United States
                   Environmental Protection
                   Agency
Atmospheric Sciences Research
Laboratory
Research Triangle Park NC 27711
                   Research and Development
EPA/600/S3-84/094  Mar. 1986
&EPA         Project Summary
                   Gaseous  HCI and  Chlorinated
                   Organic Compound
                   Emissions  from  Refuse  Fired
                   Waste-to-Energy Systems
                    Arthur B. Nunn
                     Testing of the emissions of hydrogen
                   chloride (HCI), chlorinated dibenzo-p-
                   dioxins, dibenzofurans, chloroben-
                   zenes, chlorophenols, polychlorinated
                   biphenyls, and aldehydes was conduct-
                   ed on a water-wall, mass-fired munic-
                   ipal waste incinerator and a refuse-
                   derived  fuel (RDF)-fired  municipal
                   waste incinerator.  The purpose of the
                   testing was to evaluate the measure-
                   ment methods used for sampling and to
                   determine what differences may exist in
                   the emission characteristics of the two
                   types of units and  the effect that this
                   difference would have on the applica-
                   bility of the methods employed.
                     Testing of  HCI emissions was per-
                   formed by absorption in four midget
                   impingers each containing 20 ml of 0.1
                   N NaOH. Sampling of chlorinated or-
                   ganic compound emissions was con-
                   ducted using a modified EPA Method 5
                   sampling train with a cartridge of XAD-
                   2 resin between the third and fourth
                   impingers. Analysis was  performed
                   using high resolution gas chromatogra-
                   phy/mass spectroscopy.
                     HCI emissions from the  mass-fired
                   incinerator (Site A) exceeded the HCI
                   emissions from the RDF-fired inciner-
                   ator  (Site B) by approximately 45%.
                   Chlorinated organic compound emis-
                   sions from Site A exceeded those from
                   site B by factors ranging from 12 to
                   371.  Emissions from both sources
                   exhibited a large variation from test to
                   test;  however, isomeric distributions
                   were similar.
                     The majority of all organic compounds
                   were detected in the back half of the
Method 5 sampling train with the larg-
est overall  percentage  found in the
impinger contents. Chlorobenzenes
were the only compounds of which the
majority was detected in the XAD-2
resin trap.
  This Project Summary was developed
by EPA's Atmospheric  Sciences Re-
search Laboratory, Research Triangle
Park, NC. to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).

Introduction
  Subsequent  to reports in the late
1 970's that various incineration process-
es give  rise to emissions of chlorinated
dibenzo-p-dioxms(CDDs) and chlorinated
dibenzofurans (CDFs), the United States
Environmental  Protection Agency con-
ducted  preliminary investigations of a
resource recovery incinerator, which
further  indicated the presence of these
compounds. The results of this program,
which included data of a qualitative rather
than  quantitative  nature, indicated the
need for further  investigation of the
sampling methods used and, as a result,
the present program was embarked upon
in early  1 981
  The purpose  of this program was to
evaluate measurement methods used for
sampling the emissions of CDDs, CDFs,
Chlorobenzenes, chlorophenols, polychlor-
inated biphenyls,  and aldehydes from
water-wall, mass-fired and from refuse-
derived-fuel (RDF)-fired municipal waste
incinerators These compounds were of

-------
interest not only because of the potential
lexicological properties associated with
each, but also because several of these
groups of compounds, particularly chloro-
benzenes and chlorophenols, have been
implicated as precursors to the formation
of CDDs, CDFs, and PCBs in the combus-
tion process. The purpose  of testing  the
two different types of incinerators was to
determine what differences may exist in
the emission  characteristics  of these
types of units and hence the effect these
differences would have on the applicabil-
ity of the method Water-wall, mass-fired,
and RDF-fired incinerators were chosen
for the program because they are the two
most prevalent types of waste-to-energy
systems currently in operation or likely to
become operational m the near  future.
One unit of each type was tested during
the program
  In addition to the chlorinated organic
compounds, sampling and analysis was
conducted for HCI emissionsfrom the two
units. This was conducted to evaluate the
distribution  in the  gas stream  of  the
chlorides released from the refuse during
the combustion process and the applica-
bility of the manual method.

Site Descriptions
  This program consisted of tests con-
ducted at two separate facilities. The first
facility, Site "A", is a mass-fired, water-
wall,  refuse-burning  incinerator-boiler
that is used to generate steam for various
purposes at a nearby research  center
The  unit burns  100% refuse with  no
auxiliary fuel  at  a  rate  approximately
90,000 to 135,000  kg/day (100 to 150
tons/day). The unit is operated at or near
capacity at all times except when firing
excessively wet trash.
  The second facility that was tested, Site
"B", is a Detroit Rotograte Stoker Boiler
that produces steam for  plant heating
The  unit is designed to  burn coal, on
which it is usually fired  The  facility
operators are, however, investigating the
possibility of switching to refuse-derived
fuel as a fuel, and the boiler was fired on
100% RDF during testing. During testing,
Site "B" was fired at a rate of approxi-
mately 205Kg (225 tons) of RDF per day.

Experimental Method

Sampling Methods
  Sampling for HCI concentrations in the
exhaust stream was conducted by pulling
a sample through a stainless steel probe,
through four midget  impingers each
containing 20 ml of 0.1 N NaOH, through
a diaphragm pump, and finally through a
flow meter, flow control valve, and a dry
gas meter
  The sample was  drawn at a rate of
approximately 2 L/mm and sampling was
conducted for 30 mm, thus, a sample
volume  of  approximately 006m3  was
collected The impingers  containing the
absorbing solution were submerged in an
ice bath during sampling. Following sam-
pling, the impinger contents were placed
in labeled sample jars where they were
maintained until analyzed on  site
  Organic compound sampling was  con-
ducted using a  modified  EPA Method 5
sampling system  This system consisted
of a heated glass-lined sample probe with
a stainless steel nozzle, a glassfiber filter
m a heated  glass filter  holder with a
stainless steel filter support; a Greenburg-
Smith impinger containing 100 ml dis-
tilled water followed by a modified
Greenburg-Smith impinger (dry for  mist
knockout), both  containing 5 cm of 3 to 6
mm diameter glass beads and immersed
in an ice bath, a resin cartridge containing
60 g of XAD-2 sorbent resin, and finally
an impinger containing 200 to 300 g of
silica gel  Sample flow was measured and
controlled by a conventional EPA Method
5 meter box  All glassware used in the
tests was annealed prior to being taken
into the field
  Sampling was conducted isokmetically
during all tests Tests were conducted for
180 and 300  mm  time periods  with
sample volumes ranging  from 2 5 to 6 8
m3.
Analytical Procedures
  HCI analysis  was  performed by  the
mercuric  nitrate (HgfNOs^)  method,
which involves titrating Cl~ with  a stand-
ardized  Hg(NOa)2 solution using a di-
phenylcarbazone indicator-acidifier  rea-
gent This method is known to be subject
to interference and endpomt  masking by
the presence of sulf ite (SOs2") and cationic
metals in the exhaust stream In order to
alleviate this problem,  samples were
treated with  3% hydrogen penoxide to
eliminate the sulfite interference,  and
they were passed through a Rexyn 1 01 -H
column (a registered trademark of Fisher
Scientific) to eliminate the cationic  end-
point masking.
  The procedures used to separate the
halocarbonsof interest from the matrices
in which they  were collected  involved
solvent  extraction with  the  addition of
internal  standards  of [37CU]-2,3,7,8-
TCDD,  [37CI7]-1,2,3,4,6,7,8-HPCDD,
[37CI8]-OCDD,[D6]-3,3',4,4'-tetrachlorobi-
phenyl, [13C6]-pentachlorophenyl, and
[13C6]-hexachlorobenzeneto each sample
  Further sample preparation procedures
included acid, water, basic, and further
solvent  extractions, macro-column elu-
tion, and mini-column liquid chromatog-
raphy  If preliminary GC-MS screening
analysis of a sample indicated the pres-
ence of  potential interfering compounds
or other sample matrix constituents that
eluted from the GC at  very  long times,
then additional sample clean up or frac-
tionation was required using high per-
formance liquid chromatography (HPLC)
The HPLC used for this purpose was a
Varian Model 5021 Microprocessor Con-
trolled HPLC equipped with a CDS-1 1 1 L
Data System  Dual Dupont Zorbax ODS
Columns (25  cm x 0.6 cm I D.) were
employed  Methanol was  employed  as
the mobile phase in the isocratic mode
  The instrumentation utilizedto perform
the quantitative analyses for the chloro-
carbons of interest was comprised of a
gas chromatograph-mass spectrometer-
data system (GC-MS-DS) that included a
Perkm-Elmer Sigma III  Gas Chromato-
graph coupled through a specially modi-
fied interface to a Kratos  System This
analytical system falls m  a  class  of
analysis referred to as  "high resolution
gas  chromatography/high  resolution
mass spectroscopy" (HRGC/HRMS) The
gas chromatograph was equipped with a
split/splitless injector and  a wall-coated
open-tubular  fused-silica  capillary col-
umn  The  interface mentioned  above
included provision for both direct admis-
sion of the column effluent into the mass
spectrometer source as well as admission
of the column effluent via a single-stage,
all-glass jet separator The  interface was
modified to minimize peak  broadening
due to  excessive dead  volume and to
optimize the temperature throughout the
interface  The mass spectrometer was
operated m the multiple ion monitoring
mode under the control of the computer-
based  Kratos DS55SM  data system.
Elaborate programs for the  DS55SM data
system  were developed that permitted
rapid, automated retunmg of the mass
spectrometer during analysis of a sample
extract and thus  20 to 30 separate ion
masses  were sequentially monitored
during an analysis Each of the groups of
chlorocarbons of  interest  (CDDs, CDFs,
chlorobenzenes,  chlorophenols,  and
PCBs) were analyzed separately

Results
  The  testing conducted on the twc
systems revealed considerable differenc-
es in the emission characteristics of the
mass-fired, water-wall  incinerator anc

-------
the RDF-fired incinerator with the emis-
sions from the mass-fired unit higher in
all cases A summary of the emissions
from each  unit  tested is  presented in
Table 1
  HCI  emissions from  the  mass-fired
incinerator (Site A)  exceeded the  HCI
emissions from the RDF-fired incinerator
(Site B) by approximately 45%  Chlori-
nated organic compound emissions from
Site A  exceeded  those from  Site B,
however, by factors ranging from 1 2 to
371  The overall distribution of chlorides
in the gas streams, which is presented in
Table 2, shows that a higher percentage
of total chlorides was emitted as HCI from
Site B  than from Site A thus indicating
that combustion was  more efficient for
the RDF-fired incinerator. This difference
may be due to differing moisture contents
of the  two  types of fuel. The RDF was
stored in silos, and was never exposed to
moisture thus  it  was very  dry  when
combusted  The  fuel for Site A,  on  the
other hand, was often quite wet as it was
stored in open areas where it was exposed
to rainfall prior to delivery to theplant site
where  it was stored m a covered pit.

Conclusions
  1.  The sampling and analytical meth-
     odologies used for this project per-
     formed well
  2.  Chlorinated  organic  compound
     emissions  were higher from  the
     mass-fired  incinerator than from
     the RDF-fired incinerator.
  3.  The majority of the CDDs emitted
     from both units were heptachlori-
     nated dibenzo-p-dioxins
  4  Chlorophenols were detected in the
     highest  concentrations of all types
     of chlorinated organic compounds
     evaluated.
Table 1.   Summary of Emissions Data
Compound
HCI
TeCDD
PsCDD
H£DD
HfCDD
OCDD
JeCDF
P£DF
H£DF
HPCDF
OCDF
DCLB
T,CIB
TeCIB
P5C/B
H£IB
T,CP
TeCP
PsCP
T,CB
TeCB
PSCB
HXCB
Aldehydes
Site A
Emission Rate*
(mg/h!
6?**
153
19 7
31 5
354
92
765
350
41 2
365
26
563
2280
3550
551 0
1650
2,444 0
1.127 0
5880
146
84
1 7
09
1660
Site B
Emission Rate*
(mg/h)
76**
0 1
0 1
02
1 5
08
1 3
05
09
32
04
03
58
80
399
21 6
1490
2150
3520

--
--

94
   Average of all tests at each site
   *HCI in kg/h
Table 2.
           Chloride Distribution in Gas Stream
                                               Percent of Total
                                          Chlorides in Gas Stream (%)
Compounds
HCI
Chlorophenols
Chlorobenzenes
Chlorinated Dibenzo furans
Chlorinated Dibenzo-p-Dioxins
Polychlonnated Biphenyls

Site A
99 92237
0 05233
002136
0 00235
000140
000019
WOO
Site B
99 992178
0 007650
0 000079
0 000064
0 000029
0 000000
WOO
                                           Arthur B. Nunn III is with Scott Environmental Technology, Inc , Plumsteadville,
                                             PA 18949.
                                           James L. Cheney is the EPA Project Officer (see below/.
                                           The complete report, entitled "Gaseous HCI and Chlorinated Organic Compound
                                             Emissions from Refuse Fired Waste-to-Energy Systems," (Order No PB 86-145
                                             661 /AS, Cost- $11 95, subject to change/ will be available only from:
                                                   National Technical Information Service
                                                   5285 Port Royal Road
                                                   Springfield, VA 22161
                                                   Telephone: 703-487-4650
                                           The EPA Project Officer can be contacted at:
                                                   Atmospheric Sciences Research Laboratory
                                                   U.S. Environmental Protection Agency
                                                   Research Triangle Park, NC 2777 1

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/600/S3-84/094
          OC00329   PS

          U  S ENVIR PROTECTION  AGENCY
          *?GION 5  LIBRARY
          230 S  DEARBORN  STREET
          CHICAGO              IL   60604

-------