EPA-600/2-76-009
January 1976
Environmental Protection Technology Series
ODOR CONTROL BY SCRUBBING IN THE
RENDERING INDUSTRY
Indystrial Etifironmentil Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
EPA-600/2-76-009
ODOR CONTROL
BY SCRUBBING
IN THE RENDERING INDUSTRY
by
R.H. Snow and J.E. Huff (HTRI)
and Werner Boehme
Fats and Proteins Research Foundation, Inc.
2720 Des Plaines Avenue
Des Plaines, Illinois 60018
Contract No. 68-02-1087
ROAPNo. 21AXM-062
Program Element No. 1AB015
EPA Project Officer: E.J. Wooldridge
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
January 1976
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research-and Development,
U.S. Environmental Protection "Agency, have been grouped into
five series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
Telated fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed
to develop and demonstrate instrumentation, equipment and
methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the
new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection
Agency, and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of the Agency, nor
does mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
This document is available to the public through the National
Technical Information Service, Springfield, Virginia 22161.
-------
FOREWORD
This report covers investigations performed by the IIT
Research Institute (IITRI), Project C8243, under contract with
the Fats and Proteins Research Foundation, Inc.
In this study, odor control in the rendering industry
using wet scrubbers was investigated. Multiple-stage scrubbers,
with alkaline hypochlorite scrubbing solutions were tested in
a rendering plant to control odorous emissions. Scrubbers offer
an economic solution to odor problems. The information provided
in this report will be a valuable aid in designing scrubbing
systems to achieve specified reduction in odor levels.
Those who contributed to this project were: Dr. A. Dravnieks,
J. E. Huff, Dr. R. H. Snow, W. Stepp, C. Swanstrom, and
J. Whitfield of IIT Research Institute, and Dr. W. Boehme of
the Fats and Proteins Research Foundation, Inc.
Data are recorded in Logbooks C21380, C21569, C21573,
C21633, and C21911.
The financial support of the Environmental Protection
Agency, under Contract No. 68-02-1087, and the valuable advice
and assistance of the Project Officers, Dr. Belur Murthy and
E. J. Wooldridge, is gratefully acknowledged.
^LXAv^xH
-------
INVESTIGATION OF ODOR CONTROL IN THE RENDERING INDUSTRY
ABSTRACT
Experiments were conducted at a rendering plant to obtain
data needed to design wet scrubber systems for rendering plant
odor control. Scrubber performance was measured both by odor
panel and gas chromatographic analysis.
A series of experiments in a three-stage packed-bed labora-
tory-scale scrubber at the rendering plant evaluated solutions
of sodium hydroxide and the strong oxidants sodium hypochlorite,
hydrogen peroxide, and potassium permanganate. Removals of
90% per stage were obtained with fresh alkaline sodium hypo-
chlorite solution, and this reagent was selected for the sub-
sequent longer-term tests.
A two-week test of plant-scale horizontal spray scrubber
operating on plant ventilating air showed odor removals of 83%.
The outlet odor units averaged 64, while the inlet ranged from
165 to 2,500 odor units.
A three-stage packed-bed scrubber was evaluated to replace
an existing incinerator being used to treat a process air
stream that contained from 5,000 to 50,000 odor units. A
week-long test with the scrubber gave an average odor reduction
of 85%, lower than expected. Further work is suggested to
investigate the conditions necessary to improve the results.
Data was obtained on chemicals consumption and effect of
flow variables on odor removal, and these data were used to
update computer models that can be used for design of scrub-
bers for odor removal.
-------
TABLE OF CONTENTS
Page No.
1. INTRODUCTION -- PREVIOUS WORK 1-1
2. SUMMARY 2-1
3. DESCRIPTION OF THE LABORATORY AND PLANT SCRUBBERS 3-1
4. DESCRIPTION OF ANALYTICAL TECHNIQUES .... 4-1
5. ODOR CONTRIBUTION OF PARTICULATES 5-1
6. EVALUATION OF SCRUBBING LIQUIDS 6-1
7. EFFECT OF SODIUM HYPOCHLORITE CONCENTRATION
ON ODOR REMOVAL 7-1
8. PERFORMANCE EVALUATION OF THE HORIZONTAL SPRAY
SCRUBBER 8-1
9. PERFORMANCE EVALUATION OF PACKED-BED SCRUBBERS 9-1
10. DESIGN AND COST OF COUNTERCURRENT PACKED TOWER
GAS SCRUBBING SYSTEM 10-1
11. DESIGN AND COST OF HORIZONTAL SPRAY SCRUBBER 11-1
12. DESIGN AND COST OF A PACKED TOWER-HORIZONTAL
SPRAY SCRUBBER SYSTEM 12-1
13. ACTIVATED CARBON STUDY 13-1
14. ON-SITE GENERATED OF SODIUM HYPOCHLORITE . . 14-1
15. SELECTING THE PROPER SCRUBBER DESIGN .... 15-1
REFERENCES R-l
APPENDIX 1, COMPARATIVE SCRUBBING TEST RESULTS Al-1
APPENDIX 2, DERIVATION OF CHEMICAL CONSUMPTION
EQUATIONS A2-1
APPENDIX 3, CALIBRATION OF THE HORIZONTAL SPRAY
SCRUBBER MODEL A3-1
APPENDIX 4, DEVELOPMENT OF THE PACKED-TOWER
MASS TRANSFER EQUATION A4-1
v
-------
TABLE OF CONTENTS (cont.)
Page No
APPENDIX 5, DESIGN AND COST CALCULATIONS FOR
PACKED TOWERS A5-1
APPENDIX 6, MASS TRANSFER AS A FUNCTION OF
PACKING DEPTH A6-1
VI
-------
LIST OF TABLES
Table No.
6-1
6-2
6-3
6-4
6-5
7-1
7-2
8-1
8-2
8-3
8-4
8-5
8-6
8-7
m
9-1
9-2
9-3
Page No.
SCRUBBING LIQUID COMBINATIONS
PLANT SCRUBBER COMPARATIVE SCRUBBING
TESTS
LABORATORY SCRUBBER COMPARATIVE SCRUBBING
TESTS
SUMMARY OF SCRUBBING TEST 31 LABORATORY
SCRUBBER
SUMMARY OF SCRUBBING TEST 32 LABORATORY
SCRUBBER
TEST 30 - SUMMARY
TEST 32 - SUMMARY OF RESULTS
SUMMARY OF SCRUBBING TEST 34 PLANT SCRUBBER
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 29 HOURS
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 31 HOURS
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 77 HOURS
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME : 84 HOURS
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 150 HOURS
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 175 HOURS
SUMMARY OF SCRUBBING TEST 35 LABORATORY
SCRUBBER
RESULTS OF SCRUBBING TEST 35-1 LABORATORY
SCRUBBER, TIME: 9 HOURS
RESULTS OF SCRUBBING TEST 35-3 LABORATORY
SCRUBBER, TIME: 22 HOURS
6-2
6-3
6-5
6-7
6-10
7-2
7-3
8-7
8-8
8-9
8-10
8-11
8-12
8-13
9-8
9-9
9-10
vii
-------
LIST OF TABLES
CONTINUED
Table No. Page No.
9-4 RESULTS OF SCRUBBING TEST 35-4 LABORATORY
SCRUBBER, TIME: 32 HOURS 9-11
9-5 RESULTS OF SCRUBBING TEST 35-5 LABORATORY
SCRUBBER, TIME: 48 HOURS 9-12
10-1 PRESSURE-DROP EQUATION FOR PARTICLE PACKING 10-13
10-2 CHARACTERISTICS OF RANDOM PACKINGS . . 10-15
10-3 PACKED TOWER COST SUMMARY 1-1/2-IN.
INTALOX SADDLES 10-34
10-4 OPTIMUM DESIGN OF PACKED TOWERS
1-1/2-IN. INTALOX SADDLES 10-35
11-1 RESULTS OF SCRUBBING TEST 36 - COARSE
NOZZLES PLANT SCRUBBER 11-4
11-2 RESULTS OF SCRUBBING TEST 36 - FINE
NOZZLES PLANT SCRUBBER 11-5
11-3 SPRAY SCRUBBER COST SUMMARY 11-8
11-4 OPTIMUM DESIGN OF SPRAY SCRUBBERS . . . 11-9
12-1 COMPARISON OF SCRUBBING SYSTEMS .... 12-4
13-1 ACTIVATED CARBON - ODOR PANEL RESULTS . 13-2
13-2 ACTIVATED CARBON TEST TIME: 20 HOURS. . 13-3
13-3 COST OF ACTIVATED CARBON FOR LOW ODOR
LEVELS 13-6
14-1 ON-SITE GENERATION OF SODIUM HYPOCHLORITE 14-3
15-1 COSTS OF PACKED BED VERSUS SPRAY SCRUBBERS
FOR PLANT VENTILATING AIR 15-2
V3.ll
-------
LIST OF TABLES
Table No. Page No
6-1
6-2
6-3
6-4
6-5
7-1
7-2
8-1
8-2
8-3
8-4
8-5
8-6
8-7
"M
9-1
9-2
9-3
SCRUBBING LIQUID COMBINATIONS 6-2
PLANT SCRUBBER COMPARATIVE SCRUBBING
TESTS 6-3
LABORATORY SCRUBBER COMPARATIVE SCRUBBING
TESTS 6-5
SUMMARY OF SCRUBBING TEST 31 LABORATORY
SCRUBBER 6-7
SUMMARY OF SCRUBBING TEST 32 LABORATORY
SCRUBBER 6-10
TEST 30 - SUMMARY 7-2
TEST 32 - SUMMARY OF RESULTS 7-3
SUMMARY OF SCRUBBING TEST 34 PLANT SCRUBBER 8-7
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 29 HOURS 8-8
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 31 HOURS 8-9
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 77 HOURS 8-10
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 84 HOURS 8-11
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 150 HOURS 8-12
RESULTS OF SCRUBBING TEST 34 PLANT SCRUBBER,
TIME: 175 HOURS 8-13
SUMMARY OF SCRUBBING TEST 35 LABORATORY
SCRUBBER 9-8
RESULTS OF SCRUBBING TEST 35-1 LABORATORY
SCRUBBER, TIME: 9 HOURS 9-9
RESULTS OF SCRUBBING TEST 35-3 LABORATORY
SCRUBBER, TIME: 22 HOURS 9-10
vii
-------
LIST OF TABLES
CONTINUED
Table No. Page No.
9-4 RESULTS OF SCRUBBING TEST 35-4 LABORATORY
SCRUBBER, TIME: 32 HOURS 9-11
9-5 RESULTS OF SCRUBBING TEST 35-5 LABORATORY
SCRUBBER, TIME: 48 HOURS 9-12
10-1 PRESSURE-DROP EQUATION FOR PARTICLE PACKING 10-13
10-2 CHARACTERISTICS OF RANDOM PACKINGS . . 10-15
10-3 PACKED TOWER COST SUMMARY 1-1/2-IN.
INTALOX SADDLES 10-34
10-4 OPTIMUM DESIGN OF PACKED TOWERS
1-1/2-IN. INTALOX SADDLES 10-35
11-1 RESULTS OF SCRUBBING TEST 36 - COARSE
NOZZLES PLANT SCRUBBER 11-4
11-2 RESULTS OF SCRUBBING TEST 36 - FINE
NOZZLES PLANT SCRUBBER 11-5
11-3 SPRAY SCRUBBER COST SUMMARY 11-8
11-4 OPTIMUM DESIGN OF SPRAY SCRUBBERS ... 11-9
12-1 COMPARISON OF SCRUBBING SYSTEMS .... 12-4
13-1 ACTIVATED CARBON - ODOR PANEL RESULTS . 13-2
13-2 ACTIVATED CARBON TEST TIME: 20 HOURS. . 13-3
13-3 COST OF ACTIVATED CARBON FOR LOW ODOR
LEVELS 13-6
14-1 ON-SITE GENERATION OF SODIUM HYPOCHLORITE 14-3
15-1 COSTS OF PACKED BED VERSUS SPRAY SCRUBBERS
FOR PLANT VENTILATING AIR 15-2
Vlll
-------
LIST OF FIGURES
Figure No. Page No.
3-1 LABORATORY PACKED-BED SCRUBBER 3-2
3-2 ONE-STAGE HORIZONTAL SPRAY SCRUBBER. . . 3-4
4-1 FIELD SAMPLING SYSTEM 4-2
4-2 CHROMATOGRAPH SAMPLER 4-6
4-3 CHROMATOGRAPH SAMPLING AND ANALYSIS SYSTEM 4-8
8-1 DIAGRAM OF PLANT SCRUBBER SYSTEM WITH
CONTINUOUS COUNTERCURRENT REAGENT SUPPLY 8-3
10-1 PACKED TOWER 10-5
10-2 K_a VS GAS RATE 10-10
U
10-3 PLANT SCRUBBER CHLORINE LEVELS 10-19
10-4 LABORATORY SCRUBBER CHLORINE LEVELS . . 10-20
10-5 CHLORINE CONSUMPTION RATES FOR SCRUBBER
DESIGN MODEL 10-22
12-1 CASE I -3-STAGE SPRAY SCRUBBER 12-2
12-2 PACKED-BED WITH A 2-STAGE SPRAY SCRUBBER 12-3
IX
-------
INVESTIGATION OF ODOR CONTROL IN THE RENDERING INDUSTRY
1. INTRODUCTION -- PREVIOUS WORK
The Fats and Proteins Research Foundation, Inc., and the
U.S. Environmental Protection Agency have jointly sponsored
a series of research programs to control rendering plant
odors. One of the early studies resulted in the identifica-
tion of many of the odorous compounds present in rendering
plant.air. This was accomplished by taking samples of air
at rendering plants, separating the organic compounds into
peaks by means of gas chromatography and identifying the
peaks by means of mass spectroscopy. A computer regression
analysis determined which of the peaks were correlated with
the odor intensity of the samples, as independently deter-
mined by an odor panel. In this way, most of the odorous com-
pounds were ascertained for the first time, and their chemi-
cal and physical properties were determined from handbooks.
The important peaks and their identifications were listed
in Report No. EPA-R2-72-088. They include various sulfides,
disulfides (and probably mercaptans), C4 to C7 aldehydes,
trimethyl amine and various C4 amines, quinoline; dimethyl
pyrazine and other pyrazines, C3 to C6 acids. Other compounds
are present but are not important so far as odor is concerned.
The objective of a second project was to develop an
effective and economic means of odor control. Before under-
taking laboratory experiments, known and potential odor con-
trol methods were compared based on limited data from the
literature, and on reports of performance and costs from mem-
ber companies of the Fats and Proteins Research Foundation.
This comparison indicated that scrubbing was probably the
most economic method, especially if one or more effective
scrubbing reagents could be found. Incineration was also
feasible, but was more costly for plant ventilating air.
The study also indicated that commercially available control
1-1
-------
equipment was not designed on a reliable basis, partly be-
cause of the lack of quantitative data on the odorous com-
pounds in rendering odors. Rendering companies were not in
a position to rationally select from among the control equip-
ment offered by manufacturers.
Computer design studies were performed, including cal-
culations for capital and operating costs for odor control
by incineration, and by packed bed scrubbers and horizontal
spray scrubbers. The incinerator performance was assumed
to be governed by an overall combustion reaction, the rate
constants for which were determined as functions of tempera-
ture from available plant data. The design of scrubbers was
based on the idea that reagents could be found that react
rapidly with the odorants, so that the scrubbing process
would be mass-transfer controlled. Based on these concepts,
the computer programs were able to explore or optimize the
operating conditions to find the design having minimum annual
cost. Results of design and cost estimates for each control
method were presented in Report No. EPA-R2-72-088 in tabular
form for a series of air flow rates and odor reduction ratios.
As a result of the information presented, we concluded that
scrubbers offer the most economical odor control method for
plant ventilating air, provided that effective scrubbing rea-
gents can be found. At that time (1972) , it appeared that
incinerators might be competitive for a small, highly con=-
centrated process air stream, but subsequent increases in
fuel costs have since changed this conclusion.
The next steps in the investigation were to determine
the performance of various candidate scrubbing reagents. The
experiments were of an exploratory nature, and were conducted
in the laboratory using the pure odor ingredients identified
in the previous step of the project. A first series of
experiments (called bubbler experiments) were designed to
measure the ultimate reactivity of the reagents with no
1-2
-------
mass-transfer limitations. A second set of experiments mea-
sured the removal of odorous compounds from an air stream
using a laboratory-scale, packed-bed scrubber. The scrubber
was designed with sufficient mass transfer to reduce odors
by 90%, assuming that the reaction rate is not limiting.
Odor compound reduction was measured by gas chromatography.
Reductions of 80-90% were achieved with water for highly
soluble odorants such as C4 amines and acids. Sodium car-
bonate was found to be no more effective than water, and
sodium hydroxide gave only slightly higher removals. Sodium
hypochlorite removed 90% of a number of odorants, including
dipropyl sulfide and trimethyl amine. Potassium permanganate,
another strong oxidizer, reacted slower but attacked some
odorants not handled by hypochlorite. Hydrogen peroxide
and sodium persulfate were comparable to hypochlorite.
Sodium bisulfite removed 90% of aldehydes.
Based on the results of these experiments, it was deter-
mined that several of the strong oxidants should be evalua-
ted as scrubbing reagents using actual rendering plant air.
1-3
-------
2. SUMMARY
2.1 Objectives of Research Program
The objective of this program was to develop the necessary
information to control rendering plant odors using wet scrubbers
based on experiments with actual rendering plant air. The
specific objectives of this program were as follows:
to determine the most effective and economical com-
bination of scrubbing liquids
to obtain experimental data on the performance of
scrubbers using actual rendering plant ventilating
air, which typically contains between 100 to 1,000
odor units
to develop computer models to assist in determining
the optimum design for packed-bed and horizontal
spray scrubbers based upon the experimental data
collected
The scope of the program was amended to include
developing a scrubber system to handle high-intensity
odors (5,000-50,000 odor units) that are presently
treated by incinerators, and thus avoid problems
of energy shortage and high cost of fuels required
by incineration.
2.2 Odor Sampling and Analysis Methodology
A methodology for sampling and analysis of scrubber inlet
and outlet air had to be established prior to testing scrubber
performance. Two methods, sensory (odor panels) testing and
gas chromatographic analysis, were used. The first method
was already established (Dravnieks and Prokop, 1973) which
utilized the collection of the odorous samples in thick-walled
polyethylene bags and used the dynamic olfactometer for analy-
sis . The gas chromatographic sampling method was developed
on the previous project for plant air testing (Dravnieks, et
al., 1971). To improve the accuracy of the results, this method
was modified and tested in the first few months of the project.
Small metal samplers, fitted with chromatographic adsorbent ma-
terial, were developed for collecting the gas chromatographic
samples.
2-1
-------
The measurement of odors presents special difficulties,
and it cannot be carried out with the accuracy expected in
usual analytical chemistry work. Odor panel measurements with
butanol have shown an accuracy of +307o, but the results in
this report with rendering plant air appear less accurate than
this. A few repeat tests have fallen within the range of
a factor of 2. In the case of gas chromatography, the
measurements with pure compounds were found to be reproducible
within 29%.- The reproducibility with the same rendering air
sample is similar. However, with different rendering plant
air samples, the relative amounts of various peaks changes
from sample to sample; since the peaks may overlap, it is
difficult to determine the baseline for graphical integration.
From observing the variations of individual peaks within the
chromatograms, we estimate that the chromatograph data has
about, the same accuracy as the odor panel data.
The only way to establish the accuracy of the plant
measurements is to repeat each odor removal test and statis-
tically analyze the data; not enough repeat experiments were
done to permit such an analysis of the data.
However, the reliability of the results is increased by
combining the results from both methods. For example, if the
standard deviation of odor measurement by each method is
a factor of 2, or a 347o chance that the error is more than
a factor of 2, then the probability that the error of both
methods of measurement exceeds a factor if 2 is 0.34 x 0.34,
or 12%. This means that 88% of the time the error will be
less than a factor of 2. Thus, a 9070 removal (10% odor
residual) will have an 88% chance of lying between 80 and 95%
odor removal (5 to 20% odor residual). To increase the
accuracy of the results will require carrying out many more
repeat odor removal measurements. In many cases, the
agreement between the two methods was better than this, but
in a few critical cases, it was comparable to this example.
2-2
-------
2.3 Determination of the Optimum Scrubber Reagent
Combination
A three-stage, full-scale horizontal spray scrubber and
a three-stage, laboratory packed-bed scrubber were used
throughout this program. By using multiple stages, a different
scrubbing reagent can be utilized in each stage. A series of
tests were performed to evaluate the effectiveness of several
scrubbing reagents and combinations thereof. Three different
combinations resulted in odor removals of over 9870 when scrubbing
the high-intensity odors (5,000-50,000 odor units) with the
three-stage packed-bed scrubber. Reagent life tests were per-
formed on all three combinations to ascertain which combination
was the most effective. The results of these tests indicated
that multiple stages of alkaline sodium hypochlorite gave the
highest odor removal (96% through a two-stage packed-bed as
compared to 77% and 8970 for the other two combinations) at a
lower cost than other scrubbing combinations.
2.4 Full-Scale Plant Scrubber Tests
The final step in this program was the testing of the
performance on plant ventilation air of the plant horizontal
spray scrubber with the selected reagent combination over a
two-week period. The plant scrubber was modified to add
sodium hypochlorite and water continuously to the system, and
to bleed off a small fraction of the scrubbing solution to
the sewer continuously. This test was performed during the
hottest two weeks of the year, providing severe test conditions.
Throughout the test period, no odor complaints were
received from the surrounding community, confirming satis-
factory performance of the scrubber. The reduction in odor
level averaged 83% (92% if one test is excluded where we
believe the effluent contained a chlorine odor) while the
2-3
-------
gas chromatographic results showed an average reduction in
odorant concentration of 857o. The outlet odor level averaged
64 odor units for the entire test, while the inlet ranged from
165 to 2,500 odor units.
2.5 Scrubbing of High Intensity Odors
As an additional objective of this program, the feasibility
of treating incinerator inlet air streams with packed-bed
scrubbers was determined. Preliminary tests on the laboratory
packed-bed scrubber demonstrated that removals of 90% per
stage could be obtained with fresh alkaline hypochlorite solu-
tions. Additional tests indicated that odor removals of 8870
per stage could be achieved after many hours of scrubbing,
even when the hypochlorite concentration had decreased 75?0.
A week-long laboratory packed-bed test was conducted at
the plant to collect performance data on the continuous make-
up and blow-down system. The performance of the scrubber
was considerably below the anticipated level, since the
average odor component reduction was only 857». The reasons
for the lower than anticipated removals were attributed to
the following problems: (1) blow-down rate was initially
set too low; (2) caustic addition rate was initially set too
low, resulting in lower than required pH levels; and (3)
air temperature, which reduced the capacity of the scrubbing
reagents. Problems 1 and 2 were partially corrected during
the week-long test, and the scrubber efficiency was improving
near the end of the test, 98 and 79% removal compared to 59
and 34% for the first 2 measurements. Considerable data was
obtained from this test needed for completing the computer
design model for packed-bed scrubbers.
2.6 Computer Design Models and Calculations
Computer design models were developed during the previous
program (Report No. EPA-R2-72-088) for horizontal spray scrub-
bers and packed-bed scrubbers. Using the new results from the
2-4
-------
scrubbing tests performed at the rendering plant, the mass
transfer equations were checked and improved to agree with
the plant scrubbing data. The predicted residual odors should
be within the accuracy of the measurements (a factor of 2).
The cost information was also revised and expanded to agree
with current practice in construction of scrubbers. Such
preliminary cost estimates are expected to be accurate within
30%. A series of design calculations was then carried out for
typical plant conditions. The results from the computer
models provide the rendering industry with the necessary design
and cost information for selecting scrubber systems. From
these models, scrubbers can be designed to achieve a specified
reduction in odor level within the limits of test results.
2.7 Activated Carbon Alternatives
For achieving extremely low odor levels, the use of
activated carbon was evaluated. The cost of using carbon
without regeneration was estimated, and found to be more
expensive than scrubbing, even for low inlet odor levels of
100 odor units. For normal plant ventilating air with an
odor level of 100 to 1,000 odor units, the cost of carbon would
be even higher. If regional carbon regeneration plants become
available, carbon as a polishing step will be more attractive.
2.8 On-Site Generation of Sodium Hypochlorite
The economics of on-site generation were also developed.
For rendering plants currently purchasing sodium hypochlorite,
on-site generation provides an alternative. The use of chlorine
gas and caustic for preparing the alkaline sodium hypochlorite
solution is typically 15% more economical than on-site
generation in most parts of the country. Future price
increases in chlorine gas will improve the relative economics
of on-site generation.
2-5
-------
3. DESCRIPTION OF THE LABORATORY AND PLANT SCRUBBERS
Two scrubbers were used throughout this program to evalu-
ate different scrubbing liquid combinations, and to determine
the effect of various operating conditions on performance.
A counter-current packed tower designated the "laboratory
scrubber" and a horizontal spray scrubber known as the "plant
scrubber" were used throughout the test program.
3.1 Laboratory Scrubber
A three-stage laboratory scale, counter-current packed
bed scrubber was designed, constructed, and installed at the
rendering plant in Des Moines, Iowa. (The laboratory scrubber
used on the previous project comprised one stage of this sys-
tem. ) The system was designed to determine the performance
of a variety of combinations of scrubbing liquids and to pro-
vide data on mass transfer rates. The scrubber was constructed
so that all three columns could be operated in series, or the
last two columns could be operated in parallel. The experi-
mental scrubbing system with the last two stages connected
in parallel is shown in Figure 3-1. Initially, 4-ft long,
6-in. diameter glass columns were installed with 2 ft of
packing in each column. The packing used throughout this pro-
ject was 1/2-in. Intalox saddles. For the last phase of the
project, the 4-ft columns were replaced with 6-ft columns,
and the packing depth in each stage was increased to 4 ft.
The scrubbing liquids were placed in 25-gal polyethylene drums.
Polyvinyl chloride (PVC) pipe was used for both the gas and
the .liquid piping. The gas flow rate was measured by an
orifice with a manometer. Typically, 17 cfm of gas was passed
through the scrubber, and each scrubbing solution was circulated
at 0.6 gpm.
Odorous air for the laboratory scrubber was obtained from
two sources: the inlet to the plant scrubber, and the inlet
to the incinerator which handles the cooker noncondensibles
3-1
-------
Odorous
air
typically
H-28 acfm
r*—x
=X ]0rifice
meters
u
Ball vdlve
Deodorized
ai r
Fan
Centrifugal
pump (typically 0.6 gpm)
Figure 3-1
LABORATORY PACKED-BED SCRUBBER
-------
and continuous cooker shrouding off-gases. Either air source
was connected to the laboratory scrubber using 4-in. poly-
ethylene pipe. Usually, the incinerator inlet stream
(process air) ranged from 5,000 to 50,000 odor units, while
the plant scrubber inlet stream (plant ventilating air)
ranged from 100 to 1,000 odor units.
3.2 Plant Scrubber
The rendering plant at Des Moines, Iowa, has a full-scale
horizontal spray scrubber designed by Air Conditioning Corp. ,
Greensboro, North Carolina.* This scrubber is a three-stage
unit that treats approximately 67,000 acfm of plant ventilating
air. (This flow is based on actual velocity measurements.)
This type of scrubber has been installed in a number of ren-
dering plants. A drawing of a one-stage scrubber is shown
in Figure 3-2.
Approximately 275 nozzles provide a relatively coarse
spray in each stage. The nozzles are arranged in banks, with
all the nozzles facing directly into the flow of air. A sin-
gle fan is located after the first scrubbing stage. Each
stage has a mist eliminator to minimize entrainment of scrubbing
solution from one stage to the next. Separate scrubbing solu-
tion tanks are provided for each stage; however, the second
and third stage tanks are also used by another similar hori-
zontal spray scrubber located on the roof of the plant. The
second spray scrubber is only used during the hottest part
of the year to increase the flow of plant ventilation air.
The liquid piping is all made of PVC, while the shell of
n
the scrubber contains type 304 and type 316 stainless steel.
*Tradenames are mentioned for identification purposes and not
to endorse any particular product.
3-3
-------
DIFFUSER
FAN
DIFFUSER
FINAL CONTACT &
ELIMINATION SECTION
ACCESS CHAMBER
SPRAY CHAMBER
ACCESS CHAMBER
ENTRANCE DIFFUSION BAFFLES
Figure 3-2
ONE-STAGE HORIZONTAL SPRAY SCRUBBER
3-4
-------
4. DESCRIPTION OF ANALYTICAL TECHNIQUES
In order to determine the most effective scrubbing com-
bination and to evaluate the effect of different operating
parameters, a series of comparative tests was performed.
The efficiency of the scrubber had to be measured in order
to ascertain the most effective scrubbing system.
To determine scrubber performance, a quantitative method
of odor measurement based on previous work was selected.
The following two methods were used throughout this program:
odor panel measuring the subjective odor intensity
gas chromatograph measuring the concentrations
of known odorous compounds.
The gas chromatograph techniques used in the previous
project were modified for field measurements. The recovery
of transfer from the chromatographic samplers was improved
from 60% to 90%, when measured with pure compounds. The
sampling procedure and odor identification methods, which are
considered a significant advance in odor measurement, are
discussed below.
4.1 Field Sampling Procedure
Since all scrubber testing was performed at the Des Moines
rendering plant and samples were analyzed in Chicago, proce-
dures for collecting, transmitting, and analyzing odor sam-
ples were developed. Bags of odorous air for the odor
panel and gas chromatograph (GC) samples were collected for
each test. This required the development of sampling probes.
For the plant scrubber, 4-ft long sampling probes were
constructed with small ports along the entire length of the
probes. A diagram of the sampling system is presented in
Figure 4-1. Except for the inlet sample probe, a demister
was required for all probes because of the possibility of
entrained water in the gas stream. From the demister, the
4-1
-------
Vacuum
pump
Rotameter
GC
sampler
Demister
Tubing pump
Bag sample
Sampling probe
^ j ° • ^
Plant scrubber
Figure 4-1
FIELD SAMPLING SYSTEM
4-2
-------
gas stream was split into two portions. Part of the gas
passed through a peristaltic pump into the sample bag, while
the remaining gas stream passed through a GC sampler, a
rotameter, and a vacuum pump.
A timer was used to activate a common vacuum pump which
pulled the gas through all of the GC samplers for the same
period and length of time. All bag samples were collected
simultaneously with the GC samples.
In the laboratory scrubber, no special sampling probes
were required. The samples were obtained from the 1-1/2-in.
diameter PVC pipe used to pass the gas stream between the
scrubbing stages. No entrained moisture was found in the
laboratory gas streams, and so no demisters were necessary.
Bag samples were collected at all sampling points simultaneously
with the GC samples.
4.2 Sensory Odor Determination
One method of ascertaining odor removal efficiencies
utilized the odor panel. The methods of sampling and measure-
ment have been described (Dravnieks and Prokop, 1973).
Techniques for sampling and evaluating these samples are
summarized here. The samples for the odor panel were collected
in 20-liter (0.67 ft3) thick-walled polyethylene bags. Prior
to actually taking the samples, the peristaltic pumps were
started and the bags partially filled. The bags were then
emptied and the sampling begun. This initial filling and
flushing permitted odor adsorption onto the container
walls prior to the actual sampling. This conditioning tech-
nique stabilized the odor intensity of the gas samples until
the bags were analyzed.
The gas samples were transferred from the steel sampling
probe to the polyethylene bags through food-grade Tygon tubing.
The Tygon tubing also adsorbed odorous compounds, and the
initial filling and flushing of the bags conditioned the
4-3
-------
tubing. Replacement of the Tygon tubing was necessary after
two to four tests, and also after every sampling of the incin-
erator inlet stream.
The collected gas samples in the polyethylene bags were
utilized for determination of odor intensity by the dynamic
method developed by Dravnieks and Prokop (1973). The results
were expressed in terms of ED5ff , which is the number of dilu-
tions at which 50% of the panelists begin to detect odor
during the dynamic test.
4.3 Gas Chromatograph and Sampling System
In a previous project (Burgwald, et al., 1971), a gas
chromatograph procedure and sampling system were developed
for identifying the odorous components of plant air.
In the earlier project, the gas was passed through a
sampler containing 5 g of Chromasorb 102 adsorbant in a
stainless steel collector tube. The adsorbed odorants were
than transferred to a chilled tube, and then into the chroma-
tograph column. This method pre-concentrated the sample,
and resulted in a higher sensitivity o.f analysis; however,
there was a question as to the quantitative transfer ability
from the adsorbant to the chilled tube and then to the
chromatograph.
In the present project, the previous method was modified
to make the transfer operation more quantitative. A smaller
stainless steel collector tube was developed, and the sample
was transferred directly to the chromatograph by electrically
heating the walls of the collector. This method resulted in
a simpler, more controllable transfer operation, and is dis-
cussed in the following paragraphs.
4.3.1 GC Sampler Design
The first GC samplers constructed were 3/8 in. diameter
by 6 in. long. Heat transfer from the sampler walls to the
4-4
-------
packing was too slow, resulting in poor separation of the
peaks. For more rapid heating, 1/4-in. thin-wall, stainless
steel tubing 12 in. long was fabricated into samplers. These
samplers heated rapidly, but resulted in excessive pressure
drops when sampling the scrubbers. The final samplers were
made of 5/16-in. tubing, 8 in. long, with 0.010-in. thick
walls. These contained the same amount of adsorbant, 2.8 gm,
but had only half as much pressure drop as the 1/4-in.
samplers. A diagram of the GC sampler is shown in Figure 4-2.
Breakthrough tests were performed by connecting two GC
samplers in series and passing 10 I of rendering plant air
through the samplers. The second sampler connected in
series contained only small quantities of the lighter
molecular weight components (less than 2% of the quantity
found in the first sampler).
Initially, the GC samplers were packed with Chromasorb 102,
but various components from this adsorbant were found to
bleed off during the transfer operation. This resulted in
reduced senstitivty of the chromatograph. Tests performed
with Poropak Q resulted in 90% less bleed-off compared to
tests with Chromasorb 102. Thus, except for the first batch
of plant tests, Poropak Q was used in the GC samplers.
To assure removal of all measurable material before
reuse, the samplers were conditioned at 185°C in a laboratory
oven for a minimum of 12 hours while passing nitrogen through
the samplers. Periodic checks were made on the cleaned sam-
p^ers before reusing. No gas chromatograph peaks were observed
from the cleaned samplers.
4.3.2 The Gas Chromatograph System
The chromatograph used in this program was a Victoreen
two-column Model 3000-2. Initially, two 10-ft 1/8-in.
diameter columns were used. After a few tests, the 10-ft
columns were replaced with 20-ft columns to obtain better
4-5
-------
1/8-in.
S. S. tubing
5/16-in. diameter
stainless steel tubing
0.010-in. wall
Packed with 2.8 g
Poropak Q adsorbant
8 in.
1/8-in.
Swagelok connector
5/16-in. S.S. tubing
Electrical
connector
1/4 NC thread
2 in.
Figure 4-2
CHROMATOGRAPH SAMPLER
4-6
-------
separation of the many peaks found in rendering air. The
chromatograph was further modified by inserting a stainless
steel heat shield with a stainless steel cooling water coil
silvered-soldered to the shield. This modification allowed
the temperature programming to start at 30°C instead of
60°C, resulting in improved separation of the lighter molecular
weight compounds. The lower portion of the chromatograph
recorder trace was spread over a 12-min interval compared to
a 2-min interval previously.
The procedure for injecting a sample from the odorant
collector was also optimized by a trial process. The tempera-
ture of sample transfer from the GC samplers to the chromato-
graph column was critical.
A trade-off is involved in selection of the transfer
temperature. We found that at a temperature of 180°C, the
comp9unds having molecular weights up to at least C20 (as
previously identified by mass spectrograph) were eluted.
At higher temperatures than this, some bleed-off from the
sampler was produced even when no sample was adsorbed, sugges-
ting that erroneous signals would beproduced by heating to
a higher temperature than 180°C.
The procedure that evolved is as follows: (refer to
Figure 4-3) with the oven cooled to 20°C, clear the GC sampler
connector tube of air by passing helium through the sampler.
Then attach the sampler to the helium tube. Attach the elec-
trical heater leads and thermocouple controller to the sampler.
Insert the sampler exit tube into the injection port using
the Conax fitting on the port. The helium carrier gas stream
is still entering the chromatograph via a tee at the injection
port. Wait until the system becomes pressurized and the
recorder output is stable -- 5-10 minutes. Then, almost simul-
taneously do the following four operations: (1) switch the
helium carrier gas through the sampler; (2) turn off the
cooling water solenoid valve; (3) start the chromatographic
4-7
-------
Injection
port
•p-
i
oo
115 V
3-way
solenoid
valve
Chromatograph
L
S detector
_n
Sniffing
port
FID
He
Figure 4-3
CHROMATOGRAPH SAMPLING AND ANALYSIS SYSTEM
-------
temperature programmer beginning at 0°C at a rate of
4°C/min; and, (4) turn on the sampler heating system switch.
The sampler heater variac is preset so that it requires 2 min
for the sampler to reach the set point of 180°C. Continue
operating the system until the chromatograph programmer reaches
its set point of 200°C.
As a result of the procedure outlined above, the initial
pressure peak due to opening the system to connect the sampler
is completed before the sample was transferred to the column.
Furthermore, the rate of heating of the sampler is slow enough
so that changes in the volume of gas in the sampler due to
heating have little effect on chromatograph performance. The
samples are transferred to the column before the programmer
reaches the actual starting temperature of 30°C. At this tem-
perature, the column can adsorb and hold all the odorants that
are of importance (except gases such as NHa and H2S).
Several other modifications were necessary during the early
part of the program to improve the quality of the chromatograms
The line from the chromatograph to the Tracer sulfur detector
had to be heated to prevent condensation of the components
in the line. A sniffing port was also installed which aided
in the identification of the odorous components. The accuracy
of GC results was discussed in Section 2.3 above.
4.3.3 Odorant Calibration and Accuracy of GC Measurements
The identification of odorous compounds present in ren-
dering plant air was reported in a previous project (Report
No. EPA-R2-72-088). Because the chromatograph procedure
was modified in this project to improve quantitative measure-
ment, the peak positions changed, and the new peak retention
times for the odorous compounds had to be identified on the
chromatograms. This work was completed by injecting known
compounds into both chromatographs and comparing the time
required before the compound was detected by each flame
4-9
-------
ionization detector. Using this procedure, each of the key
odorous components were identified on the new chromatograms.
The peak area for different quantities of odorants was
determined by injecting known concentrations of compounds
into the chromatograph and measuring the area of the response.
Monthly, the calibration was checked by injecting known con-
centrations of different components and measuring the response.
The calibration was used to convert peak area to yg of odorant.
A computer program was used with a Hewlett-Packard desk
computer to integrate the peak areas and multiply by the
calibration factor and sample size factor. The area was read
into the computer by moving a sensor over the chart recorder
trace of the peak, after first drawing in the estimated base-
line of the peak.
Calibration runs with known amounts of pure compounds
were made at various times during the development of the
chromatographic procedure. In one case, 13 determinations
were made using n-propyl sulfide samples, giving a mean
calibration of 704 in. sq. of recorder chart per yg n-propyl
sulfide, with a standard deviation of 203, or 29% of the
mean. One another day, 5 determinations were done with n-propyl
sulfide giving a mean of 1,788 and a standard deviation of
105; and 3 determinations were done with amyl alcohol giving
a mean of 1,261 and a standard deviation of 32. In this
project, we are concerned only with the relative error, since
in each experiment, we are comparing the percent reduction
of the outlet versus the inlet concentration to the scrubber.
The standard deviation of 29% can therefore be taken as a
measure of the relative accuracy that can be expected from
these measurements.
Rendering odors comprise a complex mixture of compounds
that cannot be completely resolved into peaks with a reasonable
amount of effort. For example, the sulfur detector sometimes showed
4-10
-------
a peak from a sulfur compound, while the flame ionization
detector showed only background signal or a peak that was
partially obscured by peaks from other compounds. The success
of making chromatographic measurements on rendering odors de-
pended on being able to identify selected peaks in the chro-
matogram, even when the relative intensity of the various
peaks changed from sample to sample, and even when the elu-
tion time of the peaks changed by as much as a minute from
sample to sample. Without the sulfur detector, it would have
been impossible to identify the peaks in this way; with the
detector, there were sometimes ambiguities, but the smaller
number of sulfur peaks made identification easier. The only
measure we have of the accuracy of the rendering odor measure-
ments made with the chromatograph is the apparent consistency
of the overall results of the scrubbing experiments. This
can be judged by examining the tables of results, but we do
not have any independent measure of accuracy other than the
calibration tests with pure compounds.
In spite of these limitations, the chromatographic results
were valuable as a check on the odor panel data, and especially
to explain some of the results of the scrubbing experiments.
This is discussed below, in the presentation of experimental
scrubber results.
4.4 Analyzing the Gas Chromatograms
Up to 16 peaks were identified from each chromatogram
and the concentration of each determined. Each odorous com-
pound contributes to the over-all odor level, but ascertaining
each individual contribution is a monumental task. Ideally,
the odor contributions from each compound could be added
together to ascertain the total odor level; however, only
9 of the 16 compounds have been identified and the effect of
interaction between the compounds on the over-all odor level
4-11
-------
is also not known. Thus, it was not possible to assign a
fraction of the over-all odor level to each of the compounds
present. If one assumes that the threshold odor level for
each compound is approximately at the same concentration, then
the inlet and outlet concentrations could be added and a single
percent removal determined; however, the threshold odor levels
vary by more than a factor of 1,000, and this method would
result in the two or three highest concentration compounds
determining the entire reduction in odor level. In other
words, the low concentration, extremely odorous compounds,
would have no influence on the calculated removals.
In order to provide equal weight to each of the odorous
compounds, the removal of each compound was determined. Then,
all of the individual percent removals were averaged to pro-
vide a single average percent removal of odorous compounds.
Ideally, in assessing the effect of compound removal on
odor removal, one should weigh each compound by its contribu-
tion to odors. We know from the previous project which peaks
are correlated with odor panel results (Report No. EPA-R2-72-
088), but the correlation data are not quantitative. Therefore,
we applied a weight of 1 to each compound when averaging the
removals.
4-12
-------
5. ODOR CONTRIBUTION OF PARTICULATES
The plant air was tested to determine the odor contribu-
tion of the particulates present in the air. A preliminary
test was run to demonstrate the equipment and procedure. The
results from the preliminary test indicated that particulates
contributed less than 870 of the total odor present in rendering
plant air.
For the principal test, two Millipore filters were attached
to probes positioned directly in the gas flow at the plant
scrubber inlet. Two additional Millipore filters were also
used to sample the outlet from the plant scrubber. The inlet
to the plant scrubber contained 1,600 ug of particulates per
liter of air. The outlet sample increased in weight by
3,300 ug/liter of air sampled. Presumably the outlet sample
contained considerable entrained scrubbing liquid, as the
samples were not dried before weighing. This accounted for
the weight increase.
To determine the odors associated with the particulates,
the second set of Millipore filters were placed in polyethylene
bags. The bags were then inflated with nitrogen gas from a
lecture bottle. The use of nitrogen eliminated the possibility
of obtaining an odorous air source in filling the bags. The
odor panel found only one odor unit in the inlet particulate
sample, as compared to a scrubber inlet air odor level of
150-360 odor units. The outlet particulate sample contained
15 odor units, as compared to an outlet air odor level of
35-50 odor units.
The very low odor level of the particulate inlet sample
indicates that only a very small amount of the odor can be
attributed to the particulates. The higher outlet particulate
odor levels strongly suggest that the scrubbing solution drop-
lets trapped on the filter impart some odor. This odor could
be due to both sodium hypochlorite and odorous compounds that
5-1
-------
had accumulated in the scrubbing solution. More discussion
of this outlet odor level is presented in Chapter 9.
6. EVALUATION OF SCRUBBING LIQUIDS
In order to optimize a wet scrubbing system, the most
effective combination of scrubbing liquids must be found.
Comparative scrubbing tests were performed using various
liquids in both the laboratory packed tower and the plant
spray scrubber. Considerable preliminary work in this area
was accomplished in an earlier project (Report No. EPA-R2-72-
088). The previous work is summarized, and the tests performed
using actual rendering plant air are described below.
The objective of this part of the work was not to quan-
titatively establish the performance of each scrubbing liquid,
but to compare them and determine which liquids merited fur-
ther investigation on a more complete basis. "More complete"
means repeating the experiments a number of times to establish
a statistical measure of variation.
6.1 Previous Studies
The preliminary work (EPA-R2-72-088) used two types of
scrubbing reaction studies with known odorous compounds.
First, exploratory bubbler experiments were performed to deter-
mine the reagent capacity to consume the odorants. . Second,
a packed scrubbing column was used to measure the rate of
reaction of the reagent with different odorants.
The previous laboratory experiments indicated that the
following strong oxidizers are capable of reacting with those
odorants which contain reactive functional groups: sodium
hypochlorite, hydrogen peroxide, potassium permanganate,
and sodium persulfate. For example, sulfides, aldehydes,
and amines reacted with these oxidants. Quantitative results
of the experiments are given in Report No. EPA-R2-72-088.
The degree of removal of course depends on the configuration
6-1
-------
of the scrubber, as well as the reactivity of the reagent.
Highly soluble odorants such as acids were removed by water.
6.2 Initial Evaluation of Scrubbing Combinations
Because of the results from the previous project, most
of the current test work concentrated on combinations using
strong oxidizers. A total of six sifferent scrubbing combina-
tions were evaluated at the rendering plant on plant ventila-
tion air. A description of each combination is presented in
Table 6-1.
Table 6-1
SCRUBBING LIQUID COMBINATIONS
Combination First Stage
A NaOH-pH 11
B NaOH-pH 11
C NaOH-pH 12
+ NaOCl
D NaOH-pH >11
E NaOH-pH 12
+ NaOCl
F NaOH-pH 12
1% NaOCl
Second Stage
NaOH-pH 12
NaOH-pH 12
+ 1% NaOCl
NaOH-pH 12
+ NaOCl
NaOH-pH 12
+ NaOCl
NaOH-pH 12
+ NaOCl
NaOH-pH 12
Third Stage
C12, pH <7
1% HzSO*
+ 1% H202
NaOH-pH 12
+ NaOCl
NaOH-pH 12
+ NaOCl
NaOH-pH 12
1.5% KMnO,,
A list of comparative scrubbing tests using the plant
horizontal spray scrubber is presented in Table 6-2. Because
of potential corrosion problems with the aluminum fan after
the first stage, the number of combinations that could be
evaluated using the plant scrubber was limited, until the
final test,when the fan was coated with plastic.
6-2
-------
ON
I
to
Table 6-2
PLANT SCRUBBER COMPARATIVE SCRUBBING TESTS
Combination
A
Caustic
Caustic
Chlorine
B
Caustic
Hypochlorite
Peroxide
D
Caustic
Hypochlorite
Hypochlorite
Test No.
2
3
5
6
8
12
13
15
16
17
Date
9/18/73
9/18/73
10/04/73
10/04/73
10/23/73
11/15/73
11/15/73
11/19/73
11/19/73
11/20/73
Inlet
ED50
330
90
25
30
320
360
150
660
300
-
Outlet
EDso
150
50
20
20
20
35
50
25
70
-
Odor panel
% removal
55
44
20
33
94
90
67
96
77
-
GC
7o removal
-
-
-
--
™
88
-
-
70
71
-------
The rendering plant was using combination A when this
program began. This combination was included in the tests to
provide a comparison with future test work. The plant tests
found combination A much less effective than combination D
(removals of 44 and 55% compared to 90, 67, 96, and 77% in
Table 6-2). The results from combination B using the plant
scrubber were inconclusive because the inlet odor levels in
Tests 5 and 6 were too low to obtain a good comparison.
Test 8, however, gave a removal of 9470 with combination B.
Because of the low odor levels frequently experienced
with ventilation air in the plant scrubbers, most of the com-
parative scrubbing tests were performed using the laboratory
packed bed scrubber. The laboratory scrubber could be fed
with high intensity process air normally going to an incinera-
tor. By using this odor source, the experiments could be
continued during the cooler months of the year, and still have
adequate odor levels to test the scrubber system (at least
300 odor units). The test results from the laboratory scrub-
ber are presented in Table 6-3. As with the plant scrubber
tests, combination A was found less effective in reducing the
odor intensity (removals of 54 and 64%,). Combinations B, C,
E, and F all gave removals in the range 98 to 99.9%,, except
for one test where the inlet odor was only 15 odor units.
The results in one test with combination D were unusually
low (21%, although the chromatograph showed 81% in one test).
Based upon the initial comparative tests, three different
combinations (B, C, and E) were found to be equally satisfac-
tory. Because of the variable inlet odor levels, the results
did not allow a ranking of the scrubbing liquid effectiveness
of these three combinations. Thus, it was difficult to deter-
mine the optimum combination. This experimental problem was
solved by developing a parallel scrubbing test procedure, which
provided a direct comparison of reagents.
6-4
-------
Table 6-3
LABORATORY SCRUBBER COMPARATIVE SCRUBBING TESTS
Combination
A
Caustic
Caustic
Chlorine
B
Caustic
Hypochlorite
Peroxide
C
Hypochlorite
Hypochlorite
Hypochlorite
D
Caustic
Hypochlorite
Hypochlorite
E
Hypochlorite
Hypochlorite
Caustic
F
Hypochlorite
Caustic
Test No.
2
3
5
7
10
11
25
26
12
21
23
24
28
29
Date
9/18/73
9/18/73
10/04/73
10/04/73
10/23/73
10/23/73
1/21/74
1/21/74
11/15/73
12/15/73
1/21/74
1/21/74
1/21/74
1/21/74
Inlet
ED50
330
65
15
17,000
34,000
63,000
8,300
190
11,500
12,500
4,500
4,300
Outlet
ED50
120
30
10
45
460
550
20
150
20
10
75
120
. Odor panel
% removal
64
54
33
99.7
98.6
99.1
99.8
21
99.8
99.9
98.3
97
GC
7o removal
-
98
96
81
97
94
67
Permanganate
-------
6.3 Parallel Scrubber Life Tests
In order to obtain direct comparisons between different
scrubbing liquids, the laboratory scrubber was modified so
that the last two stages were operated in parallel, as shown
below:
28 cfm
1st Stage
2nd Stage
3rd Stage
14 cfm
14 cfm
By operating with the last two stages in parallel, fur-
ther evaluation of three-stage scrubber performance was not
possible; however, without the direct comparison tests, the
optimum combination could not be conclusively determined.
All three combinations had demonstrated the capability to
remove odorous compounds when the scrubbing solutions were
fresh. By testing the scrubber performance under extended
operation, an indication of the rate of oxidation, the holding
capacity of the solutions for refractory compounds, and the
potential odor from the scrubbing liquor (especially hypo-
chlorite) could be ascertained.
In the first parallel scrubbing experiment, Test 31,
combinations B and C were compared. The scrubbing test was
performed over a 10-hr period to get a good indication of the
performance and life of the two systems. The incinerator in-
let air was used for the odorous source, but the odor level
was considerably lower than anticipated. Over the 10-hr
test, the hypochlorite concentration in the first stage only
dropped from 1.17» to 0.99%. No drop was observed in the
hypochlorite level or the hydrogen peroxide level in the final
scrubbing stages.
The results from Test 31 are presented in the Appendix,
Tables A-l through A-5, and are summarized in Table 6-4. The
6-6
-------
11st STAGE
NaOCl
-»
2nd STAGE
NaOCl
3rd STAGE
H202
Table 6-4
SUMMARY OF SCRUBBING TEST 31
LABORATORY SCRUBBER
1 •*• 2 stage
Sample round*
Odor ED 50 inlet
outlet
Odor ED .
GLC peaks
1. Dimethyl sulfide
2. Pungent U.P. 90
3. Dimethyl dlsulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9. Rendering 1335
10. Fatty 1.545
11. Burnt 1687 (quinoline)
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
Average percent removal of
each component
1
700
25
96
94
98
97
92
95
98
97
86
93
94
96
-
94
92
94
95
94
2
-
-
90
97
76
86
89
87
71
-
91
90
82
-
80
79
90
89
86
3
15,500
45
% removal
99.7
56
84
95
63
82
90
-
-
78
58
-
-
-
33
76
70
71
4
590
45
92
93
95
94
71
-
96
74
78
87
75
60
-
59
44
-
-
77
Avg
96
83
94
90
78
89
93
81
82
87
79
79
-
78
62
87
85
83
1 -»• 3 stage
1
700
265
2
-
3
15,500
370
4
590
175
Avg
-
% removal
62
82
7.9
73
84
86
70
70
63
85
93
55
-
82
84
85
91
79
-
97
89
94
99.0
99.6
76
94
-
97
98
99.3
-
98.8
99.2
97
98.8
95
98
76
56
86
81
82
73
-
-
84
75
-
-
-
75
84
68
76
70
90
86
99.1
88
-
78
64
89
96
96
97
-
94
93
-
-
89
77
86
78
88
88
89
74
76
76
90
90
84
-
92
88
89
86
85
* Sample rounds 1 to 4 were collected at the following times: 1/4, 1.5, 3, and 9 hr, respectively.
-------
chromatograph results indicate that the two scrubbing combina-
tions are about equally effective. The H202 combination re-
moved an average of 85% of each of the 15 odorous compounds,
while the two-stage hypochlorite system removed an average
of 83% of each of the compounds; however, the odor panel
found the hypochlorite combination much more effective. The
odor panel found a 96% average reduction in odor level in
the hypohclorite combination as compared to only 77% reduction
with the H202 combination. A careful review of the chromato-
grams found no new compounds or unusually large unidentified
peaks. Tables A-l through A-4 clearly show that the hypochor-
ite combination is much more effective in removing two com-
pounds , Pungent U.P. 90 and Bitter 655. The high levels of
these two compounds in the outlet from the H202 combination
could explain the higher odor levels observed.
Another possible explanation for the results might be
that one solution allows certain refractory compounds to build
up in the scrubbing liquor. To test this idea, fresh outside
air was passed through the scrubbing system after the 10 hours
of operation. The results from this test are presented in
Table A-5. Prior to running this portion of the test, all the
tubing for obtaining the bag samples was replaced with fresh
tubing to minimize residual odors. The odor panel found 220
odor units coming from the H202 solution as compared to only
25 odor units coming from the hypochlorite solution. The
chromatograms, however, indicated the quantity of most of the
compounds bleeding off the hypochlorite solutions was greater
than off the H202 solution, except for two compounds, Pungent
U.P. 90 and Bitter 655. The higher odor level in the outlet
from the H202 stage is probably attributable to these compounds,
which are not effectively removed by the H202, or the formation
of a new compound between the H202 and one of the odorous
compounds which is not shown on the chromatograms. The test
i - - -
demonstrated that two stages of sodium hypochlorite with
6-8
-------
excess caustic is more effective than a stage of sodium hypo-
chlorite followed by a stage of hydrogen peroxide with sulfuric
acid. The use of two separate scrubbing reagents would increase
labor costs for handling the solutions. Furthermore, hydrogen
peroxide costs $0.26/lb (100%, 1973 basis), while chlorine
costs $0.11/lb (1974 basis). Both materials can be hazardous
to handle, but the use of one reagent limits the hazards to
those of the one reagent; furthermore, chlorine is already in
routine use in rendering plants. For these reasons, as well as
the performance results, we decided not to conduct further
experiments with hydrogen peroxide.
In the last experiment, Test 32, two stages of 1% sodium
hypochlorite was compared to a stage of 1% sodium hypochlorite
followed by a stage of 1% caustic. Prior to the addition of
odorous air, clean air was pulled through the system and bag
samples were collected on the outlets. The odor panel found
very low odor levels, 12 and 10 odor units respectively from
stages 2 and 3. This indicates there is very little residual
odor in the piping and that little if any odor can be attributable
to fresh scrubbing solutions.
Odorous air from the incinerator inlet was passed through
the scrubber for a 24-hr period. The results are presented
in Appendix Tables A-6 through A-10. Table 6-5 summarizes the
results. The odor panel found an average of 97% reduction in
odor level from the hypohclorite combination, as compared to
an 89% reduction from the caustic combination. The chromatograms
also indicated that the hypochlorite combination was more effec-
tive, showing an average component removal through the hypo-
chlorite combination of 85% compared to a 71% removal in the
caustic system. The multiple stages of sodium hypochlorite
was found to be the most effective scrubbing combination.
From the chromatograms, it appeared that the caustic system
was becoming saturated with certain compounds toward the end of
the run. The removal efficiency decreased 15-20%, while no
6-9
-------
I
U-l
o
,-* 2
NaOCl ^ 3
Sample round*
Odor EDso Inlet
Outlet
Odor ED
LC peaks
1. Dimethyl sulfide
2. Pungent U.P. 90
nd STAGE ^
NaOCl Table 6-5
rd STAGE SUMMARY OF SCRUBBING TEST
NaOH > LABORATORY SCRUBBER
1 •* 2 Stage
2345 6 Avg
21,500 2700 5000 19,000 16,000
15 35 450 300 240
% removal
99.9 98.7 91 98.4 98.5 97
92 87 84 87 89 88
95 87 98 93 94 93
3. Dimethyl disulfide 92 99.4 99.9 95 91 95
4. Butyric acid
39 56 98.5 94 87 75
5. Propylene sulfide (cabbage 880) 41 93 95 72 88 78
6. Bitter 655
96 97 98 93 - 96
7. Burnt 1233A (pyrazine) 97 92 99.5 92 - 95
8. Burnt 1233B (pyrazine B) - 75 81 98 - 85
9. Rendering 1335
0. Fatty 1545
34 82 91 82 85 75
62 83 97 81 73 79
1. Burnt 1687 (quinoline) 68 74 98 87 92 84
2. Sulfur 660
3. Cheesy 2002
4. Unpleasant 2582
5. Fatty acid 1980
6. Fatty acid 2160
_ _
61 73 94 79 81 78
69 78 96 76 79 79
59 91 97 70 88 81
58 87 98.4 96 98.2 87
Average percent removal of
each component 69 84 95 86 87 85
32
1 -»• 3 Stage
2345 6 Avg
21,500 2700 5000 19,000 16,000
35 600 1400 370 360
% removal
99.8 78 72 98.1 98 89
57 68 64 3 37 46
50 26 21 3 10 22
82 96 96 17 39 66
70 82 76 84 7 64
82 95 95 64 92 85
57 39 43 2 - 35
34 26 75 4 - 35
82 80 94 - 85
89 90 94 86 75 87
89 89 95 79 75 86
98 90 98 90 87 93
_ _
93 90 94 83 78 88
98 94 96 85 78 90
99 95 95 81 88 92
90 88 98.9 85 98 92
78 77 81 57 64 71
* Sample rounds 2 to 6 were collected at the following times: 0.5, 2.5, 8.5, 17.5, and 24 hr, respectively.
-------
deterioration in removal efficiency in the hypochlorite system
was evident. An example of this is dimethyl disulfide. Caustic,
when fresh, removed 82-96% of dimethyl difulfide, but with time
the removal efficiency dropped to 17-39%.
During the experiment, caustic was periodically added to
all three stages to maintain a pH above 12. In the sodium
hypochlorite stages, maintaining the pH above 12 is critical for
optimum odor removal, especially in the final stage of the
scrubber. Earlier in this project, an experiment (Test 22) was
performed placing the sodium hypochlorite solution that had been
used for several hours in the first stage into the last stage.
In Test 22, the odor panel found only a 17% reduction in odor
removal through the entire three stages, while the chromatograms
indicated an average removal of 867o. The results of Test 22
suggested that chlorine was being emitted from the solution. At
the time, we were not set up to determine pH and can only hypo-
thesize that the pH had dropped well below 12, resulting in a
chlorine odor; however, this last test (Test 32) clearly demon-
strates that the sodium hypochlorite solutions are effective
over long periods of time as long as the pH is maintained above
12. In Test 32, the pH of the second stage hypochlorite-caustic
solution was initially 12.8, and it was 12.2 at 9 hr and 25 hr,
with 6 Ib of NaOH being added over the duration of the test.
Table A-11 presents the results of blowing fresh air
through the scrubber after the 24-hr test. These results show
excellent agreement with the previous findings. More material
was being emitted from the caustic solution, especially the
following compounds:
Pungent U.P. 90
Dimethyl disulfide
Bitter 655
Sulfur 660
Unpleasant 2582
6-11
-------
As with the previous test, all tubing was replaced prior
to taking the "fresh" air samples. The odor panel found 130 odor
units in the outlet from the caustic stage as compared to 30
odor units from the hypochlorite stage. These results indicate
that the caustic solution was more saturated with odorous com-
pounds than was the hypochlorite. This is similar to the con-
clusions from Table 6-5, where the average percent removal of
components through the caustic combination was 57 and 64 during
the last two samplings compared to an average of 79 removal during
the first 3 samplings.
Thus, based upon the parallel scrubbing tests, multiple
stages of sodium hypochlorite with excess caustic was found to
be the optimum scrubbing combination. Prior to this project,
it was believed that using different scrubbing solutions in each
stage would be the most efficient method of odor control. The
findings from this phase of the project improves the economics
of scrubbing systems, because one relatively inexpensive solu-
tion can be used in all stages. The use of one scrubbing solu-
tion also allows a plant to utilize a counter-current chemical
blow-down and make-up scheme, which reduces the chemical costs
even further. The continuous blow-down and make-up scheme is
described in detail in Section 10.3.
6-12
-------
7. EFFECT OF SODIUM HYPOCHLORITE CONCENTRATION ON ODOR REMOVAL
As discussed in the previous chapter, sodium hypochlorite
solution at a pH greater than 12 was found to be the most
effective scrubbing liquid. In order to properly design the
"optimum" scrubbing system, the effect of concentration on
odor removal must be established.
In all of the analytical work during this project, hypo-
chlorite concentration was expressed as wt percent equivalent
available chlorine (C12) based on the following reaction for
producing the sodium hypochlorite (Palin, 1973).
C12 + 2NaOH - NaOCl + NaCl + H20
Murthy (1973) studied the effect of reagent concentration
on scrubbing efficiency of specific odorous compounds. Murthy
used reagent concentrations from 1 to 5 wt%. In general, no
measurable difference in removal efficiency was observed by
increasing the reagent concentration from 1 to 5%. Based upon
Murthy*s findings and discussions with chlorine manufacturers,
it was postulated that a critical minimum reagent concentration
exists, above which a negligible increase in removal efficiency
occurs.
During the comparative scrubbing tests, the reagent con-
centrations were usually near 1 wt70 concentration; however, some
indication of the effect of concentration can be obtained from
the longer test runs in which the chlorine level was allowed
to decrease with time. In Test 30, two stages of the labora-
tory scrubber were used with sodium hypochlorite in the first
stage and caustic in the second stage. Samples were collected
before and after the first stage to monitor the performance
of the sodium hypochlorite stage. The results from Test 30
are presented in Table 7-1. This table indicates that there
was no decrease in removal efficiency, even though the chlorine
level dropped from 0.81% to 0.65%.
7-1
-------
Table 7-1
TEST 30 - SUMMARY
Inlet
NaOCl
•s
s
K
NaOH
Outlet
Time
(min)
30
90
210
% C12 in
1st Stage
0.81
0.78
0.65
Odor Concentration, ED50
Inlet (a)
3,550
10,500
14,200
Position (b)
590
465
Position (c)
60
60
% Removal
1st Stage
94
97
2nd Stage
90
87
In Test 32, two stages of sodium hypochlorite were opera-
ted in series for 25 hours. During the 25 hours, the chlorine
level in the first stage decreased 827o, while the chlorine
level in the second stage decreased 1470. The odor removal
through the two-stage sodium hypochlorite scrubber in test 32
showed no trend with time. The odor removal averaged 8470
for the entire test, while the last 2 measurements were 86
and 877o. The results from Test 32, which are summarized in
Table 7-2, clearly show the efficiency of the scrubber is
not controlled by the chlorine concentration maintained in
the scrubbing solutions; at least above some critical minimum
concentration.
In order to maintain the reagent concentration above the
critical minimum concentration, reagents must be added to the
solution, either on a batch or continuous basis. To determine
reagent feed rates, it is necessary to obtain data on the
chemical consumption rate. Because of the large variation
in odor intensity in a rendering plant (a tenfold change in
odor level in both the plant ventilating air and in the high-
intensity odor stream was observed in samples collected in
less than one hour apart), there is also a large variation in the
chlorine consumption rate. If the chlorine level is maintained
7-2
-------
I
u>
Table 7-2
TEST 32 - SUMMARY OF RESULTS
Time,
hr
0.5
2.5
8.5
17.5
24
Inlet >
70 C12 in
1st stage
0.93
0.88
0.71
0.36
0.17
7o C12 in
2nd stage
0.94
0.92
0.87
0.86
0.81
DC1 *• Outlet
70 removal
odor panel ED50
99.9
98.7
91
98.4
98.5
GC
average component
7o removal
69
84
95
86
87
-------
at too low a concentration, a high odor level in the scrubber
inlet for several hours would deplete the available chlorine.
If the chlorine is completely exhausted, drastically reduced
odor removal efficiencies can be expected. This may occur at
a time when maximum efficiency is required because of the
higher than average inlet odor levels. Thus, even though a
scrubber will operate efficiently at low chlorine levels, the
solutions must be maintained with sufficient chlorine to allow
for fluctuations in odor levels.
If the scrubber has multiple stages, a lower chlorine
concentration can be maintained in the first stage. If the
chlorine level drops to zero in a one-stage scrubber, the
resulting drop in efficiency would be drastic. A two-stage
system, however, would still have chlorine in the second stage,
so that the over-all efficiency would not be reduced as much
as with a one-stage system.
The lowest available chlorine level used in the experi
ments was 0.17%. We do not know how effective a lower chlorine
level would be although we suspect that the removal will
decrease less than proportional to concentration. However,
one should design a plant scrubber for a concentration greater
than the minimum, since fluctuations in plant odor load may
cause a temporary reduction in actual chlorine concentration in
the scrubbing liquor. Further discussion of the design concen-
trations and the cost of chemicals to achieve given chlorine
levels is given in Section 10.3.
7-4
-------
8. PERFORMANCE EVALUATION OF THE HORIZONTAL SPRAY SCRUBBER
Alkaline sodium hypochlorite solution was found to be
the most effective scrubbing solution. Using the hypochlorite
solution, a two-week evaluation of the full-scale horizontal
spray scrubber was performed. This evaluation should provide
a good indication of the efficiency of spray scrubbers, 30
of which are being used by rendering plants for treating
ventilating air.
8.1 Converting to a Continuous Blow-Down, Make-Up System
The current industry practice is to operate the liquid
scrubbing tanks on a batch basis. The spent solutions are
ordinarily dumped to the sewer every day or two, and fresh
scrubbing solutions are prepared. This practice requires con-
siderable labor and also results in the loss of significant
quantities of reagents to the sewer, as much as 100 pounds
per day as C12. Rapid dumping of the tanks to the sewer
can result in excessive pollutant peaks in the sewer. As
the batch solutions age, more refractory odorants and par-
tially oxidized odorants can accumulate in the solution, and
the scrubber odor removal efficiency can decrease.
The use of the same scrubbing solution in all stages lends
itself to a continuous blow-down and make-up system. There
are several advantages to the continuous blow-down approach:
reduced labor cost, because tanks do not have to
be emptied and cleaned every day or two
more consistent scrubber performance. The solutions
are kept fresh enough so that scrubber efficiency
approaches the maximum.
reduced chemical costs. In a multiple system, a
high degree of chemical utilization is realized;
very little is wasted to the sewer in the nearly
spent purge stream.
8-1
-------
reduced pollutant surges to the sewer. With the
continuous blow-down, a steady 2-4 gpm is sent
to the sewer. The load placed on the sewer from
this small stream is negligible compared to other
plant discharges to the sewer.
Because of the benefits described above, the scrubbing
system was modified to a continuous blow-down, make-up
system. In designing this system, the objectives were to
automatically supply the proper quantities of reagents when-
ever the scrubber was operating, and to do this without
operator intervention. The basic idea was to meter the
reagents into the third stage, allow the solution to over-
flow by gravity to the second stage, and then to the first
stage. The overflow from the first stage (blow-down) was
directed to the sewer. A diagram of the system is shown in
Figure 8-1.
8.2 Control System and Feed Rates
The water system was modified by adding a solenoid valve
and rotameter into the water line for the third-stage liquid
tank. The solenoid valve was open whenever the scrubber fan
was in operation. PVC pipe was used in modifying the water
make-up line and for all of the overflow piping.
Chlorine gas was introduced through an eductor, into the
third stage liquid pump. Caustic solution was also pumped
into the last stage nine minutes out of every hour while the
scrubber was operating. The intermittent feeding allowed
setting an easily controlled flow rate. This was accomplished
by using a 60-minute cycle timer that was activated whenever
the scrubber fan was in operation. The timer was wired to
a barrel pump which was used to transfer the caustic directly
from the 55-gallon drums to the third-stage liquid tank. This
avoided any manual handling of caustic solution. Samples
from each scrubbing solution tank were collected for pH
determination a minimum of twice.per day. The intermittent
8-2
-------
Spray scrubber
1 t
Odor sources
00
i
U)
(
w—
[^
>
-
•
Tank 1
»? -
*i
Fan
Liquor
bleed -
(
Stage 2
>
i
Tank 2
' VC11U
i deod
f air
— - i
Liquor
l_bleed f
(
j
Stage :
,
Tank 3 make-up
Spent liquor
to sewer
Figure 8-1
DIAGRAM OF PLANT SCRUBBER SYSTEM
WITH CONTINUOUS COUNTERCURRENT REAGENT SUPPLY
-------
caustic feeding rate was adjusted to maintain the pH above
a target level of 12.
8.3 Description of Plant Test
The plant test was performed from July 8 through July 18.
The temperature in Des Moines through the entire two-week
period averaged over 90°F, with a range of 70 to 105°F. The
horizontal spray scrubber was operated 24 hours a day except
for the week-end, when the scrubber was shut down for 46 hours.
The continuous cooker was in operation an average of 16-1/2 hours
each day during the test period.
The high temperatures resulted in an evaporation loss
from the scrubbing solutions that averaged 2.7 gpm, based on
the measured make-up water added to the third stage less the
blow-down from the first stage. The evaporation rate followed
a diurnal pattern, being considerably higher during the after-
noon than it was at night. Because of the varying evaporation
•rate, the blow-down rate varied from less than 0.1 gpm in the
afternoons to 4 gpm very early in the morning. The odor levels
were the highest during the early morning when the plant
processes the "worst" material. Thus, the diurnal pattern
was actually advantageous because during the period of highest
odor levels, the blow-down was at a maximum rate (4-6 gpm) .
Some foaming occurred in all three liquid tanks, although
excessive foaming only occurred when the pH was greater than
12.5. The foam in the first stage contained many dead insects
and probably a large amount of the organics removed from the
air. With the overflow-type blow-down system, the foam was
continuously being skimmed off the first stage tank. This
prevented large quantities of dead insects from accumulating
in the first stage.
It was also found necessary to drain 10 to 20 gallons
off the bottom of the first tank each day because of the
heavier particulates that were settling on the bottom of the
8-4
-------
tank. As a routine procedure, 10-20 gallons were drained off
the bottom of each tank each day, although the first tank con-
tained by far the most sediment. Routine draining of 10-20
gallons should extend the operating periods between cleaning
of the nozzles.
The chlorine concentration in each tank was measured
several times each day by titration. After the first two days
of operating, a 14 Ib/hr chlorine rotameter was installed in
the chlorine make-up line because the original 8 Ib/hr rotameter
was too small. The chlorine concentrations in each tank
throughout the two-week test are presented in Chapter 10,
Figure 10-3. Most of the chlorine demand occurs in the first
stage of the scrubber where the greatest quantity of odorous
compounds are removed. The large chlorine demand in the first
stage is evident from Figure 10-3, where there is very little
difference in the chlorine concentration between tanks 2 and
3, but there is a large drop in the chlorine level from tank 2
to tank 1. Also from Figure 10-3, a large fluctuation in the
chlorine level in tanks 2 and 3 is evident. This is caused by
the varying odor loading on the scrubber and the varying blow-
down rate.
The pH in the first tank remained between pH 9 and 10
throughout the entire test. The pH in tank 3 varied from pH
9.5 to 12.5; however, throughout most of the test the pH in
tank 3 remained between pH 11 and 12 which was lower than the
target value of pH 12. This was due to the malfunctioning
of a pH meter that was brought to the plant.
In order to check the build-up in alkalinity. pH titrations
were performed six times on each solution. During the first
two days there was an increase in the alkalinity of each solu-
tion; however, after the first two days no further increases
were observed, indicating that the rate of accumulation of
material was equal to the loss of material in the blow-down.
8-5
-------
Thus, steady conditions were achieved in the continuous blow-
down and make-up system.
8.4 Results of the Performance Test^
This test using the continuous blow-down system was very
successful. The odor removal results are presented in Tables 8-2
through 8-7 and are summarized in Table 8-1. The system operated
for over 175 hours and no performance deterioration was observed
with time. Throughout the entire test the outlet odor level
averaged 57 odor units, a level that is very satisfactory for
most rendering plants.
The plant ventilating air entering the scrubber contained
odor levels from 165 to 2,500 odor units. The outlet from the
scrubber contained from 20 to 135 odor units through the test
period. For sample round #7 at 175 hours, the odor panel found
165 odor units in the inlet and 100 odor units in the outlet.
The pH in the last stage had dropped to pH 9.5 when the samples
were taken. According to the gas chromatograms, a high level
of removal of the compounds occurred. Thus, it appears that
chlorine contributed a large part of the outlet odor level
measured by the odor panel. During all the other samplings, the
pH in the last stage was above pH 10, and usually above pH 11.
This result emphasizes the importance of the high pH in the
final stage of the scrubber system.
The average removal of each component for the entire test
was 857o (standard deviation of 11%) according to the chromatograph,
while the odor level reduction averaged 92% (standard deviation
of 3%) excluding the last sampling. Throughout the entire
two-week test, no odor complaints were received by the rendering
plant from the surrounding community. The extremely hot July
temperatures and the heavy load of rancid, dead animals processed
by the plant should produce the most severe conditions that
rendering plants must operate under. The sodium hypochlorite
scurbbing solution with the continuous blow-down system was
8-6
-------
Table 0-1
SUMMARY OF SCRUBBING TEST 34
PLANT SCRUBBER
00
i
-•j
s
Sample No.
Time
Inlet odor, ED50
Outlet odor, EDso
Odor ED
GLC peaks
1. Dimethyl sulfide
2. Pungent U.P, 90
3. Dimethyl disulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9. Rendering 1335
10. Fatty 1545
11. Burnt 1687 (quinoline)
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
Average removal of each component
1st Stage
T1TT
LEVELS 2nd Stage
3-rd Stage
1
29 hr
2500
135
2
31 hr
380
25
3
77 hr
180
20
4 5
84 hr 125 hr
190 550
20 40
6
150 hr
-
7
175 hr
165
100
% Removal
95
-
-
-
93
83
-
94
99.1
99
97
97
-
97
98
79
91
93
9-10
9-10
> 10
93
49
-
-
-
40
-
81
50
-
53
89
-
67
-
-
-
61
9-10
9-10
> 10
89
58
-
-
99
99
-
90
99
89
85
86
-
92
79
93
93
88
9-10
> 10
V 10
89 93
87
-
-
94
88
_
61
85
97
97
87
-
97
82
78
96
87
9-10 9.3
> 10 9.3
> 1O 10.9
-
99.3
-
-
94
82
_
88
93
95
75
90
-
90
72
52
-
85
9.1
9.4
11.2
(39)
97
-
-
71
83
-
83
99
83
99.8
75
-
86
82
-
-
86
9.4
9.5
9.5
Average
57
92(83)
78
-
-
90
79
_
83
88
93
84
87
-
88
83
76
93
85
-------
Table 8-2
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 29 HOURS
oo
I
00
Odor EDso
GLC peaks, ug/liter air
1. Dimethyl sulfide
2. Pungent U.P. 90
3. Dimethyl disulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9. Rendering 1335
10. Fatty 1545
11, Burnt 1687 (quinoline)
12, Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
Inlet
2500
Outlet
135
% Removal
95
0.75
0.47
2.07
6.52
266.2
15.41
9.31
9.48
4.48
5.22
4.43
0
0
0
0
3
0
0
0
0
1
0
.05
.08
.13
.06
.10
.50
.30
.32
.11
.09
.39
93
83
94
99.1
98.8
97
97
97
98
79
91
Average percent removal of each component
93
-------
Table 8-3
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 31 HOURS
6. Bitter 655
Outlet % Removal
Odor ED50 380 25 93
GLC peaks, yg/liter air
1. Dimethyl sulfide 0.92 0.47 49
2. Pungent U.P. 90
3. Dimethyl disulfide -
4. Butyric acid - -
5. Propylene sulfide (cabbage 880) 0.10 0.06 40
7. Burnt 1233A (pyrazine) 0.16 0.03 81
8. Burnt 1233B (pyrazine B) 0.34 0.17 50
9. Rendering 1335 - -
10. Fatty 1545 0.45 0.21 53
11. Burnt 1687 (quinoline) 0.87 0.10 89
12. Sulfur 660 -
13. Gheesy 2002 0.21 0.07 67
14. Unpleasant 2582 - -
15. Fatty acid 1980 - -
16. Fatty acid 2160 - -
Average percent removal of each component 61
-------
Table 8-4
oo
H-»
O
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 77 HOURS
Odor ED so
Inlet
180
After first stage
65
Outlet
20
Total
% Removal
89
GLC peaks, Ug/liter air
1. Dimethyl sulfide
2. Pungent U.P. 90
3. Dimethyl disulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9. Rendering 1335
10. Fatty 1545
11, Burnt 1687 (quinoline)
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
0.76
0.07
0.32
58
0.30
0.14
0.06
0.64
24.05
2.21
0.72
1.26
0.34
1.07
1.30
0.09
0.07
0.04
0.26
8.27
0.82
0.30
0.79
0.48
0.77
1.25
< 0.01
< 0.01
< 0.01
< 0.01
2.54
0.34
0.10
0.10
0.07
0.07
0.09
99
99
90
99
89
85
86
92
79
93
93
Average percent removal of each component
88
-------
Table 8-5
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 84 HOURS
Inlet Outlet % Removal
Odor EDSo 190 20 89
GLC peaks, Mg/liter air
1. Dimethyl sulfide 1.07 0.14 87
2. Pungent U.P. 90 -
3. Dimethyl disulfide -
4. Butyric acid 0.70 0.04 94
5. Propylene sulfide (cabbage 880) 0.92 0.11 88
6. Bitter 655 -
7. Burnt 1233A (pyrazine) Q.18 0.07 61
8. Burnt 1233B (pyrazine B) 1.06 0.16 85
9, Rendering 1335 70.60 1.85 97
10. Fatty 1545 7.08 0.21 97
11. Burnt 1687 (quinoline) 3.38 0.45 87
12. Sulfur 660 -
13. Cheesy 2002 5.06 0.13 97
14. Unpleasant 2582 1.81 0.32 82
15. Fatty acid 1980 0.64 0.14 78
16. Fatty acid 2160 2.48 0.09 96
Average percent removal of each component 87
-------
Table 8-6
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 150 HOURS
Inlet Outlet % Removal
Odor EDso - -
GLC peaks, yg/liter air
1. Dimethyl sulfide 7.42 0.05 99.3
2. Pungent U.P. 90 -
3. Dimethyl disulfide -
4. Butyric acid 1.25 0.07 94
5. Propylene sulfide (cabbage 880) 0.34 0.06 82
6. Bitter 655 - -
7. Burnt 1233A (pyrazine) 0.33 0.04 88
8. Burnt 1233B (pyrazine B) 3.35 0.23 93
9. Rendering 1335 79.81 4.27 95
10. Fatty 1545 9.00 2.24 75
11, Burnt 1687 (quinoline) 1.76 0.18 90
12, Sulfur 660 - -
13. Cheesy 2002 2.41 0.25 90
14, Unpleasant 2582 1.30 0.37 72
15. Fatty acid 1980 1.94 Q.94 52
16. Fatty acid 2160 -
Average percent removal of each component 85
-------
Table 8-7
RESULTS OF SCRUBBING TEST 34
PLANT SCRUBBER, TIME: 175 HOURS
Inlet Outlet % Removal
Odor ED50 165 100 39
GLC peaks, yg/liter air
1. Dimethyl sulfide 7.22 0.20 97
2. Pungent U.P. 90 -
3. Dimethyl disulfide _
4. Butyric acid 0.24 0.07 71
5. Propylene sulfide (cabbage 880) 0.23 0.04 83
6. Bitter 655 -
oo
i 7. Burnt 1233A (pyrazine) Q.12 0.02 83
w 8. Burnt 1233B (pyrazine B) Q.93 <0.01 99
9. Rendering 1335 35.08 5.86 83
10. Fatty 1545 5.86 <0.01 99.8
11. Burnt 1687 (quinoline) 1.46 0.36 75
12. Sulfur 660 -
13. Cheesy 2002 2.10 0.29 86
14. Unpleasant 2582 0.34 0.06 82
15. Fatty acid 1980 _
16. Fatty acid 2160 _
Average percent removal of each component 86
-------
extremely successful in maintaining an acceptable quality scrub-
ber discharge (less than 150 odor units) for the entire two-
week test. This two-week test clearly demonstrated the effec-
tiveness of the continuous blow-down system in maintaining
fresh solutions and good odor reductions.
8-14
-------
9. PERFORMANCE EVALUATION OF PACKED-BED SCRUBBERS
An additional objective of this program was to evaluate
the performance of packed-bed scrubbers treating the high
odor intensity air streams, typically 5,000 to 50,000 odor
units. Evaluating the performance of a scrubber system re-
quires a determination of empirical operating characteristics
of the scrubbing unit and the liquid solution, guided by
theoretical analysis. In this section, the performance of
the scrubbing solutions, the mass transfer characteristics
of odor removal, and other design features of packed-bed
scrubbers will be discussed.
9.1 Summary of Comparative Tests
The laboratory scrubber installed at the Des Moines ren-
dering plant was used to evaluate the odor removal of packed-
bed scrubbers. Comparative scrubbing tests, which were des-
cribed in Chapter 6, were utilized to determine the optimum
combination of scrubbing solutions.
According to the test results, using a single scrubbing
reagent, sodium hypochlorite, in all stages provided the
highest degree of odor reduction for both low- and high-inten-
sity odors. Three separate evaluations of this scrubbing
reagent were performed during the initial test work, and the
results are summarized below.
Test 11
Test 25
Test 26
Odor Level (ED50 )
Inlet
63,000
8,300
Outlet
550
20
70 Removal
99.1
99.8
GLC
Average %
Removal of
Each Component
96
9-1
-------
Based upon the outstanding results shown above, further
tests were performed to directly compare the hypochlorite
combination with other promising combinations. These results,
which are also described in detail in Chapter 6, are summarized
below.
Test 31
Test 32
Length of
Test, hr
10
24
Average
Odor Levels (ED50 )
Inlet
5,600
12,800
Outlet
38
208
Average
% Removal
96
97
GLC
Average %
Removal of
Each Component
83
85
These results were obtained with a two-stage packed scrub-
ber, as compared to a three-stage scrubber used in the previ-
ous tests. Even after the solutions had been subjected to the
high intensity odors for 24 hours in Test 32, the scrubbing
solutions were still very effective. The pH was maintained
above 12 throughout Test 32, and this was considered an impor-
tant factor in achieving these high removals. Throughout
Test 32, the pH in both stages was maintained above pH 12 by
adding caustic.
As discussed in Chapter 6, maintaining the pH above 12
was found to be essential based on the results of Tests 22
and 32. In Test 22, the pH was not maintained and the odor
panel found only a 17% reduction in the odor level. The
chromatograms in Test 22 indicated an average removal of 86%,
suggesting that a major part of the odor smelled by the panel
was chlorine or the reaction products between the chlorine
and the odorous compounds. In Test 32, the pH was maintained
above pH 12 for a 24-hr test, and no deterioration in removal
was reported. The difference in these two tests emphasizes
the importance of maintaining a pH of 12 to achieve odor
removals greater than 90%.
9-2
-------
9 .2 Mass Transfer as a Function of Packing Depth
In designing a packed-bed scrubber, the two most impor-
tant parameters that must be specified are the gas flow-rate
to be treated and the desired removal efficiency. The removal
efficiency in gas-liquid absorbing systems is often written
in terms of mass transfer units, where
NTUG = An
where
NTUG is number of mass transfer units
YO is inlet odor level
YE is outlet odor level
To obtain a greater reduction in odor, the scrubber must pro-
vide more mass transfer units, (assuming that the increase
does not have an effect on any chemical reactions occurring) .
Mass transfer theory for absorbing pure compounds from the
gas to the liquid phase states that if the packing depth
doubles , the number of mass transfer units also doubles , as
long as the liquid and gas flow rates are held constant. To
ascertain if the mass transfer theory is valid for odor
removal, the height of the packing in the laboratory scrubber
was increased from 2 to 4 feet. The number of mass transfer
units for Test 33 with the taller columns are compared below
with the mass transfer units for the shorter columns (Tests 31
and 32) under similar operating conditions, using sodium
hypochlorite with the pH greater than 12.
9-3
-------
Test No.
31 & 32
33
No.
of Stages
2
1
Height of
packing
per stage
Ft
2
4
Odor
YO
10,000
9,650
Units
YE
140
277
Average
NTUG
per stage1
2.13
4.12
These data suggest that doubling the column height effec-
tively doubles the number of mass transfer units, NTUG. This
result is in agreement with mass transfer theory.
Another method of increasing the number of mass transfer
units involves a reduction of the gas flow rate. If a smaller
volume of gas is being treated, the contact time between the
gas and the liquid increases, resulting in a higher removal.
Using just one 4-ft stage of the scrubber, the gas flow rate
was reduced from the normal 17 cfm to 11.7 cfm. This experi-
ment was performed twice, and the following results were ob-
tained using just the one stage.
Test 33a
Test 33b
Gas Flow
(cfm)
11.7
11.7
Odor Level
Inlet
15,000
4,300
Outlet
20
60
?0 Removal
99.9
98.6
The results indicate that odor removals greater than 9870 can
be achieved in a single-stage packed scrubber. This conclu-
sion is based on a limited number of tests, and should be
repeated for confirmation. If confirmed, it would show that
Average NTUG =
n
Z (NTUG)
=l
n
where n = total number of samplings
See Appendix 6 for method of calculation.
9-4
-------
the selection of a single- or multi-stage scrubber system is
one of economics. The determination of the economic optimum
number of stages is discussed in Chapter 10, based on a
mass transfer model.
9.3 Laboratory Scrubber Liquor Life Test
Tests using the three-stage packed-bed scrubber resulted
in odor removals greater than 997o, as shown in Table 6-3.
However, only Tests 31 and 32 were performed for an adequate
period of time to provide an estimate of the life of the
scrubbing solutions. Both of these earlier tests were per-
formed utilizing batch scrubbing solutions. Additional
data were needed using the continuous blow-down scheme to de-
termine reagent consumption rates.
Using the high strength cooker off-gas, a five-day
test was undertaken to achieve the following objectives:
demonstrate that packed bed scrubbers could
successfully deodorize high-strength gases from
a rendering plant
establish the chemical make-up and blow-down
rates for treating high-strength gases.
The following tests suggest that when treating incinera-
tor inlet air, the blow-down rate is controlled by a different
mechanism than when treating plant ventilating air. For
plant ventilating air, the blow-down rate should be set to
maintain a chlorine residual in the first stage. For incin-
erator inlet air, the blow-down rate should be set to purge
the solutions of odorous refractory compounds, which would
otherwise accumulate. Although the concentrations of these
compounds are not known, the objective should be to find a
blow-down rate that will prevent any evidence of their
effect on the odor removal of the scrubber.
9-5
-------
9.3.1 Test Procedure
Three 6-ft glass columns were installed, each with 4 ft
of packing to provide a large number of mass transfer units.
The continuous blow-down and make-up system was simulated by
manually transferring a portion of the solution from tank to
tank every hour. The process was similar to the horizontal
spray scrubber continuous blow-down system shown in Figure 8-1.
Since all previous work had been performed using batch
scrubber solutions, the necessary blow-down rate and chemical
make-up requirements were estimated and adjusted throughout the
duration of the test, as explained below.
At the start of Test 35, the following conditions were
measured.
Tank #1
Tank #2
Tank #3
NaOCl (as C12)
wt %
0.25
0.5
0.75
pH
11.8
12.1
12.4
After 10 hours of operation, the following concentrations
were measured in each tank.
Tank #1
Tank #2
Tank #3
NaOCl (as C12)
wt %
0.02
0.45
0.74
PH
9.5
9.7
10.7
At the end of 10 hours, the make-up water and caustic addition
rates were increased. Because of the pH meter malfunction
described in the previous chapter, analysis of pH had to be
performed at IITRI during the following week. Thus, proper
control of caustic addition was not achieved during the test
9-6
-------
run. At the end of the 23rd hour of operation, a residual ran-
cid chlorine odor in all of the scrubbing solutions was very
noticeable (the odor panelists also noted a chlorine-like odor
in the outlet bag samples). so the water make-up rate was in-
creased from 1.5 gal/hr to 2.0 gal/hr and the caustic increased
from 0.1 Ib/hr to 0.14 Ib/hr. Throughout the remainder of the
test, the following conditions prevailed.
Tank #1
Tank #2
Tank #3
NaOCl (as C12)
7o
0.10-0.14
0.35-0.42
0.40-0.45
PH
9.3- 9.6
9.5- 9.8
9.9-10.6
The pH was still not maintained above pH 12 in the last
stage even when the dosage rate was increased. It was also
observed that the residual odor in all of the tank solutions
persisted for the remainder of the test, even with the increase
in the water make-up rate; however, no residual odor was evident
in the scrubbing solutions used during the full-scale plant test
for plant ventilating air described in Chapter 8. These obser-
vations, and GLC data given below support the earlier statement
that the mechanism that controls the blow-down rate, varies
depending upon the strength of the odor.
9.3.2 Laboratory Scrubber Test Results
The results of the laboratory scrubber liquor-life test
(Test 35) are presented in Tables 9-2 through 9-5 and are
summarized in Table 9-1. The gas chromatograph results show
an average removal of each component for the entire test of 88%.
The removal was expected to be considerably greater than
88% based on the previous batch tests, #31 and 32. In Tests 31
and 32, which were performed utilizing only two stages of 2-ft
packing, the odorous compounds were reduced an average of 83%
and 857o respectively. In this liquor life test, taller columns
9-7
-------
Table 9-1
vo
i
00
SUMMARY OF SCRUBBING TEST 35
LABORATORY SCRUBBER
1 2 3
Sample round
time, hrs. 9 hours 14 hours 22 hours
inlet 18
outlet 62
30 * 2750 1700
DOj 1120 1150
4
32 hours
4800
95
5 Average
48 hours
35,500
7,600
% removals
Odor ED
GLC peaks
1. Dimethyl sulfide 98
2. Pungent U.P. 90 95
3. Dimethyl disulfide 99
4. Butyric acid
5. Propylene sulfide (cabbage 880) 97
6. Bitter 655 99
7. Burnt 1233A (pyrazine) 87
8. Burnt 1233B (pyrazine B) 98.
9, Rendering 1335 36
10. Fatty 1545 67
11. Burnt 1687 (quinoline) 28
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582 73
15. Fatty acid 1980 75
16. Fatty acid 2160
59 32
2 - -
-
6 - 99.9
98.9
98.9
9-55
99
4-99
97
98
96
-
95
95
95
92
98
94
54
99.9
99
99
47
-
74
98.4
98
99.7
-
98.2
96
98.7
99.6
79 67
91 94
90 80
92 98
99
98
32 58
93
90
77
88
75
-
97
88
90
96
Average percent removal of
each component
74
94
90
76
88
Either these measurements were in error, or the samples were mislabeled.
-------
I
VO
Table 9-2
RESULTS OF SCRUBBING TEST 35-1
LABORATORY SCRUBBER, TIME: 9 HOURS
Inlet
Odor ED so
GLC
1.
2.
3.
4.
5.
6.
7.
8.
9,
10.
11.
12.
13.
14.
15.
16.
peaks, Ug/liter air
Dimethyl sulfide
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
1800
21
1
2
-
1
48
1
9
21
2
1
-
0
1
2
_
.60
.68
.68
.64
.24
.17
.28
.08
.37
.13
.42
.31
.72
Outlet
% removal
6200
0.
0.
<0
-
0.
<0.
0.
0.
13.
0.
0.
-
-
0.
0.
1.
39
08
.01
05
01
15
15
46
78
81
36
68
19
98.
95
99.
-
97
99.
87
98.
36
67
28
-
-
73
75
_
2
6
9
4
Average percent removal of each component 74
Make-up Rates
Water 4.5 liters per hour
Caustic 42 ems per- Viour- (as NnOVO
-------
I
H*
O
Table 9-3
RESULTS OF SCRUBBING TEST 35-3
LABORATORY SCRUBBER, TIME; 22 HOURS
Inlet
Odor ED so
GLC peaks, jag/liter air
1. Dimethyl sulfide
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Ave:
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
rage percent removal of each component
1700
-
15
1
3
52
0
1
207
26
6
_
6
4
11
12
.36
.85
.54
.66
.27
.27
.7
.13
.68
.78
.08
.21
.90
Outlet
1150
-
<0.
0.
0.
23.
<0.
<0.
5.
0.
0.
_
0.
0.
0.
0.
01
02
04
78
01
01
79
64
26
34
19
60
98
% removal
29
-
99
98
98
55
99
99
97
98
96
_
95
95
95
92
94
.9
.9
.9
Make-up Rates
Water 5.7 liters per hour
Caustic 52 gms per hour (as NaOH)
Chlorine 52 gms per hour (as
-------
VO
I
Table 9-4
RESULTS OF SCRUBBING TEST 35-4
LABORATORY SCRUBBER, TIME: 32 HOURS
Odor ED50
GLC peaks, ug/liter air
1. Dimethyl sulfide
2. Pungent U.P. 90
3. Dimethyl disulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9, Rendering 1335
10. Fatty 1545
11. Burnt 1687 (quinoline)
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
Average percent removal of each component
Make-up Rates
Water 7.6 liters per hour
Caustic 70 gms per hour (as NaOH)
Inlet
4800
22.10
15.37
7.38
0.83
1.11
36.99
-
1.54
80.01
9.84
2.86
-
2.78
1.09
4.67
5.09
Outlet
95
1.42
7.06
<0.01
<0.01
<0.01
19.50
-
0.40
1.30
0.22
0.01
-
0.05
0.04
0.06
0.02
% removal
98
94
54
99.9
99
99
47
-
74
98.4
98
99.7
-
98.2
96
98.7
99.6
90
-------
Table 9-5
RESULTS OF SCRUBBING TEST 35-5
LABORATORY SCRUBBER, TIME: 48 HOURS
Inlet Outlet % removal
Odor ED50 35,500 7600 79
GLC peaks, pg/liter air
1. Dimethyl sulfide 12.03 1.10 91
2. Pungent U.P. 90 22.15 2.19 90
3. Dimethyl disulfide 10.62 0.82 92
4. Butyric acid -
5. Propylene sulfide (cabbage 880) -
vo
jL 6. Bitter 655 50.47 34.14 32
7. Burnt 1233A (pyrazine) -
8. Burnt 1233B (pyrazine B) -
9. Rendering 1335 -
10. Fatty 1545 -
11. Burnt 1687 (quinoline) -
12. Sulfur 660 -
13. Cheesy 2002 -
14. Unpleasant 2582 -
15. Fatty acid 1980 -
16. Fatty acid 2160 -
Average percent removal of each component 76
Make-up Rates
Water 7.6 liters per hour
Caustic 70 gms per hour (as NaOH)
Chlorine 52 gms per hour (as C12)
-------
were utilized with 4 ft of packing. Table 9-1 shows that the
removal efficiency for Bitter 655 as well as for several other
odorous compounds decreased throughout the entire test. (For
Bitter 655, the removal decreased from an initial 99.9% to a
final removal of 32%.) This indicates that the blow-down rate
was indeed too low, which resulted in the accumulation of cer-
tain compounds in the solution. The limited removal of specific
compounds suggests that either the compounds accumulated more
rapidly than the oxidation rate, or that sodium hypochlorite
was not oxidizing these compounds at all.
The odor panel reported low removals (67% average, excluding
sample 35-1) and also reported a strong smell of chlorine in
all of the outlet bag samples. Both the chlorine smell and
poor odor level reductions can be attributed to the occurrence
of a low pH (9.9-10.7) in the last stage. When the caustic
and the blow-down rates were increased after 23 hours of opera-
tion, there was an improvement in the performance. The odor
panel found that the outlet odor levels were lower (98% and 79%
removal) for samples taken after this change.
9.3.3 Discussion
Because of the unexpected different scrubbing behavior of
incinerator inlet air compared with plant ventilating air, in-
sufficient data were collected to conclusively determine the
optimum scrubbing reagent concentrations and purge rates for
high intensity odors. Therefore, the estimates for reagent
costs for incinerator inlet are not as clearly defined. The
complex nature of odors results in a very large number of con-
trol variables (i.e., make-up rates, reagent concentrations,
liquid mass flux, gas mass flux). Additional field work is
still needed before a clear understanding of high intensity
odor removal can be obtained.
Although this test did not demonstrate that odor removal
above 90% and preferably 99% could be sustained using a continu-
ous blow-down system, a great deal of design information was
9-13
-------
obtained. It is clear now that the initial blow-down and caus-
tic rates were too low and the system never fully recovered from
the initial settings. However, the performance of the system
was improving after the blow-down and caustic rates were adjusted,
which suggests that higher rates should be further investigated.
The earlier comparative tests (#31 and #32) demonstrated
that removals of 97% could be sustained even in a batch system
for considerable lengths of time. The comparison below shows
the different results obtained.
Test 32
Test 35
Two stages, 2 ft of
packing, batch test
24-hr test, GC removal
averaged 85%. Odor panel
reported 97% removal.
pH maintained above
pH 12
three stages, 4 ft of
packing, continuous
blow-down and make-up
GC results averaged 88%
removal. Odor panel
reported 67% removal
pH level below pH 11
throughout most of the
test
The difference in efficiency between Test 32 and Test 35 must
be attributed to the difference in the pH levels maintained.
Temperature is one other factor that could have affected
the performance of the scrubbing liquids. In previous tests,
the gas and liquid temperatures were measured in the range of
55 to 70°F. In Test 35, the temperatures of gas and liquid
were measured at 90 to 100°F, as a result of the warmer ambient
temperature at the time. The effect of temperature is to change
the solubility at a given vapor pressure of the odorant compounds;
for example, increasing the temperature from 50 to 86°F reduces
the amount of ammonia that water can absorb by over 50%
(Treybal, 1967, p. 221). The effect of temperature on the
scrubbing results cannot be deduced from the data because the
pH also varied at the same time. In any future work, the
possible effect of temperature should be further considered.
9-14
-------
9.4 Conclusions
The objective of the week-long Test 35 was to verify the
performance of packed-bed scrubbers to treat high odor intensity
(5,000 to 50,000 odor units) process air streams. The test
did not successfully demonstrate the design. Instead, it
showed that we have not yet found the optimum values of pH,
reagent concentration, and purge rate in scrubbing of high-
intensity odors. The importance of maintaining the pH above
12 in the last stage was clearly demonstrated, and increasing
the purge rate above 1 gal/hr for 20 cfm air flow was also
indicated.
The scrubber design, which this test was intended to verify,
was based on the idea that removal of odorous compounds by
packed-bed scrubbers follows the same mass transfer principles
as scrubbing of pure compounds, provided that conditions can
be found such that chemical reaction rates do not limit the
scrubbing process. The necessary mass transfer data was
obtained, as discussed in Chapter 10, and includes the effect
of gas and liquid flux on the mass transfer coefficient. We
still feel that a scrubber designed on these mass-transfer
principles will perform as expected, but experimental verifica-
tion requires further measurements to determine the optimum
values of pH, reagent concentration, and purge rates in scrubbing
of high-intensity odors.
9-15
-------
10. DESIGN AND COST OF COUNTERCURRENT PACKED TOWER GAS
SCRUBBING SYSTEM
A computerized design procedure for packed towers was
developed in a previous project (Report No. EPA-R2-72-088).
This scrubber design and cost model has been updated based
on the data collected from the present project. When the
odor reduction desired and the number of stages are speci-
fied, the computer model will determine the least cost design.
A description of the model and examples of utilization are
provided within this report.
10.1 Input Data
The following list itemizes and describes the required
input data for use of the design model.
1. Type of Packing
The size and type of packing must be specified.
All of the work to date has been with Intalox saddles,
although the program is capable of handling many
types of packing.
2. Type of material of construction to be used for the
tower.
Any material can be specified. Fiberglass-reinforced
plastic has been used for all of the design calcu-
lations because it is available for commercially-
built scrubbers and is corrosion-resistant for the
recommended scrubbing liquids.
3. Cost Index (CI)
To keep the program current, all cost data that
were collected was converted to 1957-59 costs. By
utilizing the Chemical Engineering Plant Cost Index
(1957-59 = 1.00) (published in Chemical Engineering)
only one factor must be adjusted to formulate the
present cost of the installed equipment.
All cost estimates in this report are based on a
cost index of 1.75, which is the projected index
for December 1974.
10-1
-------
4. Material Cost Factor (MCF)
Depending upon the type of material of construction
specified in Item 2, a material cost factor must be
determined. The cost equations in the model use as
a basis carbon steel =1.0. Thus, the relative
fabricated cost of the material specified as compared
to carbon steel must be determined. For stainless
steel, MCF =2.25 (Popper, 1970) and for fiberglass,
MCF =1.50 (Graham, 1971; Process Equipment Corp.,
1974).
5. Maximum Liquid Flux (LMAX)
The maximum liquid rate (lb/hr-ft2) must be supplied.
LMAX is a function of the type and size of packing
used. For 1-1/2-in. Intalox Saddles, LMAX =
40,000 lb/hr--ft2.
6. Henry's Law Constant (HENRY)
Henry's law constant is defined as the partial pres-
sure of the gas divided by the mole fraction of the
solute in the liquid at equilibrium. For compounds
that are very soluble in the liquid, the mole frac-
tion in the solute is much greater than the partial
pressure at low gas concentrations; thus, Henry's
law constant approaches zero. For odor control,
HENRY can be taken equal to zero.
7. Efficiency of the Pumps (EP)
This is used to calculate the pump horsepower. An
EPA =0.5 was used throughout the entire study.
8. Efficiency of the Fan (EF)
The fan efficiency was used to calculate the fan
horsepower requirements. A value of 0.55 was chosen
for the design model.
9. Operating Hours (HRS)
HRS is the number of hours the scrubber operates
per year. A value of 4,000 hours was selected.
This is equivalent to two shifts per day, five days
per week.
10. Temperature (T) °F
The temperature of the gas must be specified. In
the examples provided, the temperature was set at
70°F.
10-2
-------
11. Cost of Electricity (CE)
A value of $0.025/KWH was assigned, based on 1974
industrial electricity rates of Commonwealth Edison
in Chicago.
12. Cost Factor for Maintenance (CM)
CM is a constant used to calculate the annual main-
tenance cost. A value of 8 for CM was used in this
study.
13. Amortization Factor (AMF)
This is based on the estimated useful life of the
scrubber. The useful life was projected to be
15 years at 10% interest on the capital investment,
or 13% per year.
14. Porosity of the Packing (e)
Porosity is the void space of a cubic foot of
packing. This is a function of the type and size
of packing, and data are available from manufac-
turers .
15. Surface Area of the Packing (Ap) ft2/ft3
This is also a function of the type and size of
packing specified.
16. Equivalent Diameter of a Particle Packing (PS) ft
This is also a function of the type and size of
packing.
17. Constants Used for Computing the Pressure Drop
(Alpha) (Be"taT
A function of the type and size of packing.
18. Installed Cost of Packing (CPA) $/ft3
19. Number of Stages in the Scrubber (NS)
20. Ratio of the Mass Transfer Rate (KGA) of the
Packing Used, Compared to the Mass Transfer Rate
of 1-1/2-in. Intalox Saddles (PKGATT
The mass transfer equation in the model was based
on 1-1/2-in. Intalox saddles. Comparative KGA
values can be obtained from the packing manufac-
turer's literature. The use of comparing mass
transfer rates for various packings in the design
of scrubbing systems is described by Norton (1969).
10-3
-------
21. Actual Volumetric Gas Flow Rate (CFM) ft3/min
22. Inlet Odor Concentration (YO) odor units
23. Outlet Concentration from the First Stage of the
Scrubber (YE) odor units
24. Increment (DELTA)
During the optimization routine, the computer model
changes the liquid mass flux by this amount.
Suggested value = 1,500
25. Tolerance (EPS)
During the optimization routine, the cost is mini-
mized until the % change in the cost for the last
two iterations is less than this amount.
10.2 Scrubber Design
10.2.1 Design Equation
Consider the packed tower shown in Figure 10-1. If a
material balance for an odorous compound is performed on
an increment of volume within the tower, the following
equation applies :
-G^dY = L^dX = KGa(P - P±)AdZ
= KGaPT(Y - Y±)AdZ
where
Z is height, ft
Si & LM are the mo^ar Sas and liquid rates per
unit area, Ib moles /hr-ft2
A is the cross-sectional area, ft2
Y & X are mole fraction of pollutant in the gas
phase and liquid phase
Kga is the mass transfer coefficient,
Ib moles/hr-ft3-atm
P is partial pressure of pollutant, atm
10-4
-------
Liquid inlet
LM' X0
T
Gas outlet^.
GM> YE
Gas inlet
GM» Y0
t
dz
yz+dz Xz+dz
x
Liquid outlet
LM' XE
Figure 10-I
PACKED TOWER
10-5
-------
P. is the interfacial pressure of pollutant,
1 atm
P_ is total pressure, atm
Y. is mole fraction of pollutant in the gas
1 at the liquid gas interface.
Eq 1 can be rearranged and integrated over the entire height
of the tower:
f dz G»
If' a to
JS-^O-C rp
\j JL
riu
dY
Y - Y.
Z =
0" YE
where YO and YE are the entering and exiting mole fractions
of the pollutant.
For pollutants that are very soluble in the liquid phase
or that react very rapidly once in the liquid phase, the mole
fraction of the pollutant at the interface, Y.^, will approach
zero*. (Effect of reaction rate on K^a is considered below.)
Vj
Therefore:
rYO
G
G
Z =
m
dY
Y
YE
(3)
or
*»« <*>
10.2.2 Mass Transfer Coefficient
In the final report of the previous project (Report
No. EPA-R2=-75-088) , the mass transfer coefficient, Kg, and
the wetted area fraction, "a", were separately determined.
This approach was abandoned because wetted area data for
* In fact, a considerable part of the experimental work was
aimed at finding purge rates and reagent concentrations
for which this assumption is true.
10-6
-------
various types and sizes of packing at different liquid and
gas flow is extremely limited. From the actual mass transfer
rates observed with the laboratory scrubber, a relationship
was developed for the product K^a as a function of the
following variables:
type of packing
size of packing
liquid mass flux, lb/hr-ft2
gas mass flux, lb/hr-ft2
10.2.2.1 Type and Size of Packing
The manufacturers' literature (U.S. Stoneware, 1963,
1970; Ceilcote) contains extensive data on the relative
mass transfer rates for different types and sizes of packings.
During this project, we have used Intalox saddles because
of the large amount of design data in the literature for
Intalox saddles.
The computer model is capable of using different types
and sizes of packing. U.S. Stoneware (1963,1970) has
developed mass transfer data for many types and sizes of
packing. In the typical operating range, the mass transfer
curves for the different packings have similar shapes.
This allows one to use the same equation, and to include
a multiplier expressing the ratio of the mass transfer rate
of the given packing with that of 1-1/2-in. Intalox saddles.
Thus, the mass transfer equation used to size packed
towers can be written in the following form:
K~a = Constant LaG~bPKGA
\j
where PKGA is the ratio of the mass transfer rates of the
'type and size of packing used to the rate of 1-1/2-in.
Intalox saddles.
10-7
-------
The size of the packing has a strong influence on the
mass transfer rate attained; however, the pressure drop is
also strongly affected. For full-scale scrubbers, 1-1/2-
or 2-in. packing is normally used as a compromise between
reasonable pressure drop and mass transfer rate. The cost
of scrubbing with various types and sizes of packing can
be explored by use of the computer model.
10.2.2.2 Liquid Mass Flux
The liquid flow rate per unit of tower area (mass flux)
also has a strong influence on the mass transfer rate. As
the liquid rate increases, the mass transfer rate increases,
regardless of the size and type of packing. The mass
transfer rate for most systems is related to the liquid mass
flux by the following general relationship (Eckert, 1970):
KGa « a)0'1'0-3
For C02 scrubbing, the exponent for L is 0.27 (U.S. Stone-
ware, 1970; Perry, 1963). For many other scrubbing opera-
tions, J:his exponent is in the range of 0.25 to 0.29 (U.S.
Stoneware, 1970). Because the effect of L on the mass
transfer rate has not been experimentally investigated for
odors, the value established for C02 removal was used for
removal of rendering plant odors:
V « (L)°-27
10.2.2.3 Gas Mass Flux
The mass flux of the gas, G, refers to the flow rate per
unit of cross-section of packed bed area. Since the total
air flow is fixed for a given application, the mass flux
determines the diameter of the scrubber chosen. It is impor-
tant to know the effect G will have on the mass transfer rate,
since the equipment size and cost will be strongly influenced
by its effect.
10-8
-------
The effect of the gas mass flux on mass transfer is not
easy to predict. For gases such as ammonia, where the heat
of solution and solubility is high (solubility 38 wt70) , the
gas mass flux has a significant effect on the mass transfer
rate. For ammonia (Perry, 1963):
m°-2
Kpa « £
V? I Vj I
V /
For pollutants that are not readily soluble in the liquid
phase, such as C02 (solubility 0.169 wt% at 20°C), the mass
transfer rate is not affected by the gas flux, but is deter-
mined by the amount of liquid the gas comes in contact with.
Thus, in the scrubbing of C02 with dilute caustic, above a
minimum gas flux, G has no influence on the mass transfer rate
for this system.
0
V
An example of this behavior is presented in Figure 10-2
(U.S. Stoneware, 1963). The solubility of certain rendering
odor compounds is low, especially for example the sulf ides ,
and they are expected to behave like C02 •
Test 33 was conducted with the laboratory scrubber to
determine the effect of G on the mass transfer rate for ren-
dering plant odors. The second and third stages were operated
in parallel with the same liquid mass flux, but stage No. 2
had nearly twice the gas through-put . The results from this
test are presented below.
Gas Flux
lb/hr/ft2
Reduction in Odor Units
ED 50
Sample Rd No. 1
ED 50
Sample Rd No. 2
Scrubber No. 2
Scrubber No. 3
543
275
15,000 -»• 680
15,000 -»• 20
4,300 -> 350
4,300 + 60
10-9
-------
m "
I-'
L.
-
of
I
v
? CO
M LJ
0 _|
O
2
CO
CO
_J
o
o
8
6
4
2.
(i**
.5
1 f J
.8
.6
4
li
1 1 I 1 1 1
-
L KGa VS GAS RATE
~ I" PALL RINGS (METAL)
mm
— — _ -- _ - . — — "~ X
^i — •*' " ^
^,^ «•*•* ^xx
^r**'^ — ->"***^
•^•M ^^P
^^
^S
s
~ / 1" TELLERETTES (H.D. POLYETHYLENE )
— / IOLB/CU.FT BED DENSITY
91 D / /~*\ \ r" "P nr" rx rxr* K i r^ i T* \ /
LB./CU. FT BED DENSITY
MBMW
LIQUID RATE - 1500 LBS./FT2, HR.
MM^^B
DA'* Of r^KERT f-~r-TE AND ."'JNTiNGTON
'. S. STOKFWARC tNC-:MEL:CiN!.;. .i? ' R^TO'I :S
1 1 1 1 1 1 1 ! 1 1 1 1 1 1
>0 200 300 400 500 6DO 6-CO
—
—
—
—
__
__
—
—
I I
fO(
GAS RATL-- L3S./FT2. HR.
(from U.S. Stoneware, 1963) Figure 10-2
-------
The number of mass transfer units (NTUG) is defined as:
NTUG = Zn(YO/YE)
where
YO is the inlet odor level
YE is the outlet odor level
The height of a transfer unit (HTUG) is defined as:
HTUG =
NTUG
where Z is the total height of the packing. From Section 10.2.1
the following equation was developed.
KGa ~ ZP
m_
T
At atmospheric pressure, and converting moles to pounds, this
can be rewritten as:
KGa 29 HTUG
where
29 is the molecular weight of air
K,,a is the mass transfer coefficient,
Ib moles/hr-ft3-atm
G is the gas mass flux, lb/hr/ft2
HTUG is the height of a single mass transfer unit
Using the equations listed above, the following values were
calculated.
Scrubber
No.
2
3
G
lb/hr/ft2
543
275
NTUG
2.80
5.45
HTUG
(ft)
1.43
0.73
Kra
(a
Ib moles/hr-ft3-atm
13.09
12.99
10-11
-------
From the above table, doubling the gas throughput resulted
in less than 170 change in the mass transfer rate, K~a. The
result of this single experiment indicates that the mass
transfer rate is independent of the gas mass flux, or:
0
V <* [S
10.2.2.4 Final Mass Transfer Equation
By using the data collected on the laboratory packed
scrubber, a value for the mass transfer rate was determined.
The development of the proportionality constant is presented
in Appendix 4. The final form of the equation utilized in
the model is as follows: (Note that it does not involve G,
in accordance with the result just obtained)
uu
0.27
KGa ~ 4'5[TOOOJ
(PKGA)
For 1-1/2-in. Intalox saddles, PKGA = 1.0 by definition.
10.2.3 Correlation for Pressure Drop
The pressure drop per foot of packing is described in
the following equation from U.S. Stoneware (1963).
AP
[Alpha] ,1nN (Beta L)/3600
i PP. J
where
AP is pressure drop, in. of water
PG is gas density. lb/ft2
G is gas mass flux, lb/hr-ft2
L is liquid mass flux, lb/hr-ft2
Alpha, Beta are constants as given in Table 10-1.
10-12
-------
o
I
M
U>
Table 10-1
PRESSURE-DROP EQUATION FOR PARTICLE PACKING
TYPE OF PACKING
INTALOX SADDLES
RASCHIG RINGS
BERL SADDLES
PALL RINGS
PALL RINGS
RASCHIG RINGS
^32 WALL
RASCHIG RINGS
1^16 WALL
MAT'L
CERAMIC
CERAMIC
CERAMIC
PLASTIC
METAL
METAL
METAL
Q
fl
Q
0
Q
&
Q
2
V
0
Q
0
Q
0
NOMINAL PACKING SIZE (INCHES)
1
4
3
8
1
2
1.04
0.37
1.96
0.56
1.16
0.47
1.59
0.29
5_
8
1.31
0.39
0.43
0.17
1.20
0.28
1.01
0.39
3_
4
0.52
0.25
0.82
0.38
0.56
0.25
0.80
0.30
1
0.52
0.16
0.53
0.22
0.53
0.18
0.22
0.14
0.15
0.15
0.53
0.19
'4
I3-
'8
0.14
0.13
'i
0.13
0.15
0.31
0.21
0.21
0.16
0.08
0.16
0.29
0.20
2
0.14
0.10
0.23
0.17
0.16
0.12
0.10
0.12
0.06
0.12
0.23
014
3
0.18
0.15
EQUATION
Ap =ox IO^L(~i~) (LIMITED TO REGION BELOW FLOODING)
Ap=PRESSURE DROP - INCHES OF H2O/FT. OF PACKING
G =GAS MASS VELOCITY - LBS./SEC. FT2
L = LIQUID MASS VELOCITY - LBS/SEC. FT2
y°G =GAS DENSITY - LBS/FT 3
a AND 0 = CONSTANTS
(from U.S. Stoneware, 1963)
-------
10.2.4 Correlation for Gas Flooding Velocity
The gas flooding velocity is presented in a mathematical
form by Sawistowski (1957).
GF = 3600
(Ap)
0.2
exp
0.125
where
GF is flooding gas mass flux, lb/hr-ft2
GR is acceleration due to gravity, 32.2 ft/sec2
e is void fraction of the packing
p, is density of the liquid, lb/ft3
LI
PG is density of the gas, lb/ft3
Ap is surface area of packing per unit volume,
ft2/ft3
y is viscosity of water (1 centipoise)
y-, is viscosity of the liquid (centipoise)
L is liquid mass flux, lb/hr-ft2
G is gas mass flux, lb/hr-ft2
The values of E and Ap for different packings are listed in
Table 10-2.
10.2.5 Design Gas Flux
In order to insure stable operation and proper wetting of
the packing, G is usually set at some fraction of the gas
flooding rate, Gp. For Intalox saddles, G is fixed between
65-85% of the gas flooding rate (Perry 1963) . Where the mass
transfer rate is liquid-phase controlled, such as with odor
removal using a sodium hypochlorite solution, the mass transfer
rate is essentially independent of G. Thus, G should be as
large as possible to minimize the tower diameter and minimize
10-14
-------
Table 10-2
CHARACTERISTICS OF RANDOM PACKINGS*
schig rings:
.'cramic:t
c,
af
Vlctal:
'fo-in. wall:
c,
•
e
°»
1 rings:
>laO ii- •
c,
a,,
vletal :
(
",
alo.x saddles.
eramic
€
«„
rl saddles,
c,
€
«„
Nominal size, in.
,
1,000
0.73
240
700
0.69
236
600
0.75
300
900
0.60
274
•>„
750
0.68
155
i ..
640
0.63
III
300
0.84
128
340
0.73
118
265
0.78
190
380
0.63
142
'•: »
380
0.68
100
258
290
97
0.88
110
71
0.902
131.2
?|
255
0.73
80
185
0.88
83.5
230
0.78
71.8
130
0.77
102
170
0.66
82
I
160
0.73
58
115
0.92
62.7
145
0X5
56.7
52
0.91'
63.0
48
IU
Us
125
0.74
45
•
110
0.87
49.3
0.938
66.3
95
0.71
38
82
0.90
41.2
32
0.905
39
28
0.953
48.1
98 '52
0.775 i 0.81
78
110
0.69
76
59.5
65
0.75
1 44
2
65
0.74
28
57
0.92
31.4
25
0.91
31
20
0.964
36.6
40
0.79
36
45
0.72
32
^
37
0.78
19
37
0.95
20.6
3',,
16
23.4
Tellcrcttes,§ 3i .•• 2 in., plastic:
High density:
C, - 57
t ~ 0.87
Low density:
C, 65
e - 0.83
t Table from the United States Stoneware Company. Data arc for wet dumped packing in 16- and
30-in. ID towers.
t Nominal si/c and wutt thickness, in inches, Tor ceramic Rasclug rings arc, respectively, ! i, \-M;
?», Hb; !i to »i, 3aa; 1. '„; I11 and I',, •',t; 2,' (: 3,';.
§ Teller and'Ford: Intl. Eng. Cliem., 50, 1201 (1958).
*from Treybal, 1967)
10-15
-------
the cost of the system. Therefore, in the computer model, G
was set equal to 85% of the gas flooding rate, Gp .
10. 3 Chemical Consumption
Based on the test results during this project, sodium
hypochlorite with excess caustic (pH > 12) was found to be
the most effective scrubbing combination. The most economi-
cal way to prepare the scrubbing solution is to purchase
chlorine gas and 50% liquid caustic. The sodium hypochlorite
can be purchased already prepared or it can be generated on
site; however, the use of chlorine gas is more economical.
On-site generation of sodium hypochlorite will be discussed
in detail in Section 14.
In the model, a continuous counter -current blow-down
system as shown in Figure 8-1 was assumed. As discussed in
Chapters 6 and 9, the pH in the third stage must be maintained
at or above pH 12. Also, as discussed in Section 7, a mini-
mum chlorine concentration (0.17% C12 was the lowest concen-
tration tested) must be maintained in all stages for maximum
odor removal.
The chemical consumption is related to the quantity of
odorants removed from the air. During this program, the
odor was measured both by gas chroma tography and by an odor
panel. The odor panel measures the odor intensity in odor
units, which is the ratio of the dilution volume to sample
volume required to reduce the sample to the odor threshold
for 50% of the panelists (ED50). Thus, odor units is an
indirect measure of the concentration of odorants present
in the gas stream. To obtain a measure of the quantity of
odorants present, the concentration can be multiplied by
the volume of air processed by the scrubber.
Odor Quantity = (^^^ S) (ft3 of air treated)
10-16
-------
Some of the odorous compounds removed from the gas stream
will be oxidized by the hypochlorite, while some of the compounds
will not react and will accumulate in the solution. Thus, the
"Odor Quantity" is a measure of the oxidized by-products plus
the unreacted compounds. The ability to remove these compounds
from the gas into the scrubbing solution can be measured:
Scrubbing Capacity = "^ ^umicxujr
Gallon or solution used
10.3.1 Chlorine Consumption
The total chlorine used is attributable to two distinct
mechanisms:
1. Chlorine reacting with the odorants
2. Chlorine blow-down loss.
10.3.1.1 Chlorine Reacting with the Odorants
The amount of chlorine consumed is directly related to the
odor quantity removed by the scrubber system. This chlorine
consumption due to odorous compounds has been checked by using
batch scrubbing tests without any blow-down. Thus by measuring
the odor levels and the chlorine concentration decrease with
time, a relationship was developed. This relationship, which
is derived in the Appendix, is written in terms of the Odor
Quantity.
0 32
Annual chlorine consumed for odor removal = yfrg— (Odor Quantity)
where
-., -. . .,. . ., odor } fTotal volume of air
Odor Quantity is the [units^e^vedj [treated per year
10.3.1.2 Chlorine Blow-down Loss
A certain fraction of the solution in the first stage must
be continuously discharged to the sewer to purge the refractory
and oxidized by products from the scrubbing solutions, and to
cause a flow of liquors from stage to stage to supply chlorine
10-17
-------
needed in the first stage. It will be shown that the blow-
down rate must be higher in a single-stage scrubber than
with multiple stages to keep the concentration of refractory
compounds low, since there are no subsequent stages to remove
refractory compounds. Since the blow-down rate is higher in
a one-stage scrubber, there is more wastage of reagents to
the sewer and the scrubbing capacity is less.
The scrubbing capacities were experimentally determined
from tests in which the scrubber was operated as a 1-, 2-,
and 3-stage unit. These data are analyzed in the Appendix.
The scrubbing capacities are inversely proportional to the
blow-down rates. The following table was also developed in
the Appendix for a scrubber treating 66,000 cfm of plant
ventilating air, based on test results of both the laboratory
and plant scrubber. (The scrubbing capacity values were
not verified for scrubbing of high-intensity odors during
the two-week test described in Chapter 9.)
Number of
Stages
3
2
1
Scrubbing
Capacity
14 x 106
7 x 106
3 x 106
Required
Blow- down Rate
2 gpm
4 gpm
9 gpm
The chlorine wasted to the sewer is a function of the
blow-down rate and the chlorine concentration maintained in
the first stage. Figures 10-3 and 10-4 show the chlorine
levels in all three stages for the plant and laboratory life
tests. The concentration of chlorine in the first tank equals
the concentration of chlorine in the blow-down to the sewer.
From the figures and using the blow-down rates that were used
for each life test, the amount of chlorine lost to the sewer
was computed. For the three-stage scrubber tests performed
using the continuous blow-down, 8% of the chlorine was lost
to the sewer.
10-18
-------
CHLORINE RESIDUAL VS TIME
(Test 34)
o
I
0.7 --
0.6
c
O
o
c
O
O
0.5 --
0.4 —
« 0.3--
o
o
0.2
0.1 ..
Larger
chlori ne
rotameter
i nstalled
Tank 3
72 96 120
Hours of operation
144
168
192
Figure 10-3
PLA1IT SCRUBBCR CHLORIMK LEVELS
-------
O Q
CHLORINE RESIDUAL VS TIMF
(Test 35)
0.6
Tank 3
O
I
•ro
o
-------
In a one-stage system, the blow-down rate is very high
compared to a three-stage system. The amount of chlorine
lost to the sewer will depend on the chlorine concentration
maintained in the tank. In a one-stage system, if the chlorine
level falls due to fluctuating odor concentrations, there
is not the back-up of the following stages. In a one-stage
system, it is imperative that a chlorine residual is always
present, but the required residual is not known exactly.
As a reasonable compromise between maintaining a chlorine
residual and wasting excessive chlorine to the sewer, a
chlorine concentration of 0.3% was assumed in the one-stage
model. More research is needed to better define the minimum
chlorine level that is required.
For a two-stage system, the chlorine residual in the
first stage should be maintained somewhere between the
levels used for the one- and the three-stage systems. The
model uses 0.2% chlorine in the first stage as the chlorine
control in the two-stage system. Figure 10-5 shows the chlorine
levels for all three scrubber systems. Using Figure 10-5,
the percent chlorine lost to the sewer was computed:
Chlorine loss
1 stage 54%
2 stages 25%
3 stages 8%
As stated earlier, the blow-down rate is determined by
the odor quantity removed. The above percentages must be
converted to pounds of chlorine consumed. This can be done
by relating the chlorine loss to the chlorine consumed for
odor removal. The derivation is presented in the Appendix
with the final equation shown below:
Annual chlorine loss fAnnual chlorine for] / 0 , ^ OQMO\
to'sewer = [ odor removal J 4'2 exp(-1.28NS)
10-21
-------
26 Ib/hr CT
9 gpm H20
16 Ib/hr
4 gpm H20
12 Ib/hr
2 gpm H20
0.3%
Chlorine
Ib/hr C12
9 gpm H20
1-stage system
0.4%
Chlorine
0.2%
Chlorine
\
4 Ib/h
r
r C12
k qpm H20
2-stage system
0.5?
Chlorine
Chlorine
0.18
Chlorine
1 Ib/hr C12
2 gpm H20
3-stage system
Figure 10-5
CHLORINE CONSUMPTION RATES FOR SCRUBBER DESIGN MODEL
10-22
-------
10.3.1.3 Total Chlorine Consumption
The total chlorine consumed is the sum of the chlorine
utilized plus chlorine lost.
Annual chlorine = Annual chlorine consumed for odor removal
+ annual chlorine loss to sewer
Note that only the chlorine loss to the sewer is a function
of the number of stages.
10.3.2 Caustic Consumption
Caustic consumption is attributable to four distinct
mechanisms.
1. Caustic consumption due to reaction with chlorine
make-up.
2. Caustic consumption due to C02 absorption.
3. Caustic consumption due to reaction with odorous
compounds.
4. Caustic consumption required to raise the pH of
the incoming water to 12.
10.3.2.1 Caustic Consumption Due to Chlorine Reaction
The caustic consumption for reaction with chlorine make-up
is directly proportional to the chlorine used, as shown by
the following reaction.
C12 + 2NaOH -»• NaOCl + NaCl + H20
For every mole of chlorine (71 Ib) consumed, 2 moles of
caustic (2 x 40 Ib) will be used to form sodium hypochlorite.
Thus, the following relationship can be used for predicting
this part of the caustic consumption:
Annual NaOH for C12 = 1.12 x (annual chlorine)
10.3.2.2 Caustic Consumption Due to C02 Removal
Some caustic is consumed by reacting with the C02 scrubbed
from the air. This caustic consumption will be a function of
10-23
-------
the air flow rate and is independent of the odor level. Data
(Eckert, 1970) for C02 scrubbing shows that the removal is
not complete, but increases with increasing pH of the caustic
solution. Thus, C02 absorption varies directly with the NaOH
concentration and the amount of Na2C03 in equilibrium with
the NaOH. The equilibrium between NaOH and Na2C03 depends
on the caustic blow-down and make-up rates.
From the plant test using the continuous blow-down sys-
tem, it was observed that when the final stage was near pH 12,
there was a drop in pH to the next stage of over 2 pH units.
In aqueous caustic at pH 12 there is 0.04% free NaOH. At
pH 10, there is only 0.0004% free NaOH. Thus, in the scrub-
ber system, the amount of free caustic in the first and
second stages is negligible compared to the caustic in the
third stage.
Eckert (1970) presented mass transfer rates for C02
removal using various concentrations of NaOH and Na2C03.
With 0.05% free NaOH and assuming 0.25% Na2C03 in the solution,
a mass transfer coefficient K.-.a of 0.19 lb moles/(hr-ft2-atm)
(j
is obtained from Eckert's data. In Section 10.2.2, the
mass transfer rate for odorous compounds was developed.
For 1-1/2-in. Intalox saddles, the mass transfer rate is
typically 7 lb-moles/(hr-ft2-atm) . In these experiments,
the system was designed to remove 90% of the odor units.
Such a system would also remove:
Kra C02 n 1Q
C02 removal = 0.9 ~ -^ =09 ~ = 2 47
KQa odorants u>* T7(T ^/0
A system designed to remove 90% of the odorous compounds will
remove 2.4% of the C02, or 0.8 lb C02 per 106 ft3 of air.
If it were not for this mass-transfer limitation, the C02 in
the air would neutralize much greater amounts of caustic.
10-24
-------
For every Ib-mole of C02 removed, 2 Ib-moles of caustic
are consumed, or 1.44 Ib NaOH/106 ft3 of air. The above
equation is for 90% removal ; a generalized model is presented
below:
Annual NaOH for C02 = - (CFM) (NTUG) (HRS)
where
CFM is volumetric air flow rate, ft3/min
NTUG is number of transfer units per stage
= £n(YO/YE)
HRS is number of hours per year the scrubber
is operated.
10.3.2.3 Caustic Consumption Due to Reaction with Odorants
By comparing the liquor life tests on the plant and
laboratory scrubbers, a good indication of the caustic consump-
tion due to reaction with odorous compounds can be obtained.
This caustic consumption is a function of the odor levels
and gas throughput. The following relationship was developed
from the life tests performed in Test 32. The derivation is
presented in the Appendix.
Annual NaOH for odor removal = (YO - YE ) QQQQOO (HRS) (0.10)
where
CFM is the volumetric air flow rate, ft3/min
YO is the inlet odor concentration, odor units
YE is the scrubber outlet odor concentration,
odor units
HRS is the number of hours per year scrubber is
in operation.
10.3.2.4 Caustic Consumption Required to Raise the
Incoming Water to pH 12
The make-up water must be raised to pH 12 to maintain
the pH in the last stage, where the make-up is introduced.
10-25
-------
The make-up rate is equal to the blow-down rate.* In distilled
water, 0.01 Normal NaOH has a pH of 12. In a typical tap
water 0.015 Normal NaOH will be required for pH 12. Thus,
0.015 Normal or 0.06% by weight NaOH must be introduced into
the make-up water to maintain the pH. The water blow-down
rate times 0.06% will yield the Ib of caustic required.
Annual NaOH for 0.0006
make-up water
blow-down rate, gal
8.34 Ib
1ST
60 min
TTF
HRS
= (0.3)(HRS)(Blow-down rate, gpm)
10.3.2.5 Total Caustic Consumption
The annual caustic consumption is the sum of the four
items presented above.
Total Ib NaOH/yr
Annual NaOH for C12
+ Annual NaOH for C02
+ Annual NaOH for odors
+ Annual NaOH for make-up water
10.3.3 Design Odor Levels for Calculations
A scrubbing system should be designed to handle the maxi-
mum odor levels anticipated; however, when computing annual
chemical costs, the average odor levels anticipated should
be used.
In order to satisfy both the design and the chemical cost
requirements, the inlet odor levels are the anticipated average
levels, but the desired outlet values are lower than needed.
Thus, the scrubbers are designed with the necessary mass
transfer units (NTUG) for handling maximum odor levels.
* The make-up rate also must include any water lost through
evaporation; however, no caustic is required for the evapora-
tion loss because the caustic is left in solution when the
water evaporates.
10-26
-------
10.4 Costs
10.4.1 Capital Costs
This section details the cost equations used in the compu-
ter model. The equations include the purchase and installa-
tion costs for the entire system. All cost data taken from
the literature were changed to 1957-1959 costs; these are
then multiplied by the Cost Index to convert to current costs.
10.4.1.1 Tower
The purchased cost of fiberglass reinforced plastic
towers is given by Ceilcote (1973). The installation cost
of the towers was developed using construction data presented
by Popper (1970). Included in the installation cost is the
concrete, piping, structural support, instrumentation, elec-
trical, and labor costs. The installation portion of the
cost was assumed independent of the material of construction.
The equation derived from the data provided by Popper
and Ceilcote for one tower is presented below:
Tower(1) = (51[MCF] + 69)(CI)(Diam)^2(Z + 4)°'65
where
MCF is material cost factor
CI is cost index
Diam is diameter of the tower, ft
Z is packing depth, ft
Tower(1) is installed cost of the tower, $
As the tower diameter increases above 10 ft, the installed
cost increases much faster as a function of diameter. This
is because the tower will have to be field fabricated.
Structural support costs will also increase much faster than
10-27
-------
diameter. Thus, for towers greater than 10 ft, the following
relationship was used.
Tower(1) = [12(MCF) + 16](CI)(Diam)1'85(Z + 4)°'65
Using Popper's cost estimating technique, it was estimated
that additional installed towers would cost 80% of the first
tower. Thus, the total cost of a multiple tower installation
is:
Tower = Tower(l) + (NS - 1) (0.8)[Tower(1)]
10.4.1.2 Tanks
The size of the liquid holding tanks is a function of
the liquid pumping rate. The following relationship was
developed, assuming a holding capacity equal to four times
the liquid pumping rate (4 x GPM) and data supplied by
Ceilcote (1973).
r \ 0.6
Tanks = 300(CI)(NS)(MCF) §7T
where
Tanks is installed cost of the tanks, $
CI is cost index
NS is number of stages
MCF is material cost factor
GPM is liquid pump rate, gal/min
10.4.1.3 Internals
The internals include a demister, a liquid distribution
system, and a support plate for the packing for each stage.
The cost relationship for the internals was also developed
from data supplied by Ceilcote (1973).
Internals = 3.38(CI)(NS)(Diam)2'6
where Internals is installed cost of the internals, $.
10-28
-------
10.4.1.4 Fan
The installed cost of the fan, excluding the motor, is a
function of the gas throughput (Aries et al., 1955). The
assumption was made that only one fan would be required
regardless of the number of stages. Aluminum fan blades
coated with glass-reinforced vinyl ester to resist corrosion
were assumed.
Fan = 0.25(CI)(CFM)°-77 + 50
where
Fan is installed cost of the fan, $
CFM is gas flow rate, ft3/min.
10.4.1.5 Pumps
Each stage of the scrubber will have its own centrifugal
stainless steel pump. The installed cost of all the pumps,
excluding motors, is a function of the liquid flow rate
(Popper, 1970).
Pump = 9(NS)(CI)(GPM)°'6 + 50
where
Pump is installed cost of the pumps, $
GPM is liquid flow rate in each stage, gal/min.
10.4.1.6 Motors (Open Type)
The cost of the motors required for the liquid pumps
and the fan is a function of the horsepower required
(Peters et al., 1968)
0.67 CT™ >>0.67
Motors = 350(CI) -
[HP
pump
(NS) +
fan
10-29
-------
where
Motors ±s installed cost of all the motors, $
NS is number of stages in the scrubber
HP um is pump horsepower
HP.c is fan horsepower
ran
CI is cost index
10.4.1.7 Packing
The packing cost is a function of the packing volume
and the unit cost of the packing.
Packing = (NS)(CPA)(Volume)
where
Packing is installed cost of the packing, $
CPA is unit cost of the type and size of
packing, $/ft3
Volume is packing volume of each stage
NS is number of stages.
10.4.2 Operating Costs
10.4.2.1 Pump Horsepower
The power required for pumping the liquid to the top of
the tower is given by:
HP = 8.34(Z + 10)(GPM)
pump 3.3 x 10* EP
where
HP is pump horsepower requirement
Z is packing depth, ft. 10 ft is allowed for
distributor level above packing and
miscellaneous pressure losses
GPM is liquid pumping rate, gal/min
EP is efficiency of the pump
10-30
-------
10.4.2.2 Fan Power
The power required to draw the gas through the entire
scrubber is
HPfan - 1.57 x 10- A
where
HPf is fan horsepower requirement
AP is pressure drop across the entire scrubber,
in. of H20
CFM is volumetric gas flow rate, ft3/min
EF is efficiency of the fan
10.4.2.3 Power Cost
The annual cost of power is
Power = 0.746(HRS)(HP)(CE)
where
Power is annual power cost, $
HRS is the number of hours in operation per year
0.746 is conversion factor from HP to kwatts
CE is cost of electricity, $/KWH
HP is the total power required for the pumps
and the fan
10.4.3 Chemical Costs
The cost of buying chlorine gas and sodium hydroxide
depend^ upon the quantity purchased and the location of the
plant. A major part of the cost of both chemicals is the
delivery charge. For the computer model, the assumption was
made that chlorine would be purchased in one-ton cylinders
for $0.11/lb. The caustic, delivered in tank truck loads
i • :
will cost $0.085/lb of sodium hydroxide. (These costs are
higher than the tank-car prices ($0.05/lb) which are often
10-31
-------
quoted.) In order to purchase the caustic at $0.085, a
2,500-gal caustic storage tank will be necessary; the cost
of such a storage tank is not included in the computer model.
The cost figures used in the model are the expected prices
in the Chicago area during the next year (K. A. Steel, 1974).
10.4.4 Maintenance Cost
As the size of a scrubber system increases, so does the
maintenance cost. In addition, as the number of stages
increases, the required maintenance will also increase. The
annual maintenance cost assumed in the model is presented
below:
Maint = CM /CFM(NS)
where
Maint is annual maintenance cost, $
CM is maintenance cost factor, used 8 in this
study
CFM is gas throughput, ft3/min
NS is number of stages.
10.5 Scrubber Design and Cost Estimate Results
With the computer model described in the previous sections,
the least-cost design can be computed. In order to use the
model, the user must specify the number of packed towers
in series, the desired inlet and outlet odor level in odor
units, the gas volumetric flow rate, and the number of hours
in operation per year. In addition, the user must specify
the type of material of construction and the current construc-
tion cost index.
A series of packed scrubber designs were calculated for
various volumetric flow rates and odor levels. A construction
cost index of 1.75 was used; this is the estimated index for
December 1974 (Chemical Engineering). Fiberglass reinforced
10-32
-------
plastic was selected as the material of construction for
the packed towers. Vinyl ester resin is recommended for
alkaline hypochlorite solutions by Dow Chemical Co. Intalox
saddles (1-1/2-in.) were selected for the packing because
their mass transfer/pressure drop behavior is one of the
best available. In addition, 4,000 hours of operation per
year was assumed for the scrubber.
The model reflects the current state-of-the-art in regard
to type and quantity of chemicals consumed and in regard to
the mass transfer rate. Most of the chemical consumption and
mass transfer data were obtained using both the plant scrubber
and the laboratory packed scrubber at Des Moines.
10.6 Discussion of Computed Costs
The results at 4,000 hours per year are presented for
various conditions in Tables 1 through 14 in Appendix 5. The
results are summarized in Table 10-3 and the optimum design
under each set of conditions is presented in Table 10-4. As
would be expected, the cost per 1,000 cfm-hr decreases as the
gas throughut increases. Tables 10-3 and 10-4 also show that
the cost per 1,000 cfm-hr increases from $0.23 to $0.39 for
50,000 cfm as the inlet odor level increases, due to the
increase in chemical consumption.
There are two major costs in using packed scrubbers in
odor control; the initial capital investment and the chemical
cost. One major assumption used in the model was that the
same liquid reagents will be used in each stage. Thus, by
operating the liquid tanks countercurrent to the air flow,
as shown in Figure 8-1, the chemical consumption can be re-
duced. On the other hand, a single-stage scrubber can provide
a larger mass transfer volume at a lower capital cost than
a multiple-stage scrubber; but, it cannot bleed the liquid
from stage to stage countercurrently. For odor levels that
average less than 500 odor units, a single-stage packed
10-33
-------
Table 10-3
PACKED TOWER COST SUMMARY
1-1/2-IN. INTALOX SADDLES
o
i
co
•P-
Average Design maximum
r-i „. odor units odor units
cfm Inlet Outlet Inlet Outlet
5,000 8,000 40 50,000 250
10,000 8,000 20 50,000 125
25,000 500 10 5,000 100
25,000 2,000 20 10,000 100
50,000 500 10 5,000 100
50,000 1,000 20 10,000 100
Number of
stages
1
2
3
3
1
2
3
1
2
3
1
2
2
3
Capital
investment,
$
25,800
33,600
39,500
67,200
75,700
96,100
113,700
83,800
103,100
120,800
144,200
181,400
197,800
230,300
Chemical
cost,
$/yr
25,000
15,800
13,100
26,300
9,000
5,500
4,500
32,300
20,200
16,700
18,000
11,000
40,400
33,400
Total
annual
cost,
$/yr
29,800
22,000
20,400
39,200
23,700
24,100
26,400
48,500
40,500
40,500
45,700
46,000
78,600
78,200
$/1000
cfm-hr*
1.49
1.10
1.02
0.98
0.24
0.24
0.26
0.48
0.40
0.40
0.23
0.23
0.39
0.39
*Based on 4,000 hours of operation per year and equipment depreciated 13% a year using the
st-raigYit—line depreciation method. See AMF factor, p. 10-3.
-------
01
Table 10-4
OPTIMUM DESIGN OF PACKED TOWERS*
1-1/2-IN. INTALOX SADDLES
Flow,
cfm
5,000
10,000
25,000
25,000
50,000
0 50,000
i
LO !
Average
odor units
Inlet
8,000
8,000
500
2,000
500
2,000
Outlet
40
20
10
20
10
20
Number of
stages
3
3
2
3
2
3
Flow rates,
lb/hr/ft2
Liquid
4750
4750
5125
4000
4750
4000
i
Gas
1100
1100
1060
1180
1100
1180
'acking depth,
ft
9.7
11.0
10.3
9.6
10.8
9.6
Diameter,
ft
5.1
7.2
11.6
11.0
16.2
15.6
$/1000 cfm-hrt
1.02
0.98
0.24
0.40
0.23
0.39
tBased on 4000 hours of operation per year and equipment depreciated 13% a year using
the straight-line depreciation method.
*0ptimum combination selected based upon annual rate of return on incremental capital
expenditure for additional stages. For example,
. c ^ c * • o £ i ,. A(0perating cost, 2-3)
Annual rate of return for 3 stages, instead of 2 stages = A(Capitai expenditure, 3-2)
Assumed a minimum annual rate of return of 10% required for additional stages.
-------
tower will be the most economical, according to the calcula-
tions. Above an average inlet odor level of approximately
2,000 odor units, the economics favor a three-stage system.
Packed scrubbers offer a viable alternative to incinera-
tors for gas streams containing high concentrations of odorous
compounds. In some cases, a higher capital expenditure will
be required for a scrubber system (Doty, et al., 1972), but
the savings in fuel costs more than offset any increase in
capital expenditure. One potential use of a packed scrubber
would be treating the non-condensibles in the cooker off-gas.
Scrubbing of concentrated odor streams can result in substan-
tial savings as compared to incineration. For example, a
three-stage 10,000-cfm scrubber treating the cooker off-gas
would use $26,300 of chemicals a year, compared to an incinera-
tor, which would use $64,000 of natural gas1 each year. Thus,
the savings of a scrubber system can be substantial. A fur-
ther savings can be attained by incorporating a small packed
tower for the concentrated odor streams, with a ventilating
spray scrubber for the entire plant. The concept of a total
scrubbing system is developed in Chapter 12.
The cost data presented in this section was based upon
performance tests at a rendering plant. However, test 35,
which was performed to verify the computer model, was unsuc-
cessful, due to the low water and chemical make-up rates
used. The water and chemical make-up rates were increased
for the model based upon the results of test 35; however,
these new rates have not been verified. Therefore, the
chemical costs presented in this section are only estimates
and not actually measured rates.
1Assumes a natural gas price of $1.00/1,000 scf.
10-36
-------
11. DESIGN AND COST OF HORIZONTAL SPRAY SCRUBBERS
11.1 Introduction
A computerized design procedure for horizontal spray
scrubbers was developed in the previous project (Report EPA-
72-R2-088). This model has been revised based on the infor-
mation collected at the rendering plant in Des Moines during
this project. The model is based on the spray scrubber design
of Air Conditioning Corporation, Greensboro, North Carolina,
but variations in the design are explored.
The scrubber consists of a large duct which varies in
size with the gas throughput, into which the plant air flows,
as shown in Figure 3-2 (page 3-4 ). A bank of spray nozzles
is mounted near the entrance of the duct. The liquid is pumped
under pressure (5-10 psig) through the spray nozzles, where
the scrubbing liquid is atomized. The model uses the penetra-
tion theory to calculate the mass transfer of odorous compounds
from the gas phase to the liquid droplets. Performance data
collected on the spray scrubber at Des Moines was compared with
the predicted performance from the model in Appendix 3. The
penetration theory model predicted the dimensions of the spray
scrubber within 13% of the actual dimensions using the Des Moines
performance data as input. A detailed description of the
penetration theory is provided in Report EPA-R2-72-088 and
a summary is given in Appendix 3.
11.2 Nozzle Pressure Effect on the Mass xranster Kate
In the earlier computer model, the mean droplet diameter
was assumed constant at 400 ym. The mean droplet size is not
constant, but a function of the nozzle pressure and the liquid
flow through each nozzle. The actual effect of the pressure
and flow rate on droplet size will depend on the nozzle design.
Data from the Delavan Manufacturing Company (1972) and the
11-1
-------
Spraying Systems Company (1966) were averaged to develop the
following relationship.
/nrf. 10.17f lOl0'23
DDrop = 40C [LNOZ j [PSTJ
where
DT-V is mean droplet diameter; ym
Drop r
LJJQ^ is liquid flow rate through the nozzle, gpm
PSI is pressure across the nozzle, lb/in.2
Using the above relationship, the mass transfer rate can
be determined using the penetration theory based on the tran-
sient diffusion of odorant through the gas layer adjacent to
each drop, given in Appendix 3. During this project, an
attempt was made to verify the above model and its effect on
the mass transfer rate.
In the last stage of the plant scrubber, all 270 of the
coarse nozzles (1/4-in. orifice) were replaced by finer
(1/8-in. orifice) nozzles to produce a smaller droplet size.
The liquid flow through both stages was adjusted to the same
•rate; so that the only variation would be the finer spray in
the third stage. The nozzle pressures were 18 psig and 26 psig
in the second and third stages, respectively. The scrubber
was operated as a two-stage scrubber, using the first two
stages (both coarse nozzles) and then using stages 1 and 3
(with fine nozzles in stage 3).
Prior to the test, 3 gallons of caustic were added to the
first stage, and 14 gallons of caustic were added to the second
and third stages. Chlorine was added to just the second and
third stages. At the beginning of the test, the following
conditions were measured.
pH % chlorine
Tank #2
Tank #3
11.25
11.25
0.157
0.147
11-2
-------
For the first half-hour of the test, the third stage was
turned off and the performance of the original low pressure
nozzles was measured. Then, after the first half-hour, the
second stage was turned off and the third stage was placed
in service with the first stage. The results from this test
are presented in Tables 11-1 and 11-2,
The inlet odor level averaged 4,700 odor units during
the first half of the test, but only averaged 265 odor units
for the second half of the test. The large variation in
the inlet odor levels makes any conclusion that can be drawn
very questionable. The finer nozzles removed an average of
80% of each component as compared to 75% for the coarse
nozzles; however, the reduction in odor units was much higher
for the low pesssure nozzles, primarily due to the very high
inlet odor level.
The equation for the droplet diameter developed from
the literature shows that the droplet diameter does not change
much with increasing pressure; however, the pumping cost
increases directly with nozzle pressure. To increase the mass
transfer rate, it is more economical to increase the number
of nozzles and operate at the lowest operable pressure drop
(10-15 psig) across the nozzles, rather than to increase the
nozzle pressure. In addition, the finer nozzles are more sus-
ceptable to plugging. Thus, the maintenance cost for cleaning
finer nozzles is higher.
The mass transfer data developed for horizontal spray
scrubbers was verified using the plant scrubber in Des Moines
with the coarse nozzles. The limited data available indicates
that these are more economical than fine nozzles operating at
higher pressure.
11.3 Other Model Modifications
The cost equations have been modified to include the con-
struct^on cost index and material cost factor as described in
11-3
-------
Table 11-1
RESULTS OF SCRUBBING TEST 36 - COARSE NOZZLES
PLANT SCRUBBER
Inlet Outlet % removal
Odor EDSO 4700 160 97
GLC peaks, Mg/liter air
1. Dimethyl sulfide 1.31 0.39 70
2, Pungent U.P. 90 -
3. Dimethyl disulfide - _'....
4. Butyric acid 0.52 0.09* 83
5. Propylene sulfide (cabbage 880) - -
^ 6. Bitter 655 - -
7. Burnt 1233A (pyrazine) - -
8. Burnt 1233B (pyrazine B) 1.15 0.27 77
9. Rendering 1335 35.06 7.78 78
10. Fatty 1545 4.39 1.51 66
11. Burnt 1687 (quinoline) -
12. Sulfur 660 -
13. Cheesy 2002 - -
14. Unpleasant 2582 - -
15. Fatty acid 1980 3.79 0.85 78
16. Fatty acid 2160 - -
Average percent removal of each component 75
-------
Table 11-2
RESULTS OF SCRUBBING TEST 36 - FINE NOZZLES
PLANT SCRUBBER
Outlet % removal
Odor EDso 265 110 58
GLC peaks, yg/liter air
1. Dimethyl sulfide 0.54 0.21 61
2. Pungent U.P. 90 -
3. Dimethyl disulfide -
4. Butyric acid 0.08 <0.01 90
5. Propylene sulfide (cabbage 880) -
6. Bitter 655 -
7. Burnt 1233A (pyrazine) -
8. Burnt 1233B (pyrazine B) 0.34 0.02 94
9. Rendering 1335 7.95 0.69 91
10. Fatty 1545 7.74 0.08 99.0
11. Burnt 1687 (quinoline) -
12. Sulfur 660 -
13. Cheesy 2002 -
14. Unpleasant 2582 -
15. Fatty acid 1980 0.13 0.07 46
16. Fatty acid 2160 - -
Average percent removal of each component 80
-------
Section 10.1. The cost equations are the same ones presented
in the packed tower section (10.4), except for the scrubber
cost. The spray scrubber cost equations are the same as used
in the previous project (except that a misunderstanding about
the length of a scrubber module has been corrected). The
chemical consumption equations are the same ones developed
in Section 10.3 for the packed towers.
11.4 Spray Scrubber Design and Cost Estimate
The computer model is programmed to determine the least-
cost design by varying the liquid flow rate through each
nozzle. For each liquid flow rate, the scrubber cross-sectional
area required to obtain the specified removal is then calcu-
lated. Assuming the scrubber height and width are equal,
the scrubber cost is then determined for each given liquid
flow rate per nozzle. After completing the costs for five
different liquid flow rates, the least cost design is then
selected by the computer program. There are several other
variables that will have an effect on the least cost design
and which must be specified in the input. Each of these
variables is discussed in the following paragraphs.
The nozzle density or the number of nozzles per square
foot will have an affect on the optimum design. In general, a
high nozzle density is desirable; this will allow for a lower
liquid flow per nozzle and a lower pressure drop across the
nozzles. The nozzle density and the flow rate per nozzle are
inversely proportional for a given total liquid flow. Because
the flow rate per nozzle is varied in the program, the nozzle
density was held constant for this work at 6 nozzles/ft2.
As with packed towers, the mass transfer rate increases
with increasing contact time between the liquid and the gas.
In this model, the contact chamber length is specified, and
the chamber cross-sectional area is calculated for the
desired odor reduction. The actual chamber lengths used for
any installation will depend upon the available space for the
11-6
-------
scrubber. The capital investment for a longer contact cham-
ber is about the same as for the shorter chamber because of
the resulting difference in the cross-sectional area. Thus,
available space will be the prime factor in determining the
contact chamber length.
Table 11-3 is a summary of the computer runs performed
under various conditions. The optimum design for each flow
rate and odor level is presented in Table 11-4. The actual
computer output for each set of conditions is shown in Tables 15
through 29 in Appendix 5. The optimum capital investment
varies from $2/cfm for a 25,000 cfm unit to $1.43/cfm for a
150,000 cfm scrubber. These capital investments are for
type 316 stainless steel scrubbers*, plus all of the necessary
auxiliary equipment. The optimum design for all conditions
is a multiple-stage scrubbing unit, reflecting the significant
chemical cost savings with multiple-stage units. At the higher
inlet odor levels, the three-stage units are the most economi-
cal. Where the average inlet odor level is greater than
1000 odor units, the third stage is economically justifiable;
below 1000 odor units, a two-stage unit should be purchased.
In one set of conditions, the nozzle pressure was increased
to 25 psi. No change in the capital investment was calculated
with the higher pressure; however, the electricity cost increased
$330/yr, for a 25,000-cfm scrubber. The increase in electri-
cal costs would be proportionately greater for larger scrubbers.
A comparison of the cost of spray scrubbers with the costs
of packed towers reveals that packed towers are more expensive.
The chemical and operating costs are about the same, but the
major difference is in the inital capital investment. Packed
*The use of stainless steel in scrubbers presents corrosion
problems. An alternative material is discussed in Chapter 15.
Stainless steel was used in the model because cost data are
readily available, and this is the material currently being
used.
11-7
-------
oo
Table 11-3
SPRAY SCRUBBER COST SUMMARY
Flow,
cfm
25,000
25,000
50,000
50,000
100,000
150,000
Average
odor units
Inlet
500
2000
500
2000
500
500
Outlet
10
20
10
20
10
10
Number of
stages
1
1
2
2
2
3
1
2
3
2
3
3
2
2
2
Chamber
length,
ft
8
16
8
16
16
8
8
8
8
8
8
8
8
16
16
Total
length,
ft
20
28
36
52
52
52
20
36
52
36
52
52
36
52
52
Nozzle
pressure,
psi
15
15
15
15
25
15
15
15
15
15
15
15
15
15
15
Scrubber height
and width,
ft
5.6
4.7
4.8
3.7
3.8
4.2
5.9
5. 1
4.4
7.1
6.2
6.6
10.7
8.6
11.0
Chemical
cost,
$
9,000
9.UOO
5,500
5,500
5,500
4,500
32,300
20,200
16,700
11,000
8,900
33,400
22,000
22,000
33,000
Capital
investment ,
$
39,200
39,200
49,800
47,700
47,800
57,500
40,900
52,100
60,100
87,200
100,300
104,600
155,900
150,700
214,100
Total cost,
$/yr
17,700
17,400
17,100
16,800
17,100
18,500
41,400
32,300
31,300
31,300
33,500
59,000
58,500
57,300
83,100
$/1000 cfm-hr*
0.18
0.17
0.17
0.17
0.17
0.18
0.41
0.32
0.31
0.16
0.17
0.30
0.15
0.14
0.14
*Based on 4000 hours of operation per year and equipment depreciated 13% a year using the straight-line depreciation method.
-------
Table 11-4
OPTIMUM DESIGN OF SPRAY SCRUBBERSt
Flow,
cfm
25,000
25,000
50,000
50,000
100,000
150,000
Average
odor units
Inlet
500
2000
500
2000
500
500
,_i t Optimum combination
^ For example,
Outlet
10
20
10
20
10
10
selected
Number of Chamber length,
stages ft
2
3
2
3
2
2
based upon annual
8
8
8
8
16
16
rate of return on
Scrubber height
and width,
ft
4.8
4.4
7.1
6.6
8.6
11.0
incremental capital
A (Operating
GPM/ nozzle
3.0
3.0
2.5
2.5
3.0
2.5
expenditure
costs, 2-3)
Liquid rate
per stage,
gpm $/1000-cfm-hr*
414
352
764
648
1320
1820
for additional
0.17
0.31
0.16
0.30
0.14
0.14
stages.
A(Capital expenditure, 3-2)
Assumed a minimum annual rate of return of 10% required for additional stages.
*Based on 4000 hours of operation per year and equipment depreciated 13% a year using the straight-line depreciation method.
-------
towers require over twice the capital investment of horizontal
spray towers. Thus, for plant ventilating air (odor levels
less than 2,000 odor units), spray scrubbers are recommended.
Packed towers should be restricted to small-volume high-inten-
sity streams where a large reduction (greater than 9070) in
odor level is required.
11-10
-------
12. DESIGN AND COST OF A PACKED TOWER-HORIZONTAL SPRAY
SCRUBBER SYSTEM
It was concluded from the information presented in the
previous sections, that both packed towers and spray scrubbers
can be designed to obtain specified odor reductions. If
process air and plant ventilating air are both to be treated
by wet-scrubbing, what is the most economical design? Spray
scrubbers require less capital investment to obtain the
same removals for large gas volumes; however, should one large
spray scrubber be designed to treat both sources, or should
the process air first be treated with a packed tower; and then
sent to the ventilating air scrubber? To evaluate these
alternatives, an economic determination of the two different
systems was performed.
Figures 12-1 and 12-2 depict the two designs considered
for a rendering plant with a 10,000 cfm high intensity gas
stream (8,000 odor units) and 90,000 cfm of ventilating air
(500 odor units). Both scrubber systems were designed to
achieve a maximum outlet odor level of 100 odor units. The
continuous blow-down and make-up scheme was used in both sys-
tems to reduce chemical costs*. Required removal efficiencies
varied in each design. In Case I, the three-stage horizontal
spray scrubber was required to achieve a 99% reduction of the
odorants, while in Case II, the packed tower needed only an
87.5% removal, with the spray tower removing 98%.
The costs associated with each design are presented in
Table 12-1. Pre-treatment of the concentrated odor stream
in a packed tower proves to be more economical than use of
one spray scrubber to handle the entire load. The primary
saving is the $4,500 per year power costs. The capital ex-
penditure is also lower for Case II by $8,300, while the
*Note that in existing installations similar to Case II the
packed-bed scrubber has its own chemical supply system,
increasing the chemicals cost.
12-1
-------
Avg
Max
90,000 cfm
500 odor uni ts
5000 odor units
3-stage horizontal
spray scrubber
100,000 cfm
.Avg 1250 odor units
Max 9500 odor units
10,000 cfm
Avg 8,000 odor units
Max 50,000 odor units
I
100,000 cfm
^ Avg 13 odor units
Max 100 odor units
Chemical
make-up
Blow-down
to sewer
Figure 12-1
CASE I - 3-STAGE SPRAY SCRUBBER
12-2
-------
Avg
Max
90,000 cfm
500 odor units
5000 odor units
2-stage horizontal
spray scrubber
100,000 cfm
[Avg550 odor unitjs
5125 odor ur
its
10,000 cfm
1000 odor units
Max 6250 odor units
1
100,000 cfm
.Avg 11 odor units
Max 100 odor units
T
^
-OL— .
^
Chemical
make-up
10,000 cfm
tog 8,000 odor units
Hax 50,000 odor units
Blow-down
to sewer
Figure 12-2
PACKED-BED WITH A 2-STAGE SPRAY SCRUBBER
12-3
-------
Table 12-1
COMPARISON OF SCRUBBING SYSTEMS
10
I
Spray Scrubber Design
Chamber length, ft
Height and width, ft
Total length, ft
Liquid flow rate, gpm/stage
Packed Tower Design
Tower diameter, ft
Packing depth, ft
Liquid flow rate, gpm
Costs
Capital investment, $
Spray scrubber
Packed tower
Case I
3-stage spray scrubber
8
9.9
52
1,174
Total
Chemical cost, $/yr
Power, $/yr
Maintenance, $/yr
$/1000 cfm-hr
183,100
183,100
42,600
17,500
4,400
0.221
Case II
2-stage spray scrubber
and 1-stage packed tower
16
8.6
52
1,316
7.9
8.6
690
148,500
26,300
174,800
42,600
13,000
4,400
0.207
-------
project maintenance and chemical costs for both systems are
equal.
The use of the packed tower for treating process air
streams increases the reliability of the system. The technology
for reducing high intensity odors by at least 87% in a single-
stage packed tower was consistently demonstrated throughout
this program. The discharge from the packed bed only increases
the average odor intensity to the ventilating scrubber from
500 to 550 odor units. Thus, the spray scrubber is still
treating a dilute gas stream very similar to just ventilating
air, which is the most advantageous application of spray
scrubbers.
12-5
-------
13. ACTIVATED CARBON STUDY
Although the emphasis of this project focused on developing
wet scrubber systems for odor control, a small study on the
effectiveness of activated carbon was also performed. This
study measured the performance of a carbon system and was used
to determine the adsorptive capacity of the carbon using actual
rendering plant air.
13.1 Laboratory Carbon Test
A three-stage carbon adsorber was constructed by the Calgon
Corporation for this test. Each stage in the adsorber contained
540 gm of carbon. The adsorber was connected in parallel with
the laboratory scrubber so both could be utilized to treat the
concentrated odor stream normally sent to the incinerator. A
flow of 17 cfm was passed through the adsorber for a period of
20 hr. Bag samples were collected on the inlet and outlet at
2 hr, 10 hr and 20 hr. Gas chromatograph samplers were also
used for the 20-hr sample on both the inlet and outlet.
The odor removal results are presented in Tables I3-I
and 13-2. The data show that carbon was very effective in
deodorizing the gas for the first 10 hr; however, by the 20th
hour the odor removal had dropped off to 54%, indicating that
the carbon was becoming saturated. The gas chromatograph re-
sults tabulated in Table 13-2 indicate an unusual pattern.
The lighter molecular weight compounds were still being
effectively removed after 20 hours, while the heavier com-
pounds appeared to be passing through the carbon. This is
extraordinary since in carbon applications involving a homologous
series, the lighter molecular weight compounds are the first to
break through. The heavier compounds begin to replace the lighter
compounds in the bed forcing the lighter compounds to pass
through the bed more quickly. These experimental results
indicate that carbon has a high adsorption affinity for the
13-1
-------
Table 13-1
ACTIVATED CARBON .- ODOR PANEL RESULTS
Time,
hr
2
10
20
Odor levels
Inlet Outlet
2750 45
1700 50
4800 2200
Odor panel
7o removal
98.4
97
54
GC
% removal
-
-
74
particular lighter odorous compounds found in rendering plant
air. Three of the five lighter compounds contain sulfur, and
carbon has a .high adsorptive capacity for many sulfur compounds.
This may explain the apparent affinity of carbon for these com-
pounds .
After the 20-hr test was completed, the carbon was removed
•from each stage, dried, and weighed. The following results
were obtained:
7o increase in weight
1st stage 29
2nd stage 20
3rd stage 16
These results demonstrate that the adsorption capacity for
rendering plant process air odors is greater than the 10%
weight adsorption capacity typically used for designing carbon
adsorbers. The large quantity of organics adsorbed per pound
of carbon was at least in part due to the high inlet odor
concentration in this experiment. Some caution should be used
if the results are applied to plant ventilating air, which
contains lower odor concentrations. An adsorptive capacity
of 5-10 wt% rather than 29 wt% would be a safer assumption
for plant ventilating air, because as the inlet concentration
decreases, the adsorptive capacity also decreases.
13-2
-------
Table 13-2
ACTIVATED CARBON TEST
TIME: 20 HOURS
Inlet
Odor ED so
GLC
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
peaks, Mg/liter air
Dimethyl sulfide
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
tlnpleasant 2582
Fatty acid 1980
Fatty acid 2160
4800
22
15
7
0
1
36
-
1
80
9
2
-
2
1
4
5
.10
.37
.38
.83
.11
.99
.54
.01
.84
.86
.78
.09
.67
.09
Outlet
2200
0
0
0
0
0
0
-
0
32
2
1
-
1
0
0
1
.53
.17
.09
.41
.14
.27
.55
.37
.93
.13
.45
.58
.88
.54
% removal
54
98
98
98
51
87
99
-
64
60
70
60
-
48
47
81
70
.9
.8
.3
Average percent removal of each component
74
-------
Table 13-1
ACTIVATED CARBON .- ODOR PANEL RESULTS
Time,
hr
2
10
20
Odor levels
Inlet Outlet
2750 45
1700 50
4800 2200
Odor panel
% removal
98.4
97
54
GC
% removal
-
-
74
particular lighter odorous compounds found in rendering plant
air. Three of the five lighter compounds contain sulfur, and
carbon has a .high adsorptive capacity for many sulfur compounds,
This may explain the apparent affinity of carbon for these com-
pounds .
After the 20-hr test was completed, the carbon was removed
'from each stage, dried, and weighed. The following results
were obtained:
% increase in weight
1st stage 29
2nd stage 20
3rd stage 16
These results demonstrate that the adsorption capacity for
rendering plant process air odors is greater than the 10%
weight adsorption capacity typically used for designing carbon
adsorbers. The large quantity of organics adsorbed per pound
of carbon was at least in part due to the high inlet odor
concentration in this experiment. Some caution should be used
if the results are applied to plant ventilating air, which
contains lower odor concentrations. An adsorptive capacity
of 5-10 wt% rather than 29 wt% would be a safer assumption
for plant ventilating air, because as the inlet concentration
decreases, the adsorptive capacity also decreases.
13-2
-------
Table 13-2
ACTIVATED CARBON TEST
TIME: 20 HOURS
CO
i
OJ
Odor ED so
GLC peaks, ug/ liter air
1. Dimethyl sulfide
2. Pungent U.P. 90
3. Dimethyl disulfide
4. Butyric acid
5. Propylene sulfide (cabbage 880)
6. Bitter 655
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9, Rendering 1335
10. Fatty 1545
11. Burnt 1687 (quinoline)
12. Sulfur 660
13. Cheesy 2002
14. Unpleasant 2582
15. Fatty acid 1980
16. Fatty acid 2160
Average percent removal of each component
Inlet
4800
22.10
15.37
7.38
0.83
1.11
36.99
-
1.54
80.01
9.84
2.86
-
2.78
1.09
4.67
5.09
Outlet
2200
0.53
0.17
0.09
0.41
0.14
0.27
-
0.55
32.37
2.93
1.13
-
1.45
0.58
0.88
1.54
% removal
54
98
98.9
98.8
51
87
99.3
-
64
60
70
60
-
48
47
81
70
74
-------
There is certainly a need for additional research on this
subject before the economics of activated carbon can be developed
as thoroughly as those of wet scrubbing; however, the results
from this one test are very encouraging. Not only does the
adsorption capacity of carbon at low odor levels need to be
determined, but also the effect of regeneration on the adsorp-
tion capacity.
13.2 Regeneration of Activated Carbon
The economics of carbon adsorption depends on the method
of regeneration, and the degree of regeneration that is possible.
Some loss in absorptive capacity often occurs with each regen-
eration jcycle. Only by experimentally repeating the adsorption-r
regeneration cycle many times can the rate of deterioration
of the Carbon be accurately estimated.
Several methods of regenerating carbon are available.
Regeneration by passing"hot gases through the bed to drive off
the -adsorbed material requires an incinerator to deodorize the
hot gases. This method is costly for rendering plants (Report
No. EPA-R2-72-088, p. 32). Where available, regional regenera-
tion plants operated by the carbon manufacturers would probably
be the most economical method. One other method that could also
prove very economical, if successful, would be in-situ chemical
regeneration. The capital investment would be minimal, and if
the carbon could be chemically regenerated anywhere near its
original capacity, the economics would be greatly improved.
Perhaps a solution containing the hypochlorite ion, either a
high pH or low pH solution would be very effective. Until more
work in this area is performed, the cost of activated carbon
for odor control can only be estimated on a throw-away basis,
and this is done in the next section.
13.3 Estimated Cost of a Carbon System Without Regeneration
For low odor intensity levels (less than 100 odor units),
the economics of using carbon on a throw-away basis have been
13-4
-------
estimated. One potential use of carbon is on the outlet of
a scrubber to achieve odor levels of 10-20 odor units at all
times. A carbon bed 2-3 in. thick could be placed after the
final demister for a minimal cost and with a small increase
in the over-all pressure drop. From the results obtained from
this test, the carbon adsorbed 0.773 Ib of organics for the
24-hr test. Assuming an average odor reduction of 3000 odor
units during the 20-hr test, the total quantity of odorants
removed was
3000 odor units
17 ft:
60 min
20 hr
Odorants removed = ^ __._ ^
= 61 x 106 odor-units-ft3
Thus, 61 x 106 odor-units-ft3 must be equivalent to the quan-
tity of organics adsorbed by the carbon, 0.773 pounds, or
80 x 106 odor units-ft3 = 1 Ib organics
Using this number, the annual cost of carbon as a polishing
step was estimated for several conditions. Because of the
lower inlet odor levels, adsorption capacities of 5 and 10%
were used. The results are presented in Table 13-3. The
cost of the carbon is obviously directly proportional to the
gas flow rate and odor reduction. At the higher gas flow
rates, carbon costs are too high to be economically con-
sidered on a throw-away basis. Even at 10,000 cfm, the
annual replacement cost of the carbon will range between
$4,200 and $8,400, depending on the true adsorption capacity.
The costs presented do not include the capital expenditure
or labor cost.
Using carbon on a throw-away basis does not appear to
be economically atractive, even for low-intensity odors
(less than 100 odor units) . More research is needed to deter-
mine how the carbon can be economically regenerated.
13-5
-------
Table 13-3
COST OF ACTIVATED CARBON FOR LOW ODOR LEVELS
Flow,
cfm
10,000
25,000
50,000
100,000
Average
Inlet
50
100
50
50
50
100
odor levels
Outlet
10
20
10
10
10
20
Annual carbon
5%
capacity
8,400
16,800
21,000
42,000
84,000
168,000
cost ($/yr)*
10%
capacity
4,200
8,400
10,500
21,000
42,000
84,000
*Carbon cost of $0.35/lb, no regeneration, carbon bed operated
4000 hr/yr.
13-6
-------
14. ON-SITE GENERATION OF SODIUM HYPOCHLORITE
Rendering plants that are currently using sodium hypo-
chlorite, typically purchase chlorine gas and bubble the gas
into a caustic solution to produce sodium hypochlorite. This
method, although the most economical, has several disadvantages,
Today there is a shortage in the United States of both chlorine
gas and sodium'hypochlorite. In addition, storage of chlorine
gas presents a safety problem. Many cities have banned the
storage of bulk quantities of chlorine gas within the city
limits, thus forcing users to purchase sodium hypochlorite
in solution form.
Purchased in one-ton cylinders, chlorine gas currently
sells for $0.11/lb delivered. The price depends on location,
and may vary from less than $0.11 to greater than $0.20/lb.
In tank-car quantities, chlorine gas can be purchased for
$0.06 to $0.07/lb. The large difference in price is due
to the handling and delivery charges.
When a plant switches to purchased sodium hypochlorite
solution for safety reasons, there is an even more severe
economic penalty. Purchased sodium hypochlorite typically
sells for $0.28/lb equivalent chlorine in 5,000-gal shipments.
The price of sodium hypochlorite also varies with geographic
location.
Delivered caustic soda currently sells for $0.085/lb,
but again delivery costs will determine the actual price.
The cost in 55-gal drums of 50% caustic solution is higher
than the cost of caustic solution in bulk. It will pay the
larger rendering plants to provide an indoor storage tank
and purchase caustic solution in tank-truck quantities. The
tank must be protected from cold weather to prevent crystalli-
zation from occurring.
14-1
-------
Because of the safety problems and the shortage of chlorine
gas, an alternative was evaluated. Several firms have developed
packaged units for on-site generation of sodium hypochlorite
solution. On-site generation units utilize salt water for
the feed material. The salt water is pumped through electro-
lytic cells, where an electric current is passed through the
water, producing a dilute NaOCl solution.
The economics of on-site generation depend upon the costs
of chlorine, electricity, and salt. Table 14-1 presents the
costs associated with on-site generation as compared to pur-
chasing chlorine gas and sodium hypochlorite. On-site genera-
tion can provide savings to plants currently purchasing sodium
hypochlorite. For plants purchasing chlorine gas at or below
$0.11/lb, on-site generation is not economically competitive.
The capital investment in an on-site generation is con-
siderable, ranging up to $94,000 for a 50,000-cfm scrubber
treating an average odor intensity of 3,000 odor units. In
Table 14-1, the total cost-per-year values were computed assuming
a 15-year equipment life and a 1070 interest rate, and that the
maintenance cost is negligible. Thus, the total cost-per-year
column can be compared directly with the "purchased chemical
cost" columns because the time value of money is taken into
account. For the last example presented in Table 14-1, on-site
generation costs only $1,100 per year more than purchasing
chlorine. If the price of chlorine increases to $0.12/lb,
on-site generation will become economically competitive. Of
course, a guaranteed long-term supply of cheap electricity
or salt would also improve the competitive position of
on-site generation. Each installation should evaluate on-
site generation using the anticipated chemical and electrical
costs. Larger installations will find on-site generation
more attractive.
On-site generation units are typically limited to produc-
tion of 0.6 to 0.8 wt% hypochlorite solution. On-site generator
14-2
-------
I
00
Table 14-1
ON-SITE GENERATION OF SODIUM HYPOCHLORITE
On-site generation
Flow,
cfm
5,000
10,000
25,000
25,000
50,000
Average
odor units
Inlet
8000
8000
500
2000
2000
Outlet
40
20
10
20
20
Purchased chemical cost
Number of
stages
3
3
1
2
2
C12 +.NaOH
$/yr(z7
13,100
26,300
9,000
20,200
40,400
NaOCl f NaOH
$/yrtz)
17,300
34,700
11,500
26,500
53,000
Capital
investment,
$
40,000
71,000
32,000
55,000
94,000
Annual savings of on-site
Operating. cost, Total cost. generation over purchased
$/yrm per yearu; NaOCl, $
10,500
19,600
7,700
15,400
29,300
15
28
11
22
41
,700
,800
,900
,600
,500
1
5
3
11
,600
,900
-
,900
,500
'includes 3.5 Ib of salt and 2.5 KWH/lb of chlorine; operating cost does not include depreciation.
2Unit costs, C12 - $0.11/lb
NaOH - $0.085/lb
NaOCl - $0.28/lb
Salt (NaCl) - $0.015/lb
Electricity - $0.025/KWH.
Operating cost plus depreciation cost (13% of capital investment).
-------
units require very little space, are easy to operate, and
require minimal operator attention. Most of the units avail-
able commercially have a modular design, so that if more
hypochlorite is required in the future, another module can
be added without costly modifications. Only about 30% of
the salt solution is converted to the sodium hypochlorite;
thus, the scrubbing solution will also contain high levels
of dissolved solids. Sewer ordinances for dissolved solids
must be checked to insure that on-site generation will not
create an effluent violation.
In order to obtain sufficient concentration of the hypo-
chlorite ion, the brine solution must be recirculated through
the generator. Ideally, the scrubbing solution used in the
last stage could be recycled through the generator; however,
this may result in fouling and the high pH affects the perfor-
mance of some of the commercial generators. Thus, unless a
guarantee from the vendor can be obtained, it is recommended
that a separate tank be installed for recirculating the solu-
tion, as shown below.
Make-up
salt solution
Sod i urn
hypochlorite
generation
Reci r—
culatior
tank
Caustic
make-up
4
1
Last stage
scrubbing
1 iquor
Hypochlorite
make-up
14-4
-------
In commercially available hypochlorite generators, hypo-
chlorite solution is produced by electrolysis of sodium chloride
solution. Chlorine is generated at the anode:
2C1~ •* C12 (dissolved) + 2e~
and hydroxyl ion at the cathode:
2H20 + 2e~ -> 20H" + H2 (gas)
The chlorine and hydroxyl ion combine in solution:
C12 + 20H" + OC1" + Cl~ + H20
producing one molecule of hypochlorite ion (OC1~) , and reforming
one chloride ion (Cl~). The overall reaction is:
NaCl + H20 -»• NaOCl + H2 +
2e"
Note that this is the equivalent of adding chlorine gas and
caustic soda to the solution. If a pH of 11 or 12 is wanted,
then additional caustic must also be added to the product from
the electrolyzer.
In conclusion, on-site generation provides an economic
alternative to rendering plants currently purchasing sodium
hypochlorite. On-site generation is also attractive where
shortages of chlorine gas exist, where safety standards are
forcing plants to switch from chlorine gas , and where the cost
of chlorine gas is greater than the $0.11/lb assumed in this
report. If the price of chlorine increases in the next few
years, on-site generation may find wide-spread acceptance
throughout the rendering industry.
14-5
-------
units require very little space, are easy to operate, and
require minimal operator attention. Most of the units avail-
able commercially have a modular design, so that if more
hypochlorite is required in the future, another module can
be added without costly modifications. Only about 30% of
the salt solution is converted to the sodium hypochlorite;
thus, the scrubbing solution will also contain high levels
of dissolved solids. Sewer ordinances for dissolved solids
must be checked to insure that on-site generation will not
create an effluent violation.
In order to obtain sufficient concentration of the hypo-
chlorite ion, the brine solution must be recirculated through
the generator. Ideally, the scrubbing solution used in the
last stage could be recycled through the generator; however,
this may result in fouling,and the high pH affects the perfor-
mance of some of the commercial generators. Thus, unless a
guarantee from the vendor can be obtained, it is recommended
that a separate tank be installed for recirculating the solu-
tion, as shown below.
Make-up
salt solution
Sod i urn
hypochlorite
generation
Reci r-
culatior
tank
Caustic
make-up
1
Last stage
scrubbing
1iquor
Hypochlori te
make-up
14-4
-------
In commercially available hypochlorite generators, hypo-
chlorite solution is produced by electrolysis of sodium chloride
solution. Chlorine is generated at the anode:
2C1~ + C12(dissolved) + 2e~
and hydroxyl ion at the cathode:
2H20 + 2e~ ->• 20H~ + H2(gas)
The chlorine and hydroxyl ion combine in solution:
C12 + 20H" -»• OC1~ + Cl" + H20
producing one molecule of hypochlorite ion (OC1 ) , and reforming
one chloride ion (Cl ). The overall reaction is:
NaCl + H20 ->• NaOCl + H2 +
2e~
Note that this is the equivalent of adding chlorine gas and
caustic soda to the solution. If a pH of 11 or 12 is wanted,
then additional caustic must also be added to the product from
the electrolyzer.
In conclusion, on-site generation provides an economic
alternative to rendering plants currently purchasing sodium
hypochlorite. On-site generation is also attractive where
shortages of chlorine gas exist, where safety standards are
forcing plants to switch from chlorine gas, and where the cost
of chlorine gas is greater than the $0.11/lb assumed in this
report. If the price of chlorine increases in the next few
years, on-site generation may find wide-spread acceptance
throughout the rendering industry.
14-5
-------
15. SELECTING THE PROPER SCRUBBER DESIGN
In this chapter a comparison of packed beds and spray
scrubbers is provided. Some of the more practical design
problems associated with scrubbers are also discussed. Al-
though the scope of the project was limited to these two
types of scrubbers, much of the information obtained during
this project can be extended to other types of scrubbing
systems.
15.1 Comparison of Packed Bed and Spray Scrubbers
In Chapter 10, cost information is provided for packed-
bed scrubbers operating at low gas volumes (5,000-10,000 cfm)
with process air, and also at higher gas volumes (25,000-
50,000 cfm) with plant ventilating air. Cost information for
spray scrubbers is presented in Chapter 11 for gas throughputs
greater than 25,000 cfm. No spray scrubber cost information
is provided at the lower gas throughputs with higher odor
levels because packed towers are more economic in this range.
The main difference in the costs of the two types of
scrubbers is the difference in capital cost. Table 15-1
is a comparison of the two scrubber types at moderate gas
flows and odor levels. It is apparent from Table 15-1 that
the capital investment for packed-bed scrubbers is approximately
twice that for a spray scrubber system. The electrical
power cost of spray scrubbers is also lower than packed bed
scrubbers if low-pressure nozzles (15 psig) are used. The
available space for a scrubbing system will also influence
the cost considerations and ultimate selection. Packed-bed
scrubbers can be placed inside a building or outdoors. Spray
scrubbers can be placed on a roof with minimal additional
structural support or can be built inside a plant. Based on
total over-all cost, spray scrubbers are more economical for
volumetric gas throughputs greater than 25,000 cfm.
\
15-1
-------
Table 15-1
COSTS OF PACKED BED VERSUS SPRAY SCRUBBERS FOR PLANT VENTILATING AIR
Operating conditions
Flow,
cfm
25,000
25,000
50,000
50,000
Average
odor units
Inlet Outlet
500 10
2000 20
500 10
2000 20
Number of
stages
2
3
2
3
Packed
Capital
investment,
$
96,100
120,800
181,400
230,300
bed scrubber
Power cost,
$/vr
4,360
5,860
8,910
11,730
$/1000 cfm-hr(l)
0.24
0.40
0.23
0.39
Number of
stages
2
3
2
3
Spray
Capital
investment ,
S
49,800
60,100
87,200
104,600
scrubber
Power cost,
S
3,290
4,540
6,460
8,900
$/1000 cfm-hr(l)
0.17
0.31
0.16
0.30
I
10
1Based on 4000 hours operation per year and equipment depreciated 13% per year.
-------
In Chapter 12, a total scrubbing system was presented
for treating rendering plant odors. The high-intensity odors
are more economically treated by a small packed-bed scrubber
followed by a large plant spray scrubber, than by sending the
high intensity odors directly to the ventilating air spray
scrubber. The major cost savings of the dual scrubber system
are in the capital investment and in the power cost.
15.2 Use of Venturi Scrubbers
Several rendering plants have installed venturi scrubbers
prior to packed-bed scrubbers to treat cooker noncondensibles.
The purpose is not to remove gaseous odors, but to remove
particulates. In this project, we attempted to determine
whether particulate matter is an important source of odors.
"The experiments described in Chapter 5 indicated that particu-
lates were not an important source of odor in plant ventilating
air. In spite of this finding, venturi scrubbers might still
be needed as a pretreatment step before packed bed scrubbers,
to prevent entrained grease particles from fouling the packing.
In our laboratory scrubber experiments at Des Moines, we observed
no particulate deposits on the packing even when scrubbing
high-intensity air that normally goes to an incinerator; but
we did observe some deposits of grease in a 200-ft line that
connected our scrubber with the air source.
Prokop (1974) reported that 5 Ib/min of particulates were
removed from the off-gases from a poultry feather dryer. Such
a loading would rapidly foul a packed-bed scrubber.
We suspect that the need for a venturi scrubber depends
on the type of condenser used. A direct-contact (barometric)
condenser will remove much of the greasy particulates from
the condenser off-gases. Mill, et al. (1963) reported that a
direct-contact condenser removed 99% of the odor from cooker
off-gases. An indirect (tube-and-shell) condenser was reported
to remove 50% of the odor, and it will probably remove less
15-3
-------
of the particulates. On the other hand, the indirect conden-
ser allows the grease removed to be recycled to the process,
whereas the direct condenser results in a larger effluent
stream that may cause a wastewater disposal problem.
The venturi scrubber acts as a second condenser of the
direct-contact type. Plain water is usually used for the
venturi scrubbing solution, and it is sent directly to the
drain, just as in a direct-contact condenser. Thus, a
venturi scrubber may be needed as a pretreatment before a
packed-bed scrubber for process air, only if the plant does
not have a direct-contact condenser. The decision should be
based on tests of particulate level at the individual plant.
The cost of the venturi scrubber will not be excessive if it
treats only the noncondensibles from the cooker, which usually
amounts to only a few hundred cfm.
Another reason for using a venturi scrubber might be to
further lower the temperature of the gas to a packed-bed
scrubber. Tests reported elsewhere in this report indicate
the holding capacity of scrubber solutions may be reduced when
the temperature of the solution increases. Any type of spray
scrubber could provide precooling with once-through water.
15.3 Design Odor Levels
The design of pollution control equipment must be based
on achieving acceptable emission quality during the-periods
of maximum loading. During our tests at Des Moines, we found
that the air odor levels varied by a factor of 100, and some-
times by a factor of 10 within a period of an hour.
Target emission odor levels depend on the dispersion char-
acteristics of the plant environs, and the distance to neigh-
bors. These matters are discussed in a recent EPA Report
(Osag, et al., 1974). For example, if a plant is located one
mile from its nearest neighbors, an emission of 25,000 cfm
at 100 to 200 odor units will be diluted by atmospheric dispersion
15-4
-------
to acceptable levels. The atmospheric dispersion can be aided
under most weather conditions if the scrubber air is emitted
from a stack 50 ft high. The use of a stack without a scrubber
will not eliminate odor problems during periods of adverse meteoro-
logical conditions. For larger emissions (100,000-150,000 cfm),
odor levels lower than 100 odor units may be necessary, especially
where the plant is located very close to a residential or light
business area.
The design models presented in Chapters 10 and 11 are
based on the measured odor levels at the Des Moines rendering
plant. The cooker off-gas after the condenser averaged 8,000
odor units during the tests performed, but the maximum odor level
was slightly greater than 50,000 odor units. The plant ventila-
ting air averaged 500 odor units, with a maximum level of
5,000 odor units. Every plant will have different odor levels,
depending upon how many of the high odor intensity sources are
collected and treated separately from the ventilating air.
15.4 Corrosion Resistance
The.use of sodium hypochlorite in an alkaline solution as
the scrubbing liquid requires that the materials of construction
be selected very carefully. (This is also true of some of the
scrubbing liquids tried by certain plants such as chlorine water.)
It is known that stainless steels are subject to pitting corro-
sion with solutions that contain chlorides. Fiberglass-reinforced
plastic (FRP). is recommended for the liquid storage tanks and
the packed-bed scrubbers. Vinyl esters are generally recommended
by suppliers as the material from which FRP should be made for
exposure to NaQCl solutions. Fabrication of all FRP structures
should conform with the proposed Product Standard for Custom
Fabricated Reinforced Polyester Corrosion Resistant Process
Equipment TS-5687. In particular, they should be designed with
adequate^strength in case the entire tower is flooded with the
scrubbing liquid. Manufacturers of FRP tanks should be familiar
with these proposed standards.
15-5
-------
PVC has generally been used for process piping for rendering
plant scrubbers. No corrosion problems have been reported, but
failure due to vibration has been known to occur.
The horizontal spray scrubber at the Des Moines plant is
one of 30 installations of this type. Part of the scrubber is
type 304 stainless steel, while part of the scrubber is type
316 stainless steel. An aluminum fan is used after the first
stage. The use of sodium hypochlorite, with excess caustic in
these units will reduce the expected equipment life because of
high corrosion rates.
Any entrainment from the first stage onto the aluminum fan
will result in excessive corrosion rates. Rabald (1968)
reported that even dilute concentrations of caustic and/or
hypochlorite results in high corrosion rates of aluminum. The
problem has been solved by custom-coat ing the fan with vinyl
ester resin.
Caustic presents no problem to stainless steel; however,
severe pitting of type 304 stainless steel will occur when
subjected to sodium hypochlorite solutions. Type 316 stainless
steel is also subject to pitting, but not nearly to the extent
of type 304. For many situations where sodium hypochlorite
solutions are used, type 316 will be suitable (Mellan, 1966).
The pitting rate will be a function of the hypochlorite concen-
tration, the pH, and other chemicals present in the water.
No quantitative information could be found in the literature
concerning the effect of hypochlorite concentration and the pH
on corrosion rates. A representative of a chlorine company
(S.P.D. Services, Forest Park, Illinois) was contacted to
ascertain if they had had any experience in this area. Although
they had no published data, it has been their experience that
the corrosion rates increase as the hypochlorite level increases
for both type 304 and 316 stainless steel. This firm felt that
15-6
-------
the alkaline pH reduced the corrosion rates and that at concen-
trations from 0.03 to 0.17» hypochlorite the pitting problems
would be minimal for type 316 stainless steel.
In the recommended continuous blow-down system, the hypo-
chlorite level is maintained above 0.1%; thus, neither type 304
nor 316 stainless steel is suitable for constructing new scrub-
bers. Serious consideration should be given to the construction
of horizontal spray scrubbers of fiberglass reinforced plastic
(FRP). Process Equipment Corporation (1974) estimated that the
total cost of FRP construction for horizontal scrubbers would
be slightly less than constructing the scrubbers of type 304
stainless steel. The expected life of the scrubber would be
extended significantly with the FRP because of the reduced
corrosion rates.
For existing stainless steel scrubber installations, a
plastic coating can be sprayed on the inside of the scrubber to
protect the steel from the liquid solutions. The cost of
spraying a 100,000 cfm, three-stage scrubber is approximately
$10,000 for the complete job. The aluminum fan can be spray-
coated with the same material. Before coating, the metal must
be thoroughly sand-blasted to prepare the surface. The recom-
mended coating thickness is 35 to 40 mils, and this results in
a coating that is both mechanically and chemically resistant.
Coatings recommended by manufacturers for alkaline hypochlorite
solutions are vinyl ester containing glass flake filler, or
polyester containing similar filler. It is standard practice to
spark-test the coatings to insure that there are no pin-holes.
Scrubbers that have already developed leaks can be coated,
since the leaks are usually from pinholes, and the metal is still
sound.
Existing scrubbers that have not been coated can be
operated with sodium hydroxide scrubbing solutions at pH > 12.
I
Although! this liquid is not as effective as hypochlorite, test
data in this report indicate it is effective against most
15-7
-------
odorants. Low concentration hypochlorite may be used in a type
316 stainless scrubber section, although quantitative corrosion
data are not available. In such cases, sodium silicate has
been suggested as a corrosion inhibitor, although data are
not available. Uncoated aluminum fans must not be subjected
to mist from either alkaline or acid solutions.
15.5 Specifying the Auxiliary Equipment
When designing a scrubber system, careful attention should
be paid to the following matters:
Liquid Pumps
Pumps should be properly sized to deliver the
specified amount at the specified head loss.
Over-design of pumps leads to throttling the
discharge valve to obtain the proper rate.
Throttling the discharge valve results in an
increase in head loss and the consumption of
excess electricity.
Minimizing the Liquid Head Loss
When scrubbers are placed on the roof of
a building and the liquid tanks are placed
on the floor inside the building, the pumps
must lift the liquid in excess of 30 ft.
When the liquid pumping rate is high, which
is the case with large scrubbers, the annual
power costs can be considerable. The system
should be designed to minimize the height
difference from the top of the liquid tank
and the scrubber nozzles to reduce the pump
horsepower, and thus the annual power cost.
Liquid Drainage from Tower
The return line from the tower to the liquid
holding tank must be properly sized to prevent
the liquid from building up in the tower.
When the liquid accumulates in a tower, the
liquid pumping rate must be reduced and the
scrubber efficiency decreases. Of course,
the use of larger nozzles than intended can
overtax the return line capacity.'
15-8
-------
pH Control
Because of the importance of maintaining
the pH in the last stage at pH 12, auto-
matic pH control should be included in each
plant.
15.6 Water Pollution Implications
The use of sodium hypochlorite scrubbing solutions with
an alkaline pH does not present any serious water pollution
problems. Any sodium hypochlorite discharged to the sewer
will continue to oxidize organics in the wastewater. (Prechlor-
inating in the pumping stations and in the primary sedimentation
basins is common practice in municipal wastewatp.r treatment
plants to inhibit the production of H2S, to prevent
corrosion and septic odors in the sewerage system.)
The alkaline pH will not present any problems if a continu-
ous blow-down approach is utilized. However, if a batch system
is utilized, and the tanks are rapidly discharged to the sewer,
any grease in the sewer will be emulsified. Emulsified grease
will not be removed through the grease collection system.
Thus, alkaline solutions (pH 11-12) must be neutralized prior
to discharging if the tanks are rapidly dumped.
15.7 Summary
Based on the information provided in this report, a render-
ing plant should be able to estimate the cost of a scrubber
system. The accuracy of preliminary cost estimates based on
general equations such as are presented in this report is
usually +30%. The cost information can be used to select the
optimum equipment design, but the final cost will be determined
by obtaining bids for the chosen scrubber.
15-9
-------
REFERENCES
Aries, R. S. and Newton, R. D., Chemical engineering cost
estimation, McGraw-Hill, New York, 1955.
Burgwald, T., Dravnieks, A., Krotoszynski, B. K., Whitfield, J.,
and O'Donnell, A., High-speed collection of organic vapors from
the atmosphere, Env. Sci. & Tech. 5 (Dec.) 1220-22 (1971).
Ceilcote Company, Standards and costs - gas absorption and
pollution control equipment, Bulletin 1200, Berea, Ohio.
Ceilcote Company, "Tellerette manual," Berea, Ohio.
Delavan Mfg. bulletin 1118C-372, Des Moines, Iowa, 1972.
Doty, D. M. , Snow, R. H. and Reilich, H. G., Investigation of
odor control in the rendering industry, EPA Report R2-72-088
1972.
Dravnieks, A. and Prokop, W. H., Source emission odor measure-
ment by a dynamic forced-choice triangle olfactometer, j.
Air Poll. Control Association. 25_ (1), 28-35, 1975.
Eckert, J. S., No mystery in packed-bed design, Oil Gas J.,
Aug. 24, 1970.
Graham Mfg. Co., Bulletin 87OA, 1971.
K. A. Steel Corp., private communication, August 1974.
Chemical Engineering, current issues, McGraw-Hill, New York.
Mellan, I., Corrosion resistant materials handbook, Noyes
Development Corp., Park Ridge, New Jersey, 1966.
Michalak, S. E. and Leite, F. B., On-site generation of hypo-
chlorite, J. Water Pol. Control Fed. 44 (9) (1972).
Mills, J. L., Walsh, R. T., Luedtke, K. D. and Smith, L. K.,
Quantitative odor measurement, J. Air Pol. Control Assoc. 13
(Oct) 467-476 (1963). —
Murthy, B. N. , Odor control by wet scrubbing: selection of
aqueous reagents, 66th Ann. Mtg. of APCA, Chicago, 111., 1973.
Norton Chemical Process Products Division, Bulletin S-32,
Akron, Ohio, 1969.
Palin, A. T. , Chemistry and control of modern chlorination,
LaMotte Chemical Products Co., Chestertown, Maryland 21620, 1973
R-l
-------
Perry, J. H., ed., Chemical engineers' handbook, 4th Ed.,
Chapter 18, 1963.
Peters, M. S. and Timmerhaus, K. D., Plant design and economics
for chemical engineers, McGraw-Hill, New York, 1968.
Popper, H., Modern cost engineering techniques, McGraw-Hill,
New York, 197TT
Process Equipment Corp., private communication, August 1974.
Prokop, W. H., Wet scrubbing of inedible rendering plant odors,
Paper presented at Odor Control Technology Specialty Conf.,
APCA, Pittsburgh, Pa., March 1974.
Rabald, E., Corrosion guide, Elsevier Publishing Co., New York
1968.
Sawistowski, H., Flooding velocities in packed columns operating
at reduced pressures, Chem. Eng. Sci. 6_, 138-40 (1957).
Spraying Systems Co., DWG #11825-9, Bellwood, Illinois, 1966.
Treybal, R. E., Mass transfer operation, McGraw-Hill, New York,
1967.
U.S. Stoneware Co., Packed towers, Akron, Ohio, 1963.
U.S. Stoneware Co., Packed towers, Akron, Ohio, 1970.
R-2
-------
APPENDIX 1
COMPARATIVE SCRUBBING TEST RESULTS
-------
^ 1 2nd STAGE ^
1 1 Nann "^"
1st STAGE _ „ , , .
NaOCl Table A
3rd STAGE RESULTS OF SCRUBB
^^^^^^ IT /\ ^3^^ iXU L> W J-J A W NX J> k^ V* J.VW U U
Hz°2 LABORATORY SCRUBBER.
ING TEST 31-1
TIME: 1/2 HOUR
Outlet from
Inlet #2 (NaOCl)
Odor EDso 700
GLC peaks, yg/liter air
1. Dimethyl sulfide 79.34
2. Pungent U.P. 90 21.41
3. Dimethyl disulfide 1.43
4. Butyric acid 7.65
5. Propylene sulfide (cabbage 880) 4.09
6. Bitter 655 37.40
7. Burnt 1233A (pyrazine) 2.98
8. Burnt 1233B (pyrazine B) 6.30
9. Rendering 1335 214.0
10. Fatty 1545 67.84
11. Burnt 1687 (quinoline) 23.48
12. Sulfur 660
13. Cheesy 2002 34.70
14. Unpleasant 2582 12.43
15. Fatty acid 1980 10.26
16. Fatty acid 2160 9.94
25
4.47
0.47
0.04
0.65
0.21
0.93
0.10
0.90
15.91
4.27
1.00
-
2.13
1.03
0.61
0.50
Outlet from
#3 (H202)
265
14.48
4.60
0.38
1.20
0.58
11.40
0.90
2.30
32.70
4.47
10.52
-
6.23
1.97
1.56
0.88
1 •* 2
% Removal
96
94
98
97
92
95
98
97
86
93
94
96
-
94
92
94
95
1 -»• 3
% Removal
62
82
79
73
84
86
70
70
63
85
93
55
-
82
84
85
91
Average percent removal of each component
94
79
-------
>
N>
2nd STAGE
Mann ~"~
1st STAGE _ T ,
NaOCl Tab
L^ 3rd^STAGE _^ RESULTS OF SC
Inlet
Odor EDSO
GLC peaks, yg/liter air
1. Dimethyl sulfide 93.55
2. Pungent U.P. 90 18.62
3. Dimethyl disulfide 3.22
4. Butyric acid 23.33
5. Propylene sulfide (cabbage 880) 7.14
6. Bitter 655 25.72
7. Burnt 1233A (pyrazine) 4.50
8. Burnt 1233B (pyrazine B) 14.67
9. Rendering 1335 345.43
10. Fatty 1545 114.64
11, Burnt 1687 (quinoline) 78.20
12. Sulfur 660
13. Cheesy 2002 71.93
14. Unpleasant 2582 35.12
15. Fatty acid 1980 9.12
16. Fatty acid 2160 30.10
le A-2
RUBBING TEST
BER, TIME: 1
Outlet from
#2 (NaOCl)
_
9.53
0.65
0.77
3.37
0.75
3.25
1.29
-
31.62
11.75
14.11
-
14.65
7.45
0.92
3.19
31-2
-1/2 HOURS
Outlet from
#3 (H202)
_
3.13
2.03
0.19
0.24
0.03
6.17
0.26
0.29
11.05
2.40
0.53
-
0.83
0.28
0.26
0.35
1 •* 2
% Removal
^
90
97
76
86
89
87
71
-
91
90
82
-
80
79
90
89
1 -»• 3
% Removal
_
97
89
94
99.0
99.6
76
94
-
97
98
99.3
-
98.8
99.2
97
98.8
Average percent removal of each component
86
95
-------
>
1st STAGE _j --' T ,
NaOCl .,.
>3rd STAGE ^ RTTQTTT T
-------
2nd STAGE
""*" Mann "*"
1st STAGE , m -LI
NaOCl " Table
A- 4
> 3rdRSQAGE _>. RESULTS OF SCRUB^TMn TTT.QT
J-*Jk^*xj Ji.uk'.i.
TIME- 9
y J. J~iL JJ.J • j/
Outlet from
Inlet #2
Odor ED 50 590
GLC peaks, lag/liter air
1. Dimethyl sulfide 69.13
2. Pungent U.P. 90 25.28
3. Dimethyl disulfide 5 52
4. Butyric acid 4.63
5. Propylene sulfide (cabbage 880) 1.88
6. Bitter 655 37.21
7. Burnt 1233A (pyrazine) 1.02
8. Burnt 1233B (pyrazine B) g 91
9. Rendering 1335 193.77
10. Fatty 1545 40.93
11. Burnt 1687 (quinoline) 32.67
12. Sulfur 660
13. Cheesy 2002 28.32
14. Unpleasant 2582 11.93
15. Fatty acid 1980
16. Fatty acid 2160
(NaOCl)
45
4.73
1.32
0.33
1.35
2.40
1.32
0.27
2.19
25.81
10.22
12.94
_
11.53
6.72
-
—
31-4
HOURS
Outlet from
#3 (H202)
175
7.02
3.55
0.05
0.55
0.13
8.20
0.37
1.10
6.88
1.62
1.13
__
1.82
0.83
_
_
1 •* 2
% Removal
92
93
95
94
71
_
96
74
78
87
75
60
_
59
44
_
_
1 •*• 3
% Removal
70
90
86
99.1
88
_
78
64
89
96
96
97
_
94
93
—
_
Average percent removal of each component
77
89
-------
Table A-5
RESIDUAL ODORS FROM SPENT SCRUBBING SOLUTIONS - TEST 31-5
LABORATORY SCRUBBER
Inlet
Odor ED 50
GLC
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
peaks, ug/liter air
Dimethyl sulfide
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
Outlet from
#2 (NaOCl)
160
1
6
*0
1
0
<0
0
0
19
2
3
-
2
0
-
-
.34
.11
.01
.62
.58
.01
.34
.42
.85
.13
.15
.51
.78
8
3
<0
2
0
0
0
1
26
8
8
-
5
3
-
_
25
.20
.57
.01
.17
.33
.70
.44
.85
.85
.95
.43
.65
.17
Outlet from
#3 (H202)
220
0
0
<0
0
0
3
0
0
1
0
-
-
0
-
-
-
.44
.90
.01
.07
.09
.90
.15
.23
.17
.43
.10
-------
~, 2nd STAGE _>>
\ NaOCl
i 1st STAGE _ T
Nanrl
L^, 3rdMS™GE ->- RESULTS OF
NaUH LABORATORY SCR
Inlet
able A- 6
SCRUBBING TEST
UBBER, TIME: I/
Outlet from
#2 (NaOCl)
Odor EDSO 21,500 15
GLC peaks, pg/liter air
1. Dimethyl sulfide 11.82
2. Pungent U.P. 90 8.80
3. Dimethyl disulfide 1.46
4. Butyric acid 0.33
5. Propylene sulfide (cabbage 880) 0.22
6. Bitter 655 7.82
7. Burnt 1233A (pyrazine) 1.74
8. Burnt 1233B (pyrazine B) 1.75
9. Rendering 1335 19.25
10. Fatty 1545 3.74
11. Burnt 1687 (quinoline) 2.96
12. Sulfur 660 <0.01
13. Cheesy 2002 4.64
14. Unpleasant 2582 0.64
15. Fatty acid 1980 1.03
16. Fatty acid 2160 2.50
0.95
0.44
0.11
0.20
0.13
0.30
0.06
1.14
12.72
1.43
0.94
0.51
1.80
0.20
0.42
1.04
32-2
2 HOUR
Outlet from
#3 (NaOH)
35
5.09
4.40
0.26
0.10
0.04
3.38
1.15
-
2.07
0.41
0.06
3.37
0.34
<0.01
<0.01
0.25
1 -»> 2
% Removal
99.9
92
95
92
39
41
96
97
-
34
62
68
-
61
69
59
58
1 -*• 3
% Removal
99.8
57
50
82
70
82
57
34
-
89
89
98
-
93
98
99
90
Average percent removal of each component
69
78
-------
H2nd STAGE .
NaOCl
1st STAGE _ Table A
Mann laoo. «
3, 3rd STAGE _^ RESULTS OF SCRUBS
NaOH LABORATORY SCRUBBER.
ING TEST
TIME: 2
Outlet from
Inlet H2 (NaOCl)
Odor ED50 2,700
GLC peaks, yg/liter air
1. Dimethyl sulfide 6.73
2. Pungent U.P. 90 3<85
3. Dimethyl disulfide 5.29
4. Butyric acid 0.68
5. Propylene sulfide (cabbage 880) Q 75
6. Bitter 655 13.48
7. Burnt 1233A (pyrazine) 0.93
8. Burnt 1233B (pyrazine B) j<88
9. Rendering 1335 33.45
10. Fatty 1545 6<80
11. Burnt 1687 (quinoline) 2.15
12, Sulfur 660 2.40
13. Cheesy 2002 3.55
14. Unpleasant 2582 2.32
15. Fatty acid 1980 2.33
16. Fatty acid 2160 2.45
35
0.87
0.50
0.03
0.30
0.05
0.47
0.07
0.47
5.94
1.16
0.56
0.99
0.97
0.54
0.20
0.32
32-3
-1/2 HOURS
Outlet from
#3 (NaOH)
600
2.17
2.84
0.23
0.12
0.04
8.21
0.69
0.33
3.38
0.73
0.22
6.11
0.38
0.15
0.12
0.30
1 -»• 2
% Removal
98.7
87
87
99.4
56
93
97
92
75
82
83
74
-
73
78
91
87
1 -»• 3
% Removal
78
68
26
96
82
95
39
26
82
90
89
90
-
90
94
95
88
Average percent removal of each component
84
77
-------
,00
^ 2nd STAGE ^
•^^^^ BM^^^
^Jfi Of 1
Table
^ m ,
^ 3rd STAGE _^ RESULTS OF SCRUB
A- 8
BING TEST
, TIME: 8
Outlet from
Inlet #2 (NaOCl)
Odor ED50 5,000
GLC peaks, yg/liter air
1. Dimethyl sulfide 7.6!
2. Pungent U.P. 90 6.55
3. Dimethyl disulfide 5.63
4. Butyric acid 2.74
5. Prbpylene sulfide (cabbage 880) 4.37
6. Bitter 655 19.10
7. Burnt 1233A (pyrazine) 2.16
8. Burnt 1233B (pyrazine B) 4.21
9. Rendering 1335 64.72
10. Fatty 1545 36.06
11. Burnt 1687 (quinoline) 21.42
12. Sulfur 660 1.32
13. Cheesy 2002 17.71
14. Unpleasant 2582 9.16
15. Fatty acid 1980 6.89
16. Fatty acid 2160 56.09
450
1.21
0.16
<0.01
0.04
0.20
0.42
0.01
0.79
5.63
1.07
0.52
0.55
1.11
0.40
9.19
0.88
32-4
-1/2 HOURS
Outlet from
#3 (NaOH)
1400
2.77
5.15
0.23
0.65
0.20
10.83
0.55
0.85
3.98
1.67
0.46
4.94
1.05
0.41
0.36
0.59
1 -»• 2
% Removal
91
84
98
99.9
98.5
95
98
99.5
81
91
97
98
-
94
96
97
98.4
1 •* 3
% Removal
72
64
21
96
76
95
43
75
80
94
95
98
-
94
96
95
98.9
Average percent removal of each component
95
81
-------
I
VO
p_> 2nd STAGE ^
> 2nd STAGE >
MnnfM
1st STAGE _ ' _ ,,
NaOCl Table
*-* 3rloHAGE -* RESULTS OF SCRU
LABORATORY SCRUBBER
Inlet
Odor ED50 19,000
GLC peaks, ug/liter air
1. Dimethyl sulfide 8.75
2. Pungent U.P. 90 7.02
3. Dimethyl disulfide 3.25
4. Butyric acid 0.68
5. Propylene sulfide (cabbage 880) 0.39
6. Bitter 655 18.92
7. Burnt 1233A (pyrazine) 0.48
8. Burnt 1233B (pyrazine B) 1.52
9. Rendering 1335 26.65
10. Fatty 1545 5.83
11. Burnt 1687 (quinoline) 2.80
12. Sulfur 660 4.23
13. Cheesy 2002 3.47
14. Unpleasant 2582 0.91
15. Fatty acid 1980 1.08
16. Fatty acid 2160 9.13
A- 9
BEING
TIME:
TEST 32-5
17-1/2 HOURS
Outlet from Outlet from
#2 (NaOCl) *3 (NaOH)
300
1.10
0.51
0.16
0.04
0.11
1.40
0.04
0.03
4.67
1.12
0.36
5.73
0.72
0.22
0.33
0.42
370
8.45
6.80
2.71
0.11
0.14
18.57
0.46
0.09
3.75
1.22
0.27
30.26
0.60
0.14
0.20
1.40
1 + 2
% Removal
98.4
87
93
95
94
72
93
92
98
82
81
87
-
79
76
70
95
1 -»• 3
% Removal
98.1
3
3
17
84
64
2
4
94
86
79
90
-
83
85
81
85
Average percent removal of each component
86
57
-------
>
J 2nd STAGE
. , .,. * NaOCl
1st STAGE _ ToKiQ
NaOCl Table
"-> ^M o»AGE -*• RESULTS OF SCRU
NaUH T.AKnRATnPY SPPTTURFP
Inlet
Odor ED50 16,000
GLC peaks, yg/liter air
1. Dimethyl sulfide 3.56
2. Pungent U.P. 90 4.20
3. Dimethyl disulfide 0.33
4. Butyric acid 0.30
5. Propylene sulfide (cabbage 880) 0.64
6. Bitter 655 9.35
7. Burnt 1233A (pyrazine)
8. Burnt 1233B (pyrazine B)
9. Rendering 1335 49.95
10. Fatty 1545 3.65
11. Burnt 1687 (quinoline) 4.08
12. Sulfur 660 0.49
13. Cheesy 2002 3.34
14. Unpleasant 2582 1.17
15. Fatty acid 1980 4.13
16. Fatty acid 2160 18.44
A-10
BEING TEST 32-6
TIME: 24 HOURS
Outlet from
#2 (NaOCl)
240
0.39
0.25
0.03
0.04
0.08
0.48
0.03
-
7.34
0.98
0.32
0.98
0.62
0.25
0.49
0.33
Outlet from
#3 (NaOH)
360
2.23
3.78
0.20
0.28
0.05
10.86
0.31
-
12.45
0.90
0.54
23.51
0.72
0.26
0.48
0.46
1 -»• 2
% Removal
98.5
89
94
91
87
88
-
-
-
85
73
92
-
81
79
88
98.2
1 -" 3
% Removal
98
37
10
39
7
92
-
-
-
75
75
87
-
78
78
88
98
Average percent removal of each component
87
64
-------
.J2nd STAGE
N^nn
1st STAGE _
NaOCl .
3rd STAGE
^ NaOH ^ Table A- 11
RESIDUAL ODORS FROM SPENT SCRUBBING SOLUTIONS - TEST
LABORATORY SCRUBBER
Outlet from
Inlet #2 (NaOCl)
Odor ED50 2500 30
GLC peaks, yg/liter air
1. Dimethyl sulfide 0.33
2. Pungent U.P. 90 0.05
3. Dimethyl disulfide 0.02
4. Butyric acid 0.10
5. Propylene sulfide (cabbage 880) 0.04
6. Bitter 655 0.02
7. Burnt 1233A (pyrazine) 0.04
8. Burnt 1233B (pyrazine B) 0.03
9. Rendering 1335 1.59
10. Fatty 1545 0.13
11. Burnt 1687 (quinoline) 0.16
12. Sulfur 660 0.90
13. Cheesy 2002 0.13
14. Unpleasant 2582 0.03
15. Fatty acid 1980 0.31
16. Fatty acid 2160 0.26
32-7
Outlet from
#3 (NaOH)
130
0.34
0.73
0.15
0.12
0.11
1.19
0.54
1.78
0.34
0.32
2.18
0.19
0.31
0.16
0.27
-------
APPENDIX 2
DERIVATION OF CHEMICAL CONSUMPTION EQUATIONS
-------
DERIVATION OF CHEMICAL CONSUMPTION EQUATIONS
1. CHLORINE REACTING WITH THE ODORANTS
During the plant life test (Test #34) the average chlorine
feed rate was 12 Ib/hr. From Figure 10-4,p.10-20, the average chlorine
concentration discharged to the sewer was 0.0870 during a 24-hr
period after the system attained equilibrium. The average
blow-down rate was 2 gpm. Thus, the total chlorine used for
reacting with odorants was:
12 Ib 0.000812 gal I 60 mini 8.34 Ib = ,, ~ Ib
HR I min | HR | gal HR
The above chlorine consumption was based on the scrubber's
operation of 24 hours a day. During the day, when the plant
was not processing any material the odor removal was assumed
to be 100 odor units. During the average 13 hours a day the
plant processed material,plus one hour for clean-up, the measured
odor levels were averaged. The actual data are presented in
Section 8. The average odor removal during the evening was
720 odor units. The average odor removal over the 24-hr
period would be:
Average odor removal = [14 hr(720 o.u.)1 +JW hr(100 o.u.)]
= 462 odor units
Thus, assuming the chlorine reaction is a function of the Odor
Quantity (odor units removed x ft3) the following general rela-
tionship would be valid.
Chlorine demand = (CFM> (Hours) (Constant)
Now, if hours = 1 and combining the with (Constant)
Chlorine demand = f0018 (CFM) (Constant ')
A2-1
-------
Substituting the known chlorine demand, odor units removed, and
CFM, the constant can be evaluated
Constant' = (462) (67 ,000) = T0~6~
Similar calculations were made for several of the other tests.
All the computed constants came out between 0.31 and 0.36, with
the average value of 0.32.
Thus , the annual chlorine consumed for odor removal becomes
(CFM) (YO - YE ) (HRS)
where
CFM is the volumetric gas throughput, ft3/min
YO is the inlet odor level, odor units
YE is the scrubber outlet odor level, odor units
HRS is the total number of hours the system is
operated a year
A2-2
-------
2. SCRUBBING CAPACITIES
2.1 One-Stage Scrubber
During Tests 19 and 20, the laboratory scrubber was used to
investigate the performance of two stages of sodium hypochlorite
followed by one stage of caustic. After Test 20, the first
stage hypochlorite solution was transferred to the last stage,
and a fresh batch of caustic was prepared for Tank #1. On the
next test, no bag samples were collected; however, the gas
chromatograms indicated very poor removals across the final
stage, with several compounds showing an actual increase in
concentration across the third stage. The next test, Test 22,
found over half of the compounds increasing across the final
stage and gave a totally unacceptable outlet odor level of
38,000 odor units. The data from Tests 19-22 are presented in
Tables A2-1 through A2-4.
The, solution was exposed to the incinerator inlet air
for a period of 2-1/2 hours while it was in the first stage.
The additional amount of odorants absorbed into this solution
in the third stage was very small in comparison. Thus, the odor
removal across the first stage averaged 30,000 odor units.
30,000 o.u.
2.5 Hr
60 min
HR
17 ft3
Min
Odor Quantity = 76,500,000 odor units - ft3
This is the amount absorbed by the 25 gallons of first stage
solution; thus,
o ,v • ^ ..,_ 76,500,000 o.u.-ft3
Scrubbing Capacity = 25 gal
= 3x106
This value was checked using Test 30 data, where an average
9000 odor units were removed effectively for a 5-hr period from
a 28-cfm gas stream. Test 30 also yielded a scrubbing capacity
of 3x106.
A2-3
-------
j>
i>
Table A2-1
RESULTS OF SCRUBBING TEST 19
LABORATORY SCRUBBER
(a)
Inlet
Odor EDso
GLC peaks, yg/liter air
1. Dimethyl sulfide
2. Pungent U.P. 90
3.
4.
5.
6.
7.
8.
9,
10.
11.
12.
13.
14.
15.
16.
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
(b)
After
1st stage
7400
Samplers
overloaded
1
2
-
1
1
66
15
7
-
14
3
3
3
.59
.57
.80
.61
.72
.79
.91
.62
.45
.41
.40
0
0
-
0
0
7
1
0
-
0
0
-
-
.14
.56
.26
.37
.25
.56
.48
.83
.39
(c)
After
2nd stage
4
3
0
0
0
-
0
0
1
0
0
-
0
0
0
0
.03
.07
.38
.08
.85
.09
.13
.37
.72
.28
.47
.15
.10
.28
(d)
Outlet
400
2.56
2.28
0
0
0
-
0
0
0
0
0
-
0
0
0
0
.47
.05
.57
.05
.07
.49
.12
.12
.08
.06
.13
.09
% removal
95
-
97
78
-
97
96
99
99
98
-
99
98
96
97
.3
.2
.5
.5
.3
Average percent removal of each component
96
-------
fe
Ul
Table A2-2
RESULTS OF SCRUBBING TEST 20
LABORATORY SCRUBBER
Odor EDso
GLC
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
peaks, Ug/liter air
Dimethyl sulfide
Pungent D.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
(a)
Inlet
76,000
tt
0)
•d
cd
0
rH
ft)
£
8
1
cd
OT
(b)
After
1st stage
-o
0)
•o
cd
o
M
-------
N>
Table A2-3
RESULTS OF SCRUBBING TEST 21
LABORATORY SCRUBBER
(a)
Inlet
Odor ED so
GLC peaks, ug/ liter air
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Dimethyl sulfide
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
-
4
0
0
0
-
0
1
12
2
1
-
4
0
1
2
.46
.71
.38
.12
.88
.14
.82
.35
.48
.67
.51
.74
.97
(b)
After
1st stage
3.
1.
0.
0.
0.
-
0.
0.
3.
1.
0.
-
1.
0.
0.
0.
12
84
34
20
17
80
50
95
65
67
34
38
29
84
(c)
After
2nd stage
2
1
0
0
0
-
0
0
0
0
0
-
0
0
0
0
.75
.26
.33
.12
.04
.39
.21
.64
.26
.12
.06
.08
.41
.22
(d)
Outlet
2
0
0
0
0
-
0
0
0
0
0
-
0
0
0
0
.69
.78
.11
.13
.03
.23
.14
.41
.15
.13
.13
.18
.06
.12
% removal
-
83
85
66
75
-
74
88
97
94
91
-
97
65
97
96
Average percent removal of each component 85
-------
NJ
I
Table A2-4
RESULTS OF SCRUBBING TEST 22
LABORATORY SCRUBBER
(a)
Inlet
Odor ED so
GLC
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
peaks, ug/liter air
Dimethyl sulfide
Pungent U.P. 90
Dimethyl disulfide
Butyric acid
Propylene sulfide (cabbage 880)
Bitter 655
Burnt 1233A (pyrazine)
Burnt 1233B (pyrazine B)
Rendering 1335
Fatty 1545
Burnt 1687 (quinoline)
Sulfur 660
Cheesy 2002
Unpleasant 2582
Fatty acid 1980
Fatty acid 2160
46
13
4
1
2
1
-
0
4
53
16
6
-
9
9
1
6
,000
.25
.64
.67
.22
.52
.75
.53
.14
.50
.02
.56
.36
.92
.05
(b)
After
1st stage
(c)
After
2nd stage
(d)
Outlet
38,000
5
2
0
0
0
-
0
1
10
1
0
-
1
0
0
1
.96
.57
.70
.22
.44
.66
.59
.37
.86
.45
.51
.51
.31
.35
2
1
0
0
0
-
0
0
2
0
0
-
0
0
0
0
.92
.30
.24
.08
.32
.19
.96
.11
.37
.08
.23
.08
.07
.22
4.
0.
0.
0.
0.
—
0.
0.
7.
1.
1.
-
1.
0.
0.
0.
63
60
14
21
09
26
52
15
22
18
66
30
25
25
% removal
17
65
87
92
91
94
-
65
89
87
93
80
-
83
97
87
96
Average percent removal of each component 86
-------
2.2 Two-Stage Scrubber
In Test 32, a two-stage sodium hypochlorite system was
operated for 24 hours while effectively removing the odorous
compounds. During this period of time, the odor level reduction
averaged 12,000 odor units across the two-stage laboratory
scrubber. Each scrubbing solution contained 20 gallons.
Odor Quantity =
12,000 o.u./17 ft3/24 HR/60 min
/ min
HR
= 293,000,000 o.u.-ff3
Scrubbing Capacity - Odor ^nsX
= 7x106
2.3 Three-Stage Scrubber
During the Plant Liquor Life Test, utilizing the continuous
blow-down system, the average odor reduction was 540 odor units,
as developed in Appendix 1.
Odor Quantity = 540 odor units 67,000
= 36,000,000 o.u.-ft3/min
Scrubbing Capacity - °^ {*%£%,
gpm
= 18x106
Because only one or two samples were measured each day, the
average odor reduction is subject to considerable error. The
laboratory liquor life test using the continuous blow-down and
make-up system was also checked.
Odor Quantity - 15.000 o.u./l^/eO^in
= ISxlO6 o.u.-ft3/hr
Scrubbing Capacity =
Odor Quantity
1.5 gal/hr blow-down
= 10x106
A2-8
-------
The wide variation for scrubbing capacities is probably
due to the uncertainty in the average odor reduction. In
Chapter 9, additional reasons why the scrubbing capacity in the
laboratory liquor life test was lower than in the plant test are
provided. However, an average of the two values will be used.
Three stage scrubbing capacity = 14xl06
A2-9
-------
3. ANNUAL CHLORINE LOSS TO SEWER
Based on Section 10.3.1.2, the chlorine loss to the sewer
as a function of the total chlorine consumed is
Chlorine loss
1 stage 54%
2 stages 25%
3 stages 8%
The amount of chlorine reacted with the odorants (C12-RX) is
determined by the odor level and gas throughput. The chlorine
loss to the sewer (C12-LOSS) can be written in terms of C12-RX.
Based on the above table, the following was developed:
Ratio
Katl°
- °'33
3 Stages 100 -8* = °'085
The following equation can be written
C12-LOSS = f(x)(Cl -RX)
where f(x) should equal C12-LOSS/C12-RX as shown in the above
table. Plotting C12-LOSS/C12-RX versus the number of stages on
semi-log paper, a straight line is developed having a slope of
-1.278 and an intercept of Jin 4.2. Thus, the following rela-
tionship holds
C12-LOSS = (C12-RX)4.2 exp(-l.278NS)
where NS is the number of stages.
A2-10
-------
4. CAUSTIC CONSUMPTION DUE TO REACTION WITH ODOROUS COMPOUNDS
In Test 32, using the laboratory scrubber, the caustic
levels were monitored throughout the test ".using pH titrations .
The pH in both the first and the second stages was held at the
same level, thus both would remove the same amount of C02 . The
different caustic consumption rates between the first stage and
the second stage would be due to the greater reaction of odorous
compounds in the first stage where odor levels are higher.
For the last six hours of Test 32, the caustic level dropped ..
1.4 Ib in the first stage and 1.2 Ib in the second stage.
Because the first stage was removing an average of 12,000 odor
units per cfm (93%) while the second stage was removing an
average of only 800 odor units per cfm (9370) , the organic content
in the second stage was insignificant (less than 10%) as com-
pared to the organic content in the first stage. Thus, 0.2 Ib
were consumed in the first stage due to the higher odor level.
0.2 Ib NaOH = Constant(YO - YE) (CFM) (HRS)
or
_ _ 0.2 pounds _
_
(12,000 odor units) (28 cfm) (6 hours)
0.1
Thus , the final equation becomes :
Annual caustic consumption due _ 0.1 /vr> v-,v /n?vr\ /•uT>c^
to reaction with odorants ~ TD^
-------
APPENDIX 3
CALIBRATION OF THE HORIZONTAL SPRAY
SCRUBBER MODEL
-------
CALIBRATION OF THE HORIZONTAL SPRAY
SCRUBBER MODEL
1. INTRODUCTION
As mentioned in Chapter 11, the model for spray scrubbers
utilizes the penetration theory to calculate the mass transfer
of odorous compounds from the gas phase to the liquid droplets
Data collected during the first part of this program was used
to check the model.
2. NOMENCLATURE FOR SPRAY SCRUBBER
a Area available for mass transfer,
surface area of drops, ft2/ft3
6 Contact time of drops in scrubber, hr
Diff Diffusivity of odorant vapor in air, ft2/hr
Flow of liquid per nozzle, gal/min
Height of scrubber, ft
Length of a scrubber stage, ft
Xx Log of odor reduction ratio
k Mass transfer coefficient, Ib-mol/hr-ft2-atm
O
GM Molar gas flow rate, lb-mol/hr-ft2
n. Number of collisions per drop per sec
W Width of scrubber, ft
N Number of nozzles per sq ft of scrubber cross
oz section
T Temperature of gas stream, °R
CFM Total air flow, cu ft/min
PT , Total pressure, atm
A3-1
-------
3. DESIGN EQUATION FROM PENETRATION THEORY
The success of the horizontal spray scrubber depends on the
use of low air velocities such that the droplets remain in the
scrubber for a second. Previously designed spray scrubbers
either have insufficient droplet area/unit volume or insufficient
contact time. Although the nozzles point upstream, the stopping
distance of droplets is such that they soon acquire the air
velocity and drift down the scrubber to the exit. Therefore,
the droplet contact time was set equal to the air contact time
in the scrubber.
According to the methods of calculation used, each droplet
acts independently of the others. Actually the droplets can
collide and coalesce. This effect was computed to see if it is
significant, but it was not compensated for.
The equation for mass transfer is the same as that for a
packed scrubbing tower (Eq 4, p. 10-6). Only the method of
calculating the mass transfer coefficient is different.
The mass transfer coefficient is predicted from penetration
theory based on the transient diffusion of odorant through the
gas layer adjacent to each drop. The result is:
Ib-mol
g RT /V ir6 hr-ft^-atm
The contact time, Q of droplets in the air depends on the
length of scrubber ^ and the gas (and droplet) velocity.
The resulting design equation is:
XLVPT 2 T a /DiffXLW Z
_g
Gm 460 />/ 60TT cfm
/DiffXjW
/V 60ir eft
We now design the scrubber by fixing the length X, = 8 ft,
as is done by Air Conditioning Corporation, and vary the area
WZ. In doing so, we allow the number of nozzles per unit of
cross-sectional area N Q.to remain constant; the total number
OZd.
increases with area. The liquid flow per nozzle L is also
*• *• Tir>^
A3-2
-------
fixed; so changing the height changes the liquid flow and
increases the droplet area. It also decreases the air velocity
and increases the contact time. As a result, the capacity
3/2
depends on area ' . These substitutions, are included in the
design computer program.
The flow rate per nozzle, L___, is another variable that
T1OZ
affects the design, and hence the scrubbing cost. The program
provides for reading in four values of L ; it computes the
scrubber design and cost for each, and chooses the most economic
one. Other variables, such as~spray nozzle pressure, affect
droplet size and power. These can be explored by trial calcula-
tions with the program.
Costs of each component of equipment and operating costs
are similar to those used for the packed tower scrubber. These
equations were given in Report No. EPA-R2-72-088 and are
included in the computer program.
The drop collision rate is estimated by applying equations
from the kinetic gas theory. In most cases, the decrease in
number of drops due to collision is small, but it could be
large if a high droplet concentration is produced.
A3-3
-------
4. COMPARISON OF MODEL WITH EXPERIMENT
The user of the model specifies the desired performance,
the number of stages, and the length of each contact chamber.
The model then computes the cross-sectional area (and associated
number of nozzles) required to achieve the specified removal.
The scrubber at the Des Moines rendering plant had the following
characteristics:
52 ft2 cross-sectional area
6 nozzles/ft2
1.2 gpm/nozzle (measured)
60,000 acfm (measured)
3 stages, each with 8-ft long contact chamber
Using the above design features, one need only specify the
desired odor removal and compare the cross-sectional area
obtained. At the time of this verification, only ten scrubbing
tests had been performed using the plant scrubber. The removals
from the best four tests were averaged to represent the expected
performance utilizing the optimum scrubbing solutions.
Test No.
8
12
15
16
Inlet ED50
320
360
660
300
Outlet ED50
20
35
25
70
Average % Removal
O'dor Panel
7o Removal
94
90
96
77
89%
Thus, by specifying a removal of approximately 90%, one
would expect the model to predict a required cross-sectional
area near 52 ft2 if the model is accurate. The results of the
computer run, shown in Table A3-1, predicted a required cross-
sectional area of 58.8 ft2 for 90% removal. Thus, the model
predicted within 13% of the actual dimensions of the Des Moines
scrubber.
A3-4
-------
APPENDIX 4
DEVELOPMENT OF THE PACKED-TOWER
MASS TRANSFER EQUATION
-------
DEVELOPMENT OF THE PACKED-TOWER
MASS TRANSFER EQUATION
In Section 10.2, the mass transfer equation for packed
towers was developed in detail, with the resulting final
equation:
r T )0.27
KGa = Constant U-^y PKGA
v * /
where
Kj-.a is mass transfer rate, Ib-moles/hr-ft3-atm
L is liquid mass flux, lb/hr/ft2
PKGA is ratio of the mass transfer rate (K^a)
of the packing used, compared to
the mass transfer rate of 1-1/2-in.
Intalox saddles
In order to determine the value of the "Constant", the experi-
mental results from the three-stage laboratory scrubber were
utilized. The laboratory scrubber utilizes 1/2-in. Intalox
saddles; thus:
K,,a 1/2-in. Intalox saddles
PKGA =
Ua 1-1/2-in. Intalox saddles
Lr
at the same conditions. Norton,(1969) presents mass transfer
data for both the 1/2-in. and the 1-1/2-in. Intalox saddles.
At a liquid mass flux rate of 2,000 lb/hr/ft2, which is the
same rate used in the laboratory scrubber, the ratio of KGa
is given at 2.06. This ratio only changes 10% when increasing
the liquid rate to 6,000 lb/hr/ft2. Thus, a value of PKGA
of 2.06 was utilized to determine the constant.
At the time of the development of the "Constant", only
a limited number of tests had been performed using the labora-
tory scrubber. As various scrubbing liquor combinations were
being compared for their effectiveness, only the most efficient
removals were averaged; these values are tabulated below.
A4-1
-------
Odor Panel
Test No. % Removal
7
10
11
25
23
Average
99.7
98.6
99.1
99.8
99.8
99.4
For the computer model, the inlet and outlet odor levels are
specified and the optimum scrubber dimensions are computed.
Using an inlet odor level of 10,000 odor units and an outlet
odor level of 58 odor units (99.4% removal), the "Constant"
term was varied until the proper scrubber dimensions were
predicted. Using the following equation, the computer model
predicted that 3 stages of 2 ft of packing would be required:
L I0'27
= 4-5 PKGA
The results of the computer model are presented in Table A4-1
A4-2
-------
Table A4-1
LABORATORY SCRUBBER DESIGN CALCULATIONS
i.rrprp OF STAPFG IN ST.rnn^r ir, 3
TYPF OF PAfKIMR AMP ITS PPPPFPTIF1 "" "
1/2 INTAI.CX SAPPI FS
POPOSITY APFA FOIMVAIFNT CONSTANTS FOP DFITAP
SPFT/PITT PIAf'FTFP(FT) Al PI'A RFTA
•75P 2TF.P .PS 1.01(0 .370
OPFPATINn PONPITIPr'S FOP
FI.PW RATF PONPFNTrATIPr' HPMPY SPIIMIPTNP .,,.. ,.
~"N INI.FT PI'TI FT prnprrs r
17. _1.9POO. _ 5f. 0. 2.053 7fit
UNIT PPSTS FPP $r
POl/FP I'Alf'TFMANPP PAPKINP
*/KI;II */snpT(PFf')
.025 8.000 ?f
i/All'FS Pl'Plf'P PPTIf'IZAflPN
Aff.'l.'AI. POST
.II^IMP PA^ FipnpifP PATF PAP HTCPCFT) K(R)/1
1^3.
iro.
15°.
j: is.".
.«-- 15!?-
~ PPFPATINP
TIIF I1ATFPIAI. PF
1000.
1250.
1500.
1750.
2000.
PPNPITIPm PF
PorsTriM'TinN
53".
1(72.
ii " f
fill.
3PC.
Tni.'Fp-r/1!
.J,S_PL/»;-.S
nil? .
s^c .
si:.°
513.
I; 8 7 .
PHI ATFP
n3r. 1 . ° ' " 1.2375
n:7. 1.H7I: 9.0532
101. 1.270 11.1U?I:
p°r. i.?.i;i 11. me
PAPK. INR PPPTI! PIA.'TTFf
3.3 .1)3
2 . r . i, :,
2.2 .liS
2 . 0 . 5 t
. __.
TOUFP UlAf'F.TFP TPl.'FI1 llhlni'f "~ 10?\''\ ft: I'/^TPS " tnfAl Fl ()l;
I:KKT i-KM i.us/nr-sprT n/'i;(i>f) i iPuii'i
.M! 2.1! 3!'b.," 20IM). J./. i.
PUESSIT.F DI'OH FAN Hl'liiih.I'Ol.'H' Kl'f'P Hun.SKHPrH*
l.bCl ,n?n
PHFIMPAI
PAUSTIP PHI or IMF
ID/YFAP IB/YFAP
3P. 90.
PAPITA.I POSTS FOP S
TOVFR INTFPNAIS PAPKINP FAN PI'NP fnjOPS TANKS T^T/I
551. 2. 33. 51. 50. 51. 15/<.
Al POST FOP 5rpl''Trr SYRTFN
r pr>\-^r nAiNT*"NAt'PF rnpcir/vL TOTAL ?/icoo crn-np
no. ii. 57. 7. ISP. i.rss
G003 Ft'n OF PPOFPAM
-------
Table AA-1 Concinued
CHAMBFP
NOZZLFS/SQ FT LFNGTH UNITS/SFPIES
6.0 8.0 3.0
RPM/NOZZLE
1.200 10.000 15.000 20.000
TOTAL
LENGTH
52.0
25.000
NOZZLE PS I
14.0
POWFR
*/KWH
.025
FLOW RATE
CFH
60000.
WIT
MAINTENANCE
«/SORT(CFM)
8.000
CONCENTRATION
INLET OUTLET
1000. 50.
COSTS FOP
NOZZLES
* FA
5.000
SrPUBBFR SYSTEM
AMORTIZATION
FACTOR
.100
HENPY SCHMIPT NO
0.
0.
TEMPERATURE
OFRPEFS F
70.
INTERMEDIATE
ANNUAL COST CAPITAL
29715.
29715.
32074.
33399.
34556.
35590.
29715.
29715.
108301*.
88192.
87114.
86791.
86784.
108304.
VALUES DURING OPTIMIZATION
TIM-SFC HEIGHT GPM/NOZ
1.4
.4
.3
.3
.3
l.'i
1.2
10.0
15.0
20.0
25.0
1.2
FT2/FT3 NO
4.1 785122
Nl
774675
7.3
8.1
8.P
Q.II
4.1
686I»25
669015
656930
647708
785122
6460I»9
617289
595400
577667
774675
OPERATING CONDITIONS OF TOWFP-rALPUlATFD
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBREP HFIGHT TOTAL FIOV.' RATES
FEET FEET GAS(CFM) LIQUID(GPM)
7.0 8.4 60000.0 424.7
NOZZLES/BANK
354.
BANKS/UN IT
1.
PDROP UM
381.
PRESSURE DROP
4.500
FAN HOPSEPOWFP
77.073
I'NTS/SFPIFS
3.
PUMP HOPSFPOWFP
CHEMICAL CONSUMPTION
CAUSTIC CM LOR INF
LB/YEAR I.B/YFAP
66025. 33241.
SHELL
47501.
AMORTIZATION
10830.
CAPITAL COSTS FOR SCRUBBER SYSTEM
NOZZLES FAN PUMP MOTORS
5308. 891. 1680. 4092.
ANNUAL COST FOP SCRUBBFp SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
10527. 3394. 4963. 20715.
TANKS
7454.
EXTERNALS
41378.
TOTAL
108304.
*/1000CFM-HR
-------
APPENDIX 5
DESIGN AND COST CALCULATIONS
FOR PACKED TOWERS
-------
Table A5-1
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 1 .
NUMBER OF STAGES IN SCPUBBFP IS 1.
TYPE OF PACKINR AND ITS PROPERTIES
1 1/2 INTALOX SAnniFS
POROSITY ARFA FOUIVALENT
SOFT/CUFT DIAMFTFR(FT)
.810 52.0 .Ifi
CONSTANTS FOP DFLTAP
ALPHA BETA
.130 .150
OPERATING CONDITIONS FOR TOWFP.
FLOW RATE
CFM
5000.
CONCENTRATION
INLtT OUTLET
8000. l»0.
HENRY SCHMIDT NO
0.
2.053
TEMPFRATURF
DFRREFS F
70.
UhMT COSTS FOP
POWFR MAINTENANCE
*/KWH «/SQRT(CFM)
.025 8.000
SCRURBFR SYSTEM
PACK I MR AMORTIZATION
«/CII-FT FACTOR
«.200 .130
INTERMEDIATE VALUES DUPINfi OPTIMIZATION
ANNUAL COST
33778.
3106R.
30202.
29980.
2(5860.
LOADINR RATES(LB/HP-SOFT)
LIOIJIP
1000.
2500.
4000.
5500.
7000.
8500.
RAS
18
Ulfc.
1031.
HOP.
811.
CAPITAL-*
33216.
2R3«»7.
2f5R7".
25«)P?.
2572".
25778.
HTMR(FT)
13.RP2
6:261)
"f.HRP
J«.121
3.I»RF
K(R)A
It.5000
5.7R31
7.1303
R.'ri17
PACKINR DFPTH
73.R
kk.P
33.2
26.U
21.P
IS.5
DIAMFTFR
98
50
91
5.27
5.61
5.95
OPFRATINR CONDITIONS OF
THE MATERIAL OF CONSTRUCTI ON
TOWFR DIAMETFP PACK INC DFPTH
FEET FEFT
5.95 18.5
TOWFP-CALCIIIATFP
IS FIRFPRLASS(FPP)
LOADINR RATFS TOTAL FLOW RATES
I.BS/HR-SOFT RAS(CFM) LIOUID(RPM)
810.7 8500. 5000. 1*70.
PRESSURE DROP
3.7
FAN HORSFPOWFP
5.2
PUMP HORSEPOWER
6.8
CHEMICAL CONSUMPTION
CAUSTIC CHtOPIUP
LB/YEAP IB/YFAP
151U33. 110553.
TOWFR
16350.
CAPITAL COSTS
II'TEPWLS
610.
FOP SCPMBP.PP SYSTFM
PACK INC FAN PI'HP Mr>TOf>S
U?0«». 35R. 6P1. 86P.
TANKS
27PF.
TPTAL
2577R.
6121.
6003
ANNUAL COST FOP smppprp SYSTEM
AMORTIZATION POWFR MAINTPNAMCF CHFfMCAl
3351. 895. 5PF. 25033.
END OF PPOPPAM
THTAI
298I|I|.
-------
Table A5-2
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 2
NUMBER OF STAGFS IN SPPUBBFR IS 2.
TYPF OF PACK I MR AND ITS PROPERTIES
1 1/2 INTALOX SADDI.FS .
POROSITY ARF.A FQUIVALENT
SOFT/CUFT DIAMFTFR(FT)
.810 52.0 .IB
CONSTANTS FOR DFI.TAP
ALPHA BFTA
.130 .150
FLOW RATF
CFM
5000.
CONCENTRATION
INLET OUTLET
ROOD. 1*0.
OPERATING CONDITIONS FOP TOWFP
HFNPY SCHMIDT NO
0. 2.P53
TFMPFPATURF
DFOPFFS F
70.
UMIT COSTS FOP SCPIIRRPP SYSTF*'
POWER MAINTENANCE PACKINR AMORTIZATION
S/KWH S/SORTCCFM) -t/PU-FT FACTOR
.025 8.000 s.2on .130
INTERMEDIATE VALUES OUPINR OPTIMIZATION
UI
I
to
ANNUAL COST
25U2I*.
22805.
22U9.
21970.
21996.
LOADING RATFSUB/HP-SOFT)
LIQUID
1000.
2500.
1*000.
5500.
7000.
HAS
1812.
1189.
1031.
909.
CAPITAL-!":
3P226.
33898.
3293P.
32995.
33572.
HTUR(FT)
13.P82
8.U58
6.261*
It.986
U.121
K(0)A
I*.5000
5.7631
6.5U29
7.1303
7.6101
PACKINR DEPTH
36,
22.
16,
13,
10."
DIAMETER
3.18
l».50
4.91
5.27
5.61
OPERATING CONDITIONS OF TOV.'EP.-CALCUI.ATED
THE MATERIAL OF CONSTRUCTION IS FlRFPGLASS(FRP)
TOV/FR DIAMETER PACKINR DFPTH
FEET FEET
5.61 10.9
LOADINR RATES TOTAL FLOW RATES
LBS/HR-SOFT RAS(CFM) LIOUID(RPM)
90H.5 7000. 5000. 31*5.
PRESSURE DROP
TOWER
210U6.
6121.
FAN HORSFPOWFR
6.7
PUMP HORSEPOWER
3.F
CHEMICAL CONSUMPTION
CAUSTIC CHLOPINF
LB/YFAR LB/YFAP
98285. 67550.
CAPITAL COSTS FOR SCRUBBER SYSTEM
INTERNALS PACKINR FAN
1050. l»l»29. 358.
PUMP
1099.
MOTORS
1093.
TANKS TOTAL
l»l»96. 33572.
ANNUAL COST FOP SCRUBBER SYSTFM
AMORTIZATION POWFP MAINTENANCE CHFMICAL
U36U. 101*7. 800. 15785.
6003 END OF PRORRAM
TOTAI S/1000 CFM-HP
21996. 1.100
-------
Table A5-3
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 3
NUMRFP OF STAPFS IN SC^TRR" IS 3.
TYPE OF PACKINP AN" ITS PPPPFPTIPf:
1 1/2 INTALPy SAPPI.FS
PPPPSITY A"PA PPUIVAIFMT
SPFT/CUFT PIAMP~FP(FT)
.PIP 52.P .IP
CONSTANTS FOO pr|
AIPMA HFTA
.13P .150
FLOW PATF
CFM
5000.
CONCENTRATION
ir'LFT nifTLFT
8000. «»0.
PPFPATIPP CONPITIPNS TOP TOV'FP
HFNPY SCHMIPT NO TFMPFPATUPF
0.
7.053
UNIT COSTS rnr
MAIWTFNANrF
P.POO
POWFP
J/KWH
.025
INTFPMFPIATF VAIUFS Dl'nINR
ANNUAL COST I.OAP I NP PATFS( I R/HP-SOFT)
70.
AMpnjl7AT|ON
«/ri'-FT
Ul
1
c>>
23553.
210I»2.
20513.
?. 0 Ii 7 3 .
20R51.
2051*2.
201*69.
201*1*8.
I.I 0(1 in
1000.
2500.
1*000.
5500.
7000.
P250.
5500.
1*750.
PAS
im.
1 li J I» .
IIP".
1031.
nfio m
°5°.
1P?3.
10f»F.
^APITAI -*
UPPF5.
300P?.
MTIlP(rT)
13.°P2
P.l:5P
5.
5.7P31
6.5U?°
7!l3P3
7I3P07
7.1303
R.P53P
111.11
11.1
P-.P
7.7
7."
P.. 7
PIAMFTFP
3. OR
5.27
5.61
5.1*7
5.29
5.11
OPERATING CONDITIONS OF
THF MATFPIAL OF CONSTP.ITTION
TOWFR PIAMFTFR PACKING DFPTH
FFET FFFT
5.1J P.7
TCWP-r-AlCUl ATFD
IS FIPFPPLARSfFPP)
LOAPINP PATFS TOTAL FIOW RATFS
LBS/HP-SOFT fiAS(CFM) LlOUIP(PPr)
10»»F.5 1*750. 5000. 101,.
PPFSSURF PROP
7.1*
CHEMICAl
CAUSTIC
LB/YFAP
821*15.
FAN HOHSFPPV'FP
10.P.
CONSUMPTION
CHLOP INF
IR/YFAP
555RH.
PUMP HOPSFPOWFP
1."
CAPITAL COSTS FOP SOPl'PRPP SYSTEM
TOWFP INTFRNAIS PACKINP FAN PUMP MOTOPS
25756. 1235. l»915. 35R. llfiS. 171*7.
TANKS
lt777.
TOTA1
ANNUAL COST FOP SCPUBPPP SYSTFM
AMORTIZATION POV'FP MAINTENANCF CHFMICAI TOTAL
5120. 1?22. 9PO. 1311P. 20l*i*R.
<;/1000 CFM-HP
1.022
-------
Table A5-4
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 4
NUMBFR OF STAGES IN SCRUBBER IS 3.
TYPF OF PACKING AND ITS PROPERTIES
1 1/2 IN.TALOX SADPI.FS
POROSITY APFA FOUIVALFNT
SnFT/rilFT PIAMFTFP(FT)
.CIO 52.0 .16
CONSTANTS FOR DFI.TAP
ALPHA BFTA
.130
FLOW RATE
CFH
10000.
CONCENTRATION
INLET nilTLFT
8000. 20.
OPERATING CONPITIONS FOR TOHEP
HFNRY SCHMIPT NO
0. 7.P53
TEMPERATURE
DFRPFFS F
70.
WIT COSTS FOP
POWER MAINTENANCE
S/KWH «/SORT(CFM)
.025 8.000
SCRURRFP SYSTEM
PACK I MR AMORTIZATION
VHI-FT FACTOR
8.200 .330
ANNUAL COST
INTEPMEPIATF VALUFS Dl'PIMP OPTIMI7ATIOM
LOAD IMP RATES (I.R/HP-SPFT)
UOB75.
30380.
3011}?.
31357.
30202.
33131.
3H1PJ?.
L101! I P
1000.
2500.
14000.
5500.
7000.
6250.
5500.
U750.
RAS
1812.
HilU.
11RP.
1031.
PAPITAL-*
75052.
67733.
FH702.
05".
1023.
68P30.
P7P03.
6721P.
HTIT(FT)
13.PP2
R.l;58
P.2FI:
lt.°rF
It.121
l».«»P2
U.nl»1
5.F17
K(C)A
U.5000
5.7T31
6.FU2"
7.1303
7.F101
7.3P07
7.13P3
6.P53H
PACK IMP DFPTH
27.7
8.?
DIAMETER
5.63
6.37
6O ti
• J t
7.1»C
7. °.k
7\71
7.U8
7.23
OPERATING CONDITIONS OF TnV.'FP-CAl Cli| ATEP
THE MATERIAL OF CONSTRUCTION IS FIRPPPLASS(FPP)
TOWFP DIAMETER PACKING PFPTH LOAPINP PATFS TPTAI. Finvr RATFS
FEET FEET LRS/HP-SPFT GA?(CFM) LIOUIP(GPM)
7.23 11.0 1P1F.F U750. 10000. 3P8.
PRESSURE PROP
8.U
FAN
PUMP
U.I
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
IB/YFAP IR/YFAP
165B3U.
CAPITAL COSTS FOP SCRURRFP SYSTEM
TOWFP INTERNALS PACKING FAN PUMP MOTORS TANKS TOTAL
U1378. 30til. 1112P. 57fi. 173n. 211U. 72U1. 67218.
ANNUAL TOrT FPP SCPUBBFR SYSTEM
AMORTIZATinv PPV.'FP MAI h'TFMANCF CMFfMTAL TOTAI
873P. 2709. 13R6. 2G335. 391F8.
$/!POP PFM-HP
.979
-------
Table A5-5
DESIGN AND COST CALCULATIONS FOR-PACKED TOWERS
Case 5
NUMBFP OF STARFS IN SCRUBBER IS 1.
TYPF OF PACKING ANP ITS PROPF.PTIFS
1 1/2 INTALOX SAPPLES
POROSITY APFA FOUIVALENT
SOFT/CUFT DIAMFTFR(FT)
.810 52.0 .16
CONSTANTS FOR DFLTAP
ALPHA BFTA
.130 .150
OPERATING CONDITIONS FOP TOV.'FR
FLOW RATF
CFM
25000.
CONCENTRATION
INLET OUTLET
500. 10.
HENRY SCHMIDT NO
0.
2.053
TEMPERATURE
OFGREFS F
70.
UNIT COSTS FOP SCRUBBER SYSTEM
POWFR MAINTENANCE PACKING AMORTIZATION
$/KWH $/SOPT(CFM) $/CU-FT FACTOR
.025 8.000 8.200 .130
INTERMEDIATE VALUFS DUPING OPTIMIZATION
Oi
i
Ln
ANNUAL COST
35699.
2671*0.
2U510.
23809.
23737.
21*010.
23836.
2372U.
23697.
LOADING PATFS(LR/HP-SOFT>
LIOUIP
1000.
2500.
1*000.
5500.
7000.
8500.
7750.
7000.
6250.
GAS
1812.
Hill*.
118".
1031.
909.
811.
850.
901.
950.
CAPITAL-*
8P019.
75873.
71*322.
7I»832.
763°2.
78653.
77701.
76675.
75710.
HTUR(FT)
13.882
8.1*58
6.261*
I*. 121
3.1*86
3.71*7
I*. 085
I*. 1*80
PACKING PFPTH
U.5000
5.7631
6.5U29
7.1303
7.6101
8.01"7
7.8221
7.61P1
7.3807
Sit.
3?.
21*.
16
13
II*
IF.n
17.5
DIAMFTER
8.89
10.07
10.98
11.79
12.55
13.30
12.99
12.61
12.23
OPERATING CONDITIONS OF TOWFP-CALCUIATEP
THF MATERIAL OF CONSTRUCTION IS FlBFRGIASS(FPP)
TOWER PIAMFTER PACKING PFPTH LOADING RATFS TOTAL FlOW RATFS
FF.FT FFET LBS/HP-SOFT GAS(CFM) LIOUIP(nPM)
12.23 17.5 958.9 6250. 25000. 1U60.
PRESSURE DROP
3.9
TOWER
I*I»915.
FAN HOPSFPOWFP
28.0
PUMP HOPSFPOWFR
20.3
CHFMICAL CONSUMPTION
CAUSTIC CHI op INF
LB/YFAP LB/YFAP
61670. 31*027.
CAPITAL COSTS FOP SCPUBBFP SYSTF"
INTERNALS PACKING FAN PUMP MOTORS
3971. 16862. 1115. 1297. 2206.
ANNUAL COST FOP SCPUBRrp SYSTFM
AMORTIZATION POWFR MAINTENANCE CHFMirAI
981*2. 3605. 1265. 89P5.
TPTAI
TANKS
531* li.
TOTAI <;/lPOP CFf'-HP
2369'. . *>37
-------
Table A5-6
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 6
MUMBFR OF STARFS IN SCPl'BBFP. IS 2.
TYPF OF PACK I NO AMP ITS
1 1/2 INTALOX SAPDIFS
POROSITY APFA
SQFT/CIIFT
.P10 52,0
PPOPFPTIFS
EOUIVAIFNT
DIAMFTFP(FT)
.1R
CONSTANTS FOP DFLTAP
ALPHA BFTA
.130 .150
OPFRATINR CONDITIONS FOP TOl'-FP
FLOW RATE
TFM
25000.
CONCENTRATION
INLET OUTLET
500. 10.
HFNPY SPHMIP-T NO
0.
2.053
TFMPFPATUPF
OFRPFFS F
70.
UNIT COSTS FOP
POWER MAINTENANCE
*/KWH */SOPT(CFM)
.025 P.000
SCPURRFP
PACK 1 NR
8.200
SYSTEM
AMORTIZATION
T FACTOP
.130
i
o>
INTERMEDIATE V
ANNUAL COST
3«*R09.
2601*9.
21*338.
2U206.
21*756.
21*1*22.
21*196.
21*133.
21*302.
21*196.
2 1* 1 1* 1* .
21*11*8.
LOADIMR
LIOUIP
1000.
2500.
1*000.
5500.
7000.
6250.
5500.
1*750.
1*000.
U375.
1*750.
5125.
VALIIFS PUP I MR OPTIMIZATION
RATES(LB/HP-SOFT)
OAS
1812,
Ull».
1189.
1031.
909.
959.
1023.
CAPITA! -«
11P1.
111*1.
1100.
1062.
91901.
93385.
97028.
101865.
99R61*.
97330.
95278.
93597.
91*276.
95155.
96128.
HTCP(FT)
13.P82
8.U5R
I».9P6
5000
7631
5U29
1303
7.F101
5.517
6.225
5.86P
5.535
5.236
7.3807
7.1303
6.8536
6.5U29
6.7031
6.R536
P.9957
PACK I MO DFPTH
27.3
16.6
12.3
s!i
8.R
9.7
10.P
12.2
11.5
10. 9
10.3
OPERATING CONDITIONS OF TOVIFP-CALCUIATFP
THF MATERIAL OF CONSTRUCTION IS Ft RFPRI.ASS(FRP)
TOWER DIAMETER PACKIN.R DFPTH LOAPINR PATFS TOTAl FI.OV RATFS
FFFT FFFT LBS/HP-SOFT RAS(CFM) I.IOIMPC'PM)
11.62 10.3 1062.3 5125. 25000. 1PPO.
PRESSURF DROP
5.1
FAN HORSFPOWFP
3P.3
PUMP HOPSFPOWFP
11.1
TOWER
56360.
CHFMICAL CONSUMPTION
CAUSTIC CHI OP INF
LB/YEAP LB/YFAP
37836. 20800.
CAPITAL COSTS FOP SCPUP.RFP SYSTEM
INTERNALS PACKINR FAN
6951. 178R2. 1115.
PUMP
2132.
MOTORS
2761*.
TANKS
8921*.
TOTAL
9F128.
ANNUAL COST FOP SCPUBBFP SYSTEM
AMORTIZATION POV.'FR MAINTENANCE CHF^ITAL TOTAL
12U97. U359. 1789. 550I». 2U1U8.
!»/1000 CFM-HR
DIAMETER
8.P9
10.07
10.98
11.79
12.55
12.22
11.83
11.1*3
11.02
11.21
11.U1
11.62
-------
DESIGN AITO COST CALCULATIONS FOR PACKED TOWERS
Case 7
NUMBER OF STARFS IN SCPURBFP IS 3.
TYPE OF PACKINR AMP ITS
1 1/2 INTALOX SAOPIFS
POROSITY APFA
v SOFT/CUFT
.810 52.P
PROPERTIES
FOUIVALENT
PIAMFTFP(FT)
.16
CONSTANTS FOP PH TAP
ALPHA BPTA
.130 .150
OPERATING rONPITIONS FOR TPHFP
FLOW RATF
CFM
25000.
CONCENTRATION
INLFT OUTLET
500. 10.
HFNRY SPHMIPT NO
0.
2.053
TFMPFPATURF
DFCPFES F
70.
UNIT COSTS FOP
POWER MAINTENANCE
S/KWH */SORT(CFM)
.025 8.000
SCRUBBER SYSTEM
PACK I NO AMORTIZATION
*/CU-FT FACTOR
8.200 .130
INTERMEDIATE VALUES OUR I MR OPTIMIZATION
ANNUAL COST
35627.
27558.
261*01.
26860.
26501*.
26379.
26622.
26l»56.
26389.
26*11.
LOADINR RATFSdB/HP-SOFT)
LIOUIP
1000.
2500.
I»000.
5500.
l»750.
1*000.
3250.
3625.
ItOOO.
l»375.
RAS
1812.
CAPITAL-*
118".
1031.
1007.
11P1.
1?82.
1733.
11P5.
1P7289.
1110P1.
11R8R8.
115595.
1122B7.
llOPPfi.
1121I»3.
113731.
HTUR(TT)
13.8*2
8.1*58
6.2PU
5.518
6.225
5.P68
UP
K71
K(R)A
l*. 5000
5.7P31
6.5I»2<»
7.1303
6.853G
PACKING DFPTH
18
11
8
P. 1862
6.3713
6.5U70
P. 7031
7.2
8.7
9.l»
8.7
?.2
7.7
DIAMETER
8.89
10.07
10.98
11.79
11.43
11.02
10.57
10.78
11.00
11.21
OPEPATINR CONPITIONS OF TOWFP-CALCUIATFP
THE MATERIAL OF CONSTRUCTION IS Fl BPPRI.ASS(FPP)
TOWER DIAMETER PACKINR DFPTH I.OAPW RATFS TOTAL FIOH RATFS
FFFT FEET I.BS/HR-SOFT RAS(TFM) UOUIP(RPM)
11.21 7.7 lll»0.7 1*375. 75000. P5<1.
PRESSURE DROP
6.1
FAN HOPSFPni-fFp
1*3.6
PUMP HOPSFROV.'FP
7.7
TOWER
66852.
CHEMICAL CONSUMPTION
CAUSTIC CHI OP INF
LB/YEAP LB/YFAP
301*18. 17110.
CAPITAL COSTS FOR SCPUBBFP SYSTEM
INTERNALS PACKINR FAN PUMP
9505. 1861*2. 1115. 2771.
MOTORS
3181.
TANKS
11661*.
TOTAL
113731.
6121.
ANNUAL COST FOP SCPUBBFP SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICAL
11*785. »*967. 2101. 1*1*68.
6003 END OF PPORPAM
TOTAL */1000 CFM-HR
261(11. .261*
-------
Table A5-8
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 8
h'UMBFP OF STAOFR IN SPfH'RPFP IS 1.
TYPF OF PAPKINP AMP ITS PPOPFPTIFP
i 1/2 IWTAI.OX SAnniFS
POPOSITY APFA FOUIVAIFHT
SOrT/PI'FT PIAMFTFP(FT)
.P.JO 52." .IF
PPNSTANTS FOP DFITAP
ALPHA BFTA
.130 .150
OPFPATIMO POMPITIOMS FOP
FLOW P-ATF P.ONPFNTPATIPN HFMPY SPHMIPT MO TFMPFPATI'PF
PFH INLET Ol'TLFT PFP-PFFS \
25000. 2000. 2P. 0. 2. 053 7C.
?
00
PPWFP
S/KWH
.025
UNIT CORTS FOP SPPI'RPF
MAINTFMANPF PAP" IN
«/SPPT(pFM) «/PI'
8.000 ?,. ?no
INTEPHFPIATF VALL'FS Dl'PINrt OPT If
ANNUAL POST I
.oAPirn PATFS(I n/HP-r-opT)
P SYSTFt
n M
-FT
I7ATION
LIOIJIP PAS PAPITAI-<
Fi2''lF.
52332.
l*1B2i».
U8701.
I»850F.
U8705.
l»8557.
l»8l»86.
IPOO. iri2.
2500. IMli.
UOOO. 11P1.
55"0. 1031.
7000. 00^.
8500. 811.
7750. R50.
7000. 001.
9F5"P.
RU505 .
P2723.
P7301 .
P35P5.
R5R07.
8I|7P3.
P3R03.
•
"OPTI7ATIOM
FAPTOP
.130
^ MTI'P(FT)
13. °P?
p . l| 5 j»
F..2FI-
t) 0 P P
U T 2 1
3 . '' P- F
3.71)7
l». ORF
K(P-)A
I). 'lOOO
5.7K31
P. fiU21
7.?303
7.F101
P. n3 17
7. "221
7.F101
PAPKirn DFPTH
63." x
30. r
2R. F
23.0
1°. r
1P..1
17.3
1P.P
DIAMFTFP
8. P°
10.07
in. op.
11. 7P
12.55
13.30
12.99
12.61
OPEPATIM.fi COMPITIONS OF TOHFP-PALPUIATFP
THF MATFPIAL OF POMSTPITTIOM IS Fl BFPRI.ASS(FPP)
TOV.'EP niAMFTFP PAPK I N^ DFPTH LOAP I NP PATFS TOTAl FIOH
FFFT FFFT LBS/HP-SOFT PAS(PFM) LI PUIP. (Q.PM)
12.61 IP..? 001.F 7000. 25000. 173°.
PRESSURE DPOP
FAN HOPSFPOV.'FP
28.fi
Pl'f'P HOPSFPOV'FP
25.3
CHFMIPAL POUSt'MPTIOM
PAUSTIP PHIPPINF
I.B/YFAR LB/YFAP
2021I»6. 137lt°7.
TOWEP
l»P383.
PAPITAL PO?TS FOP
PACKINfi
1"251.
SYSTFM
FAM Pl'MP
1115. U35.
MOTOPS
237".
ANNUAL POST Fon SPPl'PPFP
AMORTIZATION POV'FP M/IKT^NANPF PHFMIPAI
U01". 1?PF. 323^7.
TANKS
5°37.
TOTAL
P3P03.
-------
DESIGN AUD COST CALCULATIONS FOR PACKED TOWERS
Case 9
NUMBFP OF STAHFS IN SPPURRFP IS 2.
TYPF OF PACK IMP AMP ITS
1 I/? IMTALOy SAOfMFS
POPOSITY A"FA
SOFT/PI'FT
.PIO s?.o
PPPPFPTIFS
FOPIVALFNT
HAMFTFP(FT)
.IP
CONSTANTS
AIPHA
.130
FOP nH.TAP
RFTA
.150
OPFPATIMP coNoiTioMs FOP TOVFP
FLOW PATF CONCENTRATION HFNPY
CFM INLET OUTLFT
25000. 2000. 20. 0.
NO
2.053
TFMPFPATIJHF
P.FRPFFS F
70.
POV.'FP
*/KWH
.025
WIT COPTS FOP
MAINTENANCE
P. 000
STPI'RBFP SYSTFM
PACK IMP At'OPTI/^TIOM
*/(M!-rT FACTOR
P.200 .130
It'TFRMFOIATF VAIUFS
OPTU'1 7ATIOF'
Ul
i
VO
ANNUAL COST
5318U.
U3030.
UORUfi.
UOU02.
I.OADIMR PATFSCLR/HP-SOFT)
I(0fi31.
UOI»7I».
Ii0503.
I I Oil ID
1000.
7500.
1*000.
5500.
7000.
6?50.
5500.
l»750.
RAS
1P12.
11)11).
IIP".
1031.
nno.
050.
1023.
loop.
CAPITAI •
110337.
1007F?.
10U72.
HTI'R(FT)
13.PP7
107101.
101^52.
103110.
U.121
U.ltP2
Ij.01,0
5.F17
P DFPTH
i*.5000
5.7F31
P.51*2"
7.3303
7.F10]
7.3P07
7.1303
R.P53F
32.0
11.F
9..r
10.?
i?',7
niAMFTFP
8.P9
IP.07
10.9P
11.70
12.55
12.22
11.P3
11.1*3
or TOVF^-CALCIMATFP
THF MATFPIAL OF COf'STPUCTIOM IS FI PFpni ASS(FPP)
TOWFP niAMFTFP PACK I NO PFPTH lOAPIMR PATFS TOTAI F|0V RATFS
FFFT FFFT IBS/HP-SOFT RAS(CFM)
ll.l»3 12.7 10HR.F. «(750. 2500P. P70.
PPFSSURF DROP
6.U
TOWER
FAN HOPSFPOWFP
I»R.O
PUMP HOPSFPOWFR
11.1
CHFMICAL CONSUMPTION
CAUSTIC CHLORIMF
LB/YEAR LB/YFAP
12911(3. 81*011*.
CAPITAL COSTS FOR SCRIIBBFP SYSTEM
INTERNALS PACKINP FAN PUMP MOTOPS
6fi71. 2137P. 1115. 2002. 3010.
ANNUAL COST FOP SCPl'RBFP SYSTFM
AMORTIZATION POWFP MAINTFMANCF CHF^'ICAL
13«*OI». 5091. 17H9. 20219.
TANKS TOTAL
S365. 103110.
TOTAI */1000 CFM-HP
l»0503. .1*05
-------
Table A5-10
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 10
NUMBFR OF STARFS IN SCPUBRFR |s 3.
TYPF OF PACK I MR AMP ITS PROPERTIES
i 1/2 IMTALOX SAPPIFS
POPOSITY APFA FOUIVALFNT
SOFT/CUFT niAMFTFP(FT)
.810 52.0 .IP
CONSTANTS
AlPHA
.130
FOP PFI.TAP
BFTA
.150
FLOW RATE
CFM
25000.
CONCENTRATION
INLET OUTLFT
2000. 20.
OPFRATINR CONDITIONS FOP TOWFP,
HENPY SCHMIDT MO
0. 2.P53
TFMPFPATUPF
PFRREFS F
70.
UMIT COSTS FOP SCPUBBFR SYSTEM
POWER MAINTENANCE PACK I MR AMORTIZATION
*/KWH */SQPT(CFM) */HI-FT FACTOR
.025 8.000 8.200 .130
>
(Jl
INTERMEDIATE VALUES DUPING OPTIMIZATION
ANNUAL COST LOAD I MR RATES(LB/HR-SOFT)
LIOUIP PAS CAPITAI-*
5185U. 1000. 1812. 1231*78.
U2H9. 2500. 1U1U. 11PPRF.
l»050«». ItOOO. 11RO. 120513.
U0733. 5500. 1031. 12fi«ni.
l»OU6lt. U750. 10H7. 123RI»5.
U0l»71. l»000. 11P1. 120822.
HTUR(FT)
13.R82
8.1(58
6.26U
5.51P
R.225
K(R)A
I).5000
5.7631
6.5U2"
7.1303
6.853P
6.5U2Q
PACKING DFPTH
21.3
13.0
n.6
7.7
8.5
9.6
DIAMETER
8.89
10.07
10.98
11.79
11.1*3
11.02
OPFRATINR CONDITIONS OF T^IPP-CALCUIATFP
THE MATERIAL OF CONSTRUCTION IS FtBFPRLASS(FPP)
TOWER DIAMETFR PACK INR DFPTH I.OAPINR PATFS TOTAI FIOW PATFS
FEET FEET I.BS/HP-SOFT RAS(CFM) IIOIHP(PPM)
11.02 °.6 1181.2 l»000. 2FOOO. 75P.
PRESSURE PROP
7.H
TOWER
713U1.
FAN HOPSFPOWFP
5R.1
PUMP HOPSFPOV/FP
7.5
CHEMICAL CONSUMPTION
CAUSTIC CHLOPINF
IB/YEAR LB/YFAP
107122. 60118.
CAPITAL COSTS
INTERNALS
9081*.
FOR SCRUBBER SYSTFM
PACKINR FAN PUMP MOTORS
221*22. 1115. 2575. 3U61.
TANKS
1082U.
TOTAL
120822.
ANNUAL COST FOP SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICAL TOTAL
15707. 5865. 2191. 16708. UOU71.
*/1000 CFM-HR
.«»05
6121.
6003 END OF PPDPRAM
-------
Table A5-11
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 11
NUMBFP OF STAfiFS IN SCPURRFP IS 1.
TYPF OF PACKINfi AMP ITS PPOPFPTIFS
1 1/2 INTALOX SAPPIFS
POROSITY APFA FOIHVAI.FNT
SOFT/OUFT PIAMFTFH(FT)
.RIO 57.0 .]F
CONSTANTS FOP DH.TAP
AIPHA RFTA
.330 .150
FIOW RATF
CFM
50000.
CONCENTRATION
INLET OUTI.FT
500. 10.
OPERATING CONPITIONS FOP TOHFP
HFNPY SCHMPT NO TFf'PFPATl'RP
0. 2.053 70.
UNIT COSTS FOP
POWFR MA I K'TFMAMrF
J/KWH */SOPT(CFM)
.025 R.OOO
PACK IMP
*/rtt-FT
8.20P
AMOPTIZATIOM
FACTOR
.130
INTFRMF.DIATE VAI.UFS Dl'PING OPTIMIZATION
Ul
i
ANNUAL COST
69317.
51890.
l»5765.
U5930.
45739.
LOAD INP RATES(IB/HP-SOFT)
LIOUIP
1000.
2500.
1*000.
5500.
7000.
8500.
7750.
7000.
6250.
OAS
1812.
mil*.
11R9.
1031.
909.
811.
850.
901.
959.
CAPITAL-*
161720.
11*5339.
11*0371.
11*7712.
11*570".
HTUO(FT)
13.882
R.l*58
6.261*
I*.121
3.1(86
3.71*7
I*.085
l».5000
5.7631
6.51*29
7.1303
7.6101
8.P107
7.8221
7.F101
7.3R07
PACKING
51*. 3
33.1
21*. 5
jo. 5
16.1
13.6
11*. 7
16. P
17.5
DFPTH
DIAMFTER
12.58
1I».2U
15.53
16.67
17.75
18.80
18.36
17.83
17.29
OPERATIC COMDITIOMS OF TOWFP-rAl.rW ATFO
THF MATERIAL OF CONSTRUCTION IS FIRFPRLASS(FPP)
TOWFR DIAMETER PACK I NO DFPTH LOAHINn PATFS TOTAI FIOW RAT^S
FFFT FFFT LBS/HP-SOFT OAS(CFM) IIOUIP(C-PH)
17.2') 17.5 "5P.Q fi?50. 50000. 2"19.
PRESSURE DPOP
3.9
TOWER
85280.
FAN HOPSFPOWPP
56.0
PUMP HOPSFPOWFP
IiO.P
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
12331*0. 68051*.
CAPITAL COSTS FOR SCRUBBER SYSTEM
INTERNALS PACKINR FAN
9777. 33721*. 1866.
PUMP
191*0.
MOTORS
3510.
TANKS TOTAL
8101. 1UU197.
ANNUAL COST FOP SCPUBRFR SYSTEM
AMORTJZATION POWER MAINTENANCE CHEMICAL
1871*6. 7210. 1789. 17970.
TOTAI Jt/1000 CFM-HR
1*5711*. .229
-------
Table A5-12
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 12
NUMBFR OF STARFS IN SCRl'BBFR IS 2.
TYPF OF PACKING ANP ITS PROPFPTIFS
1 1/2 INTALOX SAPPLFS
POROSITY ARFA EOIMVAI.FNT
SOFT/CIIFT OIAMFTFR(FT)
.«10 52.0 .If,
CONSTANTS FOP DFLTAP
AlPHA BFTA
.130 .150
FlOW RATF
CFM
50000.
CONCENTRATION
INLFT OUTLFT
500. 10.
OPERATING CONDITIONS FOP THHFP
HFNPY SCHMIPT NO
0. 2.053
TFf'PFPATUPF
OFRPFFS F
70.
UNIT POSTS FOR SCPUBBFP SYSTFM
POWFR MAINTENANCE PACKINR AMORTIZATION
*/KWH VSORT(CFM) t/ni-rr FACTOR
.025 P.000 8.200 .130
INTERMEDIATE VALUES IH'P.INR OPTIMIZATION
V/l
i
H1
10
ANNUAL COST
66621.
50008.
LOADINR PATES(LB/HP-SOFT)
'(7088.
l»60«5.
l»6030.
LIOUIP
1000.
2500.
l»000.
5500.
7000.
6250.
5500.
If750.
PAS
1P.12.
CAPITAL-*
1189.
1031.
909.
959.
1023.
17577S.
178057.
HTUR(FT)
13.R82
8.U58
6.2PU
J».9PF
1P3135.
1C9231.
1P5052.
K(R)A
U.5000
5.7631
6.5U29
7.1303
7.6101
7.3«07
7.1303
PACKING DEPTH
5.517
6.R536
27.3
16.F
12.3
9.P
8.1
8.P
9.7
10.P
PIAMFTFR
12.58
U.2I»
15.53
16.67
17.75
17.29
16.71*
16.17
OPERATING CONDITIONS OF TOWFP-CALCIM ATFP
THE MATERIAL OF CONSTRUCT I ON IS FIBFPRLASS(FPP)
TOWER DIAMETER PACKING PFPTH LOADING RATFS TOTAL Fl 0V.' RATES
FFET FFFT I.BS/HP-SOFT RAS(TFM) LIOUIP(RPM)
16.17 10. P 100R.5 l»750. 50000. 19UO.
PRESSURE DROP
5.5
TOV/FR
106511.
FAN HORSFPOK'FP
78.5
PUMP HOPSFPOWFP
CHEMICAL CONSUMPTION
CAUSTIC CHI OP INF
LB/YFAR IB/YEAP
75672. M600.
CAPITAL COSTS FOP SCPURBFP SYSTEM
INTERNALS PACKINO FAN PUMP
16U2F. 3P505. 1PRF. 300R.
ANNUAL COST FOP SCPUBBFP SYSTEM
AMORTIZATION POWFP MAINTENANCE CHEMICAL
23583. P908. 2530. 11008.
TANKS
126PO.
TOTAL
IKHll.
TOTAL */iooo CFM-HP
l»6030. .230
6121. 6003 END OF PPOGPAM
-------
Table A5-13
DESIGW AUD COST CALCULATIONS FOR PACKED TOWERS
Case 13
NUMBER OF STARFS IN SCPI'BRFP I? 2.
TYPF OF PACKINR AMD ITS PROPERTIES
1 1/2 U!TAI,OX SAPPLFS
POROSITY ARFA FOUIVALFNT
SOFT/CUFT DIAMFTF-P(FT)
.RIO 52.0 .16
CONSTANTS FOP DFLTAP
ALPHA BFTA
.130 .150
OPERATING CONOITIOMS FOP
FLOW RATE
CFM
50000.
CONCENTRATION
INLET OUTLFT
2000. 20.
HFNP.Y SCH'MPT NO
0.
2.053
TFMPFPATIJPF
PFRPFFS F
70.
UNIT COSTS FOP
POWER MA I NTFNANCE
$/KWH .«/SOPT(CFM)
.025 8.000
PACKU'P
SYSTFf
AMOPTI7AT|Ofi
T FAPTOP
.130
INTERMEDIATE VALHFS Dl'PIMR OpTUMZATIOM
ANNUAL COST
103625.
83888.
LOAPINR PATES(IB/HP-SOFT)
78B18.
7<)3«»5.
7*832.
78580.
78B8R.
78609.
LIO.UIP
1000.
2500.
UOOO.
5500.
7000.
6250.
5500.
1»750.
5125.
PAS
1812.
118°.
1031.
909.
95°.
1023.
109F.
1062.
CAPITAL-.*
207633.
1«>287F.
1P3P61.
MTUR(FT)
13.882
B.2BU
2071°6.
2035PU.
13P757.
107790.
5.517
5.23P
K(R)A
U.5000
5.7631
1303
6101
3P07
1303
853P
6.9957
PACK I MR DPPTH
32.0
10.5
ll'.5
9.5
10.3
11.k
12.7
12.3
DIAMFTFP
12.58
Ik.2k
15.53
16.67
17.75
17.29
16.7k
16.17
OPERATIMR COfiDITIONS OF TOV.'EP-CALCULATFO
THE MATERIAL OF PONSTRUCTION IS FIRFPRLASS(FRP)
TOWEP DIAMFTER PAPKINR DFPTH LOAPINR RATES TOTAL FLOW PATES
FFET FFFT LBS/HP-SOFT RAS(CFM) LIOUIP(CPM)
16.l»3 l?.l 1062.2 5125. 50000. 21P1.
PRESSUPF DROP
5.9
CHEMICAL
CAUSTIC
LB/YFAP
258287.
FAN HOPSFPOV'FP
TONFP
PUMP HOPSFPPWFP
2I».J
CHLOPINE
LB/YFAP
1PP028.
CAPITAL COSTS FOR SCRUBBER SYSTFM
INTERNALS PACK I NO FAN PUMP MOTORS
17118. I»1PR5. 1RFF. 320F. l»775.
TANKS
ANNUAL
AMORTIZATION
25713.
TOST FOP
POHFP
9929.
SCPUPRFP
253P.
SVRTFM
CHPMICAI
I|OI»*7.
TOTAI
7RFP9.
«/1000 CFM-HP
-------
Table A5-14
DESIGN AND COST CALCULATIONS FOR PACKED TOWERS
Case 14
NUMBFP OF STARFS IN SCPUBRFP IS 3.
TYPF OF PACK I MR AND ITS PROPERTIES
1 1/2 INTALOX SAPPI FS
POPOSITY APpA F01MVAIFHT
S"FT/n
-------
Table A5-15
DESIGN AHD COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 1
CHAMBER ««».* TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH
6.0 8.0 1.0 20.0
GPM/NOZZLE
1.000 2.000 3.000 ft.000 5.000
NOZZLE PS I
15.0
POWER
$/KWH
.025
FLOW RATE
CFM
25000.
UNIT COSTS FOR SCRUBBER SYSTEM
MAINTENANCE NOZZLES AMORTIZATION
$/SQRT(CFM) $ EA
8.000 5.000
CONCENTRATION HENRY
INLET OUTLET
500. 10. 0.
FACTOR
.130
SCHMIDT NO
0.
TEMPERATURE
DEGREES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
UAL COST
18351.
18351.
17860.
17860.
17722.
17722.
17690.
17690.
17702.
17690.
17690.
CAPITAL TIM-SEC
49461.
43471.
40775.
39177.
38101.
39177.
1.3
.9
.7
.6
.5
.6
HEIGHT
8.
6.
6.
5.
5.
5.
2
8
0
6
2
6
GPM/NOZ
1.
2.
3.
k.
5.
k.
0
0
0
0
0
0
FT2/FT3 NO
9
11
13
Ik
15
Ik
.6
.7
.1
.1
.0
.1
2051449
1963302
1913506
1878943
1852561*
1878943
Nl
1912313
1752764
1648077
1567999
1502437
1567999
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
5.6 5.6 25000.0 7U5.6
NOZZLES/BANK
186.
BANKS/UNIT
1.
DDROP UM
461.
PRESSURE DROP
1.500
FAN HORSEPOWER
10.705
UNITS/SERIES
1.
PUMP HORSEPOWER
20.756
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
61670. 3U027.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS
17711. 932. 519. 883. 16UO. 3929,
AMORTIZATION
5693.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
23U7. 1265. 8985. 17690.
EXTERNALS
13563.
S/1000CFM-HR
.177
TOTAL
39177,
A5-15
-------
Table A5-16
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 2
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH NOZZLE PS I
6.0 16.0 1.0 28.0 15.0
GPM/NOZZLE
1.000 2.000 3.000 I*.000 5.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3 NO Nl
181*59. 5175U. 2.0 7.3 1.0 7.6 1628217 11*911*96
181*59.
17820. 1*5091*. 1.1* 6.0 2.0 9.3 1558256 1353U33
17820.
17591. 1*201*7. 1.1 5.l» 3.0 10. k 1518733 12625U9
17591.
17U92. U02U. .9 5.0 U.O 11.2 11*91301 1193189
171*92.
171*1*9. 38961*. .8 i».7 5.0 11.9 1U7036I* 11366U1
171*1*9.
171*1*9. 38961*. .8 I*.7 5.0 11.9 11*70361* 11366U1
171*1*9.
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
I*.7 I*.7 25000.0 653.8
NOZZLES/BANK BANKS/UNIT DDROP UM UNITS/SERIES
131. 1. 1*78. 1.
PRESSURE DROP FAN HORSEPOWER PUMP HORSEPOWER
1.500 10.705 17.902
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
61670. 31*027.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
18232. 651*. 519. 820. 151*6. 3631. 13563. 38961*.
ANNUAL COST FOR SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICALS TOTAL $/1000CFM-HR
5065. 2131*. 1265. 8985. 171*1*9. .171*
A5-16
-------
Table A5-17
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 3
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH NOZZLE PS I
6.0 8.0 2.0 36.0 15.0
GPM/NOZZLE
1.000 2.000 3.000 l*. 000 5.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3 NO Nl
17667. 59273. 1.6 6.5 1.0 6.1 129671*8 1260111*
17667.
17175. 52721. 1.1 5.1* 2.0 7.1* 121*1029 1181*193
17175.
17067. U98U5. .9 I*.8 3.0 8.3 1209552 11361*21
17067. ?f)
17070. U8178. .8 <*.!» l*.0 8.9 1187701* 110050U
17118. 1*7083. .7 I*.2 5.0 9.5 1171030 1071278
17067. »*98l*5. .9 1*.8 3.0 8.3 1209552 1136U21
17067.
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
k.8 It.8 25000.0 1*11*.5
NOZZLES/BANK BANKS/UNIT DDROP UM UNITS/SERIES
138. 1. 1*39. 2.
PRESSURE DROP FAN HORSEPOWER PUMP HORSEPOWER
3.000 21.1*09 11.376
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
37836. 20800.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
2211*5. 1382. 519. 1222. 2356. 5525. 16697. 1*981*5,
i_ ANNUAL COST FOR SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICALS TOTAL $/1000CFM-HR
61*80. 3291*. 1789. 5501*. 17067. .171
A5-17
-------
Table A5_i8
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 4
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH NOZZLE PS 1
6.0 16.0 2.0 52.0 15.0
GPM/NOZZLE
1.000 2.000 3.000 U.OOO 5.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST
177UO.
1771*0.
17070.
17070.
16853.
16853.
16773.
16773.
16751*.
16751*.
16751*.
16751*.
CAPITAL TIM-SEC
61616.
51*283.
50989.
1*901*1.
1*7733.
U7733.
2.6
1.8
1.4
1.2
1.1
1.1
HEIGHT
5.8
l*.8
l».3
3.9
3.7
3.7
GPM/NOZ
1.0
2.0
3.0
l*.0
5.0
5.0
FT2/FT3 NO
U. 8
5.9
6.6
7.1
7.5
7.5
1029217
98U99U
960011
9U2671
9291*36
929U36
Nl
992852
928829
888011*
857106
83181*8
8318U8
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
3.7 3.7 25000.0 1*13.3
NOZZLES/BANK
83.
BANKS/UNIT
1.
DDROP UM
1*78.
PRESSURE DROP
3.000
FAN HORSEPOWER
21.1*09
UNITS/SERIES
2.
PUMP HORSEPOWER
11.116
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
37836. 20800.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
20621. 827. 519. 1220. 2335. 5515. 16697. 1*7733.
AMORTIZATION
6205.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
3256. 1789. 5501*. 16754.
A5-18
I/1000CFM-HR
-------
Table A5-19
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 5
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH NOZZLE PS I
6.0 16.0 -?2.0 52.0 25.0
GPM/NOZZLE
2.000 3.000 it.000 5.000 6.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 . .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3 NO Nl
17222. 52753. 1.6 4.6 2.0 6.1 1300300 1228882
17222.
17102. 49663. 1.3 4.1 3.0 6.? 1267320 1175707
17102.
17098. 47841. 1.1 3.8 4.0 7.4 1244429 1135484
17098.
17H1. 1*6623. 1.0 3.6 5.0 7.9 1226958 1102635
17207. 45747. .9 3.4 6.0 8.3 1212866 1074622
17098. 47841. 1.1 3.8 4.0 7.4 1244429 1135484
17098.
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
3.8 3.8 25000.0 345.5
NOZZLES/BANK BANKS/UNIT DDROP UM UNITS/SERIES
86. 1. 409. 2.
PRESSURE DROP FAN HORSEPOWER PUMP HORSEPOWER
3.000 21.409 13.326
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
37836. 20800.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
21203. , 864. 519. 1101. 2505. 4953. 16697. 47841,
* ANNUAL COST FOR SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICALS TOTAL $/1000CFM-HR
6219. 3585. 1789. 5504. 17098. .171
\
A5-19
-------
Table A5-20
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 6
NOZZLES/SQ FT
6.0
GPM/NOZZLE
1.000 2.000
CHAMBER
LENGTH
8.0
UNITS/SERIES
3.0
3.000
l*.000
TOTAL
LENGTH
52.0
5.000
NOZZLE PS I
15.0
POWER
$/KWH
.025
FLOW RATE
CFM
25000.
UNIT COSTS FOR SCRUBBFP SYSTEM
MAINTENANCE NOZZLES AMORTIZATION
$/SQRT(CFM) $ EA
8.000 5.000
CONCENTRATION HENRY
INLET OUTLET
500. 10. 0.
FACTOR
.130
SCHMIDT NO
0.
TEMPERATURE
DFGREES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3
NO Nl
19085.
19085.
18551*.
1855U.
18U51.
181*51.
181*68.
1853U.
181*51.
181*51.
67871.
6061*1*.
5751U.
55721*.
5U561*.
575H*.
1.9
1.3
1.0
.9
.8
1.0
5.7
U.7
l». 2
3.9
3.6
I*. 2
1.0
2.0
3.0
I*.0
5.0
3.0
I*.P
5.6
6.3
6.S
7.3
6.3
9P8881
91*6391
922387
905726
893011
922387
9721*1*1
920693
889111*
865830
81*7138
889111*
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
l*.2 I*.2 25000.0 316.1
NOZZLES/BANK
105.
BANKS/UNIT
1.
DDROP UM
1*39.
PRESSURE DROP
I*.500
FAN HORSEPOWFR
32.111*
UNITS/SFRIES
3.
PUMP HORSFPOWFR
8.578
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
30«*18. 17110.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
21*971*. 1580. 519. 151*1*. 2996. 701*3. 18857. 575H*.
AMORTIZATION
71*77.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
1*315. 2191. 1*1*68. 18U51.
$/1000CFM-HR
.185
A5-20
-------
Table A5-21
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 7
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UMITS/SERIES LENGTH NOZZLE PS I
6.0 8.0 1.0 20.0 15.0
GPM/NOZZLE
1.000 2.000 3.000 U.OOO 5.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 2000. 20. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3 NO Nl
1*2112. 52005. l.U 8.6 1.0 10.7 2287139 20971*53
U2112.
1*1596. 1*5560. 1.0 7.1 2.0 13.0 2188865 190U51U
1*1596.
1*11*57. 1*2660. .8 6.U 3.0 1«*.6 213331*8 17775H*
1*11*57.
<*1(*29. 1*091*0. .7 5.9 l».0 15.8 209U8K* 1680570
1*11*29.
1*11*1*8. 39783. .6 5.5 5.0 16.8 2065U05 1601508
1*11*29. l»09UO. .7 5.9 1*.0 15.8 2091*811* 1680570
1*11*29.
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
5.9 5.9 25000.0 831.3
NOZZLES/BANK BANKS/UNIT DDROP UM UNITS/SERIES
208. 1. 1*61. 1.
PRESSURE DROP FAN HORSEPOWER PUMP HORSEPOWER
1.500 10.705 23.272
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
20211*6. 1371*97.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
18967. '1039. 519. 939. 1720. l*19U. 13563. 1*091*0,
ANNUAL COST FOR SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICALS TOTAL $/1000CFM-HR
5322. 2535. 1265. 32307. 1*11*29. .1*11*
A5-21
-------
Table A5-22
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 8
CHAMBER TOTAL
NOZZLES/SQ FT LENGTH UNITS/SERIES LENGTH NOZZLE PS I
6.0 8.0 2.0 36.0 15.0
GPM/NOZZLE
2.000 2.500 3.000 it.000 5.000
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TEMPERATURE
CFM INLET OUTLET DEGREES F
25000. 2000. 20. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST
32372.
32372.
32297.
32297.
32267.
32267.
32279.
32338.
32267.
32267.
CAPITAL TIM-SEC
55Ht6.
53360.
52056.
50266.
<»9088.
52056.
1.
1.
1.
•
•
1.
2
1
0
8
7
0
HEIGHT
5.
5.
5.
k.
k.
5.
7
3
1
7
k
1
GPM/NOZ
2.
2.
3.
k.
5.
3.
0
5
0
0
0
0
FT2/FT3 NO
8.
8.
9.
9.
10.
9.
2
7
2
9
6
2
1378867
1359509
13t*389l*
1319620
1301093
13U3891*
Nl
13017U1
1271013
121*1*991
120201*2
1166955
121*1*991
OPERATING CONDITIONS OF TOWFR-CALCUlATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STFEl
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
5.1 5.1 25000.0 1*60.5
NOZZLES/BANK BANKS/UNIT DDROP UM UNITS/SERIES
151*. 1. l»39. 2.
PRESSURE DROP FAN HORSEPOWER PUMP HORSEPOWER
3.000 21.1*09 12.700
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
1291U3. 8I*01U.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
23661*. 1535. 519. 1298. 21*58. 5885. 16697. 52056.
ANNUAL COST FOR SCRUBBER SYSTEM
AMORTIZATION POWER MAINTENANCE CHEMICALS TOTAL $/1000CFM-HR
6767. 31*92. 1789. 20219. 32267. .323
A5-22
-------
Table A5-23
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 9
NOZZLES/SQ FT
6.0
GPM/NOZZLE
2.000 3.000
CHAMBER
LENGTH UNITS/SERIES
8.0 3.0
3.500
it.000
TOTAL
LENGTH
52.0
5.000
NOZZLE PS I
15.0
POWER
$/KWH
.025
FLOW RATE
CFM
25000.
UNIT COSTS FOR SCRUBBER SYSTEM
MAINTENANCE NOZZLES AMORTIZATION
$/SQRT(CFM) $ EA FACTOR
8.000 5.000 .130
CONCENTRATION HENRY
INLET OUTLET
2000. 20. 0.
SCHMIDT NO
0.
TEMPERATURE
DEGREES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SFC HEIGHT GPM/NOZ FT2/FT3 NO
31357.
31357.
31257.
31257.
31262.
31286.
31365.
31257.
31257.
63U91*.
60127.
5901*9.
58200.
56951.
60127.
I.I*
1.1
1.0
1.0
.8
1.1
l*.9
«*.U
k.2
l».l
3.8
«*.«*
2.0
3.0
3.5
i».o
5.0
3.0
6.3 1053200
7.0 1026U87
7.3 1016510
7.6 100791*6
8.1 993795
7.0 10261*87
Nl
1018011
981020
9661*36
953531
931338
981020
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
I*.I* I*.I* 25000.0 351.8
NOZZLES/BANK
117.
BANKS/UNIT
1.
PRESSURE DROP
I*.500
DDROP UM
1*39.
FAN HORSEPOWER
32.1U
UNITS/SERIES
3.
PUMP HORSEPOWER
9.587
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
107122. 69118.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS
26715. 1759. 519. 161*3. 3125. 7510,
AMORTIZATION
7816.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
1*51*1. 2191. 16708. 31257.
EXTERNALS
18857.
$/1000CFM-HR
.313
TOTAL
60127.
A5-23
-------
Table A5-24
DESIGN AtfD COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 10
NOZZLES/SQ FT
6.0
GPM/NOZZLE
2.000 2.500
CHAMBER
LENGTH
8.0
UNITS/SERIES
2.0
3.000
3.500
TOTAL
LENGTH
36.0
U.OOO
NOZZIE PS
15.0
POWER
$/KWH
.025
FLOW RATE
CFM
50000.
UNIT COSTS FOR SCRUBBER SYSTEM
MAINTENANCE NOZZLES« AMORTIZATION
$/SQRT(CFM) * EA ?i- FACTOR
8.000 5.000 .130
CONCENTRATION
INLET OUTLET
500. 10.
HENRY SCHMIDT NO
0.
0.
TEMPEPATUPE
DEGREES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3
31398.
31398.
31336.
31336.
31337.
31375.
311*31*.
31336.
31336.
89902,
87230
85279
83783
82596
87230
1.1
1.0
.9
.8
.8
1.0
7.6
7.1
6.8
6.5
6.3
7.1
2.0
2.5
3.0
3.5
2.5
7.k
NO Nl
121*1072 1181*231
7.8 122361*8 1158313
8.3 1209591* 11361*56
8.6 1197837 11171*1*1
8.9 11P77I*5 1100537
7.8 122361*8 1158313
OPERATING CONDITIONS OF TOWER-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(PFM) LIQUID(GPM)
7.1 7.1 50000.0 761*.2
NOZZLES/BANK
306.
BANKS/WIT
1.
DDROP UM
1*25.
PRESSURE DROP
3.000
FAN HORSEPOWER
1*2.818
UNITS/SFRIF.S
2.
PUMP HORSEPOWER
21.R77
CHEMICAL
CAUSTIC
LB/YEAR
75672.
CONSUMPTION
CHLORINE
LB/YEAR
1*1600.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS
36521. 3057. 81*9. 171*1. 3693. 7971*.
AMORTIZATION
1131*0.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
61*58. 2530. 11008. 31336.
EXTERNALS
33395.
*/1000CFM-HR
.157
TOTAL
87230.
A5-24
-------
Table A5-25
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 11
CHAMBFR TOTAI
NOZZLES/SO FT LFNGTH U»MTS/SP»I PS I.FNPTH N.r>7.7.ir PS 1
6.0 R.n 3.0 52.0 15.0
GPM/NOZZLE
2.000 2.500 3.000 3.500 I*. 000
UNIT COSTS FOR SCPUBPPP SYSTEM
POWFR MAINTENANCE NOZZLES AMORTIZATION
*/KWH «/SQRT(CFM) * FA FA.CTOD
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HCNPY SCHMIOT NO TEMPERATURE
CFM INLET OUTLET DFOPFES F
50000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HFIPHT GPM/NOZ FT2/FT3 NO
3359U. 103232. 1.3 6.6 2.0 5.6 9l*6«*2«*
33591*.
33539. 100310. 1.1 6.2 2.5 6.0 933137
3353^.
33553. 981P9. 1.0 5.9 3.0 6.3 9221*10
33607. 96573. .Q 5.7 3.5 6.6 Q13U53
3368U. 95299. .9 5.5 i*.o 6.P 905758
33539. 100310. 1.1 fi.2 2.5 6.0 933137
33539.
Nl
920721*
903500
876770
8F5857
903500
OPERATING COMOITIONS OF
THE MATERIAL OC CONSTPU^TION IS STAINLFSS STF^l.
SCRUBBER WIDTH SPPUBBFR HFIRHT TOTAL FLOW PATFS
FEET PFET ^AS(rFM) IIOUin(RPM)
6.2 6.2 50000.0 5P2.P
NOZZLES/BANK
233.
BANKS/UN IT
1.
DPPOP UM
I»25.
PRESSURE DROP
l».500
FAN HORSFPOWEP
6U.227
UNITS/SFPIFS
3.
PUMP HORSEPOWER
16.U17
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YFAR
60836. 31*219.
CAPITAL COSTS FOP SCRU3BFR SYSTEM
SHELL NOZZLES FAN PUMP MOTOPS TANKS EXTERNALS TOTAL
«*1187. 31*97. 8l*Q. 2?06. 1*691. 10166. 37711*. 100310.
AMORTIZATION
1301*0. \
ANNUAL COST FOP SCPUBBPP SYSTEM
POWER MAINTENANCE CHPM|PALS TOTAL
-------
Table A5-26
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 12
NOZZLES/SQ FT
6.0
GPM/NOZZLE
2.000 2.500
CHAMBER
LENGTH
8.0
UNITS/SERIES
3.0
3.000
It.000
TOTAL
LENGTH
52.0
5.000
NOZZLE PS I
15.0
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
$/KWH $/SQRT(CFM) $ EA FACTOR
.025 8.000 5.000 .130
FLOW RATE
CFM
50000.
CONCENTRATION
INLET OUTLET
2000. 20.
HENRY SCHMIDT NO
0.
0.
TEMPERATURF
DEGREES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SEC HEIGHT GPM/NOZ FT2/FT3
5906U.
59061*.
59017.
59017.
590U1*.
5920U.
591*30.
59017.
59017.
10771*0.
101*593.
102308.
99191*.
97168.
101*593.
1.1*
1.2
1.1
1.0
.8
1.2
7.0
6.6
6.3
5.8
5.1*
6.6
2.0
2.5
3.0
«*.o
5.0
2.5
NO Nl
6.3 1053236 10180U3
6.7 1038U50 997911
7.0 1026522
7.6 1007981
8.1 993829
6.7 1038U50
981051
953560
931366
997911
OPERATING CONDITIONS OF TOWFR-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL FLOW RATES
FEET FEET GAS(CFM) LIQUID(GPM)
6.6 6.6 50000.0 61*8.5
NOZZLES/BANK
259.
BANKS/UNIT
1.
DDROP UM
1*25.
PRESSURE DROP
I*.500
FAN HORSFPOWFR
61*.227
UNITS/SERIES
3.
PUMP HORSEPOWER
18.381
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
2U2l*5. 138237.
CAPITAL COSTS FOR SCRUBBER SYSTEM
SHELL NOZZLES FAN PUMP MOTORS TANKS EXTERNALS TOTAL
1*1*057. 3891. 8U9. 23«*9. 1*892. 108UO. 37711*. 101*593.
AMORTIZATION
13597.
ANNUAL COST FOR SCRUBBER SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
8905. 3098. 33U17. 59017.
$/1000CFM-HR
.295
A5-26
-------
Table A5-27
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 13
NOZZLES/SQ FT
6.0
RPM/NOZZLE
2.000 2.500
CHAMBER
LENGTH
8.n
UNITS/SFPIFS
2.n
3.000
3.500
TOTAL
I FMPTH
3F.n
I*.
NOZZLp PS I
15.0
POWER
UNIT COSTS FOR SCRUBBFP SYSTEM
MAINTENANCE NOZZIES AMORTIZATION
*/SORT(CFM) $ FA FACTOR
.025 8.000 5.000 .130
FLOW RATE
CFM
100000.
CONCENTRATION
INLET OUTLET
500. 10.
HFMPY SCHMIDT
0.
0.
TEMPERATURE
DFGRFES F
70.
INTERMEDIATE VALUFS DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SFC HEIGHT GPM/NOZ FT2/FT3 NO Nl
58528.
58528.
58550.
58659.
58817.
59002.
58528.
58528.
155855
151526
11*8363
11*5937
11*1*011
155855
1.1
1.0
.0
.?
.8
1.3
10.7
10.1
9.6
9.2
8.Q
10.7
7.1* 12«»1115 118U26R
2.5
3.0
3.5
i*.0
2.0
7.8 1223691
8.3 1209636
P.fi 1107879
8.9 11877P7
7.1* 12U1115
11583U9
11361*91
Ill7k7k
11005P9
11PU2P*
OPERATING COMPITIONS OF TOWFR-CALCULATFD
THE MATERIAL OF CONSTRUCTION IS STAINLESS STFPL
SCRUBBER WIDTH SCRUB^n HEIGHT TOTAL FLOW RATFS
FFET FFFT GAS(CFM) LIOUIP(GPM)
10.7 10.7 100000.n 1383.1*
NOZZLES/BANK
692.
BANKS/UNIT
1.
DDPOP UM
PRESSURE DROP
3.000
FAN HORSEPOWER
85.636
UNITS/SFRIFS
2.
PUMP HORSEPOWER
«*2.122
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YFAP
15131*5. 83199.
CAPITAL COSTS FOR SCPUBBFP SYSTEM
SHELL NOZZLES FAN PUMP MOTOPS TANKS
61092. 6917. U13. 2U65. 5793. 113P6,
AMORTIZATION
20261.
ANNUAL COST FOR SCPURBFR SYSTEM
POWER MAINTENANCE CHEMICALS TOTAL
12673. 3578. 2201F. 58528.
EXTERNALS
667PO.
*/inOOCFM-HP
TOTAL
155855,
A5-27
-------
Table A5-28
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 14
NOZZLES/SO FT
6.0
GPM/NOZZLE
1.000 1.500
CHAMBFR
LENGTH
16.0
UNITS/SFPIES
2.0
2.000
2.500
TOTAL
LENGTH
52.0
3.000
NOZZLE PSi
15.0
UNIT COSTS FOR SCRUBBER SYSTEM
POWER MAINTENANCE NOZZLES AMORTIZATION
S/KWH */SQRT(CFM) * FA FACTOR
.025 8.000 5.000 .130
FLOW RATE CONCENTRATION HENRY SCHMIDT NO TFMPFPATURF
CFM INLET OUTLET DEGREES F
100000. 500. 10. 0. 0. 70.
INTERMEDIATE VALUES DUPING OPTIMIZATION
ANNUAL COST CAPITAL TIM-S^C HEIGHT RRM/NOZ FT2/FT3 NO
58793. 1776P2. 2.6 11. P l.n U.P 1020?po
58793.
57929.
57929.
575U7.
575U7.
57381.
57381.
57329.
57329.
57329.
57329.
166055.
15Q029.
151*213.
150659.
150659.
2.
1.
1.
1.
1.
1
P
6
I*
k
IP.
ot
9.
8.
8.
(*
R
n
P
P
1
2
2
3
3
.5
. n
.5
.0
.0
5.
5.
P.
P.
6.
It
Q
9.
R
P
10P31P?
9P5PP?
971233
°PP07R
960078
Nl
95P217
92PPRP
9PP7P.8
88POPF
88806F
OPERATING CONDITIONS OF TOWEP-CALCUI.ATED
THE MATERIAL OF CONSTRUCTION IS STAINLESS STEEL
SCRUBBER WIDTH SCRUBBER HEIGHT TOTAL PLOW RATES
PEET FEET GAS(CFM) UOUIPCGPM)
8.6 8.R 100000.0 1316.0
NOZZLES/BANK
l»39.
BANKS/UNIT
1.
DPPOP UM
U39.
PRESSURE DROP
3.000
FAN HORSEPOWER
85.P.36
Uh'ITS/SERIFS
2.
PUMP HORSEPOWER
3S.P1I*
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YEAR
1513U5. 83199.
CAPITAL COSTS FOR S
SHELL NOZZLES FAN PUMP
59016. l»387. 1U13. 239«».
SYSTEM
MOTORS
Sfill.
TANKS
AMORTIZATION
19586.
ANNUAL COST FOP SCPUBREP SYSTFM
POWER MAINTENANCE THCMICALS TOTAL
12150. 3578. 22016. 57329.
EXTERNALS
66790.
«/1000CFM-HP
TOTAI
150P50,
A5-28
-------
Table A5-29
DESIGN AND COST CALCULATIONS FOR SPRAY SCRUBBERS
Case 15
NOZZLES/SO FT
6.0
GPM/NOZZLE
1.500 2.000
CHAMBER
LENGTH
16.0
UN ITS/SEP I
2.0
2.500
3.000
TOTAL
LENGTH
52.0
3.500
NOZZLE PS I
15.n
POWER
$/KWH
.025
FLOW RATE
CFM
150000.
UNIT COSTS FOP SCRUBBFR SYSTEM
MAINTENANCE NOZZLES AMORTIZATION
*/SQRT(CFM) $ EA FACTOR
8.000 5.000 .130
CONCENTRATION HFNPY
INLET OUTLET
500. 10. 0.
SCHMIDT NO
0.
TFMPEPATURF
DFfiPFES F
70.
INTERMEDIATE VALUES DURING OPTIMIZATION
ANNUAL COST CAPITAL TIM-SFC HEIGHT GPM/NOZ FT2/FT3
83653.
83653.
83258.
83258.
83132.
83132.
83U6.
8321*0.
83132.
83132.
229779.
220502.
21U1U3.
209t>52.
205818.
21UU3.
2.
1.
1.
1.
1.
1.
1
P.
6
U
3
6
12.
11.
11.
in.
10.
11.
7
7
0
5
0
0
1.
2.
2.
3.
3.
2.
5
0
t;
0
5
5
5.
5.
6.
6.
P.
6.
k
Q
?
F
8
2
1003203
9»5082
971252
9FHO°7
9507R5
971252
95623U
928T3
90BROU
88P082
87173P
906POU
OPERATING CONDITIONS OF TOWPP-CALCULATED
THE MATERIAL OF CONSTRUCTION IS STAJMLESS STFFL
SCRUBBFR WIDTH SCRUBBPP HFIGHT TOTAL FLOW RATES
FEET FFFT GAS(CFM) LIOUIP(GPM)
11.0 11.0 150000.0 1819.7
NOZZLES/BANK
728.
BANKS/UNIT
1.
DDROP UM
U25.
PRESSURE DROP
3.000
FAN HORSFPOWER
128.U55
UNITS/SERIES
2.
PUMP HORSFPOWFR
55.659
CHEMICAL CONSUMPTION
CAUSTIC CHLORINE
LB/YEAR LB/YFAP
227017. 12U799.
SHELL
81193.
CAPITAL COSTS F0» SCPtlR*rP
NOZZLES
\7279.
\
FAN
1912.
PUMP
28°P.
MOTORS
725R.
TANKS
131»21.
FXTFPNALS
******
ANNUAL COST FOP S^PMB^FP SYSTF"
AMORTIZATION POWFR MAINTENANCE CHEMICALS TOTAL «/1000CFM-HR
27839. 17887. l»3*2. 3302U. 83132. .139
TOTAL
A5-2.9
-------
Appendix 6
MASS TRANSFER AS A FUNCTION OF PACKING DEPTH
A6-1
-------
MASS TRANSFER AS A FUNCTION OF PACKING DEPTH
In tests 31 and 32, two stages of alkaline sodium hypo-
chlorite were tested. Each stage contained 2 ft of packing.
The Number of Mass Transfer Units (NTUG's) for each test were
calculated, and the average NTUG's computed:
Test No.
Inlet ED 50
Outlet ED.
NTUG*
31-1
31-2
31-3
32-1
32-2
32-3
32-4
32-5
Average
700
15,500
590
21,500
2,700
5,000
19,000
16,000
10,000
25
45
45
15
35
450
300
240
140
3.33
5.84
2.57
7.26
4.35
2.41
4.15
4.20
4.26
In test 33, alkaline sodium hypochlorite was tested in a
one-stage scrubber, containing 4 ft of packing. The average
NTUG is calculated below:
Test No.
33A
33A
33B
33B
Average
Inlet ED.
15,000
15,000
4,300
4.300
9,650
Outlet ED so
680
20
350
60
277
NTUG*
IT, Inlet ED so
Outlet ED 50
A6-2
-------
TECHNICAL REPORT DATA
(Please read Jnuructions on the rc\ cnc before completing)
NO.
EPA -600/2 -76 -009
2.
iLE AND SUBTITLE
Odor Control by Scrubbing in the Rendering Industry
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
January 1976
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) .
Richard H. Snow and James E. Huff (ITTRI);
and Werner Boehme
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORQANIZATION NAME AND ADDRESS
Fats and Proteins Research Foundation, Inc
2720 Des Plaines Avenue
Des Plaines, Illinois 60018
10. PROGRAM ELEMENT NO.
1AB015: ROAP 21AXM-062
11. CONTRACT/GRANT NO.
68-02-1087
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF FIEPORT AND
Final; 6/73-11/75
PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
repOrt gjyes results of experiments conducted at a rendering plant to
obtain data needed to design wet scrubber systems for rendering plant odor control.
Scrubber performance was measured by both odor panel and gas chromatographic
analysis. Experiments in a three-stage packed-bed laboratory-scale scrubber at the
rendering plant evaluated solutions of sodium hydroxide and the strong oxidants
sodium hypochlorite , hydrogen peroxide, and potassium permanganate. Since
removal of 90% per stage was obtained with fresh alkaline sodium hypochlorite sol-
ution, this reagent was selected for subsequent longer-term tests. A 2-week test of
a plant-scale horizontal spray scrubber, operating on plant ventilating air, showed
odor removal of 83%. The outlet odor units averaged 64: the inlet ranged from 165
to 2500 odor units. A three-stage packed-bed scrubber was evaluated to replace
an existing incinerator being used to treat a process air stream that contained from
5000 to 50,000 odor units. A week- long test with the scrubber gave a lower-than-
expected average odor reduction of 85%. Further work is suggested to investigate
conditions necessary to improve the results. Data was obtained on chemicals con-
sumption and effect of flow variables on odor removal; these data were used to up-
date computer models that can be used to design scrubbers for odor removal.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Industrial Plants
Fats
Oils
Scrubbers
Odor Control
Inc ine r ators
Gas Chromatography
Sodium Hydroxide
Sodium Hypochlorite
Hydrogen Peroxide
Potassium Perman-
ganate
Mathematical Models
b.lDENTIFIERS.'OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Rendering Plants
Odor Panel
c. COSATI Field/Group
13B
06A
07A
J2A
07D
07B
= IBUTION STATEMEN
Unlimited
19. SECURITY CLASS (Tills Report)
Unclassified
21. NO. OF PAGES
219
20. SECURITY CLASS (Thispage/
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
A6-3
------- |