EPA-600/2-76-146
October 1976
Environmental Protection Technology Series
WASTEWATER SAMPLING, TRANSFER AND
CONDITIONING SYSTEM
Municipal Environmental Research 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 five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconornic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-76-146
October 1976
WASTEWATER SAMPLING, TRANSFER
AND CONDITIONING SYSTEM
by
Louis S, DiCola
Raytheon Company
Portsmouth, R.I. 02871
Contract No. 68-03-0250
Project Officer
Robert H. Wise
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
The Environmental Protection Agency was created because of
increasing public and government concern about the dangers of pollu-
tion to the health and welfare of the American people. Noxious air,
foul water, and spoiled land are tragic testimony to the deterioration
of our natural enviroment. The complexity of that environment and
the interplay between its components require a concentrated and
integrated attack on the problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact,
and searching for solutions. The Municipal Environmental Research
Laboratory develops new and improved technology and systems for the
prevention, treatment, and management of wastewater and solid and
hazardous waste pollutant discharges from municipal and community
sources, for the preservation and treatment of public drinking water
supplies, and to minimize the adverse economic, social, health, and
aesthetic effects of pollution. This publication is one of the
products of that research; a most vital communications link between
the researcher and the user community.
To help implement the above, this study describes development
of an automatic on-line sampling, transfer and conditioning system for
monitoring wastewater-treatment process streams.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
CONTENTS
Page.
DISCLAIMER ii
FOREWORD iii
FIGURES vii
TABLES viii
ACKNOWLEDGEMENTS ix
I INTRODUCTION
II SUMMARY
III CONCLUSIONS 5
IV RECOMMENDATIONS 7
V PRELIMINARY INVESTIGATION AND ACCEPTANCE TESTING q
OF COMPONENTS
Sample Transfer Pump ^
Homogenizer
Filter n
Component-Testing Manifold 13
Test Location 13
Acceptance Testing of Components 13
First Tests of Particle Size -,0
Further Tests of Particle Size ^g
Discussion of Particle-Size Testing 22
Pipe-Size Consideration 22
Filter Tests 23
Some Observations 23
Conclusions
25
VI FINAL SYSTEM DESIGN
Establishing a Sampling Procedure 25
-------
Monitoring Both TOG and SOC 26
System Description 27
Use of Dilution 27
Adapting the Dilution Pump 36
Overcoming Intermittent Flow 37
Timing of Samples 38
Comparison with Standard Analytical Methods 39
VII TEST RESULTS 41
Reference Tests 41
Sample Transfer and Conditioning System Test Data 41
Performance of Automatic Analyzers 41
Comparison of Source and Interface Values 46
Test Results from Automatic Analyzers 46
VIII REFERENCES 48
APPENDIX A - Statistical Analysis 49
APPENDIX B - Operation and Maintenance 52
APPENDIX C - Design Specification Guidelines 62
APPENDIX D - List of Equipment 67
GLOSSARY OF TERMS AND ABBREVIATIONS 69
vi
-------
LIST OF FIGURES
Number Page
1 Hydr-0-Grind Pump 10
2 Raytheon In-Line Homogenizer 12
3 Preliminary Sampling Manifold Flow Diagram 14
4 Site Description 15
5 Millipore Filter Apparatus 17
6 Sieve Assembly 19
7 Raw Sewage, Particle-Size Reduction with Homogenizer. ... 20
8 Flow Diagram of the System 28
9 Sampling System, Front View 29
10 Sampling System, Rear View 30
11 Control Panel, Front View 31
12 Control Panel, Rear View 32
13 Manifold Assembly, Front View 33
14 Typical Hydr-0-Grind Pump Installation 34
15 Typical Duplex Dilution Pump Installation 35
16 Sampling Sequence 40
B,l Dummy Plug Wiring to Skip SOC Mode 54
B.2 Timing Diagram 55
B.3 Typical Valve Pair 58
B.4 Ladder Wiring Diagram 59
vii
-------
LIST OF TABLES
Number Page
1 PRESERVATION OF THE INTEGRITY OF CHEMICAL
COMPOSITION DURING THE COURSE OF TRANSPORT AND
CONDITIONING OF RAW SEWAGE 21
2 RESULTS OF FILTER TESTS 23
3 SAMPLING MATRIX 25
4 SAMPLE TRANSFER DATA 38
5 SAMPLE TRANSFER AND CONDITIONING SYSTEM TEST DATA: 42
SECONDARY EFFLUENT 42
PRIMARY EFFLUENT 43
RAW INFLUENT. 44
MIXED LIQUOR 45
RETURN ACTIVATED SLUDGE . . . 45
B. 1 TABLE OF OPERATION SEQUENCE 45
vlli
-------
ACKNOWLEDGMENTS
The support of Anthony Ventetuolo, Superintendent of the Water Pollution
Control Facility, Cranston, Rhode Island, is acknowledged with sincere thanks.
A. Joseph Mattera, Foreman, also provided valuable assistance.
Sincere thanks to Walter Schuk of the EPA for his guidance and assistance at
the outset of system testing.
The preliminary investigations, as well as all system designing, construction,
and testing, were performed with the invaluable assistance of H. Duane Evans.
ix
-------
SECTION I
INTRODUCTION
The purpose of this project was to develop automatic, on-line equipment for
sampling, transferring and conditioning wastewater-treatment process streams
for automated analyses for total and soluble organic carbon, ortho- and total
hydrolyzable phosphate, ammonia, nitrate, and nitrite. Furthermore, these
sampling, transferring and conditioning steps were to be accomplished without
causing unacceptable chemical changes in the sample prior to any of the
analyses. Ultimately, automated sample-handling equipment of this type will
be a necessary component of completely automated, wastewater-treatment
processes and plants.
The major factors requiring consideration were: a) the necessity of limiting
the size of suspended solids particles in samples that could not be filtered
prior to analysis, b) fabrication of a suitable automatic manifolding and
switching network, c) assurance that all samples would be representative
of the process streams from which they were taken, and d) total system
reliability. A brief discussion of each of these factors follows.
Limiting the Size of Suspended Solids Particles
The suspended solids and refractory matter present in wastewater prevent the
use of typical off-the-shelf colorimetric or organic carbon analyzers without
sample pretreatment.
The total organic carbon (TOC) analyzers presently used in most wastewater-
treatment laboratories accept samples of 50 inicroliters, or less. These
relatively small volumes are required by the small injector assemblies and
combustion reactors incorporated in such laboratory analyzers. In addition,
multiple injections of a well-homogenized sample are required for obtaining
reliable data. Regardless of TOC-analyzer design, however, incoming samples
must meet suspended solids limitations not only on average particle size,
but also on particle-size range_; otherwise, not only would the analyzer tend
to clog, but the analyzer's data would contain intermittent and unpredictable
outliers or data "spikes" that could ruin much of the total output.
Colorimetric analyzers provide false data when high suspended solids concen-
trations are present in the sample streams. Large particles tend to clog
the colorimeter's automatic delivery system, while small particles limit light
transmission. Finely suspended material causes backscatter (Tyndall effect),
and this also creates artificially high absorbance values. Thus, in either
case the true value is masked.
-------
Accurate data can be obtained only through proper conditioning (i.e.,
homogenization or filtration) of the sample before the sample enters
the analyzer.
Automatic Manifolding and Switching
The extended retention time of a typical wastewater-treatment process dampens
the short-term changes in most measured variables. This permits analytical
time sharing whereby a single analyzer can be used to monitor several process
streams without significant loss of data; although occasionally, full-time
monitoring of a process variable may be necessary.
The optimum situation is to supply a continuous flow of several different
types of samples to a centrally located, valved manifold. At this location,
the sampling sequence, sampling time and sample conditioning would be controlled
by an automatic switching device, with a manual override for experimental work.
A centralized monitoring system offers the following benefits:
1. fewer analyzers to purchase,
2. fewer analyzers to maintain,
3. reduced chemical consumption,
4. easier surveillance,
5. easier isolation from the hostile environment of a wastewater-
treatment plant.
Centralization, however, often requires long transfer distances which can
cause unreliable data. Therefore, proper sizing of transfer lines to obtain
optimal flow must be taken into consideration.
Assuring Representative Samples
Of primary importance in any sampling and conditioning system is whether or
not the sample taken is representative of its source, and whether the sample
has undergone a chemical change as a result of the conditioning process. This
factor is dealt with by proper sampling, sample transfer (i.e., transport),
and sample homogenization or filtration.
Reliability
Most automatic sample transfer and conditioning systems have been so poorly
designed and/or mechanically unreliable that the chemical integrity of the
transported sample has received little or no attention. To be successful, an
automatic sampling system must utilize essentially troublefree hardware that
has been thoughtfully integrated into a highly reliable system capable of
continuous unattended operation.
With these facts in mind, Raytheon designed and constructed an automatic, on-
line, wastewater-sample transfer and conditioning system to make automated
process stream analysis and process control possible, and thus fulfill a need
which is daily becoming more urgent.
-------
Test Location
The Water Pollution Control Facility for the City of Cranston, Rhode Island,
was chosen as the test facility for this program. The facility utilizes an
activated sludge process and has a daily flow of 6.84 million gallons (26,500
cu m). The plant's size and population it serves (75',000 persons) make it a
typical operation. The proximity of the plant to Raytheon1s facility, as
well as the good working relationship between Raytheon and the City of
Cranston through the years, were additional factors that influenced the
decision to conduct the study at that particular location.
Although the Cranston facility has adequate laboratory and office space to
accommodate the project's requirements, piping the sample transfer lines to
the central laboratory would have been most awkward. Also, such an arrange-
ment would have created a safety hazard for plant operators and maintenance
men since the piping would be crossing heavily traveled areas. Instead, a
Raytheon owned, environmentally controlled (heated and air conditioned)
instrument trailer was set up at the Cranston facility to house the system to
be tested.
This program, which extended over a 15-month period, was carried out in two
phases. Descriptions of both are reported herein:
Phase I - Preliminary Investigation and Qualification of Components
Phase II- Design, Implementation, Testing, and Evaluation of the
Final Sampling System
-------
SECTION II
SUMMARY
This report describes the construction and field evaluation of an automatic
on-line hardware system for reliably sampling, transferring, and conditioning
various wastewater-treatment process streams such that the resulting transferred
and conditioned samples are suitable for interfacing with automatic on-line
colorimetric and total organic carbon analyzers. Process streams to which
this hardware system was successfully applied included raw sewage, primary
effluent, secondary effluent, aeration tank mixed liquor, and return activated
sludge. Primary sludge could not be sampled at the field-testing site because
the sludge had become too thick at its only feasible access point. Analytical
parameters used to evaluate the hardware system included both total and soluble
organic carbon, orthophosphate, total hydrolyzable phosphate, and ammonia
nitrogen. Nitrate and nitrite were not included; however, the hardware system's
performance with the soluble parameters studied indicate that nitrate and
nitrite should present no special difficulties.
-------
SECTION III
CONCLUSIONS
General
The sample transfer and conditioning program described in this report has
demonstrated that various streams within a typical municipal wastewater-
treatment plant can be monitored remotely and reliably for TOC, SOC, o-PO^
hydrolyzable P04, NH3~N, and N03/N02-N.
Sampling, transferring and conditioning was accomplished
reliably and continuously without affecting the representative
nature of the sample except for particle size distribution.
Comparison of wastewater sources and interface sample discharge
concentrations, as measured by reference laboratory procedures,
demonstrated very satisfactory agreement.
The agreement in reference laboratory TOC values for source and
interface proves that sampling of streams containing particulate
matter need not be a problem if the following simple rules are
followed in designing the system:
1. Fluid velocities at, or greater than, 2 ft /sec should be
maintained in the sample-transport lines.
2 . A sampling manifold that keeps all sample streams flowing
continually must be provided.
3. For the automatic modes, a sampling sequence must be established
for sampling the cleanest stream first, then sampling progressively
dirtier streams. At the end of each such sequence of samples,
a complete flushing of the system with clean water must be carried
out.
4. All fittings, pipes and other wetted components in the sample-
transport and manifold systems must be designed to eliminate
restrictions and dead zones wherever possible.
Sample dilution is a viable approach and, if implemented correctly,
offers the following benefits: a) multi-stream monitoring, using a
single transfer system for high solids and low solids sources,
b) minimization of transfer-line contamination by diluting at the
source, rather than at the interface, and c) quick multi-stream
switching with relatively short purge time (this is feasible because
proper dilution minimized the transfer system's solids loading).
-------
The sample transport and conditioning system provides satisfactory
continuous sampling of a single process source; therefore, if
satisfactory automated analyzers were to be dedicated to only one
process source, continuous on-line remote analyses for that source would
be furnished. For multiple-source operation as developed in this
study, the on-line colorimetric analyzers restricted the sampling
frequency to one process source per hour. The time for transport
and conditioning of each sample, however, was only 17 minutes which
would have permitted analyses of approximatley three different
process sources per hour if sufficient colorimetric analyzers had
been added and suitably employed.
In any interfacing of a sampling manifold with an automatic analyzer,
transfer velocities and/or distances within the laboratory space are
just as important as are those used to deliver the samples to the
laboratory. All automatic analyzers should be as close to the sampling
manifold as possible, especially if the sample to be analyzed contains
suspended material.
Where it is not possible to attain optimum analyzer location, analyzer
input velocities should be increased to insure that a representative
sample is actually being supplied to the analyzers within minimum
transport time. This requirement means increase of sample delivery
rates to the analyzers, either by changing the sample pump (or pump
speed), reducing the diameter of the sample lines, or inserting an
additional sample pump to obtain, in each case, a resultant increase
in velocity.
The interface results for most of the automated on-line analyzers tested
did not satisfactorily agree with the interface results from reference
laboratory methods; the one satisfactory on-line analyzer was that
for orthophosphorus. Further development of reliable automatic
on-line analyzers is necessary.
-------
SECTION IV
RECOMMENDATIONS
With the initial phases of this program accomplished successfully, there are
several areas inviting further investigation:
How would the costs of completely automated sampling compare with
corresponding costs for existing manual sampling techniques?
Would data reliability be improved by eliminating all human
influences and human biases from the sample-collecting operation?
If preceded by a reliable automated sampling system, could on-line
analyzers operate continuously for extended periods without failure?
Would the availability of real-time analytical data influence plant
operation in such a way as to improve effluent quality significantly?
Could it likewise be used to decrease plant operating cost
significantly?
Could the concept of "quality assurance" be realized by improving
data reliability via appropriate combinations of automatic sampling
and automatic analyses?
What would be the magnitude of improved reliability resulting from
the substitution of a three-way motor-driven valve for each pair of
two-way valves?
What effect would changes in arrangement of the analyzers (i.e.,
relative to the ST & C system's homogenizer and filtration system)
have on the consistency of the data, and what (if any) limitations are
there in making such rearrangements?
Plant Expansions and New Plant Installation
In modifying and expanding existing plants, and for plants to be built in
the future, installation of a permanent automatic sampling system would seem
to be the more viable approach. It is quite likely that in the planning stages
of these facilities, much thought would be given to the centralization of
-------
sample streams for future monitoring purposes and, ultimately, for automatic
process control. Tapping into lines would be no major problem because a new
plant could provide for readily accessible sampling ports; however, flow
regulation would require some design effort. The flow rates within a large
plant are quite high (thousands of gallons per minute), whereas the Raytheon
Sample Transfer and Conditioning System requires only 5-6 gallons per minute
(19.4 - 23.3 1/min). Careful design of the entire sampling system would be
required to achieve a representative sample. In addition, to make the
homogenizer function properly and to reduce the likelihood of plugging within
the sampling system, the sample would have to be pre-conditioned to reduce
occasional large particles to no more than 1/4-inch (6-mm) diameter. An in-
line grinder pump, rather than the drop-in type used for this project, would
be more suitable for such an application.
The concepts have now been proven; with judicious effort, the problem areas
stated above do not appear to present any insurmountable obstacles.
Raytheon recommends that further wastewater transfer and conditioning studies
be performed to answer these questions.
Portable Installations at Existing Plants
An effective sampling strategy must be adaptable to existing plants, as well
as to those constructed or modified in the future. A single sampling system,
adaptable to both, may be unnecessarily flexible and expensive. The most
practical solution to this problem would be to design a mobile system for
investigating existing plants. With the type of equipment developed during
this program, an investigator could go into a treatment plant with a trailer,
housing the sampling manifold and a battery of automatic analyzers, and
within a very short time, he should be able to assess the plant's efficiency
and initiate steps to rectify problem areas. Being portable, such an
analyzer system would require no major on-site construction. The mobile
facility might be owned by the EPA and leased to municipalities as required.
This is the most practical approach for existing plants since their piping
is not readily accessible. Such an approach would allow rational investigation
of the possible cost benefits of a permanent sampling system; it could also
be coupled with existing automated plant controls without excessive capital
outlays. In either case, such a portable system could help answer many of
the questions posed in the foregoing part of this Section.
-------
SECTION V
PRELIMINARY INVESTIGATION AND ACCEPTANCE TESTING OF
COMPONENTS
The sample transfer and conditioning system was fabricated from a number of
components. These components were a sample transfer pump, a homogenizer,
a filter, and a sampling manifold. A discussion of the selection and design
criteria for each of the critical components is given below.
Sample Transfer Pump
There is great diversity in the physical makeup of wastewater-treatment
streams. A pump must be able to handle clean streams, as well as streams
that contain high amounts of foreign materialplastics, paper, fibers, and
wood chips. Such foreign material provides a formidable deterrent to
continuous pump operation. Raytheon has utilized the Hydr-0-Grind pump
manufactured by the Hydromatic Pump Company in previously developed systems,
and has found it perfectly suited to sample raw influent, primary effluent,
and secondary effluent.
The Hydr-0-Grind is a submersible centrifugal pump, possessing a grinder unit
mounted on the input; see Figure 1. The pump impeller is manufactured from
ductile iron and is cadmium plated. The grinder's stationary and rotary
cutters are made of hardened, ground, stainless steel. The pump and grinder
are mounted on a stainless steel shaft, supported by ball and sleeve bearings
that are oil lubricated. No additional lubrication of the motor or seals is
required.
The Hydr-0-Grind's pump can be operated continuously at a regular flow
of 1 to 30 gpm (3.79 to 113.54 1/min) against a maximum head of 90 feet
(27.43 m). The motor is 1-1/2 horsepower, 3-phase, and 209 to 230 volts.
The motor winding, rotor, and bearings are completely sealed in oil that
lubricates the bearings and transmits heat from the windings to the outer
shell.
The working elements of the grinder pump are a grinder ring and impeller
that macerate gross solids and a secondary cutter/impeller that further
macerates these solids to a reduced particle size of 1/4 inch (6 mm) for
pumping by the centrifugal pump.
The complete front end of the grinder pump (inlet, outer impeller, grinder
ring, inner impeller and centrifugal impeller) can be removed without
affecting the seals, motor or installation.
-------
CENTRIFUGAL
IMPELLER
OUTER
IMPELLER
DISCHARGE
INNER
IMPELLER
GRINDER
RING
INLET
Figure 1. Hydr-O-Grind Pump
10
-------
Homogenizer
Waatewater streams contain all types of particulate matter (plastics, paper,
fibers wood chips, etc.) as well as domestic sewage. This material must be
continuously and reliably reduced to finely divided, uniform, particle sizes
in order to be assimilated by any TOG analyzer.
The Raytheon Homogenizer (patent pending), used on transported Hydr-0-Grind
effluent, was selected on the basis of its proven capability to fill the above
requirement. This homogenizer (Figure 2)consists of the following parts:
An electric motor 1 hp, 115V, single phase, 3600 rpm
Sealed bearings
A micrometer adjustment wheel to regulate homogenizer-effluent
particle size
A housing manufactured of a material that is impervious to corrosion
An abrasive stator and rotor
In operation, the sample is pumped at a prescribed flowrate of 3-6 gph
(11.36 - 22.71 1/hr) through the inlet of the homogenizer and is then processed
between the abrasive stones of the rotor and stator. The design of the rotor-
stator abrasive stones permits the reduction of solids to small particles
without buildup of homogenized solids on the grinding surfaces; in effect, the
abrasive stones are self-cleaning.
Field experience with this type of homogenizer has demonstrated that it is
capable of reducing such difficult materials as plastics to a fine particle size
on a continuous basis without any buildup on the grinding surfaces, a problem
typically associated with solids blenders.
Filter
As mentioned previously, filtering plays an important role in determining the
success or failure of colorimetric analyses. This system's pretreatment
filtration unit incorporates an automatically controlled backwash sequence that
may be initiated by a manual push-button, or hy automatic internal sensing
elements.
The Raytheon Pretreatment Assembly (Model 2550) utilizes a two-stage filtration
process. The first stage is a self-cleaning wash-flow filter which eliminates
the large particles. The second-stage filter is a fixed-media bed which
reduces the filtrate from the first stage to particles of 12 micrometers, or
less.
Component-Testing Manifold
To expedite testing of the filter and other system components, a preliminary
11
-------
INDICATOR GAp
ADJUSTER
0-RING
rTT.--' - '--_J STATOR ,-.v.-.
4- COVER
DO NOT DISTURB
T THESE SCREWS
FIGURE 2. RAYTHEON IN-LINE HOMOGENIZER
-------
sampling manifold was constructed. Figure 3 illustrates the flow diagram.
The wash-flow filter, the fixed-media bed, and the filter's pump were piped
to a test manifold consisting of a series of ball valves capable of directing
various samples to the desired locations. Gauges were installed at appropriate
locations on the manifold to provide pressure and temperature data which
provided additional information to establish effective acceptance criteria
for the individual system components. This preliminary manifold served as a
valuable investigative tool in the acceptance testing of all system components.
Its flexibility enabled Raytheon personnel to test variations in particle-
size reduction as a function of sample flowrate merely by changing orifice
sizes.
It should be noted that the preliminary manifold was essentially a test
vehicle and that the flow diagram shown in Figure 3 in no way reflects the.
final configuratidn of the sampling system described later in the "Final System
Design" Section.
Test Location
Checkout and testing of the design of the sample transfer and conditioning
system were conducted at the Water Pollution Control Facility for Cranston,
R.I. A mobile laboratory was located opposite the grease floatation unit as
indicated in Figure 4. The location was selected so that unused samples
could be exhausted back into the system without affecting the plant operation.
Acceptance Testing of Components
To demonstrate that a representative sample could be taken, transferred, and
conditioned without altering the chemical composition of the original sample,
appropriate analyses were performed on paired samples. Sets of two grab
samples (i.e., sample pairs) were taken: one sample from the stream source
and another sample following sample transport and conditioning. Each of
these sample pairs was then characterized for particle-size distribution and
for TOG value, and the values for each pair of samples were "cross-compared".
First Tests of Particle Size
To determine particle-size reduction by homogenization, four tests were set
up, using the following generally-accepted techniques: a) settleable solids
measurements, b) suspended solids measurement, c) microscopic examination, and
d) sieving.
Settleable solids measurements were made in an attempt to demonstrate that
differences in settling rate had a direct relationship to particle-size
reduction. This test proved inconclusive.
Suspended solids measurements were also made in an attempt to show that solids
content remained unchanged during sample transport and conditioning. A Milli-
pore filter apparatus (Figure 5) was used for this purpose. This test also
gave inconclusive results.
13
-------
FIXED MEDIA
FILTER
SAMPLE COLLECTION FOR
FILTER PERFORMANCE
EVALUATION
20 CC/MIN
PERISTALTIC
X ''PUMP
LEGEND
V = 2-WAY BALL VALVE
G = PRESSURE GAUGE
T = TEMPERATURE GAUGE
TO
DRAIN
TO
DRAIN
SAMPLE COLLECTION
FOR ESTABLISHING
HYQR-0-GRIND
PERFORMANCE
TO
DRAIN
TO
DRAIN
3-WAY BALL VALVE
HOMOGENIZER
ORIFICE
INPUT FROM
HYDR-0-GRIND
SAMPLE COLLECTION
FOR PARTICLE
SIZE REDUCTION
EVALUATION
FIGURE 3.
PKELIMINARY SAMPUHG MANJFOLD FLOW DUGSAM
-------
GUT
IEHOV1L
CHlHIEt
SAMPHH6 POINT ACTIVATED SLUDGE THIS
^SAMPLING POINT
RAW INFLUENT
SAMPLING POINT «««H |
CHiMIEI
«%rrVfP-S£^
^-rfrr=a
^ * i
7- "Tv-hii-lTi
a
L^...I
S c
X \ I /
SLUDGE COtTIOL IOON
ilD HETEI IOOM-
SlUCGE OISESTIOH
Tt»$
FIGURE 4. SITE DESCRIPTION (Sheet 1 of 2)
-------
A Raw Sludge Pumps
B Sludge Density Meter
C Returned Activated Sludge Pumps
D Waste Activated Sludge Pumps
E High Level Air Blowers
F Low Level Air Blowers
G Motor Drives for Blowers
H Gas Engine Drive for Blower
J Grease Well
K Concentrated or Raw Sludge Pumps
L Heated Sludge Recirculation Pumps
M Sludge Heater
N Sludge Well
P Digested Sludge Pump
Q Elutriated Sludge Pump
R Filter Pumps
S Conditioning Tanks and Vacuum Filters
T Sludge Incinerator
U Supernatant Liquor Pumps
V Digestion Gas Booster Pump
W Plant Heater
X Waste Gas Burner
LIQUIDS SHOWN THUS:
SEWAGE SOLIDS SHOWN THUS:
GAS SHOWN THUS:
1 Raw Sewage Influent to Treatment Plant
2 Raw Sewage Influent to Grit Removal Chamber
3 Comminutor Effluent to Grease Removal Chamber
4 Grease Removal Effluent to Primary Settling Tanks
5 Primary Effluent to Aeration Tanks
6 Aeration Effluent to Final Settling Tanks
7 Final Effluent to Chlorine Contact Tank
8 Chlorine Influent
9 Treatment Plant Effluent to Pawtuxet River
10 Raw Sludge or Scum from Primary Settling Tanks
11 Sludge or Grease to Sludge Heater
12 Heated Returned Sludge to Digest Tanks
13 Activated Sludge from Final Settling Tanks
14 Returned Activated Sludge to Aeration Tanks
15 Waste Activated Sludge to Concentration Tank
16 Concentrated Sludge to Sludge Water
17 Waste Activated Sludge from Final Settling Tanks
18 Waste Activated Sludge to Primary Settling Tanks
19 High Level Air to Aeration Tanks
20 Low Level Air to Aeration Tanks
21 Grease to Sludge Heater
22 Digested Sludge to Elutriation Tanks
23 Elutriated Sludge - Tank No. 1 to Tank No. 2
24 Elutriated Sludge to Vacuum Filters
25 Filtered Sludge to Truck or Incinerator for Disposal
26 Ash from Incinerator to Truck
27 Supernatant Liquor from Sludge Digestion Tanks
28 Sludge Digestion Gas to Plant Heaters and Incinerator
29 Elutriate to Primary Settling Tanks
30 Grit Removal
NOTE: Plant Water Piping, By-passes and Tank Drains
are not shown.
Figure
site Description (Sheet 2 of 2)
-------
\
[FILTER! 1110 MICRONI210MICRONH 420 MICRON]
PAPER I ISCREEN ISCREEN ISCREEN
FIGURE 5. MILLIPORE FILTER APPARATUS
17
-------
Microscopic examination with a Bausch and Lomb microscope (//XL10BU-FW with
100-power magnification) was used to study particlesize distribution:
however, it did not yield quantitative data as to particle-size reduction. The
microscopic data were quite subjective since they varied significantly from
individual to individual. In addition, the field of view was quite limited.
and any single determination could be influenced by the occasional presence
of large particles. In essence, the volume of sample examined (one drop)
was too small to yield good results. Therefore, this method was abandoned
for particle-size determination, but was retained to confirm conclusions from
other methods.
Sieve analyses were carried out with a commercially available sieve assembly
(Figure 6). These sieves were too heavy to be accurately weighed using
laboratory balances; therefore, results from initial tests proved unsuccessful.
Further Tests of Particle Size
With all the selected tests yielding inconclusive and unreliable results,
Raytheon had to devise its own method to determine particle size.
The utilized method consisted of filtering a sample through a series of
wire mesh filters. The filtration was conducted such that the sample passed
through the filters in the direction of coarse to fine. The filters were
dried and tared prior to use, and the dry weight was again determined after
filtration. The weight of solids retained on each screen was used to
characterize the particle-size distribution in the sample. Three sizes of
screen were used: 420 urn, 210 ym and 110 vim. The filtration was carried out
in the modified Millipore filtration apparatus (Figure 5). This method
proved very satisfactory and reliable.
Figure 7 demonstrates how typical data were utilized; it also demonstrates
particle-size reduction on a percentage basis. As can be seen from inspection,
the particle-size distribution changed drastically when the homogenizer was
used. A comparison of one point on the curve (e.g., 420 micrometers) indicates
that unprocessed raw sewage possessed approximately 87% of its solids with
particle sizes equal to, or less than, 420 micrometers. With a single
homogenization this number increased to 99.2%. Along with this determination,
suspended solids values were also obtained. In each case, the values for total
suspended solids before and after homogenization agreed within 2%, proving
that no appreciable loss in solids occurred as a result of the conditioning
process. This type of evaluation was performed on the six streams in
question: a) secondary effluent, b) primary effluent, c) raw influent,
d) mixed liquor, e) return activated sludge, and f) primary sludge.
Discussion of Particle-Size Testing
Good particle-size reduction data were obtained for all six streams% However,
for effective sampling of mixed liquor, return activated sludge, and* primary
sludge, the samples had to be diluted prior to introduction into the
homogenizer because these streams were very high in particulate matter and
also very viscous, making it inadvisable to pump these streams with a
Hydr-0-Grind. Dilution pumps were not available for this phase of the project;
18
-------
FIGURE 6. SIEVE ASSEMBLY
-------
HOMOGENIZER
EFFLUENT
INPUT TO
HYDR-0-GRIND '
PUMP
0 RAW SEWAGE, HYDR-0-GRIND INFLUENT
[0 RAW SEWAGE, HOMOGENIZER EFFLUENT
100
200 300 400
PARTICLE SIZE, MICROMETERS
500
FIGURE 7. RAW SEWAGE, PARTICLE-SIZE REDUCTION WITH HOMOGENIZER
20
-------
therefore, dilution was accomplished by hand to prepare the input sample to
the homogenizer for particle-size reduction tests for these streams. The
results from these tests proved the acceptability of sample dilution prior
to homogenization.
By using sample dilution at the source, high solids loading within transfer
lines was avoided, and the probability of successful sample transfer was
increased. The method used will be covered in the "Final System Design"
Section of this report.
Demonstration of good particle-size reduction only proves that effective
conditioning of a sample for TOG analysis can be accomplished. Questions arise,
however, whether or not solids have been lost in the conditioning process, or
will be lost during transfer from homogenizer to analyzer. To demonstrate
that no solids were being lost, TOC analyses were conducted, utilizing the
Beckman 915 Total Organic Carbon Analyzer. The TOC of the sample stream was
measured in the vicinity of the Hydr-0-Grind pump. This TOC value was compared
with the TOC value of the Hydr-0-Grind effluent and also with the TOC value of
the Homogenizer effluent; such comparisons demonstrated no loss in TOC and
proved that transfer of the sample can be performed effectively.
Table 1 demonstrates quite vividly that, in every instance, integrity of
chemical composition is preserved from the sample's point of origin through
the sample-conditioning step. This data, accumulated from monitoring raw
sewage, is typical of data obtained by sampling the other five process streams.
TABLE 1
PRESERVATION OF THE INTEGRITY OF CHEMICAL
COMPOSITION DURING THE COURSE OF
TRANSPORT AND CONDITIONING OF RAW SEWAGE
Date Source TOC, mg/1 Ortho-PO^, mg/1 N0_, mg/1
1-14-74 Grinder influent 310 6.52 8.3
1-14-74 Grinder effluent 305 7.67 8.3
1-14-74 Homogenizer effluent 280 6.30 8.3
1-15-74 Grinder influent 510 9.72 8.2
1-15-74 Grinder effluent 495 10.19 8.14
1-15-74 Homogenizer effluent 500 9.72 7.32
1-17-74 Grinder influent 335 6.55 5.42
1-17-74 Grinder effluent 335 6.44 5.72
1-17-74 Homogenizer effluent 340 6.41 5.52
21
-------
Pipe-Size Consideration
Sample flowrate and pipeline size are related because, taken together, they
determine sample velocity; hence, they must be considered together. The
sample-line size must be large enough to give assurance that there will be
no plugging or clogging anywhere within the sample train. However, the line
size must also be small enough to furnish high transport velocities so that
complete transfer of suspended solids is assured. Obviously, upward velocities
of particulate matter in any vertical section of the sampling train must well
exceed the settling velocity of the maximum size particle to be sampled.
Settling of solids is an important consideration. Sizing of transfer pipes
to obtain sample velocities that will preclude settling of solids was one of
the design goals addressed. An EPA report by Shelley and Kirkpatrick states
that the minimum line size for transfer of samples from a stream or combined
sewer should be 3/8 to 1/2-inch (9.53 to 12.7 mm) inside diameter. Also stated
was that minimum line velocities should be in the 2 to 3-ft/sec (.61 to 91-
m/sec) range.
With these basic guidelines defined, the transfer lines were selected. In all
cases a minimum velocity of 2 ft/sec (.61 m/sec) was strived for, but this
value could not always be obtained. However, no transfer-line contamination
due to settling was noticed during the course of this study.
F_il_ter_Tests
To demonstrate an effective filtration apparatus, Raytheon set up a two-stage
filtration unit. The first stage consisted of a self-cleaning wash-flow filter
which served to eliminate the larger particles; the second stage was a down-
flow filter which consisted of a fixed-media bed which further reduced the
sample's particulate matter to a diameter of 12 micrometers, or less. Flow
was supplied by the grinder pump to the filter block which housed a nominally
rated, 10-micrometer, nylon filter disk. Sample across the filter disc is
continuous at a rate of 3 gal./min (11.35 1/iain). A small portion of this
flow is drawn through the nylon,wash-flow, disc filter using a peristaltic
pump adjusted to 20 ml/min. This pump in turn feeds the fixed-media gravity
filter. The final filtrate is collected in an overflow cup from which the
sample is drawn by the various analyzers.
Table 2 demonstrates results from the two-stage filter test. As can be seen,
suspended solids loading varied a great deal; however, the filtering apparatus
was able to remove (worst case) 94.3% of all particulate material in the
stream with an average removal of 97.5%.
22
-------
TABLE 2
RESULTS OF FILTER TEST
Suspended Solids Particulate Material
Loading, mg/1 Removed, %
440 97,1
520 98.1
570 97.7
580 98.8
610 94.3
620 98.7
650 96.9
680 98.6
1100 99.4
1160 98.5
Some Observations
In addition to providing the necessary flows, the preliminary manifold proved
to be an invaluable design vehicle for establishing design "ground rules" such
as the following:
1. Ball valves should be used only in a full-on or full-off position.
Intermediate positions cause dead spots in the flow passages, and
this eventually causes line plugging.
2. Solenoid valves are prone to plugging and therefore are unreliable.
3. Settling in transfer lines is virtually eliminated if no sharp
restrictions or stagnation points exist in the lines and if transfer
velocities are maintained at no less than 2 ft/sec (.61 m/sec). For
instances where this is not feasible, it should be experimentally
determined if lower transfer velocities are acceptable.
Preliminary Conclusions
From the results of the preliminary tests, certain conclusions could be drawn:
1. Use of Hydr-0-Grind pumps eliminates the need for homogenization
prior to transferring samples over long distances. This confers
significant cost advantages, particularly for those applications
requiring continuous sampling of numerous process streams.
23
-------
2. The Hydr-0-Grlnd pump reduces particle sizes sufficiently to enable
quantitative sample transfer and efficient operation of only a
single, centrally located, horaogenizer unit.
3. The Raytheon homogenizer effectively reduces particle sizes for
on-line TOC analyses without unacceptably altering the sample's
chemical composition.
4. The filtration unit removes an average of 97.5% of all particulate
matter in a flowing stream, and produces a filtrate possessing a
maximum particle size of 12 micrometers.
5. The measured temperature rise through the homogenizer is less
than one degree Celsius.
6. The measured pressure drop through the homogenizer is 2 psi
(13.79 kN/sq m).
24
-------
SECTION VI
FINAL SYSTEM DESIGN
Establishing a Sampling Procedure
Following the design specifications established in Phase I, Raytheon set out
to design an effective sampling system that would be continuous, reliable,
and easily maintained without altering the chemical composition of the original
sample. The first step was to establish a sampling matrix capable of guiding
the collection of data required to properly test the system. Table 3 illus-
trates this matrix.
TABLE 3
SAMPLING MATRIX
*Sampling Points
Chemical Test
Total organic carbon
Soluble organic carbon
Orthophosphate
Hydrolyzable
phosphate
Ammonia nitrogen
Nitrate
Nitrite
Sec.
Effl.
CD
X
X
X
X
X
X
X
Prim.
Effl.
(2)
X
X
X
X
X
X
X
Ret.
Raw Mixed Act. Prim.
Infl. Liq. Sludge Sludge
(3) (4) (5) (6)
X XX X
X
X
X
X
X
x
*Streams are listed in order of expected contaminate concentration
and were s«im>led *-n thi-8 order, as mentioned earlier in the text.
From Table 3 it can be seen that streams 1, 2, and 3 (secondary effluent,
primary effluent, and raw influent, respectively) demanded the majority of
the sampling requirements. Since the sampling-system design was greatly
influenced by these requirements, answers to the following questions were
required before system design could proceed:
25
-------
1. How can one analyzer monitor TOG and soluble organic carbon (SOC)
for the same stream?
2, What are the optimum modes of operation for the colorimetric
analyzers and Pretreatment Unit after sampling of stream (3) has
been completed; i.e., while streams (.4), (5) and (6) are being
sampled?
3. Can the system handle high solids loadings without adversely
affecting either data reliability or the system's self-cleaning
capabilities?
4. Is any flushing of the sampling manifold and/or the transfer lines
needed? If so, then what are the flushing requirements?
Monitoring Both TOC and SOC
To solve the problem of monitoring both TOC and SOC for the same stream,
Raytheon modified its 'Model 2600" on-line TOC analyzer by adding a second
rough sample pump within the analyzer. For the first cycle (i.e., secondary
effluent sample), the "Model 2600" monitored TOC by utilizing the rough
sample pump connected to the homogenizer output. Upon command from the
control panel, that pump was shut off. The second pump (connected to the
Pretreatment Unit output) was then turned on, and a filtered portion of
secondary effluent sample was supplied to the "Model 2600" for SOC analysis,
and to the various colorimeters for determinations of phosphate, ammonia,
and nitrate/nitrite. Following completion of this first cycle, the system
was switched to the next sampling point (i.e., primary effluent) and the
first cycle was repeated. This sequence also was followed for the raw
influent sample.
On completion of the third sample cycle (raw influent), no further colorimetric
or SOC analyses were required; only TOC monitoring was required for streams
4, 5, and 6. A decision had to be made as to what modes the Pretreatment
Unit and colorimeters would be left in during the interim because, if the
units were shut down for the three hours required for analyzing streams
4, 5, and 6, subsequent startups would require operator attention. Therefore,
it was decided to continue operating the units, but to use flush water as
the "sample" stream. This approach yielded a twofold benefit:
1. The analyzers would not be operating without sample input (That
type of operation is not recommended).
2. The analyzers would automatically record a zero point during
each complete run provided the flush water were not contaminated.
Following the completion of cycle 6, an additional cycle was employed,
whereby the sampling manifold, the homogenizer, and its associated plumbing
were all flushed with tap water. The flush lasted for only one cycle interval;
after which, the system was again ready to start sampling and conditioning
Sample No. 1 (secondary effluent). It should be noted that each cycle could
be aborted at any time by switching to the manual mode.
26
-------
System Description
The resultant flow diagram for the sample transfer and conditioning
system is shown in Figure 8, The system, which was designed for simple
construction and operation, has two modes of operation: automatic and
manual. Sample sequencing in the automatic mode is controlled by a timer
which can be adjusted for cycle times of 1 to 60 minutes. A second timer was
incorporated to start a data acquisition system for automatic data logging.
The sample-switching system consists of pairs of motorized ball valves which
operate in tandem to select a sample for analysis upon a command from the
control panel. At any given time, each sample is connected (via the valve
pairs) to one of two manifolds: sample manifold or drain manifold.
Both manifolds are fabricated utilizing standard PVC fittings chemically
bonded together so that all joints provide the minimum amount of obstruction
to the flow. The motorized ball valves are also PVC and have union-type
pipe connections. This type of valve connection allows for easy disassembly
of either the manifolds or the valves should a problem arise. The assembled
system, plus the more important sub-systems, are shown in Figures 9 through
13.
The sampling assembly is supplied with samples from six remote points, utilizing
Hydr-0-Grind pumps for streams 1, 2, and 3 (Figure 14) and duplex dilution
pumps for streams 4, 5, and 6 (Figure 15). Each sample stream is continuously
fed to the sampling assembly; however, only one stream can be monitored at a
time. Therefore, the other five sample streams are directed to the drain
manifold, and back to the head of the plant. This bypass systemgkeeps all
the sample streams constantly flowing, and eliminates deadending (Deadending
a stream results in a stoppage of flow and possible deposition of particulate
matter. If a deadended stream were selected for monitoring, good quantitative
data could not be obtained because of the excessive suspended solids loading
that would occur when the flow started up again. Such a sudden scouring
of the lines could easily produce a temporary, but heavy, overload of
suspended particles that might also impair the long-term performance
of the homogenizer, Pretreatment -Unit or automatic analyzers). A second
benefit conferred by a bypass system is a reduction in the time lag required
for system purging by the next sample to be analyzed. The bypass system
thus has greater useable analytical time because of greatly diminished purge
times.
Use of Dilution
Another important aspect of the sampling system is the manner in which mixed
liquor, return activated sludge, and primary sludge are conditioned and
transported. Raytheon utilized the dilution concept in sampling these three
streams for the following reasons:
1. TOC values in streams 4, 5, 6 were much higher than in streams
1, 2, and 3; therefore, the need for a second TOC analyzer (or an
analyzer capable of automatic range selection) was eliminated.
27
-------
FLUSH >
WATER S
V7A&
1 V8B
1
1
i
1
1
1
1
3
h
0 V8A
PRETREATMENT
ro
oo
SAMPLE
MANIFOLD
soc
V9A
TOC
V9B
N03
+.
__L_
DATA
ACQUISITION
SYSTEM
-V i
I
TRAILER
ELECTRICALLY OPERATED
BALL VALVES
0 - DRAIN
HYDR-0 GRIND
PUMPS
SAMPLE DILUTION
PUMPS
FIGURE 8. FLOW DIAGRAM OF THE SYSTEM
-------
HOMOGENIZER
FIGURE 9. SAMPLING SYSTEM, FRONT VIEW
29
-------
»«*<» *
INPUT LINES TO
I SAMPLING SYSTEM
FIGURE 10. SAMPLING SYSTEM, REAR VIEW
30
-------
WASTEWATER SAMPLE TRANSFER AND CONDITIONING SYSTEM
DATA ACQUISITION SYSTEM
FIGURE 11. CONTROL PANEL, FRONT VIEW
31
-------
NJ
FIGURE 12. CONTROL PANEL, REAR VIEW
-------
INCOMING
FLUSH WATER
ACTUATORS FOR
ELECTRICALLY
OPERATED BALL
VALVES, IN TANDEM
SUPPLY LINE TO
PRETREATMENT ASSY
FLUSH WATER
SUPPLY LINES
ELECTRICALLY OPERATED
BALL VALVES
TO PRETREATMENT
:AND ANALYZERS
TO SAMPLE
MANIFOLD
SUPPLY LINE
ITOHOMOGENIZER
SAMPLE MANIFOLD
EXHAUST
MANIFOLD
INCOMING SAMPLES
FROM HYDR-0-GRINDS
FIGURE 13. MANIFOLD ASSEMBLY,
FRONT VIEW
INCOMING SAMPLES FROM
DILUTION PUMPS
-------
u>
GRINDER
PUMP
FIGURE 14. TYPICAL HYDR-0-GRIND PUMP INSTALLATION
-------
FLOW RATE \ I
CONTROLS
STATIC
MIXER
u>
DILUENT
(TAP WATER)
MANIFOLD
" CHECK VALVE
ASSEMBLY
1" CHECK VALVE
ASSEMBLY
FIGURE 15. TYPICAL DUPLEX DILUTION PUMP INSTALLATION
-------
2. Because suspended solids concentrations were so very high for
streams 4, 5, and 6, if these streams had been transported in
undiluted condition, the risk of sample-line contamination would
have been significantly increased. This would have affected the
success of the program by reducing the likelihood of accurate
measurement.
Dilution systems have been applied previously1 to analyzer inputs (i.e. after
the samples had already reached the analyzer). While this approach may have
been acceptable for sample analysis, sample-line contamination was always a
problem.
The decision was therefore made.to dilute at the source, rather than at the
analyzer interface, to minimize solids loading in the transport lines. The
streams were diluted in such a way that they approximated the physical
characteristics of primary effluent. A BIF, Series 1722, Duplex "Propsuperb"
metering pump was selected because of expected quick delivery and estimated
suitability for this application. The pump is a positive-displacement
hydraulically-actuated diaphragm pump, with a manually adjustable stroke to
change the flowrate. The pump has two sides which act independently but
which are driven by a common drive. This arrangement assures a constant
dilution ratio even though the drive speed varies. A typical installation
of a dilution pump has already been illustrated by Figure 15; that illustration
shows the monitoring of mixed liquor at an aeration basin. The sample dilution
ratio was 1:5. The sample stream was drawn through the 3/4-in. (19-mm) check-
valve assembly on the left side of the pump, while the diluent (tap water)
was drawn through the 1-in. (25-mm) check-valve assembly on the right. Both
streams were fed into a common manifold.
The resultant output was a pulsating non-homogeneous flow of both sample
and diluent. An in-line static mixer (Kenics P/N 37-08-136) was installed at
the pump discharge to counteract this phenomenon. This proprietary in-line
mixing assembly employs a series of fixed helical elements enclosed within a
tubular housing. The Internal geometric design of the unit produces a unique
pattern of simultaneous flow division and radial mixing.
Subsequent to static mixing and transfer to the trailer, the sample was
homogenized just prior to TOC analysis. -r
Adapting the Dilution Pump
To install a pump of this type, certain preliminary tests must be performed
and certain conditions must be maintained. Initially, for each specific
application, a performance curve (output vs. control setting) must be
experimentally obtained by each pump after it has been installed. The
dilution ratio can then be set by adjusting the pump strokes according to
the empirically developed performance curves.
This particular type of positive displacement pump requires a non-varying
back-pressure to operate reproducibly. When the back-pressure is low or
variable, erratic operation occurs. A back-pressure valve is normally
36
-------
installed to provide unvarying back pressure. However, this would also place
an undesirable obstruction in the discharge line. To simulate back-pressure
but eliminate the unwanted obstruction, additional (i,e,, excessive) lengths
of hose were used on the discharge side of the pump to create backpressure
by increasing the head-loss across the sample line. This simple modification
solved the back-pressure problem.
In addition to being back-pressure sensitive, the pump was found to be input-
pressure sensitive. The pump relies on proper check-valve sealing to get
proper pumping action. Therefore the pump will not operate acceptably if
there is positive input pressure. As a result, this type of pump has to draw
its feed from a sample reservoir which is essentially at ambient atmospheric
pressure. If the sample is being piped under pressure, as was the case with
the return activated sludge and primary sludge, the pump cannot be connected
directly to the pipe. The sample must first drop into an overflow reservoir
(i.e., cup) that is constantly being fed with fresh sample, then the pump
will draw the sample as required from the reservoir.
Initial results demonstrated a fair amount of intermittent flow (i.e., flow
discontinuities and stoppages) caused by the presence of fibrous material in
the stream. The fibrous material affected the sealing capabilities of the
check valve and prevented proper pumping. Cleaning of the check valves brought
the pump back on-line.
The problems encountered using these pumps are characteristic of all pumps
using check valves,
Overcoming Intermittent Flow
Intermittent flow can be minimized, if not eliminated, by using low dilution
ratios; i.e., relatively high flows through the "sample side" of the pump.
If dilution ratios are selected so that sample flowrates are high, the
subsequent flow velocities through the pump's check valves are also high; such
relatively high velocities are advantageous because they increase the tendency
of fibrous material to pass cleanly through check valves.
A slightly different approach was tried to monitor primary sludge. From tests
conducted in the Preliminary Phase, the TOC values were high and dilutions of
100:1 were anticipated. With dilution ratios that high, it was a certainty
that sample flowrate would be very low. As mentioned previously, this is not
desirable. Hence, the same type of pump was used, but it employed a variable-
speed B.C. motor rather than a fixed-speed A.C. motor. This arrangement makes
possible total flow changes without altering the dilution settings, and this
in turn affords the operator greater latitude for establishing optimum pump-
operating conditions.
The particular piping configuration at the Cranston site generated a unique,
primary sludge, sampling problem: the high solids level of primary sludge
(5-6% by weight) prevented continuous flow of sample to a reservoir. Many
different piping configurations were tried, but continuous flow could not be
attained. Without major rework of the plant's piping and an accompanying
disruption of plant operation to gain access to unthickened primary sludge,
37
-------
this problem could not be rectified. The basic sample-taking concept is valid,
however. The primary sludge was only one of the six sample streams involved
in this project, and it had only one parameter of interest (TOG); therefore,
monitoring of this stream was a very small part of the total program, and
deletion of this sample stream detracted only modestly from the main purpose
of the project. There is no reason to suspect that primary sludge, properly
supplied to a dilution assembly of the type described above, would cause any
significant problems during dilution, transfer, conditioning, and analysis.
The EPA concurred with our decision to delete primary sludge sampling,
particularly since control of this plant was out of our hands.
Timing of Samples
Timing is an important, but easily overlooked, consideration in the design
of a stream-switching system. Of course, the frequency at which each point
is analyzed must be acceptable to process-control requirements. In addition,
the dwell time on each sample must be of sufficient duration to purge the
sample manifold and passageways into the analyzers, and to allow sufficient
time for analyzer response. There are three times which must be addressed:
a) transfer time, b) conditioning time, and c) analyzer response time.
Transfer times were obtained by measuring the flowrate of each pump and by
using the appropriate formula for the velocity in each line. With the
transfer velocities and distances known, transfer times were easily calculated.
These data are shown in Table 4.
TABLE 4
SAMPLE-TRANSFER DATA
Sample
Stream
'Secondary
effluent
Primary
effluent
Raw
influent
Mixed
liquor
Return
activated
sludge
Primary
sludge
Flow rate,
gal/min
(1/min)
4.87(18.4)
5.46(20.6)
5.82(22.0)
.46(1.74)
.44(1.67)
^ _ r| -r-r
Flow Velocity,
ft/min
(m/min)
119.8(36.2)
114.7(44.1)
153.0(46.6)
79.5(24.2)
75.6(23.0)
Transfer
Distance,
ft (m)
475(144.8)
55(16.8)
75(22.9)
210(64.0)
250(76.2)
^^
Transfer
Time,
Min
3.95
.38
.49
2.64
3,30
38
-------
Because it was unnecessary to separate conditioning time from analyzer
response time, these two times were measured in one combined test. To
measure the response time of the analyzers in conjunction with sample-
conditioning time, a flush cycle was initiated and a stable zero was reached
on all monitoring equipment, thus providing a zero datum. Primary effluent
was selected manually, and the time was recorded. Within ten minutes the
TOG analyzer was reading 100% of final value, but the colorimetric analyzers
required forty minutes to reach 100% of the final value. Based on this
data, the sample cycle time was set at one hour. Thus, during a one-hour
cycle, the sample for TOC is taken from the continuous flow sample stream
30 minutes after preceding portions taken for colorimetric analyses.
Comparisons with Standard Analytical Methods
To assess properly the success or failure of the sampling system, manual
grab samples were taken and analyzed by standard methods; these values were
also compared with the corresponding values obtained from the automatic
analyzers in order to determine analyzer performance. To assure a valid
evaluation of the sampling system, two grap samples were taken: one at the
source of the process stream (before any automatic sampling, transferring,
or conditioning) to establish a reference point, and a second at either
the homogenizer's exit port (interface value for TOC analyzer) or at the
filter assembly's exit port (interface value for colorimetric analyzers).
Because of the large differences in analyzer response times (ten minutes
for the TOC analyzer, and forty minutes for the colorimetric analyzers) it
was necessary to take two sets of grab samples at different times.
As previously described, the control, system has two timers: one for controlling
the sampling cycle and one for starting up auxiliary equipment (i.e., the
data-acquisition system). To obtain representative grap samples for
comparison with automatic analyzers, the samples had to be obtained in
advance of the actual readout times; hence, readout time was established as
the last five minutes of each sampling cycle. A grab sampling procedure was
established and is shown diagramatically in Figure 16.
39
-------
Sequence of events during a sampling cycle:
TIME INTO
'CYCLE
DATA-READOUT
PERIOD
47
-f-
SEQUENCE
OF EVENTS
-40 WIN**-
45
10MIN*
FIGURE 16. SAMPLING SEQUENCE
1. Sample stream selected,
2. Grab samples taken at sample source and from Pretreatment Unit effluent;
these were referee samples for the colorimetric analyses and for SOC.
3. Grab samples taken at sample source and homogenizer output; these were
referee samples for the TOC analyses.
4. Data-acquisition system was started up, and data were recorded.
5. As of this point in time, analyzer values and the corresponding grab
samples should agree.
6. Cycle completed. New stream selected. Data-acquisition system turned
off.
NOTE: Sampling sequence is the same for process streams 1, 2 and 3.
However, for process streams 4, 5 and 6, only one set of grab samples
was taken at event (3) since TOC was the only parameter of interest
for these three streams.
For a more detailed description of the system's operationg and main-
tenance procedures, refer to Appendix B.
* Time lag between sample input to TOC analyzer and the analyzer's
corresponding readout of TOC.
** Time lag between sample input to colorimetric analyzers and their
corresponding readouts of NH_, o-PO,, etc.
-------
SECTION VII
TESTS RESULTS
Reference Tests
With all the equipment installed, operating, and a sampling sequence
established and proven, the next step was to establish the referee tests to
be performed on the grab samples.
To measure TOC and SOC, a Beckman "Model 915" Total Organic Carbon Analyzer
was used. The Beckman "Model 915" analyzes discrete, 50-microliter samples
that must be injected into the instrument by means of a microsyringe.
Samples containing any significant amounts of suspended matter must be
blended, acidified, and sparged prior to injection into the instrument. For
this project, referee TOC and SOC samples were manually acidified, and a
Waring blender was used to sparge out the C0« and simultaneously blend the
sample's suspended solids. For colorimetric analyses, methods specified by
"Standard Methods for the Examination of Water and Wastewater", 13th Edition,
were used:
1. Phosphate - Method 223, "Ascorbic Acid Method"
2. Ammonia - Method 132B, "Direct Nesslerization Method"
3. Hydrolyzable Phosphate - Method 233F with a preliminary hydrolyzation
step whereby the sample was acidified to a pH of 1 and the solution
was boiled for one hour.
A Bausch and Lomb "Spectronic 70" spectrophotometer was used to perform the
manual, colorimetric, reference analyses for orthophosphate, ammonia nitrogen
and hydrolyzable phosphate.
Sample Transfer and Conditioning System Test Data
Table 5 shows the test results for the sample transfer and conditioning
system when it was evaluated by Raytheon at the Cranston Water Pollution
Control Facility. The amount of data collected is sufficient for a preliminary
evaluation of the sampling system. Note, however, that most of the erratic
data in Table 5 were produced by the various on-line analyzers being employed!
Performance of Automatic Analyzers
TOC, SOC, orthophosphate and ammonia nitrogen were the only parameters
monitored "on-line". The on-line hydrolyzable phosphate analyzer did not
41
-------
TABLE 5.
S3
SAMPLE TRANSFER AND CONDITIONING SYSTEM TEST DA1]
SECONDARY EFFLUENT (Sheet 1 of 4)
'A:
Date
(1974)
11-5
11-8
11-11
11-12
11-13
11-14
11-15
11-18
11-20
11-22
11-25
11-26
11-27
TOC, mg/1
_, Inter- Ana-
Source
face lyzer
13.2 13.4
34.0 33.0 21.5
30.0 27.0 31.0
21.0 24.0 19.0
22.0 20.0 24.0
26. 0 25. 0 52. 0
23.0 21.0 45.0
- - -
30. 0 28. 0 29. 0
33.0 30.0 25.0
30.3 30.3 11.3
24. 0 26. 0 30. 0
22.0 26.0 20.5
25.0 22.5 21.0
SOC*. mg/1
0 Inter- Ana-
Source
face lyzer
14.8 11.8
31.0 33.0 21.0
20.0 52.0 30.0
18.0 17.0 13.0
16.0 15.0 21.0
19.0 19.0 43.0
20.0 18.0 43.0
- - -
17.0 17.0 20.0
25.0 34.0 22.9
22.6 24.3 15.0
14.0 17.0 26.0
24.0 18.0 21.0
17.0 17.0 20.5
o-PO , mg PO /I
Source
17.1
19.2
18.0
16.6
22.5
20.6
17.9
15.8
17.2
16.4
17.3
17.2
22.6
22.1
18.4
17.6
13.4
12.8
22.0
19.5
16.7
17.0
16.0
16.3
Inter-
face
16.8
18.7
16.0
16.4
22.1
19.8
18.0
16.8
16.4
16.1
16.6
17.0
17.4
21.6
17.4
17.5
10.8
11.2
20.3
12.2
16.3
16.5
15.0
15.9
Ana-
lyzer
16.5
18.4
-
17.1
23.5
21.2
17.1
16.0
14.4
17.0
17.0
16.9
20.9
22.8
17.7
18.0
10.6
11.8
19.4
20.2
17.8
17.5
17.8
18.2
NH , mg NH -N/l
Source
5.1
5.0
-
-
-
-
15.2
13.4
-
-
15.9
-
-
-
11.8
11.0
16.1
17.7
9.5
8.96
Inter- Ana-
face lyzer
3.3
3.9
-
-
-
-
15.4
13.6
23.7
25.2
15.6 16.4
16.4
-
-
12.7
10.8
15.6 18.3
16.6 18.6
11.4 17.8
6.8 9.0
12.1 12.1 18.3
11.3 11.2 16.5
15.5 14.8
14. 8 14. 1
Hyd. P04
Source
20.5
20.8
22.8
18.3
_
22.5
18.7
16.8
18.2
17.2
18.1
-
24.2
23.6
20.1
18.8
_
-
26.4
22.4
_
-
16.9
17.1
mgP04/l
Inter -
7.96
14.4
19.7
18.5
_
21.4
18.7
17.8
17.4
16.9
17.4
_
18.6
23.1
18.9
18.7
_
-
24.6
14.0
-
15.9
16.7
*SOC particle size was less than 12 micrometers
-------
U!
TABLE 5. SAMPLE TRANSFER AND CONDITIONING SYSTEM TEST DATA
PRIMARY EFFLUENT (Sheet 2 of 4)
Date
(1974)
11-5
11-8
11-11
11-12
11-13
11-14
11-15
11-18
11-20
11-22
11-25
11-26
11-27
TOC, mg/1
Source
86.0
145.0
307.0
125.0
92.0
138.0
94.0
-
113.0
121.0
119.0
133.0
104.0
102.0
Inter-
face
93.0
131.0
109.0
113.0
84.0
127.0
94.0
-
107.0
126.0
114.0
110.0
109. 0
96.0
Ana-
lyzer
-
131.0
115.0
81.0
133.0
197.0
115.0
-
87.5
105.0
19.0
105.0
80.0
80.0
Source
106.0
106.0
103.0
75.0
54.0
59.0
68.0
-
67.0
62.0
63.0
49.0
48.0
51.0
SOC*,
Inter-
face
64.0
101.0
104.0
55.0
53.0
60.0
66.0
-
95.0
83.0
89.0
64.0
73.0
58.0
mg/1
Ana-
lyzer
-
105.0
93.0
81.0
128.0
213.0
173.0
-
78.0
70,0
107.5
85.0
65.0
61.0
o-PO , mg
Source
14.9
15.0
15.6
17.8
13.8
16.0
16.4
13.2
13.5
-
-
15.4
14.2
15.3
16.2
19.3
20.1
18.1
14.0
15.4
14.3
14.6
Inter-
face
15.0
14.7
14.0
17.2
13.4
16.1
16.1
12.5
13.0
-
-
15.0
13.7
13.4
15.2
18.3
17.4
17.0
12.4
13.6
11.2
13.5
PO /I
4
Ana-
lyzer
15.5
15.1
14.6
17.4
16.2
15.8
15.9
14.1
15.2
-
-
16.1
14.3
15.3
15.6
16.3
17.3
17.8
14.2
15.4
15.8
16.8
NH3>
Source
33.0
32.2
-
-
-
29.2
24.6
_
-
30.0
27.2
-
25.2
25.7
25.8
23.9
28.4
27.5
25.4
29.9
26.8
26.8
mg NH -N/l
Inter-
face
33.6
32.0
-
_
-
28.4
25.6
_
-
29.2
2.74
-
25.2
24.3
23.8
24.0
26.7
27.7
25.6
27.0
26.4
26.5
Ana-
lyzer
_
58.5
-
_
-
_
33.8
46.9
42.4
36.8
-
-
-
33.3
29.3
27.3
27.7
54.3
58.8
_
-
Hyd. P04, mg PO A
Source
25.1
27.5
26.8
_
23.4
25.7
27.9
22.2
22.9
21.7
26.6
24.6
22.7
26.8
_
32.6
30.8
-
_
23.3
24.8
Inter-
20.2
21.7
24.1
_
23.2
20.8
23.6
21.0
22.9
16.8
21.4
24.3
22.1
23.6
_
28.2
28.9
-
18.3
22.9
*SOC particle size less than 12 micrometers
-------
TABLE 5. SAMPLE TRANSFER AND CONDITIONING SYSTEM TEST DATA
RAW INFLUENT (Sheet 3 of 4)
Date
(1974
11-5
11-8
11-11
11-12
11-13
11-14
11-15
11-18
11-20
11-22
11-25
11-26
11-27
TOC, mg/1
Source
130.0
172.0
235.0
202.0
206.0
155.0
178.0
-
229.0
220.0
153.0
136.0
177.0
139.0
Inter-
face
116.0
157.0
150.0
142.0
165.0
158.0
166.0
-
168.0
163.0
116.0
116.0
132.0
133.0
Ana-
lyzer
118.0
140.0
137-. 0
203.0
240.0
245.0
-
113.0
119.0
130.0
95.0
85.0
-
SOC*. mg/1
Inter- Ana-
Source ,
face lyzer
138.0 53.0
97.0 96.0 86.0
77.0 82.0 124.0
- - -
70.0 62.0 213.0
97.0 80.0 223.0
97.0 173.0 205.0
- - -
76.0 98.0 85.0
95.0 97.0 105.0
-
- - -
73. 0 84. 0 81. 0
223.0 75.0 81.0
o-P
Source
13.0
14.5
14.4
-
16.1
12.3
12.8
13.2
13.5
18.1
16.3
14.2
14.9
14.0
15.6
14.5
14.6
13.3
11.6
16.5
13.0
D4> mgPO4/l
Inter-
face
12.9
14.4
13.4
-
16.8
12.0
13.1
11.8
13.5
19.2
16.3
15.6
12.1
12.9
12.5
13.9
14.4
12.0
7.8
15.8
13.4
Ana-
lyzer
12.6
14.8
14.0
-
14.5
11.9
12.9
15.5
17.0
15.9
20.2
16.7
16.7
12.5
16.8
15.0
14.6
12.7
18.7
NH3,
Source
22.4
19.4
-
-
-
20.7
20.4
_
18.0
21.2
-
25.2
23.0
22.2
19.5
18.2
24.9
20,4
24.0
19.2
mgNH3-N/l
Inter-
face
22.8
20.4
-
-
-
21.6
20.3
_
22.2
18.1
-
18.8
21.9
21.5
21.2
18.0
24.5
21.7
25.9
21.1
Ana-
lyzer
48.3
29.8
-
-
-
25.3
21.6
30.8
26.2
16.0
-
-
22.5
20.4
18.2
47.8
43.5
-
Hyd. P04,
Source
25.4
20.2
28.8
-
28.9
27.5
23.6
25.1
24.3
26.6
23.0
31.2
29.9
27.3
-
35.1
32.8
-
23.1
24.7
mg P04/l
Inter-
face
21.7
18.4
26.9
-
29.8
26.9
20.8
22.4
24.3
28.2
23.0
34.3
24.3
25.2
-
38.2
32.4
_
34.7
25.5 !
*SOC particle size less than 12 micrometers
-------
TABLE 5
SAMPLE TRANSFER AND CONDITIONING SYSTEM TEST DATA
(Sheet 4 of 4)
MIXED LIQUOR
Date
(1974)
11-5
11-8
11-11
11-12
11-13
11-14
11-15
11-18
11-20
11-22
11-25
11-26
11-27
11-29
Source
75.0
142.0
132.0
102.0
120.0
118.0
-
114.0
119.0
188.0
151.0
109.0
-
120.0
118.0
127.0
_ * e A.
TOC*, mg/1
Inter-
face
79.1
125.0
153.0
112.0
125.0
149.0
-
129.0
115.0
178.0
145.0
35.0
-
80.0
99.0
104.0
Ana-
lyzer
97.1
125.0
99.0
162.0
203.0
165.0
87.5
88.0
77.5
183.0
191.0
30.0
-
99.0
88.2
103.5
. 4-~U1 «
RETURN ACTIVATED SLUDGE
Date
(1974)
11-5
11-8
11-11
11-12
11-13
11-14
11-15
11-18
11-20
11-22
11-25
11-26
11-27
11-29
TOC,
Source
-
-
-
2530
1850
2080
-
2200
2890
1510
1510
2090
2050
2460
2510
2170
2300
mg/1
Inter-
face
-
-
-
1450
1670
2350
-
2650
3470
1970
1950
2060
2320
2810
2530
2970
2370
Ana-
lyzer
-
-
-
2080
2350
2540
1950
1400
2230
1950
1740
1350
1300
2520
2358
2619
2430
NOTE: The "Source" samples for these two tables were composites of several,
rapidly collected, grab samples taken directly from the mixed liquor
basin, or, from the return activated sludge line. Prior to analysis,
each composite was diluted by a factor representing the known
dilution factor (i.e., "dilution ratio") of the appropriate on-line
dilution pump.
45
-------
perform reliably and was not used. The data shown for this parameter were
obtained by grab sample analyses.
The continuous ammonia analyzer, which has the capability to monitor
ammonia nitrogen, nitrate and nitrite with the substitution of different
manifolding arrangements and reagents, did not determine ammonia satisfactorily
for this project. This ammonia analyzer was equipped by the manufacturer with
an outmoded, automatic-analyzer, wet-chemistry system (i.e., direct nessleri-
zation) which greatly promoted rapid fouling of the analyzer's optical
components; this, in turn, led to intolerable maintenance requirements and
excessive analyzer downtime. A great amount of time was expended on attempting
to obtain acceptable on-line ammonia data; hence, little time or funds were
left to monitor nitrate or nitrite. These latter parameters finally had to
be excluded from the field study.
Comparison of Source and Interface Values
Adequate agreement was achieved between source values and analyzer interface
values for almost all streams investigated; i.e., deviations generally fell
within the combined errors due to grap sampling and to the standard method
of analysis being used (see Appendix A).
The primary effluent stream (stream 2) yielded the most consistent data. This
stream was high in suspended and colloidal solids; therefore, small losses
or gains of solids as the primary effluent was being transferred did not
significantly affect the results.
The secondary effluent, on the other hand, was a very clean stream for which
any loss or gain of solids would greatly affect the results. This was
particularly true for total organic carbon analyses, and was probably the
reason why the standard deviation was greater for the secondary effluent
data than for the primary effluent data.
The raw influent values varied for another reason. This stream, at the
front end of the plant, was subject to sudden and wide changes in contaminant
concentration and composition. The variety and distribution of floating and
suspended material made the obtaining of representative samples extremely
difficult. As expected, the raw influent stream measurements showed the
greatest variance.
Analysis of the data (again see Appendix A) demonstrated that streams
containing very high solids concentrations can be monitored effectively when
dilution pumps are properly applied.
Test Results^from Automatic Analyzers
Although various difficulties, as noted below, were encountered with some of
the automatic analyzers, the results obtained helped verify (to a limited
extent) the performance acceptability of the sampling, transfer and condition-
ing system's components and the integrated system's operating reliability.
Primarily, however, the performance of most of these automatic on-line
analyzers merely emphasized* that commercially available and truly reliable
46
-------
instrumentation of this type (applicable to wastewater-treatment process
streams), is severely limited, both in variety and in number of suppliers.
TOC-SOC; Initial TOG problems made it necessary to install two TOG analyzers
simultaneously (Raytheon Company's TOG analyzer was ultimately selected for
final testing of the sample transfer and conditioning system) . Unfortunately,
the use of two TOG analyzers in the limited space of the Experimental Trailer
brought about an unfavorable positioning of several analyzers, both TOG and
colorimetric. This "unfavorable positioning" involved the placement of most
of the analyzers further from their sample interfaces than was desirable for
optimum results. The TOG measurements, especially, were adversely affected
by this situation. It should be noted, however, that these TOG variations
were not unidirectional; instead, the data exhibited both high and low biases.
Unduly long transfer lines linking TOG interface (i.e., homogenizer effluent)
to the TOG analyzer allowed solids to settle out slowly. At first, such
solids settling would tend to produce slightly low values. However, when a
sufficient amount of solids had settled out and the transfer lines had thus
become narrowed and non-uniform in bore, the resultant sporadic increases in
sample velocity would suddenly scour the lines and cause entrainment of
deposited solids by the sample stream. This in turn would produce occasional
"high" TOG values.
Qrthophosphate; The orthophosphate analyzer was properly located, hence it
operated most consistently. All three values, (source, interface, and
automatic analyzer) agreed very well.
Hydrolyzable phosphate; Measurements were not made with an automatic analyzer.
Ammonia nitrogen; The ammonia measurements suffered because of improper
placement of the ammonia analyzer, but even more because of random equipment
malfunction due to the obsolete wet-chemistry system furnished by the
colorimetric analyzer's manufacturer.
Nitrate and Nitrite; Measurements were not taken because the automatic
analyzer was devoted almost solely to ammonia-nitrogen samples.
Statistical Analysis
A statistical analysis of some of this project's final test results was
conducted by the EPA, and the statistical findings are the bases of the
claim for the acceptability of the sample transfer and conditioning system's
on-line performance. The statistical analysis is included as part of
Appendix A.
47
-------
SECTION VIII
REFERENCES
1. Sugar, J.W., and Brubaker, J.H., "Development of Sample Conditioning
Systems for Automatic Environmental Instrumentation." Presented at the
19th Annual ISA Analysis Instrumentation Symposium, St. Louis, Missouri,
April 24-26, 1973.
2. Shelley, P.E., and Kirkpatrick, G.A., "An Assessment of Automatic Sewer
Flow Samplers." EPA-R2-73-261, June 1973.
3. Houser, E.A., Principles of Sample Handling and Sampling Systems Design
for Process Analysis. Instrument Society of America, Pittsburgh, Pa.,
1972.
4. American Public Health Association, Standard Methods for the Examination
of Water and Wastewater, 13th ed., APHA, New York, 1971.
48
-------
APPENDIX A - STATISTICAL ANALYSIS
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
o,,n,ror Testing for Statistical Difference
SUBJECT: Between the Sample Source and the Sample °ATE: Janu*ry 7, 1975
Interface Methods of Data Measurement
FROM*
R. G. Eilers and Ella Hall
Systems & Economic Analysis Section
Robert H. Wise
Pilot & Field Evaluation Section
This analysis is intended to determine statistically
if there exists any evidence of a systematic difference
between the sample source and the sample interface methods
for measuring concentrations of TOC, SOC, O/PO4, NH3, and HP
(hydrolyzable phosphorus) in wastewater. A statistical test
was applied to 17 paired sets of measurement data, and the
results appear in Table 1.
The reference for this procedure is the book entitled
"Statistical Analysis in Chemistry and the Chemical Industry"
by Carl Bennett and Norman Franklin, John Wiley and Sons (1966),
pages 180-182.
In order to illustrate the statistical theory involved
here, a detailed calculation for the TOC-Secondary measure-
ments will be given. The raw data consisted of the following
13 paired measurements along with their respective differences:
Observation Source Interface Difference, d
1 13.2 13.4 -0.2
2 34.0 33.0 1.0
3 30.0 27.0 3.0
4 21.0 24.0 -3.0
5 22.0 20.0 2.0
6 26.0 25.0 1.0
7 23.0 21.0 2.0
8 3O.O 28.0 2.0
9 33.0 30.0 3.0
10 30.3 30.3 0.0
11 24.0 26.0 -2.0
12 22.0 26.0 -4.0
13 25.0 22.5 2.5
49
-------
The mean of the differences, d = .5615, and the standard
deviation of the differences, s = 2.2948, are both calculated
and Student's t - Test is applied according to the equation:
- .5615 (13)'5 _ .5615 (3.6056) _
2.29482.2948"
where n = 13 is the sample size with (n-1) = 12 degrees of
freedom. Referring to the Students t - Distribution Table,
the values of t12 .05 = 2»179 (12 degrees of freedom, 5% level
of significance, two-tailed distribution) and t-j^ 0-^ = 3.055
(12 degrees of freedom, 1% level of significance,'two-tailed
distribution) are selected. Since Itl = .8822 < 2.179 and
(t| = .8822 < 3.055, it can be concluded that there is no
evidence of a systematic difference between the two methods of
measurement at both the 5% and 1% levels of significance. What
this means, simply, is that if |t| > 2.179 the possibility of
the two methods being statistically equivalent is only 5% or
less. Similarly, if |t| > 3.055, the possibility of the two
methods being statistically equivalent is only 1% or less.
In practice a level of significance of .05 or ,O1 is
customary, although other values can be used. If, for example,
a 5% level of significance is chosen in designing a test of
hypothesis (the hypothesis in this case is that the two methods
are statistically equivalent), then there are about 5 chances in
100 that the hypothesis would be rejected when it should be ac-
cepted, i.e., 95% confidence exists that the right decision has
been made. In such a case it is said that the hypothesis has
been rejected at a .05 level of significance, which means a
.05 probability of being wrong.
50
-------
TABLE 1
RESULTS OF STATISTICAL COMPUTATIONS COMPARING
SAMPLE SOURCE AND SAMPLE INTERFACE
I t I t(n-l),.Q5 t(n-l)..01
Conclusion
TOC-Secondary
SOC-Secondary
o/PO, -Secondary
NHL-Secondary
HP-Secondary
TOC-Primary
SOC-Primary
o/PO, -Primary
NH3-Primary
HP-Primary
TOC-Raw
SOC-Raw
o/PO -Raw
4
NH -Raw
HP-Raw
TOC-Mixed Liquor
TOC-Return Sludc
.882
1.014
3.117
1.399
2.295
1.181
.809
1.847
2.615
6.099
4.744
.737
2.031
.378
.031
1.065
e 1.147
2.179
2.179
2.069
2.145
2.110
2.179
2.179
2.093
2.131
2.120
2.179
2.262
2.093
2.145
2.120
2.160
2.179
3.055
3.055
2.807
2.977
2.898
3.055
3.055
2.861
2.947
2.921
3.055
3.250
2.861
2.977
2.921
3.012
3.055
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 significant diff,
.01 significant diff,
.05 no difference
.01 no difference
.05 significant diff.
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 significant diff.
.01 no difference
.05 significant diff.
.01 significant diff.
.05 significant diff.
.01 significant diff.
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
.05 no difference
.01 no difference
51
-------
APPENDIX B
OPERATION AND MAINTENANCE
The Sampling Assembly consists of two main sub-assemblies: the exhaust
manifold and the sample manifold. These two manifolds are connected
together by means of two-way motor-driven ball valves. These valves are
"True Union" ball valves which provide for assembly and disassembly
without additional pipe unions. Each pair of ball valves is operated as a
unit to act as one three-way valve; when one is open, the other is closed.
They are numbered in the following manner (note that the "A" position of
any two-valve pair always results in a sample entering the sample manifold,
while the "B" position always diverts that sample to the exhaust manifold):
VI(A) permits secondary effluent to enter sampling manifold
VI(B) permits secondary effluent to enter exhaust manifold
V2(A) permits primary effluent to enter sampling manifold
V2(B) permits primary effluent to enter exhaust manifold
V3(A) permits raw influent to enter sampling manifold
V3(B) permits raw influent to enter exhaust manifold
V4(A) permits mixed liquor to enter sampling manifold
V4(B) permits mixed liquor to enter exhaust manifold
V5(A) permits return activated sludge to enter sampling manifold
V5(B) permits return activated sludge to enter exhaust manifold
V6(A) permits primary sludge to enter sampling manifold
V6(B) permits .primary sludge to enter exhaust manifold
V8(A) prevents sample from going to the Pretreatment Assembly
V8(B) allows flush water to be supplied to the Pretreatment Assy, while
streams 4, 5, and 6 are being monitored.
V7(A) allows flush water to be supplied to the sample manifold and
homogenizer plumbing following the sampling of stream 6.
52
-------
The purpose of valve pairs VI through V6 is to introduce dynamic samples,
one at a time, to the sample manifold through "valve A". In this way, all
sample-supply pumps run continuously to avoid settling of solids in the
transfer lines.
Valve 7 is used to introduce tap water to the homogenizer plumbing; this
simultaneously flushes the plumbing, homogenizer and TOC-analyzer input
lines. In the manual mode, any one of these sample streams can be intro-
duced to the sample manifold by depressing the push button associated with
the desired sample stream. When it is desired to select another stream,
the "All to Drain" button (located in lower right quadrant of the control
panel) is pushed, then this is followed by depressing the button associated
with the desired alternate stream. However, if the stream desired is the
next stream in the programmed sequence only the button associated with that
sample should be pushed. The circuitry permits forward sequencing, but
requires a reset ("All to Drain") to go backwards in the sequence. Note:
when the sequence is performed manually, the buttons must be held until all
valve-drive motors have completed their cycles (approximately 2-4 seconds).
Valve-pair V8 cannot be manually actuated from the front panel. It operates
so that sample streams 1, 2 or 3 (when selected) will be furnished to the
Pretreatment Unit through Valve 8A. When sample streams 4, 5 or 6 are
selected, Valve 8A closes and tap water is introduced through Valve 8B to
the Pretreatment Unit and the colorimeters.
Valve-pair V9 is not found on the manifold, but is located in the TOG analyzer.
Its purpose is to furnish to the TOG analyzer one of two samples: a homo-
genized sample to monitor TOG, or a filtered sample from the Pretreatment
Unit to monitor SOC. In the installation at Cranston, ball valves were not
used for V9. Two peristaltic pumps were installed with a common output
connection. Pump "B" is energized to sample for TOG and it pumps the sample
from the homogenizer. Pump "A" is energized to sample for SOC and it pumps
the sample from the pretreatment output. The pump not energized acts as a
closed valve and prevents mixing of the two samples.
This was designed to be a function of the TOG analyzer because of the
impracticability of switching small streams with ball valves, and the relative
ease of doing it with a pump in the TOC analyzer. The function, V9, cannot
be selected or controlled from the front panel. Any time a new sample is
selected, V9 will operate in the "B" mode which furnishes homogenizer output.
Only when the system is in the automatic mode and has run through a timing
cycle on samples 1, 2 or 3 will the SOC mode be selected for the next cycle
on the same sample. If SOC mode is not desired, then the TOC unit's first
input pump should be energized from its normal supply. Placing the wired
"Dummy Plug" (Figure B.I) in J9 on the back of the Control Panel will allow
the system to skip the SOC cycle completely.
53
-------
13579
2 4 6 8 10
T T-^T T
Figure B. 1 Dummy Plug Wiring to Skip SOC Mode
(Numbers shown are the pin numbers of
the dummy plug actually used).
'In automatic operation the system will progress through the sample
streams in increasing numerical order starting with the one selected
by depressing a pushbutton. After stream #7 (flush water) the system
will revert back to stream //I. The time duration of each sampling
cycle is controlled by the setting of the left timer (facing the unit).
Figure B.2 shows the timing diagram for the system, and Table B.I shows
the operational sequence when a stream is selected.
54
-------
VALVE T0 T, 1
N°- ON ' '
I. SECONDARY
EFFLUENT
OFF-
ON-
2. PRIMARY
EFFLUENT
OFF-
ON-
3. RAW
INFLUENT
OFF-
ON-
4 MIXED
LIQUOR
OFF-
5. RETURN ON-
ACTIVATED
SLUDGE
OFF-
ON-
6. PRIMARY
SLUDGE
OFF
ON-
7 FLUSH
WATER
OFF-
C A fclDI C"
2 T(3 T4 T5 T6 T7 T
8 T9 T0
lr
n
a PRETREATMENT
INPUT FLUSH J
WATER J
9, TOC HOMOG"
ANALYZER
INPUT pRETREAT_
ST/
CY(
kRT
:LE
^"^
Tfl
(11
i
ITAL
JTIM
"^
SAMP
ERIN
;
LING
TER\
CYC
fALS)
LC
n_
n
i
.
_
END'OI
CYCLE
Figure B.2 Timing Diagram
55
-------
TABLE B.I. TABLE OF OPERATION SEQUENCE
Ul
Time
j reference
To
T
1
T2
T3
T4
T5
T6
T7
T8
T9
Stream
Secondary
effluent
Secondary
effluent
Primary
effluent
Primary
effluent
Raw
influent
Raw
influent
Mixed
liquor
Return
activated
sludge
Primary
sludge
Flush
water
Relays
energized
K1A, K8
K1A.K1B.K8,
K9
K2A, K8
K2A.K2B.K8,
K9
K3A, K8
K3A.K3B.K8
K9
K4A, K4B
K5A, K5B
K6A, K6B
K7
Relays
deenergized
Remainder
Remainder
Remainder
Remainder
Remainder
Remainder
K8, remainder
K8, remainder
K8, remainder
Remainder
Valves
opened
VIA, V8A,
all other "B"
valves
V1B.V8A.V9A,
all other "B"
valves
V2A, V8A,
all other "B"
valves
V2A.V8A.V9A,
all other "F1
valves
V3A, V8A,
all other "F'
valves
V3A, V8 A, V9A,
all other "F1
valves
V4A,V8B,V9B,
all other "F1
valves
V5A.V8B.V9B,
all other "F1
valves
V6A.V8B.V9B,
all other "B"
valves
V7,
all"F' valves
Valves
closed
V1B, V8B,
all other "A"
valves
V1B.V8B.V9B,
all other "A"
valves
V2B, V8B,
all other "A"
valves
V2B,V8B,V9B,
all other "A"
valves
V3A, V8B,
all other "A"
valves
V3B.V8B.V9B,
all other "A"
valves
V4B.V8A.V9A,
all other "A"
valves
V5B.V8A.V9A,
all other "A"
valves
V6B.V8A.V9A,
all other "A"
valves
All "A" valves
Action
accomplished
Sec. effl. to sample manifold,
TOC, o-PO , NH -N, hyd. PO,
..,43 4
monitored
Sec. effl. to sample manifold,
SOC, o-PO,, NH -N, hyd. PO.,
4 O 4
monitored
Prim. effl. to sample manifold,
TOC, o-P04, NH3-N, hyd. PO4>
monitored
Prim. effl. to sample manifold,
SOC, o-P04, NH3-N, hyd. PO4,
monitored
Raw infl. to sample manifold,
TOC, o-P04> NH3-N, hyd. PO4,
monitored
Raw infl. to sample manifold,
SOC, o-P04, NH3-N, hyd. PO4,
monitored
Flush water to P/T and ana-
lyzers, TOC monitored
Flush water to P/T and ana-
lyzers, TOC monitored
Flush water to P/T and ana-
lyzers, TOC monitored.
All streams to drain, flush
water to entire system
-------
Theory of Operation
Each valve pair is controlled by the two double-throw contacts of its
associated relay; i.e., valve-pair VI ("A" and "B") is plugged into Jl where
it will be controlled by K1A, etc. 115 volts AC is furnished through the
de-energized relay contacts to close VIA and open VlB. When the relay is
energized, 115VAC is furnished to open VIA and close VlB. When the valves
are operating normally, one valve of a pair will always be open and the
other closed.
When the unit is first turned on, all sequence relays are deenergized. The
unit's operator can depress any one pushbutton, SI thru S7, to energize its
associated relay. (Note: SI is associated with valve-pair VI ("A" and "B"),
Jl, and Kl, etc.) This opens the associated Valve A (closing Valve B) and
permits that sample stream to flow to the sample manifold.
Jacks Jl thru J9 are wired so that the "A" Valves are controlled through the
even-numbered contacts (reference schematic shown in Figure B.3 for a typical
valve pair). Each valve is powered with a 115V motor by means of a cam-
operated double-throw micro-switch. The voltage which drives the motor is
returned to the panel through the activated micro-switch when the valve is
in its selected position. The return voltage lights an indicator lamp that
shows the status of the valve (green light indicates valve open, and red
light indicates valve closed). For valve-pairs VI through V7, the return
voltages fed back from the "A" valve are used to control the relay logic
for the automatic sequential operation.
On the ladder diagram (Figure B.4) find K11A, K11B, K11C, K12, and K14.
Observe that K11A, K11B and K11C operate together as one relay of eleven
contacts. When Kll is energized, its normally open (N.O.) contacts connect
the coil of each relay (Kl thru K7) to the green-light circuit of the "A"
valve that precedes it (K7 to K6, K6 to K5, etc.). Relay Kll is energized
when the timer K12 completes its time cycle. K12 (N.O.) contacts close,
energizing Kll. Kll contacts 9 to 5 close, energizing K14. K14 contacts
1 to 4 open, resetting timer K12 and deenergizing Kll and K14. K14 is a
delay-on-release relay; this delay is necessary to allow K12 time to reset.
Each time Kll is operated, it applies 115V to the coil of the next relay
(Kl through K7) in the sequence. Actuating any relay from Kl through K7
causes its associated valves to change condition: "A" valves open and "B"
valves close. As valve "A" starts to open, its microswitch S2 changes con-
dition, removing the holding voltage from the immediately preceding relay
in the sequence. Deenergizing that relay causes its valves to assume the
condition of valve "A" closed - valve "B" open.
The holding circuits of each relay (Kl thru K7) are wired from the red light
of the next valve in the sequence through normally closed (N.C.) contacts of
K21A and K21B. When depressed, switch S10 ("All to Drain"), energizes K21,
opening the holding circuits of all sequence relays and deenergizing any
energized relays. Energizing K21 also resets the K12 timer.
57
-------
LI
L2
liSVAC
(HI)
11" 7f. /,, , ( 12
v v \ Ji i v o
KIA | 1
L ' '
1 1 Ol ' 3
I. J !
1 THROUGH 6 8 8. FOR VALVE PAIR 2 USE K2A 8 J2,
FOR VALVE PAIR 3 USE K3A 8 J3 8 ETC.
KIA SHOWN OEENERGIZED. VALVE A HAS RUN TO
A CLOSED POSITION.VALVE B HAS RUN TO AN
OPEN POSITION.
i 1 ! *
112 1 > 41 f. .( Iz
9 Til <; Jl 4 <. (j>
KIA ' 1
A <* ii - f A
8] 3
I12 84 tJT\( A7
T T OTK T
K7 { |
L _ !! 1 '
i9
T
1
1 ^~~
1
L >Jll> O
,11
*- 1 J I
f !
-------
WHEN VALVE A IS CLOSED WHEN VALVE A IS OPEN
S2 CLOSES PIN 3 TO PIN 5 /SI CLOSES PIN 2 TO PIN 4
LI
L2
t /
12
K
II
\
t
||
K7 V7
IA x ._ . x 2
'A-
SI
4 v « -.\_
f2 J4^)-
SI
|4,,,7V_
P J3-7/
IA-S2
/fe-i .K-QV
IA
f '
1
KIB
"1
1
-f
!|
1
7
3
H
10
^
5
5
H
II
K
9
5
\
5
K
12
,._
K2
j
21
V
3
7
?4
1
h
10
1
5
1
9
t
9
t
II
<3
K2I
B
7
K
II
B
f 5
^| CJ5-3 « } | x ,- _x ^ |
KA4 V4
ill |7 f...( 2
U| | A-
H
SI
4 ^ < T^
S2 K2I
^>JC3v I0|>
B
1
10
5
ll/
:i/s
(IE
n,
IK
:2,
:2i
n/
53
IK
:a
;3i
HE
34
III
"
55
III
C5
,
6
J
9
\
7
5 ,
9
5
9
8 .
6
k
9
i
5
5 ,
*
7 ,
9
-I
6
i
9
,. isfl14
U
KIA
°ni4
u
KIB
K2A
K2B
K3A
K3B
I3r|l4
2 ~ 7
K4B
K5A
I3FI4
2 ~7
K5B
SWITCHES OF "A" VALVES ARE SHOWN WITH SYSTEM IN "ALL TO DRAIN"
MODE ;l.e..ALL "A" VALVES ARE CLOSED ("B" VALVES OPEN)
Figure B.4 Ladder Wiring Diagram (Sheet I of 2)
59
-------
LI
L2
S6
H
II
12
!5A V5A-SI KIIB
l7 x.. . f 2 |« \ .ITS "1 7
' VP-I * " ' > J» '/
<7 V7-S2 K2IB K6A
>W / x 3l X5 v.--v lll> 3 5 9
S7
!
II
II
<6A V6A-SI KIIB
I7 f ir l f 2 I4 ^ JC Tl l2 8
<|A VA-S2 K2IB K7
xf3 x.. , x 3 xfc v .. -,v 12 |4 5J |9
P| 4 1 ? 1 II
K4B K5B K6B K7
yfe lU3 lU3 II xft
n -r i i r i
KIB
12 8
K2B
12 8
K3B
12 8
KI2N.O.
SIO
Sll
1 2
x^]
KI4 K2IA Sll x
|>4 9U1 3 1-4
Kl
Af
L
(
K
13
13
KIIB
13
KIIC
K
13
13
K
LI C
\
CL2 /"
V.
KI2
2l
K
Il4
K7
KB
0s
K9
14
14
14
K2IA
14
14
M
L2
51 19
KIIA
Figure B.4 Ladder Wiring Diagram (Sheet 2 of 2)
60
-------
The selection of TOC/SOC by valve pairing is achieved by a modification of
the basic logic of Kl, K2, K3, and K4. K1A, K2A, and K3A control valve pairs
VI, V2 and V3 respectively; K1B, K2B and K3B control V9 through K9. If K9 is
deehergized (i.e., if V9 is functioning in the TOG mode), and if K1A or K2A
or K3A is suddenly energized, the actuation of Kll causes the associated K1B
or K2B or K3B to energize concurrently. Energizing K1B, K2B or K3B causes K9
to energize (through N.O. contacts 12 to 8), thus actuating V9 to the SOC
mode. When K9 is energized by energizing a valve-pair (V1A-V1B, V2A-V2B, or
V3A-V3B), the immediately preceding actuation of Kll causes the "A" relay
of the next valve pair to energize. The sequence is as follows: K1A, K1A &
K1B, K2A, K2A & K2B, K3A, K3A & K3B, K4A & K4B, K5A & K5B, K6A & K6B, K7,
then back to K1A, etc. (K4A & K4B operate together as one 5-contact relay
as do K5A & K5B and K5A & K6B.) If the dummy plug shown in Figure B.I is
inserted into J9 (instead of a valve pair), the voltage from J9 (pin 7) is
returned instantaneously to K1B or K2B or K3B contacts 11 to 7; this ener-
gizes the next "A" relay in sequence, causing that valve pair to actuate.
The sequence then is Kl, K2, K3, K4, K5, K6, K7, then back to Kl, etc.
K8 is energized when K4, K5, K6, & K7 are deenergized. Operating any
relay K4 through K7 opens a series-arranged contact pair which deenergizes
K8. When K8 is deenergized, it operates valve-pair V8 to allow flush water
to flow to the pretreatment unit.
J10, Jll, and J16 are trouble-shooting aids, and are hard-wired so that any
valve pair (when plugged into) will do the following:
J10 - opens valve A, closes valve B
Jll - opens valve B, closes valve A
J16 - closes both valves
The voltages from the divider network (IV. to 7V.) are made available at
J13-3 (Hi) & J13-1 ("0"V) to give remote indication of the program status.
1 volt indicates that the TOC for sample 1 is being monitored, 1.5V indi-
cates that the SOC for sample 1 is being monitored, 2V Indicates sample 2
TOC monitoring, etc.
The contacts of timer K13 actuate K15 and K16 to operate auxiliary equipment
(i.e., a data acquisition system and a paper-tape-punch recorder). For the
period of time that K13 (N.O. contact) is closed (before K12 completes its
time cycle), 115V AC is furnished to J14 to operate the tape punch, and the
circuit from J13 (pin 23) to J13 (pin 27) allows the data acquisition system
to print data. K16 is a delay-on-energized relay. When K15 and K16 are
energized, the circuit from J13 (pin 10) to J13 (pin 16) closes for 1 second
(delay of K16), and the circuit from J13 (pin 13) to J13 (pin 19) opens for
1 second; this operates the "All Channel Buzz" of the tape punch, a solenoid
protection device peculiar to this particular punch mechanism. K13 is reset
along with K12 when K12 "times out" and operates Kll, K14.
61
-------
APPENDIX C
DESIGN SPECIFICATION GUIDELINES
T. GENERAL
A. This specification applies to the design of an on-line hardware
system which will automatically sample, transfer and condition
all types of wastewater-treatment process streams for automatic
analysis without the occurrence of unacceptable chemical change
in the samples prior to their analysis.
B. Application
1. Municipal wastewater-treatment plants
C. The sampling system shall consist of a series of pumps and associ-
ated piping at appropriate locations within a wastewater-treatment
plant. These pumps shall supply their various samples to a
centrally located sampling assembly which has the capability to
select any one of the streams and condition it sufficiently so
that the processed sample is adaptable to automatic on-line TOG
and colorimetric analyzers. It shall have an interval flush cycle
which can flush the total system after the completion of a total
sampling cycle. The sampling system shall be controlled by means
of a control panel which can be operated either manually or auto-
matically.
D. To minimize transfer-line contamination, dilution pumps shall be
utilized when the streams being sampled have suspended solids
loadings of 1000 mg/1 or greater.
II. TECHNIQUE
Each sample shall be supplied continuously to pairs of two-way valves
which operate in tandem so as to simulate a three-way valve. The valve
pairs shall be connected by means of union-type plumbing fittings to a
sampling assembly which shall consist of an exhaust manifold and a
sampling manifold. These valve pairings shall permit each sample to
flow continuously, either to the sampling manifold or to the exhaust
manifold. Once the sample stream reaches the sample manifold, it shall
be valved to a Pretreatment Assembly (for removal of all particulate
matter in preparation for colorimetric analysis) or to a homogenizer
(in preparation for TOC analysis). The sampling of all process streams
to be analyzed shall be a sequential operation. When the sampling
62
-------
sequence has been completely traversed, the sampling assembly shall be
automatically flushed with tap water prior to the initiation of another
sampling sequence.
III. OBJECTIVES
The system's objectives shall be:
A. To provide for multiple stream monitoring within a wastewater-
treatment plant without altering any sample's initial chemical
composition.
B. To provide control signals to allow process automation.
IV. GENERAL PERFORMANCE CONSIDERATIONS
A. To handle the liquid streams within a wastewater-treatment plant
(raw influent, primary effluent, and secondary effluent), cen-
trifugal grinder pumps shall be used. These pumps shall be
capable of reducing occasional large particles of suspended
solids to a size of 1/4 inch (6.35 mm) which is small enough
to allow an in-line homogenizer to function continuously. How-
ever, when the stream contains a large amount of fibrous material
(specifically raw influent), a screen shall be utilized to prevent
entrance of these fibers into the pump.
B. To handle streams having higher solids loading (mixed liquor,
return activated sludge and primary sludge), a dilution pump
arrangement shall be used. The sample shall be diluted at the
origin to minimize solids loading within the transfer lines. Dilu-
tion shall be accomplished by utilizing a duplex pump driven by a
common drive; this maintains a constant dilution ratio, even though
rotor speed may vary. Each side of the pump shall have an adjustable
stroke-setting to vary flowrates for desired dilution ratios. The
total flow and the dilution ratios required will determine the size
of the check valves. The pump shall be driven by a standard,
constant speed, AC-drive motor or a suitable speed-controlled motor.
The latter configuration will allow for varying total flow once a
dilution ratio is established.
C. The in-line homogenizer shall utilize an abrasive rotor-stator
combination to reduce particle size. The homogenizer must be able
to reduce all types of particulate matter (i.e., plastics, paper
fibers, and woodchips), as well as sewage, to a finally divided size
on a continuous basis. Rotor clearance shall be adjustable by alter-
ing the gap between the rotor and stator to achieve a wide range of
particle sizes.
D. The Pretreatment Unit shall be a packaged filtration system which
provides continuous flow of representative samples for up to six,
on-line, water quality, monitoring instruments, while removing
virtually all solid particles above 10 micrometers in size.
63
-------
V. ELECTRICAL SPECIFICATIONS
A. Input Power
The central location which contains the sampling assembly and
conditioning equipment shall have provision for 100-ampere, 3-phase,
230-volts, 60-HZ, AC power. The sampling system requires 50 amperes;
the additional power is required for automatic analyzers.
B. Output Signals
The control panel shall have a voltage divider network serving as an
indicator of which sample is being monitored in the following manner:
Indicator
Voltage Sample Stream Parameter Monitored
1.0 Secondary effluent Colorimetric analysis & TOC
1.5 Secondary effluent Colorimetric analysis & SOC
2.0 Primary effluent Colorimetric analysis & TOC
2.5 Primary effluent Colorimetric analysis & SOC
3.0 Raw influent Colorimetric analysis & TOC
3.5 Raw influent Colorimetric analysis & SOC
4.0 Mixed Liquor TOC
5.0 Return activated sludge TOC
6.0 Primary sludge TOC
7.0 Flush water Complete System Flush
VI. MECHANICAL SPECIFICATIONS
A. The sample assembly, along with the control panel and homogenizer,
shall not exceed 4 ft. (1.22m) width, 2-1/2 ft. (.76m) depth and
6 ft. (1.83m) height.
B. Flush-water requirements shall not exceed 5 gal/min (18.92 1/min) at
20 psi (137.9 kN/nr) (Note: this water is not used on a continuous
basis but must always be available.)
C. The sampling manifold shall be firmly mounted to the floor.
D. Positioning of automatic analyzers is very important. Automatic
analyzers shall be located as close as possible to the source of the
conditioned sample. Where this is not practical, analyzer input
velocities shall be investigated and the associated plumbing shall
be adjusted to minimize line contamination. (Minimum velocity shall
be no less than 1 ft/sec. [.30m/sec.].)
64
-------
VII. SAMPLE REQUIREMENTS
A. The dilution pumps must not be operated with positive input
pressure. Therefore, the high-solids streams shall have a
reservoir-type feed so that the pumps may draw the required
sample. The reservoir must be continually replenished so that
a representative (up-to-date) sample is always available.
VIII. ADDITIONAL SERVICES REQUIRED
A. Additional services are dependent upon the requirements of the
automatic analyzers selected; i.e., reagents, bottled gas, etc.
IX. CONTROL AND INDICATORS
A. The control panel shall be graphically representative of the
flow diagram. Red (no flow) and green (flow) indicator lights
shall be incorporated to display the sampling status.
B. The system shall have two modes: automatic and manual
1. The automatic mode shall be controlled by a timer located
on the front panel. Cycle time shall be manually selectable
for times up to 1 hour.
C. A second timer shall be incorporated to start up auxiliary
equipment (e.g., data-acquisition system) at any intermediate
point within the cycle period.
X. ENVIRONMENTAL
A. Ambient Temperature:
34°F (1 °C) to 104°F (40°C)
B. Humidity:
0-95%, non-condensing
C. The grinder pumps shall be submersible pumps. The dilution pumps
shall withstand adverse weather conditions.
NOTE: If the temperature goes below freezing, adequate flow must
be maintained to prevent freezing within tha lines.
D. The sampling system shall be contained within a shelter which is
environmentally controlled (heated and air-conditioned).
XI. SAFETY PROVISIONS
A. In the event of a leak in a sample line within the shelter, all
sample flow shall be directed to drain by pushing the "ail to drain"
button on the control panel.
65
-------
B. Each pump shall have its own, independent, overload circuit
located within the shelter.
66
-------
APPENDIX D
LIST OF EQUIPMENT
Estimated Costs (1976 dollars)
Item
D
2)
3)
4)
5)
6)
7)
8)
9)
10)
11)
12)
13)
14)
15)
Pump, Grinder
Control Box
Pump Dilution
with Control Box
Pump, Dilution with
Control Box
Mixer, Static
Homogenizer
Hose, 1"
Hose, 3/8"
Assembly, Control
Panel and Rack
Assembly, Grinder
Pump Cable
Assembly, Dilution
Pump Cable
Assembly, Connector
Panel
Pump, Peristaltic
As semb ly , Samp ling
and Exhaust
Cabinet, Reagent
Pretreatment Unit
Total Equipment Cost
Manufacturer
Hydr-0-Matic
Pump Co.
BIF
BIF/Seco
Kenics
Raytheon
B. F. Goodrich
B. F. Goodrich
Raytheon
Raytheon
Raytheon
Raytheon
Randolph
Raytheon
Raytheon
Raytheon
Part #
SPG-150A2
1722-92-9517
1722-92-9510
37-08-136
2650
BFG300
BFG300
Special
Special
Special
Special
Special
Special
2590
2550
Qty. Unit
3 875
2 690
1
3 145
i ___
670 ft. .94/ft.
1,000 ft. .36/ft.
1
1,000 ft. 1.54/ft.
1,000 ft. .66/ft.
1
1
1
2 650
1
Total
2,625
1,380
1,020
435
1,350
630
360
1,320
1,540
660
320
126
4,598
1,300
3.950
$21,614
-------
APPENDIX D - LIST OF EQUIPMENT (Cont'd)
Name
TOC Analyzer
Qrthophosphate Analyzer
Monitor IV
Data Acquisition System
TOC Analyzer
Microscope
Spectrophotometer
AUXILIARY EQUIPMENT
Manufacturer Part
Raytheon 2600
Raytheon
Technicon
Esterline Angus D2020
Beckman 915
Bausch & Lomb XL1
Bausch & Lomb
Qty
1
1
1
1
1
1
1
68
-------
Contamination
Deadending
Grab Samples
Hyd. P04
Interface
o-P04
P/T
SOC
Source
Sparging
Time-Lag
TOC
GLOSSARY OF TERMS AND ABBREVIATIONS
Settled particles in transfer lines.
Allowing a stream to stop flowing.
Samples taken by hand.
Hydrolyzable phosphate expressed as phosphate.
That point at which all sample transporting and
conditioning have been performed.
Ammonia expressed as nitrogen.
Orthophosphate expressed as phosphate.
Pretreatment Assembly (Filtration Unit).
Soluble organic carbon.
That point (usually a unit process) at which sampling
originates.
Process by which inorganic carbon is removed from a solu-
tion by agitation with a CO.-free gas.
Time during which valid data cannot be obtained.
Total organic carbon.
69
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-76-146
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Wastewater Sampling, Transfer and Conditioning System
S. REPORT DATE
October 1976
(Issuing date)
6. PERFORMING ORGANIZATION CODE
7, AUTHOR(S)
Louis S. DiCola
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Raytheon Company
Submarine Signal Division
P.O. Box #360, Portsmouth, Rhode Island
10. PROGRAM ELEMENT NO. 1BB043
ROAP 21 - ASC; Task No. 20
11. CONTRACT/GRANT NO.
02871
Contract No. 68-03-0250
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio ';5268
13. TYPE OF REPORT AND PERIOD COVERED
Final: .Tnng 1Q7^-Marrh
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report describes the construction and field evaluation of an automatic
on-line hardware system for reliably sampling, transferring, and conditioning
various wastewater-treatment process streams such that the resulting transferred
and conditioned samples are suitable for interfacing with automatic on-line
colorimetric and total organic carbon analyzers. Process streams to which this
hardware system was sucessfully applied included raw sewage, primary effluent,
secondary effluent, aeration tank mixed liquor, and return activated sludge.
Primary sludge could not be sampled at the field-testing site because the sludge
had become too thick at its only feasible access point. Analytical parameters
used to evaluate the hardware system included both total and soluble organic
carbon, orthophosphate, total hydrolyzable phosphate, and ammonia nitrogen.
Nitrate and nitrite were not included; however, the hardware system's performance
with the soluble parameters studied indicate that nitrate and nitrite should
present no special difficulties.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
Sewage
Waste water
Sampling
Continous sampling
Sequential sampling
Chemical analysis
Sequential analysis
Process control
Automatic control
Automatic sampler
In-line sampling system
Automatic analysis
Sample transport system
Sample transfer system
In-line homogenizer
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
80
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
70
1976 757-056/5407 Region 5-11
------- |