United States
                Environmental Protection
                Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
                Research and Development
EPA/600/SR-94/129   September 1994
EPA        Project  Summary
                 Stormwater Pollution
                Abatement  Technologies
                Richard Field, Michael P. Brown, and William V. Vilkelis
                  The report summarized here presents
                information regarding best management
                practices (BMPs) and pollution abate-
                ment  technologies that  can provide
                treatment of urban stormwater runoff.
                The text includes a general  approach
                that considers small storm hydrology
                and watershed practices that cover pub-
                lic education, regulations, and  source
                control of pollutants. Also covered are
                source treatments of pollutants, which
                include vegetative BMPs  and  infiltra-
                tion practices.  Uses and modifications
                of installed drainage systems, types of
                end-of-pipe treatments including bio-
                logical, chemical, and  physical types
                and storage and reuse of stormwater
                are also covered.
                  This Project Summary was developed
                by EPA's Risk Reduction  Engineering
                Laboratory, Cincinnati, OH, to announce
                key findings  of the research  project
                that is fully documented in a separate
                report of the same  title  (see  Project
                Report ordering information at back).

                Introduction
                  The full report covers the control and
                treatment of stormwater in relation to the
                removal or reduction  of the stormwater
                pollutant loads.  Many  of the pollution
                abatement technologies discussed will help
                attenuate stormwater flows.  As they are
                generally designed for small storm events,
                however, they will not provide sufficient
                capacity for the large events.  Although
                prevention  of stormwater flooding is not
                discussed, a  drainage  system  design
                should consider both pollutant  and flood-
                ing aspects of stormwater.
  Strategically, the  best way to control
and treat urban stormwater  runoff is
through  a  combination  of  regulations,
BMPs, and treatment processes. The op-
timal combination will be site specific and
depend on site characteristics, specific
pollutants involved, and cost  consider-
ations.
  Regulations  and  BMPs  are  effective
tools in controlling urban stormwater run-
off because they tend to be preventive in
nature.  Mandating effluent limits and cre-
ating zoning laws are regulatory examples.
BMPs may include upgrading current sys-
tems,  developing proper management
techniques, using the existing  drainage
systems for in-line or in-sewer storage, or
creating off-line storage facilities.
  Designing devices that intercept or infil-
trate stormwater runoff back  into the
groundwater system before it is introduced
into the  stormwater or combined sewer
conveyance system can greatly save costs
in the  design and construction  of treat-
ment facilities.  Examples of such devices
are swales, filter strips, porous pavement,
and stormwater wetlands.

General Approach and Strategy

Small Storm Hydrology
  The  selection  of suitable abatement
technologies requires an understanding of
the size and distribution of storm events.
Generally, the  smaller storm events are
the  critical  storms to consider,  because
for many parts of the country, 85%  of all
the rains are less than 0.6 in. (15 mm) in
depth and can generate about 70% of the
total annual storm runoff. The character-

-------
istics  of small and  large storms  can be
very different in terms of the runoff gener-
ated,  pollutant load,  and receiving water
impacts.   Frequent  small storms  have a
more  persistent effect, whereas less fre-
quent large  storms  have a larger impact
but allow time for recovery between events.
For small  storms, inaccurately estimating
the initial abstractions and the soil infiltra-
tion rates can significantly change  the cal-
culated storm runoff pollutant load.  Initial
abstractions include the rainfall depth re-
quired to satisfy surface wetting,  surface
depression storage,  interception by hang-
ing vegetation, and evaporation. Together
with soil infiltration  rates,  the initial ab-
stractions need to be accurately estimated
to calculate the storm runoff volume.

Strategy
  Traditional wastewater treatment meth-
ods (i.e., secondary treatment processes)
tend to operate under conditions closer to
steady state and are  usually  unsuitable
for the fluctuating loads of stormwater run-
off. On the other hand, technologies used
to control and treat combined sewer over-
flows (CSOs) are  more  suitable  for
stormwater runoff. Successful stormwater
management to control urban  storm run-
off pollution requires an areawide approach
combining prevention,  reduction, and treat-
ment  practices/technologies. It is  unlikely
one method will provide the best solution
to control the widespread diffuse nature of
stormwater runoff and achieve the water
quality required.
  Establishing an urban storm runoff pol-
lution  prevention and control plan requires
a structured strategy that should include:
defining existing conditions; setting site-
specific  goals;  collecting and analyzing
data;  refining site-specific goals; assess-
ing and  ranking problems; screening and
selecting BMPs and treatment technolo-
gies;  and, implementing,  monitoring, and
reevaluating the plan.
  This strategy  will  provide  the  control
goal(s) to  be achieved — the goal(s) that
are then used as the basis for selection of
suitable technologies or approaches.  The
goal(s) should initially be broad and not
specific  as the process of  reviewing the
technologies or approaches available will
in itself generate information to focus and
refine the goal(s) to  meet  cost, level of
control, public opinion, feasibility, and other
restraints.
  A flexible  approach, which, through an
iterative process of review and adjustment
is focused to a specific action plan, is the
only real method by which the  complexity
of urban  stormwater can  be managed.
The specific action plan must also be sub-
jected to reassessment once feedback on
implementation is available.
  The  report is concerned with  an  over-
view of the abatement technologies  avail-
able and reviews  the  technologies  by
separating the drainage  system into  three
physical areas:

  • watershed area  (i.e., storm runoff
    generation/collection  area),

  • installed  and/or  modified/natural
    drainage system (i.e., conveyance
    pipes, channels, storage, etc.), and

  • end-of-pipe (i.e., point source).

  Technologies applicable to each of these
areas can be  divided into  structural and
nonstructural.  The nonstructural technolo-
gies cover approaches such as public edu-
cation, regulations,  and local  ordinances
and mainly apply to the upstream collec-
tion area. The structural approaches are
the main options for the drainage system
and end-of-pipe areas and tend to be the
more expensive items.
  The  optimal solution is likely to be an
integrated approach that employs several
practices and technologies. The manage-
ment  of the watershed  using BMPs to
prevent or control pollution at the source
is apt to offer the most cost effective solu-
tion and tends to  be the basis  of  many
stormwater  management  plans.  BMPs,
the preferred  option,  are,  however, not
always  feasible  or sufficient  to achieve
the control objectives by themselves. For
older and more heavily  urbanized areas,
BMPs  are likely to  have limited applica-
tion,  and  some form of treatment before
discharge may be required.
  Implementation of any stormwater man-
agement program needs to meet financial
and, probably,  schedule restraints. There-
fore, an early review and improved use of
existing facilities offers several advantages.
These  options, probably the quickest and
least costly to  be implemented, must also
meet the  objectives developed from the
earlier stormwater management  planning
process.  Examples might include the en-
forcement of existing  regulations to con-
trol soil  erosion  during  construction
activities and adaptation  of  existing
stormwater storage intended for flood con-
trol so  that it also provides quality control
for small storm events.  New installations
should consider design for both flood con-
trol and pollutant removals.
Watershed Area Technologies
and Practices
  As already stated, BMPs are  not suit-
able in every situation.  It is important to
understand which  BMPs are suitable  for
the site conditions and can also achieve
the required goals.  The realistic evalua-
tion for each practice includes: the techni-
cal feasibility, implementation costs, and
long-term maintenance requirements and
costs.  It is also  important to appreciate
that the reliability and performance of many
BMPs  have  not  been well established,
with  most BMPs still in the development
stage.  This is not to say that BMPs can-
not be effective  but rather that they  do  not
have a  large enough  bank of historical
data on which to base design to  be confi-
dent that the performance  criteria will  be
met under the local conditions. The most
promising and best understood BMPs  are
detention and extended detention  basins
and ponds.  Less reliable in terms of pre-
dicting  performance, but showing  prom-
ise, are  sand filter beds, wetlands, and
infiltration basins.
  The reported  poor performance of some
of the BMPs is  likely to be a function  of:
the design, installation,  maintenance, and/
or suitability of  the area.   Greater atten-
tion to these details is  apt to significantly
reduce the failure rate of BMPs.  Other
important design  considerations include:
safety for maintenance  access and  opera-
tions, hazards to the general public through
safety or  nuisance, acceptance by the pub-
lic,  and  assuming conservative perfor-
mances  in the  design  until the historical
data can justify a higher  reliable  perfor-
mance.
  The  previously mentioned  goals for a
stormwater management  plan  can  be
achieved in the watershed  area via three
basic avenues:
  Regulations,  Local  Ordinances, and
Public Education.   This should be the  pri-
mary objective  because it probably is  the
most cost effective.  Mainly nonstructural
practices will be involved, and application
to new developments should be particu-
larly effective.
  Source  Control of  Pollutants.   Both
nonstructural and structural practices can
be used to prevent pollutants coming into
contact with  the  stormwater and  hence
storm runoff.  Management and  structural
practices include:  flow  diversion  (keeping
uncontaminated stormwater from contact-
ing contaminated  surfaces  or water by a
variety of structural means);  exposure mini-
mization  (minimizing stormwater contact
with pollutants by structure and manage-

-------
ment); mitigation (plans to recover released
or spilled pollutants in  the advent of a
release); prevention (monitoring techniques
intended to  prevent  releases); control  of
sediment and erosion; and  infiltration.
  Source Treatment, Flow Attenuation,
and Storm Runoff Infiltration.  These are
mainly structural practices to provide up-
stream  pollutant removal  at the  source,
controlled stormwater release to the down-
stream  conveyance  system, and ground
infiltration or  reuse of the stormwater.
Upstream pollutant removal provides treat-
ment of stormwater runoff at the specific,
highly polluting locations where  it enters
the stormwater conveyance system.  Ar-
eas of this type include but are not limited
to vehicular parking  areas, vehicular ser-
vice stations, bus depots,  industrial load-
ing  areas, etc.

Source Treatment, Flow
Attenuation, and Storm  Runoff
Infiltration

Vegetative BMPs
  These practices  have been the subject
of many publications in the last  20 yr.
Existing  urbanized areas  are  unlikely  to
have the land space available for installa-
tion  of  many of these  practices  and,  in
these situations, their application will be
restricted.
  Swales   are   generally   grassed
stormwater conveyance channels that re-
move pollutants by  filtration through the
grass and infiltration through the soil.  A
slow velocity of flow, <1.5 ft/s (<46 cm/s),
a nearly flat longitudinal slope, <5%, and
a vertical stand  of dense vegetation higher
than the water surface, >6 in.  (15  cm)
total height,  are important for effective op-
eration.
  Filter strips are vegetated strips of land
that  act  as  "buffers" by accepting storm
runoff as overland  sheet flow from up-
stream developments before discharge to
the storm drainage system.  Filter strips
provide  potential  treatment mechanisms
similar to that of swales.
  Stormwater wetlands can  be  natural,
modified natural, or constructed wetlands
that  remove pollutants by sedimentation,
plant uptake,  microbial decomposition,
sorption, filtration,  and exchange capac-
ity.   Note that  natural wetlands  are  cov-
ered by  regulations  that limit discharges
to the wetland  and  limit modifications  to
enhance the wetland performance.

Detention Facilities
  One of the most common structural  con-
trols for urban  storm runoff and  pollution
loading  is the construction of local ponds
(including wetlands) to collect storm run-
off, hold  it  long  enough  to  improve its
quality, and  release it to receiving waters
in  a controlled manner.   The basic re-
moval  mechanism is  through settling of
the suspended solids (SS) with any asso-
ciated  pollutants, but controlled release
will also  attenuate the stormwater flows,
which can benefit receiving  streams that
suffer from  erosion  and  disturbance of
aquatic  habitat during peak flow condi-
tions.
   Extended detention dry ponds tempo-
rarily detain  a portion of stormwater runoff
for up  to 48 h (24 h is more common)
using  an outlet  control.   They provide:
moderate but variable removal of particu-
late pollutants; negligible soluble pollutant
removal;  and quick accumulation of de-
bris and  sediment.  Performance can be
enhanced by  using  a forebay  to  allow
sedimentation  and  easier removal   from
one area.
   Wet ponds have greater capacity  than
the permanent volume of the pond; this
permits  storage of the stormwater runoff
and controlled release of the mixed  influ-
ent and permanent pond water. They can
provide moderate to high removal of par-
ticulate  pollutants and  reliable  removal
rates with pool sizes ranging from 0.5 to
1.0 in.  (12.7 to 25.4 mm)  of storm runoff
per impervious acre.   Wet  ponds   offer
better removals and less maintenance than
do dry ponds. But they need to be well
designed to  ensure beneficial use and not
cause aesthetic, safety, or mosquito breed-
ing problems.   A forebay here  also im-
proves performance and maintenance.

Infiltration Practices
  These practices have a high  potential
to  control stormwater  runoff  by disposing
of  it at a local site. Infiltration in its  sim-
plest form involves maximizing the pervi-
ous area of available  ground  to  allow
infiltration of stormwater and  minimize the
storm runoff.  This can  be enhanced by
directing  storm  runoff from  impervious
paved and roof areas to pervious areas,
assuming sufficient infiltration capacity ex-
ists.  Regulations  that encourage the in-
corporation of a high proportion of pervious
areas, particularly for new developments,
can be effective;  however, soil and water
table conditions   have to  be  suitable,  a
conservative design has to be used, and
maintenance has to be undertaken to mini-
mize the  possibility of system failure.  The
possible  effects   the storm  runoff could
have on  the groundwater must also be
considered.   These could range from a
relatively minor local raising  of the water
table that results in  reduced  infiltration
rates to more serious pollution of the
groundwater, particularly if this is also used
as  a water source.   In  many  cases
stormwater runoff will have low levels  of
pollution; however,  the  long-term effects
of pollutant buildup  in the  soil and/or
groundwater from storm runoff infiltration
is not well known. Therefore, infiltration of
urban storm runoff,  especially from indus-
trial and commercial areas that have higher
levels of pollution,  should be treated with
caution.
  Infiltration of storm runoff can  offer sig-
nificant  advantages of  controlling storm
runoff at the source, reducing the risk  of
downstream flooding, recharging ground-
water,  and supplying  groundwater  to
streams  (i.e.,  low-flow  augmentation  or
maintaining stream flow during dry-weather
periods). These  advantages  need to be
judged  against any pollution  risks from
urban runoff.
  Infiltration trenches,  infiltration basins,
and porous pavement are all applications
of infiltration practices.   Performance  of
these  applications can  be  improved
through  regular  maintenance,  protective
practices against clogging  (e.g., protec-
tive screening from nearby construction),
grass filter strips to filter out particulates,
and  sub-surface piping  installed to direct
the stormwater away.

Installed Drainage  System
  Control practices  that can be applied  to
the drainage system are relatively limited,
especially for  existing systems, and in-
volve the removal of illicit or inappropriate
cross-connections,  catchbasin  and  inlet
cleaning, critical  source area  treatment
devices, infiltration,  and in-line and off-line
storage.
  New  separate (or combined) systems
can take advantage of increasing the pipe
size and gradient to provide  in-line stor-
age and  self cleaning, respectively. Exist-
ing  separate  (or  combined)  drainage
systems  can be modified for in-line stor-
age by adding flow control devices (weirs,
flow regulators, etc.).
  Established  urban areas with  separate
stormwater  drainage  systems  are  most
likely to have an existing stormwater pol-
lution problem that needs to be rectified.

Critical Source Area Treatment
Devices
  Research into the source of stormwater
pollutants has shown that certain  critical
source areas can contribute a significant
portion of the total urban storm runoff pol-
lutant load.  Treatment of the critical source
areas can, therefore, offer the  potential
for  a greater benefit than end-of-pipe  or

-------
drainage system control, to reduce down-
stream  pollutant  loads. Potential  critical
sources include: vehicle service, garage,
or parking areas; storage  and transfer
yards;  and industrial materials  handling
areas exposed to precipitation.

In-line Storage
  In-line storage uses the unused volume
in the drainage system network of pipes
and  channels to  store storm  runoff that
can  also  be  provided  by  storage tanks,
basins,  tunnels,  or  surface ponds con-
nected to the conveyance network.   In-
line  storage  will  probably  not offer any
treatment  in itself as the intent will be to
make the  system self-cleaning to reduce
maintenance  requirements.  If storage  is
combined with  an end-of-pipe treatment,
however,  the flow  attenuation will help
equalize the load to the treatment process
and, hence, optimize the treatment plant
size  and costs. Other cost  effective solu-
tions might be found if existing treatment
facilities can be used, such as  connection
to an existing wastewater system.
  The  degree to  which the existing con-
veyance system can be used for storage
is a  function of:  pipe  or  channel sizes;
pipe or channel  gradient  (relatively flat
lines provide the most storage  capacity
without  susceptibility to flooding  low ar-
eas); suitable locations for installation of
control devices; and the reliability of the
installed control.  It is essential that accu-
rate details of the existing  system be col-
lected  from  field surveys  and as-built
drawings.  This allows the assessment of
the storage capacity, number and loca-
tions of controls,  and risk of upstream
flooding.  In new drainage system design,
conveyance pipes and channels can be
up-sized and  hydraulic controls  can  be
designed  into the system for added sys-
tem storage and routing.
  Controls to restrict flow  can either be
fixed or adjustable.   Fixed systems will
probably  be  cheaper  and  require  less
maintenance.  Some examples  of fixed
regulators are: orifices, weirs (lateral and
longitudinal),  steinscrews,  hydrobrakes,
wirbeldrossels,  swirls,  and stilling-pond
weirs.
  Adjustable  systems  can offer the ad-
vantage of being connected to  a  real-time
control (RTC) system,  which can be ad-
justed to hold back or release stormwater,
to maximize storage capacity of the whole
drainage system.  The sophistication of
an RTC system is unlikely to offer a cost
effective solution for a separate storm
drainage system  unless there is a large
in-line  storage  capacity  and the stored
runoff is to be treated. Typical examples
of adjustable regulators  are:  inflatable
dams,  tilting plate regulators,  reverse-
tainter  gates, float-controlled gates,  and
motor-operated  or hydraulic gates.
  Some of the  above are relatively inex-
pensive,  quick to install,  and  effective
means of increasing storage. Also some
will concentrate the heavier solids in the
stored  storm runoff for a more  concen-
trated later release.

Off-line Storage
  This refers to storage that is achieved
by diverting flow from the drainage con-
veyance system when a certain flowrate
is exceeded.  The diverted water is stored
until sufficient capacity is available down-
stream. Storage can be provided by any
arrangement of basins, tanks, tunnels, etc.
If gravity filling and emptying are not pos-
sible, pumping  the water into or out of
storage is  involved.
  Off-line storage can be designed to be
relatively self-cleaning or to have facilities
to  resuspend the settleable solids.  It can
also be used to  provide treatment by sedi-
mentation with the sludge either collected
or diverted to a wastewater treatment plant
(WWTP).
  Flow balance method (FBM) provides a
means of storing discharged urban storm
runoff in the receiving water. This is done
by forming  a  tank with the use of flexible
plastic  curtains suspended from pontoons.
The curtains  are anchored to the receiv-
ing water bottom by concrete weights and
the base  of  the tank  is formed by the
receiving water bed.  The relatively low
cost of the materials and construction gives
this system cost advantages over conven-
tional concrete and steel tank systems.
  The  FBM requires  a suitable location
and  has limits on performance: a certain
amount of mixing exists with the receiving
water,  not  all the  stored volume will be
pumped back,  and settleable solids re-
quire regular pumpback  of  the  accumu-
lated sediment.  The quick construction
potential of the  FBM  could favor the use
of this  system as a temporary measure in
cases  of a severe problem that needs
attention.  Since the FBM  uses the exist-
ing receiving  water, permits will  probably
be required.

Maintenance
  Regular  maintenance should  be con-
ducted for  the drainage system and the
controls to work efficiently. This generally
consists of removing sediments from con-
trol devices, flushing drainage lines,  and
conducting inspections to identify any prob-
lems.  Maintenance minimizes buildup of
materials that can  be  flushed  out by  a
surge  from a  large  storm  event  and,
thereby,  minimizes the shock loading
caused by intermittent storm events.

End-of-Pipe Treatment

Use of Existing Treatment
Facilities
  Use  of existing facilities  is apt to  pro-
vide cost effective treatment as long as
an  economic  means  of connecting  the
stormwater drainage system to the facility
is possible.
  Spare  capacity at the WWTP is  one
option, particularly if storage can be  pro-
vided to  equalize the storm  runoff load.
Even if the biological  system  has very
little capacity, the primary treatment  sys-
tems can often function well at somewhat
higher  overflow  rates  that, if combined
with disinfection  of the discharged storm
runoff, will  offer significant treatment.
Stormwater  also tends to have a higher
percentage  of heavier solids than does
sanitary sewage, which will  benefit remov-
als  at higher overflow rates.
  An alternative  could be to construct ad-
ditional primary treatment at a WWTP to
run in  series with existing facilities during
dry-weather flow  (DWF) for improved treat-
ment of DWF and to run in  parallel during
wet-weather flow for some control  over
the total flow. Also, use of any storage
facilities, either at an end-of-pipe or at an
upstream location, can provide treatment
by sedimentation or storage to be released
when treatment capacity is  available.

Physical/Chemical Treatment
  Physical/chemical treatment processes
generally offer: good resistance to shock
loads,  ability to consistently produce a low
SS  effluent,  and  adaptability to automatic
operation. Those described  below are  only
suitable for removal of SS and associated
pollutants.  Other treatment methods  (de-
scribed more fully  in the  report),  which
may apply to a wider variety of stormwater
pollutants,   are  high gradient  magnetic
separation  and  powdered  activated  car-
bon-alum coagulation.  The extent of re-
movals  will   depend   on   the   SS
characteristics and the level of treatment
applied.
  Screening can be divided into four cat-
egories with the size of the SS removed
directly related to the screen aperture  size
(Table  1).

-------
Table 1. Screening Categories

Screen Type         Opening Size
Bar screen
Coarse screen
Fine screen
Microscreen
>1 in. (>25.4 mm)
3/16-1 in. (4.8-25.4 mm)
1/250-3/16 in. (0.1-4.8 mm)
<1/250 in. (<0.1 mm)
  Bar and coarse screens have been used
extensively in WWTP at the headworks to
remove large objects.  Depending  on the
level of treatment required for the storm
runoff,  the  smaller aperture  sized coarse
screens  may  be  sufficient;  however, a
higher  level of treatment can  be achieved
using the bar and coarse screens in  con-
junction  with  the fine  or microscreens.
Design of screens can  be similar to that
for WWTP and CSO. Consideration, how-
ever, must be given to  stormwater char-
acteristics of intermittent operation  and to
possible very high initial loads, which may
not reflect  WWTP operation  characteris-
tics.   A  self-cleaning system should be
included for static screens to save manual
cleaning during storm events together with
automatic start up and  shut  down.   Cat-
enary screens (a coarse screen) are rug-
ged and  reliable  and commonly  used for
CSO facilities.  Therefore, they are likely
to be a good  screen for use with storm
runoff.
  Fine screens  and microscreens have
been developed  and used for SS removal
from CSO.   The removal  efficiency of
screening devices  is dependent on  the
aperture  (size of opening) of the screen
placed on the unit, making these devices
very versatile.  The efficiencies of a screen
treating a waste  with a typical distribution
of particle sizes will increase as the screen
aperture  decreases.
  Solids removal efficiencies are affected
by two  mechanisms:   straining  by  the
screen and filtering of smaller particles by
the mat deposited by the initial straining.
Suspended matter removal increases with
the increasing thickness of filter mat be-
cause  of the  filtering action of  the mat
itself.  This also   increases  the headless
across the screen.  One study showed
(on  a  23  |im  aperture  microscreen
[Microstrainer*]) that with a large variation
in the influent SS, the effluent SS stayed
relatively constant (e.g., both 1000 mg/L
and 20 mg/L influent SS would give a 10
mg/L effluent SS). Accordingly, treatment
efficiencies vary  with influent concentra-
tion.
  Generally,  microscreens and  fine
screens remove 25% to 90% of the SS,
* Mention of trade names or commercial products does
 not constitute endorsement or recommendation for
 use.
and 10% to 70% of the BOD5, depending
on  the  screen aperture  used  and the
wastewater being treated.
  Dual-media high-rate filtration (DMHRF)
(>8 gal/ft2/min [20 nf/nf/h]) removes small
particulates that  remain  after  screening
and floe remaining after  polyelectrolytes
and/or coagulants are added. As implied,
this provides a high level of treatment that
can be  applied after screening together
with automated  operation  and limited
space requirements.  Typically  a  unit is
composed  of 5 ft of No. 3  anthracite  coal
(effective size 0.16 in. [4.0  mm])  placed
over 3 ft of No.  612 sand (effective size
0.08 in.  [2.0 mm]).  This arrangement was
shown superior to both coarser and finer
media tested separately.
  Information is available  on the use and
design of DMHRF for treatment of drink-
ing  water,  but a number  of pilot studies
have also been done with the use of CSO,
which should provide more relevant infor-
mation.   The studies on CSO used vari-
ous diameter filter columns, with anthracite
and sand media with and  without various
dosages of coagulants and/or polyelectro-
lytes.  Removal efficiency for the filter unit
was about  65% for SS,  40% for BOD5,
and 60% for chemical  oxygen demand
(COD).   The addition of polyelectrolyte
increased the SS removal  to  94%, the
BOD5 removal to 65%, and  the  COD re-
moval to 65%.  The average filtration run
was 6 h at a hydraulic loading  of 24 gal/
ft2/min (59 m3/m2/h). SS removal increased
as influent SS concentration increased and
decreased  as hydraulic loading increased.
  Dissolved air flotation (DAF) separates
solid particles or  liquid droplets from a
liquid phase by introducing  fine air bubbles
into the liquid  phase.   As  the bubbles
attach to the solid particles, the buoyant
force  of the combined  particle and  air
bubbles is great enough to cause the par-
ticle to  rise.  Once  the  particles have
floated to the surface, they  are  removed
by skimming. The most common process
for forming  the  air bubbles is to dissolve
air into  the waste stream  under pressure
and then release the pressure to allow the
air to come out of solution.  The pressur-
ized flow carrying the dissolved air to the
flotation tank is either the entire stormwater
flow, a portion of the stormwater flow (split
flow pressurization),  or recycled DAF ef-
fluent.
  Higher overflow rates (1.3 to  10.0 gal/
ft2/min  [3.2 to 25  m3/m2/h]) and shorter
detention times (0.2 to 1.0  h) can be used
for DAF when compared with conventional
settling  (0.2 to 0.7 gal/ft2/min [0.5 to  1.7
m3/m2/h]; 1.0 to 3.0 h).  Studies for CSO
have shown that a treatment system con-
sisting of screening (using  a  297|im aper-
ture  with a  hydraulic loading  rate of 50
gal/ft2/min [122.3  m3/m2/h]) followed by
DAF can offer an  effective level of treat-
ment. The addition of chemical flocculent
in the form of ferric chloride and cationic
polyelectrolyte was also shown to improve
the removals. There are no data available
for treatment  of  separate storm  runoff;
however, from the CSO  data,  it  would
appear  that, except  for  sedimentation,
screening DAF is the most expensive treat-
ment system.
  Disinfection of  storm runoff requires a
different  approach from conventional dis-
infection  because the flows have charac-
teristics of intermittency, higher rates,  high
SS  content, wide temperature variation,
and  variable bacterial quality.  Residual
disinfecting capability  may not be permit-
ted, as chlorine residuals and compounds
discharged to natural waters may be harm-
ful to human  and aquatic life.  Coliform
counts are increased  by surface runoff in
quantities unrelated to pathogenic organ-
ism concentration. Total or fecal coliform
levels may not be the most useful indica-
tion of disinfection requirements and effi-
ciencies. Discharge points  requiring
disinfection are often at outlying points on
the   drainage  system and require  un-
manned,  automated installations. In addi-
tion,  a number of nonenteric  pathogens
found  in stormwater  runoff  have  been
linked to  respiratory illnesses and skin in-
fections.
  Table 2 shows disinfectants that might
be used  for storm flow disinfection. Con-
ventional municipal  sewage  disinfection
generally involves the use of chlorine gas
or sodium  hypochlorite as the  disinfec-
tant.  To  be  effective for disinfection  pur-
poses, a contact time of not less than 15
min  at peak flowrate and a chlorine re-
sidual of 0.2 to 2.0 mg/L are commonly
recommended.
  The characteristics  of storm  runoff  (i.e.,
intermittent and often  high flows) together
with the need to minimize capital costs for
a treatment operation  lend themselves fa-
vorably  to  use of high-rate disinfection.
This  refers  to achieving  either a given
percent or a given bacterial count  reduc-
tion through the  use of: decreased disin-
fectant contact  time,  increased  mixing
intensity, increased disinfectant concen-
tration, chemicals having higher oxidizing
rates, or various  combinations of  these.
Where contact times are less than 10 min
(usually  in  the range 1 to 5  min), ad-
equate mixing is a critical parameter; it
provides  complete dispersion of the disin-
fectant and forces disinfectant contact with
the maximum number of microorganisms.
Mixing can be done  by mechanical flash
mixers at the  point where  disinfectant is

-------
Table 2. Characteristics of Principal Storm Flow Disinfection Agents
Characteristics
                               Chlorine
                                                         Hypochlorite
                                                                                   Chlorine Dioxide
                                                                                                             Ozone
Stability

Reacts with ammonia
to form chloramines

Destroys phenols

Produces a residual

Affected by pH


Hazards
Stable

Yes


At high concentrations

Yes

More effective
atpH<7.5

Toxic
6-mo half-life

Yes


At high concentrations

Yes

More effective
atpH<7.5

Slight
Unstable

No


Yes

Short-lived

Slightly


Toxic; Explosive
Unstable

No


Yes

No

No


Toxic
added and  at  intermittent points,  or by
specially designed plug flow contact cham-
bers containing closely spaced, corrugated
parallel baffles that create a meandering
path for the wastewater.
  Swirl regulators/concentrators are com-
pact flow-throttling and solids-separation
devices that also collect  floatable mate-
rial.  Swirls are compact  units that func-
tion  as both a regulator  for flow control
and  as a solids  concentrator and, when
combined with  treatment  of the relatively
heavy settleable solids,  can  provide an
effective treatment system.  Performance
of swirls is very dependent on the settling
characteristics  of the  solids  in  the
stormwater. The EPA swirl is most effec-
tive at removing solids with characteristics
similar to grit (>0.008 in.  [0.2 mm] effec-
tive diameter,  2.65 specific gravity).   It is
important to appreciate this aspect of swirl
devices and to not expect significant re-
movals of fine  and low  specific gravity
solids.
               The three most common configurations
             are  the EPA swirl  concentrator,  the
             Fluidsep™ vortex separator, and the Storm
             King™ hydrodynamic separator.  Although
             each  separator is configured differently,
             operation  and the mechanism for solids
             separation  are similar.   Flow enters the
             unit tangentially and follows the perimeter
             wall of the cylindrical shell, creating a swirl-
             ing,  quiescent vortex flow pattern.  The
             swirling  action throttles the  influent flow
             and  causes solids to be concentrated at
             the bottom of the unit.  A degritter version
             of the EPA swirl has also been developed
             that  has no underflow and only removes
             the grit (detritus) portion.

             Beneficial Reuse of Stormwater
               The reuse of municipal wastewater for
             industry, nonpotable domestic usages, and
             groundwater recharge has been  practiced
             for many years.  In 1971,  an  EPA nation-
             wide survey estimated that current reuse
             of treated municipal wastewater for indus-
                              trial water supply, irrigation,  and ground-
                              water recharge was 53.5 billion  gal/yr, 77
                              billion  gal/yr,  and 12  billion  gal/yr (200
                              million m3/yr,  290 million  m3/yr, and  45
                              million m3/yr), respectively.   It is reason-
                              able to expect that reuse of treated waste-
                              water and/or stormwater will  increase  in
                              the future.
                                Many of the treatments  discussed are
                              apt to produce an effluent  quality of a
                              higher standard than that required to meet
                              a stormwater permit.  The intended reuse
                              will govern the  level of treatment required,
                              but careful selection,  design,  and use  of
                              pilot studies should result in  the required
                              effluent quality.
                                Increasing demands  on  potable water
                              supplies, in particular where a nonpotable
                              water  quality standard is required, will
                              make the concept of  reuse an  increas-
                              ingly more viable option.

-------
  Richard Field (also the EPA Project Officer, see below), Michael P. Brown,
    and William V. Vilkelis are with the Risk Reduction Engineering Laboratory,
    Edison, NJ 08837-3679.
  The complete report, entitled "Stormwater Pollution Abatement Technologies,"
      (Order No. PB95-100053AS; Cost: $19.50, subject to change) will be
      available from:
         National Technical Information Service
         5285 Port Royal Road
         Springfield, VA 22161
         Telephone: 703-487-4650
      Copies will also be available free of charge until supply is exhausted from:
         ORD Publications
         U.S. EPA-CERI,  G-72
         Cincinnati, OH 45268
         Telephone: 513-569-7562
         FAX: 513-569-7566
  The EPA Project Officer can be contacted at:
         Risk Reduction Engineering Laboratory
         U. S. Environmental Protection Agency
         Edison, NJ 08837-3679
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268

Official Business
Penalty for Private Use
$300
      BULK RATE
POSTAGE & FEES PAID
         EPA
   PERMIT No. G-35
EPA/600/SR-94/129

-------