ESTR-80-04
APRIL 1980
t
RESULTS OF SOURCE EMISSIONS CHARACTERIZATION
AT THE HEMPSTEAD, NY REFUSE ENERGY RECOVERY SYSTEM
Submitted to:'
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
U& ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE RWK, NC 27711
UNDER CONTRACT 68-02-2566
Submitted by
NORTHROP SERVICES. INC.
ENVIRONMENTAL SCIENCES
P.O. BOX 12313
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27709
-------
ES-TR-80-04
APRIL 1980
RESULTS OF SOURCE EMISSIONS CHARACTERIZATION
AT THE HEMPSTEAD, NY REFUSE ENERGY RECOVERY SYSTEM
Prepared by:
Barry Dellinger
Chris Fortune
Jeff Lorrain
Environmental Chemistry and Emissions Sciences
Northrop Services, Inc.
Environmental Sciences
Submitted to:
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
UNDER CONTRACT 68-02-2566
Reviewed and Approved by:
w 0 c
4\m A. Stikeleather, Manager
Environmental Chemistry and
Emissions Sciences
Gary F.ฃerio, Manager
Environmental Research
NORTHROP SERVICES, INC.
ENVIRONMENTAL SCIENCES
P.O. BOX 12313
RESEARCH TRIANGLE PARK. NC 27709
-------
Environmental Sciences Center
ES-TR-80-04
DISCLAIMER
This report has been reviewed by Northrop Services, Inc.-Environmental
Sciences, and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
-------
'M Environmental Sciences Center ES-TR-80-04
FOREWORD
This report presents the results of work performed by Northrop Services,
Inc.-Environmental Sciences, under Contract Number 68-02-2566 for the Stationary
Sources Emissions Research Branch, Environmental Sciences Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina.
The source work was conducted in response to Technical Directive 3.3-52, while
the laboratory work was conducted under Technical Directive 3.3-1.
iii
-------
Environmental Sciences Center
ES-TR-8Q-Q4
ABSTRACT
In response to a request from Region II of the U.S. Environmental Protec-
tion Agency, Northrop Services, Inc.-Environmental Sciences was contracted by
the Agency's Environmental Sciences Research Laboratory to conduct source sam-
pling at the Henpstead Resources Recovery Corporation Facility in Hempstead,
New York. Source samples were collected from July 24 to July 26, 1979, and
extensive laboratory studies were undertaken to validate the results of sample
analysis.
Although sampling was hampered by considerable down time at the facility,
the measurements for chloride and sulfur oxide emissions indicate low concen-
tration levels for these species. Sampling for organic species yielded an
average total organic emission rate in excess of 25 lb/h. The majority of
organic emissions consisted of commonly-occurring, innocuous compounds, but
several materials posing potential hazards were also detected: substituted
phthalate isomers, chlorinated phenols, chlorinated biphenyls and related
compounds. The results of additional sample analyses currently in progress
will be reported at a later date.
Analysis by ion chromotography of sample fractions from the sulfur oxide
sailing train detected the presence of formate and acetate. Laboratory
studies indicated that these species were not the result of artifact forma-
tion, but rather could be formed by the oxidation of formaldehyde and acetalde-
hyde via the hydrogen peroxide impinger solutions. However, the possibility
that the decomposition or oxidation of other species is responsible for the
formate and acetate cannot be ruled out entirely.
Laboratory studies also showed that the collection efficiency of the
sampling train for formaldehyde was less than 100%, which means the reported
iv
-------
Environmental Sciences Center
ES-TR-80-04
emission rates may be considered lower limits. Since the aldehyde results are
based on one sampling run, their representativeness cannot be determined with-
out additional source testing.
-------
~
Environmental Sciences Center ES-TR-80-04
CONTENTS
SECTION PAGE
Disclaimer ii
Foreword iii
Abstract iv
Figures viii
Tables ix
Abbreviations and Symbols x
Acknowledgment xi
1 Introduction 1
2 Conclusions 3
3 Recommendations 5
4 Process Description 7
Recycling Processes 7
Energy Recovery Processes 8
5 Results 11
Source Test Results 12
Laboratory Results 21
6 Discussion 31
Inorganic Determinations 31
Organic Determinations 32
Formate and Acetate 33
Collection Efficiency Studies 34
Interference Determinations 36
Relative Impinger Distributions of HCOO and OAc 39
Concluding Remarks 40
References 43
Appendices
A. Sailing and Analytical Procedures 45
B. Calculations 56
vii
-------
Environmental Science* Center
FS-TP-an-na
FIGURES
NUMBER PAGE
1 Wet process energy recovery system 9
2 Gas chromatogram of the impinger extract from Organic Run #4 . 15
3 Gas chromatogram of XAD-2 cartridge extract from Organic
Run #4 16
4 Original IC analysis of CCS-l-Hempstead sample fractions
using a strong eluent system 18
5 IC analysis of CCS-l-Hempstead sample fractions using
a weak NaHCO^ eluent system 19
6 IC analysis of CCS-l-Hempstead sample fractions using
a weak Na^B^O^ eluent system 20
7 Effect of OH concentration on oxidation of CCS-S-^O^
impinger sample 23
8 IC analysis of standards in 0.6% H^O^ using a weak NaHCO^
eluent system 24
9 IC analysis of standards is 16* IPA using a weak NaHCO^
eluent system 25
10 IC analysis of standards treated with H^O^/NaOH using a
weak NaHCO^ eluent system 26
11 IC analysis of laboratory sample #6 before and after
treatment 27
12 IC analysis of CCS-l-Hempstead sample before and after
treatment 30
A-l Schematic drawing of manual acid condensation system.
A sampling pump and dry gas meter are contained within
the pumping meter box 49
viii
-------
jJClJG Environmental Sciences Center ES-TR-80-04
TABLES
NUMBER PAGE
1 Hempstead Master Sampling Data 12
2 Chloride Results (ppm) 12
3 Sulfur Oxide Results (ppm) 13
4 Organic Results (mg) 13
5 Acetate and Formate Results (mg) 13
6 Inorganic Species Mass Emission Rates (lb/h) 14
7 Organic Species Mass Emissions Rates (lb/h) 14
8 Major GC Peaks from Organic Run #4 14
9 MS Identification of Gas Chromatogram of XAD-2 Cartridge
extract from Organic Run #4 17
10 Laboratory Aldehyde Master Sampling Data 21
11 Formate and Acetate Results CCS Laboratory Samples 22
12 Summary of Carbonyl Interference Studies 28
13 XRF Analyses of CCS-l-Hempstead Sample 28
14 Flue Gas Formaldehyde Content of Various Sources by
Formate Analysis of CCS Samples 29
ix
-------
Environmental Sciences Center
ES-TR-80-04
ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
CCS Controlled Condensation System
D.I. H2O deionized water
DNPH 2,4-dinitrophenylhydrazine
EPA -U.S. Environmental Protection Agency
ESP electrostatic precipitator
GC gas chromatography
HC1 hydrochloric acid
HCOOH formic acid
HPLC high performance liquid chromatography
IC ion chromatography
Imp impinger
IPA isopropyl alcohol
MeOH methyl alcohol
MS mass spectrometry
MW molecular weight
NSI-ES Northrop Services, Inc.-Environmental Sciences
OAc acetate
PCB's polychlorinated biphenyls
PCP's polychlorinated phenols
PP probe plug
PW probe wash
RDF refuse-derived fuel
RT retention.time
XRF x-ray fluorescence
QA quality assurance
RDF refuse-derived fuel
RT retention time
VFR volumetric flow rate
XRF x-ray fluorescence
SYMBOLS
Cl~ chloride
CH2O formaldehyde
C2HJ+0 acetaldehyde
c2h6ฐ2 ethylene glycol
FeCl^ ferric chloride
HCOO formate
H2SO4. sulfuric acid
Na2C03 sodium carbonate
NaHC03 sodium bicarbonate
NaOH sodium hydroxide
Na2Bi,07 sodium borate
SO sulfur oxides
x
SO2 sulfur dioxide
-------
3
Environmental Sciences Center ES-TR-80-04
ACKNOWLEDGMENT
It is with pleasure that we acknowledge the work of Mr. Gary Grotecloss
and Mr. Mike Pleasant who conducted the source sampling and participated in
the preparation of our initial report. With equal pleasure we acknowledge the
assistance of our colleagues, Dr. John Windsor and Mrs. Sandy Parks, for ob-
taining and interpreting the gas chromatography/mass spectrometry results
presented in this report. We also acknowledge the work of our EPA colleagues,
Mr. Jim Homolya and Mr. Jim Cheney, for their many helpful discussions.
-------
Environmental Sciences Center
ES-TR-80-04
SECTION 1
INTRODUCTION
Northrop Services, Inc.-Environmental Sciences (NSI-ES) conducted source
sampling on Furnace #2 at the Hempstead Resources Recovery Corporation Facility
in Hempstead, New York from July 24 to July 26, 1979. The purpose of this sam-
pling was to characterize hydrochloric acid (HC1) and organic emissions pro-
duced from the combustion of nonreclaimed refuse. Sampling was also conducted
by the U.S. Environmental Protection Agency (EPA) for sulfuric acid (H^SO^)
and sulfur dioxide (SO^), while NSI-ES conducted supplementary oxygen monitor-
ing. Since the plant was down (out of operation) for considerable periods
of time during the week of sampling, only a few runs were completed.
EPA Region II has requested follow-up laboratory work in order to validate
the results obtained from analyses of the Hempstead field samples. This paper
reports this work, conducted by NSI-ES and EPA, and gives the results of addi-
tional laboratory experiments conducted by NSI-ES.
1
-------
Environmental Sciences Center
Fs_TR_an-n4
SECTION 2
CONCLUSIONS
The results of source sampling on Furnace #2 at the Hempstead Refuse
Energy Recovery System indicated low concentrations of chloride (CI ) and
sulfur oxides (SO ) in the flue gas. Because of the considerable time that
x
the plant was down during the test, sampling performed on July 26, 1979 is
considered to be most representative of the plant operation. At an operating
capacity of about 75%, the average mass emission rate for total CI was only
28 lb/h (15 lb/h particulate matter (PM), 13 lb/h gas). The mass emission
rate of SO^ averaged 51 lb/h (4.5 lb/h PM, 46 lb/h SO^ gas). Based on past
experience with other sources, these levels are not considered environmentally
harmful.
Source test results for organic emissions were consistent over three days
of sampling, producing an average emission rate of 26 lb/h for total organics.
This value must be taken as a lower limit, however, because the absolute col-
lection efficiency of the sampling train has not been determined. The samples
were qualitatively analyzed using gas chromatography (GC) and GC/mass spectro-
metry (GC/MS).
The samples were found to contain several potentially hazardous materials,
including substituted phthalates, chlorinated phenols and chlorinated biphenyls.
Since the analysis was not quantitative, whether or not these species were
present at levels great enough to pose human/environmental risk could not be
determined. Additional analyses have been performed on these samples to
identify any other species undetectable by GC/MS surveys. These results will
be reported at a later date.
3
-------
Environmental Sciences Center
ES-TR-S0-04
SECTION 3
RECOMMENDATIONS
For future field collection of flue gas samples at sources similar to
the Hempstead plant, methods development and additional source characterization
are recommended within the following framework:
Development of an organic sampling system designed to collect
low molecular weight species quantitatively. This system may
be constructed using a cooled resin cartridge and thermal de-
sorption recovery.
Development and rigorous testing of a source sampling train for
aldehyde emissions, possibly using hydrazine impinger solutions
with high performance liquid chromatography (HPLC) analysis
.techniques.
Additional source testing at the Hempstead plant using the
developed systems, and simultaneous testing using the techniques
within this study. An effort should be made to sample over an
extended period since the composition of the refuse-derived
fuel (RDF) and consequent stack emissions may vary substantially.
Onsite analysis of collected samples, if feasible, when using new
or prototype techniques.
5
-------
Environmental Sciences Center
ES-TR-80-04
Formate (HCOO ) and acetate (OAc ) were detected in the hydrogen peroxide
(HjOj) impinger of the Controlled Condensation System (CCS) by ion chromato-
graphy (IC). The identities of these species were verified through numerous
laboratory detection methods, confirming that these species are indeed formate
and acetate.
NSI-ES proposes that the species present in the flue gas of the Hempstead
plant were emitted as formaldehyde (CH^O) and acetaldehyde (C^H^O), but were
subsequently oxidized to formate and acetate, respectively, in the H2ฐ2
pinger of the CCS sampling train. The efficiency of this oxidation process
is less them 100%, however. Literature and laboratory studies have not revealed
other species which could be responsible for the observed IC peaks, though
time constraints did not permit a complete laboratory interference study.
Hence the involvement of another species cannot be ruled out.
The lower limits for the mass emissions rates of formaldehyde andซace-
taldehyde are calculated at 28 lb/h and 28 lb/h, respectively, although these
values may not be truly representative since they are based on only one sample.
4
-------
Environmental Sciences Center ES-TR-80-04
SECTION 4
PROCESS DESCRIPTION
The Hempstead recovery plant converts municipal refuse into two usable
forms: recyclable and combustible materials (The Black Clawson-Parsons and
Whittemore Organization 1974). The plant operations described are the re-
cylcing processes and the energy recovery processes.
RECYCLING PROCESSES
Raw waste entering the plant is first fed into a solid-waste-type hydra-
pulper, where it is ground, shredded, and eventually pumped away as a water
slurry. Unshredded material heavy enough to sink against a countercurrent of
water is removed from the pulper. This material is then treated to remove
ferrous metals, nonferrous metals and glass. The ferrous metals are removed
by a magnet, and the nonferrous metals are removed by a high-speed, overhead,
rotating electromagnetic drum. The nonmagnetizeable fraction drops into the
nonferrous hopper, where it is passed over a grizzly screen that removes large
nonrecoverable residue. The smaller items are recycled to the hydrapulper for
further processing.
The original material carried off by the water is treated to recover glass,
aluminum, other nonferrous metals, and an assortment of inorganic materials that
includes buttons, stones, pieces of broken china, etc. The heavy inorganic
fraction is separated in a liquid cyclone and discharged through an opening in
the bottom. This heavy fraction is washed to remove any residual light material
(which is returned to the hydrapulper), then is rewashed and screened. The
over-sized particles are processed again, but the fine particles are dewatered
and conveyed to fuel storage.
7
-------
m
Environmental Science* Center
ES-TR-80-04
The plastics and lighter materials are separated from the heavy fraction
by surface skimming in a heavy media separator. This light material is conveyed
to fuel storage, where it is dried and sized. A magnet removes residual fer-
rous metals and a shaking table removes aluminum foil. Nonconductors are re-
moved from conductors by high voltage electrodes that permit the nonconductors
to fall into a recovery drum. The remaining nonconducting fraction is mostly
glass, china, porcelain and occasional small stones. The glass is separated
at a transparency sorter and is actually separated by color with the use of
colored filters to identify the clear, amber and green fractions.
The energy recovery processes begin with the feeding of the cyclone-treated
slurry from the hydrapulper to a large surge tank (see Figure 1). The slurry
is pumped into a two-stage dewatering press apparatus where the solid content
is raised to about 50%. The water discharge is processed and returned to the
cycle; the solid is transferred to the fuel storage area for burning (usually
within a week). The fuel is burned in two steam-generating furnaces with a
nominal capacity of 200,000 lb/h of steam. The maximum fuel feed rate is
90,000 lb/h. The particulate emissions from the furnaces are controlled by the
cyclones and electrostatic precipitators (ESP's) listed below:
ENERGY RECOVERY PROCESSES
Cyclones
Manufacturer
FLAKT
Type
12 x CKOB 180
Number of units 12 per furnace
Electrotatic Precipitators
Manufacturer
FLAKT
Type
FAA 323212090-1-SP
Number of units 1 per furnace
8
-------
BARREL PRESS
(THICKENER)
GENERATOR
LOW PRESSURE
STEAM
TO PROCESS
u>
TO RECOVERY 0T
ALUMINUM
SMALL FERROUS METALS
GLASS(B? COLO")
TIPPING FLOOR
NONFERROUS MATERIALS
RETURN TO PROCESS
|<~1^*1 RECOVERED
FERROUS METAL
Figure 1. Wet process energy recovery system.
(ZD
O]
5
a
3
ft
3
?
(/>
O
5*
3
O
*
O
ft
3
tn
i
H
sO
I
CO
0
1
o
p*
-------
Environmental Sciences Canter
ES-TR-80-04
In summary, the solid waste that is burned as fuel contains about 25% mois-
ture, 20% inorganics, and 35% organic combustibles such as paper, wood, plastics
and food waste. Additionally, the water used in the process has been treated
with a biocide that is certainly retained in the 25% moisture fraction of the
fuel.
10
-------
JSO Environmental Sciences Center ES-TR-80-04
SECTION 5
RESULTS
In this section the results from the Hempstead study are given in two
parts: source testing and laboratory investigations. The first subsection
contains all results thus far obtained from Furnace #2 of the plant pertaining
to chloride, sulfur oxides, and other organic emissions of higher molecular
weight. Since velocity traverses were not made, a volumetric flow rate previ-
ously obtained for Furnace #1 (4.4 x 10ฎ SCF/h) was used for calculating mass
emission rates (New York Testing Laboratories 1979). Process load data in
Tables 1 and 6 are given so that runs can be compared (Ogg, private communica-
tion) . Results of IC analyses on field samples for the formate and acetate
ions are also presented in this section.
Additional studies were requested by EPA to determine the presence of for-
maldehyde and acetaldehyde in the flue gas. Laboratory investigations into the
collection and analysis of these species are presented in the latter subsection.
11
-------
333 Environmental Sciences Center Eo-TR-80-04
SOURCE TEST RESULTS
TABLE 1. HEMPSTEAD MASTER SAMPLING DATA
Date
Time
HC1
CCS
HC1-ESP
Organic
% 09
% Operating
Run #
Run #
Run #
Run #
c
Capacity*
7-24-79
14:30-15:30
1
1
11.5
49
16:30-17:30
1
49
7-25-79+
09:30-10:30
2
56
09:30-10:30
l
56
12:30-12:45
7-26-79
10:30-11:30
3
2
76
15:30-17:00
4
3
7.3
74
16:30-18:00
5
4
9.1
74
*
Based on 200,000 lb/h steam generation.
^Sampling was interrupted at 10:30 and resumed at 12:30 due to a plant shutdown
from 10:00 to 11:50 and from 12:35 to 13:25 on 7-25-79.
TABLE 2. CHLORIDE RESULTS (ppm)
HC1 Sampling Train
Kun w
SCF
PP
PW x 10"2 Imp x 10"2
Tot x 10~2
l
1.874
8.4
1.2
.87
2.1
2
.910
2.5
o
o
.00
.025
3
.865
8.2
.37
.31
.77
4
.985
6.5
.29
.30
.66
5
1.331
7.0
.24
.34
.65
HC1-ESP Sampling Train
, ESP NaOH , NaOH
SCF PU ESP x 10'^ Filter Imp x 10'^ Imp Tot x 10
1
26.241 .63
.55
1.2
2.5
3.2
3.1
12
-------
jjgg _ ... ES-TR-S0-04
Environmental Sciences Center
TABLE 3. SULFUR OXIDE RESULTS (ppm)
n ^ HC1 Sampling Train
SCF PR PW Imp x TO"2 Tot x 10"2
1 1.874 .99 .26 1.8 1.8
2 .910 .44 .16 .00 .0060
3 .865 .95 5.5 .45 .52
4 .985 .76 2.5 .46 .49
5 1.331 .94 2.9 .47 .51
CCS Sampling Train
SCF PW
Special
Filter Plug
I PA
Imp
h2ฐ2 _2
Imp x 10"
Tot x 10~2
1
9.121 .09
.37 .18
.00
1.5
1.5
TABLE 4. ORGANIC RESULTS (mg)
Run (J Orajnics.
SCF
Imp
XAD-2 Column
1
19.471
*
5.3
2
21.771
3.5
5.3
3
19.798
5.4
4.6
4
23.657
3.1
6.2
f1 Tot x 10"1
5.3
5.7
5.1
6.5
*Saiople discarded.
TABLE 5. ACETATE AND FORMATE RESULTS (mg)
Formate
Acetate
IV JII It
I PA
Imp
H2ฐ2
Imp
(x TO"1)
Tot
(x 10"1)
I PA
Imp
H202 Tot x 10'1
Imp
(x 10"1)
1 CCS
1.8
3.6
3.8
2.3
3.3 3.5
13
-------
ฃ]ฃ][] _ ES-TR-80-04
Environmental Sciences Center
TABLE 6. INORGANIC SPECIES MASS EMISSION RATES (lb/h)
Run #
% Operating
Capacity
pm cr
(x 10"1)
Gaseous CI
(x 10"1)
Tot cr
(x 10"1)
PM*
S04
Gaseous SOg
(x 10-2)
Tot S0X
(x 10"2)
1 HCl
49
5.2
3.5
8.7
1.4
1.4
1.4
2 HCl
56
.11
.00
.11
. 66
.00
.0066
3 HCl
76
00
1.3
3.1
7.0
.33
.40
4 HCl
74
1.5
1.2
2.7
3.5
.33
.37
5 HCl
74
1.2
1.4
2.6
4.2
.35
.39
*
Mass emission rates for oxides of sulfur are based on the analysis of the
HC1 sampling train.
TABLE 7. ORGANIC SPECIES MASS EMISSION RATES (lb/h)
Run # % Operating CHgO x 10"1 CgH^O x 10"1 Organics x 10"1
Capacity
1 CCS 49 2.8 2.8
1 ORG 49 2.6
2 ORG 76 2.5
3 ORG 74 2.5
4 ORG 74 2.7
Figure 2 represents the GC spectrum of the impinger catch of Organic Run
#4. Table 8 gives the major peaks in the chromatogram, identified through MS.
TABLE 8. MAJOR GC PEAKS FROM ORGANIC RUN #4
Peak #
Identification
1.3
2
Polychlorinated phenols (di-, tri-, tetra-, penta-)
Phthalates and phthalate isomers
14
-------
LIU
l_JL)
in
7-
t 1 1 1 1 1 1 1 1 1 1 1 r
5 10
MINUTES
T 1 . r
15
t 1 1 1 1 1 1 1 r
20 25
Figure 2. Gas chromatogram of the impinger extract from Organic Run #4.
CO
I
I
TO
l
oo
0
1
o
-------
G\
11
ULAJ
~T~
15
i r
Figure 3. Gas chromatogram of XAD-2 cartridge extract from organic run #4.
LJ
QD
r~m
3
3
3
ft
3
sr
o
5*
3
n
ft
u
O
ft
3
CO
I
i
yo
I
CD
O
I
o
-F*
-------
_JBl] Environments! Sciences Center ES-TR-80-04
TABLE 9. MS IDENTIFICATION OF GAS CHROMATOGRAff OF
XAD-2 CARTRIDGE EXTRACT FROM ORGANIC RUN =4
Peak # Identification
1 Butylbenzene isomers
2 c10h12 isomers
3 c11H16
4 C10H12
5 Divinyl benzene isomer (tentative)
6 c11h16
7 Ciqh10 isomer
8 c12h18 isomers
9 Napthalene
10 C10H10O
11 Methyl napthalenes
12 Biphenyl
13 Ethyl napthalene
14 Methyl biphenyl
15 Alkylated tetrahydronapthalenes
16 Methyl biphenyl
17 Ci3H1602 e.g. (tentative)
18 6r?-0 isomer
19 m 0ch,-^Cm. or isomer
20 isomers of
21 isomers of 0~^-c'M5
22 Ci7H10 multialkylated biphenyl isomers
H
e-9-
CHj~0 I
Crt
23 Ci7Hio multialkylated biphenyl isomers
24 Anthracene/phenanthrene
25 Methyl anthracene/phenanthrene
26 Dichlorobenzophenone
17
-------
300 ES-TR-80-04
Environmental Sciences Center
CALIBRATION STANDARD
A ฆ F" (5 ppm), 2.50 min
B - CI" (10 ppm), 3.79 min
C - SO3 (30 ppm), 6.43 min
D SO4 (50 ppm), 10.38 min
CCS-I-H2O2 (Hempstead)
A Unknown, 2.73 min
B - Unknown, 3.31 min
C CI", 3.84 min
D -SO4, 10.10 min
INSTRUMENTAL CONDITIONS
Columns:
El Eluent:
Flow Rate:
Sample Loop:
Meter Setting:
3 x 150 mm pre-column
3 x 500 mm Separator column
3 *6!50 mm Supressor column
.006 M Na2C03
130 ml/h (25%)
100 m*
10 Mmho/cm full scale
CCS - I - IPA (Hempstead)
A Unknown, 2.72 min
B CI", 3.66 min
CALIBRATION STANDARD
A - HCOO" (10 ppm), 2.73 min
5 0 15 20
Figure 4. Original IC analysis of CCS-l-Hempstead sample fractions using a
strong eluent system.
18
-------
309
Environmental Sciences Center
ES-TR-80-04
B
A
A
B
rj
A
I
1 s
J L
M
10
15
20
CALIBRATION STANDARD
A - F" (5 ppm), 3.45 min
B - HCOO" (30 ppm), 5.21 min
C Cf (20 ppm), 16.88 min
CALIBRATION STANDARD
A - OAc' (40 ppm), 4.04 min
INSTRUMENTAL CONDITIONS
CCS I H202 (Hempstead)
A - CO3 =, 2.99 min
B - Oac", 3.91 min
C- HCOO", 5.16 min
D-CI\ 16.84 min
Columns:
Eluent:
Flow Rate:
Sample Loop:
Meter Setting:
3 x 150 mm pre-column
3 x 500 mm Separator column
6 x 250 mm Supressor column
.0015 M NaHC03
103 ml/h (20%)
100
10/imho/cm full scale
CCS - I I PA (Hempstead)
A - CO3 =, 2.96 min
B - HCOO", 4.72 min
C-CI", 14.92 min
Figure 5. IC analysis of CCS-l-Hempstead sample fractions using a weak
NaHCOg eluent system.
19
-------
309
Environmental Sciences Center
ES-TR-80-04
B
B
I
A,
CALIBRATION STANDARD
A - F" (4 ppm), 2.21 min
B HCOO' (20 ppm), 3.48 min
C CI' (10 ppm), 13.21 min
CALIBRATION STANDARD
A OAc* (40 ppm), 2.67 min
INSTRUMENTAL CONDITIONS
Columns:
Eluent:
Flow Rate:
Sample Loop:
Meter Setting:
3 x 150 mm pre-column
3 x 500 mm Separator column
6 x 250 mm Supressor column
.005 M Na2B407
156 ml/h (30%)
100 ui
10 /^mho/cm full scale
CCS I HqO? (Hempstead)
A OAc", 2.64 min
B - HCOO', 3.46 min
C- CI", 12.54 min
A.
CCS ฆ I - IPA (Hempstead)
A - OAc", 2.46 min
B - HCOO', 3.31 min
C -CI", 10.13 min
D - CO3 =, 17.44 min
10 15 20
Figure 6. IC analysis of CCS-l-Hempstead sample fractions using a weak
NagB^Oy eluent system.
20
-------
~SB
Environmental Sciences Center
ES-TR-80-04
LABORATORY TEST RESULTS
TABLE 10. LABORATORY ALDEHYDE MASTER SAMPLING DATA
Run #
Sampling Train
Description
Sample
Source
standard CCS train:
1st imp 80% IPA
2nd imp 3% ii2ฐ2
same
ambient air
formalin
(3.7% soln)
same
acetaldehyde
(liq)
modified CCS train: formalin
2nd imp 3% H^/.IN NaOH (37% soln)
modified CCS train: same
3rd imp 3% HjOj/S Fe
4th imp DNPH/HC1 (aqueous soln)
modified CCS train:
3rd and 4th inฎ DNPH/MeOH
formalin
(7.4% soln)
21
-------
~so
Environmental Sciences Center
FS-TR-Sn-04
TABLE 11. FORMATE AND ACETATE RESULTS - CCS LABORATORY SAMPLES
Run #
Sample Description
Imp #
Imp
Contents
Analysis for HCOO"/OAc
Before Purge
Untreated w/added ^
H202/Na0H'
After Purge
Untreated w/added .
H202/Na0HT
1
2
1
2
1
2
3
1
2
1
2
3
80% IPA
3% H_0
2 2
tot
80% IPA
3% H202
tot
80% IPA
3% H202 (1/24/80)
3% H202 (2/29/80)
tot
80% IPA
3% H202/.l N NaOH
tot
80%.IPA
3* h2o2
3*8,0.,/
5ppm Fe
DNPH/HC1
(aqueous)
14.6
19.3
69.3
52.7
547
201
199
(52.7)
0
0
0
0
0
0
8.2
169
137
18.6
16.1
8.2
17.4
0
0
0
26.6
14.9
41.5
263
566
214
74.9
(17.4)
tot
155.9
999.7
60.3
872.3
1
80% IPA
-
246
16.3
249
2
3% H202
-
68.6
3.8
75.1
3
DNPH/MeOH
-
-
1.0
(1.0)
4
DNPH/MeOH
-
-
0.15
(0.15)
tot
314.6
21.3
325.3
* ,
Total mg/10 ft3 sample.
TIPA - sample adjusted to 0.6% H 0 (V/V) and .005 M NaOH; - sample adjusted
to .005 M NaOH.
22
-------
u Environmental Sciences Center
ES-TR-80-04
[ OH"] moles/liter
Figure 7. Effect of OH" concentration on oxidation of CCS-S-H^ impinger
sample.
23
-------
000
Environmental Sciences Center
ES-TR-80-04
A^;
/ I
I 1
A i
A'
10
15
20
CALIBRATION STANDARD
A - OAc' (20 ppm), 3.73 min
B - HCOO" (20 ppm), 4.56 min
INSTRUMENTAL CONDITIONS
REAGENT BLANK
(0.6% H2O2 - v/v)
A - HO2, 2.47 min
Columns:
Eluent:
Flow Rate:
Sample Loop:
Meter Setting:
QA STANDARD
(30 ppm HCOO" in 0.6% H2O2)
A - HO2, 2.48 min
B - HCOO", 4.59 min
QA STANDARD
(1.6% I PA in 0.6% H2Q2-v/v)
A - HO2, 2.55 min
3 x 500 mm Separator column
6 x 250 mm Supressor column
.0015 M NaHC03
130 ml/h (25%)
100 nl
10 /^mho/cm full scale
Figure 8. IC analysis of standards in 0.6X ^2 using a weak NaHCC^ eluent system.
24
-------
3S3
Environmental Sciences Center
ES-TR-80-04
INSTRUMENTAL CONDITIONS
Columns: 3 x 500 mm Separator column
6 x 250 mm Supressor column
Eluent: .0015 M NaHC03
Flow Rate: 130ml/h (25%)
Sample Loop: 100^
Meter Setting: 10|imho/cm full scale
QA STANDARD
(30 ppm HCOO" in 16% I PA ฆ v/v)
A - HCOO", 4.55 min
QA STANDARD
(16% I PA in D.I. H;Q - v/v)
(peak not detected)
QA STANDARD
(16% I PA/0.6% H2Q2 - v/v)
A HO"2 ฆ v/v) , 2.53 min
Figure 9. IC analysis of standards is 16% IPA using a weak MaHCO^ eluent system.
25
-------
m
Environmental Sciences Center
ES-TR-80-04
B
ar
A B
\A
_J
aB!
QASTANDARD
(0.6% H202/.Q05M NaOH)
A H02, 2.53 min
B - OH", 3.83 min
QA STANDARD
(20 ppm HCOO" in 0.6% H2O2/ 0Q5 M NaOH)
A - HOj, 2.56 min
B - HCOO', 4.34 min
INSTRUMENTAL CONDITIONS
Columns:
Eluent:
Flow Rate:
Sample Loop:
Meter Setting:
QA STANDARD
(16% I PA/0.6% H202/.005 M NaOH)
A - H02- 2.56 min
B-OH', 3.45 min
3 x 500 mm Separator column
6 x 250 mm Supressor column
.0015 M NaHC03
130 ml/h (25%)
100 (J
10 /xmho/cm full scale
QA STANDARD
(30 ppm HCOO' in 0.6% H2Q2/16% IPA/.005 M NaOH)
A H02, 2.62 min
B - OH', 3.48 min
C- HCOO", 4.41 min
10
15
20
Figure 10. IC analysis of standards treated with HpOp/NaOH using a weak
NaOH eluent system.
26
-------
Environmental Sciences Center
ES-TR-80-04
B
kj
10x DILUTION
A B
B
2x DILUTION
A I 1
_nL
10
15
20
CCS - 6 - I PA
(Before Treatment)
A - IPA, 3.40 min
B - HCOO', 4.24 min
CCS-6- IPA
(After Treatment)
A - HO2, 2.56 min
B HCOO", 4.26 min
CCS -6-H202
(Before Treatment)
A H02.2.46 min
B HCOO'. 4.22 min
CCS-6-H202
(After Treatment)
A HOjj, 2.55 min
B - HCOO", 4.36 min
INSTRUMENTAL CONDITIONS
Columns: 3 x 500 mm Separator column
3 x 250 mm Supressor column
Eluent: .0015 M NaHCO,
Flow Rate: 130 ml/h (25%)
Sample Loop: 100/4
Meter Setting: 10 jimho/cm full scale
Figure 11. IC analysis of laboratory sample #6 before and after treatment.
27
-------
Environmental Sciences Center
TABLE 12. SUMMARY OF CARBONYL INTERFERENCE STUDIES
Analytical Results
Sample
Standard
Standard
After ~24 h
After ~96 h
Concentration
(ug/ml)
RT+ (min)
RT
(min)
HCOO" OAc"
(ug/ml) (ug/ml)
RT
(min)
HCOO" OAc"
(ug/ml) (ug/ml
formaldehyde
100
4.41
4.35
7.577
4.40
22.510
acetaldehyde
100
3.61
3.56
1.072
3.57
9.693
propion-
aldehyde
100
3.85
4.34
0.541
4.34
4.139
acetone
100
-
3.92
0.244
3.70
4.28
0.219
0.318
butyraldehyde
100
3.99
4.31
1.046
4.35
2.610
benzaldehyde
100
24.01
4.27
0.372
3.53
4.38
0.476
1.380
*
500 mg/ml in 3% H 0^ purged 15 min
prior to analysis).
with
zero air (sample
diluted 4:1 at H^O
'standards were prepared solutions of the acid or salt in
D.I. H20.
TABLE
13. XRF ANALYSIS
OF CCS-1-HEMPSTEAD
Element
IPA Imp*
(ug/ml)
H70-/Imp
(ug/ml)
Na
138
107
Mg
.11
.09
A1
.17
.01
S
.16
47
K
.14
.09
V
.04
.02
Mn
.09
.08
Co
.14
.11
Cu
.20
.09
Zn
.20
.11
Br
.71
.18
Cd
.015
-
Ba
.02
.01
Pb
.95
.91
*
Analysis performed on diluted impinger catches (dilution volume = 100 ml).
28
-------
Environmental Sciences Center
TABLE 14. FLUE GAS FORMALDEHYDE CONTENT OF VARIOUS SOURCES BY
FORMATE ANALYSIS OF CCS SAMPLES
Source
Sample
HCOO-
Found (mg) Converted to ^0 in
Stack (ppm)
Untreated
w/added H202/NaOH
HCOO"
ch2o
HCOO"
CH20
refuse-fired
IPA imp
1.9
3.9
14.0
29.0
boiler
H.O. imp
48.7
101
75.7
157
(Hempstead)
tit2
50.6
104.9
89.7
186.0
hogged-fuel-
IPA imp
2.0
4.4
18.4
40.8
fired
H_0 imp
349
773
474
1050
boiler
/. i.
tot
351.0
777.4
492.4
1090.8
coal-fired
IPA imp
0
0
1.0
1.6
boiler
HO imp
13.7
22.6
14.3
23.5
tit2
13.7
22.6
15.3
25.1
*
IPA imp - sample adjusted to 0.6* H.O (V/V) and .005 M NaOH, HO imp - sample
adjusted to .005 M NaOH.
29
-------
Environmental Sciences Center
ES-TR-80-04
-C
Ab
B
B
CCS I - I PA (Hempstead)
(Before Treatment)
A CO3, 3.61 min
B - CI', 13.10 min
CCS - I - I PA (Hempstead)
(After T reatment)
A - CO3/OH', 3.65 min
B - HCOO', 4.24 min
C- CI', 12.91 min
INSTRUMENTAL CONDITIONS
Columns: 3 x 500 mm Separator column
6 x 250 Supressor column
Eluent: .0015 M NaHC03
Flow Rate: 130ml/h (25%)
Sample Loop 100 til
Meter Setting: 10 At mho/cm full scale
CCS I - H2O2 (Hempstead)
(Before Treatment)
A - CO f/OAc, 3.62 min
B - HCOO', 4.47 min
C - CI", 13.47 min
CCS - I - H2O2 (Hempstead)
(After Treatment)
A - CO3/OHVOAC, 3.80 min
B - HCOO', 4.70 min
0 5 10 15 20
Figure 12. IC analysis of CCS-l-Hempstead sample before and after treatment.
30
-------
f*]~1 Environmental Sciences Center E$-TR-งQ-0'1
SECTION 6
DISCUSSION
For clarity, the results of source and laboratory testing have been
divided by compound class: (1) inorganics, (2) organics of high molecular
weight, and (3) formate and acetate. The inorganic work has been completed;
additional analysis of the organic samples is under way and will be reported
at a later date. Because of the pertinence to source samples collected at
Hempstead, results of a limited laboratory study on the collection and analysis
of aldehydes are also reported here.
INORGANIC DETERMINATIONS
The results presented in Tables 2, 3 and 6 (pp. 12-14) indicate that the
emission levels of chloride and sulfur oxides are indeed quite low in com-
parison to power plants and other incinerators that have been characterized
(Jahnke et al. 1977, Homolya et al. 1976). The average total chloride con-
centration emitted as stack effluent at Hempstead was only 120 ppm, while the
average gas phase concentration was only 73 ppm.
The results from Run #1 of the HC1 train appear inconsistent with the re-
sults obtained in Runs #3, 4 and 5 on July 26, 1979. The results of both Run
#1 of the HC1-ESP train and Run #2 of the HC1 train appear to be anomalously
low. (Since these runs were interrupted by plant down time, they cannot be
considered representative of the source emissions.) In addition, the results
obtained from all runs made on July 24 appear to be inconsistent with the runs
completed on July 26. Although thorough checking of the process data revealed
that the plant was operating under a steady load, the percent operating capacity
31
-------
Environmental Sciences Center
ES-TR-80-04
was only 49% and the % 0^ was a high 11.5. Run $1 may thus not be representa-
tive of the source. If we can exclude the abo^e samples, we obtain average
total chloride and gaseous chloride concentrations of 69 ppm and 32 ppm,
respectively.
The CCS sampling of SO^ was performed only on July 24. Concentrations
were very low, with SO^ measured at 150 ppm and H2Sฐ4 at on^Y ppm. Again,
the results obtained on July 24 and 25 are not taken as representative of
plant operation. Past simultaneous sampling efforts with the HC1 and CCS
train have shown that the total sulfur oxides collected agree quite well. If
Runs #3, 4 and 5 of the HC1 train for sulfur oxide analysis are assumed valid,
an average sulfur oxide concentration of 51 ppm is obtained.
Although the existence of irregularities in the samples taken on July 24
(interruption in sampling) and July 25 (apparent disagreement with July 26
samples) has been emphasized, these differences observed during the three days
of sampling may be due to the varying composition of the RDF. Since many para-
meters may be affecting the source sampling, no one sample should be construed
as fully representative of the source emissions.
ORGANIC DETERMINATIONS
The results of sampling for organic emissions are presented in Tables 4,
7, 8 and 9 (pp. 13-14, 17). The total amounts reported for all four samples
are quite consistent, and yield an average mass emission rate of 26 lb/h. The
samples were collected in two fractions: species condensable at 0ฐC, and
species trapped in the Amberlite XAD-2 cartridge at ambient temperature.
Preliminary GC/MS analysis of Organic Run #4 indicated that polychlo-
rinated phenols (PCP's) and phthalates were the primary components of the
impinger catch. Small quantities of polychlorinated biphenyls (PCB's) were
also collected in the cartridge; all other compounds detected were relatively
innocuous species. The occurrence of these species in small quantities is not
32
-------
I0B3 Environmental Sciences Center
ES-TR-80-04
unexpected, considering the chemical makeup of the RDF and the similar findings
at other incinerators that have been reported (Eiceman et al. 1979). The source
of the chlorophenols was thought to be the biocide used in treating the refuse
(Betz CSD), until an investigation proved this chemical to be a chlorohvdantoin,
which could not directly form the chlorophenols. Alternatively, the chloro-
phenols may be formed by pyrolysis of phenolic polymers (Bakelite and Formica)
and subsequent chlorination by a chlorinating agent, such as chlorohydantoin.
The presence of chlorophenols is of prime concern, since these compounds
have been reported to undergo condensation to form dioxins under the proper
conditions (Buser 1978, Gribble 1974). Additional analyses performed on these
samples will be reported at a later date.
FORMATE AND ACETATE
During the course of routinely analyzing CCS samples for sulfate, a major
unidentified peak was observed in both the IPA and H202 impingers at a reten-
tion time of 2.73 min (see Figure 4, p. 18). A minor peak of similar retention
time had also been observed in an earlier run, and formate was suggested as
being the species present. Analysis of a formate standard gave the same reten-
tion time as the unknown peak in the Hempstead sample. Since the original
eluent system was not suitable for analysis of fast-eluting species, a weaker
eluent system was devised for running the samples. Results for both a weak
bicarbonate (NaHC03) and a weak borate (Na2B40?) eluent system are presented
in Figures 5 and 6 (pp. 19-20). Since the samples were initally prepared in
a carbonate eluent (in order to eliminate the water dip which would interfere
with analyses of chloride), a reagent blank was run first. This blank did
not include peroxide because previous experience has shown that most of the
peroxide either reacts or is thermally and photochemically destroyed in field
S
samples before analysis can be performed. The retention times for CO^ in the
NaHC03 and Na2B40? eluent systems were 2.98 and 19.54 min, respectively, and
therefore did not interfere in the HC00~/0Ac~ analysis. In addition, dilute
IPA solutions have been found to shorten the retention times of all ions. The
argument of retention times is actually very good when all the eluent systems
33
-------
JSO Environmental Sciences Center
ES-TR-8D-04
are considered, and more firmly identifies these peaks as formate and acetate.
Quantitative results are reported in Table 5 (p. 13) and mass emissions rates
in Table 7 (p. 14).
The formate and acetate were not detected, nor expected, in the NaOH im-
pinger of the HC1 train, since these species are acids or salts that would not
be collected in a basic solution. Due to the oxidizing nature of anc*
ease of oxidation of aldehydes, however, a logical source for the species would
be the parent compounds formaldehyde (CH^O) and acetaldehyde (C^H^O). The
presence of these species has been reported in low-temperature, inefficient
combustion sources (Harrenstien et al. 1979). The fact that the % Ofor the
CCS run was relatively high may indicate less efficient combustion during this
period of operation and further supports the aldehyde hypothesis.
Collection Efficiency Studies
Due to recent concern over the possible carcinogenicity of formaldehyde,
NSI-ES was requested by EPA to validate the source samples through additional
laboratory testing. The first phase of this testing was to determine the possi-
bility of artifact formation in the sample train and the collection efficiency
of the train for formaldehyde and acetaldehyde. A description of six laboratory
runs of the CCS system and the sample source is presented in Table 10 (p. 21) .
Table 11 (p. 22) lists the analysis results. No results are listed for analysis
of the probe wash, filter and special plug sample fractions, since no HCOO or
OAc" was detected in any of these fractions. Run #1 consisted of a blank in
which ambient air (laboratory) was sampled to test for artifact formation. In
other runs, formaldehyde (g) was generated by bubbling air through an impinger
containing formalin. Acetaldehyde (g) was generated and sampled in Run #3.
Initially, no formate or acetate was detected in the blank or in the formalde-
hyde run, and only small quantities of acetate were detected in the acetaldehyde
run.
The initial results of these tests indicated either that: (1) a very
small amount of CH^O was being generated, even with use of the 37% formalin
34
-------
Environmental Sciences Center
FS-TQ-an-ru
solution; (2) the IPA and Himpingers have a very poor collection effi-
ciency for CH^O; (3) the CH20 is collected by the impinger solutions, but is
only very slowly oxidized to HCOO ; or (4) some combination of the above.
Of the four possibilities, collection of CH20 with slow oxidation to
HCOO appeared to be most likely. Consequently, a test was conducted to deter-
mine the effect of acid, base, and transition-metal ions on the oxidation of
CH^O in a 3% H2ฐ2 solution. A test solution containing a small amount of
37% formalin in 3% H^O^ was prepared, and three aliquots of this solution
were treated individually: one with HC1, another with NaOH, and the third
with with ferric chloride (FeCl^), maintaining an untreated aliquot as a con-
trol. The samples were analyzed on the following day, with the result that
the degree of oxidation of CH^O to HCOO was reduced by a factor of 2 upon
treatment with HC1, but was increased by a factor of 15 upon treatment with
both NaOH and Fe+3. Since treatment of samples with Fe+3 involves addition
of the CI ion (a slowly-eluting ion in weak eluent systems) to the sample,
NaOH was used to speed oxidation.
A test was conducted to determine the concentration of OH ion required to
ensure complete oxidation of CH^O to HCOO in 0.6% H^O^ solution. Using the
3% H^O^ impinger catch from CCS Run #5 (after dilution to 1 liter with D.I.
H^O) as a control, aliquots of the sample were treated with additions of .5 N
NaOH such that the OH concentration ranged from .0005 to .01 M. The samples
were allowed to stand overnight prior to analysis for the HCOO ion. The re-
sults of this test, presented in Figure 7 (p. 23), indicate that a OH ion
concentration of .005 M is sufficient for complete oxidation of CH^O.
Based on these findings, the laboratory samples were routinely treated
with NaOH and analyzed for HCOO . Treatment of the IPA impinger with NAOH
yielded a slight increase in the amount of HCOO detected, but when these
samples were also treated with H202 (0.6%), the increase in HCOO was quite
dramatic (Table 11, p.22). Reanalyses of the system blank sanqples (Run #1)
after the treatment described above yielded no formate or acetate in either
impinger catch. Reanalysis of Run #2 samples after treatment showed the
-------
jJB3 Environmental Sciences Center . ES-TR-80-04
presence of HCOO , whereas none was detected in the original analysis. More
than 1 month after the original analysis, testing of the ^2^2 imP^n9er catc^
from Run #3 showed that of the total acetate detected in the treated sample,
64% was detected in the untreated sample, and only 3% of the total acetate was
detected in the original analysis of January 24, 1980.
At this point we may conclude that neither formate nor acetate is the re-
sult of artifact formation in the sample train. Both formaldehyde and acetalde-
hyde are collected in IPA and peroxide impingers and are converted to their cor-
responding acids by the action of 3% hydrogen peroxide, but this reaction is not
complete. Thus more aldehyde is collected than is apparent in analysis of the
samples as received from the field. The collection efficiency of the train
was evaluated by using treated impinger solutions and backup impingers to
collect breakthrough.
With the exception of the impinger solutions, Runs #4 and 5 were carried
through the same conditions as Runs #1, 2 and 3. Addition of NaOH to the
impinger in Run #4 produced a large increase in the amount of formate
detected relative to the untreated samples. Run #5 indicated that 75% of the
formate was collected in the IPA and H202 impingers as configured in the CCS
train at Hempstead.
In Run #6 an attempt was made to duplicate the level of formate detected
in the H202 impinger at Hempstead (80 mg) to more accurately evaluate its
collection efficiency. For this run a collection efficiency of nearly 100%
was attained by the first two impingers, compared to 75% for Run #5 where
considerably more formaldehyde was generated. Apparently the saturation point
was attained in Run #5, beyond which the collection efficiency dropped rapidly.
Interference Determinations
Formaldehyde and acetaldehyde are collected in the sampling train with
efficiency near 100%, but cannot be detected by IC with the same efficiency
36
-------
J35 Environmental Sciences Center . ES-TR-80-04
until subjected to the stronger oxidation of the NaOH/H^C^ solution. The final
task of our study was then to determine the potentially-interfering compounds
(i.e., other species) that may have been converted to formate and acetate by
the action of the ^2^2 so^-ut^on-
A thorough search of the literature revealed few potentially-interfering
species. Only ethylene glycol (C H O ) was clearly documented to be converted
2 6 2
to formate by the action of dilute (Miner and Dalton 1973). A brief
mention in the older literature of CO^ reduction by peroxide to produce for-
maldehyde or formic acid (HCOOH) was_ not confirmed or even mentioned elsewhere
(Fry and Payne 1930). Only the aldehydes appeared to be rapidly oxidized to
their acid forms. The breaking of a carbon-carbon bond would be necessary to
produce formic acid, for which enough energy does not seem available in the
system.
Throughout this laboratory study, the following steps were taken to test
for possible artifact formation and/or interferences to the analytical tech-
nique due to sample preparation and treatment: (1) routine preparation of
control standards containing the same amounts of each constituent as in the
samples being analyzed, (2) consistency in handling procedures, and (3) analy-
sis of each sample under the same conditions on the same day. An example of
such a control standard would be a sample prepared to contain 16% IPA (V/V),
0.6% H^O^ (V/V) and .005 mole/liter NaOH in D.I. B^O, and then allowed to
stand overnight. This standard would be analyzed along with an IPA impinger
sample (200 ml 80% IPA diluted to 1 liter) that had been treated with H^O^
and NaOH. Figures 8-10 (pp. 24-26) are representations of chromatograms of
typical laboratory samples and control standards generated in this study.
Figure 11 (p. 27) represents chromatograms of a laboratory sample before and
after treatment. Analysis of control standards containing IPA/H^O^ and
IPA/H202/Na0H of various representative concentrations indicated no oxidation
of IPA to HCOO or OAc . The presence of OH ion results in a positive inter-
ference to OAc determinations with the NaHCO^ eluent (but not with the
NaB,0_ eluent); however, this problem was not too serious since the laboratory
2 4 7
work mostly involved analysis of HCOO ion.
37
-------
Environmental Sciences Center
ES-TR-3Q-Q4
Tests were conducted to determine the possibility of interference to HCOO
due to either oxidation by H2ฐ2 comPoun<3s other than CH^O, or oxidation by
of compounds yielding products having the same retention time as the HCOO
ion. Laboratory samples of acetone and various aldehydes of higher molecular
weight were prepared at a concentration of 500 ug/ml in 3% ^2^2 ^ aS comParec*
to ~400 ug/ml HCOO in 3% H2ฐ2 ^or CCS-l-H^O^, Hempstead), then purged for
15 min using a cylinder of zero air, and allowed to stand overnight. Prior
to analysis, aliquots of the samples were diluted 4:1 with D.I. H^O, again
duplicating the actual sample recovery procedure. The samples were analyzed
a second time approximately 4 days after the initial preparation using the
procedure outlined above.
The results of these tests and the retention times of the corresponding
acids are given in Table 12 (pp. 28). Any confusion with formate or acetate
resulting from conversion of these aldehydes to their corresponding acid
should be very slight. In most cases, a slow appearance of peaks did seem
to occur that could possibly be misinterpreted as formate or acetate, though the
degree of interference is probably negligible. Nonetheless, benzaldehyde also
yielded a peak at the acetate retention time! Under no circumstances could
benzaldehye be oxidized by dilute H2ฐ2 t0 ^onn acetaldehyde, leading us to
conclude that the observed peaks are due to formaldehyde and acetaldehyde
impurities in the chemicals.
The fact that the retention times of the acids of these aldehydes are
significantly different from the retention times of formate and acetate, and
the fact that the original Hempstead sample identification was confirmed in
borate eluent, lead NSI-ES to propose that oxidation of other aldehydes is
not responsible for the results obtained at Hempstead. Borate could not be
used to analyze these samples due to the interference of ^2^2 in freshly-pre-
pared samples.
Since the flue gas of a combustion source contains a relatively large
amount of carbon dioxide (~10%), tests were conducted to determine the possi-
bility of the reduction of CO^ to formic acid in the CCS impinger train. Using
38
-------
Environmental Sciences Center FS-TR-flO-flfl
a cylinder of CO^ (5%) in air, a sample was collected in which the I?A impinger
was treated with HC1 and the impinger with H2Sฐ4 This step was taken to
duplicate the time-averaged concentrations of CI and SO^ that were determined
in analyses of the IPA and impinger catches of the CCS-l-Hempstead sample,
respectively. A second sample was collected under the same conditions as the
first, except that both impingers were spiked with vanadium (0.1 ppm), and the
IPA impinger with zinc (1 ppm). Selection of these two metals was based on
analysis of the actual field samples by X-ray fluoresence spectrometry (XRF).
(The results are presented in Table 13, p. 28.) Analysis of these samples for
HCOO yielded negative results in all cases, as well as verified that no reduc-
tion of CO^ occurs under these conditions.
Relative Impinger Distributions of HCOO" and OAc"
The only major discrepancy between the results of anlayses of the labora-
tory samples and those of the actual field samples is the relative distribution
of HCOO and OAc in the impingers. The two impinger samples from the field
test were, of course, reanalyzed using the H^O^/NaOH pretreatment procedure.
The values of both HCOO~ and OAc were found to be significantly greater in
both samples; however, the amount found in the IPA impinger catch remained
quite low with respect to the amount found in the H^O^ impinger catch. Our
studies have not shown this condition to be attributable to artifact formation
or interferences in the analytical method.
Any plausable explanation for changes in the formate and acetate concentra-
tions must take into account the highly reactive nature of aldehydes. An IPA
impinger catch obtained from am actual source test necessarily contains a com-
plex mixture of both organic and inorganic species. It has a relatively long
residence time in its concentrated form prior to dilution and subsequent analy-
sis. Conversely, an IPA impinger catch obtained from a laboratory-generated
sample remains an essentially pure solution, and it has a relatively short
residence time in its concentrated form prior to dilution and subsequent
analysis. Thus, in the absence of a suitable oxidizing agent such as H^O^
and given sufficient time, reactions involving a large percentage of any
39
-------
!I]SG Environmental Sciences Center
FS-TR-an-ru
aldehyde species condensed in the IPA solution will probably occur that de-
crease their original concentration significantly.
As an example of the above, aldehydes are known to be reduced in the
presence of a zinc catalyst in acid solution (Morrison and Boyd 1979). XRF
analysis detected zinc at 20 ppm; IC analysis detected the presence of the
chloride ion, which, as HC1, is sufficient to make the solution acidic. Thus
this reaction would not be unexpected. Assuming this explanation of aldehyde
chemistry to be valid, the calculated collection efficiency of the various
impingers of the CCS train based on the laboratory tests cannot be considered
strictly representative of the efficiency of the system in actual field use.
Concluding Remarks
A complete interference study could not be accomplished within our allotted
time period, but the compounds we considered the most likely candidates to
cause interference were proven to pose none. We do not discount that another
species may be responsible for the formate and acetate peaks, but we do feel
that formaldehyde and acetaldehyde are the most likely candidates, if only from
the standpoint that they are the most likely combustion products exhibiting the
observed properties. Other studies have shown that formaldehyde is a key inter-
mediate in low-temperature combustion sources and is the major emitted aldehyde
(Harrenstien et al. 1979). In fact, formaldehyde may be formed from thermal
decomposition of phenolic polymers, which may be responsible for the observed
chlorophenols.
If another species is responsible for the observed results, then it must
have the following properties: 1) oxidization to formate and/or acetate by
3% H2ฐ2' no aPPreciat>le solubility in IPA/H^C^ or great enough reactivity
to be destroyed before analysis, and 3) no appreciable solubility in basic
solution (no formate or acetate was observed in the NaOH impingers of the HC1
train) or great enough reactivity to be destroyed before analysis. Clearly,
a systematic survey of all compounds exhibiting these properties would be a
major undertaking.
40
-------
jl03 Environmental Sciences Center PVTR-ftO-O^1
As a final point of reference, we analyzed other samples available for
formate and acetate production. These sources included a hogged-fuel-firec
boiler, and a coal-fired power plant. These results along with zhe Hempstead
results are summarized in Table 14 (p. 29). The results both before and after
treatment are given, and in Figure 12 (p. 30) representations of chromatograms of
the Hempstead sample before and after treatment are shown. Since treatment with
NaOH may possibly oxidize other species besides aldehydes, the untreated value
should be considered a lower limit for the formaldehyde content of the flue gas.
Table 14 readily illustrates that the lowest values were obtained for the coal-
fired boiler, and by far the highest values for the hogged-fuel-fired boiler.
Considering the relative efficiency of the combustion processes, the order of
formaldehyde emissions would not appear to be unreasonable.
A means of verifying these results would be comparison of our sampling
apparatus to another type of aldehyde train. Problems with this approach are
that aldehydes are difficult to collect and stabilize, and other aldehyde trains
are known to have numerous interfering compounds (Smith et al. 1972). Unique,
undocumented problems may also exist with the Hempstead-type of emissions
source.
41
-------
j3S3 Environmental Sciences Center ES-TR-80-04
REFERENCES
Black Clawson-Parsons and Whittemore Organization, The. 1974. Proposal for
Town of Hempstead, New York, for Resource Recovery Plant. Project NY-1369,
Vol. I. Prepared for Hempstead Resources Recovery Corporation, Hempstead,
New York. October.
Buser, H. 1978. Polychlorinated Dibenzo-p-dioxins and Dibenzofurans: Forma-
tion, Occurrence and Analysis of Environmentally Hazardous Compounds.
Dept. of Org. Chem., Univ. of Umea, Sweden and Swiss Fed. Res. Station,
Waedenswil, Switzerland. 449 pp.
Cheney, J.L. and J.B. Homolya. 1979. Sampling Parameters for Sulfate Measure-
ment and Characterization. Environ. Sci. and Tech., 13:584-588.
Eiceman, G.A., R.E. Clement and F.W. Karasek. 1979. Anal. Chem., 51:2343-2350.
Fry, H.S. and J.H. Payne. 1930. The Action of Hydrogen Peroxide on Simple
Carbon Compounds. I: Methyl Alcohol, Formaldehyde and Formic Acid. J.
Am. Chem. Soc., 53:1973-1980.
Gribble, G.W. 1974. TCDD - A Deadly Molecule. Chemistry. 747:15-18.
Harrenstien, M.S., K.T. Rhee and R.R. Adt, Jr. 1979. Determination of
Individual Aldehyde Concentrations in the Exhaust of a Spark Ignited
Engine Fueled by Alcohol/Gasoline Blends. Paper No. 790952, SAE
Technical Paper Series, Society of Automotive Engineers, Inc., Warrendale,
PA.
43
-------
idcJu Environmental Sciences Center
Homolya, J.B., H.M. Barnes and C.R. Fortune. 1976. A Characterization of the
Gaseous Sulfur Emissions from Coal- and Oil-Fired Boilers. Fourth National
Conference on Energy and the Environment, Cincinnati, OH. October 4-7.
Jahnke, J.A., J.L. Cheney and C.R. Fortune. 1977. A Research Study of Gaseous
Emissions from a Municiple Incinerator. JAPCA, 27:747-753.
Lappin, G.R. and L.C. Clark. 1951. Colorimetric Method for Determination of
Traces of Carbonyl Compounds. Anal. Chem. 23:541-542.
Miner, C.S. and N.N. Dalton. 1953. Glycerol. Reinhold Publishing Corp., New
York, NY.
Morrison, R.T. and R.N. Boyd. 1973. Organic Chemistry, 3rd ed. Allyn and
Bacon, Boston, MA.
New York Testing Laboratories, Inc. 1979. Results of Particulate Emission Tests
on One Incinerator Stack for Hempstead Resources Recovery Corporation,
Lab No. 79-554441.
Smith, R.G., R.J. Bryan, M. Feldstein, B. Levadie, F.A. Miller, E.R. Stephens
and N.G. White. 1972. Tentative Method of Analysis for Low Molecular
Weight Aliphatic Aldehydes in the Atmosphere. Health Lab. Science,
9:75-78.
44
-------
3D Environmental Sciences Center ES-TR-80-04
APPENDIX A
SAMPLING AND ANALYTICAL PROCEDURES
This section describes the sampling and analytical procedures implemented
in characterizing selected source emissions at the Hempstead Resources Recovery
Corporation Plant in Hempstead, New York, from July 24 to July 26, 1979. Also
presented are the procedures followed during the course of follow-up laboratory
investigations.
HYDROCHLORIC ACID DETERMINATIONS
The method used for chloride measurement was designed to collect gaseous
and particulate chlorides emitted from stationary sources. The method employs
a midget impinger sampling train to collect HC1 in the flue gas via passage
through a series of impingers containing 0.1 N NaOH; the chloride collected
is then analyzed using IC.
Apparatus
The apparatus consisted of a heated quartz-lined probe, midget impinger
sampling train immersed in an ice bath, a pump, and a dry gas meter.
Sampling
The train was prepared by loading 15 ml of 0.1 N NaOH into the first two
midget impingers and 15 ml of 3% hydrogen peroxide *nto t*ie third im-
pinger. The fourth impinger remained empty. A piece of Pyrex glass wool was
inserted into the inlet of the probe to filter out particulate matter. Im-
pinger solutions were allowed to equilibrate in the ice bath and the system
45
-------
jaa Environmental Sciences Center ES-TR-8Q-04
was leak checked. Each sample was pulled for approximately 1 n at a flow rate
of about 1 liter/min. The probe was allowed to cool prior to sample recovery.
Sample Recovery
Recovery of the sampling train entailed collecting three separate frac-
tions: (1) the probe plug wash, (2) the probe wash, and (3) the combined 0.1 N
NaOH impinger catches. Each fraction was recovered and stored in a separate
125 ml Nalgene polypropylene sample bottle.
Sample Analysis
All samples were quantitatively transferred to appropriately-sized volu-
metric flasks, proportionally diluted, and subsequently analyzed using IC. The
IC employed a 3 x 500 mm column in combination with a 3 x 150 mm precolumn.
Samples were analyzed on the 10 umho/cm range using 0.006 M sodium carbonate
(J^CO^) as the eluent. Results were reported in total milligrams per sample.
HYDROCHLORIC ACID - ELECTROSTATIC PRECIPITATOR DETERMINATIONS
The current method for characterizing chloride emissions draws the sample
first through a glass wool filter plug and then through a solution capable of
retaining HC1. Recent studies indicate that significant amounts of HC1 can be
retained by the glass wool filter media. In an attempt to alleviate this in-
herent source of bias, an electrostatic precipitator (ESP) was utilized (Cheney,
personal communication).
An integrated sampling system was developed implementing both an ESP to
remove particulate chlorides, and impingers containing 0.1 N NaOH to remove HC1.
The flow rate required for efficient operation of the ESP necessitated the use
of a larger Greenberg-Smith impinger train.
46
-------
30]
Environmental Sciences Center ES-TR-80-04
Apparatus
The equipment consisted of a heated quartz-lined probe, an electrostatic
precipitator, a sampling module, a sampling pump, and a dry gas meter. The
sampling module contained an ice bath for the Greenburg-Smith impingers, along
with voltage controls to heat the probe and filters.
Samp!ing
Sampling preparations involved loading 200 ml of 0.1 N NaOH into the
first impinger; 200 ml of 3% H2ฐ2 ^"nto t*ie secon<^ impinger; and filling the
last impinger with silica gel. All components were allowed to reach operat-
ing temperature and the impinger train was leak checked.
The flow controller was adjusted to a flow rate of approximately 15 ft^/h
for a sampling time of 2 h. After concluding a run, the probe and ESP were
allowed to cool prior to sample recovery.
Sample Recovery
Recovery of the sampling train required the collection of five different
fractions: (1) the probe wash, (2) the ESP wash, (3) the ESP filter, (4) the
0.1 N NaOH impinger catch, and (5) the 3% impinger catch. Fractions were
recovered and stored in separate polypropylene sample bottles. Cooled fractions
from the ESP were rinsed with three 15-ml portions of distilled deionized water.
The washings were saved and stored in a separate polypropylene sample bottle.
The ESP filter was stored in a plastic petri dish.
Sample Analysis
All samples were quantitatively transferred to appropriately-sized volu-
metric flasks and diluted to the mark with distilled deionized water. Solutions
were analyzed using an IC employing a 3 x 500 mm column combined with a 3 x 150 mm
precolumn. Samples were analyzed on the 10 uho/cm range using 0.006 H Na^CO^
as the eluent. Results were reported in total miligrams per sample.
47
-------
Environmental Sciences Center
ES-TR-80-04
SULFUR OXIDE DETERMINATIONS
The sampling method utilized for the measurement of sulfur oxides was the
Controlled Condensation System (CCS) (Cheney and Homolya 1979), designed to
separate and collect the various sulfur oxides from stationary sources. Using
a Goks^yr-Ross type condensation approach, a heated quartz filter, and a
Greenburg-Smith impinger train, the CCS was designed to collect and separate
sulfur oxides belonging to four main categories: (1) particulate sulfates
and sulfites, (2) sulfuric acid, (3) sulfur trioxide gas, and (4) sulfur dioxide
gas.
Apparatus
The apparatus utilized in the CCS system consisted of a heated quartz-
lined probe, a glass filter holder, a glass plug holder, a sampling module, a
sampling pump, and a dry gas meter (see Figure A-l). The sampling module con-
tained an ice bath for the Greenburg-Smith in^ingers, along with a water bath
for the plug holder, and voltage controls for heating the probe and filter.
Sampling
The system was prepared for sampling by measuring 200 ml of 80% IPA into
the first impinger and 200 ml of 3% H2ฐ2 ^nto t*ie second impinger; the last im-
pinger was filled with silica gel. All components were allowed to reach operat-
ing temperature and the impinger train was leak checked. During sampling all
temperatures were monitored and readjusted as necessary.
The flow controller was adjusted to maintain a flow rate of 10 liter/min
for a sampling time of 0.5 h. Following the conclusion of each run, the im-
pingers were purged for 15 min at a 10 liter/min flow rate. The probe and
filter were allowed to cool before sample recovery.
48
-------
INSULATION
ID
\
HEATING MANTLE
ป. * \
(BE
PLUG HOLDER
""ฆt
FILTER HOLDER
1
PUMPING
METER BOX
SAMPLING MODULE
LJ
ED
L-ZKJ
<
3
a
3
sr
C/>
O
S*
3
o
C0
O
ป
3
U>
Figure A-l. Schematic drawing of manual acid condensation system. A sampling pump and dry gas
meter are contained within the pumping meter box.
ฆ
00
0
1
o
-------
Environmental Sciences Center
ES-TR-80-04
Sampling Recovery
The sampling train was recovered as five fractions in separate bottles:
(1) probe wash, (2) filter, (3) plug wash, (4) 80% IPA impinger catch and (5)
3% impinger catch. Each fraction was recovered and stored in a polypropy-
lene sample bottle. The filter was stored in a plastic petri dish.
Sample Analysis
All samples were quantitatively transferred to volumetric flasks and
diluted to a known volume. The samples were analyzed by IC, employing a
3 x 500 mm column, combined with a 3 x 150 mm precolumn. Samples were analyzed
on the 10 ymho/cm range using 0.006 M Na^CO^ as the eluent. Results were re-
ported in total milligrams per sample.
ALDEHYDE DETERMINATIONS
While analyzing the CCS-1-H202 sample fraction for sulfur oxides by IC,
a large peak was detected at a retention time of ~2.70 min in the .006 M Na^CO^
element system. This retention time does not correspond to those of standards
used for calibration (F , CI , etc.). Investigations were conducted to determine
the identity of this unknown species by using standard solutions of various ions.
The peak was thus tentatively identified as the formate ion (HCOO ).
Verifications of the peak's identity was achieved by analysis of the sample
in two eluent systems using sodium formate reference standards. One system con-
sisted of a 0.0015 M NaHCO^ element at a flow rate of 130 ml/h in conjunction
with a 3 x 150 mm precolumn, a 3 x 500 mm anion separator column, and a 6 x
250 mm suppresor column. Samples were analyzed on the 10 umho/cm range using
an external standard method of analysis. The second system consisted of
0.005 M Na^^O^ element at a flow rate of 156 ml/h in conjunction with a
3 x 150 mm precolumn, a 3 x 500 min anion separator column, and a 6 x 250 mm
50
-------
303 Environmental Sciences Center Liz TR-80-04
suppressor column. Samples were analyzed on the 10 umho/cm range using an
external reference method of analysis.
With the much greater resolution of fast-eluting ions attributable to
using these weak eluents, the unknown peak originally detected as a solitary
peak by the strong eluent (.006 M Na^CO^) was resolved into two peaks. These
peaks were positively identified as formate and acetate. All aldehyde determi-
nations reported were based on analysis of samples collected using the CCS.
LABORATORY ALDEHYDE DETERMINATIONS
This section describes the sampling and analytical procedures implemented
in the laboratory to examine the reliability of using the CCS for aldehyde emis-
sions characterization.
Apparatus
The apparatus utilized in the laboratory consisted of a sampling train
identical to the system utilized in field for the sampling of sulfur oxide.
The system consists of a heated quartz-lined probe, a glass filter holder, a
glass plug holder (condenser), a sampling module, a sampling pump, and a dry
gas meter. The sampling module contains an ice bath for the Greenburg-Smith
impingers, a constant temperature water bath for the glass plug holder, voltage
controls for heating the probe and filter, and a thermocouple meter for monitor-
ing the temperature of the various components of the system.
Sampli ng
In conducting the laboratory studies, care was taken to duplicate as
nearly as possible the operating parameters used in the field when sampling with
the CCS. During sampling the probe and filter were maintained at a temperature
of 520ฐ F, the plug holder at 140ฐF, and the impinger was immersed in an ice
bath. Sampling was conducted for a period of 30 min at a sampling rate of
10 liter/min.
51
-------
. Environmental Sciences Center
ES-TR-80-04
The system was prepared for sampling by measuring 200 ml of 80% IPA into
the first impinger and 200 ml of 3% into the second impinger. For the
blank run, three impingers were used, with the third containing dried silica
gel. This arrangement duplicates the impinger train as utilized in the Hemp-
stead sampling. In other runs conducted, the solutions contained within the
third and fourth impingers depended upon the particular experiment being per-
formed. All components were allowed to reach operating temperature and the
impinger train was leak checked. During sampling all temperatures were moni-
tored and readjusted as necessary.
Following each run, the impingers train was purged for 15 min at a
10 liter/min flow rate. The probe and filter were allowed to cool before
sample recovery.
Formaldehyde gas was generated and sampled by connecting a .5 in Tygon
line to the probe liner on one end with a teflon connector, and to the outlet
of a Greenburg-Smith impinger with a ground glass joint connector on the other
end. The impinger was then changed with 200 ml of a formalin solution in a
hood at ambient temperature. Sampling was conducted by pulling ambient air
through the impinger solution and into the heated probe of the CCS train. The
initial blank run consisted of pulling laboratory air through the Tygon line
and into the heated probe.
Acetaldehyde gas was generated and sampled in a similar manner, except
that the Greenburg-Smith impinger was filled with pure acetaldehyde ("*30 ml)
to a level below the glass center tube in the impinger. The incoming air was
therefore passed over the liquid surface, rather than bubbled through the
liquid.
In selected sample runs, 10-ml aliquots of the IPA and ^mP^n^er
solutions were extracted for separate analysis after the conclusion of the
sample run and prior to the 15 min purge of the impinger train.
52
-------
Environmental Sciences Center
ES-TR-Rn-04
In order to determine the effects of high carbon dioxide concentrations
(~10% CO^ in stock concentration) in air sampled by the CCS train, a cylinder
containing 5% CO^ in air was employed to collect samples using one 80% IPA im-
pinger, followed by one 3% H2^2 ^-mP^n(3er- T^e impingers were immersed in an ice
bath and the cylinder gas was bubbled through the impinger train at a rate of
10 liter/min for 30 min.
Sample Recovery
Each sample run was recovered in four separate fractions as follows:
(1) the probe wash, (2) the filter, (3) the glass-wool plug wash, and (4) the
impinger train (collected separately), each impinger being rinsed with D.I. H20.
Except for storage of the filter in a plastic petri dish, all sample fractions
were placed in polypropylene sample bottles.
Sample Analysis
All samples were diluted to a known volume (inyinger catches to 100 ml/
all other fractions.to 100 ml) with D.I. H^O, and then analyzed for formate and/
or acetate ions using a Dionex Model 10 IC. The IC method of analysis consisted
of a 0.0015 M NaHCO^ eluent in conjunction with a 3 x 500 mm anion separator
column and a 6 x 250 mm anion suppressor column. The samples were analyzed
using the 10 umho/cm range of the instrument with the eluent flow rate adjusted
to 120 ml/h. An external standard method of analysis was employed using stan-
dards prepared from stock standard sodium formate and sodium acetate solutions.
During the course of the study, selected samples were pretreated with
and/or NaOH prior to analysis. Known amounts of standard stock and/
or NaOH solutions were added to a 50-ml aliquot of the sample and mixed
thoroughly, then allowed to stand overnight prior to analysis.
The sample, consisting of 10-ml aliquots of the IPA and impinger ex-
tracted prior to purging, were diluted to 50 ml with D.I. H^O, thereby monitor-
ing the 4:1 dilution ratio used for the individual impinger catches (i.e.,
200 ml were diluted to 1000 ml).
53
-------
jJdQ Environmental Sciences Center F5-TR-80-04
Analysis of the DNPH/HC1 impinger catch (CCS Run #9) was performed gravi-
metrically; analysis of the DNPH/MeOH impingers (CCS Run #11) was performed by
a colorimetric technique (Lappin and Clark 1951).
ORGANIC DETERMINATIONS
The sampling method utilized for the measurement of total organic emissions
from stationary sources was a prototype sampling system. The method was designed
to retain the organic emissions in an Amberlite XAD-2 column. Samples were
analyzed using GC and GC/MS.
Apparatus
The apparatus utilized in this sytem consisted of a heated quartz-lined
probe, an Amberlite XAD-2 column, a Greenberg-Smith impinger train, a pump,
and a dry gas meter.
Sampling
The sampling train was prepared by inserting the Amberlite XAD-2 column
(8 in long x 1/2 in diameter) between the first and second impinger. The first
impinger was left empty to condense out water. The second impinger contained
200 ml of 3% H2ฐ2" silica was use<3 in the last impinger.
The impinger solutions were allowed to equilibrate in 2m ice bath. The
system was leak checked prior to each run. Each run consisted of pulling a
gas sample through the heated probe, which contains a glass-wool probe plug,
and through the impingers and Amberlite XAD-2 column. The sample was pulled
at a rate of approximately 25 ft^/h for 1 h.
Sampling Recovery
Recovery of the sampling train entailed the collection of two separate
fractions (1) the first impinger catch, (2) the Amberlite XAD-2 column catch.
54
-------
ฆm
Environmental Sciences Center
fs.tr.an-na
The contents of the first impinger were transferred to a one-liter glass sam-
ple bottle. The impinger was rinsed several times with distilled-deionized
water. The washes were stored in the glass sample bottle along with the ori-
ginal impinger catch. The ends of the Amberlite XAD-2 column were sealed with
glass ball and socket joints.
Sample Preparation and Analysis
All samples were extracted using methylene chloride. The extracts were
quantitatively transferred to 100-ml volumetric flasks and diluted to the
mark using methylene chloride. The sample fractions were concentrated down
to approximately 1 ml. The samples were transferred to individual, clean,
tared 5-ml glass vials with teflon liners. The sample fractions were completely
dried under a nitrogen environment. The vials were reweighed and the mass
of organic residue in each vial was recorded for the calculation of mass emis-
sion rates.
Samples were then diluted, using methylene chloride, to a concentration of
15 to 20 ug of organic residue per microliter of methylene chloride. Sample
fractions were analyzed by a Hewlett-Packard Model 5985A GC/MS under the follow-
ing operating parameters:
Maximum
Cut-off Temp Desired Actual
Temperature, initial ฐC
Temperature, final ฐC
Injection port temperature, ฐC
FID temperature, 0 C
TCD temperature, :ฐC
AUX temperature, ฐC
350
350
400
400
300
300
50
275
275
300
275
275
51
275
275
300
275
274
Rate, ฐC/min
Chart speed
Zero
Attenuation, 2T
FID Signal
Slope sensitivity
Area reject
Flow A, mil/min
Flow B, ml/min
6.00
1.00
10.0
4
+B
0.10
1000
8.1
41.2
55
-------
Environmental Sciences Center
ES-TR-80-04
Purification of XAD-2/Resin
The Amberlite XAD-2 resin was purified prior to sampling using the follow-
ing procedure:
(1) The resin was washed in a 2-liter roundbottom flask with two 1-
liter portions of distilled-deionized water. Care was taken to
ensure that the mixture was well shaken.
(2) Next, the resin was refluxed with 1 liter of reagent-grade
menthanol for 7 h.
(3) After removing the methanol, 1 liter of "glass distilled" methy-
lene chloride was added to the mixture, which was then allowed
to sit for 17 h. The mixture was refluxed for 7 h and the methy-
lene chloride decanted.
(4) Step 3 was repeated.
(5) Finally, the resin was dried in the oven in a glass container at
80ฐ C for 1 h.
(6) The resin was stored in a sealed, all-glass container until ready
for use.
56
-------
CJSS Environmental Sciences Center ES-TR-80-04
APPENDIX B
CALCULATIONS
Standard Cubic Feet (SCF) = K x a x V. x /T. (Eq. B-l)
dgm bar dgm
where K = 17.636 ฐR/irt Hg,
a = dry gas meter calibration factor,
V, = dry gas meter volume,
agin
P, = barometric Pressure, in Hg, and
oar
T, = dry gas meter temperature, ฐR.
dgm
K, x mg
1 (Eq. B-2)
Parts Per Million (PPM) = x scp
where K. = 849.52 PPM x 9 x ft
1 mg
mg = number of milligrams of pollutant,
MW = molecular weight of pollutant, and
SCF = standard cubic feet of sample.
Mass emissions rate (lb/h) = K2 x mg x VFR/SCF (Eq. B-3)
where = 2.205 x 10 6 lb/mg
mg = number of milligrams of pollutant,
VFR = volumetric flow rate of stack, 4,4 x 106 SCF/h
SCF = standard cubic feet.
The percent operating capacity was calculated based on a steam generation
rate of 200,000 lb/h, equivalent to a 100% operating capacity.
57
-------
Environmental Sciences Center
rs-TR-an-n4
A volumetric flow rate (VFR) for the stack had to be assumed in calculating
the mass emission rates. The value used, 4.4 x 106 SCF/h, was the average value
calculated from three tests (Method 5) performed on Furnace #1 by New York Test-
ing Laboratories, Inc. (1979), from April 30 to May 1, 1979 while the unit was
operating at 96% capacity. The results in this report are based on samples
collected from Furnace #2.
The mass emission rates for sulfur oxides are based on the analysis of
sulfur oxides collected in the HC1 sampling train.
REFERENCES
Cheney, J.L. and J.B. Homolya. 1979. Sampling Parameters for Sulfate Measure-
ment and Characterization. Environ. Sci. and Tech., 13:584-588.
Lappin, G.R. and L.C. Clark. 1951. Colorimetric Method for Determination of
Traces of Carbonyl Compounds. Anal. Chem. 23:541-542.
New York Testing Laboratories, Inc. 1979. Results of Particulate Emission
Tests on One Incinerator Stack for Hempstead Resources Recovery Corpora-
tion, Lab No. 79-554441.
58
------- |