vc,EPA
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/PS7-80-149 Sept 1980
Project Summary
Environmental
Assessment of a Waste-to-
Energy Process: Braintree
Municipal Incinerator
Mark A. Golembiewski, K. P. Ananth, T. Sutikno, and Harry M. Freeman
Field tests were conducted at the
Braintree Municipal Incinerator in
Braintree, Massachusetts for the pur-
pose of providing data on multi-media
emissions to help define potential en-
vironmental impacts and pollution
control technology needs. The
waterwall incinerator tested was fired
with unprocessed refuse at a rate of
about 5 tons/hr. Air pollution control
was provided by an ESP. Primary em-
phasis was placed on evaluating air
emissions, including criteria
pollutants, as well as hazardous trace
metals and organic compounds.
Trace elements were found to be par-
ticularly concentrated in the in-
cinerator bottom ash. Levels of major
quality parameters in effluent from
the bottom ash quench did not appear
to be of concern. Analysis of fly ash
collected by the ESP showed the
presence of chlorides, sulfates, and
certain trace elements. Stack emis-
sions of NOX (54 ppm), SO2 (48 ppm).
hydrocarbons (11 ppm), and chlorides
K120 mg/Nm3) were low. Paniculate
emissions average 0.24 gr/dscf, cor-
rected to 12% CO2 which was higher
than expected. However, the high
emissions were subsequently found
to be related to deficiencies in plant
operation. Multi-media emissions
were evaluated using EPA's SAM-1A
protocol.
Introduction
The EPA's Industrial Environmental
Research Laboratory in Cincinnati is
presently supporting a large-scale
research program to conduct an en-
vironmental assessment of various waste-
to-energy conversion systems. As part of
this program, field tests were carried out
by Midwest Research Institute (MRI) at
the Braintree Municipal Incinerator in
Braintree, Massachusetts in January,
1978. The sampling study was primarily
designed to provide information on un-
controlled and controlled air emisssions
so that control technology needs could be
identified. A secondary objective was to
conduct multi-media sampling to obtain
data for an overall environmental assess-
ment of the incineration process, in-
cluding air, water, and solid waste ef-
fluent streams.
This paper presents a description of the
incinerator facility, a summary of the
sampling and analytical methods used, a
discussion of the test results and conclu-
sions of the study.
Description of the Facility
The Braintree Municipal Incinerator is a
mass-burn facility, firing municipal refuse
which is collected from the town of
Braintree and surrounding communities.
The plant, which was constructed in
-------
1971, consists of twin waterwall combus-
tion units, each with a design capacity of
120 tons of refuse per 24-hr period. A por-
tion of the steam produced (20% to 35%)
is supplied to neighboring small manufac-
turers and the remainder is condensed.
Each furnace is equipped with an elec-
trostatic precipitator for emission control
and both ESP's exhaust to a common
stack.
The "as received" refuse is fed to a
charging chute for each furnace by an
overhead crane. The material is first dried
on an inclined stoker grate before being
depositied on the horizontal, travelling
grate. Underfire air is supplied to the com-
bustion bed. Unburned and noncombusti-
ble materials at the discharge end of the
grate are quenched with water before
removal to a landfill.
The Riley Stoker boilers are of a single
pass design and have about 895 ft2 of
waterwall heating surface prior to the
boiler tub section. Total heating surface
area is 3,305ft2. Rated capacity of each
furnace is 30,000 Ib of steam per hour at
400°F and 250 psig.
The electrostatic precipitator is a single
field, 12 passage unit, with a specific col-
lection area of 125 ft2/1000 acfm. It was
designed for a collection efficiency of
93%. A bypass duct connects the inlet
ducts to the two precipitators so that the
exhaust gases from one boiler can be
diverted through both collectors in a
parallel arrangement or through the op-
posite precipitator.
Sampling and Analysis
Methodology
The environmental assessment pro-
gram carried out at the Braintree incin-
erator consisted primarily of the determi-
nation of flue gas constituents. The
sampling efforts also included
measurements of the refuse feed and the
solid and liquid effluent streams.
All tests were conducted while the in-
cinerator was operated at its design refuse
feed rate at 4.5 mg/hr (5 tons/hr). Actual
feed rates during the 3 days of testing
ranged form 4.1 to 4.7 mg/hr. Testing
was carried out on Boiler No. 1 with the
flue gases diverted through the bypass ar-
rangement to the No. 2 ESP. This was
done to provide more suitable sampling
conditions (smoother velocity profile) at
the inlet to the collector.
The multi-media sampling matrix for
this program is illustrated in Figure 1. The
general sampling and analysis schemes
for each feed/effluent stream are briefly
described next.
Refuse Feed— Once each test day, a 90
Kg(200 Ib) sample of raw refuse was taken
and manually sorted into its metal, glass,
and combustible components. After
each fraction was weighed, a 1 ft3
sample of combustible material was
retained for analysis. Each refuse sample
was analyzed for moisture content and
higher heating value. Also, proximate
and ultimate analyses were carried out.
Elemental concentrations in the samples
were determined using Spark Source
Mass Spectrometry (SSMS).
Bottom Ash—A. grab sample of bottom
ash was collected for each hour of
testing. At the end of the day, the
samples were composited and then
segregated into metal, glass, and com-
bustible fractions. A 1 ft3 portion of the
combustible material was retained for
analysis. Bottom ash analysis consisted of
moisture determination and elemental
analysis by SSMS.
Quench Water— Grab samples of ef-
fluent from the bottom ash quench
trough were obtained each day and
analyzed for BOD, COD, TSS, pH,
phenols, and oil and grease.
ESP: Inlet— Uncontrolled paniculate
emissions were measured using EPA
Method 5. One sampling run was con-
ducted each day. Sample analysis in-
cluded determination of the filterable par-
ticulate catch and also the organic and
inorganic components of the condensible
paniculate. Paniculate collected on the
filters was analyzed for general elemental
composition by SSMS and for seven
specific metals (Fe, As, Sb, Hg, Pb, Cu,
and Cd) by atomic absorption spec-
trometry (AAS).
EPA Ash—Samples of fly ash were
collected hourly from the ESP hoppers
and composited at the end of each test
day. A 1 ft3 sample was then extracted
from each composite for analysis, which
included elemental composition by
SSMS, analysis for specific anion (Ce~,
F", Br", S04=, N03~, and CM"), and deter-
minations for polychlorinated biphenyl
(PCS) and polycyclic aromatic hydro-
carbon (PAH compounds)
Stack Emissions—Mr emissions of the
ESP outlet were sampled and analyzed for
a variety of paniculate and gaseous con-
stituents.
Paniculate sampling at the ESP outlet
was carried out simultaneously with the
sampling runs at the precipitator inlet,
again using EPA Method 5. The resulting
filter samples were analyzed for elemental
composition by SSMS and AAS.
An Andersen cascade impactor v
used to measure particle size distributi
in the outlet gases. Two sizing runs wi
conducted each test day.
Gaseous composition of the flue gas
included Orsat analysis for 02 and C
and continuous monitoring for 02, N(
SO2, CO, and total hydrocarbons.
An absorption sampling train, based
guidelines presented in the Los Ange,
APCD Source Testing Manual, was us
to obtain two daily samples which we
analyzed for carbonyl materials. Anotr
absorption sampling train was used
determine levels of Cl, F, Br, and C
anions, as well as vaporous mercury.
Finally, the Source Assessment Sai
pling System (SASS) developed by EF
was used to provide one sample for tl
EPA Level 1 analytical matrix.
Discussion of Test Results
This section presents the results of tl
sampling and analysis efforts. An evalu
tion of the emissions using the Soun
Analysis Model (SAM) is also summ
rized.
Refuse Feed
Data obtained from analysis of tf
refuse samples are presented in the con
plete project report. The refuse sampk
had an average heat of combustion i
16,780 kJ/kg (7,214 Btu/lb). Correctior
for a moisture content and percentage <
glass, metal, and inerts showed that tr
"as received" solid waste had a heatin
value of 10,200 kJ/kg. The combustibl
fraction of the refuse contained 0.33*3
sulfur and 4.6% ash, based on results c
the proximate/ultimate analysis.
Bottom Ash
Spark Source Mass Spectrometr
(SSMS) analysis showed an increase i
the concentration of nearly every elemer
in the bottom ash relative to the refus
feed. Even some of the more volatil
elements, which would be expected ti
vaporize and be carried off with the flu
gases (e.g., As, Sb, Pb), were found ii
greater concentrations in the bottom ash
For example, a ten-fold increase in PI
was noted after the mass ratio of bottorr
ash to input refuse was determined
However, it should be noted that the
hand sorting technique used to separate
the combustible materials from the raw
refuse could have allowed lead
containing metals to be incompletely
separated from the paper and plastics.
-------
ESP INLET
STACK EMISSIONS
REFUSE
Sampling-
Collect one 200 Ib sample each
day. Hand sort into metal, glass
and "other" Weigh each
Extract 1 ft3 of "other fraction"
Analysis
Determine H20 content of
"other" fraction. Proximate/
ultimate heating value.
SSMS
Samp/ing and Analysis
a Method 5 particulate - 1 per
day. Determine paniculate
loading, front and back half
/organic extraction and
impmger evaporation) Also
analyze for trace metals (Fe,
As, Sb. Hg, Pb, Cu, Cd) by A A
and elemental composition
by SSMS.
b. ORSA T (02 and CO2)
c. Particulate sizing (Brink) -
2 per day
d. ESP operation monitored
by SRI
Stack
Emissions
Refuse input-
Quench water
Waterwall
Incinerator
QUENCH WA TER
Bottom ash
BOTTOM ASH
Samp/ing and Analysis.
a Method 5 particulate - 1 per
day Determine particulate
loading, front and back half
(organic extraction and
impmger evaporation) Also
analyze for trace metals (Fe,
As, Sb, Hg, Pb, Cu, Cd) and
elemental composition by
SSMS
b. ORSATfOz and C02)
c Particle sizing (Andersen) -
2 per day
d Continuous analyzers (02,
NO*. SO*, HC, CO)
e Carbonyls - 2 per day
f. Mercury train - 2 per day
Also analyze KOH impmger
for Cl, F, Br and CN
g SASS train - 1 test Analyze
all portions per Level 7
requirements Additional
analysis for PCB and PAH
materials
Sampling.
Collect 1 -Liter grab sample
each day
Analysis.
BOD, COD, TSS, TDS, pH.
phenol, oil and grease •
Sampling:
Collect 1 ft3 sample each hour.
Composite and hand sort
into metal, glass and "other".
Weigh and extract 1 ft3 of
"other fraction"
Analysis:
Determine H2O content
SSMS
ESP ASH
Sampling:
Take 1 -Liter grab sample each
hour. Mix and extract 1 -Liter
composite
Analysis-
Anions (Cl, F, Br, CN, S04,
N03)
PCB and PAH
SSMS
Figure 1. Sampling and analysis matrix for the Bramtree Municipal Incinerator.
Quench Water
In general, the water quality analysis
results revealed moderately low concen-
trations of BOD, COD, oil and grease, and
TSS. TDS averaged 710 mg/liter at the
outlet. Levels of phenol were all <0.1
mg/liter.
ESP Inlet
Uncontrolled particulate emissions
averaged 0.82 g/dscm (0.36 gr/dscf), or
0.60 g/MJ heat input (1.4 lb/108 Btu). In
terms of an uncontrolled emission factor,
about 11.3 Ib of particulate were dis-
charged for every ton of refuse charged to
the incinerator.
Analysis of the particulate catch for
metals by atomic absorption (AA) showed
relatively high levels of Pb in the uncon-
trolled particulate. Pb concentrations
averaged 11.9 mg/dscm. Concentrations
of Hg and As were low, (10 and 50
ng/dscm, respectively), while levels of
Sb, Cd, Cu, and Fe were in the range of
100 to 1,000fig/dscm.
In addition to these values, elemental
analysis of the particulate samples by
SSMS indicated average concentrations
of Bi, Sn, Br, Zn, Ca, Ti, K, P, Si, Al, Na,
and Mg which were greater that 1,000
/jg/g (about 500 ^g/dscm). The data also
showed levels of Cl, F, and Br in this same
range.
The concentrations of 02 and C02 in
the flue gas at the ESP inlet were 16.8%
and 4.8% respectively.
ESP Ash
Grab samples of the fly ash collected in
the ESP hoppers were taken during each
test period and analyzed for anions,
polychlorinated biphenyl (PCB) and
polycyclic aromatic hydrocarbon (PAH)
compounds, and elemental composition.
Concentrations of chlorides and
sulfates averaged 41 and 10.4 mg/g,
respectively. Levels of cyanides and
nitrates were below the limit of detection
while F~ and Br~, concentrations were
about 0.5 mg/g.
No PCBs were detected in the ESP ash.
Concentrations were all .below the detec-
tion limit of 0.4 mg/g. Four PAH com-
pounds were identified in the ash
samples, but their levels were below the
-------
range of reliable quantitative analysis (0.3
to 0.7 ^g/g).
Several elements in the SSMS elemen-
tal composition data had concentrations
in excess of 1%. They included Ca, K, Cl,
P, Si, Al, Mg, and Na. The concentrations
of many elements increased in the col-
lected fly ash relative to the uncontrolled
particulate. This would seem to indicate
that these elements were predominantly
associated with larger sized particles.
Conversely, there were also elements
whose concentrations showed a definite
decrease in the collected ash, namely Bi,
Sb, Sn, Cd, Ag, Br, Se, As, Ga, Cu, and
F. It is likely that these elements may be
associated with smaller particles.
Stack Emissions
Gaseous analysis data of the stack ef-
fluent is shown in the complete report on
this project. This data shows low concen-
trations of hydrocarbons (11.3 ppm), NOX
(54 ppm), and S02 (48 ppm), but high
levels of carbon monoxide. CO levels
ranged as high as 750 ppm and averaged
475 ppm. Emission factors for the criteria
pollutants were calculated in terms of
pounds per ton of refuse charged. These
were as follows: particulate — 3.1
Ib/ton, NOx — 1.6 Ib/ton, SO2 — 2.0
Ib/ton, CO — 8 5 Ib/ton; and hydrocar-
bons — 0.12 Ib/ton. Two other pollutants
of interest, chlorides and lead, had
emission factors of 1.92 Ib/ton and
0.086 Ib/ton, respectively
Filterable particulate emissions, which
are tabulated in the complete project
report, averaged 0.55 g/dscm corrected
to 12% C02, which exceeded the ap-
plicable state and federal emission regula-
tions of 0.23 and 0.18 g/dscm, respec-
tively. The total uncorrected concentra-
tion was 0.22 g/dscm (0.094 gr/scf) of
which 11% was condensible particulate.
The high particulate loadings can be
traced to an apparent low collection effi-
ciency of the ESP. ESP efficiency during
MRI testing was determined to be only
-74%, while its design efficiency was
reported to be 93%. Monitoring of the
ESP during sampling did not indicate any
abnormalities in it operation. However,
plant personnel claimed subsequently that
the high particulate levels were most likely
the result of mechanical problems in the
operation of the incinerator. A new series
of compliance tests have been conducted
since the MRI tests and the emission
levels are reported to be within the com-
pliance regulations.
Andersen impactor particle size analysis
showed an average particle size of about
S^m at the ESP outlet. Approximately
22% of the particles were smaller than
O.Gpim and about 23% were 10 ^m or
larger, indicating a fairly uniform distribu-
tion of particle sizes in the stack emis-
sions.
Trace metal concentrations in the outlet
particulate showed basically the same
relative distribution as was seen in the in-
let particulate but at generally reduced
levels. Pb, at 5,400 ^m/dscm, had the
highest concentration while As and Hg
had the lowest (16 and 14 ^m/dscm,
respectively). In terms of ppm (/^g/g), the
concentrations of Pb, As, and Hg were
47,600, 145, and 124, respectively.
The concentration of mercury in the
particulate at the outlet was higher than
the concentrations of the inlet. At the
outlet, mercury concentration averaged
0.014 mg/dscm.
SSMS results showed that the concen-
trations of most elements decreased
slightly or remained essentially the same
in the stack emissions when compared to
the uncontrolled emissions. Some
elements increased in concentration and
the most prevalent elements were Pb, Zn,
Si, and Al.
Results of the sampling for carbonyl
compounds showed very low concentra-
tions, averaging 1.3 ppm.
Another absorbent train was used to
sample for vaporous mercury and
selected anions (Cl~, F', Br~, and CI\T).
Vaporous Hg concentrations varied
widely from 30 to 1,800 ng/m3. The
average was 560 ng/m3. The wide varia-
tion in Hg levels could have been the
result of fluctuations in mercury-bearing
materials in the refuse. Results of the
anion analysis showed all concentrations
at or near the limits of detection. The
highest chloride concentration was 190
ng/m3 (130 ppm).
Stack emissions sampling conducted
using SASS equipment indicated that 43%
of the particulate was 1 fjrn or smaller
in size. PCB analysis of the SASS organic
module revealed observable levels only in
the XAD-2 resin absorbent. About 100/ug
of PCB (reported as DCB) were detected,
which corresponds to a concentration of 4
ng/m3. The PCB concentration in the par-
ticulate was below the detectability of the
GC/MS analysis technique used. Four
PAH compounds were quantitatively
identified in the XAD-2 resin and aqueous
condensate. All levels, however, were
quite low. Highest observed concentra-
tions was 4^g/m3.
Level 1 analysis for the vaporous metals
As, Sb, and Hg, showed the highest con-
centrations in the particulate sample frac
tions, indicating possible adsorption c
these metals on the particulate. Sb show
ed the greatest increase in concentration
in the finer particle size.
Organic analysis of the SASS trail
components was carried out in accord
ance with EPA Level 1 protocol. Thesi
data were interpreted using EPA's Soura
Analysis Model (SAM-1A) which i;
discussed below.
SAM-1A
Because of the difficulty involved in in
terpretating the Level 1 analysis results
the environmental assessment work was
extended to include application of the
methodology known as SAM-1A, re
cently developed by EPA.
This methodology was applied to the
Braintree data for the four efflueni
streams: bottom ash, quench water, ESF
ash, and stack emissions. Results, shown
in Table 1, indicated that the stack emis-
sions had the highest degree of hazard (H
value), but this was primarily due to
groups of organic compounds which
could not be individually identified and
therefore were assigned a conservatively
low Minimum Acute Toxicity Effluent
(MATE) value in accordance with the
SAM-1A methodology. The bottom ash
stream had the highest Toxic Unit
Discharge Rate (TUDR), due to several
metals present in high concentrations.
This finding would seem to indicate that
the bottom ash stream should receive the
highest priority for control or removal of
specific metallic constituents. However,
considering the physical nature of the bot-
tom ash and the methods for its disposal,
further work should be done to better
assess its environmental hazard potential.
Conclusions
Based upon the data obtained from this
study, the following conclusions sum-
marize the environmental assessment of
the Braintree Municipal Incinerator. They
are presented in general order of the plant
operation for each effluent stream.
• Elemental analysis of the glass-
qnd metal-free bottom ash reveal-
ed an overall increase in the
elemental concentrations when
compared to the refuse feed.
• Levels of BOD, COD, oil and
grease, TSS, and TDS in the bot-
tom ash quench water do not ap-
pear to be of concern. The
phenolic content was found to be
<0.1 mg/liter in all samples.
-------
Table 1. Summary of Results from SAM-1A Methodology
Health based Ecological based
Degree of hazard
bottom ash
luench water
FSP ash
clue gas
Toxic unit discharge rate
Bottom (g/sec)
Quench water (liter/ sec)
FSP ash (g/sec)
clue gas (m3/sec)
3,800
<20
1,300
14,000
1,700,000
<76
6,900
137,000
1,740,000
<0.2
200,000
110
780,000,000
<0.76
1, 100,000
1,100
The collected fly ash contained
levels of chlorides, sulfates, and
several trace elements which may
be of concern. However, the
leachability of the ash should be
investigated to determine a more
realistic hazard potential.
PCBs were not detected in the
ash recovered by the ESP and on-
ly very minimal levels of four PAH
compounds were identified.
Stack emissions of hydrocarbons,
NOX, and S02 were low.
However, CO levels were high.
This could not be explained con-
sidering the large quantities of ex-
cess air used in the Braintree in-
cinerator.
Paniculate concentrations ex-
ceeded both federal and State of
Massachusetts regulations
governing municipal incinerators.
Average paniculate concentration
was 0.55 g/dscm (0.24 gr/dscf),
corrected to 12% C02. The ESP
removed only about 74% of the
incoming paniculate, while its
design efficiency was reported to
be 93%. Subsequent tests con-
ducted for compliance purposes
reported an outlet loading of only
0.074 g/dscf.
Levels of gaseous chlorides and
other halides were low.
The presence of PCBs was con-
firmed only in the SASS train
XAD-2 resin, which yielded a con-
centration of 3.6 ng/m3.
Results of the SAM-1 A en-
vironmental assessment pro-
cedure show the incinerator stack
emission to have the highest ap-
parent degree of health hazard.
Further analysis, however, is
needed to determine the exact
composition of the organic com-
ponents of the stack emissions to
ascertain its true hazard potential.
SAM-1A also showed that the
bottom ash effluent had the
largest toxic unit discharge rate
due primarily to the abundance of
phosphorus and metals contained
in this stream.
Mark A. Golembiewski, K, P Ananth, and T. Sutiknoare with Midwest Research
Institute, Kansas City, MO 64110.
Harry M. Freeman is the EPA Project Officer (see below)
The complete report, entitled "Environmental Assessment of a Waste-to-Energy
Process, Braintree Municipal Incinerator." (Order No. PB 80-219421,
Cosf $14.00, subject to change) will be available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U S. Environmental Protection Agency
Cincinnati, OH 45268
i US GOVERNMENT PRINTING OFFICE 1981 -757-064/0230
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use S300
------- |