Ultrasonic Filtration
   Combined Sewer Overflows

The Water Pollution Control  Research Reports describe the
results and progress in the  control  and abatement of
pollution of our Nation's waters.   They provide a central
source of information on the research,  development and
demonstration activities of  the Water Quality Office of
the Environmental  Protection Agency, through in-house
research and grants and contracts  with  the Federal, State,
and local agencies, research institutions, and industrial

Triplicate tear-out abstract cards are  placed inside the
back cover to facilitate information retrieval.   Space
is provided on  the card for  the user's  accession number
and for additional key words.   The abstracts utilize the
WRSIC system.

Inquiries pertaining to Water  Pollution Control  Research
Reports should  be  directed to  the  Head, Project Reports
System, Planning and Resources Office,  Research and
Development, Water Quality Office, Environmental  Protec-
tion Agency, Washington, D.  C. 20242.

Previously issued  reports on the Storm  and Combined Sewer
Pollution Control  Program:

                    Storm Water Pollution  from Urban Land
                    Acti yi ty
                    Combined Sewer Regulator Overflow
                    Faci1i ties
                    Selected Urban Storm Water Abstracts,
                    July 1968  - June 1970
                    Combined Sewer Overflow Seminar Papers
                    Combined Sewer Regulation and Manage-
                    ment - A Manual  of  Practice
                    Retention  Basin  Control  of Combined
                    Sewer Overflows
                    Conceptual Engineering Report -
                    Kingman  Lake Project
                    Combined Sewer Overflow Abatement
                    Alternatives - Washington, D.C.
                    Chemical Treatment  of  Combined Sewer
                    In-Sewer Fixed Screening of Combined
                    Sewer Overflows
                    Selected Urban Storm Water Abstracts,
                    First Quarterly  Issue
                    Urban Storm Runoff  and Combined sewer
                    Overf1ow Pol 1ution
— — —
                      Continued  on  inside  back  cover  ...

       Ultrasonic Filtration of Combined Sewer Overflows

       American Process Equipment  Corporation
             3309 West El  Segundo  Boulevard
               Hawthorne,  California  90250
                       for the


                WATER QUALITY OFFICE
             Program  Number  - 11023DZF
             Contract Number 14-12-195
                      June, 1970
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price 60 cents
                       Stock Number 5601-0071

               EPA/WQO Review Notice
This report has been reviewed by the Water Quality Office
and approved for publication.  Approval does not signify
that the contents necessarily reflect the views and
policies of the Water Quality Office, nor does mention
of trade names or commercial products constitute endorse-
ment or recommendation for use.

A 250,000 gpd compact ultrasonically cleaned microfi1tra-
tion system was tested with simulated combined sewer
overflows and primary treatment plant effluent.  Twenty
35 micron porous polyethylene elements of 0.8 sq ft area
each comprised the system.

Simulated settled combined sewer overflows at Atlanta,
Georgia test site quickly clogged filters with high
ccfncentrations of rust in both influent and fresh back-
wash water.  Thus, feasibility of treating combined sewer
overflows was not demonstrated.  Study indicates use of
more costly porous stainless steel elements would obviate
this problem.

Sufficient test data are available to predict performance
in more suitable water pollution control  applications.
With influent BOD and suspended solids levels of 100 mg/1 ,
or less, ultrasonic filtration using 35 micron plastic
elements can reduce BOD and SS in raw settled sewage,
or effluents from primary and secondary plants by 50%.
A 25 M6D system requires less than 20,000 square feet, and
cost for treating 1,000 gallons is $0.08, excluding pre-

Contractor's novel flotation vortex separator reduced  raw
sewage BOD 45% and influent solids 80%.  This "VORSEP"
could continuously treat combined sewer overflows, and/or
pretreat influent to the ultrasonic microfi1tration system.
A 25 MGD Vorsep treats 1,000 gallons for  $0.03, and requires
10,000 sq ft of space.

This report is submitted in fulfillment of Contract 14-12-
195 between the Environmental Protection  Agency,Water
Quality Office and American Process Equipment Corporation.

Key Words:    Ultrasonic Filtration
              Combined Sewer Overflows
              Vortex Separator

Section 1
Section 2
Section 3
Section 4
 Phase I
 Phase II
Section 5
Section 6
Section 7
Section 8
 Part  I

 Part  II
- Conclusions
- Recommendations
- Introduction
- Discussion
- Dual Canister Test Program
- Design, Fabrication, and
  Operation of Field Unit
- Patents
- Acknowledgements
- Glossary of Terms and
- Appendix
- Phase I Hydraulic Test Data:
- Phase I & Phase II Laboratory
  Analyses: Hyperion & Panama City
Page No

   i i i






 1      PHASE  I  DUAL-CANISTER  SYSTEM                   9

 2      HYPERION  TEST  AREA                             9

 3a     PHASE  II  TRAILOR  -  END VIEW                   15

 3b     TYPICAL  BANK OF 4  FILTER  CANISTERS            15





 8      BASIC  CANISTER DESIGN                         18

 9      PHASE  II  SYSTEM IN  ASSEMBLY                   18


11      PHASE  II  SYSTEM -  SCHEMATIC OF                20

12      VORSEP SKETCH                                 31

13      VORSEP PHOTO                                  32

                      SECTION 1


1. Filtration of settled combined sewage or simulated
combined sewer overflows at a test site In Atlanta, Georgia,
was unsuccessful.  The cause of failure was Identified as
excessive, Irreversible filter clogging by rust particles
Introduced 1n part by corrosion 1n the pretreatment tanks
and piping and In part from the fresh water supply used for
backwashing.  It has subsequently been determined that the
clogging problem was aggravated by the use of Inexpensive
porous polyethylene filter elements having an unusually
large affinity for rust particles as compared to much more
costly porous sintered stainless steel elements.

2.  Effective continuous filtration required waste water
pretreatment by settling or other means to provide an
Influent with no more than 100 ppm of suspended solids and

3.  At the Atlanta test site, a particle size distribution
analysis conducted by the Georgia Tech Research Institute
showed that suspended solids remaining after pretreatment
were primarily unicellular bacteria of less than 5 microns
diameter.  To capture these smaller particles would have
necessitated use of finer-pored filter elements having a
reduced flow efficiency.

4.  The semi-automatic mlcrof11tratlon system 1s best
suited for application to an influent of relatively uniform
quality day-by-day.

5.  Inexpensive, porous polyethylene filter elements were
demonstrated to be functional in rust-free applications as
permanent, recleanable filter media, and are preferred to
more expensive sintered porous stainless steel elements
initially evaluated.

6.  The semi-automatic mlcrof11tratlon system was successful-
ly operated to meet completely its designed operating para-
meters:  Total flow rate of 250,000 gpd; automatically
controlled periodic pressurized backwash; automatically
controlled periodic 1n situ application of ultrasonic energy
to enhance cleaning oT filter elements during backwash; and
filtration efficiencies with 35 micron porous filter elements
as high as 15 gpm/sq ft.

7.  The filtration system (above) removed up to SOX of BOD
and suspended solids from a domestic sewage primary treat-
ment effluent.

8.  A novel vortex separator recently developed by the
Contractor and tested in prototype form both with raw sewage
and mixed liquors obtained after secondary treatment shows
substantial promise In future full-scale installations as a
continuous sedimentation device capable of providing effect-
ive pretreatment of high volume combined sewer overflows.

                      SECTION 2


1. The Atlanta test site, with Us Installed availability
of pumps, power, tanks, and other auxiliary equipment, should
be utilized for further study of combined sewer overflow pre-
treatment processes with the vortex-type flotation system.

2.  The trailer-mounted mlcrof11tration system should be
refurbished to eliminate sources  of filter-plugging rust
and to eliminate excessive maintenance of certain unreliable
automatic valves controlling forward and backwash flows by
substituting more reliable equivalents.  The system should
be refitted with stainless steel  elements in applications
where rust may be encountered.

3.  The refurbished microf11tration system should be further
evaluated for its efficacy in removing pollutants from
selected waste waters; e.g., secondary effluents, industrial
wastes with fibrous or non-colloidal partlculate contaminants,
or chemically flocculated wastes.

4.  Waste waters should first be  analyzed for suspended solids,
per cent dissolved BOD, etc., to  determine the applicability
for utilizing this treatment process.

                      SECTION 3


This project was authorized to provide for the design,
fabrication, and evaluation of a semi-automatic ultrasonic
microf11tration system, of nominal 250,000 gpd capacity, for
field use in treating combined sewer overflows.  The field
site was to be completely equipped with the ultrasonic micro-
filtration unit, a pretreatment unit, and all auxiliary
equipment necessary for a complete demonstration of the

The Contractor's test program was divided Into two phases:

PHASE I -  A test program utilizing a contractor-owned
           dual-canister laboratory scale ultrasonic
           filtration system to provide design data for
           the field unit.

PHASE II-  Performance testing of the field unit at an
           FWQA-approved sewage treatment plant to develop
           operational procedures and evaluate system
           reliability.  The field unit was subsequently to
           be moved to an FWQA-approved field site for
           performance testing with actual combined sewer

The semi-automated ultrasonic microf11tration system was
selected for evaluation in treatment of combined sewer over-
flow discharges because of Its potential in providing ef-
ficient removal of particulate pollutants from the large
volumes of such waste waters encountered during periods of
heavy rainfall.  This potential is Inherent in the system
design, which features permanent, micro-porous filter
elements, automatic forward and backwash flow controls, and
in situ application of ultrasonic energy to enhance cleaning
of the filter elements during periodic short or long backwash
cycles.  The system can provide filter efficiencies of 10-15
gpm/sq ft continuously, provided the influent has received
requisite pretreatment.  An explicit objective of the test
program was to define "requisite pretreatment.M

The dual-canister system had previously been evaluated in
Its earlier development stage with various Influents, includ-
ing raw sewage, under FWQA Contract No. 14-12-23 with
Acoustlca Associates, Inc., the company from which American
Process Equipment Corporation evolved as an Independent
entity in March of 1968.  During this previous program, the

performance of a 7 gpm ultrasonic filtration system with 20
and 50 micron pore size stainless steel  elements was evalu-
ated with ordinary city water and with raw sewage variously
diluted with city water to simulate combined sewer overflows.
With 2 parts of water and one part of raw sewage passed
through 50 micron elements, for example, reductions of 39*
In BOD and 55% 1n suspended solids were realized.  Similar
results were obtained with other filter pore sizes and
dilution ratios.

Based upon the results of the testing accomplished under
Contract No. 14-12-23, the present program to more thoroughly
evaluate the dual-canister system and expand upon Its flow
capabilities was Initiated.

                      SECTION 4


PHASE I - Dual Canister Test Program

A.   Laboratory-Scale Ultrasonic Filtration Unit

The Contractor-owned dual canister laboratory scale ultra-
sonic filtration system consisted Of the following:

     Two stainless steel canisters, each fitted with ASCO
     solenoid-operated valve Influent port, effluent port,
     backwash port, drain port and air vent.

     Canisters manifolded to permit either one or the other
     canister to maintain forward flow with In situ cleaning
     while the other 1s either undergoing a ToVg&r backwash
     cycle or 1s maintained Idle in the clean condition*

     Each canister containing a mlcroporous filter element of
     either porous stainless steel or plastic.  Of cylindrical
     construction, each element typically measured 10 Inches
     long x 2.75 Inches outer diameter and thus had an effec-
     tive filtration surface area of 86 square Inches or
     0.6 square feet.  Each element 1s sealed at the bottom
     and fitted at the top with a connection to the effluent
     and backwash manifold.  Influent flow 1s Into the canis-
     ter, through to the Interior of the filter element, and
     out the top fitting of the filter element to the mani-

     An electrostrictive ultrasonic transducer, driven by a
     350 W output, 22 kHz ultrasonic generator, on the bottom
     plate of each canister.  The transducer 1s activated
     during frequent, short cycles ("sprltz") one second out
     of each 10 or 20 seconds, and during a long-clean
     backwash upon manual command or at any desired Interval
     set by a cam timer system.

     Dial pressure gauge on the Inlet side of each canister.

     An Influent pump to the settling tank rated at 10 gpm
     flow at 30 psl.

     A square box 20-mesh influent screen of stainless steel
     to remove gross objects, such as rags, sticks, etc.

     A backwash  pump  rated at 15 gpm at 100  ps1  maximum.

     Flow meters 1n  the  drain and backwash  lines.

     An electronic  control box Including stepping  switches,
     cam timers  and  electronic timing circuits,  manual  over-
     ride switches  and  Indicator lights showing  the  opera-
     tional  status  of the  filters at any moment.

     A 300-gallon capacity cylindrical  steel  settling tank
     used for pretreatment of raw sewage Influents and  a
     second  Influent  pump  for Introducing settled  sewage
     Into the filtration system Initially rated  at 10 gpm
     flow at 30  ps1  and  later doubled 1n flow capacity.

All  of the foregoing  system components  were  assembled for
Installation 1n  the  detrltor building at the  Hyperion
Treatment Plant, of  the  City of Los  Angeles,  where sources
of both raw  sewage  and  effluent from a  secondary plant
were available conveniently.  F1gs.  1 and 2  show the equip-
ment and the test setup  comprising the  Hyperion  Installation.

B.  Test Program Objectives and Results

Objectives of the dual-canister test program  Involved hydrau-
lic  measurements of  flow rate and head  loss  pressure; maxi-
mizing forward flow;  optimizing pressure and  application  of
backwash water;  and  determining the  need, 1f  any,  for pre-
treatment processes,  Including screening, storage  and floc-
culatlon, all determined with various types  of porous filter
elements.  These tests  were made with a variety  of Influents
Including settled and non-settled raw sewage  variously
diluted with fresh  water and effluent from  the secondary
treatment plant  at  Hyperion.

    (1)  System  Description

    Before discussing detailed test  results,  a brief
    discussion of the operation of the  ultrasonic  filtra-
    tion system  Itself  will be presented.  In this type of
    system using permanent-type elements, rapid  plugging
    of the elements  can  be expected  unless  means for reduc-
    ing and/or eliminating buildup of cake  on the  filters
    are incorporated.  In  early work with the Contractor-
    owned system employed  to filter  drinking  water,  periodic
    Initiation of the "sprltz" sequence was  adequate to
    maintain each filter element 1n  a "like-new1* condition
    during forward  flow.  That sprltz sequence Involved the
    momentary stoppage  of  forward flow  for  1  sec every  10
    to 20 sec during  which Interval  the drain to the canister


was opened simultaneous with the application of ultra-
sonic energy to the Interior of the canister.  Cake
building up on the surface of the filter was dislodged
and placed 1n suspension 1n the quiescent liquid surround-
ing the element, while a small  portion of the Influent
within the canister was exhausted to the drain by gravity
as the solenoid-operated air vent opened.  Under light
conditions of Influent loading, typically 25 mg/1 of
mineral-type suspended solids and sand, the sprltz
renewed forward flow to the point where a periodic
backwash of longer duration was needed only several
times per day.  Similar results were obtained when
filtering various petroleum fuels.

In the Hyperion tests, however, it  became evident short-
ly after tests commenced that because of the far heavier
solids loadings encountered in  settled raw sewage, the
sprltz must be augmented by adding  a pressurized backwash
during the short sprltz sequence.  The addition of pres-
surized sprltz provided a very  significant and substan-
tial improvement in system operation and became standard
in all further Phase I and Phase II work.  With the
addition of this pressurized sprltz sequence aided by
ultra-sound, all subsequent tests evaluating various
filter elements and sequencing  techniques were based upon
maintaining the forward flow rate as close to constant as
possible using pressurized sprltz only, with the longer
backwash being delayed in time  as long as was feasible.

(2)  Initial Phase I Tests

Initial tests during the Phase  I effort Involved the use
of sintered porous stainless steel  elements of 50 and
100 micron pore size.   Such elements are commercially
available in open cylinders and must be joined together
by welding to fabricate elements of the 10 Inches length
used.  Thereafter, end plates and the pipe fittings must
be affixed.  Ordinary  welding techniques proved unsatis-
factory, as the heat involved distorted the sintered
bodies that cracked thereafter.  This difficulty was
rectified by substituting more  expensive electron beam
welding techniques.

Complete data obtained in each  test at Hyperion are
included in the Appendix.  The  sequence of hydraulic
tests can be summarized in the  order they were Investi-
gated as follows:

     a.  Non-pressurized sprltz with stainless steel
         elements - premature clogging.

          b.  Pressurized backwash spritz with stainless
              steel elements - decided improvement.

          c.  Pressurized backwash spritz with Teflon
              elements - premature clogging.

          d.  Pressurized backwash spritz with polyethylene
              elements - superior performance.

Numerically, the hydraulic results for the best of the runs
1n each of the four categories listed above when conducted
with settled raw sewage are shown in the following Table:

PARAMETER                          CATEGORY
WATER/SEWAGE                10:1    1:1    1:1    1:1

FORWARD FLOW @ START,GPM     5.5    5.5    7.5    5.2

START, GPM/SQ FT             9.4    9.4   12.5    8.7

FORWARD FLOW @ END, GPM      1.0    5.2    1.0    5.2

(3)  Phase I System Optimization

During the various tests listed In the Appendix, varia-
tions 1n settling time, sprltz sequence, and backwash
pressure were made as part of the system optimization
process.  Recalling that tests under category (d) re-
sulted in a constant flow situation without need for a
longer backwash cycle from time-to-tlme, 1t can be seen
that this condition required backwash water during
sprltz alone amounting to 15% of forward flow.  Both
clear water and filtered effluent were used for back-
washing without an appreciable change 1n performance.
This amount of backwash water could have been reduced
by allowing the filter elements to clog up gradually
during forward flow to a point where a more frequent
backwash cycle would be required.  This periodic back-
wash procedure was programmed by the control system to
last for three minutes during which time the canister
was  fully emptied and refilled with backwash water pres-
surized at 50 psi three successive times while the
ultrasonic energy was applied continuously.  This seq-
uence restored clogged filter elements to a like-new
condition in every case and was employed routinely at
the  end of each day's testing.  Thereafter, the canis-
ters were filled with fresh water to which liquid
chlorine was added to Inhibit clogging caused by organ-
isms growing during shut-down periods.  Total amount of
water consumed during each complete backwash cycle was
approximately 6 gallons.

An Important objective of the Phase I program was to
select the optimum pore size for the filter elements.
Initial tests with stainless steel elements In the 50
and  100 micron pore size range revealed equal hydraulic
performance for both.  Smaller pored elements were
rejected because of poor flow characteristics.  Thus,
the  50 micron pore size was used.  Subsequent work with
porous polyethylene permitted the use of 30 to 35 micron
pore size elements, a standard value obtained from the
manufacturer.  Ten micron pore size elements were
investigated but were not used, primarily since they
were mechanically unstable in the early stages of the
program, and, secondarily, because of their lower flow
efficiency.  Since the 35 micron elements performed
adequately, they were selected for use 1n the Phase II
system.  It should be recognized that a final selection
of filter pore size depends strongly on the interaction
between actual Influent contamination levels and

hydraulic variables involving the desired forward flow
efficiency, the spritz sequence, the backwashing
sequence and the backwash pressure.  Filtration systems
of this type are best operated under stable influent
conditions, such as those encountered in treating
ordinary domestic sewage.  With widely varying contam-
ination levels in the influent, the system operating
parameters may have to be changed accordingly if
premature clogging is to be avoided.

(4)  Phase I Laboratory Analyses

Also included in the Appendix are measurements of 600
and suspended solids in the filtration system's influent
and effluent.  Using 35 micron elements and settled raw
sewage, a BOD reduction of 39% and a reduction in sus-
pended solids of 15% were obtained.  With similar filter
elements and an influent obtained from the effluent of
a secondary treatment plant, a 50% reduction in BOD was
obtained along with a 35% reduction in suspended solids.
It was clear that the crude settling used with raw -
sewage in itself produced a significant reduction in
suspended solids introduced thereafter to the filter,
as evidenced by the formation of a settled sludge at
the bottom of the sedimentation tank.

(5)  Phase II System Guidelines

As a result of the data obtained during Phase I tests,
the design of the Phase II canisters was enhanced some-
what by increasing the size of the filter elements to
12 inch length and 3 inch OD, an increase in surface
area amounting to 0.8 sq ft as opposed to 0.6 sq ft for
the smaller elements.  The polyethylene filter elements
finally selected were one-piece cylinders with a 3/16
inch wall thickness having plastic-welded end pieces
and fittings.  Previously evaluated polyethylene
sleeves 1/16 inch diameter, which were slipped over
open plastic support frames, were discarded because of
weak structural strength.  Otherwise, the Phase II
system was designed to be similar in all respects to
the Phase I system in terms of ultrasonic power,
control sequencing and general appearance.

PHASE II - Design. Fabrication, and Operation of the
           Phase II Field Unit

A.   Serol-Automatic Ultrasonic Filtration Unit (Fig.3* & 3b.)

The nominal  250,000 gpd filtration unit Included the follow-

     20 stainless steel canisters, each fitted with Influent
     port, effluent port, backwash port, drain, and air vent.

     Canisters manifolded 1n 5 banks of four canisters each,
     each bank 1n turn connected to the system manifold
     through automatic and manual  valves.

     Each canister containing a microporous polyethylene
     filter element.   The cylindrical  elements are sealed
     at the bottom, and fitted at the top with a connection
     to the effluent and backwash manifold.   Influent flow
     1s into the canister body, through the filter element,
     and out the top  fitting to the manifold.

     An electrostrictive ultrasonic transducer, driven by
     a generator in the electrical control  console, on the
     bottom plate of  each canister.  The transducer 1s
     activated during frequent, short backwash cycles
     ("spritz") one second out of  each 10 or 20 seconds,
     and during a long-clean backwash 3 minutes out of
     every 15 minutes.

     Dual pressure gauges:  One of the Inlet side of one
     canister in each bank;  one on each system manifold

     A system manifold fabricated  from 3" Iron pipe, to
     supply Influent  and backwash  flows to  each bank mani-
     fold, and to remove filtrate  and "spr1tzM drain flows
     from each bank manifold.

     An air compressor and air storage tank, supplying
     control air to solenoid valves for operation of the
     automatic flow control  valves.

     An influent pump of 200 gpm capacity,  to supply pre-
     treated sewage to the filters.

     A backwash pump.

     Flow meters in the drain and  backwash  lines on one
     filter bank.

   Fig. 3a - End View - Phase II Trailer
F1g. 3b - Typical Bank of 4 Filter Canisters


     A power supply panel, to supply the control  electrical
     power to the pumps, air compressor, and electronic
     control panel.

     An electronic control panel, Including two 700 watt
     22 kHz ultrasonic generators, stepping switches,
     control circuits, timers, manual  over-rides, and  a
     lighted control  panel array showing the operational
     status of each bank of filters at any moment.

     A pump to inject hypochlorite disinfectant into system
     at time of shutdown, through the  backwash line.

All the foregoing system components were assembled  and
mounted on a 38 foot  flat bed trailer  to provide mobility
for testing at a sewage treatment plant and at a field test
site.  Figs. 4, 5, 6, and 7 illustrate the Atlanta  test
site and the raajtfr system components Installed.

The following figures and drawings Illustrate the design
and fabrication of the trailer-mounted filtration unit.

     F1g.   8                Basic Canister Design
     Fig.   9                Phase II System in Assembly
     Fig.  10                Phase II System Schematic  of
                            Piumbi ng
     Fig.  11                Phase II System Schematic  of
                            Electrical Controls

Operating  Instructions - Microf11tration Unit

1.   Open  all manual  valves between system manifolds and
     each  bank of canisters.

2.   Turn  on power supply.

3.   Start air compressor, set control at 70-75 psi.

4.   Open  valves between system manifold and Influent
     supply, backwash supply tanks.

5.   Turn  on power to electronic control panel.

6.   Open  all vents,  using over-ride switches on
     electronic control panel.

7.   Start influent pump.

8.   Start backwash pump.

                                 F1g. 5 - SHAKER SCREEN S TRASH BOX

                                 CIRCULAR FILTER
                                ' CYLINDER
                  J        L
                    SUPER SONIC

                                     PHASE  II  SYSTEM - SCHEMATIC OF PLUMBING
        FIGURE 10
     "RED VALVE"


                 INFLUENT FROM PUMP  •*-

                 BACKWASH FROM PUMP  -»•

                 FILTERED INFLUENT    <*-

                 DRAIN ("3PRITZ")      «*-

                                TO PREVENT
                                                                       AIR  \ /  AIR
                                                                      DRAIN  HBACKFLUS
                                                                      VALVE / \ VALVE
                                                                                  AIR COMPRESSOR
                                                                                  ELECTRIC MOTOR
              PHASE n SYSTEM-


 9.   Start automatic control sequencing system ( 1 minute
     and 3 minute timers) on electronic control panel.
     Return vents to automatic control.

10.   Using manual valves, adjust Influent pressure to each
     bank of canisters to 25-35 psi.

11.   Using manual valves, adjust backwash pressure to 50 psi.

12.   Observe control panel light array to assure that all
     banks and valves are operating 1n proper sequence.

13.   Observe pressure gauges and flow meters to assure that
     operation 1s normal and stable.

The system now should be 1n fully automatic operation.  The
operator will accordingly have ample time to operate and
check on auxiliary equipment, take samples, maintain log
data for operations.

Shut-Down Procedure

1.   Turn off influent pump.

2.   Turn off air compressor and air to solenoids.  This
     will leave all automatic valves open.  Shut all
     manual valves except backwash valves.

3.   Turn off electronic controls at panel.

4.   Backwash should now be flowing through all 20 canisters.
     Introduce two gallons of hypochlorlte disinfectant
     solution Into backwash stream.  Continue flow until
     chlorine can be detected exiting from the drain line.
     (Disinfectant need not be Introduced for short shut-
     downs) .

5.   Turn off backwash pump.

6.   Shut backwash manual valves.  Shut Influent and back-
     wash valves between system and supply tanks.

7.   Turn off power to electronic control panel.

8.   Turn off power to electrical power supply panel.

B.   Testing of Ultrasonic MlcrofHtratlon System at a
    Panama City, Florida. Sewage Treatment Plant

The 20-canister, 250,000 gpd trailer-mounted filtration
system was temporarily Installed at the M1llv11le Treatment
Plant, Panama City, Florida.  Key features of the Installa-
tion were:

    Influent to system was primary treatment effluent,
    without further pretreatment.

    Backwash was city water pumped at 80-90 psi from a
    pool  reservoir.

    "Spritz" cycle was 1.7 seconds on, 9 seconds off.

Before introducing sewage into the system, operation was
checked using an influent of city water directly from a
main.  All aspects of the system performed according to
design ... automatic cycling, "spritz" timing, long clean
periods with backwash.  Total forward flow through the
system exceeded design capacity, averaging more than 15 gpm
per canister, an efficiency equal to 19 gpm/sq ft.  The
system was operated for 6 hours at a time for the greater
part of a week.

Following this demonstration that the system was mechan-
ically and electrically performing at or  exceeding design
goals, the system was operated with an influent of primary
treatment effluent.  The influent pump installed on the
system was used.  Again, forward flow was 15 gpm per can-
ister, or 240 gpm for the 16 canisters operating at any
given time.  This corresponds to a daily flow rate of
345,600 gpd, well 1n excess of nominal design capacity of
250,000 gpd.

During these filtration tests using primary sewage from
the Millville Plant, two sets of influent and filtrate
samples were taken.  Five day BOD analyses performed by
Millville Plant personnel showed an average reduction of
BOD across the filter system of 51%, from about 90 ppm
to 45 ppm.  All canisters were fitted with 35 micron porous
polyethylene elements.

C.  Field Site Installation and Phase II Test Program

It was decided to site the Phase II twenty canister system
at an appropriate pollution point 1n the Southeast* selected
from a half-dozen potential locations mentioned during
discussions between FWQA and the Contractor.  These Included
Augusta,Savannah, Columbus and Atlanta, Georgia, plus two
others too Inconvenient to reach 1n the ordinary course of
the program.  The selection of the appropriate site was
governed 1nthe main by the availability of a nearby qual-
ified laboratory testing firm, since 1t was not felt proper
to Impose upon the facilities used by the various municipal-
ities for their own analyses.  A further consideration was
to locate an Independent organization to monitor the
Phase II test program and provide an Impartial analysis of

These various considerations narrowed the selection of a
site to Atlanta and the retention of the Georgia Institute
of Technology's Research Institute to provide both program
monitoring and necessary analytical laboratory services.
Accordingly, the Contractor executed a subcontract with the
Georgia Tech Research Institute and obtained as the princi-
pal Investigator the services of Dr. Robert S. Ingols, an
authority in waste water treatment and analysis.  In con-
junction with Dr. Ingols and local FWQA representatives,
the Contractor visited with the Georgia Water Resources
Council and cognizant engineers of the City of Atlanta and
agreed upon a Phase II test site at the MqDaniel Street
outfall not far from Five Points, the heart of the City.
At this outfall, dry weather sewage normally diverted to the
South River Treatment Plant emerges from an antiquated sewer
main built 1nthe 1880's.  During storm conditions, the mix-
ture of sewage and storm water overflows a weir and passes
down Into a creek behind a number of homes on Its way to a
river system near Savannah.  Such overflows occur routinely
throughout years of normal rainfall, but during the extra-
ordinarily dry summer just ended, overflows occurred only

The location of the McDanlel Street outfall 1s on a City
owned cul-de-sac some 1,000 ft from the nearest home and
power and water.  The Contractor was permitted to site his
various equipments on a gravel-filled area adjacent to the
manhole diverting dry weather flow to the treatment plant.

The trailer-mounted ultrasonic mlcrof11tratlon unit was set
1n place at the site.  Required electrical service was
provided, a limited fresh water supply was obtained through
a garden hose from a nearby Boy Scout building, and lighting,
security fencing, and telephone service were Installed.


Auxiliary equipment was procured and installed as follows:

     A WEMCO torque-flow pump to lift sewage from a city-
     owned manhole.

     A booster pump to permit pumping raw sewage at the
     required rates against the high head required by the

     A boom and hoist to permit removal  of the WEMCO pump
     as periodically required to provide access to the
     manhole for city sewer maintenance employees.

     A Link-Belt 20-mesh shaker screen to remove gross
     solids from the raw sewage, and appropriate hoses,
     rakes, and solids collection box.

     Two hemi-cylindrical  settling tanks, connected in
     series to provide up  to two hours'  settling time prior
     to filtration.  Nominal capacity, 10,000 gallons.

     A 300 gallon  fresh water supply tank for backwash,
     sewage dilution, etc.

     Piping and hose connections to bring sewage to the
     field unit and to return filtrate and drain discharges
     to the sewer.

     Weir boxes to measure  flows of the filtrate and drain

     At a later date, a 3-inch water line was installed from
     the nearest city main  over 600 feet away to provide
     fresh water volumes needed for backwashing and pre-
     paring simulated combined sewer overflows.

     The Georgia Tech Research Institute, as subcontractor,
     installed analytical  equipment and monitoring instru-
     ments, including a Honeywell No. W-1Q river monitor,
     pH meter, turbidimeter,  conductivity meter, and a
     non-recording rain gauge.

     A protective  enclosure was built over the control
     panels on the trailer; steps, catwalks, and guard rails
     were added as required for safe operations; and a
     storage shed  for tools and materials was erected.

As noted earlier, the site was selected because of a history
of frequent combined sewer overflows during heavy rainfall
periods.  However, during the time available for operating
the field unit, rainfall was very infrequent and of short
duration, so that no real test with actual combined sewage
was possible.  Attempts to achieve meaningful results with
combined sewage during two very short rainfall periods were

Discussions with the Project Officer at this point led to
a decision to modify the field unit to permit operation with
dry weather sewage, diluted to simulate combined sewer over*
flows.  A first approach was to dilute screened dry weather
sewage with filtered effluent.  This approach was unsucces-
sful.  Detailed analyses of extended runs made with this
configuration were provided by the Georgia Tech Research
Institute.  During these runs, the filtration unit appeared
to be operating normally, but recirculation of the filtrate
to dilute the raw sewage resulted in a rapid build-up of
unfilterable suspended solids within the system.  As a
result, reductions of BOD, COD, and bacteria count across
the filter were low -- in the range of 5 to 25%.  The data
cannot be considered significant as a measure of filter
performance, however, because of the recirculation system
used.  During the course of these runs, turbidity of the
filtrate actually increased, due to a build up in unfilter-
able  suspended solids.

The subcontractor concluded that recirculation of filtrate
as diluent would preclude any meaningful results relative
to filter efficacy.  Accordingly, and in view of a contin-
uing  dry summer and lack of actual combined sewer overflow
availability, it was decided to use fresh water for both
diluent and backwash.  To enable such operation, an addition-
al fresh water supply was necessary, and the contractor
installed a 3-inch line from the nearest city main more than
600 feet away.

Several runs were made with raw, dry weather sewage diluted
with  twice its volume of fresh water.  Fresh water was also
used  for backwash.  Mechanically, the field unit was operat-
ing in accordance with design; and, in fact, throughout the
field test period, the electronic control system, the ultra-
sonic cleaning devices, and other components of the trailer-
mounted filter system performed reliably.  An exception to
this was frequent rupture of rubber sleeves on the air-
operated automatic "Red Valves.11  Replacement of those
valves to provide more acceptable reliability under operat-
ing conditions is recommended for future test programs.

Subcontractor analyses of runs made with fresh water dilu-
tion to simulate combined sewage were at first extremely
encouraging; BOD was reduced 66 to 80%, suspended solids
were reduced to as low as 2 ppm with an Influent containing
72 ppm!

It was, of course, suspected that such data resulted from
a malfunction of the overall system, and, accordingly,
these data were rejected and not reported upon herein.
The Contractor and Subcontractor hypothesized that somehow
the "filtrate" quality was so excellent as a result of
dilution rather than of filtration removal of contaminants,
that further experiments were devised to determine exactly
what was happening.  Conductivity  of Influent, filtrate, and
drain streams was measured, and flow rates of each were
monitored to provide a reasonable materials balance.

The results of these tests proved our hypothesis to be
correct.  The "filtrate" stream was 1n fact 50 to 80% fresh
water, and the drain ("spritz") effluent was 65 to 85%
unflltered influent.  The reason for this anomolous per-
formance was sought and identified.  Fine rust particles
had become lodged permanently in the porous polyethylene
filter bodies, causing a 75 to 90% reduction 1n filtration
efficiency.  Automatic operation of the filter system was
thus effecting a flow-switching process, in which the major
portion of Influent was by-passing the filters and exiting
through the drain line.  Likewise, the major portion of
fresh-water backwash was by-passing and exiting through the
filtrate Hne.

Sources of filter-clogging rust particles were identified.
The pretreatment settling tanks and the system manifold
piping were the major sources, with substantial corrosion
caused by the high pH and other attributes of the excessively
variable industrial-domestic dry weather sewage at the field
site.  A second source of rust particles was the city fresh
water supply itself.  Filter discoloration from rust was
noticeable after passage of only 15 gallons of city water
through a new element.   The plugging rust was not removed
by the normal ultrasonic cleaning action of the system,
and in fact would be removed only by dissolution 1n
concentrated hydrochloric or nitric add.

Particle size distribution analyses of screened settled
sewage at the McDaniel  Street field site, which were made
by Georgia Tech, gave further insight into why the partic-
ular filter did not achieve meaningful  Improvement In
waste water quality.  Following pretreatment (settling),
95% ofthe suspended solids in the sewage were shown by

laboratory analyses to be unicellular bacteria of less
than 5 microns diameter that passed through the 35 micron
filter pores unimpeded.  To capture such small particles
(in the absence of the otherwise overpowering rust) would
have required the use of 1 micron stainless steel elements
of far higher cost and substantially lessened flow capacity.

The clear conclusion from these studies 1s that the Con-
tractor's ultrasonic microflltratlon system 1s not yet
suitable for effective treatment of combined sewer overflows
at the Atlanta site until more 1s known about pretreatment
and rust-handling requirements.  Without pretreatment, the
sewage 1s so variable and at times so strong that filter
plugging precludes reliable operation.  With conventional
time-dependent gravity settling pretreatment, the remaining
suspended solids analyzed 1n Atlanta were primarily uni-
cellular bacteria of less than 5 microns diameter that are
not subject to effective filtration at the desired efficiency
of 10-15 gpm/sq ft.  These conclusions reinforce the
Contractor's recommendation that low cost, Instantaneous
pretreatment processes, such as the proposed vortex separators,
be ultlUzed for such combined sewer overflow treatment
problems followed, 1f necessary, by an ultrasonic filtration

Subsequent to the conclusion of this program and during the
period when the final report was under preparation, a reason
for the rapid, Irreversible plugging of the polyethylene
filter elements by rust became more apparent.  An analysis
of various filtration systems of possible value 1n treating
fresh water, which was conducted by personnel of the United
States Public Health Service, revealed an extreme tendency
of polyethylene to absorb rust by the polyelectrolyte
effect.  Since one of the main causes of fresh water con-
tamination is rust, it has been concluded that stainless
steel elements are to be preferred over polyethylene in
applications where rust is present.

This finding Indicates that polyethylene filters have
value in treating rust-free domestic sewage, but that stain-
less steel must be substituted in applications where rust
can be encountered.  The question remaining unanswered
at this time 1s whether the premature clogging of the
polyethylene filter elements 1n Atlanta could have been
obviated by substitution of stainless steel elements and,
if so, how effective would the filtration system have been
treating combined sewage when compared to its proven
ability to filter raw settled domestic sewage at its full
design capabilities.

D.  Pretreatment Considerations

One of the main factors determining the utility of an
ultrasonic microf11tration system, or any comparable filtra-
tion process, 1s the level of Influent contaminents.  The
results of the Phase I and Phase II tests have Indicated
an approximate upper limit of 100 mg/1 for both BOD and
suspended solids 1n  the influent above which performance of
the filtration system suffers in terms of reduced flow and
added backwash requirements.   Below these levels, filtra-
tion of raw settled  sewage and effluent from primary treat-
ments was accomplished most economically at filtration
efficiencies approaching 19 gpm/sq ft with reasonable
amounts of backwashing water  in the order of 15% of the
forward flow, or less.  This  waste water/fresh water ratio
could possibly be raised substantially by shortening the
sprltz duration or by sacrificing the filtration efficiency,

With these parameters 1n mind, the need for pretreatment
of highly contaminated influents becomes mandatory,
especially when the  levels of contaminents can vary widely
during operating conditions.   It 1s also clear that con-
ventional time-dependent sedimentation processes used for
pretreatment obviates the main advantage of the ultrasonic
filtration system, namely its ability to treat waste water
continuously on an "Instantaneous" basis.  The program has
also Indicated that  sedimentation treatment of storm water
is not only Inefficient in terms of the space required for
holding tanks; the actual sedimentation time must be varied
to accommodate an influent that changes Its consistency
almost continuously.

During April of 1968 when the Santa Barbara oil Incident
occurred, the Contractor, at  Its own expense, developed a
novel device for Instantaneously removing oil from sea
water using a highly simplified, gravity-assisted vortex
separation technique and has  been successful In obtaining
FWQA support to build a pilot system rated at 1 MGD.  This
new system is unique in being able to remove floating or
floatable material with a minimum of pumping power, space
and cost.

The Contractor's so-called "VORSEP" oil-water separator
has been modified recently by the addition of compressed
air to act as an Instantaneous air flotation unit which
appears to have great promise 1n treating storm water and
a host of other waste waters, Including raw sewage.

In early December of 1969, after concluding the contract
being reported upon, tests of a Contractor-owned prototype
VORSEP with air flotation added were made under relatively
crude operating conditions at Hyperion.  These tests demon-
strated a 21% reduction in BOD and a 50% reduction in the
volume of settleable solids passed through this device.

During February, 1970, the performance of this prototype
VORSEP was greatly enhanced by the use of an improved pump
that permitted the raw sewage influent to be more properly
aerated.  The test results obtained during the third and
most efficient set of runs in which laboratory data were
obtained are as follows:

10 gpm   214

 7 gpm   115
(Reduction = 45%)
CONCENTRATE   3gpm    227

 3.0 ml/1

 0.6 ml/1
 (Reduction = 80%)

11.0 ml/1
The above noted data were obtained during a test run using
a nominally rated system of 50 gpm.  The lesser flow of
10 gpm was the maximum obtainable with limited pumping
and power facilities available at the time.  It is known,
however, from previous tests of the same VORSEP in the
oil-water separation mode of operation, that at 50 gpm
flow, the concentrate flow remains constant at the 3 gpm
figure mentioned, or less.  Thus, by obtaining the full
50 gpm flow of raw sewage influent, the ratio of influent
to concentrate will approximate 17, instead of 3.3.

Recently concluded analyses and tests now indicate that
a proposed VORSEP incorporating both means for instantan-
eously removing gross matter, such as grit and sand, from
one exit port and floating materials, including grease, oil
and particles of aerated sewage, through a second,
centrally-located "vortex tube" exit port is entirely
feasible in capacities up to 1 MGD or more.  The diameter
of a 1 MGD unit would be under 10 feet and the height under
15 feet.  Aside from the use of a single low horsepower
recirculation pump and an insignificant amount of compres-
sed air, the unit would have no other moving parts except
for a float control system.  The cost of power is estimated

at 24/1.000 gallons, while the total  20-year  amortized  cost
of a 1  MGD unit plus operating power  is estimated  to  be 3i
per 1,000 gallons, or less.   In its optimized form,  the
VORSEP  with flotation appears capable of removing  up  to 75%
of the  suspended solids in combined sewage of virtually any
concentration, a performance figure surpassing that  ordin-
arily obtained in today's time-dependent primary sedimenta-
tion processes.

Figs. 12 and 13 illustrate the general  design features  and
appearance of typical VORSEP systems.

Should  such a system in practice achieve its  75% SS  removal
goal, 1t could precede the ultrasonic filtration system in
applications such as that encountered in Atlanta.   Even
greater removal possibly may be achieved with the VORSEP
1n conjunction with flocculants deliberately  Introduced 1n
the Influent line.

It 1s clear that a pressing  need exists for some type of
Instantaneous pretreatment system and that a  device,  such
as the  VORSEP may provide measureable treatment at low cost
and revolutionize the entire field of waste water treatment.

Without such a device, filtration systems are best used
where some conventional type of pretreatment  system  1s
available to reduce the levels of contaminents to below
100 mg/1 for both BOD and suspended solids or where  the
Influent itself 1s relatively lightly loaded  and consistent
1n composition.

E.   Economic Evaluation of  Combined  Sewage Treatment Systems.


     (a)  ESTIMATED COST                        $  750,000

          TOTAL INVESTMENT IN HARDWARE           $1,500,000

          TOTAL INVESTMENT*  PER DAY             $      205
          SERVICE CHARGE, PER DAY               $       27

          ENERGY CHARGE, PER DAY                $      628

          (70% of ENERGY CHARGE)                $	440

                                                   " Av" &-.-.V. O*-J
                      CO»«4\C.K\. S S PB R./V'T O «.

                                  Fig. 13 - 50 GPM VORSEP  IN OPERATION

          TOTAL POWER COST, PER DAY (AVERAGED)  $      561
          POWER COST, PER 1,000 GALLONS         $        0.02

     (c)  ESTIMATED COST PER POUND OF BOD       $        0.32
          REMOVED ASSUMING 200 ppm BOD IN
          INFLUENT & 50% REMOVAL (2,500 LBS
          BOD FOR $766)

          REMOVED ASSUMING 200 ppm SS IN
          INFLUENT & 50% REMOVAL                $        0.32

          GALLONS                               $        0.03



     (a)  ESTIMATED COST LESS VORSEP            $2,600,000


          TOTAL INVESTMENT IN HARDWARE          $5,200,000

          TOTAL INVESTMENT, PER DAY             $      712

          & 3,500 KW LOAD INTERMITTENTLY)
          SERVICE CHARGE, PER DAY               $      112

          ENERGY CHARGE, PER DAY                $    1,160

          (70% of ENERGY CHARGE)                J	812

          TOTAL POWER COST, PER DAY (AVERAGED)  $    1,050

          POWER COST, PER 1,000 GALLONS         $        0.04

          REMOVED ASSUMING 100 ppm BOD IN
          INFLUENT & 50% REMOVAL (1,250 LBS
          BOD FOR $1.760)                       $        1.40

          REMOVED ASSUMING 100 ppm SS IN
          INFLUENT & 502 REMOVAL                $      1.40

          FOR BACKWASHING                       $      0.07

          WASH e 10*/1,000 GALLONS              $      0.08


The two previous cost analyses indicate that a system for
treating  combined sewage in two steps, first through the
VORSEP for pretreatment and thereafter through the ultra-
sonic filtration system, can be accomplished for a total
of $0.12  per 1,000 gallons in a 25 MGD system, including
costs for personnel and maintenance.   Power rates quoted
are based on those of the Department  of Water and Power,
of the City of Los Angeles, which are  among the lowest in
the United States.  In other localities, the additional
cost of power up to 100% above the Los Angeles rates should
be taken  into consideration.  These additional charges  for
power could roughly double the cost of removing one pound
of BOD and suspended solids in each system.  The cost for
treating  1,000 gallons would rise more in the case of the
VORSEP, wherein the cost of power per day amounts to 75%
of the total daily cost.  For the ultrasonic filtration
system, power represents some 60% of  the daily cost.

Some mention should be made of demand charges, since a
combined  sewage treatment system might well be used only
intermittently.  During shutdown periods, the cost of power
is 70% of that used in operation, whether the system is
used or not.  Should power be available without this demand
charge, the operating costs drop significantly for both
systems by as much as 50%.

In these  estimates, both systems have been analyzed based
upon a 50% removal of BOD and SS each, with the influent to
the VORSEP being estimated as having  200 ppm of both BOD
and suspended solids.  Since the VORSEP can accommodate  far
greater loads, the cost of removing these contaminents  can

urop airectiy with the additional level of contamination as
no extra power or equipment is required.*  On the other
hand, the ultrasonic filter cannot at this moment perform
above the Influent loadings for BOO and SS of 100 ppm and,
thus, the removal costs are as stated.

Consideration should also be given to the space required by
this 25 MGD treatment plant.  Twenty-five 1 MGD VORSEPS have
been selected for this application, each approximately
12 feet 1n diameter maximum.  Thus, the ground area consumed
by 25 units would be roughly approximate 10,000 square feet.
The ultrasonic filtration system would occupy about 20,000
square feet in addition to make the total 30,000 square feet.
A conventional time-dependent primary-secondary treatment
plant would easily require ten times this space at a time
when real estate is more expensive than ever.  The Instantan-
eous treatment capability greatly reduces space and makes
possible the installation of fully-enclosed equipment that
cannot be of an objectionable nature to nearby homes and

Lastly, the cost analyses have omitted consideration of
sludge treatment, since the requirements for sludge disposal
vary widely.  For treating combined sewage during infrequent
overflows, sludge can be stored in holding tanks for delivery
to nearby waste treatment plants after surge conditions have
passed.  In any event, the requirements for sludge disposal
using the VORSEP and the filter together could involve the
disposal of 7,500 pounds of BOO and suspended solids per day
which may be mixed with approximately 4 MGD of concentrate
liquor.  Of this 4 MGD, 2.5 MGD are exhausted by the filtra-
tion system.  It is advantageous to include on the plant
site a conventional sludge thickener or possibly several
additional VORSEPS working in this function or a 4 million
gallon surge holding basin for each day's continuous flow
of sludge.
*Footnote No. 1 - With Influent BOD and SS levels of 1,000 ppm
 each, the cost of removing one pound of contaminent will drop
 from $0.32 to $0.06.

                    SECTION 5

During the course of the program, there were no applications
entered for patents covering any portion of the system.  The
basic patent covering the concept of ultrasonic filtration
was applied for previous to the start of this contract effort
and has recently been Issued by the United States Patent
Office.  This patent 1s owned by Acoustlca Associates, Inc.,
and 1s sub-licensed to the Contractor.  This patent 1s
Identified as follows:
Serial No.
U.S. Patent    F. D. DeLuca, Jr.  Acoustic Filtration
No. 3,478,883                     Apparatus

The VORSEP separation system described 1n this report has
been developed by the Contractor under corporate funding
and Is presently the subject of a patent application to be
filed with the United States Patent Office.

                    SECTION 6


The Contractor is Indebted to the Cities of Los Angeles,
California; Panama City, Florida; and Atlanta, Georgia,
for making available at various stages of the program test
facilities 1n their respective cities.  Personnel  at the
Hyperion Treatment Works, of the City of Los Angeles, under
the direction of Mr. Alfred Leipzig, Chief Engineer, and
Mr. William Garber, Assistant Chief Engineer, and  the
Millvllle Treatment Plant of the City of Panama City, were
most cooperative in making available at no cost various
items of equipments and working space which facilitated
testing of the Phase I and II systems respectively.   The
Hyperion facilities were made available under authority of
the City Council by Mr. Norman B. Hume, Director of  the
Department of Public Works, of the City of Los Angeles.

Full  cooperation of the City of Atlanta was obtained
through the efforts of Messrs. Richard W. Respess, Deputy
Director, and J. W. Cameron, Engineer of Sewers, of  the
Public Works Department.

Acknowledgement of the services furnished by the Georgia
Tech Research Institute under the direction of Dr. F.
Bellinger are 1n order.  The Contractor's sub-contracted
effort to Georgia Tech was under the direction of
Dr. Robert S. Ingols, who expended considerable time and
effort at the McDaniel Street Site, in Atlanta, analyzing
the waste water present and evaluating the performance
of the Contractor's microf 11 tratlon system.  Considerable
of the data and assumptions contained 1n this Report have
been extracted from test data and reports furnished  by
Dr. Ingols and his staff.  Mr. Edwin Lomasney, of  the
FWQA Regional Office in Atlanta, was also most helpful
in guiding the Contractor's efforts in Georgia.

                    SECTION 7


The following abbreviations are used in this report:






Biochemical Oxygen Demand

Chemical Oxygen Demand

KiloHertz, frequency In
thousands of cycles


Suspended Solids
The following uncommon terms are used in this report:

TERM               DESCRIPTION
  Sprltz Cycle
The sequence 1n which forward flow
through a filter element 1s interrupted
momentarily during which interval  the
filter element 1s cleaned ultrasonically
while immersed 1n filter Influent and
during which interval a small amount of
the influent surrounding the outer
surface of the element is discharged.

A trade name applied to a vortex
separation system developed by the

                    SECTION 8
PART I - Phase I Hydraulic Test Data:  Hyperion
During the period from November 11, 1968, to November 25,
1968, seven days of testing were conducted to establish
basic operating characteristics.  All tests of the Phase I
system during this time had similar operating parameters.
They were as follows:
     Equipment          Phase I Dual Tank System
     Filter Element     100 micron Stainless Steel
                        10" long, 2-3/4" OD, 3/16"
     Influent Pumping   One 3/4 hp Jabsco Pump
     Influent           Raw Sewage thru 20 mesh screen
                        settled for 1/2 hour and
                        diluted 10:1 with water
     Sprltz Timing      1 sec/10 sec
Average results of the above tests were as follows:
     Average Starting Influent Pressure          10-15 ps1
     Average Time Until Head Loss • 30 ps1        7- 9 m1n
     Average Starting Flow                       5.5 gpm
     Average End Flow                            1.0 gpm
     (Influent pressure consists of piping losses plus
      filter element back pressure.)
     System always clogged up after short running time.
     Phase I system modified to Include a pressurized back-
     wash during the normal sprltz time.  During the short
     sprltz Interval of approximately 1 second on to 9
     seconds off, the filter element received a pressurized
     backwash plus an ultrasonic cleaning.

     The  other  operating  parameters  were  as  follows:

     Filter  Element      50  micron  Stainless  Steel
                        10"  long.  2-3/4"  OD,  3/16"  thick

     Influent  Pumping    One  3/4  hp Jabsco  Pump

     Influent            Raw  sewage thru  20 mesh  screen
                        settled  for  1/2  hour

     Sprltz  Timing       1  sec/10 sec,  50  ps1
     and  Pressure

     The  results of  the  above  test were  as follows:

     Starting  Influent  Pressure                   10-15  ps1

     Time until  Head  Loss  •  30 ps1         Never  obtained
                                          as  pressure  only
                                          built  up  to  20

     Starting  Flow                                 5.5 gpm

     Ending  Flow                                   5.2 gpm

     Addition  of power  sprltz  deemed major improvement.


     Operatlng  parameters  were the same  as on  12/3/68;
     similar results  were  obtained.


     Operatlng  parameters  same as  on 12/3/68,  system  had
     dirty filter  elements  from  previous  run.  After  back-
     washing system  with  strong  hypochloMte  solution,  flow
     results of 4  gpm  and  an Inlet pressure  of 21  ps1 were
     obtained.   This  test  showed that  the  system  could  be
     recleaned  after  being  severely  contaminated.


     Operatlng  parameters  same as  on 12/3/68.  20  mesh
     screen  on  influent  hose ruptured  thus allowing
     unscreened sewage  as  influent.  System  performance:
     Flow 4.2  gpm,  influent  pressure 17  psi.   Use  of
     unscreened sewage  as  Influent did not detract  from

     Operating parameters same as 12/3/68, except spritz
     timing equal  to 1.4 sec/10 sec.   System performance:
     Flow 4.5 gpm,  influent pressure  16 psi.
     Operating parameters same as 12/3/68,  except timing
     variable.  System performance:   Filter element clogged,
     therefore system was backwashed with strong hypochforite
     solution and solution was left  to sit  in tank.
     Operating parameters same as 12/3/68,  except spritz  time
     equal  to 2 sec/10 sec.   System performance:   Filter
     elements still  clogged  therefore filters  were pulled
     and cleaned in  acid.
     Operating parameters same as 12/3/68,  except influent
     was fresh water.  This test run was made to  check  for
     cleaned filters.  System performance indicated  that
     filters had been cleaned.  Flow 6.50 gpm,  influent
     pressure at 10.00 psi Tank #1,  5.00 psi  Tank #2.


     Operating parameters same as 12/13/68.   System  perform-
     ance:   Flow 5.4 gpm, influent pressure  17.00 psi.  Use
     of unsettled sewage at end of test run  indicated that
     influent should be settled before filtration Is attempted,


     Operatlng parameters:

     Filter element -         40 micron Teflon, 10:  long,
                              2-3/4" 00, 1/16"  thick

     Influent Pumping         Two 3/4 hp Jabsco Pumps

     Influent                 Raw sewage thru 20  mesh screen
                              settled for 1/2 hour

     Spritz Timing and        Variable, 35  psi

    System Performance:   Flow 7,5  gpm,  Influent pressure
                         28 psi

    Filter clogged  up rapidly;  increase 1n Influent pressure
    due mainly to using  two Jabsco pumps  on the Influent.
    Operatlng parameters  same as  12/23/68,  except for use of
    30 micron polyethylene filter,  10"  long,  3"  ID,  1/16"
    thick,  and spritz  timing  equal  to  1.6  sec/10 sec.  System
    performance:   Flow 9.4 gpm,  Influent  pressure 18 psi.
    Polyethylene  filter superior  in flow  characteristics.


    Operating parameters:   Tank  #1, 30  micron polyethylene
    filter,  Tank  #2,  40 micron  Teflon  filter.   System
    performance:   Tank #1  flow  5  gpm,  influent pressure
    approximately 33  psi,  Tank  #2 flow  6.5  gpm,  influent
    pressure 27 psi.   Polyethylene  filter  element appeared
    to be deteriorating in Its  filtering  ability.

1/2/69- MORNING

    Operating parameters:

    Filter  elements        Tank  #1,  30 micron  Kynar 10"
                          long,  2-3/4"  OD,  1/16"  thick.

                          Tank  #2,  50 micron  Stainless
                          Steel  10" long,  2-3/4"  OD,
                          3/16"  thick

    Influent Pumping       Two 3/4 hp Jabsco Pumps

    Influent              Raw sewage thru  20  mesh screen
                          then  settled  for  2-1/2  hours.

    Spritz  Timing
    and Pressure           1.6 sec/10 sec,  80  psi

    System  performance:   Tank #1  end flow  5.4 gpm, end
    influent pressure  27  psi; Tank  #2 end  flow 6.50  gpm,
    end influent  pressure  27  psi.

    Kynar filter  deteriorated rapidly  1n  Its  flow

1/2/69 - AFTERNOON

    Operating parameters same as 1/2/69, Morning,  except
    Tank #1  had 30 micron polyethylene filter 10"  long
    3" ID, 1/16" thick.  System performance:   Tank #1  flow
    7.75 gpm. Influent pressure 25 ps1.  Polyethylene  filter
    again showed superior flow capabilities.

1/3/69 - MORNING

    Operating parameters same 1/2/69, Afternoon.   Only tank
    with polyethylene filter was tested.  Influent was taken
    from non-operating detrltor.  System performance:   Flow
    7 gpm Influent pressure 29 p$1.   Polyethylene  filter
    continuously maintaining Us high flow capabilities.

1/3/69 - AFTERNOON

    Operating parameters same as 1/2/69, Afternoon.  System
    performance:  Tank #1 flow 7 gpm, Influent pressure
    28.25 psi, Tank #2 flow 7.6 gpm  Influent  pressure  23 psl,

1/6/69 -

    Operating parameters same as 1/2/69, Afternoon.  System
    performance:  Tank II flow 6 gpm, Influent pressure
    34 psi,  Tank #2 flow 6.8 gpm, Influent pressure  26 psi.
    Filter elements performed basically the same.   Differ-
    ence in  micron rating (30 microns for polyethylene
    against  50 microns for Stainless Steel) still  Indicated
    polyethylene to be better filter.


    Operating parameters:

    Filter Element - 30 micron Polyethylene 10"  long,
                     3" ID, 1/16" thick

    Influent Pumps   Two 3/4 hp Jabsco Pumps

    Influent         Plant secondary

    Spritz Timing    1.0 sec/10 sec, 100 psi
    and Pressure

    System performance:  With sprltz turned off,  system
    tested for maximum flow capabilities.  After  30
    minutes  pressure remained constant at 17  psi  and flow
    rate was approximately 12.25 gpm.  Thus an obtainable
    maximum  flow rate was established of around  12 gpm.

2/12/69 -
    Operating parameters same as 2/5/69, except Influent
    was raw sewage thru 20 mesh screen and not settled.
    System performance:  No relevant Information obtained
    from this test as polyethylene filter element ruptured
    and therefore was not actually filtering.
2/13/69 -

    Operating parameters:

    Filter Element
    Influent Pumping

    Spritz Timing
    and Pressure
                 35 micron polyethylene,  10"  long
                 3" OD, 5/16" thick

                 Two  3/4 hp Jabsco Pumps

                 Raw  sewage thru 20 mesh  screen
                 settled for 1/2 hour

                 1 .5  sec/10 sec, 100 psi
    System performance:   Test was ended due to backflush

2/25/69 -
>m  performance:   Test was enae
 failure, no relevant results.
    Operating parameters same as 2/13/69,  except Tank #2
    contains 30 micron polyethylene,  1/16"  thick filter
    element mounted on plastic cage support mechanism.
    System performance:   Tank #1, flow 4.5  gpm influent
    pressure 31 psi; Tank 12, flow 2.75 gpm, influent
    pressure 32 psi.  Results of this test  showed 5/16"
    polyethylene filter  superior to 1/16"  polyethylene.
3/4/69 -
    Operating parameters same as 2/13/69, except Influent
    was plant secondary.  System performance:   Flow
    approximately 8 gpm, influent pressure 21  psi.   Good
    performance from 5/16" thick filter element.
3/5/69 -
    Operating parameters same as 3/4/69.  System performance
    Flow 8.5 gpra, Influent pressure 20 psi.  Increase in
    flow rate from 3/4/69 due mainly to new rotor being put
    into one Jabsco pump.

3/6/69 -

    Operating parameters same as 2/13/69, except influent
    settled for 2 hours.  System performance:  Flow 5,2 gpm
    influent pressure 28 psi.  This test represents perform-
    ance of the finalized Phase I system.  Average flow of
    5.2 gpm indicates filter efficiency of approximately
    10 gpm per square foot of filter area.

PART II - Phase I Hydraulic Test Data:  Hyperion

Laboratory tests for BOD and tests for Suspended Solids
were made by Truesdail Laboratories, Inc., of Pasadena,
California, and by the Millville Treatment Plant in Panama
City, Florida, for BOD only.

Samples were kept refrigerated pending delivery to the
laboratory late in the afternoon following the tests.

At Hyperion, as stated previously, BOD was reduced an
average of 39% and the suspended solids were reduced an
average of 15% on settled raw sewage screened through
20 mesh screens.

On secondary sewage run directly through the filter system
without screening, average BOD reduction was approximately
50% and suspended solids were reduced by approximately 35%
using the 35 micron polyethylene filter element.

In Panama City the system was tested using the effluent
from the Millville Primary Treatment Tank, operating
continuously at a high flow rate of approximately 15 gpm
per canister, with results showing a 51% reduction in  BOD.

                        BIBLIOGRAPHIC.  Thi Amirlcin Process Equipment Corporitlen
                            Ultrasonic Filtration at Combined Sewer Ovtrflowi EPA/WQO

                        ABSTRACT! A 290,000 gpd compict ultriionlcilly cleaned
                            mleroflltrttlan IviUm wll tested with simulated combined
                            Mwer overflow! ind prlmiry treatment punt effluent.
                            Twenty 35 micron porous polyithyline elementl of O.I IQ
                            ft irll uch comprllld thi lyinm.

                            Slmuratld tattled comblnid »wtr Ovtrflowi it Atunti.
                            Oiorgii tut ilt« quickly cloggid fntiri wltn nign
                            concentritloni of ruit In both Influint ind fmh bickwiin
                            water.  Thui, fMilblllty of trutlng combined iiwer over-
                            flowl wll not dimonstnted. Study Indicates UH of mori
                            coitly poroui ttilniess mil eiiminti would obvlite thli

                            Sufficient Hit dttt  iri ivilltbli to predict performance
                            In mori lultable wittr pollution control ippllcitloni.
                            Wltn Influent BOO and luipindid tolldi ituli of 100 mg/1,
                            or His, ultrnonlc flltntlon ullng  39 micron pltnlc
                            •Hmtntl ctn rMuci BOO ind SS In riw Killed Hwigl,
                            or iffluintl from prlmiry ind Mcondlry plinti by 90%.
                            A 29 MOO lyitim riqulrfll till tnin 20,000 iquiri fiat, ind
                            colt for treating 1,000 gilloni li »0.0«, including pre-

                            Contrictor'i novii tiotitlon vortitc upintor nducad nw
                            Mwigl BOO 49% ind Influint solids 10%. This  "VORSEP"
                            could contlnuouily  trill comblnid sewer .overflows, ind/or
                            prltrut Influint to  tni ultmonlc mlcroflltritlon lyitim.
                            A 29 MOO Vorup trull 1,000 gilloni for 10.03. ind requires
                            10,000 iq ft of IPICI.

                            Thli riport li lubmlttM In fulfillment of Contrict 14-12-199
                            bltwiin thi Envlronmintll Protictlon Agincy Witlr Quillty
                            Offlci ind Amirlcin Process Equipment Corporltlon.




                        BIBLIOORAPHICi  Tni Amirlcin Proem Equlpmint Corporltlon
                           Ultrltonlc Flltritlon of Comblnid Slwir Overflow! EPA/WQO

                        ABSTRACTi A 250,000 gpd compict ultriionlcilly clunM
                           mlcroflltritlon lyitim wll tlltld wltn HmulltM comblnid
                           Hwir ovirflowl ind prlmiry trlitmint punt effluent.
                           Twinty 39 micron poroui polyitnylini elementl of 0.1 iq
                           ft eru ucn comprlild tni tyitim.

                           Slmulilid Mtllld comblnid Mwir overflowi it Atlinti,
                           Qiorgli tilt llti quickly cloggM flltlri wltn nign
                           concintritioni of  ruit In botn Influint ind frlin bickwiin
                           water. Thui. tullblllty of trilling comblnid iiwlr ovir-
                           flowl wil not dimonilritld. Study Indlcitll uu of morl
                           coitly poroui itiinini itiel elementl would obvliti tnii

                           Sufflclint tut dltl in ivillibli to prMlcl pirfornunci
                           In mori lultlbll wltir pollution control ippllcitlont.
                           Wltn Influint BOO ind luipindid lolldi livin of 100 mi/1,
                           or Mil, ultriionlc flltritlon utlng 39 micron pintle
                           elementl cin rMucl BOO ind SS In riw Hilled  Hwigl,
                           or iffluintl from primary ind ncondiry plinli by 90%.
                           A 29  MOO lyitim requires l«l tnin 20,000 tquira fMt, ind
                           colt for trutlng 1,000 gallon! li (0.01, including prl-
                           trut mint.

                           Contriclor'l novil flotltlon vortix upiritor riducid nw
                           MWIXBOO «9% and Influint  lolldi (0%.  Thli "VORSEP"
                           could contlnuouily trut comblnM Hwar ovirflow/, and/or.
                           prltrut Influant to tn« ultmonlc mlcroflllratlon lyltam.
                           A 29  MOO Vomp trull 1,000 galloni for S0.03, and riqulrn
                           10,000 w ft of ipaca.

                           Thli raport II lubmlttad In fulflllmint of Contrict 14-12-K9
                           bitwun tn« Environmintll Prolactlon Agincy Watar Quillty
                           Offlci and American Proem Equlpmint Corporltlon.




                         BIBLIOGRAPHIC!  Tne American Proem Egulpmim Corporltlon
                            Ultretonlc Filtration of Combined Sewer Ovirflowl EPA/WQO

                         ABSTRACTi A 250.000 gpd compict ultraionlcally cluned
                            mlcroflltratlon lyitam wn tntid with ilmulatid combined
                            lawer ovirflowl ind prlmiry trutment plint effluent.
                            Twenty 39 micron poroui polyethylene element! of O.t H
                            ft iru ucn comprlud  the tyitem.

                            Slmulltld llttled comblnid uwer ovirflowl it Atlinti,
                            Oeorgla ten lite quickly clogged fllleri with high
                            concentratloni of ruit In botn Influint ind frein bickwiin
                            water. Thui, feailblllty of treating combined fewer over-
                            flowl wai not demonitrated. Study Indicitu uu of more
                            coitly poroui italnlaii iteel elementi would obviate thii

                            Sufficient tell dill iri ivallible to predict performance
                            In more tunable water pollution control ippllcitloni.
                            With Influint BOD ind impended lolldi levin of 100 mg/1,
                            or laii, ultrasonic flltntlon uilng 3ft micron pintle
                            •lements can reduce BOO and SS In raw settled  tewage,
                            or effluents from primary and secondary plints  by 90%.
                            A 29 MOO tyitem requlrei Im thin 20,000 iquire fut, and
                            con for treating 1,000 gallon! Is (0.01, including ore-

                            Contrictor's novil flotltlon vortex septritor riduced raw
                            wwage BOD 49% ind Influint wild! 10%.  TKIt "VORSEP"
                            could contlnuouily trut comblnid uwer overflows, and/or
                            pretrut Influint to the ultrasonic mlcroflltratlon system.
                            A 25 MOO Vortep treatl 1,000 gilloni for »O.OJ, ind riquln*
                            10,000 iq ft of space.

                            This report Is submitted In fulfillment of Contrict 14-12-199
                            between the Envlronmentu Protection Agency Watar Quality
                            Office and American Process Equipment Corporltlon.





    xlcce.s-siori Number
             Subject Field & Group
                                                SELECTED WATER  RESOURCES ABSTRACTS
                                                      INPUT TRANSACTION FORM
1Q Authors)


Project Designation
     Descriptors (Starred First)
     Identifiers (Starred First)
                          Ultrasonic Filtration
                          Combined Sewer Overflows
                          Vortex Separator
A 250,000 gpd ultrasonically cleaned microfiltration system was
unsuccessfull in treating combined  sewer overflows at an Atlanta, Georgia
test  site.   High concentrations of  rust  clogged the porous polyethylene
filter  elements.  It is anticipated that with stainless steel filter
elements,  influent BOD and Suspended Solids concentrations of 100 mg/1
or less  could be reduced by 50$.  A novel flotation vortex separator  is
described,  which could serve as a pretreatment device for the filter.

This  report is submitted in fulfillment  of Contract lU-12-195 between
the Environmental Protection Agency, Water Quality Office and the
American Process Equipment Corporation.
        Darwin R. Wright
                       Environmental Protection Agency, Water Quality Office
 WR:102 (REV. JULY 1969)
                                       U.S. DEPARTMENT OF THE INTERIOR
                                       WASHINGTON, D. C. 20240
                                                                                * SPO: 1969-359-339

Continued from inside front cover ....
11022 --- 08/67

11023 --- 09/67

11020 --- 12/67

11023 --- 05/68
11020 DWF 12/69

11000 --- 01/70

11020 FKI 01/70

11024 DOK 02/70

11023 FDD 02/70

11024 DMS 05/70

11023 EVO 06/70

11024 --- 06/70
Phase I - Feasibility of a Periodic
Flushing System for Combined Sewer
Demonstrate Feasibility of the Use of
Ultrasonic Filtration in Treating the
Overflows from Combined and/or Storm
Problems of Combined Sewer Facilities
and Overflows, 1967, (WP-20-11)
Feasibility of a Stabilization-
Retention Basin in Lake Erie at
Cleveland, Ohio
The Beneficial Use of Storm Water
Water Pollution Aspects of Urban Runoff,
Improved Sealants for Infiltration
Control, (WP-20-18)
Selected Urban Storm Water Runoff
Abstracts, (WP-20-21)
Sewer Infiltration Reduction by Zone
Pumping, (DAST-9)
Strainer/Filter Treatment of Combined
Sewer Overflows, (WP-20-16)
Polymers for Sewer Flow Control,
Rapid-Flow Filter for Sewer Overflows
Design of a Combined Sewer Fluidic
Regulator, (DAST-13)
Combined Sewer Separation Using
Pressure Sewers, (ORD-4)
Crazed Resin Filtration of Combined
Sewer Overflows, (DAST-4)
Storm Pollution and Abatement from
Combined Sewer Overflows - Bucyrus,
Ohio, (DAST-32)
Control of Pollution by Underwater
Storm and Combined Sewer Demonstration
Projects - January 1970
Dissolved Air Flotation Treatment of
Combined Sewer Overflows,  (WP-20-17)
Proposed Combined Sewer Control by
Electrode Potential
Rotary Vibratory Fine Screening of
Combined Sewer Overflows,  (DAST-5)
Engineering Investigation of Sewer
Overflow Problem - Roanoke, Virginia
Microstraining and Disinfection of
Combined Sewer Overflows
Combined Sewer Overflow Abatement