EPA-600/4-76-059
December 1976
Environmental Monitoring Series
PERFORMANCE INVESTIGATION OF THE MANNING
MODEL S-4000 PORTABLE WASTEWATER
SAMPLER AND THE MODEL F-3000
DIPPER FLOWMETER
Environmental Monitoring and Support Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into 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. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL MONITORING series.
This series describes research conducted to develop new or improved methods
and instrumentation for the identification and quantification of environmental
pollutants at the lowest conceivably significant concentrations. It also includes
studies to determine the ambient concentrations of pollutants in the environment
and/or the variance of pollutants as a function of time or meteorological factors.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/4-76-059
December 1976
PERFORMANCE INVESTIGATION OF THE
MANNING MODEL S-4000 PORTABLE
WASTEWATER SAMPLER AND THE
MODEL F-3000 DIPPER FLOWMETER
by
Richard P. Lauch
Instrumentation Development Branch
Environmental Monitoring and Support Laboratory
Cincinnati, Ohio 45268
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and
Support Laboratory - Cincinnati, U.S. Environmental Protection Agency,
and approved for publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
11
-------
FOREWORD
Environmental measurements are required to determine the quality of
ambient waters and the character of waste effluents. The Environmental
Monitoring and Support Laboratory - Cincinnati conducts research to:
Develop and evaluate techniques to measure the presence
and concentration of physical, chemical, and radiological
pollutants in water, wastewater, bottom sediments, and
solid waste.
Investigate methods for the concentration, recovery, and
identification of viruses, bacteria and other microbio-
logical organisms in water. Conducts studies to determine
the responses of aquatic organisms to water qualtiy.
Conduct an Agency-wide quality assurance program to assure
standardization and quality control of systems for moni-
toring water and wastewater.
The Instrumentation Development Branch, EMSL, has provided functional
designs relating to water quality instrumentation systems. This report,
which investigates an automatic wastewater sampler and flowmeter, provides
considerations for field personnel in acquiring samples for wastewater
monitoring.
Dwight G. Ballinger
Director
Environmental Monitoring and Support Laboratory
Cincinnati
111
-------
ABSTRACT
Performance of the Manning model S-4000 wastewater sampler and the
model F-3000 flowmeter was investigated.
The S-4000 wastewater sampler was tested at temperatures of 2, 20, and
35C to determine accuracy and precision of the timer and sample volumes.
The multiplexer function of delivering multiple aliquots per bottle was
tested. Tests for ability to fill up to four bottles with the same sample
were made. Battery endurance was determined. Discrete sample temperatures
versus time were recorded under iced conditions to determine preservation
capability. Field tests were performed to determine representative collec-
tion of suspended solids and ability of the unattended sampler to collect
raw sewage samples over a 24-hour period.
The F-3000 flowmeter was tested within the laboratory for accuracy and
precision of tracking, analog to digital conversion, deadband, and elec-
tronic drift caused by temperature change and battery decay. Accuracy of
the flow chart and integrator was determined.
Manufacturer's claims were mostly confirmed, however improvement is
warranted for some functions of the sampler and flowmeter.
IV
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vii
Acknowledgments viii
1. Conclusions
S-4000 Sampler 1
F-3000 Flowmeter s 3
2. Recommendations
S-4000 Sampler 5
F-3000 Flowmeter 6
3. Description of Sampler and Flowmeter 7
4. Equipment Used and Method of Testing 10
5. Results of Performance Tests for the S-4000 Sampler
Timed Runs 14
Multiplexer - 16
Multiple Bottles Per Sample 18
Battery Operation and Endurance 20
Sample Preservation With Ice 25
Field Tests 25
Dependability 25
Sample Representativeness 29
6. Results of Performance Tests for the F-3000 Flowmeter
Tracking 34
Analog to Digital Conversion 36
Drift 37
Deadband 42
Overall Accuracy and Precision 43
Possible Theoretical Inaccuracy 43
7. Discussion 45
8. References 46
9• Appendix 47
-------
FIGURES
Number Page
1 S-4000 Sampler 7
2 Model F-3000 Flowmeter 9
3 Schematic Drawing Illustrating Operation of the Model
F-3000 Flowmeter 9
4 Typical "Set-up" for Testing in the Laboratory 11
5 Thermocouple Locations 11
6 Perintown Contact Stabilization Plant 12
7 Schematic Diagram Showing Method of Taking Isokinetic
and Manning Samples Simultaneously 13
8 Discrete Sample Temperature Versus Time (Chamber
Temperature at 35C) 26
9 Discrete Sample Temperature Versus Time (Chamber
Temperature at 35C) 27
10 Discrete Sample Temperature Versus Time (Chamber
Temperature at 21.5C) 28
11 Linearity of Analog to Digital Signal Conversion 38
12 Results of Source Voltage Change (Flowmeter at 100%) 40
13 Results of Source Voltage Change (Flowmeter at 50%) 41
VI
-------
TABLES
Number
1 Sampler Accuracy (Timed Runs) 15
2 Synchronization of Sampler 16
3 Timed Multiplexer Operation (Run 2) 17
4 Timed Multiplexer Operation and Battery Endurance (Run 26) 18
5 Multiple Bottles Per Sample 19
6 Multiple Bottles Per Sample 20
7 Timed Multiplexer Operation and Battery Endurance (Run 27) 21
8 Battery Endurance (Run 38) , 22
9 Battery Endurance (Run 44) 23
10 Sampler Iced and Battery Endurance 24
11 Dependability Tests at Perintown Influent 30
12 Dependability Tests at Perintown Effluent 31
13 Sample Representativeness at Perintown Influent 32
14 Sample Representativeness at Perintown Effluent 33
15 Head-Tracking Ability of the F-3000 Flowmeter at 2 and 35C 35
16 Tracking (Probe Descending and Ascending) 36
17 Tracking (Room Temperature) 36
18 Linearity of Analog to Digital Signal Conversion (Run 97) 37
19 Electronic Drift with Temperature Change 39
20 Deadband (Backlash) 42
21 Overall Accuracy and Precision for Circular Pipe 44
22 Overall Accuracy and Precision for 90°V-notch Weir 44
23 Overall Accuracy and Precision for Six-Inch Parshall Flume 44
VI1
-------
ACKNOWLEDGMENTS
The author thanks Manning Environmental Corporation for supplying
electronic schematics of the sampler and flowmeter and for their prompt
courteous reply to questions and problems that arose during the investi-
gation. Thanks to Dr. D. F. Bender for performing chemical analysis on
the field samples. My appreciation is extended to Mr. Anthony Clark,
Superintendent of Clermont County Sewers, for permitting us to test the
sampler at the Perintown Sewage Treatment Plant and to Mr. Ollie Cohorn,
Operator, for his assistance at the plant. Technical advice given by
Drs. R. N. Kinman and J. D. Eye, University of Cincinnati, and Mr, A. F.
Mentink, EPA, is greatly appreciated.
Vlll
-------
SECTION 1
CONCLUSIONS
S-4000 SAMPLER
Accuracy and Precision
The timing function was accurate and precise. Average time per sample
for a 24-hour run, set to collect one sample/hr, was 60.023 min/sample with
a standard deviation of 0.014 min for the 24 samples. In tests made before
the sampler was returned to the factory because of a malfunction, even better
accuracy was achieved.
Approximate volume settings of up to 500 ml/sample are possible by
adjusting the siphon tube in the measuring chamber. The quantity and preci-
sion of volumes collected were satisfactory, during this part of the
investigation, and the standard deviation for 24 bottles ranged from 0.69 to
2.25 ml over 13 runs.
One deficiency was that the sampler's timer started counting before the
main switch was turned on if the spout was stepped to the first bottle by
making the battery connection instead of using the bottle-advance button.
This impaired the time of the first cycle and the last sample was skipped.
Additional bottles at the end of the run were skipped for each cycle time
period that passed before the main switch was activated. When the sampler
was stepped to the number one position with the bottle-advance button,
operation was satisfactory.
These runs were made at temperatures of approximately 2, 20, and 35C,
and no significant difference in accuracy or precision due to temperature
change was noted.
Multiplexer
Multiplexer runs allow up to 5 aliquots to be placed in the same bottle.
During some of these runs the first aliquot for the first bottle was skipped,
but the rest of the bottles received the correct volume. Elapsed time for
the first and second aliquots of the first bottle was not always accurate,
but it was accurate and precise thereafter.
-------
Multiple Bottles
The sampler can be programmed to fill four bottles in succession during
each cycle. During some of the tests, the time of the first cycle was not
accurate; for example, a run programmed to cycle every 180 min, took the
first sample after 127.5 min. Quantity of sample and precision of volume
were satisfactory. i*
Battery Endurance
Tests showed that battery power was sufficient for the most severe run
that could be made. These tests included runs at 2, 20, and 35C. Light
weight of the battery/sampler combination is desirable. If batteries are
charged before each run and checked with a hydrometer there should be no
problem.
Sample Preservation with Ice
Sample preservation tests made within an environmental chamber at 22
and 35C showed that sample temperatures did not reach 4C, as recommended by
EPA.* Ice within the sampler did not last for 24 hr, but melted after 5 to
7 hr. Tests showed that better heat transfer from the samples was obtained
when the lower portion of the bottle is covered with ice water instead of
air. As ice melted, empty bottles floated and hit the stepping spout, and
the remaining bottles were left empty. "0" rings were not strong enough to
hold the bottles down, and it was necessary to tie the retainer with string.
Representative Sample
Tests were made at a sewage treatment plant influent, and there was no
significant difference between Manning and isokinetic samples for nonfilter-
able solids (NFS) and total organic carbon (TOC). Samples collected at the
effluent were not significantly different with regard to NFS. TOC samples
of the effluent were slightly different from isokinetic. These results were
analyzed statistically using the T-test at the 95 percent confidence level.
Reliability
The sampler performed satisfactorily when left unattended for 24-hour
periods at both sewage treatment plant influent and effluent points. One
discrepancy is that sample volume variation for the influent was higher
(standard deviation = 7.79 ml) than volume variation for the effluent (stan-
dard deviation =1.76 ml). There were also a few other times during the
laboratory investigation when one of the 24 samples was too small. Present
Manning samplers use a pressure sensing detector on the measuring chamber
instead of the resistive sensor that was tested and this may have corrected
this problem.
Miscellaneous Problems
During these tests both battery leads pulled apart at the clips where
the leads are fastened to the battery terminals.
-------
During chemical analysis of samples, two full plastic sample bottles fell
from the lab bench and broke as if made of glass. This was because they were
cold. Full warm bottles were dropped from a height of 3 ft and they did not
break. Bottle cap inserts came out when they got wet, however the bottles
did not leak when the inserts were left out.
CONCLUSIONS
F-5000 FLOWMETER
Tracking
No significant error in head readings from the liquid level dial were
detected at" temperatures of 2, 20, and 35C. The liquid level dial has 1/4-
in. and 0.5-cm graduations, and these should be changed to a least 0.1 in.
and 0.2 cm. Also the percent flow dial has numbers at 5-percent intervals
but no graduation marks. There should be marks at 1 percent intervals.
Markings on the percent flow chart should be at 5-percent intervals rather
than 10. The chart arm has an error with time of about 15 min per instan-
taneous full-scale swing.
Analog to Digital Conversion
Analog to digital conversion (signal from flow rate pot to final output
on counter) was linear at constant temperature, hence the electronic cir-
cuitry in this part of the instrument performed satisfactorily.
Drift
Electronic drift due to changing temperature from 2 to 35C averaged 1.1
percent.
No flowmeter drift was seen on the counter when the source voltage was
changed from 12.9 to 11.5. Output variation at 100 percent of maximum flow
was 1.02 to 1.053 cycles/min over the source voltage range of 12.9 to 11.5
volts.
Manning specifies an output of 1 cycle/min at maximum flow and this was
1 percent high.
Deadband
The most serious fault with the flowmeter was caused by deadband or back-
lash in the gearing. This error ranged from -11.63 to 6.14 percent of
reading at 6 in. head on a 90° V notch weir when the instrument was cali-
brated at 15 in. full scale.
-------
Overall Accuracy and Precision
Error ranged from ±11.6 percent of reading at low flows to 0-3 percent
at the calibration point (full scale). These tests showed that most of the
error and lack of precision was due to backlash in the gearing.
An additional source of error is incorporated if the instrument is used
with primary flow measuring devices that do not exactly follow H*>/2} H^/2
or, the Manning equation. For example, slight error will be incorporated
when the instrument is used with small parshall flumes. For permanent
installations a cam or forms of electronics, such as a functional amplifier
or microprocessor, that follow the exact equation for the primary device
should be required.
-------
SECTION 2
RECOMMENDATIONS
S-4000 SAMPLER
These tests showed that the timing function is accurate and precise.
Perfect accuracy was obtained at first and after the instrument was returned
to the company for a repair, it was off 1.4 sec/hr. This is satisfactory
for most sampling operations, and quality control should be maintained at
the factory to keep the timer accurate.
The sampler must be wired so that the timer cannot start before the
main switch is actuated. It should not make any difference whether the
spout is stepped to the first bottle with the advance button or by con-
necting the battery to the sampler. An additional button may be required
that synchronizes all circuits to zero at the start of a run.
There is a need to synchronize the start of sampling on both multi-
plexer and multiple-bottle sample runs. The first sample should contain
the correct number of aliquots, and it should start being collected at the
set time. It should be possible to set the sampler so that the first sample
is taken immediately with proper increments of time thereafter or the first
sample should start exactly after the set time increment has elapsed.
!
The device should be able to cool samples to 4C and maintain them for
24 hours at that temperature, therefore a larger ice space, clearance
between bottles for ice water to flow, and better insulation are required.
It is necessary to secure sample bottles better so that they do not float
as the ice melts. , '
Excessive variation in sample volume was noted during raw sewage col-
lection and a few other times during these tests. Precise volume is
important for composite samples and flow-proportional samples. This problem
may have been solved since Manning changed from a resistive to a pressure
sensing level detector.
Clips used to fasten battery leads to the terminals should be sturdier.
Cold plastic bottles should be more ductile so that they do not break
if accidentally dropped.
-------
F-3000 FLOWMETER
It is recommended that the liquid level dial graduations be made
smaller, 0.1 in. instead of 1/4 in. and 0.2 cm instead of 1/2 cm. The per-
cent dial should include graduations, and they should be 1 percent apart.
Markings on the percent flow chart should be changed from 10 to 5 percent,
and it is recommended that the chart be synchronized properly with time by
slightly relocating the pivot point for the pen arm.
Better quality control should be maintained at the factory so that the
counter cycles exactly once/min at full scale instead of 1.01 times/min.
Deadband or backlash due to play between the gears must be eliminated.
This could be accomplished by installing antibacklash gears that incorporate
springs or by using forms of electronics such as a functional amplifier or
microprocessor that have no gears and cams. The latter modification is
recommended.
Overall accuracy which is ±11.6 percent at low flows (most of which is
caused by backlash) must be improved. Instrumentation for detecting head
and converting head to flow should approach near perfect accuracy and preci-
sion. Fabrication and installation of the primary flow measuring device may
inherently incorporate a slight error, but the present state of the art in
electronics can assure that the instrument used for detecting head and
recording flow is accurate and precise.
Flow measuring equipment installed at permanent installations should
incorporate the exact equation for the primary device.
The flowmeter incorporates a desiccant. Mechanisms and electronics
should be designed to perform satisfactorily in all humidity conditions
without a desiccant. Desiccants in field equipment are easily forgotten
and therefore they are seldom recharged.
These recommendations are meant to be objective and hopefully investi-
gations of this nature will help improve wastewater samplers in general.
The S-4000 sampler and F-3000 flowmeter performed satisfactorily for the
most part and should be adequate for general sampling.
-------
SECTION 3
DESCRIPTION OF SAMPLER AND FLOWMETER
The model S-4000 sampler pictured in Figure 1 weighs 40 Ib (18.6 Kg.)
including a YUASA (model 12N12A-4A) 12-volt battery and tray of 24 sample
bottles (500 ml each).
MEASURING CHAMBER
^CONTROLS SOLENOID
COMPRESSOR VALVE
SPOUT
-SAMPLE
BOTTLES
STEPPING
MOTOR
a) Photograph of sampler
b) Schematic diagram.
Figure 1. S-4000 sampler (courtesy Manning Environmental Corp.).
Intake tubing is 3/8 in. ID reinforced tygon; it is 22 ft long and
terminates at a plexiglass measuring chamber. Sampling cycles are initiated
from either an internal timer or an external contact closure that originates
from a flowmeter. Twenty-four discrete bottles can be filled at intervals
of 15 min to 24 hr. Multiplexing allows each bottle to be composited of
from one to five samples. It is also possible to fill four consecutive
bottles one right after another, a capability that permits the addition of
different preservatives to the same sample. Controls are all solid state
electronics, and they incorporate a quartz crystal controlled oscillator and
digital logic to provide the sampling intervals. The sequence of events
during sampler operation is as follows (refer to Figure Ib). The controller
initiates a cycle; the solenoid valve is positioned so that the compressor
-------
clears the intake line for a few seconds; the solenoid valve changes
position and a sample is sucked into the measuring chamber; an electronic
level sensor within the measuring chamber sends a signal to the control that
changes the solenoid valve position so that excess sample is forced out of
the siphon; the pinch valve opens and the measured amount of sample is
forced into the sample bottle; the pinch valve closes and compressed air
clears the intake line; the compressor turns off, and the spout steps to the
next bottle. This completes one cycle.
Figure 2 dipicts the F-3000 flowmeter and Figure 3 is an electro-
mechanical schematic that illustrates its operation. The unit must be
installed upstream from a primary device, such as a weir or flume. It can
also be installed on a circular pipe if its diameter, slope, and roughness
factor are known. The dipper tracks the level of the liquid above the
primary device, and the voltage from the level pot is fed into a variable
gain amplifier that allows selection of maximum head over a four-to-one
range (Figure 3). Output of the variable gain amplifier controls a servo
system that causes a cam follower to rotate proportional to percent flow
rate, according to the characteristic curve of the cam being used. The
cam follower is mechanically coupled to a pen arm that records percent flow
rate, a dial that reads percent flow rate, and a potentiometer (flow rate
pot). The output of the flow rate pot is proportional to percent flow rate
and this signal is integrated and fed to a digital pulse circuit that con-
trols a totalizing mechanical counter. The counter adds one count for each
flow increment equivalent to maximum (100%) flow. In addition to the
counter, another digital circuit allows for the accumulation of a multiple
(switch selectable) number of maximum flow units. When this preset number
of maximum flow units is reached, a switch closes and this signal can be
used to start a cycle on the S-4000 sampler. The flowmeter can, therefore,
be used with the sampler to make it flow proportional on a constant-sample-
volume, variable-time basis.
Three cams are incorporated within the flowmeter, but each is used
singly and represents one characteristic curve. The desired cam is easily
rotated into position with a screwdriver. The standard set of cams includes
characteristic curves for: 1) V-notch weirs (H^/2); 2) flumes and rectangu-
lar weirs (H^/^); and 3) circular pipes.
-------
Figure 2. Model F-3000 flowmeter.
LIQUID
LEVEL
DIAL
PERCENT
FLOW
CONTROL
)POT
LEVEL
POT
VARIABLE
GAIN
AMP
L*.
SERVO
CONTROL
SYSTEM
SERVO MOTOR
TO SAMPLER
Figure 3.
Electromechanical schematic of dipper flowmeter (courtesy
Manning Environmental Corp.)-
-------
SECTION 4
EQUIPMENT USED AND METHOD OF TESTING
The following equipment was used to test the sampler and flowmeter:
1. Honeywell Electronik 15 recorder (span to 12 volts with supression).
2. Esterline Angus recorder (span to 100 volts).
3. Honeywell Electronik 16, 12-point thermocouple recorder.
4. Keithley (model 616) digital electrometer.
5. Universal Electronics regulated, variable, power supply (model C22-2).
6. Simpson (model 379) battery tester.
7. YUASA Syringe Hydrometer (model 404-14A).
8. EPCO (model 6130) water current meter.
9. Webber Manufacturing Company, Inc., environmental chamber.
10. Matheson thermometer, -1 to 51C, 1/10 division.
11. Starett 24" Vernier Height Gage.
12. Flotec Inc. (model F4P1-3100) Pump with Reliance 1/3 Hp., 1725 RPM
motor and Scrambler (model PM1) variable speed control).
13. BARCO Portable Master Meter, range 0-50 (meter no. BR-L0500-00-01).
14. Barco Venturi (model 1/2" - 402, VI, BR-12402-08-31).
15. Marshalltown pressure gage^O-30 psi.
16. Various graduates and flasks.
Testing took place in the laboratory at ambient conditions, within an
environmental chamber at temperatures of 2, 20, and 35C and in the field at
a small wastewater treatment plant located at Perintown, Ohio. A typical
laboratory "set-up" is pictured in Figure 4.
Sampler performance was tested initially without the flowmeter. The
HoneywelJ. recorder was connected to the sampler pump leads and used as an
event recorder to determine the accuracy and precision of the sampler's
timing function. Its ability to preserve samples under iced conditions was
determined by placing thermocouples in eight of the 24 bottles at locations
shown in Figure 5. Thermocouples were fastened to a small plexiglass insert
that kept them centrally located within the sample bottles. Temperatures
were traced on a Honeywell 12-point temperature recorder.
Field tests on the S-4000 sampler were made in Perintown, Ohio. Figure
6, a flow diagram of the perintown sewage treatment plant, shows the sampler
locations.
10
-------
S-4000 SAMPLER
91.4 cm.
WATER^
LEVEL
F-3000 FLOWMETER
SIGNAL
CABLE
ENVIRONMENTAL
CHAMBER
BATTERY
VERNIER
HT. GAGE
7
Al PLATE OR
WATER BEAKER
' J
a) METHOD OF TESTING SAMPLER b) METHOD OF TESTING FLOWMETER
FIGURE 4. Typical "set-up" for testing
in the laboratory.
22
21
BOTTLES
CONTAINING
THERMOCOUPLES
17
16
15
14
10
FIGURE 5. Thermocouple locations.
11
-------
BY
j
i
IN
PASS GRIT
, CHAMBERS
BAR \
I SCREEN
FLUENT
GATES
COMMINUTER PARSHALL
GATES GATES FLUMES
-e-
STABILIZATION
AERATOR
•/-
BAR SCREEN
DIGESTOR
CONTACT AERATOR
AERATOR 4 CLARIFIER
t
_L
SLUDGE
CHLORINATION
\
EFFLUENT
I
TO DRYING PITS
FIGURE 6. Perintown contact stabilization plant.
Manning sampler compared to isokinetic (locations B
and C), dependability tests (locations A and C).
Tests for dependability were made at locations A and C and tests for
sample representativeness were made at locations B and C. Sample repre-
sentativeness was tested by comparing the S-4000 sampler to isokinetic
samples. Isokinetic samples were obtained by the method illustrated in
Figure 7. The Manning intake was strapped to the isokinetic intake as
shown. Velocity of flow in the waste stream was measured initially with an
EPCO electromagnetic current meter. Velocity entering the isokinetic intake
was then set equal to that of the waste stream by adjusting the variable
speed pump, until pressure drop across the venturi gave the correct flow.
A graph shown in Figure 1 of the Appendix was used to convert velocity in
the 0.620 in. ID isokinetic intake to AP. Inlets of both intakes were
facing directly into the direction of flow. Sample velocity within the
Manning sampler was higher than that of the waste stream for both influent
and effluent samples; therefore, to obtain simultaneous samples, it was
necessary to wait for the proper residence time within the isokinetic unit
before collecting this sample. Waste stream velocities at the influent and
effluent were 1.7 and 0.7 feet/sec (fps), respectively. Isokinetic samples
were collected 20 and 50 sec for influent and effluent, respectively, after
12
-------
starting the Manning sampler. Comparison samples for representativeness
were analyzed for NFS and TOC.
The flowmeter was tested in the laboratory under static conditions, as
shown in Figure 4b. Head settings, accurate to .001 in., were made with a
vernier height gage. Either an aluminum plate or a small beaker of water
was attached to the slide of the height gage and connected to the negative
battery terminal of the flowmeter. Test voltages were read with a Keithley
electrometer, and an Esterline Angus recorder was used to record battery
voltage. Most of the flowmeter tests were made with the sampler connected
to the flowmeter, and the Honeywell recorder was connected to the sampler;
hence the Honeywell chart gave an accurate record of flowmeter counts versus
time.
VARIABLE
SPEED PUMP
\
GAGE
ISOKINE
INTAKE
WASTE
Cl O\*/ ""'
TIC
\STRAP
^^^ _j
*~^
— <
L-91
^
3/8"
LSPACER
.VENTURI VALVE
MANNING
SAMPLER*-
MANNING INTAKE
SAMPLE
BOTTLE
FIGURE 7. Schematic diagram showing
method of taking isokinetic and Manning
samples simultaneously.
13
-------
SECTION 5
RESULTS OF PERFORMANCE TESTS OF THE S-4000 SAMPLER
TIMED RUNS
Accuracy of the timing function and volume delivery was tested at
approximately 2, 20, and 35C. Table 1 summarizes this part of the original
data, which are included elsewhere in the body of the report or in the
Appendix. For example, Table 1 shows that run 1 was made with the sampler
and water bath at 21.7C. Average volume (X) of the 24 bottles collected was-
277.7 ml (standard deviation Sx = 0.97 ml), and the average time (t) between
bottles was 30-00 min. Sampler settings were 300 ml and 30 min for volume
and time, respectively. For this run, the average volume collected was
satisfactory, and volume variation was not excessive. Average time between
samples (30.00) was essentially perfect, hence the sampler has an accurate
timing device. Standard deviation for time between discrete samples was not
calculated because the variation was so small that it could not be deter-
mined on the strip chart. The accumulated error for the entire 720-min run
was only 0.025 min and hence insignificant.
Runs 21 and 22 in Table 1 show that the controller malfunctioned, as
indicated by failure to cycle at the correct times. The controller was
returned to the factory and runs 23 through 43 were made after the repair.
When the accumulated error for runs 1 and 2 is compared with that for runs
23 through 43, it is seen that the sampler was a little less accurate after
being repaired. For example, runs 1 and 32 were both set to cycle every 30
min, and the accumulated error for run 1 was 0.025 min versus 0.3 min for
run 32. The total error of 0.3 min in 12 hr is not too significant, however
the accumulated error of 0-025 min for run 1 shows that almost perfect
accuracy was possible with the Manning timing function.
Run 33 was made with the recorder moving at a higher chart speed (0.8
in./min), and it was possible to estimate the average time/cycle and the
standard deviation of discrete cycles. During this run the average time of
a cycle was 60.023 min, and the standard deviation was 0.014 min. The con-
plete 24-hr cycle took 0.56 min (33.6 sec) too long. This is satisfactory
for most sampling purposes, however it is possible to do even better, as
shown by run 1. Complete data for run 33 are shown in the Appendix.
In Table 1, runs 23 through 43 were made with the equipment operating
at temperatures of approximately 2, 21, and 35C. The accumulated error for
all of these runs is about the same (0.5 to 0.7 min for 24-hr runs and 0.3
min for 12-hr runs). Therefore, temperature change did not affect timer
performance.
14
-------
Table 1. SAMPLER ACCURACY (TIMED RUNS)'
Volume (ml)
Run
1
2
21
22
Temp.
(0)
21.7
21.6
24.3
22.3
set
300
400
400
400
X
277.7
424.8
371.2
375
sx
0.97
0.85
1.28
set
30
30V
30
15
Time (min . )
t
30.00
30.00
18.49
8.996
St.
t
t
.45
.19
accumulated
error
0.025
0.0
276.24
144.1
Sent controller back for repair
23
24
25
32
33t
34
35
36
41
42
43
23.3
22.8
22.8
2
2
21
22
31
35
35
21
400
400
400
400
400
400
400
400
425
425
425
379.5
380.5
382.9
387.5
384.6
X
380
383.4
414.6
413.8
414.1
0.88
0.83
0.99
1.10
0.72
X
0.69
0.85
1.47
2.25
1.05
60
60
60
30
60
60
60
60
60
60
60
60.025
60.026
60.026
30.013
60.023
60.02
60.03
60.02
60.03
60.03
60.03
t
t
T
t
.014
t
t
t
t
t
t
0.6
0.62
0.63
0.3
0.56
0.5
0.7
0.5
0.6
0.7
0.6
*Averages (X and t) and standard deviations (Sx And St) were calcula-
ted from 22, 23, or 24 samples.
tVariation was too small to read.
XBottles floated up and hit indexing arm, see run 34 in the Appendix.
VMultiplexer.
fRun made at faster chart speed (0.8 in./min); able to detect more
precise time between cycles.
Table 2, run 28, shows that the last sample was skipped and that time
for the first cycle was inaccurate. Additional data (run 28A) show that
this happened when the indexing spout was stepped to the first position by
connecting the battery instead of pushing the bottle-advance button. The
sampler was turned on 5 min after the battery connection was made, and this
is the reason that the first sample was collected after approximately 10 min
instead of 15. Hence, the battery connection will index the spout to the
first bottle and start the digital timer, but samples are not collected
until the switch is turned on. Run 28B was started by stepping the spout to
the first bottle with the bottle-advance button, and this run was satisfac-
tory. The sampler should be wired so that nothing starts until the switch
is turned on, and all circuits should also reset to zero at this time. It
may be necessary to incorporate a button to "reset" before activating the
main switch.
15
-------
Table 2. SYNCHRONIZATION OF SAMPLER
Sample
1
2
*
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run
Cycle
time
(min)
13.7
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
28*
Volume
(ml)
83
84
83
83
82
83
82
83
83
83
83
83
83
84
83
83
81
83
83
83
82
82
83
skipped
Run
Cycle
time
(min)
9.8
15
15
15.05
15
15
15
15
15.05
15
15
15
15.05
15
15
15
15.05
15
15
15
15
15
15
28At
Volume
(ml)
345
347
348
347
349
350
350
349
349
348
347
347
347
348
346
261 «-
348
347
247
247
349
349
348
skipped
Run '.
Cycle
time
(min)
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
zset
Volume
(ml)
404
404
404
404
405
405
404
404
404
405
405
403
403
404
404
404
388 «-
402
404
403
405
404
405
404
*Stepped to
bottle #1 with
battery con-
nection.
tStepped to bottle
#1 with battery
connection; waited
5 min before
turning on switch.
flndexed through
24 bottles and to
bottle #1 with
bottle advance
button.
MULTIPLEXER
Tables 3 and 4 give the results of multiplexer runs 2 and 26, which
were made up of 2 and 4 aliquots per bottle, respectively. It is seen that
in run 2 only one aliquot was taken during the first cycle and that the
sampler then went on to the next bottle. For run 26, the second aliquot
was taken after 9.9 min and the third after 15.3 min. Only three aliquots
were taken into the first bottle instead of four, and after the first cycle
the run was 19.8 min ahead of schedule. Other than these discrepancies,
the remainder of the multiplexer run was satisfactory.
16
-------
Table 3. TIMED MULTIPLEXER OPERATION (RUN 2}'
Time (min)t
Sample
if
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Aliquot 1
15.0
15.0
15.0
15.0
15.0
14.9
14.9
15.0
15.0
14.9
15.0
15.0
15.0
15.0
15.0
15.1
15.0
15.0
15.0
15.0
15.0
15.0
15.0
Aliquot 2
15.0
15-0
15.0
15.0
15.0
15.0
15.1
15.1
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
15.0
Sample total
15.0
30.0
30.0
30.0
30.0
30.0
30.0 <
30.0
30.0
30.0
29.9
30.0
30.0
30.0
30-0
30.0
30.1
30.0
30.0
30.0
30.0
30.0
30.0
30.0
Volume (mi;
215
425
425
423
425
424
427
425
425
425
425
425
425
425
423
423
425
425
425
425
425
425
425
425
X = 424.8
= 0.85
*Sampler programmed to collect two 230-ml aliquots/bottle at 15 min
intervals. Room temperature, 21.6C; sample water temperature, 18.2C.
tAll cycle times were very close to 15 min. Readings of 14.9, 15.1,
29.9, and 30.1 are not exact, but mean slightly low and slightly
high.
fFirst bottle contained only one aliquot.
17
-------
Table 4. TIMED MULTIPLEXER OPERATION AND BATTERY
ENDURANCE (RUN 26)*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Aliquot
1
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Time
Aliquot
2
9.9
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
(min)
Aliquot
3
15.3
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
4
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Sample
total
40.2
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
Volume
(ml)
243t
324
323
323
324
321
323
322
321
320
321
323
322
324
323
322
322
321
325
323
327
324
324
325
*Sample programmed to collect four 100-ml aliquots/bottle at 15-min
intervals. Slight variations in the intervals were not detectable
because recorder chart speed was too slow (0.1 in./min). Entire
run took 39 sec too long. Room temperature, 23.4G; sample water
temperature, 21.8C. Initial charge by Simpson meter on 15 V full
scale, 85%. Final charge by Simpson meter on 15 V full scale,
79%.
tFirst bottle contained only three aliquots and time between aliquots
for this bottle was in error.
MULTIPLE BOTTLES PER SAMPLE
Runs 9, 17, 19, and 20 (Tables 5 and 6) were set up to take four bot-
tles, one right after the other, and then wait for a timed period before
taking the next sample. This function is useful if more than one preserv-
ative is required. In runs 9 and 17, the timing was correct, and the
average sample size and standard deviation were satisfactory. The time of
the first sample was not correct for runs 19 and 20. This is related to
18
-------
the same problem as encountered with the multiplexer. A reset button that
would synchronize the start may be required and proper timing would then
occur after activating the main switch.
Table 5. MULTIPLE BOTTLES PER SAMPLE*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run 9t
Cycle time Volume
(min) (ml)
376
378
379
379
15 380
380
380
380
15 380
380
378
379
15 379
380
378
378
15 378
378
378
379
14.9 379
379
379
379
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run 17
Cycle time
(min)
15
15
15
15
,
15
15
Volume
(ml)
381
382
382
383
383
382
382
382
382
382
382
382
381
381
382
380
382
382
380
382
382
382
381
380
X = 378.88
SY = .992
A
X = 381.67
5V = .816
*Sampler programmed to put sample into four consecutive bottles
every 15 min. Sample volume set at 400 ml.
tRoom temperature, 22.4C; sample water temperature, 20.3C.
19
-------
Table 6. MULTIPLE BOTTLES PER SAMPLE'
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run 19f
Cycle time
(min)
127.6
180
180
180
180
180
Volume
(ml)
381
380
380
380
380
380
380
381
380
379
380
380
x 380
378
379
378
378
378
377
378
379
378
378
377
Sample
no.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run 20*
Cycle time
(min)
127.5
180
180
180
180
180
Volume
(ml)
379
379
378
378
379
377
379
380
380
380
379
380 .
379
378 -
377
379
378
378
378
377
377
378
380
378
X = 379.13 \ = 378.54
sx = 1-191 Sx = 1.021
tRoom temperature, 24C fRoom temperature, 22.3C
Sample water temperature, 19.1C. Sample water temperature, 19.1C.
*Sampler programmed to sample into four consecutive bottles every
180 min. Sample volume set at 400 ml.
BATTERY OPERATION AND ENDURANCE
Tables 4 and 7 show the results of runs 26 and 27, which were made at
room temperature (approximately 23.5C). The sampler was programmed to take
four aliquots into each bottle at 15-min intervals. The fact that the
first aliquot was skipped in runs 26 and 27, as mentioned earlier, was a
problem of reset and was not the fault of the battery. These runs were
started with a new YUASA model 12N12A-4A, 12-volt battery that was charged
for over 20 hr with YUASA's 500 ma charger. Gassing was seen in the bat-
tery electrolyte before the charger was removed. Therefore the battery was
assumed to be fully charged. Initial charge was read at 85 percent on the
15-volt scale (12.75 volts) of a Simpson Model 379 battery tester.
20
-------
Table 7. TIMED MULTIPLEXER OPERATION AND BATTERY ENDURANCE
(RUN 27)*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Aliquot
1
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
2
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15X
Time (min)
Aliquot
3
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
4
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Sample
total
45
60
60
60
60
60
60
60
60
60
60
60
60
60t
60
60
Volume
(ml)
245
333
329
330
330
330
330
331
332
333
336
333
339
soot
*Sampler programmed to collect four 100-ml aliquots/bottle at 15-min
intervals. Slight variations in the intervals were not detectable
because recorder chart speed was too slow (0.1 in./min). Timing
error was a little more than 1 sec/bottle. Room temperature, 23.7;
sample water temperature, 22.4C. Initial charge by Simpson meter
on 15 V full scale, 79%. Final charge by Simpson meter 15 V full
scale, 18%.
tFailed to step to next bottle.
fOverflowing.
XTimer continued to function accurately to this point.
Run 26 was completed and the battery failed while bottle 14 of run 27
was receiving a sample. When battery power is low the unit fails to index,
therefore this bottle overflowed, but the timer continued to function for
a few more cycles. Final charge, as read by the Simpson meter, was 18
percent of 15 volts or 2.7 volts. The total number of accurate samples
collected was 146. Since the maximum setting on the sampler is five sam-
ples/bottle for 24 bottles or 120 samples, a freshly charged battery has
enough power for one run of the most severe type at normal temperatures.
21
-------
Manning Corporation noted that its tests averaged 170 samples at room
temperature and 140 samples at 0.5C. Run 38 (Table 8) shows that battery
power was sufficient to complete a similar type run at 31.5C. Initial and
final battery voltage (as read by the Simpson Tester) was 85 percent and
79 percent for 12.75 and 11.85 volts, respectively. Low temperature is
usually the most severe condition for batteries, and run 44 (Table 9) was
made at 2C. Initial and final charge and specific gravity of battery
electrolyte are included in Table 9. The battery's power was satisfactory
during this run, but the specific gravity of its electrolyte at the end of '
this run was only 1.15 to 1.16. It would be undesirable to run the sampler
much longer than this because voltage would start to drop off rapidly.
Table 8. BATTERY ENDURANCE (RUN 38)*
Time (min)
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Aliquot
1
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
2
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
3
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
4
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Sample
total
45
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
Volume
(ml)
315
422
422
421
422
421
422
423
423
425
424
424
424
423
421
422
424
425
424
423
424
428
425
425
*Sampler programmed to collect four 125-ml aliquots/bottle at 15 min
intervals. Slight variations in the intervals were not detectable
because recorder chart speed was too slow (0.1 in./min). Entire
run took 0-5 min too long. Chamber temperature, 31.5C. Started
run with new battery. Initial charge by Simpson meter on 15 V full
scale, 85%; final charge by Simpson meter on 15 V full scale, 79%.
22
-------
Table 9. BATTERY ENDURANCE (RUN 44)*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Aliquot
1
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
2
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Time (min)
Aliquot
3
15
15
15
15
15
15
15
15
15
15 '
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Aliquot
4
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
Sample
total
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
Volume
(ml)
486
485
485
485
485
488
485
486
485
486
487
486
485
487
487
487
484
487
489
488
488
478
489
488
*Sampler programmed to collect four aliquots/bottle at 15-min
intervals. Slight variations in the intervals were not detect-
able because recorder chart speed was too slow (0.1 in./min).
Entire run took about 0.5 min too long. Chamber temperature,
2C. Initial charge by Simpson meter on 15 V full scale, 86%;
final charge by Simpson meter on 15 V full scale, 78%.
Specific gravity of electrolyte in cells:
Start run 1.29 1.27 1.27 1.27 1.30 1.29
End run -- 1.16 1.16 1.16 1.15
Level of electrolyte in first and last cells was too low to
draw sample.
The specific gravities taken after runs 41, 42, and 43 (Table 10) show
that the battery has more than enough power for the usual type of run when
multiplexing is not used.
23
-------
Table 10. SAMPLES ICED AND BATTERY ENDURANCE*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Run
Cycle
time
(min)
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
X =
sx =
41
Volume
(ml)
410
415
415
416
415
410
415
415
415
415
415
415
415*
414
415
415
415
415
415
415
415
415
415
416
414.6
1.469
Run
Cycle
time
(rain)
60
60
60
60
60
60
60
60
£0
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
X =
sx =
42
Volume
(ml)
413
413
413
415
414
415
414
415
415
415
415
415
414
404
414
414
415
415
414
412
415
414
414
413
413
2.25
Run
Cycle
time
(min)
60
60
60
60
60
60
60
60
60
60
60
60
60
61
60
60
60
60
60
60
60
60
60
60
X
sx
43
Volume
(ml)
412
415
415
414
415
415
414
414
415
415
415
415
415
413
414
413
412
414
413
413
413
415
t
t
= 414.05
1.046
*Slight variation in cycle time was not detectable because recorder
chart speed was too slow (0.1 in./min). Each run took approximately
0.6 min too long. Temperatures for runs 41, 42, and 43 were 35, 35,
and 20C, respectively.
tTurned off intentionally.
Specific gravity of battery electrolyte:
Start End Start End Start End
1.26 1.20 1.27 1.21 1.25
1.26 1.21 1.27 1.21 No 1.25
1.26 1.21 1.27 1.22 data 1.25
1.26 1.15 1.27 1.23 1.21
1.26 1.21 1.27 1.22 1.25
1.26 1.23 1.27 1.21 1.25
24
-------
In summary, battery power was sufficient for the most severe run that
could be made with this sampler. If batteries are charged before each run
and checked with a hydrometer, there should be no problem.
SAMPLE PRESERVATION WITH ICE
Manning's sample bottle tub has space for ice in the center, and the
outer surface is made of an insulating material. The purpose of these runs
was to determine the temperature of the iced samples and the length of time
that they remain cold. EPA methods1 require a preservation temperature of
4C for some parameters. The tests were run as mentioned in Section IV of
the report. Thermocouples were located in bottles 1, 2, 7, 8, 13, 14, 19,
and 20, as shown in Figure 5. Figures 8 and 9 are plots of sample tempera-
ture versus time for 35C environmental chamber runs. Figure 8 shows the
results of a run in which the center of the tub was packed with ice and
cold water was poured over the ice so that the lower portion of the sample
bottles were immersed in ice water. Figure 9 gives the results of a run in
which only ice was packed into the center compartment. Figure 8 shows that
the temperature of samples 1 and 2 dropped rapidly after the sample was
taken and reached a minimum point (IOC) after 5 hr. Figure 9 shows that it
took 7 hr for samples 1 and 2 to reach a minimum point of only 14.5C. In
summary, these graphs show that it is best to have the bottles partially
submerged in ice water for better heat transfer. A larger compartment with
more ice is also required if the sample is to reach 4C and remain there for
24 hr. Better insulation may also be needed.
Figure 10 depicts the results of an environmental chamber run at 21.SC.
This run was made with the center compartment filled with ice and cold
water covered the lower portion of the sample bottles. The temperature of
samples 1 and 2 dropped to 9.2C and then rose as the ice melted. The first
dip in the curve (i.e., sample 1 going from 14.2 to 14.5) show that the
temperature of the sample rose slightly as the bottle next to it was filled.
Bottles within the sampler are shaped to fit tightly together as shown in
Figure la. Better heat transfer would be obtained if a little space were
left between the bottles to accommodate ice water.
FIELD TESTS
Dependability
The sampler was taken to a small wastewater treatment plant at Perin-
town, Ohio, and tested for ability to collect a representative sample and
ability to run dependably and without clogging for a 24-hr period. Figure
6, a flow diagram of the plant, shows sampler locations.
Dependability investigations were made by running the sampler for 24-
hr periods at locations A and C. Controls were set to collect samples at
1-hr intervals, and the center compartment was packed with ice. Freshly
charged batteries were used during both tests, the results of which are
shown in Tables 11 and 12. The sampler ran throughout the 24-hr period,
samples were collected at 60-min intervals, and no samples were missed.
Table 12 shows that the volume variation for effluent samples was not
25
-------
to
en
SAMPLE NO. 1
i i i i i i
2 34 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
ELAPSED TIME (HR)
FIGURE 8. Discrete sample temperature versus time (chamber
temperature at 35C) additional water poured over ice
so that lower part of sample bottles were in ice water.
-------
N>
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
ELAPSED TIME (HR)
FIGURE 9. Discrete sample temperature versus time (chamber
temperature at 35C) no additional water poured
over ice or around bottles.
-------
N>
00
23
22J- SAMPLE NO. 1
NO. 7 NO. 13
**- NO. 8 T
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
ELAPSED TIME (HR)
FIGURE 10. Discrete sample temperature versus time (chamber
temperature at 21.5C) additional water poured
over ice so that lower part of sample bottles were
in ice water
-------
significant (S^ = 1.76 ml, range 405 to 410 ml). Volume variation for the
influent samples (Table 11) was more significant (Sx = 7.79, range = 361
to 408 ml). The most variation occurred on the first two samples, and the
volumes of the third thraugh 24th bottles were more precise. -The quantity
of sample collected in all bottles was satisfactory for analysis. Precise
volumes are important for composite and flow-proportional samples.
All ice was melted in the center of the sampler before the runs were
completed, and the final temperature in both runs was about 12.5C. Maxi-
mum ambient temperature for both runs was only 27.5C, therefore better
cooling of samples is required.
Sample Representativeness
These tests were made as described in Section IV. Figure 7 shows the
method of collecting simultaneous samples for comparison of Manning to
isokinetic. Comparison samples were analyzed for NFS and TOC. Influent
and effluent results are given in Tables 13 and 14. A statistical T-test^
was used to determine if there was a difference between isokinetic and
Manning samples. Influent samples for NFS and TOC showed no significant
difference at the 95 percent confidence level. Effluent samples did show
a difference for TOC at the 95 percent confidence level. Observing these
data in a rational manner and considering variations for NFS that have
been detected during other studies leads to the conclusion that there
was no serious difference between isokinetic and Manning for these tests.
29
-------
Table 11. DEPENDABILITY TESTS AT PERINTOWN INFLUENT
Date
1975
11/3
11/4
Approx .
time
10:OOA
llrOOA
12:OON
1:OOP
2: OOP
3: OOP
4: OOP
5: OOP
6 : OOP
7:OOP
8:OOP
9:OOP
10:OOP
11:OOP
12:OOM
1:OOA
2:OOA
3:OOA
4:OOA
5:OOA
6:OOA
7:OOA
8:OOA
9:OOA
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Volume
(ml)
361
408
397
397
398
400
395
397
398
397
393
395
395
395
395
395
393
396
395
395
395
395
393
394
Temperature (C) Remarks
air
16
22
25
23.5
23
22.5
20.5
15.5
influent
sewage
18 •*- Manual cycle. Initially,
20 ice was placed inside
20 sampler with about one
19 . 5 cup of water .
19.5 Lift at the influent
20 was about 1 ft.
20
17
X = 394.7
SX = 7.79
X = 396.1 First sample omitted
Sx = 3-11 from calculation.
X = 395.6 First and second samples
SX = 1-69 omitted from calculation,
The following temperatures were taken when the sample was collected at
9:OOA on 11/4/75.
Sample no. Temperature C
1 12.5
16 12.5
24 15.5
Cold water in center of sampler 13 No ice was remaining.
30
-------
Table 12. DEPENDABILITY TESTS AT PERINTOWN EFFLUENT
Date
1975
11/5
11/6
Approx.
time
8:48A
9:48A
10:48A
11:48A
12:48P
1:48P
2:48P
3:48P
4:48P
5:48P
6:48P
7:48P
8:48P
9:48P
10:48P
11:48P
12:48A
1:48A
2:48A
3:48A
4:48A
5:48A
6:48A
7:48A
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Volume
(ml)
405
408
408
410
408
410
410
408
407
406
407
405
408
407
405
405
405
405
405
405
405
405
407
406
Temperature (C) Remarks
air
14.5
20
23
25.5
27.5
27.5
25
23
10
effluent
sewage
17 -e- Manual cycle. Initially,
18 ice was placed inside
19 sampler with about one
19 cup of water.
19 Lift at the effluent
19 was about 6 ft.
19
19
11.5
X = 406.7
Sx = 1.76
The following temperatures were taken when the sample was collected at
9:OOA on 11/6/75."
Sample no. Temperature C
1 12
6 12
12
18
24
Cold water in center of sampler
12
12
12.5
12.5
No ice was remaining
31
-------
Table 15. SAMPLE REPRESENTATIVENESS AT PERINTOWN INFLUENT
Sample Total nonfilterable solids
(mg/1)
Isokinetic Manning Diff(d)
1 89.4 96.3 -6.9
2 147.0 138.0 9.0
3 162.2 143.0 19.2
4 170.9 151.7 19.2
5 174.7 152.5 22.2
6 154.1 158.8 -4.7
7 203.4 199.9 3.5
8 195.6 202.3 -6.7
9 185.9 191.0 -5.1
10 151.9 153.5 -1.6
11 141.1 142.5 -1.4
12 147.4 150.0 -2.6
d = 3.675
S, =10.926
d
H0: yd = 0
HI: yd + 0
a = 0.05
Critical region: T<-2.201, T>2.201
where T = (d - 0)/S2.201
where T = (d - 0)/Sd//12
T = 11 degrees of freedom
rnmniitit-i nn • T - .-.?.'/„.
V-iUlllpULu. LJ-UIl . 1 — -I n ^>
lo . /.
T = 0.234
Conclusion: Accept Ho and con-
clude that the Manning method of
sample collection for TOC at the
influent is not significantly
different from isokinetic.
32
-------
Table 14. SAMPLE REPRESENTATIVENESS AT PERINTOWN EFFLUENT
Sample Total nonfilterable solids
no. (mg/1)
Isokinetic Manning Diff(d)
13 4.9 4.2 0.7
14 4.7 4.3 0.4
15 5.0 4.6 0.4
16 4.5 4.8 -0.3
17 6.3 6.0 0.3
18 6.6
19 5.3 5.9 -0.4
20 5.3 5.3 0.0
21 5.4 5.0 0.4
22 4.6 5.7 -1.1
23 6.1 5.5 0.6
24 5.1 5.6 -0.5
d = 0.045
Sd = 0.557
Ho: ^d = °
HI: Vd 4 0
a = 0.05
Critical region: T<-2.228, T>2.228
where T - (d - 0)/Sd/ /IT and
Y = 10 degrees of freedom
r <- <-• T 0.45 y/TT
computations . i — Efcv
. 3D /
T = 0.268
Conclusion: Accept Ho and con-
clude that the Manning method of
sample collection for NFS at the
effluent is not significantly
different from isokinetic.
Total organic carbon
fme/n
Isokinetic Manning Diff(d)
81.7 75.6 6.1
83.8 81.4 2.4
84.7 82.3 2.4
83.8 85.9 -2.1
85.6 81.4 4.2
84.4
85.3 82.0 3.3
85.0 83.8 1.2
81.2 79.1 2.1
85.0 82.9 2.1
83.2 84.7 -1.5
85.0 84.4 0.6
d = 1.89
Sd = 2.35
V Vd = 0
HI: Vd+'°
a = 0.05
Critical region: T<-2.228, T>2.228
where T = (d - 0)/Sd/ /lT and
1^ = 10 degrees of freedom
Pnnrnul" 1 1~ i nn ^ * T — ^
L . JO
T = 2.667
Conclusion: Reject Ho and con-
clude that the Manning method of
sample collection for TOC at the
effluent is different from
isokinetic.
33
-------
SECTION 6
RESULTS OF PERFORMANCE TESTS OF THE F-3000 FLOWMETER
TRACKING
The ability of the liquid level dial to track head accurately and
precisely is important because this dial is used in calibration. Tracking
tests were made with the aid of a vernier height gage and either a small
beaker of water or an aluminum plate attached to the slide of the height
gage (Figure 4b). The liquid level dial was read as accurately as possible
but, since graduations were marked only every 1/4 in. and 1/2 cm, it was
impossible to read the dial as precisely as the vernier (0.001 in.). Table
15 gives the results of static tests made within an environmental chamber
at 2 and 35C. The liquid level dial followed changes in the vernier height
gage satisfactorily, and no significant difference due to temperature
change was detected. Readings shown in Table 15, such as 17.95 for the
level dial versus 18 for the height gage during run 72, mean that the level
dial was just a little low, since the 1/4 in. graduations made it impossi-
ble to read as close as 17.95. Graduations on the level dial of 0.1 in.
and 0.2 cm would be more acceptable. Table 15 also shows that the percent
flow dial was always a little higher than the chart, and better agreement
is required. Also some backlash or deadband is indicated in Table 15 for
the percent flow dial and this will be explained later.
Table 16 shows a test in which the dipper was made to follow an ascend-
ing and decending height gage, and satisfactory agreement was obtained.
Table 17 illustrates another tracking test and no serious difference
was detected.
In summary, tracking is satisfactory, but graduations on the "level
dial" should be at least 0.1 in. and 0.2 cm. Also the percent flow dial
has numbers every 5 percent but no graduations. Graduations should be
included for every 1 percent of flow. Graduations on the percent flow
chart should be at least every 5 percent instead of every 10 percent. Also
the time markings on the chart paper are not properly synchronized with the
pen arm (error is about 15 min for an instantaneous full scale swing on the
chart). There is an adjustment for length of the pen arm, but this is not
adequate; the pivot point needs to be relocated slightly.
34
-------
Table 15. HEAD-TRACKING ABILITY OF THE F-3000 FLOWMETER*
AT 2 AND 35C
Vernier
Run Temperature height gage
(C) (in.)
72 2 5
9
12
15
18
21
23
5
73 35 5
9
12
15
18
21
23
5
23
5
74 2 5
9
12
15
18
21
23
5
Liquid
level dial
(in.)
5
9
12
15
17.95
21
23.05
5
5
9
12.05
15.03
18
21.05
23.05
5.05
23.08
5.05
5.05
9.05
12.05
15.05
18
21.05
23.08
5.05
Flow dial
C*)
2
7.5
15
27
44
67
82
2
2
7.5
13
25
42
66.5
--
2.5
86
2.5
2.5
5.5
13
27.5
41
67.5
83
2.5
Chart
(%)
0
6
13
24
41
63
80
1
0
6
12
23
41
65
--
1
82.5
1
1
5
12
25
39
64
82
1
Calibrated with flow dial at 100% and liquid level dial at 24 in.
(V-notch weir).
Vernier height gage with dipper striking metal plate.
Liquid level dial initially set at 5 on run 72 and not reset
throughout runs 72, 73, and 74.
35
-------
Table 16. TRACKING (PROBE DESCENDING AND ASCENDING*)
Probe ascending
Vernier Level dial
(in.) (in.)
4 4
7 7
13 13
19 19
Probe descending
Vernier Level dial
(in.) (in.)
4 4
7 6.95
13 13
Probe ascending
Vernier Level dial
(in.) (in.)
4 4
7 6.95
13 13.03
19 19
*Calibrated with flow dial at 100% and liquid level dial at 9 in.
(flume). Vernier height gage with dipper striking water beaker.
Liquid level dial initially set at 4 in. and not reset throughout
run.
Table 17. TRACKING (ROOM TEMPERATURE):
Vernier height gage Liquid level dial
(in.) (in.)
4 4
7 7
13 13.05
19 19.05
Calibrated with flow dial at 100% and liquid
level dial at 9 in. (round pipe). Vernier
height gage with dipper striking water beaker.
Liquid level dial initially set at 4 in. and
not reset throughout run.
ANALOG TO DIGITAL CONVERSION
The analog signal from the flow rate pot (Figure 3) is integrated and
fed to a digital pulse circuit that controls the totalizing counter. The
object of this test was to determine if the output frequency at the counter
was linearly proportional to the voltage signal from the wiper of the flow
rate pot. Results are given in Table 18 and plotted in Figure 11; the
latter shows that analog to digital conversion was linear. These tests
were made with the flowmeter connected to the water sampler, and the sam-
pler was connected to a recorder so that the chart gave a record of sampler
cycles versus time (hence flowmeter counts versus time).
36
-------
Table 18. LINEARITY OF ANALOG TO DIGITAL
SIGNAL CONVERSION (RUN 97)
Date Time Test
10/6/75 3:55 A
B
C
D
E
F
G
H
I
J
10/6/75 5:35 K
10/7/75 10:30A L
M
N
0
Battery
voltage
12.565
12.58
12.565
12.565
12.56
12.56
R101 Cen-
ter tap
(volts)
.291
.341
.454
.656
.812
1.093
1.365
1.733
2.15
2.63
3.10
3.09
1.76
.841
.342
Cycles
(min) Probe
.086 u
.091
.143
.197
.25
.339
.435
.552
.696
.860 i
1.013 u
p
P
1.013 down
.567 {
.260 ')
.093 down
DRIFT
Drift Caused by Temperature Change
Tests for electronic drift caused by temperature change were made in
groups of three within an environmental chamber whose temperature was set
at 2, 35, and 2C. During the runs the flowmeter was connected to the
sampler, which was connected to a recorder; the strip chart from this
recorder gave cycles/min. All runs were made with the flowmeter on cali-
brate at 100 percent or in the operate position with the dipper set to
100 percent. Table 19 shows that the average variation in output due to
changing temperature from 2 to 35C was 0.011 cycle/min or 1.1 percent;
hence flow readings would be slightly higher at the cold temperature.
Manning states that the output should be 1 cycle/min at the 100 percent
setting, but the table shows that the output is slightly higher.
37
-------
1.0
0.9
0.8
07
'i
\
w
0.6
05
O
0.4
0.3
0.2
0.1
• - dipper rising
X- dipper lowering
1 2 3
INPUT AT FLOWRATE POT (VOLTS)
FIGURE 11. Linearity of analog to digital signal conversion
38
-------
Table 19. ELECTRONIC DRIFT
Run
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
Temperature
(C)
2
35
2
2
35
2
2
35
2
2
35
2
2
35
2
Cycles
(min)
1.016
1.01
1.018
1.018
1.01
1.022
1.022
1.009
1.017
1.019
1.007
1.020
1.032
1.023
1.031
Average
WITH TEMPERATURE CHANGE
Range
(cy/min)
.008
.012
.013
.013
.009
.011
Type of Operation
Circular pipe
dipper at
maximum flow
Circular pipe
flowmeter on
calibration
V-notch
flowmeter on
calibration
Parshall flume
flowmeter on
calibration
Parshall flume
dipper at
maximum flow
Drift Caused by Battery Voltage Decay
A variable DC power supply was used to simulate battery decay from 12.9
to 11.5 volts. Results of these tests are plotted in Figures 12 and 13.
Battery decay is represented as source voltage on the abscissa. Pin 10 to
ground is the wiper of the flow rate pot (input to operational amplifier),
and test point 5 to ground is the output of the operational amplifier. As
mentioned earlier, this signal is integrated converted to a digital signal
and displayed on a counter as total flow (represented on the ordinate as
output cycles/min). Figures 12 and 13 show that the analog input at the
flow rate pot (pin 10 and Test Point 5) drifted linearly downward as the
source voltage decayed, however the output signal did not drift although
some variability is shown in the graph. Figure 12 shows that variability
of the output signal was greater at 100 percent of flow than at 50 percent
flow (Figure 13). Although the output did not drift with source voltage,
the output variation of 1.02 to 1.053 cycles/min (Figure 12) at the 100
percent point should be improved.
39
-------
3.2
O
> 3.1
a
z
3
O
ee
O
°3.0
2.9
2.8
2.7
6.4
6.3
6.2
o
>6.0
0
§5.9
at
O
25.8
5.7
5.6
5.5
5.4
5.3
5.2
1.06
1.05
\
/TV
- .£ 1.04
X
L- 5 1.03
Q.
I—
- O 1.02
1.01
h i.oo
I
TEST PT. 5
TO GROUND
-X
h
PIN 10
TO GROUND
OUTPUT
I
I
I
11.5 11.6 11.7 11.8 11.9 12.0 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8 12.9
SOURCE VOLTAGE
FIGURE 12. Results of source voltage change (flowmeter on calibrate at 100%)
-------
6
8
o
z
1.4
1.31_
r-Q 32
-z
1"
h5«
i
g 2.9
2.8
27
1 < 1 1 1 1 1 1 1
1 I I 1
OUTPUT
TEST *O!NT 5
TO GROUND
PIN 10 TO GROUND
1
j_
I
I
11.5 11A 117 11.8 11.9 12.0 12.1 125 12.3 12.4 125 12A 127 12.8 12.9
SOURCE VOLTAGE
FIGURE 13. Results of source voltage change (flowmeter on calibrate at 50%)
-------
DEADBAND (BACKLASH)
These tests were made with the flowmeter calibrated at 15 in. full
scale on the h^' cam. An aluminum plate was set on a stand below the
dipper and connected to the negative battery terminal. When the dipper
was touching the plate, the liquid level dial was set to 6 in. Readings
were taken after the dipper was lowered to touch the plate, raised to
touch the plate, and finally lowered to touch the plate. Results of
these tests are given in Table 20. Run 94, for example, was made after
the dipper descended to the 6-in. level. The center tap of the flow rate
pot gave 0.391 volts. Percent flow was 9 percent on the flowmeter dial
and 7.5 percent on the chart. Flow at 6-in head for a 90° V-notch weir,
according to Manning's manual, is 200 gal/min (gpm). Total flow after
operation for 155.6 min is 31.125 gal and 10.29 percent of maximum flow.
The instrument's integrator, as read from the counter after multiplying
by maximum flow (1943.3 gpm), gave 33,036.1 gal; therefore, the resulting
error was 6.14 percent of reading or 0.6 percent of full scale. This
same procedure was followed with the dipper ascending and gave an error
of -11.63 percent of reading. Readings were taken again after the elec-
trode had descended, and the error was 5.5 percent of reading. This
error was due to deadband or backlash in the mechanical gearing of the
instrument (gearing is illustrated in Figure 3). The error is most
obvious from voltage readings of the flow rate pot (0.391, 0.337, 0.389).
Voltage should be the same at 6 in. head, regardless of whether the
dipper descends or ascends to reach its destination. Error of this type
was most significant at lower flows, as shown in Table 20 by comparing
percent of reading to percent of full scale. The least error was at
maximum flow where percent of reading equals percent full scale (-1.2
percent to 0.6 percent). Error as percent of reading becomes progres-
sively more significant as flow decreases (-11.63 to 6.14 at 6 in. head).
Backlash in this instrument was too great and improvement is warranted.
Some of this error would be eliminated if anti-backlash gears were used;
the use of forms of electronic devices such as a functional amplifier or
microprocessor instead of gears and cams would completely eliminate error
of this type.
Table 20. DEADBAND (BACKLASH)*
Integrator
Error
Run
94
95
96
Battery
(volts)
12.62
12.59
12.59
Flow rate
pot (volts)
0.391
0.337
0.389
Liquid
level dial
(in)
6*
61-
64-
Percent flow
dial chart
9 7.5
7.5 6.5
9 7
AT
(min)
155.6
153.9
18.4
Flow from
tables
(gal/min)
200
200
200
Total
flow
(gal)
31,125
30 , 785
3,684
Percent
of max.
flow
10.29
10.29
10.29
Total
flow
(gal)
33036.1
27206.2
3886.6
%
Read-
ing
6.14
-11.63
5.5
%
Full
scale
0.6
-1.2
0.6
"Instrument calibrated at 15 inches maximum flow for 90°V notch weir.
•(•Descending electrode.
^Ascending electrode.
42
-------
OVERALL ACCURACY AND PRECISION
Overall accuracy and precision are illustrated in Tables 21, 22, and
23. These runs were made with the equipment at 2, 22, and 35C. In Table
21, for example, run 85 shows that the vernier height gage was set at 6
in. and the dipper stopped slightly above 6 in. The length of this run
times the flow rate from Manning's tables gave total flow as 37,925 gal,
whereas the counter read 39,331 for an error of 3.7 percent. Error as
percent of full scale was 1.72 percent. The instrument dial and chart
read 46 percent and 45 percent, respectively, compared to the reading
from the tables of 46.5 percent. It is seen from the tables that error,
as a percent of reading, was greatest at low heads and became less signi-
ficant as maximum flow was approached. Most of this error, as mentioned
earlier, was caused by deadband or backlash. Tables 21, 22, and 23 also
show that the percent of maximum flow on the dial was usually a little
lower than the reading from the tables, and the chart reading was lower
than the dial. The dial and chart need to be adjusted for better accu-
racy. The elimination of backlash is required for better precision.
POSSIBLE THEORETICAL INACCURACY
The F-3000 flowmeter converts head into flow rate by using three cams
that are machined as a function of (h)3'^ (h)^/^, and the Manning equa-
tion for circular pipe. Some error will exist if the equation for the
specific primary device has a slightly different exponent. For example,
the equation for small Parshall flumes are:
3" Parshall flume Q = .992H1'547 (1)
6" Parshall flume Q = 2.06H1-58 (2)
9" Parshall flume Q = 3.07H1'53 (3]
Q = flow rate (ft3/sec)
H = head(ft)
When Manning's (h)1'5 cam is used for small Parshall flumes, slight
error will exist. Equations for many primary devices, such as suppressed
rectangular, Cipolletti and V-notch weirs, have exponents of 3/2 and 5/2.
The F-3000 flowmeter is convenient for portable applications, but perma-
nently installed instrumentation for flow measurement should follow the
exact equation of the primary device.
43
-------
Table 21. OVERALL ACCURACY AND PRECISION FOR CIRCULAR PIPEt
Run
84
47
81
85
45
82
86
46
83
Temperature
(0
2
22
35
2
22
35
2
22
35
Level
Vernier
4
4
4
6
6
6
11.25
11.25
11.25
(in)
Dial
4*
4
4*
6.05
6*
6
11.3
11.25
11.25
Table
(gal)
35,314
41,007
35,795
37,925
39,505
36,874
147,512
302,966
150,376
Total flow
Counter Error, % of
(gal) Reading Full
37,621
39,331
37,621
39,331
39,331
37,621
150,483
306,095
150,483
6.5
-4.1
-0.4
2.6
1
0.07
1
-0
1
1
-0
1
2
1
0
scale
.45
.91
.14
.72
.19
.21
.0
.0
.07
Percent
From
tables
22.3
22.3
22.3
46.5
46.5
46.5
100
100
100
of maximum
Dial
22
20
22
46
45
46.5
100
100
100
flow
Chart
19.5
19
18
45
42
45
100
100
99
tStatic tests, dipper touching beaker of water or Al plate attached to slide of vernier height gage.
*Set point.
Table 22. OVERALL ACCURACY AND PRECISION FOR 90°V-NOTCH WEIRt
Run Temperature
Level (in)
Total flow
Percent of maximum flow
78
50
75
79
48
76
80
49
77
(0
2
22
35
2
22
35
2
22
35
Vernier
6
6
6
10
10
10
15
15
15
Dial
6*
6
6*
10
10*
10
15
15
15
Table
(gal)
38,535
48,402
38,320
45,149
92,585
43,654
86,525
173,129
83,416
Counter
Error, % of
(gal) Reading Full
42,753
44,696
42,753
42,753
87,449
40,809
89,392
176,840
85,505
10.9
-7.7
11.6
-5.3
-5.5
=6 iJ>
3.3
2.1
2.5
1
-0
1
-1
-2
-2
3
2
2
scale
.12
.79
.19
.94
.01
.38
.3
.1
.5
From
tables
10.3
10.3
10.3
36.6
36.6
36.6
100
100
100
Dial
9.5
7.5
10
33
33.5
35
101
102
104
Chart
7.5
7.5
7.5
31.5
32
31.5
100
100
101
tStatic tests, dipper touching beaker of water or Al plate attached to slide of vernier height gage.
*Set point.
Table 23. OVERALL ACCURACY AND PRECISION FOR SIX-INCH PARSHALL FLUMEt
Run
51
52
53
Temperature
(C)
22
22
22
Level (in)
Vernier Dial
3 3*
6 6.1
9.01 9.05
Table
(gal)
12,825.6
26,797
105,266
Total flow
Counter Error ,
(gal) Reading
13,409 4.5
26,235 -2.1
106,689 1.4
Percent of maximum flow
% of
Full scale
0.86
-1.14
1.14
From Dial
tables
19.2 18
54.5 52.5
100 101
Chart
16
50
100
tStatic tests, dipper touching beaker of water or Al plate attached to slide of vernier height gage.
*Set point.
44
-------
SECTION 7
DISCUSSION
The S-4000 sampler and F-3000 flowmeter are well engineered and
designed. They incorporate most features that are desired in portable
equipment. They are light in weight, fairly rugged and easy to handle.
The sampler's design is good in that it incorporates solid state elec-
tronics. Flowmeter electronics are also solid state and include digital
output circuitry. The use of some form of microelectronics within the
flowmeter such as a microprocessor to eliminate cams and other mechanical
components should be considered. Greater use of "plug in" components and
circuits should be incorporated. A small event recorder to mark the time
that the sample was taken on both time- and flow-proportional runs would
be helpful. Improvements are indicated in the conclusions, and it is
hoped that they will be effected.
In general, the overall design and performance of the S-4000 sampler
and F-3000 flowmeter was above average when compared to other equipment
of this type.
45
-------
SECTION 8
REFERENCES
1. Methods Development and Quality Assurance Research Laboratory,
National Environmental Research Center, Cincinnati, Ohio, "Methods
for Chemical Analysis of Water and Wastes," U.S. Environmental
Protection Agency, Technology Transfer, Washington, D.C., 1974.
2. Livingston, J., "Battery Application Note" 5-475, Manning Environ-
mental Corp., Santa Cruz, Calif.
3. Walpole, R. E., and Myers, R. H., "Probability and Statistics for
Engineers and Scientists," The Macmillan Company, 1972, p 244, ex
7.4.
4. Harris, D. J., and Keffer, W. J., "Wastewater Sampling Methodologies
and Flow Measurement Techniques," U.S. Environmental Protection Agency,
Region VII, June 1974.
46
-------
SECTION 9
APPENDIX
50
40
30
25
20
15
5"
CM
I 10
.1 9
I 8
< 7
6
5
4
3
2.5
2
13
1
TT
BARCO
3/4-425
I I I I I I
I I
I
.4 5 h 7 .8 .9 1.0 1.5 2 15 3 4 5 6 7 8 9 10
VELOCITY (ft/sec)
FIGURE 1. Velocity versus AP for 0.620 in. I.D. pipe
(isokinetic sampling unit).
47
-------
Table 1. ORIGINAL DATA
RUN 1, TIMED TEST*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Cycle time
(min)
manual
29.73
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
30
Volume X
(ml)
279
278
277
277
276
279
279
277
277
277
278
277
278
279
279
279
277
278
278
277
277
276
278
X = 277.7
Sx = .974
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
RUN 21, TIMED TEST*
Cycle time
(min)
20.4
18.8
18.7
18.7
18.6
18.7
18.6
18.5
18.5
18.4
18.4
18.3
18.3
18.3
18.3
18.3
18,3
18.3
18.3
18.2
18.2
18,2
18.2
18.2
t = 18.49
Sx = .448
Volume X
(ml)
369
373
372
370
372
373
372
373
373
372
371
373
372
370
370
370
370
370
370
370
370
371
371
372
X ~ 371.21
Sx = 1.28
*Room temperature: 21.7C
Water temperature: 19.8C
Volume set at approximately 300 ml.
*Room temperature: 24.3C
Water Temperature: 19C
Volume set at approximately 400 ml.
Cycle time set at 30 min.
Battery startt: 86%
Battery endf: 82%,
tSimpson meter 15V scale.
48
-------
Table 2. ORIGINAL DATA
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
RUN 22, TIMED
Cycle time
9.7
9.2
9.2
9.2
9.0
9.0
9.0
9.Q
8.9
8.9
8.8
9.0
8.9
9.0
9.0
8.9
9.0
9.0
9,0
8.9
8.8
8.8
8.8
8.9
t «» 8.996
S«. = 0.190
t
TEST* •
Volume X
(ml)
376
i
o
S-i
fa,
3- -
e
•P O
W*«J
•^^
f-i O
0)
•08-
fl} C
(/) rH
rt ri
g 3
fi 10
•H
+J >
-------
Table 3. ORIGINAL DATA
RUN
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
— -
24, TIMED
• ' '• •' *^mm,^mm, , , ^l
TEST*
Cycle time Volume X
(min) (ml)
60
60
60
60
60
60
60
60
60
$ 60
e 60
•* 60
g 60
£ 60
< 60
60
60
60
60
60
60
60
60
60
380
381
381
381
380
381
380
383
381
379
380
380
380
381
380
380
382
380
380
380
380
381
380
381
X = 380.5
SY = 0.843
A.
RUN
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 '
20
\ 21
22
23
24
25, TIMED
Cycle time
(min)
60
60
60
60
60
60
60
60
60
$ 60
1 60
'x 60
g 60
& 60
< 60
60
60
60
60
60
60
60
60
60
nil 1 ill I I II 1 '-
TEST*
Volume X
(ml)
382
383
383
384
383
384
383
383
384
384
384
385
382
382
383
382
383
382
383
383
381
381
382
383
X = 382.9
Sx = 0.992
Entire run took 0.6 min too long.
*Room temperature: 22.8C
Water temperature: 21.6C
Volume set at approximately 400 ml.
Used battery and 500 ma charger.
Entire run took 0.6 min too long.
*Room temperature: 21.3C
Water temperature: 22.8C
Volume set at approximately 400 ml,
50
-------
Table 4. ORIGINAL DATA
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
RUN 32, TIMED
Cycle time
(min)
30
30
30
30
30
30
30
30
30
® 30
I 30
'330
£30
&30
•* 30
30
30
30
30
30
30
30
30
30
TEST*
Volume X
(ml)
387
390
390
388
388
388
387
388
388
388
387
386
387
388
385
388
386
387
387
387
387
388
387
387
X = 387.5
SY = 1.103
X
RUN
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
t
St
33, TIMED
Cycle time
(min)
60.000
60.013
60.025
60.013
60.063
60.038
60.013
60.025
60.038
60.025
60.025
60.025
60.025
60.025
60.025
60.038
60.025
60.025
60.000
60.025
60.013
60.013
60.038
60-000
= 60.023
= 0.014
TEST*
Volume X
(ml)
384
384
386
386
386
384
384
384
385
384
384
385
384
385
384
384
385
384
385
385
385
384
385
384
X = 384.6
Sx = 0.717
Entire run took 0.3 min too long.
*Chamber temperature: 2C
Water temperature: 13C
Volume set at approximately 400 ml,
Battery startt: 84%
Battery endt: 80%
tSimpson meter 15V scale.
Entire run took 0.55 min too long.
*Chamber temperature: 2C
Water temperature: 3C
Volume set at approximately 400 ml.
Used Honeywell recorder.
Chart speed: 0.8 in./min.
51
-------
Table 5. ORIGINAL DATA
RUN
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Entire run
/•«_ i j_
34, TIMED
Iced cycle
time (min)
60
60
60
60
60
60
60
60
60
o 60
« 60
.S 60
g 60
& 60
^ 60
60
60
60
60
60
60
60
60
60
-
TEST*
Volume X
(ml)
383
386
384
386
510t
•£
o
ft
to •
CD
•P i-l
§>*£
.tt O
,0
§ x
CD
O
tJ -M
0)
rt C
O -H
f""( PH
M-l PH
0)
O
O f-l
PQ 4-1
RUN
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
35, TIMED
Iced cycle
time (min)
60
60
60
60
60
60
60
60
60
o 60
rt 60
•H 60
o 60
| 60
<"' 60
60
60
60
60
60
60
60
60
60
took 0.5 min too long.
TEST*
Volume X
(ml)
380
378
380
380
380
381
380
380
380
379
381
380
380
380
380
380
379
379
380
380
380
381
380
381
X = 380
SY = 0.69
A
Water Temperature: 20C
•(•Overflowing
Battery startt: 85%
Battery endf: 82%
tSimpson meter 15V scale.
Entire run took 0.7 min too long.
*Chamber temperature: 21.9C
Water temperature: 20C
Volume set at approximately 400 ml.
52
-------
Table 6. ORIGINAL DATA
RUN 56, TIMED TEST*
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Iced cycle
time (min)
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
Volume X
(ml)
382
383
384
383
384
383
383
384
384
385
385
384
384
384
384
384
383
383
383
383
383
385
385
385
X = 383.8
SY = 0.85
A
Entire run took approximately 0.5 min
too long.
*Chamber temperature: 31C
Water temperature: 32C
Volume set at approximately 400 ml.
Battery startt: 84%
' Battery endt: 83%
tSimpson meter 15V scale.
53
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/4-76-059
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
PERFORMANCE INVESTIGATION OF THE MANNING MODEL S-4000
PORTABLE WASTEWATER SAMPLER AND THE MODEL F-3000 DIPPER
FLOWMETER
5. REPORT DATE
December 1976 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Richard P. Lauch
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Lab.
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
- Gin., OH
10. PROGRAM ELEMENT NO.
1HD621
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
In-House
Same as above
14. SPONSORING AGENCY CODE
EPA/600/06
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Performance of the Manning model S-4000 wastewater sampler and the model F-3000
flowmeter was investigated.
The S-4000 wastewater sampler was tested at temperatures of 2, 20, and 35C to
determine accuracy and precision of the timer and sample volumes. The multiplexer
function of delivering multiple aliquots per bottle was tested. Tests for ability
to fill up to four bottles with the same sample were made. Battery endurance was
determined. Discrete sample temperatures versus time were recorded under iced condi-
tions to determine preservation capability. Field tests were performed to determine
representative collection of suspended solids and ability of the unattended sampler
to collect raw sewage samples over a 24-hour period.
The F-3000 flowmeter was tested within the laboratory for accuracy and precision
of tracking, analog to digital conversion, deadband, and electronic drift caused by
temperature change and battery decay. Accuracy of the flow chart and integrator was
determined.
Manufacturer's claims were mostly confirmed, however improvement is warranted
for some functions of the sampler and flowmeter.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Group
Samplers, Water Pollution, Acceptance
sampling, Continuous sampling, Data
sampling, Sequential sampling, Flowmeters.
Sampler evaluation,
Evaluation, Sewer samp-
ler evaluation, Effluent
sampler evaluation,
Water sampler evaluation
Water sampler.
13B
8. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (This page I
UNCLASSIFIED
_62_
22. PRICE
EPA Form 2220-1 (9-73)
ft U.S. GOVERNMENT PRINTING OFFICE; 1977— 757-056/5547
54
------- |