EPA-600/4-77-023
April 1977
Environmental Monitoring Series
INVESTIGATION OF A HONEYWELL DISSOLVED
OXYGEN PARAMETRIC SYSTEM
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/4-77-023
April 1977
INVESTIGATION
OF A
HONEYWELL DISSOLVED OXYGEN PARAMETRIC SYSTEM
by
A. F. Mentink
J. 0. Patterson
T. E. Hickman
Instrumentation Development Branch
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and
Support Laboratory - Cincinnati, U.S. Environmental Protection Agency,
and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
ii
-------
FOREWORD
Environmental measurements are required to determine the quality of
ambient waters and the character of waste effluents. The Environmental
Monitoring and Support Laboratory - Cincinnati conducts research to:
develop and evaluate techniques to measure the presence
and concentration of physical, chemical, and radiological
pollutants in water, wastewater, bottom sediments, and
solid waste;
investigate ways to concentrate, recover, and identify
viruses, bacteria, and other microbiological organisms
in water, and to study the responses of aquatic orga-
nisms to water quality; and
assure Agency-wide standardization and quality control
of systems used to monitor water and wastewater.
This investigation of the Honeywell submersible model dissolved oxygen
parametric system, one of many instrumentation systems investigated by
the Instrumentation Development Branch, was pursued to determine the
effectiveness of that system in Environmental Protection Agency
research sewage treatment programs.
Dwight G. Ballinger
Director
Environmental Monitoring and Support Laboratory
Cincinnati
iii
-------
ABSTRACT
A Honeywell dissolved oxygen parametric system was investigated for
possible application in EPA's research on sewage treatment. Labora-
tory and field data were accumulated. Summaries on selected background
and theoretical aspects of the measurement have been included for those
unfamiliar with this type of instrumentation.
iv
-------
CONTENTS
Page
Abstract iv
List of Tables vi
List of Figures vii
Acknowledgement x
Sections
I Introduction 1
II Conclusions 2
III Recommendations 5
IV Overview of the Honeywell DO Parametric System 8
Sensor Assembly 8
MV/V Transmitter 11
V Laboratory Investigation 19
Sensor Assembly Investigation 19
MV/V Investigation 53
VI Field Investigation 72
Introduction 72
Final Settling Basin (Secondary) 75
Performance in the Aeration Basin 78
Effluent Chamber (Chlorinated) 80
VII Evaluation of Manuals 87
Overall Appraisal of Manuals 88
Sensor Manual 88
MV/V Transmitter Manual 88
VIII References 90
Appendices
A. Background on Dissolved Oxygen Measurements 91
References 100
B. Theory of Dissolved Oxygen Measurements 102
References 109
-------
TABLES
No. Page
1 Components of DO Sensor 8
2 MV/V Functional Components 12
3 Room Temperature Zero and Span Calibration for the
MV/V 19
4 Calibration Data of the Honeywell DO Parametric
System Employing Sensor 1 20
5 Transient Response Procedures 24
6 Calibration Data for the Honeywell DO Parametric
System Employing Sensor 2 29
7 Selected Response Points for Membrane Only 43
8 Thermal Compensator Resistance 43
9 Manufacturer's Specifications for Sensor Assembly 52
10 Manufacturer's Specifications for MV/V Transmitter 54
11 MV/V Thermal Shift 61
12 Linearity of MV/V 30 Days Following Independent
Linearity Tests 64
13 Plant Design Data 73
14 Measurements Made in Final Settling Basin 77
15 Measurements Made in Aeration Tank 81
16 Measurements Made in Chlorinated Effluent 86
B-l Saturated DO for Selected Values of temperature 108
and Chlorides
vl
-------
FIGURES
No.
1 Submersible Type DO Sensor, Wiring Diagram with
External Circuits in Measuring Instrument 9
2 Simplified Schematic of the Honeywell Sensor's
Circuitry 9
3 Functional Layout of MV/V 13
4 Simplified Diagram of MV/V Filter Network 14
5 Theoretical Frequency Response of MV/V Input Filter 16
6 Total Functional Circuit of MV/V. 18
7 Forma Bath for Temperature-Compensation and Thermal
Responses 21
8 First Sensor Assembly Configuration 22
9 Pipe Support System for Sensor Assembly 23
10 Short-Term Response of Sensor 1 25
11 Sludge Accumulation on Cathode of Sensor 1 27
12 Photographs of Sensors 1 and 2 28
13 Temperature Compensation of Sensor Assembly 30
14 Room-Temperature Stability of Sensor Reduction
Potential 32
15 Twenty-four Hour Total Sensor Response (RT-*CT) 33
16 Twenty-four Hour Total Sensor Response (CT->RT) 34
17 Twenty-four Hour Total Sensor Response (RT-»HT) 35
18 New and Old Electrolytes 37
19 Sensor Head 38
20 Twenty-four Hour Membrane Response (RT^CT) 39
-vii
-------
FIGURES (continued)
No. Page
21 Twenty-four Hour Membrane Response (CT-HIT) 40
22 Twenty-four Hour Membrane Response (RT-»HT) 41
23 Twenty-four Hour Membrane Response (HT+RT) 42
24 Twenty-four Hour Thermistor Only Response (RT-»CT) 44
25 Twenty-four Hour Thermistor Only Response (CT-^-RT) 45
26 Twenty-four Hour Thermistor Only Response (RT-^HT) 46
27 Twenty-four Hour Thermistor Only Response (HT-*RT) 47
28 Twenty-four Hour Total Sensor Response (RT-KTT) 48
29 Twenty-four Hour Total Sensor Response (CT->RT) 49
30 Twenty-four Hour Total Sensor Response (RT-»-HT) 50
31 Twenty-four Hour Total Sensor Response (HT-»RT) 51
32 Temperature Stability of MV/V (Before and During
Runaway) 55
33 Temperature Stability of MV/V (Immediately After
Runaway) 56
34 Temperature Stability of MV/V (After Runaway) 57
35 Temperature Stability of MV/V (1 Month After Runaway) 59
36 Temperature Stability of MV/V (1 Month After Runaway) 60
37 Measuring Circuit Components of MV/V 61
38 MV/V End Point Linearity 63
39 Independent Linearity of the MV/V 55
40 Independent Linearity of Panel Meter 66
41 Independent Linearity of the MV/V 67
viii
-------
FIGURES (continued)
No. Page
42 Step-Change Response of the MV/V 69
43 Frequency Response of the MV/V 70
44 Sewage Flow Diagram 74
45 Sensor Installed in Secondary Settling Basin 76
46 Sensor Installed in Aeration Basin 79
47 Sensor Output in Activated Sludge Following Manual
Cleaning 83
48 Sensor Installed in Chlorinated Effluent Chamber 84
A-l Simplified Concept of Membrane "Canals" 94
-------
ACKNOWLEDGEMENT
The authors acknowledge J. Roesler,* for recommending that the Honeywell
system be investigated and L. Koppel, A. Gealt, R. Metarko, and J.
Scarlett of Honeywell Inc. for loaning the instrumentation, for permit-
ting Honeywell's drawings, schematics, specifications, and data to be
reproduced, and for their informative and constructive comments. The
authors thank Mr. H. Augustine, Superintendent, for permitting this
investigation to be pursued at the Hamilton, Ohio, Sewage Treatment
Plant, and Mr. T. Harrel, Supervisor of the plant for his assistance
throughout the field tests. The authors also extend their gratitude to
the plant operators for logging data 'during the second and third work
shifts.
*Wastewater Research Division, Municipal Environmental Research Labora-
tory, Office of Research and Development, U.S. Environmental Protection
Agency, Cincinnati, Ohio.
-------
SECTION I
INTRODUCTION
The basic objective of this study was to determine through laboratory
and field investigations whether Honeywell's submersible model dis-
solved oxygen parametric system,* is suitable for use in EPA's programs,
especially those involving sewage treatment plants.
Honeywell has been a major supplier of this type of instrumentation and
provided this particular system for application and investigation by
EPA.
It is evident that an instrument system employed in field applications--
whether for permit program, enforcement, regulatory, or for process
control--must be ruggedly designed, perform within tolerance, require
minimal maintenance, and be readily serviceable. Additionally, the
manufacturer must be able to assist the customer in emergency situations.
These areas have been given consideration and are also discussed.
*By definition, a parametric system includes the detecting function
(sensor) and the signal conditioning function (in this instance, the
MV/V). The detecting function includes temperature compensators.
-------
SECTION II
CONCLUSIONS
OVERALL
The system can be productively employed in a treatment process provided
the manufacturer's instruction are followed.
SENSOR
1. Two sensors were investigated and both failed after 3 months of
operation. Neither operated over its expected "life without
maintenance of 6 months." After 3 months, sludge accumulated at
the cathode and dismantling of the sensor was required to clean
the sludge and obtain satisfactory operation.
2. The "0" ring must be sealed exactly in the sensor-head groove to
prevent water seepage into the electronics.
3. The recessed stem-nut requires lubrication to be removed to permit
sensor maintenance.
4. Temperature compensation potentiometers located in the sensor head
prevent an exacting temperature compensation calibration.
5. Long-term transient response (2-24 hours) is dependent upon
initial and final conditions.
6. Thermistors in the sensor head must be verified for proper posi-
tioning, and cable connection between the MV/V and different sensors
must be identified in accordance to the Honeywell catalog.
7. Manual cleaning of the. sensor caused 60 percent full scale spikes
to occur which lasted several hours. Spikes did not occur when
sensor maintenance was performed in a low pH (2.5) solution with
minimal mechanical wiping.
8. Mechanical mounting of the sensor in the aeration basin determines
the amount of air bubbles trapped on the sensor membrane. Near
horizontal mounting of the sensor i-s preferred.
The sensors performed within the specifications for transient
response, and were within tolerance for the temperation compen-
^ rt ^ -1 >^*\ ^ S* ^ ^ ^
9-
sation tests
-------
10. With daily maintenance, sensor performance in activated sludge
was acceptable—the unbiased standard deviation between the
Winkler determination and the Honeywell reading was 0.256.
11. The sensors can be operated without maintenance for over a week
in a chlorinated effluent.
MV/V*
1. The MV/V displayed an apparent shift of 40.3 to 40.9 percent at
35F (1.7C), and 41.2 to 42.9 percent at 65F (18.3C) or ±1.3
percent between the worst conditions at each temperature before
the chamber overheated.
2. The apparent shift after the chamber temperature excursion was
42.2 to 42.5 percent at 35F (1.7C), 42.6 to 43.1 percent at 65F
(18.3C), and 43.5 to 44.0 percent at 95F (34C). The apparent
shift between the worst cases for the total temperature range was
±0.9 percent, which verified Honeywell's opinion that the 200F
(93.4C) temperature in the chamber had not damaged the MV/V.
3. The independent linearity of the panel meter for the low and high
ranges was a nominal ±1.0 percent and ±2.0 percent, respectively.
4. The panel meter display could not be easily seen from several
feet away which meant it could not be mounted on a wall where
physical contact could not be made. Panel meter indications are
on the plastic cover plate, a position that increases the proba-
bility of error due to parallax.
5. Because of mechanical arrangement and location, the potentio-
meters on the range card could not be easily adjusted,.
6. Parametric system performance in the final settling basin at the
treatment plant was not satisfactory because there was no flow
past the sensor. However, manual movement of the sensor produced
performance comparable to other instrumental measurements of DO.
7. The sensor reduction potential was stable within 0.25 percent
over a 2-week period and the MV/V response was in tolerance for
step changes, and for filtering noise. The multivibrator 200 Hz
frequency chosen was optimum in comparison to 100 Hz and 400 Hz.
*Millivolt to voltage converter function.
-------
Although overall linearity was within specification (±2 percent),
it varied by 0.5 percent between laboratory and field work. Per-
formance employing different cable lengths met the manufacturer's
specifications. And lastly, attenuation of power frequency noise
fully met the manufacturer's specifications.
MANUALS AND SERVICE
The Honeywell manuals met or exceeded Section 5.00 of EPA's specifi-
cations for an integrated system and contained a summary of technical
information to enable a broad spectrum of users to effectively employ
the product. The local and main offices* of Honeywell were available
for either hardware or technical explanation as the need occurred.
"Honeywell, Fort Washington, Philadelphia, Pennsylvania.
-------
SECTION III
RECOMMENDATIONS
The recommendations summarized are directed to three product areas:
A. The manufacturer should:
1. Provide the customer with two options: automatic sensor
cleaning and a small submersible pump to maintain sample velocity
(parallel) over the membrane. Both devices would be attachable
to the submerged sensor.
2. If the sensor must be cleaned manually, define the exact
method to reduce "spikes" following maintenance.
3. Relocate the temperature compensating potentiometer in the
MV/V. If this cannot be done, simplified temperature calibra-
tion procedures for submerged sensor operation must be developed.
4. Clarify the notations on the circuit diagrams so that there
is only one subscripted variable, such as TP-1, per system.
5. Reconsider the overall design since the same measurement can
be effected with fewer electronic components.
6. Simplify the page numbering system in the manual.
7. Spare "0" rings and sealant material for the sensor head
should be provided.
8. Define methods of installation unique to the application and
employ drawings or their equivalent.
9. Locate meter markings on the panel face rather than on the
cover plate or provide a mirror at the back to eliminate parallax.
A digital readout would be helpful.
10. Determine the linearity of the MV/V at 35, 65, and 95F (1.8,
18.3, 35C) and publish the data in the manual.
11. Provide a screwdriver in the kit to adjust the MV/V potenti-
ometers .
12. Reevaluate the manuals and correct small errors without losing
the tutorial characteristics.
-------
B. The prospective user should:
1. Define basic requirements for DO measurement so that Honeywell
can provide appropriate system functions.
2. Study the manuals thoroughly before attempting to use system.
3. Random errors occur in both Winkler and instrument determina-
tions, so that an exact correlation between the two methods is
not to be expected.
4. Establish a strong communication link with the manufacturer.
This will save money over a period of time.
5. Expect to achieve results similar to those summarized in this
document.
6. Expect periodic manual maintenance although automatic cleaning
is also provided.
C. The final user should:
1. Develop a well-disciplined maintenance program based on the
manufacturer's recommendations. If the system is comparable to
the one investigated and the sensor is installed in activated
sludge, daily maintenance is required. Weekly maintenance is
recommended if it is placed in a chlorinated effluent. More
frequent maintenance may be required if the sensor is installed
in an effluent raceway (if untreated samples are diverted there).
2. Keep logs on application, performance, method of installation
(contact the manufacturer), flow, liquor constituents, air volume,
and other appropriate parameters. Inform the manufacturer of
measurement and sample characteristics; this information can be
considered when new models are being designed.
3. Follow the manufacturer's written recommendations exactly.
If the system fails to provide the performance expected, contact
the manufacturer.
4. If auxiliary flow is provided, discuss with the manufacturer
the proper flow vector in relation to the membrane because the
angle of flow directed at the membrane depends on the physical
design (Honeywell recommends a 45° angle).
5. Determine the presence of air bubbles on sensor. These cause
spikes in output and can cause incorrect control (see No. 2 under
Recommendations for Manufacturer).
-------
6. Analyze probable fault before making potentiometric adjust-
ments .
7. Consider having the manufacturer carry out maintenance
operations under contract if manpower is in short supply.
8. Install the MV/V in a pump house to protect it from the
weather and calibrate the unit monthly.
-------
SECTION IV
OVERVIEW OF THE HONEYWELL DO PARAMETRIC SYSTEM
The Honeywell parametric system consists of an in situ sensor
assembly, 50 feet (15 m) of four-conductor-shielded cabling, and a
MV/V transmitter.
SENSOR ASSEMBLY
Honeywell's sensor design philosophy permits a. general-purpose MV/V
to be used because the sensor assembly produces a mv potential signal
that is proportional to the DO concentration. The sensor assembly has
two main functional components, the DO detector and the temperature-
compensating network.^>* The components are listed in Table 1.
Table 1. COMPONENTS OF DO SENSOR
Cathode
Anode
A/C ratio
Body
Thermistor
Membrane
Electrolyte
Reduction potential
Temperature-
compensating poten-
tiometer
Other
(RQ § 65F
1/8" diameter: silver-
platinum-gold
Silver
400, nominal
PVC
Exposed to sample
500 n)
1 ml teflon
0.5 N KOH + 2.0 N KC1
800 mv, fixed
Adjustable
Discussed in patent.
A wiring diagram of the sensor assembly is shown in Figure 1, and a
simplified schematic is shown in Figure 2.
-------
TJ
IE
tn
S
)
o
CO
CO
^^ ••
I
1—
Female
Connector
/\ / A YD \ A
•§
or
CJ
IB •••
1 I VucuWr.n /* 1 1
&»*1U
^x
[
Male
Connector
1
1
"1
!
1
.C
g
~
-
Re
ti
u
OJ
CO
- -
id
roM
"8
cc
CJ
Blac
east
2
S
Q
! n
' 1 i
Electronic
0.8V D-C
Power
Supply
k
iring Instrument
Temperature
Compensator
I Potentiometer
FIGURE 1. SUBMERSIBLE TYPE DO SENSOR, WIRING DIAGRAM WITH
EXTERNAL CIRCUITS IN MEASURING INSTRUMENT.
°B
FIGURE 2. SIMPLIFIED SCHEMATIC OF THE HONEYWELL SENSUR'S
CIRCUITRY.
-------
If it is assumed that the impedance connected to terminals B and A
is much greater than Rpot + ^th f°r a^ temperatures, the steady-
state output potential is:
vo = i CDO,TD x (Rpot + Rth (T)) (i)
where V = sensor assembly signal, volts
i(DO,T) = current proportional to DO concentration and tempera-
ture, amperes
R = trim potentiometer, ohms
R , (T) = temperature-sensitive thermistor, ohms
T = degrees Kelvin
For fresh water and considering the temperature dependency noted by
equations B-7 and B-8 in Appendix B equation 2 is obtained.
-£ / +!\
T T I
V = iKe x\R . + R e / (2)
o \ pot o / *• J
where K, J, and B = constants of proportionality
R = reference resistance, ohms
Equation 2 can be rewritten:
-J B-J
Vo = iKRpote T + iKRoe T C3)
10
-------
It appears, therefore, that the Honeywell sensor* is not completely
temperature compensated, since, assuming B and J are equal:
Vo = iKRpote + 1KRo (4)
However, equation 2 is theoretically ideal and Honeywell (as well as
most other manufacturers) follows R. Bates '3 philosophy of an opera-
tional definition employed in pH determinations: the total parametric
system, including temperature compensation, is calibrated in accordance
with the Winkler process"1" and at saturated conditions. Thus, if the
sensor is immersed in a comparable sample it produces an output con-
sidered to be proportional to the DO concentration, within acceptable
tolerances .
EPA's long-term practice has been to temperature compensate and cali-
brate the parametric system under fresh water "saturated" conditions
at 35, 65, and 95F (1-7, 18.4, 35C) . (The results will be discussed
later.) In the field, trim on the gain control would be made follow-
ing comparison of the Winkler and instrument determinations.
By schematically interchanging the temperature compensator and
potentiometer, it would be possible to relocate the potentiometer
within the analyzer thereby minimizing manhours involved in tempera-
ture compensation. Since the analyzer is for general purposes, it
may be necessary to intercept the black wire from Terminal A in
Figure 1 and insert two separate plug-in posts, providing means for
potentiometer insertion or a short circuit. (Honeywell uses a
similar DO sensor design in its tank-type systems.)
MV/V TRANSMITTER
The MV/V converts a source mv potential into an output potential in
the 0-5 VDC range. Table 2 lists some of the components displayed
in Figure 3.
*When R0»Rpot» tne first term in equation 4 becomes negligible (within
sensor tolerance) so that the output potential is effectively indepen-
dent of temperature. And, additionally, for temperatures of interest,
J»T.
tThe process involves wet chemical techniques that culminate in the
titration of free iodine. See Appendix A for further discussion.
11
-------
Table 2. MV/V FUNCTIONAL COMPONENTS
1. Power distribution 5. Auxiliary power
2. Sensor power supply supply
(800 mv) 6. Measuring circuit
3. Power supply 7. Amplifier assembly
analyzer 8. Isolation module
4. Control/alarm module 9. Output function
1. (Refer to Manual).
2. Sensor power supply - This two-stage, zener-regulated DC supply
provides a fixed 800 mv potential to the sensor. Conveniently
located test points TP^ and T?2 are used to verify the reduction
potential. As noted by Gealt and Metarko, a slight change in the
reduction potential has little or no effect on sensor output,
therefore no trim pots are included.
3. (Refer to Manual).
4. Control/alarm - This component was damaged in shipment, but since
the authors had not planned to investigate it, the component was
left unchanged. The alarm circuitry provides (1) an on - off
controller function; (2) a separate Hi - Low alarm, and (3) a
combination of (1) and (2). EPA specifications provide for (2)
within its Sample Taker parametric system but not for (I).5
These features could be employed to control aeration in sewage
treatment plants.
5. (Refer to Manual).
6. Measuring circuit - This function consists of potentiometers and
divider networks having offset and gain capability. Additionally,
the measuring circuit, which is connected to the negative terminal
of the signal source, can be provided with input suppression and
thermocouple compensation when required. The measuring circuit
is provided with a 5-volt supply, whose temperature-stability data
are discussed later. It is apparent that the MV/V has greater
flexibility than a standard DO-type microammeter available in
other designs. Thus, the available components should be matched
with the engineering application to obtain maximum benefits.
7. Amplifier assembly - This is the main component in the MV/V, and
it consists of a filter section, a solid state modulator-
demodulator driven by a free running multivibrator, an AC amplifier,
12
-------
INPUT
(See enclosed
wiring diagrams
for specific terminal
connections)
0-5 VDC ON-ISOLATED
* T1 is omitted and W11 and W12 added only when used on 120V
primary and NO isolator or alarm-control circuit.
CODE
i These components are not used with 0-5 VDC, non-isolated, output feature.
• These components are not used with 0-5 VDC, isolated, output feature.
• These components are not used with 4-20 MA DC, isoJated, output feature.
**R5 is used as noted, but with 0-5 VDC, isolated, output feature a jumper wire, W22
(W22. not shown), is used in place of R5.
0-50 mv for voltage output feature
10-50 mv for current output feature
I 1
550-142
FIGURE 3. FUNCTIONAL LAYOUT OF WIV/V.
-------
and an integrating operational amplifier. The output drops the
AC amplifier input to zero through the measuring circuits so
that any error developed by a change in DC input causes a change
in the potential of the integrating capacitor, hence change in
output.
Sixty Hz noise has been a periodic problem in this type of instru-
mentation, but the following analysis shows that most of it is
attenuated by the filter section.
v
I ^V
- v
) ^J
i
N
'no
FIGURE 4. SIMPLIFIED DIAGRAM OF MV/V FILTER NETWORK.
14
-------
Equation 5 and 6 are derived from Figure 4 as follows
- 12
(R2
where R = R = 10K, ohms
Cj = C2 = 4.7 yfarads
i1 and i» = instantaneous loop currents, amperes
It is assumed that en is a 60-Hz sinusoidal noise and that the
load impedance approaches an op'en circuit.
Employing RMS notation, steady-state conditions are:
En ' ll "4 * - 1
0 ' -1 f-' * '2 »2 * - * ' C8)
Equations 7 and 8 yield the current through £2, and the approximate
steady-state output voltage, EOn, is then given by:
E.
E =1 . = - - - (9)
°n 2 Ja)C2 1 - (RajQ2 + j3Ra)C
An assumed noise of 3 mv yields a filter output of approximately
9.6 microvolts. Thus, if noise appears in a recording function,
it is probably caused by output shielding rather than the input
filter- -presuming that all filter components are performing within
tolerance. A plot of equation 9 is shown in Figure 5.
The amplifier, which is equipped with a panel meter display, can
provide either current or voltage outputs, depending on bid speci-
fications and application. The MV/V investigated provided '0-50 mv
and 0-5 VDC outputs.
15
-------
CO
cd
—
CJ
HI
Q
oc —.
LU
(3
O
O
-------
Simply stated, the amplifier assembly can be thought of as a
gain/offset device described by equation 10:
DO = KG(jco)V. + MVo (10)
where DO = dissolved oxygen in mg/1 (output)
K,G,m = constants of proportionality
V. = input potential, mv
V = offset potential, mv
ju) = complex notation, variable frequency
For the DC steady-state condition, equation 10 reduces to:
DO = K2Vt + mVO (11)
where K7 = constant of proportionality.
Hence, the amplifier is a two-control device having span
and offset (m) potentiometers (Figure 6).
8. (Refer to Manual).
9. (Refer to Manual).
17
-------
INPUT TERMINALS
ON MOTHER BOARD
+ INPUT
00
CONTROL AND/OR
ALARM MODULE
Assy. 30679448-TAB
Schematic 30703145
.Sensor Power Supply
Assy. 30677598-002
Schematic 30703230
Power Supply Assy.
i Assy. 30677596-001
j Schematic 30703155 TP7
5-I1OTp1 | R30
30V DCCW
Meter
mpeden
Matching
Circuit
SIGNAL
COM.
• SI Meter Signal
js-isr-:
[Isolator
I Assy, 30677936-001
Amplifier
! Assy .30677592-001
j Schematic 30703168
! Schematic 30703266
• A.C. D.C.
AMPL
A.C. AMPL. AMPLIFIER !
p Meter"dtust (wTtn
i without suppression)
I Assy. 3067941S-TAB
I Schematic 30703369
l B |—» +--,
AND DRIVER SUPPLY
SUPPLY <
Measuring Circuit
Assy.30677592-TAB
Schematic 30703181
FIGURE 6. TOTAL FUNCTIONAL CIRCUIT OF MV/V.
DESCRIPTION
Soldered wire connection to P/W Board
Test Point
Screw Terminal
Multi-contact Connector
Single Contact Connector
KEY
-O-
0-
O
-«-
-&
-------
SECTION V
LABORATORY INVESTIGATION
To accommodate EMSL's test procedures, Honeywell field changed the
MV/V range card from 0-5 and 0-10 mg/1 DO to 0-12 and 0-24 mg/1 on
the low and high ranges, respectively, a modification that required
a two-point recalibration. The panel meter markings were left
unchanged, therefore 10 corresponds to 100 percent full scale for
both ranges. The data for this recalibration are shown in Table 3.
Data between these limits for linearity are shown further on.
Table 3. ROOM TEMPERATURE ZERO AND SPAN
CALIBRATION FOR THE MV/Va
T T l
Input (10 volts) Range PM Output (10 volts)
0
0
6.6
13.2
low
high
low
high
0
0
10
10
10
10
50
50
o
Honeywell recommends that the sensor assembly and the MV/V
be calibrated as an integrated unit. For this test and
several others, it was convenient to work with each sepa-
,rately.
Measured between TP-1 and TP-2.
SENSOR ASSEMBLY INVESTIGATION
Two sensor assemblies were investigated for their temperature-
compensation and short- and long-transient responses. The MV/V was
employed in temperature-compensation tests. Two assemblies were
investigated due to failure noted below. Three Forma baths similar
to the one shown in Figure 7 were used. Each contained 7.5 liters of
distilled water that was adjusted to 41F, 65F, and 95F (5C, 18.3C,
35C) and held constant within l.OF (0.5C). Each bath was aerated
with atmospheric gases, and diffusers provided a reasonably consistent
distribution.
Several test procedures were explored before satisfactory data were
obtained. The first employed the polyethylene bottle shown in Figure
8 that was modified several times, such as adding holes to the top
19
-------
permitting escape of trapped gases. The second modification employed
a horizontal aluminum frame but slight dimensional differences in the
Forma baths allowed movement of the frame and change in relative posi-
tion of the sensor tip with respect to the sample stream. Transient
and long-term stability data for sensor 2 were obtained with the last
and most favorable modification shown in Figure 9. Each Forma bath
was marked for the precise location of the supporting pipe so that
regardless of the bath employed, the sensor tip was always in the
main stream line of the sample flow with the stream line 10° below
membrane surface.
The sample velocity from each flow nozzle, as measured by a March-
McBirney velocity meter Model 201, was 1.8 feet per second at cold,
room, and high temperatures.
Sensor 1
One problem encountered in designing DO parametric systems is providing
an inexpensive and conveniently located temperature-compensation cir-
cuit to correct for the membrane permeability factor illustrated by
equations 1-4.
The temperature-compensation technique was investigated by adjusting
the zero and span controls in the MV/V and the potentiometer Rpot i-n
the sensor assembly. Data taken following initial experimentation
and after final potentiometric adjustments are shown in Table 4.
Table 4. CALIBRATION DATA OF THE HONEYWELL DO PARAMETRIC
SYSTEM EMPLOYING SENSOR 1
Temperature
C
6
19
38
Winkler DO
determination
mg/1
12.50
9.40
6.80
Output-high range
indicated DO mg/1
before after
12.6 12.60
9.5 9 . 30
6.7 6.60
Temperature compensation was difficult to achieve, because after each
adjustment the sensor head had to be replaced and validity of the test
setup had to be reconfirmed. This occurred several times for each
measurement.
20
-------
io
-------
FIGURE 8. FIRST SENSOR ASSEMBLY CONFIGURATION (THE POLYETHYLENE
BOTTLE WAS LATER MODIFIED).
-------
FIGURE 9. PIPE SUPPORT SYSTEM FOR SENSOR ASSEMBLY.
23
-------
The deviation between instrument data and the Winkler DO determination
was less than 1.0 percent of full scale for both the "before" and
"after" tests. (Honeywell calibrated the sensor when it was partially
submerged, but the data in Table 4 were obtained while the sensor was
fully submerged.)
Transient Response Procedure
Transient response data magnify any quirk in (1) design, (2) manufac-
ture, or (3) the investigation process. Transient data accumulated
on the Honeywell sensor assembly reflected the influence of all three:
(1) the trim pot should be re-located into the MV/V; (2) sensor 1 failed,
and the thermistor was not properly installed in sensor 2--it rotated
during the preliminary tests; (3) tests were repeated several times
because of unsatisfactory conditions, such as the flow moving the sensor
and the improper alignment of the assembly over the flow. The latter
yielded a sinuosoidal superimposed on an exponential that invalidated
the response. The latter was eventually corrected by the piping
arrangement shown in Figure 9.
Transient responses to temperature and DO concentration were obtained
by rapidly transferring the sensor assembly from one bath to another.
The procedures listed in Table 5 were followed in investigating both
sensor assemblies.
Table 5. TRANSIENT RESPONSE PROCEDURES
Sensor in bath Winkler taken from
RT (initial)
RT CT before transfer
CT RT after "
CT RT before "
RT CT after "
RT HT before "
HT RT after "
HT RT before "
RT HT after "
RT (final)
CT = cold temperature = 5C (varied between 1.7C and 5C,
and noted, due to ice formation on the refrigerant
coils).
RT = room temperature = 18C.
HT = hot temperature = 35C.
Sensor transfer = RT->CT-»RT-»HT-»RT
24
-------
LU
Q
RT * 18.3°C
65.(fF
LU
CC
CT = 5.0°C
41.0°F
I
WINKLER= 12.20
2 4
ELAPSED TIME (MINUTES)
10
LU
RT = 18.3°C
65.0°F
I
HT
= 35.0°C
95.0°F
234
ELAPSED TIME (MINUTES)
WINKLER-6.80
•30=-
10
O
LU
LU
DC
HT=35.0°C
r 95.0°F
7
RT=18.3°C
65.0°F
WINKLER=9.30
i—ff-
1
ELAPSED TIME (MINUTES)
FIGURE 10. SHORT-TERM RESPONSE OF SENSOR 1.
10
25
-------
Results of Transient Response Investigation - The short-response
(10-min) capability of sensor 1 is displayed in Figure 10. Slight
irregularities were caused by imperfect test facilities, but the
output appeared to be within Honeywell's tolerance at 2 minutes as
related to the 10-minute value. Data were collected on an Esterline
Angus Speed Servo recorder operated at 0.75 inch per minute chart
speed.
The response curves are comparable to those recorded during previous
investigations, except for the top graph in Figure 10; the overshoot
was much less than that observed for other sensor designs (see dis-
cussion on Sensor 2).
Failure - Long-term (>60 min) responses were attempted but were unsat-
isfactory because of: (1) changes in the sensor's orientation; (2)
recorder inking problems; and (3) a "noisy" sensor.
Problems (1) and (2) were resolved in-house, but (3) could not be
corrected, and a replacement sensor was provided.
Figure 11A illustrates the cathode before the sensor was placed in
operation. Figures 11B and 11C show its condition after several
tests over a 2-week period.
Following discussions with Honeywell,^ the pH of the electrolyte was
determined to be 10, but the discoloration of the cathode was of a
sludge nature and not a "plating-out" phenomenon. To accumulate more
residue at the cathode, the sensor was energized by 80 mv and im-
mersed in a saturated, distilled-water sample at 95C (35C). After
2 weeks, the cathode appeared as shown in Figure 11D; once again, the
discoloration was of a sludge nature, but slightly more had accumu-
lated. It appeared that either the screen described in Gealt's patent
did not function as expected or that it had been improperly installed
(no work was performed to determine the cause of the accumulation,
presumably AgCl).
(In subsequent discussions, Gealt said that several sensors, including
the one tested, were manufactured employing a different assembly
technique--only one end was supported during assembly, and this had
caused the screen and cathode element to be slightly out of line. As
a result, more frequent (within 12-15 weeks) service was required.
He indicated that a revised assembly process was being employed.)
Photographs of sensors 1 and 2 appear in Figure 12.
26
-------
A. SENSOR CATHODE AS RECEIVED
1/8" i
GOLD COLOR
3/32"
HAS UNDERGONE ONLY FACTORY CALIBRATION
B. SENSOR CATHODE AFTER TRANSIENTS
MOTTLED
SILVER-GOLD
COLOR
LIGHT BLACK
COLOR
TO THERMISTOR
HAS UNDERGONE LABORATORY
CALIBRATION, TEMPERATURE COMPENSATION,
AND TRANSIENT RESPONSES
C. SENSOR CATHODE 2 WEEKS AFTER
EXPOSURE TO 95°F(35°C)
D. SENSOR CATHODE 4 WEEKS AFTER
PROLONGED EXPOSURE TO 95°F(35°C)
LIGHT
BLACK
COLOR
MOTTLED
SILVER-GOLD
COLOR
DEEP BLACK
- COLOR
MOTTLED
SILVER-GOLD
COLOR
LIGHT
BLACK
COLOR
DEEP BLACK
COLOR
FLAKY BLACK
PATCHES
TO THERMISTOR
DISTILLED H20 BATH, D.O. SATURATED,
AT 95°F(35*fc)
FINAL APPEARANCE AT 6/2Q/74
FIGURE 11. SLUDGE ACCUMULATION ON CATHODE OF SENSOR 1.
27
-------
S3
OD
FIGURE 12. PHOTOGRAPHS OF SENSORS 1 AND 2.
-------
Sensor 2
Sensor 2 was subjected to similar tests but their duration and
sequence differed:
1. Laboratory (initial calibration of parametric system).
2. Field investigation (parametric system).
3. Service on sensor (sensor assembly).
4. Field investigation (parametric system).
5. Laboratory (sensor assembly only):
(a) 10- and 60-minute response and 24-hour stability
(b) service on sensor
(c) with and without pressure diaphragm.
Items 2, 3, and 4 are discussed later.
Laboratory (Initial Calibration of Parametric System) - Saturated
samples maintained at a constant temperature, as described earlier,
were used. All available controls (Rp0t> zero, and span) were
employed, and the MV/V panel meter reading was recorded. The sensor
assembly was maintained in each sample for a nominal 24 hours. The
results are summarized in Table 6.
Table 6. CALIBRATION DATA FOR THE HONEYWELL DO
PARAMETRIC SYSTEM EMPLOYING SENSOR 2
(STEADY STATE, 'AFTER 24 HOURS)
Temperature, C Winkler DO, mg/1 PM-Hi range DO, mg/1
CT
RT
HT.
4.9
18.3
35.0
12.75
9.45
6.95
12.60
9.41
6.89
Data in Table 6 are within EPA's tolerance of ±0.24 mg/1.
Output from the sensor assembly was recorded on an Esterline Angus
recorder, and the steady-state values are illustrated in Figure 13.
Laboratory (Sensor Assembly Only) - Following the field work, the
sensor assembly was stored in the laboratory for several weeks during
which time the stability of the 800 mv reduction supply was investi-
gated. This potential "lies on the center" of the polarographic
29
-------
60.0
50.0
- 40.0
LU
LU
DC
D 30.0
S5
Q
DC
O
(J
£ 20.0
4.9°C (FROM 18.3°C)
18.3°C (FROM 4.9°C)
18.3°C (FROM 35.0°C)
35.0°C (FROM 18.3°C)
10.0
NOTE A)FULLSCALE=24Mg/L
B) 50% RELATIVE OUTPUT = 12 Mg/L
I
3.0 6.0 9.0 12.0 14
D.O. IN Mg/L
FIGURE 13. TEMPERATURE COMPENSATION OF SENSOR ASSEMBLY.
30
-------
plateau, which reduces the effects caused by slight voltage
fluctuations. Honeywell does not, therefore, provide potentio-
metric adjustments. Figure 14 illustrates the stability of the
800 mv supply.
The response and the 24-hour stability of the sensor assembly are
illustrated in Figures 15-17 for the RT-H]T, CT+RT, and RT->HT modes.
Inspection of Figures 15, 16, and 17 reveals that at 2 minutes the
sensor output was within 98 percent of the output recorded at
approximately 1 hour and that in all cases the sensor output was
within Honeywell's tolerances at 24 hours.
An apparent maximum overshoot approached 3 percent at 10 minutes for
the RT->CT transient, and those for the CT->RT and RT+HT modes approxi-
mated 1 percent. The CT DO level decreased 1 percent during the last
23 hours of the stability run, therefore the overall overshoot for
the RT-»CT test may have been less than 3 percent.
The temperature and the DO concentration at room temperature varied
and caused a slight change in the output. The differences between
the Winkler DO determination and the Honeywell measurement were less
than 1 percent.
All three response curves indicate, therefore, that the thermal
response was within Honeywell's tolerances and met the 24-hour stabi-
lity criterion.
The initial decrease in Figure 15 is caused by the teflon as the
diffusion of oxygen through it is more rapidly affected by tempera-
ture than the compensating resistance of the thermistor. Treating
the sensor as a constant-current generator for given conditions,
the output voltage across the thermistor-potentiometer circuit
increased, as displayed by the graph.
Regardless of the final DO level, several tests indicated that the
RT-HTT response, although within 98 percent of steady-state at 2
minutes, did not reach 100 percent steady-state for several hours.
This was thought to be related to the total sensor heat capacity (PVC,
electrolyte) as well as to possible thermal affects of the teflon. No
work was performed to isolate and study these possibilities.
The response curves in Figures 16 and 17 are similar since in both
cases the sensor assembly was transferred to a warmer sample. The
initial positive pulse apparently occurred because the thermal
response of the membrane resulted in greater diffusion of DO to the
31
-------
K)
1.0
0.9
0.8
0.7
5
> 0.6
z
£ 0.5
Q_
3
85 0.4
1
C/3
0.2
0.1
0
I
0.5
I
I
16.5
1.0 16
TIME (HOURS)
FIGURE 14. ROOM-TEMPERATURE STABILITY OF SENSOR REDUCTION POTENTIAL.
-------
0'
CT=4.80°C
13.05 PPM
CHART SPEED CHANGED
tt
CT=4.80°C
12.80 PPM
\
1 2
'TIME IN MINUTES
22 23
TIME IN HOURS
ELAPSED TIME
FIGURE 15. TWENTY-FOUR TOTAL SENSOR RESPONSE (RT-CT).
-------
100
O-l
CT=4.80°C
12.80 PPM
RT=18.20X
9.60 PPM
CHART SPEED CHANGED
i
4-
1 2
- TIME IN MINUTES
RT=18.20*C
9.60 PPM
•4-
4-
21 22
TIME IN HOURS -
ELAPSED TIME
FIGURE 16. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (CT-RT).
23
24
-------
CM
C/l
80- RT=18.10°C
9.20 PPM
HT=35.00°C
6.60 PPM
CHART SPEED CHANGED
4-
TIME IN MINUTES
10
-»
t-tf
4-
HT=35.00°C
6.60 PPM
4-
21 22
TIME IN HOURS •
23
ELAPSED TIME
FIGURE 17. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (RT-HT).
4-
24
-*1
-------
cathode and generated a higher output current. As the thermistor
approached the higher temperature, its reduced resistance produced
a lower output potential from the DO sensor.
During the transient-response tests, the sensor performed erratically
when the HT->RT run was attempted, which indicated that it needed to
be reconditioned. This proved impossible to do, however, because
the retaining nut on the stem of the cathode could not be loosened
and the pressure diaphragm in the electrolyte was not removed out of
fear of tearing it. However, the electrolyte was exchanged and much
of the accumulation drained from the sensor. Contrast between the
new and old electrolyte can be seen in Figure 18.
The diaphragm was tested to determine its impedance by employing
electrolyte on both sides. The results was satisfactory—the imped-
ance was 10 megohms.
After exchange of electrolyte and membrane, the output varied several
percent for similar conditions and is consistent with the variations
suggested by Figure A-l in the Appendices.
After the sensor was partially reconditioned, an acceptable RT->HT
response was obtained but the reverse run was unsatisfactory because
the "0" ring had been improperly seated and water leaked into the
electronic compartment of the sensor head (Figure 19).
Uncompensated Sensor - Honeywell provided a replacement plug without
diaphragm so that the investigation of the sensor's components could
be pursued.
Thus, following sensor maintenance (clean cathode, new electrolyte,
and new membrane), response data for an uncompensated sensor were
obtained. The thermistor was replaced with a fixed resistor whose
low-temperature coefficient had previously been verified.
The temperature-compensating potentiometer was bypassed, therefore
only the electrochemical and mechanical components were involved.
Procedures outlined in Table 5 were followed, and transient data
for 10 and 60 minutes and for 24 hours were obtained. Segments of
these continuous recordings are shown in Figures 20-23.
The graphs indicate that response times were longest when the sensor
was transferred from lower to higher temperatures (note values at
break in graph). Table 7 summarizes several pertinent response
points. (Note that the output failed to return to initial condi-
tions .)
36
-------
FIGURE 18. NEW AND OLD ELECTROLYTES.
-------
FIGURE 19. SENSOR HEAD.
-
-------
o
IU
9-
I 8'
_j
<
5 7-
LJJ
2
i— 6*
O 5-
o 4-
cc
a
LU
H 3-
m
S 2-
0
, j j , . ! , ^ , , _
• •
• •
RT
W=9.3 PPM
•/
•
CT
W=13.5 PPM
\
-------
o
CT
W=13.6 PPM
\
CHART SPEED CHANGED
+
RT
W=9.4 PPM
-4-
2 ' ' 9
TIME IN MINUTES
10
2))
-TIME IN HOURS
ELAPSED TIME
FIGURE 21. TWENTY-FOUR HOUR MEMBRANE RESPONSE (CT-RT).
23
24
H
-------
to
CHART SPEED CHANGED
4-
4-
HT
W=6.7 PPM
2 > J 9
TIME IN MINUTES
10
1 2
TIME IN HOURS
23
ELAPSED TIME
FIGURE 22. TWENTY-FOUR HOUR MEMBRANE RESPONSE (RT-HT).
24
-------
1-0
10
7- -
6- •
O
Q_
h-
D_
8 54-
i-
D
LU
z
S 3--
LU
2- •
1 --
W =6.9 PPM
CHART SPEED CHANGED
10
-TIME IN MINUTES'
1
RT
W=9.4 PPM
-TIME IN HOURS-
ELAPSED TIME
FIGURE 23. TWENTY-FOUR HOUR MEMBRANE RESPONSE (HT-RT).
23
24
-------
Table 7. SELECTED RESPONSE POINTS FOR MEMBRANE ONLY
Transfer
RT-K:T
CT->RT
RT+HT
HT+RT
RT = 18. 3C
CT = 1.7C
HT = 35. OC
W = Winkler
Minutes
0 2
Amplitude, %
52.0 33.5
32.0 50.0
52.0 85.0
88.0 58.0
WRT = 9.3
WCT =13.5
WHT = 6.8
10
»
33.7
50.8
86.0
52.5
Hours
2
Amplitude,
32.0 31
53.0 52
88.0 87
54.0 53
24
%
.8
.0
.5
.0
determination of dissolved oxygen.
Thermal Compensator - Response of the thermal compensator was inves-
tigated by replacing the sensor with an 8K resistor mounted in the
MV/V, and measuring the potential developed across the output network
as a function of temperature and time. Response and long-term stabi-
lity are displayed in Figures 24-27.
Table 8 summarizes the steady-state thermistor resistance for the
test temperatures noted.
Table 8. THERMAL COMPENSATOR RESISTANCE
Temperature (C)
5.0
18.6
35.0
Resistance (K)
2.10
0.94
0.39
In concluding the sensor investigations, the cold-bath flow system
was replaced because excessive overshoots continued to be recorded.
Figure 28 reflects the values obtained after this was done. The
data displayed in Figures 29-31 were recorded before the plumbing
was corrected but in all cases, the flow was approximately of the
same magnitude. The slight overshoots detected were apparently
caused by the test set up and the main stream line of flow with
respect to the assembly. Regardless, the output was within 98 per-
cent of steady state at 2 minutes, as noted previously.
43
-------
50
CHART SPEED CHANGED
ELAPSED TIME
FIGURE 24. TWENTY-FOUR HOUR THERMISTOR ONLY RESPONSE (RT-CT).
-------
Cn
CHART SPEED CHANGED
10
TIME IN MINUTES-
RT=18.3"C
\
4-
2 ' ' 23
-TIME IN HOURS
ELAPSED TIME
FIGURE 25. TWENTY-FOUR HOUR THERMISTOR ONLY RESPONSE (CT-RT).
24
-------
50
45.-
40 -
35--
t 30-- RT=18.3°C
25--
8
u
cc
|
|
15- -
10- •
5- •
1
•TIME IN MINUTES-
CHART SPEED CHANGED
1
HT=35.0°C
1
tff
+
2 ' 23 24
-TIME IN HOURS »••{
ELAPSED TIME
FIGURE 26. TWENTY-FOUR HOUR THERMISTOR RESPONSE (RT-HT).
-------
50
CHART SPEED CHANGED
ELAPSED TIME
FIGURE 27. TWENTY-FOUR HOUR THERMISTOR ONLY RESPONSE (HT-RT).
-------
C»
10
9- •
8--
3! 7.-
2
i—
i
6- •
5-.
D 4
o
cc
o
s 3
CO
LU
to
2- •
W=9.4 PPM
RT=18.3t
CHART SPEED CHANGED
10
•TIME IN MINUTES.
W = 12.5 PPM
CT=3.1°C
23
-TIME IN HOURS-
ELAPSED TIME
FIGURE 28. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (RT-CT).
24
-------
10
CHART SPEED CHANGED
ELAPSED TIME
FIGURE 29. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (CT-RT).
-------
U1
o
CHART SPEED CHANGED
10
1
W=6.8 PPM
HT=35.0*C
23
TIME IN MINUTES'
-TIME IN HOURS'
24
•H
ELAPSED TIME
FIGURE 30. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (RT-HT).
-------
Ui
W=6.8 PPM
HT=35.0°C
4-
4-
1 2
TIME IN MINUTES-
T
CHART SPEED CHANGED
W= 9.4 PPM
RT=i18.3°C
TO
Hi-
1 2 ' ' 23
'TIME IN HOURS-
ELAPSED TIME
FIGURE 31. TWENTY-FOUR HOUR TOTAL SENSOR RESPONSE (HT-RT).
24
H
-------
Table 9 summarizes Honeywell's specifications for the DO sensor
assembly.
Table 9. MANUFACTURER'S SPECIFICATIONS FOR SENSOR ASSEMBLY
Input range -
Output -
Output impedance
Accuracy
(reference) -
Response time -
Long-term -
(30 days)
Normal operating
conditions -
Operative limits -
Material specifi-
cations -
0-2 to 0-25 mg/1.
Nominal 0.55 mv/mg/1 (±5%).
3200 ohms maximum at OC.
190 ohms minimum at 45C.
Within 0.1 mg/1 or 1.0% of reading (whichever
is larger).
1 rain, maximum to reach 98% of span change step
in DO; 2 min. maximum to reach 98% of tempera-
ture change step.
0-2 mg/1 typical maximum.
Water temperature: 0 to 45C; compensation
adjustable to within 0.10 mg/1 over entire
temperature range.
Air temperature: 0 to 45C.
Relative humidity: 0 to 100% less than ±0.25%
of reading effect for any combination of
temperature and humidity.
Polarizing power supply: 0.8 volts DC ±0.1
volt.
Sample/flow velocity: Any constant velocity
(±20% from 1.5 to 11 fps).
Water temperature: 0 to 45C.
Air temperature: 0 to 50C.
Transportation and storage - air temperature:
5-60C; mechanical shock - 50g for 30 msec.;
vibration - 5g over 0 to 60 Hz.
Electrodes: Gold cathode, silver anode-both
immersed in the electrolyte.
Electrolyte: 0.5 normal potassium hydroxide
and 2.0 normal potassium chloride in distilled
water.
Cell body: unplasticized PVC.
Memb rane: teflon.
"0" rings: Buna N.
Thermistor guard: stainless steel.
Membrane cap nut: stainless steel.
52
-------
Table 9. MANUFACTURER'S SPECIFICATIONS FOR DO SENSOR ASSEMBLY
(continued)
Physical data - Connecting cable.
The self-powered tank type is 24 in. long and
terminates in a Cannon XLR-3-12c connector.
The remote-powered tank type is 24 in. long
and terminates in a MS310E-14S-2P connector.
The submersible type is 50 ft. long and termi-
nates in a MS3106E-14S-2P. Extension cable is
available.
Weight: tank type 2.5 Ib. (1.14 kg).
Submersible type: 3.5 Ib. (1.50 kg).
MV/V INVESTIGATION
Operational performance of the MV/V transmitter was investigated in
the laboratory at selected temperatures for:
1. stability
2. linearity
3. response
(a) step input
(b) frequency
to determine whether it met certain items of the manufacturer's speci-
fications listed in Table 10.
The temperature stability was investigated to establish the most
favorable environment for operating the MV/V, i.e., under controlled
ambient conditions, or in an unheated, steel-roofed shelter in which
the ambient temperature would normally follow the outside ambient
temperature (OC to 40C). The linearity, which would indicate whether
the output would be directly proportional to the DO concentration was
determined before and after the field investigation. This would also
verify equation 11 in Section IV (MV/V Overview).
The response was determined experimentally by two methods: a sudden
input following shorted input terminals, and frequency response for
selected chopper frequencies.
53
-------
Table 10. MANUFACTURER'S SPECIFICATIONS FOR MV/V TRANSMITTER
Range -
Source impedance
Output -
Load impedance -
Accuracy -
Response time -
Long-term drift -
(30 days)
Normal operating
conditions -
Operative limits -
Alarm -
DO; fixed extended range 0-5 and 0-10 mg/1.
0 to 5000 ohms.
0-5 volts DC, non-isolated; as specified.
250 K ohms minimum for 0-5 volts DC output;
0 to 550 ohms maximum for 4-20 ma output.
±0.25% FS.
2.5 sec. to attain 99% FS step change.
0.25% FS maximum.
Ambient temperature and relative humidity (RH):
±0.5% FS shift for any combination of -5 to 50C
with 0 to 90% RH.
Power supply: ±0.01% FS/volt from 110 to 125
VAC, 50 or 60 Hz ±1 Hz.
Ambient temperature: -30 to 60C.
Power supply: 107 to 127 VAC, 50 or 60 Hz ±10%.
Vibration: 0.2g for 10 to 60 Hz.
Mechanical shock: 5g for 30 msec.
Separate adjustable "hi-Lo" alarm.
Contact rating: 5 amperes, 120 VAC and 2.5
amperes, 220 VAC, resistive load.
Alarm contacts: Hi - one normally closed or
normally open (selectable); lo - one normally
closed or normally open (selectable).
Temperature Stability
Temperature stability was investigated at 1.7, 18.3, and 35.OC
employing a constant input of 6.65 mv. During the test, control
circuits in the environmental chamber failed which caused a tempera-
ture runaway shown by Figures 32 and 33. Data displayed by Figure
34 were obtained immediately after this occurred, and the data for
temperature interval Ty were recorded at an uncontrolled ambient of
25C.
For the total stability test, temperatures were measured at several
points within the chamber and near the converter's chassis. This
data indicated absence of thermal gradients under steady-state con-
ditions.
There is insufficient data to draw a precise conclusion from the
trasistory runaway data although the runaway offers an opportunity
for speculation.
54
-------
47.0
46.0
45.0
.2
CD
u
CO
- 44.0
O
oc 43.0
LU
42.0
41.0
Note: Input = 6.65mVDC
Zero output is a nominal 10mVDC
Full scale output is a nominal 50mVDC
KEY: Temperatures at data points are
Ti.01.7t
T2=018.3°C
T3-A88t (estimated)
(Bin transition
T4=«26.7°C
40.0
12 18 0
5-28
6 12 18
6 12 18
12 18 0 6
6-1
0 6
5-29 5-30 5-31
TIME (hours and days of the year 1974)
FIGURE 32. TEMPERATURE STABILITY OF MV/V (BEFORE AND
DURING RUNAWAY).
12 18
55
-------
47.0
46.0
45.0
as
o
ui
44.0
oc
LJJ
t 43.0
1/3
I
42.0
41.0
40.0
k T3
Note: Input = 6.65mVDC
Zero output is a nominal 10mVDC
Full scale output is a nominal 50mVDC
KEY: Temperatures at data points are
T3=A88°C (estimated)
®in transition
14. • 26.7°C
T5-H18.3°C
18 0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18 0
6-1 6-2 6-3 6-4 6-5 6-6
TIME (hours and days of the year 1974)
FIGURE 33. TEMPERATURE STABILITY OF MV/V (IMMEDIATELY
AFTER RUNAWAY).
56
-------
47.O
46.0
45.0
44.0
43.0
E
t
42.0
41.
40.0
Note: I nput - 6.65 mVDC
Zero output is a nominal 10mVDC
Full scale output is a nominal 50mVDC
KEY: Temperatures at data points are
T5 *E18.3°C
® in transition
T6 =ffl35.0°C
1- -Vambient uncontrolled 25*t
6 12 18
6-6
0
12 18 0 6 12
6-10
6 12 18 0 6 12 18 0 6
6-7 6-8 6-9
TIME (hours and days of the year 1974)
FIGURE 34. TEMPERATURE STABILITY OF WIV/V AFTER RUNAWAY.
57
-------
The data displayed in Figure 32 suggest that the MV/V was temperature-
dependent during intervals TI and ^2- Change in output at 1.7C
(interval TI) was between 40.3 and 40.9 percent, within ±0.5 percent.
At room temperature (interval T2), the output varied between 41.2 and
42.9 percent for an overall shift of ±0.85 percent. The overall shift
for the maximum excursion (worst case) between each interval was 2.6
percent, or ±1.3 percent over both intervals.
Very little can be said regarding the runaway itself since thermal
gradients existed throughout the chassis, and a steady-state condi-
tion was not reached until the chamber was returned to 26.7C in
interval T^.
In interval T5, the chamber was returned to a controlled 18.3C
temperature, which was maintained for approximately 42 hours. The
output remained constant. For temperature intervals Tg (35C) and
T7 (25C), the output varied between 41.7 percent and 41.9 percent,
and 41.3 and 41.9, respectively, and within the expected ±0.5 per-
cent tolerance. The maximum variation between both intervals was
41.3 to 41.9, which was within the expected ±0.5 percent tolerance.
Since Honeywell personnel agreed that the MV/V had not been damaged,
other testing was pursued.
One month later, following recalibration, the MV/V was again exposed
to temperature-stability tests. Data comparable to those displayed
in Figures 30, 33, and 34 were obtained and are presented in Figures
35 and 36. At a constant input, the overall spread (worst case) of
the output in the controlled environmental chamber was between a
high of 44.0 percent at 35C to a low of 42.2 percent at 1.7C. Thus,
the shift was actually ±0.9 percent or ±0.4 percent more than
expected, but ±0.5 percent less than before the runaway. (Data
contained within the dashed sectors indicate a transition and are
excluded from the discussion.)
The one basic difference between the method used to obtain the
environmental chamber temperature changes presented in Figures 32
and 33 and Figures 35 and 36 was in the employment of the propor-
tional controller. In the first sequence, the slope is steeper
since 100 percent of the heat cycle was added, whereas a nominal 50
percent of the heat cycle was employed in the second sequence. The
reason for the reduced thermal shift in the output after the runaway
was not investigated, but it may have been related to semiconductors,
capacitors, or the circuitry and its power supply. The measuring
circuit is shown in Figure 37, and Table 11 summarizes the mid-point
output for each temperature interval. (Since the analyzer was
"recalibrated" to obtain the data in Part II, there is not equiva-
lence with Part I.)
58
-------
44.0
43.5
(0
o
Ifl
I
£ 43.0
O
I
Q
I42-5
cc
O
42.0
KEY:
A=T2 = 35.0°C
H = T3*18.3°C
© = T4 = 1.7^C
®= in transition
12
7-15
0 12
7-16
0 12
7-17
0 12
7-18
0 12
7-19
0 12
7-20
0 12
7-21
0 12
7-22
0 12
7-23
TIME (hours and days of the year 1974)
FIGURE 35. TEMPERATURE STABILITY OF MV/V (1 MONTH
AFTER RUNAWAY).
0 12
7-24
-------
44.0 -
= 18.3°C
= T4 = 1
= T5 = 18.3SC
in transition
TIME (hours and days of the year 1974)
FIGURE 36. TEMPERATURE STABILITY OF MV/V (1 MONTH
AFTER RUNAWAY).
-------
FIGURE 37. MEASURING CIRCUIT COMPONENTS OF MV/V.
63
-------
Table 11. MV/V THERMAL SHIFT
Temperature
Runaway C Interval
Part I
1. before
2. before
3. after
4. after
5. after
6. after
Part II
1.7
18.3
26.7
18.3
35.0
25.0
T4
Duration
hours
48
35
24
42
27
54
Output
range % mid-point
40.3-40.9
41.2-42.5
43.1-43.5
41.2-42.2
41.7-41.9
41.3-41.9
40.6
42.3
43.2
41.4
41.8
41.7
7.
8.
9.
10.
11.
after
after
after
after
after
18.3
35.0
18.3
1.7
18.3
Tl
T2
T3
^4
T5
36
48
51
93
90
42.6-43.1
43.5-44.0
42.6-43.1
42.2-42.5
42.8-43.1
42.9
43.7
42.9
42.4
42.9
Linearity
Before the field work was carried out, end-point linearity data for
the high range of the MV/V were obtained at 41F, 65F, and 95F (1.7C,
18.3C, 35.OC) (Figure 38). The data were obtained by exposing the
MV/V to controlled temperatures in the environmental chamber and
employing a Fluke Model 341A voltage source. The input and output
(output measured between TP-1 and TP-2) were measured by a Keithley
Model 616 digital electrometer. The input was limited to a maximum
of 8.0 mv, the equivalent of approximately a 14.5 mg/1 DO concentra-
tion, and within the expected range of operation.
The data in Figure 38 indicate that the end-point linearity for each
temperature was within a nominal ±0.5 percent FS, but the spread
(gain) at full scale over the temperature range was 3 percent or
±1.5 percent from room temperature. This spread converges to zero
for zero input.
Although the input signal was increased by 20 percent, the apparent
increase in gain with temperature is consistent with the stability
data shown in Figures 32 and 36.
62
-------
50.00
> 30.00
A AFTER 24 HOURS AT 35.0°C
AFTER 24 HOURS AT 18.3°C
O AFTER 24 HOURS AT 1.7°C
MV/V IN HIGH RANGE ONLY.
TEST RUN BEFORE FIELD EVALUATION
1.00
2.00
3.00 4.00 5.00
INPUT VOLTAGE ( IN MV DC)
FIGURE 38. MV/V END-POINT LINEARITY.
6.00
7.00
8.00
-------
After the field work was completed, additional linearity data for
the potential output and panel meter were obtained to determine what
effects, if any, time (approximately 4 months), field calibration,
and temperature compensation had had on the unit (Figures 39-41).
The MV/V was maintained at 69.8F (21C), and data were obtained for
both ranges. The method of investigation was similar to that shown
in Figure 38, but the input range was extended to 14 mv and adjusted
to provide panel meter deflections that corresponded to integral
markings. The resultant millivolt output and the required input
potential were plotted as dependent variables. The independent
linearity (output voltage was measured between TP-1 and TP-2) was
±0.3 percent, but the panel meter approached a nominal offset of
±2.5 percent when near zero, and it had a nominal independent linear-
ity of ±1.0 percent.
After the unit was operated in the laboratory for a month, its
linearity was again determined for both ranges (Table 12). Except
for full scale, the input was adjusted to integral numbers, and the
output potential and panel meter deflection were tabulated as depen-
dent variables. (Some meter-reader error may have occurred.)
Table 12. LINEARITY OF MV/V 30 DAYS FOLLOWING
INDEPENDENT3- LINEARITY TESTS
Input Ei Output Eo Output less offset, P Gain
(millivolts) (millivolts) Eo' (millivolts) PM Eo'/Ei
Low range:
0.00
1.00
2.00
3.00
4.00
5.00
6.00
6.63
High range :
0.00
2.00
4.00
6.00
8.00
10.00
12.00
13.00
13.63
9.80
15.42
21.40
27.40
33.50
39.70
46.00
50.00
9.75
15.24
21.00
26.80
32.80
38.90
45.00
48.00
50.00
0.0
5.62
11.60
17.60
23.70
29.90
36.20
40.20
0.00
5.49
11.25
17.05
23.05
29.15
35.25
38.25
40.25
0.20
0.75
1.49
2.25
3.00
3.75
4.51
5.00
0.50
1.49
2.90
4.40
5.80
7.30
8.75
9.50
10.00
5.62
5.80
5.87
5.93
5.98
6.03
6.06
:
2.75
2.81
2.84
2.88
2.92
2.94
2.94
2.95
Liata taken at room temperature.
64
-------
Ui
KEY A HIGH RANGE
O LOW RANGE
NOTE; MV/V AT 21.0 C
TEST RUN AFTER FIELD EVALUATION
SCALE OF ERRORS 10/1
2.00
4.00
6.00
8.00
10.00
12.00
14.00
INPUT VOLTAGE (IN MV DC)
FIGURE 39. INDEPENDENT LINEARITY OF THE IWV/V.
-------
•f
KEY A HIGH RANGE (0-10)
O LOW RANGE (0-5)
NOTE; TEMPERATURE 21.0°C
TEST RUN AFTER FIELD EVALUATION
2.00
4.00
10.00
6.00 8.00
INPUT VOLTAGE (IN MV DC)
FIGURE 40. INDEPENDENT LINEARITY OF PANEL METER.
12.00
14.00
-------
KEY A INCREASING VOLTAGE
O DECREASING VOLTAGE
NOTE; SCALE OF ERRORS; 10/1
TEMPERATURE =23.6°C
TEST RUN ON 2/24/75
EACH INPUT HELD CONSTANT FOR 90 SECONDS
2.00
4.00
6.00
8,00
10.00
12.00
14.00
INPUT VOLTAGE (IN MV DC)
FIGURE 41. INDEPENDENT LINEARITY OF THE MV/V.
-------
The data in Table 12 indicate that the low-range gain increased from
5.62 to 6.06 (or approximately 7.5 percent for the worst case) and
that the high-range gain increased a nominal 7.2 percent. The simi-
larity of the gain values displayed in Figures 39-41 and those
presented in Table 12 suggests that the long-term (30-day) stability
of the MV/V was satisfactory.
The total MV/V linearity data suggest that a nonlinear effect of a
nominal ±0.5 percent existed at a given temperature. Exact correla-
tion between ranges was neither intended nor investigated. Although
the MV/V had a tendency to shift with temperature before the runaway
occurred, the fact that the MV/V was exposed to 200F (93.5C) raises
questions as to its exact nonlinear characteristics. It should be
recalled that the system was intended to operate over a DO range of
0-5 and 0-10 mg/1 but that this was changed in the field at EPA's
request to 0-12 and 0-24 mg/1. This modification may have affected
the design and also the overall tolerances listed in Table 10.
The final independent linearity data taken indicates that the MV/V's
nonlinearity is approximately ±0.27 percent (=±0.3 percent). The
data also indicate that hysterisis was negligible because the input
was varied from zero to full scale and back to zero. As in the case
of Figure 39, the deviation in Figure 40 was multiplied by a factor
of 10.
Response
The MV/V is provided with an input RC filter to dampen sudden changes
in signal. Transient (time) response data (and the effect of the
filter) when the MV/V was exposed to an input signal that produced
a nominal 200 percent increase in output signal are displayed in
Figure 41. Initial conditions consisted of shorted input terminals,
and final conditions responded to an output of about 48 percent,
including offset.
Full response was attained within 1.3 seconds for the signal increase,
and full response of a nominal 1.2 seconds was attained when the input
was short-circuited. Honeywell's specifications call for a response
to 99 percent of full-scale step change within 2.5 seconds.
The filter action at 60 Hz was of particular interest since Honeywell
indicated that:
1. The filter was "notched" at 60 Hz (i.e., actually attenuated).
2. The primary noise originated from power sources (60 Hz).
68
-------
vo
NQ!E;TEMPERATURE=65° F,18.3°C
( f
INPUT SHORT CIRCUITED INITIALLY
5.65 MV INPUT APPLIED
4-
5.65 MV INPUT REPLACED
BY SHORT CIRCUIT
3 '' 11 12
ELAPSED TIME (SECONDS)
FIGURE 42. STEP-CHANGE RESPONSE OF MV/V.
13
14
15
-------
5.00C
u
Q
{2
d
a
cc
o
LL.
O
CO
O
O
NOTE; INPUT AT DC 0.01370 VOLTS DC
INPUT AT 3-100 HERTZ 0.0137 VOLTS RMS
0.30
< 0.200- •§
0.100-•
40
KEY MULTIVIBRATOR FREQUENCY
A 102 Hz
—-O 204 Hz
O 431 Hz
80
100
-0.100-1-
INPUT FREQUENCY (IN HERTZ)
FIGURE 43. FREQUENCY RESPONSE OF THE MV/V.
70
-------
The frequency response was obtained for the three multivibrator
frequencies shown in Figure 43. (The data in Figure 43 were
obtained after the chamber temperature ranaway and the MV/V was
maintained at room temperature.) Three frequencies were chosen to
determine if multivibrator drift would affect AC noise attentuation.
(No other performance characteristics of the multivibrator frequency
were investigated.) The RMS value of the input AC signal was held
constant at 13.7 mv, the same magnitude as that of the initial DC
potential.
Although each multivibrator frequency produced a slightly different
output, the change was so slight that 60 Hz-induced noise is appar-
ently unrelated to the frequencies selected.
The apparent resonance that seemed to occur between 7 and 10 Hz had
no effect on overall performance.
71
-------
SECTION VI
FIELD INVESTIGATION
INTRODUCTION
Field investigation of the Honeywell DO parametric system was
conducted at the Hamilton Sewage Treatment facility located on the
Great Miami River south of Hamilton, Ohio.
The plant is designed to service a population of 75,000 and can treat
12 million gallons of sewage daily. Design data are summarized in
Table 13, and a flow diagram of the plant is displayed in Figure 44.
The sensor was tested in the final settling basin, aeration tank, and
effluent chamber.
The final settling basin was chosen to determine how the unit per-
formed in a climate in which the sample velocity across the sensor
membrane was minimal. Since the sensor was provided with neither a
self-contained sample agitator nor a self-cleaning device, such as a
mechanical wiper, it was expected that the output would degenerate
with time, as indicated by equation A-l in Appendix A.
Performance in the aeration tank was of primary concern since
Honeywell had designed the system primarily for this application.
It was expected that sample flow and turbulence created by aeration
would produce adequate sample velocity over the membrane and that a
daily cleaning of the membrane would prevent signal degeneration.
The effluent chamber provided two characteristics favorable to making
continuous DO concentration measurements: (1) the velocity of the
effluent was high (»1.8 ft/sec[54.9 cm/sec]); (2) the chlorine used
would inhibit slime growth on the membrane.
Circles shown in the flow diagram of the treatment plant displayed
by Figure 44 indicate location of the sensor during the field tests.
Although reference is made to instrumental differentials, the reader
is reminded that investigation at Hamilton Treatment Plant was not a
comparison study of several DO parametric systems, but an investiga-
tion of one Honeywell design concept. It should also be recalled
that the YSI was calibrated each time it was used, and agitated, to
obtain a maximum reading. It should further be recalled that the
Delta Scientific provided with an agitator had been installed by
72
-------
Table 13. PLANT DESIGN DATA*
Item Characteristics
1 Description:
Split Activated Sludge Treatment Plant with separate
sludge digestion, sludge filtering and chlorination
of effluent.
2 Design Average Flow:
12 MGD (million gallons per day)
3 Design Maximum Flow:
18 MGD
4 Design Sewage and Waste Characteristics:
Suspended Solids - 350 PPM (parts per million)
5-Day B.O.D. (Biochemical Oxygen Demand)
5 Expected Overall Removal of Suspended Solids at Design
Flow of 12 MGD:
75%
6 Expected Overall Removal of 5-Day B.O.D. at Design Flow
of 12 MGD:
65%
7 Design Population (1970):
75,000
8 Two Raw Sewage pumps:
Rated: 12 MGD
Capacity of each pump: Maximum: 20 MGD
9 Two Sewage Pump Engines:
Horsepower of each engine: 175 brake horsepower (at 760
R.P.M.)
10 Two Air Blowers
Capacity of each blower: 7000 cu. ft. per minute
11 Two Air Blower Engines
Horsepower of each engine: 300 brake horsepower (at
690 R.P.M.
12 Detention Period in Grit Chambers at 12 MGD Flow:
10 minutes
13 Detention Period in Primary Tanks at 12 MGD Flow:
2 hours
14 Detention Period in Aeration Tanks based on Settled Sewage
Flow of 6 MGD and 30% Return Sludge:
6 hours
15 Detention Period in Final Settling Tanks based on Average
Flow of 6 MGD:
2.8 hours
16 Detention Period in Chlorine Contact Chamber at Maximum
Flow of 18 MGD:
16 minutes
17 Total Digester Volume:
486,000 cubic feet
18 Digester Volume per Capita Based on Design Population:
6.5 cubic feet
19 Two Digested Sludge Filters
Capacity of each filter: 1250 Ibs. digested sludge (dry
solids basis) per hour
20 Capacity of Gas Storage Sphere at 40 P.S.I.:
120,000 cubic feet
*Taken directly from Hamilton's Sewage Treatment Plant Brochure.
73
-------
MECHANICAL
SEWAGE SCREEN
RAW SEWAGE
PUMPS
CHLORINE
MI^MV^^^^MIIH
CONTACT
CHAMBER
MEASURING
FLUMES
MEASURING
FLUME
PRIMARY
SETTLING
TANKS
0
SLUICE •JEASURING1 FLUMES T
GATE A A A
+H
4-!
H
•^-i
^-*MI
T
AERATION
TANKS
FIGURE 44. SEWAGE FLOW DIAGRAM.
r'QT
^^ SLUDGE ^-A^
^f FILTERS «V
SLUDGE
CAKE
SLUICE
GATE
D
aaQ
D
A D
• aaO
-------
Hamilton Sewage Treatment Plant as a primary DO indicator so that
deviations between the two instruments (Delta Scientific and
Honeywell) are not intended to infer that one or the other is refer-
ence data. In spite of difficulty in sampling, the Winkler is
reference for the field work. Nonetheless, these peripheral mea-
surements are included for informational purposes to those requiring
this type of data.
FINAL SETTLING BASIN (SECONDARY)
Figure 45 displays the sensor installed in the settling basin, and
the data accumulated during the test are summarized in Table 14.
The "before turbulence" data were accumulated as follows: the
Honeywell MV/V output—including both the panel meter and the strip
chart recorder was recorded; the YSI sensor was calibrated and
inserted adjacent to the Honeywell, stirred, and its output recorded;
and, simultaneously, a sample was drawn for the Winkler determina-
tion of the DO concentration.
The "after turbulence" data acquisition was similar: within 3
minutes, the above process was repeated, but the Honeywell sensor
was manually agitated by raising and lowering it. The maximum
deflection noted was then compared with the second Winkler determi-
nation.
The differentials between the Honeywell and Winkler measurements
were significantly reduced "after turbulence," but the Honeywell
deflections were not consistently within tolerance. This may have
occurred because: (1) the sample taken for chemical analysis was
not the same as the sample measured by the sensor; (2) chemical
components in the sample caused the Winkler to indicate a higher DO
concentration level; (3) insufficient sample velocity across the
membrane increased the diffusion layer into the sample and reduced
the equivalent oxygen current; (4) slime growth on the sensor
reduced the signal level.
A device designed and built by plant personnel containing a BOD
bottle was inserted in the basin adjacent to the sensor so that the
sample for chemical analysis was taken within 2 inches (5 cm) of the
sensor. Similarly, the YSI was inserted adjacent to the Honeywell
sensor so that we assume an equivalent sample for all measurements.
No tests were run to determine whether interference affected the
Winkler process. Plant personnel indicated that interference (to
the Winkler process) would more likely occur in the activated sludge
where the suspended solids were highly agitated and, hence, would
become a significant part of the sample for chemical analysis.
75
-------
FIGURE 45. SENSOR INSTALLED IN SECONDARY SETTLING BASIN
-------
Table 14. MEASUREMENTS MADE IN FINAL SETTLING BASIN
Date § time
Before turbulence
Winkler YSI1'2
After turbulence
Honeywell Winkler YSI Honeywell
PM E! PM £2
8/21
1:00 p.m.
2:00 p.m.
3:00 p.m.
8/22
9:30 a.m.
4.0
4.1
4.2
1.4
3.7
4.0
4.0
NA
1.9
2.4
2.6
-2.1
-1.7
-1.6
0.3 -1.1
3.9
4.0
4.1
1.5
4.0
4.0
4.0
NA
3.8
3.6
3.6
Sensor cleaned after the "Before turbulence" 10:00 a.m. reading
10:00 a.m.
11:00 a.m.
12:00 N
:00
:00
1:
2:
p.m.
p.m.
3:00 p.m.
8/23
9:00 a.m.
1.6
1.8
2.5
3.0
3.5
3.5
1.8
1.6
1.7
2.5
3.2
3.6
3.6
NA
0.8
0.6
0.9
1.9
1.7
-0.8
-1.2
-1.6
-1.1
-1.8
2.3 -1.2
0.8 -1.0
1.7
Sensor cleaned after the "Before turbulence" 9:30 a.m. reading
-0.
-0,
-0,
1.0 -0.5
1.6
1.7
2.5
2.8
3.6
3.6
1.2
1.8
2.4
2.9
3.6
3.5
1.2
1.6
2.2
2.6
3.2
3.6
-0.4
-0.1
-0.3
-0.2
-0.4
0.0
1.8 1.6 -0.1
9:30
10:00
11:00
12:00
a.m.
a.m.
a.m.
N
1
1
2
2
.7
.9
.1
.9
1
1
1
2
.8
.5
.7
.5
1.2
1.1
1.0
1.5
-0.5
-0.8
-1.1
-1.4
1.8
1.9
2.2
2.7
1.6
1.6
1.7
2.6
1.6
1.6
1.8
2.7
-0.2
-0.3
-0.4
0.0
1. Yellow Springs Instruments
2. The precalibrated YSI was always agitated to obtain maximum output.
£l,£2 = Difference between the Winkler and Honeywell panel meter (H-W).
77
-------
Minimal sample flow (factor(s) above) across the membrane appeared to
be the primary cause of differences - presuming Winkler determinations
as reference data. The YSI - following maintenance and calibration -
was always agitated and therefore produced data more nearly equal to
the Winkler for the "before turbulence" case.
The turbulence data suggest that the sensor was inadequately agitated,
but no direct conclusion can be drawn from the cleaning. It is appar-
ent for this type of application that sensors must be provided with:
(1) a submerged pump to push water over the membrane surface at a
high velocity; or (2) an automatic cleaning device that eliminates the
need for daily manual membrane cleaning; (3) both. Delta Scientific
and Weston and Stack integrated these functions so that cleaning
action is obtained by the sample agitator.
PERFORMANCE IN THE AERATION BASIN
Figure 46 shows the Honeywell sensor installed in the aeration basin,
and the data obtained are summarized in Table 15. (Immediately before
testing started, the sensor was cleaned, and the parametric system was
recalibrated at room conditions, employing sodium bisulphite for zero
and saturated tap water for span.)
The DO concentration was periodically determined by the Winkler process,
logged from the panel meter of a Delta Scientific instrument in continu-
ous operation, and periodically logged from the YSI panel meter. The
deflections of the Honeywell panel meter were logged, and the output
from standard test points TP-1 and TP-2 was continuously recorded on
an Esterline Angus Speed Servo recorder. The Honeywell sensor was
installed within 1 foot (30.5 cm) of the Delta Scientific sensor in
the aeration basin and readings were taken with the portable YSI
within immediate vicinity of the two sensors. Plant personnel logged
data during the second and third work shifts so that data were obtained
on a 24-hour basis. During these periods Hamilton Sewage Treatment
Plant logged the Delta Scientific and Honeywell outputs.
To reduce the effect that suspended solids might have on the chemical
determination, 2 ml of 10 percent copper sulfate (10 grams/100 ml H20)
were added as soon as the sample was brought to the cement walkway.
Chemicals for the Winkler determination were then added, with the
sulphuric acid being added after settling had occurred. The sample
was taken within 2 inches (5 cm) of the Honeywell sensor.
Since three different DO parametric systems were employed, one would
not expect each to indicate identical values in an activated sludge
environment due to flowing sample, variations in turbulence, inability
78
-------
FIGURE 46. SENSOR INSTALLED IN ACTIVATED SLUDGE BASIN.
79
-------
to measure identical samples, inconsistency in the air bubbles from
the aeration system, and tolerances within each design. Occasionally
the indicated DO varied during meter reading so that "reader" error
would also be expected.
Except for three data points, all Honeywell measurements were within
±0.3 mg/1 of the DO concentration determined by the Winkler process.
The Honeywell reading exceeded the Winkler determination by 0.6 mg/1
on two occasions, and was low by 0.4 mg/1 in one case.
Seventy-nine percent of the differentials between the YSI and
Honeywell were within ±0.3 mg/1. Four data points indicated a dif-
ference of ±0.5 mg/1 between the two instruments.
Seventy-seven percent of the differences between the Delta Scientific
and Honeywell instruments were within ±0.3 mg/1 and 23 percent of the
differentials were within ±0.6 mg/1 excepting one that was 1.8 mg/1.
The latter was discounted.
For all data accumulated, including 56 data points during the preced-
ing week, the differentials between the Honeywell and Delta Scientific
were within ±0.4 mg/1 90 percent of the time.
Employing the Winkler as reference, the mean and standard deviations
for the Honeywell system were 0.118 and 0.256, respectively.
Several problems occurred in obtaining the data shown in Table 15.
The inking system on the recorder failed several times, and factory
replacements had to be installed. Frequently, samples taken for the
Winkler process were too turbid, and the titration was invalidated.
The addition, however, of copper sulfate reduced this interference.
On frequent occasions, a series of spikes was recorded following
sensor cleaning; a typical series is displayed in Figure 47. This
was corrected by dipping the tip of the sensor into a sulphuric acid
solution (pH = 2.5) and not mechanically wiping the sensor.
EFFLUENT CHAMBER (CHLORINATED)
Before the sensor was installed in the effluent chamber, its cap was
removed and sludge (oxides of silver) was removed from the cathode.
The parametric system was then recalibrated employing sodium bisul-
phite for zero and saturated tap water for span. The sensor was
installed and moved to an approximate 45-degree angle by the force
of the water, which had a minimum velocity of 15 feet per second (4.8
m/sec) at the sensor (Figure 48). The sensor was not cleaned during
the test. Samples were drawn periodically for the Winkler determina-
tion, and peripheral measurements by the YSI were recorded. The YSI
80
-------
Table IS. MEASUREMENTS MADE IN AERATION TANK
00
Date § time
8/29/74
11:00 a.m.
12:00 N
1:00 p.m.
2:00 p.m.
3:00 p.m.
4:00 p.m.
6:00 p.m.
8:00
10:00 p.m.
12:00 M
8/30/74
2:00 a.m.
4:00 a.m.
6:00 a.m.
8:00 a.m.
9:30 a.m.
10:00 a.m.
11:00 a.m.
12:00 N
1:00 p.m.
2:00 p.m.
4:00 p.m.
6:00 p.m.
8:00 p.m.
10:00 p.m.
12:00 M
8/31/74
2:00 a.m.
4:00 a.m.
6:00 a.m.
8/31/74
8:00 a.m.
10:00 a.m.
12:00 N
1:00 p.m.
2:00 p.m.
4:00 p.m.
6:00 p.m.
8:00 p.m.
10:00 p.m.
12:00 M
W
6.4
6.2
6.5
6.3
6.4
—
—
—
—
—
—
—
New RCDR
6.3
6.5
6.8
7.2
6.5
6.1
—
—
—
—
—
—
—
—
—
—
—
3.5
YSI*
6.2
6.2
6.3
6.4
6.3
—
—
—
—
—
—
—
—
pen
—
—
6.8
6.8
6.2
5.9
—
—
—
—
—
—
—
—
—
—
—
3.5
A*
6.3
6.1
6.2
6.2
6.3
6.0
5.4
4.9
4.1
4.1
4.5
4.9
5.4
6.4
6.6
7.0
7.1
6.4
6.1
5.0
5.1
5.0
4.4
3.9
3.6
3.4
3.0
3.1
3.7
3.6
3.4
2.7
2.1
1.5
1.5
2.0
2.2
PM
6.4
6.3
6.4
6.4
6.4
6.0
5.5
4.8
4.3
4.3
4.3
4.9
5.5
6.6
6.7
7.3
7.3
6.6
6.1
5.1
5.1
5.1
4.3
3.9
3.6
3.2
3.2
3.2
3.6
3.6
3.4
2.6
2.4
1.2
1.2
1.5
2.2
RCDR
6.5
6.4
6.4
6.4
6.5
6.0
5.5
4.9
4.4
4.5
4.7
5.0
5.3
6.6
6.8
7.4
7.3
6,7
6.1
5.1
5.0
4.8
4.4
3.7
3.7
3.0
2.7
3.0
3.3
3.4
3.3
2.3
1.7
1.1
0.9
1.4
1.9
el
0.1
0.2
-0.1
0.1
0.1
—
—
—
—
—
—
---
0.3
0.3
0.6
0.1
0.2
0.0
—
—
—
—
—
—
—
—
—
—
—
-0.2
E2
0.3
0.2
0.1
0.0
0.2
—
—
—
—
—
—
—
—
0.6
0.5
O.S
0.2
—
—
—
—
—
—
—
—
—
—
—
-0.2
e3
0.2
0.3
0.2
0.2
0.2
0.0
0.1
0.0
0.3
0.4
0.2
0.1
-0.1
0.2
0.2
0.4
0.2
0.3
0.0
0.1
-0.1
-0.2
0.0
-0.2
0.1
0.1
-0.3
-0.1
-0.4
-0.2
-0.1
-0.4
-0.4
-0.4
-0.6
-0.6
-0.3
e
-------
Table 15 (continued). MEASUREMENTS MADE IN AERATION TANK
00
Date § time
9/1/74
2:00 a.m.
4:00 a.m.
6:00 a.m.
8:00 a.m.
10:00 a.m.
12:00 N
2:00 p.m.
4:00 p.m.
6:00 p.m.
8:00 p.m.
10:00 p.m.
12:00 M
9/2/74
2:00 a.m.
4:00 a.m.
6:00 a.m.
8:00 a.m.
10:00 a.m.
9/2/74
12:00 N
2:00 p.m.
4:00 p.m.
6:00 p.m.
8:00 p.m.
10:00 p.m.
12:00 M
9/3/74
2:00 a.m.
4:00 a.m.
6:00 a.m.
8:00 a.m.
10:00 a.m.
10:30 a.m.
11:30 a.m.
12:30 p.m.
1:30 p.m.
2:00 p.m.
W YSI* A*
2.8
3.2
4.4
5.7
6.6
6.5
5.8
4.0
2.8
2.8
3.2
4.0
5.0
5.4
5.5
5.4
6.0
6.5
5.7
4.5
4.2
3.7
— — —
3.7
4.0
4.1
4.5
6.0
6.9
7.2 6.9 7.1
6.4 6.8 6.6
6.1 6.1 5.9
5.5 5.4 5.2
Sensor cleaned with H2S04
4.9 4.9 4.7
PM
2.6
3.2
4.3
6.0
6.7
6.3
6.0
4.2
2.8
2.9
3.2
2.9
5.1
5.7
5.7
5.5
6.2
6.7
6.2
4.8
4.2
3.7
—
3.9
4.1
4.1
4.5
6.2
7.4
7.3
6.9
6.1
5.3
solution
4.8
RCDR E! £2
2.3
2.8
2.6
ink failed — —
6.8
6.8
5.9
4.2
2.4
2.6
3.0
3.7
5.0
5.5
5.6
5.6
6.1
6.7
6.1
4.7
4.0
3.6
3.6
3.6
3.8
4.0
4.4
6.1
7.3
7.4 0.2 0.5
7.0 0.6 0.3
6.1 0.0 0.0
5.1 -0.4 -0.3
4.8 -0.1
x -v 0.118 0.207
s -»• 0.256 0.270
63
-0.5
-0.4
-1.8
0.2
0.3
0.1
0.2
-0.4
-0.2
-0.2
-0.3
0.0
0.1
0.1
0.2
0.1
0.2
0.4
0.2
-0.2
-0.1
—
-0.1
-0.2
-0.1
-0.1
0.1
0.4
-0.3
0.5
0.2
-0.1
0.1
-0.030
0.341
Ei* E5
— —
—
— —
— —
— —
—
— —
— —
— —
— —
— —
— —
— —
— —
-0.1 0.3
0.2 +0.4
-0.2 0.0
-0.3 -0.1
-0.2 0.0
-0.059 -0.09
0.154 0.20
PM and RCDR refer to output of Honeywell DO parametric system.
EJ = RCDR - W E3 = RCDR - A eg = YSI - W
E2 = RCDR - YSI E4 = A - W
-------
100
75' •
UJ
o
CO
UJ
u
50- •
25- •
DC
O
0-
SENSOR CLEANED
0 1
ELAPSED TIME (HOURS)
FIGURE 47. SENSOR OUTPUT IN ACTIVATED
SLUDGE FOLLOWING MANUAL
CLEANING.
83
-------
FIGURE 48. SENSOR INSTALLED IN CHLORINATED EFFLUENT.
84
-------
was calibrated on a daily basis employing the manufacturer's procedure
as outlined in the YSI manual (Model 51A). Similar to the previous
tests, the YSI was inserted adjacent to the Honeywell, and samples for
the Winkler were extracted from a position within 2 inches (5 cm) of
the Honeywell sensor. Data for the test are summarized in Table 16.
Although data were accumulated continuously on the recorder, only data
recorded from the Honeywell panel meter are tabulated. Midway during
the test, the recorder developed a 2.5 percent down-scale offset; the
fault was corrected, disrupting the continuity in data.
The ability of the Honeywell sensor to operate unattended, continu-
ously, and successfully (well within ±0.3 mg/1) for 1 week was
traceable to two basic facts: the chlorinated condition of the
effluent and its high discharge velocity (greater than 15 feet per
second [4.8 m/sec] at the sensor).
85
-------
Table 16. MEASUREMENTS MADE IN CHLORINATED EFFLUENT
Date $ time
9/4/74
4:30 p.m.
9/5/74
10:00 a.m.
11:00 a.m.
12:00 N
1:00 p.m.
2:00 p.m.
3:00 p.m.
9/6/74
9:30 a.m.
10:00 a.m.
11:00 a.m.
12:00 N
1:00 p.m.
2:00 p.m.
3:00 p.m.
9/8/74
12:30 p.m.
9/9/74
12:00 N
1:00 p.m.
2:00 p.m.
3:00 p.m.
9/10/74
3:00 p.m.
Winkler
2.0
4.5
4.0
4.1
3.6
3.6
3.5
3.5
3.4
3.2
2.4
2.5
2.6
2.6
3.6
2.8
2.8
2.8
2.8
2.8
Standard
YSI
—
4.3
4.1
3.8
3.8
3.6
3.2
2.5
2.5
2.6
2.6
3.7
2.9
2.8
2.9
2.8
Mean =
deviation =
PM
2.2
4.4
3.9
4.0
3.8
3.6
3.6
3.4
3.4
2.9
2.4
2.4
2.5
2.5
3.7
2.8
2.9
2.8
2.8
2.8
X
s(x)
«1
+0.2
-0.1
-0.1
-0.1
+0.2
0.0
0.0
-0.1
0.0
-0.3
0.0
-0.1
-0-1
-0.1
+0.1
0.0
+0.1
0.0
0.0
0.0
-0.0200
0.115
£2
-0.4
-0.1
0.0
-0.2
-0.1
-0.3
-0.1
-0.1
-0.1
-0.1
0.0
-0.1
+0.1
-0.1
0.0
0.0
-0.100
0.121
-3
+0.3
0.0
+0.2
+0.2
+0.1
—
0.0
+0.1
0.0
0.0
0.0
+0.1
+0.1
0.0
+0.1
0.0
0.0800
0.0941
Key:
YSI = Yellow Spring Instrument DO system
PM = Panel Meter of Honeywell Instrument DO system
E! = PM - Winkler
E2 = PM - YSI
£3 = YSI - Winkler
q
The Delta Scientific Corporation dissolved oxygen sensor was not
transferred to the effluent raceway for this test.
86
-------
SECTION VII
EVALUATION OF MANUALS
Honeywell provided two manuals on the DO parametric system: (1)
#830-10 operator's manual for the sensor assembly, and (2) #542012-
11-121-200001-12 operator's manual for the MV/V. Each manual is of
a modular design and contains the following basic sections: descrip-
tion, installation, operation, maintenance, theory, service, parts
list.
The MV/V manual contains many more sections than the sensor assembly
manual, because the MV/V is electronically more sophisticated and
contains many devices, as noted earlier.
As a guide to this Section (VII), Section 5.00 of reference 5 was
employed.
The following line items are intended for the beginner as it points
out several areas that the Honeywell manuals either exceed EPA's
current requirements, are not applicable, or emphasizes an area that
the authors believe are necessary for this type of instrumentation.
Section 5:00
Section 5.04
Section 5.07
Section 5.08
Section 5.10
Not applicable for this particular project. However,
Honeywell has been known to provide as many copies of
each manual as the customer requires.
Meets - This section is instructional (both manuals).
Meets - This section is instructional (both manuals).
Sensor Manual - The Honeywell manual indicates that
the sensor "is not calibrated separately but as a
parametric system with its matching conditioner."
Due to the basic design philosophy, it was possible
and convenient to investigate each separately for
given tests.
Honeywell provides many additional engineering draw-
ings and some exploded diagrams that clarify any
photographic vagueness.
Section 5.12 All parts are available from Honeywell.
87
-------
Section 5.14 Cost not included.
Section 5.15 Cost not included.
Section 5.16 Cost not included.
OVERALL APPRAISAL OF MANUALS
Overall, the Honeywell manuals provide more than adequate information
to permit successful use of the product and they are also tutorial.
Because Honeywell manufactures a variety of products that employ
similar components, the modular sections can be inserted in other
manuals. This affects the numbering of figures and causes a certain
amount of inconsistency.
SENSOR MANUAL
There are several typographical errors regarding the chemicals needed:
Section 131-33, Page 3, Physical data; 535-3, Page 2: 636-9, Page 3;
639-9, Page 6.
The summary of material specifications-, Section 131-33, Page 3, might
be more easily understood if the type of material followed the func-
tion: Cell Body - Unplasticized PVC, etc., i.e., a horizontal rather
than a vertical format.
The page numbering required because of the modular arrangement is not
as straightforward as other manuals that are used for similar designs.
It would be better to number pages consecutively, although there is
merit in giving each section its own numbering system. Perhaps each
manual could be provided with a unique table of contents. This could
be accomplished via the Honeywell Data Processing function and merely
key-in the functional descriptions required. The printout would
provide the table of contents and the computer could print the entire
manual.
MV/V TRANSMITTER MANUAL
The MV/V transmitter manual reveals excessive design strategy and a
generalized product suitable for many applications.
1. Costs have been omitted because of variations in economics. GSA
schedule prices are available to Federal Agencies and open-market
prices are available from the Honeywell Offices.
88
-------
The manual is also modular in format, and this has caused several
details to be omitted. Of particular concern is the numerical sub-
script for the components, i.e., there are several TP-l's, Ql's, RI's,
etc., since the numbering is unique to each circuit card. A double
subscript or alpha-numeric subscript system might be more direct.
Additionally, a clearer interconnection drawing between cards would be
desirable - for example, the feedback through the measuring circuit
is not obvious.
Modularity and variety of options make schematic interpretation diffi-
cult . i
Page numbering comments are similar to those for the sensor manual.
Honeywell personnel have been exceptionally liberal in providing
design information so that anyone desiring to work successfully with
this product should have no problems.
89
-------
SECTION VIII
REFERENCES
1. Patent #3,510,421 issued to A. Gealt, Honeywell Inc., Fort
Washington, Pa. 19034.
2. Patent #3,235,477 issued to A. H. Keyser, et al, Honeywell Inc.,
Fort Washington, Pa. 19034, October 12, 1962.
3. Bates, R. G., "Electrometric pH Determinations - Theory and
Practice," Wiley $ Sons, Inc., New York, 1954.
4. Gealt, A. E., and Metarko, R. P., Honeywell Inc., "Reliable
Measurement of Dissolved Oxygen in Polluted Waters," presented
at ISA Conference and Exhibit, Philadelphia, Pa., October 1970.
5. Mentink, A. F., "Specifications for an Integrated Water Quality
Data Acquisition System - 8th Edition," FWPCA, Division of Pol-
lution Surveillance, U.S. Department of the Interior, Cincinnati,
Ohio 45268.
6. Private communication, A. Gealt, Honeywell Inc., and the authors,
June 1974.
7. Norton, H. N., "Handbook of Transducers for Electronic Measuring
Systems," Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
8. Yellow Springs Instrument Company, Inc., Yellow Springs, Ohio
45287.
9. Delta Scientific Corporation, Linderhurst, New York 11757.
90
-------
APPENDIX A
BACKGROUND ON DISSOLVED OXYGEN MEASUREMENTS
DO is theoretically present in all open bodies of water, but waters
void of DO are generally heavily polluted, so that DO is one of the
more frequently measured parameters. With some exceptions (such as
an anaerobic operation) DO is a basic ingredient to the sustaining
of aquatic life and is one of the primary measurements made in
sewage treatment plants.
In an aerobic domestic (organic) sewage treatment plant, i.e., one
in which the waste, air (or "pure" oxygen), and microorganisms are
mixed, oxygen is one of the variables which determines how effec-
tively a treatment function operates. If saturation is reached,
additional oxygen bubbles through the mix to the atmosphere. Under
some conditions additional aeration could be uneconomical so that
measurement of DO within the aeration tank can be used to control
the aerators saving oxygen, energy, machinery wear, and cost.
Overall, D. G. Ballinger's* considerations for employing unattended
instruments (quiescent monitoring) in water pollution control pro-
grams can also be used in process control .^-1 in his paper he states
that "the first factors to be considered are the specific objectives
of the control (legal) programs."
"Among the objectives may be:
1. To collect information on sources of pollution and the impact
on water quality, in support of regulatory activities. [Treat-
ment effluents are being regulated.]*
2. To obtain real-time data for the protection of public water
supplies. ["Raw material" sources for water supplies are
generally rivers and lakes into which industrial effluents
have been discharged.]
3. To obtain sufficient data to develop mathematical models for
water quality management. [Sewage treatment facilities can be
fully automated and computer controlled once favorable models
are developed.]
*Director, Environmental Monitoring and Support Laboratory - Cincinnati,
Office of Research and Development, U.S. Environmental Protection Agency.
tRemarks in brackets are those of the authors.
91
-------
4. To record long-term trends in the quality of resources." [This
could easily include long-term trends related to sewage plant
influents and effluents.]*
DO has been measured in rivers, streams, lakes, and oceans for a
number of years to identify many problem areas. One of the major
pollution sources causing excessive BOD and reduced DO levels is
overburdened domestic sewage treatment plants.
Oxygen is dissolved in streams, lakes, and oceans in accordance with
Henry's Lawt. The solubility of oxygen is a function of the water's
temperature, pressure, and salt content.^-2 As the temperature
increases, the solubility of oxygen decreases, and as the salt con-
tent increases, a "salting out" phenomenon also reduces the solubility.
METHODS USED TO MAKE DO DETERMINATIONS
There are several methods for determining DO in water samples:
1. Winkler process
2. Application of paramagnetic properties of DO
3. Galvanic cell, exposed electrodes
4. Galvanic cell, membrane covered assembly
5. Passive cell, membrane covered assembly.
1. The Winkler process employs wet chemical techniques that culmi-
nate in the titration of free iodine. Although fully documented
in Reference A-2, it is well to consider the following:
a) Several chemical modifications to the Winkler process are
necessary to compensate for ionic interferences unique to
the samples.
b) Generally, low-temperature, nearly saturated samples are
titrated almost to the end point (1-2 mg/1) before starch
or thyodene is added to obtain a near purple-like color.
If they are added too early, colored, undissolved particles
remain, which affect the true end point.
*Remarks in brackets are those of the authors.
t"The mass of a slightly soluble gas that dissolves in a definite
mass of a liquid at a given temperature is very nearly directly
proportional to the partial pressure of that gas. This holds for
gases which do not unite chemically with the solvent," (Handbook
of Chemistry and Physics, 43rd Edition).
92
-------
c) In some applications, such as in sewage aeration basins, it
is difficult to obtain a representative sample and to obtain
consistent results because of the suspended solids present.*
2. The Hays Company developed a DO analyzer based on the paramagnetic
properties of oxygen, but the device employs a rather cumbersome
magnet. This instrument was investigated by J. WeeksA~3 and
requires no further comment.
3. HertzA~4 and Mancy^~ have summarized the history and operation
of galvanic cells. Based on their investigations and those of
others,A-4,A-5 companies such as Union CarbideA~6 and American
LimneticsA~7 developed DO sensors that employed thallium cathodes.
The Union Carbide sensor assembly used a silver-silver chloride
reference electrode but performance was unsatisfactory. In
these two designs, the exposed thallium became poisoned by
impurities that affected overall response. The Union Carbide
sensor response was also affected by particles of thallium that
dropped from the base metal, and left voids. The fault was
traced to manufacturing procedures.A~^
4. Mancy and WestgarthA and others pioneered the development of
membrane-covered galvanic cells. Generally, the galvanic cell
employs a platinum cathode, lead anode, either polyethylene or
teflon membranes, and either KI or KOH electrolyte.
Weston and Stack, one of many companies that manufacture a DO
parametric system, employs a galvanic cell, teflon membrane,
platinum cathode, lead anode, potassium iodide electrolyte, and
a self-contained sample agitator.A'10 In the absence of oxygen,
the output of the sensor is virtually zero.A"11
5. The passive DO sensor (electrolytic cell) is provided with
devices similar to those found in a membrane-covered galvanic
cell, but an external reduction potential is applied and, in the
presence of oxygen, current is generated.
Membrane
Both teflon and irradiated polyethylene membranes have been
employed. Honeywell employs 1-mil teflon in the sensor
*J. Winkler indicated that the Hamilton, Ohio, plant has been employing
instrumental DO parametric systems instead of the Winkler process for
these reasons.
93
-------
investigated. Although somewhat overly simplified, the "paths"
("canals") through which oxygen diffuses in the membrane to the
cathode seem to be randomly located in teflon membranes but more
systematically located in irradiated polyethylene. A sketch of
this simplified concept is illustrated by Figure A-l, without
proof,A-12 ancj each section is symbolic of a unique membrane.
A1
A2
O o
°0
o
O
o
o
0^
n °
0°'
A3
"CANALS'
n
H_
1 1
u
n
| |
B_
i
i
• —
i
I
u
u
n
i.
i
i
u
u
A3 POLYETHYLENE
TEFLON
FIGURE A-1. SIMPLIFIED CONCEPT OF MEMBRANE "CANALS".
It is evident that section Al of the teflon membrane has a dif-
ferent sensitivity than does section A4. Since the polyethylene
membrane exhibits more consistency among sections, it is more
consistently sensitive to the diffusion of molecular oxygen.
Although considered to be more favorable than polyethylene from
a maintenance viewpoint, teflon appears to have "plastic memory"
because temperature cycling affects "sensitivity" (see Laboratory
Investigation Section). Polyethylene membranes lost their sensi-
tivity within several hours when used to measure DO under
supersaturated (20 mg/1) conditions in the Little Miami River.
The loss was traced to the growth of rotifers, stalk protozoa,
diatoms, and fibers* on the membrane.
*Identified by Dr. C. Weber, Chief, Aquatic Biology Section, EMSL-
Cincinnati, Ohio.
94
-------
Sample Velocity at Membrane
Oxygen that has diffused through the membrane and is reduced at
the cathode must be replenished at the sample-membrane interface
to maintain measurement integrity. This is effected by providing
sample velocities greater than 1.0 foot per second (30.5 cm/sec)
across the membrane. The effect of sample velocity on diffusion
current is approximated by
i(v) = K(l-Ae"9v) (A_D
where i(v) = current produced by reduced oxygen,
amperes
K, A, 3 = constants of proportionality
v = velocity of sample, cm/sec
Most manufacturers consider 1.8 feet per second (55 cm/sec)
adequate velocity but the Dutch company, Philips, employs 6.6
feet per second (2 m/sec) velocity over the sensor area which
also aids in scrubbing the sensors.
Some manufacturers provide vertical flow patterns (this is also
recommended by Honeywell)* perpendicular to the membrane surface,
but in some cases the vertical impingement on the sample has
caused mud and sand particles to adhere to or penetrate the mem-
brane, which damages it and lowers its sensitivity. For one DO
parametric system,A-13 ^he authors designed a nozzle which
provided an almost horizontal flow over the membrane, and the
main stream line was directed at the center of the sensor
membrane starting 10° below the horizontal. Flow cells designed
by the Schneider Instrument Company followed a similar philosophy
employing tangential influent and effluent paths. By mounting
the Schneider sensors off center in these cylindrical cells,
maximum benefits of "high velocity" at nearly horizontal flow
are obtained.A-14
Sensor Electrodes
In accordance with EPA's specification/-15 the Honeywell employs
a gold cathode and silver anode.
selected applications.
95
-------
Platinum cathodes have been used in some other sensors, but
experiments suggested that oxides form on the cathode and de-
creases its sensitivity.
Reasonably high surface-area ratios (anode/cathode) are pre-
ferred for long-term stability, but ratios between 25 to 100
have proven adequate for 6 to 12 weeks of operation. Sensor
maintenance frequency rises as the ratio decreases. The
Honeywell sensor has a ratio of approximately 400 and employs
a gold cathode alloyed with small quantities of platinum-silver
to improve its machining capability. The button is approximately
1/8 inch (0.318 cm) in thickness -to 3/32 inch (0.237 cm) at the
center. There is no apparent difference between the performance
of this cathode working against its silver anode and the gold-
silver electrode assemblies.
In several experiments performed by the authors on designs employ-
ing gold cathodes, silver ions plated-out on the gold shoulder
exposed to the electrolyte and produced a zero offset current.
Extreme care must, therefore, be exercised when the assembly is
manufactured.A"12 In the absence of high quality control, an
auxiliary zero offset circuit compensates for "silver ion"
current.
The migration of silver ions to the" cathode was predominant in
one design that employed carbon, because silver crystalized near
the cathode-membrane junction. This eventually led to a short-
circuited sensor, because the anode and cathode were only 1/2
inch (1.77 cm) apart.A-13 -p^g manufacturer attempted to "filter"
the silver crystals and/or silver ions by adding Koalin to the
electrolyte but did not succeed. Using a similar concept,
however, Gealt developed a screen that inhibits oxides of silver
(sludge) from settling on the cathode, thus preventing reduction
in sensitivity, spurious signals, and offset currents.A~16
Maclnnes describes a silver coulometer which is functionally
similar.A-17 The coulometer consists of a platinum cathode bowl
filled with a silver nitrate electrolyte, a silver anode, and a
cup of porous ceramic material that surrounds the silver. He
states, "During electrolysis, the anode disintegrates to some
extent so that particles become detached and drop off. At the
same time a dense 'anode slime1 is formed, the composition of
which is still in doubt. The porous cup protects the cathode by
96
-------
catching the particles which fall from the anode and it also
obstructs the diffusion of the anode slime into the cathode
chamber." The effectiveness of the screen (cup) depends on the
manufacturing techniques discussed later.
Electrolyte
There are several types of electrolytes, including KC1 and KOH.
The normalitv of a given electrolyte affects the transient
response.A" Additionally, Mancy, et al, also note that car-
bonates form within the electrolyte because the membranes
discussed are permeable to carbon dioxide, which reduces their
sensitivity. ~-L°
Honeywell employs a combination of KC1 and KOH (2N KCL and 0.5N
KOH).
In the absence of KOH, Honeywell says that hydrogen peroxide will
probably form and raise the output during sensor break-in.
Tests indicate that sensors require a 1- to 4-hour break-in; this
is normally performed at the company before installation. Under
selected conditions, it would seem that the KOH generated (with
KC1 electrolyte) during break-in would be sufficient to reduce
the amount of hydrogen perioxide formed and lower the output
increase observed by Honeywell. The authors did not, however,
run tests to substantiate this possibility.
Temperature Compensation
The gas-permeable membrane is temperature sensitive so that at a
constant pressure and DO concentration, an increase in tempera-
ture causes more oxygen* to diffuse through the membrane, thereby
increasing the sensor output. Thus, in saturated samples, an
uncompensated sensor produces more output at 95F (35C) than at
35F C1-7C).
Honeywell locates the total compensating network at the sensor
to generate an mv potential that is proportional to the DO con-
centration (in fresh water). This approach is similar to that
reported by Carritt and Kanwisher.A~^
Other designers locate the thermal components with the sensor,
but schematically show them as being in the signal-conditioning
device. Related network components are actually located in the
si'gnal conditioner.
*0xygen activity also is affected as noted in the references.
97
-------
Thermal response depends on the total mass of the sensor assembly,
including the electrolyte, body, thermistor, and membrane. To
achieve a balanced thermal response, Honeywell has mounted the
thermistor so that it is exposed to the sample. This arrangement
obtains 98 percent of a step change in temperature (and DO) within
2 minutes.
Reduction Potential
In the presence of 800 mv, oxygen that has diffused through the
membrane into the sensor assembly is reduced at the negative gold
cathode. Experiments have shown that the output for selected sen-
sors remains relatively constant when the reduction potential
plateau lies between 550 and 900 mv.
However, in experiments performed in EMSL on 10 DO sensors
obtained from the Schneider Instrument Company, the transient
response was improved when the reduction potential was reduced
from 800 to 650 mv. The authors thought the improvement may have
been related to the manufacturing technique used and the fact
that the diffusion layer emanated away from the sensor cathode
into the immediate electrolytic film. At 800 mv, excessive over-
shoot occurred when the sensor was transferred from the RT to the
CT mode (saturated, 65F [18.3C] to 35F [1.7C]). At 650 mv, the
overshoot was reduced, and the response was within tolerance in
less than 2 minutes compared to 3-5 minutes when the reduction
potential was higher. (In these tests, the temperature compensa-
tion and analyzer were the same, as were all other conditions,
except the sensor.)
With increased reduction potential, other gases can be reduced
and the output current increased.A-21
In one design, when the anode was internally connected to about
5.0 volts DC during a test, gases formed and the increased
internal pressure caused the membrane to balloon, ~ but this
problem was eliminated by changing EPA's specifications and
design parameters.
In the Honeywell design, the sensor assembly generates a poten-
tial that can be connected or not connected. Hence, prolonged
temperature-stability tests of the analyzer employing the
"check" will not affect the sensor.
98
-------
Sensor Body
In cooperation with the Schneider Instrument Company in an
experiment employing a teflon body, EMSL personnel found that
the response was erratic during transient analysis. A
smooth response was obtained under similar conditions when a
PVC body was tested. Improved performance with PVC was traced
to its lower temperature coefficient (89 ycm/cm/°C). Teflon,
which has a temperature coefficient of 147 ycm/cm/°C contracted
at lower temperature and this caused the membrane to balloon.
This, in turn, increased the electrolytic film between the
cathode and membrane and affected the response.
Most companies currently employ PVC as the base material for the
DO sensor.
OTHER CONSIDERATIONS
In the beginning of this appendix, single point measurement of dis-
solved oxygen for possible aerator control application was implied.
The Honeywell system lends'itself to multipoint (m) measurements and
"average point" control philosophy. The sensor assembly produces a
temperature compensation potential linear in dissolved oxygen so
that several sensor outputs can be summed in an operational ampli-
fier with a gain of 1/m, and then connected to the input of the MV/V.
The single output of the MV/V would serve as the "average control"
potential.
99
-------
APPENDIX A
REFERENCES
A-l. Ballinger, D. G., "Decisions to be Made in the Use of Automatic
Water Quality Monitors," presented at the International Symposium
on Identification and Measurement of Environmental Pollutants,
Ottawa, Canada, June 14-17, 1971.
A-2. American Public Health Association, "Standard Methods for the
Examination of Water and Wastewater," 13th Edition, New York,
1971.
A-3. Weeks, J. D., "Evaluation of the Field Performance of a Hays
Dissolved Oxygen Analyzer," National Water Quality Network
Applications and Development Report #6, Water Quality Section,
Basic Data Branch, DWSSPC, U.S. Public Health Service, Cincinnati,
Ohio 45268, October 1962.
A-4. Hertz, P., "Galvanic Analysis," from Advances in Analytical
Chemistry and Instrumentation, Vol. 3, J. Wiley, New York, 1964.
A-5. Mancy, K. H., and Jaffe, T., "Analysis of Dissolved Oxygen in
Natural and Waste Waters," PHS Grant WP 00566, Public Health
Service Publication No. 99-WP-37, April 1966.
A-6. O'Herron, R. J., "Performance of the Union Carbide Dissolved
Oxygen Analyzer," Report #EPA-670/4-73-018, NERC, Office of
Research and Development, U.S. Environmental Protection Agency,
Cincinnati, Ohio 45268, December 1973.
A-7. O'Herron, R. J., "Performance Evaluation of the American
Limnetics Instruments Dissolved Oxygen Meter," unpublished
report, NERC, Office of Research and Development, U.S. Environ-
mental Protection Agency, Cincinnati, Ohio 45268, December 1973.
A-8. Private communication between R. Yaeger, Union Carbide, and the
authors.
A-9. Mancy, K. H., and Westgrath, W. C., "A Galvanic Cell Oxygen
Analyzer," Journal Water Pollution Control Federation, October
1962.
A-10. Weston and Stack, Inc., Malvern, Pa., R. Evangelista, General
Manager.
100
-------
A-ll. Unpublished notes on experiments performed in the EMSL by the
authors on the Weston and Stack dissolved oxygen parametric
system.
A-12. From experiments performed with The Schneider Instrument Com-
pany on several contracts such as PH-86-62-90 and WA 67-17'A.
A-13. From experiments performed in cooperation with the ProTech
Company on Contract No. 47-4-11-63.
A-14. From unpublished experiment performed at the EPA's Great Miami
River Research Station, Hamilton, Ohio, by J. Teuschler, J.
Griffith, and the authors.
A-15. Mentink, A. F., "Specifications for an Integrated Water Quality
Data Acquisition System - 8th Edition," FWPCA, Division of Pol-
lution Surveillance, U.S. Department of the Interior, Cincinnati,
Ohio 45268, 1968.
A-16. Patent #3,510,421 issued to A. Gealt, Honeywell Inc., Fort
Washington, Pa.
A-17. Maclnnes, D. A., "The Principles of Electrochemistry," Reinhold
Publishing Corporation, International Textbook Press, Scranton,
Pa., 1939.
A-18. Mancy, K. H., Okun, D. A., and Reilley, C. N., "A Galvanic Cell
Oxygen Analyzer," Journal of Electroanalytical Chemistry, Vol.
4, 1962.
A-19. Private communication, A. Gealt, Honeywell Inc., and the authors,
June 1974.
A-20. Carritt, D. E., and Kanwisher, J. W., "An Electrode System for
Measuring Dissolved Oxygen," Analytical Chemistry, Vol. 31, No.
1, January 1959-
A-21. Gealt, A. E., and Metarko, R. P., Honeywell Inc., "Reliable
Measurement of Dissolved Oxygen in Polluted Waters," presented
at ISA Conference and Exhibit, Philadelphia, Pa., October 1970.
101
-------
APPENDIX B
THEORY
Only selected considerations on theory of sensor components behavior
are provided below as the literature included theoretical aspects in
great detail.B-l,B-2
When molecular oxygen diffuses through the membrane and the electro-
B-3
lytic film to the cathode (-800 mv), the following reaction occurs.
4e + 02 + 2H20 -*• 40H" (B-l)
in which e is a negative charge.
When the electrolyte is KC1, the +800 mv silver anode reacts with
the chloride ion as follows. "^
4Ag + 4C1 -»• 4AgCl + 4e (B-2)
When the electrolyte is KOH, the silver anode is oxidized:
4Ag + 40H~ -> 2Ag20 + 2H + 4e (B-3)
The current measured can be calibrated in terms of the oxygen content
(mg/1) of the sample, dependent upon design as the reduced oxygen is
a function of oxygen activity, temperature, and partial pressure.
Mancy notes that the total solubility of oxygen at constant pressure
is related to oxygen activity;B-5
P = HAQ = H6C (B-4)
where P = partial pressure
H = Henry's Law constant
AQ = activity of molecular oxygen
3 = activity coefficient
C = concentration of DO
102
-------
He further shows that the diffusion current (proportional to reduced
oxygen) is given by
(B-5)
where id = diffusion current, amperes
F = Faraday constant
z = the number of electrons transferred
A = cathode area
P = membrane permeability
b = membrane thickness
Briggs shows a similar relationship, except that Ao2 is replaced
with Cs, the partial pressure of oxygen in the sample.B-6
Because large anode/cathode surface ratios have been noted,B~^ an
element of qualification is necessary in considering equation B-5.
A high ratio obtained with a small cathode diameter tends to produce
a lower sensitivity, which requires a change in the analyzer design.
Additionally, if an offset current exists, its magnitude could be
comparable to that of the signal current and also have an effect on
the design. Normally, a cathode having a nominal diameter of 0.20
inches (0.51 cm) and an anode/cathode ratio of 40 or more is suitable
for continuous field monitoring.
If it assumed that the electrolytic film between the inside of the
membrane wall and the cathode surface is infinitesimal and that
instantaneous reduction takes place as the oxygen passes from the
membrane, the inside surface of the membrane acts as a sink under
steady-state conditions, and the gradient in the permeation equation
of Pick's Law becomes a constant,*
*The reader is referred to Reference B-2 to solve Pick's second law
for the transient response.
103
-------
P . -D CB-6)
where P = permeation
D = diffusion coefficient
3C
— = concentration gradient of oxygen through the membrane
x (axis per pendicular to membrane surface) .
Hence, the permeation can be rewritten as
C - C DC
P = D
a i
where GI = incoming oxygen concentration and assumed constant
C? = outgoing oxygen concentration
H = membrane thickness
Under the assumption presented above, C2 is zero and P is reduced
to the approximation.
This is obviously simplified since some manufacturers provide thin
tissue between the membrane and cathode to assure that electrolyte
is present at the cathode. In absence of tissue, capillary action
provides electrolyte at the cathode.B-8 The cathode need not be
restricted to a line, as implied by equation B-6, but can be curved,
so that spherical coordinates may be a closer approximation.
Barrer&~9 summarizes the principal characteristics of the permeation
process regarding the passage of gases through membranes and states:
j\r
"(1) The Pick diffusion Law P = D -p- (P denotes the
permeability and D the diffusion constant) is
true in the stationary state. [The authors
call this 'steady-state..']
(2) Stationary flow is established in a period of
minutes, the actual time depending upon the
temperatures. [Refer to the transient response
data on the Honeywell sensor.]
104
-------
(3) The premeation rates are usually proportional
to the pressure and inversely proportional to
the thickness of the membrane [see equation
B-5 above] .
(4) The velocity of diffusion is only slightly
altered by roughening the outgoing surface.
[This is not to imply that diffusion occurs
through a choked 'canal' in the membrane.]
(5) The passage of an electric discharge through
either the glass wall or through the gas
(hydrogen or helium) has no effect upon the
diffusion rate.
(6) The process of permeation through these (glass)
membranes is highly selective and markedly tem-
perature dependent ...... " [Note discussion on
temperature compensation.]
Oxygen profiles in membranes and electrolyte are discussed by Mancy
as a time function. After the transient response has been com-
pleted, Mancy 's approximation for steady-state DO concentration
through the membranes agrees with that derived from equation B-6.
Briggs and Viney^-10 show how temperature affects the DO cell output
current by the following equation:
i(T) = Ke"J/T (B-7)
where i(T) = cell output current
K, J = constants for a particular cell geometry and
cell membrane
T = temperature, degrees Kelvin
and indicate that J has a value of 4500°K for polyethylene membranes.
Temperature effects of the DO membrane are eliminated by multiplying
equation B-7 by
B/
and setting the magnitudes of B2 and J equal
105
-------
This form of compensation is inherent in most DO parametric systems.
Data consistent with Briggs and Viney~ were also obtained by the
authors on EPA contracts. "^ Comparable data were obtained for the
Honeywell sensor with and without temperature compensation.
In samples having relatively high salt content, DO determinations by
the Winkler process indicate a lower value than that obtained from
the sensor. MancyB~12 notes that current due to reduced oxygen is
an exponential function of ionic strength, AI, of the sample:
K AI
i = ACe s (B-9)
where i = sensitivity at a reference state, amperes per
mg/1 DO
C = DO concentration, mg/1
K = salting out coefficient
o
AI = ionic strength
B'-13
The ionic strength is defined as :
2
Ym.Z.
z •£-• i i
where m. = molar concentration of ions
Z. = ionic charge
To reduce the oxygen current, i, to direct proportionality in AC and
thereby eliminate the salting out effects, it is evident that an
equivalent miltiplier of
-F AI
B e s
o
be generated. This can be developed by a special conductance para-
metric system (MancyB"12 notes that there is a linear relationship
106
-------
between conductance and ionic strength under selected conditions)
and analog multiplication module. The Honeywell system investigated
was not provided with salt compensation.
The philosophy of instrumental DO measurements is to approximate*
Winkler determinations (as accurately as practical under selected
conditions). Temperature compensation has, therefore, become an
integral component of field measurement systems. Compensation for
salinity is more sophisticated and costly, and its use has been
avoided. In offshore oceanographic applications unaffected by
estuaries, the conductivity is reasonably constant, and salinity
effects on the DO concentration can be corrected by periodic field
calibration.
Depending on the DO measurement tolerances, salinity correction may
be necessary in estuarial studies, or in expanded monitoring programs
involving selected industrial discharges. In these operations, com-
pensation can be effected by: (1) periodic, manual instrument
recalibration; (2) automatic compensation.
Since the conductivity could be reasonably predicted for estuarial
studies (or actually measured by a conductance instrument along with
the DO measurement) either manual or automatic corrrection at the data
processing center could be employed. Thus, salinity compensation in
this application is a convenience, but manpower (for frequent cali-
bration checks) may not be available.
Saline discharges from selected industries may not be as predictable
as tidal effects, therefore automatic compensation has merit.
Mancy has provided an instrumental method for temperature and salinity
compensation for DO measurement.8"3 His circuit first compensates
sensor output for a change in salinity and lastly for a change in
temperature. It should be noted that both membrane permeability and
conductance are temperature dependent.
Table B-l summarizes the levels of saturated DO as a function of tem-
perature and chlorides.
*It is well known to the laboratory chemist that in some instances
the Winkler determination must be modified to accurately determine
the DO concentration--hence, the chemical components of the water
sample must be known. All DO parametric systems are, however,
calibrated in accordance with the Winkler determination.
107
-------
Table B-l. SATURATED DO FOR SELECTED VALUES OF
TEMPERATURE AND CHLORIDES*
Temperature Chloride concentration - mg/1
C 0 5,000 10,000 15,000 20,000
Dissolved oxygen - mg/1
0
5
10
15
20
25
30
35
14.6
12.8
11.3
10.2
9.2
8.4
7.6
7.1
13.8
12.1
10.7
9.7
8.7
8.0
7.3
13.0
11.4
10.1
9.1
8.3
7.6
6.9
12.1
10.7
9.6
8.6
7.9
7.2
6.5
11.3
10.0
9.0
8.1
7.4
6.7
6.1
*From Standard Methods, 13th Edition, Pages 480-481.
108
-------
APPENDIX B
REFERENCES
B-l. Mancy, K. H., Okun, D. A., and Reilley, C. N., "A Galvanic Cell
Oxygen Analyzer," Journal of Electroanalytical Chemistry Vol
4, 1962.
B-2. Mancy, K. A., and Okun, D. A., "Oxygen in Waste Treatment - The
Galvanic Cell Oxygen Analyzer," special report on Research Grant
WP 131 (C7), National Institute of Health, University of North
Carolina, Chapel Hill, North Carolina, December 1961.
B-3. Mancy, K. H., and Jaffe, T., "Analysis of Dissolved Oxygen in
Natural and Waste Waters," PHS Grant WP 00566, Public Health
Service Publication No. 999-WP-37, April 1966.
B-4. McKeown, J. J., Brown, L. C., and Gove, G. W., "Comparative
Studies of Dissolved Oxygen Analysis Methods," JWPCF, August
1967.
B-5. Mancy, K. H., "Water Quality Monitoring - The Sensor System,"
Proceedings of the Specialty Conference on Automatic Water
Quality Monitoring in Europe, March 1971.
B-6. Briggs, R., "Monitoring Water Quality in the United Kingdom,"
Proceedings of the Specialty Conference on Automatic Water
Quality Monitoring in Europe, March 1971.
B-7. Correspondence to W. Westgarth, Oregon State Board of Health
from authors, October 1965.
B-8. Hertz, P., "Galvanic Analysis" from Advances in Analytical
Chemistry and Instrumentation, Vol. 3, J. Wiley, New York,
1964.
B-9. Barrer, R. M., "Diffusion In and Through Solids," University
Microfilms, Inc., Ann Arbor, Michigan, 1962.
B-10. Briggs, R., and Viney, M., "The Design and Performance of
Temperature Compensated Electrodes for Oxygen Measurements,"
Water Pollution Research Laboratory, Stevange, Herts, England,
1963.
B-ll. From experiments performed with The Schneider Instrument Com-
pany on several contracts such as PH-86-62-90 and WA 67-17A.
109
-------
B-12. Mancy, K. H., "Instrumental Analysis for Water Pollution
Control," Ann Arbor Science Publishers, Inc., Ann Arbor,
Michigan, 1971.
B-13. Skoog, D. A., and West, D. M., "Fundamentals of Analytical
Chemistry," Second Edition, Vol. 1, Reinhart and Winston, Inc.,
1969.
110
-------
TECHNICAL REPORT DATA
(tlease read Instructions on the reverse before completing)
REPORT NO.
EPA-600/4-77-023
3. RECIPIENT'S ACCESSION-NO.
TITLE AND SUBTITLE
INVESTIGATION OF A HONEYWELL DISSOLVED OXYGEN
PARAMETRIC SYSTEM
5. REPORT DATE
April 1977 issuing date
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
A. F. Mentink, J. 0. Patterson, and T. E. Hickman
8. PERFORMING ORGANIZATION REPORT NO.
t. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Lab.
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Gin., OH
10. PROGRAM ELEMENT NO.
1HD621
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Same as above
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/06
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A Honeywell dissolved oxygen parametric system was investigated for
possible application in EPA's research on sewage treatment. Labora-
tory and field data were accumulated. Summaries on selected background
and theoretical aspects of the measurement have been included for those
unfamiliar with this type of instrumentation.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
:. COS AT I Field/Group
Dissolved oxygen; Instrumentation;
Measurement of DO in sewage treatment
plant; Investigation of DO parametric
system; Water Pollution.
DO measurement in sewage
plant effluent;. DO
measurement in aeration
tank; Background on DO
measurement.
13B
18. DISTRIBUTION STATEMENT
Release to public
^•^MMMMMMMM^^M^
EPA Form 2220-1 (9-73)
19. SECURITY CLASS (ThisReport)'
Unclassified
121
20. SECURITY CLASS (This page)
22. PRICE
Ill
•fr U. S. GOVERNMENT PRINTING OFFICE: I977-757-056/559& Region No. 5-11
------- |