PCDD / Fs Inhibition by Sludge Decomposition Gases : Effects of Sludge Dosage , Treatment Temperature and Oxygen Content

Sludge decomposition gases (SDG) have been identified as potential PCDD/F-inhibitors, since they contain high concentrations of NH3 and SO2 as well as various unidentified Nand S-compounds. The effects of sludge dosage, its treatment temperature and oxygen content on PCDD/Fs suppression were observed during test runs using a model fly ash (MFA) system, generating high chlorinated PCDD/Fs. The experiments were carried out in two parts: a first set with SDG of undefined composition, and a second with injection of NH3 and SO2 combined in three different concentration ratios. It was observed that the SDG evolving from dried sludge (DS) could suppress the formation of 2,3,7,8-substituted PCDD/Fs even with very high efficiency (> 99%), both in weight units and in I-TEQ units. Such suppression was somewhat stronger for PCDD than PCDF. It was also found that 300°C might be a suitable sludge decomposition temperature when using SDG as suppressant in PCDD/Fs inhibition. Moreover, the oxygen content only slightly influenced upon the suppression capabilities. The results revealed that SO2 was more effective than NH3 in suppressing PCDD/Fs formation, with suppression efficiencies of 61.9% and 38.6%, respectively. Remarkably, the formation of OCDD/Fs was inhibited least and even further formation could occur when both NH3 and SO2 were injected together as simulated flue gas. The results exposed that NH3 and SO2 might be the primary PCDD/Fs suppressants resulting from mild thermal decomposition of sludge. However, further parameters should be evaluated and adjusted before SDG could be used in large-scale applications.


INTRODUCTION
The amount of municipal solid waste (MSW) generated in China is increasing impressively, from 148.6 million ton in 2003 to 170.8 million ton in 2012, in line with the figures for economic development (National Bureau of Statistics of China, 2004 and2013).MSW Incineration is already widely applied in MSW treatment in China: in 2012 over 25% of the MSW was disposed of by burning.However, incineration is still controversial for it might cause secondary pollution, in particular by the emission and eventual deposition of polychlorinated dibenzo-pdioxins and dibenzofurans (PCDD/Fs) (Olie et al., 1977;Chen et al., 2008;Gao et al., 2009;Ngoc and Schnitzer, 2009;Xu et al., 2009a, b).Much attention is paid to such PCDD/Fs because of their potential environmental impacts (Suzuki et al., 2005;Shih et al., 2006;Lee et al., 2007;Zheng et al., 2008).In most thermal processes, including incineration and pyrometallurgy, dioxins are mainly formed by heterogeneous reactions, in which fly ash is the key catalyst carrier.Numerous studies have been performed on the inhibition of PCDD/Fs by chemical suppressants, especially N-or S-containing compounds, such as NH 3 , (NH 4 ) 2 SO 4 , urea and SO 2 (Takacs and Moilanen, 1991;Addink et al., 1996;Raghunathan and Gullett, 1996;Samaras et al., 2000;Kuzuhara et al., 2005;Pandelova et al., 2007;Kasai et al., 2008).The suppressants are added to the raw MSW or injected into incinerator flue gases.In addition, Hunsinger et al. (2007) designed a system on recirculation of SO 2 for suppression of PCDD/F and that this avoids the continuous addition of sulphur or nitrogen compounds.However, the cost of supplying such substances is tangible, so finding cheap and effective suppressants has been a topical subject for R&D initiatives.
Sludge decomposition gases (SDG), evolving from mild thermal decomposition of sewage sludge, contain NH 3 and SO 2 as well as other, unidentified N-and S-compounds (Lindbauer et al., 1992;Hong et al., 2009;Lu et al., 2013;Chen et al., 2014).It was observed that these SDG are able to suppress up to 69.5% of the PCDD/Fs amount and 78.7% of TEQ (Yan et al., 2012), yet the dosage, temperature, moisture and oxygen content during sludge decomposition could all affect the release characteristics of actively inhibiting N-and S-compounds.When the sludge decomposition temperature was raised from 140°C to 170°C, the emission rate of NH 3 could expand 120% (Deng et al., 2009).When sludge was treated at 260°C, Chun et al. (2012) observed that the concentration of NH 3 was ca.three times higher than that at 225°C (Chun et al., 2012).Accordingly, the oxygen content of the carrier gas used to heat sludge could influence upon the emission characteristics of N-and Scompounds (Cusidó and Soriano, 2011;Lu et al., 2013;Widman et al., 2013).Therefore, the eventual suppression of PCDD/Fs and I-TEQ by vapours generated during sludge decomposition could be influenced by sludge dosage, its treatment temperature, and the moisture and oxygen content of the decomposition gases.
The suppression effect of NH 3 and SO 2 was studied before by several researchers.Adding compounds containing either N or S at times led to relatively limited inhibitory effects (Ruokojärvi et al., 2004;Pandelova et al., 2007;Hajizadeh et al., 2012;Wu et al., 2012).For instance, only 70% of PCDD/Fs inhibition efficiency was achieved using 390 ppm SO 2 as an additive (Ogawa et al., 1996), whereas 1% w/w of amidosulfonicacid (ASA) would lead to 96% reduction of PCDD/Fs.
In metallurgical processes sulfur is less desirable.British Steel patented the use of urea in inhibiting PCDD/F formation.A series of nitrogen compounds were tested in the suppression of PCDD/F formation from highly reactive filter dust, sampled from an iron ore sintering belt: triethanolamine (TEA; 0.2, 0.5, 1 and 5%), Ca(OH) 2 (2 and 5%), NaOH (2 and 5%), urea (1%) and ammonia (200 ppm) added as aqueous ammonia to the gas phase.The best effect was reached with strongly basic substances, yet gaseous ammonia was found to be ineffective (Hell et al., 2000b).In another study simple housekeeping measures combined with suppression successfully reduced the PCDD/F-emissions arising from a manganese ore sintering belt (Brown et al., 2003).
This paper examines the changes in concentration and profile of the toxic 2,3,7,8-substituted PCDD/F congeners obtained while using model fly ash (MFA) samples in suppression tests related to the dosage, temperature and oxygen content during sludge decomposition.Most importantly, these experiments also established the appropriate conditions to inhibit the formation of PCDD/Fs efficiently.In addition, the effects of injecting NH 3 together with SO 2 into the flue gas on the concentration of PCDD/F congeners have been considered.The results could confirm the important role of both NH 3 and SO 2 in the suppression of PCDD/Fs by SDG.

Experimental Materials
In this study a synthetic model fly ash was used to simulate PCDD/Fs formation.This model fly ash was chosen with similar characteristics to the fly ash produced from MSWI (Hell et al., 2000a) to allow an unequivocal interpretation of eventual changes of fly ash state and composition.This model fly ash is composed of (in wt.%) 91.8% of SiO 2 , 3% of activated carbon, 5% of NaCl and 0.2% of CuCl 2 , corresponding to a fly ash load of ca.3.1% Cl, 0.1% Cu, and 3% C.
The sludge samples were collected from a municipal wastewater treatment plant in Shanghai.The raw sludge samples were dried for 24 h in an oven at 105°C, and the resulting sample is then referred to as dried sludge (DS).The dried samples have a high N content (4.87 wt.%) and a low S content (0.79 wt.%).Their moisture content and ultimate or elemental analysis are presented in Table 1.

Experimental Set-up
Thermal decomposition of the DS samples were accomplished using the experimental set-up described in Fig. 1.The same apparatus was also used for the PCDD/Fs' suppression experiments.It comprises a tubular furnace consisting of three sections a, b, and c (0.5 m for each section) and featuring independent heaters and temperature controllers to maintain the desired temperature profile.The sludge and the model fly ash were each separately located, between two quartz cotton layers situated in a short quartz tube and in the internal quartz tube respectively.The external and internal diameter of the quartz reactor tube was 45 and 30 mm.The simulated flue gas (300 mL/min; 12% O 2 in N 2 ) was divided into two parts (Gas A and Gas B). Gas A consisted of nitrogen (264 mL/min) yet it featured different contents of oxygen when it was used to prepare the sludge decomposition gases (SDG) used as suppressant.Gas B contained the oxygen necessary to conduct a PCDD/F generation test at a fixed concentration of 12 vol.%.The SDG and Gas B were well mixed at the end of the short quartz tube and then these mixed gases passed through the fixed bed of model fly ash, to inhibit the formation of PCDD/Fs.The experimental test conditions are listed in Table 2 and are further explained in the Experimental Design (see 2.4.).In general, laboratory experiments with model fly ashes not necessarily reflect the situation in real application.
The experimental procedure was established to take into account the unsteady evolution of volatile suppressants during the additional heating of dried sludge.Before each experiment, this DS was preheated for 3.5 min up to 300°C (or to 350°C and 250°C, for test runs B-4 and C-4, respectively) and from then on the suppressant gases were directed to flow over the model fly ash.After 15 minutes a second portion of sludge is pushed into the heating zone, so that a new surge of suppressant gases is produced.
The conditions of the de novo tests (A) and concomitant suppression tests are: 300 °C, 50 min, 12 vol.% O 2 。 the different PCDD/Fs congeners.The GC temperature program was optimized as follows: splitless injection of 1 µL at 150°C, initial oven temperature of 150°C for 1 min, then increased at 25 °C/min to 190°C, finally increased at 3 °C/min to 280°C and held for 20 min at the final temperature.All tests were conducted in duplicate, except for the tests A, which were repeated five times; the arithmetic average of the results was further used in our analysis of suppression.

Experimental Design
Test A is considered to be a blank test, i.e. a test without the addition of suppressing sludge-derived S-and N-compounds, supplied as part of the carrier gas stream during the de novo PCDD/Fs formation tests on model fly ash.
Then, in the series B, and C sludge samples were tested as suppressants: the series B and C compare a double SDG dosage (B) to a single (C), whereas series D tests the effect of specific mixes of NH 3 and SO 2 in suppressing PCDD/F formation (cf.Table 2); for samples DS (Shanghai dried sludge, high in N, low in S) eight distinct test conditions were established, namely B (1 g + 1 g DS, per 2 g of model fly ash) and C (0.5 g + 0.5 g DS, per 2 g of model fly ash).
The tests use three different treatment temperatures (250°C, 300°C and 350°C) to heat the DS, and also the oxygen content of the carrier gas used during sludge decomposition is varied (0, 6, and 12 vol.%).However, the oxygen content of the simulated flue gas flowing over the model fly ash was always 12 vol.%.
Lastly, in order to confirm the inhibition effect of NH 3 and SO 2 and study the inhibition mechanism of SDG, three distinct conditions are tested, namely D. D-1 and D-2 focus on investigating the inhibition effect of NH 3 and SO 2 separately.The purpose of D-3 is to study the eventual synergetic effect of NH 3 and SO 2 further.
In the experiments A the amount of PCDD/Fs generated was determined separately in both the gas phase and the residue.Since the PCDD/Fs largely remained in the model fly ash (ca.99.9%), however, the fly ash and gas phase were analyzed together as a lumped sample in the further tests.

PCDD/F Synthesis and Inhibition
Experiment A was conducted five times to confirm the total PCDD/Fs produced from the model fly ash.The average concentration of total PCDD/Fs was 2534 ± 222 ng/g (or 77.7 ± 32.9 ng I-TEQ/g).The PCDD/F congener distribution is dominated by the PCDF, with 70.0 ± 2.9% of the total PCDD/F, and also by OCDF (41.3 ± 4.1%) and OCDD (25.2 ± 1.65%).The best defined congeners are OCDD (± 9%) and 1234678-HpCDF (± 10%).Unexpectedly, the least reproducible congener was 23478-PeCDF, which explains the width of the confidence interval of the I-TEQ-value.The blank experiment showed an unusually high weight average degree of chlorination, with values of 7.81 for Cl-PCDD and 7.31 for Cl-PCDF.Such a remarkably high average level of chlorination is typical for model fly ash incorporating CuCl 2 as a catalyst.
The I-TEQ values of the sludge samples were always quite low (Stevens et al., 2001;Lu et al., 2012) and almost no PCDD/Fs would desorb from sludge and enter with the SDG into the fly ash model system.
During the suppression tests with SDG (experiment series B and C) the reduction efficiencies of 2,3,7,8-substituted PCDD/Fs always exceeded 80%; especially experiments B-4, with a reduction of 99.9% in the PCDD/Fs concentration and of 99.2% in the I-TEQ value scored highly (Table 2).The high concentration of NH 3 and SO 2 may account tentatively for this result, yet the effect of unidentified Sand N-compounds cannot be excluded.Similar suppression effects could be achieved bythe injection of, e.g., sulfur dioxide, ammonia, dimethylamine, and methylmercaptan on particle-phase PCDD/Fs (Ruokojärvi et al., 1998).These results also show that several factors of influence, including the sludge dosage, and the temperature and oxygen content during sludge decomposition, could directly or indirectly affect the suppression efficiencies obtained for the PCDD/Fs.
The least favorable results were obtained during the suppression tests with gaseous NH 3 and SO 2 solely (experiments D).Suppression efficiency values computed on a basis of PCDD/Fs were quite low during these tests, especially for D-3, namely 0.3%.However, in some experiments also a dramatic increase in chlorination level occurred, as shown in Table 2. Accordingly, the values of I-TEQ decreased from 77.7 ng I-TEQ/g to less than 19.0 ng I-TEQ/g with the combined addition of gaseous NH 3 and SO 2 .This result shows that a disappointing effect on the total amount of PCDD/Fs may be accompanied by a remarkable decrease in the toxicity equivalent I-TEQ.

PCDD/F Inhibition by SDG Effect of DS Dosage
The amount of inhibitor applied could greatly influence the suppression effect of PCDD/Fs.For instance, 0.5 wt.% of sodium ammonium hydrogen phosphate (NAHF) reduced the particle-phase PCDD/Fs by 40% and this value would be raised to 90% when 1.0 wt.% of NAHF was added.Furthermore, the reduction efficiencies of urea on PCDD/Fs formation at 0.1%, 0.5% and 1.0% of the fuel feed were 64%, 75% and 90%, respectively (Ruokojärvi et al., 2001).Most importantly, the efficiency of reducing PCDD/Fs formation by addition of sulfur is strongly influenced by the S/Cl mole ratio (Lindbauer et al., 1992, Raghunathan and Gullett 1996, Duo and Leclerc 2004), and thus the DS dosage could further influence the suppression efficiency of PCDD/Fs.
In these experiments the suppression efficiency was monitored after addition of 1 + 1 g (B-1, B-2 and B-3) or 0.5 + 0.5 g (C-1, C-2 and C-3) DS heated at 300°C.The three B-experiments attained a suppression efficiency of 95.9 to 99.94% (PCDD/Fs) or of 98.84 to 99.33% (I-TEQ).Even when reducing the DS amount from 1 + 1 g to 0.5 + 0.5 g, the PCDD/F inhibition efficiency was still above 80%.
In almost all above cases the suppression efficiency was somewhat stronger for PCDD than for PCDF, except for B-2 and B-4, where essentially equally high values of > 99.9% were attained.This trend was consistent with the results reported by Chang et al. (2006), who added elemental sulfur to the fly ash from a municipal waste incinerator, which was reacted at 300°C in a gas stream.However, when added to the flue gas sulfur has a greater inhibitory effect on PCDF formation than on PCDD formation.
In general, NH 3 was the primary PCDD/F suppressant resulting from DS decomposition and a clear decrease in the concentration of both PCDD and PCDF congeners occurred when NH 3 was used as an additive.
No characteristic changes in congener distribution would be observed in earlier studies (Addink et al., 1996;Hajizadeh et al., 2012).In this study this congener distribution is of paramount importance: because of the very high average chlorination level of PCDD and PCDF any deviation from full octa-chlorination assumes great importance.
The blank test sample is composed of 30.0 ± 2.9% PCDD and 70.0 ± 2.9% PCDF, forming the balance.
The B-series shows the deepest suppression of the I-TEQ value, with values ranging from 98.84 to 99.33%.During the C-series the SDG is halved, reducing the suppression efficiency to ca. 90% (test runs C-1 to C-3).Even the Dseries still attains impressive inhibition of the I-TEQ values (Table 3).

Effect of DS Decomposition Temperature
The inhibitory effects of SDG on PCDD/Fs formation could be due to many distinct factors, including the type, the concentration, and the emission characteristics of Nand S-compounds (Furrer et al., 1998;Stieglitz et al., 1998).Since the composition of DS is complex and the evolving S-and N-compounds unidentified and possibly highly variable, the decomposition temperature might influence the ability of SDG in suppressing PCDD/Fs formation.
The results of Yan et al. (2012) indicated that decomposition at 200°C was high enough for PCDD/F inhibition (Yan et al., 2012).In general, the higher the decomposition temperature applied, the shorter the emission time period of N-and S-compounds would last.Comparison between the experiments B-3 and B-4 shows that the suppression effects of PCDD/Fs when DS was heated at 350°C were still stronger than at 300°C.A comparison between the experiments C-3 (300°C) and C-4 (250°C) also shows better results for the higher decomposition temperature.Considering energy savings and emission reduction, the sludge decomposition temperature should not be selected too high, and thus 300°C was selected as the most appropriate decomposition temperature.
Comparison between B-3 (C-3) and B-4 (C-4) also shows an increase in chlorination levels of both PCDD and PCDF when the sludge decomposition temperature was changed: the fraction of OCDD and OCDF increased (see Figs. 3(a) and (b)).In the presence of NH 3 the suppression of lower chlorinated PCDDs was larger than for highly chlorinated ones (Hajizadeh et al., 2012).Hence, these results were in agreement with some previously reported results using sulfur compounds as inhibitors (Lee et al., 1998).However, no clear trend was found in the reduction of PCDF congeners in the presence of NH 3 .Conversely, SO 2 suppression effects were found to be higher for high chlorinated PCDF (Pandelova et al., 2007).The results in this study revealed that the reduction of low substituted congeners with high toxic equivalency value were more favorable at 250°C and 350°C than at the sludge decomposition temperature at 300°C.
A comparison between B-3 (C-3) and B-4 (C-4) also shows an increase in chlorination levels of both PCDD and PCDF when the sludge decomposition temperature was varied: the fraction of OCDD and OCDF increased (see Figs. 3(a) and 3(b)).In the presence of NH 3 the suppression of lower chlorinated PCDDs was larger than for highly chlorinated ones (Hajizadeh et al., 2012).Hence, these results were in agreement with some previously reported results using sulfur compounds as inhibitors (Lee et al., 1998).However, no clear trend was found in the reduction of PCDF congeners in the presence of NH 3 .Conversely, SO 2 suppression effects were found to be higher for high chlorinated PCDF (Pandelova et al., 2007).The results in this study revealed that the reduction of low substituted congeners with high toxic equivalency value were more favorable at 250°C and 350°C than at the standard sludge decomposition temperature at 300°C.
By comparison and analysis, we found that sludge could emit more PCDD/Fs suppression gases, such as NH 3 , SO 2 and HCN, in case the decomposition temperature was raised step by step.However, further work is needed to establish the best mode of temperature rise in terms of PCDD/F inhibition results.

Effect of Oxygen Content during DS Decomposition
The emission characteristics of these compounds liberated during DS decomposition could be affected by the oxygen content and thus the suppression efficiencies would be varied.In an industrial context it is plausible to use the residual heat of flue gas to treat the sludge and thus to produce a large amount of SDG.In addition, after a deep decomposition the sludge will act as a fuel supplement: dry sludge has the same calorific value as lignite and a low chlorine content (Werther and Ogada, 1999;Francisca Gómez-Rico et al., 2005).
The sludge decomposition atmosphere also could influence the emission characteristics of PCDD/Fs suppression gases, e.g., by a partial oxidation of active suppressants.In general, 100% N 2 enhanced the emission of NH 3 and the emission of SO 2 increased under the simulated flue gas condition.
As described in Fig. 2, no significant and systematic changes were observed in the PCDD/F suppression efficiencies when the oxygen content of the sludge decomposition atmosphere was varied (0%, 6% and 12%), especially for 1 + 1 g DS.Comparison between C-1, C-2 and C-3 showed that C-1 and C-3 had a stronger suppression effect on PCDD/Fs than C-2, revealing that DS heated in 6% O 2 of carrier gas may release much less N-and S-compounds.
The result for the PCDD/F congeners was very different, yet without any clear trends in the reduction of their concentration in the presence of SDG.Interestingly, the chlorination level of PCDD/Fs when DS heated in 6% of decomposition atmosphere was higher than that of 0% and 12%.As for I-TEQ value, no obvious differences were observed, and the suppression efficiencies of them were ca.90%.
These tests with SDG can be concluded with the following statements: 1.A deep reduction (99.9%) of PCDD/F formation was realized during two duplicated test runs: B-2 and B-4.These tests feature a somewhat lower reduction of the I-TEQ values, of 98.9 and 99.2%, respectively.
2. The identity and the nature of the active suppressants are still unidentified.Two plausible contenders for such a role are gaseous NH 3 and SO 2 .

PCDD/F Inhibition by Gaseous NH 3 and SO 2
During the previous tests SDG has proved to be a highly efficient suppressant of PCDD/F-formation.Since the precise composition of SDG is basically unknown and also varies with the origins, composition and thermal treatment conditions of the sludge three supplemental tests were also conducted using a mixture of 1000 ppm NH 3 and SO 2 of known composition: 900 ppm NH 3 and 100 ppm SO 2 , 100 ppm NH 3 and 900 ppm SO 2 , and 500 ppm NH 3 and 500 ppm SO 2 .

Injection of 900 ppm NH 3 and 100 ppm SO 2
The results obtained with (mainly) NH 3 injection are presented in Table 3.The suppression efficiency recorded on PCDD/Fs was 38.7%, which was well within the range (34-75% for the solid phase and 21-40% for the gas phase) reported by Hajizadeh et al. (2012).Interestingly, suppression was much stronger for PCDD than PCDF: 54.9% and 31.6%,respectively.Moreover, when compared with the base case both chlorination levels increased, with a significant decrease in I-TEQ value as a consequence.The reduction of lower chlorinated congeners was greater than that for high chlorinated ones, for both PCDD and PCDF congeners (see Figs. 3(a) and 3(b)).For example, 1,2,3,7,8-PeCDD and 1,2,3,7,8-PeCDF were reduced by 92.9% and 93.8% respectively, whereas OCDD and OCDF reduced by 48.1% and 11.0% respectively.These results were also comparable with the results reported by Ruokojärvi et al. (1998), who found that the total PCDD/F concentration was reduced between 42% and 78%.However, varying the NH 3 concentration and the temperature of the fly ash would affect the PCDD/Fs congener distribution.So results from this work showed some difference in PCDD/F congeners with other similar works.
In contrast with the addition of SDG, 2,3,4,7,8-PeCDF is not the major contributor to I-TEQ, since it accounts for only 13.8%, whereas the percentage of 2,3,4,6,7,8-HxCDF and 1,2,3,4,6,7,8-HpCDF are 19.5% and 24.6% respectively.In addition, PCDF is still the major contributor to I-TEQ, which accounts for 93.8% of total I-TEQ.This could be mainly attributed to the exceptionally high levels of PCDF in terms of I-TEQ.

Injection of 100 ppm NH 3 and 900 ppm SO 2
The suppression efficiency of PCDD/Fs was found to be 61.9%,which was within the range (42-75% for the solid phase and 24-57% for the gas phase) reported by Hajizadeh et al. (2012).It also agrees with the work of Shao et al. (2010), who found that the addition of SO 2 to the inlet gas of a model ash treatment test at 350°C could reduce the PCDD/F levels by about 90%, with a S/Cl ratio of 0.25.In this test, the S/Cl ratio exceeded 20, but it is unclear why the reduction of PCDD/Fs was far lower than in the results of Shao et al. (2010).Furthermore, SO 2 was more effective than NH 3 in suppressing PCDD/Fs formation, consistent with the results reported by Hajizadeh et al. (2012).
In contrast with the addition of 900 ppm NH 3 , the suppression effects on PCDD and PCDF were quite close, 66.0% and 60.2%, respectively.Moreover, the reduction efficiencies of OCDD and OCDF were only 60.7% and 46.4%, whereas the average values of suppression efficiencies of the lower chlorinated exceed 90%.Therefore, the I-TEQ Table 3. Results regarding the suppression of PCDD and PCDF.value decreased by 94.8%, even more than with the SDG produced from 0.5 + 0.5 g DS.In addition, 1,2,3,4,6,7,8-HpCDF was also the major contributor to I-TEQ, which accounted for ca.31.7%.
Injection of 500 ppm NH 3 and 500 ppm SO 2 Remarkably, during the two tests with a joint injection of 500 ppm NH 3 and 500 ppm SO 2 a negative reduction for the sum PCDD/Fs was observed.Although a clear decrease in the concentrations of PCDD/Fs occurred when either 500 ppm NH 3 or 500 ppm SO 2 was injected (Ruokojärvi et al., 1998;Ke et al., 2010;Wu et al., 2012), only 0.3% of PCDD/Fs reduction was achieved in these experiments when they were injected together.The experiments were conducted in duplicate and the results were well reproducible, with values for the standard variation of 11% for PCDD, 8%% for PCDF and 16% for I-TEQ.
Strangely enough, a strong reduction was observed on all lower chlorinated PCDD, ranging from 83.9% for 2378-TeCDD to 62.3% for 1234678-HpCDD, yet only 0.4% for OCDD.The situation was fairly similar for PCDF, with strong reductions for lower chlorinated PCDF, culminating in a 92.5% reduction for 23478-PeCDF and lowest for the two main PCDF congeners: only 20.8% for 1234678-HpCDF and a negative reduction of -41.1% for OCDF.Also, the reductions on the lower chlorinated congeners were comparable with the suppression effects of NH 3 and SO 2 .Thus, the suppression efficiency of total I-TEQ would reach 75.6%.
Thus this test has led to a very strange situation: although the total amount of suppression calculated on PCDD/Fs was negligible, the effect on I-TEQ was quite rewarding.The following two hypotheses may help to explain this strange result.One might be that the acidic effect of SO 2 and the basic effect of NH 3 and that equal amounts neutralizes somehow.Another possible reason is that the concentrations of NH 3 and SO 2 were not large enough to suppress the activity of formation of PCDD/Fs in our fly ash model system.Most importantly, the abatement of SO 2 , NH 3 and other nitrogen and sulphur compounds need to be considered if such gases are used for suppression of PCDD/F formation.Therefore levels of 500 or 900 ppm addition might have consequences for the flue gas cleaning a therefore cost.

CONCLUSIONS
Five de novo tests were conducted using model fly ash (MFA) to establish the average level of PCDD/F formation under standard conditions (300°C, 12 vol.%O 2 ).Such standard PCDD/F formation was suppressed during the following eleven inhibition tests, each conducted in duplo, and using either sludge decomposition gases (SDG) rich in nitrogen-and sulfur compounds, or else a mixture of NH 3 and SO 2 as suppressant.The sludge dosage, and the treatment temperature and oxygen content during sludge decomposition might influence the emission characteristics of N-and Scompounds, which are effective PCDD/Fs suppressants.The most suitable experimental condition identified in the paper was probably: using 0.5 + 0.5 g of dried sludge DS, heated at 300°C, to treat 2 g of MFA.The oxygen content of the decomposition atmosphere did not play an important role in PCDD/Fs suppression.Clear reductions of PCDD/Fs were also observed with the injection of NH 3 and SO 2 .
Obviously, the suppression by gaseous NH 3 and SO 2 passes over a poisoning or other form of neutralization of the CuCl 2 catalyst active in converting carbon into PCDD/Fs.A first step in such a deactivation process is adsorption.It seems likely that larger S-and N-molecules easier condense and adsorb onto the reacting system than do NH 3 and SO 2 .
According to the relevant literature the S/Cl-ratio is of paramount importance.Moreover, the signature of the PCDD/F would be very similar in the original standard system and its inhibited counterparts.The results obtained in this study demonstrate that minor differences in fingerprint may be very consequential in the results obtained, in particular when these are expressed in I-TEQ units.
Further work is needed to evaluate the PCDD/Fs suppression effect on pilot scale tests in incinerators, to apply this inhibition technique using SDG to real MSWI process.

Fig. 3 .
Fig. 3. Congener profiles of the PCDD (a) and PCDF (b) without and with various inhibitors.

Table 1 .
Moisture content and ultimate analysis of the dried sludge sample before the experiment (wt.%).

Table 2 .
Experimental design conditions.