EMB REPORT NO. 74-SSI-3
C3
o-
AIR POLLUTION
EMISSION TEST
PISCATAWAY SEWAGE .;SLUDGE INCINERATOR
ACCOKEEK, MD.
SEPTEMBER, 1975
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Emission Measurement Branch
Research Triangle Park, North Carolina
-------
SEWAGE SLUDGE INCINERATOR EMISSION TESTS
PISCATAWAY SEWAGE SLUDGE INCINERATOR
ACCOKEEK, MARYLAND
A REPORT PREPARED FOR
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK
NORTH CAROLINA 27711
EPA REPORT NO. 7/»~SS1-3
CONTRACT'NO. 68-02-0239
TASK NO. 8
PREPARED BY:
COTTRELL ENVIRONMENTAL SCIENCES INC.
A SUBSIDIARY OF RESEARCH-COTTRELL INC.
P.O. BOX 750
BOUND BROOK, NEW JERSEY 08805
-------
TABLE OF CONTENTS
PAGE
INTRODUCTION 1
SUMMARY AND DISCUSSION OF RESULTS 2
I. STACK EMISSIONS
A. PARTICULATE
B. LEAD
C. MERCURY
D. BERYLLIUM
E. VISIBLE EMISSIONS
II. PROCESS SAMPLES 8
III. MATERIAL BALANCES FOR MERCURY AND LEAD 10
PROCESS DESCRIPTION AND OPERATION 12
LOCATION OF SAMPLING PORTS 14
SAMPLING AND ANALYTICAL PROCEDURES 16
LIST OF FIGURES
FIGURE I.: SEWAGE SLUDGE INCINERATOR PROCESS AT 13
ACCOKEEK, MD.
FIGURE II: STACK LAYOUT AND TEST PORT LOCATION 15
FIGURE III: LOCATION OF SAMPLING POINTS 17
LIST OF TABLES
TABLE I: PARTICULATE SUMMARY 3
TABLE II: LEAD SUMMARY 4
TABLE III: MERCURY SUMMARY 5
TABLE IV: BERYLLIUM SUMMARY 6
TABLE V: VISIBLE EMISSIONS 7
TABLE VI: PROCESS SAMPLE ANALYSIS SUMMARY 9
TABLE VII: MERCURY AND LEAD MASS BALANCE DATA 11
-------
APPENDICES
APPENDIX A: COMPLETE RESULTS AND SAMPLE CALCULATIONS
A.I. PARTICULATE
A.2. LEAD
A.3. MERCURY
A.4. BERYLLIUM
APPENDIX B: FIELD DATA
APPENDIX C: VISIBLE EMISSIONS DATA
APPENDIX D: PROCESS DATA SHEETS AND LOG
APPENDIX E: LABORATORY REPORT
E.I. SAMPLE IDENTIFICATION
E.2. LABORATORY ANALYSES
E.3. OES ANALYSIS OF SLUDGE SAMPLES
APPENDIX F: TEST LOG
APPENDIX G. PROJECT PARTICIPANTS
-------
-1-
INTRODUCTION
Under Section 112 of the Clean Air Act/ the Environmental
Protection Agency is required to develop background infor-
mation and set National Emission Standards for Hazardous
Air Pollutants where applicable.
In order to determine the emissions from a representative
sewage sludge incinerator process, emission testing was
performed on one of the three (3) parallel incinerator
units at the Piscataway Sewage Sludge Facility operated by
the Washington Suburban Sanitary Commission at Accokeek,
Maryland.
The incinerators at this facility are fluid-bed type, with
venturi scrubbers for air-pollution control.
Three (3) separate test runs were made on the exhaust stack
servicing the west most incinerator at the facility. During
each run, samples were collected by Cottrell Environmental
Sciences personnel to determine particulate, mercury, and
beryllium emissions. The particulate samples were analyzed
for total mass and lead content. Process samples were col-
lected during each run in order to obtain data to complete
a material balance for mercury and lead for the system.
During each run, Walden Research Corporation personnel re-
corded visible emission data in order to compare opacity
results with particulate mass emissions.
Testing was performed during February 20-21, 1974.
-------
-2-
SUMMARY AND DISCUSSION OF RESULTS
I. STACK EMISSIONS
A. Particulate
The results of the three (3) particulate tests are summarized
in Table I. The reported particulate weights are corrected
for a filter blank only. Acetone blanks and water blanks are
not included.
The mass concentration measurements, based on dry filterable
alone, ranged between 6.178 mg/Nm3 and 9.84 mg/Nm3 with an
average value of 7.628 mg/Nm3. Total particulate catch
averaged 18.46 mg/Nm3 with a range of 10.298 mg/Nm3 to 34.096
mg/Nm*. There is no known reason for the discrepancy between
Run 1 and Runs 2 and 3. There were no recorded process up-
sets or process changes, during the tests.
B. Lead
Table II contains the results of the lead analysis. The
lead content was determined by analysis of the particulate
residue after mass analysis. Sample blanks were calculated
based on estimated reagent and water volumes.
The lead emissions, based on total catch, varied between
0.754 and 1.08 grams per day with a 0.951 gram per day
average. The concentration, based on total catch, varied
between 3.58 and 5.19 ugm/Nm3. The average concentration
was 4.53 jagm/Nm3.
C. Mercury
Mercury emissions averaged 44.94 grams per day with a range
of 43.09 to 48.61 grams per day (Table III). Mercury con-
centration ranged from 194.6 to 229.4 jag/Nm3 with an average
concentration of 207.1 jug/Nm3. At no time during the tests
were any problems encountered with either the sampling train
or process operations.
D. Beryllium
Table IV summarizes the beryllium test results. A computation
to estimate a minimum detectable level produced a value of
0.21 grams per day. Each of the three (3) samples obtained
was below detectable limits. Therefore, it cannot be ex-
plicitly stated that any beryllium was actually present.
E. Visible Emissions
The average opacity for Tests #1, 2 and 3 was 0.27, 0.00,
and 0.38 percent respectively (Table V). Except for short
intervals during Tests #1 and 3, the opacity remained zero.
-------
TABLE I
PARTICULATE SUMMARY
ENGLISH AND METRIC UNITS
^fcTN NUMBER
V
TEST NUMBER
™TE
VOLUME OF GAS SAMPLED - DSCF(a)
Nm3 (b)
PERCENT MOISTURE BY VOLUME
AVERAGE STACK TEMPERATURE - °F
°C
STACK VOLUMETRIC FLOW RATE - DSCFM(c)
Nm3/min(d)
STACK VOLUMETRIC FLOW RATE - ACFM(e)
m3/min(f )
PERCENT ISOKINETIC
PERCENT EXCESS AIR
1
7
2-20-74
89.034
2.521
9.42
115
46
5189.3
146.94
6328.8
179.2
90.18
40.85
2
8
2-20-74
93.131
2.637
10.95
115
46
5108.0
144.64
6337.2
179.4
96.25
34.95
3
9
2-21-74
96.71
2.738
9.96
115
46
5162.0
146.17
6332.8
179.3
98.78
. 39.95'
EMISSIONS, PROBE, CYCLONE, FILTER CATCH
gr/DSCF
mg/Nm3
gr/ACF
mg/m3
Ib/hr
Kg/hr
Ib/ton DRY SLUDGE
24.8
.0043
9.84
.0035
8.009
0.191
0.087
0.235
18
.0030
6.865
.0024
'5.'492
0.131
0.059
0.161
17
.0027
•6.178
.0022
5.034
0.119
0.054
0.147
EMISSIONS, TOTAL CATCH
mg
gr/DSCF
mg/Nm3
gr/ACF
mg/m3
Ib/hr
Kg/hr
Ib/ton DRY SLUDGE
86.2
0.0149
34.096
0.0122
27.918
0.663'
0.301
0.817
(a) Dry standard cubic feet at 70° F, 29.92 in. Hg.
{b) Dry normal cubic meters at 21.1° C, 760 mm Hg.
(c) Dry standard cubic feet per minute at 70° F, 29.92
(d) Dry normal cubic meters per minute at 21.1° C, 760
(e) Actual cubic feet per minute
£g^ Actual cubic meters per minute
29.3
0.0048
10.984
.0039
8.925
0.210
0.095
0.259
in. Hg.
mm Hg.
28
0.0045
10.298
.0037
8.467
0.199
0.090
0.245
-------
TABLE II
LEAD SUMMARY
(PARTICULATE SAMPLES)
-4-
AlN NUMBER 123
DATE 2-20-74 2-20-74 2-20-74
LEAD EMISSIONS, PROBE, CYCLONE, FILTER CATCH
jag* 3.78 9.90 6.38
ugm/Nm3, Dry 1.50 3.75 2.33
grams/day 0.317 0.782 0.490
LEAD EMISSION, TOTAL
;ig* 12.16 13.68 9.81
pgm/Nm3, Dry 4.82 5.19 3.58
grams/day 1.02 1.08 0.754
* Blanks based on estimated reagent and
wash volumes have been subtracted.
-------
TABLE III
MERCURY SUMMARY
ENGLISH AND METRIC UNITS
^N NUMBER
TEST NUMBER
DATE
VOLUME OF GAS SAMPLED - CU. FT.
@ STACK CONDITIONS CU. M.
AVERAGE STACK TEMPERATURE - °F
UC
PERCENT MOISTURE
STACK VOLUMETRIC - DSCFM(a)
FLOW RATE NmVmin(b)
STACK VOLUMETRIC - ACFM(c)
FLOW RATE mVmin(d)
TOTAL WEIGHT OF HG. COLLECTED -/«3.
CONCENTRATION - ,«g/Nm3
EMISSIONS - ^rams/day
RRCENT ISOKINETIC
i
4
2-20-74
99.92
2.83
115
46
10.87
5453.6
154.4
6755.3
191.3
611.5
229.4
48.61
96.64
2
5
2-20-74
97.96
2.77
.115
46
11.44
5214.9
147.7
6501.2
184.1
550.7
194.6
43.12
98.10
3
6
2-21-74
99.29
2.81
115
46
10.34
5438.0
154.0
6696.1
189.6
544.5
197.4
43.09
97.08
(a) Dry standard cubic feet per minute at 70° F, 29.92 in. Hg.
(b) Dry normal cubic meters per minute at 21.1° C, 760 mm Hg.
(c) Actual cubic feet per minute
(d) Actual cubic meters per minute
-------
TABLE IV
BERYLLIUM SUMMARY
-6-
A* NUMBER
-^j- _ . .. ,.
TEST NUMBER
DOTE
1
1
2-20-74
2
2
2-20-74
3
3
2-21-74
TOTAL WEIGHT OF BERYLLIUM COLLECTED -^g -*0.1 <:0.1 *0.1
SAMPLE VOLUME - ml
23
24
24
MINIMUM DETECTIBLE LEVEL
RUN NUMBER 1
FILTER WEIGHT = ^0
(Assumed
SOLUTION WEIGHT = ^O.l^.g = <; 4 . 3 x 10"3 ALS. * 500 ml Volume)' = ^o 2
23 ml <=.^
EMISSION RATE =
^ 2.2^.g x 179 m3 x 1440 min x 1 gram = -£0.21 grams
2.66 m-3 mTn da 10° /
day 10° //g
day
-------
TABLE V
VISIBLE EMISSIONS
-7-
TEST NUMBER PERCENT OPACITY EMISSIONS
Lb/Hr.
1 0.27 0.663
2 0.00 0.210
3 0.38 0.199
-------
-8-
When visible emissions were presentf the opacity wa,s jce^
ported to be five (5) percent and on one occasion during
Test #3 the opacity reached ten QO) percent. There were
no reported process upsets which could account for the
visible emission increase.
Due to the extremely low opacity readings, no attempt can
be made to correlate opacity with mass emissions..
II. PROCESS SAMPLES
Table VI presents a summary of the mercury, beryllium and
lead analyses of the process samples. A complete laboratory
analysis report is located in Appendix E.
Process samples were obtained from three C31 system input
locations: Plant effluent water Cscrubber feed water 1-, fuel
oil and sewage sludge; and from two 02} system output locations:
Concentrated ash water and constant cooling water overflow
(Figure I). Each sample was analyzed for mercury, lead and
beryllium content. In addition, the sludge samples were sub-
mitted for Optical Emission Spectrograph Analysis to deter~
mine trace metal content. The results of this analysis are
presented in Appendix E-3.
Samples were obtained at approximately 1/4 hour intervals
at the five (5) locations. The 1/4 hour interval samples
(aliquots), at each location, were composited into a total
volume of approximately 400 ml during each test.
The plant effluent water contained an average of 0.067 ug/ml
of lead, 0.02 PPM by weight of mercury and 0.5 ug of bery-
llium. The lead content of the fuel oil was 36,79 jug/ml.
The sewage sludge lead content was determined as 278.8 ug/gm
dry and 0.71 PPM by weight for the mqrcury content. The
lead contents for the ash water and scrubber cooling water
overflow were 3.54 and 0.06 ug/ml, respectively.
The sludge mass flow rate was determined at approximately
1.4 hour intervals by measuring, to the nearest second, the
time required for the incinerator wet sludge feed line to
fill a pre-weighed (empty) five-gallon bucket. The filled
bucket was then pre-weighed and a net weight computed. This
weight and the measured time was used to calculate a wet
sludge flow rate.
Since there was no access available to directly determine
the scrubber water flow rates, pump curves were used to
determine flow rates. These values were supplied by the
Chicago, Illinois, location of the Chicago Bridge and Iron
Co., which is the incinerator contractor.
The fuel oil flow, concentrated ash water flow and cooling
water overflow were not directly measured. These flows were
read from the flow control gauges.
-------
TABLE VI
PROCESS SAMPLE ANALYSIS SUMMARY
-9-
EFFLUENT WATER
No. 2 FUEL OIL
SLUDGE
ASH WATER
SCRUBBER OVERFLOW
RUN NUMBER
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
MERCURY
PPM by WT.
0.01
0.02
0.03
0.2
MISSING
0.01
0.83
0.74
0.76
0.02
0.05
0.01
0.01
0.03
0.10
BERYLLIUM •
TOTAL
0.5
0.5
LOST
0.02
0.02
0.02
0.1
0.1
0.1
0.5
0.5
0.5
0.5
0.5
0.5
LEAD
ug/ml
0.04
0.10
0.06
32.53
39.94
37.91
282.90*
274.87*
278.64*
2.81
3.62
4.19
0.09
0.04
0.06
* ;ag/gm/dry
-------
-10-
III. MATERIAL BALANCES FOR MERCURY AND LEAD
Table VII presents the calculated mercury and lead mass
balance data. The rate of lead input to the incinerator
was computed to be 5137 grams per day and the output from
the incinerator was computed as 3544 grams per day. Based
on these figures, 69.0% of the lead entering the incinerator
was accounted for as leaving the incinerator. The accuracy
of mass determinations depends upon factors such as the ex-
istance of equilibrium mass flow conditions of process mat-
erials and upon the accuracy of measuring mass flow. The
unaccounted 31.0% of the lead input to the incinerator can-
not be attributed to any one specific factor. The cause
may have been sample contamination, the methods of sample
acquisition, handling, storage or analysis. The other pos-
sibilities are inaccurate mass flow rates or the incinerator
system may not have yet reached equilibrium.
The quantity of mercury entering the incinerator was cal-
culated as 73.71 grams per day. The output was computed as
47.38 grams per day.
-------
TABLE VII
MERCURY AND LEAD MASS BALANCE DATA
-11-
OL MATERIAL
(FTgure I)
t.
Plant effluent water
C. Fuel oil
D. Sewage sludge
E. Concentrated ash-laden water
F. Constant cooling water overflow
A. Stack gases
LEAD gram/day
MERCURY gram/day
Input
*
262
4875
0
0
0
Output
0
0
0
3543
*
0.951*
Input
0.66
73.05
0
0
0
Output
0
0
0
2.44
—
44.94 ,
* The concentration of lead in the water samples taken at (B) ,
(F), and the water portion of the samples at (E) are such
that within the accuracy of the analysis, the lead contents
of the water flow in and out of the system are equal/ thereby
canceling each other in the mass balance.
-------
-12-
PROCESS DESCRIPTION AND OPERATIONS
Primary and activated sludge from the wastewater treatment
plant is concentrated by vacuum filtration to 20% solids
and incinerated in fluid bed incinerators. The three (3)
incinerators at the facility are essentially identical and
each have a capacity of 32.2 Ibs. of dry solids per minute.
This amounts to approximately 160 pounds per minute of
sludge with a solids content of 20%. The sludge is fed
into the top of the incinerator and immediately subjected
to temperatures of approximately 1400° F. Rapid and uni-
form evaporation of the entrained water and almost instan-
taneous combustion results when the solids contact the hot
fluidized sand bed. The combustion products, fluidizing
air, and ash are ducted out of the incinerator and are passed
through the shell side of a shell and tube, air to air heat
exchanger. In this combustion air preheater, the incinerator
exhaust gases are cooled from approximately 1400° F to 950°
F and the combustion air is heated from 150° F to about 1000° F.
The partially cooled incinerator exhaust gases then flow
through a flooded throat venturi scrubber where particulate
matter is removed. The scrubber is designed to operate
with a pressure drop across the scrubber of between 10 inches
w.g. and 100 inches w.g. The pressure drop can be varied by
varying the size of the venturi throat. The optimum pres-
sure drop across the scrubber to meet the gas cleaning re-
quirements and lowest operating costs is approximately 40
inches w.g. In the scrubber the fine, ash particles are
wetted and collected while the gases are cooled to about
180° F. The exhaust gases then pass through an entrainment
separator and a gas cooler and are emitted to the atmosphere
at a temperature of between 90° and 115° F through a 30 inch
I.D. stack. The gas is cooled to lessen the white steam
plume which is normally emitted when gases are released dir-
ectly from the scrubber entrainment section.
The ash-water slurry from the entrainment separator is pump-
ed to ash-water separation devices called hydroclones. The
hydroclones separate most of the ash from the slurry. The
overflow from the cyclones is returned to the venturi sect-
ion of the scrubber, and the ash-laden underflow is deposited
in a drag-out type ash classifier. The heavier ash settles
to the bottom of this device, and the liquid overflows and
is recycled to the wastewater treatment plant. The ash is
removed from the classifier vessel by means of a drag-out
rake mechanism and it is subsequently deposited at a land-
fill site. Figure I is a schematic of the incinerator process.
Preliminary estimates indicated that the water-saturated
gases would be emitted from the stack at a rate of 4990
ACFM at 90° F. Constituents of the gases are primarily air
and combustion products.
-------
-13-
Sewage Sludge
156#/HR
©
No. 2
Fuel Oil
Plant
Effluent
Water
1t
I nc i ne ra t<
Venturi
Scrubber
Cooli ng
Dem i s t i ng
De-Ashed
Water
Water
Cyc1 one,
for
De-Ashi n(
Ash Lad<
Water
Concen t ra ted
Ash Laden
Water
T
Cons tan t
Scrubber Overflow
Water
FIGURE I: SEWAGE SLUDGE INCINERATOR PROCESS AT ACCOKEEK, MARYLAND
-------
-14-
LOCATION OF SAMPLING PORTS
1. STACK
Stack height above the roof was forty-four (44) inches. A
five (5) foot stack extension of eighteen (18) gauge gal-
vanized steel was installed to decrease downstream turbul-
ance and minimize the number of points required to sample.
Samples were taken from a point four (4) inches above the
top of the present stack. The extension, at this elevation,
had four (4)- 3 inch diameter sampling ports placed at 90°
around the stack (Figure II).
As a result of locating the sampling ports at this height,
the sampling trains were mounted on platforms. The plat-
forms were constructed so that the probes should easily
slide in and out of the sample ports. When a port was not
in use, it was properly covered to prevent erroneous readings,
2. PROCESS STREAMS
Process samples were gathered at five (5) locations desig-
nated B, C, D, E, and F in Figure I. The samples at points
B and F were obtained at the cooling, demisting column. A
sewage sludge sample, point D, was obtained from-the incin-
erator feeder.
-------
rs-
FIGURE II: STACK LAYOUT -AND TEST PORT LOCATION
Sta^ck Extension
3" Ports . ^
^^"v^
•^
f-30IJ->
*-D-»
mta^tmm
-£2
\
Min
\
(
•
Min.
>
1
. 2D
80
r
^"""^"1
r *i
Q £ 1 •
•miujji
p
West Inc i nera tor
East -Incinerator
-------
-16-
SAMPLING AND ANALYTICAL PROCEDURES
Three (3) sample trains were operated simultaneously in a
coordinated pattern so that the probes could be inserted
into the ports without interfering with each other, (Figure
III). A total of twelve (12) points were, sampled (six (6)
per port) by each train. Because the particulate train was
easier to move, it was used to sample three (3) points at
each of the four (4) ports; whereas the mercury and bery-
llium trains were used to sample two (2) ports each 90° apart.
Stack gas samples were collected for the determination of
total mercury by Method 101, beryllium by Method 104, Fed-
eral Register Volume 36, No. 66, April 6, 1973, and part-
iculate by Method 5, Federal Register, Volume 36, No. 247,
December 23, 1971.
In addition to collecting samples of the exhaust gases,
various process samples were obtained. In reference to
Figure I, samples were collected in mercury free containers
at points B, C, D, E, and F. The containers were prepared
according to the Federal Register 38, No. 66, Part II, April
6, 1973, Method 101, Sections 2.3.1 and 4.5.1. The samples
at points B, C, D, E, and F were obtained at fifteen (15)
minute intervals during the testing at point A (stack).
Each total sample from points B, C, D, E, and F were a
composite of all the approximately equal portions from the
fifteen (15) minute samplings. At points B, C, E, and F a
total sample of 400 ml was obtained while at point D a
minimum total of 400 ml or 400 grams was collected.
Process samples and samples collected at point A (stack)
were labeled according to EPA requirements and given to
Mr. J. Peoples of EPA for analysis by the Source and
Fuels Analysis Section at Research Triangle Park, North
Carolina. The particulate analyses were performed by CES
personnel.
In coordination with the tests performed by CES, the Walden
Research Corporation was contracted to perform visible
emissions readings. These were taken by two (2) ringlemann
readers on the roof of the incinerator building at distances
between twenty-five (25) and fifty (50) feet southeast of
the stack. These locations were chosen to provide adequate
backgrounds consisting of wooded hills and blue skies.
-------
FIGURE III LOCATlW OF SAMPLING POINTS
THREE SAMPLE TRAINS IN ONE STACK- FOR ONE - TWO HOUR TEST,
EACH POINT SAMPLED FOR TEN (10) MINUTES.
0-30 MINUTES
BERYLLIUM
P-ARTICULATE : ,.
_o-o- MERCURY i '
60-50 MINUTES
=S**A-<._
f • , cil. ' s
< : '• t
-------
W.A. Navickas
Source Testing Supervisor
WAN/nk
N.R. Troxefl, P.E.
Manager
Source Testing Operations
------- |