EMISSION TESTING REPORT
EPA REPORT 74-LIM-3-A
WOODVILLE LIME AND CHEMICAL CO.
WOODVILLE, OHIO
PEDCo ENVIRONMENTAL
-------
PEDCo-EN VI RON MENTAL-
SUITE is • ATKINSON SQUARE
CINCINNATI. OHIO 4524.6
513 I-7-7 1-433O
EMISSION TESTING REPORT
EPA REPORT 74-LIM-3-A
WOODVILLE LIME AND CHEMICAL CO.
WOODVILLE, OHIO
Prepared by:
William G. DeWees
Richard W. Gerstle, P.E.
PEDCo-Environmental Specialists, Inc,
Suite 13, Atkinson Square
Cincinnati, Ohio 45246
Contract No. 68-02-0237
Task Order No. 26
-------
I. TABLE OF CONTENTS
Page
II. INTRODUCTION 1
III. SUMMARY OF RESULTS . 3
IV. PROCESS DESCRIPTION 9
V. LOCATION OF SAMPLING POINTS 11
VI. PROCESS OPERATION . 13
VII. SAMPLING AND ANALYTICAL PROCEDURES 16
APPENDICES
A. PARTICULATE RESULTS AND EXAMPLE CALCULATIONS
.B. GASEOUS RESULTS AND EXAMPLE CALCULATION
C. VISIBLE EMISSION RESULTS
D. OPERATIONAL RESULTS
E. FIELD DATA
F. LAB REPORT
G. SAMPLING METHODS
H. TEST LOG
I. RELATED REPORTS
J. PROJECT PARTICIPANTS
K. SUMMARY OF TESTING COST
-------
II. INTRODUCTION
Under the Clean Air Act of 1970, as amended, the Environ-
mental Protection Agency (EPA) is charged with, the establishment
of performance standards for stationary sources 'which may con-
tribute significantly to air pollution. A performance standard
is based on the best emission reduction systems which have been
shown to be technically and economically feasible.
In order to set realistic performance standards, accurate
data.on pollutant emissions must be gathered from the stationary
source category under consideration.
Woodville Lime and Chemical Company in Woodville, Ohio, was
designated as a well-controlled stationary source in the lime
producing industry and was thereby selected by EPA for an emission
testing program. The tests were conducted during the period of
May 20 to 21, 1974. The process under investigation in this test
series was the No. 1 lime kiln at the Woodville plant. Emissions
from the lime kiln are controlled by a cyclone in series with an
electrostatic precipitator. During the test period difficulties
encountered with process operation and above-normal production
rates resulted in nontypical emissions; the test program was
therefore rescheduled for a later date (July 8, 1974).
Results obtained in preliminary testing on May 20 and in
one complete test on May 21 are presented here. The complete
-------
test run included measurements of filterable particulates, total
particulate, oxides of nitrogen, and sulfur dioxide in effluent
from the exit .stack of the lime kiln. In addition, moisture
content and dry molecular weight of the exit gas were determined,
and opacity of visible emissions was recorded by two trained
observers and also with a continuous opacity transmissometer.
-------
III. SUMMARY OF RESULTS
A summary of data on particulate emissions from the No. 1
lime kiln ESP exit stack is presented in Table 1. The pre-
liminary test (A), which was conducted on May 20, 1974, covered
only 40 of the 48 traverse points; the second test (No. 1), con-
ducted on May 21, 1974, covered only 44 of the 48 traverse
points. In addition, the plant was operating at an above-normal
production rate, which caused overloading of the ESP emission
control device. Therefore, the emissions data presented in
this report are considered to be questionable and nonrepresentative
of a well-controlled lime producing process.
The EPA Process Engineer decided to conduct several tests
on the effluent gas stream from the ESP even though the process
rate exceeded normal operational parameters. This decision was
based on the need for information with which to determine.
correlation of emissions concentrations and visible opacity
observations. EPA had transported to the plant site a continuous
monitoring transmissometer for opacity measurements; the emission
measurements therefore were performed for purposes of correlation
with readings from this instrument, even though emission rates
were known to be above normal for this plant.
The filterable particulate, as measured by the probe and
filter catch, averaged 30.35 pounds per hour at a concentration of
-------
Table 1. SUMMARY OF PARTICULATE RESULTS
LIME KILN
Run Number
Date
Volume of Gas Sampled,-DSCFa
Average Stack Temperature, °F
Percent Moisture by Volume
Stack Volumetric Flow Rate, DSCFM
Stack Volumetric Flow Rate, ACFMC
Percent Isokinetic
Feed Rate , ton/hr
Particulates - probe, bypass, and
filter catch
mg
gr/DSCF
gr/ACF
Ib/hr
Ib/ton
Particulates - total
mg
gr/DSCF
gr/ACF
Ib/hr
Ib/ton
A
5/20/74
56.155
675
12.17
26994
66504
99.3
5/21/74
121.814
686 681
10.65 11.41
27142 27068
66504 66504
98.2
30.62
277.2 1463.4
0.0762 0.1854
0.0310 0.0757
17.6 43.1
1.41
635.2 1586.8
0.1746 0.2010
0.0710 0.0820
40.4 46.8
1.53
0.131
0.0533
30.35
0.188
0.0765
43.6
Dry- standard cubic feet at 70°F, 29.92 in Kg.
b Dry standard cubic feet per minute at 70°F, 29.92 in. Hg
0 Actual cubic feet per minute.
-------
0.131 grains per DSCF. Total particulate emissions averaged 43.6
pounds per hour at a concentration of 0.188 grains per DSCF.
Results of the second test yielded higher concentrations of
filterable particulate than were obtained in the first test,
probably because of discontinuities in process operation and
production rates.
Emissions of oxides of nitrogen and sulfur dioxide were
determined only during the second test. A summary of oxides
of nitrogen data is presented in Table 2. These data show
average emissions of 481 ppm by volume and 93.2 pounds per hour.
The method for sampling of sulfur dioxide was intended to
measure emissions in the 1000 ppm range. Analysis of the samples
obtained,.however, showed sulfur dioxide concentrations much
lower than those expected, about 20 ppm; sample volume was there-
fore inadequate to yield measurable results. This low concen-
tration was attributable to the neutralization of the sulfur
dioxide by the alkaline dust produced in the process.
Determinations of visual opacity of emissions from the lime
kiln stack were performed independently by two PEDCo personnel.
using Federal Register Method 9. Opacity was also measured by the
Lear Siegler continuous transmissometer* Data on visible emissions
are summarized in Table 3. The average opacity was 14 percent
for all readings during the test periods. A 20 percent opacity
level was exceeded about 8.5 percent of the test time. This high
visible opacity was another indication of emissions much higher
than those expected.
-------
Table 2. SUMMARY OF NITROGEN OXIDES EMISSIONS FROM
LIME KILN
Test no.
Date :
Time, 24 hr
Flow rate, DSCFM
Sample volume, ml
Milligrams of NO2
Concentration, ppm
Emissions, Ib/hr
5/21/74
1249
27142
1025.46
0.979
500.7
96.94
5/21/74
1339
27142
1145.04
1.009
462.2
89.48
Average
column
27142
481.5
93.21
-------
Run No. 2
Date V21/74
Table 3. 'Summary of Visual Opacity Readings
Lear
Siegler
Interval of Observations
1
Start
• -End
Duration of Observation (niin)
Total No. of Readings
No. of Readings Unobservable
No. of Readings @ 0% Opacity
5%
10%
15%
' 20%
'25%
30%
35%
• 50%
Percent Readings Unobservable
Percent Readings (? .0% Opacity
5%
10%
1 5%
• 20%
Percent Readings Exceed ing 20%
Obs, 1
1158
1441
163
653
-
0
0
189
388
54
21
1
0
0
0
0
-
0
0
28.9
59.4
8.3
3.4
Obs. 2
1158
1458
180
720
-
• ' . 4
118
'201
270
64
22
15
18
6
2
0
-
0.6
16.4
27.9
37.5
8.9
8.8
Obs. 1
2
"~l
1530
1550
20
80
-
0
0
42
24
14
0
0
6
0
0
0
- .
0
0
52.5
30.0
17.5
0
Obs. 1
Obs.
24-hour clock start and end times
p .
"Excluding the time that readings were not recorded for period- of observation.
Readings recorded at 15-second intervals unless other v/ise noted.
Observer 1 - W. DeWees
Observer 2 - R. Amick
-------
The testing series was terminated by OAP personnel before
the desired number of tests were run because the lime kiln.was
running at a high production rate, and visible emissions were
exceeding the average 5 percent opacity level that was normal
for the No. 1 lime kiln exit stack.
-------
IV. PROCESS DESCRIPTION
Limestone consisting primarily of calcium carbonate or combina-
tions of calcium and magnesium carbonate with varying amounts of
impurities is quarried at the Woodville Plant. The limestone is
calcined or burned to form lime, commonly divided into two basic pro-
ducts—quicklime and hydrated lime. Calcination expels carbon
dioxide from the raw limestone, leaving calcium oxide (quicklime).
With the addition of water, calcium hydroxide (hydrated lime) is
formed.
The basic processes in production are: (1) quarrying the lime-
stone raw material, (2) preparing the limestone for kilns by crushing
and sizing, (3) calcining the limestone, and (4) optionally processing
the quicklime further by additional crushing and sizing followed by
i
hydration. The majority of lime is produced in rotary.kilns which
can be fired by coal, oil, or gas. Rotary kilns have the advantage
of producing high production per man-hour and a more uniform product.
However, they do require higher capital investment and unit fuel
costs than most vertical kilns.
The Woodville Lime and Chemical plant has two rotary kilns
each equipped with a Buell electrostatic precipitator. The kilns
are almost identical. The feed for both is a dolomitic stone,
quarried on the site and fed in sizes ranging from 1 inch to
2 1/4 inches at a rate of about 700 tons per day. There is no
-------
preheater. Normally the kiln is fueled with a mixture of
95 percent Number 6 fuel oil and 5 percent natural gas. Both
kilns have two heat transfer sections, each 20 feet long. The
product, about 350 tons per day, is cooled in a.Neims cooler before
storage. There is no product crushing, but undersize material
is separated and returned to the kiln. The majority of the
product is used in the steel industry, mostly in basic oxygen
furnaces;, none of the product is hydrated.
The electrostatic precipitator on kiln Number 1 was put in
operation in July 1971. In this kiln the main process fan is
located before the ESP, with a cyclone before the fan to reduce
fan blade erosion. The precipitator on kiln Number 2 was put in
operation in December 1973; The main process fan is after the
ESP and there is no .cyclone.
In both systems the inlet gas to the precipitators is cooled
to about 600°F with a combination of water injection and/or
tempering air. Each precipitator has 28,800 square feet of
collecting surface area, which includes one cell and two fields;
design gas velocity is 1.5 feet per second and treatment time,
10.0 seconds.
At present the dust collected from the precipitators is
disposed of in the quarry. It is expected that in the future the
dust will be granulated and used as a component of dry mix
fertilizers that are blended in another part of the complex.
10
-------
V. LOCATION OF SAMPLING POINTS
Figure 1 shows the sampling ports and sampling points used
in the No. 1 lime kiln exit stack. The sampling ports were
located in a 63.5-inch inside-diameter vertical stack, 4 feet
(0.75 diameter) from the stack exit, and 12 feet (2.26 diameters)
from the nearest downstream disturbance. In order to meet the
sampling requirements of Methods 1 and 5 of the Federal Register,
Vol. 36, No. 247, it was necessary to install a stack extension
on the ESP exhaust outlet. Forty-eight traverse points (24
along each of two perpendicular diameters) were used as described
in the Federal Register Method 1. Additional sampling points in
the existing stack at a lower site were used for some of the gas
sampling.
11
-------
Scaffolding
Edge of roof
W
1 1
CROSS SECTION
All Dimensions in Fool Arc Approximate
Port 3 " Diameter.
S Porl 6" Diameter J
_ 87' lo ground
ELEVATION
Figure 1. Test Site-No. 1 Kiln Precipitator Outlet.
12
-------
VI. PROCESS OPERATION
The EPA project engineer arrived at the plant at 3:00 p.m.
on Monday, May 20. At this time the plant was running at 5 to
10 percent over the normal production rate. Because the
Pfizer Chemical Company lime plant in Perrysburg, Ohio, had
just gone on strike,.production at the Woodville plant had
been increased to the absolute maximum in anticipation of
increased demand. Neither stack was clear (about 20% opacity),
and the ESP voltage and amps were reading low. .'It was assumed
that the higher opacity was due to the increased production rate,
which caused the process to emit greater concentrations of dust.
The kiln feed rate was calculated by multiplying the factor of
0.7291666 times the total indicated tonnage recorded from the
control room stone totalizer meter (see reference letter in
Appendix D).
On the morning of May 21, kiln No. 1 was processing an
average of 30.62 tons of stone per hour. The ESP voltage and
amp readings were still low, and opacity readings still high.
(about 15%, with some as high as 30%). A water spray was added
to the feed end of the kiln to increase the conductivity .of the
gas stream in the ESP, but no effect on collection performance
was noticed. Plant officials believed that increasing the
13
-------
production capacity may have resulted in "heavy .combustibles"
(oil) passing into the ESP. This oil possibly combined with
dust to build up a coating on the wires and plates. This
coating would eventually be removed by the rappers but this
process might require several weeks. Further testing was post-
poned until the time when the plant and the ESP resumed normal
operation. The operating variables are summarized in Table 4.
14
-------
Table 4. SUMMARY OF OPERATING VARIABLES
Test No. 1
Stone feed rate , (tons/hr) 30.62
Oil rate, (gal/hr) 388
Gas to oil ratio, (BTU .Basis)
Mid kiln temp.,(°F) 1490-1530
Feed end temp.,(°F) 1040-1100
Before ESP temp.,(°F) 690-740
Stack temp.,(°F) 675-710
Electrostatic Precipitator Data
A Field
Primary current,(amps) 35-46
Primary voltage,(volts) 210-245
Precipitator current,(amps) 0.17-0.21
B Field
Primary current,(amps) 22-39
Primary voltage,(volts) 200-245
Precipitator current,(amps) 0.12-0.21
NOTE: No operational data were collected during the
preliminary run on May 20, 1974.
15
-------
VII. SAMPLING AND ANALYTICAL PROCEDURES
Sampling procedures were designated by EPA. Analyses of
collected samples were performed by PEDCo. Appendix G presents
detailed sampling and analytical procedures.
Velocity and Gas Temperature
Gas velocities were measured with a calibrated type S
pitot tube and inclined draft gage. Velocities were measured
at each sampling point across the stack diameter to determine
an average value according to procedures described in the
Federal Register - Method 2. Temperatures were measured with
the use of a thermocouple.
Molecular Weight
A 4-hour integrated sample of the stack gases was collected
during test 1 by pumping the gas into a Tedlar plastic bag at
the rate of approximately 0.005 CFM. This bag sample was then
analyzed with an Orsat analyzer for C02, 0-, and CO as described
in the Federal Register, Method 3. . .
Particulates
2
Method 5 as described in Federal Register, was used to
measure particulate matter. A rigid train consisting of a heated
glass-lined probe, a 3" glass fiber filter, and a series of Green-
burg-Smith impingers, as shown in Figure 2, was employed in all
particulate tests.
1) Federal Register, Vol. 36, No. 247, December 23, 1971.
2) Federal Register, Vol. 36, No. 159, August 17, 1971.
16
-------
FILTER
THERMOMETER
HEATED
GLASS
PROBE
Jj_qO_m.L_ OF WATER
THERMOMETERS
UMBILICAL
CORD
CALIBRATED ORIFICE
MANOMETER- I
Figure 2. Particulate Sample Train
-------
Sampling was conducted under isokinetic conditions
by monitoring the velocity with a pitot tube and adjusting
the sampling rate accordingly. Because of the heavy dust
loading during a 4-hour test, two filter changes were required
during Test 1. After 131 minutes of testing on Test 1 the
high temperature and high vacuum pulled a hole in the poly-
vinyl vacuum line in the meter box. The hole was detected
immediately because of the vacuum that was required to sample.
The test was terminated at this point.
Sample recovery consisted of triple-rinsing the nozzle,
probe, cyclone by-pass, and front half of the filter holder
with acetone into a glass container. The back half of the
filter holder, impingers, and connecting tubes were first
rinsed with distilled water, then placed in a glass container
along with the impinger contents. These components were then
triple-rinsed with acetone and the washings placed in another
glass container. The filter was placed in a separate con-
tainer. Blank samples of water and acetone were also taken.
N0x
Nitrogen oxides were collected in evacuated 2-liter
flasks containing 25 ml of a dilute sulfuric acid/hydrogen
peroxide absorbing solution. The sampling and analytical
procedure was as described in Method 7 of the Federal Register
except that the final flask vacuum was read immediately after
sampling.
1) Federal Register, Vol. 36, No. 247, December 23, 1971,
18
-------
The samples were analyzed by the PDSA method. Two samples
were lost in the lab by breakage.
so2
Sulfur dioxide was determined by passing a measured volume
of flue gas through a set of midget impingers at a rate of
approximately 0.025 cubic foot per minute. The first impinger
contained 15 ml of 80 percent isopropanol; the second and third
impingers contained 15 ml each of hydrogen peroxide in water
solution. The hydrogen peroxide solution was titrated with
barium chloride using a Thorin indicator to determine sulfur
dioxide (S0?). Because the concentration was so low, less than
1 ml of titrant was required and an accurate measurement was
not possible.
19
------- |