lution
Apparaf.i
Continuous Delivery
jus Concentrations
n Water
ENVIRONMENTAL HEALTH
SERIES
Water Supply
and Pollution Control
-------
A SERIAL-DILUTION APPARATUS
FOR CONTINUOUS DELIVERY
OF VARIOUS CONCENTRATIONS
OF MATERIALS IN WATER
Donald I. Mount
Aquatic Biology Section
Basic and Applied Sciences Branch
Robert A. Taft Sanitary Engineering Center
and
Richard E. Warner
Engineering Science, Inc.
Oakland, California
U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
Division of Water Supply and Pollution Control
Cincinnati, Ohio
June 1965
-------
The ENVIRONMENTAL HEALTH SERIES of reports was estab-
lished to report the results of scientific and engineering studies of
man's environment: The community, whether urban, suburban, or
rural, where he lives, works, and plays; the air, water, and earth he
uses and re-uses; and the wastes he produces and must dispose of in a
way that preserves these natural resources. This SERIES of reports
provides for professional users a central source of information on the
intramural research activities of Divisions and Centers within the
Public Health Service, and on their cooperative activities with State
and local agencies, research institutions, and industrial organizations.
The general subject area of each report is indicated by the two letters
that appear in the publication number; the indicators are
WP - Water Supply
and Pollution Control
AP - Air Pollution
AH - Arctic Health
EE - Environmental Engineering
FP - Food Protection
OH - Occupational Health
RH - Radiological Health
Triplicate tear-out abstract cards are provided with reports in
the SERIES to facilitate information retrieval. Space is provided on
the cards for the user's accession number and key words.
Reports in the SERIES will be distributed to requesters, as
supplies permit. Requests should be directed to the Division identi-
fied on the title page or to the Publications Office, Robert A. Taft
Sanitary Engineering Center, Cincinnati, Ohio 45226.
Public Health Service Publication No. 999-WP-23
-------
CONTENTS
Page
Abstract v
Introduction 1
Materials 2
General Principle of Operation 2
The Serial-Dilution System 2
Water Delivery System 6
Toxicant-Metering System 9
Calibration 11
Timing 12
Trouble Shooting 13
Performance 14
Reference 16
-------
ABSTRACT
This paper describes a serial-dilution apparatus designed to
deliver continuously a series of different concentrations of a material
in water. The materials needed for construction normally would be
available in a. chemistry laboratory. No electrical power is needed
for operation, and the apparatus will remain accurate even if the in-
fluent waterflow varies over a wide range. It maintains accuracy of
10 percent or less for periods of time up to 30 days or more with
very little servicing or adjustment, and the cost is $50 or less.
-------
A SERIAL-DILUTION APPARATUS
FOR CONTINUOUS DELIVERY
OF VARIOUS CONCENTRATIONS
OF MATERIALS IN WATER
Introduction
The evaluation of adverse effects of toxicants or polluting agents
on aquatic animals is receiving increasing attention as pollution of the
nation's water becomes more severe. This evaluation requires that
the aquatic animals be exposed to an environmental concentration
rather than to an injected dose of the agent. It is necessary, there-
fore, to use tests in which the concentration in water of the material
being tested is known and maintained constant. These conditions can-
not be met in a static or nonrenewed test because the toxicant may
volatilize, precipitate, adsorb onto suspended material, or be ab-
sorbed by the test organisms, and thus the amount of toxicant in the
water would be constantly changing.
Static tests often require additional aeration to maintain the
necessary dissolved oxygen (DO) in the water, and test animals
cannot be fed because the water becomes foul. Moreover, static
tests cannot be continued for more than a few days because metabolites,
such as ammonia, accumulate in the water. Periodic renewal of the
test -water offsets some of the objections to the static test, but the
process is laborious and does not sufficiently maintain controlled
conditions.
Continuous-flow tests are the best tool available at present to
evaluate toxic effects under controlled conditions. For laboratory
studies, they most closely approach actual pollutional situations,
and are, therefore, being used more commonly by workers in the
field. The Newtown Laboratory of the Robert A. Taft Sanitary En-
gineering Center has been engaged in continuous flowthrough testing
for 3 years, and the problem of maintaining concentration control on
100 or more flowthrough test chambers has been troublesome and has
required much costly time of technicians. The cost of metering
pumps and valves needed to do the job adequately have increased,
and malfunctions have occurred regularly. Nearly two full-time
positions have been required to maintain and check the systems each
day, including weekends and holidays. The most critical fault of the
systems in use--a fault that became impossible to tolerate consisted
in a metering system's delivering toxicant solution when the waterflow
slowed or failed. The result was the death of the fish in the test
chambers. These problems became acute when long-term studies on
fish reproduction were begun; a year's \vork could be lost in one night
in this manner. The apparatus described in this paper is a result of
efforts to improve dependability of the dosing process.
In 1962, Mr. Richard Warner of Engineering Science, Inc.,
Oakland, California, in an annual report on a Public Health Service
contract, described a serial-dilution apparatus that he had built to
conduct flowthrough tests. Serial dilution has many advantages:
1
-------
It is more accurate in most cases, fewer controlling valves are
needed, only one stock solution is required, and there are fewer small
constrictions to become clogged. We have used the principles from
Mr. Warner's apparatus to build diluters suitable for our use. We
now have 13 of these diluters in operation; they have improved our
accuracy of concentration control, at least 40 man-hours per week
have been saved, the number of malfunctions has dropped sharply,
and space requirements have been reduced. In addition, the toxicant
dose is dependent on waterflow, the cost of the metering system has
been reduced from $600 to $800 to $50 or less, and the system can
handle water with plankton or suspended solids of surprisingly large
particles and still not clog.
MATERIALS
Supplies normally available in a chemistry laboratory have been
used to construct the apparatus. Crystallizing dishes and sidearm
vacuum flasks are useful for making the various cells. Epoxy glue
is especially convenient for gluing glass tubes in place and making
watertight seals. A polyethylene hose "X" can be used for the valve
housing.
GENERAL PRINCIPLE OF OPERATION
The serial-dilution apparatus described here consists of three
functional components: the water delivery system, the toxicant feed
system, and the serial-dilution system. In essence, water delivery
is accomplished by filling the water-metering cells, stopping the
waterflow to the cells, emptying the metering cells, and restoring
the waterflow. This sequence must be completed twice to produce
one cycle of the serial-dilution system; water containing the appro-
priate toxicant concentration is delivered once per cycle. The
sequence of delivering 1 unit volume of -water from each metering
cell, together with the associated changes in the serial-dilution
system, is termed a half cycle.
THE SERIAL-DILUTION SYSTEM
To grasp better the major functional concepts of this apparatus,
temporarily disregard the water delivery and toxicant feed systems
and consider only the serial-dilution system. Refer to Figures 1,
2, and 3 and cells M-l, T-l, and M-2. For illustration, assume a
desired volume of 100 milliliters of dosed water for each concentra-
tion per cycle. If a 2:1 dilution ratio is used, twice the desired
volume of the first concentration must be made to have a sufficient
quantity for making other dilutions. At the first half cycle, 100 milli-
liters is discharged into chamber M-l along with an appropriate
quantity of toxicant, Figure 3A. Chamber M-l will then be one-
half full, and during the next half cycle, an additional 100 milli-
liters of water is added along -with toxicant, giving a total volume of
200 milliliters, a volume that fills cell M-l and causes the siphon
to start. T-l is adjusted so that the transfer volume (volume of
water contained between the levels of the standpipe and the end of the
2 A SERIAL-DILUTION APPARATUS
-------
VACUUM MANIFOLD
WATER SOURCE
i- VALVE
VACUUM VENTURI
Figure I. Diagram of the serial-dilution apparatus. Flasks I to 6 at the top are metering
cells, M-l to M-5 are mixing cells, T-l to T-4 are transfer cells, and V-l to V-4
are venturi tubes. The toxicant-metering system is not shown
The Serial-Dilution System
-------
venturi siphon tube, Figure 2) is 100 milliliters. As the siphon of M-1
discharges water into T-l, 100 milliliters remains in T-l and 100
milliliters goes down the standpipe to a test chamber or other location,
Figure 3B.
During the next half cycle, 50 milliliters of water is delivered
through transfer venturi 1 into cell M-2 (Figure 1). As the 50 milli-
liters passes through the venturi, the siphon from T-l to M-2 starts
and 100 milliliters of the previous concentration (presently in T-l)
is added, making cell M-2 three-fourths full (Figure 3C). Finally,
50 milliliters of water is again added to M~2. Since T-l contains no
VENTURI SIPHON TUBE
TRANSFER CELL
FROM FUNNEL
TRANSFER VENTURI
VENTURI FLARE
_J
- MIXING CELL .
AUTOMATIC
" SIPHON
D
Figure 2. Details of the transfer venturi, transfer cell, and mixing cell.
water that can be siphoned by the venturi siphon tube, none is trans-
ferred from T-l to M-2 during this half cycle. The addition of the
second 50-milliliter volume to M-2 fills it, and the siphon starts and
empties 200 milliliters of solution into T-2, Figure 3D. This 200-
milliliter volume is one-half as concentrated as the first 200-milli-
liter volume in M-l. During the two half cycles when water was
being added to M-2, 100-milliliter volumes were being added to M-l
A SERIAL-DILUTION APPARATUS
-------
so that when M-2 is full, M- 1 is full again, and both empty simul-
taneously and fill T-l and T-2. The same sequence is continued for as
many concentrations as are desired. Two-hundred milliliters of the
last concentration will be discharged from the "M" cell because no
water is transferred downward to make additional concentrations.
Water is discharged on half cycles rather than once each cycle
to simplify timing problems. One can see from Figure 3 that all of
the 50 milliliters entering the venturi on T-l must clear the venturi
before the 200 milliliters from M-1 fills T-l again. This is easily
accomplished by half-cycle operation, but if water were delivered
Figure 3. General diagram illustrating the operation of the serial-dilution system.
only once per cycle, 100 milliliters of water from the metering cell
and 100 milliliters from T-l would have to arrive in M-2 before any
of the 200 milliliters from M-1 reached T-l. If this did not happen
and the venturi siphon on T-l was still discharging water to M-2
when M-l began emptying into T-l, all of the water emptied from M-l
to T-l would subsequently be transferred to M-2 and then on to T-2,
and so forth. Water delivery each half cycle makes possible a safety
factor that prohibits the action referred to above and that is called
"double siphoning. "
The Serial-Dilution System
-------
In addition, by using half-cycle delivery, the diluter need not
remain "in phase. " "In phase" refers to sequences in which all "T"
cells are full and all "M" cells are empty. After the next "in phase"
half cycle, all "M" cells are half full and all "T"-cell transfer volumes
are empty. By using half-cycle delivery, cells T-l and T-3 may be
full while the transfer volumes of T-2 and T-4 are empty. Conse-
quently M-l, M-3, and M-5 would be empty and M-2 and M-4 would be
half full when the "T" cells are as described above. The condition
that must be met is that when a "T" cell is full, the "M" cell immedi-
ately above it must be empty, and when the transfer volume of a given
"T" cell is empty, the "M" cell immediately above must be half full.
Each diluter usually operates in a particular pattern called the phase
sequence pattern. The latter one described above is called the alter-
nating sequence pattern and is the more common because timing is
less critical.
In practice, the total volume need not be used on any cell. One
can shorten the siphon tube in the "M" cells, for instance, and use
only half of their volume. In this instance the "M" cells would never
empty completely. If a dilution of 1:3, or any other ratio, is desired,
the volume transferred from the "T" chambers to the "M" chambers
is changed accordingly to give the desired ratio. An example would be
a transfer volume of 75 milliliters and a half-cycle water delivery of
112.5 milliliters to M-l and 75 milliliters to the others.
Figure 2 shows the details of the "T" cell and the "T"-cell
venturi. A very efficient venturi can be made from either "T"- or
"U"-shaped glass connecting tubes if they are cut and modified as
shown. One important precaution is that the venturi flare must have
the same or slightly larger bore at its upper end as the bore of the
venturi "U" tube. If the upper end of the flare is smaller, a back
pressure will develop on the venturi siphon tube and the siphon will
not start.
Another suggestion for better venturi operation is the use of a
tube bore for the funnel siphon and funnel siphon tube (Figure 4) that
is slightly smaller than the bore of the "U" tube venturi. The funnel
siphon tube must fill with water without air for proper venturi opera-
tion. The primary purpose of the funnel siphon is to keep air out of
the funnel siphon tube. Many of our diluters operate without these
funnel siphons because the funnel siphon tube is small enough and the
metering cells empty fast enough to prevent air from entering the
funnel siphon tube.
Finally, the venturi flare must not be excessively large or else
the water and air mixture (the air enters from the venturi siphon tube
until the water from the "T" cell arrives) passing through the flare
will not completely fill it and there will be no venturi. An improper
taper on the flare or a nonwettable film such as grease can cause the
same type of trouble.
WATER DELIVERY SYSTEM
Other types of water delivery systems \vere tried in conjunction
with the serial-dilution system, but these would not deliver water with
6 A SERIAL-DILUTION APPARATUS
-------
WATER DELIVERY TUBE
TO VACUUM MANIFOLD
WORKING
VOLUME
WATER METERING CELL
WATER INLET TUBE
FROM WATER MANIFOLD
VOLUME ADJUSTMENT COLLAR
GLASS "U" TUBE
FUNNEL SIPHON
'.-MINIMUM WATER LEVEL
FUNNEL SIPHON TUBE
TO TRANSFER VENTURI
Figure 4. Detail of a water delivery unit.
Water Delivery System
-------
an accuracy comparable to that achieved by the serial-dilution system.
This system has the advantage that one can visually determine whether
or not the proper volume of water is being delivered, with an error
of 10 percent or less. More important, the volume delivered each
half cycle is not affected by the rate at which water is fed to the system
and there are no small water passages that can clog with solid material.
As shown in Figures 1 and 4, the water delivery system is com-
posed of a series of units, each consisting of a metering cell with a
water inlet, air tube and water delivery tube, a venturi, and a funnel.
There is a common vacuum manifold, vacuum venturi, and water shut-
off valve.
. n ,
\7_ STOPPER
WATER FROM CELL NO. 6
I
VAUE SPRING
SLICED STOPPER
- GLASS NEEDLE
TO CONTROL OUTLET TO WATER MANIFOLD
Figure 5. Detail of the main water valve and vacuum venturi.
Frequent reference to Figures 1, 4, and 5 will assist in under-
standing the operation. A water manifold delivers water to each of
the metering cells, which fill until the water level is at the end of the
air tube. Cell Number 6 fills last because the inlet to it is higher than
that of the others, and it empties first because its outlet is lowest.
As water from the last cell fills the bucket on the needle valve arm,
the valve is closed by the additional weight. When the bucket is filled
A SERIAL-DILUTION APPARATUS
-------
to the level of the siphon tube, the water travels through the vacuum
venturi, producing a partial vacuum in the vacuum manifold, and is
then discharged to the control chamber. The reduced pressure in the
vacuum manifold causes water to rise from both ends of each water
delivery tube (from the funnel and from the metering cell). If distance
A' is less than distance A" (Figure 4), the water will rise over the arch
of the tubes and activate the siphons, which will empty the metering
cells down to the end of the \vater delivery tube. By the time the si-
phons break, the bucket on the needle valve arm will be nearly empty,
the valve will open, and the metering cells will fill again.
The sequence in which the metering cell siphons start is de-
termined by the realtive distance of A' of each cell. Figure 1 illus-
trates the most desirable order, namely, Numbers 1 and 5 start at
nearly the same time and then Numbers 4, 3, and Z begin in that
order. Number 1 should begin emptying early because if a 2:1 dilu-
tion is used the first cell must deliver twice as much water as the
other cells.
The "funnel end" of each water delivery tube must remain under
water at all times so that a partial vacuum can be created in the
vacuum manifold. A bypass spout on the funnel is used to measure the
water delivered by the cell during operation. The funnel siphon and
bypass spout are positioned so that when the bypass is open enough
water remains in the funnel to cover the end of the water delivery
tube and so that all the water delivered goes through the bypass and
none goes over the funnel siphon to the transfer venturi.
Major volume adjustments of the metering cells are accomplished
by moving the volume adjustment collar (Figure 4), and minor ones
are made by moving the air tube. Caution must be exercised, how-
ever, since moving the air tube changes distance A', which could
change the sequence of delivery. If changed enough, the distance
may exceed A" and then the siphon will not start.
Care should be taken in constructing the vacuum venturi to see
that proper tube sizes and tapers are used to produce sufficient vacuum.
If it is properly made, more than enough vacuum can be produced
with ease.
TOXICANT-METERING SYSTEM
The toxicant-metering system is not necessarily an integral
part of the apparatus. We have used several other types such as
metering pumps and controlled dripping rates but all have been in-
ferior to the one shown in Figure 6. It has the important advantage
that it will not deliver toxicant unless waterflow is near normal.
The system as shown in Figure 6 is nearly self-explanatory.
The solution level in the funnel is kept constant by some means such
as a Mariotte bottle or float valve. If the toxicant is a suspension
rather than a solution, settling may occur. A small peristaltic
pump may be used to pump the suspension to the funnel with a bypass,
overflow line back to the bottle. In this way the level in the funnel is
maintained and the suspension is mixed and replaced, which eliminates
settling.
Toxicant-Metering System 9
-------
For use on the diluter apparatus, the plastic bucket of the meter-
ing system is placed beneath the first funnel of the diluter (Figure 1)
and when water flows through the funnel siphon tube the bucket is
filled with water. This weight causes the bucket arm to drop and the
opposite arm rises upward taking with it a predetermined volume of
toxicant, which runs into M-1 through the tube arm.
The components are easily assembled and the tube is simply
made by heating an appropriately sized piece of glass tubing in a
narrow area, stretching it until the bore is sufficiently reduced, cool-
ing, and then bending the two angles. The bore is constricted so that
minor fluctuation in the solution level in the funnel will not significantly
change the volume delivered.
MAfilOTTE BOTTLE
DIRECTION Of ROTATION
Figure 6. Detail of toiicant-metering system. Bottle is not in proportion to drawing.
The accuracy of this system equals or exceeds the accuracy of
systems using metering pumps and constant waterflows because de-
livery is balanced against a water volume rather than a flow; thus a
major source of variation (waterflow) is bypassed. Obviously, for
best calculations one should know how many times the metering cell
empties and likewise how many times toxicant is added. Since it did
not seem advisable to put a counter switch on the metering system
because of its small size, a microswitch was placed on the needle
valve arm (It moves each time the Number 1 cell empties. ), giving an
actual count of the number of times chemical is delivered. (Mechani-
cal counters are also very good because no electricity is needed. )
10
A SERIAL-DILUTION APPARATUS
-------
By using a calibrated toxicant bottle, the total volume of toxicant
delivered can be read, the number of counts can be obtained from the
counter, and the average volume of toxicant delivered each half cycle
can, therefore, be calculated. Then, by measuring the volume of
water delivered by metering cell 1, the calculated accuracy of the first
concentration can be determined.
CALIBRATION
Calibration of the various volumes seems to be a laborious
task, but it can be completed in 10 to 15 minutes with experience.
After the wetting characteristics of the tubes have stabilized, the
calibration needs to be checked only every 2 to 4 weeks to ascertain
that the maximum error is 10 percent or less. The diluter should be
checked visually every day to insure good operation and detect mal-
functions such as that caused by a broken component.
Only two sets of volumes need to be known in order to calculate
the error for each concentration, the first excluded. These are the
amounts of diluent water being added at each half cycle and the trans-
fer volume (the volume of the next higher concentration being mixed
with the diluent water from the water-metering cells).
The volume of diluent water added is checked by opening the
bypass on the funnel and catching the volume for one or more half
cycles for each metering cell. One must be careful to see that all
water delivered goes through the bypass and that none goes over the
funnel siphon.
The transfer volume is checked as follows: Assume a desired
volume of 100 milliliters. Fill a 250-milliliter graduated cylinder to
some volume greater than 120 milliliters. Wait until the "T" cell
to be checked is just emptied by the venturi siphon. Pour water into
the "T" cell until it is full, then add 20 milliliters or more in addition
and catch the water that runs down the standpipe before the next half
cycle begins. Return this water to the graduated cylinder. Read
the volume now remaining in the cylinder; the difference between the
present volume and the original volume represents the transfer volume
of the "T" cell being checked. When adding the water to the "T" cell,
one should fill it at approximately the same rate as it is filled by the
"M" cell siphon above it in order to simulate actual operation.
After all volumes are recorded, the error for each concentration
can be calculated. Obviously, the error for the first concentration
must be determined by using the volume of water delivered from
metering cell 1 and the volume of toxicant delivered by the metering
system.
Each diluter will usually operate in a typical phase sequence.
Obviously, checking the volumes as described above disrupts this
sequence because water is being added or removed. It is best,
therefore, to check the transfer volumes first, starting with the
lowest transfer cell, and then check the metering cell volumes. In
this way each transfer volume is checked when the diluter is in
its typical phase sequence. The checks can be made while the diluter
Calibration 1 1
-------
is operating, and the removal or addition of water for checking will
cause one or at most two errors. The diluter will correct itself and
continue properly, but possibly in a different phase sequence. If a
diluter is well timed it will stay in any phase sequence in which it is
placed. This pattern offers a useful visual check of proper operations.
If the pattern keeps changing every few minutes, the diluter is making
errors.
Other visual checks that can be made for proper operation are:
1. Determine whether, when an "M" cell is empty, the "T"
cell immediately below it is full. (This is a check for double
siphoning. See page 5.
2. Observe the sequence in which water runs down the stand-
pipes of the "T" cells. This should occur on each second
half cycle. A malfunction may indicate that a "T" cell
venturi is not working.
3. Observe whether the needle valve is turning off the water com-
pletely and for a period long enough to allow the metering
cells to dump.
4. Observe whether each metering cell is filling and emptying
completely each half cycle. Other checks will be learned by
experience. These checks may seem tedious, but with
experience, a 2-minute daily visual check is sufficient to
insure good operation regularly.
TIMING
In performing the calibration, one may find that the system may
need to be timed more accurately. This may also be necessary after
the wetting characteristics have stabilized. Timing adjustments solve
the double-siphoning problem mentioned on page 5. When a diluter
is optimally timed, the serial-dilution system can be disrupted for
some reason such as for a calibration check or accidentally by dirt,
air, and so forth, and the system will make only one error and then
return to proper operation. This ability to correct mistakes is
highly important because during normal operation things will occur that
cause an error and only if the system can correct errors and not
repeat them can it be considered satisfactory.
To review: Double siphoning occurs when all the water from a
given metering cell has not passed the transfer venturi as the asso-
ciated "T" cell begins to fill from the next higher "M" cell. The
result is that the venturi siphon begins to remove water from its
"T" cell when it should not.
There are three timing adjustments: (1) The sequence in which
the metering cells empty, (2) the volume of water required to start
the funnel siphons, and (3) the volume required to start the "M" cell
siphons. If the sequence in which the siphons of the metering cells
start is as described earlier, then frequently no other adjustment is
needed. If difficulty is experienced in getting the desired sequence
or if an additional safety factor is desired, or both, the next best
12 A SERIAL-DILUTION APPARATUS
-------
timing adjustment is the "M" cell volume. The volume of the cell is
reduced by shortening the siphon tube or by adding volume displacers
such as marbles, in order to make the cell empty sooner, or the tube
length is increased, or volume displacers are removed to make it
empty later. As a last resort, and this is rarely necessary, one
can change the volume of water required to start the funnel siphon in
a manner as described above for the "M" cells.
As a general rule, the greatest safety factor for timing should be
given to the higher cells. In practice, calibration and timing must be
done more or less simultaneously, starting at the top "M" cell and
working downward. After the metering cells have been calibrated and
one feels that the apparatus is properly timed, the serial diluter should
be intentionally fouled by randomly adding water in order to be certain
that the diluter can correct errors and resume proper operation.
TROUBLE SHOOTING
Experience in the use of the diluter has shown how to correct
minor malfunctions with ease by using "tricks of the trade. " Some
of these do not fit well into the text and are presented here for the
reader's convenience.
1. Symplon: One or more metering cells are not filling or
emptying completely.
Trouble: Dirt in the water or air tubes.
2. Symplon: Last metering cell (No. 6 on Figure 1) fills and
empties but the others remain full.
Trouble: One of the funnels has insufficient water in it and
the water delivery tube is getting air, or there is dirt or a
break in the vacuum venturi or manifold.
3. Symplon: One or more metering cells are delivering water
continuously and other cells are not filling.
Trouble: Needle valve is not closing tightly, or is not re-
maining closed long enough for all cells to empty, or a par-
ticle of dirt caused the valve to malfunction. Correction may
involve adjusting valve seating, providing more water to in-
crease closed period of valve, or improving efficiency of
vacuum venturi.
4. Symplon: Waler rises over water delivery tubes before valve
buckel fills and closes valve.
Trouble: There is loo much waler pressure for exisling ad-
justments or there is dirl in waler tube leading to melering
cell 6. Correclion may involve increasing Ihe dislances A'
(Figure 4), lowering Ihe level of cell 6, increasing Ihe bore
of Ihe waler manifold and inlel lubes lo melering cells, or
lowering curve of waler delivery lube of cell 6.
5. Symplon: Cell 6 fills before Ihe olher cells are full.
Trouble: Cell 6 needs lo be raised slightly.
6. Symplon: Volumes delivered by one or more melering cells
vary.
Trouble: Valve is nol closing properly.
Trouble Shooline 13
-------
10.
Symplon: Diluter is double siphoning.
Trouble: Improper timing.
Symplon: Diluter does nol hold a phase sequence pattern.
Trouble: Timing is improper, one venturi is not operating,
one metering cell does not always empty, or one of the water
volumes (either from metering cell or a transfer volume) is
grossly wrong.
Sympton: A transfer venturi does not work every time.
Trouble: There is air in funnel siphon tube, dirt in water
line, venturi "U "tube is cracked, flare is not completely
filling with air-water mixlure, or end of flare is under water
of next lower "M" cell part of the time.
Sympton: Transfer volumes vary from day to day.
Trouble: Venturi siphon tubes are not held securely.
PERFORMANCE
One may expect no more than 10 percent calculated error (1 to
3 percent for any one concentration) from a diluter if it is calibrated
once every Z to 4 weeks with daily observations as described previously.
One advantage is that the errors tend to be both positive and negative so
that they cancel each other. Table 1 illustrates the usual accuracy ob-
tained from a diluter with no special efforts or precautions. These
data were recorded in a record book during routine testing and do not
represent unusual accuracy nor were they recorded for demonstration
purposes.
Table 1. THE CALCULATED AND POLAROGRAPHICALLY MEA-
SURED CONCENTRATIONS OF ZINC IN mg/liter FROM A
SERIES OF THREE 4-day TESTS. DAILY SAMPLES WERE
COMPOSITED FOR 4 days AND THEN THE ZINC ANALYSES
WERE MADE. WATER FLOW WAS 100/ml/min/concentra-
tion. THERE WERE 10 FISH IN EACH 10-liter TEST
CHAMBER. ALL SOURCES OF ERROR INCLUDING DILUTER
ERROR, ERROR IN THE TOXICANT-METERING APPARATUS,
WATERFLOW VARIATION, REMOVAL BY THE TEST
ORGANISM, AND ANALYTICAL ERROR ARE REFLECTED
IN THESE MEASUREMENTS.
Test 1
Calculated
33.5
17.0
8. ,4
4.2
2. 1
1.0
Measured
34
18.7
10.5
5. 3
2.4
1. 1
Test 2
Calculated
33.4
17.2
8.4
4. 3
2. 1
1. 1
Measured
34.9
18.5
8.5
3.4
2. 1
1.2
Test 3
Calculated
41. 3
20. 4
11.0
5.8
2.9
1. 5
Measured
41.0
20.5
10.9
5. 5
3.0
1.4
14
A SERIAL-DILUTION APPARATUS
GPO 820573-3
-------
The choice of optimum cell size can be made by a convenient rule
of thumb as follows: Determine the flow desired for each concentra-
tion per minute and use "T" and "M" cells that hold four times that
volume. In this way one can obtain a 2:1 dilution ratio by using a
1-minute half-cycle time. The flow can be doubled by decreasing the
half-cycle time to one-half minute, about the maximum practical speed,
and, of course, the flow can be decreased to any desired volume by
simply increasing half-cycle time. One-minute half cycles seem to be
optimum.
Figure 7. A serial-dilution apparatus designed to deliver 300 ml /min of five concentrations
and a control. Several minor changes are used on this diluter.
Performance 15
-------
Some of our dilute rs have run for 6 months with no more than one
or two adjustments during that period. Of course there were occasions
during that period when foreign material or an accident caused a. signi-
ficant error for a short period. Figure 7 illustrates a 5-concentration
diluter designed to deliver up to 300 milliliters per minute.
One may logically expect to build and calibrate one of these di-
luters in a week, but after several are built, this time may be re-
duced to 3 days or even less. Four professional biologists in the
Newtown Laboratory have become proficient with these systems and
are thoroughly convinced of their utility. Most recently -we have been
studying effects of silt in the water, and the diluter appears to be
suited for that work as well.
The experiences of our people with the diluter have been very
similar; after the diluter is built, a short period is needed for final
adjustments and for solving minor problems. Then, as the operation
becomes smooth, satisfaction is felt because technical problems are
minimal and the biological investigation can proceed.
REFERENCE
1. Lemke, A. E. , and D.I. Mount, Some effects of alkyl benzene
sulfonate on the Bluegill, Lepomis macrochirus. Trans. Am.
Fish. Soc., 92(4): 373-378, Oct. 1963.
16 A SERIAL-DILUTION APPARATUS
GPO 82O5732
------- |