Dioxin emissions after installation of a polishing wet ... - BVSDE

Dioxin emissions after installation of a polishing wet ... - BVSDE Dioxin emissions after installation of a polishing wet ... - BVSDE

bvsde.paho.org
from bvsde.paho.org More from this publisher
18.11.2013 Views

Chemosphere 61 (2005) 405–412 www.elsevier.com/locate/chemosphere Dioxin emissions after installation of a polishing wet scrubber in a hazardous waste incineration facility Carl-Johan Löthgren * , Bert van Bavel Department of Natural Sciences, Man-Technology-Environment Research Centre, Örebro University, SE-701 82 Örebro, Sweden Received 22 December 2003; received in revised form 10 January 2005; accepted 15 February 2005 Available online 13 April 2005 Abstract Dioxin levels measured after wet scrubbing systems have been found to be higher than levels measured before the scrubber. It is believed that there is an adsorption of PCDD/Fs on plastic materials in the scrubber. The PCDD/F levels after a polishing wet scrubber were followed continuously for 18 months using long-time sampling equipment at a hazardous waste incineration facility in Sweden. Each sampling period lasted two weeks. It was found that the levels during and shortly after start-up periods were elevated. The decline was very slowly, which supports a memory effect in the scrubber. Further, a multivariate model showed that the relation between different homologues changed over time, which is in agreement with a desorption model, taking into account the vapour pressures for different congeners. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: PCDD/PCDF; Memory effect; Flue gas cleaning; MVDA; Incineration 1. Introduction Dioxin formation from incineration is known since 1977 when dibenzo-dioxins and dibenzo-furans were found in fly ashes and flue gases (Olie et al., 1977). Their toxic effects are well known and dioxins have been in focus in the environmental discussions for almost 20 years (Van den Berg et al., 1998). Throughout these years efforts have been made to minimize the formation and emission of dioxins from incineration facilities, both by means of primary measures, i.e. combustion control and by secondary measures such as the supply of additives or catalysts to the flue gas treatment system. One * Corresponding author. Tel.: +46 1930 5100; fax: +46 1957 7027. E-mail address: carl-johan.lothgren@sakab.se (C.-J. Löthgren). of the most used methods is the injection of activated carbon into the flue gas channel just before a bag house filter, where the carbon together with the fly ash form an adsorption layer on the bag on which dioxins are adsorbed. Emissions after such an installation are known to be well below the EU emission limit of 0.1 ng/m 3 (Chang and Lin, 2001). According to the EU directive 94/67/EG and EU directive 2000/76/EC emission measurements of PCDD/Fs only have to take place once or twice yearly and such measurements are made under normal, i.e. good operation conditions. Typical CO levels are well below the stipulated 50 mg/m 3 normal dry gas. The emission levels under other operating conditions have been poorly studied until recently, when Belgian MSW incinerators started continuous flue gas sampling for dioxin analyses. Blumenstock et al. (2000) found an increase in PCDD/F levels after a period with bad operating conditions in a pilot scale incinerator. This 0045-6535/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2005.02.015

Chemosphere 61 (2005) 405–412<br />

www.elsevier.com/locate/chemosphere<br />

<strong>Dioxin</strong> <strong>emissions</strong> <strong>after</strong> <strong>installation</strong> <strong>of</strong> a <strong>polishing</strong> <strong>wet</strong><br />

scrubber in a hazardous waste incineration facility<br />

Carl-Johan Löthgren * , Bert van Bavel<br />

Department <strong>of</strong> Natural Sciences, Man-Technology-Environment Research Centre, Örebro University, SE-701 82 Örebro, Sweden<br />

Received 22 December 2003; received in revised form 10 January 2005; accepted 15 February 2005<br />

Available online 13 April 2005<br />

Abstract<br />

<strong>Dioxin</strong> levels measured <strong>after</strong> <strong>wet</strong> scrubbing systems have been found to be higher than levels measured before the<br />

scrubber. It is believed that there is an adsorption <strong>of</strong> PCDD/Fs on plastic materials in the scrubber. The PCDD/F levels<br />

<strong>after</strong> a <strong>polishing</strong> <strong>wet</strong> scrubber were followed continuously for 18 months using long-time sampling equipment at a hazardous<br />

waste incineration facility in Sweden. Each sampling period lasted two weeks. It was found that the levels during<br />

and shortly <strong>after</strong> start-up periods were elevated. The decline was very slowly, which supports a memory effect in the<br />

scrubber. Further, a multivariate model showed that the relation between different homologues changed over time,<br />

which is in agreement with a desorption model, taking into account the vapour pressures for different congeners.<br />

Ó 2005 Elsevier Ltd. All rights reserved.<br />

Keywords: PCDD/PCDF; Memory effect; Flue gas cleaning; MVDA; Incineration<br />

1. Introduction<br />

<strong>Dioxin</strong> formation from incineration is known since<br />

1977 when dibenzo-dioxins and dibenzo-furans were<br />

found in fly ashes and flue gases (Olie et al., 1977). Their<br />

toxic effects are well known and dioxins have been in<br />

focus in the environmental discussions for almost 20<br />

years (Van den Berg et al., 1998). Throughout these<br />

years efforts have been made to minimize the formation<br />

and emission <strong>of</strong> dioxins from incineration facilities, both<br />

by means <strong>of</strong> primary measures, i.e. combustion control<br />

and by secondary measures such as the supply <strong>of</strong> additives<br />

or catalysts to the flue gas treatment system. One<br />

* Corresponding author. Tel.: +46 1930 5100; fax: +46 1957<br />

7027.<br />

E-mail address: carl-johan.lothgren@sakab.se (C.-J. Löthgren).<br />

<strong>of</strong> the most used methods is the injection <strong>of</strong> activated<br />

carbon into the flue gas channel just before a bag house<br />

filter, where the carbon together with the fly ash form an<br />

adsorption layer on the bag on which dioxins are adsorbed.<br />

Emissions <strong>after</strong> such an <strong>installation</strong> are known<br />

to be well below the EU emission limit <strong>of</strong> 0.1 ng/m 3<br />

(Chang and Lin, 2001).<br />

According to the EU directive 94/67/EG and EU<br />

directive 2000/76/EC emission measurements <strong>of</strong><br />

PCDD/Fs only have to take place once or twice yearly<br />

and such measurements are made under normal, i.e.<br />

good operation conditions. Typical CO levels are well<br />

below the stipulated 50 mg/m 3 normal dry gas. The<br />

emission levels under other operating conditions have<br />

been poorly studied until recently, when Belgian MSW<br />

incinerators started continuous flue gas sampling for dioxin<br />

analyses. Blumenstock et al. (2000) found an increase<br />

in PCDD/F levels <strong>after</strong> a period with bad<br />

operating conditions in a pilot scale incinerator. This<br />

0045-6535/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.<br />

doi:10.1016/j.chemosphere.2005.02.015


406 C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412<br />

was later confirmed by several other authors (Gullett<br />

et al., 2000; Zimmermann et al., 2001; Gass et al., 2002).<br />

Gass et al. (2002) investigated the <strong>emissions</strong> from a full<br />

scale incinerator and obtained the same results as earlier<br />

were shown in pilot scale.<br />

Wet scrubbing, i.e. the physical or chemical absorption<br />

<strong>of</strong> gases in a liquid, has been used for a long time<br />

for the removal <strong>of</strong> acid components in flue gases from<br />

incineration facilities because <strong>of</strong> its good removal efficiency<br />

(Onnen, 1972; Korell et al., 2003). High dioxin<br />

<strong>emissions</strong> have, however, been found <strong>after</strong> <strong>wet</strong> scrubbers<br />

when they have been placed as the final step in a flue gas<br />

cleaning system. It has further been shown that this is<br />

caused by the so called memory effect, where dioxins<br />

are adsorbed on plastic details such as scrubber fillings<br />

when the formation and emission from the incinerator<br />

is high. The adsorbed dioxins are then desorbed slowly<br />

when the plant is running under more stable conditions.<br />

A comparative study using a pilot scale scrubber made <strong>of</strong><br />

glass showed no such increased emission levels (Hunsinger<br />

et al., 1998). This memory effect can be distinguished<br />

from high <strong>emissions</strong> caused by bad combustion<br />

through its different congener pattern (Kreisz et al.,<br />

1996; Gass and Neugebauer, 1999; Adams et al., 2001;<br />

Giugliano et al., 2002).<br />

To comply with the hazardous waste incineration<br />

directive (EU directive 94/67/EG), which came in force<br />

on the first <strong>of</strong> July 2000 in Sweden, a hazardous waste<br />

treatment facility in Kumla, Sweden was equipped with<br />

an enhanced flue gas cleaning. A <strong>wet</strong> scrubber was installed<br />

<strong>after</strong> the existing semi-dry system, which consisted<br />

<strong>of</strong> a spray dryer and a fabric filter with activated<br />

carbon injection. Already five weeks <strong>after</strong> the new scrubber<br />

was taken into operation, elevated PCDD/F <strong>emissions</strong><br />

were observed. In order to learn more about the<br />

dynamic behaviour <strong>of</strong> the scrubber as to dioxin <strong>emissions</strong><br />

a continuos sampling device was installed in July<br />

2001. The results from these measurements are presented<br />

in this article.<br />

2. Material and methods<br />

The hazardous waste treatment facility studied<br />

mainly consists <strong>of</strong> an incineration line with a rotary kiln,<br />

secondary combustion chamber, boiler and a flue gas<br />

cleaning system. The flue gas cleaning consists <strong>of</strong> a<br />

semi-dry absorber, a fabric filter and the new <strong>wet</strong> scrubber.<br />

In the absorber lime milk is injected to neutralize<br />

the main part <strong>of</strong> HCl and SO 2 present in the flue gas. Between<br />

the absorber and the fabric filter activated carbon<br />

is injected for dioxin removal. The incineration temperature<br />

in the rotary kiln is above 1100 °C. Outlet temperature<br />

from the boiler is 280 °C. In the semi-dry flue gas<br />

cleaning system the temperature is lowered to 160 °C<br />

and in the scrubber the temperature is 70 °C.<br />

The scrubber consists <strong>of</strong> a quench and a 28 m high<br />

scrubbing tower, 5.5 m in diameter, filled with filling<br />

material <strong>of</strong> polypropylene (Rauschert hiflow rings type<br />

50–6). The scrubber and quench shell are made <strong>of</strong><br />

glass fibre reinforced plastic (GRP) Derakane 470.<br />

After the scrubber the flue gas fan is situated and the<br />

gases leave the plant through a 60 m high chimney,<br />

which interior surface also is made <strong>of</strong> GRP. A schematic<br />

sketch <strong>of</strong> the plant and the sampling points are shown in<br />

Fig. 1.<br />

Fig. 1. Schematic sketch <strong>of</strong> the plant and indication <strong>of</strong> sampling points for dioxins. The layout <strong>of</strong> the Amesa sampling system is shown<br />

in the bubble.


C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412 407<br />

The purpose <strong>of</strong> the <strong>wet</strong> scrubber is to polish the emission<br />

levels <strong>of</strong> HCl and SO 2 to comply with the 1/2-hour<br />

averages prescribed by the Swedish implementation <strong>of</strong><br />

the European Hazardous Waste incineration directive.<br />

This is achieved by introducing H 2 O 2 to the scrubber<br />

liquid, the so called MercOx-process. The H 2 O 2 content<br />

in the scrubber liquid also increases the removal efficiency<br />

for mercury. The MercOx process has been described<br />

in detail by Korell et al. (2003).<br />

Five weeks <strong>after</strong> the first start-up <strong>of</strong> the plant with<br />

the new scrubber, dioxin levels were measured before<br />

and <strong>after</strong> the scrubber. High concentrations were found<br />

<strong>after</strong> the scrubber even though the concentrations before<br />

the scrubber were low. Further measurements during the<br />

autumn 2000 confirmed these results. In order to learn<br />

more about the dynamic behaviour <strong>of</strong> the scrubber as<br />

to dioxin <strong>emissions</strong> a continuos sampling device was installed<br />

in July 2001. Since then samples have been taken<br />

continuously whenever the plant has been in operation.<br />

The commercial sampling equipment used has been described<br />

by Funcke and Linnemann (1992) and Mayer<br />

et al. (2000) and consists <strong>of</strong> a titanium probe, a XAD-<br />

2-cartridge and a control cabinet. (Fig. 1). Flue gases<br />

are taken out isokinetically, pass the cartridge and are<br />

transferred to the control cabinet where flue gas volume<br />

and water content is measured.<br />

The sampling period for each analysis has varied between<br />

one to two weeks. Sample preparation and analyses<br />

have been made by Eur<strong>of</strong>ins GfA-laboratories in<br />

Germany according to EN1948. Eur<strong>of</strong>ins GfA is accredited<br />

according DIN EN ISO/IEC 17025:2000 and is<br />

participating on a regular basis in international intercalibration<br />

studies.<br />

3. Multivariate data treatment<br />

Multivariate data analysis (MVDA) has been used<br />

successfully in a number <strong>of</strong> investigations where complex<br />

data sets are to be evaluated, also in the field <strong>of</strong> dioxin<br />

research (Pitea et al., 1994; Buekens et al., 2000;<br />

Wang et al., 2003). Multivariate data analysis (MVDA)<br />

consists <strong>of</strong> several methods to extract information from<br />

large data matrices. In MVDA the samples (objects k),<br />

in our case the weekly dioxin measurements, are seen<br />

as points in a multidimensional space with as many<br />

dimensions as variables (i) measured on the samples.<br />

In principal component analysis (PCA) the points in<br />

the multidimensional space are projected down onto a<br />

line, a plane or a hyperplane to give the best low-dimensional<br />

representation <strong>of</strong> the data. The data matrix (X) is<br />

now approximated by a systematic part (the loading and<br />

the score vectors) and a random part (the residuals)<br />

which consists <strong>of</strong> errors in the measurements and imperfections<br />

in the approximation. The dimension <strong>of</strong> PCA is<br />

determined by cross validation.<br />

The visualisation <strong>of</strong> the PCA is normally done in diagram<br />

form. The variables are plotted in a diagram with<br />

the loading vectors as the axes. The greater the absolute<br />

value (positive or negative) <strong>of</strong> a variable is, the greater<br />

impact it has on its loading vector. Variables correlated<br />

with a specific vector can thus be identified. The objects<br />

are plotted in a diagram with the score vectors as the<br />

axes. Objects that have similar patterns will be located<br />

near each other and objects with different patterns will<br />

be more apart.<br />

Before MVDA was performed on the data set, all<br />

data were centered and scaled to unit variance. The calculations<br />

were performed on a PC with the s<strong>of</strong>tware program<br />

Simca 7.01 (Umetri AB, Umeå, Sweden). PCA is<br />

described in detail elsewhere (Wold et al., 1983; Wold<br />

et al., 1984).<br />

4. Result and discussion<br />

The sum <strong>of</strong> the different homologues for all measurements<br />

performed in the flue gas since 1998 are shown as<br />

series <strong>of</strong> time in Fig. 2. The first part shows the measurements<br />

made before the <strong>wet</strong> scrubber and the second part<br />

shows the measurements made <strong>after</strong> the <strong>wet</strong> scrubber,<br />

both short-time and long-time samplings. Each sampling<br />

is named according to the time when it was performed<br />

with the year and the consecutive week number <strong>of</strong> that<br />

year in the form yyww. When two measurements have<br />

been accomplished the same week, the individual measurements<br />

have been denoted a and b. Two maintenance<br />

stops in April and October 2002 are also indicated in the<br />

figure. The TEQ values for the measurements before the<br />

<strong>wet</strong> scrubber varied between 0.003 ng/m 3 (W9839a) and<br />

0.033 ng/m 3 (W9935). A significant higher level <strong>of</strong> TeC-<br />

DDs (0.19 and 0.5 ng/m 3 ) explains much <strong>of</strong> the rise in<br />

1999. No further investigations have been made in this<br />

study as to the cause <strong>of</strong> the results before the <strong>wet</strong> scrubber.<br />

The values shall be seen in comparison to the results<br />

from the measurements performed <strong>after</strong> the <strong>wet</strong><br />

scrubber.<br />

In the year 2000 when the <strong>wet</strong> scrubber was installed,<br />

sampling was performed simultaneously before and <strong>after</strong><br />

the <strong>wet</strong> scrubber. The results before the <strong>wet</strong> scrubber as<br />

to TEQ values were 0.010, 0.006 and 0.005 ng/m 3 , which<br />

are in the same range as earlier results, i.e. well below<br />

0.1 ng/m 3 . The TEQ-values <strong>after</strong> the <strong>wet</strong> scrubber in<br />

the sampling campaigns in year 2000 were much higher,<br />

0.18, 0.10, 0.36 and 0.45 ng/m 3 . During the first sampling<br />

campaign in May (W0021), the difference between<br />

the TEQ value before and <strong>after</strong> the scrubber is about<br />

one order <strong>of</strong> magnitude, 0.01 ng/m 3 before the scrubber<br />

compared to 0.18 and 0.10 ng/m 3 <strong>after</strong> the scrubber. In<br />

the campaign in October (W0040) this difference has increased<br />

to almost two orders <strong>of</strong> magnitude, 0.006 and<br />

0.005 ng/m 3 before the scrubber compared to 0.36 and


408 C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412<br />

Measurements<br />

100<br />

before scrubber<br />

Measurements<br />

<strong>after</strong> scrubber<br />

Maintenance<br />

stop spring<br />

2002<br />

Maintenance stop<br />

autumn 2002<br />

10<br />

ng/m 3<br />

1<br />

0.1<br />

0.01<br />

0.001<br />

9839a<br />

9839b<br />

9934<br />

9935<br />

0021<br />

0040a<br />

0040b<br />

0021a<br />

0021b<br />

0040a<br />

0040b<br />

0127<br />

0131<br />

0136<br />

0137<br />

0138<br />

0139<br />

0140<br />

0141<br />

0146<br />

0147<br />

0148<br />

0149<br />

0150<br />

0151<br />

0152<br />

0201<br />

0202<br />

0203<br />

0204<br />

0205<br />

0206<br />

0207<br />

0209<br />

0211<br />

0213<br />

0215<br />

0218<br />

0220<br />

0222<br />

0224<br />

0226<br />

0228<br />

0230<br />

0232<br />

0234<br />

0235<br />

0236<br />

0237<br />

0238<br />

0240<br />

0242<br />

0244<br />

0247<br />

0249<br />

0251<br />

0252<br />

Sum TeCDD<br />

Sum PnCDD<br />

Sum HxCDD<br />

Sum HpCDD<br />

OCDD<br />

Sum TeCDF<br />

Sum PnCDF<br />

Sum HxCDF<br />

Sum HpCDF<br />

OCDF<br />

TEQ<br />

Fig. 2. Results <strong>of</strong> dioxin measurements 1998–2002 for all homologues and the TEQ value. To the left <strong>of</strong> the dotted line the results<br />

before the scrubber <strong>installation</strong> are shown. Maintenance stops are indicated with solid lines. Each sampling period is named with year<br />

and week number in the form yyww.<br />

0.45 ng/m 3 <strong>after</strong> the scrubber. It is clear that the differences<br />

in TEQ values before and <strong>after</strong> the scrubber above<br />

all can be explained by the drastically higher levels <strong>of</strong><br />

low-chlorinated homologues <strong>after</strong> the scrubber. For<br />

example TeCDD increased from the range 0.02 to<br />

0.14 ng/m 3 before the scrubber to 0.1–0.7 ng/m 3 <strong>after</strong><br />

the scrubber, TeCDF increased from the range 0.01 to<br />

0.07 ng/m 3 before the scrubber to 3.5–5.7 ng/m 3 <strong>after</strong><br />

the scrubber and PnCDF increased from the range<br />

0.02 to 0.07 ng/m 3 before the scrubber to 1.54–3.11<br />

ng/m 3 <strong>after</strong> the scrubber. OCDD, on the other hand,<br />

was in the range 0.03–0.15 ng/m 3 before the scrubber<br />

and 0.01–0.04 ng/m 3 <strong>after</strong> the scrubber. The same<br />

applies for OCDF, which was in the range 0.02–0.19<br />

ng/m 3 before the scrubber and 0.01–0.05 ng/m 3 <strong>after</strong><br />

the scrubber.<br />

This change in composition is illustrated in the fingerprint<br />

plots in Figs. 3 and 4. Especially the fraction<br />

levels <strong>of</strong> TeCDFs have risen. Before the scrubber the<br />

average fraction <strong>of</strong> TeCDF is 8% and <strong>after</strong> the scrubber<br />

it is 49%. The somewhat deviating pattern for the measurements<br />

in 1999 can also be seen. At this time the<br />

TeCDD fractions are the most deviating homologue<br />

with values <strong>of</strong> 44% and 52% compared to 5–14% for<br />

other measurement periods before the scrubber.<br />

The fingerprint plot in Fig. 4 shows clearly decreasing<br />

values with the degree <strong>of</strong> chlorination. For the PCDFs<br />

the average level <strong>of</strong> the TeCDF is 49% and that is<br />

approximately double the amount <strong>of</strong> the PeCDFs,<br />

which is 23%, i.e. more than the double <strong>of</strong> HxCDFs<br />

(9%). HpCDFs have an average fraction <strong>of</strong> 2% and<br />

OCDFs have an average fraction <strong>of</strong> 0.2%. For PCDDs<br />

the trend is similar but to a lesser extent (TeCDD 6%,<br />

PeCDD 6%, HxCDD 4%, PnCDD 1% and OCDD<br />

0.2%).<br />

In August 2000, as soon as the results from the measuring<br />

campaign in May 2000 were available, an electrostatic<br />

precipitator (ESP) used as by-pass for the fabric<br />

filter during start-up and shut down was closed since it<br />

was suspected that PCDD/Fs were loaded into the<br />

scrubber during its operation periods and slowly released<br />

<strong>after</strong>wards, the ‘‘so called’’ memory effect described<br />

in the literature by Kreisz et al. (1997); Wevers<br />

and De Fré (1998); Giugliano et al. (2000); Adams


C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412 409<br />

Fig. 3. Homologue fingerprint for the measurements made before the <strong>wet</strong> scrubber. This includes both measurements made before and<br />

<strong>after</strong> the scrubber <strong>installation</strong>. In the years 1998–2000.<br />

Fig. 4. Homologue fingerprint for the measurements made <strong>after</strong> the <strong>wet</strong> scrubber in the period May 2000 to July 2001. High ratio <strong>of</strong><br />

low chlorinated congeners can be observed due to memory effects in the <strong>wet</strong> scrubber.


410 C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412<br />

et al. (2001). When the ESP was in operation no carbon<br />

could be injected so it seemed logical that dioxins were<br />

transported to the <strong>wet</strong> scrubber during these periods.<br />

As can be seen in Fig. 2 the dioxin levels were elevated<br />

also in the measuring campaign in October 2000 (0.36<br />

and 0.45 ng/m 3 TEQ in W0040). When the continuous<br />

sampling equipment was installed in July 2001 the<br />

TEQ-value <strong>after</strong> the scrubber was 0.52 ng/m 3 , which is<br />

in the same range as in the October measurement. In<br />

Fig. 2 it can be seen that all the homologues except<br />

OCDF and OCDD fell from this high level to a level<br />

comparable to the levels measured before the scrubber<br />

(0.003–0.03 ng/m 3 ). This decline took almost 15 months<br />

from the shut down <strong>of</strong> the ESP. In W0147, which is in<br />

the beginning <strong>of</strong> November 2001 the first result below<br />

0.1 ng/m 3 TEQ <strong>after</strong> the scrubber was obtained. The result<br />

was 0.06 ng/m 3 TEQ and two weeks later the level<br />

had decreased further to 0.04 ng/m 3 .<br />

During the period <strong>of</strong> continuos sampling presented<br />

here, two dramatically risings <strong>of</strong> TEQ levels were observed,<br />

both in the period just <strong>after</strong> a maintenance stop<br />

<strong>of</strong> the plant. At the first occasion (W0215) the TEQ<br />

value increased from 0.02 ng/m 3 to 0.25 ng/m 3 and at the<br />

second occasion (W0240) the TEQ-value increased from<br />

0.03 to 0.15 ng/m 3 . It is believed that during the start-up<br />

periods high levels <strong>of</strong> dioxins are generated which normally<br />

is not recognized since sampling normally takes<br />

place only during stable operating conditions.<br />

The high concentrations just <strong>after</strong> start-up and the<br />

somewhat exponential decline <strong>after</strong>wards indicate that<br />

dioxins are adsorbed on the olefin plastics in the scrubber<br />

and released slowly. Otherwise a sharp peak would<br />

be expected just <strong>after</strong> start-up. It is also typical that<br />

the concentrations <strong>of</strong> OCDD/F do not vary over time<br />

since it is known that the absorption for these homologues<br />

are much stronger than for the others, (Kreisz<br />

et al., 1996). The different absorption pattern for different<br />

congeners can be explained by their different vapour<br />

pressures. The strong dependence on vapour pressures<br />

with the degree <strong>of</strong> chlorination has been described by<br />

Rordorf (1989).<br />

The change in composition can be seen even better<br />

when the data is studied using multivariate statistics.<br />

The concentrations <strong>of</strong> the 17 congenes that are used in<br />

the calculation <strong>of</strong> the TEQ value and the results in<br />

Fig. 2 were evaluated using PCA. At first a general<br />

PCA analysis <strong>of</strong> the data was made. This model gave little<br />

information as to the memory effect, since most <strong>of</strong> the<br />

variation in the data set was used to explain the changes<br />

in concentrations between the different sampling periods.<br />

If, however, the individual results for every congener<br />

was divided by the total amount <strong>of</strong> all dioxins and<br />

furans in that sample, the influence from variations in<br />

concentration could be eliminated.<br />

A new PCA model was calculated using the new variables<br />

and it was found that principal components 1 and<br />

3 <strong>of</strong> this model best described the variations in congener<br />

pattern that had been observed. The score and loading<br />

plots for these vectors are shown in Figs. 5 and 6<br />

respectively.<br />

7<br />

6<br />

5<br />

4<br />

3<br />

4<br />

3<br />

2<br />

t [3]<br />

1<br />

0<br />

-1<br />

-2<br />

-3<br />

2<br />

1<br />

10<br />

8<br />

50<br />

45 44<br />

49<br />

4743<br />

46 48 40<br />

3536 29 32 33 4239<br />

38<br />

31<br />

20 26 34<br />

24 25 30<br />

41 37<br />

55<br />

56<br />

53<br />

57 22<br />

9<br />

1119<br />

52<br />

21 181716<br />

5<br />

23 27 51 1314<br />

54<br />

12 15<br />

7<br />

28<br />

-4<br />

-5<br />

6<br />

-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8<br />

t [1]<br />

Fig. 5. Score plot vector 1 vs vector 3. Observations 1, 2, 3, 4, 6, 8, 10 are measurements before the <strong>wet</strong> scrubber. Maintenance stops<br />

are between obs. 36–37 and obs. 50–51. The ellipse indicates the Hotelling T2 95% confidence interval.


C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412 411<br />

0.40<br />

totdioxin<br />

TEQ<br />

TCDD<br />

0.30<br />

0.20<br />

2378TCDD<br />

HxCDD<br />

PnCDD<br />

0.20<br />

p [3]<br />

0.00<br />

-0.10<br />

-0.20<br />

HpCDD<br />

1234678hpCDD<br />

1234789hpCDF<br />

HpCDF 123789hxCDF<br />

OCDD OCDF<br />

1234678hpCDF<br />

123789hxCDD<br />

123478hxCDD<br />

123678hxCDD123478hxCDF<br />

23478pCDF<br />

HxCDF<br />

123678hxCDF<br />

12378pCDD<br />

2378TCDF<br />

234678hxCDF<br />

12378pCDF<br />

PnCDF<br />

TCDF<br />

-0.30<br />

TotFuran<br />

-0.30<br />

-0.20 -0.10<br />

0.00<br />

p [1]<br />

0.10<br />

0.20<br />

Fig. 6. Loading plot vector 1 vs vector 3. Low-chlorinated congeners are situated to the right and PCDDs are situated to the top,<br />

PCDFs to the bottom <strong>of</strong> the diagram.<br />

To the left in Fig. 5, with low values in score vector 1,<br />

all the measurements before the scrubber are located<br />

(obs. nos 1,2,3,4,6,8,10) and the measurements <strong>after</strong><br />

the scrubber are located more to the right. Two maintenance<br />

stops are located between obs. nos 36 and 37 and<br />

between nos 50 and 51. A rapid shift in score vector 1<br />

can be observed during the two maintenance stops. Further<br />

it is visible how the observations at the beginning <strong>of</strong><br />

the continuos sampling period (obs. nos 12–16) are all<br />

located close together. During a transition period <strong>of</strong><br />

four weeks the observations have moved to a new point<br />

in the diagram, indicating a new composition in the<br />

emission pattern. The emission pattern is then relatively<br />

stable until the maintenance stop between observations<br />

36 and 37 occurs.<br />

Observation 12, which is the first measurement with<br />

the continuos sampling device, and observation 37 almost<br />

have the same score in vector 1 but differ in vector<br />

3. From observation 37, which is the first observation<br />

<strong>after</strong> one <strong>of</strong> the maintenance stops, the process again<br />

moves from high values in vector 1 to lower values over<br />

a couple <strong>of</strong> weeks. The same shift can be seen even better<br />

between observations 50 and 51 which is the second<br />

maintenance stop during the period covered in this study.<br />

The loading plot (Fig. 6) shows a tendency that the<br />

degree <strong>of</strong> chlorination is described by loading vector 1<br />

and the ratio <strong>of</strong> PCDD to PCDF is described by loading<br />

vector 3. Low-chlorinated congeners have higher loadings<br />

in vector 1 and PCDFs are situated toward lower<br />

values in loading vector 3. Implementing this in the<br />

score diagram it seems that just <strong>after</strong> a maintenance stop<br />

there is a big loss <strong>of</strong> dioxins and furans with four chlorine<br />

atoms. The difference in score vector 3 for the observations<br />

12 and 37 would then indicate that the ratio<br />

PCDD/PCDF is higher <strong>after</strong> the maintenance stop than<br />

it was when the continuos samplings were started. The<br />

PCA model also shows that there is a shift between<br />

PCDDs and PCDFs over time since the loadings for<br />

total-PCDD and total-PCDF are situated on one <strong>of</strong><br />

the diagonals in Fig. 6.<br />

5. Conclusions<br />

• The high PCDD/F <strong>emissions</strong> found <strong>after</strong> the <strong>installation</strong><br />

<strong>of</strong> the <strong>wet</strong> scrubber was caused by memory<br />

effects due to high formation in the incinerator during<br />

start up <strong>of</strong> the plant.<br />

• The homologue pattern <strong>after</strong> the <strong>wet</strong> scrubber differs<br />

significantly from the pattern before the scrubber. A<br />

relation between the degree <strong>of</strong> chlorination, the<br />

vapour pressure and the emission level for a specific<br />

congener was found with the use <strong>of</strong> multivariate<br />

statistics.<br />

• Start-up periods, e.g. <strong>after</strong> maintenance stops, are the<br />

main cause <strong>of</strong> elevated dioxin <strong>emissions</strong> from an<br />

incinerator. This can only be shown by continuos<br />

monitoring where all modes <strong>of</strong> operation are covered.


412 C.-J. Löthgren, B. van Bavel / Chemosphere 61 (2005) 405–412<br />

Acknowledgments<br />

The authors would like to thank the staff at Sydkraft<br />

SAKAB for their assistance and acknowledge the analyses<br />

work <strong>of</strong> Eur<strong>of</strong>ins, Münster Roxel.<br />

This study was financially supported by the Swedish<br />

Knowledge foundation and SAKAB Kumla<br />

Miljöstiftelse.<br />

References<br />

Adams, B., Goemans, M., Joannes, J., 2001. Elevated <strong>Dioxin</strong><br />

Emission from a MSWI Due to the Memory Effect in the<br />

Wet Scrubbers. Seghers Bettertechnology for Solids + air.<br />

Willebroek, Belgium.<br />

Blumenstock, M., Zimmermann, R., Schramm, K.-W., Kettrup,<br />

A., 2000. Influence <strong>of</strong> combustion conditions on the<br />

PCDD/F-, PCB-, PCBz- and PAH-concentrations in the<br />

post-combustion chamber <strong>of</strong> a waste incineration pilot<br />

plant. Chemosphere 40, 987–993.<br />

Buekens, A., Cornelis, E., Huang, H., De<strong>wet</strong>tinck, T., 2000.<br />

Fingerprints <strong>of</strong> dioxin from thermal industrial processes.<br />

Chemosphere 40, 1021–1024.<br />

Chang, M.B., Lin, J.J., 2001. Memory effect on the dioxin<br />

<strong>emissions</strong> from municipal waste incinerator in Taiwan.<br />

Chemosphere 45, 1151–1157.<br />

Funcke, W., Linnemann, H., 1992. Sampling <strong>of</strong> polychlorinated<br />

dibenz<strong>of</strong>urans (PCDF) and dibenzo(p)dioxins (PCDD) in<br />

<strong>emissions</strong> from combustion facilities using an adsorption<br />

method. Chemosphere 24, 1563–1572.<br />

Gass, H.C., Lüder, K., Wilken, M., 2002. PCDD/F-<strong>emissions</strong><br />

during cold start-up and shut-down <strong>of</strong> a municipal waste<br />

incinerator. Organohalogen Compd. 56, 193–196.<br />

Gass, H.C., Neugebauer, F., 1999. Change <strong>of</strong> PCDD/PCDF<br />

pr<strong>of</strong>iles in the <strong>wet</strong> scrubber system <strong>of</strong> a fluidized bed<br />

incinerator used for municipal solid waste. Organohalogen<br />

Compd. 41, 153–156.<br />

Giugliano, M., Cernuschi, S., Grosso, M., Miglo, R., Aloigi, E.,<br />

2000. PCDD/F concentrations and mass balance in the flue<br />

gas cleaning section <strong>of</strong> a state-<strong>of</strong>-the-art MSW incineration<br />

plant. Organohalogen Compd. 46, 146–149.<br />

Giugliano, M., Cernuschi, S., Grosso, M., Miglio, R., Aloigi,<br />

E., 2002. PCDD/F mass balance in the flue gas cleaning<br />

units <strong>of</strong> a MSW incineration plant. Chemosphere 46, 1321–<br />

1328.<br />

Gullett, B.K., Touati, A., Lee, C.W., 2000. Formation <strong>of</strong><br />

chlorinated dioxins and furans in a hazardous-waste-firing<br />

industrial boiler. Environ. Sci. Technol. 34, 2069–2074.<br />

Hunsinger, H., Kreisz, S., Seifert, H., 1998. PCDD/F behavior<br />

in <strong>wet</strong> scrubbing systems <strong>of</strong> waste incineration plants.<br />

Chemosphere 37, 2293–2297.<br />

Korell, J., Seifert, H., Paur, H.-R., Andersson, S., Bolin, P.,<br />

2003. Flue gas cleaning with the Mercox process. Chem.<br />

Eng. Tech. 26, 737–740.<br />

Kreisz, S., Hunsinger, H., Vogg, H., 1996. Wet scrubbers—a<br />

potential PCDD/F source? Chemosphere 32, 73–78.<br />

Kreisz, S., Hunsinger, H., Vogg, H., 1997. Technical plastics as<br />

PCDD/F absorbers. Chemosphere 34, 1045–1052.<br />

Mayer, J., Linnemann, H., Becker, E., Rentschler, W., Jockel,<br />

W., Wilbring, P., Gerchel, B., 2000. Certification <strong>of</strong> a long<br />

term sampling system for PCDFs and PCDDs in the flue gas<br />

from industrial facilities. Chemosphere 40, 1025–1027.<br />

Olie, K., Vermeulen, P.L., Hutzinger, O., 1977. Chlorodibenzop-dioxins<br />

and chlorodibenz<strong>of</strong>urans are trace components <strong>of</strong><br />

fly ash <strong>of</strong> some municipal incinerators in the Netherlands.<br />

Chemosphere 6, 455–459.<br />

Onnen, J.H., 1972. Wet scrubbers tackle pollution. Environ.<br />

Sci. Technol. 6, 994–998.<br />

Pitea, D., Bonati, L., Collina, E., Cosentino, U., Lasagni, M.,<br />

Moro, G., Todeschini, R., 1994. Chemometric approaches<br />

in environmental-problems concerning PCDD and PCDFdata<br />

interpretation and source correlation—mechanisms <strong>of</strong><br />

formation and destruction in MSW combustion process.<br />

Fresenius J. Anal. Chem. 348, 111–120.<br />

Rordorf, B.F., 1989. Prediction <strong>of</strong> vapor pressures, boiling<br />

points and enthalpies <strong>of</strong> fusion for twenty-nine halogenated<br />

dibenzo-p-dioxins and fifty-five dibenz<strong>of</strong>urans by a vapor<br />

pressure correlation method. Chemosphere 18, 783–788.<br />

Van den Berg, M., Birnbaum, L., Bosveld, A.T.C., Brunstrom,<br />

B., Cook, P., Feeley, M., Giesy, J.P., Hanberg, A., Hasegawa,<br />

R., Kennedy, S.W., Kubiak, T., Larsen, J.C., van Leeuwen,<br />

F.X.R., Liem, A.K.D., Nolt, C., Peterson, R.E., Poellinger,<br />

L., Safe, S., Schrenk, D., Tillitt, D., Tysklind, M., Younes,<br />

M., Waern, F.T., Zacharewski, T., 1998. Toxic equivalency<br />

factors (TEFs) for PCBs, PCDDs, PCDFs for humans and<br />

wildlife. Environ. Health Perspect. 106, 775–792.<br />

Wang, L.-C., Lee, W.-J., Lee, W.-S., Guo-Ping, Chang-Chien,<br />

Tsai, P.-J., 2003. Effect <strong>of</strong> chlorine content in feeding wastes<br />

<strong>of</strong> incineration on the emission <strong>of</strong> polychlorinated dibenzop-dioxins/dibenz<strong>of</strong>urans.<br />

Sci. Total Environ. 302, 185–198.<br />

Wevers, M., De Fré, R., 1998. <strong>Dioxin</strong> emission reduction <strong>of</strong> a<br />

municipal waste incinerator by injection <strong>of</strong> activated<br />

carbon—the abatement <strong>of</strong> memory effects. Organohalogen<br />

Compd. 36, 343–346.<br />

Wold, S., Albano, C., Dunn III, W.J., Esbensen, K., Hellberg,<br />

S., Johansson, E., Sjöström, M., 1983. Pattern recognition:<br />

finding and using regularities in multivariate data. In:<br />

Hmartens, , Russwurm, H. (Eds.), Food Research and Data<br />

Analysis. Applied Science, London and New York, pp. 147–<br />

188.<br />

Wold, S., Albano, C., Dunn III, W.J., Esbensen, K., Hellberg,<br />

S., Johansson, E., Sjöström, M., Edlund, U., Geladi, P.,<br />

1984. Multivariate data analysis in chemistry. In: Kowalski,<br />

B. (Ed.), Chemometrics: Mathematics and Statistics in<br />

Chemistry. Riedel, Dordrecht, Netherlands, pp. 17–95.<br />

Zimmermann, R., Blumenstock, M., Heger, H.J., Schramm,<br />

K.-W., Kettrup, A., 2001. Emission <strong>of</strong> nonchlorinated and<br />

chlorinated aromatics in the flue gas <strong>of</strong> incineration plants<br />

during and <strong>after</strong> transient disturbances <strong>of</strong> combustion<br />

conditions: delayed emission effects. Environ. Sci. Technol.<br />

35, 1019–1030.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!