Sampling of polychlorinated dibenzofurans (PCDF) and dibenzo(p)dioxins (PCDD) in emissions from combustion facilities using an adsorption method

Sampling of polychlorinated dibenzofurans (PCDF) and dibenzo(p)dioxins (PCDD) in emissions from combustion facilities using an adsorption method

Chemosphere, Voi.24, No.ll, pp 1563-D72, 1992 Printed in Great Britain 0045-6535/92 $5.00 + 0.00 Pergamon Press Ltd. SAMPLING OF POLYCItLORINATED DI...

402KB Sizes 0 Downloads 29 Views

Chemosphere, Voi.24, No.ll, pp 1563-D72, 1992 Printed in Great Britain

0045-6535/92 $5.00 + 0.00 Pergamon Press Ltd.

SAMPLING OF POLYCItLORINATED DIBENZOFURANS (PCDI0 AND DIBENZO(P)DIOXINS (PCDD) IN EMISSIONS FROM COMBUSTION FACILITIES USING AN ADSORI~ION M E T I t O D W e r n e r F u n c k e and H e i n z L i n n e m a n n

GfA - Gesellschaftffir Arbeitsplatz- und UmweltanalytikmbH, Otto-Hahn-StraBe 22, D - 4400 Mfinster-Roxel,Fed. Rep. of Germany Keywords: PCDD, PCDF, sampling, adsorption method

Abstract A sampling method for PCDF/D in emissions from combustion and other industrial facilities is presented in which these substances are deposited in one collection unit consisting of a combination of quartz wool and a polymeric adsorbent (pretreated XAD-2 resin). In combination with a cooled probe, this sampling method can be used for flue gas at temperatures up to more than 1,000 °C, with dust contents up to more than I g/m 3 and water contents of more than

300

g/m 3.

Performance

characteristics

referring

relative

standard

deviations

from

parallel determinations and comparison with other sampling methods are presented for PCDF/D flue gas concentrations between 2 and 0.02 ng

(TE)/m3 (Toxicity equivalent according to

NATO/CCMS).

I

Introduction

In t h e

Federal

ling

methods

loped the

and

for

in

these

the

described given The

sampling

minate

or

mentioned this

in

in

method

time

facilities

series 1 -

at

part

for PCDF/D

2,

it w a s at

from

4 of

this

guideline

concentrations

here

flue

gas

>

1 ng

was

in

than

planned

1988 250

1563

samp-

were

preliminary

deve-

drafts

characteristics

show

good

( T E ) / m 3.

for pre-

comparability For

the

characteristics

and

in

1987

of

the

method

are

with

of

i00

the

aim

VDI-guideline

optimized

samplings

concentrations

[5,6].

different

also

( T E ) / m 3.

disadvantages

used

more

series

of

performance

of < 0.i ng

the

first in

The

as

four

facilities

performance

presented

tested

combustion

[I - 4].

corresponding

It was

last years

3499

minimize

PCDF/D

gas

the

exist

concentrations

least

above.

flue

over

descriptions

PCDF/D

device

at

Germany

The

parts

methods

of

PCDF/D

tested.

VDI-guideline

sented of

Republic

at to

until

1991.

different <

0.005

to

eli-

methods During

combustion ng

(TE)/m 3

1564

2

Expct'imenlal

2.1

Sampling Equipment

The sampling

equipment

which

according to VDI-guideline

allows

isokinetic

collection

of flue gas dusts

2066, part 1 [7], is shown in figure I.

The equipment contains the following parts: a) Probe: tures PCDF/D ted

nozzle,

> 80

elbow

°C double

collecting

XAD-2,

flexible

phase:

D0-frit;

spherical

with glass

support

tube

to

joints;

via

gas

glass

cooler

tubes

for

wool, with

extensive

b)

pretreagastight,

removal

of

d) gas transporting system with volume meter.

Sampling

The PCDF/D-sampling meter

and

concerning

determination

of

isokinetic

flue

gas

suction,

tube

spiked

is cleaned

with

or

13C-labelied

> 80 °C to over 1,000

renewed,

°C a water-chilled

at flue

gas temperatures

used.

In case

of

ries.

Dependent

tection

flue

gas

on flue

of < 80

dust

is recommended

is

drying tower,

realized

pretreated

For

flue

gas

XAD-2

suction tube is used °C an uncooled

concentrations

limit for PCDF/D-congeners

PCDF/D

to

of

is used,

temperatures

> 1 g/m 3 a

concentration

of

(according to

suction tube

can be

larger quartz

or a glass cyclone has to be connected

gas velocity,

gas-

according

(ca. 4 mm ID) inside the suc-

especially

PCDF/D-standard(s).

[3]);

wool cartridge

and

pump,

parameters

[2,7]. Prior to each sampling the glass insert tion

at flue gas tempera-

cooling of the probe;

stuffed with quartz

probe c)

inserts;

for water

glass-cartridge

connection ground

flue gas humidity;

2.2

(titanium)

cased

in se-

and desired

de-

a suction volume flow of 0.5 to 2 m3/h is

adjusted.

2.3

Work-up and Analysis

Usually only the glass cartridge, from the sampling. ded

in

the

The glass

work-up

if

dust

usually avoided or minimized the

condensate

washed

which

results

into the glass

further clean-up

results

insert inside the suction tube is rarely includepositions

are

visible.

Such

depositions

by using a glass tube with suitable from

cartridge.

and analysis

filled with quartz wool and XAD-2,

cooling

the

The extraction

is realized

tube,

by which

of the matrices

analogous

are

ID, and by the

dust

is

as well

as

to the descriptions

gi-

ven in [2]. At a sampling volume of 10 m 3 and application of a HRGC/HRMS-system

(e.g. VG

AutoSpec)

-

for

e.g.

ng/m 3 is reached.

2,3,7,8-TetraCDD,

This way TE-values

a

detection

limit

of

0.0005

0.001

of 0.01 ng/m 3 can be determined without

1565

r,

v

r"

o

'o

r" -,4 .,-I

~

. ~ 1

.-4

@

¢" -I~ '~J ~:1~>~

o o o . ~ , ~ , ~ 4.~ . ~ -,..~

~., ~:~ ,~ q.~ ~t

o /t

t.-,.

1566

being

influenced

blank values

decisively

by

not

can be kept below

detected

0.001

ng

PCDF/D-congeners.

The

PCDF/D

(TE)/m 3 (referred to i0 m 3 sampling

volume). If 13C-I,2,3,4-TetraCDD

is used as standard for spiking the XAD-2, mean reco-

very rates of usually 90 % are observed.

This is higher than for other samp-

ling methods used so far [i - 3].

Results and Discussion

3.1

PCDF/D Dispersion

To verify

the PCDF/D

tial PCDF/D cartridge);

inside c)

back-up

each

in the sampling

from XAD-2

the

probe;

condensate

behind

XAD-2

cartridge

collection

centration

dispersion

break-through

glass-insert d)

in the Sampling Device

phase

level

shown

of ca.

b)

in

5 ng

figure

unit

unit

condenser 2

(quartz

(cooled

unit).

are

and to check poten-

phases were analysed:

adsorption

adsorption

(behind

device

the following

flue gas;

The PCDF/D

referred

(TE)/m 3, a suction

to

a)

wool/XAD-2 20

°C);

portions

a

PCDF/D

rate of 2 m3/h,

in

con-

a sampling

volume of ca. I0 m 3 and a XAD-2 temperature of 20 °C. The PCDF/D-portion deposited tal

PCDF/D.

adsorption

The

performance

case of OctaCDF. backup

XAD-2

in the filter/adsorption

differentiation with

homologeous

increasing

The additional

cartridge)

of

chlorination

collection phases

contain

only

very

unit is > 99 % for to-

groups

low

shows

order

down

(probe tube, and

a

decay

to

95

in

%

condensate

therefore

in +

irrelevant

PCDF/D-portions which slightly increase with degree of chlorination.

3.2

Standard Deviations

from Parallel Determinations

[5, 8] The relative lysis) in

case

PCDF/D

standard

from parallel of

PCDF/D

groups

for the complete method

determinations concentrations

concentrations

homologeous

deviations

around

and single

0.02

shown of

2

ng

congeners

in table

ng

i and 2, are about

(TE)/m 3 and

(TE)/m 3.

(sampling and ana-

The

slightly

standard

are higher;

this

i0 %

higher

deviations

at for

is especially evi-

dent for the lower PCDF/D concentration range. It

has

errors sulting

to

be

noted

from sampling, from

be excluded.

that

standard

clean-up

local or temporary

deviations

and analysis,

do

not

only

include

but also systematic

non-homogeneities

random

errors re-

in the stack which cannot

1567

PCDF/D-Percentage (Differentiated by Homologeous Groups) in Probe Tube, Adsorption Unit and Condensate + Backup XAD-2 Cartridge. The Total of the PCDF/D Concentrations in all Collection Phases is Equivalent to i00 %; Mean Values and Ranges out of 6 Samplings in the PCDF/D Concentration Range of 5 ng TE)/m 3

Figure 2:

54,54 3.5

PCDF/D- P ~ T I ~ (~) IN GLASS- TUBEINSIOE~OBE

3

2.52 1,5

i

10.5-

~

0

'.

"

'

"

PCDF/D- PORTION(%)

INADSORPTIONUNIT

I00 l 99~

9a! 9796. 95.

94. 93. 92

91 90,

PCOF/D- PORTION(X) IN CONDENSATE(AFTERADSORBENT) AND IN BACK-UPXAD-2 CARTRIDGE

109. 8

7 6 5 4

O ' ~ '





i

-- ,--L,. . . . . . (T4COO"PSCDD H6CDOH/CO00CDO

i

T4,COFPSCOFHBCOF"H7COFOCDF

The

standard

levels are [i

-

deviations,

documented

in the same range

for

corresponding

as specified

PCDF/D

4-8 COF/O

concentration

for other PCDF/D sampling

methods

3].

C o m p a r i s o n M e a s u r e m e n t s with a Known S a m p l i n g M e t h o d

3.3

The

sampling

[2]

at

method

relatively

presented high

The best comparability

and

here was compared

relatively

is shown

low PCDF/D

especially

of ca. 1 ng (TE)/m 3 (see figure 3).

to the method flue

for higher

gas

PCDF/D

described

in

concentrations. concentrations

1568

In case 4),

of

lower

partly

HeptaCDF

PCDF/D

higher

concentrations

deviations

and TetraCDD

are

are

of

ca.

observed.

not real

0.02

The

because

ng

(TE)/m 3

highest

the totals

the

are considerably

in-

As mentioned earlier,

geneities

have

the

estimation

thods.

flue

of

gas

the

The relative

channel

concentration

PCDF/D

to

be taken

differences

concentration

4

PCDF/D concentration

observed

differences

levels

possible non-homo-

into consideration for

shown

4 are in the range of the relative standard deviations at corresponding

figure for

fluenced by not detected congeners. in the

(see

deviations

the

for

two

in figures

me-

3 and

for the two methods,

(see tables i, 2 and [2, 5]).

Conclusion

The results

presented

here show that the new sampling method

for the determination This

includes

of PCDF/D

PCDF/D

in flue gas of e.g.

concentration

levels

of

0.02

the verification

of the threshold value of 0.i ng

the 17. BImSchV

[9].

the

sampling

equipment

equipment

can

be

almost

network

in dust sampling techniques

measurements

are

following

measurement

equipment

after

solvents)

is

eliminated

unnecessary;

in

the

contact

glass cartridge; case

of

the

adsorbent, dust with

g)

with

and

the

e)

in

[i - 4],

advantages:

a) the

(up to

1,000

(over i g/m 3) of the flue at the sampling

location;

c) the requirements

measurements;

minutes;

sampling

possible

value

of

gas

extensive

location

secondary the

are

methods

solution), resin;

cartridge

13C-PCDF/D-standard)

described

temperature

d)

limited

(probe

only in

with

can

one

case

is

to

the

sent

probe

filter off

by

of

for a

of

the

inflammable is

largely

minimized glass

because

tube

and

the

collection matrices

dust

collection

dust

be

tube,

the

with

contamination

equipment

for

exchange

cleaning

(e.g.

in contrast to up to 5 different

XAD-2

glass

only

the

flue

conventional

washing

filter

here;

f)

blank

dust

assures

(including dust filter and XAD-2 resin)

requires at

to

(TE)/m 3 which

following the

is sufficient;

comparable

sampling

and

the of

of the sampling equipment

glass tube and glass cartridge

surfaces

here has

independent

(over 300 g/m 3) and dust content

simple handling

knowledge

ng

facilities.

(TE)/m 3 in accordance with

to at least one method

presented

used

°c), water content gas; b)

Compared

is appropriate

combustion

filter,

matrix

tube)

(cartridge

has

to

and

adsorbent

the

analytical

be

with

analysed

(e.g.

spiked

laboratory;

spiking of the equipment at the sampling location is no longer necessary; costs for

for

the

sampling

costs

for

sampling

equipment

preparations

analysis

are

are

reduced

are

reduced as

relatively and

well,

collection phases have to be extracted.

due

low; to

because

i)

that only

time or

h)

requirements

sampling one

in

condensate,

in

costs; case

j) two

1569

Table

i:

Mean Values (~), A b s o l u t e and Relative the Complete Method from 8 Parallel Work-up and Analysis) at Relatively High

Parameter

I

s

i

Standard Deviation (s) of Determinations (Sampling, PCDF/D Concentrations

Srel

m

PCDF/PCDD Total TetraCDF Total PentacDF Total HexaCDF Total HeptaCDF OctaCDF Total Tetra- to OctaCDF 2378-TetraCDF 12378-/12348-PentacDF a 23478-PentaCDF 123478-/123479-HexaCDF a 123678-HexaCDF 123789-HexaCDF 234678-HexaCDF 1234678-HeptaCDF 1234789-HeptaCDF Total TetraCDD Total PentaCDD Total HexaCDD Total HeptaCDD OctaCDD Total Tetra- to OctaCDD 2378-TetraCDD 12378-PentaCDD 123478-HexaCDD 123678-HexaCDD 123789-HexaCDD 1234678-HeptaCDD Total Tetra- to OctaCDF/D TE (NATO/CCMS 1988)

ng/m 3

ng/m 3

40,7 27,0 12,1 4,0 0,6 84,4

3,6 3,3 1,1 0,5 0,3 7,1

9 12 9 13 50 8

8 8 8 8 5 8

0,17 0,30 0,19 0,17 0,18 0,04 0,06 0,38 0,06

ii ii 14 10 i0 25 8 13 29

8 8 8 8 8 8 8 8 8

0,4 0,6 0,5 0,3 0,4 1,6

ii 15 16 13 13 i0

8 8 8 8 8 8

0,03 0,06 0,04 0,03 0,03 0,12

ii I0 19 11 8 10

8 8 8 8 8 8

8

8

i0

8

1,54 2,63 1,32 1,71 1,74 0,16 0,71 3,04 0,21 3,6 3,9 3,1 2,4 3,0 16,1 0,28 0,62 0,21 0,28 0,40 1,26 100,4 2,11

8,5 0,22

%

For totals and toxicity equivalents (TE} not detected congeners or homologeous groups are not included a gaschromatographically not separable on SP 2331

1570

Table

2:

Mean Values (~), A b s o l u t e and Relative the Complete Method from 5 Parallel Work-up and Analysis) at Relatively Low

Standard Deviation (s) o f Determinations (Sampling, PCDF/D Concentrations

Parameter

~

s

PCDF/PCDD

ng/m 3

ng/m 3

%

Total TetraCDF Total PentacDF Total HexaCDF Total HeptaCDF OctaCDF Total Tetra- to OctaCDF

0,140 0,189 0,142 0,070 0,034 0,575

0,095 0,060 0,054 0,029 0,024 0,106

68 32 38 41 71 18

5 5 5 5 3 5

2378-TetraCDF 12378-/12348-PentaCDF a 23478-PentaCDF 123478-/123479-HexaCDF a 123678-HexaCDF 123789-HexaCDF 234678-HexaCDF 1234678-HeptaCDF 1234789-HeptaCDF

0,0122 0,0233 0,0119 0,0351 0,0162 0,0026 0,0069 0,0487 0,0051

0,0082 0,0099 0,0007 0,0131 0,0061 0,0006 0,0038 0,0194 0,0028

67 42 6 37 38 23 55 40 55

5 5 5 5 5 5 5 5 4

Total TetraCDD Total PentaCDD Total HexaCDD Total HeptaCDD OctaCDD Total Tetra- to OctaCDD

0,022 0,032 0,030 0,010 0,019 0,113

0,013 0,006 0,016 0,005 0,004 0,022

59 19 53 50 21 19

5 5 5 5 5 5

2378-TetraCDD 12378-PentaCDD 123478-HexaCDD 123678-HexaCDD 123789-HexaCDD 1234678-HeptaCDD

0,0025 0,0050 0,0020 0,0029 0,0021 0,0050

0,0015 0,0013 0,0012 0,0013 0,0017 0,0025

60 26 60 45 81 50

5 5 4 5 5 5

Total Tetra- to OctaCDF/D

0,688

0,096

14

5

TE (NATO/CCMS

0,0207

0,0022

11

5

1988}

Sre 1

For totals and toxicity equivalents (TE) not detected congeners or homologeous groups are not included a gaschromatographlcally not separable on SP 2331

1571

Figure 3:

Percentage Differences in P C D F / D C o n c e n t r a t i o n s determined w i t h the New M e t h o d (Set to i00 %) and the M e t h o d d e s c r i b e d in [2] at P C D F / D Flue Gas C o n t e n t s of ca. 1 ng (TE)/m 3. Mean V a l u e s and R a n g e s from 5 P a r a l l e l D e t e r m i n a t i o n s 50'

3020 10 0

II

• .,ll

'1 I1'11

,,.,.,ll,il=

I r

-10-20-30 -40 -50 T4COF" P5CDF H6CDF H7CDF OCDF

T4CDD

P5CDD

H6CDD

H7CDO

OCDD

4-8

CDF/D

Figure 4:

Percentage Differences in P C D F / D C o n c e n t r a t i o n s D e t e r m i n e d w i t h the N e w M e t h o d (Set to I00 %) and the M e t h o d D e s c r i b e d in [2] at P C D F / D Flue Gas C o n t e n t s of ca. 0.02 ng (TE)/m 3. M e a n V a l u e s and R a n g e s from 5 P a r a l l e l D e t e r m i n a t i o n s 100 80 604020 0

,,.= |

i

i

,

'i

-20 -40 -60-80-

-100 T4COF PSCDF H6COF H7CDF OCDF

T4CDD P5CDO HGCOD H7COD OCOD

4-8

cOr/o

1572

Finally

it has to be pointed out that not only PCDF/D but further,

relati-

vely unpolar substances of comparable volatility can be detected with this equipment as well: polybrominated dibenzofurans (PBDF), polybrominated dibenzo(p)dioxins (PBDD), polybrominated/-chlorinated dibenzofurans (PBCDF), polybrominated/-chlorinated dibenzo(p)-dioxins (PBCDD), trito decachlorinated

biphenyls

(Tri- to DecaCB),

(Penta- and HexaCBz),

penta- and hexachlorinated

tri- to pentachlorinated

polychlorinated naphthalines

phenols

benzenes

(Tri- to PentaCPh),

(PCN), polycyclic aromatic hydrocarbons

(PAH).

Acknowledgements We would like to thank the Ministerium fur Umwelt, Raumordnung und Landwirtschaft of North Rhine Westphalia for financial support and Dr. Br~ker of the Landesanstalt fur Immissionsschutz,

Essen,

for professional consultation du-

ring execution and development of this project.

L~erature [i]

VDI-guideline 3499, part 1 (draft), DUsseldorf,

[2]

VDI-guideline 3499, part 2 (preliminary draft), DUsseldorf,

1990

[3]

VDI-guideline 3499, part 3 (preliminary draft), DUsseldorf,

1990

[4]

VDI-guideline 3499, part 4 (preliminary draft), DUsseldorf,

1988

[5]

GfA-report schaft:

to

the

Ministerium

"Entwicklung

polychlorierten industrieller

fur

Umwelt,

1990

Raumordnung

eines vereinfachten Verfahrens

Dibenzofuranen

Feuerungsanlagen".

und

Dibenzo(p)dioxinen

This procedure

und

Landwirt-

zur Probenahme von

contains

in

Emissionen

a pending pa-

tent. [6]

W. Funcke and H. Linnemann: unpublished results.

[7]

VDI-guideline 2066, part i, DUsseldorf,

1975

[8]

VDI-guideline 2449, part i, DUsseldorf,

1970

[9]

"17.

Verordnung

zur

Durchfflhrung

(Verordnung Uber Verbrennungsanlagen Stoffe - 17. BImSchV)",

des

Bundes-Immissionsschutzgesetzes

fur Abf~lle und ~hnliche brennbare

Bundesgesetzblatt,

2545 (1990)

(Received in Germany 2 March 1992; accepted i0 April 1992)