EPA/832-B-96-006
October, 1996
ASSESSMENT OF VORTEX SOLIDS SEPARATORS
FOR THE CONTROL AND TREATMENT OF
WET-WEATHER FLOW
U.S. Environmental Protection Agency
Office of Wastewater Management
Municipal Support Division
Municipal Technology Branch
Washington, DC 20460
and
National Risk Management Research Laboratory - Cincinnati
Water Supply and Water Resources Division
Urban Watershed Management Branch
Edison, NJ 08837
-------
NOTICE
This document has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer review and administration review policies and approved for publication.
The material presented is for informational purposes only. This information should not be used
without first obtaining competent advice with respect to its suitability to any general or specific
application. References made in this document to any specific method, product or process does
not constitute or imply an endorsement, recommendation or warranty by the U.S. Environmental
Protection Agency.
-------
Acknowledgements
ACKNOWLEDGEMENTS
This manual is the product of the efforts of many individuals. Gratitude goes to each person
involved in the preparation and review of the document.
Contributors , .
Richard Field and Thomas P. O'Connor, U.S. EPA, National Risk Management Research
Laboratory, Urban Watershed Management Branch, Wet-Weather Flow Research Program,
Edison, NJ.
Peer Reviewers
*" ' ' ' ' - . "
Hugh Masters and Chi-Yuan (Evan) Fan, U.S. EPA, National Risk Management Research
Laboratory, Urban Watershed Management Branch, Edison, NJ.
Technical Direction and Coordination
Joseph Mauro, Work Assignment Manager,U.S. EPA, Municipal Technology Branch, Office
of Wastewater Management, Washington, DC.
-------
-------
Table of Contents
TABLE OF CONTENTS
Page
EXECUTIVE SUMMARY
Background . . . ........... . , . . . . ES-1
Approach and Scope ES-2
Findings and Conclusion ES-3
>-•-'••• ' • »
SECTION. 1 - INTRODUCTION TO VORTEX SOLIDS SEPARATORS
Background Information 1-1
Current Status .x. . . . . . .... .... .1-2
Process Description "... 1-9
Particles Removed by Vortex Separators 1-10
SECTION 2 - DESIGN CRITERIA
General 2-1
Swirl .'..'...'...'.'.• 2-4
Fluidsep™ ...... . ... 2-6
-Storm King™ . . . . ; 2-9
Degritters ..............r............... 2-12
Monitoring and Analysis . . ...... ... . .,. . . . . 2-15
Disinfection . . . 2-16
SECTION 3 - TECHNOLOGY ASSESSMENT
Performance Equations . 3-1
CSO Control Applications 3-6
Stormwater Control Applications 3-18
Swirl Degritter . . 3-25
Operations and Maintenance 3-25
Other Pollutant Removals 3-27
-------
Table of Contents
TABLE OF CONTENTS
(Continued)
Page
SECTION 4 - TECHNOLOGY COSTS
Design Costs . 4-1
Capital Costs '.. 4-2
Operation and Maintenance (O&M) Costs 4-5
SECTION 5 - SUMMARY OF FINDINGS
Performance Summary 5-1
Findings :.;....... 5-3
Recommendations 5-8
SECTION 6 - REFERENCES , '. 6-1
SECTION 7 - BIBLIOGRAPHY 7-1
LIST OF FIGURES
Figure 1-1 Alternative Process Arrangements (CSO) . 1-11
Figure 1-2 Alternative Process Arrangements (Stormwater) 1-12
Figure 1-3 Settling Curves 1-13
Figure 2-1 Swirl Settleable Solids Removal 2-3
Figure 2-2 Swirl Isometric 2-5
Figure 2-3 Fluidsep™ Isometric 2-8
Figure 2-4 Storm King™ Isometric . . . 2-10
-------
Table of Contents
TABLE OF CONTENTS
(Continued)
Page
Figure 2-5 Swirl Degritter . 2-13
Figure2-6 GritKing™ ............................. . 2-14
Figure 3-1 Swirl CSO Contol Performance, Washington, DC ......... . 3-11
Figure 3-2 Fluidsep™ CSO Control Performance, .
Tengen, Germany 3-14
Figure 3-3 Storm King™ CSO. Control Performance,
James Bridge, England * 3-17
Figure 3-4 Site Plan-for the West Roxbury, MA Facility ............. 3-20
. i • • • , "
Figure 3-5 Swirl Sampling Locations for West Roxbury, MA .... ... . . . . 3-22
Figure 3-6 Swirl Stormwater Control Performance,
West Roxbury, MA ............ . ... 3-24
Table 1-1
Table 1-2
Table ,1-3
Table 3-1
Table 3-2
fable 3-3
LIST OF TABLES
Swirl CSO Installations 1-4
Fluidsep™ CSO Installations 1-5
Storm King™ CSO Installations ... 1-6
Swirl CSO Control Performance, Washington, DC
Northeast Boundary Facility -•'. ....... 3-9
Fluidsep™ CSO Control Performance,
Tengen, Germany 3-13
Storm King™ CSO Control Performance,
James Bridge, England , 3-16
-------
Table of Contents
TABLE OF CONTENTS
(Continued)
Page
Table 3-4
Table 4-1
Table 5-1
Table 5-2
Swirl Stormwater Control Performance,
West Roxbury, MA 3-23
Comparative Unit Cost 4-4
Vortex Solids Separator Performance Summary '..... 5-2
Potential Pollutants Amendable to Treatment in
Vortex Solids Separators .5-5
IV
-------
Executive Summary
EXECUTIVE SUMMARY
ASSESSMENT OF VORTEX SOLIDS SEPARATORS FOR
THE CONTROL AND TREATMENT OF WET WEATHER FLOW
BACKGROUND
This document presents results of a technical evaluation of vortex solids separators for
the treatment of wet-weather flows (WWF). This evaluation provides information in support
of the Municipal Technology Branch (MTB) and the National Risk Management Research
v '
Laboratory (NRMRL) of the U.S. Environmental Protection Agency (EPA) mission to collect,
evaluate, and disseminate technical information on WWF control and treatment practices which
achieve the goals of the Federal Clean Water Act.
These evaluations explore design-related issues, identify specific weaknesses or
limitations, provide cost data and are beneficial in resolving operation and maintenance (O&M)
problems. In addition, the results of these evaluations identify a specific range of conditions
under which the processes or technologies demonstrate levels of performance efficiency. These
evaluations are an essential first step in disseminating actual data on the selected processes or
techniques. As such, this report is an assessment of vortex solids separators for the control
and treatment of WWF from separate storm-sewer systems. Because the preponderance of data
is from combined sewer overflow (CSO) application, CSO data was used.
ES-l
-------
Executive Summary
APPROACH AND SCOPE
Originally the primary objective of this report was to assess vortex solids separators for
the control and treatment of separate storm-sewer discharges. However, only a limited amount
of effectiveness data for storm-sewer applications is currently available. Therefore, CSO (and
other) data will be presented along with stormwater data. It should be understood that vortex
treatability is principally a function of particle settling velocity; therefore, use of data from
different types of flow, i.e., stormwater, CSO and river water, are justified and valid. This
report is prepared for engineers and scientists who wish to obtain a basic understanding of
vortex solids separators. It identifies current applications of this technology, presents
limitations, describes certain types of units, provides general cost information, and evaluates
performance using available information and data. This evaluation was derived from careful
consideration of data from:
• Vendors, developers, and
• Demonstration studies.
This project focused on existing data from the three commercially available types of
vortex solids separator units: the EPA swirl flow regulator/settleable-solids separator (swirl),
the Storm King™, and the Fluidsep™. No direct data collection was conducted. Nevertheless,
this evaluation represents a careful assessment of vortex solids separators. As additional vortex
solids separators are installed and new data become available, periodic reevaluations of this
technology are recommended.
ES-2
-------
Executive Summary
FINDINGS AND CONCLUSIONS
Based on data from CSO applications, the mass suspended solids (SS) removal from
vortex solids separators varied between an average of 38% at the Washington, DC full-scale
swirl facility to 54% at the full-scale Fluidsep™ facility in Tengen, Germany. The mass SS
I - ' , " ~
net removal (defined as the mass SS removal less the percentage of the underflowrinflow ratio)
was found to be lower. Average mass SS net removals varied between 7% at the Tengen,
Germany full-scale CSO Fluidsep™ facility and 17% at the pilot-scale stormwater swirl
facility at West Roxbury (Boston), MA. The average mass SS removal was 26% for separate
stormwater treatment in West Roxbury.
i
It was concluded that mass SS net removal by concentration into the underflow (actual
treatment) is approximately 30% or less. In CSO applications, most of the mass is removed
by regulation (flow splitting of the underflow) not by concentration. However, the vortex may
capture small storms that would otherwise result in overflows. This additional CSO
management advantage may not be obvious by focusing on the net removal of mass SS. It was
impossible to make any conclusions about the performance of one type of vortex unit relative
to the other due to site-specific flow characteristics and treatment objectives and the limited
available data..
Vortex solid separators performed better in terms of mass SS removal under lower
hydraulic loading (HL). Lower HL did not Improve mass SS net removal.
ES-3
-------
Executive Summary
Based on available information, capital costs in 1994 (ENR 5450) dollars attributable
to vortex solids separators were between $3,900 and $25,000/MGD of design flowrate
depending on specific site requirements. .
ES-4
-------
Section I - Introduction to Vortex Solids Separators
SECTION!
INTRODUCTION TO VORTEX SOLIDS SEPARATORS
The degradation of our nation's water bodies can partly be attributed to pollution in
. ' . • . • . '• V-
stormwater runoff and combined sewer overflow (CSO). During the past two decades,
significant research has been conducted to identify and test technologies to control wet-weather
sources of pollution that contribute to the degradation of the water bodies. One technology that
has shown promise for achieving pollution control while consuming minimal land space is the
vortex solids separator.
This section will introduce the concept of vortex separation, describe available units and
their application, and defines particles removed. Design criteria for the swirl and the
commercially available vortex separators are provided in Section 2. A performance assessment
of the technology for control of both CSO and separated stormwater is presented in Section 3.
Section 4 presents the costs of various technologies. Section 5 summarizes the performance
/ '- . ' • . • , - ,
characteristics, process limitations and recommendations.
BACKGROUND INFORMATION
Vortex separators were initially studied in Bristol, England, in the early 1960s as dual
purpose CSO regulator/suspended solids (SS)-liquid separator devices. Experiments were
1-1
-------
Section 1 - Introduction to Vortex Solids Separators
conducted to determine the hydraulic and performance characteristics of the vortex mechanism.
Results of early studies indicated that vortex separation was effective at regulating CSO and
concentrating SS into the underflow for further treatment while producing a significantly cleaner
overflow for discharge to a receiving water.
Li the 1970s, the U.S. Environmental Protection Agency (EPA) and the American Public
Works Association (APWA) expanded vortex separation technology with their own tests relative
to North American practice. The original purposes were to hydraulically regulate CSO as well
as to provide concentration of the settleable SS. Through hydraulic modeling studies the EPA
and the APWA developed design specifications for a triple-purpose (flow regulator/settleable-
solids separator/floatables collector) vortex device or swirl to control CSOs. Thorough design
information on swirls for CSO and stormwater control is available in an EPA design manual
(Sullivan et al, 1982). In the late 1970s the first full-scale swirl for CSO control was
constructed in Lancaster, Pennsylvania (Pisano et al., 1984). EPA also developed the swirl
degritter (Sullivan et al., 1974; Sullivan et al., 1977; Shelley et al. ,1981; Sullivan et al., 1982)
for settleable solids separation without flow regulation.
CTJRRENT STATUS
In addition to the swirl, two other vortex separators developed by private companies will
be emphasized in this report. One is the Fluidsep™ patented by a German firm, Umwelt-und
1-2
-------
Section 1 - Introduction to Vortex Solids Separators
Huid-Technik((UFT) and the other is the Storm King™ patented by Hydro International Limited
(HIL), a British firm. Each of the three units will be described in Section 2.
Vortex units are most often used for CSO control although they have been used to treat
stormwater runoff at two sites in the U.S. and one in England. There are at least 19 full-scale
swirl units in the U.S. and four in Japan, as shown in Table 1-1. Of the 24 swirls listed, 23 are
used to control CSOs. A swirl pilot unit was also tested for flow regulation and treatment of
stormwater in a separate storm-sewer system in West Roxbury (Boston), Massachusetts (Pisano
et al., 1984) (in the late 1970's, but has since been disassembled). Swirls are also located in
other countries, e.g., Holland, France, and Norway. As of this writing, there are 13 full-scale
Fluidsep™ units in the U.S. and Europe, as shown in Table 1-2, with additional units planned
for construction. The Fluidsep™ unit has only been applied to CSOs. There are no full-scale
Storm King™ units in operation in the U.S. at this time, however, there are more than 100
1 ""/
Storm King™ units in operation in Europe and Canada, as shown in Table 1-3. Full-scale Storm
King™ units are planned for the City of Columbus, Georgia, to treat CSOs. Stormwater
treatment by the HIL's Storm King™ has only been demonstrated at pilot scale in Bradenton,
Florida and by HIL's Grit King™, a full-scale degritting unit, in Surrey Heath, B.C., England.
1-3
-------
Section 1 - Introduction to Vortex Solids Separators
TABLE 1-1 SWIRL CSO
Location
United States
Auburn, IN
Brownsburg, IN
Brownsburg, IN
Decatur, IL
Decatur, IL
Decatur, IN
Lancaster, PA
Oswego, NY
Presque Me, ME
Syracuse, NY
Toledo, OH
Washington, DC
West Roxbury (Boston), MA~
Yonkers, NY
Japan
Nerima-Shiyakuji Prk
Chuo WWTP
Itabashi
Ouji
Diameter
(ft)
28.0
25.0
28.0
.25.0
44.0
18.0
24.0
36.0
18.5
12.0
32.0
57.0
10.5
19.0
36.1
97.1
30.2
78.7
Number
of Units
1
1
1
1
1
i
1
1
1
1
1
3
3.
1
3
1
1
1
1
INSTALLATIONS
Design Flowrate
Per Unit
(MGD)
32.0
25.0
36.0
40.0
113.0
37.0
26.0
60.0
14.0
6.9
51.7
133.0
3.9
25.7
8.4
78.1
NA
61.6
Hydraulic Loading'*
Per Unit
(gpm/ft2)
36
31
41
57
52
101
40
41
36
40
45
36
31
63
6
7
NA
9
" Hydraulic Loading (HL) is the design flowrate divided by the vortex chamber plan area and it is a nominal value since it does not account for
the reduction in area due to the overflow weir arrangement.
** Stormwater pilot-scale application.
Source: Sullivan et al.t 1982; H.I.L. Technology, 1993, and NKK Corporation, 1987.
1-4
-------
TABLE i-2 FLUIDSEP™GSO
Location
UNITED STATES
Burlington, VT
Decatur, IL
Decatur, IL
Decatur, IL
Saginaw, MI
Saginaw, MI
EUROPE
Tengeh, Germany
Number
.Diameter of Units
(ft)
40.0 1
45.0 1
44.0 4
27.0 1
36.0 3
36.0 1
10.0 2
INSTALLATIONS
Design
Flowrate
Per Unit
(MGD)
80.0
113.0
104.0
20.0
64.6
130.0
10.8
.
Hydraulic
Loading
Per Unit*
(gpm/ft2)
44
49
47
, 24
44
89
95
* See footnote for Table 1-1.
Source: Pisano, 1993.
1-5
-------
Section 1 - Introduction to Vortex Solids Separators
TABLE
Location
ENGLAND
Abersychan
Armagh West (NI)
Ashington
Bank Parade
Bargoed
Bexhill
BexhUl
Bexhill
Blaenau Ffestinog
Bluther Burn
Bluther Burn
Bowerfield D
Bum Beach
Burnham
Caroline Street, Langholm
Castleford
Clyde Park
Coatbridge
Coatbridge
Cowes (IOW)
Crewkerne
Crossways Parkj Caerphilly
Culcheth
Denmead
Dock road
DosthUl
Ely VaUey
Exwick and Redhills
Fountain Road
GeUi
Grange Lane
Grove Lane
1-3 STORM
Internal
Diameter
(ft)
29.5
10.0
19.7
• 18.0
21.3
9.8
13.1
16.4
25.0
14.8
13.1
24.6
14.8
263
5.9
21.3
33.8
24.0
29.4
14.8
13.1
16.4
21.3
13.1
17.2
16.4
16.4
17.3
18.0
16.4
21.3
25.0
KING™ CSO
Number
of Units
1
2
1
1
1
1
2
2
1
1
1
1
1
1
3
1
1
1
1 •
1
1
1
1
1
2
1
1
3
1
1
2
1
INSTALLATIONS
Design Flow
Per Unit
(MOD)
38.8
3.7
17.3
15.0
15.5
2.9
6.2
11.0
22.8
5.2
3.5
25.3
10.3
14.8
1.5
26.8
54.6
42.7
78.9
8.0
9.0
12.5
16.0
6.1
18.2
11.4
9.0
0.9
17.3
13.7 '
28.5
31.9
.
Hydraulic
Loading
Per Unit*
(gpm/ft2)
39
33
39
41
30
27
32
36
32
21
18
37
42
19
38
52
42
66
81
32
46
41
31
31
54
37
30 "
3
47
45
56
45
1-6
-------
Section 1 - Introduction to Vortex Solids Separators
.
Location
Haverfordwest
Invergowrie
James Bridge
Kirkby Stephen
Ladye Bay
Lamberhurst
Lanark
Langwith
Lochaline
Lochgelly
Manthorpe
Mayland
Middleton Cheney
Milbome St. Andrew
Moor Row (Cleater Moor)
Neath Link
New Road
Newport (IOW)
OldTebay
Oxford St. Maerdy
Oxford STW
Porterbrook
Portsmouth Relief D
Portsmouth Relief D
Portsmouth Relief D
Queensway
Rivacre
RMALake"
Sealstrand Dalgety Bay
Shenfield
SK Research \
Sneyd Lane
South Ballachulish STW
Southern Orbital
TABLE
Internal
Diameter
(ft) .
2,0
21.3
17.1
12.0
19.7
5.9
29.4
5.9
5,9
23.0
29.5
9.8
6.9
11.0
10.0
5.9
14.8
25.0 \
5.9
16.4
17.2
16.4
21.3
23.0
21.3
16.4
16.4
19.7
16.4
23.0
4.8
17.3
5.9
6.9
1-3 (Continued)
Number
of Units
2
1
2
1
1
1
1
3
1
2
2
1- '?
I
1
1
1
1
1
1
1
1
1
A 1
1
1
2
. 1
1
1
2
1
1
1
1
Design Flow
Per Unit
(MOD)
0.2
10.9
8.6
4.6
18.2
2.3
91.2 ,
1.9
1.0
27.6
29.1
2.1
1.4
1.6
4.2
1.6
10.7
105.2
0.5
11.8
18.2
12.3
17.1
26.2
21.7
19.7
11.4
16.0
14.0
10.6
1.4
7.4
1.3
2,3
Hydraulic
Loading
Per Unit*
(gpm/ft2)
44
21
26
28
41
58
93
48
25
46
30
19
26
12
37
41 .
43
149 .
13
39
54
40
33
44
42
65
37
36
46
18
54
22
33
43
1-7
-------
Section 1 - Introduction to Vortex Solids Separators
Location
Spa Slaithwaite
Spodden Valley
Stoke Canon
Summerhill
Swansea Road
Tatlers Farm
Totnes
Treorchy
Upminster
Warley Road
Wellingborough
Wellingborough
Wellingborough
Wellingborough
Wellingborough
Wenvoe
Wenvoe
West Pontnewydd D
White Bridge
Whitecliffe
Wick
Wigton Bypass
TABLE
Internal
Diameter
(ft)
23.0
21.3
9.8
8.9
19.7
12.0
• 13.1
23.0
14.9
16.4
8.9
5.9
7.9
8.9
8.9
5.9
5.9
14.8
5.9
8.2
16.4
10.0
1-3 (Continued)
Number
of Units
1
1
1
1
1
1
1
1
1
3
1
1
1
1
1
1
1 •
1
1
2
1
1
Design Flow
Per Unit
(MOD)
31.6
22.8
1.0
2.6
16.0
5.7
4.7 '
25.1
5.0
19.4
4.1
1.7
2.7
3.7
3.9
2.0
2.0
9.6
1.4
4.2
10.4
5.7
Hydraulic
Loading
Per Unit*
(gpm/ft2)
53
44
9
• 29
36
35
24
42
20
64
46
43,
38
41
44
51
51
39
36
55
34
50
CANADA
Gander, Newfoundland
29.5
Note: All Storm Kings™ are used for SS removal.
" See footnote for Table 1-1.
** Stormwater application.
Source: H.I.L. Technology, 1993.
1-8
-------
Section 1 - Introduction to Vortex Solids Separators
PROCESS DESCRIPTION
. ^
Vortex SS separators have no moving parts. The cylindrical chamber configuration
induces rotational forces that cause the separation and removal of settleable solids. During
storm-flow conditions, flow enters the unit - tangentially and a vortex is induced which
concentrates SS into the underflow and thereby reduces SS concentration in the clarified liquid
overflow. Vortex separation occurs when settleable SS circulating in the stationary unit are
directed tangentially outward from the fluid fiowfield arid downward by gravity. In CSO
applications, the concentrated SS are removed, from the bottom of the unit and conveyed via the
intercepting sewer to a wastewater treatment plant (WWTP). In separate stormwater
applications, the concentrated underflow may be routed to a holding tank or pond or can also
be routed to the WWTP if capacity (including sewerlines) is available.
In the case of the swirl degritter or the Grit King™ (HIL's vortex degritter) there is no
- • - • i • .| • , • . -
underflow. The grit collection zone is at the bottom of the vortex unit. The Surrey Heath Grit
King™ in England treats separate stormwater runoff from roof and highway runoff. Periodic
emptying of deposits is required from a bottom hopper.
For CSO applications, vortex separators may be used in-line or off-line. Dry-weather
flow (DWF) passes unimpeded through an in-line unit. Off-line units receive flow only when
1-9
-------
Section 1 - Introduction to Vortex Solids Separators
storm flows are diverted to the unit. In both in-line and off-line modes, the units can be used
in combination with other CSO control facilities (i.e., storage tanks or ponds). Figures 1-1 and
1-2 illustrates alternative arrangements that can be used with vortex flow regulators/separators
for CSO and separately sewered stormwater applications, respectively.
PARTICLES REMOVED BY VORTEX SEPARATORS
Vortex SS separators have been used for many CSO control applications, and for a few
separate stormwater applications. As a result, the bulk of existing information on vortex units
pertains to CSO applications. Although there are intrinsic differences between the two types of
wet-weather overflows, the water quality characteristics of CSO and urban stormwater runoff
are similar.
The design and performance of vortex-solids separators are based on solids' settling
characteristics. This characterization aspect of WWF is at least as important as the actual
concentration of solids for processes that depend on inertial separation. In the early 1970s EPA
provided curves shown in Figure 1-3 that was the basis of design for the swirls.
1-10
-------
Section 1 - Introduction to Vortex Solids Separators
FIGURE 1-1
ALTERNATIVE PROCESS ARRANGEMENTS (CSO)
OFF-LINE
ON-LINE
! 3
I
TVPE A - VORTEX' SEPARATOR
t
. TYPE B - FIRST FLUSH TO VORTEX SEPARATOR
TYPE C - PARALLEL
R—S
TYpr p - FIPST FljUSH TO ^TOPM TANK
VORTEX SEPARATOR
DIVERSION WEIR
II i ' STORAGE TANK
Source: Adapted from Pisano, 1993.
l-il
-------
Section 1 - Introduction to Vortex Solids Separators
FIGURE 1-2
ALTERNATIVE PROCESS ARRANGEMENTS (STORMWATER)
TREATMENT
PLANT
. SANITARY
INTERCEPTOR.
.—SMALL CONCENTRATE
TANK
STORM DRAIN
NETWORK
X f X
Source: Field and Masters, 1977.
1-12
-------
Section 1 - Introduction to Vortex Solids Separators
FIGURE 1-3
. SETTLING CURVES
U.S. Standard Steve Numbers
400 300 140 70 40 30 10
4 3 3/8
0.01
0.03
0.04 0.08
10.0
0.60 1.0
Particle diameter, mm
8.0
1 in. - 2.54 -cr.
Source: Sullivan er a/., 1982.
1-13
-------
Section 1 - Introduction to Vortex Solids Separators
Swirl SS separation efficiency also depends upon its design fiowrate, Q, and the fraction
of settleable solids included in the total storm-flow influent SS. The swirl was developed
through hydraulic modeling studies which used representative settleable model particles (based
on the Froude Number and Stokes Law) to simulate grit [fine sand] (specific gravity (SG) equal
to 2.65 and de from 0.2 to 2 mm) and relatively heavy organics (SG equal to 1.2 and de from
approximately 0.2 to 5.0 mm) (Sullivan et al, 1982). Floatables were also simulated (SG
range of 0.90 to 0.96 and d/s from 5 to 50 mm), it is important to appreciate this aspect of the
swirl's development and not expect significant removals of fine-grained and/or low-specific
gravity particles.
The swirl design manual (Sullivan et al. 1982) stating removal values of 70%, 80%,
90%, and 100% are for the removal of synthetic settleable solids used for swirl development.
A major portion of these simulated particles have settling velocitoies of 2.6 cm/sec or greater.
The swirl will also concentrate particles with lower settling velocities but with decreasing
effectiveness. The design manual (Sullivan et al. 1982) indicates the limit of SS removal
effectiveness is for particles with a settling .velocity of 0.14 cm/sec.
1-14
-------
Section 2 - Design Criteria
SECTION 2
DESIGN CRITERIA
Prior to designing a vortex solids separator, regardless of the type or multi-purpose
function, the mass and concentration of pollutants to be removed and the design flowrate, Q,
must be established. Other considerations include location and site structural limitations,
operation and maintenance strategies, and regulatory requirements.
Vortex separators should be designed to achieve the desired level of SS removal for a
statistical envelope of settling velocity distributions. Several other design parameters affect the
performance of the vortex separator and must be considered so that the desired control is
achieved. These parameters include the Qd, which should be verified by long-term continuous
modeling and the underflow rate. However, the SS settleability characteristics and dissolved
solids fraction of the incoming flow is of primary importance. Since vortex separators are
designed to remove grit-like solids and heavier organic particles, they will not concentrate the
finer SS found in wet weather flows. The desired removal efficiency of SS and associated
'pollutants will dictate the Q, allowable for the measured SS settling characteristics in conjunction
with the underflow rate. Typically, the underflow is designed to capture between 5 and 10
percent, of the Qj. ' >
2-1
-------
Section 2 - Design Criteria
Howxates less than the Qd result in higher removals by way of gravity separation and
flow reduction (a greater portion of the flow is diverted to the underflow). Figure 2-1 illustrates
decreased settleable-solids removal as a function of increasing flowrate. A swirl is still capable
of a reasonable degree of setfleable solids concentration when its influent flowrate is below twice
the Qj. Howrates that exceed twice Qd convey the settleable solids through the unit too quickly
and keep the particles suspended i.e., not removed by swirl concentration.
Small storm events may be fully captured in the underflow reducing the number or
volumetric quantity of overflows. The underflow is drained to the sanitary intercepting sewer.
In the absence of a sanitary sewer system or inadequate interceptor carrying capacity, a holding
tank or compartment would need to be part of the swirl, Fluidsep™, or Storm King™ system
and the tank would isolate the concentrated material for further treatment and disposal. Another
option is the swirl degritter, Grit King™ or other vortex degritter design that does not have an
underflow.
Although the performance of the three vortex units is based on a similar vortex SS
separation mechanism, each has its own design criteria. Therefore, the background and design
of the three types of vortex separators are discussed individually in the following text.
2-2
-------
Section 2 - Design Criteria
FIGURE 2-1
SWIRL SETTLEABLE SOLIDS REMOVAL
100
I
90.
80
70.
60.
•a so-
I
1 40,
I
& 30
20
10
0
Curve valid for both
- Grit larger than 0.35mm
- SettieaHe Solids larger than 1.0mm
Qd
Hydrograph Peak Discharge
2Qd
Source: Sullivan etal., 1972.
2-3
-------
Section 2 - Design Criteria
SWIRL
The swirl design, developed in the 1970s, is based on settleability studies and hydraulic
modeling (Sullivan et al., 1972). EPA derived settleability curves that were intended to
represent CSO SS settling velocity distributions using the Froude Number and Soke's Law to
scale-up the model for full-scale (prototype) results. The optimum design configuration was
based on 90% removal of the SS based on specified settling 'velocity distributions. Various
design curves are found in the EPA design manual (Sullivan et al, 1982). This allows a unit
to be designed for a desired removal and given paniculate settling velocity distribution after
determination of site-specific settling velocities and Qd.
The swirl configuration is shown in Figure 2:2. ' Flow enters through a tangential inlet
at the bottom of the unit. A flow deflector is located at the entrance ramp so that flow
completing its first revolution is deflected off the wall and inward preventing short circuiting.
The flow completes an additional revolution thereby following the longest path. SS and
floatables separation occurs as the flow circulates within the unit. The foul-sewer outlet conveys
DWF and storm flow having concentrated SS to the WWTP (for CSOs) or holding tank
(typically for stormwater application for periods when the sanitary intercepting sewer/WWTP
does not have enough capacity to accept the underflow). The original design for the swirl was
for CSO application, therefore, a primary floor gutter was included in the design. The primary
floor gutter conveys DWF from the inlet ramp to the foul-sewer outlet. This was done to
provide confined DWF and thus prevent solids deposition on the floor of the unit.
2-4
-------
Section 2 - Design Criteria
FIGURE 2-2
SWIRL ISOMETRIC
Inflow
Overflow
A Inlet ramp
B Flow deflector
C Scum ring
D Overflow weir and weir plate
E Spoilers
F Floatables trap
G Foul sewer outlet
H Floor gutters
I Downshaft ,
J Secondary overflow weir
K Secondary gutter
Source: Sullivan et al., 1982.
2-5
-------
Section 2 - Design Criteria
Buoyant materials quickly float to the top of the unit where they are directed to the
floatables trap. The mini-vortex in the floatables trap draws the captured floatables under the
weir plate where the floatables are contained by the weir skirt and weir plate arrangement. The
floatables eventually exit into the foul-sewer outlet during DWF drawdown.
Treated flow exits over the overflow weir and onto the weir plate. Spoilers on the weir
plate reduce rotational energy of the flow, thus increasing the overflow capacity of the downshaft
and improving the separation efficiency (Field and Masters, 1977). Flow that exits through the
downshaft is conveyed to further treatment, a holding tank, or the receiving water body (where
permitted).
The swirls installed in the U.S. (see Table 1-1) represent a wide range of sizes and
design criteria with diameters from 12 to 57 ft, nominal hydraulic loadings (HL) from 24 to 101
gpm/ft2, and Qd from 3.9 to 134 MGD. The swirls in Japan are designed with significantly
lower HL, all being under 10 gpm/ft2.
FLUJDSEP™
The basis of Fluidsep™ design is similar to the swirl in that SS settling velocity
distribution curves are used to determine the most appropriate unit dimensions. The Fluidsep™
design requires that site-specific settling velocity distribution curves and modeling be performed
2-6
-------
Section 2 - Design Criteria
by the proprietor. Once the settling characteristics are established, a unit is designed for that
site. The settling characteristics may take from three months to one year to establish.
How enters the Fluidsep™ through a tangential inlet near the bottom of the unit. The
inlet is designed to dampen the incoming flow velocities. Unlike the swirl, the Fluidsep™ does
not have floor gutters, spoilers, or flow deflectors.. This allows for an unimpeded vortex flow
pattern.
The unit has a conical baffle as shown in Figure 2-3. The conical baffle stabilizes an
inner vortex separation core where smaller particles have greater opportunity to be entrained in
the core and swept toward me foul-sewer outlet that is in alignment with the vessels rotational
axis. The unit is constructed with an angled floor (6% to 8%) that slopes toward the foul-sewer
outlet in the center of the unit. The slope as well as the smooth finish, are intended to enhance
the separated SS removal and facilitate washdown.
Clear flow exits the Fluidsep™ between the guiding, conical baffle, and the
scumboard/weirboard, as shown in Figure 2r3., The discharge exits the unit on the opposite side
it entered. An adjustable weirbarid allows the effluent flow to exit in a uniform peripheral
fashion preventing short circuiting before the flow is collected in a trough.
Floatables are trapped in an air cushion that slowly rotates under the cover. The air
cushion at the top of the vessel is created by the scumboard and the vent with the dip pipe.
2-7
-------
Section 2 - Design Criteria
FIGURE 2-3
FLUIDSEP™ ISOMETRIC
1
2
3
4
5
6
inflow from sewer
tangential inlet
vortex chamber
underflow
flow restrictor
interceptor to
8
overflow to re-
ceiving water
8 adjustable weir
band
9 aircushionfor
floatablestrap •
10 vent with dip-pipe
11 guiding baffle
12 scum board
13 cover
14 electronic level
monitor with data
logger
15 pressure pick up
16 outlet cone
17 outlet plug with
lever
18 depot for coarse
sediments
19 screen
20 support leg
21 inner processing
vortex core
13
im
Source: Brombach er c/., 1993.
2-8
-------
Section 2 - Design Criteria
After a storm event the fioatables are removed via the foul-sewer outlet and conveyed to the
WWTP. Fioatables trap capture effectiveness varies with the flowrates. During extremely
highflowrates the fioatables often escape the unit. The unit can be modified to include screening
devices above the overflow outlet to trap any escaping fioatables.
The Fluidsep™ units installed in the U.S., as shown in Table 1-2, also represent a wide
range of sizes and design criteria. The diameters range from 27 and to 44 ft and are typically
2.5 times greater than the height, but vary between 0.5 and 3.0 times the height (Pisano, 1993).
The HL range from 32 to 95 gpm/ft2 with Q, (per unit or group of units) from 25 to 416 MOD.
STORM KING™
Prior to installing a Storm King™ at a site, a pilot study should be performed to
determine dimensions most appropriate for the site. It is not always necessary to perform a pilot
study if the particle settling velocity distribution for the drainage area is known. Pilot studies,
however, reduce the possibility of installing an incorrectly sized unit. Pilot units are usually 3
to 6 ft in diameter. Information required for the pilot study would include the desired SS (or
pollutants) removal, Qd, and the particle settling velocity distribution. The HL is varied during
the pilot study to establish the HL at which optimum removal is achieved.
The Storm King™'s configuration consists of a cylindrical vessel with a solid central
cone, a sloped floor, and a top assembly, as shown in Figure 2-4. A majority of the units in
2-9
-------
Section 2 - Design Criteria
FIGURE 2-4
STORM KING™ ISOMETRIC
DIP PLATE AND SPILLWAV
ASSEMBLY —
SUPPORT FRAME
-OP BAFFLE
BAFFLE PLATE
VENTURI PLATE — ..
CENTER SHAFT . _X\
OVERFLOW CHAMBER _
OVERFLOW TO OUTFALL
CENTER CONE
CONCRETE FILL
BENCHING .
CONE BASE
FOUL OUTLET
PIPE TO SANITARY SEWER
- TANGENTIAL INTAKE PIPE
_ INTAKE DEFLECTOR PLATE
(optional)
SANITARY OUTLET PIPE SPIGOT
CONCRETE CHAMBER POURED IN PLACE OR
PRECAST SEGMENTAL SHAFT RINGS
leg Cnarcon On* Paul
CONCRETE BASE (ooureo in puce)
Source: H.I.L. Technology, Inc., 1991.
2-10
-------
Section 2 - Design Criteria
England are prefabricated, however, cast-in-place units are available. Dimensions of the units
will vary, but generally, the diameter is twice as large as the depth (Hedges et aL, 1992). How
enters the unit tangentially through an entry port located halfway up the vessel wall. Similar to
the swirl, a deflector plate can be constructed at the entrance that will prevent the heavily
polluted storm flows from bypassing treatment and immediately exiting via the overflow. SS
removal is also aided by the slope of the floor, which is between 10° and 30°. This slope
provides an additional benefit of minimizing solids collecting on the benching during drawdown.
The concentrated ^underflow exits via a helical channel, which is identified as benching in Figure
2-4, and is conveyed to the WWTP or holding tank. The benching is located midpoint between
the outside and center axis of the unit, therefore decreasing the energy required to move the SS
to the outlet. Furthermore, the outlet is located beneath the dip plate where the shear zone forms
and the greatest vortex activity and separation occurs. This differs from the other two vortex
designs that have exit points in or near the center of the unit.
The flow that was directed down the perimeter of the unit is then directed toward the
center of the unit and up the center cone. The flow rotates at a slower velocity during this
action than the velocity that occurred during the downward flow. The clear flow rises up toward
the baffle plate and exits the chamber between the baffle plate and the dip plate and is then
conveyed to the overflow chamber. The dip plate locates the shear zone, which is the interface
between the outer, downward flows and the inner, upward flows.
2-11
-------
Section 2 - Design Criteria
Floatables are also removed by the Storm King™. The buoyant materials move upward
and outward and become trapped behind the dip plate. When the storm flow ends and the unit
drains down, the floatables exit out the foul-sewer outlet and are conveyed to the WWTP or
storage tank.
The Storm King™ installed in England, as shown in Table 1-3, have diameters from 2
to 33.8 ft, HL from 3 to 149 gpm/ft2 and Qd from 0.2 to 105.2 MGD.
DEGRTTTERS
Vortex-type degrittefs come in various forms, two of which are:
• EPA swirl degritter, and
• HIL Grit King™
The swirl degritter does not have a continuous underflow to the WWTP or holding tank. Instead
a relatively dry mass of settleable solids (grit/detritus) collects in a 60° conical-bottom hopper
for intermittent removal. Degritters have been used in CSO, stormwater, potable water and
river water intake applications. Figures 2-5 and 2-6 contain plan and elevation views of the
swirl degritter and the Grit King™.
2-12
-------
Section 2 - Design Criteria
FIGURE 2-5
SWIRL DEGRTTTER
INLET
" CRIT
. CHAMBER
WASHWATER
OVERFLOW WEIR
WASH WATER
OUTLET
CRIT WASHER
AND ELEVATOR
^' * V-^'-•'.*•'—''•••••^-•V--:-
SECTION A-A
Source: Sullivan et al., 1982.
2-13
-------
Section 2 - Design Criteria
FIGURE 2-6
GRTTKING™
I
CENTER SHAFT & SUPPORT —
FRAME (REMOVABLE FOR \
INSPECTION 4 MAINTENANCE) ~
— 'NSPECT10N
SUPPORT LEGS
•.* No/
-i mr*Ti/->Ki OVERFLOW OUTLET
-J.FVATION CHANNEL W,TH
Fl>NCED CONNECTION
TWL UPSTREAM , N.
OVERFLOW
CHANNEL
INSPECTION
PANEL.
NOTES:
. TIC UNCUT a CZMMt. MO
ruoici OW.Y.
:. TIC
tnmtn. MO ajtJrcr re
I»CH src Tconc vournxxn.
3. THC OMtMTnnOH (OH BUM) Cf MUKt. WORCTT.
v ounzr MC * norccnoM MMXS ow ac
TO 9UT TIC *
*OJUSTCD TO SUIT THE OCWCOV
Source: H.I.L. Technology, Inc., 1991.
2-14
-------
Section 2 - Design Criteria
MONITORING AND ANALYSES
Proper sampling, flow measuring, and analysis are a must for design and treatability
evaluation. Prior to selecting the swirl or vortex separator for a combined-sewerage or
separately-sewered stormwater system, adequate volumes of representative samples of the storm
flow should be collected by use of appropriate sampling techniques. The particle settling
velocity distributions of these samples as related to total solids and SS and associated pollutant
content should then be determined. This analysis is essential for assessing the applicability of
vortex separators. If the storm flow does not contain enough SS with grit-like particles (SG >
2.65 and de from 0.2 to 2 mm) and relatively-heavy-organic particles (SG > 1.2 and de from
approximately 0.2 to 5.0 mm) then swirl and vortex technology may be inappropriate and
alternative technologies should be used. As previously noted, the swirl will also concentrate
particles with lower settling velocities down .to 0.14 cm/sec but with decreasing effectiveness.
The variable nature of storm flow and sewer .slope influence suspended-/settleable-solids
concentration and particle-settling-velocity distribution. In addition, the build up of these settieable
solids in the sewer system; is usually a function of the length of the antecedent dry-weather period.
Furthermore, suspended-/settleable-solids concentrations will vary with time during the storm event.
.These storm flow variations require that sampling be done for the duration of the storm event and
for several storms in order to develop a long-term average of the settleable-solids concentration and
particle settiing'-velocity distribution. _
2-15
-------
Section 2 - Design Criteria
Sampling devices must be able to capture the heavier SS or settleable solids (i.e., that fraction
of the SS that the swirl was developed to remove) and not manifest biassied results due to
9 ; ' •
stratification. For an automatic sampling device, this means that its intake velocities and ports must
be greater than the main stream velocity and must be placed at multiple levels, respectively, in order
to capture the heavier particles near the channel invert.
After samples have been collected and analyzed for SS, two particle settling characteristic
analyses should be conducted. One for settleable solids (gravimetric) and the other for settling-
velocity distribution. These analyses will enable a site estimate of the percent of SS the swirl is
capable of removing.
If particle-settling velocities indicate that swirl technology will remove an acceptable
percentage of the particles in the storm flow, then hydrological and hydraulic studies should be
conducted to determine the Qd. This analysis of flow should be done on a long-term continuous
basis using mathematical modeling and then directly measuring flowrates for calibration and
verification to achieve the best Qd and settleable-solids removal prediction.
DISINFECTION
Swirl/vortex separators used in CSO applications can'be often modified to include
disinfection. The mode of disinfection to be applied may require a higher solids removal than
the mandated discharge requirements.
2-16
-------
Section 2 - Design Criteria
Swirl/vortex separators can be placed upstream and/or downstream of disinfectant
addition. A benefit of disinfectant addition upstream of the swirl is that mixing by the swirling
action will increase collisions between the disinfectant and, the microorganisms, potentially
resulting in a more effective kill per unit contact time. This was done using chlorination at
Lancaster, PA (Pisano et al, 1984) and Syracuse, NY (Drehwing et al, 1979). However,
additional laboratory analysis is necessary to determine the effectiveness of disinfection due to
protective particles in the overflow. Microorganisms may survive in the interstices of the larger
organic particles and in the micro-fractures of soil grain's. The swirl/vortex units would not
remove all these particles and therefore would allow a portion of the particles and their occluded
microorganisms, to overflow to the receiving waters.
2-17
-------
-------
Section 3 - Technology Assessment
SECTIONS
TECHNOLOGY ASSESSMENT
Data from performance tests were obtained on each of three types of vortex separators
and a swirl degritter. The data discussed in this section are from a stormwater treatment
demonstration project in West Roxbury (Boston), Massachusetts (swirl-pilot scale); three full-
scale CSO control demonstrations in Washington, DC (swirl), Tengen, Germany (Fluidsep™),
and James Bridge, England (Storm King™); and a full-scale demonstration in Tamworth,
-( . - ' *
N.S.W., Australia (swirl degritter) of pretreatment of intake river water for potable supply.
""*• , = •» p " '
PERFORMANCE EQUATIONS
The performance data available for evaluation in this report were presented so that similar
comparisons could be made between each of the facilities except the previously mentioned
degritting units which do not have an underflow. The fraction of the influent SS (and associated
pollutants) that are concentrated by the vortex separators into the underflow indicates the portion
of SS diverted from the effluent (overflow) to the WWTP. The three performance indicators
for swirl and vortex treatment are: "
• Removal
• Net Removal
• Treatment Factor (TF)
3-1
-------
Section 3 - Technology Assessment
The aforementioned performance indicators are defined by the terms in these storm-flow-
event summation equations:
(1)
(2)
C =
(3)
where M = storm-flow-event pollutant mass loading (mass), V = storm-flow-event flow volume
(volume), C = storm-flow-event flowrate-weighted-average pollutant concentration (flow-
weighted-average concentration), cfej) = average pollutant concentration between samples,
~ average flowrate between samples, and A/ = time interval between samples.
These terms can be combined to form this equation for Removal:
Removal = —-
cv,
M.-M,
x 100% = —! - » 100%
M,
(4)
where M{ = QVf = mass of untreated influent and Me = CeVe = mass of treated effluent.
3-2
-------
Section 3 - Technology Assessment
For most treatment unit operations, e.g., settling, screening, and filtration, the volumes
of the influent and effluent are equivalent and therefore may be canceled out of Equation 4.
Removal (4) can then be rewritten in terms of concentration alone. However, this approach,
using only concentration to measure performance, cannot be used for the swirl because the swirl
treatability evaluation is complicated by the continuous and relatively dilute underflow. The
volume of the underflow remains significant throughout the event and its percentage of the
influent is highly .variable (5% to 10% at Qd and as high as 100% for the smaller storms which
are completely captured in the swirl chamber without causing effluent). Therefore a different
approach is necessary to evaluate the concentrating effect of the swirl and other vortex units.
Assuming there is no net concentrating effect by the swirl or vortex, i.e., SS
concentration of the influent, effluent, and underflow are all equal, then C would cancel out of
Equation 4. The equation would then only reflect the flow splitting nature of these devices and
is termed the Reduction:
V.-V
Reduction = ' e x 100%
(5)
where: V{ = influent volume and Ve = effluent volume. This non-concentrating flow
phenomenon is similar to what occurs during the operation of conventional CSO flow
regulators. Accordingly this can further be thought of as the CSO pollution reduction resulting
from a conventional flow regulator.
3-3
-------
Section 3 - Technology Assessment
Removal (4) and Reduction (5) can now be combined to define the Net Removal and TF
which represent the pollution Removal above and beyond the Reduction gained by conventional
! '
CSO flow regulation. The equation for Net Removal is:
Net Removal = Removal - Reduction
(6)
A positive Net Removal indicates that SS (and associated pollutant) concentration has taken place
in the unit. The TF equation is :
Removal
Reduction
(7)
where Ct = influent flow-weighted-average concentration and Cu = underflow flow-weighted-
average concentration. TFs greater than 1 indicate that the vortex separator is concentrating SS
to the underflow. The higher the TF the better the vortex device concentrates pollutants. A
negative Net Removal or a TF less than 1, indicates an anomaly or faulty sampling and
monitoring techniques.
3-4
-------
Section 3 - Technology Assessment
These equations define the terms used in the tables presented in the following subsections.
In the case of the swirl degritter and Grit King™, as with most forms of treatment, the influent
- /
volume can be treated as equivalent to the effluent volume.
The volume of underflow, a function of the Qd for the unit, is very significant in the
calculation of performance. The Net Removal and TF for a storm-flow event that does not
overflow in the swirl or vortex unit will be 0% and 1, respectively, while Removal is 100%.
Events that barely overflow the unit will have high Removals with low Net Removals and TFs
due to the proportionately larger volume in the underflow. As the storm flow increases towards
the Qj, Removal will begin to decrease while Net Removal and TF should increase due to
, significantly decreased volume (5 to 10% at QJ of the underflow.
Two important factors that will effect performance measurements are:.
• sampling and flow measuring techniques
• variation of SS loading and influent flowrate
To provide a "true" measurement of performance, sampling devices must be able to
capture the heavier influent SS that the swirl/vortex was designed to remove. As previously
mentioned in Section 2, the intake ports of automatic sampling devices need intake velocities
greater than the main stream'velocity of the influent SS being sampled and must be placed at
.3-5
-------
Section 3 - Technology Assessment
multiple levels in order to capture the heavier particles at the channel invert without reflecting
bias due to stratification. Sampling of SS and other pollutants must also be synchronized with
flow measurements.
The data presented in this report only represent averaged Removals^ Net Removals, and
TFs for specific events and the total average of these events. To gain a better understanding of
the capture of pollutants and the most polluted segment of the storm flow please refer to the
individually referenced reports which display the capture of pollutants through a storm event.
CSO CONTROL APPLICATIONS
Evaluation of the Swirl
Three swirls were evaluated as part of Washington DC's CSO abatement program. The
abatement program was required because of a high sediment oxygen demand downstream and
depletion of dissolved oxygen in the Anacostia River. These two factors resulted in frequent fish
Mils and the elimination of game fish. In addition, public health standards for coliform bacteria
resulted in restriction of water contact recreation.
The swirls are in an enclosed facility. In addition to the swirls, an automated inflatable
3-6
-------
Section 3 - Technology Assessment
weir system was installed upstream to provide in-line storage at nine of the largest, overflows
in the conveyance system. Flow from approximately 4,000 acres are treated by either the swirls
and/or stored upstream of the weirs for subsequent treatment by the Blue Plains WWTP.
The automated inflatable weir system was designed to maximize the storage of wet-
weather flows within the sewer system. During extreme high flows, the weirs are deflated when
the level in the sewer becomes too high, thus preventing upstream flooding. The swirls and weir
system operate automatically. A telemetry system allows for real-time control at the Bureau of
Sewer Services and at the Blue Plains WWTP.
Dry-weather flows are conveyed from the Northeast Boundary (NEB) CSO Swirl
Treatment Facility to the Blue Plains WWTP, which has a capacity for complete treatment of
740 MGD with an additional capacity of 336 MOD for primary treatment. However, when the
flow exceeds 15 MGD .during wet-weather events, the flow is diverted and gravity fed to the
swirls. The swirls have a total Qd of 400 MGD, although they have been operated at 500 MGD
on occasion.
- * ' , .
Minor modifications were made to the swirl design due to site constraints. For example,
the depth of the swirl was decreased and the diameter was increased, while maintaining the same
capacity in accordance with the design specifications (Sullivan et al, 1982). The swirls were
installed as part of a CSO abatement system, which includes the upstream inflatable weirs, an
3-7
-------
Section 3 - Technology Assessment
upstream bar screen which captures the larger debris (i.e., cans, bottles, leaves) and high-rate
disinfection following the swirls.
The three swirls are identical in design. They are ,57 ft in diameter, 6.5 ft deep, have
a Qd of 134 MGD for a combined total of 400 MGD, and have a per unit hydraulic loading (HL)
of 36 gpm/ft? (O'Brien and Gere, 1992). Flow to the foul-sewer lines is controlled and
approximately 8 MGD per unit (or 5% of the Qj) are conveyed as underflow to a downstream
pumping station. Each unit also is equipped with an automated washdown system.
During the first year of operation sampling and analyses were performed to determine
the swirls' effectiveness. Samples were collected at the screening, the influent chamber, the
downshaft, the overflow weir, and at the foul sewer (underflow). The monitoring was originally
set up to be done automatically, but conditions, e.g., SS stratification, necessitated the
monitoring to be performed manually.
Settleability tests were performed which indicated that the CSO contained a large fraction
of SS with settling velocities lower than what the swirl was developed to remove. SS Removals
based on HL were predicted for each storm event, as shown in Table 3-1. The SS Removals
for all but one storm event exceeded the predicted removals.
3-8
-------
Section 3 - Technology Assessment
TABLE 3-1 SWIRL CSO CONTROL PERFORMANCE, WASHINGTON, DC
NORTHEAST BOUNDARY FACILITY
Storm
Number
1
3
4
5
7
8
9-SW2
10
11
12
Average
Flow
Date Range
(MGD)
3129191 18 - 90
5/6/91 17-30
6/16/91 8-50
6/18/91 15 - 103
8/27/91 10 - 60
9/24/91 63-100
10/6/91 11-68
10/17/91 23 - 51
11/22/91 8 - 58
12/2/91 45 - 95 '
Avg
Avg Foul Volume
Flow Flow Reduction
(MOD) (MGD) (%)
62 9.4 IS
21 11.6 55
23
50 7.7 15
35 9.2 26
72 8.4 , 12
27
33 . " '
29 8.4 29
63 7.9 13
42 8.9
Influent
FlowWgtd
Avg SS
(mg/1)
183
297
211
250
180
365
510
140
104
231
Mass SIS Mass
MassSS Predicted SS Net
Removal Removal Removal-
(56) (%) (%)
28.3 16.6 13.1
67.5 67.0 12.3
39.4 ,
27.9 22.2 12.5
42.0 .. 26.0 15.7
35.6 . 10.5 23.9
48.9 .
73
29.9 25.7" 0.9
21.1 32.1 8.6
37.6 28.6 12.4
TF
1.87
1.22
1.81
1.60
3.05-
34.9
1.03
1.68
1.75
Note: No settling characterization performed for storms 4, 9, and 10. Storm #3 contains data from relatively low flowratei. Design
flowrate of 134 MGD with an HL 36 of gpm/fi*. SS Mass Removals, Predicted Removals, and Net Removals are determined using
mast loadings. , , '
Source: O'Brien and .Gere, 1992 . -
3-9
-------
Section 3 - Technology Assessment
The average Net Removal for the storm events was 12.4%. This indicates that SS are
being concentrated into the underflow and a portion of the Removal (Net Removal) is not simply
due to flow splitting. Figure 3-1 illustrates Removals, Net Removals and predicted removals for
several storm events sampled in 1991. Performance was also evaluated by determining TF. All
TFs were greater than 1 with the average being 1.75, which indicates that the swirl is
concentrating SS to the underflow.
Evaluation of the Fluidsep™
Only one thorough evaluation of a full-scale Fluidsep™ has been performed. This
evaluation took place at the Tengen, Germany facility. This facility was constructed in 1987 to
treat CSOs from a 25.3 acre drainage area and contains two Fluidsep™ units with a total design
flowrate of 19 MOD. Dry-weather flowrate was Ql2 MGD, or approximately one hundredth of
the design flowrate. Their chamber diameters are approximately 10 ft and have an overall
specific volume of 254 ftVacre, which includes upstream wet-weather flow pipe storage.
Underflow from the units, which peaks at a combined rate of 0.81 MGD, discharges to a trunk
sewer and is then conveyed to a WWTP (Brombach et aL, 1993).
3-10
-------
Section 3 - Technology Assessment
FIGURE 3-1
SWIRL CSO CONTROL PERFORMANCE, WASHINGTON, DC
70
60
50
O
ill
DC
40 -
30 --
20 •
10 -
i
, , 3 4 5 7 8 9- 10 11
SW2
EVENT
I PREDICTED REMOVAL Q REMOVAL ^ NET REMOVAL
12
3-11
-------
Section 3 - Technology Assessment
The first phase (1987-1988) of the performance evaluation had 134 facility storm
activations. Of these only 80 events, or 47.8%, resulted in overflow events with discharge to
the receiving stream. The remainder of the events did not result in overflow events due to the
upstream and Fluidsep™ storage (Brombach et aL, 1993).
During the second phase of the evaluation, samples from the underflow and overflow
were collected during five storm events. Each of the samples were analyzed for SS, settleable
solids, COD, and phosphorus. The SS settling velocities were also determined. Removals, as
well as the HL for each event and other pertinent data, are presented in Table 3-2. Figure 3-2
illustrates Removal and Net Removal for the storm events sampled.
The average Net Removal and the TF, as shown in Table 3-2, for the storm events were
6.9% and 1.19, respectively, indicating limited treatment. The ZPand Net Removalare low due
to the high volume of underflow which ranged from 24% to 82% of the inflow.
Evaluation of the Storm King™
There are two 17.3 ft diameter Storm King™ units operating in parallel at the James
Bridge facility in Walsall, England. This facility treats CSOs from a 39 acre drainage area at
a design flowrate of 16.8 MGD. Due to a high drainage area perviability, runoff was
3-12
-------
Section 3 - Technology Assessment
TABLE 3-2 FLUIDSEP™ CSO CONTROL PERFORMANCE,
TENGEN, GERMANY
Antecedent •
Storm
Number
1
2
3
4
5
Average
Dry
Period •
(d»y«)
7
14
5
1.1
0.2
Rain
Depth
(in.)
03
03
0.9
0.2
0.7
0.5
Rain
Duration
0>r)
2.6
2.3
4.5
33
5.5
3.6
Overflow
Duration
-------
Section 3 - Technology Assessment
<
o
LL1
DC
100
90
80
60
50
30 -
FIGURE 3-2
FLUIDSEP™ CSO CONTROL PERFORMANCE,
TENGEN, GERMANY
-10
EVENT
D TOTAL REMOVAL • NET REMOVAL
3-14
-------
Section 3 - Technology Assessment
significantly reduced to the separators. During the study, one of the units had to be shut down
to force an overflow in the operating unit.
The number of overflow events were infrequent as a result of an unusually dry year.
Data from only three of the events were analyzed and are presented in Table 3-3. As shown,
events 2 and 3 were so small that greater than 90% of the flow ended up in the underflow and
this is reflected in the TF which indicate that significant SS concentration does not occur.
Due to the insufficient storm runoff, an additional six tests were conducted using river
water. Sanitary sewage was added to the river water to simulate a CSO. The volume of the
underflow for the six tests ranged between 9% and 36% of the inflow volume. Table 3-3 and
Figure 3-3 show the results of the second phase.
/ • •
Performance results using Net Removals and TF are shown in Table 3-3. The combined
average Net Removals and the TF for the storm events and river water tests were 14.3% and
1.65, respectively, which shows concentration occurred. Bar graphs of Removals and Net
Removals for the storm events and river water test are shown in Figure 3-3.
3-15
-------
Section 3 - Technology Assessment
TABLE 3-3 STORM KING™ CSO CONTROL PERFORMANCE
JAMES BRIDGE, ENGLAND
Test
Number
Storm Events
SI
S2
S3
Peak
How
(MGD)
6.8
5.7
3.7
Volume
Reduction
(%)
35.9
93.6
91.9
Hydraulic
Loading
(gpm/ft2)
12.9
1.0
0.9
MassSS
Removal
<*)
43.1
93.4
95.0
MassSS
Net Removal
W
7.2
-0.2
3.1
TF
1.20
1.00
1.03
River Water Tests
Dl
D2
D3
D4
DS
D6
Average
2.7
2.0
2.5
5.0
5.2
4.6
4.3
32
36
19
19
9
14
39
5.4
3.7
6.1
11.9
13.9
11.6
7.5
65.9
. . 64.1
27.1
32.4
13.9
43.8
53.2
33.9
28.1
8.1
13.4
4.9
29.8
14.3
2.06
1.78
1.43
1.71
1.54
3.13
1.65
Note: Design flowrate of 8.6MGD with a hydraulic loading of 26 gpm/ft2.
Source: Hedges et aL, 1992.
3-16
-------
Section 3 - Technology Assessment
~ FIGURE 3-3 :
STORM KING™ SS REMOVALS FOR CSO CONTROL, JAMES BRIDGE, ENGLAND
100
o^
CO
—1
O
UJ
:CC
90
80
70
60
50
40
30
20
10
0 -
-10
S1 S2 S3 D1 D2 D3 D4 D5 D6
EVENT
O REMOVAL M NET REMOVAL
3-17
-------
Section 3 - Technology Assessment
STORMWA1
CONTROL APPLICATIONS
Swirl
The West Roxbury (Boston), MA pilot study was conducted to demonstrate the
effectiveness of swirls as a treatment technology for separate urban stormwater. The unit was
tested between 1979 and 1981, and is no longer in existence. Removals, Net Rempvals, and TF
were determined for each of the events.
The 160 acre drainage area served by the demonstration facility was located in a
moderate income residential neighborhood served by a completely separate storm-sewer system.
In addition, due to the gravity operation of the unit, considerations had to be made for the
vertical elevation available. The project also evaluated the helical-bend regulator/concentrator
(another concentrating unit that uses similar secondary-fluid motion solids separation principles
as that of the swirl). The project layout is shown in Figure 3-4.
Rainfall records from two local rainfall gauging stations were analyzed for a period of
10 years to determine the appropriate Qd for the units. Discrete storm events were identified,
as well as the year, month, and day the storm began; the hour of day that the storm began; the
length of the antecedent dry period; the total amount of rainfall in the storm event; the duration
3-18
-------
Section 3 - Technology Assessment
of the storm event; the maximum hourly rain in the storm; and the hour of the maximum rain
in the storm.
Hydrologic modeling, of the catchment area was then performed using the EPA Storm
. Water Management Model (SWMM). This was done to determine an overall estimate of the
runoff coefficient for the watershed. The runoff coefficient was required so that the swirls
design flowrate frequencies could be estimated.
The Rational Formula (Q, = CIA) was used to determine the design flowrate, Q,,, for
the watershed. The runoff coefficient, C, was determined to be 0.41. A time series of
maximum average hourly intensities for the observed rainfall events at one of the rainfall
gauging stations and a drainage area, A, of 160 acres were used in the calculation. The resultant
design flowrate was 3.9 MGD per unit (swirl and helical-bend regulator), with a maximum
flowrate to each unit of 7.8 MGD.
The design of the swirl was based on the initial guidelines established by the EPA
(Sullivan et al., 1972). The unit was sized to provide 80% removal of settleable solids. The
inlet and outlet pipes were 2 ft in diameter and the unit diameter was 10.5 ft. Flow entering the
foul-sewer outlet was regulated by a Hydro Brake™, which only allowed up to 0.1 MGD to
discharge. This is equivalent to 3% of the Qd. SS from the a foul-sewer outlet were discharged
3-19
-------
Section 3 - Technology Assessment
FIGURE 3-4
SITE PLAN FOR THE WEST ROXBURY, MA FACILITY
Source: Pisano et al., 1984.
3-20
-------
Section 3 - Technology Assessment
to a foul sewer tank 10.5 ft in diameter and 6 ft in depth that could be pumped to either a 120
1 ' ' • ' , '/
in. diameter storm drain or into a 27 in. sanitary sewer.
Flow and quality measurements were performed at the West Roxbury facility.
Monitoring and sampling was done at the influent, the effluent, and the foul-sewer discharge.
The sampling locations are identified in Figure 3-5. Sampling consisted of both manual and
automatic procedures. Samples were analyzed for SS, volatile SS, settleable solids, and volatile
settleable solids.
Although additional monitoring occurred, only seven storm events were selected for
detailed analysis. Removals, Net Removals, and TF were determined for each of the events and
are shown in Table 3-4. The SS Removals varied between 6% arid 36% and averaged 28.1%.
The Net Removal and TF averaged 17.0% and 3.4, respectively which indicates that SS
concentration had taken place. Removal and Net Removals for several storm events are shown
in Figure 3-6. '
3-21
-------
Section 3 - Technology Assessment
FIGURE 3-5
SWIRL SAMPLING LOCATIONS FOR WEST ROXBURY, MA
Source: Pisano et al.t 1984.
3-22
-------
Section 3 - Technology Assessment
TABLE 3-4 SWIRL STORMWATER
CONTROL PERFORMANCE,
, WEST ROXBURY, MA
Storm
Number
1
2
3
4
5
'
6
7
Date
6/29/80
7/29/80
10/3/80
10/25/80
6/9/81
6/22/81
8/4/81
*
Average
Flowrate
(MOD)
1.3
1.9
3.9
1.9
0.5
1.9
1.4
1.3
3.9
7.8
Volume
Reduction
(%)
12.8
7.8
4.0
8.3
31.3
8.0
11.0
1-2.0
4.0
0.0
Mass SS
Removal
(%)
21.0
35.3
36.0
29.7
34.0
27.0
27.0
33.0
9.5.
5.8
Mass SS Net
Removal
(%)
8.2
. 27.5
32.0 .
2.1.4
2.8
19,0
16.0
21.0
5.5
5.7
TF
1.64
4.53
9.00
3.58
1.09
3.38
2.45
2.75
2.38
58.00
Average
2.0
11.0
28.1
17.0
3.4
Note: Design flowrate of 3.9 MOD with a HL of 31 gpm/ft2. SS mass Removals and Net
Removals are determined using loadings. Reductions are estimated.
* Data is extreme and not included in the average.
Source: Pisano etal., 1984.
3-23
-------
Section 3 - Technology Assessment
FIGURE 3-6
SWIRL .STORMWATER CONTROL PERFORMANCE, WEST ROXBURY, MA
40
£
CO
o
UJ
cc
35
30 -
25
20
15 -
10 •
5 •
1
1
1
1
1-
I
3 4 5 6
EVENT
D TOTAL REMOVAL • NET REMOVAL
I
3-24
-------
Section 3 - Technology Assessment
SWIRL DEGRTTTER ;
A full-scale 16.5 ft diameter swirl degfitter was demonstrated in Tamworth, N.S.W.,
Australia as a cheaper alternative for pretreatment of river water as a way to reduce wear and
tear on raw-water pumps. It was designed to operate between 3.4 MOD and 19.4 MOD. The
swirl degritter efficiently removed the portion of SS for which it was designed (defined as
inorganic material > 0.2 mm in diameter with a specific gravity of 2.65), achieving an average
of 95.5% (Shelley et al., 1981).
OPERATION AND MAINTENANCE
Vortex separators do not have any moving parts and, accordingly, are not maintenance
intensive. However, washdowns are required following every CSO event to prevent shoaling
and foul odors from developing. Some of the units are designed to be self-cleansing or have
automated washdown systems. Washdown may not be necessary after every storm event for
separated stormwater treatment applications since the residual solids tend to be less putrescible.
The length of time that automated washdown operates varies with the time at which the
washdown takes place following the overflow event. Typically, the washdown system for the
Washington, DC swirl will run for approximately one-half of an hour if the washdown occurs
immediately following an overflow. Longer washdowns will be required as the time increases
3-25
-------
Section 3 - Technology Assessment
between the end of the overflow and the washdown. For a CSO application, the washdown
water and the residuals caught in the water flow are generally conveyed via the underflow to the
s
sanitary sewer. *
The amount of SS or floatables that reach the vortex separator vary due to several factors
including the antecedent-dry period; sewer flowrate, volume and duration; drainage area and
sewer length, topographic and sewer slopes; and season (e.g., autumn will generate added solids
due to leaf fall). Any pretreatment (i.e., bar screens) that exists will decrease the quantity of
coarse solids reaching the vortex separators. This may include street cleaning, to reduce those
wastes that get flushed by the stormwater runoff.' If bar screens are used, these require regular
maintenance (i.e., cleaning and residuals disposal).
The Huidsep™ in Tengen, Germany and the Storm King™ in Surrey Heath, England
have not reported any malfunctions in their 4 years and 1.5 years of operation, respectively.
Experience with the DC swirls has also indicated that the wash down system would be more
effective with a few design modifications. The operators of the units have recommended that
the floor be sloped toward the underflow. The original vortex unit in Bristol, England has been
in operation for approximately 30 years w/o bar screen and has required very little maintenance.
3-26
-------
Section 3 - Technology Assessment
OTHER POLLUTANT REMOVALS
Most available data on the performance of vortex solids separators focused on SS
Removal. Limited information is available on other pollutants of concern. The Tengen,
Germany CSO Fluidsep™ evaluation determined IF for COD averaged 1.8 with a range of 1.0,
to 8.0. Again, it is indicated that a TF > 1 indicates some Net Removal of COD by vortex
concentration. COD Removal in Tengen was lower than the SS Removal as indicated by the
rFof2.1forSS. ' ,
3-27
-------
Section 4 - Technology Costs
SECTION 4
TECHNOLOGY COSTS
This section presents the costs associated with the design, construction, and operation of
vortex solids separators. Each cost type is discussed in the following subsections. All costs
have been converted to 1994 dollars (ENR 5450).
DESIGN COSTS
As discussed in Section 2, Design Criteria, each of the three different vortex solids
separator units have varying design protocols. The swirl design is available in the public domain
from an EPA design manual (Sullivan et al., 1982). Additional design cost associated with this
unit include settleability analysis and possibly a pilot-scale demonstration. The Storm King™
design is based on pilot-scale units that can be rented at $2,000 a month without including piping
and treatability studies (HEL, 1995). The pilot-scale testing is performed on units 3 to 6 ft in
diameter. Pilot testing is necessary to select the appropriate unit dimensions which best suits
the intended application. A Storm King™ pilot-scale steel unit (3ft diameter) may be purchased
for $12,000 as was done in the case of Columbus, Georgia (HIL, 1995). The Fluidsep™ design
is based on site-specific settleability studies. The Fluidsep™ design study costs typically vary
between $25,000 to $100,000, with actual study cost depending on the site and size of the
4-1
-------
Section 4 - Technology Costs
facility. The study takes from three months to one year to complete. The swirl, Storm King™,
and Fluidsep™ require hydrologic studies for proper Qd selection.
! ^
Cost estimates were determined for the Bradenton, Florida pilot study using two 5 ft
diameter Storm Kings™. The costs were calculated with pumping or for gravity feed. The
costs for pumped and gravity fed units were $54,000 and $37,500 (1990 costs), respectively
(Smith and Gillespie, 1990). Other costs include the pumps, maintenance on the pumps,
electricity for the pumps, excavation, piping .and valves. This amounts to $335 - $482 /acre of
*••• " • . ...
drainage area or $110,000 - $158,000 /mgd design flow (1990 costs).
CAPITAL COSTS
The capital cost for vortex solids separator facilities are very dependant on site-specific
characteristics. Commonly, vortex solids separators are used with other treatment technologies
e.g., disinfection. The adjusted capital costs (ENR 5450) for several vortex CSO control
facilities are shown in Table 4-1 as a function of Qa, diameter, and volume. The general range
of capital costs for vortex facilities in the U.S. varies between $3,500 and $25,000/MGD.
The swirl CSO facility in Washington,- DC had a capital cost of approximately
$16,200,000 of which $9,100,000 were attributable to the swirl (O'Brien an Gere, 1992). The
total facility contains three 57 ft diameter swirls* automated bar screens, and chlorination-
4-2
-------
Section 4 - Technology Costs
dechlorination. The capital costs attributable to the swirl facility are $2,116/acre drainage area
or $22,750/MGD Qd. Earlier swirl demonstration projects in Syracuse, NY (Drehwing et al.,
1979) and Lancaster, PA (Pisano et al., 1984) are also listed.
Capital cost data is available for several vortex separator installations. Fluidsep™ facility
costs were between $3,500 and $14,600/MGD. Capital costs of the Fluidsep™ units are
available for the Decatur, IL Oakland and 7th Ward projects (Pisano, 1994). The Fluidsep™
for the 7th Ward costs approximately $5,200/MGD for the separator alone as shown in Table
4-1. The Oakland unit at $14,600 reflects cost including foul-sewer pumping. The capital costs
of the vortex units alone are approximately 10 to 12% of the total capital costs of the CSO
abatement facilities including the vortex units. Additional costs are a result of the improvements
made to the overall facility (e.g., grading, piping, bypass structures, outfalls, and mechanical
screening). Capital costs for Storm King™ units are also provide in Table 4-1 (Boner, 1994).
The costs include the conveyance systems, tanks, pumps, etc, .
Generally, above-ground units will be less costly than the underground units do to
excavation costs. However^ pumping should also be avoided to reduce capital and operational
costs. In a comparison of annual operation and present worth, respectively, the swirl degritter
cost approximately 10% and 20% less than the conventional aerated grit chamber (Sullivan et
al., 1982). In its first application, swirl degritter construction costs in Australia were 30% less
than that of an equivalent aerated grit chamber (Shelley et al., 1981).
4-3
-------
Section 4 - Technology Costs
TABLE 4-1. COMPARATIVE UNIT COST
LOCATION
ENR/Year
COST
per MGD per $-
per ft3
SWIRL
Syracuse, NY*
Lancaster, PA
Washington, DCf
FLUIDSEP™
Saginaw, MI
Burlington, VT
Decatur, IL - 7th Ward
Decatur, IL — Oakland*
STORM KING™
Hartford, CT*
Columbus, GA*
5450
5450
5450
$10,700
$5,900
$22,900
$5,900
$6,400
$53,500
1994
1994
$27,000
$20,000
$27,000
$15,000
* 0 's= diameter.
Includes foul-sewer pumping.
* Reflects total cost attributable to swirl.
' * Estimated cost, includes internal components and piping.
$170
5450
5450
5450
5450
$4,600 -
$3,500
$5,200
$14,600
$8,200
$7,000
$13,000
$10,800
$20
$14
$20
$47
$29
4-4'
-------
Section 4 - Technology Costs
OPERATION AND MAINTENANCE (O&M) COSTS
Vortex solids separator units require minimal energy expenditures unless pumping or
automated washdown systems are required. Fluidsep™ and Storm King17"1 units which are
deeper than the swirl, are much more likely to require pumping. The units are designed to
operate without moving parts which are less likely to require replacement. Operating expenses
primarily include labor for washdown or minimal, intermittent energy to support an automated
washdown system. -
HEL's Grit King™ in Surrey Heath England, which treats separate stormwater runoff
t
from roof and highway runoff, has a grit collection zone at the conical base of the vortex unit
which requires periodic emptying once the level of the grit accumulates. Periodic maintenance
to clean out the grit collection zone is estimated to cost between $300 and. $450 per cleaning.
4-5
-------
Section 5 - Summary of Findings
SECTIONS
SUMMARY OF FINDINGS
PERFORMANCE SUMMARY
Selected design information, Removals, Net Removals, and TF for the vortex solids
separator units evaluated in this report are summarized in Table 5-1. The storm events were
screened to eliminate data from storm events with appreciably less flow than the design flowrate.
Net removal determinations from storm events that are significantly below design flowrate are:
misleading because the underflow component is disproportionately large relative to the overflow.
The performance data indicates that vortex separators can separate SS from the influent
and concentrate them in the underflow. Average Removals varied from a low of 37.6% at the
DC swirl to 54% at the Tengen, Germany Fluidsep™. Average Net Removals varied from 6.9%
at the Tengen, Germany Fluidsep™ and to 17% at the West Roxbury, MA pilot-scale study.
The data on the performance of swirl and vortex separators was too specific to site and
storm event and too limited in scope to assess the performance of one vortex separator design
relative to the others, More evaluation is needed :with a set of full-scale units to make a
meaningful comparison. The stormwater application of swirl and vortex units exhibited
Removals and Net Removals within the same range as for units treating CSO.
5-1
-------
Section 5 - Summary of Findings
TABLE 5-1 - VORTEX SOLIDS SEPARATOR PERFORMANCE SUMMARY
Unit Type &
Location
Swirl
Washington, DC
Design Design
Hydraulic Flow
Loading (MGD)
(gpm/ft?)
36
West Roxbury, MA" 31
Fluidsep™
Tengen,Germany
95
,TU
Storm King'
James Bridge
Walsall, UK
26
134
3.9
10.8
8.6
Storm Flow Mass SS
Event* (MGD) Removal
Net SS Treatment
Removal Factor
3/29/91
9/24/91
12/2/91
10/3/80
8/4/81
62
72
63
3.9
3.9
28.3
35.6
21.1
36.0
9.5
13.1
23.9
8.6
32.0
5.5
1.87
3.05
1.68
9.00
2.38
No. 2
No. 3
3.2
2.3
32
40
-7.1
15.8
11/29/88 6.8
43.1
7.2
* Storm events selected because respective HL is closest to design HL out of sample storms.
~ Evaluation based on separate stormwater flow.
All available data from sample storms is provided in Section 3.
0.82
1.65
1.20
5-2
-------
Section 5 - Summary of Findings
Because vortex solids separation removes particles by concentrating SS by inertia!
separation, the settling velocity distribution of the SS affect the performance of the unit. SS in
storm flow, whether stormwater or CSO, that exhibit settling velocities > 0.14 cm/sec, are
more amenable to removal by vortex separation than SS with lower settling velocities.
However, the design of vortex solids separator units are targeted for particle with high settling
velocities (e.g., 2.65 cm/sec or greater for the swirl, as previously discussed). Vortex unit
design should be based on site-specific settleability and pilot-scale testing.
The available data indicate that vortex solid separators had higher Removals under
conditions with lower HL. Lower HL generally decrease Net Removals as a result of the large
capture of underflow during storm events well below the design flowrates. A balance must be
struck in choosing Q, between determining the greater benefits of Removal or Net Removal for
each particular installation.
FINDINGS
Applications
The selection of the best treatment or control technology for any wet-weather flow
application must be based upon knowledge of these three components: pollutants of concern,
pollutant characteristics (particle settling velocity or size distributions), and performance
requirements (permitted discharge levels or removal efficiencies). General information on the
•-'•'•. 5-3 . . -
-------
Section 5 - Summary of Findings
potential of pollutant classes to be removed by settleability-based treatment such as vortex
separators are presented in Table 5-2, Vortex separators are an attractive process where high-
rate SS separation of gritty materials, heavy particles, or floatables is required. A significant
portion of the pollutants of concern have settling velocities < 0.14 cm/sec, are dissolved or
colloidal, or close to the density of water, vortex separators are not an appropriate technology.
The majority of vortex separator applications have been for CSO. There has only been
limited testing for separated stormwater. Vortex separators provide similar SS Removals for
separated stormwater and CSO. Vortex units have been used for CSO because they offer several
advantages apart from the level of SS Removal provided. For example, swirls provide a
secondary purpose of flow regulation. The vortex separator can capture smaller storms,
providing storage for the polluted flow segment or entire flow volume, and then allow discharge
of the stored volume to the underflow, thus eliminating the major portion of smaller CSO events.
This latter concept has been successfully applied at the Washington, DC swirl facility where the
numbers of CSO events have been reduced. By operating the three swirl units in parallel rather
than providing limited treatment through one unit, smaller storms are completely captured. This
. f
advantage is not available from storm-flow controls that do not have an underflow to a treatment
facility, or those which have limited underflow holding facilities.
5-4
-------
Section 5 - Summary of Findings
TABLE 5-2
POTENTIAL POLLUTANTS AMENABLE
TO TREATMENT IN VORTEX SOLIDS SEPARATORS
Wet-Weather
Water Quality
Concerns
Potential Performance
by Vortex Solid Separator*
Suspended solids and
Sediment
Floatables
Good removal of heavier particles. Fair to good •
removal can be expected for total suspended solids.
Good floatables capture up to design flowrates can be
expected.
Oil and grease
Oxygen demanding organics
Nutrients
• ' N
Metals
Toxic organics
Fair removal
Fair to good removal of heavy organic material.
Poor or no removal for light particles and
dissolved organics.
Fair removal of phosphorus associated with particles
is possible. Poor removal of dissolved forms of
nitrogen.
Fair to good removal of metals which sorb to
particles or are in solid form. Poor removal of metals like
zinc and nickel.
Poor removal of dissolved organics. Toxic
organics that bind to particulates may exhibit fair to
good removals.
Source: Randall, et a/., 1983 and Whipple and Hunter, 1981.
* This table is based upon the pollutants general response to settlebality based treatment. Additional research is
necessary to document actual vortex solids separator performance. Approximate range of pollutant removal is
the following: excellent > 90%; Good 60-80%; Fair 30-60%; poor < 30%.
5-5
-------
Section 5 - Summary of Findings
Vortex separators for stormwater control and treatment are most appropriate when
used in conjunction with other controls. For example, a vortex separator can be used prior
to a detention pond or wetland to lessen deposits and floatables and the associated operation
and maintenance requirements.
Vortex separator units, when designed as satellite systems, are relatively compact
units. They are often suitable for underground installation. Satellite units can be used over a
broad expanse of the collection system, minimizing the high cost of conveyance systems
needed for centralized treatment facilities. "••'.-
The swirl degritter or HDL's Grit King™ can be used when coarse or preliminary
treatment is the only objective and flow regulation or flow splitting is not required. Here
detritus (grit and heavy organics) is collected in a bottom hopper for later removal. To
decrease downstream wear, the swirl degritter can also be placed in series with a swirl to
remove detritus from the foul-sewer underflow (especially in cases where the underflow has
to be pumped back to the WWTP). This was done for CSO in the Lancaster, PA project
(Pisanoera/., 1984).
. • 'l
Limitations
Characteristics of the pollutants of concern and the nature of SS carried in the storm
flow should be identified in preliminary investigations to determine if a vortex separator is
5-6
-------
Section 5 - Summary of Findings
the appropriate control for each site. Vortex treatment effectiveness will be poor for storm
flow having a relatively small percentage of particles with high settling velocities. Many
water quality pollutants adhere to fine particles or are dissolved. The vortex unit will only
provide minimal abatement of these pollutants.
Treatment objectives must match the capability of the vortex solids separator. If a
specific application necessitates high-level SS removal (say 80%), vortex separators would
not likely be an applicable control. However, if the vortex separator does not achieve the
desired treatment level as a stand-alone unit, it can still be appropriate to use as part of the
overall storage/treatment system.
1 '• • ' ' ••
Vortex units should be placed underground so that they are gravity fed. Inflow
pumping is costly and breaks up particles rendering them less settleable and treatable.
Therefore, site conditions may restrict the use of vortex units. Sites must have the
appropriate depth and stability to structurally support the unit. Sites that require blasting will
significantly increase construction costs. If the flows from the underflow must be pumped
from.the vortex unit, the capital and O&M cost associated with pumping increase the cost of
using this technology.
Vortex separators are not maintenance intensive; however, washdown should be
performed following every significant storm-flow event. Experience with vortex separators
used for CSO treatment in Washington, DC indicates that if maintenance is not performed,
5-7
-------
Section 5 - Summary of Findings
odors occur. Manpower or automated washdown should be available, particularly for CSO
applications.
RECOMMENDATIONS
• Additional studies on the effect of design variables (e.g., settling velocity,
hydraulic loading, and underflow to influent ratio) on unit performance would
ensure better design.
• More data is necessary to assess the ability of vortex separators to treat
pollutants other than SS.
• Results from side-by-side studies of all three commercially available units will
help to determine the effects of the design differences and comparative unit
performance. A demonstration of this type is under development for CSO by
the New York City Department of Environmental Protection.
5-8
-------
Section 6 - References
SECTION 6
REFERENCES
Boner, M., 1994. Personal Communication.
Brombach, H. etal., 1993. Experience with Vortex Separators for Combined Sewer Overflow
Control. Water Science Technology, Volume 27, No. 5-6, pp 93-104. ,
Dalrymple, R.J. etal., 1975. Physical and Settling Characteristics of Particulates in Storm and
Sanitary Wastewaters. EPA-670/2-75-011. NTIS PB24200L
Drehwing, FJ. et al., 1979. Disinfection/Treatment of Combined Sewer Overflows. EPA-
\
600/2-79-134. NTIS PB80-113459.
Field, R. and Masters, H. 1977. Swirl Device for Regulating and Treating CSOs. EPA
Technology Transfer Capsule Report. EPA-625/2-77-012. NTIS PB299571. N
Hedges, P.D. et al., 1992. A Field Study of an Hydrodynamic Separator CSO. Novatech 92,
Lyon, France.
H.I.L. Technology, 1991. Information brochures.
H.I.L. Technology, 1993. Information brochures and memos.
H.I.L. Technology, 1994. Personnel Communication.
NKK Corporation, 1987. Solid-Liquid Separation by Swirl Concentration. Brochure.
O'Brien and Gere, 1992. CSO Abatement Program Segment 1: Performance Evaluation.
Prepared for the Water and Sewer Utility Administration, Washington, D.C..
6-1
-------
Section 6 - References
Pisano, W.C. et al, 1984. Swirl and Helical Bend Regulator/Concentrator for Storm and CSO
Control. EPA-600/2-84-151. NTIS No. PB85-102523.
Pisano, W.C., 1993. Summary: The Fluidsep™ Vortex Solids Separator Technology. WKInc.
Marketing Brief, Belmont, Massachusetts.
Pisano, W.C., 1994. Personal communication.
Randall, C. W. et al., 1983. "Urban Runoff Pollutant Removal by Sedimentation."
Proceedings of the Conference on Storm Water Detention Facilities. American Society
of Civil Engineers. New York, New York.
Shelley, G. J. et al., 1981. Field Evaluation of a Swirl Degritter at Tamworth, New South
Wales, Australia. EPA-600/2-81-063. NTIS No. PB81-187247.
Smith and Gillespie Engineers, Inc., 1990. Engineer's Study for Storm water Management
Demonstration Project No. 2 for Evaluation of Methodologies for Collection, Retention,
Treatment and Reuse of Existing Urban Storm water. S&G Project No. 7109-133-01.
Sullivan R.H. et al., 1972. The Swirl Concentrator as a Combined Sewer Overflow Regulator
Facility. EPA-R2-72-008. NTIS No. PB214687.
Sullivan, R.H., et al. 1974. The Swirl Concentrator as a Grit Separator .
Device. EPA Report No. EPA-670/2-74-026.
Sullivan, R.H., etal. 1977. Field Prototype Demonstration of the Swirl Degritter. EPA-600/2-
77-185. NTIS No. PB-272654.
Sullivan, R.H. et al. 1982. Swirl and Helical Bend Pollution Control Devices. EPA-600/8-82-
013. NTIS No. PB82-266172.
t
Whipple, W. and J. V. Hunter, 1981. "Settleability of Urban Runoff Pollution." Journal of the
Water Pollution Control Federation. Vol. 53. No. 12.
6-2
-------
Section 7 - References
SECTION 7
BIBLIOGRAPHY
Drehwing, FJ. ei al., 1979. Combined Sewer Overflow Abatement Program - Volume II: Pilot
Plant Evaluations. EPA-600/2-79-031b NTIS PB 80-159 262.
Field, R., 1973. The Dual Functioning Swirl Combined Sewer Overflow Regulator/
Concentrator. Report EPA-670/2-73-059. NTIS PB 227 182.
' - !
Field, R., 1974. Design of a Combined Sewer Overflow Regulator/Concentrator, Journal Water
Pollution Control Federation. Vol, 46, No. 7.
Field, R., 1986. State-of-the-Art Update on Combined Sewer Overflow Control, Critical
Reviews in Environmental Engineering, Vol. 16, No. 2.
Sullivan, R.H. et al., 1974. Relationship between Diameter and Height for Design of a Swirl
Concentrator as a Combined Sewer Over-flow Regulator, American Public Works
Association, Chicago, IL. EPA-670/2-74-039. NTIS PB 234 646.
Sullivan, R.H. et al., 1975. The Helical Bend Combined Sewer Overflow Regulator, American
Public Works Association, Chicago, IL. EPA-600/2-75-062. NTIS PB 250 619.
Sullivan, R.H. et al., 1976. The Swirl Concentrator for Erosion Runoff Treatment American
Public Works Association, Chicago, IL. EPA-600/2-76-271. NTIS PB 266 598.
Sullivan, R.H. et al., 1978. The Swirl Primary Separator: Development and Pilot
Demonstration, American Public Works Association, Chicago, IL. EPA-600/2-78-122.
NTIS PB 286 339/7.
, 7-1
-------
Section 7 - References
Walker, D. et al., 1993. Manual: Combined Sewer Overflow Control, U..S. Environmental
Protection Agency, Office of Research and Development. EPA/625/R-93/007. NTTS PB
93-144649.
7-2
------- |