during the co-combustion of lignite coal and solid waste when small amount of PVC is additionally added to the mixture. Recent laboratory scale study investigates also the role of PVC in CuCl 2 H 2 O mixture showing that chlorine in PVC is once converted to HCl and contributes to PCDD/F formation [9]. However another experiment of co-combustion of coal and PVC in bubbling fluidised bed boiler suggested that an increase in PVC content of 1% to 2% resulted in a decrease of PCDD/F emissions in the flue gas as well as in the particle phase [10]. During co-combustion of coal with 2% PVC about 1.4 pg/nm 3 PCDD/F I-TEQ was detected that remained below settled European limit value of PCDD/F in flue gas. Current results showed that the metal oxides group investigated herein has no inhibitory effect. Some of the additives such as Cr 2 O 3, l 2 O 3 and TiO 2 even increased the PCDD/F emissions by catalysing likely by de novo formation. They produced very high amounts of the most toxic congeners and high amounts of other less chlorinated PCDD/Fs. Some other substances belonging to the metal oxides groups such as NaVO 3, Na 2 MoO 4 2H 2 O and Na 2 WO 4 2H 2 O showed low inhibitory effects, especially for PCDDs. These substances were active for catalytic oxidation. Their chemical structure consisted of a strong oxide base such as VO 3 -, MoO 4-2, WO 4-2. Probably due to that base, a better performance of incineration 6

7 during the experiments occurred. Regardless, these additives did not reduce PCDD/F and PCB formation sufficiently. lthough the compounds containing either N or S were not very effective as inhibitors, those containing both N and S seemed to be able to strongly reduce PCDD/Fs and PCBs flue gas emission if used as a 10 % additive to a fuel containing lignite coal, solid waste and PVC. The mechanism has been shown to be a complex interaction of the inhibitors with the catalytic active sites of the fly ash [11]. In our study, we showed that sulfur and nitrogen compounds like (NH 4 ) 2 SO 4 and (NH 4 ) 2 S 2 O 3 were the most efficient inhibitors. (NH 4 ) 2 SO 4, present at 3% of the fuel, could reduce the PCDD/F and PCB emissions by 90%. These compounds are low cost and non-toxic materials, making them applicable for use in full-scale combustion units. CKNOWLEDGMENTS This work received financial support from the European Union under contract number ECSC 7220-PR-108. REFERENCES 1. Cleaner Coal Technology Programme, Waste / biomass co-gasification with coal Technology status report 017, Thomé Kozmiensky K. J. (1994) Thermische bfallbehandlung, Berlin, EF-Verl. für Energie und Umwelttechnik, ISBN EP/600/6-88/00C United States Environmental Protection gency. 4. EP 30-R United States Environmental Protection gency.. Schramm K.-W., Merk M., Henkelmann B., Kettrup. (199) Leaching of PCDD/F from fly ash and soil with fire-extinguishing water Chemosphere, Vol. 30, Nr. 12, pp Henkelmann B., Schramm K.-W., Klimm C., Kettrup. (1996) Quality criteria for the isotope dilution method with HRGC/HRMS Fresenius Z. nal., Vol. 34, pp Multi -Variate Statistical Package Version Kovach Computing Services, 8. Pandelova M., Lenoir D., Kettrup., Schramm K.-W. (200) Primary measures for reduction of PCDD/F in co-combustion of lignite coal and waste: effect of various inhibitors Environmental Science Technology, Vol. 39, pp Hatanaka T., Kitajima., Takeuchi M. (200) Role of chlorine in combustion field in formation of polychlorinated dibenzo-p-dioxins and dibenzofurans during waste incineration Environmental Science Technology, Vol. 39, pp Sánchez-Hervás J.M., rmesto L., Ruiz-Martínez E., Otero-Ruiz J., Pandelova M., Schramm K.-W. (200) PCDD/PCDF emissions from co-combustion of coal and PVC in a bubbling fluidised bed boiler Fuel, Vol. 84, pp Dickson L. C., Lenoir D., Hutzinger O., Naikwadi K. P., Karasek F.W. (1989) Inhibition of chlorinated dibenzo-p-dioxin formation on municipal incinerator fly ash by using catalyst inhibitors Chemosphere, Vol. 19, Nr. 8-9, pp