ON1'/
United States
Environmental Protection
Agency
Water Engineering Research
Laboratory
Cincinnati, OH 45268
Research and Development
 EPA/600/S2-85/126  Dec. 1985
Project  Summary
Evaluation  of Erosion
Feed  Chlorinators
Kenneth E. Olson
  Erosion chlorinators were evaluated
to determine their reliability in delivering
a constant chlorine dose for disinfecting
potable water sources. Several erosion
chlorinators are on the market, but only
one has  been  approved  for use with
potable water systems in the United
States.
  The erosion chlorinator was shown to
provide unstable dose rates when oper-
ated in a continuous-flow mode. Inter-
mittent-flow operation provided a more
stable dose rate. But the greatest degree
of dose stability resulted from a flow
rate  of 40 gpm  (150  L/min)  with
operating periods of 10 min on and 10
min off. The use of erosion chlorinators
should therefore be limited to intermit-
tent operation. Additional studies are
needed to determine the behavior of the
tablets used in the chlorination process.
  This Project Summary was developed
by EPA's Water Engineering Research
Laboratory. Cincinnati. OH. to announce
key findings of the research project that
is fully documented in a separate report
of the same title (see Project Report
ordering information at back).

Introduction
  Disinfection is an essential barrier to
the transmission of pathogens in a potable
water system. Of the disinfection agents
available,  chlorine  is the most widely
used. This project  was undertaken to
evaluate the reliability of erosion chlori-
nators for delivering a constant chlorine
dose under determined conditions and to
evaluate their overall performance and
operation  in the laboratory and in the
field.
  Erosion feed chlorinators use pressed
chlorine tablets that are eroded (or dis-
solved) as water passes over their surface.
The unit generally consists of a canister
that stores a supply of tablets and posi-
tions them in a moving stream of water
and a contact chamber that provides an
interface between the tablets and water.
Adjustments of the chlorine dose rate are
made by changing the tablet surface area
immersed in the water stream. Presently,
only one erosion feed chlorinator  is
marketed for potable water systems—the
Water Sure* chlorinator (in two models),
manufactured by World Water Resources,
Rockville, Maryland.
  The erosion feed chlorinator provides
several advantages over gas chlorinators
and hypochlorinators. The use of tablets
reduces operator exposure to caustic
chlorine dust. Tablets are easier to handle
and store than are gas or powder. Safety
equipment is not required, and the opera-
tion of the chlorinator does not require as
much technical knowledge. However, the
capabilities of the erosion feed chlorinator
needed to be determined.

Test Procedures
  Testing was  conducted  at the U.S.
Department of Agriculture Forest Service
Equipment  Development  Center, San
Dimas, California, in two phases: (1) con-
tinuous flow and (2) intermittent flow.
  Continuous-flow testing was conducted
with  two models of the  Water  Sure
chlorinator  in  modified installations.
Model 101 was installed to ensure con-
tinuous flow. The Model 050 installation
was truncated, with only the chlorinator
head installed in the test stand. Fourflow
rates (7,15, 25, and 40 gpm) (26, 57,95,
and 150 L/min) and two chlorine dose
'Mention of trade names or commercial products
 does not constitute endorsement or recommenda-
 tion for use.

-------
rates  (1  and  5 mg/L) were  used  to
evaluate performance.
  Two flow/dose rate combinations were
tested simultaneously for 4 consecutive
days.  Chlorine analyses were performed
at alternating  10-min  intervals.  The
continuous-flow testing was performed
to obtain performance data, even though
the manufacturer does not recommend
continuous flow operation.
  Intermittent flow testing was conducted
with the Model 050 chlorinator. A sole-
noid valve was installed upstream at the
chlorinator to interrupt the water flow for
desired intervals (10, 30,  and  60 min).
The desired flow interval, or on cycle, was
determined at the beginning of this test
phase. Free  chlorine  residuals  were
measured with a  continuous  ampero-
metric analyzer and recorded on a strip
chart. Grab samples obtained from the
test stand were a nalyzed with an ampero-
metric titrator and  compared with the
continuous monitor reading at the time of
sampling.  The accuracy of the instru-
ments was checked at the beginning of
the test and periodically thereafter with
known solutions of calcium hydrochlorite
that were  analyzed with the titrator and
run through the analyzer. Method 408C
from Standard Methods (1) was used in
the analysis procedures.

Results and Discussion

Continuous-Flow Tests
  Results  of the continuous-flow tests
were  not encouraging. When first oper-
ated with  fresh tablets, the chlorinator
achieved a stable dose rate within 4  hr
and could be adjusted as desired. After 2
days,  however, chlorine concentrations
fluctuated over a wide span and became
uncontrollable, especially at the high rate.
Tablet bridging (or jamming) in the
canister occurred frequently. Examination
of the chlorinator revealed condensation
inside the canister on the lid, walls, and
tablets. The initial theory was that mois-
ture caused premature release of the
chlorine, which entered the water stream
at an uncontrolled  rate.  Operating the
chlorinator without the lid provided only
limited correction of the problem. Con-
sistent dose rates  were achieved for a
week, and tablet bridging was signifi-
cantly reduced; but fluctuations recurred,
and the  dose  settings again became
uncontrollable.

Presence of Moisture—
  During  testing,  visual  observations
were  made of the top layer of tablets in
the canister each time fresh ones were
added. The amount of moisture present
on  these  tablets varied from  slight to
heavy, depending on the flow rate and the
humidity.  At a flow rate of 7  gpm  (26
L/min) and low humidity, the top layer of
tablets showed a slight discoloration but
was dry to the touch. This condition was
detected after a day's operation and could
also have been  caused by the higher
humidities at night. At a flow rate of 40
gpm  and  low humidity,  discoloration
occurred within 3 to 5 hr, and dampness
could be detected within 5 to 8 hr. As the
humidity increased, the amount of mois-
ture present on  the  tablets increased
significantly. During periods of fog or
rain, water droplets 1/16 to 1/8 in. in
diameter were discovered within 2 to 4 hr
after  fresh  tablets were added to  the
canister.
  Visual inspections of the tablets layered
throughout the canister showed that the
bottom layer of tablets (1 to 1.5 in.) was
saturated. Tablet erosion occurred within
this layer.  Above  this erosion zone, mois-
ture present on the tablets was approxi-
mately the same as that on the top layer of
tablets. While there was an increase in
moisture with depth, the difference did
not appear greatly different.
Tablet Bridging—
  Moisture present in the canister was
absorbed by the tablets. This absorption
in turn caused tablet bridging and may
have  contributed  to the  formation  of
calcium deposits  on the tablets, which
interfered with the release of chlorine.
During testing, tablet bridging was gen-
erally  associated  with high  moisture
levels. Absorption of moisture  by the
tablets created a paste-like coating on the
tablet surface and also  caused them to
swell. A slightly adhesive  quality of the
paste, the increased tablet size, and the
inward taper of the canister caused the
tablets to bond together. When still wet,
the bridge was easily broken by tapping
the canister; but if drying  occurred, the
tablets were difficult to break. Evaluation
of the Water Sure showed that the tapered
canister contributed to bridging. In chlo-
rinators with untapered canisters, tablet
bridging was reported  to occur when
swelling caused the tablet to stick in the
canister. The general result of a tablet
bridge was an interruption of chlorination
when the tablet supply below the bridge
was depleted.  The  saturated layer  of
tablets  in  the erosion zone  was not
affected by bridging. Along with  service
interruptions, the occurrence of bridges  A
was detected during testing by tapping  *
the canister and noting any significant
settling of the tablets. A settling of about
0.5 in. was generally due to tablet shifting.
Bridges resulted in a settlement of 1 in. or
more,  usually 1.5 to 2 in.  Undetected
bridges may also have occurred and may
have been broken by the weight  of the
tablets above the  bridge or  by  some
disturbance to the chlorinator.
Calcium Deposits on the
Tablet Surface—
  Although tablet bridging was a definite
interference, it was not significant in the
overall performance of the chlorinators.
Of more importance was the apparent
deposition of calcium on the tablet sur-
face.  Observations of the chlorinator
revealed that only a portion  of the total
water flow entered the contact chamber.
The amount depended on the depth of the
canister in the water stream, which varied
according to the flow rate and canister
adjustment. Conditions within the can-
ister were conducive to the formation of a
calcium coating, with a calcium concen-
tration at or above the saturation level.
Evaluations of  the chlorinator,  events
occurring during testing, and the envi-
ronment of the tablets supported this  m
theory.  However, tablet analysis could  ~
not be performed to confirm this behavior.
  As the calciu m coating covers the tablet
surface, it reduces the exposure of hypo-
chlorite ions to the  water flow in the
canister. Although the coating is not
watertight, it acts as a barrier, reducing
the amount of water in contact with the
calcium hypochlorite surface and restric-
ting the flow of hypochlorite ions from
that surface. The continued formation of
this  barrier would result in  decreased
chlorine concentrations  in the  water
stream. A periodic sloughing of portions
of the  barrier would temporarily expose
fresh hypochlorite surfaces and cause
fluctuating chlorine concentrations.


Disturbances to the Chlorinator—
  Disturbances to the chlorinator resulted
in increased chlorine doses. The  magni-
tude and duration of these increases
depended on the severity of the disturb-
ance. The addition of tablets, sudden flow
rate changes, canister adjustment, and
movement of  the  chlorinator  caused
chlorine concentrations to increase. The
most common disturbance during testing
was rapping the canister to settle tablets
or to release a tablet bridge.

-------
Intermittent-Flow Tests
  Intermittent-flow testing revealed im-
proved chlorinator performance, primarily
because of the interruption of water flow
through the unit. The ideal conditions for
this test phase and for overall operation of
the chlorinator were found to be a flow
rate  of 40 gpm (150 L/min) with an
operating  period of 10-min on, and 10-
min off. Increasing both the on and off
cycles resulted in greater dose fluctua-
tions, as did reducing the flow rate. The
various weather conditions encountered
during testing revealedthat dose stability
was heavily influenced by humidity levels.
Erratic chlorine concentrations were ob-
served only during periods of high humid-
ity.
Tablet Moisture—
  Visual inspection  of tablets  in the
canister did not show that moisture levels
changed substantially  with flow rates.
With similar humidity conditions, tablet
moisture  appeared  constant. Surface
textures, observed visually and by touch,
retained the same graininess and pasty
feeling. The depth of the moisture, de-
tected by scraping the tablet surface, did
not appear to change at the various flow
rates.  Increased  moisture levels were
detected only during periods of increased
humidity. As moisture  levels increased,
the tablet surface became grainy and
paste-1 ike, but not to the degree observed
in continuous-flow testing. Only during
periods of heavy rain  or fog  did water
droplets (1/16 to 1/8 in. in diameter)
occur. The absorption of tablet moisture
was primarily governed  by humidity
levels.  Once tablets absorbed moisture,
however,  they tended  to retain  it. The
anticipated drying of tablets during the
different periods of no-flow (off cycle) was
not observed to be significant. Although
the tablets were exposed to air for up to
60 min, appreciable improvement in the
appearance of the tablet surface was not
detected.  The only  improvement noted
was the disappearance of water droplets
from the top layer of tablets, which could
have been due to their absorption by the
tablet.  The lack of drying during  the off
cycles  was attributed to the lack of air
flow through the chlorinator. Depending
on the humidity, drying occurred only
when the'test stand was shut down for 1
day or  more.  However, this  procedure
occasionally resulted in the  bonding  of
tablets. Overall, less tablet moisture was
observed during  intermittent  flow than
during  continuous  flow. Intermittent
operation reduced moisture on the tab-
lets.
  Interrupting the water flow through the
chlorinator reduced the exposure of tab-
lets to moisture by allowing the erosion
zone to drain and dry, at least partially.
The periodic draining of the erosion zone
reduces the migraton of water into the
upper tablet level by reducing the water
tablet contact time. Also, any free water
on  the tablet  surface  just above  the
erosion zone would have  drained and
been exposed to air. Though tablet drying
in the  upper  level of tablets was  not
apparent, it did occur in the erosion zone.
After a 10-min off cycle, the tablets in the
latter zoned to 2 in.) had a loose, granular
texture that  was  easily dislodged and
could not be measured. Regardless of the
flow rate, the same texture was observed.
Tablet drying in the erosion zone was
attributed to this texture, which permitted
a greater surface area to be exposed to a
larger volume of air. In the upper tablet
level, air was present  only in the voids
between the tablets.

Chlorinator Turbulence—
  Observations of the chlorinator's oper-
ation showed an initial turbulence at the
start  of each on  cycle,  which quickly
subsided to the quiescent flows observed
previously. Fluctuating water surfaces
and splashing accompanied the entry of
water into the contact chamber. Although
the event could not be  observed with
tablets  in the canister, the initial turbu-
lence dislodged the loose tablet surfaces.
This fact was indicated by the presence of
tablet residue found in the contact cham-
ber, the mix tank, and the carbon filter.
Although residue was found during con-
tinuous-flow testing, its frequency was
less than that during this  test phase.
Tablet  material  may  also  have been
dislodged by jarring of the canister at the
beginning of each flow cycle. A vibration
of the canister was observed at 40 gpm
(150 L/min), but  its  strength was  not
great and decreased with flow so that it
was not detected at 7 gpm (26 L/min).
  Turbulence at the beginning of each on
cycle tended to improve the performance
of the chlorinator by removing loose tablet
material and providing relatively uniform
calcium hypochlorite surface in the ero-
sion zone. Although  turbulence was
important to achieving dose stability, its
impact was diminished by humidity.

Dry Weather  Operation
  Greater variations in chlorine concen-
trations generally occurred during periods
of high humidity. Data patterns for dry
weather operation are interrupted, but
data are sufficient to indicate a relative
stability of chlorine doses during intermit-
tent operation of the chlorinator in dry
periods.
  Comparisons of data from periods of
low humidity indicate that the drying of
the erosion zone results in  expanded
fluctuation ranges for chlorine  concen-
trations. Decreased flows with a 10-min
off  cycle did not result in a significant
degradation of dose stability. The data
from these periods showed stable average
concentrations with  deviations  ranging
from 0.4 to 0.8 mg/L (Figure 1). Greater
fluctuations (about 1 to 1.5 mg/L) were
also seen during periods of high humidity
when the off cycle was increased (Figure
2).
  Increasing the off cycle duration re-
sulted in drying of  the  erosion zone.
Although this  increase was expected to
be a benefit, it proved otherwise. Since
the continued  exposure to moisture fol-
lowed by drying increased the amount of
loose tablet material, the onset of the next
on cycle resulted in increased exposure to
calcium hypochlorite  surfaces.  The
amount of loose material present depend-
ed on the amount removed during the
previous on cycle and the rates of mois-
ture absorption and evaporation. Fluctu-
ating amounts of loose tablet material
from cyclic buildup and removal would
result in fluctuating chlorine concentra-
tions.


Wet Weather Operation
  As humidity levels increased, a number
of related actions occurred to alter the
impact of the off cycle. Tablets in the
upper level of the canister were softened
to various  degrees,  depending on  the
humidity. These tablets had varying initial
erosion rates. The amount of moisture
absorbed during the  off cycle increased
and the evaporation rate slowed. The
impact of this depended on the flow rate.
At 40 gpm (150 L/min), the undulating
chlorine concentrations were attributed
to changes in  humidity levels  during
testing. The changes in chlorine concen-
tration  between  successive on cycles
were reduced  by the turbulence associ-
ated with this flow rate. With reduced
flow rates, the dose became erratic. The
data patterns of wet weather operation
resemble those seen in continuous-flow
testing. These results were attributed to
the  interference of calcium deposits and
the  same moisture  actions seen  with
continuous flow. The presence  of  high

-------
    6-
 !
 .§'
tio

*>
 §
 c
 o
 o
 .c
ee Ch
             Mean
                                                                      Mean
6 a.m.
                 12 noon
6p.m.
                                            12 mid.
                           6a.m.
                                                                   12 noon
Figure 1 -    Free chlorine concentrations for intermittent flow at 40 gpm with 10 min on and 10
           min off.
 O
 
-------

-------
     Kenneth E. Olson is with U.S. Department of Agriculture Forest Service, San
       Dimas,CA91773.
     Gary S. Logtdon is the EPA Project Officer (see below).
     The complete report, entitled "Evaluation of Erosion Feed Chlorinators," (Order
       No. PB 86-118 882/AS; Cost: $9.95, subject to change) will be available only
       from:
             National Technical Information Service
             5285 Port Royal Road
             Springfield. VA 22161
             Telephone: 703-487-4650
     The EPA Project Officer can be contacted at:
             Water Engineering Research Laboratory
             U.S. Environmental Protection Agency
             Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300

EPA/600/S2-85/126
              0000329   PS
                                        It.    60604

-------