v>EPA
United States
Environmental Protection
Agency
Office of Water
Washington, D.C.
EPA 832-F-99-079
September 1999
Decentralized Systems
Technology Fact Sheet
Recirculating Sand Filters
DESCRIPTION
A recirculating sand filter (RSF) system is a
modified version of the old-fashioned, single-pass
open sand filter. It was designed to alleviate the
odor problems associated with open sand filters.
The noxious odors were eliminated through
recirculation, which increases the oxygen content in
the effluent that is distributed on the filter bed.
RSFs are a viable addition or alternative to
conventional methods of treatment when soil
conditions are not conducive to proper treatment or
wastewater disposal through percolative
beds/trenches. Sand filters can be used on sites that
have shallow soil cover, inadequate permeability,
high groundwater, and limited land area. RSFs
commonly serve subdivisions, mobile home parks,
rural schools, small municipalities, and other
generators of small wastewater flows.
Sand filters remove contaminants in wastewater
through physical, chemical, and biological
processes. Although the physical and chemical
processes play an important role in the removal of
many particles, the biological processes play the
most important role in sand filters.
Figure 1 shows the three basic components of a
RSF system. These three components are a
pretreatment unit, a recirculation tank, and an open
sand filter.
Wastewater first flows into a septic tank (or in the
Pretreatment
To disinfection/discharge
Float valve
Recirculation
pump discharge
Filter media
Layered
support gravel
FIGURE 1 TYPICAL RECIRCULATING SAND FILTER
Sources: Mines, 1998 and NSFC, 1997.
-------
case of a clustered or community system, a number
of septic tanks) for primary treatment. A standard
concrete or fiberglass septic tank can be used, with
size being relative to the home/facility served.
The partially clarified effluent from the pretreatment
tank then flows into a recirculation tank. The
volume of the recirculation tank should be
equivalent to at least 1 day's raw wastewater flow
(or follow local jurisdiction requirements). In the
recirculation tank, raw effluent from the septic tank
and the sand filter filtrate are mixed and pumped
back to the sand filter bed.
APPLICABILITY
Stonehurst Development in Martinez,
California
The Stonehurst development is a small residential
subdivision near the City of Martinez in Contra
Costa County, California. This subdivision is
located in a hilly, rural area that did not have a
wastewater collection system. Thus, an innovative
decentralized wastewater system was designed to
provide for wastewater collection, treatment,
disinfection, and reuse.
The innovative system combines the use of septic
tanks, screened effluent filter vaults, high-head
effluent pumps, small-diameter variable grade
sewers, pressure sewers, a recirculating granular
medium filter, an ultraviolet (UV) disinfection unit,
a subsurface drip irrigation system for wastewater
reuse, and a community soil absorption field for
wintertime disposal. The principle elements for
treatment consisted of two sections of recirculating
granular filter followed by disinfection.
Each filter was 24 inches deep with 3 millimeter
gravel (washed and rounded with less than 2%
fines) sandwiched between layers of drain rock,
which was coarse, washed gravel approximately 1
to 2.5 inches in diameter. The wastewater was
pumped from the recirculating tank to the filters for
five minutes every half hour, and circulated
approximately five times through the filter. Since
one half of the filter was used during the time the
study was conducted, the hydraulic loading was 1.2
gal/ft2.
Performance data was calculated for 28 months
from June 1994 to September 1996, based on an
average of at least two samples per month for
five-day BOD, and at least four samples per month
for TSS, chemical oxygen demand (COD), pH, and
total coliform. Table 1 summarizes the performance
data of effluent samples that passed through the
recirculating gravel filter and the UV system.
To date, the Stonehurst decentralized wastewater
system has exceeded all expectations by performing
beyond required standards.
TABLE 1 PERFORMANCE DATA FOR
STONEHURST WASTEWATER
TREATMENT SYSTEM
Constituent
BOD5
COD
TSS
PH
Total coliform
NH4
N03
TKN
Oil and grease
TDS
EC
Range
0 - < 5 mg/L
1 -18 mg/L
2-15 mg/L
6.96 - 8.65 unitless
<2-12.5MPN/100mL
0-15 mg/L
3.55 -37 mg/L
0 - 3 mg/L
0-12 mg/L
340 - 770 mg/L
433 - 1,200 ^mhos/cm
* TDS - total dissolved solids, EC = electrical conductivity,
Aimhos/cm - micro mhos per centimeter
Source: Crites et al., 1997.
Elkton, Oregon
A RSF system was installed and monitored for a
community in Elkton, which is located on the
Umpqua River in Southwestern Oregon. The
population of this community was 350, mostly
residential with some commercial establishments.
The wastewater generated from stores, restaurants,
schools, and about 100 residences was first
pretreated and screened in individual septic tanks.
Partially clarified effluent was then collected and
-------
conveyed by an effluent pressure sewer system to a
RSF and finally pumped to a drainfield for final
treatment and disposal.
The sand filter was 60 feet x 120 feet with four
cells, 36 inches deep, and designed to treat 30,000
gallons per day (gpd). A recirculation tank of
29,500-gallon capacity was used with four
one-horsepower pumps. Each pump dosed one cell
at the rate of 130 gallons per minute. Two pumps
alternately dosed during each cycle. The actual
recirculation ratio was 3.2:1, and during low
periods, a motorized valve allowed 100%
recirculation.
Effluent quality data obtained from February 1990
through October 1997 are presented in Table 2.
It was concluded from this study that the RSF
produced a high quality effluent, thus protecting the
river nearby at an affordable cost. Capital costs for
RSFs range from $3 to $10 per treated gallon. The
annual operating costs are very low. For example,
at Elkton, the annual O&M cost for the RSF is less
than $5,000, which includes $780 for electricity.
Use of a smaller media (< 3.0 nm) would have
resulted in better nitrification, but this was not a
concern when the design was made.
TABLE 2 ELKTON'S RSF EFFLUENT
QUALITY DATA
Wastewater
Characteristics
BOD
TSS
NH3-N
NO3-N
Influent
(mg/L)
123
37
51
2
Effleunt
(mg/L)
4
9
10
26
Source: Orenco Systems, Inc., 1998.
ADVANTAGES AND DISADVANTAGES
Advantages
No chemicals are required.
RSFs provide a very good effluent quality
with over 95% removal of biochemical
oxygen demand (BOD) and total suspended
solids (TSS).
The treatment capacity can be expanded
through modular design.
RSFs are effective in applications with high
levels of BOD.
RSFs are easily accessible for monitoring and
do not require a lot of skill to maintain.
A significant reduction in the nitrogen level is
achieved.
If sand is not feasible, other suitable media
could be substituted that may be found
locally.
Less land area is required (1/5 of the land area
of a single-pass sand filter) for RSFs than for
single-pass sand filters.
Disadvantages
If appropriate media are not available locally,
costs could be higher.
Weekly maintenance is required for the
media, pumps, and controls.
Design must address extremely cold
temperatures.
DESIGN CRITERIA
The RSF system is an open sand filter with a sand
media depth of 2 feet. A layer of graded gravel
(about 12 inches) is provided under the sand for
support to the media and to surround the underdrain
system. A portion of the mixture (septic tank
effluent and sand filtrate) is dosed by a submersible
-------
pump through a distribution system that applies it
evenly over the sand filter. The dosing frequency is
controlled by a programmable timer in the control
panels.
The filtrate from the sand filter is collected by
underdrains that are located at the bottom of the
bed. The filter discharge line passing through the
recirculation tank is located near the top of the tank.
Figure 1 shows a ball float valve connected to a
downturned "T" on the discharge line, in which is
housed a rubber ball with a diameter slightly larger
than that of the pipe. As the filter effluent rises in
the tank, it forces the rubber ball firmly against the
bottom of the downturned leg, thus discharging the
effluent for further treatment or disposal. Other
control mechanisms may be used, but care must be
taken to ensure that the recirculation tank does not
run dry.
Table 3 gives typical design specifications for RSFs.
In very cold climates, the RSF design must include
elements that prevent freezing of standing water.
Distribution lines must drain between doses and
tanks, and the filter should be insulated.
PERFORMANCE
RSFs produce a high quality effluent with
approximately 85% to 95% BOD and TSS removal.
In addition, almost complete nitrification is
achieved. Denitrification also has been shown to
occur in RSFs. Depending on modifications in
design and operation, 50% or more of applied
nitrogen can be removed.
The performance of a RSF system depends on the
type and biodegradability of the wastewater, the
environmental conditions within the filter, and the
design characteristics of the filter. Temperature
affects the rate of microbial growth, chemical
reactions, and other factors that affect the
stabilization of wastewater within the RSFs.
Other parameters that affect the performance and
design of RSFs are the degree of wastewater
pretreatment, the media size, media depth, hydraulic
loading rate, organic loading rate, and dosing
TABLE 3 TYPICAL DESIGN CRITERIA
FOR RSFS
Item
Design Criteria
Pretreatment
Filter medium
Material
Effective size
Uniformity coefficient
Depth
Underdrains
Type
Slope
Bedding
Hydraulic loading
Organic loading
Recirculation ratio
Recirculation tank
Distribution and dosing
system
Dosing
Time on
Time off
Frequency
Volume/orifice
Minimum level: septic
tank or equivalent
Washed durable
granular material
1.0 to 3.0 mm
<4.0
24 in
slotted or perforated
pipe
0-0.1%
Washed durable
gravel or crushed
stone (0.25 -1.50 in)
3.0to5.0gpd/ft2/
(forward flow)
0.002-0.008 Ib/ft2/day
3:1 to 5:1
Volume equivalent to
at least 1 day's raw
wastewater flow
Pressure-dosed
manifold distribution
system and spray
nozzles where
permitted
< 2-3 minutes
Varies
48-120 times/day or
more
1-2 gal/orifice/dose
Source: Adapted from Crites and Tchobanoglous with
permission from The McGraw-Hill Companies, 1998.
techniques and frequency.
The effectiveness of a granular material as filter
media is dependent on the size and uniformity of the
grains. The size of the granular media affects how
-------
much wastewater is filtered, the rate of filtration, the
penetration depth of paniculate matter, and the
quality of the filter effluent. The finer the grain, the
slower the rate and higher the quality of the effluent.
High hydraulic loading rates are typically used for
filters that receive higher quality wastewater. The
accumulation of organic material in the filter bed
affects the performance of RSFs. As with hydraulic
loading, an increase in the organic loading rate
results in shorter filter life.
OPERATION AND MAINTENANCE
RSFs require routine maintenance, although the
complexity of maintenance is generally minimal.
Primary O&M tasks include monitoring the influent
and effluent, inspecting the dosing equipment,
maintaining the filter surface, checking the di scharge
head on the orifices, and flushing the distribution
manifold annually. The surface of the sand bed
should be kept weed free.
In addition, the septic tank should be checked for
sludge and scum buildup and pumped as needed.
The recirculation tank should also be inspected and
maintained.
The pumps should be installed with quick
disconnect couplings for easy removal. A duplicate
recirculation pump should be available for backup.
Listed in Table 4 are the typical O&M requirements
for RSFs.
COSTS
The cost of RSFs depends on the labor, materials,
site, capacity of the system, and characteristics of
the wastewater. One of the most significant factors
that affects the cost of sand filters is media cost.
Therefore, using locally available materials for the
media is usually the most cost-effective option.
Table 5 shows the costs for RSFs with sand media
and black beauty sand media used in a facility
treating 5,000 gpd. These are typical costs, actual
costs will vary from site to site and among different
designs. Local regulatory requirements and labor
rates will affect cost as well. The cost data in Table
5 includes the labor and machinery necessary to
TABLE 4 RECOMMENDED O&M FOR
RSFS
Item
O&M Requirement
Pretreatment Depends on process;
remove solids from septic
tank or other
pretreatment unit
Dosing chamber
Pumps and controls Check every 3 months
Timer sequence
Appurtenances
Filter media
Other
Check and adjust every 3
months
Check every 3 months
If continuous hydraulic or
biological overloading
occurs, the top portion of
the media can clog and
may need to be replaced
if not corrected in time
Weed as needed
Monitor/calibrate
distribution device as
needed
Prevent ice sheeting
Source: U.S. Environmental Protection Agency, 1980.
install media, plumbing, and tankage in the
excavation and landscape, the same should be noted
for the recirculation tank (minus the media).
The cost of the pretreatment unit(s) for a RSF
system will depend on the waste stream
characteristics specific to the site application.
Effluent sewer systems incorporate individual or
community septic tanks to pretreat wastewater
before it flows into the recirculation tank.
Developments that include commercial
establishments may require higher levels of
pretreatment in the form of additional septic tank
storage, surge capacity, grease traps, and possibly
aerobic digestion.
Suggested maintenance for RSFs range from weekly
inspections (15 to 30 minutes) to monthly
inspections (for approximately 1 hour).
The Ashco Rock Filter Storage II (RFSII) sand
filters consists of three different gradations of
-------
media; high spec black beauty sand, Ashco's Bottom
Zone, and spray grids with spray nozzles to
distribute the recycled filtrate evenly over the media,
all contained in 75 square foot precast concrete
cells.
TABLE 5 COST ESTIMATES FOR A 5,000
GPD FACILITY USING TWO DIFFERENT
MEDIA
Cost ($)
Item
Sand1
Black Beauty
Sand2
Capital Costs
Construction costs
Pretreatment May vary May vary
Recirculation 10,000 9,000
tank and pumping
system
Sand filter
Non-component costs
Engineering
Contingencies
Land
Total Capital Costs
Annual O&M Costs
Labor 20/hr 20/hr
Power May vary May vary
Sludge disposalฎ 10 50/yrb 50/yrb
cents/gal
Note: Non-component costs include piping and electrical.
Engineering and contingency each equal approximately 15%
of construction costs. Costs toward land, labor, and power
may be different from site to site and system to system.
a Design does not include precast concrete cells.
b Average pumping frequency is every 5 years.
Source: (1) Orenco Systems, Inc., 1998. and (2)
10,000a
May vary
3,000
3,000
May vary
26,000
43,100
May vary
7,800
7,800
May vary
67,700
REFERENCES
1. Anderson, D. L.; R. L. Siegrist; and R. J.
Otis. 1985. Technology Assessment of
Intermittent Sand Filters. U.S.
Environmental Protection Agency (EPA).
Municipal Environmental Research
Laboratory. Cincinnati, Ohio.
2. Ball, J. L. and G. D. Denn. 1997. Design of
Recirculating Sand Filters Using a
Standardized Methodology. Site
Characterization and Design of Onsite
Septic Systems. American Society for
Testing Materials. Fredericksburg, Virginia.
3. Crites, R.; C. Lekven; S. Wert; and G.
Tchobanoglous. Winter 1997. A
Decentralized Wastewater System for a
Small Residential Development in
California. The Small Flows Journal, vol. 3.
no. 1.
4. Crites, R. and G. Tchobanoglous. 1998.
Small and Decentralized Wastewater
Management Systems. The McGraw-Hill
Companies. New York, New York.
5. Hines, M. and R. E. Favreau. Dec. 9-10,
1974. Recirculating Sand Filter: An
Alternative to Traditional Sewage
Absorption Systems. Proceedings of the
National Home Sewage Disposal
Symposium, pp. 130-136. Chicago, Illinois.
6. Hines, M. Sept. 29-Oct. 1, 1975. The
Recirculating Sand Filter: A New Answer
for an Old Problem. Proceedings of the
Illinois Private Sewage Disposal
Symposium, pp. 68-78. Champaign, Illinois.
7. Martin, E. J. and E. T. Martin. 1991.
Technologies for Small Water and
Wastewater Systems. Environmental
Engineering Series, pp. 285-291. Van
Nostrand Reinhold (now acquired by John
Wiley & Sons, Inc.). New York, New York.
-------
8. Orenco Systems, Inc. 1998. Elkton,
Oregon: A Case Study. Orenco Systems,
Inc. Sutherlin, Oregon.
9. U.S. Environmental Protection Agency.
1980. Design Manual: Onsite Wastewater
Treatment and Disposal Systems. EPA
625/1-80-012. EPA Office of Water. EPA
Office of Research & Development.
Cincinnati, Ohio.
The mention of trade names or commercial products
does not constitute endorsement or recommendation
for use by the U.S. Environmental Protection
Agency.
ADDITIONAL INFORMATION
Infiltrator Systems Inc.
Technical Sales and Services Department
P.O. Box 768
Old Saybrook, CT 06475
Dr. Bruce J. Lesikar, Associate Professor
Agricultural Engineering Department
Texas A&M University System
201 ScoatesHall
College Station, TX 77843-2117
David L. Lindbo
Assistant Professor, Non-Agricultural Soil Science
Vernon G. James Research and Extension Center
NC State University, Dept of Soil Science
207 Research Station Road
Plymouth, NC 27962
Anthony Tarquin
University of Texas at El Paso
Civil Engineering Department
El Paso, TX 79968
David Vehuizen, P.E.
5803 Gateshead Drive
Austin, TX 78745
For more information contact:
Municipal Technology Branch
U.S. EPA
Mail Code 4204
401 M St., S.W.
Washington, D.C., 20460
IMTB
Extelence fh tompfance through optfhial tethnftal solutfons
MUNICIPAL TECHNOLOGY ~
------- |