United States
 Environmental Protection
 Agency
 Municipal Environmental
 Research Laboratory
 Cincinnati OH 45268
 Research and Development
 EPA-600/S2-84-114   Sept. 1984  ''i
 Project Summary
 Filtration  of  Giardia  Cysts  and
 Other  Substances
 Volume  1:   Diatomaceous  Earth
 Filtration

 Kelly P. Lange, William D. Bellamy, and David W. Hendricks
  How effective  is filtering drinking
 water through diatomaceous earth to
 remove Giardia lamblia cysts, total
 coliform bacteria, standard plate count
 bacteria, turbidity, and particles? We
 evaluated the process for a range of
 operating  conditions  and  simulated
 ambient conditions. Hydraulic loading
 rates imposed were 2.44,  4.88, and
 9.76 m/hr (1,2. and 4gpm/ft2). Seven
 grades of diatomaceous earth were
 used. Temperatures were from 5° to
 19°C; concentrations of Giardia cysts
 ranged from 50 to 5000 cysts/L; and
 bacteria densities were varied from 100
 to 10,000/100 mL.
  The results of this study showed that
 diatomaceous earth  filtration  is  an
 effective process for water treatment.
 Giardia cyst removals were greater than
 99.9 percent for all grades of diatoma-
 ceous earth tested, for hydraulic loading
 rates of 2.44 to 9.76 m/hr, and for all
 temperatures tested. Percent reduction
 in total coliform  bacteria, standard
 plate count bacteria, and turbidity are
 influenced strongly by the grade of dia-
 tomaceous earth  used. The coarsest
 grades of  diatomaceous  earth
 recommended for water  treatment
 (e.g., C-545®)* will  remove  greater
 than 99.9 percent of Giardia cysts, 95
 percent of cyst-sized particles, 20 to 35
 percent of coliform bacteria, 40 to 70
 percent of heterotrophic bacteria, and
 12 to 16 percent of the turbidity from
 Horsetooth Reservoir water. The use of
•Mention of trade names or commercial products
 does not constitute endorsement or recommenda-
 tion for use.
 the finest grade of diatomaceous earth
 (i.e., Filter-Cel®), or alum coating on the
 coarse  grades,  will increase the
 effectiveness of the process, resulting
 in 99.9 percent removals of bacteria
 and 98 percent removals of turbidity.
  This Project Summary was developed
 by  EPA's  Municipal Environmental
 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
  This study was conducted at Colorado
 State  University  under  a  cooperative
 agreement with the U.S. Environmental
 Protection Agency (EPA) to determine the
 effectiveness  of  diatomaceous  earth
 filtration for removal of Giardia cysts. At
 the same time, removals of turbidity, total
 coliform  bacteria, standard plate count
 bacteria, and particles were determined.
 Operating conditions examined included
 the   grade  of  diatomaceous  earth,
 hydraulic  loading  rates,  influent
 concentrations of bacteria and Giardia
 cysts,  head loss, run time, temperature,
 and   the  use  of  alum-coated
 diatomaceous earth.
  Giardia lamblia is a  protozoan
prevalent in  the  clear, cool waters
characteristic  of the Rocky Mountain
region. This organism causes giardiasis,
a harmful but nonfatal intestinal disease.
Many communities use water from these
Rocky  Mountain  streams,  which  are

-------
considered pristine pure because  they
look aesthetically pleasing and will meet
the  1-NTU  turbidity  water  quality
standard. How to treat these waters has
become an important concern over the
last few years as outbreaks of giardiasis
have occurred. Economical and effective
filtration systems are  needed to remove
Giardia cysts. Designs  appropriate for
small  water  systems are  particularly
needed.
  Diatomaceous  earth   filtration  was
introduced in 1942 as a technology for
water  treatment.   The  process  was
adopted by the U.S. Army for field use in
1944 after being  shown effective for
removal  of  Endamoeba  histolytica
cysts.The basic principles of the process
were outlined in the 1940's and 1950's,
and further studies were made in the
1960's.
  The diatomaceous earth filtration pro-
cess consists of three basic operations:
(1) precoating, (2) filtering, and (3) clean-
ing. In precoating, an initial filter  cake
consisting  of  a 3- to  5-mm layer  of
powder-sized diatomaceous earth  filter
medium  is   applied  to   a  support
membrane called a septum. The cake is
applied by circulating  flow from  the
precoat tank  through  the filter, causing
the slurried  diatomaceous earth  to  be
deposited   on  the  filter  septum.  In
filtering.the second step of the process,
raw water  combined  with  bodyfeed
passes  through  the  filter  cake.  The
bodyfeed  consists of a filter  medium
slurry metered into the raw water stream
during filtration. The continuous addition
of  bodyfeed  maintains the filter  cake
permeability.   In the  third  operation,
cleaning, the filter cake is removed from
the septum and  discarded.
  Operating  parameters  for  diatoma-
ceous earth filtration include the grade of
diatomaceous   earth   (a   commercial
designation of particle size),  hydraulic
loading rate, precoat thickness, bodyfeed
concentration,  terminal  headloss, and
run  time.  Chemical  coating  of  the
diatomaceous earth may be used under
some circumstances to improve removal
effectiveness.
  The  protozoan Giardia lamblia  is of
interest because itcausesgiardiasis. This
organism   has  been  identified  as  a
pathogen  only  recently. Giardiasis  is
considered  a  serious problem  in
mountainous and forested regions of the
United  States  where the  organism is
endemic; the Giardia cysts shed by dogs,
humans,  and   animals, such  as  the
beaver, are believed to be of the Giardia
lamblia species.
  The Giardia lamblia trophozoite (Figure
1 a) can reside in the intestine of a variety
of warm-blooded   animals.  The  cyst
(Figure  1b)  is the  form  shed  and
transmitted. An infected person  may
shed up to 900 million cysts per day. The
cyst form of the organism is hardy and
may remain viable  for a long period (2
months, for example), particularly in cold
water. Infection is  caused by ingestion;
an infective dose may be from 1  to  10
cysts  and the incubation period is 1 to 2
weeks. A surface water supply source is a
vehicle for  cyst transmission.

Materials and Methods
Design of Tests
  The objective of the experimental pro-
gram  was  to  evaluate  the  removal
effectiveness  of   diatomaceous  earth
filtration for the  dependent  variables
(Giardia cysts,  total coliform bacteria,
standard plate count bacteria, turbidity,
particle counts, and  headloss)  as  a
function  of the independent  variables
(grade of diatomaceous earth,  run  time,
headloss, hydraulic loading rate, temper-
ature, influent coliform  concentration,
and alum coating).  This was achieved by
changing a given  independent variable
over a range of magnitudes and observing
the effect on the filtration performance
(i.e., the dependent variables).
  Tests were terminated either at the
predetermined  head  loss  across the
septum of 40 psi, or after a given period of
operating time. The precoat application
was established at  1 kg/m2(0.2lb/ft2)for
all tests. The bodyfeed concentration was
established  by   using  successive
concentrations until a linear headloss
versus time relationship was found.
                           5-15 fjrn
 9-21
                Flagella
                                          The pilot plant was used mostly in the  m
                                        laboratory,  but  final  confirming  tests  ~
                                        were conducted in the field. The filter, the
                                        main element of the system (Figure 2),
                                        consisted of a 1 -ft2 septum enclosed in a
                                        pressure  housing. The septum used in
                                        this work was stainless steel wire mesh
                                        of 110 x 24 wires per in.2. The operations
                                        were controlled  by the ancillary pumps,
                                        valves, and gauges.

                                        Experimental Procedures
                                           Giardia testing began with the process-
                                         ing of Giardia   cysts  from  dog  fecal
                                        samples.  The processing consisted of
                                        adding  the  infected feces  to distilled
                                        water,  straining the  feces  and  then
                                        making a count of cysts in the  fecal
                                        concentrate.
                                          A known  concentration of the Giardia
                                        concentrate was then added to a 1400-L
                                        filter feed tank. This tank was a modified
                                        milk cooler that could be maintained at 2
                                        to 15 *1°C. The filter feedtank wasfilled
                                        with  Horsetooth Reservoir   water  and
                                        cooled before the addition of the cysts.
                                        Primary settled sewage was added also to
                                        increase  the  concentration  of  total
                                        coliform bacteria.
                                          Preparation for a test run began with
                                        the precoat  step. After precoating,  10  ^
                                        mg/L chlorine was  added for dismfec-  ^
                                        tion, and the recycle of precoat water was
                                        continued for 10 min  The chlorine was
                                        purged by operating in the filtering mode
                                        for 30 min  Sampling  was started after
                                        the 30-min washout period.
                                          Samples were obtained from the filter
                                        feed tank and from the effluent side of the
                                        filter for  measurements of turbidity,
                                        particle counts,  total coliform bacteria,
                                        standard  plate   count  bacteria,   and
                                                    Nuclei
                                                   Claw-like
                                                   Median Bodies
                                                                      8-12
Figure 1 .
            Sketches of aj trophozoite and b) cyst stages of Giardia lambda (Jakubowski and II
            Hoff, 1979).                                                          ^

-------
Giardia  cysts.  Grab  samples  were
collected  for  all   parameters  except
Giardia  cysts.  Other  measurements
included elapsed time from beginning of
run, headloss, hydraulic loading rate, and
water temperature.
  The  Giardia cyst  sampling  technique
used a 142-mm-diameter polycarbonate
membrane filter with a 5- (im pore size to
remove and concentrate the cysts from
the sampled water. After the sample was
concentrated, the membrane  filter was
washed  and  the  wash  water  was
analyzed  for   cysts   by   microscopic
counting. The influent water in the fitter
feed tank and the  diatomaceous earth
effluent were  both  sampled for Giardia
cysts in this manner. The influent sample
volume ranged between 2  and 10 L, and
the  effluent  sample  volume   ranged
between 86 and 174/L
  Field tests to verify laboratory findings
were conducted  in April and May 1983.
The April test  used  raw water from the
Cache La Poudre River, and the May tests
used  water from  Straight Creek and
Laskey  Creek at the  Dillon,  Colorado
water treatment  plant.

Leak  Testing
  During the initial phases of experimen-
tation, the need to determine whether a
leak was present in the filter septum or its
manifold  became  apparent.  Thus  a
technique was developed to test for leaks.
First the filter  was precoated  with 2
kg/m2  of the finest grade of diatoma-
ceous earth, Filter Cel.  This grade  was
determined to be capable of removing
100 percent of the applied coliform bac-
teria. Then the  filter was operated at 1
gpm/ft2 for 1  hr with a  high influent
coliform concentration. If any coliforms
were  detected  in   the  effluent, the
equipment  was  assumed to have a
mechanical problem  resulting in a leak,
and the problem was corrected.  This
technique was used as a quality control
measure throughout all testing.

Results
  The diatomaceous earth  filtration pro-
cess was found to be effective for remov-
ing Giardia cysts  under virtually all
operating  conditions tested  No  cysts
were detected in the filter effluent  when
normal water treatment practices were
simulated. Note, however, that removals
of bacteria, turbidity, and particles  in the
6.35-A«m to 12.67-/ym  size range were
functionally dependent on: (1) grade of
diatomaceous earth, (2) use of chemicals,
(3) hydraulic loading rate, and (4) influent
concentrations.
    Overflow
      Cup
   Effluent
 Overall Process Effectiveness
   Giardia cyst  removals were greater
 than 99.9 percent, regardless of grade of
 diatomaceous earth, filtration rate, tem-
 perature, duration of test, or influent
 concentration of  Giardia  cysts  (when
 cysts  are fewer than  10,000/L).  The
 single breakthrough occurred at  a high
 influent concentration of 33,600cysts/L.
   Figure 3 illustrates the uniformly high
 removals of Giardia cysts. Testing was
 done only for the water treatment grades,
 since removals would have been at least
 as much for the finer grades.
   The grade of diatomaceous earth was
 not a factor in removal of Giardia cysts,
 even  with  grades C-545 and C-535.
 These grades create a filter cake with
 reported median pore sizes of 17 and 13
 fjm, respectively, which are larger than
 Giardia cysts. However, many pores were
 apparently  smaller than the cysts  and
 blocked their passage.
   Removal of coliform bacteria,  standard
 plate  count   bacteria  and   turbidity
 approached 100 percent for the smallest
 grades of diatomaceous earth and fell
 below 40 percent for the coarsest grades
 (Figure 3).  Removals of total coliforms
 and standard plate count bacteria would
 follow this trend regardless of the water
 source.  Turbidity  removal,  however,

Recycle
                             Manometer
                               Spray Jet
                                Nozzles
                        Septum ~
                        Drain

                                                                 Filter
                                                              Flowmeter
                                             ,  Jet
                                              Flow
                                   1 w
                                   Filler
                                   Unit
                            Sample
                              Tap
                   Membrane
                  Sample Filter
                                 Bodyfeed
                                 Flowmeter
       • Bodyfeed
    P }   Pump

          Pressure
           Relief
            Valve
                    Dampener
Figure 2.   Layout of diatomaceous earth filtration pilot plant.

-------
would depend on the size of the particles
making   up  the  turbidity;   improved
removal  would  result  if the turbidity
consisted of larger particles.

Effects of Operating Conditions
on Removals
  Operating conditionsexammed include
grade of diatomaceous  earth, hydraulic
loading rate, Giardia cyst concentration,
bacteria   concentration,  temperature,
duration  of  filtration  run,  and  alum
coating of the diatomaceous  earth. The
influence of each condition on removals
of turbidity,  bacteria, Giardia cysts, and
particles  is   reported in the  following
paragraphs.

Grade of Diatomaceous  Earth
  As  mentioned   earlier,  removals  of
Giardia   cysts  are greater than  99.9
percent for all earth tested, but removals
of turbidity, standard plate count bacteria,
and total cohform bacteria are strongly
influenced   by   grade   Removals  of
particles in  the 6.35- to 12.67-//m size
range were  uniformly  high  (87  to  94
percent) for the water treatment grades of
diatomaceous earth, thus indicating no
correlation of particle removal to grade.

Hydraulic  Loading Rate
  Some scatter  occurs in the data on the
effects of hydraulic loading on removals
of  particles,  standard  plate   count
bacteria,  coliform  bacteria, and turbidity;
but the  trends are  toward  declining
removals  with  increasing   hydraulic
loading  rate. Percent removals of total
coliform bacteria were affected the most,
standard  plate  count bacteria declined
nominally,  and particles and turbidity
were  affected only moderately. The fine
clays constituting most of the raw water
turbidity  passed readily through the C-
503  and C-545  diatomaceous  earth
grades at all hydraulic loading  rates.
Hydraulic loading  rate had no detectable
influence on removal of Giardia  cysts,
since all but  one test resulted in complete
removal  of the  influent  cysts. Because
the cysts are larger than  some of the
pores in  the  filter  cake  (which vary
statistically), they will be blocked by some
pore as they are  convected by the flow
within the cake.

Giardia Cyst  Concentrations
  No  discernible  relationship existed
between  the Giardia cyst influent con-
centration  and cyst  removal.  Influent
Giardia cyst concentrations ranged from
500 to 10,000  cysts/L,  with one excep-
O
I
6
•Q
«J
to
10
CO
13

10
•6
-5

    700
     90
     80
     70
60
50
     40
     30
     20
     10
                      
-------
grades. This phenomenon is reasonable,
since the finer grades can  completely
strain coliform bacteria.
  Turbidity did  not vary enough in the
waters tested to  determine  whether a
relationship existed  between  influent
levels of bacteria  and their removal. For
the  laboratory  tests with  Horsetooth
Reservoir water,  the influent turbidity
levels ranged only between 4.5 and 5.4
NTU.
  Most turbidity in Horsetooth water was
caused by particles  smaller than 1  //m.
Consequently, a 1-NTU turbidity standard
could be attained only with the finest
grade of diatomaceous earth or by using
alum-coated diatomaceous earth. Remov-
als of coliform  bacteria were therefore
higher, since they were larger than much
of the turbidity. The small turbidity parti-
cles found in Horsetooth Reservoir are
not  necessarily  characteristic  of  all
waters,  since many sources have been
treated to meet a 1-NTU standard with
water treatment grades of diatomaceous
earth and without chemicals.


Temperature
  Water temperature did not appear  to
affect any of the parameters measured.
Because   the  removal  process   in
diatomaceous earth filtration is physical,
slightly   poorer   removals  might  be
expected at lower temperatures because
shear forces are higher. This possible
effect was not noticeable, however, since
it was masked by the significant changes
in  removal caused by  variations  in
influent concentration and flow rate.

Alum Coating
  Alum-coated  diatomaceous earth
removed significantly more total coliform
bacteria, standard plate count bacteria,
and  turbidity  than  did the uncoated
diatomaceous  earth. Coarse grades  of
diatomaceous earth were used (e.g., C-
545 and C-503), and both the precoat and
bodyfeed were coated  while in slurry
form. Removals of total coliform bacteria
ranged from 96 to 99.9 percent compared
with 30 to 70 percent without alum
coating.  Removals  of  standard  plate
count bacteria ranged from  79 to 99.5
percent  compared with  38 to 56 percent
without  alum coating. Turbidity removals
ranged from 66 to 99 percent compared
with 11  to 17 percent without alum
coating.  These results  demonstrate  a
marked  improvement in treatment with
alum  coating  and  illustrate  how
diatomaceous  earth filtration  can  be
applied in otherwise marginal situations.
  Note, however,  that not  all water
sources can be  treated  to  a  1-NTU
standard   with  the  normal  water
treatment grades of diatomaceous earth.
To obtain a 1 -NTU effluent for these spe-
cial  cases,  the  advantages  of alum
coating must be weighed against using a
finer grade of diatomaceous earth.  Both
techniques will reach the desired goal,
but only pilot plant studies and economic
considerations  will  determine  which
approach should be taken.

Field Testing
  Field tests were conducted to confirm
the  laboratory  results.  Water  was
obtained from the Cache La Poudre River
and from  the  raw water  intake at the
Dillon  water treatment plant.  Turbidity
conditions were about 4 NTU for tests
conducted April 17,1983, at the Cache La
Poudre River, and about 0.6 NTU for the
May 1983 tests at Dillon. Despite the use
of  these  different  water sources,  no
Giardia cysts   were  detected   in   any
filtered water  samples.  Removals  of
turbidity, total coliform bacteria, and total
plate count bacteria were all consistent
with laboratory results.


Conclusions
  1. Diatomaceous  earth filtration  is
    virtually 100  percent  effective in
    Giardia cyst removal for all grades of
    diatomaceous earth over  a wide
    range of conditions.

  2. Grade of diatomaceous earth is the
    most important factor in the removal
    of bacteria and turbidity. Removals
    effected by coarse and fine grades
    were,   respectively,   17   and  98
    percent for turbidity, 28 and  99.9
    percent for coliform bacteria, and 38
    and 99.8 percent for standard plate
    count bacteria.

  3. Increasing the  hydraulic  loading
    rate causes a decrease in removals
    of  bacteria and  turbidity  for  the
    water  treatment  grades   of
    diatomaceous earth. The effect was
    strongest  for coliform bacteria and
    weakest  for  turbidity.  Hydraulic
    loading rate showed no effect on the
    removal of Giardia cysts.

  4. Water temperature did not influ-
    ence  the effectiveness of diatoma-
    ceous earth  filtration, as  demon-
    strated by testing over the range of
    3.5° to 15°C. The results are not,
    however,  conclusive.
  5. Bacteria  removal decreased  with
     increased influent concentrations of
     bacteria, especially for the coarser
     grades of diatomaceous  earth. A
     three-log increase in influent coli-
     forms reduced removals from 77 to
     39 percent for C-545; but for C-512,
     a  two-log  increase in  conforms
     reduced removals only from 96 to 92
     percent.

  6. Bacteria removals decreased slight-
     ly with increasing filtration time--
     fro m 87 to 79 percent in 5.5 hr for C-
     503  and from  >99.98  to 99.92
     percent for Standard Super-Cel.

  7. Alum-coated  diatomaceous  earth
     filtration  removed significantly
     more  turbidity and  bacteria than
     diatomaceous earth filtration  with
     no alum.  The use of alum coating
     increased removals  from 17 to 99
     percent for turbidity, from 30 to 96
     percent for total  coliform  bacteria,
     and from  56 to 99.5 percent for
     standard plate count bacteria.

  8. Increased removals of turbidity and
     bacteria can be accomplished either
     by chemically coating the diatoma-
     ceous earth or by using a smaller
     grade.

  9. Field testing with two different raw
     waters yielded the same results as
     laboratory tests.

 10. Pilot-plant testing should be done
     before  implementing any full-scale
     application of diatomaceous earth
     filtration. Applicability,  design
     criteria, and  operating conditions
     cannot be determined without pilot
     tests.

 11. Periodically, a diatomaceous earth
     filtration system should be checked
     for leaks by applying Filter-Cel and
     then filtering  a coliform-contamin-
     ated water. In production of potable
     water, this should be done only as a
     part  of  a  routine  performance
     evaluation program in which careful
     controls are set up to ensure that a
     cross connection is not possible.

  The  full  report was  submitted  in
fulfillment of Cooperative Agreement No.
CR808650-02  by   Colorado  State
University under the sponsorship of the
U.S. Environmental Protection Agency.
                                                                                              *USGPO:  1984-759-102-10668

-------
      Kelly P. Lange. William D. Bellamy, and Davis W. Hendricks are with Colorado
        State University, Fort Collins, CO 80523.
      Gary S. Logsdon is the EPA Project Officer (see below).
      The complete report, entitled "Filtration of Giardia Cysts and Other Substances:
        Volume 1: Diatomaceous  Earth Filtration," (Order No. PB 84-212 703; Cost:
        $19.00, 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:
             Municipal Environmental Research Laboratory
             U.S. Environmental Protection Agency
             Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
     BULK RATE
POSTAGE & FEES PAI
        EPA
   PERMIT No. G-35
Official Business
Penalty for Private Use $300
                                                                                       00
                                                                                Of-  —IOO
                                                                                         C5
                                                                                        >«
                                                                                         o

-------