WATER POLLUTION CONTROL RESEARCH SERIES • 17020 GDN 07/71
         Improving
Granular Carbon Treatment
 U.S. ENVIRONMENTAL PROTECTION AGENCY

-------
          WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes the
results and progress in the control and abatement of pollution
in our Nation's waters.  They provide a central source of
information on the research, development and demonstration
activities in the Environmental Protection Agency, through
inhouse research and grants and contracts with Federal, State,
and local agencies, research institutions, and industrial
organizations.

Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications Branch
(Water), Research Information Division, R&M, Environmental
Protection Agency, Washington, D.C. 20460.

-------
              IMPROVING GRANULAR CARBON TREATMENT
                         FMC Corporation
                  Princeton, New Jersey   08540
                              for the
                 ENVIRONMENTAL PROTECTION AGENCY
                       Project #17020 GDN
                      Contract #14-12-901
                             July  1971
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.00

-------
              EPA Review Notice
This report has been reviewed by the EPA,
and approved for publication.  Approval does
not signify that the contents necessarily
reflect the views and policies of the Environ-
mental Protection Agency, nor does mention of
trade names or commercial products constitute
endorsement or recommendation for use.
                     ii

-------
                       ABSTRACT
The magnitude and effects of biological activity in expanded
carbon beds used for direct clarification/adsorption treat-
ment of wastewater were investigated.  Major aspects of the
project involved comparisons of the relative effectiveness
of aerobic and anaerobic conditions in the expanded-bed
systems, and a comparison of the relative treatment effec-
tiveness of expanded carbon beds of high and low sorptive
activity under aerobic operating conditions.  The use of
short polishing columns to remove traces or organics  escaping
from on-line adsorbers was also evaluated.

The feed, primary effluent from a treatment plant near Trenton,
New Jersey, was coagulated with FeCl^, clarified and filtered.
Clarification removed over 60% of the charged total organic
carbon  (TOG) and 55% of the BOD.  The clarified primary
effluent was fed to the carbon columns.  Carbon activity
comparisons with and without addition of oxygen were made
in identical series of four columns containing a total of
24 ft  of carbon.

TOC removal in activated carbon beds operated constantly with
6 to 10 mg/1 of dissolved oxygen dropped from 81% originally
to 66% during nine months of operation, and averaged 77.6%
for the entire 889,600 gal. (8896 bed volumes)  of feed
treated.  During 3-1/2 months in similarly aerated columns,
beds of unactivated coal removed 44.3% of charged TOC,
entirely due to biological activity-  Activated carbon beds
operated without oxygen  (anaerobically) removed 66.9% of the
TOC, and presented problems of H2S evolution.  Anaerobic
conditions should be avoided.

Over the nine-month operating period, combined chemical clari-
fication plus aerobic activated carbon treatment (in 24-ft-deep
beds at a 5-gpm/sq ft flow rate) removed 87% of the TOC and
over 90% of the BOD.  The product had a TOC of 8 mg/1 or less,
and carbon column operation was continuous except for three
short interruptions to cleanse the columns.  With inactive
coal, combined coagulation and aerobic treatment gave 80%
TOC removal and 90% BOD removal.  The product averaged about
10 mg/1 BOD or TOC.  Estimated treatment costs for combined
clarification/adsorption treatment at 10 mgd are 20C/1,000 gal.
with activated carbon, and about lSC/1,000 gal. with the un-
activated coal.
                          111

-------
Polishing columns were effective in reducing effluent TOC
to about 2 mg/1, but to be economically practical  regenera-
tion must be very low cost.  The best regenerants studied
were not effective enough to offer promise for developing
an economically practical process embodying this technique
for producing an organic-free water.

This report was submitted in fulfillment of Project Number
17020GDN, Contract 14-12-901, under the sponsorship of
the Water Quality Office, Environmental Protection Agency.
                          IV

-------
                          CONTENTS

Section                                                  Page

  I       Conclusions                                       1

  II      Recommendations                                   3

  III     Introduction                                      5

  IV      Description of Pilot Plant                        9

          A.   Pilot Plant Installation                     9

          B.   Clarification System                         9

          C.   Carbon Adsorption Systems                   11

          D.   Product Sampling and Analytical Methods     J.6

  V       Objectives                                       19

  VI      Experimental Results                             21

          A.   Comparison of Aerobic and Anerobic
               Column Activities                           21

          B.   Comparison of Aerobic and Combined
               Anaerobic-Aerobic Carbon Column
               Activities                                  32

          C.   Comparison of Activated Carbon and
               Non-Activated Coal Derivative Under
               Aerobic Conditions                          4l

          D.   Summary of Carbon Column Operations         47

          E.   Regeneration of Polishing Column Carbon     50

  VII     Miscellaneous Analyses                           59

  VIII    Discussion of Results                            65

          A.   Effect of Aerobic Conditions Within
               Carbon Columns                              65

          B.   Effect of Carbon Properties on
               Column Efficiencies                         65

-------
                     CONTENTS  (Cont'd)

Section
          C.   Comparison of Results  from Current
               and Previous Studies                        66

          D.   Cleansing of Carbon in the Columns         66

          E.   In Situ Biologic Regeneration              67

          F.   Polishing Column                            70

          G.   Treatment Costs                             71

  IX      Acknowledgements                                 79

  X       References                                       8l

  XI      Publications                                     83
                             VI

-------
                        FIGURES

                                                       Pag

 1     Flow Diagram of Clarification System             ^.0

 2     Flow Diagram of Carbon Bed System                ^3

 3     Pilot Adsorber Column Detail                     ]_ij

 4     Clarification of Primary Effluent With
       140 ppm FeCl3                                    23

 5     Turbidities of Primary Effluent and of
       Clarified Primary Effluent                       24

 6     Removal of TOG with Oxygenated Clarified
       Feed in Expanded Beds of Activated Carbon        25

 7     Reduction of TOC with Anaerobic Clarified
       Feed in Expanded Beds of Activated Carbon        26

 8     Cumulative Removal of TOC in 24 Ft
       Expanded Beds of Activated Carbon in Aerobic
       and Anaerobic Operations                         29

 9     Cumulative Removal of TOC in 12 Ft Expanded
       Beds of Activated Carbon in Aerobic and
       Anaerobic Column Operations                      OQ

10     Removal of TOC by Activated Carbon in
       Aerobic and Combined Anaerobic-Aerobic
       Operation                                        35

11     Cumulative Removal of TOC in 24 Ft Expanded
       Beds of Activated Carbon  in Aerobic and
       Anaerobic-Aerobic Operations                     39

12     Cumulative Removal of TOC in 12 Ft Expanded
       Beds of Activated Carbon in Aerobic and
       Anaerobic Operation                              JJQ

13     Relative TOC Removal in Non-Activated and
       Partially Spent Activated Carbon Columns         42

14     Cumulative Removal of TOC in 24 Ft Expanded
       Beds of Activated and Non-Activated Carbons
       in Aerobic Operation                             ^5
                           VI1

-------
                    FIGURES (Cont'd)

                                                       Page

15     Cumulative Removal of TOC in 12 Ft Expanded
       Beds of Activated and Non-Activated  Carbons
       in Aerobic Operation                             46

16     Schematic Interpretation of the Mode of
       In Situ Biologic Extension of Adsorption
       Capacity                                         69

17     Proposed Scheme of Treatment of Raw Sewage
       By Chemical Clarification and Adsorption
       on Activated Carbon                              72

18     Proposed Arrangement of Processing Units         73
                           Vlll

-------
                         TABLES

                                                       Page

 1     Mesh Size Distribution and Physical Analyses
       of Activated Carbon and Non-Activated Coal
       Derivative                                       15

 2     Dissolved Oxygen Concentrations in Aerobic
       and Anaerobic Column Streams                     22

 3     Turbidities of Carbon-Treated Effluents          31

 4     Hydrogen Sulfide Concentrations in Carbon-
       Treated Effluent Streams                         32

 5     Dissolved Oxygen Concentrations in Streams
       From Aerobic and Anaerobic-Aerobic Columns       33

 6     TOC Removed by Carbon                            36

 7     H2S Concentrations in Effluents from Aerobic
       and Combined Anaerobic-Aerobic Carbon Systems    38

 8     Removal of TOC with Non-Activated and Spent
       Activated Carbons                                i|3

 9     Analyses of Spent Activated and Non-Activated
       Carbons from the Pilot Operations                48

10     Removal of TOC and SOC from Primary Effluent
       by Chemical Clarification and Carbon Treatment   ^9

11     Polishing Column Field Study with a 4-Foot
       Bed of CAL Carbon - Run A                        52

12     Polishing Column Field Study with a Low
       TOC Feed - Run B                                 53

13     Removal of TOC from On-Line Column Effluent
       with Fresh Activated Carbon in Laboratory
       Column                                           55

14     Removal of Organics by Polishing Carbon
       Regenerated with 11.0 pH Sodium Hydroxide
       Solution                                         55

15     Removal of Organics by Polishing Column
       Regenerated with 4% by Wt. H202                  56
                            IX

-------
                    TABLES  (Cont'd)

                                                        Page

16     BOD Removal in Aerated and Unaerated
       Activated Carbon Columns                         60

17     BOD Removal in Spent Activated Carbon
       and Non-Activated Carbon Columns                 60

18     Stabilities of Effluents from Aerobic
       and Anaerobic Columns Relative to
       Turbidities                                      6l

19     Iron in Column Effluents                         62

20     Total Phosphorus of Effluent Streams             62

21     Nitrogen Analyses of Effluent Streams            63

22     Estimated Capital Costs for Treatment
       of Clarified Raw Sewage in Activated
       Carbon and Unactivated Coal                      yc

23     Estimated Annual Operating Costs for
       Treating Clarified Municipal Sewage
       in Aerated Carbon Beds                           76

-------
                       SECTION I

                      CONCLUSIONS


1.   In continuous, nine-month pilot-scale studies, aerobic
expanded-bed activated carbon columns effectively and
economically treated a chemically coagulated and clarified
primary effluent to produce a high quality effluent.

2.   Overall organic carbon removal by the combined
coagulation and activated carbon treatment averaged almost
87.2% for the entire nine-month period.  BOD removal was
over 90%.

3.   Combined coagulation and treatment in the aerobic non-
activated coal columns removed over 80% of the organic
carbon in primary effluent.  BOD removal was over 90%.

4.   Activated carbon removed 75% of organic carbon in
chemically coagulated and clarified primary effluent; non-
activated carbon removed about 50% after 2 to 4 weeks when
biological activity developed inside the columns.

5.   In activated carbon beds, more organics were removed with
aerobic operation than with anaerobic.  Aerobic products retained
their clarity but anaerobic products clouded on standing.

6.   A combined anaerobic-aerobic system was also effective.
However, the anaerobic section gave a much poorer organic
removal than did the corresponding aerobic section.

7.   The effectiveness of the expanded-bed adsorbers was
enhanced by occasional air scrubbing and backwashing to remove
excess biomass and precipitated coagulant from the carbon beds.

8.   Anaerobic effluents consistently contained 8 to 13 mg/1
H2S, but effluents from aerobic columns were essentially free
of H2S.

9.   With only 24-ft-carbon-depth treatment, efficiencies
dropped in severely cold weather, probably due, in part, to
higher fluid viscosity, lower sorption rates, and a decline
in bacterial activity-

10.  NaOH and 4% #2^2 were the most active regenerants for
partially-spent polishing column carbons.  However, neither
regenerant restored the original activity of the lightly-loaded
carbon.
                          - 1 -

-------
                       SECTION II

                    RECOMMENDATIONS
     The following studies should be conducted to develop
an economical and effective expanded-bed physico-chemical
process:

     1.   Study ways of aerating streams to and between
          columns to devise a plant-operable system
          capable of introducing at least 6 mg/1 of Q>2
          into the stream before it enters the succeeding
          column.

     2.   Determine whether low activity sorbents can be
          operated more effectively in expanded-bed
          adsorbers for long periods of time.  The use
          of such materials would offer major cost
          economies.

     3.   Prepare flow sheet designs and economic
          evaluations for a demonstration plant of
          about  1-mgd capacity.
                         -  3 -

-------
                      SECTION III

                      INTRODUCTION
     Under the sponsorship of the FWPCA and EPA, the FMC
Corporation has been conducting investigations on advanced
physico-chemical methods for wastewater treatment for
over three years.  This work was primarily a study of
expanded-bed contacting for effective utilization of
granular activated carbon on the treatment of both secondary
and primary clarified sewage effluents.

     In expanded-bed operation, sewage is passed up through
a bed of carbon at a sufficient velocity to expand the bed
to at least 115% of the volume of the carbon at rest.  Weber1
had shown the feasibility of using expanded-bed adsorbers to
treat secondary effluent.  Expanded-bed adsorbers gave
essentially the same organic removal as fixed beds of
carbon,2 but the latter required frequent interruptions
for backwashing.2'^'^

     The expanded-bed adsorption system was also used to
treat primary effluent.5  With a chemically coagulated and
clarified primary effluent feed, the expanded-bed carbon
adsorption system produced a clear, high quality effluent.
Thus, the combined chemical clarification and adsorption
systems removed over 95% of the organics, essentially all
suspended solids and turbidity, and over 90% of the phos-
phates in the primary effluent.  Treatment cost for a 10-mgd
plant, based on 1969 values, was estimated to be 20^/1,000 gal.
of product water.  During four months of steady operation,
the expanded-bed treatment system needed little or no cleaning
or backwashing and operated with a low and relatively con-
stant feed pressure requirement.

     The results not only demonstrated the effectiveness of
expanded-bed carbon contacting, but also pointed up the
potential advantages of physico-chemical water treatment.
A modern well-operated biological treatment plant might
reduce suspended solids  (SS) by 95% and yield a product with
a BOD of 20 mg/1 or less.  However, such performance is
difficult to maintain on a continuous basis.  By contrast,
the products from the pilot-plant expanded-bed system con-
stantly had a BOD of under 10 mg/1 and the performance of this
system equalled or exceeded the performance of conventional
plants in other characteristics as well.

     The growth in the concern for the quality of surface
waters has accentuated the search for more efficient methods
of treating wastewaters.  Conventional biological treatment

-------
plants will be able to meet more stringent demands for
better water quality or more effective pollution control
only by adding tertiary treatment processes to existing
biological plants.  Such processes increase sewage treatment
costs significantly and require land areas that are becoming
increasingly difficult to acquire in developed areas.  In
addition, the effective operation of such a tertiary treat-
ment system is still dependent upon consistent and efficient
operation of the biological secondary treatment process,
which is subject to problems arising from changing waste
composition, fluctuations in flow, and the presence of
materials that are toxic to bacteria.

     Physico-chemical processes, on the other hand, produce
the desired higher treatment quality at costs approaching
those for efficient biological processes, and with smaller
land requirements and greater resistance to those upsets
that plague biological processes.

     During the course of the previous studies, the formation
and presence of biological growth on the carbon was noted.
The presence of such growth was attributed to the high con-
centration of adsorbed organic material accumulated on the
carbon.  There was no evidence that the biological activity
hindered the adsorption process.  On the contrary, the
presence of such activity enhanced the capacity of the
system for removal of organics, permitting longer treatment
periods before exhaustion of the activated carbon.

     To take full potential advantage of this phenomenon,
the relative effect of biological activity on the carbon
in expanded-bed columns was studied.  Although the studies
noted above indicated that such activity increased the
capacity of carbon for removal of organic matter from clari-
fied primary effluent, no work had been performed or reported
to establish its magnitude under varying conditions of
operation.  The work described here evaluated the effect
of operating expanded-bed carbon columns treating primary
effluent under conditions which would enhance biological
activity in order to assess the relative value of operating
a physico-chemical treatment process under such conditions.
After the preferred conditions for bacteriological growth were
established, they were evaluated with carbons of different
sorptive capacities to determine if adsorption on carbon
surfaces affected biological activity within a carbon column.

     In virtually all carbon treatment processes, there is a
leakage of organic compounds through the activated carbon
column.3  The exact nature of this leakage is unknown, but
it probably is due to dead cell fragments or small highly
degraded, polar materials that are adsorbed poorly.  Studies
                          _ t> _

-------
using small, fresh carbon beds as a polishing adsorber showed
that this leakage was at a minimum when treated effluent was
first passed through fresh carbon, but that it increased to
4 to 5 mg/1 in a few days, presumably due to a chromatographic
effect in which the more easily adsorbed materials displace
those that are adsorbed weakly.  The action and potential
utilization of the polishing column, as a means of extending
the effectiveness of the physico-chemical treatment system, were
also examined in this study.
                           -  7 -

-------
                         SECTION IV

                 DESCRIPTION OF PILOT PLANT
A.   Pilot Plant Installation

     The pilot plant used for the experimental work was the same
as that described in the FWQA report of the preceding study
on the treatment of secondary effluent in expanded-bed
adsorbers.5  The plant was located at the sewage treatment
facility of the Ewing-Lawrence Sewerage Authority  (ELSA) near
Trenton, New Jersey which serves residential, commercial, and
industrial areas within the two townships of Ewing and Lawrence.
The sewage consists of about 25% industrial waste and 75% domestic
waste.  This is a trickling filter plant which includes four
circular, primary sedimentation basins.

     For this study, primary effluent was taken from one of
these basins and siphoned through a 300-ft  long, 1.5-in.
polyethylene pipe to a pump at the pilot plant site.  Primary
effluent was used as feed because of the presence in the raw
sewage of varying amounts of large solids which would have
been difficult to handle in the small lines and valves of the
pilot system.  The primary basins provide settling for raw
sewage plus the liquid from return sludge and the supernatant
from the anaerobic sludge digester.

     The experimental apparatus at the pilot plant was set on
a poured concrete slab installed next to the ELSA return pump,
and controls were located in a 10-ft by 16-ft building con-
structed on the slab.  The carbon adsorbers were internally
coated steel pipe columns resting on the slab and supported by
an angle iron frame.  The filter and other tanks were located
on the slab next to the building.  Connections between the
columns and valves were made with rubber hoses which passed
through the building walls.


B.   Clarification System

     The clarification system consisted of two 55-gal. drums
for rapid mix, coagulation, and flocculation followed by an up-
flow clarifier and dual media filter, as shown in Figure 1.
The primary effluent was pumped through a float valve which
controlled the level in the rapid mix compartment  in the upper
part of the first 55-gal. drum.  The coagulant, a  30 weight-
percent aqueous solution of ferric chloride, was fed by  a metering
pump into the stream of primary effluent, discharging from a
5/8-in. nozzle into an elbow to impart a circular motion in the
                             -  9  -

-------
o
i
                                   RAPID  MIX
                                   CHAMBER

V.


,^
c

<\
)
(D

ATORS-^
i — i

gpm
                                                                         1
            9'COAL
            9" SAND
         6"COARSE SAND
            GRAVEL
                                                     SLUDGE PUMP
DRAIN
    AIR-*-*—i
 WATER-«HX>—L
 FOR BACKWASH
             PUMP
                                                UP-FLOW CLARIFIER
                                                45'.d.4'STRAIGHT  SIDE
                                                 60°CONE  BOTTOM
           DUAL  MEDIA
            FILTER
             38'id.
                       BACKWASH
                       DRAIN
                                                                                                                   5.5
                                                                                                                      gpm
                                                                                                    CLARIFIED FEED
                                                                                                      RESERVOIR
                                 FIGURE  I —FLOW   DIAGRAM  OF  CLARIFICATION   SYSTEM

-------
rapid-mix compartment.  A motor driven propeller was used to
provide additional rapid mixing.  After an average detention
time of two minutes at the design flow of 5.5 gpm, the mixture
flowed successively into the lower part of the first drum for
flocculation with slow mixing, into the bottom of the second
drum for further flocculation with slow mixing, and then out at
the top and into the clarifier.  The slow stirring in both tanks
was provided by 24-in. by 2-in. redwood paddles, with the 24-in.
dimension in the vertical position, mounted at a 7-in. radius
to a vertical shaft driven by a constant speed motor.  The
first flocculation tank was also fitted with three vertical
redwood stators, 24-in. by 1.5-in. by 0.5-in. attached to the
side of the tank.  At 5.5 gpm flow, the detention times were
7 min. in the first flocculation chamber and 9 min. in the
second.  The motor-driven paddles in these tanks could be
operated at various speeds to provide different degrees of
mixing.

     The up~flow clarifier was designed for this project.  It
was a 400-gal. capacity, shop fabricated, steel cylindrical
tank, 3 ft 9 in. in diameter with a 4-ft high straight section,
and a 60° cone bottom.  Flocculated water entered a central
8-in. diameter chimney, discharged at a depth of about 4-ft
below the surface, then flowed upward to the over-flow trough
at the surface.  The detention time at 5.5 gpm was approximately
one hour.  The product water from the clarifier flowed to
the dual-media filter, which consisted of 9 in. of anthracite
coal  (effective size 0.59 mm) over 9 in. of filter sand
(effective size 0.62 mm) supported on gravel with a pipe
underdrain.  Filtered water was pumped to a 250-gal. reservoir
to provide feed to the pumps for the carbon column systems.

     Sludge was pumped from the clarifier by a positive pressure
pump attached to the bottom of the cone and operated by a
cycle timer to remove and discard sludge at predetermined
intervals.  Usually this pump was operated for one minute,
three times an hour, to discharge about 15 gal. of sludge per
hour.
C.   Carbon Adsorption Systems

     1.   Multi-Bed Adsorbers

          The carbon adsorbers were vertical columns constructed
of internally coated 10-in.-diameter steel pipes with flanges
at each end.  The pilot plant had eight such columns which
were operated as two parallel systems of four columns each.
                            -11 -

-------
          The flow diagram for each system is shown in Figure 2.
The first two columns in each system were 9-ft steel columns
with a 1-ft section of ll-in.-i.d. transparent pipe to permit
observation of the expanded carbon beds, and to provide a
visual check to ensure that carbon was not being washed out
of the column.  The last two columns in each system were
11 ft steel columns.  Figure 3 shows a schematic view of the
columns with the transparent pipe.  The all-steel pipes had
identical construction except for the absence of the trans-
parent section.  Hose connections were provided at the tops
and bottoms of the columns.

          Each column was charged with 3.25 cu ft carbon
which provided a 6-ft-deep bed of settled carbon.  The carbon
was supported on a 6-in. layer of gravel and coarse sand
over an inverted 5-in.-diameter cone shaped distributor.
The two carbons used in this investigation were Pittsburgh
Filtrasorb 400 activated carbon, marketed by the Calgon Cor-
poration, Pittsburgh, Pennsylvania, and "Filt-0-Cite" No. 1
filter media, an anthracitic material marketed by the Shamokin
Filter Company, Shamokin, Pennsylvania.  Both carbons were
charged as received.  The analyses of these carbons is shown
in Table 1.

          Feed was supplied to the columns by constant dis-
placement pumps driven by electric motors through variable
speed drives.  The feed charge rate was 5 gpm/sq ft.
Reinforced rubber hose, 5/8-in.  i.d., was used for the
connecting lines to and between the columns to provide for
ease of installation.  All flow controls, including in-line
valves, pressure gauges, flow meters (water meters), and
solenoid sampling valves were mounted on one central operating
panel within the pilot building.  Stream flows were maintained
at constant rate by adjustment of the pump drives.  Treated
water from the expanded-bed columns was discharged into a
drum so that any carbon particles carried out could be
collected and returned to the column if necessary.  All
product water was returned to the sewage treatment plant.

          When added to column influents, oxygen was fed into
the feed lines just downstream from the column solenoid sampling
valves inside the pilot building.  The oxygen was bled from
high-pressure storage cylinders through calibrated rotameters
at rates designed to give oxygen concentrations of 6 to 10 mg/1
in column feed lines.  Samples were taken from the bottoms of
each column to monitor their dissolved oxygen contents.
                            - 12 -

-------
OJ

 I
          PRIMARY
          EFFLUENT
                    CHEMICAL
                   CLARIFICATION
                                                                 TO  DISPOSAL OR
                                                                 POLISHING COLUMN
         FIGURE  2—FLOW  DIAGRAM  OF  CARBON  BED  SYSTEM

-------
                                HOSE  CONNECTION FOR
                                COLUMN FEED  OR PRODUCT
                                 II in. O.D. ACRYLIC PIPE
                                 3/8 in. WALL
                                 1/8 in. RUBBER GASKETS
                                 3/4 in. BOLTS
                                 10 in.  STEEL PIPE 1/4 in. WALL
                                 I50lb. FLANGES
                                 BED DRAIN


                                 3/4 in.  I.P. COUPLING

                                 HOSE CONNECTION  FOR
                                 COLUMN FEED  OR PRODUCT
                                 PLASTIC CONE Sin. DIA.

                                 COVERED  WITH GRAVEL
FIGURE 3 —  PILOT  ADSORBER  COLUMN  DETAIL

-------
                          Table 1

      MESH SIZE DISTRIBUTION AND PHYSICAL ANALYSES OF
     ACTIVATED CARBON AND NON-ACTIVATED COAL DERIVATIVE
                       Filtrasorb 400'
Sieve Analysis
 Sieve
U.S. No.

   16
   20
   30
   40
   50
  Pan
Retained

   9.0
  42.6
  33.0
  13.2
   2.0
   0.2
 Filt-0-Cite No. 1
 Sieve       %
U.S. No.  Retained
                                                 8
                                               10
                                               16
                                               20
                                               30
                                              Pan
            10.8
            23.9
            54.1
             9.6
             1.4
             0.2
Density-Ib/ft

     Bulk

% Moisture
% Volatile Matter
% Ash
Iodine No. mg/g
             24.6

              1.2
              1.0
              6.6

          1,270.
                         52.5

                          9.6

                        < 1.0
                         19.3

                        150.
a Purchased from Calgon  Corp.,  Pittsburgh,  Pa.

b Purchased from Shamokin  Filler  Co.,  Shamokin,  Pa.
                             -  15  -

-------
          In the up-flow operations used, the carbon seldom
required cleaning.  When it did become necessary, however, to
dislodge solids trapped in the carbon beds, the most effective
procedure consisted of lowering the water level to within a
foot of the top of the carbon bed, injecting air into the
bottom of the column for 5 to 10 min. and then backflushing
with clean water to sweep away the dislodged solids.  The
sludge resulting from the carbon cleaning operation was
collected in a drum so that any activated carbon lost from
a column could be returned to that column.

          The entire system was designed for essentially
automatic operation.  A technician visited the plant daily
to adjust flows, perform routine maintenance, and take samples
for analysis.


     2.   Polishing Column

          The polishing column consisted of a 6-in.-i.d. glass
column containing 20 Ib of fresh Filtrasorb 400.  This amount
of carbon formed a 4-ft-long bed in the columns.  The
column was washed upflow with city water to purge approxi-
mately 0.5 Ib of dust that rose to the surface before each
run.  Feed for this column was effluent from the 24-ft columns.
During polishing column tests, the column effluent was
collected in a clean, polyethylene 5-gal. bottle which
served as the reservoir for the polishing column feed.  This
feed was charged up-flow at 1 gpm, which was the same 5-gpm/sq
ft rate used in the large columns.

          Regeneration studies on spent carbons for the polishing
columns were conducted in the laboratory.  The adsorption
effectiveness of the treated carbons was evaluated in glass
columns having a cross-sectional area of 1 sq in. in beds
18 in. long.  Lffluent from the 24-ft carbon beds at
the pilot plant was collected in clean 5-gal. polyethylene
bottles daily and brought to the laboratory.  This effluent
was passed upflow through the laboratory glass columns at a
rate of 1 gpm/sq ft.


D.   Product Sampling and Analytical Methods

     Composite samples of the primary effluent feed to each
carbon column and the product water from each column were collected
automatically by timer-controlled solenoid valves which opened
at 15-min. intervals to draw 60-ml samples.  These samples were
composited in 5-gal. polyethylene bottles in an acid medium  to
maintain stability and prevent deterioration or biological
activity over the sampling periods.  Spot samples were
collected by hand.
                           - 16 -

-------
     Chemical and biochemical analyses were performed at the
FMC Chemical Research Center.  Analytical determinations on
composited samples included TOC, soluble organic carbon (SOC),
and suspended solids  (SS).  BOD, turbidity, dissolved oxygen,
pH and all analyses not performed routinely were run on spot
samples brought unacidified to the laboratory for immediate
analysis.

     Suspended solids concentration of the primary effluent
and treated water samples was  measured by a procedure^ in-
volving the use of 0.45-micron membrane filters*  The membrane
filters, which in manufacture are treated with an organic
conditioning agent, were washed in distilled water before
use to remove this agent.  They were then dried in individual
desiccators to constant weight.  After filtration of a sample,
each filter was dried again in the same desiccator and
weighed to determine weight gain by retention of suspended
solids.  The filtrates were collected to provide samples for
SOC analyses.

     Well-mixed composite samples were analyzed on the Beckman
Carbonaceous Analyzer for organic carbon analysis.  TOC was
determined directly,  and SOC on the filtrate from the sus-
pended solids determination.

     Dissolved oxygen was determined by a Precision Scientific
Company galvanic cell oxygen analyzer.  Hydrogen sulfide in
aqueous streams was determined by titration with standardized
iodine solutions.

     BOD of the unacidified spot samples was determined by
the dilution procedure described in "Standard Methods".7

     All turbidity determinations were made with a Hach Model
2100 Photoelectric Turbidimeter.

     Total phosphate was determined by ASTM Procedure D-515-
60T on samples that had been digested to convert all phosphate
to the ortho form.8

     Nitrate was determined by ASTM Procedure D-992-52.
Ammonia and organic nitrogen were determined by  the  Kjeldahl
Procedure as outlined in "Standard Methods"."7
                           -  17 -

-------
                         SECTION V

                         OBJECTIVES
     The primary objective of this investigation was to study
the effects of biological activity in expanded-beds of carbon
treating clarified primary effluent.  In each phase of the
program parallel tests provided direct comparison between
tests under different conditions.

     The specific objectives of the projects were:

     1.   To compare the relative effectiveness of aerobic
          and anaerobic conditions.

     2.   To compare the relative effectiveness of aerobic
          and combined anaerobic-aerobic conditions.

     3.   To compare the behavior of an activated and a
          non-activated carbon under aerobic conditions.

     4.   To evaluate the concept of using a short bed
          of fresh activated carbon as a final polishing
          adsorber to produce an essentially organic-free
          effluent.

     5.   To determine if a simple, effective technique
          can be developed to make polishing carbon
          treatment economical.
                           -  19  -

-------
                         SECTION VI

                    EXPERIMENTAL RESULTS
A.   Comparison of Aerobic and Anaerobic Column Activities

     The relative effectiveness of aerobic versus anaerobic
column operation was studied for two months.  Clarified feed
was pumped from a common storage vessel to the two sets of
columns which had the same geometry and quantity of activated
carbon.  Precautions were taken to minimize splashing in the
clarification and filtration vessels so that the feed would
pick up a minimum of dissolved oxygen.  Usually, this
amounted to less than 2 mg/1 Cu  (see Table 2).  A dissolved
oxygen content of 6 to 10 mg/1 02 was obtained in the
aerobic columns by bleeding oxygen into the feed entering
the columns.  Originally, 0.1 cfh of oxygen was bled into
each of the four columns.  Later it was found that the
desired oxygen content could be maintained by adding the
same quantity of oxygen to only the first and third columns.
The oxygen did not dissolve completely, as oxygen bubbles
were observed in the transparent sections at the top of the
columns.  At the start of the run these oxygen bubbles
entrained tiny particles of carbon to the top of the column.
This ceased after a few days.

     Clarifying primary effluent with approximately 140 mg/1
of ferric chloride in the coagulation-clarification system
reduced the TOC by about 50% during this part of the work,
as shown in Figure 4.  The TOC of the primary effluent
averaged about 60 mg/1 and the clarified primary effluent
showed a TOC of approximately 30 mg/1.  Figure 5 shows the
reduction in the turbidity of the primary effluent due to
clarification.  Primary effluents generally had turbidities
of between 40 to 50 Jackson Turbidity units  (JTU).  The
clarified effluent generally had turbidities of between
1 and 7 JTU.  The only two high values, 11 and 17 JTU res-
pectively, were recorded in periods when the clarifier
became septic.

     The removal of TOC in the aerobic and anerobicc columns
is shown in Figures 6 and 7, respectively.  The solid line
shows the TOC of product leaving the fourth column, and the
triangles show the TOC of product leaving the second column.

     Effluent from the aerobic column contained less than
8 mg/1, except when approximately 65,000 and 175,000 gal. of
effluent feed were passed through the columns.  These were
the two aforementioned periods in which the clarifier became
                            - 21  -

-------
                                               Table  2

              DISSOLVED OXYGEN CONCENTRATIONS  IN  AEROBIC  AND ANAEROBIC COLUMN STREAMS


                                       Dissolved  oxygen,  mg/1
ru
ru
Date        Feed

 9/28
 9/30
10/2
10/5
10/7
10/12 am 0.3
      pm 0 . 4
        10/14
        10/16
        10/19
        10/21
        10/23
        10/26
        10/28
        10/30
        11/2
         0
         0
         0.8
         0.8
         0
         0
         0.6
         0.7
         4.2
CPE

4.8
3.5
2.8
2.5
2.2
2.5
0.4
0.6
0.4
0
0
0
0
0
0.6
3.1
 5
 5
,6
 2
 5
Stream
Aerobic
A-l
7.0
7.0
3. 8
8.0
4.1
5.5
-
6.9
6.6
5.0
6.2
7.2
5.6
7.7
7.0
7.6
A- 2
4.6
4.9
7.3
4.3
7.0
7.2
9.5
11.2
9.2
10.3
10.6
8.7
6.9
12.5
7.6
7.6
A- 3
4.5
5.2
9.4
9.4
12.5
7.6
-
9.9
5.3
7.4
7.2
7.7
5.1
5.0
3.0
10.8
A- 4
5.0
5.5
7.7
12.5
12.7
8.6
7.0
9.2
5.6
9.3
7.4
7.0
3.4
1.8
0.5
10.5
U-l
3.7
2.6
1. 8
2.8
1.6
0.7
-
0.8
1.2
0.7
1.4
0.7
0.5
0.9
0.7
07
Anaerobic
U-2
-
-
-
-
-
0.6
0.9
0.6
0.6
0.5
0.7
0.5
0.4
0.6
0.5
Deratic
U-3
-
-
-
-
-
0.6
-
1.0
0.5
0.8
0.6
0.7
0.4
0.7
0.6
DH stoi
U-4
3.5
1.5
1.3
1.8
1.4
0.3
0.8
0.6
0.4
1.1
1.0
0.8
0.4
0.6
0.3
oped
        PE - Primary effluent
       CPE - Clarified primary effluent

        A-l, etc. - First aerobic carbon  column,  etc.
        U-l, etc. - First anaerobic  carbon  column,  etc

-------
ro
UJ
              120 r
              100
            o>
            E


            O
            CO
            (E
            <
            O
80
            O 60
            o
            cc
            O
            O
              40
              20
    CLARIFIED
   hPRIHARY
    EFFLUENT
                                            I
                                       I
                   I
I
                         50
                    100
150       200      250      300

  VOLUME TREATED, 1000 GAL.
        350
400
                 FIGURE  4 — CLARIFICATION  OF  PRIMARY EFFLUENT WITH  140 ppm  FeCI3

-------
IX)
£=•
              50
              40
           :D
           h-
           -3
           CD
           tr
              20
              10 -
               0
                        25
    PRIMARY EFFLUENT
                                      CLARIFIED  PRIMARY  EFFLUENT
50
75       100      125       150

 VOLUME   TREATED, 1000 GAL.
175
200
                FIGURE  5-TURBIDITiES  OF  PRIMARY EFFLUENT  AND  OF  CLARIFIED  PRIMARY

                            EFFLUENT

-------
            50 r
ro
ui
            40
          o>
          E
            30
          O
          m
          o
          O
            20
          o
          o
          I-  10
                      THRU 24' BED
                      J	I
                      25
50
75      100      125      150

  VOLUME  TREATED, 1000 GAL.
175
200
             FIGURE  6 —REMOVAL OF  TOC  WITH  OXYGENATED CLARIFIED  FEED  IN  EX-
                        PANDED BEDS OF  ACTIVATED  CARBON

-------
  50
I

rvj
  40
o
m


< 30
u
o
o
a: 20
O
o
  10
           THRU 24'BED

          _J	L
           25
                            50
75      100      125      150

 VOLUME  TREATED, 1000 GAL.
175
200
   FIGURE  7—REDUCTION  OF  TOC  WITH  ANAEROBIC  CLARIFIED  FEED  IN  EX-

              PANDED  BEDS  OF ACTIVATED CARBON

-------
septic.  While the first 50,000 gal. was passed through the
column, the 12-ft. column was nearly as effective as the
24-ft column, but the 12-ft column then lost effectiveness
and the effluent was definitely inferior to that from the
24-ft columns by the time 100,000 gal. of feed had been
treated.

     In the anaerobic system, the quality of the effluent
was not nearly as consistent as that of the aerobic system.
Thus, the TOG of effluent rose after 10,000 to 15,000 gal.
of feed had been charged, and when approximately 85,000 to
115,000 gal. were charged,in addition to the two aforementioned
periods when the septic conditions were encountered.  The
occasional higher values with the anaerobic system may be
indicative of a more unstable system.  It should be noted,
however, that despite these occasional high values, the
effluent from the anaerobic system had low TOG several
times during this study.

     A comparison of the data in Figures 6 and 7 shows that
when the first 50,000 gal. was passed through the columns
containing fresh carbon, the removal of TOG in the aerobic
and anaerobic columns was not greatly different.  Thereafter,
the TOG removal in the aerobic carbon columns was signicantly
greater than in the anaerobic columns.  Except for two
readings, the aerobic column effluent TOG was consistently
under 10 mg/1 during that period when between 60,000 to
135,000 gal. was charged to the columns, but the TOG of the
effluent from the anaerobic columns was under 10 only twice
during that period.

     After 135,000 gal. had been pumped through both columns,
the clarification vessel became septic and the entire
system had to be shut down and cleaned thoroughly.  Following
this cleaning, both the aerobic and anaerobic columns again
showed essentially the same degree of TOG reduction until
175,000 gal. had been treated.  Thereafter, the TOG of
effluent from the anaerobic columns again became significantly
higher than those of the aerobic columns and this continued
until the end of the run.

     The data in Figures 6 and 7 show that the bulk of the
TOG was removed in the first 12 ft of the carbon.  After
150,000 gal. of feed had been passed through the anaerobic
columns, removal of TOG by the first 12 ft dropped markedly.
However, both systems still had considerable residual sorptive
capacity in the first 12 ft of the respective columns.

-------
     The  cumulative   removal of TOG in the 24-ft aerobic
and anaerobic columns during the two-month period studied is
presented in Figure 8.  The aerobic columns removed 77%
of the charged TOC, the anaerobic columns 67%.  Both
systems removed a larger fraction of the charged TOC after
the system had been cleaned and aerated.  The cleaning
became necessary because the clarification vessel became
septic.  This probably is a further indication of the harmful
effect of septicity in a carbon sorption system.  The improved
removal of TOC after aeration was not manifested, however, in
the first 12-ft beds of carbon.  Thus, plots of the amount
of TOC removed vs that applied in the 12-ft beds of carbon
show no significantly increased rate of sorption of TOC
after 32 pounds of TOC had been applied and the system
was backwashed and aerated  (see Figure 9).

     Of the first 50 pounds of TOC charged to both columns,
the 12-ft anaerobic columns sorbed 54% as opposed to 62% for
the aerobic.  This relative difference is virtually the same
as the 67% and 77%, respectively, found in the 24-ft beds.

     The turbidities of various streams of the carbon
treating system are shown in Table 3.  Clarification of
the primary effluent with 140 mg/1 FeCl3 removed over 90%
of the turbidity.  Product from the first aerated column
usually was more turbid than the clarified primary effluent.
Product from the last aerated column was almost always more
turbid than product from the unaerated columns, but the
latter usually became more turbid on standing.  Thus, samples
of fresh unaerated product that had turbidities of one or two
JTU, showed turbidities of 30 or 40 JTU after standing 24
hours.  By contrast, the turbidities of the aerated samples
dropped significantly after standing 24 hours and this clari-
fication was accompanied by a settling of grayish solids
which were presumed to be bacterial remains removed from
the carbon.

     Typical hydrogen sulfide contents of carbon-treated
streams are shown in Table 4.  The clarified primary effluent
normally contained up to 4 mg/1 of material which reacted
with iodine and was reported as PUS even though no H~S odor
could ever be detected in these streams.  Effluents from the
second and fourth aerated columns contained approximately
0.5 ppm of material titratable with iodine, as compared
with 0.4 ppm for city water.  Samples from the second and
fourth anaerobic columns consistently contained from 7 to 13
mg/1 of H2S and had a pronounced H2S odor.  After standing
at room temperature for 24 hours, samples of this H2S-containing
product contained less than 0.6 mg/1 H2S and never more than
a trace of H2S odor.
                           - 28 -

-------
  35
o 30
Q.
o
bJ


o

UJ
CE
  25
1 20
o:
<
o
tr
o
   10
                                      AEROBIC
                                                    ANAEROBIC
                  I
                        _L
_L
_L
_L
_L
_L
                 10      15     20     25     30     35     40

                      TOTAL  ORGANIC CARBON  APPLIED,  POUNDS
                                                               45
                                       50
                                       55
    FIGURE  8 — CUMULATIVE REMOVAL OF TOC IN 24FT  EXPANDED BEDS  OF ACTI-

                VATED CARBON  IN  AEROBIC  AND  ANAEROBIC  OPERATIONS

-------
LO
O
o


o

z
          cr
          o
          i-
          o
            35
30
          Q

          i
          o
          CL
          Q
          UJ
          LL)
          o:
          O 20
          m
          cr
 15
             10
             0
              0
                                  AEROBIC  COLUMNS
                                                   ANAEROBIC COLUMNS
                                               J_
                                                _L
                                                         _L
               10      15     20     25     30     35     40

                       TOTAL  ORGANIC CARBON  APPLIED, POUNDS
                                                                45
50
55
              FIGURE 9 —CUMULATIVE REMOVAL  OF TOC  IN 12 FT  EXPANDED  BEDS OF ACTIVATED

                         CARBON IN AEROBIC  AND  ANAEROBIC  COLUMNS OPERATIONS

-------
 Date
                          Table 3
          TURBIDITIES OF CARBON-TREATED EFFLUENTS
PE
                      Turbidities, JTU
                    CPE
                     A-l
A-4
U-l
U-4
9/2
9/4
9/11
9/14
9/21
9/23
9/25
9/28
9/30
10/2
10/5
10/12
10/14
10/16
10/19
10/21
10/23
10/26
10/28
10/30
50
35
40
40
43
44
48
39
42
42
38
43
54
46
47
49 b
42(7. 2)D
44
43
48
8.
5.
4.
4.
3.
7.
10
4.
4.
4.
3.
3.
1.
0.
6.
3.
1.
5.
6.
6.
0
0
0
0
2
5

0
0
0
5
8
0
8(32)
8
7(9.0)
1(13.0)
0
6
5
3.
3.
2.
7.
13
7.
8.
15
15
5.
7.
7.
14
13
10
14
22
17
0
0
4
5

0
0


8
3
5






3
3
2
2
6
4
5
8
18
11
10
19
29
18
11
17
16
19
.0
.0
.0
.8
.5

.5
.0


.5
.0(7.8)a

(3.5)
(1.8)





1.
2.
4.
4.
6.
5.
5.
2.
1.
1.
5.
1.
0.
4.
3.
2.


5
8
0
0
0
0
0
3
2
3
0
8
7
6
2
3


1.0
1.5
2.0
2.5
4.0
4.0
2.5
1..2
1.2
1.6
4.0
1.5
11.3
4.8
2.3
1.5











(31)a

(48)
(5.6)



a Numbers in parenthesis are values obtained in samples
  kept 24 hours at room temperature.

b After settling 72 hours.
PE    Primary Effluent
CPE   Clarified Primary Effluent
A-l   Effluent from first aerated column
A-4   Effluent from 4th aerated column
U-l   Effluent from first anaerobic column
U-4   Effluent from 4th anaerobic column
                           - 31 -

-------
                          Table 4

             HYDROGEN SULFIDE CONCENTRATIONS IN
               CARBON-TREATED EFFLUENT STREAMS

                  H2S Concentration, mg/1
Date       PE       CPE       A-2       A-4       U-2       U-4

10/22     2.0       3.1       0.6       0.4       9.2       13.4
10/23     0.9       1.7       	       0.6       	       7.9
10/26     1.6       1.7       	       0.4       	       7.0
10/28     1.5       3.9       0.7       0.6       9.9       11.9
10/30     2.0       2.0       0.5       0.5       9.9       12.5
     No differences were found in the pH of the aerobic and
anaerobic products.  Virtually all final products from both
the aerobic and anaerobic columns had pH of between 6 and 7,
and the variations in pH between aerobic and anerobic products
taken in the same day rarely exceeded 0.2.

     Except for short periods during which the clarifier was
down for cleaning, operation of the carbon columns was not
interrupted during the entire 2-month aerobic-anaerobic evalua-
tion period during which time 210,000 gal. of feed (2100 bed
volumes) were treated.  There was no evidence of any pressure
build-up inside the columns at any time during this period.
This confirmed the anticipated benefit from up-flow, expanded-
bed operation.


B.   Comparison of Aerobic and Combined
     Anaerobic-Aerobic Carbon Column Activities

     The anaerobic four-carbon column system was operated
with two anaerobic and two aerobic columns to determine if
this type of operation offered any advantages over a completely
aerobic system.  The aerobic system was operated as described
above without interruption but the combined anaerobic-aerobic
operation was started about two weeks after anaerobic operation
had been suspended.

     In anaerobic-aerobic operation, the first two columns were
operated exactly as described above for anaerobic operation,
and 0.1 cfm of oxygen was bled into the last two columns.
Table 5 shows that this provided sufficient oxygen to generally
maintain the desired 6 to 10 mg/1 02 in the last two columns
of the anaerobic-aerobic system.  The aerobic system had a
higher oxygen level than that shown in Table 2, even though
                           - 32 -

-------
                     Table  5

DISSOLVED OXYGEN CONCENTRATIONS IN STREAMS FROM
     AEROBIC AND ANAEROBIC-AEROBIC COLUMNS

             Dissolved oxygen,  mg/1
   Feed
Columns
Aerobic
1970
11/11
13
16
18
20
23
25
30
12/ 2
4
7
9
11
14
16
18
23
29
1971
I/ 4
8
11
13
20
22
25
27
29
2/ 1
5
PE-
1.0
2.0
2.2
2.3
2.0
3.7
0.3
0.1
0.6
0.4
0.4
1.1
0.6
1.4
3.2
3.6
2.8
2.0

1.7
7.2
5.0
	
3.2
2.4
0.8
4.7
4.6
3.0
6.2
CPE
1.7
1.4
4.2
3.5
1.6
2.8
0.1
0.7
0.6
0.4
0.7
0.5
1.4
1.4
1.2
2.2
0.7
1.0

1.7
2.2
4.0
0.8
0.6
2.9
1.2
0.3
1.7
1.0
1.2
A-l
12.3
14.5
11.5
13.5
12.0
12.5
12.6
17.5
25.6
19.0
24.0
14.3
15.4
17.0
13.6
13.0
17.0
12.0

21.0
11.2
13.0
	
15.2
15.7
6.6
	
15.6
15.0
9.7
A- 2
4.4
10.5
3.0
4.2
7.4
6.2
4.0
4.5
5.1
3.6
6.6
5.9
11.1
12.0
9.2
10.0
9.0
4.0

9.2
12.5
5.0
8.0
5.8
1.2
2.1
6.3
6.4
6.5
4.1
A- 3
6.0
10.0
10.0
10.4
9.6
10.5
14.0
6.1
17.8
10.0
20.0
7.3
5.7
13.5
8. 8
12.0
15.0
17.0

26.7
20.0
13.5
	
17.0
15.2
7.9
	
14.5
21.0
11.7
A-
8
13
10
10
9
10
15
7
15
6
13
7
6
11
7
10
17
19

28
18
14
21
18
18
9
24
16
-
11
4
.0
.1
.0
.4
.2
.0
.0
.2
.8
.6
.0
.6
.0
.2
.6
.5
.0
.0

.6
.7
.1
.0
.0
.8
.5
.5
.0
--
.2
Anaerobic- Aerobic
U-l
1.5
2.2
2.5
2.1
1.6
4.4
1.0
1.2
2.8
0.6
1.2
0.9
1.7
2.9
2.0
5.8
1.2
1.1

1.1
2.5
4.0
	
1.0
0.9
1.3
	
1.7
1.5
1.2
U-2
1.0
	
1.6
1.9
1.4
2.5
0.6
0.8
1.0
0.6
0.7
0.6
1.4
2.3
2.0
4.0
1.2
1.1

1.1
2.2
4.0
1.7
1.0
0.7
0.8
1.6
0.8
0.7
0.9
U-3
6,0
16.0
4.3
8.5
4.2
8.1
4.0
2.7
4.4
2.6
10.0
4.0
8.4
9.4
4.0
8.0
1.3
1.3

4.2
4.4
6.7
	
3.2
1.7
5.8
	
4.0
5.1
3.8
U-4
7.5
17.5
8.8
10.5
10.0
11.0
10.0
4.1
6.4
3.7
7.1
4.6
10.0
13.5
4.4
8.4
8.0
1.2

6.7
15.6
10.2
8.0
9.7
5.7
6.6
7.9
7.5
15.2
7.7
                      - 33 -

-------
oxygen was added at the same rotameter setting.  The higher
oxygen content could have been due to a lower oxygen con-
sumption by the bacteria during the winter weather and to
the greater solubility of oxygen in water at low temperatures.
Oxygen solubility is 43.4 mg/1 at 20°C (68°F), and 60.7 mg/1
at 5°C (41°F), the water temperature that was attained quite
often during the winter.  The dissolved oxygen content was
particularly high in January when temperatures were the
 lowest.   The oxygen rotameter setting was too low to make
the critical adjustments with the control valve that would
have given 6 to 10 mg/1 in cold weather.

     The relative removal of TOG by the aerobic and anaerobic-
aerobic systems is shown in Figure 10.  Except for the two
sharp peaks for the anaerobic-aerobic effluent at the
270,000 and 430,000-gal. regions, the effluents from the
two systems had essentially the same TOC values.  The fact
that anaerobic-aerobic activity, unlike anaerobic, was
equivalent to aerobic activity constituted additional evidence
for the beneficial effects of aerobic carbon column operation.

     TOC removal generally was poorer during this evaluation
than during the aerobic-anaerobic comparison, as shown in
Table 6.   Between August 28 and October 30 when the aerobic
and anaerobic systems were compared, TOC removal averaged
77.5% through the aerobic columns, and 67% in the anaerobic
columns.   Between October 31 and November 11, the aerobic
columns still removed 78% of the charged TOC, but removals
fell steadily from December 1 through February 8, and were
significantly lower than during September through November.

     The drop in TOC removal was due to at least four factors.
These were:

     1.   The much colder weather in winter reduced bacterial
          activity within the columns.

     2.   Rates of adsorption of organic matter from aqueous
          phase are decreased at lower temperatures.

     3.   During December, several clarified effluents had
          TOC of 20 to 22 mg/1 as compared with 27 to 35 mg/1
          that was characteristic of earlier operation.  Since
          the TOC values of column effluents obtained with
          this weaker feed were not lower than those obtained
          with the stronger feed, the actual percent of TOC
          removed by the columns dropped.

-------
UJ
un
             o>
             E
              40
            O
            00
               30
            o
            o
            z
               20
            <
            o
               10
                                                          I
                                                                                  _l_
225      250     275      300     325      350     375     400      425

                                      VOLUME  TREATED, 1000 GAL.
                                                                                          450
475
500
                                                                                                                  525
                 FIGURE 10-REMOVAL  OF TOC  BY  ACTIVATED  CARBON  IN  AEROBIC  AND  COMBINED  ANAEROBIC-AEROBIC
                           OPERATION

-------
Table 6

Column


Aug 28-Sep
Oct 1-Oct
Oct 31-Nov
Nov 12-Nov
Dec 1-Dec
Dec 30-Jan
Jan 30- Feb
Feb 9 -Feb


Ib TOG
Charged
30
30
11
30
29
29
8
26
24.6
24.6
7.6
13.2
19.7
21.7
7.8
11.9

Aerobic
Ib TOG
Removed
17.7
20.4
5.9
9.9
12.5
10.8
3.5
7.9
TOG REMOVED BY CARBON

Anaerobic An aerobic- Aerobic
wt% TOG Ib TOG Ib TOG wt% TOG Ib TOG
Removed Charged Removed Removed Charged
72
83
78
75
64
50
45
67
24.3 14.9 62
27.3 19.5 72









13.6
18.1
21.9
8.7

Ib TOG wt% TOG
Removed Removed



8.3 61
12.0 66
10.7 51
3.8 44


-------
     4.   in both December and January, coincident with the
          two sharp peaks observed in Figure 10, a marked
          buildup of solids occurred within the columns.
          This was manifested by a rise in effluent TOG to
          10 mg/1 or more, a gradual buildup of pressures
          within the columns, and the presence of iron
          hydroxide precipitate in the column effluents.
          The solids were dislodged from the carbon bed by
          air scouring and backwashing.

     The cleaning of the carbon involved draining the water
down to about 1 ft above the top of the bed, blowing a
stream of air through the column to scour the carbon, and
backwashing with city water at a flow rate of about 10 gpm/
sq ft, about double the u$ual treating rate.  The run was
then continued in the normal manner.

     The cleaning treatment removed considerable iron from
the last two columns, but little from the first two.  It is
believed that some iron used in coagulation entered the columns
and was precipitated as ferric hydroxide inside the last two
columns.

     After the air scouring, the effluent TOC dropped to the
7 to 8 mg/1 level that had been obtained before the upset,
and the pressure inside the columns dropped significantly.
The pressures in the four aerated columns dropped from 22 to
15 psig before washing, to 16 to 9 psig after washing.

     During the entire two-week period before the second
cleaning in February, the TOC of effluents from both columns
averaged about 15 mg/1.  The near zero weather was too cold
to permit stopping feed flow through the lines to air-scour
the columns.  During this period much grey material, which
was believed to be bacterial remains, was carried from the
column.  This may have been due to the destruction of a large
fraction of the bacterial population due to the severe weather.

     After the second cleaning, 7 to 8 mg/1 TOC product was
again obtained from the aerobic columns and the product quality
remained good throughout the balance of February (see Table 6).
From February 12 through March 1, TOC removal in the columns
averaged 67% of the TOC charged.  During this period, the
average TOC of the primary effluent was 71 mg/1 and the average
for the effluent from the carbon column was 9 mg/1, giving an
overall TOC removal by the combined coagulation and carbon
treatment of 89%, which approached that which was attained when
the columns were started in September.  During this period,
the amount of clarified primary effluent charged to the columns
increased from 535,000 to 604,000.
                            - 37 -

-------
     Figures 11 and 12 show the cumulative quantities of TOC
removed by the aerobic and anaerobic-aerobic systems.  Overall,
the two systems removed nearly the same amount of TOC during
the test period, (36 vs 34 Ib of the 60 Ib charged), but the
12-ft anaerobic section was much less effective than the 12-ft
aerobic section.  Thus, the aerobic 12-ft section removed
25 Ib or 42% of the charged TOC, whereas the anaerobic removed
only 17 Ib or 28%.   The relatively good overall activity of
the anaerobic-aerobic system was therefore due to the fact
that the second 12 ft, which was aerobic, removed 17 Ib of
TOC as compared to only 11 Ib in the last 12 ft of  the aerobic
system.

     Product from the anaerobic-aerobic systems had neither
the H2S odor nor the instability that characterized the
product from the all-anaerobic system.  Table 7 shows that
the H?S content of the anaerobic-aerobic effluent was not
significantly higher than that of the aerobic effluent,
despite the pronounced H^S content of product from  the anaerobic
sections.  On standing 24 hours at room temperature, the
anaerobic-aerobic effluents did not develop the haze observed
in anaerobic effluents, as described above.
                          Table 7

        H2S CONCENTRATIONS IN EFFLUENTS FROM AEROBIC
       AND COMBINED ANAEROBIC-AEROBIC CARBON SYSTEMS

                  H2S Concentration, mg/1
          Feed
                         Columns
Date

11/11
   13
   16
   18
   23
   30
PE
3.0
1.4
1.0
1.0
1.2
1.2
CPE
1,
1
0,
0,
Aerobic
A-l
2.1
0.9
0.4
	
	
A-
1.
0.
0.
0.
0.
2
7
7
7
9
7
A- 3
0.5
0.7
0.5
	
	
A-
0.
0 .
0.
0.
0.
4
4
3
3
3
5
An aerobic- Aerobic
U-l
9.0
5.5
1.2
3.1
2.8
U-
12.
6.
2.
5.
4.
2
1
9
4
2
5
U-
0.
0.
0.
0.
0.
3
9
7
7
5
5
U-4
0 .9
0.5
0 .5
0.5
0.7
0.7  	
1.0  0.9  0.5  0.7  0.7
4.1  5.8  0.7  0.5
                            -  38  -

-------
 I


00
           35-
         m
           30-
         Q
         O
         cr
         u
         Q- 25
UJ
H


Z

cr
           20
           15
         O
         UJ
         UJ 10
         a:


         o
         o
                                                   AEROBIC
             0      5     10     15     20     25     30     35    40     45     50    55     60

                                   TOC  CHARGED  DURING  TEST  PERIOD, LB

            55     60     65     70     75     80     85     90    95     100    105    110    115

                                    TOTAL  TOC  APPLIED  TO COLUMNS, LB


             FIGURE  II —CUMULATIVE  REMOVAL OF  TOC  IN  24 FT EXPANDED  BEDS  OF ACTIVATED

                        CARBON  IN  AEROBIC AND ANAEROBIC-AEROBIC  OPERATIONS

-------
o
I
        CD
            55
 5


60
10


65
15     20     25     30     35     40
 TOC APPLIED  DURING TEST PERIOD, LB

70    75     80     85     90     95
  TOTAL  TOC  APPLIED  TO COLUMNS, LB
                                                                      45
100
       50
105
      55
no
      60
115
             FIGURE 12—CUMULATIVE REMOVAL  OF TOC  IN  12 FT EXPANDED BEDS OF ACTIVATED
                        CARBON IN  AEROBIC  AND ANAEROBIC  OPERATION

-------
>-.   Comparison of Activated Carbon and Non-Activated
     Coal Derivative Under Aerobic Conditions	

     After aerobic bacterial activity had been found to give
improved removal of organics from sewage streams by carbon,
a study was begun to determine the extent to which properties
of the carbon affected bacterial activity within a column.
For this study, the active carbon in the anaerobic-aerobic
system was replaced with an anthracitic, unactivated filtering
material.

     The properties of the coal-derived product are compared
with those of activated carbon in Table 1.  The activated
carbon was 16 x 40 mesh, and the coal 8 x 20 mesh.  The coal
had a packed density of 56.4 Ib/cu ft, which was nearly double
that of the activated carbon.  Presumably because of its
coarser size, the higher density coal columns did not give
higher back pressures than the carbon in the upflow operations.

     The major differences between these two carbons was in
their sorptive properties.  The iodine numbers of fresh
activated carbon and coal, 1270 and 150, respectively, were
indicative of the relative adsorptive properties of these
two carbons.  By operating the two sets of columns aerobically
under identical conditions, it was possible to determine the
relative organic removal efficiencies of carbons having good
and poor adsorption properties.  However, at the time when
the comparison with the coal was started, the activated
carbon column had already treated 535,000 gal.  (5,350)
volumes of clarified primary effluent in 5.5 months of steady
operation.

     The coal behaved completely differently from fresh
activated carbon in the columns.  Unlike the fresh activated
carbon which gave products with TOG of 5 mg/1 or less for the
first 100,000 gal. treated,  (see Figure 6) the initial pro-
ducts attained with fresh coal had a TOC of over 20 mg/1, and
this dropped rather slowly, perhaps indicating a gradual
buildup of sorbed organics and bacteria within the columns
(see Figure 13).
                                                    *
     As the comparison continued, the performance of the un-
activated carbon columns improved and approached that of the
partially spent activated carbon columns.  The average TOC
removal in March through May is shown in Table 8.  In the
March through May period, TOC removal by the partially spent
carbon was higher than the 45 to 50% shown in Table 6 for the
December 30 to February 8 period.

-------
-C:
rv>
           o>
           E

           z
           o
           m
           tr
o
z
             40
             30
             20
             10
                       10
                                                                     ACTIVATED

                                                                     CARBON

                                                                         I
                     20       30      40      50       60      70

                                            DAYS   ON  STREAM
80
                                                                                         90
                 100
110
               FIGURE  13—RELATIVE  TOC  REMOVAL   IN  NON-ACTIVATED  AND PARTIALLY  SPENT  ACTIVATED

                           CARBON  COLUMNS

-------
                           Table  8




REMOVAL OF TOG WITH NON-ACTIVATED AND  SPENT  ACTIVATED CARBONS







                            TOC of Effluents,mg/l     % TOG Removal
Average TOC
of Feed,
Month mg/1
March 6 4
April 57
May 74
i
Co
i
Average TOC of Non- Spent Non- Spent
Clarified Feed, Activated Activated Activated Activated
mg/1 Carbon Carbon Carbon Carbon
16.8 9.5 7.7 43.5 54.3
15.5 7.7 6.7 50.4 57.8
21.6 11.9 9.7 45.0 54.7



-------
     The carbon column effluents showed the same low  (6-7 mg/1)
average TOC throughout March and April that had been obtained
several months earlier, but these low TOC values probably
were due to the low TOC of the column feeds.  Whereas,
clarified primary effluents normally had a TOC of about
30 mg/1, the TOC averaged about 16 mg/1 during much of March
and April.  The low TOC of the feed lowered the TOC removal
efficiency, as was also noted above on page 29.

     Late in April, clarified feed in the storage tank developed
a haze, and slime appeared on the walls of the vessel.  This
condition had been observed during hot weather periods in
work conducted on previous contracts, but when it occurred
in the current study the weather was not hot.  During the
period in which this unstable feed was charged to the column,
the column product had a TOC of almost 10 mg/1, and turbi-
dities of 10 to 15 JTU, compared to the normal 1 to 4 JTU
range.

     Throughout this study, the filter was backwashed
routinely twice a week to prevent the formation of septic
growths within the bed.  The backwash water normally contained
much reddish iron oxide.  During this period when the feed
appeared to be unstable, the filter backwash contained black
material which would normally indicate a septic condition.
Septicity did cause poorer TOC removal in the past, and
could have accounted for the poor TOC removals during this
period.  However, unlike other septic conditions, this one
could not be corrected by repeated backwashings.  The Ewing-
Lawrence Sewerage Authority reported abnormal operations at
the same time, so that the problem may have been associated
with some contaminant in the feed.  The only abnormality
noted in our feed analyses was a number of high pH values
for the primary effluent.  This stream normally had a pH
between 6.5 and 7.5, but in May readings between 8.0 and
8.8 were common.

     The activated carbon columns were backwashed on June 1
to determine whether the activity of the carbon could again
be improved.  This backwashing brought no improvement in
the carbon column activity, and the project was discontinued
before another backwash could be attempted.

     The relative cumulative TOC removals by the activated
and unactivated carbons are shown in Figures 14 and 15.  The
TOC removals that had been achieved when the activated carbon
was fresh are plotted as dotted lines for comparison.  The
partially spent 24-ft bed of activated carbon, which had
already removed 84 Ib of TOC when this comparison was started,
removed 56% of the charged TOC, during this test period, as

-------
VJ1

I
              0

              120
5

125
10     15     20     25     30     35     40
    TOC APPLIED  DURING COMPARISON PERIOD, LB
130     135    140    145     150    155     160
    TOTAL  TOC  APPLIED  TO ACTIVATED  CARBON, LB
45
165
50

170
55

175
                                                                                               82
              FIGURE 14—CUMULATIVE  REMOVAL  OF TOC IN 24FT  EXPANDED  BEDS OF ACTIVATED  AND
                         NON-ACTIVATED  CARBONS  IN  AEROBIC  OPERATION

-------
CD
CO
z
o
CO
cc
  30
25
  20
CO
UJ
cc
>-
m
Q
UJ
(T
O
O
   15
   10
   0
                                                                       PARTIALLY SPENT
                                                                       ACTIVATED CARBON
                                                     FRESH UNACTIVATED
                                                     CARBON
86
                                                                                     81
                                                                                        76
                                                                                        71
                                                                                        O
                                                                                        m
                                                                                        cc
                                                                                        z
                                                                                        UJ
                                                                                        Q.
   >

   _l
   <

   I-
   (T

   Q.
                                                                                        >-
                                                                                        m
                                                                                        UJ
                                                                                        >
                                                                                        o
                                                                                        66
                                                                                   I
    0       5       10      15      20     25      30     35      40     45      50     55

                      TOC  APPLIED   DURING  COMPARISON  PERIOD, LB.

    120     125     130    135     140    145     150     155     160     165     170     175

                      TOTAL  TOC  APPLIED  TO  ACTIVATED  CARBON,  LB


    FIGURE  15—CUMULATIVE  REMOVAL  OF TOC  IN  I2FT  EXPANDED  BEDS  OF  ACTI-

                 VATED  AND  NON-ACTIVATED   CARBONS  IN  AEROBIC   OPERATION
                                                                                     61
                                                                                           o
                                                                                           o

-------
compared to 44% for the unactivated material.  Fresh activated
carbon removed an average of 80% of the first 50 Ib of TOG
charged, which was double the removal achieved with the
unactivated carbon.  In 12-ft beds, the fresh activated
carbon also removed twice as much TOC as the unactivated
carbon, the respective removals being 64 and 32%.  The 12-ft
sections of spent activated carbon removed 42% of the charged
TOC.

     The analyses of the spent activated carbon and coal are
shown in Table 9.  The activated carbon taken 'at the end of
the runs from both the anaerobic-aerobic and the aerobic
columns had essentially the same analyses, despite the fact
that the aerobic columns had been on-stream almost twice as
long as the anaerobic-aerobic columns.  Both carbons gained
considerable weight during use.  After devolatilization at
750°C, weight of the carbon increased from 10 to 20%.  This
probably was due to accumulated iron salts within the beds.
As anticipated, the iodine numbers of the dried and devolati-
lization carbons dropped 50 and 30%, respectively.

     Unlike the activated carbons, the non-activated carbon
showed no gain in volatile matter or ash, showed a much
smaller gain in weight during use, and showed no drop in
iodine number during use.  These trends were not associated
with differences in on-stream lives, since the non-activated
carbon columns were on-stream as long as the anaerobic-
aerobic columns.

     It is not known why the devolatilized non-activated
carbon samples had lower iodine numbers than dried samples.
This may be a peculiarity inherent in the iodine number
determination for relatively inactive materials.


D.   Summary of Carbon Column Operations

     The aerobic columns were operated without interruption
for nine months during which time 889,607 gal.  (8,896 carbon
bed volumes) of chemically clarified primary effluent were
treated and in which the three comparison runs described
above were conducted.  Data for the removal of TOC and SOC
in these runs are summarized in Table 10.  In the three test
periods listed, the aerobic activated carbon system removed
90.1, 85.2 and 87.2%, respectively, of the charged TOC, and
81.2, 72.4 and 74.6, respectively, of the charged SOC.  Lowest
removals of both TOC and SOC occurred during the colder months.
For the entire nine-month period, average TOC removal was
87.2%, and average SOC removal was 75.5%.

-------
                                          Table 9

     ANALYSES OF SPENT  ACTIVATED AND NON-ACTIVATED CARBONS FROM THE PILOT OPERATIONS
Carbon
Source
                                        Iodine Numbers
Proximate Analysis
     Pittsburgh
     Filtrasorb
         400

i     Anaerobic-
     Aerobic
co    Activated
i     Carbon

     Aerated
     Activated
     Carbon
Filt-0-
 Cite #1

Aerated Non-
Activated
Carbon
Column

1
2
3
4
1
2
3
4

] 1
I 2
3
J 4
Mois-
ture
1.
41.
43.
43.
46.
45.
43.
49.
47.
9.
23.
23.
20.
17.
3
0
2
8
8
2
6
7
4
6
0
9
6
4
Vola-
tile
Matter,
Wt. %
Dry
Basis
1
16
13
15
9
19
14
11
10
4
5
5
4
4
.0
.0
.3
.5
.2
.2
.7
.7
.1
.1
.6
.8
.3
.5
, Ash ,
Wt.%
Dry
Basis
6.
10.
12.
8.
17.
10.
11.
16.
16.
19.
16.
15.
18.
15.
75
6
9
8
7
0
3
1
8
3
5
4
2
3
Drained
Wt. ,
Ib.

232.
200.
187.
188.
220.
219.
212.
190.
	
202.
233.
181.
163.

a
7
i
8
9
4
4
9

9
5
6
5
Dry
Wt. ,
Ib.
85
136
116
107
103
123
126
109
102
154
158
178
144
135
a
. 1
. 8
.7
.0
.8
.5
.8
.8
a
•
.3
.0
.0
.0
Devola-
tized
Wt. ,
Ib.

114.4
96.5
92.0
94.5
101.2
109.2
98.0
93.3
	
	
	
	
	
                                                                       1270

                                                                         450
                                                                         610
                                                                         610
                                                                         740

                                                                         410
                                                                         490
                                                                         660
                                                                         680
                                          150

                                          155
                                          155
                                          145
                                          155
                                                 800
                                                 870
                                                 850
                                                 850

                                                 750
                                                 750
                                                 860
                                                 870
                                                                                      70
                                                                                      70
                                                                                      70
                                                                                      70
  Weight charged to each  column.

-------
                                                   Table 10
vo
 I
            Date
            System
            Charge,  Gal.
            Organics Charged
            Removed by Clarifi-
              cation
            Removed in Columns
            Total Removed

            Date
            System
              arge, Gal.
            Organics Charged
            Removed by Clarifi-
              cation
            Removed in Columns
            Total Removed

            Date
            System
            Charge, Gal.
            Organics Charged
            Removed by Clarifi-
              cation
            Removed by Columns
            Total Removed
MOVAL OF TOC AND SOC FROM PRIMARY EFFLUENT BY
CHEMICAL CLARIFICATION AND CARBON TREATMENT
August 28, 1970 to November 2, 1970
Anaerobic
TOC
Ib
118
66
34
100
.
217,
%
Carbon
107
SOC
Ib.
.0 — 55.0
.4 56.3 9.4
.5 29.2 30.8
.9 85.5 40.2
November 10 ,
%
17.
55.
73.
1970
Aerobic Carbon
TOC

Ib.
212,
464
SOC
% Ib . %
115.5
1 65.0 56.2
9 39.2 33.9
0 104.2 90.1
to February 8 ,
53.8
10.8
32.9
43.7
1971
20.0
61.2
81.2
Anaerobic-Aerobic Carbon
TOC
172
109
34
144
.2
.9
.8
.7
299,
810
SOC
79.1
63.8 28.5
20.2 28.4
84.0 56.9
February 8 ,
Aerobic
Ib
178
126
22
149
323,
. TOC
.0
.5
.8
.3
71.1
12.8
83.9
Coal
002
Ib.
73.7
27.3
22.6
49.9
TOC
174.2
36.0 110.8
35.9 37.7
71.9 148.5
1971 to June
SOC
37.
30.
67.

0
7
7
Ib.
199.
141.
32.
173.
303,
348
80.
63.6 28.
21.6 29.
85.2 58.
6, 1971
361,
TOC
2
4
3
7
71.0
16.2
87.2
693
Ib
82.
30.
30.
61.
SOC
0
6
4
0
6
7
9
6
35.6
36.8
72.4
SOC
37.2
37.4
74.6

-------
     During the nine-month period in which the aerated
activated carbon columns were operated, a total of 488.5 Ib
of TOC was passed through the adsorption system.  Clarification
removed 317.2 Ib or 64.9% of the TOC, and the carbon 109.2 Ib
or 22.3%.  The carbon thus removed 63.6% of that TOC not re-
moved in the clarification step.  During this same period,
the feed contained 216.4 Ib of SOC.  Clarification removed
70.1 Ib of SOC or 32.5%, and carbon adsorption 93.2 Ib or
43.0%.  This is equivalent to 64.4% removal of SOC not re-
moved in the coagulation step.  One peculiarity in the data
in Table 10 is the improvement in clarification efficiency
which occurred with time.  Since the clarification system
and coagulant dosage were not varied during this study, this
is probably associated with changes in the nature of the feed.

     Table 10 also shows that with fresh carbons the anaerobic
carbon system removed 11% less TOC and 9% SOC than the aerobic
system.   The unactivated carbon removed 44.3 and 48.7%,
respectively, of the TOC and SOC not removed by clarification.
During this same period the partially spent carbon which had
already treated over 515,000 gal. removed 55.9 and 59.6%
of the charged TOC and SOC, respectively.  By contrast, fresh
activated carbon removed 71 and 76% of the charged TOC and
SOC, respectively,in treating the first 300,000 gal.  of feed
pumped through the columns.


E.   Regeneration of Polishing Column Carbon

     When used in an adsorption system, fresh activated carbon
initially removes essentially all dissolved organics;  but in
a relatively short time, the effluent from the columns con-
tains from 4 to 6 mg/1 of TOC.  In the earlier carbon contacting
studies cited previously,  small columns containing 1-ft deep
beds of fresh activated carbon effectively removed for a
period of 24-hours most of the residual TOC that was  not
removed by 24-ft of partially spent carbon.   The organics
adsorbed in short carbon beds evidently were adsorbed only
weakly and at a slow rate.  If removal of these organics
and regeneration of the carbon could be achieved easily and
inexpensively, the use of a short regenerable polishing column
would offer a treatment scheme suitable for those applications
which demanded nearly complete removal of organics from water
for  reuse or recycle.

     Since the earlier studies had already demonstrated that
polishing columns removed residual traces of dissolved organics,
the present work dealt primarily with studies that would
establish the weight of material that could be sorbed on the

-------
polishing column carbon before significant breakthrough
occurred, and if simple regeneration techniques which would
make polishing column operation economically  feasible could
be developed.  The relatively short active on-stream life
of polishing columns made the thermal reactivation of the
carbon impractical.

     Typical operating data are summarized in Tables 11 and
12.  As expected, the active life of a polishing column
carbon increased as the organic content of the feed decreased.
Thus, with feed in Run B averaging less than  7 mg/1 TOG,
Table 12, the effluent contained less than 2 mg/1 TOG after
10,000 gal. had been treated, and was under 3 mg/1 after
treating 20,000 gal.  By contrast, such high quality product
was seldom achieved in Run A with feed which almost always
had TOG of at least 9.5 mg/1, Table 11.

     The spent carbon from the 6-in. glass polishing columns
was taken to the laboratory for regeneration studies.  The
activity of the regenerated material was measured in 24-in.
long glass columns having a one-square-inch cross-sectional
area.  Results were compared with those for fresh activated
carbon.  The test solution was the final effluent from the
carbon columns which had TOG of 7 to 9 mg/1.

     Table 13 shows typical removal of TOG from column
effluent with fresh activated carbon in the laboratory columns.
Initially, virtually all organics were removed by the carbon
bed, but a small leakage occurred after only  4 liters were
treated, and product TOG increased until the effluent had
about the same organic content as the charge.  Organic re-
moval in the laboratory column was less than that achieved
in the pilot polishing column, possibly due to leakage and
channeling in the smaller columns.

     The two best regenerants found in this study were sodium
hydroxide solution and 4% hydrogen peroxide.

     Although sodium hydroxide removal adsorbed organics
readily, as evidenced by the immediate buildup of organics
in caustic regeneration solutions, it was difficult to wash
from the carbon.  Thus, the wash water used to flush 300 ml
of carbon that had been treated with 12.4 pH caustic still
had a yellow color after the 35th wash with 500 ml of cold water.

     Table 14 shows results obtained with a carbon that had been
regenerated with 11.0 pH NaOH solution.  Unlike the fresh
carbon, the regenerated carbon never gave an effluent with
less than 2.5 mg/1 TOG.  When the same batch of carbon was
regenerated a second time with caustic solution, a minimum TOG
of 4 mg/1 resulted.  The iodine numbers of carbon from the
first and second regeneration were 1030 and 1000, respectively,
as compared to the original value of 1040.


                            - 51 -

-------
                                         Table  11
          Date
                          Time
Started  2/17/71  10:50 AM
 I

VJ
         2/17
         2/18
         2/19a
         2/20
         2/22
         2/23D
         2/24
         2/26
         2/27
                        11:38 AM
                         9:00 AM
                        10:15 AM
                         5:15 AM
                         9:30 AM
                        10:55 AM
                         9:30 AM
                         4:45 AM
                         8:50 AM
FIELD STUDY WITH A 4-
Min.
Elapsed
per Interval
48
1,282
1,295
1,860
2,415
1,645
1,355
3,315
1,850

Gal.
Charged
60
2,095
1,609
1,298
4,178
2,251
976
5,327
1,032
Cumulative
   Gal.
 Charged
     60
  2,155
  3,764
  5,062
  9,240
 11,491
 12,467
 17,794
 18,826
   Rate
gal./min,
   1.3
   1.6
   1.3
   0.7
   1.7
   1.4
   0.7
   1.6
   1.1
   TOC
Feed  EfT.
 9.5
 9.5
10.0
12.5
10.0
15.0
10.0
13.0
 8.5
5.0
3.5
2.0
2.0
2.5
5.0
3.5
7.5
4.0
           Column down 3 hours on 2/19.

           Column down 10 hours on 2/23 due to blocked  feed  lines.
         NOTE:  When the rate deviated from  the specified  1.0  gpm rate,  the column
                was operated for a minimum of  30 minutes at  the  correct  rate
                before sampling for TOC analyses.

-------
                           Table 12
Date
1971

3/25
3/26




3/27



3/28



3/29


3/30



3/31




4/1

4/2
ISHING COLUMN FIELD STUDY


Time
9 :30 AM
10:00 AM
11:00 AM
12:00 Noon
2:40 PM
3:40 PM
4:30 PM
7:15 PM
10:30 PM
8:15 AM
10:15 AM
Off for 4.
7:40 AM
5:11 PM
8:45 PM
2:05 AM
12:10 PM
8:35 PM
9 :50 AM
8:35 PM
8:18 AM
4:37 PM
9:18 PM
8:20 AM
10:57 AM
Off for 9
7:40 PM
9:04 AM
4:20 PM
8:50 PM
Min.
Elapsed per
Interval
30
30
60
60
160
60
50
165
195
585
120
5 hours
1,015
570
215
320
605
505
735
705
700
500
280
660
157
hours
700
804
436
270
WITH A

Gal.
Charged
34
33
65
56
205
63
52
197
249
687
139

971
651
226
333
711
443
703
397
753
383
275
526
156

1,155
1,047
534
331
Cumulative
   Gal.
 Charged


    34
    67
   132
   188
   393
   456
   508
   705
   954

 1,641
 1,780
 2,751
 3,402
 3,628

 3,961
 4,672
 5,115

 5,818
 6,215

 6,968
 7,351
 7,626

 8,152
 8,308
 9,463

10,510
11,044
11,375
Rate
Gal./
Min.
1.1
1.1
1.1
0.9
1.3
1.0
1.0
1.2
1.3
1.2
1.15
.95
1.15
1.05
1.05
1.15
.85
.95
.55
1.05
.75
1.0
0.8
1.0
*+
1.65
1.3
1.2
1.2
TOC
Feed
6.0
6.0
6.0
4.8
4.2
3.6
3.6
4.7
5.3
3.6
3.6
11.5
7.0
6.0
6.5
4.0
11.2
7.5
5.7
4.7
7.1
7.0
8.5
7.7
7.7
6.5
8.2
9.0

Eff .
1.8
1.2
Q.6
	
1.2
0.6
0.6
1.2
1.2
1.8
1.2
1.2
1.8
2.4
2.4
1.8
3.0
1.2
6.0
0.6
1.2
1.2
1.2
3.6
2.4
1.2
1.8
2.4
                             -  53 -

-------
Table 12  (Cont'd)
POLISHING COLUMN FIELD STUDY WITH A
Min .
Date Elapsed per Gal.
1971 Time Interval Charged
4/3
4/4

4/6
4/7
4/8
4/9

4/14
4/15
4/16
4/17
4/18
4/19
4/20
4/21
4/22
7 :57 AM
4:07 PM
9:05 PM
11:58 AM
9:10 PM
Off for 35
4:07 PM
10:00 PM
8:40 AM
8:03 PM
8:20 AM
8:05 AM
5:00 PM
Off for 3
8:17 AM
9 :15 PM
8:35 AM
9:00 PM
8:40 AM
9:00 PM
4:50 PM
12 :05 AM
9:40 PM
8:30 AM
10:30 PM
8:30 AM
9:30 PM
8:30 AM
4 :14 PM
8 : 70 AM
9 :45 PM
667
490
300
893
552
hours
480
353
640
683
737
1,425
535
days
1,440
780
680
745
700
740
1,190
1,155
575
650
840
600
780
660
465
965
805
771
525
225
1,023
575

514
360
538
629
808
1,121
566

1,870
590
785
1,033
896
900
1,432
1,387
675
760
1,125
545
850
714
482
1,117
909
LOW TOC FEED - RUN
Cumulative Rate
Gal. Gal./
Charged Min.
12,146
12,671
12,896
13,919
14,494

15,008
15,368
15,906
16,535
17,343
18,464
19 ,030

20,900
21,490
22,275
23,308
24,204
25,104
26,536
27,923
28,598
29,358
30,483
31,028
31,878
32,592
33,074
34,191
35,100
1.15
1.07
.75
1.15
1.05

1.05
1.0
0.85
.95
1.1
0 .8
1.05

1.3
.75
1.15
1.4
1. 3
1.2
1.2
1.2
1.15
1.15
1.35
0 .9
1.1
1.1
1.05
1.15
1.15
B
TOC
Feed
8.0
11.2
16.5
6.0
22.4

10.0
12.3
8.2
4.7
4.7
7.0
6.0

8.2
9.4
10.
11. 8
8.2
7.0
7.0
6.5
7.0
6.0
11.8
8.9
7.0
6.5
7-0
8.9
9.4
Eff .
2.4
2.4
8.2
2.4
12 .5

3.6
4.2
3.0
2 .4
1. 8
2.4
3.0

3.0
4.2
3.6
8.2
6.0
2.4
4.7
5.3
5.3
3.6
6.5
4.7
4.2
2.4
4.2
4.7
5.3

-------
                          Table  13

     REMOVAL OF TOC FROM  ON-LINE  COLUMN EFFLUENT WITH
       FRESH ACTIVATED  CARBON IN  LABORATORY COLUMN
     Volume of
     9 TOC Feed
     Charged, 1.

        2.5
        1.5
        1.0
        1.5
        1.
        1.
        1.0
         Cumulative
           Volume
         Charged, 1.

            2.5
            4.0
            5.0
            6.5
            7.5
            8.5
            9.5
         TOC of
         Product

         < 1.0
           1,
           2,
  0
 .0
2.5
5.5
7.5
9.0
                           Table  14

    REMOVAL OF ORGANICS BY POLISHING CARBON REGENERATED
           WITH 11.0 pH SODIUM HYDROXIDE  SOLUTION
   First Regeneration
                        Second Regeneration
Cumulative
Volume of
9 mg/1 TOC
   Feed
Charged, 1

    4
   19
   20
 TOC of
Product,
 mg/1

  2.5
  3.0
  4.5
Cumulative
Volume of
9 mg/1 TOC
   Feed
Charged, 1

    4
    5
     TOC of
    Product,
     mg/1

       4
       7
                            -  55  -

-------
     Ho way was found to improve the action of NaOH solutions
on spent carbon.  An 8.0 pH NaOH solution effected no regenera-
tion.  Carbon treated with 9.0 pH NaOH solution for 60 min.
and then washed in a beaker with running tap water for 30 min.
removed only 35% of the charged organics.  Repeated washings
with hot water did not cut the washing time and gave the same
organic removal.  Incompletely washed carbon that had been
treated with 12.5 pH NaOH solution removed no organics.  The
same result occurred when the incompletely washed carbon was
neutralized with dilute HCl and washed with water until the
effluent had a pH of 6.7.

     All other basic solutions studied were less effective
than sodium hydroxide.  Carbon treated with 10% Na2C03 removed
a maximum of 50% of the organics, and the Na2C03 was no easier
to remove by washing than NaOH.  Carbons washed with concen-
trated ammonia and saturated lime water removed no organics.
Carbons treated with KOH solutions of 11 and 12 pH removed
less than 50% of the charged organics.

     Spent polishing column carbons reacted vigorously with
hydrogen peroxide solutions.  Spent carbon reacted with 15%
by weight H202 and backwashed removed no TOC.  Carbon treated
with 4% H202 and backwashed for 30 min. with tap water initially
gave a product with a TOC of 5.5 mg/1, but then produced 2 1
of effluent with 3.5 mg/1 TOC and 2 1 with 3.0 mg/1 TOC as
shown in Table 15.  It is not known why the first 5 1 charged
to the column had a relatively high TOC.
                          Table 15

          REMOVAL OF ORGANICS BY POLISHING COLUMN
               REGENERATED WITH 4% BY WT. HO
Run No.             I                           II
            Unacidified Carbon         Acid-Treated Carbon
         Cumulative                  Cumulative
         Volume of       TOC of      Volume of       TOC of
         9 mg/1 TOC     Product,     9 mg/1 TOC     Product,
          Charged        mg/1        _Charged        mg/1

             5            5.5            3            4.5
             7            3.5            5            1.5
             9            3.0            6            2.5
            11            3.0           12            3.5
                                        13            5.0

-------
     Part of the reaction between carbon and H202 was
attributed to reaction with iron salts that were associated
with the carbon.  Therefore, a sample of carbon was treated
with dilute HC1 to dissolve the iron and was washed before
being reacted with 4% H202.  Following a period of poor TOC
removal, this carbon gave the best organics removal obtained
in any regeneration study, as shown in Run No.II, in Table 15,
but its effective sorptive life was short.  Thus, only three
liters of product with TOC of 1.5 to 2.5 mg/1 were obtained.
After the acid-H2C>2 treatment this carbon had an iodine
number of 1040, the same as fresh carbon.

     Washing repeatedly with hot tap water gave no regenera-
tion of spent carbon.  Carbon heated in boiling water for
30 min. reduced the TOC of five liters of feed from 8.5 to
4.5 mg/1, and of an additional four liters to 5.5 mg/1.  After
blowing steam down a bed of carbon inside the 1-sq.in. column
for 5 hours, the carbon reduced the TOC of feed from 12 to 9
mg/1.  During the steaming, the TOC of the condensate dropped
steadily from 105 mg/1 in the first hour to 57 mg/1 in the
fifth.

     Carbon treated with HC1 and H2S04 removed less than
one-third of the charged TOC.  Even poorer results were
obtained with carbon soaked in one-tenth normal potassium
permanganate and ammonium persulphate solutions.

-------
                        SECTION VII

                   MISCELLANEOUS ANALYSES
     Miscellaneous analyes were run at various times during
this study to observe effects resulting from the conditions
studied.

     Table 16 shows BOD removal accomplished by chemical
clarification and treatment in aerobic and anaerobic columns.
Except for the January values when the columns were known
to be performing poorly, the system including aerobic columns
consistently removed about 90% of the BOD and the system in-
cluding anaerobic columns only slightly less.  Table 17 shows
that clarification and treatment with partially spent
carbon, which had treated over 700,000 gal.  (7000 volumes) of
sewage when the tests were made, removed 93% or more of the
BOD.  A system including unactivated carbon gave an average
removal of .93%.

     The relative stabilities of the various effluent streams
are indicated in Table 18 which shows the turbidities of the
various streams at the time they were sampled and after being
allowed to stand exposed to air for 18 to 30 hours.  Some of
these data were shown in Table 3.  Primary effluents
generally become clearer on standing due to settling of
the suspended material, but clarified primary effluents
became more turbid.  Effluents from the aerated columns
invariably became clearer on standing, due to settling of
grayish solids which were presumably biological remains.  The
effluents from the unaerated columns showed a different
behavior before and after the last two columns were made
aerobic.  When the entire system was anaerobic, the turbidi-
ties of samples from the fourth column increased from 1 to
as much as 48 JTU on standing.  However, after the first two
months when the last two carbon columns were run aerobically,
the effluents showed no increase in turbidity on standing.

     Although 140 mg/1 of FeCl3 were used to clarify the
primary effluent, the column effluents contained only about
3 mg/1 of iron (Table 19).  The concentration of ferrous
iron in the water increased as the effluent passed from the
second to the fourth column.

     Essentially no phosphorus reduction occurred in the columns,
as column effluents had about the same phosphorus contents
as clarified primary effluent  (Table 20).  The amount of
phosphorus removed in the clarification of the primary effluent

-------
                            Table  16

                   BOD REMOVAL  IN  AERATED AND
               UNAERATED ACTIVATED CARBON COLUMNS

                           BOD, mg/1
Sample

Primary Effluent
Clarified P.E.
Aerated Carbon Co1.2
Aerated Carbon Col. 4
% BOD Removed
-naerated Carbon
  Col.2
Unaerated Carbon
  Col. 4
% BOD Removed
       10/22/70 10/30/70 11/12/70 12/8/70 1/21/71
(1)
99
44
19.5
8.0
92
130
62
16
13
90
113
48
18
12.5
91
92
35
21.5
11.0
88
70
40
25
16
77
          26.5

          14.0
          86
29
21

14
88
27.5

10
90
34.5

17
76
                            Table  17

             BOD REMOVAL IN SPENT  ACTIVATED CARBON
                AND NON-ACTIVATED  CARBON  COLUMNS

                           BOD, mg/1
Sample

Primary Effluent
Clarified P.E.
Aerated Activated
  Carbon, Col. 4
% BOD Removal
Aerated Non-Activated
  Carbon, Col.4
% BOD Removed
         3/3/71   4/2/71   4/14/71   4/23/71
124
24.5
4.5
97
4.5
97
103
26
4
97
15
86
68
17
5
93
4.5
93
140
26
5
96
4.
97




5

                              -  oO  -

-------
                                                 Table  18

                                STABILITIES OF EFFLUENTS FROM AEROBIC  AND
                                ANAEROBIC COLUMNS RELATIVE TO TURBIDITIES


                                    10/13/70          10/16/71         10/23/70          11/19/70         2/24/71
                               Initial   30 hrs  Initial 24 hrs   Initial   18 hrs   Initial  30 hrs  Initial  18 hrs

Primary Effluent

Clarified Primary Effluent

Aerated Carbon
   Column 4

Unaerated Carbon*
   Column 4

Unactivated Carbon
   Column 4                                                                                           0.5      0.8



All turbidity values  in  this  table  are  in Jackson Turbidity Units (JTU).

* After 11/10/70, the  first two columns  were kept anaerobic and  the
  last two columns were  kept  aerobic.
-
0.5
10.5
1.0
11.
11.
4.
48.
0
0
5
0
0.8
0.8
19.0
1.6
32.0
32.0
7.8
31.0
42.0
1.1
11.0
1.3
7.2
13.0
1.8
5.6
40.0
1.3
4.8
3.0
18
16
2
3
.0
.0
. 8
.8
46
0
6

.0
.4
.5
_
15.0
1.5
3.0
_

-------
                          Table  19

                  IRON IN COLUMN EFFLUENTS

                        May 20,  1971


Iron Content, ppm                      Fe             Total

Effluent

     12-ft Carbon                      0.59           1.25
     24-ft Carbon                      1.69           2.83
     12-ft Coal                        0.15           2.26
     24-ft Coal                        2.05           2.64
                          Table 20

            TOTAL PHOSPHORUS OF EFFLUENT  STREAMS

                  Total Phosphorus, mg/1  P


                      10/15/70 11/24/70 1/20/71  2/2/71 2/4/71

Primary Effluent         9.6       5.1     13.4    >25.0  >20.0
Clarified P.E.           0.7       0         1.7    >17.0    9.5
Aerated Carbon
   12-ft                 0.7       0.3       1.7    >17.0    7.8
   24-ft                <0.3       0.5       2.3    >17.0    8.5
Anaerobic Carbon
   12-ft                 0.7       0.3       1.8    >17.0    8.8
   24-ft                 1.0       0         2.2    >17.0-    9.5

-------
varied with the phosphorus concentrations in the feed. The con-
centration of ferric chloride used in the clarification removed vir-
tually all phosphorus from primary effluents having about 5 mg/1 P;
but as the phosphorus content of the primary effluent increased,
the concentration of phosphorus in the clarified primary
effluent also increased.

     Table 21 shows the effect of carbon treatment on the
concentration of nitrogen compounds in effluent streams.
Ammonium nitrogen concentrations underwent little change,
but organic nitrogen contents dropped as organic compounds
were removed by the carbon beds.  A small reduction in nitrate
concentrations occurred in both the aerobic and anaerobic
columns.
                          Table 21
           NITROGEN ANALYSES OF EFFLUENT STREAMS
Nitrogen, mg/1

Stream, 10/15/70
    Primary Effluent
    Clarified Primary Effluent
    Aerated   - 12-ft Carbon
              - 24-ft Carbon
    Anaerobic - 12-ft Carbon
              - 24-ft Carbon
    Ammonium    Organic
       8.6
       8.8
       8.7
      10.6
       8.4
       7.7
        3.2
        1.7
        1.7
        1.2
        3.2
        1.0
                    Nitrates as N, mg/1

                                   10/15/70  1/7/71  2/2/71
Primary Effluent
Clarified P.E.
Aerated Carbon
    12-ft
    24-ft
Anaerobic Carbon
    12-ft
    24-ft
Anaerobic-Aerobic Carbon
    12-ft
    24-ft
<0
<0

<0
<0
<0.5
<0.5
3.5
5.3

4.0
2.7
          2.0
          0.5
10.0
 4.8

 2.5
 2.0
        2.5
        2.0

-------
                        SECTION VIII

                   DISCUSSION OF RESULTS


A.   Effect of Aerobic Conditions Within Carbon Columns

     In this study, aerobic conditions inside a carbon column
gave better performance than anaerobic conditions.  The
aerobic columns removed more of the carbonaceous impurities
in sewage and produced more stable products.  Unlike anaerobic
effluents, aerobic effluents contained no hydrogen sulfide.
The combined anaerobic-aerobic operation, in which the first
one-half of the carbon column system was operated anaerobi-
cally and the other aerobically, produced essentially the
same organic removal and product quality as the all aerobic
system.  However, since the activity of the 12-ft anaerobic
section was much less than that of the corresponding aerobic
section, the effectiveness of the combined anaerobic-aerobic
system depended essentially on the effectiveness of the
aerobic section, and there was no advantage in operating
with the mixed anaerobic-aerobic system.


B.   Effect of Carbon Properties on Column Efficiencies

     In aerobic columns, the treating effectiveness of fresh
activated carbon was much superior to that of unactivated
carbon, as the fresh activated carbon columns removed approxi-
mately 50% more TOC than the unactivated coal product.

     The different behavior of these two types of carbon
showed that at least two different mechanisms are involved
in carbon treatment.  The low organic removal obtained ini-
tially with fresh non-activated coal material indicated that
very little adsorption occurred on this material and that
only gradually did effective biological activity develop on
its surfaces.  On the other hand, the excellent removal obtained
immediately with the activated carbon indicated that this
material rapidly adsorbed organics and facilitated rapid
development of bacteriological activity within the columns.
The high degree of organic removal observed at the beginning
of the run with activated carbon was due principally to
adsorption.  However, the fact that the first 12-ft of this
material was able to remove essentially its own weight of
organics indicated that some of the material removed by
adsorption was subsequently degraded by biologic activity
at the carbon surfaces, which in turn freed these surfaces
for further adsorption.

-------
C.   Comparison of Results from Current and Previous Studies

     In the period in which clarified primary effluent was
tested in the earlier study,^ product with 3 to 5 mg/1 TOC
was obtained consistently during four summer months.  Although
product from the current study had TOC of 6 to 8 mg/1 after a
few weeks, these higher values are not indicative of poorer
carbon column performance than that experienced previously.
On the contrary, overall TOC removal in the columns from
August 28 to November 30 averaged 76.9%, compared with 70.4%
during the earlier study.  Because the TOC of the feed was
higher during this study, 70 Ib of TOC was charged to the
columns in the three months between August 28 and November 30,
1970; whereas only 62.1 Ib was charged during the four-month
period between April 29 and September 2, 1969 in the previous
study.  The improved TOC removal in this study may reflect
better control of aerobic conditions within the columns.
It may also be that higher concentrations of organics in the
feed promote better organics removal in the carbon beds.


D.   Cleansing of Carbon in the Columns

     During the nine months in which the carbon columns were
operating, the carbon was cleansed by air scrubbing and
backwashing three times.  The air scouring and backwashing
probably aided column operation by improving the hydraulic
characteristics of the system, and by freeing the adsorptive
surfaces of dead, non-active biomass and inorganic precipi-
tate.  After long periods of operation ratholing and ineffi-
cient contacting probably occurred within the columns,
tending to reduce the effectiveness of the carbon treatment.
Further, ferric hydroxide that precipitated in the columns
likely interfered  with flow and/or covered the carbon with
a coating which inhibited its sorptive activity.   This is
supported by the observation that much red ferric hydroxide
was recovered from the columns after backwashing.

     Cleansing of the carbon improved its effectiveness and
increased its useful life.  To take full advantage of this
finding, columns should be designed so that they can be
backwashed and air scoured efficiently.  The carbon chosen
should have sufficient abrasion resistance to withstand
several cleansings.  Otherwise, large amounts of fine
material produced during the air and water scouring would
be carried from the columns.
                            - oo -

-------
E.   In Situ Biologic Regeneration

     One of the most notable findings of this study is the
remarkable increase in the effective capacity of activated
carbon facilitated by aeration in expanded-bed operation.
Carbon has been found to function efficiently as an adsor-
bent for markedly longer periods than those predicted either
by saturation data obtained from measurements of adsorption
isotherms or from reports by others of sorption capacities
obtained in the treatment of wastewaters in packed-bed
adsorption systems.  The enhanced effective capacity of
the carbon observed in the present studies is, as suggested
by Weber, Hopkins, and Bloom,5/10 in earlier studies of
the treatment of wastewaters in expanded-bed adsorbers,
apparently attributable to bacterial activity on the
surfaces of the carbon.  This activity can be tolerated,
indeed encouraged, in expanded-bed adsorbers,  whereas
packed-bed adsorbers tend to clog and foul as a conse-
quence of biological growths.  In expanded-bed adsorbers,
biological activity can be enhanced by addition of air or
oxygen, a treatment which could cause air-binding of packed-
bed adsorbers.

     As noted earlier in this report, maintenance of aerobic
conditions in the solution phase passing through adsorbers
was determined to enhance the quality of the effluent, both
from the point of view of TOC reduction and the prevention
of H2S generation.  It was not immediately obvious, however,
whether aerobic solution conditions would affect the
operating capacity of the carbon.  The data obtained in
this study indicated that the biological activity within
the aerobic columns increased the useful life of the carbon
significantly.

     The precise mechanisms by which "in situ" biologic
regeneration functions to achieve the remarkably effective
sorption capacities observed in the current study, are
difficult to isolate and characterize.  However, some
logical deductions can be made from the data observed earlier
by Weber et al.5/l° and the results of the present pilot
investigation.

     First, the principal separation process operating in
these systems is adsorption from solution onto the surfaces
of the activated carbon.  Evidence of this is the rapidity
with which organic matter is removed to a high degree from
the aqueous phase, and observations on the lower effective-
ness of other solid materials, such as sand and anthracite
coal, which have less capacity for adsorption than does
                            - 67 -

-------
activated carbon.  Active adsorbents do not function merely
as expanded-bed trickling filters.  Once organic material
is adsorbed on the carbon, a highly enriched substrate
environment is created at the surfaces.  This environment
fosters the growth of micro-organisms which utilize the
organic matter as food source, and, in so doing, free the
surface of the carbon for continued function as an adsorbent.
This then extends the effective operating life, or apparent
capacity, of the carbon.

     Adsorbed organic substances can be held at the surface
of the carbon for long periods of time.  The effective
detention time of such materials in the sorption system may
be several days or several weeks, rather than the few minutes
residence time of the fluid from which they are adsorbed.
With this long residence time, micro-organisms on the carbon
surface have sufficient time to acclimate to even relatively
non-degradable or biologically-resistant materials.  Further,
and by the same reasoning, because the biologic activity takes
place on the carbon surface rather than in the bulk solution
and because the substrate is so abundant, the micro-
organisms are protected, or buffered, against shock loads
of toxic materials and/or wide variations in waste composi-
tion.

     The microbiologic growth responsible for substrate
breakdown at the surface and consequent surface regeneration
might be anaerobic.  There are several observations which
support this supposition.  First., there is no evidence of
rapid and massive sludge accumulation on the carbon, as
would be expected for a completely aerobic system.  The
sludge that does build up does so gradually and in relatively
small amounts.  Second, the accumulated sludge scoured from
the carbon by vigorous agitation contained dark material that
could have been an anaerobic residue.  Third, even though
the introduction of air or oxygen enhances adsorption
effectiveness and apparent capacity, the amount of oxygen
required is much less than that which would be required for
aerobic oxidation of the mass of organic matter extracted
from solution and concentrated on the carbon surfaces.
Based on the dissolved oxygen analyses of the column feeds,
an oxygen mass balance does not support an hypothesis of
aerobic regeneration.

     One possible explanation of the observed phenomena may
be that shown in Figure 16.  It is assumed for this explana-
tion that two biologically active films surround each carbon
particle.  The interior film is anaerobic; the external film
is aerobic.  Adsorbable molecules pass through the films to
the carbon surface.  At the surface there is an opportunity
                           - 68 -

-------
             ANAEROBIC
              .BOUNDARY \
                   FILM
                           \
                             \
                        AEROBIC \
                       BOUNDARY \
                           RLM
BULK
SOLUTION
                                            C02 + H20
EVENT SEQUENCES'  I. DIFFUSION  OF  LARGE  ADSORBING ORGANIC MOLE-
                     CULE @ TO SURFACE  OF CARBON.

                   2. ANAEROBIC DEGRADATION OF LARGE MOLECULE
                     ® TO SMALL  MOLECULE  © .

                   3. DIFFUSION  OF  SMALL  NON-ADSORBING  ORGANIC
                     MOLECULE  (§)   AWAY  FROM SURFACE OF CARBON.

                   4. AEROBIC  DEGRADATION  OF  SMALL  MOLECULE (§)
                     TO C02  AND H20.
FIGURE 16 -SCHEMATIC  INTERPRETATION  OF  THE MODE OF  IN
           SITU  BIOLOGIC  EXTENSION  OF ADSORPTION CAPACITY

-------
for anaerobic degradation.  The anaerobic degradation will
;.ot Le complete, however, and low molecular weight degrada-
tion products, such as organic acids and alcohols, will form.
rlr.e low molecular weight degradation products so formed
nave an inherently low relative energy for adsorption, and
thus will back-diffuse from the surface through the external
boundary film to the bulk solution.  Under anaerobic conditions
in the bulk solution there would be no aerobic film and these
degradation products would join other materials which
were never adsorbed from solution, and pass out in the
effluent from the sorbers.  If, however, oxygen or air is
added to the expanded-beds of carbon, the solution phase
as well as the outer layer of the boundary film on the carbon
can be maintained in an aerobic state.  Aerobic micro-organisms
in the outer film could oxidize the outward diffusing products
of the anaerobic decomposition.  This is consistent with the
observation that the effluent from the aerobic adsorbers con-
tains less TOG leakage thaa does that from the anaerobic
adsorbers.  The leakage in the aerobic case would be com-
prised principally only of relatively non-adsorbable organics
initially present in the wastewater.  This interpretation is,
of course, speculative.  Considerable further study is required
before the effective mechanisms can be unequivocally defined.


F.   Polishing Column

     The rapid breakthrough of organics in effluents from
carbon columns indicated that the carbon quickly became
saturated with some fraction of organic matter.  One method
for achieving a closer approach to an organic-free renovated
water would consist of passing effluent from an on-line
adsorber through a short polishing column of fresh carbon
which is changed or regenerated frequently to remove the
residual TOC comprised of weakly adsorbing, easily displaced
organic substances.  To be practical, this concept requires
a low-cost, rapid regeneration.

     None of the regeneration studies conducted offered promise
for ultimate development of a rapid, simple reactivation
system.  Sodium hydroxide and 4% F^C^ were the best carbon
regenerants found.  However, neither restored more than 75%
of the original sorption capacity of the spent carbon, and
such activity losses could not be tolerated because of the
low sorptive efficiency use of the carbon in a polishing
operation.  In addition, the regenerants were washed from
the carbon with such difficulty that in many cases the volume
of wash water greatly exceeded the volume of feed treated.
                             -  70 -

-------
G.   Treatment Costs

     On the basis of results obtained previously, a physico-
chemical wastewater treatment system was proposed.5  That
system, shown in Figures 17 and 18, included coagulation of
raw sewage with lime and treatment of the clarified effluent
with carbon adsorption units to remove dissolved organics.
Lime was the chosen coagulant because it gave good clarifi-
cation and a rapid-settling sludge.  Unlike FeCl3, lime can
be recovered from settled sludge by incineration.  The
organic solids in the sludge would be destroyed in this
incineration.

     Although the beneficial effects of aeration were only
partially realized previously, that process design included
open tanks with trough-type overflows at the surface of the
contacting basin to induce aeration of the wastewater during
treatment, thus ensuring aerobic conditions inside the
columns.  The 6 to 10 mg/1 oxygen concentration used in
the current study was selected because this is in the range
of the solubility of oxygen in water aerated at ambient
temperatures.

     The investment and operating costs for a 10-mgd com-
bined coagulation-activated carbon treatment plant, like
that shown in Figures 17 and 18, were presented in detail
previously.   The prices of the individual items in that cost
estimate were adjusted to an ENR Index of 1300, which was
representative of anticipated prices in 1970.  The current
estimate is based on an ENR Index of 1400.  Nothing in the
current study affected clarification costs, but the change
in the ENR Index increased capital costs for this section of
the treating plant from 1.93 to 2.08 million dollars.  This
increased clarification treatment operating costs from the
11.6<=/1000 gal. cited previously to 12.5C/1000 gal.

     The current study did make changes in the projected
active life of the carbon and in the potential carbon treatment
costs.  The earlier study had assumed a carbon dosage of 500
Ib/million gal. treated.  Experience with aerated carbon
columns has shown that much longer life can be obtained with-
out sacrificing product quality.  Thus, during April when the
340 Ib of carbon in the aerated columns had already -Created
approximately 800,000 gal. of water, the TOC of the product
averaged less than 8 mg/1.  Almost 90% of the TOC was removed,
and over 90% of the BOD was removed in the total treatment system.
Actually, the carbon treated 890,000 gal. of water satisfactorily
without reactivation.  If all the carbon were to be reactivated,
                            -  71 -

-------
          RAW SEWAGE
r\j
I
                              -AIR
GRIT
                     AERATED GRIT CHAMBER
                     AND FLOCCULATOR
            TREATED
            WATER
                                                  CLARIFIER
HP0
                JHrn
          LIME
          SLURRY
          TANK
    MULTI-
    HEARTH
    FURNACE
                             DRUM
                             FILTER
                       SLUDGE
                       THICKENER
                                                                      2-STAGE
                                                                      CARBON  CONTACTORS
                                                                      EXPANDED BEDS
                                                             I  CARBON REGENERATION
                                                       D   D
rn
                                                              DRAIN
                                                              TANK
                                                       MULTI-   STORAGE
                                                       HEARTH  TANK
                                                       FURNACE
              FIGURE 17 — PROPOSED  SCHEME  OF TREATMENT  OF  RAW SEWAGE  BY  CHEMICAL CLARI-
                         FICATION  AND  ADSORPTION  ON ACTIVATED  CARBON

-------
                         RAW SEWAGE
       COAGULANT
FILTER
          L FURNAC
SLUDGE

THICKENER
                              GRIT REMOVAL
«•	AIR



     FLOCCULATION CHAMBER
                                 SETTLING CHAMBERS
                                                              gl
                            Q

                            3

                            9
                        —'I v  ADSORBER FEED

        !  EXPANDED BED AD-  NfANKS 8 PUMPS
        CARBON

        REGENERATION
                      OOOOO


                      OOOOO
                       ALTERNATE
                       PACKED BED ABSORBERS
                              TREATED WATER
   FIGURE 18 	 PROPOSED  ARRANGEMENT OF  PROCESSING UNITS
                          - 73 -

-------
the dosage would be 380 Ib/million gal.  In a system of
contactors in series, however, only one half or even one
quarter of the carbon need be reactivated at one time.  If,
in this case, part of the carbon had been reactivated, and
those contactors placed after the contactors in which the
carbon was not reactivated, more acceptable product water
could have been obtained.  The carbon dosage would then have
been less than 380 Ib/million gal.  It is impossible to obtain
an accurate figure for the "steady state" carbon dosage
without actually carrying out further operation with reacti-
vation.  It can probably be assumed, however, that the
dosage would be about 250 Ib/million gal.  It should be
emphasized that this value is for the particular waste-
water used in the study.  Variations with wastewater from
other locations can be expected.

     Table 22 shows the estimated capital costs for the
carbon-treating systems, and Table 23 shows the estimated
operating costs for treating clarified municipal sewage in
aerobic carbon beds.  The first column of Table 23 is based
on the same carbon dosage of 50-0 Ib/million gal. used in
the earlier estimate.  The second column shows the estimated
cost with a projected dosage of 250 Ib/million gal. of effluent
treated.  The savings accruing from the longer treating life
of the carbon are reflected in a 50% reduction in the amount
of makeup carbon needed and a similar cut in the fuel and
power required to regenerate the carbon.  A small additional
reduction resulting from a smaller regeneration system was
not included.  The net effect of these economies is to lower
the treatment costs for the carbon adsorption treatment by
approximately 0.7^/1,000 gal.  Based on a 7.6
-------
                          Table 22

      ESTIMATED CAPITAL COSTS FOR TREATMENT OF CLARIFIED
      RAW SEWAGE IN ACTIVATED CARBON AND UNACTIVATED COAL

                      Material Columns
Basis:  10 mgd
Activated
 Carbon
              Unactivated
              Coal Product
Equipment
    Adsorption System
    Regeneration System

Piping
    Adsorption System
    Regeneration System
            Total

Instrumentation
Painting and Insulation
Buildings and Structures
            Physical Costs

Engineering
    Home Office
    Field
    Contractors
            Base Cost

Contingency, 15% of Base Cost
Auxiliary Facilities
    Power
    Fuel Oil
    Roads, Walks and Fence
Total Plant Cost
Carbon Charge
$  185,900
   139,600
              $  185,900
325
209
14
224
$ 549
26
10
136
$ 174
$ 723
$ 126
123
43
$ 292
$1,016
152
$1,168
$ 25
20
60
$ 105
$1,273
288
,500
,600
,500
,100
,600
,900
,700
,500
,100
,700
,600
,000
,000
,600
,300
,400
,700
,000
,000
,000
,000
,700
,000
185,
209,
209,
$ 395,
19,
1,
98,
$ 126,
$ 522,
$ 91,
88,
31,
$ 211,
$ 733,
110,
$ 843,
$- 25,
60,
$ 85,
$ 928,
50,
900
600
600
500
800
900
900
600
100
400
800
300
500
600
000
600
000
000
000
600
000
                                      $1,561,700
              $  978,600
                            - 75 -

-------
                                        Table  23

                     ESTIMATED ANNUAL OPERATING COSTS  FOR TREATING
                   CLARIFIED MUNICIPAL SEWAGE  IN AERATED  CARBON  BEDS

                                                                           Unactivated
 Sorbent                                           Active Carbon	     Coal Product

 Ib/million gallons treated                        500          250
 1.  Operational Labor*                          $  43,300      $  43,300        $ 43,300
 2.  Maintenance Labor, 3% Plant
       Physical Costs                              20,100        20,100          15,800
 3.  Maintenance Materials, 2% Plant
       Physical Costs                              13,500        13,500          10,500
 4.  Maintenance Supplies, 15% of 2 & 3             5,000         5,000           3,900
 5.  Supervision, 15% of 1                          6,500         6,500           6,500
 6.  Payroll Overhead, 15% of 1 & 2                 9,500         9,500           8,900
 7.  General Overhead, 30% of 1, 2 & 6             21,900        21,900          20,400
 8.  Insurance, 1% of Plant Physical Costs          7,200         7,200           5,700
 9.  Carbon Makeup, 5% @ 28C/lb                    27,500        13,750             	
10.  Fuel, @ $0.50/million Btu                     13,000         7,000             	
11.  Power, @ $0.01/kwh                            13,500         6,750           6,700
12.  Amortization, 24 yrs @ 6%                   124,200       124,200          77,800
13.  Coal Makeup, 2% @ $50/ton                   		      		           1,000

         Total Treatment Costs                   $305,200      $278,700        $200,500
         Treatment Costs - C/1000 gal.              8.36          7.64           5.50


* 2 Shift Men -I- 2 day men @ $4.00/hr.

-------
for the carbon system.  Columns with unactivated carbon or
other inert material could probably be operated at a cost
of 5.5C/1,000 gal- and the combined coagulation-carbon
treatment cost would be about 18C/1,000 gal.
                          - 77  -

-------
                         SECTION IX

                      ACKNOWLEDGMENTS
     This report was prepared by Louis D. Friedman,
Walter J. Weber, Jr., Ralph Bloom, Jr. and Charles B.
Hopkins.  Pilot plant operations and physical and
chemical analyses were performed by W.H. Behn, E.M.
DiPolvere and E.R. Smith.

     Dr. Carl A. Brunner of the Environmental Protection
Agency was Project Officer.

     The authors express their appreciation to Dr. Brunner
and L. Seglin of FMC Corporation for their constructive
criticisms during the course of the work and on this
report.
                           -  79 -

-------
                          SECTION X

                          REFERENCES


 1.    Weber, W.J., Jr., "Fluid-Bed Columns for Sorption of
      Persistent Organic Pollutants".   Proc.  Third Inter-
      national Conference on Water Pollution Research, Vol.  I,
      Water Pollution Control Federation, Washington, D.C.,
      1966.

 2.    Hopkins, C.B.,  Weber, W.J., Jr., and Bloom, R., Jr.,
      "A Comparison of Expanded-Bed and Packed-Bed Adsorption
      Systems", Report No. TWRC-2, 1968, 77 pp., R.A. Taft
      Water Research  Center, U.S. Department of the Interior,
      Cincinnati, Ohio.

 3.    Parkhurst, J.P., Dryden, F.D., McDermott, G.N., and
      English, J., "Pomona Activated Carbon Pilot Plants",
      JWPCF, 39, 10,  part 2, R70, (October 1967).

 4.    Schlechta, A.F., and Gulp, G.L., "Water Reclamation
      Studies at the  South Tahoe Public Utility District",
      JWPCF, 39, 78  (May 1967).

 5.    Hopkins, C.B.,  Weber, W.J., Jr., and Bloom, R., Jr.,
      "Granular Carbon Treatment of Raw Sewage", Water Pollu-
      tion Control Research Series ORD17050DAL 05/70, U.S.
      Department of the Interior, Federal Quality Administration.

 6.    Winneberger, J.H., Austin, J.H., and Klett, C.P.,
      "Membrane Filter Weight Determination", FWPCA Publi-
      cation, U.S. Department of the Interior, 1967.

 7.    "Standard Methods for the Examination of Water and
      Wastewater", 12th Edition, American Public Health Asso-
      ciation, New York, 1965.

 8.    "Manual on Industrial Water and Industrial Wastewater",
      American Society for Testing Materials, Philadelphia,
      Pennsylvania, 1963.

 9.    "Industrial Water Pollution Control", W.W. Eckenfelder,
      Jr., McGraw-Hill Book Company, New York, N.Y.  (1966),  p.149.

10.    Weber, W.J., Jr., Hopkins, C.B., and Bloom, R., Jr.,
      "Expanded Bed Adsorption Systems for Treatment of Sewage
      Effluents".  Paper presented at 63rd Annual Meeting of
      American Institute of Chemical Engineers, December 1,
      1970, Chicago,  Illinois.
                             - 81 -

-------
                         SECTION XI

                        PUBLICATIONS
Friedman, L.D., Hopkins, C.B., Behn, W.H., and Bloom, R.,
Jr., "Improving Effectiveness of Adsorption Systems for
Sewage Treatment".   Presented at the Annual Meeting of
the New Jersey Water Pollution Control Association,
May 6, 1971, Atlantic City, New Jersey.
                             -  83  -

-------
   SELECTED WATER
   RESOURCES ABSTRACTS
   INPUT TRANSACTION FORM
                   1. Report No.
                                    3. Accession No.
                                    w
   4.  Title

   IMPROVING GRANULAR CARBON TREATMENT
   7.  Author(s) Friedman,  L.D., Weber,  W.J.  Jr.,
   Bloom, R. Jr.,  and Hopkins,  C.B.
   9.  Organization
   FMC Corporation
   Princeton, New Jersey


  12.  Sponsoring Organization

  15.  Supplementary Notes
                                    5. Report Date

                                    6.
                                    8. Performing Organization
                                      Report No.

                                    10. Project No.

                                    #17020 GDN	
                                    11. Contract/Grant No.

                                    #14-12-901

                                    13. Type of Report and
                                      Period Covered
  16. Abstract
  The magnitude  and effects of biological activity in expanded carbon beds
  used for direct clarification/adsorption treatment of wastewater wre
  investigated.
  Primary sewage effluent was coagulated with  FeCl3, clarified, filtered
  and charged  to carbon columns.   Clarification  removed 60%  of the total
  organic carbon (TOO  and 55% of the BOD.  Carbon activity  comparisons
  were made in identical series  of four columns  containing 24 ft of carbon,

  Aerobic conditions achieved by bleeding 6 to 10  mg/1 Q£ into the feed,
  enhanced activated carbon's activity and prolonged its effectiveness.   In
  a nine-month pilot-scale study,  overall TOC  removal was 87.2%, BOD re-
  moval exceeded 90%, and effluent TOC or BOD  averaged 8 to  9 mg/1.
  Aerobic activated carbon columns removed 14% more TOC than anaerobic,  and
  over 20% more  TOC than aerobic coal columns.   Products from anaerobic
  columns contained I^S and clouded on standing;  aerobic products did not.
  Estimated treatment costs for  combined clarification/adsorption treatment
  at 10 mgd are  20C/1000 gal. with activated carbon and 18C/1000 gal. with
  coal, which  yields a product with a TOC or BOD of about 12 mg/1.
  17a. Descriptors
  *Carbon Treatment, *Clarification, *TOC  Removal, BOD Removal,
  Activated  Carbon, Unactivated Carbon, Coal,  Aeration, Carbon Regenera-
  tion, Costs,  Carbon Reactivation.
  17b. Identifiers
  17c. COWRR Field & Group  05 D
  18. Availability
19.  Security Class.
   (Report)

20.  Security Class.
   (Page)
  Abstractor   L.D.  Friedman
    21. No. of
      Pages

    22. Price


Institution
                                               Send To :
                                               WATER RESOURCES SCIENTIFIC INFORMATION CENTER
                                               U.S. DEPARTMENT OF THE INTERIOR
                                               WASHINGTON, D. C. 20240
                        Corp., Princeton,  N.J.
WRSIC 102 (REV. JUNE 1971)
                                                GPO 913.261

-------