THE MEASUREMENT AND CONTROL OF
FOULING IN FINE PORE DIFFUSER SYSTEMS
by
Edwin L. Barnhart and Michael Collins
Southern Methodist University
Dallas, TX 75212
Cooperative Agreement No. CR812167
Project Officer
Richard C. Brenner :
Water and Hazardous Waste Treatment Research Division
Risk Reduction Engineering Laboratory i
Cincinnati, Ohio 45268 i
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
Development of the information in this report has
funded in part by the U.S. Environmental Protection Agency
Cooperative Agreement No. CR812167 by the American Society
Civil Engineers. The report has been subjected to Agency
and administrative review and approved for publication
document. Mention of trade names or commercial products
constitute endorsement or recommendation for use.
been
under
of
peer
as an EPA
does not
-------
FOREWORD |
Today's rapidly developing and changing technologies and
industrial products and practices frequently carry with them the
increased generation of materials that, if improperly dealt with,
can threaten both public health and the environment. The U.S.
Environmental Protection Agency (EPA) is charged by Congress with
protecting the Nation's land, air, and water resources'. Under a
mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance
between human activities and the ability of natural systems to
support and nurture life. These laws direct EPA to perform
research to define our environmental problems, measure' the
impacts, and search for solutions. l
i
The Risk Reduction Engineering Laboratory is responsible for
planning, implementing, and managing research, development, and
demonstration programs to provide an authoritative, defensible
engineering basis in support of the policies, programs; and
regulations of EPA with respect to drinking water, wastewater,
pesticides, toxic substances, solid and hazardous wastes, and
Superfund-related activities. This publication is one;of the
products of that research and provides a vital communication link
between the researcher and the user community. \
As part of these activities, an EPA cooperative agreement
was awarded to the American Society of Civil Engineers!(ASCE) in
1985 to evaluate the existing data base on fine pore diffused
aeration systems in both clean and process waters, conduct field
studies at a number of municipal wastewater treatment facilities
employing fine pore aeration, and prepare a comprehensive design
manual on the subject. This manual, entitled "Design Manual -
Fine Pore Aeration Systems," was completed in September 1989 and
is available through EPA's Center for•Environmental Research
Information, Cincinnati, Ohio 45268 (EPA Report No. EPA/625-1-
89/023) . The field studies, carried out as contracts under the
ASCE cooperative agreement, were designed to produce reliable
information on the performance and operational requirements of
fine pore devices under process conditions. These studies
resulted in 16 separate contractor reports and provided critical
input to the design manual. This report summarizes the results
of one of the 16 field studies. '•
i
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
111
-------
PREFACE i
In 1985, the U.S. Environmental Protection Agency' funded
Cooperative Research Agreement CR812167 with the American Society
of Civil Engineers to evaluate the existing data base on fine
pore diffused aeration systems in both clean and process waters,
conduct field studies at a number of municipal wastewajter
treatment facilities employing fine pore diffused aerajtion, and
prepare a comprehensive design manual on the subject. ; This
manual, entitled "Design Manual - Fine Pore Aeration Systems,"
was published in September 1989 (EPA Report No. EPA/725/1-89/023)
and is available from the EPA Center for Environmentalj Research
Information, Cincinnati, OH 45268. I
i
i
As part of this project, contracts were awarded under the
cooperative research agreement to conduct 16 field studies to
provide technical input to the Design Manual. Each of' these
field studies resulted in a contractor report. In addition to
quality assurance/quality control (QA/QC) data that may be
included in these reports, comprehensive QA/QC information is
contained in the Design Manual. A listing of these reports is
presented below. All of the reports are available from the
National Technical Information Service, 5285 Port Royal Road,
Springfield, VA 22161 (Telephone: 703-487-4650). :
1. "Fine Pore Diffuser System Evaluation for the Green Bay
Metropolitan Sewerage District" (EPA/600/R-94/093) by J.J.
Marx . ;
2. "Oxygen Transfer Efficiency Surveys at the Jones Island
Treatment Plants, 1985-1988" (EPA/600/R-94/094) by R.
Warriner j
3. "Fine Pore Diffuser Fouling: The Los Angeles Studies"
.(EPA/600/R-94/095) by M.K. St exist ram and G. Masutani
4. "Oxygen Transfer Studies at the Madison Metropolitan
Sewerage District Facilities" (EPA/600/R-94/096) by W.C.
Boyle, A. Craven, W. Danley, and M. Rieth
5. "Long Term Performance Characteristics of Fine Pore Ceramic
Diffusers at Monroe, Wisconsin" (EPA/600/R-94/097) by D.T.
Redmon, L. Ewing, H. Melcer, and G.V. Ellefson '<
6. "Case History of Fine Pore Diffuser Retrofit at Ridgewood,
New. Jersey" (EPA/600/R-94/098) by J.A. Mueller and P.D.
Saurer
i
iv l
-------
7. "Oxygen Transfer Efficiency Surveys at the South :Shore
Wastewater Treatment Plant, 1985-1987" (EPA/600/R-94/099) by
R. Warriner ;
8. "Fine Pore Diffuser Case History for Frankenmuth,i Michigan"
(EPA/600/R-94/100) by T.A. Allbaugh and S.J. Kang
9. "Off-gas Analysis Results and Fine Pore Retrofit information
for Glastonbury, Connecticut" (EPA/600/R-94/101) by R.G.
Gilbert and R.C. Sullivan ;
10. "Off-Gas Analysis Results and Fine Pore Retrofit Case
History for Hartford, Connecticut" (EPA/600/R-94/105) by
R.G. Gilbert and R.C. Sullivan !
11. "The Measurement and Control of Fouling in Fine Pore
Diffuser Systems" (EPA/600/R-94/102) by E.L. Barnhart and M.
Collins ;
12. "Fouling of Fine Pore Diffused Aerators: An Interplant
Comparison" (EPA/600/R-94/103) by C.R. Baillod and K.
Hopkins
13. "Case History Report on Milwaukee Ceramic Plate Aeration
Facilities" (EPA/600/R-94/106) by L.A. Ernest
14. "Survey and Evaluation of Porous Polyethylene Media Fine
Bubble Tube and Disk Aerators" (EPA/600/R-94/104); by D.H.
Houck :
15. "Investigations into Biofouling Phenomena in Fine^ Pore
Aeration Devices" (EPA/600/R-94/107) by W. Jansen, J.W.
Costerton, and H. Melcer :
16. "Characterization of Clean and Fouled Perforated Membrane
Diffusers" (EPA/600/R-94/108) by Ewing Engineering Co.
-------
ABSTRACT
The purpose of the study was two-fold: first, to'define the
efficiency of various methods of cleaning fine pore diffusers
and, second, .to develop a methodology that could be used to
evaluate the efficiency of the cleaning techniques. Dirty fine
pore domes from the North Texas Municipal Water District were
cleaned by a variety of techniques, and the improvement in oxygen
transfer efficiency was measured. The domes were reinstalled in
the aeration tanks and withdrawn at various time intervals
thereafter. The deterioration in oxygen transfer efficiency was
then noted. The cleaning techniques were repeated, and the
improvement in transfer was recorded.
Overall, the domes form the North Texas Plant did not show
severe fouling. Low pressure hosing appeared to be as'effective
as any other method in cleaning the domes. The domes :
deteriorated promptly after they were reintroduced into the
aeration tank, but the deterioration in oxygen transfer was not
severe enough to impose an unacceptable aeration cost.
The technique of using an off-line aeration tank for
studying the cleaning techniques provided mixed results. The
comparison of cleaning techniques appeared to be properly
described in this small test tank, but the degree of fouling that
had actually occurred in the full-scale plant appeared;to be
underestimated. This probably resulted from the breakdown of
slimes and fouling materials during dome transportation and
handling. :
The cost of cleaning domes by various techniques is
difficult to estimate because of a variety of site specific
factors. A method was developed for estimating the cost that
would be encountered in a typical case. The cost for simple
cleaning was found to vary from approximately $1.20 a dome for
small plants to somewhat under $0.80 a dome for large plants.
This report was submitted in partial fulfillment of
Cooperative Agreement No. CR812167 by the American Society of
Civil Engineers under subcontract to Southern Methodist
University under the partial sponsorship of the U.S. !
Environmental Protection Agency. The work reported herein was
conducted over the period of 1985-1987. ;
VI
-------
CONTENTS
Foreword ...................... ;. . . . iii
Preface ........... ......... ....... iv
Abstract ...................... .. . . . vi
Figures
x
Tables ....................... i ..... x
Acknowledgements
xi
Introduction .................... I . . 1
i
Study Purpose .................... I .... 3
Description of SMU Test Tank ................. 5
Initial Problems .................. i. . . . 9
The Pressure Measurements .............. • 17
Testing With Detergents Added ............ ; . . . 20
Uniformity of Test Tank Results ..... ...... : . . . 24
!
Studies at North Texas Municipal Water District, Rowlette
Creek Plant .................... .... 27
i
Dome Cleaning Operations ; ....... ...... 1 ... 33
Dome Collection ................ 1 ! ] ! 33
Low Pressure Hosing .............. ]'.'.'. 35
High Pressure Hosing .............. . . . . . 36
Milwaukee Method ................ i . ! ! 36
Steam Cleaning .... ....... ......]... 37
Kilning .............. _ ..... . ! ! ! 38
Microscopic Photographs of the Domes ........ j 39
Study Results of Cleaning Investigations ...... .... 45
i
Off-Gas Testing ................... : . . . 52
Clean Water Studies ......... ......... • 54
vii :
-------
CONTENTS (continued)
Observations on Diffuser Cleaning . j 55
General Discussion of Results ........ 58
Evaluation of Dome Air Flow Characteristics 60
Special Studies - New Domes . . 63
Cost of Dome Cleaning .• 68
i
Factors That Influence Tank Cleaning Costs ;..... 71
The Unit Operations of Tank Cleaning ' . . . 73
General Observations ' . . . 78
Overall Discussion of Results . . 81
References I ... 82
Appendices
Appendix 1 ; . 84
Appendix 2 ' , . 112
Appendix 3 • . . . 118
-------
FIGURES
i
Number : Page
1 Schematic of Shop Test Tank System at SMU . ... I .... 6
2 Photo of Test Tank 7
3 Effect of Airflow Rate on OTE Preliminary Study . I . . . 10
i
4 Norton Dome Diffusers Assembly 1 ... 12
5 Effect of Airflow Rate on OTE After Proper Sealing'of the
Domes 1' . 15
6 Pressure Loss With the Time After Installation ..... 18
7 Comparison of Kla @ Locations . . . 25
8 Rowlette Creek Regional Wastewater . . . 28
9 Aeration Tank Under Study ...:... 29
10 Dome Layout in the Test Section 34
lla Electron Micrograph of a Dirty Dome i 41
lib Electron Micrograph of a Dome After Low Pressure Hosing . 42
lie Electron Micrograph of a Dome After Acid Washing > . . . 43
lid Electron Micrograph of a Dome After Kilning ....... 44
12 Dynamic Wet Pressure of Contaminated Domes ...:... 61
13 Dynamic Wet Pressure of Cleaned Domes j . . 62
IX
-------
TABLES I
Number ' Page
1 Impact of Air Flow Rate on OTE ; ... 14
i
2 Impact of Detergent Level on Oxygen Transfer Rate '
Coefficient j 22
}
3 Operating Condition at Study Plant:, NTWD ........ 31
4 "aSOTE" of Domes Before and After Internal Cleaning as
Tested in Detergent Solution .; . . . 46
5 "aSOTE" of Clean Domes After 9 Months of Use ...;... 47
6 "aSOTE" of Domes After 21 Months 48
7 Summary of Diffuser Cleaning Data Detergent Testing . . .. 50
8 Rowlette Creek Aeration Study ...... 53
9 Recleaning After 21 Months Service ; 55
10 Clean Dome Study 64
11 New Dome Study - Triplicate Runs in Clean Water . . . . . 66
-------
ACKNOWLEDGEMENTS ;
The work reported herein was supported, in part, by a grant from
the U.S. EPA through the American Society of Civil Engineers. Most
i
of the laboratory testing was completed under the supervision of
Dr. Michael Collins of Southern Methodist University. ; Two graduate
stxidents, Randall Covington and Ramarao Vuddagiri, performed the
actual tests.
The cooperation of the management and staff of the North Texas
Municipal Water District is gratefully acknowledged, i
-------
INTRODUCTION
Each year the United States spends more than I$500 million
to transfer oxygen to waste liquor during various wastewater
treatment processes (1). The cost of aeration is increasing
with the demand for high-level waste treatment.! To achieve
\
the most efficient oxygen transfer, many existing treatment
plants have installed fine pore diffused aeration systems.
Most new municipal plants are also employing this
i
technology. Investigations at some of these plants
indicated that diffuser fouling may have a significant
i
impact on the oxygen transfer efficiency (2). :The lowering
of transfer efficiency by fouled diffusers will add
substantially to the wastewater treatment cost, i
To better define the conditions contributing to diffuser
fouling and to develop methods for evaluating and techniques
i
for controlling this problem, the American Society of Civil
Engineers, under a grant from the Environmental Protection
Agency has undertaken a study program to develop information
and guidelines to improve the design and application of fine
pore diffused aeration systems.
-------
Southern Methodist University (SMU), Center for-Urban Water
Studies, received a grant from the study program> as well as
support from several municipal districts interested in
promoting developing information on oxygen transfer in fine
pore aeration systems. The SMU studies have been directed
toward methods of quantifying the degree of deterioration in
oxygen transfer and evaluating various cleaning techniques
for diffuser systems.
-------
STUDY PURPOSE
The purpose of the study was two-fold: first, to develop a
procedure to evaluate diffuser cleaning techniques and
second, to investigate the improvement in oxygen transfer
efficiency achieved by the various techniques. The
advantage of knowing the efficiency of various cleaning
i
methods is clear; however, the information is not totally
useful unless the rate of fouling after a dome is cleaned
using the various techniques can also be determined. The
function of defining the rate of fouling is important for
modeling the process or determining an economic optimization
of the process. !
i
The first objective of the studies, the development of a
procedure to evaluate dome cleaning techniques, is quite
i
important. Existing field data suggest that the optimum
method of cleaning may well be different 'at different
plants. If this is the case, a method of evaluating
cleaning techniques that can be applied to'a particular
plant is needed. Because of limited resources, conducting
large-scale studies on a wide variety of cleaning methods is
impractical for many plants. A more practical technique
-------
would be a shop evaluation of the cleaning techniques.
E
The study program presented in this report involved removing
fine pore dome diffusers from the North Texas Municipal
Water Treatment Plant in Rowlette. These domes were cleaned
by various techniques, and their oxygen transfer efficiency
was evaluated in a shop scale tank located at SMU. The
cleaned domes were then placed back in service for periods
of up to 21 months. Selected domes were removed from the
i
aeration tank at approximately 10 months and 21 months and
retested at SMU to evaluate the deterioration of oxygen
transfer efficiency. 1
-------
DESCRIPTION OF SMU TEST TANK
The SMU Test Tank is a steel tank coated with an epoxy
lining. The tank is 20 feet long by 3 feet 6< inches wide
and has a sidewall depth of 9 feet 6 inches. The operating
volume is approximately 17,200 liters. The tank; is equipped
with glass windows located at several points! so that the
aeration process can be observed and photographed from
outside the tank. The tank is shown schematically in Figure
1 and in a photo in Figure 2. :
Air is supplied to the tank from a central' compression
i
system that contains a large reservoir so that the air can
be fed at a constant temperature and pressure. Air from the
compression system flows through a series of metering and
control valves and finally through a dual rotameter system
that allows precise air flow measurement over a wide range.
The rotameters from the air system are tested at regular
intervals in the adjacent "Hydraulic Measurements
Laboratory", which contains accurate and precise equipment
for instrument calibration. The tank is also equipped with
a pressure measuring device so that the exact head loss
through the aeration equipment can be measured, i
-------
Ǥs
k
|
o>
I
\"T
"Tv ,-8*»ty
fe
I
|>_\ """.
^
I
^
^V
a
3
3
/
u
11
1
t
I
o»
II
IDi
:>:
COi
UJ!
CO!
o
X
CO
u_
o
o
bJ
n:
o
CO
bJ
CE
O
U_
-------
FIGURE 2. PHOTO 'OF TEST TANK
-------
During oxygen transfer testing, three YSI dissolved oxygen
probes are placed to measure representative portions of the
total tank volume. Oxygen transfer testing is conducted in
accordance with the procedures described in the ASCE "A
Standard for the Measurement of Oxygen Transfer in Clean
Water." (3) ;
For the purposes of this study the tank was fitted with a
four-inch air header containing 10 diffuser assemblies for
Norton Domes. Any combination of these assemblies could be
used to install domes. The assemblies that were not used
were plugged during the tests.
j
A detailed description of the study procedures iis presented
as Appendix 1. \
-------
INITIAL PROBLEMS
i
Initial studies were conducted using clean water. This was
done to establish a baseline for the system and to assure
that the test apparatus could duplicate conditions observed
by other investigators. Oxygen transfer efficiency was
studied over a range of air flows from 0.5 to 2.5 cubic feet
per minute (cfra) per dome. Each run was:conducted in
triplicate. The results of these initial tests are
presented in Figure 3. '•
f
Previous investigations (4) have shown that although some
oxygen is transferred during formation and bursting of the
bubbles, this effect is relatively minimal when!dealing with
fine pore diffusion systems. Over the range of,8 to 16 feet
of water depth, the oxygen transfer per foot of 1 depth should
be almost constant. As shown in Figure 3, these tests did
not match the performance estimates of the manufacturer. A
comparison" of the initial test data shows that at low air
flow rates the observed performance in the SMU aeration
system was close to that reported by the equipment
manufacturer. However, as the air flow rate increased,
deviation from the manufacturer's reported; performance
-------
2.0 -r-
Q_
Ul
Q
LJ
O
1.5 -r
1.0 --
Manufacturers
Data
-------
increased. This anomalous" behavior was investigated by
photographing the submerged domes under conditions of
increasing airflow. The pressure drop associated with
various airflows was also evaluated. The photographs showed
that as the air flow increased so did the percentage of
large bubbles. Increased air flow resulted in little
increase in pressure loss. These investigations determined
that the gasket between the header base plate'and the dome
was not providing a proper seal. The dome mounting system
is shown schematically in Figure 4. As the air flow
increased, the air leakage around the gasket also increased.
A dome mounting system was set up outside the1test tank to
evaluate the mechanics of gasket sealing. ;The dome was
fastened to the mounting apparatus by a brass[bolt passing
through the center of the dome. This bolt is;tightened to
compress the gasket between the base plate:and the dome
until a seal is obtained. Laboratory studies determined
that compressing the gasket to effect a tight seal was
impossible without cracking the dome. This indicated that
the gaskets being employed were much too rigid. The rigid
gaske'ts were replaced with a more ductile gasket that
properly sealed the system. ,
11
-------
UJ
O
3
mi
2
LiJ
00
001
<
00
02
UJI
001
51
Ld
2
o
Q
Z!
O
£
O
LJ
C£
^>
O
u_
12
-------
The increased ductility of the gasket was needed to
compensate for warping of the base plate. The degree of
warp was determined using a small micrometer wheel to
measure the level of the base plate around its perimeter.
By moving the micrometer slowly around the surface, the
degree to which the surface was not flat could be
determined. Investigation of ten separate units showed a
typical unit to be out of flat by approximately :0.05 inches.
Individual units showed warping as much as 0.1; inches from
the high to the low point on the plane. :
The gaskets in the test tank were then replaced with the
more ductile gaskets. Care was taken to ensure that no
leaking would occur in the system. After the; new gaskets
were installed and fully checked, a second set of clean
water tests (Figure 5). The test study results are also
presented in Table 1. When the system is properly sealed,
the performance was virtually identical to those' reported by
the manufacturer. ;
One other problem worthy of note developed during the
I
initial test program. The City of Highland Park, which
13
-------
TABLE 1
IMPACT OF AIR FLOW RATE ON OTE
Air Flow
CFM/Dome
0.5
1.0
1.5
2.0
2.5
Clean Water Studies
*
KLa
l/hr@20°
1.10
2.40
3.35
4.50
5.30
@ 8.5' Depth
i
OTE
!
16.0 !
17.1 .
16.1 :
16.0 I
15.2 i
OTE /FT
1.9
2.0
1.9
1.9
1.8
14
-------
Q.
LU
Q
>+—
O
U_
2.0 —
1.5 -L-
1.0 --
0.5 --
0.5
1.0
1.5
2.0
AIR FLOW - CFM/DOME
2.5
O Norton Domes ® 16' Depth
O Present Study, Norton Domes & 8.5'
FIGURE 5, EFFECT OF AIRFLOW RATE ON OTE
AFTER PROPER SEALING OF THE DOMES
15
-------
provides potable water for Southern Methodist University,
utilizes ferrous sulfate as a coagulating chemical. The
dosage of this chemical is higher during the warmer months
in response to their increased coagulation needs.
i
As a result, tests conducted in the late spring and early
summer were influenced by this change in water chemistry.
Observation of the tank indicated that a darkened color was
developing when adding the test chemicals. Investigation of
i
the water chemistry revealed that an iron: complex was
precipitating. This iron complex had a slight absorptive
effect on the cobalt, which is a catalyst used during the
test. Consequently, if a slight excess of cobalt was not
added, the effective cobalt concentration in the tank would
drop below the minimum specified for good testing. The
problem was resolved by increasing the cobalt concentration
by approximately 0.3 mg/1. j
The problem of color persisted and made it:difficult to
provide accurate photographic evidence of transfer during
the period when the higher chemical use was in effect at the
water plant. \
16
-------
THE PRESSURE MEASUREMENTS
t
Pressure loss across the diffuser devices were carefully
measured during each test-run. A bubbler tube was inserted
into the tank at an elevation equal to the center•of the gas
manifold. This pressure reading was subtracted from the
pressure drop across the diffuser system to determine the
actual pressure loss across the diffusers. ;
i
Initial studies of the diffuser pressure loss indicated that
a significant time period was required for the diffusers to
come to *equilibrium. The magnitude of pressure loss
increased as dry domes became saturated and decreased as
I
previously wetted domes dried under airflow; conditions.
Figure 6 shows a summary of pressure loss data after time of
aeration. :
The investigation concluded that domes must be* operated at
the intended airflow for approximately 24 hours before a
true equilibrium pressure is obtained. Evaluating the
impact of changing air pressure relationships on oxygen
transfer were not practical. Physical observation of the
systems indicated that the air flow from domes changed
17
-------
.2
Domes Initally Saturated
Final Pressure
Domes Initaliy Dry
0 2 4 6 8 10 12 14 16 18 20 22 24
Hours
FIGURE 6, PRESSURE LOSS WITH TIME AFTER INSTALLATION
18
-------
somewhat as they reached equilibrium. This observation
indicates that the oxygen transfer capability of a dome
system is influenced, to some degree, by the dome's
i
i
condition and the period that it is allowed to operate
before testing. In the present investigations, domes to be
tested were installed in the tank and allowed tjo aerate for
periods in excess of 12 hours before studies were conducted.
In most cases, the domes were aerated at least overnight to
allow the equilibrium to be established. ;
19
-------
TESTING WITH DETERGENTS ADDED
It was recognized from experience that the most valid
comparison of the cleaning efficiency could be obtained by
testing in water that simulated, to the degree possible, the
conditions that were actually observed in the field. The
decision was made to adopt a test fluid that contained
approximately 5 mg/1 active detergent; essentially similar
to that proposed by the British researchers (5). This fluid
would be the basis of comparison to be used throughout the
study. Parallel studies were also conducted in clean water
to provide a basis for comparison to the detergent tests.
A stock detergent solution was prepared using[a mixture of
household laundry detergent and dishwasher detergent that
has low foaming characteristics. Three to' 5 mg/1 of
detergent was found to be of an acceptable range so that a
significant impact on the oxygen transfer ^process with
minimum foaming was observed. ,
t
Analytical testing by the methylene blue extraction method
(6) for the presence of the detergents proved to be erratic.
Laboratory concentrations of the detergent ;taken before
20
-------
testing indicated concentrations generally equal to those
calculated from the stock mixture. However, after one or two
tests, the concentration of detergent seemed to vary
t
randomly. This variance is likely because of entrainment
and the reaction of the materials in the fluid. '
Investigations indicated that a much more satisfactory
method of tracking the presence of surface active materials
in the aeration system is to perform periodic evaluations of
the surface tension of the fluid. The intended range of
detergent concentration corresponded to a surface tension of
approximately 65 dynes per square centimeter as measured by
a surface tensiometer. Surface tension was chosen as the
preferred method for tracking the condition of the test
fluid.
Table 2 presents the studies conducted to determine the
impact of the detergent concentration on oxygen transfer
efficiency. The study indicated that the :alpha of the
detergent system was approximately 0.67 at a surface tension
of 63 to 65 dynes per square centimeter '(dynes/cm2).
Because this level was determined to be an acceptable level
that corresponds well with alpha values observed in the
21
-------
TABLE 2
IMPACT OF DETERGENT LEVEL ON
Intended
SAA Cone
(mg/1)
0
1.
2
4
6*
OXYGEN
KLa
(1/hr 20
2.36
1.77
1.8
1.6
1.4
TRANSFER RATE
Surface
0 ) Tension
(dynes/cm2
72
69
67
64
60
i
COEFFICIENT !
Measured '
SAA CONC.
I
) (mg/1)
o !
1-2
1-2 :
2-4 ;
3-6 !
a
0.75
0.76
0.67
0.60 '
* sustained foaming observed.
All runs are the average of duplicate studies,
22
-------
effluent of well performing treatment plants, a surface
tension of approximately 65 dynes/cm2 was used^ as the basis
for comparison during the remainder of this study.
23
-------
UNIFORMITY OF TEST TANK RESULTS
To evaluate the uniformity of the test tank results, 20 test
runs on clear water were performed and analyzed. In each
run five domes and flow rates of 1 and 2 cfm/dome were used.
> s
Three probes were located in the test tank' as shown in
Figure 7, the first probe was located at the left end of the
tank, approximately 1/4 depth above the bottom. The second
probe was located at the middle of the tank, while the third
probe was located at the upper right corner of the tank.
The meters and probes themselves were rotated^ on a random
basis so that the same meter and probe were not usually in
the same location on consecutive runs. Probes were
calibrated at the beginning of each run, and the! membrane on
probes for all systems were changed at regular intervals.
i
The results of a comparative study of the mass transfer
coefficient calculated at each location are presented as a
graphic summary in Figure 7. The average volumetric mass
transfer coefficient (KLa) for the tank was 1:. 80/hr. The
variation from point to point was less than 3 percent. The
two end locations were slightly less than the average, and
24
-------
Probe Locations
KlA/>ir - 1.79
-t-12/-8
-------
the middle location was slightly above average. '
Examination of the individual data shows that Station 1 was
higher than the average 12 times and lower 8 times. Station
2 values were higher 13 times and lower 7, and ,at Station 3
values were higher 11 times and lower 9. Overall, these
data describe a very uniform test tank where each of the
points exhibits essentially the same value of
26
-------
STUDIES AT NORTH TEXAS MUNICIPAL WATER DISTRICT,
ROWLETTE CREEK PLANT
The North Texas Municipal Water District, Rowlette Creek
Plant, serves the Cities of Piano and Allen, Texas. A
schematic of the waste treatment plant is presented as
Figure 8. Figure 9 shows the aeration tank;under study.
The wastewater receives primary settling and then is pumped
i
to the aeration system. The plant was treating
approximately 15 million gallons per day (MGD) during the
study period. The aeration system consists of two basins,
each with a volume of approximately 2.4 million gallons.
The flow pattern to the aeration basin varies;depending on
the rate of flow entering the treatment plant. ;
i
Returned sludge is introduced into the head of the aeration
basin. Under average flow conditions, the waste is
introduced into aeration Basin 2 where it mixes with the
return sludge and proceeds through Basin 3. ,If the flows
become high, because of peak demand or rainfall, the
influent is diverted to a second influent point in Tank 1.
This had the effect of providing additional detention time
for treatment. This process is initiated automatically by
i
the positioning of the inlet structures in the tank.
27
-------
H
LUI
IS
-------
<
a.
LU
LU
CC
O
LU
^
o
a:1
UJI
Cd
LU
en
LU
CE
Z)
o
i
29
-------
The mixed liquor suspended solids in the aeration basins are
normally maintained in the 2,500-mg/l range, and uptakes
observed in the test segment are generally in Ithe 25 to 50
mg/l/hr. Table 3 summarizes the operating conditions during
the study period. The plant does not keep separate records
of activated sludge wastage so sludge age cannot be
calculated directly. Indirect calculations indicate that
the sludge age during the study period varied from 20 to 14
days.
The Norton Domes located in the aeration •• basin were
installed in 1982 and had never been cleaned except for
periodic washdown before the present study. , No detailed
history of the system is available but discussions with
plant personnel indicate that continued problems with breaks
in lines were encountered after the original start up. The
i
system was overhauled, and all faulty piping was replaced in
the test section in 1983. 1
The air flow to the plant is supplied by either1two or three
2,250-cfm blowers. Depending on oxygen demand, there are no
individual air flow meters within the aeration system so
30
-------
TABUS 3
OPERATING CONDITIONS AT SIDDY PLANT: HTMWD
DATE
NOV 85
DEC 85
JAN 86
FEE 86
MAR 86
APR 86
MAY 86
JUNE 86
JULY 86
AUG 86
SEPT 86
OCT 86
NOV 86
DEC 86
INF BOD
CONG MG/1 F/M
156
160
182
148
188
138
183
99
140
150
134
153
150
132
0.25
0.25
0.23
0.18
0.18
0.18
0.27
0.13
0.17
0.21
0.18
0.23
0.20
0.19
% BOD
A REDUCTION
10
10
12
17
17
17
8
21
17
14
13
12
15
16
95%
93%
93%
92%
94%
94%
94%
87%
93%
94%
94%
92%
93%
90%
1 1 INF BOD % BOD
|| DATE CONG MG/1 .F/M A REDUCTION
If JAN 87 132 b.20 15 90%
II
|| FEE 87 140 0.25 10 90%
1 1
1 1 !
|| MAR 87 120 0.23 12 92%
II I
| | APR 87 163 0.23 12 93%
1 1
1 1
|| MAY 87 122 0.18 17 93%
II !
|| JUNE 87 110 0.16 19 92%
1 1
1 1
|| JULY 87 144 0.2 15 95%
1 1
II ;
|| AUG 87 146 0.24 11 94%
1 1
1 1
H.
.
H . ;
H!
.
II
H:
•
ii ;
Hp
!
II
A = ESTIMATED SLUDGE AGE
(DAYS)
31
-------
airflow is adjusted by observing the dissolved oxygen level
in the tanks and adjusting the airflows until the system
balances. With two blowers running, the system provides
approximately 1 cfm/dome and with three blowers operating,
the air flow is 1.5 cfm/dome. Aeration Basin 2, where the
studies are conducted, has a volume of 0.225 million
gallons. i
32
-------
DOME CLEANING OPERATIONS
In the summer of 1985 a cleaning program to prepare the
domes for testing was undertaken. An area containing a 150
domes in the center of the aeration basin was selected as
the test area. A detailed drawing of the test segment is
presented as Figure 10. Five methods of cleaning were
selected for testing. These included the following: Low
pressure hosing, High pressure hosing, Steam cleaning, Acid
washing under the Milwaukee Method (7), Kilning. Because
sonic cleaning and soaking in bleach had been evaluated in a
previous study on similar domes and did' not prove
particularly effective for the effort involved, they were
not chosen for further study during this investigation.
Dome Collection ;
i
i
After the tank was dewatered, the domes to be tested in a
contaminated state were carefully •removed and placed in
plastic Ziplock bags. The domes were stored- in an ice chest
for transportation and in a refrigerator until they were
placed in the test tank. Even with careful handling, much
33
-------
o
V
1.
(N
H:
i—
I
.-'
s1
1
1
-
•
•
>
'!
— :
•
n
"8
<
sz
1*
pH
t
ii!
c
! 1
;i
jj
U_l
1:
C
3 ;
n
06
1
1
1
1
1
i
i
T
ifi
1
i
C 1
I?
II
p
1
1
1
1 ,
l<
i;
i
i-,
i
•&1
pg
MB
MR
'
'
O>
lO
(N
CS
O
E
"o
c
.2
o
o>
00
o
UJ
00
00
Z)
O
LU
S
O
Q
LU
OH
z>
o
34
-------
of the growth on the dome surface broke away from the
immediate surface of the domes and was lost before testing.
The domes to be cleaned were hosed off in the tank and then
i
were cleaned by the various methods. :
Low Pressure Hosing
All domes being cleaned received low pressure hosing as the
initial step in their cleaning. This was.necessary to
remove the loose slimes and other materials from the tops of
the domes. Low pressuring hosing consisted of washing the
domes from the floor of the tank using the standard water
pressure, approximately 40 pounds per square inch (psi),
available in the plant's main water system. A standard
I
hosing nozzle was used, and each dome as washed for 10 to
30 seconds depending on the time required'to clean the
surface. Air continued to flow through the domes during the
hosing process. No attempt was made to maintain any
particular airflow rate through the domes during cleaning.
35
-------
High Pressure Hosing - ;
A section of domes that were washed at high pressure were
cleaned in place in the tank. A water supply system of
approximately 85 psi is available at the plant site. This
water system was extended into the tank, and an individual
washed the surface of each dome at a distance of
approximately 1 foot for 1 minute per dome. ; Attempts to
come closer to the dome resulted in splashing and were not
continued. Domes were washed until they appeared to be
clean. ' •
Milwaukee Method
For the Milwaukee Method of cleaning, the domes were washed
with high pressure hosing similar to the procedure described
above. Thereafter, the air was turned off and each dome was
saturated with a solution of muriatic acid (14% hydrochloric
acid solution) and was allowed to set for 30 minutes. No
j
initial reaction appeared to take place, although some small
i
amount of frothing did occur on individual domes. After 30
minutes, no reaction was obvious, and the application of a
36
-------
small amount of additional acid did not appear:to cause any
additional reaction. The domes were then hosed, using high
pressure water, for approximately one minute- The air was
i
then turned back on and hosing of all dome;surfaces was
carried out for another 10 to 20 seconds. i .
Steam Cleaning
No steam cleaning apparatus could be gotten into the test
bays at Rowlette. The domes were, therefore, removed and
taken to SMU where a small steam generator is'available in
the maintenance area. The units were mounted on a temporary
header and exposed to steam for approximately 30 to 40
seconds at a steam pressure of approximately 150 psi.
While removing the domes for treatment, it was noted that
some domes had a significant amount of material on the
inside. The material appeared to be dried activated sludge
particles that had somehow entered the chamber under the
dome. :
i
Brushing appeared to remove most of this material easily.
This material caused concern. Interviews! with plant
37
-------
personnel indicated that line damage might ' have caused
similar problems with all the domes in the tank.' Because of
this possibility, all domes in the test sequence were,
thereafter, examined, and all loose material from the
interior side of the domes was removed by simple brushing
techniques.
Kilning :
The domes to be kilned were removed from the test tank and
brought to the Art Department at SMU where a large kiln is
available. Domes were placed in the kiln, and the
temperature was raised gradually to a 980°C over
approximately 12 hours. The temperature was then held at
980°C for 4 hours. Then the kiln was allowed to cool, which
took approximately 12 hours more. The domes were then
removed, brushed free of any obvious accumulated ash or
other materials, and returned to the test tank, i
38
-------
MICROSCOPIC PHOTOGRAPHS OF THE DOMES
The entire cleaning operation took about 4 days;and required
two individuals working most of that time. Six domes
cleaned by each, method were retained for initial study and
investigation in the test tank. The tank was placed back in
service on September 19, 1985. \
\
Having some detailed physical method of evaluating the
impact of cleaning on the dome materials.- After some
investigation, electron micrographs of the1 domes were
s
determined to provide the best insight into examining the
surface and the penetration of particles into the dome
structure. To achieve effective photography of the dome
interior, the domes were held approximately 1,foot above a
concrete floor and dropped on their bottom side. This
resulted in cracking of the dome without any significant
introduction of foreign particles into the dome structure.
One dome " for each condition was taken toithe electron
microscope located in the SMU Anthropology: and Geology
!
Department, and each dome was photographed. These
photographs are shown in Figure 11. Most noteworthy in the
photographs is that in both the acid wash and the kilning
39
-------
operations a significant number of particles appear to be
remain in the void spaces within the domes. Considerable
!
further investigation of pictures and other information on
the domes was conducted. i
All that can be said with certainty is that some minor
penetration of particles into the domes doesl occur under
conditions of vigorous cleaning. Under uncleaned conditions
it is unusual for particles to be more than 2 to' 3 grains of
aggregate below the surface. After kilning, ash was found
5 to 7 grains deep in the stones. This method, of analysis
will not likely provide any quantitative method of
estimating the efficiency of dome cleaning. !
40
-------
FIGURE lla: ELECTRON MICROGRAPH OF A DIRTY DOME
41
-------
FIGURE lib: ELECTRON MICROGRAPH OF A DOME
AFTER LOW PRESSURE HOSING
42
-------
FIGURE lie: ELECTRON MICROGRAPH OF A DOME
AFTER ACID WASHING
43
-------
FIGURE lid: ELECTRON MICROGRAPH OF A DOME AFTER KILNING
44
-------
STUDY RESULTS OF CLEANING INVESTIGATIONS
Table 4 presents the results of the oxygen transfer studies
conducted on new domes, the dirty domes removed from the
North Texas aeration chambers, and the domes after cleaning
by the five selected test methods. The data are presented
for airflows of 1 and 2 cfm per dome. The test water for
all tests .contained detergent in a sufficient concentration
to lower the surface tension to approximately 65
dynes/cm2.
i
The test shows that the dirty domes are transferring
approximately 75 percent of the oxygen of the new units.
Cleaning appears to restore the domes to between 80 and 90
percent of their original performance level. There does not
appear to be any significant difference between the cleaning
efficiency that is achieved by the various methods.
The oxygen transfer efficiency of clean domes was improved
to about 85 percent of the value of new domes', which is a
10 percent improvement over the 75 percent transfer observed
for dirty domes. Tables 5 and 6 show the results of similar
testing after 9 months and 21 months of exposure of the
45
-------
TABLE 4
"O.SOTE" OF DOMES1 BEFORE AND AFTER INTERNAL CLEANING AS
TESTED IN DETERGENT SOLUTION* ;
DOME AIR FLOW
CONDITION CFM/DOME
NEW
DIRTY
KILNED
LOW PRESSURE
HOSED
HIGH PRESSURE
HOSED
STEAM
CLEANED
ACID WASHED
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
aSOTE
12.75
9.78
9.35
7.65
11.5
8.5
11.9
7.9
11.0
7.8
10.2
8.6
10.2
7.9
OTE/FT '
1.5 ;
1.15 |
1-1 !
0.9
1.35. <
. 1.0
1.4 j.
0.93 :
1.3 ;
0.92
1.2 ;
1.03 !
1.2 1
0.93 :
% OF
NEW DOMES
73
78
90
87
93
82
87
81
80
90
80
80
*A11 test conducted in a solution with a surfiace tension =
65 dynes/cm2. (measured average 3.5 mg/1 SAX). All runs
were conducted in duplicate. '
1 Domes were on service for 3 years before testing.
46
-------
TABLE 5
aSOTE OF CLEANED DOMES AFTER 9 MONTHS OF; USE
DOME AIR FLOW SURFACE aSOTE
CLEANING CFM TENSION
METHOD DYNE /CM2
LOW PRESSURE 1 65 8.05
HOSING 2 64 6.5
HIGH PRESSURE 1 64 7.15
HOSING 2 65 6.23
ACID 1 63 9.65
WASHED 2 65 6.7
aSOTE/FT OF
WATER DEPTH
0.95
0.77
! .84
.73
1.13
0.79
KILNEDJ
DATA ON KILNED DOMES WERE NOT INCLUDED
DUE TO ANALYTICAL PROBLEMS.
47
-------
DOMES
TABLE 6
aSOTE OF DOMES AFTER 21 MONTHS
JUNE 1987
AIR FLOW SURFACE TENSION aSOTE: aSOTE/FT OF
DYNE/CM2 WATER DEPTH
DIRTY
LOW PRESSURE
HOSING
HIGH PRESSURE
HOSING
ACID
WASHED
KILNED
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
63
63
62
62
63
63
62
62
64
64
8.
7.
7.
8.
7.
8.
8.
8.
6.
6.
1
2
7
6
4
9
5
0
5
q
0
0.96
0.905
0.9
0.99
0.93
.1.00
0.94
1.0
0.71
' 0.71
48
-------
domes to the tank conditions. The data for steam cleaned
" I
domes are not available in the latter periods. During the
test program the header that contained the steam cleaned
domes was broken loose, and this unit had to!be replaced.
In view of the data that had been collected to date, a
decision was made not to reinstall the steam cleaned units.
Table 7 summarizes the data for the 1 cfm dome testing
i
i
study. The data suggest that the newly cleaned domes
returned to their former condition within the 9 >months after
cleaning. Thereafter, little deterioration in the systems
were noted. Visual inspection of the domes tended to
support this conclusion. No difference was observed in the
pressure required to pass air through the individual domes
at the desired flow rate.
Concern must be expressed regarding the efficacy of the test
method. When the dirty domes were placed in:the aeration
tank for'study, the activity of handling the domes resulted
in a disturbance of the films. Under aeration in the
detergent solution, particles of growth broke off the domes
in a random manner. This had the impact of'changing the
appearance of the. aeration pattern in the tank somewhat
49
-------
TABLE 7
SUMMARY OF DIFFUSER CLEANING DATA
DETERGENT TESTING
DOME
TYPE
NEW
DIRTY
LOW PRESSURE
HOSING
HIGH PRESSURE
HOSING
ACID WASHED
KILNED
STEAM CLEANED
OCT 851
12.75
9.35
11.9
11.0
10.2
11.5
10.2
aSOTE @ 1.0 CFM/DOME
JULY 86 . ' JUNE" 87
9 MONTHS I 21 MONTHS
i 11.9
8.0
8.05 8.0
7.15 ; 8.2 .
9.65 • 8.2
1 6.0
__
-"• Newly cleaned domes; these domes were 3 years old at time
of cleaning.
50
-------
during the studies.
Close observation of the domes indicates that even with
careful handling, some degree of anaerobiosis develops under
thick films. This anaerobic process most likely results in
a condition where the bond between the slimes and the stone
is broken. When these units are placed in :the aeration
tank, a cleaning process begins. The degree to which
sluffing and cleaning occurs appears to be random. This
occurs in spite of vigorous efforts to maintain the units in
proper condition. This changing condition on the surface of
the domes leads to the conclusion that this test method is
not particularly suitable for evaluating dirty domes.
51
-------
OFF-GAS TESTING
A field study was conducted to evaluate the oxygen transfer
in the tank under study. The three tanks'. in the study
portion of plant were evaluated using "off-gas1.1 techniques
to determine the system efficiency. The data ^re presented :
in detail in Table 8. In the test section, the air
.»
flow/dome is 1.19 cfm/unit and the oxygen transfer
efficiency, as aSOTE, averages 6.75 percent or 0.45 percent ,
per foot (ft) of depth based on 15 feet of depth. This is
substantially below the value of approximately 1.0
percent/ft observed in the SMU shop testing of the domes. .
i
l
The lower transfer efficiency is most likely the result of a
lower a in the waste and the generally dirtier condition of
most of the domes in the tank. The fact that;the aSOTE is
much lower in the actual tank, compared to that in the shop
tests, raised concerns regarding the efficacy: of using the
."•''. i
shop test data except in a comparative mode. \
52
-------
ua •• eu x:
s = s s
*s — cu m
UJ v tjit (jj
eu ^o o
— — cu
ru c-» f— — eu '-a
-a- vs 3- -a- u? -j-
en —• en cr* — c» — eu ITJ
•i c» ru -j- OK eu -o — o- o» -S
eu —• en -o cr- -a u? eu eu aa m CD o- o- «j —
! = "
= = £S
i -o -o ~f eu rtj en eu t~- m *n en T- en -o -«r ,»« cu en ~f -o vi en eu ca en —• ca en 9
< •«*• eu -o cr» -r e~- O -a cr- eu -o eu CB -a cu -a eu ea en ^ vxi -a en ^9 c-» _»• — 03 -o
• -O •» « « -O •» <3^-3>.9<».^9 & -»^-9e3^ O -Z«Z»-J«ocd
•=. » -9 «» O
ITS co e— en o -»• ui
s s s s s s s
o- -a —• —• cu en
en co u-3 o — o- i
g s s.
t— r- -a -J3
s a '
eucucucurucunjcu
i ira m •
cu cu eu cu cu eu eu
9 -^ o <^ O <3
cu cr> oa ~y -^ —• eo
-a- u? -o -o 5~- o> oa
2 S
S Sg
cu cu eu cu cu
S CU
S
03 S S
«» -3> i^
O «» «>
cucucucurucucucu
•^ «^ *=»•=> oa o* i
u-3 ui ^3 r- <=> -o
s s s s 5 2
en -a c» ^- cu i
-a- en -a « -3* cu cu
-r —• 'CB ^r ca c^ -r-
EH W
CJ
w
EH
EH
W
§
u.=»— -o
UJ 3= -^ <=-
<=. O -3- 0 —
OOOO9
I CU —— -•* •
:SS§:
C-. -a- m -a —
s s: s 3: s
~a r- ira en -a cy- «u
or- o- <3 >s> o ^ ^
o- cr- o o o ca <»
• o- eS o- o- a* o»
§ §gggS
! -I -Z -I o o •=>
CD ^ \n cu en cu era en P*- c^- r«- -o en ^ irt -*• — o- "3- co cu o- o b*3 —>
—- *au-a-4-4*acacu — u? u-3 «A ^s ^3 CD en eu c^^-eneueucB^ct» ca
tr> •»• m •— va en cu cu cu cu —• cu —• as
ru tu eu cu cu eu tra
eu CU cu eu cu cu -r
Scu cu eu cu cu iS
eu cu cu cu CU -^
co co -«• en a-* en trt
oC oC cd — o^
m tn i-a- -a- — i
' trt en ^ e— r^-
f »—
^ eg nj cu cu cu nj eu cu cu
cu cu cu cu cu cu eu
eu cu cu eu cu eu eu
cu cu cu eu cu cu eu cu
-o « irt
en en en c*^ cu cu m cu
<=
s _ „ ,,,* i
C r r ~ ~
53
-------
CLEAN WATER STUDIES
At the end of the plant scale studies, a set .of domes was
removed from the tank, and the domes were recleaned by low
pressure hosing, high pressure hosing, acid wash and
kilning. These domes were tested in clean water and
compared with unclean domes from the respective sections.
The 'results of these studies are presented in Table 9.
j
Old domes that had not been cleaned at any time during the
program were also tested as part of this evaluation. The
I
results of the study were erratic.
In general, the domes after cleaning returned to within 10
percent of the original test level. Variability in the data
makes it difficult to draw detailed conclusions concerning
any of the individual cleaning methods. As a general
observation, all the methods worked well.
54
-------
TABLE 9
RECLEANING AFTER 21 MONTHS SERVICE l
i
CLEAN WATER I
DOME
TYPE
OLD DOMES
NOT CLEANED
LOW PRESSURE
HOSING
HIGH PRESSURE
HOSING
ACID
WASHED
KILNED
1.
1.
2.
1.
2.
1.
2.
1.
2.
0
0
0
0
0
0
,0
0
0
BEFORE
RECLEANING
DETERGENT CLEAN
ctSOTE WATER
aSOTE
cfm/d 8.0 9.25
cfm/d 8.0 11.7
cfm/d
cfm/d 8.2 8.5
cfm/d
cfm/d 8.2 8.4
cfm/d
cfm/d 6.0 11.5
cfm/d
: AFTER
' RECLEANING
: CLEAN
; WATER
aSOTE
1
I
: 10.65
' 12.65
12.65
1
12.4
: 11.5
10.6
9.3
: 10.3
NEW DOMES 12.6% @ 1 cfm/dome
55
-------
OBSERVATIONS ON DIFFUSER CLEANING
"' -U11"-1" - .'"•—• in i i j
Overall observations suggest that, in the test plant,
fouling of the diffusers was not a major problem;. • The major
diffuser problems appear to have been caused by failures in
the system that allowed broad-based contamination. The
growths that developed on the exterior of the domes did not
appear to dramatically reduce the OTE over the study period.
Over a substantial period fouling does develop on the domes
and periodic cleaning is recommended. Field investigations
suggest that conscientious low pressure hosing is an
acceptable routine technique for diffuser cleaning. High
pressure hosing could be helpful periodically to improve
cleaning of the systems. In the study system it is probably
not justified each time the tank is taken down, i
I
Although the data are not conclusive, we believe that the
use of acid washing system, such as the Milwaukee Method,
should be used ,at the North Texas plant, at intervals of
possibly 3 to 5 years. A note of caution: this technique
should be used with care to avoid possible harm to
employees. :
56
-------
More elaborate cleaning techniques such as kilning require
removing the domes from the mounts and should be avoided.
The problems associated with removing domes, handling them,
and replacing them outweigh the benefits that:appear to be
associated with these types of techniques. The effort
involved in such a program does not appear to be justified,
based on the observed results at North Texas. :
57
-------
GENERAL DISCUSSION OF RESULTS
The proposed test method, based on removing specific devices
i
from an aeration tank and performing shop tests,' seems to be
limited in value. Even if the method can be carried out in
i
a vigorous, controlled atmosphere, the changes in conditions
and the variations introduced by handling the devices appear
to significantly influence results. •
There is a significant variation in test data, which is
magnified when a relatively shallow tank is employed.
Variations appear to occur among individual aeration
devices, which suggests that a large number of samples would
be required to develop a statistically ( significant
evaluation of the transfer capacity. This does not appear
to be very practical in a small test tank. :
Using detergent in the test solution allows [the fluid to
more closely represent the actual field conditions under
which transfer occurs; however, including detergent appears
to cause some degree of cleaning of the ;domes. This
inter-reaction changes the concentration of surface active
materials and causes some sluffing of slimes from the dirty
58
-------
units. The importance of this phenomenon is hard to
quantify, but observation indicates that it is significant.
The clean water testing shows the efficiency of domes.
Although the methods very imprecise it does not overwhelm
i
the data interpretation. Overall, using a test tank of
small size, and evaluating a limited number of domes, does
not appear to provide an efficient, effective means of
estimating the need for diffuser cleaning, i On the other
hand, using such a tank appears to be appropriate in
comparing the relative degree of cleaning ; that can be
achieved by various methods. Such a comparison is helpful
i
in determining the appropriate cleaning techniques that
should be used for diffuser maintenance. !
59
-------
EVALUATION OF DOME AIR FLOW CHARACTERISTICS
At the end of the study, five domes from each cleaning group
were removed from the Rowlette Creek treatment system and
were sent to the University of Wisconsin for evaluation.
The detailed data collected from this survey arq included as
!
Appendix 2. Evaluation of these data suggests that the
domes vary widely in all measurable characteristics. No
discernible pattern or correlation between -the physical
variables and the transfer characteristics was apparent.
Further investigation of the meaning of the 'data and the
relationships between diffuser fouling and oxygen transfer
will be required before this information can be meaningfully
related to plant performance.
Figure 12 shows the Dynamic Wet Pressure (DW,P) variation
from clean units to the average of each fouled unit. Figure
13 shows the variation of DWP from unit to iunit for the
cleaned domes.
60
-------
25-
20-
1 j-_j
15-
Q.
Q
10-
5 -
0
Dynamic Wet Pressure j •
Before Cleaning ;
-f Units Never Cleaned
A High- Pressure Housed
• Kilned ;
O Cleaned (New Domes)
0
O
A
O
1 2
CFM
FIGURE 12: DYNAMIC WET PRESSURE OF CONTAMINATED DOMES
61
-------
CD
25-
20-
.. r- I
15-
10-
5 -
Dynamic Wet Pressure
After Cleaning
D Never Previously Cleaned
+ High Pressure Housed
A Low Pressure Housed
• Kilned i
CFM
FIGURE 13: DYNAMIC WET PRESSURE OF CLEANED DOMES
62
-------
SPECIAL STUDIES - NEW DOMES
A special study was conducted to examine the variations that
would occur by virtue of the cleaning techniques on new
domes. To accomplish this study, 24 new domes were selected
from a lot available at Fort Worth Village Creek Treatment
Plant. These domes were divided into four groups. Each
group of domes were installed in the test itank and was
tested at 1 and 2 cfm. Each test was carried out in
triplicate. In the first set of tests, the domes were
evaluated as received. After that, the tank water level was
lowered, and the domes were washed according to the low
pressure hosing procedure. The run was then repeated. The
domes were also high pressure washed, acid washed and
kilned, and then retested. \
Only two sets of kilned domes were studied because the Art
Department had to shut down the kiln during the latter
portions of the study. ;
The detailed analytical data from this investigation are
presented in Appendix 3. Table 10 summarizes the run for
the KLa and the standard oxygen transfer efficiency (SOTE)
63
-------
§§
-^ TO c-» -*• r-* -^
CU CO O- —• -*• CO
- - - - •
u *m
O - O
•&• -*• -a- tn cu m
c -a- — ru cu in -a*
-a .-. cu cu o o cu cu
QJ Ul « 0 <=> • • O «
co-oruirs-ocucn—•nj-a-
in oa
en CP>
• •«•
cu *
co -a- -a« en en ^
Sen ]
-
CO
w
, 1 rj
<3 o
H Q
c: co — i
• • ^ o • ^ _ , „ „ ~ *~j ^-^ ^-^
*tj ^S o^** otS'O-^oo^'rJ
CU iuT
g
-O *Q
3 S
S 5
cu co -o -o —i .* <»•
- O CO U-3 Z3 ^ ^
— ^ en « • *n
rO CU
m cu u*i -a-
^ c»a —^ cu
• o- co <=>
co in -^ ^o o' cu
UT ^o m r-* cu cu
— <=• — co ^ en
O <=>
CO —
co en
CO O-
— cu
o- ITS CD CD e*~ cu in
*Tk «-^ <=>,=>.=>•=>•=>
c: ^- en to,
• -— C*» CO O
-
c: —^
o- ca in
,2
ca -*-*
CUCOCU-J3— e-eu
^o cu 1?
S ^ S o "^ <=Sl-
" • - - <=> «•
— ru —
J= 5 ^ ^£ **
64
-------
for each study condition. The mean and standard deviations
of each set of studies are presented in that table. The
data from the initial run on clean domes are 'inconsistent
with all other observations. The reason for this change is
believed to be related to system problems in airflow. Field
observations, as well as the values obtained, 'suggest.that
these values are not representative. The average for the
new dome set is calculated by deleting ^the initial
triplicate set of data.
Table 11 compares the new domes with the various cleaning
methods. The data present a curious picture. The new domes
in this test showed an oxygen transfer efficiency less than
observed in all previous testing. The new domes would be
expected to transfer approximately 12.8 percent!at 1 cfm and
12.2 percent at 2 cfm. The first run with'new domes is
suspect because of erratic air flow; all other runs should
be considered valid. The lower than expected values for
clean domes may have resulted from a film or coating on the
f
dome surface. :
The values obtained for the low pressure, high pressure, and
acid testing are statistically significantly higher than the
-------
TABLE 11 I
NEW DOME STUDY - TRIPLICATE RUNS IN CLEAN WATER
SCFM/UNIT
NEW DOMES
LOW PRESSURE
HOSING
HIGH PRESSURE
HOSING
ACID
WASHED
KILNED
1
2
1
2
1
2
1
2
1
2
MEAN
10.0
11.3
12.8
11.1
13.5
13.5
13.6
12.3
10.96
9.86
SOTE STANDARD DEVIATION
2.
2.
1.
1.
0.
1.
1.
1.
0.
0.
1
8
4 :
9 ;
2
9 ;
6 :
8
5
55 . :
71
66
-------
values obtained for new domes. They are statistically equal
to the values that would have been expected for the new
domes based on previous testing. The kilned values are
lower than would have been expected. !
Examination of the new domes showed no physica'l reason why
the lower performance was observed. Examination of the data
and study techniques did not indicate any problem. The fact
that these runs were run at the end of the study is also an
important consideration. Techniques had been :refined, the
r
personnel were well trained, and the test conditions were
ideal. The only possible problem was that the 'temperatures
were somewhat lower than desirable for testing, but the
clean water runs were all above 9°C. This problem is not
i
believed to have interfered significantly with test data
i
gathering. We do not know the reason for :this seeming
anomaly in the information. [
67
-------
COST OF DOME CLEANING ,
To collect significant information on the cost of dome
cleaning, the data available from several treatment plants
were reviewed. Examination of the information showed a
wide range of costs associated with this process. Further
examination of the data revealed that each plant tabulates
costs in a unique fashion; grouping together!both cost of
!
dome cleaning and a wide variety of related activities.
Activities such as draining tanks, cleaning tanks,
inspecting and repairing dome systems, and carrying out
other required maintenance are usually reported as an
integral part of the total cost of a dome cleaning
operation. This procedure makes it very difficult to
provide any specific information on the individual unit
operations associated with this process. ;
!
The cost of dome cleaning is also significantly influenced
by the work rules and procedures employed at, a particular
plant. The impact of work rules on the total time required
for a project can be exemplified in the following example,
Case 1.
68
-------
CASE 1 - TYPICAL WORK DAY FOR DOME CLEANING AT ONE PLANT
7:00 a.m. Arrive, dress for days work, review
work assignment with the
supervisor, walk to tank.
7:35 a.m. Enter tank and begin work.
9:15 a.m. Out of tank and coffee |break.
9:40 a.m. Reenter tank and work.
11:00 a.m. Leave tank, walk to lunch room,
wash-up, half-hour for lunch,
redress, walk back to tank.
12:00 p.m. Reenter tank.
!
1:20 p.m. Leave tank, 10-minute break.
1:30 p.m. Reenter tank. •
2:30 p.m. Leave tank, walk to locker room,
wash-up, fill out sheets on daily
activities.
3:00 p.m. End of shift. ;
69
-------
The actual time spent in the tank washing domes during the
8-hour day in Case 1 came to approximately 5.3' hours. In a
second case, a contractor was able to have 7.2 hours of
I
actual work during a day. Although these cases are well
within standard operating efficiency expectations, they
represent a significantly different effort in the given
period, which impacts the cost information obtained.
i
Another major factor that must be considered when evaluating
i
time estimates is the availability of mechanical equipment
to assist in the cleaning process. For example, in one
case, the plant is able to lower a small front-end loader
into the aeration tank to assist with removing grit and
other debris; however, in another case, the removal of
debris is manual with shovels and buckets. The difference
in man-hours associated with these two operations is
obviously very significant.
70
-------
FACTORS THAT INFLUENCE TANK CLEANING COSTS
i
A wide variety of factors influence the cost of dome
cleaning. These factors relate to the design and' operating
load at the treatment system. :
To carry out any cleaning operation, the tank must be first
dewatered. The relative ease of dewatering depends on the
plant's design. In many instances, the tanks can be drained
by gravity with a minimum of inconvenience. In other cases,
complicated rerouting of the sewage and or return sludge
flow is required to dewater a.tank. In other instances, the
tanks must actually be pumped to achieve effective drainage.
The time required for dewatering a tank can vary from 1 to
2 man-hours up to 8 to 10 man-hours depending on the
considerations of the individual plant. ;
After the tank is dewatered, the condition of ithe floor is
the next factor of major concern. In most plants, grit and
. I
i
other heavy materials will precipitate in the zone, below the
diffusers. If the diffusers are placed more than 1 foot
above the floor, significant deposits of material will
normally be observed at least in the front end of the
71
-------
system. The . relative ease or difficulty of removing these
materials will depend on the forethought of the engineering
design. If the system has been designed to allow easy
operation and cleaning, this material can be hosed down or
bucketed out. If the units' have been placed !too close to
each other and if the piping is complicated, 'removing this
material can be a long and tedious task. In some plants
mechanical equipment must be brought in to effectively
handle the removal of grit and other solids. iOther plants
have used vacuum type pumps to facilitate cleaning.
Before a tank can be returned to service,:the aeration
I
system should be inspected and repaired as necessary. The
age and condition of the aeration system will obviously
impact this particular activity in a significant manner.
72
-------
THE UNIT OPERATIONS OF TANK CLEANING
i
To come up with some effective and reasonable cost estimates
for cleaning and operations, a study was performed on the
unit operations associated with the process. This was done
i
by observing, in the field and in a laboratory, the
work-time required for each process associated with dome
cleaning. The data presented in this ; report are
generalizations and not intended to be precise. They are
presented to allow the engineer to make a reasonable
estimate of the costs associated with various cleaning
operations. :
UNIT OPERATION
(1) Tank Dewatering
DESCRIPTION
Depends on Design—Withdrawal or
pumping will influence time.
Normal needs, 4 to 6 man-hours.
(2) Tank Cleaning
Depends on location of the
diff users, the height off the
floor, and whether or hot the plant
has
primary
treatmnt.
Twenty
man-hours per 1,000 diff users.
73
-------
(3) Cleaning—Low
Pressure Hosing
The initial unit operation in any
cleaning system must be low
pressure hosing to remove the loose
growths from the system. Ten
i
man-hours for setup and ten
man-hours per 1,000 domes,
including pipes and supports.
(4) Inspection and
Repair
Any obvious system breaks will be
i., i i
observed and noted while a small
i.
amount of water still covers the
diffusers. Inspection :of the
system and repair of it depends on
the system's age. Ten man-hours
per 1,000 domes.
(5) High Pressure
Hosing
(6) Steam Cleaning
Can be accomplished after low
i
pressure hosing. Twenty-five
man-hours per 1,000 domes.
Should be accomplished '• after low
74
-------
pressure hosing. Assume that
mechanical equipment ; is available
to lower equipment • into tanks;
i
forty man-hours per 1,000 domes.
(7) Acid Washing
(8) Dome Removal and
Replacement
This process requires approximately
60 man-hours per 1000 domes.
Removing domes from the system
for any cleaning and operation and
subsequently replacing them will
require significant time and some
degree of equipment : replacement.
Eighty-five man-hours per 1,000
domes, plus replacement of 5
percent of equipment. ;
(9) Kilning
If kilning is to take place,
estimate $5 per dome above the cost
of removal and replacement.
75
-------
Using these data to estimate the cost of ' cleaning is
demonstrated in the following example:
EXAMPLE 1: '. •
TYPICAL CASE--LOW PRESSURE HOSING
1,000 DOMES/5,000 DOMES
(1) 1,000 domes would treat approximately '1.5 million
gallons of sewage per day. ;
1,000 DOMES >: 5,000 DOMES
(1) Dewater Tank 4 mh .; 4 mh
\
(2) Clean Tank 20 mh ' 100 mh
(3) Low Pressure Hosing • 20 mh 60 mh
(4) Inspect and Repair 10 mh , 50 mh
i
(5) Refill Tank 4 mh ; 4 mh
TOTAL 58 mh : 218 mh
i
For 1,000 Domes: :
58 man-hours x $7.50/hr pay x 2.1 (indirect costs including
benefits) = $913.50 x efficiency factor of 1.3 = $1,187.55,
or approximately $1.19 per dome. ;
76
-------
For 5,000 Domes: i
The cost per dome is $0.89 per dome.
The 2.1 (indirect cost.including benefits factor) used in
the above example accounts for the in-direct manpower costs
including supervision, administrative, payroll, and benefit
costs.
The efficiency factor of 1.3 is used to relate the time
spent cleaning domes to the total hours worked by the
individual. This considers such time as preparation time,
breaks, and washup. !
i
If two men are employed the total estimated time for
!
cleaning a tank of 1,000 domes is approximately 1 week.
Experience suggests that such operations usually occupy the
full time allotted for the task. Adding a thifd man to the
staff, for example, is unlikely to result in the tank's
being cleaned any faster. It will most likely result in a
i
more thorough job of cleaning and inspection. ;
77
-------
EXAMPLE 2: _ ;
[
If the 1,000 domes from Example 1 are to be acid washed, in
addition to previous cleaning, an additional sixty man-hours
would be required. This would result in a total1 cost of one
hundred-eighteen man-hours or $2.40 per dome.
Although these formulations are not considered to be precise
or scientific, they do provide a reasonable estimate of the
i
costs associated with cleaning. These numbers have been
checked against the actual data available from field studies
and correlate realistically. '
i
GENERAL OBSERVATIONS , ;
...... . r
The following observations have been made concerning the
cleaning process and may be helpful in actual plant
operations:
(1) Domes should be regularly cleaned; once a year appears
to be desirable. The operations are likely to be more
i
smooth if they are planned in advance rather than being
undertaken when the occasion presents itself.
78
-------
(2) When dewatering the tank, leave the air on until the
water level reaches the domes. This policy may require
adjusting other tank conditions to maintain airflow
throughout the system. If possible, stop the
dewatering when the water is about 1 to,2 feet above
the domes. At this point, inspect the system from the
tank edge to identify any discontinuity or breaks in
the aeration system^ Carefully map the location of any
problems so that they can be corrected later when the
tank is completely dewatered. '
(3) If possible, do the low pressure washing from within
the tank and close to the domes. Water,;particularly
from fire hoses, cascading on domes from the top of the
tank, has an adverse affect on the units and in some
cases causes cracking in the housings.
(4) Do not loosen or move domes unless it is absolutely
necessary. Reseating domes in a proper manner is a
difficult and time-consuming operation. ;
(5) Acid washing of the domes every several years is
probably desirable. There is, however;, no absolute
79
-------
evidence to reenforce this belief under the operating
conditions observed at the North Texas plant.
80
-------
OVERALL DISCUSSION OF RESULTS
The studies conducted on the cleaning of domes suggest that
the costs vary significantly from plant to plant. The cost
differences are usually associated with the ancillary
operations attendant to dome cleaning rather than the
cleaning itself. These variations are caused bbth by design
features and work rules. In case, the cost of' cleaning the
domes are relatively modest. For domes operated in the
i
i
range of 1 cfm/unit, the cost of providing air for a year is
estimated at $18.50 per unit. If the efficiency can be
improved by 10 percent each year by cleaning, the costs
savings would be about $1.85, which is roughly equivalent to
r
the cleaning costs of the unit. Improving the efficiency by
!
20 percent would certainly be a good investment!.
Keeping an aeration system in top condition will lead to
better overall operation of the treatment plant and better
effluent Duality. This consideration alone is [sufficient to
justify the investment in maintenance and upkeep of aeration
equipment.
81
-------
REFERENCES;
(1) BARNHART, E.L. "An Overview of Oxygen Transfer
Systems" Proc. Workshop on Aeration System Designs,
Operation, Testing and Control, U.S.EPA, EPA
600/9-85-005, Cincinnati, Ohio (January, 19|85).
i
I
(2) BOYLE, W.C. and REDMON, D.T. "Biological Fouling of
I
Fine Bubble Diff users: State-of-Art" ASCE Journal of
the Environmental Engineering 109,5,991-1005 (1983).
I
(3) "A Standard for the Measurement of Oxygen Transfer in
Clean Water" American Society of Civil Engineers,
Oxygen Transfer Standards Committee, New Yo|rk, New York
(1984). '
I
(4) BARNHART, E.L. "Transfer of Oxygen [ in Aqueous
Solutions" ASCE Journal of the Sanitary, Engineering
Division, 95, 3, 645-661 (1969). :
(5) DOWLING, A.L. and BOON, A.G. "Oxygen Transfer in the
Activated Sludge Process", In Advances In Biological
i
Waste Treatment, Ed W.W. Eckenfelder, 'Jr. and B.J.
i
McCabe Pergammon Press, New York, New York;(1963).
82
-------
(6) Standard Methods for the Examination of. Water and
Wastewater, 16th Edition (1985).
(7) Quality Assurance Program Plan ASCE/.EPA "Design
Information On Fine Bubble Diffused Aeration":
C. Robert Baillod, Section A12.0, 1985. '
83
-------
TEST T-A1STK.
84
-------
SMU TEST TANK STUDIES
Measurement of Oxygen Transfer Rate in Clean Water and
Detergent Tests. This method of procedure was Iemployed for
all testing.
1. Scope ,
i
This method covers the measurement of the oxygen transfer
rate, OTR, as a mass of oxygen per unit time is dissolved in
a- volume of water by an oxygen transfer system operating at
* ;
a given gas flow rate. It is intended to measure the rate
of oxygen transfer from diffused gas oxygenation devices to
relatively large volumes of water. Although the method is
•intended primarily for clean water, it applies to water
containing surface active agents and low concentration of
salts. \
The study results are expressed as the Standardized Oxygen
Transfer Rate, (SOTR), a hypothetical mass of oxygen
transferred per unit time at zero dissolved oxygen
concentration, a water temperature of 20°C, and a
85
-------
barometric pressure of 1.00 atm, under specified gas rate
i
and power conditions. The results can be used to estimate
oxygen transfer rates at process conditions.
2. Summary of Method ; .
i
The Test method is based on removing dissolved oxygen (DO)
from the water volume by sodium sulfite followed by
reoxygenation to near the saturation level. The DO
inventory of the water volume is monitored during the
reaeration period by measuring DO concentration's at several
determination points selected so that each point senses an
equal tank volume. These DO concentrations may ;be sensed in
situ using membrane probes. The method specifies a minimum
number, distribution, and range of DO measurements at each
determination point. '
The data obtained at each determination point are then
analyzed by a simplified mass transfer model to 'estimate the
apparent volumetric mass transfer coefficient, KLa, and the
saturation concentration, C*. The basic model is defined
oo
as follows: ;
C = C* - (C* - Cn) exp (-Kra t) 1
86
-------
where:
C = DO concentration, m L~3
C* = DO concentration attained as time approached
00 infinity, m L~3 '
C0 = DO concentration at time zero, m L
= Apparent volumetric mass transfer coefficient
t"-"-, defined so that \
rate of mass transfer per unit volume
KLa = 1
C - C
Nonlinear regression is employed to fit Equation 1 to the DO
profile measured at each determination point during
reoxygenation. In this way, estimates of Kra and C as are
; 00
obtained at each determination point. These estimates are
adjusted to standard conditions, and the standardized oxygen
transfer rate (mass of oxygen dissolved per unit time at an
hypothetical- concentration of zero DO) is obtained as the
product of the average adjusted KLa value, the average
adjusted C* value, and the tank volume.
87
-------
3. Significance and Limitations •
Oxygen transfer rate measurements are useful for comparing
the performance and energy efficiency of t oxygenation
devices operating in clean water. Performance of these
devices in process water may significantly differ from the
performance in clean water, and the amount of difference
[
will depend on the device and on the nature of; the process
water. >
s
4. Definitions and Nomenclature I
4.1 Mass Transfer Terms :
4.1.1 Oxygen Transfer Rate (OTR). Mass of oxygen
per unit time dissolved in a volume of water by an
I
oxygen transfer system operating under given
conditions of temperature, barometric pressure,
power, gas rate and dissolved oxygen
concentration.
88
-------
4.1.2 Oxygen Transfer Rate at Zero DO (OTR0).
OTR when the DO concentration is equal to zero at
all points in the water volume.
f
4.1.3 Oxygen Transfer Rate in Process Water
(OTRf). OTR for the oxygenation system operating
at a specified average DO concentration and
temperature in wastewater. i
4.1.4 Standardized Oxygen Transfer Rate (SOTR).
OTR in clean water when the DO concentration is
zero at all points in the water volume, the water
temperature is 20°C, and the barometric pressure
is 1.00 atm. ;
4.1.5 Aeration Efficiency (AE). OTR per unit
total power input. Power input may be; used either
on delivered power or wire power.
4.1.6 Standardized Aeration Efficiency (SAE).
SOTR per unit standard power input; may be based
on Total 'Delivered Standard Power or Wire Standard
Power.
89
-------
4.1.7 Oxygen Transfer Efficiency (OTE). Fraction
f
of oxygen in an injected gas steam dissolved under
given conditions of temperature, barometric
pressure, gas rate, and DO concentration-.
.4.1.8 Oxygen Transfer Efficiency !at Zero DO
(OTE0). OTE when the DO concentration is equal to
zero at all points in the water volume.
4.1.9 Standardized Oxygen Transfer Efficiency
(SOTE). OTE0 when the water temperature is 20°C
t
and the barometric pressure is 1.00 atm.
5. Apparatus and Methods
5.1 Tank. The SMU test tank is 20 feet long, 3.5 feet
foot of free
water.
wide and 9.5 feet. deep. Allowing for 1
board, the system contains 17,800 liters
of
5.2 Water. For determination of a standardized OTR,
the water to which oxygen was transferred was potable
public water from the City of Highland Park, Texas.
90
-------
5.3 Oxygenation Device. This method was;applied to a
i
variety of oxygenation devices installed in the tank
including Norton Domes, WYSS tubes, and flat plate
diffusers.
5.4 Dissolved Oxygen Measurement
:
5.4.1 In situ Membrane Electrode Measurement of
DO was employed with Section 421F of Standard
Methods (6) . - '
5.5 Temperature Measurement. Water! temperature
measurement was in accordance with Section 212 of
Standard Methods (6).
5.6 Deoxygenation Chemicals
I
5.6.1 Sodium Sulfite. Technical :Grade sodium
!
sulfite (Na2SC>3) was used for deoxygenation in
accordance with Section 6.8
91
-------
5.6.2 Cobalt Catalyst. Either reagent grade
cobalt chloride, CoCl2, was used to catalyze the
deoxygenation reaction in accordance Vith Section
6.8. ;
5.7 Electronic Computer. A digital computer was used
for running the nonlinear regression method of
1
parameter estimation described in Section 7-2.1.
5.8 Gas Flow Measurement Apparatus. ' Rotameters,
calibrated at regular intervals, were used for all air
flow measurements. ,
6.0 Procedure
6.1 Water Quality
6.1.1 General and Total DissolvedjSolids. The
water supplied for the tests was a potable public
water supply. Repetitive testing was conducted in
the water only twice so that - the TDS did not
exceed 1,500 mg/1 in any case. ;
-------
6.1.2 Temperature. The water temperature should
! •
be between 5° and 30°C. Low temperatures were :
recognized to slow the deoxygenation reaction,
which may introduce some error. :A standard 8
value of 1.024 was employed to adjust for
temperature. Appreciable error can be introduced
when the actual 9 value differs from this and the
temperature difference is more than 5°C. The
water temperature did not change by more than 2°C
during a single unsteady state test.
6.1.3 Water Quality Analyses. Initial Analyses:
Before beginning the testing ^program, a
representative sample of the water was tested and
analyzed for TDS, alkalinity, sulfite, iron,
manganese, residual chlorine, pH, total organic
carbon or chemical oxygen demand, cobalt, >
surfactant (MBAS), and temperature. ;
. , . , i | .
6.2 System Stability. The aeration : system was
operated to achieve-steady state hydraulic conditions
before starting the oxygen transfer evaluation. The
hydraulic mixing regime was established in the test
93
-------
tank for each test condition before deoxygenation.
i
6.3 Deoxygenation Chemicals. Technical grade sodium
sulfite (Na2SC>3) was used for deoxygenation. The
1
sulfite was essentially cobalt free and contained no
impurities that would alter the OTR analysis. Sodium
sulfite was added in solution by dissolving the sulfite
in a separate mixing tank before adding it to the test
tank.
i
The sulfite deoxygenation reaction is ; catalyzed by
cobalt. The cobalt source utilized was technical grade
cobalt chloride, CoCl2. The cobalt was dissolved
before adding it to the test tank. Careiwas taken to
ensure that the cobalt salt was completely' dissolved.
6.4 Addition of Deoxygenation Chemicals !
6.4.1 Cobalt Addition. A solution of cobalt salt
i
was added to the test tank to achieve a soluble
cobalt concentration between 0.3 and 0.5 mg/1 in
i
the test water.
94
-------
The cobalt solution was added before the beginning
i
of the oxygen transfer testing with ithe aeration
system operating. The solution was uniformly
distributed into the test tank. The cobalt
solution was dispersed throughout ithe tank by
i
operating the aeration system for longer than 30
minutes. The cobalt catalyst was normally added
once for each test water. ' '
!
6.4.2 Sulfite Addition. The theoretical sodium
sulfite requirement for deoxygenation !is 7.88 mg/1
per 1.0 mg/1 DO concentration. Sulfite additions
were made in 130 percent excess of stoichiometric
amounts.
i
Sufficient sulfite solution was added to depress
j
the DO level below 0.50 mg/1 at all points in the
test water. In most cases, the DO concentration
reached zero at all sample points and remained at
!
zero at least several minutes prior\to beginning
the run.
95
-------
Sodium sulfite was dissolved in a small mixing
tank outside the test tank and distributed
uniformly and rapidly into the test tank. Care
was exercised to assure adequate dispersion and
dissolution in the test tank. : .
6.5 Determination of Dissolved Oxygen, at Various
i
Points in the Tank During the Unsteady Statie Test. The
DO concentration was determined at various points in
the tank and at various times during the unsteady state
test. This determination shall be carried out by in
situ measurement of dissolved oxygen in the tank by
membrane probes.
6.5.1 Location of Dissolved Oxygen Determination
Points. Three determination points were used:
One at a shallow depth, one at a deep location<
i
and one at mid-depth. The points were mid-tank.
The determination points were located;so that each
senses an equal portion of the tank volume and
i
were distributed vertically and horizontally to
best represent the tank contents. '•
96
-------
6.5.2 Times of Dissolved Oxygen Determination. A
minimum of 20 DO values were measured at
i
prescribed times at each determination point
during the unsteady state test. In most cases, 30
points were obtained. :
6.5.3 Run Duration and DO Saturation. DO data
were obtained over as wide a range 'as possible.
Data at : DO levels of less than 10
1
percent of C* were truncated to avoid lingering
oo !
i
effects of the deoxygenation technique.
i
All test runs were continued for a period
approximately equal to 4 divided by the
anticipated value of KLa. This is equivalent to
continuing the run until the DO concentration is
98 percent of the saturation concentration, C*;
00
the system was allowed to run overnight to obtain
data on C*.
oo
Measured values and tabulated values of DO surface
saturation concentrations were ' used for
97
-------
comparative information only and were; not used as
model parameters for calculation: of oxygen
transfer rates.
6.6 Dissolved Oxygen Measurements
r
i
6.6.1 Measurement by In Situ and Sample Line DO
Probes. The in situ DO probes were fast response
probes with 1.0-mil membranes, and care was taken
!
to ensure that the water velocity was sufficient
past the probe. The probes were calibrated using
the Winkler procedure with test tank water and
checked for linearity against Winkler procedure
titrated samples. The calibration and linearity
• were established before every two runs. Probe
calibration and _ , linearity check can be
conveniently accomplished by comparing probe
readings with Winkler measurements on discrete
samples taken at the probe locations. '
7.0 Data Analysis !
i
!
7.1 Nonlinear Regression Method. This method is based
98
-------
on nonlinear regression of the model (Equation 1)
through the DO versus time data as prepared for
i
analysis in Section 7.1. The best estimates of the
parameters Kra, C* and Cn are selected as the values
•" oo w
that drive the model equation through the prepared DO
concentration versus time data points with a minimum
residual sum of squares. The parameter estimates are
selected so that the sum of the squares of the
residuals is minimized. A "residual" refers to the
difference in concentration between a measured DO value
at a given time and the DO value predicted
at the same time.
by the model
The data were calculated employing the computer program
. - - - • - j
attached as Attachment 1 to this section.
I
8.0 Interpretation and Reporting of Results
8.1 .Standardized Oxygen Transfer Rate\ (SOTR). By
convention, the oxygen transfer capacity of an
oxygenation system is expressed as the rkte of oxygen
transfer predicted by the model at zero dissolved
oxygen under standard conditions of temperature and
99
-------
pressure, usually 1.00 atmosphere and 20:°C. This is
termed the Standardized Oxygen Transfer Rate (SOTR).
It should be noted that the SOTR is a hypothetical
value based on zero dissolved oxygen in the oxygenation
zone that is not usually desirable in real, oxygenation
systems operating in process water. The SOTR value
shall be determined by correcting the values of KLa and
C* estimated for each determination point to standard
oo ;
conditions by: :
KLa20 = KLa 8 (20-T) : 2
C*20 = C* 1 ' 3
00^U 00 / \
where:
KLa = determination point value of KLa estimated
according to Section 7.2.1 or Section 7.2.2.
KLa20;' = determination point value of KLa corrected to
20°C. !
6 = empirical temperature correction factor,
defined by Equation 16; shall be taken equal
to 1.024 unless proven to have a different
value for the aeration system and' tank tested
100
-------
C* = determination point value of C*
C 20 = determination point value of C corrected to
00 20°C and a standard pressure °° ;of 1.00 atm.
= temperature correction factor = C*
1st
C*
;S20
C = tabular value of dissolved oxygen surface
st saturation concentration, mg/1, at the test
temperature and a standard total pressure of
1.00 atm, (5) ;
C* = tabular value of dissolved oxygen surface
saturation concentration, mg/1, at 20°C and a
standard total pressure of 1.00 atm, shall be
taken as 9.07 mg/1 (5) \
Q. = Pressure correction factor = |
pb + Ywtde - Pv20 !
ps + Ywsde - Pv20 !
Pjj _,.= Barometric pressure during test,|f/l2.
PV2Q = saturated vapor pressure of water at 20°C,
i
p_ = standard barometric pressure of 1.00 atm,
3 f/1'.
101
-------
Ywt = weight density of water at test conditions,
£ / J. .
i
Yws = weight density of water at 20.0°C f/L3.
Pvt = saturated vapor pressure of water;at the test
temperature, f/12.
de = effective saturation depth at infinite time,
defined by:
de = 1 C* i '
^^ sv\
Ywt c sT
The average values of KLa2Q and 0*20 shall be
calculated by averaging the values at each of the n
determination points by: <
n
Average KLa = KLa2Q = i s KLa20
n 1
*
Average C = C = 1 n C
°°20 ~20
The Standard Oxygen Transfer Rate (SOTR) shall be
!
computed by:
SOTR = KLa2o C*°°20°v
where: V = volume of water in the test tank
102
-------
The individual and average values of KLa2Q, C*oo20' ^e'
and the actual test temperature and tank :volume shall
be reported along with the SOTR. For subsurface gas
injection systems, the value of SOTE should also be
reported (See Section 8.4). If possible, the standard
deviations of the KLa, C*^,, parameter estimates should
i
also be reported.
8.2 Spatial Uniformity and Reproducibility of K^a,
C*«,2Q/ Values- Replicate tests, conducted \sequentially
under the same conditions of temperature and pressure
and the replicate KLa, C20 values can be compared
directly without temperature and pressure adjustments.
i
8.3 Oxygen Transfer Efficiency (OTE). Oxygen transfer
f
efficiency (OTE) refers to the fraction of.oxygen in an
injected gas stream, dissolved under given conditions.
The Standardized Oxygen Transfer Efficiency (SOTE),
which refers to the OTE at a given gas irate, a water
temperature of 20°C, a DO of zero, and a barometric
103
-------
pressure of 1.00 atm, is calculated. For a! given flow
rate of air, this is given by:
SOTR Ib/hr
SOTE =
1.034 Qs
where: Qs = volumetric air flow rate, scfm
104
-------
ATTACHMENT 1
Basic Program for Non Linear Regression
This Attachment gives the BASIC computer language adaptation
!
of the FORTRAN non linear estimation program. :
105
-------
10
20
30
40
50
60
70
30
90
100
110
120
130
140
150
160
170
180
190
200
210
RINT
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
400
410
420
430
440
450
460
470
480
490
500
510
520
530
540
NON-LINEAR LEAST SQUARES PROGRAM IN APPLE II BASIC'
FOR OXYGEN TRANSFER PARAMETERS
OUTPUT SETUP FOR 40 POSITION' CRT/MONITOR
: CALL - 936: REM CLEARS SCREEN
NON-LINEAR ESTIMATION FOR"
UNSTEADY-STATE OXYGEN TRANSFER"
BY"
LINFIELD C. BROWN & GEORGE R. FISETTE" i
VERSION 1.0-NOVEMBER 11, 1979" -;
PRINT "THE VALUES ARE TRUNCATED": PRINT "AND^ NOT ROUNDED OFF.'
OS = CHRS
REM
REM
REM
REM
TEXT
REM
REM :::::::::::::
REM STEP 1
REM WRITE TITLES
REM ::::::::::;:
REM
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
PRINT
INVERSE
: NORMAL
REM
REM PROGRAM HAS MAXIMUM LIMIT OF 30 DATA POINTS
REM
DIM C(30),T(30),F(30),R(30)
INPUT "IS DATA IN DISK FILE;Y/N?";A$
INPUT "INPUT NAME OF DATE FILL?";N$
IF AS = " " GOTO 650: REM GET DATA FROM DISK FILL
INPUl "DO YOU WANT INPUT DATA SAVED ON DISK,Y/N?" ;A$
PRINT "INPUT DATA IN TIME,DO DATA PAIRS"
PRINT "INPUT 999,999 AS LAST DATA PAIR"
FOR I = 1 TO 30
INPUT T(i),C(I)
IF T(I) = 999.0 GOTO 360
NEXT I
ND = I - 1.0
INPUT "BEST ESTIMATE
INPUT "BEST ESTIMATE
INPUT "BEST ESTIMATE
XK = XK / 60.0
IF AS = "N" GOTO 790
REM
REM WRITE DATA TO DISK FILE
REM-.-SPECIFIC FOR APPLE/MICROSOFT BASIC
REM
PRINT DS;"OPEN "N$;",VO,L15"
FOR I = 1 to NO
PRINT D$;"WRITE "N$;",BO,R";I
PRINT T(I): PRINT C(I)
NEXT I
"WRITE "N$;",BO
FOR C-STAR OR USE 10.0 MG/L?";CS.i
FOR C-ZERO OR USE 0.0 MG/L?";CO '
FOR KLA-PP.IME OR USE 4.0 1/HR?";XK
PRINT D$;"WRITE "N$;", BO, RO"
PRINT ND
PRINT D$;"WRITE "N$;",BO,R";ND
PRINT CS
1.
106
-------
=50 PRIM i OS:"WRI~E '"IS:" ,30,R" ;ND + ? '
560 PRINT CO • ;
570 PRINT DS;"WRITE "MS;" ,30 ,R" ;.ND * 3 ' ;
530 PRINT XK :
590 .PRINT OS;"CLOSE "MS >:
600 GOTO 790 ',
510 REM |
620 REM READ DISK FILE FOR DATA ;
630 REM SPECIFIC FOR APPLE/MICROSOFT BASIC '
640 REM
650 PRINT OS;"OPEN "NS;" ,VO,L15"
650 PRINT OS;"READ "NS:",30,RO"
570 INPUT ND ;
630 FOR I = 1 TO ND ;
690 PRINT OS;"READ "NS ;" ,BO,R";I
700 INPUT T(I),C(I)
710 NEXT I
720 PRINT OS: "READ "NS;",80,R";ND * 1
730 INPUT CS i
740 PRINT OS;"READ "NS;" ,BO,R";ND + 2 •
750 INPUT CO
760 PRINT D$;"READ "NS;" ,BO,R";ND + 3.
770 INPUT XK ;
780 PRINT D$;"CLOSE "N$ ' .
790 PRINT : FLASH : INPUT "HIT RETURN FOR ITERATIONS.";IS: NORMAL
800 CALL 936: PRINT : PRINT " DATA SET ";NS- PRINT
810 PRINT "ITERATION" TAB( 11)"C-STAR" TAB( 18)"C-ZERO" TAB( !26)"KLA" TAB ( 33)"
SUM OF" :
820 PRINT TAB( 2)"NUMBER" TAB( 26)"PRIME" TAB( 33)"SQUARES" '
830 PRINT TAB ( 11)"(MG/L)" TAB( 18)"(MG/L)" TAB ( 26)"(1/HRV"
840 PRINT , ;
850 REM
860 REM :::::::::::::::::::::: , '
870 REM STEP 2
880 REM INITIALIZATION OF VARIABLES
890 REM DO ITERATION CALCULATIONS
900 REM ::::::::::::: ::::::;:;
910 REM
920 K2 «= 0
930 OS = 0.0 ;
940 FOR I = 1 TO ND . ;
950 F(I) = CS - (CS - CO) * EXP ( - XK * T(I)) \
960 R(I) = C(I) - F(I)
970 OS = OS-H- R(I) * R(I)
980 NEXT I ;
990 ZZ$ = STR$ (CS) :VA = 5.: GOSUB 2900 \ •
1000 CS$ = ZZ$:ZZ$ = STRS (CO): GOSUB 2900 i
1010 COS = ZZ$:ZZ$ = STR$ (XK * 60.): GOSUB 2900 I
1020 XK$ = ZZ$:ZZ$ = STRS (OS): GOSUB 2900 . I
1030 OSS = ZZ$
1040 PRINT TAB( 4JKS TAB( 10)CS$ TAB( 18)CO$ TAB( 26)XK$ TAB( 33)OS$
1050 GOTO 1070
1060 REM i
1070 REM CALCULATION LOOP - INITILIZE VARIABLES !
107
-------
1080 REM :
1090 1C = KS * 1 ;
1100 Al = 0.0
1110 A2 = 0.0
1120 A3 =0.0 i
1130 A4 =0.0
1140 A5 = 0.0
1150 A6 = 0.0
1160 Cl = 0.0 :
1170 C2 = 0.0 !
1130 C3 = 0.0 :
1190 SQ = 0.0 i
1200 REM '
1210 REM :::::::::::::::::::::: '<
1220 REM STEP 3
1230 REM SETUP NORMAL EQUATIONS FOR LINEARIZED MODEL
1240 REM USING CURRENT LEAST SQUARE ESTIMATES . !
1250 REM ::::::::::::::::::::::
1260 REM ;
1270 FOR I = 1 TO ND i
1280 Z2 = EXP ( - XK * T(I))
1290 Zl = 1.0 - Z2
1300 Z3 = T(I) * Z2 * (CS - CO) i
1310 Al = Al + Zl * Zl
1320 A2 = A2 + Zl * Z2
1330 A3 = A3 + Zl * Z3 !
1340 A4 = A4 + Z2 * Z2
1350 A5 = A5 + Z2 * Z3 i
1360 A6 = A6 + Z3 * Z3 i
1370 F(I) = CS - (CS - CO) * 12
1380 R(I)'= C(I) - F(I) :
1390 Cl = Cl + R(I) * Zl
1400 C2 = C2 + R(I) * Z2
1410 C3 = C3 + R(I) * Z3 •
1420 NEXT I :
1430 REM • i
1440 REM ::::::::::::::::::::::
1450 REM STEP 4 i
1460 REM SOLUTION OF NORMAL EQUATIONS FOR CORRECTIONS
1470 REM TO THE PRIOR LEAST SQUARES ESTIMATES
1480 REM ::::::::::::::::::::::
1490 REM :
1500 Dl = Al * A4 - A2 * A2
1510 ' D2' = Al * C3 - A3 * Cl
1520 D3 = Al * A5 - A3 * A2 ;
1530 D4 = A6 * Al - A3 * A3
1540 D5 = Al * C2 - A2 * Cl
1550 XN = Dl * D2 - D3 * 05
1560 XD = Dl * D4 - D3 * D3
1570 X3 = XN / XD ;
1580 YN = D5 - D3 * X3
1590 X2 = YN / Dl
1600 XI = (Cl - A2 * X2 - A3 * X3) / Al
1610 REM :
108
-------
1620 RE! ::::::::::::::::
1630 REM STEP 5
16-10 REM UPDATE ESTIMATES, SUM OF ^QUARE"
1650 REM :::::::::::::::
i6cu RE;:
1670 71 = XI r CS
1630 72 = X2 -i- CO
1690 T3 = ;<3 + ;<:<
1700 FOR [ = 1 TO NO
1710 F(I; = Tl - (-! - T2Y * EXP (
1720 R(i; = CO - F{[^
1730 SQ = SQ -i
17*0 NEXT !
1750 REM
1760 REM :::::::::: .'
1770 REM STEP 6
J780 REM TEST FOR CONVERGENCE - PARAMETERS 1 PART IN 100,000
1800 REM i
Q82160 IF Ul ' T1 ^ °-00001) AND (X2 ! T2 - 0.00001) AND (X3 / T3 ^ 0.00001) GOT:
1820 REM
1830 REM PARAMETERS NOT CONVERGED,
1840 REM TEST SUM OF SQUARES - 1 PART IN 1,000,000
1850 REM
1860 IF ABS ((OS - SQ)/SQ).<0.000001 GOTO 2160 !
1870 REM
1880 REM SUM OF SQUARES NOT CONVERGED, ' <
1890 REM TEST NO. OF ITERATIONS
1900 REM :
1910 IF (KZ 10.) GOTO 2090 '
1920 ZZ$ = STR$ (Tl) : GOSUB 2900 ,
1930 Tl$ = ZZ$:Z2$ = STR$ (T2) : GOSUB 2900
1940 T2$ = ZZ$:ZZ$ = STR$ (T3 * 60.): GOSUB 2900 '.
1950 T3$ = ZZ$:ZZ$ = STR$ (SQ) : GOSUB 2900
1960 SQ$ = ZZ$ . .
1970 PRINT TAB( 4)KS TAB( 10)T1$ TAB( 18)T2$ TAB( 26)T3$ TAB(;
1990 REM NEW ESTIMATES !
2000 REM i
2010 CS = Tl ;
2020 CO = T2
2030 XK = T3 . :
2040 OS = SQ
2050 GOTO 1090
2060 REM '
2070 REM OUTPUTS
2080 REM
2090 PRINT :
2100 PRINT "SOLUTION NOT CONVERGED IN 10 ITERATIONS'" !
2110 PRINT "CHANGE VALUE IN LINE 2670 TO TRY MORE ITERATIONS "
2120 END
2130 REM i
2140 REM OUTPUT PARAMETER ESTIMATES >
109
-------
2150
2160
2170
2180
2190
2200
2210
2220
2230
2240
2250
2260
2270
2280
2290
2300
2310
2320
2330
2340
2350
2360
2370
2380
2390
2400
2410
2420
2430
2440
2450
2460
2470
2480
2490
2500
2510
2520
2530
2540
2550
2560
2570
2580
2590
2600
2610
2620
2630
2640
2650
2660
2670
STRS (!
ZZS:ZZS
ZZS:ZZS
ZZS:ZZS
ZZS
: GOSUB 2900
SIRS (12): GOSUB ?900
SIRS (T3 * 60.): SOSUB 2900
SIRS (SQ): GOSUB 2900
REM
ZZS =
T1S =
T2S =
T3S =
SQS =
PRINT
PRINT
REM
REM
REM
REM
REM
REM
XF = ND - 3.0
RS = SQ / XF
ER = SQR (RS)
PRINT "STD DEVIATIONS OF PARAMETER ESTIMATES-
PRINT
DP = Al
TAB( 26)T3S TAB( 33)SQS
STEP 7 " '-.
COMPUTE STANDARD DEVIATIONS OF PARAMETER ESTIMATES-
DN
DT
Fl
F2
F3
VI
V2
V3
SI
Al
DP
A4
Al
Al
A6
A5
+ 2.0 * A2 * A3 * A5
A4 * A3 * A3 + A6 * A2 * A2
A5 * A5
A3 * A3
A2 * A2
RS
RS
RS
A4 *
AS *
DN
A6 -
A6 -
A4 -
(Fl / DT) *
(F2 / DT) *
(F3 / DT) *
SQR (VI)
S2 = SQR (V2)
S3 = SQR (V3)
ZZS = STR$ (S1):VA = 5.: GOSUB 2900
51$ = ZZ$:ZZ$ = STRS (52): GOSUB 2900
S2S = ZZ$:ZZ$ = STRS (S3 * 60.): GOSUB 2900
S3$ = ZZS
PRINT " UNITS" TAB( 10)51$ TAB( 18)S2S TAB( 26)S3$
SI = SI / CS * 100.0
S2 = S2 / CO * 100.0
S3 = S3 / XK * 100.0
ZZ$ = STRS (S1):VA = 3.: GOSUB 2900
SIS = ZZS:ZZS = STR$ (52.): GOSUB 2900
S2$ = ZZ$:.ZZ$ = STRS (s3): GOSUB 2900
S3$ = ZZS"
PRINT "% OF LSE" TAB( 10)51$ TAB( 18)S2$ TAB( 26)S3S
PRINT
ZZS = STRS (ER):VA = 4.: GOSUB 2900
ERS = ZZS
PRINT "ESTIMATE OF ERROR = ";ER$
REM
REM ::::::::::::::::::::-
REM STEP 8
REM WRITE SUMMARY
REM ::::::::::::::::::::::
110
-------
2530 REM :
2590 • PRINT . '
2700 FLASH : INPUT "HIT RETURN FOR SUMMARY OF DATA.";IS: NORMAL
2710 CALL - 936: PRINT : PRINT : RE-1 CLEARS SCREEN •
2720 PRINT TAB( 12) "SUMMARY OF DATA"
2730 PRINT : PRINT !
27*0 PRINT TAB( 3) "TIME" TAB( 16)"COHC" TAB( 22)"F" VALUE'1 TA.B( 32V'RESIDUAL:'
2750 PRINT TABf 3)"(MIN)" TABf 15."(MG/i V- TAB( ?3)"MG/L"
2760 PRINT ' !
2770 FOR I = 1 TO ND :
2780 ZZS = STRS (F{I)):VA = 4.: 60SUB- 2900 i
2790 HIS = ZZS:ZZS = STRS (R(I)): GOSUB 2900 !
2800 H2S = ZZS :
2810 PRINT TAB( 2)1 TAB( 8)T(I) TAB( 16)C(I) TAB( 25;H1S TAB( 33)H2S
2820 NEXT I ;
2830 PRINT : PRINT :
2840 PRINT "*********************************•*•***•*••*•''
2850 END i
2860 REM j
2870 REM OUTPUT FORMATTING ROUTINES i
2880 REM SPECIFIC FOR APPLE/MICROSOFT BASIC i
2890 REM :
2900 LL = LEN (ZZS) •
2910 IF LL<12 THEN ZZS = LEFTS (ZZS.VA): RETURN \
2920 IF MID$ (ZZS,LL - 2,1) = "+" THEN ZZ$ = LEFT$ (ZZ$,VA - 3) + RIGHT$ (ZZ
$ 3): RETURN :
2930 CC = 2.: IF LEFTS (ZZ$,1) = "-" THEN CC = 1.
2940 IF MID$ (ZZ$,LL - 3,1) = "E" THEN EE = VAL (RIGHT$ (ZZ$,2)):NN$ = MIDS
(ZZ$,CC,1): FOR J = 1 TO EE:NN$ = "0" + NM$:: NEXT J:ZZ$'= "." NN$ + MID$ (Z
Z$,CC + 2,LL - 4): IF CC = 2. THEN ZZ$ = "-" + ZZ$ !
2950 ZZ$ = LEFTS (ZZ$,VA): RETURN ;
2960 REM ' ' 1
2970 REM NON-LINEAR LEAST SQUARES PROGRAM FOR
2980 REM UNSTEADY-STATE OXYGEN TRANSFER i
2990 REM LY LINFIELD C. BROWN & GEORGE R. FISETTE I
3000 REM VERSION 1.0-NOVEMBER 11, 1979 '
3010 REM COPYRIGHT BY ASCE !
Ill
-------
AJE> :E> EISTID i x
112
-------
2
So
i as
i iii
X 1 M
X 1 0
i
01 1
1
> 1 PI
a i o-
• 1 N
si
•X 1
.p 1 —
u * 1
a -a \ >.
^!i
•SSI0
Q 1
•f fu3«r
Type Conditio
ton
Q 1 O
!z
* » i pi
a a i PI
i t
o-
n
o
o
CO
n
•H
Q
C
a
CD
PI
PI
•t
n
o
to
0-
Q
C
0
o
N
00
3
«r
PI PI
o o
S 2
" n
PI 0
* S
Q a
c c
o o
1 1
CO CD
0 0
M K)
4- 03 00
?. 4. „, _ _
-. u
c
O PI PI tv . >. X >-
Q Q Q Q Q
1
X
3
U.
U
M.
iS
3E.Q:
o
L IT
4
£
X
3
IL
4? u7 r/
L cr
3 E
O M-
U
* 10 t
in tr -o in
TJ X -1 -*
TJ e
~l M-
z: u
L H-
ffl tr PI PI
c «
C X
>-< e
M-
u
-< PI
* *
Q O
iii In -
07 f7
!
K n -o
PI -<
I
«' i 5
Q Q Q
i
,
1
1
1
1
1
1
1
I
1
1
1
1
CO 1
li. 1
U. 1
»-• 1
a i
i
iii i
(D 1
co ! a.
•-• i in
| g
O 1
U. 1
1
1
1
il.
x i e
tn i o
1 L
> 1 U.
(L 1
FOULANT SUMMA
ueera Recieved
1 H-
1 H-
I a
&
PI
«r
Recei vedi
o
4J
*8
Q
« C
TJ TJ -
II -
*"a S
il! tn x
-
m
•p e
§-
U 4J
1. *
0. O
= «;
-.Ox
> —
« C
^^ TJ ~4
v ~ a
0 O S
i- tn x
te Diffuser
ted Type
1J 81
Q U
(—
CO
0
o
d
N
CO
o
0
d
,
in
o
d
'87 Sanitaire
* Dirty HI
pi
pi
x
o
d
S
M
O
O
d
s
d
'87 Sanitairo
* Dirty »2
PI
PI
•t
PI
0
6
pj
d
W
PI
0
d
CO
o
d
'87 Sanitaire
* 'Dirty t»3
•0
PI
x
'oo
0
d
o-
"
o
o
d
to
8
d
'87 Sanitaire
* Dirty t»4
Q
to
•t
PI
PI i i
o ~.
0 TJ -0 5
X — CT
— O Ul
I u. tn x
i CT
V)
0 *
n *< 01
c —
01 44
01 L i»
1 3 >
•Q
oil
0 TJ J.5
O 1 U 4J •« O-
1 —"ox
1 O OT O1
t 1
PI
o - * S
d « 2 **
1 4J — O-
o o n
i i- tn x
j ~
1 I
i i
II ! L
v. n 01
•-. * no
* 3 a
•P >- 1 H- >
•** 4J 1 H- h-
C L ••«
li -H Q
in Q
N
CO
O 1
n i
«f
CC «f
<£ O
Ul O
•d
a: «
a: •
ui -a
or in
Lt PI
tu o
o
0
d
cc t
cc o
Ul O
d
* Sanitaire
Dirty #1
* Sanitaire
Dirty »2
o
0
d
pi
•t
00044
O
o
0
d
* Sanitaire
Dirty #3
00
o
, o
d
*
IN
IN
(
i
OOO17
O
CO
8
d
i
1
- -* Sanitaire
Dirty *4
i
o
o
d
b-
PI
pi
M
§
d
c
8 S
0 01
• .p
o
TJ
U
ig
M
a
* Sanitaire
Dirty *5
* indicat
113
-------
in
Ul
§
u.
u.
o
_j
IB
o>
D
HI
%
IL.
or
o
u.
f-
m
a
01
>
g
01
H
z
oi c
-I Ul ~4
.- 13
jj — cr
•0 -4 81
-.Ox
O tfl .
a
.£
m a
*
„
c -
01 ~4
U JJ
01 i. a
-4 01-4
£1 LL 0
3 >
0
01
c
ai c
T) -.10 -.
•-4 -.4 13
u jj — cr
v
ui 2
flj ft
*J ^ cr
o o n
i- en x
(7
L
01
n ii
3 a
•«- >-
4- h-
a
0~ LC.' 1 '! • 0" C 1
O OC — i CI O
O LU O O O
d O i O ' O
1
[
o* i^ f** ; *V
B »
U 13
HI V
(C. >
in 3
I. U
01 01
ui o:
3
M- 01
H- -*J
Q Q
1
i PI • -o re -o re
1 PI CO PI 0 PI
X | _ _ r., g. r,
X 1 o O O O O
Ul 1
1
1
1
§1 -0 111 *t «t -4
i m PI in PJ o
PI 03 -0 PI |x
• 1 0- CI PI PI l-i
O1 I rt — 01 -4
-------
1-
a:
3
t
*J
8-
a
Ul
g
1
N
CO
•« X
ON
tJ
•o
U
>
U 13
a v
cc. >
* t
« a
• V
n cc
3
•*• •
» *J
D Q
x
X
0)
a:
01
,
en
•4- (-
a
a •
v n
a D
h~
n
0-
*^
d
0-
CO
n
K>
N
^
*
X
a.
c
o
I
CO
X
M
t
rt
O xj
[M N - M
f
1
I
;
M :« N 10
N N N *
* w * *
1
i
N to «f n
* * * *
XXII
a. a. a. a.
i I i I
1
i
i
i
i
i
i
i
i
i
i
i
i
i
01 1
IT 1
Ul 1
01 1
D 1
U. 1
LL 1
i- 1
O
Ul
Or 1
D 1 -J
ai i a
1 •
Ul 1 a
5 ui
u. 1 E
1 U
1
1
1 D
21
<£ at
o
i
i
1 N
t- CO
Ul X
x! i »
01 1 O X
1 L *
> 1 U.
O 1 tt
1 U Tl
Hit a
z i a: >
C 1 -•
_i i w a
3 1 L. U
O 1 9 U
u. i n cc
1 3
i — «
i a D
^
» c
•D -a -
X rt U
— o m
U. 01 X
D
4J O
C —
« .-
U 4-1
» -
>
«.=
— -a
A> — cr
H -1 «
« 0 X
O • 9
> ~
m c
-, -a -.
i> ^ cr
1- 01 x
u
n •
_3 tt
•*• i-
a
TJ
• •
4J 4J .
* •
Q «
f-
o — N n *
PI _ ^ _ n
o o o o o
o o o o o
£ £ £ m n
f CO h- »^ O
•-< M ri M ti
* r> in V*4 'CCJ TiK) IQ^ 4j|n
4J* -U* 4J* 4J* 4J*
CX CI CI CI CI
«a. *a. to. ma. ma.
uix cox co i cnx cox
aj co co oo co
X XX X X
N M -0 O O
CM M « M «
X X X X X
* * * * *
_
in c
•a TI -.
• x -H cr
•-. 0 ai
u. to x
ot
•p ffi
c •-•
U ^J
01 L. us
rt nj -.
§ >
ut
c
•rt ^ TI
u -u — cr
— a x
O W O1
m-
r4 Tl -4
* -1
*< « cr
o o n
K U) x
en
t
3 a.
M- K
a
>
Oi f t ': T m
tc o o i o o
UJ O O , O O
d d d d
i
ce N Fo co ii
CC • • ' • •
LU IN CO -0 CD
i
CC CD * i N 0-
£n M , M ix
o 0,0 o
o 0:0 o
0 0 i 0 0
d d d d
•V4
a: m * «r o- a
cc o o o o m
ul o o o o u
• . • • * 4J
O O i O O
| -a
u
; is
m
HI
O 31 II 11 ffi in
L L. L. . ' L L- U
'S^ 'SN "i n ~* T 'Sn T!
4J* -P* 4J*,4J* -U* C
?X CI CX CX CI
ma. aa. « CL « a.
-------
in
us
in
£
u.
0
1U
m
01
o
Ul
z
a!
g
u.
til
in
X
tn
|
cn
|
4J
a
2
U
3
£
tn
fv.
*x
E CM
O x
U *
U.
'g
II
U T>
f£ >
• e
iu
II
* •
*** "li
Q Q
^
» C
•D T) -i
x ~ tr
i o n
u. tn x
*• a
C -i
u •-.
U 4J
L if
e —
a. o
^
8 C
•* Tl "
— Ox
one?
> ~-
« c
— -D ._
If ~«
O O «
t- tn x
S1
L
II
3 a.
* t-
0
« *
« * I
I
in oa jj *• r.
o ci ~ r- CM
o o o o o
o o o O o
«i n « r> *
10 CO l"5 tM IT
n n w N to
n on eo - CM
§-• O CM —
0000
d d o d d
.
N O * CM r>
O » M 0> KJ
O O 0 0 O
d o d d d
« CM M
'3
91
C
" * 2
TJ ~ » —
— ,- -a
u ii — tr
r-l O X
0 * *
^
» c
* —
*» *M o <*• a-
8 — o n o
o o o o
00600
^ N N N M
j;
3
tn
4J
Q.
£
u
E
U
tn
..?>
E -
O N
L. x
U. «»•
TJ
e
>
i*
U Tl
a a
cn >
in ai
L U
(U II
lit IT
H- 9
o a
fe
M-
o
L
U
n
3
1^
s
a
fi
Q
x
w
>
a;
£0
.
?
c
o
(1
(J
•v4
H-
4J
•5
S
^
4J
Tl
0
-
e
g
(_
i
«l
« i
O 1
h- i
i
i
i •*•
i M
1 M
1 .
i °
j
1
1 -0
1 **
\ -
1
1
I
1
I
1
i
i "
i *
I 3
I
i B
u
c
a
z
•to
rj
^~
n
w
o
s
K
CO
^
PI
*
~°
3
Tl
u
c
o
§.
03
X
n
CM
X
•t
-0
d
—
CO
10
f^
*
Tl
O
c.
*
«
3
•o
u
c
a
L
O
Z
N
CO
X
•0
N
X
in
o-
d
HI
^
o
_
•^
o-
•n
«
£
3
TJ
U
X
n
d
n
•o
Q.
in
*
TJ
e
c
n
3
TJ
«
g
a
z
00
x
-0
IN
V.
in -H o o- w
O r^ *^ CD 03 tt
•^ • > n fcl ^ 0
-o «o ^d , d
O 6 -> *4
• H-
M U
o o «t r)
• o e -d (h n ^3
Q> • H*
* ••-" U
c
^^
„. N 0 CM -
ane n * n ,, *
• u u
2
u
X
rt XTJ 6
U. 3 -i M-
. r £ °
L U. v
5
61 **'
4* 4J
If L M-
"ssr
X C X
3 « E
^^ M-
u. u
-i « 0- CM
«t o r*v (IT
Kl CM K) <}•
.
o ri rt n
N CD 0* Ix
CM CM — rf
' 5
U
if
, L
; U
r** ^ o o
o m o CM
CM M n CM'
* * * * *
TJ TJ Tl TJ TJ
a •. • u a
c. £l c. c jc
• « « n i!
« * i n o n
3 33 3 3
TJ TJ ' TJ Tl TJ
U U U u U
C
-------
1
o
cc
ui
i-
LL
O
a
u
a
ui
E
U
N
CO
1.
U. •
QC >
n •
X
X
cn
SL
m
•
s
„
s e
lit!
a a
n
r-i
0
n
PI
•o
*
TJ
01
C
*
c
o
1
N
CD
M
N
«»•
Ob
in
C1
d
M
«t
»«
M
ft
3
C
*
c
o
^
a
z
N
m
PI
PI
4-
o
3
41
C
01
o
L
CO
M
*
TJ
II
C
*
g
1
N
CO
-0
«•
111
M
£«
Ill
0-
CD
*H
*
«:
TJ
C
*
C
o
a
z
rx
CO
-0
M
t-
0-
n
O
PI
CD
PI
in
*
TJ
01
C
*
C
o
a
z
N
CO
•0
PI
«r
i
0 N
in n -o
or*- « -o
4J «i CC O 0
1) u. OQ
Q
O -O
c-i i>
o e -«
PI u
0 CO
" o e o- N
O> . 4-
3 - u
o
C I
-I 4J
•0 -0 C'
CL —
3 il 6 CO N
a f* t- 01
• U X
a
^_
a
PI N
S
n M-
• u
-« N M
* * *
T) TJ TJ
U II 01
C C C
525
N n
<»• CD
0 0
PI 0-
. th N
0- PI
1^ O
N CO
N 0-
K> -a
r~! ai
«»• n
* *
TJ TJ
0 01
C C
* *
11
m
cc
x
3
U.
ti
a
If "
CL 4>
n
z a:
o
^H X
LL 3
L LL
3
m
4J
a
cc
K
3
LL
^j
a
91
TJ
TJ
L,
e
c
c
H4
S K S;
cr • • !
in to o-i ;
6 ',
U |
^_
cr o- r-.
X PI PI ,
E
M- O
U I
•^ 4J
1 C
a
, "a
n
E
U
" PI • «
* » *
TJ TJ TJ
01 a *
c c c
i
i
N CD
O 117
«r pi
-0 N
CO CO
0- tt
0- ^
PI •»
<»• n
* *
TJ TJ
a ii
c c
2 5
cn
cc.
D
U.
U.
a
Ul
a>
o
3
at
Ul
z
»-*
LL
i
U.
f-
Ul
T
cn
^.
cc
E
cn
i-
z
Q
LL
1
a
a
a
§
3
to
^
m
E N
O X
L *
LL
^J
*
•
s >
*r+
M ft
L U
S 01
tt CC
3
4- 01
M- 4J
— «Q
a a i
,_,
ffl c
TJ TJ —
x — cr
— * o «t
U. to x
n
•u e
c —
01 —
U 41
L. a
u M
CL a
. >
,^
J? * 5
22 cr
-OX
O « O"
n c
** TJ -^
4 -^
*> -< cr
o a «
i- en x
L.
01
m ft
3 a
M- X
M- 1-
' Q
TJ
n D
9 W
Q «
1-
n m o to
-< -> c-i n
o o o o
do o d
x j*s i^ X
N Ch in fx
NO »1 >t
N to M i
i
•0 r~- -H -o
§o c -< o
o — o o
d d c d d
a
0
m
_ ^ ^ o-
PI PI MM
O O O O
0 O O 0
• ~< OIPJ «l M 114- V n
L* L * 1. « L,» L*
ISTJ 4TJ (S13
o
u
c
T, ".£
-^ — TJ
^ t!Sff
^. o -^
O M O1
> —
-
* C
^1 TJ T»
« -<
** -* cr
o*
i_
V •
3 a.
M- X
H- h-
S
tx CO CO 0-
O 0 0 ~
Q O O O
do i d d
J
j
(
•*? x 'MM
0- O -0-0
in si -o -r
r
O i
I •
•M '
PI * n -
«• >«• c n o-
o o — o o
o o o o
o o c • o o
d d ^ : d d
0
L 0"
a : c
IX 0- . i CO O 4>
o o ; o N «
o o o o o
0 O ; O O
TJ
1 -r4
U
4
m
11
a -. ON UK), lit en 4
L* L,* 1- * U* L.* U
4TJ 413 4TJ, 4TJ 4TJ TJ
4J 11 4J H -U • 4J B JJ « C
--C — C -
-------
AE> E> EISTID X 2C
118
-------
NEW DOMES CLEANED
BY NOTED METHODS
ASCE/EPA
DATA SUMMARY
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
LPH1
LPH
LPH
LPH
LPH
LPH
HPH2
HPH
HPH
HPH
HPH
HPH
AW3
AW
AW
AW
AW
AW
AIR FLOW
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
-•• 1.0
2.0
2.0
2.0
T (°C)
14
14
14
10
10
10
10
10
10
10
10
10
5
5
5
5
5
5
4
4
4
4
4
4
C*
10.67
12.55
12.85
10.46
11.20
11.74
10.61
10.50
11.10
10.50
11.17
10.75
12.40
12.50
11.77
11.84
12.16
12.30
13.20
13.36
13.90
13.20
13.70
13.70
KLat !
0.0107 ;
0.0075 i
0.0083 :
0.0276
0.0263 1
0.0250 >
0.0369
0.0240
0.0230 '
0.0370 ;
0.0350 '
0.0410
0.0220 i
0.0194 :
0.0220 ;
0.0480 ;
0.0430
0.0400 '
0.0206 !
0.0206 I
0.0204
0.0421 ;
0.0396 '.
0.0380
KLa20
0.0152
0.0108
0.0112
0.0399
0.0375
0.0358
0.0300
0.0320
0.0315
0.0510
0.0477 .
0.0560
0.0314
0.0277
0.0316
0.0685
0.0613
0.0570
0.0300
0.0300
0.0298
0.0615
0.0580
0.0560
LPH1 - LOW PRESSURE HOSING
HPH2 - HIGH PRESSURE HOSING
AW3 - ACID WASHED
119
-------
NEW DOMES CLEANED
BY NOTED METHODS
ASCE/EPA
DATA SUMMARY
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
LPH1
LPH
LPH
LPH
LPH
LPH
HPH2
HPH
HPH
HPH
HPH
HPH
AW3
AW
AW -
AW
AW
AW
AIR FLOW
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
• 1.0
2.0
2.0
2.0
T (°C)
11
11
11
11
11
11
9
9
9
9
9
9
6
6
6
6
6
6
4.5
4.5
4.5
4.5
4.5
4.5
C*
10.66
10.20
11.00
11.01
10.83
11.08
10.88
11.07
10.77
10.82
10.96
10.95
12.79
13.00
12.87
12.67
12.74
13.20
13.40
'14.00
14.30
13.76
14.40
13.30
KLat I
0.0196 i
0.0236 !
0.0197 !
0.0350
o.oseo ;
0.0306
'
0.0190
0.0195 ',
0.0200
0.0384 ;
0.0360 ;
0.0380 j
0.0230
0.0210 !
0.0210 ;
0.0430 '
0.0380
0.0350 ;
0.0260 '
0.0220 ,
0.0205
0.0430 I
0.0400 ;
0.0460 ;
KLa20
0.0266
0.0321
0.0270
0.0480
0.0490
0.0420
0.0240
0.0250
0.0260
0.0500
0.0460
0.0490
0.0320
0.0290
0.0300
0.0600
0.0540
0.0480
0.0370
0.0320
0.0296
0.0625
0.0580
0.0660
LPH1
HPH2
AW3
LOW PRESSURE HOSING
HIGH PRESSURE HOSING
ACID WASHED
120
-------
NEW DOMES CLEANED
BY NOTED METHODS
ASCE/EPA
DATA SUMMARY
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
LPH1
LPH
LPH
LPH
LPH
LPH
HPH2
HPH
HPH
HPH
HPH
HPH
AW3
AW
AW
AW
AW
AW
AIR FLOW
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
• i.o
2.0
2.0
2.0
T (°C)
9
9
9
9
9
9
4
4
4
4
4
4
5
5
5
5
5
5
7
7
7
7
7
7
»»*
C"
11.58
12.15
14.30
11.32
10.30
10.04
12.60
12.10
10.88
11.90
12.02
11.85
12.80
12.50
13.20
12.77
12.49
12.24
13.10
13.96
14.00
13.26
14.50
13.93
KLat i
0.0150 :
0.0150 ;
0.0100
0.0360
0.0530 i
0.0560 ;
;
i
0.0150
0.0180 '
0.0240 i
0.0360 !
0.0350 .
0.0360 ;
0.0210 i
0.0240 !
0.0210 i
0.0420 I
0.0530
0.05101
0.02401
0.0199
0.0190.
0.0420^
0.0310!
0.04001
KLa20
0.0200
0.0190
0.0130
0.0460 ;
0.0690
0.0730
0.0230
0.0265
0.0360 :
0.0530
0.0510
0.0530
0.0305
0.0350
0.0300
0.0590
0.0760
0.0730
0.0330
0.0270
0.0250
0.0570
0.0420
0.0540
LPH1
HPH2
AW3
LOW PRESSURE HOSING
HIGH PRESSURE HOSING
ACID WASHED
121
-------
NEW DOMES CLEANED
BY NOTED METHODS
ASCE/EPA
DATA SUMMARY
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
CLEAN
LPH1
LPH
LPH
LPH
LPH
LPH
HPH2
HPH
HPH
HPH
HPH
HPH
AW3
AW
AW
AW
AW
AW
AIR FLOW
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
1.0
2.0
2.0
2.0
1.0
1.0
, 1.0
2.0
2.0
2.0 t
T (°C)
14
14
14
14
14
14
10
10
10
10
10
10
7
7
7
7
7
7
4
4 •
4
4
4
4
»'*
C"
10.23
10.50
11.30
10.19
10.79
10.65
10.25
10.39
10.29
11.34
10.88
11.73
12.50
13.36
13.34
12.76
12.53
13.15
13.15
13.97
14.60
13.50
14.30
14.70
KLat ,
0.0162
0.0175 i
0.0150 ;
0.0330 :
0.0330 ;
0.0340 ;
i
0.0260 :
0.2480 !
0.0260 ;
0.0370 '
0.0480 :
0.0303
0.0250 •
0.0214 :
0.0210 !
0.0450 :
0.0470 i
0.0420
i
0.0270
0.0210 :
0.0170 i
0.0450 •
0.0380 ;
0.0310 :
KLa20
0.0220
0.0240
0.0210
0.0450
0.0450
0.0470
0.0350
0.0324
0.0340
0.0480
0.0640
0.0403
0.0343
0.0290
0.0280
0.0610
0.0640
0.0580
0.0390
0.0310
0.0250
0.0650
0.0550
0.0460
LPH1
HPH2
AW3
LOW PRESSURE HOSING
HIGH PRESSURE HOSING
ACID WASHED
122
-------
w
1
w <2
1-1
Ed
1
CO
LU
CD
cez
UJ '
UJ
1
i
Lt_
O-
_
UJ
CJ
LU
tJ
E
s
u-
CO
Ow S
DC
CU UJ
° a
fe
en
en
DC
C3
•- ILJ CJQ r-» -a c*- -o *r? -o -a
«*< o o -o o> <^0<=>
03 c— c— c-* -a in bra
co co ca o> -3- -d- in
co t*- - o <=»<=><=•«=><=»
ru cu cu cu ru ru ru
CU O- CO -d- -Jl- — « CO
•**• tn -o -^i c— a-- 03
o- o^ o- a- o- o- o-
o o <=. c> -o o <=»
«o to «o CD •<*• cu w
en ~a at *r-t •&- *~o ca
cu eo **• iff -*• U"3 -^
«—«•-« ro -o co ITS ~o
co — « -o en o- <*• ira
e> <=• o o <:• o o
in in in in in tn tn
eu cu ru cu ru cu ru
o o <=> o •=?«=> o-
cu eu ru cu ru cu cu
cr- e>- o- o* e>- er» o-
O O O O ^> ^> ^7
JT m -o m •*• ru ru
0 0 0 <» O ^ <=•
ru co -a- -•»• in m in
gggfgSg
^ inm*A«a-cicoco eu
eu cu cu cu eu cu tn
ru eu ru ru ru ru
C3C3CUCUCUCUCUCUCUCU
ru ru ru ru eu eu ru
cu cu ru cu cu ru cu
cucucueucururucu
frj
-o
en
-O
^oc^f^b^inbi^a-
cocoeocncurucucu
S
ru
------- |